Previous studies of the spotted bat Euderma maculatum have demonstrated that this bat emits echolocation calls that are lower in frequency, shorter in duration and fainter in intensity compared with those of most other insectivorous bats, acoustic characteristics which should render it less conspicuous to eared moths. We tested this prediction by monitoring electrophysiologically the ears of sympatric noctuoid (noctuid, arctiid and notodontid) moths in a site in western Canada. Auditory threshold curves demonstrate that most of the moths tested are less responsive to the calls of Eu. maculatum than to those of another sympatric bat, Eptesicus fuscus. Playbacks to moth ears of pre-recorded search-and approach-phase echolocation calls of Eu. maculatum and Ep. fuscus further demonstrate that the calls of Eu. maculatum are poorly detectable to moths and, in some cases, completely inaudible. We estimate that, in the wild, an average noctuoid moth would detect the calls of Eu. maculatum at distances of less than 1 m as opposed to the calls of Ep. fuscus which should be first heard at distances of 20–25 m. Although most moths are unable to adequately hear Eu. maculatum, the observation that two individuals possessed ears sensitive to this bat’s calls suggests the existence of auditory pre-adaptation to this type of echolocation.

Many species of moths defend themselves against the attacks of hunting bats with simple ears that detect the bats’ echolocation calls (Roeder, 1967; Miller, 1983; Fullard, 1987a; Surlykke, 1988). Moth ears are used primarily for bat detection (but see Surlykke and Fullard, 1989) and their frequency sensitivities are syntonic (Fullard, 1987a,b, 1988) with (i.e. matched to) the echolocation characteristics of those bats that comprise their main predatory threat (Fullard, 1982). Although insectivorous bats and their species-specific echolocation calls cover a wide range of frequencies (Fenton and Bell, 1981), moth ears are most sensitive from 20 to 50 kHz, the bandwidth used by most insectivorous bats (e.g. Roeder, 1970; Fenton and Fullard, 1979; Fullard, 1979, 1982, 1984a,b, 1987a; Fullard and Barclay, 1980; Fullard and Thomas, 1981; Fullard et al. 1983; Fullard and Belwood, 1988; Surlykke and Fullard, 1989; Faure et al. 1990, 1993). To frequencies below and above this range (allotonic frequencies) most moths are relatively insensitive and although bats could theoretically increase their foraging success on eared moths by evolving allotonic echolocation (Fenton and Fullard, 1979; Fullard, 1990; Jones, 1992; Rydell et al. 1995), there are costs associated with using such calls. The severe atmospheric attenuation of very high frequencies and the poor target resolution in the echoes of very low frequencies (Griffin, 1971; Lawrence and Simmons, 1982) place limitations on the acoustic efficacy of allotonic echolocation. Nevertheless, some bats do use allotonic frequencies as the main component of their echolocation calls. Certain old-world hipposiderid bats emit very high frequencies (e.g. Cloeotis percivali, 212 kHz; Fenton and Bell, 1981) and feed heavily upon moths (Whitaker and Black, 1976) suggesting a benefit to allotonic echolocation. The North American vespertilionids, Myotis evotis and M. septentrionalis also emit high-frequency calls that sympatric noctuid moths (Catocala spp.) cannot detect as well as those of other bats (Faure et al. 1990, 1993) and it has been argued that this allows them to feed heavily upon moths. There is, however, another reason for using very high frequencies. Bats hunting in dense vegetation experience problems with acoustic ‘clutter’ (Neuweiler, 1990) resulting from the multiple echoes returning from such environments, and the use of very high frequencies improves the resolving power of the echoes for these bats. This acoustic consideration provides an alternative explanation for the use of allotonically high-frequency echolocation rather than as an evolved response to the auditory defences of moths (Rydell et al. 1995).

To date, few studies have examined the sensory ecology of eared moths and allotonic low-frequency bat species. Certain bats use very low frequencies for their echolocation calls (e.g. the molossids Otomops martiensseni: 13 kHz, Fenton and Bell, 1981; Tadarida australis: 12.6 kHz, Fullard et al. 1991 and T. teniotis: 11 kHz, Zbinden and Zingg, 1986) and some appear to feed heavily on moths (Jones and Rydell, 1994). The spotted bat Euderma maculatum is a western North American vespertilionid that echolocates using an extremely low (for bats) dominant frequency of 9–12 kHz (Woodsworth et al. 1981; Fenton and Bell, 1981; Leonard and Fenton, 1984; Obrist, 1995) and has been reported to prey heavily on moths (Easterla, 1965, Ross, 1967; Poché, 1981; Wai-Ping and Fenton, 1989). Woodsworth et al. (1981) and Barclay and Brigham (1991) hypothesized that Eu. maculatum is acoustically inconspicuous to eared moths and the purpose of our study was to test this to examine the possibility that allotonic echolocation is adaptive as an acoustic countermanoeuvre against the auditory defences of moths.

Study area and animals

We conducted this study in the southern Okanagan Valley near Oliver, British Columbia, Canada (42° 12′N, 119° 33′W). We collected eared moths (Noctuoidea: Noctuidae, Arctiidae and Notodontidae) using ultraviolet and mercury-vapour lamps erected in front of white cotton sheets facing ‘open’ (Fenton, 1990) habitats (fields and hillsides). These collection sites were chosen to select for moths that fly in the flight zones used by Eu. maculatum (Leonard and Fenton, 1983; Wai-Ping and Fenton, 1989). Examples of the moths used in our study were subsequently deposited in the Canadian National Collection (Ottawa). Echolocation signals from wild sympatric bats were recorded as they foraged near the Okanagan Falls provincial campground (Ep. fuscus) and on the banks of the Water Dog lake on the property of the Osoyoos Indian Band (Eu. maculatum). Auditory examinations were performed in the field at the Department of Geology station, University of British Columbia and subsequent data analyses including acoustic analyses of the bat calls were performed at the laboratories of the Queen’s University Biological Station, Chaffey’s Lock, Ontario, Canada.

