LEARNINU BY MARINE INVERTEBRATES
39
others do not (Stephenson, 1936). On balance, there now seems to be
no doubt that certain species, (Patella vulgata, P . granularis L. and
Acmaea scabra Gould, for example) do regularly return to " scars )'
which remain their home bases for long periods of time. Siphonaria,
among the pulmonates, appears to do the same (Ohgushi, 1966). The
evidence for homing arises from direct observation of the tracks of
individual limpets (Hewatt, 1940 ; Thorpe, 1963), from repeated
observations of individuals found foraging away from home and
subsequently on their bases (Stephenson-see Thorpe, 1963), and from
the shape of the shell, which grows to fit the animal's particular home
scar. (Hewatt (1940) filed away the edge of a limpet's shell and it
returned to its old base, i.e. it seems not just to be a matter of the
animal wandering about until it finds a place where it fits.) Return to a
home base can be taken as evidence that an animal has learned,
provided that it can be shown that the animal is not simply homing on a
beacon (the sight of the home, or its smell) or returning along a track
that it lays on the outward journey. The former type of explanation is
difficult to eliminate experimentally, since to do so necessitates aasumptions about the stimuli the animal is detecting. But it appears unlikely
that limpets see or smell their home from a distance since their limited
visual horizon must mean that they move rapidly out of sight of it and,
at least for forms that normally move when covered by water, the usefulness of chemical cues must be very limited in an environment where
current direction is so variable. Return along the outward track has
often been observed but is by no means invariable (for references, see
Fischer, 1960 ; Thorpe, 1963). Moreover, attempts to upset homing by
scrubbing out the track (Davis, 1896) or altering the topography
(Pieron, 1909 ; Hewatt, 1940) have not prevented return by most of the
animals concerned.
Displacement experiments, mainly carried out with Patella, have
given variable results, and in general, it seems that a Patella, forcibly
removed from its scar and replaced more than 10 cm from home, has
less than a 60% chance of finding its way home again (Fischer, 1898 ;
Morgan, 1894 ; Pieron, 1909). Unfortunately these experiments have
never been done with individuals whose normal feeding range was
known or estimated from the condition of the algae around them.
The performance of the naked pulmonate Onchidiurn, which returns
not to a scar but to a " nest," a hole in the rocks where a dozen or more
individuals may live together, is evidently much more reliable than
that of the prosobranch limpets or Siphonaria. Arey and Crozier
( 192 1) report almost invariably successful homing in many experiments
with individual Onckidium displaced 60-100 cm, including returns to
39
others do not (Stephenson, 1936). On balance, there now seems to be
no doubt that certain species, (Patella vulgata, P . granularis L. and
Acmaea scabra Gould, for example) do regularly return to " scars )'
which remain their home bases for long periods of time. Siphonaria,
among the pulmonates, appears to do the same (Ohgushi, 1966). The
evidence for homing arises from direct observation of the tracks of
individual limpets (Hewatt, 1940 ; Thorpe, 1963), from repeated
observations of individuals found foraging away from home and
subsequently on their bases (Stephenson-see Thorpe, 1963), and from
the shape of the shell, which grows to fit the animal's particular home
scar. (Hewatt (1940) filed away the edge of a limpet's shell and it
returned to its old base, i.e. it seems not just to be a matter of the
animal wandering about until it finds a place where it fits.) Return to a
home base can be taken as evidence that an animal has learned,
provided that it can be shown that the animal is not simply homing on a
beacon (the sight of the home, or its smell) or returning along a track
that it lays on the outward journey. The former type of explanation is
difficult to eliminate experimentally, since to do so necessitates aasumptions about the stimuli the animal is detecting. But it appears unlikely
that limpets see or smell their home from a distance since their limited
visual horizon must mean that they move rapidly out of sight of it and,
at least for forms that normally move when covered by water, the usefulness of chemical cues must be very limited in an environment where
current direction is so variable. Return along the outward track has
often been observed but is by no means invariable (for references, see
Fischer, 1960 ; Thorpe, 1963). Moreover, attempts to upset homing by
scrubbing out the track (Davis, 1896) or altering the topography
(Pieron, 1909 ; Hewatt, 1940) have not prevented return by most of the
animals concerned.
Displacement experiments, mainly carried out with Patella, have
given variable results, and in general, it seems that a Patella, forcibly
removed from its scar and replaced more than 10 cm from home, has
less than a 60% chance of finding its way home again (Fischer, 1898 ;
Morgan, 1894 ; Pieron, 1909). Unfortunately these experiments have
never been done with individuals whose normal feeding range was
known or estimated from the condition of the algae around them.
The performance of the naked pulmonate Onchidiurn, which returns
not to a scar but to a " nest," a hole in the rocks where a dozen or more
individuals may live together, is evidently much more reliable than
that of the prosobranch limpets or Siphonaria. Arey and Crozier
( 192 1) report almost invariably successful homing in many experiments
with individual Onckidium displaced 60-100 cm, including returns to