Evaluation of Bioversity’s Efforts to Promote Millets in India Mitch Renkow North Carolina State University March 2019 1 Table of Contents 1. INTRODUCTION ..................................................................................................................... 2  2. BACKGROUND ........................................................................................................................ 4  3. PROJECT INTERVENTIONS .................................................................................................. 9  4. OUTCOME ANALYSIS ......................................................................................................... 12  5. DISCUSSION ........................................................................................................................... 19  ACKNOWLEDGEMENT ............................................................................................................ 22  REFERENCES ............................................................................................................................. 23  ANNEX A: CONTACTS ............................................................................................................ 44  2 1. INTRODUCTION This study evaluates key outcomes associated with Bioversity’s work promoting minor millets in India. That work, part of a broader IFAD-funded effort to promote neglected and underutilized species globally, has been pursued jointly with the MS Swaminathan Research Foundation (MSSRF), as well as with other partners. The Bioversity-MSSRF project, as I will refer to it throughout this document, has been on-going since 2001 and currently is in its fifth multi-year “phase.” Bioversity’s interests in promoting neglected and underutilized species (NUS) generally, and minor millets in India specifically, date back to when it was known as the International Plant Genetic Resources Institute (IPGRI). Those interests relate to the Institute’s core missions of conserving genetic resources, promoting biodiversity, and—because millets in India tend to be grown in less-favored production environments by poor people—enhancing the welfare of limited resource farmers. As in other parts of the world, millet production in the sub-continent had been declining for decades, replaced by other cereals (especially wheat and rice) or cash crops (Grovermann, et al. 2018). Inter alia that meant a corresponding decline in the genetic diversity of the millets being produced, as well as the diversion of agricultural research resources toward more favored crops. Both of these tended to widen pre-existing gaps in the well-being of millet farmers vis-à-vis farmers in other, more favored production environments. Arresting and/or reversing these trends was identified as a means of achieving those missions (Bala Ravi, et al. 2010). It bears mention at the outset that this current evaluation was intended originally to examine outcomes from a particular policy achievement that is very much associated with the research and extension activities associated with the Bioversity-MSSRF project—namely, a key provision of India’s National Food Security Act of 2013 that mandates millets as a scheduled commodity in India’s public food distribution system. Specifically, the National Food Security Act established an enabling environment for government procurement of millets for purposes of distribution to the hundreds of millions of India’s poorer citizens. Clearly, this represents a potentially huge positive shock to overall demand for millets, with attendant benefits to millet producers. However, early on in the course of this evaluation it became clear that those potential benefits are yet to materialize. Only one state (Karntataka) has actually incorporated millets into its public distribution system, and procurement of meaningful quantities only occurred in one year (2015) due to poor harvests in subsequent years (Rajshekar and Raju 2017). In short, it was determined that evaluating the outcomes of the National Food Security Act would be premature at this time. Instead, the current evaluation describes key outcomes attributable to the efforts of the Bioversity-MSSRF project that have occurred in the sites in which the project was implemented. Village-level and household outcomes related to agricultural production, consumption, marketing, and income are analyzed. The basis for those analyses lies in a farm household data set that was collected in the summer of 2018 from two locations in southern India where Bioversity-MSSRF teams have been conducting research since the project’s beginning. The 3 survey used was based on similar surveys that were administered in 2007 and 2010 (at the beginning and end of Phase II of the Bioversity-MSSRF project). Comparison of the 2018 data with the earlier data provides insights into the extent to which the Bioversity-MSSRF project has met the “meta-goals” of promoting genetic conservation of millets and improving livelihoods of limited resource millet farmers. It also enables us to make some salient observations on the extent to which future demand increases for millet—e.g., due to government procurement for the PDS—may be met by the limited resource farm households that currently account for the bulk of millets production in India. Carrying out this evaluation involved meeting with Bioversity and MSSRF personnel who had been involved in the project over the years; reviewing written outputs of the research; extensive email exchanges and telephone conversations with individuals having knowledge of the various research activities; and a field visit to India in December 2017. Interlocutors during the field visit included current and former Bioversity and MSSRF staff, government officials, academics, members of the international donor community, and villagers in one of the two study locations (Kolli Hills). Annex A provides a listing of these individuals. The report is laid out as follows. The following section provides background information on the project. This section also discusses the National Food Security Act, focusing particularly on how the Bioversity-MSSRF project contributed to its salutary provisions with respect to millets. Following this, I explore the changes that have occurred in the study areas. Section 3 describes the various interventions that were promoted. Section 4 examines key village-level outcomes. The concluding section summarizes key findings and discusses what these outcomes imply as regards the welfare effects of future demand increases for millets on the kinds of limited resource millet farmers impacted by the Bioversity-MSSRF project. 4 2. BACKGROUND The IFAD-NUS Project Bioversity’s collaboration with MSSRF on neglected and underutilized species (NUS) began in February 1999 when MSSRF hosted an IFAD-supported workshop on NUS at their headquarters in Chennai. That meeting, entitled “Enlarging the Basis of Food Security: The Role of Underutilized Species,” was held under the auspices of the CGIAR’s Genetic Resources Policy Committee, and was attended by representatives from IFAD (among other selected donors). Subsequent to that meeting, IFAD indicated its support of a comprehensive global effort to promote research on neglected and underutilzed species, as well as the extension of that research to various stakeholders (farmers, researchers, policymakers). Shortly thereafter, Bioversity— then the International Plant Genetic Resources Institute (IPGRI) —organized three regional exploratory workshops. Out of this background work emerged the first of several phases of a long-term program of grants from IFAD to a variety of institutions for purposes of promoting so- called “Neglected and Underutilized Species. Those grants continue to fund research undertaken by the Bioversity-MSSRF team up to the present (Table 1). The IFAD-NUS grants target the conservation and use of neglected and underutilized species in Latin America, South Asia and Western Asia. Targeted crop species included Andean grains (quinoa and amaranth); medicinal and aromatic plants in West Asia; and minor millets1 in South Asia. All of these target crops played a key role in nutritional security and/or income generation for specific limited-resource—and frequently culturally disadvantaged—rural populations. As in other locations, the work on minor millets in India has been guided by a ‘holistic value chain approach’ that “sought to improve sustainability of production and income generation from target crops, as well as to secure nutrition benefits for local communities to achieve livelihood and conservation outcomes” (Bioversity International 2007). This has given rise to a range objectives—and attendant mechanisms for reaching them—across multiple dimensions (Bioversity International 2018):  Increase the contributions of millets to improved livelihoods. The mechanisms for achieving this objective centered on promoting best agricultural practices for sustainably increasing productivity, maintaining local biodiversity, enhancing the capacities of stakeholders in value-adding technologies, and strengthening value chains within which millets and millet-based products are marketed.  Identify the contribution that key species can make to dietary diversity and to improved nutrition, particularly in children and expectant women. In the case of millets in India, this included promoting small-scale technologies for processing millets that require substantially less effort on the part of women. 1 Minor millets grown in India include finger millet (Eleusine coracana), foxtail millet (Setaria italica), proso millet (Panicum miliaceum), kodo millet (Paspalum scrobiculatum), little millet (Panicum sumatrense), and barnyard millet (Echinochloa colona) (Riley, et al. 1993, pg. 557). Of these, finger millet is by far the most widely grown and has been the subject of most attention by crop breeders (including ICRISAT). Note that pearl millet is not considered a “minor millet.” Throughout this report, I use the term “millets” to refer to the group of minor millets noted above. 5  Integrate minor millets into an increased range of production systems, ensuring their improved use and biodiversity maintenance in these systems  Conservation of genetic resources of millets, both on-farm and ex-situ.  