Integrating more efficient renewable energy

technologies into food systems in Central

Mozambique: implications to food and
nutrition security

D ISSERTATIO N
zur Erlangung des akademischen Grades
Doctor of Philosophy (Ph.D.)

eingereicht an der Lebenswissenschaftlichen Fakultät

der Humboldt-Universität zu Berlin

von

Custódio Efraim Matavel, M.Sc.

Präsidentin der Humboldt-Universität Dekan der Lebenswissenschaftlichen

zu Berlin Fakultät der Humboldt-Universität zu
Berlin
Prof. Dr. Julia von Blumenthal Prof. Dr. Dr. Christian Ulrichs

Gutachter/innen Tag der mündlichen Prüfung

1. PD Dr. Stefan Sieber 25. Mai 2023
2. Prof. Dr. Dr. Christian Ulrichs
3. Prof. Dr. Sandro Luis Schlindwein
Integrating more efficient renewable energy technologies into
food systems in Central Mozambique: implications to food and
nutrition security
PhD thesis submitted by Custódio Efraim Matavel, M.Sc.
 3

Table of contents
I |Acknowledgements ............................................................................................................................... 4
II |Abbreviations .......................................................................................................................................... 5
III |List of Tables ....................................................................................................................................... 6
IV |List of Figures...................................................................................................................................... 7
V |Summary ................................................................................................................................................... 8
VI |Zusammenfassung ......................................................................................................................... 11
VII |List of featured publications ...................................................................................................... 14
VIII |Additional Peer-reviewed Articles with Relevance to the PhD Topic........................ 15
1 | Introduction ......................................................................................................................................... 16
1.1 State of food and nutrition security in Africa ................................................................... 16
1.2 The concept of food and nutrition security (FNS).......................................................... 18
1.2.1 Food and nutrition security measurement ................................................................... 20
1.2.2 Linkages between energy and FNS .................................................................................. 23
1.3 Food and nutrition security situation in Mozambique ................................................. 25
1.4 Problem definition, research objectives and hypotheses............................................ 27
2 |Synopsis of this work ........................................................................................................................ 29
3 |Research methodology ..................................................................................................................... 31
3.1 Study area ...................................................................................................................................... 31
3.2 Research design ........................................................................................................................... 32
3.3 Statistical analysis....................................................................................................................... 34
4 |Results ..................................................................................................................................................... 36
4.1 Understanding the drivers of food security among agriculture-based
households in Gurué District, Central Mozambique ................................................................... 36
4.2 Experimental evaluation of a passive indirect solar dryer for agricultural
products in Central Mozambique....................................................................................................... 68
4.3 Toward energy saving and food safety in Central Mozambique: the role of
improved cook stoves and heat retention boxes ......................................................................... 91
4.4 Effect of passive solar drying on food security in rural Mozambique ..................116
4.5 Stronger adoption at lower costs: individual capacity building as a pivotal tool
for improved cookstoves adoption in Central Mozambique .................................................144
5 |Synthesis ..............................................................................................................................................169
6 |Study limitations and further research needs .......................................................................173
7 |Overall Conclusion............................................................................................................................176
8 |Policy Recommendations ..............................................................................................................178
9 |References ...........................................................................................................................................180
Statutory Declaration ................................................................................................................................190
 4

I |Acknowledgements
Completing this thesis would not have been possible without the financial, scientific and
emotional support I received from several individuals and institutions. My journey as a
doctoral student has indeed had its ups and downs. I am deeply grateful to my parents,
Graciete and Efraim, my partner Dina, and my son Kevin who, even though I was far away,
were always present in my life, supporting me unconditionally.

I would like to thank Prof. Dr. rer. pol. Klaus Müller, PD Dr. Stefan Sieber and Prof. Harald
Kächele for always guiding and encouraging me to improve and become a better scholar
throughout the past years.

A special thanks to Dr. João Salavessa, Dr. Constance Rybak and Dr. Harry Hoffmann for
believing in me. You are the reason I became part of one of the best research teams I've
ever met. Furthermore, my gratitude goes explicitly to the Federal Ministry of Food and
Agriculture (BMEL) and Federal Office for Agriculture and Food (BLE) for funding my PhD
project. I must thank Karin Stahl, Annett Jahn and Anika Kulling, who always supported
me when administrative issues came up along the way.

Dr. Götz Uckert, Dr. Jonathan Steinke and Dr. Johannes Hafner, thank you for the early and
tireless support that helped me to find my way out of countless dead ends I have been to.

I also want to thank my co-authors, who helped me with their constructive contributions
during the articles' conceptualization, writing, and editing. My gratitude extends to the
anonymous reviewers who took the time to help me to improve my work. Many thanks to
Dr. Carsten Paul for accepting to be my mentor and for the guidance.

Many thanks to all my fellow PhD colleagues in the SusLand group for the friendship and
the time we spent together, sharing good and challenging times.

I would also like to thank Lúrio University, the local administrative offices of the case
study sites in Gurué district for the kind cooperation that made this research possible.
Thanks to Diofino, Baltazar and Ernesto for supporting the data collection as
enumerators. A special thanks also goes to the people in the villages who kindly
cooperated and provided the data that made this research possible.
 5

|
II Abbreviations

CCT: Controlled Cooking Test

FAO: Food and Agriculture Organization of the United Nations

FAS: Food Availability Scores

FI: Food Insecurity

FNI: Food and Nutrition Insecurity

FNS: Food and Nutrition Security

FS: Food Security

HFIAS: Household Food Insecurity Access Scale

HRB: Heat Retention Box

HRS: Heat retention System

ICS: Improved Cookstove

IFPRI: International Food Policy Research Institute,

MAHFP: Months of Adequate Household Food Provisioning

OSD: Open-Sun Drying

PHLs: Post-Harvest Losses

PISD: Passive Indirect Solar Drying

PSD: Passive Solar Drying

SDG: Sustainable Development Goals

SSA: Sub-Saharan Africa

TSF: Three-Stone Fire Stove

UNHCR: United Nations High Commissioner for Refugees

UNICEF: United Nations International Children's Emergency Fund

WB: The World Bank

WDDS: Women Dietary Diversity Score

WFP: World Food Programme

WHO: World Health Organization

 6

|
III List of Tables

Table 1.1: Food and nutrition security metrics. Adapted from Pangaribowo et al. (2013),
Jones et al. (2013) and Herforth and Ballard (2016) ..................................................................... 22
Table 4.1: Food Groups and Item used for HFCS calculations .................................................... 43
Table 4.2: Explanatory variables used in the regression analysis ............................................. 45
Table 4.3: Socioeconomic characteristics of the households in each village ........................ 48
Table 4.4: Percentage of households in each food security category ...................................... 50
Table 4.5: Estimation results of GOLM ................................................................................................. 51
Table 4.6: Marginal effects of explanatory variables for FS level .............................................. 52
Table 4.7: Fresh leaves loaded during successive drying days .................................................. 74
Table 4.8: Design features of the ICS ..................................................................................................... 97
Table 4.9: Food used in the CCT .............................................................................................................. 97
Table 4.10: Stove performance indicators .......................................................................................... 98
Table 4.11: Average values of fuel and time consumption for different types of food and
stoves (n=12). ..............................................................................................................................................101
Table 4.12: Average time (min) the food temperature took to reach 60 ̊C ..........................101
Table 4.13: Household food inventory. ..............................................................................................122
Table 4.14: Descriptive statistics of surveyed households (Continuous/count variables)
...........................................................................................................................................................................124
Table 4.15: Descriptive statistics of surveyed households (dummy variables) ................124
Table 4.16: Prevalence of food security .............................................................................................127
Table 4.17: Determinants of predicted likelihood of using solar dryer (probit model
output) ............................................................................................................................................................128
Table 4.18: Effect of using solar dryer on food security based on PSM ................................130
Table 4.19: Effect of using solar dryer on food security based on ESR .................................130
Table 4.20: Individual characteristics of the households (continuous variables) ............152
Table 4.21: Random effect probit estimates of ICS adoption ....................................................157
Table 4.22: Marginal effect of the training approach on the probability of ICS adoption
...........................................................................................................................................................................157
Table 4.23: Cost per adopter for ICS dissemination approaches .............................................158
 7

|
IV List of Figures

Figure 1.1: Interlinkages between food system, socio-economic systems and Food and
Nutrition Security. Adapted from Mbow et al. (2019) ................................................................... 20
Figure 2.1: Research concept: proposed mechanisms to improve food and nutrition
security ............................................................................................................................................................. 30
Figure 3.1: Study area location ................................................................................................................ 31
Figure 4.1: Location map for the study area ...................................................................................... 40
Figure 4.2: Percentage (%) of households for: gender; non-farm income; and good land
quality ............................................................................................................................................................... 48
Figure 4.3: Percentage (%) of households producing specific crop in each village ........... 49
Figure 4.4: Schematic view of passive indirect solar dryer ........................................................ 73
Figure 4.5: Variation of the temperature and relative humidity at the ambient and inside
the drying chamber and solar radiation intensity for amaranth leaves ................................. 76
Figure 4.6: Average solar energy input per unit water removal and thermal efficiency for
amaranth leaves ............................................................................................................................................... 77
Figure 4.7: variation of the temperature and relative humidity at the ambient and inside
the drying chamber and solar radiation intensity for maize ....................................................... 77
Figure 4.8: Average solar energy input per unit water removal and thermal efficiency for
maize ................................................................................................................................................................. 78
Figure 4.9: Respondents preference for amaranth leaves sensory attributes ..................... 79
Figure 4.10: Respondents preference for amaranth leaves sensory attributes ................... 80
Figure 4.11: Study area .............................................................................................................................. 96
Figure 4.12: A - improved biomass cook stoves; B - three-stone stoves................................. 96
Figure 4.13: Means of heat retention: A - leftover charcoal; B - heat retention box, C -
banana leaves ................................................................................................................................................. 99
Figure 4.14: Respondents preference for bean curry ..................................................................102
Figure 4.15: Respondent’s preference for maize flour porridge..............................................103
Figure 4.16: Conceptual framework for causality between solar dryer use and food
security ...........................................................................................................................................................121
Figure 4.17: Distribution of estimated propensity scores and region of common support
...........................................................................................................................................................................129
Figure 4.18: Study area location. (a): Africa, (b): Mozambique and (c) Study area. .........149
Figure 4.19: Individual characteristics of the households (categorical variables) ...........153
Figure 4.20: Adoption rates of ICS .......................................................................................................156
 8

|
V Summary

Globally, the number of people experiencing food insecurity (FI) and undernourishment
is rising due to the unprecedented effects of climate change, the COVID-19 pandemic,
population growth, the unaffordability of healthy diets, as well as poverty and inequality.
This situation is even worse in Africa, where the number of people facing hunger is more
than double the proportion of any other region in the world. Facing this challenge requires
empowering smallholder farmers – who supply up to 80% of the food – through
facilitating access to improved or more efficient technologies to increase productivity and
ensure adequate energy access. The potential effects of access to energy-efficient
technologies on food and nutrition security (FNS) are manifold. Energy technologies are
required throughout food systems to produce, process and preserve food. They are also
necessary for cooking meals; however, most smallholder farmers in rural areas still rely
on traditional uses of energy, such as open-sun drying and the three-stone fire stoves. As
a result, a substantial amount of harvested food is not adequately utilised and does not
meet the appropriate standards in terms of quality, volume and safety, which hinders
smallholder farmers’ ability to achieve FNS.

This dissertation explores the effects of optimally designed processing energy usage on
food quality and the respective impacts on FNS in rural Mozambique. It adds to the current
literature, as it provides 1) a holistic understanding of the nature of FNS and its underlying
drives. Furthermore, 2) it adds to the understanding of the technical performance of
optimally designed food processing and preparation technologies. 3) The dissertation
provides insights concerning users’ acceptability of the food processed and prepared
through these technologies. 4) The effects of the newly introduced food processing
technology on FNS are presented. Last – but not least – 5) it discusses the importance of
the dissemination approach for the success of clean cooking adoption. This PhD
dissertation is comprised of 8 chapters.

The first chapter presents an overview of the food and nutrition situation in Africa and
Mozambique. The concept of food and nutrition security and its linkages with energy are
also presented in this chapter. Furthermore, the problem definition, research objectives
and hypothesis are presented. The second chapter provides a synopsis of the dissertation,
 9

and chapter 3 provides an overview of the research methodology. In chapter 4, the

research results are presented in five subchapters. In the first subchapter, household food
security (FS) is assessed using three standard indicators: Household Dietary Diversity
Score (HDDS), Household Food Consumption Score (HFCS), and Months of Adequate
Household Food Provisioning (MAHFP). The results provided evidence of the seasonal
nature of FS in rural Mozambique.

Moreover, the subchapter shows the importance of staple crops such as maize, cassava
and rice for the household FS in the study region. Given that these staple crops require
drying after harvesting, the second subchapter presents a passive indirect solar dryer
(PISD) that was designed and tested as an alternative to open sun drying (OSD). In
addition, a sensorial analysis was performed to evaluate the respective quality attributes
of dried amaranth and maize grains. Compared to OSD, the PISD reduces drying time,
increases thermal efficiency, and produces products with higher sensory quality.
Subchapter 3 presents the performance data of an improved cookstove (ICS) combined
with a heat retention box (HRB). The results show that ICS increases time and energy
efficiency, and the use of HRB has the potential to maintain adequate food storage
temperatures (above 60°C) compared to traditional technologies.

The fourth subchapter focusses on the determinants of farmers’ choice to apply solar
drying technologies and evaluates the effect of a passive solar dryer on food security. The
results provide evidence that solar drying – with associated training – significantly
increases the food security status of participants by increasing households’ food
availability, women’s dietary diversity, and months of adequate household food provision
while simultaneously decreasing the household food insecurity access scale.

The fifth subchapter analyses strategies to effectively disseminate improved and efficient
cooking technologies. It investigates the effect of a) individual video training, b) on-site
group training and c) the combination of both on the adoption of ICS. The results indicate
higher rates of adoption and lower monetary costs for individual training. Moreover,
individual training stimulated early and sustained adoption compared to group-based on-
site training. These findings provide strong evidence that the effectiveness of a training
approach might be context-dependent, and extension agencies should be aware of cost-
effectiveness and adjust extension approaches to achieve the highest impact per unit of
resource expenditure.
 10

The results presented in the individual subchapters are subsequently synthesized

(chapter 5), and study limitations (chapter 6) are presented. The overall conclusion and
policy recommendations are presented in chapters 7 and 8, respectively. The general
conclusions of this thesis are that 1) the improved technologies presented in this study
are sustainable and cost-effective means to substitute, at least temporarily, the prevailing
traditional methods of food processing and food preparation; 2) energy provision through
the use of passive solar drying is an essential component in the fight against food and
nutrition insecurity and 3) governments and relevant stakeholders involved in energy
and food security programs are advised to consider the local context to identify the most
adequate training or information delivery technique.
 11

|
VI Zusammenfassung

Weltweit steigt die Zahl der Menschen, die von Ernährungsunsicherheit und
Unterernährung betroffen sind aufgrund der beispiellosen und gekoppelten
Auswirkungen des Klimawandels, der COVID-19-Pandemie, des
Bevölkerungswachstums, der Nicht-Finanzierbarkeit gesunder Ernährung sowie generell
von Armut und Ungleichheit. Zweifellos am gravierendsten ist die Situation auf dem
afrikanischen Kontinent, wo die Zahl der Hungernden mehr als doppelt so hoch ist wie in
jeder anderen Weltregion. Um diese Herausforderung zu bewältigen, müssen die
Kleinbauern - die bis zu 80 % der Nahrungsmittel liefern - gestärkt werden, auch indem
ihnen der Zugang zu verbesserten oder effizienteren Technologien erleichtert wird, mit
dem Ziel die Produktivität zu steigern aber auch um einen angemessenen Zugang zu
Energie zu gewährleisten. Die potenziellen Auswirkungen des Zugangs zu
energieeffizienten Technologien auf die Ernährungssicherheit sind vielfältig.
Technologien im Energiesektor werden in allen Ernährungssystemen benötigt, um
Lebensmittel zu produzieren, zu verarbeiten und zu konservieren. Sie sind aber auch für
die Zubereitung von Mahlzeiten notwendig, wobei die meisten Kleinbauern in ländlichen
Gebieten noch immer traditionelle Energieformen nutzen wie beispielweise das Trocken
von Ernteprodukten in der offenen Sonne oder die Zubereitung von Nahrung mit Hilfe
von sogenannten „Drei-Stein-Öfen“. Infolgedessen wird ein großer Teil der geernteten
Lebensmittel nicht optimal verwertet und erfüllt nicht die gewünschten Standards in
Bezug auf Qualität, Menge und Sicherheit, was wiederum die Ernährungssicherung von
Kleinbauern negativ beeinträchtigt.

In dieser Dissertation werden die Resultate eines optimal gestalteten Energieeinsatzes

bei der Lebensmittelverarbeitung auf die Lebensmittelqualität und die entsprechenden
Auswirkungen auf die Ernährungssicherheit im ländlichen Mosambik untersucht. Sie
ergänzt die aktuelle wissenschaftliche Literatur, da sie 1) ein ganzheitliches Verständnis
von Ernährungssicherung und der zugrundeliegenden Treiber vermittelt. Darüber hinaus
trägt sie 2) zum Verständnis der technischen Leistungsfähigkeit optimal konzipierter
Lebensmittelverarbeitungs- und -Lebensmittelzubereitungstechnologien bei. Diese
Dissertation liefert zudem 3) Erkenntnisse über die Verbraucherakzeptanz, der mit
 12

diesen Technologien verarbeiteten und zubereiteten Lebensmittel. Weiterhin stellt sie 4)

die Auswirkungen von neu eingeführten Lebensmittelverarbeitungstechnologie auf die
Ernährungssicherheit dar. Schließlich wird 5) die Bedeutung des „Verbreitungsansatzes“
für den Erfolg der Einführung von „Clean Cooking“ Technologien diskutiert. Um all diesen
Aspekten gerecht zu werden besteht diese Dissertation aus acht Kapiteln.

Das erste Kapitel liefert einleitend einen Überblick über die Lebensmittel- und
Ernährungssituation in Afrika und speziell in Mosambik. Das Konzept der Lebensmittel-
und Ernährungssicherheit wie auch seine Verknüpfung zur Energiefrage werden
ebenfalls in diesem Kapitel vorgestellt. Außerdem zeigt es die Problemdefinition, die
Forschungsziele und die Hypothesen dieser Dissertation auf. Das zweite Kapitel gibt einen
vertieften Überblick über die Dissertation und Kapitel 3 erläutert die angewendeten
Forschungsmethoden. In Kapitel 4 werden die Forschungsergebnisse in fünf einzelnen
Unterkapiteln vorgestellt. Im ersten Unterkapitel wird die Ernährungssicherheit der
untersuchten Haushalte anhand von drei Standardindikatoren bewertet: „Household
Dietary Diversity Score“, „Household Food Consumption Score“ und „Months of Adequate
Household Food Provisioning“. Die Ergebnisse belegen den saisonalen Charakter der
Ernährungssicherheit im ländlichen Mosambik. Darüber hinaus unterstreicht dieses
Unterkapitel die Bedeutung von Grundnahrungsmitteln wie Mais, Maniok und Reis für die
Lebensmittelversorgung in der Untersuchungsregion. Da diese Grundnahrungsmittel
nach der Ernte getrocknet werden müssen, wurde im zweiten Unterkapitel ein passiver,
indirekter Solartrockner für landwirtschaftliche Produkte als Alternative zur offenen
Sonnentrocknung entwickelt und im Feld getestet. Darüber hinaus wurde eine
sensorische Analyse durchgeführt, um die entsprechenden Qualitätsmerkmale von
getrockneten Amaranth- und Maiskörnern zu bewerten. Im Vergleich zur offenen
Sonnentrocknung verkürzt der getestete Solartrockner die Trocknungszeit, erhöht die
thermische Effizienz und generierte Produkte mit höherer sensorischer Qualität. In
Unterkapitel 3 wird die Leistung eines verbesserten Kochofens in Kombination mit einer
Wärmespeicherbox vorgestellt. Die Ergebnisse zeigen, dass der verbesserte Kochofen die
Zeit- und Energieeffizienz erhöht und die Verwendung einer Wärmespeicherbox das
Potenzial hat, im Vergleich zu traditionellen Technologien angemessene
Lagertemperaturen für Lebensmittel über längere Zeit aufrechtzuerhalten. Das vierte
Unterkapitel konzentriert sich auf die Entscheidungsdeterminanten der Landwirte für die
Nutzung von Solartrocknern und analysiert die Auswirkungen eines passiven
 13

Solartrockners auf die Ernährungssicherheit. Die Ergebnisse zeigen, dass die Nutzung der
Solartrocknung inklusive einer dazugehörigen Schulung den
Ernährungssicherheitsstatus der Teilnehmer signifikant verbessert. Die Gründe hier sind,
dass die Verfügbarkeit von Nahrungsmitteln im Haushalt generell, die Quantität und
Qualität der Ernährung vor allem bei Frauen wie auch die Anzahl der Monate, in denen
sich ein Haushalt mit Nahrungsmitteln versorgen kann erhöht wird. Das fünfte
Unterkapitel befasst sich mit Strategien zur effektiven Verbreitung von verbesserten und
effizienten Kochtechnologien. Es untersucht die Auswirkungen von individuellem
Videotraining, Gruppentrainings vor Ort und einer Kombination aus beiden Methoden auf
die Akzeptanz von effizienten Kochern. Die Ergebnisse zeigen, dass individuelle
Schulungen zu höheren Akzeptanzraten und geringeren finanziellen Kosten führen.
Darüber hinaus förderte die Einzelschulung im Vergleich zur Gruppenschulung vor Ort
eine frühzeitige und dauerhafte Einführung. Diese Ergebnisse sind ein deutlicher Hinweis
darauf, dass die Wirksamkeit eines Schulungsansatzes kontextabhängig sein kann und
dass Beratungsstellen die Kosteneffizienz im Auge behalten und ihre Beratungsansätze
anpassen sollten, um eine hohe Akzeptanzrate zu erreichen.

Die Ergebnisse der einzelnen Unterkapitel werden im folgenden Kapitel 5 in einen

generellen Kontext gesetzt, wobei in Kapitel 6 dann die entsprechenden Limitationen des
Forschungsansatzes aufgezeigt werden. Die Zusammenfassung wie auch die abgeleiteten
politischen Handlungsempfehlungen werden in den Kapitel 7 und 8 dargestellt.

Die generellen Ergebnisse dieser Dissertation sind, dass erstens die in dieser Arbeit
dargestellten Technologien nachhaltig und kosteneffizient genug sind um, zumindest
temporär, die vorherrschenden traditionellen Methoden der Lebensmittelverarbeitung
und Lebensmittelzubereitung zu ersetzen. Zweitens ist der Zugang zu Energie durch
passive Solartrockner eine wichtige Komponente im Kampf gegen den Hunger und
generelle Ernährungsunsicherheit. Drittens sollten Regierungen und relevante Akteure in
den Kampf gegen Hunger und Energieunsicherheit die lokalen Kontexte in ihre Planungen
einbeziehen und die entsprechenden und angemessenen Bildungs- oder
Informationsansätze wählen.
 14

|
VII List of featured publications

Paper 1:
Matavel, C., Hoffmann, H., Rybak, C. Steinke, J., Sieber, S., & Müller Klaus (2022).
Understanding the drivers of food security among agriculture-based households in Gurué
District, Central Mozambique. Agriculture & Food Security 11, 7.
https://doi.org/10.1186/s40066-021-00344-3

Paper 2:

Matavel, C. E., Hoffmann, H., Rybak, C., Hafner, J. M., Salavessa, J., Eshetu, S. B., & Sieber, S.
(2021) Experimental evaluation of a passive indirect solar dryer for agricultural products
in Central Mozambique. Journal of Food Processing and Preservation, DOI:
https://doi.org/10.1111/jfpp.15975

Paper 3:

Matavel, C.E., Hafner, J. M., Hoffmann, H., Uckert, G., Massuque, J., Rybak, C., & Sieber, S.
(2022) Toward energy saving and food safety in Central Mozambique: the role of
improved cook stoves and heat retention boxes. Energy, Sustainability and Society 12, 26.
https://doi.org/10.1186/s13705-022-00352-6

Paper 4:

Matavel. C. E., Kächele, H., Steinke, J., Rybak, C., Hoffmann, H., Salavessa, J., Sieber, S. &
Müller, Klaus. (2022) Effect of passive solar drying on food security in rural Mozambique.
Scientific Reports 12, 17154. https://doi.org/10.1038/s41598-022-22129-9

Paper 5:

Matavel, C. E., Kächele, H., Hafner, J. M., Rybak, C., Uckert, G., Hoffmann, H., Kipkulei, H.,
Massuque, J., Steinke, J., Sieber, S. (2023) How to increase cookstove adoption? Exploring
cost-effective dissemination techniques in Central Mozambique. Energy Research & Social
Science 100, Article 103082. https://doi.org/10.1016/j.erss.2023.103082
 15

|
VIII Additional Peer-reviewed Articles with
Relevance to the PhD Topic
Matavel, C., Hoffmann, H., Rybak, C. et al. Passive solar dryers as sustainable alternatives
for drying agricultural produce in sub-Saharan Africa: advances and challenges. Discover
Sustainability 2, 40 (2021). https://doi.org/10.1007/s43621-021-00049-4

Berrang-Ford, L., Siders, A. R., Lesnikowski, A., Fischer, A. P., Callaghan, M. W., Haddaway,
N. R., Mach, K. J., Araos, M., Shah, M. A. R., Wannewitz, M., Doshi, D., Leiter, T., Matavel, C.
E., […] Abu, T. Z. (2021) A systematic global stocktake of evidence on human adaptation
to climate change. Nature Climate Change 11, 11, 989–1000.
https://doi.org/10.1038/s41558-021-01170-y

Massuque, J., Matavel, C. E., Trugilho, P. F. (2021) Outlook for the biomass energy sector
in Mozambique: policies and their challenges. Journal of Energy in Southern Africa 32, 4,
1-10.

Massuque, J., Matavel, C. E., de Paula Protásio, T., Trugilho, P. F. (2021) Combustion
performance of charcoal: a comparative study on Miombo woodland native species and
Eucalyptus grandis. Biomass Conversion and Biorefinery.

Massuque, J., De Assis, M.R., Loureiro, B.A., Matavel, C. E. & Trugilho, P. F. (2021).
Influence of lignin on wood carbonization and charcoal properties of Miombo woodland
native species. European Journal of Wood and Wood Products. 79, 527–535.
https://doi.org/10.1007/s00107-021-01669-3
 16

|
1 Introduction

1.1 State of food and nutrition security in Africa

Recent estimates indicate a rising trend in the number of people experiencing FI and
undernourishment globally (UN, 2021a). Between 2019 and 2020, the prevalence of
undernourishment increased from 8.4% to around 9.9% (FAO et al., 2021) due to the
unprecedented effects of climate change, the COVID-19 pandemic, population growth,
unaffordability of healthy diets as well as poverty and inequality (Carducci et al., 2021,
Holleman et al., 2020, Herforth et al., 2020, Springmann, 2020, Tian et al., 2021). In Africa,
the number of people facing hunger is more than double the proportion of any other
region. According to FAO, around 60% of the African population was facing moderate or
severe FI in 2020. Projections also indicate that while Asia is expected to experience a
substantial reduction in hunger, Africa will face a significant increase, from more than 280
to 300 million people by 2030 (FAO et al., 2021), with small-scale farmers being the most
affected (Ng’endo et al., 2017). They supply up to 80% of the food (Ng’endo et al., 2017),
but the low productivity constrains them compared to large-scale producers. The average
annual income for small-scale farmers in African countries such as Malawi, Mozambique
and Niger is 300 US dollars, while in Albania, Guatemala, Iraq, and Serbia is more than
3000 US dollars (UN, 2021b). In addition, post-harvest losses (PHLs) still remain a
persistent problem (Affognon et al., 2015, Stathers et al., 2020).

A substantial amount of harvested food does not meet the appropriate standards in terms
of quality, volume and safety due to inadequate post-harvesting management, food
preservation capacities, processing technologies, storage infrastructure and market
facilities (Zorya et al., 2011) triggered also by insufficient energy access (Makungwa et al.,
2013, Sola et al., 2016). Therefore, eliminating FI is a complex challenge that is unlikely to
be achieved by solely increasing quantitative food production. On the contrary, it requires
a holistic solution that involves empowering smallholder farmers through access to
improved or new technologies to increase productivity and ensure effective food
processing and preservation to reduce losses. Latter can guarantee value-added products
for income-generation and self-employment (Tefera et al., 2011, Fasoyiro and Taiwo,
2012, Salami et al., 2010, Wambugu et al., 2018, Neme et al., 2021), especially for women,
Introduction 17

who are traditionally responsible for many postharvest activities (Stathers et al., 2020).
If adequately performed, food processing – including cooking – can help to destroy food-
borne microbes and toxins, improve the bioavailability of nutrients, extend shelf life
(thereby reducing PHLs), and improve sensory characteristics (Van Boekel et al., 2010,
Weaver et al., 2014, Augustin et al., 2016).

Despite the considerable efforts that have been put in place to increase agricultural
productivity in sub-Saharan Africa (SSA) (Adetutu and Ajayi, 2020), the current rates of
progress towards the sustainable development goal (SDG) 2 (end hunger by 2030) will
still need to increase three- to four-fold (Nicolai et al., 2016). Moreover, an increase in
agricultural productivity can indeed increase food supply but is also likel to increase the
pressure on the limited natural resources and thus create new or boost existing
environmental problems (Evans, 2003). Therefore, special attention is also given to
reducing PHLs, which is central to achieving SDG 12 (to ensure sustainable consumption
and production patterns) (FAO, 2019). Reducing PHL is viewed has a sustainable means
to improve food available and preserve limited natural resources (Zorya et al., 2011,
Hodges et al., 2011). However, smallholder farmers in SSA generally use traditional post-
harvest handling methods, resulting in substantial losses (Jennifer et al., 2019). Improved
post-harvest technologies have been promoted to overcome the challenges imposed by
traditional methods and increase food security (FS). However, the low adoption of these
technologies – partialy due to high initial costs – has also led to declining investments in
this area (Zorya et al., 2011). Additonal developments towards FNS have also taken place
in the cooking sector. Cooking processes are likely to influence the content of nutrients
and bioactive compounds in food naturally present in the raw materials (Lobefaro et al.,
2021). Thus, they directly affect the food utilization dimension of FNS. However, there has
been limited research focusing on the linkages between access to cooking energy and FNS.

Knowledge and awareness are essential to increase demand for FS-related technologies
(Muro and Burchi, 2007). However, despite various developments that have taken place
in terms of designing scalable technologies for rural areas, creating demand is still a
challenge. Therefore, Goodwin et al. (2015) suggest a deeper understanding of the
effectiveness of training approaches used to disseminate innovation since the evidence is
still mixed. While some literature suggests that group-based delivery approaches can be
most effective in transferring innovation (Darr and Pretzsch, 2008, Hoffmann et al., 2007),
others indicate that individual delivery approaches are superior, especially in situations
Introduction 18

where the innovation is relatively complex (Feder et al., 2010, Gerster-Bentaya and
Hoffmann, 2011). Rogers (2003) highlighted the complexity of technology adoption
processes by showing that different categories of adopters can be found within
communities. Thus, training strategies or rural extension programs can potentially
strengthen technology dissemination by targeting individuals with the power to influence
others’ decision to adopt (Van den Bulte and Joshi, 2007).

1.2 The concept of food and nutrition security (FNS)

The current concept of FNS results from a sequence of definitions and paradigms.
Hoddinott (1999) indicated the existence of approximately 200 definitions and 450
indicators of FS. Initially, the concept of FS was developed and referred to the adequacy
of the overall food supply at a regional, national, or even global scale (Weingärtner, 2009).
However, the adequacy of food supply does not necessarily result in adequate food intake
by certain groups of people or individuals; thus, the concept has also evolved to be applied
at the community, local, household or individual levels (Foster, 1992). A widely accepted
definition states that FS “is a situation that exists when all people, at all times, have
physical, social and economic access to sufficient, safe and nutritious food that meets their
dietary needs and food preferences for an active and healthy life” (FAO, 2000).
Traditionally, FS focuses on food availability, access and stability (CFS, 2012). Food
availability refers to the physical existence of food, which depends on food production and
food purchase in the markets (Fasoyiro and Taiwo, 2012). On the national level, it is a
combination of domestic food production and imports, food aid, and domestic food stocks,
as well as the underlying determinants of each of these factors (Weingärtner, 2009).
Access focusses on households’ ability to generate sufficient income, which, together with
own food production, is used to meet food needs. Thus, food access be achieved even if
the household is not self-sufficient in food production (Hoddinott, 1999). In the African
context, where most of the population facing food insecurity (FI) practice smallholder
farming, food processing and associated activities are essential factors in promoting food
access through increased income resulting from the commercialization of value-added
foods (Fasoyiro and Taiwo, 2012). However, the role of value-addition technologies in
attaining food security has not been adequately addressed in research so far (Alamu and
Mooya, 2017, Neme et al., 2021). Stability is the temporal dimension of FS and refers to
the ability of a population, household or individual to have access to adequate food at all
times (FAO, 2006).
Introduction 19

The acknowledgement of the importance of crucial nutrition concerns such as care and
feeding practices, public health and sanitation issues led to the combination of the term
“food security” with “nutrition security”, latter particularly focussing on individual and
household food consumption and how food is utilized (CFS, 2012, El Bilali et al., 2019).
Food utilization in turn refers to the human body's ability to ingest, digest, and metabolize
food. In addition to an adequate diet, it requires a healthy physical environment, adequate
sanitary facilities (to avoid diseases) and at least a basic understanding of proper health
care, food preparation and storage processes (Weingärtner, 2009). Food utilization can
also refer to instances where food is available but is not perceived as “good” food and thus
is not consumed (Wijesinha-Bettoni and Mouillé, 2019). Recently, the term “food and
nutrition security” emerged out of these debates and has since been widely used in
academia and many organizations such as the International Food Policy Research
Institute (IFPRI), Food and Agriculture Organization of the United Nations (FAO), the
World Food Programme (WFP) and the United Nations Children's Fund (UNICEF) to
emphasize the importance of nutrition for achieving food security (El Bilali et al., 2019).
Actions to ensure that food systems provide all households with stable access to sufficient,
appropriate and safe food should be complemented with nutrition-oriented action to
ensure that households and individuals have the capacity and supportive health as well
as environmental conditions necessary to obtain adequate nutritional benefits from the
food (CFS, 2012). Following Weingärtner (2009), FNS is a condition under which
“adequate food (quantity, quality, safety, socio-cultural acceptability) is available and
accessible for and satisfactorily used and utilized by all individuals at all times to live a
healthy and active life”. In short, achieving FNS requires that food is available, accessible,
stable (i.e. constantly available and accessible) and appropriately utilized.

FNS is an outcome of food systems (Figure 1.1) and thus depends on investments in
agriculture and its associated sectors, such as production, storage, aggregation, post-
harvest handling, transport, processing, distribution, marketing, disposal and
consumption of food (FAO, 2000, FAO, 2021).
Introduction 20

Figure 1.1: Interlinkages between food system, socio-economic systems and Food and
Nutrition Security. Adapted from Mbow et al. (2019)

One of the critical steps in the fight against FI is an investment in research on how it is
distributed within a community and households (Gundersen and Ziliak, 2018). According
to Pangaribowo et al. (2013), food and nutrition policies or interventions require
information on the following:

a) The current absolute and relative status and evolution of FNS,

b) The causes of change in FNS,
c) The possible actions and possible impacts of those actions, and
d) The cost-effectiveness and actual implications of FNS intervention.

The appropriate metric for collecting the information above depends on the
conceptualization of the construct to be measured and the intended use of the data to be
collected (Jones et al., 2013).

1.2.1 Food and nutrition security measurement

The FNS concept incorporates economic, social, nutritional, and psychological
considerations (Cafiero et al., 2014), which has also resulted in the evolvement of
Introduction 21

respective food security metrics (Jones et al., 2013). Different organizations, at different
scales (at global, national, household and individual levels), have proposed and applied
multiple tools (or indicators) across various disciplines to identify and monitor those
most in need of food security interventions. However, these indicators are, as a result on
their multiple application foci, heterogeneous and are thus used for a broad variety of
purposes, often producing mixed results (Maxwell et al., 2014, Santeramo, 2015, Jones et
al., 2013). However, this multitude of indicators and instruments are necessary
(Pangaribowo et al., 2013) since a single measure cannot capture all FNS dimensions.
Furthermore, different instruments are applied by different bodies with different
capacities (e.g. supranational bodies vs. local NGOs) (Jones et al., 2013, Hendriks et al.,
2016). In fact, policy intervantions in the agricultural sector can affect several key FNS
outcomes thus multiple indicators have been used to understand the pathways through
which agriculture can improve nutrition outcomes (Herforth and Ballard, 2016). Table 1.1
summarizes the FNS indicators commonly used in agricultural projects.
 Introduction
Table 1.1: Food and nutrition security metrics. Adapted from Pangaribowo et al. (2013), Jones et al. (2013) and Herforth and Ballard (2016)
Indicator
Stunting
Wasting
Underweight
Overweight/obesity
Maternal BMI
Vitamins and micronutrients
status
Household Dietary Diversity Score
(HDDS)
Food Consumption Score
(household-level indicator)
Women's dietary diversity (usually
WDDS)
Individual Dietary Diversity Score
(IDDS) – young child
Household Food Insecurity Access
Scale (HFIAS)
Household Hunger Scale (HHS)
Coping strategies
Months of inadequate household
food provision
Meal frequency
Nutrition knowledge or
behaviours
Women's empowerment
22
Purpose Description Level of measurement
Used to assess childhood malnutrition Low height-for-age Individual
Used to assess the prevalence of childhood malnutrition Low weight-for-height Individual
Use of weight to determine whether a person is within the Low weight-for-age Individual
healthy weight range
Use of weight to determine whether a person is within the Excessive fat accumulation due to excess intake of dietary energy Individual
healthy weight range
Used to assess the wellbeing of the pregnant mother and, by Assessment of body fat based on height and weight Individual
inference of her baby
To assess consumption or deficiency states of essential Nutritional status based on measurements of nutrient stores, Individual
micronutrients (vitamins and minerals) functional forms, excreted forms, and/or metabolic functions
Used to assess household's economic access to a variety of Sum of equally weighted response data on the consumption of 12 Household
foods food groups; score obtained from 0 to 12
Establish the prevalence of food insecurity Sum of differently weighted response data on the consumption Household
of different food groups consumed by a household during the seven
days before the survey
Used to assess women's access to a variety of foods Sum of equally weighted response data on the consumption of 9 Individual
food groups
Used to assess children (6-23 months) access to a variety of Number of food groups consumed by the child the previous day Individual
foods
Assess FS status within regions or households Sums responses to 9 questions related to four domains of food Household
security, including 4-level frequency response questions; a score
from 0 to 27 is obtained and may be categorized into a 4-level
variable
Assess hunger status within and across contexts Sums responses from three questions related to hunger and lack of Household
food, including 3-level frequency response questions; a score from 0
to 6 is obtained and may be categorized into a 3-level variable
Used to identify vulnerable households Locally adapted list of coping strategies and the frequency of their Household
use; severity weightings are assigned to each strategy
Measure seasonal differences and/or changes in households' Sums total number of months in the past year the household did not Household
abilities to address food vulnerability have enough food to meet the family's needs
To obtain frequency and, in some cases, portion size Number of meals per given time period Household, individual
information about food and beverage consumption over a
specified period of time, typically the past month or year.
Used to assess awareness and level of usage of nutrition Knowledge of the concepts and processes related to nutrition and Household
information. health, including diet and health, diet and disease, dietary guidelines
and recommendations
Used to assess women’s capacity to cope with food Women’s access and control of income, their time and labor, Household, individual
insecurity. agricultural land and agricultural
inputs.
Introduction 23

Evidently, each indicator has strengths and limitations that must be acknowledged before
selection. For example, choosing a household level over an individual-level indicator will
result in less time needed and thus fewer resources necessary to collect data. However,
intra-household food allocation, an important measure to capture food utilization, can
only be accessed through individual-level indicators. Thus, for identifying the most
adequate indicators/instruments to measure FI, it is essential to know, among other
factors, explicitly:

a) The dimension of FNS that is intended to be captured (e.g. food availability, food
access, utilization of food, or stability);
b) The purpose of the data (e.g., early famine warning, targeting food aid, screening
at-risk households for intervention, monitoring changes in food security status
over time, monitoring program utilization, evaluating programs and policies, or
advocacy);
c) The periodicity to be evaluated (e.g., whether there will be repeated measures);
d) The scale of measurement (national, regional, household, individual);
e) The available resources for planning, data collection, analysis, and the intended
application of the findings.

1.2.2 Linkages between energy and FNS

The importance of energy as a development trajectory has been highlighted by SDG 7:
“Ensure access to affordable, reliable, sustainable and modern energy for all”.
Improvements in energy availability can make a big difference to human development
indicators (Brew-Hammond, 2012). Despite limited evidence, linkages between energy
access and FNS have been occasionally reported (Sola et al., 2016) and include mainly the
aspect of cooking (energy), processing and storing. Many food products require cooking
before consumption. Thus, food consumption can be negatively affected by a lack of fuel,
which prevents people from being able to cook food properly or disinfect water by boiling
(WFP, 2019), with negative implications on the dimension of food utilization. Cooking is
strongly linked to the ability of the human body to digest and absorb nutrients from the
available food (Boko et al., 2007, Hoffmann, 2016). Rural communities affected by cooking
energy scarcity may stop cooking protein-rich food, such as beans that need simmering,
in favor of fast-cooking low-protein meals (Scheid et al., 2018). In addition, energy use for
water purification might be suspended (Kees and Feldmann, 2011, Murphy, 2001). FAO
(2013) also reports that the quantity, quality and nutritional value of the food consumed
Introduction 24

are affected due to the lack of safe access to sufficient and appropriate cooking fuel. This
forces people to change their eating and cooking habits, move to less preferred and lower
quality food or reduce the number or undercooking meals. These measures have
detrimental impacts on nutrition (WFP, 2019). Moreover, energy is critical for food
preservation and storage (Lee and Okos, 2011), as the absence of affordable food
processing methods and technologies leads to food waste, loss of nutritive properties and
health problems (toxicity) (Bradford et al., 2020).

On the one hand, suboptimal processing techniques, including inefficient cooking

strategies (MacCarty et al., 2010, Adkins et al., 2010, Sutar et al., 2015), as well as not well-
adopted preservation strategies for high-quality foods such as vegetables, negatively
affect FNS (Mulokozi and Svanberg, 2003). On the other, these suboptimal strategies are
often without any alternatives, as energy scarcity prevents food processing strategies that
preserve food adequately and effectively. Restricted access to energy at each activity of
the food value chains (production, processing and preservation) limits the ability of
farmers and agri-enterprises to raise productivity, cut losses and cope with a changing
climate and other shocks (IRENA and FAO, 2021).

Access to energy opens up new economic opportunities in rural areas thus boosting
resilience and strengthening livelihoods. Energy can power food processing tasks that
increase the quality of the final product, increase storage time and ease handling (e.g.
refrigeration, drying, smoking, pasteurization, fermentation, canning and packaging) to
reduce post-harvest food loss, increasing availability of nutritious foods at the household
level and enabling farmers to determine the timing of crop sales, in turn improving
household income (WFP, 2019).

It is noteworthy that most poor households rely on traditional uses of forest biomass and
solar energy for cooking and food processing. The traditional uses of solar energy – mainly
for drying food – have significant limitations since it leads to high product losses due to
inadequate drying, fungal growth, and encroachment of insects, birds and rodents
(Purohit et al., 2006). This can lead to aflatoxin contamination and other food safety
challenges that lower food utilization and stability of the food supply (Mabiso et al.,
2014b). Biomass energy has detrimental effects, especially for women and children in
rural areas who walk several kilometres and hours to collect firewood for cooking
(Ugembe et al., 2022). Scheid et al. (2018) found that, due to fuel scarcity, a high
percentage of the households in their case study areas are forced to increase the
Introduction 25

frequency of collection and respective walking distances. In Mozambique, women and

girls spend hours cooking (MISAU and INE, 2018, Ugembe et al., 2022), reducing the time
for other subsistence activities or recreation substantially. According to IEA (2017),
households relying on forest-sourced biomass for cooking spend around 1.4 h/day
collecting firewood and several hours cooking and are exposed to indoor pollution.
Atanassov et al. (2012) found that conditions such as acute respiratory infections, chronic
obstructive lung disease, and lung cancer are common among long-time biomass users in
Mozambique. Clean cooking technologies can prevent cancer, chronic obstructive
pulmonary disease, respiratory infections, ischemic heart disease, and stroke (Gill-Wiehl
and Kammen, 2022).

Thus, the availability and access to affordable energy technologies for agricultural and
domestic uses is a crucial component of FNS. Measures to increase access and energy
efficiency in food systems are valuable for raising productivity, strengthening supply
chains, reducing food and income losses, and improving food security (IRENA and FAO,
2021), especially in poor rural communities. The use of efficient stoves has been
promoted due to associated positive implications for users, especially women, through
changes in cooking habits and time–labor allocation (Nielsen et al., 2015). Besides the
strongly reduced indoor air pollution (Desai et al., 2004, Fullerton et al., 2008, Naeher et
al., 2007) – which is responsible for 600,000 annual deaths in SSA (IHME, 2022) – women
and children can decrease their workload as especially wood quantities needed for
providing cooking energy decline substantially. This allows affected individuals to invest
more time in food and nutrition security activities, such as farming.

1.3 Food and nutrition security situation in Mozambique

Despite the significant achievement of halving the number of hungry people in 2015, FI
and malnutrition are significant problems in Mozambique (WFP Mozambique, 2022). The
country has reduced hunger from 56% in 2003 to 24% in 2015 (Macassa et al., 2018).
However, 80% of the population cannot afford an adequate diet, and 42.3% of children
under five are stunted, with peaks of 46% in rural areas. These levels fall within the WHO
threshold of ≥40% for “very high” stunting (WHO and UNICEF, 2021). Chronic FI is
estimated at 24% and malnutrition at 25% (WFP Mozambique, 2022), which is directly
associated with 19% of school failures and 25.6% of infant mortality (FAO et al., 2022).

There are regional differences regarding the prevalence of chronic malnutrition in

children under the age of 5. The provinces in the central and northern regions have
Introduction 26

chronic malnutrition rates above 40%, while in the south, the proportions are below 30%
(Castigo and Salvucci, 2017). Such high levels of malnutrition, especial in the northern and
central regions, are partially caused by insufficient micronutrient intake due to poor diets
(USAID, 2022). Moreover, low rates of production technologies’ use and the consequent
low productivity combined with high PHLs and low-quality products due to poor handling
and storage practices add to the major causes of food and nutrition insecurity (FNI). For
example, up to 24% of the total produced maize is lost in the central provinces of Manica
and Zambézia (Delgado et al., 2021).

Staple food production is and will continue to play a significant role in the Mozambican
economy and is a fundamental determinant of people’s diets (Ferrão et al., 2018). In fact,
cereals account for 72% of energy (kcal) in the diet (FAO et al., 2022). Therefore, high
losses of staple foods due to inadequate post-harvest handling significantly impact FNS.
Other factors, such as limited (economic) access to quality foods, lack of nutrition
knowledge, as well as factors such as the women’s heavy workload and duties that limit
child feeding and care, are also important causes of persistently high levels of
malnutrition (WFP Mozambique, 2022, Ferrão et al., 2018).

