SPATIAL AND TEMPORAL RECRUITMENT PATTERNS OF DUNGENESS CRAB IN THE NORTHEAST PACIFIC GLEN 5S. JAMIESON AND DAVID A, ARMSTRONG Jamieson, G.S.and Armstrong, D.A. 199] (1901; Spatial and temporal recruitment patterns af Dungeness crab in the northeast Pacific, Memoirs of the Queensland Museum 31; 365-381, Brisbane. ISSN 0079-8835, The Dungeness crab. Cancer magister, is the main crustacean species exploited in the northeastern Pacific from central California to Kodiak, Alaska, Abundance along the coast south of British Columbia has Nuctualed ina generally cyclical manner which a number of studies have tried to expla. but several unreluled competing, hypotheses currently remain preventing resolution of this question. A combination of mechanisms seems a likely pense and additional data appears necessary before understanding is achieved. There Ss geneyal agreement that {luctuation in calch is a reflection of variable yvear-class strength and recent studies of farval and (+ crab have investigated the importance of abiotic and biotic factors. Dungeness crab are regionally umque in that while many of their pelagic larvae move tens of kilometres offshore. they musi return to shallow waler Lo survive as juveniles. Oceanographic and meteorological conditions seem to be particularly influential in determining the magnitude ofd ispersal and unshore movement, and the conditions which allow successful settlement off Vancouver Island have now been described. Strong settle- ment does not necessarily equate wilh a strong year class al harvesting, though, and biotic factors primarily determine survival of juvenile crab, Finally, there is increasing evidence that the crab population in Georgia Strait and Puget Sound, i.e, in the 8inland sea9 inside of Vancouver Island, may be a distinct stock with dispersal, recruitment and population dynamics characteristics unique from the population found on the open outer coast. Comparison of common features betwee the two stocks is allawing evaluation of the relative importance of major factors influencing population abundance and ultimately, landings. [] Dungeness crah, Cancer magisrer, fishery, reeruitment, larvae, dispersal, pupulation dynamics, northeast Pacific. Glen S. Jamieson, Pepariment uf Fisheries and Oceans, Bialagical Sciences Branch, Pacific Biological Station, Nonainio, B.C, Canada, VOR SKO; DavidA . Armstrong, School of Fisheries, University of Washington, Seattle; WA, USA 98195; 6 July, 1990. Studies of larva) and juvenile stages of com- mate adult abundanee from measures of larval mercial Decapoda have generally been done for abundance and female fecundity. Nichols er al, purposes of describing ecology, reproductive bi- (1987) calculated female stock biomass of Ne- ology, population dynamics, habitat require- phrops lobster in the Irish sea from estimates of ments, and predator/prey relationships. While larval production and female fecundity and in- quantitative studies of catch-per-unit-effort corporated these data into a multi-species model (CPUE) and density have been used to estimate of lobster and cod interaction. Application oft his the next year9s recruitment for incorporation into same technique was.used by Nichols and Thom- fisheries. management plans, data have been pson (1985) to estimate stock size of the edible rarely amenable for accurate predictions. in crab, Cancer pagurus, Similarly, Incze et al. excess of one year (Cobb and Caddy, 1989). (1987) accounted for significant differences in Although predictive capability is a desirable interannual densities of larval tanner crab aspect for management, in the case of Decapoda, (Chionoecefes optlio) based on quantitative this is often frustrated because of high seasonal changes in number of adult females. However, and spatial variability, as well as tremendous their suggested relationship did not hold in all interannual variability, caused by a suite of biotic years of their study, nor did it apply well to a and abiotic factors (Jamieson, 19863, b; Botsford congener, C. bairdi. et al., 1989). This is particularly the case for Use of indices of juvenile abundance in Deca- quantitative data on Jarvae, and such data are poda to predict adult spawning stocks or relative typically of limited use in definition ofs tack-re- strength of fisheries have only been developed cruitment relationships or prediction of ultimate successfully in a few instances, Relative year- year-class strength in the fishery- class strength of juvenile rock lobster (Panulirus Only a few attempts have becn made to esti- cygnus), aS