Changes in species' distribution: variations in the occupation of UTM 10x10 squares between atlases

For a particular species, the quantification of changes in its distribution can be estimated by means of an analysis of the changes in occupied UTM 10x10 squares. However, this approach generates a problem: a temporal variation in sampling effort often creates a significant bias in the estimates of distribution changes. Unfortunately, the data collected for the first atlas (Muntaner et al. 1984) was not standardised in relation to the sampling effort applied to each UTM 10x10 square. As well, despite the timed censuses conducted in the UTM 10x10 squares, a large portion of the data obtained for the new atlas comes from observations gathered in a non-standardised way.

The methodological approach used to resolve the problems originating from changes in the sampling effort consisted of indirectly estimating the effective sampling time in every UTM 10x10 square for both atlases. These estimates were then given as values to the co-variable effort in the analyses of the changes in species' distribution between atlases. This approach was made through timed censuses conducted on a sample of UTM 1x1 squares located within each of the UTM 10x10 squares. Curves of species accumulation dependent on the time spent bird surveying were drawn from these data and so it was also possible to estimate the effective surveying time for a particular species richness by reversing the process.

Distribution map (Griffon vulture). Distribution in the period 1999-2002 on a UTM 10x10 km grid. This map presents evidence of breeding in three categories: possible, probable-sure and non-reproductive summer visitors. This distribution is superimposed on the map from the 1975-1983 period (Muntaner et al. 1984). Alongside appear the number of 10x10 squares in which the species has been detected in each atlas and the trend in the number of 10x10 UTM squares occupied by the species, assessed by means of a statistical analysis that monitors the differences in sampling effort between atlases.