Search

Agroforestry systems for cocoa production are commonly promoted for biodiversity conservation, climate change mitigation and adaptation as well as for food security and risk mitigation. Generally, these systems include timber, legume or fruit trees. Sucessional or dynamic agroforestry systems represent a special type of design and management approach, using high densities and diversity of trees and crops occupying different strata and with varying life cycles. Here we present results on absolute and potential yields of 3 organic cocoa production systems entering the mature stage from a long-term trial in Bolivia: A complex successional dynamic agroforestry system (SAFS), a simpler but diversified agroforestry system (AF ORG) and a cocoa monoculture (MONO ORG). Cocoa yields were highest in MONO ORG, followed by the AF ORG, and lowest in SAFS. Total system yields in both types of agroforestry systems (dry matter) were 3.5 to 4 times higher than in monocultures over the 11 years. This was mainly due to banana production in AF ORG and from a multitude of by-crops in SAFS. In mature SAFS peach palm was the by-crop with the highest dry matter production. The results demonstrate how different crops can be associated with cocoa while still reaching elevated cocoa yields. The potential of agroforestry systems to sustainably intensify production on one surface is high and could still be optimised with the use of improved and locally selected varieties of cocoa and fruit trees. Different designs of agroforestry systems can contribute to diversification of diets and incomes of producing families and regions.

Agroforestry systems offer many benefits, such as diversifying incomes and reducing input costs. However, planning and managing diversified systems is knowledge intensive, and needs constant observation and adaptation to specific contexts by the farmers. Therefore, sharing knowledge between farmers, technicians and researchers critical for adoption of agroforestry systems that fit the respective situation and objectives. The dynamic agroforestry community has developed some principles that can be applied for regenerating degraded land with agroforestry systems by combining a variety of species, focussing on diversity and synergies instead of simplification and competition. These concepts, together with practical experiences, and research results are a complex set of information to be complemented with local knowledge of farmers (e.g. on species’ use, available resources, local markets). Here we share experiences from elaborating training materials and conducting a training of trainers based on research findings on cocoa agroforestry and practical experiences from working with dynamic agroforestry systems and organic cocoa for more than 25 years. We used a set of training materials and methods for awareness raising for the environmental and social benefits of agroforestry systems, for example simplified research methods like infiltration measurements. Some practical visualizations e.g. for erosion on surfaces after slash-and-burn vs. slash-and-mulch, or dry run exercise for the installation and development of dynamic agroforestry plots. These methods help to understand processes that are happening on larger spatial or temporal scale. Field visits to research plots as well as farmers fields and discussions based on associated data were central. In conclusion, sharing of scientific results combined with practical and visual experiences give technicians the confidence to multiply the experiences with their farmers. While demoplots are crucial for working with farmers, additional videos describing the design and management of agroforestry systems, farmers’ testimonials, and some research results directly from the field can be helpful.

Cacao agroforestry systems offer a wide range of ecosystem services beside food and timber production, such as biodiversity, carbon sequestration, microclimate regulation, water infiltration etc. which makes them interesting for farmers as well as for society. In many cases the mid to long-term benefit for farmers from the shade trees is not known. To assess the added value of timber and fruit trees, 16 smallholder cacao agroforestry fields in Alto Beni, Bolivia were selected in 2017 with the criteria of having dynamic agroforestry or diverse agroforestry systems implemented. Fields were georeferenced and tree inventories established. Quantitative and qualitative data of the timber trees were taken. The farmers were interviewed concerning the fruit tree productivity and use of the harvest. The regulative governmental entity for timber tree management was interviewed for assessment of the timber value chain. The cacao agroforestry plantations have an average age of 18 years. For the total of 2’941 timber trees the species were identified (72), 20% of it were Swietenia macrophylla, which makes it the most popular timber species. Followed by Myroxylon balsamum, Amburana cearensis and Centrolobium ochroxylum. The average timber tree density was 154 trees/ha and the standing timber volume was 38.8 m³/ha. The value of the standing timber was evaluated in average of 12’947 USD/ha. This value needs to be reduced due to lack of professional timber processing. Most numerous fruit are banana, followed by the fruit trees of Bactris gasipaes, Oenocarpus bataua, citrics and Garcinia gardneriana.There is a marked difference in fruit tree density according the agroforestry system, dynamic agroforestry systems harbor much more fruit trees compared to the simpler agroforestry systems which have the focus on timber. With the aim to increase farmer’s income from timber as well from fruit trees the value chains need to be strengthened.

