Prey pursuit strategy of Japanese horseshoe bats during an in-flight target-selection task

The prey pursuit behavior of Japanese horseshoe bats (Rhinolophus ferrumequinum nippon) was investigated by tasking bats during flight with choosing between two tethered fluttering moths. Echolocation pulses were recorded using a telemetry microphone mounted on the bat combined with a 17-channel horizontal microphone array to measure pulse directions. Flight paths of the bat and moths were monitored using two high-speed video cameras. Acoustical measurements of returning echoes from fluttering moths were first collected using an ultrasonic loudspeaker, turning the head direction of the moth relative to the loudspeaker from 0° (front) to 180° (back) in the horizontal plane. The amount of acoustical glints caused by moth fluttering varied with the sound direction, reaching a maximum at 70°–100° in the horizontal plane. In the flight experiment, moths chosen by the bat fluttered within or moved across these angles relative to the bat’s pulse direction, which would cause maximum dynamic changes in the frequency and amplitude of acoustical glints during flight. These results suggest that echoes with acoustical glints containing the strongest frequency and amplitude modulations appear to attract bats for prey selection.     

On-board telemetry of emitted sounds from free-flying bats: compensation for velocity and distance stabilizes echo frequency and amplitude

To understand complex sensory-motor behavior related to object perception by echolocating bats, precise measurements are needed for echoes that bats actually listen to during flight. Recordings of echolocation broadcasts were made from flying bats with a miniature light-weight microphone and radio transmitter (Telemike) set at the position of the bat's ears and carried during flights to a landing point on a wall. Telemike recordings confirm that flying horseshoe bats (Rhinolophus ferrumequinum nippon) adjust the frequency of their sonar broadcasts to compensate for echo Doppler shifts. Returning constant frequency echoes were maintained at the bat's reference frequency +/-83 Hz during flight, indicating that the bats compensated for frequency changes with an accuracy equivalent to that at rest. The flying bats simultaneously compensate for increases in echo amplitude as target range becomes shorter. Flying bats thus receive echoes with both stabilized frequencies and stabilized amplitudes. Although it is widely understood that Doppler-shift frequency compensation facilitates detection of fluttering insects, approaches to a landing do not involve fluttering objects. Combined frequency and amplitude compensation may instead be for optimization of successive frequency modulated echoes for target range estimation to control approach and landing.     

Discrimination of fluttering targets by the FM-batPipistrellus stenopterus?

Summary Five bats of the speciesPipistrellus stenopterus were trained in a two-alternative forced-choice procedure to discriminate between two fluttering targets. The positive target simulated an insect with a 50 Hz wingbeat rate. The negative target was varied between 0 and 48 Hz.The bats were able to discriminate a target with 41 Hz from a target with 50 Hz with 75% correct choices. In the discrimination task, they typically emitted echolocation calls of 2–4 ms duration sweeping from 60 kHz to 30 kHz. The duty cycle (i.e. fraction of time filled with echolocation sounds) increased when the targets fluttered, but was always lower than 3%.The performance ofP. stenopterus in discriminating fluttering targets is comparable to that of bats emitting longer sounds with constant-frequency (CF) components and a higher duty cycle. The FM-sounds ofP. stenopterus are short compared with the period of the fluttering targets, and therefore make it difficult for the animal to measure the time interval between two acoustic glints. Other cues may be prominent, such as the frequency modulation by Doppler shifts from the moving blades.     

High duty cycle pulses suppress orientation flights of crambid moths

Bat-and-moth is a good model system for understanding predator–prey interactions resulting from interspecific coevolution. Night-flying insects have been under predation pressure from echolocating bats for 65 Myr, pressuring vulnerable moths to evolve ultrasound detection and evasive maneuvers as counter tactics. Past studies of defensive behaviors against attacking bats have been biased toward noctuoid moth responses to short duration pulses of low-duty-cycle (LDC) bat calls. Depending on the region, however, moths have been exposed to predation pressure from high-duty-cycle (HDC) bats as well. Here, we reveal that long duration pulse of the sympatric HDC bat (e.g., greater horseshoe bat) is easily detected by the auditory nerve of Japanese crambid moths (yellow peach moth and Asian corn borer) and suppress both mate-finding flights of virgin males and host-finding flights of mated females. The hearing sensitivities for the duration of pulse stimuli significantly dropped non-linearly in both the two moth species as the pulse duration shortened. These hearing properties support the energy integrator model; however, the threshold reduction per doubling the duration has slightly larger than those of other moth species hitherto reported. And also, Asian corn borer showed a lower auditory sensitivity and a lower flight suppression to short duration pulse than yellow peach moth did. Therefore, flight disruption of moth might be more frequently achieved by the pulse structure of HDC calls. The combination of long pulses and inter-pulse intervals, which moths can readily continue detecting, will be useful for repelling moth pests.     

