Joint species distribution modelling and bioregionalisation to shed light on the drivers of amphibian distribution and species assemblages at the local scale

Joint species distribution modelling and bioregionalisation to shed light on the drivers of amphibian distribution and species assemblages at the local scale | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 26 September 2025 V1 Latest version Share on Joint species distribution modelling and bioregionalisation to shed light on the drivers of amphibian distribution and species assemblages at the local scale Authors : Martino Flego 0009-0000-8368-9330 [email protected] , Alice Funk , Thomas Dadda , Valentina Rumo , Roberto Sacchi 0000-0002-6199-0074 , and Marco Mangiacotti 0000-0001-7144-3851 Authors Info & Affiliations https://doi.org/10.22541/au.175889185.54955555/v1 872 views 218 downloads Contents Abstract Abstract Key words Introduction Materials and methods Occurrence data and fieldwork Environmental predictors Statistical analysis Results Model evaluation Response to environmental variables and residual associations Regions of Common Profile Discussion Bibliography Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Joint Species Distribution Models are multispecies distribution models which explicitly acknowledge the multivariate nature of communities, modelling all species at once and considering their interactions. Besides giving insights on the specie niches, they can be used to individuate regions based on species occurrence probabilities (Regions of Common Profile, RCPs), supplying a key to understand and manage a study area. Despite their potential, these methodologies are rarely applied, especially regarding terrestrial vertebrates and small geographical scale. In this work we started filling the gap, relying on the framework Hierarchical Modelling of Species Communities (HMSC) for investigating amphibian populations at a lowland-mountains interface (Po Valley-Northern Apennines, Italy). The aim is to demonstrate the potential of these methods and broaden their application by answering ecological questions regarding the study system. We specifically investigated: environmental drivers of species distribution; cooccurrence pattern; bioregionalization of the study area. Species showed differential effects of environmental covariates, and mean annual temperature, broadleaf forests and vineyards came out to be the most relevant ones. Significant cooccurrence patterns were highlighted, probably not due to biotic interactions but to factors not explicitly included in the model, such as kind of breeding site. We identified three RCPs: a lowland region, dominated by farmland adapted species; a transitional one which lies in the lower Apennine and hosts widespread species, quite tolerant but probably not enough for the over-exploited Po plain; and a mountain region with microtherm and forest amphibians. Alarmingly, the first two RCPs are dominated by the alien marsh frog Pelophylax ridibundus, which reached the area in 2005 and is now widespread throughout. The work supported the applicability of the HMSC framework to amphibians and the definition of RCPs at a local scale. Finally, the results obtained can be a useful basis for conservation planning and long-term studies. Abstract Joint Species Distribution Models are multispecies distribution models which explicitly acknowledge the multivariate nature of communities, modelling all species at once and considering their interactions. Besides giving insights on the specie niches, they can be used to individuate regions based on species occurrence probabilities (Regions of Common Profile, RCPs), supplying a key to understand and manage a study area. Despite their potential, these methodologies are rarely applied, especially regarding terrestrial vertebrates and small geographical scale. In this work we started filling the gap, relying on the framework Hierarchical Modelling of Species Communities (HMSC) for investigating amphibian populations at a lowland-mountains interface (Po Valley-Northern Apennines, Italy). The aim is to demonstrate the potential of these methods and broaden their application by answering ecological questions regarding the study system. We specifically investigated: environmental drivers of species distribution; cooccurrence pattern; bioregionalization of the study area. Species showed differential effects of environmental covariates, and mean annual temperature, broadleaf forests and vineyards came out to be the most relevant ones. Significant cooccurrence patterns were highlighted, probably not due to biotic interactions but to factors not explicitly included in the model, such as kind of breeding site. We identified three RCPs: a lowland region, dominated by farmland adapted species; a transitional one which lies in the lower Apennine and hosts widespread species, quite tolerant but probably not enough for the over-exploited Po plain; and a mountain region with microtherm and forest amphibians. Alarmingly, the first two RCPs are dominated by the alien marsh frog Pelophylax ridibundus , which reached the area in 2005 and is now widespread throughout. The work supported the applicability of the HMSC framework to amphibians and the definition of RCPs at a local scale. Finally, the results obtained can be a useful basis for conservation planning and long-term studies. Key words ecological modelling, multispecies, HMSC, community ecology, species distribution modelling, vertebrates Introduction One of the main aims of ecology is to understand the factors affecting species distribution and abundance (Smith, 1968). This issue, historically approached in an anecdotal and descriptive manner, is now being answered using quantitative methods. Several approaches have been applied in order to identify these factors and quantify their importance. Among them, the ever-developing family of species distribution models (SDMs) has emerged some decades ago and has established itself as the most widely used (Elith & Leathwick, 2009; Guisan & Thuiller, 2005). Besides individuating drivers of distribution, SDMs allow predicting such distributions, calculating species occurrence probability or “suitability” in each unit of the study area (Elith & Leathwick, 2009). As a result, SDMs have immediately emerged as a major tool used for a wide range of tasks, from basic ecology to conservation and management, e.g.: studying niches characteristics; understanding ecological requirements of endangered species; figure out biogeographical pattern; evaluating invasiveness of introduced species; predicting effects of climate and human-mediated changes on populations; planning protected area and reintroductions (Araújo et al., 2019; Franklin, 2023; Hirzel & Le Lay, 2008; Lissovsky et al., 2021; Peterson, 2006; Srivastava et al., 2019). One of the intrinsic constrain of the early SDM approaches was that, whatever the proposed algorithm, they could model one species at a time, thus limiting the possible community level analysis to the assembling of several single-species models (Guisan & Rahbek, 2011). Such additive approach to multi-species SDMs inevitably results simplistic, as it ignores information about species interactions in modelling distribution, thus implicitly assuming that the process leading to the observed spatial pattern has not been directly affected by any biotic factor (the ”B” in the BAM diagram; (Sillero, 2011; Soberón, 2007). To overcome this issue, true multi-species approaches, has been subsequently proposed and defined as Joint Species Distribution Models, which give the opportunity to explicitly consider interactions between species, and to borrow information from one species and another, improving modelling of their niches (Ovaskainen et al., 2017; Ovaskainen & Soininen, 2011; Warton et al., 2015). This kind of multispecies models, embedded in the framework Hierarchical Modelling of Species Communities (HMSC), can explore species-specific responses to environmental variables and simultaneously account for interactions across taxa, and allows answering several ecological questions at once (Cranston et al., 2024; Ovaskainen et al., 2017; Soanes et al., 2024). Further, HSMC can greatly improve typically community-level analysis such as “bioregionalisation” (Foster et al., 2013; Ovaskainen & Abrego, 2020), which can be of great ecological and conservational interest (Foster et al., 2013; Hill et al., 2017). Bioregionalisation corresponds to the identification of discrete ecologically meaningful units within a given area, each characterized by unique and defined species assemblages (Lomolino et al., 2016), and inevitably requires a community-level approach. One of the