Figure 2

Mean relative bias (dots/ triangles), together with 2.5% and 97.5% quantiles (solid lines) in the estimated (a) detection parameters \({\theta }_{det}\), (b) abundances \(\lambda\), (c) site-use frequencies \({\lambda }_{use}\) and (d) trends in \(\lambda\). Results are displayed for all combinations of model used (BernP, BP, PP), the submodel and covariate structure for \({\theta }_{det}\) (constant, spatial variation (naïve model), spatial variation (covariate model)), population size N (120, 240), \(\rho\) (0.7: dots, 0.95: triangles) for \(\sigma\) = 300 (emerging home range area of 2.61 km²). Line of equality (dotted line). Average relative bias for each HM (dashed line). For visual clarity, x-scales are different for the subpanels in (b) and (c).