The process of crop domestication was initiated during the Neolithic period, around 12-10,000 thousand years ago. It involved the selection of traits suiting human needs from wild plant species – ancestors – imposing selective pressures that resulted on the one hand in bottlenecks that strongly reduced the genetic diversity (Gross and Olsen, 2010; Smýkal et al., 2018), but on the other hand in a further phenotypic diversification that allowed the adaptation of crops and emergence of new varieties in a complex and multi-stage process (Meyer and Purugganan, 2013). According to FAO (2009) the concept Plant Genetic Resources (PGR) includes any plant material of actual or potential value for food and agriculture. Thus, PGR encompass the cultivated plants – commercial varieties, modern cultivars, landraces, etc. – and also their ancestors and wild relatives. This diversity of resources for food and agriculture is part of the agrobiodiversity, that includes also other biological fundamental resources for the agriculture (Thrupp, 2000). The selection of traits and the search for highest yields reached a milestone in the so-called Green Revolution. The Green Revolution started in the decade of the 1950’s, when modern varieties started to be developed, mainly in wheat and rice but also in other crops, which allowed to increase production worldwide resulting in a gain in health and life expectancy (Evenson et al., 2003). However, it also brought homogeneity in crop genetic diversity, leading to genetic erosion, that is, the loss of genetic diversity of unique value (FAO, 1996). Genetic erosion has been identified as a major threat for food security (Esquinas-Alcázar, 2005; Khoury et al., 2014) mainly due to the adoption of monocropping and the genetic uniformity of cultivated varieties, which aggravates the threat (Conway and Barbie, 1988; Ehrlich et al., 1993).

Trends in global climate change are confirmed by the provisional report of the World Meteorological Organization (WMO, 2021) in which tendencies of strong increments in global mean temperatures are ratified by all analyzed datasets and extreme climate events – droughts, heat and cold waves, floods and disruption in precipitation patterns – are shown to be more frequent. Additionally, it also reports a peak in world risk of famine in 2020 which will probably be higher in 2021 due to humanitarian crises and the impact of climate change on cultivated plants and agricultural areas. The vulnerability of agriculture to climate change is known since decades ago. Whereas some countries at higher latitudes could be benefited in some aspects by climate change, some others, especially those at developing countries would be strongly affected by it (Rosenzweig and Parry, 1994; Kotschi, 2006). Nevertheless, most of the predictions on climate change impact on crops project world scenarios in which crop yields would be reduced in many world areas (Rosenzweig et al., 2014). Particularly in the Mediterranean area, climate change is hitting with force and all climatic indicators are worrying: just to cite some, mean temperatures have increased at a higher rate than global mean temperatures, precipitation has reduced over the years and droughts are more frequent (MedECC, 2019). These impacts are expected to increase food insecurity in the area, reducing yields in many crops and specially in the Southern areas (MedECC, 2019). Spain – the main studied area in this Thesis – is especially vulnerable to climate change (MTERD, 2021), and the impact in agricultural practices and crops cycles has already been assessed (García-Mozo et al., 2010; Vargas-Amelin and Pindado, 2014) mainly with negative effects, although it depends on the crops (Iglesias et al., 2000). To overcome such impacts and protect future food security, the development of new varieties, resilient and adapted to new environmental and social conditions becomes critical (Massawe et al., 2016) and actions to confront climate change should be addressed at the national or local level despite the global significance of climate change (Hasegawa et al., 2018).

