1. Background The global human population is expected to reach 9.3 billion by 2050, at a time when climate change is going to adversely affect crop production, challenging food security. Solving this problem could be difficult, as the highly productive cultivars currently used in agriculture have a narrow genetic base and, in many cases, lack the adaptation mechanisms to respond to environmental changes.
The Food and Agriculture Organization (FAO) has emphasized the relevance of plant genetic resources for food and agriculture (PGRFA) due to the important role they play to ensure global food and nutrition security. Among plant genetic resources, crop wild relatives (CWR) and landraces are key components, because they provide a wide genepool of potential gene donors, due to their specific adaptation to their local environmental conditions. However, despite their importance, their diversity is increasingly threatened and, thus, they must be safeguarded with the highest priority. In this sense, the Second Global Plan of Action for PGRFA proposed 18 priority activities to take in response to the lacks and needs identified by the Second Report on the State of the World’s PGRFA.
In the context of plant genetic resources, the analysis of ecogeographical data is a very useful tool due to its versatility and low cost, particularly in developing countries, where financial resources are scarce. The main applications of ecogeographical data on conservation of PGRFA have been the development of ecogeographical maps to describe the different potential adaptive scenarios of a species in a certain territory, the generation of core collections and the assessment of the representativeness of ex situ collections. Ecogeographical information has also been used to identify the most appropriate sites to collect wild germplasm and to characterize germplasm in a predictive way, by generating sets of landraces that have a higher probability than random subsets of possessing a given target trait.
2. Objectives The main objective of this thesis is to assess the applicability of ecogeographical data on conservation and use of PGRFA, and specifically CWR and landraces. We hypothesized that ecogeographical data may be useful in the conservation of CWR and landraces, and can enhance their utilization. To achieve this main objective, three interconnected specific objectives were set, based on some of the tasks proposed by the Second Global Plan of Action on PGRFA: (i) Development of a methodological approach for the systematic identification of the most appropriate areas for landraces on-farm conservation, (ii) Development of germplasm optimized collecting designs for CWR species, and (iii) Study of applicability of predictive characterization methods on CWR.
3. Methodology A multidisciplinary approach was used to comprehensively address the objectives exposed above. Firstly, tomato was selected to gain further insight into the potential of mono-specific approaches in the identification of the most appropriate areas (MAPA) for on-farm conservation of landraces in Spain (Chapter 1). Through the contribution of a panel of experts, the most important criteria in the assessment of the MAPA were identified. Then, a specific GIS layer for each important criterion was generated, and six different prioritization strategies were applied to finally assign an average priority value for on-farm conservation per county.
Secondly, and to develop germplasm optimized collecting designs for CWR species, data about conserved and non-conserved populations of Aegilops spp. were compiled, and ecogeographical variables that might be important for the distribution of these target taxa in Spain were identified (Chapter 2). Then, ecogeographic land characterization (ELC) maps for each species were produced to carry out an assessment of the ecogeographical representativeness of ex situ collections, with the final purpose of identifying priority ecogeographical gaps in these collections.
One additional step was taken in Chapter 3 to propose a new type of optimized collecting design to determine priority areas for collecting CWR to capture the maximum genetic diversity of adaptive value at a minimum cost, by adding the principle of complementarity to the methodology applied on Aegilops spp.. By using a set of 98 priority species related to cereal and legume crops used for food, and considering Spain as the area of study, ecogeographical gaps of the targeted taxa were input as taxon-ELC category combinations in a complementarity analysis. This analysis provided a set of 10x10 km complementary areas ranked by decreasing richness of ecogeographical gaps. Finally, the coverage of the top 10 complementary areas of ecogeographical gaps by the Sites of Community Importance (SCI) of the Natura 2000 network in Spain was checked, with the aim of identifying which ecogeographical gaps within these top 10 complementary areas were not covered by the Spanish SCI network and, thus, could be considered priority for germplasm collection.