Support the development of multi-sectorial national policies to promote an enabling environment for conservation, production, and profitability of millets, as well as to strengthen collaboration among institutions working on millets. The Bioversity-MSSRF Collaboration Bioversity’s work on millet in India has been conducted in complete partnership with the MS Swaminathan Research Foundation. MSSRF entered that partnership having been involved in a variety of development projects in millet-growing areas. These projects were centered around community development and women’s empowerment, particularly as they related to the well- being of the limited-resource, culturally disadvantaged groups who remain the primary millet cultivators in India. In practical terms, this meant that MSSRF had solid working relationships with a large number or stakeholders multiple communities in many millet-growing regions.2 For its part, Bioversity brought a significant catalytic role, both scientifically and organizationally. On the science side, Bioversity provided a guiding hand in terms of overall research strategies, methodologies and approaches, monitoring of research activities, and synthesis of results. Organizationally, Bioversity facilitated collaborative partnerships with a variety of research institutions in academia, the CGIAR, government and the NGO sector. The Bioversity-MSSRF collaboration sought to meet the objectives listed above by (a) developing networks providing participatory systems and procedures to support the improved production and use of millets; (b) forming operational alliances among selected network partners to overcome production and marketing constraints (e.g. seed supply systems, processing groups, distribution cooperatives); (c) linking to existing rural and economic development projects where millets contribute to incomes and nutrition; (d) increasing the capacity of marketing associations and producer groups linked to millets; and (e) raising awareness among policy-makers of issues and options for improved policies and laws affecting millets (IPRGRI 2000). A comprehensive overview of the many activities, achievements, and partners in the Bioversity- MSSRF project are laid out in Table 2, which is an adaptation of a somewhat more comprehensive listing in Padulosi, et al. (2015). The outcome analysis reported in Sections 4 and 5 presents information on the cumulative effect of these various activities. The National Food Security Act of 2013 The National Food Security Act (NFSA) of 2013 enacted a sweeping rearrangement of India’s Public Distribution System (PDS)—a keystone of India’s social safety net for its hundreds of millions of eligible, poor people. The headline feature of the NFSA is that it institutionalized an 2 Institutionally, MSSRF tends to orient its efforts more toward development and extension activities than research per se, according to several sources. By this accounting, the project’s focus on the community development and women’s empowerment issues were more in MSSRF’s bailiwick, whereas facilitating research on farm-level productivity or production enhancement lay more in the wheelhouse of Bioversity. 6 entitlement of 5 kg per person per month for up to 75% of rural dwellers and 50% of urban dwellers.3 Of key interest here is the fact that millets were included with rice and wheat as grains to be procured by the government and distributed through the PDS. This, of course, represents a potentially huge increase in demand for millets. In addition to changes to the PDS and the system of procurement that underpins it, the NFSA addressed a range of other safety net programs, most notably operation of Mid-Day (school) Meal scheme and the Integrated Child Development Services program serving pregnant women, lactating mothers, and children under 14 years of age. These, too, represent potentially very large sources of institutional demand for millet. Inclusion of millets in the PDS represented the culmination of years of effort to promote millets by, among others, individuals associated with the NUS project under consideration here (see Box 1). In particular, Professor Swaminathan himself was a tireless advocate for millets. His advocacy was particularly effective due to his prominence as the “Father of the Indian Green Revolution,” as well as his political connections (he served as a member of Parliament from 2007-2010, and chaired India’s National Commission on Farmers from 2004-2014). But there is no doubt that the research-based evidence on the nutritional, environmental, and economic benefits of millet production developed as part of the NUS project greatly focused the power and persuasiveness of those advocacy efforts as well (e.g., Bioversity 2007; King, Nambi, and Nagarajan 2008; Padulosi, et al. 2009; Bhag Mal, Padulosi, and Bala Ravi 2010; Padulosi, et al. 2015). The research forming the basis for these advocacy efforts touched on the various dimensions along which millets are superior dominant cereals cultivated in India (rice and wheat). These included:  Production benefits – millets mature quickly, grow well under varying day lengths and soil quality.  Environmental benefits – millets are less water-intensive than rice and wheat, and hence are less sensitive variable weather conditions than other cereals. They tend to be grown in rainfed conditions, so do not put pressure on groundwater resources like irrigated crops. They are also typically cultivated with minimal fertilizer inputs  Nutritional and health benefits – compared with other cereals, millets possess superior nutritional characteristics due to greater availability of minerals, polyphenols, anti- oxidants, and soluble fibers.4  Social benefits – millets are generally grown by limited resource—and commonly, socially disadvantaged (lower caste) farmers living in marginal production environments. 3 Note that under the NFSA, millions of below-the-poverty-line households covered under the pre-existing Antyodaya Anna Yojana program would continue to be eligible to receive their larger entitlements of up to 35 kg per household per month. Note further that the NFSA included a number of provisions related to logistics, organization, operation, and institutional accountability, all of which betokened an expansion of India’s food security safety net (Government of India 2013). 4 Dr. Thingnganng Longvah (Director of the National Institute for Nutrition), pers. comm. 7 Box 1. Policy Precursors to the National Food Security Act Key members of the Bioversity-MSSRF collaboration can credibly lay claim to having played an important role in the formation of several national agricultural policies related to plant genetic resources for agriculture (GPRFA).  Biological Diversity Act of 2002  Protection of Plant Variety and Farmers’ Rights Act (PPV&FR Act) 2001  Patents Act (most recently amended in 2005)  Modifications to the Seeds Act (1966) A recent paper by Notaro et al. (2017) discusses each of these specific policies in greater detail. In brief, the these pieces of legislation represent successive (clarifying) components the legal framework governing intellectual property rights in the areas of biodiversity and plant varietal protection. Importantly, millets were included in the roster of covered crops in these Acts—in very large measure due to the persuasiveness and political influence of Professor Swaminathan himself. Indeed, one well-placed interlocutor contended that the “master plan” of Prof. Swaminathan was first to promote institutionalizing biodiversity protections—in the form of the Biodiversity Act and the PPV&FR Act, and then to promote greater government demand for millets (as manifested in the Food Security Act). To date, there is no evidence of on-the-ground impacts of these pieces of legislation on growers of millets. I was told by a number of interlocutors that this is because there have been as of yet no legal tests of the IPR protections—i.e., that they remained “on paper” at least as of December 2017. Moreover, it remains highly unclear how to identify—let alone operationalize compensating for access rights—the “owners” of property rights over traditional varieties of millets that have been bred collectively over long periods of time. 8 While the NFSA was national legislation, it largely left the mechanics of its implementation of the India’s massive food safety net—both procurement and distribution—to individual states. On the demand side, States have substantive authority over:  determining which households are eligible to participate;  what commodities to include in the foods made available to PDS recipients—i.e., whether or not to include minor millets;  the size composition of food bundles available to PDS recipients (for example, at a PDS shop we visited in Bangalore PDS beneficiaries were entitled to monthly allotments of 5kg of rice and 2 kg of either wheat or finger millet); and  the (subsidized) prices charged for different commodities made available to PDS recipients. On the supply side, India’s states have long had significant latitude in procurement of grains under the decentralized grain procurement scheme that was introduced in 1997 (Notaro, et al. 2017). In particular, states are authorized to offer commodity-specific premiums over nationally- determined minimum support prices that are announced annually for a range of cereals (including finger millet) and other crops. Such premiums are frequently required in order for State governments to meet procurement targets in a (relatively) timely fashion. Of particular relevance here is the fact that that MSPs for coarse grains (including millets) historically have been set at levels below prevailing market price in most years, and hence have been ineffective (Notaro, et al. 2017). The NFSA was enacted into law in 2013. Since that time, Karnataka is the only State in India that has incorporated minor millets—specifically, finger millet—into the PDS.5 And to date there has only been one year in which meaningful quantities of finger millet were available to through the PDS (2014-15). Poor harvest in subsequent years severely curtailed the Government of Karnataka’s ability to procure finger millet, and hence there was virtually none provided to PDS recipients. Consequently, there is no scope at this time for ex post evaluation of the impacts of the NFSA on specific populations of interest; rather, the NFSA represents a source of potential impacts that merit future examination. Of particular interest here would be an evaluation of the extent to which different types of producers and consumers of minor millets will be affected by the NFSA, and by how much. In the next sections of this report, I take steps towards informing such an evaluation with regard to farmers reached by the Bioversity-MSSRF research teams in their work in the Tamil Nadu and Odisha. 5 Two other States—Tamil Nadu and Chhattisgarh—have engaged in procurement and distribution of pulses through the PDS. And pearl millet (jowar) has been procured and distributed in a number of states. But in general, rice and wheat are overwhelmingly the dominant commodities transacted through the PDS in all States. 