Sustainable energy provision for staple foods processing and preparation can help
overcome some of the aforementioned challenges. However, Mozambique is challenged
by a lack of access to electricity, particularly in rural areas, where only 6% have access
(Salite et al., 2020). The poor cannot always afford conventional forms of energy and, thus,
use traditional means for cooking and processing, which have been associated with
negative environmental, economic, social, cultural and health impacts. Access to
affordable and reliable energy services is critical to ensuring that individuals properly
utilize food and might stimulate de production of goods and services that may encourage
the sustainable production and preservation of the food needed to fight hunger (Ugembe
et al., 2022). Thus, it is essential to implement market-based, sustainable energy
approaches that strengthen food security by providing people with the means to process,
cook and boost resilience activities that support local food value chains (WFP, 2022).
Nevertheless, interventions seeking to promote efficient technologies and/or the
substitution of the traditional forms of energy with cleaner and improved sources in
Mozambique have had limited success or are at a very early stage (Nyambane et al., 2020,
Otte et al., 2018, Samuelsson and Deslandes, 2017, Bernardo et al., 2021, Pailman et al.,
Introduction 27

2018). Moreover, energy provision for post-harvest handling is still a widely neglected
area in FNS research (Cugala et al., 2012).

1.4 Problem definition, research objectives and hypotheses

The state of FI in Mozambique is alarming, and there is already an extensive amount of

academic literature trying to address the mechanisms by which food systems can
guarantee food and nutrition (de Brauw et al., 2019, Ferrão et al., 2018, Zhou and
Hendriks, 2017, Mabiso et al., 2014a, Charrua et al., 2021, Aabø and Kring, 2012).
Nevertheless, challenges regarding post-harvest losses are still persistent (Delgado et al.,
2021), and staple crops production is the primary source of diet, particularly for people
living in rural areas (Charrua et al., 2021). Moreover, despite the recognized importance
of reliable access to efficient energy technologies in transforming food systems, empirical
research has paid little attention to its role on FNS, especially for smallholder farmers in
rural areas, where access to conventional forms of energy is limited. Many studies focus
on the production side and overlook post-harvest activities and food preparation
practices (Leonardo et al., 2015, Nyyssölä et al., 2014, Mango et al., 2017). The attempts
that have been made so far in orde to assess the technical performance of energy
technology innovations (Bernardo et al., 2021, Samuelsson and Deslandes, 2017) have
paid little attention to social and cultural factors and other aspects related to consumer
needs and preferences (for example, the quality of the final product). While the
development of technologies is an essential step towards improving the food systems,
analysing aspects related to the demand side is also crucial. The demand for technology
is complex and thus requires a proper understanding of the role of different actors,
extension services, and institutional reforms (Sulaiman and Hall, 2002). This would
provide greater certainty on the way forward and avoid costly technology development
and distribution, which may lead to economic losses due to a lack of acceptance and
adoption. In line with this, diverse techniques using different training approaches have
been used to support and motivate users to adopt new energy technologies (Lindgren,
2020, Champion et al., 2017, Evans et al., 2018, Lewis et al., 2015, Namagembe et al., 2015,
Furszyfer Del Rio et al., 2020, Otte et al., 2018, Kitinoja and Barrett, 2015). However, the
role of the dissemination approaches in fostering the adoption of energy technologies by
smallholder farmers is still not well understood. A holistic and quantitative assessment
that integrates the demand and supply of technology innovation is missing. This work fills
this gap by addressing two separate but essential lines: 1) Food preservation techniques
Introduction 28

and technologies – especially solar drying (e.g. (Ayua et al., 2017)) – and 2) cooking energy
provision and retention for food preparation (e.g. (IEA, 2006)) in central Mozambique
(Zambézia province), a region with highest needs to improve FNS, were malnourishment
(e.g. stunting and anaemia) is above the average national level. This research will allow
stakeholders involved in food security programs (governments, policy makers, non-
governmental organizations, donor organizations, and smallholder farmers, among
others) to understand the potential of energy technologies for improving FNS and design
the best strategies and approaches to benefit from technological innovations.

This PhD work aims to investigate to which extent potentially (more) efficient energy
technologies might foster FNS in specific target groups. Therefore, the following research
objectives were defined and addressed within five subchapters:

a) To explore the role of staple foods production on FNS at household-level;

b) To test the performance of a passive solar drying technology;
c) To assess the performance of improved cooking and heat retention technologies
with regard to fuel and time saving and adequate food storage temperatures;
d) To analyze the effects of food processing technologies on FNS.
e) To understand the effect of the training approach on technology adoption;

The underlying hypotheses guiding the work are:

a) Staple foods production plays a significant role in improving household FNS

situation;
b) Solar drying technology increases drying efficiency and results in a high-quality
product;
c) Energy-efficient technologies reduce cooking time and household energy
consumption compared to traditional uses of biomass energy;
d) The adoption of food processing technologies results in improvements in the
household FNS situation.
e) The training approach plays a significant role in stimulating the early adoption of
energy technologies;
 29

|
2 Synopsis of this work

The present research started with an assessment of the food security situation and its
drivers to further understand the role of staple foods production on households’ ability to
guarantee FNS. A holistic approach was followed to identify the most critical points for
intervention. Previous research has shown the seasonal nature of household FNS in
Mozambique (Selvester et al., 2008). Some technological innovations that hold the
potential to induce change in the FNS situation through energy provision (a prerequisite
for a sustainable food system) were then proposed. The immediate impacts of energy
access on FS refer to the effects on food production (availability), income (economic
access to food) and food conservation (stability) and preparation (utilization) (Candelise
et al., 2021). Stimulating agriculture and food systems innovation fast-tracks sustainable
solutions (Augustin et al., 2021, Langendahl et al., 2016). Despite the efforts to increase
agricultural production (Chilundo et al., 2020), post-harvest losses (PHLs) in Mozambique
limit the poor population’s access to quality food, thereby exacerbating food insecurity
(FI) (Popat et al., 2020). Thus, this research focused on post-harvest handling,
preparation, storage and consumption of food. We excluded other food system subsectors,
namely, aggregation, transport, distribution, and marketing, since the research was
conducted in an area were more than 90% of all agricultural land is cultivated by small-
scale farmers who have 1.5–2.5 ha, mainly for subsistence purposes (Baumert et al.,
2019). Furthermore, the analysis was also extended to improve the understanding of the
extension delivery approach to increase the effectiveness of the behavior change
techniques in strengthening technology dissemination.

This cumulative dissertation addresses the research objectives in five individual research
papers. The following chapter (Chapter 3) describes a brief overview of the research
methodology. In chapter 4, the research results are presented, focusing on 1) the state of
the role of staple foods on the FNS situation in Guruè district; 2) the performance of solar
drying compared to the traditional open-sun drying; 3) the performance of an improved
cooking stove and a heat retention box compared to the traditional techniques for cooking
and food storage; 4) the effect of passive solar drying on FNS; 5) the dissemination
strategies for clean cooking technologies. Chapter 5 presents a synthesis of the results,
Synopsis of this work 30

and Chapter 6 acknowledges the study's limitations and future research needs. The
overall conclusion and specific policy recommendations are presented in chapters 7 and
8, respectively. The research concept showing the proposed mechanisms through which
FNS can be improved in the context of our areas of study is presented in Figure 2.1.

Figure 2.1: Research concept: proposed mechanisms to improve food and nutrition
security
 31

|
3 Research methodology

3.1 Study area

The primary data used in this PhD dissertation was collected in Gurué district, which is
located in Zambézia Province, Central Mozambique (Figure 3.1). Gurué district was
selected since it generates a large surplus during harvest season, but the local farmers still
relay on traditional post-harvesting methods (Matavel et al., 2020, Alemu and Van
Schalkwyk, 2008, Norfolk and Hanlon, 2012). In this area, despite the high proportion of
the population practising agriculture and animal husbandry as the main occupations (INE,
2018), few households have regular cash income (Arimond et al., 2011), and there is a
high prevalence of food insecurity (UNICEF, 2018). The local diets mainly consist of maize
and cassava (Arimond et al., 2011). Agricultural activities occupy 147,760 ha, of which
93% is primarily used for small-scale farming, and the remaining 7% is used for large-
scale commercial production (Soares, 2017).

Figure 3.1: Study area location

Research methodology 32

3.2 Research design

The defined innovations (improved solar drying, efficient stoves and heat-retention box)
were implemented based on a participatory action research methodology (Mapfumo et
al., 2013). Different performance monitoring missions continuously and closely
monitored the action research methodology. A subsequent comparison of different
innovations - and their effects in contrast to the status quo in the control village – was
realized by a set of specific indicators, which allowed a comparative analysis of
achievements and triggered changes.

Before implementation, a baseline survey was conducted as the starting point to analyze
the status-quo in selected case study villages. The survey was administered to selected
agriculture-based households in the study area. The aim was to collect data that allows a
multifaceted characterization of individual households regarding demographics, income
activities, agricultural performance and FS situation. Latter was measured in the
harvesting and pre-harvesting seasons to account for within-year variations of FS.
Multiple indicators were selected to capture all four dimensions of FNS (availability,
accessibility, stability and utilization). These household characterizations were crucial to
design the research on the dissemination approaches and the impact of implemented
technologies. In total, the baseline survey covered 296 households. Subsequently,
meetings with the local villagers were held to jointly agree on the design and testing
procedures.

The overall performance and efficiency of the implemented innovations were assessed as
compared to the locally existing techniques. The evidence presented in this dissertation
was generated through multiple interconnected activities described in detail in the
methods parts of the result chapter. This section gives a brief overview of these activities.
Improved technologies that could potentially improve FNS on local and regional levels
were identified and selected as research foci. These technologies were implemented and
compared with locally applied techniques throughout the research. The improved
technologies included: 1) improved solar drying (indirect sun exposition) – a wooden and
locally produced dryer that allows an indirect drying of food – which conserves heat and
UV radiation-sensitive nutrients as best as possible under given conditions; 2.) improved
and efficient mud stoves, a simple technology already implemented and tested in rural
Tanzania (Uckert et al., 2017), which decreases fuelwood consumption and additionally
frees time due to less cooking tasks; and 3) a heat retention box, insulated with layers of
Research methodology 33

PE foam and aluminium foil to improve storage conditions of cooked food and prevent the
multiplication of pathogenic microorganisms. The local techniques are: 1) traditional
solar drying (direct sun exposition) – basically the only food preservation technique
currently applied – in which food is stored on the ground and left in the sun to dry, even
though the UV radiation destroys essential micronutrients (Mulokozi and Svanberg,
2003); 2) traditional cooking methods (three-stone fire); an ancient, inefficient technique
whereby cooking takes place above an open fire (Jetter et al., 2012), and 3) locally used
heat-retention techniques, comparable to the heat retention box, even though produced
from locally available materials (banana leaves and leftover firewood).

Six prototypes were constructed and tested simultaneously with the traditional open sun
drying. The tests included technical performance and consumer acceptability.
Controllable factors such as the drying time, solar radiation intensity, relative humidity
and temperature were recorded. During the technical performance assessments of the
implemented technologies, sensory analyses were performed by randomly selected
groups of adult people – both female and male – from different communities. The
participants individually evaluated the texture, aroma, and colour of dried food, following
standard guidelines (e.g. ASTM-International (2015)). People did not have prior
knowledge of the method used to dry the product so as not to influence their choices. The
evaluators specified their level of acceptance using a five points Likert scale (1 = very bad
(dislike very much) to 5 = very good (like very much)). After the tests, we used and
experimental approach to analyze the effects of solar drying on FNS. Hence, 21
communities were selected, and solar dryers were allocated. These solar dryers were
made available to any interested community member. The food security situation of the
households who benefited from solar dryers between August 2020 and January 2022 was
compared to those who did not use the solar dryer. The specific FNS indicators included
household food availability, women's dietary diversity, household food availability and
household food insecurity access scale.

Likewise, 12 non-portable mud-improved cookstoves and heat retention boxes were

constructed and tested against the traditional three-stone-fires and heat retention
systems, respectively. The cooking time and energy consumption were used to assess the
performance of the improved cookstove, and the heat retention ability was used to test
the heat retention box. A group of randomly selected individuals also tested cooked food's
taste, texture, aroma, and color.
Research methodology 34

Furthermore, this study follows another experimental procedure in which local farmers
were instructed on how to construct, maintain and use the ICS using different training
approaches. The aim was to analyze the effect of the dissemination approach on the
dissemination of improved cookstoves. The training also included information on the
advantages of using the ICS. The heads of selected households attended workshops in
groups of ten people for three days or received a 30 min visit for individual video-based
training. An experienced male trainer conducted training sections on construction and
previously trained local female residents hosted demonstrations on how to use and
maintain the ICS.

3.3 Statistical analysis

In the first subchapter of the results section, descriptive statistics were used to summarise
the food security status and the socio-economic characteristics of the study area. Three
indicators were used to assess FS status, namely, Household Dietary Diversity Score
(HDDS), Household Food Consumption Score (HFCS) and Months of Adequate Household
Food Provisioning (MAHFP). To identify the factors affecting these FS indicators, a
generalized ordered logit model with partial proportional odds was fitted in Stata with a
user-written program, gologit2 (Williams, 2006). In subchapter 4.2, the performance of a
PSD was assessed by calculating the thermal efficiency of the dryer and the solar energy
required for 1 kg of moisture removal. The overall thermal efficiency was then obtained
by calculating the average of all daily efficiencies (Singh et al., 2021).

We calculated the total fuelwood consumed and the total cooking time to assess the ICS
performance in subchapter 4.3. A Mann–Whitney U test was used to determine if there is
a significant difference in mean fuelwood consumption between the three-stone fire stove
and ICS. The length of time before food temperature falls below 60 °C was the
performance indicator used for HRS. We used a one-way analysis of variance (ANOVA)
followed by the Tukey test for multiple comparisons to compare the different HRS
performances as we wanted to examine whether there was a difference between the mean
of all possible pairwise comparisons (Lee and Lee, 2018). In subchapter 4.4, Propensity
score matching (PSM) was performed to assess the impact of using a solar dryer on four
food security indicators: household food availability, women’s dietary diversity, months
of adequate household food provision and the household food insecurity access scale. An
endogenous switching regression (ESR) model was used to complement the propensity
score matching and check the robustness of the results. A random effects probit model
Research methodology 35

was applied to examine the relationships between the adoption of ICS and the training
approach in subchapter 4.5. Probit is a valuable model for analysing panel data in which
the dependent variables are a participants’ binary choice, and the independent variable is
randomly assigned by the experimenter (Bland and Cook, 2019).
Results 36

|
4 Results

4.1 Understanding the drivers of food security among agriculture-based

households in Gurué District, Central Mozambique

Agriculture & Food Security (2022), volume 11, Article number: 7.

Custodio Matavel1,2; Harry Hoffmann1; Constance Rybak1; Jonathan Steinke1,2; Stefan

Sieber1,2; Klaus Müller1,2

1Leibniz Centre for Agricultural Landscape Research (ZALF), Eberswalder Str. 84, 15374 Müncheberg,
Germany

2 Department of Agricultural Economics, Faculty of Life Sciences, Humboldt University of Berlin, Germany

© The Author(s) 2022. Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/.

https://doi.org/10.1186/s40066-021-00344-3

Abstract

Background: The prevalence of food insecurity in Mozambique is alarming, despite

progress made during the 2010s. Several studies apply different proxy indicators of food
security (FS) to assess the FS situation. However, these studies overlook the factors
affecting FS, using only a single data point that results in an incomplete picture of FS. Food
security is expected to fluctuate, being better and worse than what studies suggest. Using
a sample of 296 households to assess FS, key drivers conditioning households’ capacity to
achieve FS in Gurué District, Central Mozambique, are identified. Data were collected in
the pre-harvest period and during the harvest period to capture relevant interseasonal
variation of FS. Household FS is assessed using three standard indicators: Household
Results 37

Dietary Diversity Score (HDDS), Household Food Consumption Score (HFCS), and Months
of Adequate Household Food Provisioning (MAHFP).

Results: Each household was classified into a specific FS status depending on the
indicator applied. Generally, most households were classified as being severely or
moderately food insecure during the pre-harvest season, while during the harvest season,
medium and high levels of FS predominated. Nevertheless, varying outcomes were found
depending on the indicator used to assess FS. MAHFP and HDDS are more related to the
consumption of farm-sourced food, while HFCS responds more strongly to purchased
food. Gender and age of the household head, geographic location, size and quality of land,
staples production (especially cassava), livestock and crop diversity, as well as cash crops
had a statistically significant effect on FS indicators.

Conclusions: The study concludes that the decision whether farmers should rely on
staple foods production for increasing their FS status or specialize on cash crops
production to generate income and buy food depends on the indicator used to assess FS,
since each indicator captures a specific domain of food security. Thus, one central
recommendation derived from our results is that policy makers should promote a balance
between market-oriented agriculture and subsistence production to achieve FS.

Keywords: Food consumption, Cash crops, Crop diversity, Staple food, Adequate food
provisioning

Background children under the age of 5 in Gurué

district, which is also located in Central
Nearly 19% of the population in Sub-
Mozambique, is estimated at
Saharan Africa (SSA) is undernourished
approximately 50% [5, 6]. Reasons for
[1]. In Mozambique, despite the progress
this include frequently occurring natural
made in reducing chronic food insecurity
disasters [7], climate change [8], and
(FI), it is estimated at 24% [2] and 42.3%
destructive crop pests that hindered
of children under the age of 5 are stunted
adequate agricultural production [9].
[3]. A study conducted by [4] finds that
The latter negatively affects livelihoods
42–67% of households in Central
in multiple ways, since approximately
Mozambique have experienced hunger
80% of the economically active
during the pre-harvest period. The
population work in the agricultural
prevalence of undernourishment for
Results
sector [10] and, therefore, may be at risk
of FI. This situation is likely to worsen in
the context of the COVID-19 crisis [11],
potentially triggering irreversible long-
term consequences [12].
Studies on the impacts of FI on people’s
well-being indicate that it is associated
with reduced physical health [13],
maternal and child underweight [14],
poor mental health [15], stress [16], as
well as high-risk sexual behavior (e.g. in
SSA) [17]. It is likely to also negatively
affect educational outcomes [18] and the
ability of mothers to adopt exclusive
breastfeeding practices [19]. Therefore,
achieving food security is necessary and
urgent. It is prominently acknowledged
by the United Nations through
Sustainable Development Goals 2 and 3
(end hunger; ensure good health and
well-being, respectively) [20].
Nevertheless, persistent food crises in
the Global South has led many
stakeholders, including policy makers
and academics, to redefine the concept of
FS. This has also resulted in various
changes to the approaches used by
governments and aid organizations to
address FI [21].
The debate over the FS concept has
evolved from the adequacy of country-
level food supplies to dietary energy
adequacy at the household and
38
individual levels, plus, currently,
consideration of the economic, social,
nutritional, and psychological factors
[22]. FAO [23] define FS as access to an
adequate supply of sufficient, safe, and
nutritious food that meet people’s
dietary needs and food preferences for
an active and healthy life. It involves
three physical dimensions, namely,
availability, accessibility, and utilization,
along with one temporal dimension,
stability [24].
A critical aspect for strategies to achieve
FS is the identification of food insecure
households or individuals and the
characterization of the nature of their
insecurity through measurements. This
provides the basis for monitoring the
progress and impact of FS programs [25].
Thus, multiple indicators have been
proposed and applied as a way to identify
and monitor those most in need of food
security interventions. However, these
indicators are quite heterogeneous [26]
and a single measure cannot capture all
its dimensions [27], thus producing
mixed results [28]. In Mozambique, for
example, several studies, conducted in
different regions and applying different
proxy indicators of FS, have found
different results. Proxies applied include
households’ perception of FS [29], food
expenditures [30, 31], number of calories
Results
obtained by each household [32], the
frequency with which households have
experienced different food access
challenges [33], and household food
insecurity access scale [34].
Furthermore, the literature on FS
measurement categorize FS indicators
into two groups, one group uses indirect
approaches to measure adequacy of food
consumption (e.g. dietary diversity and
food consumption scores), while the
other directly measures behaviors and
lived experiences of household food
security (e.g. food insecurity experience
scale) [22, 35]. Hence, the application
and the comparison of different FS
indicators as well as the use of mixed
methods are required to holistically
analyze and describe FS [36]. Moreover,
some indicators capture only one FS
dimension, while others combine two or
more of these dimensions. Nevertheless,
there are indicators that are not clear
about which FS dimension they measure
[35] and there is still lack of consensus on
the effectiveness of most of indicators
used to monitor the progress of FS [22].
All indicators have strengths and
weaknesses; thus, one must take into
account the trade-offs while selecting an
indicator. The selection of appropriate FS
indicators should always take into
account the dimension of food security
39
that is intended to be measured and if the
purpose is to take a more holistic view of
the food security situation, the use
multiple indicators is preferable over a
single indicator [35, 36].
Mawoko et al. [37] and Fitawek and
Hendriks [38], use multiple FS indicators
to provide a holistic picture of FS status
at the household-level in different
regions of Mozambique, including Gurué
district, and examined the effects of
large-scale agricultural investments on
household FS. However, these studies
only used data collected at one point of
time and overlooked other key drivers
conditioning households’ capacity to
achieve FS; thus, giving an incomplete
picture of FS, which is expected to
fluctuate to the better or worse than
what is shown in these studies. In our
study, we attempt to close this gap: (1)
using data collected in the pre-harvest
period and during the harvest period to
capture relevant interseasonal variation
of FS; (2) combining three FS indicators
to capture household access to sufficient
food quantity (energy), food quality
(nutrient adequacy), and stability over a
one year period; and (3) exploring the
underlying drivers of FS at household-
level in Gurué District, Central
Mozambique.
Results 40

Study area and 40°C, and a dry season with,

temperatures ranging from 17°C to 20°C.
The study was conducted in two villages
The average annual rainfall is about
situated in Gurué district: Lioma
1800 mm. The main activities practiced
(15°10′33.3"S 36°48′21.8"E) and
by the population are agriculture and
Mepuagiua (15°46′26.4"S 37°03′38.5"E)
animal husbandry (chickens, ducks, and
(Fig. 4.1). The district is located in the
pigs) [39]. The agricultural activities
north of Zambézia province, central
occupy a total area of 147,760 ha, of
Mozambique, bordering the Republic of
which 10,080 ha (7%) are used primarily
Malawi. The total area of Gurué is
for large-scale commercial production;
565,000 hectares and the population is
the remaining 137,680 ha (93%) is used
estimated at 431,000 inhabitants, which
for small-scale farming. The majority of
corresponds to 7.7% of total population
the small-scale cultivated area is under
in Zambézia province and a population
food crop cultivation, mostly maize,
density of 76 individuals per km2 [39].
cassava, and sorghum [40].
This area has two seasons: a rainy
season, with temperatures between 30°C

Figure 4.1: Location map for the study area

Results 41

Methods
1.962 ∗ 0.99 ∗ 0.1
𝑛𝑀 = = 152 (3)
Study and sample design 0.052

In this study, the purpose is to assess the Where 𝑛𝐿 = total sample size in Lioma and
prevalence and the factors associated 𝑛𝑀 = total sample size in Mepuagiua.
with FS during the period immediately Nevertheless, one and five additional
prior to the harvest and during the households were included in the samples
harvest season, thus, a two-wave panel in Lioma and Mepuagiua, respectively.
study design is used to collect both Thus, the survey covered 296 households
qualitative and quantitative data. Study (n=157 in Mepuagiua and n=139 in
participants were randomly selected Lioma).
from lists provided by the local
Data collection
administrative office and training
We use panel data that was collected in
records provided by local extension
two waves through a semi-structured
services. Before selecting the
household survey questionnaire. The
households, we combined the two lists
first wave took place in February and
and removed duplicates. The sample size
March 2020 and the second wave in May
was determined using the equation by
and June 2021. Priority was given to
[74]:
household heads as responding
𝑍2 ∗ 𝑝 ∗ 𝑞 individuals. Nonetheless, the head of the
𝑛= (1)
𝑑2
household was not available in 58 (32 in
Where 𝑛 = sample size; 𝑍 = abscissa of the Mepiagiua and 26 in Lioma) and 47 (24
standard normal curve (Z = 1.96 for 95% in Mepiagiua and 23 in Lioma) of the
confidence); 𝑝 = proportion of selected households in the first and
agriculture-based households (0.9 for second waves, respectively. Thus, in
Lioma e 0.99 for Mepuagiua); 𝑑= error these households, we interviewed any
set at 0.05; and 𝑞 = 1 − 𝑝. The resulting available adult household member (> 18
sample sizes are demonstrated in year old). All respondents agreed to
Equations 2 and 3. participate in the survey and signed a
consent form, translated into the
1.962 ∗ 0.9 ∗ 0.1
𝑛𝐿 = = 138 (2) Portuguese language. Through the
0.052
survey, we collected data on household
socio-economic characteristics and
demographics, crop production and food
Results 42

consumption patterns focusing on vegetables, milk and milk products,

specific FS indicators (HDDS, HFCS and fruits, oil/fats, meat/poultry,
MAHFP). To access the full dataset, see sugar/honey, eggs and miscellaneous.
Additional file 2. Values for the food groups were either 1
(consumed) or 0 (not consumed). Thus,
Indicators of food security
HDDS is the total number of food groups
In this study, we select indicators that
consumed by household members.
capture household access to sufficient
Households were classified as having low
food quantity and quality, namely,
(HDDS ≤ 3), medium (HDDS = 4–6), or
Household Dietary Diversity Score
high dietary diversity (HDDS = 7–12).
(HDDS), Household Food Consumption
Household food consumption score
Score (HFCS) [35, 75]. The stability of FS
(HFCS): The HFCS is a widely applied
over a period of 1 year is assessed using
food security indicator that is calculated
Months of Adequate Household Food
using the frequency of consumption of
Provisioning (MAHFP) [73]. These
different food groups by a household
indicators are detailed in the following
during a 7 days period [79]. Household
sub-sections.
members were asked the question: “How
Household dietary diversity score (HDDS):
many days in the past 7 days prior to the
The HDDS is a proxy measure of
survey did the household eat each of the
household nutrient adequacy [76]. It lists
food items presented in Table 4.1”.
the number of different food groups
Following this, the data were grouped
consumed over a given reference period
into three food consumption groups,
[77], thus capturing dietary quality
subsequently combined into a composite
within a household [76]. Households
score using standardized weights
were asked to report the foods (meals
(Table 4.1). We created the food
and snacks) that they ate or drank during
consumption groups using standard
the 7 days prior to the survey to capture
thresholds, classifying a household’s food
variability in intake. Subsequently, the
consumption as being poor (HFCS ≤ 21),
food items were grouped into 12 food
borderline (HFCS = 21.5–35), or
groups, as defined by Kennedy et al. [78],
acceptable (HFCS > 35) [79]. HFCS is
to calculate the HDDS for each
calculated by Eq. 4 (respective
household: Cereals, fish and seafood, root
abbreviations are displayed in the
and tubers, pulses/legumes/nuts,
subsequent Table 4.1).
Results 43

HFCS = (MS ∗ 2) + (P ∗ 3) + V + F + (MF ∗ 4) + (D ∗ 4) + (Fa ∗ .5) + (SH ∗ .5) (4)

Table 4.1: Food Groups and Item used for HFCS calculations
Food Group Food Items Weight
Main staples (MS) Maize, Sorghum, Cassava, Rice, Potato, Sweet
2
Potato, Bread, noodles, yam
Pulses (P) Beans, Cowpeas, Mung bean, Groundnuts,
3
Pigeon peas, soybean, Bambara Nuts
Vegetables (V) Green leafy veg, Tomato, okra 1
Fruit (F) Any fruit consumed during the seven day period 1
Meat/Fish (MF) Beef, poultry, pork and fish 4
Dairy (D) Milk, yogurt and other diary 4
Sugar/honey (SH) Sugar and sugar products, honey 0.5
Fats (Fa) Oils, fats and butter 0.5

Months of adequate household food The first step of data analysis was to
provisioning (MAHFP): The MAHFP is an determine the profile of farming
indicator that captures the consistency of activities and the socio-economic
food availability throughout a 12 months characteristics of the study area. We
period [73]. Respondents were asked to computed basic descriptive statistics to
identify in which of the last 12 months provide summary statistics of the data.
they had access to sufficient food to meet Frequency distributions and percentage
their household needs, jointly defined were used for categorical variables.
with villagers as the months they could Mean, minimum, maximum, and
have at least three meals a day. Values for standard deviation were used for
each month were either 1 (yes) or 0 (no). continuous variables. Due to the
Therefore MAHFP was the total number differences in sample sizes, Welch’s t test
of months all household members could was used to compare the means of the
have three or more meals a day. two study sites [80].
Households were classified as most food Generalized ordered logit model (GOLM)
insecure (MAHFP ≤ 5), moderately food
The food security indicators used is this
insecure (MAHFP = 6–9), or least food
study were categorized into three levels
insecure (MAHFP = 10–12).
(cf. Table 4.4), coded as 0 = lower,
Data analysis 1 = medium and 2 = higher level. Thus,
Descriptive statistics these indicators represent the level of FS
Results 44

in an ordinal scale with the level j = 0 be the same for all equations. The general
being the minimum value of the model is presented in Eq. 5.
indicators (low = 0 for HDDS; poor = 0 for
exp(𝛼𝑗 + 𝑋𝑖1 𝛽𝑗1 + 𝑋𝑖2 𝛽2 )
𝑃(𝑌𝑖 > 𝑗) = ,
FCS; most FI = 0 for MAHFCS). For such 1 + [exp(𝛼𝑗 + 𝑋𝑖1 𝛽𝑗1 + 𝑋𝑖2 𝛽2 )] (5)
ordinal dependent variables, ordered 𝑗 = 1,2, … , 𝑚 − 1

logit model is generally suggested [81].

Where 𝑌𝑖 is the recorded FS category for
However, the ordered logit model must
household 𝑖. 𝑃(𝑌𝑖 > 𝑗) is the probability
meet the proportional odds assumption
of a household 𝑖 be in a given FS category.
(also known as parallel line assumption),
𝑗 is the number of categories or cut
such that the coefficients of explanatory
points. 𝛼𝑗 is the regression intercept of
variables on different levels of the
each category. 𝑚 is the number of
dependent variable are the same across
categories of the FS indicators (c.f Table
different cut points [82], but this is
4.4). 𝑋𝑖1 is the vector of explanatory
usually not the case [41]. In this study, we
variables that violate the constraint of
initially applied the ordered probability
parallel line assumption. 𝛽𝑗1 is a vector
models, however, the Brant test [83]
of regression coefficients that varies
suggested that the parallel line
across the category. 𝑋𝑖2 is the vector of
assumption was violated, which means
the rest of the explanatory variables with
that a subset of variables has a varying
a vector of regression coefficients 𝛽2.
coefficient. Thus, to identify the factors
that affect the different food security As we intended to have more

indicators, we applied the generalized parsimonious layout to easily pinpoint

ordered logit models with partial the variables violating the assumptions,

proportional odds model (PPOM) for we used a gamma parameterization (see

each FS indicator (MAHFP, HFCS, and Additional file 1 for detailed output

HDDS) using the program gologit2 in results), an equivalent form of partial

Stata [41]. This model is already applied proportional odds model proposed by

in studies assessing factors associated to Peterson et al. [86]. The model is

food security, e.g., Akbar et al. [84] and presented in Eq. 6.

Ayele et al. [85]. PPOM relaxes the

restriction of parallel line assumption, exp(𝛼𝑗 + 𝑋𝑖 𝛽 + 𝑇𝑖 𝛾𝑖 )
𝑃(𝑌𝑖 > 𝑗) = ,
1 + [exp(𝛼𝑗 + 𝑋𝑖 𝛽 + 𝑇𝑖 𝛾𝑖 )] (6)
allowing one or more coefficients to
𝑗 = 1,2, … , 𝑚 − 1
differ across equations while others can
Results 45

Where 𝑇𝑖 is the vector of explanatory parallel line assumption has one 𝛽

variables that violate the assumption of coefficient and 𝑚 − 1 𝛾 coefficients [41].
proportional odds. 𝛽 is the effect of Thus, if 𝛾𝑖 = 0 the model would reduce to
variables that have the same coefficients the ordered logit model [87]. The
for all possible pairs of FS categories, explanatory variables used in PPOM and
while 𝛾𝑖 is the differential effect of the the expected signs are summarized in
variables on each pair of FS categories Table 4.2. These variables were selected
and indicate the extent to which the due to their potential explanatory power
parallel regression assumption is for food security, based on prior
violated by the variable. Each identified publications.
explanatory variable violating the
Table 4.2: Explanatory variables used in the regression analysis
Expected
Variable Description
signs
Age Age of household head in years +

Education Numbers of schooling years of household head +

Gender Gender of household head (1 = male, 0 = female) +

Household size Number of household members -

Extra income Off-farm/non-farm income (1 if yes, 0 if no) +

Livestock diversity Number of animal species raised by the household +

Location Household location (Lioma=1; Mepuagiua=0) +/-

Farm size Farm size in hectares +

Land quality Perception of having good land fertility (good =1; bad =0) +

Crop diversity Number of crop species produced by the household +

Maize Maize production (1 if yes, 0 if no) +

Sorghum Sorghum production (1 if yes, 0 if no) +

Rice Rice production (1 if yes, 0 if no) +

Cassava Cassava production (1 if yes, 0 if no) +

Cash cropsa Soybean and/or Tobacco production (1 if yes, 0 if no) +

Period of data collection (1 = harvest season, 0 = pre-harvest

Season +
season)
aSoybean and tobacco are the major cash crops in the study area
Results
The age of the household head is a
continuous variable that is used as a
proxy for farming experience [88, 89].
Older farmers are assumed to be more
experienced and resource endowed as
compared to younger farmers [55]. Thus,
we expect in our study that FS increases
with age. Education of the household
head is expected to have a positive
impact on FS, as educated farmers are
better able to obtain information on
improved agricultural techniques and
new economic opportunities, thereby
increasing their productivity [90, 91].
Moreover, it is already demonstrated
that education positively affects FS
[92,93,94]. The knowledge gap and
gender differences in access, control, and
use of assets is a major concern of gender
studies in agriculture [95, 96]. Therefore,
we expect differences between male and
female-headed households with regards
to food security status. This is a dummy
variable equaling 1 if the household head
is male and 0 if the household head is
female. The size of the household is the
total number of people that depend and
live in a household. Although some
literature suggests that lager households
may reflect household labor available for
agricultural activities [51], we expect
households with a larger number of
people to be less food secure, since larger
households have a higher burden to feed
46
[61, 88]. Geographic location is a dummy
variable that is equal to 1 for a household
located in Lioma and 0 for households in
Mepuagiua. Due to differences in some
socio-economic characteristics, we
expect differences in food security status
between the two villages.
Households with off-farm and cash crop
income can purchase more appropriate
and nutritious foods, thus improving
food security [97]. As such, we expect
these households to be more likely to be
food secure than those without non-farm
and cash crops income. These are dummy
variables that equal 1 for households
with non-farm and cash crops (soybean
and/or tobacco) income and 0 otherwise.
Livestock and crop diversity represent
the number of animal and crop species
raised and produced by a household,
respectively. Livestock ownership
represents an additional source of
subsistence, income, and nutritional
requirements [88]. Furthermore, several
studies demonstrate the link between
crop diversity and FS [54, 56, 57].
Therefore, we expect these variables to
have a positive relationship with FS.
Farm size is a continuous variable and
measured in hectares. In this study, we
expect households with larger farm sizes
to have higher probability of being food
secure, since they can diversify
Results 47

production and produce more quantities Results

[60, 61]. Moreover, we expect the Household characteristics
production of staple foods to be
The majority of households, 71% in
significantly related to FS. Staple foods
Mepuagiua and 80% in Lioma, are
represent the foundation for food
headed by men. Only 10% of the
security and an adequate diet [98]. They
households in Mepuagiua have a non-
are sources of marketable surpluses that
farm income source, while in Lioma, 23%
can improve income and the ability to
of the households practice activities
purchase foods other than those
other than agriculture (Fig. 4.2).
produced within the household [69].
According to the results presented in
Rice, sorghum, maize and cassava were
Table 4.3, the mean age of household
selected in this study, as they are the
heads in Mepuagiua (37 years) was
most produced and consumed staple
higher than that of Lioma (33 years)
crops in the study region. Each staple
(p < 0.05); nevertheless, in Lioma,
crop represented a dummy variable that
farmers are better educated (p < 0.001).
equals 1 if the crop was produced by the
The mean farm sizes are not statistically
household and 0 otherwise. Quality of
different and overall crop diversity is
land is a dummy and represent the
higher in Lioma (p < 0.001). Moreover,
perception of having good land fertility. It
there are no statistically significant
is equal to 1 if the household assumes to
differences with regard to mean
have good land quality and is expected to
household size between both study
be positively related to food security. The
areas. In Mepuaguia, there is more
definition of “good” is based on the
diverse staple foods production with
individual farmers’ assessment of their
dominance of sorghum and very little
own land. Season is an important factor
cash crop production (Fig. 4.3).
for food security in Mozambique [4],
thus, we expect an increase of food
security during the harvest season.
 Results 48

100 Lioma Mepuagiua

90

80
Percentage of households (%)

70

60

50

40

30

20

10

0
% of female-headed households % of households with good land quality % of households with off-farm income

Figure 4.2: Percentage (%) of households for: gender; non-farm income; and good land
quality

Table 4.3: Socioeconomic characteristics of the households in each village

Mepuagiua (n=157) Lioma (n = 139) t-test
Variables
Average Min Max Sd Average Min Max Sd tCritical p(T<=t)

Age of household head 36.9 18 83 14.72 33.2 18 65 10.4 1.969 0.013**

Size of the household 4.89 1 11 2.12 5.2 1 13 2.01 1.968 0.122

Farm size (ha) 1.32 0.2 5 0.85 1.39 0.2 5.5 1.01 1.969 0.565

Livestock diversity 2.73 0 4 0.56 3.06 0 6 1.14 1.971 0.671

Education of household
4.71 0 10 2.57 5.9 0 16 2.54 1.968 <0.001***
head

Crop diversity 3.27 1 9 1.31 3.87 1 9 1.68 1.969 <0.001***

Significant levels: ***1%, **5%, *10%.

Results 49

Mepuagiua Lioma

Maize

Cash crops

Sunflower

Cowpeas

Onion

Rice

Beans

Sorghum

Mung bean

Groundnuts

Tomato

Cassava

Green leafy veg

Okra

Potato

Sweet Potato

Bambara Nuts

0 10 20 30 40 50 60 70 80 90 100
Percentage of households (%)

Figure 4.3: Percentage (%) of households producing specific crop in each village
Household food security status In Lioma, the majority (50%) present
high dietary diversity and only 5% fall
The household FS categories, classified
into low dietary diversity category. In
via different FS indicators, are shown in
Mepuagiua, households also increase
Table 4.4. The majority of households in
their dietary diversity during the harvest
Mepuagiua (52%) fall into the low
period, with 45% having medium diet
dietary diversity category when the
diversity and 37% having high diet
HDDS is applied in the pre-harvest
diversity.
period, whereas in Lioma, the majority
(68%) has a medium dietary diversity.
Only 2% and 1% of the households fall
within high dietary diversity category in
Mepuagiua and Lioma, respectively, and,
thus, can be assumed as food secure in
pre-harvest period. However, during the
harvest period, many households move
into the high dietary diversity category.
Results 50

Table 4.4: Percentage of households in each food security category

Mepuagiua (n=157) Lioma (n=139)
FS indicators Categories
Pre-harvest Harvest Pre-harvest Harvest
High 2% 37% 1% 50%
HDDS Medium 46% 45% 68% 45%
Low 52% 18% 31% 5%
Acceptable 0% 1% 0% 2%
HFC Borderline 11% 54% 12% 57%
Poor 89% 45% 88% 41%
Least FI 3% 27% 8% 17%
MAHFP Moderately FI 32% 31% 21% 41%
Most FI 65% 42% 71% 42%

Moreover, results reveal similar trends
between both study sites with regards to
HFCS. The majority of the households
(88% in Lioma and 89% in Mepuagiua)
had a poor HFCS (HFCS ≤ 21) in the pre-
harvest period. However, during the
harvest period, the majority of the
households are categorized
borderline. High food insecurity levels
were also found with respect to MAHFP
in the pre-harvest period. The majority of
the households (65% in Mepuagiua and
71% in Lioma) had less than 5 MAHFP,
while only 3% in Mepuagiua and 8% in
Lioma were least food insecure. In the
harvest period, the level of food insecure
households decreased to 42% in both
study sites.
Factors affecting household food
security
We used generalized ordered logit
models (GOLM) with partial proportional
odds (PPOM), fitted in Stata with a user-
written program, gologit2 [41] (see
Additional file 1 for detailed output
results). According to the PPOM results
presented in Table 4.5, the variable Age
violates the parallel line assumption in
the model fitted for the dependent
as variables HDDS and MAHFP, meaning
that their effect change across equations.
The size of household and season also
violates the parallel line assumption for
HDDS and MAHFP, respectively.
Therefore, age, size of household, and
season have gamma (𝛾) coefficients that
are statistically significant (p-value <
0.05). To obtain the coefficient for these
variables, 𝛾 coefficients must be added to
beta (𝛽) coefficients; for example, the
coefficient of season on the low and
medium dietary diversities versus high
dietary diversity is 4.451 (1.976 + 2.475).
The two alphas (𝛼) represent the
Results 51

intercepts of each cut-off points of the for at least one of the FS indicators used
logit function. All variables, except size of in this study (cf. Table 4.5). The marginal
household, sorghum production, and off- effects of each explanatory variable for
farm income, were statistically FS indicators are presented in Table 4.6.
significant at the 5% of significance level

Table 4.5: Estimation results of GOLM

Variables HDDS MAHFP HFCS

Beta
Season 1.976 (0.238)*** 1.291 (0.175)*** 3.433 (0.310)***
Gender -0.174 (0.210) -0.078 (0.204) -0.659 (0.286)**
Age -0.026 (0.009)*** -0.010 (0.008) -0.008 (0.010)
Geographic location 2.164 (0.433)*** 1.078 (0.429)** -0.015 (0.598)
Size of land -0.114 (0.102) 0.236 (0.099)** -0.134 (0.137)
Quality of land 0.788 (0.283)*** 1.090 (0.298)*** -0.235 (0.397)
Maize 0.391 (0.253) 0.315 (0.255) 0.907 (0.368)**
Sorghum 0.041 (0.278) -0.024 (0.276) 0.314 (0.385)
Cassava 1.076 (0.314)*** 0.689 (0.303)** 1.150 (0.431)***
Rice 0.726 (0.235)*** 0.551 (0.228)** 0.327 (0.305)
Size of household -0.077 (0.044)* -0.045 (0.046) -0.079 (0.060)
Off-farm income 0.089 (0.242) -0.070 (0.242) 0.603 (0.324)*
Livestock diversity 0.053 (0.016)*** 0.043 (0.014)*** -0.003 (0.018)
Education 0.066 (0.038)* 0.043 (0.037) 0.041 (0.052)
Cash crop 0.116 (0.254) 0.349 (0.259) 0.980 (0.350)***
Crop diversity 0.152 (0.085) * -0.204 (0.084) ** 0.854 (0.122)***
Alpha
_cons_1 -1.470 (0.594)** -1.8827 (0.579)*** -6.590 (0.887)***
_cons_2 -7.429785 (0.829) *** -5.148 (0.689) *** -13.082 (1.210)***
Gamma_2
Age 0.022 (0.011)** 0.020 (0.010)**
Size of household n.a 0.129 (0.048)***
Season 2.475 (0.572) *** n.a
n.a= not applicable; Standard Errors are presented in parenthesis; Significant levels: ***1%, **5%, *10
 Results 52

