IEH Instituto de Estudios del Hambre ANALYSIS OF CLIMATE CHANGE IMPACTS ON COFFEE, COCOA AND BASIC GRAINS VALUE CHAINS IN NORTHERN HONDURAS

Transcription

1 IEH Instituto de Estudios del Hambre ANALYSIS OF CLIMATE CHANGE IMPACTS ON COFFEE, COCOA AND BASIC GRAINS VALUE CHAINS IN NORTHERN HONDURAS January 10th 2013

2 Table of contents Abbreviations and acronyms...4 Weights and measures...5 ABSTRACT INTRODUCTION Background Methodology Limitations of the study SIMULATING CLIMATE SCENARIOS FOR THE PROJECT AREA Technical requirements for future climate scenarios Use of the most recent Climate Models Provision of daily series Provision of local information Handling of uncertainties Full verification and validation studies Downscaling methodology Verification and Validation Results Future local climate scenarios COFFEE VALUE CHAIN Mapping: process and actors Value chain steps and scheduling State of coffee cultivation in the project area Actors in the chain Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios Flowering phase Ripening stage Harvest stage Post-harvest stage COCOA VALUE CHAIN Mapping: process and actors Value chain steps and scheduling State of cocoa cultivation in the project area Actors in the chain Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios The whole crop cycle Flowering phase Ripening phase Harvest phase Post-harvest phase BASIC GRAINS VALUE CHAIN Mapping: process and actors Value chain steps and scheduling Maize and beans situation in the project area Actors in the chain

3 5.2 Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios MAIZE VALUE CHAIN Sowing phase Germination and development phase Flowering and ripening phase Harvesting phase Post-harvest phase BEANS VALUE CHAIN Whole crop cycle Sowing phase Germination and development phase Flowering and ripening phase Harvesting phase Post-harvest phase CONCLUSIONS Projected climate change Coffee value chain Cocoa value chain Basic grains value chains RECOMMENDATIONS Coffee value chain Cocoa value chain Basic grains value chain...95 REFERENCES...97 ANNEXES ANNEX I TERMS OF REFERENCE ANNEX II - LIST OF PEOPLE NTERVIEWED ANNEX III - DATA USED TO PRODUCE FUTURE CLIMATE LOCAL SCENARIOS ANNEX IV - TECHNICAL REQUIREMENTS FOR FUTURE CLIMATE SCENARIOS ANNEX V: FICLIMA STATISTICAL DOWNSCALING METHODOLOGY ANNEX VI. VERIFICATION AND VALIDATION RESULTS 3

4 Abbreviations and acronyms ACDI Canadian International Development Agency ADECAFEH Honduran Coffee Exporters Association AECID Spanish Agency of Development Cooperation CAN Canada climate model CATIE Topical Agricultural research and Higher Education Center CM Climate model CMIP5 Coupled Model Intercomparison Project Phase 5 ESM Earth System Model FHIA Honduras Agricultural Research Foundation FIC Climate Research Foundation FUNDER Rural Enterprise Development Foundation GFDL Geophysical Fluid Dynamics Laboratory - United States climate model GHG Greenhouse gas IEH Institute of Hunger Studies IFAD International Fund for Agricultural Development IHCAFE Honduras Coffee Institute INE National Statistics Institute IPCC Intergovernmental Panel on Climate Change MT Metric Ton NASA National Aeronautics and Space Administration NCAR National Centre for Atmospheric Research NCEP National Centres for Environmental Prediction NOR Norway - climate model PET Potential Evapotranspiration PRONAGRO National Programme for Agrifood Development RCP Representative Concentration Pathways SAG Secretary of Agriculture and Livestock SERNA Secretary of Natural Resources and Environment TRMM Tropical Rainfall Measuring Mission USAID United States Agency for International Development Indicators ABO ACOS COS ICF ICFT ICC ICCF ICGP ICGT ICIPF ICMP ICMTF ICP-SF IDTemp IEST IF* IFFP Fruit Abortion Indicator Harvest Accumulation Indicator Harvest and Transport Conditions Indicator Quantity of Flowering Indicators Storm-caused Flower Fall Indicator Harvesting Conditions Indicator Harvesting Conditions Indicator (beans) Indicator of Germination and Establishment Conditions Associated with Water Shortage/Excess Indicator of Germination and Establishment Conditions Associated with Low Temperatures Indicator of Establishment Conditions Associated with Excessive Rainfall Minimum Bean Crop Rainfall Conditions Indicator Minimum Temperature Conditions Indicator Pre-sowing Conditions Indicator Differences in Temperature Indicator Dry Season Indicator Star Flower Indicator Flowering and Bean Formation Conditions Indicator 4

5 IIF III IIPC IIPrp IP-CF IPCh IRT ISF ISH ISM ITMin PCOS SEC SuEn SuMon SUS SUSH Flowering Induction Indicators Rainy Season Onset Indicator Pollen Non-viability Indicator Non-viability Due to Irregular Rainfall Indicator Post-harvest Indicator (beans) Post-harvest Indicator Temperature Severity Indicator Sowing Indicator Water Saturation Indicator Sowing Indicator (maize) Non-viability Due to Minimum Temperatures Indicator Post-harvest Indicator Pre-drying Indicator Vulnerability to Disease Indicator Vulnerability to Monilia Indicator Vulnerability to Diseases Indicator Indicator of Pests and Diseases Due to Absence of Cold Weights and measures 1 manzana = 0.7 Hectares 1 quintal = Kilograms 1 quintal per manzana = 32 kilograms per hectare 5

6 ABSTRACT This study applies an innovative methodology designed to analyse climate change impacts on the areas prioritized by the Northern Horizons project in Honduras and make recommendations that make it possible to strengthen the resilience of project beneficiaries in the coffee, cocoa, maize and bean value chains. The proposed methodology sets out and applies a range of minimum requirements for the reliable generation of climate change scenarios through the use of the most advanced climate models and historical series of daily data. It also quantifies uncertainties, verifies and validates the methods and applies regionalization in order to downscale the projected changes to a local scale. By mapping the value chains and consulting national experts, this methodology identifies the critical elements most vulnerable to climate change, formulates and verifies indicators to predict how future climate will affect the value chains and analyses its impact, proposing adaptation measures. In the case of the coffee crop, the expected impacts are negative due to the increase in temperatures that will provoke changes in the crop cycle, with consequences ranging from higher vulnerability to some diseases to more complicated harvesting and post-harvest tasks. On the other hand, higher temperatures projected by the scenarios will favour cocoa growing, although there will also be negative impacts associated with better conditions for monilia disease development. In the maize crop, in general terms, the future climate will be beneficial to most of the studied areas, though in some areas higher rainfall will complicate certain farming phases, including the first stages of development and harvesting, which will increase vulnerability to diseases. Something similar will happen with beans: increased rainfall will make sowing, flowering and grain filling more difficult. Recommendations for the coffee crop include improving the existing varieties and crop management, supporting investment in infrastructure such as irrigation systems or drying facilities, encouraging more efficient associations, and doing research on the relationships between crop and climate. In the case of the cocoa crop, recommendations include expansion of the cultivated area and replacement of old plants by better adapted varieties, improved cultural practices in order to reduce disease impact and enhance quality, diversification of cocoa grower income sources with timber species that also provide protection from higher temperatures, and support for the small-scale producer to gradually incorporate more addedvalue activities to the chain. The latter requires more training, technical support and improved associations. In general terms, more investment in meteorological stations is suggested in order to ensure that enough meteorological data are available. It is also recommended to include additional research activities in the Northern Horizons project to deepen climate change impact analysis at local levels as well as to further study the climatic impact on specific stages of the value chain. Finally, dissemination of the findings of this study and its replication in other sectors is recommended, as well as in other IFAD programmes in different countries. 6

7 1. INTRODUCTION 1.1 Background In 2011, the Government of Honduras asked IFAD to prepare a project proposal aimed at increasing the climate resilience of agricultural production chains, protecting smallholder farmers and their crops from the impacts of climate variability in that country. This proposal would complement the Competitiveness and Sustainable Rural Development Project in the Northern Zone (Northern Horizons), a pro-poor value chain development investment co-funded by IFAD, which will also support capitalization and asset building in production, value aggregation and transformation, sales process and better market access. Among others, Northern Horizons will support the following value chains: coffee, cocoa and basic grains (maize and beans). The concept note of this additional proposal was approved in November The project is expected to last 4 years and aims to reduce the effects of climate change on smallholder farming system productivity and profitability through (a) the promotion of adaptation techniques and introduction of technologies that can make agricultural value chains more resilient, and (b) awareness raising and reinforcement of capacities for better management of climate change risks. Northern Horizons and the additional adaptation proposal will operate in municipalities with high concentrations of rural poverty in the departments of Atlántida, Cortés and Santa Bárbara. Climate model forecasts are essential elements of any adaptation strategy because they make it possible to plan better through the anticipation of future impacts. However, efforts are currently scattered, and few, if any, institutions are systematically applying climate modelling to agricultural scenarios. For most regions of the world the impact of an increasingly variable climate and more frequent extreme weather events on agribusiness and production chains is unknown, making the recommendation of viable, practical adaptation paths even harder. Fundación para la Investigación del Clima (Climate Research Foundation, FIC) and Instituto de Estudios del Hambre (Institute for Hunger Studies, IEH) have been working since 2010 on the application of climate simulation to enhance food security in rural areas in Central America 1. To this end, these institutions have signed a letter of agreement with IFAD to detect and analyse the expected impacts of climate change in the territories prioritized in the Northern Horizons project, making preliminary recommendations on how to better protect the value chains included in the project from climate risks, while contributing to increased resilience among the project beneficiaries (Terms of Reference in Annex I). 1.2 Methodology The methodological process applied is based on the three stages necessary to address climate change adaptation: 1. Description of potential future climate conditions; 2. Evaluation of how this future climate will impact the value chains covered by the study; 3. Recommendation of what to do to minimize the adverse impacts identified, including effective measures of adaptation to climate variability. These stages must be developed on a local scale that is, in the three departments included in the project area: Santa Bárbara, Cortés and Atlántida because many of the adaptation interventions are defined at this level. 1 See, for example, Ribalaygua, J. et al. (2011). Enfoques innovadores en la simulación del cambio climático y su impacto en la seguridad alimentaria. La experiencia de Nicaragua. Managua: Universidad Centroamericana de Nicaragua 7

8 The methodology followed these steps: 1. Generating future climate scenarios for the project area These elements were necessary in order to generate future climate scenarios: a) Model selection: Four climate models were applied, all of them internationally recognized and used by the IPCC in its Fifth Evaluation Report. These models were developed in Canada (CAN), Norway (NOR), the United States (GFDL), and Germany (Max Planck Institute, MPI). The different ways that humanity may develop in the future and the extent to which reducing/mitigating its effect on climate change will be considered important were represented by four RCPs (Representative Concentration Pathways), also used in the forthcoming IPCC5. b) Gathering historical observations. The input data were daily series of rainfall and maximum and minimum temperatures. The temperatures were collected from 9 temperature monitoring stations and the rainfall from 14 stations and 110 grid points (NASA s TRMM database Tropical Rainfall Measuring Mission). The data period was 1951 to 2012 (although many stations and the TRMM provided data for some years only). See Annex III. c) Application of a downscaling methodology. Downscaling methodologies adapt the low resolution information provided by Global Climate Models in order to project local-scale surface effect (rainfall, temperature). These are the local climate scenarios required to be able to develop local adaptation strategies. The downscaling methodology named FICLIMA was used to this end; it is based on a two-step analogue method developed by the FIC. See AnnexV. d) Verification process. The methodology s reliability in simulating the variables of interest is determined through verification. To do this, the downscaling methodology must be applied to an atmospheric re-analysis, which offers information about the state of the atmosphere in the past. By comparing the simulated series with the real recorded data series, it is possible to evaluate the methodology s potential to translate low-resolution atmospheric information into rainfall and temperature on a local scale. See Annex VI. e) Validation process. Once the methodology has been successfully verified, that is, when its capacity to simulate surface effects based on observations of the atmospheric situation at low resolution has been checked, the validation process takes place. This aims to show whether the Climate Models faithfully represent the current climate. This process was carried out applying the FIC downscaling methodology to the so-called historical experiment or control simulation, obtaining a series of simulated rainfalls and temperatures for each model, which was compared to the past climate recorded. See Annex VI. f) Generation of future scenarios. Once the methodology for climate model downscaling and validation was verified it was applied to the model outputs with the different RCP for a future period (from now to 2100), obtaining the simulated rainfall and temperature values for this period at every observatory under study and for each of the 14 model outputs (two models with three RCPs each, and another two with four RCPs). 2. Mapping the value chains of coffee, cocoa and basic grains and identifying the critical elements particularly vulnerable to climatic events Taking the value chain as the subject of analysis, the value chains for the four crops studied (coffee, cocoa, maize and beans) were mapped in order to obtain information necessary for the rest of the study. Mapping includes information on the chain stages, the scheduling, the crop situation in the project area and the main actors involved in the different stages of the chain and their interrelations. This exercise started with a meeting at the offices of the Secretary of Agriculture and Livestock (SAG in Spanish) with representatives of the two ministry units involved in the FIDA project: the SAG, involved through the National Programme for Agrifood Development (PRONAGRO), and the Secretary of Natural Resources and Environment (SERNA), involved through the National Climate Change Directorate. Both institutions identified the primary sources of information and presented their views on the study s development. The first step was to carry out interviews with technical staff and organisations from each sector in 8

9 the three departments covered by the study and consult secondary sources to map the working of each value chain as completely as possible. After describing the value chain components and how they work, the critical elements most vulnerable to climatic events were identified. A critical element is any aspect of the crop cycle or further stage in the value chain that may eventually have a significant impact on product quantity or quality, and which is particularly sensitive or vulnerable to climatic events. The critical elements were selected based on value chain activities, particularly the crop cycle, to determine the most climate-sensitive stages. Moreover, a desk review of similar studies in other countries was carried out and discussed with national experts in workshops (see the list of people interviewed in Annex II) in order to validate the critical elements selected and jointly determine how they are affected by climatic events. 3. Analysing the vulnerability of the critical elements identified in the value chains and the impacts of future climate on them The definition of critical elements and the future climate scenarios generated were the basis for analysing the potential future climate impacts on the coffee, cocoa and basic grains value chains. For the analysis it was necessary to build a set of indicators based on climate information (rainfall and temperature) to measure how every critical element is affected by climate, programming them using R language (programming environment for graphic and statistical analysis). The indicators were applied to the rainfall and temperatures observed and discussed in workshops with farmers and national experts in order to ensure that the results reflected what actually happened in Northern Honduras. Once verified, these indicators were applied to the generated scenarios, making it possible to determine the temporary evolution of indicators and their implications on the value chains through four different climate models and four RCPs. The use of different models and RCPs made it possible to reduce the uncertainties when applying the climate change predictions to each of the critical value chain elements. This method of translating rainfall and temperature into information useful to the value chain analysis (including the quantity of flowering depending on the rainfall, fertilization depending on the beginning of the rainy season or the presence of infestations and diseases depending on temperature) is necessary in order to evaluate the impact of future climate on the value chains and make recommendations for minimizing unwanted impacts. 4. Making recommendations regarding each of the critical elements to minimize negative impacts and reinforce positive impacts The results of the analysis were discussed with national experts in order to jointly analyse the potential future climate impacts on the coffee, cocoa and basic grains value chains and propose adaptation actions that minimize negative impacts and boost opportunities. The discussion conclusions have been included in the recommendations of this study. 1.3 Limitations of the study First of all, it should be highlighted that limited resources and time were allocated for the present study, and therefore the analysis has been carried out in record time. As a result, it has not been possible to achieve the desired level of depth to obtain more detailed conclusions and recommendations. Along with this significant limitation, the following challenges were also present during the realization of the study: Obtaining meteorological data from the government agency responsible for climatic information was particularly difficult. Other government agencies having relevant 9