Acoustic irradiation produced by flying moths (Lepidoptera,Noctuidae)

Characteristics of acoustic waves accompanying the flight of noctuid moths (Noctuidae) were measured. The low-frequency part of the spectrum is formed of a series of up to 17 harmonics of the wingbeat frequency (30–50 Hz) with a general tendency toward the decrease in the spectral density and the increase in the sound frequency. The root-mean-square level of the sound pressure from flapping wings was found to be 70–78 dB SPL. Besides low-frequency components, the flight of moths was accompanied by short ultrasonic pulses, which appeared with every wingbeat. Most of the spectral energy was concentrated within a range of 7–150 kHz with the main peaks at 60–110 kHz. The short-term pulses were divided into two or more subpulses with different spectra. The high-frequency pulses were produced at two phases of the wingbeat cycle: during the pronation of the wings at the highest point and at the beginning of their upward movement from the lowest point. In most of the specimens tested, the peak amplitude of sounds varied from 55 to 65 dB SPL at a distance of 6 cm from the insect body. However, in nine noctuid species, no high-frequency acoustic components were recorded. In these experiments, the acoustic flow from the flying moth within a frequency range of 2 to 20 kHz did not exceed the self-noise level of the microphone amplifier (RMS 18 dB SPL). Probable mechanisms of the high frequency acoustic emission during flight, the effect of these sounds on the auditory sensitivity of moths, and the possibility of their self-revealing to insectivorous bats are discussed. In addition, spectral characteristics of the moth echolocation clicks were more precisely determined within the higher frequency range (>100 kHz).     

Classification of insects by echolocating greater horseshoe bats

Summary Echolocating greater horseshoe bats (Rhinolophus ferrumequinum) detect insects by concentrating on the characteristic amplitude- and frequency modulation pattern fluttering insects impose on the returning echoes. This study shows that horseshoe bats can also further analyse insect echoes and thus recognize and categorize the kind of insect they are echolocating.Four greater horseshoe bats were trained in a twoalternative forced-choice procedure to choose the echo of one particular insect species turning its side towards the bat (Fig. 1). The bats were able to discriminate with over 90% correct choices between the reward-positive echo and the echoes of other insect species all fluttering with exactly the same wingbeat rate (Fig. 4).When the angular orientation of the reward-positive insect was changed (Fig. 2), the bats still preferred these unknown echoes over echoes from other insect species (Fig. 5) without any further training. Because the untrained bats did not show any prey preference, this indicates that the bats were able to perform an aspect-anglein-dependent classification of insects.Finally we tested what parameters in the echo were responsible for species recognition. It turned out that the bats especially used the small echo-modulations in between glints as a source of information (Fig. 7). Neither the amplitudenor the frequencymodulation of the echoes alone was sufficient for recognition of the insect species (Fig. 8). Bats performed a pattern recognition task based on complex computations of several acoustic parameters, an ability which might be termed cognitive.Abbreviations AM amplitude modulation - CF constant frequency - FM frequency modulation - S+ positive stimulus - S- negative stimulus     

Dominant glint based prey localization in horseshoe bats: a possible strategy for noise rejection

