1 Analysis of the diversity and productive potential of timber trees in cocoa agroforestry systems in

2 Alta Verapaz, Guatemala

3
4 Carlos E. Villanueva-González1,2, José A. Ruiz-Chután1,3, Zbyněk Polesný1, Marie Kalousova1,
5 Cristobal Villanueva4, Edwin E. Vaides-López2, Hermes E. Bol-Kloth5, Carlos A. Chavarría-
6 Ramírez2, Bohdan Lojka1, *
1
8 Department of Crop Sciences and Agroforestry, Czech University of Life Sciences Prague, Kamýcká
9 129 165 00 Praha – Suchdol, Czech Republic
2
10 Facultad de Ciencias Ambientales y Agrícolas, Universidad Rafael Landívar, Campus San Pedro Claver,
11 S.J. de la Verapaz, 16010, carretera a San Juan Chamelco, Alta Verapaz, Guatemala
3
12 Facultad de Agronomía, Universidad de San Carlos de Guatemala, Ciudad Universitaria Zona 12,
13 010012, Guatemala
4
14 Centro Agronómico Tropical de Investigación y Enseñanza, 19-7170, Turrialba, Cartago, Costa Rica
5
15 Consorcio Regional de Investigación Agropecuaria, Instituto Interamericano de Cooperación para la
16 Agricultura, Guatemala
*
17 Corresponding author (lojka@ftz.czu.cz).
18
19 ABSTRACT
20
21 Timber production in agroforestry systems can improve environmental and economic sustainability,
22 especially for families living in rural areas. Despite its importance in Guatemala, little is known about the
23 potential of wood and its contribution to the sustainability of traditional agroforestry systems. This study
24 aims to analyze the timber resources in cocoa agroforestry systems (CAFS) of different ages, evaluating
25 variables such as floristic composition, species diversity, and the potential to obtain timber from the
26 species present in the evaluated CAFS. Twenty temporary sampling plots with a size of 2500 m 2 were
27 established, and dendrometric variables were measured in all plots, including total height (th), commercial
28 height (ch), and the diameter at the breast hight (DBH) of the inventoried species. Species occurrence was
29 statistically analyzed by cross-tabulation and Pearson's Chi-square test (χ2) to determine species'
30 frequency distribution among the evaluated agroforestry systems ages. 827 trees of 38 species belonging
31 to 19 families were identified. The most represented species in the age range of the evaluated CAFS were
32 Gliricidia sepium (35.07%), Swietenia macrophylla (19.11%) and Inga sapindoides (7.62%). The total
33 shade trees occupied a basal area of 33.29 m 2, and a total volume of 352.35 m 3 was recorded, of which
34 148.9 m3 were for commercial use. The most frequent uses of the wood were firewood (38.21%), thin
35 plank (30.23%), and poles (22.85%), while the use of thick plank was the least common (8.71%). In

1 1
36 conclusion, cocoa agroforestry systems in the studied area are characterized by a wide diversity of shade
37 tree species. These results provide a solid basis for promoting sustainable practices that foster the
38 productivity and preservation of these cocoa agroforestry systems, thus contributing to the sustainability
39 and well-being of farmers and the environment.
40
41 Key words: cocoa agroforestry; dendrometric variables; floristic composition; plant biodiversity; timber
42 production
43
44 INTRODUCTION
45
46 Currently, agroforestry presents multiple opportunities to improve ecosystem services to people and
47 communities and at the same time decrease pressure on the environment
(Braga et al., 2019; Manaye et al., 2021)
48 while providing number of useful tangible products to farmers. Several studies have identified
49 the positive impact of agroforestry practices implemented by smallholder farmers on production and
50 resource management in their plots (Maza et al., 2016; Arrazate et al., 2021) . Integrating woody and non-
51 woody species, such as fruit, is commonly employed by smallholders to diversify land use in agroforestry
52 systems (AFS) (Ordoñez Espinosa and Rangel, 2020) . Trees play a fundamental role in agroforestry
53 production and farmers' livelihoods in Latin America, as they provide a wide range of ecosystem services,
54 such as climate and water cycle regulation, soil protection, pollination, and disease regulation, which are
55 essential for maintaining the health and productivity of agricultural and forestry systems, and improving
56 the resilience and adaptation of rural communities to the impacts of climate change
(Salvador-Morales et al., 2020; Morán-Villa et al., 2022; Ra
57 . The diversification of the tree component
58 in these systems is influenced by edaphoclimatic conditions, as well as by the needs and beliefs of
59 producers (Detlefse and Somarriba, 2012) . In Mesoamerica, about 85% of cocoa production is grown in
60 association with shade trees and annual and perennial plants (Matey et al., 2013). In Mexico, for example,
61 46 tree species have been recorded in traditional cocoa-agroforestry systems (CAFS)
(Roa-Mero et al., 2009)
62 . Furthermore, in countries such as Costa Rica, Colombia, and Peru, studies of floristic diversity and
63 structure in cocoa agroforestry systems (CAFS) have revealed the presence of 56, 61, and 33 species,
64 respectively (Suatune et al., 2003; Vebrova et al., 2014a; Ordoñez Espinosa and Rangel, 2020). According
65 to Sonwa et al. (2017), besides providing shade, tree species cultivated in CAFS are frequently used as
66 food, medicine, firewood, timber, and cultural value.
67 In these systems, trees are highly valued for their timber potential and the opportunities they
68 provide to diversify household income and livelihoods. For example, in the Kichwa indigenous
69 community of Ecuador, in the CAFS called Chakras, Cordia alliodora (Ruiz & Pav.) Oken and Cedrela

