Habitat-based statistical models for predicting the spatial distribution of butterflies and day-flying moths in a fragmented landscape

1.  Most species' surveys and biodiversity inventories are limited by time and money. Therefore, it would be extremely useful to develop predictive models of animal distributions based on habitat, and to use these models to estimate species' densities and range sizes in poorly sampled regions.
2.  In this study, two sets of data were collected. The first set consisted of over 2000 butterfly transect counts, which were used to determine the relative density of each species in 16 major habitat types in a 35-km² area of fragmented landscape in north-west Wales. For the second set of data, the area was divided into 140 cells using a 500-m grid, and the extent of each habitat and the presence or absence of each butterfly and moth species was determined for each cell.
3.  Logistic regression was used to model the relationship between species' distribution and predicted density, based on habitat extent, in each grid square. The resultant models were used to predict butterfly distributions and occupancy at a range of spatial scales.
4.  Using a jack-knife procedure, our models successfully reclassified the presence or absence of species in a high percentage of grid squares (mean 83% agreement). There were highly significant relationships between the modelled probability of species occurring at regional and local scales and the number of grid squares occupied at those scales.
5.  We conclude that basic habitat data can be used to predict insect distributions and relative densities reasonably well within a fragmented landscape. It remains to be seen how accurate these predictions will be over a wider area.     

Novel methods improve prediction of species' distributions from occurrence data

Prediction of species' distributions is central to diverse applications in ecology, evolution and conservation science. There is increasing electronic access to vast sets of occurrence records in museums and herbaria, yet little effective guidance on how best to use this information in the context of numerous approaches for modelling distributions. To meet this need, we compared 16 modelling methods over 226 species from 6 regions of the world, creating the most comprehensive set of model comparisons to date. We used presence-only data to fit models, and independent presence-absence data to evaluate the predictions. Along with well-established modelling methods such as generalised additive models and GARP and BIOCLIM, we explored methods that either have been developed recently or have rarely been applied to modelling species' distributions. These include machine-learning methods and community models, both of which have features that may make them particularly well suited to noisy or sparse information, as is typical of species' occurrence data. Presence-only data were effective for modelling species' distributions for many species and regions. The novel methods consistently outperformed more established methods. The results of our analysis are promising for the use of data from museums and herbaria, especially as methods suited to the noise inherent in such data improve.     

Predicting habitat distribution and frequency from plant species co-occurrence data

Aim Species frequency data have been widely used in nature conservation to aid management decisions. To determine species frequencies, information on habitat occurrence is important: a species with a low frequency is not necessarily rare if it occupies all suitable habitats. Often, information on habitat distribution is available for small geographic areas only. We aim to predict grid‐based habitat occurrence from grid‐based plant species distribution data in a meso‐scale analysis. Location The study was carried out over two spatial extents: Germany and Bavaria. Methods Two simple models were set up to examine the number of characteristic plant species needed per grid cell to predict the occurrence of four selected habitats (species data from FlorKart, http://www.floraweb.de ). Both models were calibrated in Bavaria using available information on habitat distribution, validated for other federal states, and applied to Germany. First, a spatially explicit regression model (generalized linear model (GLM) with assumed binomial error distribution of response variable) was obtained. Second, a spatially independent optimization model was derived that estimated species numbers without using spatial information on habitat distribution. Finally, an additional uncalibrated model was derived that calculated the frequencies of 24 habitats. It was validated using NATURA2000 habitat maps. Results Using the Bavarian models it was possible to predict habitat distribution and frequency from the co‐occurrence of habitat‐specific species per grid cell. As the model validations for other German federal states were successful, the models were applied to all of Germany, and habitat distribution and frequencies could be retrieved for the national scale on the basis of habitat‐specific species co‐occurrences per grid cell. Using the third, uncalibrated model, which includes species distribution data only, it was possible to predict the frequencies of 24 habitats based on the co‐occurrence of 24% of formation‐specific species per grid cell. Predicted habitat frequencies deduced from this third model were strongly related to frequencies of NATURA2000 habitat maps. Main conclusions It was concluded that it is possible to deduce habitat distributions and frequencies from the co‐occurrence of habitat‐specific species. For areas partly covered by habitat mappings, calibrated models can be developed and extrapolated to larger areas. If information on habitat distribution is completely lacking, uncalibrated models can still be applied, providing coarse information on habitat frequencies. Predicted habitat distributions and frequencies can be used as a tool in nature conservation, for example as correction factors for species frequencies, as long as the species of interest is not included in the model set‐up.     