most direct way to define such geographical entities relies on the concept of Regions of Common Profile (RCPs): a RCP is a portion of the environmental space where the vector of predicted species’ probabilities of being found are (relatively) constant and distinct from other regions (Foster et al., 2013). By providing vector of occurrence probability estimated by HSMC, it should be possible to obtain more realistic RCPs compared to those derived by separately estimating probability for each species. The use of interrelated species occurrence probabilities, in turn, may improve knowledge of the drivers of local beta-diversity, with a significant impact on conservation management. Although the HMSC framework is increasingly being applied (e.g. Bonfim et al., 2024; Custer et al., 2024; König et al., 2024; Seoane et al., 2023) and the same is occurring for the concept of data-driven bioregionalization (Hill et al., 2017; Woolley et al., 2020), as far as we are aware, the combination of these tools has never been attempted, at least concerning land environments and vertebrate taxa . Further, bioregionalization based on terrestrial vertebrates has been applied to macroecological studies at the global level (Falaschi et al., 2023; Ficetola et al., 2021; Holt et al., 2013), but never at local scale, where it could directly support management and conservation actions by providing a more complete picture than the widespread umbrella species concept (Roberge & Angelstam, 2004; Wang et al., 2021). To start filling the gap, the present work has the purpose of demonstrating the potential for the combination of HMSC and RCP approaches, giving support to the use of these emerging tools to enhance both ecological knowledge and conservation strategies; thus, expanding the data-driven bioregionalisation concept to the small geographic scale and showing its applicability to vertebrate communities with a limited number of species. The study system was set up on amphibian populations in the Oltrepò Pavese district, a small region in northern Italy in between the Po Valley and the Northern Apennines. This allows us to deepen the ecological knowledge of an interesting and heterogenous area, answering specific ecological questions. Amphibians, as well known, are the most threatened class of vertebrates at world-level (IUCN, 2024). Their decline, despite encouraging cases of both successful conservation actions (Moor et al., 2022) and innovative solutions (Waddle et al., 2024), is not stopping globally (Luedtke et al., 2023). In the last global evaluation, climate change have been recognized has the main driver of status deterioration, and northern Italy as an hotspot of threat, especially for potential spread of diseases (Luedtke et al., 2023). Italy is characterized by high levels of endemism across taxa (Amori & Castiglia, 2018; Massa & Fontana, 2020; Selvi et al., 2023) and amphibians make not an exception (Sillero et al., 2014). Besides conservation issues, amphibians also provide a good study case to test both HMSC and RCP approaches, since their need for water to succeed in reproduction constrains species to co-occur and share reproductive sites, naturally leading to species assemblages (López‐de Sancha et al., 2025). Nonetheless, HMSC has been applied to amphibians in just two, very different cases, both oriented towards assessing species interaction, in one case in congener salamanders (Gould & Peterman, 2024) and in the other with wetland birds and fishes (Kačergytė et al., 2023). The Oltrepò Pavese (Lombardy, Northern Italy) is a relatively small area (about 1097 km 2 ) set at the interface between the Po plain and the Northern Apennine, thus encompassing two ecoregions with very different characteristics: Central (Po) Plain Subsection and Toscana and Emilia-Romagna Apennine Subsection, belonging to the Po Plain Province and Apennine Province, respectively (Blasi et al., 2014, 2018). Amphibian communities and their drivers have been studied in the Po plain (Francesco Ficetola & De Bernardi, 2004; Ildos & Ancona, 1994) such as in the Apennines (Biancolini et al., 2024; A. Romano et al., 2012), although in their northern section studies focused on single species (Canessa et al., 2013; Mangiacotti et al., 2025) or environments (Manenti & Pennati, 2016). However, there is a lack of research investigating the distribution and ecological preferences of amphibians at the interface between these two different but closely connected regions, giving us the opportunity to apply the RCP framework on a relatively small scale. Combining HMSC and RCP approach, in this study we want to specifically address three different issues: i) identifying climatic and land cover factors driving amphibian assemblage distribution; ii) shading light on the role, if any, of species interactions in shaping cooccurrence pattern; iii) defining Regions of Common Profile, comparing them with recognized ecoregions (Blasi et al., 2018) and with special attention to location and species composition of a possible transition zone. A side point of interest is to assess the distribution and role in species assemblages of an harmful alien species, the marsh frog Pelophylax ridibundus (Pille et al., 2024). Materials and methods Study region We conducted our study in Oltrepò Pavese (Lombardy, Northern Italy, Fig.1). Due to its position which embraces plain and mountain areas, the Oltrepò is characterized by a great heterogeneity of environments and notable gradients both of habitat and climate. The altitudinal gradient ranges from about 50 m above sea level to 1754 (Mt. Lesima), along a north-south axis. The climate regime is temperate/mesothermal according to the Koppen classification (Bordoni et al., 2019), with differences within the area mainly related to altitude. The main land cover types are arable land and impervious surfaces in the lowland with an increasing proportion of broadleaved forest, paired with a fragmented distribution of permanent grasslands, rising in altitude; the lower hills are characterized by a high proportion of vineyards, constituting a cultural landscape typical of the study area (Brambilla & Ronchi, 2020). Target species We aimed to map as best as possible the distribution of all amphibian species found in Oltrepò Pavese, building the basis for subsequent modeling. Species consist of 6 Caudata : Triturus carnifex , Lissotriton vulgaris , Ichthyosaura alpestris , Salamandra salamandra , Salamandrina perspicillata , Speleomantes strinatii and 7 autochthonous Anura: Hyla intermedia , Bufo bufo , Bufotes viridis , Rana italica , Rana dalmatina , Rana temporaria , Pelophylax synkl. esculentus (Bernini et al., 2004, Fig.1). The alien marsh frog Pelophylax ridibundus is also present and abundant in the study area. This species has been introduced in northwestern Italy in 1941, spread in past decades and is expected to continue to do so (Falaschi et al., 2018; Lanza, 1962), with high risks for native amphibians (Pille et al., 2024). In the study area, the species first appeared in 2005 and rapidly spread across the plain portions and the bottom of the valley of the main stream (Razzetti et al. 2010). From now on we will refer to the non-native group as Pelophylax sp.; autochthonous taxa are marginal in the study area, and, even where present, are often mixed with the alien one. Among other species, several are endemic or subendemic to the Italian peninsula at the species (i.e., T. carnifex , S. perspicillata , S. strinatii, H. intermedia , R. italica ) or subspecies level (i.e., I. alpestris apuana , B. viridis balearicus ) (Sindaco & Razzetti, 2021). Figure 1. Location and map of the study area (Oltrepò Pavese). The target species included in the analysis are shown on the left and bottom edges. From the top left: Triturus carn ifex, Lissotriton vulgaris , Rana dalmatina , Pelophylax ridibundus , Ichthyosaura alpestris , Bufo bufo , Hyla intermedia , Bufotes viridis , Rana italica , Salamandra salamandra . Occurrence data and fieldwork An intensive campaign was conducted in 2024 to collect amphibian presence points, covering all species in the study area and their entire breeding season (February-July), for a total of 73 survey days. Field activities can be divided into 4 main tasks: (i) Monitoring of wetlands known by previous projects. Most of the wetland locations were obtained from the pond census conducted by Pellitteri-Rosa et al. (2010) which individuated 194 wetlands in Oltrepò Pavese. This number was supplemented by ad hoc surveys in the months preceding the field activity, reaching a total of 252 sites. Each site was surveyed 1-3 times between February and July. (ii) Searching for new breeding sites, by walking trails, especially