Rhinolophidae or Horseshoe bats emit long and narrowband calls. Fluttering insect prey generates echoes in which amplitude and frequency shifts are present, i.e. glints. These glints are reliable cues about the presence of prey and also encode certain properties of the prey. In this paper, we propose that these glints, i.e. the dominant glints, are also reliable signals upon which to base prey localization. In contrast to the spectral cues used by many other bats, the localization cues in Rhinolophidae are most likely provided by self-induced amplitude modulations generated by pinnae movement. Amplitude variations in the echo not introduced by the moving pinnae can be considered as noise interfering with the localization process. The amplitude of the dominant glints is very stable. Therefore, these parts of the echoes contain very little noise. However, using only the dominant glints potentially comes at a cost. Depending on the flutter rate of the insect, a limited number of dominant glints will be present in each echo giving the bat a limited number of sample points on which to base localization. We evaluate the feasibility of a strategy under which Rhinolophidae use only dominant glints. We use a computational model of the echolocation task faced by Rhinolophidae. Our model includes the spatial filtering of the echoes by the morphology of the sonar apparatus of Rhinolophus rouxii as well as the amplitude modulations introduced by pinnae movements. Using this model, we evaluate whether the dominant glints provide Rhinolophidae with enough information to perform localization. Our simulations show that Rhinolophidae can use dominant glints in the echoes as carriers for self-induced amplitude modulations serving as localization cues. In particular, it is shown that the reduction in noise achieved by using only the dominant glints outweighs the information loss that occurs by sampling the echo.     

Detection of sonar signals in the presence of pulses of masking noise by the echolocating bat,Eptesicus fuscus

Summary Two big brown bats (Eptesicus fuscus) were trained to report the presence or absence of a virtual sonar target. The bats' sensitivity to transient masking was investigated by adding 5 ms pulses of white noise delayed from 0 to 16 ms relative to the target echo. When signal and masker occurred simultaneously, the bats required a signal energy to noise spectrum level ratio of 35 dB for 50% probability of detection. When the masker was delayed by 2 ms or more there was no significant masking and echo energy could be reduced by 30 dB for the same probability of detection. The average duration of the most energetic sonar signal of each trial was measured to be 1.7 ms and 2.4 ms for the two bats, but a simple relation between detection performance and pulse duration was not found.In a different experiment the masking noise pulses coincided with the echo, and the duration of the masker was varied from 2 to 37.5 ms. The duration of the masker had little or no effect on the probability of detection.The findings are consistent with an aural integration time constant of about 2 ms, which is comparable to the duration of the cries. This is an order of magnitude less than found in backward masking experiments with humans and may be an adaptation to the special constraints of echolocation. The short time of sensitivity to masking may indicate that the broad band clicks of arctiid moths produced as a countermeasure to bat predation are unlikely to function by masking the echo of the moth.Abbreviations SPL sound pressure level - SD standard deviation - SE standard error - BW bandwidth     

Acoustic Aposematism and Evasive Action in Select Chemically Defended Arctiine (Lepidoptera: Erebidae) Species: Nonchalant or Not?

Tiger moths (Erebidae: Arctiinae) have experienced intense selective pressure from echolocating, insectivorous bats for over 65 million years. One outcome has been the evolution of acoustic signals that advertise the presence of toxins sequestered from the moths’ larval host plants, i.e. acoustic aposematism. Little is known about the effectiveness of tiger moth anti-bat sounds in their natural environments. We used multiple infrared cameras to reconstruct bat-moth interactions in three-dimensional (3-D) space to examine how functional sound-producing organs called tymbals affect predation of two chemically defended tiger moth species: Pygarctia roseicapitis (Arctiini) and Cisthene martini (Lithosiini). P. roseicapitis and C. martini with intact tymbals were 1.8 and 1.6 times less likely to be captured by bats relative to those rendered silent. 3-D flight path and acoustic analyses indicated that bats actively avoided capturing sound-producing moths. Clicking behavior differed between the two tiger moth species, with P. roseicapitis responding in an earlier phase of bat attack. Evasive flight behavior in response to bat attacks was markedly different between the two tiger moth species. P. roseicapitis frequently paired evasive dives with aposematic sound production. C. martini were considerably more nonchalant and employed evasion in fewer interactions. Our results show that acoustic aposematism is effective at deterring bat predation in a natural context and that this strategy is likely to be the ancestral function of tymbal organs within the Arctiinae.     