9 3. PROJECT INTERVENTIONS This section describes key interventions undertaken by the Bioversity/MSSRF project in two locations in South India: (a) the Kolli Hills area of Namakkal district, Tamil Nadu; and (b) Kundura tehsil of Koraput district, Odisha. Both locations have been the locus of research and extension activities since the outset of the project. Kolli Hills is regarded as a secondary center of origin for minor millets. Prior to the initiation of the IFAD-NUS project, the area had seen a steady erosion of both area devoted to millets, and an attendant attenuation of genetic diversity of millet populations (Takeshima and Nagarajan 2012). Kundura is similarly an area in which a number of minor millets were grown historically and in which millet—particularly finger millet—is an important staple. The set of traditional agricultural practices in the Koraput region that includes Kundura has been designated a Globally Important Agricultural Heritage System by the FAO (Down to Earth 2015). Both the Kolli Hills and Kundura are populated by limited resource farmers; tribal and scheduled castes compose the majority of populations in both areas. When the IFAD-NUS project was initiated, MSSRF had operated in both locations for a considerable amount of time, working directly with those limited resource populations on projects across a variety of areas related to biodiversity conservation, sustainable agricultural production, nutrition, and the building of grass-roots institution (Nampoori and Parida 2015). By all accounts, this pre-existing network of contacts and relationships greatly facilitated relatively smooth implementation the many household level research and extension activities associated with the project. As noted earlier, the NUS project sought operated on four key dimensions: conserving genetic resources, enhancing the productivity of millet cultivation, promoting increased millet consumption, and facilitating expansion of commercial activity involving millets and millet- based products. Figure 1 provides a summary overview of the types of project interventions and activities related to these four dimensions. Figure 2 presents a rough timeline over which these activities occurred. Specific interventions are described below. Genetic Conservation At the initiation of the project, millet production had declined and relatively few varieties were being planted. Interventions related to genetic conservation sought to address this issue by evaluating, testing and promoting utilization of a broader array of germplasm resources. These included the following activities engaged in over the period from 2002 to 2007:  Collection and ex-situ storage of samples of local landraces of different millets. In Kolli Hills these were primarily finger millet, little millet, and foxtail (Italian) millet; in Kundura they included finger millet and little millet.  Field testing of 6,000 varieties of finger millet, little millet, and foxtail millet. In addition to local landraces, accessions of improved varieties were made primarily from ICRISAT’s gene bank, but also from the All India Coordinated Small Millet Improvement Project (AICSMIP) and from the holdings of the University of Agricultural Sciences-Bangalore. Field testing of these varieties took place on farmers’ fields. 10  Participatory varietal selection (PVS). Joint—and largely congruent—evaluations by farmers and scientists, identified a number of varieties with desirable characteristics. These included some landraces and some improved varieties, depending on agronomic circumstances. Generally, landraces of little and foxtail millet were found to dominate improved varieties, whereas improved varieties of finger millet were generally favored. Agronomic Interventions A variety of participatory crop management experiments gave rise to a set of recommended agronomic interventions. As with PVS, the agronomic trials were conducted on farmers’ fields, and subsequent determination of recommended practices were conducted over the period 2002- 20010.6 Demonstrations on farmers’ fields were oriented around comparing improved and traditional practices. Improved practices included the following (Bala Ravi, et al. 2010):  Use of farmer preferred millet varieties  Use of quality seed  Row planting instead of broadcast sowing  Altered (lower) seeding rates  Standardized row ratios for intercrops  Use of farmyard manure or fertilizer  Regulating plant population density via thinning  Inter-cultivation and (need-based) top dressing with fertilizers  Novel intercrop systems— Seed Production and Management At the initiation of the project, the variety, quantity, and quality of available millet seeds was identified as a substantial constraint—both on efforts to conserve genetic resources, and on efforts to improve productivity. In response, quality seed production training was undertaken in the early years of the project. Those trainings included lessons on variety characteristics and optimal handling of seeds at various stages of production (harvest, threshing, drying and storage). Following up on those training activities, the project has actively sponsored and promoted the institution, operation, and management of village seed banks. Seed banks handling quality seed of various types of millets have been established in a number of village—typically under the aegis of local, cooperatively-managed self-help groups (SHG’s). Seed banks generally operate under a system whereby individuals are required to return 2 kg of quality seed for each kg that they “borrow” from the seed bank. Most seed banks that have been established under the auspices of the project continue to operate to this day. Table 3 provides data on seed bank usage 6 Note, however, that various agronomic field research projects occurred well beyond this period. For example, I observed on-farm pigeon pea-millet intercrop trials and bio-fertilization experiments during a site visit to the Kolli Hills in December 2017. 11 in the Kolli Hills between 2001 and 2014 (comparable data were not available for Kundura). These indicate that with the exception of a few years, seedbanks have been widely used throughout that time period. Most transactions involved little millet, foxtail millet, and finger millet, although small quantities of Kodo and Proso millets also have been transacted, particularly since 2010. Household Consumption and Marketing The preceding set of interventions related to millets production and the genetic resources that underpin it. Substantial effort has also been devoted to various aspects of household millets consumption and marketing. These centered around three areas of effort: (a) reducing drudgery associated with processing millet grain; (b) developing an array of millet-based products to be produced and sold from own-produced grain; and (c) nutritional training aimed at promoting in- home consumption of millets and millet-based products.  Processing. Historically, millets were an important staple in project areas. But household consumption has always been constrained by the fact that processing millet grain into flour by traditional means—i.e., via manual pounding and grinding—is physically demanding and highly time-consuming. The project invested in innovating small-scale processing machinery that was inexpensive and easy for limited resource persons to use. Diffusion of suitable processing machinery to villages was then facilitated by trainings, as well as through the development of cooperative SHG’s to operate and manage village grinding mills.  Value-added products. An important driver of project activities is the belief that cultivating sustained demand for millet over the long run is essential to sustaining any advances made on the production side. Consequently, considerable attention was paid to building value chains for millets and millet-based products. At the village level, this involved substantial training of village women in the production of so-called value-added products—e.g., cookies, dumplings, malt beverages, cooked breakfast cereals. The focus of these trainings ranged from recipe development to use of mechanical inputs for production and packaging to marketing, generally under the aegis of dedicated SHG’s.  Nutritional training. Training of village women on the nutritional benefits of millet consumption were a standard aspect of value-added trainings noted above. A goal of those trainings was to promote better nutrition (especially of children) via greater household consumption of millet products.7 In addition, on-going efforts promote substituting finger millet for rice in school lunches which, aside from promising nutritional benefits to children, would also hold the potential for cultivating an important source of institutional demand for millets.8 7 Interestingly, and somewhat paradoxically, nutritionists involved in these trainings would appear to have been simultaneously promoting two opposing uses (hopefully greater) of millet output: greater own-consumption and greater sales. Such an come is possible, of course; it is more, though, some households would simply gravitate toward one or the other. 8 The idea of partially substituting finger millets for rice in school lunch programs as a means for the national government to promote (finger) millet value chains was a common discussion point in conversations with many persons interviewed during field visits, including the Secretary to the Minister of Agriculture, Dr. S.K. Pattanayak. 12 4. OUTCOME ANALYSIS In this section I present information on village-level outcomes from a variety of interventions implemented by the Bioversity-MSSRF team in five villages in the Kolli Hills and six villages in Kundura (see Box 2 for a more detailed description of these data). The outcomes of interest relate to agricultural production, consumption, marketing, and income. The interventions of interest are those summarized in the previous section. These took place during the first two phases of the NUS project—i.e., the ones that had been the focus of the 2007 and 2010 surveys. Treatments Tables 4a and 4b indicate the number of households in each village that were involved in the various interventions. The 2018 survey asked respondents whether or not their household had ever had adopted a practice participated; and whether they were still involved in the practice. In general, these data indicate that substantial fractions of respondents had been exposed to, and had adopted some recommended practices, in most villages. In the Kolli Hills, between one half and two-thirds of sample households took part in the participatory varietal selection (PVS) and seed quality training activities, depending on the village. Adoption of a couple of crop management activities—thinning and weeding—was even greater among surveyed households (80%), whereas for the other agronomic interventions listed adoption was rather low. There was some rather modest disadoption of improved crop management interventions in evidence; but depending on the intervention, roughly 80-90% of respondents who indicated that they had ever adopted were still engaged in the activity in 2018. In Kundura, household participation in PVS and seed quality training was even greater than in Kolli Hills: 85% of surveyed households reported having participated, and participation was greater than 75% in all villages. As in Kolli Hills, some improved crop management practices were adopted by high proportions of the surveyed households, while others were not (in Kundura fertilizer use was more widely adopted, whereas thinning was less adopted). And as in the Kolli Hills, some disadoption of improved crop management techniques was reported in Kundura as well. But in general, uptake of improved crop management practices was substantial and sustained in both locations. It is interesting to consider seed bank usage reported by survey respondents. Seed bank usage provides a signal as to the effectiveness of seed quality training, and the attendant knowledge transmitted about the benefits of good management of genetic resources. Roughly 60% of households in Kolli Hills reported that they had used community seed banks, compared with only about 21% of households in Kundura. Moreover, of the households in Kolli Hills that had ever “borrowed” from seed banks, 80% reported still utilizing them as a source of seeds. In Kundura, by contrast, most of the (relatively few) households who