Explaining detection heterogeneity with finite mixture and non-Euclidean movement in spatially explicit capture-recapture models.

Landscape structure affects animal movement. Differences between landscapes may induce heterogeneity in home range size and movement rates among individuals within a population. These types of heterogeneity can cause bias when estimating population size or density and are seldom considered during analyses. Individual heterogeneity, attributable to unknown or unobserved covariates, is often modelled using latent mixture distributions, but these are demanding of data, and abundance estimates are sensitive to the parameters of the mixture distribution. A recent extension of spatially explicit capture-recapture models allows landscape structure to be modelled explicitly by incorporating landscape connectivity using non-Euclidean least-cost paths, improving inference, especially in highly structured (riparian & mountainous) landscapes. Our objective was to investigate whether these novel models could improve inference about black bear (Ursus americanus) density. We fit spatially explicit capture-recapture models with standard and complex structures to black bear data from 51 separate study areas. We found that non-Euclidean models were supported in over half of our study areas. Associated density estimates were higher and less precise than those from simple models and only slightly more precise than those from finite mixture models. Estimates were sensitive to the scale (pixel resolution) at which least-cost paths were calculated, but there was no consistent pattern across covariates or resolutions. Our results indicate that negative bias associated with ignoring heterogeneity is potentially severe. However, the most popular method for dealing with this heterogeneity (finite mixtures) yielded potentially unreliable point estimates of abundance that may not be comparable across surveys, even in data sets with 136-350 total detections, 3-5 detections per individual, 97-283 recaptures, and 80-254 spatial recaptures. In these same study areas with high sample sizes, we expected that landscape features would not severely constrain animal movements and modelling non-Euclidian distance would not consistently improve inference. Our results suggest caution in applying non-Euclidean SCR models when there is no clear landscape covariate that is known to strongly influence the movement of the focal species, and in applying finite mixture models except when abundant data are available.

Harvest and density-dependent predation drive long-term population decline in a northern ungulate.

The relative effect of top-down versus bottom-up forces in regulating and limiting wildlife populations is an important theme in ecology. Untangling these effects is critical for a basic understanding of trophic dynamics and effective management. We examined the drivers of moose (Alces alces) population growth by integrating two independent sources of observations within a hierarchical Bayesian population model. We used one of the largest existing spatiotemporal data sets on ungulate population dynamics globally. We documented a 20% population decline over the period examined. There was negative density-dependent population growth of moose. Although we could not determine the mechanisms producing density-dependent suppression of population growth, the relatively low densities at which we documented moose populations suggested it could be due to density-dependent predation. Predation primarily limited population growth, except at low density, where it was regulating. After we simulated several harvest scenarios, it appeared that harvest was largely additive and likely contributed to population declines. Our results highlight how population dynamics are context dependent and vary strongly across gradients in climate, forest type, and predator abundance. These results help clarify long-standing questions in population ecology and highlight the complex relationships between natural and human-caused mortality in driving ungulate population dynamics.

Seven decades of southern range dynamics of Canada lynx.

The range of the Canada lynx (Lynx canadensis) has contracted substantially from its historical range. Using harvest records, we found that the southern range of the lynx in Ontario in the late 1940s collapsed and then, in a short period of time, increased to its largest extent in the mid-1960s when the lynx range spread south of the boreal forest for a decade. After this expansion, the southern range contracted northwards beginning in the 1970s. Most recently, there has been a slight expansion between 2010 and 2017. We have attributed these dynamics on the southern range periphery to the fluctuation of the boreal lynx population in the core of the species' range. In addition, connectivity to boreal lynx populations and snow depth seemed to condition whether the lynx expanded into an area. However, we did not find any evidence to suggest that these changes were due to anthropogenic landscape disturbances or competition. The boreal lynx population does not reach the peak abundance it once did, without which we would not expect to see large expansions of the southern lynx range as in the mid-1960s. Our results suggest that the southern lynx range in Ontario has been driven by the magnitude of the boreal lynx population cycle, connectivity to the boreal forest, and snow conditions. Future persistence of lynx in the southern range periphery will likely depend on dynamics in the range core.

