Multiscale patterns of functional connectivity across a Mediterranean periurbanized landscape in Timon lepidus

Multiscale patterns of functional connectivity across a Mediterranean periurbanized landscape in Timon lepidus | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Multiscale patterns of functional connectivity across a Mediterranean periurbanized landscape in Timon lepidus Johan Ludot, Véronique Arnal, Benoit Charrasse, Aurélie Coulon This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9550783/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Functional connectivity plays a key role in shaping gene flow and population persistence through human-modified environments, yet its drivers are understudied at spatial scales relevant to periurban planning.Using a landscape genetics framework, we investigated the multiscale effects of landscape composition and configuration on gene flow in the ocellated lizard ( Timon lepidus ), a cavity-dependent reptile inhabiting a Mediterranean periurban landscape. We combined individual-based genetic data from 202 genotyped individuals with resistance surface optimization to assess how environmental features influence functional connectivity across two nested spatial scales: a continuous periurban landscape and a broader surrounding rural matrix. Genetic structure was weak and continuous at both scales, with no evidence of sharp genetic discontinuities, and spatial autocorrelation analyses revealed limited dispersal, coherent with a gradual isolation-by-distance (IBD) pattern. Resistance-based models however consistently outperformed IBD models, indicating that landscape features influence gene flow despite the overall weak genetic structuring. At the broader scale, rivers and agricultural cover emerged as the main contributors to resistance, while the availability of natural shelters and steep slopes facilitated connectivity. At the periurban scale, urban land cover and road infrastructures did not markedly restrict gene flow. Our findings indicate that ocellated lizards’ weak genetic structure across spatial scales reflects historically high connectivity. Although gene flow is influenced by landscape resistance, its overall effect is modest and suggests that moderate urbanization, if associated with the presence of anthropogenic cavities, can maintain functional connectivity in this species. Threatened species Ocellated lizard landscape genetics gene flow landscape resistance resistance optimization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Landscape functional connectivity has been demonstrated to be a critical determinant of population structure and dynamics (e.g. Crooks and Sanjayan, 2006 ; Kool et al., 2013 ). Defined as a species-specific measure of an organism’s capability to move among resource patches through a landscape (Taylor et al., 1993 ), it can be assessed for a wide range of movement types, from daily foraging to migration. As a result, connectivity is also process-specific (McClure et al., 2016 ), although it has been more commonly investigated in studies focusing on dispersal (Baguette et al., 2013 ). The strength and directionality of dispersal flow between connected habitat patches can affect, among other outcomes, genetic diversity, genetic drift, inbreeding depression, and local colonization and extinction events (Keyghobadi, 2007 ). These ecological and evolutionary processes unfold within naturally dynamic landscape systems, in which ecological successions as well as natural perturbations can modify connectivity and, as a result, dispersal trajectories. However, the contemporary acceleration of human modifications of landscape composition and configuration are the main drivers of connectivity alterations (Schlaepfer et al., 2018 ). Indeed, the habitat fragmentation caused by linear infrastructures and land-use changes in rural and urban areas directly impacts habitat quality but also reduces its connectedness, i.e. the structural links connecting habitat patches (Miles et al., 2019 ). These constraints can materialize as physical barriers, such as dams for river systems (Nilsson et al., 2005 ), or be spatially diffuse, such as artificial light pollution (Korpach et al., 2022 ; Seewagen et al., 2023 ). They can impose an increased cost to movements (energy expenditure) and/or increased risks (predation, road collision, competition), resulting in reduced or redirected movements within the landscape (Cassimiro et al., 2023 ; Masden et al., 2010 ). For example, the construction of wind farms in Sweden greatly reduced the use by reindeer of their historical migration routes located close to the project development areas (Skarin et al., 2015 ), and wolves’ selection of anthropogenic linear features increased the predation risk for caribou near these features (Whittington et al., 2011 ). Conversely, anthropized areas can facilitate movements for some species and human transportations can be used as a vector for dispersal (Crispo et al., 2011 ; e.g. extensive urbanization as dispersal facilitator in feral pigeons: Carlen and Munshi-South, 2021 ; linear infrastructure as dispersal corridor in urban foxes: Kimmig et al., 2020 ; human-mediated transport in western black widow spider: Miles et al., 2018 , and tropical house gecko: Phillips et al., 2024 ). However, urbanization may only favor persistence and connectivity for a small group of generalist and urbanophile species, to the detriment of native specialists relegated to fragmented suburban and fringe habitats (McKinney, 2006 ; McKinney and Lockwood, 1999 ). Preserving connectivity has thus become a major issue for the conservation of functional ecosystems in human-modified landscapes (Lookingbill et al., 2022 ), either through the protection of existing connections or the creation of ecological corridors where connectivity has been disrupted (Rudnick et al., 2012 ). From adapting the location of large-scale infrastructure projects (e.g., offshore windfarm) to maintain transcontinental migratory routes to implementing localized wildlife crossings over roads (Dunn et al., 2019 ; Soanes et al., 2024 ), this objective has been integrated into spatial planning and operational strategies at different spatial scales (Hilty et al., 2020 ; Sordello et al., 2021 ). Indeed, even at small spatial scales, populations can be strongly and rapidly isolated, depending on species dispersal ability and matrix resistance to movements (Beninde et al., 2016 ; Schulte et al., 2013 ). Fine-scale connectivity analyses can thus contribute to preserve and integrate more effective ecological networks within human-modified environments. Among these environments, the periurban fringes, i.e., landscapes within which urban, rural and natural habitats are intertwined without any one dominating the surface area, represents transitional areas where ecological planning is crucial for conserving connectivity. In ecological research, assessing and mapping connectivity with empirical data is a common approach to assess the effects of landscape characteristics and management on the functionality of ecological continuities (Creech et al., 2024 ; McCluskey et al., 2024 ; Zeller et al., 2012 ). Within the connectivity modeling toolbox, landscape genetics (LG) has emerged as a powerful tool, used to relate landscape characteristics with gene flow, genetic structure and effective dispersal (Balkenhol et al., 2015 ). Populations that are well connected tend to exhibit higher genetic similarity, whereas isolated populations often become genetically distinct over time due to limited gene flow. By linking genetic variability (genetic distances between individuals or populations) with landscape composition and configuration, LG offers a direct and quantitative evaluation of functional connectivity (Zeller et al., 2018 ). Specifically, some LG approaches allow the inference of landscape resistance to movements, through the optimization of resistance surfaces (e.g., continuous surfaces representing how the landscape impedes movements). This can be done by the use of algorithms that identify the sets of resistance values assigned to landscape variables that maximize the correlations between genetic distances between individuals or populations and the resistance-based distances among them (Bauder et al., 2021 ; Kimmig et al., 2020 ). Prior studies have demonstrated the utility of landscape genetics in informing urban planning decisions, particularly by quantifying connectivity and identifying barriers and corridors critical to species movements (Baptista et al., 2025 ; Brunt and Smith, 2025 ; Jha and Kremen, 2013 ; Soanes et al., 2018 ; van Strien et al., 2014 ). However, landscape genetic studies remain scarce at the scale of periurban development units, mainly due to the associated time and cost constraints of this type of studies (LaPoint et al., 2015 ). Consequently, connectivity assessments integrated into environmental impact assessments (EIA) of infrastructure projects are most often conducted at a single spatial scale and rely on model-based or expert-driven approaches, which may overlook scale-dependent processes shaping gene flow (Cumming and Tavares, 2022 ). Accounting for functional connectivity across multiple spatial scales is therefore a key challenge to better capture the ecological processes underlying species movements and population genetic structure, and to improve the relevance of connectivity assessments for spatial planning. In this study, we focus on the multiscale effects of landscape on gene flow and population genetic structure of the ocellated lizard Timon lepidus (Daudin, 1802). As a low-mobility and cavity-dependent species able to inhabit anthropized areas, in which it exploits cavities embedded in the built environment (dry stone-walls, cracks under concrete surfaces, underground piping systems, building façades, etc…), this terrestrial lizard is an interesting species to study the effects of urbanization level and configuration on gene flow and connectivity (Ludot et al., 2025 ; Renet et al., 2022 ). Space use and movement patterns of this species during its active reproductive period have been studied in various natural areas. However, little is known about its movement ability in periurban landscapes, and, more generally, about its dispersal capacity (Grillet et al., 2010 ; Renet et al., 2022 ; Thirion et al., 2009 ). Our main objectives were to identify environmental drivers of gene flow across a periurban landscape within which the species is continuously distributed, and across a broader area, including this periurban landscape but also the surrounding Mediterranean rural landscape, in which the species is discontinuously distributed. We hypothesized that, at the broad spatial scale, landscape resistance would have a greater influence on Timon lepidus gene flow than distance alone, with genetic structure primarily shaped by rivers, as well as by large Mediterranean forested areas, agricultural fields and areas with low shelter densities. At the finer periurban scale, throughout the spatially continuous distribution, we predicted low levels of genetic structure, with urban cover and the road network being the main factors hindering gene flow, while the density of linear infrastructures providing anthropogenic shelters could facilitate it. 2. Material and methods 2.1 Study species and area The ocellated lizard range spreads from the Iberian Peninsula to northwestern Italy. In France, in its eastern Mediterranean geographic distribution, it favors hilly landscapes of semi-open garrigue shrubland, rocky steppes and dry grassland, as well as traditional agricultural lands, such as olive orchards and vineyards, belted by stonewalls and embankments. The ocellated lizard is an opportunistic forager that predates directly in the neighborhood of its shelter and which spends a lot of time thermoregulating at its entrance. Its home ranges are hence small (~ 1 ha), centered around a dense network of fewer than a dozen shelters in which it hibernates (Cheylan et al., 2015 ). Our study area is centered around the CEA Cadarache research center, in southern France (Fig. 1 ). It is a 9-km² periurban site comprising 2-km² of impervious areas, and extends across 220-km² of Mediterranean forests, shrublands, agricultural fields and small cities on the left bank of the Durance River. The periurban research center is known to host an important population of ocellated lizards, that opportunistically use anthropic cavities embedded in the built environment. 2.2 Sample collection We established an individual-based multi-scale sampling scheme which includes (i) a continuous sampling within the area covering the Cadarache periurban site and contiguous favorable habitat (hereafter the Central Study Area, CSA), (ii) a discontinuous sampling in an approximate 10-km wide radius around the Cadarache plateau (hereafter called “extended study area (ESA)”), targeting known species presence locations. The sampling campaigns occurred from April to August 2023 and 2024. DNA samples were either obtained from buccal swabbing upon individual capture, from scats and moults found near occupied shelters or from tissues collected off road kills. The locations of all samples were recorded (using the qfield application on mobile device) as UTM (Universal Transverse Mercator) coordinates. To capture individuals, we used tunnel traps positioned at the entrance of shelters consequently to lizard sightings, although we also resorted to opportunistic hand capture. The traps were 80-mm width PVC tubes of 50-cm to 1-meter length (according to shelter configuration) and were equipped with a one-way entrance. The upper-side of the tube was cut and replaced by a 50-mm mesh net except for the entrance edge, which was left whole. The net also covered the exit of the tube. This design impeded the lizards going inside the tube to leave it, while letting the light flow inside. Following set-up, the traps were checked every 30-min and were removed after 4 to 6 unsuccessful checks depending on climate conditions. Upon capture, we used a sterile swab to collect DNA (Puritan Medical Products, Sterile Polyester Tipped Applicator) and stored samples in a DNA stabilization solution (Zymo Research, DNA/RNA Shield). To avoid sampling twice the same individual, we took photographs of the right and left flanks of all individuals for identification. At a later stage, we also checked all genotypes for duplicates (see 2.4). Scat samples were either found at sight during the searches of individuals or during specific sessions with the assistance of a detection dog trained to detect ocellated lizard scats (Olivier et al., 2017 ). The scat samples were filtered to remove feces that did not match the size criteria to distinguish them from Lacerta bilineata feces. All samples were stored at − 20°C upon collection until shipment in dry ice to Microsynth ecogenics (Balgach, Switzerland) for genotyping. Capture permits were delivered by the French Government (permit numbers: n°2022-094-004). 2.3 Genotyping DNA extraction of swabs, moults and tissues was performed with Qiagen DNeasy Blood and tissue kits and DNA extracts from tissues were diluted to 1 ng/µl with 10 mmol Tris-HCl pH 8 before used in PCR reactions. DNA extraction of scat samples was performed with ZymoBIOMICS kit. All individuals were genotyped at 15 microsatellite loci, using markers TL_303574, TL_361715, TL_684747, TL_265456, TL_966180, TL_1079095, TL_276028, TL_619091, TL_742562, TL_1006729, TL_750235, TL_741815, TL_985672, TL_748432, TL_1291540 (Online Resource 1 and 2). Primers were labelled with FAM, ATTO-532, ATTO-550, ATTO-565. The multiplex PCR amplification was performed in a total volume of 10 µl using a mix of 1.5 µl double distilled water, 5 µl Multiplex Master Mix (stock concentration 2 x; Qiagen catalog no. 206143), 0.3 µl of each the forward and reverse primer (stock concentration 10 µM), and 2–10 ng of genomic DNA (stock concentration 1–5 ng/µl). The cycling protocol started with an initial denaturation step of 15 min at 95°C, followed by 40 cycles with 30 s at 94°C, 90 s at 56°C, and 60 s at 72°C, and ended with a final extension step of 30 min at 72°C. Amplifications were performed on a Programmable Thermal Controller (Sensoquest Labcycler). The PCR products were run on an ABI 3730XL DNA Analyzer (Thermo Fisher Scientific) and fragment lengths were scored with the Size Standard GeneScan LIZ500 (Applied Biosystems). 2.4 Data cleaning DNA amplification was not equally successful depending on sample types, with moult and scat samples having a lower genotype completion rate (Online Resource 3). Hence, as a trade-off between missing information and spatial distribution of samples (some locations being predominantly represented by moult and scat samples), we applied a cut-off threshold allowing a 0.3 missing rate for genotypes. Due to financial constraints, we were not able to perform multiple genotyping of scat and moult samples, but we performed several analyses to ensure that they were not diverging from buccal and tissue samples, and could therefore be considered reliable (Online Resource 4). We computed the triadic likelihood relatedness index with COANCESTRY through the R package related using a 0.7 threshold score to identify duplicates and we removed them from the final dataset (Wang, 2011 ). The frequencies of null alleles and their associated confidence intervals were estimated using the method of Chakraborty et al. ( 1994 ) in PopGenReport (Adamack and Gruber, 2014 ). We estimated linkage disequilibrium (LD) by calculating the pairwise (rd) index of association with 500 permutations in poppr (Agapow and Burt, 2001 ; Kamvar et al., 2014 ). We adjusted the p-values using the False Discovery Rate method (Benjamini and Hochberg, 1995 ) to account for multiple testing. Deviations from Hardy-Weinberg equilibrium (HWE) were calculated using the exact test based on Monte Carlo permutations in the pegas package (Paradis, 2010 ). 2.5 Genetic diversity and structure For each locus, we calculated the number of alleles (A), the effective number of alleles (Ae), and the observed (Ho) and expected heterozygosity (He), using the adegenet package (Jombart, 2008 ). To identify genetic structure at the two spatial scales, we conducted a spatial principal component analysis (sPCA, Jombart et al., 2008 ), a cluster analysis (STRUCTURE, Pritchard et al., 2000 ) and a spatial autocorrelation and isolation by distance (IBD) analysis. The sPCA is a spatial multivariate method that describes genetic structure by measuring spatial autocorrelation and genetic variability in the dataset without assigning individuals to clusters. In contrast, STRUCTURE is an aspatial Bayesian model-based clustering method. We applied sPCA using the spca function implemented in the ADEGENET package (Jombart, 2008 ). We set the connection network using the inverse of pairwise Euclidean distances between individuals and the significance of local and global spatial structures was tested performing eigenvalue tests (nperm = 9999, Montano and Jombart, 2017 ). We used the admixture model with correlated allele frequencies in STRUCTURE and ran Markov chain Monte Carlo simulations with 500 000 iterations after 50 000 iterations of burn-in. A preliminary testing phase with 100 000 iterations after 50 000 iterations of burn-in was used to select the ALPHAPROPSD and Lamba parameters, setting K from 1 to 5 clusters and using 3 replications. We then set the number of clusters K from 1 to 10, with 10 replications for each K (ALPHAPROPSD = 1.0, lambda = 0.640). The best K value was determined based on the delta K method (Evanno et al., 2005 ). The replicate runs with the best K value were then averaged using CLUMPP (Jakobsson and Rosenberg, 2007 ), and individuals were assigned to clusters with a 0.6 threshold membership. We applied sPCA and STRUCTURE analyses at both the central and extended study areas. The spatial autocorrelation analysis was performed in the GenAlEx software (Peakall and Smouse, 2006 ), following the approach developed by Smouse and Peakall ( 1999 ). GenAlEx computes an autocorrelation coefficient (r) using pairwise genetic and geographic distance matrices. The observed distribution of the (r obs ) coefficient is compared to a generated distribution (r rd ) under the assumption of random mating, using permutations among individuals, to detect spatial structure at specified distance intervals. We used distance intervals of 80-m, which approximates a mean ocellated lizard home range (Cheylan et al., 2015 ; Renet et al., 2022 , Ludot et al. in prep ). Only the CSA was considered for this analysis as the sparser and discontinuous sampling at the ESA did not allow a robust estimation of spatial autocorrelation across distance classes. Significance was calculated based on 10,000 permutations and bootstraps. Finally, we tested for the presence of isolation by distance (IBD) at both sampling scales with maximum likelihood population effects (MLPE) mixed-effects models, using the R package resistanceGA (Peterman, 2018 ). Inter-individual genetic distances were based on a PCA-based metric (Shirk et al., 2017 ) using the dudi.pca function from the R package ade4 (Thioulouse et al., 2018 ) and we computed geographic pairwise Euclidean distances in SPAGEDI (Hardy and Vekemans, 2002 ). Geographic distances were log-transformed. 