Table 4.6: Marginal effects of explanatory variables for FS level

HFCS Categories HDDS Categories MAHFP Categories

variables
Poor Borderline Acceptable Low Medium High Most FI Moderately FI Least FI
Season -0.570 0.568 0.002 (0.001) -0.323 -0.083 (0.044)* 0.406 -0.309 0.182 0.126
(0.038)*** (0.038)*** (0.035)*** (0.031)*** (0.039)*** (0.027)*** (0.020)***
Gender 0.126 -0.125 0.000 (0.000) 0.028 (0.033) -0.016 (0.018) -0.012 (0.015) 0.019 (0.051) -0.012 (0.031) -0.008 (0.020)
(0.059)** (0.059)**
Age 0.001 (0.002) -0.001 (0.002) 0.000 (0.000) 0.004 -0.004 -0.000 (0.000) 0.003 (0.002) -0.003 (0.002)* 0.001 (0.001)
(0.001)*** (0.001)***
Geographic 0.003 (0.105) -0.003 (0.105) 0.000 (0.000) -0.341 0.174 0.167 -0.261 0.154 0.107
Location (0.067)*** (0.049)*** (0.052)*** (0.099)*** (0.056)*** (0.046)**
Size of land 0.024 (0.024) -0.023 (0.024) 0.000 (0.000) 0.019 (0.017) -0.011 (0.010) -0.008 (0.007) -0.058 0.036 0.022
(0.024)** (0.015)** (0.010)**
Quality of 0.041 (0.068) -0.040 (0.067) 0.000 (0.000) -0.123 0.066 0.057 -0.265 0.150 0.037*** 0.116
land (0.042)*** (0.025)*** (0.026)** (0.069)*** (0.036)***
Maize -0.136 0.136 0.000 (0.000) -0.068 (0.047) 0.045 (0.034) 0.023 (0.015) -0.077 (0.061) 0.049 (0.040) 0.028 (0.021)
(0.047)*** (0.046)***
Sorghum -0.055 (0.068) 0.055 (0.067) 0.000 (0.000) -0.007 (0.045) 0.004 (0.027) 0.003 (0.018) 0.006 (0.068) -0.004 (0.042) -0.002 (0.026)
Cassava -0.207 0.207 0.001 (0.000) -0.170 0.094 0.077 -0.169 0.101 0.068
(0.079)*** (0.079)*** (0.049)*** (0.032)*** (0.029)*** (0.073)** (0.043)** (0.032)**
Rice -0.060 (0.057) 0.059 (0.057) 0.000 (0.000) -0.108 0.053 0.055 -0.136 0.079 0.058
(0.032)*** (0.019)*** (0.023)** (0.056)** (0.031)** (0.026)**
Size of 0.014 (0.011) -0.014 (0.011) 0.000 (0.000) 0.013 (0.007)* -0.008 (0.005) -0.005 (0.003) 0.011 (0.011) -0.019 0.008 (0.005)
household (0.009)**
Off-farm -0.117 (0.068)* 0.117 (0.068)* 0.000 (0.000) -0.014 (0.038) 0.008 (0.021) 0.006 (0.017) 0.017 (0.059) -0.011 (0.037) -0.007 (0.022)
income
Livestock 0.000 (0.003) 0.000 (0.003) 0.000 (0.000) -0.009 0.005 0.004 -0.011 0.007 0.004
diversity (0.003)*** (0.002)*** (0.001)*** (0.004)*** (0.002)*** (0.001)***
Education -0.007 (0.009) 0.007 (0.009) 0.000 (0.000) -0.011 (0.006)* 0.006 (0.004) 0.004 (0.003) -0.011 (0.009) 0.007 (0.006) 0.004 (0.004)
Cash crops -0.191 0.190 0.001 (0.000) -0.019 (0.040) 0.011 (0.023) 0.008 (0.018) -0.086 (0.064) 0.051 (0.037) 0.035 (0.028)
(0.074)** (0.073)**
Crop -0.150 0.149 0.000 (0.000) -0.025 (0.014)* 0.015 (0.009)* 0.010 (0.006)* 0.050 -0.031 -0.019
Diversity (0.022)*** (0.022)*** (0.021)** (0.013)** (0.008)**
Standard Errors are presented in parenthesis; Significant levels: ***1%, **5%, *10%
Results
Two variables are consistently
associated with all three FS indicators (p
value < 0.05), namely, the season and the
production of cassava. The three FS
indicators used in this study are higher
during the harvest season and for
households producing cassava, which
can also be seen from the negative
marginal effects on the lower categories
for all three indicators (cf. Table 4.6).
Geographic location, quality of land,
livestock diversity, and rice production
are statistically associated with two of
the FS indicators (p value < 0.05).
Farmers located in Lioma are more likely
to have a higher HDDS and MAHFP, as
shown by the negative marginal effects
for the lower categories and positive
marginal effects for the
categories. The higher the quality of land,
the higher the HDDS and MAHFP. The
production of rice also increases HDDS
and MAHFP. Households with higher
livestock diversity are more likely to
have higher HDDS and
Nevertheless, crop diversity
statistically and positively related to
MAHFP and HFCS.
Size of land is an important factor for
ensuring adequate food
throughout the year but does not have
any influence on HDDS and HFCS. Gender
is negatively associated with HFCS,
53
which is also represented by the positive
marginal effect for the lower category
(0.126) and negative for the borderline
(−0.125). This suggests that female-
headed households are more likely to
have a medium (or borderline) food
consumption score compared to male-
headed households. The age of the
household head only affects HDDS, while
the positive and negative signs of the
marginal effect for the low and medium
dietary diversity, respectively, indicate
that the older the head of the household,
the lower the dietary diversity. The
production of maize and cash crops only
have an effect on HFCS.
Discussion
This study assesses the FS situation
higher during the period immediately prior to
the harvest and during the harvest
period. It also deepens the analysis of the
key drivers conditioning household
capacity to achieve food security in
Gurué district. Three FS indicators
MAHFP. categorized into three levels were used
is to the capture multidimensionality and
the prevalence of more nuanced patterns
of FS among the households, which are
neglected in those studies that sort the
provision households into only two categories, e.g.
food secure and food insecure [42].
Generally, we observe comparably low
dietary diversity and predominantly
Results
poor household food consumption
during the pre-harvest period and a
predominance of households with
medium and high dietary diversity and
food consumption during the harvest
season. This is in line with our
expectation, as our data were collected
between February and March, the period
when food reserves are already scarce,
and during the harvest period (May and
June), when access to farm-sourced food
is high [43]. As stated by [4], in
Mozambique household, food security is
sensitive to seasonal variations.
Therefore, our results may represent
both the lower and upper margins of FS
in Gurué district.
It is noteworthy that gender is only
associated with HFCS. According to the
results presented in Table 4.6, female-
headed households have higher food
consumption scores. Due to the
weighting system applied by this FS
indicator, it responds more strongly to
animal products, which are generally
purchase in the study area. HDDS
weights all food groups equally, whereas
HFCS applies different weights to the
different food groups [44]. Fish, which is
mainly purchased in our study area, have
the highest weight (4), whereas the farm-
sourced vegetables and main staple, have
a low weight (cf. Table 4.1). Thus, using
54
HFCS may result in the highest food
insecurity levels for household relying on
farm-sourced food items. Although, in
general, female-headed households are
often behind male-headed households
with regard to FS status [42, 45],
behavioral differences between women
and men result in differences with
respect to how financial resources are
spent in male and female-headed
households [46]. While female decision-
makers may choose to invest their
financial resources in food expenditures,
male decision-makers may spend their
financial resources on non-food items. In
line with this, studies also find that
low-income women with children are
motivated to improve the nutritional
quality of their families and are more
likely to consume a nutritious diet [47,
48].
Surprisingly, livestock diversity is
statistically significant only for HDDS and
MAHFP. Since HFCS weighs food groups
differently, we expected households
engaged in animal husbandry to have
higher HFCS. However, this may imply
that livestock is not used for own
consumption, but rather for commercial
and other purposes. Moreover, off-farm
income did not have a statistically
significant impact on all FS indicators
(Table 4.6), although it is often reported
Results 55

as one of the most important drivers of Although aged farmers are, in general,
FS [49]. Thus, the results of our study more experienced and resource
may be in line with [50], who suggested endowed as compared to younger
that in many regions, poor households farmers [55], they may also be more
tend to spend their income on non-food likely to work fewer hours per day during
items such as clothing, household the harvest season compared to younger
appliances, etc., and not necessarily on household heads, therefore, losing their
more nutritious foods. Notably, the ability to diversify the diet [53].
production of cash crops has a positive The geographic location of the household
influence on HFCS. This may suggest that is a statistically significant factor of food
cash crops are important sources of security. Households from Lioma are
income to purchase non-farm sourced more food secure according to MAHFP
foods, especially those with the highest and HDDS than those from Mepuagiua.
weights (cf. Table 4.1). According to [51], This is partially explained by the higher
households with cash crops income can crop diversity in Lioma than Mepuagiua,
purchase more appropriate and
likely contributing to greater dietary
nutritious foods, thus being more likely diversity. The results indicate that crop
to have improved food security. diversity is one of the main drivers of
Nevertheless, the fact that the production food security, according to MAHFP and
of cash crops has a positive influence on HFCS. In line with this, many existing
HFCS while off-farm income does not, studies demonstrate the link between
reinforces the importance of crop diversity and FS [56, 57]. Another
psychological and behavioral factors on possible explanation is that the
household income expenditure and food percentage of households that claim to
consumption patterns [52]. Perhaps the have “good land quality” is higher in
source of income plays a role in the Lioma (86%) than in Mepuagiua (74%),
decision whether to invest in food or not. even though the local definition of “good
The effect of age is statistically significant land quality” might differ. Studies
for HDDS, suggesting that older farmers demonstrate that this factor significantly
have lower dietary diversity than affects agricultural productivity [58, 59]
younger farmers. This negative and is positively associated with MAHFP
relationship between age and HDDS is and HDDS (cf. Table 4.5).
also found in other studies [53, 54].
Results
The size of land only drives MAHFP.
According to [60] and [61], households
with larger farm size can have a
comparably higher level of production
diversify and produce comparably high
quantities of food, therefore having a
higher probability of being food secure.
However, in our study area, a large farm
size does not necessarily result in a
higher dietary diversity. Rather, it is an
essential mechanism to ensure that food
is quantitatively available in most
months of the year.
Our analysis further indicates that the
production of the staple food crops,
except sorghum, is significantly related
to at least one of FS indicators. Sorghum
is usually less productive and less
marketable than other staples [62], thus
it does not have a significant impact on
FS. Rice, maize, and cassava are the most
consumed crops in Mozambique [63].
Cassava ranks first in terms of average
daily caloric intake by households in
Mozambique (678 kilocalories per capita
per day), followed by maize (478 kcal per
capita per day) and rice (166 kcal per
capita per day) [64]. According to our
results, maize has a positive relationship
only with HFCS, rice is associated to
MAHFP and HDDS, while cassava is
positively related to all three indicators.
This might indicate that, although all food
56
crops play an important role in ensuring
at least one dimension of household food
security, cassava is used as both a
subsistence crop (ensuring household
access to calories) and a market crop
(allowing households to sell surplus and
re-invest into other food groups). In
addition, cassava is a drought-tolerant
crop and has a low-cost vegetative
propagation. Thus, it might improve
households’ ability to absorb weather-
induced failures in the production of
other staple crops. Maize, however, is a
seasonal crop, prone to weather shocks
such as droughts and floods [65].
Moreover, the simple in-ground storage
and perennial nature of cassava offers a
flexible harvesting calendar that enables
farmers to adjust harvested quantities
throughout the year [66]. Likewise, local
rice is resistant to floods [67] and usually
goes through manual processing [68],
thus it remains stored for a slightly
longer period, which increases its
availability to the households over the
year but is less likely to simultaneously
satisfy the households needs and
produce marketable surplus. Although
the staple food crops are mostly grown
for household consumption [40], maize
and cassava are positively related to
HFCS. This is because they are important
sources of marketable surpluses that can
improve income and the ability to re-
Results
invest into other food groups [69]. In fact,
more than 50% of households sell at least
a part of the maize they produce. In
general, however, in our study region,
farmers consume approximately 60%
and sell 40% of their agricultural
products to cover non-food household
expenses [62].
The percentage of food secure
households differs depending on the
individual indicator applied (Table 4.4).
This underlines that each indicator
reflects the eventual output of the
different FS dimensions (especially
availability, access, and stability) [27].
Maxwell et al. [28] likewise find that
different food security measures can
produce divergent results. Although,
HFCS and HDDS share a common
emphasis on dietary diversity as a proxy
for household food access and are both
correlated with total per capita food and
non-food expenditures, they are not
interchangeable [44]. Unlike HDDS, the
weights applied by HFCS do not reflect
per capita calories intake but rather
reflect the quality of the diet [70], since
including animal source foods, fruits,
vegetables, and dairy products in the
diets increases the intake of essential
micronutrients [71]. This is an important
aspect as programs to overcome FI may
follow different agricultural production
57
strategies, depending on the individual
food security indicator applied or
weighted most. Policies aiming to raise,
for example, diet quality would promote
more market-oriented agriculture with
the aim of improving HFCS, while an
increase of calories intake may be
achieved when production for own
consumption is promoted. Hence, our
findings support the idea of a fair balance
between the production of cash crops
and crops for own consumption [72].
MAHFP has a recall period of 12 months
and, thus, it is a useful tool to capture
food stability [73], as it shows
households’ ability to address
vulnerability over the year.
Food utilization is an essential dimension
of FS that encompasses the nutritional
quality of food within households and the
bioavailability of nutrients in those
foods. The traditional proxy measure of
food utilization is the use anthropometric
measurements, e.g. nutritional status, to
understand whether food is allocated
equally to all individual household
members [35]. Nevertheless, our study is
conducted at household level, thus
unable to capture this dimension.
Therefore, it is essential for future
research to combine the self-reported
consumption patterns with some FS
outcomes, i.e., anthropometric measures,
Results 58

to understand whether food is allocated strongly to purchased food. As such,

equally to all individual household combining different food security
members. This generates more complete indicators is an important strategy to
information for policy makers and holistically assess the food security
development agencies. The short recall status of the local population. In our
period applied by HDDS and HFCS (7 analysis, we find that gender and age of
days) represents another limitation of the household head, geographic location,
this study, though these indicator are size and quality of land, staples
very useful and sensitive enough to show production (especially cassava),
dietary diversity at the household level livestock, crop diversity and cash crops
[27]. are important drivers of at least one of FS
indicator. Thus, one central
Conclusions
recommendation derived from our
In this study, we investigate the FS
results is that relevant stakeholders can
situation and its drivers at the household
improve FS by promoting a balance
level in Gurué District, Zambézia
between market-oriented agriculture
province, central Mozambique by
and subsistence production, as well as by
applying different indicators (HDDS,
supporting farmers whose land is of low
HFCS, MAHFP). In general, we find very
quality, e.g., by teaching composting
critical levels of food insecurity among
techniques to enhance agricultural
households during the pre-harvest
productivity. To reduce the seasonality of
season and relatively high levels of food
food security, policy-makers and
security during the harvest season.
relevant agencies must direct their
However, varying outcomes are found
efforts at increasing access to irrigation
depending on the indicator used to
technologies to offset the
assess FS. Thus, the decision whether
drought-related shocks that occur during
farmers should rely on staple foods
the dry season.
production for increasing their FS status,
Declarations
or specialize in cash crop production to
generate income and buy food items Ethics approval and consent to
depends on the indicator used to assess participate: Informed consent was
FS. For instance, MAHFP and HDDS are obtained from all individual participants
more related to the consumption of farm- in the study.
sourced food, while HFCS responds more Consent for publication: Not applicable.
Results
Availability of data and material: All
data generated and analyzed during this
study are included in Additional file 2.
Competing interests: The authors
declare that they have no competing
interests.
Funding: The project is supported by
funds of the Federal Ministry of Food and
Agriculture (BMEL) based on a decision
of the Parliament of the Federal Republic
of Germany via the Federal Office for
Agriculture and Food (BLE).
Author contributions:
Conceptualization: Custodio Matavel,
Harry Hoffmann, and Constance Rybak;
Methodology: Custodio Matavel, Harry
Hoffmann, Constance Rybak, and
Jonathan Steinke; Formal analysis and
investigation Custodio Matavel; Writing -
original draft preparation: Custodio
Matavel and Harry Hoffmann; Writing -
review and editing: Jonathan Steinke,
Harry Hoffmann, Constance Rybak,
Stefan Sieber, and Klaus Müller ; Funding
acquisition: Harry Hoffmann, Constance
Rybak, and Stefan Sieber; Supervision:
Stefan Sieber, Klaus Müller, and Harry
Hoffmann.
Acknowledgments: Sincere thanks to
the Leibniz Centre for Agricultural
Landscape Research (ZALF)
Müncheberg, Germany, for their
59
logistical support in conducting the
research in Gurué district, Central
Mozambique. Further, we thank the
farmers who participated in the survey.
References
1. FAO, ECA, AUC. Africa Regional
Overview of Food Security and Nutrition
2019. Accra, Ghana: FAO; 2020.
2. WFP. WFP Mozambique: Country Brief
September2019. WFP Mozambique
Country BriefSeptember2019; 2019.
3. USAID, WFP. USAID Mozambique:
World Food Program Fact Sheet,
February 2020. USAID; 2020.
4. Selvester K, Fidalgo L, Ballard T,
Kennedy G, Dop M, Mistura L, et al.
Report on use of the household food
insecurity access scale and household
dietary diversity score in two survey
rounds in Manica and Sofala Provinces,
Mozambique. FAO Project. 2008:1-23.
5. Rose ES, Blevins M, González-Calvo L,
Ndatimana E, Green AF, Lopez M, et al.
Determinants of undernutrition among
children aged 6 to 59 months in rural
Zambézia Province, Mozambique: results
of two population-based serial cross-
sectional surveys. BMC Nutrition.
2015;1(1):41.
6. UNICEF. Nutrition situation in
in
Mozambique 2016
Results 60

https://www.unicef.org/mozambique/e 13. Hadley C, Stevenson EGJ, Tadesse Y,

n/nutrition. Accessed 19/05/2020. Belachew T. Rapidly rising food prices
and the experience of food insecurity in
7. Artur L, Hilhorst D. Everyday realities
urban Ethiopia: Impacts on health and
of climate change adaptation in
well-being. Social Science & Medicine.
Mozambique. Global Environmental
2012;75(12):2412-9.
Change. 2012;22(2):529-36.
14. Maitra C, Sethi V, Unisa S, Shankar S.
8. Adhikari U, Nejadhashemi AP,
Household Food Insecurity and Maternal
Woznicki SA. Climate change and eastern
and Child Nutritional Status: Evidence
Africa: a review of impact on major crops.
from Maharashtra. Review of Income and
Food and Energy Security.
Wealth. 2019;65(S1):S63-S101.
2015;4(2):110-32.
15. Diamond KK, Stebleton MJ, delMas
9. USAID. Food Assistance Fact Sheet -
RC. Exploring the Relationship Between
Mozambique - June 12, 2019. USAID;
Food Insecurity and Mental Health in an
2019.
Undergraduate Student Population.
10. Bey A, Jetimane J, Lisboa SN, Ribeiro
Journal of Student Affairs Research and
N, Sitoe A, Meyfroidt P. Mapping
Practice. 2019:1-15.
smallholder and large-scale cropland
16. Whittle HJ, Sheira LA, Wolfe WR,
dynamics with a flexible classification
Frongillo EA, Palar K, Merenstein D, et al.
system and pixel-based composites in an
Food insecurity is associated with
emerging frontier of Mozambique.
anxiety, stress, and symptoms of
Remote Sensing of Environment.
posttraumatic stress disorder in a cohort
2020;239:111611.
of women with or at risk of HIV in the
11. Barrett CB. Actions now can curb food
United States. The Journal of Nutrition.
systems fallout from COVID-19. Nature
2019;149(8):1393-403.
Food. 2020.
17. Masa R, Graham L, Khan Z, Chowa G,
12. García Cruz LM, González Azpeitia G,
Patel L. Food insecurity, sexual risk
Reyes Súarez D, Santana Rodríguez A,
taking, and sexual victimization in
Loro Ferrer JF, Serra-Majem L. Factors
Ghanaian adolescents and young South
associated with stunting among children
African adults. International Journal of
aged 0 to 59 months from the central
Public Health. 2019;64(2):153-63.
region of Mozambique. Nutrients.
2017;9(5):491.
Results
18. Cady CL. Food Insecurity as a Student
Issue. Journal of College and Character.
2014;15(4):265-72.
19. Maitra C. A review of studies
examining the link between food
insecurity and malnutritio. Rome: FAO;
2018.
20. UN. The 17 Sustainable
Development Goals: United Nations;
2015 https://sdgs.un.org/goals.
Accessed; cited 2020 09/09.
21. Haysom G, Tawodzera
“Measurement drives diagnosis and
response”: Gaps in transferring food
security assessment to the urban scale.
Food Policy. 2018;74:117-25.
22. Cafiero C, Melgar-Quiñonez HR,
Ballard TJ, Kepple AW. Validity and
reliability of food security measures.
Annals of the New York Academy of
Sciences. 2014;1331(1):230-48.
23. FAO. World food summit plan of
action. Rome: FAO; 1996.
24. Gross R, Schoeneberger H, Pfeifer H,
Preuss H-J. The four dimensions of food
and nutrition security: definitions and
concepts. 2000. Contract No.: 20.
25. Hoddinott J. Choosing outcome
indicators of household food security:
Citeseer; 1999.
61
26. Santeramo FG. On the Composite
Indicators for Food Security: Decisions
Matter! Food Reviews International.
2015;31(1):63-73.
27. Hendriks SL, van der Merwe C, Ngidi
MS, Manyamba C, Mbele M, McIntyre AM,
et al. What are we measuring?
Comparison of household food security
indicators in the Eastern Cape Province,
South Africa. Ecol Food Nutr.
2016;55(2):141-62.
28. Maxwell D, Vaitla B, Coates J. How do
G.
indicators of household food insecurity
measure up? An empirical comparison
from Ethiopia. Food Policy. 2014;47:107-
16.
29. Pitoro R, Chagomoka T. Food Security
Dynamics and its Drivers in Rural
Mozambique. International Journal of
Sciences. 2017;6(05):55-65.
30. Garrett JL, Ruel MT. Are Determinants
of Rural and Urban Food Security and
Nutritional Status Different? Some
Insights from Mozambique. World
Development. 1999;27(11):1955-75.
31. Nyyssölä M, Pirttilä J, Sandström S.
Helping poor farmers to help themselves:
Evidence from a group-based aid project
in Mozambique. WIDER Working Paper;
2012. Report No.: 9292305522.
32. Mabiso A, Cunguara B, Benfica R.
Food (In)security and its drivers: insights
Results 62

from trends and opportunities in rural Regression Model. Land.

Mozambique. Food Security. 2021;10(3):323.
2014;6(5):649-70. 39. INE. Folheto estatistico distrital
33. McCordic C, Abrahamo E. Family gurue 2018. Quelimane: Instituto
structure and severe food insecurity in Nacional de Estatística; 2018.
Maputo and Matola, Mozambique. 40. Soares MG. Relação entre as
Sustainability. 2019;11(1):267. mudanças de uso e cobertura de terra e
34. Riley L, Caesar M. Urban household as queimadas em florestas de Miombo,
food security in China and Mozambique: Gurué, Moçambique: Universidade
a gender-based comparative approach. Eduardo Mondlane; 2017.
Development in Practice. 41. Williams R. Generalized Ordered
2018;28(8):1012-21.
Logit/Partial Proportional Odds Models
35. Jones AD, Ngure FM, Pelto G, Young for Ordinal Dependent Variables. The
SL. What Are We Assessing When We Stata Journal. 2006;6(1):58-82.
Measure Food Security? A Compendium 42. Gebre GG, Isoda H, Amekawa Y, Rahut
and Review of Current Metrics. Advances DB, Nomura H, Watanabe T. What
in Nutrition. 2013;4(5):481-505. Explains Gender Gaps in Household Food
36. Carletto C, Zezza A, Banerjee R. Security? Evidence from Maize Farm
Towards better measurement of Households in Southern Ethiopia. Social
household food security: Harmonizing Indicators Research. 2021;155(1):281-
indicators and the role of household 314.
surveys. Global Food Security. 43. Matavel C, Hoffmann H, Rybak C,
2013;2(1):30-40. Sieber S. Can Subsistence Farming Help
37. Mawoko Z, Hendriks S, Reys A. The to Achieve Household Food Security?
influence of large-scale agricultural Evidence from Gurue, Central
investments on household food security Mozambique. TROPENTAG;
in the Gurue and Monapo districts of Germany2020.
Mozambique. 2018. 44. Kennedy G, Berardo A, Papavero C,
38. Fitawek W, Hendriks SL. Evaluating Horjus P, Ballard T, Dop M, et al. Proxy
the Impact of Large-Scale Agricultural measures of household food
Investments on Household Food Security consumption for food security
Using an Endogenous Switching assessment and surveillance:
Results 63

comparison of the household dietary availability in sub-Saharan Africa based

diversity and food consumption scores. on big data from small farms.
Public Health Nutrition. Proceedings of the National Academy of
2010;13(12):2010-8. Sciences. 2016;113(2):458.

45. Broussard NH. What explains gender 50. Banerjee AV, Duflo E. Poor
differences in food insecurity? Food economics: A radical rethinking of the
Policy. 2019;83:180-94. way to fight global poverty: Public
Affairs; 2011.
46. Carranza M, Niles MT. Smallholder
Farmers Spend Credit Primarily on Food: 51. Mango N, Zamasiya B, Makate C,
Gender Differences and Food Security Nyikahadzoi K, Siziba S. Factors
Implications in a Changing Climate. influencing household food security
Frontiers in Sustainable Food Systems. among smallholder farmers in the Mudzi
2019;3(56). district of Zimbabwe. Development
Southern Africa. 2014;31(4):625-40.
47. Evans A, Chow S, Jennings R, Dave J,
Scoblick K, Sterba KR, et al. Traditional 52. Silva A, Caro JC, Magaña-Lemus D.
Foods and Practices of Spanish-Speaking Household food security: Perceptions,
Latina Mothers Influence the Home Food behavior and nutritional quality of food
Environment: Implications for Future purchases. Journal of Economic
Interventions. Journal of the American Psychology. 2016;55:139-48.
Dietetic Association. 2011;111(7):1031- 53. Huluka AT, Wondimagegnhu BA.
8. Determinants of household dietary
48. Dubowitz T, Acevedo-Garcia D, diversity in the Yayo biosphere reserve
Salkeld J, Cristina Lindsay A, of Ethiopia: An empirical analysis using
Subramanian SV, Peterson KE. sustainable livelihood framework.
Lifecourse, immigrant status and Cogent Food & Agriculture.
acculturation in food purchasing and 2019;5(1):1690829.
preparation among low-income mothers. 54. Jones AD, Shrinivas A, Bezner-Kerr R.
Public Health Nutrition. 2007;10(4):396- Farm production diversity is associated
404. with greater household dietary diversity
49. Frelat R, Lopez-Ridaura S, Giller KE, in Malawi: findings from nationally
Herrero M, Douxchamps S, Djurfeldt AA, representative data. Food Policy.
et al. Drivers of household food 2014;46:1-12.
Results 64

55. Etwire PM, Martey E, Dogbe W. 62. Khatiwada L, Mussagy IH, dos Anjos
Technical Efficiency of Soybean Farms Luis A, Siniquinha A, Biedronski J.
and Its Determinants in Saboba and Zambézia Market Analysis Report.
Chereponi Districts of Northern Ghana: A Washington, DC: Research Technical
Stochastic Frontier Approach. 2013. Assistance Center; 2021.

56. Rajendran S, Afari-Sefa V, Shee A, 63. Popat M, Tostão E, Fontes F,

Bocher T, Bekunda M, Lukumay PJ. Does Vilanculos OC. Monitoring Price
crop diversity contribute to dietary Incentives for Rice in Mozambique. 2017.
diversity? Evidence from integration of 64. NET F. Mozambique Staple Food
vegetables into maize-based farming Market Fundamentals, September 2018.
systems. Agriculture & Food Security. USAID 2019.
2017;6(1):50.
65. Salazar C, Ayalew H, Fisker P.
57. Snapp SS, Fisher M. “Filling the maize Weather Shocks and Spatial Market
basket” supports crop diversity and Efficiency: Evidence from Mozambique.
quality of household diet in Malawi. Food The Journal of Development Studies.
Security. 2015;7(1):83-96. 2019;55(9):1967-82.
58. Wiebe KD. Linking land quality, 66. Haggblade S, Andersson Djurfeldt A,
agricultural productivity, and food Banda Nyirenda D, Bergman Lodin J,
security. USDA-ERS Agricultural Brimer L, Chiona M, et al. Cassava
Economic Report. 2003(823). commercialization in Southeastern
59. Sanchez PA. Soil Fertility and Hunger Africa. Journal of Agribusiness in
in Africa. Science. 2002;295(5562):2019- Developing and Emerging Economies.
20. 2012;2(1):4-40.

60. Van der Veen A, Gebrehiwot T. Effect 67. Salazar-Espinoza C, Jones S, Tarp F.
of policy interventions on food security Weather shocks and cropland decisions
in Tigray, Northern Ethiopia. Ecology and in rural Mozambique. Food Policy.
Society. 2011;16(1). 2015;53:9-21.

61. Aidoo R, Mensah JO, Tuffour T. 68. Vellema S, Beekman W. Rice farmers,
Determinants of household food security local markets and rice trade: A case study
in the Sekyere-Afram plains district of of local market dynamics in Mozambique
Ghana. European Scientific Journal. and its implications for a cooperative
2013;9(21).
Results 65

model. Wageningen International/DGIS; 75. Leroy JL, Ruel M, Frongillo EA, Harris
2011. J, Ballard TJ. Measuring the Food Access
Dimension of Food Security:A Critical
69. Mango N, Makate C, Mapemba L, Sopo
Review and Mapping of Indicators. Food
M. The role of crop diversification in
and Nutrition Bulletin. 2015;36(2):167-
improving household food security in
95.
central Malawi. Agriculture & Food
Security. 2018;7(1):7. 76. Swindale A, Bilinsky P. Household
dietary diversity score (HDDS) for
70. Wiesmann D, Bassett L, Benson T,
measurement of household food access:
Hoddinott J. Validation of the world food
indicator guide. Washington, DC: Food
programme s food consumption score
and Nutrition Technical Assistance
and alternative indicators of household
Project, Academy for Educational
food security: Intl Food Policy Res Inst;
Development. 2006.
2009.
77. Hoddinott J, Yohannes Y. Dietary
71. Wiesmann D. A global hunger index:
diversity as a food security indicator.
Measurement concept, ranking of
2002.
countries, and trends: Intl Food Policy
Res Inst; 2006. 78. Kennedy G, Ballard T, Dop M.
Guidelines for measuring household and
72. Ntakyo PR, van den Berg M. Effect of
individual dietary diversity. Rome: FAO;
market production on rural household
2010.
food consumption: evidence from
Uganda. Food Security. 79. WFP. Food consumption analysis:
2019;11(5):1051-70. Calculation and use of the food
consumption score in food security
73. Bilinsky P, Swindale A. Months of
analysis. World Food Programme; 2008.
adequate household food provisioning
(MAHFP) for measurement of household 80. Sakai T, editor Two sample t-tests for
food access: indicator guide: Food and ir evaluation: Student or welch?
Nutritional Technical Assistance Project, Proceedings of the 39th International
Academy for Educational …; 2007. ACM SIGIR conference on Research and
Development in Information Retrieval;
74. Cochran WG. Sampling Technique.
2016.
2nd Edition ed. New York: John Wiley
and Sons Inc.; 1963. 81. Walker SH, Duncan DB. Estimation of
the probability of an event as a function
Results 66

of several independent variables. motorway accident severity in England

Biometrika. 1967;54(1):167-79. using the generalised ordered logistic
regression model. Journal of Safety
82. Wang J, Wang Y, Peng Y, Lu JJ.
Research. 2015;55:89-97.
Examining partial proportional odds
model in analyzing severity of high- 88. Bogale A, Shimelis A. Household level
speed railway accident. Smart and determinants of food insecurity in rural
Resilient Transport. 2021;3(1):12-24. areas of Dire Dawa, Eastern Ethiopia.
African Journal of Food, Agriculture,
83. Brant R. Assessing proportionality in
Nutrition and Development. 2009;9(9).
the proportional odds model for ordinal
logistic regression. Biometrics. 89. Tauer LW, editor Farmer productivity
1990;46(4):1171-8. by age over eight US census years.
International Farm Management
84. Akbar M, Niaz R, Amjad M.
Association Conference; 2017.
Determinants of households’ food
insecurity with severity dimensions in 90. Asfaw A, Admassie A. The role of
Pakistan: Varying estimates using partial education on the adoption of chemical
proportional odds model. Health & Social fertiliser under different socioeconomic
Care in the Community. environments in Ethiopia. Agricultural
2020;28(5):1698-709. economics. 2004;30(3):215-28.

85. Ayele AW, Kassa M, Fentahun Y,
Edmealem H. Prevalence and associated
factors for rural households
insecurity in selected districts of east
Gojjam zone, northern Ethiopia: cross-
sectional study. BMC Public Health.
2020;20(1):202.
86. Peterson B, Harrell FE. Partial
Proportional Odds Models for Ordinal
Response Variables. Journal of the Royal
Statistical Society Series C (Applied
Statistics). 1990;39(2):205-17.
87. Michalaki P, Quddus MA, Pitfield D,
Huetson A. Exploring the factors affecting
91. Weir S, Knight J. Externality effects of
education: dynamics of the adoption and
food diffusion of an innovation in rural
Ethiopia. Economic development and
cultural change. 2004;53(1):93-113.
92. Fisher M, Lewin PA. Household,
community, and policy determinants of
food insecurity in rural Malawi.
Development Southern Africa.
2013;30(4-05):451-67.
93. Idrisa Y, Gwary M, Shehu H. Analysis
of food security status among farming
households in Jere Local Government of
Results 67

Borno State, Nigeria. Agro-Science. 96. Kilic T, Winters P, Carletto C. Gender

2008;7(3). and agriculture in sub-Saharan Africa:
introduction to the special issue.
94. Makombe T, Lewin P, Fisher M. The
Agricultural Economics.
determinants of food insecurity in rural
2015;46(3):281-4.
Malawi: Implications for agricultural
policy. International Food Policy 97. Kuma T, Dereje M, Hirvonen K,
Research Institute (IFPRI); 2010. Minten B. Cash crops and food security:
Evidence from Ethiopian smallholder
95. Croppenstedt A, Goldstein M, Rosas N.
coffee producers. The Journal of
Gender and agriculture: Inefficiencies,
Development Studies. 2019;55(6):1267-
segregation, and low productivity traps.
84.
The World Bank Research Observer.
2013;28(1):79-109. 98. FAO. The Special Programme for
Food Security. FAO; 1996.
Results 68

4.2 Experimental evaluation of a passive indirect solar dryer for agricultural

products in Central Mozambique

Journal of Food Processing and Preservation (2021). Volume 45, Issue 11.

Custodio E. Matavel1,2; Harry Hoffmann1; Constance Rybak1; Johannes M. Hafner1,2; João

Salavessa3; Shibire Bekele Eshetu1; Stefan Sieber1,2

1Leibniz Centre for Agricultural Landscape Research (ZALF), Müncheberg, Germany

2Department of Agricultural Economics, Faculty of Life Sciences Thaer-Institute, Humboldt-Universität

zu Berlin, Berlin, Germany

3Faculty of Health Sciences, Universidade Lúrio, Nampula, Mozambique

© 2021 The Authors. This is an open access article under the terms of the Creative
Commons Attribution NonCommercial License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited and is not used
for commercial purposes.

https://doi.org/10.1111/jfpp.15975

Abstract

Post-harvest losses are one of the major livelihood challenges for farmers in the Global
South. The use of drying technologies to preserve agricultural products has been
promoted as a particular solution to address this challenge. In this regard, we designed
and tested a passive indirect solar dryer for drying agricultural products as an alternative
to open sun drying (OSD) in Gurue district, Central Mozambique, using amaranth leaves
and maize. In addition, a sensorial analysis was conducted by randomly selecting a group
of 60 adults who evaluated the texture, aroma and color of dried amaranth and maize
grains. Compared to OSD, the passive indirect solar dryer reduced drying time and
increased the thermal efficiency. Evaluation of sensory quality attributes showed that
passive indirect solar drying outperforms OSD.
Results 69

Practical applications

This study evaluates the performance of a passive indirect solar dryer, a sustainable
alternative to conventional food preservation technologies (e.g., refrigeration) that are
not affordable to resource-constrained communities. The use of passive solar dryers, if
carried out correctly, creates the possibility for poor rural households to safely store and
increase shelf life of food. In addition, the acceptability of products dried in the passive
indirect solar dryer is evaluated. Thus, the study also provides insights on passive solar
dryer potential for preserving the quality of the final product.

Introduction faces major challenges due to climate

change. Moreover, reliance on traditional
Post-harvest losses (PHL) are significant
food processing and preservation
threats to food security in the global
techniques exacerbates their
South (Affognon et al., 2015; Hodges et
vulnerability to food insecurity (FI)
al., 2011), and their reduction is an
(Adeyeye, 2017). As reported by FAO, in
important factor in achieving the
2016, about 14% of food produced in SSA
sustainable development goals (SDG),
was lost from post-harvest to
particularly, SDG 2 (Zero Hunger) and
distribution. In addition, SSA has the
SDG 12 (ensure sustainable consumption
highest caloric losses globally due to PHL
and production patterns) (FAO, 2019). In
and presents the highest levels of
addition, the African Union has
malnutrition (FAO, 2019; FAO et al.,
committed to halving PHL by 2025 under
2020), which is likely to worsen for
the Malabo Declaration (AUC, 2014).
several reasons, including the growing
These targets are particularly important
food needs of the increasing population
to smallholder farmers in Sub-Saharan
(Gerland et al., 2014; Valin et al., 2014),
Africa (SSA) as they are among the
poor dietary habits (Afshin et al., 2019),
poorest populations in the world (WB,
potential impacts of Covid-19 (Saccone,
2016) and rely largely on food
2021), and challenging global economic
production for their livelihoods (Sheahan
conditions (FAO & ECA, 2018).
& Barrett, 2017). Indeed, more than 80%
Therefore, finding solutions to minimize
of rural households in many SSA
PHL is crucial to address FI in SSA.
countries depend to some extent on
agriculture (Davis et al., 2017), even A particular solution to this challenge is
though agricultural production in SSA the adoption and application of more
Results
efficient drying methods for the
preservation of agricultural products.
Drying—defined as the removal of
moisture from food to prevent the
growth of molds, yeasts, fungi, and
bacteria—can be accomplished using
various energy sources such as fossil
fuels, electricity, natural gas, biomass,
and solar energy (Lingayat,
Chandramohan, Raju, et al., 2020;
Prakash & Kumar, 2017). However,
energy sources such as fossil fuels,
electricity, and natural gas are in
generally comparatively expensive, and
their access is often unreliable for rural
households in developing countries.
Therefore, given the abundant and freely
available solar radiation in tropics and
sub-tropics (Mustapha et al., 2014;
Mustayen et al., 2014), the use of solar
energy, one of the oldest methods of food
preservation (Janjai & Bala, 2012), takes
a leading role in the sustainable drying of
agricultural products for smallholder
farmers in the Global South (Esper &
Mühlbauer, 1998; Lamidi et al., 2019).
Smallholder farmers in SSA use solar
energy to dry their produce after harvest
as an attempt to ensure safe storage of
food until the next harvest (Stathers et
al., 2013). However, the use of traditional
open solar drying (OSD) to dry
agricultural produce (Karekezi, 2002;
70
Ndukwu et al., 2018) results in
significant losses such as contamination
from dust, rain, spoilage, insects and
other pests (Affognon et al., 2015;
Kaminski & Christiaensen, 2014; Kumar
et al., 2016; Udomkun et al., 2020).
Therefore, the advances in solar energy
research and especially the
corresponding applications in
agriculture have led to the design and
development of a variety of solar dryers
to overcome the limitations of OSD
(Mustayen et al., 2014; Ssemwanga et al.,
2020).
Solar dryers need to supply more heat to
the product than is available under
ambient conditions to promote
evaporation of moisture from inside the
crop (Kalogirou, 2014). They are
generally classified into direct, indirect
or mixed mode, based on how heat is
transferred to the food and passive or
active, based on differences in the
circulation of air used for drying (Chavan
et al., 2020; Fudholi et al., 2010; Mohana
et al., 2020; Visavale, 2012). In addition,
solar dryers can also be classified as
hybrid solar dryers, if they are designed
with an additional heat sources—such as
solar assisted auxiliary thermal storage
system, wood, gas or electricity—to
enable a higher crop drying rate and
consequently a higher product quality
Results
(Khaing Hnin et al., 2019; Reyes et al.,
2014; Udomkun et al., 2020).
In direct solar dryers (DSD), food is
directly exposed to solar radiation that
penetrates through a transparent cover
(Islam et al., 2018; Kumar et al., 2016).
This cover reduces heat losses, and
minimizes the product contamination by
rain, dust and insects (Sandali et al.,
2019). However, sensory properties,
such as color and certain vitamins, may
be affected by direct sunlight (Al-Juamily
et al., 2007; Mustayen et al., 2014). The
use of indirect type solar dryers (ISD)
leads to improved product quality
compared to DSD (Mohana et al., 2020)
as they minimize color changes and loss
of specific vitamins (Tomar et al., 2017).
They consist of a separate solar air heater
in which solar energy is collected. The
heated air circulates through trays in a
drying chamber where the agricultural
products are placed (El-Sebaii et al.,
2002). An intermediate solution are
mixed solar dryers (MSD), where the
product is heated by both transparent
drying chamber and also separate air
heater (Shalaby et al., 2014).
In passive solar dryers (PSD), the heated
air is circulated through the food
products by buoyancy forces or as a
result of pressure differences (Basunia &
Abe, 2001; Ekechukwu & Norton, 1999).
71
They are completely dependent on solar
energy, while in active solar dryers
(ASD), the circulation of heated air is
done externally with the help of electric
fans or pumps, which increases the
drying rates (Lingayat, Chandramohan,
Raju, et al., 2020). PSD play an important
role in the drying sector because many
rural areas in SSA have poor access to the
electricity grid (Duran et al., 2015) and
because of their low cost (Mustayen et al.,
2014) compared to ASD. In fact, the
requirements of ASD are not affordable
to many rural households in developing
countries due to the additional costs of
the external energy source (Bala & Janjai,
2012; Veremachi et al., 2015).
Therefore, the performance of PSD needs
to be further improved and evaluated.
Erick César et al. (2020) designed and
evaluated a PSD with the option operate
as an ISD or MSD for drying tomato slices
and found an overall efficiency of 8.8%
and 10.7% for the ISD and MSD,
respectively. Mahapatra and Tripathy
(2019) tested the thermal performance
of PSD under no load conditions and
found efficiency of 31.4%, 27.6%, and
41.4% for DSD, ISD, and MSD,
respectively. Musembi et al. (2016)
designed and tested a passive indirect
solar dryer (PISD) for drying fresh apples
and found an overall dryer efficiency of
Results
17.9%. Several other studies have also
shown high performance of PSD as
compared to OSD (Arunsandeep et al.,
2018; Dasin et al., 2015; Ghaffari &
Mehdipour, 2015; Irtwange & Adebayo,
2009; Mohammed, Fatumah, et al., 2020;
Tedesco et al., 2018; Yadav et al., 2018).
Notably, solar dryers exhibit variations
in their overall performance, attributed
to factors such as the solar dryer’s design
(Nabnean et al., 2016), unstable ambient
temperature, relative humidity, hours of
sunshine, available solar radiation,
frequency and duration of rain, and wind
speed (Shahi et al., 2011). Therefore,
research efforts aimed at adapting solar
dryers to specific site conditions are
critical for more effective use of solar
drying systems. In addition, to avoid
underutilization of a new technology, it is
crucial to consider the preferences of
potential users (Foster & Rosenzweig,
2010), especially the sensory
characteristics of food (Leng et al., 2017).
Attributes such as taste, texture, smell or
72
appearance often tend to be less
negotiable than other values (Furst et al.,
1996). Thus, in this study we first
constructed and tested a PISD for drying
agricultural products as an alternative to
OSD in Guruè district, Central
Mozambique—where there is a limited
access to costly modern food processing
technologies—and secondly analyzed
the acceptability of food dried with PISD
using consumers’ preference analysis.
Material and methods
Description of the solar dryer
Since the constructed dryer is to be used
in a rural area with limited access to
fossil fuels, electricity and natural gas, a
passive solar dryer design suitable for
this environment was chosen
(Figure 4.4). It was made of wood and
consisted of a solar collector and a
separate drying chamber with five drying
trays. The dimensions of the solar
collector are 0.3 × 0.75 × 1.90 m and the
volume of the drying chamber is 1 m3.
Results
Figure 4.4: Schematic view of passive indirect solar dryer
The solar collector was tilted (22°) so
that its surface is perpendicular to the
solar radiation, which ensures that more
solar energy is collected and allows the
hot air to rise naturally into the drying
chamber (Forson et al., 2007; Handoyo et
al., 2013). The base of the collector was
vented to allow the entry of air that needs
to be heated for drying. The venting was
evenly distributed across the entire
width of the collector base to prevent
individual areas in the collector from
overheating. The top of the collector was
completely open to the bottom of the
drying chamber. The absorber plate was
suspended between the top clear cover
and the base plate, providing two
73
channels through which air flowed in the
same direction on both sides of the
absorber plate, thus, creating twice the
surface area for heat transfer to the air
(Forson et al., 2007). There were doors
on the drying chamber that allowed
access to the crop. At the top of the drying
chamber was an exit air vent to allow
natural air circulation throughout the
solar dryer.
Experimental procedure
Six solar dryers were constructed and
tested simultaneously with the OSD for
drying amaranth leaves (Amaranthus
gangeticus) and maize (Zea mays) in
August and September 2020,
Results 74

respectively. Maize was selected because
it is the main food crop grown in the
region (Chichongue et al., 2016) and
amaranth was selected among other
most commonly consumed green leafy
vegetables in collaboration with local
farmers. The experiments were
conducted in Guruè district (latitude:
15°10′46.9″S; longitude: 36°48′32.6″E),
Central Mozambique, one of the study
sites embedded in the “Vegi-Leg project”
(Vegi-Leg, 2019).
Drying of amaranth leaves
Fresh amaranth leaves were obtained
from the local market in Guruè, washed
in clean water, blanched in water at 90℃
for two minutes as described by Traoré
et al. (2017), and cooled on an open stand
for five minutes. Then, on the first day,
about 7 kg were evenly divided and
layered in the five drying trays. To avoid
the voids in the trays caused by product
shrinkage, we performed semi-
continuous drying according to Singh et
al. (2021). Thus, a total of 21.2 kg of fresh
leaves were loaded in each solar dryer
during the 7 days of the experiment. The
amount of fresh leaves loaded on
different days of the experiment is shown
in Table 4.7. Observations were made at
5:00 p.m. each day and the leaves that
were already dried were removed. As a
control, a tray with an area of 1 m2 was
placed in the ground as practiced by the
local residents for OSD.

Table 4.7: Fresh leaves loaded during successive drying days

Experiment Weight of fresh leaves Weight at 8:00 Weight at Weight of dried leaves
day loaded (g) a.m. (g) 17:00 (g) removed (g)
1st 7,000 7,000 5,600 0
2nd 50 5,460 3,696 0
3rd 900 4,925 3,004 137
4th 2,250 5,326 2,856 356
5th 4,200 5,934 3,256 317
6th 2,671 5,134 2,967 319
7th 4,200 5,324 3,467 297

were periodically mixed at two-hour

Drying of maize
intervals. No voids were observed in the
Freshly harvested maize grains were
trays during the maize experiment;
loaded and stored in the PISD until they
hence, the semi-continuous mode of
reach ≤14% of moisture according to
drying was not followed. Similar to
Bern et al. (2013). The loading per unit
amaranth leaves, OSD was carried out as
aperture area was 13.5 kg/m2. The grains
Results
a control and 13.5 kg of maize was placed
in a 1 m2 tray.
Data collection and analysis
Ambient relative humidity and
temperature, relative humidity and
temperature inside the drying chamber
were recorded during the experiment
using the EL-USB-2-LCD Temperature
and Relative Humidity Data Logger. Solar
radiation on the aperture was measured
hourly using a solar power meter
(Tenmar TM 207). An electronic balance
(±0.1 g accuracy) was used to measure
the weight of dried products in each tray
at start and end of each day. The standard
oven method (Aoac, 1990) was used to
determine the moisture content (M) of
the amaranth leaves. An Agratronix MT-
16 Grain Moisture Tester was used to
determine the moisture content of maize.
Equations 1 and 2 were used to calculate
thermal efficiency of the dryer (η) and
the solar energy (S) required for 1 kg of
moisture removal, respectively. The
overall thermal efficiency was obtain by
calculating the average of all daily
efficiencies (Singh et al., 2021).
𝑚𝐿 (1)
Ƞ= 𝑥100
𝐼𝑎𝑣 𝐴𝑡
𝐼𝑎𝑣 (𝐴)𝑡 1 (2)
𝑆= 𝑥 between the acceptances of the products
𝑚 1000
Where: Ƞ is thermal efficiency of the
solar dryer (%); 𝐼𝑎𝑣 is daily average solar
75
radiation intensity (W/m2); 𝑚 is the
mass of water evaporated (kg); 𝐿 is
Latent heat of water (kJ/kg K); 𝐴 area of
solar collector (m2) and 𝑡 the time during
drying day (s).
Product acceptance
One day after the end of the drying
experiment, a sensorial analysis was
performed. We randomly selected a
group of 60 adult people (43% female
and 57% male) from six different
communities who individually evaluated
the texture, aroma, and color of dried
amaranth and maize grains from both
PISD and OSD following the ASTM
standard guide for two-sample
acceptance and preference testing with
consumers (ASTM-International, 2015).
People did not have prior knowledge of
the method used to dry the product so as
not to influence their choices. The
evaluators specified their level of
acceptance using a five points Likert
scale (1 = very bad to 5 = very good). Each
subject evaluated four samples in total,
two from each dryer, with blinded codes.
The frequency distribution of the scale
categories was calculated and the
Wilcoxon signed-rank test was used in
order to examine the differences
dried using the PISD and the OSD.
Results 76

Results ambient air and inside the drying

chamber during drying ranged from 41%
Drying of amaranth leaves
to 61% and 27% to 37%, respectively.
The values of temperature and relative
The solar radiation varied from an
humidity variation for the ambient air,
average of 254 to 765 W/m2. At end of
temperature and relative humidity in the
3rd, 4th, 5th, 6th, and 7th drying days, an
drying chamber and intensity of solar
average of 137, 356, 317, 319, and 297 g
radiation over time were recorded and
of dried leaves were removed,
their mean values are shown in Figure
respectively (Table 4.7). Thus, an
4.5. The ambient temperature ranged
average of 1,430 g of dried amaranth
from 20 to 27℃, while the temperature in
leaves were obtained from each of the six
the drying chamber varied from 27 to
solar dryers in 7 drying days.
38℃. The relative humidity for the
70 900
Temperature (oC) and Relative Humidity (%)

800
60
700
50

Solar radiation(W/m2)
600

40 500

30 400

300
20
200
10
100

0 0
8:00:00 AM 9:00:00 AM 10:00:00 11:00:00 12:00:00 1:00:00 PM 2:00:00 PM 3:00:00 PM 4:00:00 PM 5:00:00 PM
AM AM PM
Time

Ambient RH Ambient T Inside RH Inside T SR

Figure 4.5: Variation of the temperature and relative humidity at the ambient and inside
the drying chamber and solar radiation intensity for amaranth leaves

The thermal efficiency from day 1 to 7 5th day. This is 58.7% more drying time
was 13%, 18%, 19%, 23%, 27%, 18%, compared to PISD. The average thermal
and 16% (Figure 4.6). Thus, the overall efficiency for OSD was 7%, and the total
thermal efficiency was 19%. Initial dried amaranth leaves obtained by the
moisture content of the fresh leaves was end of the drying experiment was 850 g.
83.4% and final moisture content after The average solar energy input per unit
drying was 10.5%. In OSD, the first dried water removal was 12.53 MJ/kg for PISD
leaves (126 g) were obtained only on the and 24.1 MJ/kg for OSD (Figure 4.6).
Results 77

20.00 30%
Solar energy input per unit water
18.00
25%

Thermal efficiency (%)

16.00
14.00 20%
removal (MJ/kg)

12.00
10.00 15%
8.00
10%
6.00
4.00 5%
2.00
0.00 0%
Day1 Day2 Day3 Day4 Day5 Day6 Day7 Day1 Day2 Day3 Day4 Day5 Day6 Day7
Day# Day#

Figure 4.6: Average solar energy input per unit water removal and thermal efficiency for
amaranth leaves

Drying of maize ambient and indoor temperatures during

The drying of maize took 5 consecutive
days to reduce the grain moisture from
26% to 14%. The variations in ambient
air temperature and relative humidity,
drying
relative humidity, and solar radiation
were recorded, and their mean values
plotted as shown in Figure 4.7. The
drying ranged from 16.8 to 26.8℃ and
27.6 to 48.2℃, respectively. The ambient
relative humidity ranged from 40.8% to
59.4% and the drying chamber relative
humidity ranged from 26.8% to 38%. The
chamber temperature and
average solar radiation during different
drying days ranged from 225.0 to 724.6
W/m2.