2 2
 70 odorata L. were identified as the species with the highest timber potential (22 m 3 ha-1) (Jadán et al., 2015),
71 while, in Cardenas, Mexico, the species Erythrina poeppigiana (Walp.) O.F. Cook (64.4 m3 ha-1), E.
72 americana Mill. (40.3 m3 ha-1), and C. odorata (15.5 m3 ha-1) recorded the highest values
(Sánchez Gutiérrez et al., 2016)
73 . Recent studies indicate that timber production is experiencing an increase among
74 small producers in indigenous communities around the world. This trend is mainly due to the scarcity of
75 resources and the importance of wood in the development of productive activities, as well as in the supply
76 of commodities for the construction of houses, roof thatching, fence posts, and firewood
(Sousa et al., 2015; Tapia Vera et al., 2021; Morán-Villa et al., 202
77 .
78 Due to the growing interest in cocoa production in AFS, numerous studies have been carried out
79 to evaluate their functionality and sustainability. Current studies in Latin America have evidenced the
80 complex structure of CAFS (Suatune et al., 2003; Navarro Garza et al., 2012) , the floristic diversity
81 (Mejia-Rueda et al., 2023) and the contribution of the tree component to biodiversity conservation
82 ( Manaye et al., 2021).
83 According to Bullock et al. (2020) the total deforested area in Guatemala between 2000 and 2017
84 was around 854,137 ha of forest. Deforestation has reduced species richness and diversity, as well as the
85 timber supply in the country. Agroforestry is considered an effective strategy in several regions of
86 Guatemala to address various socioeconomic and environmental phenomena, including deforestation
87 (Braga et al., 2019). However, little is known in Guatemala about the timber potential and its contribution
88 to the sustainability of traditional CAFS. This study aims to examine the timber resource in CAFS of
89 different ages and evaluate variables such as floristic composition, species richness, and timber productive
90 potential of woody species present in the CAFS located in four municipalities of the department of Alta
91 Verapaz, Guatemala.
92
93 MATERIALS AND METHODS
94 Study area
95 The study was carried out in CAFS located in the municipalities of Cahabón, Cobán, Lanquín and Panzós
96 in the department of Alta Verapaz, Guatemala (Figure 1). These municipalities were chosen due to the
97 importance of cocoa cultivation in the region (MAGA, 2016) . The climatic zone in the study area
98 corresponds to the very humid subtropical (warm) forest (bmh-S (c)) (Holdridge, 2000). The area is
99 situated along the altitudinal gradient from the lowlands at an elevation of around 100 m a.s.l. to montane
100 zones exceeding 2000 m a.s.l. (IDES, 2012). Although conditions vary between municipalities in Alta
101 Verapaz, cocoa in the department is generally grown at altitudes ranging from 100 to 800 meters above
102 sea level. The temperature varies slightly over the year (from 24 to 28.1 °C) with an annual average of
103 25.6 °C. The average annual precipitation is estimated at 2199 mm, varying from 1426 to 4071 mm.

3 3
104 Study plot and sampling procedure
105 Twenty temporary sampling units (n=20) were established, each measuring 50 x 50 meters. Five samples
106 were taken in each municipality (Ramírez-Meneses et al., 2013). A non-probabilistic selection method of
107 study units was used, considering specific selection criteria such as time availability of producers, age of
108 the CAFSs, management, and crop area (>2500 m 2). The CAFS in the age range of 3 to 25 years were
109 evaluated, with a predominance of plots (11) in the age range of 7 to 12 years. These CAFS were located
110 at different altitudes, ranging from 132 to 500 m asl and they occupied areas ranging from 0.5 to 5 ha.
111
112 Floristic composition and dendrometric variables
113 The main dendrometric variables of all woody specimen were measured in each sampling unit, including
114 total height (th), commercial height (hc) and diameter at breast height (DBH) ≥5 cm. The formulas
115 recommended by Sánchez Gutiérrez et al. (2016) were applied as follows to calculate basal area (BA),
116 commercial volume (VC), and total volume (VT): BA=0.7854 x DAP2, where 0.7854 represents the
117 constant. Total volume and merchantable volume (VT, VC, m 3) were calculated using the equation V=BA
118 x ff x H, where: ff= shape factor (0.70) and H= total or merchantable height. Total height and DBH were
119 measured using a clinometer and a diametric tape.
120 Botanists and local producers assisted in identifying the scientific and common names of the species in the
121 field. For this purpose, duplicates of each identified species were collected corresponding to botanical
122 specimens. Utilized various resources to verify the species' identity, including the Flora of Guatemala
123 (Standley and Steyermark, 1946), the Tropics nomenclature database (www.tropicos.org, 2023), GBIF.org
124 (www.gbif.org, 2023), and the Guide for Identification of Common Trees in Guatemala by
Guerra-Centeno et al. (2016)
125 . The confirmation process was conducted by the Basic Sciences Laboratory at
126 Universidad Rafael Landívar in Guatemala.
127
128 Characterization of the vertical structure and uses
129 The method proposed by Somarriba (2004), based on the forest inventory results, was used to stratify the
130 shade canopy. The trees were classified into three strata: 1) the lowest stratum (indicated by trees and their
131 seedlings/saplings 1-8 m tall), 2) the middle stratum (indicated by trees 9-24 m in height), and 3) the
132 tallest stratum (indicated by emergent trees 25-35 m in height). To estimate timber production, the DBH
133 variable was classified into different diameter categories: firewood (5-10 cm), poles (10-15 cm), thin
134 boards (15-30 cm) and thick boards (>30 cm; (Sánchez Gutiérrez et al., 2016)).
135
136 Data analysis

