Скачать книгу

Eastern Highlands.

      In addition, riparian forest plays an important role in bat distributions. Riparian fringes along the Limpopo and Zambezi rivers and their major tributaries extend the ranges of several species deep into the semi-arid savanna of southern and northern Zimbabwe. This factor can be invoked to explain outlying records of several species, including Epomophorus dobsonii and Myotis bocagii, and especially Pipistrellus rueppellii. Available data on the diversity of bat communities reveal that bat communities in high tropical forests are richer in species (e.g. at least 59 species in a Liberian study area) (Monadjem et al. 2013b) compared to the richest bat assemblages recorded in the savanna woodlands of southern and Central Africa. In savanna, total species richness ranges from 40 to 45 species at three intensive study sites: Pafuri, Kruger National Park (Aldridge and Rautenbach 1987), Soutpansberg Mountains (Monadjem et al. 2018b) and Sengwa Wildlife Research Area, Zimbabwe (Fenton 1975, 1985). These differences represent the greater diversity of clutter foragers and also frugivorous species in rainforests (Monadjem et al. 2018b).

      The majority of vertebrates rely on vision to perceive their environment. Even nocturnal predators such as owls and lions principally use their eyes to navigate and hunt. In contrast, most bats find food and avoid obstacles at night with great ease using an alternative sensory mechanism, called echolocation. Although usually associated with bats, other animals such as toothed whales, porpoises, some species of shrews and tenrecs, oilbirds, and several species of swiftlets also use echolocation.

      Echolocating bats emit sound pulses and analyse the returning echoes to detect, characterise, and localise objects that reflect the impinging pulse as an echo (Fenton 1990, Schnitzler and Kalko 2001, Fenton et al. 2016) (Figure 35). Sound pulses are generated in the larynx (except in Rousettus species, which produce echolocation pulses by repeatedly clicking their tongue against the palate), and emitted through the mouth (e.g. Vespertilionidae, Miniopteridae, Cistugidae, Emballonuridae, and Molossidae) or nose (e.g. Rhinolophidae, Hipposideridae, Rhinonycteridae and Nycteridae). The ears, or pinnae, receive returning echoes, which are then funnelled and processed into the rest of the bat’s hearing system.

image

      Not all vocalisations produced by bats are echolocation calls. For example, the audible squeaks bats make in their roosts, the calls that mother and young make to one another, or the calls flying bats make to defend their foraging territories, are not echolocation calls. Instead, these are usually referred to as social calls and are less well understood than echolocation calls.

      The echolocation frequencies of most bat species are ultrasonic (i.e. above the range of human hearing), and peak echolocation frequencies (i.e. the frequencies with the highest intensity) usually fall within 20–60 kHz (Fenton 1990). This may be due to the frequency-dependent effects of atmospheric attenuation and target strength (Jones and Rydell 2004). In contrast, many social calls are audible to humans.

      Echolocation coupled with flight enables bats to capture nocturnal flying insects in a variety of habitats. This ability probably explains how the radiation of bat species has manifested into the highest trophic diversity among mammals (Patterson et al. 2004, Roemer et al. 2019). Nevertheless, not all bats echolocate (e.g. fruit bats from the family Pteropodidae, except Rousettus species), nor do all echolocating bats use the same type of echolocation.

      Two different echolocation systems – high and low duty-cycle echolocation – evolved independently in the Chiroptera (Eick et al. 2005). Low duty-cycle echolocation bats emit narrowband or broadband sound pulses separated by inter-pulse intervals that are much longer than the duration of the emitted pulses. Such species therefore separate the emitted pulse from the returning echo in time (Fenton 1990).

      High duty-cycle bats emit long, narrowband pulses that are separated by much shorter inter-pulse intervals. Consequently, the emitted signals often overlap with the returning echo. However, the overlap does not produce masking effects (Schnitzler and Kalko 2001), because Doppler-shift compensation keeps the target echo in the range of the neurons of the acoustic fovea – a group of neurons sharply tuned to a very narrow frequency band, a few kHz higher than the peak frequency of the emitted signal (Neuweiler 1990). The Doppler effect is the change in frequency of a sound wave as perceived by a listener moving relative to the source of the sound. For example, the frequency of the siren of a passing emergency vehicle will start out high, slide down as it passes by the listener, and continue lower as it recedes into the distance. In other words, high duty-cycle bats separate the emitted pulse from the returning echo in frequency rather than time (Fenton 1990).

      Broadband, low duty-cycle, frequency-modulated (LD-FM) echolocation pulses typically sweep downward through as much as an octave for a short duration of time (Fenton 1990, Schnitzler and Kalko 2001) (Figure 36). At the same time, the bat senses with increasing precision the range and position of the object in space – its localisation (Simmons and Stein 1980). This is because LD-FM signals sweep rapidly through the corresponding neural filters, and are therefore reliable time markers to determine the range of the target from the foraging bat (Moss and Schnitzler 1995). In addition, the neuronal filters are activated across a broad frequency range, which increases the reliability of the monaural and binaural cues the bat uses to localise the target in space (Schnitzler and Kalko 2001). Nevertheless, LD-FM signals are less suited for the detection of distant and/or weak echoes, because the neuronal filters are activated for only a short time (Schnitzler and Kalko 2001).

image

      Narrowband, low duty-cycle pulses composed of constant frequency (LD-CF) or shallow frequency-modulated (LD-QCF) components (Figure 36) are not suitable for localisation of a hunted target, but are well suited to detection, because they activate the neuronal filters of the corresponding narrow frequency band during the entire echo (Schnitzler and Kalko 2001). In addition, acoustic ‘glints’ (short prominent amplitude peaks in the echo that are created when a fluttering insect’s wing is perpendicular to the incoming sound wave) can be 20–30 decibel (dB) stronger than the echo from the body of the insect; it is these glints off the flying insect’s wings that further increase the likelihood of its detection by the bat (Kober and Schnitzler 1980, Moss and Zagaeski 1994).

      In contrast to low duty-cycle bats, Doppler-shift compensation combined with a specialised auditory system enables constant frequency high duty-cycle (HD-CF) echolocating bats to localise and classify fluttering insects in dense (clutter) habitats (Schnitzler and Kalko 2001). HD-CF bats can classify insects by listening to the unique acoustic glints imprinted by the fluttering wings of different insects onto the echoes of their CF calls (Schnitzler 1987, von der Emde and Menne 1989, von der Emde and Schnitzler 1990) (Figure 36).

      Bat species exhibit great diversity in how and where they fly. The most important ecological constraint on foraging insectivorous bats is clutter, that is, the number of obstacles a bat has to detect and avoid (Fenton 1990). Vegetation structure has an overriding control over the relative clutter of the habitats

Скачать книгу