Under this context of climate change and food insecurity, crop wild relatives (CWR) appear as a source of novel diversity worth using to enhance genetic diversity in crop species. CWR are plants intimately related to crop plant species and to which can transfer genetic material (Heywood et al., 2007; Maxted et al., 2012), which makes them specially valuable for breeding for traits related to adaptation. CWR have evolved under natural conditions, with non-directed selective pressures, and continue doing so, providing a magnificent source of adaptations to current environmental conditions (Brozynska et al., 2016). Although its use in breeding programs is becoming more apparent, the utilization of CWR in breeding programs is not widely adopted. Some of the main constrains to use them systematically are the lack of data on genotypic and phenotypic characterization and the linkage drag (introgression of undesired characteristics into the pure lines) (Dempewolf et al., 2017). However, in this review of the past and future usage of CWR, the authors provide several arguments that lead to an optimistic forthcoming increase of their use in pre-breeding programs, such as the improved documentation of materials, advances in biotechnological tools or the development of introgressed materials, just to cite some. In this line, it is worth mentioning the efforts made to use them in crops such as beet (Thurau et al., 2010), tomato (Willits et al., 2005; Bleeker et al., 2012), pea (Aryamanesh et al., 2012) or sunflower (Seiler et al., 2017) just to cite some among many others (see Ashraf (2010); Hunter et al. (2012); Redden (2015), Dempewolf et al. (2017) and references therein). Also encouraging is the interest in the development of new tools to enhance their use such as the introgressiomics approach (Prohens et al., 2017) that aims to the mass scale development of breeding materials from CWR, or the Predictive Characterization approach (Thormann et al., 2014a) that aims to facilitate the identification of useful germplasm or wild populations to be used with breeding purposes, applied in this PhD thesis and which will be further explained below.

Along with this growing interest and probably driven by it, CWR have been the subject of numerous research projects over the last two decades funded by the European Commission (Sixth and Seventh Framework Programmes and Horizon 2020) and has motivated the creation of international bodies aimed at their conservation (e.g., the CWR Specialist Group of the International Union for the Conservation of Nature (IUCN) or the Wild Species Conservation in Genetic Reserves Working Group of European Cooperative Programme for Plant Genetic Resources (ECPGR)). CWR have raised the attention and concern of international bodies such as the Food and Agricultural Organization of the United Nations (FAO), who included them as valuable resources for the sustainability of food security in the First and Second Plan of Action for the Conservation of PGR (FAO, 1996, 2010). Similarly, the first contract of the Convention on Biological Diversity (CBD) recognizes PGR as a component of biodiversity, advocating for their conservation and sustainable use, explicitly mentioning CWR as part of PGR (article 1 and Annex 1 of the CBD contract) (UN, 1992). Furthermore, CWR are included in the Strategic goal C, target 13 of the Strategic Plan for Biodiversity 2011-2020 and the Aichi Biodiversity Targets agreed by the CBD in 2010 (UN CBD, 2010). This increasing interest has culminated in Spain with the development and publication of the National Strategy for the Conservation of Crop Wild Relatives and Wild Food Plants (Molina et al. 2022, in press), supported by the Ministry of Agriculture, Fisheries and Food of Spain.

The effective conservation of CWR should start with a clear delimitation of species aimed for conservation, that will depend on the scope of the action (global, regional or national) (Heywood et al., 2007). The creation of inventories or checklists delimiting CWR subject to conservation actions can follow an up-down or floristic approach that identifies CWR from the flora of the area of interest or a bottom-up or monographic approach, that firstly set priority crops and then identify their CWR in the targeted area (Maxted et al., 2011; Magos Brehm et al., 2013). Subsequently, if the generated CWR list is large in terms of number of species and might difficult management and conservation planning, CWR checklists should be further prioritized. The incorporation of conservation assessment information (in situ or ex situ) or any other relevant knowledge about their use in breeding, endemicity or threat status (just to cite some), would finally turn the CWR checklist into a CWR inventory. Nowadays, we count on global, regional and national inventories and/or checklists that have been developed over the last two decades. Vincent et al. developed in 2013 the first web-enabled global inventory of CWR containing 1667 taxa related to 173 priority crops important at the global level and identified gaps in ex situ collections as well as areas for collecting and fulfilling those gaps. In the same vein, Castañeda-Álvarez et al. (2016) also developed a global inventory that contained 1076 taxa related to 81 crops and likewise identified gaps in ex situ collections and priority areas for collecting missing diversity in genebanks. We found also efforts for inventorying CWR at the regional level, retrieving a Catalogue of CWR for Europe and the Mediterranean area (Kell et al., 2008) that recognizes 25,687 taxa as CWR with potential or actual socio-economic value (80% of the flora of the area). In the Fertile Crescent region around 4% of the flora (835 species) were identified as CWR and further prioritized, generating an checklist of 220 CWR species (Zair et al., 2018). North and South Africa regions also have their corresponding CWR inventories with 5780 CWR taxa (Lala et al., 2017) and 1900 CWR taxa (Allen et al., 2019) respectively, the latter further prioritized into 745 taxa. At the national level, we can see notable advances in the generation of CWR inventories and checklists (please see Iriondo et al., 2016 and Labokas et al., 2018). Apart from Portugal which developed its inventory for CWR and wild harvested plants in 2008, the major advances in inventorying CWR have been achieved in the last decade. Since then, many countries have boosted the development of inventories or checklists of CWR. Europe has been very prolific in this sense and we find published inventories for Finland (Fitzgerald, 2013), Cyprus (Phillips et al., 2014), , England (Fielder et al., 2015a), Scotland (Fielder et al., 2015b), Norway (Phillips et al., 2016), the Czech Republic (Taylor et al., 2017), The Netherlands (van Treuren et al., 2017) and Italy (Ciancaleoni et al., 2021) who updated the work done by Landucci et al. (2014). African countries such as Benin (Idohou et al., 2013) and Zambia (Mwila et al., 2019) also have inventories, as well as the Asiatic Israel (Barazani et al., 2008), China (Kell et al., 2014) and Turkey (Tas et al., 2019). In America there are published CWR lists for Venezuela (Berlingeri and Crespo, 2012), the United States (Khoury et al., 2013) and Mexico (Contreras-Toledo et al., 2019).

