Climate gradients, and patterns of biodiversity and biotic homogenization in urban residential yards

groups had many intermediate forms), succinids were identified to family, and individuals that combined characteristics of both Gastrocopta pentodon and Gastrocopta tappanianna were designated as G. pentodon. Immature snails and shells were counted as unidentified. The paucity of historic records in the regions of the 12 towns combined with the wide cli- matic range across towns made determination of native versus non-native species challenging. A conservative approach was used; specifically, species recorded as present in Oklahoma and/ or Kansas in Hubricht [28] were considered native, even though many species have restricted distributions with the state. For example, Anguispira alternata was considered native in both states even though its native distribution is restricted to the eastern-most tier of towns [28] and its presence in the middle of Oklahoma [Norman: 23] is a disjunct, likely introduced population. Distance between pairs of towns was measured between town centers using Google Earth Pro. Data analysis Data from visual surveys and soil samples were combined; hence abundance and richness include both live snails and shells, unless otherwise specified. In addition, data from the 10 sur- veyed yards in each town were pooled to produce a single snail composition data set per town. Patterns between climatic factors (mean monthly low temperature for January and annual rainfall) and snail abundance and richness were analyzed using linear regression (Systat within SigmaPlot 12.0 software). Effects of town population size on snail richness were also tested

using linear regression. Taxonomic composition of snail assemblages across towns was com- pared using one-way PERMANOVA (Primer 6 plus Permanova+; Primer-E software). Snail compositions within towns were characterized using SIMPER (Primer 6) to determine the spe- cies contributing 90% of the within-town Bray Curtis similarity. The relationships between snail assemblages and climatic variables (mean cold temperature and annual rainfall) were analyzed using DistLM (Distance-based linear models; Primer 6 plus Permanova+). Snail assemblages among the 66 combinations of pairs of towns were compared using three similarity indices: Sørensen and Simpson indices are both presence/absence indices and Bray- Curtis is an abundance-based metric. Equations and references for all three similarity indices are found in Koleff et al. [38]. Sørensen and Bray-Curtis indices were calculated using Primer 6 and the Simpson index was manually calculated using an Excel spreadsheet and the equation from Koleff et al. [38]. The Sørensen index was used to assess changes in assemblage similarity across distance (= distance decay), which was analyzed using exponential decay regression (Systat software). Although the Jaccard Index is often used for distance decay, the Sørensen and Jaccard indices are similar and a regression of our data using the 66 possible pairwise comparisons produced an R2 of 0.991. Sørensen similarity is also commonly used similarity index for measuring beta diversity, indicating biotic homogenization and, in combination with the Simpson similarity

index, is used to distinguish spatial turnover and spatial nestedness components of biotic homogenization [39, 40]. The effect of non-native species on the homogenization versus differentiation of communi- ties was examined using the homogenization index described in Horsa´k, Lososova´ (16). This index is the difference between the similarity based on all taxa present and that based on only native species, and produces values ranging from -1 to +1. Positive values indicate that non- native species contribute to homogenization by having a relatively low similarity whereas neg- ative values indicate differentiation, with the similarity within non-native species being higher than for all species combined [16]. PLOS ONE Diversity patterns of yard-dwelling snails across climate gradients PLOS ONE | https://doi.org/10.1371/journal.pone.0234830 August 28, 2020 5 / 18