Biogeography & Spatial Ecology

Olive fly (Bactrocera oleae) is widely recognized as an important insect pest affecting olive cultivation in Mediterranean agroecosystems, with significant consequences for crop yield and oil quality. Existing management strategies primarily focus on population dynamics within olive groves, while the potential influence of adjacent non-crop habitats on pest ecology remains understudied. In the context of increasing temperatures and more frequent heat events associated with climate change, the role of landscape heterogeneity and microclimatic variability in shaping olive fly movement and persistence warrants further investigation. This study examined the occurrence of B. oleae in olive groves and adjacent forest habitats within a Mediterranean mosaic landscape on Lesvos Island, Greece. Baited McPhail traps were deployed along a spatial gradient extending from grove interiors to forest interiors (up to 300 m), while microclimatic conditions were monitored using fine-scale temperature measurements during the summer and autumn periods. Adult olive flies were consistently detected within olive groves, along grove–forest interfaces, and at distances up to 100 m inside pine forests, whereas captures at 300 m were minimal. Forest habitats exhibited lower temperatures than olive groves, and generally had conditions considered suitable for olive fly activity and survival. In addition, olive fly abundance near grove edges and forest-adjacent locations showed a negative association with temperature, suggesting increased utilization of cooler habitats during periods of elevated thermal stress. These findings indicate that adjacent forest habitats may provide temporary microclimatic shelter for olive flies and facilitate limited cross-habitat movement under high-temperature conditions. The study further demonstrates the ecological relevance of surrounding non-agricultural habitats in olive fly dynamics and highlights the importance of incorporating landscape context and microclimatic variation into integrated pest management approaches under future climatic conditions.

Mediterranean islands are increasingly experiencing ecological restructuring driven by the expansion of introduced mammalian herbivores, whose establishment within fragmented insular landscapes may generate profound consequences for vegetation dynamics, agricultural systems, and ecosystem functioning. The European rabbit, Oryctolagus cuniculus, has exhibited substantial spatial expansion on Lemnos Island, Greece, where its growing distribution across heterogeneous agroecosystems and semi-natural habitats has raised increasing ecological and management concerns. Here, we applied a species distribution modelling framework based on Maximum Entropy algorithms to investigate habitat suitability patterns of O. cuniculus using an extensive occurrence dataset derived from systematic field surveys. Environmental predictors included land-use composition, agricultural cover, vegetation structure, topographic variables, climatic conditions, and anthropogenic landscape features in order to identify the ecological factors associated with rabbit occurrence. Model outputs demonstrated strong predictive performance and revealed substantial spatial heterogeneity in habitat suitability across the island. Habitat suitability was strongly associated with agricultural landscapes, cultivated areas, agro-pastoral systems, and lowland environments providing abundant trophic resources and accessible foraging opportunities. Areas characterized by crop availability and heterogeneous resource-rich mosaics exhibited particularly high predicted suitability, indicating the importance of anthropogenic food subsidies in facilitating rabbit persistence and expansion. Spatial projections identified extensive environmentally suitable areas across large portions of the island, suggesting that current land-use configurations may facilitate continued expansion of rabbit populations within Mediterranean insular ecosystems. The predicted distributional patterns indicate that anthropogenic resource availability rather than climatic limitation may constitute the primary driver shaping rabbit habitat suitability at the spatial scale examined. These findings highlight the critical role of agricultural landscape structure in promoting the establishment and persistence of introduced herbivorous mammals and provide a spatially explicit framework for future monitoring and management strategies in insular Mediterranean environments.

