Optimum design in monitoring common birds and their habitat

According to recent development in landscape ecology, monitoring bird populations within defined habitat patches may not be sufficient to evaluate population changes at a larger scale. In order to develop a quick-and-easy regional bird monitoring system capable of evaluating the effects of both population and habitat changes, we planned and tested a random sampling design over a study area as wide as 23,842 km(2), corresponding to the administrative region of Lombardy (Northern Italy). A sampling design based on randomly selected point counts was drawn. The region was divided into a number of districts, each of them composed of eight census units. A breeding bird survey was then organised over a period of eight years, to cover one unit per district in one year so that, at the end of the project, a quantitative breeding atlas should also be attained. Two pilot surveys were run in 1992 and 1995, by performing respectively 417 and 749 point counts, and by detecting 24,347 and 26,619 birds (pertaining to 128 and 129 species). During the field work we also recorded habitat and environmental variables, in order to relate trends to habitat changes. After data standardisation, we obtained population indices of 32 common species. We validated our indexing system by analysing changes in population levels at sub-regional and "between habitats" scales