Environmental Effects on the Echolocation Call Structure of Bats

Christopher Ryan Allen, Dr. Scott Burt, Dr. Jason Miller

Truman State University NSF-STEP Program, 2007

Introduction-

Order Chiroptera (bats) contains approximately 25% of the total known mammal species with over 1,000 documented species worldwide (Broders et al 2004; Fenton 1983; Findley 1993). The vast majority of bats use echolocation to obtain information about their surroundings and in finding prey (Kalko and Schnitzler 1989; Obrist 1995; Robbins et al 2001). The development of ultrasonic detectors and recorders has provided great potential to further the study of the echolocation of bats. Though these detectors have been in use for over two decades (Broders et al. 2004), a standardized, quantitative identification system for distinguishing between individual species has not been developed. The use of this non-intrusive tool to quantitatively identify individual species could provide a wealth of information such as habitat range, population size and distribution, habitat use, social patterns, foraging habits, etc., all of which are key basis for wildlife management and conservation (Betts 1998).

Echolocation calls of bats have traditionally been put into three distinct categories: search-phase (detection calls), approach-phase (pursuit of prey), and terminal-phase calls (feeding buzzes just prior to capture) (Griffin et al. 1960; Kalko and Schnitzler 1989; Robbins et al. 2001). Of these calls, search-phase calls are best suited for acoustical identification of bats. They are the most common call and the one most encountered in the field. Search-phase calls also exhibit some consistency throughout a call sequence (several calls recorded from one bat) and have been shown to have species-specific characteristics (Ahlen 1981; Fenton and Bell 1981; O’Farrell et al. 1999; Robbins et al. 2001). Although search-phase calls appear to be the clearest indicator to distinguish amongst species, the maturation of a quantitative identification system has been hindered by intraspecific differences in call structure as well as a close relationship in the call structures of different species within some genera (i.e. Myotis).

Plasticity of call structure within a species is thought to be necessary for a bat to be able to efficiently navigate under a range of environmental and clutter conditions (Boonman and Jones 2002; Broders et al. 2004; Kalko and Schnitzler 1993; Miller and Treat 1993; Obrist 1995; Siemer et al. 2001). Though many studies have been conducted to show that there is indeed intraspecific differences in search phase call structures due to factors such as age, sex, location, habitat, and social context (Betts 1998; Obrist 1995), few have attempted to identify the effects of structural clutter, air temperature, wind speed, relative humidity, and barometric pressure on the call structure of a species. This study was conducted in order to further increase knowledge of the effects of environmental variables that could significantly advance the ability to quantitatively identify free flying bats. It was also part of a larger ongoing project to quantitatively identify all species of bats present in Missouri through the use of echolocation calls. This has become of vital importance as there are two endangered species (Myotis sodalis and M. grisescens) present in the state. The project has run into difficulty accurately distinguishing between species, especially within the genus Myotis. The results of this study continue to amplify the difficulty in achieving the overall goal. The desire to improve the accuracy by which species can be identified by helping to explain the intraspecific variations in call structures and reducing the overlap between species has further been muddled by the bats ability to change its call structure. It is worth mentioning that a concurrent study was done that will address some of the biological differences such as sex, age, and body mass; and we did not operate completely independent of each other.

Methodology

Area of Study

Bat calls were collected on 22 nights in nine locations across Northern Missouri. Sites included Sugar Creek Conservation Area, Adair County; Deer Ridge Conservation Area, Lewis County; Union Ridge Conservation Area, Sullivan and Adair Counties; and several pieces of private land throughout Adair and Sullivan Counties. To obtain passive calls, sites were selected based on known bat activity in the area and the distance from tree line. All recordings in high clutter environments were at least 10 meters from the closest clearing in a forested environment and all recordings in low clutter environments were at least 20 meters from the nearest tree or other structure in open fields. To obtain active calls from hand released bats, mists nets were deployed perpendicular to a known flyway of bats in order to capture them. These flyways are typically found above creek beds, streams, hiking trails, and open clearings. All captured animals were handled and released in accordance with the State of Missouri collector’s permit and followed the protocols of the American Society of Mammalogists.

