EFFECTS OF FENCES ON PREDATION, HABITAT SELECTION, AND CONNECTIVITY OF BLACK-TAILED DEER IN CALIFORNIA

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: HATCH
• Reporting Frequency: Annual
• Accession No.: 1009064
• Project Start Date: Jan 6, 2016
• Project End Date: Sep 30, 2020
• Program Code: [(N/A)]- (N/A)

Recipient Organization
UNIVERSITY OF CALIFORNIA, BERKELEY
(N/A)
BERKELEY,CA 94720

Performing Department
Ecosystem Sciences

Non Technical Summary
Black-tailed deer (Odocoileus hemionus columbianus) have significant social and economic importance in California. In the pastseveral decades, black-tailed deer populations in the United States have fluctuated widely and experienced overall decline.California populations similarly have shrunk from their peak in the 1960s. The drivers of deer decline are diverse and remainpoorly understood. To help understand these declines, we propose to study 1) how fences impact deer predation bymountain lions, black bears, and coyotes, and 2) how fences change connectivity patterns and habitat selection fordeer. The widespread fencing of land for growing grapes and for minimizing agricultural disease transmission has profoundlyaltered the permeability of many California habitats for big game. Countless hunters, hikers, and wildlife and land managers inCalifornia have raised concerns about the effects of fencing on deer and other wildlife populations, and yet there has been littleto no rigorous research on the impact of fencing on game species in California or elsewhere. We propose to address this criticalgap through a collaborative study that combines cutting-edge technology with tradition wildlife management methods.

Animal Health Component

80%

Research Effort Categories

Basic

10%

Applied

80%

Developmental

10%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
135	0899	1070	50%
136	0830	1070	25%
121	0850	3100	25%

Knowledge Area
135 - Aquatic and Terrestrial Wildlife; 136 - Conservation of Biological Diversity; 121 - Management of Range Resources;

Subject Of Investigation
0850 - Wildlife habitats; 0830 - Wild animals; 0899 - Wildlife and natural fisheries, general/other;

Field Of Science
1070 - Ecology; 3100 - Management;

Keywords

wildlife management

black-tailed deer

fences

predation

agriculture

biodiversity conservation

Goals / Objectives
Black-tailed deer (Odocoileus hemionus columbianus) have significant social and economic importance in California. In the pastseveral decades, black-tailed deer populations in the United States have fluctuated widely and experienced overall decline.California populations similarly have shrunk from their peak in the 1960s. The drivers of deer decline are diverse and remainpoorly understood (Wittmer et al., 2014).To help understand these declines, we propose to study 1) how fences impact deer predation by mountain lions, blackbears, and coyotes, and 2) how fences change connectivity patterns and habitat selection for deer. The widespreadfencing of land for growing grapes and for minimizing agricultural disease transmission has profoundly altered the permeabilityof many California habitats for big game. Countless hunters, hikers, and wildlife and land managers in California have raisedconcerns about the effects of fencing (particularly the now common woven wire fence) on deer populations, and yet there hasbeen little to no rigorous research on the impact of fencing on deer or other game species in California or elsewhere. Wepropose to address this critical gap through a collaborative study that uses cutting-edge technology and methods with which weare very familiar. This research can be summarized in two general questions.Question 1: How do fences alter spatial patterns and frequency of deer predation?Predation is a key factor in black-tailed deer declines, particularly when multiple predators target diverse deer life stages, asthey do in much of California (Wittmer et al. 2014). Fences alter spatial patterns of predator-prey interactions, as carnivores usefences to trap prey or select ambush sites beside high-use trails along fence lines. For example, bears in California have beenobserved to run their prey into fences, including elk calves (S. Koller, pers. comm.) and deer fawns (B. Davis, pers. comm.).Camera traps have shown that multi-strand and wire-mesh fences separate does from their fawns and thereby likely increaserisk of predation (e.g., Brashares and Keiffer unpublished). Anecdotal evidence further suggests that pumas target deer along fence trails and have a higher kill rate in these "pinch points" for movement and connectivity (B. Davis pers. comm.). By creatingadditional high predation risk areas, fences may also shape the landscape of fear for prey species, changing prey behavior andeffectively altering resource availability (Brown et al. 1999). By studying the ways in which fences alter patterns of deerpredation, we will refine our understanding of one of the primary sources of deer mortality.Question 2. How do fences alter patterns of deer habitat use and landscape connectivity?Fences are likely to alter patterns of landscape connectivity both for individual deer (through effects on movement and habitatuse) and for populations (through effects on population demography, metapopulation dynamics, and gene flow). Connectivity isa critical factor in determining distribution, population growth, and depredation patterns (Rudnick et al. 2012). Nevertheless, littleresearch has examined connectivity for black-tailed deer and their predators in California's mosaic landscape. Some fences,particularly those associated with agriculture and other intensive land uses, may create impermeable barriers to the movementof black-tailed deer, limiting access to preferred habitats and impeding optimal foraging. Even where animals can cross fences,the energy expenditure associated with jumping or climbing through a fence is costly. Animals often choose instead to movealong fences, creating high-traffic movement paths. At large scales, networks of various types of fences may greatly affectlandscape connectivity for black-tailed deer across California, with implications for population and metapopulation dynamics.Our broad study questions can be broken down into the following objectives:1. Determine the spatial effects of fences on interactions between deer and their predators (mountain lions, black bears, andcoyotes). Specifically, we will:• Compare fence permeability/fence-crossing ability between black-tailed deer and large carnivores;• Determine if fences exclude predators and create spatial refugia for deer;• Determine if fences exclude or concentrate deer and create zones of attraction for predators along fences;• Examine if predation events occur disproportionally along fences.2. Determine the effect of fences on black-tailed deer movement, resource selection patterns, and landscape connectivity.Specifically, we will:• Assess the frequency with which deer cross fences (relative to fence crossing opportunities) to determine fencepermeability;• Compare relative permeability of different types of fences for black-tailed deer;• Examine how fence crossing behavior differs with age and sex of deer;• Explore how deer respond to changes in fence design and layout;• Construct landscape connectivity models that account for fence effects.