in areas with fewer already known wetlands. We walked a total of 274 km, taking the GPS location of observed amphibians and breeding sites (including potential ones). (iii) Monitoring of species with notable calling activity: Hyla intermedia , Bufotes viridis , Pelophylax sp. for which call survey is a suggested census technique (Stoch & Genovesi, 2016). We drove a total of 435 km by night at low speed, stopping near wetlands or ditches, or whenever we heard calling, for listening session of at least 10 minutes. iv) Targeted surveys on Rana italica , searching for adults and tadpoles along streams. This endemic species is strictly related to running waters where individuals spend most of the year (Lanza et al., 2007), given that adults are elusive and primarily nocturnal, searching for tadpoles is an advantageous methodology (Stoch & Genovesi, 2016). The occurrence data were made unique within pixels of 500 × 500 m. We choose this grain as a compromise between high resolution species distribution modelling, avoiding duplicates in the case of larger wetlands, reducing spatial autocorrelation and allowing sufficient aggregation between points of occurrence of different species. Such a scale was used in other studies which applied species distribution modelling to amphibians (e.g. Gerick et al., 2014; Préau et al., 2018). We included occurrence points beyond administrative limits of Oltrepò Pavese for a maximum of 5 km. Environmental predictors We aimed to individuate drivers of amphibian distribution and community assemblage both among climatic and land cover variables, since both can represent key factors and using their combination has been shown to outperform climate-only models when applying species distribution modeling to amphibians (Seaborn et al., 2021). Climatic data were obtained from the free software ClimateEU v4.63 (Marchi et al., 2020) which allows downscaling climatic variables at the requested scale, providing coordinates and altitude as input. We provided as locations the centroids of a 500 × 500 m grid built on the study area and derived monthly average temperature and rainfall for the decade 2011-2020. We used the output to calculate 4 bioclimatic variables, bio01 (mean annual temperature), bio04 (temperature seasonality = standard deviation × 100), bio12 (cumulative annual precipitation), bio15 (precipitation seasonality = coefficient of variation), as indicated in the WorldClim database (Fick & Hijmans, 2017). In this way we obtained a measure of the two principal climatic factors with a dispersion index for both. However, all these bioclimatic variables turned out to be highly correlated with each other, so we kept exclusively bio01 as predictor. We derived land cover variables from the Land Cover Map 2022 produced by ISPRA, the Italian Institute for Environmental Protection and Research (https://groupware.sinanet.isprambiente.it/uso-copertura-e-consumo-di-suolo, accessed on 7 th November 2024). This map is obtained by integrating data from Copernicus Land Monitoring Services (https://land.copernicus.eu/en, accessed on 7 th November 2024) and the National Map of Land Consumption produced by ISPRA itself (Luti et al., 2021). The result is a raster layer with high spatial (10 × 10 m) and thematic (13 land cover classes) resolution. We discarded the land cover classes underrepresented in the study area, reducing them to 5: broadleaf forests, coniferous forests, periodic herbaceous (crops), permanent herbaceous, shrubs; we split the latter category in natural shrubs and vineyards, intersecting the map with the ISPRA Land Use Map (https://groupware.sinanet.isprambiente.it/uso-copertura-e-consumo-di-suolo, accessed on 7 th November 2024), resulting in a total of 6 categories. Percentage coverage within the 500 × 500-meter pixels was calculated for each class. Besides climate and land cover, landscape characteristics and heterogeneity can have an important role in predicting species occurrence and have recently claimed to be included in species distribution modelling (Riva et al., 2024). To address this need, we included a measure of thematic diversity in our predictors, by calculating Shannon’s index on land cover proportions (McGarigal & Marks, 1995; Shannon, 1948). Statistical analysis All the analyses were run in R environment (v. 4.4.0) (R Core Team, 2024). We managed and prepared spatial data using the package terra (Hijmans, 2020). We modelled species distribution and responses to environmental variables in the framework Hierarchical Modelling of Species Communities (HMSC) (Ovaskainen et al., 2017). In practice, Bayesan joint species distribution models were fitted using the R package Hmsc (Tikhonov et al., 2020). We modelled amphibians’ presence/absence using a probit distribution. The presence matrix was obtained from the 500 x 500 m pixels with at least one presence point, considered as the sample units (rows), while the species constituted the columns. Before running the models, we checked for multicollinearity among variables, and we had to remove a land cover class (periodic herbaceous); after this removal all remaining variables had a VIF below 5 (Menard, 2002). So, we ultimately included in the models as fixed effects the coverage of 5 land cover classes, the Shannon index and the mean annual temperature (bio01). Each covariate was scaled and included as a second-degree polynomial, to consider possible quadratic effects and species optima at intermediate levels. HMSC enables modelling of species associations by measuring the covariance structure of model residuals. Species associations are estimated in the random effect part of the HMSC model (Ovaskainen & Abrego, 2020). The random effect can be spatially explicit to control for the spatial structure of the residuals (Ovaskainen et al., 2016). We preliminary run a set of models, using the default prior distributions and sampling the posterior distribution with 4 short Markov Chain Monte Carlo chains (samples = 250, thinning = 10, burn-in = 1250) to assess the importance of random effects inclusion in the HMSC models and whether fixed effects provide more information than the mere spatial structure of the data: ENV (only fixed effects), ENVRL (fixed effects and sample unit-level random effect), ENVSPAT (fixed effects and spatial random effect), SPAT (only spatial random effect, without fixed effects). We compared these models using three metrics: root-mean-square-error (RMSE) between predicted and observed values, AUC (Area Under a relative operating characteristic Curve; Pearce & Ferrier, 2000), and Tjur R 2 (Tjur, 2009). We chose the best model and sampled the posterior distribution with 4 MCMC chains, each running for 375000 iterations, removing the first 125000 as burn-in and with a thinning interval of 1000. Thus, we kept 250 samples per chain, for a total of 1000. We analyzed the drivers of species distribution and community assemblage in four steps: (i) Partitioning of the explained variation to identify the most important factors among covariate fixed effects and random effect. (ii) Individuating responses to the environmental variables by exploring β parameter estimates and response curves; we described distribution of each β calculating Maximum A Posteriori probability estimate (MAP) and Highest Density Interval (HDI) with a credible interval of 0.89 (Kruschke, 2015), using the package bayestestR (Makowski et al., 2019) (iii) Computing residual species associations to individuate patterns of cooccurrence. (iv) Individuating Regions of Common Profile (Foster et al., 2013), clustering species based on their predicted occurrence probabilities (Ovaskainen & Abrego, 2020). For the latter purpose we chose the number of clusters using the Elbow method as implemented in the R packages factoextra and Nbclust (Charrad et al., 2014; Kassambara & Mundt, 2016). Results Occurrence data We collected 840 amphibian occurrences in total. We removed the duplicates within 500 × 500 m pixels. Discarding species with less than 15 presences at this resolution ( Salamandrina perspicillata , Speleomantes strinatii, Rana temporaria ), this resulted in a set of 275 pixels with a minimum of 1 amphibian species and a maximum of 6. The most common taxon is by far the alien Pelophylax , widespread in almost the entire study area, while the less commonly encountered is Salamandra salamandra (Tab. 1). Table 1. Number of occurrences per species in 500 m × 500 m pixels. Model evaluation The preliminary analysis of whether to include a random part in the model and whether to consider it spatially structured showed clear indications. The inclusion of a random effect increased the performance of the model by all metrics, especially