The contributions of topoclimate and land cover to species distributions and abundance: fine‐resolution tests for a mountain butterfly fauna

Aim Models relating species distributions to climate or habitat are widely used to predict the effects of global change on biodiversity. Most such approaches assume that climate governs coarse‐scale species ranges, whereas habitat limits fine‐scale distributions. We tested the influence of topoclimate and land cover on butterfly distributions and abundance in a mountain range, where climate may vary as markedly at a fine scale as land cover. Location Sierra de Guadarrama (Spain, southern Europe) Methods We sampled the butterfly fauna of 180 locations (89 in 2004, 91 in 2005) in a 10,800 km² region, and derived generalized linear models (GLMs) for species occurrence and abundance based on topoclimatic (elevation and insolation) or habitat (land cover, geology and hydrology) variables sampled at 100‐m resolution using GIS. Models for each year were tested against independent data from the alternate year, using the area under the receiver operating characteristic curve (AUC) (distribution) or Spearman's rank correlation coefficient (r_s) (abundance). Results In independent model tests, 74% of occurrence models achieved AUCs of > 0.7, and 85% of abundance models were significantly related to observed abundance. Topoclimatic models outperformed models based purely on land cover in 72% of occurrence models and 66% of abundance models. Including both types of variables often explained most variation in model calibration, but did not significantly improve model cross‐validation relative to topoclimatic models. Hierarchical partitioning analysis confirmed the overriding effect of topoclimatic factors on species distributions, with the exception of several species for which the importance of land cover was confirmed. Main conclusions Topoclimatic factors may dominate fine‐resolution species distributions in mountain ranges where climate conditions vary markedly over short distances and large areas of natural habitat remain. Climate change is likely to be a key driver of species distributions in such systems and could have important effects on biodiversity. However, continued habitat protection may be vital to facilitate range shifts in response to climate change.     

Building better wildlife-habitat models

Wildlife-habitat models are an important tool in wildlife management today, and by far the majority of these predict aspects of species distribution (abundance or presence) as a proxy measure of habitat quality. Unfortunately, few are tested on independent data., and of those that are, few show useful predictive skill. We demonstrate that six critical assumptions underlie distribution based wildlife-habitat models, all of which must be valid for the model to predict habitat quality. We outline these assumptions in a meta-model, and discuss methods for their validation. Even where all six assumptions show a high level of validity, there is still a strong likelihood that the model will not predict habitat quality. However, the meta-model does suggest habitat quality can be predicted more accurately if distributional data are ignored, and variables more indicative of habitat quality are modelled instead.     

Relating habitat and climatic niches in birds

Predicting species' responses to the combined effects of habitat and climate changes has become a major challenge in ecology and conservation biology. However, the effects of climatic and habitat gradients on species distributions have generally been considered separately. Here, we explore the relationships between the habitat and thermal dimensions of the ecological niche in European common birds. Using data from the French Breeding Bird Survey, a large-scale bird monitoring program, we correlated the habitat and thermal positions and breadths of 74 bird species, controlling for life history traits and phylogeny. We found that cold climate species tend to have niche positions in closed habitats, as expected by the conjunction of the biogeographic history of birds' habitats, and their current continent-scale gradients. We also report a positive correlation between thermal and habitat niche breadths, a pattern consistent with macroecological predictions concerning the processes shaping species' distributions. Our results suggest that the relationships between the climatic and habitat components of the niche have to be taken into account to understand and predict changes in species' distributions.     

Testing a Mahalanobis Distance Model of Black Bear Habitat Use in the Ouachita Mountains of Oklahoma

Abstract: Regional wildlife-habitat models are commonly developed but rarely tested with truly independent data. We tested a published habitat model for black bears (Ursus americanus) with new data collected in a different site in the same ecological region (i.e., Ouachita Mountains of Arkansas and Oklahoma, USA). We used a Mahalanobis distance model developed from relocations of black bears in Arkansas to produce a map layer of Mahalanobis distances on a study area in neighboring Oklahoma. We tested this modeled map layer with relocations of black bears on the Oklahoma area. The distributions of relocations of female black bears were consistent with model predictions. We conclude that this modeling approach can be used to predict regional suitability for a species of interest.     

