Fish on Farms Phase 2: Family Farms for the Future Cambodia

Timothy Green, Zaman Talukder, Kristina D. Michaux, Najma Moumin

Research output: Book/ReportCommissioned report

Abstract

1. EXECUTIVE SUMMARY
This final report is the culmination of two CIFSRF projects carried out in Cambodia, colloquially known as ‘Fish on Farms’ (FoF) and ‘Family Farms for the Future’ (FF4F) that together spanned over six years. The overall aim of both projects was to “to improve household food security and nutrition outcomes, livelihoods, and women’s empowerment through an enhanced homestead food production (EHFP) model using an environmentally sustainable approach”. In part one, FoF, we conducted a cluster randomised control trial (cRCT) of an integrated model of EHFP in which 900 women headed households in Prey Veng province were randomized to either plant-based EHFP, EHFP plus fishponds, or control. The participants received many benefits such as reduced food insecurity, increased fruit and vegetable production, increased fish production in households with fishponds, and increased income that was controlled by women and used to purchase animal-source foods or for children’s education. However, a number of limitations were noted: women were not given a choice in the type of EHFP they received and may have been less motivated resulting in high attrition; almost all the costs of the intervention were borne by the project which in addition to not being sustainable may have led to less engagement by women farmers; insufficient time for EHFP to reach its full potential; issues with capturing the seasonality of crop production and nutrient intake; and lack of information on how EHFP might work in other areas in Cambodia, such as costal, highlands, and peri-urban areas.
The second phase of the project, FF4F, was designed to overcome some of these shortcomings. Here we aimed to develop a scalable, cost-effective model of EHFP for Cambodia. We incorporated successful components and lessons learned from FoF, refined tools, methodologies, and practices for scale-up in diverse agro-ecological zones and provinces (Kampot, Kampong Cham, Prey Veng, and peri-urban Phnom Penh) that included 4600 households. These households were able to choose the form of EHFP they wanted, which included a mix of gardens, poultry, and fishponds, but were expected to contribute more to the establishment of these homestead farms. In addition to the scale up, we conducted a pragmatic-delayed cluster randomized control trial (PDcRCT) nested within the larger FF4F study in Kampot province. In short, villages were randomized to EHFP of the households choosing, or control. During the first year, EHFP was introduced into the intervention villages (any type of EHFP) but no measurements were made. In the control group production was quantified monthly and detailed 24-hour dietary recalls were conducted with recipients (women and children) during lean (May) and peak (Dec) production periods in year one. In year two, we quantified production and dietary intake in the groups receiving EHFP.
With the stage set for the background we now focus on the key outputs and learnings from the last three years. Initially, households were expected to contribute 50% of the costs of the inputs required to set up their EHFP farms, but with extensive feedback from beneficiaries, this was reduced to 30% to ensure greater uptake of the program. Some households relied on microcredit but most relied on personal family funds. We targeted 4600 households, but through spill-over we estimate that the project reached another 15 households per study village. Members of these households either attended training, education, and/or demonstration events. In total, we estimate that 3,500 non-target households received benefits from this project, thus benefiting an estimated 17,500 family members overall. Attrition was high in non-surveillance households at over 40%. The main reasons given for attrition were migration for other economic opportunities and lack of time and labour resources. Fortunately, attrition occurred early and households could be replaced before considerable resources were spent. All provinces except the city of Phnom Penh took up aquaculture ranging from 23% in Kampong Cham to 54% in Kampot. Poultry raising took place in excess of 90% of houses in all provinces. In addition to initially poor-quality seeds, poultry death was high and plans were taken to mitigate this high mortality. Another strength of the study was that we followed up with both inactive and active households to explore the main factors contributing to retention and adoption of practices promoted by the project.
Based on data from our PDcRCT trial, we found significant differences in agriculture production and prevalence of inadequacy for key micronutrients between households that received the EHFP intervention (regardless of model) versus the control group. Specifically, of producing households in each category, on average, households in the EHFP group produced significantly more fruit (507 vs. 306 kg), large fish (23 vs. 15 kg), eggs (117 vs. 99 units), and live birds (99 vs. 89 units) than the control group, however, cumulative vegetable production was similar between groups (267 vs 192 kg). For food intake, we observed a statistically significant reduction in the prevalence of inadequate intake for zinc, thiamin, riboflavin, and vitamin A in women and children in the EHFP group compared to control, during both the lean and peak agricultural seasons. Over the two-year intervention, the project was cost neutral, when accounting for only targeted households. However, if projected out over a ten-year period the incremental net benefit was nearly USD500. Further, if you take into consideration the spill-over effect, the project is likely to have even larger positive net benefit.