Linear transportation infrastructure increasingly reshapes ecological processes by generating spatially heterogeneous mortality risks across landscapes. Despite growing recognition of wildlife–vehicle collisions as a major conservation concern, spatially explicit assessments of road mortality in Mediterranean herpetofauna remain limited, particularly at large temporal and landscape scales. This study investigates the spatial configuration, clustering dynamics, and seasonal concentration of amphibian and reptile road mortality within the catchment basin of Spercheios River, a protected Mediterranean ecosystem of high conservation importance included in the Natura 2000 network. Using a long-term dataset (1997–2021), we analyzed 544 wildlife–vehicle collision records involving 23 herpetofaunal species to identify spatial mortality concentrations and landscape-scale collision patterns. Spatial analytical frameworks incorporating Kernel Density Estimation and hotspot analysis were employed to detect non-random aggregations of mortality events and delineate persistent collision corridors. Species-specific road mortality intensity was additionally quantified through a relative mortality index to evaluate differential vulnerability among taxa. Analyses revealed pronounced spatial autocorrelation in road mortality distribution, with distinct mortality hotspots concentrated along ecotonal zones linking wetlands, agricultural mosaics, and riparian habitats. Several statistically significant collision clusters persisted across years, indicating the existence of stable mortality corridors associated with recurrent movement pathways of herpetofauna. Seasonal analysis demonstrated a marked intensification of hotspot formation during spring, suggesting strong associations between road mortality patterns and seasonal behavioral ecology. Mortality intensity was disproportionately higher for large-bodied snake species, particularly the Eastern Montpellier snake, Grass snake, Caspian whipsnake, and the European glass lizard. Our findings demonstrate that herpetofauna road mortality within Mediterranean riverine systems is spatially structured rather than randomly distributed, emphasizing the importance of spatial statistical approaches for identifying high-risk road segments and prioritizing mitigation measures. Integrating hotspot detection and density-based spatial analyses provides a robust framework for conservation-oriented road planning and for reducing biodiversity losses in ecologically sensitive Mediterranean landscapes.

Wildlife–vehicle collisions (WVCs) constitute a spatially structured form of anthropogenic mortality that can disproportionately affect insular mammal populations, particularly within Mediterranean landscapes characterized by ecological heterogeneity and fragmented habitat mosaics. Despite increasing concern regarding road-driven biodiversity loss, the spatial organization and temporal persistence of mammalian road mortality in island ecosystems remain poorly understood. This study examines the spatial ecology of WVCs on Lesvos through a long-term survey conducted across the island's 488.7 km primary road network between 2018 and 2024. Surveys focused on five terrestrial mammalian species: Red Fox, Beech Marten, Northern White-breasted Hedgehog, Least Weasel, and Persian Squirrel. To investigate the spatial structure of mortality events, we employed an integrated spatial-statistical framework combining Kernel Density Estimation (KDE), Getis–Ord Gi* hotspot analysis, and Anselin Local Moran’s I statistics. These approaches enabled the detection of persistent mortality hotspots, statistically significant spatial clusters, and localized outlier events across the transportation network. The analyses revealed strong spatial dependence in WVC occurrence, with recurrent hotspot concentrations primarily distributed in the central and southeastern sectors of the island. Several hotspot zones persisted through multiple survey years, suggesting the existence of chronic collision-risk corridors shaped by road configuration and surrounding landscape structure. Species-specific spatial patterns reflected ecological and behavioral differentiation among taxa. Mortality records of Persian Squirrel were strongly clustered near traditional olive grove systems, whereas Red Fox and Beech Marten exhibited broader and more spatially diffuse mortality distributions. By integrating multi-year monitoring with advanced spatial statistics, this study provides a geographically explicit framework for identifying collision-risk landscapes and informing spatially targeted mitigation planning in Mediterranean insular ecosystems.

Subterranean burrow systems created by fossorial mammals represent concealed forms of biological architecture capable of influencing environmental conditions beyond the underground spaces they occupy. Despite their ecological importance, the extent to which these hidden structures generate organized thermal patterns at the land surface remains poorly understood. In the European rabbit, warrens form spatially complex subterranean networks whose ecological expression is typically inferred from the distribution of visible entrances alone, potentially overlooking thermal patterns associated with the underlying burrow system. To address this gap, we combined infrared thermography with geospatial analyses to investigate whether rabbit warrens generate structured thermal mosaics detectable at fine spatial scales. Thermal imagery was collected from multiple independent warrens and converted into high-resolution thermal surfaces. Spatial analyses were subsequently applied to quantify thermal clustering, local heterogeneity, and the spatial configuration of temperature anomalies surrounding burrow entrances and across entire warrens. In parallel, entrance density and spatial arrangement were used as indicators of subterranean structural complexity in order to examine whether variation in surface thermal organization reflects differences in hidden burrow architecture. Our analyses revealed that rabbit warrens generate distinct and spatially structured thermal mosaics rather than isolated thermal anomalies. Burrow entrances consistently corresponded to localized thermal hotspots relative to the surrounding substrate; however, these anomalies were embedded within broader heterogeneous thermal fields characterized by clear spatial clustering and non-random thermal organization. Furthermore, warrens exhibiting greater entrance density and more complex entrance configuration displayed increasingly pronounced thermal heterogeneity and stronger spatial structuring of surface temperatures. These findings suggest that surface thermal patterning captures important aspects of the organization of subterranean burrow systems. Overall, our study demonstrates that infrared thermography integrated with geospatial analysis provides a powerful non-invasive framework for detecting the thermal imprint of subterranean ecosystem engineering and advances the spatial understanding of animal-generated microclimatic heterogeneity.