Global habitat suitability models of terrestrial mammals

Detailed large-scale information on mammal distribution has often been lacking, hindering conservation efforts. We used the information from the 2009 IUCN Red List of Threatened Species as a baseline for developing habitat suitability models for 5027 out of 5330 known terrestrial mammal species, based on their habitat relationships. We focused on the following environmental variables: land cover, elevation and hydrological features. Models were developed at 300 m resolution and limited to within species' known geographical ranges. A subset of the models was validated using points of known species occurrence. We conducted a global, fine-scale analysis of patterns of species richness. The richness of mammal species estimated by the overlap of their suitable habitat is on average one-third less than that estimated by the overlap of their geographical ranges. The highest absolute difference is found in tropical and subtropical regions in South America, Africa and Southeast Asia that are not covered by dense forest. The proportion of suitable habitat within mammal geographical ranges correlates with the IUCN Red List category to which they have been assigned, decreasing monotonically from Least Concern to Endangered. These results demonstrate the importance of fine-resolution distribution data for the development of global conservation strategies for mammals.     

Body mass explains characteristic scales of habitat selection in terrestrial mammals

Niche theory in its various forms is based on those environmental factors that permit species persistence, but less work has focused on defining the extent, or size, of a species' environment: the area that explains a species' presence at a point in space. We proposed that this habitat extent is identifiable from a characteristic scale of habitat selection, the spatial scale at which habitat best explains species' occurrence. We hypothesized that this scale is predicted by body size. We tested this hypothesis on 12 sympatric terrestrial mammal species in the Canadian Rocky Mountains. For each species, habitat models varied across the 20 spatial scales tested. For six species, we found a characteristic scale; this scale was explained by species' body mass in a quadratic relationship. Habitat measured at large scales best-predicted habitat selection in both large and small species, and small scales predict habitat extent in medium-sized species. The relationship between body size and habitat selection scale implies evolutionary adaptation to landscape heterogeneity as the driver of scale-dependent habitat selection.     

Transferability and model evaluation in ecological niche modeling: a comparison of GARP and Maxent

We compared predictive success in two common algorithms for modeling species' ecological niches, GARP and Maxent, in a situation that challenged the algorithms to be general – that is, to be able to predict the species' distributions in broad unsampled regions, here termed transferability. The results were strikingly different between the two algorithms – Maxent models reconstructed the overall distributions of the species at low thresholds, but higher predictive levels of Maxent predictions reflected overfitting to the input data; GARP models, on the other hand, succeeded in anticipating most of the species' distributional potential, at the cost of increased (apparent, at least) commission error. Receiver operating characteristic (ROC) tests were weak in discerning models able to predict into broad unsampled areas from those that were not. Such transferability is clearly a novel challenge for modeling algorithms, and requires different qualities than does predicting within densely sampled landscapes – in this case, Maxent was transferable only at very low thresholds, and biases and gaps in input data may frequently affect results based on higher Maxent thresholds, requiring careful interpretation of model results.     

Dynamic distribution modelling: predicting the present from the past

Confidence in projections of the future distributions of species requires demonstration that recently-observed changes could have been predicted adequately. Here we use a dynamic model framework to demonstrate that recently-observed changes at the expanding northern boundaries of three British butterfly species can be predicted with good accuracy. Previous work established that the distributions of the study species currently lag behind climate change, and so we presumed that climate is not currently a major constraint at the northern range margins of our study species. We predicted 1970–2000 distribution changes using a colonisation model, MIGRATE, superimposed on a high-resolution map of habitat availability. Thirty-year rates and patterns of distribution change could be accurately predicted for each species (κ goodness-of-fit of models >0.64 for all three species, corresponding to >83% of grid cells correctly assigned), using a combination of individual species traits, species-specific habitat associations and distance-dependent dispersal. Sensitivity analyses showed that population productivity was the most important determinant of the rate of distribution expansion (variation in dispersal rate was not studied because the species are thought to be similar in dispersal capacity), and that each species' distribution prior to expansion was critical in determining the spatial pattern of the current distribution. In future, modelling approaches that combine climate suitability and spatially-explicit population models, incorporating demographic variables and habitat availability, are likely to be valuable tools in projecting species' responses to climatic change and hence in anticipating management to facilitate species' dispersal and persistence.     