However, the concern for the conservation of PGR is not new and emerged in the 1920s when the Russian botanist Nicolai Ivanovitch Vavilov was asked to gather germplasm for the maintenance and breeding of food and industrial crops (Bacchetta et al., 2008). He started thirty-years world expeditions aimed to the collection of PGR germplasm, including crop ancestors and their wild relatives, that were subsequently ex situ preserved, in the first germplasm bank created in the world (Brush, 1989). Soon after, in 1947 the United States also established seed collections of crop species and, since the 1970s, the ex situ conservation of PGR was spread around the world (Rajasekharan, 2015). The ex situ conservation has significant benefits, just to cite some, it gives the possibility of storage for the mid and long term with relatively low costs, grants rapid and easy access for the characterization of the stored germplasm and the rapid access to useful germplasm and allows the conservation of wide genetic diversity (Hawkes et al., 2000). It is also a way to back up the in situ diversity, allowing to reintroduce or reinforce species populations if needed. However, evolutionary processes take place in natural circumstances, when species are exposed to different conditions that exert selective pressures, so we can infer that the ex situ preserved material is an static picture of the genetic diversity in the moment of sampling. In fact, the ex situ conservation must be carefully designed as the genetic diversity of samples can be compromised (Hamilton, 1994). Regarding CWR, ex situ conservation has traditionally been the main pathway to conserve their diversity (Meilleur and Hodgkin, 2004).

Even so, the in situ conservation of CWR is pointed out as the first choice for preserving their diversity along with the habitats where they occur, in an efficient manner to conserve also the genetic diversity in continuous evolution, as reviewed by Meilleur and Hodgkin (2004). Diverse approaches and concepts for the implementation of such type of conservation targeting CWR have been developed especially in the last two decades (Heywood and Dulloo, 2005; Maxted and Kell, 2009; Maxted et al., 2015). Nowadays, the best strategy for the implementation of in situ conservation actions for CWR protection is through the establishment of Genetic Reserves. The genetic reserve concept (Maxted et al., 1997b) involves the active managing of the sites together with the design of strategies that would guarantee their long-term endurance (Kell et al., 2012). The rationale for the establishment of a genetic reserve is the dynamic conservation of the genetic diversity of the targeted CWR (Iriondo et al., 2012). In this sense, a genetic reserve must comply with a minimum of quality standards that involve the identification of suitable localities, with a sufficient extent to take the action, a concrete list of targeted taxa and populations and a management plan (Iriondo et al., 2012). Furthermore, the genetic reserve approach enables the conservation of multiple species in the same place, following a multispecies approach. A multi-species approach would – under the same costs – target multiple taxa which will increment the efficiency of the action. Ideally, the establishment of a genetic reserve must be accompanied by ex situ planning measures that ensure a safe back up of the genetic diversity of species protected in the reserve. That is the case, for example of the three recent created genetic reserves for CWR in the Biosphere Reserve Sierra del Rincón, in Madrid (Spain) (OAPN, 2020). For the establishment of these genetic reserves under the protection of the biosphere reserve, 15 CWR species were identified, populations inventoried and biotic and abiotically characterized and main risks for the populations survival identified (OAPN, 2021). Furthermore, the seeds from the 15 CWR species were sampled and sent to the germplasm bank ‘César Gómez Campo’, belonging to the Polytechnique University of Madrid for their long-term storage. This procedure of exploiting the synergies between in situ and ex situ techniques, is known as complementary conservation. The complementary conservation is endorsed by the CBD in its nineth article, that mandate the contracting countries to, “as far as possible and as appropriate”, complement in situ measures with ex situ actions (UN, 1992). Therefore, it is the recommended pathway to fully conserve CWR diversity and make it accessible to users (Maxted et al., 1997a; Fielder, 2015 and references therein). However, and despite the extensive literature and growing interest in CWR protection, their conservation has been neglected during decades, both ex situ (Hunter et al., 2012) and in situ (Maxted, 2003): still very few coordinated actions regarding their in situ conservation have been released (Álvarez-Muñiz et al., 2021) and their ex situ conservation is poorly addressed (Castañeda-Álvarez et al., 2016).