Obtaining Active Calls

The active technique of capturing and releasing bats was used for several reasons; to identify the types of species present in the area, to have a known call library in order to reference the passive calls to, and to assist in two concurrent studies dealing with both the biological differences in call structure and the endangered M. sodalis. Once free-flying bats were captured pertinent data about the specimen was recorded, including species, sex, weight, and relative age. Then they were transported to a predetermined release point in cloth bird bags. In general, the release point was a clearing at least 50 meters in diameter, not more than 0.5km from capture site, and pre-sampled for minimal bat activity, thus reducing the likelihood of recording unknown free-flying individuals (Robbins et al. 2001). At the time of release, a bat was placed on the out stretched hands of an investigator and allowed to fly away on its own prerogative. All persons present remained quiet in order to minimize background noise (Broders et al. 2004). One to two individuals used a spotlight (Clulite, Cluson Engineering Limited, Hampshire, England) to follow the flight path of the bat. Another individual recorded the calls using either an Anabat II with a Zero Crossing Analysis Interface (ZCAIM) attached or an Anabat SD1 with a built in ZCAIM (Titley Electronics, Ballina NSW, Australia). For the purposes of this study all Anabats were set to a division ratio of sixteen. Call sequences were then stored on a compact flash card, sent to a laptop computer, or sent to a PDA (HP, IPAQ, Pocket PC) running Analook or Anapocket software respectively. Upon our return to the lab, all data was backed up to the hard drive and a Truman State University Network folder for later analysis. None of the active calls obtained were used in this study except as a reference point for selecting calls qualitatively.

Obtaining Passive Calls

Passive calls were obtained using the identical equipment used in obtaining active calls, except that the Anabat microphone was placed on a table approximately 30 minutes prior to sunset and orientated vertically. The information initially was stored on a compact flash card and collected at the end of each night (approximately 02:30 CST). Upon return to the lab, the location was judged for its effectiveness in obtaining calls by analyzing both the quantity and quality of the calls.

Obtaining weather information

Weather data was collected and stored in the field using Kestrel 4000 and 4500 pocket weather trackers (Nielsen-Kellerman, Boothwyn, PA) attached to the top of a 10 ft. pole at the passive call stations. The Kestrel units were set to record data every 20 seconds and the time-tag was used to match the data to the correct call file. Data recorded was: Air temperature, Wind Speed, Barometric Pressure, Altitude, Relative Humidity, Density Altitude, Wet Bulb temperature, Heat Index, Wind Chill, and Dew Point.

Selecting and Cleaning Calls

Of the passive calls obtained three species were used for the comparisons in this study. They include Eptesicus fuscus, Lasiurus borealis, and Lasiurus cinereus and were chosen for their easily identifiable qualitative call structures. To eliminate some of the subjectivity in qualitatively choosing calls, three separate individuals chose calls all of whom have extensive knowledge in analyzing echolocation calls. Only the calls agreed upon by all three individuals were used. In addition, the following parameters were used in the selection process with at least one chirp in a call sequence needing to meet all guidelines:

Lasiurus cinereus -- Fmin < 23kHz or Fmin < 30kHz, call duration > 4.5ms,

Fmax < 50kHz, and some portion of call exhibits a flat slope or constant frequency component.

Lasiurus borealis – Fmin between 35kHz and 45kHz, Fmin is variable, Fmin contains a

slight to heavy hook or constant frequency component.

Eptesicus Fuscus – Fmin between 25kHz and 30kHz, Fmin is consistent, call duration >

4.5ms, Fmin may contain a slight hook but is not a constant frequency.

After selection, the calls were cleaned using Analook 4.9j run in Dos. The Analook Software is able to identify specific characteristics of the call structure including; call duration (ms), maximum frequency (kHz), minimum frequency (kHz), mean frequency (kHz), frequency of the knee (frequency at the start of the flattest or least sloped portion of the call; kHz), characteristic frequency (frequency at the end of the knee; kHz), duration of call to the knee (ms), duration of call to the characteristic (ms), initial slope (octaves/s), and characteristic slope (octaves/s; refer to O’Farrell et al. 2000 for complete definitions). Initially, the “Z key” was used, a built in feature of the software, and then the sequences were lightly cleaned using the “mark off dots” option. All portions not deemed to be part of a search phase call were deleted. After extracting the data from cleaned calls into spreadsheet format, the data was sent through and algorithm to eliminate the fragmented call portions. This final form of the data was then matched to the environmental conditions present within 10 seconds of when the call was recorded.