Analysis of factors implicated in the recent decline of Australia's mammal fauna

Aim To assess whether eight factors thought to be involved in the extinction process can explain the pattern of recent decline in Australia's mammal fauna. Location Australia. Methods We compiled the first comprehensive lists of mammal species extant at the time of European settlement in each of Australia's 76 mainland regions, and assigned a current conservation status to each species in each region to derive an index of faunal attrition. We then sought to explain the observed region‐to‐region variation in attrition (the dependent variable) by building a series of models using variables representing the eight factors. Results A strong geographically based pattern of attrition emerged, with faunal losses being greatest in arid regions and least in areas of high rainfall. The Akaike information criterion showed support for one model that explained 93% of the region‐to‐region variation in attrition. Its six variables all made independent contributions towards explaining the observed variation. Two were environmental variables, namely mean annual rainfall (a surrogate for regional productivity) and environmental change (a measure of post‐European disturbance). The other four were faunal variables, namely phylogenetic similarity, body‐weight distribution, area (as a surrogate for extent of occurrence), and proportion of species that usually shelter on the ground (rather than in rock piles, burrows or trees). Main conclusions In combination with historical evidence, the analysis provides an explicit basis for setting priorities among regions and species. It also shows that the long‐term recovery of populations of many species of Australian mammals will require introduced predator suppression as well as extensive habitat management that includes controlling feral herbivores. Specifically, habitat management should restore aspects of productivity relevant to the types of species at risk and ensure the continual availability of suitable refuges from physiological stressors.     

Validation of outlier loci through replication in independent data sets: a test on Arabis alpina

Outlier detection and environmental association analysis are common methods to search for loci or genomic regions exhibiting signals of adaptation to environmental factors. However, a validation of outlier loci and corresponding allele distribution models through functional molecular biology or transplant/common garden experiments is rarely carried out. Here, we employ another method for validation, namely testing outlier loci in specifically designed, independent data sets. Previously, an outlier locus associated with three different habitat types had been detected in Arabis alpina. For the independent validation data set, we sampled 30 populations occurring in these three habitat types across five biogeographic regions of the Swiss Alps. The allele distribution model found in the original study could not be validated in the independent test data set: The outlier locus was no longer indicative of habitat‐mediated selection. We propose several potential causes of this failure of validation, of which unaccounted genetic structure and technical issues in the original data set used to detect the outlier locus were most probable. Thus, our study shows that validating outlier loci and allele distribution models in independent data sets is a helpful tool in ecological genomics which, in the case of positive validation, adds confidence to outlier loci and their association with environmental factors or, in the case of failure of validation, helps to explain inconsistencies.     

Modelling invasion for a habitat generalist and a specialist plant species

Predicting suitable habitat and the potential distribution of invasive species is a high priority for resource managers and systems ecologists. Most models are designed to identify habitat characteristics that define the ecological niche of a species with little consideration to individual species' traits. We tested five commonly used modelling methods on two invasive plant species, the habitat generalist Bromus tectorum and habitat specialist Tamarix chinensis , to compare model performances, evaluate predictability, and relate results to distribution traits associated with each species. Most of the tested models performed similarly for each species; however, the generalist species proved to be more difficult to predict than the specialist species. The highest area under the receiver-operating characteristic curve values with independent validation data sets of B. tectorum and T. chinensis was 0.503 and 0.885, respectively. Similarly, a confusion matrix for B. tectorum had the highest overall accuracy of 55%, while the overall accuracy for T. chinensis was 85%. Models for the generalist species had varying performances, poor evaluations, and inconsistent results. This may be a result of a generalist's capability to persist in a wide range of environmental conditions that are not easily defined by the data, independent variables or model design. Models for the specialist species had consistently strong performances, high evaluations, and similar results among different model applications. This is likely a consequence of the specialist's requirement for explicit environmental resources and ecological barriers that are easily defined by predictive models. Although defining new invaders as generalist or specialist species can be challenging, model performances and evaluations may provide valuable information on a species' potential invasiveness.     