Auditory sensitivity examinations

Auditory examinations were similar to those described in Fullard (1984a) and Faure et al. (1993) and are described briefly below. Moths collected each night were used the following day. After exposure of the moth’s auditory nerve (IIIN1b; Nüesch, 1957) we used stainless-steel hook electrodes to record extracellularly the responses of the two auditory receptor cells (A1 and A2). The stimuli were 10 ms tone pulses (1–30 kHz in 1 kHz increments) generated by a Wavetek function generator (model 23), shaped to a 1 ms rise/fall time (Coulbourn S84–04), amplified (National Semiconductor LM1875T) and broadcast at 1 s−1 from a low-frequency (InterTan 40–8428) (for stimulus frequencies of 1–5 kHz) or a high-frequency (Technics EAS-10TH400B) (for stimulus frequencies of 5–30 kHz) loudspeaker mounted 30 cm from the moth (total system frequency response ±3 dB, 1–40 kHz). The moth was oriented in a ‘normal’, wings-extended, dorsal-side-up position for these exposures and the loudspeaker was directed towards the posterio-lateral margin of the moth (i.e. its most sensitive region; Payne et al. 1966). Thresholds were determined as those intensities that elicited three action potentials per stimulus pulse for four consecutive pulses (Faure et al. 1993). Intensities were recorded as millivolts peak-to-peak and were later converted to peak equivalent sound pressure levels (peSPL; Stapells et al. 1982) from equal-amplitude continual tones using a Brüel and Kjær (B&K) type 4135 1/4 inch microphone and type 2610 B&K measuring amplifier following calibration with a B&K type 4228 pistonphone. The playback system, including calibration equations for converting voltages into decibels, was controlled by a customized MS-DOS program written by J. W. D. A stimulus sound duration of 10 ms was chosen for these tests to obtain an estimate of the moths’ best hearing abilities since this duration is in excess of the integration time of noctuid moth ears (Surlykke et al. 1988). Bats use a wide range of echolocation call durations (Novick, 1977; Fenton and Bell, 1981) so any particular stimulus duration will not match all those naturally encountered by flying moths.

Echolocation playback trials

Echolocation calls were recorded (see below) from wild bats as they foraged for insects over streams or lakes during the same period that the auditory examinations were performed. These recorded calls were played from a RACAL Store 4DS tape recorder (cf. Waters and Jones, 1996) running at 76 cm s−1, amplified and broadcast from a Technics EAS-10TH400B loudspeaker mounted 30 cm from the moth auditory preparations (the low-frequency cut-off of this loudspeaker filtered out most of the recorded background noise). We recorded the moth’s auditory receptor responses and the echolocation pulses onto separate channels of another RACAL tape recorder. One fly-by echolocation sequence from each bat (‘bat pass’; Fenton, 1970) was chosen as the playback stimulus and played back at five amplification levels to provide a range of echolocation pulse intensities. Avoidance flight in moths commences in response to searching or cruising bats (Roeder, 1967) so we selected bat passes for playbacks that exhibited these phases and did not culminate in a terminal echolocation sequence. The voltages of the pulses were later converted to dB peSPL values by broadcasting a 10 kHz (for Eu. maculatum) or 30 kHz (for Ep. fuscus) continuous tone from the same loudspeaker while measuring intensities with the B&K measuring equipment described above.

The amplified bat echolocation pulses and corresponding auditory receptor spikes were digitized and stored using MS-DOS graphics analysis software (AxoTape 2.02, Axon Instruments, Foster City, CA, USA) by J. H. F. and later measured by a technician who was unaware of the acoustic characteristics of the two bat species chosen for the playbacks.

Bat recordings and analyses

Foraging bats (see above) were recorded using a Larson–Davis type 2520 1/4 inch microphone preamplified by a type 2200C power supply connected to a RACAL Store 4DS tape recorder running at 76 cm s−1. Calls were later examined using a customized Fast Fourier Transform (FFT) spectral analysis program written by J. W. D. (see Faure et al. 1993 for details). The following variables were measured: duration (ms), dominant frequency (the value, in kHz, of the maximum spectral peak), lowest and highest frequencies (the kHz values of the spectrum −18 dB below and above dominant frequency, respectively) and the spectral bandwidths at −6 and −18 dB below dominant frequency.

