Journal of ISSN 0972-1045

Vol. 11, No. 1, January-June, 2009

Applied Appl Hort

Horticulture

Journal of
THE SOCIETY FOR ADVANCEMENT OF HORTICULTURE
JOURNAL OF APPLIED HORTICULTURE
Vol. 11, No. 1, Januar y-June, 2009
CONTENTS
Applications of GIS to Citriculture in South Texas 3
Reginald S. Fletcher (USA).
Extracting within-experiment precision of horticultural experiments useful for
meta-analysis 10
Guido Knapp, Bimal K. Sinha and Dihua Xu (USA).
Is CropSyst adequate for management-oriented simulation of growth and yield of
processing tomato ? 17
Onofri Andrea, Beccafichi Catia, Benincasa Paolo, Guiducci Marcello and Tei Francesco (Italy).
Starch degradation characteristics in relation to physiological and biochemical
properties during growth and maturation of apple fruit 23
Manasikan Thammawong and Osamu Arakawa (Japan).
Water usage and water use efficiency of drip-irrigated tomato under deficit irrigation 31
Berhanu Kebebew and Ketema Tilahun (Ethiopia).
Effects of antibrowning agents on the shelf life of fresh-cut green jackfruit
(Artocarpus heterophyllus Lam.) 41
Boodia Navindra, Ruggoo Arvind and Boodoo B. Hassina (Mauritius).
Sucrose synthase and acid invertase activities in relation to the floral structures abortion
in pepper (Capsicum annuum L.) grown under low night temperature 35
Néji Tarchoun Salah Rezgui and Abdelaziz Mougou (Tunisia).
Use of plastic shades to regulate growth of korarima (Aframomum corrorima
(Braun) P.C.M. Jansen) 46
S. Eyob (Ethiopia).
Effect of growth regulators on in vitro plant regeneration of female papaya using axillary
bud as an explant 50
Renu Singh, Ram C. Yadav and Neelam R. Yadav (India).
Effects of the addition of clinker ash to the propagation medium on rooting of
rabbiteye blueberry cuttings 54
T. Ban, H. Kitazawa, S. Matsumoto, N. Kobayashi, K. Tokumasa, M. Kobatake and T. Asao (Japan).
Persian walnut (Juglans regia L.) grafting as influenced by different bench grafting
methods and scion cultivars 56
Babak Dehghan, Kourosh Vahdati, Darab Hassani and Reza Rezaee (Iran).
Extraction and determination of α-solanine in eggplant fruits 59
Zhiwen Li, Baoli Zhou, Yuwen Ding and Xiang Liu (China).
Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum
under stress and normal conditions 64
M. Altaf Wani, G.R. Lawania, R.A. Bhat, Iffiat Fayaz, A. Nanda and Gazenfar Gani (India).
Growth and interference of invasive Russian knapweed on “Valcatorce INTA” onion 68
Carlos R. Bezic, Armando A. Dall Armellina, Omar A. Gajardo, Lucrecia M. Avilés and
Silvia L. Cañón (Argentina).
Cauliflower hybrids for spring production in a southern mediterranean area 73
M. Sciortino and G. Iapichino (Italy).
The effect of different sowing times on development and efficiency of some 78
Chinese cabbage varieties (Brassica campestris sbsp. pekinensis)
Funda Eryilmaz Acikgoz (Turkey).
Forthcoming Papers
ISSR, anthocyanin content and antioxidant activity analyses to characterize strawberry genotypes-Samir C
Debnath and Elodie Ricard (USA).
The effect of high daytime temperatures on inhibition of flowering in ‘Koroneiki’ olives (Olea europaea L.)
under chilling and non-chilling nighttime temperatures-Nasir S.A. Malik and Joe M. Bradford (USA).
Irrigation management in greenhouse tomato production on peat substrate-G.A. Peyvast and N. Mayer
(Germany).
Efficacy and physical properties of ground, composted rice hulls as a component of soilless substrate for
selected bedding plants-C. Song, Paul V. Nelson, Carl E. Niedziela Jr., and D. Keith Cassel (USA).
Extraction of anthocyanins from strawberries (Fragaria × ananassa) and analysis using HPLC/Q-TOF mass
spectrometry-Qiong Zhang, Hong-qing Wang and Le-xin Jia (China).
Low cost hydroponic devices and use of harvested water for vegetable and flower cultivation-A. Das, and
D. Sing Majhi (India).
Evaluation of the FAO CROPWAT model for deficit-irrigation scheduling for onion crop in a semiarid region
of Ethiopia-Samson Bekele and Ketema Tilahun (Ethiopia).
An evaluation of health of cactus collection available in botanical garden of Cluj-Napoca, Romania-Gyorgy
Feszt and Lucica Mihalte (Romania).
Performance of asparagus under the desert conditions of Arabian Peninsula: A pilot study-N. Kameswara
Rao and Mohammed Shahid (UAE).
Effect of ripening concentrate on ripening and sensory properties of locally cultivated mangoes (Mangifera
indica L.)-William Ofori Appaw , Ibok Oduro, William Otto Ellis (Ghana).
Horizontal and vertical soilless growing systems under Cyprus conditions-Damianos Neocleous, Charalambos
Kaittanis, Nicos Seraphides and Polycarpos Polycarpou (Cyprus).
Expression of gene-related anthocyanin accumulation in berries of cv. Tannat (Vitis vinifera L.)- O. Borsani,
G. Gonzalez-Neves, M. Ferrer and J. Monza (Uruguay).
Caffeine, phenol and protein contents of thirty-seven clones of Nigerian robusta coffee (Coffea canephora
Pierre ex. Froehner)-S.S. Omolaja (Nigeria).
Growth, nutrient uptake and nitrogen use efficiency of Ficus hawaii grown on nutrient film techniques (NFT)
,
using different N-sources- Mohamed, M. El-Fouly, A.A. El-Sayed A.A. Fawzy, B.M. Mansour H.A. Bosila and
Hassan. A. Hamouda (Egypt).
Influence of fungicides and Phytophthora capsici-resistant/tolerant cultivars on bell pepper yield and farm-
gate revenues- Jamie, R. Stieg, S. Alan Walters, Jason P. Bond, and M. Babadoost (USA).
Growth, yield and productivity responses of okra-pawpaw mixture to sequence of components in S.W.
Nigeria-O.O. Olubode, I.O.O. Aiyelaagbe and J.G., Bodunde (Nigeria).
Constraints as perceived by the vegetable growers regarding the adoption of IPM technologies in Cauliflower
cultivation: an empirical study- Prabuddha Ray and Sarthak Chowdhury (India).
The effect of high daytime temperatures on inhibition of flowering in ‘Koroneiki’ olives (Olea europaea L.)
under chilling and non-chilling nighttime temperatures- Nasir S.A. Malik and Joe M. Bradford (USA).
Leaf N and P in different growth habits of peach: Effects of root system morphology and transpiration-
Thomas Tworkoski A, Ralph Scorza, and D.Michael Glenn (USA).
In vitro flowering and shoot multiplication of gentiana triflora in air-lift bioreactor cultures-Yaser Hassan
Dewir, Nisha Singh, Siveshni Govender and Pragashnee Pillay (Egypt).
Comparison of drying characteristics of bitter gourd using cabinet drier, microwave drier and combination
of cabinet and infrared drier-K.A. Athmaselvi (India).
Induction of multiple shoots in Amomum hypoleucum Thwaites – A Threatened Wild Relative of Large
Cardamom- M. Bejoy, M. Dan, N.P. Anish, Githa Ann George and B.J. Radhika (India).
Effects of grafted eggplants on allelopathy of cinnamic acid and vanillin in root exudates-Chen Shaoli, Zhou
Baoli, Wang Ruhua, Xi Haijun (China).
Ripening in fruit of three papaya cultivars: Solo Sunrise, Tainung #2 and Red Lady at two temperatures-
Protain, S., M. Mohammed and L.A. Wilson (West Indies).
 Journal
Journal of Applied Horticulture, 11(1):3-9, January-June, 2009
Appl

Applications of GIS to Citriculture in South Texas

Reginald S. Fletcher
United States Department of Agriculture, Agricultural Research Service, Kika de la Garza Subtropical Agricultural
Research Center, 2413 E. Hwy. 83, Weslaco, Texas 78596, E-mail: reginald.fletcher@ars.usda.gov

Abstract
The South Texas citrus industry needs an inventory of soil properties within existing citrus (Citrus spp.) orchards, wants data at the
county level showing soils that are suitable for citrus production, and would value any information related to the establishment of citrus
orchards. This study discusses integration of citrus, soil survey geographic data (SSURGO), and U.S. Census spatial and tabular data
with geographical information system (GIS) technology for citriculture. For this study, Hidalgo County Texas was evaluated because it
is the major citrus producing county in South Texas. The spatial and tabular data and commercial GIS software were used to inventory
selected soil chemical and physical properties within citrus groves, to identify orchards that may be affected by urban expansion, and
to select potential sites for establishing new citrus orchards. Results indicated that citrus, SSURGO, and U.S. census spatial and tabular
data integrated with GIS technology can be a powerful tool for citriculture. The information provided in this study should appeal to
producers, extension agents, scientists, and government agencies within the U.S. and abroad.
Key words: Citriculture, citrus, Citrus spp., Geographic Information System (GIS), soil survey geographic data (SSURGO), Topologically
Integrated Geographic Encoding and Referencing (TIGER) data, grapefruit, Citrus paradisi, oranges, Citrus sinensis

Introduction decisions in reclamation and in management of salt affected soils

In the Lower Rio Grande Valley of Texas, the citrus industry needs
inventories on chemical and physical properties of soils within
citrus groves to assist them in making management decisions.
The life expectancy of citrus trees ranges from 30 to 50 years.
Establishment of new groves on soils having a high potential for
citrus production and a low potential for commercial buildings
and homes would be ideal because these locations can remain
in citrus for extended periods, reducing the loss of established
orchards to urban expansion.
Spatial and tabular soil data are available in digital formats,
allowing analysts to incorporate it into a geographic information
system (GIS), a computer system designed to store, retrieve, and
manipulate geographic data in various formats. The soil survey
geographic data (SSURGO) created by the Natural Resource
Conservation Service (NRCS) provides the most comprehensive
soil information for the United States (U.S.). It is designed for
multiple purposes (National Soil Survey Center, 1995). The U.S.
Census Bureau has also created spatial and tabular data in digital
format to support its mapping needs for the decennial census
and other bureau programs. This data contains line features
(e.g., roads, railroad, transportation, etc.,); boundary features
(e.g., statistical-census tracts and blocks, government-places and
counties, administrative-congressional districts, school districts,
etc.,); and landmark features (e.g., point-schools and churches,
area-park and cemeteries, and key geographic locations-apartment
building and factories).
Researchers have used GIS technology and spatial and tabular data
provided by government agencies to assist them in assessing and
in solving problems affecting agricultural and natural resources.
Mandal and Sharma (2006) demonstrated the application of GIS
to develop maps of salt-affected soils in India at the regional, state,
and local level. Maps were useful for planning and for making
in the Indo-Ganetic Plain. Boonyanuphap et al. (2004) employed
GIS-based analysis to identify suitable land for establishing
banana plantations in Thailand. Wu et al. (2001) showed that the
use of SSURGO data, integrated with remotely-sensed and other
GIS layers were beneficial for planning and for managing natural
resources in Finney County Kansas. Wu et al. (1997) also found
SSURGO data to be an excellent source for determining erodible
areas and for developing conservation practices. For subtropical
South Texas, Richardson et al. (1996) utilized GIS technology to
estimate sugarcane production, to map soil type, to manage farms,
to classify vegetation communities, to enforce the cotton stalk
destruction program, and to monitor the migration of Africanized
honey bees into the southern U.S.
The development of SSURGO, along with other digital GIS
layers, has been an important advancement. Little information is
available on integrating SSURGO and U.S. Census data with GIS
technology for citriculture. The objectives of this study were to
use citrus, SSURGO, and U.S. Census data with GIS technology
(i) to inventory selected chemical and physical soil properties
within active citrus groves; (ii) to assess potential of soils for
establishing citrus orchards; and (iii) to address alternatives for
citrus grove establishment. For this study, Hidalgo County Texas
was evaluated because it is the major citrus producing county in
South Texas.
Materials and methods
County description: Hidalgo County is located at the southern
tip of Texas (Fig. 1). The climate is subtropical. Annual rainfall
average is 58.42 cm. The elevation ranges from 12.2 to 114 m
with the greatest elevation occurring in the western half of the
county. The most valuable resources are soil, water, natural gas,
and the mild climate.
4 Applications of GIS to Citriculture in South Texas
Fig. 1. (a) Map of the continental United States showing the location
of Texas and the study area (Enclosed in black square). (b) Close-up
showing the Lower Rio Grande Valley and the study area-Hidalgo
County.
Agriculture is a major source of income for the county. For
the state of Texas, the highest vegetable production occurs in
Hidalgo County. Irrigated land is primarily used for agricultural
production. Producers intensely farm the irrigated land, and their
methods are highly specialized. Irrigated crop production includes
cotton (Gossypium hirsutum L.), citrus, vegetables, grain sorghum
Fig. 2. Flowchart showing steps used to create citrus soil database.
(Sorghum bicolor L.), and sugarcane. Grapefruit (Citrus paradisi
L.) and oranges [C. sinensis (L.) Osbeck] are the primary citrus
grown in the county. Drainage and salinity are the major soil
problems affecting agriculture.
Development of GIS database: Illustrated in Fig. 2 is a flowchart
showing the development of the spatial and tabular citrus-soil
database. Spatial and tabular soil data of Hidalgo County, Texas,
and a spatial database template were downloaded from the Soil
Data Mart website (http://soildatamart.nrcs.usda.gov/), the
official NRCS website for SSURGO data. SSURGO is the most
comprehensive soil data in digital format provided by NRCS. The
spatial component contained information needed to display and
to analyze the data with GIS software. The tabular component
included a set of American Standard for Information Interchange
fields and text delimited files. The text delimited files matched a
table within the SSURGO database template, which is a Microsoft
Access database file containing empty tables. For this study, the
soildb_US_2002.mb template (Microsoft Access 2002 format)
was downloaded for further analysis. The spatial SSURGO data
were transferred to the Manifold GIS (version 7.0) software.
The template table and the text delimited data were merged
using Micrsoft Access 2002. The SSURGO database template
has a macro (a rule specifying how a certain input sequence
should be mapped to an output sequence according to a defined
procedure) that allows the user to import the text delimited data
into the SSURGO database. To use the macro and to complete
the importing procedure, the analyst has to use Microsoft Access.
Detailed instructions for this procedure were provided in the
SSURGO manual.
The majority of the feeder roots for mature citrus trees occur
within 61 cm of the soil surface (Ray and Walheim, 1980). For
the soil inventory, chemical and physical properties based on
this depth were determined with Soil Data Viewer (version 5.1,
 Applications of GIS to Citriculture in South Texas
http://nrcs.usda.gov), a freeware designed to create soil-based
thematic maps when linked with ArcMap (GIS software) and
tabular output for systems not equipped with ArcMap. The tabular
data was used in this study. Soil Data Viewer only works with
Microsoft Access databases, hence, the need to merge the text
with the tabular template using Microsoft Access.
To calculate the information needed for the 0 to 61 cm depth,
the soils data had to be aggregated. The following rules were
used to estimate the values: (1) the dominant component, (2) the
high tie break rule, and (3) converting null values to zeroes. The
dominant component aggregation method returns the attribute
value associated with the component occupying the highest
percent composition in the map unit. The tie-break rule indicates
which value should be selected from a set of multiple candidate
values, or which value should be selected in the event of a percent
composition tie. For this study, the component having the greatest
value was used in tie break situations. Interpret nulls as zeroes
is used to determine if a null value for a component should be
converted to zero before aggregation occurs. This conversion
happens only if a soil map unit has at least one component
where this value is not null. For this study, null values were not
interpreted as zeroes. Aggregated data were transferred to the GIS
software and incorporated into the soil spatial data file table.
Soil parameters estimated with the Soil Data Viewer were clay
content, pH, organic matter, and electrical conductivity. These
properties were selected because they are important to citrus
production. Brief definitions of the selected categories are as
follows. Clay is mineral material composed of particles that are
less than 0.002 mm in diameter. pH is a measure of the degree
of acidity or alkalinity of a soil. Soils with pH values below
6.6 are considered acidic; soils with pH values above 7.3 are
characterized as alkaline; and soils with pH values equal to or
within the extremes are characterized as neutral (USDA, 1981).
Electrical conductivity (EC) is a measure of the concentration of
water-soluble salts in soils. Organic matter is the plant and animal
residue in the soil at various stages of decomposition. Output
for soil properties were manually entered into the soil database
residing in the GIS.
Other soil parameters calculated with the Soil Data Viewer
included the soil map unit potential for building small commercial
buildings and for building dwellings without basements. The
dominant component and high tie break rule were used to assign
the map units to a class: not limited, somewhat limited, and very
limited.
Brief definitions for small commercial buildings and dwellings
without basements categories are as follows. Small commercial
buildings and dwellings without basements are structures less than
three stories high and do not have basements. Ratings are based on
soil properties that affect the capacity of the soil to support a load
without movement and on the properties that affect excavation
and construction costs. Properties influencing load-supporting
capacity include depth to a water table, ponding, flooding,
subsidence, linear extensibility (shrink-swell potential), and
compressibility (which is inferred from the Unified classification
of the soil).
Ratings are both verbal and numerical. Rating class terms
indicate the extent to which the soils are limited by all of the
5
soil features that affect the specified use. Verbal ratings include
the following classes, not limited, somewhat limited, and very
limited. Not limited encompasses soils having very favourable
characteristics for the specified use. Good performance and very
low maintenance can be expected. Somewhat limited includes
soils having features that are moderately favourable for the
specified use. The limitations can be overcome or minimized by
special planning, design, or installation. Fair performance and
moderate maintenance can be expected. Very limited indicates
that the soil has one or more features that are unfavourable for
the specified use. The limitations generally cannot be overcome
without major soil reclamation, special design, or expensive
installation procedures. Poor performance and high maintenance
can be expected.
No information related to the potential of soils for citrus grove
establishment appeared in the SSURGO tables. This information
was available in the hard copy version of the soil survey (USDA,
1981). Soils were grouped into the following classes for citrus
grove establishment, high, medium, low, questionable, or not
suitable. High potential indicated that the performance is at
or above the level of locally established standards, the cost of
measures for overcoming soil limitations are judged locally to
be favourable in relation to the expected performance or yields,
and soil limitations that continue after corrective measures are
installed do not detract appreciably from environmental quality or
economic returns. Medium potential signifies that production or
performance is somewhat below locally established standards, the
costs of measures for overcoming soil limitations are high, or soil
limitations that continue after corrective measures are installed
detract from environmental quality or economic returns. Low
potential indicates that production or performance is significantly
below local standards, measures that are required to overcome
soil limitations are very costly, or soil limitations that continue
after corrective measures are installed detract appreciably
from environmental quality or economic returns. Questionable
indicates that not enough evidence existed to determine if the
land could or could not be used for citrus grove establishment.
Unsuitable means that these soils are not adequate for citrus
grove establishment.
The soil survey did not group five soil map units into a potential
class. Several steps were taken to try to group the map units into
a class; the steps are as follows. The descriptions of the map units
were evaluated thoroughly and compared with map units having
the same surname. For example, the soil survey did not rate the
Harlingen clay-urban complex into a potential class for citrus
grove establishment. However, it did provide information on
Harlingen clay, which received a rating of not suitable for citrus
grove establishment. The Harlingen clay-urban complex consisted
of urban areas developed on Harlingen clay soil; therefore, the
Harlingen clay-urban complex was categorized as unsuitable for
citrus production.
If the nonrated soil could not be placed into a category using the
surname procedure, then it was assigned to one of two additional
classes, not rated-yield reported and not rated-yield not reported.
To create these classes, the Soil Data Viewer was used to determine
if citrus yields had been reported for the soil map unit in question.
If yield information was reported, then the soil was categorized
as not rated for citrus grove establishment-citrus yield reported
6 Applications of GIS to Citriculture in South Texas

(NRYR). If yield information was not provided, then the soil map Note: The spatial data utilized for the study were projected to a
unit was rated as not rated for citrus grove establishment-no citrus coordinate system and datum by the various entities that created
yield reported (NRNYR). The suitability data were transferred to the data. For this study, the final coordinate system and datum
the soil spatial table residing in the Manifold software package. used for the maps were Universal Transmercator (Zone 14 N)
and North American Datum 1983, respectively.
The citrus database was created in a joint venture between U.S. It was difficult to display the citrus polygons onto thematic maps.
Department of Agriculture-Animal Plant Inspection Service and For illustrative purposes, centroids (points) extracted from the
U.S. Department of Agriculture- Agricultural Research Service. polygons of the citrus-soil spatial layer was used to represent the
The 2006 updated version of the database was employed in this locations of citrus groves.
study. It consists of spatial and tabular data. The former has
polygons representing areas planted to citrus trees and the latter Results and discussion
contains attributes for the citrus polygons. Spatial and tabular data Citrus inventory by type: Summarized in Table 1 is the overall
were imported directly into the GIS software (Fig. 2). Attributes statistics created with the combined citrus-soil database. The total
used in this study were citrus type and active orchards. area encompassed by active orchards was 9214.7 ha. Grapefruit
To extract the urban area information of Hidalgo County, the and oranges were the dominant citrus types, covering 66% and
TIGER (Topologically Integrated Geographic Encoding and 33% of the total area, respectively. Tangelos (C. x tangelo Ingram
Referencing)/Line 108th CD Census 2000 data were downloaded and Moore), tangerines (C. reticula Blanco), and lemons [C. x
from the U.S. Census Bureau website (www.census.gov). The limon (L.) Burm.f.] were included in the citrus survey; their total
TIGER/Line files were created from the Census Bureau’s TIGER area of coverage was less than 1%.
database of selected geographic and cartographic information. Figs. 3-5 show how the GIS can be used to display the spatial
TIGER was developed by the U.S. Census Bureau to support distribution of the orchards at the county level based on citrus
the mapping and related geographic activities required by the type. Grapefruit and orange groves were scattered throughout
decennial and economic censuses and sample survey programs. the south central region of the county (Figs. 3 and 4). Lemons,
The TIGER data were imported into the GIS software for further tangelos, and tangerines were established in isolated areas within
analysis (Fig. 2). the county (Fig. 5).
For this study, information related to Urban Areas was employed Soil physical data of citrus groves: The soil physical data
to assist in answering management questions. The Census Bureau indicated that producers established their groves on soils with
uses the term Urban Area (UA) to refer collectively to Urbanized high potential for citrus grove establishment (Table 1). Thirty-four
Areas (UZA) and Urban Clusters (UC). Urbanized Areas (UZA) percent of citrus production occurred on soils with lower ratings.
is a statistical geographic entity consisting of a central core and The dominant soil mapping units for citrus grove establishment
adjacent densely settled territory that together contain at least was Brennan fine sandy loam, 0 to 1% slopes; followed by
50,000 people, generally with an overall population density of Hidalgo sandy clay loam, 0 to 1% slopes; Hidalgo fine sandy
at least 386 people per square km. Urban Clusters (UC) is a new
statistical geographic entity designated by the Census Bureau
for the 2000 Census, consisting of a central core and adjacent
densely settled territory that together contains between 2,500
and 49,999 people. Typically, the overall population density is
at least 386 people per square km. Urban Clusters are based on
Census block and block group density and do not coincide with
official municipal boundaries.
The soil data was merged with the citrus data by employing the
topology overlay function of the GIS software. This function
created new polygons containing soil and citrus attributes. Using
the newly created data layer and structure query language, the
inventory of the selected chemical and physical properties of soils
within citrus groves were completed. Structure query language
and visual basic script were used throughout the study to select
appropriate information from tables, estimate area, or perform
mathematical calculations.
At the county level, a thematic map showing the potential of
the soil map units for citrus grove establishment was developed
to assist in selecting new areas for establishing citrus groves. A
thematic map showing the census urban area data, citrus location
data, and high potential soil citrus grove establishment data versus
the building site development data was evaluated to determine Fig. 3. Spatial distribution of active grapefruit orchards within Hidalgo
alternative sites for establishing citrus groves. County.
 Applications of GIS to Citriculture in South Texas 7

Table 1. Overall statistics for active citrus groves in Hidalgo County, Texas, including area of citrus by variety, soil map unit ranked by area*, soil
clay content ranked by area*, soil organic matter ranked by area*, soil potential, soil pH level, soil electrical conductivity, and soil characteristics for
building site development (small commercial buildings and dwellings without a basement)
Electrical Conductivity Value Range (S m-1) Area (ha)
Non-saline 0.00 to 0.20 9151.0
Slightly-saline 0.21 to 0.40 45.6
Moderately saline 0.41 to 0.80 1.0
Strongly-saline 0.81 to 1.60 17.1

pH pH Values Area (ha) Building site development Area (ha)

Slightly acid 6.1 to 6.5 41.1 Not limited 3480.0
Neutral 6.6 to 7.3 753.5 Somewhat limited 5240.0
Mildly alkaline 7.4 to 7.8 2967.5 Very limited 494.2
Moderately alkaline 7.9 to 8.4 5452.1
Strongly alkaline 8.5 to 9.0 0.5

Potential Area (ha) Organic matter g kg-1 Area (ha)

High potential 8033.2 15.3 266.6
Medium potential 597.6 5.7 276.6
Low potential 355.2 16.9 401.9
Not suitable 163.1 6.6 2893.2
Questionable 17.3 16.3 4705.1
NRYR 48.4
NRNYR 0.0

Clay g kg-1 Area (ha) Variety Area (ha)

193 270.7 Grapefruit 6124.5
224 274.5 Oranges 3045.1
203 2342.1 Tangelo 38.1
256 2363.1 Tangerine 6.8
180 2889.2 Lemon 0.2

Soil Map Unit Area (ha)

Raymondville clay loam
Brenan fine sandy loam, 1 to 3% slopes
Hidalgo fine sandy loam, 0 to 1% slopes
Hidalgo sandy clay loam, 0 to 1% slopes
Brennan fine sandy loam, 0 to 1% slopes
*Only the top five soils by area are included in the table. NRYR = Not rated yield reported and NRNYR = not rated no yield reported.
loam, 0 to 1% slopes; Brennan fine sandy loam, 1 to 3% slopes;
and Raymondville clay loam (Table 1). Raymondville clay loam
ranked number five in area listed under citrus production. It had
a low potential rating for citrus grove establishment; while the
top four soils had a high potential ranking.
Most of the groves appeared on soils that were categorized as
somewhat limited to building development, followed by not
limited and very limited (Table 1). Groves occurring on soils rated
not limited and somewhat limited for building site development
may succumb to urban expansion in the future.
Clay content ranged from 70 to 455 g kg-1. Increases in clay content
represent heavier textured soils. Eighty-two percent of the groves
occurred on soils having clay contents of 186, 203, and 256 g kg-1.
Shrink swell capacity of the soils played a role in determining if
they were appropriate for agricultural production and for building
site development. There was not a pattern between clay content
and potential rating for establishing citrus groves.
Soil chemical data-citrus groves: Citrus production occurred on
strongly alkaline (pH values: 8.5-9.0) to slightly acid soils (pH
values: 6.1-6.5) (Table 1). The majority of the production was
established on moderately alkaline soils (pH values: 7.9-8.4; Table
1). At pH values above 7.8, macronutrients such as phosphorus
266.5
331.6
2271.5
2351.9
2557.6
and micronutrients such as copper, iron, zinc, and manganese
are not readily available for plant uptake. Therefore, producers
have to supply these nutrients to the trees. Macronutrients such
as calcium, magnesium, and potassium would normally be high
in these soils.
The statistics indicated that producers have chosen to grow citrus
trees on non-saline soils (0 to 0.2 S m-1), limiting the effect of
soil salinity on citrus yields. Organic matter content ranged from
0.5 to 20 g kg-1 with the most hectares of citrus appearing on
soils having 16.3 g kg-1 of organic matter. Organic matter totals
seemed to fluctuate; for example, in comparison of the top soils
by area for organic matter content, the values ranged from 6.6 g
kg-1 to 16.3 g kg-1. It is well known by producers in this county
that they have to increase the organic matter content in soils used
for agricultural production.
Spatial distribution of soil potential at the county level for
citrus grove establishment: Fig. 6 shows the spatial distribution
of the soil potential for active citrus grove establishment and
the location of citrus groves within the county. Soils with high
potential for citrus grove establishment occupy most of the county.
Based on this map, producers can expand citrus production into
the north central section of the county. If they have to settle for
8 Applications of GIS to Citriculture in South Texas
Fig. 4. Spatial distribution of active orange orchards within Hidalgo
County.
Fig. 5. Spatial distribution of active lemon, tangelo, and tangerine
orchards in Hidalgo County.
using soils with medium potential for citrus grove establishment,
then areas in the extreme western and the central north east section
of the county are adequate. Soils having low potential would be
their last alternative. These areas are scattered throughout the
county.
Spatial distribution of high soil potential for citrus vs building
site development, urban area, and citrus grove location:
A thematic map showing high potential soils for citrus grove
establishment versus the building development categories are
Fig. 6. Thematic map showing the soil potential for citrus grove
establishment and the location of active citrus groves based on citrus
polygons and soil interactions in Hidalgo County.
Fig. 7. Thematic map showing the high soil potential for citrus grove
establishment vs soil potential for building site development, urban
area identified by the 2000 census, and the location of citrus groves
based on citrus polygons and soil interactions in Hidalgo County. HPC
= high potential for citrus grove establishment, NLB = not limited for
building site development, and SWLB = somewhat limited for building
site development.
shown in Fig. 7. High potential soils were chosen for map
development because they were the soils of choice for growers
to establish citrus groves (Table 1). Combinations mapped were
high potential for citrus grove development and some what
 Applications of GIS to Citriculture in South Texas
limited potential for building site development (92300 ha) and
high potential for citrus grove establishment and the some what
limited for building development combination (71300 ha). None
of the soils were classified into the high potential for citrus
grove establishment and very limited potential for building site
development class. Soils ranked into this category would have
been an ideal choice for establishing citrus groves. Results specify
that prime land for citrus grove establishment can have a rank of
not limited or somewhat limited for building site development.
The urban area 2000 census map overlaid onto the citrus and
soil layers was helpful in indentifying citrus groves appearing in
urban areas (Fig. 7). Thirty-four percent of the citrus production
occurred in urban areas. In these areas, buildings are continuously
being built because of the increase of population. Therefore,
some of these groves may be lost to commercial building over
time. In the future, citrus producers may have to get accustomed
to the concept of urban farming, the growing, processing, and
distribution of food through intensive plant cultivation in and
around cities (Bailkey and Nasr, 1999). This practice could limit
treatments commonly used to control pests in citrus orchards.
To reduce potential losses to urban expansion, it is believed
that new groves should be established as far as possible from
the urban area boundaries. It appears that the northwest central
and the northeast central regions of the county would be ideal
choices for establishing new groves. The overall results of this
study concurred with others in that GIS technology is valuable
for assessing and for managing agricultural resources (Richardson
et al., 1996; Boonyanuphap et al., 2004; Mandal and Sharma,
2006).
Results of this study showed that citrus, SSURGO, and U.S.
census spatial and tabular data integrated with GIS technology
can be a powerful tool for citriculture. This information was
employed to inventory soil chemical and physical properties
9
within citrus groves of Hidalgo County, Texas, to estimate and to
identify orchards that may be affected by urban expansion, and
to select potential sites for establishing new citrus orchards. The
techniques used in this study and findings of the study should
appeal to the citrus industry in the U.S. and abroad.
Acknowledgments: I thank Dr. David Bartells and Mr. Russel
Sheets for supplying the citrus database.
References
Bailkey, M. and J. Nasr, 1999. From brown fields to green fields:
Producing food in North American cities. Community Fd. Security
News, Fall 1999/Winter 2000: 7.
Boonyanuphap, J., D. Wattanachaiyingcharoen and K. Sakurai, 2004.
GIS-based land suitability assessment for Musa (ABB group)
plantation. J. Appl. Hort., 6(1): 3-10.
Mandal, A.K. and R.C. Sharma, 2006. Computerized database of salt-
affected soils for agro-climatic regions in the Indo-Gangetic Plain
of India using GIS. Geocarto Intl., 21(2): 47-57.
National Soil Survey Center, 1995. Soil Survey Geographic (SSURGO)
Data Base: Data Use Information. USDA NRCS Misc. Publ. no.
1527. Natural Resources Conservation Serv., Fort Worth, TX.
Ray, R. and L. Walheim, 1980. Citrus, How to Select, Grow, and Enjoy.
Hort. Publishing Co., Inc. Tucson, AZ.
Richardson, A.J., C.L. Wiegand, G.L. Anderson and A.H. Gerbermann,
1996. Six exemplary applications of GIS technology to subtropical
Texas agriculture and natural resources. Gecarto Intl., 11(1): 49-
60.
USDA. 1981. Soil Survey of Hidalgo County, Texas. U.S. Govt. Printing
Office, Washington, D.C.
Wu, J., M.D. Nellis, M.D. Ransom, K.P. Price and S.L. Egbert, 1997.
Evaluating soil properties of CRP land using remote sensing and GIS
in Finney County, Kansas. J. Soil Water Conserv., 52: 352-358.
Wu, J., M.D. Ransom, G.J. Kluitenberg, M.D. Nellis and H.L. Seyler,
2001. Land-use management using a soil survey geographic database
for Finney County, Kansas. Soil Sci. Soc. Amer. J., 65: 169-177.
 Journal
Journal of Applied Horticulture, 11(1): 10-16, January-June, 2009
Appl

Extracting within-experiment precision of horticultural

experiments useful for meta-analysis

Guido Knapp1, Bimal K. Sinha2 and Dihua Xu2

1
Department of Statistics, TU Dortmund University, 44221 Dortmund, Germany, 2Department of Mathematics and
Statistics, University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA

Abstract
For combining results from independent experiments, it is essential that information about the precision of the estimates of treatment
effects is available. In publications of horticultural experiments, the results of multiple comparisons tests are often reported without
sufficient information about the precision of the experiments. Based on limited information of the precision of an experiment such as
treatments with the same letter are not significantly different, we develop a method for extracting a possible range of the precision of
the experiment which can then be used for meta-analysis. The procedure is demonstrated using a real data example where alternatives
to methyl bromide are studied in pre-plant soil fumigation. We also provide an R program which computes the possible range of the
precision.
Key words: Duncan’s multiple range test, Student-Newman-Keuls multiple range test, Fisher’s LSD test, standardized mean differences,
ratio of means, random effects meta-analysis.

Introduction (2006) simply assumed that the between-experiment variability is

Meta-analysis (MA) has become a common and widely accepted
tool in different spheres of sciences including horticulture for
combining tests of significance as well as methods for comparing
differences between treatments. Olkin and Shaw (1995) presented
methods for the latter, the standardized difference of two normal
means is the effect size of interest and they also presented the
statistical meta-analysis in the fixed effects model, that is, they
assumed homogeneous effect sizes in all the experiments eligible
for meta-analysis. A major obstacle to conducting meta-analysis is
the non-availability of appropriate experimental results. As Olkin
and Shaw (1995) said “The minimum information required for
quantitative research synthesis includes means, sample sizes, and
either standard errors or standard deviations.”
Shaw and Larson (1999) conducted a meta-analysis of strawberry
yield response to preplant soil fumigation with combinations of
methyl bromide-cloropicrin and several alternative systems. They
followed the lines of Olkin and Shaw (1995) and an inclusion
criterion was that means, sample sizes, and either standard errors
or standard deviations of the treatments had to be reported in the
published articles. Since the published results of four studies,
basically eligible for that meta-analysis, lacked these necessary
parameters for inclusion, these four studies were omitted.
The quality of reporting results in the published articles determines
the quality of the meta-analysis. Unfortunately, Olkin and Shaw’s
summation about the minimum information required for meta-
analysis is still often ignored by presenting results of horticultural
experiments in publications nowadays. The lack of information
about the precisions of the estimated means in most published
articles was a major obstacle for the meta-analysis of Porter et
al. (2006) on validating the yield performance of alternatives
to methyl bromide for preplant fumigation. To circumvent the
problem of missing within-experiment precision, Porter et al.
so large compared to within-experiment variability that the latter
one can be ignored in the analysis. Obviously, this assumption
is a very critical one and gives all the different experiments the
same weight in the meta-analysis.
But sometimes the results of multiple testing procedures for
comparing the treatment means are reported using the presentation
that groups of means with the same letter are not significantly
different. For instance, Student-Newman-Keuls’ or Duncan’s
multiple range test are applied in the analysis of horticultural
experiments. In this paper we present a method to demonstrate
how to extract information on the within-experiment precision
when only the results of a multiple range test are reported besides
the estimated means of the treatments.
The paper is organized as follows: In Section 2 we briefly
summarize the basic ideas of multiple range tests. In Section
3 we present the extraction method using a simulated data set.
Section 4 contains the results of the extraction method using the
reported results from two published articles. Since the sample
sizes (number of observations or replicates) are often not available
from the published articles, we present a meta-analytical approach
in Section 5 using the ratio of means as effect size of interest. In
Section 6, we give some concluding remarks and show how to
extract the precision when simultaneous test procedures like Fisher
LSD, Scheffe, or Tukey test were used in the statistical analysis
of the horticultural experiments. In the Appendix, an R code is
given to extract the precision from experiments when the results
of Duncan’s multiple range test or Fisher LSD are known.
Multiple Range Tests
Let Yij, i=1, …, r, j=1, …, n, be r independent samples of n
independently, normally distributed random variables with a common
variance σ2 and expectations E(Yij) = μi, i=1,…r, j=1, ..., n.
 Extracting within-experiment precision of horticultural experiments useful for meta-analysis 11

Following Miller (1981), the basic credo of multiple range tests further testing. All the other null hypotheses are rejected.
is: the difference between any two means in a set of r means Table 2. Grouping for Student-Newman-Keuls (SNK) and Duncan’s
is significant provided that the range of each and every subset multiple range test
which contains the given means is significant according to an Grouping Mean Treatment
αp-level studentized range test where p is the number of means A 512.2 1
in the subset concerned.
A 511.4 2
Let S2 be the error mean sum of squares from one-way analysis of B 492.2 5
variance and assume S2 is a multiple of a χ2-random variate with B 490.7 4
r(n-1) degrees of freedom. The αp-level studentized range test is B 483.9 3
then conducted by comparing the range (divided by S / n) of the p
Now, let us assume that only the results from Table 2 are available.
means involved with the critical value of the studentized
The question is: can we extract any information about the within-
range distribution.
experiment precision, that is, S / n , from this table? Note that
In Student-Newman-Keuls multiple range test, the αp-levels are it holds in the above example
chosen as S 85.5244
= = 2.924456.
p = , p = 2,…, r , (1)
n 10
whereas Duncan’s multiple range test uses, see Miller (1981), Recall that we reject the null hypothesis at level αp if

= 1 − (1 − ) p −1 , p = 2,…, r. Y max − Y min > q

( p) ( p)
p (2) p , r ( n −1) , p , (4)
( p)
S/ n
Given a subset of p means, if Y denotes the largest mean in
max that is, if
( p)
this subset and Y min the smallest one, the corresponding null
Y max − Y min > q p ,r ( n −1), p S / n
( p) ( p)
hypothesis is then rejected at level αp when (5)

Y max − Y min > q

( p) ( p) or if
Y max − Y min > S / n .
( p) ( p)
p , r ( n −1) , p .
(3)
S/ n (6)
q p ,r ( n −1), p
Extracting method using a simulated data set
Conversely, we do not reject the null hypothesis at level αp if
Y max − Y min ≤ q
( p) ( p)
Let us consider a simulated data set with five treatments each (7)
with ten replications. The standard deviation is σ = 10 and we p , r ( n −1) , p ,
S/ n
choose the following true means: μ1 = μ2 = 510 and μ3 = μ4 = that is, if
μ5 = 490.
Y max − Y min ≤ q p ,r ( n −1), p S / n
( p) ( p)
(8)
Then, starting with the global null hypothesis, that is, p = 5, the
or if
Y max − Y min ≤ S / n .
( p) ( p)
following hypotheses are sequentially tested:
H05: μ1 =μ2 =μ3 =μ4 = μ5 q p ,r ( n −1), p (9)
p=5
p=4
4,1 μ =μ =μ =μ
H0 : 1 2 3 4 H4,2 0 : μ2 =μ3 =μ4 = μ5 It is obvious from (5) and (8) that for each subset of p means the
3,1 μ =μ =μ 3,2 μ =μ =μ critical range is identical, namely
p=3 H0 : 1 2 3 H0 : 2 3 4 H3,3 0 : μ3 =μ4 = μ5

p=2
2,1 μ =μ 2,2 μ =μ 2,3 μ =μ
H0 : 1 2 H0 : 2 3 H0 : 3 4 H2,4 0 : μ4 =μ5
q p ,r ( n −1), p S / n .
In case, for instance, the null hypothesis H3,1
0 cannot be rejected at Furthermore, if a hypothesis is rejected at level αp, then we know
level α3, the hypotheses H2,1
2,2
0 and H 0 are also not rejected without from (6) that the standard error S / n is at most
Y max − Y min ,
( p) ( p)
further testing.
q p ,r ( n −1), p
We used the function RAND(’NORMAL’) in SAS 9.1.3 and the
ROUND function to obtain the data given in Table 1. that is, each rejected hypothesis gives an information about an
upper bound of the standard error. Calculating the upper bounds
Using SAS PROC GLM, we obtain an error mean square from all rejected hypotheses and taking the minimum of all
of S 2 = 85.5244. For Student-Newman-Keuls (SNK) and possible upper bounds, we obtain a sharp upper bound of the
Duncan’s multiple range test, we get the same grouping, see within-experiment standard error. To facilitate the computation,
Table 2. The grouping in Table 2 means that treatments with it is sufficient to consider only the rejected hypothesis with the
the same letter are not significantly different. Thus we can smallest range of means for subsets of magnitude p, as this range
conclude that we cannot reject and provides the smallest upper bound for the within-experiment
Consequently, we also accept and without standard error.
Table 1. Simulated data set for five treatments
Treatment Observations
1 506 509 502 514 504 513 514 538 506 516
2 501 513 518 505 524 498 516 512 512 515
3 479 488 482 491 487 479 479 504 482 468
4 489 485 490 498 500 487 496 479 501 482
5 473 496 495 498 498 482 501 491 482 507
 12 Extracting within-experiment precision of horticultural experiments useful for meta-analysis

Conversely, if a hypothesis is not rejected at level αp, then we more informative for the upper bound of the standard error than
know from (8) that the standard error is at least Duncan’s multiple range test.
Y max − Y min ,
( p) ( p)
In the above calculation we have used the information that the
q p ,r ( n −1), p number of replications is ten for each treatment. Sometimes,
that is, each non-rejected hypothesis gives an information about a this information is not available from the published articles and,
lower bound of the standard error. Calculating the lower bounds thus, no information about the error degrees of freedom can be
from all non-rejected hypotheses and taking the maximum of deduced.
all possible lower bounds, we obtain a sharp lower bound of the In Table 3, critical values of Student-Newman-Keuls multiple
within-experiment standard error. To facilitate the computation, range test are presented for p=2(1)10 and several error degrees
it is again sufficient to consider only the non-rejected hypothesis of freedom (ddf). The corresponding critical values of Duncan’s
with the largest range of means for subsets of magnitude p, as this multiple range test are given in Table 4. For given p and increasing
range provides the largest lower bound for the within-experiment ddf, the critical values decrease but converge to a certain value.
standard error. In case no information on the ddf are available, a possible choice
Let us now apply this method to the results from Table 2. The critical would be using the limiting value, that is, Cp, ∞,α which can be,
values of Student-Newman-Keuls multiple range test are for instance, determined with the SAS function PROBMC [e.g.,
q5,45,.05=4.0184, q4,45,.05=3.7727, q3,45,.05=3.4275 and q2,45,.05=2.8484 x = PROBMC(’RANGE’, ., 1- alpha , ., p) with alpha = 1-(1-α)p-1
and the critical values of Duncan’s multiple range test are or alpha = α].
q5,45,.1855=3.1617, q4,45,.1426=3.0919, q3,45,.0975=2.9954 and q2,45,.05=2.8484 Applying the limiting values in the above calculations we obtain
In our example only the hypotheses and the following possible lower bounds :
were tested and not rejected. Consequently, we can extract the H0 Range SNK Duncan
following possible lower bounds: H 30,3 Y 5−Y 3 2.5042 2.8440
H0 Range SNK Duncan H 02,1 Y 1−Y 2 0.2886 0.2886
H 30,3 Y 5−Y 3 2.4216 2.7709 The possible upper bounds are given as
H 02,1 Y 1−Y 2 0.2809 0.2809 p Range SNK Duncan
For the determination of the upper bound, we consider, for each 5 Y 1 − Y 3 7.3361 9.1606
p, the smallest observed range of the rejected null hypotheses 4 Y 1 − Y 3 5.9177 7.1271
and calculate the bounds using the critical values of both tests. 3 Y 1 − Y 4 6.0341 6.8530
This results in 2 Y 3 − Y 5 6.9269 6.9269
p Range SNK Duncan
The resulting range for possible values of the standard error is
5 Y 1 − Y 3 7.0426 8.9509
[2.5004,5.9177] using Student-Newman-Keuls multiple range
4 Y 1 − Y 4 5.6989 6.9536
test, and [2.8440,6.8530] using Duncan’s test. With respect
3 Y 1 − Y 5 5.8351 6.6768 to the upper bound, the use of the limiting critical value is a
2 Y 2 − Y 5 6.7407 6.7407 conservative choice.
The resulting range for possible values of the standard error is
We have explicitly demonstrated the idea of the extraction
[2.4216,5.6989] using Student-Newman-Keuls multiple range
method in this section, but our example is restricted to the
test and [2.7709,6.6768] using Duncan’s test.
situation that the means can be divided into disjunct subgroups
Note that for p > 2, Student-Newman-Keuls multiple range test is of non-significant means (or treatments). In practice, however,
Table 3. Critical values Cp, ddf, 0.05 of Student-Newman-Keuls multiple range test
ddf Number of means p
2 3 4 5 6 7 8 9 10
10 3.1511 3.8768 4.3266 4.6543 4.9120 5.1242 5.3042 5.4605 5.5984
20 2.9500 3.5779 3.9583 4.2319 4.4452 4.6199 4.7676 4.8954 5.0079
50 2.8405 3.4159 3.7584 4.0020 4.1904 4.3437 4.4727 4.5839 4.6814
100 2.8058 3.3646 3.6950 3.9289 4.1093 4.2557 4.3785 4.4842 4.5768
1000 2.7752 3.3194 3.6393 3.8647 4.0379 4.1781 4.2954 4.3962 4.4843
10000 2.7721 3.3150 3.6338 3.8584 4.0309 4.1704 4.2872 4.3875 4.4751
∞ 2.7718 3.3145 3.6332 3.8577 4.0301 4.1696 4.2863 4.3865 4.4741
Table 4. Critical values Cp, ddf, 0.05 of Duncan’s multiple range test
ddf Number of means p
2 3 4 5 6 7 8 9 10
10 3.1511 3.2928 3.3763 3.4297 3.4652 3.4891 3.5052 3.5156 3.5218
20 2.9500 3.0965 3.1896 3.2546 3.3026 3.3392 3.3678 3.3905 3.4086
50 2.8405 2.9876 3.0843 3.1544 3.2082 3.2511 3.2862 3.3155 3.3403
100 2.8058 2.9527 3.0505 3.1217 3.1771 3.2219 3.2590 3.2904 3.3173
1000 2.7752 2.9218 3.0200 3.0925 3.1494 3.1957 3.2345 3.2676 3.2964
10000 2.7721 2.9188 3.0170 3.0896 3.1466 3.1931 3.2320 3.2653 3.2943
∞ 2.7718 2.9184 3.0167 3.0893 3.1463 3.1928 3.2317 3.2651 3.2941
 Extracting within-experiment precision of horticultural experiments useful for meta-analysis 13

the subgroups of non-significant means (or treatments) usually
overlap. In the Appendix a computer program is given written in
R (R Development Core Team, 2008) which can be used to extract
the possible range of standard errors in the general scenario of
overlapping subgroups.
Application to real data examples
Since there is one replicate less in the second year, one expects
that the result of the first year is more precise. This is reflected
in the deduced ranges of the standard error.
Note that we can easily extract the error mean sum of squares
if the number of replicates (blocks) in a completely randomized
block design is known.