Noctuid moths show neural and behavioural responses to sounds made by some bat-marking rings

Coloured rings are often used for marking bats so that specific individuals can be recognized. We noticed that the rings of mouse-eared bats, Myotis myotis and Myotis blythii, in a combination of one plastic-split and one metallic ring on the same forearm, emitted sounds that were largely ultrasonic each time the rings met in flight. We recorded the ring sounds and the echolocation calls produced by the bats, and played them back to neural preparations of lesser yellow underwing moths, Noctua comes, while making extracellular recordings from the moths' A1 auditory receptors. The peak energy of the ring sounds occurred much closer in frequency to the moth's best auditory frequency (the frequency at which the moth has the lowest auditory threshold) than the peak energy of the calls, for both bat species, and the ring sounds were detected at a threshold 5-6 dB peSPL lower than the calls. Moths performed evasive manoeuvres to playbacks of ring sounds more frequently than they did to control (tape noise) sequences. These neural and behavioural responses imply that certain bats should not be marked with two rings on one wing, as this may make the bat more apparent to tympanate insects, and may therefore reduce its foraging success. Copyright 1999 The Association for the Study of Animal Behaviour.     

Persistence and regression of hearing in the exclusively diurnal moths, Trichodezia albovittata (Geometridae) and Lycomorpha pholus (Arctiidae)

Abstract.  1. Auditory sensitivities and ultrasound avoidance behaviour of two exclusively diurnal moths were examined to test the prediction that total isolation from the predatory effects of echolocating bats will result in the regression of these sensory systems and/or the defences they evoke.
2. The silent geometrid, Trichodezia albovittata , possessed large ears with auditory neural thresholds similar to or better than those of a sympatric, exclusively nocturnal geometrid moth. Trichodezia albovittata readily responded with evasive flight to ultrasound and it is suggested that if this moth has become completely isolated from bats its ears are functionally vestigial, at least in the population studied here.
3. In contrast, while the sound-producing arctiid, Lycomorpha pholus , had low auditory sensitivity based on neural thresholds, it still responded with flight changes to ultrasound. It did not, however, produce sounds when stimulated ultrasonically. It is suggested that the ears of this moth are functionally vestigial for bat-detection purposes but may be used for short-distance social communication.     

Convergence of temporal and spectral information into acoustic images of complex sonar targets perceived by the echolocating bat,Eptesicus fuscus

1. FM echolocating bats (Eptesicus fuscus) were trained to discriminate between a two-component complex target and a one-component simple target simulated by electronically-returned echoes in a series of experiments that explore the composition of the image of the two-component target. In Experiment I, echoes for each target were presented sequentially, and the bats had to compare a stored image of one target with that of the other. The bats made errors when the range of the simple target corresponded to the range of either glint in the complex target, indicating that some trace of the parts of one image interfered with perception of the other image. In Experiment II, echoes were presented simultaneously as well as sequentially, permitting direct masking of echoes from one target to the other. Changes in echo amplitude produced shifts in apparent range whose pattern depended upon the mode of echo presentation. 2. Eptesicus perceives images of complex sonar targets that explicitly represent the location and spacing of discrete glints located at different ranges. The bat perceives the target's structure in terms of its range profile along a psychological range axis using a combination of echo delay and echo spectral representations that together resemble a spectrogram of the FM echoes. The image itself is expressed entirely along a range scale that is defined with reference to echo delay. Spectral information contributes to the image by providing estimates of the range separation of glints, but it is transformed into these estimates. 3. Perceived absolute range is encoded by the timing of neural discharges and is vulnerable to shifts caused by neural amplitude-latency trading, which was estimated at 13 to 18 microseconds per dB from N1 and N4 auditory evoked potentials in Eptesicus. Spectral cues representing the separation of glints within the target are transformed into estimates of delay separations before being incorporated into the image. However, because they are encoded by neural frequency tuning rather than the time-of-occurrence of neural discharges, the perceived range separation of glints in images is not vulnerable to amplitude-latency shifts. 4. The bat perceives an image that is displayed in the domain of time or range. The image receives no evident spectral contribution beyond what is transformed into delay estimates. Although the initial auditory representation of FM echoes is spectrogram-like, the time, frequency, and amplitude dimensions of the spectrogram appear to be compressed into an image that has only time and amplitude dimensions.(ABSTRACT TRUNCATED AT 400 WORDS)     

Tiger moths and the threat of bats: decision-making based on the activity of a single sensory neuron

Echolocating bats and eared moths are a model system of predator–prey interaction within an almost exclusively auditory world. Through selective pressures from aerial-hawking bats, noctuoid moths have evolved simple ears that contain one to two auditory neurons and function to detect bat echolocation calls and initiate defensive flight behaviours. Among these moths, some chemically defended and mimetic tiger moths also produce ultrasonic clicks in response to bat echolocation calls; these defensive signals are effective warning signals and may interfere with bats'' ability to process echoic information. Here, we demonstrate that the activity of a single auditory neuron (the A1 cell) provides sufficient information for the toxic dogbane tiger moth, Cycnia tenera, to decide when to initiate defensive sound production in the face of bats. Thus, despite previous suggestions to the contrary, these moths'' only other auditory neuron, the less sensitive A2 cell, is not necessary for initiating sound production. However, we found a positive linear relationship between combined A1 and A2 activity and the number of clicks the dogbane tiger moth produces.     

