Bats of Southern and Central Africa. Ara Monadjem

Чтение книги онлайн.

СКАЧАТЬ group). These bats have short, broad wings and low wing loading. Bats that hunt insects between vegetation (clutter-edge functional foraging group) have intermediate-sized wings, wingspans, and wing loading (e.g. Scotophilus). Fruit-eating bats (e.g. Rousettus) have large wings and high wing loading (modified after Neuweiler 2000). Ef Epomops franqueti; Eh Eptesicus hottentotus; Na Neoromicia anchietae; Hc Hipposideros caffer; Mn Miniopterus natalensis; Ra Rousettus aegyptiacus; Rcl Rhinolophus clivosus; Sd Scotophilus dinganii; Ta Tadarida aegyptiaca; Tm Taphozous mauritianus.

      Open-air foragers that feed on flying prey high above the ground and far from vegetation have no problems with clutter echoes and obstacles. Masking problems are not likely to affect echolocation signals, as long as the emitted signal does not overlap with the returning echo. Nevertheless, bats that forage in open habitats must find relatively small prey in a big space. Hence, their echolocation signals are optimised for detection, and their wing morphology is geared for speed and agility (Norberg and Rayner 1987, Schnitzler and Kalko 1998, 2001). Some of the species in this functional group fly at considerable heights above ground; in Zimbabwe, Fenton and Griffin (1997) revealed that six species of molossids and Taphozous mauritianus feed at heights over 500 m above ground.

      Clutter-edge foragers are bats that hunt for insects near the edges of clutter such as forest edges and gaps. Clutter-edge bats therefore experience perceptual and mechanical constraints at the same time. They must distinguish insect echoes from clutter-edge echoes, and navigate along these edges avoiding collision. Clutter-edge bats solve these problems by using mixed signals (Schnitzler and Kalko 1998, 2001). Typically, search-phase echolocation consists of low duty-cycle QCF signals alternated with FM signals. QCF signals increase detection distance, while FM signals allow the bats to localise and distinguish between clutter echoes and insect echoes. Wing morphology is variable, but most species have average wingspans and wing areas. Consequently, some bats are very flexible in their foraging and echolocation behaviour and often switch between open and clutter-edge habitat space (Fenton 1990).

      Clutter foragers are bats that search for and capture insects in highly cluttered space close to the ground or vegetation, and therefore experience more perceptual and mechanical constraints than clutter-edge foragers. Clutter foragers must distinguish insect echoes buried in clutter echoes, and at the same time know their exact spatial position to navigate and avoid collision with the clutter. Bats have evolved two different echolocation sensory strategies to solve this problem (Schnitzler and Kalko 2001). On the one hand, HD-CF bats use overlap-insensitive CF signals of long duration to hunt fluttering insects. On the other hand, LD-FM bats, which glean prey from surfaces, use overlap-sensitive uni- or multi-harmonic FM signals of short duration at low intensities (< 100 dB). Short-duration and low-intensity calls do not overload the sensory system with clutter echoes (Schnitzler and Kalko 2001). In addition, many of these bats have long ears that listen for prey-generated acoustic cues, such as the calls of crickets and frogs (Tuttle and Ryan 1981, Bell 1982, Fenton et al. 1983). All clutter-foraging bats have similar short and broad wings associated with the slow, manoeuvrable flight necessary to hunt in clutter habitats (Norberg and Rayner 1987).

      In this book, open-air foragers include insectivorous bat species from the families Molossidae and Emballonuridae. These LD-CF and LD-QCF bat species are characterised by long and narrow wings with high wing loading and aspect ratio (> 10.9), coupled with narrowband echolocation calls at low frequencies (< 30 kHz) and of long duration (> 8 ms).

      Clutter-edge foragers include insectivorous bat species from the families Vespertilionidae, Cistugidae and Miniopteridae. In general, these bat species are characterised by average wingspans, and wing areas with low to intermediate wing loading and aspect ratio (between 7 and 10.9), coupled with echolocation calls at intermediate frequencies (30–70 kHz), ranging from broadband (> 20 kHz) to narrowband signals (< 20 kHz) of intermediate duration (3–8 ms).

      Clutter foragers are divided into two groups: HD-CF bats of the families Hipposideridae, Rhinonycteridae and Rhinolophidae, which emit CF signals of long duration (10–100 ms), and medium to high peak frequency (> 30 kHz); and LD-FM bats of the families Nycteridae and Megadermatidae, which use FM signals of short duration (1–3 ms) at low intensities (< 100 dB). Despite having very different echolocation systems, the wings of all clutter feeders are relatively short and broad with low wing loading and aspect ratio (< 7).

RECORDING ECHOLOCATION CALLS

      Sound originates when matter vibrates in a medium such as air or water, and is perceived by the sense of hearing. Physically, sound is the compression and rarefaction of particles as a wave in the medium through which the sound is travelling. A microphone converts the wave motion of the particles into an electric signal. Special ultrasound microphones are required to record ultrasonic bat frequencies.

      Before a complex sound such as a bat’s echolocation call can be manipulated or analysed with a digital computer, the signal must be acquired or digitised by an analog-to-digital (A/D) converter. The A/D converter repeatedly samples the voltage amplitude of the electric input signal at a particular sampling rate, typically tens or hundreds of thousands of times per second. The quality of the resultant digitised signal depends on the rate at which amplitude measurements are made (the sampling rate), and the number of bits used to represent each amplitude measurement (the sample size).

      The sampling rate must be more than twice as high as the highest frequency contained in the original signal. Otherwise, the digitised signal will have (phantom) frequencies that were not present in the original signal. This is called aliasing. For example, an accurate recording of an echolocation pulse with a peak frequency of 100 kHz requires a sampling rate of at least 200 kHz.

      The precision with which the digitised amplitude represents the actual amplitude at the instant the sample is taken depends on the sample size or number of bits used. Some recording models have an 8-bit sampling limit, while others allow a choice between 8-bit and 16-bit samples. An 8-bit sample can resolve 256 (= 2⁸) different amplitude values; a 16-bit converter can resolve 65,536 (= 2¹⁶) values.

      For the echolocation recordings in this book, the sampling rate (250 or 500 kHz) was always more than twice the highest frequency contained in the original bat signal, and the sample size was 16-bit.

SPECTRUM ANALYSIS

      Any acoustic signal can be graphically or mathematically represented in two ways: time domain and frequency domain. A time-domain graph shows how a signal changes over time, while a frequency-domain graph shows how much of the signal lies within each given frequency band over a range of frequencies. A pure tone is called a sinusoid because its amplitude is a sine function in the time domain. In the frequency domain, it is a vertical line. Any continuous sound, no matter how complex, can be represented as the sum of sinusoidal components, because each pure tone has a particular amplitude and time relationship relative to the other pure tone components.

      Spectrum analysis is the process of converting the time-domain signal to a frequency-domain signal, showing how different frequency components contribute to the sound. Fourier transformation is the mathematical function that is often used to convert the time-domain form to a frequency-domain representation or spectrum. An individual spectrum contains no information about temporal changes in frequency composition of the spectrum. A spectrogram shows how the frequency composition of a signal changes over time. The spectrograms in this book are based on a Fast Fourier Transform algorithm. A Hanning window was used to eliminate the effects of background noise (Figure 38).

СКАЧАТЬ