4 4
137 To analyze the contribution of each species to the different quantitative measures, we utilized the ggplot2
138 statistical package (Wickham, 2016) to generate a bar chart. The frequency of the species was analyzed
139 using a statistical cross-tabulation approach to determine the frequency distribution of species across the
140 different age categories of each population where the CAFSs were located. To assess the relationship
141 between the variables included in the cross-tabulation, Pearson's Chi-square test (χ2) was used. These
142 analyses were conducted using the ggstatsplot package (Patil, 2021) implemented in R.
143
144 RESULTS
145
146 Floristic diversity and composition
147 Eight hundred twenty-seven tree individuals corresponding to 38 species belonging to 19 botanical
148 families were found cultivated within all sampled CAFS plots (Table 1). The behavior of the specific
149 richness is determined between 11 and 25 species per municipality, with the municipality of Panzós being
150 the area with the highest richness. At the same time, Cobán has the lowest species record. This difference
151 originates from the prioritization by the producers of Cobán in managing trees of high commercial value
152 in their agroforestry designs. In the other municipalities, a more traditional management approach
153 prevails. On average, a density of 165.4 trees ha -1 was found, varying from 36 to 364 trees ha -1 (Table 2).
154 The Fabaceae and Meliaceae were the most represented families with 45.7% and 25.6% of the total
155 number of species, respectively. Gliricidia sepium (Jacq.) Kunth. (290; 35.1%), Swietenia macrophylla G.
156 King (158; 19.1%), Inga sapindoides Willd. (63; 7.6%), C. odorata (54; 6.5%), and C. alliodora (42;
157 5.1%) were the most abundant species.
158
159 Frequency of trees by CAFS-age
160 According to dominant species of CAFS tree component, we can distinguish four types: (i) CAFS in
161 Cahabón, where G. sepium and S. macrophylla area the predominant species; (ii) CAFS in Cobán,
162 characterized by the presence of S. macrophylla and I. sapindoides; (iii) CAFS in Lanquín, where G.
163 sepium and Protium copal (Schltdl. & Cham.) Engl. are the dominant species; and (iv) CAFS in Panzós,
164 where G. sepium and C. odorota prevail. A pattern of species occurrence by age was observed in all
165 zones, suggesting that species are not distributed indistinctly from the age of the CAFS in each locality
166 (p<0.05) (Figure 2).
167
168 Tree basal area reflecting particular species and CAFS age
169 The results were classified according to the municipality, the age of establishment of the CAFS evaluated,
170 and the basal area (BA) variable for each trees species inventoried. A total of 827 shade trees were

5 5
171 recorded, occupying a BA of 33.29 m2. The mean BA was 6.65 m2 ha-1, ranging from 0.62 to 13.57 m2 ha-
1
172 . The highest BA value (13.57 m2 ha-1) has been determined for 10-year-old CAFS, while a 13-year-old
173 CAFS presented the lowest BA value (2.23 m 2 ha-1, Figure 3). The species G. sepium, S. macrophylla, C.
174 odorata, and C. alliodora represented 71% of the recorded BA. In the three and 4-year-old CAFS, S.
175 macrophylla was the dominant species in the BA with a density average of 254 trees ha -1, while in the 10,
176 11, 12, 14, and 25-year-old systems, G. sepium prevailed with 172 trees ha-1. The locality with the highest
177 BA was Panzós, with a value of 10.99 m2 ha-1, while Lanquín presented the lowest BA with 4.02 m 2 ha-1
178 (Figure 4).
179
180 Total and commercial volume of timber
181 A total volume of timber 352.3 m 3 was recorded in the whole sample, consisting of 5 ha with a mean of
182 70.4 m3 ha-1, varying from 4.9 m3 ha-1 to 171.4 m3 ha-1. As for the commercial volume, 148.9 m 3 was
183 reached in the total area, with an average of 29.7 m 3 ha-1 per plot, ranging from 2.9 m3 ha-1 to 73.1 m3 ha-1.
184 The five species that stand out in the present study, in terms of total volume per species and localities, are
185 G. sepium (25.1 m3 ha-1), S. macrophylla (9.4 m3 ha-1), C. odorata (8.2 m3 ha-1), C. alliodora (7.1 m3 ha-1)
186 and Vochysia guatemalensis Donn. Sm. (5.3 m3 ha-1), followed by M. indica (2.4 m3 ha-1), I. sapindoides
187 (2.3 m3 ha-1), Persea americana Mill. (1.6 m3 ha-1), and Vatairea lundellii (Standl) (1.4 m3 ha-1). The
188 remaining 29 species presented values < 1 m 3 ha-1. The 25- and 13-year-old CAFS reported the highest
189 and lowest values, with 133.3 m3 ha-1 and 18.4 m3 ha-1, respectively. The same analysis by locality showed
190 that Panzós had the highest total volume with 133.8 m 3 ha-1, while Lanquín had the lowest value with 30.1
191 m3 ha-1.
192 The results of our study indicate that the most relevant species for commercial timber production
193 are S. macrophylla (5.6 m3 ha-1), C. odorata (5.1 m3 ha-1), G. sepium (5.1 m3 ha-1), C. alliodora (4.3 m3 ha-
1
194 ) and V. guatemalensis (3.2 m3 ha-1). In contrast, the remaining 33 species presented figures <1 m 3 ha-1,
195 suggesting their potential for commercial timber production is much lower than the five species
196 mentioned. The 3-year- old CAFS (46.6 m 3 ha-1) and 14-year-old (8.0 m3 ha-1) presented the maximum and
197 minimum values, respectively (Figure 5). This difference is because the 3-year-old CAFS harbored timber
198 species of high commercial value with outstanding DBH and H.
199
200 Description of vegetation structure
201 The tree inventory confirmed the presence of three main strata in the study area. The lowest stratum (1-8
202 m), representing 19.3% of the total number of recorded trees, with an account of 160 individuals, the most
203 representative species being G. sepium, I. sapindoides, and P. copal. These results confirm the presence of
204 short-cycle species in this canopy, which is attributed to enrichment practices and management of natural

