Parameterizing mechanistic niche models in biophysical ecology: a review of empirical approaches

Mechanistic niche models are computational tools developed using biophysical principles to address grand challenges in ecology and evolution, such as the mechanisms that shape the fundamental niche and the adaptive significance of traits. Here, we review the empirical basis of mechanistic niche models in biophysical ecology, which are used to answer a broad array of questions in ecology, evolution and global change biology. We describe the experiments and observations that are frequently used to parameterize these models and how these empirical data are then incorporated into mechanistic niche models to predict performance, growth, survival and reproduction. We focus on the physiological, behavioral and morphological traits that are frequently measured and then integrated into these models. We also review the empirical approaches used to incorporate evolutionary processes, phenotypic plasticity and biotic interactions. We discuss the importance of validation experiments and observations in verifying underlying assumptions and complex processes. Despite the reliance of mechanistic niche models on biophysical theory, empirical data have and will continue to play an essential role in their development and implementation.

Biophysical models accurately characterize the thermal energetics of a small invasive passerine bird

Effective management of invasive species requires accurate predictions of their invasion potential in different environments. By considering species' physiological tolerances and requirements, biophysical mechanistic models can potentially deliver accurate predictions of where introduced species are likely to establish. Here, we evaluate biophysical model predictions of energy use by comparing them to experimentally obtained energy expenditure (EE) and thermoneutral zones (TNZs) for the common waxbill Estrilda astrild, a small-bodied avian invader. We show that biophysical models accurately predict TNZ and EE and that they perform better than traditional time-energy budget methods. Sensitivity analyses indicate that body temperature, metabolic rate, and feather characteristics were the most influential traits affecting model accuracy. This evaluation of common waxbill energetics represents a crucial step toward improved parameterization of biophysical models, eventually enabling accurate predictions of invasion risk for small (sub)tropical passerines.

Mechanistic models project bird invasions with accuracy

Invasive species pose a major threat to biodiversity and inflict massive economic costs. Effective management of bio-invasions depends on reliable predictions of areas at risk of invasion, as they allow early invader detection and rapid responses. Yet, considerable uncertainty remains as to how to predict best potential invasive distribution ranges. Using a set of mainly (sub)tropical birds introduced to Europe, we show that the true extent of the geographical area at risk of invasion can accurately be determined by using ecophysiological mechanistic models that quantify species' fundamental thermal niches. Potential invasive ranges are primarily constrained by functional traits related to body allometry and body temperature, metabolic rates, and feather insulation. Given their capacity to identify tolerable climates outside of contemporary realized species niches, mechanistic predictions are well suited for informing effective policy and management aimed at preventing the escalating impacts of invasive species.

Behavioural responses of a large, heat-sensitive mammal to climatic variation at multiple spatial scales

Climate warming creates energetic challenges for endothermic species by increasing metabolic and hydric costs of thermoregulation. Although endotherms can invoke an array of behavioural and physiological strategies for maintaining homeostasis, the relative effectiveness of those strategies in a climate that is becoming both warmer and drier is not well understood. In accordance with the heat dissipation limit theory which suggests that allocation of energy to growth and reproduction by endotherms is constrained by the ability to dissipate heat, we expected that patterns of habitat use by large, heat-sensitive mammals across multiple scales are critical for behavioural thermoregulation during periods of potential heat stress and that they must invest a large portion of time to maintain heat balance. To test our predictions, we evaluated mechanisms underpinning the effectiveness of bed sites for ameliorating daytime heat loads and potential heat stress across the landscape while accounting for other factors known to affect behaviour. We integrated detailed data on microclimate and animal attributes of moose Alces alces, into a biophysical model to quantify costs of thermoregulation at fine and coarse spatial scales. During summer, moose spent an average of 67.8% of daylight hours bedded, and selected bed sites and home ranges that reduced risk of experiencing heat stress. For most of the day, shade could effectively mitigate the risk of experiencing heat stress up to 10°C, but at warmer temperatures (up to 20°C) wet soil was necessary to maintain homeostasis via conductive heat loss. Consistent selection across spatial scales for locations that reduced heat load underscores the importance of the thermal environment as a driver of behaviour in this heat-sensitive mammal. Moose in North America have long been characterized as riparian-obligate species because of their dependence on woody plant species for food. Nevertheless, the importance of dissipating endogenous heat loads conductively through wet soil suggests riparian habitats also are critical thermal refuges for moose. Such refuges may be especially important in the face of a warming climate in which both high environmental temperatures and drier conditions will likely exacerbate limits to heat dissipation, especially for large, heat-sensitive animals.