Auditory analyses

Auditory sensitivity curves (audiograms) were derived for the following 18 noctuoid moths caught in the field: Apamea antennata (Noctuidae), N=2; Leuconycta lepidula (Noctuidae), N=2; Hyphantria cunea (Arctiidae), N=3; Clostera albosigma (Noctuidae), N=3; Gluphisia septentrionalis (Noctuidae), N=6; Nadata gibbosa (Notodontidae), N=2. Fig. 1 illustrates the combined audiograms of these moths. For the present study, we chose frequencies relevant to the bats used in the playback experiments and exposed the moths to 5–30 kHz (for audiograms of British Columbian moths that extend beyond 30 kHz, see Fullard et al. 1983). The dashed lines in Fig. 1 indicate 12.4 and 27.8 kHz, the dominant frequencies emitted by Eu. maculatum and Ep. fuscus, respectively. Low-frequency thresholds in these moths were generally high, ranging (at 10 kHz) from 55 to 105 dB with most thresholds exceeding 70 dB. All but one moth had higher thresholds for the frequencies of Eu. maculatum (12 kHz) than for those of Ep. fuscus (28 kHz) and there was an average threshold difference across all of the moths of 27 dB (range 2–44 dB). One individual of H. cunea did show a lower threshold to 12 kHz than to 28 kHz but there was no evidence that these ears were most sensitive to (i.e. tuned to) the frequencies emitted by Eu. maculatum. Fig. 1 further illustrates that inter-individual auditory thresholds for most species are variable, particularly at low (allotonic) frequencies. Threshold variability, especially at frequencies outside the best frequency, is commonly observed in moths (Surlykke, 1986; Faure et al. 1993; Fullard, 1994) and does not appear to be an artifact of experimental conditions.

Fig. 1.

Audiograms of British Columbian moths captured for this study. For frequencies where the maximum loudspeaker output could not obtain a threshold value, an arbitrary value of 110 dB is assigned; for frequencies where the lowest output of the loudspeaker still elicited receptor activity, an arbitrary value of 30 dB is assigned. For each audiogram, the approximate frequencies of the searching echolocation calls of Eu. maculatum (12.4 kHz) and Ep. fuscus (27.8 kHz) are indicated by dashed lines.

Fig. 1.

Audiograms of British Columbian moths captured for this study. For frequencies where the maximum loudspeaker output could not obtain a threshold value, an arbitrary value of 110 dB is assigned; for frequencies where the lowest output of the loudspeaker still elicited receptor activity, an arbitrary value of 30 dB is assigned. For each audiogram, the approximate frequencies of the searching echolocation calls of Eu. maculatum (12.4 kHz) and Ep. fuscus (27.8 kHz) are indicated by dashed lines.

calls of Eu. maculatum will result in lower dB SPL levels for this species compared to Ep. fuscus, resulting in a conservative estimate of the difference between the moths’ responsiveness to the two bats (see Discussion). Fig. 2 illustrates that the auditory receptors in both moths respond to the calls of Ep. fuscus more vigorously than to those of Eu. maculatum and with a difference as to which receptor neuron is activated. Whereas both A1 and A2 receptors respond to the calls of Ep. fuscus, the A2 cell responds only rarely to those of Eu. maculatum. The results illustrated in Fig. 2 are typical of those observed for most of the moths investigated, although some individuals (e.g. Lacinipolia illaudabilis) revealed only single receptor spikes to even the most intense of the calls of Eu. maculatum and these were never those of the A2 cell. The results from all the moths’ receptor responses to the bat pulses in the playback experiments are shown in Fig. 3. For all moths tested, the auditory responses (calculated as the total number of A1 and A2 receptor cell spikes) were less to Eu. maculatum (mean number of spikes per pulse was 2.4) than to Ep. fuscus (mean of 7.3 spikes per pulse). We caution that the points in Fig. 3 arise from repeated measurements from single individuals and are therefore pseudoreplicates. Although quantitative statistical comparisons are not possible for these data, we contend that the differences between the responses to the two bats are obvious with the calls of Eu. maculatum eliciting fewer spikes than those of Ep. fuscus at all intensities used in the playbacks. In addition, the narrower bandwidths of the calls of Eu. maculatum (see below) will result in lower total energy levels per call, further reducing their conspicuousness to the moths.

Fig. 2.

Representative oscillograms (upper traces in each set) of the auditory nerve responses of an insensitive moth, Nadata gibbosa and a sensitive moth, Euxoa flavicollis to the echolocation calls (lower traces) of the two bat species (Ep. fuscus and Eu. maculatum) used in the playback trials. In both illustrations the most common spike is that of the A1 receptor while the less-sensitive A2 receptor (indicated by asterisks) responds more commonly to the calls of Ep. fuscus than Eu. maculatum.

Fig. 2.

Representative oscillograms (upper traces in each set) of the auditory nerve responses of an insensitive moth, Nadata gibbosa and a sensitive moth, Euxoa flavicollis to the echolocation calls (lower traces) of the two bat species (Ep. fuscus and Eu. maculatum) used in the playback trials. In both illustrations the most common spike is that of the A1 receptor while the less-sensitive A2 receptor (indicated by asterisks) responds more commonly to the calls of Ep. fuscus than Eu. maculatum.

Fig. 3.

Auditory receptor responses of the moths used in the playback trials to varying intensities of echolocation pulses from the two bat species (Ep. fuscus and Eu. maculatum) used in the playback trials. Each moth (N=1) was exposed to a number of echolocation pulses whose peSPL was determined from the recorded voltages. The shaded zone from 0 to 1 spike per bat pulse indicates spontaneous firing activity.

Fig. 3.

Auditory receptor responses of the moths used in the playback trials to varying intensities of echolocation pulses from the two bat species (Ep. fuscus and Eu. maculatum) used in the playback trials. Each moth (N=1) was exposed to a number of echolocation pulses whose peSPL was determined from the recorded voltages. The shaded zone from 0 to 1 spike per bat pulse indicates spontaneous firing activity.

