Potato diversity at height: multiple dimensions of farmer-driven in-situ conservation in the Andes

In-situ conservationTwo types of in-situ conservation of crop genetic resources can be distinguished: farmer-driven andexternally driven. The first is subject of this thesis and refers to the persistence of potato genetic resourcesin areas where everyday practices of farmers maintain diversity on-farm. The second concerns the morerecent phenomenon of Research & Development (R&D) interventions which aim to support in-situconservation by farmers. In this study, farmer-driven in-situ conservation of the potato in the central Andesof Peru is investigated at different system levels from alleles, cultivars, and botanical species up to the levelof the landscape, as well as the interconnected seed and food systems. Dimensions of time and space areinferred upon by taking both annual and longer-term spatial patterns into account. Further, diversity islinked to selected farmer-based and external drivers.Objective and study areaThe overall objective of the study is to enhance our understanding of farmer-driven in-situ conservationand the context in which it takes place. The main field research was conducted between 2003 and 2006 ineight farmer communities following a north-south transect through the department of Huancavelica.Communities were selected on the basis of distribution and distance along the north-south transect,tradition of potato cultivation, ethnicity, and relative distance from major markets or cities. Depending onthe specific dimension of farmer-driven in-situ conservation investigated, a range of different methodsand tools were used. Chapter 1 provides a brief description of the study area and an overview of the researchmethods used.Species, cultivar and allelic diversityIn chapter 2 the species, morphological and molecular diversity of Andean potatoes in Huancavelica istreated at different scales of conservation: farmer family, community, geographically distanced, regional,in-situ and ex-situ subpopulations. The infraspecific diversity of in-situ collections was characterized usingmorphological descriptor lists and 18 polymorphic microsatellite markers (SSR). Botanical species weredetermined through ploidy counts in combination with morphological keys. Datasets were used fordescriptive statistics, (dis)similarity analysis, dendrogram construction, cophenetic analysis, matrixcorrelations calculations (Mantel tests), and Analysis of Molecular Variance (AMOVA).Results show that farmers in Huancavelica maintain high levels of species, morphological and moleculardiversity. All cultivated potato species with the exception of Solanum phureja and Solanum ajanhuiri provedto be present. Tetraploid species were most abundant followed by diploids, triploids and pentaploids. Atotal of 557 morphologically unique cultivars were identified based on the morphological characterizationof 2,481 accessions belonging to 38 in-situ collections. Genetic fingerprinting of 989 accessions belongingto 8 in-situ collections resulted in the identification of 406 genetically unique cultivars. AMOVA shows thatthe principal source of molecular variation is found within rather than between geographically distancedand farmer family subpopulations. No evidence of genetic erosion was found as the contemporary regionalin-situ population and a geographically restricted subset of CIP´s ex-situ core collection share 98.8% ofallelic diversity. Yet, in-situ collections contain numerous unique genotypes.Indigenous biosystematicsThe indigenous biosystematics of potatoes (folk taxonomy, folk descriptors and nomenclature) isinvestigated in chapter 3. The chapter includes an extensive literature review on the subject. Folk taxonomywas investigated with the use of grouping exercises with farmers, participant observation, and comparisonof farmer-recognized groups with formal classification based on morphological descriptors and 18polymorphic microsatellite markers (SSR). Analysis of the latter was based on (dis)similarity analysis,dendrogram construction and consequent levels of coherent clustering by folk taxonomic entity (folkspecific and varietal taxon). Ethnobotanical free and indicated listing exercises with farmers were used forresearch concerning folk descriptors. Descriptive statistics were used for analysis and interpretation.Nomenclature was investigated by applying nomenclature surveys, participant observation and basicethnolinguistic analysis of regional names.Folk taxonomy of the potato consists of no less than five ranks. The folk generic rank is composed ofthree taxa: Araq Papa (semi-wild / consumed), Papa Tarpuy (cultivated / consumed), and Atoq Papa (wild /not consumed). Folk specific taxa (= cultivar groups) and varietal taxa (= cultivars) within the generic taxonof Papa Tarpuy are abundant. Use categories and agroecological criteria are of little importance in the folktaxonomical system of the potato. Folk varietal taxa cluster well when using formal morphologicaldescriptors; folk specific taxa less so. A moderate concordance, albeit with considerable exceptions, existsbetween folk specific or varietal taxa and their genetic make-up as characterized with molecular markers(18 SSR microsatellites). The coherence of clustering in a dissimilarity tree varies for each folk specific orvarietal taxon considered. Farmers use 22 plant and 15 tuber folk descriptors with recognized characterstates in the Quechua language. Farmers are well able to recognize specific cultivars based on abovegroundplant parts only (without