measured by an index of puerulus MEMOIRS OF THE QUEENSLAND MUSEUM settlement, Was correlated to the strength of cam- megalopal stage before settlement to the ben- mercial fisheries four years later (Caputi and thos. Total larval period is about 110 days at Brown, 1986; Phillips, 1986, Phillips and ambient temperatures (Poole, 19466; Lough, Brown, 1989), A similar approach was used in 1976; Reilly, 1983), with about 28 days spent as slock-recruitment analyses of blue crab (Cal- megalopac (Hatfield, 1983), [t is somewhat linectes sapidus) populations in Chesapeake Bay unique among nearshore benthic species in that (Tang, 1985). A more thorough study was re- a portion of its larvae is commonly found con- cently presented by Lipcius and Van Engel siderable distances offshore, with later stage (1990) based ona thirty year time series of data, zova and early stage megalopae tending lo be from which they concluded thal a significant found furthest offshore (Reilly, 1983; Jamieson correlation exists between juvenile abundiince and Phillips, 1988; Jamieson ef al., 1989). Late and spawning stock size. intermoult stage megalopae are found in abun- In this paper, we review recruilment of Dunge- dance progressively closer inshore (Hatfield, ness crab, Cancer magister Dana, considering 1983; Jamicson and Phillips, 1988), but mecha- all three major life stages: embryo, larvae and nisms which would bring megalopac located juveniles. We define recruitment as abundance more than about 30-40 km offshore to appro- change between consecutive life history stages, priate nearshore locations (<64 m depth) for culminating ultimately in an annual increase in enhanced survival as juveniles (Carrasco ev al., abundance of the fished population since larger 1985) have not yet been satisfactorily deter- crabs only moult once annually. We emphasise mined (Jamieson ef al., 1989), Offshore move- recruitment to both larvae and juvenile life his- ment presumably facilitates larval dispersal, but lory stages, since these are periods of great mor- i may be that most nearshore settlement results lality, and generally the times when relative from those larvae which remained shoreward of yeur-class Strengths at recruitment to the fishery region-specific, mostly as yet undetermined, are typically established for this species. Events oceanographic boundaries. progressively occurring throughout the life his- Most larval settlement is typically in May and tory cycle are discussed as these will bear on the June along the outer coast, with selllement in abundance of crabs at each successive life his- both estuarine and nearshore locations. Much tory stage leading to recruitment to the fishery. recent study in Washington (Gunderson ef ai., We also consider recruitment in four major geo- 1990) has been focused on the relative impor- raphic regions: 1) the outer coast from San tance of some of the region's major estuaries tancisco, California, to Cape Flattery, Wash- (Willipa Bay and Grays Harbor, Washington), ington; 2) off the west coast of Vancouver Island: compared with the arca shoreward of the 50 m 3) north of Vancouver Island to Kodiak Island, isobath along the outer coast, in terms of their Alaska: and 4) the Georgia Strait-Puget Sound habitat contribution to overall regional recruit- (GS-PS) complex. These specific locations were ment, Juvenile dynamics of Dungeness crab in selected on the basis of broad-scale oceano- both these locations have been relatively well graphic boundaries, unique oceanographic sin- desertbed, and it has become evident that consid- pularities, and available data on local Dungeness erable annual variation can occur, Female crab crab population dynamics. extrude their first egg mass as 2 y-olds at about 115 mm, notch-to-notch carapace width (CW), DUNGENESS CRAB LIFE HISTORY while males are larger than females at puberty and begin mating successfully at about 140 mm CW (3-yolds}. Males mostly recruitt ot he fished Dungeness crab ranges from the Pribilofs Is- population al 34 y of age. lands to Magdalena Bay, Mexica, in the north- eastt¢rn Pacific (Hart, 1982; Jensen and THE PHYSICAL ENVIRONMENT Ammstrong, eet} and is commercially exploited from northern California to Kodiak, Alaska, This spatial distribution overlaps generally recog- REGIONAL CIRCULATION nised oceanographic domains (Dodimeud et al, The general surface current pattern over the 1963; Thomson, 1981; Ware and McFarlane, continental shelf has been described by a number 1989) for coastal areas of the norjheastern of recent reports, including