Foraging Patterns of Bats in Southeast Alaska through Acoustic Monitoring Carrie Talus 1 Matt Heavner 1 1 Dept. Natural Sciences, University of Alaska.

Foraging Patterns of Bats in Southeast Alaska through Acoustic Monitoring Carrie Talus1 Matt Heavner1

1Dept. Natural Sciences, University of Alaska Southeast, [email protected], [email protected]

AbstractMuch is still unknown about the five species of bats found in southeast Alaska (Parker 1996), and bioacoustics may be an effective way to learn more. During 2005 and 2006 summers, over 80 hours of acoustic observations of bats were made in southeast Alaska. Using this data, the nightly activity of bats were examined for 20 different sites located in two different areas of southeast Alaska, Prince of Wales Island and Juneau. An Anabat II detector recorded overnight (pre-sunset to post-sunrise) and each site was analyzed for presence of bats, activity patterns, and total activity, using the number of pulses per ten minute period as a measurement of activity.

MethodsThis study was conducted in southeast Alaska during the summers of 2005 and 2006. Study sites were located on northern and central Prince of Wales (POW) island, and in Juneau, Alaska. Southeast Alaska is a part of the Tongass National Forest which is a north-temperate coniferous rainforest. The climate is strongly maritime, with cool summers, mild winters, and abundant precipitation throughout the year.

IntroductionThe most recent and extensive effort to document the occurrence of bats in Alaska confirms that five species, Myotis lucifugus, M. keenii, M. californicus, M. volans, and Lasionycteris noctivagans, are known to be part of the regular fauna of Southeast Alaska (Parker 1996), accounting for about 13% of all regional species of terrestrial mammals. However, much remains uncertain about the geographic range, seasonal distribution, and population size and dynamics of these northern bats.

Acknowledgements

In 2005 recordings took place at 5 sites on Prince of Wales Island. In 2006, recordings took place at a total of 12 sites on POW and 5 sites in Juneau. An Anabat II detector and a laptop running the recording software Raven 1.2 (Cornell Ornithology Laboratory) were used to record bats. Recording began each night at dusk (just after sunset). The software was allowed to run until sunrise or for as long as the battery lasted on the laptop. Data was taken for each site on elevation, karst, forest type, and proximity to water source.

We focused on recording in riparian areas but we are interested in how the neighboring or surrounding ecosystem may affect the bat use of the riparian zone. We looked at riparian areas found within old growth, second growth and clearcut forest types. Because bat activity patterns in riparian areas differ during lactation and post-lactation (Parker et al 1995), for 2005 and 2006 we analyzed only those recordings made during the month of July. Recordings were analyzed using custom software for the presence or absence of bats and nightly activity patterns. We measured the nightly activity patterns as the change in the number of pulses per ten minutes from sunset to sunrise.

Figure 2. Anabat and laptop setup.

Thank you to:

• REU students Allan Hall and Liz Mallot and UAS student Dan Gaston for their work on this project.

• The Alaska Department of Fish and Game, National Science Foundation and University of Alaska Southeast for funding and support.

• Steve Lewis, Julia Boland, Doreen Parker and the US Forest Service for their help and support.

In order to better understand bats in southeast Alaska as a part of improving the Alaska Department of Fish and Game’s non-game management plan, we are developing a remote acoustic monitoring system that will use software designed to recognize bat signals and automatically identify and eliminate background noise (see Heavner poster). Based on analysis from data from two field seasons we find unexpected patterns in bat activity throughout the night.

Bats typically forage during two feeding periods between dusk and dawn in riparian areas, the first feeding period falling before midnight and the second between midnight and dawn (Anthony and Kunz, 1977). It is not known, however, if the bats of southeast Alaska follow these activity patterns. It is likely that the shortened period of darkness during Alaska’s summer may affect these feeding patterns.

On POW, Parker et al. (1995) found activity in riparian areas began just after sunset and in July she found 2 additional activity peaks within 4 hours of sunset and almost no activity the next 3 hours. Riparian habitat has been found to have the highest proportion of bat use compared with other habitat types (Parker et al 1995). This is due to the fact that bats forage along water sources where insects congregate.

Figure 1. Map of study area located in southeast Alaska.

Figure 3. Recording with two Anabats and laptops from bridge.