We consider some results from Bartual et al. (2002) for A small meta-analysis
demonstrating the extraction method in a real data scenario.
Bartual et al. (2002) studied alternatives to methyl bromide in Olkin and Shaw (1995) used the standardized mean difference
preplant soil fumigation. Seven treatments were investigated, as the effect size for combining differences of treatments from
where treatment 1 is a non treated control, treatment 2 is the several independent experiments. In our scenario, this combination
standard methyl bromide treatment, and treatments 3 to 7 are method is possible if all the experiments provide the number of
alternative treatments. We refer to Bartual et al. (2002) for replicates besides the treatment means and the results from the
details. multiple range test. In case the number of replicates (blocks) is
unknown, we can use the extracted information of the standard
The published article contains the information that the errors to combine the results of several experiments using the ratio
experimental design consisted of two years crop with a complete of means as the effect size for comparing treatments. Combining
randomized block with three replications in the first year. The results using the standardized mean difference or the ratio of
treatments were repeated on the same plot for a second year but
means will be demonstrated in the next two sections. We will
with only two replicates. Duncan’s multiple range tests were done
consider only the combination of two independent experiments.
for statistical comparison among treatments.
Let and be generally two effect size estimates of a common
Several outcomes were measured like marketable yield, first effect size, say θ, with estimated variances and ,
quality fruit yield, first quality fruit size, and percentage of second respectively. Then, the combined estimator of θ is given as
quality fruit yield. The observed outcomes for marketable yield
and first quality fruit size along with the results of Duncan’s
multiple range tests are reproduced in Table 5.
With seven treatments and three replicates in a completely with and wˆ 2 = 1 − wˆ1 .
randomized block design, the denominator degrees of freedom
are ddf = 12 for the first year. Only two replicates in the second Since the extraction method provides a range of possible values
year provide ddf = 6. of the standard errors, we only will consider point estimates of
Table 5. Marketable yield and first quality fruit size in the first and the common effect θ in the present paper. For further statistical
second year of planting inference on θ, we refer to the textbooks of Hedges and Olkin
Treatment Marketable yield First quality fruit size (1985) and Hartung, Knapp, and Sinha (2008), or to the overview
First year Second year First year Second year of Olkin and Shaw (1995).
1 319 C 392 C 17.6 C 17.3 B
2 544 A 738 A 19.4 A 19.5 A Standardized mean difference
3 513 A 683 AB 19.6 A 20.1 A
Recall that the standardized mean difference is defined as
4 562 A 579 AB 18.7 B 18.2 B
μ1- μ2
5 554 A 542 BC 18.6 B 18.2 B δ= σ ,
6 427 B 410 C 18.6 B 17.7 B
7 284 C 193 D 18.4 B 16.1 C where μ1 is the expected value in the first group (treatment),
Applying the extraction method from the previous section, we μ2 the expected value in the second group (treatment), and σ
obtain the standard errors of estimated marketable yield in both the common standard deviation of both groups (treatments).
years as [14.7927,27.910] (first year) and [44.3336,47.1219] An estimator of δ, called Hedges’s g, is given as (Hartung et al.
(second year) (2008))
Y1- Y2
Replacing the denominator degrees of freedom by infinity (∞) g= S ,
yields [16.2432,31.0267] (first year) and [54.4815,57.9080] where is the sample mean in the first group (treatment),
(second year) the sample mean in the second group (treatment), and S the
The interval for the second year is relatively tight, because we pooled standard deviation of both groups (treatments) or
know that the critical range of two means for p = 3 is between the root error mean sum of squares in case of more than two
159 and 169. treatments considered in an experiment. Since g is biased for δ,
an approximately unbiased estimator of δ is given as (Hartung
For the first quality fruit size, the limits of the standard errors are et al. (2008))
[0.0906,0.2272] (first year) and [0.2466,0.3468] (second year)
Replacing the denominator degrees of freedom by infinity (∞)
( 3
g*= 1- 4N - 9 g )
yields [0.0994,0.2525] (first year) and [0.2983,0.4329] (second with N = n1 + n2 and ni, i=1,2, is the number of replicates in the
year) ith group (treatment).
14 Extracting within-experiment precision of horticultural experiments useful for meta-analysis
For the variance of g, it holds that (Hartung et al. (2008))
which can be easily estimated by
where, . possible values of the estimated variances for estimated effect
Let us now consider the estimated marketable yield in the first
and second year from the real data example of Section 4. We first
want to determine the effect of the (active) standard treatment
(No. 2) compared to the control treatment (No. 1). Since the
ranges of the standard errors are [14.7927,27.9101] in the first
year and [44.3336,47.1219] in the second year with knowing the
number of replicates, the ranges for the root error mean sum of
squares are [25.6217,48.3417] (first year) and [62.6972,66.6404]
(second year). Consequently, the true observed standardized mean
differences (Hedges’s g) lies in the range [4.6544,8.7816] (first
year) and [5.1920,5.5186] (second year).
Applying the bias correction to Hedges’s g, see Hedges and
Olkin (1985) or Hartung, Knapp, and Sinha (2008), we obtain
the possible range of observed values for g* as [3.7235,7.0253]
(first year) and [2.9669,31535] (second year)
Note that the larger the estimated standardized mean difference
the larger the estimated variance given a fixed sample size. The
estimated variances for Hedges’s g are [5.9247,19.3844] (first
year) and [225.6405,254.792] (second year)
For the bias corrected estimates g* we obtain the following estimated
variances [4.0318,12.6460] (first year) and [74.3514,83.871]
(second year)
Let us combine the results for the marketable yield of the two
years considering the extreme cases, that is, using the limits of
the extracted intervals. Combining the results using Hedges’s g,
we obtain the following range of possible values of the common
effect size: [4.6682,8.5509]. Using the bias corrected g*’s yields
the range: [3.6846,6.5180].
Ratio of means: Let be the ratio of two (normal) means
and
be the logarithm of ρ. An estimate of ξ is readily given as
Using the delta method, an estimate of the variance of p can be
deduced as
where S2 is the pooled sample variance of both groups (treatments)
or the error mean sum of squares in case of more than two
treatments considered in an experiment. Note that for identical
replications, that is, n1 = n2 = n, the variance estimate reads
and the knowledge of the standard error besides the treatment
means is sufficient to calculate this variance estimate.
Let us combine the results for the first quality fruit size of the
two years for the ratio of means of treatment 2 and 1 using the
extracted standard errors with infinite degrees of freedom, that is,
assuming that the number of replicates is unknown. Recall that
the extracted standard errors are [0.0994,0.2525] (first year) and
[0.2983,0.4389] (second year)
Since the effect size here depends only on means, we obtain, in
contrast to the standardized mean difference, exactly one estimate
of the effect size from each experiment. But we get a range of
sizes. For the first year of the first quality fruit size, we get
with
and for the second year
with
Clearly, the weighted average lies between 0.0974 and 0.1197.
We obtain the minimum value of all possible weighted averages
by using the smallest possible value of and the largest
possible value of . Conversely, the maximum value of all
possible weighted averages is given by using the largest possible
value of and the smallest possible value of .
This leads to the following range of possible estimates of the
common effect size: [0.0985,0.1067].
Backtransforming the results to the original scale we obtain the
range of estimates of ρ as
[exp(0.0985), exp(0.1067)]=[1.1035,1.1126]
Concluding Remarks
In this paper we have demonstrated how information about the
within-experiment variability can be extracted when several
treatments are compared and only the treatment means and the
results of a multiple range test, either Student-Newman-Keuls or
Duncan, are reported. Based on the results of the multiple range
test we can deduce a lower and an upper bound of possible values
of the root error mean sum of squares.
The extracted within-experiment variability can be used in meta-
analysis, when the results of several independent experiments
should be combined. Possible effect sizes are the standardized
mean difference and the ratio of means. For using the standardized
mean difference, the number of replicates has to be additionally
known to determine the common variance estimate or the error
mean sum of squares. Moreover, we can only calculate a range of
possible effect size estimates leading also to a possible range of
variance estimates. Fortunately, there is a one-to-one relationship
between effect size estimate and variance estimate. Using the
ratio of means, we always get one estimate of the effect size per
experiment but a range of possible variance estimates.
The extraction method described in this paper is a little bit
elaborate as the critical ranges of the multiple range tests vary
with the number of means. Since Tukey’s and Scheffe’s multiple
comparisons or Fisher’s LSD tests are simultaneous comparisons
for all treatments and consequently have only one critical range,
the interval of the standard error can be easily determined using
 Extracting within-experiment precision of horticultural experiments useful for meta-analysis
the largest range of two means which do not lead to rejection of
the null hypothesis and the smallest range of two means which
lead to a rejection of the null hypothesis.
To demonstrate the less elaborate extraction method for the
simultaneous multiple comparison methods, let us consider
Fisher’s LSD test. Let and , i ≠ j, be two sample means. Then
Fisher’s LSD test rejects the null hypothesis of equal means if
and tv, α/2 is the upper critical value of tv .
If LSD is explicitly given, then it holds
Since the error degrees of freedom ν are unknown, we approximate
the standard error by
If LSD is not explicitly given but only the information non-
significant then a lower bound of the standard error is
which, when ν is unknown, can again be approximated by
If LSD is not explicitly given but it is known which differences
of means are significant and which ones are not, then we proceed
as follows. Consider the largest range of two means which do not
lead to rejection of the null hypothesis, say , i ≠ j, and the
smallest range of two means which lead to a rejection of the null
hypothesis, say , l ≠ k. Then we have
Consequently,
and we can approximate the sought-after interval by
References
Bartual, R., V. Cebolla, J. Bustos, A.Giner and J.M. Lopez-Aranda,
2002. The Spanish project in alternatives to methyl bromide(2):
The case of strawberry in the area of Valencia. Acta Horticulturae,
567: 431-434.
Hartung, J., G. Knapp and B.K. Sinha, 2008. Statistical Meta-Analysis
with Applications. Wiley, New York.
Hedges, L. and I. Olkin, 1985. Statistical Methods for Meta-Analysis.
Academic Press, Orlando, Fl.
Miller, R.G.J. 1981. Simultaneous Statistical Inference. Springer, New
York.
Olkin, I. and D.V. Shaw, 1995. Meta-analysis and its applications in
horticultural science. HortScience, 30: 1343-1348.
Porter, Ian J., L. Trinder, D. Partington, J. Banks, S. Smith, M. Hannah
and N. Karavarsamis, 2006. Validating the Yield Performance of
Alternatives to Methyl Bromide for Pre-Plant Fumigation. UNEP,
Nairobi, Kenya.
15
R Development Core Team 2008. R: A language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.
org.
Shaw, D.V. and K.D. Larson, 1999. A meta-analysis of strawberry yield
response to preplant soil fumigation with combinations of methyl
bromide-chloropicrin and four alternative systems. HortScience,
34: 830-845.
Appendix: Computer code in R
Extracting the Standard Error from Multiple Range Test
Description: multi is used to extract the standard error from Duncan’s
Multiple Range Test or Fisher’s LSD Test.
Idea: For one study, we compare every possible pair of treatments to
find out if they have at least one letter in common. If this happens, we
do not reject the null hypothesis for that pair of treatments, thus enabling
us to get a lower bound for the standard error. On the other hand, if they
have no letter in common, we get an upper bound.
Usage: multi(y.mean, y.mrt, method, method=c(“Duncan”, “Fisher”),
alpha=.05)
Arguments:
y.mean: a vector of mean values reported in the study.
y.mrt: a vector of statistical comparison results reported in the
study.
method: method used to report comparison results in the study, either
Duncan’s Multiple Range Test or Fisher’s LSD Test. For Fisher’s Test,
we only consider the situation that neither LSD nor “not significant” is
reported.
alpha: level for the reported statistical comparison results, default
is .05
Details:
# Function for comparing two strings
# if they have at least one letter in common return “L”, else return “U”
charcomp <- function(x, y) {
nsep <- nchar(x) + nchar(y)
# summing up number of letters in x, and number of letters in y
xy <- c( strsplit(x,split=character(0))[[1]], strsplit(y,split=character(0
))[[1]])
# combine x and y into a new vector, letters by letters
# for example, x = “abc”, y=”a”, then xy = c(“a”,”b”,”c”,”a”)
nbind <- length(unique(xy))
# report number of unique elements in xy
# for previous example, will be “a”, “b”, “c”, then return 3.
if(nsep==nbind) {return(“U”)}
# the two values are equal means x and y have no letter in common
else {return(“L”)}
}
# Function for extracting the standard error
# from Duncan’s Multiple Range Test or Fisher’s LSD Test
multi <- function(y.mean, y.mrt, method=c(“Duncan”,”Fisher”),
alpha=.05) {
n <- length(y.mean)
n.pair <- choose(n, 2) # number of possible pairs
c.pair <- combn(n, 2) # all possible combinations of n choose 2
bound <- decision <- rep(NA, n.pair)
if(method==”Duncan”) {
y.rank <- 13- rank(y.mean, ties.method=”min”)
for (i in 1:n.pair) {
j <- c.pair[1, i]
k <- c.pair[2, i]
16 Extracting within-experiment precision of horticultural experiments useful for meta-analysis

mean.diff <- abs(y.mean[j]- y.mean[k]) decision[i] <- charcomp(y.mrt[j], y.mrt[k])

rank.diff <- abs(y.rank[j]- y.rank[k]) }}
alpha.p <- 1- (1- alpha)^rank.diff c(max(bound[decision==”L”], na.rm=T), min(bound[decision==”U”],
bound[i] <- mean.diff / qtukey(1-alpha.p, rank.diff+1, 100000) na.rm=T))
decision[i] <- charcomp(y.mrt[j], y.mrt[k]) }
}} Value: The lower bound and upper bound of the standard error will be
if(method==”Fisher”) { returned.
for (i in 1:n.pair) { Example:
j <- c.pair[1, i] yield <- c(392, 738, 683, 579, 542, 410, 193)
k <- c.pair[2, i] dmrt <- c(“c”, “a”, “ab”, “ab”, “bc”, “c”, “d”)
mean.diff <- abs(y.mean[j]- y.mean[k]) multi(yield, dmrt, method=“Duncan“, alpha=0.05)
bound[i] <- mean.diff /(qnorm(1-alpha / 2) * sqrt(2)) [1] 54.48149 57.90800
 Journal
Journal of Applied Horticulture, 11(1): 17-22, January-June, 2009
Appl

Is CropSyst adequate for management-oriented simulation

of growth and yield of processing tomato?

Onofri Andrea*, Beccafichi Catia, Benincasa Paolo, Guiducci Marcello

and Tei Francesco
Department of Agricultural and Environmental Sciences, University of Perugia, Borgo XX giugno, 74, I-06121,
Perugia (Italy). *E-mail: onofri@unipg.it

Abstract
The model CropSyst has proven useful for management-oriented simulations of growth and yield of cereals and other field crops, but
no scientific information is available with reference to processing tomato. The aim of this paper was to parameterise and validate the
crop module of CropSyst for the simulation of potential fruit production in processing transplanted tomato (Lycopersicon esculentum
Mill.). Parameterisation and calibration were performed by using field data from an experiment carried out in 1997 in Central Italy.
The same set of parameters was validated against five independent experiments, carried out on the same location in 1998, 1999, 2000,
2001 and 2002. The simulation of aerial biomass was always very good, with RRMSE values ranging from 7.5 to 13.4% and modelling
efficiencies (EI) always above 0.976. The simulation of LAI was very good during the first part of growing season (up to 40-50 days
after transplanting), while the decreasing trend in the final part of growing cycle was not always reliably simulated. Indeed, RRMSE
for LAI ranged from 13.5 to 26.8% and EI ranged from 0.849 to 0.966. The differences between simulated and observed final fruit
yield were below 10%, except in one year (18% in 2001), confirming the practical value of this model, for management and legislative
purposes. For research purposes, it is confirmed that the simulation of dry matter partitioning is a crucial issue in vegetable crops such
as tomato, wherein the growth of sources and sinks coexists for a main part of crop cycle.
Key words: Processing tomato, CropSyst, simulation, modelling, dry matter partitioning

Introduction
Tomato (Lycopersicon esculentum Mill.) is one of the most
widespread summer crops in Mediterranean environments and
it would be helpful to make predictions of yield, as affected by
farming practices and environmental conditions. Among available
models, CropSyst (Stöckle et al., 2003) may be particularly
useful for practical applications oriented to field management
and decision making, thanks to the very user-friendly interface,
to the possibility of including several management events either
on specific dates or synchronised with crop phenology, and to the
possibility of simulating crop rotations.
With respect to other models, CropSyst introduces several
conceptual simplifications and thus works with a smaller set of
input parameters. For example, the core of the simulation engine
for crop growth is based on two simple functions for radiation-
and transpiration-dependent growth (Stöckle and Nelson, 2003),
which rely on two input parameters, i.e. the light-to-biomass
conversion coefficient (LtBC, as kg MJ-1), and the water-to-
biomass conversion ratio (BTR, as kg m-3 kPa).
The approach to dry matter partitioning is also very simple and
based on one empirical equation, with two main input parameters,
the ‘leaf area/plant biomass’ ratio at the early growth stages (LAR,
as m2 leaves kg-1 plant) and the stem-leaf partition coefficient
(SLP, as m2 kg-1), that accounts for the sharp decline of LAR as
biomass accumulates over time (Stöckle and Nelson, 2003). On
the other hand, dry matter partitioning to commercial yield is
very simply simulated by multiplying final accumulated biomass
by the harvest index (HI), eventually corrected by water stress
during flowering and fruit ripening.
It has been shown that LtBC, BTR, LAR and SLP, together with
other phenological parameters, are those that more strongly affect
simulation results and thus must be chosen with care (Confalonieri
and Bechini, 2004; Donatelli et al., 1997; Pala et al., 1996).
The above-mentioned simplicity may be regarded as an advantage,
because CropSyst is easily parameterised and calibrated. This may
contribute to a high level of diffusion, outside research institutions
and with very practical aims (legislative support, technical advice
and so on). However, the use of CropSyst without an appropriate
validation may easily lead to unreliable conclusions.
This is particularly true for tomato and other vegetable crops that
show several ecophysiological differences with respect to cereals
and other field crops, wherein CropSyst has been more extensively
validated. In particular, the conceptual shortcuts introduced by
CropSyst with reference to dry matter partitioning may hold
for cereals, but may represent a problem in tomato, wherein the
development of sources and sinks overlaps for a main part of
crop cycle. It should not be forgotten that models developed for
specific use in tomato adopt a more complex approach to biomass
partitioning with respect to CropSyst (Van Keulen and Dayan,
1993; Heuvelink, 1996; Scholberg et al., 1997; Heuvelink, 1999;
Ramirez et al., 2004; Boote and Scholberg, 2006). Scientific
information on the reliability of CropSyst simulations with
reference to growth and yield of tomato is not available.
Therefore, the aim of this study was to parameterise and
validate the crop module of CropSyst, by using a series of field
experiments carried out in Central Italy, on conventionally grown
processing tomato.
18 Simulating growth and yield of processing tomato with CropSyst
Materials and methods
Field experiments: Six field experiments were carried out in
processing tomato (cv. PS1296) from 1996 to 1997 and from
1999 to 2002, at Papiano (Perugia, Central Italy, 43°N, 165
m a.s.l.) on a silty-loam soil with 1.3 % organic matter. These
experiments compared development and growth of processing
tomato at different N-fertilisation and density levels, on plots
of 100 m2 size. Results have been already published elsewhere
(Tei et al., 1999, 2001, 2002; Benincasa et al., 2006). From these
experiments, we selected only those experimental treatments
wherein the crop was grown following ordinary practices with
reference to seedbed preparation, transplanting (from 25 May to
3 June), plant density (always 3.2 plants m-2 with rows 0.9-1.2 m
apart), N-fertilisation (always 200 kg N ha-1) and pest control. No
limitations of growth were introduced by nutrient shortage, weeds
or pests. All the details can be found in the cited papers.
In all the experiments 4-6 plants per plot were sampled throughout
the growing season at approximately weekly intervals. At each
sampling date, above ground dry weight and leaf area index
(LAI) were determined. The main phenological indexes were also
recorded as well as final commercial fruit yield at harvest time.
Daily meteorological data (maximum and minimum temperature,
rainfall, global solar radiation and wind speed) were also collected
from a station inside the experimental site.
Parameterisation, calibration and validation of CropSyst:
Simulations were run by using CropSyst version 3.04.08 (29
March 2005). Potential evapotranspiration was estimated by using
the Priestley-Taylor equation, while soil water redistribution was
simulated by the cascade model.
CropSyst was parameterised by using default values, literature
data and the experimental dataset obtained in 1997. A LtBC
value of 2.4 g MJ-1 was chosen from Cavero et al. (1998), that is
perfectly in line with values found by Scholberg et al. (2000a,
b) and Tei et al. (2002). Optimal mean temperature for plant
growth was set at 20 °C (Boote and Scholberg, 2006), while a
value of 0.55 was chosen for the average extinction coefficient
(k), following Ramirez et al. (2004) and Tei and Guiducci
(unpublished data; see also Acock et al., 1978; Jones et al., 1991;
Cavero et al., 1998).
On the other hand, LAR and SLP were estimated by fitting into the
dataset of 1997 the following equation (Stockle et al., 2003):
LAR · DW
LAI = (1)
1+SLP · DW
where DW is the accumulated above-ground biomass (kg m-2),
as recorded with increasing LAI values.
Initial and maximum LAI were taken from the 1997 experiment,
as well as final LAI values with respect to maximum LAI. The
harvest index (HI) was obtained from the same experiment, as the
observed ratio between dry commercial fruit yield and total dry
biomass at final harvest (HI = 0.67). Such values are consistent
with findings of Scholberg et al. (2000a) and Battilani (2006).
Also the phenological parameters required by CropSyst, i.e.
Growing Degree Days [GDD = Tmean - Tbase; where Tmean =
(Tmax + Tmin)/2] for flowering, for reaching maximum LAI and
for physiological maturity, were calculated from data observed
in 1997 (Stöckle and Nelson, 2003). A Tbase of 10°C (Scholberg
et al., 2000a) and a Tcutoff of 35°C (Boote and Scholberg, 2006)
were assumed.
All the other parameters were initially set to default values and
simulations were performed by using the dataset of 1997 to
calibrate the BTR and maximum water daily uptake (mm).
After calibration, the model was validated by applying the
calibrated set of parameters to all the other experiments (1996,
1999, 2000, 2001, 2002). The same set of parameters was always
used, except for the initial LAI, that was regarded as an input
datum for the model and was always set to the observed value.
Such a decision was taken considering that in transplanted tomato
the size of plants at transplanting may be very different from year
to year; thus the same initial LAI does not hold in practice and
may lead to considerable differences in modelling results (Tei et
al., 1996a, b)
The agreement between observed and predicted values was
expressed by using the Relative Root Mean Squared Error
(RRMSE: minimum and optimum value = 0) and the modelling
efficiency (EF: optimal value = 1), as indicated by Martorana
and Bellocchi (1999).
Results and discussion
The whole set of crop parameters as used in simulations is
reported in Table 1. The calibrated value for the water to biomass
conversion coefficient (BTR) is much higher than those reported
in the CropSyst manual (maxima of 6.0 and 8.5 kg m-3 kPa
respectively for C3 and C4 species; Stöckle and Nelson, 2003).
Unfortunately, it was not possible to lower the BTR value below
10.72 kg m-3 kPa without severely degrading the quality of
simulations.
The relationship between LAI and accumulated dry biomass
(Eq. 1) in 1997 (Fig. 1) was used to obtain a reliable estimate of
LAR and SLP as shown by low standard errors (12.194 ± 0.873
and 1.368 ± 0.249).
l
Fig. 1. Relationship between above ground biomass and LAI. Symbols
show observed data (1997), solid line shows fitted curve.
 Simulating growth and yield of processing tomato with CropSyst 19

Aerial biomass (t dm ha-1)

Days after transplanting

Fig. 2. Above-ground total dry biomass of processing tomato during six field experiments. Symbols show
observed data, solid lines show CropSyst simulations by using the parameters reported in Table 1. Vertical lines
show standard errors.
The parameterisation of CropSyst led to a good simulation of (Fig. 3). The time course of LAI is always well simulated at the
aerial biomass (Fig. 2). As expected, the simulation is particularly beginning of crop cycle, approximately until 40-50 days after
good for the 1997 data, that were used for calibration (Table 2), transplanting, while the simulation quality decreases afterwards
but simulations were also good for 1996, 1999, 2000, 2001 and and, in some cases, the maximum LAI and the final decreasing trend
2002 (Table 2). are not fully reproduced by CropSyst. As a consequence, RRMSE
Simulations of LAI are not as good as those for aerial biomass and EF values are less favourable than those observed for aerial
20 Simulating growth and yield of processing tomato with CropSyst

Leaf Area Index

Days after transplanting

Fig. 3. LAI of processing tomato during six field experiments. Symbols show observed data, solid lines show
CropSyst simulations by using the parameters reported in Table 1. Vertical lines show standard errors.
biomass. A similar behaviour was observed when modelling the mechanistic approaches adopted in other models (Marcelis et al.,
growth of sweet pepper (Tei et al., 1996b). This confirms that the 1998; Scholberg et al., 1997; Boote and Scholberg, 2006).
simulation of dry matter partitioning represents a crucial issue with However, it should be emphasized that such deviations from
vegetables characterised by the contemporary growth of sources observed data occur during a period of time when the impact
and sinks (Benincasa et al., 2006). Indeed, the empirical approach on biomass yield may not be very relevant, as it would be at the
used in CropSyst may be too simple with respect to the more beginning of crop cycle. Furthermore, it is important to notice
 Simulating growth and yield of processing tomato with CropSyst 21

Table 1. Crop model parameters for transplanted tomato (cv. PS1296) and source of information
Parameter Ref. 1 Value Units
Photosynthetic pathway - C3 -
Perennial - no -
Land use - row crop -
Aboveground biomass-transpiration coefficient (BTR) 1997 10.72 kg m-3 kPA
Light to aboveground biomass conversion (LtBC) 1997 2.38 g MJ-1
Actual/potential transpiration ratio limiting leaf area growth D 0.95 0-1
Actual to potential transpiration ratio that limits root growth D 0.50 0-1
Optimum mean daily temperature for growth (Topt) L 20 °C
Maximum water uptake D 7 mm d-1
Leaf water potential at the onset of stomatal closure D -700 J kg-1
Wilting leaf water potential D -1200 J kg-1
Maximum rooting depth L 1 m
Initial green LAI 1997 0.010 m2 m-2
Maximum expected Leaf Area Index (LAI) 1997 2.8 m2 m-2
Fraction of maximum LAI at physiological maturity 1997 0.81 0-1
Initial “LAI/DW” ratio (LAR) 1997 11.516 m2 kg-1
Stem/leaf partition (SLP) 1997 2.504 1-10
Leaf Area Duration 1997 575 °C d-1
Extinction coefficient for solar radiation (k) L 0.55 0-1
Leaf duration sensitivity to water stress D 0 0-3
ET crop coefficient at full canopy L 1.05 mm mm-1
Degree days to overcome transplanting crisis 1997 15 °C d-1
Degree days from emergence to flowering 1997 300 °C d-1
Degree days from emergence to maximum LAI 1997 800 °C d-1
Degree days from emergence to begin fruit filling 1997 400 °C d-1
Degree days from emergence to maturity 1997 1300 °C d-1
Base temperature L 10 °C
Cutoff temperature L 35 °C
Phenological sensitivity to water stress D 0 0-3
Unstressed harvest index 1997 0.63 0-1
Table 2. Goodness of simulations expressed as Relative Root Mean when used for management and legislative purposes. This usage
Squared Error (RRMSE) and modelling efficiency (EF) for aerial biomass
and LAI (Fig. 1 and Fig. 2), simulated and measured fruit yield (kg d.m. is encouraged by the relatively small set of parameters required
ha-1) of processing tomato during the six experiments. All the simulations to run the simulations and by the very user-friendly interface. The
are based on the parameterisation in Table 1 set of parameters hereby reported seems to be quite robust for
Year Aerial biomass LAI Simulated Measured the environmental conditions of Central Italy and may serve as a
RRMSE EF RRMSE EF yield yield useful starting point to run simulations in processing tomato.
1996 13.4 0.976 17.3 0.92 6509 6082
1997 8.3 0.991 13.5 0.97 7968 8165 Acknowledgements
1999 9.3 0.998 14.9 0.95 9093 9180 The authors wish to thank Dr. Luca Bechini (University of Milano)
2000 12.4 0.976 17.6 0.94 8999 8257
for critically reviewing the manuscript.
2001 12.4 0.976 15.7 0.96 5706 6716
2002 8.4 0.990 26.8 0.85 6060 6249
References
that the year to year variations on LAI growth were reproduced in Acock, B., D.A. Charles-Edwards, D.J. Fitter, D.W. Hand, L.J. Ludwig,
a satisfactory manner, which is a very interesting and promising J. Warren Wilson and A.C. Whiters, 1978. The contribution of
result. leaves from different levels within a tomato crop to canopy net
photosynthesis: An experimental examination of two canopy models.
In all the cases, the simulation of final fruit yield is very good Journal of Experimental Botany, 29(4): 815-827.
(Table 2) and differences between observed and simulated Battilani, A. 2006. Growth indexes for fertigated processing tomato in a
values are rather small, always below 10% apart from 2001. This Mediterranean sub-humid area. Acta Horticulturae, 724: 103-110.
result follows the relatively constant value of the harvest index Benincasa, P., C. Beccafichi, M. Guiducci and F. Tei, 2006. Source-sink
throughout years, ranging from 0.63 to 0.67. relationship in processing tomato as affected by fruit and nitrogen
availability. Acta Horticulturae, 700: 63-66.
In conclusion, in spite of the scientific shortcuts introduced by Boote, K.J. and J.M.S. Scholberg, 2006. Developing, parameterising and
CropSyst, it seems that, following a careful parameterisation, the testing of dynamic crop growth models for horticultural crops. Acta
practical value of this model may hold also in processing tomato, Horticulturae, 718: 23-34.
22 Simulating growth and yield of processing tomato with CropSyst
Cavero, J., R.E. Plant, G. Shennan, J.R. Williams, J.R. Kiniry and V.W.
Benson, 1998. Application of EPIC model to nitrogen cycling in
irrigated processing tomatoes under different management systems.
Agricultural Systems, 56(4): 391-414.
Confalonieri, R. and L. Bechini, 2004. A preliminary evaluation of
the simulation model CropSyst for alfalfa. European Journal of
Agronomy, 21(2): 223-237.
Donatelli, M., C. Stockle, E. Ceotto and M. Rinaldi, 1997. Evaluation
of CropSyst for cropping systems at two locations of northern and
southern Italy. European Journal of Agronomy, 6(1-2): 35-45.
Heuvelink, E. 1996. Dry matter partitioning in tomato: Validation of a
dynamic simulation model. Annals of Botany, 77(1): 71-80.
Heuvelink, E. 1999. Evaluation of a dynamic simulation model for tomato
crop growth and development. Annals of Botany, 83(4): 413-422.
Jones, J.W., E. Dayan, L.H. Allen, H. Van Keulen and H. Challa, 1991. A
dynamic tomato growth and yield model (TOMGRO). Transactions
of the American Society of Agricultural Engineers, 34: 663-672.
Marcelis, L.F.M., E. Heuvelink and J. Goudriaan, 1998. Modelling
biomass production and yield of horticultural crops: A review.
Scientia Horticulturae, 74: 83-111.
Martorana, F. and G. Bellocchi, 1999. A review of methodologies to
evaluate agroecosystem simulation models. Italian Journal of
Agronomy, 3(1): 19-39.
Pala, M., C.O. Stockle and H.C. Harris, 1996. Simulation of Durum
Wheat (Triticum turgidum ssp. durum) growth under different
water and nitrogen regimes in a Mediterranean environment using
CropSyst. Agricultural Systems, 51(2): 147-163.
Ramirez, A., F. Rodriguez, M. Berenguel, and E. Heuvelink, 2004.
Calibration and validation of complex and simplified tomato
growth models for control purposes in the Southeast of Spain. Acta
Horticulturae, 654: 147-154.
Scholberg, J.M.S, K.J. Boote, J.W. Jones and B.L. McNeal, 1997.
Adaptation of the CROPGRO model to simulate the growth of field-
grown tomato. In: Systems approach for sustainable agricultural
development: Application of systems approach at the field level
(Kropff et al., eds.), Kluwer Academic Publishers, Doordrecht, The
Netherlands, 133-151.
Scholberg, J.M.S., B.L. McNeal, J.W. Jones, K.J. Boote, C.D. Stanley
and T.A. Obreza, 2000a. Growth and canopy characteristics of field
grown tomato. Agronomy Journal, 92: 152-159.
Scholberg, J.M.S, B.L. McNeal, K.J. Boote, J.W. Jones, S.J. Lo Cascio
and S.M. Olson, 2000b. Nitrogen stress effect on growth and
nitrogen accumulation by field grown tomato. Agronomy Journal,
92: 159-167.
Stöckle, C.O., M. Donatelli and R. Nelson, 2003. CropSyst, a cropping
system simulation model. European Journal of Agronomy, 18:
289-307.
Stöckle, C.O. and R. Nelson, 2003. Cropping System Simulation Model:
User’s Manual. Washington State University, Biological Systems
Engineering Department, Washington.
Tei, F., D.P. Aikman and A. Scaife, 1996a. Growth of Lettuce, Onion and
Red Beet. 2. Growth Modelling. Annals of Botany, 78(5): 645-652.
Tei, F., A. Onofri and M. Guiducci, 1996b. Simulation of sweet pepper
growth based on INTERCOM: validation of the model in central
Italy. In: “System analysis and simulation in Agricultural Sciences:
a contribution of the RAISA Project” (Eds F. M. J. Goudriaan and
H. H. van Laar), Firenze, 25-33.
Tei, F., P. Benincasa and M. Guiducci, 1999. Nitrogen fertilisation on
lettuce, processing tomato and sweet pepper: yield, nitrogen uptake
and the risk of nitrate leaching. Acta Horticulturae, 506: 61-67.
Tei, F., P. Benincasa and M. Guiducci, 2001. Determination of a critical
nitrogen dilution curve for processing tomato. In: 2nd International
Symposium on Modelling Cropping Systems, Book of proceedings
(Bindi M., Donatelli M., Porter J., Van Ittersum M.K eds), Florence
(Italy), 33-34.
Tei, F., P. Benincasa and M. Guiducci, 2002. Critical nitrogen
concentration in processing tomato. European Journal of Agronomy,
18: 45-55.
Van Keulen, H. and E. Dayan, 1993. TOMGRO, a greenhouse tomato
simulation model. Wageningen, The Netherlands: Simulation report
CABO-TT No. 29, Wageningen University and Research Centre,
1-48.
 Journal
Journal of Applied Horticulture, 11(1): 23-30, January-June, 2009
Appl

Starch degradation characteristics in relation to

physiological and biochemical properties during growth and
maturation of apple fruit

Manasikan Thammawong* and Osamu Arakawa**

Faculty of Agriculture and Life Science, Hirosaki University, Bunkyocho, Hirosaki 036-8561, Japan.
E-mail: *manasika@hotmail.co.jp, **oarakawa@cc.hirosaki-u.ac.jp

Abstract
Fruit maturity indices, i.e. respiration rate and ethylene production, amylose (AM) and amylopectin (AP) content, total hydrolytic
activity, and sugar content were investigated during the growth and maturation of ‘Tsugaru’ (early-maturing) and ‘Fuji’ (late-maturing)
apples (Malus domestica Borkh.). Different starch degradation characteristics during the growth and maturation processes were
observed between ‘Tsugaru’ and ‘Fuji’. By iodine staining, the loss of starch in ‘Tsugaru’ was observed earlier than in ‘Fuji’. The
different degradation patterns of starch were also demonstrated through the observations on AM and AP content. In ‘Tsugaru’, AM
and AP degraded rapidly between 95 to 110 days after full bloom (DAFB) and almost all starch were lost rapidly at 125 DAFB with
simultaneous increases in rate of respiration and production of ethylene. However, in ‘Fuji’, starch degraded gradually throughout
growth and maturation process and was clearly degraded at 170 DAFB with a low level of ethylene production and decreased respiration.
In both the cultivars, content of AM and AP were highest in the outer cortex and lowest in the inner cortex. Starch degradation was
observed simultaneously in 3 different tissue zones and there was little difference in the total hydrolytic activity among tissue zones
in both cultivars. These results suggest that starch hydrolysis in the apple flesh began simultaneously rather than preferentially in any
one tissue zone. For sugar content, although differences among tissue zones were not clear, it increased distinctly with loss of starch
content. Moreover, sugars from the degradation of accumulated starch and sugar translocation seem to influence mainly the sweetness
quality as the fruit ripens.
Key words: Amylopectin, amylose, Malus domestica Borkh., starch degradation, total hydrolytic activity.