70 800

700
60
Temperature (oC) and Relative Humidity (%)

600
50
Solar radiation(W/m2)

500
40

400

30
300

20
200

10
100

0 0
8:00:00 AM 9:00:00 AM 10:00:00 AM 11:00:00 AM 12:00:00 PM 1:00:00 PM 2:00:00 PM 3:00:00 PM 4:00:00 PM 5:00:00 PM
Time

Ambient RH Ambient T Inside RH Inside T SR

Figure 4.7: variation of the temperature and relative humidity at the ambient and inside
the drying chamber and solar radiation intensity for maize
Figure 4.8 shows the daily average solar and the thermal efficiency of the PSID.
energy input per unit water removal (S) The values of S varied from 12.99 MJ/kg
Results 78

in the first day of drying to 28.65 MJ/kg days to reach the desired moisture
in the 5th day. Thermal performance at content of ≤14%. Therefore, compared to
the 1st, 2nd, 3rd, 4th, and 5th days were OSD, PISD reduced drying time by 29%.
17%, 18%, 15%, 14% and 8%, The overall thermal efficiency of OSD was
respectively. Thus the average thermal 4.6%.
efficiency was 14%. For OSD, it took 7
35.00 20%
Solar energy input per unit

Thermal efficiency (%)

30.00
water removal (MJ/kg)

25.00 15%

20.00
10%
15.00
10.00 5%
5.00
0.00 0%
Day1 Day2 Day3 Day4 Day5 Day1 Day2 Day3 Day4 Day5
Day# Day#

Figure 4.8: Average solar energy input per unit water removal and thermal efficiency for
maize
Product acceptance 40% of the people rated the amaranth
leaves dried in the PISD as very good and
We found that 40%, 27%, and 23% of the
30% as good. None of the evaluators
evaluators consider that the OSD had
found the dried amaranth from the OSD
poor texture, aroma and color,
very good and only 23% found it good.
respectively (Figure 4.9). None of the
The Wilcoxon signed-rank test was
evaluators considered the sensory
significant (pvalue < 0.01) for all sensory
attributes of amaranth dried in the PISD
attributes tested, indicating that most
to be bad or very bad. Most of them
people preferred the dried amaranth
(more than 68%) indicated that the
leaves from the PISD.
sensory properties were good or very
good. In terms of overall acceptability,
Results 79

TEXTURE AROM A

Solar Dryer OpenSun Solar Dryer OpenSun

50% 50% 43%

40% 38% 40%
35% 37%
40% 32% 33% 40% 33%
27%
30% 22% 30%
20%
20% 20%
10% 10%
0% 0% 0% 0% 0% 0% 0% 0%
0% 0%
Very bad Bad Average Good Very Good Very bad Bad Average Good Very Good

COLOUR OVERALL ACCEPTABILIT

Solar Dryer OpenSun Solar Dryer OpenSun

45% 40% 47%

50%
40% 45% 40%
35% 32% 40%
28%30% 28%
30% 35% 30%30% 30%
23%
25% 30% 23%
18% 25%
20%
20%
15%
15%
10% 10%
5% 0% 0% 0% 5% 0% 0% 0% 0%
0% 0%
Very bad Bad Average Good Very Good Very bad Bad Average Good Very Good

Figure 4.9: Respondents preference for amaranth leaves sensory attributes

With regards to dried maize, the
evaluators also found the sensory
attributes of maize dried in PISD better
than that dried via OSD (p-value < .01).
The percentage distribution of the
evaluators by score category is shown in
Figure 4.10. In general, the percentage of
people who assign values 4 (good) and 5
(very good) for maize dried in PISD
varies between 65%
depending on the sensory attribute. As in
the evaluation of amaranth, none of the
evaluators found the sensory attributes
of maize dried in PISD bad or very bad,
while 28%, 32%, and 33% of the
evaluators found the texture, aroma, and
color of the maize dried in the OSD bad,
respectively. Furthermore, none of the
evaluators found the sensory attributes
and 75%, of maize dried in the OSD to be very good.
Results 80

TEXTURE AROM A

Solar Dryer OpenSun Solar Dryer OpenSun

50% 43% 45% 40%

38%
45% 38% 40%
40% 37% 32%
35% 30%30% 30%
35% 28% 28% 30%
30% 25%
25%
25%
20%
20%
15% 15%
10% 10%
5% 0% 0% 0% 0% 5% 0% 0% 0% 0%
0% 0%
Very bad Bad Average Good Very Good Very bad Bad Average Good Very Good

COLOUR OVERALL ACCEPTABILIT

Solar Dryer OpenSun Solar Dryer OpenSun

47% 50% 45%

50%
38%
40% 33% 35%35% 40% 35%
32% 32%
27%
30% 30% 23%
18%
20% 20%

10% 10%
0% 0% 0% 0% 0% 0% 0% 0%
0% 0%
Very bad Bad Average Good Very Good Very bad Bad Average Good Very Good

Figure 4.10: Respondents preference for maize sensory attributes

Discussion which most fungal species cannot grow

as they require relative humidity above
In this study, six solar dryers were
60% (Arundel et al., 1986). In OSD, the
constructed and tested simultaneously
observed relative humidity during the
and their performance was evaluated.
experiments was around 60%, making
They were tested for drying amaranth
the crop more susceptible to attack by
leaves and maize grains. In general, the
fungi. The comparatively lower relative
ambient and internal temperature of the
humidity in the drying chamber of PSD
drying chamber and the solar radiation
allows for greater removal of moisture
increased to a peak in the afternoon and
from the products being dried, as the
decreased in the evening (cf. Figures 2
reduction in humidity increases the
and 3). This is in agreement with the
diffusion of moisture from the product
results of Ayua et al. (2017) and
into the air, which accelerates the drying
Ssemwanga et al. (2020). Similar to
process (Aravindan et al., 2017). The
Nimrotham et al. (2017), the relative
drying temperatures in the drying
humidity decreased to its minimum
chamber ranged from 27.6 to 48.2℃.
value during the afternoons. The relative
These values are similar to those
humidity values observed inside PSD’s
observed in other studies on PISD (Jain &
drying chamber during the experiments
Tewari, 2015; Ssemwanga et al., 2020;
were below 40%, which is a range in
Results
Vijayan et al., 2016). However, a very
wide variation was observed. For
example, Jain and Tewari (2015) found
values that varied between 40 and 45℃
and A. Lingayat et al. (2017) observed
drying air temperature that ranged
between 44 and 55℃. Nevertheless,
Ahmad Fudholi et al. (2014) found
Drying chamber air temperature ranging
from 28 to 55℃. This variation is due to
changes in the intensity of solar radiation
and can be even higher depending on the
position of the tray in the drying chamber
(Lingayat, Chandramohan, & Raju, 2020).
In addition, factors such as the design of
the dryer and the season in which the
analysis was carried out can also
influence the temperature variation in
the drying chamber.
The average thermal efficiency of PISD
was higher than the efficiency of the OSD.
Compared to the results of other studies,
it was lower than that of Lingayat,
Chandramohan and Raju (2020)
(21.57%) and Mahapatra and Tripathy
(2019) (27.55%) and higher than the
efficiency found by Erick César et al.
(2020) (2.61%). According to Kumar et
al. (2016), the average drying efficiency
of a PISD is 13%–25%. Thus, the results
of this study are in an acceptable range
for a PSD. The differences in the
efficiency reported in different studies
81
are, according to Shahi et al. (2011) due
to differences in the ambient
temperature, relative humidity, hours of
sunshine, available solar radiation,
frequency and duration of rain, and wind
speed.
The sensory attributes are important
aspects when a food is presented for
consumption. Therefore, it is important
for food processors to know the sensory
characteristics of their products (Geel et
al., 2005). Therefore, we conducted a
consumer acceptance analysis to find out
which drying method resulted in a
product with comparatively better
sensory properties (taste, aroma, color
and overall acceptability). The results
showed that both amaranth and maize
from PISD were rated better than those
from OSD. The advantages of PSD over
OSD in terms of final products quality
have been demonstrated in several
studies (Hii et al., 2019; Irtwange &
Adebayo, 2009; Mohammed, Edna, et al.,
2020; Udomkun et al., 2020).
Nevertheless, further studies are needed
to explore factors such as demographic
profile, health status, personality,
knowledge, exposure, perceived quality
and mood as they may have an impact on
consumers’ perception and evaluation of
a particular product (Owureku-Asare et
al., 2017; Rozin & Tuorila, 1993).
Results
According to Steenkamp et al. (1994),
despite its sensitivity, the human sensory
system cannot distinguish minor
differences between products. However,
in the case of the present study, our
results are supported by the fact that the
use of solar dryers as a substitute for OSD
increases drying air temperatures,
thermal energy and drying rate (Kumar
et al., 2016; Orphanides et al., 2016),
which consequently can improve the
organoleptic quality of the dried
products (Mohammed, Edna, et al.,
2020).
Conclusion
The performance of a passive indirect
solar dryer as compared to OSD was
evaluated, using amaranth and maize. In
addition, the sensory quality attributes
were evaluated by a group of randomly
selected individuals. The results showed
that in the passive indirect solar dryer,
the average drying air temperature was
higher and the relative humidity lower, in
comparison to the OSD. The sensory
quality attributes evaluation showed that
passive indirect solar dryer
outperformed OSD in terms of texture,
aroma, color and overall consumer
acceptability. Thus, based on the
outcome of this study, it is concluded that
the use of passive indirect solar dryer is a
sustainable way of drying agricultural
82
products and its use is recommended
rather than OSD.
Acknowledgments: The authors would
like to thank the Leibniz Centre for
Agricultural Landscape Research (ZALF)
in Müncheberg, Germany, for their
logistical support in conducting the
research in Gurué district, central
Mozambique and Federal Ministry of
Food and Agriculture (BMEL) for the
financial support. Open access funding
enabled and organized by ProjektDEAL.
Funding information: The study is
financially supported by the Federal
Ministry of Food and Agriculture (BMEL)
based on the decision of the Parliament
of the Federal Republic of Germany and
managed by the Federal Office for
Agriculture and Food (BLE
Conflict of interest: The authors have
declared no conflicts of interest for this
article.
Author contributions: Custodio Efraim
Matavel: Conceptualization; Formal
analysis; Investigation; Methodology;
Writing-original draft; Writing-review &
editing. Harry Hoffmann:
Conceptualization; Funding acquisition;
Project administration; Supervision;
Writing-review & editing. Constance
Rybak: Conceptualization; Funding
acquisition; Project administration;
Results
Writing-review & editing. Johannes M.
Hafner: Writing-original draft; Writing-
review & editing. João Salavessa:
Writing-review & editing. Shibire Bekele
Eshetu: Writing-review & editing. Stefan
Sieber: Conceptualization; Supervision;
Writing-review & editing.
References
Adeyeye, S. A. O. (2017). "The role of food
processing and appropriate storage
technologies in ensuring food security
and food availability in Africa." Nutrition
& Food Science.
Affognon, H., C. Mutungi, P. Sanginga and
C. Borgemeister (2015). "Unpacking
Postharvest Losses in Sub-Saharan
Africa: A Meta-Analysis."
Development 66: 49-68.
Al-Juamily, K. E. J., A. J. N. Khalifa and T. A.
Yassen (2007). "Testing of
performance of a fruit and vegetable
solar drying system in Iraq." Desalination
209(1): 163-170.
Aoac (1990). Association of official
analytical chemists. Official methods of
analysis, AOAC Arlington, VA.
Aravindan, V., A. Dineshkumar,
Giriprasath, V. Karthikeyan and D.
Ebenezer (2017). "Moisture removal rate
of solar dryers–A review." Journal of
Chemical and Pharmaceutical Sciences
ISSN 974: 2115.
83
Arundel, A. V., E. M. Sterling, J. H. Biggin
and T. D. Sterling (1986). "Indirect health
effects of relative humidity in indoor
environments." Environmental health
perspectives 65: 351-361.
Arunsandeep, G., A. Lingayat, V. P.
Chandramohan, V. R. K. Raju and K. S.
Reddy (2018). "A numerical model for
drying of spherical object in an indirect
type solar dryer and estimating the
drying time at different moisture level
and air temperature." International
Journal of Green Energy 15(3): 189-200.
AUC (2014). Malabo Declaration on
Accelerated Agricultural Growth and
Transformation for Shared Prosperity
and Improved Livelihoods. Malabo,
World
Equatorial Guinea African Union
Commission.
Ayua, E., V. Mugalavai, J. Simon, S. Weller,
the
P. Obura and N. Nyabinda (2017).
"Comparison of a mixed modes solar
dryer to a direct mode solar dryer for
African indigenous vegetable and chili
processing." Journal of Food Processing
and Preservation 41(6): e13216.
Babar, O. A., A. Tarafdar, S. Malakar, V. K.
B.
Arora and P. K. Nema (2020). "Design and
performance evaluation of a passive flat
plate collector solar dryer for
agricultural products." Journal of Food
Process Engineering 43(10): e13484.
Results
Bala, B. K. and S. Janjai (2012). Solar
Drying Technology: Potentials and
Developments. Energy, Environment and
Sustainable Development. M. A. Uqaili
and K. Harijan. Vienna, Springer Vienna:
69-98.
Basunia, M. A. and T. Abe (2001). "Thin-
layer solar drying characteristics of
rough rice under natural convection."
Journal of Food Engineering 47(4): 295-
301.
Calín-Sánchez, Á., A. Figiel, F. Hernández,
P. Melgarejo, K. Lech and Á. A. Carbonell-
Barrachina (2013). "Chemical
Composition, Antioxidant Capacity, and
Sensory Quality of Pomegranate (Punica
granatum L.) Arils and Rind as Affected
by Drying Method." Food and Bioprocess
Technology 6(7): 1644-1654.
Chavan, A., V. Vitankar, A. Mujumdar and
B. Thorat (2020). "Natural convection
and direct type (NCDT) solar dryers: a
review." Drying Technology: 1-22.
Dasin, D., N. Godi and O. Kingsley (2015).
"Experimental investigations of the
performance of passive solar food dryer
tested in Yola-Nigeria." International
Journal of Energy Engineering 5(1): 9-15.
Duran, G., M. Condorí and F. Altobelli
(2015). "Simulation of a passive solar
dryer to charqui production using
84
temperature and pressure networks."
Solar Energy 119: 310-318.
Ekechukwu, O. V. and B. Norton (1999).
"Review of solar-energy drying systems
II: an overview of solar drying
technology." Energy Conversion and
Management 40(6): 615-655.
El-Sebaii, A. A., S. Aboul-Enein, M. R. I.
Ramadan and H. G. El-Gohary (2002).
"Experimental investigation of an
indirect type natural convection solar
dryer." Energy Conversion and
Management 43(16): 2251-2266.
Elias, N. d. F., P. A. Berbert, M. A. B. d.
Molina, A. P. Viana, R. G. Dionello and V.
A. V. Queiroz (2008). "Nutritional and
sensory evaluation of Osmo-convective
dried Fuyu persimmons." Food Science
and Technology 28(2): 322-328.
Erick César, L.-V., C.-M. Ana Lilia, G.-V.
Octavio, P. F. Isaac and B. O. Rogelio
(2020). "Thermal performance of a
passive, mixed-type solar dryer for
tomato slices (Solanum lycopersicum)."
Renewable Energy 147: 845-855.
Esper, A. and W. Mühlbauer (1998).
"Solar drying - an effective means of food
preservation." Renewable Energy 15(1):
95-100.
FAO (2019). The State of Food and
Agriculture 2019: Moving forward on
Results 85

food loss and waste reduction. Rome, Geel, L., M. Kinnear and H. L. de Kock
FAO. (2005). "Relating consumer preferences
to sensory attributes of instant coffee."
FAO, ECA and AUC (2020). Africa
Food Quality and Preference 16(3): 237-
Regional Overview of Food Security and
244.
Nutrition 2019. Accra, Ghana, FAO.
Gerland, P., A. E. Raftery, H. Sevčíková, N.
Forson, F. K., M. A. A. Nazha, F. O. Akuffo
Li, D. Gu, T. Spoorenberg, L. Alkema, B. K.
and H. Rajakaruna (2007). "Design of
Fosdick, J. Chunn, N. Lalic, G. Bay, T.
mixed-mode natural convection solar
Buettner, G. K. Heilig and J. Wilmoth
crop dryers: Application of principles
(2014). "World population stabilization
and rules of thumb." Renewable Energy
unlikely this century." Science (New
32(14): 2306-2319.
York, N.Y.) 346(6206): 234-237.
Foster, A. D. and M. R. Rosenzweig
Ghaffari, A. and R. Mehdipour (2015).
(2010). "Microeconomics of technology
"Modeling and Improving the
adoption." Annu. Rev. Econ. 2(1): 395-
Performance of Cabinet Solar Dryer
424.
Using Computational Fluid Dynamics."
Fudholi, A., K. Sopian, M. H. Ruslan, M. A.
International Journal of Food
Alghoul and M. Y. Sulaiman (2010).
Engineering 11(2): 157-172.
"Review of solar dryers for agricultural
Handoyo, E. A., D. Ichsani and Prabowo
and marine products." Renewable and
(2013). "The Optimal Tilt Angle of a Solar
Sustainable Energy Reviews 14(1): 1-30.
Collector." Energy Procedia 32: 166-175.
Fudholi, A., K. Sopian, M. H. Yazdi, M. H.
Hii, C., S. Ong, C. Chiang and A. Menon
Ruslan, M. Gabbasa and H. A. Kazem
(2019). A review of quality
(2014). "Performance analysis of solar
characteristics of solar dried food crop
drying system for red chili." Solar Energy
producst. IOP Conference Series: Earth
99: 47-54.
and Environmental Science, IOP
Gbaha, P., H. Yobouet Andoh, J. Kouassi
Publishing.
Saraka, B. Kaménan Koua and S. Touré
Hodges, R. J., J. C. Buzby and B. Bennett
(2007). "Experimental investigation of a
(2011). "Postharvest losses and waste in
solar dryer with natural convective heat
developed and less developed countries:
flow." Renewable Energy 32(11): 1817-
opportunities to improve resource use."
1829.
Results
The Journal of Agricultural Science
149(S1): 37.
Irtwange, S. and S. Adebayo (2009).
"Development and performance of a
laboratory-scale passive solar grain
dryer in a tropical environment." Journal
of Agricultural Extension and Rural
Development 1(2): 042-049.
Islam, M. K., M. S. Karim, N. N. Begum and
K. Z. Uddin (2018). "Fabrication and
performance study of a direct type solar
dryer." International Journal of
Engineering Research 9(2): 565-569.
J. Bern, C., A. Yakubu, T. J. Brumm and K.
A. Rosentrater (2013). Hermetic storage
systems for maize stored on subsistence
farms. 2013 Kansas City, Missouri, July
21 - July 24, 2013. St. Joseph, MI, ASABE:
1.
Jain, D. and P. Tewari (2015).
"Performance of indirect through pass
natural convective solar crop dryer with
phase change thermal energy storage."
Renewable Energy 80: 244-250.
Janjai, S. and B. K. Bala (2012). "Solar
Drying Technology." Food Engineering
Reviews 4(1): 16-54.
Kalogirou, S. A. (2014). Chapter 1 -
Introduction. Solar Energy Engineering
(Second Edition). S. A. Kalogirou. Boston,
Academic Press: 1-49.
86
Kaminski, J. and L. Christiaensen (2014).
"Post-harvest loss in sub-Saharan
Africa—what do farmers say?" Global
Food Security 3(3): 149-158.
Karekezi, S. (2002). "Poverty and energy
in Africa—A brief review." Energy Policy
30(11): 915-919.
Khaing Hnin, K., M. Zhang, A. S. Mujumdar
and Y. Zhu (2019). "Emerging food
drying technologies with energy-saving
characteristics: A review." Drying
Technology 37(12): 1465-1480.
Kumar, M., S. K. Sansaniwal and P. Khatak
(2016). "Progress in solar dryers for
drying various commodities." Renewable
and Sustainable Energy Reviews 55: 346-
360.
Lamidi, R. O., L. Jiang, P. B. Pathare, Y. D.
Wang and A. P. Roskilly (2019). "Recent
advances in sustainable drying of
agricultural produce: A review." Applied
Energy 233-234: 367-385.
Leng, G., R. A. Adan, M. Belot, J. M.
Brunstrom, K. de Graaf, S. L. Dickson, T.
Hare, S. Maier, J. Menzies and H. Preissl
(2017). "The determinants of food
choice." Proceedings of the Nutrition
Society 76(3): 316-327.
Lingayat, A., V. Chandramohan and V.
Raju (2017). "Design, development and
performance of indirect type solar dryer
Results 87

for banana drying." Energy Procedia 109: Journal of Stored Products Research 88:
409-416. 101634.

Lingayat, A., V. P. Chandramohan and V. Mohana, Y., R. Mohanapriya, T.

R. K. Raju (2020). "Energy and Exergy
Analysis on Drying of Banana Using
Indirect Type Natural Convection Solar
Dryer." Heat Transfer Engineering 41(6-
7): 551-561.
Anukiruthika, K. S. Yoha, J. A. Moses and
C. Anandharamakrishnan (2020). "Solar
dryers for food applications: Concepts,
designs, and recent advances." Solar
Energy 208: 321-344.

Lingayat, A. B., V. P. Chandramohan, V. R. Musembi, M. N., K. S. Kiptoo and N. Yuichi

K. Raju and V. Meda (2020). "A review on (2016). "Design and Analysis of Solar
indirect type solar dryers for agricultural Dryer for Mid-Latitude Region." Energy
crops – Dryer setup, its performance, Procedia 100: 98-110.
energy storage and important Mustapha, M. K., T. B. Ajibola, A. F. Salako
highlights." Applied Energy 258: 114005. and S. K. Ademola (2014). "Solar drying
Mahapatra, A. and P. P. Tripathy (2019). and organoleptic characteristics of two
"Thermal performance analysis of tropical African fish species using
natural convection solar dryers under no improved low-cost solar driers." Food
load condition: experimental Science & Nutrition 2(3): 244-250.
investigation and numerical simulation." Mustayen, A. G. M. B., S. Mekhilef and R.
International Journal of Green Energy Saidur (2014). "Performance study of
16(15): 1448-1464. different solar dryers: A review."
Mohammed, S., M. Edna and K. Siraj Renewable and Sustainable Energy
(2020). "The effect of traditional and Reviews 34: 463-470.
improved solar drying methods on the Nabnean, S., S. Janjai, S. Thepa, K.
sensory quality and nutritional Sudaprasert, R. Songprakorp and B. K.
composition of fruits: A case of mangoes Bala (2016). "Experimental performance
and pineapples." Heliyon 6(6): e04163. of a new design of solar dryer for drying
Mohammed, S., N. Fatumah and N. Shadia osmotically dehydrated cherry
(2020). "Drying performance and tomatoes." Renewable Energy 94: 147-
economic analysis of novel hybrid 156.
passive-mode and active-mode solar Ndukwu, M. C., L. Bennamoun and F. I.
dryers for drying fruits in East Africa." Abam (2018). "Experience of Solar
Results
Drying in Africa: Presentation of Designs,
Operations, and Models."
Engineering Reviews 10(4): 211-244.
Nimrotham, C., R. Songprakorp, S. Thepa
and V. Monyakul (2017). "Experimental
Research of Drying Red Chili by Two
Methods: Solar Drying and Low -
Temperature System Drying." Energy
Procedia 138: 512-517.
Orphanides, A., V. Goulas and V. Gekas
(2016). "Drying technologies: vehicle to
high-quality herbs." Food Engineering
Reviews 8(2): 164-180.
Owureku‐Asare, M., R. K. Ambrose, I.
Oduro, C. Tortoe and F. K. Saalia (2017).
"Consumer knowledge, preference, and
perceived quality of dried tomato
products in Ghana." Food science &
nutrition 5(3): 617-624.
Pangavhane, D. R., R. L. Sawhney and P. N.
Sarsavadia (2002). "Design, development
and performance testing of a new natural
convection solar dryer." Energy 27(6):
579-590.
Reyes, A., A. Mahn and F. Vásquez (2014).
"Mushrooms dehydration in a hybrid-
solar dryer, using a phase change
material." Energy Conversion
Management 83: 241-248.
Rozin, P. and H. Tuorila (1993).
"Simultaneous and temporal contextual
88
influences on food acceptance." Food
Food Quality and Preference 4(1): 11-20.
Sandali, M., A. Boubekri, D. Mennouche
and N. Gherraf (2019). "Improvement of
a direct solar dryer performance using a
geothermal water heat exchanger as
supplementary energetic supply. An
experimental investigation and
simulation study." Renewable Energy
135: 186-196.
Shahi, N. C., J. N. Khan, U. C. Lohani, A.
Singh and A. Kumar (2011).
"Development of polyhouse type solar
dryer for Kashmir valley." Journal of
Food Science and Technology 48(3): 290-
295.
Shalaby, S. M., M. A. Bek and A. A. El-
Sebaii (2014). "Solar dryers with PCM as
energy storage medium: A review."
Renewable and Sustainable Energy
Reviews 33: 110-116.
Sheahan, M. and C. B. Barrett (2017).
"Review: Food loss and waste in Sub-
Saharan Africa." Food Policy 70: 1-12.
Singh, S., R. S. Gill, V. S. Hans and M. Singh
(2021). "A novel active-mode indirect
solar dryer for agricultural products:
and Experimental evaluation and economic
feasibility." Energy 222: 119956.
Ssemwanga, M., E. Makule and S. I.
Kayondo (2020). "Performance analysis
of an improved solar dryer integrated
Results 89

with multiple metallic solar Veremachi, A., B. C. Cuamba, O. J. Nydal, J.

concentrators for drying fruits." Solar Lovseth and A. Zia (2015). Direct
Energy 204: 419-428. illuminated rock-bed heat storage a
potential component of a solar thermal
Stathers, T., R. Lamboll and B. M. Mvumi
system for food preservation and space
(2013). "Postharvest agriculture in
heating in rural areas of Mozambique.
changing climates: its importance to
ISES Solar World Conference, Korean
African smallholder farmers." Food
Solar Energy Soc, Daegu, South Korea,
Security 5(3): 361-392.
Nov 08-12, 2015, International Solar
Steenkamp, J.-B. E., H. C. Van Trijp and J.
Energy Society (ISES).
M. T. Berge (1994). "Perceptual mapping
Vijayan, S., T. Arjunan and A. Kumar
based on idiosyncratic sets of attributes."
(2016). "Mathematical modeling and
Journal of Marketing Research 31(1): 15-
performance analysis of thin layer drying
27.
of bitter gourd in sensible storage based
Tedesco, F. C., A. J. Bühler and S.
indirect solar dryer." Innovative food
Wortmann (2018). "Design,
science & emerging technologies 36: 59-
Construction, and Analysis of a Passive
67.
Indirect Solar Dryer With Chimney."
Visavale, G. (2012). "Principles,
Journal of Solar Energy Engineering
classification and selection of solar
141(3).
dryers." Solar drying: Fundamentals,
Tomar, V., G. N. Tiwari and B. Norton
Applications and Innovations, Ed. Hii, CL,
(2017). "Solar dryers for tropical food
Ong, SP, Jangam, SV and Mujumdar, AS,
preservation: Thermophysics of crops,
Published in Singapore: 1-50.
systems and components." Solar Energy
WB. (2016). "A Year in the Lives of
154: 2-13.
Smallholder Farmers." Retrieved 29/3,
Udomkun, P., S. Romuli, S. Schock, B.
2021, from
Mahayothee, M. Sartas, T. Wossen, E.
https://www.worldbank.org/en/news/f
Njukwe, B. Vanlauwe and J. Müller
eature/2016/02/25/a-year-in-the-lives-
(2020). "Review of solar dryers for
of-smallholder-farming-families.
agricultural products in Asia and Africa:
Yadav, S., A. B. Lingayat, V. P.
An innovation landscape approach."
Chandramohan and V. R. K. Raju (2018).
Journal of Environmental Management
"Numerical analysis on thermal energy
268: 110730.
storage device to improve the drying
Results 90

time of indirect type solar dryer." Heat

and Mass Transfer 54(12): 3631-3646.
Results 91

4.3 Toward energy saving and food safety in Central Mozambique: the role of
improved cook stoves and heat retention boxes

Energy, Sustainability and Society (2022), volume 12, Article number: 26.

Custodio E. Matavela,b*, Johannes M. Hafnera,b, Harry Hoffmanna, Götz Uckerta, Jonas

Massuquec, Constance Rybaka, Stefan Siebera,b

aLeibniz Centre for Agricultural Landscape Research (ZALF), Eberswalder Str. 84, 15374 Müncheberg,
Germany

bDepartment of Agricultural Economics, Faculty of Life Sciences Thaer-Institute, Humboldt-Universität

zu Berlin, Unter den Linden 6, 10099 Berlin, Germany

c Faculty of Agrarian Sciences, Universidade Lúrio, Department of Environment and Nature Conservation,

Unango Campus, Sanga District, Niassa Province, Mozambique.

© The Author(s) 2022. Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/.

https://doi.org/10.1186/s13705-022-00352-6

Abstract

Background: Almost 80% of the population in sub-Saharan Africa relies on traditional

biomass for cooking, which is typically associated with negative environmental, health,
economic, and social impacts. Thus, many stakeholders, including development agencies
and national governments in the Global South are promoting the use of the improved
cookstove in order to save cooking time, save financial assets, maximize fuel efficiency,
and reduce indoor air pollution. However, little attention is paid to the heating practices
among households, which can determine food safety levels. Specifically, cooked food
should be kept at temperatures above the danger zone (from 5 to 57 °C) prior to its
consumption to prevent its contamination by bacteria and other unhealthy contaminants.
In general, many studies address food preparation and storage separately, despite being
Results 92

complementary. In this study, we attempt to understand whether, the use of improved

cookstove combined with heat retention box would result in improvements with regard
to fuel and time saving, and adequate food storage temperatures. Furthermore, we
examine the acceptability of food prepared with these two systems based on consumers’
preference analysis. Involving 122 participants, the study was conducted in Gurué district,
central Mozambique.

Results: The use of improved cookstove resulted in energy savings of 9% and 17% for
cooking maize porridge and beans curry, respectively. The overall time consumption for
cooking decreased by 14% (beans curry) and 24% (maize porridge). The use of heat
retention boxes shows a better heat retention ability as compared to the locally used heat
retention systems (leftovers, banana leaves).

Conclusions: The study concludes that improved cookstove is a sustainable mean for
saving cooking time and fuel. Heat retention box has a potential to maintain adequate food
storage temperatures. Both improved cookstove and heat retention box present a
superior performance compared to traditional technologies; thus, can easily be diffused
for not affecting the quality of food.

Keywords: Biomass energy, Foodborne illness, Time saving, Fuel saving, Gurué

Background collection are additional health concerns,

Almost 80% of the population in sub-
Saharan Africa (SSA)—roughly
million people—relies on traditional
biomass for cooking [1, 2], which is
typically associated with negative
environmental, health, economic, and
social impacts [3,4,5,6,7,8].
traditional uses of biomass, for example,
result in more than 600,000 deaths
annually due to ambient air pollution in
SSA [9, 10]. Beyond the indoor air
pollution, gender-biased time [11] and
physical burdens associated to fuelwood
especially for females [12]. Women often
have to carry heavy loads, e.g., 14–36 kg
780
[13], which results in musculoskeletal
pain [14]. As increasing time is spent on
fuelwood collection, there is less time for
other subsistence activities or recreation
[15]. Moreover, the consumption of
The
fuelwood increases the pressure on
forest resources and the consequent
forest degradation [16], which threatens
the biodiversity conservation [17].
To reduce these negative effects, it is
crucial to invest in more efficient and
Results
clean cooking technologies such as fuel-
efficient cookstoves, since in
Mozambique, for example, the current
average efficiency of fuelwood use in
three-stone-fires (TSF) is estimated to be
around 10% [18], while the efficient
cookstoves can reach efficiency levels of
between 20 and 30% [19, 20].
In this context, many stakeholders,
including development agencies and
national governments in the Global
South, are promoting the use of the
improved cookstove (ICS) in order to
reduce the adverse environmental,
health, economic and social impacts
associated with the traditional solid
biomass. These stoves can be
constructed of ceramic, mud, or metal
[21]. In general, the disseminated stoves
were primarily designed to maximize
thermal and, thus, fuel efficiency [22]
while reducing indoor air pollution [23,
24], although the current designs do not
yet meet the World Health Organization
air quality guidelines [25]. Moreover,
other more advanced stove designs also
exist [26]. Associated advantages, such as
(potentially) reducing deforestation and
forest degradation, saving cooking time,
and saving financial assets to purchase
fuels are connected to the use of the ICS
[27,28,29]. Yet, the total replacement of
93
traditional stoves in developing
countries remains far from reality [30].
There are several test protocols to assess
the performance of the ICS. Some are
laboratory-based, e.g., water boiling test
(WBT) [31], while others are field-based,
e.g., kitchen performance test (KPT) [32].
The controlled cooking test (CCT) was
developed as a mix of these approaches
[33]. It allows for comparing traditional
stoves with ICS regarding fuel
consumption and cooking time under
controlled conditions [34]. Moreover, it
is recommended for testing stove
performance when the aim is to
understand how stoves perform with
local foods, fuels, and cooking practices
[35].
From a health perspective, while it is
important to reduce particle emissions
during the cooking process, thus
lowering indoor air pollution [36], on the
other hand it essential enhance the
digestibility, taste, texture, and shelf-life
of food [37]. Latter is particularly
ensured by access to potable water,
adequate food handling and storage
conditions, adequate sanitation and
hygiene [38]. Thus, it is also critical to
look beyond food preparation as, for
example, the heating practices and
storage of cooked food at ambient
temperature for extended periods can
Results
determine the level of food safety [39], a
very serious but neglected health
problem. In this study, we refer to food
safety as actions aimed at protecting
foods from biological and physical
hazards that may occur during
preparation and consumption [40]. By
heating food, the multiplication of
pathogenic microorganisms and,
therefore, food-borne illnesses—which
have an economic loss from deaths of
approximately 39 billion USD in SSA
[41]—can be reduced [42]. In fact, the
health burdens caused by food-borne
diseases in SSA are comparable to
malaria, HIV/AIDS and tuberculosis [41].
According to Ricci et al. [43], food should
generally be maintained at temperatures
below 5 °C or above 60 °C, to avoid the so
called “temperature danger zone” (from
5 to 57 °C [44]) in which pathogenic
microorganisms grow quickly. To keep
cooked food at temperatures above this
danger zone, technology that can retain
thermal energy at least for some time is
needed. The alternative of maintaining
temperatures below 5 °C through
refrigeration is challenging in rural
Mozambique because access to
electricity remains low: only 5.7% of the
population has access to electricity in
rural areas [45]. Furthermore, people in
rural areas may be very unlikely to warm
their food before consumption, after it
94
cools down [46]. A promising solution to
this challenge is the use of heat retention
systems (HRS), here defined as systems
that can temporary hold thermal energy
in the form of hot substances for later use
[47]. The advantages of these systems
include keeping food at temperatures
above 60 °C for several hours after the
cooking pot is taken off the heat source.
Furthermore, some of these systems can
easily be manufactured and all that is
required is to place the pots inside [48].
Therefore, combining ICS with HRS can
lead to very positive results in terms of
energy savings [49] and the prevention
of foodborne illnesses [50]. However, the
use and suitability of HRS in rural areas
is poorly documented.
Although several (meta-)studies analyze
the performance of different ICS designs
[28, 51,52,53,54], there are very few
analyses on the lapse of time between the
food being ready and its actual
consumption, especially in rural areas of
developing countries [49]. Only a limited
number of studies, including Tiffany et al.
[55] and Kaushik [56], assess the
performance of HRS, especially those
designed for domestic use in rural areas
of developing countries. There is still no
standardized test protocol for domestic
HRS. The commonly applied method to
evaluate the performance of an HRS is by
Results 95

monitoring temperature change inside Secondly, we examine the acceptability of

the HRS for a certain period [55, 57]. In food prepared with these systems based
addition, studies generally address food on consumers’ preference analysis. The
preparation and handling separately, results of this study present a more
despite being complementary. holistic view of the necessary
interventions to reduce cooking energy
It could be argued that positive health
consumption and increase food safety in
and environmental effects are enough for
poor and rural areas.
ICS and HRS implementation. However,
sensory and function factors are Methods
essential for acceptance and adoption of Study area
the respective technologies [58].
The study was conducted in six different
Familiar taste, texture, and appearance
communities (Mocha, Mulapane,
are the primary goals for the consumers
Nauouoro, Muranuco, Sewere and Muala
[59]. As found by Malakar et al. [60] and
Oripa) in Gurué district of central
Wang et al. [61], for example, people may
Mozambique, where the majority of
find food cooked with firewood in
population use TSF to prepare their
traditional stoves more tasty compared
meals (Fig. 4.11). The district borders of
to that cooked in improved cooking
the Republic of Malawi lay between the
stoves and fuels, since smoke may infuse
latitudes 15° 18′ 55.2ʺ S and 36° 54′ 45.6ʺ
food with distinctive flavor [59].
E. It occupies a total area of 5646 km2 and
Therefore, in this study, we focus on two
the population is estimated at 431,000
essential aspects. First, we attempt to
inhabitants, which corresponds to a
understand whether, given the local
population density of 76 individuals per
conditions and prevailing cooking
km2 [62, 63]. The main activities are
practices, the use of ICS combined with
subsistence agriculture and animal
HRS would result in improvements with
husbandry [63].
regard to fuel and time saving, and
adequate food storage temperatures.
Results 96

Figure 4.11: Study area

Performance of the ICS constructed—2 in each area—and were
located inside the kitchens of the local
Field testing is an important step toward
leaders’ houses. The stove design used in
acceptability of an ICS [32, 64]. We
this study was based on the ICS
conducted CCT between June and
implemented in Idifu village, Tanzania
December 2020, using both ICS and TSF
[65], as it is proven successful under field
in order to analyze how efficient ICS is
conditions. The design features of the ICS
compared to TSF (Fig. 4.12). A total of 12
are presented in Table 4.8.
non-portable mud-ICS was

A B

Figure 4.12: A - improved biomass cook stoves; B - three-stone stoves

Results 97

Table 4.8: Design features of the ICS

Construction features Dimensions
Stove height 40 cm
Length of the stove body 107 cm
Width of the stove body 56 cm
Wood entry slot and combustion chamber 12 X 12 cm
Diameter of the smaller saucepan 16 cm
Diameter of the bigger saucepan 26 cm
Diameter of the chimney 12.5 cm

ugali or nsima in other regions of East

We follow Bailis et al. [66] testing
Africa (e.g., Kenya and Tanzania). To
protocols on CCT to compare the
ensure that the tests were uniform, not
performance of ICS to TSF in
only did we use locally available fuels and
standardized cooking tasks. The cooking
pots, but the quantity of ingredients were
task consisted of cooking two types of
also equivalent to those regularly used in
food regularly consumed in the
a household, jointly agreed upon with the
community, namely, bean curry and
villagers during an initial village meeting
xima, a type of maize flour porridge
(Table 4.9).
locally consumed; this is also known as

Table 4.9: Food used in the CCT

Dish Ingredients Ingredients Weight (g)
Bean curry Beans 500
Carrots 64
Onion 43
Tomato 205
Water 3000
Xima Corn flour 800
Water 2750

ensure that any water remaining in the

Prior to the tests, the cooks (mainly
stove evaporates, so that the ICS were
female) who prepared the food were
dry enough to be tested. This was done to
trained and allowed to use ICS several
avoid potential bias, given the fact that
times (within one month) in order to gain
the individuals who normally cook in the
experience on how to operate the ICS and
Results 98

households did not have any experience
with ICS at the time of the experiment. To
ensure that the CCT protocol was
precisely followed, the tests
supported by an assistant,
interfered only to a minimum amount
and only to safeguard testing protocols.
The cooking tasks were identically
performed for both ICS and TSF. For both
ICS and TSF, the beans were cooked first,
then the maize flour porridge. The end of
the cooking tasks was defined as the
points in time when the beans could be
mashed easily between two fingers or
with a fork and when the xima had a
consistency in which one could stick a
knife and it could stay upright without
falling. Both definitions reflect daily
reality in the study villages.
The fuelwood consisted of a mix of five
species commonly used in
(Swartzia madagascariensis,
Julbernardia globiflora,
curatellifolia, Pterocarpus angolensis
and Uapaca kirkiana). Before the CCT,
wood was sun dried for 7 days to lower
the moisture content (MC), as commonly
practiced by the study area residents. A
digital Wood Hygrometer model MD-2G
was used to measure wood moisture
were content. It has two sensor pins at the top
who that were pushed into the wood to
determine the percentage value of the
water content. The average value of MC
was 12%. At the end of the cooking task,
the unburned wood and the leftovers
were removed and weighed directly.
After the CCT, measurements were used
to calculate performance indicators (cf.
Hafner et al. [67]). We calculated the total
fuelwood consumed (Δf) by the
difference between the final fuelwood
(𝑓𝑓 ) and the initial fuelwood (𝑓𝑖 ). W was
calculated by adding the type of
ingredient (Ci) used in grams. The total
cooking time (Δt) is the difference
between the final time (𝑡𝑓 ) and the initial
Gurué time (𝑡𝑖 ) of the cooking process. The
respective formulas are presented in
Parinari Table 4.10. A Mann–Whitney U test was
used to determine if there is significant
difference in mean fuelwood
consumption between TSF and ICS.

Table 4.10: Stove performance indicators

Indicators Formulas Equation no.
Total fuelwood consumed (g) 𝛥𝑓 = 𝑓𝑓 − 𝑓𝑖 (7)
𝑛

Weight of ingredients (g) 𝑊 = ∑ 𝑐𝑖 (8)

𝑖=1
Total cooking time (min) 𝛥𝑡 = 𝑡𝑓 − 𝑡𝑖 (9)
Results 99

stove to use the leftovers (LO) from the

Performance of heat retention
cooking process. The newly implemented
systems
system is a heat retention box (HRB),
Locally used HRS were tested against a
which can easily be manufactured at
newly implemented system, the heat
relatively low cost. It is insulated with
retention box (Fig. 4.13). There are two
layers of PE foam and aluminum foil
techniques used at the study site to retain
facing toward the hot cooking pot [48].
heat in the food. One technique is to cover
Each HRS was tested two times in each
the hot cooking pot with banana leaves
community and 20 to 21 households per
(BL) to keep food warm until serving
community were involved.
time, the other is keep the pot on the

A B C

Figure 4.13: Means of heat retention: A - leftover charcoal; B - heat retention box, C -
banana leaves
After the cooking sections, 300 g of beans multiple comparisons to compare the
and 500 g of xima were immediately different HRS performance as we wanted
transferred to each of the heat retention to examine whether there was a
systems. The initial food temperature difference between the mean of all
was measured, then for the next 6h, possible pairwise comparisons [68]. The
temperature changes were recorded data were normally distributed and
every 30 min [55, 57], using a presented equal variances according to
temperature data logger. The amount of Shapiro–Wilk and Bartlett’s tests
time the system can hold the food at (pvalue > 0.05).
above 60 °C was determined. Consumers’ preferences
The length of time before food Consumer acceptability tests were
temperature falls below 60 °C was the conducted [69, 70] using the method of
performance indicator used for HRS. We central location test [71]. We tested
used one-way analysis of variance acceptability of food from (1) only TSF,
(ANOVA) followed by Tukey test for
Results 100

(2) only ICS, (3) TSF and LO, (4) TSF and
BL, (5) TSF and HRB, (6) ICS and LO, (7)
ICS and BL, and (8) ICS and HRB. A total
of 122 participants were randomly
chosen for the tests. In two of the testing
areas, we had 21 participants each and in
the remaining 4 areas we had 20
participants.
Food items were kept in heat retention
systems until ready for serving, defined
as the absolute time between the end of
the cooking task and actual food
consumption, jointly defined with the
villagers. The consumer preference tests
included representatives of selected
households (preferably the heads of
households) who did not have prior
knowledge about which stove was used
to cook the food they were about to taste.
“the visual or tactile characteristics and
appearance of the food”.
To analyze the consumers’ preferences,
we computed frequency distribution
(percentages of responses) of the Likert
scale categories (dislike very much,
dislike, neither like nor dislike, like, like
very much). Given that we wanted to
investigate people’s preference of the
foods from different combinations of ICS
and HRS, Chi-square tests of
independence were used to examine the
extent to which the different types of ICS
and HRS influenced the respondents’
perceived food attributes.
Results
Performance of the ICS
The results in Table 4.11 are the average

The participants were asked to maintain values of fuel and time consumption for

some distance from each other and not to different types of food and stoves. These

communicate during the session. Every results indicate that, as compared to TSF,

participant tasted a portion of the cooked ICS consume less fuelwood and time

food and evaluated the acceptability of (pvalue < 0.05). TSF presented higher

food in terms of taste, texture, aroma, overall means for the calculated

color, and overall acceptability [72], variables, namely, total fuelwood

according to a 5-point Likert scale consumed and total cooking time (Δt).
(1 = dislike very much, 2 = dislike, ICS reduced time and fuelwood
3 = neither like nor dislike, 4 = like, consumption by 14% and 17% during
5 = like very much). Prior to tasting, all beans cooking, respectively. During
participants were explained the meaning maize flour porridge cooking process, the
of the food attributes, e.g., texture is the ICS saved 24% of time and 9% of
fuelwood.
Results 101

Table 4.11: Average values of fuel and time consumption for different types of food and
stoves (n=12).