10 information, such as the Water Resources Directorate, were very collaborative, providing rainfall information from five meteorological stations. Given the gaps in weather information in the study area and the difficulties faced in accessing this information, available data from stations or grid spots near the project areas (in some cases in other municipalities) were used instead. Conclusions obtained from these data have been extrapolated to the municipalities included in the study. As regards temperature, very few data have been obtained, but since temperature is more geographically homogeneous the extrapolation is more robust. In the case of rainfall, which has higher spatial variability and thus requires more data density, there were more meteorological stations and the RMSS grid has acceptable coverage of all the territory. There are very few data on temperature in high altitude stations, which has made it difficult to analyse the coffee value chain. In the area of study, Virgilio Enamorado Station has temperature data, but the data series is very short and not homogeneous. Therefore, in order to analyse indicators including temperature, Santa Rosa de Copán Station data were used instead; it is located 1,000 meters above but outside the area of study. Statistics on production of cocoa and basic grains (quantity, coverage, number of farmers) for the area of study are not available. The INE has data on basic grains for 2009, but only covering large producers. Some key informants did not respond to information requests, which added to the difficulty of gathering data for the value chain mapping. The collaboration of experts has proved extremely difficult. Most of the specialists contacted did not participate in the workshops organized to validate indicators, perform scenario analysis and make recommendations. As a result, few experts have provided their knowledge and opinion for the study. Nevertheless, those who participated were highly knowledgeable in their respective areas, which ensures the accuracy of the collected information, although it was not possible to make an in-depth contrast of their opinions. In the case of coffee and cocoa, the interviews were carried out during the harvest season, which inhibited farmers from participating in the meetings. Even though there has been some research on climatic impact on some crops in Honduras (such as the FHIA research on cocoa), no research done in Honduras has been found with the level of detail required for verifying the indicators. Therefore, this study has been based on research from other countries and knowledge of the experts consulted, which is often empirical. 2. SIMULATING CLIMATE SCENARIOS FOR THE PROJECT AREA 2.1 Technical requirements for future climate scenarios The causes of Climate Change are global (greenhouse gas emissions all over), but the consequences will be local (very different changes for points located quite close to each other, depending on topographical influences). Adaptation to climate change requires information about future climate on a local scale (Trigo and Palutikof, 2001). Climate Models (CM) are the main tool used to simulate future climate conditions. They provide valuable information on general future atmospheric characteristics but they have limitations with regard to suggesting surface effects (e.g. precipitation, temperature) and resulting local details. These limitations are mainly due to CMs low resolution (cells of km), which prevents them from properly representing topography details that greatly influence local climates (Von Storch, 1994; Nakicenovic and Swart, 2000). For example, in order to provide adaptation strategies for coffee or cocoa in specific areas in the North of Honduras we need to know the climate expected at that point, and it will not be enough to know how the CM grid cell (i.e., 200km x 200km) will warm on average. 10

11 Therefore, we must extract the information that makes it possible to do an impact assessment and adaptation recommendation on a local scale (local weather effects precipitation, temperature, etc.) from the valuable information provided by CMs (low resolution atmospheric configurations). This procedure is called downscaling. More information on CMs, their limitations, and downscaling is provided in Annex III. Robust climate scenarios and downscaling have specific requirements that have been fulfilled in this study (more details on these requirements are provided in Annex IV): Use of the most recent Climate Models Climate Models are renewed and improved continuously, and a new version usually appears every 4 to 6 years, which is used for the corresponding IPCC-AR (Intergovernmental Panel on Climate Change Assessment Report). It is very important to use the most recent CMs to ensure the most robust future climate simulations possible. The most recent CMs available are those of CMIP5 (Coupled Model Intercomparison Project Phase 5), and their results will be used for IPCC AR5. One of the main new features introduced by CMIP5 is related to the CMs: most of them are Earth System Models (ESM), a new generation of CMs. Another important new feature is the way future radiative forcing (which depends on society s evolution) is taken into consideration: the traditionally-used GHG emission scenarios (A2, B1, A1B...) have been replaced by so-called Representative Concentration Pathways (RCPs), which introduce relevant differences. Studies using scenarios obtained with previous models will be probably considered obsolete in the near future. For all these reasons, CMs belonging to CMIP5 have been used in this study Provision of daily series Daily series are needed to explain many essential features of the climate: how the distribution of precipitation (number of days without precipitation, maximum accumulated precipitation over five days, maximum precipitation in 24 hours...) or temperature (the effect of several days of extreme temperatures on health or agriculture, for example) within a month affects specific characteristics of a value chain. Daily series are necessary for running many impact assessment models: (on hydrology, agriculture/food security and phytoclimatology). In this study, we have used simple models that require daily data to calculate specific indicators for each critical element of the value chain, most of them defined to take into account the influence of daily data distribution. Temperature and precipitation daily series from 14 meteorological stations and 110 TRMM grid points (Tropical Rainfall Measuring Mission) have been used for this study (see Annex III) Provision of local information Future climate will imply changes with regard to the present climate that will be very different for locations very close to one another, depending on the influence of topography. Also, local information about future climate is required for many adaptation activities (for example, to define the variety of coffee to be planted). Despite being the most powerful tools available today for the simulation of future climate, Climate Models (CM) are not able to represent local climate details. To solve this problem, what are known as downscaling techniques have been developed. These techniques obtain the local scale surface effects (precipitation, temperature) required for impact assessment and adaptation from the valuable information provided by the climate models (low-resolution atmospheric fields). Some tools provide information with local detail, but it is often a simple interpolation or redistribution of the data provided by the Climate Models (CMs). The use of more sophisticated 11

12 downscaling methodologies is required. The downscaling tool used in this study, called FICLIMA, is one of the most robust methodologies, as has been shown by the results of different national and international projects. See Annex V Handling of uncertainties The quantification of the uncertainties inherent in any climatic simulation is one of the areas in which the scientific community is focusing significant efforts. To work appropriately with uncertainties, rather than managing them initially to build a single "future" (for example, doing a weighted average of all the future projections), it is more advisable to select multiple "futures" and work with all of them to assess impacts, as has been done in the present study Full verification and validation studies The establishment of projects which perform systematic, rigorous verification and validation procedures for all the downscaling tools and all the CMs available (e.g. the European STARDEX project or the Spanish one, ESTCENA) is of the utmost interest. These projects make it possible to identify the strengths and weaknesses of the different tools and CMs and seek complementarities between them. At the same time, they provide a lot of useful information for making a subsequently appropriate use of the scenarios. In this study we performed detailed verification and validation procedures before producing future scenarios, with good results in general (although more detailed analysis on the local scale should be done.) 2.2 Downscaling methodology The Climate Research Foundation (FIC) has developed a statistical downscaling technique called FICLIMA that has been successfully verified in several national and international projects. The statistical downscaling techniques consist of establishing relationships between large-scale atmospheric fields (called predictors) and high-resolution surface variables such as temperature and precipitation (called predictands). The scenarios are built applying those relationships to the outputs provided by the CM. FICLIMA methodology and its application in this study comply with the aforementioned requirements: the newest Climate Models from CMIP5 have been used; FICLIMA works on a daily scale, and provides daily series of maximum and minimum temperature and precipitation for each CM projection; it provides local information for the observatories and the grid points used in the study; uncertainties have been considered and quantified by means of downscaling 14 projections (4 CM, with 3 or 4 RCPs each); and detailed verification and validation processes have been undertaken for each variable, observatory or grid point and CM. Annex III includes information about the data used in the study: predictand observations from stations provided by NOAA (National Oceanic and Atmospheric Administration, USA) and by the General Directorate of Water Resources (Dirección General de Recursos Hídricos) of SERNA (Secretary of Natural Resources and Environment), and from grid points in the TRMM (Tropical Rainfall Measuring Mission); NCEP/NCAR Reanalysis ( observations of predictors); and Climate Model outputs. Scarcity of observation data has been one of the most important limitations in this study, because information was only available in few stations and historical series were not complete. More details about FICLIMA downscaling methodology are provided in Annex V. 2.3 Verification and Validation Results FICLIMA statistical downscaling methodology has been tested with excellent results on various 12

13 projects of national and international scope, and in different regions such as Europe, Central America and South America (after the methodology s corresponding adaptation). Despite the good results already obtained, if the methodology is applied to a new area of study (Northern Honduras), verification and validation processes are performed to assess how well the methodology and Climate Models are simulating the present climate. Results of the methodology's verification process In the verification process, simulations of the predictands obtained by applying the downscaling technique to "observations" of the predictors (NCEP/NCAR reanalysis, are compared to observations of those predictands. This process evaluates how well the methodology translates low-resolution atmospheric fields (predictors) into local surface effects (predictands). The results of the verification process for temperatures (both maximum and minimum) are very accurate. As shown in Figure 1 (top graphics), which represents the monthly averages for , the methodology successfully reproduces the annual temperature cycle for observed data (blue) and for simulations obtained from downscaling NCEP/NCAR reanalysis (red). The temperature extremes can be observed in April-May and September for maximum temperatures and in July for the minimum. These figures represent the average of all the observatories included in this study. Individual results for each observatory are provided in Annex VI. Figure 1- Results of the verification process for maximum temperature, minimum temperature, precipitation and the number of days with precipitation, for the average of all the observatories included in this study and for the common period For precipitation (Figure 1, bottom graphics), the verification results are satisfactory. The methodology tends to underestimate precipitation in October to February while it overestimates it in March to May, and tends to slightly overestimate the number of days with precipitation. It is remarkable that the methodology reflects the phenomenon of the canícula (summer drought: the dryer period in July-August). 13

14 Results of the methodology's validation process The validation process is similar to that of verification, but in this case it compares the simulated data obtained by downscaling the NCEP/NCAR reanalysis with those obtained by downscaling the CM control output (using the Historical Experiment, which is the CM simulation for the period ). The model's ability (together with the downscaling methodology) to simulate present climate is checked through the validation process (some models tend to simulate the climate as colder/warmer or wetter/drier than it really is). Figure 2 shows the validation results for maximum and minimum temperature, precipitation and number of days with precipitation. The black curve corresponds to the data obtained by downscaling the NCEP/NCAR reanalysis and the other curves correspond to each of the models used in the study. As shown in the figure, in the case of temperature (maximum and minimum), all models are able to reproduce the annual cycle and reflect temperature maximums and minimums, although these peaks are more or less marked depending on the model. Validation results for precipitation are acceptable, although most of the models tend to undersimulate precipitation and to soften the effect of the canícula (and the Norwegian model, NorESM1, and the GFDL-ESM2M, USA, are not even able to reproduce this phenomenon). The MPI-ESM-MR model (from the Max Planck Institute, Germany) is the best one in the whole validation process. Figure 2 - Results of the validation process for maximum temperature, minimum temperature, precipitation and number of days with precipitation for the average of all the observatories included in this study and for the period Verification and validation results correspond to the average of observatories; however accurate impact evaluations and definition of adaptation strategies for a specific area require a more detailed research at local scale for each of the observatories. Figures 3 and 4 illustrate the importance of local analysis. Figure 3 represents the validation of the GFDL-ESM2M CM for precipitation in Santa Rosa de Copan, where the CM (dark grey line) is not able to accurately 14

15 simulate climate observations (light grey line) and more concretely June s precipitation and the phenomenon of canicula. However, Figure 4 for La Mesa in San Pedro Sula the simulation of this CM is more accurate. Analysis at local level would conclude that ESM2M CM can be used for simulating future climate in San Pedro but not in Santa Rosa, because CM is not able to simulate the climate observed in this station Figure 3 and Figure 4: GFDL-ESM2M CM validation (and climate projections, coloured lines) for precipitation in Santa Rosa de Copan and in La Mesa (San Pedro Sula). Observations (downscaling NCEP/NCAR Reanalysis, light grey) and simulation downscaling GFDL-ESM2M historical experiment (dark grey) 2.4. Future local climate scenarios Once the verification and validation processes have been completed and the results have been found to be sufficiently satisfactory, the future climate scenarios are produced. Daily series of maximum temperature, minimum temperature and precipitation are simulated throughout the 21 st century. In this study, future climate scenarios for four climate models belonging to IPCC5 and three/four RCPs have been produced (Table 1). Climate Model Spatial resolution Temporal resolution Calendar days/year MPI-ESM-MR 1.8ºx1.8º daily Gregorian GFDL-ESM2M 2ºx2.5º daily 365 CanESM2 2.8ºx2.8º daily 365 NorESM1 1.8ºx2.5º daily 365 Table 1. Climate Models used in this study Available outputs Historical RCP26 RCP45 RCP85 Historical RCP26 RCP45 RCP85 RCP60 Historical RCP26 RCP45 RCP85 Historical RCP26 RCP45 RCP85 RCP60 Research Centre Max Planck Institute for Meteorology (MPI-M) Germany National Oceanic and Atmospheric Administration (NOAA) USA Canadian Centre for Climate Modelling and Analysis (CC-CMA) Canada Norwegian Centre (NCC) Norway Climate The future climate scenarios are analysed as expected increases compared to the control period The following graphics show the 30-year moving averages of the four RCPs for all of the observatories: thick lines show the average for all the observatories, and dashed lines (where needed) show the 95 th percentile for all the observatories and CMs used in the study. Shaded areas represent the interval between the 95 th and 5 th percentiles illustrating the 15

16 differences between observatories and CMs. It is generally expected that maximum and minimum temperatures will increase gradually over the entire century for all emission scenarios. For the middle of the century, the highest increases are expected in maximum temperature, with values of up to 2ºC in the most extreme RCP (on average for all the observatories and CMs) and 1ºC in the coldest RCP. The minimum temperature increases expected to take place are between just under 0.8 C and 1.7ºC for the same period (figures 5 and 6). Precipitation is expected to increase gradually over the whole century, reaching values of about 0.25 mm/day (approximately 90 mm/year) by mid century, which represents an increase of less than 10% by mid-century (figures 7 and 8). Future climate scenarios for maximum and minimum temperature Figure 5 Changes in mean annual maximum temperature expected throughout the 21 st century (30-year moving averages). Different colours for different RCPs. Thick lines represent the average for all the observatories and Climate Models, and dashed lines (where needed) represent the 95 th percentile of that population (all observatories for all CMs). Shaded areas represent the interval between the 5 th and 95 th percentiles. Temperature changes in C. Figure 6 Changes in mean annual minimum temperature expected throughout the 21 st century (30-year moving averages). Different colours for different RCPs. Thick lines represent the average for all the observatories and Climate Models, and dashed lines (where needed) represent the 95 th percentile of that population (all observatories for all CMs). Shaded areas represent the interval between the 5 th and 95 th percentiles. Temperature changes in C. 16

17 Future climate scenarios for precipitation Figure 7 Changes in annual mean daily precipitation expected throughout the 21 st century (30-year moving averages). Different colours for different RCPs. Thick lines represent the average for all the observatories and Climate Models, and dashed lines (where needed) represent the 95 th percentile of that population (all observatories for all CMs). Shaded areas represent the interval between the 5 th and 95 th percentiles. Precipitation changes in mm/day. Figure 8 Relative changes (in %) in annual mean daily precipitation expected throughout the 21st century (30-year moving averages). Different colours for different RCPs. Thick lines represent the average for all the observatories and Climate Models, and dashed lines (where needed) represent the 95 th percentile of that population (all observatories for all CMs). Shaded areas represent the interval between the 5 th and 95 th percentiles. Precipitation changes in %. These results apply to average data of all the observatories and CMs used in the study throughout the year. Once more, an analysis at the local level is needed because climate change will have different impacts on different locations within the territory and at different times throughout the year. In order to illustrate these differences, Figure 9 shows the German MPI- ESM-MR CM s annual rainfall simulation. The dark grey line is the historical (CM corrected with observations) and the three coloured lines are mid-century simulations ( averages) for three RCPs for La Ceiba (787050) and Tela (787060). Figures show that the German model simulates a relevant rainfall increase in July and August in La Ceiba and from June to November in Tela. This example illustrates how a CM is predicting very different changes in climate at diverse times of the year and in different stations. 17