Introduction Although maturity indices of the apple such as respiration rate,

As starch is degraded when the fruit ripens (Blankenship
and Unrath, 1988; Dinar and Stevens, 1981; Prabha and
Bhagyalakshmi, 1998), determination of starch hydrolysis by
iodine staining (SI) is widely used as a maturity index of an apple
to provide an estimation of the starch content (Fan et al., 1995;
Ingle and D’Souza, 1989; Lau, 1988).
The starch index value provides valuable guidance to the level
of fruit maturity and to the appropriate time of harvesting for
immediate consumption or for long-term storage. Despite the
common use of SI to monitor starch loss during fruit maturation,
SI results varied widely between cultivars and it has been reported
that it does not relate well to starch concentration (Fan et al.,
1995; Watkins et al., 1993).
Starch concentration of fruit flesh differed between tissue zones,
with the highest starch content in the outer cortex and the lowest
content in the core (Brookfield et al., 1997; Ohmiya and Kakiuchi,
1990). Apple starch consists of amylose (AM) and amylopectin
(AP) which react differently with I2-KI solution. AM is reacted
most efficiently with I2-KI to produce blue-black pigment
(Fan et al., 1995; McCready and Hassid, 1943). The different
degradation patterns of starch in the fruit cortex may be due to
the different ratio of AM and AP. However, the relationship of
starch degradation pattern and AM or AP content in the fruit flesh
has not been clearly explained.
ethylene production and starch rating had been extensively
studied (Dinar and Stevens, 1981; Kader, 1985; Prabha and
Bhagyalakshmi, 1998; Watkins, 2003), other aspects including
AM and AP composition, sugar content and the enzyme actions,
necessary to understand starch degradation mechanisms and
quality development of fruit flesh, are still poorly understood.
Knowledge of the relationship between these properties would
be valuable for the improvement of fruit quality, not only for
plant physiologists but also for pomologists and fruit growers.
Therefore, the objective of this study was to investigate the
degradation mechanism of starch among different tissue zones
of the apple fruit flesh during growth and maturation in relation
to the physiological and biochemical properties using ‘Tsugaru’,
which produces large amounts of ethylene, and ‘Fuji’, which
produces very low amounts of ethylene.
Materials and methods
Two cultivars of apples, ‘Tsugaru’ fruits from a 6-year-old tree,
and ‘Fuji’ fruits from a 31-year-old tree grafted on Marubakaido
(Malus prunifolia Borkh.) rootstock, were obtained from the
experimental orchard of the Faculty of Agriculture and Life
Science, Hirosaki University, Japan. Fruits were picked on five
to six different days from July to November 2005, at 10-15 days
intervals. The ‘Tsugaru’ fruits were harvested on 70, 80, 95, 110,
125 days after full bloom (DAFB); and the ‘Fuji’ fruits on 95,
24 Starch degradation during growth and maturation of apple fruit
110, 125, 140, 155, 170 DAFB. Fifteen fruits of each cultivar per
harvesting crop were used. The last harvest of each cultivar was
a week before commercial harvest.
Measurements
Respiration rate, ethylene production and starch rating: To
determine CO2 production, each fruit was weighed and sealed
in 1.4 L plastic boxes for 2 h, and 1 mL of headspace gas was
injected into a gas chromatography (model GC-18A, Shimadzu
Co., Ltd., Japan) equipped with a molecular sieve column (60/80
mesh, GL Sciences, Inc.), and a thermal conductivity detector.
Helium was the carrier gas. The injector, oven, and detector
temperatures were set at 60oC. To determine the ethylene content,
1 mL of the headspace gas was removed and injected into a gas
chromatography (model GC-8A, Shimadzu Co., Ltd., Japan)
equipped with an activated alumina column (30/60 mesh, GL
Sciences, Inc.) and a flame ionization detector. Nitrogen was the
carrier gas. The injector, oven, and detector temperatures were
set at 120, 100, and 120oC, respectively.
The starch distribution was measured by dipping an apple slice
taken from the equatorial region in I2-KI solution (10 g/25 g in
1 L distilled water), and the starch-iodine rating was done using
the generic starch-iodine index chart for comparison (Watkins,
2003). This method uses a 1 to 8 scale, with 1 = all starch and
8 = no starch.
Starch, AM, and AP contents: Samples were taken by a cork
borer (1 cm ø), and were peeled, cored, and cut. The fruit flesh
between the core and the outer skin were separated into three
equal portions, with each sample containing 3-5 g of fruit flesh.
The tissues were frozen in liquid N2 and stored at -80oC before
being freeze-dried. Samples were reweighed after freeze-drying,
and then ground in a TI-100 vibrating sample mill (Heiko Sample
Mill, Heiko Seisakusho Ltd., Japan).
Measurements of starch, AM, and AP were obtained through
methods applied in Fan et al. (1995) and Magein and Leurguin
(2000). Briefly, 1 mL of 18% HCl was added to each tube of
ground sample (60 to 80 mg). The ground cortex tissue and 18%
HCl were thoroughly mixed and then incubated at 20oC for 30
min. 10 mL distilled H2O was added to each tube; the supernatant
was mixed and centrifuged at 3,800×g for 10 min. Supernatant of
0.2 to 1.0 mL was used, and 5.0 mL of 1.8% HCl was added to
each tube. After mixing, 200 μL of I2-KI (2 mg I2 and 20 mg KI
per mL H2O) solution was added. Each sample was mixed, and 10
min later, absorbance was measured at 530 nm (for AP) and 606
nm (for AM) using the spectrophotometer (U-2000 Spectronic,
Hitachi, Ltd., Japan).
AM (Sigma Chemical Co., St. Louis) and AP (MP Biomedicals,
Inc., Ohio) standards were dissolved in 18% HCl for 30 min
and then diluted 10-fold. The two standards were mixed in
various ratios, and absorbance was measured at 530 and 606
nm to generate a standard curve. Absorbance coefficients (A),
for AM and AP standards, were AAM606 = 6.88, AAM530 = 4.54,
AAP606 = 5.00, AAP530 = 6.99, and a typical standard curve with
an r2 value of 0.92 was obtained. The amount of AM and starch
concentration (SC) in the samples were calculated using equations
derived by Magel (1991); AP concentration can then be obtained
by subtracting the AM concentration from the SC.
Sugar content: For sugar determination, 100 mg of the dried
sample was extracted three times, each over a duration of 20
min, with 2 mL of 80% (v/v) ethanol at 80oC. The homogenates
were centrifuged at 15,000×g for 10 min to give ethanol-soluble
and ethanol-insoluble fractions. The ethanol soluble fractions
were pooled and evaporated to dryness with a concentrator, and
resolublized in 2 mL of de-ionized water. The soluble fraction
was then filtered. Glucose, fructose, sucrose, and sorbitol were
separated with a high-performance liquid chromatograph (HPLC)
(Shimadzu Co., Ltd., Japan). De-gassed, distilled, de-ionized
water at 1 mL min-1 and 80oC was used as the mobile phase. A
refractive index detector (RI-98, Laboratory System Co., Ltd.,
Tokyo, Japan) was used to quantify sugar’content following the
separation. Sugar content was determined by comparison with
standard samples of known concentration of glucose, fructose,
sucrose, and sorbitol.
Activity of total hydrolytic enzyme: The total hydrolytic
activities of ‘Tsugaru’ and ‘Fuji’ (inner, middle, and outer parts)
were investigated by the method described by Steup (1990).
Briefly, apple flesh (approximately 5 g fresh weight) was
homogenized in 30 mL ice-cold grinding medium which consisted
50 mM 2-(N-morholino)ethane-sulphonic acid (Mes), brought
to pH 6.0 with NaOH, 5mM CaCl2, and 5% (v/v) glycerol. The
homogenate was filtered through several layers of Miracloth
and the filtrate was centrifuged for 15 min at 40,000×g. The
supernatant was passed through a Sephadex G-25 gel which had
been previously equilibrated with grinding medium.
The incubation mixture contained 1 mL 2% (w/v) soluble starch,
0.9 mL grinding medium, and 0.1 mL of the filtered supernatant.
A blank was prepared by mixing 2% (w/v) soluble starch with
an equal volume of grinding medium. Mixtures were incubated
at 30oC. At 5 min, the incubation mixture was added to an equal
volume of alkaline colour reagent, mixed thoroughly, and heated
for 5 min in a boiling water bath. Samples were then cooled to
room temperature and stored for at least 30 min. Absorbance at
546 nm was measured against a reference (2 mL blank plus 2 mL
alkaline colour reagent, treated as above). The alkaline colour
reagent was prepared by dissolving 1 g 3,5-dinitrosalicylic acid
in a mixture of 40 mL 1 N NaOH and approximately 30 mL H2O
at an elevated temperature. Solid potassium sodium tartrate (3g)
was added and dissolved. After cooling to room temperature, the
mixture was brought to a final volume of 100 mL.
For calibration purposes, varying amounts of maltose were reacted
with the alkaline colour reagent. Samples (2 mL each) which
contained 0-2.0 mM maltose, 0.2 mL 2% (w/v) soluble starch and
grinding medium were prepared. Each sample was mixed with 2 mL
alkaline colour reagent and processed as described above. A typical
standard curve of maltose has an r2 value of 0.99 (Fig. 7A).
Data analysis: Analysis of Variance (ANOVA) with Completely
Randomized Design (CRD) using tissue zones as a factor was
performed using SPSS (SPSS, IL, USA), and Tukey’s multiple-
range test was used to test significant difference at the 95%
confidence level of each variable.
Results
Fruit maturity: Starch index value, respiration rate, and ethylene
production were investigated during the growth and maturation
 Starch degradation during growth and maturation of apple fruit
Fig. 1. Starch rating scores (A), iodine staining patterns (B; ‘Tsugaru’ and
C; ‘Fuji’), respiration rate (D), and ethylene production (E) of ‘Tsugaru’
and ‘Fuji’ apples. Each value is the mean of five replicates.
processes of the fruit samples. The starch degradation of fruit flesh
and different degradation patterns between ‘Tsugaru’ and ‘Fuji’
were observed by monitoring the starch index values. Starch in
‘Tsugaru’ started degrading at 95 DAFB, and there was great loss
at 110 to 125 DAFB, but the starch content of ‘Fuji’ degraded
gradually throughout growth and maturation and obvious loss
was observed at 170 DAFB (Fig. 1A).
Starch rating, by iodine staining, was quantified at five and
six stages of growth and maturation for ‘Tsugaru’ and ‘Fuji’,
respectively (Fig. 1B, 1C). The starch-staining pattern of ‘Tsugaru’
showed slight loss of starch at the core area between 80 and 95
DAFB (Fig. 1B-b, 1B-c). Starch degradation was observed all
over the fruit flesh, with starch content of the core area completely
lost at 110 DAFB (Fig. 1B-d) and almost all starch completely
lost at 125 DAFB (Fig. 1B-e). Contrarily, starch in ‘Fuji’ was lost
25
Fig. 2 . Amylose content of ‘Tsugaru’ (A) and ‘Fuji’ (B) apples during
growth and maturation. Each value is the mean of five replicates. Mean
values with the same letter are not significantly different at P=0.05 by
Tukey’s multiple-range test.
gradually during growth and maturation. Starch of the core area
was completely degraded while some loss of starch in the middle
cortex was observed at 125 DAFB (Fig. 1C-c). Almost all starch
was degraded at the last harvest at 170 DAFB (Fig. 1C-f).
The respiration rate of ‘Tsugaru’ decreased initially, followed by a
steep increase between 110 and 125 DAFB. Ethylene production
of ‘Tsugaru’ also increased significantly between 110 to 125
DAFB, demonstrating the climacteric rise of the fruit. However,
in ‘Fuji’, respiration rate decreased gradually and low level of
ethylene production was observed during investigation period
(Fig. 1D, 1E).
AM and AP contents: AM content of the two cultivars was
highest in the outer cortex and lowest in the inner cortex (P≤0.05).
The changes in AM content were observed simultaneously in 3
different fruit tissues; inner, middle, and outer parts. AM content
of ‘Tsugaru’ changed slightly at the first two harvest dates, and
dropped significantly between 95 to110 DAFB. Almost all starch
was degraded at 125 DAFB (Fig. 2A).
However, AM content of ‘Fuji’ was highest at the first harvest
(95 DAFB) and then decreased gradually during the investigation
26 Starch degradation during growth and maturation of apple fruit
Fig. 3. Amylopectin content of ‘Tsugaru’ (A) and ‘Fuji’ (B) apples during
growth and maturation. Each value is the mean of five replicates. Mean
values with the same letter are not significantly different at P=0.05 by
Tukey’s multiple-range test.
period (Fig. 2B). AP content also decreased during growth and
maturation; however, there was no significant difference in AP
content among tissue zones in ‘Tsugaru’ (Fig. 3; P≤0.05).
Sugar content: Total sugar content (sum of sucrose, glucose,
and fructose) of ‘Tsugaru’ and ‘Fuji’ gradually increased during
growth and maturation (Fig. 4). The individual sugar component
of these two cultivars was shown in Fig. 5 and Fig. 6. Although the
difference among tissue zones was not clear, total sugar content
of ‘Tsugaru’ increased during growth and maturation, and it
increased greatly between 110 and 125 DAFB (Fig. 4A). For the
individual sugar content of ‘Tsugaru’, sucrose content was high
between 110 and 125 DAFB, and was highest in the outer cortex
(Fig. 5A). Glucose content increased slightly, and was highest
in the inner tissue zone (Fig. 5B). Fructose and sorbitol contents
changed slightly and the difference among tissue zone was not
clear (Fig. 5C, 5D).
In ‘Fuji’, total sugar and sucrose contents tended to increase
during maturation (Fig. 4B, 6A). The glucose and fructose
contents changed slightly with the highest content in the inner
cortex (Fig. 6B, 6C). The sorbitol content of ‘Fuji’ was higher
than ‘Tsugaru’, and the highest sorbitol content of the inner cortex
was observed between 155 and 170 DAFB (Fig. 6D).
Fig. 4. Total sugar content of ‘Tsugaru’ (A) and ‘Fuji’ (B). Each value
is the mean of five replicates. Mean values with the same letter are not
significantly different at P=0.05 by Tukey’s multiple-range test.
Total hydrolytic activity: The total hydrolytic activity of ‘Tsugaru’
changed slightly between 70 to 95 DAFB, but dropped significantly
at 110 DAFB. In ‘Fuji’, however, the total hydrolytic activity
changed slightly and decreased gradually throughout the growth and
maturation processes. There was a small difference observed in the
tissue zones among both cultivars (Fig. 7B, 7C; P≤0.05).
Discussion
From a previous study, the role of ethylene in starch degradation
of a detached apple fruit was shown to differ between cultivars and
their harvest stages, and is related to ripening and physiological
characteristics of the fruit (Thammawong and Arakawa, 2007).
Although starch degradation of the detached fruit has been studied
in various apple cultivars, the characteristics of starch degradation
in the fruit flesh on the tree during growth and maturation is not
well studied. Moreover, the relationship between cultivar variation
and starch degradation mechanism, which includes the role of
physiological aspects and cellular components such as starch and
sugar content, and enzyme activity in each tissue zone of the fruit
flesh, is still unclear.
Starch index value can be used as a tool to indicate fruit maturity
based on the level of starch degradation and hence indicate the
appropriate time for harvesting (Kingston, 1992; Knee et al.,
1989; Lau, 1988; Watkins, 2003). However, the characteristics
 Starch degradation during growth and maturation of apple fruit
Fig. 5. Sucrose (A), glucose (B), fructose (C), and sorbitol (D) content of ‘Tsugaru’ during growth and development. Each
value is the mean of five replicates. Mean values with the same letter are not significantly different at P=0.05 by Tukey’s
multiple-range test.
of starch patterns shown by iodine staining vary according to
growing conditions, canopy environment, seasonal changes, and
cultivar variation (Reid et al., 1982; Smith et al., 1979; Watkins
et al., 1982; Watkins et al., 1993). From our results, ‘Tsugaru’
and ‘Fuji’ revealed different patterns of starch loss. The loss
of starch in ‘Tsugaru’ (early-maturing cultivar) was rapid and
complete between 110 to 125 DAFB (Fig. 1B). However, iodine
staining of ‘Fuji’ showed slight changes in starch degradation
until commercial harvesting began. This difference between
cultivars suggests that iodine staining is recommended more for
determining the maturation of late-maturing cultivars such as
‘Fuji’ rather than the early-maturing ‘Tsugaru’.
During starch degradation, physiological aspects of the ripening
characteristics were observed to differ between ‘Tsugaru’ and
‘Fuji’. In ‘Tsugaru’, starch content changed slightly between 70-
80 DAFB, but loss of starch content occurred rapidly between
95 to 125 DAFB with simultaneous increases in the rate of
respiration and ethylene production (Fig. 1A, 1D, 1E, 2A). On
the other hand, in ‘Fuji’, degradation of starch occurred gradually
throughout growth and maturation processes with a low level
of ethylene production and decreased respiration. As ethylene
has been suggested to play a role in stimulating physiological
changes and in the conversion of starch to sugar (Kader, 1985;
27
Watkins, 2003), the rapid change of starch content in ‘Tsugaru’
might be due to the induction of internal ethylene production and
respiration rate. Furthermore, the response of the apple fruit to
endogenous and exogenous ethylene for climacteric induction
and starch degradation varied according to cultivar variations
(Thammawong and Arakawa, 2007).
The changing patterns of AM and AP content of the ‘Tsugaru’ and
‘Fuji’ showed the highest accumulated starch in the outer cortex
and lowest in the inner cortex. The action of starch degrading
enzymes has been suggested to play a role in the loss of starch
in the apple fruit (Beck and Ziegler, 1989; Frenkel et al., 1968;
Garcia and Lajolo, 1988; Jackson, 2003; Zhang and Wang, 2002).
Since it has been suggested that starch hydrolysis generally
proceeds from the core (carpel) towards the skin (Poapst et al.,
1959), the degradation of AM and AP content in both cultivars
was observed simultaneously in 3 different regions of the cortex
tissue. However, there was only small difference in the total
hydrolytic activity among different tissue zones in ‘Tsugaru’ and
‘Fuji’ cultivars (Fig. 7B, 7C). This suggests that starch degradation
began simultaneously rather than preferentially in any one
tissue zone. The important factors affecting starch degradation
patterns are probably the amount of starch in the tissue and the
rate of degradation. If starch concentration in the core region
28 Starch degradation during growth and maturation of apple fruit
Fig. 6. Sucrose (A), glucose (B), fructose (C), and sorbitol (D) content of ‘Fuji’ during growth and development. Each value is the mean of five
replicates. Mean values with the same letter are not significantly different at P=0.05 by Tukey’s multiple-range test.
is very low, the lack of iodine staining will occur sooner rather
than later during fruit maturation and ripening. Additionally,
this supports the hypothesis that different degradation patterns
between ‘Tsugaru’ and ‘Fuji’ during maturation might be due to
the physiological aspects of the ripening effect such as rate of
respiration and ethylene production.
As the sweetness of a ripe apple fruit is associated to its cellular
sugar components, sugar from the degradation of accumulated
starch and sugar translocated from the leaves were both taken
into account. Jackson (2003) suggested that starch hydrolysis is
generally accompanied by the appearance of sucrose which is
then slowly hydrolyzed to produce more glucose and fructose.
However, Whiting (1970) suggested that the amount of sucrose
far exceeded the amount produced by starch degradation alone.
Our results revealed that the total sugar content (sucrose,
glucose, and fructose) increased during maturation. The
individual fructose content was the highest in the fruit flesh of
both cultivars; it might have accumulated from the conversion
of translocated sorbitol from leaves by sorbitol dehydrogenase
(SDH, EC 1.1.1.14) (Bieleski and Redgwell, 1985; Teo et al.,
2006). There was a predominant increase of sucrose content that
occurred simultaneously with the loss of starch. Sugar content
increased significantly in ‘Tsugaru’ between 110 and 125 DAFB
while a gradual increase during growth and maturation was
observed in ‘Fuji’. As such, it can be assumed that the high
amount of sucrose in ‘Tsugaru’ may be due to the hydrolysis
of accumulated starch in the fruit tissue. However, an attached
fruit may also gain sucrose from the sugar translocation process
as it is transported to the fruit sink together with sorbitol. The
simultaneous accumulation of sugar from translocation during
starch degradation in the maturation process is clearly supported
by the occurrence of sorbitol translocation in ‘Fuji’ (Fig. 6D).
To clarify this phenomenon, however, further study of sugar
translocation and individual enzyme expression of each cultivar
are required. Moreover, in order to improve fruit eating quality,
the accumulation of sugar from both starch hydrolysis and
translocation should be taken into consideration.
Overall, physiological and biochemical characteristics of the fruit
development and maturation process, including respiration rate,
ethylene production, AM and AP contents, and sugar content,
seem to account conjointly for the degradation pattern of starch in
the fruit flesh. ‘Tsugaru’, an early-maturing cultivar with a short
growth and maturation period, produced a high amount of ethylene
and had increased rate of respiration in order to induce degradation
of starch and production of sugar to develop fruit sweetness as the
fruit ripened. However, in the late-maturing ‘Fuji’, gradual starch
 Starch degradation during growth and maturation of apple fruit
Fig. 7. Standard curve of maltose at different concentrations (A), and
total hydrolytic activity of ‘Tsugaru’ (B) and ‘Fuji’ (C) during growth
and development. Each value is the mean of five replicates. Mean values
with the same letter are not significantly different at P=0.05 by Tukey’s
multiple-range test.
degradation occured simultaneously with low levels of respiration
and ethylene production, and sugar translocation seemed to be
the main factor in sweetness development.
29
Refrences
Beck, E. and P. Ziegler, 1989. Biosynthesis and degradation of starch in
higher plants. Annu. Rev. Plant Physiol. Plant Mol. Biol., 40: 95-117.
Bieleski, R.L. and R.J. Redgwell, 1985. Sorbitol versus sucrose as
photosynthesis and translocation products in developing apricot
leaves. Austra. J. Plant Physiol., 12: 657-668.
Blankenship, S.M. and C.R. Unrath, 1988. Internal ethylene levels and
maturity of ‘Delicious’ apples destined for prompt consumption. J.
Amer. Soc. Hort. Sci., 113: 88-91.
Brookfield, P., P. Murphy, R. Harker and E. Macrae, 1997. Starch
degradation pattern indices; interpretation and relationship to
maturity. Postharvest Biol. Technol., 11: 23-30.
Dinar, M. and M.A. Stevens, 1981. The relationship between starch
accumulation and soluble solids content of tomato fruits. J. Amer.
Soc. Hort. Sci., 106: 415-418.
Fan, X., J.P. Mattheis, M.E. Patterson and J.K. Fellman, 1995. Changes
in amylose and total starch content in ‘Fuji’ apples during maturation.
HortScience, 30: 104-105.
Frenkel, C., I. Klein and D.R. Dilley, 1968. Protein synthesis in relation
to ripening of pome fruits. Plant Physiol., 43: 1146-1153.
Garcia, E. and F.M. Lajolo, 1988. Starch transformation during banana
ripening: the amylase and glucosidase behaviour. J. Food Sci., 53:
1181-1186.
Ingle, M. and M.C. D’Souza, 1989. Fruit characteristics of ‘Red
Deliciuos’ apple strains during maturation and storage. J. Amer. Soc.
Hort. Sci., 114: 776-780.
Jackson, J.E. 2003. Eating quality and its retention. In: Biology of apples
and pears. J.E. Jackson (ed.), Cambridge University Press, New
York, pp. 341-383.
Kader, A.A. 1985. Ethylene-induced senescence and physiological
disorders in harvested horticultural crops. HortScience, 20: 54-57.
Kingston, C.M. 1992. Maturity indices for apple and pear. Hort. Rev.,
13: 407-432.
Knee, M., S.G.S. Hatfield and S.M. Smith, 1989. Evaluation of various
indicators of maturity for harvest of apple fruit intended for long
term storage. J. Hort. Sci., 64: 403-411.
Lau, O.L. 1988. Harvest indices, dessert quality, and storability of
‘Jonagold’ apples in air and controlled atmosphere storage. J. Amer.
Soc. Hort. Sci., 113: 564-569.
Magein, H. and D. Leurguin. 2000. Changes in amylase, amylopectin
and total starch content in ‘Jonagold’ apple fruit during growth and
maturation. Acta Hort., 517: 487-491.
Magel, E. 1991. Qualitative and quantitative determination of starch by
a colorimetric method. Starch, 43: 384-387.
McCready, R.M. and W.Z. Hassid, 1943. The separation and quantitative
estimation of amylose and amylopectin in potato starch. J. Amer.
Chem. Soc., 65: 1154-1157.
Ohmiya, A. and N. Kakiuchi, 1990. Quantitative and morphological
studies on starch of apple fruit during development. J. Jap. Soc.
Hort. Sci., 59: 417-423.
Poapst, P.A., G.M. Ward and W.R. Phillips, 1959. Maturation of McIntosh
apples in relation to starch loss and abscission. Can. J. Plant. Sci.,
39: 257-263.
Prabha, T.N. and N. Bhagyalakshmi, 1998. Carbohydrate metabolism
in ripening banana fruit. Phytochem., 48: 915-919.
Reid, M.S., C.A.S. Padfield, C.B. Watkins and J.E. Harman, 1982. Starch
iodine pattern as a maturity index for ‘Granny Smith’ apples. 1.
Comparison with flesh firmness and soluble solids content. N. Z. J.
Agric. Res., 25: 229-237.
Smith, R.B., E.C. Lougheed, E.W. Franklin and I. McMillan, 1979. The
starch iodine test for determining stage of maturation in apples. Can.
J. Plant Sci., 59: 725-735.
Steup, M. 1990. Starch degrading enzymes. In: Methods in Plant
Biochemistry. Vol. 3; Enzymes of primary metabolism. P. J. Lea
(ed.), Academic Press Limited, London, pp.103-128.
30 Starch degradation during growth and maturation of apple fruit
Teo, G., Y. Suzuki, S.L. Uratsu, B. Lampinen, N. Ormonde, W.K.Hu, T.M.
Dejong and A.M. Dandekar, 2006. Silencing leaf sorbitol synthesis
alters long-distance partitioning and apple fruit quality. Proc. Natl.
Acad. Sci. USA, 103: 18842-18847.
Thammawong, M. and O. Arakawa, 2007. Starch degradation of detached
apple fruit in relation to ripening and ethylene. J. Jap. Soc. Hort.
Sci., 76: 345-350.
Watkins, C.B., M.S. Reid, J.E. Harman and C.A.S. Padfield, 1982. Starch
iodine pattern as a maturity index for ‘Granny Smith’ apples. 2.
Differences between districts and relationship to storage disorders
and yield. N. Z. J. Agric. Res., 25: 587-592.
Watkins, C.B., P.L. Brookfield and F.R. Harker, 1993. Development of
maturity indices for the ‘Fuji’ apple cultivar in relation to watercore
incidence. Acta Hort., 326: 267-275.
Watkins, C.B. 2003. Principles and practices of postharvest handling and
stress. In: Apples: Botany, production and uses. D.C. Ferree and I. J.
Warrington (eds.), CABI Publishing, Massachusetts, pp.585-614.
Whiting, G.C. 1970. Sugars. In: The Biochemistry of Fruits and Their
Products, Vol. 1. A.C. Hulme (ed.), Acad. Press, London, pp.1-32.
Zhang, D. and Y. Wang, 2002. β-amylase in developing apple fruits:
activities, amounts and subcellular localization. Sci. In China (series
C), 45: 429-440.
 Journal
Journal of Applied Horticulture, 11(1): 31-34, January-June, 2009
Appl

Water usage and water use efficiency of drip-irrigated

tomato under deficit irrigation

Berhanu Kebebew1 and Ketema Tilahun2

Oromia Irrigation Development Authority, Addis Ababa, Ethiopia. 2Haramaya University, Ethiopia.
1

Present address: School of Agricultural and Wine Sciences, Charles Sturt University, Locked Bag 588, Wagga Wagga,
NSW 2678, Australia. E-mail: ketematilahun@yahoo.com

Abstract
Efficient irrigation is essential for sustainable use of available water resources. A field experiment was conducted on two tomato cultivars
(Melka Shola and Melkassa Marglobe) and four irrigation deficit levels (0%ETc, 25%ETc, 50%ETc, and 75%ETc). The objective was
to determine crop factor (Kcf) and water use efficiency (WUE). The Kcf values of 0.62, 0.65, 0.70, and 0.71 during the respective four
growth stages of the crop were determined. The highest (91.23 kg ha-1 mm-1) and lowest (81.62 kg ha-1 mm-1) water use efficiencies
were recorded in 25 and 0% deficit levels, respectively. The yield and WUE of Melka Shola cultivar was higher than that of Melkassa
Marglobe. Generally, it was found that irrigating the tomato crop with 75% of ETc (i.e. 25%ETc deficit) is the best irrigation practice
in the area. In terms of both yield and WUE, Melka Shola tomato cultivar was found to perform better than Melkassa Marglobe.
Key words: Crop factor, drip irrigation, Ethiopia, tomato, water use efficiency.

Introduction
Irrigation is one of the most important inputs for agricultural
production. However, limited water resources and increasing
water demands for other uses are causing a decrease in the
quantity of water available for agriculture. The use of water
saving technologies such as drip irrigation is therefore essential.
The possibility of applying water at a very low rates offers the
drip irrigation system the means to deliver water to the soil in
small and frequent quantities at a relatively low cost compared
to other pressurized systems (Cetin et al., 2002).
Irrigation scheduling can be established by using several approaches
such as soil water balance estimates, plant stress indicator and pan
evaporation. Irrigation scheduling with pan evaporation is one
of the irrigation scheduling methods that has been used widely
because of its simplicity and low cost (FAO, 1995). It can also be
operated by farmers. Changes in weather conditions that cause
variation in pan evaporation will have a similar, but not identical,
impact on potential evapotranspiration from a (reference) crop. As
such, pan evaporation (Ep) measurements can be used to estimate
both reference evapotranspiration (ETo) using a pan factor (Kp)
and potential crop evapotranspiration (ETc) using a crop factor
(Kcf) (Paul, 2001). The crop factor use depends on the growth
stage of the crop and crop type.
The upper limit for yield is set by soil fertility, climatic conditions
and management practices. Where all of these are optimal
throughout the growing season, yield reaches the maximum value
as does evapotranspiration. Any significant decrease in soil water
storage from field capacity water content has an impact on water
availability to crops, and subsequently, on evapotranspiration and
yield (Vaux and Pruitt, 1983). In order to increase the productivity
of irrigation water, there is a growing interest in deficit irrigation,
an irrigation practice where water supply is reduced below
maximum level and mild stress is allowed with minimal effects on
yield. Under conditions of scarce water supply and drought, deficit
irrigation can lead to greater economic gains by maximizing yield
per unit of water for a given crop.
In order to supplement the income and nutritional intake of
Ethiopian farmers who live on very fragmented land holdings of
less than a hectare, currently there is a great interest in rain water
harvesting for family-level vegetable production. In an effort to
use the harvested rain water efficiently, gravity drip irrigation
system is also being made available to the farmers on credit basis.
However, there is no documented study on this irrigation package
which can help prepare guideline to be used by the farmers. The
objectives of this study were to determine (i) crop factor using pan
evaporation, and (ii) water use efficiency of drip-irrigated tomato
using deficit irrigation at Awash Melkassa, Ethiopia.
Materials and methods
Site description: The study was conducted at Melkassa
Agricultural Research Center in the Central Rift Valley of
Ethiopia. It is located at 8o24´ N latitude and 39o21´ E longitude
and has an elevation of 1552 m above mean sea level. The mean
annual rainfall is 950 mm. The mean maximum and minimum
temperature is 28 and 14 oC, respectively. The soil of the
experimental site is loam (sand 37%, silt 42%, and clay 21%).
The field capacity and permanent wilting point of the soil is 38%
and 22%, respectively.
Field experiment: Runoff water was harvested in a dome-shaped
under-ground water storage structure. It was then pumped using
a treadle pump into an elevated tanker 1.5 m above ground.
Treadle pump is a simple, low cost, foot-operated water lifting
device. Low head or gravity type drip irrigation set was used
for the experiment. In addition to the pump and elevated water
32 Water usage and water use efficiency of drip-irrigated tomato under deficit irrigation

tanker, the system has a main line, filter, submains, manifiolds, was obtained from Melkassa Agricultural Research Centre
and laterals. Each submain has a length of 3 m and supplies water weather station.
to four drip laterals spaced 0.80 m apart.
Pan factor Kp was determined from Ep and ETo using Eq. (4).
The experimental treatments were factorial, consisting of two Tomato crop coefficients available in literature (Allen et al.,
tomato cultivars (Melka Shola and Melkassa Marglobe) and four 1998) were used to determine potential crop water requirement
irrigation water deficit levels: 0 (optimal), 25, 50, and 75% (most ETc from ETo (Eq. 5). Crop factor Kcf was determined from Ep
stressed) expressed as percentage reductions from the potential and ETc using Eq. (7).
crop water requirement ETc. The treatments were conducted under
ETo = Ep x Kp (4)
three replications. The selected cultivars are known for their
higher yield and disease resistance. The plants were transplanted ETc = ETo x Kc (5)
in plots of 3 x 5 m at 0.30 m plant spacing and 0.80 m row spacing.
ETc = Ep x Kp x Kc (6)
Each treatment plot consisted of four rows of tomato with total
number of 68 plants per plot. Irrigation water was applied equally ETc = Ep x Kcf (7)
for all experimental plots for the duration of plant establishment
after transplanting i.e. for 10 days. There after, the plots were Crop water requirements determined by conventional methods
irrigated with drippers according to their respective treatment should be corrected by a reduction factor Kr for drip irrigation. Kr
levels. Irrigation was carried out with drip emitters of an average = Gc/0.85 or 1.0, whichever is the smallest where Gc is percentage
flow rate 350 mL hr-1 at 1.5 m operating head. ground cover (Vermeiren, 1998). Net irrigation water requirement
(IRn) was calculated as
Fertilizer was applied as per agronomists’ recommendation as:
DAP was side dressed at a rate of 200 kg ha-1 at transplanting and IRn = ETc * K r − R (8)
100 kg ha-1 urea was applied in split at transplanting and 45 days where R is rainfall during the growing period. Gross irrigation
after transplanting. To protect disease infection, Ridomil Gold RZ water requirement (IRg) can be calculated as
63% was applied as 3.5 kg ha-1. For insect protection Cypermethin
ETc * K r (9)
or Karate was used at a rate of 100 g ha-1. IR g = −R
Ea
Dripper characterization: Dripper emission uniformity (EU)
was calculated as follows Ep * K p * Kc * Kr
IR g = −R (10)
Ea
⎛q ⎞ (1)
EU = ⎜⎜ n ⎟⎟100
⎝ qa ⎠ E p * K cf * K r * A (11)
IR g = −R
where EU = emission uniformity (%), qn = average low quarter Ea
emitter flow rate (L h-1), qa = average emitter flow rate (L h-1).
where Ea = irrigation system application efficiency, A = the wetted
The application efficiency of the drippers was calculated as area allocated to each plant.
(Vermeiren, 1998)
Water use efficiency: Crops’ response to deficit irrigation is
E a = K s * EU (2) different due to the difference in their ability to tolerate water
where, Ks is a coefficient which expresses the storage efficiency of stress during their growth period (Vaux and Pruitt, 1983). Crop
the soil as (average water stored in the root volume/average water yield response to water deficit can be described (Doorenbos and
applied). It takes into account the losses of drip irrigation water Kassam, 1979) as:
application (Ks = 1 or 100% for loam soil) (Vermeiren, 1998).
Ya ⎡ ETa ⎤ (12)
1− = Ky ⎢1 − ⎥
Crop water requirement and irrigation requirement: Pan Ym ⎣ ETm ⎦
evaporation Ep was measured using pan evaporimeter installed
where Ya = actual yield (kg ha-1), Ym = maximum yield (kg
just by the side of the experimental plots. Reference crop
ha-1), ETa = actual evapotranspiration (mm), ETm = maximum
evapotranspiration ETo was determined using FAO Penman
evapotranspiration (mm) or ETc, and K y = yield response
Monteith equation (Allen et al., 1998) as:
factor.
900
0.408Δ( Rn − G ) + U 2 (ea − ed ) The crop water use efficiency, WUE, can be determined as
ETo = T + 273 (3)
[Δ + (1 + 0.34U 2 )] ⎡ ⎤
Where, ETo = reference crop evapotranspiration (mm day-1), Y ⎢ (K y − 1)⎥⎥ Ym (13)
Rn = net radiation at crop surface (MJ m-2 day-1), WUE = a = ⎢
ETa ⎢ ETa ⎥ ETm
G = soil heat flux (MJ m-2 day-1), ⎢ ET ⎥
T = average temperature (oC), ⎣ m ⎦
U2 = wind speed measured at 2m height (m/s), Field data collection: To determine plant growth, water use and
ea-ed = vapor pressure deficit (kpa), water use efficiency, plants were marked at random locations
Δ = Slope vapor pressure curve (kpa oC-1), along the entire plots in each replication. Two plants per row
γ = Psychometric constant (kpa oC-1), were marked, giving a sample size of eight plants per plot for
900 = conversion factor. Daily weather data used in Eq. (3) both tomato cultivars. Plant heights were measured every week
 Water usage and water use efficiency of drip-irrigated tomato under deficit irrigation 33

starting from the beginning of the treatment period to the end
of the midseason of growth stage. The two middle rows of each
treatment were harvested and weighed at the end of the season.
Results and discussion
Dripper characterization: The emitter flow rate collected from
randomly selected drippers was used to calculate the performance
of the irrigation system. The average low quarter flow rate (9 out
of 36 sample flow rates collected in this study) was 311 mL h-1 and
the average flow rate was 329 mL h-1. The emission uniformity
EU was calculated to be 95% (Eq. 1). For the loam soil with Ks
= 1, the application efficiency was found to be 95% (Eq. 2).
Crop water use: The crop water requirement and irrigation
requirement for the tomato crop during the growing season
determined on two-day basis is presented in Table 1a-d for the
months of October, November, December and January. Eqs. 3,
4, 5, 7 and 8 were used to calculate ETo, Kp, ETc, Kcf, and IRn.
The crop coefficient Kc values were interpolated for the growing
stages of the tomato crop from Doorenbos and Pruitt (1977).
During the crop establishment period (until 28 October), equal
amount of water (67 mm) was applied to all the treatment plots.
The total amount of water applied to the non-stressed (0% deficit)
treatment is 528 mm (Table 1a-d) + 67 mm = 595 mm. The amount
of water applied to the 25, 50, and 75% deficit levels is 467 mm,
331 mm, and 199 mm, respectively.
The pan factor Kp and tomato crop factors Kcf determined for
the growth stages and the total growing season are presented
in Table 2. The Kcf values are of great importance for farmers
interested to optimally schedule the limited harvested water using
pan evaporation data. The Kcf value varies during the growing
season with the average seasonal value indicating that about
70% of pan evaporation can be considered as potential crop
evapotranspiration.
The water use efficiency of the experiment was significantly

Table 1. Two-day average tomato water requirement at Melkassa Agricultural Research Center
a. October (mm day-1)**
Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31
ETo 6.00 5.59 6.56 6.40 6.49 6.73 6.33 6.43
Ep 8.95 7.85 9.18 9.68 9.87 10.98 8.60 8.84
Kp 0.67 0.71 0.71 0.66 0.67 0.62 0.74 0.73
Kc 0.75 0.79 0.81 0.83 0.85 0.87 0.89 0.91
ETc 0.50 0.56 0.58 0.55 0.57 0.54 0.65 0.66
Kcf 4.47 4.40 5.32 5.42 5.60 5.93 5.59 5.83
R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
IRn 4.82 4.75 5.75 5.85 6.05 6.40 6.03* 6.30
* The day on which irrigation treatments started and total amount of water applied upto this day was 67.12 mm.
b. November (mm day-1)
Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31
ETo 5.59 6.61 6.40 6.44 6.44 6.36 5.57 4.00 4.53 4.55 5.07 6.57 7.14 5.62 5.94
Ep 8.48 10.17 9.77 9.72 9.30 9.51 7.75 6.16 5.22 7.79 7.28 10.40 10.01 9.02 8.65
Kp 0.66 0.65 0.65 0.66 0.69 0.67 0.72 0.65 0.87 0.58 0.70 0.63 0.71 0.62 0.69
Kc 0.93 0.95 0.97 0.99 1.01 1.03 1.05 1.07 1.09 1.11 1.13 1.15 1.15 1.15 1.15
ETc 0.61 0.62 0.64 0.66 0.70 0.70 0.76 0.68 0.95 0.65 0.79 0.73 0.82 0.72 0.90
Kcf 5.17 6.30 5.95 6.54 6.51 6.66 5.89 4.19 4.96 5.06 5.75 7.59 8.20 6.49 7.79
R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.60 4.80 0.00 0.00 0.00 0.00 0.00
IRn 5.58 6.80 6.43 7.07 7.03 7.19 6.36 4.53 5.35 5.47 6.21 8.20 8.86 7.01 8.40
c. December (mm day-1)
Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31
ETo 5.62 5.16 5.41 5.77 5.58 6.17 5.57 5.48 5.20 5.95 5.34 4.86 5.73 5.92 4.94 5.43
Ep 8.74 8.97 8.10 8.86 8.13 9.56 8.66 7.97 9.07 8.13 7.81 7.87 8.25 9.24 7.02 8.08
Kp 0.65 0.58 0.67 0.65 0.69 0.64 0.64 0.69 0.57 0.73 0.68 0.62 0.70 0.64 0.70 0.67
Kc 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15
ETc 0.74 0.66 0.77 0.75 0.79 0.73 0.74 0.79 0.66 0.84 0.79 0.71 0.80 0.74 0.81 0.77
Kcf 6.47 5.92 6.24 6.65 6.42 6.98 6.37 6.29 5.98 6.84 6.17 5.59 6.60 6.84 5.69 6.22
R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
IRn 6.98 6.39 6.74 7.18 6.94 7.54 6.55 6.80 6.46 7.39 6.66 6.03 7.13 7.38 6.14 6.72
d. January (mm day-1)
Date 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 31
ETo 5.79 6.09 6.56 5.12 5.59 5.53 6.30 6.60 6.16
Ep 8.78 8.14 7.94 7.64 8.13 8.27 8.55 7.91 8.43
Kp 0.66 0.75 0.70 0.67 0.69 0.67 0.74 0.83 0.73
Kc 1.15 1.06 1.06 1.02 0.98 0.93 0.89 0.84 0.80
ETc 0.76 0.79 0.74 0.68 0.67 0.62 0.65 0.70 0.58
Kcf 6.67 6.46 5.88 5.22 5.45 5.12 5.56 5.54 4.89
R 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
IRn 7.29 6.97 6.35 5.64 5.88 5.53 6.00 5.98 5.28
** The values given in this table are average of two consecutive days expressed as mm day-1 for concise presentation of the data.
34 Water usage and water use efficiency of drip-irrigated tomato under deficit irrigation

Table 2. Average values of the crop water use and crop factor parameters Table 5. Average plant growth performance (mm) for deficit irrigated
during the growth stages of tomato tomato cultivars
Parameter Crop growth stages Character Deficit level (%)
Development Mid Late Total 0 25 50 75
season season growing Melka Shola 358 380 357 314
season Melkassa Marglobe 468 453 368 297
ETo (mm day-1) 6.12 5.56 6.30 5.99
Ep (mm day-1) 9.05 8.44 8.18 8.57 Pan evaporation data can be used to calculate potential crop
Kp (mm day-1) 0.68 0.66 0.77 0.69 evapotranspiration once crop factor is determined. In this study,
ETc (mm day )-1
5.60 5.50 5.73 5.61 crop factor values of 0.62, 0.65, 0.70 and 0.71, respectively were
Kcf 0.62 0.65 0.70 0.71 determined for crop development, midseason, late season, and
total growing season of tomato. As the water stress level increases,
different (P<0.05) for tomato cultivars (Table 4). The WUE of the yield of tomato decreases. The 25%ETc deficit level resulted in
25% ETc deficit was significantly different from the others (Table the highest water use efficiency. Melka Shola cultivar has higher
3). In both cases, the value was lower for irrigation treatment yield and water use efficiency compared to Melkassa Marglobe
with high amount of water application. The 25% deficit level cultivar. Besides being less water stress resistant, Melkassa
has got higher water use efficiency which was 1.12 times that of Marglobe was easily susceptible to leaf diseases and bacterial
fully irrigated (0% deficit level) and the 75% deficit follows it wilting during the early growth stages. Yield response factor of
with relative water use efficiency value of 1.06 times that of fully Melka Shola is lower than that of Melkassa Marglobe. In the
irrigated tomato (Table 4).
scenario of water shortage, the use of 25%ETc deficit and Melka
Table 3. Influence of moisture deficit levels on yield, water use efficiency Shola cultivar can result in higher water use efficiency.
of tomato
Parameter Deficit level (%)
0 25 50 75
Acknowldegement
Total yield (t ha-1) 45.113a 41.555a 24.685b 15.757c The authors gratefully acknowledge Dr. Dawit Zerihun and
WUE (kg ha-1 mm-1) 81.62a 91.23b 82.96a 86.11a Dr. Tirusew Assefa for suggestions and improvement in the
Relative WUE 1.12 1.02 1.06
Ky 0.80b 1.07a 1.08a
manuscript.

Table 4. Yield, water use efficiency and yield response factor of two
Reference
cultivars of tomato Allen, R.G., L.S. Pereira, D. Raes and M. Smith, 1998. Crop
Characters Tomato cultivars evapotranspitation. FAO Irrigation and Drainage Paper, No. 56,
Melka Shola Melkassa Marglobe Rome.
Total yield (t ha )
-1
32.688 a
30.868a Cetin, O., O. Yildirim, D. Uygan and H. Boyaci, 2002. Irrigation
WUE (kg ha-1 mm-1) 91.67a 79.04b scheduling of drip-irrigated tomatoes using class A pan evaporation.
Ky 0.69 a
0.78a Turkish J. Agric. For., 26: 171-178.
* Means within each row followed by the same letter are not statistically Doorenbos, J. and A.H. Kassam, 1979. Yield response to water. Irrigation
significant at the 1% level according to Duncan’s multiple range test. and Drainage Paper, No. 33. FAO. Rome.
As crop yield response factor (Ky) increases, the crop water use Doorenbos, J. and W.O. Pruitt, 1977. Guidelines for predicting crop
efficiency decreases, which in turn implies that benefit from deficit water requirements. Irrigation and Drainage Paper, No. 24. FAO.
Rome.
irrigation is unlikely. Only those crops and growth stages with a
lower crop yield response factor (Ky < 1) can generate significant FAO. 1995. Irrigation in Africa in Figs.. Water Report No. 7. FAO.
Rome.
savings in irrigation water through deficit irrigation. Seasonal Ky
value for tomato is 1.05 (Doorenbos and Kassam, 1979) which is Paul, E. 2001. Regulated deficit irrigation and partial root zone drying.
Land and Water Resources Department, Australia.
close to the Ky value of the 50% depletion level in this study.
Vaux, H.J. and W.O. Pruitt, 1983. Crop-water production. Advances in
From Table 5, it can be seen that the height of tomato was Irrigation, 2: 61-93.
influenced by water deficit level with a general decrease in crop Vermeiren, I. 1998. Localized irrigation design, installation, operation,
height as water stress increases. evaluation. FAO Irrigation and Drainage Paper, No. 36. Rome.
 Journal
Journal of Applied Horticulture, 11(1): 35-40, January-June, 2009
Appl

Sucrose synthase and acid invertase activities in relation to

the floral structures abortion in pepper (Capsicum annuum L.)
grown under low night temperature

Néji Tarchoun1*, Salah Rezgui2 and Abdelaziz Mougou2

1
Centre Régional des Recherches en Horticulture et Agriculture Biologique BP47- 4042 Sousse, Tunisia,
2
Institut National Agronomique de Tunisie (INAT) 43, av. Charles Nicolle 1082 cité mahrajène Tunis- Tunisia.
*E-mail: ntarchoun@yahoo.fr/ tarchoun.neji@inat.agrinet.tn

Abstract
Effects of low night temperature were investigated on two local hot pepper varieties (‘Beldi’ and ‘Baklouti’) grown at day/night
temperature of either low night temperature regime (25°C/10°C) or optimum night temperature regime (25°C/20°C). The negative
effect of low night temperature on floral structure differentiation was registered on both varieties. The deleterious effect was more
sensitive on bud stage than on flower buds stage. Abortion of these structures was less important in ‘Beldi’ than in ‘Baklouti’. Floral
structure abortion induced by low night temperature was negatively and significantly correlated with soluble acid invertase activity
on ‘Beldi’ (r=-0.82), while on ‘Baklouti’, both sucrose synthase and insoluble acid invertase activities were correlated with floral
abortion (r=-0.78). Under low night temperatures, sucrose synthase and soluble acid invertase activities were reduced to 50%, while
the insoluble acid invertase activity was reduced by more than 90%. Enzymatic activities and flowers abortion correlation show a
differential response between these two parameters and the developmental stages of flowers.
Key words: Abortion, bud and flower, hot pepper, low night temperature, sucrose synthase, acid invertase.