Choosing among correlative, mechanistic, and hybrid models of species' niche and distribution

Different models are available to estimate species' niche and distribution. Mechanistic and correlative models have different underlying conceptual bases, thus generating different estimates of a species' niche and geographic extent. Hybrid models, which combining correlative and mechanistic approaches, are considered a promising strategy; however, no synthesis in the literature assessed their applicability for terrestrial vertebrates to allow best-choice model considering their strengths and trade-offs. Here, we provide a systematic review of studies that compared or integrated correlative and mechanistic models to estimate species' niche for terrestrial vertebrates under climate change. Our goal was to understand their conceptual, methodological, and performance differences, and the applicability of each approach. The studies we reviewed directly compared mechanistic and correlative predictions in terms of accuracy or estimated suitable area, however, without any quantitative analysis to support comparisons. Contrastingly, many studies suggest that instead of comparing approaches, mechanistic and correlative methods should be integrated (hybrid models). However, we stress that the best approach is highly context-dependent. Indeed, the quality and effectiveness of the prediction depends on the study's objective, methodological design, and which type of species' niche and geographic distribution estimated are more appropriate to answer the study's issue.

Field data confirm the ability of a biophysical model to predict wild primate body temperature

In the face of climate change there is an urgent need to understand how animal performance is affected by environmental conditions. Biophysical models that use principles of heat and mass transfer can be used to explore how an animal's morphology, physiology, and behavior interact with its environment in terms of energy, mass and water balances to affect fitness and performance. We used Niche Mapper™ (NM) to build a vervet monkey (Chlorocebus pygerythrus) biophysical model and tested the model's ability to predict core body temperature (T_b) variation and thermal stress against T_b and behavioral data collected from wild vervets in South Africa. The mean observed T_b in both males and females was within 0.5 °C of NM's predicted T_bs for 91% of hours over the five-year study period. This is the first time that NM's T_b predictions have been validated against field data from a wild endotherm. Overall, these results provide confidence that NM can accurately predict thermal stress and can be used to provide insight into the thermoregulatory consequences of morphological (e.g., body size, shape, fur depth), physiological (e.g. T_b plasticity) and behavioral (e.g., huddling, resting, shade seeking) adaptations. Such an approach allows users to test hypotheses about how animals adapt to thermoregulatory challenges and make informed predictions about potential responses to environmental change such as climate change or habitat conversion. Importantly, NM's animal submodel is a general model that can be adapted to other species, requiring only basic information on an animal's morphology, physiology and behavior.

Modeling Dragons: Using linked mechanistic physiological and microclimate models to explore environmental, physiological, and morphological constraints on the early evolution of dinosaurs

We employed the widely-tested biophysiological modeling software, Niche Mapper™ to investigate the metabolic function of the Late Triassic dinosaurs Plateosaurus and Coelophysis during global greenhouse conditions. We tested a variety of assumptions about resting metabolic rate, each evaluated within six microclimate models that bound paleoenvironmental conditions at 12° N paleolatitude, as determined by sedimentological and isotopic proxies for climate within the Chinle Formation of the southwestern United States. Sensitivity testing of metabolic variables and simulated “metabolic chamber” analyses support elevated “ratite-like” metabolic rates and intermediate “monotreme-like” core temperature ranges in these species of early saurischian dinosaur. Our results suggest small theropods may have needed partial to full epidermal insulation in temperate environments, while fully grown prosauropods would have likely been heat stressed in open, hot environments and should have been restricted to cooler microclimates such as dense forests or higher latitudes and elevations. This is in agreement with the Late Triassic fossil record and may have contributed to the latitudinal gap in the Triassic prosauropod record.

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.

