Latin America and the Caribbean are regions rich in biodiversity, hosting nearly 60% of the world's biological diversity (UNEP-WCMC, 2016). Within this region, Mesoamerica is recognized as one of the main centers of origin, diversification, domestication, and biological plant diversity of various species of agricultural interest and animal consumption (Fortuny-Fernández et al., 2017). The complex evolutionary history, phylogenetics, geology, biogeography, and climatic variability are some factors that enhance the diversity in this area (UNEP-WCMC, 2016). This condition is essential to ensure food, socioeconomic, and cultural sovereignty for sustainable development and offers a large number of ecosystem services (FAO et al., 2019).

In this sense, tomato (Solanum lycopersicum L.) is one of the most cultivated vegetables due to its wide distribution and environmental adaptation in warm, subtropical, and tropical regions with nutritional and commercial importance worldwide (Peralta et al., 2008; Ramírez-Ojeda et al., 2021a). Regarding the place of origin and diversification of tomato, Peru is considered the center of origin with two transitions that involve tomato diversification process; the first one in South America, from wild species S. pimpinellifolium L. to a partially domesticated species S. lycopersicum L. var. cerasiforme (SLC); the second transition occurred in Mesoamerica from SLC to the completely domesticated species S. lycopersicum L. var. lycopersicum. However, new findings indicate that the origin of SLC may be prior to its domestication since many typical characteristics of tomatoes grown in South America come from this species; SLC is subsequently considered to have been lost or declined once the partially domesticated forms extended to the north (Razifard et al., 2020).

Wild species related to cultivated tomatoes are essential genetic resources in breeding programs focused on food security to face future challenges. Therefore, it is of strategic importance to study the climatic and edaphic factors that help to understand their current distribution patterns, as well as to establish the best indicators predicting possible effects of climate change and natural or anthropic environmental alterations. This is why it is necessary to undertake national and regional strategies for the conservation and use of cultivated and wild tomato genetic resources (Sandoval-Ceballos et al., 2021).

Based on an integrative taxonomy, which includes multiple evidences, the classification of wild tomatoes and their wild relatives was proposed: Solanum section Lycopersicon (Mill.) Wettst. comprises cultivated tomato (S. lycopersicum L.) and 12 wild tomato species: S. arcanum Peralta, S. cheesmaniae (L. Riley) Fosberg, S. chilense Dunal, S. chmielewskii (C. M. Rick, Kesicki, Fobes, and M. Holle), D. M. Spooner, G. J. Anderson, and R.K. Jansen, S. corneliomulleri J. F. Macbride, S. galapagense S. C. Darwin and Peralta, S. habrochaites S. Knapp and D. M. Spooner, S. huaylasense Peralta, S. neorickii D. M. Spooner, G. J. Anderson and R. K. Jansen, S. pennellii Correll, S. peruvianum L., and S. pimpinellifolium L. Four phylogenetically related Solanum species are also considered in the present study: S. juglandifolium Dunal, S. ochranthum Dunal (Solanum sect. Juglandifolia (Rydb.) A. Child), S. lycopersicoides Dunal, and S. sitiens I. M. Johnston (Solanum section Lycopersicoides (A. Child) Peralta) (Peralta et al., 2008; Causse et al., 2016; Tropicos.org, 2021).

Evidence of phylogenetic relationships of these species have been studied in detail by Peralta et al. (2008), who have proposed a six-group classification of wild tomatoes and phylogenetically closely related species: Section Lycopersicon: “Lycopersicon group” (S. pimpinellifolium, S. cheesmaniae, and S. galapagense), “Arcanum group” (S. arcanum, S. chmielewskii, and S. neorickii), “Eriopersicon group” (S. habrochaites, S. huaylasense, S. corneliomulleri, S. peruvianum, and S. chilense), “Neolycopersicon group” (S. pennellii); and two outgroups: Section Juglandifolia (S. juglandifolium and S. ochranthum) and Section Lycopersicoides (S. lycopersicoides and S. sitiens). This classification has been verified by molecular, genomic, and transcriptomic evidence of wild tomatoes (Rodríguez et al., 2009; Aflitos et al., 2014; Pease et al., 2016) and recently has been used for ecogeographic studies with satisfactory results (Ramírez-Ojeda et al., 2021a).