Convergent evolution of anti-bat sounds

Bats and their insect prey rely on acoustic sensing in predator prey encounters—echolocation in bats, tympanic hearing in moths. Some insects also emit sounds for bat defense. Here, we describe a previously unknown sound-producing organ in Geometrid moths—a prothoracic tymbal in the orange beggar moth (Eubaphe unicolor) that generates bursts of ultrasonic clicks in response to tactile stimulation and playback of a bat echolocation attack sequence. Using scanning electron microscopy and high-speed videography, we demonstrate that E. unicolor and phylogenetically distant tiger moths have evolved serially homologous thoracic tymbal organs with fundamentally similar functional morphology, a striking example of convergent evolution. We compared E. unicolor clicks to that of five sympatric tiger moths and found that 9 of 13 E. unicolor clicking parameters were within the range of sympatric tiger moths. Remaining differences may result from the small size of the E. unicolor tymbal. Four of the five sympatric clicking tiger moth species were unpalatable to bats (0–20 % eaten), whereas E. unicolor was palatable to bats (86 % eaten). Based on these results, we hypothesize that E. unicolor evolved tymbal organs that mimic the sounds produced by toxic tiger moths when attacked by echolocating bats.     

Listening for bats: pulse repetition rate as a cue for a defensive behavior inCycnia tenera (Lepidoptera:Arctiidae)

Summary The tympanate, arctiid moth,Cycnia tenera responds to pulsed, 30 kHz acoustic stimuli resembling bat echolocation signals by emitting trains of clicks. This phonoresponse was used to determine that this moth is maximally sensitive to stimulus pulse repetition rates of 30–50 pulses/s, rates typically emitted by bats shortly before they close with their targets. At rates both above and below this optimum moths exhibit higher thresholds and reduced responsiveness. These data suggest thatC. tenera is capable of using the repetition rate emitted by an approaching bat as a cue in determining the relative proximity of the bat. The use of repetition rate information should allow this moth both an unambiguous indication of a bat at very close range as well as the ability to distinguish sources of nocturnal, high-frequency sounds not emitted by predators.     

Female greater wax moths reduce sexual display behavior in relation to the potential risk of predation by echolocating bats

Female greater wax moths Galleria mellonella display by wing fanning in response to bursts of ultrasonic calls produced bymales. The temporal and spectral characteristics of these callsshow some similarities with the echolocation calls of batsthat emit frequency-modulated (FM) signals. Female G. mellonellatherefore need to distinguish between the attractive signalsof male conspecifics, which may lead to mating opportunities,and similar sounds made by predatory bats. We therefore predictedthat (1) females would display in response to playbacks of male calls; (2) females would not display in response to playbacksof the calls of echolocating bats (we used the calls of Daubenton'sbat Myotis daubentonii as representative of a typical FM echolocatingbat); and (3) when presented with male calls and bat callsduring the same time block, females would display more whenperceived predation risk was lower. We manipulated predationrisk in two ways. First, we varied the intensity of bat callsto represent a nearby (high risk) or distant (low risk) bat.Second, we played back calls of bats searching for prey (lowrisk) and attacking prey (high risk). All predictions weresupported, suggesting that female G. mellonella are able todistinguish conspecific male mating calls from bat calls, andthat they modify display rate in relation to predation risk.The mechanism (s) by which the moths separate the calls ofbat and moth must involve temporal cues. Bat and moth signalsdiffer considerably in duration, and differences in durationcould be encoded by the moth's nervous system and used in discrimination.     