Climate-driven environmental change is increasingly altering the spatial dynamics of biological invasions, particularly for species tightly associated with freshwater systems and human-modified landscapes. The semi-aquatic rodent Myocastor coypus has expanded rapidly across parts of southern Europe, where its burrowing activity, herbivory, and occupation of riparian habitats generate substantial ecological and economic impacts. In Greece, however, the potential for large- scale range expansion under future climatic conditions remains insufficiently explored despite growing evidence of establishment in lowland wetlands and drainage networks. A spatial modelling framework based on Maximum Entropy algorithms was employed to reconstruct the environmental niche of M. coypus using a nationwide occurrence dataset combined with climatic, hydrological, edaphic, and land-cover predictors. Variable selection focused on hydroclimatic stability, thermal tolerance, moisture availability, and landscape connectivity in order to capture the processes governing dispersal across heterogeneous Mediterranean environments. Model performance indicated strong predictive capacity, revealing that hydrological proximity and persistent surface moisture overwhelmingly structure habitat suitability, while warmer thermal regimes substantially reduce climatic barriers to establishment. Spatial projections under multiple future climate trajectories revealed a marked reorganization of invasion-prone landscapes during the twenty-first century. Environmentally suitable areas are projected to expand far beyond the species’ currently documented distribution, particularly along riparian corridors, irrigated agricultural plains, artificial water networks, and ecologically sensitive wetland systems. Several projected expansion zones intersect with protected habitats and biodiversity-rich aquatic ecosystems, indicating an increasing likelihood of ecological conflict between invasive spread and conservation priorities. The emerging spatial configuration additionally suggests the formation of interconnected dispersal pathways capable of accelerating secondary colonization between previously isolated hydrographic basins. The projected redistribution of suitable habitat indicates that ongoing hydroclimatic change may transform Greece into an increasingly favorable landscape for the persistence and spread of this invasive mammal, reinforcing the urgency for spatially targeted surveillance and anticipatory management strategies.

Understanding the mechanisms governing species distributions is fundamental for predicting biodiversity responses to environmental change, particularly within insular ecosystems where spatial constraints and landscape heterogeneity strongly influence ecological processes. We investigated the habitat suitability patterns of the Persian squirrel (Sciurus anomalus), a regionally Vulnerable arboreal mammal, on Lesvos Island, Greece, using a species distribution modeling (SDM) framework based on the MaxEnt algorithm. To evaluate the relative influence of landscape and climatic drivers, we developed two competing SDMs: an environmental model including landscape composition and anthropogenic variables, and a second model integrating environmental predictors with bioclimatic variables associated with temperature and precipitation regimes. Both models produced high predictive performance, identifying clear spatial gradients in habitat suitability across the island. Suitable habitats were strongly concentrated in areas characterized by heterogeneous vegetation structure and persistent semi-natural landscape mosaics, whereas intensively modified or spatially homogeneous areas exhibited substantially lower predicted suitability. Jackknife analyses revealed that landscape-related variables contributed disproportionately to model gain and predictive stability, while the addition of climatic predictors resulted in only marginal improvements in explanatory power. Response curves further indicated that habitat suitability declined sharply beyond specific thresholds of landscape simplification and increasing isolation from structurally complex habitats. The spatial predictions also revealed a fragmented distribution of highly suitable habitat patches, suggesting potential restrictions in landscape connectivity for insular populations of S. anomalus. In contrast, climatic gradients showed weak spatial correspondence with predicted suitability patterns, indicating limited macroclimatic control at the scale examined. Our findings demonstrate that landscape complexity and habitat configuration are stronger determinants of Persian squirrel distribution than broad-scale climatic conditions within Mediterranean island ecosystems, highlighting the importance of integrating fine-resolution landscape predictors into SDMs for insular mammals and underscores the conservation value of maintaining heterogeneous and structurally connected landscapes for the long-term persistence of vulnerable arboreal species.