18 The analysis of effects on the value chains will capture these differences, which will affect coffee, cocoa, maize and beans in the different stages of production, harvest and post-harvest. Future research will be needed to deepen the analysis of the consequences and impacts of different local future scenarios in specific locations of particular interest for development of the value chains. 3. COFFEE VALUE CHAIN 3.1 Mapping: process and actors Value chain steps and scheduling The coffee value chain is complex because it includes a great variety of activities and actors. The central link of this value chain is made up of the activities related to the physical handling of the product; these steps are described in the following figure: Figure 10: Central link in the coffee value chain. Source: The authors. In addition to this central link, the value chain includes activities related to support services and the institutional framework that will be analyzed in less detail, given that they are not subject of this study. Production It was assumed that the plants are more than three years old and that they are in a productive phase. The following diagram shows the productive cycle: 18

19 Figure 11: Schedule of the coffee production cycle at Northern Honduras. Source: The authors. Harvesting takes place after weeks since flowering (between September and February) depending on the weather conditions and the altitude. At Cortes, in the lake area, harvesting may start in September while in the higher areas of Santa Barbara it may go through May. Wet milling Figure 12. Parchment coffee production routes in the three Northern Honduras departments analyzed. Source: The authors. Choosing to sell coffee as cherries or wet parchment depends on the farmer s capacity and the wet mill capacity. The mill occasionally reaches its maximum capacity, forcing the grower to process the coffee at his own farm or outsource this service to other farmer. Collecting and drying Figure 13: Collecting and drying. Source: The authors Wet parchment can be dried using solar or mechanical procedures. Transformation into dried parchment which contains no more than % humidity requires 32 hours of mechanical drying or 3 to 4 days of solar drying. Transportation to the San Pedro Sula dry parchment 19

20 storage area is mostly done by the middleman or through a transport service outsourced by the exporter. 40% of the coffee milled by the grower is sold to local middlemen, 45% to regional middlemen and 15% to exporters. In the case of milling facilities, most coffee goes to exporters (85%) and the remaining 15%, generally lower quality coffee (beginning or end of harvest, lower quality grains or mill tail, etc.) is sold to regional middlemen in nearby departments. In turn, 80% of the coffee sourced by local middlemen is supplied to regional middlemen, and the remaining 20% goes directly to exporters. Figure 14: Coffee value chain, simplified diagram. Source: The authors Note. The stages in green are carried out in the production areas while those in orange are carried out in nearby departments or in San Pedro Sula. Preparation for export Exporters store the coffee and, by threshing it, transform dry parchment into green or raw coffee. After classifying it mechanically or manually by weight, size (depending on client specifications) and colour, it is bagged and ready to be shipped. This stage is sometimes extended to as late as September of the following year for commercial reasons State of coffee cultivation in the project area The three project area departments represent 16.78% of the national producers, 16.57% of the country s cultivated area and 12.51% of the national production. Average yields in this area are lower than the national average (10.25 quintals per manzana compared to at the national level 2 ). Nevertheless, there are municipalities in these three departments where the productivity is much higher, such as San Francisco de Yojoa at Cortés, with more than 20 quintals per manzana (approximately 640 kilograms per hectare), or close to Ocotepeque -the most productive department in the country - with more than 23 quintals per manzana (736 kilograms per hectare). As can be seen in figures 6, 7 and 8 3, the department of Santa Bárbara is outstanding for both the number of farmers and for its cultivated area and production. Additionally, the number of farmers in this department has grown in recent years, with a total increase of 28% from the harvest to that of At Cortés and Atlántida, however, the number of farmers has decreased kilograms per hectare compared to 434 kilograms per hectare 3 Note: Atlántida figures are very much below those of Santa Bárbara and Cortés, so they are shown as zero in the figures. For instance, there were 185 farmers in 2010/2011, compared to 1,817 in Cortés and 15,048 in Santa Bárbara. 20

21 Figure 15. Evolution in number of farmers by department Source: IHCAFE, unpublished data, and calculations by the authors. In terms of cultivated area, the departments of Cortes and Atlantida show a sharper decrease than Santa Barbara, and this variable is the steadiest of those analysed. Considering how the number of farmers has increased in Santa Barbara, it can be concluded that the farms are being divided, in order to be sold to new producers or to pass on as inheritance. Figure 16. Evolution in cultivated area by department (manzanas) 4 Source: IHCAFE, unpublished data, and calculations by the authors. Production appears to be decreasing in all three departments, most markedly in Atlántida, after several years of more or less sustained growth in this department. This information, however, must be viewed with caution. According to some experts consulted, part of the production may not have been declared. This would explain the sharp fall in the figures corresponding to the campaign in Santa Bárbara. Figure 17. Evolution in production by department (quintals) Source: IHCAFE, unpublished data, and calculation by the author. As observed in Figure 18, Santa Barbara is also the most productive department. It shows an increase in yields starting from the campaign, followed by a drop in due to a sharp decline in production, which, as was mentioned, may not be real. Whatever the case, the yields obtained in the campaign have not been equalled. 4 1 manzana = 0.7 hectares 21

22 Figure 18. Evolution in yields by department (quintals per manzana 5 ) Note: Yields are calculated by cultivated area, not by area under production. Source: IHCAFE, unpublished data, and calculation by the author. The evolution observed over the last 12 years indicates that, while Santa Barbara is the most important department in terms of production, number of farmers and cultivated area, these variables behave unevenly. This makes it impossible to identify a clear tendency regarding productivity and production Actors in the chain Growers: 98.40% 6 of farmers are small-scale, which means that their fields are smaller than 14 hectares 7. They occupy 81.4% of the total surface and account for 80.9% of production, with yields of approximately 13.6 quintals per manzana (435 kilograms per hectare). These farmers work with traditional methods, so they do not use technologies such as soil or leaf analysis; most of them do not use certified seeds; their use of pesticides is low; the workforce is more the family than external; most of their fields are adult (although the good prices in recent years have stimulated some renovation); they lack of financial power (they depend on middlemen and input providers for financing); and they own limited productive infrastructure, so the quality offered is very mixed due to inadequate post-harvest handling. Nevertheless, small-scale farmers spend more time on their farms because of their small size, the fact that they live on the farm itself and their dependence on the crop as their source of income. As a result, the yields they obtain are higher than expected considering the conditions described above. Medium-scale farmers cultivate fields of between 14 and 52 hectares (over 20 and 75 manzanas), represent 1.5% of the total and own 13.2% of the cultivated area. They are responsible for 14.5% of production, with yields of 14.9 quintals per manzana (477 kilograms per hectare). Their better technical, productive and handling conditions explain this higher yield compared to small-scale farmers. Finally, large-scale producers 0.1% of the total own farms larger than 52 hectares (75 manzanas), account for 5.4% of the total land, 4.6% of the production and obtain average yields of 11.3 quintals per manzana (362 kilograms per hectare). Within this group we should distinguish between large and very large farmers those owing more than 300 manzanas (210 hectares). In general, large-scale producers have characteristics similar to medium-scale producers, but with higher financial capacity. This implies better supplies and management, reaching yields of 15.1 quintals per manzana (483 kilograms per hectare). However, the very large-scale producers yields are a lot lower, even lower than small-scales. This is probably explained by the fact that in most cases coffee is not their main source of income and they do not cultivate the whole farm, only a portion of it. 5 1 quintals per manzana = 32 kilograms per hectare 6 This data and the next in this section correspond to the national average, unless otherwise indicated, according to unpublished IHCAFE data. It is assumed that they behave similarly in the three departments studied. 7 The criteria for segmentation on small, medium- and large-scale producers follow FLO Cert and C.A.F.R. Practices (Starbucks) parameters, with small-scale farmers having less than 12 hectares, medium-scale having between 12 and 50 hectares and large-scale more than 50 hectares. Nevertheless, producer segmentation data are only available in manzanas, not hectares. The intervals used by IHCAFE in manzanas make it necessary to take 14 hectares (20 manzanas) as the limit for small-scale farmers and 52 hectares as that of medium-scale farmers. 22

23 Therefore, the biggest group of farmers, in terms of both surface area and production, is the one made up of small-scale farmers. At the same time, they are the group most vulnerable to the potential negative impacts of climate change. This is due to their dependence on the crop, their lack of financial capacity and, as a consequence, their limited resilience to adverse situations that may damage or ruin their crops. According to the experts consulted with regard to associativity, around 40% of farmers are members of some type of association, mostly cooperatives. This does not respond to a perception of associations as the best way to organize and collaborate, but more as a channel for receiving services and financing, given that certain development and financial institutions require the beneficiaries of their programmes to be legal entities. On the other hand, a culture of individualism and tradition of limited cooperation prevails in this sector. A common fact supporting this statement is that although the farmer has committed to delivering his entire production to the cooperative, he sells to a middleman if offered a better price, only delivering the quantity to the cooperative that is necessary to pay back his debts to it, if he has any. Neither is there a cooperative culture among cooperatives themselves. There are no second floor cooperatives in this area, even though many of the existing cooperatives take part in higher-level initiatives, such as the fair trade cooperatives committee, which hasn t got any legal standing. There is some degree of competition to attain greater production percentages. Milling facilities A milling facility collects the coffee cherries in order to pulp, ferment, wash and pre-dry them, obtaining wet parchment coffee. There are 7 mills counted in Atlántida, all of them small 8 ; in Cortés there are 2 large mills and 8 medium ones; and in Santa Bárbara there are around 100 mills: 4 of them large, 60 medium and 36 small. They generally belong to cooperatives, are located near the production area, and have limited organisational and financing capacity. As stated above, part of the milling (30%) is performed by growers on their own farms. Of the remaining 70%, 30% is processed by small mills, 60% by medium and 10% by large ones. Local middlemen There are three big groups of local middlemen. On the one hand, the individual traders that operate legally; on another, the physical persons who operate as their suppliers but who are not officially registered; and finally, some companies constituted as intermediaries. The companies collect approximately 5% of the total coffee managed by local middlemen and the remaining amount goes through individual traders. For these reasons, it is difficult to know how many local middlemen operate in the study area. It can be affirmed, however, that they usually receive the coffee milled by producers, because the coffee milled in milling facilities goes directly to regional intermediaries or exporters. 40% of coffee milled by growers is collected by local middlemen. Very often, the farmer pays back his debts to the middleman by delivering coffee, which helps the middleman secure a significant volume and meet his commitments to the exporter, who will in turn require debt repayment from the middleman in coffee instead of cash. These local middlemen only collect wet parchment, and the volume of coffee they can deliver, as dry parchment is not significant. 80% of the coffee they handle is delivered wet to regional intermediaries, and the remaining 20% to exporters. Local middlemen, who usually live in the production areas, may also be growers. In general, they have limited financing capacity (they depend on the exporter for financing), have vehicles for transportation and do not have specific professional training. They usually work exclusively with coffee, although exceptionally they may have other minor dealings. Local middlemen work independently, and examples of collaboration have not been found. On the contrary, through the prices and financing they can offer growers they compete to collect coffee. 8 Those processing less than 1,000 quintals of wet parchment per year are considered small, between 1,000 and 5,000 quintals are considered medium and those processing more than 5,000 quintals are large. 23

24 Regional intermediaries Most regional intermediaries are individual traders (60%), located in municipal and departmental capitals, or at sites that are strategic due to their central location relative to the different production areas. Grower cooperatives sometimes act as regional intermediaries, collecting and drying coffee, a role played by companies as well. In any case, regional intermediaries receive coffee from growers, wet milling facilities and local middlemen. As indicated above, 45% of the coffee milled by growers goes directly to regional intermediaries, and another 32% arrives indirectly through local middlemen. Additionally, they receive 15% of the coffee processed at milling facilities. In Santa Bárbara, there are around 48 regional intermediaries and 3 cooperatives performing this function. Approximately 24 regional intermediaries operate in Cortés (including Atlántida). Sometimes they do not only operate in these departments, but also in others. These actors are mainly characterized by greater financial capacity and solid credit guarantees. They often work exclusively with coffee, although some of them also deal with general trading activities and finance growers by harvest, among other things. Given their limited storage and drying capacity, they are forced to sell part of the collected coffee as wet parchment. Just as occurs among local middlemen, a certain degree of competition based on prices has been observed among regional intermediaries. Nevertheless, they have recently created a national association with regional representatives, which may contribute to achieving better collaboration within this link of the chain. Exporters There are 50 exporters operating in the country and they may or not be located in the departments studied (close to 50% are in the project area), although most of them operate in San Pedro Sula. They are transnational and national companies or producers groups registered as exporters. Individual traders are almost absent, given the financing capacity required to operate. The exporter, as was mentioned before, collects, dries and processes the coffee through the green (raw) coffee stage. There is virtually no roasting capacity in the country, given that clients prefer to buy green coffee because it is less perishable than roasted coffee. The exporter is the main financial backer for the chain activities. It provides financial resources to intermediaries, who in turn make loans to farmers, almost always with higher rates than those applied by financial institutions. Exporters are organized under a national association, ADECAFEH (Honduran Coffee Exporters Association). There are some examples of collaboration among exporters, always commerciallyoriented (for instance, when an exporter does not have the capacity to respond to an order, he may outsource part of the process). There is also high competition in this part of the chain. Exporters compete based on prices, focusing more on volume than quality, with harmful consequences to the sector because farmers are not as sensitive to quality as would be desirable. Other actors There are other actors involved in the coffee value chain, including input suppliers, financial entities, research centres, service and technical assistance companies, etc. Along with them, there are institutions and projects linked to this product. This section contains an overview of these actors in order to determine the extent to which their availability is adapted to the sector s support needs. First, there are different equipment and input importers operating in the study area, which supply local providers of agro-services. Among them are Fundidora del Norte, SEAGRO, DISAGRO, FERTICA, Atlántica Agrícola and BAYER. To judge by the presence of points of sale, the availability of these elements appears to be sufficient, although the farmers ability to afford their cost is a different question. It has been already mentioned that this segment of small-scale farmers lacks the financial capacity to purchase the inputs recommendable for their activity with their own means. Concerning financial institutions, as discussed below, their role in this value chain is marginal 24

25 because financial resources tend to be provided by the value chain actors themselves. Nevertheless, over recent years it has been observed that some regional intermediaries are applying for formal credits, turning to the exporter as a last resort. Only medium- and large-scale growers apply to financial institutions in order to finance their investments, unlike the smallscale farmer, who usually resorts to self-financing, which limits his investment capacity to a great extent. The main reason for avoiding financial institutions lies in the guarantee requirements that they impose, given that small-scale farmers lack the necessary backing to use this less costly alternative. There are also institutions providing quality technical assistance to coffee growers, with IHCAFE playing an important role. Exporters also provide technical assistance for field certification. In addition to that, some development projects in the area, financed by AECID or USAID, complement the offer of technical assistance services. In other words, there is some technical assistance available in the area, although the shortage of specialized technical staff prevents it from reaching every farmer. Relations between the links in the chain Most negotiating power is concentrated with the exporter, because it finances every activity in the chain and imposes its conditions concerning quantity, quality and schedule on the other actors. Producers make up the weakest link. Their intense competition makes them virtually pricetakers, although they may partially take advantage of the competition between intermediaries and exporters, who play with slight differences in offered prices in order to grab production. Producers, therefore, do not sell their product: it is bought. Indebted to other chain actors (intermediaries or exporters) instead of financial entities for financing, growers are tied to the financer-buyer and have limited capacity to look for better offers. Something similar happens in the relationship between producers and cooperatives, although in this case the farmer has a more favourable position than with the intermediary because the cooperatives tend to offer better financing terms and more transparent prices, allowing the grower to decide whether to sell through this channel. As for the links between farmers and input suppliers, these last also concentrate negotiating power. The farmer, not being associated, must accept the price offered by the company. A cooperative has better bargaining power in the face of the supplier, thus benefitting the farmer in the end. 3.2 Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios A critical element is defined as any aspect of the crop cycle or further phase in the value chain that shows significant sensitivity or vulnerability to climatic events, meaning that the impact of climate change on the product s quantity or quality could be considerable. Fertilization, for instance, would be a critical element: if rainfall is insufficient to facilitate good access of nutrients to the plant at the right time, fertilization is less effective and the number of beans produced will be lower. Another example of a critical element is the number of flowers that bloom on the tree. This depends on the humidity and temperature experienced by the plant at certain moments, among other factors. The number of flowers determines the quantity of fruits, and therefore the volume of production that will be obtained. The selection of critical elements shown below was arrived at by analysing the value chain activities, particularly those of the crop cycle, in order to identify the phases most sensitive to climate. Additionally, a bibliographic review of similar studies performed in other countries was done and some national experts were consulted in order to validate the critical elements and jointly determine how they are affected by climatic events. It is essential to consider the moment that a climatic event occurs in relation to the phases and activities in the value chain, because a delay or advance usually has important implications on 25

