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

Residential yards constitute a substantive biodiverse greenspace within urban areas. This biodiversity results from a combination of native and non-native species and can contribute to biotic homogenization. Geographical climatic patterns affect the distribution of native species and may differently affect non-native species. In this study, we examined biodiversity and biotic homogenization patterns of yard-dwelling land snails across 12 towns in Oklahoma and Kansas (USA). The 3 x 4 array of towns incorporated a N-S winter temperature gradient (mean low January temperature range = -8.4 to 0.1°C) and an E-W annual rainfall gradient (annual rainfall range = 113.8 to 61.3 cm/yr). Ten yards per town were surveyed. We hypothesized that mild winter temperatures and greater annual rainfall would be associated with greater snail abundance and richness, and that the presence of non-native species would contribute to biotic homogenization. Non-native snails were present and often abundant in all towns. Snail communities varied with both rainfall and cold temperature. Contrary to our prediction, snail abundance was inversely related to annual rainfall–likely because drier conditions resulted in greater yard watering that both augmented rainfall and maintained moist conditions. Sørensen similarity between towns for the entire land snail community and for only non-native species both showed distance-decay patterns, with snail composition becoming less similar with increasing distance—patterns resulting from species turnover. The biotic homogenization index also showed a distance-related pattern, such that closer towns were more likely to have biotic homogenization whereas more distant towns tended to have biotic differentiation. These results support the concept that biotic homogenization is more likely regionally and that climatic changes over distance result in species turnover and can reduce spatially broad biotic homogenization.


38
Biotic homogenization occurs when biotas become more similar through the processes of the 39 loss or expansion of native species and the establishment of non-native species [1] and is 40 considered a global phenomenon [1,2]. Urbanization has a high potential for producing biotic 41 homogenization, given the high levels of disturbance and severe habitat modification, combined 42 with multiple routes of species introduction through the movement of goods and people [3]. 43 Studies of urban biota have supported both biotic homogeneity and biotic differentiation [e.g., 4, 44 5-7], illustrating trade-offs between changes in native species occurrence and abundance and the 45 establishment of non-native species. Cities may be located in biodiverse regions and in some 46 places retain significant native diversity, including rare species [8,9], and some native species 47 may benefit from urbanization and thereby increase biotic homogenization [3,10,11]. 48 Introduced non-native species adapted to urban environments can establish and increase biotic 49 homogenization [3] or contribute to biotic differentiation [7]. 50 Urbanization produces homogenization of physical habitats, with similar composition of 51 roads, residential areas, commercial areas and aquatic features across cities [3,12,13]. Within 52 this overall urban homogenization is a gradient of habitat types from seminatural parks to 53 industrialized areas or city centers with little vegetation. These habitat types act as environmental 54 filters [14] that result in differential patterns of biodiversity [11,15] and in some cases 55 differential patterns of biotic homogeneity and biotic differentiation [16]. 56 Climatic differences, specifically in temperature regimes and rainfall characteristics, 57 strongly affect species distributions -however the effects of spatial climatic factors on biotic 58 homogenization are poorly known. 59 To reduce the effects of comparing multiple habitats across cities, this study concentrated 60 on residential yards or gardens that surround dwellings. These yards typically comprise the  Land snails are an informative model system to study biotic homogeneity in human-67 influenced landscapes, including yards. Snail populations in urban areas can be quite speciose, 68 comprised of both local native species and a suite of non-native species [11,16,21,24,25].

69
Although snails move slowly on their own, they can be frequent hitch-hikers -dispersing 70 through such routes as the plant trade [24,26,27] or landscaping materials.

71
In this study, we examine the biotic homogeneity of land snails in the residential yards of 72 12 towns in Oklahoma and Kansas (USA) across two axes of climatic gradient: rainfall and 73 temperature. Both rainfall and cold winter temperatures influence the distribution and abundance 74 of land snails, resulting in a pattern of species replacement among native species [28]. Non-75 native species often have wide environmental tolerances and this study examines the patterns of 76 native versus non-native species across climatic gradients to determine patterns of biotic 77 homogeneity versus biotic differentiation relative to climatic factors. 78 We anticipated that the widespread introduction of land snails combined with 79 disturbance-related loss of native species and expansion of tolerant native species would result in 80 significant biotic homogeneity of land snail faunas. We further hypothesized that biotic 81 homogeneity would vary with climatic factors. Specifically, we hypothesized that biotic 82 homogeneity will be lower in towns with different climatic features (more distant towns) than in 83 towns with more similar climates (closer towns). Texas, which are much larger than Oklahoma). With the addition of sites in Kansas to increase 90 the temperature range, we captured a relatively large climatic range within a relatively small 91 area.

194
Towns are arranged from north to south, starting with the east tier, then the mid-and west tiers.   Ardmore, and was highest in southwestern Oklahoma (Fig 4A). The regression of annual rainfall 214 and low winter temperature with snail abundance was significant (F 2,9 = 7.253; p = 0.013). Snail    Snail species richness displayed a spatially complex pattern (Fig 4B). Richness was low 228 in the northwest and southeast portions of the survey and high in the southwest, which is similar 229 to the pattern of total snail abundance (Fig 4B versus Fig 4A). Snail richness showed little  A distance decay pattern was evident as the Sørensen dissimilarity of snail assemblages 253 increased with increasing distance among towns (p < 0.001, R 2 = 0.330; Fig 7A). Simpsons 254 dissimilarity, a measure of species replacement, also increased with distance (p < 0.001, R 2 = 255 0.389; Fig 7B). Nestedness (within the dissimilarity index) was little related to distance among 256 towns (p = 0.50, R 2 = 0.007; Fig 7C).   Table 1) and coldest mean monthly temperature (R 2 = 0.379).

266
Together, rainfall and cold temperature explained 50.1% of the variation in taxonomic 267 composition among towns.  Snail species typical of yards in each town (based on SIMPER analysis; Fig 6) range Non-native snail species were among the typical species in 11 of 12 towns (Fig 6). The
14 292 The homogenization index among pairs of towns ranged from moderate homogeneity 293 (0.231 in a scale of -1 to +1) to high differentiation (-0.516). Of the 66 pairs of towns, 26 pairs 294 supported homogenization, 39 supported differentiation, and 1 was neutral (S3 Table). The 295 regression between the distance between towns and the homogenization index indicated a 296 distance-decay type of relationship, in which closer towns were more likely to show biotic 297 homogenization and more distant towns were more likely to have biotic differentiation (p = 298 0.004; R 2 = 0.11; Fig 8). Land snail populations in residential yards were speciose and displayed an overall 307 distance-decay spatial pattern based on similarity of taxonomic composition; that is, reduced 308 similarity (or increased dissimilarity) with increasing distance among towns.  relative to the all-taxa dataset). Species replacement is apparent for both entire snail assemblages 375 and for the non-native subset but is much stronger for the entire assemblage. Species 376 replacement is consistent with spatial patterns of climate and (water) management and mirrors 377 spatial changes in the native fauna.

378
Biotic homogenization and biotic differentiation were related to the distance between 379 towns. In general, biotic homogenization was more frequently associated with nearby towns, 380 whereas distant towns had higher frequency of biotic differentiation. This spatial relationship is 381 consistent with the spatial patterns of climate and species turnover. That is, as distance between 382 towns increased, climate differences increased -resulting in snail species turnover and increased 383 biotic differentiation.
18 384 385 Acknowledgements 386 We are extremely grateful to the many residents who allowed us to survey their yards and 387 appreciated their frequent offers of hospitality. We also thank the state extension agents and 388 others who helped find participating residents. Matthew Carman helped with field work, snail 389 sorting and identification.