had ever used their community’s seedbank reported not using it in 2018.9 9 These differences across locations reflects more widespread development of SHG’s in Kolli Hills than in Kundura. Designated SHG’s for seed banks have proven to be very effective for effectuating seed bank management and operation (King, Kumar, and Padulosi 2015) 13 Box 2. Data Collection Midway through this consultancy we uncovered two agricultural household data sets collected in 2007 and 2010 (see Attachments A and B). The data were collected at the beginning and end of Phase II of the IFAD-NUS project. To the best of my knowledge, and for reasons that are unclear, the data have never been highlighted in published research.1 The data were collected in two locations in which the participatory research, training and extension activities of the Bioversity-MSSRF teams were focused—Kolli Hills, Tamil Nadu and Kundura, Odisha. A total of 246 households were surveyed in 2007: 148 households in five villages in the Kolli Hills; and 98 households spread across 6 villages in Kundura. The 2010 survey re- surveyed 134 households in Kolli Hills and 87 households in Kundura. The surveys had collected information on production, consumption, income, and marketing both from households who had been involved in the project activities and from households who had not. Examination of the data indicated that it had been competently assembled, and that it was presented in a way that was easily accessible. In short, it appeared to represent a useful benchmark against which to measure changes in household production, consumption and marketing behavior in areas which the Bioversity-MSSRF teams had been most active. Following up on this discovery, in mid-January 2018 Bioversity contracted with MSSRF to oversee a follow-up study to collect comparable data from as many of the same households as could be located. A letter of agreement was signed on March 1. Over the next couple months the survey instrument was crafted (see Attachment C); and MSSRF hired and trained enumerators in the study locations. Data collection took place in May and June. 202 of the 221 households that had been surveyed in both 2007 and 2010 were re-located—127 in Kolli Hills and 75 in Kundura. All agreed to participate in the 2018 survey. These data were cleaned and assembled in July, and the final data set was delivered at the beginning of August 2018. ____________________________ 1. Curiously, the data were found on a hard drive at Bioversity Headquarters in Rome. Bioversity and MSSRF colleagues had, at best, vague recollections of the data having been collected. Neither did other interlocutors with whom I interacted over the course of this research. 14 Also of interest is that roughly one-fifth of households in Kolli Hills, and one-sixth in Kundura, reported having undertaken training in production of value-added products. As noted earlier, significant emphasis was placed on such value-added production as both an impetus to stimulating greater millet production and, relatedly, as a contributor to household incomes. Before turning to considering what the data say about production, consumption and incomes, one caveat regarding the data merits mention. While there is a healthy mix of households that were and were not exposed to these various interventions, it is probably inappropriate to consider them as treatment and control populations in a quasi-experimental setting. Seed quality and value- added training activities extended knowledge that could easily be transferred to others afterwards (e.g., via participation in SHG’s). Likewise, knowledge about improved crop management practices no doubt passed from neighbor to neighbor over the years. Knowledge of yield increases associated with varieties selected via PVS would no doubt be widely shared, as would access to those seeds (via seed banks as well as more informal mechanisms). Instead, the approach taken here will be to examine mean outcomes on a village-by-village basis—i.e., I assess what might be termed “village treatment effects.” Our primary purpose is to track (a) trends in production, consumption, marketing of millets across the three survey years (2007, 2010, an 2018); (b) the relative importance of millets vis-à-vis other competing crops; and (c) the footprint of millets in household livelihood strategies—i.e., how earnings from millets cultivation (including value-added activities) compare to other sources of earnings. For all these purposes, examining central tendencies at village-level is appropriate. Area This section presents information gleaned from the surveys on trends in area cultivated in the various villages. For Kundura, only rice and millets production data were collected in the 2007 and 2010 surveys. The Kolli Hills data set contains a richer array of production data for all years. Of particular interest here is ascertaining the extent to which millet area increased over time (an explicit goal of the project); and also to gauge the relative importance of millet production vis-à-vis production of rice and, in Kolli Hills, cash crops. It is important to note that in both areas—but particularly in Kolli Hills—farms are generally quite small. In 2018, for example, only 14 of 127 surveyed households in the Kolli Hills cultivated more than 5 acres (with the maximum being 9 acres); and 49—roughly 40% farmed less than 2.5 acre. Total area cultivated in Kundura is greater on average. But even there, roughly two-thirds of surveyed household operated 5 acres or less; and only 8 farms cultivated 10 acres or more (with a maximum of 17 acres). Table’s 5a and 5b present sample area means by village for the two locations. In the Kolli Hills sample, millets area more than doubled over time. There was some inter-village variation, but substantial percentage increases were observed in all villages. Finger millet has consistently been the dominant millet type grown, although modest increases in area under little millet and foxtail millet were in evidence as well. Note, however, that despite this increase in millet cultivation, the area devoted to millets in 2018 was considerably smaller than the area devoted to rice across all Kolli Hills households (and for all but one village). And the area devoted to cash crops— especially tapioca, but over time a more diverse mix—dwarfs area devoted to cereals. 15 In Kundura, where the average acreage cultivated is substantially larger than in Kolli Hills, a different pattern is evident in the data. In all villages, finger millet area was the same or less in 2018 than in earlier years—in several cases by quite a bit. Part of this decline represents switching into cultivation of little millet; but total area under millets area fell in all but one village. Paddy area also displayed a similar pattern of declining acreage, but the decline in paddy acreage was not as steep. Interestingly, reported finger millet area had been greater than rice area in most Kundura villages in 2007, but by 2018 overall rice area exceeded that of finger millet area. Another way of gauging changes in the importance of millets is represented by the number of farmers planting millets. The data suggest that the footprint of millets has definitely increased in each location, but in different ways. As indicated in Table 6, the number of Kolli Hills farms growing finger millet increased from 42 in 2007 to 76 in 2018; and average area devoted to finger millet among those farms was about three-tenths of an acre in 2018. The small number of Kolli Hills farmers growing little millets was the same in 2018 as it was in 2007, but average area planted on those farms doubled (from 0.33 to 0.66 acres); and about 5% of farmers had begun growing foxtail millet by 2018 (none was planted in earlier years). In Kundura, both rice and finger millet were cultivated on nearly all farms in all years. Reported finger millet acreage fell over time. But this decline was paired with an increase in acreage devoted to little millet—in 2018, over half of the Kundura farmers interviewed reported growing little millet (none had grown any previously). Those farmers, all of whom also grew finger millet, appear to have substituted to some extent the little millet production for finger millet production. The indication here is that the importance of millets—both the number of farmers choosing to grow them and/or the variety in different types of millets cultivated—has grown over time. But the scale of millet production is small by any reasonable standard, a reflection of the socio-economic circumstances of the populations living in the study areas. Production and sales Tables 7a and 7b present data on output of the main crops grown. The quantity of millets produced has generally increased over time.10 Consistent with area allocation patterns discussed above, finger millet output greatly exceeds output of other millets. Interestingly, more little millet was produced than finger millet in a couple Kundura villages in 2018. However, it is also clear from Tables 7a and 7b that the output of millets is considerably smaller than that of paddy, a reflection of the fact that millet yields considerably lower than paddy.11 In addition, the evidence from Kolli Hills highlights the fact that volumes of cash crops produced, especially tapioca, dwarf cereal production. Additional insight into the size distribution of finger millet production across years and locations is presented in Table 8. There I have tabulated the number of households in each location 10 An exception to this is that most millet growers in Kundura produced very well in 2010. Paddy output in that year was also high relative to other years, suggesting that 2010 was a particularly good year for farming. 11 In 2018, millet yields averaged 628 and 237 kg/acre in Kolli Hills and Kundura, respectively, among households growing millets. In comparison, paddy growers achieved average yields 1,106 and 1,256 kg/acre, respectively. 16 growing different amounts of finger millets. In the Kolli Hills practically all finger millet- growing households produced less than 250kg in all three years. Production has clearly increased there, as evidenced by the substantially greater number of farmers growing more than 100 kg in 2018 compared with earlier years (as well as the larger number of finger millet producers). In Kundura, the pattern is a bit skewed by the fact that 2010 appears to have been a bumper year for finger millet production (with 12 farmers there reporting having harvested in excess of 1 ton in that year). Comparison of the distributions for 2007 and 2018 indicates that production outcomes were more widely dispersed in 2018: a substantially larger number of households produced more than 250 kg than in 2007; but at the same time, a larger number of households in 2018 reported producing very little output (less than 100 kg) compared with 2007. Table 9 presents information on the millet sales. In line with production data, the number of households engaging in millet sales—and the average quantities sold—were substantially smaller in the Kolli Hills. In the Kolli Hills in 2018, only eight out of 76 finger millet producers sold any of their output. Among those who did sell, the average quantity sold was rather small (a little over 50 kg); and that average was dominated by one household that sold 140 kg (most of the rest sold about 30 kg). Both the number of sellers and the mean quantity sold were very similar to what was observed in 2007.12 Also in 2018, a couple farmers sold rather sizeable quantities of little millet (200kg and 300kg). Compared with Kolli Hills, production and sales have consistently been greater in Kundura. Over 40% of finger millet producers in Kundura sold some fraction of their output in 2018. Of these, about half sold 100kg or less and about a quarter marketed 300 kg or more. Unlike the Kolli Hills, both the number of finger millet producers selling finger millets and the volume of those sales increased substantially compared to 2007. The most striking aspect of the data presented in Table 9 is that marketing of little millets was substantial in Kundura in 2018. This is a striking result, given that no little millet production was observed in earlier years. 40 of the farms in our sample grew little millets in 2018. Nearly all of them (37 of 40) sold some of their output, and the fraction kept for home consumption was quite small: The entire group of sellers sold well over 90% of their output (and 20 of them so sold all of their little millet output). Moreover, average quantities sold were significantly greater than those for finger millets. Roughly 40% of selling households sold 100 kg or less; but substantial fraction (one-third) sold amounts exceeding 300 kg, and five households sold 500 kg or more. The clear implication here is that Kundura farmers growing little millet in 2018 were doing so for commercial—not consumptive—purposes. Consumption Tables 10a and 10b present information on average monthly cereal consumption for the two study locations. Focusing first on the 2018 data, it is clear that currently rice is a far bigger component of average diets than are millets. In Kolli Hills, households consumed roughly 14 times more rice than millet; in Kundura the ratio is six to one. Focusing on millets specifically, 12 A somewhat greater number sellers were observed in 2010, and average amounts sold were smaller. Again, this is a reflection of the fact that 2010 appears to have been a year of good millet harvests in which some modest amount of surplus (over consumption demands) was available for sale. 17 most consumption in 2018 was of own-produced (as opposed to purchased) finger millet. Finger millet is the only millet consumed in the Kolli Hills. A small amount of little millet consumption was reported in some villages in Kundura, but finger millet dominates there as well. With regard to consumption trends, rather different patterns are in evidence between the two locations. In Kolli Hill average millet consumption has increased steadily over time in practically all villages. As noted above, though, millet consumption remained considerably less important than rice consumption, since reported rice consumption also increased over time. In contrast, average millet consumption among surveyed households in Kundura was considerably smaller in 2018 than in previous years. What explains this fall is unclear. But one possibility that presents itself is that availability of rice through the PDS may well have contributed to a shift from millets to rice among some households. The data in Table 10b reveal that by 2018 the PDS replaced other market sources of rice pretty much completely in Kundura—hardly a surprise given the significant subsidies on rice purchased through the PDS. Households in Kolli Hills also procured substantial quantities of rice from the PDS. There, however, it would appear that PDS rice largely replaced own-produced rice. Income Surveyed households were asked to provide information on crop-specific cost of production and gross income.13 Gross income data were cross-checked against reported production and price information. Households were instructed to include the value of their own labor in their cost estimates, although it cannot be known how respondents valued their time. All data were deflated by the All-India Rural CPI (2010=100). Tables 11a and 11b presents mean income by source and village. In both locations earnings from agriculture, both via net returns from farming and agricultural wage labor income, dominated overall earnings. Interestingly, in Kolli Hills the importance of (off-farm) agricultural wage income has grown considerably; by 2018 it had surpassed crop production as the number one source of earnings among sample households. This presumably reflects the fact that many individuals in the Kolli Hills routinely engage in seasonal migration to work on farms in other locations (O. King, pers. comm.). Of particular note is that in both locations, earnings from cultivation of millets has consistently been much smaller than earnings from competing crops. In the Kolli Hills, cash crops like tapioca dominate total earnings from cultivation. But paddy production, too, appears to have been substantially more important income source. In Kundura, paddy production has been substantially more remunerative in nearly all villages in all years. For 2018—the only year for which production data were available for other crops in Kundura—net earnings from other crops also exceed earnings from millets production in four of six village. Finally, relatively small value-added earnings were observed for 2018 in all villages in Kolli Hills, and all but one village in Kundura. In terms of overall averages, the scale of these earnings is low—but in Kolli Hill, of a comparable size to reported net returns from millets 13 Households were instructed to include the value of their own labor in their cost estimates. 18 cultivation. Early adoption of production of value-added millet products was also observed in Kundura in 2010. A more in-depth view of the footprint of value-added production is provided in Tables 12a and 12b. One-third of surveyed households in Kolli Hills indicated that they engaged in such activities; and earnings therefrom averaged under 500 rupees. For Kundura, fewer households engaged in value addition. but average earnings for those that did so were substantially higher— 1,583 rupees in 2018 (largely due to large earnings from value-addition reported in one village in particular). Moreover, participation in this activity declined significantly in Kundura over time: more than half of the households who had been involved in value-added production there in 2010 had ceased to do so by 2018. 19 5. DISCUSSION Key Findings The analysis presented in the previous section suggests, that taken as a whole, the many interventions promulgated by the Bioversity-MSSRF project have contributed to the scale, productivity, and diversity of millets production in the villages in which those interventions were undertaken:  Generally speaking, uptake of improved agronomic practices was substantial and sustained in both Kolli Hills and Kundura, involvement in participatory varietal selection was extensive, and seed bank use was widespread (more so in Kolli Hills).  The importance of millets—as indicated by the number of farmers choosing to cultivate them and/or the variety in different types of millets cultivated—has risen over time.  Historically, finger millet has been the dominant type of millet grown in both locations, and that continues to generally be the case. However, the data indicate that the area and production of little millet has increased significantly, particularly in a couple of villages in Kundura.  In Kundura, where farm sizes and area under millets tends to be larger, the number of farmers selling millets (both finger millet and little millet) has risen over time; and in the case of little millets, these appear to have been largely grown for commercial purposes. These findings bespeak significant success on the part of the Bioversity-MSSRF partnership in reaching key goals motivating the project—augmenting dwindling stocks of local genetic resources (through expanding the types and varieties of millets grown); and arresting the decline in the production of minor millets in India. The results summarized above indicate clearly that in the locations in which the project operated, the project’s participatory approaches to promoting millets have moved the needle in the desired direction. At the same time, however, it is important to note that in both locations, but especially the Kolli Hills, the scale of millet production is small by any reasonable standard. Farm sizes are small, and the footprint of millets on those farms—both in terms of area and output—remain relatively small compared to other crops grown. Regarding contributions to income, the data are similarly quite clear that millets occupy a much less important position than other livelihood sources—both on-farm and off-farm—in both study areas. The composition of household earnings from agriculture production varies substantially from location to location—cash crops and earnings from off-farm agricultural wage labor dominate in Kolli Hills; paddy cultivation and agricultural wage labor dominate in Kundura. But the relative unimportance of millets’ contribution to household overall earnings is consistent across areas. And while substantial fractions of households had engaged in value-addition, in the both areas earnings from value-addition were similarly modest in 2018—although a few households in one Kundura do seem to be earning sizeable profits from value-added production. Finally, the evidence indicates quite clearly that rice remains, by far, the most important food consumed by surveyed households in both locations. Finger millet is the only minor millet consumed in the Kolli Hills, and its consumption has increased somewhat over time. In Kundura, 20 a small amount of little millet consumption was reported in some villages, but finger millet dominates there as well. Average consumption of both types of millet actually declined in Kundura. Meanwhile, in both locations average rice consumption increased over time. Interestingly, in 2018 households in both areas appear to have substituted rice purchased from PDS shops for rice that previously came from other sources (open-market purchases in Kundura, own-production in Kolli Hills). This bears implications for possible outcomes that might arise if and when millets were to become available in PDS shops in places like the study locations. Implications What then do these findings suggest for the future—in particular, a future in which substantial government procurement of millets (for provisioning PDS shops) were to take place. As has been noted, the inclusion of millets in the National Food Security Act represents a potentially huge boost to millet demand. So what might such an increase in demand mean for the sorts of limited-resource farmers that have been the beneficiaries of the Bioversity-MSSRF project? To begin with, it is almost certainly the case that in the foreseeable future government procurement would concentrate on finger millet exclusively. There are a few reasons for this. First, in nearly all areas, production of finger millet is substantially larger than that of other minor millets. This reflects, among other things, the fact that varietal improvement research for finger millet has a longer and more successful track record (compared to