Source: UNIVERSITY OF VERMONT submitted to
ESTIMATING THE EFFECTS OF FOREST FRAGMENT SIZE, ISOLATION, AND SURROUNDING HUMAN DEVELOPMENT ON CARNIVORES IN THE CHAMPLAIN VALLEY, VERMONT
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
TERMINATED
Funding Source
MCINTIRE-STENNIS
Reporting Frequency
Annual
Accession No.
0224345
Grant No.
(N/A)
Project No.
VTZ00123
Proposal No.
(N/A)
Multistate No.
(N/A)
Program Code
(N/A)
Project Start Date
Oct 1, 2010
Project End Date
Sep 30, 2012
Grant Year
(N/A)
Project Director
Murdoch, J.
Recipient Organization
UNIVERSITY OF VERMONT
(N/A)
BURLINGTON,VT 05405
Performing Department
School Of Natural Resources
Non Technical Summary
Continuous forest once covered much of the Champlain Valley of Vermont, but has been fragmented into several small patches due to the expansion of agriculture and urban development. The fragmentation of forests represents a potential conservation concern for carnivores, which serve important ecological roles in forest ecosystems and are often sensitive to habitat disturbance. This project aims to understand the effects of forest fragmentation on carnivores in the Champlain Valley. The project will examine the effects of forest fragment size, degree of isolation, and surrounding human developments, including roads, housing, and agriculture, on the likelihood of carnivore occupancy. The project will survey carnivore occupancy at 60 sites in two seasons using camera traps and track-plate stations and estimate forest fragment characteristics and human development using aerial photos and state and national land-cover data. Multiple-season occupancy models will be used to examine the effects of these characteristics on carnivore occupancy. The results will provide a baseline measure of forest fragment and human factors that influence the carnivore community. They will provide a foundation for developing management strategies for state Species of Greatest Conservation Need, such as the bobcat (Lynx rufus), gray fox (Urocyon cinereoargenteus), and long-tailed weasel (Mustela frenata), and informing forest management decisions in the region.
Animal Health Component
0%
Research Effort Categories
Basic
(N/A)
Applied
(N/A)
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1350830106020%
1350840106015%
1350850106015%
1360830106020%
1360840106015%
1360850106015%
Knowledge Area
135 - Aquatic and Terrestrial Wildlife; 136 - Conservation of Biological Diversity;

Subject Of Investigation
0840 - Fur-bearing animals; 0850 - Wildlife habitats; 0830 - Wild animals;

