Variation in habitat selection among individuals differs by maternal status for moose in a region with low calf survival

Foraging behaviors often involve trade-offs between predation risk and access to forage. Risk-forage trade-offs may be particularly acute for maternal female ungulates, whose nutritional needs are high and whose calves are highly vulnerable to predation. In moose, the selection of calving habitat is one way in which females can respond to these trade-offs. Our objective was to compare among-individual variation in selected habitat for maternal and nonmaternal female moose during the calving season. We hypothesized that, compared to nonmaternal females, maternal females would exhibit a greater range of variation among individuals, which may signal differential responses to risk-forage trade-offs. Meanwhile, we expected nonmaternal females to show comparatively less variation among individuals, consistent with a group primarily maximizing forage intake. To test our hypotheses, we used a path selection framework and a set of continuous remotely sensed map covariates to build predictive models and corresponding spatial predictions for maternal and nonmaternal groups. We then calculated the range of variation among individuals within each group along a relative unitless axis, which we call the “maternal difference index” and define as the divergence of predicted maternal habitat selection from nonmaternal habitat selection. We included 10,080 GPS collar locations for 24 female moose over three consecutive years. Our predictive models had high levels of accuracy (>75%) based both on independent test partitions of a nested cross-validation and on independent very high frequency (VHF) location data, each including spatial and temporal replication. Both groups of females preferred areas where primary forage species were abundant, diverse, and within foraging height. Habitat selected by the maternal group both overlapped and was broader than habitat selected by the nonmaternal group. Based on the maternal difference index, maternal individuals were less consistent in their habitat selection than nonmaternal individuals. Given that habitat selection behaviors are one way in which animals respond to potential risk-forage trade-offs and that maternal individuals in our study differed along a continuum in their selection for primary forage species, we suggest that the concept of maternal trade-offs, as it relates to habitat selection, is most useful when seen as individually determined and variable, rather than group-determined and discrete.