Premium
Integrated modeling to estimate population size and composition of mule deer
Author(s) -
Furnas Brett J.,
Landers Russ H.,
Hill Scott,
Itoga Stuart S.,
Sacks Benjamin N.
Publication year - 2018
Publication title -
the journal of wildlife management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.94
H-Index - 111
eISSN - 1937-2817
pISSN - 0022-541X
DOI - 10.1002/jwmg.21507
Subject(s) - odocoileus , transect , wildlife , mark and recapture , abundance (ecology) , aerial survey , population , environmental science , range (aeronautics) , telemetry , habitat , population size , statistics , abundance estimation , ecology , geography , biology , remote sensing , demography , mathematics , computer science , telecommunications , materials science , sociology , composite material
Estimating population size, age composition, and sex ratio of mule deer ( Odocoileus hemionus ) is important to conservation and managed hunting of this species in the western United States. Increasingly, wildlife agencies are estimating abundance of deer using fecal DNA (fDNA), especially in forested habitats where aerial surveys are not feasible. These same data can be used to estimate overall sex ratio but require additional data on age structure to quantify adult‐ and fawn‐specific sex ratios, which are expected to differ substantially. We demonstrate an integrated modeling approach to estimating population sizes of adult females, adult males, and fawns from 3 sources of data: fDNA, camera stations, and global positioning system (GPS) telemetry. We conducted the study on an 11,500‐km 2 forested region in northern California, USA, corresponding to 3 hunt management zones. Within a Bayesian framework, we used spatial capture–recapture (SCR) modeling of fDNA samples and prior information on home range sizes from telemetry to estimate sex‐specific densities, and N‐mixture modeling of camera detections to separate adult and fawn densities. We estimated 29,317 adult females (90% CI = 24,550–34,592), 10,845 adult males (90% CI = 7,778–14,858), and 19,587 fawns (90% CI = 15,340–24,430) within the study area. The inclusion of telemetry increased precision of our results, and cameras provided comparable estimates of density when we calibrated them on the SCR results. Based on these results, we recommend a monitoring program of fDNA transects repeated once every 5 years, camera stations repeated at half of transects every year, and telemetry data from 1 deer for every 2 transects on average. We estimated an average annual cost of $1,316 (U.S.) per transect to sustain this endeavor. The integration of cameras with fDNA to combine age structure data with sex‐specific abundance data represents a novel and significant step forward in the capacity to estimate deer population parameters. © 2018 The Authors. Journal of Wildlife Management published by Wiley Periodicals, Inc. on behalf of The Wildlife Society.