In crabs and their relatives the metamorphic moult is usually (but not always) considered as the change from the zoea larva to megalopa. After a few more moults, the megalopa moults to the first crab instar (juvenile) stage. Most research studies done on settlement and recruitment of crabs concerns Dungeness crabs Cancer magister and this species is treated here separately.  Studies on other species are considered in a separate section entitled OTHER DECAPOD SPECIES. 

NOTE for this reason most authors refer to the megalopa as a developmental "stage" and not a larva

NOTE  studies of a strictly mariculture nature for this commercial species are mostly omitted

Laboratory tests on larvae of Cancer magister show that both zoeae and megalopae are perceptive to changes in light, gravity, and pressure.  The first two are unsurprising because of the presence from an early age of well-developed eyes and statocysts, but pressure-sensitive organs are not known in crustacean larvae.  Plankton tows through 24-h periods show that zoea larvae migrate vertically from 25m depths during the day to the sea surface at night.  Early zoeal instars are deeper during the day than later ones, and megalopae tend to remain at the surface both day and night.  The author suggests that such behaviours put the larvae in position to be transported shorewards by currents and prevailing winds.  Jacoby 1982 Mar Behav Physiol 8: 267.

The major Dungeness crab fishing areas on the west coast of the U.S. are indicated on the map. Further north these fishing areas extend into Puget Sound, Washington and to Haida Gwaii (Queen Charlotte Islands), British Columbia.  In the southern regions during the middle of the last century and possibly even now, catches of Cancer magister cycled on a 9-10y pattern. The explanation for this, as suggested by researchers, is that after hatching in early winter the larvae are carried northwards in the Davidson Current (about 90km in width).  In springtime (Mar/Apr) when the winds shift to the south, the Davidson Current weakens on the surface and is replaced by a southwards flow.  This current may tend to move the larvae and meglalopae back onshore.  Although the authors at the time of publication do not have sufficient data to determine how wind and current patterns actually affect recruitment patterns in C. magister, the study is one of the first of its kind to attempt such a correlation.  Johnson et al. 1986 Can J Fish Aquat Sci 43: 838.

A later study on wind effects on distribution of megalopae of Cancer magister along the west coast of North America (Washington-California) shows good correlation of onshore water transport with the nearshore abundance.  The study uses 4y of data collected from 120 offshore stations distributed from 40-48^oN latitude (mid-north California to northern Washington).  Note in the histograms that when onshore transport values are high, nearshore (<50km) densities of megalopae are also high, and vice versa.  Hobbs et al. 1992 Can J Fish Aquat Sci 49: 1379.

NOTE onshore-transport distances are calculated for the 45-d period prior to sampling for megalopae

A large-scale¹ series of plankton collections off the coasts of Washington, Oregon, and northern California provide good information on development and daily migrations of larvae of Cancer magister.  Results show that late-stage megalopae are most abundant in the neuston² at night, with occasional peaks in the afternoon and early evening. During the day the larvae hang out in deeper depths (not sampled in the study).  Early-stage megalopae are less abundant in the neuston, but weakly display the same pattern of diel migration as late-stage forms.  Zoeae mostly reside below the depth of neuston, except for a 2-3h period in the evening. The extent of vertical migration is independent of cloud cover or sea state.  The metamorphosis³ from last-stage zoea to megalops occurs over a period of less than a month in early spring at any given latitude, and the process seems to proceed latitudinally from south to north. Most megalopae are found within 75km of the shore, but some drift out to more than 300km from shore.  Also of interest in the study is the extent to which changes in vertical position of the megalopae interact with vertical shear to result in different degrees of horizontal movement.  Hobbs & Botsford 1992 Mar Biol 112: 417.