All project objectives were met. Briefly, we refined technologies and practices from FoF and adapted them for different geographic regions and scale. Different types of fish and stocking densities were tested and small-scale hatcheries were encouraged such that now we feel that Cambodia has a sustainable model for fish hatcheries moving forward, which can be managed by the Fisheries Administration (FiA). We increased agricultural productivity such that the number of households with cultivated gardens rose from 60 to nearly 100%. To date, FF4F farms have contributed an estimated 28,000 kg of fish, 260,000 poultry eggs and 6,000,000 kg of fruit and vegetables to the local food supply. Women’s empowerment increased although we were not able to measure all dimensions. By the end of the study women were making 90% of the decisions around farming; a recommendation was made to CARD that gender be incorporated into their trainings and future policies. We improved access to resources, and/or markets and income. A series of business tools and initiatives have also been piloted and shown to improve crop selection decision-making, beneficiary financial planning, and market access. Using zinc as the model nutrient we were able to show the incidence and mortality rates of zinc deficiency could be reduced by 23.4%
We had in mind that our ultimate our goal is to transfer the skills and learnings to the government, private industry, and civil society to benefit as many Cambodians as possible. As major outputs, a dissemination workshop was carried out where Cambodian representatives reaffirmed their support for EHFP. A policy brief has been prepared and a number of manuscripts have been published or are under consideration. HKI is sharing their learnings with neighbouring countries through regional workshops and are implementing parts of the innovative surveillance to better track production of EHFP outputs and costs incurred (inputs) by households. HKI is a key informant to Cambodia’s National Strategy for Food Security and Nutrition which will be developed for the next five-year period covering 2019-2023.
In summary, this project has had great successes, a few failures, but with important lessons learned. Agricultural interventions such as EHFP will struggle to show improvement in nutrition indicators, due in part to the short duration of most research projects. Improvements in stunting are inter-generational and include not only greater access to better food but improvements in socio-economic status as a whole. A lot of our modelling is based on a ten-year projection. The reality is we don’t know what happens to these farms when donors pull out. We hope government and other organizations step up to fill the gap, but we simply don’t know the long-term impact of EHFP. It would seem essential to return to these households in five years to see the longer-term impact.
Finally, we would like to highlight some of the spin-off successes that stem directly from this project. Through this study we were able to bring the leading expert from Harvest Plus to train Cambodian and international nutrition researchers on how to conduct proper 24-dietary hour recalls in resource-poor settings. We held focus groups to establish common recipes and created a Cambodian Food Composition Database, that is now in the public domain. This will be essential for future researcher planners, including those wanting to do National Nutrition Surveys. The initial aim of FoF was to reduce iron deficiency anemia in Cambodian women. Through careful study we determined that anemia in Cambodia was largely caused by genetic blood disorders. Working with HKI and funds from the Canadian Institute of Health Research, we proved that iron was not the cause of anemia and iron might be doing harm. In response the Cambodian government ceased weekly iron and folic acid supplements for non-pregnant women. We also noted that there were high cases of infantile beriberi (thiamine deficiency) in Cambodia. With funding from Grand Challenges Canada and with generous in-kind support from HKI we were able to fortify fish sauce with thiamine and give it to pregnant women to improve both theirs’s and their infant’s thiamine levels in blood. This is now being applied to salt and the team has been given nearly a million USD in additional funding. Directly and indirectly through this project we have helped trained the next generation of Canadian nutrition researchers with skills to work in low income countries; this includes two PhDs, three MSc, three research assistants, and countless undergraduate nutrition, business, and public health students.
LanguageEnglish
Number of pages46
Publication statusPublished - 26 Jun 2018