6 6
205 regeneration. The middle stratum (9-24 m), with a total of 648 individuals, represented 78.3% of the total,
206 and two dominant species were identified, G. sepium and S. macrophylla. On the other hand, the high
207 stratum (25-35 m), represented by 19 individuals and concentrating only 2.3% of the total, had C.
208 alliodora as the dominant species. The average total height of the trees was 11.9 m, with a variability
209 ranging from 2 m for G. sepium to 27 m for V. guatemalensis.
210
211 Potential use of the trees as a function of DBH
212 The statistical analysis of the DBH variable revealed that the average diameter of trees associated with
213 CAFS in Alta Verapaz is 20.2 cm, ranging from 5 cm to 90 cm. The distribution of individuals showed
214 that the highest concentration was found in the diameter interval ≥5 cm and <15 cm, with a total of 316
215 individuals, while the lowest concentration was observed in the diameter interval ≥35 cm, with a total of
216 72 individuals. Upon analyzing the results, it is evident that the primary use of trees associated with CAFS
217 in Alta Verapaz is obtaining firewood, accounting for 38.1% of the cases. According to interviews,
218 farmers in the region use these trees as a source of fuel. Additionally, significant use of the trees was
219 observed for obtaining thin boards at 30.2% and posts at 22.8%. These materials are used to construct
220 rural houses, especially for making roofs, walls, and fences to restrict properties and agricultural areas.
221 Furthermore, they produce furniture such as tables, chairs, and beds. On the other hand, the least frequent
222 use was obtaining thick boards, representing only 8.71% of the cases.
223
224 DISCUSSION
225 Floristic composition and diversity
226 In terms of species diversity, the CAFS of Alta Verapaz are similar to other cocoa-growing areas of the
227 Mesoamerican region (Suarez-Venero et al., 2019; Salvador-Morales et al., 2020) . However, studies
228 carried out in Colombia by Jadán et al. (2015) and in Ecuador by Ordoñez et al. (2020) suggest that
229 traditional CAFS managed by smallholders have an even greater diversity at species and family levels,
230 with records of up to 110 species and 68 families, and 61 species and 26 families, respectively. Compared
231 to the results of our study, there are other cocoa-growing regions with higher richness and diversity of
232 species; this could be because, in the Amazon, cocoa plantations are established in areas of cleared forests,
233 while in Alta Verapaz, most of the shade trees in the CAFS were intentionally planted after clearing the
234 forests. Additionally, some trees already existing in the place at the time of the establishment of the CAFS
235 were conserved. Despite this difference, these results are valuable for the conservation of biodiversity
236 since they indicate that the management of CAFS by small producers can foster a greater diversity of
237 species and families compared to other cultivation systems (Sol-Sánchez et al., 2018; Niether et al., 2020).
238 These findings are significant since biodiversity is essential for the maintenance of healthy ecosystems

7 7
239 and for the provision of ecosystem services that are critical to human well-being, such as pollination
240 (Barrios et al., 2018) , pest control (Delgado-Vargas et al., 2022) and firewood supply
(Sibelet et al., 2019)
241 .
242 In Mesoamerica, various studies have identified the potential of certain tree species cultivated in
243 diverse AFS designs to produce wood; these studies have also highlighted the possibility of selling this
244 wood in local markets as a source of income for farmers
(Detlefsen and Somarriba, 2012; De Sousa et al., 2015)
245 . In the context of agroforestry arrangements in Alta Verapaz, the frequency of species (Figure 2),
246 can have substantial economic implications for local communities that depend on agricultural production.
247 A concrete example is found in the municipalities of Lanquín and Cahabón, where the sale of resins
248 extracted from P. copal trees generates complementary income for family sustenance during most of the
249 year. In the case of Cobán, the producers manage the S. macrophylla and C. odorota trees, before final
250 harvesting, by pruning and thinning, obtaining economic resources simultaneously. Therefore, it is
251 essential to carefully consider species selection in agroforestry designs to promote resilience and long-
252 term sustainability. During interviews with cocoa producers in Alta Verapaz, their interest in preserving
253 tree species in these designs that are disease-resistant, climate-tolerant, and highly productive was evident,
254 particularly regarding the supply of firewood and timber to meet their basic needs. Therefore, it is
255 essential to take into account economic, ecological, and social aspects when selecting and managing
256 species in these agroforestry arrangements (Núñez et al., 2021).
257
258 Wood productive potential
259 Wood production in the CAFSs of Alta Verapaz is crucial because many people in rural Guatemala
260 depend on firewood for cooking and other household needs (Pineda, 2022). In addition to covering these
261 basic needs, timber production in these systems can generate additional income for farmers and their
262 families, especially in communities where paid employment opportunities are limited
(Sibelet et al., 2019; Tscharntke et al., 2022)
263 . Well-designed AFSs can provide a significant volume of wood and by-products
264 when correctly managed (Soncim et al., 2021) . This management may involve pruning practices, careful
265 species selection, and crop rotation, which contribute to increasing soil productivity (Niether et al., 2018).
266 The richness of tree species recorded in this study represents a distinctive CAFS trait in the Alta Verapaz
267 department. This finding provides a valuable perspective on the potential of the tree component in CAFS
268 to offer a wide range of ecosystem services for the population (Bukomeko et al., 2019) . One of the key
269 indicators to understand the relevance of shade trees in the evaluated CAFS is the total basal area
270 recorded, which reached 33.29 m2. It is essential to highlight that the 10-year-old CAFS stood out by
271 exhibiting the highest basal area value, reaching 13.57 m 2 ha-1 (Figure 3), suggesting that this specific
272 system has effectively promoted the growth and development of shade trees. In contrast, in Lanquín, the

8 8
273 13-year-old AFS presented the lowest basal area value, with only 2.23 m 2 ha-1 (Figure 3). This disparity
274 could be attributed to several factors, among which silvicultural management and the diversity and density
275 of recorded trees stand out. These factors have been mentioned in previous studies as possible causes of
276 the variability in the values of BA in AFSs
(Garza et al., 2012; Haggar et al., 2015; Suarez-Venero et al., 2019; Tapia Vera et al., 2021)
277 .
278 In the Panzós case, producers preferred to preserve some remaining trees from the secondary forest within
279 the CAFS. This decision positively affected the reported total wood volume, with species such as V.
280 guatemalensis and C. alliodora standing out. On the other hand, it was found that the timber potential in
281 Lanquín differs from that of the evaluated CAFS in other municipalities. This is due to the farmers'
282 preference for fast-growing trees with less canopy coverage, which, in turn, have the capacity to provide
283 shade and improve soil fertility (Matey et al., 2013; Villanueva-González et al., 2023) . Among the most
284 prominent species in this preference are G. sepium, P. Copal, I. Sapindoides, and T. Bicolor, which are
285 used both for firewood production and other essential products for the subsistence of the local population.
286 According to Nicli et al. (2019) report, the versatility, and multiple benefits these species offer to
287 agricultural communities have been crucial for their adoption, leading to the recognition of their
288 importance in the department's management and conservation of agroforestry systems.
289
290 Potential use of trees based on DBH
291 Our results indicate that wood's potential for producing thick planks is lower than that recorded by other
292 authors (Sánchez Gutiérrez et al., 2016) who report that up to 27% of the total volume is used for this
293 purpose. The fact that most of the trees associated with the CAFS are used for producing firewood, thin
294 boards, and poles suggests that the wood has a low value in the local market and is not considered a
295 critical income for the producers. In addition, the low frequency of use of wood to produce thick planks
296 indicates a low demand for this type of wood in the area, which limits business opportunities for
297 producers. To increase the diversity of timber trees and ensure usable volumes in the future in those CAFS
298 with low commercial yields, it is essential to improve agroforestry designs and apply silvicultural
299 planning tools (Esche et al., 2023) . A key strategy is to prioritize the combination of trees intended for
300 subsistence and those of commercial value, aiming to enhance forest production and the satisfaction of
301 basic family needs (Méndez et al., 2013; Villareyna et al., 2020) . However, it is essential to highlight that
302 the lack of economic valuation of timber in the local market does not necessarily imply that wood has no
303 value in ecosystem services (Sol-Sánchez et al., 2018). Trees in AFS provide a series of benefits, such as
304 carbon sequestration (Ibrahim et al., 2007) , climate regulation (Peralta-Rivero, 2022) soil conservation,
305 and biodiversity (López-Baez et al., 2015) , which can have a positive impact on the sustainability of the
306 production system. In addition, there are opportunities to explore other markets, such as crafts, and access