26 production. For instance, if a given climatic event means a large quantity of the product ripens in a short period of time, this will affect the producers capacity to harvest and dry the coffee, having an impact on the final price. Therefore, this analysis considers the time when a given climatic event occurs, along with its effect on the critical elements identified in the value chain. The value chain has been divided in the following phases in order to identify its critical elements: flowering, ripening, harvesting and post-harvest. The main critical elements, the indicators that will make it possible to measure how this critical element will be affected and the impacts of future scenarios corresponding to each indicator are presented below. As mentioned before, it is important to consider the following climate projections 9 : Significant changes in the current rainfall pattern are not expected, so indicators dependent on rainfall alone will not undergo important changes. Changes in temperature will be very significant, so indicators depending on temperature will undergo important changes and will be essential to explain impacts on the coffee value chain Flowering stage Two critical elements were identified in the flowering phase 10 : 1. Flowering induction (onset of flowering) 2. Flowering quantity (the number of flowers) 1. Flowering induction The onset of flowering is important because it affects the subsequent phases in the chain. If flowering takes place earlier or later, the subsequent phases will also occur earlier or later, and at that time they may lack the temperature and rainfall conditions necessary for proper development. On the other hand, a delay in flowering may also mean later harvesting, with the consequent impact on the coffee price received by the grower. National experts consulted agree on the ideal conditions for the onset of coffee flowering in Honduras: average temperatures of between 14 and 25ºC and a light rainfall event that produces a relative humidity above 70%. Blossoming may be triggered by just one light rainfall event and adequate temperatures. Once it has started, coffee blooms every 15 days approximately. Experts concur that flowering takes place in the project area between March 1 st and June 1 st, when temperatures are within the mentioned range and there is light rainfall. The indicators that were identified, analysed and verified with experts by applying observed Flowering Induction Indicator (IIF1): First day between March 1st and June 1st on which the average daily temperature is between 14ºC and 25ºC and on which there is 3 mm of rain meteorological data are as follows: This indicator is heavily influenced by rainfall and to a lesser extent by temperature, given that a range of 14-25º is very common at this time of the year in the areas analysed. It also suggests a very early flowering date (the beginning of March), very likely because temperature and rainfall are not restrictive conditions. For these reasons, use of this indicator is not recommended for impact analyses. 9 Virgilio enamorado (in Santa Barbara) and Santa Rosa de Copan have been the only two observatories with available temperature observations for coffee. Rainfall data have been also provided from TRMM grid points in Santa Barbara: Callejones (R2_05), Macuelizo (R2_06), SO Santa Barbara (R3_04) and NE San Luis (R3_05); and in Cortés: East Yojoa Lake (R5_04). Stations and grid points used are shown in annex III. 10 As pointed out by Arcila Coffee flowering is strongly linked to regional climatic conditions and usually is registered as the time of anthesis, when flowers open. Nevertheless, it is important to consider that flowering is a complex development process that starts from 4 to 5 months before the flowers opening (Arcila, 2007:37). 26

27 Flowering Induction Indicator (IIF2): Tenth day after an accumulated rainfall of over 5mm once the PET accumulated since October 1 st has reached 350 mm In this case, potential evapotranspiration (PET) will determine flowering 11 induction, so that once a certain level of PET has been reached, a slight rainfall is enough to trigger anthesis. A tendency to flower earlier is clearly observed, both in Santa Rosa and Virgilio Enamorado (Santa Bárbara), as figures 19 and 20 show. Figure 19. IIF2 evolution foreseen in Santa Rosa de Copán Figure 20. IIF2 evolution foreseen in Virgilio Enamorado Flowering Induction Indicator (IIF3): Tenth day after accumulated rainfall exceeds 5 mm once 1600 day degrees have been surpassed This indicator is based in the fact that the plant needs an accumulated temperature amount for 11 Studies performed in Brazil (Camargo and Camargo, 2001; Zacharias, Camargo and Fazuoli, 2008; and Camargo, 2010) show that, in the Arabica variety, temperature rises and a reduction in rainfall are associated an earlier rupture in the flower buds dormancy period. When potential evapotranspiration (PET) starting from the March equinox (shorter days) reaches a certain level (around 350 mm), followed by approximately 7 mm of rain, flower buds complete ripening and become prepared for anthesis (blossoming). This criterion has been applied to Honduras, where studies allowing verification of the PET level for flowering induction are not available. The Thornthwaite method was also applied to identify existing trends, taking the latitude and different daylight distribution into account. 27

28 flowering 12 induction and, as with IIF2, once this accumulated temperature has been reached a light rain is enough to cause anthesis. In this case acceleration is even greater, as observed in figures 21 and 22. Figure 21. IIF3 evolution foreseen in Santa Rosa de Copán Figure 22. IIFE evolution foreseen in Virgilio Enamorado When applying IIF2 we may conclude that by mid century flowering would take place earlier, by the first week of March, while with IIF3 it could even occur at the end of February, under intermediate emissions scenarios such as the Norwegian with RCP 45. As a consequence, harvesting would start earlier, so in areas below one thousand metres above sea level harvesting would finish by September and in highlands by December. This implies harvest difficulties due to the likely concentration of production. Costs would possibly increase because of the need to engage labour and transportation in a shorter time, the drying process would be more complicated due to the lack of adequate facilities to receive more volume in less time, etc. Additionally, the harvest period could coincide with a rainier period and with the school season, implying worse conditions for performing this task. 2. Flowering quantity Although several drivers influence the volumes of coffee produced, as analysed below, the quantity of flowers is the first driver (in chronological order) influencing the quantity of beans. Among other factors, flowering will depend on: 12 Jaramillo suggests the need for accumulated thermal time (Jaramillo, 2005) and Zacharias establishes a correlation between PET and day degrees (Zacharias, Camargo and Fazuoli, 2008) 28

29 a) the plant experiencing water stress at the right time. According to the experts consulted, at least 30 days of water stress per quarter are required for flower buds to mature and during dormancy 13. This means that the rainier the winter, the less flowering. Therefore, the first indicator formulated is as follows: Quantity of Flowering Indicator (ICF1): Quantity of accumulated precipitation between January 1 st and March 1 st divided by 150 mm If during this period the rainfall exceeds 150 mm the quotient will be over 1, and it is considered that the water stress conditions are not sufficient for the plant to blossom adequately. The ICF1 indicator scenarios show that significant changes in rainfall are not foreseen during this phenological stage. Therefore, negative evolution associated with rainfall and humidity is not expected, as observed in figure 23. Figure 23. ICF1 evolution foreseen in Santa Rosa de Copán As rainfall will be stable and evapotranspiration will increase (see figures 3 and 4 for IIF2), plants will be more frequently exposed to water stress during dormancy, which will have a positive effect and prepare the plant for flowering. b) the plant having enough water during the last stages of flowering for anthesis to occur. For this, a certain level of humidity is required from March to May, so if there is little rainfall and temperatures are high it will be harder for the plant to blossom. ICF2 is formulated this way: Quantity of Flowering Indicator (ICF2): Number of days between March 1 st and May 30 th with daily average temperatures above 24ºC and where accumulated rainfall on that day and the four previous days is under 2 mm The different scenarios show the negative impact of temperature changes when the plant is moving from dormancy to anthesis. The ICF2 indicator shows significant results, which means more days with temperatures above 24ºC and drier conditions (figures 24 and 25). This may generate worse conditions for flowering, affect flowering quantity and, ultimately, the quantity of product obtained. 13 Period during which growth, development and physical activity are temporary suspended. 29

30 Figure 24. ICF2 evolution foreseen in Santa Rosa de Copán Figure 25. ICF2 evolution foreseen in Virgilio Enamorado Furthermore, an increase in the frequency of events when temperature is above 24ºC during flowering may increase the frequency of star flower. c) frequency of the star flower phenomenon, which takes place after flowering when relative humidity is below 20% and the temperature is above 24ºC. Here, the flower is lost as it is not pollinated due to a lack of humidity or, if it is pollinated, the fruit may also be lost during its initial setting stage ( pimentilla ). Star Flower Indicator (IF*): Number of degree-days accumulated above 24ºC of daily average between days 1 to 30 after flowering. Note: Despite the three Flowering Induction indicators (IIF1, IIF2 and IIF3) were used in the analysis, only IIF2 and IIF3 were used to estimate the star flower due to the greater reliability of their estimations.. Scenarios for this indicator show an increase in star flower situations with data taken from Santa Rosa de Copán observatory, as shown in figures 26 and

31 Figure 26. IF2* evolution in Santa Rosa de Copán Figure 27. IF3* evolution in Santa Rosa de Copán The potential incidence of star flower seems to be lower if IIF3 is used, given that at the beginning of the year temperatures above 24ºC are less likely and it is therefore less probable that star flower will occur. Figures show that the number of accumulated degree days could double (IIF3) or even triple (IIF2) by 2050, so the percentage of star flower will therefore increase. As a consequence, the quantity of viable flowers will decline and production will drop. This decrease in production comes in addition to the drop reflected in the previous indicator, and while they reduce the incidence of harvest accumulation projected by IIF2 and IIF3 because there will be less beans to harvest they will probably not prevent the problems related to earlier flowering induction. d) stormy weather during flowering, because strong and continued rainfall can cause a significant number of flowers to fall. While there are no records of this having occurred, the experts consulted agree that it could have a very negative impact on the quantity of fruits generated and, accordingly, on final production if it happened. Storm-caused Flower Fall Indicator (ICFT): Maximum accumulated rainfall (in millimetres) in 7 consecutive days from March 15 th to May 30 th divided by 300 mm. As it deals with maximum precipitation, this indicator locates the largest storm that has taken place during the period considered (a storm is defined as any precipitation greater than 300 mm for 7 consecutive days). If the indicator exceeds 1 it means that at least one storm has taken 31

32 place, and the higher this value, the more intense the storm. As regards the ICFT indicator, scenarios suggest that storm events are not expected within the flowering stage, as is currently the case, according to the observed data in figure 28 for Santa Rosa de Copan, with a light and non significant increase and always below 1. Figure 28. ICFT Evolution foreseen in Santa Rosa de Copán Ripening stage Two critical elements were identified in this stage: 1. the timing of fertilizing 2. the level of vulnerability to pests and diseases 1. Timing of fertilizing The plant needs nutrients to mature. If they are not available at the right time, it may abort part of its fruits, maintaining only those that are viable. Fertilization or the supply of nutrients is essential to achieve an adequate number of fruits, but the soil must have a minimum level of humidity for them to be assimilated. A delay in primera rainfall may mean that the necessary humidity level will not be reached by the first fertilization, which takes place between May and June. The indicator to measure the probability of fruit abortion analyses the changes in the onset of the winter ( invierno ) or the primera rainy season. Fruit Abortion Indicator (ABO): Starting from April 1 st, the first day (with more than 2 mm of rainfall) when a cycle begins of 4 days of accumulated rainfall that is more than twice the April-May period s daily average The criterion applied is that the rainy season begins sometime after April 1 st, when the first rainy day is followed by 4 days with some level of precipitation (not counting any isolated rainfall events that don t necessarily represent the onset of the rainy season). ABO scenarios remain stable over time, so the onset of the rainy season is not expected to come earlier or later in either Santa Rosa or Virgilio Enamorado (figures 29 and 30). 32

33 Figure 29. ABO evolution foreseen in Santa Rosa de Copán Figure 30. ABO evolution foreseen in Virgilio Enamorado Very significant changes in the onset of the rainy season are not, therefore, expected. However, an earlier onset of flowering, as suggested by the previous indicators, would imply the need for earlier fertilizing to ensure that nutrients get to the fruits at the right time to avoid abortions. This fertilization, however, will not be as accessible to the plant, given that the onset of the first rainfalls of the season is not projected to occur earlier. Therefore, there may be nutrient deficit problems if measures to mitigate this impact are not implemented. 2. Vulnerability to pests and diseases During ripening, the plant is vulnerable to pests and diseases that reduce the amount and quality of beans and force the grower to allocate resources in order to combat them. Excessive humidity along with high temperatures favours the emergence and proliferation of fungi, while some of them, such as the rust, also thrive where there is a lack of moisture. Moreover, if the temperature is not low enough during the coldest months (November to January), vulnerability to diseases also increases. Three indicators have been developed to measure this. The first two indicators measure the excess of humidity linked to high temperatures which is the main cause of coffee plant diseases in the project area. The first indicator specifically measures vulnerability to rust, because this fungus spreads when the temperature is above 20ºC and the weather is wet: Vulnerability to Diseases Indicator (SUS1): Maximum accumulated rainfall in 30 days between July 1 st and October 30 th with average temperatures above 20ºC, divided by 300 mm. 33

34 When this indicator exceeds 1 it means that the average temperature has been above 20ºC and more than 300 mm have accumulated in 30 days. At the two analysed observatories a slight but not very significant increase has been observed. Figure 31 shows the example of Virgilio Enamorado. Figure 31. Evolution foreseen in Virgilio Enamorado A second indicator measures the conditions favourable to coffee diseases and has a higher temperature requirement (>24ºC): Vulnerability to Diseases Indicator (SUS2): Number of days between July 1 st and October 30 th when daily average temperatures are above 24ºC and accumulated rainfall during that day and the four previous days exceeds 30 mm. The scenarios applied to this indicator show very significant changes. For instance, by 2050 in Santa Rosa de Copán the number of days when these conditions are met will increase from 3 to 15 (for intermediate RCPs) and therefore vulnerability to diseases including rust and other fungi will be much higher, as well as to pests such as the berry borer and other insects. This will cause higher harvest losses and increased costs to deal with those pests and diseases (figure 32). Figure 32. SUS2 evolution foreseen in Santa Rosa de Copan A third indicator to consider is related to the absence of cold. The cold stops development and spreading of fungi, which means that if the temperature does not drop enough during the coldest months, this phenomenon does not happen. Indicator of Pests and Diseases Due to Absence of Cold (SUSH): "Number of degrees by which the average temperature in the coldest natural month (30 consecutive days) exceeds 18ºC" 34

35 The higher the indicator value, the more likely than pests and diseases will remain in viable condition during the coldest months. The scenarios generated for SUSH indicator also show very significant changes in Santa Rosa and Virgilio Enamorado, as shown in figures 33 and 34. Figure 33. SUSH evolution foreseen in Santa Rosa de Copan Figure 34. SUSH evolution foreseen in Virgilio Enamorado The conclusion drawn is that vulnerability to pests and diseases will tend to be higher; fungal, viral and bacterial diseases will be more frequent, strong and widespread, as will insect infestations and, consequently, the impact on both production volume and quality will be significant Harvest stage In this stage, the main critical element is related to the last phases of coffee ripening, so that when ripening accelerates the harvesting period becomes shorter and the harvested beans accumulate. The harvest accumulation occurs when the dry season starts earlier. When ripening accelerates, the grower: must find additional workforce, which causes an added expense in salaries. must cover additional transportation costs, because the needs are higher in less time. has to sell the harvest quickly if he/she does not have enough capacity to dry or store the grain, thus obtaining a lower price as he/she has less negotiating power. Moreover, producers must deal with an excess of supply if there are more producers in the same situation (which is very likely in the same area). 35