Cocoa agroforestry systems, combining cocoa with timber and fruit trees, are mainly promoted for their environmental benefits. Knowledge gaps exist about the economic performance of such systems. Here we present the productivity and income of 5 cocoa production systems from planting to entering the mature stage 11 years).

Cocoa (Theobroma cacao) is commonly produced in full-sun monoculture cropping systems to increase yields in the short term. Nevertheless, cocoa is a suitable crop for production under shaded conditions and is traditionally cultivated in agroforestry systems in Latin America. To ensure productivity and profitability, however, the development of best practices for shade management is crucial, but shade tree pruning is not commonly practised. This study investigates the influence of pruning shade trees in cocoa-based organic agroforestry systems in Bolivia on agronomic and economic performance. Four organic agroforestry farms were selected, where shade trees were not pruned for at least 10 years. At each site, half of the plot was kept unpruned and the other half was pruned, while all other management practices were kept equal. Data on yield formation were collected subsequently for two harvesting seasons. The trial results show a significant increase in cocoa yield under pruning conditions ranging from 28 % to 82 % compared to unpruned plots. This is attributed to an increase in flowering and fruit set in pruned plots. No differences in the incidence of pests and diseases in the pods were found between both treatments. To evaluate, whether pruning shows an economic benefit for farmers, different scenarios of yield increase based on the minimum, average, and maximum yield of local cocoa producers were used. Other sources of income, such as by-crops, were not considered in the calculations. For the average yield level of 287.4 kg ha−1 (dry), an increase of 51 % in two consecutive years will cover the pruning costs. Despite the promising results and indication, that the yield increase will last for more than two years, the initial costs for pruning might still discourage farmers. Therefore, financing programs for farmers that support agroforestry tree pruning are necessary to increase both cocoa production and farmer’s income.

Agroforestry systems for cocoa production are commonly promoted for biodiversity conservation, climate change mitigation and adaptation as well as for food security and risk mitigation. Generally, these systems include timber, legume or fruit trees, and sometimes additional crops. Knowledge gaps exist about the economic performance of cocoa based agroforestry systems, including the by-crops. Here we present the economic performance of 5 cocoa production systems from planting to entering the mature stage (11 years). In a long-term trial in Bolivia, a gradient of complexity from monocultures, agroforestry systems to successional agroforestry systems (SAFS) is studied. Additionally, for monocultures and agroforestry, conventional and organic management are compared, while SAFS are managed organically. Income was calculated taking into account yields of cocoa, fruit trees and by-crops with farm gate prices. Only for cocoa organic premium prices were reached and taken into account. Labour time was registered for management, input preparation and post-harvest. Cocoa yields were lower in the agroforestry systems compared with monocultures, and lowest in SAFS. For monocultures, they were higher under conventional management, while in agroforestry systems management had no influence. Total system yields in agroforestry systems (dry matter) were 3–4 times higher than in monocultures. This was mainly due to banana production in agroforestry systems and from a diversity of by-crops in SAFS. Income over all years was comparable among all systems. In agroforestry systems, cocoa was responsible for more than 50 % of the income, while in SAFS the share of cocoa was smaller, as some crops like pineapples or peach palm had good markets in the region. The income generated per workday invested during the whole period did not differ between the systems. Income analysis shows the importance of cocoa as a cash crop, but also the potential of by-crops depending on the development of their markets. On the other hand, agroforestry systems contribute to food security and mitigate risks of price or yield fluctuations in the cash crop. In conclusion, the data show that with different strategies and plantation design, the same level of income and income per work day invested can be reached. More information on the trial layout can be found here: https://systems-comparison.fibl.org