Project Methods
Study Site: Situated at the interface of suburban and wild, the Hopland Research and Extension Center (HREC) is the site wepropose to explore how black-tailed deer and their predators respond to fencing. HREC and its neighboring private and publicproperties are latticed with fences of varying age and construction, representing the great majority of fence types used byfarmers, ranchers, and vineyards in California (e.g., woven-wire sheep fence, 3- and 4- strand barbed wire fences, 9-strandhigh-tensile electrified fencing, 7' woven-wire vineyard fencing). HREC also provides a rare opportunity to experimentally alterthe permeability of different fence types and observe wildlife responses to various fence configurations. Additionally, new fenceshave just been erected at HREC and its neighboring properties, creating a timely opportunity for research on their ecologicaleffects. We will collect all data in accordance with HREC protocols, and after obtaining approval from the University of California- Berkeley Animal Care and Use Committee (ACUC) and the Committee for Protection of Human Subjects (CPHS).GPS collars: Recent advances in GPS positioning and accelerometer technologies provide a novel and critically importantopportunity to study the effects of fencing on landscape connectivity for deer populations. We propose to use high-resolutionmovement data from GPS collars as well as observations from camera traps to examine effects of fences on deer. Past GPSstudies, generally using fix rates between 1 and 4 hours, have lacked sufficient spatial-temporal resolution to examine the finescalemovement of deer and their predators around fences. To gain fine spatiotemporal resolution on movement patterns andresponses to fences, we will place GPS collars on 20 adult female black-tailed deer and proximity tags on 10-15 fawns. We willalso collar 5-10 individuals of each of the three chief predator species of the black-tailed deer: 10 coyote (Canis latrans), 5mountain lions (Puma concolor), and 5 American black bears (Ursus americanus. Collars will be programmed to take fixes every1 minute near fences and every 5 minutes elsewhere. Accelerometers will record activity levels every 1 minute at a 10 Hz fixrate for a period of 10 seconds following each GPS fix. Collars will be programmed to drop off when batteries are exhausted,after 4-6 months depending on animal size (the high fix rate leads to more rapid battery depletion than in typical GPS studies),and then re-deployed.Fence Mapping and Camera Trapping: We will combine ground-truthing and other data sources to produce a map of all fencetypes at HREC. We will set up 30 camera traps at locations along fencelines to identify resident individuals and examine deerbehavior at fence crossing points. We have already placed 10 cameras at HREC and collected preliminary data on speciesoccupancy and behavior (Figure 3).Fence Manipulation: Over the course of the study, we will experimentally alter the permeability of fences by adding gaps andgates of various configurations, and alternating the location of fence types (e.g., 3-strand wire, 5-strand electrified, and wiremesh). The willingness of HREC to support such experiments creates a great opportunity for this study.