And thirdly, two different approaches were utilized to study of applicability of predictive characterization methods on CWR. On the one hand, the predictive characterization filtering method was applied to search for the ecogeographic gaps in Spanish Aegilops spp. ex situ collections with higher probability of containing phenotypes with a high tolerance to drought and saline soils (Chapter 2). For this purpose, ecogeographic gaps were filtered using the Lang aridity index and topsoil salinity of their occurrence sites. On the other hand, the predictive characterization calibration method was utilized to model the expression of cyanogenesis in white clover at a global scale (Chapter 4). Data on genebank accessions and other population occurrences were divided into two subsets, one including accessions that had been evaluated for this trait, and the other with those that had not. Then, some ecogeographical variables selected by a panel of experts as potentially influential in cyanogenic response were utilized to construct several alternative models in which cyanogenesis of the evaluated set was used as the dependent variable, and the ecogeographic variables were used as explanatory variables. After that, the model with the best predictive power was run to predict the levels of cyanogenesis in the non-evaluated set, to identify populations of white clover that may have desirable low cyanogenic levels. In addition, this model was projected on the area of study, to identify the areas where environmental conditions would favor acyanogenesis. And finally, to validate the methodology applied, 18 accessions from the top 40 USDA accessions in the ranking of probability of acyanogenesis were evaluated to determine the percentage of plants that were cyanogenic.
4. Results 4.1. Chapter 1 Among a wide set of factors, the panel of experts determined that the most important criteria in the assessment of the MAPA were, in decreasing order of relevance: phenotypic diversity of tomato landraces, number of tomato landraces, number of species (other than tomato) with landraces, environmental conditions that favour the cultivation of tomato landraces with potential adaptation to high temperatures and saline soils, vulnerability of tomato landraces to climate change, ecogeographical diversity, number of non-govermental organizations involved in promotion and conservation of agrobiodiversity, percentage of population > 65 years old, percentage of the area within the action range of a “Km0 Restaurant” of “Slow Food”, percentage of the area dedicated to tomato cultivation, permission of tomato production with protected geographical indication, percentage of disadvantaged area, distance to an urban center of more than 10,000 inhabitants, number of landraces included in the Project “Ark of Taste” of “Slow Food”, average area of the horticultural properties, and average population age.
The strategy selected to obtain a final priority value per cell in the area of study was strategy 16W, that involved the consideration of the 16 GIS important criteria’ layers, taking into account the weighted average importance value calculated per criterion from experts’ responses. The priority counties, in accordance with the selected strategy, were Rincón de Ademuz (in the Valencian Community); Azuaga, Almendralejo and Llerena (in Extremadura); Goierri, Urola Costa, Bajo Deba and Alto Deba (in the Basque Country); and Pallars Jussà (in Catalonia).
4.2. Chapter 2 Isothermality and altitude were selected, in four of the five Aegilops species, as the most relevant variables to describe different adaptive scenarios of each species.
Among the 2614 populations reported by external sources (i.e., populations not conserved ex situ), 2571 were identified as spatial gaps, and Ae. geniculata Roth was identified as the species with the largest number of spatial gaps. Within spatial gaps, 393 populations were identified as priority ecogeographical gaps in ex situ collections of Aegilops in the Spanish Network of Plant Genetic Resources, given that they occur in ecogeographical categories that are not represented by the corresponding species in the Spanish Network. Ae. biuncialis Vis. was the species whose ex situ ecogeographical representativeness needed the most improvement, as 80% of the available external sources were identified as high priority gaps. On the contrary, only 10% of the analyzed externalsources of Ae. geniculata and Ae. triuncialis L., were identified as priority ecogeographical gaps.
Among the 393 populations identified as priority ecogeographical gaps, 223 populations inhabited sites with a Lang index value < 40, and thus were potentially adapted to arid environments. Twenty percent of these 223 populations with the highest values of topsoil salinity for each species constituted the predictive characterization subset of Aegilops populations of potential interest, due to their potential tolerance to drought and saline soils.
Aegilops geniculata, Ae. neglecta Req. ex Bertol. and Ae. triuncialis were the species that contributed with more populations.
4.3. Chapter 3 The most relevant variables to describe different adaptive scenarios for each species were, with regard to the bioclimatic component, isothermality and October precipitation (in 15 and 10 taxa, respectively). Topsoil reference bulk density and topsoil calcium sulfate (gypsum) content were the most important variables (in 33 taxa) in the edaphic component.
Among the 37,426 populations of the target taxa reported by external sources, 27,907 corresponded to spatial gaps. Hordeum murinum L. and Vicia sativa L. had the largest number of spatial gaps (3839 and 2530, respectively), whereas Vicia leucantha Biv. had the smallest (1). Among the 27,907 spatial gaps, 12,449 (45%) populations corresponded to ecogeographical gaps. Hordeum murinum and Vicia tenuifolia Roth had the largest number of ecogeographical gaps (1018 and 868, respectively), whereas Lathyrus sylvestris L., Lens ervoides (Brign.) Grande and Vicia leucantha had the smallest (1).