exposing tubers). Nomenclature is regionally consistent for common cultivars,while inconsistent for scarce cultivars. Primary cultivar names (nouns) generally refer to a folk specific taxonthrough predominant metaphorical reference to tuber shape. Secondary cultivar names (adjectives)predominantly provide direct reference to tuber color.Annual spatial patternsAnnual spatial management of the potato consists of cropping and labor calendars, field scattering practices,and genotype by environmental management. These three dimensions of agrobiodiversity managementare explored in chapter 4. A structured survey was conducted to investigate the potato cropping and laborcalendars. Participatory cartography resulted in the detailed mapping of 601 scattered potato fields,including their cultivar content, belonging to a total of 122 households. A genotype by environment (GxE)experiment employing 4 environments and 31 cultivars was conducted following an altitudinal transect.Data obtained was analyzed and interpreted using descriptive statistics, correlation analysis, GeographicalInformation Systems (GIS), Additive main Effects and Multiplicative Interaction (AMMI) analysis, and analysisof variance (ANOVA).The annual distribution of tasks and labor is primarily an adaptation to the single-season rain-fedcharacter and climate extremes of high-altitude agriculture. Three different footplough-based tillagesystems allow farmers to efficiently manage scarce labor availability for soil preparation. Native-floury, nativebitterand improved potato cultivars show considerable overlap concerning their altitudinal distributionpatterns. The notion that these cultivar categories occupy separate production spaces (so-called “altitudinalbelts”) is rejected as results show that differences between the altitudinal medians for areal distribution byaltitude of the different cultivar categories are modest (chapter 4). Field scattering is based on a combinedlogic which results in a patchy distribution of potato genetic diversity across the agricultural landscape.Depending on the community, farmers annually crop an average of 3.2 to 9.1 potato fields measuringbetween 660 to 1,576 m² and containing up to a hundred cultivars per field. However, neither field scatteringnor the management of high levels of diversity by farmers is a direct consequence of niche adaptation asmost cultivars are versatile (chapter 4). Rather, it is suggested that farmers conduct annual spatialmanagement by deploying combined tolerance and resistance traits imbedded in particular cultivarcombinations in order to confront the predominant biotic and abiotic stresses present in differentagroecologies. Andean farmers manage GxE adaptation for overall yield stability rather than fine-grainedenvironmental adaptation of native cultivars.Dimensions of land useThree specific dimensions of potato land use were researched in order to gain insights into possiblecontemporary changes affecting the in-situ conservation of potato genetic resources: land use tendencies,rotation designs and their intensity, and sectoral fallowing systems (chapter 5). The main research methodinvolved participatory cartography using printed poster-size high-resolution Quickbird satellite imagescombined with in-depth consultation through interviews and focus group meetings with members of thecommunities. A total of 4,343 fields and their 1995-2005 crop contents were mapped. The evolution over a30-year time-span (1975-2005) of traditional sectoral fallow systems (“diversity hotspots”) was documentedfor each community. Data was analyzed using descriptive statistics and Geographical Information Systems(GIS). Processes of change and adaptive innovation were documented by building case studies.Land use tendencies between 1995 and 2005 shows that the total cropping area dedicated to improvedcultivars has grown fast while the area dedicated to native-floury and native-bitter cultivars has remainedmore or less stable. Reduced fallow periods for existing fields and the gradual incorporating of high-altitudevirgin pasture lands sustain areal growth. Areal growth is particularly fast at extreme altitudes between3,900 and 4,350 m. However, fallow periods at these altitudes are still relatively long compared to fields atlower altitudes. Results show that fallowing rates increase by altitude for all cultivar categories, but tend tobe lowest for improved cultivars followed by native-floury and native-bitter cultivars. There is no evidenceof a straightforward replacement of one cultivar category by another resulting in the replacement and lossof infraspecific diversity. Inquiry into the dynamics of sectoral fallow systems over a 30 year period evidencesthe gradual disintegration and abandonment of these systems rich in cultivar diversity. They are replacedby more individualist management regimes based on household decision making. Nowadays, the spatialpatterning of potato genetic diversity within the agricultural landscape is increasingly characterized bypatchy distribution patterns rather than its concentration within a single communal sector. Where sectoralrotation designs survive local innovations have been adopted.Farmer seed systemsFarmer seed systems can be conceived as an overlay of crop genetic diversity determining its temporaland spatial patterning. Chapter 6 investigates the relation between selected farmer seed system components(storage, health and procurement) and infraspecific diversity of potato in Huancavelica. A sampling exercisewas carried out in farmer seed stores in order to gain insight