1 INTRODUCTION

Mediterranean ecosystems harbour diverse herpetofauna, representing ecologically distinct communities with contrasting evolutionary histories that originate from various biogeographic regions (Salvador, 2014; Werner, 2016). Some species are microendemic and highly vulnerable to processes that rapidly alter habitat composition (Monasterio et al., 2010). These vulnerabilities, coupled with centuries of human-driven landscape change and harmful side effects, such as climate change and the increasing presence of alien species (Genovesi, 2005; Winter et al., 2016), make identifying the environmental factors that negatively impact reptile diversity in the Mediterranean region a priority (Cox et al., 2006).

Habitats in this region range in light availability from full sun on rocky outcrops with minimal obstruction to almost complete shade in mature forests (Gallaher et al., 2019). Squamates living at opposite ends of this light diffusion spectrum have developed distinct morphological and physiological characteristics. Species that occupy sparsely vegetated habitats face rapid overheating and increased detection, leading to adaptions for high-speed running and inconspicuous colouration (Chen et al., 2013; Kohlsdorf et al., 2001). In contrast, squamates in forest habitats demonstrate traits such as the ability to climb on vertical surfaces, viviparity and the capacity to function at lower body temperatures while remaining active (Higham et al., 2011; Lynch, 2009; Meek, 2005).

Many Mediterranean squamate species display broad habitat tolerance, using areas ranging from scattered shrubland to open forests, but are rarely found in habitats at either extreme (Castilla & Bauwens, 1992; Reading & Jofré, 2009; Salvador, 2014). This suggests that dense forests may not be optimal for many squamate species that thrive in Mediterranean temperate ecosystems, and where they do occur in these habitats, they are likely confined to open gaps or discontinuities (Delgado-García et al., 2007; Meek, 2014).

Forest cover in southern Europe has gradually expanded in recent decades due to the decline of traditional agriculture (Bracchetti et al., 2012; Mantero et al., 2020; Otero et al., 2015). Part of this land has been used for intensive tree plantation of fast-growing species such as Eucalyptus spp., Pinus radiata, Platanus x hispanica, and Populus spp. (981.406 ha in Spain, 2020; https://www.miteco.gob.es). Frequently, these reforestation efforts have not prioritised the preservation of local biodiversity or the protection of threatened species (Vadell et al., 2022). As a result, rapid reforestation poses a significant conservation concern for squamate diversity, potentially leading to population declines and even local extinctions, as seen in other parts of the world (Goodman, 2009; Kellermann et al., 2021).

In this study, we evaluated the hypothesis that densely canopied habitats with greater stand height (i.e., trunk height) are unfavourable for most Mediterranean reptiles. Lower light levels in such forests reduce the efficiency of basking behaviour required for thermoregulation (Suzaki et al., 2003). This hypothesis was tested in the order Squamata (amphisbaenians, lizards and snakes), a monophyletic clade that includes most of the reptile species present in southwestern Europe (Pyron et al., 2013).

2 MATERIALS AND METHODS

2.1 Study region

The study region spanned the Iberian Peninsula, southeastern France, the Italian Peninsula and Sicily (Figure 1). This part of Europe is notable for its high diversity of squamate species compared to other temperate regions of the Palearctic (Araújo et al., 2006). However, extensive human activity has substantially reshaped this landscape over time (Falcucci et al., 2007). The main land cover in this region is a heterogeneous mosaic composed of agricultural fields, urbanised areas and fragmented forests (Virgós et al., 2002).

Data collection was carried out each year from March to October between 2014 and 2023, aligning with the seasonal activity periods of the target species (Escoriza et al., 2021, 2023). Most species were encountered opportunistically during random surveys, with habitat selection determined by accessibility rather than a pre-defined sampling strategy. However, an exception was made for microendemic lizard species, which were specifically targeted within their known distribution ranges. Fieldwork teams, consisting of two to four surveyors, characterised 3 to 6 localities per day during 8-h sampling periods. Squamates were detected through visual transects carried out during daylight hours and by lifting large objects such as stones and logs (McDiarmid et al., 2012). In total, 960 sites were assessed across an elevation range of 1 to 2553 metres (Figure 1).

3 RESULTS

A total of 56 squamate species were recorded during sampling (2 species of amphisbaenians, 37 species of lizards and 17 species of snakes) (Figure 2 and Supporting Information file II). No forest-obligate species were identified, although one snake, Zamenis longissimus, was more commonly found in habitats with canopy density over 50% (present at more than 75% of sites). Two other species were predominantly found in habitats with over 20% canopy cover, also appearing at over 75% of sites: Chamaeleo chamaeleon (88%) and Vipera aspis (88%). Although most species did not show strong preferences for specific habitat types (Figure 3), some were predominantly associated with sparsely wooded environments featuring 5–19% canopy cover. This group mainly comprised cursorial (Acanthodactylus erythrurus, Psammodromus edwarsianus) and xeric specialists (Timon nevadensis) (see Figures 3 and 4).

The first two axes of the OMI explained 89.69% of the total inertia (Table 2). The first axis (inertia = 68.15%) separated the sites based on their abiotic conditions, following a gradient determined by opposing temperature and precipitation vectors (Figure 4). This axis differentiated Mediterranean thermophilic species from subalpine temperate species (Figure 4). The second axis (inertia = 21.54%) separated the sites based on plant community diversity, distinguishing squamate species that occupy habitats with dense and complex vegetative communities (e.g., native forests) from those that inhabit compositionally simpler habitats, such as steppes and subalpine meadows (Figure 4).