by Tjur R 2 (Tab. 2). There were not remarkable differences between considering the random effect spatially structured or not. We therefore chose not to consider it structured, keeping the model simpler and reducing the computational effort. Table 2. Comparison between the four models: ENV (only fixed effects), ENVRL (fixed effects and sample unit-level random effect), ENVSPAT (fixed effects and spatial random effect), SPAT (only spatial random effect, without fixed effects). The final model, run with longer MCMC chains, showed good convergence. The effective sizes were mean 1033 ± 10 S.E. while the potential scale reduction factors for the β parameters were all close to 1 and under the threshold of 1.1 (Gelman & Rubin, 1992). Response to environmental variables and residual associations Partitioning of the variation showed that most of it was explained by land cover, while the climatic variable contributes 4 to 45 per cent, and the random effect of the sample unit was the less important driver for most species, excluding Triturus carnifex and Lissotriton vulgaris (Fig. 2). F igure 2. Proportion of explained variation among the climatic variable (mean annual temperature), land cover ones, and sample unit random effect. Looking at the single variables’ importance, contributions varied considerably among species, with bio01 and vineyard the only ones above 5% for all of them (Supporting information; Table 2). Regarding the effects and their signs, below we report the ones of variables with at least one of the two β (first- or second-degree term) that have 89% credibility interval, computed as the Highest Density Interval (HDI) of posterior distributions, that did not contain zero. Bio01 had a positive effect on the probability of occurrence of Hyla intermedia and Bufotes viridis , while Ichthyosaura alpestris was more likely to occur at low temperatures; Triturus carnifex , Pelophylax sp., Rana dalmatina , and Rana italica presented a negative second-degree term, indicating a temperature optimum, in the case of the last two species shifted towards lower values while centered on higher values for Pelophylax . Vineyards positively affected Triturus carnifex , Lissotriton vulgaris , Rana dalmatina and Bufo bufo ; in the case of Pelophylax sp. the positive sign of first-degree β was associated with the negative of second-degree one, with a drop of occurrence beyond high values; finally, Bufotes viridis and Hyla intermedia lose probability of occurrence beyond fairly low coverage values. Permanent herbaceous had negative second-degree β for Lissotriton vulgaris , Bufo bufo and Rana italica (in the latter case associated with a positive first-degree one); for Pelophylax sp. , βs were negative and positive for the first- and second-degree respectively. Broadleaf forests had a positive effect on Rana dalmatina , I c hthyosaura alpestris , Rana italica and Salamandra salamandra . Needle-leaved forests had a linear positive effect on Salamandra salamandra, Rana dalmatina and Pelophylax sp. , for the latter two associated with a negative quadratic one; Hyla intermedia is instead negatively affected. Shrubs only affected Rana italica , with a negative second-degree term. Finally, Bufotes viridis was negatively influenced by Shannon index, with a slightly positive second-degree term. All the response curves are available in the Supporting information (Supporting information: Fig.1); some of the most notable ones, concerning bio01, vineyards, and broadleaf forests, are provided below as examples to highlight the differential responses of species (Fig.2). Figure 3. Curves representing the predicted responses to variable variations. The values on the y axis are occurrence probabilities of the mentioned species, the credibility intervals are delimited by the first and third quartile. Here, some of the most notable curves are reported, to highlight species-specific responses to environmental gradients. All curves are available in Supporting Information. Analyses of residual associations among species revealed a positive cooccurrence pattern, with a support level above 0.95 credibility, among Rana dalmatina , Triturus carnifex , Lissotriton vulgaris and Ichthyosaura alpestris . These species were all negatively associated with Rana italica (Fig. 4). Figure 4. Residual species associations. Associations estimated to be positive with a probability of at least 0.95 are shown in yellow, those estimated to be negative in blue. In white, associations not statistically supported. Regions of Common Profile According to the Elbow method, the optimal number of clusters was three, thus leading to the definition of three RCPs where the probability of occurrence of the species varied sharply. Region 1 was dominated by Pelophylax sp. , with a significant presence of Hyla intermedia and Bufotes viridis ; in Region 2 the most abundant species did not change but the probability of presence of Rana dalmatina , Bufo bufo and the two newts Triturus carnifex and Lissotriton vulgaris increased. Finally, in region 3, Ichthyosaura alpestris , Rana italica and Salamandra salamandra became the most common species. Observing their projection in space, Region 1 corresponded to the plain part of the study area, Region 2 was distributed mainly in the hilly belt, while Region 3 was purely Apennine (Fig. 5). Evaluating the correspondence with the ecoregions of Italy (Blasi et al., 2018), we see that region 1 belongs to the Po Plain Province while 2 and 3 belong to the Apennine Province. Figure 5. Map of the three Regions of Common Profile in which the study area is divided. The thick black line represents the boundary between the Po Plain Province and the Apennine Province. The table on the right shows the five most common amphibian species in each region, with the probability of occurrence. The color palette is from the R package scico (Pedersen & Crameri, 2018). Discussion In this work we combined the use of joint species distribution modelling (HSMC) and bioregionalisation at local scale (RCP) to: i) analyze the factors affecting amphibian distribution at the interface between a lowland and a mountain environment in Northern Italy, ii) shedding lights on residual cooccurrence patterns; iii) identify species assemblages and their turn-over based on the predicted occurrence probabilities. The combination of the two approaches gave ecologically meaningful results, both considering the single-species and community perspective. HSMC simultaneously pinpointed the main distribution drivers of the studied species, disentangling the effects of climatic and land cover variables. The species that exhibit a higher importance of climate are the most stenothermic ones. Ichthyosaura alpestris is the only real cold-adapted species considered in this study (Chiocchio et al., 2017), and climate account for 45% of its explained variation. Although occurring at relatively low altitudes, R. italica is mainly related to hilly and mountainous areas (Sindaco et al., 2006), and the values of variation explained by climate is 31%. On the other hand, Hyla intermedia and Bufotes viridis , two lowland and thermophilic species (Derakhshan & Nokhbatolfoghahai, 2015; Dufresnes et al., 2018), shows values of 41% and 37%, respectively, highlighting the importance of temperature for both cold- and warm-adapted species. Although the above proportions should be interpreted with caution, also due to the different number of variables considered in the two blocks (1 climatic and 5 land cover), they are consistent with the species ecology (Lanza et al., 2007). The alpine newt shows strong responses to climate, with probability of occurrence falling to zero beyond species specific temperature thresholds. On the contrary, species with a broader altitudinal range showed not significant response or bell-shaped response curves, such as in the case of Rana dalmatina (Lanza et al., 2007). Land cover also showed particularly important effects (more than 50% variation explained), especially for species requiring specific habitats. Notably, the positive effect of broadleaf forests on I c hthyosaura alpestris , Rana italica, Rana dalmatina and Salamandra salamandra confirms the strong relation of the latter three species with woodlands; also I. alpestris , even if adaptable to many different types of breeding sites, is related to forest habitat, especially for its terrestrial phase (Denoël & Ficetola, 2008; Naumov et al., 2020). The positive effect of vineyards on Triturus carnifex , Lissotriton vulgaris , Rana dalmatina and Bufo bufo is probably due to the availability of artificial ponds related to viticultural activities. These ponds are usually stable and quite deep (more than one