NOTE¹ 5 annual cruises with sampling at 120 stations

NOTE² planktonic organisms associated with the air-water interface.  In some collections, 95% of developmental stages are megalopae of C. magister

NOTE³ the authors consider the moult from zoea to megalopa to be metamorphosis, rather than the more commonly held idea that metamorphosis occurs from megalopa to 1st-instar juvenile. The first view makes a lot of sense, and perhaps the issue needs to be clarified

Much west-coast research on Cancer magister has been conducted in Grays Harbor estuary, Washington.  The explanation for this focus begins in 1990 when dredging by the U.S. Army Corps of Engineers to widen and deepen the main navigation channel in the Harbor killed many thousands of Dungeness crabs (over 160,000 by one estimate).  A pilot enhancement project done shortly afterwards showed that addition of oyster-shell cultch led to significantly increased settlement and survival of Cancer magister.  The results of this pilot project became the impetus for the Corps in 1992 to construct 8ha of intertidal oyster-shell habitat. This turned out to be a highly successful enhancement operation, and represented the first such habitat mitigation for a decapod fishery.  Dumbauldt et al. 1993 Can J Fish Aquat Sci 50: 381; for a review of the project see Dumbauld et al. 2000 J Shellf Res 19 (1): 379.

In Grays Harbor Cancer magister mates in May-June, females extrude eggs onto their pleopods in September-October, and the protozoeae hatch Jan-Mar.  Larval life including the megalopa stage is 130-150d and settlement is in May-June of the following year, with sometimes a smaller pulse later in the year.  Metamorphosis involves a single moult to a 1st-instar juvenile.  After that, during their first year of life, an individual will moult several more times.  Studies over several years show that settlement numbers in Grays Harbor may vary 40-fold in intensity related to the extent of northward alongshore transport, which is often seasonally persistent.  The authors suggest that Washington populations may frequently depend on larvae originating in the south for their recruitment.  McConnaughey et al. 1992 Can J Fish Aquat Sci 49: 2028.

In Grays Harbor the major settlement of Cancer magister megalopae occurs in May-June.  In other areas, such as Puget Sound, Washington, additional, smaller, recruitments have been recorded.  In 1988, for instance, researchers record a primary settlement in May in areas of the Sound closest to the Strait of Juan de Fuca, a significantly smaller settlement in June around the Hood Canal, and an even smaller settlement in July/August widely distributed in Puget Sound from the south near Seattle north to the Canadian border. Larvae in early cohorts are larger-sized than in later cohorts, and may originate from outer coast stocks entering Puget Sound via the Strait of Juan de Fuca.  The authors note that such occurrences, however, are rare.  Dinnnel et al. 1993 Mar Biol 115: 53.

NOTE do the larvae recognise this boundary, do you think, or did the authors for some reason not go across to have a look?

Map showing the location of Grays Harbor and other
locations of interest with respect to settlement of
Dungeness crabs Cancer magister

Survival and, therefore, recruitment of Cancer magister from the major May-June settlement of larvae in Grays Harbor is significantly higher in “virgin” oyster-shell habitats than in ones already containing 2nd- and 3rd-instar juveniles, owing to cannibalism.  The larger crabs also out-compete the smaller ones for shelter, forcing the little ones to seek out new habitats and thus exposing them to other kinds of mortality.  Regulation of density in the oyster-shell habitats, then, appears to be primarily through cannibalism and space competition.  Fernandez et al. 1993 Can J Fish Aquat Sci 50: 2100; Iribarne et al. 1994 J Exp Mar Biol Ecol 183: 259; see also Iribarne et al. 1995 J Exp Mar Biol Ecol 192: 173.

Research on larval settlement and habitat selection by juvenile Cancer magister in Grays Harbor shows also that survival is enhanced in oyster-shell debris through protection from fish predators.  Use of tethered juvenile crabs reveal that in this area staghorn sculpins are principal predators, and the 3-dimensional interstices of the shells provides good protection for the tiny crabs.  Fernandez et al. 1993 Mar Ecol Progr Ser 92: 171.

