Correlative Species Distribution Models Research Articles

Estimate global risks of a forest disease under current and future climates using species distribution model and simple thermal model – Pine Wilt disease as a model case

Predicting distributions of a pest species is an important part of Pest Risk Analysis, but it is not always an easy task. Expert knowledge may help, yet, validation with field data is essential. Pine Wilt Disease (hereafter PWD) caused by the presence of the Pine wood nematode (hereafter PWN), Bursaphelenchus xylophilus is one of the most severe forest diseases in the world. The symptom development of this disease only occurs in a warmer region while the nematode itself could be widely distributed in cooler regions in its native range. Isotherm approaches have been used to estimate the distribution of this disease, but these models have not been well tested with field data. A correlative species distribution model, MaxEnt, is used to evaluate the climatic suitability for PWD at a global scale, along with a simple thermal model. The MaxEnt analysis indicated that most of the current distribution of PWD is explained by the average monthly mean temperatures of the warmest three months and aridity. The thermal model suggested larger PWD risk zones, particularly in hot and arid areas. The current distribution of some susceptible host species (e.g., Pinus sylvestris) is largely outside the area suitable for the development of PWD, but results using projected future climate showed that half of those areas become at risk from PWD in future. Species distribution models such as MaxEnt are useful tools for the evaluation of not only the likely potential distributions of pests but also areas where symptom development could occur under various climatic conditions, even exotic pests not yet present in a region. Our results suggest that PWD will undoubtedly pose a major threat to European pine species if climate change proceeds as currently projected.

Modelling the spatial variation of vital rates: An evaluation of the strengths and weaknesses of correlative species distribution models

AbstractAimSpecies distribution models based on breeding occurrence data allow for identifying both environmental drivers and geographic areas potentially relevant for breeding. However, the interpretation of model predictions in terms of reproductive performance should be further investigated, as this information is crucial for conservation planning. We evaluated the strengths and weaknesses of a correlative modelling approach based on breeding occurrence data (presence–absence) against another approach based on vital rates’ data (breeding success) for gaining insights on species persistence in the case of Great Bustards (Otis tarda).LocationSpain.MethodsBreeding occurrence and breeding success were independently modelled using generalized linear models and multimodel inference analyses. Sensitivities to the way in which the population parameter (breeding success) was defined were explored by building five versions of the dependent variable. We evaluated differences in model performance and identified areas of congruence for breeding occurrence and breeding success.ResultsThe agreement between the spatial predictions achieved by breeding occurrence and breeding success models differed substantially across databases, with the largest differences in occupied breeding areas. The deviance explained by the breeding occurrence model was 64.98% and ranged from 7.83% to 62.27% for the breeding success models. Model performance was higher for models calibrated within potential than within occupied breeding areas.Main conclusionsThe combination of data on both breeding occurrence and breeding success into a species distribution modelling framework showed the limitations of breeding occurrence models for inferring reproductive parameters. The definition of the population parameter as dependent variable was a key factor that strongly affected the inference of vital rates’ models. The approach allowed for discriminating between areas and landscape attributes necessary for the long‐term species persistence from others that may be relevant, but not so much for reproductive performance.

Integrating demography, dispersal and interspecific interactions into bird distribution models

Species’ ranges are primarily limited by the physiological (abiotic) tolerance of the species, described by their fundamental niche. Additionally, demographic processes, dispersal, and interspecific interactions with other species are shaping species distributions, resulting in the realised niche. Understanding the complex interplay between these drivers is vital for making robust biodiversity predictions to novel environments. Correlative species distribution models have been widely used to predict biodiversity response but also remain criticised, as they are not able to properly disentangle the abiotic and biotic drivers shaping species’ niches. Recent developments have thus focussed on 1) integrating demography and dispersal into species distribution models, and on 2) integrating interspecific interactions. Here, I review recent demographic and multi‐species modelling approaches and discuss critical aspects of these models that remain underexplored in general and in respect to birds, for example, the complex life histories of birds and other animals as well as the scale dependence of interspecific interactions. I conclude by formulating modelling guidelines for integrating the abiotic and biotic processes that limit species’ ranges, which will help to disentangle the complex roles of demography, dispersal and interspecific interactions in shaping species niches. Throughout, I pinpoint complexities of avian life cycles that are critical for consideration in the models and identify data requirements for operationalizing the different modelling steps.

