Vibration Communication in Vertebrates
139
the sandfish lizard appears to be able to localize the vibratory cues while buried,
and thereby gains an additional advantage over his prey.
3.6 Eremitalpa granti namibensis - the Namib Desert Golden
Mole
The 18 species of golden moles, placed in the family Chrysochloridae, comprise
seven genera that are found exclusively in sub-Saharan Africa, south of a line
from Cameroon to Uganda. They are image-blind (DuBost 1968), and for the most
part nocturnal, surface-foraging mammals, some genera possessing massively
hyper-trophied mallei which presumably confer low-frequency, substratevibration sensitivity through inertial bone conduction (Mason 1998). (Inertial bone
conduction results from the vibration of the skull which results in relative
movement between the stapes footplate and the oval window due to ossicular
inertia.)
Unlike many other subterranean mammals, the Namib Desert golden mole,
Eremitalpa granti namibensis, lacks a permanent burrow system and can forage
five or more km per night (Holm 1969) for its prey on the dune surface (Fielden et
a!. 1992). Foraging animals typically move between miniature sand dunes or
mounds topped with dune grass (Stipagrostis) which contain the large majority of
living biomass in the Namib Desert (Fielden et a!. 1990). Recently, two
hypotheses were directly tested regarding the foraging strategies adopted by E. g.
namibensis: that foraging is random in these animals, that is, encounters with food
resources are purely stochastic events, and that the mole's ability to detect prey is
effective over short distances only (Narins eta!. 1997).
To test the first hypothesis, a simple statistical test was performed to determine
the probability that the golden mole would visit (for example), seven food patches
in succession if it were indeed foraging randomly. For every foraging trail tested,
the calculated cumulative probabilities of randomly encountering seven particular
mounds in succession were so small that the hypothesis of random foraging could
safely be abandoned. Consistent with this result, foraging trail segments between
visited mounds appear remarkably straight, showing no evidence of a random
search pattern. Rather, the data suggest an alternative working hypothesis, namely
that the golden mole exhibits sensory-guided foraging behavior.
Foraging trails consist of footprints interspersed with characteristic sand
disturbances in which the animal either head dips or "swims" under the sand. The
function of these behaviors is not known but we suspected that they may be used
to obtain a seismic "fix" on the next mound to be visited. To test this, the local
seismic vibrations both on the top of a mound and in the flat sand at a distance of
at least 20 m from the mound under test were measured. Resulting representative
velocity spectra are shown in Fig. SA. The spectrum recorded on the flat shows a
relatively low-amplitude peak at about 120 Hz, whereas the spectral peak recorded
from the mound is nearly 17 dB greater in amplitude and centered at 310 Hz. The
139
the sandfish lizard appears to be able to localize the vibratory cues while buried,
and thereby gains an additional advantage over his prey.
3.6 Eremitalpa granti namibensis - the Namib Desert Golden
Mole
The 18 species of golden moles, placed in the family Chrysochloridae, comprise
seven genera that are found exclusively in sub-Saharan Africa, south of a line
from Cameroon to Uganda. They are image-blind (DuBost 1968), and for the most
part nocturnal, surface-foraging mammals, some genera possessing massively
hyper-trophied mallei which presumably confer low-frequency, substratevibration sensitivity through inertial bone conduction (Mason 1998). (Inertial bone
conduction results from the vibration of the skull which results in relative
movement between the stapes footplate and the oval window due to ossicular
inertia.)
Unlike many other subterranean mammals, the Namib Desert golden mole,
Eremitalpa granti namibensis, lacks a permanent burrow system and can forage
five or more km per night (Holm 1969) for its prey on the dune surface (Fielden et
a!. 1992). Foraging animals typically move between miniature sand dunes or
mounds topped with dune grass (Stipagrostis) which contain the large majority of
living biomass in the Namib Desert (Fielden et a!. 1990). Recently, two
hypotheses were directly tested regarding the foraging strategies adopted by E. g.
namibensis: that foraging is random in these animals, that is, encounters with food
resources are purely stochastic events, and that the mole's ability to detect prey is
effective over short distances only (Narins eta!. 1997).
To test the first hypothesis, a simple statistical test was performed to determine
the probability that the golden mole would visit (for example), seven food patches
in succession if it were indeed foraging randomly. For every foraging trail tested,
the calculated cumulative probabilities of randomly encountering seven particular
mounds in succession were so small that the hypothesis of random foraging could
safely be abandoned. Consistent with this result, foraging trail segments between
visited mounds appear remarkably straight, showing no evidence of a random
search pattern. Rather, the data suggest an alternative working hypothesis, namely
that the golden mole exhibits sensory-guided foraging behavior.
Foraging trails consist of footprints interspersed with characteristic sand
disturbances in which the animal either head dips or "swims" under the sand. The
function of these behaviors is not known but we suspected that they may be used
to obtain a seismic "fix" on the next mound to be visited. To test this, the local
seismic vibrations both on the top of a mound and in the flat sand at a distance of
at least 20 m from the mound under test were measured. Resulting representative
velocity spectra are shown in Fig. SA. The spectrum recorded on the flat shows a
relatively low-amplitude peak at about 120 Hz, whereas the spectral peak recorded
from the mound is nearly 17 dB greater in amplitude and centered at 310 Hz. The