Progress 01/06/16 to 09/30/20

Outputs
Target Audience:This research targetedacademic, agency and public stakeholders. During the term of thisproject we reached each of these target audiences throught the following:• Publication of three graduate theses, publication or submission of sixpeer-reviewed papers, and three reports.• Participation and presentations (poster or contributed talk) at seven regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and three UC-ANR conferences. • Participation in two workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through two one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), 21 blog pieces by independent bloggers covering our research, and two 'public open house' events at the Hopland Research and Extension Center. • Public education through 11 presentations to K12 school groups in Mendocino and Sonoma Counties Changes/Problems:In July, 2018, the Mendocino Complex Fire burned over 400,000 acres in our study area, making it the largest wildfire in California history. The fire burned 2/3rds of the Hopland Research and Extension Center, the location of this project. We tried to find opportunity to pursue sciencein what otherwise was atrajedy for local businesses and homeowners. We did this by broadening our study to examine the effects of fire on deer and other wildlife. Our choice was to adjust to the changes created by the fireor cancel the project; we chose to adjust. The 2019-2020 project year brough another challenge in the form of the Covid-19 pandemic. As true for countless other studies, our activiites in the field and lab abruptly stopped in March 2020 as we faced the scale of the Covid disaster. With careful management, detailed protocols, and smaller crews we were able to resume fieldwork in June 2020 and managed to complete nearly 90% of our intended project goals for the year. What opportunities for training and professional development has the project provided?This project supported fulltime training of six PhD students, two MSc students and five technicians, and part-time training of 14underraduate students. We have also led three workshops at the Hopland Research and Extension Center providing training on non-lethal strategies for managing wildlife conflict. Attendance was 50+ at each workshop and included land managers, owners, outreach coordinators, state and federal agents and interested members of the public. How have the results been disseminated to communities of interest?This research targetedacademic, agency and private stakeholders. During the term of thisproject we disseminated our results to each of these target audiences throught the following:• Publication of three graduate theses, publication or submission of sixpeer-reviewed papers, and three reports.• Participation and presentations (poster or contributed talk) at seven regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and three UC-ANR conferences. • Participation in two workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through two one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), 21 blog pieces by independent bloggers covering our research, and two 'public open house' events at the Hopland Research and Extension Center. • Public education through 11 presentations to K12 school groups in Mendocino and Sonoma Counties What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Goal 1: Assess and integrate camera and DNA-based estimation methods. Before and during this project period, we maintianed a camera grid in coordination with CDFW and have been collecting camera data since May 2016. There are a total of 36 Recoynx HyperFire cameras in a hexagonal grid. With the camera grid established, fieldwork now consists of routine camera maintenance every 3-4 months. We remove excess grass that can cause false triggers, test and replace batteries, and download and clear memory cards from cameras as needed. So far, all cameras are working without defect. At Berkeley, with the help of undergraduate volunteers, we sort photos by species to format them for statistical analyses. We have so far classified species in >500,000 photos and cleaned data in a .csv file with the location, date, time, and species for all records. Deer are the primary species occurring on the camera traps, representing approximately 70% of camera activity after false triggers have been removed. We have completed our first round of four iterative runs of our fecal DNA transects, following CDFW protocols. We have collected well over 2,000 samples in this process. Extractions continue apace, and even in spite of some rain, our success rate remains high (see previously submitted report for more detail on methods and success rates of extractions). Members of our team attended a workshop on Spatially Explicit Capture-Recapture models (SECR) at The Wildlife Society's annual conference. This model can easily be used in parallel to existing CDFW approaches for further comparison and crossvalidation of population estimates. Goal 2: Build spatially explicit deer occupancy and density models. The main emphasis of task 2 this year has been the deployment of GPS collars, which we paused during the hunting season but have continued since. We have now collared 87does, 16 bucks and placed ear tags on an additional 12 fawns, and have conducted 147 total captures (including recaptures). Ten of the doe collars we deployed included additional custom made GPS and accelerometer units to collect a second stream of extremely fine-scale data (10s GPS, 50hz accelerometer) on deer movement and behavior to further improve our understanding of their movement, behavior, and distributions. While clover traps have proven a safe and effective Report Date 02/23/2021 Page 2 of 4 United States Department of Agriculture Progress Report Accession No. 1009064 Project No. CA-B-ECO-0159-H means of collaring does, bucks have been less interested in entering the traps. We have built a GIS database by inputting spatial information from deployed collars, historical records at HREC, fecal DNA samples, camera trap data and fence locations. In August 2020, we completed our 6th consecutive year of providing HREC hunters with GPS trackers. In combination with deer telemetry data and historical data on deer harvest locations at HREC (now digitized and compiled), this hunter telemetry data will provide the basis for the development of a statewide hunting pressure model. Goal 3: Develop an integrated population model accounting for fence effects. The data being collected from deer GPS collars, camera traps, and fecal DNA sampling are now being fed directly into updates of the integrated population model. DNA and camera trap data are providing initial estimates of sex ratios and age classes (the latter from camera traps only). Our initial estimate of sex ratios from fecal DNA is approximately 3 males per 10 females. Contemporary and historical harvest data from HREC have been digitized, but other predation information is difficult to capture without GPS collars on predators as well. As data continues to stream in with increasing regularity - especially as more collars are deployed - specific updates to the population model will be developed to satisfy task 3 deliverables.

