Learn About Crabs & Relatives: Feeding, foods, & digestion


Decapod crustaceans have a variety of feeding modes including CARNIVORY, considered in this section, and HERBIVORY, SUSPENSION-FEEDING, and OMNIVORY/SCAVENGING considered in other sections.

This part of CARNIVORY deals with the Dungeness crab METACARCINUS MAGISTER, while related topics of MECHANICS OF CRUSHING and carnivory in CANCER PRODUCTUS are dealt with elsewhere.

Metacarcinus magister

Research study 1

photograph of Dungeness crab Cancer magister

In Boundary Bay, British Columbia Metacarcinus magister subsists on crustaceans (shrimps, barnacles, small crabs, isopods), molluscs (clams, mussels), polychaetes, and other small invertebrates. Only rarely are seaweeds eaten. A mark-recapture study in the Prince Rupert, British Columbia area in 1930 reveals that the crabs tend to be home-bodies. The greatest distance travelled in a year by any individual in the study group is 1.6km. MacKay 1942 Bull Fish Res Bd Can 62: 1.

Metacarcinus magister in an aquarium tank with shell fragments of prey mussels 0.6X

Research study 2

histogram of % prey frequencies in gut contents of Dungeness crabs Cancer magister in Humboldt Bay, California

A large and comprehensive analysis of gut contents of Metacarcinus magister in Humboldt Bay, California and nearby open-coast locations discloses 40 different items eaten. The author combines the prey items into several major faunal groups, of which clams (33% of all prey items), crustaceans (31%), and fishes (22%) are the most preferred. One finding of interest is that cannibalism is quite rare in this area in comparison with higher incidences reported in other areas. Not surprisingly, prey preferences vary with depth, most notably with polychaetes being eaten more by crabs in deeper waters (72-90m), and fishes more by crabs in shallower waters (0-18m). The reason for the latter is not known, although the author suggests that in the more murky shallows, crabs may have greater success in capturing fishes than in deeper, clearer waters. Gotshall 1977 Calif Fish Game 63: 43.

NOTE a total of 208 stomachs are examined over a 3-y period: 39 from Humboldt Bay and the remainder from the open coast. Of the 208, only 28 stomachs are recorded as being empty. Size range examined is 67-200mm carapace width. Fish species are identified by otolith morphology

Research study 3

Dungeness crabs Metacarcinus magister collected from the Strait of Juan de Fuca area eat a variety of shellfish, including clams, mussels, and barnacles. Pearson et al. 1979 J Exper Mar Biol Ecol 39: 65.

Research study 4

Prey or scavengable matter is perceived at a distance by Dungeness crabs Metacarcinus magister through chemosensory stimulation of the antennules or close in by chemotactile stimulation of the antennae, walking legs, claws, or maxillipeds. The antennules flick almost continuously, at a rate of about 30 . min-1 The action splays out the sensory bristles and exposes the chemosensory surfaces to maximum contact with passing water. If clams Protothaca staminea are freeze-dried, powdered, made up in a series of dilutions with seawater, and directed onto crabs in aquarium tanks via a small-bore tube, the crabs can : 1) detect the clam extract at a concentration of 10-10 g . liter-1, and 2) dig for perceived prey at a concentration of 10-2 g . liter-1. Note that the detection threshold, defined as 50% response by the test crabs, is 8 orders of magnitude smaller than the feeding/digging threshold. Pearson et al. 1979 J Exper Mar Biol Ecol 39: 65.

NOTE lit. “chemical touch”. The sensory organs involved are scattered over the surface of the exoskeleton. Each consists of a chemosensory pit in the exoskeleton with an associated touch-sensitive bristle. Humans have no comparable chemosensory organ. Our fingers have good touch perception, but no chemical sensitivity and, while our tongues are sensitive to both chemicals and touch, they are generally not adapted for employment outside of the mouth

NOTE detection is defined by the author as an abrupt change in orientation of the antennules within 30sec from application of the clam extract, accompanied by an increase in rate of flicking for an ensuing 1-min period that is 1.5-times the rate exhibited 1min before application of the extract

Research study 5

The way that crabs shred and masticate their prey makes identification of soft-bodied species difficult, but sufficient hard parts may remain to estimate frequency of occurrences and masses. In a study on Metacarcinus magister from Gray’s Harbor, Washington, researchers integrate the frequency of prey occurrence, percentage of total number of prey items, and percentage of total mass of all prey contents in individual stomach samples to obtain an overall indication of dietary importance, which they term index of relative importance (IRI). IRI is shown here for small M. magister (31-58mm carapace width) collected at 2 sites in springtime. The data show that conspecifics (M. magister, 34% IRI) represent the most important dietary component. Other important groups, in order, are bivalves (28% IRI) and acorn barnacles (25% IRI), followed by other taxa of less importance. The large representation of conspecifics in the diet further emphasises the importance of cannibalism to this species, at least in this geographic area.Interestingly, fishes figure much less importantly in the diets of these Grays Harbor crabs (1%) than in Humboldt Bay crabs (Research Study 2 above). No data are provided on relative occurrence of the prey items in the habitat, so no measure of food preference can be derived from the data. The authors note that this is the first use of the IRI method of stomach analysis for a crustacean. Stevens et al. 1982 Mar Biol 72: 135.

