Long-term population monitoring of northern Idaho ground squirrel: sampling design, pilot implementation, and 2013 population estimates

Call Number:

Keywords:

Abstract:

In 2013 we piloted a new sampling design for northern Idaho ground squirrel (NIDGS) long-term population monitoring. We used a layered monitoring approach that combined grid-based line-transect distance sampling within a patch occupancy framework to provide a statistically valid, repeatable approach for estimating population size and trend each year for a time frame of 20–30 years. From a universal 100 m-by-100 m grid we defined a sampling frame as those grid cells that overlapped currently occupied NIDGS sites by ≥40%. We selected cells to be surveyed from a ranked list generated by Generalized Random Tessellation Stratified (GRTS) sampling, an equal probability sampling procedure. We aimed to survey 1,000 cells, or ~57% of the available cells. We conducted distance sampling along transects and analyzed survey data with program DISTANCE. Analyses were based on 941 grid cells that were carried forward into the final sampling frame. We detected 1,280 NIDGS at 375 of these 941 cells (40%), and estimated a density of 1.26 squirrels/ha and a total population size of 2,093. We compared the new grid-based design to site-based surveys used in prior years by conducting both types of surveys at 10 sites. Population estimates were equivalent for the two methods. We provided a method to quantify the magnitude of ‘missed squirrels’ by revisiting line transect survey data and mark-recapture data from 10 sites sampled with both methods in 2012. We concluded that field implementation of the new sampling design was straightforward and yielded comparable population estimates as those obtained in previous years from site-based approaches. We recommend that the grid-based sampling design using distance sampling along transects within a patch occupancy framework be adopted. Results from 2013 should be the baseline of long-term monitoring, with the caveat that changes to the sampling frame between the 2013 and 2014 field seasons could influence interpretation of changes in density estimates in subsequent years. To continue to validate estimates of population size from distance-based line-transect surveys, we recommend repeating a comparison of survey to mark-recapture estimates every 3–5 years.

Notes:

ELECTRONIC FILE - Zoology, Mammals

Cooperative Agreement No. F11AC00175

Diane's citation for this document (in her 2015 report):
Evans Mack, D., C. Gillen, and L. McDonald. 2013. Long-term population monitoring of northern Idaho ground squirrel: sampling design, pilot implementation, and 2013 population estimates. Unpublished report to the U.S. Fish and Wildlife Service, threatened and endangered species projects F11AP00417 and F12AP00815 and cooperative agreement no. F11AC00175. December 31, 2013. Idaho Department of Fish and Game, Boise, Idaho, USA.