36 The following indicator measures the harvest accumulation: Harvest Accumulation Indicator (ACOS): First date starting from October 1 st when during 8 consecutive days the cumulative rainfall is equal to or less than 2 mm and the average temperature during the 8 days exceeds the average climatic October- December temperature This indicator measures the onset date of the dry season. The dry season is considered to have begun when there are 8 consecutive days with almost no rain and with high temperatures; if this happens earlier, coffee ripening accelerates. The later this happens, the more gradual the harvest will be, causing fewer problems for the producer. The ACOS indicator shows that the harvest accumulation will tend to be slightly less, although the changes are not very significant (they seem to respond more to the absence of data during the first years of the historical period). Figure 35. ACOS evolution foreseen in Santa Rosa de Copán On the other hand, when there is excessive rainfall the growers may face problems related to the transportation of the harvested beans due to bad road conditions. The following indicators have been formulated for this: Harvest Conditions Indicator (ICC1): Maximum accumulated rainfall (mm) in 7 consecutive days between December 15 and January 15 divided by 70 mm. If during the harvest it rains 70 mm per day for 7 days, the soil cannot drain the water, complicating both the harvesting and the beans transportation. The farms are usually located in the highlands and the state of the roads makes them sometimes impassable. This indicator is unfavourable when it is above 1 and favourable when it is below 1. Indicator ICC1 does not seem to change significantly, as shown in figure 36. However, considering that flowering will occur earlier, harvesting could also move forward and coincide with a rainier period, which would mean facing transportation problems. 36

37 Figure 36. ICC1 evolution foreseen in Virgilio Enamorado Harvest Conditions Indicator (ICC2): Number of days with minimum temperatures above 19ºC and rainfall greater than 10mm between December 15 and January 15 The ICC2 indicator contemplates nighttime temperatures more frequently above this limit, which means more harvest accumulation in December and January and a more complicated task. This accumulation of harvested volume complicates the collection and further processing of the beans: part of them may deteriorate if they are not processed promptly and appropriately. Along with this, a higher incidence of bacteria may provoke the unpulped beans to ferment more, both in the tree and once harvested. Figure 37. ICC2 evolution foreseen in Virgilio Enamorado Post-harvest stage During the post-harvest stage, the most relevant critical element has to do with pre-drying in the production areas in order to extract as much humidity as possible, as required by the stockpiler. As explained earlier, this process is carried out by the grower in his own backyard ( patio ) or at wet-milling facilities. In the event of rain, even a little, the coffee cannot be outdoors because it may become wet and the quality may be affected. Furthermore, during the fermentation process fungi are produced. In a limited quantity they are beneficial because they add aroma, but in higher quantities they reduce the coffee s quality due to unwanted fermentation. High temperatures and humidity favour fungi development. If there is a problem of excessive humidity and fungal infection, it s difficult for the stockpiler to resolve 37

38 them because he has no ability to handle the product under these conditions. Two complementary indicators have been verified and applied for the pre-drying stage: Pre-drying Indicator (SEC1): Number of days with more than 1 mm of rain between December 1 st and February 28 th This indicator is intended to measure the trend toward a higher or lower number of days that complicate the pre-drying process, so it counts the number of problematic days, those when it rains even a minimal amount. Scenarios for SEC1 do not show significant changes, although a slight increase in the rainy days during this period has been observed (just a few days in 3 months, see figure for Santa Rosa and grid point 04_07). Figure 38. SEC1 evolution foreseen in Santa Rosa de Copán Figure 39. SEC1 evolution foreseen at grid point R04_07 Pre-drying Indicator (SEC2): Number of events (periods) of at least 8 consecutive days when accumulated rainfall is below 2 mm between December 1 st and February 28. This indicator takes the opposite view: if there are several virtually rain-free shorter or longer periods, the producer will dry the beans without problems. It is assumed that 8 days are enough to dry a batch. 38

39 The SEC2 indicator also shows stable evolution with a slight tendency to decrease, so it is assumed that conditions will remain more or less as they currently are. Notwithstanding the above, earlier flowering may also cause the harvest to move up, affecting the timing of drying and moving into potentially more rainy periods, such as November or the beginning of December. Figure 40. SEC2 evolution foreseen in Santa Rosa de Copán 4. COCOA VALUE CHAIN 4.1 Mapping: process and actors Value chain steps and scheduling The cocoa value chain is less complex than the coffee value chain. The activities are more concentrated in certain actors and there are less direct links in the commercial stages. In this case, the complexity lies in crop management, harvest and post-harvest. The central link of the cocoa chain is made up of the phases shown in Figure There are other activities supporting the value chain, including technical assistance, institutional support, etc., that are part of the cocoa value chain along with the central link. Figure 41: Cocoa value chain links. Source: The authors. 14 Only the first two phases (in orange) are the subject of this study. 39

40 Production We assumed that the plants are more than three years old and in a productive phase. In fact, in the project area there is a high percentage of very old trees: 45% of plantations in Atlántida and 23.3% in Cortés are more than 24 years old (Mejía and Canales, 2010:11). This is different from Santa Bárbara, where almost all plantations are new. It must be taken into account that plant age is a negative factor concerning productivity and vulnerability to pests and diseases. Figure 42: Productive cycle scheduling in Northern Honduras. Source: The authors. The cocoa productive cycle consists of three main phases: flowering, ripening and harvesting. There are two harvest periods per year: the first one from approximately September to January and the second one from February to June. In Atlántida, the first one accounts for 35% of the total and the second one for 65%. In the Cortés area, the first harvest accounts for 70% and the second one for 30%, gradually tending to decrease. The crop is not usually fertilized, and the main cultivation practices are limited to manual weed control, called comaleo (75% of farms) in order to make harvesting more efficient; and pruning (98% of farms) after harvesting, to renew productive tissue and eliminate useless shoots (chupones) (Mejía and Canales, 2010:13). In the event of pests or diseases, crop control consists of collecting the affected fruits. The risk of pests increases with rainfall and cold temperatures. When the cocoa pods are ripe, the farmer cuts them, extracts the beans and transports them to the local points of sale, usually using his own beasts of burden, while some organisations support producers by transporting the product to the stockpiling centres in their vehicles. Post-harvest and processing The stockpiler can be an export intermediary or a producers organization like a cooperative. It collects the beans from farmers and supplies them to the processor. Figure 43: Cocoa processing stages. Source: The authors. After drying, the beans are classified and stored for transport to the exporter centres, which deal with storage and sales to destination countries. Processing mainly takes place in the intermediaries and exporters facilities located in San 40

41 Pedro Sula (88%). The rest of the process (12%) takes place in stockpiling centres that belong to farmers organisations in the production areas. Figure 44. Cocoa value chain, simplified diagram. Note. Stages in green are carried out in production areas while those in orange are carried out in San Pedro Sula. Source: The authors State of cocoa cultivation in the project area Cocoa behaviour in Honduras has been very unstable in recent years. This is due to natural disasters which affected coastal cultivation areas, mainly hurricane Mitch, along with diseases including monilia disease. The crop has been recovering, with new plantations and restoration of old plantations. However, it has not yet achieved pre-mitch production levels, when Honduras produced more than 5,000 tonnes annually and was the foremost export country in Central America. Most plantations are concentrated in five departments in the Atlantic region: Atlántida, Cortés, Gracias a Dios, Santa Bárbara and Yoro, producing an annual average of 1,000 tonnes of cocoa that are exported to the United States, Central America and Switzerland.15 However, there is a lack of reliable data about the number of farmers, production volumes, cultivated area, etc. Sources consulted offer unequal and even conflicting figures. To compensate for this, we have chosen to use data provided by APROCACAHO, FHIA and FUNDER as our main source. Figure 45: Cocoa producing departments in Honduras Cocoa nurseries Cocoa plantations Source: Orlando Mejía and Marlon Canales (2010) 15 Switzerland has been the leading buyer since

42 In Atlántida, there are approximately 380 farmers (26.2% of the three-department total), 40 of them certified as organic. In Cortés there are about 800 (55.2% of the total) with 160 certified organic, and there are 270 in Santa Bárbara (18.6%) including 175 certified organic. Thus, according to these figures, there are a total of 375 farmers with organic certification, 25.9% of the total in the three departments. Low average farm size makes the use of fertilizers not very cost-effective, and this, along with the use of low-input cultural practices, facilitates obtaining organic certification to a great extent. Certification levels have also been driven by the growing interest and demand for this type of certification in final destination markets Actors in the chain Growers As stated above, there are around 1,450 farmers in the three departments, most of them smallscale (95%), who manage farms smaller than 3 manzanas (approximately 2.1 hectares). Farm size prevents the producer from depending exclusively on cocoa, so this crop is usually combined with other activities. It is estimated that 30% of the farmers income comes from this product; 50% from other crops, essentially basic grains for personal consumption; and the remaining 20% from selling their labour power. Moreover, activity is focused on basic grains during certain periods of the year (June and October, respectively). This means that cocoa gets limited attention and poor management, so farmers obtain low yields as a consequence. Other factors, including the weather, soil types or the varieties cultivated may also contribute to this. These farmers are characterized by their use of traditional practices. In other words, they do not apply technologies such as soil or leaf analysis, they plant crop stock that is not resistant to pests and diseases and tend to concentrate the harvesting time in short periods. There is some division of tasks by gender: women are involved in harvesting, plucking and opening the pods and drying the sap, while men deal with carrying the bags. With adequate training, women could have a more relevant role and also participate in decision-making. In general the plantations are mature, farmers lack access to financing, the productive infrastructure is poor and, as a consequence, quality is very uneven due to inadequate postharvest management. According to the experts consulted with regard to associativity, approximately 25% of farmers are members of some type of association, mostly cooperatives. In Atlántida, 10.5% of farmers are affiliated; in Cortés, 20%; and in Santa Bárbara 65% of farmers belongs to the cooperative COAGRICSAL, which has promoted cocoa in this department. The low overall participation can be explained because cocoa is not the main source of income for most producers, who just grow it and are not willing to invest more time in this product. Those who join probably expect to get more profits or receive some incentive for being organized (for instance, more support from institutions or international development agencies). Regarding cooperatives, there is a stronger culture of cooperation than among coffee growers. APROCACAHO plays a significant role in coordinating grower activity and supporting initiatives that may ultimately improve interrelations across the sector. Cooperatives are expanding their cultivation areas, so they do not grow at the expense of others, which reduces competition. Stockpiling centres 12% of production in the three departments goes through stockpiling centres. They receive cocoa in a sweet pulp stage (baba de cacao 16 ) from associated farmers and proceed to ferment and dry it. There is one stockpiling centre in Atlántida, three in Cortés and one in Santa Bárbara. They generally have poor drying facilities and lack capacity. Their mechanized drying systems sometimes are not validated for cocoa, which causes losses in quality and production. They also have significant organisational weaknesses. 16 "Baba de cacao" is a sweet pulp surrounding the cocoa beans. 42

43 Small middlemen Small middlemen are informal individual traders (unlike in coffee, there is no formal registry of their activity in the cocoa sector). These middlemen are local and establish a route in the production area to buy sweet pulp cocoa, mostly from independent farmers. Small middlemen sell untransformed cocoa to medium or large intermediaries. They receive financing from the exporter and usually pay it back with product. They work independently, and there is a high level of competition to collect cocoa through price. There are no collaborative experiences among them. Medium middlemen In the three departments there are over 10 medium-sized intermediaries, on a regional or departmental scale. They are also individual traders, with some financial capacity (and if they do not have it, they are financed by the larger intermediaries). There is some competition among them because they operate all over the coast, competing for the same farmers. Large intermediaries/exporters There are four in total in this region, and they are exporters as well: three individual traders and one company. They have good logistics thanks to the network of small middlemen they have set up in each village. They have a long history in this activity: they got established even before the producers organisations. These actors take care of fermenting and drying the beans sourced from medium or small intermediaries. Finally, they put the cocoa into bags and store it until it is traded, usually to destinations in Central America. So far, intermediary-exporters only compete with APROCACAHO for conventional cocoa markets in the Central American region, but future competition for international markets is expected. There are no association initiatives among intermediary-exporters and, given the limited offer of cocoa in Honduras, a high level of competition is assumed. Export organisations Currently, only APROCACAHO operates as exporter. Being a second-floor organisation, it receives cocoa processed by stockpiling centres and sells it. It is the only exporter that differentiates between high and regular quality cocoa. As the exporter, it also puts the cocoa into bags and stores it. High quality cocoa (qualities A and B) is usually destined to international markets (Chocolates HALBA), whereas regular cocoa is sold in Central America. Other actors Given the almost absolute lack of use of phytosanitary products (even pests and diseases are controlled traditionally), other actors linked to the chain have to do with support services and the regulatory framework. Regarding support services, institutions providing technical assistance to the sector include FHIA, FUNDER and TECHNOSERVE. Others provide training and technology transfer, including CATIE, PROCOMER, PRONAGRO-SAG, FHIA and international development institutions such as SWISSCONTACT and CHOCOLATES HALBA (which is a commercial link that provides assistance through HELVETAS). Additionally, other institutions provide certification and financing services, although with a less important role. There are also import companies, such as XOCO, that drive the production, storage and marketing of high quality cocoa in Honduras. Their role is focused on technology transfers to improve quality through creating stable relationships with farmers. They are rehabilitating some areas and establishing new ones. Links in the chain The greatest part of the negotiating power is concentrated in the exporter, because it finances every activity in the chain through the intermediary level (the farmer does not receive financing, 43

44 he/she sells for cash) and imposes its conditions on the intermediaries with regard to quantities and deadlines. As happens with coffee, farmers constitute the weakest link. The intense competition among them makes them virtually price-takers. They have slightly more negotiating power than coffee producers, because they are less dependent on financing and there is a surplus in export demand. Sales of their product are almost guaranteed. Most of them are not affiliated with any cooperative, so they do not work exclusively for one client. Both independent and organized producers depend on external support to renew or expand their plantations. For instance, XOCO supports independent growers, providing them with genetic material and some inputs to start the activity, subject to the condition that the farmer sells his full production during a certain period determined by the contract. In Santa Bárbara, independent farmers finance planting and renewal with their own resources (this department is where the most recent investments have been made in planting and, to a lesser extent, in renewal). Nevertheless, organized producers usually receive support from institutions and international development agencies, including FHIA-ACDI, which acts as an incentive for independent farmers to join an organisation. 4.2 Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios In order to indentify its critical elements, the value chain has been divided in the following phases: flowering, ripening, harvesting and post-harvest. A section has been added to these phases in order to evaluate crop viability along the whole cycle, as the plant requires certain temperature and rainfall conditions during the year for it to be viable and produce fruits. The main critical elements and selected indicators corresponding to each phase are explained below, along with the impacts of future scenarios for each indicator. For this analysis it is important to consider, as mentioned in section 2, that significant changes in temperature and some changes worth highlighting in rainfall will take place in some observatories areas of influence. In general, the detected changes in rainfall that will affect cocoa in the three stations of reference are as follows: in La Ceiba, rainfall will increase in June, July and August; in Tela it will increase in October and November; and in La Mesa in San Pedro Sula rainfall will increase from June/July to November The whole crop cycle Three critical elements were identified in the cocoa crop cycle: 1. minimum temperature threshold 2. differences in temperature 3. non-viability due to prolonged lack of precipitation 1. Minimum temperature threshold When the temperature goes below 16ºC, all physiological activity in the cocoa plant stops. If this occurs on isolated days followed by periods with average temperatures above 20ºC the effect is minimum. But if this happens during periods of more than 2 days and happens often, the plant is not able to synthesise the nutrients (carbohydrates) it needs to produce fruits. The following indicator was verified in order to measure whether the crop is unviable due to minimum temperature: 17 Along with the main observatories with rainfall and temperature information La Ceiba, Tela and San Pedro Sula- the analysis has been performed for the grid points shown in annex III 44

