Recipient Organization
SOUTH DAKOTA STATE UNIVERSITY
PO BOX 2275A
BROOKINGS,SD 57007
Performing Department
Natural Resource Management
Non Technical Summary
Bobcats are an economically important furbearer that are distributed across a range of habitats in North America, but tend to prefer forested habitats. Bobcats are sought by trappers and hunters for their valuable pelts, but harvest of bobcats may decrease overall population survival. Thus, it is important for managers to effectively monitor bobcat populations to ensure long-term persistence and sustainable harvest. Effective monitoring of any harvested species requires the delimitation of population boundaries, which can be used to identify biologically relevant and appropriate management units. The composition and configuration of forest habitats may influence how bobcat populations are structured across the landscape. Furthermore, natural landscape features such as rivers may limit bobcat movement, as can features associated with human development such as roads and urban areas.In South Dakota, bobcat harvest was historically permitted statewide, but was restricted to areas west of the Missouri River from 1977 to 2011. Since 2012, bobcat harvest has been allowed east of the Missouri River, but harvest is regulated differently in this region as compared to areas west of the river. West of the Missouri River, the bobcat harvest season is ~8 weeks long, harvest is permitted in all 22 counties, and there is no limit to the number of bobcats that a person can harvest. In contrast, east of the Missouri River the bobcat harvest season is ~4 weeks long, harvest is restricted to 10 of 44 counties, and only 1 bobcat can be harvested per person. Little is known about how bobcat populations are structured in South Dakota; it is unknown how the composition and configuration of forested habitats, rivers, roads, or other landscape features influence bobcat connectivity, dispersal, or population structure. Identifying discrete populations would allow managers to identify areas over which population estimates (e.g., abundance) relate, effectively delineate areas for setting defensible harvest quotas, and better understand the importance of forested habitats in the region to bobcat populations. Identifying discrete populations would also help managers better determine if there is sufficient justification for having disparate harvest restrictions for bobcats on either side of the Missouri River, or if there is justification for more fine-scale harvest management of bobcats across the state.This project will use genetic samples collected by managers from hunter or trapper harvested bobcats in South Dakota. These genetic samples will be used to characterize the genetic diversity among bobcats, identify the number of unique genetic bobcat populations in South Dakota, and to characterize and estimate dispersal patterns for bobcats (i.e., how far do they typically disperse, does one sex disperse further than the other). Additionally, we will combine information from previous research on the movements of radio-collared bobcats with information on the distribution of forested and non-forested habitats, rivers, roads, and other landscape features, to develop models of landscape connectivity for bobcats. We will then combine information on the population genetic structure and landscape connectivity, to formally evaluate the influence of forested habitats and landscape features (e.g., rivers, roads) on bobcat populations.Ultimately, this project will increase our understanding of how forest habitats, natural landscape features, and anthropogenic features influence the population structure and connectivity for the economically and ecologically important bobcat. This knowledge will improve our understanding of how habitat fragmentation influences bobcats and will produce actionable recommendations that can improve management practices for bobcats. In particular, information on population structure can be used to effectively delineate areas for setting defensible harvest quotas, whereas information on connectivity and the identification of areas of high importance for connectivity can inform and guide habitat management.
Animal Health Component
(N/A)
Research Effort Categories
Basic
50%
Applied
(N/A)
Developmental
50%
Goals / Objectives
Problem StatementA primary objective of wildlife management agencies is to establish sustainable harvest levels for game populations. An important component of sustainable management practices is the identification of appropriate management units. Management units are best defined as discrete populations that are independent of other populations (i.e., local population growth rate depends primarily on local vital rates-survival and reproduction-rather than on immigration and emigration; Allendorf and Luikart 2007). Identifying discrete populations allows managers to (1) effectively delineate areas for setting defensible harvest quotas, and (2) identify areas over which population monitoring estimates (e.g., abundance) relate.The bobcat (Lynx rufus) is an economically and ecologically important furbearer in North America (Woolf and Neilsen 2001) that shows strong preference for forested habitats (Reding et al. 2013). Bobcats are commonly considered one of the most valuable furbearers, and they produce the most heavily-traded pelts among felids worldwide (Kapfer and Potts 2012). Due to their similarity in appearance to endangered lynx (Lynx spp.), bobcats are regulated by the Convention on International Trade in Endangered Species (CITES), and agencies responsible for managing bobcats are required to strictly monitor and document their harvest. Bobcat harvest has commonly been perceived as an additive source of mortality (Anderson and Lovallo 2003) and, therefore, managers require detailed information on populations to help minimize the impacts of harvest and ensure sustainable populations and harvest opportunities.In South Dakota, bobcat