Publications

• Type: Theses/Dissertations Status: Submitted Year Published: 2021 Citation: J. Dorcy. Estimating black-tailed deer populations using integrated camera and genetic models. MSc Thesis, Univ. of California, Berkeley
• Type: Journal Articles Status: Accepted Year Published: 2021 Citation: Identifying individual ungulates from fecal DNA: A comparison of field collection methods to maximize efficiency, ease, and success. Mammalian Biology.
• Type: Journal Articles Status: Published Year Published: 2020 Citation: Fence ecology: Frameworks for understanding the ecological effects of fences A McInturff, W Xu, CE Wilkinson, N Dejid, JS Brashares BioScience 70 (11), 971-985
• Type: Journal Articles Status: Accepted Year Published: 2021 Citation: Patterns of coyote predation on sheep in California: A socio?ecological approach to mapping risk of livestockpredator conflict A McInturff, JRB Miller, KM Gaynor, JS Brashares Conservation Science and Practice, e175
• Type: Journal Articles Status: Accepted Year Published: 2021 Citation: Response of black-tailed deer to megafire in California. Journal of Applied Ecology.

Progress 10/01/18 to 09/30/19

Outputs
Target Audience:This research targets academic, agency and public stakeholders. During this reporting period we reached each of these target audiences throught the following: • Participation and presentations (poster or contributed talk) at seven regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and three UC-ANR conferences. • Participation in two workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through two one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), 21blog pieces by independent bloggers covering our research, and two 'public open house' events at the Hopland Research and Extension Center. • Public education through 11 presentations to K-12 school groups in Mendocino and Sonoma Counties Changes/Problems:In July, 2018, the Mendocino Complex Fire burned over 400,000 acres in our study area, making it the largest wildfire in California history. The fire burned 2/3rds of the Hopland Research and Extension Center, the location of this project. We have tried to find opportunity in this trajedy for local businesses and homeowners by broadening our study to examine the effects of fire on deer and other wildlife. Perhaps more honestly, our choice was to include this dimension or cancel the project. What opportunities for training and professional development has the project provided?This project has supported fulltime training of six PhD students, one MSc student and four technicians, and part-time training of eleven underraduate students. We have also led two workshops at the Hopland Research and Extension Center providing training on non-lethal strategies for managing wildlife conflict. Attendance was 50+ at each workshop and included land managers, owners, outreach coordinators, state and federal agents and interested members of the public. How have the results been disseminated to communities of interest?Our results have been disseminated to academic, agency and public stakeholders throught the following: • Participation and presentations (poster or contributed talk) at six regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and two UC-ANR conferences. • Participation in workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through two one-hour radio interview on Mendocino County Public Radio ("Ecology Hour"), several blog pieces by independent bloggers covering our research, and a 'public open house' event at the Hopland Research and Extension Center. • Public education through presentations to K-12 school groups in Mendocino County and Sonoma County What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period we plan to continue collecting data to address the questions outlined in our project proposal. This will involve maintaining GPS collars on 15-20% of our study deer population such that we can quantify movement responses to fences. It also will involve capturing and collaring predators of deer in our study area. We have secured approval from CDFW to begin captures of mountain lions and still await approvals for bear and coyote captures.