By considering plant genetic resources as the biological foundation for maintaining and improving crop productivity (Kantar et al., 2015), wild tomato species constitute an important gene pool due to the presence of genes with tolerance and resistance to biotic and abiotic factors (Arellano-Rodríguez et al., 2013; Cervantes-Moreno et al., 2014; Nosenko et al., 2016; Razali et al., 2018; Dinh et al., 2019) with potential use for breeding programs. Additionally, several questions arise about these gene pools, such as current distribution, population dynamics in situ or ex situ, and how are they used directly or as sources of genes to generate new varieties that respond to current and future basic problems of tomato cultivation (for example, climate change, diseases, pests), including the contribution of genes capable of conferring a greater nutritional–nutraceutical quality to new varieties (Chávez-Servia et al., 2011; Hernández-Bautista et al., 2014).

Identification of variables that derive in adaptation and speciation processes requires a large amount of field data of significant variables in natural populations. Recent developments and the use of remote sensing technologies, as well as a great availability of environmental information derived from Geographic information systems (GIS), have made it possible to identify patterns of species environmental variations at different scales (Nakazato et al., 2010). These tools and the availability of databases, with passport information of specimens collected in natural areas, allow for verification of the presence of species in a geographic range, as well as possible ecological descriptors, that is, to describe in detail the environmental conditions associated with the distribution of natural populations (Nakazato et al., 2010; Sánchez-González et al., 2018; Vilchez et al., 2019; Ministerio del Ambiente, 2020; Ramírez-Ojeda et al., 2021a).

One way to identify the adaptive ranges and most relevant variables that determine species distribution of valuable genetic resources is through ecogeographic studies, focusing on collection, conservation, characterization, documentation, and use of these resources (Parra-Quinajo et al., 2012; Pease et al., 2016), with the purpose of describing and explaining spatial patterns and processes involved in biodiversity distribution through time and space (Martiny et al., 2006; Tofalo et al., 2013; Délices et al., 2019). Ecogeographic studies of plant genetic resources allow the identification of the adaptive ranges of the species and the most relevant environmental variables that define their distribution (Parra-Quinajo et al., 2012; Ramírez-Ojeda et al., 2021a). Through ecogeographic studies, it is also possible to predict the environmental characteristics of the accession sites (Steiner and Greene, 1996) from ecological descriptors obtained through GIS tools using the geographical location and environmental variables (Lobo-Burle et al., 2013; Sánchez-González et al., 2018; Ramírez-Ojeda et al., 2021a, 2021b).

Currently, several information sources about geographical distribution of tomato species can be found in public databases (GBIF, 2021; Solanaceae Source, 2021; TGRC, 2021), conservation programs and gene banks (Córdoba-Téllez and Molia-Moreno, 2006; Florido et al., 2009; Magallanes-López et al., 2020), and genetic resources baseline studies (Ministerio del Ambiente, 2020), as well as some studies on geographic distribution patterns and ecological and climatic descriptors of wild tomato species (Peralta et al., 2008; Chetelat et al., 2009; Nakazato et al., 2010; Grandillo et al., 2011; Gonzá lez et al., 2013; Vilchez et al., 2019; Ramírez-Ojeda et al., 2021a). However, information regarding edaphic conditions of the sites where these species are located is limited or unknown (Balaguera-López et al., 2009).

Soil, a finite and nonrenewable natural resource, is of great importance in a large number of environmental services such as food and biomass production, climate regulation, carbon fixation, water storage and filtration, biogeochemical cycles, biodiversity reserve, and human physical and cultural environment (Burbano-Orjuela, 2016). Therefore, when considering edaphic together with climatic characteristics, it allows having a better understanding of the ecological and distribution patterns of the species.

Due to the limited edaphic information available regarding optimal characteristics for development of wild tomato species, the aim of the present work was to study ecological descriptors associated with soil characteristics and their relationship and the statistical association with climatic variables. Likewise, it was also analyzed whether the classification of wild tomatoes is related to the edaphoclimatic descriptors and supports the proposed groups of species.