Unlike cacao monocultures, cacao agroforestry systems offer a wide range of additional ecosystem services and values to the farmers, such as (i.a. regulating pests and diseases, balancing the understory climate and carbon sequestration) and enable a higher independence from the main crop by generating a potential additional income. To assess the standing timber volume and value, a tree inventory was conducted in 2017 in 16 smallholder cacao agroforestry plots in Alto Beni, Bolivia. Farmers and experts were interviewed to identified the challenges for the timber production in these agroforestry systems. The timber trees on the plots had mainly an age of about 10 20 years (mean 15.5 years) but some trees were also in the range between 2–40 years. A total number of 2’941 trees were counted on all the plots and 20 % of it were Swietenia macrophylla, which makes it the most popular timber species. Other very common species were Myroxylon balsamum (12 %), Amburana cearensis (11 %) and Centrolobium ochroxylum (10 %). The average timber tree density was 230 trees/ha and the standing timber volume was 46 m3/ha. The standing timber per plot had an estimated average value of 12’947 USD/ha at the time. Because of lack of professional timber processing, such as timber transportation and sawmill, the loss in timber volume is estimated around 40 %. Additionally, farmers are challenged with trimming and pruning of trees as well as with the legal requirements. With the aim to increase farmer’s income from timber trees we suggest the following measures at three levels: (1) improving plantation layout (density, layout, species) and tree management (criteria for selection for cut off trees, trimming and pruning); (2) to support a more professional timber logging and processing to decrease losses and (3) to create service providers such as farmer owned cooperatives for logging, sawing, registration of trees and logging permits.

La coca es uno de los cultivos más antiguos en los Yungas de La Paz. Hoy en día, en su mayoría se planta en monocultivos y muchas veces en laderas inestables. El monocultivo de coca degrada los suelos, causa deslizamientos de terrenos, pérdida de biodiversidad y se desarrolla con el uso de agroquímicos potencialmente dañinos para la salud humana, los suelos y los cuerpos de agua. Sin embargo, en los Yungas persiste también un conocimiento ancestral sobre formas más sustentables de producción de coca con cultivos diversificados con árboles, arbustos y hierbas de diferentes usos. En este estudio se identificaron iniciativas y experiencias de recuperación de conocimientos ancestrales del cultivo sostenible de la coca, que integran técnicas de la agroforestería dinámica. La creciente demanda para coca orgánica puede ser una oportunidad para incentivar cultivos más sustentables y adaptados al cambio climático, que al mismo tiempo contribuyan a la seguridad alimentaria de las familias. Actualmente, el cultivo de coca con certificación orgánica en agroforestería dinámica ya representa una vía para más de 300 familias cocaleras en las Yungas de La Paz, y en miras de la degradación de suelos y la alta demanda de nuevas tierras para cultivar coca que está amenazando las áreas protegidas, es un potencial económico y ecológico para las más de 13 500 familias cocaleras de los Yungas paceños. Sin embargo, el cultivo de la coca en agroforestales dinámicos necesita apoyo técnico e institucional en el manejo de los sistemas agroforestales, en la investigación y en la comercialización tanto de la coca agroecológica como otros productos agroforestales., Acta Nova, 8 (4), ISSN:1683-0768

To estimate the cocoa dry bean weight from the fresh weight a standard dry bean factor (DBF) ranging from 32% to 40% is usually applied. This factor, i.e. the relation between the dry bean weight (fermented and dried) and the fresh weight at pod opening, is used by cooperatives and factories that buy the fresh beans of the farmers. However, this factor is subjected to seasonal fluctuations and can be highly influenced by rainfall. Therefore, for instance, higher values are expected during the dry season compared with the rainy season. The objective of this study was to evaluate the accuracy of the use of a standard fix DBF value compared with the use of actual DBF values periodically collected over the season for assessing the performance of 12 cocoa cultivars (clones and hybrids). We hypothesised that using the same DBF for different cultivars that may have the production peak at different times of the season can lead to over- or underestimations of the yield performance. The study was performed in a long-term cocoa trial in Alto Beni, Bolivia, which comprises full-sun monocultures, agroforestry and successional agroforestry productions systems. In 2014, for each harvest date, cultivar and production system, a subsample of the fresh beans was collected, put in a mesh bag and weighted. After the fermentation and drying process, the subsamples were weighted again and the DBF was then calculated at 8% water content of the beans. The total performance of each cultivar at the end of the harvest season was estimated using the actual DBF for each cultivar, system and harvest date, and the results were compared with the performance estimated using the DBF values averaged across systems, harvest dates and varieties and also with the fixed DBF of 35%. The results only showed minor differences in the overall performance of each cultivar when comparing the different methods used. However, for all the cultivars, using the 35% fix value gave the poorest estimation compared with any of the other DBF tested, mainly due to underestimation of the performance of the early maturing cultivars.

