Climate change disproportionately threatens alpine species, by reducing available habitat and by isolating their populations. These pressures are particularly relevant for rear‐edge populations, which typically occupy more marginal habitat compared to populations at the core of species' ranges. We studied Caucasian grouse Lyrurus mlokosiewiczi in the Caucasus ecoregion, a global biodiversity hotspot where this species is endemic, to understand potential climate change impacts on the species. Specifically, we assessed how climate change impacts rear‐edge populations and how important these populations are for understanding range shifts and adaptive capacity under climate change. We used maximum entropy modeling to assess changes in the distribution of climatically suitable habitat under present and 2070 climate conditions for the representative concentration pathways 8.5 (RCP8.5). Our results revealed that ignoring rear‐edge populations leads to a significant underestimation of the future range (by about 14,700 km ² ). Rear‐edge populations were better adapted to warmer climates compared to core populations, and ignoring them, therefore, also underestimates adaptive capacity. Preventing the loss of rear‐edge populations should, therefore, be a priority for conservation planning in the face of climate change. Because the Caucasian grouse is associated with alpine mountain tops, conservation should focus on establishing connectivity between rear‐edge and core populations (e.g., via transboundary corridors or assisted colonizations). Our study reveals how species distribution modeling can highlight the importance of rear‐edge populations for mitigating climate change impacts on species of conservation concern.

Context: The identification of suitable habitats and the assessment of connectivity are important to preserve key areas for small isolated, endangered populations. The brown bear, Ursus arctos, needs connectivity to supply the primary habitat requirements including food, water, shelter and space and to provide gene flow among all populations in the Iranian Caucasus. Aims: In the present study, we investigated the status and habitat requirements of an endangered brown bear population within the Iranian Caucasus. Methods: We applied an approach of consensus species distribution modelling to estimate the distribution of suitable habitats for brown bears using uncorrelated environmental variables. We then used the concept of circuit theory on resultant breeding patches to evaluate regional patterns of connectivity among these patches. Key results: We predicted that ~9.10% of the study area is suitable for the brown bear at present. Ten patches (7.95% of the study area) were detected as suitable for breeding populations, where some populations are not able to survive without connectivity. The results indicated that habitat connectivity is sometimes widely affected by a high concentration of human activities such as roads, settlements and mining activities. Our findings showed that existing conservation areas could not safeguard the connectivity of brown bear habitats across the Iranian Caucasus. Conclusions: Our results can help target fine-scaled planning approaches for the maintenance of bear meta-population structure, as well as facilitate the movement of individuals by protecting different landscape features. Implications: The populations of brown bear are among the first to be harmed by the loss of habitat and connectivity, and, thus, this species is an appropriate focal species for linkage design that is beneficial for threatened populations of other co-existing species such as Persian leopard, grey wolf and Eurasian lynx. Moreover, the brown bear is among the most popular flagship species for conservation planning, which might increase public support for the restoration of habitat and linkages.

The need to obtain information on population size to inform management actions for conservation is imperative. Despite this, reliable data on Caucasian grouse abundance is scarce in Iran. The goal of this study was to explore the potential distribution of Caucasian grouse using an ensemble of small models with outstanding performance for modelling rare species’ distributions to estimate the potential population size in Iran. We fitted an ensemble of small models with generalized boosted model (GBM) and maximum entropy (MaxEnt), and then built a final ensemble prediction by averaging across these two ensembles of small models.

Micro-habitat characteristics of breeding display Sites (Leks) of Caucasian Black Grouse (Tetrao mlokosiewiczi) were studied in the mountains of Arasbaran Region, East Azerbaijan, from October 2006 to June 2008. We quantified micro-habitat characteristics around active and inactive Caucasian black grouse lek sites and random points in the 3 areas (Biosphere reserve, Kalan protected area (one of core zones of Arasbaran biosphere reserve) and unprotected areas)). Our objective was to compare micro-habitat characteristics among these sites. We also developed a logistic regression model to identify micro-habitat factors have important effect in the active lek sites. In the micro-habitat scale, we measured both biotic factors (canopy cover (%), height (cm) and frequency of vegetations) and abiotic (physical) factors (slope (%), altitude (m), aspect and etc.). Slope and altitude were different significantly between sampling sites (p = 0.05). Mean of slope in the active leks (= 57.9%) and inactive leks (= 52.0%) were higher than random sites (= 40.1%). The mean aspect differed between active leks ( $ \bar X $ = 284°, S.D = 54.1°) and random sites ( $ \bar X $ = 33°, S.D = 76.18°) (Watson’s U²-test, U _{0.05, 22, 27} ² = 0.717). Also, variance analysis (ANOVA) revealed that grasses and forbs cover, frequency of deciduous shrubs and forbs, canopy cover and frequency of deciduous woody plants > 12 cm diameter at breast height were different significantly between sampling sites. The best model for discriminating between active lek and random sites included rock-pavement cover (parameter estimate [B] = 0.286, SE = 0.219), canopy cover of grasses (parameter estimate [B] = 0.880, SE = 0.453), distance to the nearest tree (parameter estimate [B] = −0.047, SE = 0.027).