Aim. When studying the effects of global change on biodiversity, it is far more common for the effects of climate change and land-use changes to be assessed separately rather than jointly. However, the effects of land-use changes in recent decades on species richness in areas affected by climate change have been less studied. We assess the temporal turnover in species richness of an avian community between a historical period and a modern one as a consequence of global change.

Location. Semiarid Mediterranean ecosystem (Southeastern Spain).

Method. We fitted a hierarchical multi-species occupancy model for each period (1991-1992, and 2012-2017), obtaining avian species-specific estimates of occupancy probability in relation to environmental covariates (elevation and forest cover). We analyze the relationships between changes in the bird community and environmental variables, analysing the temporal turnover of the species richness and the richness-based species-exchange ratio.

Results. The estimated species richness accounting for detectability was higher than observed species richness, and decreased in the more recent period. Following our hypotheses, we observed a dual pattern of species richness increase associated with different elevations, showing different species turnover rates due to the joint effects of climate change and land-use change. There is a trend toward greater species richness with higher elevations that is associated with climate change, where the species turnover rate is low. Also, species richness increased towards lower elevations, but with a high turnover rate. The latter can be due to species expansions throughout new habitat configurations in bordering forest systems associated with anthropic land-use changes.

Conclusions. Our study is of great interest to understand the temporal turnover of avian species richness associated with areas experiencing both climate and land-use change.

This field work was developed as a basis for assessing the state of the region's forest heritage and the effects of global change on its biodiversity, concentrated in two periods: a "historical period" (years 1991 and 1992) and a "modern period" (years 2012 and 2017). The survey protocol was characterized by transects of 1 kilometer length, which were distributed across all the forested areas in the study area in a random fashion and were surveyed once during a breeding period (May-July). Each transect survey was conducted by walking and recording the number of each species detected (by sight or song/call) along it. A total of 377 1-km transects were conducted for the historical period as part of a forestry plan in the region, developing a new avian survey of 198 transects with a similar protocol for the modern period.

We considered the 1-km transects of each bird survey as nested within a "site", i.e., as spatial observations replicated within the site, thus enabling estimation of detection probability in an occupancy model. We defined sites as cells in a grid that covers our study area. We selected cell size to be 2x2 km, based on the species in our bird community (most of them passerines with small home ranges), which may be suitable for evaluating the effects of environmental covariates on bird species occupancy at a regional scale. Therefore, a total of 377 and 198 1-km spatial transects were conducted and grouped in 226 and 139 sites for the historical and modern periods, respectively.