|
Sarah The World Through a Bat's Ears
|
Continued...
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Table 3a. The results of a cross-validated DFA showing the percent of calls identified correctly to setting using the call data I collected. Here, setting 1 is the open situation, setting 2 is vegetation, setting 3 is the mine entrance, and setting 4 is inside the mine. The percentage on the diagonal is the percentage of correct classifications. Overall, the cross-validated DFA correctly assigned calls by setting 75% of the time.
Table 3b. The results of a cross-validated DFA showing the percent of calls identified correctly to setting using data only on Dur and FME. Here, setting 1 is the open situation, setting 2 is vegetation, setting 3 is the mine entrance, and setting 4 is inside the mine. The percentage on the diagonal is the percentage of correct classifications. Overall, the cross-validated DFA correctly assigned calls by setting 55.6% of the time.
In the next DFA (Table 3c), I used all call features and compared open versus closed (cluttered) situations. The DFA correctly associated calls with setting 95.8% of the time. In this analysis, when I just used Dur and FME (Table 3d), calls were correctly identified 97.2% of the time. It is interesting that in both of these analyses (Table 3c and 3d), the cluttered calls were identified 100% of the time, but the calls produced in the open were not. In examining any echolocation call, I do not know exactly where the bat was located at the time of call production. Although my experimental setup was designed to minimize the chances of a bat in the open hearing clutter from vegetation and fixed objects, other bats may also be important sources of clutter. Table 3c. The results of a cross-validated DFA demonstrating the percent of calls correctly identified as either open or closed using all the call features. Settings 2, 3, and 4 are classified as closed (2). The open situation is (1). The percentage on the diagonal is the percentage of correct classifications. The DFA correctly assigned calls 95.8% of the time.
Table 3d. The results of a cross-validated DFA showing the percent of calls correctly identified as either open or closed using only Dur and FME. Settings 2, 3, and 4 are classified as closed (2). The open situation is (1). The percentage on the diagonal is the percentage of correct classifications. The DFA overall correctly assigned calls 97.2% of the time.
|
My results demonstrate that clutter influences the structure of bat echolocation calls. Free-flying M. lucifugus adjusted their calls according to clutter. The results confirm my hypothesis that bats vary the features of their echolocation calls according to the settings in which they are operating. My findings are based on my identification of different forms and degrees of clutter (from an open field to a forest to a mine adit) and my analyses of bats' responses to it, and support earlier suggestions of biologists such as Kalko (1995) and Neuweiler et al (1987), who have proposed a connection between echolocation call design and clutter.
While Siemers and Schnitzler (2004) describe how different species of European Myotis adjusted their echolocation calls when hunting prey in clutter, and suggest that high-frequency signals are a good adaptive strategy for successfully capturing prey in clutter, their bats were hunting in captivity. Mine were in the wild, and were not hunting but rather approaching or entering a mating site (located in the mine). My findings also indicate that increases in clutter for bats operating in the wild (from an open field to a forest to a mine adit) are met by increasing adjustments in bats' echolocation calls. In cluttered settings, in which there are more rapidly returning echoes, the bats shortened the duration of their calls and the inter-pulse interval. M. lucifugus appear to be using shorter calls, as well as higher frequencies, to obtain more information about close objects in cluttered settings. Quantifying bats' responses to varying measures of clutter in the wild presents unique challenges. More detailed conclusions would benefit from analyses of the complexity of vegetation contours or foliage echoes and their impact on bat echolocation. Challenges in the wild also include the need to take adequate account of other bats as sources of clutter. Future research would benefit as well from an increase in the sample size (number of calls and number of sequences). It is clear that bats adjust the design of their echolocation calls according to their situation. Situation-specific calls alert other bats to those conditions in which the bat is operating. Other bats within earshot of the Renfrew site could have used the calls I recorded to locate the entrance to the mine. This helps biologists to understand how bats can find important sites such as mating and hibernation sites. In other settings, biologists have demonstrated that free-flying bats in the wild listen to one another and keep track of what is happening to other bats within earshot. In the future, the effects of clutter on the echolocation calls of other bats in other places could be examined. Larger, narrow-winged, fast-flying bats, such as some free-tailed bats, would be expected to be more wary of clutter and to adjust to it with greater difficulty. Recent suggestions that time of night could influence bat echolocation calls are unexplained and intriguing (Biscardi et al, in press). I would like to experiment with other circumstances, such as the darkness of the setting, which might have an impact on bats' echolocation calls. How bats cope with clutter can have implications for the world beyond bats. Emulating bats, some people have designed sonar aids for the blind. The sonar aid has been successful in allowing users to navigate environments that humans would call cluttered. Kay's Auditory Spatial Perception Aid (KASPA), for example, is a sonic guide in the form of a device worn as a headband, or as miniature earpieces connected to a small unit attached to a cane. Ultrasonic waves are sent in the form of a high-frequency, narrow band of sound, and the echoes are received and transformed into electrical signals that are then processed to produce sounds audible to humans. Just as my results indicate that bats cope with clutter by producing shorter, more frequent calls of higher frequency, this sonar aid technology has controls that change the pulsating tone to higher frequencies of shorter duration for short-range use. Humans can use the same techniques as bats to "see" with their ears. Learning more about how bats handle clutter can provide valuable lessons for further development of sonar aids. Improvements to sonar aids which give the user greater control over the outgoing signal, and refine additional dimensions of the signal, may adjust the information that comes back. An answer for the next time someone calls bats "creepy"! |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||