Field Of Science
1060 - Biology (whole systems);
Goals / Objectives
Habitat fragmentation often results in substantial species- and community-level changes and has been cited as a major threat to biodiversity. Human activities, such as the conversion of land to agriculture and urban development, represent the main cause of fragmentation in many areas. In forested ecosystems, fragmentation often results in negative effects on carnivores, which require relatively large tracts of habitat and are regarded as sensitive to human disturbance. As carnivores serve several key ecosystem functions, such as exerting regulatory top-down forces on lower trophic levels, the loss of carnivore species and communities represents a potential conservation concern. Island Biogeography Theory provides a framework for predicting the effects of fragmentation by relating species occurrence and richness to patch size and degree of isolation. According to theory, large habitat patches should have more species than small ones and connected patches should have more species than isolated ones. However, the effects of fragment size and isolation on carnivores are not clear. For example, a review of the effects of fragmentation on several carnivores concluded that patch area and isolation were surprisingly poor predictors of occupancy, whereas a study of 11 sympatric carnivore species in California suggested that fragment area and isolation were the two strongest predictors of distribution and abundance. These conflicting accounts suggest that other landscape factors affect species and community occupancy of habitat fragments. Forest ecosystems once covered much of the Champlain Valley of Vermont. However, agricultural expansion and urbanization have fragmented the area into a patchwork of small forest fragments. The effects of fragmentation are largely unknown and may be severe for some species, especially carnivores, some of which are listed as Vermont Species of Greatest Conservation Need. This project will examine the effects of forest fragmentation on carnivore species. It will explore the effects of fragment size and relative isolation as well as surrounding human development on the likelihood of carnivore occupancy. The project aims to address the following questions: (1)Is carnivore occupancy in forest fragments affected by fragment size and isolation (2)Is carnivore occupancy in forest fragments affected by surrounding human development Project outputs will include articles detailing results that will be submitted to peer-reviewed scientific journals for publication. Other outputs will include presentations of the results at professional conferences, graduate student training, and the publication of a Masters thesis. The results will also be presented to local stakeholders including the Vermont Division of Forestry, Vermont Fish and Wildlife Department, and non-profit organizations. The project will provide a baseline measure of how forest fragment characteristics affect carnivores. While the project may be completed with the funds requested, it has the strong potential to attract further funding, especially given increasing rates of fragmentation in the Champlain Valley and the conservation priorities of the state.
Project Methods
Study area. The project will occur in the Champlain Valley in northeastern Vermont. The region is bounded on the east by the foothills of the Green Mountains and west by Lake Champlain. The northern border occurs in Colchester town and the southern border is approximately route 73 from Brandon town west to Lake Champlain. The Champlain Valley represents one of the most developed regions of Vermont and includes a mosaic of urban, suburban, agricultural, and rural farmland. Several forest patches occur in the Valley, some of which lie in protected areas such as Snake Mountain Wildlife Management Area and Mount Philo State Park, while others occur on private land and in urban and suburban environments. The Champlain Valley is characterized by the Champlain Lowlands physiographic region (Johnson 1998). Surveys. The project will select 60 fragmented forest sites in the study area. These sites will be surveyed to estimate carnivore occupancy using camera traps for larger species, such as gray fox, red fox (Vulpes vulpes), coyote (Canis latrans), and bobcat, and track-plate stations for smaller species, such as long-tailed weasel (Mustela frenata). Surveys of all sites will occur monthly in two seasons including summer (May to September) and winter (November to March). Traps will be deployed at the beginning of each season and monitored monthly. The project will estimate species presence/absence at each site. Occupancy estimation. The likelihood of occupancy will be estimated for each species using occupancy models as described by Mackenzie et al. (2006). Occupancy will be estimated using multiple-season models and based on detection probability calculated from repeated monthly surveys of each site. To address the first objective, the project will estimate the size of each forest fragment and degree of isolation (distance to nearest forest patch) and examine whether these two factors affect occupancy. To address the second objective, the project will measure three human developments surrounding forest fragments, including road cover, housing density, and amount and type of agriculture at multiple spatial scales. Forest fragment size, degree of isolation, and surrounding human development will be estimated from ground surveys, aerial photos and state and national Geographic Information Systems (GIS) databases. These characteristics will be incorporated into models as covariates to evaluate their influence on occupancy. Analyses will be performed with the occupancy modeling program PRESENCE v. 3.0 (Hines 2006).

