Farmer management of gene flow: The impact of gender and breeding system on genetic diversity and crop improvement in The Gambia

In many tropical countries we can distinguish two seed systems: a formal seed system (comprising breeding companies and national institutes) and an informal seed system, often called farmer seed system (comprising of all farmer activities related to the transfer of seeds). These two systems are intertwined in various degrees for different crops and different regions. Related to these two seed systems are formal and informal crop development systems. In recent years, it is suggested to link these two systems in a more structural way, for example, through Participatory Variety Selection and Participatory Plant Breeding. Various research projects have shown that the integration of these two systems is feasible and provide various advantages compared to formal crop development, such as varieties that better fit farmer' preferences and are better adapted to their conditions, a higher adoption rate of varieties, enhancement of biodiversity and germplasm conservation. It is often suggested that farmer varieties are well adapted to local conditions and that farmers maintain high levels of diversity in their fields, enabling conditions for the creation of new genetic diversity and new varieties. However, how exactly farmers manage such processes, or whether they actually manage these processes at all is not much studied. With better information it will be possible to identify opportunities to maintain and enhance farmer selection practices, and to prevent elimination of local processes of value in maintaining genetic diversity. Hence, the objective of this study was to get a better understanding of local gene flow and its effects on crop genetic diversity (i.e. availability, genetic make-up and diversity of varieties) and as such on the farming system as a whole, and whether these effects are related to or influenced by breeding system, gender, farmer selection, or possibly other agricultural practices, socio-economic factors or cultural understandings, consciously or tacitly.The three main research questions that follow from this objective were:1.     How is gene flow managed by farmers and which factors play a crucial role?2.     What is the impact of gene flow on the adaptability and resilience of the farming system?3.     Based on the answers to questions 1 and 2, what suggestions can be made to integrate formal and informal crop improvement?To answer these questions, the following model of gene flow is used: Gene flow = (seed flow           +         pollen flow)     *         selection pressure To achieve a better understanding how the various factors related to seed flow, pollen flow and selection pressure interact an interdisciplinary approach is used combining natural and social science. A comparative approach is chosen focusing on two key-variables: crop breeding system and gender. In The Gambia millet (an outbreeding crop) is grown by men and rice (an inbreeding crop) grown by women. During the fieldwork, farmer cultivation practices, seed selection and use and management of varieties are compared between millet and rice and between men and women. In the analyses, quantitative and qualitative plant data and socio-economic data are integrated. Chapters 2 to 9 described the various aspects of the model above.In chapter 12, the concluding chapter, these aspects are brought together and the short formula on gene flow (presented in Chapter 1) further elaborated. With this information, and information from Chapters 10 and 11, the second research question is answered. Finally, this information is used to illuminate the third question through some suggestions on how formal and informal crop improvement might be better integrated. Below, a few of the major findings are highlighted.It seems logical that the higher the rate of cross-pollination, the higher the chances are for new varieties to develop in farmers' fields. In Chapter 7, however, it is shown that for pollen flow to result in the development of new genotypes, the level of pollen flow should not be too high. Distinct, stable genotypes develop more easily in crops with a low cross-pollination rate, like rice. Hence, it is also easier for farmers to select off-types in strong inbreeding crops, and develop them into new varieties, than in strongly outbreeding crops. Other factors determining the chances of the development of new stable genotypes are the distances between fields, the time of flowering, the number of varieties in different fields and the number of mixed-in varieties within fields. Farmers do not mind mixed-in varieties in their seed as they consider it insurance. Because of all these factors, chances of cross-pollination between rice varieties are much higher within fields than between fields. Since within a village almost all farmers grow the same variety of millet chances of pollen flow between millet varieties are very low.Although many people (both farmers and scientists) believe that women are better in distinguishing varieties and plant types, Chapter 9 shows there is no difference in capability between men and women. Additionally, both male and female farmers think off-types are caused by God and they do not have a clear understanding of what exactly causes change