| [ report an error in this record ] | basket (0): add | show |
 |
| Green and blue infrastructure but not anthropogenic drivers affect invertebrate communities in urban ecosystems Morpurgo, J.; Remme, R.P.; Mudde, Q.D.; Didaskalou, E.A.; Serwatowska, K.J.; Trimbos, K.B.; Hu, M.; van Bodegom, P.M. (2025). Green and blue infrastructure but not anthropogenic drivers affect invertebrate communities in urban ecosystems. Ecol. Indic. 176: 113650. https://dx.doi.org/10.1016/j.ecolind.2025.113650
In: Ecological Indicators. Elsevier: Shannon. ISSN 1470-160X; e-ISSN 1872-7034
|   |
| Available in | Authors |
| |
| Author keywords |
Biodiversity; Green infrastructure; Remote sensing; Species distribution model; Urban ecology; Urban planning |
| Authors | Top | |
|
|
| Abstract |
Rapid urbanization alters habitat quality and connectivity, influencing species dispersal and ultimately shaping community assembly. Increasingly, urban environments and their vegetation are shown to be important for the conservation of biodiversity. However, urban species distributions remain poorly understood. By integrating DNA-based sampling with species distribution models (SDMs), we aim to quantitatively assess urban community assembly while using a novel green infrastructure classification framework intended to assess both urban biodiversity and ecosystem services. We sampled invertebrate distributions in The Hague, the Netherlands, using two complementary DNA-based methods: traditional bulk trapping (n = 205) and environmental DNA (eDNA) sampling (n = 207). Species were identified using DNA sequencing and Operational Taxonomic Units, with presence-absence data used to develop SDMs driven by vegetation and anthropogenic indicators. The SDMs generally outperformed random models (59.5 %) and performed strong during calibration (90.4 %, AUC > 0.70). They highlighted that vegetation density, structure, and proximity to water are the primary drivers of invertebrate distributions, while direct anthropogenic pressures play a minimal role. At the same time, very few models (1.3 %) performed well during validation. This finding indicates challenges in predicting species distributions and suggest that dispersal is not a limiting factor in The Hague’s urban environment, as current green infrastructures appear sufficient to support many species. The model overfitting and low validation performance also indicate the need to refine biodiversity indicators for urban environments, as typically used vegetation indicators do not predict species distributions well. The absence of dispersal limitations that suggest that the urban environment acts as one large meta-community, indicates that ensuring sufficient green infrastructure in the urban environment should be the first priority to enhance biodiversity in the urban environment. |
| Top | Authors |
