Abstract
Urbanization generates a distinctive suite of environmental filters -- elevated temperatures, impervious surfaces, artificial
light at night (ALAN), noise pollution, and chemical contamination -- that systematically restructure animal communities
along urban-rural gradients. This study quantifies the effects of urbanization intensity on four taxonomic groups (breeding
birds, ground beetles, bats, and herpetofauna) across urban-rural gradients in Helsinki, Uppsala, and Amsterdam using
standardised multi-taxon surveys at 72 sites spanning the full built environment intensity range (2021-2023; n = 19,284
individual records across 294 taxa). Urbanization index (UI; 0-1 composite of impervious surface %, ALAN, noise, and
human population density within 500 m) was the primary predictor of species assemblage composition (permanova R2 =
0.48, p < 0.001). Total species richness declined significantly with increasing UI across all groups (mean R2 = 0.54; all p
< 0.001), but functional diversity showed a non-monotonic response: functional richness declined at high UI but functional
evenness peaked at intermediate UI (0.4-0.6) where urban generalist and specialist species co-occur. Biotic
homogenisation was confirmed across all three cities: Bray-Curtis dissimilarity between sites decreased by 42.4 +- 6.8%
from rural to urban cores. ALAN intensity was the strongest single predictor of bat activity reduction (beta = -0.72 +- 0.10,
p < 0.001), while impervious surface cover was strongest for ground beetles and herpetofauna. Urban green space
characteristics -- particularly connectivity, tree canopy cover, and distance from impervious surfaces -- significantly
moderated biodiversity loss at all three sites, with highly connected green spaces retaining 64.8% of rural reference
richness versus 28.4% in isolated urban parks. These findings provide quantitative evidence for urban
biodiversity-sensitive planning standards under EU Urban Greening Strategy and city-level biodiversity action plan
frameworks.