NOTE IRI is calclated as: (% total number of prey items in a stomach sample + % of total dry mass of prey items in a stomach sample) x (% of stomachs containing the prey item). Overall "% gut-ratio content" is 0.95 (i.e., most of the guts are 95%full)

NOTE a moment’s scrutiny of the data discloses separate entries for “Decapoda” and “Cancridea”, which could be combined with “Metacarcinus magister” to make a slightly larger overall total for “crabs”. Similarly, “Macoma sp.”, “Tellinidae”, “Myidae”, and “Mya arenaria” could be combined with "Bivalves" to make a larger overall total for “bivalves”. The authors explain that the unspecified category “Bivalves” and several similar entries indicate that many of the prey items are not identifiable

Research study 6

Large Cancer spp. and other crabs readily dig up and eat shallow-burrowing clams. The prey is sensed both by touch and by perception of chemical emanations from the burrow. Studies in Coos Bay, Oregon show that juvenile M. magister (10-30mm carapace width) will readily seek out and eat a variety of bivalve species as long as they are small-sized. If Cancer juveniles are place in an aquarium containing sand with the small bivalve Transennella tantilla (<5mm shell length, see photo) they immediately use their chelae and walking legs to sift through the sediment. When a clam is found it is brought to the mouth region and rotated into proper position for separating the shells, a process that photograph of clam Transennella tantillatakes a few seconds to several minutes. Asson-Batres 1986 Calif Fish Game 72: 144.

NOTE for example, cockles Clinocardium nuttalli, and other bivalves Cryptomya sp., Protothaca staminea, and Macoma sp. are readily eaten at shell lengths of 5-13mm

NOTE it is not clear from the author’s description how this is accomplished. However, it appears that separation of the valves may sometimes involve crushing them

Bivalve Transennella tantilla

Research study 7

graph showing times for Dungeness crabs Cancer magister to break the shells of littleneck clams Protothaca staminea

Dungeness crabs Metacarcinus magister dig up littleneck clams Protothaca staminea, crack their shells with their claws, and eat them. One would predict that a crab that forages optimally would obtain most nutrients and energy from the largest possible clam, and would preferentially select the biggest prey possible. However, results of a study in British Columbia show that this is not the case. The largest clams simply take too long to break open. Thus, a crab that can break a 3-cm clam in 120sec might take 400sec to break a 4-cm clam, or not be able to break it at all (see graph). Since most of the energy cost to the crab is in breaking open the shell of its prey, it is energetically better off to eat small clams. The small clams are also easier for the crab to dig up because they live proportionally shallower. Finally, In laboratory feeding tests over 2-5d periods, adult crabs (presumably experienced at eating clams) given a selection of 3 size-classes of Protothaca will preferentially select and eat the smaller sizes. Juanes & Hartwick 1990 Ecology 71: 744.

Research study 8

Cannibalism is not uncommon amongst crabs and has been reported many times for Dungeness crabs Metacarcinus magister. In Grays Harbor, Washington where oyster cultch is placed on mud habitat to provide 3-dimensional refuge for young-of-the-year Metacarcinus magister, cannibalism by older crabs living in the cultch may explain recruitment failure of later cohorts of megalopae. Laboratory studies show that predation by older crabs on young-of-the-year recruits (9-18mm carapace width) is highest in mud habitat, intermediate in low-density oyster-shell cultch, and lowest in high-density oyster-shell cultch (see histogram). The study confirms that "contrived" shell habitat enhances production of young-of-the-year crabs by increasing their survival. Fernández 1999 Mar Ecol Progr Ser 182: 221.

NOTE low density = 20% shell cover on mud; high density = 100% shell cover on mud

NOTE the histogram shows mortality of 0+ (young-of-year) crabs in 3 habitats by 3 size classes of conspecific crabs. Overall mortality is lowest for 0+ crabs being eaten by 0+ conspecifics because the "predators" are only slightly larger than their "prey" (bars at extreme Left of histogram). 1+ = 1-yr-old crabs; 2+ = 2-yr-old crabs

Research study 9

In a study of movement of crabs at Bamfield, British Columbia, Dungeness crabs Metacarcinus magister are fitted with ultrasonic telemetry tags and tracked during a single tidal cycle. The extent of wandering in this species is strongly dependent on nutritional status, with unfed crabs moving up to 1370m during a 6-h period and fed crabs settling and moving little during the first 48h after release. Bernatis et al. 2007 Mar Biol 150: 941.