Species richness and functional diversity of isopod communities vary across an urbanisation gradient, but the direction and strength depend on soil type

Urbanisation involves major changes in environmental conditions such as light, temperature, humidity and noise levels, but the effect of urbanisation on soil conditions and soil biodiversity has received less attention. The reported effects on species richness across a rural to urban landscapes are not unequivocal. Positive, negative and neutral effects have been found, but what is causing this ambiguity in the relationship between species richness and urbanisation is poorly understood. Here we tested whether and how urbanisation affects isopod diversity and whether this depends on habitat soil type. We recorded isopod species presence and soil type in each 1 km2 grid cell in a 500 km2 area in and around the city of Amsterdam, the Netherlands. We analysed the relationship between the number of species and the most important traits (body size and drought- and inundation resistance), the percentage built-up area, and soil type (peat or clay) across a rural-urban landscape gradient. Because many studies find a negative effect of urbanization on biodiversity, we expected a decrease in isopod species richness by increasing amount of built-up area. We found a negative association between isopod species richness and the amount of built-up area across the rural-urban landscape gradient when rural soils contained clay, while for rural soils containing peat the highest species richness was found at intermediate levels of urbanisation. Hence, the relationship between isopod species richness and the level of urbanisation was mediated by soil type. We also found a significant relationship between percentage built-up area and functional traits, irrespective of soil type. Specifically, the percentage built-up area showed a positive correlation with isopod drought resistance as well as body size. These results together demonstrate (1) that soil type variation introduces an important, but often overlooked context-dependency in the effect of urbanisation on biodiversity and (2) that a functional trait approach is an important tool for understanding general effects of environmental change on biodiversity, irrespective of context-dependency.