Foraging Activity Pattern Is Shaped by Water Loss Rates in a Diurnal Desert Rodent

Although animals fine-tune their activity to avoid excess heat, we still lack a mechanistic understanding of such behaviors. As the global climate changes, such understanding is particularly important for projecting shifts in the activity patterns of populations and communities. We studied how foraging decisions vary with biotic and abiotic pressures. By tracking the foraging behavior of diurnal desert spiny mice in their natural habitat and estimating the energy and water costs and benefits of foraging, we asked how risk management and thermoregulatory requirements affect foraging decisions. We found that water requirements had the strongest effect on the observed foraging decisions. In their arid environment, mice often lose water while foraging for seeds and cease foraging even at high energetic returns when water loss is high. Mice also foraged more often when energy expenditure was high and for longer times under high seed densities and low predation risks. Gaining insight into both energy and water balance will be crucial to understanding the forces exerted by changing climatic conditions on animal energetics, behavior, and ecology.

Validation of a Mechanistic Model for Non-Invasive Study of Ecological Energetics in an Endangered Wading Bird with Counter-Current Heat Exchange in its Legs

Mechanistic models provide a powerful, minimally invasive tool for gaining a deeper understanding of the ecology of animals across geographic space and time. In this paper, we modified and validated the accuracy of the mechanistic model Niche Mapper for simulating heat exchanges of animals with counter-current heat exchange mechanisms in their legs and animals that wade in water. We then used Niche Mapper to explore the effects of wading and counter-current heat exchange on the energy expenditures of Whooping Cranes, a long-legged wading bird. We validated model accuracy against the energy expenditure of two captive Whooping Cranes measured using the doubly-labeled water method and time energy budgets. Energy expenditure values modeled by Niche Mapper were similar to values measured by the doubly-labeled water method and values estimated from time-energy budgets. Future studies will be able to use Niche Mapper as a non-invasive tool to explore energy-based limits to the fundamental niche of Whooping Cranes and apply this knowledge to management decisions. Basic questions about the importance of counter-current exchange and wading to animal physiological tolerances can also now be explored with the model.

Applications and implications of ecological energetics

The ecological processes that are crucial to an animal's growth, survival, and reproductive fitness have energetic costs. The imperative for an animal to meet these costs within the energetic constraints of the environment drives many aspects of animal ecology and evolution, yet has largely been overlooked in traditional ecological paradigms. The field of 'ecological energetics' is bringing comparative physiology out of the laboratory and, for the first time, is becoming broadly accessible to field ecologists addressing real-world questions at many spatial and temporal scales. In an era of unprecedented global environmental challenges, ecological energetics opens up the tantalising prospect of a more predictive, mechanistic understanding of the drivers of threatened species decline, delivering process-based modelling approaches to natural resource management.

Simulating polar bear energetics during a seasonal fast using a mechanistic model

In this study we tested the ability of a mechanistic model (Niche Mapper™) to accurately model adult, non-denning polar bear (Ursus maritimus) energetics while fasting during the ice-free season in the western Hudson Bay. The model uses a steady state heat balance approach, which calculates the metabolic rate that will allow an animal to maintain its core temperature in its particular microclimate conditions. Predicted weight loss for a 120 day fast typical of the 1990s was comparable to empirical studies of the population, and the model was able to reach a heat balance at the target metabolic rate for the entire fast, supporting use of the model to explore the impacts of climate change on polar bears. Niche Mapper predicted that all but the poorest condition bears would survive a 120 day fast under current climate conditions. When the fast extended to 180 days, Niche Mapper predicted mortality of up to 18% for males. Our results illustrate how environmental conditions, variation in animal properties, and thermoregulation processes may impact survival during extended fasts because polar bears were predicted to require additional energetic expenditure for thermoregulation during a 180 day fast. A uniform 3°C temperature increase reduced male mortality during a 180 day fast from 18% to 15%. Niche Mapper explicitly links an animal’s energetics to environmental conditions and thus can be a valuable tool to help inform predictions of climate-related population changes. Since Niche Mapper is a generic model, it can make energetic predictions for other species threatened by climate change.

