Land-cover patterns surrounding Caucasian grouse leks in Arasbaran region, East Azerbaijan, Iran

Abstract

Context To create management strategies with the goal of sustaining a species such as Caucasian grouse (Lyrurus mlokosiewiczi), it is important to identify the habitat requirements of species, not just in terms of a correlation with a given habitat feature, but also the relationship between species presence and vegetation coverage, proximity to other habitat types, and importance at different spatial scales. Aims To predict the proportions and spatial configuration of major habitat types that are associated with high probabilities of Caucasian grouse lek occurrence. Methods Using minimum mapping-unit scale (i.e. grain) for land cover, we applied spatial analysis at three spatial extents (472-, 702- and 867-m-radius circles) to assess how the importance of different land-cover patterns and patch characteristics surrounding leks of Caucasian grouse changed with scale within the Arasbaran landscape (316.56 km2) in East Azerbaijan, Iran. A set of a priori models has been developed on the basis of landscape metrics linked to hypotheses that could explain the spatial pattern of Caucasian black habitat use at each scale. We used an information-theoretic approach based on Akaike’s information criterion (AIC) within a general additive models framework to model habitat selection, so as to compare the values of landscape metrics calculated for Caucasian grouse lek sites (n = 22) with those calculated for non-lek points (n = 44). Key results The probability of lek occurrence at each of the spatial scales increases with a larger amount of open, young forests in the landscape. At each scale, we could indicate the landscape composition and structure required to create an ideal habitat mosaic for Caucasian grouse. Such an ideal habitat mosaic within mountain forests of Arasbaran, for a 702-m-radius area around a potential lek site, would consist of non-square (i.e. more geometrically complex) patches of rangeland cover and deciduous stands with canopy cover of <50%, which encompass over 30% of landscape. Conclusions Our results identified differences in black grouse requirements at several scales within the landscape. We believe this will help managers improve the habitat focusing on the area around existing or inactive leks, to adapt the landscape to species requirements, and to encourage targeting new sites. Implications These findings demonstrated that not only can we identify important landscape requirements at a range of scales, but by characterising landscape composition and structure across these scales, forest managers can help prioritise combinations of habitats that best serve the conservation of the target species.

Differential habitat associations in peripheral populations of threatened species: The case of the Caucasian grouse

“…Even though Habibzadeh and Rafieyan () clarified that the CG occurrence probability increases with the large amount of open and young forests in Iranian CG range, our models in this study only suggested the CG requirements for grassland cover at 2‐km scale in LCse region. These differences might be due to the existence of cross‐correlation and redundancy issues among habitat variables that entered the models, the research scales or modeling techniques.…”

“…These differences might be due to the existence of cross‐correlation and redundancy issues among habitat variables that entered the models, the research scales or modeling techniques. Habibzadeh and Rafieyan () used a priori generalized linear modeling to discriminate the response of CG to only land‐cover variables at fine (30‐m) resolution and local extent (Iran). However, our study's variables involved land cover, topographic and climatic ones at resolutions of 300, 90 and 1,000‐m, respectively, over regional scale extent.…”

“…The CG distribution of Iran, Azerbaijan and Armenia was included in LCse region (Figure ). We used species occurrences provided by previous studies (Gavashelishvili & Javakhishvili, ; Habibzadeh & Rafieyan, ; Isfendiyaroğlu, Welch, & Ataol, ) as CG presence data, which were based on field surveys in various parts of Georgia, Russia, Armenia and Turkey (Gavashelishvili & Javakhishvili, ; Isfendiyaroğlu et al, ) in 2004–2005, and from Iran (Habibzadeh & Rafieyan, ) in 2013–2014, as well as 200 radio‐telemetry presence locations captured from eight adult males and six adult females in Georgia's Greater and Lesser Caucasus regions during 2004–2006 (Gavashelishvili & Javakhishvili, ). After removal of duplicate records, the total number of presence locations was 564 (218 GCw, 185 GCe, 100 LCnw and 61 LCse).…”

“…To describe environmental conditions of CG habitat, we calculated land cover, topographic and climatic variables according to their acknowledged relevance for the ecology of the species (Gavashelishvili & Javakhishvili, ; Gottschalk, Ekschmitt, Isfendiyaroglu, Gem, & Wolters, ; Habibzadeh et al, ; Habibzadeh, Karami, & Tarinejad, ; Habibzadeh & Rafieyan, ). We used the Climate Change Initiative‐Land Cover (CCI‐LC) data version 1.6 provided by the European Space Agency (ESA) at ~300˗m spatial resolution for 2010 epoch, which contained 22 primary and 16 sublevel land‐cover classes (ESA, ).…”

“…CG habitat is highly fragmented (Gavashelishvili & Javakhishvili, ), both naturally and due to habitat loss and conversion, such as intensive grazing (Etzold, ; Storch, ). The species is found on steep slopes of subalpine and alpine meadows covered by widely scattered dwarf shrubs (e.g., rhododendron thickets [ Rhododendron caucasicum ]) but in proximity to deciduous broad‐leaf forest (Gavashelishvili & Javakhishvili, ; Habibzadeh & Rafieyan, ; Klaus et al, ; Klaus, Bergmann, Wiesner, Etzold, & Vitovich, ). Subalpine coppice formations such as birch ( Betula litwinowii ), oak ( Quercus macranthera ), beech ( Fagus orientalis ), juniper ( Juniperus communis ) and rose ( Rosa spp.)…”