Mountainous regions are regarded as significant biodiversity hotspots, offering a diverse range of vital ecosystem services to the communities who reside in them and the surrounding plains. The black grouse (Lyrurus tetrix), a galliform species emblematic of the European Alps, is currently threatened by habitat change. The estimation of population dynamics, and in particular the reproductive success of this species, represents a significant challenge for the long-term conservation of the black grouse. In this study, we attempted to map black grouse brood habitat suitability (BHS) at the scale of an Alpine bioregion, coupling a species distribution model (SDM) with multi-source remote sensing data. To extract landscape composition features likely to influence BHS, convolutional neural networks (CNNs) were employed to characterise very high spatial resolution (VHSR) SPOT6-7 imagery. Altitude, phenological indices derived from Sentinel-2 time series (NDVImax, NDWI1max), and a texture feature derived from the SPOT6-7 images (Haralick entropy) were used to refine landscape characterisation. An SDM based on a random forest ensemble model was used to map black grouse BHS. Altitude, ericaceous heathland and NDVImax emerged as the three most significant variables, consistent with the ecological needs of black grouse. The proportion of ericaceous heathland was especially representative of the foraging needs of female black grouse; the main ecological determinant of habitat suitability for brood rearing with sufficient vegetation cover. This study highlights the effectiveness of integrating VHSR and multispectral time series, together with the advantages offered by deep learning techniques, in extracting species-specific information tailored to conservation issues. The BHS map, produced on a regional scale, constitutes a significant advances in the monitoring of the current population dynamics of black grouse at the western limit of the species’ geographical distribution.

Sensitivity of lidar metrics to scan angle can affect the robustness of area-based approach (ABA) models, and modelling the interplay of scan geometry and terrain properties can be complex. The study hypothesises that neural networks can manage the interplay of lidar acquisition parameters, terrain properties, and vegetation characteristics to improve ABA models. The study area is in Massif des Bauges Natural Regional Park, eastern France, comprising 291 field plots in a mountainous environment with broadleaf, coniferous, and mixed forest types. Field plots were scanned with a high overlap from multiple flight lines and the corresponding point clouds were considered independently to expand the standard ABA dataset (291 observations) to create a dataset containing 1095 independent observations. Computation of lidar, terrain and scan metrics for each point cloud associated each observation in the expanded dataset with the scan information in addition to the lidar and terrain information. A multilayer perceptron (MLP) was used to model basal area and total volume to compare the predictions resulting from standard and expanded ABA datasets. With expanded datasets containing lidar, terrain and scan information, the R² for the median predictions per plot were higher (R² of 0.83 and 0.85 for BA and Vtot) than predictions with standard datasets(R² of 0.66(BA) and 0.71(Vtot)) containing only lidar metrics. It also outperformed an MLP model for a dataset with lidar and terrain information (R² of 0.77 (BA and Vtot)). The MLP performed better than RF regression, which could not sufficiently exploit additional terrain and scan information.