Introduction Sucrose synthase and acid invertase were found to regulate

Flowers retention and fruit set are highly sensitive to environmental
factors in many species (Van Doorn and Stead, 1997). High
temperature, drought, low light condition or failure of pollination
are important factors that may induce abortion (Aloni et al.,
1997; Heuvelink and Korner, 2001; Marcelis et al., 2004). These
environmental stresses may alter photosynthetic activity (Havaux,
1993; Kitroongrung et al., 1992) and carbohydrate partitioning
(Aloni et al., 1991a; Schaffer et al., 1987) causing an imbalance
between source-sink structures (Geiger et al., 1996; Minchin and
Thorpe, 1996).
Relationships between carbohydrate partitioning and floral
structures are studied in few species. Photosynthetic competition
between reproductive and vegetative sink seems to play a role
in the abortion of Vecia faba flowers (Aufhammer et al., 1987).
Turner and Wien (1994) and Aloni et al. (1996) indicated that
competition for assimilates between flowers and adjacent young
leaves may partially determine flowers abortion in pepper. Then,
abortion seems to be dependant not only on the source strength but
also on sink strength (assimilate demand) of competing organs.
Several studies indicate that reproductive organs abortion depends
on differentiation stages of these organs and the stress type. In
fact, under shade conditions, Wien et al. (1989) noted that open
flowers were the most susceptible organs to abortion, while Aloni
et al. (1991b), applying heat stress on sweet pepper, concluded
that flower buds were more susceptible to abortion. Applying
shade and heat stress at different stages of flower differentiation,
Marcelis et al. (2004) noted that flowers/fruits of sweet pepper
were susceptible to abortion few days before anthesis.
phloem unloading (Geiger and Servaites, 1991). The magnitude
of the activities of both enzymes was suggested as a reliable
measurement of sink strength (Black, 1993; Jenner and Hawker,
1993) but was highly dependant on environmental stress (Roitsch,
1999; Sturm and Tang, 1999). Complex mechanisms have to be
assumed which integrate the expression of the enzymes involved
in carbohydrates production in source tissue and the regulation
of source-sink relationship (Roitsch, 1999).
Investigation on low night temperature effect on abortion
of vegetable floral structures is however scarce; the present
experiment was conducted to determine the effect of low night
temperature on sucrose synthase and acid invertase activities in
relation to the floral structures abortion in hot pepper.
Material and methods
Plant material and growth conditions: Seeds of two hot pepper
varieties (‘Beldi’ and ‘Baklouti’ from INRAT, Tunisia) were
sown in elevated trays containing fertilized peat (NPK, 12-
14-24) and allowed to germinate in a growth chamber at 25°C
± 2°C. Seedlings, at 6- 8 leaf stage, were transplanted into 12
liters pots, one plant per pot, containing the same substrate with
a layer of about 2 cm of clay (Argex 4/10) and transferred to
growth chamber with day/night temperatures of either low night
temperature regime (25°C/10°C) or optimum temperature regime
(25°C/20°C). The photoperiod was 16 h with light intensity of
250 ± 5 μmol m-2 s-1 (PAR). The relative humidity was maintained
at about 70 ± 5%. Ten plants per variety were placed, at random,
in each chamber and were watered when needed and fertigated
with Nutri chem (N:P:K 22:5: 11) at 1 g L-1.
36 Relationships between pepper flower abortion and enzymes activity under low temperature
Bud
1 2 3 Soluble and insoluble acid invertase activities were determined
Fig. 1. Floral growth stage based on the dimension and morphological
criteria: bud (stage A) [1] , flower bud (stage B) [2] and flower bud (stage
C) [3] corresponding Fig.1.
Floral structures sensitivity to low temperature (experiment
1): Evaluation of the floral structures sensitive to low night
temperature was performed by the transfer of 3 plants of each
variety from optimal to low night temperature regime (25/20°C
to 25/10°C) and inversely. These plants were acclimated at
least seven days in either condition prior to transfer to growing
condition. Three developmental stages were carried out using a
dimensional and morphological criteria:
- Stage A corresponding to the bud stage,
- Stage B corresponding to the floral bud with 3-4 mm diameter and 4-5
mm height (green petals welded to sepals, 4-5 days before anthesis),
- Stage C corresponding to floral buds with diameter ≥ 4 mm, height ≥
5 mm (white petals lightly welded to sepals, 2-3 days before anthesis)
(Fig. 1).
Under low night temperature, these three floral structures present,
usually, a highly abortion percentage compared to the flowers at
anthesis stage; so the late developmental stage was not considered
in this experiment.
Based on the developmental period of either stage (Tarchoun,
2003), the transfer period of plants was determined to seven
nights. Before transferring the plants of ‘Beldi’ and ‘Baklouti’
varieties, 10 to 15 buds and 6-10 floral buds for each stage were
studied to follow their evolution. The treatment was repeated
three times.
Effect of temperature was evaluated using abortion percentage of
buds (stage A) while abortion percentage of flower buds (stages
B and C) was calculated based on flower buds development of
stage B to stage C and the flower buds development of the stage
C to pre-anthesis flowers (stage D). The abortion percentage was
calculated three and five days after transfer.
Enzyme activity essays (experiment 2): Determination of
sucrose synthase and acid invertase activities was carried out on
floral structures at stage A, B and C and compared to the ovary
at anthesis stage. Tissue samples of 200-300 mg fresh weight,
collected in morning from each growth chamber, were ground
in the mortar containing 3 mL of medium extraction: 25 mM
HEPES buffer (N-2hydroxyethylpepirazine-N2-2ethansulphonic
acid) pH 7.2; 2 mM DDT (DL-Dithiothreitol) ; 5 mM MgCl2
and 3 mM DIDCA (diethyldithiocarbamic acid) as antioxidant
(Aloni et al., 1991b).
Sucrose synthase activity, in cleavage sense, was determined as
follows: After extraction, the mixture was centrifuged at 10000
g for 30 min at 4°C. Aliquots of 200 μL of the supernatant were
incubated for 30 min at 37°C in the medium containing 200 mM
sucrose, 5 mM UDP, 50 mM HEPES buffer (pH 7). Reaction was
stopped after boiling for 3 min. Fructose content was determined
by the addition of 3 mL of dinitrosalicylic acid reagent (1%
dinitrosalicylic acid, 0.2% phenol, 0.05% bisulfate sodium and
1% NAOH) according to Schaffer et al. (1987). To reactivate
the reaction, tubes were placed on boiling water bath for 5 min;
before cooling, 1 mL of Rochelle salt (potassium sodium tartrate)
was added to stabilize the reaction (Miller, 1959).
by comparable extraction procedure as described for sucrose
synthase and the supernatant was collected and placed at 0°C.
The pellet was washed three times by the same buffer and
centrifuged at 10000 g at 4°C, the supernatant was eliminated
in each centrifugation. The later pellet was suspended in 3mL of
the extraction medium added to NaCl (1M) and placed at 0°C
for 1 hour.
Aliquots of 100μL of both the supernatant (soluble acid invertase)
and the suspended pellet (insoluble cell wall acid invertase) were
incubated in 1 mL 0.1N phosphate citrate buffer (pH 5) and 20
mM sucrose. The incubation was carried out for 1h at 37°C and
was terminated by the addition of 1 mL dinitrosalicylic acid
reagent (Aloni et al., 1991b). After boiling for 5 min, 1 mL of
Rochelle salt was added (Miller, 1959). Sucrose synthase and
acid invertase activities were determined colorimetrically at 580
nm of the optic density.
Statistical analysis: The experiment was carried out as split-split-
plot design with temperature as main factor and varieties as the
second factor (sub-plot) and the floral structures represent the
smallest experimental unit (sub-sub-plot). Analysis of variance
was performed by SAS system (1985), means were separated by
LSD, and floral structures abortion was analyzed per date. The
relationship between enzymatic activity and floral structures
abortion was estimated by Pearson correlation coefficient using
proc corr of SAS (1985).
Results
Experiment 1
Buds sensitivity to low night temperature: Fig. 2 illustrates
the buds development to flowers-buds on ‘Beldi’ and ‘Baklouti’
varieties after the transfer of plants from 25/20°C to 25/10°C
and vice versa. Temperature effect was significant from the
third day of treatment and was stabilized at the fifth day for both
temperatures regimes and varieties. The transfer from optimal
night temperature to low thermal-conditions caused an increase
in buds abortion on the two varieties especially on ‘Baklouti’.
Fifteen percent of buds aborted after three days and reached 82
and 88% at the fifth day for ‘Beldi’ and ‘Baklouti’, respectively
(Fig. 2.1). The reciprocal transfer favoured significant decrease
of buds abortion and the abortion rate depended on varieties
(Fig. 2.2).
Flower buds sensitivity to low night temperature: The
development evolution of flower buds is represented in Fig. 3.
Transferring pepper plants from 25/20°C to 25/10°C and 3 days
after treatment, flower buds expressed increased abortion of 56
and 69% for ‘Beldi’ and ‘Baklouti’ respectively. This abortion was
more important at the fifth day of treatment and reached 86 to 91%
for ‘Beldi’ and ‘Baklouti’, respectively (Fig. 3.1). Although at the
end of treatment period, flower buds at stage C aborted less than
those at stage B; differential behaviour was noted on both local
 Relationships between pepper flower abortion and enzymes activity under low temperature 37

7 5
(50%) * (56-69%)*

Number of flower buds

Flower bud development

6 (1)
4
(1)
5 (66-70%)

at stage C
3
4
(82-88%) 2
3 (86-91%)
2 1
1
0

1 2 3 4 5 6 7 1 2 3 4 5 6 7
10 Beldi (2) 5
(24%) (2)
Flower bud development

Baklouti

Number of flower buds

8 (36%) 4
(49%)

at stage B
6 3 (65-75%)
(45%)
4
2
(60%)
2
1

.
(90%) (63-70%)
1
1 4 2 5 3 6 7 0
Days 1 2 3 4 5 6 7
Fig. 2. Development evolution of buds to flower buds on ‘Beldi’ and 5
‘Baklouti’ varieties following plant transfer from 25/20°C to 25/10°C (36-39%)*
(1) and vice versa (2). * Numbers between brackets represent the bud
Number of flower buds

abortion percent after 3, 4 and 5 days of treatment. 4 (3)

varieties. These flower buds at stage C seem to be more tolerant

at stage C

3
to low night temperature than those at stage B (Fig. 3.2).
The reciprocal transfer (from 25/10°C to 25/20°C) appeared to 2
stabilize abortion at the fifth day of treatment as well as for the (63-72%)
two-stage of flower buds with a slight difference between both 1
varieties (Fig. 3.3 and 3.4).
0
Floral structures sensitivity after three and five days of the 1 2 3 4 5 6 7
treatment: The lowest percentage abortion of floral structure was 5 Beldi
(4)
noted under constant optimal night temperature 25/20°C after
Number of flower buds

Baklouti
either 3 or 5 days of treatment, while the highest abortion percentage 4 (53%)
was recorded following the transfer from 25/20°C to 25/10°C (Table
at stage B

1). Low night temperature enhanced abortion of floral structure 3

that varied from 47 to 74% after 3 and 5 days respectively, while (48%)
2
the reciprocal transfer (from low to the optimal night temperature
regime) reduced this abortion percentage to 34%. 1
The floral structure sensitivity was strongly dependent on
varieties (Table 2). ‘Beldi’ showed lower abortion percentage 0
1 2 3 4 5 6 7
than ‘Baklouti’ that seems to be more sensitive to low night Days
temperature; it presented 48% of floral structure abortion after Fig. 3. Development evolution of flower buds, on ‘Beldi’ and ‘Baklouti’
5 days of treatment. Temperature x varieties interaction was not varieties, at stage C and at the pre-anthesis stage (D) following plant
significant. These results describe the response of both varieties transfer from 25/20°C to 25/10°C (1and 2) and vice versa (3 and 4).
*numbers between brackets represent the flower bud abortion percent
grown during winter season under unheated greenhouse. The after 3 and 5 days of treatment.
most pronounced effect of low night temperature was noted on
Table 1. Abortion percentage of floral structure after 3 and 5 days of the buds and flower buds at stage B compared to those at stage C, as
treatment at different temperature regime, constant 25/20°C, transfer well as after 3 or 5 days of the transfer. The highest abortion of
from 25/20°C to 25/10°C and inversely
Treatment 3 days 5 days
floral structure was recorded on buds (stage A) that ranged from
constant 25/20°C 13.9b* 25.9c 43.5 to 50.2% after 3 and 5 days of the treatment, respectively,
Transfer from 25/20°C to 25/10°C 47.2a 74.5a while the lowest percent abortion was noted on flower buds at
Transfer from 25/10°C to 25/20°C 42.2a 33.9b stage C (Table 3). The differential sensitivity of floral structure
LSD 6.9 4.7 to low night temperature could be explained by the enzymatic
* means not followed by the same letter are significantly different at P≤ 0.05 activity variation.
38 Relationships between pepper flower abortion and enzymes activity under low temperature

Table 2. The average of floral structure abortion on ‘Beldi’ and ‘Baklouti’ acid invertase appears to be more important than other enzymes
after 3 and 5 days of the transfer from optimal night temperature regime for all floral structures and it is characterized by an increasing
(25/20°C) to low night temperature regime (25/10°C)
gradient according to the evolution of these structures. The
Variety 3 days 5 days
insoluble acid invertase activity shows a different behaviour; it
‘Beldi’ 31.5b* 41.2b
indicates a decrease between buds stage and flower buds stage
‘Baklouti’ 37.4a 48.3a
and takes an increasing trend until the ovary.
LSD 5.7 5.4
* means not followed by the same letter are significantly different at Inspite of the similar activity for the sucrose synthase, buds
P≤ 0.05 and flower buds express an opposite activity for the soluble and
Table 3. Differential sensitivity of floral structure abortion to low
night temperature after 3 and 5 days of the transfer from optimal insoluble acid invertase. On the other hand, sucrose synthase
night temperature regime (25/20°C) to low night temperature regime and soluble acid invertase activities were more intense in flower
(25/10°C) buds at stage C as compared to the previous structures. Abortion
Stage/organ 3 days 5 days of these floral structures seems to be controlled differentially by
Buds (stage A) 43.5a* 50.2a one or the other type of enzymes.
Flower buds (Stage B) 37.7b 50.1a
Low night temperature reduced sucrose synthase activity in
Flower buds (Stage C) 2.1c 33.8b
‘Baklouti’ by 53% and by 38% in ‘Beldi’ (Table 7). Under low
LSD 3.8 3.1
* means not followed by the same letter are significantly different at night temperature of 10°C, reduction of soluble and insoluble acid
P≤ 0.05 invertase activities was more pronounced than sucrose synthase
Experiment 2 activity. The insoluble fraction of acid invertase was more affected
Effect of low night temperature on sucrose synthase and acid in both ‘Beldi’ and ‘Baklouti’ varieties with 1 to 0.7 μmol gfwt-1
invertase activities: The activity of the sucrose synthase (the min-1, respectively.
sucrose cleaving enzyme) and soluble and insoluble acid invertase Table 6. The enzymatic activity expressed on μmol gfwt-1 min-1 on four
(expressed on a fresh weight basis) was strongly dependent on different pepper flower structures, buds (stage A), flower buds (stage B
and C) and flower ovaries at anthesis
temperature regime (Table 4). On the other hand, 50 to 90%
of the acid invertase activity was found in the soluble fraction Stage/organ ctures Sucrose Soluble acid Insoluble acid
synthase invertase invertase
either at optimal or at low night temperature regime. The lowest Buds (stage A) 2.9c* 10.7c 4.0b
activity was noted on the insoluble part of acid invertase under Flower buds (stage B) 2.0c 15.9b 2.3c
temperature regime of 25/10°C. Flower buds (stage C) 4.6b 19.4a 4.8b
Effect of varieties on enzymatic activity: Enzymatic activity Ovaries 10.5a 20.2a 14.2a
depends on varieties. In fact, enzymatic activity of ‘Beldi’ was LSD 1.0 2.7 1.1
found to be superior to that on ‘Baklouti’ (Table 5). Sucrose * means not followed by the same letter are significantly different at
P≤ 0.05
synthase activity was greatly suppressed in ‘Baklouti’ (3.4 μmol
gfwt-1 min-1) compared to ‘Beldi’ (6.6 μmol gfwt-1 min-1 ). Insoluble Table 7. Sucrose synthase and acid invertase activities expressed on μmol
gfwt-1 min-1 on ‘Beldi’ and ‘Baklouti’ varieties grown under optimal night
and soluble acid invertase followed a similar pattern, but less temperature regime (25/20°C) or low night temperature (25/10°C)
pronounced for the soluble fraction.
Enzyme ‘Beldi’ ‘Baklouti’
Enzymatic activity in different floral structures: Table 6 shows 25/20°C 25/10°C 25/20°C 25/10°C
that enzymatic activity depends on the floral structures. In fact, Sucrose synthase 8.2±2.1* 5.1±0.2 4.7±1.0 2.2±0.3
in flower ovaries (at the anthesis stage), this activity was more Soluble Acid invertase 26.1±7.1 11.8±4.9 17.9±5.2 10.4±3.2
intense than in buds (stage A) and flower buds (stage B) and less Insoluble Acid invertase 15.7±6.0 1.0±0.2 7.9±1.5 0.7±0.08
pronounced in flower buds at stage C. The activity of the soluble * means ± SE (n= 12 replications)
Table 4. The effect of low night temperature on sucrose synthase and acid Relationships between floral structures abortion and
invertase activities expressed on μmol gfwt-1min-1 evaluated on different enzymatic activity: Correlation coefficient between floral
pepper flower structures
structures abortion in ‘Beldi’ and ‘Baklouti’ grown under optimal
Temperature Sucrose Soluble acid Insoluble acid
synthase invertase invertase night temperature regime (25/20°C) or low night temperature
25/20°C 6.5a* 22.0a 11.8a (25/10°C) and enzymatic activity revealed that, under low night
25/10°C 3.6b 11.1b 0.8b temperature, the abortion of ‘Baklouti’ floral structures was
LSD 1.7 2.5 0.9 associated negatively with sucrose synthase and insoluble acid
* means not followed by the same letter are significantly different at invertase (r=-0.78**), while for ‘Beldi’, this coefficient was only
P≤ 0.05 significant for soluble acid invertase (r=-0.82**). However, under
Table 5. The average enzymatic activity expressed on μmol gfwt-1 min-1 optimal temperature regime (25/20°C) floral structures abortion
evaluated on pepper flower structure of two local hot pepper varieties
‘Beldi’ and ‘Baklouti’ of ‘Beldi’ and ‘Baklouti’ was associated to the insoluble and
soluble acid invertase, respectively (Table 8). Moreover, the
Variety ties Sucrose Soluble acid Insoluble acid
synthase invertase invertase abortion of different floral structures seems to be dependant on
‘Beldi’ 6.6a* 18.9a 8.4a the enzyme type. In fact, buds abortion was associated mainly
‘Baklouti’ 3.4b 14.2b 4.3b to the acid invertase activity as well as in ‘Beldi’ and ‘Baklouti’,
LSD 0.9 3.3 2.1 while sucrose synthase activity controlled the flowers buds (stage
* means not followed by the same letter are significantly different at B) abortion. Although abortion of flower buds (stage C) depended
P≤ 0.05 on the varieties; soluble and insoluble acid invertase controlled
 Relationships between pepper flower abortion and enzymes activity under low temperature
Table 8. Pearson correlation coefficients between floral structures abortion in ‘Beldi’ and ‘Baklouti’ grown under optimal night temperature regime
(25/20°C) or low night temperature (25/10°C) and enzymatic activity expressed as μmol g fwt-1 min-1
Tem ‘Beldi’
Sucrose synthase Soluble acid Insoluble acid
invertase
25/20°C -0.13 ns -0.44 ns
25/10°C -0.12 ns -0.82**
* , ** significant differences at P<0.05 and P<0.01 respectively, ns- differences not significant at P>0.05
this abortion in ‘Baklouti’ while only insoluble fraction of acid
invertase was associated with flower buds abortion in ‘Beldi’.
Discussion
The physiological processes like photosynthesis, metabolism
of sugars and translocation of assimilate have been studied
under stress conditions. Important progress has been made in
quantifying and modeling synthesis and distribution of assimilates
in leaves of fruits tree species (Grossman and Dejong, 1994;
Wermelinger et al., 1991) and of the vegetable species (Heuvelink,
1995; Marcelis, 1996; Marcelis et al., 1998). However, the
floral structures: buds and floral buds, more particularly of the
vegetable species cultivated under low night temperature, had
little attention.
The floral structures abortion at various differentiation stages
has been more common under low light intensity and/or high
temperature in several Solanaceous crops, including pepper
(Aloni et al., 1996; 1997) and tomato (Heuvelink, 1996). The
sensitivity of the floral structure to low night temperature was
exhibited when plants were transfered from optimal condition
(25/20°C) to low night temperature (25/10°C) and reciprocally
(experiment 1). A greater abortion rate had been noted after
three days of this treatment. This result suggests an early effect
of temperature on these structures. A similar response has
been recorded by Aloni et al. (1991b) who observed that the
development of buds is affected by the high temperatures (>30°C)
during the first 6 hours post treatment and a complete growth
stopping took place after 24h of heat stress.
The abortion of buds and flower buds at stages B and C has been
found to be dependent on temperature regime and variety, and
within the same variety, on the floral differentiation stages (Fig.
2, 3). The low night temperatures (25/10°C) enhances buds and
flower buds abortion which is more pronounced for buds than
flower buds at stage B and less for those at stage C. The abortion
of floral structures was more pronounced in ‘Baklouti’ than in
‘Beldi’. This differential behaviour of varieties tested has been
found in close association with enzymatic activity that was
strongly reduced by the night temperature of 10°C (Table 4).
Indeed, the activities of both enzymes, sucrose synthase and the
soluble and insoluble acid invertase, were significantly higher
under temperature regime of 25/20°C and were more expressed
in ‘Beldi’ than in ‘Baklouti’. This result suggests that, in addition
of the temperature effect, other factors such as, the genetic aspect
seems to regulate the metabolic activity. In fact, under light stress
conditions (shading of 60%), Shiffriss et al. (1994), studying
the abortion of flowers of inbred lines and hybrids F1 and F2 of
pepper in segregation, had noted variable abortion rates between
these different types of plant material and concluded that some
genetic factors are responsible.
39
‘Baklouti’
Sucrose synthase Soluble acid Insoluble acid
invertase invertase invertase
-0.70* -0.13 ns -0.72* -0.49 ns
-0.17 ns -0.78** -0.35 ns -0.78**
Geiger et al. (1996) showed that distribution of assimilates is
controlled by at least two enzymes: sucrose synthase and acid
invertase and this distribution is controlled by the strength of
sink organs. It seems, however, that this distribution is governed
by the intensity of the organ strength. Buds (stage A) and flower
buds at stage B presented the weakest enzymatic activity in
comparison to ovaries at anthesis stage, while flower buds at stage
C had intermediate position (Table 6). The differential abortion of
these structures would be, particularly, attributed to the activity of
these two enzymes that may serve as an indicator for the organs
strength (Sun et al., 1992).
Heuvelink (1995) and Marcelis et al. (1998) reported that the
temperature is an important factor affecting the distribution of
assimilates in plants, while light and the CO2 level affect the
strength of the source organs. On the other hand, Bertin (1995)
suggested that the abortion of the tomato inflorescences, before
the anthesis may be due to the competition for assimilate between
the young vegetative organs and the last inflorescences. For other
researchers (Sato et al., 2001) the floral buds abortion (before
the anthesis) at the tomato seems to result in a competition of
assimilates between fruits and flowers of the same bunch or the
superior bunch. In this investigation, result suggests that the floral
structure abortion is attributed to a poor translocation capacity of
assimilates. Furthermore, a direct effect of the temperature regime
is suggested. Varying sink strength by changing the number or
position of early-formed fruits affected abortion in sweet pepper
flowers and the abortion rate showed a linear relationship with
the growth rate of the earlier formed competing fruits (Marcelis et
al., 2004). This abortion induction may be caused by competition
for available assimilates by dominance due to the production
of plant growth regulators from the developing fruits, or by a
combination of them (Heuvelink and Korner, 2001; Marcelis and
Baan Hofman-Eijer, 1997)
The competition for the assimilate translocation between the
young leaves and flowers also constitutes a reason of the flower
abortion (Van Doom and Stead, 1997) since these young leaves
constitute a more powerful sinks than the adjacent flowers (Aloni
et al., 1991a, 1991b). However, Turner and Wien (1994) did
not report such effect, since leaves suppression did not improve
flowers retention. This controversy assumes that the floral
structure abortion would rather be assigned to a competition
between the floral structures at different differentiation stages.
Flowers, at anthesis stage, have been considered as a strong sinks
(Black 1993; Marcelis 1996). In the present study (Table 6),
except for insoluble acid invertase, the most intense enzymatic
activity has been found in the ovaries; this could explain their
low abortion under low night temperature. Working under high
temperature, Aloni et al. (1997) found more intense activity of
sucrose synthase at post-anthesis stage.
40 Relationships between pepper flower abortion and enzymes activity under low temperature
The analysis of the association between the floral structures
abortion and the enzymatic activity revealed that, at low night
temperature, abortion control suggests a genetic effect; thus,
the soluble acid invertase seems to control abortion in ‘Beldi’
variety, whereas on ‘Baklouti’ variety the combination of sucrose
synthase and insoluble acid invertase controls this phenomenon
(Table 8).
The abortion of buds seems to be associated with acid invertase
activity especially its soluble fraction and to a lesser extent with
insoluble fraction either in ‘Beldi’ or ‘Baklouti’. Sucrose synthase
seems to be responsible of the flowers buds (stage B) in both
varieties. Compared to the flowers at anthesis stage, studies on
floral structure abortion at the first stages of differentiation (buds
and flower buds) in relation to the metabolic activity are scarce.
The amplitude of variation of flower structures to abortion is
attributed to the simultaneous effects of the endogenous factors
(metabolic activity) and of the exogenous factors (environmental
factors). The later factor optimize the availability and the
distribution of assimilates between the different organs of the
plant. The abortion of buds and flowers buds under low night
temperatures of 10°C, associated to a low enzymatic activity,
support this hypothesis.
References
Aloni, B., T. Pashkar and L. Karni, 1991a. Nitrogen supply influences
carbohydrates partitioning in pepper seedlings and transplant
development. J. Amer. Soc. Hort. Sci., 116: 995-999.
Aloni, B., T. Pashkar and L. Karni, 1991b. Partitionning of [14]-C sucrose
and acid invertase activity in reproductive organs of pepper plants in
relation to their abortion under heat stress. Ann. Bot., 67: 371-377.
Aloni, B., L. Karni, Z. Zaidman and A.A. Schaffer, 1996. Changes of
carbohydrates in pepper (Capsicum annuum L.) flowers in relation
to their abortion under different shading regimes. Ann. Bot., 78:
163-168.
Aloni, B., L. Karni, Z. Zaidman and A.A. Schaffer, 1997. The relationship
between sucrose supply, sucrose cleaving enzymes and flower
abortion in pepper. Ann. Bot., 79: 601-605.
Aufhammer, W., I. Gotz and M. Peter, 1987. Yield performance in field
bean (Vicia faba L.) in relation to interactions between inflorescence
at different nodes. J. Agri. Sci., 105: 662-664.
Bertin, N. 1995. Competition for assimilates and fruit position affect fruit
set in indeterminate greenhouse tomato. Ann. Bot., 75: 55-65.
Black, C.C., 1993. Sink strength: it is real or measurable ? Plant Cell
Envir., 16: 1037-1038.
Deng, X., S.A. Weinbaum, T.M. DeJong and T.T. Muraoka, 1991.
Pistillate flower abortion in ‘Serr’ walnut associated with reduced
carbohydrate and nitrogen concentrations in wood and xylem sap.
J. Amer. Soc. Hort. Sci., 116: 291-296.
Geiger, D.R., K.E. Koch and W.J. Shieh, 1996. Effect of environmental
factors on whole plant assimilate partioning and associated gene
expression. J. Exp. Bot., 47: 1229-1238.
Geiger, D.R. and J.C. Servaites, 1991. Carbon allocation and response
to stress. In: Response of Plants to Multiple Stress. Mooney H.A. et
al. (Eds) Academic press, San Diego. p. 103-125.
Grossman, Y.L. and T.M. Dejong, 1994. Peach: a simulation model of
reproductive and vegetative growth in peach trees. Tree Physiol.,
14: 329-345.
Havaux, M. 1993. Rapid photosynthetic adaptation to heat stress
triggered in potato leaves by moderately elevated temperatures.
Plant Cell Environ., 16: 461-467.
Heuvelink, E. 1995. Effect of temperature on biomass allocation in
tomato (Lycopersicon esculentum). Physiol. Plantarum, 94: 447-
452.
Heuvelink, E.1996. Dry matter portioning in tomato: validation of a
dynamic simulation model. Ann. Bot., 77: 71-80.
Heuvelink, E. and O. Korner, 2001. Parthenocarpic fruit growth reduces
yield fluctuation and blosson-end rot in sweet pepper. Ann. Bot.,
88: 69-74.
Jenner, C.F. and J.S. Hawker, 1993. Sink strength: soluble starch synthase
as a measure of a sink strength in wheat endosperm. Plant Cell
Environ., 16: 1023-1024.
Kitroongrung, N., S. Jodo, J. Hisai and M. Kato, 1992. Photosynthesis
characteristics of melon grown under high temperature. J. Jap. Soc.
Hort.. Sci., 61: 107-114.
Marcelis, L.F.M. 1996. Sink Strength as a determinant of gray matter
partitioning in the whole plant. J. Exp. Bot., 47: 1281-1291.
Marcelis, L.F.M. and L.R. Baan Hofman-Eijer, 1997. Effect of seed
number on competition and dominance among fruits in Capsicum
annuum L. Ann. Bot., 79: 687-693.
Marcelis, L.F.M. , E. Heuvelink and J. Goudriaan, 1998. Modeling
biomass reduction and yield of horticultural crops: a review. Scientia
Hortic., 74: 83-111.
Marcelis, L.F.M., E. Heuvelink, L.R. Baan Hofman-Eijer, J. Den Bakker
and L.B. Xue, 2004. Flower and fruit abortion in sweet pepper in
relation to source and sink strength. J. Exp. Bot., 55: 2261-2268.
Miller, G.L. 1959. Use of dinitrosalycilic acid reagent for determination
of reducing sugars. Ann. Chem., 3: 426-428
Minchin, P.E.H. and M.R. Thorpe, 1996. What determines carbon
partitioning between competing sites? J. Exp. Bot., 47: 1293-
1296.
Roitsch, T. 1999. Source-sink regulation by sugar and stress. Current
Opinion Plant Biol., 2: 198-206.
Statistical Analysis System, 1985. SAS User’s guide Statistics. Ed.
Cary, NC.
Sato, S., M.M. Peet and R.G. Gardner, 2001. Formation of parthenocarpic
fruit, undeveloped flowers and aborted flowers in tomato under
moderately elevated temperatures. Scientia Hortic., 90: 243-254.
Schaffer, A.A., B. Aloni and E. Fogelman, 1987. Sucrose metabolism and
accumulation in developing fruit of sweet and non-sweet genotypes
of Cucumis. Phytochemistry, 26: 1883-1887.
Shiffriss, C., M. Pilowsky and B. Aloni, 1994. Variation in flower
abortion of pepper under stress shading conditions. Euphytica, 78:
133-136.
Sturm, A. and G.Q. Tang, 1999. The sucrose-cleaving enzymes of plants
are crucial for development, growth and carbon partitioning. Trends
Plant Sci., 4: 401–407.
Sun, J., T. Loboda, S.S. Sung and C.C. Jr Black, 1992. Sucrose synthase
in wild tomato Lycopersicon chmielewskii, and tomato fruit strength.
Plant Physiol., 98: 1163-1169.
Tarchoun, N. 2003. Etude de la floraison et de la nouaison en conditions
de basses températures nocturnes et leur déterminisme génétique
chez le piment (Capsicum annuum L.). Thèse de Doctorat d’Etat en
Sciences Agronomiques de l’INAT- Tunisie, 188pp.
Turner, A.D. and H.C. Wien, 1994. Photosynthesis, dark respiration and
bud sugar concentrations in pepper varieties differing in susceptibility
to stress-induced bud abortion. Ann. Bot., 73: 623-628.
Van Doom, W.G. and A.D. Stead, 1997. Abortion of flowers and floral
parts. J. Exp. Bot., 48: 821-837.
Wermelinger, B., J. Baumgartner and A.P. Gutterrez, 1991. A
demographic model of assimilation and allocation of carbon and
nitrogen in grapevines. Ecol. Modeling, 53: 1-26.
Wien, H.C., A.D. Turner and S.F. Yang, 1989. Hormonal basis for low
light intensity induced flower bud abortion of peppers. J. Amer. Soc.
Hort. Sci., 114: 981-985.
 Journal
Journal of Applied Horticulture, 11(1):41-45, January-June, 2009
Appl

Effects of antibrowning agents on the shelf life of fresh-cut

green jackfruit (Artocarpus heterophyllus Lam.)

Boodia Navindra*, Ruggoo Arvind and Boodoo B. Hassina

Faculty of Agriculture, University of Mauritius, Réduit, Mauritius. *E-mail: n.boodia@uom.ac.mu

Abstract
Green mature jackfruits were minimally processed into cubes, dipped in solution of citric acid (0 and 1%) and ascorbic acid (0, 1 and
2%), vacuum packed at 550 mbar atmospheric pressure in 80 μm laminated low density polyethylene vacuum pouches and stored at
2-4°C for 15 days. A control was prepared, using water. Quality parameters like colour, firmness, pH, titratable acidity and total soluble
solids were determined during storage. Colour parameters indicated increase in browning during storage. A significant increase (P<0.05)
in titratable acidity and significant decrease (P<0.05) in pH were observed in all treatments. Texture significantly decreased (P<0.05)
in all treatments during storage. Combinations of the browning inhibitors were more effective than when applied individually. Citric
acid and ascorbic acid when applied together resulted in non-significant change (P>0.05) in microbial counts, browning, and colour
lightness. Treatment of 1% citric acid and 2% ascorbic acid in combination with moderate vacuum packaging and low temperature
storage was found most effective in inhibiting browning and deterioration of fresh-cut green jackfruit for up to 15 days.
Key words: Antibrowning agents, citric acid, ascorbic acid, Artocarpus heterophyllus, minimal processing, green jackfruit, moderate
vacuum packaging.

Introduction Asian community. Therefore, presentation of jackfruit as a fresh-

Fresh-cut products are defined as fruits or vegetables that have
been freshly cut, washed and packaged that offer consumers a
nutritious, convenient and fresh-like product (Gimenez et al.,
2003). There is significant growth in the market for these products
worldwide. However, although minimally processed products
satisfy consumer demand towards fresh, healthy and convenient
products, minimal processing such as peeling, cutting, shredding
and grating renders the product highly perishable (Laurila and
Ahvenainen, 2002). Fresh-cut processing operations result in a
wound response that leads to chemical reactions such as increased
respiration rates, cut surface browning, increased ethylene
production, water loss, and off-flavour development, which in turn
results in texture loss and reduces product quality and shelf-life
(Escalona et al., 2005; Gonzalez-Aguilar et al., 2000; Agar et
al., 1999). Apart from biochemical causes, spoilage of fresh-cut
products is also caused by the growth of micro-organisms, the
growth of which is enhanced in the processing process due to
the availability of cell nutrients and increased surface area for
growth.
Fresh-cut fruits and vegetables are normally packaged in film bags
to reduce the respiration rate and slow the growth of anaerobic
spoilage micro-organisms on the product. Creation of a modified
atmosphere of low oxygen and elevated carbon dioxide levels
inside the package may retard browning, spoilage and maintain
the fresh appearance (Gonzalez-Aguilar et al., 2000; Farver
and Dodds, 1995). Modified atmosphere packaging is used as
a supplement to low temperature storage to reduce the rate of
enzyme activity. Moderate vacuum packaging of shredded lettuce
with polyethylene (80μm) resulted in a shelf life greater than 10
days at 5°C (Heimdall et al., 1995).
Green jackfruit, as a cooked vegetable, is an appreciated dish by the
cut product would be very convenient to reduce preparation time
in the kitchen and would be a product welcomed by consumers.
However, enzymatic browning was found to be the major limiting
factor in fresh-cut jackfruit commercialisation. Enzymatic
browning occurs when phenolic compounds are oxidised by the
copper-containing enzyme, polyphenol oxidase (PPO) (Paul and
Palmer, 1972). During fresh-cut preparation, cells are ruptured
causing the intermixing of PPO with phenolic substrates, which
are normally compartmentalised, resulting in an undesirable brown
colour (Dong et al., 2000; Sapers et al., 2002).
Browning and decay in fresh-cut fruits and vegetables can be
reduced by the use of natural antibrowning agents (Gonzalez-
Aguilar et al., 2000). As an alternative to sulphites, which has
been found to cause allergic reactions in asthmatics (Sapers,
1993), the most studied alternative has been ascorbic acid (Laurila
and Ahvenainen, 2002). Since ascorbic acid eventually oxidises
to dehydro-ascorbic acid, it is best used in combination with
other antibrowning agents like citric acid. Ascorbic acid and
citric acid have been found to be effective enzymatic browning
inhibitors. Langdon (1987) found that immersion of peeled and
sliced potatoes in solutions of ascorbic and citric acid followed
by vacuum packaging, resulted in product shelf life of greater
than 14 days.
The objective of this study was to investigate the effects of
ascorbic acid and citric acid as antibrowning agents on the quality
and microbial load of moderately vacuum packed fresh-cut green
jackfruit stored at 2-4°C.
Materials and methods
Plant materials: Disease free jackfruits of the same maturity were
hand-harvested from registered growers in Mauritius. Harvested
42 Preservation of fresh-cut green jackfruit

jackfruits were immediately transported from the field to the
University of Mauritius.
Sample preparation: Jackfruits were manually cut at the base
and the latex allowed to flow. The fruits were then cut into rings,
peeled, cored and finally cut into cubes of 25 mm. The jackfruit
cubes were dipped in the test solutions, which included different
concentrations of ascorbic acid and citric acid (Table 1). These
cubes were drained and blotted dry with paper towels, placed in
laminated polyethylene vacuum bags of thickness 80 μm (oxygen
transmission rate: 35 cm3/m2/24h at 23°C; Linpac Ltd., Pontivy,
France) and vacuum packaged at 550 mbar atmospheric pressure
using the Multivac vacuum packaging machine (Multivac,
Wolfertschwenden, Germany). Each pack contained about 100-
125 g cut fruit. The packaged fresh-cut fruits were stored at 2-
4°C for 15 days. All utensils, containers and work surfaces were
sanitized with 3% Oxonia® solution.
Table 1. Composition of antibrowning dip solutions used on jackfruit
slices
Treatment Concentration
Control (T1) Water
Citric acid (T2) 1%
Ascorbic acid (T3) 1%
Ascorbic acid + Citric acid (T4) 1%+1%
Ascorbic acid (T5) 2%
Ascorbic acid + Citric acid (T6) 2%+1%
Physico-chemical determinations: Destructive analysis was
carried out in duplicate on days 0, 3, 7, 10, 13 and 15. Two
packages selected at random were subjected to physical and
chemical analyses. Diced jackfruits selected randomly from
each package were blended and then squeezed manually through
cheesecloth. The homogenate was used for titratable acidity (TA)
determination and the juice analyzed for total soluble solids
content (TSS) and pH. The TSS was measured using a standard
hand refractometer, corrected at 20°C, and expressed as °Brix
(Askar and Treptow, 1993). The pH was determined using a
pH meter as per AOAC method (1995). For TA determination,
10g of pulp homogenate was boiled with 50 ml hot water for 5
minutes. The mixture was filtered and the filtrate titrated with 0.1
N NaOH (Askar and Treptow, 1993) in triplicate. The mean titre
value obtained was used to calculate TA, which was expressed
in g of citric acid/ 100 g of sample.
Colour measurement: Browning of the jackfruit cubes was
measured as L*, a*, b* using a Minolta CR-300 colorimeter
(Minolta Company Ltd, Osaka, Japan). Measurements were taken
at six different points on each of the cut surface of five dices of
the green jackfruit. The L* values for colour indicate lightness,
whereby an increase in lightness is indicated by an increase in
the L* values. Positive a* values indicate redness and negative a*
values indicate greenness. Positive b* values indicate yellowness
and negative values blueness.
of 10g of jackfruit was taken from each replicate of the six
treatments. This was added to 270 ml sterile (0.1 %) peptone water
and blended in a stomacher for at least 2 minutes. Serial dilutions
(10-1-10-4) were carried out using 1 ml of the macerated sample
and 9 ml aliquots of peptone water. The pour plate technique
was used to inoculate the medium. Plate count agar and potato
dextrose agar were used for enumeration of total viable count and
yeast and mould count, respectively. The inoculated plates were
incubated at 25°C for 3-5 days. The counts were expressed as log
colony-forming units per g (Log10 CFU/g) of sample.
Statistical design and analysis: The statistical design used was a
factorial Randomized Block Design with two levels of citric acid
(0 and 1%) and three levels of ascorbic acid (0, 1 and 2%). Days
in storage (0, 3, 7, 10, 13 or 15) were used as the blocks. Two
replicates per treatment combination were subjected to destructive
analysis. Data was analyzed using MINITAB Version 13.1. All
Least Significant Differences (LSD) were computed at the 5%
level of significance.
Results and discussion
Titratable acidity and pH: Increase in TA (as g citric acid 100 g-1
fresh tissue) was associated with a decrease in pH during storage
(Fig. 1). High acidity was due to the presence of both the acids.
Ascorbic acid as well as citric acid was found to have a significant
effect on TA (P<0.05). A significant interaction effect (P<0.05)
was also noted between citric and ascorbic acids, whereby TA
was found to be higher in fresh-cut jackfruit treated with both
acids (Table 2). This could be due to respiration.
Utilization of organic acids during respiration was found to cause
a decrease in TA in fresh-cut apples (Fan et al., 2005). Increase in
acidity during storage could be due to ripening of the fruit, where
the degradation of pectin in the cell wall results in the formation
of galacturonic acids (Eskin, 1990). Another factor contributing
to increased acidity could be due to the fixation of carbon dioxide
formed during respiration into organic acids (Wang, 1990).
Table 2. Two-way table of treatment means for TA
Citric acid (%) Ascorbic acid concentration (%)
0 1 2
0 0.120 0.097 0.135
1 0.124 0.156 0.160
SED (interaction) = 0.007
Both citric acid and ascorbic acid led to a significant decrease
(P<0.05) in the pH during storage (Table 3). There was no
evidence (P>0.05) of any interaction between the two treatments
indicating that they were acting independently on the pH during
storage.
4.9 0.18
4.8 0.16
4.7 0.14
4.6 0.12
Mean pH

Mean TA

4.5 0.1
Determination of firmness: Firmness of five jackfruit dices 4.4 0.08
chosen randomly from each package was determined using a hand- 4.3 0.06
mean pH mean TA
held penetrometer at six different places on each dice (HP-Fff; 4.2 0.04
4.1 0.02
Tracer 0.25 cm2; Bareiss Prufgeratebau GmbH, Oberdischingen, 4 0
Germany) as per Askar and Treptow (1993). 0 3 7 10 13 15
Days
Microbiological analysis: Microbiological analysis was carried Fig 1. Changes in mean TA and pH during storage of the diced jackfruit
out in triplicate for each treatment on days 0, 6 and 12. A sample LSD (pH) = 0.2303; LSD (TA) = 0.0138
 Preservation of fresh-cut green jackfruit 43