2.6 Effects of environmental features on gene flow 2.6.1 Landscape covariates Using a 10-m resolution land cover map (CoSIA IGN, 2018–2023, https://geoservices.ign.fr/telechargement-api/COSIA ), we created a categorical land cover surface with seven thematic classes (urban, grassland, agricultural, shrub, tree, water and bare ground) by merging the 15 original classes based on shared properties relatively to ocellated lizard habitat requirements (Online Resource 5a). We used OpenStreetMap (2015) to create a categorical road surface combining motorways, primary, secondary and tertiary roads; and a categorical river surface. We dilated the linear features (rivers and roads) by 10 m to avoid breaks caused by the rasterizing of the vector data. We used a 1-m elevation raster (RGE ALTI®, geoservices.ign.fr) to compute a 10-m continuous slope surface using the package terra (Hijmans, 2024 ). We used a 10-m resolution surface representing the occurrence probabilities of natural shelters. This surface was built using a habitat suitability distribution modeling approach, as described in a previous study investigating ocellated lizard’s suitable habitat within the same periurban landscape (Ludot et al., 2025 ). Lastly, we used a 10-m resolution raster representing an underground electrical infrastructure running through the CSA. This linear feature has been described as positively associated with the species presence due to its numerous cavities (Ludot et al., 2025 ) and as such could facilitated dispersal movements. We used these 10-m resolution rasters for the analyses conducted in the CSA, but to minimize the computational time of the analyses conducted in the ESA, we converted these surfaces to 40-m pixel rasters. We used these surfaces to calculate several landscape variables, to be used in the analysis of landscape resistance to gene flow. To do so, we computed the percentage of each land cover class and of roads in 80, 160 and 320-m buffers around each pixel, and the mean of slope and natural shelter occurrence probabilities in the same buffers. As mentioned above, 80-m is the approximate average diameter of an ocellated lizard home range. The 160 and 320-m distances approximated respectively the mean and maximum distances from home-range centroids to the farthest location of adult male ocellated lizards in Cadarache research center (Ludot et al., in prep .). We also computed a continuous surface representing the distance to the closest rivers. For clarity, some variables are abbreviated in the tables and figures; the abbreviations for each variable are listed in Online Resource 5b. 2.6.2 Resistance surface optimization To evaluate functional connectivity, we used the RESISTANCEGA package in R (Peterman, 2018 ) to optimize resistance surfaces at the two spatial scales. Resistance surfaces represent estimates of landscape resistance to movements, associated with landscape structure and composition. The RESISTANCEGA workflow calculates pairwise resistance distances between individuals and uses a genetic algorithm from the GA package (Scrucca, 2013 ) to iteratively calibrate resistance values. The optimization uses the log-likelihood (or AIC) from a linear mixed-effects model with a MLPE parameterization to maximize the fit of resistance surfaces to the genetic data. To calculate the pairwise resistance distances between individual sampling locations, we used the random-walk commute-time method from the GDISTANCE package (Etten, 2017 ; McRae et al., 2008 ). This function models the landscape (our continuous raster surfaces) as a network of connected cells, where movement between adjacent cells occurs probabilistically according to their resistance values. The commute-time distance represents the expected cost of all possible random-walk paths connecting two sampling locations. This measure is conceptually similar to the effective resistance computed by circuit theory and both were shown to be highly correlated. First, we used the SS_optim function to conduct single-surface optimizations for each variable independently across four buffer sizes (0, 80, 160, and 320-m), comparing them against an isolation-by-distance (IBD) model based on Euclidean geographic distance. A variable was retained if its best model had an AICc at least 2 units lower than that of the IBD model and the best-performing buffer size was selected for each retained variable. We used the removeCollinearity function to assess the pairwise correlations among the retained variables and proceeded to variable exclusion if correlation was larger than 0.7. Second, we performed a multi-surface optimization, combining the retained variables at their selected buffer size, and used AICc to compare performance across all the existing combinations of the retained variables. We performed five independent optimization replicates to account for the stochastic nature of the resistanceGA optimization algorithm and to ensure the robustness of model selection. For both steps, all types of relationships between resistance and the environmental variables available in resistanceGA were tested. All analyses were based on the continuous raster surfaces described above. 3. Results 3.1 Sample collection and genetic data We collected 454 samples, among which 151 buccal swabs, 274 scats, 20 moults, and 9 tissue samples (Fig. 1 ). In natural habitat, outside of the periurban research center, 85.15% of scat samples (138 out of 162) were found with the assistance of the detection dog. Among all the genotypes, pairwise permutation tests did not reveal any significant linkage disequilibrium (Online Resource 6). Deviations from Hardy–Weinberg equilibrium were observed in seven loci and four of them had a significant likelihood of null alleles (Online Resource 7). Due to the high frequency of null alleles detected for TL_361715 (0.20), and its high percentage of missing data (0.14), we excluded this locus and therefore based our analyses on the 14 remaining microsatellite loci. The other loci had null allele frequencies < = 0.08, which were considered as low enough to keep them in the analyses. Following removal of TL_361715 , secondary genotype missing rate cut-off and duplicate filtering, we retained 202 genotyped samples in our final dataset, with an overall rate of missing data of 3.35%. Among these samples, 170 occurred inside the CSA (Fig. 1 ). 3.2 Genetic diversity and structure Over the 202 samples, the number of alleles in the 14 retained microsatellite loci ranged from 3 to 8 (mean = 4.86), effective number of alleles was 1.05 to 2.83 (mean = 2.12), observed heterozygosity was 0.02 to 0.66 (mean = 0.42) and expected heterozygosity was 0.02 to 0.69 (mean = 0.46). In the extended study area, the sPCA revealed a significant global spatial structure (global test, p < 0.001), while no local structure was detected (local test, p = 0.998). We retained the first three positive axes (0.10 and 0.09) and no negative axis. For the first global score, positive values were concentrated in the south of the ESA and the southwestern part of the research center, and negative values were in the north and east of the ESA and in the northeast of the research center (Fig. 2 ). For the second global score, no clear spatial pattern emerged (Fig. 2 ). In the central study area, the sPCA revealed a significant global spatial structure (global test, p < 0.001), while no local structure was detected (local test, p = 0.984). We retained the three first positive axes (0.11 and 0.10) and no negative axis. For the first global score, there was no strong spatial pattern, although there were only negative scores in the southern part of the CSA (Online Resource 8). For the second global score, there was no strong spatial pattern either (Online Resource 8). In the ESA, the optimal number of genetic clusters inferred by the Evanno method applied to STRUCTURE results was 4 (ΔK = 7.18, mean likelihood = -4735.89; Online Resource 9). However, using a threshold assignment membership of 0.60, most individuals (n = 174, 86%) were classified as admixed, and one of the four clusters had no individual assigned to it (Online Resource 10). Considering STRUCTURE might have inferred ghost clusters (i.e. a cluster with no individual, Guillot et al., 2005 ), we looked at K = 2, which obtained the second highest delta K score and the highest mean likelihood with low variability among runs (ΔK = 5.92, mean likelihood = -4716.34, Online Resource 9). Considering the assignment probability bar plot (Online Resource 10), we applied a more severe threshold of 0.70, to avoid including individuals with uncertain cluster memberships. There was no clear spatial boundary between the two clusters but rather a gradient, with the individuals sampled in the east being assigned to one group, those in the west to the other, and the central zone (Cadarache research center) with individuals from both groups (Fig. 3 ). In the CSA, the highest mean likelihood, with low variability among runs, was obtained for K = 1 (mean likelihood = -3851.63, Online Resource 11). Given that, by construction, the Evanno method cannot infer K = 1 as the best solution (Evanno et al., 2005 ), we did not apply it to STRUCTURE results for this spatial scale and considered that the most likely number of genetic clusters in the CSA was 1. The spatial autocorrelation analysis revealed fine-scale positive autocorrelations up to 720-m (r = 0.028, P = 0.006, Fig. 4 ) and negative spatial autocorrelation arose from 3600-m (r = -0.027, P = 0.034). Finally, we found a significant positive effect of geographic distance on genetic distance at both spatial scales (ESA: slope coeff. = 0.033; CSA: slope coeff. = 0.026; Online Resources 12 and 13). 3.3 Effects of environmental features on gene flow Extended study area . During the single-surface optimization step, seven variables generated resistance distances that exceeded the performance of the IBD model (∆AICc > 2, Online Resource 14): percentage of water cover, probability of natural shelter occurrence, mean slope, percentage of roads and agricultural surfaces, all within a 320-m buffer; percentage of tree cover within a 160-m buffer; and distance to rivers. None of these variables were correlated (correlation coefficient < 0.7). Testing all combinations of these seven variables in the multi-surface optimization, the combinations that best explained inter-individual genetic distances were quite consistent across the five replicates, with seven combinations ranking as equivalently best performing (∆AICc < 2) (Online Resource 15, Table 1 ). Five of these combinations were present in all replicates, one in four out of five replicates, and one in three replicates (Table 1 ). The part of variance of inter-individual genetics distances explained by the resistance distance varied between 0.10 and 0.12 in the top-performing models. Table 1 Summary of the most performant variable combinations (∆AICc < 2) during the multi-surface optimization step, and mean contribution of each covariate, across the 5 replicates, in the extended study area. N: number of covariates in the combination, x/5: number of replicates that ranked the combination within the most performant among the 5 replicates. Within each replicate, variable contribution was estimated by the genetic algorithm during the multi-surface optimization and reflects the relative influence of each variable on the composite resistance distance. N mean contribution across replicates combination presence across replicates agricultural river shelter road slope 2 agricultural * river 5/5 0.45 0.55 3 agricultural * river * shelter 5/5 0.20 0.24 0.56 2 river * shelter 4/5 0.25 0.75 3 agricultural * river * slope 5/5 0.24 0.28 0.49 2 agricultural * shelter 5/5 0.24 0.76 2 road * slope 3/5 0.58 0.42 2 river * slope 5/5 0.43 0.57 Two variables never appeared in any top combination: water percentage and tree percentage. Conversely, the percentage of agricultural land, the distance to river, and the probability of natural shelter occurrence were the variables most frequently represented across all combinations, with slope and road being the least represented respectively. Across replicates, the relative contributions of variables to resistance surfaces varied among combinations (Table 1 ). Natural shelter occurrence consistently had the highest contribution when included, with mean values ranging from 0.56 to 0.76. Distance to rivers and agricultural land contributed to the same amount when appearing both within the same combination, ranging from 0.20 to 0.55. Slope contributed twice as much (0.49) as river and agricultural land (0.28 and 0.24, respectively) in their three-variable combination, but contributed similarly as road and river in their respective two-variable combinations. The relationship between variables and estimated resistance was mostly consistent across replicates and models but the effect size varied (Fig. 5 ). Across most combinations, the relationship between agricultural cover and scaled resistance was asymptotic, with resistance increasing rapidly and reaching a plateau at approximately 70% scaled resistance when agricultural cover reached around 15%. In contrast, for the river-agricultural combination, the relationship was linear, with increasing agricultural cover associated with lower resistance compared to other combinations. Conversely, increases in occurrence probability of natural shelters, slope, road surface and distance to the river were associated with decreased resistance (Fig. 5 ). For natural shelter occurrence, resistance declined non-linearly across all combinations, with a steep decrease at low probabilities followed by a gradual flattening at higher values. Similarly, resistance decreased with increasing distance to rivers, showing a rapid decline at short distances and reaching an asymptote at around 1.5 km. Increasing slope was associated with lower resistance across combinations, but the relationship was less consistent. Finally, road surface cover exhibited a monotonic decrease in resistance, indicating lower resistance at higher levels of road cover within the road – slope combination. As no combination distinguished itself from the others in terms of log-likelihood, R² and AICc scores (Online Resource 15), we used all the retained combinations to construct the resistance surface of the study area. First, for each replicate, we constructed a resistance surface by simply scaling and averaging the resistance surfaces from each of the best performing combinations (∆AICc < 2). We then created a final resistance surface as the mean of the resistance surfaces across replicates. We also created a raster of resistance uncertainty by computing the standard deviation of the resistance surfaces across replicates (Fig. 6 ). Central study area . In the CSA, five variables, when taken alone, generated resistance distances that exceeded the performance of the IBD model (∆AICc > 2, Online Resource 16): probability of natural shelter occurrence, percentage of road and agricultural surfaces within a 320-m buffer; percentage of tile network within a 160-m buffer; and distance to the river. None of these variables were correlated (correlation coefficient < 0.7). The combination analysis yielded the same four combinations across the five replicates. These included two single-variable surface combinations: agricultural surface and distance to river; and 2 two-variable combinations: agricultural – distance to river and agricultural – natural shelter occurrence (Online Resource 17). Among the two-variable combinations, the agricultural surface variable always had the lowest relative contribution across replicates compared to the distance to river and natural shelter occurrence variables (agricultural mean contribution: 0.37 and 0.36 for each combination respectively, Online Resource 18). Of the five variables that were selected in the single optimization step, two variables never appeared in any top combination: the proportion of surface covered by the tile network, and the proportion of surface covered by roads. The part of variance of inter-individual genetic distances explained by the resistance distance varied between 0.09 and 0.11 in the top-performing models (Online Resource 17). Increasing agricultural cover was associated with higher resistance. In contrast, increasing natural shelter occurrence and greater distance to rivers were associated with lower resistance (Online Resource 19). The scaled resistance assigned to values of distance to river and of the occurrence probability of natural shelters was overall much lower in the CSA compared to the ESA. The same method as for the ESA model was used to construct the resistance surface. Within the boundaries of the CSA, the resistance surfaces produced by the two models were similar (Online Resource 20). 4. Discussion We used landscape genetics to characterize at two spatial scales the impact of a periurban Mediterranean landscape on the genetic connectivity of a population of ocellated lizards. Overall, we observed a weak genetic structure at both spatial scales, suggesting an historically well-connected and/or continuously distributed population. There was significant positive genetic autocorrelation at short distance classes (< 800m), consistent with limited natal dispersal and local gene flow. Resistance-based models outperformed isolation by distance alone, indicating that landscape features influence gene flow, although the variance explained by resistance remained modest. These results partly support our hypotheses with landscape resistance affecting gene flow at broader scales, but did not support an impact of the urban environment on functional connectivity at the periurban scale. Genotyping We originally developed our sampling strategy based on the fact that we would collect faecal samples, and hence developed microsatellite markers that were likely to persist in this type of samples. However, faecal genotyping revealed poor results, despite the preliminary genotyping analyses yielding encouraging results, with scat samples exhibiting a high amplification rate and high fidelity with their buccal sample counterparts from the same individual. These preliminary results were obtained using fresh scat samples collected during ocellated lizard handling following capture. It was difficult to date the scat samples during fieldwork, and the samples were rarely fresh, having likely been exposed for several days to external climatic conditions (sun, drought and rain) that could have degraded the DNA before collection. This led us to develop a sampling strategy aimed at capturing individuals to collect more reliable buccal samples. This type of samples would have eventually allowed the use of SNP markers, which would have led to a finer distinction between individuals, as the richness of our microsatellite dataset in terms of allelic richness was low to moderate. Spatial genetic structure Overall, our results did not reveal discrete genetic clusters but instead indicated weak and continuous spatial genetic structure consistent with isolation by distance. Both sPCA analyses detected significant global structure without local structure at either spatial scale, suggesting broad-scale genetic gradients rather than genetic discontinuities. Likewise, STRUCTURE did not identify well-defined genetic clusters. Most individuals were highly admixed, especially within the central study area, and inferred clusters lacked clear spatial boundaries. Spatial autocorrelation analyses revealed significant positive genetic structure at a distance of around 700 m, indicating that individuals within a radius of approximately 1 km are more genetically similar than random. This fine-scale spatial genetic structure is consistent with the species' short-range spatial ecology (Renet et al., 2022 ; Ludot et al., forthcoming). Beyond approximately 3.5 km, however, spatial autocorrelation became significantly negative, suggesting that individuals separated by several kilometers belong to different genetic neighbourhoods. This spatial signal is consistent with the isolation-by-distance pattern observed in the aspatial and spatial genetic structure analyses and spatial genetic patterns driven by limited dispersal, with gene flow declining gradually with distance. As it represents the first genetic estimate of dispersal ability for the species, these results provide a valuable baseline for identifying appropriate spatial scales for habitat management, corridor design, and impact assessment. Ensuring functional connectivity at the kilometer scale might be particularly important for sustaining population resilience in the face of ongoing landscape changes. Taken together, the analyses of genetic structure show that the periurban area did not appear to generate a genetic break, suggesting that the level and configuration of urbanization encountered throughout the site has not restrained functional connectivity. The literature shows divergent results regarding genetic differentiation among fragmented urban populations of other lizard species. While some studies have found evidence of strong and rapid genetic differentiation (< 40 years) between populations inhabiting remaining suitable patches following urbanization (e.g., Delaney et al., 2010 in Chamaea fasciata, Sceloporus occidentalis, Plestiodon skiltonianus and Uta stansburiana ; Wenner et al., 2022 in in Phrynosoma blainvilli ), there are also reports of the urban environment having only a weak effect on genetic structure, even in conditions more urbanized than those encountered in our study (e.g. Beninde et al., 2016 in Podarcis muralis ; Virens et al., 2015 in Ctenotus fallens ). These latter results have two potential explanations: either a time lag between the urbanization process and the observed genetic signal as detected by the genetic markers and analyses involved; or the ability of those species to cope with urban conditions during dispersal. At our study site, periurbanization dates back 60 years and replaced an already anthropogenic agricultural landscape, structured by stone walls and other linear features. Such land-use continuity and moderate levels of urbanization may have limited the disruption of dispersal throughout the site. In the ocellated lizard, the largest movements studied are performed by males during the reproductive season, and little is known about its natal dispersal. However, for this cavity-dependent species, routine movements heavily rely on a shelter network. It is probably also true, though to a lower extent, for dispersal movements. Moderate urbanisation levels, which provide human-made shelter resources, hence do not necessarily degrade habitat connectivity for this species. Environmental drivers of connectivity Our analyses of the resistance of landscape features to gene flow revealed that the environmental drivers of connectivity slightly differed between spatial scales but were mainly consistent. At the broad ESA scale, resistance modelling suggested that rivers and agricultural cover contributed to limiting gene flow, particularly when considered jointly. Rivers are often cited as an impediment to gene flow in terrestrial lizards (Beninde et al., 2016 ; Oliveira et al., 2018 ; Pounds and Jackson, 1981 ). While anecdotal observations of ocellated lizards fleeing underwater suggest that the species can swim, this is most likely a marginal antipredation strategy (Cheylan et al., 2015 ). This effect is likely reinforced by the fact that riverbanks are often bordered by extensive agricultural fields in our study area. Rather than acting as absolute barriers, rivers and surrounding agricultural matrices therefore appear to contribute to a gradual reduction