Poor Transferability of Species Distribution Models for a Pelagic Predator,the Grey Petrel,Indicates Contrasting Habitat Preferences across Ocean Basins

Species distribution models (SDMs) are increasingly applied in conservation management to predict suitable habitat for poorly known populations. High predictive performance of SDMs is evident in validations performed within the model calibration area (interpolation), but few studies have assessed SDM transferability to novel areas (extrapolation), particularly across large spatial scales or pelagic ecosystems. We performed rigorous SDM validation tests on distribution data from three populations of a long-ranging marine predator, the grey petrel Procellaria cinerea, to assess model transferability across the Southern Hemisphere (25-65°S). Oceanographic data were combined with tracks of grey petrels from two remote sub-Antarctic islands (Antipodes and Kerguelen) using boosted regression trees to generate three SDMs: one for each island population, and a combined model. The predictive performance of these models was assessed using withheld tracking data from within the model calibration areas (interpolation), and from a third population, Marion Island (extrapolation). Predictive performance was assessed using k-fold cross validation and point biserial correlation. The two population-specific SDMs included the same predictor variables and suggested birds responded to the same broad-scale oceanographic influences. However, all model validation tests, including of the combined model, determined strong interpolation but weak extrapolation capabilities. These results indicate that habitat use reflects both its availability and bird preferences, such that the realized distribution patterns differ for each population. The spatial predictions by the three SDMs were compared with tracking data and fishing effort to demonstrate the conservation pitfalls of extrapolating SDMs outside calibration regions. This exercise revealed that SDM predictions would have led to an underestimate of overlap with fishing effort and potentially misinformed bycatch mitigation efforts. Although SDMs can elucidate potential distribution patterns relative to large-scale climatic and oceanographic conditions, knowledge of local habitat availability and preferences is necessary to understand and successfully predict region-specific realized distribution patterns.     

Modelling species distribution in complex environments: an evaluation of predictive ability and reliability in five shorebird species

Aim  Many wader populations around the world are declining as a consequence of habitat degradation or loss. It is therefore important to identify species-specific habitat demands accurately and to define the important factors explaining species distribution, in order to develop tools that can be used in conservation planning. The aim of this study is to create reliable, functional and ecologically interpretable predictive distribution models for five breeding wader species.
Location  The archipelago of SW Finland in the Baltic Sea.
Methods  We used multivariate adaptive regression splines (MARS) to create single-species and multiresponse distribution models based on 525 study islands and 12 abiotic and biotic environmental variables. Model evaluation was carried out on independent data not used in model building (100 + 116 islands). The models were tested for discrimination with receiver-operating characteristic statistics and for calibration with Millers calibration statistics (MCS).
Results  The single-species models for the turnstone ( Arenaria interpres ), redshank ( Tringa totanus ) and oystercatcher ( Haematopus ostralegus ) showed good predictive abilities, regarding both discrimination and calibration, when evaluated on independent data. The multiresponse models for the less prevalent species, common sandpiper ( Actitis hypoleucos ) and the common ringed plover ( Charadrius hiaticula ) had better discriminative abilities than the single-species models. The most influential predictor overall was occurrence of small larids. Exposure, area of forest and low and flat areas were also important, as well as shore habitats.
Main conclusions  We found that the ability of MARS to fit non-linear and multiresponse models makes it a useful method to quantitatively relate species occurrence to environmental characteristics of a complex environment.     

Propagule vs. niche limitation: untangling the mechanisms behind plant species' distributions

Distinguishing the roles of propagule limitation and niche requirements in controlling plant species distributions is important for understanding community structure, invasion, and restoration. We used species distribution models based on plant and environmental survey data to assess the strength of species' affinities for particular environmental conditions. We hypothesized that species with statistically detectable environmental requirements were primarily niche-limited, while species with weak habitat affinities were primarily propagule-limited. We tested this hypothesis via a seeding experiment in which we compared species' reproductive fitness in occupied and unoccupied sites. Species that appeared to be niche-limited based on distribution models had lower fitness when planted in unoccupied sites, while species that models suggested were propagule-limited had equivalent fitness when planted in occupied and unoccupied sites. Our results demonstrate that within a single community, both species limited primarily by niche availability or primarily by propagule availability can be identified using observational data.     