Predicting invasion winners and losers under climate change.

Invasive plant species are human-introduced, nonnative, and damaging. They can restructure natural habitats, change community composition, disrupt food webs, and degrade ecosystem services. Global climatic changes are often thought to exacerbate invasions (1), but no consensus exists over the generality of this phenomenon (2). As a consequence, anticipating areas of high invasion risk or decreased impact in the future is still uncertain. By combining experimental biogeography with demographic models, Merow et al. (3) develop climate driven process-based models to predict population growth or decline under current and future climate scenarios for two established and damaging invasive plant species in New England, United States. They find that one invasive species will likely perform better, whereas the other will do worse in areas that are currently occupied. Forecasting the distribution and performance of invasive species under a changing climate comes with particular challenges. Invasive species have not had time to colonize all favorable habitats available in the nonnative range. This nonequilibrium situation is problematic for correlative species distribution models, which may then use incomplete occurrences in the nonnative range as inputs for predicting suitable climatic conditions. However, predictions can be improved by using occurrence data from the native range (4). A more concerning challenge is that invasives may be able to establish in environmental conditions the species has not encountered in the native range (4⇓⇓–7), even if those conditions occur but are unoccupied in the native range (6). Niche … [↵][1]1To whom correspondence should be addressed. Email: buckleyy{at}tcd.ie. [1]: #xref-corresp-1-1

Predictability in community dynamics.

The coupling between community composition and climate change spans a gradient from no lags to strong lags. The no-lag hypothesis is the foundation of many ecophysiological models, correlative species distribution modelling and climate reconstruction approaches. Simple lag hypotheses have become prominent in disequilibrium ecology, proposing that communities track climate change following a fixed function or with a time delay. However, more complex dynamics are possible and may lead to memory effects and alternate unstable states. We develop graphical and analytic methods for assessing these scenarios and show that these dynamics can appear in even simple models. The overall implications are that (1) complex community dynamics may be common and (2) detailed knowledge of past climate change and community states will often be necessary yet sometimes insufficient to make predictions of a community's future state.

Projections of future habitat use by Atlantic bluefin tuna: mechanistic vs. correlative distribution models

Climate change is likely to drive complex shifts in the distribution and ecology of marine species. Projections of future changes may vary, however, depending on the biological impact model used. In this study, we compared a correlative species distribution model and a simple mechanistic oxygen balance model for Atlantic bluefin tuna (Thunnus thynnus: ABFT) in the North Atlantic Ocean. Both models gave similar results for the recent historical time period, and suggested that ABFT generally occupy favourable metabolic habitats. Projections from an earth system model showed largely temperature-induced reductions in ABFT habitat in the tropical and sub-tropical Atlantic by 2100. However, the oxygen balance model showed more optimistic results in parts of the subpolar North Atlantic. This was partially due to an inherent ability to extrapolate beyond conditions currently encountered by pelagic longline fishing fleets. Projections included considerable uncertainty due to the simplicity of the biological models, and the coarse spatiotemporal resolution of the analyses. Despite these limitations, our results suggest that climate change is likely to increase metabolic stress on ABFT in sub-tropical habitats, but may improve habitat suitability in subpolar habitats, with implications for spawning and migratory behaviours, and availability to fishing fleets.

Comparative phylogeography in the Atlantic forest and Brazilian savannas: pleistocene fluctuations and dispersal shape spatial patterns in two bumblebees.