Cocoa production systems can vary from a mono-cropping (MC) full sun plantation to a highly diversified successional agroforestry (SAFS). Mono-cropping is still the most common one but recently agroforestry (AF) systems have raised more interest because of their expected long term resilience. Nevertheless, information about the influence of cropping systems and management methods such as organic versus conventional in cocoa production and phenology in South America is missing. This study is based on the SysCom program’s framework located in Alto Beni (Bolivia), where three cropping system (MC, AF and SAFS) are compared. Additionally, MC and AF systems are conventionally and organically managed. A total of 12 cultivars including local clones, foreign clones (from the Imperial College Selections and Trinidad Selections) and hybrids are planted in each system. The aim of the present study is to identify differences between systems, management methods, cultivars and the respective interactions on cocoa phenology. Data on flushing and number of flowers, small, medium and big fruits and cherelles are being recorded every 15 days from March 2015. The trial was designed according to a split plot design with four replicates. Principal component analysis (PCA) and cluster analysis were performed on the cultivar level. Additionally, the collected data will be statistically analysed with a generalised mixed model (Poisson distribution) to test the effect of the different cultivars, systems and their interactions on the phenological traits. Preliminary results of the PCA and cluster analysis do not show a clear grouping pattern between the three types of genotypes analysed. This suggests that there are differences between cultivars but they are not consistent within the three groups. Preliminary descriptive results of the collected traits show differences between single cultivars and systems, but not between managements (organic vs. conventional). Some of the traits seem to be more system dependent than others. For instance, the number of cherelles seems to be higher in the SAFS than in the other systems. Additionally, interactions between systems and cultivars are expected to be present. This would be positive in terms of potentiality for selection of cultivars adapted to each different system.

Cocoa cropping systems vary from full-sun monocrops to highly diversified agroforestry systems, which combine the cocoa production with by-crop such as bananas or plantains and trees with different functions. Additionally, these systems can be organically or conventionally managed. Although labour demand is one of the main activities contributing to the total production costs of the plantations and it may highly differ between production systems, there are still scarce data and information available. In this study we present the labour demands of five different production systems in a newly established cocoa plantation in Alto Beni, Bolivia. The trial was established in 2008 within the framework of the SysCom programme (www.systems-comparison.fibl.org) and comprises monoculture and agroforestry systems under organic and conventional management and one successional agroforestry system with organic management, each one replicated four times. From 2009 to 2013, the establishment phase of the plantation, the time spent in all the farming activities, mainly weeding, pruning, harvesting and fertilising was recorded separately for each plot. Overall, total labour demand was higher in the successional and agroforestry systems compared with the monocultures, mainly due to the time devoted to the management of the multifunctional trees and by crops. However, no significant differences were found between organic and conventional management under both agroforestry and monoculture systems. On the contrary, the time needed for pruning the cocoa trees was higher in the monocultures. Similarly, the time spent in cocoa harvesting was also higher in the monocultures, especially under conventional management, which well correspond to the yield obtained. Weeding was high time demanding the first years of the plantation but decreased with the time. However, no differences between the monoculture and agroforestry systems and between the organic and conventional management were found. An economic analysis of the gross margins for each system at plot level, including the input costs (herbicides, fertilisers, etc.) and the associated cost for their preparation (e.g. compost preparation), and the income generated from the harvested produces (cocoa, banana, plantain and other by-crops out of the highly diversified system ) is under evaluation.