Results

Table 1 lists the total number of species used in the analysis and the corresponding clutter condition in which those species were recorded. Although the recorders were in place in each environment for equal time, of the 267 individuals whose calls were chosen, 88% were recorded in open environments. The most likely reason for this is that bats prefer to fly in open areas in order to forage and conserve energy. In addition, atmospheric attenuation may be greater in the high cluttered environments, as well as background noise from other animals and/or insects present, causing the calls to be extremely fragmented. Nevertheless, a larger sample size should be obtained from the high cluttered environment in order to have a greater understanding of the effects of this variable on call structure. It should be suggested that an individual wishing to reproduce this study may want to double or even triple the amount of recording hours in a high cluttered environment in order to obtain equal sample sizes.

Using Spearman’s rho (table 2), non-parametric test for correlation, both the L. cinereus and L. borealis showed significant correlations between the call parameters and low/high clutter environments, while the E. fuscus did not. In all three species the general trend was for clutter to have the greatest effect on the frequency components, then slope components, with the least effect on the time components of calls. The fact that there is little correlation present when comparing call duration and environmental clutter is significant in itself and is not in line with previous findings in other manuscripts (i.e. Broders et al. 2004).

In fig. 1 there are also many general trends that are not consistent with others findings, again begging the question if the sample sizes are too small in both this study and the ones previously conducted with similar sample sizes, or is the variation exclusively due to factors other than clutter condition. Results from a fully nested analysis of variance show that when comparing individual bats and clutter condition to call parameters the clutter condition at most is responsible for up to approximately 8% of the variation while the individual can be responsible for up to approximately 88%, both in the case of Fmin. The same test run comparing overall species and clutter condition to call parameters shows that the clutter environment at most is responsible for up to approximately 5% of the variation in the case of Fmax and the species is responsible for as much as 95% of the variation in the case of Fmax. This shows that much of the variation in call structure is still primarily due to the bat itself.

Table 3 shows the correlations between call parameters and temperature, humidity, and barometric pressure. Despite significant levels of correlation throughout (p< 0.05) there seems to be no discernable pattern amongst the three species. The highest reported value is barometric pressure vs. Fmin in the E. Fuscus of 0.581 with a p-value

< 0.0001. However the same comparison done with L. borealis yields a correlation coefficient of -0.033 and a p-value of 0.302. These types of anomalies continue to occur throughout the data set. Possible explanations include, but are not limited to; other factors causing call variation such as age, body mass, sex, geographic location, social patterns, distance from equipment, atmospheric attenuation, etc.; different species having varied reactions to the same environments, lack of a sufficient sample sizes, or no actual correlations between call structure and weather exists, but are merely a result of chance.

The actual effect that the environment plays on the echolocation call structure of Northern Missouri bats is still unknown at this time. However, most environmental variables were shown to have significance on corresponding call variables, suggesting that some relationship does exists. Further studies will need to be conducted amongst very large data sets, all the while trying to eliminate some of the other possible reasons for variation. This study has shown that not all species of bats are the same and though other studies have been conducted that suggest certain results; other studies may show the opposing results using different test species. In the end, it is probably the individual bat that determines its own call structure and all of the other variables play only minor supporting roles.

Acknowledgements

I would like to thank the following people and organizations for making this research possible; National Science Foundation, Truman State University and the STEP program, all of the individuals that put time and effort into the field work, and special thanks to Dr. Scott Burt, Dr. Jason Miller, Ben Hale, Josh Kelly, and Phil Vance.

Tables

Table 1. – Number of individuals from each species and the related clutter environment in which the calls were recorded. Numbers in parentheses represent the total number of pulses from within each category.