meter), representing good breeding sites for the above-mentioned species. On the other hand, the pond features and the scarcity of other types of wetlands, such as canals or temporary pools, in this kind of agriculture, explain the negative effects on Bufotes viridis and Hyla intermedia (Ebisuno & Gentilli, 2002; Landler et al., 2023) All the above results showed that a multispecies approach provides the same outcomes in terms of climatic and habitat selection as the classical single-species approach. However, the multispecies approach has an added value: the inclusion of a random effect, which allows estimating species residual associations, increasing model performance. On the other hand, caution should be exercised when interpreting residual associations: when statistically supported, they may indicate real interactions among species, as well as variations attributable to factors not controlled by the model (Ovaskainen & Abrego, 2020). In our case, the positive association among Lissotriton vulgaris , Triturus carnifex , Ichthyosaura alpestris and Rana dalmatina are not thought to reflect species direct interactions, but rather the shared microhabitat preferences not included in the fixed part of the model; indeed, these species are the most strictly pond-breeder ones (Lanza et al., 2007). The capture of breeding site type preferences by the random part of the model is confirmed by the negative association among the latter species and Rana italica , which spawns exclusively in lotic environments. The fact that the model didn’t capture real biotic interaction but rather unmodelled microhabitat type could be seen as a disadvantage, however its ability to account for features that are difficult to explicitly include in a species distribution model (e.g. mapping the number of ponds for each cell of the study area would be almost impossible) could be a further strength of the HMSC approach. Based on the presence probabilities derived from the multispecies model, we could identify three Regions of Common Profile, geographically corresponding to three distinct zones: an Apennine one, a lowland one, and a hilly and transitional one; the latter acting somewhat as an interface between the former two, both spatially and in terms of species composition. The lowland region is dominated by the alien Pelophylax sp. and secondarily by Bufotes viridis and Hyla intermedia , while the other species are entirely marginal. The Po Valley is strongly human-dominated and the species most sensitive to habitat alteration are rare and isolated, while the most adaptable ones can be still widespread ( Ficetola & De Bernardi, 2004). In this study this trend is confirmed, showing a strong homogenization of the amphibian community in the Po Valley. In the transitional region the most common species by far is still Pelophylax sp. but there is a turnover among the other common species (17 %: Rana dalmatina ; 10%: Bufo bufo and Lissotriton vulgaris ; 7 %: Triturus carnifex ). This assemblage includes widespread species with great distribution range both at the Italian and at the European level (considering the Triturus cristatus superspecies for T. carnifex ). These species are theoretically quite tolerant and adapted also to lowland environments, but the intensive agriculture and the subsequent loss of habitat and isolation can strongly threaten them and lead to local extinction (Ficetola & De Bernardi, 2004). These phenomena are exacerbated by the presence of alien species and fishes in general, which are a major driver of population dynamics of newts in the Po Plain (Falaschi et al., 2022). Finally, Ichthyosaura alpestris is the most common species in the Apennine region, followed by two amphibians inhabiting the Apennine streams: Salamandra salamandra and Rana italica. Besides these species, Bufo bufo and Rana dalmatina also occur with a fair probability. In conclusion, the RCP approach clearly defined three regions: the lowland region, which harbors amphibian species adapted to agricultural areas; the transitional region, which hosts species not tied to a specific altitude or predominantly to the lowlands and once widespread also in the Po Valley; and the purely Apennine region, inhabited by mountainous and forest taxa. According to the ecoregions from Blasi et al. (2018), the lowland region is included in the Po Plain Province while the transitional and Apennine regions lie in the Apennine Province. This is consistent geographically but allows us to highlight an interesting phenomenon at the level of community composition. While the plain region is homogenized and even quite tolerant and lowland adapted species such as Bufo bufo , Rana dalmatina and Lissotriton vulgaris are very rare there, such species are hosted in the Apennine, becoming the most common ones in its lower part, where the most typical and upland species still fail to prevail. Decline and vanishing of amphibians, including widespread ones, is a recurring pattern in intensively cultivated lowlands in Europe (Arntzen et al., 2017; Renoirt et al., 2024). Po Plain is one of the most urbanized and cultivated areas in Europe (Romano & Zullo, 2016), and some of its most iconic amphibian taxa are declining and threatened with extinction (Andreone et al., 2004; Falaschi et al., 2022; Rondinini et al., 2022; Ficetola et al., 2023). In the case of this study the contiguity of the plain with the Apennine, affected by different trends (Falcucci et al., 2007), makes it possible to maintain a region where species are still quite common in the study area. A point of attention must be paid to the spread of the marsh frog ( Pelophylax ridibundus ). Untill 2005, this invasive species was absent (Razzetti et al. 2010) and till 2016 it was restricted to the western part of the study area (Falaschi et al., 2018), while it is now dominant everywhere but the highest peaks. This spread is dangerous for the native amphibians due to the large environmental niche of the marsh frog, which overlaps with that of many of them (Pille et al., 2024). This species is able to colonize even the Apennine streams and the ongoing global warming could foster its propagation in areas where it is currently absent ( Falaschi et al., 2018; Padilla et al., 2023). In conclusion, we showed the potential of combining multispecies spatial distribution models and bioregionalisation at local scale. HMSC can account for factors not explicitly integrated into the model and provide the basis for Regions of Common Profile identification even at the local scale. Specifically, the most crucial climatic and land-use factors for amphibian species in an under-studied but interesting region were identified. Understanding the drivers of species distribution and basically their habitat preferences can also be the basis for conservation and management actions (Cañadas et al., 2005; Gray et al., 2007; Sanderson et al., 2002). The subdivision of an area based on species probability of presence can be an important information for conservation planning too, as it allows reasoning in terms of species assemblages rather than individual species, without using a single “umbrella” species to represent all of them, at the risk of losing crucial information (Hill et al., 2017; Wang et al., 2021). Finally, having defined the distribution and composition of species assemblages in a fast-changing world can be an initial step in studying their shifts, both with a modelling and long-term study approach (Franklin et al., 2016; Miller et al., 2018). Bibliography Amori, G., & Castiglia, R. (2018). Mammal endemism In Italy: A review. Biogeographia – The Journal of Integrative Biogeography , 33 (0), 19–31. https://doi.org/10.21426/B633035335Andreone, F., Eusebio Bergò, P., Bovero, S., & Gazzaniga, E. (2004). On the edge of extinction? The spadefoot Pelobates fuscus insubricus in the Po Plain, and a glimpse at its conservation biology. Italian Journal of Zoology , 71 (sup002), 61–72. https://doi.org/10.1080/11250003.2004.9525563Araújo, M. B., Anderson, R. P., Márcia Barbosa, A., Beale, C. M., Dormann, C. F., Early, R., Garcia, R. A., Guisan, A., Maiorano, L., Naimi, B., O’Hara, R. B., Zimmermann, N. E., & Rahbek, C. (2019). Standards for distribution models in biodiversity assessments. Science Advances , 5 (1), eaat4858. https://doi.org/10.1126/sciadv.aat4858Arntzen, J. W., Abrahams, C., Meilink, W. R. M., Iosif, R., & Zuiderwijk, A. (2017). Amphibian decline, pond loss and reduced population connectivity under agricultural intensification over a 38 year period. Biodiversity and Conservation , 26 (6), 1411–1430. https://doi.org/10.1007/s10531-017-1307-yBernini, F., Bonini, L., Ferri, V., Gentili, A., Razzetti, E., & Scali, S. (2004). Atlante