Cocoa (Theobroma cacao L.) trees have a lifespan of up to 100 years in their natural environment, the lower strata of tropical alluvial forests. Sustainability of global cocoa production is at stake due to the deterioration of soil fertility, high losses due to pests and diseases (P&D) and old tree stock. Producing cocoa in shaded, low input agroforestry systems offers a potential solution for achieving long-term productivity and sustainability. However, shade-tolerant cocoa clones are needed which are productive and resistant to P&D, as well as meeting the quality standards requested by the industry. In Bolivia, the cocoa mirid (Monalonion dissimulatum) and the novel disease Frosty Pod Rot (“Monilia” Moniliophthora roreri) are among the most devastating P&D. The Research Institute of Organic Agriculture (FiBL) and its local partners have been conducting cocoa production systems research in the Alto Beni region of Bolivia since 2008. A participatory rural appraisal revealed that indeed, how to reduce incidences of the cocoa mirid and Frosty Pod Rot was the top priority of cocoa farmers. Furthermore, a need for evaluating the productivity of different cocoa clones and for the documentation of best practices was expressed. In order to address these points, several research activities were carried out: 1. 16 cocoa clones have been evaluated for productivity and susceptibility to P&D for four consecutive years (2010—13) in on-farm trials at multiple locations. 2. The yield development and prevalence of P&D in the fields of four outstanding farmers was assessed for two years (2012–13) Data analysis revealed that some locally selected clones (elite tree selection program) were not only among the most productive, but also showed earliest maturation and some degree of resistance to Monilia (i.e. with lower incidences of the disease). Thus these clones hold the potential to escape attacks by P&D which occur later on in the season and may serve as a basis for further germplasm development towards clones with resistance to Monilia. However, before disseminating information to farmers and farmers’ organisations, sound scientific data from at least five years is required.

Cocoa is a crucial export commodity for many developing countries and provides income for millions of smallholders. However, cocoa cultivation has resulted in habitat destruction, biodiversity loss and soil degradation. While much of the world’s cocoa is produced in arguably unsustainable full-sun monoculture systems, shaded agroforestry systems may be an alternative for sustainable cocoa production. However, data-based information on advantages and limitations of different cocoa production systems are limited and pairwise comparisons on the long-term performance of cocoa monocultures and agroforestry systems under conventional and organic management are literally inexistent. The Research Institute of Organic Agriculture (FiBL) is pioneering to fill this knowledge gap with a unique long term field trial in tropical Bolivia. The trial was established in 2008 and consists of six systems: two monocultures (MONO CONV/ORG) and two agroforestry systems (AF CONV/ORG) under conventional and organic management, one successional agroforestry system (SAFS, organic only) with dynamic shade management, and a fallow system of the same age serving as a reference for biodiversity and soil fertility studies. The systems aim to represent current smallholder cocoa farmers’ practices. Parameters such as the tree development, yield of cocoa and by-crops, incidences of pests and diseases, soil fertility, carbon stocks, nutrient balances, economic data and biodiversity are regularly assessed. Five years after planting, results showed significantly shorter tree circumference (18% and 33 %) in AF systems and SAFS, respectively, compared to MONO systems. Tree circumference correlated strongly with cocoa dry bean yield which was, as expected, highest in MONO CONV (603 kg ha−1). By-crops such as plantain, cassava, pineapple, etc. were harvested in AF systems and SAFS, which may compensate for lower cocoa yields in the first years. Future research will investigate cocoa performance after the establishment phase and thus provide indications on the long-term sustainability of the different systems.