Forest and Landscape Structure as Predictors of Capercaillie Occurrence

ABSTRACT Capercaillie (Tetrao urogallus) is a large, endangered forest grouse species with narrow habitat preferences and large spatial requirements that make it susceptible to habitat changes at different spatial scales. Our aim was to evaluate the relative power of variables relating to forest versus landscape structure in predicting capercaillie occurrence at different spatial scales. We investigated capercaillie-habitat relationships at the scales of forest stand and forest-stand mosaic in 2 Swiss regions. We assessed forest structure from aerial photographs in 52 study plots each 5 km². We classified plots into one of 3 categories denoting the observed local population trend (stable, declining, extinct), and we compared forest structure between categories. At the stand scale, we used presence-absence data for grid cells within the plots to build predictive habitat models based on logistic regression. At this scale, habitat models that included only variables relating to forest structure explained the occurrence of capercaillie only in part, whereas variables selected by the models differed between regions. Including variables relating to landscape features improved the models significantly. At the scale of stand mosaic, variables describing forest structure (e.g., mean canopy cover, proportion of open forest, and proportion of multistoried forest) differed between plot categories. We conclude that small-scale forest structure has limited power to predict capercaillie occurrence at the stand scale, but that it explains well at the scale of the stand mosaic. Including variables for landscape structure improves predictions at the forest-stand scale. Habitat models built with data from one region cannot be expected to predict the species occurrence in other regions well. Thus, multiscale approaches are necessary to better understand species-habitat relationships. Our results can help regional authorities and forest-management planners to identify areas where suitable habitat for capercaillie is not available in the required proportion and, thus, where management actions are needed to improve habitat suitability.     

A hierarchical Bayesian regression model that reduces uncertainty in material demand predictions

Predictions of metal consumption are vital for criticality assessments and sustainability analyses. Although demand for a material varies strongly by region and end-use sector, statistical models of demand typically predict demand using regression analyses at an aggregated global level (“fully pooled models”). “Un-pooled” regression models that predict demand at a disaggregated country or regional level face challenges due to limited data availability and large uncertainty. In this paper, we propose a Bayesian hierarchical model that can simultaneously identify heterogeneous demand parameters (like price and income elasticities) for individual regions and sectors, as well as global parameters. We demonstrate the model's value by estimating income and price elasticity of copper demand in five sectors (Transportation, Electrical, Construction, Manufacturing, and Other) and five regions (North America, Europe, Japan, China, and Rest of World). To validate the benefits of the Bayesian approach, we compare the model to both a “fully pooled” and an “un-pooled” model. The Bayesian model can predict global demand with similar uncertainty as a fully pooled regression model, while additionally capturing regional heterogeneity in income elasticity of demand. Compared to un-pooled models that predict demand for individual countries and sectors separately, our model reduces the uncertainty of parameter estimates by more than 50%. The hierarchical Bayesian modeling approach we propose can be used for various commodities, improving material demand projections used to study the impact of policies on mining sector emissions and informing investment in critical material production.     

Spatial risk assessment of livestock exposure to pumas in Patagonia, Argentina

Livestock predation and associated human-carnivore conflicts are increasing worldwide and require the development of methods and concepts for risk assessment and conflict management. Here we use knowledge on habitat preference and distribution of pumas and provide a first assessment of the spatial risk of livestock to puma depredation in Patagonian ranches, Argentina. In an initial step, we developed a rule-based habitat model in a Geographic Information System (GIS) to predict the distribution of puma habitat at a regional scale in Patagonia. We then used empirically derived puma occurrence records from Patagonian ranches 1) to test our regional habitat predictions, and 2) to evaluate if paddock characteristics (vegetation cover, topography, and distance to roads) contribute to explain puma occurrences within ranches. Finally, we simulated three livestock management scenarios differing in their spatial and seasonal allocation of livestock to paddocks, and compared the likelihood of livestock exposure to pumas among scenarios. At a regional scale, 22% of the study region was predicted to be suitable for puma home ranges. The greatest uncertainty in these predictions resulted from assumptions on woody vegetation cover requirements at the home range scale. Within ranches, puma occurrences were positively associated with paddock topography, woody vegetation cover on paddocks, and proximity to predicted regional puma habitat. Comparing the risk of predation by puma among simulated livestock management scenarios implied that rotating livestock during seasons may help to reduce the likelihood of livestock exposure to pumas. Our results show the usefulness of rule-based habitat models for describing broad-scale carnivore distributions and for aiding risk assessments to mitigate conflicts between predators and human activities.