Type of food Type of stove Δt (min) Δf (g)

ICS 77.50 7094.72
Beans TSF 89.72 8512.50
12.22a 1417.78a
Diff (TSF-ICS)
(14%) (17%)
ICS 20.61 4978.28
Maize flour porridge TSF 27.17 5448.89
TSF-ICS 6.56a (24%) 470.61a (9%)
a Differences are significant at a level of significance of 0.05

Performance of heat retention decrease is slower in the HRB when

systems compared to LO. For both meals, HRB
took 160 to 175 min to reach 60 °C,
Tests were carried out to measure the
whereas LO took less than 140 min to
time that food within the HRS took to lose
reach the temperature danger zone.
heat until it reached 60 °C. Two locally
Although the absolute values indicate
used HRS, namely, the use of leftovers
that BL was in an intermediate position,
(LO) and banana leaves (BL) were tested
BL are not statistically different from
against heat retention box (HRB) (Fig. 5).
both LO and HRB.
Table 4.12 shows that the temperature

Table 4.12: Average time (min) the food temperature took to reach 60 ̊C
Bean curry (n=12) Maize flour porridge (n=12)
Mean time (min) Std.Dev. Mean time (min) Std.Dev.
HRB and ICS 170a 24.5 160a 24.5
HRB and TSF 175a 22.6 175a 12.2
BL and ICS 145ab 12.2 140ab 15.5
BL and TSF 150ab 19.0 145ab 12.2
LO and ICS 115b 12.2 105b 16.4
LO and TSF 135b 25.1 120b 12.3
Means followed by the same letter (a or b) within a column are not significantly different by the Tukey-test at a level of significance
of 0.05

Consumers’ preferences some respondents reported a neutral

position (23 to 44% depending on the
The results presented in Figs. 4.14 and
type of stove or HRS). The Chi-square test
4.15 show that most of the consumers
of independence does not indicate a
“like” or “like very much” the food from
relationship between the type of stove or
any of the stoves or HRS. None of the
HRS and the preference for food
respondents said that they did not like
(pvalue > 0.05).
the tested food attributes. However,
Results 102

Taste Colour
80% 80%
70% 70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%

Neither like nor dislike Like or like very much Neither like nor dislike Like or like very much

Aroma
Texture
80%
80%
70%
70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%

Neither like nor dislike Like or like very much Neither like nor dislike Like or like very much

Overall acceptability
80%
70%
60%
50%
40%
30%
20%
10%
0%

Neither like nor dislike Like or like very much

Figure 4.14: Respondents preference for bean curry

Results 103

Taste Colour
80% 80%
70% 70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%

Neither like nor dislike Like or like very much Neither like nor dislike Like or like very much

Aroma Texture
80% 80%
70% 70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%

Neither like nor dislike Like or like very much Neither like nor dislike Like or like very much

Overall acceptability
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

Neither like nor dislike Like or like very much

Figure 4.15: Respondent’s preference for maize flour porridge

Discussion heat retention ability as compared to

We use CCT and heat retention tests to
estimate the performance of ICS and HRS.
The use of ICS resulted in time and fuel
savings. The use of HRB showed a better
locally used HRS.
Like KPT, CCT also has the advantage of
predicting stove performance during real
use [33]. Nevertheless, the variability in
the results is high as compared to WBT
Results
due to real world differences in user
behavior and types of fuelwood, etc. [73].
CCT is a relevant method because it has a
lower level of variability when compared
to KPT, while allowing a certain level of
repeatability given that it is performed
under controlled conditions. As such, the
results of the current study offer insights
on mud-ICS’s ability to reduce time and
fuel consumption as compared to TSF.
The ICS model used is the present study
is of easy construction and replication in
the conditions of the study site due to not
just its low costs, but also the easy
availability and handling of required
construction materials. However, despite
being statistically significant, the rate of
fuel savings found in this study is
relatively low when compared to those
found in other studies [34, 74, 75].
Negash et al. [75], for instance, report
fuel savings of 32% in a non-
transportable mud stove as compared to
the TSF, whereas we found fuel saving of
9% and 17%. This may be due to the
energetic properties of wood species
used in the study area [76] and the site
conditions [77]. In fact, properties such
as water content and density can
determine the amount of fuelwood
needed for a cooking task [33, 78]. High-
density solid fuels, for example, have
more mass per volume available for
104
burning [79]. Moreover, the cooks’
experience may also be another factor for
the low rate of fuel savings. Cooks use
less fuel and cook more quickly over time
[34], but in this study the cooks only had
about a month to become familiar with
the ICS; perhaps a longer time period
would have increased saving rates
regarding time and fuel used.
The average savings in cooking time
were 6.6 min and 12.2 min (14% and
24%) for beans and maze flour porridge,
respectively. This might show that the
performance is likely to be different
depending both on the food item and on
the quantity of food cooked [34]. In
addition, ICS performance also depends
on the number of pots. For a single pot,
TSF have higher efficiency over ICS while
ICS is superior to TSF when cooking with
two pots simultaneously [80]. It is
noteworthy that the stove used in this
study was designed to prevent direct
exposure to smoke, which is likely to lead
to an improvement in the indoor air
quality of the kitchen area [81, 82].
Food is not always consumed
immediately after cooking; thus, we
secondly evaluated HRS’s ability to keep
food above temperature danger zone
before consumption. The tests results
indicate that the loss of temperature
inside the cooking pot is more
Results
pronounced in LO, possibly due to the
lack of insulation of the pot against
exposure to wind. Thus, it can be
concluded that HRB results in less heat
losses during the lapse of time between
the food being ready and its actual
consumption because HRB is insulated
with layers of PE foam and aluminum foil,
which gives it a comparably better heat
retention capacity [83]. As a result, HRB
may have the ability to inhibit the
multiplication of microorganisms, which
ultimately increases food safety.
However, the results of this study were
lower compared to those found by
Kaushik [56], who found average heat
retention time of more than 300 min.
This was probably due to the type and
amount of food. While Kaushik [56]
tested the HRS with 300 g of rice and 990
ml of water, we used 800 g of corn flour
and 2750 ml of water. Moreover, in the
study context, we were unable to observe
and compare the microorganism growth
between the different HRS, which would
allow us to obtain evidence of the
effectiveness of HRS in preventing the
multiplication of microorganisms.
It is important to note that LO depends
very much on the amount of fuelwood
that remains after the cooking process;
therefore, it also depends on the type and
initial amount of fuel used for cooking. A
105
small initial amount of LO might
contribute to the rapid heat losses. LO
had the same efficiency as BL, but the
latter does not need fuelwood; hence, it
can be recommended as a replacement
for LO given its potential to reduce
fuelwood consumption. The use of BL as
insulation material is also not
statistically different from HRB.
Nevertheless, BL, as an organic material,
is easily degradable and must be
replaced frequently due to its poor
durability. HRB is made of a more
durable material (aluminum foil and
foam), making it more suitable for
domestic use. In the context of our study
area, people generally consume the
maize flour porridge while it is still hot
[84]. When it becomes cold, it is usually
discarded since its typical consistency
changes with cooling and its heating
becomes difficult. In general, as
suggested by Taulo et al. [46], many poor
people in rural areas may be very
unlikely to warm their food before
consumption, after it cools down.
Therefore, the use of HRB may also help
to reduce food waste.
This study also aims to test consumer
preferences to sensory food features.
Therefore, we also conducted food
preference tests to assess the
acceptability of different combinations of
Results 106

cooking systems (TSF and ICS) and HRS avoid this issue by clearly explaining the
(LO, HRB, BL). Our results show that survey items as suggested by Kulas et al.
most of the consumers “like” or “like very [92]. Studies demonstrate the reliability
much” the sensory food features of 5-point Likert scale as compared to
regardless of the stoves or HRS. Thus, the other approaches [93,94,95].
use of the newly implemented In future studies, evidence on reduction
technologies, ICS and HRB, does not of time for fuelwood collection and the
negatively affect food acceptance given health benefit of the stove design with
that textural properties of food are an regard to indoor air pollution reduction
important factor for adults’ food should be provided. In addition, the
acceptance [85] and the willingness to relationship between HRB and microbial
try [86]. As pointed out by Leng et al. growth needs further examination.
[58], a crucial aspect in modifying
Conclusion
cooking technology is to consider the
sensory characteristics of food, as this is We use CCT and heat retention tests to

essential for adoption of the respective estimate the performance of ICS and HRS.

technology. Ignoring the preferences of Additionally, we conducted consumers’

the potential users will most likely lead preference analysis to evaluate the

to an underutilization of the technologies acceptability of food from these

[87]. A study investigating the links technologies. The use of ICS resulted in

between food preferences and food fuel saving of 9% and 20% for maize

choices shows that sensory and porridge and beans, respectively. The

functional factors have a much stronger overall time consumption for cooking

impact on the selection of food than decreased by 14% (beans) and 24%

health and price [88]. In fact, sensory (maize porridge). The HRB took 160 to

perceptions often tend to be less 175 min to reach temperature danger

negotiable than other values, with these zone” (from 5 to 60 °C) in which

perceptions including taste, texture, pathogenic microorganisms grow

odor, or appearance [89]. Nevertheless, quickly, whereas LO took less than 140

the use of 5-point Likert scale may min and BL took 140 to 150 min. Thus,

promote social desirability bias [90], as the use of HRB showed a better heat

respondents may use the midpoint to retention ability as compared to LO. In

avoid selecting socially undesirable addition, the introduction of ICS and HRB

options [91]. In this study, we tried to did not statistically and significantly
Results 107

affect food acceptance. Therefore, the Competing interests: The authors

study concludes that although the rates declare no competing interests.
of fuel savings found in this study is Funding: The project is financially
relatively low compared to other studies, supported by the Federal Ministry of
ICS is a sustainable mean for saving Food and Agriculture (BMEL) based on
cooking time and fuel. HRB has a the decision of the Parliament of the
potential to maintain adequate food Federal Republic of Germany and
storage temperatures. Both ICS and HRB managed by the Federal Office for
present a superior performance and can Agriculture and Food (BLE).
easily be diffused for not affecting the
Authors' contributions: CEM
quality of food. These technologies can be
conceptualized and designed the work,
manufactured using locally available
collected the data, analysed and
materials and are suitable for energy-
interpreted the data and drafted the
poor communities. Further studies are
manuscript. JMH interpreted the data
needed to provide evidence on factors
and critically revised the manuscript. HH
like fuelwood collection time and indoor
conceptualized and designed the work
air pollution reduction. In addition, HRS
and critically revised the manuscript. GU
ability to reduce microbial growth need
interpreted the data and critically
further examination.
revised the manuscript. JM interpreted
Declarations the data and critically revised the
Ethics approval and consent to manuscript. CR conceptualized and
participate: The authors had permission designed the work and critically revised
from the local administrative authorities the manuscript. SS substantively revised
to conduct the study. Informed consent the manuscript. All authors read and
was obtained from all individual approved the final manuscript.
participants included in the study. Acknowledgments: The project is
Consent for publication: Not applicable financially supported by the Federal
Ministry of Food and Agriculture (BMEL)
Availability of data and materials: The
based on the decision of the Parliament
datasets used and analyzed during the
of the Federal Republic of Germany and
current study are available from the
managed by the Federal Office for
corresponding author on reasonable
Agriculture and Food (BLE). Sincere
request.
thanks to the Leibniz Centre for
Results 108

Agricultural Landscape Research (ZALF) 5. Pope DP, Mishra V, Thompson L, et al.

in Müncheberg, Germany, for their (2010) Risk of low birth weight and
logistical support in conducting the stillbirth associated with indoor air
research in Gurué district, central pollution from solid fuel use in
Mozambique. developing countries. Epidemiol Rev
32:70-81.
References
6. Smith KR (2006) Health impacts of
1. Nyambane A, Johnson FX, Romeu–
household fuelwood use in developing
Dalmau C, et al. (2020) Ethanol as a clean
countries. Unasylva 57:41-4.
cooking alternative in Sub-Saharan
Africa: Insights from sugarcane 7. Thakur M, Nuyts PAW, Boudewijns EA,
production and ethanol adoption sites in et al. (2018) Impact of improved
Malawi and Mozambique. In: Gasparatos cookstoves on women’s and child health
A, Naidoo M, Ahmed A, Karanja A, in low and middle income countries: a
Fukushi K, Saito O, Takeuchi K (eds) systematic review and meta-analysis.
Science for Sustainable Societies. Thorax 73:1026.
Sustainability Challenges in Sub-Saharan 8. Zorrilla-Miras P, Mahamane M,
Africa II. Springer, Singapore, pp 115-44. Metzger MJ, et al. (2018) Environmental
2. IEA (2017) Energy Access Outlook Conservation and Social Benefits of
2017. International Energy Agency, Charcoal Production in Mozambique.
Paris. Ecol Econ 144:100-11.

3. Chidumayo EN, Gumbo DJ (2013) The 9. Fisher S, Bellinger DC, Cropper ML, et
environmental impacts of charcoal al. (2021) Air pollution and development
production in tropical ecosystems of the in Africa: impacts on health, the
world: A synthesis. Energy Sustain Dev economy, and human capital. Lancet
17:86-94. Planet Health 5:e681-e8.

4. Chirwa PW, Adeyemi O (2019)
Deforestation in Africa: Implications on
Food and Nutritional Security. In: Leal
Filho W, Azul AM, Brandli L, Özuyar PG,
Wall T (eds). Zero Hunger. Springer
International Publishing, Cham, pp 1-15.
10. IHME (2022) GBD Compare. Institute
for Health Metrics and Evaluation,
Seattle.
https://vizhub.healthdata.org/gbd-
compare/. Accessed 19 May 2022
11. Farioli F, Dafrallah T (2012) Gender
Issues of Biomass Production and Use in
Results 109

Africa. In: Janssen R, Rutz D (eds). 17. Specht MJ, Pinto SRR, Albuquerque
Bioenergy for Sustainable Development UP, et al. (2015) Burning biodiversity:
in Africa. Springer Netherlands, Fuelwood harvesting causes forest
Dordrecht, pp 345-61. degradation in human-dominated
tropical landscapes. Glob Ecol Conserv
12. Adkins E, Oppelstrup K, Modi V
3:200-9.
(2012) Rural household energy
consumption in the millennium villages 18. Cuvilas CA, Jirjis R, Lucas C (2010)
in Sub-Saharan Africa. Energy Sustain Energy situation in Mozambique: A
Dev 16:249-59. review. Renew Sustain Energy Rev
14:2139-46.
13. Hoffmann HK. (2016) Bioenergy,
development and food security in Sub- 19. Clough L (2012) The improved
Saharan Africa. Berlin, Humboldt- cookstove sector in East Africa:
Universität zu Berlin, Experience from the developing energy
Lebenswissenschaftliche Fakultät. enterprise programme (DEEP). GVEP-
Global Village Energy Partnership
14. Matinga MN, Annegarn HJ, Clancy JS
International, London.
(2013) Healthcare provider views on the
health effects of biomass fuel collection 20. Honkalaskar VH, Bhandarkar UV,
and use in rural Eastern Cape, South Sohoni M (2013) Development of a fuel
Africa: An ethnographic study. Soc Sci efficient cookstove through a
Med 97:192-200. participatory bottom-up approach.
Energy Sustain Soc 3:16.
15. Kuusela O-P, Bowman MS, Amacher
GS, et al. (2020) Does infrastructure and 21. Sedighi M, Salarian H (2017) A
resource access matter for technical comprehensive review of technical
efficiency? An empirical analysis of aspects of biomass cookstoves. Renew
fishing and fuelwood collection in Sustain Energy Rev 70:656-65.
Mozambique. Environ Dev Sustain 22. Jagger P, Pedit J, Bittner A, et al.
22:1811-37. (2017) Fuel efficiency and air pollutant
16. Sedano F, Silva JA, Machoco R, et al. concentrations of wood-burning
(2016) The impact of charcoal improved cookstoves in Malawi:
production on forest degradation: a case Implications for scaling-up cookstove
study in Tete, Mozambique. Environ Res programs. Energy Sustain Dev 41:112-
Lett 11:094020. 20.
Results
23. Bardouille P (2012) From Gap to
Opportunity: Business Models forScaling
Up Energy Access. Washington, DC.
24. Fullerton DG, Bruce N, Gordon SB
(2008) Indoor air pollution from biomass
fuel smoke is a major health concern in
the developing world. Trans R Soc Trop
Med Hyg 102:843-51.
25. van Gemert F, de Jong C, Kirenga B, et
al. (2019) Effects and acceptability of
implementing improved cookstoves and
heaters to reduce household air
pollution: a FRESH AIR study. npj Prim
Care Respir Med 29:32.
26. O'Shaughnessy SM, Deasy MJ, Doyle
JV, et al. (2014) Field trial testing of an
electricity-producing portable biomass
cooking stove in rural Malawi. Energy
Sustain Dev 20:1-10.
27. Urmee T, Gyamfi S (2014) A review of
improved Cookstove technologies and
programs. Renew Sustain Energy Rev
33:625-35.
28. Burwen J, Levine DI (2012) A rapid
assessment randomized-controlled trial
of improved cookstoves in rural Ghana.
Energy Sustain Dev 16:328-38.
29. Bensch G, Peters J (2013) Alleviating
deforestation pressures? Impacts of
improved stove dissemination on
charcoal consumption in urban Senegal.
Land Econ 89:676-98.
110
30. Pailman W, de Groot J, Clifford M, et
al. (2018) Experiences with improved
cookstoves in Southern Africa. J Energy
South Afr 29:13-26.
31. Arora P, Das P, Jain S, et al. (2014) A
laboratory based comparative study of
Indian biomass cookstove testing
protocol and Water Boiling Test. Energy
Sustain Dev 21:81-8.
32. Adkins E, Tyler E, Wang J, et al. (2010)
Field testing and survey evaluation of
household biomass cookstoves in rural
sub-Saharan Africa. Energy Sustain Dev
14:172-85.
33. L'Orange C, DeFoort M, Willson B
(2012) Influence of testing parameters
on biomass stove performance and
development of an improved testing
protocol. Energy Sustain Dev 16:3-12.
34. Gebreegziabher Z, Beyene AD,
Bluffstone R, et al. (2018) Fuel savings,
cooking time and user satisfaction with
improved biomass cookstoves: Evidence
from controlled cooking tests in Ethiopia.
Resour Energy Econ 52:173-85.
35. Abeliotis K, Pakula C (2013) Reducing
health impacts of biomass burning for
cooking—the need for cookstove
performance testing. Energy Effic 6:585-
94.
36. Shankar AV, Quinn AK, Dickinson KL,
et al. (2020) Everybody stacks: Lessons
Results
from household energy case studies to
inform design principles for clean energy
transitions. Energy Policy 141:111468.
37. Baldwin DE (2012) Sous vide
cooking: A review. Int J Gastron Food Sci
1:15-30.
38. Aworh OC (2021) Food safety issues
in fresh produce supply chain with
particular reference to sub-Saharan
Africa. Food Control 123:107737.
39. Woldt M, Moy G (2015) Literature
review on effective food hygiene
interventions for households
developing countries. FHI 360/FANTA,
Washington, DC.
40. Paparella A (2020) Food Safety:
definitions and aspects. In: Al-Rub FA,
Shibhab P, Al-Rub SA, Pittia P, Paparella
A (eds). Food Safety Hazards. GAVIN
eBooks, Lisle, pp 1-4.
41. Grace D, Alonso S, Mutua F, et al.
(2018) Food safety investment expert
advice: Burkina Faso, Ethiopia, Nigeria.
International Livestock Research
Institute, Nairobi.
42. Linscott AJ (2011) Food-Borne
Illnesses. Clin Microbiol Newsl 33:41-5.
43. Ricci A, Martelli F, Razzano R, et al.
(2020) Service temperature
preservation approach for food safety:
Microbiological evaluation of
meals. Food Control 115:107297.
111
44. U.S.FDA (2017) Food Code. U.S.
Department of Health and Human
Services, Washington, D.C.
45. Castán Broto V, Baptista I, Kirshner J,
et al. (2018) Energy justice and
sustainability transitions in
Mozambique. Appl Energy 228:645-55.
46. Taulo S, Wetlesen A, Abrahamsen R,
et al. (2008) Microbiological hazard
identification and exposure assessment
of food prepared and served in rural
households of Lungwena, Malawi. Int J
Food Microbiol 125:111-6.
in
47. Abedin HA, Rosen AM (2011) A
critical review of thermochemical energy
storage systems. Open renew energy J
4:42-6.
48. Nienhuys S (2012) How to make a
heat retention box (HRB). HUYS ADVIES.
49. Bailey J, Esteves Miramont A, Quiroga
VN, et al. (2019) Efficient Cooking for
Montainous Areas: Development,
Assembly and Thermal Behavior of
Cylindrical Heat Retention Box. Sci
Technol 9:12-7.
50. Krishnan J, Gowtham M, Rahul K, et al.
(2018) An Organic Composite Phase
Change Material for Hot Food Storage: A
Review. Int J Mech Prod Eng Res Dev
8:457-70.
ready 51. Memon SA, Jaiswal MS, Jain Y, et al.
(2020) A comprehensive review and a
Results
systematic approach to enhance the
performance of improved cookstove
(ICS). J Therm Anal Calorim:1-11.
52. Suresh R, Singh V, Malik J, et al. (2016)
Evaluation of the performance of
improved biomass cooking stoves with
different solid biomass fuel types.
Biomass Bioenergy 95:27-34.
53. Mehetre SA, Panwar N, Sharma D, et
al. (2017) Improved biomass cookstoves
for sustainable development: A review.
Renew Sustain Energy Rev 73:672-87.
54. MacCarty N, Still D, Ogle D (2010)
Fuel use and emissions performance of
fifty cooking stoves in the laboratory and
related benchmarks of performance.
Energy Sustain Dev 14:161-71.
55. Tiffany HT, Chu (2013) The hot lunch
dilemma: evaluating heat retention
ability of insulated container with
macaroni and cheese. British Columbia
Institute of Technology.
56. Kaushik V (2010) Designing Fireless
Cooker of Indigenous Insulation Material
for Better Heat Retention. J Hum Ecol
30:99-104.
57. Estrada AV, Santos AJM (2019) Using
Diatomite to Enhance the Thermal
Storage of Lunchbox. Paper presented at
2019 IEEE Integrated STEM Education
Conference (ISEC), IEEE, Princeton, 16-
16 March 2019.
112
58. Leng G, Adan RA, Belot M, et al. (2017)
The determinants of food choice. Proc
Nutr Soc 76:316-27.
59. Khandelwal M, Hill ME, Greenough P,
et al. (2017) Why Have Improved Cook-
Stove Initiatives in India Failed? World
Dev 92:13-27.
60. Malakar Y, Greig C, van de Fliert E
(2018) Resistance in rejecting solid fuels:
Beyond availability and adoption in the
structural dominations of cooking
practices in rural India. Energy Res Soc
Sci 46:225-35.
61. Wang Y, Bailis R (2015) The
revolution from the kitchen: Social
processes of the removal of traditional
cookstoves in Himachal Pradesh, India.
Energy Sustain Dev 27:127-36.
62. Alfredo C. (2016) Controlo pós-
terapêutico da schistosomose e das
parasitoses intestinais e atualização da
situação malacológica nos distritos de
Quelimane e Gurué, na província da
Zambézia-Moçambique, Instituto de
Higiene e Medicina Tropical.
63. INE (2018) Folheto estatistico
distrital gurue 2018. Instituto Nacional
de Estatística, Quelimane.
64. Barnes DF, Openshaw K, Smith KR, et
al. (1993) The design and diffusion of
improved cooking stoves. World Bank
Res Obs 8:119-41.
Results
65. Uckert G, Hafner J, Graef F, et al.
(2017) Farmer innovation driven by
needs and understanding: building the
capacities of farmer groups for improved
cooking stove construction and
continued adaptation. Environ Res Lett
12:125001.
66. Bailis R, Ogle D, MacCarty N (2004)
Controlled cooking test (CCT). Household
Energy and Health Programme, Shell
Foundation.
67. Hafner J, Uckert G, Graef F, et al.
(2018) A quantitative performance
assessment of improved cooking stoves
and traditional three-stone-fire stoves
using a two-pot test design in Chamwino,
Dodoma, Tanzania. Environ Res Lett
13:025002.
68. Lee S, Lee DK (2018) What is the
proper way to apply the multiple
comparison test? Korean J Anesthesiol
71:353-60.
69. Tomlins K, Ndunguru G, Stambul K, et
al. (2007) Sensory evaluation and
consumer acceptability of pale-fleshed
and orange-fleshed sweetpotato by
school children and mothers with
preschool children. J Sci Food Agric
87:2436-46.
70. Tomlins KI, Manful JT, Larwer P, et al.
(2005) Urban consumer preferences and
sensory evaluation of locally produced
113
and imported rice in West Africa. Food
Qual Prefer 16:79-89.
71. Boutrolle I, Delarue J, Arranz D, et al.
(2007) Central location test vs. home use
test: Contrasting results depending on
product type. Food Qual Prefer 18:490-9.
72. Oyeyinka AT, Pillay K, Siwela M
(2017) Consumer awareness and
acceptability of bambara groundnut as a
protein source for use in complementary
foods in rural KwaZulu-Natal. S Afr J Clin
Nutr 30:87-92.
73. Robinson J, Ibraimo M, Pemberton-
Pigott C (2011) The uncontrolled cooking
test: measuring three-stone fire
performance in northern Mozambique.
Paper presented at the DUE (Domestic
Use of Energy) Conference, Cape
Peninsula University of Technology, Cape
Town, 11-13 April 2011.
74. Jetter JJ, Kariher P (2009) Solid-fuel
household cook stoves: Characterization
of performance and emissions. Biomass
Bioenergy 33:294-305.
75. Negash D, Abegaz A, Smith JU (2021)
Environmental and financial benefits of
improved cookstove technologies in the
central highlands of Ethiopia. Biomass
Bioenergy 150:106089.
76. Massuque J, De Assis MR, Trugilho PF
(2020) Characterization of Miombo
species used by rural communities as
Results 114

fuelwood in Northern Mozambique. 82. de la Sota C, Lumbreras J, Pérez N, et

Energy Sources A: Recovery Util Environ al. (2018) Indoor air pollution from
Eff:1-10. biomass cookstoves in rural Senegal.
Energy Sustain Dev 43:224-34.
77. Lee CM, Chandler C, Lazarus M, et al.
(2013) Assessing the Climate Impacts of 83. Abdullahi G, Muhamad R, Dzolkhifli O,
Cookstove Projects: Issues in Emissions et al. (2017) Disinfestation of cocoa
Accounting. carbon market; carbon beans infested with life stages of
accounting; household energy. Tribolium castaneum
Stockholm Environment Institute, (Herbst)(Coleoptera: Tenebrionidae)
Stockholm. using solar heat trapped in a cardboard
solar heater box. J Entomol Zool Stud
78. Ramos DC, Carneiro AdCO, Tangstad
5:226-33.
M, et al. (2019) Quality of Wood and
Charcoal from Eucalyptus Clones for 84. Wanjala W, Onyango A, Makayoto M,
Metallurgical Use. Floresta Ambiente 26. et al. (2016) Indigenous technical
knowledge and formulations of thick
79. Massuque J, De Assis MR, Loureiro
(ugali) and thin (uji) porridges consumed
BA, et al. (2021) Influence of lignin on
in Kenya. Afr J Food Sci 10:385-96.
wood carbonization and charcoal
properties of Miombo woodland native 85. Martins Y, Pliner P (2006) “Ugh!
species. Eur J Wood Wood Prod. That's disgusting!”: Identification of the
characteristics of foods underlying
80. Hafner JM, Uckert G, Hoffmann HK, et
rejections based on disgust. Appetite
al. (2020) Efficiency of Three-Stone Fire
46:75-85.
and Improved Cooking Stoves using on-
farm and off-farm fuels in semi-arid 86. Dovey TM, Aldridge VK, Dignan W, et
Tanzania. Energy Sustain Dev 59:199- al. (2012) Developmental differences in
207. sensory decision making involved in
deciding to try a novel fruit. Br J Health
81. Sharma D, Jain S (2019) Impact of
Psychol 17:258-72.
intervention of biomass cookstove
technologies and kitchen characteristics 87. Foster AD, Rosenzweig MR (2010)
on indoor air quality and human Microeconomics of technology adoption.
exposure in rural settings of India. Annu Rev Econ 2:395-424.
Environ Int 123:240-55. 88. Wądołowska L, Babicz-Zielińska E,
Czarnocińska J (2008) Food choice
Results 115

models and their relation with food agreeing nor disagreeing: Person and
preferences and eating frequency in the item contributors to middle category
Polish population: POFPRES study. Food endorsement intent on Likert personality
Policy 33:122-34. indicators. J Res Pers 47:254-62.

89. Furst T, Connors M, Bisogni CA, et al. 93. Kulas JT, Stachowski AA, Haynes BA
(1996) Food Choice: A Conceptual Model (2008) Middle Response Functioning in
of the Process. Appetite 26:247-66. Likert-responses to Personality Items. J
Bus Psychol 22:251-9.
90. Chyung SY, Roberts K, Swanson I, et
al. (2017) Evidence-Based Survey 94. Leung S-O (2011) A Comparison of
Design: The Use of a Midpoint on the Psychometric Properties and Normality
Likert Scale. Perform Improv 56:15-23. in 4-, 5-, 6-, and 11-Point Likert Scales. J
Soc Serv Res 37:412-21.
91. Raaijmakers QAW (2000)
Adolescents’ midpoint responses on 95. Adelson JL, McCoach DB (2010)
likert-type scale items: neual or missing Measuring the mathematical attitudes of
values? Int J Public Opin Res 12:209-17. elementary students: The effects of a 4-
point or 5-point Likert-type scale. Educ
92. Kulas JT, Stachowski AA (2013)
Psychol Meas 70:796-807.
Respondent rationale for neither
Results 116

4.4 Effect of passive solar drying on food security in rural Mozambique

Scientific Reports (2022), volume 12, Article number: 17154.

Custodio Matavel1,2,3*; Harald Kächele1; Jonathan Steinke1,2,; Constance Rybak1; Harry

Hoffmann1; João Salavessa4; Stefan Sieber1,2; Klaus Müller 1,2

1Leibniz Centre for Agricultural Landscape Research (ZALF), Eberswalder Str. 84, 15374 Müncheberg,
Germany

2Department of Agricultural Economics, Faculty of Life Sciences, Humboldt University of Berlin, Germany

3Faculty of Agrarian Sciences, Universidade Lúrio (Unilúrio), Campus Universitários de Unango Km 62,
Niassa, Sanga District, Mozambique

4Faculty of Health Sciences, Universidade Lúrio (Unilúrio), Nampula, Mozambique

© The Author(s) 2022. This article is licensed under a Creative Commons Attribution 4.0
International License. To view a copy of this licence, visit http:// creativecommons. org/
licenses/ by/4. 0/.

https://doi.org/10.1038/s41598-022-22129-9

Abstract

Achieving food security in Mozambique is critical, since 80% of the population cannot
afford an adequate diet. While increasing agricultural production is a necessary effort to
address this challenge, inadequate post-harvest treatment leads to storage losses and
quality degradation, with repercussions for food security. The use of solar drying is
promoted as a solution to provide efficient and reliable access to food preservation that
improves the food security situation in rural communities. However, there is a lack of
clear evidence on how the use or access to solar drying affects food security. This study
identifies the determinants of farmers’ choice to use solar drying and evaluates the effect
of a passive solar dryer on food security using survey data from 634 households. We
allocated solar dryers to selected communities and all interested individuals belonging to
these communities were eligible to use it. Propensity score matching and endogenous
switching poisson regression are used to estimate the average effect. The use of solar
Results 117

drying with associated training significantly increases the food security status of
participants by increasing household food availability, women’s dietary diversity, and
months of adequate household food provision and by decreasing the household food
insecurity access scale.

Keywords: solar energy; propensity score matching; endogenous switching regression;

Mepuagiua; Lioma.

Introduction to suffice for achieving long term FS in

Achieving food security (FS) is a complex
development challenge in Mozambique1.
According to World Food Programme,
80% of people in Mozambique cannot
afford an adequate diet — defined as a
diet that is rich in diverse foods
necessary to meet the nutritional needs
of an individual2,3 — and 42.3% of
children under the age of 5 are stunted4.
Despite efforts to increase agricultural
production5, high levels of post-harvest
losses (PHL) in Mozambique limit the
poor population’s access to quality food,
thereby exacerbating food insecurity6.
For example, PHL of maize result from
traditional storage practices, where
insects, rodents, and mold infection can
destroy up to 60% of the harvest7.
Increased productivity could potentially
mitigate PHL, but this remains
challenging for many farmers, who face
labor constraints and limited access to
modern agricultural inputs8. While
agriculture in Mozambique has ample
scope to increase its productivity, efforts
to increase farm output are thus unlikely
rural areas. Especially considering the
need for ecological efficiency, reducing
PHL seems a key feature of an integrated
FS strategy. Increasing rural residents’
access to locally viable postharvest
processing and preservation solutions
may therefore be a sustainable strategy
to improve FS9.
The introduction of appropriate
technologies to provide efficient and
reliable access to food preservation is
particularly critical for rural populations,
but current standard technologies
require access to electricity or fossil
fuels10. Solutions that require electricity
are more efficient in theory, but they are
not applicable in rural Mozambique,
where 67% of the population lives11, but
only 6% of the population has access to
electricity12. Appropriate technologies
are those that match the needs of users
with the available resources, within a
specific context13. Thus, in many remote
locations, such as in rural Mozambique, it
is crucial to provide farmers access to
locally adapted agro-processing
Results 118

technologies9 as well as affordable and proven. Solar drying, for example, results
sustainable energy solutions for in a product with comparably better
agricultural and domestic uses14, in order quality21, takes less time to finalize the
to achieve the main dimensions of FS. drying process22, reduces the
dependence on weather conditions15,
Scalable and affordable solutions that do
prevents harm to the product from
not require electricity or fossil fuels for
external factors such as rain, wind, dust,
food processing are already available.
and insects23, and reduces contamination
The use of solar energy for drying
by toxins24, among other advantages. In
agricultural produce has been proposed
light of the broad evidence on the
as a cost-effective and environmentally
technical performance of PSD,
sustainable solution to increase shelf-life,
development organizations and
minimize food and specific nutrient
governments already promote PSD to
losses and health risks, as well as add
contribute to improvements in FS25.
value to agricultural products15,16,17. As
Empirical evidence on the impacts of
opposed to refrigeration, which requires
these initiatives on the target group’s FS
a continuous supply of energy, it is
status is scarce, however. Nagwekar et
attractive since after initial drying, no
al.26 demonstrate that the use of solar
further equipment or energy input is
drying can significantly increase dietary
required to maintain product quality18.
diversity during the lean season due to
Nevertheless, the widely practiced
extended preservation period of diverse
method of drying agricultural products in
food, but the study did not investigate
rural Mozambique is the open-sun drying
effects on other aspects of FS. To date, the
(OSD)19. This method has significant
general lack of evidence limits the
limitations since it can lead to high
capacity of policy-makers and
product losses due to inadequate drying,
development practitioners concerned
fungal growth, as well as the
with rural FS to make informed decisions
encroachment of insects, birds, and
on investments and interventions. To be
rodents20.
able to prioritize locally suitable FS
Several types of devices that use solar
solutions, decision-makers need
energy to dry food products, the so called
evidence on the heterogeneous impacts
passive solar dryers (PSD), have already
of alternative intervention options,
been developed and tested, with their
including PSD. To fill this gap, this study
superiority over the often-used OSD
takes an experimental approach to
Results
analyze the effects of PSD on four
indicators of FS, reflecting the four
dimensions of FS in rural Mozambique.
By providing an integrated analysis of FS
effects across the four pillars, this study
goes beyond existing impact studies of
PSD. To inform future efforts toward
effective introduction and scaling of PSD
approaches by development
organizations, this study also aims to
identify factors that influence rural
households’ decisions to use a PSD. With
this, our analysis provides decision-
makers the evidence needed for deciding
on whether to implement
interventions, and how to maximize
adoption.
Methods
Study area
This study was conducted in Gurué
district, which is located at around 15°
South and 36° East, in Zambézia
Province, Central Mozambique between
December 2021 and January 2022. First,
Zambézia Province was selected for this
study due to its high levels of chronic
malnutrition53 and its frequent food
shortages54; thus, the need
interventions to improve the
situation. Secondly, Gurué district was
selected because it generates a large
surplus during harvest season55,56, but
faces a high risk losing it due to the
119
commonly used OSD method57. Thirdly,
two of the three administrative posts in
Gurué district, Lioma and Mepuagiua, are
also purposively selected due to their
characteristics in terms of low
urbanization rates and high proportions
of the population engaged in agricultural
activities. According to data provided by
the local authorities, Lioma is divided
into 29 communities and comprises
29,868 inhabitants, while Mepuagiua is
divided into 11 communities with 61,227
inhabitants. In this region, about 90% of
the population practices small-scale
PSD agriculture as the main occupation28 and
few households have access to
electricity58. The average farm size is less
than 2 ha and, generally, food insecure
households predominate during the pre-
harvest period; during the harvest
season there is a predominance of
households with medium and high food
security28. Local diets usually consists of
maize and cassava, both cooked as a
paste and served with beans or dark
green leaves sauces and/or dried or fresh
fish27.
Study design and sampling approach
for
FS In this study, we allocate a solar dryer to
50% of communities in each
administrative post (15 communities in
Lioma and 6 in Mepuagiua). The
communities were randomly selected
Results
and solar dryers were built and allocated
to local leaders in August 2020.
Allocation to communities (here defined
as clusters) was made since allocation to
individuals was not financially feasible59.
The solar dryer used, hereby wooden and
locally produced dryers that allow an
indirect drying of food, is already tested
and approved by local residents in the
study area. This solar dryer was intended
to be used during harvest periods, so
from implementation to the last data
collection two harvest periods were
observed. Therefore, individuals who
had utilized the dryers in at least one of
the two harvest periods were considered
as users. A detailed description of the
solar dryer is presented in previous
work22.
All interested individuals belonging to
the selected communities were eligible to
use the solar dryer. Likewise, they could
choose whether they wanted to use it, not
use it, leave or refuse to participate in the
study at any time. As such, in each
community, assignment into the
treatment group (users of the solar
dryer) was based on farmers’ self-
selection. Since farmers had no previous
experience with the implemented solar
dryer design, as this was the first
implementation, the choice to use was
voluntary and gradual. Therefore, we
120
only conducted a survey after
intervention, from December 2021 to
January 2022. This period coincides with
the period when the highest levels of
food insecurity are observed57. A total of
367 households had used solar dryers
since their construction between August
and September 2020, of which 308
(84%) agreed to be part of the study: 155
households were from Mepuagiua and
153 were from Lioma. Similarly, 350 of
the households that did not use the solar
dryer were randomly selected into the
control group (non-users), of which 326
(93%) agreed to be part of the study.
Thus, the total number of observations is
634. The survey captured household
socio-economic characteristics,
demographics, and four FS indicators to
capture the multidimensionality of FS. It
was performed in accordance with the
guidelines laid down in the ‘Declaration
of Helsinki’ and ethically reviewed by the
Mozambican National Committee of
Bioethics in Health (IRB00002657, Ref
370/CNBS/19). Informed consent was
obtained from all individuals who agreed
to participate in the study.
Conceptual framework
Figure 4.16 presents a conceptual
background summarizing the
hypothesized causal effects of using a
solar dryer on FS. A number of
Results 121

socioeconomic and demographic factors Food Availability Scores (FAS),

can influence farmer’s decision to use a Household Food Insecurity Access Scale
solar dryer. Likewise, individual (HFIAS), and Women Dietary Diversity
perceptions of the benefits of using a Score (WDDS). The use of a solar dryer
solar dryer and training can also can allow users to keep a higher share of
influence the choice to use it. their harvest at home instead of having to
Socioeconomic and demographic factors sell it in fear of losing it to pests and mold,
together with the use of the solar dryer thus contributing to stability across time.
can determine the physical availability of This temporal dimension of FS was
food, its physical access, economic captured via Months of Adequate
access, and intra-household food Household Food Provisioning (MAHFP).
allocation to individuals. These food Detailed description of each FS indicator
security dimensions were captured using is found in the following subsections.

Food Security

Availability Access Utilization Stability

(FAS) (HFIAS) (WDDS) (MAHFP)

Use of solar drying

Socioeconomic characteristics and demographics

Training Perceived benefits

Figure 4.16: Conceptual framework for causality between solar dryer use and food
security

Food availability scores (FAS) any food present in the house at time of
the survey60,61. A predefined list
To assess food availability, we used a
(Table 4.13), containing 46 foods items
Household Food Inventory (HFI), which
generally available and consumed in the
is based on participant self-reporting of
Results
study area, was used. This list was
generated following a baseline data
collection in February
Respondents were asked whether the
foods on the list were present in their
home or not and what was the source
(purchased, borrowed, own production
or donation/offered) of each respective
item. In case of any doubt,
respondents could consult someone they
Table 4.13: Household food inventory.
No. Food group
1 Starchy staples
2 Dark green leafy vegetables,
3 Other vitamin A rich fruits and
vegetable
4 Other fruits and vegetables,
5 Eggs
6 Organ meat
7 Meat and fish
8 Legumes, nuts and seeds
9 Milk and milk products
Household Food Insecurity Access Scale
(HFIAS)
The access dimension of FS was assessed
via Household Food Insecurity Access
Scale (HFIAS). This is a nine-item food
insecurity scale that uses a recall period
of four weeks to answer both occurrence
and frequency questions64.
respondents were first asked whether
the condition in the question occurred at
122
considered most likely to have the
answer or check the storage facilities or
2020. any other place where they were likely to
have the food. The FAS ranged from 0 to
46 and was calculated as the total sum of
foods available at the households. The
household was classified according to
their food availability as low (if FAS < 24)
the or high (if FAS ≥ 24), similar to Gichunge
et al.62 and Koui et al.63.
Food items
Cassava, bread, corn flour, maize, noodles, rice,
white potatoes, sorghum
amaranth, cassava leaves, cabbage, lettuce,
pumpkin leaves, sweet potato leaves
Carrots, sweet potatoes
Garlic, tomato, onion, banana, eggplant, okra
Eggs
liver, kidney, heart or other organ meats
Beef (cow), chicken, duck, pigeon, fish, goat,
sheep/lamb, pork
Beans, cowpeas, green grams, pigeon peas,
soybeans, groundnuts, coconuts, sunflower,
bambara nuts
Milk, yogurt
all in the past four weeks and, if the
answer was “yes”, they were
subsequently asked to answer whether
the condition happened rarely (once or
twice), sometimes (three to ten times), or
often (more than ten times) in the past
four weeks. The detailed questionnaire is
presented in Supplementary Table S1.
The
To ensure that the respondent
understood the meaning of interview
Results
questions, key informants (extension
officers and community leaders) were
asked to review and adapt the phrases
and definitions to local context. To
further refine the questions and gain
insights into whether the questions were
actually being understood as intended,
10 individuals who are not part of the
study sample were asked to answer the
questionnaire in order to allow the
interviewers to explore the respondent’s
own understanding of the question and
its meaning, according to the Key
Informant Interview Guide64. HFIAS
scores, ranging from 0 to 21, were
calculated for each household by
summing the codes for each
frequency-of-occurrence question. The
household were categorized into Food
Secure, Mildly Food Insecure, Moderately
Food Insecure, and Severely Food
Insecure, according to the Household
Food Insecurity Access Scale Indicator
Guide64.
Women dietary diversity score (WDDS)
Assessment of individual dietary intakes
within households allows for a more
accurate estimation of intra-household
food utilization65. Thus, in this study we
use individual dietary diversity scores
for women to capture food utilization.
These are good proxies for overall
123
dietary quality in a poor rural context
and are linked to nutritional status66.
One female respondent aged 15 years or
older was interviewed within selected
households. When there was more than
one eligible and available woman in the
household, the respondent was
randomly selected. The households that
were visited each day during the survey
were also randomly selected to minimize
day to day differences. We asked
questions about all foods they consumed
the previous day, inside and outside the
home. The food were grouped into 9
groups (see Table 4.13) and,
consequently, respondents were
assigned numbers ranging from 0 to 9
according to the number of food groups
they consumed. Women who consumed
less than 4 food groups were classified
into low dietary diversity, those who
consumed 4 to 5 food groups as medium
dietary diversity, and more than 6 as high
dietary diversity67. Appropriate wording
of the questions was agreed with local
leaders and extension officers, with the
food groups listing locally available foods
and locally recognized names for each.
Months of adequate household food
provisioning (MAHFP)
Months of adequate household food
provisioning (MAHFP) was used to
capture the stability of food consumption
Results 124

over time. Respondents were asked to
identify in which of the last 12 months all
household members could have at least
three meals a day. This allowed us to
assess food stability and classify
households into most food insecure (if
MAHFP ≤ 5), moderately food insecure
(6 ≤ MAHFP ≤ 9), or least food insecure
(10 ≤ MAHFP ≤ 12), as described in
previous work28.
Determinants of households’ choice to
use the solar dryer
A farmer is expected to use the solar
dryer if the benefits or utility gain of
using it outweighs that of not using it.
This utility, in terms of improving food
safety, can be expressed as a function the
observable vector of covariates (Z):
𝑆𝐷𝑈𝑖∗ = 𝛼𝑍𝑖 + 𝜇𝑖 , 𝑆𝐷𝑈𝑖 = 1[𝑆𝐷𝑈𝑖∗ > 0] (1)
𝑆𝐷𝑈𝑖 is a binary variable representing
use of the solar dryer by household 𝑖. 𝛼 is
a vector of parameters to be estimated, 𝑍𝑖
is a vector variables that are expected to
influence the use of solar drying, and 𝜇𝑖 is
the error term assumed to be normally
distributed (see Table 4.14 and Table
4.15). These variables are selected as
they are shown in previous studies to
potentially influence adoption of
agricultural technologies (e.g. Zhang et
al.30, Launio et al.68, Karki et al.69, Pollard
et al.70, Gitonga et al.71 and Hamza Conteh
et al.72).