The cocoa mirid (Monalonion dissimulatum) is one of the major pests in cocoa cultivation in Alto Beni, Bolivia, causing up to 50% yield losses. The most common control method is the manual removal of the nymphs sucking at the pods in the morning. This practice is very time-consuming, and farmers often do not dispose of enough time to perform these control measures. Therefore, there is an urgent need for more efficient practices to control this important pest. The present study evaluated the pathogenicity of two strains of Beauveria bassiana for the control of Monalonion dissimulatum on cocoa: a non-native, commercialized strain (Probiobass MR, Probiotec S.R.L.), and a native strain of Alto Beni, which is not yet commercialized (isolated by the local association SIEMPRE). Moreover, a silicon-based product (TECSIL PM®) was tested. In addition, the effect of different degrees of infestation with Monalonion dissimulatum on different stages of cocoa pod development was examined. In order to investigate these questions, several different field trials were carried out at the experimental station of Sapecho between June and September 2013. The foreign strain of Beauveria Bassiana was the most effective bio-pesticide, causing a mortality rate of 63.3% in adults and 49.1% in nymphs of Monalonion dissimulatum. Cocoa pods in their early stages of development were highly susceptible to attack by Monalonion dissimulatum. On the other hand, fully developed cocoa pods showed a rather strong resistance to attacks by Monalonion dissimulatum: up to about 70% of damaged tissue on the surface of the cocoa pods, no effects on cocoa wet bean yield were recorded. However, when the damage increased above 70%, it had a strong impact on yield, amounting up to 50.4% yield loss. Fully damaged pods started drying out, and if they were completely desiccated, the yield loss was total. It is concluded that the foreign strain of Beauveria Bassiana may be the most efficient to control Monalonion dissimulatum in the field. More on-farm field trials need to be conducted over longer time periods in order to elucidate whether the observed effects will be reflected in higher cocoa yields in the farmer’s context. It is recommended to perform harvesting operations at regular intervals of two weeks in order to minimize losses caused by the desiccation of damaged cocoa pods.

Cocoa is one of the most important export commodities for many developing countries and provides income for millions of smallholders. The expansion of cocoa production has resulted in habitat destruction, biodiversity loss, and soil degradation. The prevalent cocoa production systems worldwide are conventional monoculture full sun systems. Agroforestry systems are argued to be a viable strategy for sustainable cocoa production. However, data-based information on advantages and limitations of different cocoa production systems is limited. Pairwise comparisons on the long-term performance of cocoa monocultures and agroforestry systems under conventional and organic management are inexistent. FiBL is pioneering to fill this knowledge gap with a unique long-term field trial in tropical Bolivia established in 2008. The trial consists of six treatments: two monocultures (MONO CONV/ORG) and two agroforestry system (AF CONV/ORG) under conventional and organic management, one organic successional agroforestry system (SAFS) with dynamic shade management, and a fallow of the same age serving as a reference for biodiversity and soil fertility studies. The treatments are representative for current cocoa production systems of smallholders. Parameters regularly assessed include canopy openness, cocoa stem diameter and bean yield, pests and diseases, soil fertility, carbon stocks, economic data and biodiversity. Five years after planting, results showed significantly shorter tree circumference (18% and 33%) in AF systems and SAFS, respectively, compared to MONO systems. Tree circumference correlated strongly with cocoa bean yield, and highest bean yields were recorded in MONO CONV as expected. Additional products like banana/plantain, cassava, pineapple, etc. were harvested in AF systems and SAFS, which may compensate for lower cocoa yield in the first years. First results indicate that disease incidences were higher in MONO systems compared to AF and SAFS. Future research will investigate cocoa performance after the establishment phase and thus provide indications on the long-term sustainability of the different systems.

