Details - Bat surveys on USFS Northern Region 1 lands in Montana : 2006

Bat surveys on USFS Northern Region 1 lands in Montana

Title

Bat surveys on USFS Northern Region 1 lands in Montana : 2006

Title Variants

Alternative: Bat surveys on United States Forest Service Northern Region One land in Montana : 2006

By

Lenard, Susan.

Currier, Coburn, 1970-
Hendricks, P.
Maxell, Bryce A. (Bryce Alan)
Montana Natural Heritage Program.

Type

Book

Material

Published material

Publication info

Helena, Mont, Montana Natural Heritage Program, c2007

Notes

"June 2007".

Prepared for: USDA Forest Service Northern Region.

The distribution and status of bats in Montana remain poorly documented on US Forest Service Northern Region lands. The Northern Region recognized the need for additional documentation of bats on Forest Service lands and initiated bat surveys in 2005 across the Region on selected National Forest (NF) Ranger Districts (RD). In Montana, these included Bozeman RD-Gallatin NF, Swan Lake RD-Flathead NF, Townsend RD-Helena NF, Libby RD-Kootenai NF, and Judith RD-Lewis & Clark NF. In 2006, the second year of the project, increased number of surveyors in the field resulted in greater survey effort with both mist-net and acoustic sampling in the following RDs, Butte and Dillon RD, Beaverhead-Deerlodge NF, Sula and West Fork RD, Bitterroot NF, Ashland, Beartooth, and Sioux RD , Custer NF, Tally Lake RD-Flathead NF, Helena, Lincoln, and Townsend RD-Helena NF, Fortine and Rexford RD-Kootenai NF, Mussellshell RD, Lewis and Clark NF, and Superior RD, Lolo NF. Following a modified protocol based on the Oregon Bat Grid system, crews surveyed non-randomly chosen suitable habitats within randomly chosen 10 km2 sample units in each RD for a total of 75 sites surveyed on Northern Region lands in Montana. This approach was primarily targeted at identifying species richness within grid cells; inferences on rates of occupancy are limited to the percent of 10 x 10 km2 grid cells where a species was detected within each sampled RD. The 2006 field survey filled important gaps in documented distributions in Montana, adding new county records. However, a summary of all existing bat records across the region continues to show large distribution gaps for all species, underscoring the need for additional surveys. In particular, large portions of the Beaverhead-Deerlodge NF, Custer NF, Flathead NF, Gallatin NF, and Lewis and Clark NF lack records for any bat species. Even with two years of surveys only two Districts (Beartooth RD-Custer NF and Libby RD-Kootenai NF) have documented the full compliment of species predicted to occur there. Ten species of bats were captured by mist net or detected by acoustic recording during the USFS surveys between late June and early September 2006. Species recorded included Little Brown Myotis (Myotis lucifugus) at 34 sites, Western Long-eared Myotis (M. evotis) at 37 sites, Fringed Myotis (M. thysanodes) at nine sites, Long-legged Myotis (M. volans) at 25 sites, California Myotis (M. californicus) at four sites, Western Small-footed Myotis (M. ciliolabrum) at 17 sites, Big Brown Bat (Eptesicus fuscus) at 23 sites, Hoary Bat (Lasiurus cinereus) at 38 sites, Silver-haired Bat (Lasionycteris noctivagans) at 28 sites, and Spotted Bat (Euderma maculatum) at three sites. California Myotis was detected by acoustic recording at three sites outside their known distribution; these observations are considered tentative until the species is captured with mist nets in the area. Call analysis has yet to be performed on seven sites. Genetic analysis is needed for species identification for single individuals netted at three sites. Surveys at four sites detected no bats during mist-netting efforts; no acoustic sampling was done on these sites. Tentative identification was made for Yuma Myotis at mist-netting sites, but no acoustic recordings produced calls definitive for the species and no genetic analysis has been performed that confirm the species presence in the state. All previously recognized observations of Yuma Myotis appear to be misidentifications of Little Brown Myotis given recent acoustic analysis at a number of sites previously identified Yuma Myotis roost sites. The presence of this species in the state is highly questionable given the lack of definitive documentation. Detection probabilities for bats with multiple survey types (acoustic and mist-netting surveys) and survey duration were investigated as a pilot project to: (1) compare nai¨ve site occupancy rates with estimates adjusted because all species are not detected at all sites where they are present; and (2) plan future inventory and monitoring. Models that best fit the resulting data indicated that acoustic monitoring generally does a better job of detecting most bat species compared to mist netting and acoustic surveys outperformed mistnet surveys in the number of species documented per site. The average nai¨ve site occupancy rate as determined from acoustic sampling was 38.2% while the average nai¨ve site occupancy rate as determined from mist-netting totaled 18.0%. Thus, detection probabilities are clearly higher for acoustic sampling methods and allocating resources for equipment and supplies to increase acoustic monitoring efforts is an important next step in monitoring bat species in Montana. Models which best fit the data also indicated that duration of surveys has an important influence on detection of species; although not to the extent of the importance of acoustic sampling. Estimates of recommended minimum or maximum duration of surveys were not a product of this analysis. Nai¨ve site occupancy rates (range 21.2 to 78.8%) were lower than robust estimated occupancy rates (Psi) resulting from multiple surveys of grid cells (33.7 to 100%) for all species for which this comparison could be made. Lower estimates of detection probability or insufficient data for calculation of estimates were associated with a number of species with limited distributional information. Pilot surveys need to be conducted to evaluate baseline levels of site occupancy and detection probability for these and other bat species in Montana not evaluated with this pilot effort. Pilot surveys also need to address how detection probabilities vary with sampling covariates such as type and duration. This pilot survey work will place future inventory and monitoring efforts on a sound base for supporting management decisions and evaluating changes in status. We recommend the USFS Northern Region continue with a grid-based random sampling scheme stratified by ecoregion or Ranger District, with multiple surveys per grid cell allowing for valid inference of grid cell occupancy rates across each sampling stratum. While the Oregon based 10 km2 grid sampling protocol may be appropriate, other grid systems could be employed to accomplish landscape-scale bat monitoring. A bat sampling grid based upon the latilong concept would fit well with other current and historical wildlife distribution studies in Montana and would greatly simplify implementation of the sampling because 1:24,000 scale quadrangle maps fit within this scheme and could be used directly as the sampling unit. It is important to note, however, that the detection analysis shows strong support for a grid scale smaller than either the Oregon bat scheme or the latilong scheme so that a greater number of sample units could be surveyed with multiple surveys. Further investigation of the appropriate sampling unit and sampling scheme is still needed. However, a grid-based sampling scheme is an important monitoring approach that should be considered beyond USFS lands and coordinated with other partner agencies and organizations to guide effective bat management across the state. Up-to-date distribution maps for Montanas species can be queried and viewed with a variety of map layers on the Montana Natural Heritage Programs Tracker website at: http://mtnhp.org/Tracker.