An effective network design for in situ conservation is dependent on the location of the components of biodiversity, including genetic diversity (Humphries et al., 1995). The genetic component should be considered when planning conservation actions (Thomassen et al., 2011) and the genetic diversity of the species systematically monitored (Schwartz et al., 2006) to develop conservation strategies that preserve the genetic diversity of the species and its distribution within and among populations. Accordingly, populations selected for in situ and ex situ collection should be representative of the overall genetic diversity of the species (Parra-Quijano et al., 2012a) and various populations of the same species might be needed to achieve such representation. The development of conservation genomics techniques, i.e., “use of new genomic techniques to solve problems in conservation biology”, is expected to generate great advances in conservation biology (Allendorf et al., 2010). The investment in developing new and affordable techniques to characterize species and populations genetics will allow to target species that otherwise would be relegated to lower priorities (i.e. non-threatened or endemic species). Still, nowadays, the molecular characterization of all populations of every target species to evaluate their genetic diversity and link it to conservation actions is currently not feasible. Consequently, only a select group of emblematic threatened species is genetically characterized for most or all of their populations, which leaves apart many important CWR.

The minimum standards for the conservation of the genetic diversity have been object of debate in the last decades, both for its in situ or ex situ preservation. For instance, Brown and Briggs (1991) suggested that the seeds of a minimum of five populations should be collected and ex situ preserved to ensure a good representation of the genetic diversity contained by endangered plant species. In fact, this minimum five populations has been used as reference targeting CWR conservation, and applied to in situ and ex situ conservation planning and assessment (Dulloo et al., 2008; Fielder et al., 2015a; Phillips et al., 2016). On the other hand, Whitlock et al. (2016) suggested that at least 35% of the populations of a given species should be in situ conserved to meet the CBD recommendation of preserving 70% of plant species genetic diversity (Objective II, target 9 of the Global Strategy for plant conservation targets 2011-2020) (UN CBD, 2010).

Some of the pioneer studies on the role of the environment in the genetic polymorphisms of species date back more than 40 years (Hedrick et al., 1976). Throughout this period, many studies have found that environmental pressures may result in the genetic differentiation of populations (Wang and Bradburd, 2014) and thus, in genetic diversity of adaptive value. Because different environments may induce diverse adaptation patterns in the genomes, one could infer that the diversity of environmental conditions could be a proxy for the estimation of genetic diversity of adaptive value. Different climatic, geographic, soil and other environmental variables have been used to construct ecoregional maps at different levels (global, regional, national) (Omernik, 1987; Olson et al., 2001; Abell et al., 2008; Sayre et al., 2020) that have been used for management or conservation purposes. However, they have been designed with little consideration for their possible link with species genetic diversity. Even so, Egan et al. (2018) found a good association between genetic variation and climate and landscape proxies as predictors for evolutionary processes in wild woodland strawberry and ecogeography was found to be extremely useful for the discrimination of wild Helianthus species (sunflower relatives) (Kantar et al., 2015) and wild Ipomoea species (sweetpotato relatives) (Khoury et al., 2015), finding patterns of adaptations, in the search for germplasm to be used with breeding purposes. On the contrary, Thormann et al. (2016) did not find such correlations when assessing wild barley germplasm in Jordan. Notwithstanding with this, if the aim is to protect adaptive variability, it is essential to conserve populations from all environments where the target species is found (Maxted et al., 2012).