Table 3. Overall pH of diced jackfruits for citric acid and ascorbic acid Table 5. Two-way table of treatment means for a* values
treatments Citric acid Ascorbic acid (%) Mean
Citric acid (CA) pH Ascorbic acid pH (%) 0 1 2
(AA) 0 (+) 5.49 (+) 0.70 (-) 0.22 (+) 1.99
0% 4.70 0% 4.65 1 (+) 0.97 (-) 1.33 (-) 2.38 (-) 0.91
1% 4.32 1% 4.55
Mean (+) 3.23 (-) 0.32 (-) 1.30
2% 4.48
SED (interacion) = 0.581; SED (AA) = 0.411; SED (CA) = 0.335. F-
SED (CA) = 0.065, SED (AA) = 0.079 ratios: CA=74.91; AA=67.28; CA*AA=5.85
Total soluble solids: On day 0, a high °Brix was observed, which However, the F-ratio for the interaction effect was much smaller
decreased on day 3. This could be due to increased respiration compared to the main effects’ sums of squares. This indicates that
as a result of wounding during processing. Sugars and organic changes in a* were mostly driven by the main effects. The increase
acids are the main substrates for respiration in plants (Tovar in a* values during storage indicates an increased occurrence of
et al., 2000). Citric acid was found to have a significant effect browning (Fig. 2).
(P<0.05) on the TSS (Table 4). No significant interaction effect
(P>0.05) between ascorbic and citric acid was noted. Changes L* values were found to decrease during storage (Fig. 3). The
in the soluble solids may be attributed to changes occurring control was found to have the lowest L* value. This is due to
during ripening (Nakasone and Paul, 1999). However in the the fact that no antibrowning agent was used in the control,
present study, it was found that the TSS remained almost constant thereby causing greater extent of browning (Fig. 3). There was no
throughout storage. significant interaction between the effects of citric and ascorbic
acid (P>0.05). Both acids were found to have significant effect
Table 4. Overall TSS of diced jackfruits for citric acid and ascorbic
acid treatments (P<0.05) on the lightness of the fresh-cut green jackfruit (Table
6). Application of citric acid and ascorbic acid individually were
Citric acid (CA) °Brix Ascorbic acid °Brix
found to inhibit browning moderately. The combination of 1% CA
(AA)
and 2% AA proved to be the most effective treatment in reducing
0% 4.09 0% 4.22
browning for 15 days. Combination of 1% CA and 1% AA was
1% 4.33 1% 4.14
also effective in suppressing browning but to a lesser extent as
2% 4.27
slight browning was noted on day 15.
SED (CA) = 0.064, SED (AA) = 0.078
Table 6. Mean L* values of diced jackfruit for citric acid and ascorbic
Colour: Changes in a* and L* values have been used to monitor acid treatments
enzymatic browning of cut apples (Soliva-Fortuny et al., 2001).
Citric acid (CA) L* value Ascorbic acid L* value
Kim et al. (1995) correlated a* and L* values with PPO activity
(AA)
but there was no correlation with b* values. A significant
0% 75.2 0% 71.3
interaction effect (P<0.05) was noted for a* value suggesting
1% 79.9 1% 79.6
better browning control when using both acids (Table 5).
2% 81.8
SED (CA) = 0.84, SED (AA) = 1.03
10
8 T1 T2 T3 T4 T5 T6 Firmness: Firmness is considered as an important quality criterion
6 for fresh-cut products. Both citric acid and ascorbic acid had
significant effects (P<0.05) on the firmness of the fresh-cut
a* values

4
2 jackfruit where both acids were found to retain the firmness (Table
0 7). Retention of firmness could be due to cross-linking between
-2 0 3 6 9 12 15 endogenous calcium ions and the freely available carboxyl
-4 groups in the cell wall pectin, following acidification (Sapers and
-6 Days Miller, 1995). Percentage loss in firmness for all the treatments
is depicted in Fig. 4.
Fig 2. Changes in a* values during storage of the diced jackfruit. LSD
(any 2 means) = 2.072
90 60
% lo s s in firm n es s

85 50
80
40
L* values

75
70 30
65 20
60 10
55 T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6
0
50
3 5 7 9 11 13 15
0 3 7 10 13 15
Days Days

Fig 3. Changes in L* values during storage of the diced jackfruit. LSD Fig 4. Percent loss in firmness during storage of the diced jackfruit. LSD
(any 2 means) = 5.20 (any 2 means) = 9.708
44 Preservation of fresh-cut green jackfruit

Onset of ripening could be a factor responsible for the firmness and citric acid was noted (Fig. 9). This showed that combinations
loss. Fruit pulp loses firmness during ripening, which occurs as a of both acids resulted in an increase in Y:M count. Acidic
result of the decomposition of the primary cell wall constituents, environments were found to be favourable for yeast and mould
following the solubilization of pectin by several hydrolytic growth (Beaulieu and Gorny, 2002; Jay, 1992). Although the Y:
enzymes. (Eskin, 1990; Riquelme et al., 1999). In addition to M count increased during storage, the counts were still low on
ripening, loss of firmness could also be due to decreased turgor day 15. According to Debevere (1996), the recommended Y:
due to water loss (Beaulieu and Gorny, 2002). M count for fresh-cut vegetables should not exceed Log10 104
CFU/g (4.0).
Table 7. Mean firmness of diced jackfruits for main effect treatments
Table 9. Two-way table of treatment means for Y:M count in Log10
Citric acid (CA) Firmness Ascorbic acid Firmness CFU/g
(AA)
Citric acid (%) Ascorbic acid concentration (%)
0% 26.2 0% 24.0
0 1 2
1% 27.1 1% 27.7
0 2.81 2.00 2.14
2% 28.3 1 2.05 2.45 2.69
SED (CA) = 0.70; SED (AA) = 0.86
SED (interaction) = 0.162.
Microbial growth: An increase was observed in both total viable
count (TVC) and yeast and mould (Y:M) count over the storage The low TVC and Y:M counts can also be explained by the possible
period (Fig. 5). development of a high CO2 atmosphere inside the packages. High
6
CO2 atmospheres inhibit most aerobic microorganisms, especially
gram-negative bacteria that cause off-flavours and off-odours
5
(Martin-Belloso, 2006; Farver and Dodds, 1995). Elevated CO2
log CFU/g

4 environments are also fungistatic (Gorny, 1997). However lactic

3 acid bacteria are unaffected by high CO2 and they continue to
2 grow (Farver and Dodds, 1995). Growth of lactic acid bacteria
may result in a decrease in pH and may produce antimicrobial
1 Mean TVC MeanY: M count
compounds, which may suppress the growth of pathogens (Francis
0 et al., 1999).
0 3 69 12 15
Days Slow growth of bacteria, yeasts and moulds was also due to the
Fig 5. TVC and Y:M count during storage of the diced jackfruit. LSD low storage temperature. Results obtained on shredded chicory
(TVC) = 0.711; LSD (Y:M count) = 0.3619
salads, and shredded lettuce (Nguyen-The and Carlin, 1994)
Both citric and ascorbic acid significantly (P<0.05) decreased the showed that growth of mesophilic microflora was significantly
microbial load of the fresh-cut jackfruit (Table 8). No significant reduced at low storage temperatures. The Institute of Food Science
interaction effect (P>0.05) was noted between the two treatments. and Technology recommends a storage temperature in the range
Dipping fresh-cut produce in ascorbic acid/citric acid prior to of 0-5°C (Francis et al., 1999) for prepared salads to limit the
packaging significantly decreased aerobic and anaerobic counts growth of pathogens. In this study, the storage temperature was
(McLachan and Stark, 1985; O’Beirne and Ballantyne, 1987). maintained in the range of 2-4°C, thus explaining the lower plate
These had an impact on the product’s shelf life. counts observed. Low temperature storage is crucial for maintaining
Table 8. Mean TVC (in Log10 CFU/g) of the jackfruit dices for main the microbiological quality of fresh-cut green jackfruit.
effect treatments
Two percent ascorbic acid was found to prolong the shelf life
Citric acid (CA) TVC Ascorbic acid TVC of fresh-cut green jackfruit, but its effect was limited to only
(AA) 7 days at 2-4°C. The combination of 1% citric acid and 1%
0% 4.04 0% 4.29 ascorbic acid was successful in inhibiting browning up to 13
1% 3.37 1% 3.64 days at 2-4°C, while the combination of 1% citric acid and 2%
2% 3.18 ascorbic acid was the most suitable treatment in maintaining the
SED (CA) = 0.18, SED (AA) = 0.23 quality attributes of the fresh-cut jackfruit for up to 15 days at
1 % CA was found to be more effective in reducing plate count 2-4°C. This treatment also prevented microbial growth although
as compared to 1% AA. However 2% AA reduced the plate count the acidic conditions did favour the growth of yeast and mould,
in almost the same way as 1% CA. The antimicrobial effects of which were still less than the permissible amount. The use of
the acids can be explained by the fact that they lower the pH the combined preservative factors has greater effectiveness in
resulting in unfavourable conditions for the growth of bacteria preserving quality attributes and delaying microbial growth than
(Piagentini et al., 2003). Although the TVC continued to increase when used singly.
during storage, the total count was quite low and below the
recommended level of Log10 107 CFU/g (7.0) as per Mauritian References
Food Regulations (1999). Agar, I.T., R. Massantini, B. Hess–Pierce and A.A. Kader, 1999.
Postharvest CO2 and ethylene production and quality maintenance
No significant effects (P>0.05) on Y:M count were observed of fresh-cut kiwifruit slices. J. Fd. Sci., 64(3): 433-440.
when using citric acid and ascorbic acid individually. However Askar, A. and H. Treptow, 1993. Quality Assurance in Tropical Fruit
a significant interaction effect (P<0.05) between ascorbic acid Processing. Springer Laboratory, Germany.
 Preservation of fresh-cut green jackfruit
Association of the Official Analytical Chemists, 1995. Official methods of
analysis of AOAC International, Volume II. (P. Cunniff, ed). AOAC
International, USA.
Beaulieu, J.C. and J.R. Gorny, 2002, Fresh-cut fruits, <http: //www.
ba.ars.usda.gov/hb66/146freshcutfruits.pdf>
Debevere, J. 1996. Criteria en practische methoden voor de bepaling
van de houd baarhei dsdatum in the etikettering. In: Etikettering
houdbaarheid en bewaring, (G. Temmerman, G. Cremer, M. Thyssen,
J. Devebere, eds). Brugge, Belgium. Die Keure, p. 37-64.
Dong, X., R.E. Wrolstad and D. Sugar, 2000. Extending shelf life of
fresh-cut pears. J. Fd. Sci., 65(1): 181-186.
Escalona, V.H., A. Encarna and A. Francisco, 2005. Overall quality
throughout shelf life of minimally fresh processed fennel. J. Fd.
Sci., 70(1): 13-17.
Eskin, M.N.A. 1990. Biochemistry of Foods. Academic Press, Inc.
USA.
Fan, X., B.A. Niemera, J.P. Mattheis, H. Zhuang and D.W. Olson, 2005.
Quality of fresh-cut apple slices as affected by low-dose ionizing
radiation and calcium ascorbate treatment. J. Fd. Sci., 70(2): 143-
148.
Farver, J.M. and K.L. Dodds, 1995. Principles of modified atmosphere
and sousvide product packaging. Technomic Publishing Company,
Inc. USA.
Francis, G.A., C. Thomas and D. O’Beirne, 1999. The microbiological
safety of minimally processed vegetables. Intl. J. Fd. Sci. Technol.,
34: 1-22.
Gimenez, M., C. Olarte, S. Sanz, Lomas, J.F. Echavarri and F. Ayala,
2003. Influence of packaging films on the sensory and microbiological
evolution of minimally processed borage (Borrago officinalis). J. Fd.
Sci., 68(3): 1051-1057.
Gonzalez-Aguilar, G.A., C.Y. Wang and J.G. Buta, 2000. Maintaining
quality of fresh-cut mangoes using antibrowning agents and modified
atmosphere packaging. J. Agr. Fd. Chem., 48(9): 4204-4208.
Gorny, J.R. 1997. Modified atmosphere packaging and the fresh-cut
revolution. Perishables Handling Nwsl. 90.
Heimdall, H., B.F. Kuhn, L. Poll and L.M. Larsen, 1995. Biochemical
changes and sensory quality of shredded and MA packaged iceberg
lettuce. J. Fd. Sci., 60(6): 1265-1268.
Jay, M.J. 1992. Spoilage of fruits and vegetables. In: Modern Food
Microbiology. Chapman and Hall, London. p.187.
Kim, D.M., K.H. Kim, N.L. Smith and C.Y. Lee, 1995. Changes in
flesh colour and PPO activity by apple cultivars. Fd. Biotechnol.,
4(4): 222-225.
Langdon, T.T. 1987. Preventing browning in fresh prepared potatoes
without the use of sulphiting agents. Fd. Technol., 41(5): 64.
Laurila, E. and R. Ahvenainen, 2002. Minimal processing in practice. In:
Minimal Processing Technologies in the Food Industry, (T. Ohlsson
and N. Bengtsson, eds). Woodhead Publishing Ltd. p.219-233.
45
Martin-Belloso, O. and R. Soliva-Fortuny, 2006. Effect of modified
atmosphere packaging on the quality of fresh-cut fruits. Stewart
Postharvest Rev., 2(1): 1-8.
Mauritian Food Regulations, 1999. Microbiological Standard for
Specified Food. Ready to eat food other than those specified above.
Government Gazette of Mauritius No.114. Government Notice
No.173.
McLachan, A. and R. Stark, 1985. Modified atmosphere packaging of
selected prepared vegetables. Campden Food Preservation Research
Association Techical Memorandum. Chipping Campden, UK, 1985,
78 pp.
Nakasone, H.Y. and R.E. Paul, 1999. Tropical fruits. In: Crop Production
Science in Horticulture. CABI International, Hawaii. p.334-338.
Nguyen-The, C. and E. Carlin, 1994. The microbiology of minimally
processed fresh fruits and vegetables. Critical Rev. Fd. Sci., 34:
371-401.
O’Beirne, D. and A. Ballantyne, 1987. Some effects of modified
atmosphere packaging and vacuum packaging in combination with
antioxidants on quality and storage life of chilled potato strips. Intl.
J. Fd. Sci. Tech., 22: 515-523.
Paul, P.C. and H.H. Palmer, 1972. Fruits and vegetables. In: Food Theory
and Applications. John Wiley & Sons, Inc. p. 294-297.
Piagentini, A.M., D.R. Guemes and M.E. Pirovani, 2003. Mesophilic
aerobic population of fresh-cut spinach as affected by chemical
treatment and type of packaging film. J. Fd. Sci., 68(2): 602-607.
Riquelme, F., M.T. Pretel, G. Martinez, M. Serrano, A. Amoros and F.
Romajaro, 1999. Packaging of fruits and vegetables: recent results.
In: Food Packaging and Preservation, M. MathLouthi (eds.). Aspen
Publishers Inc. p.144-153.
Sapers, G.M. 1993. Browning of foods: control by sulphites, antioxidants
and other means. Fd. Technol., 47: 75-84.
Sapers, G.M., K.B. Hicks and R.L. Miller, 2002. Antibrowning agents.
In: Food Additives, A.L. Brane, P.M. Davidson, S. Salminen, J.H.
Thorngate (eds.). Marcel Dekker. p.543-561.
Sapers, G.M. and R.L. Miller, 1995. Heated ascorbic/citric acid as
browning inhibitor for pre-peeled potatoes. J. Fd. Sci., 60(4): 762-
776.
Soliva-Fortuny, R.C., N. Grigelmo-Miguel, I. Odriozola-Serrano, S.
Gorinstein and O. Matin-Belloso, 2001. Browning evaluation of
ready-to-eat apples as affected by modified atmosphere packaging.
J. Agri. Fd. Chem., 49(8): 3685-3690.
Tovar, B., L.I. Ibarra, H.S. Garcia and M. Mata, 2000. Some compositional
changes in Kent mango (Mangifera indica) slices during storage. J.
Appl. Hortic., 2(1): 10-14.
Wang, C.Y. 1990. Physiological and biochemical effects of controlled
atmosphere on fruits and vegetables. In: Food Preservation by
Modified Atmospheres, M. Calderon and R. Barkai-Golan (eds.).
CRC Press, Inc. p.199.
 Journal
Journal of Applied Horticulture, 11(1): 46-49, January-June, 2009
Appl

Use of plastic shades to regulate growth of korarima

(Aframomum corrorima (Braun) P.C.M. Jansen)

S. Eyob
Horticulture Department, Awassa College of Agriculture, Hawassa University, P.O.Box 5, Awassa, Ethiopia.
E-mail: solomoneyob70@gmail.com

Abstract
Korarima (Aframomum corrorima (Braun) P.C.M. Jansen), a slow growing and persistent under tree shade as an under-story perennial
plant, is native to Ethiopia. When it is grown in full sunny condition, all plants die off a few weeks after planting, but the effect of
different shading materials on its growth is not known. Half a year old korarima plants were planted under differently coloured plastic
shades (red, green, blue and clear) and coffee (Coffea arabica L.) tree shade to regulate the growth. The coffee tree shade was used
as control. Varying levels of photosynthetic photon flux density (PPFD) and red to far red (R/FR) ratio of light were recorded under
different shaded and open conditions. The korarima plant responded differently to the different plastic and coffee tree shades. Average
plant height, number of leaves per plant, number of sprouts per plant, chlorophyll content, leaf area, total fresh and dry weights were
significantly different when recorded at different stages of growth, highest being recorded under the blue plastic cover. The minimum
efficiency was achieved under control.
Key words: Korarima, Aframomum corrorima, Mesketo, photon flux density, plastic shade, tree shade, growth regulation.

Introduction (Coffea arabica L.). Review of indigenous production practices in

Korarima (Aframomum corrorima (Braun) P.C.M. Jansen) belongs
to the family Zingiberaceae and native to Ethiopia (Jansen, 1981;
Lock, 1997). It has been used as a spice crop in Ethiopia since
time of Queen Sheba. The seed, pod, leaf, rhizome and root of
korarima are also used as traditional medicine in Ethiopia (Eyob
et al., 2008). It is a perennial aromatic shade loving herb; with
scaly rhizomes and leafy stems reaching 1-2m height. It grows
slowly and persistent under tree canopy as an under-story plant.
It is mainly grown in the natural forests dominated by woody
species.
Geoffrey and Martin (2000) described that many changes in the
environment can occur when the gap is created in the forest.
The changes may include alteration in photosynthetically
active radiation (PAR), light intensity and quality, temperature,
competition for moisture, humidity and nutrient availability.
Perhaps the most critical management requirement is needed
for a well-adapted shade loving plant species such as korarima,
which grow under forest vegetation. Adaptation to shade does
not significantly change leaf radiation adsorption efficiency in
tree species (Baltzer and Thomas, 2005). However, since plants
convert only a small percentage of absorbed radiation into
photosynthate facilitating growth, absorption efficiency may
not be a relevant measure of adaptation benefits. Plants have
evolved very sophisticated sensory networks for monitoring the
status of several important features of their surroundings. Shahak
et al. (2004) reported that the spectral manipulation promotes
physiological responses, while light scattering improves light
penetration into the plant canopy.
Korarima is traditionally grown in home gardens under tree
shades, and in the natural forests in association with wild coffee
Ethiopia showed a decreasing trend in production areas and yield
in two decades as a result of shortage of permanent shade trees in
the household farms and destruction of natural forests. As reported
previously by Kendrick and Kronenberg (1994) the economic
success of crop production requires proper management of the
solar radiation resource as light quality has a significant effect
on plant growth, development and productivity. With respect to
korarima, it is well known fact that its cultivation is unsuccessful
where the shade trees are not available. Hence, the need for
permanent tree shades hinders large scale production of crop. Uses
of covers that block parts of the incoming radiation have been
used in areas with high solar radiation and excess temperatures.
Therefore, the objectives of this study were to determine the
comparative advantages and optimal shade management strategies
using locally produced plastic films over tree shades and open
conditions for growth of korarima.
Materials and methods
Plant material and experimental design: Approximately half a
year old planting materials from cultivar Mesketo were planted
(one plant per pot) in plastic pots (Ø 40cm) filled with forest
soil and compost mixture, and grown in the protected garden
of Horticulture Department, Hawassa University, Ethiopia.
The plants were managed properly as per agronomic practices.
Four plastic pots were replicated three times for each treatment
consisting of 12 pots per treatment in completely randomized
design. Locally produced photo selective red, green, blue, and
clear plastic films partially transparent for sun light were used
as experimental treatments. Coffee tree shades were used as
control treatment. The plastic films used for this purpose are
commonly used as shades in the local open markets and road
sides in Ethiopia. The plastic films were mounted on sticks ca
 Use of plastic shades to regulate growth of korarima 47

2m above the ground. On all sides, 1 m above the ground, the Table 2. Effect of plastic shading on plant height (cm) of korarima at
plastic shades were open to maintain free air circulation at the different stages of growth
same level as the shades from the coffee trees. The coffee trees Treatment Days after planting (DAP)
grew ca 35m from plastic shades and were used to simulate natural 60 90 120 150 180 210
growing condition of koraima. Twelve pots were placed under Red plastic 29.3 33.7 48.0 72.7 75.3 77.0
open conditions for purpose of comparison. Green plastic 32.3 49.7 61.0 86.00 88.3 81.7
Blue plastic 34.7 71.3 83.3 104.8 106.0 109.0
The photosynthetic photon flux density (μmolm-2s-1) and red to far
Clear plastic 22.7 31.3 43.0 61.7 65.7 74.7
red ratio of light were measured under different plastic shades,
Control 23.3 27.0 37.0 56.4 58.3 68.3
coffee trees and open condition by using a SKP 215 PAR Quantum
Sensor before the plants were planted (Table 1). LSD 6.3 19.9 21.1 16.2 15.3 15.4
Table 3. Effect of plastic shading on leaf number per plant
Measurements: Sixty days after planting (DAP), number of fully
Treatment Days after planting (DAP)
opened leaves per plant and number of newly emerging young
shoots at the base of mother plants and chlorophyll content were 60 90 120 150 180 210
recorded on a monthly basis from June 2005 to May 2006. A field Red plastic 8.0 8.7 9.7 11.0 11.3 12.3
portable, hand held chlorophyll content meter (Model CL-01, Green plastic 9.3 11.3 14.00 16.7 17.3 18.3
Hansatech Instruments Ltd, England) was used at day time from Blue plastic 11.3 13.7 15.7 17.3 19.0 20.0
2: 00 to 3:00 p.m. to estimate average total chlorophyll content Clear plastic 5.0 8.7 10.7 11.3 11.7 12.3
in leaves of korarima. Leaf area was measured using portable Control 3.0 4.0 6.7 8.7 10.3 11.3
leaf area meter, Model LI-3000A after removing all leaves from LSD 5 5.8 4 3.5 3.4 3.4
plants grown for destructive measurements and average leaf Table 4. Effect of different shade treatments on the leaf area, fresh weight
area was estimated for a single plant at 300 DAP. In the end of and dry weight per plant
the experiment fresh and dry weights were recorded. Total fresh Treatment Leaf area (m2) Fresh wt (g) Dry wt
weight of sampled plants was measured at 360 DAP and the dry 300 DAP 360 DAP (g)
weight was recorded after oven drying at 100°C for 2 days. Soil Red plastic 3.31 2982.5 231.4
particles and any debris were removed from roots before taking Green plastic 3.80 3075.1 262.0
measurements. Blue plastic 4.22 3766.4 394.9
Data analysis: The data were analysed as a completely Clear plastic 2.13 2029.2 193.8
randomised design with repeated measurements three times. Control 1.62 1633.9 112.7
Fisher least significant difference (LSD) test was used to compare LSD 0.23 351.1 51.2
means at P=0.05. On randomly selected plants regression analysis
plant height between treatments was significant (Table 2). For
was carried out on the number of sprouts per plant and leaf number
all the treatments, there was a gradual increase in plant height
per plant as independent and dependant variables, respectively
from 60 DAP to 210 DAP. However, shoots under the control
to see any relationship and trends in growth of korarima at 240
DAP. The software used for this statistical analysis was MINITAB treatment were weak and slow in the growth. Remarkably strong
13th edition. and erect plants were observed under blue plastic sheet compared
to other treatments. Significantly higher plant height (109.0cm)
Results was recorded at 210 DAP in blue plastic sheet compared to the
control (68.3cm) when recorded on the same date.
The varying coloured plastic sheets significantly affected PPFD
and R/FR level under plant canopy inside the shades. As a In case of leaf number also, the variation between treatments was
consequence of extreme increase of PPFD (1410 μmolm-2s-1) significant (Table 3). There was a gradual increase in leaf number
and R/FR (1.12) under open condition (Table 1), plants showed in different stages of growth. The increase under the control was
very slow rate of increase in shoot height for four weeks (data low. Significantly increased leaf number was recorded in blue
not shown) and died off afterwards while those under shade plastic (20.0) as compared to the control (11.3), clear plastic (12.3)
showed a continuous growth. Under shaded conditions, the and red plastic (12.3) at 210 DAP. Changes in leaf number for the
plants showed significant differences between the treatments clear plastic and the control were low during growth period.
with regard to growth performances at P=0.05.. The variation in The varying coloured plastic sheets induced a significant effect
Table 1. Photosynthetic photon flux density (PPFD) and red to far red on the number of sprouts per plant in almost all the treatments
(R/FR) ratio of light under different shading and open conditions used (Fig. 1). The number of sprouts were most enhanced with
Shading materials PPFD R/FR about 50% of increase in blue sheet when compared with the
(μmolm-2s-1)
control and clear sheet starting from 150 DAP to 210 DAP. The
Red plastic 680 0.98
other treatments showed about 20-30% increase over the control
Green plastic 462 0.68 and clear plastic.
Blue plastic 243 0.66
The variation in leaf area at 300 DAP due to shading effect was
Clear plastic 790 1.01
significant. Plants grown under blue plastic shade had the highest
Coffee tree shade (control) 102 0.53 leaf areas (4.22m2) compared to the control (1.62m2) and other
Open condition 1410 1.12 treatments (Table 4).
48 Use of plastic shades to regulate growth of korarima

16 Red plastic The shade treatments increased fresh and dry weight. The highest
Average number of sprouts/plant

fresh (3766.4g) and dry weight (394.9g) were recorded under blue
14 Green plastic
plastic shade (Table 4). Increase in leaf area accounted wholly for
12 Blue plastic the increase in plant fresh and dry weight.
10 Clear plastic The chlorophyll content in leaf was affected significantly by
8 Control shade treatments, the highest being recorded in blue plastic shade
when recorded at 90 DAP and the lowest was in the control (Fig.
6 2). In all treatments the chlorophyll content gradually decreased
4 after 90 DAP. However, under blue plastic shade they seemed to
increase again after 180 DAP.
2
The varying level of PPFD and R/FR induced by different shading
0 materials showed considerable difference on korarima growth.
90 120 150 180 210 60 The PPFD (102.2 μmol m-2s-1) and R/FR (0.53) under the control
Days after planting (DAP) treatments resulted in poor growth performances of korarima. On
Fig. 1. Effects of artificial shading materials on average number of sprouts average 82.8% decrease and 58% increase in PPFD by blue plastic
per plant at different stages of growth.
shade over open condition and control, respectively, regulated
18
growth of korarima. Most of the growth characteristics were
highest at interaction level of PPFD at 243 μmolm-2s-1 and R/FR
16 at 0.66 under blue plastic.
A v erage total chlorophyll content units

14
The simple linear regression analysis showed a significant
relationship between sprouts per plant as independent variable
and leaf number as a response variable when tested at P=0.05.
12
For a unit increase in sprout number per plant, there was linear
increase in leaf number by 1.964 as indicated in Fig. 3. It is clear
10 that 80.4% variability (R2=0.804) in leaf number per plant could
be explained by sprout number per plant.
8
Discussion
6
The potential importance and results obtained from growth
performance of korarima are presented in this study. Shading
4 materials significantly influenced the rate of vegetative growth of
Red plastic
korarima. In the blue plastic shade, plant vegetative growth was
2 Green plastic Clear plastic significantly promoted. The growth performance of the plant was
Control reduced in all stages of growth under tree shades indicating that
Blue plastic
0 tree canopy intercepted most part of incoming solar radiation.
60 90 120 150 180 210 This explains the existing fact that korarima is indigenous slowly
growing species in the forests of Ethiopia as they appear more or
Days after planting (DAP)
less isolated or in clumps under trees. Total death of plants under
Fig. 2. Effect of artificial shading materials on average chlorophyll
content of korarima at different stages of growth. open condition might be due to heat from a high portion of near
infra-red radiation (NIR, 700 to 1100nm).
Number of leaves per plant=
0.71+1.964 Number of sprouts per plant It was observed that the plant height under plastic shade showed
25 greater increase in length compared to those of under control. The
remarkable increase in plant height, leaf number and number of
Number of leaves per plant

20 sprout under blue plastic shade during growth period over other
treatments indicate differences in response of the rate of growth
15 under the various light regimes. In our experiment, etiolated and
very weak plants with dull green leaves were observed under
the control as compared to plastic shades explaining the need of
10
enough light sources during photosynthesis for the manufacture
R2=80.4 of food. In the shortage of light, sufficient energy is not available
5 for plants as observed in control.
Leaf area increased faster under blue plastic due to a higher
0
2 4 6 8 10 12 leaf appearance rate, which made possible the growth of young
Number of sprouts per plant sprouts at the base of mother plant, as photosynthetically active
Fig. 3. Linear relationship between sprouts number per plant and leaves leaves were available. After 90 DAP chlorophyll content in leaf
number per plant at 240 DAP. was declined gradually under all shade treatments implying the
 Use of plastic shades to regulate growth of korarima
assimilation of photosynthetic products for growth by young
growing sprouts. The increased total fresh and dry weights per
plant at maturity were associated with enhanced effect of leaf
area production under blue plastic.
Compared to other plastic colours, blue plastic decreased
PPFD and R/FR to optimum level and resulted in better growth
performance of korarima as a result of the partial control of
climate inside the plastics. The significance of the climatic
control systems inside the greenhouses have been suggested
by Arellano et al. (2006). This experimental finding shows that
artificial coloured covering materials are effective in regulating
vegetative growth. This is in conformity with finding of Donald
et al. (2005), Murakami et al. (1997), and Rajapaske and Kelly
(1992). Generally with moderate reduction of PAR, there were
increase in korarima growth performance. The overall effect of
heavy shading resulted in tiny and a poorly grown plant under
the control, which is in agreement with work of Bartlett and
Remphrey (1998).
The results reported in this work, in brief, suggest that in the
absence of enough forest gaps for sufficient light quality to
infiltrate and initiate better growth of korarima, physiological
and morphological response would be affected negatively. The
difference in growth response achieved due to shade treatments
suggests various possibilities of field management strategies
using locally available plastic sheets which could vary depending
on light source to optimise production. Moreover, korarima
response to various level of PAR suggesting a high degree of
flexibility in terms of growth and resource utilization even though
the korarima plant exist in the forest under natural conditions.
Further we suggest development of new korarima varieties
through conventional breeding and molecular techniques for
maximum and efficient utilization of total light sources under
open conditions to promote production at large scale in the future,
which also avoids cost of shading materials.
Acknowledgments
Author gratefully acknowledges research support received from
Norwegian Agency for Development Cooperation (NORAD)
research project in collaboration with Hawassa University,
Ethiopia. The author also thanks Associate Professor Maigull
Appelgren, Norwegian University of Life Sciences, Department
49
of Plant and Environmental Sciences and Associate Professor
Admasu Tsegaye, Hawassa University, Awassa College of
Agriculture, Horticulture Department for their valuable comments
and suggestions during preparation of this manuscript.
References
Arellano, M.A., S. Garcia, A. Sanchez, J. Soria-Ruiz, D.L.Valera and
M. Urrestarazu, 2006. Greenhouse microclimate and its natural
variation in two subtypes of an almería greenhouse. Acta Hort.,
719: 147-156.
Baltzer, J.L. and S.C. Thomas, 2005. Leaf optical response to light and
soil nutrient availability in temperate deciduous trees. Am. J. Bot.,
92: 214-223.
Bartlett, G.A. and W.R. Remphrey, 1998. The effect of reduced quantities
of photosynethetically active radiation on Fraxinus pennsylvanica
growth and architecture. Can. J. Bot., 76: 1359-1365.
Donald, T.K., C.H. David and M.M. Roman, 2005. Spectral properties
of selected UV-blocking and UV-transmitting covering materials
with application for production of high-value crops in high tunnels.
Photochem. Photobiol., 81: 1047-1051.
Eyob, S., M. Appelgren, J. Rohloff, A. Tsegaye and G. Messele, 2008.
Traditional medicinal uses and essential oil composition of leaves
and rhizomes of korarima Aframomum corrorima (Braun) P.C.M.
Jansen) from southern Ethiopia. S. Afr. J. Bot., 74: 181-185.
Geoffrey, G.P. and J.B. Martin, 2000. Forest canopy stratification-Is it
useful? Am. Nat., 155: 473-484.
Jansen, P.C.M. 1981. Spices, Condiments and Medicinal Plants
in Ethiopia: Their Taxonomy and Agricultural Significance.
Agricultural research reports 906. Center for Agricultural Publishing
and Documentation, Wageningen, the Netherlands.
Kendrick, R.E. and G.H.M. Kronenberg, 1994. Photomorphogenesis
in Plants. 2nd ed. Kluwer Academic Publishers, Dordrecht,
Netherlands.
Lock, J.M. 1997. Zingiberaceae. In: Flora of Ethiopia and Eritrea, Vol.
6, S. Edwards, S. Demissew and I. Hedberg (eds.). Addis Ababa,
Ethiopia and Uppsala, Sweden. p. 324-326.
Murakami, K., H. Cui and M. Kiyota, 1997. Control of plant growth
by covering materials for greenhouses which alter the spectral
distribution of transmitted light, Acta Hort., 435: 123-130.
Rajapaske, N.C. and J.W. Kelly, 1992. Regulation of chrysanthemum
growth by light quality. J. Amer. Soc. Hort. Sci., 117: 481-485.
Shahak, Y., E.E. Gussakovsky, E. Gal and R. Ganelevin, 2004.
ColourNets: Crop protection and light-quality manipulation in one
technology. Acta Hort., 659: 143-151.
 Journal
Journal of Applied Horticulture, 11(1): 50-53, January-June, 2009
Appl

Effect of growth regulators on in vitro plant regeneration of

female papaya using axillary bud as an explant

Renu Singh, Ram C. Yadav* and Neelam R. Yadav

Department of Biotechnology and Molecular Biology, CCS Haryana Agricultural University, Hisar. 125004, India
*E-mail: renuboora@gmail.com

Abstract
A study was carried out on mature female papaya (Carica papaya L.) plant of Selection 1 cultivar by using axillary bud as an explant
and media supplementation with the main aim to assess the effect of growth regulators (auxins, cytokinins) and silver nitrate on in
vitro regeneration of female papaya plant. Total of 28 media were used for shoot regeneration while for root regeneration total of eight
media were tested supplemented with different growth hormones. Based on the results of this study, for shoot proliferation, MS basal
medium supplemented with BAP (1.0 mg L-1) and BAP (2.0 mg L-1) + NAA (0.1 mg L-1) was found to give the best results while MS
medium supplemented with IBA (2.0 mg L-1) gave best rooting percentage. Besides, auxins and cytokinins, effect of silver nitrate
(AgNO3) on plant regeneration from axillary buds taken from mature female papaya plant was also carried out.
Key words: Auxins, axillary bud, benzylaminopurine (BAP), cytokinins, indole acetic acid (IAA), Carica papaya L., α-naphthalene
acetic acid (NAA), proliferation, silver nitrate (AgNO3)

Introduction
Papaya (Carica papaya L.) is a native of tropical America and
belongs to family Caricacea. The plant starts fruiting in just
one year and gives economically high yield per acre next to
banana. The total area under papaya cultivation in India was
57,000 ha (FAO, 2000). Papaya is very nutritious and has much
therapeutic value. The fruit contains fair amount of vitamin A,
B1, B12 and C. The tremendous yielding potential of this crop is
left economically unexploited due to several problems associated
with its cultivation. The main problem associated is lack of
clonal multiplication methods and in tissue culture bacterial
contaminants are the limiting factors (Litz and Conover, 1981).
It is well known that vegetative propagation of papaya through
conventional methods has not been successful with commercial
viability. Thus, there is ample scope to overcome the above
limitations through tissue culture. The plant tissue culture
technique has been successfully employed in the regeneration
of various fruit crops like almonds, strawberry, citrus, etc
(Bajaj, 1986). However, papaya is a problematic crop for
micropropagation from grown up plants as it contains a good
amount of latex interfering in proliferation. Therefore, the present
study was undertaken with a view to evolve a tissue culture
technique in papaya and to evaluate the effect of different growth
regulators on the vegetative regeneration of female papaya plant
using axillary buds.
Materials and methods
The present investigation was carried out in Tissue Culture
Laboratory of the Department of Biotechnology and Molecular
Biology, CCS Haryana Agricultural University, Hisar, India.
Axillary buds as explants were excised from the field grown
sexually differentiated female plants. The chemicals of high purity
were used throughout the course of investigation. The excised
explants i.e. axillary buds were cut (6 mm) and washed with
single glass distilled water containing few drops of teepol. The
explants were then treated with 0.5 per cent Bavistin (fungicide)
for 45 minutes. The disinfections steps after bavistin treatment
were carried out in the horizontal laminar flow cabinet. Explants
were further rinsed 4-5 times with distilled water followed by
70 per cent alcohol treatment for 30 seconds and by 50 per cent
sodium hypochlorite for 3 minutes. After this treatment, they
were rinsed 4-5 times with double distilled water finally, given
0.1% HgCl2 treatment for 4 minutes and then the explants were
again rinsed 5-6 times with autoclaved double distilled water. In
the present study, MS basal medium supplemented with different
auxins and cytokinins were initially tried for culture establishment
(Murashige and Skoog, 1962). The medium supplemented with
kinetin, NAA, BAP and AgNO3 giving better response were
further used for regeneration and those which don’t gave better
regeneration were discarded in the initial of the experiment. All
the cultures were maintained at 25 ± 1°C under 16/8 hours cycles
of light (light intensity 50 μmol m-2 s-1) and dark. The regenerated
shoots were further cultured on regeneration medium with
variable concentration of growth hormones. The multiple shoots
formed were then transferred to rooting media in order to get root
formation. The data was recorded time to time and presented as
the mean of three repeats. Data were analyzed statistically using
one-way analysis of variance (ANOVA).
Results and discussion
In general, poor response for shoot regeneration on all the media
was observed when axillary bud was used as explant (Fig.
1A). Maximum (70.1%) response was observed on medium
supplemented with BAP (Table 1) while very poor response
was observed on the media supplemented with kinetin (Table
2). No shoot regeneration was observed on most of the Media
supplemented with kinetin.
 Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant 51

Maximum (70.1%) response (Fig. 1B) was observed on medium response of BAP was superior over kinetin by taking less time
supplemented with and moderate response was observed on for induction of shoot regeneration. In the present study, highest
MS supplemented with BAP (2.0) + NAA (0.1), BAP (1.0) + frequency of shoot regeneration and axillary bud proliferation
IAA (0.2) and BAP (2.0) + NAA (0.5) (65.0, 51.1 and 45.8%, could be achieved on BAP (1.0 mg L-1) without the need of
respectively), while rest of media gave poor response (Table 1). subculture. BAP is reported to have favoured axillary shoot
In case of kinetin maximum shoot regeneration (21.4%) response
proliferation in several tree species (Eeswara, 1998; Purohit and
was observed on medium supplemented with Kinetin (2.5) and
Dave, 1996; Purohit and Singhvi, 1998). In the present study, it
minimum response was observed on medium supplemented with
Kinetin (5.0) + NAA (0.2) (11.7%). was observed that an increase in the level of cytokinin from 1.0
to 5.0 mg L-1 produced a negative effect on all the parameters
When AgNO3 was added in the media (Table 3) supplemented except shoot number (Tables 1, 2).
with other growth hormones, the regeneration response was
Table 1. Shoot induction response from axillary bud explants on media
observed maximum (59.0%) on MS medium supplemented with supplemented with BAP in papaya
BAP (1.0) + AgNO3 (2.0) + Zeatin (3.0). Moderate (50.0%) Medium Average Average Mean (%) response
regeneration response was observed on MS + BAP (2.0) + AgNO3 Composition number of number of (Mean ±S.E.)*
(2.0) + Zeatin (2.5) medium, while MS medium supplemented (mg L-1) cultured responding
with AgNO3 (2.0) and AgNO3 (1.0) gave comparatively poor explants explants
MSP 50 0 00.0 (04.05 ± 0.01)
response (24.8 and 13.2%, respectively). No response was
MS + BAP (1.0) 100 70 70.1 (57.14 ± 0.06)
observed on MS medium without any growth regulators.
MS + BAP (2.5) 70 24 34.4 (36.14 ± 0.31)
The regenerated plantlets from micropropagation experiments MS + BAP (5.0) 74 19 25.4 (30.56 ± 1.10)
grown in vitro were transferred to different rooting media for MS + BAP (1.0) + 74 38 51.1 (45.90 ±0.63)
carrying out their rooting in vitro (Fig. 1C). Roots formed were IAA (0.2)
thick and strong (Fig. 1D). No response was observed when medium MS + BAP (2.5) + 80 15 18.7 (26.01 ± 0.70)
IAA (0.2)
was devoid of any auxin. Maximum rooting frequency (67.2%) was
MS + BAP (5.0) + 80 10 12.4 (21.01 ± 0.25)
observed on MS medium supplemented with IBA (2.0); however, IAA (0.2)
it was minimum (22.2 and 27.8%) on medium supplemented with MS + BAP (2.0) + 100 46 45.8 (42.88 ± 0.74)
IBA (0.1) and IBA (0.5), respectively (Table 4). NAA (0.5)
MS + BAP (2.0) + 100 65 65.0 (53.99 ± 0.22)
In vitro micropropagation has been successfully used for NAA (0.1)
many horticultural fruit trees (Das et al., 1996). Multiple shoot MS + BAP (1.0) + 58 12 20.6 (27.32 ± 0.75)
production from axillary buds obtained from mature trees is now NAA (0.2)
recognized as a better alternative of micropropagation in fruit trees MS + BAP (2.5) + 56 10 17.8 (25.32 ± 0.15)
where fidelity of the propagules is of prime importance (Quarashi NAA (0.2)
and Mitra, 1998; Tavares et al., 1996). MS + BAP (5.0) + 78 8 10.2 (19.04 ± 0.50)
NAA (0.2)
The overall conclusion from the above regeneration experiment LSD (P=0.05)=1.641, *Transformed value
emerged that the media comprised of MS basal salt + BAP (2.0 Table 2. Shoot induction response from axillary bud explants on media
mg L-1) + NAA (0.1 mg L-1) gave the best regeneration results supplemented with kinetin in papaya
in C. papaya. As outlined by Skoog and Miller (1957), root and Medium Average number Average Mean (%) response
shoot initiation is basically regulated by the interaction between Composition of cultured number of (Mean ±S.E.)*
(mg L-1) explants responding
auxins and cytokinins. This combination of both the hormones explants
was equally effective when regeneration was done with the shoot MSP 50 00 00.0 (04.05 ± 0.01)
tips therefore, we can conclude that a proper ratio of cytokinins MS + Kin (1.0) 54 00 00.0 (04.05 ± 0.01)
and auxins helps in both the shoot and root formation from MS + Kin (2.5) 74 16 21.4 (27.91 ± 0.53)
different explants. Axillary buds were associated with relative MS + Kin (5.0) 58 10 16.9 (24. 61 ± 0.68)
high level of endogenous growth substances in juvenile tissues in MS + Kin (1.0) + 84 00 00.0 (04.05 ± 0.01)
comparison to the adult tissues. This implies the poor regeneration IAA (0.2)
of axillary buds from mature plants than the explants taken from MS + Kin (2.5) + 48 00 00.0 (04.05 ± 0.01)
the juvenile plant. The un-branched character of the papaya plant IAA (0.2)
also offers a serious limitation in using axillary buds as explants MS + Kin (5.0) + 48 00 0.00 (04.05 ± 0.01)
IAA (0.2)
in the experiment.
MS + Kin (1.0) + 60 00 00.0 (04.05 ± 0.01)
Effect of cytokinins: The most striking influence on bud-break NAA (0.2)
and shoot multiplication has been found with the use of auxins MS + Kin (2.5) + 80 14 17.5 (25.06 ± 0.55)
NAA (0.2)
and cytokinins (Normanly, 1995). The most commonly used
MS + Kin (5.0) + 84 10 11.7 (21.07 ± 1.24)
cytokinins are BAP, kinetin, 2ip and zeatin, the latter two being NAA (0.2)
natural cytokinin. Superiority of BAP over other cytokinins has MS + Kin (0.5) + 74 00 00.0 (04.05 ± 0.01)
been reported and discussed in relation to shoot proliferation in 2,4-D (2.5)
culture of trees (Bonga and Jvon Aderkas, 1992). In the present MS + Kin (1.0) + 100 13 12.0 (20.67 ± 0.28)
study, lower level (1.0 mg L-1) of BAP and of kinetin (2.5 mg L-1) 2,4-D (2.5)
induced highest frequency of shoot regeneration. Regeneration LSD (P=0.05)=1.379, *Transformed value
52 Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant

Table 3. Shoot induction response from axillary bud explants on media supplemented with AgNO3 in papaya
Medium composition Average number of cultured Average number of responding Mean (%) response
(mg L-1) explants explants (Mean ±S.E.)*
MS 60 0 00.0 (04.05 ± 0.01)
MS + AgNO3 (2.0) 120 30 24.8 (30.16 ± 0.52)
MS + AgNO3 (1.0) 106 14 13.2 (21.71 ± 0.32)
MS + BAP (1.0) + AgNO3 (2.0) + Zeatin (3.0) 100 59 59.0 (50. 48 ± 0.29)
MS + BAP (2.0) + AgNO3 (2.0) + Zeatin (2.5) 100 50 50.0 (45.27 ± 0.01)
LSD (P=0.05)=1.120, *Transformed