9 9
307 to certification programs and sustainable markets; by maintaining a diversity of trees, small producers can
308 meet certification criteria for sustainable practices, giving them access to programs and markets that value
309 responsible and environmentally friendly production. Therefore, it is vital to consider the immediate
310 economic value of timber and its importance in terms of ecosystem services and possible alternative
311 market opportunities.
312
313 CONCLUSIONS
314 The cocoa agroforestry systems in Alta Verapaz, Guatemala, are characterized by a wide diversity of
315 shade trees at different productive stages. The wood from these trees is primarily used for producing
316 firewood, thin planks, and posts. The results of this study suggest that these agroforestry systems have
317 promising potential as a viable strategy for timber production. In this way, they could significantly
318 contribute to improving the environmental and economic sustainability of the region while meeting the
319 demand for timber products from local families without resorting to destructive practices such as
320 deforestation. This information is crucial for designing strategies for the sustainable management and
321 utilization of CAFS in the department to increase productivity and, consequently, generate higher incomes
322 for the producers through a more significant contribution from the tree cover.
323
324 Author contribution
325 Conceptualization, C.E.V.-G. and B.L.; methodology, C.E.V.-G. and B.L..; software, J.A.R.-C.;
326 validation, B.L. Z.P., and C.V.; investigation, C.E.V.-G.; C.A.C.R. and H.E.B.K.; resources, E.E.V.L.;
327 data curation, J.A.R.-C.; writing—original draft preparation, C.E.V.-G.; M.K. and C.V.; writing—review
328 and editing, B.L.; Z.P. and M.K.; Visualization, J.A.R.-C.; supervision, B.L.; project administration, B.L.
329 and C.E.V.-G; funding acquisition, B.L. and C.E.V.-G.; All authors have read and agreed to the published
330 version of the manuscript.
331
332 Acknowledgements
333 This research had financial support from the budget assigned to The Internal Grant Agency of the Faculty
334 of Tropical AgriSciences, Czech University of Life Sciences Prague, Czech Republic (IGA FTZ, Project
335 No. 20223103) and Inter-American Institute for Cooperation on Agriculture (IICA), its Regional
336 Agricultural Research Consortia (CRIA) Programme.
337
338 REFERENCES

10 10
 Arrazate, C.H.A., G.M.S. Venero, A.M. López, M.M. Bolaños, J.R. Reyes, et al. 2021. Composición arbórea de
339
340 especies asociadas al cacao: selva Lacandona y sistemas agroforestales, Chiapas, México. Agronomía
341 Mesoamericana: 365–381. doi: 10.15517/am.v32i2.41630.
Barrios, E., V. Valencia, M. Jonsson, A. Brauman, K. Hairiah, et al. 2018. Contribution of trees to the
342
343 conservation of biodiversity and ecosystem services in agricultural landscapes View supplementary
344 material Contribution of trees to the conservation of biodiversity and ecosystem services in agricultural
345 landscapes. Int J Biodivers Sci Ecosyst Serv Manag 14: 1–16.
Braga, D.P.P., F. Domene, and F.B. Gandara. 2019. Shade trees composition and diversity in cacao agroforestry
346
347 systems of southern Para´, Brazilian Amazon. Agroforest Syst 93: 1409–1421.
Bukomeko, H., L. Jassogne, S.B. Tumwebaze, G. Eilu, and P. Vaast. 2019. Integrating local knowledge with
348
349 tree diversity analyses to optimize on-farm tree species composition for ecosystem service delivery in
350 coffee agroforestry systems of Uganda. Agroforest Syst 93: 755–770.
Bullock, E.L., C. Nolte, A.L. Reboredo Segovia, and E. Woodcock, Curtis. 2020. Ongoing forest disturbance in
351
352 Guatemala’s protected areas. Romote Sensing in Ecology and Conservation 2: 141–152.
Delgado-Vargas, I.A., W. Ballesteros Possú, and V. Arellano Chungana. 2022. Agrobiodiversidad de leñosas
353
354 multipropósito en sistemas productivos cafeteros. Revista de Investigación Agraría y Ambiental RIAA 13:
355 67–80.
Detlefsen, G., and E. Somarriba. 2012. Producción de madera en sistemas agroforestales de Centroamérica.
356
357 109: 2–426.
Esche, L., M. Schneider, J. Milz, and L. Armengot. 2023. The role of shade tree pruning in cocoa agroforestry
358
359 systems: agronomic and economic benefts. Agroforest Syst 97: 175–185. doi:
360 https://doi.org/10.1007/s10457-022-00796-x.
Espinoza, C.M.O., and J.O. Rangel. 2020. Composición florística y aspectos de la estructura de la vegetación en
361
362 sistemas agroforestales con cacao (Theobroma cacao L. - Malvaceae) en el departamento del Huila,
363 Colombia. Rev Acad Colomb Cienc Exactas Fis Nat 44(173): 1033–1046. doi: 10.18257/raccefyn.1183.
Garza, H.N., S.A. Santiago, M.A. Musálem, H.V. Lindeman, and Ma.A.P. Olvera. 2012. La diversidad de
364
365 especies útiles y sistemas agroforestales. Revista Chapingo Serie Ciencias Forestales y del Ambiente 18:
366 71–86.
Guerra-Centeno, D., J.C. Valdez-Sandoval, and D. Orozco. 2016. Guía para la identificación de especies de
367
368 árboles y arbustos comunes en el agropaisaje de Guatemala. Servi Pren. Guatemala.
Haggar, J., M. Asigbaase, G. Bonilla, J. Pico, and A. Quilo. 2015. Tree diversity on sustainably certified and
369
370 conventional coffee farms in Central America. Biodivers Conserv 24: 1175–1194.
Holdridge, L.R. 2000. Ecología basada en zonas de vida. Quinta rei. Instituto Interamericano de Cooperación
371
372 para la Agricultura, San José, Costa Rica.