Hickey (1979, 1989), Pacific Ocean, and this probably influences ob- Freeland er «tf, (1984), and Thomson ev al, served recrifilment patterns and makes causative (1989), The west coast of Vancouver Island generalisations inappropriate for the coast as a borders the bifurcation zone of the Subaretic whole. Current. an extensive, albeit poor defined, Dungeness crab have a relatively long pelagic zonal flowing, cross-Pacific surface current (Fig, larval period, with fiye zocal stages and one 1), Seaward of the continental shelf, this current RECRUITMENT PATTERNS OF BUNGENESS CRABS splits into the pole-ward flowing Alaska Curreni the two Domains discussed. The outflow from and the equator-ward flowing California Cur- Juande Fuca Strait, beingo fl ower density runoff rent. Direct observation of a persistent northward from rivers entering the GS4PS complex, forms near-surface flow off Vancouver Island indicates the source of the Vancouyer Island Coastal Cur- that the offshore circulation there is dominated rent{VICC), This.is a persistent pole-ward flaw- by the Alaska Current (Thomson e7 al,, 1989), ing coastal curren(, confined mostly landward of Dungeness crab occur in two of the three princi- the 100 m jsobath on the cantinental shelf, that pal oceanographic domains recognised by Ware extends Io about the northern tip of Vancouver and McFarlane (1989), namely the Coastal Up- Island (Thomson et al,, 1989), After the spring welling and the Coastal Downwelling Domains transition, this current flows counter to the pre- (Fig. 2). The former extends from Baja Cal- vailing northwesterly winds along the outer ifornia to the northern tip of Vancouver Island coast and the Shelf-break Current while after the and is defined by the normal summer paticrn of fal] transition, it flows with and merges with the Wind stress curl and Ekman divergence, 1.¢. up- Davidson Current, Welling (Parrish ef al, 1981), Generally north- west winds from May to September resull in TorOGRAPHY southward-flowing Shelf-Break Current (SBC) Washington, Oregon and California typically centred on the outer margin of the continental have broad, extensive, sandy beaches extending shelf, This causes upwelling of intermediate for scores of kilometres, with few bays or head- depth, cold water onto the continental margin lands. The coast of British Columbia and south- and offshore transport. Southwest winds domi- east Alaska is mostly rocky and fjordal, with nate this Domain during the winter, causing many headlands, tslands, small bays, and rela- downwelling, onshore transport and poleward lively small, crescent-shaped beaches. Dunge- transport of surface waters in.a seasonal current ness crab most commonly occur in a habitat of called the Davidson Current. The annual transi- extensive sand, and so available optimal habitat tion between the predominantly upwelling and is generally less north of Washington State. downwelling seasons occurs inthe spring (Mar.4 Georgia Strait and Puget Sound consist of a Apr.) and fall (Sept,4Oct,) with The seasonal variely of habitats, with Tjordal inlets predami- reversal in psevalling alongshore winds and cur- nant On Ure eastern side tn Canada and gently rents (Thomson ef al., 1989), sloping, gravel-sand bottams predominant in The Coastal Downwelling Domain extends Pugel Sound and the southeastern side of Van- from the northern tip of Vancouver Island north- couver Island. The Gulf and San Juan Islands are ward along the coast of southeast Alaska and in the middle of this region and have mostly then westward along the Aleutian Islands. The sleep, rocky shores. which are of marginal suita- Alaska Current flows adjacent to the coast of bility the Dungeness crab. North America seaward of the coastal margin. being driven by a wind stress.curl and augmented LIFE HISTORY STAGES AND THEIR by freshwater addition and an along-shore, RECRUITMENT wind-induced sea level gradient. Freshwater runoff causes a poleward Mowing coastal cur- Parent PoruLaTion AND FECUNDITY rent, extending to about 40 km offshore. The The extensive pelagic larval duration of transilion in prevailing coastal winds in the Dungeness crab in areas of strong along-shore spring is weak, bul due to the general behaviour ocean currents apparently prevents development of the cyclonic meteorological systems in the of discrete, genetically distinctive populations Tegion, wind stress tends to augment the baro- on the ouler coasi (Soulé.and Tasto, 1983). There clinic component of the coastal circulation by is no evidence 16 supgest thal crab larvae have