Ecological Constraints for Sound Communication: From Grasshoppers to Elephants 75
the noisy environment of the insect. Figure 8 shows representative periods of
neural activity of the same preparation placed within the rainforest (Panama, Barro
Colorado Island) at two times of day, at 16.00 and 18.30 h, and in response to a
signal broadcast from a distance of 10 m. In the early afternoon, under ambient
noise levels of 40 dB SPL and less, a burst of activity in this neuron was always
associated with the broadcast signal (100% hits). It never happened that such a
burst of activity occurred without a signal, i.e., a receiver using the burst criterion
would never decide incorrectly on the presence of a signal (no false alarms).
However, after another 2.5 h, shortly after sunset, the noise level had increased to
more than 60 dB SPL. The same experiment now led to quite different results (Fig.
SA, lower line).
There is so much afferent activity due to noise from many other sources in the
habitat that - based on the previous burst criterion - the receiver produced 1397
false alarms in a period of 5 min, indicating the presence of a signal where, in fact,
there had been none. The signal detection performance deteriorates with increasing
distance to the sender, then the rate of correct hits decreases, but at the same time
the rate of false alarms stays high. In addition, at greater distances the number of
cases where the neuron failed to generate a burst of action potentials in response to
a signal increases (miss).
Results of such experiments also point to one solution to many communication
systems to counteract noise, namely redundancy by signal repetition. When the
signal is repeated at a high rate, both in low and high ambient noise levels the rate
of hits is almost 100%, and there are only few false alarms even at high noise
levels (Fig. 8B). Other solutions for avoiding mistakes in signal detection is the
temporal segregation of calling activity of several signaling individuals, or the
partitioning of the carrier frequencies of their calls, thus allowing for "private
channels" of communication (Moore et al. 1989; Narins 1995; Riede 1995).
References
Arak A, Eiriksson T (1992) Choice of singing sites by male bushcrickets
(Tettigonia viridissima) in relation to signal propagation. Behav Ecol Sociobiol
30:365-372
Belwood J, Morris GK (1987) Bat predation and its influence on calling behavior
in neotropical katydids. Science 238:64-67
Cade WH ( 1975) Acoustically orienting parasitoids: fly phonotaxis to cricket
song. Science 190:1312-1313
Cade WH (1981) Field cricket spacing, and the phonotaxis of crickets and
parasitoid flies to clumped and isolated cricket songs. Z Tierpsychol 55:365-375
Dusenbery DB (1992) Sensory Ecology. WH Freeman and Company, New York
Embleton TFW (1996) Tutorial on sound propagation outdoors. J Acoust Soc Am
100:31-48
the noisy environment of the insect. Figure 8 shows representative periods of
neural activity of the same preparation placed within the rainforest (Panama, Barro
Colorado Island) at two times of day, at 16.00 and 18.30 h, and in response to a
signal broadcast from a distance of 10 m. In the early afternoon, under ambient
noise levels of 40 dB SPL and less, a burst of activity in this neuron was always
associated with the broadcast signal (100% hits). It never happened that such a
burst of activity occurred without a signal, i.e., a receiver using the burst criterion
would never decide incorrectly on the presence of a signal (no false alarms).
However, after another 2.5 h, shortly after sunset, the noise level had increased to
more than 60 dB SPL. The same experiment now led to quite different results (Fig.
SA, lower line).
There is so much afferent activity due to noise from many other sources in the
habitat that - based on the previous burst criterion - the receiver produced 1397
false alarms in a period of 5 min, indicating the presence of a signal where, in fact,
there had been none. The signal detection performance deteriorates with increasing
distance to the sender, then the rate of correct hits decreases, but at the same time
the rate of false alarms stays high. In addition, at greater distances the number of
cases where the neuron failed to generate a burst of action potentials in response to
a signal increases (miss).
Results of such experiments also point to one solution to many communication
systems to counteract noise, namely redundancy by signal repetition. When the
signal is repeated at a high rate, both in low and high ambient noise levels the rate
of hits is almost 100%, and there are only few false alarms even at high noise
levels (Fig. 8B). Other solutions for avoiding mistakes in signal detection is the
temporal segregation of calling activity of several signaling individuals, or the
partitioning of the carrier frequencies of their calls, thus allowing for "private
channels" of communication (Moore et al. 1989; Narins 1995; Riede 1995).
References
Arak A, Eiriksson T (1992) Choice of singing sites by male bushcrickets
(Tettigonia viridissima) in relation to signal propagation. Behav Ecol Sociobiol
30:365-372
Belwood J, Morris GK (1987) Bat predation and its influence on calling behavior
in neotropical katydids. Science 238:64-67
Cade WH ( 1975) Acoustically orienting parasitoids: fly phonotaxis to cricket
song. Science 190:1312-1313
Cade WH (1981) Field cricket spacing, and the phonotaxis of crickets and
parasitoid flies to clumped and isolated cricket songs. Z Tierpsychol 55:365-375
Dusenbery DB (1992) Sensory Ecology. WH Freeman and Company, New York
Embleton TFW (1996) Tutorial on sound propagation outdoors. J Acoust Soc Am
100:31-48