in connectivity. Indeed, resistance increased sharply with agricultural cover, suggesting that agricultural areas act as a resistant matrix once they surpass relatively low thresholds. Some traditional Mediterranean agricultural structures (e.g., stone walls and dry-stone buildings) provide anthropogenic cavities. However, these old structures are often abandoned and are disappearing more and more (López-Estébanez et al., 2024 ). In some parts of the ocellated lizard range, the restoration of dry-stone walls is a priority conservation action by local planning (LPO Auvergne, https://auvergne-rhone-alpes.lpo.fr/projets/lezard-ocelle/ ). The observed resistance pattern likely reflects contrasted movement conditions within agricultural landscapes. Edges and trails through agricultural plots may provide linear features and shelter opportunities that facilitate movement, whereas crossing the interior of large fields is likely more costly due to the lack of refuges, vegetation structural complexity and prey resources. Through agricultural land consolidation, the merging of agricultural plots has reduced edge density, removed hedgerows and stonewalls delimitating plots, and increased field size, potentially decreasing permeability for cavity-dependent species (Colucci, 2014 ; Michael et al., 2021 ). In their study, Guerrero-Casado et al. ( 2022 ) showed that reptile abundance and species richness were markedly lower in modern intensive trellis vineyards than in traditional vineyards, with Timon lepidus absent from trellis systems. Conversely, the availability of natural shelters strongly promoted connectivity, with areas of low shelter probability being associated with high resistance. Across combinations in which it was retained, shelter occurrence showed the highest relative contribution to resistance, and resistance declined sharply even at low probabilities of shelter presence. This pattern indicates that small increases in cavity availability can enhance permeability of the landscape, highlighting the importance of fine-scale structural features for movements in this species. In contrast, variables describing general land cover, such as shrubs or tree cover, were never retained in the top-performing models, suggesting that structural microhabitat features outweigh coarse land-cover descriptors in shaping functional connectivity. Similarly, steeper terrain facilitated gene flow, likely due to the greater density of natural cavities in such areas. Roads also contributed to connectivity at this scale, which might be due to their rocky and open embankment that can host cavities. Together, these results align with the species' strong dependence on cavity-rich habitats. At the periurban CSA scale, however, no clear resistance signal emerged throughout the research center on the optimized resistance surface, suggesting that moderate urbanization does not markedly disrupt connectivity. The absence of a strong effect may also reflect the area's relatively recent development, which may not yet have produced detectable genetic consequences. The main drivers of gene flow were similar to those operating at the broader ESA scale, but were confined to the CSA's periphery rather than its urbanized core. Two weak signals emerged during the single-surface optimization stage, indicating that anthropogenic linear features (roads and tile networks) acted as local facilitators of connectivity, although these features were not retained in the final multi-surface model. At the CSA scale, roads also contributed to connectivity, likely not because they promote direct crossing, but because individuals may move along their embankments and through managed open vegetation on roadsides, due to the research center legal obligation to clear vegetation around infrastructures. Future studies could explicitly account for road use intensity, road width, and management practices to better disentangle the role of roads as barriers and potential movement corridors. As expected, given the presence of anthropogenic cavities along its entire length, the tile network, which is present only in the peri-urban landscape, also showed a positive correlation with gene flow. Management implications Our results suggest that the landscape connectivity of Timon lepidus is more dependent on the fine-scale distribution of cavities across the landscape than on broad land-cover categories. Effective conservation measures should therefore prioritize the preservation and restoration of structures that provide cavities, such as dry-stone walls, terraces and embankments, as well as other traditional elements of Mediterranean agro-pastoral systems. This is particularly urgent given the documented decline of dry-stone wall networks in many Mediterranean regions, which has prompted EU-level restoration initiatives such as the LIFE StoneWall4Life project. At the same time, the resistance associated with large, homogeneous agricultural fields highlights the importance of promoting heterogeneity and microhabitat diversity within intensively cultivated areas. In periurban landscapes, our findings suggest that anthropogenic structures can partially compensate for the loss of natural shelters, and that retaining artificial microhabitat features (e.g., cavities along road embankments or utility infrastructures) can help to maintain functional connectivity. Integrating biodiversity objectives with the preservation of cultural landscapes and traditional land-use practices may substantially improve the social acceptability and durability of conservation actions (Bridgewater and Rotherham, 2019 ; Rollo, 2025 ). Bio-cultural heritage frameworks highlight that many species depend on long-maintained anthropogenic structures, and that conserving these elements supports both ecological connectivity and cultural identity (Rotherham, 2015 ). Aligning shelter restoration with heritage preservation thus represents a promising pathway for participatory and socially grounded conservation planning. Additionally, our results provide an interesting insight into the limited dispersal ability of the species at a fine scale, which nevertheless leads to a gradual genetic structuring. Although the genetic structure did not seem to be impacted by urbanization, given the short-range dispersal capacities of this species, the emergence of a barrier could have a strong effect on population genetic structure. Overall, as the species' spatial ecology depends on a network of shelters, identifying existing shelter resources, to be maintained or rehabilitated, as well as areas lacking shelter resources, is essential to maintain or enhance landscape connectivity for this species. These conservation efforts should enable Timon lepidus to persist and move throughout mosaic periurban environments, provided sufficient suitable habitat remains available. Given the moderate levels of genetic diversity inferred in our study area, despite the rather high density levels of the species compared to numerous other areas, such measures need to be seriously considered for implementation. Perspectives Even though the final genotyping of feces provided less data than expected, the use of dogs to find these samples was highly profitable, as the majority of feces were found in bushy areas that were undetectable by sight alone. Training dogs to detect the scent of individuals rather than feces could also be used to capture individuals or simply to monitor the presence of the species on site and find its shelters. Because our landscape genetic analyses relied exclusively on contemporary environmental covariates, the inferred resistance patterns reflect recent landscape configurations rather than historical conditions under which part of the observed genetic structure may have been established. Comparing current relationships with those derived from past landscape configurations would help disentangle historical and contemporary effects. Finally, other methodologies complement genetic approaches when it comes to estimating dispersal and movement capacities, particularly with regard to landscape connectivity. Techniques such as GPS tracking and radiotelemetry provide direct, high-resolution data on individual movements, habitat use and potential dispersal barriers, which are difficult to infer from genetic data alone. However, these approaches are often used independently, which limits our ability to fully understand population connectivity. The importance of combining genetic and movement-based data is often emphasized in global reviews, as this enables researchers to link realised dispersal patterns with actual gene flow (Brum et al., 2023 ; Tan et al., 2023 ). This reveals not only where individuals move, but also which movements result in successful reproduction. In the context of our study, combining genetic analyses with radiotracking or GPS data would provide a more comprehensive understanding of the species' dispersal ecology and help identify critical habitats, corridors and potential barriers. This integrated approach would be invaluable for informing conservation management, guiding habitat connectivity measures and predicting population responses to environmental changes. Declarations Competing Interests The authors have no relevant financial or non-financial interests to disclose. Funding sources This research has been supported by the Commissariat à l’Energie Atomique et aux Energies Alternatives (grant no. MRISQ) Author Contribution Johan Ludot: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review and editingAurélie Coulon: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – review and editingBenoit Charrasse: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – review and editingVéronique Arnal: Resources Acknowledgement We thank the "Collection BEV" (Biogéographie et Ecologie des Vertébrés collection) housed at the Centre d’Ecologie Fonctionnelle et Evolutive (CEFE) in Montpellier, France and the persons who contributed to the sampling of the individuals we used for microsatellite development, among which Alexandre Cluchier, René Celse, Marc Cheylan; and for microsatellite genotyping among which: Vincent Hallot, Jean-Baptiste Rico, Lena Delcamp, Serena Ghiles, Nicolas Fuento and Nathalie Espuno. Data Availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. References Adamack AT, Gruber B (2014) PopGenReport: simplifying basic population genetic analyses in R. Methods Ecol Evol 5:384–387. https://doi.org/10.1111/2041-210X.12158 Agapow P-M, Burt A (2001) Indices of multilocus linkage disequilibrium. Mol Ecol Notes 1:101–102. https://doi.org/10.1046/j.1471-8278.2000.00014.x Baguette M, Blanchet S, Legrand D, Stevens V, Turlure C (2013) Individual dispersal, landscape connectivity and ecological networks. Biol Rev 1 Balkenhol N, Cushman SA, Storfer AT, Waits LP (2015) Landscape genetics concepts, methods, applications. Wiley Blackwell Baptista MSP, Keuroghlian A, Tambosi LR, Côrtes MC, de Góes Maciel F, Cirino DW, Schmaedecke G, Biondo C (2025) Agriculturally developed areas reduce genetic connectivity for a keystone Neotropical ungulate. Biotropica 57:e13418. https://doi.org/10.1111/btp.13418 Bauder JM, Peterman WE, Spear SF, Jenkins CL, Whiteley AR, McGarigal K (2021) Multiscale assessment of functional connectivity: Landscape genetics of eastern indigo snakes in an anthropogenically fragmented landscape in central Florida. Mol Ecol 30:3422–3438. https://doi.org/10.1111/mec.15979 Beninde J, Feldmeier S, Werner M, Peroverde D, Schulte U, Hochkirch A, Veith M (2016) Cityscape genetics: structural vs. functional connectivity of an urban lizard population. Mol Ecol 25:4984–5000. https://doi.org/10.1111/mec.13810 Benjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol 57:289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x Bridgewater P, Rotherham ID (2019) A critical perspective on the concept of biocultural diversity and its emerging role in nature and heritage conservation. People Nat 1:291–304. https://doi.org/10.1002/pan3.10040 Brum PHR, Gonçalves SRA, Strüssmann C, Teixido AL (2023) A global assessment of research on urban ecology of reptiles: patterns, gaps and future directions. Anim Conserv n/a. https://doi.org/10.1111/acv.12799 Brunt T, Smith AL (2025) Habitat Quality and Water Availability Affect Genetic Connectivity of Platypus Across an Urban Landscape. Anim Conserv 28:648–662. https://doi.org/10.1111/acv.13011 Carlen E, Munshi-South J (2021) Widespread genetic connectivity of feral pigeons across the Northeastern megacity. Evol Appl 14:150–162. https://doi.org/10.1111/eva.12972 Cassimiro IMF, Ribeiro MC, Assis JC (2023) How did the animal come to cross the road? Drawing insights on animal movement from existing roadkill data and expert knowledge. Landsc Ecol 38:2035–2051. https://doi.org/10.1007/s10980-023-01671-5 Chakraborty R, Zhong Y, Jin L, Budowle B (1994) Nondetectability of restriction fragments and independence of DNA fragment sizes within and between loci in RFLP typing of DNA. Am J Hum Genet 55:391–401 Cheylan M, Doré F, Grillet P (2015) Le Lézard ocellé - Un géant sur le continent européen, Biotope. ed Colucci C (2014) Evidence-based evaluation of artificial reptile habitats. A case study on stone structures Creech TG, Brennan A, Faselt J, Stabach JA, Keeley ATH (2024) Validating Connectivity Models: A Synthesis. Curr Landsc Ecol Rep 9:120–134. https://doi.org/10.1007/s40823-024-00102-8 Crispo E, Moore J-S, Lee-Yaw JA, Gray SM, Haller BC (2011) Broken barriers: Human-induced changes to gene flow and introgression in animals. BioEssays 33:508–518. https://doi.org/10.1002/bies.201000154 Crooks KR, Sanjayan M (2006) Connectivity Conservation. Cambridge University Press Cumming K, Tavares D (2022) Using strategic environmental assessment and project environmental impact assessment to assess ecological connectivity at multiple scales in a national park context. Impact Assess Proj Apprais 40:507–516. https://doi.org/10.1080/14615517.2022.2031553 Daudin FM, 1802. Description of Lacerta lepida (= Timon lepidus). - Histoire Naturel, générale et particulièredes reptiles, Dufart F Paris. ed Delaney KS, Riley SPD, Fisher RN (2010) A Rapid, Strong, and Convergent Genetic Response to Urban Habitat Fragmentation in Four Divergent and Widespread Vertebrates. PLoS ONE 5:e12767. https://doi.org/10.1371/journal.pone.0012767 Dunn DC, Harrison A-L, Curtice C, DeLand S, Donnelly B, Fujioka E, Heywood E, Kot CY, Poulin S, Whitten M, Åkesson S, Alberini A, Appeltans W, Arcos JM, Bailey H, Ballance LT, Block B, Blondin H, Boustany AM, Brenner J, Catry P, Cejudo D, Cleary J, Corkeron P, Costa DP, Coyne M, Crespo GO, Davies TE, Dias MP, Douvere F, Ferretti F, Formia A, Freestone D, Friedlaender AS, Frisch-Nwakanma H, Froján CB, Gjerde KM, Glowka L, Godley BJ, Gonzalez-Solis J, Granadeiro JP, Gunn V, Hashimoto Y, Hawkes LM, Hays GC, Hazin C, Jimenez J, Johnson DE, Luschi P, Maxwell SM, McClellan C, Modest M, Di Sciara N, Palacio G, Palacios AH, Pauly DM, Rayner A, Rees M, Salazar AF, Secor ER, Sequeira D, Spalding AMM, Spina M, Van Parijs F, Wallace S, Varo-Cruz B, Virtue N, Weimerskirch M, Wilson H, Woodward L, Halpin B (2019) P.N., The importance of migratory connectivity for global ocean policy. Proc. R. Soc. B Biol. Sci. 286, 20191472. https://doi.org/10.1098/rspb.2019.1472 van Etten J (2017) R Package gdistance: Distances and Routes on Geographical Grids. J Stat Softw 76:1–21. https://doi.org/10.18637/jss.v076.i13 Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Mol Ecol 14:2611–2620. https://doi.org/10.1111/j.1365-294X.2005.02553.x Grillet P, Cheylan M, Thirion J-M, Bonnet X, Dauge C, Chollet S, Marchand M-A (2010) Rabbit burrows or artificial refuges are a critical habitat component for the threatened lizard, Timon lepidus (Sauria, Lacertidae). Biodivers Conserv 19:2039–2205 Guerrero-Casado J, Carpio AJ, Canós-Burguete M, Torrijo-Salesa M, Tortosa FS (2022) The modernization of traditional vineyards into intensive trellis systems reduces the species richness and abundance of reptiles. Span J Agric Res 20:e0302–e0302. https://doi.org/10.5424/sjar/2022202-18224 Guillot G, Estoup A, Mortier F, Cosson JF (2005) A Spatial Statistical Model for Landscape Genetics. Genetics 170:1261–1280. https://doi.org/10.1534/genetics.104.033803 Hardy OJ, Vekemans X (2002) spagedi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Mol Ecol Notes 2:618–620. https://doi.org/10.1046/j.1471-8286.2002.00305.x Hijmans RJ (2024) terra: Spatial Data Analysis. R Package Version 17–81 Httpsrspatialgithubioterra Hilty J, Worboys GL, Keeley A, Woodley S, Lausche BJ, Locke H, Carr M, Pulsford I, Pittock J, White JW, Theobald DM, Levine J, Reuling M, Watson JEM, Ament R, Tabor GM (2020) Guidelines for conserving connectivity through ecological networks and corridors. IUCN, International Union for Conservation of Nature. https://doi.org/10.2305/IUCN.CH.2020.PAG.30.en Jakobsson M, Rosenberg NA (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23:1801–1806. https://doi.org/10.1093/bioinformatics/btm233 Jha S, Kremen C (2013) Urban land use limits regional bumble bee gene flow. Mol Ecol 22:2483–2495. https://doi.org/10.1111/mec.12275 Jombart T (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24:1403–1405. https://doi.org/10.1093/bioinformatics/btn129 Jombart T, Devillard S, Dufour A-B, Pontier D (2008) Revealing cryptic spatial patterns in genetic variability by a new multivariate method. Heredity 101:92–103. https://doi.org/10.1038/hdy.2008.34 Kamvar ZN, Tabima JF, Grünwald NJ (2014) Poppr: an R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2:e281. https://doi.org/10.7717/peerj.281 Keyghobadi N (2007) The genetic implications of habitat fragmentation for animals. Can J Zool 85:1049–1064. https://doi.org/10.1139/Z07-095 Kimmig SE, Beninde J, Brandt M, Schleimer A, Kramer-Schadt S, Hofer H, Börner K, Schulze C, Wittstatt U, Heddergott M, Halczok T, Staubach C, Frantz AC (2020) Beyond the landscape: Resistance modelling infers physical and behavioural gene flow barriers to a mobile carnivore across a metropolitan area. Mol Ecol 29:466–484. https://doi.org/10.1111/mec.15345 Kool JT, Moilanen A, Treml EA (2013) Population connectivity: recent advances and new perspectives. Landsc Ecol 28:165–185. https://doi.org/10.1007/s10980-012-9819-z Korpach AM, Garroway CJ, Mills AM, von Zuben V, Davy CM, Fraser KC (2022) Urbanization and artificial light at night reduce the functional connectivity of migratory aerial habitat. Ecography 2022, e05581. https://doi.org/10.1111/ecog.05581 LaPoint S, Balkenhol N, Hale J, Sadler J, van der Ree R (2015) Ecological connectivity research in urban areas. Funct Ecol 29:868–878. https://doi.org/10.1111/1365-2435.12489 Lendrum PE, Northrup JM, Anderson CR, Liston GE, Aldridge CL, Crooks KR, Wittemyer G (2018) Predation risk across a dynamic landscape: effects of anthropogenic land use, natural landscape features, and prey distribution. Landsc Ecol 33:157–170. https://doi.org/10.1007/s10980-017-0590-z Lookingbill T, Minor E, Mullis C, Nunez-Mir G, Johnson P (2022) Connectivity in the Urban Landscape (2015–2020): Who? Where? What? When? Why? and How? Curr. Landsc Ecol Rep 7:1–14. https://doi.org/10.1007/s40823-021-00068-x López-Estébanez N, Molina-Holgado P, Allende Álvarez F (2024) Preserving the Values of Mediterranean Enclosed Fields with Dry Stone Walls, an Example of Vulnerable Natural and Rural Heritage. Heritage 7:844–872. https://doi.org/10.3390/heritage7020041 Ludot J, Coulon A, Charrasse B (2025) Fine-scale distribution modeling of ocellated lizard ( Timon lepidus ), in a peri-urban landscape. Glob Ecol Conserv 62:e03724. https://doi.org/10.1016/j.gecco.2025.e03724 Masden EA, Haydon DT, Fox AD, Furness RW (2010) Barriers to movement: Modelling energetic costs of avoiding marine wind farms amongst breeding seabirds. Mar Pollut Bull 60:1085–1091. https://doi.org/10.1016/j.marpolbul.2010.01.016 McClure ML, Hansen AJ, Inman RM (2016) Landsc Ecol 31:1419–1432. https://doi.org/10.1007/s10980-016-0347-0 . Connecting models to movements: testing connectivity model predictions against empirical migration and dispersal data McCluskey EM, Kuzma FC, Enander HD, Cole-Wick A, Coury M, Cuthrell DL, Johnson C, Kelso M, Lee YM, Methner D, Rowe L, Swinehart A, Moore J (2024) Assessing habitat connectivity of rare species to inform urban conservation planning. Ecol Evol 14. https://doi.org/10.1002/ece3.11105 McKinney ML (2006) Urbanization as a major cause of biotic homogenization. Biol Conserv Urbanization 127:247–260. https://doi.org/10.1016/j.biocon.2005.09.005 McKinney ML, Lockwood JL (1999) Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends Ecol Evol 14:450–453. https://doi.org/10.1016/S0169-5347(99)01679-1 McRae BH, Dickson BG, Keitt TH, Shah VB (2008) Using Circuit Theory to Model Connectivity in Ecology, Evolution, and Conservation. Ecology 89:2712–2724. https://doi.org/10.1890/07-1861.1 Michael DR, Moore H, Wassens S, Craig MD, Tingley R, Chapple DG, O’Sullivan J, Hobbs RJ, Nimmo DG (2021) Rock removal associated with agricultural intensification will exacerbate the loss of reptile diversity. J Appl Ecol 58:1557–1565. https://doi.org/10.1111/1365-2664.13897 Miles LS, Johnson JC, Dyer RJ, Verrelli BC (2018) Urbanization as a facilitator of gene flow in a human health pest. Mol Ecol 27:3219–3230. https://doi.org/10.1111/mec.14783 Miles LS, Rivkin LR, Johnson MTJ, Munshi-South J, Verrelli BC (2019) Gene flow and genetic drift in urban environments. Mol Ecol 28:4138–4151. https://doi.org/10.1111/mec.15221 Montano V, Jombart T (2017) An Eigenvalue test for spatial principal component analysis. BMC Bioinformatics 18:562. https://doi.org/10.1186/s12859-017-1988-y Nilsson C, Reidy CA, Dynesius M, Revenga C (2005) Fragmentation and Flow Regulation of the World’s Large River Systems. Science 308:405–408. https://doi.org/10.1126/science.1107887 Oliveira EF, Martinez PA, São-Pedro VA, Gehara M, Burbrink FT, Mesquita DO, Garda AA, Colli GR, Costa GC (2018) Climatic suitability, isolation by distance and river resistance explain genetic variation in a Brazilian whiptail lizard. Heredity 120:251–265. https://doi.org/10.1038/s41437-017-0017-2 Olivier A, Smith H, Santos R, Tatin L, Espuno N, Arnal V, Montgelard C, Miaud C (2017) Utilisation de chiens pour détecter la présence du lézard ocellé (Timon lepidus) 22 Paradis E (2010) pegas: an R package for population genetics with an integrated–modular approach. Bioinformatics 26:419–420. https://doi.org/10.1093/bioinformatics/btp696 Peakall R, Smouse PE (2006) genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6:288–295. https://doi.org/10.1111/j.1471-8286.2005.01155.x Peterman WE (2018) ResistanceGA: An R package for the optimization of resistance surfaces using genetic algorithms. Methods Ecol Evol 9:1638–1647. https://doi.org/10.1111/2041-210X.12984 Phillips PM, Langhans KE, Helmus MR, Jesse WAM, Surget-Groba Y, Behm JE (2024) Dispersal restriction and facilitation in species with differing tolerance to development: A landscape genetics study of native and introduced lizards. Divers Distrib 30:e13858. https://doi.org/10.1111/ddi.13858 Pounds JA, Jackson JF (1981) Riverine Barriers to Gene Flow and the Differentiation of Fence Lizard Populations. Evolution 35:516–528. https://doi.org/10.2307/2408199 Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945–959. https://doi.org/10.1111/j.1471-8286.2007.01758.x Renet J, Dokhelar T, Thirion F, Tatin L, Pernollet CA, Bourgault L (2022) Spatial pattern and shelter distribution of the ocellated lizard (Timon lepidus) in two distinct Mediterranean habitats. Amphib -Reptil 43:263–273. https://doi.org/10.1163/15685381-bja10095 Rollo MF (2025) Interconnected Nature and People: Biosphere Reserves and the Power of Memory and Oral Histories as Biocultural Heritage for a Sustainable Future. Sustainability 17:4030. https://doi.org/10.3390/su17094030 Rotherham ID (2015) Bio-cultural heritage and biodiversity: emerging paradigms in conservation and planning. Biodivers Conserv 24:3405–3429. https://doi.org/10.1007/s10531-015-1006-5 Rudnick D, Ryan S, Beier P, Cushman S, Dieffenbach F, Epps C, Gerber L, Hartter J, Jenness J, Kintsch J, Merenlender A, Perkl R, Perziosi D, Trombulack S (2012) The Role of Landscape Connectivity in Planning and Implementing Conservation and Restoration Priorities. Issues in Ecology. Issues Ecol Schlaepfer DR, Braschler B, Rusterholz H-P, Baur B (2018) Genetic effects of anthropogenic habitat fragmentation on remnant animal and plant populations: a meta-analysis. Ecosphere 9:e02488. https://doi.org/10.1002/ecs2.2488 Schulte U, Veith M, Mingo V, Modica C, Hochkirch A (2013) Strong genetic differentiation due to multiple founder events during a recent range expansion of an introduced wall lizard population. Biol Invasions 15:2639–2649. https://doi.org/10.1007/s10530-013-0480-5 Scrucca L (2013) GA: A Package for Genetic Algorithms in R. J Stat Softw 53:1–37. https://doi.org/10.18637/jss.v053.i04 Seewagen CL, Nadeau-Gneckow J, Adams AM (2023) Far-reaching displacement effects of artificial light at night in a North American bat community. Glob Ecol Conserv 48:e02729. https://doi.org/10.1016/j.gecco.2023.e02729 Shirk AJ, Landguth EL, Cushman SA (2017) A comparison of individual-based genetic distance metrics for landscape genetics. Mol Ecol Resour 17:1308–1317. https://doi.org/10.1111/1755-0998.12684 Skarin A, Nellemann C, Rönnegård L, Sandström P, Lundqvist H (2015) Wind farm construction impacts reindeer migration and movement corridors. Landsc Ecol 30:1527–1540. https://doi.org/10.1007/s10980-015-0210-8 Smouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82:561–573. https://doi.org/10.1038/sj.hdy.6885180 Soanes K, Rytwinski T, Fahrig L, Huijser MP, Jaeger JAG, Teixeira FZ, van der Ree R, van der Grift EA (2024) Do wildlife crossing structures mitigate the barrier effect of roads on animal movement? A global assessment. J Appl Ecol 61:417–430. https://doi.org/10.1111/1365-2664.14582 Soanes K, Taylor AC, Sunnucks P, Vesk PA, Cesarini S, van der Ree R (2018) Evaluating the success of wildlife crossing structures using genetic approaches and an experimental design: Lessons from a gliding mammal. J Appl Ecol 55:129–138. https://doi.org/10.1111/1365-2664.12966 Sordello R, Paquier F, Daloz A, Patrinat OFB (2021) Trame noire - Méthodes d’élaboration et outils pour sa mise en oeuvre Tan WC, Herrel A, Rödder D (2023) A global analysis of habitat fragmentation research in reptiles and amphibians: what have we done so far? Biodivers Conserv 32:439–468. https://doi.org/10.1007/s10531-022-02530-6 Taylor PD, Fahrig L, Henein K, Merriam G (1993) Connectivity Is a Vital Element of Landscape Structure. Oikos 68:571. https://doi.org/10.2307/3544927 Thioulouse J, Dray S, Dufour A-B, Siberchicot A, Jombart T, Pavoine S (2018) Multivariate Analysis of Ecological Data with ade4. Springer New York, New York, NY. https://doi.org/10.1007/978-1-4939-8850-1 Thirion J-M, Doré F, adamczyk A, Grillet P, Cheylan M (2009) Etude spatiale et temporelle d’une population de lézard ocellé Timon lepidus en limite nord de distribution. https://doi.org/10.13140/RG.2.2.33032.93445 van Strien MJ, Keller D, Holderegger R, Ghazoul J, Kienast F, Bolliger J (2014) Landscape genetics as a tool for conservation planning: predicting the effects of landscape change on gene flow. Ecol Appl 24:327–339. https://doi.org/10.1890/13-0442.1 Virens E, Krauss SL, Davis RA, Spencer PBS (2015) Weak genetic structuring suggests historically high genetic connectivity among recently fragmented urban populations of the scincid lizard, Ctenotus fallens. Aust J Zool 63:279–286. https://doi.org/10.1071/ZO15022 Wang J (2011) coancestry: a program for simulating, estimating and analysing relatedness and inbreeding coefficients. Mol Ecol Resour 11:141–145. https://doi.org/10.1111/j.1755-0998.2010.02885.x Wenner SM, Murphy MA, Delaney KS, Pauly GB, Richmond JQ, Fisher RN, Robertson JM (2022) Natural and anthropogenic landscape factors shape functional connectivity of an ecological specialist in urban Southern California. Mol Ecol 31:5214–5230. https://doi.org/10.1111/mec.16656 Whittington J, Hebblewhite M, DeCesare NJ, Neufeld L, Bradley M, Wilmshurst J, Musiani M (2011) Caribou encounters with wolves increase near roads and trails: a time-to-event approach. J Appl Ecol 48:1535–1542. https://doi.org/10.1111/j.1365-2664.2011.02043.x Zeller KA, Jennings MK, Vickers TW, Ernest HB, Cushman SA, Boyce WM (2018) Divers Distrib 24:868–879. https://doi.org/10.1111/ddi.12742 . Are all data types and connectivity models created equal? Validating common connectivity approaches with dispersal data Zeller KA, McGarigal K, Whiteley AR (2012) Estimating landscape resistance to movement: a review. Landsc Ecol 27:777–797. https://doi.org/10.1007/s10980-012-9737-0 Statements & Declarations CRediT authorship contribution statement Johan Ludot Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review and editing Aurélie Coulon Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – review and editing Benoit Charrasse Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – review and editing Véronique Arnal: Resources Additional Declarations No competing interests reported. Supplementary Files supmatLandGen.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 06 May, 2026 Editor assigned by journal 29 Apr, 2026 Submission checks completed at journal 29 Apr, 2026 First submitted to journal 28 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9550783","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":636735506,"identity":"24b1276c-d805-42cc-bdfc-7997e1794260","order_by":0,"name":"Johan Ludot","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyklEQVRIiWNgGAWjYBACxgbmBgYeBgY5AzDXwIKBvYGgFkaQFgNjAwZmkBYJBp4DRNgD0pK4AayFgQgtzA2MbRJvKv6kb2fvP7rhRwFQizQBPUCHtUnOOWOQu7PnMNvNHpDD+BIIa5HmbTPI3XAjme0GD1CLPQ9Bn0C0pBsAtdz8A7KFWC0JIC23eYjS0szYbDnnjLHhhjOHzW7LGEjwENRi2N588MabCjl5g+ONz26++WMjR1hLM5oAIQ0MDPIEVYyCUTAKRsEoAAA5qTqez1SdVAAAAABJRU5ErkJggg==","orcid":"","institution":"Muséum national d'Histoire naturelle, Centre National de la Recherche Scientifique, Sorbonne Université","correspondingAuthor":true,"prefix":"","firstName":"Johan","middleName":"","lastName":"Ludot","suffix":""},{"id":636735511,"identity":"99f054a9-37c8-4443-ae63-4e74fa439795","order_by":1,"name":"Véronique Arnal","email":"","orcid":"","institution":"CEFE Univ Montpellier, CNRS, EPHE, IRD","correspondingAuthor":false,"prefix":"","firstName":"Véronique","middleName":"","lastName":"Arnal","suffix":""},{"id":636735516,"identity":"a3577fbe-ea1e-41e6-82b0-d22092334477","order_by":2,"name":"Benoit Charrasse","email":"","orcid":"","institution":"CEA, DES, IRESNE","correspondingAuthor":false,"prefix":"","firstName":"Benoit","middleName":"","lastName":"Charrasse","suffix":""},{"id":636735518,"identity":"34bb59af-6101-4e90-9803-a345c69b6d3e","order_by":3,"name":"Aurélie Coulon","email":"","orcid":"","institution":"Muséum national d'Histoire naturelle, Centre National de la Recherche Scientifique, Sorbonne Université","correspondingAuthor":false,"prefix":"","firstName":"Aurélie","middleName":"","lastName":"Coulon","suffix":""}],"badges":[],"createdAt":"2026-04-28 08:23:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9550783/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9550783/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109303941,"identity":"0e6fc662-d3b1-4c34-be04-77387eb631b6","added_by":"auto","created_at":"2026-05-15 09:41:08","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":331328,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Genetic sampling within the study zone and (b) location of the study zone within France. The extent of the central study area is represented in blue. Genetic samples are represented according to their status after genotype quality filtering and duplicate filtering (red dots: retained after filtering; white dots: excluded after filtering)\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/37f6e57ddd9b66d630f8a344.png"},{"id":109303938,"identity":"7659423d-ccf8-4fce-9177-c5d51ba730e7","added_by":"auto","created_at":"2026-05-15 09:41:07","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":709948,"visible":true,"origin":"","legend":"\u003cp\u003eResults of the sPCA in the extended study area (ESA). First (a) and second (b) global scores of each individual are shown by the size and color of the squares. White indicates negative values and black positive values. The ESA is represented in light grey, and the central study area (CSA) in dark grey. The main linear landscape elements (roads and rivers) are also shown.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/6e08bd90735d40b93f678683.png"},{"id":109303959,"identity":"b0ee40f2-c1f8-423b-8e96-6e7391fb2666","added_by":"auto","created_at":"2026-05-15 09:41:13","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":985805,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of the genetic groups inferred by STRUCTURE for K = 2. Individual assignments to each group (membership threshold = 0.70) are shown by distinct markers (red diamond; blue triangle), as well as admixed individuals (hollow circle).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/eac5710ecbf7ae70ed21c59d.png"},{"id":109303952,"identity":"30de87e5-4dac-4c2a-b518-e0321a98e12b","added_by":"auto","created_at":"2026-05-15 09:41:09","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":213148,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial genetic autocorrelogram representing the influence of Euclidean distance on genetic spatial autocorrelation. Error bars bound the 95% confidence interval of r by bootstrap resampling and the dotted lines represent the 95% confidence interval of r under random distribution. Distance intervals were set to an even distance of 80-m. Significant departure from random mating distribution is indicated by an asterisk. The grey area represents distance bins with less than 50 pairs. The average number of pairs by distance bins, excluding distance bins with less than 50 pairs, was 239.9 ± 106.8.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/653be28720e325956d726fb6.png"},{"id":109303942,"identity":"1d61169b-31b8-489a-aac9-72ac0a8304d1","added_by":"auto","created_at":"2026-05-15 09:41:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":323286,"visible":true,"origin":"","legend":"\u003cp\u003eThe relationship between estimated resistance and environmental covariates, as estimated by multi-surface optimization in the extended study area, for each top ranking combination in which the covariate is involved. Each variable combination is represented by a different line type and color, and the mean (line) and standard error (ribbon) of each relationship across replicates is represented.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/dc0d5fd649ddb6dcc414fb1a.png"},{"id":109303951,"identity":"936537de-658b-40c1-93e2-313c758cacb1","added_by":"auto","created_at":"2026-05-15 09:41:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":547161,"visible":true,"origin":"","legend":"\u003cp\u003eResistance surface of the extended study area, and its associated uncertainty. The resistance surface was generated by averaging the resistance surfaces generated by the top-performing variable combinations, across replicates. Grey contours delineate areas where the standard deviation of resistance exceeds 1, indicating areas of higher uncertainty in resistance estimates. White points represent the genetic samples used to parameterize the resistance surface. The environmental gradients of the five covariates that contribute to the best combination are shown to illustrate their spatial distribution.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/b1dee4fdf12ff4086be35a00.png"},{"id":109303989,"identity":"a2ec4185-493e-4daf-b23c-ad8aa33b1a5f","added_by":"auto","created_at":"2026-05-15 09:41:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3661588,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/44fb99db-257b-43f4-b48c-70c232250295.pdf"},{"id":109303924,"identity":"5aecaf89-e157-43b2-aa7f-49faa6d5593d","added_by":"auto","created_at":"2026-05-15 09:41:07","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":4253701,"visible":true,"origin":"","legend":"","description":"","filename":"supmatLandGen.docx","url":"https://assets-eu.researchsquare.com/files/rs-9550783/v1/7132f7fefbcb45332f6c0204.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multiscale patterns of functional connectivity across a Mediterranean periurbanized landscape in Timon lepidus","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eLandscape functional connectivity has been demonstrated to be a critical determinant of population structure and dynamics (e.g. Crooks and Sanjayan, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Kool et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Defined as a species-specific measure of an organism\u0026rsquo;s capability to move among resource patches through a landscape (Taylor et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e1993\u003c/span\u003e), it can be assessed for a wide range of movement types, from daily foraging to migration. As a result, connectivity is also process-specific (McClure et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), although it has been more commonly investigated in studies focusing on dispersal (Baguette et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The strength and directionality of dispersal flow between connected habitat patches can affect, among other outcomes, genetic diversity, genetic drift, inbreeding depression, and local colonization and extinction events (Keyghobadi, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). These ecological and evolutionary processes unfold within naturally dynamic landscape systems, in which ecological successions as well as natural perturbations can modify connectivity and, as a result, dispersal trajectories. However, the contemporary acceleration of human modifications of landscape composition and configuration are the main drivers of connectivity alterations (Schlaepfer et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIndeed, the habitat fragmentation caused by linear infrastructures and land-use changes in rural and urban areas directly impacts habitat quality but also reduces its connectedness, i.e. the structural links connecting habitat patches (Miles et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These constraints can materialize as physical barriers, such as dams for river systems (Nilsson et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), or be spatially diffuse, such as artificial light pollution (Korpach et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Seewagen et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). They can impose an increased cost to movements (energy expenditure) and/or increased risks (predation, road collision, competition), resulting in reduced or redirected movements within the landscape (Cassimiro et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Masden et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). For example, the construction of wind farms in Sweden greatly reduced the use by reindeer of their historical migration routes located close to the project development areas (Skarin et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and wolves\u0026rsquo; selection of anthropogenic linear features increased the predation risk for caribou near these features (Whittington et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Conversely, anthropized areas can facilitate movements for some species and human transportations can be used as a vector for dispersal (Crispo et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; e.g. extensive urbanization as dispersal facilitator in feral pigeons: Carlen and Munshi-South, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; linear infrastructure as dispersal corridor in urban foxes: Kimmig et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; human-mediated transport in western black widow spider: Miles et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, and tropical house gecko: Phillips et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, urbanization may only favor persistence and connectivity for a small group of generalist and urbanophile species, to the detriment of native specialists relegated to fragmented suburban and fringe habitats (McKinney, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; McKinney and Lockwood, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e1999\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePreserving connectivity has thus become a major issue for the conservation of functional ecosystems in human-modified landscapes (Lookingbill et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), either through the protection of existing connections or the creation of ecological corridors where connectivity has been disrupted (Rudnick et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). From adapting the location of large-scale infrastructure projects (e.g., offshore windfarm) to maintain transcontinental migratory routes to implementing localized wildlife crossings over roads (Dunn et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Soanes et al., \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), this objective has been integrated into spatial planning and operational strategies at different spatial scales (Hilty et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Sordello et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Indeed, even at small spatial scales, populations can be strongly and rapidly isolated, depending on species dispersal ability and matrix resistance to movements (Beninde et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Schulte et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Fine-scale connectivity analyses can thus contribute to preserve and integrate more effective ecological networks within human-modified environments. Among these environments, the periurban fringes, i.e., landscapes within which urban, rural and natural habitats are intertwined without any one dominating the surface area, represents transitional areas where ecological planning is crucial for conserving connectivity.\u003c/p\u003e \u003cp\u003eIn ecological research, assessing and mapping connectivity with empirical data is a common approach to assess the effects of landscape characteristics and management on the functionality of ecological continuities (Creech et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; McCluskey et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Zeller et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Within the connectivity modeling toolbox, landscape genetics (LG) has emerged as a powerful tool, used to relate landscape characteristics with gene flow, genetic structure and effective dispersal (Balkenhol et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Populations that are well connected tend to exhibit higher genetic similarity, whereas isolated populations often become genetically distinct over time due to limited gene flow. By linking genetic variability (genetic distances between individuals or populations) with landscape composition and configuration, LG offers a direct and quantitative evaluation of functional connectivity (Zeller et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Specifically, some LG approaches allow the inference of landscape resistance to movements, through the optimization of resistance surfaces (e.g., continuous surfaces representing how the landscape impedes movements). This can be done by the use of algorithms that identify the sets of resistance values assigned to landscape variables that maximize the correlations between genetic distances between individuals or populations and the resistance-based distances among them (Bauder et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Kimmig et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePrior studies have demonstrated the utility of landscape genetics in informing urban planning decisions, particularly by quantifying connectivity and identifying barriers and corridors critical to species movements (Baptista et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Brunt and Smith, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Jha and Kremen, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Soanes et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; van Strien et al., \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). However, landscape genetic studies remain scarce at the scale of periurban development units, mainly due to the associated time and cost constraints of this type of studies (LaPoint et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Consequently, connectivity assessments integrated into environmental impact assessments (EIA) of infrastructure projects are most often conducted at a single spatial scale and rely on model-based or expert-driven approaches, which may overlook scale-dependent processes shaping gene flow (Cumming and Tavares, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Accounting for functional connectivity across multiple spatial scales is therefore a key challenge to better capture the ecological processes underlying species movements and population genetic structure, and to improve the relevance of connectivity assessments for spatial planning.