For gregarious animals the cost-benefit trade-offs that drive habitat selection may vary dynamically with group size, which plays an important role in foraging and predator avoidance strategies. We examined how habitat selection by bison (Bison bison) varied as a function of group size and interpreted these patterns by testing whether habitat selection was more strongly driven by the competing demands of forage intake vs. predator avoidance behavior. We developed an analytical framework that integrated group size into resource selection functions (RSFs). These group-size-dependent RSFs were based on a matched case-control design and were estimated using conditional logistic regression (mixed and population-averaged models). Fitting RSF models to bison revealed that bison groups responded to multiple aspects of landscape heterogeneity and that selection varied seasonally and as a function of group size. For example, roads were selected in summer, but not in winter. Bison groups avoided areas of high snow water equivalent in winter. They selected areas composed of a large proportion of meadow area within a 700-m radius, and within those areas, bison selected meadows. Importantly, the strength of selection for meadows varied as a function of group size, with stronger selection being observed in larger groups. Hence the bison-habitat relationship depended in part on the dynamics of group formation and division. Group formation was most likely in meadows. In contrast, risk of group fission increased when bison moved into the forest and was higher during the time of day when movements are generally longer and more variable among individuals. We also found that stronger selection for meadows by large rather than small bison groups was caused by longer residence time in individual meadows by larger groups and that departure from meadows appears unlikely to result from a depression in food intake rate. These group-size-dependent patterns were consistent with the hypothesis that avoidance of predation risk is the strongest driver of habitat selection.

No abstract

Our previous studies have advanced the idea that the folliculostellate cell GJA1 (gap junction membrane channel protein alpha1; previously known as connexin 43)-mediated gap junctions contribute to the establishment of an intercellular network that regulates the paracrine messages and the endocrine response within the anterior pituitary. The folliculostellate cells are targets for growth factors and cytokines that modulate hormone secretion. Proinflammatory cytokines modulate the cell-to-cell communication in many tissues of the body. The present study measured the effect of the proinflammatory cytokines tumor necrosis factor and interleukin-1 on the GJA1-mediated intercellular communication, specifically the expression, localization, degradation, and phosphorylation status of GJA1 in the folliculostellate cell line TtT/GF. The GJA1 localized to the plasma membrane and to minute cytoplasmic vesicles in the perinuclear area. Using different antibodies that recognize distinctly the nonphosphorylated from the phosphorylated forms of GJA1, we showed that nonphosphorylated GJA1 in Ser-368 (NP-GJA1) localized chiefly in the cytoplasm, whereas GJA1 phosphorylated in Ser-368 (P-GJA1) localized to the plasma membrane in controls. The cytokine treatment transiently increased 1) GJA1, NP-GJA1, and P-GJA1 levels; 2) NP-GJA1 and P-GJA1 degradation by both the lysosomal and proteasomal pathways; and 3) cell-to-cell communication in TtT/GF cells. The results suggest that the cytokine-evoked, transient enhancement of folliculostellate cell-mediated intercellular communication contributes to the coordination of the response among folliculostellate cells.

Developing tools that help predict animal distribution in the face of environmental change is central to understanding ecosystem function, but it remains a significant ecological challenge. We tested whether a single foraging currency could explain bison (Bison bison) distribution in dissimilar environments: a largely forested environment in Prince Albert National Park (Saskatchewan, Canada) and a prairie environment in Grasslands National Park (Saskatchewan, Canada). We blended extensive behavioral observations, relocations of radio-collared bison, vegetation surveys, and laboratory analyses to spatially link bison distribution in the two parks and expected gains for different nutritional currencies. In Prince Albert National Park, bison were more closely associated with the distribution of plants that maximized their instantaneous energy intake rate (IDE) than their daily intake of digestible energy. This result reflected both bison's intensity of use of individual meadows and their selection of foraging sites within meadows. On this basis, we tested whether IDE could explain the spatial dynamics of bison reintroduced to Grasslands National Park. As predicted, bison distribution in this park best matched spatial patterns of plants offering rapid IDE rather than rapid sodium intake, phosphorus intake, or daily intake of digestible energy. Because the two study areas have very different plant communities, a phenomenological model of resource selection developed in one area could not be used to predict animal distribution in the other. We were able, however, to successfully infer the distribution of bison from their foraging objective. This consistency in foraging currency across ecosystems and populations provides a strong basis for forecasting animal distributions in novel and dynamic environments.

No abstract