Ecogeographical Land Characterization maps (ELC maps) classify a given territory according to its climatic, geophysical and geographical variables but, on the contrary to the above-mentioned maps, attempt to link possible adaptation patterns in plant species to differences in environments (Parra-Quijano et al., 2012a, 2012c). To do so, the variables that may be shaping a given species genome are taken into account. Is the researcher who, based on expert advice/knowledge or on objective processes (i.e., multivariate analyses), decides which variables are most likely to be affecting the plant species ecology and distribution. Main differences with former maps rely on the inclusion of a larger number of abiotic variables to define the territory and the obtention of more reticulated maps, avoiding large, continuous and homogenous regions or units (Parra-Quijano et al., 2020), which are less likely to be operating selective pressures. The generation of an ELC map allows the researcher to decide the resolution according to the territory to be classified and its extension, which help to better adjust environmental variables and targeted plant species searching for adaptive traits. The utilization of this approach for conservation in CWR started to be widely used in the late 2000 and 2010 decades using them as a proxy for genetic diversity as shown by many recent works (Table 1) dealing with CWR conservation and access, for example: i) the development of national strategies for CWR conservation (see Labokas et al. (2018) and references therein), ii) identification of areas for in situ conservation and iii) ex situ assessment and collection design.

One of the most important advantages of the ecogeographic approaches are the association of traits of adaptive value to populations of CWR in particular environments, ready to be used in breeding (Kantar et al., 2015; Khoury et al., 2015). Such relationships were lucidly suspected to be worth explorable in the search for germplasm for crop breeding by Mackay and Street (2004), who conceived the Focused Identification of Germplasm Strategy (FIGS). The FIGS methodology aimed to identify candidate landrace accessions for breeding abiotic tolerances and biotic resistances, with higher success probabilities than if randomly selected, based on the environmental characteristics of the collecting sites. The FIGS technique was further developed and conceptually supported by the successful correlations found in nordic barley landraces between ecogeographic data and morphological traits (Endresen, 2010) and biotic stresses (Endresen et al., 2011) and the agreement between drought tolerance scores given by tested models and classification given by genebank curators (Bari et al., 2016). FIGS was also experimentally tested in many works, identifying subsets of wheat accessions resistant to powdery mildew (Bhullar et al., 2010), association of resistance to wheat stem rust to certain environments (Bari et al., 2012) and successfully predicting drought tolerance traits in Vicia faba L. accessions (Khazaei et al., 2013). Based on the FIGS methodology, Thormann et al. (2014b, 2014a) proposed the Predictive Characterization technique for its application in CWR. Mirroring back the FIGS methodology, Predictive Characterization aims to the identification of populations and accessions of CWR with higher probabilities of containing target adaptive traits that if randomly selected. They argued that associations between environment and genetic diversity of adaptive value in CWR would be stronger than for cultivated varieties and landraces, due to their evolution in natural conditions. The Predictive Characterization technique can be applied by means of two methods: the ecogeographic filtering method or the calibration method (Figure 1). The ecogeographic filtering method is established under the basis of associations between environmental conditions and the adaptive characteristics that are likely to occur in plants inhabiting those particular environments. That is, if a population of a plant species is inhabiting an arid environment, this population will be more likely to possess adaptation to drought that other populations that do not occur in arid where such conditions do not concur. On the other hand, the calibration method, also based on that hypothesis, uses mathematical models to infer the probabilities of containing the desired trait, using as training data accessions of populations already characterized for the targeted trait, for example resistance to a plant disease.

This doctoral thesis was conceived at a time where no previous studies were available on the identification of priority crop wild relatives for Spain or their conservation status. Taking into account the relevance of the problem of the genetic erosion in crops in the context of global change, the main objective of this thesis was to contribute to setting the conceptual and methodological basis for an effective and sustainable conservation, access and use of crop wild relatives in Spain.

To achieve this goal, we posed five specific objectives: I) Delimitation and prioritization of CWR in Spain, II) Assessment of conservation status of priority CWR of Spain, III) Application and evaluation of experimental approaches that focus on multiple CWR species conservation, IV) Addressing the infraspecific genetic diversity component when designing CWR conservation and collecting plans, and V) Application and evaluation of predictive characterization techniques that identify CWR populations more likely to contain desired traits for plant breeding, facilitating their access and use. These specific goals contribute to addressing challenges and gaps identified at the different stages of the conservation-access-utilization continuum (Figure 2).