Table 4.14: Descriptive statistics of surveyed households (Continuous/count variables)

Non-users (n=335) PSD Users (n=308)
Variable
Mean Std. Dev. Min Max Mean Std. Dev. Min Max
Age of household
34.46 13.55 18 83 52.72 16.56 18 82
head (years)
% of agricultural
0.59 0.19 0 0.96 0.69 0.19 0.09 0.99
output sold
Size of household 6.01 2.42 1 13 6.94 2.60 1 15
Size of land (ha) 1.45 1.12 0.2 7 2.77 1.76 0.25 7

Table 4.15: Descriptive statistics of surveyed households (dummy variables)

Variable % of Non-users % of Users
Lioma residents 47% 50%
Female household heads 40% 65%
Received training from project 44% 59%
OSD users 44% 50%
Received information from neighbor or relatives 49% 64%
Received information from extension 50% 52%
Belong to an association/cooperative 19% 43%
Beans farmers 16% 49%
Maize farmers 74% 54%
Sorghum farmers 48% 50%
Rice farmers 29% 52%
Cassava farmers 45% 52%
Results
Estimation of effects of solar dryers on
FS
To determine the FS effects of PSD, we
were interested in identifying what level
of FS would the PSD users have in its
absence by comparing the FS situation
between PSD users and non-users. Since
the decision to use the solar dryer is
based on farmers’ own perceptions about
the effectiveness and benefits of this
decision, the random allocation of
households to the users group was not
possible. In such case, self-selectivity
issues need to be addressed to minimize
biases on the impact estimations73.
Propensity score matching (PSM), along
with generalized propensity score and
instrumental variable, are econometric
approaches that address selection bias in
cross-sectional data74. Generalized
propensity score is applied in the case of
continuous treatment75. To implement a
valid instrumental variable analysis, it is
necessary to find a reliable instrument,
which is complicated and not always
possible76. Therefore, in this study, we
apply the PSM approach77 to assess the
impact of using a solar dryer on FS. It is
recommended in situations where self-
selection bias is an issue, since it
compares the difference between the
outcome variables of users and non-
users with similar inherent and
125
observable characteristics78. However,
one issue related to the use of PSM is that
unmeasured or unobserved
characteristics are likely correlated with
both the treatment and outcome
variables, which is likely to introduce
bias, thus understating or overstating the
program’s effect79. Thus, to complement
the propensity score matching and check
the robustness of the results, an
endogenous switching regression (ESR)
model is also employed.
Propensity score matching (PSM)
We first generated the propensity score
using a probit model. Second, the average
treatment effect on the treated (ATT),
based on the predicted propensity
scores, was estimated. PSM is suitable in
situation where the baseline data is not
available80. To check the robustness of
the results, we use three different
matching algorithms, namely nearest
neighbor matching, radius matching, and
kernel-based matching. PSM is expressed
as:
p(X i ) = Pr(𝑆𝐷𝑈 = 1|𝑋) = 𝐸(𝑆𝐷|𝑋) (2)
Where SDU is a dummy variable
indicating the use solar dryer. X
represents the factors that are expected
to influence the adoption of solar dryer
(see Tables 2 and 3). Two central
conditions of the PSM is the assumption
that observations with the same
Results 126

propensity score must have the same variables that can influence both
distribution of observable characteristics treatment and outcome variables. In this
independently of treatment status (Eq. 3) study, all the outcomes variables are the
and assignment to treatment is food security indicators resulting from
unconfounded (Eq. 4). the count of positive responses to the
questionnaire questions (see the
𝑆𝐷𝑈 ⊥ 𝑋| 𝑝(𝑋) (3)
Conseptual Framework). Thus, an
𝑌1 , 𝑌0 ⊥ 𝑆𝐷| 𝑋 (4)
Endogenous Switching Poisson
The user written program package
Regression (ESPR) approach was
PSMATCH2 in STATA81 was used to
adopted. The ATT was estimated using
estimate the propensity score.
the STATA command etpoisson, which
For impact evaluation, it is desirable to estimates the coefficient of an
estimate the average impact of solar endogenous binary treatment model
dryer adoption on PSD users, the so when the outcome is a count variable84,85.
called average effect of the treatment on The estimated model can be stated as
the treated (ATT)82, which can be follows:
estimated as follows:
𝐸(𝑌𝑗 |𝑋𝑖𝑗 , 𝑆𝐷𝑈𝑗 , ∈𝑗 ) = 𝑒𝑥𝑝(𝑋𝑖𝑗 𝛽𝑖 + 𝛼𝑆𝐷𝑈𝑗 +∈𝑗 (6)
𝐴𝑇𝑇 = 𝐸{𝑌1𝑖 − 𝑌0𝑖 |𝑆𝐷𝑈𝑖 = 1} =
𝐸[𝐸{𝑌1𝑖 − 𝑌0𝑖 |𝑆𝐷𝑈𝑖 = 1, 𝑝(𝑋𝑖 )}] = Where 𝑌𝑗 denotes the food security
𝐸[𝐸{𝑌1𝑖 |𝑆𝐷𝑈𝑖 = 1, 𝑝(𝑋𝑖 )} − status, 𝑋𝑖𝑗 represents the vector of the
𝐸{𝑌0𝑖 |𝑆𝐷𝑈𝑖 = 0, 𝑝(𝑋𝑖 )}|𝑆𝐷𝑈𝑖 = 1] (5) independent variables, and ∈𝑗 is the error
Where 𝜏 is the ATT, with 𝑌1𝑖 and 𝑌0𝑖 being term.
Results
the potential outcomes (FAS, HFIAS
score, MAHFP, and WDDS) in the two Food security status of PSD users and
counterfactual situations of using solar non-users
dryer and not using, respectively. The The household FS categories, classified
ATT was estimated for each outcome via Food Availability Scores (FAS),
variable separately. Women Dietary Diversity Score (WDDS),
Endogenous switching regression (ESR) Months of Adequate Household Food
model Provisioning (MAHFP), and Household
Food Insecurity Access Scale (HFIAS), are
The ESR model considers selectivity as
presented in Table 4.16. The majority of
an omitted variable problem83. Thus, it
solar dryer users (56%) has a high
was used to capture the differential
availability of food, whereas most non-
response taking into account unobserved
Results 127

users have low availability of food. Only
5% of non-users fall into the high
availability category. As for MAHFP, the
majority of users are moderately food
insecure (58%), while in the most food
insecure category the majority are non-
users. Nevertheless, we observed only a
small difference between the users and
non-users in the percentage of least food
insecure households. The classification
based on HFIAS revealed that 21% of
users and 53% of users fall within food
secure and mildly food secure categories,
respectively. The majority of non-users
(70%) are severely food insecure. As for
the WDDS, medium to high diversity is
observed among women belonging to
households that used the solar dryer,
whereas the majority of women (71%) in
the group of non-users are classified into
low dietary diversity. A Chi-square test
provides evidence of statistically
significant differences between users
and non-users in all FS indicators
(P < 0.01).

Table 4.16: Prevalence of food security

FS indicator Row Labels Non-users Users
High Availability 5% 56%
FAS
Low Availability 95% 44%
Least food insecure 6% 2%
MAHFP Moderately food insecure 29% 58%
Most food insecure 66% 40%
Food Secure 1% 21%
Mildly Food Insecure 8% 53%
HFIAS score
Moderately Food Insecure 21% 22%
Severely Food Insecure 70% 4%
High dietary diversity 7% 59%
WDDS Medium dietary diversity 22% 41%
Lowest dietary diversity 71% 0%

Determinants of households’ choice to The size of agricultural land and the

use the solar dryer percentage of agricultural sales are also
positively related to the use of PSD.
The results of probit regression present
Farmers who had training during project
the predicted likelihood of ever using
implementation or who had information
solar drying (Table 4.17). These results
from neighbors and those who belong to
indicate that households headed by
an association or cooperative are also
female or older farmers, with a larger
more likely to use the solar dryer.
family size, are more likely to use PSD.
Results 128

Production of staple foods also maize, beans, and rice increases the
influences adoption. The production of probability of adopting solar dryers.

Table 4.17: Determinants of predicted likelihood of using solar dryer (probit model
output)
Variable Coef. Std. Err.
Geographic Location (Lioma=1) -0.23 0.16
Gender of household head (Male =1) -0.28** 0.15
Age of household head (years) 0.04*** 0.00
% of agricultural output sold 1.82*** 0.36
Size of household 0.08*** 0.03
Size of land (ha) 0.29*** 0.05
Received training from project (Yes =1) 0.28** 0.13
Use drying methods (Yes=1) 0.15 0.13
Received information from neighbor (Yes=1) 0.36*** 0.13
Received info from extension (Yes=1) -0.03 0.13
Belong to an association/cooperative (Yes=1) 0.39*** 0.14
Produce beans (Yes=1) 0.62*** 0.14
Produce maize (Yes=1) -0.43*** 0.14
Produce Sorghum (Yes=1) 0.03 0.14
Produce Rice (Yes=1) 0.29** 0.13
Produce Cassava (Yes=1) -0.05 0.14
_cons -4.27*** 0.42
⁎⁎⁎
Significant at 1%; ⁎⁎ Significant at 5%; ⁎ Significant at 10%. The actual P value are presented at Supplementary Data S1

The test of assumptions indicate that the

The effect of using PSD on FS
balancing property of the propensity
Matching of PSD users and non-users was
score is satisfied. The distribution of
undertaken within a region of common
propensity scores and the region of
support in order to ensure that treated
common support is presented in
and control households are comparable
Fig. 4.17. A substantial overlap of the
in their covariates that predict use of
propensity scores of both groups can be
solar dryer (see Supplementary Data S1).
observed.
Results
Figure 4.17: Distribution of estimated propensity scores and region of common support
The difference in FS status between the
self-selected solar dryer users and non-
users is estimated using nearest
neighbor matching, radius matching, and
kernel-based matching. The Average
Treatment Effect on Treated (ATT)
estimates are presented in Table 4.18.
The use of solar dryer has a significant
positive effect on FAS and WDDS as well
as a significant negative effect on HFIAS.
The ATT for FAS is 9.57 in the nearest
neighbor matching, 14.02 in the radius
matching, and 9.38 in the kernel-based
matching. Furthermore, positive and
statistically significant ATT is found for
WDDS in all the matching methods. The
129
ATT in the nearest neighbor matching is
1.98, in the radius matching 2.74, and in
the kernel-based matching 1.96. This
results imply that the use of solar dryer
improves household food availability and
women’s dietary diversity. Furthermore,
on average, the HFIAS score of treated
households is 6.15 points lower than the
HFIAS score of matched control
households in the nearest neighbor
matching. In the radius and kernel-based
matching, the ATT is − 6.51 and − 6.54,
respectively. Nevertheless, according to
nearest neighbor and kernel matching,
the use of the solar dryer was not
sufficient to ensure that all household
Results 130

members had three or more meals a day matching, the use of the solar dryer has a
during the 12 months prior to the survey, positive and significant effect of 1.66.
as measured by MAHFP. In radius

Table 4.18: Effect of using solar dryer on food security based on PSM

Kernel-based
Nearest neighbor matching Radius matching
FS indicator matching
ATT Std. Err. ATT Std. Err. ATT Std. Err.
FAS 9.57*** 1.15 14.02*** 0.23 9.38*** 0.71
MAHFP 0.43 0.53 1.66*** 0.14 0.53 0.44
HFIAS score -6.15*** 0.76 -6.51*** 0.20 -6.54*** 0.64
WDDS 1.98*** 0.276 2.74*** 0.07 1.96*** 0.19
⁎⁎⁎
Significant at 1%; ⁎⁎
Significant at 5%; Significant at 10%. The actual P value are presented at Supplementary Data S1
⁎

The use of PSM does not account for was no presence of selection bias arising
unobservable factors. Thus, we from unobserved factors. However, for
estimated ATT from the endogenous- FAS and WDDS, the Wald test is
switching regression (ESR), which significant at 5% and 1%, respectively,
accounts for both observed and indicating possible biases from
unobserved factors. Based on the Wald unobserved factors (see Supplementary
test, the null hypothesis of no correlation Data S2 for detailed output results). The
between the treatment errors and the ATTs estimated from ESR are presented
outcome errors could not be rejected for in Table 4.19
HFIAS and MAHFP, implying that there

Table 4.19: Effect of using solar dryer on food security based on ESR
FS indicator ATT Std. Err.
FAS 14.31*** 0.46
MAHFP 2.3*** 0.65

HFIAS score -5.47** 2.14

WDDS 2.61*** 0.12

⁎⁎⁎
Significant at 1%; ⁎⁎ Significant at 5%; ⁎ Significant at 10%. The actual P value are presented at Supplementary Data S2
differences in the ATT. In general,
The ATT for FAS, MAHFP, and WDDS are
however, the use of the solar dryer
positive and statistically significant at
positively affected the FS status, which
1%, whereas the ATT for HFIAS score is
can be seen from the negative ATT for
negative and statistically significant at
HFIAS score and the positive ATT for FAS,
5%. These results are similar to those
WDDS, and MAHFP.
found in radius matching, despite small
Results 131

Discussion — may have been granted the right of

first access. Likewise, farm size might be
This study identifies the factors that
an indicator of the level of economic
influence rural households’ decisions to
resources owned by households34, thus,
use a passive solar dryer and investigates
farmers with larger farms may also be
the effect of using a passive solar dryer
more likely to adopt new technologies35.
on food security. It was conducted in a
Other studies also show a positive impact
drought-prone area27, where food
of farm size on agricultural technology
provision is dependent on rainfed
adoption36,37. Contrary to what is found
agriculture. This area has an improved FS
in previous studies38,39, female headed
during the harvest season28, thus, we
households are more likely to use the
focus our analysis on the period of
new technology than men. This is
greatest food shortage (immediately
consistent with the results in Table 4.15,
prior to the harvest).
in which 40% of non-users and 65% of
The results suggest that household
users were female. However, to capture
characteristics, such as age, gender, size
the actual gender effect, the dynamics of
of household, size of land, market
the decision-making process need
orientation, training, neighbor-to-
further examination by, for example,
neighbor communication, membership
accounting for the number male or
in an association, and the production of
female adults within the households39.
staple foods increase the likelihood of
Household size in terms of the number of
adopting solar drying. These findings are
household members, may be an indicator
consistent with previous research on the
of family labor availability40. In addition,
determinants of agricultural technology
some household members may be
adoption29,30,31,32,33.
engage in off-farm activities that provide
Older farmers are more experienced and, additional income34. As more resource-
therefore, may have greater awareness of endowed households have a better
the benefits of new technologies, as well ability to cope with production and price
as more resources to enable adoption33. risks, they are more willing to adopt new
Moreover, in the context of this study, technologies41.
one solar dryer per community may not
The importance of cooperatives in
have been enough for every interested
promoting adoption is already
farmer and the most powerful members
demonstrated30,31. Post-harvest
of the community — i.e. the older farmers
Results 132

technologies require more skills in technologies that allow them to reduce

operation and management, thus PHL and maintain the quality of their
demanding greater organizational products as long as possible. Therefore,
competence. Hence, members of a the probability of adopting the solar
cooperative can benefit from systematic dryer is higher for households producing
and frequent training as well as regular staple foods. Nonetheless, maize farmers
collaborative actions30. Farmers who sell were less likely to use the PSD than
their products are probably more others, probably because it is a very
interested in adopting post-harvest popular and important crop, with
technologies as these could help them to households already having significant
maintain the quality of their products experience with alternative post-harvest
and, consequently, their competitiveness management to reduce losses. Thus, they
in the market. Moreover, better storage may not have much interest in adopting a
quality also means that farmers do not new technology.
have to sell right immediately after The use of the solar dryer has, on one
harvest (when prices are low) but they
hand, the potential to alleviate seasonal
can keep their product and sell it at a food shortages and increase food shelf
later time, when prices are high. life, hence enhancing food availability
Neighbors can be the most influential and stability. On the other hand, it adds
source of information when choosing to value to agricultural products, thus
adopt an agricultural technology. Early increasing marketability and providing
technology adopters provide a better financial returns for farmers9,44.
community laboratory from which Household financial returns are an
neighbors can gain some experience42. important determinant of food access45.
Farmers without the required skills to Moreover, solar drying results in
operate a new technology may face some products with improved nutrient quality
difficulties to use it43. Thus, farmers who and hygiene46. PSM is employed to
got training on solar drying technology estimate the average treatment effect of
are more willing to adopt. using a solar dryer on FS. This technique
assumes that there is no selection bias
Staple foods are the biggest source of
arising from unobservable factors. The
food security in the region, thus
results suggest that using a solar dryer is
encouraging farmers who produce these
positively related to FS. Specifically,
foods to become interested in adopting
Results
there is strong evidence (significant at
the 1% level) of a positive FS effect,
measured by FAS, HFIAS, and WD.
However, evidence is mixed for MAHFP.
The nearest neighbor and kernel-based
matching did not show a statistically
significant effect of solar drying on
MAHFP, whereas in the radius matching,
the effect was positive and significant. It
is not uncommon for different matching
techniques to result in estimates of
different magnitudes. Except for MAHFP,
the PSM results are internally coherent
and in line with our expectations. The
type of PSM technique substantially
affects the magnitude of treatment effect
but none of the methods is a priori
superior to the others47. Thus, this study
also uses ESR to complement PSM and
check the robustness of the estimates.
ESR accounts for both unobservable and
observable factors, providing ATTs that
are closer to radius matching. This might
imply that nearest neighbor and kernel-
based matching underestimate the
average effect of solar drying on FS. In
general, however, both methods also
show that the use of a solar dryer positive
affects FS. This might be because the
study participants who chose to use the
solar dryer could protect themselves
from the unavailability and
inaccessibility of food in the lean months
thereby increasing their FAS and WDDS
133
and decreasing HFIAS score as compared
to non-users. This complies with results
of Nagwekar et al.26, who find that the use
of solar drying increases WDDS by 36%.
Although the impact of solar drying on
MAHFP is not significant in the nearest
neighbor and kernel-based matching, the
radius matching and ESR estimates
suggest that solar drying can provide
some FS stability by guaranteeing that
food is available over a 12 months period.
It should be noted that the 12 months
recall period also includes some months
before the intervention for those farmers
who decided to use the solar dryer
several months after the allocation.
Moreover, only a small difference (4%) is
observed between users and non-users
with regards to the percentage of least
food secure households (Table 4.16).
This could explain the non-significance in
the nearest neighbor and kernel-based
matching. Nevertheless, the use of PSD
did not show a clear effects on the
stability dimension, although the HFIAS
score also include some elements of
stability since it also analyses
respondents' perceptions of household
food insecurity experiences over a 30-
day period48.
One of the limitations of the solar dryer
implemented in the study area is that it
requires several rounds of drying to dry
Results 134

large quantities of food. In addition, since the FS outcome, such as the

it requires initial investment cost, the anthropometric measurements, may also
dryer was shared among households, provide more evidence on the impact of
which may not have allowed them to dry using solar drying on FS.
enough product for the entire lean Conclusion
period. This may discourage new users
This study demonstrates the potential of
from joining and continuing the use of
passive solar dryers to contribute to
the solar dryer. In fact, two of the main
improved food security in rural
limitations of PSD are the small amount
Mozambique. Robust evidence for
they dry per drying process and the
positive effects on all four pillars of food
intermittent nature of solar radiation
security suggests that promoting passive
availability49. Hence, there is a need to
solar dryers, along with trainings on their
develop solar dryers with relatively high
use, can be a promising intervention
efficiency that dry relatively large
strategy for improving the wellbeing of
quantities of product in the shortest time
relatively food-insecure subsistence
possible in order to minimize seasonal
farmers. This finding underscores that
food insecurity, such as the natural
food security and nutrition outcomes can
convection solar tunnel dryer and the
be improved with simple, locally
passive glass-roof type greenhouse
available technology and without the
dryer50. Moreover, the additional use of
need to intensify local production or food
adequate storage and packaging
imports, which may have negative
materials can improve moisture removal
environmental side-effects. To benefit
and thermal efficiency and prevent
large numbers of rural households,
eventual food loss during storage51,52.
however, massive introduction of
Further monitoring of food quality after
passive solar dryers, possibly in larger
the drying process is necessary. Studies
scale, would be needed. Future research
on the longevity and life of the dryer are
could evaluate the costs of such an
also necessary, since this can impact the
intervention, as well as the longevity and
feasibility and sustainability of such a
resilience of the drying devices over
technologies. FS can also affect the
extended use periods, to assess its cost–
financial returns for the farmers9,44.
benefit ratio against alternative food
Thus, we also suggest that future
security interventions. With adequate
research examines the impact of solar
support from government authorities or
drying on household income. The use of
Results 135

humanitarian organizations, the manuscript. KM substantively revised

construction and diffusion of passive the manuscript. All authors read and
solar dryers could possibly become an approved the final manuscript.
income generation model for rural youth Data availability statement: All data
and landless population. generated or analyzed during this study
are included in this published article
Acknowledgements: The project is
(and its Supplementary Information
financially supported by the Federal
files).
Ministry of Food and Agriculture (BMEL)
based on the decision of the Parliament Funding: Open Access funding enabled
of the Federal Republic of Germany and and organized by Projekt DEAL.
managed by the Federal Office for Competing interests: The authors
Agriculture and Food (BLE). Sincere
declare no competing interests.
thanks to the Leibniz Centre for
Supplementary Information: The
Agricultural Landscape Research (ZALF)
online version contains supplementary
in Müncheberg, Germany, for their
material available at
logistical support in conducting the
https://doi.org/10.1038/s41598-022-
research in Gurué district, central
22129-9.
Mozambique.
References
Author contributions: CM
conceptualized and designed the work, 1. Mabiso A, Cunguara B, Benfica R. Food

collected the data, analysed and (In)security and its drivers: insights from

interpreted the data and drafted the trends and opportunities in rural

manuscript. HK interpreted the data and Mozambique. Food Security.

critically revised the manuscript. JS 2014;6(5):649-70.

interpreted the data and critically 2. Adesuyi E. Attitude of Pregnant

revised the manuscript.
conceptualized and designed the work
and critically revised the manuscript. HH
conceptualized and designed the work
and critically revised the manuscript. JS
critically revised the manuscript. SS
conceptualized and designed the work
and substantively revised
CR Women Attending Ante-natal Clinic
towards Adequate Diet in the Prevention
of Anaemia in Selected Hospitals in
Osogbo. British Journal of Medicine and
Medical Research. 2016;14(12).
3. Bai Y, Herforth A, Masters WA. Global
variation in the cost of a nutrient-
the
Results
adequate diet by population group: an
observational study. The Lancet
Planetary Health. 2022;6(1):e19-e28.
4. WFP. Mozambique 2021 [cited 2021
02.03]. Available from:
https://www.wfp.org/countries/mozam
bique.
5. Chilundo M, de Sousa W, Christen EW,
Faduco J, Bjornlund H, Cheveia E, et al. Do
agricultural innovation platforms and
soil moisture and nutrient monitoring
tools improve the production and
livelihood of smallholder irrigators in
Mozambique? International Journal of
Water Resources Development.
2020;36(sup1):S127-S47.
6. Popat M, Griffith G, Mounter S, Cacho O.
Postharvest losses at the farm level and
its economy-wide costs: the case of the
maize sector in Mozambique. Agrekon.
2020;59(2):235-53.
7. Cugala D, Tostão E, Affognon H,
Mutungi C. Postharvest losses in Africa–
Analytical review and synthesis: the case
of Mozambique: ICIPE (International
Center of Insect Physiology and Ecology);
2012.
8. Ferrão J, Bell V, Cardoso LA, Fernandes
T. Agriculture and food security in
mozambique. Journal of Food, Nutrition
and Agriculture. 2018;1(1):7-11.
136
9. Adeyeye SAO. The role of food
processing and appropriate storage
technologies in ensuring food security
and food availability in Africa. Nutrition
& Food Science. 2017;47(1):122-39.
10. Marco C, Barbieri J, Riva F, Colombo E,
editors. Energy Technologies for Food
Utilization for Displaced People: from
identification to evaluation. 2016
UNESCO Chair Conference on
Technologies for Development: From
Innovation to Social Impact; 2016: EPFL,
Lausanne, Switzerland.
11. Dados Definitivos do IV
Recenseamento Geral da População e
Habitação, 2017 [Internet]. Instituto
Nacional de Estatística de Moçambique.
2019. Available from:
http://www.ine.gov.mz/iv-rgph-2017.
12. Salite D, Kirshner J, Cotton M, Howe L,
Cuamba B, Feijó J, et al. Electricity access
in Mozambique: A critical policy analysis
of investment, service reliability and
social sustainability. Energy Research &
Social Science. 2021;78:102123.
13. Murphy HM, McBean EA,
Farahbakhsh K. Appropriate technology
– A comprehensive approach for water
and sanitation in the developing world.
Technology in Society. 2009;31(2):158-
67.
Results 137

14. Chilundo RJ, Neves D, Mahanjane US. development through participatory

Photovoltaic water pumping systems for research. The Journal of Agricultural
horticultural crops irrigation: Education and Extension.
Advancements and opportunities 2018;24(3):285-99.
towards a green energy strategy for 20. Purohit P, Kumar A, Kandpal TC. Solar
Mozambique. Sustainable Energy drying vs. open sun drying: A framework
Technologies and Assessments. for financial evaluation. Solar Energy.
2019;33:61-8. 2006;80(12):1568-79.
15. Mustayen AGMB, Mekhilef S, Saidur R. 21. Sharma VK, Colangelo A, Spagna G.
Performance study of different solar Experimental investigation of different
dryers: A review. Renewable and solar dryers suitable for fruit and
Sustainable Energy Reviews. vegetable drying. Renewable Energy.
2014;34:463-70. 1995;6(4):413-24.
16. Tomar V, Tiwari GN, Norton B. Solar 22. Matavel CE, Hoffmann H, Rybak C,
dryers for tropical food preservation: Hafner JM, Salavessa J, Eshetu SB, et al.
Thermophysics of crops, systems and Experimental evaluation of a passive
components. Solar Energy. 2017;154:2- indirect solar dryer for agricultural
13. products in Central Mozambique. Journal
17. Eswara AR, Ramakrishnarao M. Solar of Food Processing and Preservation.
energy in food processing—a critical 2021;45(11):e15975.
appraisal. Journal of Food Science and 23. Sandali M, Boubekri A, Mennouche D,
Technology. 2013;50(2):209-27. Gherraf N. Improvement of a direct solar
18. Bradford KJ, Dahal P, Van Asbrouck J, dryer performance using a geothermal
Kunusoth K, Bello P, Thompson J, et al. water heat exchanger as supplementary
Chapter 17 - The dry chain: reducing energetic supply. An experimental
postharvest losses and improving food investigation and simulation study.
safety in humid climates. In: Kosseva MR, Renewable Energy. 2019;135:186-96.
Webb C, editors. Food Industry Wastes: 24. Vijayavenkataraman S, Iniyan S, Goic
Academic Press; 2020. p. 375-89. R. A review of solar drying technologies.
19. Otte PP, Bernardo R, Phinney R, Renewable and Sustainable Energy
Davidsson H, Tivana LD. Facilitating Reviews. 2012;16(5):2652-70.
integrated agricultural technology
Results 138

25. Udomkun P, Romuli S, Schock S, Conference Proceedings, Maputo,

Mahayothee B, Sartas M, Wossen T, et al. Mozambique, 10-13 October 2011; 2011:
Review of solar dryers for agricultural African Crop Science Society.
products in Asia and Africa: An 30. Zhang S, Sun Z, Ma W, Valentinov V.
innovation landscape approach. Journal The effect of cooperative membership on
of Environmental Management. agricultural technology adoption in
2020;268:110730. Sichuan, China. China Economic Review.
26. Nagwekar NN, Tidke VB, Thorat BN. 2020;62:101334.
Seasonal Nutritional Food Security to 31. Abebaw D, Haile MG. The impact of
Indian Women through Community-level cooperatives on agricultural technology
Implementation of Domestic Solar adoption: Empirical evidence from
Conduction Dryer. Ecology of Food and Ethiopia. Food Policy. 2013;38:82-91.
Nutrition. 2020;59(5):525-51.
32. Djibo O, Maman NM. Determinants of
27. Arimond M, Wiesmann D, Becquey E, agricultural technology adoption: Farm
Carriquiry A, Daniels M, Megan Deitchler, households evidence from Niger. Journal
et al. Dietary Diversity as a Measure of of Development and Agricultural
theMicronutrient Adequacy of Women’s Economics. 2019;11(1):15-23.
Diets in Resource-Poor Areas: Summary
33. Feyisa BW. Determinants of
of Results from Five Sites Washington,
agricultural technology adoption in
DC: USAID; 2011 [Available from:
Ethiopia: A meta-analysis. Cogent Food &
https://www.fantaproject.org/sites/def
Agriculture. 2020;6(1):1855817.
ault/files/resources/WDDP_Summary_R
eport_Jul2011.pdf. 34. Langyintuo AS, Mungoma C. The
effect of household wealth on the
28. Matavel C, Hoffmann H, Rybak C,
adoption of improved maize varieties in
Steinke J, Sieber S, Müller K.
Zambia. Food Policy. 2008;33(6):550-9.
Understanding the drivers of food
security among agriculture-based 35. Admassie A, Ayele G. Adoption of

households in Gurué District, Central improved technology in Ethiopia.

Mozambique. Agriculture & Food Ethiopian Journal of Economics.

Security. 2022;11(1):7. 2010;19(1):155-79.

29. Uaiene RN, editor Determinants of 36. Abebe Y, Bekele A. Analysis of

agricultural technology adoption in adoption spell of improved common

Mozambique. 10th African Crop Science bean varieties in the central rift valley of
Results 139

Ethiopia: A duration model approach. early 20th century agriculture.

Journal of Agricultural Economics and Explorations in Economic History.
Development. 2015;4(3):037-43. 2012;49(3):316-34.

37. Ahmed MH. Adoption of multiple 43. Challa M, Tilahun U. Determinants

agricultural technologies in maize and impacts of modern agricultural
production of the Central Rift Valley of technology adoption in west Wollega: the
Ethiopia. Studies in Agricultural case of Gulliso district. Journal of Biology,
Economics. 2015;117(1316-2016- Agriculture and Healthcare.
102848):162-8. 2014;4(20):63-77.

38. Doss CR, Morris ML. How does gender 44. Sharma A, Chen CR, Vu Lan N. Solar-
affect the adoption of agricultural energy drying systems: A review.
innovations? Agricultural Economics. Renewable and Sustainable Energy
2000;25(1):27-39. Reviews. 2009;13(6):1185-210.

39. Gebre GG, Isoda H, Rahut DB, 45. Briones Alonso E, Cockx L, Swinnen J.
Amekawa Y, Nomura H. Gender Culture and food security. Global Food
differences in the adoption of Security. 2018;17:113-27.
agricultural technology: The case of 46. Ukom A, Nwanagba N, Okereke D.
improved maize varieties in southern Effect of drying methods on the chemical
Ethiopia. Women's Studies International composition and anti-nutritional
Forum. 2019;76:102264. properties of a cocoyam (Xanthosoma
40. Noltze M, Schwarze S, Qaim M. Maffafa Schott) tuber flour and leaf
Understanding the adoption of system powder. Nutr Food Sci. 2020;2:197-203.
technologies in smallholder agriculture: 47. Baser O. Too Much Ado about
The system of rice intensification (SRI) in Propensity Score Models? Comparing
Timor Leste. Agricultural Systems. Methods of Propensity Score Matching.
2012;108:64-73. Value in Health. 2006;9(6):377-85.
41. Hardaker JB, Lien G, Anderson JR, 48. Pandey R, Bardsley DK. An
Huirne RB. Coping with risk in application of the Household Food
agriculture: Applied decision analysis: Insecurity Access Scale to assess food
Cabi; 2015. security in rural communities of Nepal.
42. Parman J. Good schools make good Asia & the Pacific Policy Studies.
neighbors: Human capital spillovers in 2019;6(2):130-50.
Results 140

49. Matavel C, Hoffmann H, Rybak C, 55. Alemu ZG, Van Schalkwyk HD. Market
Sieber S, Müller K, Brüntrup M, et al. integration in Mozambican maize
Passive solar dryers as sustainable markets: African Books Collective; 2008.
alternatives for drying agricultural 56. Norfolk S, Hanlon J, editors.
produce in sub-Saharan Africa: advances Confrontation between peasant
and challenges. Discover Sustainability. producers and investors in northern
2021;2(1):40. zambézia, Mozambique, in the context of
50. Patil R, Gawande R. A review on solar profit pressures on european investors.
tunnel greenhouse drying system. Annual World Bank Conference on Land
Renewable and Sustainable Energy and Poverty; 2012; Washington DC: The
Reviews. 2016;56:196-214. World Bank.

51. Natarajan K, Thokchom SS, Verma TN,
Nashine P. Convective solar drying of
Vitis vinifera & Momordica charantia
using thermal storage materials.
Renewable Energy. 2017;113:1193-200.
52. Giang NTN, Khai TV, Thuy NM. Effect
of thickness of polyethylene packaging
and temperature on quality of solar-
dried oyster mushroom (Pleurotus sajor-
caju). Plant Science Today.
2022;9(3):722-7.
53. Improving the Nutritional Status of
Children in Zambézia and Nampula
[press release]. Maputo: UNICEF2018.
54. Delgado L, Laborde D, Piñeiro V. Post-
harvest loss in Mozambique: Estimating
maize loss in Manica and Zambezia
provinces. Rome: Food and Agriculture
Organization of the United Nations; 2021.
57. Matavel C, Hoffmann H, Rybak C,
Sieber S, editors. Can Subsistence
Farming Help to Achieve Household
FoodSecurity? Evidence from Gurue,
Central Mozambique. Tropentag 2020:
Food and nutrition security and its
resilience to global crises; 2020.
Witzenhausen: DITSL; 2020.
58. MAEFP. Diagnóstico integrado de
infra-estruturas e serviços básicos para
os municípios da província de zambézia:
Municipio de Gurué: MAEFP; 2020
[Available from:
https://pdul.gov.mz/content/download
/207/1118/file/DIISB4_GURUE_Final.p
df.
59. Aggarwal R, Ranganathan P. Study
designs: Part 4 - Interventional studies.
Perspect Clin Res. 2019;10(3):137-9.
60. Fulkerson JA, Nelson MC, Lytle L, Moe
S, Heitzler C, Pasch KE. The validation of
Results 141

a home food inventory. International 66. Savy M, Martin-Prével Y, Sawadogo P,

Journal of Behavioral Nutrition and Kameli Y, Delpeuch F. Use of
Physical Activity. 2008;5(1):55. variety/diversity scores for diet quality
measurement: relation with nutritional
61. Beto JA, Sheth G, Rewers P. Assessing
status of women in a rural area in
Food Purchase Behavior Among Low-
Burkina Faso. European Journal of
Income Black and Hispanic Clients Using
Clinical Nutrition. 2005;59(5):703-16.
a Self-Reported Shelf Inventory. Journal
of the American Dietetic Association. 67. Kennedy G, Ballard T, Dop M.
1997;97(1):69-70. Guidelines for measuring household and
individual dietary diversity. Rome: FAO;
62. Gichunge C, Somerset S, Harris N.
2010.
Using a Household Food Inventory to
Assess the Availability of Traditional 68. Launio CC, Luis JS, Angeles YB.
Vegetables among Resettled African Factors influencing adoption of selected
Refugees. International Journal of peanut protection and production
Environmental Research and Public technologies in Northern Luzon,
Health. 2016;13(1):137. Philippines. Technology in Society.
2018;55:56-62.
63. Koui E, Jago R. Associations between
self-reported fruit and vegetable 69. Karki LB, Bauer S, editors.
consumption and home availability of Technology adoption and household
fruit and vegetables among Greek food security. Analyzing factors
primary-school children. Public Health determining technology adoption and
Nutrition. 2008;11(11):1142-8. impact of project intervention: A case of
smallholder peasants in Nepal.
64. Coates J, Swindale A, Bilinsky P.
Proceedings of Deutscher Tropentag
Household Food Insecurity Access Scale
Workshop; 2004: Citeseer.
(HFIAS) for measurement of food access:
indicator guide: version 3. 2007. 70. Pollard CM, Landrigan TJ, Ellies PL,
Kerr DA, Underwood Lester ML,
65. Jones AD, Ngure FM, Pelto G, Young
Goodchild SE. Geographic factors as
SL. What Are We Assessing When We
determinants of food security: a Western
Measure Food Security? A Compendium
Australian food pricing and quality study.
and Review of Current Metrics. Advances
Asia Pacific journal of clinical nutrition.
in Nutrition. 2013;4(5):481-505.
2014;23(4):703-13.
Results 142

71. Gitonga ZM, De Groote H, Kassie M, 77. Rosenbaum PR, Rubin DB. The central
Tefera T. Impact of metal silos on role of the propensity score in
households’ maize storage, storage observational studies for causal effects.
losses and food security: An application Biometrika. 1983;70(1):41-55.
of a propensity score matching. Food 78. Dehejia RH, Wahba S. Propensity
Policy. 2013;43:44-55. score-matching methods for
72. Hamza Conteh AM, Yan X, Moiwo JP. nonexperimental causal studies. Review
The determinants of grain storage of Economics and statistics.
technology adoption in Sierra Leone. Cah 2002;84(1):151-61.
Agric. 2015;24(1):47-55. 79. Gibson‐Davis Christina M, Foster EM.
73. Vrachioli M, Stefanou SE, Tzouvelekas A Cautionary Tale: Using Propensity
V. Impact Evaluation of Alternative Scores to Estimate the Effect of Food
Irrigation Technology in Crete: Stamps on Food Insecurity. Social Service
Correcting for Selectivity Bias. Review. 2006;80(1):93-126.
Environmental and Resource Economics. 80. Imbens GW, Wooldridge JM. Recent
2021;79(3):551-74. developments in the econometrics of
74. Shiferaw B, Kassie M, Jaleta M, Yirga program evaluation. Journal of economic
C. Adoption of improved wheat varieties literature. 2009;47(1):5-86.
and impacts on household food security 81. Leuven E, Sianesi B. PSMATCH2: Stata
in Ethiopia. Food Policy. 2014;44:272-84. module to perform full Mahalanobis and
75. Hirano K, Imbens GW. The propensity propensity score matching, common
score with continuous treatments. support graphing, and covariate
Applied Bayesian modeling and causal imbalance testing: Boston College
inference from incomplete-data Department of Economics; 2003.
perspectives. 2004;226164:73-84. 82. Gilligan DO, Hoddinott J. Is There
76. Laborde-Castérot H, Agrinier N, Thilly Persistence in the Impact of Emergency
N. Performing both propensity score and Food Aid? Evidence on Consumption,
instrumental variable analyses in Food Security, and Assets in Rural
observational studies often leads to Ethiopia. American Journal of
discrepant results: a systematic review. Agricultural Economics.
Journal of Clinical Epidemiology. 2007;89(2):225-42.
2015;68(10):1232-40.
Results 143

83. Heckman JJ. Sample Selection Bias as
a Specification Error. Econometrica.
1979;47(1):153-61.
84. Zimmer D. Using copulas to estimate
the coefficient of a binary endogenous
regressor in a Poisson regression:
Application to the effect of insurance on
doctor visits. Health Economics.
2018;27(3):545-56.
85. Cerulli G. Ivtreatreg: A Command for
Fitting Binary Treatment Models with
Heterogeneous Response to Treatment
and Unobservable Selection. The Stata
Journal. 2014;14(3):453-80.
Results 144

4.5 How to increase cookstove adoption? Exploring cost-effective dissemination

techniques in Central Mozambique

Energy Research & Social Science, Volume 100, Article 103082

Custodio Efraim Matavel1,2,3*; Harald Kächele1; Johannes M. Hafner2,4; Constance

Rybak1; Götz Uckert1; Harry Hoffmann1; Harison K. Kipkulei1; Jonas Massuque3;
Jonathan Steinke1,2; Stefan Sieber1,2

1Leibniz Centre for Agricultural Landscape Research (ZALF), Eberswalder Str. 84, 15374 Müncheberg,
Germany

2Department of Agricultural Economics, Faculty of Life Sciences, Humboldt University of Berlin, Germany

3Faculty of Agrarian Sciences, Lúrio University, Campus Universitários de Unango Km 62,

Sanga,Mozambique

4World Agroforestry (ICRAF), ICRAF Country Programme, TARI Mikocheni, P.O. Box 6226 Dar es Salaam,
United Republic of Tanzania

© 2023 Elsevier Ltd. All rights reserved. This article was published in Energy Research &
Social Science, Volume 100, Article 103082

https://doi.org/10.1016/j.erss.2023.103082

Abstract

The adoption and usage of improved cookstoves have been promoted in the Global South
to mitigate the adverse effects of relying on solid biomass for cooking. However, the
uptake and continued use of improved cookstoves remain low. Several behavioral
changing techniques have been used to increase the adoption of improved cooking stoves.
Still, the monetary costs involved in the dissemination process are generally neglected,
and few studies follow an experimental design to evaluate the impact of different stove
dissemination approaches. Thus, we conducted an experiment to explore the cost-
effectiveness of dissemination approaches and measure their effect on improved
cookstoves adoption in Central Mozambique. Forty communities were randomly assigned
to receive one of the following treatments: (1) individual video training, (2) on-site group
training or (3) a combination of both. In total, 620 households were randomly selected
Results 145

from the 40 communities to participate in the study and 510 were included in the final
analysis. The rates of adoption within each treatment were 53 % for individual video
training, 21 % for on-site group training and 47 % for the combination of both. The
monetary training costs were lower for individual video training, which stimulated the
early and sustained adoption of improved cook stoves. Our findings provide strong
evidence that the tool used for capacity building affects the adoption of improved
cookstoves. Extension agencies should be aware of cost-effectiveness and adjust
dissemination approaches to achieve the highest impact per unit of resource expenditure.

Keywords: traditional bioenergy; biomass energy; training; rural extension; cooking

demonstrations.

Introduction to utilise solid biomass more efficiently

[9]. Empirical evidence suggests that ICS
The rural areas in Mozambique, like in
can substantially reduce fuel use [10] and
most other countries in the Global South,
time for cooking [9], decrease the time
primarily depend on biomass energy.
for gathering wood [2], and the exposure
Approximately 95 % of the population
to indoor air pollution and respiratory
relies on solid fuels for cooking [1]. This
diseases [4], [11]. However, despite the
has been associated with deforestation
profound knowledge of ICS's benefits, its
[2], low energy efficiency [3], adverse
adoption and sustained use remains far
health and climate impacts [4], [5], as
from expectations [9], even when
well as women's time poverty [6]. In line
financial incentives such as lowering the
with these challenges, the United Nations
price of stoves are adopted [12], [13]. In
Sustainable Development Goal 7 calls for
many rural areas, there is free access to
universal access to affordable, reliable,
solid fuel such as tree branches or
sustainable, and modern energy by 2030
manure; thus, people may prefer to
[7].
spend their income on other priorities
It is argued that as opposed to facilitating
[14]. Moreover, individuals who decide
a transition to conventional forms of
to adopt a stove may later discontinue
energy, such as electricity, coal and gas,
the adoption [15]. Hanna et al. [16] and
which are not accessible to rural
Pine et al. [17] further suggest that after
communities [8], the use of improved
adoption, the intensity of use of an ICS
cookstoves (ICS) can enable households
may vary or decline over time. This
Results 146

undermines the achievement and the need for more effective methods for
sustainability of cookstove benefits. promoting sustained ICS adoption. In this
Therefore, facilitating the initial adoption study, we follow Troncoso et al. [28] and
of an ICS is a necessary but insufficient define the adoption of ICS with regard to
condition to achieve sustained usage the frequency of use. As such, we
[18]. consider adopters, the households using
the ICS at least two times per week.
Achieving sustained use of ICS may
require more than just the distribution of Multiple limitations of existing capacity-
stoves. It may also be necessary to building approaches have been identified
provide technical information on the in the past, requiring highly context-
stove use and comparative advantages of sensitive solutions. In particular, the
ICS relative to other stoves [19]. Jürisoo effectiveness of ICS promotion efforts
et al. [19] suggest the use of behavior can be limited by inadequate training
change techniques to support and approaches, an insufficient focus on
motivate users to develop new cooking opinion leaders, and budget constraints,
habits. Such techniques can include favouring rather superficial training
easily comprehensible manuals, approaches [29], [30], [31]. Despite the
educational initiatives, cooking vast body of academic literature testing
demonstrations, community meetings, the effectiveness of training approaches,
tutorials, and marketing or the costs associated with dissemination
communication campaigns [20], [21], are rarely considered in the ICS
[22], [23], [24], [25], [26]. In dissemination research. Moreover, there
Mozambique, several governmental and is a shortage of experimental designs to
non-governmental agencies have evaluate the impact of ICS interventions
implemented stove intervention [26].
programs in many areas using Budget constraints and cost-
techniques such as marketing campaigns, effectiveness of stove intervention are
demonstrations and training of important variables for selecting the
community-based organisations [4], dissemination approach [32]. For
[27]. However, a study by Pailman et al. example, compared to individual face-to-
[9] found that 60 % of ICS adopters still face approaches, demonstrations to
used traditional three-stone or groups are more effective in reaching out
unimproved cookstoves. This highlights to larger populations at a lower cost.
Results
Research has shown that group-based
approaches can effectively trigger the
purchase of ICS [19]. However, the
diversity of attendees can mean that not
everyone perceives or understands the
content equally. Lewis et al. [23] found
that individual and personalised
demonstrations are more effective in
disseminating clean cooking stoves. They
require high financial resources but tend
to result in higher uptake outcomes [23].
Thus, given the typical budget limits on
dissemination programs [33], locally
appropriate and cost-effective
approaches are still required to achieve
maximum impact per unit of resource
expenditure.
We argue that the dissemination tool is
also essential to enhance the cost-
effectiveness of ICS dissemination
programs. Word of mouth has been used
to trigger ICS adoption [26], [34], [35],
but other studies have suggested and
used short documentaries and videos to
promote ICS [32], [36], [37], [38]. The
literature in the field of agricultural
extension has also shown that a video can
cost-effectively reach the widest – even
geographically isolated – populations
[39], [40]. Unlike on-site demonstration,
the video does not require the facilitator
to have experience with the promoted
skills [40], [41]. This may, arguably,
147
minimize the costs required to hire
highly qualified specialists and organize
training activities.
Thus, we conducted an experiment in
Gurue district, Central Mozambique, in
which household members were trained
to construct, use and maintain an ICS.
The aim was to stimulate the adoption of
ICS and explore the cost-effectiveness of
an in-person demonstration for a group
and an individual video-based training.
With this, we intend to contribute to
informed decision-making and
investments by policymakers and
development stakeholders promoting
ICS in Mozambique.
We distinguish between the groups of
adopters since strategically targeting the
users with the highest adoption potential
can improve ICS dissemination [18].
Rogers [42] categorised adopters into
five groups based on when they adopt a
technology: innovators, early adopters,
early majority, late majority, and
laggards. Van den Bulte and Joshi [43]
further distinguish between independent
adopters and imitators. Latter's decision
to adopt is influenced by the former, and
both categories can be found during the
earlier or later stage of the adoption
process [43].
Results
Programs aiming to disseminate new
technologies must invest
dissemination techniques that encourage
early adoption from the highest number
of independent adopters since non-
adopters can also play a significant role
in the adoption process by acting as
opponents, thus, preventing others from
adopting [44]. In contrast, independent
early adopters are more likely to
positively influence those within the
social network [17], leading to the
highest diffusion outcomes.
The present study adds to the emerging
literature that includes the
effectiveness of stove interventions (see
Afridi et al. [32]). Specifically, we
demonstrate that the economic
perspective should be considered when
analyzing ICS adoption and related
promotion programmes rather than
merely assessing the effectiveness of the
training method.
Methods
Study design and sampling approach
The study was based on a participatory
action research methodology [45] and
followed an experimental approach since it
sought to analyse the causal relationship
between variables [46]. This approach is
also recommended when the aim is to
evaluate the impact of energy-related
behavior change programs and initiatives
148
[47]. We tested and introduced an ICS
in between June and December 2020 in two
localities, namely Lioma and Mepuagiua, in
Gurué district (Fig. 4.18). The ICS design –
based on an ICS implemented in Idifu
village, Tanzania [48] – can be constructed
using freely available materials. A detailed
description of its features can be found in
previous work [49]. The two areas were
purposely selected, within Gurué district,
based on similarities with regard to
household socio-economic characteristics
and lack of access to clean cooking fuels.
All communities belonging to the two
selected localities (29 in Lioma and 11 in
cost-
Mepuagiua) were included in the study.
Therefore, 40 communities were randomly
assigned to receive either group-based
training, individual training or a
combination of both. The description of
these approaches is presented in the next
subsections. In Lioma, we randomly
selected ten households in each community
and, in Mepuagiua, 30 households
participated in the training. Thus, we
selected 620 households (290 in Lioma and
330 in Mepuagiua). The difference in the
number of households selected in each
community was based on the population
size of the localities and the average number
of residents per community. Lioma has a
population of 29,868 inhabitants, and
Mepuagiua has 61,227 inhabitants.
Although this sample size was determined
Results 149

due to financial and logistical constraints, it Before the training sessions, meetings with
is within the minimum acceptable range for the local villagers were held to understand
a population size of <100,000 people within the status quo with regard to cooking energy
a precision level of +/− 5 % [46], [50]. needs and jointly agree on the stove design.