Noctuid moth auditory receptors spontaneously fire in the absence of sounds, a random activity that presumably signifies a ‘no-bat’ condition (Roeder, 1967). We assume that, in the absence of other sensory cues to the presence of an approaching bat, a moth cannot distinguish between a single receptor spike arising spontaneously and one elicited by a faint echolocation pulse. We therefore tabulated the receptor responses consisting of one spike per bat pulse as a proportion of the total response to each bat. The results (41.4 % of Eu. maculatum’s pulses versus 10.1 % of Ep. fuscus’ pulses) indicate that a greater proportion of Eu. maculatum’s pulses elicited responses equal to spontaneous firing levels, and were presumably treated as ‘no bat’, than did those of Ep. fuscus.

Bat echolocation analyses

Fig. 4 and Table 1 describe the characteristics of the echolocation pulses used in the playback experiments with moth auditory preparations. We caution that, in both cases, only the approach sequence of one bat was used and that these data are not intended to provide species-specific characteristics (for a detailed treatment of the acoustic characteristics of Eu. maculatum see Obrist, 1995). The calls of Eu. maculatum are human-audible with a fundamental harmonic of 12.4 kHz and with a second harmonic of 25–30 kHz, 10–20 dB less than the fundamental. In contrast, the calls of Ep. fuscus possess a

Table 1.

Echolocation call characteristics of the Euderma maculatum and Eptesicus fuscus passes used in the playback experiments

Echolocation call characteristics of the Euderma maculatum and Eptesicus fuscus passes used in the playback experiments
Echolocation call characteristics of the Euderma maculatum and Eptesicus fuscus passes used in the playback experiments
Fig. 4.

Acoustic analyses of the echolocation pulses of the bats used in the playback trials. For both species a time/amplitude oscillogram (inset) is shown with its Fast Fourier Transform. The following call characteristics (values given in Table 1) are defined in the upper panel: BW−6dB, bandwidth (−6 dB from spectral peak); BW−18dB, bandwidth (−18 dB from spectral peak); DF, dominant frequency (spectral maximum); HF, high frequency (−18 dB from spectral peak); LF, low frequency (−18 dB from spectral peak).

Fig. 4.

Acoustic analyses of the echolocation pulses of the bats used in the playback trials. For both species a time/amplitude oscillogram (inset) is shown with its Fast Fourier Transform. The following call characteristics (values given in Table 1) are defined in the upper panel: BW−6dB, bandwidth (−6 dB from spectral peak); BW−18dB, bandwidth (−18 dB from spectral peak); DF, dominant frequency (spectral maximum); HF, high frequency (−18 dB from spectral peak); LF, low frequency (−18 dB from spectral peak).

Playbacks

Five moths, one individual each of Halysidota maculatum (Arctiidae), Nadata gibbosa (Notodontidae), Euxoa flavicollis (Noctuidae), Lacinipolia illaudabilis (Noctuidae) and Zotheca tranquilla (Noctuidae), were exposed to the prerecorded echolocation calls of Eu. maculatum and Ep. fuscus. Fig. 2 illustrates the auditory nerve responses of an insensitive (N. gibbosa) and a sensitive (E. flavicollis) moth to the echolocation pulses of the two bats; the shorter duration of the fundamental dominant frequency of 27.8 kHz, also with upper harmonics. Furthermore, the spectral bandwidths at −6 and −18 dB of Eu. maculatum are narrower, with all frequencies lower than those of the calls of Ep. fuscus. The durations of the calls of the two species also differ, with those of Eu. maculatum being shorter (4.2 ms) than those of Ep. fuscus (9.8 ms). While our results reflect those reported for Ep. fuscus in the same region, our values for dominant frequency in Eu. maculatum were higher than those previously reported (Fenton and Bell, 1981; Leonard and Fenton, 1984; Obrist, 1995).

Results from the auditory analyses and playback experiments in the present study demonstrate that moths are less responsive to the calls of Euderma maculatum than to those of Eptesicus fuscus, an insensitivity that appears to arise from a combination of the very low frequency and short duration of the echolocation calls of Eu. maculatum. Threshold versus stimulus-pulse-length relationships for other noctuoid moths (Surlykke et al. 1988) suggest that the short calls of Eu. maculatum fall below the integration time of these ears (evoking a response equal to a longer pulse approximately 6 dB less intense) while the longer calls of Ep. fuscus will evoke the maximum response possible. The short durations of the calls of Eu. maculatum suggest that our use of a 10 ms stimulus tone resulted in higher thresholds than would be elicited by the calls themselves and the natural ability of moths to hear Eu. maculatum approach would be even less than we predict. Woodsworth et al. (1981) estimated the emitted natural intensities of the calls of Eu. maculatum to be 80–90 dB SPL at 10 cm as extrapolated from measured intensities at known distances, values considerably less than the 100–110 dB SPL at 10 cm computed for other vespertilionid bats (Griffin, 1958). Using estimates of moth auditory integration times from Surlykke et al. (1988), the calls of Eu. maculatum would be treated as if they were 6 dB less than predictions from Woodworth et al. (1981), rendering the calls even less detectable to moths.