degli Anfibi e dei Rettili della Lombardia . “Monografie di Pianura” n. 5.Biancolini, D., De Riso, L., & Romano, A. (2024). Opposite effects of climate and land‐use/cover change on Mediterranean herpetofauna: Insights from the southern Apennines. Animal Conservation , acv.12999. https://doi.org/10.1111/acv.12999Blasi, C., Capotorti, G., Copiz, R., Guida, D., Mollo, B., Smiraglia, D., & Zavattero, L. (2014). Classification and mapping of the ecoregions of Italy. Plant Biosystems - An International Journal Dealing with All Aspects of Plant Biology , 148 (6), 1255–1345. https://doi.org/10.1080/11263504.2014.985756Blasi, C., Capotorti, G., Copiz, R., Guida, D., Mollo, B., Smiraglia, D., & Zavattero, L. (2018). Terrestrial Ecoregions of Italy. Map and Explanatory notes. Terrestrial Ecoregions of Italy. Map and Explanatory Notes. Bonfim, F. C. G., Galetti, M., Benchimol, M., Morante‐Filho, J. C., Magioli, M., & Cazetta, E. (2024). Land‐use homogenization reduces the occurrence and diversity of frugivorous birds in a tropical biodiversity hotspot. Ecological Applications , 34 (4), e2980. https://doi.org/10.1002/eap.2980Bordoni, M., Vercesi, A., Maerker, M., Ganimede, C., Reguzzi, M. C., Capelli, E., Wei, X., Mazzoni, E., Simoni, S., Gagnarli, E., & Meisina, C. (2019). Effects of vineyard soil management on the characteristics of soils and roots in the lower Oltrepò Apennines (Lombardy, Italy). Science of The Total Environment , 693 , 133390. https://doi.org/10.1016/j.scitotenv.2019.07.196Brambilla, M., & Ronchi, S. (2020). Cool species in tedious landscapes: Ecosystem services and disservices affect nature-based recreation in cultural landscapes. Ecological Indicators , 116 , 106485. https://doi.org/10.1016/j.ecolind.2020.106485Cañadas, A., Sagarminaga, R., De Stephanis, R., Urquiola, E., & Hammond, P. S. (2005). Habitat preference modelling as a conservation tool: Proposals for marine protected areas for cetaceans in southern Spanish waters. Aquatic Conservation: Marine and Freshwater Ecosystems , 15 (5), 495–521. https://doi.org/10.1002/aqc.689Canessa, S., Oneto, F., Ottonello, D., Arillo, A., & Salvidio, S. (2013). Land abandonment may reduce disturbance and affect the breeding sites of an Endangered amphibian in northern Italy. Oryx , 47 (2), 280–287. https://doi.org/10.1017/S0030605311001542Charrad, M., Ghazzali, N., Boiteau, V., & Niknafs, A. (2014). NbClust : An R Package for Determining the Relevant Number of Clusters in a Data Set. Journal of Statistical Software , 61 (6). https://doi.org/10.18637/jss.v061.i06Chiocchio, A., Bisconti, R., Zampiglia, M., Nascetti, G., & Canestrelli, D. (2017). Quaternary history, population genetic structure and diversity of the cold-adapted Alpine newt Ichthyosaura alpestris in peninsular Italy. Scientific Reports , 7 (1), 2955. https://doi.org/10.1038/s41598-017-03116-xCranston, A., Cooper, N., & Bro‐Jørgensen, J. (2024). Using joint species distribution modelling to identify climatic and non‐climatic drivers of Afrotropical ungulate distributions. Ecography , e07209. https://doi.org/10.1111/ecog.07209Custer, C. A., North, J. S., Schliep, E. M., Verhoeven, M. R., Hansen, G. J. A., & Wagner, T. (2024). Predicting responses to climate change using a joint species, spatially dependent physiologically guided abundance model. Ecology , 105 (8), e4362. https://doi.org/10.1002/ecy.4362Denoël, M., & Ficetola, G. F. (2008). Conservation of newt guilds in an agricultural landscape of Belgium: The importance of aquatic and terrestrial habitats. Aquatic Conservation: Marine and Freshwater Ecosystems , 18 (5), 714–728. https://doi.org/10.1002/aqc.853Derakhshan Z, Nokhbatolfoghahai M. (2015). Thermal tolerance limits and effects of temperature on the growth and development of the green toad, Bufotes viridis . Salamandra , 51 , 129–136.Dufresnes, C., Mazepa, G., Rodrigues, N., Brelsford, A., Litvinchuk, S. N., Sermier, R., Lavanchy, G., Betto-Colliard, C., Blaser, O., Borzée, A., Cavoto, E., Fabre, G., Ghali, K., Grossen, C., Horn, A., Leuenberger, J., Phillips, B. C., Saunders, P. A., Savary, R., … Jeffries, D. L. (2018). Genomic Evidence for Cryptic Speciation in Tree Frogs From the Apennine Peninsula, With Description of Hyla perrini sp. Nov. Frontiers in Ecology and Evolution , 6 , 144. https://doi.org/10.3389/fevo.2018.00144Ebisuno, M., & Gentilli, A. (2002). Reproductive site selection and characteristics of sources and sinks in an Italian Tree Frog metapopulation (Hyla intermedia, Boulenger 1882). Revue d’Écologie (La Terre et La Vie) , 57 (3), 263–278. https://doi.org/10.3406/revec.2002.2397Elith, J., & Leathwick, J. R. (2009). Species Distribution Models: Ecological Explanation and Prediction Across Space and Time. Annual Review of Ecology, Evolution, and Systematics , 40 (1), 677–697. https://doi.org/10.1146/annurev.ecolsys.110308.120159Falaschi, M., Mangiacotti, M., Sacchi, R., Scali, S., & Razzetti, E. (2018). Electric circuit theory applied to alien invasions: A connectivity model predicting the Balkan frog expansion in Northern Italy. Acta Herpetologica , 33-42 Pages. https://doi.org/10.13128/ACTA_HERPETOL-20871Falaschi, M., Marta, S., Lo Parrino, E., Roll, U., Meiri, S., & Ficetola, G. F. (2023). Global bioregions of reptiles confirm the consistency of bioregionalization processes across vertebrate clades. Global Ecology and Biogeography , 32 (8), 1272–1284. https://doi.org/10.1111/geb.13694Falaschi, M., Muraro, M., Gibertini, C., Delle Monache, D., Lo Parrino, E., Faraci, F., Belluardo, F., Di Nicola, M. R., Manenti, R., & Ficetola, G. F. (2022). Explaining declines of newt abundance in northern Italy. Freshwater Biology , 67 (7), 1174–1187. https://doi.org/10.1111/fwb.13909Falcucci, A., Maiorano, L., & Boitani, L. (2007). Changes in land-use/land-cover patterns in Italy and their implications for biodiversity conservation. Landscape Ecology , 22 (4), 617–631. https://doi.org/10.1007/s10980-006-9056-4Ficetola, G. F., Manenti, R., Lo Parrino, E., Muraro, M., Barzaghi, B., Messina, V., Giachello, S., Melotto, A., & Falaschi, M. (2023). Decline and Extinction of the Italian Agile Frog Rana latastei from Core Areas of Its Range. Animals , 13 (20), 3187. https://doi.org/10.3390/ani13203187Ficetola, G. F., Mazel, F., Falaschi, M., Marta, S., & Thuiller, W. (2021). Determinants of zoogeographical boundaries differ between vertebrate groups. Global Ecology and Biogeography , 30 (9), 1796–1809. https://doi.org/10.1111/geb.13345Fick, S. E., & Hijmans, R. J. (2017). WorldClim 2: New 1‐km spatial resolution climate surfaces for global land areas. International Journal of Climatology , 37 (12), 4302–4315. https://doi.org/10.1002/joc.5086Foster, S. D., Givens, G. H., Dornan, G. J., Dunstan, P. K., & Darnell, R. (2013). Modelling biological regions from multi‐species and environmental data. Environmetrics , 24 (7), 489–499. https://doi.org/10.1002/env.2245Francesco Ficetola, G., & De Bernardi, F. (2004). Amphibians in a human-dominated landscape: The community structure is related to habitat features and isolation. Biological Conservation , 119 (2), 219–230. https://doi.org/10.1016/j.biocon.2003.11.004Franklin, J. (2023). Species distribution modelling supports the study of past, present and future biogeographies. Journal of Biogeography , 50 (9), 1533–1545. https://doi.org/10.1111/jbi.14617Franklin, J., Serra-Diaz, J. M., Syphard, A. D., & Regan, H. M. (2016). Global change and terrestrial plant community dynamics. Proceedings of the National Academy of Sciences , 113 (14), 3725–3734. https://doi.org/10.1073/pnas.1519911113Gelman, A., & Rubin, D. B. (1992). Inference from Iterative Simulation Using Multiple Sequences. Statistical Science , 7 (4). https://doi.org/10.1214/ss/1177011136Gerick, A. A., Munshaw, R. G., Palen, W. J., Combes, S. A., & O’Regan, S. M. (2014). Thermal physiology and species distribution models reveal climate vulnerability of temperate amphibians. Journal of Biogeography , 41 (4), 713–723. https://doi.org/10.1111/jbi.12261Gould, P. R., & Peterman, W. E. (2024). Coexistence of Stream-Associated Salamanders: Disentangling Abiotic and Biotic Interactions in Desmognathus Salamander Communities. Herpetologica , 80 (2). https://doi.org/10.1655/herpetologica-d-23-00053Gray, T. N. E., Chamnan, H., Borey, R., Collar, N. J., & Dolman, P. M. (2007). Habitat preferences of a globally threatened bustard provide support for community-based conservation in Cambodia. Biological Conservation , 138 (3–4), 341–350. https://doi.org/10.1016/j.biocon.2007.04.030Guisan, A., & Rahbek, C. (2011). SESAM - a new framework integrating macroecological and species distribution models for predicting spatio-temporal patterns of species assemblages: Predicting spatio-temporal patterns of species assemblages. Journal of Biogeography , 38 (8), 1433–1444. https://doi.org/10.1111/j.1365-2699.2011.02550.xGuisan, A., & Thuiller, W. (2005). Predicting species distribution: Offering more than simple habitat models. Ecology Letters , 8 (9), 993–1009. https://doi.org/10.1111/j.1461-0248.2005.00792.xHijmans, R. J. (2020). terra: Spatial Data Analysis (p. 1.7-83) [Dataset]. https://doi.org/10.32614/CRAN.package.terraHill, N. A., Foster, S. D., Duhamel, G., Welsford, D., Koubbi, P., & Johnson, C. R. (2017). Model‐based mapping of assemblages for ecology and conservation management: A case study of demersal fish on the Kerguelen Plateau. Diversity and Distributions , 23 (10), 1216–1230. https://doi.org/10.1111/ddi.12613Hirzel, A. H., & Le Lay, G. (2008). Habitat suitability modelling and niche theory. Journal of Applied Ecology , 45 (5), 1372–1381. https://doi.org/10.1111/j.1365-2664.2008.01524.xHolt, B. G., Lessard, J.