Effect of cytokinins (BAP and kinetin) in combination

with auxins (NAA): In the present study combined effect of
cytokinin and auxin was promotive on shoot regeneration from
nodal explant, however, an inhibitory effect was observed in the
combination where levels of cytokinins were high (5.0 mg L-1).
Auxins (NAA, 2, 4-D and IAA) were also examined for their
influence on multiple shoot production and growth of papaya
axillary bud cultures. Auxins were found to affect only the A B
growth of the cultures rather than proliferation rate. NAA (0.1-
0.2 mg L-1) resulted in better growth of the cultures and hence
included in the medium. Similar response by shoot and axillary
bud cultures with regard to auxin application were reported in
papaya (Buriken et al., 1988; Renveni, 1990). As a promotive
effect on shoot regeneration in combination of BAP/kinetin with
NAA, maximum shoot regeneration frequency in axillary bud
explant (65.0%) was observed on basal medium with BAP (2.0
mg L-1) and NAA (0.1 mg L-1) while kinetin was found to have less
promotive effect in combination with NAA. The complementary C D
effect of cytokinin and auxin has been observed by Miller and
Fig. 1. A. Nodal explants cultured on regeneration medium; B. Multiple
Drew (1990) during shoot proliferation in papaya. Highest shoot shoot formation from axillary explants; C. Regenerated shoot on rooting
production was achieved on medium containing BAP (2 μM) and medium; D. Regenerated shoots with roots on rooting medium
NAA (0.5 μM). A combination of BAP (2.0 mg L-1) and NAA (1.0
mg L-1) was found to enhance shoot bud proliferation in cultured Rajeevan and Pandey (1983), Singh et al., (2000). Although IAA
shoot tip in papaya (Litz and Conover, 1981). Combination of also induced root formation but the root development was poor
cytokinin and auxin thus has been found to promote shoot bud as compared to IBA. The strength of the basic medium employed
proliferation effectively. is reported to influence root initiation in several papaya cultivars.
However, better root induction has been obtained in papaya with
Effect of auxins: Various auxins were tried with the objective half strength MS medium (Renveni et al., 1990).
to induce roots in shoots. In this study, medium containing MS
basal + IBA (2.0 mg L-1) was found most appropriate for root Effect of silver nitrate (AgNO3): AgNO3 is an ethylene inhibitor
induction followed by MS basal + IBA (1.0 mg L-1). However, the in plants which is reported to help in somatic embryogenesis
induction of rooting in the regenerated shoots of papaya was found (Songstad et al., 1991). AgNO3 may also serve as stress agent
to be very difficult. Plantlets produced with IBA were normal inducing endogenous ABA accumulation, Ag+ being a metallic
with regard to the shoot and root growth. IBA has been found to ion may also promote somatic embryo production via an increase
be effective for root induction in papaya-regenerated shoots by in the endogenous ABA levels (Kong and Yeung, 1994, 1995). But
Bhattacharya et al., (2002), Buriken et al., (1988), Fitch (1993), in our experiment when AgNO3 was added in the media (Table
Table 4. Root induction response in shoots regenerated from axillary 3) supplemented with other growth hormones, the regeneration
bud explant on various rooting media in papaya response was observed maximum (59.0%) on MS supplemented
IAA NAA IBA Number Number of Mean (%) with BAP (1.0) + AgNO2 (2.0) + Zeatin (3.0). Thus, overall in
(mg L-1) (mg L-1) (mg L-1) of cultured responding response conclusion media containing AgNO3 has no promotive effect on
explants explants (Mean ±S.E.)*
shoot formation.
- - - 6 00 00.0 (04.05 ± 0.01)
2.0 - - 7 00 00.0 (04.05 ± 0.01)
- - 0.1 12 2 15.0 (20.43 ± 8.25)
References
- - 0.5 9 2 19.4 (23.30 ± 9.74) Bajaj, Y.P.S. 1986. Biotechnology of trees improvement for rapid
- - 1.0 9 4 44.4 (42.02 ± 3.25) propagation and biomass energy production. In: Biotechnology in
Agriculture and Forestry, Vol. 1. Trees. Bajaj,Y.P.S (ed.), Springer,
- - 2.0 11 7 63.9 (52.46 ± 3.65)
Berlin. 1-23.
- 0.5 0.1 10 3 30.5 (33.79 ± 1.74)
Banerjee, J. 2002. Tissue Culture and Transformation Studies in
- 0.5 0.5 9 3 33.3 (35.53 ± 0.01)
Indian Cultivars of Papaya (Carica papaya L.), Ph.D. Thesis,
LSD (P=0.05)=1.431, *Transformed value Biotechnology Dept, University of Pune, India.
 Effect of growth regulators on in vitro plant regeneration of female papaya using axillary bud as an explant
Bhattacharya, J., S.S. Khuspe, N.N. Renukdas and S.K. Rawal, 2002.
Somatic embryogenesis and plant regeneration from immature
embryo explant of papaya (Carica papaya L. cv. Washington and
Homey Dew.). Indian J. Exp. Bioi., 40: 624-627.
Bonga, J.M. and P. Ven Aderkas, 1992. In vitro culture of tree. Kluwer
Academic Publishers, The Netherlands. 236p.
Burikan, S., V. Chommalee and S. Attathon, 1988. Effects of plant growth
regulators on papaya ( Carica papaya) cultured in vitro. Kasetsart
Journal Natural Sciences, 22 (5): 1-6.
Chengalrayan K., V.B. Mhaske and S. Hazra, 1997. High frequency
conversion of abnormal peanut somatic embryos. Plant Cell
Report,16:783-786.
Das, S., B.J. Timir and S. Jha, 1996. In vitro propagation of cashew nut
(Anacardium occidenatle). Plant Cell Reports, 15: 615-619.
Eeswara, J.P. 1998. A standard procedure for the micropropagation of
the Neem tree (Azadirachta indica A. Juss.). Plant Cell Rept., 17:
215-219.
FAOstats. 2001. www.fao.org
Fitch, M.M.M. 1993. High frequency somatic embrygogenesis and plant
regeneration from papaya hypocotyls callus. Plant Cell Tissue Organ
Culture, 32: 205-212.
Kong, L. and E.C. Yeung, 1994. Effects of ethylene and ethylene
inhibitors on white spruce somatic embryo maturation. Plant Science,
104: 71-80.
Kong, L. and E.C. Yeung, 1995. Effects of silver nitrate and polyethylene
glycol on white spruce (Picea glauca) somatic embryo development:
enhancing cotyledonary embryo formation and endogenous ABA
content. Physiol Plant., 43: 298-304.
Litz, R.E. and R.A. Conover, 1981. Effect of sex type, season and other
factors on in vitro establishment and culture of Carica papaya L.
explants. J. Amer. Soc. Hort. Sci., 106: 792-794.
Miller, R.M. and R.A. Drew, 1990. Effect of explant type on proliferation
of Carica papaya L. in vitro. Plant Cell Tissue Organ Culture, 21
(1): 39-44.
53
Murashige, T. and F. Skoog, 1962. A revised medium for rapid growth
and bioassays with tobacco tissue cultures. Physiol. Plantarum, 15:
473-497.
Murashige, T. 1974. Plant propagation through tissue culture. Ann. Rev.
Plant Physiol., 25: 135-166.
Normanly, J. 1995. Rethinking auxin biosynthesis and metabolism. Plant
Physiol., 107: 323-329.
Purohit, S.D. and A. Dave, 1996. Micropropagation of Sterculia urens
Roxb.: An endangered tree species. Plant Cell Report, 15: 704-706.
Purohit, S.D. and A. Singhvi, 1998. Micropropagation of Achras sapota
through enhanced axillary branching. Sci. Hort., 76: 219-229.
Quarashi, A. and S.K. Mitra, 1998. Micropropagation of nodal explants
from adult trees of Cleistanthus collinus. Plant Cell Report, 17:
430-433.
Rajeevan, M.S. and R.M. Pandey, 1983. Propagation of papaya through
tissue culture. Acta Horticulturae, 131: 131-139.
Renveni, R.M. 1990. Propagation of papaya through tissue culture. Plant
Cell Tissue Organ Culture, 13: 77-83.
Singh, H.P., S. Singh, R.P. Saxena and R.K. Singh, 1993. In vitro bud
break in axillary nodal segments of mature trees of Acacia nilotica.
Indian J. Plant Physiol., 36: 21-24.
Skoog, F. and C.O. Miller, 1957. Chemical regulation of growth and
organ formation in plant tissue cultured in vitro. Sym. Exp. Biol.,
11: 118-130.
Songstad, D.D., C.L. Armstrong and W.L. Petersen, 1991. AgNO3
increases Type II callus production from immature zygotic embryos
of inbred B73 and its derivatives. Plant Cell Reports, 9: 699-702.
Tavares, A.C., M.C. Pimenta and M.T. Goncalves, 1996. Micropropagation
of Melissa officinalis L. through proliferation of axillary shoots. Plant
Cell Report, 15: 441-444.
 Journal
Journal of Applied Horticulture, 11(1): 54-55, January-June, 2009
Appl

Effects of the addition of clinker ash to the propagation

medium on rooting of rabbiteye blueberry cuttings

T. BanA*, H. KitazawaB, S. MatsumotoA, N. KobayashiA, K. TokumasaC, M. KobatakeC

and T. AsaoA
A
Faculty of Life and Environmental Science, Shimane University, 2059 Kamihonjyo, Matsue, Shimane, 690-1102, Japan,
B
United Graduate School of Agricultural Science, Tottori University, Koyama-cho Minami Tottori, Tottori, 680-8553, Japan,
C
Energia Economic and Technical Research Institute, The Chugoku Electric Power Co., Inc., Kagamiyama, Higashihiro-
shima, Hiroshima, 739-0046, Japan. E-mail: *ban@life.shimane-u.ac.jp

Abstract
The recommendation for a propagation medium of rabbiteye blueberry (Vaccinium ashei Reade) in Japan includes the incorporation
of peat moss and Kanumatsuti (a volcanic ash deposit). This experiment compared the use of coal ash (clinker ash) and Kanumatsuti
to peat moss as soil conditioner for rooting rabbiteye blueberry cutting. The numbers of cuttings survived and the root dry weight of
plants propagated in clinker ash- peat moss mixes were almost the same as cuttings propagated in kanumatsuti- peat moss mix. While
the quadratic model between the root dry weight and the clinker ash content in the medium was significant, the maximum root dry
weight was estimated to reach about 0.2 g when the proportion of clinker ash in the medium was about 40%. These findings indicate
that clinker ash can be used in the propagation medium of rabbiteye blueberry.
Key words: Clinker ash, cutting, propagation, rabbiteye blueberry, rooting

Introduction Materials and methods

Coal continues to play a vital role in electricity generation in the
world. While coal-fired electrical generation supplied 39% of
the world’s electricity in 2002, this value is likely not to change
over the next three decades (World Coal Institute, 2005a). After
burning of coal for electric purposes, a huge amount of coal
ash is generated. Coal ash mainly consists of SiO2, Al2O3 and
Fe2O3 and is classified into fly and clinker ash depending on the
production processes (Schlossberg et al., 2004). Clinker ash,
which is coarser than fly ash, is a dark gray, granular and porous
substance (Table 1).
Coal ash production will increase continuously in the coming
three decades, hence its recycling for a wide range of uses is
required. Coal ash has been utilized as building material and for
other constructional purpose (World Coal Institute, 2005b). Since
coal ash has desirable horticultural characteristics, including low
cost, good drainage and its soil amendment effect, it has been
started to be utilized as a horticultural material (Schlossberg
et al., 2004; Stevens and Dunn, 2004). Black and Zimmerman
(2002) showed the growth and yield of highbush blueberry plants
in coal ash- compost mixes were similar to or exceeded those of
plants in berryland sand. Rabbiteye blueberry can be propagated
by hardwood cuttings and have been propagated in many kinds
of medium, including peat moss, sand, sawdust, perlite and
pine bark (Austin, 1994). In New Jersey, North California and
Massachusetts, a mixture of half sand and half peat moss are used
to root blueberry plants (Mainland, 1966).
In this study, we investigated the effects of adding clinker ash to
the propagation medium on the rooting of rabbiteye blueberry
and evaluated whether it could serve as a suitable component of
the propagation medium for blueberry cuttings.
Clinker ash (3~7 mm in diameter) was obtained from the Chugoku
Electric Power Co., INC. Misumi power plant in Shimane, Japan.
The components of the propagation medium were combined in
the proportions listed in Table 2. Kanumatsuti, a volcanic ash
deposit, was used in our studies as a substitute for the control,
since a mixture of peat moss and Kanumatsuti is generally used
as the propagation medium of blueberry plants in Japan (Tamada,
1997). Mature bushes of rabbiteye blueberry ‘Tifblue’ growing
on the experimental farm of Shimane University were used. On
14 Feb, 2005, well-matured and healthy shoots of the former
growing season were collected and stored at 4˚C. Fifty cuttings
(10 cm in length) were prepared from the shoots for each medium
combination on Apr, 7, 2005. Before planting the cuttings in the
mediums, we confirmed that the mean diameter of cuttings was
not significantly different between treatments (Table 3). The
cuttings were planted in 9 cm plastic pots containing 300 mL of
each medium. After adequate watering, the pots were placed in
plastic propagation tunnels. Mist spray was applied from 08: 30
to 09: 00 and 16: 00 to 16: 30 daily.
The numbers of cuttings that formed callus, rooted and survived
cuttings were recorded when almost all of the leaves fallen
(about seven months after their planting). Root dry weight was
determined by weighting roots after oven drying at 80o for 7
days. The measurements of root dry weight were replicated 5
times per treatment.
Results and discussion
Almost all the cuttings formed callus, but the number of
cuttings that rooted and survived decreased with an increase
in the proportion of clinker ash in the medium (Table 3). It is
 Effects of the addition of clinker ash to the propagation medium on rooting of rabbiteye blueberry cuttings 55

Table 1. Physical properties of clinker ash

Location of power plant Soil phase ( vol % ) Water retention Bulk density
characteristics
Solid phase Liquid phase Air phase (pF=1.5, wt %)
Misumi in Shimane, Japan 45.2 39.3 15.5 30 0.86
Shinonoda in Yamaguchi, Japan 50.7 28.2 21.1 19.1 1.06
Table 2. Composition of medium treatments expressed as a relative 0.4
percentage
Treatment Component ( vol % )
Clinker ash Kanumatsuti Peat moss y=-6x10 - 5x 2+0.004x+0.081
1(Control) - 50 50 0.3 R2= 0.448*

Root dry weight (g)

2 0 - 100
3 25 - 75
4 50 - 50
0.2
5 75 - 25
6 100 - 0

well known that solid agents in plant propagation medium can 0.1
regulate moisture reserves and improve aeration, since these
substances have a range of porosity. The results obtained in our
study indicate that drought damages was induced with an increase
in the proportion of clinker ash in the medium. 0
0 25 50 75 100
The No. 3 and 4 medium resulted in the highest root dry weights Clinker ash component in medium (%)
and there was no significant differences in root dry weight between Fig. 1. Relationship between the root dry weight and clinker ash
the No. 1 (control), 3 and 4 mediums (Table 3). In medium No. composition in the medium. The curve was tested with a polynomial
regression analysis. * means significant at the 5% level.
Table 3. Effects of the addition of clinker ash to the propagation medium on the rooting of rabbiteye blueberry cuttings
Treatment Average diameter of Number of callus Number of rooted Number of survived Root dry weight
cuttings (mm) formed cuttings cuttings cuttings (g)
1(Control) 8.30 NSa 50 48 48 0.128 abb
2 8.17 NS 49 49 49 0.069ab
3 8.38 NS 50 50 50 0.197a
4 8.32 NS 50 46 46 0.168a
5 8.48 NS 49 43 41 0.124ab
6 8.26 NS 50 26 22 0.017b
Non significant at 5% level by the Tukey’s test. dDifferent letters within a column indicate significance at 5% level by the Tukey’s test.
a

2 and 6, the root dry weight was lower than that of the other
four mediums. Fig. 1 shows a significant relationship between
the root dry weight and the clinker ash content in the medium.
While this quadratic model was significant, the maximum root dry
weight was estimated to reach about 0.2 g when the proportion of
clinker ash in the medium was about 40%. In general, propagation
medium of cuttings requires a high moisture holding ability and
good drainage (Preece and Read, 1993). The low values in root
dry weight from No. 2 medium may be attributed to excessive
moisture level, since this medium consisted of only peat moss.
In contrast, a high proportion of clinker ash in the propagation
medium will inhibit the root growth of rabbiteye blueberry
cuttings, because its moisture holding ability will be low.
In the present study, we clarified certain effects of clinker ash
when added to the propagation medium on the rooting of rabbiteye
blueberry cuttings. The growths of cuttings in the clinker ash-
peat moss mixes was similar to or exceeded that of cuttings in
normal medium. These findings indicate that clinker ash can
be successfully used as a propagation medium for rabbiteye
blueberry cuttings.
References
Austin, M.E. 1994. Propagation. In: Rabbiteye Blueberries. AGSCIENCE,
INC., Florida, USA. p.29-33.
Black, B.L. and R.H. Zimmerman, 2002. Mixtures of coal ash and
compost as substrates for highbush blueberry. J. Amer. Soc. Hort.
Sci., 127: 869-877.
Mainland, C.M., 1966. Propagation and planting. In: Blueberry
Culture. Eck, P. and N.F. Childers (eds.). Rutgers Univ. Press, New
Brunswick, Canada. p. 111-131.
Preece, J.E. and P.E. Read, 1993. The Biology of Horticulture. John
Wiley & Sons, Inc., New York, USA. p. 323-352.
Schlossberg, M.J., C.P. Vanags and W.P. Miller, 2004. Bermudagrass sod
growth and metal uptake in coal combustion by-product-amended
medium. J. Environ. Qual., 33: 740-748.
Stevens, G. and D. Dunn, 2004. Fly ash as a liming material for cotton.
J. Environ. Qual., 33: 343-348.
Tamada, T., 1997. Guide for blueberry production [11]. Agri. Hort., 72:
92-98 (in Japanease).
World Coal Institute, 2005a. The global coal market. In: The Coal
Resource. World Coal Institute, London, U.K. p.13-18.
World Coal Institute, 2005b. How is coal used ? In: The Coal Resource.
World Coal Institute, London, U.K. p. 19-25.
 Journal
Journal of Applied Horticulture, 11(1): 56-58, January-June, 2009
Appl

Persian walnut (Juglans regia L.) grafting as influenced by

different bench grafting methods and scion cultivars

Babak Dehghan1, Kourosh Vahdati1*, Reza Rezaee2 and Darab Hassani3

1
Department of Horticulture, College of Abouraihan, University of Tehran, Tehran, Iran, 2Agricultural and Natural
Resources Research Center, West Azerbaijan, Uromia, Iran, 3Department of Horticulture, Seed and Plant Improvement
Institute (SPII), Karaj, Iran. *E-mail: kvahdati@ut.ac.ir

Abstract
The study was conducted to determine the effects of different grafting methods and scion cultivars on walnut grafting under controlled
conditions in March 2006. Four walnut cultivars (‘Z53’, ‘Hartley’, ‘Pedro’ and ‘Serr’) were grafted using three bench grafting methods
(side stub, omega and whip tongue) onto dormant two years old Persian walnut seedlings as rootstock. The plants after grafting were
covered with moist sawdust with relative humidity of 85- 90% and stored in a humid room at 26-28 ºC for 21 days. Based on the results,
the highest grafting success was observed with omega (84.33%) followed by side stub (41.89%) and whip tongue (24.31%) grafting,
respectively. Significant variations were also observed in graft take and scion growth. The differences among walnut cultivars (scion)
on grafting take and scion growth were not significant. However, the scions x grafting methods interaction was significant and ‘Hartley’
variety grafted by omega method showed the highest graft take (88.44%) among all combinations. A significant positive correlation
(R2 = 0.84) was observed between the callus quality and graft takes in all grafting methods.
Key words: Callus formation, grafting techniques, graft survival, greenhouse, sawdust, walnut cultivars.

Introduction
Persian walnut (Juglans regia L.) is an important nut crop and still
being propagated by seedlings in several countries including Iran,
which resulted to high variability and poor crop quality (Vahdati,
2000). Selection and vegetative propagation of the superior walnut
cultivars of different agroclimatic areas is the most efficient
method for increasing the production and nut quality (Solar et
al., 2001). Vegetative propagation in walnut is more difficult in
comparison with other fruit trees (Kuden and Kaska, 1997; Ozkan
and Gumus, 2001) and low success has always been considered a
drawback in massive propagation of superior walnut individuals
(Ozkan and Gumus, 2001; Vahdati, 2003). Walnut grafting success
were reported to be affected by several factors including graft
technique, temperature, humidity, phenolic compounds, hormonal
condition, nutrition of scion cultivars, and time of taking the
scions (Mitrovic, 1995; Mehmet et al., 1997).
Environmental conditions during and following grafting
have major impact on callus formation in walnut (Millikan,
1971; Rongting and Pinghai, 1993; Avanzato and Atefi, 1997;
Ebrahimi et al., 2006). In Persian walnut, for successful grafting,
temperature around the grafting point should be maintained at
about 27 °C after grafting (Avanzato and Atefi, 1997; Germain
et al., 1997).
In walnut vegetative propagation, fluctuating temperatures and
lack of sufficient humidity cause undesirable environment under
field condition responisble for poor callus formation and grafting
failure (Ebrahimi et al., 2006), thus bench grafting methods are
being usually preferred. The advantages of these techniques
include:1) table grafting could be done under controlled conditions
and gives a better result, 2) the operation could be done during
winter and 3) grafting could be mechanized to increase labour
productivity (Lantos, 1990; Tsurkan, 1990; Unal, 1995).
New methods using bench grafting techniques like hot cable
or hot callus have been developed and used by different
researchers (Avanzato and Atefi, 1997; Hartmann et al., 2001),
but requirement of more skillful workers and expensive facilities
restrict their application in most nurseries. Therefore, the main
goal of the present study was to evaluate the efficiency of modified
bench grafting methods in order to decrease the high cost of
production by grafting in walnut as well as to compare different
walnut cultivars in terms of grafting success and scion growth.
Materials and methods
The experiment was conducted in Department of Horticulture,
College of Abouraihan, University of Tehran during March
2006, to evaluate effect of factorial combination of three
grafting methods (side stub, omega and whip tongue) and four
walnut cultivars (‘Z53’, ‘Hartley’, ‘Pedro’ and ‘Serr’) on callus
formation, grafting success and scion growth. The scions were
stored at cool and moist condition, until they were used in grafting.
The seedling rootstocks were taken in late January and selected
for size and uniformity. The study was conducted using a factorial
experiment on a complete randomized design with 12 treatments
in three replications and 15 seedlings per plot.
All of the grafted combinations were carried out by the same
person and standard methods were used for grafting as described
by Hartmann et al. (2001) briefly as following. For side-stub
grafting, the basal part of the scions were cut as wedge of about
2.5 cm and an oblique cut at angle of 20º to 30º was made into
the basal part of the seedling’s stem then scion were inserted,
without any fastening material. In whip grafting, a sloping cut
(2-3 cm in length) was made at the top of the rootstock, and then a
second downward cut was made starting one-third of the distance
from the tip to the base of the first cut. Similarly, a sloping cut
 Persian walnut grafting as influenced by different bench grafting methods and scion cultivars 57

Table 1. Effects of grafting methods on callus quality, graft take, survival

and scion growth
Grafting type Callus Graft takes Graft Scion
qualityA (%) survival (%) growth (cm)
Omega 2.53aB 67.77a 84.33a 12.9a
Side stub 2.28a 58.88a 41.89b 5.38b
Whip tongue 0.97b 19.44b 24.31b 3.06b
A
Values are means of callus scoring rating from 1 (low callus) to 4
(very good callus). BMeans with different letters in each column are
significantly different at P<0.05.
efficient treatment was similar to Lantos (1990) who obtained
80% of grafting success using bench grafting method. Among
three methods of grafting, omega grafting showed the best result
considering all of studied characters which is consistent with
Fig. 1: A view of grafted plants after two months of grafting and Solar et al. (2001).
transferring to pots. A) Whip tongue grafting method, B) Omega grafting,
and C) Side stub grafting method. Significant differences was observed among the cultivars in terms
of callus quality and graft survival, but the effect of variety on
was made at the base of the scions to the same length as the cut grafting take or scion growth was not statistically significant
on the rootstock. Also a second cut was made under the first cut (Table 2). This is in contrary with the result of Rongting and
similar to the stock. Finally, the graft compounds were slipped Pinghai (1993) and Mitrovic et al. (1997) who reported the
together to be interlocked, and were tied by rubber tape tightly
without waxing. The omega grafting method was performed by
omega grafting machine (OMEGA-STAR Company of Germany)
without any fastening similar to the side stub grafting method.
Grafted plants were maintained in a greenhouse at 26-28 ºC
inside the sawdust with the moisture content of 85-90% for 21
days. After completing callusing period, data were recorded for
amount of callus formation based on a visual scale of one to four
in which 1= poor, 2= medium, 3= high and 4= very high callusing.
Subsequently, the percentage of callus formation, graft take, and
scions growth were measured in each treatment. The grafted plants
then were transferred to the black polyethylene pots (20 × 30 cm) Fig. 2. Correlation between callus quality and graft take in different
containing sand: soil: leaves compost mixture (1:2:1 w/w), and grafting methods.
were put in a humid greenhouse for about three months (Fig. 1). significant effect of walnut cultivars on graft take. Although scion
They were brought to the shade house in September, 2007, before cultivar could affect grafting result, but it seems that it is mainly
transplanting to the field condition. affected by scion quality which is a management related issue
rather than their genetic makeup (Rezaee and Vahdati, 2008).
The collected data were analyzed by SAS statistical software (SAS
Institute, North Carolina, USA) and means compared using Duncan’s The interaction of grafting methods and variety on graft take,
multiple range test (DMRT) method. Statistical significance indicates grafting survival and scion growth were significant (Table 3).
means difference (P<0.05). The highest graft take (88.44%) was achieved by omega grafting
of ‘Hartley’ and the lowest graft take by whip tongue grafting of
Results and discussion ‘Z53’ and ‘Hartley’ cultivars with 0.00 and 13.33 %, respectively.
The highest scion growth (14.66 cm) was in ‘ Z53’ variety with
The results showed that different grafting methods had omega grafting method and lowest scion growth (2.22 cm) was
significantly different effects on callus quality, graft take and obtained from ‘Serr’ variety with side stub grafting method.
grafting survival (Table 1). Among the methods, omega grafting The highest graft survival (96.67 and 90.91%) was obtained by
showed the highest callus quality (2.5 from 4), higher graft grafting of ‘Z53’ and ‘Pedro’ cultivars using omega and side stub
take (67.77%) and grafting survival (84.33 %) as well as higher grafting methods, having the highest score of callus quality of
scion growth (12.9 cm) followed by side stub and whip grafting. 2.76 and 2.75, respectively
Percentage of graft takes in side stub and whip tongue methods
were 58.88 and 19.44%, respectively. However, percentages of Table 2. Callus quality, graft take, survival and scion growth in different
walnut cultivars
grafting survival were about 41.89% in side stub and 24.31% in
Cultivar Callus qualityA Graft takes Graft survival Scion growth
whip tongue methods (Table 1). (%) (%) (cm)
As demonstrated in Fig. 2, percentage of graft take shows a ‘Serr’ 2.24aB 45.92a 48.61ab 5.82a
highly positive correlation with callus quality at different grafting ‘Pedro’ 2.08a 54.07a 73.27a 8.16a
methods and cultivars. This is in agreement with the results of ‘Hartley’ 2.01a 52.59a 42.22ab 8.57a
Rongting and Pinghai (1993) that callus quality and amount of ‘Z53’ 1.37b 42.22a 36.67b 5.90a
callus formation plays an important role in the grafting success. A
Values are means of callus scoring ratings from 1 (low callus) to 4
(very good callus). BMeans with different letters in each column are
In this experiment, the percent of grafting success in the most significantly different at P <0.05.
58 Persian walnut grafting as influenced by different bench grafting methods and scion cultivars

Fig. 3. Interlocking and quality of callus bridge formation between rootstock and scion in the studied grafting methods. A, D)
Whip tongue method, B, E) Side stub method, C, F) Omega method
The results of this study were different than Ozcan and Gumus Table 3. Interaction of grafting methods × cultivar on callus quality, graft
(2001) who found the highest success by using whip tongue take, survival and scion growth
grafting method. It could be probably related to the experimental Grafting × variety Callus Graft take Graft Scion
qualityA (%) survival growth
condition; so that tight binding of the grafting place by rubber tape (%) (cm)
in our experiment, resulted to limitation in aeration and blockage Omega × ‘Z53’ 2.76a B
64.44ab 96.67a 14.66a
of the phloem sap transportation under the graft cover. Adequate
aeration and auxins have an important roles on callus formation Omega × ‘Hartley’ 2.71a 88.44a 80abc 14.34a
and grafting success (Ronging and Pinghai, 1993; Vahdati, 2000; Omega × ‘Pedro’ 2.07ab 57.78bc 81.67abc 11.07ab
Hartmann et al., 2001; Rezaee and Vahdati, 2008). Omega × ‘Serr’ 2.57a 64.44ab 79.00abc 10.91abc
Interlocking and quality of callus formation between rootstock Side stub × ‘Z53’ 1.36cb 13.33d 11.11c 3.067bcd
and scion in the omega and side stub grafting methods were Side stub × ‘Hartley’ 2.33a 60.00ab 30.00cd 7.44abcd
approximately similar. Despite a good interlocking of grafting Side stub × ‘Pedro’ 2.75a 68.89ab 90.91ab 8.82abc
compound, as shown in Fig. 3, formation of callus bridge was
Side stub × ‘Serr’ 2.71a 44.44bcd 33.33bcd 2.22abc
restricted in whip tongue grafting method.
Whip tongue × ‘Z53’ 0.00d 0.00f 0.00d 0.00d
Considering the difficult-to-graft nature of walnuts and higher Whiptongue בHartley’ 1.00c 13.33ef 16.67d 3.93bcd
cost of specialized facilities used in methods like hot cable and
Whip tongue × ‘Pedro’ 1.43bc 35.56cb 41.22abcd 3.98bcd
hot callus (Avanzato and Atefi, 1997; Avanzato, 2001) the results
of this experiment are comparably promising, and represents a Whip tongue × ‘Serr’ 1.45bc 28.89ed 33.33bcd 4.33bcd
good alternative method to propagate walnut cultivars under A
Values are the means of callus scoring ratings from 1 (low callus) to
partially controlled condition. Further research is needed for 4 (very good callus). BMeans with different letters in each column are
significantly different at P<0.05.
improvement of grafting techniques and condition to obtain more
Mehmet, S., T. Karadanize, F. Balta and E.Tekintas, 1997. Changing of
uniform grafting success. flavan contents at some organs of walnut seedling (Juglans regia
L.) exposed to the controlled grafting conditions. Acta Hort., 442:
Acknowledgment 181-184.
Millikan, D.F. 1971. ‘Propagation of Juglans species by fall grafting.’
We thank Dr. Reza Amiri for his assistance in analyzing the Annual Report: Northern Nut Growers. Association, 61: 41-44.
data. University of Tehran and Iran National Science Foundation
Mitrovic, M. 1995. Effect of the cutting date of walnut scion wood on the
(INSF) are also acknowledged for their financial supports. take and callusing of grafts. Jugoslovensko Vocarstvo, 29: 59-63.
Ozkan, Y. and A. Gumus, 2001. Effects of different applications on
References grafting under controlled conditions of walnut. Acta Hort., 544:
Avanzato, D. 2001. Effect of different hygro-thermic environments on 515-520.
growth of potted walnut grafted seedlings. Acta Hort., 544: 459-464. Rezaee, R. and K. Vahdati, 2008. Introducing of a simple and effective
Avanzato, D. and J. Atefi, 1997. Walnut grafting by heating the graft procedure for topworking Persian walnut tree. J. Amer. Pom. Sci.,
point directly in the field. Acta Hort., 442: 291-294. 62: 21-26.
Ebrahimi, A., K. Vahdati and E. Fallahi, 2006. Improved success of Rongting, X. and D. Pinghai, 1993. A study on the uniting process of
Persian walnut grafting under environmentally controlled conditions. walnut grafting and the factors affecting. Acta Hort., 311: 160-170.
Int. J. Fruit Sci., 6: 3-12. Solar, A., F. Stampar, M. Trost, J. Barbo and S. Avsec, 2001. Comparison
Germain, E., F. delort and V. Kanivets, 1997. Precocious maturing walnut of different propagation methods in walnut (Juglans regia L.) made
population originating from central Asia: their behaviour in France. in Slovenia. Acta Hort., 544: 527-530.
Acta Hort., 442: 83-90. Tsurkan, L.P. 1990. Production technology of English walnut planting,
Hartmann, H.T., D.E. Kester, F.T. Davies and R. Geneve, 2001. Plant materializing winter table grafting. Acta Hort., 286: 65-68.
Propagation: Principles and Practices. 7th ed. Prentice Hall, New Unal, A. 1995. Studies on the effects of different practices on grafting
Jersey. success in bark, whip and top-cleft grafting in walnuts. E.U.Z.F.
Kuden, A. and N. Kaska, 1997. Studies on the patch budding of walnuts Dergisi Izmir / Turkey 32: 85-90.
in different budding periods under subtropical conditions. Acta Vahdati, K. 2000. Walnut situation in Iran. Nusis Newsletter, 9: 32-33.
Hort., 442: 299-301.
Vahdati, K. 2003. Nursery Management and Grafting of Walnut. 2th ed.
Lantos, A. 1990. Bench grafting of walnut. Acta Hort., 284: 53-56. Khaniran Publ,Tehran.
 Journal
Journal of Applied Horticulture, 11(1): 59-63, January-June, 2009
Appl

Extraction and determination of α-solanine in eggplant fruits

Zhiwen Li1, Baoli Zhou1*, Yuwen Ding1 and Xiang Liu2

College of Horticulture, Shenyang Agricultural University, Shenyang 110161, China; 2College of Agriculture and
1

Bioengineering, Tianjin University, Tianjin 300072, China. *E-mail: zblaaa@163.com

Abstract
A simple and effective high-performance liquid chromatographic (HPLC) method for determination of α-solanine in eggplant fruits
is described in our study. A new extraction method is established for extracting α-solanine in eggplant fruits. Single and orthogonal
tests were designed to analyze the effect of different extraction methods and ultrasonic wave extraction condition on extraction of
α-solanine in eggplant fruits. HPLC separation was achieved on a Waters Nova-pak C18 column with the mobile phase acetonitrile-
0.05N potassium dihydrogen phosphate (55:45, V/V). The flow rate was 0.7mL min-1 and the UV absorbance was monitored at 202
nm. The optimal extraction method was ultrasonic wave extraction in 70% methanol for 60 minutes at 50oC, and with material to liquid
ratio of 1:10. Under the optimal extraction conditions, the average content of α-solanine in skins and flesh of dried eggplant fruits was
0.107±0.006 and 0.626±0.004mg g-1, respectively. The average recovery efficiency was 97.97%.
Key words: Eggplant, α-solanine, HPLC, extraction, ultrasonic wave

Introduction these methods include isotachyphoresis (Kvasnicka et al., 1994),

Eggplants (Solanum melomgena L.) produce natural active
substances called glycoalkaloids with biological activity.
Glycoalkaloids, like many secondary metabolites, are thought
to function in plant chemical defense, acting as nonspecific
protectors or repellents against potential pest predators (Osman,
1980). Eggplants contain a type of glycoalkaloids like other
nightshade species called α-solanine (Manuchair, 2006), which
is an important active substance with many biological activities
including anti-tumor (Tai, 2002, Ji et al., 2005), lowering blood
pressure and heart stimulation (Yang, 2004). In addition, α-
solanine has been used in the treatment of asthma and epilepsy
(Zhang et al., 2002), and also has shown the effects of antifeedant
(Beier, 1990), fungicide (Allen and Kúc, 1968), and pesticide
(Birch et al., 2002).
Consideration of the above demonstrates the importance of
accurate and reliable analytical methods for α-solanine. Literatures
suggest that α-solanine could be extracted with methanol
(Kobayashi et al., 1989; Saito et al., 1990), methanol-chloroform
(2:l v/v) (Bushway and Ponnampalam, 1981), heptanesulfonic and
acetic acids (Carman et al., 1986), 1.1% acetic acid (Filadelfi and
Zitnak, 1982), and tetrahydrofuran (THF)-water-acetonitrile with
1% acetic acid (Bushway et al., 1980a, 1983, 1986). However,
the common soaking extraction procedure usually requires long
time, high temperature and more energy. As a novel technique for
sample pretreatment, ultrasound-assisted extraction has attracted
more attention in recent years. Compared with other leaching
techniques such as Soxhlet extraction, microwave-assisted
extraction and supercritical fluid extraction, ultrasound-assisted
extraction shows many potential advantages in various aspects,
such as faster extracting rate, matter quality reservation and lower
time and energy cost. It has been used in extracting anthraquinones
from Rheum palmatum L, and cyanuric acid from pet food (Chen
et al., 2008; Wang et al., 2008)
Many analytical methods for α-solanine are reported in literature,
thin layer chromatographic scanning (Ferreira et al., 1993),
various colorimetric methods, gas chromatography (Lawson et
al., 1992), countercurrent chromatography (Fukuhara and Kubo,
1991), and high-pressure liquid chromatography (HPLC) (Everard
et al., 1996; Hitoshi et al., 2005), and enzyme immunoassays
(Plhak and Sporns, 1992). Each method has advantages and
disadvantages. For example, the colorimetry and thin layer
chromatographic scanning are not proper to quantitative analysis
of glycoalkaloids because of low sensitivity and recovery; the
gas chromatography and enzyme immunoassays treatment
has complicated process, bad colour stability, low sensitivity.
Compared to other methods, HPLC is a sensitive, simple, rapid,
and relatively cheap detection method.
The objective of this study was to establish a concise, reproducible
method for the routine extraction and quantification of α-solanine
in eggplants.
Materials and methods
Materials: The seeds of eggplant ‘Liaoqie-7’ (purple eggplant)
were bought from Fuyou Seed Company in Liaoning Province,
China. The seeds were sown in soil and transplanted on 60th days
in the greenhouse in Shenyang Agricultural University. Eggplant
fruit samples were harvested at mature stage. In order to separate
skin from flesh, a knife was used to peel fruits. Then scrape the
inside of skin to remove the left flesh. So the content of α-solanine
in skin didn’t contain the flesh. And the content of α-solanine in
flesh didn’t contain the skin. The skin and flesh were porphyrized
to powder after drying in the air. The rate of dried weight to fresh
weight in samples was 1:10. The powder was then passed through
a 40-mesh screen, and stored in the extractor at 4oC.
The α-solanine stock solution (3 mg mL-1) was prepared through
dissolving α-solanine (Sigma-Aldrich chemical Co., USA) in
methanol. All solutions were prepared with ultrapure water
(Barnstead Diamind, USA). All chemicals were AnalaR grade
except for special notifications.
60 Extraction and determination of α-solanine in eggplant fruits

Extraction procedure: The extraction technique was adopted Results

from Bushway et al. (1980b) with some modifications. 25g of
dried powder was stirred with 250 mL of 70% methanol and 1mg
mL-1 sodium bisulfitein in a mixer (Corning PC-420D, USA) for
15 min. The mixture was treated by different treatments which
were ultrasonic wave extraction, microwave extraction, soxhlet
extraction, oscillation extraction and soaking extraction at 50o for
60 min, and then vacuum filtrated. The filter was collected and
concentrated to approximately 10-15 mL by rotary evaporator
(Eyela N-1000, Japan). The concentrate was then added with
10-15 mL of 0.2 NHCl while being stirred. The extraction was
centrifuged at 8000 rpm at 20o for 15 min (HERMLE Z 36 HK,
German), and the pH of the suspension was adjusted to 10-11
(Horiba F-53, Japan) with concentrated ammonium hydroxide
before the sample was placed into a water bath at 70o for 30min.
The solution was then cooled in an ice-water bath for 30 min and
centrifuged at 10000 rpm at 20o for 30 min. The precipitate was
washed with 5 mL of 1% ammonium hydroxide and centrifuged
at 10000 rpm at 20o for 15 min, and the pellet was collected. The
air-dried pellet was then dissolved in 10 mL of methanol (HPLC
grade, Merck, German) and treated in the ultrasonic treatment
unit at 50o for 30min. The suspension was filtered through a
0.45 μm filter membrane. The filtrate was used for HPLC. In this
experiment we adopted different factors which were three solvent
(70% Methanol, 5% Acetic acid and Methanol: chloroform = 2:1),
three extraction time (40, 50 and 60 min), three temperature (30,
40 and 50oC) and three material to liquid ratio (1:8, 1:9 and 1:10)
to find the best condition of extracting α-solanine from eggplant
fruit with ultrasonic wave by orthogonal testa.
HPLC Procedure. The HPLC procedure was adopted from
Friedman and Dao (1992) with some modifications. Spectra of α-
solanine in the adopted mobile phase showed maxima absorbance
at 202 nm. Therefore, 202 nm was used for quantification of α-
solanine. A Waters 600 controller HPLC system with a 600 pump
and degasser, an automatic injector with a final volume loop of
10 μL, a Nova-pak Cl8 coupled cartridges column (250×4.6mm,
5μm) and a Waters 2487 UV-vis variable-wavelength detector
was used. System management and hardware interface for data
acquisition were performed by the Millennium32 computer
software package from Waters. The mobile phase was acetonitrile
(HPLC grade, Merck, German)-0.05 N potassium dihydrogen
phosphates (55:45 v/v) adjusted to pH 4.5 with 1% phosphoric
acid. This was passed through a 0.45 μm filter and degassed for
20 min under reduced pressure. The HPLC flow rate was 0.7 mL
min-1, and the UV absorbance was measured at 202 nm.
The methanol concentrated extracts were spotted on a TLC plate
along with α-solanine standards. And the skin and flesh extracts
were analyzed. After development, all plates showed only one
spot corresponding to α-solanine (R = 0.35).
Fig. 1 shows linear relationships in the range 0.15-3 mg mL-1
between concentrations of α-solanine (r = 0.995) and the peak
height on HPLC chromatograms. Fig. 2 illustrates the separation
of α-solanine on the HPLC columns in standard solution, extract
and extract added with standard of eggplant fruit, with the
retention time of 2.8 min. These data suggested that minimal
error was related to HPLC.
The highest extracting rates of α-solanine in both skin and flesh
were obtained using ultrasonic wave (Table 1), with virtually
0.107 and 0.626 mg g-1, whereas common extraction gave only
0.0114 and 0.192 mg g-1, respectively. None of the other extraction
means used gave extracting rates as high as those obtained with
the ultrasonic wave extraction.
The effect of extracting α-solanine from eggplant fruit with
ultrasonic wave was dependent on solvent, extraction time,
extraction temperature and material to liquid ratio. For both skin
and flesh, the effect of these four factors on extracting rate was
solvent (A) > material to liquid ratio (D) > time (B) > temperature
(C) (Table 2). The best treatment was A1B3C3D3, with extraction
Table 1. Extracting rates of different extraction means on skin and flesh
of eggplanta
α-solanine extracting rate (mg g-1)
Extraction method
Skin Flesh
Ultrasonic wave extraction 0.107±0.006 a A b 0.626±0.004 a A
Microwave extraction 0.0764±0.002 b B 0.432±0.003 b B
Soxhlet extraction 0.0422±0.0004 c C 0.358±0.006 c C
Oscillation extraction 0.0405±0.001 c C 0.343±0.007 d D
Common Extraction 0.0114±0.001 d D 0.192±0.003 e E
a
Results values are means ± standard errors (n=3)
Different lowercase letters in the same column indicate significant
difference at P=0.05 and different capital letters indicate significant
difference at level of P=0.01
700
600
Peak height response (mV)

Spiking Experiments: A series of spiking experiments were 500

carried out to establish the extent of recovery of added α-solanine
from eggplants. Specifically, 20 g dried eggplant fruit powder was 400
added in different experiments with 1, 2, 3, 4 and 5 mg α-solanine,
respectively. The samples were thoroughly mixed, extracted, and 300
analyzed by HPLC for recovery of the added alkaloids.
TLC Procedure: Preparative thin-layer chromatography (TLC) 200
was performed on silica gel precoated plates, 0.25 mm×20cm×20
cm (Merck, Darmstadt, Germany). A 25-μL methanol concentrated 100
extract was spotted on the plate along with α-solanine standards.
Then the plate was developed with a bottom layer of benzene- 0
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3 3.3
methanol-1 % ammonium hydroxide (10:2 v/v). After that, the -1
dried plate was sprayed with 1% bismuth potassium iodide with Concentration ( mg mL )
visible violet red spot. Fig. 1. Standard α-solanine peak height response with HPLC
 Extraction and determination of α-solanine in eggplant fruits 61

1 1 1
400
600 a 300 b c

500
300
Peak height (mV)