As the megalopae enter shallow inshore water, either by swimming or floating, their primary function is to seek out suitable areas in which to settle. Lab studies on megalopae larvae of Cancer magister collected in Grays Harbor show that there is a significant tendency for the first moult to the juvenile instar to occur during darkness, presumably for protection.  The tendency is not strong (only 60-70%), but is consistent regardless of whether the larvae have been kept in a normal light/dark regime, or in constant-light or constant-dark regimes over a 4-d period of the experiment.  The authors add, however, that light is likely to be only one of several factors affecting moulting rhythms in crustaceans.  Fernandez et al. 1994 Mar Biol 118: 611.

NOTE  ambient 16h light: 8h dark from west-facing windows in the aquarium room

Detailed studies on the mechanisms favouring recruitment and survival of Dungeness crabs Cancer magister in different habitats in Grays Harbor confirm that shell substrata¹ are significantly preferred over mud, regardless of whether the shells are placed on mud or mud placed among shells (see SUBSTRATUM graph on Right).

Not surprisingly, location of habitat is important, with one of two test sites receiving significantly greater settlement of megalopae than the other (see LOCATION graph lower Right).   

Somewhat unexpected, however, is a considerable temporal difference in settlement of megalopae during “back-to-back” 3-day periods² (almost 4-fold in magnitude; see TEMPORAL graph on Left). 

Finally, the presence of predator-exclusion cages³ has no significant effect on the settlement of megalopae within or without the cages (see PREDATOR EXCLUSION graph bottom Left).

The authors suggest that spatial variation in larval abundance mainly owes to local differences in wind-driven surface currents.  There are other results presented in this detailed study, too many to include here, but these main sets of data suggest to the authors: 1) that use of artificial substrata in this way can provide a useful index of post-settlement density of megalopae in C. magister, and 2) that extremely small temporal scales should not be overlooked in future studies.  Eggleston & Armstrong 1995 Ecol Monogr 65: 193.

NOTE¹  test habitats consist of plastic trays filled with oyster shells or mud, respectively.  The substratum surface in these trays stands 9cm above the surrounding sediment surface.  Tests show that the elevated position of these substrata does not significantly affect settlement of megalopae over that on the natural sediment surface

NOTE²  for these particular tests the authors select a new-moon period predicted to coincide with a pulse of Dungeness crab megalopae into the estuary

NOTE³  exclusion cages are constructed of 12-mm galvanised mesh.  The mesh size restricted access of most staghorn sculpins Leptocottus armatus, a dominant predator of early-stage C. magister, but not of juvenile conspecific crabs that are known to cannibalise newly settled recruits.  The authors present convincing argument that cage effects (e.g., current modification) are unlikely to have influenced settling patterns in these experiments

Other studies at 5 sites in northern Puget Sound, Washington over a 3-y period indicate that recruitment of Cancer magister is largely from inland parental stocks (PUGET-SOUND COHORT), but with a smaller portion coming from outer-coast areas. These latter stocks (OCEAN COHORT) settle earlier in the season (late spring) at a larger first-instar size (7.2mm carapace width vs. 5.3 for the PUGET-SOUND COHORT), and grow more quickly owing to warmer seawater temperatures. In comparison, the PUGET-SOUND COHORT settles in late summer and grows little over the winter. 

Seasonal densities of crabs at all 5 sites tend to be higher in gravel/algae and eelgrass habitats than in open-sand habitats (see graph on Left). McMillan et al. 1995 Estuaries 18: 390.

Other research in Grays Harbor confirms that recruitment of Cancer magister is poor if shore crabs Hemigrapsus spp. are present in the oyster-shell cultch.  Laboratory studies at Shannon Point Marine Center, Washington show that Hemigrapsus will, indeed, eat juvenile C. magister, but only to significant extent when they are twice as large in carapace width as their intended prey (see graph on Right).  If the 2 protagonists are of equal size, then no predation occurs.