Impacts
What was accomplished under these goals? Goal 1: Assess and integrate camera and DNA-based estimation methods. Before and during this project period, we maintianed a camera grid in coordination with CDFW and have been collecting camera data since May 2016. There are a total of 36 Recoynx HyperFire cameras in a hexagonal grid. With the camera grid established, fieldwork now consists of routine camera maintenance every 3-4 months. We remove excess grass that can cause false triggers, test and replace batteries, and download and clear memory cards from cameras as needed. So far, all cameras are working without defect. At Berkeley, with the help of undergraduate volunteers, we sort photos by species to format them for statistical analyses. We have so far classified species in >500,000 photos and cleaned data in a .csv file with the location, date, time, and species for all records. Deer are the primary species occurring on the camera traps, representing approximately 70% of camera activity after false triggers have been removed. We have completed our first round of four iterative runs of our fecal DNA transects, following CDFW protocols. We have collected well over 1,000 samples in this process. Extractions continue apace, and even in spite of some rain, our success rate remains high (see previously submitted report for more detail on methods and success rates of extractions). Members of our team attended a workshop on Spatially Explicit Capture-Recapture models (SECR) at The Wildlife Society's annual conference. This model can easily be used in parallel to existing CDFW approaches for further comparison and crossvalidation of population estimates. Goal 2: Build spatially explicit deer occupancy and density models. The main emphasis of task 2 this year has been the deployment of GPS collars, which we paused during the hunting season but have continued since. We have now collared 74does, 16bucks and placed ear tags on an additional 12 fawns, and have conducted 127 total captures (including recaptures). Ten of the doe collars we deployed included additional custom made GPS and accelerometer units to collect a second stream of extremely fine-scale data (10s GPS, 50hz accelerometer) on deer movement and behavior to further improve our understanding of their movement, behavior, and distributions. While clover traps have proven a safe and effective means of collaring does, bucks have been less interested in entering the traps. Wehave built a GIS database by inputting spatial information from deployed collars, historical records at HREC, fecal DNA samples, camera trap data and fence locations. In August 2019, we completed our 5th consecutive year of providing HREC hunters with GPS trackers. In combination with deer telemetry data and historical data on deer harvest locations at HREC (now digitized and compiled), this hunter telemetry data will provide the basis for the development of a statewide hunting pressure model. Goal 3: Develop an integrated population model accounting for fence effects. The data being collected from deer GPS collars, camera traps, and fecal DNA sampling are now being fed directly into updates of the integrated population model. DNA and camera trap data are providing initial estimates of sex ratios and age classes (the latter from camera traps only). Our initial estimate of sex ratios from fecal DNA is approximately 3 males per 10 females. Contemporary and historical harvest data from HREC have been digitized, but other predation information is difficult to capture without GPS collars on predators as well. As data continues to stream in with increasing regularity - especially as more collars are deployed - specific updates to the population model will be developed to satisfy task 3 deliverables.

Publications

• Type: Journal Articles Status: Submitted Year Published: 2019 Citation: Gaynor, K, McInturff, MA, Brashares, JS. Black-tailed deer navigate contrasting spatiotemporal patterns of risk from hunters and mountain lions. Journal of Animal Ecology