Cite this

Green, T., Talukder, Z., Michaux, K. D., & Moumin, N. (2018). Fish on Farms Phase 2: Family Farms for the Future Cambodia.
Green, Timothy ; Talukder, Zaman ; Michaux, Kristina D. ; Moumin, Najma . / Fish on Farms Phase 2: Family Farms for the Future Cambodia. 2018. 46 p.
@book{085f9a1d37f84634821968d7becdd0e5,
title = "Fish on Farms Phase 2: Family Farms for the Future Cambodia",
abstract = "1. EXECUTIVE SUMMARYThis final report is the culmination of two CIFSRF projects carried out in Cambodia, colloquially known as ‘Fish on Farms’ (FoF) and ‘Family Farms for the Future’ (FF4F) that together spanned over six years. The overall aim of both projects was to “to improve household food security and nutrition outcomes, livelihoods, and women’s empowerment through an enhanced homestead food production (EHFP) model using an environmentally sustainable approach”. In part one, FoF, we conducted a cluster randomised control trial (cRCT) of an integrated model of EHFP in which 900 women headed households in Prey Veng province were randomized to either plant-based EHFP, EHFP plus fishponds, or control. The participants received many benefits such as reduced food insecurity, increased fruit and vegetable production, increased fish production in households with fishponds, and increased income that was controlled by women and used to purchase animal-source foods or for children’s education. However, a number of limitations were noted: women were not given a choice in the type of EHFP they received and may have been less motivated resulting in high attrition; almost all the costs of the intervention were borne by the project which in addition to not being sustainable may have led to less engagement by women farmers; insufficient time for EHFP to reach its full potential; issues with capturing the seasonality of crop production and nutrient intake; and lack of information on how EHFP might work in other areas in Cambodia, such as costal, highlands, and peri-urban areas. The second phase of the project, FF4F, was designed to overcome some of these shortcomings. Here we aimed to develop a scalable, cost-effective model of EHFP for Cambodia. We incorporated successful components and lessons learned from FoF, refined tools, methodologies, and practices for scale-up in diverse agro-ecological zones and provinces (Kampot, Kampong Cham, Prey Veng, and peri-urban Phnom Penh) that included 4600 households. These households were able to choose the form of EHFP they wanted, which included a mix of gardens, poultry, and fishponds, but were expected to contribute more to the establishment of these homestead farms. In addition to the scale up, we conducted a pragmatic-delayed cluster randomized control trial (PDcRCT) nested within the larger FF4F study in Kampot province. In short, villages were randomized to EHFP of the households choosing, or control. During the first year, EHFP was introduced into the intervention villages (any type of EHFP) but no measurements were made. In the control group production was quantified monthly and detailed 24-hour dietary recalls were conducted with recipients (women and children) during lean (May) and peak (Dec) production periods in year one. In year two, we quantified production and dietary intake in the groups receiving EHFP. With the stage set for the background we now focus on the key outputs and learnings from the last three years. Initially, households were expected to contribute 50{\%} of the costs of the inputs required to set up their EHFP farms, but with extensive feedback from beneficiaries, this was reduced to 30{\%} to ensure greater uptake of the program. Some households relied on microcredit but most relied on personal family funds. We targeted 4600 households, but through spill-over we estimate that the project reached another 15 households per study village. Members of these households either attended training, education, and/or demonstration events. In total, we estimate that 3,500 non-target households received benefits from this project, thus benefiting an estimated 17,500 family members overall. Attrition was high in non-surveillance households at over 40{\%}. The main reasons given for attrition were migration for other economic opportunities and lack of time and labour resources. Fortunately, attrition occurred early and households could be replaced before considerable resources were spent. All provinces except the city of Phnom Penh took up aquaculture ranging from 23{\%} in Kampong Cham to 54{\%} in Kampot. Poultry raising took place in excess of 90{\%} of houses in all provinces. In addition to initially poor-quality seeds, poultry death was high and plans were taken to mitigate this high mortality. Another strength of the study was that we followed up with both inactive and active households to explore the main factors contributing to retention and adoption of practices promoted by the project.Based on data from our PDcRCT trial, we found significant differences in agriculture production and prevalence of inadequacy for key micronutrients between households that received the EHFP intervention (regardless of model) versus the control group. Specifically, of producing households in each category, on average, households in the EHFP group produced significantly more fruit (507 vs. 306 kg), large fish (23 vs. 15 kg), eggs (117 vs. 99 units), and live birds (99 vs. 89 units) than the control group, however, cumulative vegetable