Are the Cancer megalopae able to perceive the presence of the predators either by scent or sight and take avoidance action?  This is tested in a Y-tube choice apparatus with trays of sand and eelgrass (preferred settling substrata) in either arm.  Prior to the water entering an arm of the apparatus, it either goes through an aquarium tank with 100 Hemigrapsus spp. present (experimental treatment) or an aquarium tank with Hemigrapsus absent (control treatment).  Runs consist of 400 Cancer magister megalopae being released into the stem of the apparatus and left for 48h to swim upstream to the sand/eelgrass settling sites.  In the first set of test results shown here the megalopae are able to see and scent the Hemigrapsus.  Counts of megalopae in the sand/eelgrass trays afterwards show that significantly more larvae choose the Hemigrapsus-absent channel/settling tray than the Hemigrapsus-present channel/settling tray indicating that they are responding negatively to scent and/or sight of the shore crabs (left pair of bars in the histogram).

When the larvae are tested in the absence of potential sight of the shore crabs, they avoid the Hemigrapsus-present side, indicating that they are responding to chemical cues (right pair of bars in the histogram). Another test, allowing sight of the shore crabs, but not their scent, yields results that are not significantly different from random.  This part of the study is not as convincing as the other parts, as it relies on an assumed visual stimulus that may or may not be present.  The study provides useful data for mariculturists.  Banks & Dinnel 2000 Crustaceana 73: 223.

NOTE  species used are nudus and oregonensis.  One hundred individuals corresponds roughly to estimated field densities per m² of cultch

Five species of Cancer crabs live on the northwest Pacific coast.  Transect surveys off the coast of Vancouver Island, British Columbia indicate that megalopae of C. magister and C. oregonensis are found in peak numbers in June about 50km offshore, with abundance shorewards decreasing markedly (within 28km). In fact, the megalopae appear to be concentrated in an area between two major coastal surface currents flowing in opposite directions (see orange dots on map). The authors note that given the current patterns off the Vancouver Island coast, these springtime developmental stages could have originated anywhere between northern California and southern Alaska.  Indeed, because of the barrier created by the Coastal Current, it is not known if any of these megalopae will ever settle in inshore waters. 

The localised pattern of distribution of the megalopae offshore is in deeper water during the day and at the surface during the night (see graph on Right). Note that abundances of Cancer oregonensis are 2 orders of magnitude greater than those of C. magister. Jamieson & Phillips 1988 Fish Bull 86: 525.

NOTE  the larvae are collected in special neuston (air-water interface) collectors – basically an open box with net attached that skims the surface to a depth of 35cm

NOTE  these are the northward-flowing Vancouver Island Coastal Current and the southward-flowing Outer Shelf-Break Current

Settling larvae of many species, including crabs, are affected by the hydrographic process of upwelling.  Upwelling, by moving surface water offshore, tends to reduce the chances of successful settlement of larvae, but its influence tends to differ with different species.  Thus, in areas around Bodega Bay, California, crabs (primarily Cancer spp.) tend to settle adventitiously whenever relaxation events from upwelling create warmer water temperatures, while sea urchins Strongylocentrotus spp. settle in a single large pulse during what the authors describe as atypical oceanographic conditions characterised mainly by lower than expected salinities and temperatures.

Note in the accompanying graphs that settlements of crab megalopae and sea-urchin plutei a during the 1992 period of study are negatively correlated.  The authors suggest that residence in different water masses and predation by crab megalopae on sea-urchin larvae, perhaps even after settlement, may have been involved.  Sea-urchin larvae are small and swim poorly, and are likely to be passively entrained in a water mass (of certain salinity and temperature).  In comparison, crab larvae are comparatively large, swim actively, and tend as terminal-stage megalopae to remain in surface waters where they are more likely to be carried inshore along with warmer water during relaxation events.  Wing et al. 1995 Limnol Oceanogr 40: 316.