400 200

mV mV 200
300

200 100
100

100

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Minutes
Fig. 2. Chromatograms of α-solanine standard (a), α-solanine in eggplant fruit (b) and α-solanine in mixture of eggplant fruit extracting
solution added with α-solanine standard (c). Peak: 1, α-solanine
rates 0.107 and 0.626 mg g-1, extracted at 50o for 60 minutes in
70% methanol with the material to liquid 1:10.
Spiking experiments (Table 3) revealed that the modified method
could recover a 96-99% of added α-solanine. In the range of 1- 5
mg of α- solanine / 20 g dried eggplant fruit powder (flesh and
skin), the extent of recovery increased with the amount added
before extraction. The average recovery was 97.97%.
Discussion
Since eggplant fruits contain approximately 93-94% water, the
air-dried samples provide more than 10 times concentration of
α-solanine. Air-dried samples can be easily grinded and stored for
longer periods of time prior to measurements. Considering these
advantages, air-dried fruits were used in this study.
The recovery of added and native α-solanine in skin and flesh of
eggplant fruts increased by approximately 2% in the condition of
Table 2. Results of orthogonal testa of α-solanine extracted from eggplant skin and flesh with ultrasonic waveb
Treatment A B
Solvent Time
(min)
1 A1 70% Methanol B1 (40)
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4. 0 4.5 5.0 5.5 6.0 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Minutes Minutes
Retention time (min)
without washing the precipitate with 1% ammonium hydroxide.
Washing aimed to remove most brown pigments in the final
extraction product when eggplants flesh was used as starting
materials. The brown pigment does not seem to interfere in
the analysis except for turning into dark. To minimize possible
interference of pigment in skins, the purple precipitate formed
during evaporation of the extracts was first removed by
centrifugation and filtration. The partially evaporated clear filtrate
was then analyzed for alkaloid content.
Separation and retention of the α-solanine increased by raising
the buffer pH. However, this resulted in reduced sensitivity of
the system due to low solubility of α-solanine. No alkaloid was
detected above pH 7 because of precipitation upon injection onto
the column. The optimum pH was found to be 4-5. Reducing
the amount of acetonitrile in the mobile phase also improved
separation. Using 55% acetonitrile, 45% buffer gave maximum
separation. α-solanine was insoluble at low acetonitrile ratios.
C D Extracting rate in Extracting rate in
Temperature Material to liquid skins (mg g-1) flesh (mg g-1)
(°C) (g:mL)
C1 (30) D1 (1:8) 0.0867±0.004 0.524±0.002

2 A1 B2 (50) C2 (40) D2 (1:9) 0.0958±0.003 0.554±0.003

3 A1 B3 (60) C3 (50) D3 (1:10) 0.1070±0.006 0.626±0.004

4 A2 5% Acetic acid B1 C2 D3 0.0776±0.003 0.386±0.006

5 A2 B2 C3 D1 0.0535±0.0003 0.147±0.004

6 A2 B3 C1 D2 0.0547±0.0003 0.158±0.005
A3 Methanol:chloroform
7 B1 C3 D2 0.0395±0.001 0.139±0.007
2:1
8 A3 B2 C1 D3 0.0458±0.001 0.151±0.006

9 A3 B3 C2 D1 0.0220±0.001 0.0635±0.004

Rα 0.4542 0.0708 0.0528 0.1390 — —

Rα 0.0608 0.0067 0.0044 0.0228 — —

a
L9(3)4 design was adopted in this orthogonal test
b
Results are the mean extracting rate of standard ± standard errors, n=3
62 Extraction and determination of α-solanine in eggplant fruits
Table 3. Results of recoveries of α-solanine added to dried samplea
Sample (mg) Added (mg) Detected (mg) Recoveryb (%)
15.09±0.05 1 15.90±0.05 98.82±1.1
14.27±0.05 2 16.18±0.05 99.45±0.9
14.25±0.05 3 16.70±0.05 96.81±1.3
14.65±0.05 4 18.24±0.05 97.80±0.8
14.89±0.05 5 19.29±0.05 96.98±1.2
Mean±SD 97.97±1.1
a
20 g powder was added the known amount of α-solanine. The samples
were mixed, extracted, and analyzed with HPLC.
b
Results are the mean recoveries of standard ± standard errors, n=3
The composition of the mobile phase used for HPLC was
important in ensuring full separation of α-solanine. A number of
published methods were conducted with Tris-HCl at near-neutral
pH (Jonker et al., 1992). However, under such conditions the
HPLC stability was not good and the linear relationships did not
perform well.
Many of extraction solvents were utilized in the published
methods and most were based on some solvents or salts in a
weak solution of acetic acid. Heptanesulfonic acid and aqueous
acetic acid (Everard et al., 1996) or 5% acetic acid (Hitoshi et
al., 2005) were commonly added. However, as α-solanine is
probably stored within the aqueous phase of the eggplant cell
and is readily soluble in solvents like methanol with the addition
of water through some special extraction means like ultrasonic
wave extraction, the use of acid is unnecessary. In addition, the
solution of acetic acid under ultrasonic wave condition may lead
to hydrolyzation of α-solanine resulting in a reduced recovery. In
this experiment, the extracting rate of 70% methanol as solvent
was higher than that of 5% acetic acid. Sodium bisulfite was used
to reduce oxidation of the extract (Hellenas, 1986), and also used
in these investigations.
Ultrasound is an efficient non-thermal alternative (Kim und
Zayas, 1989; Yan et al., 2002; Knorr et al., 2002). Ultrasonic
cavitation creates shear forces that break cell walls mechanically
and increase material transfer. In this experiment, ultrasonic wave
extraction with methanol as extraction solvent gave the highest
extracting rate and optimal recovery.
Some researches showed that during ultrasonic extraction, the
relationships between extraction time and extracting rate are
as follows (Guo, 1995; Guo, 1997): (1) The extracting rate
increased as time prolongs; (2) The extracting rate increasment
slowed as time prolonging to certain extent; (3) The extracting
rate decreased as time prolonging to critical stage. The reason for
extracting rate decrease is probably due to the fact that prolonged
time may decompose the active components and increase their
impurity, leading to low extracting rate (Duan and Feng, 1992).
In a preparatory experiment, the extracting rate increased rapidly
within 60 minutes extraction, slowly in 60-90 minutes extraction,
and decreased after 90 minutes.
The above method aim at a number of problems with HPLC
analysis of α-solanine, and utilizes a new extracting way to
improve sample recoveries and got good results. It can be used
in eggplant breeding, production and medical care research.
Laboratory also can use this method to investigate the response
of eggplant alkaloid concentrations to environmental conditions
during cultivation and some biological activity of eggplant.
References
Allen, E.H. and J. Kúc, 1968. α-solanine and α-chaconine as fungitoxic
compounds in extracts of Irish potato tubers. J. Phytopathol., 58:
776-781.
Beier, R.C. 1990. Natural pesticides and bioactive components in foods.
Rev. Environ. Contam. Toxicol., 113: 47-137.
Birch, N.E., I.E. Geoghegan, D.W. Griffiths and J.W. Mcnicol, 2002.
The effect of genetic transformations for pest resistance on foliar
solanidine-based glycoalkaloids of potato (Solatium tuberosuni). J.
Annals Appl. Biol., 14: 143-149.
Bushway, R.J. and R. Ponnampalam, 1981. α-chaconine and α-solanine
content of potato products and their stability during several modes
of cooking. J. Agr. Food Chem., 29: 814-817.
Bushway, R.J., E.S. Barden, A.W. Bushway and A.A. Bushway, 1980a.
The mass extraction of potato glycoalkaloids from blossoms. Am.
Potato J., 157: 175-180.
Bushway, R.J., E.S. Barden, A.M. Wilson and A.A. Bushway, 1980b.
Analysis of potato glycoalkaloids by high-performance liquid
chromatography. J. Food Sci., 45: 1088-1089.
Bushway, R.J., J.L. Bureau and J. King, 1986. Modification of the
rapid high-performance liquid chromatographic method for the
determination of potato glycoalkaloids. J. Agr. Food Chem., 34:
277-279.
Bushway, R.J., J.L. Bureau and D.F. McGann, 1983. α-Chaconine and
α-solanine content of potato peels and potato peel products. J. Food
Sci., 48: 84-86.
Carman, A.S., S.S. Kuan, G.M. Ware, O.J.J. Francis and G.P.
Kirschenheuter, 1986. Rapid HPLC determination of the potato
glycoalkaloids α-solanine and α-chaconine. J. Agr. Food Chem.,
34: 279-282.
Chen Y., L. Zhua, J. Xiaoa, H. Tanga, G. Guob, Q. Zengb and X.Wang,
2009. Ultrasonic extraction and determination of cyanuric acid in
pet food. Fd. Control, 20(3): 205-208.
Duan, G.M. and C.Q. Feng, 1992. Glycoalkaloids in potatoes. J. Plant
Physiol. Commun., 28: 457-461.
Everard, J., Edwards and H.Cobb. Andrew, 1996. Improved high-
performance liquid chromatographic method for the analysis of
potato (Solanum tuberosum) glycoalkaloids. J. Agr. Food Chem.,
44: 2705-2709.
Ferreira, F., P. Moyna, S. Soule and A. Vazquez, 1993. Rapid determination
of solanum glycoalkaloids by thin-layer chromatographic scanning.
J. Chromatogr., 653(2): 380-384.
Filadelfi M.A. and A. Zitnak, 1982. Preparation of chaconines by
enzymic hydrolysis of potato berry alkaloids. J. Phytochemistry,
21: 250-251.
Friedman, M. and L. Dao, 1992. Distribution of glycoalkaloids in potato
plants and commercial potato products. J. Agric. Food Chem., 40:
419-423.
Fukuhara, K. and I. Kubo, 1991. Isolation of steroidal glycoalkaloids
from Solanum incanum by two countercurrent chromatographic
methods. J. Phytochemistry, 30: 685-687.
Guo, X.W. 1995. Effect on extraction yield of berberine by ultrasound-
assisted extraction. J. Chin. Materia, 20: 673-675.
Guo, X.W., 1997. Comparison of effect on rutin component by
ultrasound-assisted and hot alkali extraction. J. Chin. Traditional
Herbal Drugs, 28: 88-89.
Hellenas, K. 1986. A simplified procedure for the quantification of potato
glycoalkaloids in tuber extracts by HPLC, comparison with ELISA
and a colorimetric method. J. Sci. Food Agr., 37: 776-782.
Hitoshi, K., S. Keiitsu, N. Nobumitsu, Y. Shigeo and T. Youichi, 2005.
Simple and sensitive method for determination of glycoalkaloids
in potato tubers by high-performance liquid chromatography with
chemiluminescence detection. J. Chromatogr. A, 1100(1): 26-31.
 Extraction and determination of α-solanine in eggplant fruits
Ji, Y.B., H.L. Wang and S.Y. Gao, 2005. Effect of solanine on DNA
and RNA in tumor cell of tumor-bearing mice. J. Chin. Traditional
Herbal Drugs, 36: 1200.
Jonker, H.H., A.J. Koops and J.C. Hoogendoorn, 1992. A rapid method
for the quantification of steroidal glycoalkaloids by reversed phase
HPLC. J. Potato Res., 35: 451-455.
Kim, S.M. and J.F. Zayas, 1989. Processing parameter of chymosin
extraction by ultrasound. J. Food Sci., 54: 700.
Knorr, D., B.I.O. Ade-Omowaye and V. Heinz, 2002. Nutritional
improvement of plant foods by non-thermal processing. J.
Proceedings Nutr. Soc., 61: 311-318.
Kobayashi, K., A.D. Powell, M. Toyoda and Y. Saito, 1989. HPLC
method for the simultaneous analysis of α-solanine and α-chaconine
in potato plants cultured in vitro. J. Chromatogr., 462: 357-364.
Kvasnicka, F., K.R. Price, K. Ng and G.R. Fenwick, 1994. Determination
of potato glycoalkaloids using isotachophoresis and comparison
with a HPLC method. J. Liquid Chromatogr. Related Technol., 17:
1941-1951.
Lawson, D.R., W.A. Erb and A.R. Millar, 1992. Analysis of Solanum
alkaloids using internal standardization and capillary gas
chromatography. J. Agric. Food Chem., 40: 2186-2191.
63
Manuchair, S.E., 2006. Pharmacodynamic basis of herbal medicine,
US: CRC Press, 309.
Osman, S.F., 1980. Glycoalkaloids of the Solanaceae. Recent Adv.
Phytochem., 14: 75-96.
Plhak, L.C. and P. Sporns, 1992. Enzyme immunoassay for potato
glycoalkaloids. J. Agric. Food Chem., 40: 2533-2540.
Saito, K., M. Horie, Y. Hoshino, N. Nose and H. Nakazawa, 1990. HPLC
determination of glycoalkaloids in potato products. J. Chromatogr.,
508: 141-147.
Tai, J.R. 2002. The effect of cancer prevention and resistance of main
vegetables. J. Southwest Hort., 30: 28-29.
Wang, L., D. Li, C. Bao, Z. Wang, Y. Shi and H. Zhang, 2008. Ultrasonic
extraction and separation of anthraquinones from Rheum palmatum
L. Ultrason. Sonochem., 15: 738-746.
Yan, W., S.F. Li and S.J. Tian, 2002. Ultrasound-assisted extraction
technology. J. Chem. Ind. Eng. Progress, 21: 649-651.
Yang, X.W. 2004. Alkaloids. Beijing, Chemical Industry Press, 393.
Zhang, C.Y., S. Li, X. Chen and C.M. Wang, 2002. Studies on the anti-
asthma action mechanism of nightshade alkaloids. J. Theory and
Practice of Chinese Medicine, 2002: 488-489.
 Journal
Journal of Applied Horticulture, 11(1): 64-67, January-June, 2009
Appl

Morphogenetic and biosynthetic potential of in vitro grown

Hypericum perforatum under stress and normal conditions

M. Altaf Wani1, G.R. Lawania2, R.A. Bhat3, Iffiat Fayaz1, A. Nanda3 and Gazenfar Gani3
Division of Plant Breeding and Genetics, SK University of Agricultural Sciences and Technology of Kashmir, Srinagar, India.
1

Division of Plant Breeding and Genetics, Allahabad Agricultural Institute-Deemed University, Allahabad, India.3Division
2

of Floriculture, SK University of Agricultural Sciences and Technology of Kashmir, Srinagar, India.

Abstract
Three different strains of Hypericum perforatum viz. HP-1, HP-2, and HP-3 were subjected to different levels of saline stress (0.25,
0.5, 0.75 and 1.0% with NaCl) and high pH regime (8.5, 9.0, 9.5 and 10.0 with NaOH). Gradual loss in callus growth was observed in
all the three strains in response to both kinds of stress. However, high pH showed more drastic effect than saline stress. All the three
strains showed higher content of pseudohypercin than hypercin. Change in hypercin production was negligible, however remarkable
change was observed in pseudohypercin production in response to both kinds of stress. HP-2 strain produced higher content of
hypercin than HP-1 and HP-3 strains under normal as well as under stressfull regime. Proteins were affected qualitatively as well as
quantitatively. Maximum numbers of proteins were isolated from control cultures at the retention time of five minutes. Among the three
strains maximum numbers of proteins were isolated from HP-3 strain. High pH reduced number of proteins to 12 and 3 while salinity
increased number of proteins to 42 and 52 in HP-1 and HP-2, respectively due to accumulation of low molecular weight proteins in
response to saline stress.
Key words: Hypericum perforatum, hypercin, pseudohypercin, hyperforin, HPLC, saline stress
Abbreviations: PMSF- Phenylmethylsulphonylfluoride, LC/MS-Liquid chromatography/Mass spectrometry

Introduction Materials and methods

Among the medicinal plants which recieved great attention
during the recent years from pharmaceutical industry at global
level, Hypericum perforatum ranks very high due to its secondary
metabolites (hypercin and pseudohypercin) which are found to be
anti depressant in action (Wentworth and Agostini, 2000). Besides,
anti depressant activity H. perforatum is gaining popularity due to
its antiviral (Lopez and Hudson, 1991), anti retroviral (Lavie et
al., 1989) and antibiotic action (Fritz and Schleicher, 1993).
Although vast information is available on cell cultures of H.
perforatum, however, investigations are required to understand
the biosynthetic pathway of hypercin, pseudohypercin and
hyperforin production, enhancement of production by media
manipulations and induction of gene/ protein alterations by
environmental stress (Ramgopal and Carr, 1991; Sairam and
Aruna, 2004). Abiotic stress like salinity, pH changes and drought
conditions highly influence biomass production, accumulation
of secondary metabolites and yield of the most plants. The fast
expanding research in plant molecular biology has given several
clues in understanding how plant responds under stressful regime
(Cushman et al., 1990). Great deal of success has been achieved
in unveiling gene / protein alterations associated with preparation
of plant against the abiotic stress.
Owing to the medicinal value of H. perforatum an experiment was
carried out at IIIM (CSIR), Jammu to find out the morphogenetic
and biosynthetic potential of H. perforatum under saline and
alkaline stress and to unveil the alternations in protein profiles
in response to stress.
The calli resulted from shoot tip cultures were transferred to
stress induced (saline and alkaline) solid MS culture media and
cultivated on a constant temperature of 25+2 oC with 16 hour
photoperiod. High pH stress was induced by adjusting pH of
culture media to 8.5 (T-1), 9.0 (T-2), 9.5 (T-3) and 10.0 (T-4) with
NaOH and saline stress was induced by incorporating different
concentrations of NaCl2 0.25% (T-6), 0.5% (T-7), 0.75% (T-8)
and 1.0% (T-9). The calli developed on this high pH and saline
stress media were analyzed for presence and quantification of
hypercin, pseudohypercin and hyperforin.
Morphological analysis of callus cultures: Growth was
measured in terms of fresh callus weight. All the replicates of all
the treatments were harvested at the same time. All cultures were
also visually screened for red colour formation and for friability
at different stages of growth to assess pigment formation.
Biochemical analysis of callus cultures
Sample extraction for biochemical analysis: Air-dried callus of
H. perforatum was extracted with chloroform and then filtered.
The powder from filter paper was dissolved in acetone and
concentrated on rotawave to dryness. This dried material was then
dissolved in methanol and analyzed by HPLC after filtration.
HPLC analysis: Presence and quantification of hypercins and
pseudohypercins was determined by an HPLC using Gilson
system (Gilson Medical Electronics, France) equipped with 321-
model pump, operated at room temperature (25+2oC). Separations
were performed on Lichosphere RPC-18 (5 μm, 4.6 x 150 mm)
and detection was achieved with a UV detector set at 589 nm.
 Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress 65

The chromatographic data was recorded and processed on uni- Table 1. Effect of saline and high pH on fresh callus weight/flask (Mean
point system software. Isocratic mobile phase was used with value ± Standard error)
methanol, ethyl acetate and water (pH 2.5) in the ratio of 67:16:17 Treatment HP-1 (g) HP-2 (g) HP-3 (g)
respectively. pH of water was adjusted to 2.5 with O-phosphoric T1 3.68±0.30 3.89±0.60 3.19 ±0.08
acid. The flow rate was kept constant at 1 mL min-1 and injection T2 1.55±0.18 1.88±0.04 1.03±0.25
T3 - - -
volume 25 μL using six point calibration curve generated with
authentic hypercin and pseudohypercin. T4 - - -
T5 5.62±0.22 5.49±0.27 5.27±0.33
Sample extraction for protein analysis by LC/MS: Fresh callus T6 4.45±0.24 3.87±0.32 3.61±0.33
was homogenized with 2-3 volume of 0.1M buffer carrying T7 3.89±0.13 2.69±0.28 3.24±0.30
5mM EDTA. A pinch of PMSF was added during grinding. The T8 3.15±0.35 3.15±0.15 3.77±0.46
homogenized sample was centrifuged at 4oC @ 10, 000 rpm for T9 - - -
20 minutes. The protein from supernatant was precipitated with
5 volumes of chilled (-20oC) acetone and all the samples were production of hypercin and pseudohypercin under normal (devoid
kept at -20oC for 1-2 hours. The samples were again centrifuged of stress) as well as under stress conditions (saline and high
at 4oC @ 10000 rpm for 10 minutes to separate the precipitated pH). In the present study, HP-2 strain produced higher content
proteins from the solution. The acetone was evaporated from of hypercin (0.272) than HP-1 (0.188) and HP-3 (0.196) under
the sample and the precipitate was dissolved in 500μL of HPLC control conditions. There was no remarkable effect of stress on
grade water. The protein solution was then filtered through 0.2 accumulation of hypercin; however pseudohypercin production
mesh micro filter. was highly influenced by both kinds of stress indicating that
abiotic stresses has differential effect on biosynthesis of different
LC/MS analysis: The extracted samples were analyzed before metabolites. (Table 1).
LC/MS using AGLLENT 1100 series HPLC and squin 300
(Bruker/Mass spectrometer) connected by ESI (Electro Spray Inability of undifferentiated cultures to produce flavonoids
Ionization) interface using DAD (Diode Array Detector) at under both normal and stress conditions: HPLC analysis of all
wavelength of 280.8 nm. The mobile phase was used with 0.05% the three strains revealed no signs of hyperforin production from
TFA in acetonitrile and 0.05% TFA in water. RP-8 column with the undifferentiated cultures where as field grown plants showed
diameter of 5μm and length of 4x250 mm and flow rate of 0.5 mL presence of hyperforin particularly at the flowering stage.
min-1 was used. The temperature of column was kept at 300C.
Effect of stress on protein profile: Nine selected samples
Results analyzed by LC/MS showed proteins of varied molecular weights
(1,000-12,000 Daltons). Maximum numbers of protein were
Response of callus to high pH and NaCl stress: Calli of all three isolated from saline cultures (NaCl stress) within the retention
strains investigated showed similar response in terms of callus time of 5 minutes. Among the three strains, maximum number
weight however gradual decrease in callus growth was observed of protein was isolated from HP-3 strain. High pH and salinity
in stress cultures (saline and high pH) as compared to control showed significant effect on the number of proteins. High pH
cultures (devoid of stress). Among the two stresses high pH stress reduced the number of protein to 12 and 3 while salinity increased
showed more drastic effect than saline stress. No callus growth was the number of protein to 42 and 52 in HP-1 and HP-2 strain,
observed at pH 9.5 (T3) and pH 10 (T4). Further callus showed slight respectively. The qualitative change in proteins was also indicated
increase in compactness towards the higher concentration of NaCl by change in polarity and difference in retention time. Saline stress
and decrease in compactness towards the increase in pH. in present study altered the protein profile by formation of proteins
The visual examination of callus revealed dark brown to reddish at different retention times and accumulation of low molecular
brown callus and the intensity of red clour formation showed a weight protein in response to saline stress. (Table 3).
slight increase with time. Among the three strains HP-2 strain
showed more tendency towards red colour formation. No Discussion
remarkable effect was shown by high pH on colour formation but
Our results showed that the calli of all three strains used in this study
NaCl showed slight increase in red colour formation. (Table 1)
had potential for accumulating hypercins and pseudohypercins
Accumulation of hypercin and pseudo hypercin in response under normal (devoid of stress) as well as under stress conditions
to stress: Calli of all the three strains investigated showed (saline and high pH). Mosen et al. (1993) reported the formation of
Table 2. Effect of saline and high pH stress on total hypercin and Pseudohypercin content
Treatment HP-1 (%) HP-2 (%) HP-3 (%)
Hypercin Pseudohypercin Hypercin Pseudohypercin Hypercin Pseudohypercin
T1 0.180 0.213 0.176 0.278 0.180 0.310
T2 0.172 0.229 0.184 0.246 0.176 0.164
T3 - - - - - -
T4 - - - - - -
T5 0.188 0.574 0.272 0.590 0.196 0.573
T6 0.178 0.413 0.184 0.451 0.176 0.459
T7 0.176 0.203 0.178 0.369 0.192 0.328
T8 0.184 0.178 0.176 0.196 0.180 0.164
T9 - - - - -
66 Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress

Table 3: Effect of saline and high pH stress on protein profile

Treatment Strain Retention Molecular weight of proteins (Daltons x 1000)
time 12 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11
HP-l 5 min - - 8 - 1 5 1 1 4 2
T5 25min - - - - - - - - - -
60min - - - - - - - - - -
HP-2 5 min 2 1 - 2 - 1 4 4 2 6
25min - - - - - - - - - -
60min - - - - - - I - - -
HP-3 5 min 1 2 2 4 9 6 9 6 8 5
25min - 1 - - - - - - 1 ‘. -
60min - - - - - - - - - -
T2 HP-l 5 min 2 - 1 .- - 1 I 3 3 -
25min - - - - - - - - 1 -
60min - - - I- - - - - - -
HP-2 5 min - 1 - - 1 1 - - - -
25min - - - - - - - - - -
60min - - - - - - - - - -
HP-3 5 min - 2 - 3 3 1 1 4 2 2
25min - - - - - - - - - -
60min - - - - - - - - - -
T8 HP-I 5 min - - - - - 3 - - - -
25min - - - - - 1 - - - -
60min 3 1 1 15 3 - 8 5 6 4
HP-2 5 min - - - - - - - - - -
25min 8 - - 1 2 - 7 11 15 5
60min - - - - - 6 - - - -
HP-3 5 min - - - 1 1 - - - - -
25min - - - - - 6 - - - -
60min - - - - - - - - - -

hypercin and other metabolites in undifferentiated cultures of H.
perforatum. The calli of HP-2 strain accumulated highest content
of hypercin than HP-1 and HP-3 strain under control conditions.
Cambell and May (1992 ) also confirm the difference in hypercin
content of different strains of H. perforatum.
Karting et al. (1996) reported that field grown plants of H.
perforatum produce hyperforin particularly at the flowering stage
but the production of hyperforin from undifferentiated cultures
of H. perforatum has not been documented so for. Similar results
were revealed during the present investigations thus confirms
the inability of undifferentiated cultures of H. perforatum to
biosynthesize hyperforin. This inability of undifferentiated
cultures to synthesize particular metabolite is due to loss of gene
or mutation of gene involved in particular steps of biosynthetic
pathway of the metabolite or unavailability of substrate and
enzymes or lack of specialized cells involved in metabolite
formation (Bohm, 1982).
Bohnert et al. (1995) reported that biosynthetic pathways are altered
by abiotic stresses. During the present investigations saline and
high pH stress showed negligible effect on hypercin production,
however remarkable change was observed in pseudohypercin
production (Table 2). Thus the findings are in agreement with
the previous findings of Cushman et al. (1990) who revealed that
salt and dehydration stress shows biochemical effects along with
morphological, physiological and molecular effect.
LC/MS analysis revealed that all the three strains varied from one
another in their protein content as was evident from the variation
in the number as well as in molecular weight of proteins. Changes
in gene expression at transcriptional and post-transcriptional level
are initially demonstrated by analysis of protein profiles elicited in
plant using salt treatment. These studies revealed both qualitative
and quantitative changes in pattern of polypeptides synthesized
following salt treatment (Ericson et al., 1984; Chen et al., 1991).
In the present study, high pH reduced the number of proteins
perhaps due to creation of free amino acid pool due to inhibition
of peptide bond formation; while salinity increased the number
of proteins due to accumulation of large number of low molecular
weight proteins. This coincides with Lopez et al. (1994) who
reported accumulation of protein in leaves of Raphanus sativus in
response to salt stress or water deficiency. The qualitative change
in proteins was also indicated by change in polarity and difference
in retention time. Saline stress in present study altered the protein
profile by formation of proteins at different retention times. The
change in protein profile in response to stress is also reported by
Ramgopal et al. (1991) and Sairam et al. (2004).
We conclude from our results that among the three strains used
in this study HP-2 strain showed best potential for accumulating
hypercins and pseudohypercins (antidepressant in function).
Hypercin production did not show any change in response to
stress, however pseudohypercin production was highly influenced
by both kinds of stress. High pH stress showed more drastic
effect on both morphogenetic as well as biosynthetic potential of
the plant as compared to saline stress. Change in protein profile
was observed in response to both kinds of stress however this
change could not be correlated with the change in accumulation
of secondary metabolites due to stress.
 Morphogenetic and biosynthetic potential of in vitro grown Hypericum perforatum under stress
References
Bohm, H.1982. The inability of plant cell cultures to produce secondary
metabolites. In: Plant tissue culture, Fujiwara, A. (Ed.) Japanese
Assoc. for Plant tissue culture, Tokyo. 325-326.
Bohnert, H.J., D.W. Nebson and R.G. Jonsen, 1995. Adaptation to
environmental stress. Plant Cell, 7: 1099-1111.
Cambell, M.H. and C.E. May, 1992. Variation and varietal determination
in H. perforatum. Plant Protection Quarterly, 2: 43-45.
Chen, R.D. and Z. Tabaeizadeh, 1991. Alteration of gene expression in
tomato plants by drought and salt stress. Genome, 35: 385-391.
Cushman, J.C., E.J. De Rocher and H.J. Bohnert, 1990. Gene expression
during adaptation to salt stress. Environmental inquiry of plants. Ed.
F. Kalterman, Academic Press, Bandigo; 177-203.
Ericson, M.C. and S.H. Alfinito, 1984. Protein produced during salt stress
in tobacco cell culture. Plant Physiology, 74: 506-509.
Fritz, Z. and A. Schleicher, 1993. Effect of substituting milfoil, St. Johns
Wort and lovage for antibiotics on chicken performance and meat
quality. J. Anim. Free Sci., 2(4): 189-195.
Karting, T., I. Goebel and B. Heydel, 1996. Production of hypercin and
pseudohypercin and flavanoids in cell cultures of various Hypericum
species and their chemotypes. Planta Medica, 62 (1): 51-53.
67
Lavie, G., F. Valentine, B. Levin, Y. Mazur, G. Gallo and D. Weiner, 1989.
Studies of microorganism of action of antiretroviral agents hypercin
and pseudohypercin. Proc.Natl. Acad. Sci, 86: 5963-5967.
Lopez, B.I. and J.B. Hudson, 1991. Antiviral activity of the photoactive
plant pigment hypercin. Photochem and Photobio., 54 (1): 95-98.
Lopez, F., G. Yansuyt, P. Fourcroy and F. Cassedelbart, 1994.
Accumulation of a 22 Kda protein and its mRNA in the leaves of
Raphanus sativus in response to salt stress or water deficit. Plant
Physiol., 91: 605-614.
Mosen, B.V., K. Zdunik, H.J. Woerdenbag, N. Pras and A.W. Alfermann,
1993. Production of natural products by shoots in cultures cultivated
SW in vitro. J. Pharma. World, 15: 6.
Ramgopal, S. and J.B. Carr, 1991. Sugarcane proteins and messenger
RNA regulated by salt in suspension cells. Plant Cell Environ., 14:
46-47.
Sairam R.K. and T. Aruna, 2004. Physiology and molecular biology of
salinity stress in plants. Curr. Sci., 86: 407-422.
Wentworth J.M. and M. Agostini, 2000. St. John’s Wort a herbal
antidepressant activates the steroid X receptor. J. Endocrino.,166(3):
R11-R16.
 Journal
Journal of Applied Horticulture, 11(1): 68-72, January-June, 2009
Appl

Growth interference of invasive Russian knapweed on

“valcatorce INTA” onion

Carlos R. Bezic, Armando A. Dall Armellina, Omar A. Gajardo, Lucrecia M. Avilés and
Silvia L. Cañón
Weed Ecology and Control Research Group, CURZA-University of Comahue (8500) Viedma, Río Negro province,
Argentina. E-mail: malezas@uncoma.edu.ar

Abstract
Russian knapweed is an invasive creeping perennial herb which affects crops by competition and allelopathy. Herbicides available for
use in onion are not able to control Russian knapweed in a crop context. Conversely, recommended products for Russian knapweed
are not selective for the crop. The aims of this work were to study Russian knapweed biomass production and propagation for a
range of increasing densities in an experimental onion culture and to characterize the productive response of onion plants under these
conditions. A partial additive experiment was carried out to study Russian knapweed interference (variable density, 0-64 ramet m-2)
on onion transplants (constant density, 40 pl m-2) under greenhouse conditions in Viedma, Argentina (40º 03’ S; 62º 48’ O). Although
no differences among treatments were found for weed final aboveground biomass, low density treatments (0, 2 ramet m-2) were lower
than 64 ramet m-2 for belowground biomass. Final weed density was proportional to initial conditions. For onion, total (-54%) and
commercial bulb yield (- 56 %) were reduced by weed competition with ≥ 32 ramet m-2. While size 3 bulbs (50-70 mm eq. diam.) were
less represented at weed densities higher than 16 ramets m-2, size 4 ones (70-90 mm eq. diam.) were not present in this condition. For
A. repens, traits such as the rate of vegetative propagation, high competitive ability, mainly belowground, and high propagule pressure
support its high invasive potential.
Key words: Acroptilon repens, Allium cepa, plant competition, partial additive experiment, plant invasion, irrigated agriculture.
which elongates in September, the time at which onion must be
Introduction
transplanted. The first flowers appear in November (Bezic et al.,
As a consequence of weed interference, vegetable yield and 2005). Clonal populations of Russian knapweed may be 30-100
quality are subjected to severe losses in most irrigated areas ramets m-2 in density and 40-60 cm in height, dense enough to be
in Argentina. Instead of rational management programs, weed a strong competitor for any vegetable crop (Kearney et al., 1960;
control is normally done by herbicide application in a clearly Whitson, 1987; Panter, 1991).
reactive approach.
Onion (Allium cepa L.) is the 3rd place vegetable crop in the
There are several irrigated valleys in southern Argentina whose LVRN with 500-1500 hectares annually sowed (Pozzo Ardizzi
areas are mostly dedicated to fruit and vegetable production for et al., 2005). The species is considered a poor competitor against
external markets. The Lower Valley of Rio Negro (LVRN), in weeds because of its slow initial growth associated to few narrow
particular, is located at the southeast of the Rio Negro province, and erect foliage leaves. If remain uncontrolled, weeds can
near the Negro river outlet, at 40º S. It comprises 18,500 hectares reduce onion yield by 36-96 % (Khan et al., 2003; Williams et
of irrigated lands with a high value irrigation infrastructure of al., 2005).
both concrete lined and unlined channels and also an efficient Herbicides available for use in onion are not able to control
drainage system. There are almost 500 properties among 30-120 Russian knapweed in a crop context. Conversely, recommended
hectares each. products for Russian knapweed are not selective for the crop.
The invasive weed, Russian knapweed (Acroptilon repens L.), is The aims of this work were to study Russian knapweed biomass
increasing in incidence at the LVRN. While 2,000 hectares were production and propagation for a range of increasing densities in
directly and indirectly affected by 1980 (Dall Armellina and an experimental onion culture and to characterize the productive
Iglesias, 1984) almost 50% of the properties reported patches response of onion plants under these conditions.
of different size (10 m2-100 hectares) in 2005. The affected area
was actually estimated as 6,000 hectares approximately (Bezic Materials and methods
et al., 2005).
Site and experimental methods: The experiment was carried
Russian knapweed is a creeping perennial herb which affects out in a plastic covered 120 m2 greenhouse, located at the
crops by competition and allelopathy (Watson, 1980; Whitson, Universidad Nacional del Comahue (CURZA) campus in
1987; Dall Armellina and Zimdahl, 1988; Gajardo et al., 2004). Viedma, Río Negro province, Argentina (40º 03’ S; 62º 48’ O).
It reproduces vegetatively by sprouting from roots. Each new All lateral windows remained open over the entire experimental
ramet forms an early season rosette by the end of August period (26 October, 2004-20 March, 2005) in order to avoid
 Growth and interference of invasive Russian knapweed on onion 69

excessive heating during summer. Twenty one 0.49 m2 (0.70 m normality, percentages were transformed by the usual arcsine
by 0.70 m) wooden boxes (35 cm deep), internally lined with a transformation p’ = arcsin(SQRT(p)), where p is the proportion.
200 microns black polyethylene sheet were used. Each box was Data were subjected to ANOVA and the multiple comparison test
filled with sandy-loam soil (OM 3 %; SAR 1.9) free of Russian SNK was applied. The LSD test was also employed in specific
knapweed propagule. The soil was nutrient enriched with 0.1 kg cases.
m-3 of 15-15-15 (N:P:K) at the beginning of the experiment. All
boxes were separated 20 cm from the soil line by placing them Results
on wooden blocks and the bottom of each one was perforated for
drainage purposes. Weed growth and propagation: As season was in progress new
ramets were formed by sprouting from roots. This increased the
Plant material: Several clonal plants of Russian knapweed were box weed density for every treatment. The statistical analysis
obtained by vegetative propagation of sprouting root pieces which revealed that AR2 and AR4 were lower in final weed density in
were selected for the presence of one visible bud preformed in comparison with AR32 and AR64. Intermediate treatments were
the original habitat. This plant material was collected in winter intermediate in response and no differences were detected in
2004 from a unique population colonizing an irrigated field. relation to the previous reports on this aspects(Table 1).
They were maintained under refrigeration at 5 ºC in plastic bags
until September 15, 04 when they were individually placed in Vegetative propagation rate (VPR, number of final ramets /
368 cm3 plastic pots (64 cm2) to induce sprouting. Additionally, number of initial ramets) was decreasing with initial density
late onion cv. Valcatorce INTA transplants were produced in an increase (Fig. 1). The highest VPR was calculated as 12.3 for
experimental nursery until the two true leaves stage when they AR2 and the lowest as 2.6 for AR64.
were transplanted in a regular 10 by 13 cm grid over the entire
No differences among weed densities were observed for
box surface (20 transplants per box; 408,000 plant ha-1) on 26
aboveground weed biomass at harvest (P = 0.13). The average
October, 2004.
value was calculated in 1.06 ± 0.11 ton DW ha-1 (Table 1).
Experimental design: A partial additive experimental design
The less conservative LSD test, however, indicated significant
was applied to study Russian knapweed interference (variable
differences between AR2 and AR32, AR64. Same results were
density) on onion transplants (constant density) under greenhouse
obtained for the aboveground biomass in elongated plants
conditions. The boxes were arranged in a complete randomized
block design (n=3) with 7 treatments for A. repens density (AR): (rosettes excluded) as shown in Table 1, were AR2 plots evidence
AR0 (crop check without weeds); AR2 (2 ramet m-2); AR4 (4 ramet a biomass lower than AR32 and AR64, which in turn were not
m-2); AR8 (8 ramet m-2); AR16 (16 ramet m-2); AR32 (32 ramet m-2): different between them. Intermediate densities also showed an
AR64 (64 ramet m-2). Both species were transplanted the same day intermediate response.
and they coexisted for a period of 145 days until onion harvest. Elongated ramet rate or proportion (elongated ramets/ total
The variables evaluated for onion plants included i) total and ramets) was the same among treatments, comprising the 62.8%
commercial fresh bulb yield (kg m-2), ii) non commercial bulb of the complete ramet population. The mean elongated plant
proportion by size (bulbs <35 mm diameter) in terms of biomass height was 34.4 cm and no differences among treatments were
and total number percentages, iii) onion plant survival (percentage observed (Table 1).
of dead plants) and iv) bulb size distribution by commercial
categories. Weed variables observed were: i) ramet density at Belowground biomass increased with the initial plant density in
harvest (final density, ramet m-2), ii) above and belowground a direct proportional way. Differences among treatments were
biomass at harvest (g DW m-2), iii) vegetative propagation rate observed for AR2 and AR4 with respect to AR64. Intermediate
(VPR, final ramet number/ initial ramet number). Dry biomass densities also showed an intermediate response (Table 1).
was obtained by desiccation of fresh samples in an electrical Onion bulb yield: As initial weed density increased, a decreasing
oven at 65 ºC for 72 h.
tendency was observed for both total and commercial yield (Fig.
Data analysis: While biomass data were logarithmically 2). Bulb production was not reduced under low weed density in
transformed (natural log, ln) to achieve homoscedasticity and AR2 (p>0.05). In this case the reference values were calculated
Table 1. Growth parameters of Acroptilon repens for initial densities of 0-64 ramets m-2 in onion transplanting
Initial density Final density Rate of elongated Total aboveground Aboveground Elongated plants Belowground
(ramet m-2 ) (ramet m-2) plants biomass biomass of elongated height biomass
(%) (g DW m-2) plants (g DW m-2) (cm plant-1) (g DW m-2)
2 25.2(6.7)a 76.7(6.2) 57.9(20.1)a 53.6(19.5)a 31.28(1.80) 25.8(8.8)a
4 47.6(6.7)a 60.5(5.9) 103.4(5.0)ab 101.3(11.6)ab 32.62(1.36) 47.1(7.1)a
8 87.8(30.9)ab 64.8(6.9) 117.5(34.9)ab 96.4(29.1)ab 36.57(1.48) 74.6(28.0)ab
16 81.0(3.6)ab 55.2(3.3) 96.6(7.3)ab 82.0(5.3)ab 38.00(3.39) 77.3(8.4)ab
32 164.6(52.8)b 61.6(4.7) 118.1(27.2)b 102.7(25.4)b 32.60(2.12) 105.6(44.4)ab
64 172.1(14.9)b 58.2(3.8) 142.5(12.1)b 120.7(10.8)b 35.13(4.09) 146.8(2.4)b
Average 62.8 106.0 92.8 34.37
Statistical P= 0.0089 P = 0.16 P = 0.13 P = 0.28 P = 0.42 P = 0.0219
significance LSD 0.05 = 59.90 LSD 0.05 = 58.85
Data between brackets correspond to standard error (SE) of the treatment mean value. Average values calculated when no differences were found.
 70 Growth and interference of invasive Russian knapweed on onion

20 60 240

Final weed density (ramet m-2 )

VPR (ramet ramet m-2 )

16

Total bulb yield (ton ha-1 )

180
40
12

120
8
20
4 60

0
0 20 40 60 80 0 0
0 20 40 60 80
Initial weed density (ramet m-2 )
Initial weed density (ramet m-2 )
Fig. 1. Vegetative propagation rate (VPR) of A. repens in a partial additive
competition experiment with onion cv “Valcatorce INTA”. VPR = final Fig. 4. Onion bulb yield (dot line) and final density of A. repens (full
ramet number per initial ramet. line) in a partial additive competition experiment with an initial weed
density range of 0-64 ramet m-2.
60
Bulb yield (ton ha-1 )

2.0 50

Weed dry weight (ton ha-1 )

45

Bulb yield (ton ha-1 )

40
1.5
30
AR 30
15 1.0 64
20
0
0.5
0 2 4 8 16 32 64 10
AR 4
Initial weed density (ramet m-2 ) AR
0.0 2 0
Fig. 2. Bulb yield in “Valcatorce INTA” onion under Russian knapweed 0 10 20 30 40 50 60 70
competition for a range of 0-64 ramet m-2 of weed density at transplant.
Initial weed density (ramet m-2 )
Bars indicate total (white) and commercial (black) yield (ton FW ha-1).
While statistical analysis was performed independently for each data Fig. 5. Total Onion bulb yield (full line-thick) and final A. repens
series, differences presented (SNK, P<0.05) are valid for both. aboveground (dot line) and belowground biomass (full line-thin) for
increasing A. repens initial density among 0-64 ramet m-2 in a partial
40 additive competition experiment.
35 < 35 mm Mean bulb loss was 13.7 ± 3.28 % of the total bulb number and
5 ± 1.19 % in weight, equivalent to 1.2 ton ha-1 (Table 2).
Bulb weight (ton ha-1 )