In Malaysia an intensively managed high input full sun cacao (Theobroma cacao L.) plantation can reach high yields of 1.5 to 2 t dry beans per ha. Following a high production period of 10 years the yields often decrease markedly. Focusing on leading depleted cacao plantations sustainably back to full production a research project was initiated in June 2011 using a diversification approach with agroforestry systems. The experimental site is located on a large commercial farm in the humid tropical lowlands of peninsular Malaysia, in the region of Kuala Lipis. In a field trial with a strip-split-plot design, three different production systems are compared under two tree age conditions: newly planted and old rehabilitated cacao trees, after the removal of the original canopy back to the leader structure. The examined production systems are: i) high external inputs in a mono crop full sun system representing the common practice of large cacao plantations in South-East Asia (COM); ii) medium level of external inputs in an agroforestry system of low diversity focusing on leguminous and timber trees (AF LD); iii) low external inputs in an agroforestry system of high diversity and high density shade trees including annual crops and fruit trees (AF HD). The existing twenty-two-year-old plantation with the original canopy and a high input level serves as control treatment (REF). When cocoa yields decline after the initial high production period trees are often replanted. This results in a non-productive phase lasting several years before the young trees start to develop pods and even longer before yields reach a remunerative level. Rehabilitating old low producing trees on the other hand, as practiced in the present experiment, is expected to re-establish higher yields more quickly than re-planting. The first full harvest in the trial started in September 2012, 15 months after the rehabilitation pruning. Between October 2012 and June 2013 (main harvest) an average of 471.2 kg dry beans per ha were harvested in the control treatment. The yields of the common practice treatments already amounted to 41.6 % of the control. This is a very promising result, especially in view of the development of young trees which will take at least another year before the first pod development. Yields in the agroforestry systems increased less quickly as tree development under shade and with lower fertilizer input is inherently slower.

In Malaysia an intensively managed high input full sun cocoa plantation can reach high yields of 1.5 to 2 t dry beans per ha. Following a high production period of 10 years the yields often decrease markedly. Focusing on leading depleted cacao plantations sustainably back to full production a research project was initiated in June 2011 using a diversification approach with agroforestry systems. The experimental site is located on a large commercial farm in the humid tropical lowlands of peninsular Malaysia, in the region of Kuala Lipis. In a field trial with a strip-split-plot design, three different production systems, mainly characterised by diversification levels (mono culture to high diversity agroforestry) and accordingly different external input levels (high to low), are compared under two tree age conditions: newly planted and old rehabilitated cocoa trees, after the removal of the original canopy back to the leader structure. The existing twenty-two-year-old plantation with the original canopy and a high input level serves as control treatment. When cacao yields decline after the initial high production period trees are often replanted. This results in a non-productive phase lasting several years before the young trees start to develop pods and even longer before yields reach a remunerative level. Rehabilitating old low producing trees on the other hand, as practised in the present experiment, is expected to re-establish higher yields more quickly than re-planting. The first full harvest in the trial started in September 2012, 15 months after the rehabilitation pruning. Between September 2012 and March 2013 (main harvest) an average of 462 kg dry beans per ha were harvested in the control treatment. The yields of the common practice treatment already amounted to 24.9 % of the control. This is a very promising result, especially in view of the development of young trees which will take at least another year before the first pod development. Yields in the agroforestry systems increased less quickly as tree development under shade and with lower fertiliser input is inherently slower.

Cocoa, mainly produced by 5 to 6 millions of smallholder farmers, is considered as one of the most sustainable production system in the humid tropics. Little is known about the sustainability of different cocoa production systems. A long-term experiment is set up in Alto Beni at 400m above sea level with a humid winter dry climate, 1’540 mm annual rainfall. The trial assesses the sustainability of five cocoa (Theobroma cacao) production systems with the parameters of yield and yield stability, input-output efficiency of nutrients and energy, soil fertility, biodiversity, economic result, climate change mitigation and adaptation. The two-factorial experiment is arranged in an completely randomised block design; the five cocoa treatments, based on local and international practices, are four times repeated. The production systems are differentiated by the diversity of shade canopy and by crops, from mono culture full sun cocoa to a agroforestry cocoa with leguminous species (Inga edulis, Erythrina poeppigiana) shade canopy, including fruits (e.g. Euterpe precatoria, Theobroma grandiflorum) and timber (e.g. Centrolobium ochroxylum, Swietenia macrophylla) species, and a higher diversified agroforestry system based on the natural successions of species. The management of the cocoa is conventional and organic. The five treatments are: mono culture full sun cocoa conventional, mono culture full sun organic, agroforestry conventional, agroforestry organic and successional agroforestry organic. Fallow plots and nearby forests plots are monitored for soil fertility and biodiversity. Field clearing started in 2007 followed by maize (Zea mays) crop and end of 2008 the cocoa plots (48m×48 m) were established. The results of the baseline studies concerning soil fertility show good nutrient level for cocoa production; the variance of soil parameters is documented in a soil map. According the FAO soil classification (2006) the soils are Lixisole and Luvisole with high base saturation.