This research provides a relevant ecogeographic characterization to understand the distribution patterns of wild species that complement the phenotypic and genetic information. Characterization of genetic resources through environmental characteristics associated with accession areas and use of GIS tools allows the identification of adaptive ranges and most relevant environmental factors affecting species distribution and ecological adaptation (Parra-Quinajo et al., 2012).

Likewise, through GIS and georeferenced information of species locations, it is possible to quantify geographical distances and distribution patterns of germplasm accession sites. From this perspective, it is likely to determine specific environmental conditions in which wild species and local varieties of crops have acquired their adaptive characters (Hijmans and Spooner, 2001). Therefore, the results obtained in this research constitute a source of updated and valuable information on the edaphoclimatic characteristics in which wild tomatoes and phylogenetically related species are distributed along its natural geographic range.

In general, geographical distribution of 16 wild species related to the cultivated tomato is wide, from Colombia through Peru, comprising Pacific coastal region to Chile and the Andean mountains, with an altitudinal range from sea level to 3,300 m (Peralta et al., 2008; Bergougnoux, 2014. However, within this distribution, there are overlapping areas between several species or regions with specific distribution such as the endemic species of the Galapagos Islands (S. cheesmaniae and S. galapagense) or hyper arid regions of northern Chile with other rare endemic species, S. sitiens. Within these distribution patterns, it is also possible to identify differences and similarities between the species that conform each group, for example, the similarity between S. arcanum and the species of Lycopersicon group (Figure 6), reflecting a wider distribution and adaptations to local sites of ecotypes (Peralta et al., 2008). These environmental characteristics reflect the ecological adaptation patterns and habitat preference of each species (Nakazato et al., 2010; Vilchez et al., 2019) (Figures 6, 7, Table A1 and A3 in Supplementary Material). It is worth mentioning that these results also suggest a thorough revision of the proposed groups, incorporating the new passport data as well as genetic and molecular information to corroborate the belonging of each species to the phylogenetic assigned groups. The aforementioned are under the assumption that the species are closely and genetically related and in expecting that their adaptation areas are similar.

Regarding wild tomato species and phylogenetically related species, few studies have been carried out with an ecogeographic or climatic focus. A comprehensive treatment integrates main botanical, biological, and ecological characteristics of each wild tomato and related species (Peralta et al., 2008); other studies focused on distribution of species richness and diversity through the analysis with GIS (González, 2013) and established conservation priorities (Vilchez et al., 2019); further geographical and ecological characterization have been investigated in 10 tomato species determining soil and climate variables (Nakazato et al., 2010); studies have been conducted on tomato biogeography, S. lycopersicum var. cerasiforme, in its center of origin and domestication (Délices et al. (2019); and finally climatic effects on species distribution (Lin et al., 2020) and bioclimatic characterization, and identification of ecological descriptors and patterns of climatic diversity of 12 wild tomato and 4 closely related species (Ramírez-Ojeda et al., 2021a) have been studied. In this sense, this study complements the information available, providing information on soil characteristics that had not been analyzed.

The canonical correlation analysis satisfactorily identifies climatic variables with greatest influence on edaphic variables and vice versa, with a correlation of 0.80 representing 74% of total variation in 4,649 accessions. One main conclusion is that variables related to water availability (ET, Bio12, Bio14) have a great influence on physical (BD) and chemical soil characteristics (BS, pH). This pattern is persistent in all six groups. This relationship can be better observed in group 5 (S. juglandifolium and S. ochranthum) accessions with greater availability of annual precipitation and evapotranspiration, which present lower pH, BD, and BS than the rest of species; that is, they are located in soils with the lowest agricultural quality (Figures 4, 5). This methodological approach is promising to be applied at other scales, considering the analysis at population level of each species and climatic and edaphic factors limited to smaller areas of distribution. This basis of ecogeographic characterization could incorporate information from genetic and ecological studies. A better understanding of these variables would allow the generation of projection models in different climate change scenarios (Violle and Jiang, 2009;Luebert and Weigend, 2014; Godoy-Bürki, 2016; Lin et al., 2020).