This thesis is structured in four interconnected chapters (Figure 3). The chapters have been designed to achieve the main objective of the thesis and structured following the typical configuration of research articles, to be published in international peer-reviewed journals (i.e.: Introduction, material and methods, results, discussion and conclusions). At the end of the thesis, we present a general discussion that evaluates the achievement of the general and specific objectives mentioned above.

In the first chapter we face the first specific objective, aiming to define a National Inventory of CWR for Spain, setting up a dynamic baseline of CWR species important for Spain in terms of socio-economic importance of crops, crossability to crops, endemicity or threat status. In this way, we asked which criteria should be applied for prioritization of CWR in Spain and what is the threat status and degree of endemicity of the prioritized CWR checklist. These CWR were also assessed in their abundance in the country (number of species and populations in each Autonomous Community), in their presence in national and international genebanks, and on whether they were under any legal protection in Spain tackling objective II.

In chapter two, targeting specific objective II, we aimed to gain deeper insight into the in situ and ex situ conservation status of priority CWR of Spain and identify CWR that require urgent conservation measures. To that purpose, those CWR simultaneously classified as easier to cross with crops, endemic to Spain and in any threat category of the IUCN were identified and further prioritized. To understand the needs for conservation, the National Catalogue of Protected Species of Spain, as well as the Catalogues of Protected Species for each Autonomous Communities were consulted. Those CWR that were threatened but lacked legal protection were pinpointed, finding 11 species that should be proposed into such Catalogues. Finally, herein prioritized CWR were evaluated in their current conservation status, checking their representation in protected areas and in national and international genebanks.

In the third chapter, we tackle objectives III and IV, and work towards an integrative approach for CWR conservation. We make use of existing concepts and tools that help targeting groups of species that tend to grow together (phytosociological associations) and include the genetic diversity component of adaptive value in the conservation assessments using ELC maps, answering whether phytosociological associations can be a useful approach to design and implement in situ conservation measures for CWR. To illustrate the method, we work with fodder and forage priority CWR of Spain and develop an ELC map to capture potential adaptations given by bioclimatic and other environmental conditions. Through the combination of the CWR associations with the ELC environments they inhabit, we create a new target conservation unit that includes, both the associations (species diversity) and the genetic diversity component. Subsequently, we include a gap analysis to check whether Natura 2000 network can properly conserve the target CWR conservation units and, finally, a complementarity analysis is performed to identify the priority sites for the establishment of genetic reserves for fodder and forage CWR.

Lastly, the fourth chapter addresses not only the inclusion of the genetic diversity component of the species into their populations, but also their potentiality to be used as gene donors in plant breeding processes. Thus, in this chapter we deal with objectives IV and V posing that it is possible to identify wild lentil populations tolerant to abiotic stresses, based on their ecological range and identify populations of wild lentils more probable to be tolerant to drought, soil salinity, waterlogging. We furthermore pose that ecogeographic variables associated with each population will allow to train mathematical models that will identify better wild lentil populations resistant to lentil rust than if randomly selected. To reach this goal, we apply Predictive Characterization methodologies that link environmental conditions to the probability of occurrence of the trait and thus, diminish the costs of field trials to confirm the existence of the desired trait against the random selection of populations. In order to increase the statistical robustness of the analyses and increment the chances of success, the geographic scope of this chapter was opened to Europe and Turkey and, consequently, the number of populations under analysis increased.

Crop quality and yields are known to be affected by climate change. Some models, using diverse climate change scenarios, crops and territories, predict a decrease of up to 40% in crop yields from 2010-2050 (Müller and Robertson 2014). At the same time, production risks are expected to increase, which may lead to an increase in world hunger (Tubiello and Fischer 2007). In Southern Europe, not only are temperatures expected to rise and precipitations expected to decrease under a climate change scenario, but the frequency of extreme events is also predicted to increase (Lotze-Campen 2011). In a conclusive meta-analysis, Challinor et al. (2014) report a highly significant negative impact of climate warming on crop yields that will be more pronounced in the second half of the century especially in tropical regions, as well as increases in yield variability that will further compromise food security. These hypothetical scenarios are already becoming apparent. In 2016 the greatest increase in global temperatures and variations in precipitation patterns were reported for the 137-year period of record (NOAA 2017). In Spain, the average temperature was 0.7°C higher than the mean of the reference series (1981-2010), placing 2016 among the warmest years on record (AEMET 2017).