into the internal organization of seed storesand how this relates to the management of infraspecific diversity. Virus infection rates were determined bytaking seed tuber samples of diverse cultivars from farmer’s storage facilities. ELISA tests were conducedfor APMoV, PLRV, PMTV, PVY and PVX. Seed procurement was investigated through a series of structuredsurveys focusing on household seed exchange, the role of regular markets and biodiversity seed fairs, andseed provision after severe regional frost. Data was analyzed and interpreted using descriptive statistics.Potato seed stores contain different seed lots, reflecting the rationales underlying management ofcultivar diversity at the field level and the overall structure of infraspecific diversity. Seed health of farmerconserved cultivar stocks in Huancavelica is affected by Diabrotica leaf beetle and contact transmittedviruses (APMoV, PVX) while aphid and powdery scab transmitted viruses (PMTV, PLRV, PVY) are of limitedimportance. During normal years without extreme events seed exchange of native-floury cultivars ispracticed by few households and characterized by a limited number of transactions involving smallquantities of seed of few cultivars covering relatively short distances. Native-bitter and uncommon nativeflourycultivars are rarely exchanged and generally reproduced year after year by the same householdsthat maintain them. High-altitude diversity-rich communities tend to be net seed exporters. However, thecapacity of the farmer seed system to annually widely supply and distribute infraspecific diversity is limited.Regular markets have a decentralized capacity to supply and widely distribute seed of a limited number ofwell-known cultivars. Frequencies of seed exchange at biodiversity seed fairs are low and involve smallquantities of a few uncommon cultivars. The resilience of the farmer seed system to cope with severe regionalseed stress is insufficient for households to be able to restore volumes and cultivar portfolios within ashort period of time.The potato-based food systemThe role of biodiverse potatoes within the human diet in Huancavelica is investigated in chapter 7. Analysisto determine the dry matter, gross energy, crude protein, iron (Fe) and zinc (Zn) content of 12 native-flourycultivars (fresh / boiled tuber samples) and 9 native-bitter cultivars (boiled unprocessed / boiled processedtuber samples) was conduced. Additionally, the nutritional composition of the native-floury cultivars wasdetermined after 3 and 5 months of storage under farmer conditions. A food intake study was conductedduring two contrasting periods of food availability (abundance versus scarcity) in order to quantify andcharacterize the contribution of the potato, different cultivar categories and other food sources to the dietof children between 6 and 36 months of age and their mothers. The specific method consisted of directmeasurement of food intake by weight during a 24 hour period for each household (77 households). Further,the overall nutritional status of 340 children aged between 4 and 16 years was determined. Selected culturalconnotations of the highland diet were investigated through participant and ethnographic observation,surveys, and workshops.Results show that several native-floury cultivars contain higher contents of specific nutrients (protein,iron) than those commonly reported as representative for native potato cultivars. This suggests thatinfraspecific diversity can make a valuable contribution to enhanced nutrition. Storage does not affect thenutritional quality of native-floury cultivars very significantly while traditional freeze-drying of native-bittercultivars considerably reduces protein and zinc content. The research shows that malnutrition inHuancavelica is primarily a consequence of micronutrient deficiency and secondarily of insufficient totalenergy coverage. The highland diet is heavily dependent on staple foods, particularly potato and barley,and generally short in vegetable, fruit, meat and milk intake. The potato contributes significantly to thenutritional balance and the recommended requirements for energy, protein, iron and zinc of women andchildren during periods of both food abundance and scarcity. Improved and native-floury cultivarscomplement each other as each category provides the bulk of potatoes consumed at different moments intime. The consumption of diverse potato cultivars is entangled with cultural constructions of meals andlocal perceptions of preference traits and quality. The potato itself, as a food item, is no socioeconomic classmarker. However, certain dishes or products and the overall cultivar diversity grown and used by a householdshape perceptions of relative wealth.Conclusions and implicationsChapter 8 highlights the main conclusions of the study and provides answers to the original researchquestions while taking the different system levels explored throughout the thesis into account. Selectedpriority areas of future research are identified and, where appropriate, links to other parts of the Andes aredrawn. Furthermore, the implications for externally driven R&D oriented in-situ conservation efforts seekingto support dynamic and ongoing farmer-driven conservation are discussed. It is argued that the scienceand practice of R&D oriented in-situ conservation lag behind the policy commitments to its implementationand that institutional learning from diverse projects already implemented throughout the Andes and thediffusion of key lessons is essential for the success of future interventions.