Progress 10/01/17 to 09/30/18

Outputs
Target Audience:This research targets academic, agency and public stakeholders. During this reporting period we reached each of these target audiences throught the following: • Participation and presentations (poster or contributed talk) at six regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and two UC-ANR conferences. • Participation in three workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through a one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), fifteen blog pieces by independent bloggers covering our research, and four 'public open house' events at the Hopland Research and Extension Center. • Public education through six presentations to K-12 school groups in Mendocino and Sonoma Counties Changes/Problems:In July, 2018,the Mendocino Complex Fireburned over 400,000 acres in our study area, making it the largest wildfire in California history. The fire burned 2/3rds of the Hopland Research and Extension Center, the location of this project. We have tried to find opportunity inthis trajedy for local businesses and homeowners by broadening our study to examine the effects of fire on deer and other wildlife. Perhaps more honestly, our choice wasto include this dimension or cancel the project. What opportunities for training and professional development has the project provided?This project has supported fulltime training of threePhD students and fourtechnicians, and part-time training of seven underraduate students. We have also led two workshops at the Hopland Research and Extension Center providing training on non-lethal strategies for managing wildlife conflict. Attendance was 50+ at each workshop and included land managers, owners, outreach coordinators, state and federal agents and interested members of the public How have the results been disseminated to communities of interest?Our results have been disseminated to academic, agency and public stakeholders throught the following: • Participation and presentations (poster or contributed talk) at six regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and two UC-ANR conferences. • Participation in four workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. • Public outreach through a one-hour radio interviewon Mendocino County Public Radio ("Ecology Hour"), several blog pieces by independent bloggers covering our research, and a 'public open house' eventat the Hopland Research and Extension Center. • Public education through presentations to K-12 school groups in Mendocino County and Sonoma County What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period we plan to continue collecting data to address the two questions outlined in our project proposal. This will involve maintaining GPS collars on 15-20% of our study deer population such that we can quantify movement responses to fences. It also will involve capturing and collaring predators of deer in our study area. We have secured approval from CDFW to begin captures of mountain lions and still await approvals for bear and coyote captures.

Impacts
What was accomplished under these goals? Goal 1: Assess and integrate camera and DNA-based estimation methods Before and during this project period, we maintianed a camera grid in coordination with CDFW and have been collecting camera data since May 2016. There are a total of 36 Recoynx HyperFire cameras in a hexagonal grid. With the camera grid established, fieldwork now consists of routine camera maintenance every 3-4 months. We remove excess grass that can cause false triggers, test and replace batteries, and download and clear memory cards from cameras as needed. So far, all cameras are working without defect. At Berkeley, with the help of undergraduate volunteers, we sort photos by species to format them for statistical analyses. We have so far classified species in >500,000 photos and cleaned data in a .csv file with the location, date, time, and species for all records. Deer are the primary species occurring on the camera traps, representing approximately 70% of camera activity after false triggers have been removed. We have completed our first round of four iterative runs of our fecal DNA transects, following CDFW protocols. We have collected well over 1,000 samples in this process. Extractions continue apace, and even in spite of some rain, our success rate remains high (see previously submitted report for more detail on methods and success rates of extractions). Members of our team attended a workshop on Spatially Explicit Capture-Recapture models (SECR) at The Wildlife Society's annual conference. This model can easily be used in parallel to existing CDFW approaches for further comparison and crossvalidation of population estimates. Goal 2: Build spatially explicit deer occupancy and density models The main emphasis of task 2 this year has been the deployment of GPS collars, which we paused during the hunting season but have continued since. We have now collared 51 does, 7 bucks and placed ear tags on an additional 12 fawns, and have conducted 97 total captures (including recaptures). Ten of the doe collars we deployed included additional custom made GPS and accelerometer units to collect a second stream of extremely fine-scale data (10s GPS, 50hz accelerometer) on deer movement and behavior to further improve our understanding of their movement, behavior, and distributions. While clover traps have proven a safe and effective means of collaring does, bucks have been less interested in entering the traps. We have begun building the GIS database by inputting spatial information from deployed collars, historical records at HREC, fecal DNA samples, camera trap data and fence locations. In August 2018, we completed our 4th consecutive year of providing HREC hunters with GPS trackers. In combination withdeer telemetry data and historical data on deer harvest locations at HREC (now digitized and compiled), this hunter telemetry data will provide the basis for the development of a statewide hunting pressure model. Goal 3: Develop an integrated population model accounting for fence effects The data being collected from deer GPS collars, camera traps, and fecal DNA sampling are now being fed directly into updates of the integrated population model. DNA and camera trap data are providing initial estimates of sex ratios and age classes (the latter from camera traps only). Our initial estimate of sex ratios from fecal DNA is approximately 3 males per 10 females. Contemporary and historical harvest data from HREC have been digitized, but other predation information is difficult to capture without GPS collars on predators as well. As data continues to stream in with increasing regularity - especially as more collars are deployed - specific updates to the population model will be developed to satisfy task 3 deliverables