Background Bombus morio and B. pauloensis are sympatric widespread bumblebee species that occupy two major Brazilian biomes, the Atlantic forest and the savannas of the Cerrado. Differences in dispersion capacity, which is greater in B. morio, likely influence their phylogeographic patterns. This study asks which processes best explain the patterns of genetic variation observed in B. morio and B. pauloensis, shedding light on the phenomena that shaped the range of local populations and the spatial distribution of intra-specific lineages.ResultsResults suggest that Pleistocene climatic oscillations directly influenced the population structure of both species. Correlative species distribution models predict that the warmer conditions of the Last Interglacial contributed to population contraction, while demographic expansion happened during the Last Glacial Maximum. These results are consistent with physiological data suggesting that bumblebees are well adapted to colder conditions. Intra-specific mitochondrial genealogies are not congruent between the two species, which may be explained by their documented differences in dispersal ability.ConclusionsWhile populations of the high-dispersal B. morio are morphologically and genetically homogeneous across the species range, B. pauloensis encompasses multiple (three) mitochondrial lineages, and show clear genetic, geographic, and morphological differences. Because the lineages of B. pauloensis are currently exposed to distinct climatic conditions (and elevations), parapatric diversification may occur within this taxon. The eastern portion of the state of São Paulo, the most urbanized area in Brazil, represents the center of genetic diversity for B. pauloensis. Electronic supplementary materialThe online version of this article (doi:10.1186/s12862-016-0803-0) contains supplementary material, which is available to authorized users.

A comparison of absolute performance of different correlative and mechanistic species distribution models in an independent area.

To investigate the comparative abilities of six different bioclimatic models in an independent area, utilizing the distribution of eight different species available at a global scale and in Australia. Global scale and Australia. We tested a variety of bioclimatic models for eight different plant species employing five discriminatory correlative species distribution models (SDMs) including Generalized Linear Model (GLM), MaxEnt, Random Forest (RF), Boosted Regression Tree (BRT), Bioclim, together with CLIMEX (CL) as a mechanistic niche model. These models were fitted using a training dataset of available global data, but with the exclusion of Australian locations. The capabilities of these techniques in projecting suitable climate, based on independent records for these species in Australia, were compared. Thus, Australia is not used to calibrate the models and therefore it is as an independent area regarding geographic locations. To assess and compare performance, we utilized the area under the receiver operating characteristic (ROC) curves (AUC), true skill statistic (TSS), and fractional predicted areas for all SDMs. In addition, we assessed satisfactory agreements between the outputs of the six different bioclimatic models, for all eight species in Australia. The modeling method impacted on potential distribution predictions under current climate. However, the utilization of sensitivity and the fractional predicted areas showed that GLM, MaxEnt, Bioclim, and CL had the highest sensitivity for Australian climate conditions. Bioclim calculated the highest fractional predicted area of an independent area, while RF and BRT were poor. For many applications, it is difficult to decide which bioclimatic model to use. This research shows that variable results are obtained using different SDMs in an independent area. This research also shows that the SDMs produce different results for different species; for example, Bioclim may not be good for one species but works better for other species. Also, when projecting a “large” number of species into novel environments or in an independent area, the selection of the “best” model/technique is often less reliable than an ensemble modeling approach. In addition, it is vital to understand the accuracy of SDMs' predictions. Further, while TSS, together with fractional predicted areas, are appropriate tools for the measurement of accuracy between model results, particularly when undertaking projections on an independent area, AUC has been proved not to be. Our study highlights that each one of these models (CL, Bioclim, GLM, MaxEnt, BRT, and RF) provides slightly different results on projections and that it may be safer to use an ensemble of models.

Impacts of Climate Change on the Global Invasion Potential of the African Clawed Frog Xenopus laevis.

By altering or eliminating delicate ecological relationships, non-indigenous species are considered a major threat to biodiversity, as well as a driver of environmental change. Global climate change affects ecosystems and ecological communities, leading to changes in the phenology, geographic ranges, or population abundance of several species. Thus, predicting the impacts of global climate change on the current and future distribution of invasive species is an important subject in macroecological studies. The African clawed frog (Xenopus laevis), native to South Africa, possesses a strong invasion potential and populations have become established in numerous countries across four continents. The global invasion potential of X. laevis was assessed using correlative species distribution models (SDMs). SDMs were computed based on a comprehensive set of occurrence records covering South Africa, North America, South America and Europe and a set of nine environmental predictors. Models were built using both a maximum entropy model and an ensemble approach integrating eight algorithms. The future occurrence probabilities for X. laevis were subsequently computed using bioclimatic variables for 2070 following four different IPCC scenarios. Despite minor differences between the statistical approaches, both SDMs predict the future potential distribution of X. laevis, on a global scale, to decrease across all climate change scenarios. On a continental scale, both SDMs predict decreasing potential distributions in the species’ native range in South Africa, as well as in the invaded areas in North and South America, and in Australia where the species has not been introduced. In contrast, both SDMs predict the potential range size to expand in Europe. Our results suggest that all probability classes will be equally affected by climate change. New regional conditions may promote new invasions or the spread of established invasive populations, especially in France and Great Britain.

