Food choices of land hermit crabs (Coenobita
compressus H. Milne Edwards) depend on past experience
Robert W. Thacker1 1995
Department of Biology, University of Michigan.
Ann Arbor, MI 48109, USA
Abstract
Land hermit crabs are scavengers that use olification
to locate their foods. Although they are scavengers, land
hermit crabs do not select their foods randomly. This study
demonstrates that land hermit crabs (Coenobita compressus
H. Milne Edwards) display negative preference indication when
feeding on natural foods, preferring foods that they have
not experienced recently. These induced feeding preferences
were measured at a population level by recording crab choices
among food odors in the field with olfactory attraction assays.
When the abundances of foods in their habitat were altered,
crabs preferred the odors of foods that were less abundant.
Induced feeding preferences were measured at an individual
level by recording crab choices among 3 foods in a laboratory
choice assay. Crabs more frequently chose foods that they
had not experienced during the previous 24 hours. As a consequence
of this behavior, land hermit crabs consume a broader diet
which may result in crabs obtaining a more nutritionally balanced
diet or limiting their exposure to toxins.
Keywords: Diet selection; Learning; Foraging;
Crustacean
1. Introduction
Optimal foraging theory predicts that a forager
should select its diet based on the profitability of available
food items, with profitability usially expressed as the energy
content of the food item per unit of searching and handling
time necessary to consume the food item (Stephens and Krebs,
1986). The predictions of this diet model change if the profitability
of a given food type is not constant. Searching and handling
times can change with experience, altering the profitability
ranking of food types (Hughes, 1979)
If handling times decrease with experience,
then food types with which an individual has had more experience
may be ranked higher (Croy and Hughes, 1991). The rankings
of food types may change if the nutritional value of a food
type depends on the nutritional status of the consumer (Stephens
and Krebs, 1986). A protein-deprived animal may rank an otherwise
less profitable type higher if it has a higher protein content
relative to other available food types (Simpson and White,
1990). The rankings of food types may be dependent on the
amounts of toxins or secondary compounds found in a food item
and whether a forager has recently consumed those compounds
(Stephens and Krebs, 1986). A variety of animals, including
grasshoppers (Taeniopoda eques, Bernays et al., 1992) snails
(Arianta arbustorum, Speiser and Rowell-Rahier, 1993) and
sea hares (Dolabella auricularia, Pennings et al., 1993),
may limit their exposure to toxins found in different food
types by consuming mixed diets, such that their preferences
alternate among different food types.
A narrower diet that results from experience
has been referred to as an induced feeding preference (Jermy
et al., 1968). Since a broader diet that results from experience
also can be considered an induced feeding preference, these
2 forms of preference induction can be termed positive preference
induction, which results in a more narrow diet, and negative
preference induction, which results in a broader diet. If
an individual's searching and handling times decrease for
a given food type, it will be an optimal forager if it narrows
its diet by selecting that type more frequently than other
food types (Hughes, 1979), thereby displaying positive preference
induction. If relative nutritional values change, an optimal
forager will not select another item of a recently encountered
food type, but will broaden its diet to include a larger number
of food types (Stephens and Krebs, 1986), thereby displaying
negative preference induction.
Negative preference induction has been documented
in a variety of organisms, including grasshoppers (Tueniopoda
eques, Bernays et al., 1992), locusts (Locusta migratoria,
Simpson and White, 1990), rats (Rattus norvegicus, Rozin,
1976), chameleons (Chamaeleo senegalensis, Eason, 1990) and
garter snakes (Thamnaphis sirtalis, Burghardt, 1992). Negative
preference induction can be considered a form of dietary self-selection,
which has been reviewed recently for insects (Wauldbauer and
Friedman, 1991). Despite the variety of taxa in which negative
preference induction has been observed, this behaviour has
been reported only once for a crustacean. Wellins et al. (1988)
maintained the land hermit crab Coenobita rugosus
in the laboratory on single-item diets of apple, banana, or
horse manure. Subsequently, crabs were offered choices among
these foods in odor preference trials. Crabs chose items not
included in their maintenance diet more frequently than expected
by a random model of diet choice, thereby displaying negative
preference induction. However, this behavior has not been
documented in the natural habitat of land hermit crabs.
Land hermit crabs (Coenobita
spp.) have been reported to be scavengers, eating plant and
animal material washed into intertidal areas and the leaves
and fruits of mangroves, coconuts and other plants (Burggren
and McMahon, 1988). Ball (1972) found that C.compressus
eats fungi, dead plants and animals, fruits, feces, and economically
important foodstuffs such as plantain and rice. The mechanics
of land hermit crab feeding have been well-described (Dunham
and Gilchrist, 1988). Kurta (1982) found that C. compressus
oriented visually to food items and groups of hermit crabs.
This social facilitation of foraging could improve the ability
of land hermit crabs to observe food from a distance (Kurta,
1982).
Many decapod crustaceans use chemosensation
to locate their food from a distance (Rittschof, 1992), including
aquatic decapods (Pearson et al, 1979); Derby and Atema, 1981),
and semi-terrestrial decapods (Wellins et al., 1989). The
morphological structure and the antennules of C. clypeatus
is similar to that of the chemoreceptive organs of other
decapods (Ghiradella et al, 1968). C. compressus has
been shown to detect odors from feces, fruits and fish from
a distance of at least 5m (Dunham and Gilchrist, 1988). Rittschof
and Sutherland (1986) demonstrated that C. rugosus
can locate potential foods from a distance by detecting volatile
chemical cues and can detect nonvolatile compounds by contact
chemoreception.
In this study, I determined if the land hermit
crab C. compressus H. Milne Edwards could
locate its common food items in the field by using chemoreception.
I hypothesized that these crabs would be able to detect food
items from a distance using olfaction. This information was
used in designing subsequent field and laboratory assays.
To assay induced food preferences, I first determined whether
olfactory attraction could be used to measure the food preferences
of field populations of land hermit crabs. Next, I measure
the food preferences of field populations of land hermit crabs.
Next, I manipulated in the field by changing the dietary experience
of field populations. Sequences of food choices have been
used to measure food preferences and dietary mixing in grasshoppers
(Bernays et al., 1992). To determine if this technique could
be applied to land hermit crabs, I monitored sequences of
food choices by land hermit crabs in a laboratory setting.
Based on the results of Wellins et al. (1988), I hypothesized
that land hermit crab food preferences would depend on past
experience in both the field and laboratory experiments. Specifically,
I predicted that the crabs would display negative preference
induction, i.e., crabs would be less attracted to foods that
they had recently experienced.
2. Materials and methods
2.1. Study site and organisms
These studies were conducted at the Achotines
Laboratory of the Inter-American Tropical Tuna Commission
(Los Santos Province, Republic of Panama) between January
and April, the dry season, of 1992 and 1993.
Cite as:
The Crab Street Journal: Articles