45 Non-viability Due to Minimum Temperatures Indicator (ITMin): Maximum number of consecutive days with minimum temperatures below 16ºC This indicator affects cocoa plantations in the highest areas, so the figures shown correspond to the only two observatories located in highlands where temperature data are available: Virgilio Enamorado in Santa Bárbara and Santa Rosa de Copán. The scenarios predict a rise in the minimum temperatures, so these are not expected to affect the cocoa crop. The value for this indicator diminishes significantly (see figures 46 and 47), meaning that conditions for growing cocoa in the Northern Honduras highlands will improve in the future because there will be less days with low temperatures. It is even felt that with the expected temperature increase the highlands that have been restricted from growing cocoa due to low temperatures will become suitable areas for this crop. Figure 46: ITMin evolution foreseen in Santa Rosa de Copan (Copan) Figure 47 : ITMin evolution foreseen in Virgilio Enamorado (Santa Barbara) 2. Differences in temperatures One of the important aspects with a negative impact on cocoa growing is a great difference between day and night-time temperatures. The information consulted indicates that differences of more than 9ºC between maximum and minimum temperatures have a negative effect (MAG, 1991). The indicator used is described below: Differences in Temperature Indicator (IDTemp): Number of months per year when the difference between the maximum monthly average and the minimum monthly average is more than 9ºC 45

46 The scenarios show different results in the different departments: in the case of La Ceiba in Atlántida, the scenarios show a reduction in the number of months where the difference between maximum and minimum is more than 9ºC, decreasing from 4 months currently to 3 months by the mid-century, so that the climate in this area may become increasingly suitable for growing cocoa in this regard (figure 48). In Tela, Atlántida, the difference in temperature remains more or less stable, with a slight increase in the more pessimistic RCP (figure 49. In San Pedro Sula this indicator tends to rise significantly in the future scenarios, so growing cocoa there will become more complicated (figure 50). Figure 48: IDTemp evolution foreseen in La Ceiba (Atlántida) Figure 49: IDTemp evolution foreseen in Tela (Atlántida) 46

47 Figure 50: IDTemp evolution foreseen in La Mesa (San Pedro Sula) 3. Non-viability due to prolonged lack of rainfall Cocoa needs an annual rainfall pattern of between 1,500 and 2,500 mm, with a monthly minimum of 100 mm. Cocoa also becomes less viable when rainfall is less regular and the low rainfall periods are longer. The following indicator has been formulated and validated: Non-viability due to Irregular Rainfall Indicator (IIPrp): Maximum number of consecutive days during which less than 100 mm of rainfall accumulate In the areas of Santa Bárbara (Trinidad) and San Pedro Sula, the scenarios show a significant decrease in the number of consecutive days needed to accumulate 100 mm, around 10 days less between the current situation and mid-century. This decrease will be very slight in Tela, and in the case of La Ceiba the situation is not expected to change. A reduction in this indicator means that conditions in these areas will be more suitable for growing cocoa. Figure 51: IIPrp situation anticipated in Trinidad (Santa Bárbara) 47

48 Figure 52: IIPrp situation anticipated in La Mesa (San Pedro Sula) Figure 53: IIPrp situation anticipated in Quimistán (Santa Bárbara) Figure 54: IIPrp situation anticipated in Tela (Atlántida) Flowering stage Two critical elements related to flower pollination were identified in this phase: 1. the pollen conditions 2. the conditions required for flower pollination 48

49 1. Pollen conditions Cocoa flowering continues throughout the year, and minimum temperatures below 20/21ºC may cause viability problems. Physiological studies demonstrate that short periods of low temperature do not affect flowering 18. Therefore, the indicator defined is as follows: Pollen Non-viability Indicator (IIPC): Maximum number of consecutive days with minimum temperatures below 21ºC If these temperatures persist over a long period of time pollination decreases, so less flowers are inseminated and the quantity of produced fruits diminishes. The scenarios suggest that the number of consecutive days with temperatures below 21ºC will decrease by 50%. Minimum temperature climatograms shown in figure 58 also show that minimum temperatures will be increasingly higher, so less days with minimum temperatures below 21ºC are expected. Figures 55, 56 and 57 show a future trend toward fewer pollen viability problems associated with temperatures below 21ºC, which means that the conditions for cocoa pollination tend to improve. Figure 55: IIPC evolution foreseen in Tela (Atlántida) Figure 56: IIPC evolution foreseen in La Ceiba (Atlántida) 18 Information provided by Dr. Carlos Astorga of CATIE 49

50 Figure 57: IIPC evolution foreseen in La Mesa (San Pedro Sula) Figure 58: Climatogram mínimum temperatura La Mesa (San Pedro Sula) Figure 59: Climatogram mínimum temperature La Ceiba (Atlántida) 2. Conditions for flower pollination Cocoa flowers can only be inseminated during a 24-hour period. The little Forcipomyia fly is the main cocoa pollinator (a tiny insect able to easily get the pollen from cocoa flower). This little fly proliferates in pooled water, so during long periods of drought there are few flies, especially if it is hot. The indicator taken as reference to identify the reduction in fly colonies is the same as that which has been applied to water stress situations, which also reduce the fruit and bean filling, as will be seen. 50

51 4.2.3 Ripening stage Three critical elements have been identified in this phase: 1. fruit and bean filling: dry season conditions 2. vulnerability to diseases 3. delay in fruit development due to flood conditions 1. Fruit and bean filling: dry season conditions Fruit and bean filling is more affected in plantations unprotected by shade or trees, totally exposed to the sun, where fruit development is limited to days and a long dry season with high temperatures may contribute to a reduction in pod size and bean weight. Although fruit development is not as dependent on rainfall as on soil water availability, it is important to predict periods of low rainfall and high temperatures (dry seasons), so that if the dry season is very severe but temperatures do not exceed 35ºC for a long time the fruit s development will not be affected. The following indicators have been defined: Temperature Severity Indicator (IRT): Number of consecutive days with temperatures equal to or higher than 35ºC during the period from March 1 st to May 30 th Dry Season Indicator (IEST): Number of consecutive days with accumulated rainfall on that day and the previous 5 of less than 10 mm and maximum temperatures equal to or greater than 35ºC, between March 1 st and June 30 th The temperature climatograms shown in figure 60 indicate that in San Pedro Sula (unlike Tela and La Ceiba) the temperature will be above 35ºC for many months, something that had not happened until now on a continuous basis. Figure 60: La Mesa climatogram (San Pedro Sula) The scenarios for these indicators show that in Tela (and in La Ceiba as well) the dry season will be slightly longer and there will be more days with temperatures above 35ºC (see figures 61 and 62). In the area of San Pedro Sula, however, the number of consecutive days with high temperatures will be very much higher than in the Atlántida area, and the days with both low rainfall and high temperatures will also increase significantly (figures 63 and 64). 51

52 Figure 61: IEST evolution foreseen in Tela (Atlántida) Figure 62: IRT evolution foreseen in Tela (Atlántida) Figure 63: IEST evolution foreseen in La Mesa (San Pedro Sula) 52

53 Figure 64: IRT evolution foreseen in La Mesa (San Pedro Sula) It is expected that the combination of low rainfall and high temperatures will increase during the period from March to June in the San Pedro Sula area (despite the scenarios showing some rainfall increase in June). Therefore, situations with more water stress will be concentrated from March to May and will be more intense, with consequences including a lower presence of pollinator flies as well as on fruit filling out. 2. Vulnerability to diseases There are two main cocoa diseases: black pod (Phythophthora Phythophthora palmivora and P. capsici) and monilia (Moniliophthora roreri), which affect the plant at different times during the year and therefore affect different harvests. Black pod is less severe than monilia disease, but it nevertheless undermines product quantity and quality, causing losses that may account for more than 30% of the almost mature product, when it s close to being harvested. Black pod proliferation is favoured by the presence of cold fronts during the period from October to January. When there is a week or more during that period with temperatures below 20ºC along with frequent rainfall events (even short and light), the relative humidity increases, favouring fungi development. Two additional factors must be added to these, which are relevant although they are not analysed on this study because they are defined locally: the presence of inoculums on the plantation (controlled by crop practices) and crop ventilation (well ventilated areas have less relative humidity and present less incidence of the disease). The following indicator has been formulated to measure vulnerability to black pod (and consequently to other fungal diseases): Vulnerability to Diseases Indicator (SuEn): Maximum number of consecutive days between November 1 st and February 28 th with daily average temperatures below 20ºC and accumulated rainfall that day and the previous four of more than 20 mm The scenarios show that the number of days when this favourable condition for black pod development is met will be reduced by almost half, due to the trend toward higher temperatures, with daily averages above 20ºC (figures 65, 66 and 67). 53

54 Figure 65: SuEn evolution foreseen in La Mesa (San Pedro Sula) Figure 66: SuEn evolution foreseen in Tela (Atlántida) Figure 67: SuEn evolution foreseen in La Ceiba (Atlántida) Monilia is the disease that causes the most economic harm to producers in adult cocoa plantations. It affects both the March-June harvest and that of September-January. According to the experts consulted, high temperatures favour monilia development. An indicator for each season would be as follows: Vulnerability to Monilia Indicator (SuMon1): Maximum number of consecutive days between October 1 st and January 31 st with daily average temperatures above 25ºC and accumulated rainfall for that day and the previous four of more than 20 mm 54

55 The scenarios show a significant increase in the number of consecutive days when conditions favourable to monilia development are met 19. This increase is affected by the overall rise expected in average temperatures. The gradient is steeper in La Ceiba because the scenarios show significant increases in rainfall during October and November, and gentler in San Pedro Sula because average temperatures are currently high and their expected increase doesn t have much impact on the indicator (figures 68, 69 and 70). Figure 68: SuMon1 evolution foreseen in La Ceiba (Atlántida) Figure 69: SuMon1 evolution foreseen in Tela (Atlántida) 19 Graphs for this indicator and the next one often show a limit of 5 days because, as this indicator has been defined, if it rains 20 mm only one day during the period analysed, the obtained value is 5. 55

56 Figure 70: SuMon1 evolution foreseen in La Mesa (San Pedro Sula) Vulnerability to Monilia Indicator (SuMon2): Maximum number of consecutive days between March 1 st and June 30 th with daily average temperatures above 25ºC and accumulated rainfall that day and the previous four of greater than 20 mm Weather conditions suitable for monilia development will slightly increase from March to June in the Tela and La Ceiba areas, but not in the San Pedro Sula area. The slight changes perceived can be associated with temperature increases (figures 71, 72 and 73). Figure 71: SuMon2 evolution foreseen in La Ceiba (Atlántida) 56

57 Figure 72: SuMon2 evolution foreseen in Tela (Atlántida) Figure 73: SuMon2 evolution foreseen in La Mesa (San Pedro Sula) 3. Accumulated rainfall and saturation during ripening In the case of cocoa grown in the lowlands (70% of the current situation in Honduras, and expected to increase) if it rains continuously and flood conditions are reached for two weeks, soil water saturation and the lack of aeration affect the translocation 20 of nutrients for the generation of the carbohydrates essential to form fruits, causing a delay in fruit development. This fact is more critical during the rainiest months (October to January) and in trees very laden with fruit which would have small, low-weight beans. The indicator defined and verified to approximately measure this phenomenon is as follows: Water Saturation Indicator (ISH): Maximum accumulated rainfall (mm) in 10 consecutive days between October 1 st and January 31 st divided by 200 mm If this indicator has a value above 1 it means that it rains more than 200 mm on 10 consecutive days, which can often cause floods, delaying ripening and leading to lower-weight fruits. La Ceiba is where there are the rainiest conditions that may cause soil water saturation, but the situation there is not expected to change because the scenarios do not predict a rise in rainfall from October to January in that area. In Tela, however, more intense rainfall is predicted during these months and therefore the saturation index will rise. A slight increase is also expected in the San Pedro Sula area (figures 74, 75 and 76) Transfer of nutrients to the place where they are needed by the plant. 21 This type of analysis must be complemented by soil analyses in order to discover the saturation level. 57

58 Figure 74: ISH evolution foreseen in La Ceiba (Atlántida) Figure 75: ISH evolution foreseen in Tela (Atlántida) Figure 76: ISH evolution foreseen in La Mesa (San Pedro Sula) Harvest stage Cocoa pods must be collected every ten or fifteen days, except during the most productive periods (April, May, November and December) when harvesting must be more frequent. The main critical element is that a cold front could affect the harvest at that moment, because the roads are generally unpaved and any rain, even light, is enough for them to become impassable and make it difficult to gain access to the plantation to harvest the fruit and transport the product to processing centres. 58

59 The following complementary indicators have been proposed to measure the incidence of this critical element: Harvest and Transport Conditions Indicator (COS1): Maximum accumulated rainfall (mm) in 7 consecutive days between November 1 st and December 31 st divided by 140 mm This indicator makes it possible to study the rainfall intensity over short periods of time and its effect on the cocoa crop. If it is above 1 it means that rainfall may complicate product harvesting and transportation tasks. This indicator shows that rainfall intensity during harvest time will increase (except in La Ceiba), and therefore the conditions during harvest will be more difficult in both the Tela and San Pedro Sula growing areas (figures 77, 78 and 79). Figure 77: COS1 evolution foreseen in Tela (Atlántida) Figure 78: COS1 evolution foreseen in La Mesa (San Pedro Sula) 59

60 Figure 79: COS1 evolution foreseen in La Ceiba (Atlántida) Harvest and Transport Conditions Indicator (COS2): Number of days between November 1 st and December 31 st with more than 10 mm of rainfall This second indicator makes it possible to see if the frequency of days with high rainfall during the harvest will increase, which could complicate transportation conditions. This indicator shows a similar situation of slightly worse conditions for harvesting cocoa in the Tela and San Pedro Sula areas of influence, with a slight increase in the number of days with more than 10 mm of rainfall in the most pessimistic scenario (RCP85) (see figures 80, 81 and 82). Figure 80: COS2 evolution foreseen in La Mesa (San Pedro Sula) 60

61 Figure 81: COS2 evolution foreseen in Tela (Atlántida) Figure 82: COS2 evolution foreseen in La Ceiba (Atlántida) Post-harvest stage Cocoa s final sales quality depends directly on suitable bean fermenting and drying. While fermenting usually takes place in a protected but well-ventilated place (requiring from five to seven days), the most used and recommended drying method is directly under the sun, until the beans internal moisture level is reduced to 6%. During drying the bean must not get wet, because this delays the process and reduces the final quality. The use of these systems gains importance in highly rainy areas, where sun drying faces continuous hurdles. The following indicator has been verified and applied to measure the incidence of this critical element: Post-harvest Indicator (PCOS): Number of days between December 1 st and 31 st with more than 1 mm of rainfall This indicator is intended to measure the fact that a light rainfall is enough to keep the producer from taking the beans outdoors for drying. The scenarios show that the increase in the number of days will be very slight, almost imperceptible, in San Pedro Sula. There will be a somewhat greater increase in La Masica (Atlántida) but in any case the situation does not seem to imply more adverse drying conditions than those existing to date (figures 83, 84, 85 and 86). 61

62 Figure 83: PCOS conditions predicted in La Mesa (San Pedro Sula) Figure 84: PCOS conditions predicted in Tela (Atlántida) Figure 85: PCOS conditions predicted in La Ceiba (Atlántida) 62