Hearing and evasive behaviour in the greater wax moth, Galleria mellonella (Pyralidae)

Greater wax moths (Galleria mellonella L., Pyraloidea) use ultrasound sensitive ears to detect clicking conspecifics and echolocating bats. Pyralid ears have four sensory cells, A_1?4. The audiogram of G. mellonella has best frequency at 60 kHz with a threshold around 47 dB sound pressure level. A₁ and A₂ have almost equal thresholds in contrast to noctuids and geometrids. A₃ responds at + 12 to + 16 dB relative to the A₁ threshold. The threshold data from the A‐cells give no indication of frequency discrimination in greater wax moths. Tethered greater wax moths respond to ultrasound with short‐latency cessation of flight at + 20 to + 25 dB relative to the A₁ threshold. The behavioural threshold curve parallels the audiogram, thus further corroborating the lack of frequency discrimination. Hence, the distinction between bats and conspecifics is probably based on temporal cues. At a constant duty cycle (percentage of time where sound is on) the pulse repetition rate has no effect on the threshold for flight cessation, but stimulus duration affects both sensory and behavioural thresholds. The maximum integration time is essentially the same: 45 ms for the A₁‐cell and 50–60 ms for the flight cessation response. However, the slopes of the time‐intensity trade‐off functions are very different: ? 2.1 dB per doubling of sound duration for the A₁‐cell threshold, and ? 7.2 dB per doubling of sound duration for the behavioural threshold. The significance of the results for sexual acoustic communication as well as for bat defence is discussed.     

Discrimination of insect wingbeat-frequencies by the batRhinolophus ferrumequinum

Summary Five Greater Horseshoe bats,Rhinolophus ferrumequinum, were trained in a two-alternative forced-choice procedure to discriminate between artificial echoes of insects fluttering at different wingbeat rates. The stimuli were electronically produced phantom targets simulating fluttering insects with various wingbeat frequencies (Figs. 3, 4). Difference thresholds for wingbeat rates of 50 Hz and 100 Hz were determined. For an S+ of 50 Hz the difference threshold values lay between 2.8 and 4.6 Hz for individual bats; with an S+ of 100 Hz they increased to between 9.8 and 12.0 Hz (Figs. 5, 6, Table 1).Three bats, previously trained to discriminate between a S+ of 50 Hz and a S– with a lower wingbeat rate, were tested with higher frequency stimuli. When they had to decide between their old S+ of 50 Hz and either a 60 or 70 Hz echo two bats continued to select the 50 Hz stimulus while the third bat now preferred the faster fluttering insects (Table 2).During the discrimination task the echolocation behavior of the bats was monitored. When the phantom targets were presented all bats increased their duty-cycle of sound emission from about 40% to sometimes near 70%. They did so by either emitting longer echolocation calls or by increasing the sound repetition rate (Figs. 7, 8).The results show that Greater Horseshoe bats can determine the wingbeat rate of flying insects with an accuracy between 6 and 12%. Possible cues for flutter rate determination by cf-fm bats from natural and artificial insect echoes are discussed.Abbreviations DC duty-cycle - PD pulse duration - PI pulse interval - cf constantfrequency - fm frequency modulation     

Arctiid moth clicks can degrade the accuracy of range difference discrimination in echolocating big brown bats,Eptesicus fuscus

Summary Four big brown bats (Eptesicus fuscus) born and raised in captivity were trained using the Yes/No psychophysical method to report whether a virtual sonar target was at a standard distance or not. At threshold bats were able to detect a minimum range difference of 6 mm (a t of 36 s).Following threshold determinations, a click burst 1.8 ms long containing 5 pulses from the ruby tiger moth, Phragmatobia fuliginosa (Arctiidae), was presented randomly after each phantom echo. The sound energy of the click burst was -4 dB relative to that of the phantom echo. Clicks presented for the very first time could startle naive bats to different degrees depending on the individual.The bats' performance deteriorated by as much as 4000% when the click burst started within a window of about 1.5 ms before the phantom echo (Fig. 4). Even when one of ten phantom echoes was preceded by a click burst, the range difference discrimination worsened by 200% (Fig. 9). Hence, clicks falling within the 1.5 ms time window seem to interfere with the bat's neural timing mechanism.The clicks of arctiid moths appear to serve 3 functions: they can startle naive bats, interfere with range difference determinations, or they can signal the moth's distastefulness, as shown in earlier studies.Abbreviations peSPL peak equivalent sound pressure level - sd standard deviation - FM frequency modulation - CF constant frequency - EPROM erasable programmable read only memory