In this context, farmers need to change their agricultural practices to effectively adapt to climate change, if they are to maintain and improve crop quality and yields. Such practices include adjusting planting times to avoid drought or heat stress and adopting new crop varieties, amongst others (Howden et al. 2007). However, these measures may not be sufficient (Turner and Meyer 2011), as modern cultivars may lack the ability to adapt to environmental change due to their narrow genetic base, resulting from selection applied in previous domestication and breeding processes (Stamp and Visser 2012).

Crop Wild Relatives (CWR) are wild plant species that are genetically related to crops (Heywood et al. 2007). As potential gene donors of desired traits for crops (Maxted et al. 2012), they have been successfully used in breeding for new traits and adaptations (Zamir 2001; Hajjar and Hodgkin 2007; Tester and Langridge 2010; Honnay et al. 2012). The wide range of adaptations found in CWR can be used as a genetic resource to mitigate the effects of climate change on crops, thereby helping to maintain and improve yields, and guarantee food security (Brozynska et al. 2016). The great value of CWR as a component of the Plant Genetic Resources for Food and Agriculture has been recognized in the United Nations Sustainable Development Goals (UN 2015), the Food and Agriculture Organization of the United Nations’ Second Global Plan of Action for Plant Genetic Resources for Food and Agriculture (FAO 2011) as well as the Convention of Biological Diversity (UN CBD 2010).

In global terms, CWR are seriously threatened by processes driven by human activities such as habitat fragmentation and loss, competition with invasive species, nitrogen depositions or changes in land uses, just like any other component of biological diversity (Ford-Lloyd et al. 2011; Heywood, 2011; Kell et al. 2012). The importance of CWR conservation and the best approaches to preserve these natural resources have been largely discussed (Heywood et al. 2007; Magos-Brehm et al. 2010; Maxted 2003; Pautasso 2012; Maxted et al. 2013) and over the past few years, several international projects have been implemented to conserve and manage CWR (see Online Resource 1). The in situ conservation of CWR in protected areas (Hunter and Heywood 2011), the establishment of genetic reserves (Pinheiro de Carvalho et al. 2012; Fielder et al. 2015a) and the identification of priorities and efficient sampling approaches for ex situ conservation (Khoury et al. 2015, García et al. 2017) are some of the procedures recently addressed for CWR conservation. In this context, the generation of CWR inventories is an essential first step in identifying the conservation needs of this group of species. Thus, listing and prioritizing existing CWR at the appropriate scale helps direct management efforts and underpins agrobiodiversity conservation. Two different approaches can be used to generate the inventories. The “crop list” approach uses a priority list of important crops to obtain their corresponding wild relatives. Alternatively, the “floristic” approach uses the flora of a territory as a starting point and matches it against a previously-existing catalogue of CWR species in the region or an exhaustive database of plants of economic use. To the best of our knowledge, seventeen CWR checklists have been published in the scientific literature: fifteen at the national level and two at the subnational level. These were generated for India (Arora and Nayar 1984), the United Kingdom (Maxted et al. 2007), Portugal (Magos-Brehm et al. 2008), Russia (Smekalova 2008), Israel (Barazani et al. 2008), Denmark (Bjørn et al. 2011), Venezuela (Berlingeri and Crespo 2012), Finland (Fitzgerald 2013), Benin (Idohou et al. 2013), the United States (Khoury et al. 2013), Italy (Panella et al. 2014), Cyprus (Phillips et al. 2014), China (Kell et al. 2014), the Czech Republic (Taylor et al. 2017), the Netherlands (van Treuren et al. 2017), England (Fielder et al. 2015a) and Scotland (Fielder et al. 2015b). Furthermore, a global inventory (Vincent et al. 2013), two European catalogues (Heywood and Zohary 1995; Kell et al. 2005) and a prioritized checklist of North Africa (Lala et al. 2017) have also been published. Similarly, some regional and country red lists of CWR have been generated (VMABCC-BIOVERSITY 2009; Bilz et al. 2011).