63 Figure 86: PCOS conditions predicted in La Masica (Atlántida) 5. BASIC GRAINS VALUE CHAIN 5.1 Mapping: process and actors Honduras was the foremost producer of maize and beans in the Central American region during the decades of 1970 and However, it was replaced by Guatemala as the leader for maize and Nicaragua for beans. Value chains for these two crops share many elements, so they are analysed together Value chain steps and scheduling Both maize and beans present two types of value chains: one corresponds to commercial production (by medium- and large-scale farmers) and the other to personal consumption (by small-scale farmers) 22. The former is more complex, because small-scale farmers carry out a lower number of activities. For example, they do not apply phytosanitary products and fertilization is less frequent. According to the experts consulted, over 50% of the maize grown in the three-department project area is for trading and the remaining 50% is for personal consumption. In the case of beans, the experts estimate that only 15% is traded whereas 85% is for personal consumption. The central link in the chain is made up of activities related to the physical handling of the product. The value chain steps 23 are described in the following figure: Figure 87: Central link in the maize and beans value chain. Source: The authors. While processing refers to maize, some cooperatives are currently starting to mill beans. Along with this central link, the value chain includes activities related to support services and to the institutional framework. They will be analysed in less detail, given that they are not the subject of this study. This study focuses on the first two activities shown in Figure 87, because they are the most affected by climatic events and they are carried out in the production areas. 22 A portion of maize and beans for personal consumption is also traded, as will be further analyzed. 23 Note: only the first two steps (in orange) are studied here. 63

64 Production Both maize and beans are sown at two different moments throughout the year: primera (the first sowing, from May to June) and postrera (October-November). In the case of beans, they are 24 vulnerable to droughts, so they are mainly sown in postrera (80% of production). Maize, on the contrary, is mainly sown in primera (74% of production in Northern Honduras and 83% of production in the whole country) (INE, 2009). Additionally, small-scale farmers living in the highlands use the same field to grow maize and beans for personal consumption, so they usually sow maize in primera and, after harvesting it, sow beans in postrera25. In some cases, when it rains, they may sow beans before harvesting the maize, to avoid not being able to sow them. The maize production cycle lasts 120 days. Most varieties used are traditional-variety seeds and their production cycle follows the diagram below: Figure 88: Maize production cycle. Source: The authors. The most common commercial varieties of beans, Amadeus and Deoro, have a production cycle that lasts 90 days and follows the diagram bellow: Figure 89: Beans production cycle. Source: The authors. 24 Statistics from the National Institute of Statistics (INE) published in the Encuesta Agropecuaria Básica refer to largescale producers, as indicated in the methodological section about the producer list framework: The list framework is made up of a list of large-scale producers, defined according to minimum limits of cultivated area previously defined for each basic crop, INE (2009), 25 Only maize in primera and beans in postrera are considered in the critical elements analysis, because they are the most common combination in the area studied. 64

65 The main difference between commercial production and production for personal consumption is the technical level, including the type of seed used and the phytosanitary products applied. Regarding the seeds, in the case of maize 68% are traditional, 10% are certified and 22% hybrid or genetically modified. In beans, it is estimated that 15% are traditional variety seeds, 10% certified and 75% commercial. Only half the farmers apply chemical fertilizers, usually in less quantity than the crop requires. In general, only producers with some financial capacity (medium- and large-scale) or those receiving the Bono Solidario Productivo 26 (Productive Solidarity Voucher, BSP) apply fertilizers. Small-scale farmers who do not receive the BSP obtain lower yields because their crops do not receive the necessary nutrients. The labour is provided by the family members both in the case of maize and beans at almost 70% of commercial farms (100% in the case of maize and beans for personal consumption), while 30% rely on temporary day labourers (SAG, 2012a and SAG, 2012b). Part of the production considered commercial goes to the seed market. Drying In the case of maize, the maize ears are bent while still on the stalk, approximately 90 days after sowing. This reduces their humidity and facilitates drying. After harvesting they are taken indoors to separate and dry the kernels. Kernels initially have a humidity content of 40-50% and they must end up having 12% at most. Small-scale farmers, lacking drying facilities, dry the kernels outdoors. This means they are more affected by the rainfall and low temperatures that may extend the drying process and make the kernel more vulnerable to diseases, which is why the maize is not planted in postrera in the highland areas of the study. In the case of beans, the drying process takes place in two phases. The first is in the field, after harvesting, and lasts between 10 and 15 days. After this time, or earlier if the weather conditions do not allow drying in the field, the beans are transferred to the farm facilities to finish drying, until a humidity content of only 12% is achieved. When the farmers do not have drying or storage facilities, they are forced to sell wet beans to middlemen that do have them (therefore, an excess of rainfall during drying has an immediate effect on household economies dependent on beans). This often happens with beans for personal consumption: farmers usually sell wet beans to small local traders (for instance, grocery stores) and buy them back once they re dried, for consumption. Therefore, beans for personal consumption are in fact sold and bought back at a higher price, so the small-scale farmer loses the trading margin in this operation. In the case of producers with more technical resources, they sell the dry product to local middlemen who transport it to departmental capitals and distribute it to processing plants in the case of milling (as mentioned above, this is an emerging activity in this area). Figure 90: Commercial and self-consumed beans cycle. Source: The authors. 26 BSP is a project promoted by the Honduran Government which started in 2010 with the target of distributing 150,000 vouchers annually to basic grain producers in different areas all over the country between 2010 and This programme is coordinated by the Agriculture and Livestock Secretariat (SAG) and managed by agreement by the Regional International Organization for Agricultural Health (OIRSA). The voucher consists of a 50 pound bag of beans and a fertilizer bag. See more information: 65

66 5.1.2 Maize and beans situation in the project area According to the Honduran National Institute of Statistics (INE, 2009:36), in the primera cycle of the northern region (Santa Bárbara and Cortés) produced 1,350,230 quintals on a surface area of 51,054 manzanas. In the Atlantic Coast region 335,340 quintals (approximately 15,211 MT) were produced on 13,754 manzanas (approximately 9,627 hectares). Productivity in these areas is usually around 26 quintals per manzana (832 kilograms per hectare). Through experts meetings it was confirmed that these data must be viewed with caution because they do not reflect much of the small-scale production. Interviews in this area suggest that production is higher and that average yields, at least in Santa Bárbara, may be around 35 quintals per manzana (1,120 kilograms per hectare). With more technical methods (such as those used by large-scale farmers and small-scale farmers with support from the BSP) they can even reach average yields of up to quintals per manzana (1,920-2,240 kilograms per hectare). Concerning beans, in the INE reported yields of 80,974 quintals (approximately 3,673 MT) in 6,839 manzanas (4,787 hectares) for postrera in the northern region and 22,292 quintals (1,011 MT) in 1,851 manzanas (1,295 hectares) at the Atlantic Coast, with average yields around 12 quintals per manzana (384 kilograms per hectare). Small-scale farmers supported by the Productive Solidarity Voucher produce average yields of quintals per manzana ( kilograms per hectare), whereas those not using fertilizers only obtain quintals per manzana ( kilograms per hectare) or even less (INE, 2009). Currently, both maize and bean production aimed at personal consumption in the project area are insufficient to meet the population s food needs (grain reserves last approximately six months), so it is necessary to adopt support measures that mitigate food insecurity risks. At a national level, the country is forced to import basic grains Actors in the chain Growers Almost every small-scale farmer in the project area grows maize and beans, usually with lowtech methods, unless they receive support from the state through the BSP. These producers lack the financial capacity to purchase seeds and other quality inputs, so their yields are very poor. Small-scale farmers are usually indebted to input suppliers, so they are forced to sell part of their production in order to pay back their debts, reducing the volume of basic grain available to meet household needs. It is estimated that these small-scale farmers allocate 20% of maize yields to debt repayment, 50% to personal consumption and the remaining 30% is sold to obtain incomes to meet other needs. Given their uncertain financial situation, they are forced to sell their labour power to other producers. Along with this, small farmers who own manzanas on average (0.7-1 hectare) have difficulty storing the maize, so they are forced to sell it quickly. The prices obtained are relatively low, and they are often unable to sell their product except to local retailers under unfavourable conditions, given that larger buyers (processing companies such as MASECA) demand high quality standards that small-scale farmers cannot satisfy. Medium-scale farmers, owning more than 5 cultivated manzanas, sell their product more easily to MASECA through intermediaries from San Pedro Sula who receive a commission. As for beans, small-scale farmers are in a very vulnerable situation because, lacking drying and storing capacity, they must sell wet beans to local traders and buy them back dried at higher prices. In general, producers receive very low prices for their product because they have limited bargaining power. Sales to local traders are usually on a cash basis, and the farmers are clearly at a disadvantage due to their limited financial capacity and their urgent need to sell their beans before they spoil. This vulnerable situation is exacerbated when an adverse climatic event causes a partial or total production loss. Beans are also very sensitive to diseases, and farmers lack the capacity to purchase phytosanitary products, so pests spread very fast. This damages the scanty livelihoods of these households even further. 66

67 Intermediaries As mentioned before, small-scale farmers can rarely sell their maize harvest to processing companies, so they usually sell it to local retailers. These, in turn, sell maize locally or regionally. Beans are also sold to small local traders with drying and storing capacity. Most intermediaries that take the product from the production areas in order to trade it in other departments or sell it to processing and manufacturing centres are individuals, generally financed by trading companies. They have more power than small-scale farmers and they usually impose their conditions because, since they normally operate in remote rural areas, they do not compete with other traders. Medium-scale farmers have access to middlemen working on commission for processing companies, because it is profitable for them to store the product Drying and processing companies There are maize processing companies at a national level (such as MASECA) that buy the product through middlemen and take it to the big cities (San Pedro Sula). These processing companies may dry the maize in their own facilities, although there are specific drying services in the production areas that can also be used by small-scale farmers. Very often these services are provided by other, larger farmers or by the traders themselves. As for beans, as mentioned before, some cooperatives are starting to mill them, but this is still an emergent activity Other actors There are other actors involved in the two value chains, including input suppliers, financial entities, research centres, service and technical assistance companies, etc. Along with them, there are institutions and projects linked to this product. Different companies import equipment and inputs in the departmental capitals of the study area, including SEAGRO, DISAGRO, FERTICA, FENORSA, CADELGA, FINCA, DUWEST, etc. These companies sell to local agricultural service suppliers. To judge by the presence of points of sale, the availability of these elements appears to be sufficient. Nevertheless, small-scale farmers very often lack the financial capacity to purchase these inputs in amounts suitable for their production. Concerning financial institutions, small-scale farmers hardly have access to formal credit, so when they need financing they rely on other chain actors involved in commercial links, particularly the middlemen that buy their production. They have access to credit from suppliers to a lesser extent, because low-tech farmers keep input purchases to a minimum. Along with these sources there are also financial institutions such as BANADESA, an entity specializing in the agricultural sector. DICTA provides technical and financial support (through the Productive Voucher) to the most vulnerable producers, and there are also different international development programmes such as PESA-FAO, PMA, etc., supporting the farmer technically or financially. 5.2 Most vulnerable critical elements to climate change and analysis of potential impacts under the different climate scenarios For analyzing critical elements in basic grain crops only primera maize and postrera beans have been considered, because it is the most used combination among producers in the project area MAIZE VALUE CHAIN In order to identify critical elements in the maize crop, the chain has been divided into the following phases: sowing, germination and plant development, ripening, harvesting and postharvesting. The main critical elements and selected indicators for each phase are presented below. 67

68 Sowing stage Two critical elements were identified in this phase: 1. the onset of the rainy season (primera), when it s possible to begin planting 2. the conditions for primera sowing 1. Onset of the rainy season An advance or delay in the onset of the rainy season, along with the consolidation of the beginning of this season, is key to analysing maize crop viability in primera. This indicator will be also the reference for many of the critical elements analysed for maize crops. Rainy Season Onset Indicator (III): First day with more than 2 mm of rainfall after April 20 which marks the beginning of a cycle of 4 days with accumulated precipitation of more than twice the daily average rainfall of the April 20 th -May 30 th period The projected scenarios for this indicator do not predict changes in any of the locations used as a reference for maize 27 in the three project departments, as observed in figures 91, 92 and 93. Figure 91. III evolution foreseen in Trinidad (Santa Bárbara) 27 Along with the main observatories with rainfall and temperature information La Ceiba, Tela and San Pedro Sula the following points with rainfall data have been taken as representative of the three departments chosen to use in maize scenarios: La Trinidad (point R04_05) and Quimistán (DHH010) in Santa Bárbara, Choloma in Cortés (point R05_07) and La Masica in Atlántida (DHL012). In any case, the analysis was performed for all the stations and critical points shown in Annex III. 68

69 Figure 92. III evolution foreseen in San Pedro Sula (Cortés) Figure 93. III evolution foreseen in La Ceiba (Atlántida) 2. Conditions for sowing in primera The primera sowing usually takes place 8-10 days after the onset of the rainy season, provided the adequate conditions for soil preparation (tempero) exist, including the absence of soil water saturation. The sowing indicator is designed to define these appropriate pre-sowing conditions: Sowing Indicator (ISM): First day from the onset of the rainy season with no rainfall and when a cycle of 7 consecutive days with accumulated rainfall of less than 40 mm starts The rainy season have started, so that the crop will receive the water it needs, but at the time of sowing the rainfall must not be very high. Otherwise the soil becomes compacted and the seeds find it difficult to break through the ground in order to germinate. Changes in this indicator are not expected, as observed in figures 94, 95 and 96. The sowing onset indicator responds to the same pattern as the rainy season onset indicator, because they are much linked. The difference between the two indicators shows that once the rainy season has started, neither the trends nor the onset of the sowing conditions are expected to vary. Therefore, it will be feasible to plant maize in the same primera conditions as currently. 69

70 Figure 94. ISM evolution foreseen in Trinidad (Santa Bárbara) Figure 95. ISM evolution foreseen in San Pedro Sula (Cortés) Figure 96. ISM evolution foreseen in La Ceiba (Atlántida) Germination and development stage Three critical elements were identified for crop germination and establishment: 1. minimum rainfall threshold 2. maximum rainfall threshold 3. minimum temperature threshold 70

71 1. Minimum rainfall threshold Maize germination and the first stage of development may be jeopardized by the shortage or excess of rainfall. If rainfall is scarce, moisture conditions do not meet the plants water needs. For measuring this critical element, the following indicator has been defined: Indicator of Germination and Establishment Conditions Associated with a Water Shortage (ICGP1): Accumulated rainfall (mm) during days 1-30 since sowing divided by 80 mm If this quotient is below 1 it means that during the 30 days after sowing the rainfall has been below 80 mm, implying that the seed could not find the necessary water conditions to germinate and develop in its initial stages. The scenarios do not predict increased problems due to water deficit, but there is a trend toward a slight improvement in water availability during the most sensitive period of primera maize farming, as figure 97 shows. However, in some areas including San Pedro Sula (figure 98) or La Ceiba (figure 99), although the trend is not increasing, a higher frequency of intense rainfall events after sowing can be observed. For instance, in San Pedro Sula the value 3 will be achieved more often, meaning more than 240 mm of rainfall in a month. This could have an impact on the crop if concentrated within a few days. Figure 97. ICGP1 evolution foreseen in Trinidad (Santa Bárbara) Figure 98. ICGP1 evolution foreseen in San Pedro Sula (Cortés) 71

72 Figure 99. ICGP1 evolution foreseen in La Ceiba (Atlántida) 2. Maximum rainfall threshold If rainfall and humidity are excessive, the seed may rapidly lose its chemical protection and became more vulnerable to fungal diseases. The soil can become compacted by rains, so the seedling cannot sprout. The excessive humidity may also damage already germinated plants by affecting their roots. Moreover, the rainfall can also prevent fertilization, and maize being a short cycle crop, a delay in fertilization jeopardizes the supply of nutrients necessary for germination, tilling and stem elongation. The corresponding indicator would be as follows: Indicator of Germination and Establishment Conditions Associated with Water Excess (ICGP2): Accumulated rainfall (mm) between days 1-35 after sowing divided by 200 mm If this quotient is above 1 it means that during the plant germination and first development periods accumulated rainfall exceeded 200 mm, which could result in adverse consequences for the plant (depending on soil properties, gradient, etc.) The scenarios indicate slight rainfall increases by mid-century at this time of the year (see, for instance, figure 100 for Santa Bárbara). Except in places where rainfall is already significant, such as La Masica in Atlántida (figure 101), accumulated rainfall during the first 35 days will rarely amount to 200 mm on average, because average rainfall during that period is around mm. As happened with the previous indicator, an increase in the frequency of intense rainfall is observed in the influence areas of La Ceiba and San Pedro Sula (figure 102). In the latter, the frequency of rainfall above 200 mm will increase, and therefore maize germination and establishment will became more difficult. 72