30 35-50 mm
50-70 mm
25 As observed yield differences can not be explained by differential
70-90 mm bulb loss among treatments. Plant mortality was 5.7 ± 1.4 % in
20
average without differences among treatments (Table 2).
15
The average size for commercial bulbs decreased as initial weed
10
density increased. The pattern of statistical differences was the
5 same as observed for total and commercial yield. While the mean
0 bulb diameter was 5.6 ± 0.1 cm for AR0, the value for AR32 and
0 20 40 60 80 AR64 was 20 % low (Table 2). The same grouping was observed
Initial weed density (ramet m-2 ) Table 2. Onion yield parameters under A. repens competition (0-64 ramet
m-2 at transplant time)
Fig. 3. Onion bulb yield for the most common commercial categories
(discard, < 35 mm; size 2, 35-50 mm; size 3, 50-70 mm; size 4, 70-90 A. repens Discard bulbs Discard bulbs Onion plant Commercial
mm of eq. diameter) under competition with increasing densities of the initial density rate (number rate (weight mortality bulbs
invasive weed A. repens. (ramet m-2) based) % based) % rate % diameter (cm)
by averaging AR0 and AR2, observing 42.3 and 41.1 ton ha-1 for 0 6.7(4.4) 1.6(1.1) 1.7(1.7) 5.6(0.1)a
total and commercial bulb yield, respectively. In AR32 and AR64 2 4.9(2.9) 5.1(4.6) 0.0(0) 5.5(0.3)a
the bulb production was statistically low in comparison with AR0 4 12.2(6.5) 3.0(1.7) 6.7(3.3) 5.3(0.2)a
and AR2. An average of 19.6 ton ha-1 of total bulb yield (-54%) and 8 6.7(6.7) 2.3(2.3) 5.0(0) 5.1(0.2)ab
18.1 ton ha-1 of commercial bulb yield (-56%) were calculated. 16 18.3(11.7) 7.0(3.6) 8.3(8.3) 5.0(0.2)ab
Intermediate treatments were intermediate in response as were 32 24.1(13.0) 8.5(4.3) 10.0(0) 4.6(0.1)b
cited for the other variables studied. 64 23.0(11.7) 7.5(3.9) 8.3(4.4) 4.7(0.1)b
Average 13.7 5.00 5.71
No bulb losses (discard) were observed at harvest due to
P value 0.34 0.47 0.35 0.0344 *
phytopathological disorders. Low size (bulb equatorial diameter
Data between brackets correspond to standard error (SE) of the treatment
lower than 35 mm) was the unique cause for bulb discarding. mean value. Average values calculated when no differences were
There were no differences among treatments for this variable. found.
 Growth and interference of invasive Russian knapweed on onion
for mean bulb fresh weigh (FW) at harvest. In AR32 and AR64
the calculated value was 52.97 ± 3.31 g FW bulb-1, 47.3 % lower
than AR0 (100.60 ± 4.11 g FW bulb-1). Intermediate treatments
were also intermediate in response (Table 2). Fig. 3 demonstrates
the incidence of weed density on bulb size for the most common
categories. While size 3 bulbs (50-70 mm eq. diam.) were less
represented at weed densities higher than 16 ramets m-2, size 4
ones (70-90 mm eq. diam) were not present. Low size categories
remained constant in proportion for the entire weed density
experimental range.
Discussion
Onion is a poor competitor against weeds. Soares et al. (2003)
reported a 95 % yield loss and 91 % of decrease in mean bulb
weight for a 98 days period of weed competition. The complexity
increases when perennial weeds are dominant. Creeping
perennials normally have a higher relative growth rate than
onion, with biomass accumulation, both above and belowground,
sufficiently high to strongly compete with the crop.
For instance, onion yield losses due to volunteer potato (Solanum
tuberosum) interference occur at densities commonly observed in
the field. One plant per 14.9 m-2 resulted in 10% crop yield loss
while 100% yield loss was achieved with 4 volunteer potato plants
m-2. Volunteer potato competition limits onion bulb size, resulting
in a lower quality, less valuable crop (Williams et al., 2004).
Because Russian knapweed is not controlled by herbicides in
a crop context, yield losses due to weed interference could be
severe. For 64 ramets m-2 Watson (1980) reported 75 % grain yield
loss in wheat and 88 % in corn. Other weed control operations,
like mechanical or hand weeding, are also ineffective because the
weed regrow and extend from creeping roots. In this experiment
we observed a 56 % average reduction in the onion commercial
yield for 32-64 ramets m-2 at transplant.
While an increasing tendency was observed as weed density
increased, non commercial bulb proportion was statistically
constant for the complete range of Russian knapweed initial
density. The observations reported a high coefficient of variation
(CV = 1.10) as consequence of which no significant differences
were detected among treatments. This variability was higher
than that observed for total bulb yield (CV = 0.37) and also for
commercial bulb yield (CV = 0.40).
Most reports indicate that certain plant traits are responsible for
invasive weed dominance. As cited by Mashhadi and Radosevich
(2004) and Rejmánek et al (2005) the list include i) vegetative
propagation, ii) a high competitive ability in resource foraging,
iii) high propagule pressure, iv) high seedling relative growth rate
and specific leaf area, v) seed dispersal by vertebrates, vi) high
phenotypic plasticity, and vii) allelopathy.
We suppose that Russian knapweed probably evidence various of
these traits. In this experiment we observed at least three of these:
a) Vegetative propagation: Weed density increased along the
experimental period as a consequence of vegetative propagation
and interference relationships between species resulted which
is a normal behaviour of creeping herbs. If we extrapolate the
experimental VPR, it is possible to infer that low density patches
would reach a density enough to strongly compete with onions in
71
a second year. A similar behaviour would be expected under field
conditions, at least in newly infected sites where belowground
biomass tends to increase as a function of time.
From the statistical analysis applied to final ramet density no
differences were found between the two lowest densities, AR2
and AR4, whose average value was 36.4 ramet m-2. Additionally,
no differences were observed between the two highest densities,
AR32 and AR64, for which the mean value was 168.4 ramet m-2
at harvest. Both groups were, in effect, different from each other.
As in all variables under study, intermediate treatments had an
intermediate response.
Total onion yield correlated well with both ramet density (Fig.
4, r =-0,75) and weed belowground biomass (Fig. 5, r =-0,68).
The last two variables were clearly interdependent because, at
least under our experimental conditions, ramet production was
directly associated with root growth (r = 0.87).
The relation among belowground biomass-bulb yield and ramet
density-bulb yield could not be observed for aboveground
biomass, which in turn had a similar value for the entire range of
weed density. This could be possible as a consequence of intra-
specific canopy limitation in the Russian knapweed population.
b) Competitive ability: The best predictors for competition
coefficients for pairs of species are: i) maximum plant size, ii)
root biomass allocation, iii) time of emergence and iv) seed size
(Freckleton and Watkinson, 2001). Russian knapweed is clearly
a dominant competitor in relation to onion plants when the first
three attributes from the list are considered.
It is generally assumed that root foraging traits are size symmetric
in coexistent species. This implies that the intensity of resource
acquisition is proportional to the size of the root system
(Schwinning and Weiner, 1998). In this context, we can argue
that the intensity of Russian knapweed plant competition would
be a consequence of a higher size than onion plants.
However, Rajaniemi and Reynolds (2004) working with several
herbs including Centaurea maculosa, which is similar to Russian
knapweed, demonstrate size asymmetry in weed root foraging due
to an anticipated use of limited soil nutrients. We also support
this hypothesis with respect to Russian knapweed competition. In
established patches of this weed there is a clonal root system that
is both extensive and pre-existent as compared with any direct
seeded or transplanted vegetable crop. Except for water, that is
not limited under furrow irrigation, nitrogen (N) fertilization
for instance is, both spatially and temporally concentrated.
Conventional N applications are normally localized near the
crop rows in two times along the crop cycle. Then, asymmetric
competition resulted as consequence of the extension of the weed
root system and also by an anticipated access to soil nutrients
with respect to crop plants.
c) Propagule pressure: The persistence of an important bank
of vegetative propagule allows us to understand the importance
of controlling the progress of Russian knapweed invasion in
agricultural fields at the LVRN. We must bear in mind the high
cost of land and also the loss of agricultural aptitude in invaded
soils. As cited by Soukup et al. (2004), crop losses and land
degradation are common in both natural and agricultural systems
affected by invasive weeds.
72 Growth and interference of invasive Russian knapweed on onion
In the irrigated LVRN we have demonstrated the competitive
ability of Russian knapweed over one of the most important
vegetable alternatives in the south of Argentina and 2nd national
vegetable crop with 25,000 ha cultivated area and 700,000 ton
annual production (FAO, 2006).
Characterization of the vegetative propagation ability of
Russian knapweed explained the constantly modified crop-weed
interference relationship. It is important to gain knowledge about
the importance of plant invasions and their impact in order to
persuade the public and private sectors about the urgent need to
limit the invasive weed dispersal and, additionally implement
concrete plans for controlling established populations in both
private agricultural and public sites.
Acknowledgements
This research was funded by the Universidad Nacional del
Comahue throughout its Secretaría de Investigación and it is
part of Carlos Bezic Doctoral thesis at Universidad Nacional del
Sur, Bahía Blanca, Argentina. We specially thank undergraduate
student Sergio Vazquez for field and lab support.
References
Bezic, C.R., O.A. Gajardo, S. Polo, L. Avilés, S. Cañón, W. Benitez,
S. Vazquez, S. Iribarne, D. Schwindt and A. Dall Armellina, 2005.
Distribución y abundancia del yuyo moro (Acroptilon repens L.) en
el Valle Inferior del río Negro. Actas Jornadas Interdisciplinarias
de Estudios Agrarios, Bs. As., Argentina, 8 p.
Dall Armellina, A. and H. Iglesias, 1984. Estrategias para el control de
“yuyo moro” Centaurea repens L. en parcelas hortícolas del Valle
Inferior del Río Negro. Actas X Reunión Nacional sobre malezas y
su control. Tucumán, Argentina, Tomo II: 43-46.
Dall Armellina, A.A. and R.L. Zimdahl, 1988. Effect of light on growth and
development of field bindweed (Convolvulus arvensis) and Russian
knapweed (Centaurea repens). Weed Sci., 36: 779-783.
FAO. 2006. FAOstat Prodstat module. On-line serch at http: //faostat.
fao.org.
Freckleton, R.P. and A.R. Watkinson, 2001. Predicting competition
coefficients for plant mixtures: reciprocity, transitivity and
correlations with life-history traits. Ecol. Lett., 4: 348-357.
Gajardo, O., C. Bezic, L. Avilés, S. Cañón and A. Dall Armellina, 2004.
Alelopatía del yuyo moro (Acroptilon repens L.) sobre maíz dulce.
Revista Pilquén, 6: 9-18.
Kearney, T.H., R.H. Peebles, J.T. Howell and E. Mc Clintock, 1960.
Arizona Flora. 2nd ed., University of California Press, Berkeley,
CA, 1085 p.
Khan, M.H., N. Khan and N. Badshah, 2003. Effect of weedicides and
hand weedings on the yield of onion (Allium cepa L.). Asian J. of
Plant Sci., 2(6): 464-466.
Mashhadi, H.R. and S.R. Radosevich, 2004. Invasive plants: Ecol. and
management. In: Weed Biology and Management. Inderjit (ed.)
Dordrecht, The Netherlands. Kluwer, p 1-28.
Panter, K.E. 1991. Neurotoxicity of the knapweeds (Centaurea spp.) in
horses. In: Noxious range weeds. James and Lynn F. (ed.) Boulder,
CO. Westview Press. p 316-324.
Pozzo Ardizzi, M.C., M. Abrameto, G. Pellejero, G. Aschkar, M.I. Gil,
and A. van Konijnemburg, 2005. Efecto del período de conservación
sobre algunas propiedades nutracéuticas y organolépticas en los
bulbos de cultivares nacionales de cebollas (Allium cepa L.) en el
Valle Inferior de Río Negro. RIA (Argentina), 34(3): 115-130.
Rajaniemi, T.K. and H.L. Reynolds, 2004. Root foraging for patchy
resources in eight herbaceous plant species. Oecologia, 141: 519-
525.
Rejmánek M., D.M. Richardson and P. Pyšek, 2005. Plant invasions and
invasibility of plant communities. In: Vegetation Ecology. Van der
Maarel E. (ed.). Blackwell Sci., Oxford. p-332–355.
Schwinning, S. and J. Weiner, 1998. Mechanisms determining the
degree of size asymmetry in competition among plants. Oecologia,
113: 447-455.
Soares, D.J., R.A. Pitelli, L.T. Braz, R. Gravena and R.E.B. Toledo, 2003.
Periods of weed interference with transplanted onion crop (Allium
cepa). Planta daninha, 21(3): 387-396.
Soukup, J., J. Holec, P. Hamouz and L. Tyšer, 2004. Aliens on arable
land. Proceedings of the Scientific Colloquium Weed Sci. on the Go.
Universität Hohenheim, 11-22.
Watson, A.K. 1980. The biology of Can. weeds. 43. Acroptilon
(Centaurea) repens (L.) DC. Can. J. Plant Sci., 60: 993-1004.
Whitson, T.D. 1987. Weeds and poisonous plants of Wyoming and Utah.
Res. Rep. 116-USU. Laramie, WY: University of Wyoming, College
of Agriculture, Cooperative Extension Service. 281 p.
Williams, M.M. II, C.V. Ransom and W.M. Thompson, 2004. Effect
of volunteer potato density on bulb onion yield and quality. Weed
Sci., 52: 754-758.
Williams, M.M., C.V. Ransom and W.M. Thompson, 2005. Duration of
volunteer potato (Solanum tuberosum) interference in bulb onion.
Weed Sci., 53: 62-68.
Journal of Applied Horticulture, 11(1): 73-77, January-June, 2009

Cauliflower hybrids for spring production in southern

mediterranean area

M. Sciortino1 and G. Iapichino2

1
Dipartimento di Scienze e Tecnologie Agroambientali, Università degli Studi di Bologna, Via Fanin 44, 40127, Bologna,
Italy. 2Dipartimento di Agronomia Ambientale e Territoriale, Sezione di Orticoltura e Floricoltura, Università degli Studi
di Palermo, Viale delle Scienze, 90128, Palermo, Italy. 2E-mail: giapichino@unipa.it

Abstract
New cultivars (F1 hybrids) of cauliflower (Brassica olearacea L. var. botrytis L.) were evaluated at four crop densities (1.3, 1.7, 2.2
or 3.3 plants m-2) for spring harvest crop in a Southern Mediterranean area (western coast of Sicily). The F1 hybrids (‘White-Flash’,
‘Milky-Way’ and ‘White Excel’) having white head, usually cultivated in Northern Italy and Europe in the autumn, were used. The
aim was the introduction of new varieties which can fill the gap from mid May to mid July, now existing in the Sicilian cauliflower
production, which is based on autochthonous ecotypes of green head varieties, e.g. ‘Cavolfiore Verde di Palermo’. Crop density
significantly influenced the growth and the phenology of the new hybrids. It was positively correlated to earliness, total marketable
yield and inversely to unmarketable product percentage and head size. The best crop density was found to be 2.2 plants m-2. Among
the cultivar tested ‘White Flash’ and ‘Milky Way’ appeared particularly suited for a spring harvest in the experimental environment.
They gave high yields with a minimum discard and uniform heads of approximately 1 kg of weight each.
Key words: Brassica olearacea L. var. botrytis L., planting density, cultivars, quality

Introduction
According to FAO statistics (2000-2005), Italy is the third
producer (500,000 t) of cauliflower (Brassica olearacea L.
var. botrytis L.) in the world, after China and India (7,250,000
and 4,800,000 t, respectively), followed by Spain, USA and
France. Italian production has always been commercialized also
abroad, but in the last thirty years exports gradually decreased
(Liguori,1985; Franca, 1985; Renzoni 2004). France and Spain
are particularly strong competitors. They are offering cauliflowers
of better and more constant quality standards. Moreover their
farms are generally larger than in Italy and this helps in reducing
management costs. Nowadays even traditional European
importers, i.e. The Netherlands, Great Britain and Germany, are
becoming new exporters, particularly with the introduction of cold
resistant hybrids (Ranco, 1985; Franca, 1985; Renzoni, 2004).
In the past, Italian production was based on ecotypes with loose
corymbs. Generally they lacked colour and size uniformity and
their ripening period was prolonged, aspects that heavily affected
commercialization and reduced customer satisfaction (Baldoni,
1982; Liguori, 1985; Acciarri et al., 2004). Exceptions were
rare. For example ‘Romanesco’ and ‘Verde Maceratese’ were
appreciated in the North European market for the typical shape
and for the green colour of their heads, respectively. In the last
few years Italian ecotypes lost importance, although many of
them have been genetically improved, maintaining their typical
organoleptic properties and particular taste. In Italy, most of the
cauliflower production comes from central and southern regions
(ISTAT, 2005). Campania is the first producer, followed by Puglia,
Calabria, Abruzzo, Marche and Sicily (Table 1). Marche’s crops
has always been focused towards the export. On the contrary,
most of the Sicilian products are commercialized on the home
market. Sicilian farmers generally use local varieties which
are less appreciated abroad. Their heads are usually coloured,
too heavy (about 2.5-3 kg head-1) and often scarcely uniform.
In western Sicily, the typical cultivar is ‘Cavolfiore Verde di
Palermo’ which has green, compact and medium-large corymbs,
while in the oriental part of the island there is a predominance
of ‘Cavolfiore Violetto di Catania’ that is characterized by large
and purple heads (Branca and Iapichino, 1997; Acciarri et al.,
2004). The ‘Cavolfiore Verde di Palermo’ has several ecotypes
with different earliness of production (Incalcaterra and Iapichino,
2000). These types, selected by growers during the centuries, are
now well identified and are called with the name of the harvest
month. They allow a continuous production sequence from August
to April, with a unique gap that goes from the middle of May to
mid July (Branca and Iapichino 1997, Incalcaterra and Iapichino
2000). The goal of this research was the introduction of new
cauliflower cultivars in the occidental part of Sicily for a spring
harvest in order to cover this productive gap. Today most of the
new available varieties of cauliflowers are F1 hybrid that were
bred in localities quite different from the Sicilian environments.
The adaptability of these species to many climates is confirmed
by several studies. For example, according to Wurr et al. (1993)
head formation requires a vernalization period at 9- 21°C. Grevsen
and Oelsen (1994) indicated a range between 0 and 26°C, while
Wiebe (1990) suggests maximum vernalization stimulus at 16
to 30°C. A further proof is provided by the cauliflower crops in
tropical regions, with about 25°C as yearly mean temperature.
However, the fitness of a new cauliflower variety to a particular
environment (and market) should always be thoroughly evaluated.
In this instance, a good cultivar for a spring cultivation in the
Southern Mediterranean area like Sicily should produce smaller
heads than the indigenous ecotypes in order to be appreciated
on the international market. Moreover, it should be precocious
and tolerate hot temperatures during head formation. Only thus
74 Cauliflower hybrids for spring production in southern mediterranean area
it can give new chances to the Sicilian cauliflower production
and marketing. The aim of the study was to explore the hybrid
cauliflower production during mid may to mid July in Northern
Italy and Europe.
Material and Methods
The research was carried out at the experimental farm of the
Horticulture and Floriculture branch of the Department of
Agronomy of the University of Palermo in the north-western
coast of Sicily (38°7’ N; 13°22’ E), Italy . The field experiment
was conducted on a typical Mediterranean litosoil. It consists of a
50 cm deep layer of sandy-clayey textured soil, lying on dolomitic
limestone rock, with the following main characteristics: sand
53.7%, lime 30.3%, clay 16.0%, pH 7.9, active limestone (Ca)
4.3%, Kjeldahl nitrogen (N) 1.2%, organic matter 3.1% (Lotti
Method), Olsen phosphorous (P2O5) 0.3% and exchangeable
potassium (K2O) 0.1%. Climatic parameters were continuously
recorded by a meteorological station situated at the experimental
farm. The experimental field was previously cropped with winter
melon (Cucumis melo L. var inodorus Naud.). In February 2000,
a ploughing to 0.35 m was carried out, followed by flexible
harrowing. Before transplanting, mineral fertilizer (11-22-16
NPK) was applied at 0.6 t ha-1 as a basal dressing. On March 19th
2000 the cauliflower seeds were sown in plastic trays with 45
mm diameter holes (Dufault and Waters, 1985a) and filled with
dark peat; the plantlets were placed first in a nursery and then
outdoor for a slow acclimatizing. Transplanting was made on
April 14th with 4-5 leaves/plant. The crop was hand weeded and
was irrigated throughout the season according to its need: one
irrigation per week starting from the middle of May.
The experiment consisted of a factorial comparison between
three cauliflower cultivars and four crop densities. The cultivars
were commercial F1 hybrids: ‘White-Flash’, ‘Milky-Way’ and
‘White Excel’, obtained by Sakata Co. (Japan). All of them have
white head and are widely used for autumn harvest in Northern
Italy and all over Europe. Compared densities were 1.7, 1.3, 2.2
or 3.3 plants m-2. They were obtained by maintaining constant
distance (1 m) between row and changing that along the row:
0.30- 0.45- 0.60 and 0.75 m.
Experimental plots were 18 m2 of area and were randomized
according to a complete randomized block design with three
replicates. At harvest, border plants within each plot were
discarded to avoid interferences between treatments. During
the growing season the rate of leaf appearance was assessed by
leaf counts on 15 plants for each plot at 15, 30 and 45 days after
transplanting. Moreover the time of head formation was recorded
when the change of the apices became visible and heading
earliness was expressed as number of days from transplanting.
At harvest, cauliflower heads were hand collected as soon they
were reached marketable stage i.e. sufficiently large, but still
compact. Head collection implied three passages in each plot,
throughout June. Collected corymbs were sorted into 5 diameter
classes: Ø<0.11m; 0.11-0.13m; 0.14-0.155m; 0.156-0.18m and
Ø>0.18m. This classification is linked to the number of heads
that can be placed in a standard commercial box and represents
the base of the selling price of cauliflowers in the Italian market.
Each box contains 24 corymbs of the 0.11-0.13m class, 18 pieces
of the 0.14-0.155m class, 12 pieces of the 0.156-0.18m class and
Fig. 1. Air temperatures and rainfalls from March to July 2000.
9 heads with Ø>0.18m. Ø<0.11m cauliflowers are considered
not marketable.
Data were analysed by a Two-Way ANOVA (cultivar x crop
density) procedure and the differences between means were
evaluated by Duncan’s multiple range test (SAS®, ProcAnova/
DuncanTest).
Results
During the experiment climate was very favourable to cauliflower
growth (Fig. 1). Minimum temperature varied from 3.4°C in
the first part of March to 19.5°C in last days of June, while
maximum temperature ranged between 17.0 and 34.6°C in the
same periods, respectively. Total rainfall was about 50 mm
distributed during the growing season. In particular, 26.65 mm
fell soon after transplanting (April 10 and 20) and two rainfall
(8 and 6 mm) happened during fertile stem elongation and head
ripening, respectively.
Leaf appearance rate: Leaf appearance rate was strongly reduced
by increasing crop density (Table 2). This effect can be easily
explained by considering the higher competition at root and shoot
level of densely grown plants. The parameter tested significantly
varied among cultivars. Fifteen days after transplanting ‘White
Flash’ showed more leaves per plant (1 or 3 more than the other
cultivars) and this difference was particularly marked in denser
crops. After 30 days ‘White Flash’ again had the maximum
number of leaves. In this instance, however, the differences with
the other cv. were significant only at the lowest density. Regardless
of the planting density, 45 days after transplanting, ‘Milky-Way’
produced the highest number of leaves per plant.
Head formation: The passage from vegetative to reproductive
phase varied widely (from 36.0 to 55.5 d from transplanting)
and depended mainly on genotype (Table 3). On average, ‘Milky
Way’ was the first hybrid to develop corymbs, followed by ‘White
Excel’ and ‘White Flash’. Heading response to crop density was
variable as well. The competition stress that is typical of a dense
stand accelerated head formation. However this effect was marked
in ‘White Excel’ (approximately one week between the highest
Table 1. Regional distribution of Italian cauliflower production (ISTAT,
2005)
Regions Area (ha) Production (t ha-1)
Puglia 2692 15.30
Campania 2858 29.61
Calabria 2316 33.31
Abruzzo 2117 23.73
Marche 1943 23.25
Sicilia 1804 21.86
Italian 17950 24.10
 Cauliflower hybrids for spring production in southern mediterranean area 75

Table 2. Influence of cultivar and crop density on the leaf appearance marketable yield was positively correlated to plant density. This
assessed 15, 30 and 45 days after transplanting implies that, even at the highest compared density, the plant
Crop density Number of leaves plant-1 competition was not so strong as to impair the high yield potential
(plant m )
-2
‘White Flash’ ‘Milky Way’ ‘White Excel’ of the new hybrids. Only for ‘Milky Way’ the highest density
15 days slightly reduced yield, probably because its large plants are better
3.3 8.7D 5.4G 6.7F suited to less dense plantings.
2.2 10.8B 8.6D 7.8E
Head quality: As expected, the weight of single heads (Table 5)
1.7 9.6C 9.3C 9.5C
was negatively correlated to plant density, but genotypic influence
1.3 11.6 A
11.5 A
10.8B
was equally important. ‘Milky Way’ always produced the heaviest
30 days
corymbs, which exceeded 1.3 kg head-1 at the lowest density. The
3.3 16.5BE 14.8DF 14.4DF head size was most sensitive to density in ‘White Excel’. From
2.2 19.6 AB
15.6 CE
15.0DF the lowest to the highest density the weight of its heads decreased
1.7 17.3 BE
16.1 BE
16.3BE more than half a kg.
1.3 20.8 A
18.7 CE
18.2AD
45 days Unmarketable heads were mainly due to bolting. Internode
3.3 20.9D 24.3BC 18.4E elongation, less leaves, thicker leaf lamina and thinner stem are
2.2 21.7 C
24.8 B
21.3D
all symptoms of competition stress that predispose the plant to
bolting (Table 6). Thus this adversity was clearly associated to
1.7 24.8 B
25.2 A
23.6BC
crop density. However genotype also had significant effect. In
1.3 27.3 A
28.2 A
24.4C
particular, the discard of ‘White Flash’ was always inferior to
Means followed by different letters are significantly different according 20%, even at the highest density (Table 6). On the contrary, in
to Duncan’s test at P<0.05
‘White Excel’ it consistently represented more than a quarter of
Table 3. Cauliflower heading time as affected by genotype and crop
density collected heads, with a slight variation in response to crop density.
Crop density Heading time (Days after transplanting) ‘Milky Way’ showed an intermediate behaviour. In general,
(plant m-2) the increase of crop density augmented the frequency of small
‘White Flash’ ‘Milky Way’ ‘White Excel’ cauliflowers reducing the number of the extra-large ones (Fig. 2).
In ‘White Flash’ Ø<0.11m corymbs were very few (10.8 %) at
3.3 36.0H 53.2B 41.8F 1.3 plants m-2 compared to 35.0% that were found at 3.3 plants
2.2 39.2G 54.6A 43.3E m-2, while the opposite happened for Ø>0.18m (31.3 and 5.4 %,
respectively). Also, in ‘Milky Way’ the extreme classes were
1.7 39.3G 55.3A 44.8D markedly influenced by plant density: the percentages of corymbs
1.3 39.6G 55.5A 48.1C belonging to the smallest class were 28.6 and 4.2% at 3.3 and
1.3 plants m-2 density, respectively, while those with Ø>0.18m
Means followed by different letters are significantly different according
to Duncan’s test at P<0.05 were 0.9 and 35.2%, respectively. A similar pattern, even more
Table 4. Influence of genotype and crop density on the yield of marketable Table 5. Influence of genotype and crop density on the unit weight of
cauliflower heads collected cauliflower heads
Crop density Marketable cauliflower heads (t ha-1) Crop density Unit head weight (g head-1)
(plant m-2) (plant m-2)
‘White Flash’ ‘Milky Way’ ‘White Excel’
‘White Flash’ ‘Milky Way’ ‘White Excel’
3.3 21.77 A
15.40 CDE
21.42 A
3.3 774GH 893G 711H
2.2 16.07CD 18.28B 17.28BC
2.2 993F 1023DE 796G
1.7 15.06 DE
13.99 DE
10.27 G
1.7 959EF 1186C 1015DE
1.3 13.24EF 11.54FG 8.20H
1.3 1053D 1353A 1270B
Means followed by different letters are significantly different according
to Duncan’s test at P<0.05 Means followed by different letters are significantly different according
to Duncan’s test at P<0.05.
and lowest density) and less noticeable in the other two hybrids Table 6. Percentage of unmarketable cauliflower heads as affected by
genotype and crop density
(2-3 days of difference, Table 3).
Plant density Percent unmarketable heads
Head yield: For each hybrid, the harvest period lasted less than (plant m-2) ‘White Flash’ ‘Milky Way’ ‘White Excel’
15 days and all heads were collected in three harvests. ‘White
Flash’ was the first cultivar to give a marketable production (in the 3.3 19.3EF 24.5CD 32.1A
first part of June), followed by ‘Milky Way’ and then by ‘White 2.2 11.4G 22.1DE 28.3B
Excel’, which started to be collected at the end of June.
1.7 8.5H 19.1F 25.4BC
The highest yields of marketable heads (Table 4) were obtained
with ‘White Flash’ and ‘White Excel’, both at the densest stand 1.3 6.7H 17.5F 26.6BC
(21.8 and 21.4 t ha-1, respectively). ‘White Excel’ at the lowest Means followed by different letters are significantly different according
density gave unsatisfactory production (8.2 t ha-1). On average, to Duncan’s test at P<0.05
76 Cauliflower hybrids for spring production in southern mediterranean area
a
b
c market, opening the export market to local growers.
Fig. 2. Distribution of collected heads into diameter classes (<11, 11-13,
14-15.5, 15.5-18, >18 cm) as affected by hybrid and planting density:
(a) White flash (b) Milky Way, (c) White Excel.
marked, was observed for ‘White Excel’. The percentage of
its corymbs belonging to the smallest class ranged from 2.6 to
33.2% when density increased from 1.3 to 3.3 plants m-2, while
the largest heads decreased from 30.5 to 0.7%. On the whole, the
percentage of intermediate diameter classes was more stable over
crop densities and hybrids.
Discussion
In this study, the increasing plant density significantly augmented
the marketable yield as already observed by Salter and James
(1975) in cauliflower. Our results are also consistent with those
obtained by Dufault and Waters (1985), who found that high
plant density reduced head weight in cauliflower cultivar Snow
Crown. ‘White Excel’ appeared highly influenced by density, as
at the highest stand it gave heads that were so small (less than
800 g head-1) as to be scarcely appreciated both by home and
international customers. ‘White Excel’ gave the most suitable
weights for the international market (1 kg head-1) at 1.7 plant
m-2, whereas ‘Milky Way’ and ‘White Flash’ gave satisfactory
head weight with high productivity even at 2.2 plant m-2. Higher
percentages of too small (Ø<0.11m) and too large (Ø>0.18m)
corymbs were obtained at 3.3 and 1.3 plants m-2, respectively.
Among the cultivar tested, ‘Milky Way’ can be considered the
best responsive with medium sized head of even uniformity over
a wide range of densities. ‘White Flash’ was the earliest hybrid,
followed by ‘Milky Way’ and ‘White Excel’. This earliness was
due to faster development, particularly during the initial stages of
growth, where it showed the fastest leaf formation. An increase of
crop density sped up growth. This influence, combined with the
genotype growth characteristic, can be used to achieve an earlier
production at the end of spring, which is particularly imporant
for the Sicilian cauliflower market.
The amount of unmarketable product was affected by density
through bolting; indeed at the highest density the discarded head
percentages was high. But bolting effect was influenced by hybrid
susceptibility as ‘White Excel’ (being later producing) was more
prone than ‘White Flash’ (being earlier producing).
The results indicate that the tested hybrids can be successfully
cropped in a Mediterranean climate area such as the western coast
of Sicily. Indeed, they can almost completely fill the gap in the
Sicilian cauliflower production (mid May- mid July) with a high
productivity, that was confirmed in our agronomic environment.
Both, the high productivity and the quality standard that the tested
cultivar showed might be greatly appreciated on the international
Planting density influenced both plant growth and phenology:
denser stands promoted earliness and augmented yield, but
implied smaller corymbs and high proportion of discardable
heads. Thus medium densities are recommended. On the whole,
the better quantitative and qualitative results were obtained
with ‘White Flash’ and ‘Milky Way’ transplanted at a density of
2.2 plants m-2: this density seems thus to be the best choice as
reference for further experiments.
Acknowledgments
We are grateful to Mr. La Rosa and Mr. Galati for technical
support.
References
Acciarri, N., F. Branca, E. Sabatini, S. Argento and V. Magnifico, 2004.
Miglioramento genetico dei cavolfiori a corimbo bianco e colorato.
L’Informatore Agrario, 24: 33-36.
Baldoni, G. 1982. La coltivazione del cavolfiore. Edagricole, Bologna,
pp. 83.
Branca F. and G. Iapichino, 1997. Some wild and cultivated Brassicaceae
exploited in Sicily as vegetables. Plant Genetic Resources Newsletter,
110: 22-28.
Dufault, R.J. and L. Waters Jr.,1985a. Container size influences broccoli
and cauliflower transplant growth but not yield. HortScience, 20
(4): 682-684.
Dufault, R.J. and L. Waters, Jr., 1985b. Interaction of nitrogen fertility
and plant populations on transplanted broccoli and cauliflower yields.
HortScience, 20(1): 127-128.
F.A.O. Food and Agriculture Organization of the United Nations. http://
www.fao.org/es /ess/top/commodity.jsp?lang=EN.
Franca, F. 1985. Associazionismo per difendere e riallacciare la
produzione del cavolfiore. L’Informatore Agrario, 16: 43-44.
Goldberg, D., B. Cornat and Y. Bar, 1991. The distribution of roots,
water, and minerals as a result of trickle irrigation. J. Amer. Soc.
Hort. Sci., 96: 645–648.
Grevsen, K. and J.E Oelsen, 1994. Modelling cauliflower development
from transplanting to curd initiation. J. Hort. Sci., 69: 755-766.
Incalcaterra, G. and G. Iapichino, 2000. Sowing time influences
cauliflower seed production. Acta Hort. 533: 45-52.
ISTAT Istituto Nazionale di Statistica. http://www.istat.it/agricoltura/
datiagri/ coltivazioni/anno2005.htm#italia.
 Cauliflower hybrids for spring production in southern mediterranean area
Liguori, A. 1985. Cavolfiore: come riconquistare i mercati. L’Informatore
Agrario, 22: 75-79.
Ranco, E. 1985. Difficile ma non impossibile riconquistare le posizioni
perdute sui mercati esteri. L’Informatore Agrario, 16: 21-24.
Renzoni, F., 2004. Obiettivi qualità certificata e tipicità per il cavolfiore
nazionale. L’Informatore Agrario, 24: 37-39.
Salter, P.J. and J.M. James, 1975. The effect of plant density on the
initiation, growth and maturity of curds of two cauliflower varieties.
J. Hort. Sci., 50 (3): 239-248.
77
SAS Institute, 1988. SAS/STAT User’s Guide, Release 6.03 Edition. Cary
NC, SAS Institute, Inc.
Wiebe, H.J. 1990. Vernalization of vegetables crops- a review. Acta
Hort., 267: 323-328.
Wien, H.C. and D.C.E. Wurr, 1997. Cauliflower, Broccoli, Cabbage and
Brussels Sprouts. In: The Physiology of Vegetable Crops. Wien H.C.
(Ed.). CAB International, New York, USA.
Wurr, D.C.E., J.R. Fellows, K. Phelps and R.J. Reader, 1993.
Vernalization in summer/autumm cauliflower (Brassica olearacea
var. botrytis L.). J. Exp. Bot., 44: 1507-1514.
 Journal
Journal of Applied Horticulture, 11(1): 78-80, January-June, 2009
Appl

Effect of different sowing times on development and

efficiency of some chinese cabbage varieties
(Brassica campestris sbsp. pekinensis)

Funda Eryilmaz Acikgoz

Faculty of Agriculture, Namik Kemal University, Tekirdag, Turkey. E-mail: fundaea@yahoo.com

Abstract
The aim of the study was to determine the effect of different sowing times on development and efficiency of some Chinese cabbage
varieties (Brassica campestris sbsp. pekinensis) under Corlu conditions. The study was conducted in Corlu County which has a tougher
climate than its Province Tekirdag where a similar research had been done before. The research was conducted in 2000 and three
different sowing times (15 August, 15 September and 15 October) and four domestic varieties (Tokat-2, Tokat-5, Tokat-29 and Tokat-
89) were used. The variety, Tokat-89 and the sowing time of 15 September were found to be the most suitable variety and sowing time,
respectively, The variety and time of sowing recorded superiority for head weight, level of hardness and head quality.
Key words: Chinese cabbage (Brassica campestris sbsp. pekinensis), sowing times, development, efficiency

Introduction which there were drainage holes and which were filled in with
peat. The sowings were done at the dates of 15 August (S1), 15
Chinese cabbage is a vegetable which is widely grown and September (S2) and 15 October (S3). The seedlings were planted
pleasingly consumed in the East Asian Countries; especially in in double lines on rows at 40x40 cm distances in the rows and
China, Japan, Korea and Taiwan, and which is also very well on the rows respectively.
adapted to the cool climate conditions (Vural et al., 2000).
When it is eaten fresh and raw, it is crisper, fresher and easily Production plan of the experiment
digested than the other cabbage sorts. For a working man, 100 g 1st Sowing time : 15 August
of Chinese cabbage supplies the total of daily need for calcium, 1st Planting time : 16 September
magnesium, and vitamin C while it also provides some other Fertilizing : 30 September
minerals and elements important for human nutrition (Eryilmaz 1st Harvesting time : 22 October
and Varis, 1996). Days from sowing to harvesting : 67
2nd Sowing time : 15 September
Since, Chinese cabbage has a short vegetation period as 2-2.5 2nd Planting time : 9 October
months; it can be grown as the second crop after harvesting of Fertilizing : 24 October
cereals. By using the region’s advantage of being closer to a 2nd Harvesting time : 14 December
metropolis as Istanbul, Chinese cabbage can be marketed in the Days from sowing to harvesting : 94
same season as lettuce for which its production can be done as 3rd Sowing time : 15 October
an alternative. Moreover, its production is suggested in between 3rd Planting time : 8 November
the growing periods of wheat and sunflower when the fields Fertilizing : 23 November
remain unused. Whether the results of the research conducted in 3rd Harvesting time : 18 January
Tekirdag Province in 1993 (Eryilmaz and Varis, 1994) can also Days from sowing to harvesting : 94
be used for Corlu County which has a tougher climate was the Irrigation was supplied with a filtered bucket in the seedling period
objective of the research. and in addition it was provided by furrow irrigation and raining
irrigation with a hose in the development period. (NH4)2SO4 2% N
Materials and methods was used for fertilization and it was supplied as 7 g m-2 (Eryilmaz
and Varis, 1996). The experiment was set in a double recurrence
The research was conducted in Corlu County of Tekirdag Province
and according to the experimental design of divided parcels.
in 2000 under field conditions. The varieties, Tokat-2 (V1), Tokat-
5 (V2), Tokat-29 (V3) and Tokat-89 (V4) well adapted to Tokat The average temperature was 17.7 oC, the average rainfall was
Region of the country, were used in the experiment and the seeds 44.3 mm, the average proportional humidity was 68.2 % and the
were obtained from Faculty of Agriculture; Gazi Osman, Pasa average wind speed was 2.1 m s-1 during the experimental period
University, Tokat, Turkey. (July-October).
The seeds of the varieties were sown as 2 seeds for each PE The observations and analysis methods were as per Opena and
bags of which the closed dimensions were 15 x 15 cm and the Lo (1980) and Eryilmaz and Varis (1996). Head length (cm) was
thickness was 0.15 mm, black coloured in order to check moss measured from the bottom to the top of the head. Head width was
growth and had bellows in order to stand upward, and under measued in the middle part of the head cut at length (cm). For
 The effect of different sowing times on development and efficiency of some chinese cabbage varieties 79

Table 1. The effect of sowing times on average head weight in some Chinese cabbage varieties (g)
Sowing Times Tokat-2 Tokat-5 Tokat-29 Tokat-89 Main Sowing Effect
15 August 1070.0 1475.0 1326.5 1375.0 1311.6AB
15 September 1470.0 1532.0 1610.0 1600.0 1553.0A
15 October 1500.0 1200.0 1100.0 1300.0 1275.0B
Main variety effect 1346.6 1402.3 1345.1 1425.0 1379.8
LSD for Main Sowing Effect: LSD (P=0.01)=192.7
Table 2. The effect of sowing times on the ratio of tight head in some Chinese cabbage varieties (%)
Sowing Times Tokat-2 Tokat-5 Tokat-29 Tokat-89 Main Sowing Effect
15 August 63.0 71.0 56.5 68.0 64.6A
15 September 62.0 58.0 57.5 43.0 55.1AB
15 October 51.5 60.0 43.0 60.0 53.6B
Main variety effect 58.8 63.0 52.3 57.0 57.7
LSD for Main Sowing Effect: LSD (P=0.01) = 11.2
number of leaves, all leaves were counted and averaged. For head obtained among the sowing times with respect to the averages.
weight plant head weight with outer leaves was measured after There was no significant difference within the varieties. However,
harvesting. Heading efficiency was calculated by dividing mean Tokat-29 gave the highest percentage of not forming head plant
head weight with non-wrapper leaf weight. (Table 4).
For head tightness, all plants were observed and evaluated after Hardness: The hardness of the varieties ranged from 0.39-0.42
each harvesting and they were explained as tight head, loose head (Table 3). The higher the hardness is, the tighter the head will be
and not forming head by (%). Hardness was determined by using and this is a desirable property, because the paleness is attained
following formula: Hardness (g cc-1) = Average head weight head later in plants with tight head. There was an excess weight in
volume -1. Where, Head volume (cc) = 0.524(d12 d2-1), d1 = Average the plants which have smaller volume and they are comfortably
Head width (cm), d2 = Average head length (cm). and profitably marketed (Yazgan and Edizer, 1987). Moreover,
a high value of hardness is advantageous for the producer,
Results and discussion carrier, marketer and consumer due to enduring (Yazgan and
Ece, 1990).
Head weight: Analyis of variance for testing effect of variety and Table 3. Hardness in Chinese cabbage varieties
sowing time on head weigh revealed that the mean sowing effect
Variety Hardness (g cc-1)
was significant at P=0.01 for head weight and sowing time of 15
Tokat-2 0.39
September gave the plant with the heaviest head (Table 1).
Tokat-5 0.40
Tight head: Analysis of variance indicated that the main sowing Tokat-29 0.40
effect was significant at P=0.01 for the tightness of head. Tokat-89 0.42
Consequently, the sowing time of 15 August gave the tightest head The results of the research are given as a whole in Table 4. For
(Table 2). Tight head is an important factor for Chinese cabbage Corlu conditions; the variety of Tokat-89 and the sowing time of
because of easy marketing (Gercekcioglu and Yazgan, 1989). 15 September were determined as the most convenient variety
Loose head: Analysis of variance revealed that the sowing time and sowing time, respectively, regarding: head weight, level of
hardness and head quality.
was important for loose head formation. Thus, the sowing time
of 15 October gave the plants having the loose heads.
References
Percentage of not forming head: The number of plants which Eryilmaz, F. and S. Varis, 1994. The effect of different sowing times
did not form head was quite low. Hence, variance analysis was on development and efficiency in some Chinese cabbage varieties.
not realised. From the sowing time of 15 August, 7.4 % and from The J. Trakya Univ., 3(1-2): 1-8.
the sowing time 15 October 20 % of not forming head plants were Eryilmaz, F. and S. Varis, 1996. The effect of different sowing times on

Table 4. The results of the criteria analysed (S: sowing time; V: variety)
Treatment Head length Head width Number of Head weight Seed forming Tight head Loose head Not forming Heading
(cm) (cm) leaves (g) head (%) ratio (%) ratio (%) head (%) efficiency
S 1V 1 23.8 15.8 36 1070.0 0 63 37 0 21400
S 1V 2 22.0 18.3 30 1475.0 0 71 30 0 29500
S 1V 3 24.0 16.7 33 1326.5 0 56.5 43 0 26530
S 1V 4 23.2 15.9 30 1375.0 0 68 42 0 27500
S 2V 1 26.2 16.1 33 1470.0 0 62 38 0 29400
S 2V 2 24.3 16.7 35 1532.0 0 58 41 4 30640
S 2V 3 27.4 15.2 33 1600.0 0 43 43 11 32000
S 2V 4 28.2 16.4 34 1610.0 0 57.5 32 14 32200
S 3V 1 25.5 16.3 40 1500.0 0 51.5 30 18 30000
S 3V 2 22.8 15.8 42 1200.0 0 60 17 23 24000
S 3V 3 23.4 16.1 33 1100.0 0 43 20 35 22000
S 3V 4 25.2 15.2 38 1300.0 2 60 15 25 26000
80 The effect of different sowing times on development and efficiency of some chinese cabbage varieties
development and efficiency in some Chinese cabbage varieties. The
Project of South-eastern Anatolia (GAP) 1st Agriculture of Vegetables
Symposium: 106-112, 7-10 May, 1996, Sanlıurfa, Turkey.
Gercekcioglu, R. and A.Yazgan, 1989. A research on determining the
convenient Chinese cabbage varieties (Brassica campestris L. Ssp.
pekinensis Lour Olsson) for Tokat region’s fall season. Doga TU
Tar. ve Or., 13(2): 255-264.
Opena, R.T. and S.H. Lo, 1980. Procedure for Chinese cabbage
evaluation trials. International Co-operator’s Guide, AVRDC 80-
144, 6.
Vural, H., D. Esiyok and I. Duman, 2000. Culture Vegetables (Vegetables
Growing), Ege University Prints, Izmir, Turkey. p 23-129.
Yazgan, A. and Y. Edizer, 1987. The determination of Chinese cabbage
varieties (Brassica campestris L. Ssp. pekinensis Lour Olsson)
convenient to be grown in Tokat Province’s spring and summer.
The Journal of Cumhuriyet University Faculty of Agriculture, 6(1):
127-149.
Yazgan, A. and A. Ece, 1990. The adaptation of Chinese cabbage
variety nominees Tokat-2, Tokat-5, Tokat-29 ve Tokat-89 in Tokat
conditions. The Journal of Cumhuriyet University Faculty of
Agriculture, 6(1): 439-449.