11 11
 Ibrahim, M., M. Chacón, C. Cuartas, J. Naranjo, G. Ponce, et al. 2007. Almacenamiento de carbono en el suelo
373
374 y la biomasa arbórea en sistemas de usos de la tierra en paisajes ganaderos de Colombia, Costa Rica y
375 Nicaragua. Agroforestería de las Américas (45): 27–36.
Instituo de Investigaciones Económicas y Sociales (IDES) de la Universidad Rafael Landivar. 2012. Estudio de
376
377 potencial económico y propuesta de mercadeo territorial del Valle del Polochic, Departamento de Alta
378 Verapaz.
Jadán, O., S. Günter, B. Torres, and D. Selesi. 2015. Riqueza y potencial maderable en sistemas agroforestales
379
380 tradicionales como alternativa al uso del bosque nativo, Amazonia del Ecuador. Revista Forestal
381 Mesoamericana Kurú 12: 13–22.
López-Baez, O., S.I. Ramírez-Gonzalez, S. Espinosa-Zaragosa, J.M. Villareal-Fuentes, and A. Wong-Villareal.
382
383 2015. Diversidad vegetal y sustentabilidad del sistema agroforestal de cacao en la región de la selva de
384 Chiapas, Mexico. Revista Iberoamericana de Ciencias 2(2): 56–63.
Ministerio de Agricultura, Ganadería y Alimentación (MAGA). 2016. Plan Estratégico de la Agrocadena de
385
386 Cacao de Guatemala: 1–96.
Manaye, A., B. Tesfamariam, M. Tesfaye, A. Worku, and Y. Gufi. 2021. Tree diversity and carbon stocks in
387
388 agroforestry systems in northern Ethiopia. Carbon Balance Manag 16(1): 2–10. doi: 10.1186/s13021-021-
389 00174-7.
Matey, A., L. Zeledón, L. Orozco, F. Chavarría, A. López, et al. 2013. Composición florística y estructura de
390
391 cacaotales y parches de bosque en Waslala, Nicaragua. Agroforestería en las Américas: 61–67.
Maza, O.J., B. Torres, D. Selesi, D. Peña, C. Rosales, et al. 2016. Diversidad florística y estructura en
392
393 cacaotales tradicionales y bosque natural (Sumaco, Ecuador). Colombia Forestal 19(2): 129–142. doi:
394 10.14483/udistrital.jour.colomb.for.2016.2.a01.
Mejia-Rueda, H.G., G. Montañez-Acevedo, and C. Villamizar-Quiñonez. 2023. Descripción de características
395
396 sociales, económicas y tecnológicas del sistema de producción Cacao Theobroma cacao L en los
397 municipios de Zapatoca y Betulia, pertenecientes a la zona de influencia de la central hidroeléctrica
398 Sogamoso. Brazilian Journal of Animal and Environmental Research 6(1): 503–511. doi:
399 10.34188/bjaerv6n1-044.
Méndez, V.E., C.M. Bacon, M.B. Olson, K.S. Morris, and A. Shattuck. 2013. Conservación de
400
401 Agrobiodiversidad y Medios de Vida en Cooperativas de Café Bajo Sombra en Centroamérica. Revista
402 Científica de Ecología y Medio Ambiente 22: 16–24.
Morán-Villa, V.L., A.I. Monterroso-Rivas, J.D. Gómez-Díaz, S.R. Márquez-Berber, and E. Valdes-Velarde.
403
404 2022. Composición florística y arreglo de sistemas agroforestales de cacao en Tabasco, Mexico. Tropical
405 and Subtropical Agroecosystems 25: 2–19.

12 12
 Navarro Garza, H., A.S. Santiago, M.Á.M. Santiago, H.V. Lindemann, and Ma.A.P. Olvera. 2012. La
406
407 diversidad de especies útiles y sistemas agroforestales. Chapingo Serie Ciencias Forestales y Ambiente
408 18: 71–86.
Nicli, S., J. Mantilla-Contreras, R.W. Moya Fernandez, M. Schermer, D. Unger, et al. 2019. Socio-economic,
409
410 political, and institutional sustainability of agroforestry in Alta Verapaz, Guatemala. Journal of
411 Agriculture and Rural Development in the Tropics and Subtropics 120: 105–1117.
Niether, W., L. Armengot, C. Andres, M. Schneider, and G. Gerold. 2018. Shade trees and tree pruning alter
412
413 throughfall and microclimate in cocoa (Theobroma cacao L.) production systems. Ann For Sci 75(2): 38.
414 doi: 10.1007/s13595-018-0723-9.
Niether, W., J. Jacobi, W.J. Blaser, C. Andres, and L. Armengot. 2020. Cocoa agroforestry systems versus
415
416 monocultures: a multi-dimensional meta-analysis. Environmental Research Letters 15.
Núñez, H.E.H., H.J. Andrade, J.C.S. Salazar, J.R.S. A., D.R.G. S., et al. 2021. Almacenamiento de carbono en
417
418 sistemas agroforestales en los Llanos Orientales de Colombia. Rev Biol Trop 69(1): 352–368. doi:
419 10.15517/rbt.v69i1.42959.
Ordoñez Espinosa, C.M., and J.O. Rangel. 2020. Composición florística y aspectos de la estructura de la
420
421 vegetación en sistemas agroforestales con cacao (Theobroma cacao L.-Malvaceae) en el departamento de
422 Huila, Colombía. Ciencias Naturales 44: 1033–1046.
Patil, I. 2021. Visualizations with statistical details: The “ggstatsplot” approach. J Open Source Softw 6(61):
423
424 3167. doi: 10.21105/joss.03167.
Peralta-Rivero, C. 2022. Modelos productivos de desarrollo rural y su contribución en la generación de los
425
426 servicios ecosistematicos en Bolivia. Revista Grifos 32(59): 1–24. doi: 10.22295/grifos.v32i59.7147.
Pineda, P. 2022. Biodiversidad. In: Gálvez, J., editor, Perfil Ambiental de Guatemala. Cara Paren. Guatemala.
427
428 p. 3–48
Ramírez-Meneses, E. García-López, J.J. Obrador-Olán, O. Ruiz-Rosado, and W. Camacho-Chiu. 2013. Plan
429
430 diversity in cocoa agroforestry systems in Cárdenas, Tabasco, Mexico. Universidad y Ciencia 29: 215–
431 230.
Ramos-Prado, J.M., E. Romero-Hernández, P. Sánchez-Morales, D. Jiménez-García, and E. Hipólito-Romero.
432
433 2023. Dimensiones bioculturales y socioeconómicas de la sustentabilidad en sistemas agroforestales
434 diversificados con cacao y vainilla. Rev Mex De Cienc Agric 14: 401–412. doi:
435 https://doi.org/10.29312/remexca.v14i3.3093.
Roa-Mero, H.A., M.G. Salgado-Mora, and J. Alvarez-Herrera. 2009. Análisis de la Estructura Arbórea del
436
437 Sistema Agroforestal de Cacao (Theobroma cacao L.) en el Soconusco, Chiapas-México. Acta biol.
438 Colomb. 14: 97–110.