It is possible to estimate a moth ear’s maximum detection distance for a particular bat if the moth’s auditory thresholds to the dominant echolocation call frequencies of the bat are known (Fenton and Fullard, 1979; Surlykke, 1988). Fig. 5 plots the mean maximum detection distances predicted from the moths’ thresholds (Fig. 1) to 10 kHz (i.e. Eu. maculatum) and 30 kHz (i.e. Ep. fuscus) against a range of emitted echolocation intensities (the estimated echolocation intensities for the two bats are indicated by arrows and, in the case of Eu. maculatum, two predicted dB values are given, one with and one without the correction factor from the integration time results of Surlykke et al. (1988). The maximum detection distances predict that whereas Ep. fuscus will be detected at 20–25 m, Eu. maculatum will not be detected until it is less than 1 m from the moths. Assuming that bats must be within 1–5 m of their intended targets (Kick and Simmons, 1984; Kalko, 1995) before they can detect their echoes (indicated by the dashed line in Fig. 5), only Eu. maculatum appears to be able to achieve this distance before moths have heard its calls. Whereas most moths we tested would be subject to increased predation from Eu. maculatum (certain species, such as Nadata gibbosa appear to be completely deaf to it), others (e.g. Euxoa flavicollis) have ears that should be able to hear this bat in time to effect a ‘take-cover’ (Roeder, 1967) evasive response. Leonard and Fenton (1983) observed Eu. maculatum diving to within 1 m of the ground in pursuit of unidentified insects. If the insects in these encounters were moths, it would suggest that some possess ears that are at least capable of evoking take-cover defensive responses to attacking Eu. maculatum.

Fig. 5.

Predicted maximum detection distances for the two species of bat recorded for all of the moths whose audiograms were derived (Fig. 1). The proposed in-flight intensities for both bats [Eu. maculatum: 85 dB SPL (Woodsworth et al. 1981) and 79 dB (as attenuated by the moth’s integration time; Surlykke et al. 1988); Ep. fuscus: 105 dB (Griffin, 1958)] are indicated by arrows. The dashed line indicates 5 m, the point at which both bats may first detect echoes returning from the moths. The moths detect Ep. fuscus at a distance of 20–25 m, Eu. maculatum will not be detected until less than 1 m.

Fig. 5.

Predicted maximum detection distances for the two species of bat recorded for all of the moths whose audiograms were derived (Fig. 1). The proposed in-flight intensities for both bats [Eu. maculatum: 85 dB SPL (Woodsworth et al. 1981) and 79 dB (as attenuated by the moth’s integration time; Surlykke et al. 1988); Ep. fuscus: 105 dB (Griffin, 1958)] are indicated by arrows. The dashed line indicates 5 m, the point at which both bats may first detect echoes returning from the moths. The moths detect Ep. fuscus at a distance of 20–25 m, Eu. maculatum will not be detected until less than 1 m.

Does Eu. maculatum eat more moths as a result of having inconspicuous calls? Although stomach and faecal analyses of this bat suggest that it feeds extensively on moths (Easterla, 1965; Ross, 1961, 1967; Poché, 1981; Wai-Ping and Fenton, 1989), caution is warranted in correlating these results with ours since some moth taxa are earless (Roeder, 1974; Fullard and Yack, 1993) and these may be the species fed upon by Eu. maculatum. The flight characteristics of earless moths, however, suggest that this is not the case. Morrill and Fullard (1992) and Lewis et al. (1993) observed that earless moths fly less often, more erratically and closer to the ground than eared moths, traits that presumably serve earless moths as passive defences against the attacks of bats whose echolocation calls they cannot hear. These observations suggest that earless moths do not inhabit the open zones where Eu. maculatum hunts (Woodsworth et al. 1981; Leonard and Fenton, 1983). Our collecting lights sampled from open habitats and the scarcity of earless species among those that we gathered there suggests that it is eared moths that fly with Eu. maculatum and subsequently become their prey. Eu. maculatum presumably benefits from preying on moths over other insects (e.g. caddisflies; Belwood and Fenton, 1976) since the greater body mass of moths would result in greater foraging efficiency for these bats. This suggests that Eu. maculatum may be an authentic candidate for the title of moth ‘specialist’ since this bat appears to have evolved sensory and acoustic methods to maximize its capture success of moths while sacrificing its ability to prey on other insects.

These conclusions lead to two questions: first, if Eu. maculatum poses a threat to moths because of its echolocation calls, why have these insects not evolved ears that can effectively detect them? Second, if Eu. maculatum benefits from its specialisation on moths, why is it (and other bat species that use allotonic echolocation) not more common? Fullard (1982) and Fullard and Belwood (1988) hypothesised that the tone-deaf ears of moths (Roeder, 1967) are not tuned to specific bats but to the acoustic assemblage of the echolocation frequencies of all of the echolocating bats that present a significant predation risk (i.e. selection force) and whether a particular bat species is listened for by a moth is determined by that species’ contribution to the total predator community. To a moth, common bats such as Ep. fuscus form a heavy predation potential because of the high probability of encountering them during the night. Where censuses have been taken, Eu. maculatum is a relatively rare species (Findley and Jones, 1965) and would not participate in many of the total nightly encounters a moth would make with bats. We contend that Eu. maculatum and its echolocation calls have not imparted a strong enough selection pressure on moths to have resulted in the evolution of auditory defences specific to this bat. Paradoxically, although Eu. maculatum feeds heavily on moths, moths have not evolved ears that can detect its calls. We believe that a similar situation exists for the short, faint, allotonically high-frequency calls of the gleaning bat, Myotis septentrionalis in eastern Ontario. This bat preys on moths as they sit on vegetation and uses echolocation calls which are also relatively inaudible to moths (Faure et al. 1993). Myotis septentrionalis, like Eu. maculatum, is an uncommon bat (Fullard et al. 1983) relative to the total bat community, and we contend that its predation pressure has not been intense enough to have favoured moths able to hear its approach. These bats, and other rare species with specialised foraging habits, have therefore exploited the sensory ecology that exists between moths and most bats and are consequently better able to prey heavily upon these insects.