\u003c/p\u003e \u003cp\u003eIn this study, we focus on the multiscale effects of landscape on gene flow and population genetic structure of the ocellated lizard \u003cem\u003eTimon lepidus\u003c/em\u003e (Daudin, 1802). As a low-mobility and cavity-dependent species able to inhabit anthropized areas, in which it exploits cavities embedded in the built environment (dry stone-walls, cracks under concrete surfaces, underground piping systems, building fa\u0026ccedil;ades, etc\u0026hellip;), this terrestrial lizard is an interesting species to study the effects of urbanization level and configuration on gene flow and connectivity (Ludot et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Renet et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Space use and movement patterns of this species during its active reproductive period have been studied in various natural areas. However, little is known about its movement ability in periurban landscapes, and, more generally, about its dispersal capacity (Grillet et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Renet et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Thirion et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur main objectives were to identify environmental drivers of gene flow across a periurban landscape within which the species is continuously distributed, and across a broader area, including this periurban landscape but also the surrounding Mediterranean rural landscape, in which the species is discontinuously distributed. We hypothesized that, at the broad spatial scale, landscape resistance would have a greater influence on \u003cem\u003eTimon lepidus\u003c/em\u003e gene flow than distance alone, with genetic structure primarily shaped by rivers, as well as by large Mediterranean forested areas, agricultural fields and areas with low shelter densities. At the finer periurban scale, throughout the spatially continuous distribution, we predicted low levels of genetic structure, with urban cover and the road network being the main factors hindering gene flow, while the density of linear infrastructures providing anthropogenic shelters could facilitate it.\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study species and area\u003c/h2\u003e \u003cp\u003eThe ocellated lizard range spreads from the Iberian Peninsula to northwestern Italy. In France, in its eastern Mediterranean geographic distribution, it favors hilly landscapes of semi-open garrigue shrubland, rocky steppes and dry grassland, as well as traditional agricultural lands, such as olive orchards and vineyards, belted by stonewalls and embankments. The ocellated lizard is an opportunistic forager that predates directly in the neighborhood of its shelter and which spends a lot of time thermoregulating at its entrance. Its home ranges are hence small (~\u0026thinsp;1 ha), centered around a dense network of fewer than a dozen shelters in which it hibernates (Cheylan et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Our study area is centered around the CEA Cadarache research center, in southern France (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). It is a 9-km\u0026sup2; periurban site comprising 2-km\u0026sup2; of impervious areas, and extends across 220-km\u0026sup2; of Mediterranean forests, shrublands, agricultural fields and small cities on the left bank of the Durance River. The periurban research center is known to host an important population of ocellated lizards, that opportunistically use anthropic cavities embedded in the built environment.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Sample collection\u003c/h2\u003e \u003cp\u003eWe established an individual-based multi-scale sampling scheme which includes (i) a continuous sampling within the area covering the Cadarache periurban site and contiguous favorable habitat (hereafter the Central Study Area, CSA), (ii) a discontinuous sampling in an approximate 10-km wide radius around the Cadarache plateau (hereafter called \u0026ldquo;extended study area (ESA)\u0026rdquo;), targeting known species presence locations. The sampling campaigns occurred from April to August 2023 and 2024.\u003c/p\u003e \u003cp\u003eDNA samples were either obtained from buccal swabbing upon individual capture, from scats and moults found near occupied shelters or from tissues collected off road kills. The locations of all samples were recorded (using the qfield application on mobile device) as UTM (Universal Transverse Mercator) coordinates. To capture individuals, we used tunnel traps positioned at the entrance of shelters consequently to lizard sightings, although we also resorted to opportunistic hand capture. The traps were 80-mm width PVC tubes of 50-cm to 1-meter length (according to shelter configuration) and were equipped with a one-way entrance. The upper-side of the tube was cut and replaced by a 50-mm mesh net except for the entrance edge, which was left whole. The net also covered the exit of the tube. This design impeded the lizards going inside the tube to leave it, while letting the light flow inside. Following set-up, the traps were checked every 30-min and were removed after 4 to 6 unsuccessful checks depending on climate conditions. Upon capture, we used a sterile swab to collect DNA (Puritan Medical Products, Sterile Polyester Tipped Applicator) and stored samples in a DNA stabilization solution (Zymo Research, DNA/RNA Shield). To avoid sampling twice the same individual, we took photographs of the right and left flanks of all individuals for identification. At a later stage, we also checked all genotypes for duplicates (see 2.4). Scat samples were either found at sight during the searches of individuals or during specific sessions with the assistance of a detection dog trained to detect ocellated lizard scats (Olivier et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The scat samples were filtered to remove feces that did not match the size criteria to distinguish them from \u003cem\u003eLacerta bilineata\u003c/em\u003e feces. All samples were stored at \u0026minus;\u0026thinsp;20\u0026deg;C upon collection until shipment in dry ice to Microsynth ecogenics (Balgach, Switzerland) for genotyping. Capture permits were delivered by the French Government (permit numbers: n\u0026deg;2022-094-004).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Genotyping\u003c/h2\u003e \u003cp\u003eDNA extraction of swabs, moults and tissues was performed with Qiagen DNeasy Blood and tissue kits and DNA extracts from tissues were diluted to 1 ng/\u0026micro;l with 10 mmol Tris-HCl pH 8 before used in PCR reactions. DNA extraction of scat samples was performed with ZymoBIOMICS kit. All individuals were genotyped at 15 microsatellite loci, using markers TL_303574, TL_361715, TL_684747, TL_265456, TL_966180, TL_1079095, TL_276028, TL_619091, TL_742562, TL_1006729, TL_750235, TL_741815, TL_985672, TL_748432, TL_1291540 (Online Resource 1 and 2). Primers were labelled with FAM, ATTO-532, ATTO-550, ATTO-565. The multiplex PCR amplification was performed in a total volume of 10 \u0026micro;l using a mix of 1.5 \u0026micro;l double distilled water, 5 \u0026micro;l Multiplex Master Mix (stock concentration 2 x; Qiagen catalog no. 206143), 0.3 \u0026micro;l of each the forward and reverse primer (stock concentration 10 \u0026micro;M), and 2\u0026ndash;10 ng of genomic DNA (stock concentration 1\u0026ndash;5 ng/\u0026micro;l). The cycling protocol started with an initial denaturation step of 15 min at 95\u0026deg;C, followed by 40 cycles with 30 s at 94\u0026deg;C, 90 s at 56\u0026deg;C, and 60 s at 72\u0026deg;C, and ended with a final extension step of 30 min at 72\u0026deg;C. Amplifications were performed on a Programmable Thermal Controller (Sensoquest Labcycler). The PCR products were run on an ABI 3730XL DNA Analyzer (Thermo Fisher Scientific) and fragment lengths were scored with the Size Standard GeneScan LIZ500 (Applied Biosystems).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Data cleaning\u003c/h2\u003e \u003cp\u003eDNA amplification was not equally successful depending on sample types, with moult and scat samples having a lower genotype completion rate (Online Resource 3). Hence, as a trade-off between missing information and spatial distribution of samples (some locations being predominantly represented by moult and scat samples), we applied a cut-off threshold allowing a 0.3 missing rate for genotypes. Due to financial constraints, we were not able to perform multiple genotyping of scat and moult samples, but we performed several analyses to ensure that they were not diverging from buccal and tissue samples, and could therefore be considered reliable (Online Resource 4). We computed the triadic likelihood relatedness index with COANCESTRY through the R package related using a 0.7 threshold score to identify duplicates and we removed them from the final dataset (Wang, \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The frequencies of null alleles and their associated confidence intervals were estimated using the method of Chakraborty et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1994\u003c/span\u003e) in PopGenReport (Adamack and Gruber, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). We estimated linkage disequilibrium (LD) by calculating the pairwise (rd) index of association with 500 permutations in poppr (Agapow and Burt, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Kamvar et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). We adjusted the p-values using the False Discovery Rate method (Benjamini and Hochberg, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1995\u003c/span\u003e) to account for multiple testing. Deviations from Hardy-Weinberg equilibrium (HWE) were calculated using the exact test based on Monte Carlo permutations in the pegas package (Paradis, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Genetic diversity and structure\u003c/h2\u003e \u003cp\u003eFor each locus, we calculated the number of alleles (A), the effective number of alleles (Ae), and the observed (Ho) and expected heterozygosity (He), using the adegenet package (Jombart, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo identify genetic structure at the two spatial scales, we conducted a spatial principal component analysis (sPCA, Jombart et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), a cluster analysis (STRUCTURE, Pritchard et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) and a spatial autocorrelation and isolation by distance (IBD) analysis. The sPCA is a spatial multivariate method that describes genetic structure by measuring spatial autocorrelation and genetic variability in the dataset without assigning individuals to clusters. In contrast, STRUCTURE is an aspatial Bayesian model-based clustering method. We applied sPCA using the spca function implemented in the ADEGENET package (Jombart, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). We set the connection network using the inverse of pairwise Euclidean distances between individuals and the significance of local and global spatial structures was tested performing eigenvalue tests (nperm\u0026thinsp;=\u0026thinsp;9999, Montano and Jombart, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). We used the admixture model with correlated allele frequencies in STRUCTURE and ran Markov chain Monte Carlo simulations with 500 000 iterations after 50 000 iterations of burn-in. A preliminary testing phase with 100 000 iterations after 50 000 iterations of burn-in was used to select the \u003cem\u003eALPHAPROPSD\u003c/em\u003e and \u003cem\u003eLamba\u003c/em\u003e parameters, setting K from 1 to 5 clusters and using 3 replications. We then set the number of clusters K from 1 to 10, with 10 replications for each K (ALPHAPROPSD\u0026thinsp;=\u0026thinsp;1.0, lambda\u0026thinsp;=\u0026thinsp;0.640). The best K value was determined based on the delta K method (Evanno et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). The replicate runs with the best K value were then averaged using CLUMPP (Jakobsson and Rosenberg, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), and individuals were assigned to clusters with a 0.6 threshold membership. We applied sPCA and STRUCTURE analyses at both the central and extended study areas.\u003c/p\u003e \u003cp\u003eThe spatial autocorrelation analysis was performed in the GenAlEx software (Peakall and Smouse, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), following the approach developed by Smouse and Peakall (\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). GenAlEx computes an autocorrelation coefficient (r) using pairwise genetic and geographic distance matrices. The observed distribution of the (r\u003csub\u003eobs\u003c/sub\u003e) coefficient is compared to a generated distribution (r\u003csub\u003erd\u003c/sub\u003e) under the assumption of random mating, using permutations among individuals, to detect spatial structure at specified distance intervals. We used distance intervals of 80-m, which approximates a mean ocellated lizard home range (Cheylan et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Renet et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Ludot et al. \u003cem\u003ein prep\u003c/em\u003e). Only the CSA was considered for this analysis as the sparser and discontinuous sampling at the ESA did not allow a robust estimation of spatial autocorrelation across distance classes. Significance was calculated based on 10,000 permutations and bootstraps.\u003c/p\u003e \u003cp\u003eFinally, we tested for the presence of isolation by distance (IBD) at both sampling scales with maximum likelihood population effects (MLPE) mixed-effects models, using the R package resistanceGA (Peterman, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Inter-individual genetic distances were based on a PCA-based metric (Shirk et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) using the \u003cem\u003edudi.pca\u003c/em\u003e function from the R package ade4 (Thioulouse et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and we computed geographic pairwise Euclidean distances in SPAGEDI (Hardy and Vekemans, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Geographic distances were log-transformed.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Effects of environmental features on gene flow\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.6.1 Landscape covariates\u003c/h2\u003e \u003cp\u003eUsing a 10-m resolution land cover map (CoSIA IGN, 2018\u0026ndash;2023, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://geoservices.ign.fr/telechargement-api/COSIA\u003c/span\u003e\u003cspan address=\"https://geoservices.ign.fr/telechargement-api/COSIA\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), we created a categorical land cover surface with seven thematic classes (urban, grassland, agricultural, shrub, tree, water and bare ground) by merging the 15 original classes based on shared properties relatively to ocellated lizard habitat requirements (Online Resource 5a). We used OpenStreetMap (2015) to create a categorical road surface combining motorways, primary, secondary and tertiary roads; and a categorical river surface. We dilated the linear features (rivers and roads) by 10 m to avoid breaks caused by the rasterizing of the vector data. We used a 1-m elevation raster (RGE ALTI\u0026reg;, geoservices.ign.fr) to compute a 10-m continuous slope surface using the package terra (Hijmans, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). We used a 10-m resolution surface representing the occurrence probabilities of natural shelters. This surface was built using a habitat suitability distribution modeling approach, as described in a previous study investigating ocellated lizard\u0026rsquo;s suitable habitat within the same periurban landscape (Ludot et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Lastly, we used a 10-m resolution raster representing an underground electrical infrastructure running through the CSA. This linear feature has been described as positively associated with the species presence due to its numerous cavities (Ludot et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and as such could facilitated dispersal movements. We used these 10-m resolution rasters for the analyses conducted in the CSA, but to minimize the computational time of the analyses conducted in the ESA, we converted these surfaces to 40-m pixel rasters. We used these surfaces to calculate several landscape variables, to be used in the analysis of landscape resistance to gene flow. To do so, we computed the percentage of each land cover class and of roads in 80, 160 and 320-m buffers around each pixel, and the mean of slope and natural shelter occurrence probabilities in the same buffers. As mentioned above, 80-m is the approximate average diameter of an ocellated lizard home range. The 160 and 320-m distances approximated respectively the mean and maximum distances from home-range centroids to the farthest location of adult male ocellated lizards in Cadarache research center (Ludot et al., \u003cem\u003ein prep\u003c/em\u003e.). We also computed a continuous surface representing the distance to the closest rivers. For clarity, some variables are abbreviated in the tables and figures; the abbreviations for each variable are listed in Online Resource 5b.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e2.6.2 Resistance surface optimization\u003c/h2\u003e \u003cp\u003eTo evaluate functional connectivity, we used the RESISTANCEGA package in R (Peterman, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) to optimize resistance surfaces at the two spatial scales. Resistance surfaces represent estimates of landscape resistance to movements, associated with landscape structure and composition. The RESISTANCEGA workflow calculates pairwise resistance distances between individuals and uses a genetic algorithm from the GA package (Scrucca, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) to iteratively calibrate resistance values. The optimization uses the log-likelihood (or AIC) from a linear mixed-effects model with a MLPE parameterization to maximize the fit of resistance surfaces to the genetic data. To calculate the pairwise resistance distances between individual sampling locations, we used the random-walk commute-time method from the GDISTANCE package (Etten, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; McRae et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). This function models the landscape (our continuous raster surfaces) as a network of connected cells, where movement between adjacent cells occurs probabilistically according to their resistance values. The commute-time distance represents the expected cost of all possible random-walk paths connecting two sampling locations. This measure is conceptually similar to the effective resistance computed by circuit theory and both were shown to be highly correlated. First, we used the SS_optim function to conduct single-surface optimizations for each variable independently across four buffer sizes (0, 80, 160, and 320-m), comparing them against an isolation-by-distance (IBD) model based on Euclidean geographic distance. A variable was retained if its best model had an AICc at least 2 units lower than that of the IBD model and the best-performing buffer size was selected for each retained variable. We used the \u003cem\u003eremoveCollinearity\u003c/em\u003e function to assess the pairwise correlations among the retained variables and proceeded to variable exclusion if correlation was larger than 0.7. Second, we performed a multi-surface optimization, combining the retained variables at their selected buffer size, and used AICc to compare performance across all the existing combinations of the retained variables. We performed five independent optimization replicates to account for the stochastic nature of the resistanceGA optimization algorithm and to ensure the robustness of model selection. For both steps, all types of relationships between resistance and the environmental variables available in resistanceGA were tested.\u003c/p\u003e \u003cp\u003eAll analyses were based on the continuous raster surfaces described above.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Sample collection and genetic data\u003c/h2\u003e \u003cp\u003eWe collected 454 samples, among which 151 buccal swabs, 274 scats, 20 moults, and 9 tissue samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). In natural habitat, outside of the periurban research center, 85.15% of scat samples (138 out of 162) were found with the assistance of the detection dog. Among all the genotypes, pairwise permutation tests did not reveal any significant linkage disequilibrium (Online Resource 6). Deviations from Hardy\u0026ndash;Weinberg equilibrium were observed in seven loci and four of them had a significant likelihood of null alleles (Online Resource 7). Due to the high frequency of null alleles detected for \u003cem\u003eTL_361715\u003c/em\u003e (0.20), and its high percentage of missing data (0.14), we excluded this locus and therefore based our analyses on the 14 remaining microsatellite loci. The other loci had null allele frequencies\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;0.08, which were considered as low enough to keep them in the analyses. Following removal of \u003cem\u003eTL_361715\u003c/em\u003e, secondary genotype missing rate cut-off and duplicate filtering, we retained 202 genotyped samples in our final dataset, with an overall rate of missing data of 3.35%. Among these samples, 170 occurred inside the CSA (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Genetic diversity and structure\u003c/h2\u003e \u003cp\u003eOver the 202 samples, the number of alleles in the 14 retained microsatellite loci ranged from 3 to 8 (mean\u0026thinsp;=\u0026thinsp;4.86), effective number of alleles was 1.05 to 2.83 (mean\u0026thinsp;=\u0026thinsp;2.12), observed heterozygosity was 0.02 to 0.66 (mean\u0026thinsp;=\u0026thinsp;0.42) and expected heterozygosity was 0.02 to 0.69 (mean\u0026thinsp;=\u0026thinsp;0.46).