-P., Borregaard, M. K., Fritz, S. A., Araújo, M. B., Dimitrov, D., Fabre, P.-H., Graham, C. H., Graves, G. R., Jønsson, K. A., Nogués-Bravo, D., Wang, Z., Whittaker, R. J., Fjeldså, J., & Rahbek, C. (2013). An Update of Wallace’s Zoogeographic Regions of the World. Science , 339 (6115), 74–78. https://doi.org/10.1126/science.1228282Ildos, A. S., & Ancona, N. (1994). Analysis of amphibian habitat preferences in a farmland area (Po plain, northern Italy). Amphibia-Reptilia , 15 (3), 307–316. https://doi.org/10.1163/156853894X00083Kačergytė, I., Knape, J., Żmihorski, M., Arlt, D., & Pärt, T. (2023). Community associations of birds with amphibians and fish in wetlands created for biodiversity. Biological Conservation , 282 , 110031. https://doi.org/10.1016/j.biocon.2023.110031Kassambara, A., & Mundt, F. (2016). factoextra: Extract and Visualize the Results of Multivariate Data Analyses (p. 1.0.7) [Dataset]. https://doi.org/10.32614/CRAN.package.factoextraKönig, S., Krauss, J., Classen, A., Hof, C., Prietzel, M., Wagner, C., & Steffan‐Dewenter, I. (2024). Micro‐ and macroclimate interactively shape diversity, niches and traits of Orthoptera communities along elevational gradients. Diversity and Distributions , 30 (5), e13810. https://doi.org/10.1111/ddi.13810Kruschke, J. K. (2015). Doing Bayesian data analysis: A tutorial with R, JAGS, and Stan (Academic Press). Elsevier. https://doi.org/10.1016/B978-0-12-405888-0.00001-5Landler, L., Burgstaller, S., & Schweiger, S. (2023). Land-use preferences of the European green toad (Bufotes viridis) in the city of Vienna (Austria): The importance of open land in urban environments. Frontiers in Zoology , 20 (1), 3. https://doi.org/10.1186/s12983-022-00480-xLanza, B. (1962). On the Introduction of Rana ridibunda Pallas and Rana catesbeiana Shaw in Italy. Copeia , 1962 (3), 642. https://doi.org/10.2307/1441194Lanza, B., Andreone, F., Bologna, M. A., Corti, C., & Razzetti, E. (2007). Amphibia . Fauna d’Italia. Edizioni Calderini.Lissovsky, A. A., Dudov, S. V., & Obolenskaya, E. V. (2021). Species-Distribution Modeling: Advantages and Limitations of Its Application. 1. General Approaches. Biology Bulletin Reviews , 11 (3), 254–264. https://doi.org/10.1134/S2079086421030075Lomolino, M. V., Riddle, B. R., & Whittaker, R. J. (2016). Biogeography: Biological Diversity across Space and Time (5 a ed.). Oxford University Press. https://doi.org/10.1093/hesc/9781605354729.001.0001López‐de Sancha, A., Benejam, L., Boix, D., Briggs, L., Cuenca-Cambronero, M., Davidson, T. A., De Meester, L., Fahy, J. C., Lemmens, P., Martin, B., Mehner, T., Oertli, B., Rasmussen, M., Greaves, H. M., Sayer, C., Beklioğlu, M., Brys, R., & Brucet, S. (2025). Drivers of amphibian species richness in European ponds. Ecography , 2025 (5), e07347. https://doi.org/10.1111/ecog.07347Luedtke, J. A., Chanson, J., Neam, K., Hobin, L., Maciel, A. O., Catenazzi, A., Borzée, A., Hamidy, A., Aowphol, A., Jean, A., Sosa-Bartuano, Á., Fong G., A., De Silva, A., Fouquet, A., Angulo, A., Kidov, A. A., Muñoz Saravia, A., Diesmos, A. C., Tominaga, A., … Stuart, S. N. (2023). Ongoing declines for the world’s amphibians in the face of emerging threats. Nature , 622 (7982), 308–314. https://doi.org/10.1038/s41586-023-06578-4Luti, T., De Fioravante, P., Marinosci, I., Strollo, A., Riitano, N., Falanga, V., Mariani, L., Congedo, L., & Munafò, M. (2021). Land Consumption Monitoring with SAR Data and Multispectral Indices. Remote Sensing , 13 (8), 1586. https://doi.org/10.3390/rs13081586Makowski, D., Ben-Shachar, M., & Lüdecke, D. (2019). bayestestR: Describing Effects and their Uncertainty, Existence and Significance within the Bayesian Framework. Journal of Open Source Software , 4 (40), 1541. https://doi.org/10.21105/joss.01541Manenti, R., & Pennati, R. (2016). Environmental factors associated with amphibian breeding in streams and springs: Effects of habitat and fish occurrence. Amphibia-Reptilia , 37 (2), 237–242. https://doi.org/10.1163/15685381-00003040Mangiacotti, M., Flego, M., Oneto, F., Ottonello, D., Cottalasso, R., Ferraro, G., & Sacchi, R. (2025). Climate and land use change through the eyes of two endemic amphibians: Temporal trajectories of suitability and connectivity reveal differential responses. Biological Conservation , 302 , 110971. https://doi.org/10.1016/j.biocon.2025.110971Marchi, M., Castellanos-Acuña, D., Hamann, A., Wang, T., Ray, D., & Menzel, A. (2020). ClimateEU, scale-free climate normals, historical time series, and future projections for Europe. Scientific Data , 7 (1), 428. https://doi.org/10.1038/s41597-020-00763-0Massa, B., & Fontana, P. (2020). Endemism in Italian Orthoptera. Biodiversity Journal , 11 (2), 405–434. https://doi.org/10.31396/Biodiv.Jour.2020.11.2.405.434McGarigal, K., & Marks, B. J. (1995). FRAGSTATS: Spatial pattern analysis program for quantifying landscape structure. (PNW-GTR-351; p. PNW-GTR-351). U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. https://doi.org/10.2737/PNW-GTR-351Menard, S. (2002). Applied Logistic Regression Analysis . SAGE Publications, Inc. https://doi.org/10.4135/9781412983433Miller, D. A. W., Grant, E. H. C., Muths, E., Amburgey, S. M., Adams, M. J., Joseph, M. B., Waddle, J. H., Johnson, P. T. J., Ryan, M. E., Schmidt, B. R., Calhoun, D. L., Davis, C. L., Fisher, R. N., Green, D. M., Hossack, B. R., Rittenhouse, T. A. G., Walls, S. C., Bailey, L. L., Cruickshank, S. S., … Sigafus, B. H. (2018). Quantifying climate sensitivity and climate-driven change in North American amphibian communities. Nature Communications , 9 (1), 3926. https://doi.org/10.1038/s41467-018-06157-6Moor, H., Bergamini, A., Vorburger, C., Holderegger, R., Bühler, C., Egger, S., & Schmidt, B. R. (2022). Bending the curve: Simple but massive conservation action leads to landscape-scale recovery of amphibians. Proceedings of the National Academy of Sciences , 119 (42), e2123070119. https://doi.org/10.1073/pnas.2123070119Naumov, B. Y., Popgeorgiev, G. S., Kornilev, Y. V., Plachiyski, D. G., Stojanov, A. J., & Tzankov, N. D. (2020). Distribution and Ecology of the Alpine Newt Ichthyosaura alp- estris (Laurenti, 1768) (Amphibia: Salamandridae) in Bulgaria. Acta Zoologica Bulgarica , 72 (1), 83–102.Ovaskainen, O., & Abrego, N. (2020). Joint Species Distribution Modelling: With Applications in R (1 a ed.). Cambridge University Press. https://doi.org/10.1017/9781108591720Ovaskainen, O., Roy, D. B., Fox, R., & Anderson, B. J. (2016). Uncovering hidden spatial structure in species communities with spatially explicit joint species distribution models. Methods in Ecology and Evolution , 7 (4), 428–436. https://doi.org/10.1111/2041-210X.12502Ovaskainen, O., & Soininen, J. (2011). Making more out of sparse data: Hierarchical modeling of species communities. Ecology , 92 (2), 289–295. https://doi.org/10.1890/10-1251.1Ovaskainen, O., Tikhonov, G., Norberg, A., Guillaume Blanchet, F., Duan, L., Dunson, D., Roslin, T., & Abrego, N. (2017). How to make more out of community data? A conceptual framework and its implementation as models and software. Ecology Letters , 20 (5), 561–576. https://doi.org/10.1111/ele.12757Padilla, P., Herrel, A., & Denoël, M. (2023). May future climate change promote the invasion of the marsh frog? An integrative thermo-physiological study. Oecologia , 202 (2), 227–238. https://doi.org/10.1007/s00442-023-05402-0Pearce, J., & Ferrier, S. (2000). Evaluating the predictive performance of habitat models developed using logistic regression. Ecological Modelling , 133 (3), 225–245. https://doi.org/10.1016/S0304-3800(00)00322-7Pedersen, T. L., & Crameri, F. (2018). scico: Colour Palettes Based on the Scientific Colour-Maps (p. 1.5.0) [Dataset]. https://doi.org/10.32614/CRAN.package.scicoPellitteri-Rosa, D., Cavagnini, F., Bellati, A., & Razzetti, E. (2010). Catasto di stagni e pozze per la tutela degli Anfibi in Oltrepò Pavese. Atti VIII Congresso Nazionale Societas Herpetologica Italica (Chieti, 22-26 settembre 2010) , 119-125.Peterson, A. T. (2006). Uses and Requirements of Ecological Niche Models and Related Distributional Models. Biodiversity Informatics , 3 (0). https://doi.org/10.17161/bi.v3i0.29Pille, F., Salomon, L., Comau, A.-C., Tendron, P., Duret, C., & Denoël, M. (2024). One frog to rule them all: Wide environmental niche of invasive marsh frogs induces large co-occurrence patterns with native amphibian prey in ponds. Hydrobiologia . https://doi.org/10.1007/s10750-024-05620-4Préau, C., Trochet, A., Bertrand, R., & Isselin-Nondedeu, F. (2018). Modeling potential distributions of three European amphibian species comparing ENFA and MaxEnt. Herpetological Conservation and Biology , 13 (1), 91-104.Razzetti E., Pellitteri-Rosa D., Sacchi R., Tiso E., Bernini F. (2010). Primi dati sulla presenza di Pelophylax kurtmuelleri (Gayda, 1940) in Lombardia. Atti VIII Congresso Nazionale Societas Herpetologica Italica (Chieti, 22-26 settembre 2010) , 341-347.R Core Team (2024). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.Renoirt, M., Angelier, F., Cheron, M., Jabaud, L., Tartu, S., & Brischoux, F. (2024). Population declines of a widespread amphibian in agricultural landscapes. The Science of Nature , 111 (2), 17. https://doi.org/10.1007/s00114-024-01905-9Riva, F., Martin, C. J., Galán Acedo, C., Bellon, E. N., Keil, P., Morán‐Ordóñez, A., Fahrig, L., & Guisan, A. (2024). Incorporating effects of habitat patches into species distribution models. Journal of Ecology , 1365-2745.14403. https://doi.org/10.1111/1365-2745.14403Roberge, J., & Angelstam, P. (2004). Usefulness of the Umbrella Species Concept as a Conservation Tool. Conservation Biology , 18 (1), 76–85. https://doi.org/10.1111/j.1523-1739.2004.00450.xRomano, A., Bartolomei, R., Conte, A., & Fulco, E. (2012). Amphibians in Southern Apennine: Distribution, ecology and conservation notes in the “Appennino Lucano, Val d’Agri e Lagonegrese” National Park (Southern Italy). Acta Herpetologica , 203-219 Pages. https://doi.org/10.13128/ACTA_HERPETOL-10503Romano, B., & Zullo, F. (2016). Half a century of urbanization in southern European lowlands: A study on the Po Valley (Northern Italy). Urban Research & Practice , 9 (2), 109–130. https://doi.org/10.1080/17535069.2015.1077885Rondinini, C., Battistoni, A., Teofili, C. (editors). 2022 Lista Rossa IUCN dei vertebrati italiani 2022 Comitato Italiano IUCN e Ministero dell’Ambiente e della Sicurezza Energetica, Roma Sanderson, E. W., Redford, K. H., Vedder, A., Coppolillo, P. B., & Ward, S. E. (2002). A conceptual model for conservation planning based on landscape species requirements. Landscape and Urban Planning , 58 (1), 41–56. https://doi.org/10.1016/s0169-2046(01)00231-6Seaborn, T., Goldberg, C. S., & Crespi, E. J. (2021). Drivers of distributions and niches of North American cold‐adapted amphibians: Evaluating both climate and land use. Ecological Applications , 31 (2), e2236. https://doi.org/10.1002/eap.2236Selvi, F., Campetella, G., Canullo, R., Chelli, S., Domina, G., Farris, E., Gasperini, C., Rosati, L., Wellstein, C., & Carrari, E. (2023). The Italian endemic forest plants: An annotated inventory and synthesis of knowledge. Plant Ecology and Evolution , 156 (1), 29–45. https://doi.org/10.5091/plecevo.95929Seoane, J., Estrada, A., Jones, M. M., & Ovaskainen, O. (2023). A case study on joint species distribution modelling with bird atlas data: Revealing limits to species’ niches. Ecological Informatics , 77 , 102202. https://doi.org/10.1016/j.ecoinf.2023.102202Shannon, C. E. (1948). A Mathematical Theory of Communication. Bell System Technical Journal , 27 (3), 379–423. https://doi.org/10.1002/j.1538-7305.1948.tb01338.xSillero, N. (2011). What does ecological modelling model? A proposed classification of ecological niche models based on their underlying methods. Ecological Modelling , 222 (8), 1343–1346. https://doi.org/10.1016/j.ecolmodel.2011.01.018Sindaco R, Doria G, Mazzetti E, Bernini F, (editors). (2006). Atlante degli anfibi e dei rettilli d’ltalia / Atlas of Italian amphibians and reptiles. Firenze: Edizioni Polistampa. Sindaco, R., & Razzetti, E. (2021). An updated check-list of Italian amphibians and reptiles. Natural History Sciences , 8 (2), 35–46. https://doi.org/10.4081/nhs.2021.519Soanes, K., Rytwinski, T., Fahrig, L., Huijser, M. P., Jaeger, J. A. G., Teixeira, F. Z., Van Der Ree, R., & Van Der Grift, E. A. (2024). Do wildlife crossing structures mitigate the barrier effect of roads on animal movement? A global assessment. Journal of Applied Ecology , 1365-2664.14582. https://doi.org/10.1111/1365-2664.14582Soberón, J. (2007). Grinnellian and Eltonian niches and geographic distributions of species. Ecology Letters , 10 (12), 1115–1123. https://doi.org/10.1111/j.1461-0248.2007.01107.xSrivastava, V., Lafond, V., & Griess, V. C. (2019). Species distribution models (SDM): Applications, benefits and challenges in invasive species management. CABI Reviews , 1–13. https://doi.org/10.1079/PAVSNNR201914020Stoch, F., & Genovesi, P. (2016, settembre). Handbooks for monitoring species and habitats of Community interest (Council Directive 92/43/EEC) in Italy: Animal species. Tikhonov, G., Opedal, Ø. H., Abrego, N., Lehikoinen, A., De Jonge, M. M. J., Oksanen, J., & Ovaskainen, O. (2020). Joint species distribution modelling with the r ‐package H msc. Methods in Ecology and Evolution , 11 (3), 442–447. https://doi.org/10.1111/2041-210X.13345Tjur, T. (2009). Coefficients of Determination in Logistic Regression Models—A New Proposal: The Coefficient of Discrimination. The American Statistician , 63 (4), 366–372. https://doi.org/10.1198/tast.2009.08210Waddle, A. W., Clulow, S., Aquilina, A., Sauer, E. L., Kaiser, S. W., Miller, C., Flegg, J. A., Campbell, P. T., Gallagher, H., Dimovski, I., Lambreghts, Y., Berger, L., Skerratt, L. F., & Shine, R. (2024). Hotspot shelters stimulate frog resistance to chytridiomycosis. Nature , 631 (8020), 344–349. https://doi.org/10.1038/s41586-024-07582-yWang, F., Winkler, J., Viña, A., McShea, W. J., Li, S., Connor, T., Zhao, Z., Wang, D., Yang, H., Tang, Y., Zhang, J., & Liu, J. (2021). The hidden risk of using umbrella species as conservation surrogates: A spatio-temporal approach. Biological Conservation , 253 , 108913. https://doi.org/10.1016/j.biocon.2020.108913Warton, D. I., Blanchet, F. G., O’Hara, R. B., Ovaskainen, O., Taskinen, S., Walker, S. C., & Hui, F. K. C. (2015). So Many Variables: Joint Modeling in Community Ecology. Trends in Ecology & Evolution , 30 (12), 766–779. https://doi.org/10.1016/j.tree.2015.09.007Woolley, S. N. C., Foster, S. D., Bax, N. J., Currie, J. C., Dunn, D. C., Hansen, C., Hill, N., O’Hara, T. D., Ovaskainen, O., Sayre, R., Vanhatalo, J. P., & Dunstan, P. K. (2020). Bioregions in Marine Environments: Combining Biological and Environmental Data for Management and Scientific Understanding. BioScience , 70 (1), 48–59. https://doi.org/10.1093/biosci/biz133 Information & Authors Information Version history V1 Version 1 26 September 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords biological invasion community ecology ecological modelling hmsc multispecies vertebrates Authors Affiliations Martino Flego 0009-0000-8368-9330 [email protected] University of Pavia Department of Earth and Environmental Sciences View all articles by this author Alice Funk University of Pavia Department of Earth and Environmental Sciences View all articles by this author Thomas Dadda University of Pavia Department of Earth and Environmental Sciences View all articles by this author Valentina Rumo University of Pavia Department of Earth and Environmental Sciences View all articles by this author Roberto Sacchi 0000-0002-6199-0074 University of Pavia Department of Earth and Environmental Sciences View all articles by this author Marco Mangiacotti 0000-0001-7144-3851 University of Pavia Department of Earth and Environmental Sciences View all articles by this author Metrics & Citations Metrics Article Usage 872 views 218 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Martino Flego, Alice Funk, Thomas Dadda, et al. Joint species distribution modelling and bioregionalisation to shed light on the drivers of amphibian distribution and species assemblages at the local scale. Authorea . 26 September 2025. DOI: https://doi.org/10.22541/au.175889185.54955555/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175889185.54955555/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a00cfdd1feda0708',t:'MTc3OTYzMjk1NQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();