The second question, why bats such as Eu. maculatum are not more common, may be answered by re-examining the problems associated with these types of echolocation systems. The low-frequency calls of Eu. maculatum suggest that this bat has limited itself to large prey items (Leonard and Fenton, 1984) and a foraging zone that contains relatively few insects, constraints that may have favoured the southern portions of this bat’s North American range and in which insect densities may be higher throughout the year. The primarily tropical and subtropical distributions of other low-frequency bats (e.g. Tadarida spp.) (Fenton and Bell, 1981) further suggests that populations of allotonic bats in general are determined by the insect prey base to which their echolocation signals have constrained them.

Although most of the moths we tested were insensitive to the calls of Eu. maculatum, some (e.g. Euxoa flavicollis) had thresholds that indicated that they could detect this bat in time to commence effective avoidance behaviours. High threshold variability at allotonic frequencies is common for moths (Fullard, 1984b; Surlykke and Fullard, 1989) and is likely to be a result of random variation in receptor sensitivities at these values. In the present study, we are confident that threshold variation was not a result of experimental error since low frequencies are not susceptible to confounding acoustic effects such as high reflectivity. We conclude, therefore, that some moths possess low thresholds at these frequencies because of naturally occurring character variability and not as a result of selection from low-frequency bats. If this is true, the rare allotonic sensitivity of these moths may represent a sensory preadaptation to low-frequency bats. It is unlikely that the pleisiomorphic dominant frequency of echolocating bats was lower than 10 kHz (Simmons and Stein, 1980) so the occurrence today of allotonically sensitive moths is not likely to be an evolutionary hold-over from those times.

Whether or not the poor detectability of the calls of Eu. maculatum represents a co-evolved countermanoeuvre (Fenton and Fullard, 1979) or simply a fortuitous coincidence remains speculative. The existence of extremely low-frequency echolocation in bats such as Eu. maculatum is rare and we suggest that the use of these frequencies (acoustically less efficient than typical echolocation for most insect prey sizes) serves primarily to confound moth ears. Simmons and Stein’s (1980) proposed phylogeny of echolocation also suggests that low-frequency allotonic echolocation is rare and likely to be an apomorphic (i.e. derived) trait. Regardless of the evolutionary origins of this type of echolocation, it is not employed by most of the bats in the world. Anti-defence countermanoeuvres should work best when they are not expressed by the majority of predators (Edmunds, 1974) and it is reasonable to expect that certain bats (i.e. a minority of species in any given community) have evolved inconspicuous echolocation calls as a countermanoeuvre to prey more effectively on eared moths.

We thank Drs B. Fenton, M. Brigham and R. Barclay and their students for help in the field, K. Pendlebury for her technical assistance and Dr J. L. D. Lafontaine (Agriculture Canada) for identifying our voucher specimens. We also thank the Osoyoos Indian Band and the Geology Department (University of British Columbia) for permission to use their land and facilities, and M. A. V. Venance, M. B. Fenton and (many) anonymous referees for their helpful comments on the manuscript. This study was funded by research and equipment grants from the Natural Sciences and Engineering Research Council of Canada.