The WitOMI results indicated a reduction in niche sizes in habitats with denser canopies (Figure 5). Medium forest represented 30.25% of the total area, low forests 34.00%, medium woodland 44.81%, low woodland 68.83%, other woody vegetation 80.55%, and non-woody vegetation 83.14% (Figure 5). WitOMI analysis revealed significant differences between these forest categories in terms of habitat properties: forests and woodland supported richer plant communities than other woody or non-woody vegetation (Table 3).

The WitOMI results also showed that habitats with denser canopies support lower squamate species richness and lower phylogenetic diversity (Table 4). For example, medium forests included 24 species, whereas low woodland supported 40 species (Table 4). Furthermore, species occupying forests showed greater marginality (WitOMI = 80.6‒85.3) and lower tolerance (6.9‒7.4) compared to species occupying non-forest habitats (other woody/non-woody vegetation; WitOMI = 50.0‒65.8; tolerance = 9.4‒11.4) (Table 4). Residual tolerance was consistently low (7.8‒38.6) (Table 4), indicating that predictor variables accounted for a substantial part of the environmental variance between species. The OMI parameters and their statistical significance for each vegetation category are provided in Supporting Information file III.

4 DISCUSSION

For thousands of years, human activities have shaped the ecosystems of the Mediterranean region. Traditional agriculture maintained a patchwork of habitats that supported highly diverse biotic communities (Baiamonte et al., 2015). However, industrialisation intensified forest exploitation, introduced nonnative tree species and homogenised open areas through urbanisation and intensive agriculture, leading to biodiversity loss (Brotons et al., 2018; Henle et al., 2008). Additionally, the disuse of open areas due to reduced grazing has allowed forests to reclaim land, negatively impacting species reliant on open habitats (Preiss et al., 1997; Stefanescu et al., 2011; Suárez-Seoane et al., 2002), potentially including reptiles. These trends highlight the long-term, complex effects of human activity on Mediterranean biodiversity.

In this study, we tested the hypothesis that squamate reptiles would show negative responses to denser crown canopies and greater vegetation stand height. Our results confirmed that canopy cover exerts a strong influence on squamate presence. We found that open woodland formations and open habitats with scattered trees (canopy cover between 5 and 49%) support higher species richness and phylogenetic diversity. These ecotones provide suitable environments for both generalists and specialists alike, as cursorial forms benefit from open patches, while mesic species find protection from temperature extremes within vegetation cover (Gregory & Tuttle, 2016; House & Spellerberg, 1983; Salvador, 2014). Furthermore, similar findings from transitional woodland in southern Greece (Ioannidis et al., 2008) suggest that these results may apply across the entire Mediterranean ecoregion.

We also observed that forest habitat quality measures (such as stand height, crown density and plant community diversity) are inversely related to squamate diversity. In contrast, other temperate faunal groups such as salamanders, bats, birds and insects often prefer mature primary forests, which support greater biodiversity and provide crucial habitats for rare species (Basile et al., 2021; Curtis & Taylor, 2004; Ford et al., 2002; Piksa et al., 2022; van Halder et al., 2015).

These findings align with the observation that forest canopy cover can have a significant negative impact on habitat quality for heliothermic reptiles (Thompson et al., 2018; Webb et al., 2005). Reduced sunlight exposure forces adults to exert more effort to regulate body temperature, increasing their exposure to predators and decreasing embryo survival rates (Downes, 2001; Schlaepfer, 2003). Climbing species, such as Chamaeleo chamaeleon, can thermoregulate by using the elevated and potentially sun-exposed branches of trees and shrubs (Hódar et al., 2000), while species with lower thermal requirements, like Zamenis longissimus, may have an advantage in shaded environments (Rugiero et al., 2002). However, a substantial number of Mediterranean squamates either avoid densely shaded habitats or use them by favouring discontinuities such as rock outcrops, paths or forest edges (Amo et al., 2007).

Human management of habitats can enhance squamate diversity, as demonstrated by traditional agro-pastoral oak woodland pastures, which support more species than homogeneous native forests due to a dense understory of low shrubs (Martı́n and Lopez, 2002). Conversely, areas subjected to intensive agricultural practices and aggressive reforestation efforts show a marked decrease in reptile species, likely due to the combined effects of reduced availability of refuges and food resources, as well as the widespread use of toxic chemical agents (Kazes et al., 2020; Luiselli et al., 2018; Rugiero et al., 2021).

5 CONCLUSIONS

In Europe, forests play a critical role in nature conservation by sequestering atmospheric carbon dioxide and supporting specialised biotic communities in mature forests (Nabuurs et al., 2015; Plath et al., 2024; Vergara-Tabares et al., 2018). However, maintaining a balance between forests and open natural habitats would significantly benefit biodiversity. Our results indicated that species richness peaks in open woody formations with crown canopy densities below 50%. These findings have important implications for the conservation of Mediterranean squamates. Intensive reforestation campaigns should be avoided in areas particularly rich in reptile species, such as nature reserves, Mediterranean coastal mountain ranges (Balletto et al., 2010; Lescure & de Massary, 2012) or areas with high conservation value for threatened species.

When considering strategic reforestation programmes that address soil conservation and long-term timber use within a sustainable framework, trees should be spaced appropriately to allow sunlight to filter through to the ground, or plantations should be intersected with wide gaps (Bragg et al., 2005). A promising approach could involve novel, landscape-scale reforestation strategies that integrate both woodland and open habitats (Douglas et al., 2020).

AUTHOR CONTRIBUTIONS

Daniel Escoriza: Conceptualization; formal analysis; investigation; resources; supervision; writing the original draft. Félix Amat: investigation; writing the original draft.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.