Benchmarking novel approaches for modelling speciesrange dynamics.

Increasing biodiversity loss due to climate change is one of the most vital challenges of the 21st century. To anticipate and mitigate biodiversity loss, models are needed that reliably project species' range dynamics and extinction risks. Recently, several new approaches to model range dynamics have been developed to supplement correlative species distribution models (SDMs), but applications clearly lag behind model development. Indeed, no comparative analysis has been performed to evaluate their performance. Here, we build on process-based, simulated data for benchmarking five range (dynamic) models of varying complexity including classical SDMs, SDMs coupled with simple dispersal or more complex population dynamic models (SDM hybrids), and a hierarchical Bayesian process-based dynamic range model (DRM). We specifically test the effects of demographic and community processes on model predictive performance. Under current climate, DRMs performed best, although only marginally. Under climate change, predictive performance varied considerably, with no clear winners. Yet, all range dynamic models improved predictions under climate change substantially compared to purely correlative SDMs, and the population dynamic models also predicted reasonable extinction risks for most scenarios. When benchmarking data were simulated with more complex demographic and community processes, simple SDM hybrids including only dispersal often proved most reliable. Finally, we found that structural decisions during model building can have great impact on model accuracy, but prior system knowledge on important processes can reduce these uncertainties considerably. Our results reassure the clear merit in using dynamic approaches for modelling species' response to climate change but also emphasize several needs for further model and data improvement. We propose and discuss perspectives for improving range projections through combination of multiple models and for making these approaches operational for large numbers of species.

Unpacking the mechanisms captured by a correlative species distribution model to improve predictions of climate refugia.

Climate refugia are regions that animals can retreat to, persist in and potentially then expand from under changing environmental conditions. Most forecasts of climate change refugia for species are based on correlative species distribution models (SDMs) using long-term climate averages, projected to future climate scenarios. Limitations of such methods include the need to extrapolate into novel environments and uncertainty regarding the extent to which proximate variables included in the model capture processes driving distribution limits (and thus can be assumed to provide reliable predictions under new conditions). These limitations are well documented; however, their impact on the quality of climate refugia predictions is difficult to quantify. Here, we develop a detailed bioenergetics model for the koala. It indicates that range limits are driven by heat-induced water stress, with the timing of rainfall and heat waves limiting the koala in the warmer parts of its range. We compare refugia predictions from the bioenergetics model with predictions from a suite of competing correlative SDMs under a range of future climate scenarios. SDMs were fitted using combinations of long-term climate and weather extremes variables, to test how well each set of predictions captures the knowledge embedded in the bioenergetics model. Correlative models produced broadly similar predictions to the bioenergetics model across much of the species' current range - with SDMs that included weather extremes showing highest congruence. However, predictions in some regions diverged significantly when projecting to future climates due to the breakdown in correlation between climate variables. We provide unique insight into the mechanisms driving koala distribution and illustrate the importance of subtle relationships between the timing of weather events, particularly rain relative to hot-spells, in driving species-climate relationships and distributions. By unpacking the mechanisms captured by correlative SDMs, we can increase our certainty in forecasts of climate change impacts on species.

On the dangers of model complexity without ecological justification in species distribution modeling

Although biogeographic patterns are the product of complex ecological processes, the increasing complexity of correlative species distribution models (SDMs) is not always motivated by ecological theory, but by model fit. The validity of model projections, such as shifts in a species’ climatic niche, becomes questionable particularly during extrapolations, such as for future no-analog climate conditions. To examine the effects of model complexity on SDM predictive performance, we fit statistical models of varying complexity to simulated species occurrence data arising from data-generating processes that assume differing degrees of distributional symmetry in environmental space, interaction effects, and coverage in climate space. Mismatches between data-generating processes and statistical models (i.e., different functional forms) led to poor predictive performance when extrapolating to new climate-space and greater variation in extrapolated predictions for overly complex models. In contrast, performance issues were not apparent when using independent evaluation data from the training region. These results draw into question the use of highly flexible models for prediction without ecological justification.