harvest was historically open statewide, but was restricted to areas west of the Missouri River (hereafter, West River) between 1977 and 2011. Since 2012, bobcat harvest has been permitted in select counties east of the Missouri River (hereafter, East River), but this harvest has been considerably restricted (i.e., shorter season and a bag limit of 1) relative to West River harvest (i.e., longer season and no bag limit). Little is known about how bobcat populations are structured in South Dakota and which (if any) natural or anthropogenic landscape features demarcate bobcat populations. Thus, it is unclear if East River and West River should be managed independent of one another for bobcats, or if management and harvest restrictions for bobcats should be established at finer scales than current management zones (e.g., should West River be managed as >1 bobcat unit).Bobcats are widely distributed across South Dakota, yet little is known about how bobcat populations are structured within the state. At continental-scales, bobcats exhibit complex patterns of population genetic structure influenced by historical and contemporary (i.e., environmental and anthropogenic) factors (Reding et al. 2012). Roads have been identified as significant barriers to bobcats in some regions, with larger roads being greater barriers than smaller roads (Riley et al. 2006). In South Dakota, I-90 may constitute a significant anthropogenic barrier for bobcats. It has been suggested that rivers may also restrict movement of bobcats (Nielsen and Woolf 2003), but this has been refuted by studies documenting movement across rivers (Johnson et al. 2010). In South Dakota, it is unclear if bobcat population structure is influenced by the Missouri River. Although the Missouri River may limit movement during some portions of the year, the barrier effect of the river may be minimized during the winter when it is frozen. Variation in land cover may also influence bobcat population structure. Agricultural lands may restrict movement of bobcats (Reding et al. 2013), and bobcats may be using different ecoregions of South Dakota differently. Mosby et al. (2012) found that habitat and landscape features used by bobcats in the Black Hills and Badlands differed significantly. Although they concluded that these differences reveal the adaptability of the species to a variety of landscapes, these differences among bobcats in different ecoregions may also lead to patterns of isolation by environment (IBE; Wang and Bradburd 2014, Lonsinger et al. 2015).Consequently, land managers play a critical role in managing wildlife populations. Wildlife populations are influenced by land cover and variation in land cover can influence population structure of species (Lonsinger et al. 2015). Bobcats are believed to show a strong preference for forested habitats (Reding et al. 2013), and this preference may influence: 1) how bobcat populations are structured across the landscape, 2) the connectivity of these populations, and 3) appropriate management units and activities.Project Goal & ObjectivesThe goals of this study are to (1) evaluate the population structure of bobcats in South Dakota, (2) assess the influence of forested habitats on bobcat population structure, and (3) provide managers with essential information to effectively manage bobcats and set defensible regional harvest quotas. To this end, our project objectives are to:Utilize hunter and trapper harvested samples to infer population genetic structure for West River and East River (the harvested portion) bobcatsDevelop a predictive map of landscape connectivity for bobcats to identify areas of high importance for maintaining gene flow (i.e., important corridors for connectivity)Elucidate the relative influences of land cover (e.g., forests), natural features (e.g., major rivers), and anthropogenic features (e.g., major highways, human developments) on bobcat population structureEstimate mean bobcat dispersal distances and test for patterns of sex-biased dispersal
Project Methods
The proposed project will combine hunter-harvested genetic samples, laboratory analyses, and ecological modeling. This project will employ landscape genetic approaches. Landscape genetics combines expertise in population genetics, landscape ecology, and spatial analyses to quantify the influence of landscape features on gene flow (Balkenhol et al. 2016). Sampling strategies for inferring population structure, landscape genetic patterns, and dispersal vary by objectives, species ecology, genetic variability, landscape heterogeneity, and scale of study (Balkenhol et al. 2016). Population genetic studies often require sampling of 20-30 individuals per population (Storfer et al. 2007), whereas landscape genetic studies of isolation by distance may require >300 individuals (Anderson et al. 2010). When population genetic structure is unknown, random or systematic sampling may be best to infer structure (Hall and Beissinger 2014). In contrast, clustered sampling provides greater power for investigating dispersal (Banks and Peakall 2012). Due to the disparate needs of landscape genetic studies, harvest samples are often used (e.g., bobcats, Reding et al. 2012). Harvested samples are not random but can approximate random sampling (Schwartz and McKelvey 2008) and offer advantages. Harvest can provide large samples, which would be difficult to obtain with alternative techniques for elusive species. For bobcats, harvest is likely unbiased with respect to sex and age (Knick et al. 1985). Furthermore, harvest samples are commonly collected at two scales: clustered where harvest is expected to be highest (i.e., habitats perceived as optimal) and broadly across the landscape (e.g., in South Dakota, bobcats were harvested from every county with an open season in 2017-2018). These multiple scales make harvest samples appropriate for the variety of analyses employed in landscape genetic studies.Objective 1: Utilize hunter and trapper