Introduction
Understanding how taxa within a community respond to shared environmental variation is of particular relevance in the context of climate change. Climatic changes can affect species distributions, population abundance, and evolution due to impacts on genetic drift, structuring of genetic variation, gene flow, and selection (Foden et al., 2019; Román-Palacios & Wiens, 2020; Aguirre-Liguori et al., 2021). Species’ responses depend on their ecological niche, defined as all the variables that influence organismal fitness (Hutchinson, 1957; Blonder, 2018). A species niche will determine the amount and quality of habitat available to populations, which can change independently. Habitat quantity has been commonly used as a metric of population abundance as predicted by metapopulation and island biogeography theory (Fahrig et al., 2013). However, the amount of habitat is not sufficient to describe all processes affecting a species and can be misleading, and evidence suggests that habitat quality should also be considered (Morletilli et al., 2010; Walting et al., 2020; Galán-Acedo et al., 2021; Regolín et al., 2021). Separately assessing how habitat availability and habitat quality changed over past climatic changes for a range of species offers an opportunity to understand the degree to which these patterns are coupled, and therefore offer more nuanced information about how organisms may respond to ongoing and future climate changes.
The ecological niche can be modeled in terms of the climate factors that determine the occurrence of species in space, or the “Grinnellian niche” (Soberón, 2007; Sillero et al. 2021). Most studies addressing the effect of climatic changes on species with a niche modeling approach focus on habitat quantity and distribution, and relatively fewer studies have focused on changes in habitat quality (But see Morente-López et al., 2022; Kebaïli et al., 2023). The ecological niche can be represented as an ellipsoid in multivariate space, consisting of the range of suitable conditions for a defined taxon based on a determined set of variables (Jiménez et al., 2019; Osorio-Olvera et al., 2020a). In this framework, the ellipsoid centroid (niche centroid) corresponds to high suitability conditions and high habitat quality, whereas positions in the multidimensional space near the ellipsoid borders correspond to more marginal conditions or lower habitat quality (Martínez-Meyer et al., 2012; Osorio-Olvera et al., 2020b). According to the center-marginal hypothesis (Eckert et al., 2008; Pironon et al., 2017), populations living under more suitable conditions present higher abundance and genetic diversity, whereas populations inhabiting more marginal conditions are expected to have lower abundance, lower genetic diversity, and higher drift (Lira-Noriega & Manthey, 2014; San Juan et al., 2021). Moreover, populations living at the limit of their tolerances often experience higher selection pressure and respond by adapting to those challenging environmental conditions (Aguirre-Liguori et al., 2017, Bontrager et al., 2020). Therefore, a population’s distance to the ellipsoid centroid may be proportional to selection pressure.
As climatic conditions change, populations may track suitable environmental conditions geographically, assuming that niches do not evolve (i.e., niche conservatism; Wiens et al., 2010). This can result in changes in the species’ distribution and leave a genomic signature of range expansion (Lenoir & Svenning, 2015; Tomiolo & Ward, 2018). It can also lead to founder effects and surfing of deleterious alleles at the margin of the expanding front (‘allele surfing’; Escoffier et al., 2008; Gilbert et al., 2018). Further, climate changes can affect species abundance by the reduction or increase in the amount of suitable area (Fahrig et al., 2013). Regarding habitat quality, changes in climatic conditions can affect population fitness and selection pressure because the distribution of habitats closer to the niche centroid and marginal conditions could shift. For instance, a geographic location consisting of conditions matching the niche centroid at one time could shift to more marginal conditions at a different time, while still being suitable. Therefore, populations inhabiting this area would experience a population decline, and/or decrease in fitness and/or an increase in selection pressure on traits related to variables that have become marginal at that location. This highlights that environmental changes can affect both habitat quantity and habitat quality independently and they yield different population genomic predictions.
The Baja California peninsula (BCP) is a good system to assess how different taxa respond to changes in habitat quantity and quality during different climatic conditions. It presents a wide variety of ecosystems ranging from desert scrub to high-altitude forests (Rebman & Roberts, 2012). It spans 10 degrees of latitude with stark differences in rainfall and temperature and since it is a peninsula, ecological and expansion-dispersal dynamics are constrained by its geography (Dolby et al., 2015). Native species have largely been co-distributed and isolated from the mainland since the Gulf of California finished flooding 6.3 Mya (Oskin & Stock, 2003; Darin et al., 2024). In particular, low and high amplitude glaciation cycles during the Pleistocene (~3 Mya) are expected to have had a large impact on redistributing climatic conditions and therefore the distribution and abundance of populations (Dolby et al., 2015). About 80 taxa show a diffuse north-south genetic co-divergence signal centered in the middle of the Peninsula (Dolby et al., 2015; Araya-Donoso et al., 2022) and show ecological niche divergence, suggesting potential adaptation to local environmental conditions (Cab-Sulub & Álvarez-Castañeda, 2021), which offers an opportunity to assess both species-level and clade-specific changes in habitat quantity and quality. It also allows us to test which organismal features determine responses to environmental change. For example, highland species that can resist cold may respond differently than species inhabiting lowland deserts adapted to low water availability, or taxonomic groups with different physiological requirements, such as mammals, plants, and reptiles, could also respond specifically.
Previous models and descriptions have detected contrasting patterns of species distribution to the Last Glacial Maximum (LGM) on the peninsula. Some taxa show range expansions during LGM (e.g. Graham et al., 2014; González-Trujillo et al., 2016; Harrington et al., 2017; Arteaga et al., 2020), whereas others show range contractions (Klimova et al., 2017; Valdivia-Carrillo et al., 2017). Cab-Sulub & Álvarez-Castañeda (2021) proposed that southern clades within species contracted their distribution ranges to LGM, whereas northern clades expanded. Furthermore, some studies have assessed species’ past demography with genetic data showing signatures of population contraction towards LGM (Álvarez-Castañeda & Murphy, 2014; Ferguson et al., 2017; Phuong et al., 2017; Martínez-Noguez et al. 2020), which does not agree with the range expansion patterns proposed by some distribution models. Therefore, changes in population size could be determined by not just habitat quantity but also habitat quality.
Here, we used ecological niche modeling and species distribution modeling to compare intra-specific and inter-specific distribution patterns and niche marginality of 21 taxa from the Baja California peninsula including mammals, reptiles and plants that inhabit highland and lowland environments, and have different levels of genetic divergence along the peninsula. We aimed to assess changes in habitat quantity and quality between LGM and present day and determine if organismal characteristics affected these patterns. Then, we used our models to generate predictions about the effects of historical climate change on abundance and selection pressure on natural populations that can be tested in the future with genomic data.