Publications

• Type: Conference Papers and Presentations Status: Accepted Year Published: 2018 Citation: Kaitlyn M Gaynor, Alex McInturff, Justin S. Brashares LAYERED LANDSCAPES OF FEAR: BLACK-TAILED DEER NAVIGATE SPATIOTEMPORAL VARIATION IN RISK FROM HUNTERS AND CARNIVORES
• Type: Conference Papers and Presentations Status: Accepted Year Published: 2018 Citation: Samantha E. Kreling, Kaitlyn M. Gaynor, Alex McInturff, Justin S. Brashares THE EFFECTS OF THE MENDOCINO COMPLEX FIRE ON HABITAT USE AND MOVEMENT IN BLACK-TAILED DEER AND OTHER LARGE MAMMALS.

Progress 10/01/16 to 09/30/17

Outputs
Target Audience:This research targets academic, agency and public stakeholders. During this reporting period we reached each of these target audiences throught the following: Participation and presentations (poster or contributed talk) at six regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and two UC-ANR conferences. Participation in four workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. Public outreach through two one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), three blog pieces by independent bloggers covering our research, and two 'public open house' events at the Hopland Research and Extension Center. Public education through eight presentations to K-12 school groups in Mendocino County. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?This project has supported fulltime training of two PhD students and two field technicians, and part-time training of four underraduate students. We have also led two workshops at the Hopland Research and Extension Center providing training on non-lethal strategies for managing wildlife conflict. Attendance was 50+ at each workshop and included land managers, owners, outreach coordinators, state and federal agents and interested members of the public. How have the results been disseminated to communities of interest?Our results have been disseminated toacademic, agency and public stakeholders throught the following: Participation and presentations (poster or contributed talk) at six regional or international conferences, including The Wildlife Society (Western Section and National conferences), the Society for Conservation Biology, Ecological Scoiety of America, and two UC-ANR conferences. Participation in four workshops, including two UC-ANR workshops held on human-wildlife conflict and coexistence, a pest management workshop and a Cal Dept of Fish and Wildlife workshop on deer management. Public outreach through two one-hour radio interviews on Mendocino County Public Radio ("Ecology Hour"), three blog pieces by independent bloggers covering our research, and two 'public open house' events at the Hopland Research and Extension Center. Public education through eight presentations to K-12 school groups in Mendocino County. What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period we plan to continue collecting data to address the two questions outlined in our project proposal. This will involve maintaining GPS collars on 15-20% of our study deer population such that we can quantify movement responses to fences. It also will involve capturing and collaring predators of deer in our study area. We expect approval from CDFW in the next month to begin captures of mountain lions and expect approvals for bear and coyote captures to come within the next 6 months.