73 Figure 100. ICGP2 evolution foreseen in Trinidad (Santa Bárbara) Figure 101. ICGP2 evolution foreseen in La Masica (Atlántida) Figure 102. ICGP2 evolution foreseen in San Pedro Sula (Cortés) 3. Minimum temperature threshold During the germination stage the plant is very sensitive to low temperatures, especially if they are accompanied by humidity. Therefore, if the temperature is below 21ºC for 8 days the seed becomes dormant, the phytosanitary treatment loses effectiveness, the germination phase becomes prolonged, sprouts die, etc. This situation is aggravated if it is accompanied by rain. This indicator would be as follows: 73

74 Indicator of Germination and Establishment Conditions Associated with Low Temperatures (ICGT): Number of consecutive days with minimum temperatures below 21ºC between days 1-30 from the onset of planting The scenarios indicate that this factor limiting maize cultivation will appear less and less frequently. In the case of La Ceiba (figure 103), the observatory with the lowest night-time temperatures (using currently available data from Tela, La Ceiba and San Pedro Sula), the number of consecutive cold days will decrease. Therefore, this indicator projects better conditions for growing maize. Figure 103. ICGT evolution foreseen in La Ceiba (Atlántida) Flowering and ripening stage During consultations with experts in Honduras it was not possible to get enough details to define the links existing between weather conditions and crop behaviour at this stage. More detailed studies will have to be carried out that include physiological analysis of crop sensitivity in key stages of flowering and ripening in order to extract conclusions during this phase Harvesting stage One critical element was identified in this stage, related to harvest viability after maize bending (a common practice in this area): Harvest viability after maize bending Bending takes place approximately 90 days after sowing, in order to reduce maize ear humidity and prepare it for harvest. High rainfall at that time prevents this drying and the ear becomes more vulnerable to pests and diseases. It also causes a delay in the harvesting time, which can provoke grain rotting. Intense rain complicates the access of machinery and the labour force to the field for harvesting. In the postrera sowing, the biggest problem is not humidity but low temperatures, which cause a delay in the process. To avoid this problem, maize is not grown in the highest areas in the study area. Two indicators were combined during this phase: the first one is designed to measure intense rainfall situations and the second one measures the frequency of rainfall events, and therefore the existence of persistent humidity that contributes to disease development. Harvesting Conditions Indicator (ICC1): Maximum accumulated rainfall (mm) during 5 consecutive days between days 90 and 120 after sowing, divided by 100 mm 74

75 If this quotient is above 1 it means that during the time from bending to harvesting there is at least one period of 5 consecutive days with more than 100 of accumulated rainfall, making it more difficult to harvest. In Santa Bárbara (figure 104) and La Masica (Atlántida) (figure 105), there are no trend variations detected, and therefore harvesting conditions will remain the same. Projections are different in areas such as San Pedro Sula, Tela or La Ceiba. In San Pedro Sula (figure 106) the scenarios predict a trend of 20-30% increases in the volume of accumulated rainfall during this period by mid-century. In more western areas in Atlántida, such as Tela (figure 107), this increase will be around 10-20% according to the most pessimistic scenario (085). In these areas of San Pedro Sula and Tela, higher peaks are also observed in the graphs as the century progresses, meaning that the years with extreme conditions rainfall values above 200 mm in 5 days will increase. Precipitation in this period will be increasingly intense, and the drying process will became more complicated after bending. This may have implications on the increase of pests and diseases in the maize ears and kernel rotting. Figure 104. ICC1 evolution foreseen in Trinidad (Santa Bárbara) Figure 105. ICC1 evolution foreseen in La Masica (Atlántida) 75

76 Figure 106. ICC1 evolution foreseen in San Pedro Sula (Cortés) Figure 107. ICC1 evolution foreseen in Tela (Atlántida) Harvesting Conditions Indicator (ICC2): Number of days between days 90 and 120 after the planting date with more than 5 mm of rainfall Harvesting conditions (measured by the number of rainy days) will be a bit more complicated under the most pessimistic scenario in Quimistán (figure 108), La Ceiba and La Masica (figure 109). On the other hand, they will improve at some places in Cortés including Choloma (figure 110) and will not change in the area of San Pedro Sula in Cortés or Trinidad in Santa Bárbara (figure 111). 76

77 Figure 108. ICC2 evolution foreseen in Quimistán (Santa Bárbara) Figure 109. ICC2 evolution foreseen in La Masica (Atlántida) Figure 110. ICC2 evolution foreseen in Choloma (Cortés) 77

78 Figure 111. ICC2 evolution foreseen in Trinidad (Santa Bárbara) Post-harvest stage In case of commercial production, once harvested the grain is moved indoors. Therefore it will not be affected by weather condition, providing the ears were picked with the right humidity level. In case of maize grown for personal consumption, post-harvest storage is affected by the weather because it is necessary to reduce the grain humidity level from the 40-50% it has in the field to 12%. Temperature is a key factor for reducing humidity in rural highlands 28. If the temperature is below 20ºC and it rains, the grain cannot be put in the sun outdoors and it becomes more vulnerable to pests and diseases. This may cause total harvest loss. The following indicator has been proposed to measure the occurrence of this critical element: Post-harvest Indicator (IPCh): Number of days between day 120 and 150 after sowing with more than 5 mm of rainfall This indicator is designed to measure the number of days after harvesting when these adverse conditions occur. This indicator does not suggest significant trends in San Pedro Sula (figure 112), Tela or La Ceiba, but it does in other stations, including La Masica and Quimistán (figure 113). At these stations the number of rainy days will increase during key periods for maize drying, affecting future storage in suitable conditions. Figure 112. IPCh evolution foreseen in San Pedro Sula (Cortés) 28 It was not possible to analyse it in highlands due to the lack of meteorological information from highland stations, therefore this indicator does not capture temperature data. 78

79 Figure 113. IPCh evolution foreseen in Quimistán (Santa Bárbara) BEANS VALUE CHAIN The chain has been divided in the following phases to identify its critical elements: sowing, germination and plant development, flowering and ripening, harvesting and post-harvest. In order to evaluate the crop s viability throughout its whole cycle, a first section analysing the complete bean cycle has been included, because this crop requires certain temperature and rainfall conditions to be viable. The main critical elements and selected indicators are presented below for each crop phase Whole crop cycle Two critical elements were identified in this phase: 1. minimum precipitation threshold 2. minimum temperature threshold 1. Minimum precipitation threshold Bean farming requires a minimum level of accumulated precipitation throughout the whole productive cycle in order to be viable, estimated at 300 mm. The corresponding indicator is: Minimum Bean Crop Rainfall Conditions Indicator (ICMP): Accumulated rainfall between November 1 st and February 10 th divided by 300 mm If this quotient is below 1 it means that during the plant development period accumulated rainfall did not reach the required 300 mm, so the plant will lack the moisture conditions necessary to develop adequately (many factors affect these moisture conditions, but the study only analyses rainfall). This indicator shows a growth trend in every location analysed. The most significant increases take place in Quimistán (figure 114), San Pedro Sula (figure 115) and Tela (figure 116), with a tendency in some cases to achieve or even exceed 300 mm of rainfall. Therefore, these areas will have better humidity conditions for crop development in the future. 29 Along with the main observatories with rainfall and temperature data in La Ceiba, Tela and San Pedro Sula, the following points with rainfall data have been taken as representative of the three departments studied in order to follow the bean scenarios: in Santa Bárbara the SW at Santa Bárbara (point R03_04), E at la Trinidad (R04_05), and Quimistán (DHH010); in Cortés Lago Yojoa (point R05_04) and El Llano (R05_05); and in Atlantida Jutiapa (R10_07), Arizona (R07_07) and La Masica (DHH012). During the analysis all the points mentioned in Annex III were taken into consideration for the bean crop. 79

80 Figure 114. ICMP evolution foreseen in Quimistán (Santa Bárbara) Figure 115. ICMP evolution foreseen in San Pedro Sula (Cortés) Figure 116. ICMP evolution foreseen in Tela (Atlántida) 2. Minimum temperature threshold Bean plants are very sensitive to low temperatures, especially if they are accompanied by humidity. Depending on the time when this happens, if the temperature goes below 21ºC for around 8 consecutive days, the seed does not germinate, the crop does not develop, a lower number of flowers is produced, fewer and inferior pods form and fill due to an insufficient photoperiod, there are more attacks from pests and diseases, etc. As a result, yields decrease. 80

81 The corresponding indicator verified and analysed is: Minimum Temperature Conditions Indicator (ICMTF): Maximum number of consecutive days when daily minimum temperature is below 21ºC between November 1 st and February 10 th The projected evolution of this indicator shows a very significant decrease in the number of days with temperatures below 21ºC during the bean growing months in the study area (see figure 117 for Tela, for instance, although for the rest of locations the trend is similar). This means that conditions will be more suitable for this crop, because this limitation will become increasingly uncommon. Figure 117. ICMTF evolution foreseen in Tela (Atlántida) Sowing stage Two critical elements were identified in this phase: 1. pre-sowing conditions 2. conditions for sowing 1. Pre-sowing conditions Beans are currently sown around the second fortnight in October. Something that could complicate this activity is the arrival of a cold and rainy front just before sowing, which would delay the crop cycle. This could cause problems due to the proximity of the dry season, because the plant would not receive the necessary rain in further development stages. Being a short cycle crop, it needs the right conditions at the right time to develop properly. The corresponding indicator would be as follows: Pre-sowing Conditions Indicator (ICP-SF): Maximum rainfall accumulated (mm) on 7 consecutive rainy days between October 1 st and November 1 st, divided by 150 mm If this quotient is above 1 it means that there is too much rainfall during the period suitable for sowing, which prevents it from being done. Two very different trends have been observed for this indicator. In Santa Bárbara, Cortés and Tela (Atlántida), accumulated rainfall tends to increase during the sowing month. Intense rainfall events will significantly increase in San Pedro Sula (figure 118) and Tela. This may imply that sowing must be delayed because the soil does not meet the necessary conditions. On the other hand, in La Ceiba, La Masica, Jutiapa and Arizona (eastern and central parts of Atlántida) the situation will be more suitable because these intense rainfall events tend to decrease (figure 119 corresponding to La Masica). 81

82 Figure 118. ICP-SF evolution foreseen in San Pedro Sula (Cortés) Figure 119. ICP-SF evolution foreseen in La Masica (Atlántida) 2. Conditions for sowing Also rainfall must be low for sowing to take place, because otherwise it is not possible to enter the fields. The defined indicator is: Sowing Indicator (ISF): First day between October 1 st and November 15 th with no rainfall that starts a cycle of 7 consecutive days with less than 40 mm of accumulated rainfall The projected scenarios show that the dates for sowing will not change substantially, as observed for instance in figure 120 corresponding to the grid point in the southwest of the Santa Bárbara municipality. 82

83 Figure 120. ISF evolution foreseen in Santa Bárbara-SW Municipality Germination and development stage One critical element has been identified in this phase for appropriate crop germination and establishment; it is related to maximum rainfall for crop establishment: Maximum rainfall threshold for crop establishment Bean germination and first development may be jeopardized by an excess of moisture, usually associated with an extended rainfall period during the first days after sowing. This Indicator of Establishment Conditions Associated with Excessive Rainfall (ICIPF): Accumulated rainfall (mm) between days 1 and 30 after sowing, divided by 200 mm moisture affects three aspects that may cause the loss of part of the seeds: first, vulnerability to fungi, because the seed loses its chemical protection; second, soil compacting, because it prevents the sprout from breaking the ground to emerge; and third, it inhibits fertilization and application of preventive pest control treatments that take place at sowing and 15 days later. The defined indicator is: If this quotient is above 1 it means that during the germination and first development period there are more than 200 mm of cumulative rainfall, causing adverse consequences to the plant. In general, all the analysed locations show similar results. The indicator value exceeds 1 in few cases and rainfall is not expected to increase during that period. Precipitation tends to increase at only two stations: Tela (figure 121) and San Pedro (figure 122), occasionally exceeding 200 mm during sowing time, which means that at these locations the establishment of bean crops may be jeopardized. Figure 121. ICIPF evolution foreseen in Tela (Atlántida) 83

84 Figure 122. ICIPF evolution foreseen in San Pedro Sula (Cortés) Flowering and ripening stage In this phase one critical element has been identified, relating to the importance that excessive precipitation may have on flowering and bean formation/ripening: Maximum precipitation threshold for bean flowering and formation Flowering and weed control take place between days 35 and 40 after sowing, and at that time an excess of water may have consequences on crop development. On the one hand, lower grain setting because flowering was affected, and on the other there may be more weed competition (reducing productivity). The following is the indicator defined for this critical element: Flowering and Bean Formation Conditions Indicator (IFFP): Accumulated precipitation (mm) between days 25 and 50 after sowing, divided by 125 mm If this quotient is above 1 it means that it rains more than 125 mm during the flowering and fruit formation period, which could have adverse consequences on flowering and bean nutrition. The scenarios predict an increased rainfall during flowering and grain formation, which could lead to higher weed presence and also negatively affect flowering. See for instance the case of Tela (figure 123). Figure 123. IFFP evolution foreseen in Tela (Atlántida) 84

85 Harvesting stage One critical element has been identified in this phase, relating to harvest viability due to moisture levels. Harvest viability Harvesting takes place between December 15 th and January 15 th (70 to 80 days after sowing, when this is done in the second half of October). If there is an intense rainfall event (around 30 mm per day for 5 days) it makes it difficult to enter the fields to harvest, thus delaying it, which may cause grain rotting and higher susceptibility to pests and diseases. Two complementary indicators have been suggested: the first one measures precipitation volume in a short time in order to evaluate rainfall intensity, and the second one measures the number of rainy days, evaluating rainfall persistence and therefore humidity over time: Harvesting Conditions Indicator (ICCF1): Maximum accumulated rainfall (mm) during 5 consecutive days between December 15 th and January 15 th, divided by 150 mm If this quotient is above 1 it means that during the harvesting period there is at least one period of 5 consecutive days with more than 150 mm of accumulated rainfall, which may complicate harvesting. In Tela and La Ceiba (figure 124) intense rainfall events tend to increase slightly during harvesting under the most pessimistic scenarios. This situation could complicate harvesting. On the other hand, no significant changes were observed in northern Santa Bárbara (Quimistán) or in areas in Atlántida such as La Masica, Arizona and Jutiapa. Figure 124. ICCF1 evolution foreseen in La Ceiba (Atlántida) Harvesting Conditions Indicator (ICCF2): Number of days between December 15 th and January 15 th with more than 5 mm of daily rainfall. The scenarios for this indicator show a slight increase in the number of rainy days, more steady in places where rainfall events are more intense, such as La Masica (figure 125). This situation may complicate harvesting, a task already more difficult in those locations. 85

86 Figure 125. ICCF2 evolution foreseen in La Masica (Atlántida) Post-harvest stage Only one critical element, also related to humidity, has been identified at this stage. Pre-drying viability After harvesting, the grain is left in the fields for drying for days. If this period is extended due to rainfall events, beans may be more vulnerable to fungi and can be lost. To measure the incidence of this critical element, the following indicator has been proposed, designed to measure the number of days after harvesting when these adverse conditions occur: Post-harvest Indicator (IP-CF): Number of days between January 15 th and February 15 th with more than 5 mm of precipitation The scenarios do not show any trend changes regarding the number of days without rain needed for bean drying, except a slight increase in the case of La Masica, a location which is already very rainy, under the most pessimistic emissions scenario (figure 126). Figure 126. IP-CF evolution foreseen in La Masica (Atlántida) 86