production was similar between groups (267 vs 192 kg). For food intake, we observed a statistically significant reduction in the prevalence of inadequate intake for zinc, thiamin, riboflavin, and vitamin A in women and children in the EHFP group compared to control, during both the lean and peak agricultural seasons. Over the two-year intervention, the project was cost neutral, when accounting for only targeted households. However, if projected out over a ten-year period the incremental net benefit was nearly USD500. Further, if you take into consideration the spill-over effect, the project is likely to have even larger positive net benefit.All project objectives were met. Briefly, we refined technologies and practices from FoF and adapted them for different geographic regions and scale. Different types of fish and stocking densities were tested and small-scale hatcheries were encouraged such that now we feel that Cambodia has a sustainable model for fish hatcheries moving forward, which can be managed by the Fisheries Administration (FiA). We increased agricultural productivity such that the number of households with cultivated gardens rose from 60 to nearly 100{\%}. To date, FF4F farms have contributed an estimated 28,000 kg of fish, 260,000 poultry eggs and 6,000,000 kg of fruit and vegetables to the local food supply. Women’s empowerment increased although we were not able to measure all dimensions. By the end of the study women were making 90{\%} of the decisions around farming; a recommendation was made to CARD that gender be incorporated into their trainings and future policies. We improved access to resources, and/or markets and income. A series of business tools and initiatives have also been piloted and shown to improve crop selection decision-making, beneficiary financial planning, and market access. Using zinc as the model nutrient we were able to show the incidence and mortality rates of zinc deficiency could be reduced by 23.4{\%}We had in mind that our ultimate our goal is to transfer the skills and learnings to the government, private industry, and civil society to benefit as many Cambodians as possible. As major outputs, a dissemination workshop was carried out where Cambodian representatives reaffirmed their support for EHFP. A policy brief has been prepared and a number of manuscripts have been published or are under consideration. HKI is sharing their learnings with neighbouring countries through regional workshops and are implementing parts of the innovative surveillance to better track production of EHFP outputs and costs incurred (inputs) by households. HKI is a key informant to Cambodia’s National Strategy for Food Security and Nutrition which will be developed for the next five-year period covering 2019-2023.In summary, this project has had great successes, a few failures, but with important lessons learned. Agricultural interventions such as EHFP will struggle to show improvement in nutrition indicators, due in part to the short duration of most research projects. Improvements in stunting are inter-generational and include not only greater access to better food but improvements in socio-economic status as a whole. A lot of our modelling is based on a ten-year projection. The reality is we don’t know what happens to these farms when donors pull out. We hope government and other organizations step up to fill the gap, but we simply don’t know the long-term impact of EHFP. It would seem essential to return to these households in five years to see the longer-term impact. Finally, we would like to highlight some of the spin-off successes that stem directly from this project. Through this study we were able to bring the leading expert from Harvest Plus to train Cambodian and international nutrition researchers on how to conduct proper 24-dietary hour recalls in resource-poor settings. We held focus groups to establish common recipes and created a Cambodian Food Composition Database, that is now in the public domain. This will be essential for future researcher planners, including those wanting to do National Nutrition Surveys. The initial aim of FoF was to reduce iron deficiency anemia in Cambodian women. Through careful study we determined that anemia in Cambodia was largely caused by genetic blood disorders. Working with HKI and funds from the Canadian Institute of Health Research, we proved that iron was not the cause of anemia and iron might be doing harm. In response the Cambodian government ceased weekly iron and folic acid supplements for non-pregnant women. We also noted that there were high cases of infantile beriberi (thiamine deficiency) in Cambodia. With funding from Grand Challenges Canada and with generous in-kind support from HKI we were able to fortify fish sauce with thiamine and give it to pregnant women to improve both theirs’s and their infant’s thiamine levels in blood. This is now being applied to salt and the team has been given nearly a million USD in additional funding. Directly and indirectly through this project we have helped trained the next generation of Canadian nutrition researchers with skills to work in low income countries; this includes two PhDs, three MSc, three research assistants, and countless undergraduate nutrition, business, and public health students.",
author = "Timothy Green and Zaman Talukder and Michaux, {Kristina D.} and Najma Moumin",
year = "2018",
month = "6",
day = "26",
language = "English",