Consider a non-spherical elephant: computational fluid dynamics simulations of heat transfer coefficients and drag verified using wind tunnel experiments

Animal momentum and heat transfer analysis has historically used direct animal measurements or approximations to calculate drag and heat transfer coefficients. Research can now use modern 3D rendering and computational fluid dynamics software to simulate animal-fluid interactions. Key questions are the level of agreement between simulations and experiments and how superior they are to classical approximations. In this paper we compared experimental and simulated heat transfer and drag calculations on a scale model solid aluminum African elephant casting. We found good agreement between experimental and simulated data and large differences from classical approximations. We used the simulation results to calculate coefficients for heat transfer and drag of the elephant geometry.

Patterns in avian malaria at founder and source populations of an endemic New Zealand passerine

Significant progress in our understanding of disease transmission in the wild can be made by examining variation in host-parasite-vector interactions after founder events of the host. This study is the first to document patterns in avian malaria, Plasmodium spp., infecting an endemic New Zealand passerine, Anthornis melanura, at multiple-host subpopulations simultaneously. We assess the Beaudoin hypothesis of bimodal seasonality and use AIC model selection to determine host factors associated with disease prevalence. We had the rare opportunity to test the enemy release hypothesis (ERH) after a recent colonisation event of the bellbird host. Four Plasmodium species were found to infect bellbirds. Temporal patterns of three exotic parasite lineages, including GRW06 Plasmodium (Huffia) elongatum, SYAT05 Plasmodium (Novyella) vaughani and a Plasmodium (Haemamoeba) relictum, were sporadic with low prevalence year round. The fourth species was an endemic parasite, an unresolved Plasmodium (Novyella) sp. here called ANME01, which exhibited a strong winter peak at the source subpopulations possibly indicating greater immune stressors at the densely populated source site. At the colonies, we observed bimodal seasonality in the prevalence of ANME01 with autumn and spring peaks. These infection peaks were male-biased, and the amplitude of sex bias was more pronounced at the newer colony perhaps due to increased seasonal competition resulting from territory instability. We observed a decrease in parasite species diversity and increase in body condition from source to founder sites, but statistical differences in the direct relationship between body condition and malaria prevalence between source and colony were weak and significant only during winter. Though our data did not strongly support the ERH, we highlight the benefits of 'conspecific release' associated with decreased population density and food competition. Our findings contribute to the identification of ecological and environmental drivers of variability in malaria transmission, which is valuable for predicting the consequences of both natural range expansions, as well as host re-introductions resulting from intensive conservation practices.

Biomechanics meets the ecological niche: the importance of temporal data resolution

The emerging field of mechanistic niche modelling aims to link the functional traits of organisms to their environments to predict survival, reproduction, distribution and abundance. This approach has great potential to increase our understanding of the impacts of environmental change on individuals, populations and communities by providing functional connections between physiological and ecological response to increasingly available spatial environmental data. By their nature, such mechanistic models are more data intensive in comparison with the more widely applied correlative approaches but can potentially provide more spatially and temporally explicit predictions, which are often needed by decision makers. A poorly explored issue in this context is the appropriate level of temporal resolution of input data required for these models, and specifically the error in predictions that can be incurred through the use of temporally averaged data. Here, we review how biomechanical principles from heat-transfer and metabolic theory are currently being used as foundations for mechanistic niche models and consider the consequences of different temporal resolutions of environmental data for modelling the niche of a behaviourally thermoregulating terrestrial lizard. We show that fine-scale temporal resolution (daily) data can be crucial for unbiased inference of climatic impacts on survival, growth and reproduction. This is especially so for species with little capacity for behavioural buffering, because of behavioural or habitat constraints, and for detecting temporal trends. However, coarser-resolution data (long-term monthly averages) can be appropriate for mechanistic studies of climatic constraints on distribution and abundance limits in thermoregulating species at broad spatial scales.

Can mechanism inform species' distribution models?