breeding efforts other minor millets). Second, processing of finger millet is dramatically less difficult and costly (other minor millets require de-husking, whereas finger millet does not). And finally, finger millet is generally consumed much more widely than other minor millets throughout India. As has been noted, the production of finger millets occurs at very small scales in the areas where the Bioversity-MSSRF project has taken place. But that is fairly representative: Millet production throughout India is stereotypically associated with small-scale production by limited resource farmers who market little if any of their millet output. Clearly, then, some sort of price premium would be required to incentivize those farmers to produce more marketable surpluses. Indeed, many interlocutors in India with whom I interacted over the course of this research argued that substantially increasing the minimum support price (MSP) for millets is an indispensable “requirement” for generating sufficient quantities of grain to meet government procurement targets.14 Note, however, that raising the MSP—or, as was done in Karnataka, paying a substantial premium over and above the national MSP—would benefit finger millet farmers, but only if they are selling millet. A difference in the level of market participation between the two study locations was evident. Predictably, more farmers sold finger millet in Kundura (where farm sizes are small, but much larger than the Kolli Hills). This carries implications for the distribution of benefits to farmers from boosting the price of finger millet, namely that more of those benefits would be felt in places like Kundura than in places like the Kolli Hills. 14 Corresponding to this point of view, subsidies that raised the price of finger millet 15% over the MSP were deemed essential by the Karnataka government in 2015 in order to generate sufficient marketed surpluses of finger millet (Rajshekar and Raju 2017). 21 Finally, it is important to recognize that offering a higher producer price for finger millet might produce some impacts that are inimical to core goals of the Bioversity-MSSRF project, particularly those related to biodiversity and in situ conservation of genetic resources. First, an increased price of finger millet would likely motivate existing millet farmers to substitute away from other cereals—including other minor millets that they are currently producing—in favor of more acreage allocated to finger millet. In support of this contention, note that the outcome analysis indicated that where little millet production has expanded in Kundura, it has been grown for commercial—not consumption—purposes. Putative price increases for finger millet would tend to reverse the decision-making calculus that led to the switch toward little millet. Second, and perhaps more importantly, higher prices for finger millet may be expected to generate production from new—i.e., non-traditional—growers. It is entirely likely that such new production would occur at larger scale, and on more favorable lands, than millet production in places like Kolli Hills and Kundura. Moreover, a reasonable case can be made that government procurement entities might be pre-disposed to favor such new millet producers, especially if those growers tended to produce larger quantities (that contribute to more efficient collection and bulking in procurement) and/or they are already served by government procurement entities for marketing other crops. In short, a potentially very large share of the pecuniary benefits of millet price subsides might not go to sorts of limited-resource farmers that represent Bioversity’s and MSSRF’s core constituencies. Third, if government pricing policies were altered to promote cultivation of finger millet for commercial purposes, the varietal characteristics that such farmers would value most highly may be expected to skew toward production characteristics related to yield and yield stability, and less by consumption and other characteristics, than is currently the case. Importantly, the dominance of these production characteristics in farmers’ varietal selection is commonly regarded as a key culprit underlying the dwindling genetic bases of major cereal crops like paddy and wheat. Finally, if finger millet becomes available to millet farmers at subsidized consumer prices via the PDS—as is the case for rice in the study villages that have been analyzed here—then some existing millet farmers may simply choose to re-allocate land previously devoted to millets to other, more remunerative crops. Indeed, one of the more interesting findings related to rice consumption in the study areas is that PDS-sourced rice appears to have substituted to a significant extent for own-produced rice in the Kolli Hills. A similar substitution response among millet farmers would clearly be at odds with the goal of expanding the scale millet production in such locations. 22 ACKNOWLEDGEMENT This work was implemented as part of the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS), which is carried out with support from the CGIAR Trust Fund and through bilateral funding agreements. For details please visit https://ccafs.cgiar.org/donors. The views expressed in this document cannot be taken to reflect the official opinions of these organizations . 23 REFERENCES Bala Ravi, S., S. Swain, D. Sengotuvel and N.R. Parida. 2010. “Promoting Nutritious Millets for Enhancing Income and Improved Nutrition: A Case Study from Tamil Nadu and Orissa,” in Bhag Mal, S. Padulosi and S. Bala Ravi, (eds.). 2010. Minor Millets in South Asia: Learnings from IFAD-NUS Project in India and Nepal Bioversity International, Maccarese, Rome, Italy and the M.S. Swaminathan Research Foundation, Chennai, India, 19-46. Bhag Mal, S. Padulosi and S. Bala Ravi, (eds.). 2010. Minor Millets in South Asia: Learnings from IFAD-NUS Project in India and Nepal Bioversity International, Maccarese, Rome, Italy and the M.S. Swaminathan Research Foundation, Chennai, India. Bioversity International. 2007. “Neglected No More: Achievements of the IFAD-NUS project (2001–2005) and Framework for Its Follow-up Initiative (2007–2009).” www.bioversityinternational.org/fileadmin/_migrated/uploads/tx_news/Neglected_no_more_ 1233.pdf Bioversity International. 2018. “Holistic Value Chain Approach.” Bioversity International, Rome. (www.nuscommunity.org/research/projects/ifad-nus-i-ii/). Down to Earth. 2015. “UN heritage Status for Odisha's Koraput Farming System.” www.downtoearth.org.in/news/un-heritage-status-for-odishas-koraput-farming-system--35627 Government of India. 2013. “The National Food Security Bill, 2013.” Bill No. 109-C of 2013, as passed by Lok Sabha on 26 August 2013. Grovermann, C., K. B. Umesh, S. Quiédeville, B.G. Kumar, S. Sakamma, and S. Moakes. 2018. “The Economic Reality of Underutilised Crops for Climate Resilience, Food Security and Nutrition: Assessing Finger Millet Productivity in India.” Agriculture 8(131): 1-12. King, I.O., N. Kumar, and S. Padulosi. 2015. “India: Community Seed Banks and Empowering Tribal Communities in the Kolli Hills,” in R. Vernooy, P. Shrestha, and B. Sthapit (eds.), Community Seed Banks Origins, Evolution and Prospects. Oxon, UK: Routledge. King, I.O., V.A. Nambi, and L. Nagarajan. 2008. “Integrated Approaches in Small Millets Conservation: A Case from Kolli Hills, India.” Acta Horticulture 806: 79-84 Nampoothiri, K. and P.K. Parida. 2014. “Biju Patnaik Medicinal Plants Garden and Research Centre.” Jeypore: MS Swaminathan Research Foundation. http://mssrfjeypore.org/save/wp- content/uploads/2015/07/BROCHURE_BPMPRGC_ENG.pdf. Notaro V., S. Padulosi, G. Galluzzia and I.O. King. 2017. “A Policy Analysis to Promote Conservation and Use of Small Millet Underutilized Species in India.” International Journal of Agricultural Sustainability 15(4): 393-405. 24 Padulosi, S., Bhag Mal, S. Bala Ravi, J. Gowda, K.T.K. Gowda, G. Shanthakumar, N. Yenagi and M. Dutta. 2009. “Food Security and Climate Change: Role of Plant Genetic Resources of Minor Millets.” Indian Journal of Plant Genetic Resources 22(1):1-16. Padulosi, S., Bhag Mal, I.O. King, and E. Gotor. 2015. “Minor Millets as a Central Element for Sustainably Enhanced Incomes, Empowerment, and Nutrition in Rural India.” Sustainability 7(7): 8904-8933. Rajshekar, S.C. and S. Raju. 2017. “Introduction of Millets in PDS: Lessons from Karnataka.” Unpublished LANSA report, MS Swaminathan Research Foundation, Chennai, India. Riley, K.W., Gupta, S.C., Seetharam, A., Mushonga, J.N. 1993. Advances in Small Millets. New Delhi: Oxford & IBH Publishing Company. Takeshima, H. and L. Nagarajan. 2012. “Minor Millets in Tamil Nadu, India: local Market Participation, On-Farm Diversity and Farmer Welfare.” Environment and Development Economics 17(5): 603-632. 25 Table 1. IFAD-NUS Grants, 2001-2018 Phase Years Title of Project Total Budget Target regions Target countries India Budget (USD) I 2001-2005 Enhancing the contribution of neglected and underutilized species to food security, and to incomes of the rural poor 1,410,000 USD North Africa, West and South Asia, Latin America Egypt, Yemen, Nepal, India, Peru, Bolivia, Ecuador 156,000 II 2007-2010 Empowering the rural poor by strengthening their identity, income opportunities and nutritional security through the improved use and marketing of neglected and underutilized species 1,400,000 USD West Asia and South Asia, Latin America Yemen, India, Peru, Bolivia 224,491 III 2011-2015 Reinforcing the resilience of poor rural communities in the face of food insecurity, poverty and climate change through on-farm conservation of local agrobiodiversity 975,000 USD South Asia and Latin America India, Bolivia, Nepal 259,997 IV 2013-2015 Improving smallholder farmers’ food and nutrition security through sustainable use and conservation of agrobiodiversity 500,000 USD South Asia and Latin America India, Bolivia, Nepal Incorporated with previous grant V (I) 2015-2016 Linking agrobiodiversity value chains, climate adaptation and nutrition: empowering the poor to manage risk 1,000,000 USD West and South Asia, Latin America Mali, Guatemala, India 136,000 V (II) 2016-2018 Linking agrobiodiversity `value chains, climate adaptation and nutrition: empowering the poor to manage risk 1,500,000 Euros West and South Asia, Latin America Mali, Guatemala, India 100,000 (2016) 71,000 (2017) Total budget provided to India $947,488 26 Table 2. Methods, Tools, and Key Achievements of the Bioversity-MSSRF Project Value Chain Portion Activities Participants Achievements Genetic Diversity Survey and collection of target crop diversity in situ (Focus Groups, discussions, base line surveys) MSSRF, Self Help Groups (SHGs), communities, farmers’ clubs, women’s assn’s Understanding of existing diversity on farms; access to wider genetic diversity by user groups Map out existing crop diversity, assess threats and vulnerability status (via Focus Group Discussions, base line surveys) MSSRF, SHGs, Communities, farmers’ clubs Distribution maps of crops and varieties; status of conservation and use; status of vulnerability and loss Introduction of varieties from ex situ gene banks MSSRF, SHGs, communities, ICRISAT, farmers’ clubs, women’s assn’s Wider basket of diversity options to farmers Participatory documentation of useful