The energetic gains from foraging and costs of movement are expected to be key drivers of animal decision‐making, as their balance is a large determinant of body condition and survival. This fundamental perspective is often missing from habitat selection studies, which mainly describe correlations between space use and environmental features, rather than the mechanisms behind these correlations. To address this gap, we present a novel parameterisation of step selection functions (SSFs), that we term the energy selection function (ESF). In this model, the likelihood of an animal selecting a movement step depends directly on the corresponding energetic gains and costs, and we can therefore assess how moving animals choose habitat based on energetic considerations. The ESF retains the mathematical convenience and practicality of other SSFs and can be quickly fitted using standard software. In this article, we outline a workflow, from data gathering to statistical analysis, and use a case study of polar bears Ursus maritimus to demonstrate application of the model. We explain how defining gains and costs at the scale of the movement step allows us to include information about resource distribution, landscape resistance and movement patterns. We further demonstrate this process with a case study of polar bears and show how the parameters can be interpreted in terms of selection for energetic gains and against energetic costs. The ESF is a flexible framework that combines the energetic consequences of both movement and resource selection, thus incorporating a key mechanism into habitat selection analysis. Further, because it is based on familiar habitat selection models, the ESF is widely applicable to any study system where energetic gains and costs can be derived, and has immense potential for methodological extensions.

There is strong incentive to model behaviour-dependent habitat selection, as this can help delineate critical habitats for important life processes and reduce bias in model parameters. For this purpose, a two-stage modelling approach is often taken: (i) classify behaviours with a hidden Markov model (HMM), and (ii) fit a step selection function (SSF) to each subset of data. However, this approach does not properly account for the uncertainty in behavioural classification, nor does it allow states to depend on habitat selection. An alternative approach is to estimate both state switching and habitat selection in a single, integrated model called an HMM-SSF. We build on this recent methodological work to make the HMM-SSF approach more efficient and general. We focus on writing the model as an HMM where the observation process is defined by an SSF, such that well-known inferential methods for HMMs can be used directly for parameter estimation and state classification. We extend the model to include covariates on the HMM transition probabilities, allowing for inferences into the temporal and individual-specific drivers of state switching. We demonstrate the method through an illustrative example of African zebra (Equus quagga), including state estimation, and simulations to estimate a utilisation distribution. In the zebra analysis, we identified two behavioural states, with clearly distinct patterns of movement and habitat selection ("encamped'' and "exploratory''). In particular, although the zebra tended to prefer areas higher in grassland across both behavioural states, this selection was much stronger in the fast, directed exploratory state. We also found a clear diel cycle in behaviour, which indicated that zebras were more likely to be exploring in the morning and encamped in the evening. This method can be used to analyse behaviour-specific habitat selection in a wide range of species and systems. A large suite of statistical extensions and tools developed for HMMs and SSFs can be applied directly to this integrated model, making it a very versatile framework to jointly learn about animal behaviour, habitat selection, and space use.

Sea ice habitats are highly dynamic, and ice drift may affect the energy expenditure of travelling animals. Several studies in the high Arctic have reported increased ice drift speeds, and consequently, polar bears Ursus maritimus in these areas expended more energy on counter-ice movement for station-keeping. However, little is known about the spatiotemporal dynamics of ice drift in Hudson Bay (HB) and its implications for the declining Western Hudson Bay (WH) polar bear subpopulation. Using sea ice drift data from 1987-2015 and polar bear satellite telemetry location data from 2004-2015, we examined trends in drift speeds in HB, polar bear movement relative to drift, and assessed annual and individual variation. In contrast to other areas of the Arctic, we did not find an increase in ice drift speed over the period examined. However, variability in ice drift speed increased over time, which suggests reduced habitat predictability. Polar bear movement direction was not strongly counter to ice drift in any month, and ice drift speed and direction had little effect on bear movement rates and, thus, energy expenditure. On an annual scale, we found individuals varied in their exposure and response to ice drift, which may contribute to variability in body condition. However, the lack of a long-term increase in ice drift speed suggests this is unlikely to be the main factor affecting the body condition decline observed in the WH subpopulation. Our results contrast findings in other subpopulations and demonstrate the need for subpopulation-specific research and risk evaluation.

Background There is strong incentive to model behaviour-dependent habitat selection, as this can help delineate critical habitats for important life processes and reduce bias in model parameters. For this purpose, a two-stage modelling approach is often taken: (i) classify behaviours with a hidden Markov model (HMM), and (ii) fit a step selection function (SSF) to each subset of data. However, this approach does not properly account for the uncertainty in behavioural classification, nor does it allow states to depend on habitat selection. An alternative approach is to estimate both state switching and habitat selection in a single, integrated model called an HMM-SSF. Methods We build on this recent methodological work to make the HMM-SSF approach more efficient and general. We focus on writing the model as an HMM where the observation process is defined by an SSF, such that well-known inferential methods for HMMs can be used directly for parameter estimation and state classification. We extend the model to include covariates on the HMM transition probabilities, allowing for inferences into the temporal and individual-specific drivers of state switching. We demonstrate the method through an illustrative example of plains zebra (Equus quagga), including state estimation, and simulations to estimate a utilisation distribution. Results In the zebra analysis, we identified two behavioural states, with clearly distinct patterns of movement and habitat selection (“encamped” and “exploratory”). In particular, although the zebra tended to prefer areas higher in grassland across both behavioural states, this selection was much stronger in the fast, directed exploratory state. We also found a clear diel cycle in behaviour, which indicated that zebras were more likely to be exploring in the morning and encamped in the evening. Conclusions This method can be used to analyse behaviour-specific habitat selection in a wide range of species and systems. A large suite of statistical extensions and tools developed for HMMs and SSFs can be applied directly to this integrated model, making it a very versatile framework to jointly learn about animal behaviour, habitat selection, and space use.