Maternal glucocorticoid levels during incubation predict breeding success, but not reproductive investment, in a free-ranging bird.

The hormone corticosterone (CORT) has been hypothesized to be linked with fitness, but the directionality of the relationship is unclear. The 'CORT-fitness hypothesis' proposes that high levels of CORT arise from challenging environmental conditions, resulting in lower reproductive success (a negative relationship). In contrast, the CORT-adaptation hypothesis suggests that, during energetically demanding periods, CORT will mediate physiological or behavioral changes that result in increased reproductive investment and success (a positive relationship). During two breeding seasons, we experimentally manipulated circulating CORT levels in female tree swallows (Tachycineta bicolor) prior to egg laying, and measured subsequent reproductive effort, breeding success, and maternal survival. When females were recaptured during egg incubation and again during the nestling stage, the CORT levels were similar among individuals in each treatment group, and maternal treatment had no effect on indices of fitness. By considering variation among females, we found support for the CORT-adaptation hypothesis; there was a significant positive relationship between CORT levels during incubation and hatching and fledging success. During the nestling stage CORT levels were unrelated to any measure of investment or success. Within the environmental context of our study, relationships between maternal glucocorticoid levels and indices of fitness vary across reproductive stages.

Fine-scale habitat selection by sympatric Canada lynx and bobcat.

The Canada lynx (Lynx canadensis) and the bobcat (Lynx rufus) are closely related species with overlap at their range peripheries, but the factors that limit each species and the interactions between them are not well understood. Habitat selection is a hierarchical process, in which selection at higher orders (geographic range, home range) may constrain selection at lower orders (within the home range). Habitat selection at a very fine scale within the home range has been less studied for both lynx and bobcat compared to selection at broader spatiotemporal scales. To compare this fourth-order habitat selection by the two species in an area of sympatry, we tracked lynx and bobcat during the winters of 2017 and 2018 on the north shore of Lake Huron, Ontario. We found that both lynx and bobcat selected shallower snow, higher snowshoe hare abundance, and higher amounts of coniferous forest at the fourth order. However, the two species were spatially segregated at the second order, and lynx were found in areas with deeper snow, more snowshoe hare, and more coniferous forest. Taken together, our findings demonstrate that the lynx and bobcat select different resources at the second order, assorting along an environmental gradient in the study area, and that competition is unlikely to be occurring between the two species at finer scales.

Climate connectivity of the bobcat in the Great Lakes region.

The Great Lakes and the St. Lawrence River are imposing barriers for wildlife, and the additive effect of urban and agricultural development that dominates the lower Great Lakes region likely further reduces functional connectivity for many terrestrial species. As the climate warms, species will need to track climate across these barriers. It is important therefore to investigate land cover and bioclimatic hypotheses that may explain the northward expansion of species through the Great Lakes. We investigated the functional connectivity of a vagile generalist, the bobcat, as a representative generalist forest species common to the region. We genotyped tissue samples collected across the region at 14 microsatellite loci and compared different landscape hypotheses that might explain the observed gene flow or functional connectivity. We found that the Great Lakes and the additive influence of forest stands with either low or high canopy cover and deep lake-effect snow have disrupted gene flow, whereas intermediate forest cover has facilitated gene flow. Functional connectivity in southern Ontario is relatively low and was limited in part by the low amount of forest cover. Pathways across the Great Lakes were through the Niagara region and through the Lower Peninsula of Michigan over the Straits of Mackinac and the St. Marys River. These pathways are important routes for bobcat range expansion north of the Great Lakes and are also likely pathways that many other mobile habitat generalists must navigate to track the changing climate. The extent to which species can navigate these routes will be important for determining the future biodiversity of areas north of the Great Lakes.

