Laura Finnegan

Integrating multiple analytical approaches to spatially delineate and characterize genetic population structure: an application to boreal caribou ( Rangifer tarandus caribou ) in central Canada

Individual-based clustering (IBC) methods have become increasingly popular for the characterization and delineation of genetic population units for numerous species. These methods delineate populations based on the genetic assumptions of a breeding unit which may provide a better representation of the behaviour of the species. The increasing use of IBC has resulted in the development of several analytical models all of which vary in their theoretical assumptions to infer genetic population structure. In this paper, we report a comparative strategy utilizing three IBC methods to characterize the spatial genetic structure of the boreal population of woodland caribou (Rangifer tarandus caribou) in central Canada. In addition, we implement both tests for isolation-by-distance (IBD) and frequency-based assignment tests to validate the consensus genetic clusters as defined by IBC. We also compare indirect metrics of genetic diversity and gene flow using both a priori defined herds and the IBC defined populations. Although our results show some concordance between both pre-defined herds and IBC derived genetic clusters, the IBC analyses identified a cluster that was cryptic to observation-based caribou herds and found no difference between several adjacent herds. By comparing multiple IBC methods and integrating both IBD and indirect genetic diversity metrics a posteriori, our strategy provides an effective means to delineate wildlife population structure and accurately assess genetic diversity and connectivity.

The complimentary role of genetic and ecological data in understanding population structure: a case study using moose (Alces alces)

Landscape features can influence animal movements and gene flow, so rigorous analysis of population structure should include both levels of analysis. We conducted individual-based landscape genetic analysis, using data from eight microsatellite loci, on adult female moose (Alces alces; n = 92) from two areas in southern Ontario, and also monitored movements of a subsample of animals (n = 38) to gauge complementarity of the data sources (genetic and telemetry data) in understanding patterns of population structure. Genetic data indicated that female moose constituted a single panmictic population encompassing both areas, with no evidence of isolation by Euclidean (Mantel test p > 0.38) or least-cost (Mantel test p > 0.19) distances, even when such distances were calculated from costs extracted from resource selection functions of the radio-collared individuals. In contrast, throughout the 2.5-year study radio-telemetry data failed to reveal large-scale movement of female moose. In fact, although moose tended to occupy home ranges in a wide range of sizes, their locations were largely fixed at a larger spatial scale encompassing both study areas. This finding implies that radio-telemetry data alone would not have revealed the extent of connectivity between moose populations, whereas least-cost genetic measures in the absence of telemetry data would not have shown the localised movements of adult females. Our contrasting results highlight the importance of multiple sources of information when assessing population connectivity of large mammals.

Toward the Restoration of Caribou Habitat: Understanding Factors Associated with Human Motorized Use of Legacy Seismic Lines

Populations of boreal and southern mountain caribou in Alberta, Canada, are declining, and the ultimate cause of their decline is believed to be anthropogenic disturbance. Linear features are pervasive across the landscape, and of particular importance, seismic lines established in the 1900s (legacy seismic lines) are slow to regenerate. Off-highway vehicles are widely used on these seismic lines and can hamper vegetative re-growth because of ongoing physical damage, compaction, and active clearing. Restoration of seismic lines within caribou range is therefore a priority for the recovery of threatened populations in Alberta, but a triage-type approach is necessary to prioritize restoration and ensure conservation resources are wisely spent. To target restoration efforts, our objective was to determine factors that best explained levels of off-highway vehicles use on seismic lines intersecting roads. We investigated the relative importance of local topography, vegetation attributes of seismic lines, and broad-scale human factors such as the density of infrastructures and the proximity to recreation campsites and towns to explain the observed levels of off-highway vehicles use. We found that off-highway vehicles use was mainly associated with local topography and vegetation attributes of seismic lines that facilitated ease-of-travel. Broad-scale landscape attributes associated with industrial, recreation access, or hunting activities did not explain levels of off-highway vehicles use. Management actions aimed at promoting natural regeneration and reduce ease-of-travel on legacy seismic lines within caribou ranges can be beneficial to caribou recovery in Alberta, Canada, and we therefore recommend restrictions of off-highway vehicles use on low vegetation, dry seismic lines in caribou ranges.

