As urban areas increase in extent globally, domestic yards play an increasingly important role as potential contributors to ecosystem services and well-being. These benefits largely depend on plant richness and composition of yards, which in turn are determined by both natural and anthropogenic factors, such as socioeconomic characteristics. Most of our understanding of the drivers of biodiversity and composition of urban yard floras comes from studies conducted in individual cities. Drivers of urban yard diversity at continental scales, and interactions between cities, are less well-studied despite the increasing awareness of current global change impacts on urban ecosystems. Here, we analyzed plant richness and phylogenetic composition of cultivated and spontaneous flora of 117 private yards from 6 major metropolitan areas in the US to answer the following questions: (1) what factors drive plant richness and phylogenetic composition of the cultivated and spontaneous flora in urban yards at the continental scale? And (2) to what extent do these factors interact? Yard plant richness and phylogenetic composition for each pool of species were expressed as a function of biophysical and socioeconomic variables and yard structural characteristics using spatial simultaneous autoregressive error (SARerr) models and (spatially explicit) structural equation models (SEM).
Results/Conclusions (199/200):
Our results revealed that at the continental-scale extreme temperatures largely determined yard species richness and phylogenetic composition, although this effect was often mediated by yard area. Furthermore, the explanatory power of our models was larger for all spontaneous diversity parameters than for cultivated diversity, likely reflecting the importance of homeowners’ actions in structuring cultivated yard floras. Precipitation positively predicted spontaneous richness but negatively predicted cultivated richness, supporting the idea that irrigation may be compensating for precipitation in driving cultivated yard plant diversity at the continental scale. Among various socioeconomic variables, only education level showed a significant effect on cultivated phylogenetic composition. Due to the characteristics of our dataset, this conclusion may only be valid for upper-middle class households and further research should elucidate the extent to which these results hold among households in lower socioeconomic brackets. SEM models provided a more accurate description of the influences and interactions between predictor and response variables than SARerr regressions. We show these interactions for private yards for the first time and highlight their implications at the continental scale. These findings can inform the adaptation of US urban vegetation and the management of ecosystem services in cities in the face of global change.