Integrative Taxonomy of Southeast Asian Snail-Eating Turtles (Geoemydidae: Malayemys) Reveals a New Species and Mitochondrial Introgression.

Based on an integrative taxonomic approach, we examine the differentiation of Southeast Asian snail-eating turtles using information from 1863 bp of mitochondrial DNA, 12 microsatellite loci, morphology and a correlative species distribution model. Our analyses reveal three genetically distinct groups with limited mitochondrial introgression in one group. All three groups exhibit distinct nuclear gene pools and distinct morphology. Two of these groups correspond to the previously recognized species Malayemys macrocephala (Chao Phraya Basin) and M. subtrijuga (Lower Mekong Basin). The third and genetically most divergent group from the Khorat Basin represents a previously unrecognized species, which is described herein. Although Malayemys are extensively traded and used for religious release, only few studied turtles appear to be translocated by humans. Historic fluctuations in potential distributions were assessed using species distribution models (SDMs). The Last Glacial Maximum (LGM) projection of the predictive SDMs suggests two distinct glacial distribution ranges, implying that the divergence of M. macrocephala and M. subtrijuga occurred in allopatry and was triggered by Pleistocene climate fluctuations. Only the projection derived from the global circulation model MIROC reveals a distinct third glacial distribution range for the newly discovered Malayemys species.

Uncertainty in predicting range dynamics of endemic alpine plants under climate warming.

Correlative species distribution models have long been the predominant approach to predict species' range responses to climate change. Recently, the use of dynamic models is increasingly advocated for because these models better represent the main processes involved in range shifts and also simulate transient dynamics. A well-known problem with the application of these models is the lack of data for estimating necessary parameters of demographic and dispersal processes. However, what has been hardly considered so far is the fact that simulating transient dynamics potentially implies additional uncertainty arising from our ignorance of short-term climate variability in future climatic trends. Here, we use endemic mountain plants of Austria as a case study to assess how the integration of decadal variability in future climate affects outcomes of dynamic range models as compared to projected long-term trends and uncertainty in demographic and dispersal parameters. We do so by contrasting simulations of a so-called hybrid model run under fluctuating climatic conditions with those based on a linear interpolation of climatic conditions between current values and those predicted for the end of the 21st century. We find that accounting for short-term climate variability modifies model results nearly as differences in projected long-term trends and much more than uncertainty in demographic/dispersal parameters. In particular, range loss and extinction rates are much higher when simulations are run under fluctuating conditions. These results highlight the importance of considering the appropriate temporal resolution when parameterizing and applying range-dynamic models, and hybrid models in particular. In case of our endemic mountain plants, we hypothesize that smoothed linear time series deliver more reliable results because these long-lived species are primarily responsive to long-term climate averages.

Niche expansion and temperature sensitivity of tropical African montane forests

AbstractAim Climate and land‐use change will have a dramatic impact on future ecosystems through alterations to species ranges and community composition. When forming conservation strategies, correlative species distribution models are often created to assess risks for individual species. These models are based on the assumption of climatic equilibrium, such that the modern range is representative of the full range of conditions under which species could thrive. However, the palaeo‐ecological record illustrates examples of disequilibrium in species today, and recent studies suggest that many species could occur in much broader climatic settings than previously thought. Montane ecosystems are thought to be at disproportionate risk due to temperature sensitivity and restricted geographical ranges. However, in the Afrotropics the palaeo‐ecological record shows that montane forest taxa expanded into the lowlands numerous times, suggesting a possible tolerance to warm temperatures.Location Africa.Methods We integrate palaeo‐ecological and palaeo‐climatic data in order to compare climate conditions in which species are currently found with those in the past. We use species distribution models to construct potential modern ranges for Afromontane species based on modern distributions and distributions in the palaeo‐ecological record in order to evaluate the equilibrium of species ranges.Results We show that many Afromontane trees have occupied warmer climates in the past, which suggests that the current low‐elevation boundaries are not set by climate. Interestingly, the species with the largest disequilibrium between palaeo‐ and modern distributions are those whose modern distributions show the least temperature sensitivity. Mapping of species potential ranges based on modern and palaeo‐ distributions clearly shows that suitable climate conditions exist today in the lowlands for less temperature‐sensitive species.Main conclusions These results imply that the current range of these forest trees does not necessarily inform risk from climatic change, and that human land use may be the major pressure for many species in the future.