\u003c/p\u003e \u003cp\u003eIn the extended study area, the sPCA revealed a significant global spatial structure (global test, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while no local structure was detected (local test, p\u0026thinsp;=\u0026thinsp;0.998). We retained the first three positive axes (0.10 and 0.09) and no negative axis. For the first global score, positive values were concentrated in the south of the ESA and the southwestern part of the research center, and negative values were in the north and east of the ESA and in the northeast of the research center (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). For the second global score, no clear spatial pattern emerged (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the central study area, the sPCA revealed a significant global spatial structure (global test, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while no local structure was detected (local test, p\u0026thinsp;=\u0026thinsp;0.984). We retained the three first positive axes (0.11 and 0.10) and no negative axis. For the first global score, there was no strong spatial pattern, although there were only negative scores in the southern part of the CSA (Online Resource 8). For the second global score, there was no strong spatial pattern either (Online Resource 8).\u003c/p\u003e \u003cp\u003eIn the ESA, the optimal number of genetic clusters inferred by the Evanno method applied to STRUCTURE results was 4 (ΔK\u0026thinsp;=\u0026thinsp;7.18, mean likelihood = -4735.89; Online Resource 9). However, using a threshold assignment membership of 0.60, most individuals (n\u0026thinsp;=\u0026thinsp;174, 86%) were classified as admixed, and one of the four clusters had no individual assigned to it (Online Resource 10). Considering STRUCTURE might have inferred ghost clusters (i.e. a cluster with no individual, Guillot et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), we looked at K\u0026thinsp;=\u0026thinsp;2, which obtained the second highest delta K score and the highest mean likelihood with low variability among runs (ΔK\u0026thinsp;=\u0026thinsp;5.92, mean likelihood = -4716.34, Online Resource 9). Considering the assignment probability bar plot (Online Resource 10), we applied a more severe threshold of 0.70, to avoid including individuals with uncertain cluster memberships. There was no clear spatial boundary between the two clusters but rather a gradient, with the individuals sampled in the east being assigned to one group, those in the west to the other, and the central zone (Cadarache research center) with individuals from both groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the CSA, the highest mean likelihood, with low variability among runs, was obtained for K\u0026thinsp;=\u0026thinsp;1 (mean likelihood = -3851.63, Online Resource 11). Given that, by construction, the Evanno method cannot infer K\u0026thinsp;=\u0026thinsp;1 as the best solution (Evanno et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), we did not apply it to STRUCTURE results for this spatial scale and considered that the most likely number of genetic clusters in the CSA was 1.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe spatial autocorrelation analysis revealed fine-scale positive autocorrelations up to 720-m (r\u0026thinsp;=\u0026thinsp;0.028, P\u0026thinsp;=\u0026thinsp;0.006, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) and negative spatial autocorrelation arose from 3600-m (r = -0.027, P\u0026thinsp;=\u0026thinsp;0.034). Finally, we found a significant positive effect of geographic distance on genetic distance at both spatial scales (ESA: slope coeff. = 0.033; CSA: slope coeff. = 0.026; Online Resources 12 and 13).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Effects of environmental features on gene flow\u003c/h2\u003e \u003cp\u003e \u003cem\u003eExtended study area\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eDuring the single-surface optimization step, seven variables generated resistance distances that exceeded the performance of the IBD model (∆AICc\u0026thinsp;\u0026gt;\u0026thinsp;2, Online Resource 14): percentage of water cover, probability of natural shelter occurrence, mean slope, percentage of roads and agricultural surfaces, all within a 320-m buffer; percentage of tree cover within a 160-m buffer; and distance to rivers. None of these variables were correlated (correlation coefficient\u0026thinsp;\u0026lt;\u0026thinsp;0.7).\u003c/p\u003e \u003cp\u003eTesting all combinations of these seven variables in the multi-surface optimization, the combinations that best explained inter-individual genetic distances were quite consistent across the five replicates, with seven combinations ranking as equivalently best performing (∆AICc\u0026thinsp;\u0026lt;\u0026thinsp;2) (Online Resource 15, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Five of these combinations were present in all replicates, one in four out of five replicates, and one in three replicates (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The part of variance of inter-individual genetics distances explained by the resistance distance varied between 0.10 and 0.12 in the top-performing models.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of the most performant variable combinations (∆AICc\u0026thinsp;\u0026lt;\u0026thinsp;2) during the multi-surface optimization step, and mean contribution of each covariate, across the 5 replicates, in the extended study area. N: number of covariates in the combination, x/5: number of replicates that ranked the combination within the most performant among the 5 replicates. Within each replicate, variable contribution was estimated by the genetic algorithm during the multi-surface optimization and reflects the relative influence of each variable on the composite resistance distance.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c8\" namest=\"c4\"\u003e \u003cp\u003emean contribution across replicates\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"1\" nameend=\"c9\" namest=\"c9\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecombination\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003epresence across replicates\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eagricultural\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eriver\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eshelter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eroad\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eslope\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"1\" nameend=\"c9\" namest=\"c9\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eagricultural * river\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eagricultural * river * shelter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eriver * shelter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eagricultural * river * slope\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eagricultural * shelter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eroad * slope\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eriver * slope\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTwo variables never appeared in any top combination: water percentage and tree percentage. Conversely, the percentage of agricultural land, the distance to river, and the probability of natural shelter occurrence were the variables most frequently represented across all combinations, with slope and road being the least represented respectively. Across replicates, the relative contributions of variables to resistance surfaces varied among combinations (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Natural shelter occurrence consistently had the highest contribution when included, with mean values ranging from 0.56 to 0.76. Distance to rivers and agricultural land contributed to the same amount when appearing both within the same combination, ranging from 0.20 to 0.55. Slope contributed twice as much (0.49) as river and agricultural land (0.28 and 0.24, respectively) in their three-variable combination, but contributed similarly as road and river in their respective two-variable combinations. The relationship between variables and estimated resistance was mostly consistent across replicates and models but the effect size varied (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Across most combinations, the relationship between agricultural cover and scaled resistance was asymptotic, with resistance increasing rapidly and reaching a plateau at approximately 70% scaled resistance when agricultural cover reached around 15%. In contrast, for the river-agricultural combination, the relationship was linear, with increasing agricultural cover associated with lower resistance compared to other combinations. Conversely, increases in occurrence probability of natural shelters, slope, road surface and distance to the river were associated with decreased resistance (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). For natural shelter occurrence, resistance declined non-linearly across all combinations, with a steep decrease at low probabilities followed by a gradual flattening at higher values. Similarly, resistance decreased with increasing distance to rivers, showing a rapid decline at short distances and reaching an asymptote at around 1.5 km. Increasing slope was associated with lower resistance across combinations, but the relationship was less consistent. Finally, road surface cover exhibited a monotonic decrease in resistance, indicating lower resistance at higher levels of road cover within the road \u0026ndash; slope combination.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs no combination distinguished itself from the others in terms of log-likelihood, R\u0026sup2; and AICc scores (Online Resource 15), we used all the retained combinations to construct the resistance surface of the study area. First, for each replicate, we constructed a resistance surface by simply scaling and averaging the resistance surfaces from each of the best performing combinations (∆AICc\u0026thinsp;\u0026lt;\u0026thinsp;2). We then created a final resistance surface as the mean of the resistance surfaces across replicates. We also created a raster of resistance uncertainty by computing the standard deviation of the resistance surfaces across replicates (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eCentral study area\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eIn the CSA, five variables, when taken alone, generated resistance distances that exceeded the performance of the IBD model (∆AICc\u0026thinsp;\u0026gt;\u0026thinsp;2, Online Resource 16): probability of natural shelter occurrence, percentage of road and agricultural surfaces within a 320-m buffer; percentage of tile network within a 160-m buffer; and distance to the river. None of these variables were correlated (correlation coefficient\u0026thinsp;\u0026lt;\u0026thinsp;0.7). The combination analysis yielded the same four combinations across the five replicates. These included two single-variable surface combinations: agricultural surface and distance to river; and 2 two-variable combinations: agricultural \u0026ndash; distance to river and agricultural \u0026ndash; natural shelter occurrence (Online Resource 17). Among the two-variable combinations, the agricultural surface variable always had the lowest relative contribution across replicates compared to the distance to river and natural shelter occurrence variables (agricultural mean contribution: 0.37 and 0.36 for each combination respectively, Online Resource 18). Of the five variables that were selected in the single optimization step, two variables never appeared in any top combination: the proportion of surface covered by the tile network, and the proportion of surface covered by roads. The part of variance of inter-individual genetic distances explained by the resistance distance varied between 0.09 and 0.11 in the top-performing models (Online Resource 17). Increasing agricultural cover was associated with higher resistance. In contrast, increasing natural shelter occurrence and greater distance to rivers were associated with lower resistance (Online Resource 19). The scaled resistance assigned to values of distance to river and of the occurrence probability of natural shelters was overall much lower in the CSA compared to the ESA. The same method as for the ESA model was used to construct the resistance surface. Within the boundaries of the CSA, the resistance surfaces produced by the two models were similar (Online Resource 20).\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eWe used landscape genetics to characterize at two spatial scales the impact of a periurban Mediterranean landscape on the genetic connectivity of a population of ocellated lizards. Overall, we observed a weak genetic structure at both spatial scales, suggesting an historically well-connected and/or continuously distributed population. There was significant positive genetic autocorrelation at short distance classes (\u0026lt;\u0026thinsp;800m), consistent with limited natal dispersal and local gene flow. Resistance-based models outperformed isolation by distance alone, indicating that landscape features influence gene flow, although the variance explained by resistance remained modest. These results partly support our hypotheses with landscape resistance affecting gene flow at broader scales, but did not support an impact of the urban environment on functional connectivity at the periurban scale.\u003c/p\u003e \u003cp\u003e \u003cem\u003eGenotyping\u003c/em\u003e \u003c/p\u003e \u003cp\u003eWe originally developed our sampling strategy based on the fact that we would collect faecal samples, and hence developed microsatellite markers that were likely to persist in this type of samples. However, faecal genotyping revealed poor results, despite the preliminary genotyping analyses yielding encouraging results, with scat samples exhibiting a high amplification rate and high fidelity with their buccal sample counterparts from the same individual. These preliminary results were obtained using fresh scat samples collected during ocellated lizard handling following capture. It was difficult to date the scat samples during fieldwork, and the samples were rarely fresh, having likely been exposed for several days to external climatic conditions (sun, drought and rain) that could have degraded the DNA before collection. This led us to develop a sampling strategy aimed at capturing individuals to collect more reliable buccal samples. This type of samples would have eventually allowed the use of SNP markers, which would have led to a finer distinction between individuals, as the richness of our microsatellite dataset in terms of allelic richness was low to moderate.\u003c/p\u003e \u003cp\u003e \u003cem\u003eSpatial genetic structure\u003c/em\u003e \u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eOverall, our results did not reveal discrete genetic clusters but instead indicated weak and continuous spatial genetic structure consistent with isolation by distance. Both sPCA analyses detected significant global structure without local structure at either spatial scale, suggesting broad-scale genetic gradients rather than genetic discontinuities. Likewise, STRUCTURE did not identify well-defined genetic clusters. Most individuals were highly admixed, especially within the central study area, and inferred clusters lacked clear spatial boundaries. Spatial autocorrelation analyses revealed significant positive genetic structure at a distance of around 700 m, indicating that individuals within a radius of approximately 1 km are more genetically similar than random. This fine-scale spatial genetic structure is consistent with the species' short-range spatial ecology (Renet et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ludot et al., forthcoming). Beyond approximately 3.5 km, however, spatial autocorrelation became significantly negative, suggesting that individuals separated by several kilometers belong to different genetic neighbourhoods. This spatial signal is consistent with the isolation-by-distance pattern observed in the aspatial and spatial genetic structure analyses and spatial genetic patterns driven by limited dispersal, with gene flow declining gradually with distance. As it represents the first genetic estimate of dispersal ability for the species, these results provide a valuable baseline for identifying appropriate spatial scales for habitat management, corridor design, and impact assessment. Ensuring functional connectivity at the kilometer scale might be particularly important for sustaining population resilience in the face of ongoing landscape changes.\u003c/p\u003e \u003cp\u003eTaken together, the analyses of genetic structure show that the periurban area did not appear to generate a genetic break, suggesting that the level and configuration of urbanization encountered throughout the site has not restrained functional connectivity. The literature shows divergent results regarding genetic differentiation among fragmented urban populations of other lizard species. While some studies have found evidence of strong and rapid genetic differentiation (\u0026lt;\u0026thinsp;40 years) between populations inhabiting remaining suitable patches following urbanization (e.g., Delaney et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2010\u003c/span\u003e in \u003cem\u003eChamaea fasciata, Sceloporus occidentalis, Plestiodon skiltonianus\u003c/em\u003e and \u003cem\u003eUta stansburiana\u003c/em\u003e; Wenner et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2022\u003c/span\u003e in in \u003cem\u003ePhrynosoma blainvilli\u003c/em\u003e), there are also reports of the urban environment having only a weak effect on genetic structure, even in conditions more urbanized than those encountered in our study (e.g. Beninde et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e in \u003cem\u003ePodarcis muralis\u003c/em\u003e; Virens et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2015\u003c/span\u003e in \u003cem\u003eCtenotus fallens\u003c/em\u003e). These latter results have two potential explanations: either a time lag between the urbanization process and the observed genetic signal as detected by the genetic markers and analyses involved; or the ability of those species to cope with urban conditions during dispersal. At our study site, periurbanization dates back 60 years and replaced an already anthropogenic agricultural landscape, structured by stone walls and other linear features. Such land-use continuity and moderate levels of urbanization may have limited the disruption of dispersal throughout the site. In the ocellated lizard, the largest movements studied are performed by males during the reproductive season, and little is known about its natal dispersal. However, for this cavity-dependent species, routine movements heavily rely on a shelter network. It is probably also true, though to a lower extent, for dispersal movements. Moderate urbanisation levels, which provide human-made shelter resources, hence do not necessarily degrade habitat connectivity for this species.\u003c/p\u003e \u003cp\u003e \u003cem\u003eEnvironmental drivers of connectivity\u003c/em\u003e \u003c/p\u003e \u003cp\u003eOur analyses of the resistance of landscape features to gene flow revealed that the environmental drivers of connectivity slightly differed between spatial scales but were mainly consistent.\u003c/p\u003e \u003cp\u003eAt the broad ESA scale, resistance modelling suggested that rivers and agricultural cover contributed to limiting gene flow, particularly when considered jointly. Rivers are often cited as an impediment to gene flow in terrestrial lizards (Beninde et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Oliveira et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Pounds and Jackson, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e1981\u003c/span\u003e). While anecdotal observations of ocellated lizards fleeing underwater suggest that the species can swim, this is most likely a marginal antipredation strategy (Cheylan et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). This effect is likely reinforced by the fact that riverbanks are often bordered by extensive agricultural fields in our study area. Rather than acting as absolute barriers, rivers and surrounding agricultural matrices therefore appear to contribute to a gradual reduction in connectivity. Indeed, resistance increased sharply with agricultural cover, suggesting that agricultural areas act as a resistant matrix once they surpass relatively low thresholds. Some traditional Mediterranean agricultural structures (e.g., stone walls and dry-stone buildings) provide anthropogenic cavities. However, these old structures are often abandoned and are disappearing more and more (L\u0026oacute;pez-Est\u0026eacute;banez et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In some parts of the ocellated lizard range, the restoration of dry-stone walls is a priority conservation action by local planning (LPO Auvergne, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://auvergne-rhone-alpes.lpo.fr/projets/lezard-ocelle/\u003c/span\u003e\u003cspan address=\"https://auvergne-rhone-alpes.lpo.fr/projets/lezard-ocelle/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The observed resistance pattern likely reflects contrasted movement conditions within agricultural landscapes. Edges and trails through agricultural plots may provide linear features and shelter opportunities that facilitate movement, whereas crossing the interior of large fields is likely more costly due to the lack of refuges, vegetation structural complexity and prey resources. Through agricultural land consolidation, the merging of agricultural plots has reduced edge density, removed hedgerows and stonewalls delimitating plots, and increased field size, potentially decreasing permeability for cavity-dependent species (Colucci, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Michael et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In their study, Guerrero-Casado et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) showed that reptile abundance and species richness were markedly lower in modern intensive trellis vineyards than in traditional vineyards, with \u003cem\u003eTimon lepidus\u003c/em\u003e absent from trellis systems.