Figure 4.18: Study area location. (a): Africa, (b): Mozambique and (c) Study area.

dual households and 192 single

Since the training activities involved both
households. While in Mepuagiua, 149
the construction (mainly done by men)
and usage of the stove (mainly done by dual households were trained, and the
remaining 181 were single households.
women), for each sampled household,
From a total of 373 single households, 73
both male and female representatives
were represented by female residents.
were trained together whenever possible
Thus, a total of 320 women were trained
(dual households). Where that was not
from both single and dual households,
possible (mainly due to the unavailability
which represents 52 % of the
of one of the representatives), only a
households.
female or male representative was
trained from the household (single
households). In Lioma, we trained 98
Results 150

Group training approach on construction, maintenance and use of

ICS. The use of video for this approach
Local farmers were instructed on how to
was due to the fact that a video is a time-
construct, maintain and use the ICS. The
effective tool compared to on-site
training also included information on the
demonstrations. For instance, it required
advantages of using the ICS. Two
nearly 4 h to construct the ICS used in the
hundred and ten heads of selected
present study. The video was shown to
households attended the workshop in
220 households individually using
groups of ten people for three days. Two
laptops and android tablets. Local
groups were trained per day, and the first
enumerators who have attained at least
day was dedicated to training on
the 12th grade and with a good
theoretical aspects of clean cooking,
comprehension of Portuguese and the
associated advantages and introduction
local language in the study villages were
to the ICS design. The second training
recruited and trained for at least two
section (second day) mainly focused on
days on how to deliver ICS information
demonstrations and explanations of
using the video. The enumerators were
constructing the ICS. The third day was
instructed to show the video in a quiet
dedicated to demonstrations on how to
place where they were unlikely to be
use and maintain the ICS. An experienced
disturbed during the duration of the
trainer conducted training sections, and
screening of the videos to maintain
all explanations were translated into the
optimal concentration. Preferably, the
local language. The training module also
videos were shown in the farmer's home,
involved presentations, group
but there were cases in which the video
discussions, and exercises. An
was shown on the farm. A pre-test was
experienced male trainer performed the
conducted before the video
demonstrations on construction, and
demonstration to ensure that the
previously trained local female residents
enumerators followed the instructions
hosted the demonstrations on how to use
shown during the training.
and maintain the ICS. Pamphlets were
provided to participants at the end of the In this video, an example of the ICS
training for future reference. during a cooking process with two pots is
shown. Subsequently, the trainer
Individual training approach
introduces himself and explains in
The individual training consisted of an
Portuguese the advantages of using an
approximately 10 min video illustration
Results
ICS, then introduces the materials
necessary to construct the ICS, such as
clay or anthill soil, sand, sawdust, mud
bricks, banana stem, and water. The
trainer demonstrates and explains how
the sawdust and clay (or anthill soil) are
mixed in a volumetric ratio of 1:1 and the
addition of water to the mixture to make
it mouldable for blending. Next, the
trainer explains how to prepare the
stove's ground. The location needs to be
wet, and the blended mixture of sawdust
and clay (or anthill soil) is used to
construct a 2 cm high base for the stove.
Next, the foundation bricks on the 2 cm
high mixture are also laid down.
However, if bricks are unavailable, the
sawdust-clay mixture should be used.
The video continues with
demonstrations on how to create the
pans' spaces and finish the stove
construction. After the stove
construction, the trainer explained that
the stove takes up to 4 weeks to get
completely dry and should be protected
from intrusion and unfavourable
weather conditions. Subsequently,
instructions on how to use and maintain
the stove daily were provided. The
participant also received pamphlets for
future reference.
151
Combined training approach
The 190 participants selected to receive
the combination of training approaches
were first exposed to group training.
Subsequently, they received a visit to
their homes, in which the video was
shown as described above.
Data collection
After the training phase of the ICS, two
rounds of face-to-face interviews took
place. The households were surveyed to
estimate the influence of the training
approach, including other factors, on the
adoption of ICS. To minimize social
desirability bias and validate the
reported adoption rates, we explained
the confidentiality and anonymity
procedures at the beginning of the
interview. Whenever possible,
observations of the cooking tasks and the
stove condition were made [51].
Informed consent was obtained from all
individual participants in the study. The
first round was conducted during
January 2021 (a month after the last
training activities), and the follow-up
survey was conducted during January–
February 2022 (13–14 months after
receiving the training). In the first data
collection, 277 and 313 trained farmers
in Lioma and Mepuagiua agreed to
participate in the study. In the follow-up,
80 households (31 in Lioma and 49 in
Results 152

Mepuagiua) were unavailable for the technologies. Education levels influence

interviews. Thus, the study included 510
observations (246 households in Lioma
and 264 in Mepuagiua), of which 40 %
were single households with a male
representative. It is important to note
that most households that did not agree
to participate in the data collection phase
(83 %) were those represented only by
men during the training.
In addition to the training approach,
other variables, including those related
to the individual participating in the
study and the social system or network
surrounding the participants,
collected (Table 4.20 and Fig. 4.19).
These variables were based
previously published work on factors
influencing the adoption or rejection of
ICS (e.g. Mamuye et al. [52] and Jagger
and Jumbe [25]). Age was already found
relevant for technology adoption [49]
since youth are more enthusiastic about
experimenting with new things and
Table 4.20: Individual characteristics of the households (continuous variables)
knowledge about ICSs and clean fuels
[53]. Social interactions with adopters
may stimulate adoption through
stronger peer spill-over [54]. Households
where women are the decision-makers
are expected to be more likely to adopt
ICS [55]. The household size can
positively influence adoption since more
cooking wood is expected to be
consumed in larger households.
Therefore, the fuel efficiency associated
with the use of ICS might stimulate
adoption [17]. Previous experience may
influence the perception of the potential
were
adopters with regard to the benefits of
using ICS. In addition, the geographical
on
location was recorded to understand
whether the households residing in
Lioma are more likely to adopt ICS than
those in Mepuagiua since household
location is an important determinant of
adoption [56].

Variable Mean SD Min Max

Age (Years) 42.90 13.03 19 69
Years of education 5.07 2.85 0 16
Monthly income (MZN) 5075.93 3955.65 1000 43224
Size of the household 6.48 2.60 1 15
Results 153

60%
51%
50% 48%

40%

30%
25%
22%
20%

10%
2%
0%
Networks with ICS Association Female headed People who had Household located
adopters membership households previous in Lioma
experiences with
ICS

Figure 4.19: Individual characteristics of the households (categorical variables)

We obtained data on the costs associated implementation of training by the total
with stove construction material, number of participants. Furthermore, the
ingredients for the cooking relative cost per adopter was calculated
demonstrations, and the trainers' visit to by dividing the total cost by the number
the farmers, including salary, food and of participants who adopted the ICS.
transportation costs. The total costs for Data analysis
the group training included the trainer's
We cross-tabulated the number of
salary and expenses for food (coffee and
adopters by each adoption category to
lunch breaks), construction materials,
provide summaries of the adoption rate.
food ingredients (for cooking
Four categories of adopters were
demonstrations) and transportation. The
created: non-adopters, sustained
individual training had an additional cost
adopters, late adopters, and discontinued
for the video making; however, costs
adopters. Non-adopters were defined as
associated with the construction
households that did not have an ICS in
materials, ingredients and food for the
both rounds of data collection. Sustained
participants were eliminated. This data
adopters were defined as early adopters
was used to derive the cost per farmer for
who continued to use the ICS until the
a specific training approach, estimated
end of the follow-up survey period. Late
by dividing the total costs associated
adopters were households that did not
with the preparation and
Results 154

build the ICS immediately after the of ICS and the training approach while
training but used it during the follow-up controlling for geographic location,
period. Discontinued adopters were individual and social network
defined as early adopters (those who characteristics. Probit is a useful model
adopted immediately after the training) for analyzing panel data in which the
who were not using or did not have the dependent variable is a participant's
ICS during the follow-up period (12 to 13 binary choice, and the independent
months after the first data collection). It variable is randomly assigned by the
is noteworthy that ICS stove programs experimenter [58]. One of the
usually fail to completely replace the assumptions of this model is the
traditional three-stone fire stove due to a existence of an unobserved latent
variety of reasons, including the variable [59] that can be represented as:
additional effort necessary for ICS ∗
𝑌𝑖,𝑡 = 𝛼 + 𝛽𝑋𝑖,𝑡 + 𝜀𝑖,𝑡 + 𝑢𝑖 (1)
maintenance, the mismatch between the
∗
technical features and real-world Where 𝑌𝑖,𝑡 is the latent variable, 𝑋𝑖,𝑡 is

behavior of users and stove interference the vector of independent variables (see

in local cooking habits [16], [57]. Thus, in Table 4.20 and Figure 4.19), 𝛽 is the

this study, households were not required vector of coefficients, 𝜀𝑖,𝑡 and 𝑢𝑖 are the

to completely replace the three-stone fire idiosyncratic and individual-specific

stove to be considered ICS adopters. error terms, respectively. The observed

However, only the households who used binary dependent variable occurs when
∗
the ICS at least 2 to 3 times a week to 𝑌𝑖,𝑡 exceeds a given threshold [59] and

prepare their meals were classified as can be expressed as:

adopters. These are classified as medium ∗
1 𝑖𝑓 𝑌𝑖,𝑡 ≥0
𝑌𝑖,𝑡 = { ∗ (2)
adopters according to the categories 0 𝑖𝑓 𝑌𝑖,𝑡 < 0
proposed by Troncoso et al. [28]. A chi-
Our interest is to estimate the probability
square test of independence was applied
that a randomly chosen individual from
to detect the relationship between the
the population adopts an ICS conditional
type of ICS adopters and the training
on training approach while controlling
approach.
for the effects of geographic location and
Furthermore, a random effects probit individual and social network
model was applied to examine further characteristics. Group training was used
the relationships between the adoption as the base (or reference) category. We
Results
compared the effect of receiving
individual training and the combination
with group training. We estimated the
marginal effects of receiving a specific
training approach on the probability of
adopting ICS. The random effect probit
prediction of this probability is:
𝑃𝑟[𝑌𝑖,𝑡 = 1|𝑋𝑖,𝑡 ] = Φ (
𝛼+𝛽𝑋𝑖,𝑡
) (3)
√1+𝜎𝑢2
Where Φ is the standard normal
cumulative density function [57]. 𝑌𝑖,𝑡 is
the dependent variable that equals 1 if
the household adopted the ICS and 0
otherwise. Note that this variable is equal
to 0 and 1 for non-adopters and
sustained adopters, respectively,
regardless of the data collection period.
However, for late adopters, it equals 0 in
the first data collection and 1 in the
second. In contrast, for discontinued
adopters, it is equal to 1 in the first data
collection and 0 in the second.
Results
Adoption rates
Fig. 4.20 shows that the training method
has a strong influence on adoption as
well as on whether farmers sustained the
155
usage of ICS. The chi-square test of
independence shows an association
between the type of adopters and the
training approach (Pearson chi-square =
71.0080, p-value < 0.01). The overall
adoption rate during the follow-up
survey was nearly 41 %. Generally, the
highest rate of early adoption
(discontinued and sustained adoption)
was observed among those who received
individual training and the lowest among
group training. However, part of the
household who adopted the ICS
immediately following the individual and
group training discontinued the use
within 12 to 13 months of adoption. The
level of discontinued adoption was
relatively lower among the combination
group. About 30 % of the total sampled
households never adopted the ICS.
Notably, the proportion of late adopters
is almost the same for all approaches,
which was confirmed by the chi-square
independence test, which did not show
significant differences (p-value > 0.05).
This fosters the hypothesis that the
training approach and does not influence
the adoption rates in the long run.
Results 156

100%

90% 19%
30%
80% 39%

18%
70%
15%
60%
17%
50% 29%

40%
5%
38%
30%

20% 40%
34%
10% 17%

0%
Individual training Group training Combination
Non-adopters (n = 153) Sustained adopters (n = 124) Late adopters (n = 84) Discontinued adopters (n = 149)

Figure 4.20: Adoption rates of ICS

Effect of the training method on
Adoption
The probit regression model (Table 4.21)
shows that after controlling
geographic location and individual and
social network characteristics,
farmers trained by the
approach are predicted to be more likely
to adopt the ICS than those who received
group training (on-site demonstrations).
Compared to group training, combining
both approaches also increases the
probability of adoption. The marginal
effects suggest that individual training
increases the likelihood of adoption by
20 % compared to group training (Table
4.22). The likelihood of adopting ICS
increases by nearly 14 % when the
farmer receives the combination of
for
individual and group training
approaches. Household individual
the
characteristics that play a role in the
individual
adoption process are the household
head's age, the household's size and
location, and whether the farmer belongs
to a network with adopters. Age is
negatively related to adoption, whereas
the household size, location and having
relatives, friends or neighbours who
have adopted the ICS increases the
likelihood of being an adopter.
Results 157

Table 4.21: Random effect probit estimates of ICS adoption

Variables Coefficients Standard errors p-values
Training method
Individual 0.595053 0.110587 <0.001
Combination 0.413444 0.114955 <0.001
Household characteristics
Gender -0.14574 0.114264 0.202
Age -0.05521 0.007813 <0.001
Years of education -0.01582 0.014929 0.289
Previous experiences with ICS 0.250719 0.286366 0.381
Monthly income 4.15E-09 1.76E-05 1.000
Size of the household 0.052344 0.016938 0.002
Networks with adopters 0.67681 0.099549 <0.001
Association membership 0.061947 0.144268 0.668
Household location 1.012329 0.205212 <0.001
Constant (𝛼) 1.106116 0.405616 0.006
ln(𝜎𝑢2 ) -13.8511 19.10011
𝜎𝑢 0.000982 0.009382
rho (ρ) 9.65E-07 1.84E-05
Number of obs =1020 Number of groups = 510. Wald chi-square = 184.68, Prob > chi2 = 0.0000.

Table 4.22: Marginal effect of the training approach on the probability of ICS adoption

Training method Marginal effect Standard errors

Individual 0.202019*** 0.037128
Combination 0.139545*** 0.038436
***, **, * Significant at 1%, 5%, 10%, respectively

salary. The highest total costs were

Costs estimation
The cost per participant was estimated at observed in the combination of the two

1047.6 MZN for the group training (1 approaches. The cost per adopter also

USD = 63.8776 MZN), 625.0 MZN for indicated that individual group training

individual training and 1757.9 MZN for is the most cost-effective. In the group

the combination of both approaches training, each adopter (including both

(Table 4.23). The highest cost for all sustained and late adopters) costs

training approaches was the salary of the 5945.95 MZN (approximately 93 USD) on

trainers. Individual training had the average, while in the individual training

lowest cost, while the combination of approach, the cost per adopter is 1417.53

approaches had the highest cost for MZN (approximately 22 USD).

Results 158

Table 4.23: Cost per adopter for ICS dissemination approaches

Number
of
Cost per Cost per
Training Unit cost Total cost Number of adopters
Cost items participan adopter
method (MZN) (MZN) participant (sustaine
t (MZN) (MZN)
d and late
adopters)
Individual Video Making 25,000
137,500 220 625.0 97 1417.53
training Trainers’ Salary 112,500
Construction material 2,000
Food ingredients 3,000
Group
Trainers’ Salary 195,000 220,000 210 1047.6 37 5945.95
training
Trainers’ transportation 5,000
Coffee and lunch breaks 15,000
Construction material 1500
Food ingredients 3,000
combinati
Trainers’ Salary 315,500 334,000 190 1757.9 74 4513.51
on
Trainer transportation 4,000
Coffee and lunch breaks 10,000

Discussion were identified in this study: non-

adopters, sustained adopters, late
This study shows that the approach used
adopters and discontinued adopters (Fig.
to deliver information can affect the
4.20). A higher rate of sustained adoption
adoption of clean cooking technology. We
was observed in individual video
used two face-to-face approaches to
training. This study is in line with Lewis
inform local farmers about the benefits of
et al. [23], who argue that individual and
adopting an ICS and instruct them how to
personalised approaches are more
construct, utilise and maintain it. In one
effective in disseminating clean cooking
approach, we trained the participants in
stoves.
groups; in the other, the participants
were trained individually using a video. The group-based training took three
We also trained one group of participants days per group and was, therefore, more
by combining both approaches. As cost-intensive than the individual video
expected, individual video training training. Moreover, it required people to
resulted in higher adoption rates travel to the training site and could have
compared to in-person group training been regarded as an opportunity for
and a combination of the two entertainment and social gathering
approaches. rather than a training activity.
Nevertheless, this study supports the
Nevertheless, we observed that the
idea that the effectiveness of a training
adoption pattern varied over time, and
approach is context-dependent.
therefore, four categories of adopters
Results
The combination of training approaches
also resulted in high rates of adoption
when compared to group-based training.
The additional use of one-on-one
personalised visits might be responsible
for this result. The literature suggests
that the use of multiple strategies leads
to higher rates of adoption [20], [23]. In
this study, however, we noted that the
households who participated in both
group-based and individual training had
a relatively low adoption rate compared
to those who only participated in the
individual training. This is probably
because in the group training as well as
in the combination of both approaches,
the households were mostly represented
by male decision-makers, who are
generally not responsible for cooking
activities. On-site demonstrations
required both male and female
household representatives to move to
the training sites, and since it was not
mandatory to have both in the training
activities, most households preferred to
send only one of the representatives
(mostly male). This prevented women
from participating in the trainings,
leading to lower adoption rates. While in
the individual training in which there
was no need for moving to the training
site, we were mostly able to train both
female and male representatives
together.
159
As shown in Table 4.21, geographic
location affects the adoption rates.
Cultural factors might also play a role
[21]. Individual methods are widely used
by public extension in Mozambique [61]
and may have been regarded as formal
and, therefore, more persuasive than the
group method. Perhaps farmers feel
obliged to adopt the ICS by enumerator
visit in case of individual video training.
In the present study, the individual
training approach was time and cost-
effective since every visit took 25 to 30
min and did not require additional costs
with construction materials and food
provided to participants. The training
costs per participant and adopter were
lower when compared to the other
approaches. Thus, videos might be
presented to a much larger number of
farmers than on-site training. As shown
in Table 4.21, the combination of video
and on-site positively promoted
adoption, compared to the single use of
group training; however, it is the most
cost-intensive approach.
Notably, these results must be
interpreted and generalised with caution
since this study uses different
approaches (individual versus group)
and different tools (video versus on-site).
We argue that the training tool is also
essential in knowledge transfer. As Van
Results 160

Mele et al. [62] suggest, showing a video to influence those within the social
to farmers to transfer knowledge can network positively [18]. This can ensure
eliminate cultural barriers between the that the local communities accept the
trainer and the trainee. It combines new technology and prevent the negative
visual and verbal communication reputation hindering innovation
methods [63] and can more cost- diffusion [42]. Other factors such as
effectively reach the widest – even gender, education, monthly income,
geographically isolated – populations association membership and experiences
[39]. Several studies have also used with ICS are not significant predictors of
video-based approaches to increase adoption.
farmers' knowledge and awareness (e.g. The fact that connection to a network
[32], [36], [37], [38]). In addition, due to with adopters positively influences
its visual element, a video can easily be adoption and membership in an
understood by people without formal association does not, might seem
education, even if not in the local counterintuitive since social interactions
language [64]. Unlike on-site
across rural households stimulate
demonstration, the video also does not stronger peer spill-over [54]. However,
require the facilitator to have experience in our study, participants were only
with the promoted skills [40], [41]. asked if they belonged to any association
Our analyses further revealed that or cooperative, regardless of its purpose.
household characteristics are also crucial Therefore, we argue that the non-
for stove adoption. In particular, age, significance of this variable might be
household size, connection to a network because it did not necessarily mean
with adopters, and geographic location having contact with adopters or
affect the adoption. These characteristics discussing ICS. On the other hand, having
are particularly important for stove a network with adopters is more specific
promoters since they can use them to and indicates that the adopter had
target households with characteristics contact with someone who has already
likely to foster early adoption. By had ICS experience and may have been
strategically targeting households with influenced in some way. Nevertheless,
such characteristics, improved stove we could not distinguish between
programs would better allocate time and independent adopters and imitators,
resources since early adopters are likely which would allow us to understand
Results 161

further the former's influence on the households that agreed to

latter's behavior. Moreover, this study participate in this study, only men
did not capture the reasons for were trained. This is an important
discontinued usage that, among other issue to be considered since it might
variables, may also be due to influence by have played a role in the observed
other community members who have adoption rates. Thus, the
negative perceptions about the benefits generalization of this study's
of the stove. findings should be made with
caution, and future studies should be
We expected education to affect adoption
more specific on the target audience
significantly since education levels
to avoid introducing biases in the
influence the process of capturing
study results.
knowledge related to ICSs and clean fuels
2) Although we took some precautions,
[53]. However, the non-significance of
such as explaining the
education in this study might imply that
confidentiality and anonymity
rather than formal education, awareness
procedures at the beginning of the
creation is more important since the
interview and making observations
training method positively affected the
of the cooking tasks, the study may
adoption. The non-significant association
not be exempt from social
between gender and ICS adoption is also
desirability as we were unable to
counterintuitive since it was expected
validate the responses on adoption
that households where women are the
for the entire sample.
decision-makers, would be more likely to
3) Furthermore, in this study, we
adopt ICS [55]. Nevertheless, further
compare individual and group
understanding of the decision-making
training approaches using different
process is still necessary since gender
tools (video and on-site
differences in preferences within
demonstrations), which might make
households can also play a role [65].
these approaches not perfectly
Study limitations
comparable. Thus, we suggest that
The present study has three important future studies should compare
limitations, as described below: different approaches using the same

1) In our study area, the cooking tool. For example, they could study

activities are primarily done by the effect of using video to train a

women; however, in 40 % of the

Results
group of people versus training the
participants individually.
Conclusions
This study reveals that the training
approach plays a significant role in
promoting clean cooking technologies.
As expected, the individual approach
resulted in the highest adoption rates
and is cost- and time-effective compared
to the group training approach. The
training approach had no significant
effect on the adoption among late
adopters. The study concludes that the
capacity-building approach significantly
influences the effectiveness of improved
cookstoves dissemination programs. For
better financial resource efficiency of
stove promotion programs, the extension
agencies should be aware of cost-
effectiveness and adjust dissemination
techniques to achieve the highest impact
per unit of resource expenditure.
Funding
The project is supported by funds of the
Federal Ministry of Food and Agriculture
(BMEL) based on a decision of the
Parliament of the Federal Republic of
Germany via the Federal Office for
Agriculture and Food (BLE).
162
Declaration of competing interest
The authors declare the following
financial interests/personal
relationships which may be considered
as potential competing interests:
Custodio Efraim Matavel reports
financial support was provided by
Federal Agency for Agriculture and Food
(BLE).
Acknowledgements
Sincere thanks to the Leibniz Centre for
Agricultural Landscape Research (ZALF)
in Müncheberg, Germany, for their
logistical support in conducting the
research in Gurué district, Central
Mozambique. Furthermore, we thank the
farmers who participated in the survey.
References
[1] S. Bonjour, H. Adair-Rohani, J. Wolf,
N.G. Bruce, S. Mehta, A. Prüss-Ustün, et
al., Solid fuel use for household cooking:
country and regional estimates for 1980–
2010, Environ. Health Perspect. 121 (7)
(2013) 784–790.
[2] J. Burwen, D.I. Levine, A rapid
assessment randomized-controlled trial
of improved cookstoves in rural Ghana,
Energy Sustain. Dev. 16 (3) (2012) 328–
338.
[3] J. Jetter, Y. Zhao, K.R. Smith, B. Khan,
T. Yelverton, P. DeCarlo, et al., Pollutant
Results
emissions and energy efficiency under
controlled conditions for household
biomass cookstoves and implications for
metrics useful in setting international
test standards, Environ. Sci. Technol. 46
(19) (2012) 10827–10834.
[4] S.C. Anenberg, D.K. Henze, F. Lacey, A.
Irfan, P. Kinney, G. Kleiman, et al., Air
pollution-related health and climate
benefits of clean cookstove programs in
Mozambique, Environ. Res. Lett. 12 (2)
(2017), 025006.
[5] J. Barbieri, F. Riva, E. Colombo,
Cooking in refugee camps and informal
settlements: a review of available
technologies and impacts on the socio-
economic and environmental
perspective, Sustainable Energy Technol.
Assess. 22 (2017) 194–207.
[6] D. Arora, Gender differences in time-
poverty in rural Mozambique, Rev. Soc.
Econ. 73 (2) (2015) 196–221.
[7] IEA, IRENA, UNSD, World Bank, WHO,
Tracking SDG7: The Energy Progress
Report, World Bank, Washington DC,
2022.
[8] D. Salite, J. Kirshner, M. Cotton, L.
Howe, B. Cuamba, J. Feij ́o, et al.,
Electricity access in Mozambique: a
critical policy analysis of investment,
service reliability and social
163
sustainability, Energy Res. Soc. Sci. 78
(2021), 102123.
[9] W. Pailman, J. de Groot, M. Clifford, S.
Jewitt, C. Ray, Experiences with
improved cookstoves in southern Africa,
J. Energy South Afr. 29 (2018) 13–26.
[10] J. Hafner, G. Uckert, F. Graef, H.
Hoffmann, A. Kimaro, O. Sererya, et al., A
quantitative performance assessment of
improved cooking stoves and traditional
three-stone-fire stoves using a two-pot
test design in chamwino, Dodoma,
Tanzania, Environ. Res. Lett. 13 (2)
(2018), 025002.
[11] K.R. Smith, J.P. McCracken, L.
Thompson, R. Edwards, K.N. Shields, E.
Canuz, et al., Personal child and mother
carbon monoxide exposures and kitchen
levels: methods and results from a
randomized trial of woodfired chimney
cookstoves in Guatemala (RESPIRE), J.
Exposure Sci. Environ. Epidemiol. 20 (5)
(2010) 406–416.
[12] B. Barnes, J. Rosenbaum, S. Mehta,
K.N. Williams, K. Jagoe, J. Graham,
Behavior change communication: a key
ingredient for advancing clean cooking, J.
Health Commun. 20 (sup1) (2015) 3–5.
[13] G. Rosa, F. Majorin, S. Boisson, C.
Barstow, M. Johnson, M. Kirby, et al.,
Assessing the impact of water filters and
improved cook stoves on drinking water
Results 164

quality and household air pollution: a adoption of clean cookstoves in Kenya

randomised controlled trial in Rwanda, and Zambia, Energy Res. Soc. Sci. 39
PLOS ONE 9 (3) (2014), e91011. (2018) 164–176.

[14] A. Kar, H. Zerriffi, From cookstove [20] S.A. Lindgren, Clean cooking for all?
acquisition to cooking transition: A critical review of behavior, stakeholder
framing the behavioural aspects of engagement, and adoption for the global
cookstove interventions, Energy Res. Soc. diffusion of improved cookstoves, Energy
Sci. 42 (2018) 23–33.
Res. Soc. Sci. 68 (2020), 101539.
[15] Y. Alem, S. Hassen, G. K ̈ohlin, [21] W.M. Champion, P.H. Charley, B.
Adoption and disadoption of electric Klein, K. Stewart, P.A. Solomon, L.D.
cookstoves in urban Ethiopia: evidence Montoya, Perception, culture, and
from panel data, Resour. Energy Econ. 38
science: a framework to identify in-home
(2014) 110–124. heating options to improve indoor air
[16] R. Hanna, E. Duflo, M. Greenstone, Up quality in the Navajo nation, Sci. Total
in smoke: the influence of household Environ. 580 (2017) 297–306.
behavior on the long-run impact of [22] W.D. Evans, M. Johnson, K. Jagoe, D.
improved cooking stoves, Am. Econ. J. Charron, B.N. Young, A.S.M.M. Rahman, et
Econ. Pol. 8 (1) (2016) 80–114.
al., Evaluation of behavior change
[17] K. Pine, R. Edwards, O. Masera, A. communication campaigns to promote
Schilmann, A. Marr ́on-Mares, H. Riojas- modern cookstove purchase and use in
Rodríguez, Adoption and use of lower middle income countries, Int. J.
improved biomass stoves in rural Environ. Res. Public Health 15 (1) (2018)
Mexico, Energy Sustain. Dev. 15 (2) 11.
(2011) 176–183. [23] J.J. Lewis, V. Bhojvaid, N. Brooks, I.
[18] I. Ruiz-Mercado, O. Masera, H. Das, M.A. Jeuland, O. Patange, et al.,
Zamora, K.R. Smith, Adoption and Piloting improved cookstoves in India, J.
sustained use of improved cookstoves, Health Commun. 20 (sup1) (2015) 28–
Energy Policy 39 (12) (2011) 7557– 42.
7566. [24] A. Namagembe, N. Muller, L.M. Scott,
[19] M. Jürisoo, F. Lambe, M. Osborne, G. Zwisler, M. Johnson, J. Arney, et al.,
Beyond buying: the application of service
design methodology to understand
Results
Factors influencing the acquisition and
correct and consistent use of the top-lit
updraft cookstove in Uganda, J. Health
Commun. 20 (sup1) (2015) 76–83.
[25] P. Jagger, C. Jumbe, Stoves or sugar?
Willingness to adopt improved
cookstoves in Malawi, Energy Policy 92
(2016) 409–419.
[26] D.D. Furszyfer Del Rio, F. Lambe, J.
Roe, N. Matin, K.E. Makuch, M. Osborne,
Do we need better behaved cooks?
Reviewing behavioural change strategies
for improving the sustainability and
effectiveness of cookstove programs,
Energy Res. Soc. Sci. 70 (2020) 101788.
[27] FEDESMO, FEDESMO: sustainable
energy and development forum of
Mozambique, Available from:
https://naturvernforbundet.no/getfile.p
hp/1344431-
1354528791/Dokumenter/Div.%20vedl
egg%20til%20nettsaker/Fact%20sheet
%20on%20FEDESMO.pdf, 2022 18/10.
[28] K. Troncoso, C. Armend ́ariz, S.
Alatorre, Improved cook stove adoption
and impact assessment: a proposed
methodology, Energy Policy 62 (2013)
637–645.
[29] G. Miller, A.M. Mobarak, Intra-
household Externalities and Low
Demand for a New Technology:
Experimental Evidence on Improved
165
Cookstoves, 2011. Unpublished
manuscript.
[30] W. G ́omez, H. Salgado, F. V ́asquez, C.
Ch ́avez, Using stated preference
methods to design cost-effective subsidy
programs to induce technology adoption:
an application to a stove program in
southern Chile, J. Environ. Manag. 132
(2014) 346–357.
[31] R. Seguin, V.L. Flax, P. Jagger,
Barriers and facilitators to adoption and
use of fuel pellets and improved
cookstoves in urban Rwanda, PLOS ONE
13 (10) (2018), e0203775.
[32] F. Afridi, S. Debnath, E. Somanathan,
A breath of fresh air: raising awareness
for clean fuel adoption, J. Dev. Econ. 151
(2021), 102674.
[33] R. Bailis, V. Berrueta, C. Chengappa,
K. Dutta, R. Edwards, O. Masera, et al.,
Performance testing for monitoring
improved biomass stove interventions:
experiences of the household energy and
health project, Energy Sustain. Dev. 11
(2) (2007) 57–70.
[34] S. Ramirez, P. Dwivedi, A. Ghilardi, R.
Bailis, Diffusion of non-traditional
cookstoves across western Honduras: a
social network analysis, Energy Policy 66
(2014) 379–389.
Results 166

[35] A. Tigabu, Factors associated with agriculture platforms for nutrition: a case
sustained use of improved solid fuel study of a participatory, video-based
cookstoves: a case study from Kenya, agricultural extension platform in India,
Energy Sustain. Dev. 41 (2017) 81–87. PLOS ONE 11 (10) (2016), e0164002.
[36] E. Zossou, P. Van Mele, S.D. Vodouhe, [41] S.C. Kalichman, C. Cherry, F. Browne-
J. Wanvoeke, The power of video to Sperling, Effectiveness of a video-based
trigger innovation: rice processing in motivational skills-building HIV risk-
Central Benin, Int. J. Agric. Sustain. 7 (2)
reduction intervention for inner-city
(2009) 119–129. African american men, J. Consult. Clin.
[37] A. Ergeneman, Dissemination OF Psychol. 67 (6) (1999) 959–966.
Improved Cookstoves IN Rural Areas OF [42] E.M. Rogers, Diffusion of
THE Developing World:
Innovations, 5th ed, Free Press, New
Recommendations for the Eritrea York: New York, 2003.
Dissemination of Improved Stoves
[43] C. Van den Bulte, Y.V. Joshi, New
Program, Eritrea Energy Research and
product diffusion with influentials and
Training Center, 2003.
imitators, Mark. Sci. 26 (3) (2007) 400–
[38] C.W. Noonan, E.O. Semmens, D. 421.
Ware, P. Smith, B.B. Boyer, E. Erdei, et al.,
[44] H. Cavusoglu, N. Hu, Y. Li, D. Ma,
Wood stove interventions and child
Information technology diffusion with
respiratory infections in rural
influentials, imitators, and opponents, J.
communities: KidsAir rationale and
Manag. Inf. Syst. 27 (2) (2010) 305–334.
methods, Contemp. Clin. Trials 89
(2020), 105909. [45] P. Mapfumo, S. Adjei-Nsiah, F.
Mtambanengwe, R. Chikowo, K.E. Giller,
[39] J. Bello-Bravo, E. Abbott, S.
Mocumbe, R. Maria, R. Mazur, B.R. Participatory action research (PAR) as an

Pittendrigh, An 89% solution adoption entry point for supporting climate

rate at a two-year follow-up: evaluating change adaptation by smallholder

the effectiveness of an animated farmers in Africa, Environ. Dev. 5 (2013)

agricultural video approach, Inf. Technol. 6–22.

Dev. 26 (3) (2020) 577–590. [46] B.K. Sovacool, J. Axsen, S. Sorrell,

[40] S. Kadiyala, E.H. Morgan, S. Cyriac, A. Promoting novelty, rigor, and style in

Margolies, T. Roopnaraine, Adapting energy social science: towards codes of

Results
practice for appropriate methods and
research design, Energy Res. Soc. Sci. 45
(2018) 12–42.
[47] E.R. Frederiks, K. Stenner, E.V.
Hobman, M. Fischle, Evaluating energy
behavior change programs
randomized controlled trials:
practice guidelines for policymakers,
Energy Res. Soc. Sci. 22 (2016) 147–164.
[48] G. Uckert, J. Hafner, F. Graef, H.
Hoffmann, A. Kimaro, O. Sererya, et al.,
Farmer innovation driven by needs and
understanding: building the capacities of
farmer groups for improved cooking
stove construction and continued
adaptation, Environ. Res. Lett. 12 (12)
(2017), 125001.
[49] C.E. Matavel, J.M. Hafner, H.
Hoffmann, G. Uckert, J. Massuque, C.
Rybak, et al., Toward energy saving and
food safety in Central Mozambique: the
role of improved cook stoves and heat
retention boxes, Energy Sustain. Soc. 12
(1) (2022) 26.
[50] R.V. Krejcie, D.W. Morgan,
Determining sample size for research
activities, Educ. Psychol. Meas. 30 (3)
(1970) 607–610.
[51] N. Bergen, R. Labont ́e, “Everything
is perfect, and we have no Problems”:
detecting and limiting social desirability
167
bias in qualitative research, Qual. Health
Res. 30 (5) (2020) 783–792.
[52] F. Mamuye, B. Lemma, T.
Woldeamanuel, Emissions and fuel use
performance of two improved stoves and
using determinants of their adoption in Dodola,
best southeastern Ethiopia, Sustain. Environ.
Res. 28 (1) (2018) 32–38.
[53] F. Kapfudzaruwa, J. Fay, T. Hart,
Improved cookstoves in Africa:
explaining adoption patterns, Dev. South.
Afr. 34 (5) (2017) 548–563.
[54] H.-X. Wen, C. Wang, P.-Y. Nie,
Acceleration of rural households’
conversion to cleaner cooking fuels: the
importance and mechanisms of peer
effects, Energy Policy 154 (2021),
112301.
[55] C.F. Gould, J. Urpelainen, The
gendered nature of liquefied petroleum
gas stove adoption and use in rural India,
J. Dev. Stud. 56 (7) (2020) 1309–1329.
[56] A. Karanja, A. Gasparatos, Adoption
of improved biomass stoves in Kenya: a
transect-based approach in Kiambu and
Muranga counties, Environ. Res. Lett. 15
(2) (2020), 024020.
[57] G. Bensch, J. Peters, The intensive
margin of technology adoption –
experimental evidence on improved
cooking stoves in rural Senegal, J. Health
Econ. 42 (2015) 44–63.
Results 168

[58] J.R. Bland, A.C. Cook, Random effects barriers in technology-mediated rural
probit and logit: understanding learning, J. Agric. Educ. Ext. 16 (1) (2010)
predictions and marginal effects, Appl. 75–87.
Econ. Lett. 26 (2) (2019) 116–123. [63] S. David, C. Asamoah, Video as a tool
[59] R.D. Bock, Multivariate statistical for agricultural extension in Africa: a case
methods in behavioral research, in: study from Ghana, Int. J. Educ. Dev. ICT 7
Multivariate Statistical Methods in (1) (2011) 26–41.
Behavioral Research, McGraw-Hill, New
[64] G. Karubanga, P. Kibwika, F. Okry, H.
York, NY, US, 1975 xiii, 623-xiii, 623.
Sseguya, Empowering farmers to learn
[60] R.D. Gibbons, D. Hedeker, Random and innovate through integration of
effects probit and logistic regression video-mediated and face-to-face
models for three-level data, Biometrics
extension approaches: the case of rice
53 (4) (1997) 1527–1537. farmers in Uganda, Cogent Food Agric. 2
[61] B. Cunguara, K. Moder, Is (1) (2016), 1274944.
agricultural extension helping the poor? [65] G. Miller, A.M. Mobarak, Gender
Evidence from rural Mozambique, J. Afr. Differences in Preferences, Intra-
Econ. 20 (4) (2011) 562–595. household Externalities, and Low
[62] P. Van Mele, J. Wanvoeke, C. Akakpo, Demand for Improved Cookstoves,
R.M. Dacko, M. Ceesay, L. B ́eavogui, et al., National Bureau of Economic Research,
Videos bridging Asia and Africa: 2013.
overcoming cultural and institutional
 169

|
5 Synthesis

Integrating technological innovations to provide energy for food processing and

preparation has positive impacts on FNS. Energy is needed in all subsectors of food
systems, from production to food consumption. The presented research was based on a
participatory action research approach and assessed the potential of renewable energy
technologies for improving food and nutrition security in Central Mozambique. The
analysis focused on energy provision for food processing, cooking and storage. This
section returns to the five research hypotheses to synthesize the empirical evidence
presented in the results chapter.

5.1 Hypothesis 1: staple foods production plays a significant role in household FNS
situation

To test this hypothesis, the first subchapter of the results section presents the assessment
of the FS situation during the period immediately prior to the harvest and during the
harvest. The key drivers determining household capacity to achieve food security were
identified using the generalized ordered logit model with partial proportional odds model
(PPOM) for three FS indicators (MAHFP, HFCS, and HDDS). Multiple indicators were
applied to capture the multidimensionality of FNS. The results showed that FNS is
sensitive to seasonal variations since low dietary diversity and poor household food
consumption are predominant in the pre-harvest period. In contrast, during the harvest
season, there is a predominance of households with medium and high food consumption
and dietary diversity.

The production of the staple food crops (cassava, rice and maize) is significantly related
to at least one of the FS indicators. All the three crops are important sources of calories in
Mozambique. Cassava ranks first in terms of average daily caloric intake by households
(678 kilocalories per capita per day), followed by maize (478 kcal per capita per day) and
rice (166 kcal per capita per day) (FEWS NET, 2019). Maize has a positive relationship
with household food consumption score; rice is associated with months of adequate
household food provision and household dietary diversity, while cassava is positively
related to all three indicators. This implies that each crop impacts a specific dimension of
FNS. Thus, the study did not reject the hypothesis that staple foods play a significant role
Synthesis 170

in the household FNS situation. Nevertheless, other factors, such as gender, age of the
household head, geographic location, size and quality of land, livestock and crop diversity,
and cash crops, had a statistically significant effect on the different FNS indicators.

5.2 Hypothesis 2: solar drying technology increased the drying efficiency and
resulted in a high-quality product

In subchapter 4.2 of the results, the performance of passive solar dryers that were
constructed and tested against the traditional OSD for drying amaranth leaves and maize
grains is presented. As expected, the observed internal temperatures of the solar dryer
(between 27.6 and 48.2℃) were higher than the drying temperatures observed during
open-sun drying. The relative humidity was below 40%, a range most fungal species
cannot grow. In OSD, the crop was more susceptible to attack by fungi since the relative
humidity during the experiments was around 60%. The average thermal efficiency of the
solar dryer (19% for amaranth leaves and 14% for maize) was higher than the efficiency
of the OSD.

The study also applied consumer acceptance analysis to determine which drying method
resulted in a product with comparatively better sensory properties (taste, aroma, color
and overall acceptability). The results of a blind test, based on a 5-point Likert scale,
showed that both amaranth and maize from PISD were rated better than those from OSD.

5.3 Hypothesis 3: energy-efficient technologies reduce cooking time and household

energy consumption compared to traditional uses of biomass energy;

This hypothesis is tested in subchapter 4.3, in which the controlled cooking test (CCT) was
used to assess the efficiency of ICS compared to the traditional TSF. The CCT is a
recommended method for testing stove performance when the aim is to understand how
stoves perform with local foods, fuels, and cooking practices. It was developed as a mix of
two approaches: the laboratory-based water boiling test (WBT) and the field-based
kitchen performance test (KPT). The empirical results indicate that ICS results in energy
and time savings. The average savings in energy were 17% and 9%, whereas the cooking
time was reduced by 6.6 min and 12.2 min (14% and 24%) for cooking beans and maze
flour porridge, respectively.

Furthermore, the ability of HRB to keep food above the so-called temperature danger zone
before consumption (from 5 to 57 °C) was evaluated by monitoring temperature changes
inside the cooking pot. The test results showed that using a heat retention box (HRB) is
Synthesis 171

more advantageous than local methods of retaining heat inside the cooking pot. The loss
of temperature inside the cooking pot is more pronounced in the local techniques,
especially leftover firewood, possibly due to the lack of insulation of the pot against
exposure to wind. These findings have implications from the health perspective since they
show the potential of HRB in preventing foodborne illnesses (which, in turn, has
implications for food utilization). Moreover, reducing food waste is also a potential
advantage of HRB. Many poor people in rural areas may be very likely to discard their
maize flour porridge after it cools down due to its typical consistency changes with
cooling.

5.4 Hypothesis 4: the adoption of food processing technologies results in

improvements in household FNS situation.

In subchapter 4.4, the impact of a passive indirect solar dyer (PISD) on food security was
assessed. The study used inferential statistical procedures (propensity score matching
and Endogenous-Switching Regression Modeling) to compare the food security status of
households that used the PISD and those that did not. Propensity score matching was
selected to allow a reasonable interpretation of the causal effect since comparing the
outcome variable is undertaken between households with similar characteristics (Dehejia
and Wahba, 2002). In contrast, Endogenous-Switching Regression was used to
complement the propensity score matching and check the robustness of the result. Latter
considers selectivity as an omitted variable problem (Heckman, 1979). Thus, it was used
to capture the differential response taking into account unobserved variables that can
influence both treatment and outcome variables.

The results suggest that using the solar dryer enhanced food availability and dietary
diversity, alleviated seasonal food shortages and reduced food insecurity-related events
within households. These findings imply that implementing passive solar drying in rural
areas with associated training can help rural farmers improve their food and nutrition
security, probably because the study participants who chose to use the solar dryer could
protect themselves from the unavailability and inaccessibility of food in the lean months.
Therefore, a central recommendation of this paper is that policy interventions should
equip farmers with better knowledge and understanding of solar drying, emphasizing its
use as a pathway to ensure some level of food security.
Synthesis 172

5.5 Hypothesis 5: the training approach plays a significant role in stimulating the
early adoption of energy technologies.

This hypothesis was tested in the fifth subchapter of the results section featured in this
work. Different approaches were used with different groups of randomly selected farmers
to deliver information and train them on how to construct, maintain and use an ICS. Two
face-to-face approaches, namely, group-based and individual video training, were used
independently and in combination. The group training took three days per group, and it
was, therefore, more intensive, while in the individual training approach, every visit took
25 to 30 min. The results showed that the adoption pattern varied over time, and
therefore, four categories of adopters were identified: non-adopters, sustained adopters,
late adopters and discontinued adopters.

Moreover, the number of farmers in each category varied depending on the training
approach. The highest rate of non-adopters and discontinued adopters was observed
among participants who received group training. In contrast, a higher rate of sustained
adoption was observed in the individual video training. The individual training presented
the lowest training costs per participant and adopter compared to the other approaches.

This study also identified essential household characteristics for stove adoption: age, size
of household, connection to a network with adopters and geographic location. These
characteristics are particularly important for stove promoters since they can use them to
target households with characteristics likely to foster early adoption. By strategically
targeting households with such characteristics, improved stove programs would better
allocate time and resources since early adopters are likely to positively influence those
within their social network (Pine et al., 2011). This can ensure that the local communities
accept the new technology and prevent a negative reputation that can hinder innovation
diffusion (Rogers, 2003).
 173

|
6 Study limitations and further research
needs
This work aimed to capture the effect of improved renewable energy technologies usage
on FNS. However, the results are based on selected indicators. There are hundreds of food
security indicators, each with its advantages and disadvantages. In subchapters 4.1 and
4.5, this research employed indicators such as the household dietary diversity (HDDS),
women's dietary diversity (WDDS), food consumption score (FCS), household food
insecurity access scale (HFIAS), household food availability (HFA) and months of
adequate household food provision (MAHFP). The HFA indicates household food
availability. FCS and HDDS emphasise dietary diversity as a proxy for household food
access (Jones et al., 2013). MAHFP has a recall period of 12 months; thus, it is a valuable
tool for capturing food stability (Bilinsky and Swindale, 2007). The third domain of food
security, food utilization, has traditionally been captured by anthropometric
measurements, such as height, weight, mid-upper arm circumference, etc. These
measurements were not performed in this study. Despite our aim of capturing all four
dimensions of FNS, anthropometric measurements were not performed due to limiting
circumstances imposed by the COVID-19 pandemic. Therefore, future research needs to
combine self-reported consumption patterns with some FNS outcomes, i.e.,
anthropometric measures, to understand whether food is allocated equally to all
individual household members. It is important to note that, despite this limitation,
assessment of individual dietary intakes within households (e.g. WDDS) may allow for an
accurate estimation of food utilization (Jones et al., 2013). Moreover, HFIAS is positively
associated with dietary adequacy (Becquey et al., 2010), dietary diversity (Faber et al.,
2009), and lower odds of being underweight and wasting among HIV-positive adults
(Kadiyala and Rawat, 2013).