For each site, data was also taken on elevation, karst, forest age, and proximity to water source. During 2006 insects were collected using a light trap at each location. The analysis of the data will be presented. The efficacy of long duration acoustic monitoring for addressing management issues in southeast Alaska is described.

Photo by Allan Hall

Preliminary Analysis Results

Literature Cited Anthony, E.L.P. and T.H. Kunz. 1977. Feeding strategies of the little brown bat, Anthony, E.L.P. and T.H. Kunz. 1977. Feeding strategies of the little brown bat, Myotis Myotis lucifuguslucifugus, in Southern New Hampshire. , in Southern New Hampshire. EcologyEcology, 58: 775-786., 58: 775-786.

Parker, D.I. 1996. Forest ecology and distribution of bats in Alaska. Thesis for the degree of Parker, D.I. 1996. Forest ecology and distribution of bats in Alaska. Thesis for the degree of Master of Science.Master of Science.

Parker, D. I., J. A. Cook, and S. W. Lewis. 1995. Effects of Timber Harvest on Bat Activity in Southeastern Alaska ﾕ s Temperate Rainforests in R. Barclay and M. Brigham, editors. The Bats and Forests Symposium, October 19-21, 1995, Victoria, BC.

Site Description

Table 1. Shows information on each of the 2005 and 2006 recording locations.

Recordings made in 2005 and 2006 took place at multiple sites in POW and Juneau, Alaska. Most locations were at a water source where bats would be foraging for insects. One exception, for example, is the site “Scallop cave” which is a high elevation cave located in the heart of a clearcut. Recording sites were located in differing forest types: clearcut areas, old growth areas and second growth areas.

Some Juneau sites are recorded in a residential neighborhood where bats were known to roost in man-made structures. Recording locations also include areas with and without karst, and a couple sites at a high elevation. Each of these factors may affect the foraging behavior of bats. Below is the table showing the above information for each site. Rain is also noted since precipitation may affect the availability of insects and thus the presence of bats.

Figure 4: Lower 108 Creek

Figure 5: Clearcut 2

Bat activity patterns have been analyzed for 18 or the 23 recordings sites. Activity patterns vary greatly from location to location and even night to night. Although statistical analysis of this data is in progress, it is clear that the data shows little bimodal pattern of foraging where there is a pulse of bats just after sunset and another pulse of activity within 4 hours of sunset. Out of 17 locations only one, Way Low Zina, shows a bimodal pattern.

Location Date Forest Type KarstAt

Elevation?Proximity to Water Rain?

Maybeso Creek 7/13/05 Second growth No karst NoAnabat aimed over creek Yes, light

Thorne River 7/14/05 Second growth No karst No Over river YesLuck Lake 7/15/05 Second growth No karst No Over lake No

Cavern Lake 7/16/05Old and second growth Yes No Over lake No

Turn Creek 7/18/05 Old growth Yes No Over creek No

Thorne River 7/21/05 Second growth No karst No Over riverNone at setup, light

rain laterThorne River 7/22/05 Second growth No karst No Over river NoTurn Creek 7/24/05 Old growth Yes No Over creek NoTurn Creek 7/25/05 Old growth Yes No Over creek No

Sandy Beach 7/7/06 Old and second growth No karst No On tidal creek Yes, moderate rainLower 108 Creek (Under bridge) 7/8/06 Old growth Yes No On tidal creek

No, some mist at setup

Lower 108 Creek (over bridge) 7/8/06 Old growth Yes No On tidal creek

No, some mist at setup

Cavern Lake 7/9/06 Old and second growth Yes No Over lake NoTurn Creek 7/9/06 Old growth Yes No Over creek No

Flicker Creek 7/10/06Old growth, with clearcut upstream Yes No Over river No

Clearcut 1 7/10/06 Clearcut No No

WayLowZina 7/11/06

Old growth "margin" left around stream in second growth area. Yes No Over stream

Yes, from midnight till sunrise

Scallop Cave 7/11/06Clearcut with some second growth Yes Yes, 440m No water source No

One Duck 7/12/06 Old growth No karst Yes, 350m No water source NoLittle Salt Creek 7/12/06 Clearcut No karst No Over creek Yes, very lightClearcut 2 - A little North of Little Salt Creek 7/13/06 Clearcut No karst No No water source No

