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

We conducted a rarefaction analysis to select those municipalities with adequate survey sampling for each taxonomic group. In this way, we avoid the potential effects of sampling bias, which might affect the posterior analysis and interpretations. First, we evaluate the inven- tory completeness fitting a smoothed species accumulation curve [60] with the method exact of the function specaccum in the R package vegan [61]. We used the slope of this curve as an index of how many new species appear in urban and rural areas for each municipality and a cut-off value of 0.05 (a flat curve) to exclude those municipalities with incomplete faunistic inventories from the SNIB dataset [60]. These analyses give us a better picture of urban inven- tories using these opportunistic occurrence datasets from museums and human observations. The dataset of municipalities with adequate biodiversity sampling for urban areas was a com- posite of 389 Mexican municipalities. We calculated the total species richness (i.e., the sum of species richness for all four taxa) and added the socioeconomic and environmental data previ- ously collected for each of the 389 municipalities (Table A and B in S1 Data). Although several municipalities contain a very low number of species (i.e., less than five species), these vary for each taxonomic group (e.g., for amphibians, there are 66 municipalities with less than five spe- cies, but for birds, there are only eight municipalities; see Table C in S1 Data). However, we decided to exclude these municipalities from the species-area relationships and random forest

analyses to avoid potential effects on the recovered SAR patterns and the effect size of socio- ecological drivers of urban species richness. SAR models We tested for differences in species-area relationships (SAR) across an urban-rural dichotomy for Mexican municipalities. We log-transformed species richness and area to compare slopes and intercepts between urban and rural environments in the four taxonomic groups. Then, we fit a series of SAR models for urban species richness without transforming data by adopting an information-theoretic framework to compare and evaluate the best functions that maximize the explanatory power of SAR using the small-sampled corrected Akaike information criterion (AICc). We fitted the 20 SAR models in the sars package using the sar_multi function [46]. Analyses were conducted for each taxonomic group. The selected models maximized the explanatory power using the Akaike weights derived from the AICc values. We selected the first five models with the highest Akaike weights and evaluated them using the Lilliefors Table 1. (Continued) Hypothesis Variable Prediction Explanation Number of ecoregions (NumbEcoreg) Cities that are located in regions that overlap different ecoregions are expected to have more diverse biotas and, therefore, more species. Ecoregions are units containing a distinct species assemblage and environmental conditions [103]. Climate Annual mean temperature (bio1), Maximum temperature of warmest month (bio5), Minimum temperature of coldest month (bio6), total precipitation (bio12), Precipitation seasonality (bio15).

We expect that municipalities with higher temperature, more rainfall, and low seasonality are expected to have more species Several potential mechanisms are underlying the climate-species relationships [89, 103]. For instance, regions with higher productivity (i.e., warm and humid climates) and lower seasonality (stable climates around the year) concentrate more species than other regions [104, 105]. Cloud cover Cities with a higher percentage of cloud cover are expected to have high species richness. Cloud cover is a direct proxy for rainfall in mountain regions. Itis linked with important demographic processes on several taxa and strongly predict species richness in tropical areas. Mean annual cloud frequency (%) was calculated for a period over 2000–2014. https://doi.org/10.1371/journal.pone.0308522.t001 PLOS ONE Urban biodiversity and anthropogenic pressures PLOS ONE | https://doi.org/10.1371/journal.pone.0308522 October 4, 2024 5 / 19