Subchapters 4.2 and 4.3 use consumer preference analysis to assess user`s acceptability
of implemented technologies based on a 5-point Likert scale, which may promote social
desirability bias (Chyung et al., 2017) since respondents may use the midpoint to avoid
selecting socially undesirable options (Raaijmakers, 2000). Moreover, factors such as
Study limitations and further research needs 174

demographic profile, health status, personality, knowledge, exposure, perceived quality

and mood can impact consumers’ perception and evaluation of a particular product.
Despite its sensitivity, the human sensory system cannot distinguish minor differences
between products (Steenkamp et al., 1994). These considerations need to be taken into
account, and we caution against generalizations of the present results for other case
studies.

Subchapter 4.3 is based on a CCT to determine fuel and time consumption during cooking.
This is arguably the best tool to monitor fuel and time consumption in the field under
standardized cooking situations since other tests, such as the WBT, do not provide the
actual field performance and result in unreliable predictions of the real field performance
of ICS (Lombardi et al., 2017). KPT gives the performance of ICS under real operating
conditions (Kipruto, 2011), but the uncontrolled conditions make the test results not
replicable. The tests for this study were performed in specific centres prepared for this
purpose, thus, not in real-world cooking situations. This might affect the results since
actual cooking behaviors might differ due to external effects, such as different weather
situations, varying moisture content of fuels, type of wood-based fuels used, and the
behaviors of cooks while cooking. Another aspect that would potentially influence the
result is that the cook did not have prior experience before the CCT sessions. Nevertheless,
attempts were made in this study to avoid potential bias by using locally available fuels,
using three repetitions and by using trained local cooks (mainly female), who were
allowed to use ICS several times (within one month) to gain experience on how to operate
the ICS.

Another limitation of this study is the lack of assessment of ICS and HRB’s impact on the
prevention and control of foodborne diseases. This kind of disease generally occurs if food
is incorrectly handled or cooked. Thus, it would be essential to compare the prevalence of
food-borne diseases among ICS and HRB users and non-users or to assess to microbial
load in the food stored in the HRB and the other HRS.

In subchapter 4.5, we discuss the importance of the training approach on farmers’

decision to adopt clean cooking. Nevertheless, a potential limitation of this study is that
we compare individual and group training approaches using different tools (video and on-
site demonstrations). Hence, these results must be interpreted and generalised with
caution. Ideally, further research should try to disaggregate the combinations by
Study limitations and further research needs 175

comparing different approaches using the same tool, for example, studying the effect of
using video to train a group of people versus training people individually.
 176

|
7 Overall Conclusion

In the course of this PhD work, several different aspects focusing on innovative renewable
energy technologies and their effects on FNS in Central Mozambique were addressed. In
addition, the impact of technology dissemination approaches was assessed. The outlined
research objectives were accomplished within the 5 subchapters of the results section.

A central conclusion is that energy efficient technologies such as the passive solar dryer,
improved cooking stove, and the heat retention box are central aspects of food and
nutrition security. The major reasons for this are the following:

a) Solar drying, if carried out correctly, allows poor rural households to store and
increase shelf life of food safely. It is a sustainable alternative to traditional food
preservation technologies (e.g., open-sun drying) that have several adverse effects
on food quality;
b) The improved cook stove is a sustainable means of saving cooking time and fuel.
Its use could prevent a household from applying strategies to cope with fuel
scarcity, such as reducing the number of meals or stopping cooking protein-rich
food that requires simmering. Furthermore, cooking is associated with the ability
of the human body to digest and absorb nutrients;
c) The heat retention box can potentially maintain adequate food storage
temperatures, thus, preventing foodborne diseases (that could hinder the
nutritional status of individuals).

An investigation of the FNS situation and its drivers showed very critical levels of food
insecurity among households during the pre-harvest season and relatively high levels of
food security during the harvest season. Nevertheless, mixed outcomes were found,
mainly due to different FNS indicators. This implies that each indicator captures different
dimensions of FNS. Thus, food security programs should choose carefully which indicator
to use since some indicators are more related to the consumption of farm-sourced food
(e.g. MAHFP). In contrast, others respond more strongly to purchased food (e.g. FCS). The
combination of different food security indicators allows a holistic assessment of the food
security status.
Overall Conclusion 177

Compared to OSD, the performance assessment of a PISD provided evidence that the use
of passive indirect solar drying results in higher drying efficiency. The sensory quality
attributes evaluation showed that passive indirect solar dryer outperformed OSD in terms
of texture, aroma, color and overall consumer acceptability.

Moreover, CCT and heat retention tests were performed to estimate the performance of
ICS and HRS. The results indicated that ICS results in substantial fuel and time saving
compared to OSD. Food in the HRB takes longer to reach the temperature danger zone
(from 5 to 60 °C), in which pathogenic microorganisms proliferate, compared to firewood
leftovers or banana leaves. In addition, consumer preference analysis showed that the
introduction of ICS and HRB did not affect food acceptance. Therefore, these technologies,
manufactured using locally available materials, are suitable for energy-poor communities.

The effect of solar drying on food security is analyzed. This study found that using passive
solar drying significantly improves the food security situation in households, which
implies that implementing passive solar drying in rural areas with associated training can
help rural farmers improve their own food security. Policy interventions should equip
farmers with better knowledge and understanding of solar drying, emphasizing its use as
a pathway to ensure some level of food security.

Finally, despite improved technologies' advantages, adoption is still beyond the desirable
level. Therefore, this study also showed that an individual training approach effectively
promotes clean cooking technologies. It resulted in the highest adoption rates and is more
cost- and time-effective than a group training approach. Nevertheless, it is still necessary
to consider that a training approach's effectiveness can be context-dependent.

The general conclusions of this thesis are, therefore, that (1) the improved technologies
presented in this study are sustainable and cost-effective means to substitute, at least
temporarily, the prevailing traditional methods of food processing and preparation; (2)
governments and relevant stakeholders involved in energy and/or food security
programs should consider the local context to choose the appropriate training or
information delivery technique and (3) energy provision through the use of passive solar
drying is an essential component in the fight against food and nutrition insecurity.
 178

|
8 Policy Recommendations

The policy recommendations derived from this PhD thesis are summarized as follows:

1) The dissemination of passive solar drying technologies should be fostered

Solar drying technology presents great potential as an eco-friendly method to reduce

post-harvest losses and improve the FNS situation in Central Mozambique. Therefore,
governments should support using low-cost and simple solar drying technologies at
individual and community scales. The designs should be based on practical experience,
local climate, and economic conditions, and training of users on solar drying for each crop
should be provided. Attention should be given to the capacity in terms of the daily
quantity of dried food since the low output is one of the main limitations of passive solar
dryers designs (Belessiotis and Delyannis, 2011). Moreover, a better capacity to increase
drying temperature can make it possible to reduce the significant dependence on sunlight
availability due to the associated higher daily productivity and shorter exposure time
necessary (Matavel et al., 2021).

2) Energy transition initiatives should encourage the use of efficient stove

technologies

In Mozambique, replacing firewood and charcoal with other renewable energy sources is
still far from reality. Therefore, energy policies should focus on promoting improved
techniques that can reduce wood consumption, considering users’ preferences. Replacing
traditional stoves with cleaner and more efficient ones could be a driving force.
Nevertheless, this should be taken as an intermediate stage. Investments in developing
the cleanest, most affordable options should be the ultimate goal. Currently, many
available biomass ICS do not meet the World Health Organization’s standards (Gill-Wiehl
and Kammen, 2022).
Policy Recommendations 179

3) Capacity-building programs and extension services should be aware of the cost

and time effectiveness of the selected approach

Promotional efforts are crucial for increasing the adoption and use of ICS and PISD.
Nevertheless, for better financial resource efficiency of stove promotion programs, the
extension agencies should be aware of cost-effectiveness and adjust extension techniques
to achieve the highest impact per unit of resource expenditure. Since place-based factors
such as remoteness significantly affect the ability to supply and convince households to
buy and use ICS (Lewis et al., 2015) or PISD, disseminated technologies should be based
on local materials. This would also result in employment opportunities in the
communities. Capacity-building programs and extension services should also introduce
technologies based on the needs of smallholder farmers. This would provide greater
certainty about the potential benefits of introduced technologies and avoid costly
technology development and distribution, which may lead to economic losses for both
adopters (farmers) and promoters (governments, NGO).

4) FNS programs and policies should improve access to adequate diets

Although the technologies introduced in this study can potentially improve the food and
nutrition security situation, policies that promote awareness of the nutritional quality of
foods are still necessary. In the study area, the diet is very monotonous, usually made up
of maize and cassava, cooked as a paste and served with beans or dark green leaves sauces
and/or dried or fresh fish (Arimond et al., 2011). This can impair the intake of essential
micronutrients, causing so-called "hidden hunger" (Maberly et al., 1994), which has
substantial economic effects (Stein and Qaim, 2007).

5) Food security interventions should promote a balance between market-oriented

agriculture and subsistence production

The analysis in this research suggests that the different FNS dimensions can be achieved
by applying different strategies. For instance, an adequate diet can only be achieved if
households consume a combination of products from their own production and
purchased food, providing access to essential nutrients that are unavailable in farm-
sourced staple foods. Therefore, relevant stakeholders can improve FNS by promoting a
balance between market-oriented agriculture and subsistence production.
 180

|
9 References

Aabø, E. & Kring, T. 2012. The political economy of large-scale agricultural land
acquisitions: Implications for food security and livelihoods/employment creation
in rural Mozambique. United Nations Development Programme Working Paper, 4,
1-61.
Adetutu, M. O. & Ajayi, V. 2020. The impact of domestic and foreign R&D on agricultural
productivity in sub-Saharan Africa. World Development, 125, 104690
https://doi.org/10.1016/j.worlddev.2019.104690.
Adkins, E., Tyler, E., Wang, J., Siriri, D. & Modi, V. 2010. Field testing and survey evaluation
of household biomass cookstoves in rural sub-Saharan Africa. Energy Sustain Dev,
14, 172-185 https://doi.org/10.1016/j.esd.2010.07.003.
Affognon, H., Mutungi, C., Sanginga, P. & Borgemeister, C. 2015. Unpacking Postharvest
Losses in Sub-Saharan Africa: A Meta-Analysis. World Development, 66, 49-68
https://doi.org/10.1016/j.worlddev.2014.08.002.
Alamu, E. O. & Mooya, A. 2017. Chapter 10 - Food Processing Technologies and Value
Addition for Improved Food Safety and Security. In: NHAMO, N., CHIKOYE, D. &
GONDWE, T. (eds.) Smart Technologies for Sustainable Smallholder Agriculture.
Academic Press.
Alemu, Z. G. & Van Schalkwyk, H. D. 2008. Market integration in Mozambican maize
markets, African Books Collective.
Arimond, M., Wiesmann, D., Becquey, E., Carriquiry, A., Daniels, M., Megan Deitchler,
Fanou, N., Ferguson, E., Joseph, M., Kennedy, G., Martin-Prével, Y. & Torheim, L. E.
2011. Dietary Diversity as a Measure of theMicronutrient Adequacy of Women’s Diets
in Resource-Poor Areas: Summary of Results from Five Sites [Online]. Washington,
DC: USAID. Available:
https://www.fantaproject.org/sites/default/files/resources/WDDP_Summary_R
eport_Jul2011.pdf [Accessed].
ASTM-International 2015. E2943−15: Standard Guide for Two-Sample Acceptance and
Preference Testing with Consumers. West Conshohocken: ASTM-International.
Atanassov, B., Falcão, M., Egas, A., Fernandes, A., Mahumane, G. & Nadaud, D. 2012.
Mozambique Urban Biomass Energy Analysis 2012.
Augustin, M. A., Cole, M. B., Ferguson, D., Hazell, N. J. G. & Morle, P. 2021. Perspective
article: Towards a new venture science model for transforming food systems.
Global Food Security, 28, 100481 https://doi.org/10.1016/j.gfs.2020.100481.
Augustin, M. A., Riley, M., Stockmann, R., Bennett, L., Kahl, A., Lockett, T., Osmond, M.,
Sanguansri, P., Stonehouse, W., Zajac, I. & Cobiac, L. 2016. Role of food processing
in food and nutrition security. Trends in Food Science & Technology, 56, 115-125
https://doi.org/10.1016/j.tifs.2016.08.005.
Ayua, E., Mugalavai, V., Simon, J., Weller, S., Obura, P. & Nyabinda, N. 2017. Comparison of
a mixed modes solar dryer to a direct mode solar dryer for African indigenous
vegetable and chili processing. Journal of Food Processing and Preservation, 41,
e13216 10.1111/jfpp.13216.
Baumert, S., Fisher, J., Ryan, C., Woollen, E., Vollmer, F., Artur, L., Zorrilla-Miras, P. &
Mahamane, M. 2019. Forgone opportunities of large-scale agricultural investment:
A comparison of three models of soya production in Central Mozambique. World
References 181

Development Perspectives, 16, 100145 https://doi.org/10.1016/j.wdp.2019.100145.

Becquey, E., Martin-Prevel, Y., Traissac, P., Dembélé, B., Bambara, A. & Delpeuch, F. 2010.
The household food insecurity access scale and an index-member dietary diversity
score contribute valid and complementary information on household food
insecurity in an urban West-African setting. The Journal of nutrition, 140, 2233-
2240.
Belessiotis, V. & Delyannis, E. 2011. Solar drying. Solar energy, 85, 1665-1691.
Bernardo, R., Davidsson, H., Samuelsson, P., Bengtsson, G., Döhlen, V., Olsson, J., Phinney,
R., Otte, P., Tivana, L., Andersson, M. & Rayner, M. 2021. Combined solar and
membrane drying technologies for sustainable fruit preservation in low-income
countries – prototype development, modelling, and testing. Solar Energy Advances,
1, 100006 https://doi.org/10.1016/j.seja.2021.100006.
Bilinsky, P. & Swindale, A. 2007. Months of adequate household food provisioning (MAHFP)
for measurement of household food access: indicator guide, Food and Nutritional
Technical Assistance Project, Academy for Educational ….
Bland, J. R. & Cook, A. C. 2019. Random effects probit and logit: understanding predictions
and marginal effects. Applied Economics Letters, 26, 116-123
10.1080/13504851.2018.1441498.
Boko, M., Niang, I., Nyong, A., Vogel, C., Githeko, A., Medany, M., Osman-Elasha, B., Tabo, R.
& Yanda, P. 2007. Impacts, Adaptation and Vulnerability. Contribution of Working
Group II to the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change. Africa. Climate Change, 433-467.
Bradford, K. J., Dahal, P., Van Asbrouck, J., Kunusoth, K., Bello, P., Thompson, J. & Wu, F.
2020. Chapter 17 - The dry chain: reducing postharvest losses and improving food
safety in humid climates. In: KOSSEVA, M. R. & WEBB, C. (eds.) Food Industry
Wastes. Academic Press.
Brew-Hammond, A. 2012. Energy: The Missing Millennium Development Goal. In: TOTH,
F. L. (ed.) Energy for Development: Resources, Technologies, Environment.
Dordrecht: Springer Netherlands.
Cafiero, C., Melgar-Quiñonez, H. R., Ballard, T. J. & Kepple, A. W. 2014. Validity and
reliability of food security measures. Annals of the New York Academy of Sciences,
1331, 230-248 https://doi.org/10.1111/nyas.12594.
Candelise, C., Saccone, D. & Vallino, E. 2021. An empirical assessment of the effects of
electricity access on food security. World Development, 141, 105390
https://doi.org/10.1016/j.worlddev.2021.105390.
Carducci, B., Keats, E. C., Ruel, M., Haddad, L., Osendarp, S. J. M. & Bhutta, Z. A. 2021. Food
systems, diets and nutrition in the wake of COVID-19. Nature Food, 2, 68-70
10.1038/s43016-021-00233-9.
Castigo, F. & Salvucci, V. 2017. Estimativas e Perfil da Pobreza em Moçambique, Uma
Análise Baseada no Inquérito sobre Orçamento Familiar-IOF 2014/15. Inclusive
Growth in Mozambique Working Paper.
CFS 2012. Coming to terms with terminology. Rome: The Nutrition and Consumer
Protection Division of the Food and Agriculture Organization. Committee on World
Food Security.
Champion, W. M., Charley, P. H., Klein, B., Stewart, K., Solomon, P. A. & Montoya, L. D. 2017.
Perception, culture, and science: A framework to identify in-home heating options
to improve indoor air quality in the Navajo Nation. Science of The Total
Environment, 580, 297-306 https://doi.org/10.1016/j.scitotenv.2016.11.053.
Charrua, A. B., Havik, P. J., Bandeira, S., Catarino, L., Ribeiro-Barros, A., Cabral, P., Moldão,
M. & Romeiras, M. M. 2021. Food security and nutrition in mozambique:
References 182

comparative study with bean species commercialised in informal markets.

Sustainability, 13, 8839.
Chilundo, M., De Sousa, W., Christen, E. W., Faduco, J., Bjornlund, H., Cheveia, E.,
Munguambe, P., Jorge, F., Stirzaker, R. & Van Rooyen, A. F. 2020. Do agricultural
innovation platforms and soil moisture and nutrient monitoring tools improve the
production and livelihood of smallholder irrigators in Mozambique? International
Journal of Water Resources Development, 36, S127-S147
10.1080/07900627.2020.1760799.
Chyung, S. Y., Roberts, K., Swanson, I. & Hankinson, A. 2017. Evidence-Based Survey
Design: The Use of a Midpoint on the Likert Scale. Perform Improv, 56, 15-23
https://doi.org/10.1002/pfi.21727.
Cugala, D., Tostão, E., Affognon, H. & Mutungi, C. 2012. Postharvest losses in Africa–
Analytical review and synthesis: the case of Mozambique. ICIPE (International
Center of Insect Physiology and Ecology).
Darr, D. & Pretzsch, J. 2008. Mechanisms of Innovation Diffusion Under Information
Abundance and Information Scarcity—On the Contribution of Social Networks in
Group Vs. Individual Extension Approaches in Semi-Arid Kenya. The Journal of
Agricultural Education and Extension, 14, 231-248
10.1080/13892240802207676.
De Brauw, A., Moursi, M. & Munhaua, A. B. 2019. Vitamin A intakes remain higher among
intervention participants 3 years after a biofortification intervention in
Mozambique. British Journal of Nutrition, 122, 1175-1181
10.1017/S0007114519002162.
Dehejia, R. H. & Wahba, S. 2002. Propensity score-matching methods for nonexperimental
causal studies. Review of Economics and statistics, 84, 151-161.
Delgado, L., Laborde, D. & Piñeiro, V. 2021. Post-harvest loss in Mozambique: Estimating
maize loss in Manica and Zambezia provinces, Rome, Food and Agriculture
Organization of the United Nations.
Desai, M. A., Mehta, S., Smith, K. R. & Organization, W. H. 2004. Indoor smoke from solid
fuels: assessing the environmental burden of disease at national and local levels.
El Bilali, H., Callenius, C., Strassner, C. & Probst, L. 2019. Food and nutrition security and
sustainability transitions in food systems. Food and Energy Security, 8, e00154
https://doi.org/10.1002/fes3.154.
Evans, L. T. 2003. Agricultural Intensification and Sustainability. Outlook on Agriculture,
32, 83-89 10.5367/000000003101294316.
Evans, W. D., Johnson, M., Jagoe, K., Charron, D., Young, B. N., Rahman, A. S. M. M., Omolloh,
D. & Ipe, J. 2018. Evaluation of Behavior Change Communication Campaigns to
Promote Modern Cookstove Purchase and Use in Lower Middle Income Countries.
International Journal of Environmental Research and Public Health, 15, 11.
Faber, M., Schwabe, C. & Drimie, S. 2009. Dietary diversity in relation to other household
food security indicators. International Journal of Food Safety, Nutrition and Public
Health, 2, 1-15.
FAO 2000. Food insecurity: when people live with hunger and fear starvation. The state of
food insecurity in the world. Rome: Food and Agriculture Organization of the United
Nation.
FAO 2006. Food Security. FAO.
FAO 2013. Safe access to firewood and alternative energy in humanitarian settings.
Guidance Note. Food and Agriculture Organization of the United Nations.
FAO 2021. The State of Food and Agriculture 2021. Making agrifood systems more resilient
to shocks and stresses, Rome, FAO.
References 183

FAO, European Union & CIRAD 2022. Food Systems Profile - Mozambique. Catalysing the
sustainable and inclusive transformation of food systems., Rome, Brussels and
Montpellier, FAO, French Agricultural Research Centre for International
Developmen, European Union.
FAO, IFAD, UNICEF, WFP & WHO 2021. The state of food security and nutrition in the world
2021: transforming food systems for food security, improved nutrition and affordable
healthy diets for all, Rome, FAO.
FAO, I. 2019. The state of food and agriculture 2019. Moving forward on food loss and
waste reduction. FAO, Rome, 2-13.
Fasoyiro, S. B. & Taiwo, K. A. 2012. Strategies for Increasing Food Production and Food
Security in Nigeria. Journal of Agricultural & Food Information, 13, 338-355
10.1080/10496505.2012.715063.
Feder, G., Anderson, J. R., Birner, R. & Deininger, K. 2010. Promises and Realities of
Community-Based Agricultural Extension. In: OTSUKA, K. & KALIRAJAN, K. (eds.)
Community, Market and State in Development. London: Palgrave Macmillan UK.
Ferrão, J., Bell, V., Cardoso, L. A. & Fernandes, T. 2018. Agriculture and food security in
mozambique. Journal of Food, Nutrition and Agriculture, 1, 7-11.
Fews Net 2019. Mozambique Staple Food Market Fundamentals, September 2018. USAID
Foster, P. W. 1992. The world food problem: Tackling the causes of undernutrition in the
Third World.
Fullerton, D. G., Bruce, N. & Gordon, S. B. 2008. Indoor air pollution from biomass fuel
smoke is a major health concern in the developing world. Transactions of The Royal
Society of Tropical Medicine and Hygiene, 102, 843-851
10.1016/j.trstmh.2008.05.028.
Furszyfer Del Rio, D. D., Lambe, F., Roe, J., Matin, N., Makuch, K. E. & Osborne, M. 2020. Do
we need better behaved cooks? Reviewing behavioural change strategies for
improving the sustainability and effectiveness of cookstove programs. Energy
Research & Social Science, 70, 101788
https://doi.org/10.1016/j.erss.2020.101788.
Gerster-Bentaya, M. & Hoffmann, V. 2011. Rural extension: Volume 3: Training Concepts
and Tools, CTA/GTZ/Margraf/Universität Hohenheim.
Gill-Wiehl, A. & Kammen, D. M. 2022. A pro-health cookstove strategy to advance energy,
social and ecological justice. Nature Energy, 10.1038/s41560-022-01126-2.
Goodwin, N. J., O'farrell, S. E., Jagoe, K., Rouse, J., Roma, E., Biran, A. & Finkelstein, E. A.
2015. Use of Behavior Change Techniques in Clean Cooking Interventions: A
Review of the Evidence and Scorecard of Effectiveness. Journal of Health
Communication, 20, 43-54 10.1080/10810730.2014.1002958.
Gundersen, C. & Ziliak, J. P. 2018. Food Insecurity Research in the United States: Where
We Have Been and Where We Need to Go. Applied Economic Perspectives and Policy,
40, 119-135 10.1093/aepp/ppx058.
Heckman, J. J. 1979. Sample Selection Bias as a Specification Error. Econometrica, 47, 153-
161 10.2307/1912352.
Hendriks, S. L., Van Der Merwe, C., Ngidi, M. S., Manyamba, C., Mbele, M., Mcintyre, A. M.,
Mkandawire, E., Molefe, Q. N., Mphephu, M. Q. & Ngwane, L. 2016. What are we
measuring? Comparison of household food security indicators in the Eastern Cape
Province, South Africa. Ecol Food Nutr, 55, 141-62
10.1080/03670244.2015.1094063.
Herforth, A., Bai, Y., Venkat, A., Mahrt, K., Ebel, A. & Masters, W. A. 2020. Cost and
affordability of healthy diets across and within countries: Background paper for The
References 184

State of Food Security and Nutrition in the World 2020. FAO Agricultural
Development Economics Technical Study No. 9, Food & Agriculture Org.
Herforth, A. & Ballard, T. J. 2016. Nutrition indicators in agriculture projects: Current
measurement, priorities, and gaps. Global Food Security, 10, 1-10
https://doi.org/10.1016/j.gfs.2016.07.004.
Hoddinott, J. 1999. Operationalizing household food security in development projects: an
introduction. Technical guide, 1, 1-19.
Hodges, R. J., Buzby, J. C. & Bennett, B. 2011. Postharvest losses and waste in developed
and less developed countries: opportunities to improve resource use. The Journal
of Agricultural Science, 149, 37-45 10.1017/S0021859610000936.
Hoffmann, H. K. 2016. Bioenergy, development and food security in Sub-Saharan Africa:
Pathways towards a more sustainable energy provision in Western Tanzanian
settlements. Humboldt-Universität zu Berlin, Lebenswissenschaftliche Fakultät.
Hoffmann, V., Probst, K. & Christinck, A. 2007. Farmers and researchers: How can
collaborative advantages be created in participatory research and technology
development? Agriculture and Human Values, 24, 355-368 10.1007/s10460-007-
9072-2.
Holleman, C., Rembold, F., Crespo, O. & Conti, V. 2020. The impact of climate variability
and extremes on agriculture and food security-An analysis of the evidence and case
studies.
IEA 2006. Energy for Cooking in Developing Countries. World Energy Outlook 2006. Paris:
OECD Publishing.
IEA 2017. Energy Access Outlook 2017. Paris: International Energy Agency.
IHME 2022. GBD Compare. Seattle: Institute for Health Metrics and Evaluation.
INE 2018. Folheto estatistico distrital gurue 2018. Quelimane: Instituto Nacional de
Estatística.
IRENA & Fao 2021. Renewable Energy for Agri-food Systems -Towards the Sustainable
Development Goals and
the Paris agreemen, Abu Dhabi and Rome, IRENA and FAO.
Jennifer, S., Ernest, R. M., Arnold, M. & Patrick, A. N. 2019. Post-harvest losses in maize
store-time and marketing model perspectives in Sub-Saharan Africa. Journal of
Stored Products and Postharvest Research, 10, 1-12.
Jetter, J., Zhao, Y., Smith, K. R., Khan, B., Yelverton, T., Decarlo, P. & Hays, M. D. 2012.
Pollutant Emissions and Energy Efficiency under Controlled Conditions for
Household Biomass Cookstoves and Implications for Metrics Useful in Setting
International Test Standards. Environmental Science & Technology, 46, 10827-
10834 10.1021/es301693f.
Jones, A. D., Ngure, F. M., Pelto, G. & Young, S. L. 2013. What Are We Assessing When We
Measure Food Security? A Compendium and Review of Current Metrics. Advances
in Nutrition, 4, 481-505 10.3945/an.113.004119.
Kadiyala, S. & Rawat, R. 2013. Food access and diet quality independently predict
nutritional status among people living with HIV in Uganda. Public health nutrition,
16, 164-170.
Kees, M. & Feldmann, L. 2011. The role of donor organisations in promoting energy
efficient cook stoves. Energy Policy, 39, 7595-7599.
Kipruto, W. A review of the cook stove test methods and their applicability in small scale
CDM cook stove projects. 2011. Technical report, United Nations Framework
Convention on Climate Change ….
Kitinoja, L. & Barrett, D. M. 2015. Extension of Small-Scale Postharvest Horticulture
Technologies—A Model Training and Services Center. Agriculture, 5, 441-455.
References 185

Langendahl, P.-A., Cook, M. & Potter, S. 2016. Sustainable innovation journeys: exploring
the dynamics of firm practices as part of transitions to more sustainable food and
farming. Local Environment, 21, 105-123 10.1080/13549839.2014.926869.
Lee, S. & Lee, D. K. 2018. What is the proper way to apply the multiple comparison test?
Korean J Anesthesiol, 71, 353-360 10.4097/kja.d.18.00242.
Lee, W. & Okos, M. R. 2011. Sustainable food processing systems - Path to a zero discharge:
reduction of water, waste and energy. Procedia Food Science, 1, 1768-1777
https://doi.org/10.1016/j.profoo.2011.09.260.
Leonardo, W. J., Van De Ven, G. W. J., Udo, H., Kanellopoulos, A., Sitoe, A. & Giller, K. E. 2015.
Labour not land constrains agricultural production and food self-sufficiency in
maize-based smallholder farming systems in Mozambique. Food Security, 7, 857-
874 10.1007/s12571-015-0480-7.
Lewis, J. J., Bhojvaid, V., Brooks, N., Das, I., Jeuland, M. A., Patange, O. & Pattanayak, S. K.
2015. Piloting Improved Cookstoves in India. Journal of Health Communication, 20,
28-42 10.1080/10810730.2014.994243.
Lindgren, S. A. 2020. Clean cooking for all? A critical review of behavior, stakeholder
engagement, and adoption for the global diffusion of improved cookstoves. Energy
Research & Social Science, 68, 101539
https://doi.org/10.1016/j.erss.2020.101539.
Lobefaro, S., Piciocchi, C., Luisi, F., Miraglia, L., Romito, N., Luneia, R., Foti, S., Mocini, E.,
Poggiogalle, E., Lenzi, A. & Donini, L. M. 2021. Cooking techniques and nutritional
quality of food: A comparison between traditional and innovative ways of cooking.
International Journal of Gastronomy and Food Science, 25, 100381
https://doi.org/10.1016/j.ijgfs.2021.100381.
Lombardi, F., Riva, F., Bonamini, G., Barbieri, J. & Colombo, E. 2017. Laboratory protocols
for testing of Improved Cooking Stoves (ICSs): A review of state-of-the-art and
further developments. Biomass and Bioenergy, 98, 321-335
https://doi.org/10.1016/j.biombioe.2017.02.005.
Maberly, G. F., Trowbridge, F., Yip, R., Sullivan, K. & West, C. 1994. Programs against
micronutrient malnutrition: ending hidden hunger. Annual review of public health,
15, 277-301.
Mabiso, A., Cunguara, B. & Benfica, R. 2014a. Food (In) security and its drivers: insights
from trends and opportunities in rural Mozambique. Food Security, 6, 649-670.
Mabiso, A., Cunguara, B. & Benfica, R. 2014b. Food (In)security and its drivers: insights
from trends and opportunities in rural Mozambique. Food Security, 6, 649-670
10.1007/s12571-014-0381-1.
Macassa, G., Salvador, E. M. & Da Cruz Francisco, J. 2018. Food Insecurity in Mozambique
: What Do We Know? And what Can Be Done? Working paper. Gävle: Gävle
University Press.
Maccarty, N., Still, D. & Ogle, D. 2010. Fuel use and emissions performance of fifty cooking
stoves in the laboratory and related benchmarks of performance. Energy for
Sustainable Development, 14, 161-171
https://doi.org/10.1016/j.esd.2010.06.002.
Makungwa, S. D., Epulani, F. & Woodhouse, I. H. 2013. Fuelwood supply: A missed
essential component in a food security equation. Journal of Food Security, 1, 49-51.
Mango, N., Siziba, S. & Makate, C. 2017. The impact of adoption of conservation agriculture
on smallholder farmers’ food security in semi-arid zones of southern Africa.
Agriculture & Food Security, 6, 1-8.
Mapfumo, P., Adjei-Nsiah, S., Mtambanengwe, F., Chikowo, R. & Giller, K. E. 2013.
Participatory action research (PAR) as an entry point for supporting climate
References 186

change adaptation by smallholder farmers in Africa. Environmental Development,

5, 6-22 https://doi.org/10.1016/j.envdev.2012.11.001.
Matavel, C., Hoffmann, H., Rybak, C. & Sieber, S. 2020. Can Subsistence Farming Help to
Achieve Household FoodSecurity? Evidence from Gurue, Central Mozambique. In:
TIELKES, E., ed. Tropentag 2020: Food and nutrition security and its resilience to
global crises, 2020. Witzenhausen: DITSL.
Matavel, C., Hoffmann, H., Rybak, C., Sieber, S., Müller, K., Brüntrup, M. & Salavessa, J. 2021.
Passive solar dryers as sustainable alternatives for drying agricultural produce in
sub-Saharan Africa: advances and challenges. Discover Sustainability, 2, 40
10.1007/s43621-021-00049-4.
Maxwell, D., Vaitla, B. & Coates, J. 2014. How do indicators of household food insecurity
measure up? An empirical comparison from Ethiopia. Food Policy, 47, 107-116.
Mbow, C., Rosenzweig, C., Barioni, L. G., Benton, T. G., Herrero, M., Krishnapillai, M.,
Liwenga, E., Pradhan, P., Rivera-Ferre, M.-G. & Sapkota, T. 2019. Food security.
Misau & Ine 2018. Inquérito de Indicadores de Imunização, Malária e HIV/SIDA em
Moçambique (IMASIDA) 2015. Maputo: MISAU.
Mulokozi, G. & Svanberg, U. 2003. Effect of traditional open sun-drying and solar cabinet
drying on carotene content and vitamin A activity of green leafy vegetables. Plant
Foods for Human Nutrition, 58, 1-15 10.1023/B:QUAL.0000041153.28887.9c.
Muro, P. D. & Burchi, F. 2007. Education for Rural People and Food Security: A Cross Country
Analysis, Rome, FAO.
Murphy, J. T. 2001. Making the energy transition in rural East Africa: Is leapfrogging an
alternative? Technological Forecasting and Social Change, 68, 173-193.
Naeher, L. P., Brauer, M., Lipsett, M., Zelikoff, J. T., Simpson, C. D., Koenig, J. Q. & Smith, K.
R. 2007. Woodsmoke Health Effects: A Review. Inhalation Toxicology, 19, 67-106
10.1080/08958370600985875.
Namagembe, A., Muller, N., Scott, L. M., Zwisler, G., Johnson, M., Arney, J., Charron, D. &
Mugisha, E. 2015. Factors Influencing the Acquisition and Correct and Consistent
Use of the Top-Lit Updraft Cookstove in Uganda. Journal of Health Communication,
20, 76-83 10.1080/10810730.2014.994245.
Neme, K., Nafady, A., Uddin, S. & Tola, Y. B. 2021. Application of nanotechnology in
agriculture, postharvest loss reduction and food processing: food security
implication and challenges. Heliyon, 7, e08539
https://doi.org/10.1016/j.heliyon.2021.e08539.
Ng’endo, M., Bhagwat, S. & Keding, G. B. 2017. Contribution of Nutrient Diversity and Food
Perceptions to Food and Nutrition Security Among Smallholder Farming
Households in Western Kenya: A Case Study. Food and Nutrition Bulletin, 39, 86-
106 10.1177/0379572117723135.
Nicolai, S., Hoy, C., Bhatkal, T. & Aedy, T. 2016. Projecting progress: The SDGs in sub-
Saharan Africa. Overseas Development Office, London.
Nielsen, T., Schünemann, F., Mcnulty, E., Zeller, M., Nkonya, E., Kato, E., Meyer, S., Anderson,
W., Zhu, T. & Queface, A. 2015. The food-energy-water security nexus: Definitions,
policies, and methods in an application to Malawi and Mozambique.
Norfolk, S. & Hanlon, J. Confrontation between peasant producers and investors in
northern zambézia, Mozambique, in the context of profit pressures on european
investors. Annual World Bank Conference on Land and Poverty, 2012 Washington
DC. The World Bank.
Nyambane, A., Johnson, F. X., Romeu–Dalmau, C., Ochieng, C., Gasparatos, A., Mudombi, S.
& Von Maltitz, G. P. 2020. Ethanol as a Clean Cooking Alternative in Sub-Saharan
Africa: Insights from Sugarcane Production and Ethanol Adoption Sites in Malawi
References 187

and Mozambique. In: GASPARATOS, A., NAIDOO, M., AHMED, A., KARANJA, A.,
FUKUSHI, K., SAITO, O. & TAKEUCHI, K. (eds.) Sustainability Challenges in Sub-
Saharan Africa II: Insights from Eastern and Southern Africa. Singapore: Springer
Singapore.
Nyyssölä, M., Pirttilä, J. & Sandström, S. 2014. Technology adoption and food security in
subsistence agriculture–evidence from a group-based aid project in Mozambique.
Finnish Economic Papers, 27, 1-33.
Otte, P. P., Tivana, L. D., Phinney, R., Bernardo, R. & Davidsson, H. 2018. The importance of
gender roles and relations in rural agricultural technology development: a case
study on solar fruit drying in Mozambique. Gender, technology and development,
22, 40-58.
Pailman, W., De Groot, J., Clifford, M., Jewitt, S. & Ray, C. 2018. Experiences with improved
cookstoves in Southern Africa. J Energy South Afr, 29, 13-26.
Pangaribowo, E. H., Gerber, N. & Torero, M. 2013. Food and nutrition security indicators:
a review.
Pine, K., Edwards, R., Masera, O., Schilmann, A., Marrón-Mares, A. & Riojas-Rodríguez, H.
2011. Adoption and use of improved biomass stoves in Rural Mexico. Energy for
Sustainable Development, 15, 176-183
https://doi.org/10.1016/j.esd.2011.04.001.
Popat, M., Griffith, G., Mounter, S. & Cacho, O. 2020. Postharvest losses at the farm level
and its economy-wide costs: the case of the maize sector in Mozambique. Agrekon,
59, 235-253 10.1080/03031853.2020.1721305.
Purohit, P., Kumar, A. & Kandpal, T. C. 2006. Solar drying vs. open sun drying: A framework
for financial evaluation. Solar Energy, 80, 1568-1579
https://doi.org/10.1016/j.solener.2005.12.009.
Raaijmakers, Q. a. W. 2000. Adolescents’ midpoint responses on likert-type scale items:
neual or missing values? Int J Public Opin Res, 12, 209-217 10.1093/ijpor/12.2.209.
Rogers, E. M. 2003. Diffusion of innovations, New York, New York: Free Press.
Salami, A., Kamara, A. B. & Brixiova, Z. 2010. Smallholder agriculture in East Africa: Trends,
constraints and opportunities, African Development Bank Tunis, Tunisia.
Salite, D., Cotton, M. & Kirshner, J. 2020. Electricity access and social sustainability in
Mozambique.
Samuelsson, P. & Deslandes, E. 2017. Testing of solar fruit driers in laboratory and in
Mozambique.
Santeramo, F. G. 2015. On the Composite Indicators for Food Security: Decisions Matter!
Food Reviews International, 31, 63-73 10.1080/87559129.2014.961076.
Scheid, A., Hafner, J., Hoffmann, H., Kächele, H., Sieber, S. & Rybak, C. 2018. Fuelwood
scarcity and its adaptation measures: an assessment of coping strategies applied
by small-scale farmers in Dodoma region, Tanzania. Environmental Research
Letters, 13, 095004 10.1088/1748-9326/aadb27.
Selvester, K., Fidalgo, L., Ballard, T., Kennedy, G., Dop, M., Mistura, L. & Deitchler, M. 2008.
Report on use of the household food insecurity access scale and household dietary
diversity score in two survey rounds in Manica and Sofala Provinces, Mozambique.
FAO Project, 1-23.
Singh, S., Gill, R. S., Hans, V. S. & Singh, M. 2021. A novel active-mode indirect solar dryer
for agricultural products: Experimental evaluation and economic feasibility.
Energy, 222, 119956 https://doi.org/10.1016/j.energy.2021.119956.
Soares, M. G. 2017. Relação entre as mudanças de uso e cobertura de terra e as queimadas
em florestas de Miombo, Gurué, Moçambique. Universidade Eduardo Mondlane.
References 188

Sola, P., Ochieng, C., Yila, J. & Iiyama, M. 2016. Links between energy access and food
security in sub Saharan Africa: an exploratory review. Food Security, 8, 635-642
10.1007/s12571-016-0570-1.
Springmann, M. 2020. Valuation of the health and climate-change benefits of healthy diets.
Stathers, T., Holcroft, D., Kitinoja, L., Mvumi, B. M., English, A., Omotilewa, O., Kocher, M.,
Ault, J. & Torero, M. 2020. A scoping review of interventions for crop postharvest
loss reduction in sub-Saharan Africa and South Asia. Nature Sustainability, 3, 821-
835 10.1038/s41893-020-00622-1.
Steenkamp, J.-B. E., Van Trijp, H. C. & Berge, J. M. T. 1994. Perceptual mapping based on
idiosyncratic sets of attributes. Journal of Marketing Research, 31, 15-27.
Stein, A. J. & Qaim, M. 2007. The Human and Economic Cost of Hidden Hunger. Food and
Nutrition Bulletin, 28, 125-134 10.1177/156482650702800201.
Sulaiman, V. R. & Hall, A. 2002. Beyond Technology Dissemination: Reinventing
Agricultural Extension. Outlook on Agriculture, 31, 225-233
10.5367/000000002101294119.
Sutar, K. B., Kohli, S., Ravi, M. R. & Ray, A. 2015. Biomass cookstoves: A review of technical
aspects. Renewable and Sustainable Energy Reviews, 41, 1128-1166
https://doi.org/10.1016/j.rser.2014.09.003.
Tefera, T., Kanampiu, F., De Groote, H., Hellin, J., Mugo, S., Kimenju, S., Beyene, Y.,
Boddupalli, P. M., Shiferaw, B. & Banziger, M. 2011. The metal silo: An effective
grain storage technology for reducing post-harvest insect and pathogen losses in
maize while improving smallholder farmers’ food security in developing countries.
Crop Protection, 30, 240-245 https://doi.org/10.1016/j.cropro.2010.11.015.
Tian, X., Engel, B. A., Qian, H., Hua, E., Sun, S. & Wang, Y. 2021. Will reaching the maximum
achievable yield potential meet future global food demand? Journal of Cleaner
Production, 294, 126285 https://doi.org/10.1016/j.jclepro.2021.126285.
Uckert, G., Hafner, J., Graef, F., Hoffmann, H., Kimaro, A., Sererya, O. & Sieber, S. 2017.
Farmer innovation driven by needs and understanding: building the capacities of
farmer groups for improved cooking stove construction and continued adaptation.
Environmental Research Letters, 12, 125001 10.1088/1748-9326/aa88d5.
Ugembe, M. A., Brito, M. C. & Inglesi-Lotz, R. 2022. Measuring energy poverty in
Mozambique: Is energy poverty a purely rural phenomenon? Energy Nexus, 5,
100039 https://doi.org/10.1016/j.nexus.2022.100039.
UN 2021a. The Sustainable Development Goals Report 2021. New York, NY: UN.
UN 2021b. The Sustainable Development Goals Report 2021: Extended Report - Goal 2.
New York, NY: UN.
UNICEF 2018. Improving the Nutritional Status of Children in Zambézia and Nampula.
Maputo: UNICEF.
Usaid 2022. Mozambique: nutrition profile. USAID.
Van Boekel, M., Fogliano, V., Pellegrini, N., Stanton, C., Scholz, G., Lalljie, S., Somoza, V.,
Knorr, D., Jasti, P. R. & Eisenbrand, G. 2010. A review on the beneficial aspects of
food processing. Molecular nutrition & food research, 54, 1215-1247.
Van Den Bulte, C. & Joshi, Y. V. 2007. New Product Diffusion with Influentials and
Imitators. Marketing Science, 26, 400-421 10.1287/mksc.1060.0224.
Wambugu, S. K., Karugia, J. T. & Oluoch-Kosura, W. 2018. Technology use, gender, and
impact of non-farm income on agricultural investment: an empirical analysis of
maize production in two regions of Kenya. In: DJURFELDT, A. A., DZANKU, F. M. &
ISINIKA, A. C. (eds.) Agriculture, Diversification, and Gender in Rural Africa:
Longitudinal Perspectives from Six Countries. Oxford: Oxford University Pres.
References 189

Weaver, C. M., Dwyer, J., Fulgoni, V. L., Iii, King, J. C., Leveille, G. A., Macdonald, R. S.,
Ordovas, J. & Schnakenberg, D. 2014. Processed foods: contributions to nutrition.
The American Journal of Clinical Nutrition, 99, 1525-1542
10.3945/ajcn.114.089284.
Weingärtner, L. 2009. The concept of food and nutrition security. Achieving Food and
Nutrition Security, 3, 21-52.
WFP 2019. Energy for Food Security: Enhancing people’s food security with improved
energy access. WFP.
WFP. 2022. Energy for food security [Online]. Available: https://www.wfp.org/safe
[Accessed].
WFP Mozambique 2022. Country Brief. WFP.
Who & Unicef 2021. The tenth meeting of the WHO-UNICEF Technical Expert Advisory
group on nutrition monitoring (TEAM).
Wijesinha-Bettoni, R. & Mouillé, B. 2019. The Contribution of Potatoes to Global Food
Security, Nutrition and Healthy Diets. American Journal of Potato Research, 96,
139-149 10.1007/s12230-018-09697-1.
Williams, R. 2006. Generalized Ordered Logit/Partial Proportional Odds Models for
Ordinal Dependent Variables. The Stata Journal, 6, 58-82
10.1177/1536867x0600600104.
Zhou, A. C. & Hendriks, S. L. 2017. Does food assistance improve recipients’ dietary
diversity and food quality in Mozambique? Agrekon, 56, 248-262.
Zorya, S., Morgan, N., Diaz Rios, L., Hodges, R., Bennett, B., Stathers, T., Mwebaze, P. &
Lamb, J. 2011. Missing food: the case of postharvest grain losses in sub-Saharan
Africa. Technical Report. Washington DC, USA: The International Bank for
Reconstruction and Development / The World BanK.
 190

Statutory Declaration
I hereby declare having completed the doctoral thesis independently and solely based on
the stated resources and aids.
I have not applied for a doctoral degree elsewhere and do not have a corresponding
doctoral degree. I have not submitted the doctoral thesis, or parts of it, to another
academic institution and the thesis has not been accepted or rejected. I declare having
acknowledged the Doctoral Degree Regulations of the Faculty of Life Sciences at
Humboldt-Universität zu Berlin of March 5, 2015, which underlie the procedure.
Furthermore, I declare that no collaboration with any commercial doctoral degree
supervisors took place, and that the principles of Humboldt-Universität zu Berlin for
ensuring good academic practice were abided by.

Eigenständigkeitserklärung
Hiermit erkläre ich, die Dissertation selbstständig und nur unter Verwendung der
angegebenen Hilfen und Hilfsmittel angefertigt zu haben.
Ich habe mich nicht anderweitig um einen Doktorgrad beworben und besitze keinen
entsprechenden Doktorgrad. Ich erkläre, dass ich die Dissertation oder Teile davon nicht
bereits bei einer anderen wissenschaftlichen Einrichtung eingereicht habe und dass sie
dort weder angenommen noch abgelehnt wurden. Ich erkläre die Kenntnisnahme der
dem Verfahren zugrunde liegenden Promotionsordnung der Lebenswissenschaftlichen
Fakultät der Humboldt-Universität zu Berlin vom 5. März 2015.
Weiterhin erkläre ich, dass keine Zusammenarbeit mit gewerblichen
Promotionsberatenden stattgefunden hat und dass die Grundsätze der Humboldt-
Universität zu Berlin zur Sicherung guter wissenschaftlicher Praxis eingehalten wurden.

Berlin, 7th of November 2022

_________________________________
Custodio Efraim Matavel