Progress 10/01/10 to 09/30/12

Outputs
OUTPUTS: The project examined the effects of forest fragmentation on carnivores in the Champlain Valley of Vermont, which represents the largest agricultural landscape in the state. The project used non-invasive camera traps to survey the distribution of multiple carnivore species, including red fox, gray fox, coyote, bobcat, fisher, bear, and raccoon, in forested fragments in the study area. We surveyed over three discrete time periods (summer, fall, and winter), then modeled occupancy probability at a given site. Our modeling approach accounted for imperfect detection (i.e., the probability of detecting a species), and incorporated the effects of several covariates, including the size, isolation, and surrounding habitat characteristics of forest fragments. Surveys resulted in over 22,000 photos. We summarized the data, quantified covariates (using Geographic Information Systems), and used two approaches to modeling (multinomial maximum likelihood and Bayesian). Some of the results contributed to a graduate thesis (S. Williams, Master of Science, Forest and Wildlife Sciences Program, University of Vermont) and we are currently preparing two additional manuscripts of other results that will be submitted in early 2013. We contributed all of our project data, including carnivore detection records and covariate data, to the Vermont Fish and Wildlife Department ROVER on-line database (www.vermontrover.com). The database provides long-term storage of wildlife data (from multiple state-wide projects) and data access to wildlife professionals and interested parties. The database is still partly under development and access is password controlled at the moment. Among other on-line resources include a website used to disseminate project information to the public (www.champlainvalleycarnivores.wordpress.com) and a blog for students working on the project (www.cvcarnivoreblog.wordpress.com). We also presented the results of the project to the Vermont Coverts organization (www.vtcoverts.org). Vermont Coverts is a non-profit organization composed of woodland landowners and managers in Vermont that is missioned to: encourage and effect the maintenance, enhancement and/or creation of wildlife habitat on private forest lands in Vermont. We presented the project, its results, and main recommendations for management at the group's annual conference in 2012. The project has also focused on training students. The project supported a graduate student (Master of Science) and resulted in a graduate thesis in December 2012. The student received training in project design, implementation, analysis, and write-up along with public outreach. The project also sponsored six undergraduate internships. Most of the interns were wildlife students, and we provided training in field skills and data analysis. PARTICIPANTS: Individuals that worked on the project included: James Murdoch, Ph.D., who served as Principal Investigator. Dr. Murdoch is an Assistant Professor in the Wildlife and Fisheries Biology Program at the University of Vermont. Sara Williams served as a graduate student on the project. She completed her Master of Science degree (based on project results) in December 2012. Six interns also contributed to the project. Interns received academic credit through the University of Vermont, and included: Natasha Lekach, Dylan Miles, Emily Prosser, Corey Tondreau, and Alex Potash (2 separate internships). Partner organizations included various groups that hosted survey sites on their property. Among them included state agencies (Vermont Agency of Natural Resources: Vermont Fish and Wildlife Department, Vermont Department of Forests, Parks, and Recreation), federal agencies (U.S. Fish and Wildlife Service: National Wildlife Refuge System), municipalities (Towns of Hinesburg, Monkton, and Charlotte), non-profit organizations (The Nature Conservancy, Audubon Society, Winooski Valley Park District, Vermont Coverts), and several private landowners. Training and professional development opportunities provided by the project include training of a graduate student (Sara Williams) in research design, implementation, analysis, and presentation, and undergraduate interns in field skills and data analysis. TARGET AUDIENCES: Target audiences for the project included wildlife biology students and forest landowners and managers in Vermont. Efforts for students included training in research design, implementation, analysis, and presentation to a graduate student. The student (S. Williams) completed her Master of Science degree in December 2012. Efforts included mentorship and formal training through classroom instruction, independent study, and extracurricular activities. Efforts were also geared toward training undergraduate students. Training included mentorship of interns, who were all wildlife biology students. All internships (6) were credit-bearing and included formal instruction, in-field skills development, and data analysis. Efforts aimed at forest landowners and managers included presenting project results and recommendations to the Vermont Coverts non-profit group at their annual conference in 2012. The group includes a large membership of state-wide forest landowners and managers that is missioned to: encourage and effect the maintenance, enhancement and/or creation of wildlife habitat on private forest lands in Vermont. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
Our modeling results provide a measure of how forest fragmentation affects carnivore species. Specifically they address how fragment size, isolation, and surrounding habitat characteristics (including human development) affect these species. Results vary by species and provide woodland managers with tools for decision-making about forested systems. For example, fishers represent a widespread species that provides an important ecosystem service (i.e., controlling porcupine populations through predation). Our results suggested that fisher occurrence in a forested patch is determined by patch size and ruggedness, as well as the interaction between patch characteristics and surrounding landscape characteristics. Our fisher models provide a means of predicting fisher occurrence under various landscape scenarios. For example, if a forested area was to be developed, models would allow us to determine how fishers would be affected, which could have implications for the long-term sustainability of the remaining forest in the system. We presented our model results for various species to the Vermont Coverts organization (see Outputs above) and provided recommendations to members, which consist mainly of woodland owners and managers throughout Vermont. The recommendations provide tools for better managing woodlands for wildlife. Our modeling results also yielded insight into the interactions of carnivore species. Data on species interactions in the Northern Forest are lacking, yet important from a conservation perspective. For example, our modeling results examined patterns of co-occurrence between red foxes and gray foxes, which are sympatric in Vermont. Results indicated that red foxes negatively affect the occupancy probability of gray foxes at a given location in the landscape. This is a significant finding as gray foxes are listed as a Species of Greatest Conservation Need in Vermont, and can inform management of forested systems for the species. Other key outcomes include training of wildlife students. The project supported and trained a graduate student and several undergraduate interns. The graduate student completed her degree and attained skills in wildlife management and data analysis that earned her a competitive position as a Ph.D. student at the University of Montana. Her doctoral project will involve using a similar approach to modeling to examine the effects of landscape change on whales in Alaska. One of the interns also used his experience to apply for an NSF Graduate Research Fellowship (with faculty support) and others (2) used their experiences to attain employment in the wildlife research field (e.g., federal and state wildlife agencies).