Two major approaches address the need to predict species distributions in response to environmental changes. Correlative models estimate parameters phenomenologically by relating current distributions to environmental conditions. By contrast, mechanistic models incorporate explicit relationships between environmental conditions and organismal performance, estimated independently of current distributions. Mechanistic approaches include models that translate environmental conditions into biologically relevant metrics (e.g. potential duration of activity), models that capture environmental sensitivities of survivorship and fecundity, and models that use energetics to link environmental conditions and demography. We compared how two correlative and three mechanistic models predicted the ranges of two species: a skipper butterfly (Atalopedes campestris) and a fence lizard (Sceloporus undulatus). Correlative and mechanistic models performed similarly in predicting current distributions, but mechanistic models predicted larger range shifts in response to climate change. Although mechanistic models theoretically should provide more accurate distribution predictions, there is much potential for improving their flexibility and performance.

Avian distributions under climate change: towards improved projections

Birds are responding to recent climate change in a variety of ways including shifting their geographic ranges to cooler climates. There is evidence that northern-temperate birds have shifted their breeding and non-breeding ranges to higher latitudes, and tropical birds have shifted their breeding ranges to higher altitudes. There is further evidence these shifts have affected migration strategies and the composition and structure of communities. Projections based on correlative distributional models suggest many birds will experience substantial pressures under climate change, resulting in range contraction and shifts. Inherent limitations of correlative models, however, make it difficult to develop reliable projections and detailed inference. Incorporating a mechanistic perspective into species distribution models enriches the quality of model inferences but also severely narrows the taxonomic and geographic relevance. Mechanistic distributional models have seen increased applications, but so far primarily in ectotherms. We argue that further development of similar models in birds would complement existing empirical knowledge and theoretical projections. The considerable data already available on birds offer an exciting basis. In particular, information compiled on flight performance and thermal associations across life history stages could be linked to distributional limits and dispersal abilities, which could be used to develop more robust and detailed projections. Yet, only a broadening of taxonomic scale, specifically to appropriately represented tropical diversity, will allow for truly general inference and require the continued use of correlative approaches that may take on increasingly mechanistic components. The trade-off between detail and scale is likely to characterize the future of global change biodiversity research, and birds may be an excellent group to improve, integrate and geographically extend current approaches.

Size, shape, and the thermal niche of endotherms

A key challenge in ecology is to define species' niches on the basis of functional traits. Size and shape are important determinants of a species' niche but their causal role is often difficult to interpret. For endotherms, size and shape define the thermal niche through their interaction with core temperature, insulation, and environmental conditions, determining the thermoneutral zone (TNZ) where energy and water costs are minimized. Laboratory measures of metabolic rate used to describe TNZs cannot be generalized to infer the capacity for terrestrial animals to find their TNZ in complex natural environments. Here, we derive an analytical model of the thermal niche of an ellipsoid furred endotherm that accurately predicts field and laboratory data. We use the model to illustrate the relative importance of size and shape on the location of the TNZ under different environmental conditions. The interaction between body shape and posture strongly influences the location of the TNZ and the expected scaling of metabolic rate with size at constant temperature. We demonstrate that the latter relationship has a slope of approximately (1/2) rather than the commonly expected surface area/volume scaling of (2/3). We show how such functional traits models can be integrated with spatial environmental datasets to calculate null expectations for body size clines from a thermal perspective, aiding mechanistic interpretation of empirical clines such as Bergmann's Rule. The combination of spatially explicit data with biophysical models of heat exchange provides a powerful means for studying the thermal niches of endotherms across climatic gradients.

Thermodynamic modelling predicts energetic bottleneck for seabirds wintering in the northwest Atlantic

Studying the energetics of marine top predators such as seabirds is essential to understand processes underlying adult winter survival and its impact on population dynamics. Winter survival is believed to be the single most important life-history trait in long-lived species but its determinants are largely unknown. Seabirds are inaccessible during this season, so conventional metabolic studies are extremely challenging and new approaches are needed. This paper describes and uses a state-of-the-art mechanistic model, Niche Mapper™, to predict energy expenditure and food requirements of the two main seabird species wintering in the northwest Atlantic. We found that energy demand increased throughout the winter phase in both species. Across this period, mean estimated daily energy requirements were 1306 kJ day–1 for Brünnich's guillemots (Uria lomvia) and 430 kJ day–1 for little auks (Alle alle) wintering off Greenland and Newfoundland. Mean estimated daily food requirements were 547 g wet food day–1 for Brünnich's guillemots, and 289 g wet food day–1 for little auks. For both species and both wintering sites, our model predicts a sharp increase in energy expenditure between November and December, primarily driven by climatic factors such as air temperature and wind speed. These findings strongly suggest the existence of an energetic bottleneck for North Atlantic seabirds towards the end of the year, a challenging energetic phase which might explain recurrent events of winter mass-mortality, so called `seabird winter wrecks'. Our study therefore emphasizes the relevance of thermodynamics/biophysical modelling for investigating the energy balance of wintering marine top predators and its interplay with survival and population dynamics in the context of global change.