Using a dynamic forest model to predict tree species distributions

AbstractAimIt has been suggested that predicting species distributions requires a process‐based and preferably dynamic approach. If dynamic models are to contribute towards understanding species distributions, uncertainties related to their spatial extrapolation and bioclimatic parameters need to be addressed. Here, we analyse the potential of a forest gap model for predicting species distributions.LocationPacific Northwest of North America (PNW).MethodsWe used the dynamic forest gap model ForClim, which includes climate, competition and demographic processes, to simulate the distribution of 18 tree species outside the domain of the data used for fitting. We explored model accuracy for species distributions at the regional scale by: (1) estimating species climatic tolerances so as to maximize agreement with regional distribution maps versus (2) employing a bioclimatic parameter set that produces high accuracy at the local scale. We then performed the opposite tests and simulated local forest composition in a small area in the PNW, using (3) the local bioclimatic parameters and (4) the bioclimatic parameters that produced the highest accuracy at the regional scale. We also compared the ForClim results with predictions from a standard correlative species distribution model (SDM).ResultsForClim produced regional species distributions with fair to very good agreement for 12 tree species. The optimized bioclimatic parameters consistently improved the accuracy of regional predictions compared with simulations run with the local parameters, and were consistent with SDM results. At the local scale, predictions using the local parameters conformed to descriptions of forest composition, but accuracy decreased strongly when using the regionally calibrated parameters.Main conclusionsForest gap models can predict regional species distributions, but at the cost of reduced accuracy at the local scale. Future applications of gap models to understand regional species distributions should include robust parameterization schemes and additional ecological processes that are important at large spatial scales (e.g. dispersal, disturbances).

Assessing future habitat availability for coastal lowland anurans in the Brazilian Atlantic rainforest

ABSTRACTGlobal warming is expected to cause several modifications to physical environments, and sea level rise is a certain outcome. However, assessment of the potential impacts caused by sea level rise on biodiversity is still emerging. Therefore, we assessed the combined impact of global warming and sea level rise on the potential distribution of 19 coastal lowland anurans in the biodiversity hotspot Atlantic Forest. We applied a correlative species distribution model (SDM) (BIOCLIM) and GIS-based spatial analyses. We evaluated the extent of changes of potential distributions under extreme and moderate global warming scenarios as well as two extreme sea level rise scenarios. Our results suggest wide areas of suitable habitat for most species in the future. However, for 15% of these species the SDMs predict massive losses of range extent as a result of a combination of global warming and sea level rise. Such observations highlight an immediate need to consider the potential effects of sea level rise in conservation action plans. Since the current potential distribution of these anuran species is likely underestimated, we also analyzed their environmental niche under current conditions in order to provide a baseline for further field surveys. Considering this current state of knowledge for such species, species distribution modeling to help gather further information on unknown species is desirable.

Community dynamics under environmental change: How can next generation mechanistic models improve projections of species distributions?

Environmental change is expected to shift the geographic range of species and communities. To estimate the consequences of these shifts for the functioning and stability of ecosystems, reliable predictions of alterations in species distributions are needed. Projections with correlative species distribution models, which correlate species’ distributions to the abiotic environment, have become a standard approach. Criticism of this approach centres around the omission of relevant biotic feedbacks and triggered the search for alternatives. A new generation of mechanistic process-based species distribution models aims at implementing formulations of relevant biotic processes to cover species’ life histories, physiology, dispersal abilities, evolution, and both intra- and interspecific interactions. Although this step towards more structural realism is considered important, it remains unclear whether the resulting projections are more reliable. Structural realism has the advantage that geographic range shifting emerges from the interplay of relevant abiotic and biotic processes. Having implemented the relevant response mechanisms, structural realistic models should better tackle the challenge of generating projections of species responses to (non-analogous) environmental change. However, reliable projections of future species ranges demand ecological information that is currently only available for few species. In this opinion paper, we discuss how the discrepancy between demand for structural realism on the one hand and the related knowledge gaps on the other hand affects the reliability of mechanistic species distribution models. We argue that omission of relevant processes potentially impairs projection accuracy (proximity of the mean outcome to the true value), particularly if species range shifts emerge from species and community dynamics. Yet, insufficient knowledge that limits model specification and parameterization, as well as process complexity, increases projection uncertainty (variance in the outcome of simulated model projections). The accuracy–uncertainty-relation reflects current limits to delivering reliable projections of range shifts. We propose a protocol to improve and communicate projection reliability. The protocol combines modelling and empirical research to efficiently fill critical knowledge gaps that currently limit the reliability of species and community projections.

