Source populations in the context of dispersal modelling
In conservation ecology, missing source populations are often considered the main bottleneck for the successful (re-)colonisation of (restored) habitats ([e.g. Kail et al. 2015](http://dx.doi.org/ 10.1371/journal.pone.0130228)). In theoretical ecology and population biology, sources as opposed to sinks are defined as good quality habitats with a positive net difference between immigrating and emigrating individuals (demographic surplus, [Dias 1996] (http://dx.doi.org/10.1016/0169-5347(96)10037-9)). This definition differs from source populations used for dispersal modelling (e.g. Radinger et al. 2014) that typically builds on initial starting (source) populations for e.g. diffusion processes. Here, each site with at least one specimen present that might possibly emigrate can be considered as a potential source population. The presence of species and, hence, the location of initial starting (source) populations for dispersal and (re-)colonization can be either based on (i) sampling data or on (ii) species distribution models (empirical models of species presence based on explanatory environmental variables) ([Kail et al. 2015]( http://dx.doi.org/ 10.1371/journal.pone.0130228)). However, while the first probably underestimates the number of sources since only a limited number of sites is typically sampled, the latter is probably overestimating the number of sources as suitable habitat might not be directly translated into species presence. Hence, future research should focus on appropriate methods to identify / locate source populations that can be used in dispersal modelling.