13 13
 Ruiz Solsol, H., G.G. Rivas, and I.A. Gutiérrez Montes. 2014. Huertos Familiares: Agrobiodiversidad y su
439
440 aporte en la seguridad alimentaria en territorios rurales de Guatemala. Agroecología 9: 85–88.
Salvador-Morales, P., J.L. Martínez-Sanchez, L. Cámara Cabrales, and C. Zequeira Ramos. 2020. Estructura y
441
442 carbono específico en una cronosecuencia de sistemas agroforestales de Teobroma cacao L. en Tabasco,
443 México. Madera y Bosques 26: 1–15.
Sánchez Gutiérrez, F., J. Pérez Flores, J.J. Obrador Olan, Á. Sol Sánchez, and O. Ruiz-Rosado. 2016a. Tree
444
445 structure of cocoa agroforestry system in Cárdenas, Tabasco, Mexico. Rev Mex De Cienc Agric 14: 2695–
446 2709.
Sánchez Gutiérrez, F., J. Pérez-Flores, J.J. Obrador Olan, Á. Sol Sánchez, and O. Ruiz-Rolando. 2016b.
447
448 Árboles maderables en el sistema agroforestal de cacao en Cárdenas, Tabasco, México. Rev Mex De
449 Cienc Agric 14: 2711–2723.
Sibelet, N., K.E. Posada, and I.A. Gutiérrez-Montes. 2019. Agroforestry systems provide firewood for
450
451 livelihood improvement in Guatemala. Bois et Forêts des Tropiques 340: 91–102.
Sol-Sánchez, Á., S. Alexander, López-Juárez, V. Córdova-Ávalos, and F. Gallardo-López. 2018. Productividad
452
453 potencial del SAF cacao asociado con árboles forestales. Revista Iberoamericana de Bioeconomia y
454 Cambio Climático 4: 862–877.
Somarriba, E. 2004. ¿Cómo evaluar y mejorar el dosel de sombra en cacaotales? Agroforestería de las Américas
455
456 41/42: 122–130.
Soncim, I.Q., E. de Melo Virginio Filho, C.A. Righi, and R. Shirota. 2021. Rentabilidad económica de sistemas
457
458 agroforestales con café: estudio de largo plazo en Turrialba, Costa Rica. Agroforestería en las Américas
459 51(177–184).
Sonwa, D.J., S.F. Weise, B.A. Nkongmeneck, M. Tchatat, and M.J.J. Janssens. 2017. Structure and composition
460
461 of cocoa agroforests in the humid forest zone of Southern Cameroon. Agroforestry Systems 91(3): 451–
462 470. doi: 10.1007/s10457-016-9942-y.
De Sousa, K., G. Detlefsen, E. de M. Virginio Filho, D.E. Tobar L, and F. Casanoves. 2015. Timber yield from
463
464 smallholder agroforestry systems in Nicaragua and Honduras. Agroforest Syst 90: 207–218.
Sousa, K. De, G. Detlefsen, E. de M. Virginio Filho, D.E.T. L, and F. Casanoves. 2015. Timber yield from
465
466 smallholder agroforestry systems in Nicaragua and Honduras. Agroforestry Systems 90(2): 207–218. doi:
467 10.1007/s10457-015-9846-2.
Standley, P.C., and J.A. Steyermark. 1946. Flora of Guatemala. Botany 24.
468
Suarez-Venero, G.M., C.H. Avendaño-Arrazate, P.A. Ruiz-Cruz, and P. Estrada-de-los-Santos. 2019.
469
470 Estructura e impacto de la diversidad taxonómica en cacao del Soconusco, Chiapas, México. Agronomía
471 Mesoamericana 30: 353–365. doi: 10.15517/am.v30i2.34032.

