Drivers of urban biodiversity in Mexico and joint risks from future urban expansion, climate change, and urban heat island effect

between multivariate risk and residual species richness (Figs 3 and 4). After controlling by urban area effects, most municipalities with many species are in different parts of the country (Fig 4). Still, these municipalities are in regions where tropical evergreen and semideciduous forests are dominant, suggesting that the regional species pools likely play a strong role in assembling urban biodiversity through mass effects. Discussion How humans influence the community assembly processes in Latin American cities is a rele- vant question in urban ecology and conservation today [14, 40, 73]. We found differences in urban and rural assemblages in the species-area relationships across four taxonomic groups. Urban species assemblages of 389 Mexican municipalities are strongly related to scaling effects associated with the size of urban areas, the number of ecoregions present, the network of urban protected areas, human impact degree, and the economy size. Other environmental and socioeconomic variables seem not to play a substantial role in shaping the urban assemblages. Many municipalities with small urban areas contain few species, and likely, these areas are more susceptible to stochastic events than larger urban areas where habitat filtering mecha- nisms may be more dominant [5, 30, 74]. Drivers of urban biodiversity The fitting model approach of SAR models revealed that urban species richness in Mexican municipalities may be described by a sigmoidal curve [75, 76] and several ecological processes

may be inferred for small, intermediate, and large urban areas (Fig 1). For instance, municipal- ities with small urban areas have fewer species than those with larger urban areas regardless of the surrounding habitat, biogeographical region, or specific urban planning policy. Mass effects may play a major role in shaping urban biotic assemblages in tropical regions than those in desert or xeric regions. The municipalities with few species are expected to be influ- enced by colonization-extinction dynamics, and probably urban assembly is dominated by sto- chastic events. Accordingly, more ecologically relevant habitat is necessary to allocate in these municipalities to reduce local extinction rates. Our results are in line with other large-scale studies in Europe, with rural and regional assemblages exhibiting a similar positive relationship between the urban area and species rich- ness (e.g., [41]). A sigmoidal shape in species-area relationships across the four taxonomic groups suggests that differential ecological and evolutionary processes shape larger and smaller areas. For instance, a potential limit to the number of species in larger urban areas (i.e., a potential carrying capacity) suggests that some municipalities’ urban biodiversity is likely already homogenized. Therefore, it is expected that biotic composition is more similar in these places. Further studies will be necessary to evaluate whether biotic homogenization emerges across Mexican municipalities and how urban configurations shape urban ecological patterns.

The Random Forest (RF) models corroborated the results of the SAR model fitting (Fig 2). RF is a highly flexible method that allows comparing the relative role of different drivers of urban species richness (see [55]) for a similar approach by using maize yields). We found evi- dence of scaling effects on urban species richness through different variables related to the urban area’s size. For instance, GDP is a proxy of urban development and urban area [34]. Therefore, it is expected that GDP will be a proxy of more available economic resources from private and public institutions to invest in creating urban green spaces, translating into a high number of species (Table 1). However, municipalities with high GDP are not necessarily reflecting a high and homogeneous income in the population. Accordingly, more research is necessary to establish how income differences inside urban areas across municipalities impact PLOS ONE Urban biodiversity and anthropogenic pressures PLOS ONE | https://doi.org/10.1371/journal.pone.0308522 October 4, 2024 12 / 19