in rice and millet. Although men and women have a similar understanding on the development of strange off-types in millet and rice, men have never tried to test strange millet off-types, while women sometimes test and develop strange rice off-types into new varieties. As mentioned above, new stable, distinct off-types develop more easily in rice than in millet. Distinctness of off-types is a requirement for farmers to notice them. The harvesting process enables observations of strange off-types in rice more readily than in millet. Additionally, this study found evidence that it is an aspect of male identity not to test any strange off-types. Men often know how to differentiate varieties, but seemingly this is not knowledge that is important to their self image and to their image within the community.The second main research question was whether gene flow plays a role in keeping farming systems adaptable and resilient. In Chapter 10 it was concluded that processes in relation to genetic diversity differ between rice and millet. In the case of rice, pollen flow between varieties and farmer off-type-selection have important roles in the development of new varieties. Moreover, for rice, seed exchange between communities is crucial to maintain genetic diversity at village level and regional level, whereas it is less important for millet. In millet it is primarily pollen flow within varieties and seed selection through which genetic diversity is maintained. It is likely that genetic diversity for rice fluctuates more over time and space than for millet. For rice, the effect of pollen flow and seed flow on genetic diversity shows in the development of new varieties, whereas for millet, seed and pollen flow lead to the maintenance, or in some cases an increase, of genetic diversity at population level.In Chapter 11 it was elaborated that adaptability (and resilience) can be achieved through genetic diversity and plasticity. Chapter 11 elaborated further on the impact of selection on local adaptation. It shows that millet varieties are not narrowly adapted and that variety choice seems more related to cultural preferences than to local adaptation. Although no data are available, it seems that rice varieties are also quite widely adaptable. For rice, another factor that seems to influence variety choice is the time of introduction of varieties in a village and the range of suitable varieties in that village.At farming system level pollen flow likely plays a limited role in keeping populations adaptable to rapid environmental changes. This does not mean that the process of adaptation through genetic evolution at field level should be ignored, but that it is too slow a process, compared to the decision making process of farmers who can decide in a few years (sometimes in only one year) to change varieties.In Chapter 12 the implications of these findings for the importance of gene flow are discussed.Under slowly changing conditions, a high cross-pollination rate contributes to adaptation at field level, which then in turn contributes to adaptation at farming system level. Under fast changing conditions (whether ecological or socio-economic), seed flows (exchange of varieties) are essential to maintain adaptation at farming system level. A low cross-pollination rate also contributes to the adaptability of farming systems through the development of new varieties. Hence, it can be argued that for late millet the process of adaptation is more gradual, while for rice it is more of a stop-and-go process. Whereas late millet populations slowly adapt to the changing climatic conditions in The Gambia, rice varieties tend toward replacement when the 'older' ones do not fit the conditions anymore. Hence,the millet seed system can be considered low-cost and more robust, whereas the rice seed system is more a higher maintenance and less robust system, needing constant exchange and testing of off-types to keep it adaptable.Finally, in Chapter 12 some suggestions are given how to integrate farmer and formal breeding in the context of The Gambia. There is scope to improve the formal breeding system in itself, particularly by improving funding conditions and the capacity of NARI. Important is also to stimulate two-way traffic of information and seed materials between researchers and farmers, and local testing of materials. As farmer breeding is very much a random process, which is also one of its strengths, it is difficult to improve it. There is more scope in improving farmer breeding by linking it to the formal system. One way to link farmer and formal breeding is through PVS methodologies which can be easily integrated into the 'conventional system of crop improvement'. If formal breeding is improved following the above suggestions, PPB methodologies do not have any apparent advantages in the context of The Gambia, but may be used to reflect how conventional breeding methods can be improved. In the last section it is emphasised that as long as formal crop development excludes particular crop types, farmer crop development remains to play an important role, particularly in risk-prone environments. A last remark is that crop development can only improve agriculture as part of a farming systems approach understanding the complex processes through which crops are formed, maintained and improved.