Die Hexenbesenkrankheit (Crinipellis perniciosa) ist eine Pilzkrankheit, die nur meristematisches Pflanzengewebe des Kakaobaumes befällt. Sie tritt bisher ausschließlich auf dem Lateinamerikanischen Kontinent auf. Für die Region des Alto Beni im feuchttropischen Tiefland Boliviens wurden die Interaktionen zwischen Witterung, Inokulum und Phänologie der Wirtspflanze untersucht, um effiziente und ökonomisch vertretbare Kontroll- und Bekämpfungsprogramme zu entwickeln. Von 1994 bis 97 wurden zudem die Kakaoklone ICS 1, ICS 6, ICS 8, ICS 95 und TSH 565 unter Feldbedingungen auf Hexenbesentoleranz und Ertrag untersucht. Zusätzlich wurden unterschiedliche Baumschnittfrequenzen auf ihre Effizienz hinsichtlich des Ertragsverhaltens der Klone und der Auswirkung auf Hexenbeseninfektionen überprüft. Die Ergebnisse zeigen einen Zusammenhang zwischen Niederschlagsereignissen und der Ausbildung von Fruchtkörpern an Hexenbesen. Es konnte im gesamten Untersuchungszeitraum jedoch kein kausaler Zusammenhang zwischen der Intensität von Fruchtkörperbildungen an Hexenbesentrieben und der Anzahl an Hexenbeseninfektionen der untersuchten Bäume festgestellt werden. Die fünf Klone zeigten sowohl signifikante Unterschiede in Bezug auf ihre Anfälligkeit gegenüber Hexenbesen als auch in ihrem Ertragsverhalten. Die unterschiedliche Schnitthäufigkeit zeigte dagegen weder statistisch abgesicherte Differenzen hinsichtlich der Anzahl von Hexenbeseninfektionen noch hatte sie Einfluss auf den Ertrag. Die gängigen Bekämpfungsmethoden wie Baumschnitt und der Einsatz von tolerantem Pflanzenmaterial stellen keine grundsätzliche Lösung des Problems dar. Die bisher gewonnen Erfahrungen im Landbau unter Anwendung der Prinzipien sukzessionaler Agroforstsysteme zeigten, dass diese möglicherweise eine langfristige Perspektive für die Landnutzung in tropischen Regionen und zur Reduzierung phytosanitärer Probleme darstellen könnten. The witches’ broom disease (Crinipellis perniciosa) is a fungal infection which only affects the meristematic plant tissues of the cocoa tree. To date it has been restricted to the Latin American continent. The aim of this work was to examine scientifically the measures undertaken in the Alto Beni region in the humid tropical lowlands of Bolivia, to curb the witches’ broom disease (Crinipellis perniciosa) by phytosanitary pruning measures and by the use of various tolerant clones From 1994 to 97, the cocoa clones ICS 1, ICS 6, ICS 8, ICS 95 and TSH 565 were also studied under field conditions at the Sapecho location for witches’ broom tolerance and yield. In addition, various pruning frequencies were examined for their efficiency with regard to the yields of the clones and the effects on witches’ broom infection The results show a connection between rainfall levels and the formation of basidiocarps on the witches’ broom. However, over the whole of the investigation period, no causal connection was found between the intensity of basidiocarp production on the witches’ broom shoots and the number of witches’ broom infections on the trees under investigation. The five clones showed significant differences both in relation to their susceptibility to witches’ broom and in their yields. The different pruning frequencies, on the other hand, did not show any statistically provable differences in terms of the number of witches’ broom infections, nor did they influence the yields. It has become clear that neither the own studies nor the many research works undertaken at an international level were able to find a fundamental solution to the witches’ broom problem. Agriculture using the principles of successional agroforestry systems could represent a long-term perspective for land use in these areas, with the emphasis on more than short-term economic, social or structural objectives.