Publications

  • Williams, S. 2012. Factors affecting fisher occupancy in forest fragments of an agricultural landscape. Master of Science thesis, University of Vermont, Burlington, Vermont, USA.
  • Williams, S., J. D. Murdoch, T. Donovan, and W. Kilpatrick. 2013. Effects of forest patch area and isolation on carnivores in an agricultural landscape of Vermont. In preparation.
  • Williams, S., and J. D. Murdoch. 2013. Patterns of co-occurrence of two sympatric fox species in Vermont. In preparation.


Progress 10/01/10 to 09/30/11

Outputs
OUTPUTS: The study investigates the effects of forest fragmentation on carnivores in the Champlain Valley of Vermont. The methodological approach involves modeling carnivore occurrence in forest fragments as a function of fragment characteristics, such as size, degree of isolation, and extent of surrounding human developments. Carnivore occurrence is estimated non-invasively using infrared camera-traps. During the reporting period, I completed initial fieldwork for the study. I established 51 camera-trap sites in collaboration with state and federal agencies, municipalities, non-profit organizations, community groups, private landowners, and local businesses. I completed six to eight 1-month surveys per camera-trap across two seasons and recorded all target carnivore species (except long-tailed weasel), including bobcat, coyote, red fox, gray fox, black bear, and fisher, and several non-target species including deer, moose, possum, squirrel, cottontail, and various birds of prey. I also estimated forest fragment characteristics for each camera-trap site using Geographic Information Systems. I calculated forest fragment size, isolation, and surrounding human developments at multiple spatial scales. These data will be used to develop the final model. I recruited, trained, and supervise a graduate student, who developed a Master of Science thesis project as part of the study. I also sponsored four undergraduate internships that participated in fieldwork. All photographs taken by camera-traps have been disseminated on a website. The website allows collaborators and the general public to view photographs, which are updated after each survey. Preliminary data have been presented to a graduate research symposium and a community group. PARTICIPANTS: James D. Murdoch is an Assistant Professor at the University of Vermont and serves as Principal Investigator for the study. He developed and implemented the study and assumes responsibility for its completion. Sara Williams is a Master of Science graduate student at the University of Vermont and is supervised by James Murdoch. The study forms the basis of her graduate thesis project. She has been involved with the development and implementation of the study, and actively involved in fieldwork. Several partner organizations have hosted camera-traps. Among them include state and federal agencies such as the Vermont Agency of Natural Resources and US Fish and Wildlife Service, municipalities such as the Towns of Monkton, Hinesburg, and Charlotte, and non-profit organizations such as Green Mountain Audubon, The Nature Conservancy, and Winooski Valley Park District. Several local community groups, private landowners, and businesses also hosted camera-traps. The study provided graduate training to Sara Williams and sponsored four volunteer interns (University of Vermont undergraduates), who participated in fieldwork. TARGET AUDIENCES: The primary target audience is forest managers. The study aims to provide tools that will allow managers to better manage wildlife in the developing landscape of the Champlain Valley. Target managers include state and federal agencies, municipalities, non-profit organizations, and private landowners. PROJECT MODIFICATIONS: I made no major changes or modifications to the study during the reporting period.

Impacts
The study completed initial fieldwork during the reporting period. Initial findings indicate that species detections will be high enough to model the data and effectively address the objectives. There were two surprising outcomes that will lead to further inquiry and enhance the study. The first was a relatively high rate of domestic dog (non-leashed) detections at camera-sites. Domestic dogs may exert competitive pressure on native carnivores by excluding or killing them, and will be an important consideration in the final analyses. The second was a relatively high rate of raccoon detections. Raccoons represent the primary vector of rabies in Vermont and the focus of rabies control efforts in forested ecosystems (i.e., through the use of non-invasive vaccine baits). High rates of raccoon detection will lead to additional models that will provide wildlife managers with tools to better identify target areas for vaccination efforts. There have been no publications to-date as the study is still in the initial data collection phase. Fieldwork will continue into the next reporting period. I also plan to analyze data and develop manuscripts for publication during the next period.

Publications

  • No publications reported this period