Abstract
<title>Abstract</title> <p>Urban ecological connectivity models are increasingly used to inform conservation decisions, but they often lack observational calibration data, particularly for estimating habitat requirements and landscape resistance. Connectivity parameters are therefore often derived in one city and transferred to another, yet systematic tests of their transferability across cities are lacking. Here, we use observations of the common blackbird in Munich (Germany) and Angers (France) to compare parameter estimates derived from a single city or from both jointly, testing whether habitat requirements and resistance values are transferable across cities and whether joint parameterisation improves predictive performance. We found that transferability of resistance values was limited to their relative order for barriers such as buildings and streets, while absolute values differed. However, resistances derived from the joint dataset moderately outperformed the Munich parameters when predicting independent Munich data, suggesting that combining data from different cities can improve resistance estimates and facilitate transferability to unknown data points. Habitat requirements, by contrast, were already well estimated from single-city datasets. Overall, our study provides evidence that resistance values can be transferred between cities when the aim is to identify the relative importance of barriers. The improved transferability gained by combining data from different cities indicates that blackbird movement is shaped by similar barriers across them. Pooling multiple datasets into the estimation of resistance values can thus substantially enhance the predictive value of connectivity modelling. Further research extending this approach to other species and cities would help establish how broadly these findings generalise.</p>