All top 10 selected complementary 10x10 km areas with the highest richness of ecogeographical gaps were partially or totally included in the protected areas of the Spanish SCI network. All the ecogeographical gaps present in Cells 2 and 3 in the ranking were covered by the SCI, whereas no ecogeographical gaps present in Cells 5 and 9 in the ranking were covered by the SCI. The coverage analysis found that 43% of the ecogeographical gaps identified (belonging to 514 different taxon-ELC category combinations), were included by the Spanish SCI network.
The collection of germplasm from the top 10 selected complementary areas would significantly improve the ecogeographic representativeness of the targeted CWR taxa in the Spanish Network of Plant Genetic Resources, given that it would allow the capture of 59 of the 88 target taxa and 31% of the 683 different taxa-ELC category combinations identified among the ecogeographical gaps.
4.4. Chapter 4 Based on the responses of the expert survey, five potentially influential ecogeographical variables to explain cyanogenic response were identified: solar radiation, annual mean temperature, annual precipitation, elevation and soil pH. GLM method yielded the best fit, and the most influential variable on the model was annual mean temperature, followed by annual precipitation and altitude. By projecting the selected model on the non-evaluated set, 470 populations of white clover with higher probability of being acyanogenic were identified. And by projecting the model on the area of study, Sweden, Bulgaria, Poland and the Netherlands appeared as new countries most likely to have acyanogenic white clover. Interestingly, of the 18 evaluated accessions, 17 were completely acyanogenic, whereas the other accession, which had the lowest probability of being acyanogenic of the evaluated sample, had 95% of acyanogenic plants.
5. Conclusions 1. A mono-specific methodology based on expert knowledge that combines a set of agroecological, ecogeographical, demographical and socioeconomical factors, is a useful approach to identify the most appropriate counties to focus on-farm conservation efforts on landraces.
2. The most important criteria for the assessment of the most appropriate areas to focus on on-farm conservation of tomato landraces in Spain, according to the experts’ knowledge, are related to agrobiodiversity and to ecogeographical parameters, followed by demographic and agronomic criteria.
3. The priority counties to focus on-farm conservation efforts on tomato landraces in Spain are widely distributed in rural areas across the Iberian Peninsula, mainly in the Valencian Community, Extremadura, the Basque Country and Catalonia.
4. A new type of optimized collecting design is presented to determine priority areas for collecting the genetic diversity of CWR. The new approach is based on the principles of complementarity and ecogeographical representativeness of ex situ collections, using ecogeographical variation as a proxy for genetic diversity of adaptive value. It aims to offer a collecting strategy that captures the maximum genetic diversity of adaptive value at a minimum cost 5. Almost half of the ecogeographical gaps of the Spanish ex situ collections of 98 Spanish priority CWR species related to cereal and legume crops occur within the limits of a protected area of the Spanish SCI network. To guarantee proper conservation of these resources, collecting of germplasm must be coupled with active in situ conservation measures, involving the establishment of genetic reserves.
6. The assessment of priority ecogeographical gaps showed a poor ecogeographic representativeness of the Spanish ex situ collections of Aegilops spp. A germplasm collection following an optimized collecting design would increase between 27% and 70% the ecogeographic representativeness of the Spanish ex situ collections of Aegilops spp.
7. Predictive characterization through the filtering method identified, in the western edge of their geographical distribution, 45 wild populations of Aegilops spp. that were likely to be the most tolerant to drought and saline soils. Aegilops geniculata, Ae. neglecta and Ae.
triuncialis were the species that contributed with more populations.
8. Predictive characterization through the calibration method proved to be a valid approach to successfully identify acyanogenic populations of white clover. The variables chosen by the selected model as influencing the acyanogenic status (annual mean temperature, annual precipitation and altitude) are congruent with previous scientific knowledge on the environmental factors that condition acyanogenesis.
9. The projection, at a global scale, of the selected model that better explains the relationship between the expression of cyanogenesis and the influencing ecogeographical variables, expanded the knowledge of areas previously rated as highly acyanogenic to Sweden, Bulgaria, Poland and the Netherlands.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados