Ensemble Of Small Models Research Articles

Habitat suitability modelling to predict the distribution of deep coral ecosystems: The case of Linosa Island (southern Mediterranean Sea, Italy)

In areas with limited field data, predictive habitat mapping is a valuable method for elucidating species-environment relationships and enhancing our knowledge of the spatial distribution and complexity of benthic habitats. Species distribution models (SDMs) can be an important tool to support in science-based ecosystem management. The availability of direct observations of mesophotic species, including gorgonians and black corals, during costly surveys is generally limited. Therefore, predicting the distribution of mesophotic species in relation to key physical parameters of the seafloor would help improving conservation strategies in existing and new Marine Protected Areas (MPAs). This study aims to assess the distribution of gorgonians and black corals off Linosa Island, in the Strait of Sicily, a biogeographic boundary area between the western and eastern Mediterranean. The volcanic island of Linosa represents a small, naturally preserved area, with very limited human pressure, hosting rich marine benthic biodiversity on its wide submarine portions. Distribution of the most common coral species off Linosa Island was modelled under an SDM framework, relying on direct observations collected during two research cruises in 2016 and 2017 and a series of terrain parameters acquired through geophysical techniques. We used the so-called “ensemble of small models” approach to calibrate SDMs, which achieved fair-to-excellent results (AUC >0.7). In addition to identifying depth as the primary factor influencing coral distribution, our study also highlighted ruggedness as a significant terrain variable. Specifically, the depth range of 110–230 m emerged as the critical parameter determining habitat suitability for all modelled species, also highlighting peculiar and specie–specific habitat requirements.

Guiding conservation strategies for China’s Corybas species through species distribution modeling

Corybas Salisb. is a widespread plant genus in the family Orchidaceae, known for its small, beautiful flowers. Most species in this genus are narrow endemics. China is home to six Corybas species, five of which are endangered, covering southern China from Taiwan to the western Himalayas. Niche overlap analysis was conducted on global Corybas species to assess the conservation value of Chinese Corybas species from evolutionary, adaptive, and niche diversity perspectives. The results revealed that Chinese Corybas species have formed a unique niche lineage, warranting conservation efforts. Due to challenges in field investigations, the potential distribution patterns of Chinese Corybas species remain unclear. The Species Distribution Model (SDM) using MaxEnt and Ensembles of Small Models (ESMs) was employed to analyze the potential distributions of five Corybas species in southern China and guide genus conservation. Both MaxEnt and ESMs analyses indicated that Chinese Corybas species may face significant environmental challenges in the future. This study identified gaps in fieldwork, highlighting the need to explore new populations of Corybas species in these areas. Conservation recommendations considering future climate change were proposed to aid programs aimed at preserving Chinese Corybas species.

Using an ensemble modeling to predict the potential distribution and habitat suitability of caracal (Caracal caracal) in southwestern Iran

Anthropogenic activities and climate change are degrading wildlife habitats, making it crucial to assess suitability of remaining habitat for effective conservation and management of threatened wildlife. This study aimed to determine the variables influencing habitat selection and spatial distribution of caracal (Caracal caracal) in southwestern Iran from 2019 to 2020. Ensembles of Small Models (ESM) were used based on eight species distribution models that included a combination of statistical and machine learning methods (GLM, CTA, FDA, GBM, ANN, MARS, RF, and MaxEnt). The outputs of two ensemble models were hybridized by simple averaging. In ESM 1, distance to water resources was used as a representative of environmental variables and in ESM 2, annual precipitation was used as a representative of bioclimatic variables. In both ESMs, these variables were combined with ecological variables such as the distribution of wild goat (Capra aegagrus) and livestock. The average importance of the model variables for ESM 1 and ESM 2 showed that the distribution of wild goat and livestock were the most important factors in predicting suitable habitat for caracals, respectively. Habitat suitability was higher close to the distribution of wild goat and livestock, near water sources and during periods of higher precipitation. Probabilistic predictions of caracal presence based on the hybridization of ESM 1 and ESM 2 suggest that only 1.3 % (206 ha) of the study area was highly suitable habitat, while 58.4 % (9570 ha) was unsuitable. Based on the ensemble of binary presence-absence maps, caracals were certainly absent in 76.2 % (12,490 ha) of the area, while their presence was uncertain in 21.6 % (3536 ha). The presence of caracals was certain in 2.2 % (363 ha) of the study area. Awareness of the habitat and distribution of caracal can help to make better decisions towards conservation and management and also reduce its conflict with humans.

Native American geography shaped historical fire frequency in forests of eighteenth-century Pennsylvania, USA

Researchers have debated the relative importance of environmental versus Indigenous effects on past fire regimes in eastern North America. Tree-ring fire-scar records (FSRs) provide local-resolution physical evidence of past fire, but few studies have spatially correlated fire frequency from FSRs with environmental and anthropogenic variables. No study has compared FSR locations to Native American settlement features in the eastern United States. We assess whether FSRs in the eastern US are located near regions of past Native American settlement. We also assess relationships between distance to Native American settlement, environmental conditions, and fire frequency in central Pennsylvania (PA), US, using an “ensemble of small models” approach for low sample sizes. Regression models of fire frequency at 21 locations in central PA often selected distance-based proxies of Indigenous land use. Models with mean annual temperature and Native American variables as predictors explained > 70% of the variation in fire frequency. Alongside temperature and wind speed, “distance to nearest trail” and “mean distance to nearest town” were significant and important predictors. In 18th-century central PA, fires were more frequent near Indigenous trails and towns, and further south due to increasing temperature and pyrophilic vegetation. However, for the entire eastern US, FSRs are located far from past settlement, limiting their effectiveness in detecting fire patterns near population centers. Improving understanding of historical fire will require developing FSRs closer to past Native American settlement.

Modeling the rarest of the rare: a comparison between multi‐species distribution models, ensembles of small models, and single‐species models at extremely low sample sizes

Species distribution models are useful for estimating the distribution and environmental preferences of rare species, but these same species are challenging to model on account of sparse data. We contrast a traditional single‐species approach (generalized linear models, GLMs) with two promising frameworks for modeling rare species: ensembles of small models (ESMs), which average across simple models; and multi‐species distribution models (MSDMs), which allow rarer species to benefit from statistical ‘borrowing of strength' from more common species. Using a virtual species within a community of real species, we evaluated how model accuracy was influenced by the number of occurrences of the rare species (N = 2–64), niche breadth, and similarity to more numerous species' niches. For discriminating between presence and absence, ESMs with just linear terms (ESM‐L) performed best for N ≤ 4, whereas for GLMs and ESMs with polynomial terms (ESM‐P) were best for N ≥ 8. For calibrating the species' response to influential variables, the MSDM hierarchical modeling of species communities (HMSC) and ESM‐P were best for species with niches similar to those of other species. For species with dissimilar niches, ESM‐P did best for N ≥ 8, but no model was well calibrated for smaller sample sizes. For identifying uninfluential variables, ESM‐L and species archetype models (SAMs), a type of MSDM, did well for ≤ 4, and ESM‐L for N ≥ 8. Models of species with narrow niches dissimilar to others had the highest discrimination capacity compared to models for generalist species and/or species with niches similar to other species' niches. ‘Borrowing of strength' in MSDMs can assist with some inference tasks, but does not necessarily improve predictions for rare species; simpler, single‐species models may be better at a given task. The best algorithm depends on modeling goal (discrimination versus calibration), sample size, and niche breadth and similarity.Keywords: borrowing of strength, calibration, data‐deficient species, discrimination, presence–absence, rare species

Modeling present and future distribution of plankton populations in a coastal upwelling zone: the copepod Calanus chilensis as a study case

Predicting species distribution in the ocean has become a crucial task to assess marine ecosystem responses to ongoing climate change. In the Humboldt Current System (HCS), the endemic copepod Calanus chilensis is one of the key species bioindicator of productivity and water masses. Here we modeled the geographic distribution of Calanus chilensis for two bathymetric ranges, 0–200 and 200–400 m. For the 0–200 m layer, we used the Bayesian Additive Regression Trees (BART) method, whereas, for the 200–400 m layer, we used the Ensembles of Small Models (ESMs) method and then projected the models into two future scenarios to assess changes in geographic distribution patterns. The models were evaluated using the multi-metric approach. We identified that chlorophyll-a (0.34), Mixed Layer Depth (0.302) and salinity (0.36) explained the distribution of C. chilensis. The geographic prediction of the BART model revealed a continuous distribution from Ecuador to the southernmost area of South America for the 0–200 m depth range, whereas the ESM model indicated a discontinuous distribution with greater suitability for the coast of Chile for the 200–400 m depth range. A reduction of the distribution range of C. chilensis is projected in the future. Our study suggests that the distribution of C. chilensis is conditioned by productivity and mesoscale processes, with both processes closely related to upwelling intensity. These models serve as a tool for proposing indicators of changes in the ocean. We further propose that the species C. chilensis is a high productivity and low salinity indicator at the HCS. We recommend further examining multiple spatial and temporal scales for stronger inference.

The potential distribution of the yellow monitor, Varanus flavescens (Hardwick & Gray) under multiple climate, land cover and dispersal scenarios in Nepal

Context Human-induced changes in climate and land cover have altered the distribution of fauna around the globe. Some reptiles have been found to be vulnerable to these changes; therefore, studies to identify the impact of the changes on other groups of reptiles are necessary. Aims We aimed to study the impact of climate and land cover change on the yellow monitor (Varanus flavescens) in Nepal. We also aimed to identify the current distribution range and predict the potential distribution under multiple climate change, corresponding land cover change, and dispersion scenarios in the near- and mid-future. Methods We used available presence locations with a candidate set of the least-correlated environmental variables and an Ensemble of Small Models (ESM), a Species Distribution Model (SDM) approach suitable for species with small sample size. Additionally, dispersal scenarios of 1 km, 5 km, and 10 km were added to the model to determine the future distribution under the dispersal scenarios. Key results We found soil particle size, distance to forest, precipitation of wettest quarter, bulk density, and elevation were the five most important variables contributing to the distribution of the species. The Terai lowland and wide valleys in Outer Himalayas are currently suitable but are expected to experience a substantial decrease under most future climate projections and dispersal scenarios. Conclusions The distribution is mostly dependent on soil-related variables; however, climatic variables might have a greater impact on future suitability. Implications Limiting emissions contributing to climatic changes, conserving the soil outside the protected areas, and the potential areas where the species will not experience habitat loss might contribute to the conservation of the species.

Riparian areas as a conservation priority under climate change

Identifying climatic refugia is important for long-term conservation planning under climate change. Riparian areas have the potential to provide climatic refugia for wildlife, but literature remains limited, especially for plants. This study was conducted with the purpose of identifying climatic refugia of plant biodiversity in the portion of the Mekong River Basin located in Xishuangbanna, China. We first predicted the current and future (2050s and 2070s) potential distribution of 50 threatened woody species in Xishuangbanna by using an ensemble of small models, then stacked the predictions for individual species to derive spatial biodiversity patterns within each 10 × 10 km grid cell. We then identified the top 17 % of the areas for spatial biodiversity patterns as biodiversity hotspots, with climatic refugia defined as areas that remained as biodiversity hotspots over time. Stepwise regression and linear correlation were applied to analyze the environmental correlations with spatial biodiversity patterns and the relationships between climatic refugia and river distribution, respectively. Our results showed potential upward and northward shifts in threatened woody species, with range contractions and expansions predicted. The spatial biodiversity patterns shift from southeast to northwest, and were influenced by temperature, precipitation, and elevation heterogeneity. Climatic refugia under climate change were related closely to river distribution in Xishuangbanna, with riparian areas identified that could provide climatic refugia. These refugial zones are recommended as priority conservation areas for mitigating the impacts of climate change on biodiversity. Our study confirmed that riparian areas could act as climatic refugia for plants and emphasizes the conservation prioritization of riparian areas within river basins for protecting biodiversity under climate change.

Harbour Porpoise Abundance in Portugal over a 5-Year Period and Estimates of Potential Distribution.

Simple SummaryThe Iberian porpoise inhabits the Atlantic coast of Portugal and Spain. The population is relatively small and there is a high number of individuals accidentally captured in fisheries. Using airplane surveys, we estimated that there were only 2254 porpoises overall between 2011 and 2015 in the coast of Portugal. The highest annual number of porpoises was recorded in 2013 (3207 individuals). However, in the following year, our study revealed that the population had been reduced approximately by half (when an increase in stranded individuals was also registered). The northern area of Portugal presented the most suitable habitat for the Iberian harbour porpoise, where coastal fisheries represent a particularly important socio-economic activity. Measures to decrease fishing effort are urgently needed as well as detailed information on seasonal Iberian harbour porpoises’ use of space.The Iberian porpoise population is small and under potentially unsustainable removal by fisheries bycatch. Recently, a marine Site of Community Importance (SCI) was legally approved in Portugal, but no measures ensued to promote porpoise conservation. Information about porpoise abundance and distribution is fundamental to guide any future conservation measures. Annual aerial surveys conducted between 2011 and 2015 show a low overall porpoise abundance and density (2254 individuals; 0.090 ind/km2, CV = 21.99%) in the Portuguese coast. The highest annual porpoise estimates were registered in 2013 (3207 individuals, 0.128 ind/km2), followed by a sharp decrease in 2014 (1653 individuals, 0.066 ind/km2). The porpoise density and abundance estimated in 2015 remained lower than the 2013 estimates. A potential distribution analysis of the Iberian porpoise population was performed using ensembles of small models (ESMs) with MaxEnt and showed that the overall habitat suitability is particularly high in the Portuguese northern area. The analysis also suggested a different pattern in porpoise potential distribution across the study period. These results emphasize the importance of further porpoise population assessments to fully understand the spatial and temporal porpoise habitat use in the Iberian Peninsula as well as the urgent need for on-site threat mitigation measures.

Finding what you don’t know: Testing SDM methods for poorly known species

AbstractAimA limitation of species distribution models (SDMs) is that species with low sample sizes are difficult to model. Yet, it is often important to know the habitat associations of poorly known species to guide conservation efforts. Techniques have been proposed for modelling species’ distributions from a few records, but their performance relative to one another has not been compared. Because these models are built and evaluated with small data sets, sampling error could cause severely biased sampling in environmental space. As a result, SDMs are likely to underpredict geographic distributions given small sample sizes. We perform the first comparison of methods explicitly promoted or developed for predicting the geographic ranges of species with very low sample sizes.LocationNorth Carolina, USA.TaxonSouth Mountains Grey‐cheeked Salamander (Plethodon meridianus).MethodsUsing the sparse, existing georeferenced records ofP. meridianus, we built SDMs using a range of methods that previous researchers have argued should work for low sample sizes. We then tested each SDM’s ability to accurately predict independent survey data that were not georeferenced prior to our study. We compared SDMs using omission error and AUC.ResultsRoughly half of the models successfully predicted survey records in the range centre, and all models had high omission error rates in the range exterior. In the range of interior or exterior, the ‘ensemble of small models’ technique produced SDMs with high omission error rates. Spatial filtering had a negligible impact on model performance. Most, but not all, models outperformed predictions using distance from known populations. Using one of the best‐performing methods, we developed an improved range map ofP. meridianus.Main ConclusionsGeographically peripheral populations were difficult to predict for all SDMs, though some methods were clearly inferior for our data set. We recommend that when sample sizes are low, researchers use Maxent with species‐specific model settings.

Flexsdm: An r package for supporting a comprehensive and flexible species distribution modelling workflow

Abstract Species distribution models (SDM) are widely used in diverse research areas because of their simple data requirements and application versatility. However, SDM outcomes are sensitive to data input and methodological choices. Such sensitivity and diverse applications mean that flexibility is necessary to create SDMs with tailored protocols for a given set of data and model use. We introduce the r package flexsdm for supporting flexible species distribution modelling workflows. flexsdm functions and their arguments serve as building blocks to construct a specific modelling protocol for user's needs. The main flexsdm features are modelling flexibility, integration with other modelling tools, simplicity of the objects returned and function speed. As an illustration, we used flexsdm to define a complete workflow for California red fir Abies magnifica. This package provides modelling flexibility by incorporating comprehensive tools structured in three steps: (a) The Pre‐modelling functions that prepare input, for example, sampling bias correction, sampling pseudo‐absences and background points, data partitioning, and reducing collinearity in predictors. (b) The Modelling functions allow fitting and evaluating different modelling approaches, including individual algorithms, tuned models, ensembles of small models and ensemble models. (c) The Post‐modelling functions include tools related to models' predictions, interpolation and overprediction correction. Because flexsdm comprises a large part of the SDM process, from outlier detection to overprediction correction, flexsdm users can delineate partial or complete workflows based on the combination functions to meet specific modelling needs.

Spatial risk modeling of cattle depredation by black vultures in the midwestern United States

AbstractNegative economic impacts resulting from wildlife disrupting livestock operations through depredation of stock are a cause of human‐wildlife conflict. Management of such conflict requires identifying environmental and non‐environmental factors specific to a wildlife species' biology and ecology that influence the potential for livestock depredation to occur. Identification of such factors can improve understanding of the conditions placing livestock at risk. Black vultures (Coragyps atratus) have expanded their historical range northward into the midwestern United States. Concomitantly, an increase in concern among agricultural producers regarding potential black vulture attacks on livestock has occurred. We estimated area with greater or lesser potential for depredation of domestic cattle by black vultures across a 6‐state region in the midwestern United States using an ensemble of small models (ESM). Specifically, we identified landscape‐scale spatial factors, at a zip code resolution, associated with reported black vulture depredation on cattle in midwestern landscapes to predict future potential livestock depredation. We hypothesized that livestock depredation would be greatest in areas with intensive beef cattle production close to preferred black vulture habitat (e.g., areas with fewer old fields and early successional vegetation paired with more direct edge between older forest and agricultural lands). We predicted that the density of cattle within the county, habitat structure, and proximity to anthropogenic landscape features would be the strongest predictors of black vulture livestock‐depredation risk. Our ESM estimated the relative risk of black vulture‐cattle depredation to be between 0.154–0.631 across our entire study area. Consistent with our hypothesis, areas of greatest predicted risk of depredation correspond with locations that are favorable to vulture life‐history requirements and increased potential to encounter livestock. Our results allow wildlife managers the ability to predict where black vulture depredation of cattle is more likely to occur in the future. It is in these areas where extension and outreach efforts aimed at mitigating this conflict should be focused. Researchers and wildlife managers interested in developing or employing tools aimed at mitigating livestock‐vulture conflicts can also leverage our results to select areas where depredation is most likely to occur.

Ensemble of small models as a tool for alien invasive species management planning: evaluation of Vespa velutina (Hymenoptera: Vespidae) under Mediterranean island conditions

Ecological niche models have proved to be a powerful tool in assessing invasiveness risk of alien species, allowing the optimization of control strategies. Vespa velutina (Hymenoptera: Vespidae) is an invasive species with strong ecological, economical and health impacts in Europe after it was first reported in France in 2004. It was detected for the first time on a Mediterranean island (Mallorca, Balearic Islands, Spain) in 2015, where a single nest was found in the northwest of the island. Immediately, a control plan was implemented. In this study, we analysed 30 occurrence data in Mallorca island to assess the suitability distribution predicted for Mediterranean island conditions using an ensemble of small models. We obtained high values of AUC (0.9165), Somers’ D (0.8331), Boyce (0.7611) and TSS (0.7754) as quality parameters of the final ensembled model. We show for the first time that there are suitable areas where this species can expand and stablish, mainly in steeper slopes and low isothermality zones. Likewise, the distribution suitability of V. velutina for other Mediterranean islands (Ibiza, Formentera, Menorca, Corsica, Sardinia, Sicily, Crete and Cyprus) was also explored, showing potentially suitable zones. This study provides valuable information regarding the areas in the Mediterranean islands under risk of invasion, and it could be used by both scientists and managers for an early detection and control of the invasive species due to its cost-effectiveness in terms of conservation.

Small but visible: Predicting rare bryophyte distribution and richness patterns using remote sensing-based ensembles of small models.

In Canadian boreal forests, bryophytes represent an essential component of biodiversity and play a significant role in ecosystem functioning. Despite their ecological importance and sensitivity to disturbances, bryophytes are overlooked in conservation strategies due to knowledge gaps on their distribution, which is known as the Wallacean shortfall. Rare species deserve priority attention in conservation as they are at a high risk of extinction. This study aims to elaborate predictive models of rare bryophyte species in Canadian boreal forests using remote sensing-derived predictors in an Ensemble of Small Models (ESMs) framework. We hypothesize that high ESMs-based prediction accuracy can be achieved for rare bryophyte species despite their low number of occurrences. We also assess if there is a spatial correspondence between rare and overall bryophyte richness patterns. The study area is located in western Quebec and covers 72,292 km2. We selected 52 bryophyte species with <30 occurrences from a presence-only database (214 species, 389 plots in total). ESMs were built from Random Forest and Maxent techniques using remote sensing-derived predictors related to topography and vegetation. Lee’s L statistic was used to assess and map the spatial relationship between rare and overall bryophyte richness patterns. ESMs yielded poor to excellent prediction accuracy (AUC > 0.5) for 73% of the modeled species, with AUC values > 0.8 for 19 species, which confirmed our hypothesis. In fact, ESMs provided better predictions for the rarest bryophytes. Likewise, our study revealed a spatial concordance between rare and overall bryophyte richness patterns in different regions of the study area, which have important implications for conservation planning. This study demonstrates the potential of remote sensing for assessing and making predictions on inconspicuous and rare species across the landscape and lays the basis for the eventual inclusion of bryophytes into sustainable development planning.

Current climate overrides past climate change in explaining multi-site beta diversity of Lauraceae species in China

We aimed to characterise the geographical distribution of Sørensen-based multi-site dissimilarity ( β sor ) and its underlying true turnover ( β sim ) and nestedness ( β sne ) components for Chinese Lauraceae and to analyse their relationships to current climate and past climate change. We used ensembles of small models (ESMs) to map the current distributions of 353 Lauraceae species in China and calculated β sor and its β sim and β sne components. We tested the relationship between β sor , β sne and β sim with current climate and past climate change related predictors using a series of simultaneous autoregressive (SAR err ) models. Spatial distribution of β sor of Lauraceae is positively correlated with latitude, showing an inverse relationship to the latitudinal α -diversity (species richness) gradient. High β sor occurs at the boundaries of the warm temperate and subtropical zones and at the Qinghai-Tibet Plateau due to high β sne . The optimized SAR err model explains β sor and β sne well, but not β sim . Current mean annual temperature determines β sor and β sne of Lauraceae more than anomalies and velocities of temperature or precipitation since the Last Glacial Maximum. Current low temperatures and high climatic heterogeneity are the main factors explaining the high multi-site β -diversity of Lauraceae. In contrast to analyses of the β -diversity of entire species assemblages, studies of single plant families can provide complementary insights into the drivers of β -diversity of evolutionarily more narrowly defined entities.

Impact of climate change on wheat security through an alternate host of stripe rust

AbstractIn the 21st century, stripe rust, caused by Puccinia striiformis f. sp. tritici (Pst), is still the most devastating disease of wheat globally. Despite the critical roles of the alternate host plants, the Berberis species, in the sexual reproduction and spread of Pst, the climate change impacts on the redistribution of Berberis plants, and their potential risk of incurring wheat stripe rust are scarcely discussed in the literature. This study evaluated the potential risk of Pst in wheat cultivation in Sichuan Province, China, for the first time, by analyzing the dynamics of Berberis plants and the change in Pst oversummering range under climate change scenarios. Ensembles of small models (ESMs) in combination with four algorithms were used to project the future (i.e., 2041–2060, 2061–2080, and 2081–2100) distributions of 54 Berberis species in Sichuan Province based on two shared socioeconomic pathways (SSP) scenarios. The species richness of Berberis was predicted to increase in northwestern Sichuan but decreased in southern and eastern Sichuan under both SSP245 and SSP585 scenarios. The mean temperature of the driest quarter was found to be the determinant of the distributional shifts for the majority of Berberis plants. Our overlapping assessments indicated that the risk of wheat stripe rust would be reduced in southwestern Sichuan. In contrast, it would be aggravated in the northwest due to the predicted shifts of Berberis plants northwards and stronger shrinkage of Pst oversummering range occurring in the south than in the north. In the “hotspots” of the stripe rust pathogen, risk assessment in advance could provide a scientific basis for decision‐makers and local communities in effectively controlling wheat stripe rust, thus further sustaining food security through adapting to future climate change.

Genetic structure and ecological niche segregation of Indian gray mongoose (Urva edwardsii) in Iran.

Combining genetic data with ecological niche models is an effective approach for exploring climatic and nonclimatic environmental variables affecting spatial patterns of intraspecific genetic variation. Here, we adopted this combined approach to evaluate genetic structure and ecological niche of the Indian gray mongoose (Urva edwardsii) in Iran, as the most western part of the species range. Using mtDNA, we confirmed the presence of two highly differentiated clades. Then, we incorporated ensemble of small models (ESMs) using climatic and nonclimatic variables with genetic data to assess whether genetic differentiation among clades was coupled with their ecological niche. Climate niche divergence was also examined based on a principal component analysis on climatic factors only. The relative habitat suitability values predicted by the ESMs for both clades revealed their niche separation. Between‐clade climate only niche comparison revealed that climate space occupied by clades is similar to some extent, but the niches that they utilize differ between the distribution ranges of clades. We found that in the absence of evidence for recent genetic exchanges, distribution models suggest the species occurs in different niches and that there are apparent areas of disconnection across the species range. The estimated divergence time between the two Iranian clades (4.9 Mya) coincides with the uplifting of the Zagros Mountains during the Early Pliocene. The Zagros mountain‐building event seems to have prevented the distribution of U. edwardsii populations between the western and eastern parts of the mountains as a result of vicariance events. Our findings indicated that the two U. edwardsii genetic clades in Iran can be considered as two conservation units and can be utilized to develop habitat‐specific and climate change‐integrated management strategies.

Food and habitats requirements of the Eurasian Scops Owl (Otus scops) in Switzerland revealed by very high‐resolution multi‐scale models

In Europe, agricultural practices have progressively evolved towards high productivity leading either to the intensification of productive and accessible areas or to the abandonment of less profitable sites. Both processes have led to the degradation of semi‐natural habitats like extensive grasslands, threatening species such as the Eurasian Scops Owl Otus scops that rely on extensively managed agricultural landscapes. In this work, we aimed to assess the habitat preferences of the Scops Owl using habitat suitability models combined with a multi‐scale approach. We generated a set of multi‐scale predictors, considering both biotic and abiotic variables, built on two newly developed vegetation management and orthopteran abundance models. To select the variables to incorporate in a ‘best multi‐scale model’, we chose the best spatial scale for each variable using univariate models and by calculating their relative importance through multi‐model inference. Next, we built ensembles of small models (ESMs) at 10 different scales from 50 to 1000 m, and an additional model with each variable at its best scale (‘best multi‐scale model’). The latter performed better than most of the other ESMs and allowed the creation of a high‐resolution habitat suitability map for the species. Scops Owls showed a preference for dry sites with extensive and well‐structured habitats with 30–40% bush cover, and relied strongly on semi‐extensive grasslands covering at least 30% of the surface within 300 m of the territory centre and with high orthopteran availability near the centre (50‐m radius), revealing a need for good foraging grounds near the nest. At a larger spatial scale within a radius of 1000 m, the habitat suitability of Scops Owls was negatively related to forest cover. The resulting ESM predictions provide valuable tools for conservation planning, highlighting sites in need of particular conservation efforts together with offering estimates of the percentage of habitat types and necessary prey abundance that could be used as targets in future management plans to ensure the persistence of the population.

Modeling the 20th-century distribution changes of Microgyne trifurcata, a rare plant of the southern South American grasslands

Microgyne trifurcata is a rare native plant species from one of the areas with the highest human impact on the environment in southern South America. Its habitat, mostly grasslands suitable for agriculture, has been increasingly covered by crops since the late 1800s. Microgyne trifurcata provides an excellent case study to understand how different environmental variables have affected the distribution area of a rare species. This study aims to estimate the impact of topoclimatic and land-use changes in the distribution of Microgyne trifurcata throughout the twentieth century. We carried out recent past and present distribution modeling using the Ensembles of Small Models (ESM) methodology. In this spatio-temporal study, we included climatic, topographic, and land-use variables. We classified the occurrences into two periods of the twentieth century. The first dates from 1901 to 1940, and the second, from 1960 to 2000, when the main cropping changes of the area occurred. The projected area between 1960 and 2000 provides for this species new suitable habitats toward the northeast of the area of study. Our results highlight the importance of assessing the combined impacts of climate and land-use changes on species distributions over time. This study shows that the potential area of Microgyne trifurcata decreased and underwent fragmentation throughout the twentieth century when these variables combined are used to model its distribution. Our outcomes prompt future studies on the vulnerability of Microgyne trifurcata to outline conservation strategies.

Investigating niches and distribution of a rare species in a hierarchical framework: Virginia’s Warbler (Leiothlypis virginiae) at its northeastern range limit

Ensemble of small models (ESMs) is a technique to overcome the problem of few occurrence points. Applying the ESMs in a spatially hierarchical framework could increase the accuracy of predictions and conclusions by restricting available habitat at sequentially finer spatial scales. Our objective was to show how applying ESMs in a hierarchical habitat selection framework could help to understand rare species’ niches at various scales. We compared the accuracy of ESMs made by committee averaging and weighted averaging methods. We also compared the predictive power of ESMs made by various modeling techniques for Virginia’s warbler (Leiothlypis virginiae) at its northeastern range limit. We defined biologically relevant hierarchical orders of habitat selection for Virginia’s warbler in the Black Hills, U.S.A. We modeled habitat suitabity at the broadest scale as a function of bioclimatic covariates and at finer scales as functions of landcover, soil group and landscape covariates. The performance of modeling techniques varied among scales. Using the committee averaging method led to more accurate results than weighted averaging. At the broadest order, Virginia’s warbler had a narrow climatic niche. The importance of covariates changed across finer orders, such that at broader orders many covariates were important whereas at finer orders certain covariates became more important than others. We conclude that applying ESMs within a hierarchical framework can lead to detailed information about rare species’ niches, limiting factors at each habitat selection order, and potential distribution, which could help inform multiscale management.