Fine-Scale Analysis Reveals Cryptic Patterns of Genetic Structure in Canada Geese

Abstract. In migratory birds, population-genetic structure is generally low, but philopatric species can have fine-scale patterns of differentiation. We investigated the population-genetic structure of the Canada Goose (Branta canadensis) across multiple spatial scales to determine whether genetic data support current delineations of subspecies and populations for management. We collected samples from two subspecies (B. c. interior and B. c.maxima) and four management populations across Ontario and Quebec. Using 7 microsatellites and 442 base pairs of mtDNA we found that genetic structure varied with scale. FST revealed low levels of genetic differentiation between subspecies and management populations, and individual-based clustering revealed no genetic differentiation. However, fine-scale spatial autocorrelation revealed significant levels of relatedness at distances <85 km. The lack of clear genetic structure may reflect recent human management. That our fine-scale analysis revealed significant genetic relationships suggests that genetic structure may increase and in time reflect that revealed by banding data. As our markers were unable to accurately distinguish between subspecies they will be of little use in estimating subspecific contribution to harvested stock. Alternative molecular markers under selective pressure may be more informative in assess targets for harvest.

Natural regeneration on seismic lines influences movement behaviour of wolves and grizzly bears

Across the boreal forest of Canada, habitat disturbance is the ultimate cause of caribou (Rangifer tarandus caribou) declines. Habitat restoration is a focus of caribou recovery efforts, with a goal to finding ways to reduce predator use of disturbances, and caribou-predator encounters. One of the most pervasive disturbances within caribou ranges in Alberta, Canada are seismic lines cleared for energy exploration. Seismic lines facilitate predator movement, and although vegetation on some seismic lines is regenerating, it remains unknown whether vegetation regrowth is sufficient to alter predator response. We used Light Detection and Ranging (LiDAR) data, and GPS locations, to understand how vegetation and other attributes of seismic lines influence movements of two predators, wolves (Canis lupus) and grizzly bears (Ursus arctos). During winter, wolves moved towards seismic lines regardless of vegetation height, while during spring wolves moved towards seismic lines with higher vegetation. During summer, wolves moved towards seismic lines with lower vegetation and also moved faster near seismic lines with vegetation <0.7 m. Seismic lines with lower vegetation height were preferred by grizzly bears during spring and summer, but there was no relationship between vegetation height and grizzly bear movement rates. These results suggest that wolves use seismic lines for travel during summer, but during winter wolf movements relative to seismic lines could be influenced by factors additional to movement efficiency; potentially enhanced access to areas frequented by ungulate prey. Grizzly bears may be using seismic lines for movement, but could also be using seismic lines as a source of vegetative food or ungulate prey. To reduce wolf movement rate, restoration could focus on seismic lines with vegetation <1 m in height. However our results revealed that seismic lines continue to influence wolf movement behaviour decades after they were built, and even at later stages of regeneration. Therefore it remains unknown at what stage of natural regeneration, if any, wolves cease to respond to seismic lines. To reduce wolf response to seismic lines, active restoration tactics like blocking seismic lines and tree planting, along with management of alternate prey, could be evaluated.

Motorized Activity on Legacy Seismic Lines: A Predictive Modeling Approach to Prioritize Restoration Efforts

Natural regeneration of seismic lines, cleared for hydrocarbon exploration, is slow and often hindered by vegetation damage, soil compaction, and motorized human activity. There is an extensive network of seismic lines in western Canada which is known to impact forest ecosystems, and seismic lines have been linked to declines in woodland caribou (Rangifer tarandus caribou). Seismic line restoration is costly, but necessary for caribou conservation to reduce cumulative disturbance. Understanding where motorized activity may be impeding regeneration of seismic lines will aid in prioritizing restoration. Our study area in west-central Alberta, encompassed five caribou ranges where restoration is required under federal species at risk recovery strategies, hence prioritizing seismic lines for restoration is of immediate conservation value. To understand patterns of motorized activity on seismic lines, we evaluated five a priori hypotheses using a predictive modeling framework and Geographic Information System variables across three landscapes in the foothills and northern boreal regions of Alberta. In the northern boreal landscape, motorized activity was most common in dry areas with a large industrial footprint. In highly disturbed areas of the foothills, motorized activity on seismic lines increased with low vegetation heights, relatively dry soils, and further from forest cutblocks, while in less disturbed areas of the foothills, motorized activity on seismic lines decreased proportional to seismic line density, slope steepness, and white-tailed deer abundance, and increased proportional with distance to roads. We generated predictive maps of high motorized activity, identifying 21,777 km of seismic lines where active restoration could expedite forest regeneration.