Factors driving the vascular plant species richness in urban green spaces: Using a multivariable approach

• We analyzed plant species richness in green spaces with a multivariable approach.
• Species groups were affected most by different combinations of driving factors.
• Green space size was most relevant factor driving vascular plant species richness.
• Patch shape and distance to the urban edge were of minor relevance.
• Urban green spaces larger than 6 ha should be given priority in being conserved.

Many studies have shown high vascular plant species richness in urban areas and, especially, in its green spaces. However, little is known about the factors driving the numbers and proportions of different species groups. The aim of our study was to test for the effects of patch size, patch shape, and distance to the urban edge as well as the combined effects of these factors on the numbers and proportions of total, native, non-native, endangered, ornamental, and nitrophilous vascular plant species. We conducted vascular plant surveys in 32 urban green spaces in the city of Hannover, Germany. We detected positive correlations between patch size and total, native, non-native, endangered, ornamental, and nitrophilous vascular plant species numbers and the proportion of endangered species by Spearman's rank correlations and linear regressions. A more compact patch shape, calculated by the shape index, affected the proportion of native, non-native, and ornamental species positively. Testing combined effects of factors with multiple linear regressions underlined the importance of patch size in combination with distance to the urban edge, and in combination with distance and patch shape. We conclude that in the context of recent urbanization processes, it is most important to create and conserve large urban green spaces (>6 ha) in order to maintain vascular plant species richness. As species groups were affected most by different combinations of driving factors, our study highlights the importance of using multivariable approaches for detecting effects more precisely.