Page 25 - 2016Spring

Subterranean Rodents
Figure 2 shows that the blind mole rat (Spalax ehrenbergi), naked mole rat (Heterocephalus glaber), and pocket gopher (Geomys bursarius), which are subterranean rodents, lie far below the regression line, as they do not hear nearly as high as their small functional head sizes predict. Indeed their upper limits are within the range of those found in non- mammalian vertebrates such as birds. However, these find- ings do not weaken the theory that mammals evolved good high-frequency hearing for sound localization, but are the exceptions that prove the rule—not only do all three species lack good high-frequency hearing, they also lack the ability to localize brief sound (Heffner and Heffner, 1993). Thus it appears that subterranean animals adapted to the one-di- mensional tunnels of an underground habitat have little use for sound localization and are therefore released from the selective pressure to hear high frequencies. In short, sound localization and high-frequency hearing go hand-in-hand in mammals. Mammals rely on high frequencies to localize sound and those that relinquish the ability to localize sound also give up their high frequency hearing.
Mammalian Sound-Localization Acuity
Given that mammals evolved high-frequency hearing for lo- calizing sound, the question arises as to how well they use these cues. Put another way: Does each animal localize as accurately as permitted by the locus cues available to it?
Sound-localization acuity is commonly determined by training an animal to discriminate between the same sound emitted from loudspeakers on the animal’s left versus right sides (e.g., Heffner et al., 2014). The sound is typically a 100- ms broadband noise burst that is too brief to be scanned or tracked by the animal. Such brevity is an important feature as we wish to know the ability of the auditory system to com- pute locus, not the ability of an animal to scan and track an ongoing sound to its source. An animal’s “minimum audible angle” (MAA) is determined by centering the speakers on its midline and finding the smallest angle of separation it can discriminate 50% of the time.
It was originally assumed that all animals were under selec- tive pressure to accurately localize sound, and an animal’s acuity would be determined by the magnitude of the sound- localization cues generated by its head and pinnae. At first this seemed to be the case, as among the few mammals for which sound-localization acuity was known, acuity ap- peared to improve with functional head size. For example, the MAA for humans and elephants was 1-2°, for cats 5°, and
Figure 3. MAA of 39 species of mammals. With the exception of the subterranean rodents (asterisks), the animals were tested with a brief sound, either a click or 100-ms noise bursts. The subterranean ro- dents were unable to localize brief sounds and their thresholds were obtained with a 400 ms or longer duration noise burst. Modified from Heffner et al. (2014).
for Norway rats 12°. However, this changed when it was dis- covered that the MAA of horses and cattle were 25° and 30°, respectively, showing that having a large functional head size did not necessarily produce good sound-localization acuity. Overall, mammalian MAAs range from about 1° for humans and elephants to nearly 30° for cattle, 33° for mice, and 180° for subterranean species (Figure 3). Thus, the question aris- es: If a large head that generates large physical locus cues does not necessarily result in accurate sound-localization acuity, what drives the variation in localization thresholds?
Spring 2016 | Acoustics Today | 23