}

Green, T, Talukder, Z, Michaux, KD & Moumin, N 2018, Fish on Farms Phase 2: Family Farms for the Future Cambodia.

Fish on Farms Phase 2: Family Farms for the Future Cambodia. / Green, Timothy; Talukder, Zaman; Michaux, Kristina D.; Moumin, Najma .

2018. 46 p.

Research output: Book/ReportCommissioned report

TY - BOOK

T1 - Fish on Farms Phase 2: Family Farms for the Future Cambodia

AU - Green, Timothy

AU - Talukder, Zaman

AU - Michaux, Kristina D.

AU - Moumin, Najma

PY - 2018/6/26

Y1 - 2018/6/26

N2 - 1. EXECUTIVE SUMMARYThis final report is the culmination of two CIFSRF projects carried out in Cambodia, colloquially known as ‘Fish on Farms’ (FoF) and ‘Family Farms for the Future’ (FF4F) that together spanned over six years. The overall aim of both projects was to “to improve household food security and nutrition outcomes, livelihoods, and women’s empowerment through an enhanced homestead food production (EHFP) model using an environmentally sustainable approach”. In part one, FoF, we conducted a cluster randomised control trial (cRCT) of an integrated model of EHFP in which 900 women headed households in Prey Veng province were randomized to either plant-based EHFP, EHFP plus fishponds, or control. The participants received many benefits such as reduced food insecurity, increased fruit and vegetable production, increased fish production in households with fishponds, and increased income that was controlled by women and used to purchase animal-source foods or for children’s education. However, a number of limitations were noted: women were not given a choice in the type of EHFP they received and may have been less motivated resulting in high attrition; almost all the costs of the intervention were borne by the project which in addition to not being sustainable may have led to less engagement by women farmers; insufficient time for EHFP to reach its full potential; issues with capturing the seasonality of crop production and nutrient intake; and lack of information on how EHFP might work in other areas in Cambodia, such as costal, highlands, and peri-urban areas. The second phase of the project, FF4F, was designed to overcome some of these shortcomings. Here we aimed to develop a scalable, cost-effective model of EHFP for Cambodia. We incorporated successful components and lessons learned from FoF, refined tools, methodologies, and practices for scale-up in diverse agro-ecological zones and provinces (Kampot, Kampong Cham, Prey Veng, and peri-urban Phnom Penh) that included 4600 households. These households were able to choose the form of EHFP they wanted, which included a mix of gardens, poultry, and fishponds, but were expected to contribute more to the establishment of these homestead farms. In addition to the scale up, we conducted a pragmatic-delayed cluster randomized control trial (PDcRCT) nested within the larger FF4F study in Kampot province. In short, villages were randomized to EHFP of the households choosing, or control. During the first year, EHFP was introduced into the intervention villages (any type of EHFP) but no measurements were made. In the control group production was quantified monthly and detailed 24-hour dietary recalls were conducted with recipients (women and children) during lean (May) and peak (Dec) production periods in year one. In year two, we quantified production and dietary intake in the groups receiving EHFP. With the stage set for the background we now focus on the key outputs and learnings from the last three years. Initially, households were expected to contribute 50% of the costs of the inputs required to set up their EHFP farms, but with extensive feedback from beneficiaries, this was reduced to 30% to ensure greater uptake of the program. Some households relied on microcredit but most relied on personal family funds. We targeted 4600 households, but through spill-over we estimate that the project reached another 15 households per study village. Members of these households either attended training, education, and/or demonstration events. In total, we estimate that 3,500 non-target households received benefits from this project, thus benefiting an estimated 17,500 family members overall. Attrition was high in non-surveillance households at over 40%. The main reasons given for attrition were migration for other economic opportunities and lack of time and labour resources. Fortunately, attrition occurred early and households could be replaced before considerable resources were spent. All provinces except the city of Phnom Penh took up aquaculture ranging from 23% in Kampong Cham to 54% in Kampot. Poultry raising took place in excess of 90% of houses in all provinces. In addition to initially poor-quality seeds, poultry death was high and plans were taken to mitigate this high mortality. Another strength of the study was that we followed up with both inactive and active households to explore the main factors contributing to retention and adoption of practices promoted by the project.Based on data from our PDcRCT trial, we found significant differences in agriculture production and prevalence of inadequacy for key micronutrients between households that received the EHFP intervention (regardless of model) versus the control group. Specifically, of producing households in each category, on average, households in the EHFP group produced significantly more fruit (507 vs. 306 kg), large fish (23 vs. 15 kg), eggs (117 vs. 99 units), and live birds (99 vs. 89 units) than the control group, however, cumulative vegetable production was similar between groups (267 vs 192 kg). For food intake, we observed a statistically significant reduction in the prevalence of inadequate intake for zinc, thiamin, riboflavin, and vitamin A in women and children in the EHFP group compared to control, during both the lean and peak agricultural seasons. Over the two-year intervention, the project was cost neutral, when accounting for only targeted households. However, if projected out over a ten-year period the incremental net benefit was nearly USD500. Further, if you take into consideration the spill-over effect, the project is likely to have even larger positive net benefit.All project objectives were met. Briefly, we refined technologies and practices from FoF and adapted them for different geographic regions and scale. Different types of fish and stocking densities were tested and small-scale hatcheries were encouraged such that now we feel that Cambodia has a sustainable model for fish hatcheries moving forward, which can be managed by the Fisheries Administration (FiA). We increased agricultural productivity such that the number of households with cultivated gardens rose from 60 to nearly 100%. To date, FF4F farms have contributed an estimated 28,000 kg of fish, 260,000 poultry eggs and 6,000,000 kg of fruit and vegetables to the local food supply. Women’s empowerment increased although we were not able to measure all dimensions. By the end of the study women were making 90% of the decisions around farming; a recommendation was made to CARD that gender be incorporated into their trainings and future policies. We improved access to resources, and/or markets and income. A series of business tools and initiatives have also been piloted and shown to improve crop selection decision-making, beneficiary financial planning, and market access. Using zinc as the model nutrient we were able to show the incidence and mortality rates of zinc deficiency could be reduced by 23.4%We had in mind that our ultimate our goal is to transfer the skills and learnings to the government, private industry, and civil society to benefit as many Cambodians as possible. As major outputs, a dissemination workshop was carried out where Cambodian representatives reaffirmed their support for EHFP. A policy brief has been prepared and a number of manuscripts have been published or are under consideration. HKI is sharing their learnings with neighbouring countries through regional workshops and are implementing parts of the innovative surveillance to better track production of EHFP outputs and costs incurred (inputs) by households. HKI is a key informant to Cambodia’s National Strategy for Food Security and Nutrition which will be developed for the next five-year period covering 2019-2023.In summary, this project has had great successes, a few failures, but with important lessons learned. Agricultural interventions such as EHFP will struggle to show improvement in nutrition indicators, due in part to the short duration of most research projects. Improvements in stunting are inter-generational and include not only greater access to better food but improvements in socio-economic status as a whole. A lot of our modelling is based on a ten-year projection. The reality is we don’t know what happens to these farms when donors pull out. We hope government and other organizations step up to fill the gap, but we simply don’t know the long-term impact of EHFP. It would seem essential to return to these households in five years to see the longer-term impact. Finally, we would like to highlight some of the spin-off successes that stem directly from this project. Through this study we were able to bring the leading expert from Harvest Plus to train Cambodian and international nutrition researchers on how to conduct proper 24-dietary hour recalls in resource-poor settings. We held focus groups to establish common recipes and created a Cambodian Food Composition Database, that is now in the public domain. This will be essential for future researcher planners, including those wanting to do National Nutrition Surveys. The initial aim of FoF was to reduce iron deficiency anemia in Cambodian women. Through careful study we determined that anemia in Cambodia was largely caused by genetic blood disorders. Working with HKI and funds from the Canadian Institute of Health Research, we proved that iron was not the cause of anemia and iron might be doing harm. In response the Cambodian government ceased weekly iron and folic acid supplements for non-pregnant women. We also noted that there were high cases of infantile beriberi (thiamine deficiency) in Cambodia. With funding from Grand Challenges Canada and with generous in-kind support from HKI we were able to fortify fish sauce with thiamine and give it to pregnant women to improve both theirs’s and their infant’s thiamine levels in blood. This is now being applied to salt and the team has been given nearly a million USD in additional funding. Directly and indirectly through this project we have helped trained the next generation of Canadian nutrition researchers with skills to work in low income countries; this includes two PhDs, three MSc, three research assistants, and countless undergraduate nutrition, business, and public health students.