Barclay
,
R. M. R.
and
Brigham
,
R. M.
(
1991
).
Prey detection, dietary niche breadth and body size in bats: why are aerial insectivorous bats so small?
Am. Nat.
137
,
693
703
.
Belwood
,
J. J.
and
Fenton
,
M. B.
(
1976
).
Variation in the diet of Myotis lucifugus (Chiroptera: Vespertilionidae)
.
Can. J. Zool.
54
,
1674
1678
.
Easterla
,
D. A.
(
1965
).
The spotted bat in Utah
.
J. Mammal.
46
,
665
668
.
Edmunds
,
M.
(
1974
).
Defence in animals
.
Burnt Mill, UK
:
Longman Group
.
Faure
,
P. A.
,
Fullard
,
J. H.
and
Barclay
,
R. M. R.
(
1990
).
The response of tympanate moths to the echolocation calls of a substrate-gleaning bat, Myotis evotis
.
J. comp. Physiol. A
166
,
843
849
.
Faure
,
P. A.
,
Fullard
,
J. H.
and
Dawson
,
J. W.
(
1993
).
The gleaning attacks of the northern long-eared bat, Myotis septentrionalis, are relatively inaudible to moths
.
J. exp. Biol.
178
,
173
189
.
Fenton
,
M. B.
(
1970
).
A technique for monitoring bat activity with results obtained from different environments in southern Ontario
.
Can. J. Zool.
48
,
847
851
.
Fenton
,
M. B.
(
1990
).
The foraging ecology of animal-eating bats
.
Can. J. Zool.
68
,
411
422
.
Fenton
,
M. B.
and
Bell
,
G. P.
(
1981
).
Recognition of species of insectivorous bats by their echolocation calls
.
J. Mammal.
62
,
233
243
.
Fenton
,
M. B.
and
Fullard
,
J. H.
(
1979
).
The influence of moth hearing on bat echolocation strategies
.
J. comp. Physiol.
132
,
77
86
.
Findley
,
J. S.
and
Jones
,
C.
(
1965
).
Comments on spotted bats
.
J. Mammal.
46
,
679
680
.
Fullard
,
J. H.
(
1979
).
Behavioral analyses of auditory sensitivity in Cycnia tenera Hubner (Lepidoptera: Arctiidae)
.
J. comp. Physiol.
129
,
79
83
.
Fullard
,
J. H.
(
1982
).
Echolocation assemblages and their effects on moth auditory systems
.
Can. J. Zool.
60
,
2572
2576
.
Fullard
,
J. H.
(
1984a
).
External auditory structures in two species of neotropical notodontid moths
.
J. comp. Physiol. A
155
,
625
632
.
Fullard
,
J. H.
(
1984b
).
Acoustic relationships between tympanate moths and the Hawaiian hoary bat (Lasiurus cinereus semotus)
.
J. comp. Physiol. A
155
,
795
801
.
Fullard
,
J. H.
(
1987a
).
Sensory ecology and neuroethology of moths and bats: interactions in a global perspective
. In
Recent Advances in the Study of Bats
(ed.
M. B.
Fenton
,
P.
Racey
and
J. M. V.
Rayner
), pp.
244
272
.
Cambridge, MA
:
Cambridge University Press
.
Fullard
,
J. H.
(
1987b
).
The defensive function of auditory enhancers in the neotropical moth Antaea lichyi (Lepidoptera: Notodontidae)
.
Can. J. Zool.
65
,
2042
2046
.
Fullard
,
J. H.
(
1988
).
The tuning of moth ears
.
Experientia
44
,
423
428
.
Fullard
,
J. H.
(
1990
).
The sensory ecology of moths and bats: global lessons in staying alive
. In
Insect Defenses
(ed.
D. L.
Evans
and
J. O.
Schmidt
), pp.
203
226
. SUNY Series in Animal Behavior, Albany, New York: State University of New York Press.
Fullard
,
J. H.
(
1994
).
Auditory changes in moths endemic to a bat-free habitat
.
J. evol. Biol.
7
,
435
445
.
Fullard
,
J. H.
and
Barclay
,
R. M. R.
(
1980
).
Audition in spring species of arctiid moths as a possible response to differential levels of insectivorous bat predation
.
Can. J. Zool.
58
,
1745
1750
.
Fullard
,
J. H.
and
Belwood
,
J. J.
(
1988
).
The echolocation assemblage: acoustic ensembles in a neotropical habitat
. In
Animal Sonar
(ed.
P. E.
Nachtigall
and
P. W. B.
Moore
), pp.
639
643
. NATO ASI Series, Series A: Life Sciences, vol.
156
.
New York
:
Plenum Press
.
Fullard
,
J. H.
,
Fenton
,
M. B.
and
Furlonger
,
C. L.
(
1983
).
Sensory relationships of moths and bats sampled from two Nearctic sites
.
Can. J. Zool.
61
,
1752
1757
.
Fullard
,
J. H.
,
Koehler
,
C.
,
Surlykke
,
A.
and
Mckenzie
,
N. L.
(
1991
).
Echolocation ecology and flight morphology of insectivorous bats (Chiroptera) in south–western Australia
.
Aust. J. Zool.
39
,
427
438
.
Fullard
,
J. H.
and
Thomas
,
D. W.
(
1981
).
Detection of certain African, insectivorous bats by sympatric, tympanate moths
.
J. comp. Physiol.
143
,
363
368
.
Fullard
,
J. H.
and
Yack
,
J. E.
(
1993
).
The evolutionary biology of insect hearing
.
Trends Ecol. Evol.
8
,
248
252
.
Griffin
,
D. R.
(
1958
).
Listening in the Dark.
New Haven, CO
:
Yale University Press
.
Griffin
,
D. R.
(
1971
).
The importance of atmospheric attenuation for the echolocation of bats (Chiroptera)
.
Anim. Behav.
19
,
55
61
.
Jones
,
G.
(
1992
).
Bats vs. moths: studies on the diets of rhinolophid and hipposiderid bats support the allotonic frequency hypothesis
. In
Prague Studies in Mammalogy
(ed. I. Horácek and
V.
Vohralik
), pp.
87
92
. Prague: Charles University Press.
Jones
,
G.
and
Rydell
,
J.
(
1994
).
Foraging strategy and predation risk as factors influencing emergence time in echolocating bats
.
Phil. Trans. R. Soc. Lond. B