NOTE  crab species are Cancer magister, C. productus, and C. antennarius, and sea-urchin species are Strongylocentrotus franciscanus and S. purpuratus. The 2 taxa are considered together here for convenience, as well as for their interactions

NOTE relaxation events are defined as strong increases in temperature lasting for several days.  Related studies of upwelling events and settlement of larvae of crabs Cancer spp. in the Bodega area of northern California provides evidence that in addition to direct onshore transport, there is alongshore (northward) transport during upwelling relaxation periods.  Wing et al. 1995 Mar Ecol Progr Ser 128: 199.

Recruitment of open-coast brachyuran species can be thought of as a one-stage process, in that megalopae arrive at a suitable coastal beach and then settle from the plankton.  For estuarine-inhabiting species such as crabs Cancer magister, however, the process is two-stage, for the larvae must first arrive in near-shore areas, then enter a nearby esturary.  How they do this is the subject of a study by researchers at the Oregon Institute of Marine Biology, Charleston.  Their approach is to capture megalopa larvae in light traps both offshore and within the estuary throughout the summer season, then to correlate abundances with physical variables associated with wind-driven transport, including upwelling and downwelling, and with tidal transport.  Results show that within-estuary abundances of megalopa larvae of Cancer magister are significantly correlated with outer-coast abundances, and with both wind- and tide-driven transport conditions.  The time lag to entry of the estuary is 0-4d.  The data suggest that megalopae are transported shoreward during upwelling and/or spring tides, and then carried into the estuary between spring and neap tides.  The study is one of the first of its kind to investigate simultaneously larval/juvenile abundances at outer coast and estuarine sites.  Miller & Shanks 2004 Mar Ecol Progr Ser 271: 267; see also Roegner et al. 2007 Mar Ecol Progr Ser 351: 177.

NOTE  the authors also provide data for several species of juvenile fishes, and for several other crabs, including Cancer oregonensis, C. productus, and hermit crabs Pagurus spp.  However, of the crabs, only Cancer magister provides statistically significant correlations between open-coast and estuarine abundances

Do the zoeal larvae of Cancer magister adjust their vertical positions in the water column to enhance access to shallow-water areas on favourable tides as is known for other species?  This is investigated  in Glacier Bay, Alaska by sampling 5 zoeal stages at different depths at varying distances from shore during May-September.  Results show no significant differences in larval abundances at inner- and outer-bay stations, nor at shallow or deep depths, suggesting that the answer, in these particular circumstances, is no.  The authors discuss their results in the context of findings of other researchers for the same and other species.  Park & Shirley 2008 Anim Cells Systems 12: 279 and Park & Shirley 2005 Estuaries 28 (2): 266; see also Park & Shirley 2008 Anim Cells Systems 12: 287 for information on hatching times for C. magister in areas of southeastern Alaska.

What about the megalopae of Cancer magister?  Do they adjust their vertical positions in the water column to facilitate transport into shallow-silled bays such as Glacier Bay, Alaska?  This question is addressed by researchers at the University of Alaska Fairbanks using light traps set at 1m and 10m depths at 4 sites at varying distances within the entrance to Glacier Bay (see map).  The entrance to the Bay is only 50m deep and the water column is mixed daily during maximum ebb and flood tides.  Results show that almost 100% of  the megalopae are found within 1m of the sea surface and almost none at 10m depth.  Interestingly, most larvae are captured during new and ¼ to ½ moon periods, and least are captured during ¾ and full moon periods, possibly representing an anti-predator strategy.  The authors offer these and various hydrodynamic explanations (mainly related to tides) for variations in megalopae supply to the Bay.  The larvae, then, are exported from the Bay as early zoeae and imported as megalopae, a pattern seen commonly in other areas of the west coast but apparently not previously recorded in Alaska. Herter & Eckert 2008 372: 181.

