Long-term population monitoring of northern Idaho ground squirrel: 2014 implementation and population estimates

Publication Type:

Report

Authors:

Source:

Endangered Species Section 6 Grants F12AP00815 and F13AF01075, Idaho Department of Fish and Game, Boise, p.16 plus app. (2014)

Call Number:

U14MAC01IDUS

Keywords:

northern Idaho ground squirrel, Spermophilus brunneus brunneus, Urocitellus brunneus

Abstract:

We implemented the first full year of the new northern Idaho ground squirrel (NIDGS) long-term population monitoring strategy in 2014, following a pilot year in 2013. We developed a revised sampling frame of 1,757 100-m x 100-m grid cells across occupied habitat, sampled from those cells with an equal-probability, spatially balanced sampling procedure called Balanced Acceptance Sampling (BAS), and used the resulting ranks to develop a 3-year rotating panel design of 919 cells surveyed every year, including 500 core cells sampled every year and all cells sampled within 3 years. We conducted line-transect distance surveys across 918 cells and recorded 725 NIDGS detections at 275 cells (30%). From these data, program DISTANCE estimated a density of 0.55 squirrels/ha and a total population size of 968 squirrels (95% CI: 802–1,180). We post-stratified data based on relative density (higher, lower), with resulting densities of 0.67 squirrels/ha in stratum 1 compared with 0.12 squirrels/ha in stratum 2, and unadjusted population sizes of 920 and 48, respectively. We compared the 1-year change in population estimates between 2013 and 2014 from the DISTANCE analyses of all data each year, from DISTANCE analysis of 499 grid cells surveyed both years, and from a pair-wise comparison of 260 cells from the 500 core cells intended to be surveyed every year. On average, almost 50% fewer NIDGS groups were detected on surveys in 2014 compared with 2013, and modeled densities and population estimates from various subsets of data corroborate an apparent 1-year decline. Our data suggest that the difference between 2013 and 2014 was more pronounced in stratum 2 (lower density) sites. We conducted presence–absence surveys at 5 sites where no grid cells were selected for surveys this year and at 24 sites where no squirrels had been observed during line-transect surveys to establish their status. We recommend that 2014 be considered the baseline of long-term population monitoring. 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.