Linking traits to energetics and population dynamics to predict lizard ranges in changing environments

I present a dynamic bioenergetic model that couples individual energetics and population dynamics to predict current lizard ranges and those following climate warming. The model predictions are uniquely based on first principles of morphology, life history, and thermal physiology. I apply the model to five populations of a widespread North American lizard, Sceloporus undulatus, to examine how geographic variation in traits and life histories influences ranges. This geographic variation reflects the potential for species to adapt to environmental change. I then consider the range dynamics of the closely related Sceloporus graciosus. Comparing predicted ranges and actual current ranges reveals how dispersal limitations, species interactions, and habitat requirements influence the occupied portions of thermally suitable ranges. The dynamic model predicts individualistic responses to a uniform 3 degrees C warming but a northward shift in the northern range boundary for all populations and species. In contrast to standard correlative climate envelope models, the extent of the predicted northward shift depends on organism traits and life histories. The results highlight the limitations of correlative models and the need for more dynamic models of species' ranges.

The toad ahead: challenges of modelling the range and spread of an invasive species

An ability to predict the rate at which an organism spreads its range is of growing importance because the process of spread (during invasion by an exotic species) is almost identical to that occurring at the expanding range margins of a native species undergoing range shifts in response to climate change. Thus, the methods used for modelling range spread can also be employed to assess the distributional implications of climate change. Here we review the history of research on the spread of cane toads in Australia and use this case study to broadly examine the benefits and pitfalls of various modelling approaches. We show that the problems of estimating the current range, predicting the future range, and predicting the spread rate are interconnected and inform each other. Generally, we argue that correlative approaches to range-prediction are unsuitable when applied to invasive species and suggest that mechanistic methods are beginning to look promising (despite being more difficult to execute), although robust comparisons of correlative versus mechanistic predictions are lacking. Looking to the future, we argue that mechanistic models of range advance (drawing from both population ecology and environmental variation) are the approaches most likely to yield robust predictions. The complexity of these approaches coupled with the steady rise in computing power means that they have only recently become computationally tractable. Thus, we suggest that the field is only recently in a position to incorporate the complexity necessary to robustly model the rate at which species shift their range.

Model of Japanese serow (Capricornis crispus) energetics predicts distribution on Honshu, Japan

Understanding what determines a species' range is a central objective in ecology and evolutionary biology. It has important applications for predicting species distributions and how they might respond to environmental perturbations. This paper describes a mechanistic approach to predict past and present distribution of the Japanese serow (Capricornis crispus) on Honshu, Japan. We applied state-of-the-art microclimate and animal biophysical/behavioral models coupled with climate and vegetation data to estimate the distribution of potential range expansion under protection. We tested the model results against detailed empirical distribution data from the Ministry of the Environment for a five-prefecture area in central Honshu. We also applied the models to time-series land use/cover maps to investigate the historical transitions in habitat suitability during 1947-1999 in the Arai-Keinan region. This is the first time to our knowledge that mechanistic models have successfully predicted the landscape scale distribution of a mammal species in the absence of other animal species interactions, such as predators. In this case, animal energetics/behavior-plant interactions seem to be critical. Forest cover appears to be important in summer and winter for suitable serow habitats. The energetics model results indicate that the serow can overheat in some open environments in midday hours in summer. In winter, simulation results suggested that forest cover provides effective refuge to avoid increased metabolic demands of cold temperatures and strong winds. The model simulations suggested that land use/cover changes documented during 1947-1999 resulted in increased suitable serow habitat due to expanding forest cover from agricultural marginalization and ecological succession. The models provide a unique tool for estimating species' range expansion under protection or for selecting suitable reintroduction sites.