harvested samples to infer population genetic structure for West River and East River (the harvested area) bobcatsThe proposed project will use samples collected from harvested bobcats in South Dakota. Since the 2013-2014 harvest seasons, >1700 bobcats have been harvested. Tissue samples from >900 bobcats harvested from 2015-2018 have been processed by the genetics laboratory at the University of Idaho (i.e., DNA extraction and sex identification). DNA extracted from these samples has been stored frozen. Bobcat DNA samples will be amplified for 10-15 nuclear DNA (nDNA) microsatellite loci. To reduce genotyping errors, amplification will be performed in duplicate and an additional replicate will be performed to resolve differences in scoring (Taberlet et al. 1996). Resulting multilocus genotypes will be used to quantify genetic diversity (e.g., allelic richness, heterozygosity).Population genetic structure of bobcats will be assessed through (i) Bayesian clustering algorithms in the program STRUCTURE (Pritchard et al. 2001) and (ii) principal component analyses (PCA). STRUCTURE does not require spatial data and uses genotypes to infer the number of genetically distinct clusters and classify individuals into their most likely cluster, while minimizing departures from Hardy-Weinberg and linkage equilibriums (Pritchard et al. 2000). Similarly, PCAs do not require spatial data and use genotypes to separate individuals into clusters along orthogonal axes. We will calculate standard measures of population differentiation (e.g., G'ST). Individual-based pairwise genetic distance matrices based on relatedness will be derived from genotypes.Objective 2: Develop a predictive map of landscape connectivity for bobcats to identify areas of high importance for maintaining gene flow (i.e., important corridors for connectivity)To characterize the influence of landscape and habitat features on permeability, landscape connectivity will be modeled by combining (i) habitat suitability maps previously generated for bobcats in South Dakota (Mosby 2011), and (ii) movement information collected from previously telemetered bobcats in South Dakota (Mosby 2011). Different combination of predictors believed to influence movement (i.e., a priori hypotheses) will be used to generate alternative, or competing, models of landscape permeability. Each competing model will be used to generate resistance surfaces using the program CIRCUITSCAPE, which utilizes electrical circuit theory to quantify levels of connectivity between points (e.g., individuals, population centers; McRae et al. 2006, 2008). Areas estimated as having the most consistent influence on bobcat connectivity will be identified (i.e., areas identified as important for connectivity across competing models).Objective 3: Elucidate the relative influences of land cover (e.g., forests), natural features (e.g., major rivers), and anthropogenic features (e.g., major highways, human developments) on bobcat population structureWe will use available GIS layers for landscape features and habitats to evaluate the potential drivers of observed patterns of population genetic structure. Individual-based pairwise distance matrices will be generated for geographic (i.e. Euclidian) and resistance distances. Resistance distances for competing models will be determined using CIRCUITSCAPE. Partial Mantel tests will be used to test for a correlation between individual-based pairwise genetic distances and geographic or resistance distances (Lonsinger et al. 2015). The influence of isolation by environmental will be evaluated through nonparametric multivariate ANOVA (Lonsinger et al. 2015).Objective 4: Estimate mean bobcat dispersal distances and test for patterns of sex-biased dispersalEstimates of dispersal can be difficult to obtain. Under traditional strategies, such as radio-telemetry, dispersal estimates are often derived from only a small number of collared individuals, and sample sizes tend to be small for elusive wildlife that are difficult to monitor. In contrast, molecular methods of estimating dispersal do not rely on tracking individuals, and therefore can provide more reliable dispersal estimates. Patterns of dispersal will be assessed through molecular methods which exploit spatial autocorrelation in genotypes (Smouse and Peakall 1999) and assume that increased philopatry by one sex will generate greater genetic similarities among neighboring individuals of that sex. Spatial autocorrelation analyses will be used to evaluate dispersal distances and test for sex-biased dispersal.EffortsEfforts to cause a change in knowledge will include formal mentoring and training of students, experiential learning by students, and the dissemination of results to the scientific community. The project will be used as a case study in an undergraduate "Principles of Wildlife Management" course. Efforts used to cause a change in action will include outreach to regional wildlife and land managers, agricultural landowners, and sportsmen's groups to promote practices that help managers meet their objectives.EvaluationFor each annual milestone, the indicator or metric for success is in parentheses. Year 1 milestones include graduate proposal development (detailed plan to address objectives, a priori hypotheses, and approval by a graduate committee), multiplex development (identify nDNA loci, number of loci, and PCR thermal profiles), and sample amplification (student trained on lab methods). Year 2 milestones include connectivity models (estimates of resistance surfaces), genotyping (sample amplification and scoring), and population genetic analyses (estimates of genetic diversity and genetic structure). Year 3 milestones include landscape genetic analyses (identify factors influencing gene flow), dispersal analyses (estimate dispersal by sex and between sexes), and completion of a MS thesis (incorporate final analyses, defended and approved by the committee, and prepared for submission to a journal).