Summary of the best-fit multi-scalar models for measures of bat activity and presence.

‡Indicates data extracted using simple concentric circular buffers, otherwise, a connectivity mask (a 50 m buffer around tree networks) was used.
* = P<0.05,
** = P<0.01,
*** = P<0.001.
The most parsimonious model for each response variable is shown, according to Akaike Information Criterion (AIC) values. PP indicates P. pipistrellus, PPy P. pygmaeus, NSL a group comprising N. noctula, E. serotinus, N. leisleri and My indicating M. daubentonii. Models whose AIC value≤2 of the optimum model are also included. Model type is either a generalised linear model (GLM) or a generalised additive model (GAM) with either a negative binomial (Ne) distribution for activity data or binomial (Bi) for presence. See Table 2 for variable definitions.