NOTE  the sites are located at 8, 12, and 18km from the mouth of the Bay

Recruitment of a species is affected both by biotic and abiotic factors, but the relative importance of each is often difficult to determine.  Researchers at the Oregon Institute of Marine Biology, Charleston investigate this for Dungeness crabs Cancer magister, and show that over a 5-yr period greater than 90% of the variation in adult population size is determined by success during the larval stage - as measured by number of megalopae settling 4yr earlier.  As to what major factor or factors governs larval success, the researchers review 30yr of commercial-catch data and determine that most of the variation in successful return of megalopae is explained by the timing of the spring transition.  In this case, an early transition leads to greater numbers of returning C. magister megalopae, but the same may not be true for other species of crabs.  The difference is explained by the unique life-history characteristics of C. magister.  During their 3-mo developmental period in late winter the zoea larvae move progressively offshore.  By the time the megalopal stage is reached, individuals may be more than 100km offshore.  At the time of spring transition, the megalopae are thought to be carried shorewards by the upwelling-driven deeper currents.  Once they get to near-shore environs, tidally based water movements carry the settling megalopae onto the shore.  When the transition occurs later in the spring, relatively more megalopae drift offshore and die.  Larvae of other species of crabs, such as pagurids and grapsids, have shorter developmental periods, stay closer to shore, and thus are not affected in the same way by the timing of the spring transition. The authors conclude that for C. magister, population size is principally set by the relative success of the larvae during pelagic development. Shanks & Roegner 2007 Ecology 88: 1726; for more on cross-shelf transport of crab megalopae, including those of Cancer spp., see Shanks 2006 Mar Biol 148: 1383.

NOTE  this is a seasonal shift in winds that drive the ocean currents along the U.S. west coast.  During the transition the winds shift direction and cause the coastal Davidson Current to reverse direction to the south.  This creates a general movement of surface water away from shore and upwelling of deeper water to replace it.  Mature Dungeness-crab megalopae are drawn shorewards in the deeper-moving currents

NOTE  although this is a nice story for C. magister, a later comparable investigation involving larvae of several diverse taxa including barnacles. sand dollars, mussels, ghost shrimps, and other species of crabs shows no connection between cross-shelf transport and either upwelling or downwelling. Shanks & Shearman 2009 Mar Ecol Progr Ser 385: 189.

The several months spent by larvae of Dungeness crabs Cancer magister in the plankton provides great potential for dissemination in currents.  Is it possible that larvae from the northwestern coasts of Washington and British Columbia can be transported northward to Alaska?  This hypothesis is tested by researchers at the University of Alaska by analysis of drift tracks of buoys and larval samples from the plankton at locations along the coast.  Results confirm the presence of zoea larvae in the northward flowing Alaska current (see map) and calculate that current speed is sufficient to land them on Alaskan shores in a late stage of development in May-June.  Such south-originating larvae would therefore arrive just as local larvae are beginning to hatch.  The authors confirm that temperature and other conditions are conducive for larval growth during the transport.  They also discuss the implications of a possible unidirectional gene flow between southern and northern populations of C. magister.  Park et al. 2007 Limnol & Oceanog 52 (1): 248.

NOTE  the presence of late-stage larvae in the coastal waters of southeastern Alaska during May-June had previously been noted, and ultimately contributed to the formulation of the hypothesis being tested here

Along the coasts of Washington, Oregon, and California the pattern of development of Cancer magister is for early zoeal stages to move out of estuaries, develop on the continental shelf, and return to nearshore areas as megalopae for settlement. The timing of the inshore pulses is fairly regular and predictable.  An investigation by researchers from the University of Alaska Fairbanks and Alaska Department of Fish and Game shows that a different pattern is exhibited by C. magister in the northern part of  southeastern Alaska.  Thus, owing to the geographic complexity of coastline and variable nature of the currents, there is no order to the arrival of pulses from the outer coast to the inland waterways.  The researchers propose multiple transport pathways for C. magister larvae in the northern region of southeastern Alaska.  Smith & Eckert 2011 Mar Ecol Progr Ser 429: 185.