Contemporary niche contraction affects climate change predictions for elephants and giraffes

AbstractAimClimate change assessments are largely based on correlative species distribution models (SDMs) that are sensible to spatial biases or incompleteness of input distribution data. We tested whether changes on the species' climatic niche resulting from recent human‐induced range contractions have a significant influence on SDM predictions of future species distributions.LocationAfrica.MethodsFor this study, we selected two highly detectable species with acknowledged human‐induced range contractions, namely the African savanna elephant (Loxodonta africana) and giraffe (Giraffa camelopardalis). We used presence data until the 1970s to describe each species' historical distribution, while more recent data characterized the contemporary distribution. We compared the temporal variation between these species distributions using multivariate analyses and the combination of four different SDM algorithms to predict historical, contemporary and future distributional ranges under climate change scenarios.ResultsWe show how range contraction differentially reduces the climatic variability associated with the species niche and has an important influence on the predictions of suitable climatic space and species vulnerability trend under climate change scenarios. Future predictions of the distribution of the elephant were mainly affected by the loss of occupied area at the margins of the historical distributions, resulting in a lesser predicted extent when using the contemporary dataset. As for the giraffe models, there were more dramatic consequences with large areas of West Africa failing to be predicted as suitable in the contemporary models, probably as a result from the loss of climatic information due to the species almost complete disappearance from that region.Main conclusionsOur findings support the importance of considering historical distributional ranges of species in climate change studies in order to account for their full climatic niche and to derive more reliable predictions of future distribution. This is particularly important in species for which distributional ranges have been strongly affected by human activities.

Physiology-based modelling approaches to characterize fish habitat suitability: Their usefulness and limitations

Models are useful tools for predicting the impact of global change on species distribution and abundance. As ectotherms, fish are being challenged to adapt or track changes in their environment, either in time through a phenological shift or in space by a biogeographic shift. Past modelling efforts have largely been based on correlative Species Distribution Models, which use known occurrences of species across landscapes of interest to define sets of conditions under which species are likely to maintain populations. The practical advantages of this correlative approach are its simplicity and the flexibility in terms of data requirements. However, effective conservation management requires models that make projections beyond the range of available data. One way to deal with such an extrapolation is to use a mechanistic approach based on physiological processes underlying climate change effects on organisms. Here we illustrate two approaches for developing physiology-based models to characterize fish habitat suitability. (i) Aerobic Scope Models (ASM) are based on the relationship between environmental factors and aerobic scope (defined as the difference between maximum and standard (basal) metabolism). This approach is based on experimental data collected by using a number of treatments that allow a function to be derived to predict aerobic metabolic scope from the stressor/environmental factor(s). This function is then integrated with environmental (oceanographic) data of current and future scenarios. For any given species, this approach allows habitat suitability maps to be generated at various spatiotemporal scales. The strength of the ASM approach relies on the estimate of relative performance when comparing, for example, different locations or different species. (ii) Dynamic Energy Budget (DEB) models are based on first principles including the idea that metabolism is organised in the same way within all animals. The (standard) DEB model aims to describe empirical relationships which can be found consistently within physiological data across the animal kingdom. The advantages of the DEB models are that they make use of the generalities found in terms of animal physiology and can therefore be applied to species for which little data or empirical observations are available. In addition, the limitations as well as useful potential refinements of these and other physiology-based modelling approaches are discussed. Inclusion of the physiological response of various life stages and modelling the patterns of extreme events observed in nature are suggested for future work.