Multi-species genetic connectivity in a terrestrial habitat network.

BACKGROUND: Habitat fragmentation reduces genetic connectivity for multiple species, yet conservation efforts tend to rely heavily on single-species connectivity estimates to inform land-use planning. Such conservation activities may benefit from multi-species connectivity estimates, which provide a simple and practical means to mitigate the effects of habitat fragmentation for a larger number of species. To test the validity of a multi-species connectivity model, we used neutral microsatellite genetic datasets of Canada lynx (Lynx canadensis), American marten (Martes americana), fisher (Pekania pennanti), and southern flying squirrel (Glaucomys volans) to evaluate multi-species genetic connectivity across Ontario, Canada. RESULTS: We used linear models to compare node-based estimates of genetic connectivity for each species to point-based estimates of landscape connectivity (current density) derived from circuit theory. To our knowledge, we are the first to evaluate current density as a measure of genetic connectivity. Our results depended on landscape context: habitat amount was more important than current density in explaining multi-species genetic connectivity in the northern part of our study area, where habitat was abundant and fragmentation was low. In the south however, where fragmentation was prevalent, genetic connectivity was correlated with current density. Contrary to our expectations however, locations with a high probability of movement as reflected by high current density were negatively associated with gene flow. Subsequent analyses of circuit theory outputs showed that high current density was also associated with high effective resistance, underscoring that the presence of pinch points is not necessarily indicative of gene flow. CONCLUSIONS: Overall, our study appears to provide support for the hypothesis that landscape pattern is important when habitat amount is low. We also conclude that while current density is proportional to the probability of movement per unit area, this does not imply increased gene flow, since high current density tends to be a result of neighbouring pixels with high cost of movement (e.g., low habitat amount). In other words, pinch points with high current density appear to constrict gene flow.

The relationship between least-cost and resistance distance.

Least-cost modelling and circuit theory are common analogs used in ecology and evolution to model gene flow or animal movement across landscapes. Least-cost modelling estimates the least-cost distance, whereas circuit theory estimates resistance distance. The bias added in choosing one method over the other has not been well documented. We designed an experiment to test whether both methods were linearly related. We also tested the sensitivity of these metrics to variation in Euclidean distance, spatial autocorrelation, the number of pixels representing the landscape, and data aggregation. We found that least-cost and resistance distance were not linearly related unless a transformation was applied. Resistance distance was less sensitive to the number of pixels representing a landscape and was also less sensitive than least-cost distance to the Euclidean distance between nodes. Spatial autocorrelation did not affect either method or the relationship between methods. Resistance distance was more sensitive to aggregation in any form compared to least-cost distance. Therefore, the metric used to infer movement or gene flow and the manipulations applied to the data used to calculate these metrics may govern findings.

Climate change and habitat fragmentation drive the occurrence of Borrelia burgdorferi, the agent of Lyme disease, at the northeastern limit of its distribution.

Lyme borreliosis is rapidly emerging in Canada, and climate change is likely a key driver of the northern spread of the disease in North America. We used field and modeling approaches to predict the risk of occurrence of Borrelia burgdorferi, the bacteria causing Lyme disease in North America. We combined climatic and landscape variables to model the current and future (2050) potential distribution of the black-legged tick and the white-footed mouse at the northeastern range limit of Lyme disease and estimated a risk index for B. burgdorferi from these distributions. The risk index was mostly constrained by the distribution of the white-footed mouse, driven by winter climatic conditions. The next factor contributing to the risk index was the distribution of the black-legged tick, estimated from the temperature. Landscape variables such as forest habitat and connectivity contributed little to the risk index. We predict a further northern expansion of B. burgdorferi of approximately 250-500 km by 2050 - a rate of 3.5-11 km per year - and identify areas of rapid rise in the risk of occurrence of B. burgdorferi. Our results will improve understanding of the spread of Lyme disease and inform management strategies at the most northern limit of its distribution.