Software Testing and Level of Effort Determination to Inform Indiana Bat and Northern Long-eared Acoustic Survey Guidelines and Pilot Assessment of Activity Correlates Indicative of Gray Bat, Little Brown Bat and Tri-colored Bat Status
Dates
Start Date
2019-11-15
End Date
2022-02-28
Summary
With the onset and advance of whitenose syndrome (WNS) across USFWS Regions 2-6 managers realized that then current Indiana bat presence and absence survey guidelines were likely insufficient due to population declines. In essence, the 2007 mist-netting survey protocols were not sufficient to adequately determine presence or absence of Indiana bat, and later the northern long-eared bat, stemming from the decrease in detection probability. Starting in 2013 with Region 4 as the lead, USFWS Regions 3-5 began working with the USGS Virginia Cooperative Fish and Wildlife Research Unit (Unit) to calculate necessary mist-netting and acoustic levels of efforts (duration and extent) from historic data from across the Indiana bat range. From [...]
Summary
With the onset and advance of whitenose syndrome (WNS) across USFWS Regions 2-6 managers realized that then current Indiana bat presence and absence survey guidelines were likely insufficient due to population declines. In essence, the 2007 mist-netting survey protocols were not sufficient to adequately determine presence or absence of Indiana bat, and later the northern long-eared bat, stemming from the decrease in detection probability. Starting in 2013 with Region 4 as the lead, USFWS Regions 3-5 began working with the USGS Virginia Cooperative Fish and Wildlife Research Unit (Unit) to calculate necessary mist-netting and acoustic levels of efforts (duration and extent) from historic data from across the Indiana bat range. From this, the revised “Rangewide Indiana Bat Survey Guidance” https://www.fws.gov/midwest/endangered/mammals/inba/inbasummersurveyguidance.html was developed. With the addition of an acoustic survey method, USFWS also worked with the Unit to develop an automated bat identification software testing protocol, providing initial support to administer and develop testing protocols for acoustic software program approval or disapproval for use with presence/absence surveys. This testing was and is still needed to assess the ability of a software program to identify the presence of focal species of bats at p < 0.1 maximum likelihood estimator (MLE) thresholds. To date, 18 software versions from 3 developers have been tested. Each testing session requires 3-5 days of Unit staff time to process, analyze and interpret for recommendations to the USFWS putting a considerable strain on Unit resources.
Also, owing to the rapid spread of WNS, reliance on both historical mist-netting and acoustic data created a data gap whereby acquisition of a wider array of detection probability values from mist-netting and acoustic surveys across more variable habitat conditions on the Atlantic Coast, Appalachians, Interior Low Plateau and Gulf Coastal Plain was needed to formulate range-wide survey recommendations. Targeted for the Indiana bat, the Unit led acoustic surveys and matched mist-netting at upland, riparian and field edge sites in VA, WV, OH, KY and NY in 2017 to provide data to support the 2018 and 2019 survey guidance. Based on Unit findings on measured reduction in recordable bat activity after successive years of WNS impact, additional and expanded surveys are required to help ensure guidance is based on best available information. With the threatened status of the northern long-eared bat and similar survey guidance needs, additional sites in NC and TN, along with DC, WV and NY where extant maternity colonies of this species occur could be included to develop a similar approach for this species. Conceivably too, portions of these data from KY, GA and VA could form the basis of level-of-effort for the endangered gray bat and portions from coastal VA and NC could provide similar data for the tri-colored bat. Inclusion of Midwestern sites in IN and WI, an additional Northeast site in MA and sites in the Appalachians in MD and VA would provide data sufficient of little brown bats (Myotis lucifugus) metrics. Lastly, building on acoustic data collected by the Unit as well as USFWS, along with potential modifications to proposed level of effort surveys, there is a potential opportunity to move from simple species presence or absence to a more nuanced ability to establish activity thresholds that correlate to low density, non-reproductive individuals versus activity that signifies a high likelihood of maternity activity.
Objectives:
Continue to test automated bat identification software following USFWS and USGS jointly developed protocols and standards as submitted to USFWS over the study period and report findings.
Establish and operate zero-crossing, frequency division survey sites in representative upland, riparian and field-edge habitat in areas with recent past (since advent of WNS) or currently known Indiana bat maternity colonies in KY, TN, VA, IN and NY.
Establish and operate zero-crossing, frequency division survey sites in representative upland, riparian and field-edge habitat in areas with recent past or currently known northern long-eared bat maternity colonies in NC, KY, DC, VA, WV, and NY.
Supplement survey sites in objectives 2 and 3 to provide additional coverage for little brown bats in MA, VA and WI, gray bats in KY, GA and VA and tri-colored bats in KY, NC and VA.
Develop necessary and updated level-of-effort values from acoustic sampling data and VA Unit and USFWS and cooperator-contributed mist-netting data for Indiana bats and northern long-eared bats for use in setting both acoustic and mist-netting presence and absence guideline standards; determine if similar approaches are sufficient to inform guidelines for little brown bat, gray bat and tri-colored bat.
Compare Indiana bat and northern long-eared bat acoustic data from known or strongly suspected maternity colony sites (2015-2019) with current (2020-2021) data from areas in Regions 3, 4, and 5 where combined acoustic and intensive mist-net sampling has determined absence or presence of non-reproductive individuals to delineate hourly, nightly and whole survey period activity levels with the likelihood of maternity activity.
This project was jointly supported by the Midwest (R3), Southeast (R4), and Northeast (R5) Regions. An alternate reference to this project can be found here.