Ecological descriptors obtained, despite the incorporation of new accessions, are very similar to the ranges reported by Peralta et al. (2008) and Ramírez-Ojeda et al. (2021a) and generally identify the groups of species proposed in the classification. It is important to mention that this methodology has been widely used in the study of other species (Ruiz-Corral et al., 2008; Cerda-Hurtado et al., 2018; Sánchez-González et al., 2018; Ramírez-Ojeda et al., 2021a; 2021b). With this information, it is also possible to identify those species with tolerance to extreme conditions, for example, low and high temperatures, humidity conditions, altitude, pH, BD, and all the possible conditions when associating a species with a climate type or soil unit (Table 3 and Table 4, Tables A1, A2, and A3 in Supplementary Material).

Edaphic diversity (Figure 7) tends to be more constant between species groups and sections with respect to climate diversity. In general, considering climate and soil characteristics, specific adaptation patterns for each species group can be identified: Lycopersicon group (group 1) corresponds to species with lower altitude and higher mean annual temperature; species of Juglandifolia section (group 5) are those with the highest water availability, lowest pH, BD, and base saturation; species of Lycopersicoides section (group 6) are the ones with the highest altitude, the lowest mean annual temperature, and lowest water availability, groups 4 and 6 have the lowest water availability and soils with favorable agricultural characteristics, differing by altitude. The rest of the species (groups 2 and 3) are in the transition zones with the rest of the wild tomato species. One aspect to highlight is that when combining or considering climatic and edaphic information, it is possible to characterize in a better way the different groups, being able to better identify their differences and similarities.

Among possible uses of this approach is the identification of the germplasm with tolerance to adverse biotic and abiotic factors (Foolad and Lin., 2000; Mittova et al., 2004; Venema et al., 2005; Zhao et al., 2005; Ruiz-Corral et al., 2008; Chetelat et al., 2009; Arellano-Rodríguez et al., 2013; Ruiz-Corral et al., 2013; Cervantes-Moreno et al., 2014; Chen et al., 2015; Nosenko et al., 2016; Stam et al., 2017a; Stam et al., 2017b; Flores-Hernández et al., 2017; Razali et al., 2018; Dinh et al., 2019; Vilchez et al., 2019) with potential use for genetic breeding, identification of routes of germplasm accession, and areas of high and low diversity for use and conservation (Vilchez et al., 2019). In the information contained in Table 3, Table 4, and Figures 4, 5, it is possible to identify species with extreme values that indicate tolerance or resistance to climatic and edaphic factors, with potential use as germplasm for genetic breeding.

Finally, the hot spot analysis could satisfactorily identify regions with the greatest diversity of species. These are priority areas for conservation, either due to high or low diversity. Regions identified as of great importance for conservation comprise endemism. Diversity contained in populations with few isolated individuals or with restricted distribution could be more affected by environmental and anthropic changes. This result could be explained by the quantity and geographic distance between the accessions of species studied. However, this first approximation is very useful and agrees with the diversity results obtained for wild potato species in Peru (Hijmans and Spooner, 2001).

This research determines the most important edaphoclimatic descriptors of wild tomato species and its closely related species along their natural geographic range in South America. Patterns of climatic diversity correlate with species groups and sections proposed in current classification. New edaphic characteristics analyzed in the same areas were also useful, although with less discrimination than the climatic variables. Interaction between climatic and edaphic factors allows for understanding species distribution and their adaptation patterns. Another feature to highlight is the incorporation of new data from recent collections of specimens being properly identified (Ministerio del Ambiente, 2020) that were not considered before in other studies, and thus expanding precision and reliability of these results. Most important areas for conservation of wild tomato species and related outgroups were detected. Under this premise, this contribution is promissory for further ecogeographic study of wild tomatoes and closely related species at the local population scale, especially focused in situ conservation reserves as well as in localities outside protected areas. Edaphoclimatic descriptors in addition with other abiotic or biotic factors could help to better estimate the species ecological niches and determine local ecotypes. Selected descriptors would be tested in models of current and future distribution considering the impact of climate change and anthropic activities along the distribution range of these valuable genetic resources. Finally, this research can be used as a study model to replicate in other species.