Impacts
What was accomplished under these goals? Goal 1: Assess and integrate camera and DNA-based estimation methods Before and during this project period, we established a camera grid in coordination with CDFW and have been collecting camera data since May 2016. There are a total of 36 Recoynx HyperFire cameras in a hexagonal grid. With the camera grid established, fieldwork now consists of routine camera maintenance every 3-4 months. We remove excess grass that can cause false triggers, test and replace batteries, and download and clear memory cards from cameras as needed. So far, all cameras are working without defect. At Berkeley, with the help of undergraduate volunteers, we sort photos by species to format them for statistical analyses. We have so far classified species in >300,000 photos and cleaned data in a .csv file with the location, date, time, and species for all records. Deer are the primary species occurring on the camera traps, representing approximately 70% of camera activity after false triggers have been removed. We have completed our first round of four iterative runs of our fecal DNA transects, following CDFW protocols. We have collected well over 1,000 samples in this process. Extractions continue apace, and even in spite of some rain, our success rate remains high (see previously submitted report for more detail on methods and success rates of extractions). Members of our team attended a workshop on Spatially Explicit Capture-Recapture models (SECR) at The Wildlife Society's annual conference. This model can easily be used in parallel to existing CDFW approaches for further comparison and cross-validation of population estimates. Goal 2: Build spatially explicit deer occupancy and density models The main emphasis of task 2 this year has been the deployment of GPS collars, which we paused during the hunting season but have continued since. We have now collared 22 does, 3 bucks and placed ear tags on an additional 9 fawns, and have conducted 37 total captures (including recaptures). Ten of the doe collars we deployed included additional custom made GPS and accelerometer units to collect a second stream of extremely fine-scale data (10s GPS, 50hz accelerometer) on deer movement and behavior to further improve our understanding of their movement, behavior, and distributions. While clover traps have proven a safe and effective means of collaring does, bucks have been less interested in entering the traps. We have begun building the GIS database by inputting spatial information from deployed collars, historical records at HREC, fecal DNA samples, camera trap data and fence locations. In August 2017, we completed our 3rd consecutive year of providing HREC hunters with GPS trackers. In combination with deer telemetry data and historical data on deer harvest locations at HREC (now digitized and compiled), this hunter telemetry data will provide the basis for the development of a statewide hunting pressure model. An initial draft of this model was presented at the 2017 Wildlife Society conference in Albuquerque in September, and feedback was positive and constructive. The model requires more revision, data, and validation to be considered robust. Goal 3: Develop an integrated population model accounting for fence effects The data being collected from deer GPS collars, camera traps, and fecal DNA sampling will feed directly into updates of the integrated population model. DNA and camera trap data are providing initial estimates of sex ratios and age classes (the latter from camera traps only). Our initial estimate of sex ratios from fecal DNA is approximately 3 males per 10 females. Contemporary and historical harvest data from HREC have been digitized, but other predation information is difficult to capture without GPS collars on predators as well. As data continues to stream in with increasing regularity - especially as more collars are deployed - specific updates to the population model will be developed to satisfy task 3 deliverables

Publications

• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Kaitlyn M. Gaynor; Alex McInturff; Justin S. Brashares. Layering Landscapes of Fear: the Role of Predation Risk and Human Disturbance in a California Large Mammal Community. The WIldlife Society Annual Conference, Santa Rosa, CA

Progress 01/06/16 to 09/30/16

Outputs
Target Audience:This research targets academic, agency and public stakeholders. First, we expect this research will provide a novel framework for the study of fence effects worldwide. We expect to publish findings across a range of outlets including wildlife management journals (e.g., JWM), high---impact generalist venues (e.g., Science) and popular outlets (e.g., National Geographic). We will produce, at minimum, separate manuscripts based on empirically---informed models of fence effects on deer---predator interactions, deer habitat utilization, deer movement and behavior, predator movement and behavior, relative permeability of different fence types, and landscape connectivity for deer. Second, we will present our findings at regional and national wildlife management conferences. We will make this work widely available to the public, including stakeholder groups in the area. Working with the UC Hopland Research and Extension Center, California DFW and UC Extension specialists, we have and will continue to host workshops on site and reach out to local land managers, hunting groups, and interested citizens. We plan to engage both citizen scientists and UC undergraduate students in processing camera trap data. The YouTube channels of CDFW and UC Cooperative Extension provide accessible venues for informative videos on research and related management suggestions for landowners. GPS location data from collared animals will be uploaded to a public web interface in near---real time (as deemed appropriate by CDFW), providing educators and the public with an interactive view of wildlife movements. Finally, we will publicize the research project and its findings through the UC Berkeley press office, providing press releases to local and regional outlets (e.g., SF Chronicle), and publish regular updates via blog posts on the HREC and UCB College of Natural Resources websites. Our team has extensive experience as communicators of science through popular science writing using traditional and online media. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Our research has employed and/or trained 5 technicians, 2 graduate students, 1 postdoctoral researcher and 6 undergraduate students. How have the results been disseminated to communities of interest?The project is less than 9 months old, but we have distributed early results and engaged stakeholders through 1) two workshops, 2) articles in popular media and 3) appearances on to local radio programs. What do you plan to do during the next reporting period to accomplish the goals?We have made great strides in capturing and collaring wildlife, as described in our intitial proposal, and this is a key methodological foundation on which the rest of the study is built.

Impacts
What was accomplished under these goals? In the first nine months of this project we secured all necessary federal, state and UC permits and approvals. We a) recruited two staff members, b) installed an array of 40 camera "traps" for wildlife monitoring, c) developed, tested and implemented a lab protocol for genetic mark-recapture of deer using fecal samples, d) custom designed GPS collars and deployed four collars on adult deer.

Publications