traits (Focus Group Discussions) MSSRF, SHGs, communities, farmers’ clubs, women’s assn’s, value chain actors Identification of traits useful for cultivation and use by households and markets Conservation ex situ and in situ (on farm), via creation/strengthening of SHGs and farmers’ clubs MSSRF, SHGs, farmers’ clubs, women’s assn’s, custodian farmers, community gene banks Conservation of crop diversity as key asset for building resilient livelihood systems Organize seed/food fairs for exchange of seeds and knowledge among users (via engaging SHGs and local authorities) Communities, local leaders, SHGs, farmers’ clubs, women’s assn’s, value chain actors Exchange of seeds and IK among users which also helps safeguard a community’s identity and food culture associated to nutritious millets Train trainers on new methods and tools in target communities MSSRF, SHGs, communities, women’s assn’s, Universities of Bangalore, Dharwad, and Uttarakhand, farmers’ clubs, women’s assn’s, value chain actors Strengthening skills of user groups, esp. women and vulnerable groups, in conservation practices and use enhancement of millets; training in seed collection, PVS, quality seed production, seed bank management, agronomic techniques, using tools to minimize drudgery, use of processing equipment, value addition methods, product development and marketing 27 Table 2 (cont’d) Value Chain Portion Activities Participants Achievements Selection and cultivation Participatory Variety Selection (PVS) MSSRF, communities, SHGs, Universities of Bangalore, Dharwad, and Uttarakhand, farmers’ clubs, women’s assn’s, value chain actors Selection of varieties (landraces or HYVs) showing best yields and best traits in terms of cultivation, processing, and food preparation Production of high quality seed of selected varieties MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Availability of high-quality seed of selected varieties to farmers Enhancement of agronomic practices (e.g., change of crop distances in planting, crop rotation, introduction of mechanical weeding, crop rotation, vermicomposting) MSSRF, communities, SHGs, Universities of Bangalore, Dharwad, and Uttarakhand, farmers’ clubs, women’s assn’s Reduction of drudgery in cultivation of millets Train trainers on new methods and tools in target communities MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Enhanced skills of farmers in cultivating millets. Skills in using machinery, product development, and marketing. Involvement in value addition and marketing of millets. New products reach markets. Private initiatives are booming. Harvest Improvement of harvest improvement methods MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Reduced drudgery and grain loss in post-harvest operations Value Addition De-huller machine (specify target species and level of efficacy); Pulverizer machine(specify target species and level of efficacy) MSSRF, communities, SHGs, Universities of Bangalore and Dharwad, farmers’ clubs, women’s assn’s Reduced /elimination of drudgery in processing millets Development of enhanced food preparations and novel food recipes MSSRF, communities, SHGs, Universities of Bangalore and Dharwad, farmers’ clubs, women’s assn’s, schools, and hospitals Easier food preparation that eliminates drudgery and produces food items more attractive to younger generations and modern lifestyles. Creation of SHGs to manage community-based processing centers MSSRF, communities, farmers’ clubs, women’s assn’s Enhanced capacity of communities in processing and using millets for domestic or market uses 28 Table 2 (cont’d) Value Chain Portion Activities Participants Achievements Marketing Strengthening marketing products via creation or strengthening of SHGs MSSRF, communities, farmers’ clubs, women’s assn’s Enhanced capacity of communities in marketing millets, market orientation, exposure visits to different streams of markets, provision of market information through information and communication Train trainers/SHGs on new methods, tools and practices in target communities (incl. packaging, branding and ways to obtain Government’s required certificates for marketing) MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Training in marketing and entrepreneurship; development of Kolli Hills Natural Foods Brand; Quality Control and Food Safety certification Building platforms and linkages among actors of value chains (incl. with restaurants) MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Linkages with mainstream markets; establishment of producer groups; product branding for products; product diversification; identification of niche markets; market assessments; interest free loan support programs; infrastructure development; management skills development Final Use Collection of traditional food recipes and dissemination of recipe books MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s Popularization of drudgery-free and more attractive food preparations for nutritious millets; contribution in safeguarding food culture and identity. Documented traditional recipes, promote some as marketable products Food festivals, public campaigns, involving school children, religious groups, and policy makers (via fact sheets, articles in the press, TV interviews, etc.) MSSRF, communities, SHGs, farmers’ clubs, women’s assn’s, schools, policy makers, religious groups Popularization of millets among younger generations and decision makers, as religious offerings in temples. Promotion of millets in school feeding programs MSSRF, Universities of Bangalore and Dharwad, schools, policy makers Awareness raised in decision makers over the importance of millets in school meal programs Lobbying at the Indian Parliament for the amendment of the National Food Security Bill Prof. MS Swaminathan, other MPs, government officials Awareness raised among key decision makers over the strategic role of millets in India for adaptation to climate change, nutrition security, and income generation; amendment of the bill through the inclusion of minor millets in the Public Distribution System Source: Adapted from Padulosi, et al. (2015). 29 Table 3. Seed Bank Transactions in Kolli Hills, 2001-2014 (kg) Year Little millet Foxtail millet Finger millet Kodo millet Proso millet All millets 2001 900 1,089 1,989 2002 1,389 987 24 2,400 2003 453 357 227 1,037 2004 508 435 240 1,183 2005 510 510 2006 95 279 172 43 27 616 2007 8 8 2008 17 17 2009 57 18 113 5 5 197 2010 306 55 519 20 11 911 2011 298 150 232 10 10 700 2012 176 63 246 5 7 497 2013 197 81 270 10 15 573 2014 432 353 562 15 35 1,397 Source: Padulosi, et al. 2015 30 Table 4a. Number of Households that Participated in Various Interventions, Kolli Hills Intervention Padasolai Puliampati Sempathuvalavu Thuvarapallam Tirupuli All Villages Ever? Still? Ever? Still? Ever? Still? Ever? Still? Ever? Still? Ever? Still? Seeds Varietal selection 28 22 10 6 13 12 14 2 12 3 77 45 Seedbank use 24 21 8 8 12 12 16 13 13 4 73 58 Seed quality training 28 - 10 - 14 - 16 - 11 - 79 - Crop Management  Thinning 34 26 13 9 19 15 19 18 19 16 104 84  Interculture/weeding 36 28 14 14 19 18 19 2 19 17 109 97  Interculture machine 9 9 8 8 7 7 11 10 8 7 43 41  Fertilizer after weeding 3 3 6 6 2 2 4 4 5 5 20 20 Value-Added training 4 4 1 12 4 25 Total no. of households 45 14 19 26 23 127 Table 4b. Number of Households that Participated in Various Interventions,, Kundura Intervention Banuguda Bhadraguda Chendiahilingan Dholijhiligan Heruguda Kaudiaguda All Villages Ever? Still? Ever? Still? Ever? Still? Ever? Still? Ever? Still? Ever? Still? Ever? Still? Seeds Varietal selection 9 3 8 4 15 13 14 2 5 4 12 5 63 31 Seedbank use 0 0 5 0 10 5 0 0 0 0 1 1 16 6 Seed quality training 8 - 6 - 15 - 11 - 5 - 13 - 58 - Crop Management  Thinning 6 5 8 8 4 4 2 2 5 5 3 3 28 27  Interculture/weeding 12 10 7 6 15 14 14 14 4 4 14 13 66 61  Interculture machine 4 3 7 6 14 8 5 5 4 4 7 4 41 30  Fertilizer after weeding 10 9 6 5 14 13 13 11 4 4 14 13 61 55 Value-Added training 3 4 1 0 4 1 12 Total no. of households 12 9 15 16 5 18 75 31 Table 5a. Mean Area Under Different Crops in Kolli Hills (acres) Padasolai Puliampati Sempathuvalavu Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Total Area 2.4 1.9 3.0 2.6 1.5 3.6 3.2 1.8 3.3 Paddy 0.28 0.14 0.31 0.30 0.20 1.83 0.20 0.11 0.35 All millets 0.08 0.05 0.21 0.15 0.04 0.44 0.18 0.12 0.24 - Finger 0.06 0.04 0.18 0.12 0.04 0.29 0.18 0.12 0.23 - Little 0.01 0.01 0.03 0.0 - 0.15 0.00 - - - Foxtail 0.00 - 0.01 - - - - - 0.01 Cash Crops - Tapioca 1.39 0.97 2.28 1.77 0.89 0.43 2.04 0.75 0.32 - Coffee - 0.05 0.95 - 0.06 0.98 - 0.05 - - Banana 0.05 0.00 0.14 0.04 - 0.09 0.04 0.04 0.21 - Pineapple 0.14 0.03 0.14 0 - 0.01 0.25 0.11 0.18 - Pepper - - 0.54 0 - 0.96 - - 0.26 N 45 14 19 Thuvarapallam Tirupuli All villages Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Total Area 2.3 1.6 3.0 2.9 2.0 3.1 2.6 1.8 3.1 Paddy 0.14 0.13 0.14 0.47 0.38 0.31 0.28 0.18 0.45 All millets 0.05 0.03 0.17 0.09 0.02 0.15 0.10 0.05 0.22 - Finger 0.03 0.03 0.11 0.09 0.02 0.15 0.08 0.05 0.18 - Little 0.02 - - - - - 0.01 0.00 0.03 - Foxtail - - 0.06 - - - 0.00 0.00 0.02 Cash Crops - Tapioca 0.99 0.62 0.50 1.78 1.15 0.27 1.52 0.89 1.06 - Coffee 0.04 0.04 0.27 - 0.10 0.04 0.01 0.06 0.51 - Banana 0.02 0.04 0.18 0.09 - 0.29 0.05 0.01 0.18 - Pineapple 0.42 0.29 0.52 0.17 - 0.37 0.20 0.09 0.25 - Pepper 0.00 0.00 0.04 0.0 - 0.09 0.00 0.00 0.36 N 26 23 127 32 Table 5b. Mean Area Under Different Crops in Kundura (Kundura) Banuguda Bhadraguda Chendiahilingan Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Total Area 4.6 4.6 4.8 2.6 4.5 2.7 5.1 7.0 7.7 Paddy 2.2 2.5 2.0 1.4 2.8 1.4 2.2 3.9 2.2 All millets 2.4 1.7 1.0 1.2 0.7 1.4 2.8 1.8 2.2 - Finger 2.4 1.7 0.6 1.2 0.7 0.9 2.8 1.8 1.1 - Little - - 0.4 - - 0.6 - - 1.1 Other crops* n/a n/a 1.7 n/a n/a 0.7 n/a n/a 0.9 N 12  9  15  Crop Dholijhiligan Heruguda Kaudiaguda All Villages 2007 2010 2018 2007 2010 2018 2007 2010 2018 2007 2010 2018 Total Area 6.2 6.1 4.9 6.4 6.4 5.2 6.7 6.9 4.3 5.4 6.1 5.1 Paddy 2.1 2.5 1.4 2.6 3.2 1.5 3.8 3.9 2.3 2.5 3.2 1.9 All millets 4.4 1.6 1.8 2.8 0.7 1.6 2.9 1.7 0.7 2.9 1.5 1.4 - Finger 4.4 1.6 0.9 2.8 0.7 1.2 2.9 1.7 0.5 2.9 1.5 0.8 - Little - - 0.9 - - 0.4 - - 0.1 0.0 0.0 0.6 Other crops* n/a n/a 0.9 n/a n/a 1.8 n/a n/a 0.4 n/a n/a 0.9 N 16 5 18 75 33 Table 6. Number of Farmers Growing Cereals and Average Area Cultivated Among Growers Kolli Hills (N=127) Kundura (N=75) Crop 2007 2010 2018 2007 2010 2018 Paddy - No. of growers 111 104 105 66 69 73 ‐ Avg. area (ac.) 0.32 0.22 0.55 2.83 3.49 1.90 Finger millet - No. of growers 42 47 76 74 74 71 ‐ Avg. area (ac.) 0.26 0.12 0.30 2.96 1.71 0.86 Little millet - No. of growers 5 1 5 0 0 40 ‐ Avg. area (ac.) 0.31 0.25 0.66 not grown not grown 1.12 Foxtail millet - No. of growers 0 0 6 0 0 0 ‐ Avg. area (ac.) not grown not grown 0.33 not grown not grown not grown     34 Table 7a. Mean Production of Different Crops in Kolli Hills (kg) Padasolai Puliampati Sempathuvalavu Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Paddy n/a 8 361 n/a 63 377 n/a 27 498 All millets 28 17 78 35 31 190 104 55 176 - Finger 24 16 72 26 31 109 102 55 91 - Little 4 1 6 9 <1 61 2 0 0 - Foxtail <1 0 <1 0 0 0 0 <1 1 - Other n/a 0 0 0 0 19 0 0 84 Cash crops - Tapioca n/a 75 3,169 n/a 78 1,411 n/a 58 2,663 - Coffee n/a 21 59 n/a 11 14 