AB - 1. EXECUTIVE SUMMARYThis final report is the culmination of two CIFSRF projects carried out in Cambodia, colloquially known as ‘Fish on Farms’ (FoF) and ‘Family Farms for the Future’ (FF4F) that together spanned over six years. The overall aim of both projects was to “to improve household food security and nutrition outcomes, livelihoods, and women’s empowerment through an enhanced homestead food production (EHFP) model using an environmentally sustainable approach”. In part one, FoF, we conducted a cluster randomised control trial (cRCT) of an integrated model of EHFP in which 900 women headed households in Prey Veng province were randomized to either plant-based EHFP, EHFP plus fishponds, or control. The participants received many benefits such as reduced food insecurity, increased fruit and vegetable production, increased fish production in households with fishponds, and increased income that was controlled by women and used to purchase animal-source foods or for children’s education. However, a number of limitations were noted: women were not given a choice in the type of EHFP they received and may have been less motivated resulting in high attrition; almost all the costs of the intervention were borne by the project which in addition to not being sustainable may have led to less engagement by women farmers; insufficient time for EHFP to reach its full potential; issues with capturing the seasonality of crop production and nutrient intake; and lack of information on how EHFP might work in other areas in Cambodia, such as costal, highlands, and peri-urban areas. The second phase of the project, FF4F, was designed to overcome some of these shortcomings. Here we aimed to develop a scalable, cost-effective model of EHFP for Cambodia. We incorporated successful components and lessons learned from FoF, refined tools, methodologies, and practices for scale-up in diverse agro-ecological zones and provinces (Kampot, Kampong Cham, Prey Veng, and peri-urban Phnom Penh) that included 4600 households. These households were able to choose the form of EHFP they wanted, which included a mix of gardens, poultry, and fishponds, but were expected to contribute more to the establishment of these homestead farms. In addition to the scale up, we conducted a pragmatic-delayed cluster randomized control trial (PDcRCT) nested within the larger FF4F study in Kampot province. In short, villages were randomized to EHFP of the households choosing, or control. During the first year, EHFP was introduced into the intervention villages (any type of EHFP) but no measurements were made. In the control group production was quantified monthly and detailed 24-hour dietary recalls were conducted with recipients (women and children) during lean (May) and peak (Dec) production periods in year one. In year two, we quantified production and dietary intake in the groups receiving EHFP. With the stage set for the background we now focus on the key outputs and learnings from the last three years. Initially, households were expected to contribute 50% of the costs of the inputs required to set up their EHFP farms, but with extensive feedback from beneficiaries, this was reduced to 30% to ensure greater uptake of the program. Some households relied on microcredit but most relied on personal family funds. We targeted 4600 households, but through spill-over we estimate that the project reached another 15 households per study village. Members of these households either attended training, education, and/or demonstration events. In total, we estimate that 3,500 non-target households received benefits from this project, thus benefiting an estimated 17,500 family members overall. Attrition was high in non-surveillance households at over 40%. The main reasons given for attrition were migration for other economic opportunities and lack of time and labour resources. Fortunately, attrition occurred early and households could be replaced before considerable resources were spent. All provinces except the city of Phnom Penh took up aquaculture ranging from 23% in Kampong Cham to 54% in Kampot. Poultry raising took place in excess of 90% of houses in all provinces. In addition to initially poor-quality seeds, poultry death was high and plans were taken to mitigate this high mortality. Another strength of the study was that we followed up with both inactive and active households to explore the main factors contributing to retention and adoption of practices promoted by the project.Based on data from our PDcRCT trial, we found significant differences in agriculture production and prevalence of inadequacy for key micronutrients between households that received the EHFP intervention (regardless of model) versus the control group. Specifically, of producing households in each category, on average, households in the EHFP group produced significantly more fruit (507 vs. 306 kg), large fish (23 vs. 15 kg), eggs (117 vs. 99 units), and live birds (99 vs. 89 units) than the control group, however, cumulative vegetable production was similar between groups (267 vs 192 kg). For food intake, we observed a statistically significant reduction in the prevalence of inadequate intake for zinc, thiamin, riboflavin, and vitamin A in women and children in the