Table 2. – Correlations between high/low clutter environments and measured call parameters.

Figure 1. – Minimum frequency, maximum frequency, and call duration of the three different species in both low and high clutter environments. The bulls-eyes represent mean values, the boxes represent quartiles, the horizontal lines represent median values, and the asterisks represent residuals.

Table 3 – Spearman’s rho correlations of call parameters and the weather variables; temperature, relative humidity, and barometric pressure.

Literature Cited

Ahlen, I. 1981. Identification of Scandinavian bats by their sounds. Swedish University of Agricultural Sciences, Department of Wildlife Ecology Report 6.

Betts B. J. 1998. Effects of interindividual variation in echolocation calls on identification of big brown and silver-haired bats. Journal of Wildlife Management.    62:1003–1010.

Boonman A., G. Jones. 2002. Intensity control during target approach in echolocating bats; stereotypical sensori-motor behaviour in Daubenton's bats, (Myotis daubentonii) Journal of Experimental Biology. 205:2865–2874.

Broders, Hugh G., C. Scott Findlay, and Ligang Zheng. 2001. Effects of Clutter on Echolocation Call Structure of Myotis septentrionalis and M. lucifugus.          Journal of Mammalogy, Vol. 85, No. 2. (Apr., 2004), pp. 273-281.

Fenton, M.B., and G. P. Bell. 1981. Recognition of species of insectivorous bats by their echolocation calls. Journal of Mammalogy 62:233-243.

Fenton M. B. 1983. Just bats. University of Toronto Press, Toronto, Ontario, Canada.

Fenton M. B. 2000. Choosing the ‘correct’ bat detector. Acta Chiropterologica. 2:215–224.

Fenton M. B., S. Bouchard, M. J. Vonhof, J. Zigouris. 2001. Time-expansion and zero-crossing period meter systems present significantly different views of echolocation calls of bats. Journal of Mammalogy. 82:721–727.

Findley J. S. 1993. Bats: a community perspective. Cambridge University Press, Cambridge, United Kingdom.

Griffin, D. R., F. A. Webster, and C. R. Michael. 1960. The echolocation of flying insects by bats. Animal Behaviour 8:141-154.

Kalko, E. K. V., and H. U. Schnitzler. 1989. The echolocation and hunting behavior of Daubenton’s bat, Myotis daubentoni. Behavioral Ecology an Sociobiology 24:225-238.

Kalko E. K. V., H.-U. Schnitzler. 1993. Plasticity in echolocation signals of European pipistrelle bats in search flight: implications for habitat use and prey detection. Behavioral Ecology and Sociobiology. 33:415–428.

Krusic R. A., M. Yamasaki, C. D. Neefus, P. J. Pekins. 1996. Bat habitat use in White Mountain National Forest. Journal of Wildlife Management. 60:625–631.

Miller L. A., A. E. Treat. 1993. Field recordings of echolocation and social sounds from the gleaning bat, M. septentrionalis Bioacoustics. 5:67–87.

Obrist M. K. 1995. Flexible bat echolocation: the influence of individual, habitat and conspecifics on sonar signal design. Behavioural Ecology and Sociobiology. 36:207–219.

O’Farrell, M. J., B. W. Miller, and W. L. Gannon. 1999. Qualitative identification of free-flying bats using the Anabat detector. Journal of Mammalogy 80:11-23.

O'Farrell M. J., C. Corben, W. L. Gannon. 2000. Geographic variation in the echolocation calls of the hoary bat (Lasiurus cinereus). Acta Chiropterologica. 2:185–196.

Robbins, Lynn W., Kevin L. Murray, and Eric R. Britzke. Variation in Search-Phase Calls of Bats. Journal of Mammalogy, Vol. 82, No. 3 (Aug., 2001), pp. 728- 737.

Siemers B. M., E. K. V. Kalko, H.-U. Schnitzler. 2001. Echolocation behavior and signal plasticity in the neotropical bat Myotis nigricans (Schnitz, 1821) (Vespertilionidae): a convergent case with European species of Pispitrellus? Behavioral Ecology and Sociobiology. 50:317–328.