346
,
445
455
.
Kalko
,
E. K. V.
(
1995
).
Insect pursuit, prey capture and echolocation in pipistrelle bats (Microchiroptera)
.
Anim. Behav.
50
,
861
880
.
Kick
,
S. A.
and
Simmons
,
J. A.
(
1984
).
Automatic gain control in the bat’s sonar receiver and the neuroethology of echolocation
.
J. Neurosci.
4
,
2725
2737
.
Lawrence
,
B. D.
and
Simmons
,
J. A.
(
1982
).
Measurements of atmospheric attenuation at ultrasonic frequencies and the significance for echolocation by bats
.
J. acoust. Soc. Am.
71
,
585
590
.
Leonard
,
M. L.
and
Fenton
,
M. B.
(
1983
).
Habitat use by spotted bats (Euderma maculatum, Chiroptera: Vespertilionidae): roosting and foraging behaviour
.
Can. J. Zool.
61
,
1487
1491
.
Leonard
,
M. L.
and
Fenton
,
M. B.
(
1984
).
Echolocation calls of Euderma maculatum (Vespertilionidae): use in orientation and communication
.
J. Mammal.
65
,
122
126
.
Lewis
,
F. P.
,
Fullard
,
J. H.
and
Morrill
,
S. B.
(
1993
).
Auditory influences on the flight behaviour of moths in a Nearctic site. II. Flight times, heights and erraticism
.
Can. J. Zool.
71
,
1562
1568
.
Miller
,
L. A.
(
1983
).
How insects detect and avoid bats
. In
Neuroethology and Behavioral Physiology
(ed.
F.
Huber
and
H.
Markl
), pp.
251
266
.
Berlin
:
Springer–Verlag
.
Morrill
,
S. B.
and
Fullard
,
J. H.
(
1992
).
Auditory influences on the flight behaviour of moths in a Nearctic site. I. Flight tendency
.
Can. J. Zool.
70
,
1097
1101
.
Neuweiler
,
G.
(
1990
).
Auditory adaptations for prey capture in echolocating bats
.
Physiol. Rev.
70
,
615
641
.
Novick
,
A.
(
1977
).
Acoustic orientation
. In
Biology of Bats
(ed.
W.
Wimsatt
), pp.
73
287
.
New York
:
Academic Press
.
Nüesch
,
H.
(
1957
).
Die Morphologie des Thorax von Telea polyphemus Cr. (Lepid.). I. Nervensystem
.
Zool. Jb.
75
,
615
642
.
Obrist
,
M. K.
(
1995
).
Flexible bat echolocation: the influence of individual habitat and conspecifics on sonar signal design
.
Behav. Ecol. Sociobiol.
36
,
207
219
.
Payne
,
R. S.
,
Roeder
,
K. D.
and
Wallman
,
J.
(
1966
).
Directional sensitivity of the ears of noctuid moths
.
J. exp. Biol.
44
,
17
31
.
Poché
,
R. M.
(
1981
).
Ecology of the spotted bat (Euderma maculatum) in southwest Utah
.
Utah Dept Nat. Res. Publ.
81
1
.
Roeder
,
K. D.
(
1967
).
Nerve Cells and Insect Behavior.
Cambridge, MA
:
Harvard University Press
.
Roeder
,
K. D.
(
1970
).
Episodes in insect brains
.
Am. Sci.
58
,
378
389
.
Roeder
,
K. D.
(
1974
).
Acoustic sensory responses and possible batevasion tactics of certain moths
. In
Proc. Can. Soc. Zoologists Ann. Meeting, Fredericton, N.B., June 2–5, 1974
(ed.
M. D. B.
Burt
), pp.
71
78
.
Fredericton
:
University of New Brunswick Press
.
Ross
,
A.
(
1961
).
Notes on food habits of bats
.
J. Mammal.
42
,
67
.
Ross
,
A.
(
1967
).
Ecological aspects of the food habits of insectivorous bats
.
Proc. West. Found. Vertebr. Zool.
1
,
204
263
.
Rydell
,
J.
,
Jones
,
G.
and
Waters
,
D.
(
1995
).
Echolocating bats and hearing moths: who are the winners?
Oikos
73
,
419
424
.
Simmons
,
J. A.
and
Stein
,
R. A.
(
1980
).
Acoustic imaging in bat sonar: echolocation signals and the evolution of echolocation
.
J. comp. Physiol.
135
,
61
84
.
Stapells
,
D. R.
,
Picton
,
T. W.
and
Smith
,
A. D.
(
1982
).
Normal hearing thresholds for clicks
.
J. acoust. Soc. Am.
86
,
1318
1332
.
Surlykke
,
A.
(
1986
).
Moth hearing on the Faroe Islands, an area without bats
.
Physiol. Ent.
11
,
221
225
.
Surlykke
,
A.
(
1988
).
Interactions between echolocating bats and their prey
. In
Animal Sonar
(ed.
P. E.
Nachtigall
and
P. W. B.
Moore
), pp.
551
566
. NATO ASI Series, Series A: Life Sciences, vol.
156
.
New York
:
Plenum Press
.
Surlykke
,
A.
and
Fullard
,
J. H.
(
1989
).
Hearing in the australian whistling moth, Hecatesia thyridion
.
Naturwiss.
76
,
132
134
.
Surlykke
,
A.
,
Larsen
,
O. N.
and
Michelsen
,
A.
(
1988
).
Temporal coding in the auditory receptor of the moth ear
.
J. comp. Physiol. A
162
,
367
374
.
Wai-Ping
,
V.
and
Fenton
,
M. B.
(
1989
).
Ecology of spotted bat (Euderma maculatum) roosting and foraging behavior
.
J. Mammal.
70
,
617
622
.
Waters
,
D. A.
and
Jones
,
G.
(
1996
).
The peripheral auditory characteristics of noctuid moths: responses to the search-phase echolocation calls of bats
.
J. exp. Biol.
199
,
847
856
.
Whitaker
,
J. O.
and
Black
,
H. L.
(
1976
).
Food habits of cave bats from Zambia, Africa
.
J. Mammal.
57
,
199
204
.
Woodsworth
,
G. C.
,
Bell
,
G. P.
and
Fenton
,
M. B.
(
1981
).
Observations on the echolocation, feeding behaviour and habitat use of Euderma maculatum in southcentral British Columbia
.
Can. J. Zool.
59
,
1099
1102
Zbinden
,
K.
and
Zingg
,
P.
(
1986
).
Search and hunting signals of echolocating European free–tailed bats, Tadarida teniotis, in southern Switzerland
.
Mammalia
50
,
9
25
.