EHFP group compared to control, during both the lean and peak agricultural seasons. Over the two-year intervention, the project was cost neutral, when accounting for only targeted households. However, if projected out over a ten-year period the incremental net benefit was nearly USD500. Further, if you take into consideration the spill-over effect, the project is likely to have even larger positive net benefit.All project objectives were met. Briefly, we refined technologies and practices from FoF and adapted them for different geographic regions and scale. Different types of fish and stocking densities were tested and small-scale hatcheries were encouraged such that now we feel that Cambodia has a sustainable model for fish hatcheries moving forward, which can be managed by the Fisheries Administration (FiA). We increased agricultural productivity such that the number of households with cultivated gardens rose from 60 to nearly 100%. To date, FF4F farms have contributed an estimated 28,000 kg of fish, 260,000 poultry eggs and 6,000,000 kg of fruit and vegetables to the local food supply. Women’s empowerment increased although we were not able to measure all dimensions. By the end of the study women were making 90% of the decisions around farming; a recommendation was made to CARD that gender be incorporated into their trainings and future policies. We improved access to resources, and/or markets and income. A series of business tools and initiatives have also been piloted and shown to improve crop selection decision-making, beneficiary financial planning, and market access. Using zinc as the model nutrient we were able to show the incidence and mortality rates of zinc deficiency could be reduced by 23.4%We had in mind that our ultimate our goal is to transfer the skills and learnings to the government, private industry, and civil society to benefit as many Cambodians as possible. As major outputs, a dissemination workshop was carried out where Cambodian representatives reaffirmed their support for EHFP. A policy brief has been prepared and a number of manuscripts have been published or are under consideration. HKI is sharing their learnings with neighbouring countries through regional workshops and are implementing parts of the innovative surveillance to better track production of EHFP outputs and costs incurred (inputs) by households. HKI is a key informant to Cambodia’s National Strategy for Food Security and Nutrition which will be developed for the next five-year period covering 2019-2023.In summary, this project has had great successes, a few failures, but with important lessons learned. Agricultural interventions such as EHFP will struggle to show improvement in nutrition indicators, due in part to the short duration of most research projects. Improvements in stunting are inter-generational and include not only greater access to better food but improvements in socio-economic status as a whole. A lot of our modelling is based on a ten-year projection. The reality is we don’t know what happens to these farms when donors pull out. We hope government and other organizations step up to fill the gap, but we simply don’t know the long-term impact of EHFP. It would seem essential to return to these households in five years to see the longer-term impact. Finally, we would like to highlight some of the spin-off successes that stem directly from this project. Through this study we were able to bring the leading expert from Harvest Plus to train Cambodian and international nutrition researchers on how to conduct proper 24-dietary hour recalls in resource-poor settings. We held focus groups to establish common recipes and created a Cambodian Food Composition Database, that is now in the public domain. This will be essential for future researcher planners, including those wanting to do National Nutrition Surveys. The initial aim of FoF was to reduce iron deficiency anemia in Cambodian women. Through careful study we determined that anemia in Cambodia was largely caused by genetic blood disorders. Working with HKI and funds from the Canadian Institute of Health Research, we proved that iron was not the cause of anemia and iron might be doing harm. In response the Cambodian government ceased weekly iron and folic acid supplements for non-pregnant women. We also noted that there were high cases of infantile beriberi (thiamine deficiency) in Cambodia. With funding from Grand Challenges Canada and with generous in-kind support from HKI we were able to fortify fish sauce with thiamine and give it to pregnant women to improve both theirs’s and their infant’s thiamine levels in blood. This is now being applied to salt and the team has been given nearly a million USD in additional funding. Directly and indirectly through this project we have helped trained the next generation of Canadian nutrition researchers with skills to work in low income countries; this includes two PhDs, three MSc, three research assistants, and countless undergraduate nutrition, business, and public health students.

M3 - Commissioned report

BT - Fish on Farms Phase 2: Family Farms for the Future Cambodia

ER -

Green T, Talukder Z, Michaux KD, Moumin N. Fish on Farms Phase 2: Family Farms for the Future Cambodia. 2018. 46 p.