n/a 28 17 - Banana n/a 9 183 n/a 0 154 n/a 6 1,481 - Pineapple n/a 2 395 n/a 0 0 n/a 24 1,834 - Pepper n/a 2 219 n/a 2 38 n/a 4 58 N 45 14 19 Thuvarapallam Tirupuli All villages Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Paddy n/a 23 339 n/a 219 507 n/a 58 405 All millets 32 13 56 29 7 74 41 21 100 - Finger 23 12 28 28 6 50 37 21 66 - Little 9 0 0 0 0 0 5 <1 9 - Foxtail <1 1 1 0 1 0 <1 <1 1 - Other 0 0 27 0 0 24 <1 <1 24 Cash crops - Tapioca 0 47 11,052 0 8 1,513 <1 68 4,159 - Coffee 0 27 25 0 21 1 <1 22 30 - Banana 3 530 0 0 743 <1 5 540 3 - Pineapple 22 3,065 0 0 1,180 <1 9 1,236 22 - Pepper 3 20 0 5 41 <1 3 93 3 N 26 23 127 35 Table 7b. Mean Production of Different Cereals in Kundura Banuguda Bhadraguda Chendiahilingan Crop 2007 2010 2018 2007 2010 2018 2007 2010 2018 Paddy 408 1,758 1,867 139 2,300 1,233 423 3,323 3,437 All millets 127 740 122 92 605 1,234 154 784 556 - Finger 127 721 82 92 568 267 154 762 290 - Little 0 19 40 0 37 967 0 19 40 Maize 0 11 1,088 0 29 114 0 11 1,088 N 12 9 15 Crop Dholijhiligan Heruguda Kaudiaguda All Villages 2007 2010 2018 2007 2010 2018 2007 2010 2018 2007 2010 2018 Paddy 153 2,175 1,569 280 2,260 880 526 3,556 3,300 344 2,690 2,319 All millets 130 801 389 180 599 456 87 735 112 123 735 314 - Finger 130 631 167 180 590 400 87 713 70 123 681 182 - Little 0 171 222 0 9 56 0 22 42 0 54 132                         Maize 0 35 0 0 0 0 0 23 0 0 0 0 N 16 5 18 75 36 Table 8. Size distribution of finger millet production Location 2007 2010 2018 Kolli Hills < 100 kg 21 40 30 100-249 kg 18 6 44 250-499 kg 2 1 1 500-749 kg 1 0 1 750-999 kg 0 0 0 1,000-2,000 kg 0 0 0 3,000 kg 0 0 0 Kundura < 100 kg 16 1 26 100-249 kg 54 7 27 250-499 kg 4 18 10 500-749 kg 0 25 6 750-999 kg 0 11 2 1,000-2,000 kg 0 10 0 3,000 kg 0 2 0 Table 9. Average Millet Sales 2007 2010 2018 No. of Sellers/ Growers Mean quantity sold (kg) No. of Sellers/ Growers Mean quantity sold (kg) No. of Sellers/ Growers Mean quantity sold (kg) Kolli Hills - Finger millet 8/42 55 13/47 27 8/76 54 - Little Millet 0/1 - 0/5 - 2/5 250 Kundura - Finger millet 16/74 47.9 41/74 347.0 30/71 177.3 - Little Millet not grown not grown not grown not grown 37/40 237.0 37 Table 10a. Mean Monthly Cereal Consumption in Kolli Hills (kg) Padasolai Puliampati Sempathuvalavu Cereal 2007 2010 2018 2007 2010 2018 2007 2010 2018 Rice 43.3 52.8 46.3 44.2 48.4 70.6 46.0 51.5 49.5 - Own 18.5 30.3 13.3 14.1 21.4 14.6 20.3 30.9 10.4 - Purchased 24.8 22.5 7.8 30.1 27.1 28.9 25.7 20.6 22.8 - PDS 0 0 25.2 0 0 27 0 0 16.3 Finger millet 1.4 3.8 5.1 1.2 2.3 4.4 2.8 2.1 2.9 - Own 1.4 1.5 4.8 1.2 1.1 4.4 2.8 1.6 2.6 - Purchased 0.0 2.4 0.3 0 1.1 0 0 0.5 0.4 Little millet 0.1 0 0 0.4 0 0 0 0 0 - Own 0.1 0 0 0.4 0 0 0 0 0 - Purchased 0 0 0 0 0 0 0 0 0 Wheat 0 1.8 3.6 0 1 4.5 0 2 4.6 - Purchased 0 1.8 0 0 1 0 0 2 0 - PDS 0 0 3.6 0 0 4.5 0 0 4.6 N 45 14 19 Thuvarapallam Tirupuli All villages Cereal 2007 2010 2018 2007 2010 2018 2007 2010 2018 Rice 42.6 47.8 43.1 56.5 55.0 65.3 46.1 51.5 52.3 - Own 12.1 28.7 3.1 32.6 31.1 16.6 19.5 29.2 11.5 - Purchased 30.5 19.1 20.0 23.9 23.9 28.8 26.5 22.3 18.7 - PDS 0 0 20 0 0 19.9 0 0 22.0 Finger millet 1.0 1.8 2.7 1.9 2.2 2.2 1.6 2.7 3.7 - Own 1.0 0.7 0.9 1.8 1.3 0.4 1.6 1.3 2.8 - Purchased 0 1.1 1.8 0.1 1.0 1.7 0 1.4 0.8 Little millet (own) 0.4 0 0 0 0 0 0.2 0 0 Wheat 0 1.4 3.8 0.1 1.6 4.0 0.0 1.6 4.0 - Purchased 0 1.4 0 0.1 1.6 0.2 0 1.6 0 - PDS 0 0 3.8 0 0 3.7 0 0 3.9 N 26 23 127 38 Table 10b. Mean Monthly Cereal Consumption in Kundura (kg) Banuguda Bhadraguda Chendiahilingan Cereal 2007 2010 2018 2007 2010 2018 2007 2010 2018 Rice 13.3 39.2 55.9 15.1 43.9 57.8 20.3 40.7 58.5 - Own 13.3 25.0 28.0 13.9 21.1 27.8 19.3 24 46.8 - Purchased 0 14.2 0 1.2 22.8 0 1 16.7 0 - PDS 0 0 27.9 0 0 30 0 0 11.7 Finger millet 21.6 49.2 6.4 17.8 48.3 6.8 18.5 45.9 16.5 - Own 21.6 49.2 5.6 16.7 46.1 6.8 18.5 45.9 16.5 - Purchased 0 0 0.8 1.1 2.2 0 0 0 0 Maize* 0 22.1 0 0 21.7 0 0 12.7 0 Little millet* 0 0 0 0 0 0 0 0 0.6 N 12 9 15 Cereal Banuguda Bhadraguda Chendiahilingan All Villages 2007 2010 2018 2007 2010 2018 2007 2010 2018 2007 2010 2018 Rice 17.3 32.2 46.0 13.0 42.0 38.0 22.5 38.6 49.2 18.0 38.6 51.7 - Own 14.5 14.1 25.1 13.0 24.0 20 22.5 17.5 40.3 17.0 20.1 33.5 - Purchased 2.8 18.1 8.7 0 18.0 0 0 21.1 0 0.9 18.5 1.9 - PDS 0 0 12.2 0 0 18 0 0 8.9 0.0 0.0 16.3 Finger millet 20.8 42.5 6.0 21.8 40.0 6.0 22.5 47.8 5.4 20.6 46.0 8.1 - Own 20.8 42.5 6 21.8 40 6 22.5 47.8 5.1 20.4 45.8 7.9 - Purchased 0 0 0 0 0 0 0 0 0.3 0.1 0.3 0.2 Maize* 0 26.3 0 0 4 0 0 21.1 0 0.0 19.6 0.0 Little millet * 0 0 1.2 0 0 0 0 0 0 0.0 0.0 0.4 N 16 5 18 75 * All little millet and maize consumption was met from own-production 39 Table11a. Mean Annual Income from Various Sources in Kolli Hills (2010 Rupees) Source Padasolai Puliampati Sempathuvalavu 2007 2010 2018 2007 2010 2018 2007 2010 2018 Livestock 3,220 3,156 2,001 3,634 286 2,042 5,235 1,805 1,959 Ag. Wages 7,175 1,088 22,712 9,796 43 29,652 1,204 43 27,134 Nonfarm 5,817 1,047 1,627 0 154 10,231 18,807 13,148 0 Crops 15,033 23,203 10,163 25,860 29,942 5,333 21,787 22,120 25,753 - Paddy -100 1,442 750 103 2,193 468 -109 2,458 1,135 - Tapioca 14,452 19,374 2,905 26,279 26,718 4,892 21,483 16,084 4,921 - Millets 80 -13 364 173 280 202 -279 494 166 - Other* 681 2400 6,065 -695 751 -229 692 3,084 19,531 VA products 0 0 165 0 0 174 0 0 94 N 45 14 19 Source Thuvarapallam Tirupuli All villages 2007 2010 2018 2007 2010 2018 2007 2010 2018 Livestock 3,468 1,404 1,168 4,535 543 1,890 3,856 1,806 1,809 Ag. Wages 2,071 171 30,181 4,879 2,775 30,582 5,110 934 27,093 Nonfarm 17,538 9,437 3,012 13,482 22,967 259 10,907 8,446 2,368 Crops 24,525 20,586 30,056 19,624 25,602 21,958 20,012 23,682 18,172 - Paddy 36 1,487 202 -481 2,767 192 -120 1,926 563 - Tapioca 13,261 12,790 6,533 15,909 21,037 6,293 16,828 18,645 4,782 - Millets 78 33 -48 34 23 3 28 111 167 - Other* 11,176 6,276 23,369 4,162 1,775 15,470 3,276 3,001 12,660 VA products 0 0 149 0 0 188 0 0 156 N 26 23 127 * Includes peppers, pineapple, coffee, banana, primarily. a. These are wages received from off-farm sources. b. These combine both non-farm wage income and self-employment income. 40 Table 11b. Mean Annual Income from Various Sources in Kundura (2010 Rupees) Source Banuguda Bhadraguda Chendiahilingan 2007 2010 2018 2007 2010 2018 2007 2010 2018 Livestock 449 0 695 523 311 596 1,498 630 1,032 Ag. Wagesa 1,133 3,117 8,988 2,692 2,094 7,941 2,602 2,363 9,381 Nonfarmb 0 0 620 748 5,422 0 0 5,160 1,588 Crops n/a n/a 21,512 n/a n/a 19,178 n/a n/a 30,730 - Paddy 523 8,954 11,490 -527 13,128 6,419 728 16,038 22,573 - Millets -39 2,983 1,226 -124 3,217 4,308 77 3,747 5,765 - Other* n/a n/a 8,796 n/a n/a 8,451 n/a n/a 2,392 VA products 0 25 25 0 100 132 0 227 1,032 N 12 9 15 Source Dholijhiligan Heruguda Kaudiaguda All Villages 2007 2010 2018 2007 2010 2018 2007 2010 2018 2007 2010 2018 Livestock 930 125 637 269 300 0 2,393 0 174 1,225 210 567 Ag. Wagesa 2,229 2,594 4,906 1,211 3,300 2,144 2,019 3,739 7,131 2,066 2,893 7,168 Nonfarmb 0 3,281 0 0 0 0 0 10,222 1,476 90 4,836 771 Crops n/a n/a 23,044 n/a n/a 24,384 n/a n/a 20,682 n/a n/a 23,394 - Paddy -777 9,381 10,702 -603 7,470 4,610 1,937 16,194 19,009 425 12,602 14,276 - Millets 25 2,663 3,259 502 2,900 4,765 -522 4,411 1,052 -92 3,433 3,132 - Other* n/a n/a 9,083 n/a n/a 15,009 n/a n/a 621 n/a n/a 5,986 VA products 0 0 0 0 0 60 0 567 271 0 197 295 N 16 5 18 75 * “Other” includes grams, maize, cashews, pumpkin, mango, cowpeas and eucalyptus. Other crop income only available for 2018. a. These are wages received from off-farm sources. b. These combine both non-farm wage income and self-employment income. 41 Table 12a. Participation and Mean Annual Income from Value-added Products in Kolli Hills (in 2010 Rupees) Padasolai Puliampati Sempathuvalavu Thuvarapallam Tirupuli All Villages Mean income from VA products, 2018 621 306 596 774 309 473 No. of participants/Total sampled hh’s 12/45 8/14 3/19 5/26 14/23 42/127 Table 12b. Participation and Mean Annual Income from Value-added Products in Kundura (in 2010 Rupees) Banuguda Bhadraguda Chendiahilingan Dholijhiligan Heruguda Kaudiaguda All Villages Mean income from VA products, 2018 298 596 3,871 0 298 814 1,583 No. of participants/Total sampled hh’s 1/12 2/9 4/15 0/16 1/5 6/18 14/75 Mean Income from VA products, 2010 300 225 340 0 0 600 462 No. of participants/Total sampled hh’s 1/12 4/9 10/15 0/16 0/5 17/23 32/75 42 Figure 1. Interventions of note in study villages Conservation Consumption Cultivation Commercialization • Seed identification • Accessions • Seed banks • Diffusion of small-scale processing machinery • Nutritional training • Participatory varietal selection • Quality seed production • Crop management research • Value-added training • Larger-scale processing • Supply chain management • National Food Security Act 43 Figure 2. Timeline of Key Activities in Study Villages, 1999-2018 IFAD‐NUS  Phases 3 & 4 IFAD‐NUS  Phase 2 IFAD‐NUS  Phase 1 IFAD‐NUS  Phase 5 IPGRI,MSSRF  begin to  collaborate 1999 2001 2005 2007 2010 2011 2015 2018 Community Seed Banks Participatory varietal selection Crop management interventions (spacing, row planting, inter‐cropping, etc.) Promote processing technologies (milling, threshing, de‐hulling) Promote processing technologies Supply chains  Climate‐smart ag   Nutrition  44 ANNEX A: CONTACTS Dr. S. Bala Ravi, Former National Coordinator of IFAD-NUS Project Phases II and III, Hyderabad Dr. Mauricio Bellon, Coordinator of Agrobiodiversity Studies, CONABIO, Mexico City Shri. Krishna Byre Gowda, Minister of Agriculture, Government of Karnataka Dr. S.B. Dandin, Liaison Officer, Bioversity International, Bangalore Dr. B. Dayakar Rao, Principal Scientist, Indian Institute of Millets Research, Hyderabad Dr. Joanna Kane-Potaka, Director, Strategic Marketing and Communication, ICRISAT, Hyderabad Dr. Leslie Lipper, Executive Director, Independent Science and Partnership Council, CGIAR Dr.Thingnganing Longvah, Director, National Institute of Nutrition, Hyderabad Dr.Prem Mathur, Honorary Research Fellow, Bioversity International, New Delhi Dr. Latha Nagarajan, Senior Economist, International Fertilizer Development Corporation Dr. Krishna Kumar Nallur, Regional Representative, South and Central Asia, Bioversity International, New Delhi Dr. Stefano Padulosi, Senior Scientist, Bioversity International, Rome Dr. S.K. Pattanayak, Secretary, Department of Agriculture & Cooperation, Ministry of Agriculture and Farmers Welfare, New Delhi Dr.Prabhakar, Project Coordinator, AICRIP on Small Millets, University of Agricultural Sciences, Bangalore Dr. Melinda Smale, Professor, Department of Agricultural, Food, and Resource Economics, Michigan State University Dr. Subramanya, Professor & Head, Centre for Excellence, Millets Post Harvest Technology, University of Agricultural Sciences, Bangalore Mr. Vikram Sunlak, Chairman, Agro Industries, Coimbatore Prof. M.S. Swaminathan, Founder & Chairman, M S Swaminathan Research Foundation Dr. Madhura Swaminathan, Chairperson, M S Swaminathan Research Foundation Dr. M.N. Timma Gowda, Agronomist, AICRIP on Dryland Agriculture, University of Agricultural Sciences, Bangalore Dr. D. Vijaylakshmi, Professor of Nutrition, University of Agricultural Sciences, Bangalore Dr. Nirmala Yenagi, Professor of Food Sciences, University of Agricultural Sciences, Dharwad