Harris Camp 7/13/06 Old growth No karst NoAimed over Harris River No

Lower Jordan Creek, Juneau 7/18/06 Riparian No karst No Over creek NoUpper Jordan Creek, Juneau 7/20/06 Second and old growth No karst No Over creek No

Douglas Residential 1, Juneau 7/22/06 Residential No karst No No water source Yes, light mistMontana Creek, Juneau 7/25/06 Second growth No Karst No Over creek NoMontana Creek, Juneau 7/26/06 Second growth No Karst No Over creek

No rain at setup. Light rain at sunrise

Mark's Residential 2, Juneau 7/31/06 Residential No Karst No No water source Yes, light

Residential 1, Juneau 072206

0

500

1000

1500

2000

2500

3000

0 0.2 0.4 0.6 0.8 1 1.2

Time (0=sunset, 1=sunrise)

Pulses per 10 minutes

Turn Creek 070906

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 0.2 0.4 0.6 0.8 1 1.2

Time (0=sunset, 1=sunrise)

Pulses per 10 minutes

Cavern Lake 070906

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Time (0=sunset, 1=sunrise)

Pulses per 10 minutes

Flicker Creek 071006

0

200

400

600

800

1000

1200

1400

0 0.2 0.4 0.6 0.8 1 1.2

Time (0=sunset, 1=sunrise)

Pulses per 10 minutes

Way Low Zina 071106

0

500

1000

1500

2000

2500

0 0.2 0.4 0.6 0.8 1 1.2

Time (0=sunset, 1=sunrise)

Pulses per 10 minutes

Turn Creek is an old growth area with karst. This is an example of a night where there is one peak of bat activity early in the night and then no activity the remainder of the night.

Cavern Lake is an area with high levels of activity mid night with lower levels close to sunset and sunrise. This is an old and second growth forest with karst. Note that this recording does not last until sunrise.

Flicker Creek is in an old growth forest with clearcut areas just upstream. This night is an example of high pulses of bat activity for the first 3 hours of the night with activity decreasing toward sunrise.

Residential 1 is located in a residential area where bats are known to roost in someone’s house. This is an example of continuous activity from sunset to sunrise.

Below are some examples of individual results of the bat activity for specific sites.

Way Low Zina shows an initial pulse of activity an hour after sunrise with additional pulses 6 or more hours after sunrise. This is an old and second growth area with karst.

The Normalized Cumulative Pulse Count plot graphically summarizes the results of this study to date. In order to overplot all 18 nights of observations, both the horizontal time axis and the vertical pulses-per-ten-minute-bin axis have been normalized. The time axis has been normalized such that 0 represents sunset and 1 represents sunrise. Length of night can change more than five minutes per day at the latitude of southeast Alaska (depending on time of year). The cumulative pulse vs. time plot allows easy visual determination of changes in bat activity rate through the night. For example, the Turn Creek 2006/07/09 plot shows an early evening pulse of activity only. The Mark's 2006/07/31 plot shows continuous and steady bat activity through the night. The plot from Dave's 2006/07/22 shows activity at a steady rate through the night but with a break

from about 55%-80% of the way through the night.

This data is being analyzed for statistical differences in activity pattern between sites. In particular, we will look at differences between sites with different forest type. In this preliminary analysis it is interesting to note that we are not seeing a bimodal pattern of bat activity as seen by others.

Normalized Cumulative Pulse Count

0

0.2

0.4

0.6

0.8

1

1.2

-0.2 0 0.2 0.4 0.6 0.8 1 1.2

Time (0=sunset, 1=sunrise)

Normalized Cumulative Pulse (10 minute bins)

Turn Creek 2006/07/09

Cavern Lake 2005/07/16

Cavern Lake 2006/07/09

Flicker Creek 2006/07/10

Dave's 2006/07/22

Mark's 2006/07/31

Montana Creek 2006/07/25

Montana Creek 2006/07/26

One Duck 2006/07/12

Scallop 2006/07/11

Luck Lake 2005/07/15

Cavern Lake 2005/07/15

Thorne River 2005/07/14

Thorne River 2005/07/21

Turn Creek 2005/07/18

Turn Creek 2005/07/18

Turn Creek 2005/07/25

WayLowZina 2006/07/11