14 14
 Suatune, P., E.J. Somarriba, C.A. Harvy, and B. Finegan. 2003. Composición florística y estructura de bosques
472
473 y cacaotales en los territorios Indígenas de Talamanca, Costa Rica. Agroforestería de las Américas 10: 37–
474 38.
Tapia Vera, C.A., F.D. Sanchez-Mora, G.H. Vásconez-Montúfar, A.E. Barrera-Alvarez, R. V. Mora-Yela, et al.
475
476 2021. Evaluación temporal de sistemas agroforestales de cacao en el trópico húmedo ecuatoriano. Revista
477 Bionatura 6(4): 2295–2299.
Tscharntke, T., C. Ocampo-Ariza, J. Vansynghel, B. Ivañez-Ballesteros, P. Aycart, et al. 2022. Socio-ecological
478
479 benefits of fine-flavor cacao in its center oforigin. Society for Conservation Biology 16(1): 1–10.
Vebrova, H., B. Lojka, T.P. Husband, M.E.C. Zans, P. Van Damme, et al. 2014a. Tree diversity in cacao
480
481 agroforests in San Alejandro, Peruvian Amazon. Agroforest Syst 88: 1101–1115.
Vebrova, H., B. Lojka, T.P. Husband, M.E.C. Zans, P. Van Damme, et al. 2014b. Tree diversity in cacao
482
483 agroforests in San Alejandro, Peruvian Amazon. Agroforestry Systems 88(6): 1101–1115. doi:
484 10.1007/s10457-013-9654-5.
Villanueva-González, C.E., J.A. Ruiz-Chután, K. Marie, R.W. Moya Fernandez, C. Villanueva, et al. 2023.
485
486 Botanical diversity, structure and composition in cocoa agroforest systems in Alta Verapaz, Guatemala.
487 Scientia Agropecuaria 14(2): 223–234. doi: https://doi.org/10.17268/sci.agropecu.2023.020.
Villareyna, R., J. Avelino, and R. Cerda. 2020. Adaptación basada en ecosistemas: efecto de los árboles de
488
489 sombra sobre servicios ecosistémicos en cafetales. Agronomía Mesoamericana 31: 499–516. doi:
490 doi:10.15517/am.v31i2.37591.
Wickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York.
491
492
493
494 Tables and Figures
495
496
497
498
499

15 15
500
501 Figure 1. Location of the study area within the department of Alta Verapaz, Guatemala.
502
503 Table 1. Average value of dendrological variables and number of families and species registered in cocoa
504 agroforestry systems, Alta Verapaz, Guatemala.
CAFS Height BA ( Density ( Number of Taxa
Age # (m ) 2
m ha )
−1
Trees ha )
−1
Families Species
3 2 11.36 9.5 202 7 10
4 1 11.12 7.9 364 3 3
7 3 12.78 3.7 117 11 13
8 2 13.56 6.5 128 7 9
10 1 12.17 13.5 212 5 6
11 2 10.95 5.1 126 7 10
12 3 10.22 5.8 208 13 22
13 1 8.58 2.2 72 3 4
14 1 11.06 3.2 124 6 8
20 1 14.23 3.2 60 3 4
25 3 14.43 10.5 196 14 23
505
506
507
508
509
510
511

16 16
512
513 Figure 2. Frequency description and Chi-square test of species distribution by the age of cocoa
514 agroforestry systems among the study localities.
515

17 17
516
517 Figure 3. Distribution of the basal area of each species by the age of the cocoa agroforestry systems
518 studied in the localities of Alta Verapaz, Guatemala.
519

520

18 18
521 Figure 4. Distribution of the basal area (BA) in the cocoa agroforestry systems studied in the localities of
522 Cahabón, Cobán, Lanquín, and Panzós, Alta Verapaz, Guatemala.
523

524
525 Figure 5. Distribution of the volume of each species by age of the CAFS between study locations.
526
527 Supplementary Table S1. List of tree species and their abundance associated with cacao (Theobroma
528 cacao L.) agroforestry systems in Alta Verapaz, Guatemala.

529
Family Species Common name Abundance Sites
Fabaceae Gliricidia sepium (Jacq.) Kunth. Madre cacao 290 4
Meliaceae Swietenia macrophylla G.King Caoba 158 4
Fabaceae Inga sapindoides Willd. Cuje 63 4
Meliaceae Cedrela odorata L. Cedro 54 4
Cordiaceae Cordia alliodora (Ruiz & Pav.) Oken Laurel 42 2
Burseraceae Protium copal (Schltdl. & Cham.) Engl. Copal Pom 27 2
Vochysiaceae Vochysia guatemalensis Donn.Sm. San Juan 24 3
Urticaceae Cecropia peltata L. Guarumo 23 4
Malvaceae Theobroma bicolor Humb. & Bonpl. Pataxte 18 2

19 19
 Lauraceae Persea americana Mill. Aguacate 13 3
Bignoniaceae Roseodendron donnell-smithii (Rose) Miranda Palo Blanco 13 3
Malpighiaceae Byrsonima crassifolia (L.) Kunth Nance 12 3
Anacardiaceae Mangifera indica L. Mango 10 3
Rutaceae Citrus ×aurantium L. Naranja 9 1
Sapotaceae Pouteria sapota (Jacq.) H.E.Moore & Stearn Zapote 9 3
Fabaceae Inga paterno Harms Paterna 7 3
Burseraceae Bursera simaruba (L.) Sarg. Palo de Jiote 5 2
Calophyllaceae Calophyllum brasiliense Cambess. Santa María 4 1
Sapotaceae Manilkara zapota (L.) P.Royen Zapotón 4 2
Malvaceae Ochroma pyramidale (Cav.) Urb. Palo Balsa 4 2
Lauraceae Persea schiedeana Nees Coyou 4 2
Myrtaceae Pimenta dioica (L.) Merr. Pimienta gorda 4 1
Fabaceae Platymiscium dimorphandrum Donn.Sm. Hormigo 4 1
Fabaceae Vatairea lundellii (Standl.) Killip Medallo 4 2
Fabaceae Cojoba arborea (L.) Britton & Rose Cola de Coche 3 3
Pinaceae Pinus caribaea Morelet Pino Blanco 3 2
Fabaceae Tamarindus indica L. Tamarindo 3 1
Annonaceae Annona muricata L. Guanabana 2 1
Annonaceae Annona squamosa L. Anona 2 1
Anacardiaceae Anacardium occidentale L. Jocote de marañon 1 1
Moraceae Artocarpus altilis (Parkinson) Fosberg Mazapan 1 1
Fabaceae Cassia grandis L.f. Mucut 1 1
Fabaceae Dalbergia stevensonii Standl. Rosul 1 1
Fabaceae Enterolobium cyclocarpum (Jacq.) Griseb. Puntero 1 1
Fabaceae Leptolobium panamense (Benth.) Sch.Rodr. & Chichipate 1 1
A.M.G.Azevedo

Magnoliaceae Magnolia mexicana DC. Palo de peña 1 1

Anacardiaceae Spondias purpurea L. Jocote de mico 1 1
Bignoniaceae Tabebuia rosea (Bertol.) DC. Matilisguate 1 1
530
531

20 20