Governments and institutions dealing with wild and cultivated biodiversity conservation should take responsibility for the in situ and ex situ conservation of CWR. This is especially relevant for Spain, as it is one of the countries with the greatest number of plant species in Europe (7071 species, Aedo et al. 2013). In fact, it is the second country with the greatest number of CWR in Europe, hosting 26% of the Euro-Mediterranean CWR species (Kell et al. 2008). In addition, many of them are only found in this country, as the Iberian Peninsula is one of the two main centers of biodiversity in the Mediterranean Basin and presents a high number of endemics (Medail & Quezel 1999). The interest of the research community in CWR native to Spain is not new. Significant efforts have been made in the last few decades to explore, conserve and characterize wild gramineae (Soler et al. 1997), Brassica L. (Gomez-Campo et al. 2005), Vitis L. (De Andrés et al. 2012) and Medicago L. (Prosperi et al. 2006), among others. Although CWR conservation in Spain has been historically neglected by both the departments of wildlife conservation and agriculture at the national and autonomous community levels, its importance is becoming more widely recognized. Consequently, CWR have been included in the Spanish National Strategy of Plant Conservation (MAGRAMA 2014) and the Spanish Plant Genetic Resources Center has set some initiatives for the collection and preservation of CWR seed accessions. CWR germplasm is also stored in the César Gómez Campo genebank at the Polytechnic University of Madrid and the Agrifood Research and Technology Center of Aragón. REDBAG, the network of seedbanks associated with the Ibero- Macaronesian Association of Botanical Gardens, also preserves CWR seeds as part of their efforts to preserve threatened plant species. In parallel, native and exotic CWR have been actively used in breeding in Spain. For instance, Pico et al. (1999), Pérez de Castro et al. (2005), Caro et al. (2015) and Campos et al. (2017) worked on the development and evaluation of breeding tomato lines partially resistant to Tomato yellow leaf curl Sardinia virus and Tomato yellow leaf curl virus derived from Solanum chilense (Dunal) Reiche and S. peruvianum L.. Similarly, Martín-Sánchez et al. (2003) and Fernández-Martínez et al. (2000) used different wild accessions of Aegilops and Helianthus species stored in Spanish germplasm banks to deal with Hessian fly pests in wheat and Sunflower broomrape, respectively. So far, we have barely begun to explore the potential use of crop wild relatives from Spain. Some examples of how they are being used for breeding are found in Table 1. Taking all this into account, from a national strategy perspective, it is essential to list the main CWR taxa that occur in the country and prioritize the CWR species in need of active conservation measures.

The aim of this paper was to develop a checklist of CWR of importance in Spain and a prioritized list for the implementation of conservation plans. In this context, we asked: a) What criteria should be applied to prioritize CWR in Spain? b) What is the threat status of the prioritized CWR checklist at the national and European levels? c) Are these species under any legal protection in Spain? d) What are the levels of endemicity of the prioritized CWR checklist? e) How are their populations distributed and preserved ex situ in Spain?

The process involved in the generation of the prioritized CWR list is summarized in Figure 1.

Currently published CWR checklists and inventories provide a good background that exposes the regional and global importance of these species and the idiosyncrasy of each country in their development process. Our study fills an important gap in this area, as Spain has one of the largest and most diverse flora in the Euro-Mediterranean region (Médail and Quézel 1997; Molina-Venegas et al. 2015).

Existing CWR species, along with the rest of biodiversity components, should be conserved using strategies based on the establishment and management of protected areas and the sustainable use by humans of the rest of the territory, as well as species-specific approaches. In this context, the proper identification of priority CWR is essential. The generated Spanish Checklist of Crop Wild Relatives should be coordinately managed by the agriculture and environment departments of the public administration, and continuously revised in a participatory way to include species with real potential that meet the needs of the changing trends in agriculture and plant breeding.

However, the mere generation of a CWR checklist does not assure proper conservation. The conservation status of priority CWR should be properly assessed, and species-specific in situ and ex situ conservation actions should be implemented when needed. In this sense, the Spanish Checklist and Prioritized Checklist of CWR are already being used to assess the ex situ conservation of CWR in the collections of the Spanish Plant Genetic Resources Center (De la Rosa et al. 2013). Furthermore, they are also being used to plan new CWR seed collecting campaigns to improve their ex situ conservation status (García et al. 2015).