Derek A. Arnold

Connectivity is determined by the degree to which inhospitable habitat matrices are permeable to gene flow. Successful gene flow requires that dispersing individuals travel to new populations, survive there, and finally, reproduce. Consequently, examinations of connectivity must consider both the permeability of the landscape and individual survival patterns, as surviving individuals are the only ones that have the potential to reproduce. We investigated functional connectivity of Canada lynx (Lynx canadensis) populations in the northwestern boreal forest using an unprecedented dataset of 207,957 relocations from 142 individual lynx. As a specialist predator of snowshoe hare (Lepus americanus), lynx populations follow a 10‐ to 12‐year population cycle, which lags a few years behind the population cycles of hares. We used integrated step selection analyses to quantify habitat selection, and Bayesian counting process implementation of wrapped Cauchy survival models to evaluate mortality risk during dispersal. We compared movement patterns and survival across behavioral states and population phases, following the a population decline. Our results showed few complete physical barriers to dispersal in the study region, with many lynx dispersing greater than 3000 km. Step‐selection analysis of lynx movement revealed a higher tolerance for poor quality habitats in dispersing lynx as compared to resident lynx, primarily in selection for rugged terrain and closer proximity to roads. These differences resulted in a higher estimate of landscape permeability for dispersing lynx compared to residents. When survival patterns were considered, there were increasingly stark differences between resident and dispersing lynx during population declines. Dispersing lynx faced markedly lower survival rates than resident lynx. Overall, this paints a picture of a more connected landscape during population highs, which becomes increasingly fragmented during population lows. Canada lynx populations appeared to have few impermeable barriers across the landscape, but they face periods of low functional connectivity driven by lower survival during population declines. This emphasizes the significance of maintaining high‐quality refugia habitats from which population rebounds can occur following cyclic lows.

Cyclical population dynamics are a common phenomenon in populations worldwide, yet the spatial organization of these cycles remains poorly understood. In this study, we investigated the spatial form and timing of a population collapse from 2018 to 2022 in Canada lynx (Lynx canadensis) across the northwest boreal forest. We analyzed survival, reproduction, and dispersal data from 143 individual global positioning system (GPS) collared lynx from populations across five study sites spanning interior Alaska to determine whether lynx displayed characteristics of a population wave following a concurrent wave in snowshoe hare (Lepus americanus) abundance. Reproductive rates declined across the study sites; however, site-level reproduction declined first in our easternmost study sites, supporting the idea of a population wave. Despite a clear increase in percent of dispersing lynx, there was no evidence of directional bias in dispersal following a hare population wave. Analysis did show increasingly poor survival for lynx dispersing to the east compared to combined resident and westward dispersal. This pattern is consistent with a survival-mediated population wave in lynx as the driver of the theorized population wave. The combination of these factors supports the idea of a hierarchical response to snowshoe hare population declines with a drop in lynx reproduction followed by increased dispersal, and finally reduced survival. All of this evidence is consistent with the expected characteristics of a population undergoing a traveling wave and supports the hypothesis that lynx presence may facilitate and mirror the underlying wave patterns in snowshoe hare.

Wildfire is the most important disturbance regime in North American boreal forest communities, driving forest composition and structure across this region. Recent climate models predict that increasing fire intensity and frequency will result in a shift from a largely coniferous forest to one with a greater dominance by deciduous species. We investigated how an iconic predator of the boreal system, the Canada lynx (Lynx canadensis), moves through a range of burn scars (2-78 years old) and if these movements reflect habitat selection of their main prey, the snowshoe hare (Lepus americanus). Using GPS collars at 4-h fix rates, we fitted integrated step selection models to lynx movements across an 80-year post-fire chronosequence to assess habitat selection in both deciduous and coniferous forests. We predicted that lynx would primarily select intermediately-aged spruce, but young deciduous stands, mirroring the habitat selection of snowshoe hares. We found, however, that lynx habitat selection peaked at intermediately aged stands in both forest types. We hypothesize that this difference is driven by selection for habitats that maximize capture probability as opposed to simply representing the highest hare density. This idea is supported by observed seasonal changes in peak selection in deciduous stands, with selection for younger stands in winter, when leaf loss reduces cover. There was no seasonal change in coniferous stands as they experience little change in cover across seasons. Together these results show that lynx can be resilient to short-term shifts toward intermediate-aged stands. However, these benefits will likely diminish in the longer term as the decrease in fire return interval may reduce the prevalence of intermediate-age stands.