\u003c/p\u003e \u003cp\u003eConversely, the availability of natural shelters strongly promoted connectivity, with areas of low shelter probability being associated with high resistance. Across combinations in which it was retained, shelter occurrence showed the highest relative contribution to resistance, and resistance declined sharply even at low probabilities of shelter presence. This pattern indicates that small increases in cavity availability can enhance permeability of the landscape, highlighting the importance of fine-scale structural features for movements in this species. In contrast, variables describing general land cover, such as shrubs or tree cover, were never retained in the top-performing models, suggesting that structural microhabitat features outweigh coarse land-cover descriptors in shaping functional connectivity. Similarly, steeper terrain facilitated gene flow, likely due to the greater density of natural cavities in such areas. Roads also contributed to connectivity at this scale, which might be due to their rocky and open embankment that can host cavities. Together, these results align with the species' strong dependence on cavity-rich habitats.\u003c/p\u003e \u003cp\u003eAt the periurban CSA scale, however, no clear resistance signal emerged throughout the research center on the optimized resistance surface, suggesting that moderate urbanization does not markedly disrupt connectivity. The absence of a strong effect may also reflect the area's relatively recent development, which may not yet have produced detectable genetic consequences. The main drivers of gene flow were similar to those operating at the broader ESA scale, but were confined to the CSA's periphery rather than its urbanized core. Two weak signals emerged during the single-surface optimization stage, indicating that anthropogenic linear features (roads and tile networks) acted as local facilitators of connectivity, although these features were not retained in the final multi-surface model. At the CSA scale, roads also contributed to connectivity, likely not because they promote direct crossing, but because individuals may move along their embankments and through managed open vegetation on roadsides, due to the research center legal obligation to clear vegetation around infrastructures. Future studies could explicitly account for road use intensity, road width, and management practices to better disentangle the role of roads as barriers and potential movement corridors. As expected, given the presence of anthropogenic cavities along its entire length, the tile network, which is present only in the peri-urban landscape, also showed a positive correlation with gene flow.\u003c/p\u003e \u003cp\u003e \u003cem\u003eManagement implications\u003c/em\u003e \u003c/p\u003e \u003cp\u003eOur results suggest that the landscape connectivity of \u003cem\u003eTimon lepidus\u003c/em\u003e is more dependent on the fine-scale distribution of cavities across the landscape than on broad land-cover categories. Effective conservation measures should therefore prioritize the preservation and restoration of structures that provide cavities, such as dry-stone walls, terraces and embankments, as well as other traditional elements of Mediterranean agro-pastoral systems. This is particularly urgent given the documented decline of dry-stone wall networks in many Mediterranean regions, which has prompted EU-level restoration initiatives such as the LIFE StoneWall4Life project. At the same time, the resistance associated with large, homogeneous agricultural fields highlights the importance of promoting heterogeneity and microhabitat diversity within intensively cultivated areas. In periurban landscapes, our findings suggest that anthropogenic structures can partially compensate for the loss of natural shelters, and that retaining artificial microhabitat features (e.g., cavities along road embankments or utility infrastructures) can help to maintain functional connectivity. Integrating biodiversity objectives with the preservation of cultural landscapes and traditional land-use practices may substantially improve the social acceptability and durability of conservation actions (Bridgewater and Rotherham, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Rollo, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Bio-cultural heritage frameworks highlight that many species depend on long-maintained anthropogenic structures, and that conserving these elements supports both ecological connectivity and cultural identity (Rotherham, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Aligning shelter restoration with heritage preservation thus represents a promising pathway for participatory and socially grounded conservation planning.\u003c/p\u003e \u003cp\u003eAdditionally, our results provide an interesting insight into the limited dispersal ability of the species at a fine scale, which nevertheless leads to a gradual genetic structuring. Although the genetic structure did not seem to be impacted by urbanization, given the short-range dispersal capacities of this species, the emergence of a barrier could have a strong effect on population genetic structure. Overall, as the species' spatial ecology depends on a network of shelters, identifying existing shelter resources, to be maintained or rehabilitated, as well as areas lacking shelter resources, is essential to maintain or enhance landscape connectivity for this species. These conservation efforts should enable \u003cem\u003eTimon lepidus\u003c/em\u003e to persist and move throughout mosaic periurban environments, provided sufficient suitable habitat remains available. Given the moderate levels of genetic diversity inferred in our study area, despite the rather high density levels of the species compared to numerous other areas, such measures need to be seriously considered for implementation.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePerspectives\u003c/em\u003e \u003c/p\u003e \u003cp\u003eEven though the final genotyping of feces provided less data than expected, the use of dogs to find these samples was highly profitable, as the majority of feces were found in bushy areas that were undetectable by sight alone. Training dogs to detect the scent of individuals rather than feces could also be used to capture individuals or simply to monitor the presence of the species on site and find its shelters.\u003c/p\u003e \u003cp\u003eBecause our landscape genetic analyses relied exclusively on contemporary environmental covariates, the inferred resistance patterns reflect recent landscape configurations rather than historical conditions under which part of the observed genetic structure may have been established. Comparing current relationships with those derived from past landscape configurations would help disentangle historical and contemporary effects.\u003c/p\u003e \u003cp\u003eFinally, other methodologies complement genetic approaches when it comes to estimating dispersal and movement capacities, particularly with regard to landscape connectivity. Techniques such as GPS tracking and radiotelemetry provide direct, high-resolution data on individual movements, habitat use and potential dispersal barriers, which are difficult to infer from genetic data alone. However, these approaches are often used independently, which limits our ability to fully understand population connectivity. The importance of combining genetic and movement-based data is often emphasized in global reviews, as this enables researchers to link realised dispersal patterns with actual gene flow (Brum et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Tan et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This reveals not only where individuals move, but also which movements result in successful reproduction. In the context of our study, combining genetic analyses with radiotracking or GPS data would provide a more comprehensive understanding of the species' dispersal ecology and help identify critical habitats, corridors and potential barriers. This integrated approach would be invaluable for informing conservation management, guiding habitat connectivity measures and predicting population responses to environmental changes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding sources\u003c/h2\u003e \u003cp\u003eThis research has been supported by the Commissariat \u0026agrave; l\u0026rsquo;Energie Atomique et aux Energies Alternatives (grant no. MRISQ)\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eJohan Ludot: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing \u0026ndash; original draft, Writing \u0026ndash; review and editingAur\u0026eacute;lie Coulon: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing \u0026ndash; review and editingBenoit Charrasse: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing \u0026ndash; review and editingV\u0026eacute;ronique Arnal: Resources\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank the \"Collection BEV\" (Biog\u0026eacute;ographie et Ecologie des Vert\u0026eacute;br\u0026eacute;s collection) housed at the Centre d\u0026rsquo;Ecologie Fonctionnelle et Evolutive (CEFE) in Montpellier, France and the persons who contributed to the sampling of the individuals we used for microsatellite development, among which Alexandre Cluchier, Ren\u0026eacute; Celse, Marc Cheylan; and for microsatellite genotyping among which: Vincent Hallot, Jean-Baptiste Rico, Lena Delcamp, Serena Ghiles, Nicolas Fuento and Nathalie Espuno.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAdamack AT, Gruber B (2014) PopGenReport: simplifying basic population genetic analyses in R. Methods Ecol Evol 5:384\u0026ndash;387. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.12158\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.12158\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgapow P-M, Burt A (2001) Indices of multilocus linkage disequilibrium. Mol Ecol Notes 1:101\u0026ndash;102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1046/j.1471-8278.2000.00014.x\u003c/span\u003e\u003cspan address=\"10.1046/j.1471-8278.2000.00014.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaguette M, Blanchet S, Legrand D, Stevens V, Turlure C (2013) Individual dispersal, landscape connectivity and ecological networks. Biol Rev 1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalkenhol N, Cushman SA, Storfer AT, Waits LP (2015) Landscape genetics concepts, methods, applications. Wiley Blackwell\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaptista MSP, Keuroghlian A, Tambosi LR, C\u0026ocirc;rtes MC, de G\u0026oacute;es Maciel F, Cirino DW, Schmaedecke G, Biondo C (2025) Agriculturally developed areas reduce genetic connectivity for a keystone Neotropical ungulate. Biotropica 57:e13418. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/btp.13418\u003c/span\u003e\u003cspan address=\"10.1111/btp.13418\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBauder JM, Peterman WE, Spear SF, Jenkins CL, Whiteley AR, McGarigal K (2021) Multiscale assessment of functional connectivity: Landscape genetics of eastern indigo snakes in an anthropogenically fragmented landscape in central Florida. Mol Ecol 30:3422\u0026ndash;3438. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.15979\u003c/span\u003e\u003cspan address=\"10.1111/mec.15979\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeninde J, Feldmeier S, Werner M, Peroverde D, Schulte U, Hochkirch A, Veith M (2016) Cityscape genetics: structural vs. functional connectivity of an urban lizard population. Mol Ecol 25:4984\u0026ndash;5000. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.13810\u003c/span\u003e\u003cspan address=\"10.1111/mec.13810\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol 57:289\u0026ndash;300. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.2517-6161.1995.tb02031.x\u003c/span\u003e\u003cspan address=\"10.1111/j.2517-6161.1995.tb02031.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBridgewater P, Rotherham ID (2019) A critical perspective on the concept of biocultural diversity and its emerging role in nature and heritage conservation. People Nat 1:291\u0026ndash;304. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/pan3.10040\u003c/span\u003e\u003cspan address=\"10.1002/pan3.10040\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrum PHR, Gon\u0026ccedil;alves SRA, Str\u0026uuml;ssmann C, Teixido AL (2023) A global assessment of research on urban ecology of reptiles: patterns, gaps and future directions. Anim Conserv n/a. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/acv.12799\u003c/span\u003e\u003cspan address=\"10.1111/acv.12799\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrunt T, Smith AL (2025) Habitat Quality and Water Availability Affect Genetic Connectivity of Platypus Across an Urban Landscape. Anim Conserv 28:648\u0026ndash;662. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/acv.13011\u003c/span\u003e\u003cspan address=\"10.1111/acv.13011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarlen E, Munshi-South J (2021) Widespread genetic connectivity of feral pigeons across the Northeastern megacity. Evol Appl 14:150\u0026ndash;162. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/eva.12972\u003c/span\u003e\u003cspan address=\"10.1111/eva.12972\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCassimiro IMF, Ribeiro MC, Assis JC (2023) How did the animal come to cross the road? Drawing insights on animal movement from existing roadkill data and expert knowledge. Landsc Ecol 38:2035\u0026ndash;2051. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-023-01671-5\u003c/span\u003e\u003cspan address=\"10.1007/s10980-023-01671-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChakraborty R, Zhong Y, Jin L, Budowle B (1994) Nondetectability of restriction fragments and independence of DNA fragment sizes within and between loci in RFLP typing of DNA. Am J Hum Genet 55:391\u0026ndash;401\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheylan M, Dor\u0026eacute; F, Grillet P (2015) Le L\u0026eacute;zard ocell\u0026eacute; - Un g\u0026eacute;ant sur le continent europ\u0026eacute;en, Biotope. ed\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColucci C (2014) Evidence-based evaluation of artificial reptile habitats. A case study on stone structures\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCreech TG, Brennan A, Faselt J, Stabach JA, Keeley ATH (2024) Validating Connectivity Models: A Synthesis. Curr Landsc Ecol Rep 9:120\u0026ndash;134. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40823-024-00102-8\u003c/span\u003e\u003cspan address=\"10.1007/s40823-024-00102-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrispo E, Moore J-S, Lee-Yaw JA, Gray SM, Haller BC (2011) Broken barriers: Human-induced changes to gene flow and introgression in animals. BioEssays 33:508\u0026ndash;518. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/bies.201000154\u003c/span\u003e\u003cspan address=\"10.1002/bies.201000154\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrooks KR, Sanjayan M (2006) Connectivity Conservation. Cambridge University Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCumming K, Tavares D (2022) Using strategic environmental assessment and project environmental impact assessment to assess ecological connectivity at multiple scales in a national park context. Impact Assess Proj Apprais 40:507\u0026ndash;516. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/14615517.2022.2031553\u003c/span\u003e\u003cspan address=\"10.1080/14615517.2022.2031553\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaudin FM, 1802. Description of Lacerta lepida (=\u0026thinsp;Timon lepidus). - Histoire Naturel, g\u0026eacute;n\u0026eacute;rale et particuli\u0026egrave;redes reptiles, Dufart F Paris. ed\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelaney KS, Riley SPD, Fisher RN (2010) A Rapid, Strong, and Convergent Genetic Response to Urban Habitat Fragmentation in Four Divergent and Widespread Vertebrates. PLoS ONE 5:e12767. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0012767\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0012767\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunn DC, Harrison A-L, Curtice C, DeLand S, Donnelly B, Fujioka E, Heywood E, Kot CY, Poulin S, Whitten M, \u0026Aring;kesson S, Alberini A, Appeltans W, Arcos JM, Bailey H, Ballance LT, Block B, Blondin H, Boustany AM, Brenner J, Catry P, Cejudo D, Cleary J, Corkeron P, Costa DP, Coyne M, Crespo GO, Davies TE, Dias MP, Douvere F, Ferretti F, Formia A, Freestone D, Friedlaender AS, Frisch-Nwakanma H, Froj\u0026aacute;n CB, Gjerde KM, Glowka L, Godley BJ, Gonzalez-Solis J, Granadeiro JP, Gunn V, Hashimoto Y, Hawkes LM, Hays GC, Hazin C, Jimenez J, Johnson DE, Luschi P, Maxwell SM, McClellan C, Modest M, Di Sciara N, Palacio G, Palacios AH, Pauly DM, Rayner A, Rees M, Salazar AF, Secor ER, Sequeira D, Spalding AMM, Spina M, Van Parijs F, Wallace S, Varo-Cruz B, Virtue N, Weimerskirch M, Wilson H, Woodward L, Halpin B (2019) P.N., The importance of migratory connectivity for global ocean policy. Proc. R. Soc. B Biol. Sci. 286, 20191472. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rspb.2019.1472\u003c/span\u003e\u003cspan address=\"10.1098/rspb.2019.1472\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Etten J (2017) R Package gdistance: Distances and Routes on Geographical Grids. J Stat Softw 76:1\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.18637/jss.v076.i13\u003c/span\u003e\u003cspan address=\"10.18637/jss.v076.i13\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEvanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Mol Ecol 14:2611\u0026ndash;2620. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-294X.2005.02553.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-294X.2005.02553.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrillet P, Cheylan M, Thirion J-M, Bonnet X, Dauge C, Chollet S, Marchand M-A (2010) Rabbit burrows or artificial refuges are a critical habitat component for the threatened lizard, Timon lepidus (Sauria, Lacertidae). Biodivers Conserv 19:2039\u0026ndash;2205\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuerrero-Casado J, Carpio AJ, Can\u0026oacute;s-Burguete M, Torrijo-Salesa M, Tortosa FS (2022) The modernization of traditional vineyards into intensive trellis systems reduces the species richness and abundance of reptiles. Span J Agric Res 20:e0302\u0026ndash;e0302. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5424/sjar/2022202-18224\u003c/span\u003e\u003cspan address=\"10.5424/sjar/2022202-18224\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuillot G, Estoup A, Mortier F, Cosson JF (2005) A Spatial Statistical Model for Landscape Genetics. Genetics 170:1261\u0026ndash;1280. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1534/genetics.104.033803\u003c/span\u003e\u003cspan address=\"10.1534/genetics.104.033803\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHardy OJ, Vekemans X (2002) spagedi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Mol Ecol Notes 2:618\u0026ndash;620. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1046/j.1471-8286.2002.00305.x\u003c/span\u003e\u003cspan address=\"10.1046/j.1471-8286.2002.00305.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHijmans RJ (2024) terra: Spatial Data Analysis. R Package Version 17\u0026ndash;81 Httpsrspatialgithubioterra\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHilty J, Worboys GL, Keeley A, Woodley S, Lausche BJ, Locke H, Carr M, Pulsford I, Pittock J, White JW, Theobald DM, Levine J, Reuling M, Watson JEM, Ament R, Tabor GM (2020) Guidelines for conserving connectivity through ecological networks and corridors. IUCN, International Union for Conservation of Nature. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2305/IUCN.CH.2020.PAG.30.en\u003c/span\u003e\u003cspan address=\"10.2305/IUCN.CH.2020.PAG.30.en\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJakobsson M, Rosenberg NA (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23:1801\u0026ndash;1806. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/bioinformatics/btm233\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btm233\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJha S, Kremen C (2013) Urban land use limits regional bumble bee gene flow. Mol Ecol 22:2483\u0026ndash;2495. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.12275\u003c/span\u003e\u003cspan address=\"10.1111/mec.12275\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJombart T (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24:1403\u0026ndash;1405. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/bioinformatics/btn129\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btn129\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJombart T, Devillard S, Dufour A-B, Pontier D (2008) Revealing cryptic spatial patterns in genetic variability by a new multivariate method. Heredity 101:92\u0026ndash;103. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/hdy.2008.34\u003c/span\u003e\u003cspan address=\"10.1038/hdy.2008.34\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKamvar ZN, Tabima JF, Gr\u0026uuml;nwald NJ (2014) Poppr: an R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2:e281. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj.281\u003c/span\u003e\u003cspan address=\"10.7717/peerj.281\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeyghobadi N (2007) The genetic implications of habitat fragmentation for animals. Can J Zool 85:1049\u0026ndash;1064. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1139/Z07-095\u003c/span\u003e\u003cspan address=\"10.1139/Z07-095\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKimmig SE, Beninde J, Brandt M, Schleimer A, Kramer-Schadt S, Hofer H, B\u0026ouml;rner K, Schulze C, Wittstatt U, Heddergott M, Halczok T, Staubach C, Frantz AC (2020) Beyond the landscape: Resistance modelling infers physical and behavioural gene flow barriers to a mobile carnivore across a metropolitan area. Mol Ecol 29:466\u0026ndash;484. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.15345\u003c/span\u003e\u003cspan address=\"10.1111/mec.15345\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKool JT, Moilanen A, Treml EA (2013) Population connectivity: recent advances and new perspectives. Landsc Ecol 28:165\u0026ndash;185. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-012-9819-z\u003c/span\u003e\u003cspan address=\"10.1007/s10980-012-9819-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorpach AM, Garroway CJ, Mills AM, von Zuben V, Davy CM, Fraser KC (2022) Urbanization and artificial light at night reduce the functional connectivity of migratory aerial habitat. Ecography 2022, e05581. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ecog.05581\u003c/span\u003e\u003cspan address=\"10.1111/ecog.05581\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaPoint S, Balkenhol N, Hale J, Sadler J, van der Ree R (2015) Ecological connectivity research in urban areas. Funct Ecol 29:868\u0026ndash;878. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2435.12489\u003c/span\u003e\u003cspan address=\"10.1111/1365-2435.12489\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLendrum PE, Northrup JM, Anderson CR, Liston GE, Aldridge CL, Crooks KR, Wittemyer G (2018) Predation risk across a dynamic landscape: effects of anthropogenic land use, natural landscape features, and prey distribution. Landsc Ecol 33:157\u0026ndash;170. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-017-0590-z\u003c/span\u003e\u003cspan address=\"10.1007/s10980-017-0590-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLookingbill T, Minor E, Mullis C, Nunez-Mir G, Johnson P (2022) Connectivity in the Urban Landscape (2015\u0026ndash;2020): Who? Where? What? When? Why? and How? Curr. Landsc Ecol Rep 7:1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40823-021-00068-x\u003c/span\u003e\u003cspan address=\"10.1007/s40823-021-00068-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026oacute;pez-Est\u0026eacute;banez N, Molina-Holgado P, Allende \u0026Aacute;lvarez F (2024) Preserving the Values of Mediterranean Enclosed Fields with Dry Stone Walls, an Example of Vulnerable Natural and Rural Heritage. Heritage 7:844\u0026ndash;872. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/heritage7020041\u003c/span\u003e\u003cspan address=\"10.3390/heritage7020041\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLudot J, Coulon A, Charrasse B (2025) Fine-scale distribution modeling of ocellated lizard (\u003cem\u003eTimon lepidus\u003c/em\u003e), in a peri-urban landscape. Glob Ecol Conserv 62:e03724. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.gecco.2025.e03724\u003c/span\u003e\u003cspan address=\"10.1016/j.gecco.2025.e03724\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMasden EA, Haydon DT, Fox AD, Furness RW (2010) Barriers to movement: Modelling energetic costs of avoiding marine wind farms amongst breeding seabirds. Mar Pollut Bull 60:1085\u0026ndash;1091. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.marpolbul.2010.01.016\u003c/span\u003e\u003cspan address=\"10.1016/j.marpolbul.2010.01.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcClure ML, Hansen AJ, Inman RM (2016) Landsc Ecol 31:1419\u0026ndash;1432. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-016-0347-0\u003c/span\u003e\u003cspan address=\"10.1007/s10980-016-0347-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Connecting models to movements: testing connectivity model predictions against empirical migration and dispersal data\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCluskey EM, Kuzma FC, Enander HD, Cole-Wick A, Coury M, Cuthrell DL, Johnson C, Kelso M, Lee YM, Methner D, Rowe L, Swinehart A, Moore J (2024) Assessing habitat connectivity of rare species to inform urban conservation planning. Ecol Evol 14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ece3.11105\u003c/span\u003e\u003cspan address=\"10.1002/ece3.11105\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcKinney ML (2006) Urbanization as a major cause of biotic homogenization. Biol Conserv Urbanization 127:247\u0026ndash;260. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biocon.2005.09.005\u003c/span\u003e\u003cspan address=\"10.1016/j.biocon.2005.09.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcKinney ML, Lockwood JL (1999) Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends Ecol Evol 14:450\u0026ndash;453. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0169-5347(99)01679-1\u003c/span\u003e\u003cspan address=\"10.1016/S0169-5347(99)01679-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcRae BH, Dickson BG, Keitt TH, Shah VB (2008) Using Circuit Theory to Model Connectivity in Ecology, Evolution, and Conservation. Ecology 89:2712\u0026ndash;2724. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1890/07-1861.1\u003c/span\u003e\u003cspan address=\"10.1890/07-1861.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMichael DR, Moore H, Wassens S, Craig MD, Tingley R, Chapple DG, O\u0026rsquo;Sullivan J, Hobbs RJ, Nimmo DG (2021) Rock removal associated with agricultural intensification will exacerbate the loss of reptile diversity. J Appl Ecol 58:1557\u0026ndash;1565. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2664.13897\u003c/span\u003e\u003cspan address=\"10.1111/1365-2664.13897\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiles LS, Johnson JC, Dyer RJ, Verrelli BC (2018) Urbanization as a facilitator of gene flow in a human health pest. Mol Ecol 27:3219\u0026ndash;3230. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.14783\u003c/span\u003e\u003cspan address=\"10.1111/mec.14783\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiles LS, Rivkin LR, Johnson MTJ, Munshi-South J, Verrelli BC (2019) Gene flow and genetic drift in urban environments. Mol Ecol 28:4138\u0026ndash;4151. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.15221\u003c/span\u003e\u003cspan address=\"10.1111/mec.15221\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontano V, Jombart T (2017) An Eigenvalue test for spatial principal component analysis. BMC Bioinformatics 18:562. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12859-017-1988-y\u003c/span\u003e\u003cspan address=\"10.1186/s12859-017-1988-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNilsson C, Reidy CA, Dynesius M, Revenga C (2005) Fragmentation and Flow Regulation of the World\u0026rsquo;s Large River Systems. Science 308:405\u0026ndash;408. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.1107887\u003c/span\u003e\u003cspan address=\"10.1126/science.1107887\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOliveira EF, Martinez PA, S\u0026atilde;o-Pedro VA, Gehara M, Burbrink FT, Mesquita DO, Garda AA, Colli GR, Costa GC (2018) Climatic suitability, isolation by distance and river resistance explain genetic variation in a Brazilian whiptail lizard. Heredity 120:251\u0026ndash;265. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41437-017-0017-2\u003c/span\u003e\u003cspan address=\"10.1038/s41437-017-0017-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlivier A, Smith H, Santos R, Tatin L, Espuno N, Arnal V, Montgelard C, Miaud C (2017) Utilisation de chiens pour d\u0026eacute;tecter la pr\u0026eacute;sence du l\u0026eacute;zard ocell\u0026eacute; (Timon lepidus) 22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParadis E (2010) pegas: an R package for population genetics with an integrated\u0026ndash;modular approach. Bioinformatics 26:419\u0026ndash;420. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/bioinformatics/btp696\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btp696\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeakall R, Smouse PE (2006) genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes 6:288\u0026ndash;295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1471-8286.2005.01155.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1471-8286.2005.01155.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeterman WE (2018) ResistanceGA: An R package for the optimization of resistance surfaces using genetic algorithms. Methods Ecol Evol 9:1638\u0026ndash;1647. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.12984\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.12984\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePhillips PM, Langhans KE, Helmus MR, Jesse WAM, Surget-Groba Y, Behm JE (2024) Dispersal restriction and facilitation in species with differing tolerance to development: A landscape genetics study of native and introduced lizards. Divers Distrib 30:e13858. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ddi.13858\u003c/span\u003e\u003cspan address=\"10.1111/ddi.13858\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePounds JA, Jackson JF (1981) Riverine Barriers to Gene Flow and the Differentiation of Fence Lizard Populations. Evolution 35:516\u0026ndash;528. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/2408199\u003c/span\u003e\u003cspan address=\"10.2307/2408199\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945\u0026ndash;959. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1471-8286.2007.01758.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1471-8286.2007.01758.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRenet J, Dokhelar T, Thirion F, Tatin L, Pernollet CA, Bourgault L (2022) Spatial pattern and shelter distribution of the ocellated lizard (Timon lepidus) in two distinct Mediterranean habitats. Amphib -Reptil 43:263\u0026ndash;273. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1163/15685381-bja10095\u003c/span\u003e\u003cspan address=\"10.1163/15685381-bja10095\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRollo MF (2025) Interconnected Nature and People: Biosphere Reserves and the Power of Memory and Oral Histories as Biocultural Heritage for a Sustainable Future. Sustainability 17:4030. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su17094030\u003c/span\u003e\u003cspan address=\"10.3390/su17094030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRotherham ID (2015) Bio-cultural heritage and biodiversity: emerging paradigms in conservation and planning. Biodivers Conserv 24:3405\u0026ndash;3429. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10531-015-1006-5\u003c/span\u003e\u003cspan address=\"10.1007/s10531-015-1006-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudnick D, Ryan S, Beier P, Cushman S, Dieffenbach F, Epps C, Gerber L, Hartter J, Jenness J, Kintsch J, Merenlender A, Perkl R, Perziosi D, Trombulack S (2012) The Role of Landscape Connectivity in Planning and Implementing Conservation and Restoration Priorities. Issues in Ecology. Issues Ecol\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchlaepfer DR, Braschler B, Rusterholz H-P, Baur B (2018) Genetic effects of anthropogenic habitat fragmentation on remnant animal and plant populations: a meta-analysis. Ecosphere 9:e02488. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ecs2.2488\u003c/span\u003e\u003cspan address=\"10.1002/ecs2.2488\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchulte U, Veith M, Mingo V, Modica C, Hochkirch A (2013) Strong genetic differentiation due to multiple founder events during a recent range expansion of an introduced wall lizard population. Biol Invasions 15:2639\u0026ndash;2649. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10530-013-0480-5\u003c/span\u003e\u003cspan address=\"10.1007/s10530-013-0480-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScrucca L (2013) GA: A Package for Genetic Algorithms in R. J Stat Softw 53:1\u0026ndash;37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.18637/jss.v053.i04\u003c/span\u003e\u003cspan address=\"10.18637/jss.v053.i04\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeewagen CL, Nadeau-Gneckow J, Adams AM (2023) Far-reaching displacement effects of artificial light at night in a North American bat community. Glob Ecol Conserv 48:e02729. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.gecco.2023.e02729\u003c/span\u003e\u003cspan address=\"10.1016/j.gecco.2023.e02729\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShirk AJ, Landguth EL, Cushman SA (2017) A comparison of individual-based genetic distance metrics for landscape genetics. Mol Ecol Resour 17:1308\u0026ndash;1317. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.12684\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.12684\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSkarin A, Nellemann C, R\u0026ouml;nneg\u0026aring;rd L, Sandstr\u0026ouml;m P, Lundqvist H (2015) Wind farm construction impacts reindeer migration and movement corridors. Landsc Ecol 30:1527\u0026ndash;1540. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-015-0210-8\u003c/span\u003e\u003cspan address=\"10.1007/s10980-015-0210-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82:561\u0026ndash;573. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/sj.hdy.6885180\u003c/span\u003e\u003cspan address=\"10.1038/sj.hdy.6885180\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoanes K, Rytwinski T, Fahrig L, Huijser MP, Jaeger JAG, Teixeira FZ, van der Ree R, van der Grift EA (2024) Do wildlife crossing structures mitigate the barrier effect of roads on animal movement? A global assessment. J Appl Ecol 61:417\u0026ndash;430. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2664.14582\u003c/span\u003e\u003cspan address=\"10.1111/1365-2664.14582\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoanes K, Taylor AC, Sunnucks P, Vesk PA, Cesarini S, van der Ree R (2018) Evaluating the success of wildlife crossing structures using genetic approaches and an experimental design: Lessons from a gliding mammal. J Appl Ecol 55:129\u0026ndash;138. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2664.12966\u003c/span\u003e\u003cspan address=\"10.1111/1365-2664.12966\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSordello R, Paquier F, Daloz A, Patrinat OFB (2021) Trame noire - M\u0026eacute;thodes d\u0026rsquo;\u0026eacute;laboration et outils pour sa mise en oeuvre\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan WC, Herrel A, R\u0026ouml;dder D (2023) A global analysis of habitat fragmentation research in reptiles and amphibians: what have we done so far? Biodivers Conserv 32:439\u0026ndash;468. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10531-022-02530-6\u003c/span\u003e\u003cspan address=\"10.1007/s10531-022-02530-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor PD, Fahrig L, Henein K, Merriam G (1993) Connectivity Is a Vital Element of Landscape Structure. Oikos 68:571. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/3544927\u003c/span\u003e\u003cspan address=\"10.2307/3544927\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThioulouse J, Dray S, Dufour A-B, Siberchicot A, Jombart T, Pavoine S (2018) Multivariate Analysis of Ecological Data with ade4. Springer New York, New York, NY. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-1-4939-8850-1\u003c/span\u003e\u003cspan address=\"10.1007/978-1-4939-8850-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThirion J-M, Dor\u0026eacute; F, adamczyk A, Grillet P, Cheylan M (2009) Etude spatiale et temporelle d\u0026rsquo;une population de l\u0026eacute;zard ocell\u0026eacute; Timon lepidus en limite nord de distribution. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13140/RG.2.2.33032.93445\u003c/span\u003e\u003cspan address=\"10.13140/RG.2.2.33032.93445\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Strien MJ, Keller D, Holderegger R, Ghazoul J, Kienast F, Bolliger J (2014) Landscape genetics as a tool for conservation planning: predicting the effects of landscape change on gene flow. Ecol Appl 24:327\u0026ndash;339. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1890/13-0442.1\u003c/span\u003e\u003cspan address=\"10.1890/13-0442.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVirens E, Krauss SL, Davis RA, Spencer PBS (2015) Weak genetic structuring suggests historically high genetic connectivity among recently fragmented urban populations of the scincid lizard, Ctenotus fallens. Aust J Zool 63:279\u0026ndash;286. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/ZO15022\u003c/span\u003e\u003cspan address=\"10.1071/ZO15022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J (2011) coancestry: a program for simulating, estimating and analysing relatedness and inbreeding coefficients. Mol Ecol Resour 11:141\u0026ndash;145. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1755-0998.2010.02885.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1755-0998.2010.02885.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWenner SM, Murphy MA, Delaney KS, Pauly GB, Richmond JQ, Fisher RN, Robertson JM (2022) Natural and anthropogenic landscape factors shape functional connectivity of an ecological specialist in urban Southern California. Mol Ecol 31:5214\u0026ndash;5230. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.16656\u003c/span\u003e\u003cspan address=\"10.1111/mec.16656\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhittington J, Hebblewhite M, DeCesare NJ, Neufeld L, Bradley M, Wilmshurst J, Musiani M (2011) Caribou encounters with wolves increase near roads and trails: a time-to-event approach. J Appl Ecol 48:1535\u0026ndash;1542. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-2664.2011.02043.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2664.2011.02043.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeller KA, Jennings MK, Vickers TW, Ernest HB, Cushman SA, Boyce WM (2018) Divers Distrib 24:868\u0026ndash;879. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ddi.12742\u003c/span\u003e\u003cspan address=\"10.1111/ddi.12742\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Are all data types and connectivity models created equal? Validating common connectivity approaches with dispersal data\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeller KA, McGarigal K, Whiteley AR (2012) Estimating landscape resistance to movement: a review. Landsc Ecol 27:777\u0026ndash;797. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-012-9737-0\u003c/span\u003e\u003cspan address=\"10.1007/s10980-012-9737-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStatements \u0026amp; Declarations\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCRediT authorship contribution statement\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohan Ludot Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing \u0026ndash; original draft, Writing \u0026ndash; review and editing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAur\u0026eacute;lie Coulon Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing \u0026ndash; review and editing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenoit Charrasse Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing \u0026ndash; review and editing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eV\u0026eacute;ronique Arnal: Resources\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"conservation-genetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"coge","sideBox":"Learn more about [Conservation Genetics](https://www.springer.com/journal/10592)","snPcode":"10592","submissionUrl":"https://submission.nature.com/new-submission/10592/3","title":"Conservation Genetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Threatened species, Ocellated lizard, landscape genetics, gene flow, landscape resistance, resistance optimization","lastPublishedDoi":"10.21203/rs.3.rs-9550783/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9550783/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFunctional connectivity plays a key role in shaping gene flow and population persistence through human-modified environments, yet its drivers are understudied at spatial scales relevant to periurban planning.Using a landscape genetics framework, we investigated the multiscale effects of landscape composition and configuration on gene flow in the ocellated lizard (\u003cem\u003eTimon lepidus\u003c/em\u003e), a cavity-dependent reptile inhabiting a Mediterranean periurban landscape. We combined individual-based genetic data from 202 genotyped individuals with resistance surface optimization to assess how environmental features influence functional connectivity across two nested spatial scales: a continuous periurban landscape and a broader surrounding rural matrix. Genetic structure was weak and continuous at both scales, with no evidence of sharp genetic discontinuities, and spatial autocorrelation analyses revealed limited dispersal, coherent with a gradual isolation-by-distance (IBD) pattern. Resistance-based models however consistently outperformed IBD models, indicating that landscape features influence gene flow despite the overall weak genetic structuring. At the broader scale, rivers and agricultural cover emerged as the main contributors to resistance, while the availability of natural shelters and steep slopes facilitated connectivity. At the periurban scale, urban land cover and road infrastructures did not markedly restrict gene flow. Our findings indicate that ocellated lizards\u0026rsquo; weak genetic structure across spatial scales reflects historically high connectivity. Although gene flow is influenced by landscape resistance, its overall effect is modest and suggests that moderate urbanization, if associated with the presence of anthropogenic cavities, can maintain functional connectivity in this species.\u003c/p\u003e","manuscriptTitle":"Multiscale patterns of functional connectivity across a Mediterranean periurbanized landscape in Timon lepidus","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-15 09:40:53","doi":"10.21203/rs.3.rs-9550783/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-05-06T12:45:51+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-29T11:00:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-29T11:00:17+00:00","index":"","fulltext":""},{"type":"submitted","content":"Conservation Genetics","date":"2026-04-28T08:14:26+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"conservation-genetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"coge","sideBox":"Learn more about [Conservation Genetics](https://www.springer.com/journal/10592)","snPcode":"10592","submissionUrl":"https://submission.nature.com/new-submission/10592/3","title":"Conservation Genetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"9b3c3c15-b3bb-454c-bd5e-1d70c4848494","owner":[],"postedDate":"May 15th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewersInvited","content":"3","date":"2026-05-06T12:45:51+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-29T11:00:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-29T11:00:17+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-15T09:40:54+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-15 09:40:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9550783","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9550783","identity":"rs-9550783","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}