confining it close ta shore. As a consequence, the navigational abihty to 8home9 and return to from central British Columbia north ta about the specific: location where they were hatched, In Kodiak Island, there is a generally persistent the Coastal Upwelling Domai, larvae hatch in downwelling except for a few months in the January-February during a period of strong, summer (Ware and McFarlane, 1989), The northward-flowing nearshore currents, which southern boundary of this Domain is not sharply aboul two months later typically reverse to flaw defined from a strictly oceanographic viewpaint- equally strongly in asoutterly direction. With an The continental shelf off Vancouver Island, approximalely four month larval period, this although not a recognised Domain itself, being may resull in extensive larval dispersal. Larvae part of the Coastal Upwelling Domain, has. a hatching off northem Calilormia could theoreti- Unique oceanographic feature which for Dunge- cally move as far north as British Columbia ness crab makes this area intermediate between before currents reverse, Similarly, larvae hatch- 368 MEMOIRS OF THE QUEENSLAND MUSEUM 132° 126° RECRUITMENT PATTERNS OF DUNGENESS CRABS 369 son and Armstrong, unpubl.). Jamieson (1986a) 7 discussed in detail the available data supporting assumptions relating to crab fecundity, and con- cluded that if maximising reproductive potential is an identified goal, then insufficient data are presently available to establish what relative level of progeny production is being achieved. Nevertheless, although commercial landings have varied substantially over time (Fig. 3), de- tailed surveys of larval occurrence (Jamieson and Phillips, 1988; Jamieson et al., 1989; Jami- WEST WIND eson, unpubl.) have consistently found wide- DRIFT spread high levels of larval abundance over the continental shelf off Vancouver Island and CENTRAL PACIFIC a GYRE » %, Washington, and in Georgia Strait in the five ¥¢ years studied to date. Annual settlement of larvae 4 has varied substantially during this period (Jami- 1 ro 2 3 eson et al., 1989), suggesting that factors other COASTAL TRANSITION COASTAL than overall larval production are the major de- DOWNWELLING ZONE UPWELLING terminants in establishing year-class strengths. 180° wi70° W 160" W 150° W 140° W 130° W 120° W 110° Ww 100° Ww LARVAL SETTLEMENT FIG. 2. Approximate areas of oceanic domains and Some Dungeness crab larvae occur at great prevailing current directions in the northeast Pacific distances offshore and while this no doubt facil- Ocean (modified from Ware and McFarlane, 1989). itates and/or is the result of species dispersal, there is always the risk of larval wastage in that ing near the southern boundary of the Coastal many larvae may never return successfully to Downwelling Domain could move to its north- geographic areas favourable to juvenile survival. ern boundary, while larvae hatching in the north- It has not been established how far offshore most ern part of the Coastal Upwelling Domain could larvae which do settle and survive have actually move into the Coastal Downwelling Domain, gone, but the relative lack of early-stage mega- entirely because of passive transport. The con- lopae, in abundance, in close proximity to shore sequence is that considerable mixing of progeny in outer coast areas, at least off British Columbia hatched at different sites probably occurs, and Washington (Jamieson et al., 1989; Jamie- making it impossible to clearly identify the son, unpubl.), does suggests that movement in at parent population ofj uveniles that recruit at any least the kilometre scale can be expected. As- particular location. As discussed by Jamieson sumed aspects relating to this have been dis- (1986a), the extent to which a local population cussed in detail by Jamieson ef al. (1989), and may contribute to local recruitment is unknown, they, along with Jamieson and Phillips (1988), but it appears with available data to be slight. have demonstrated that juvenile recruitment pat- It is illegal to harvest female Dungeness crab terns off the outer coast of Vancouver Island, in Alaska, Washington, Oregon and California. British Columbia, are substantially different Although this can legally be done in British from that off Washington, Oregon and Cal- Columbia if females exceed the minimum legal ifornia (Fig. 3). size (MLS) of 155 mm CW (= 165 mm, spine- Outflow from rivers emptying into the GS4PS to-spine carapace width), in practice, few are complex is predominantly through surface out- harvested because of poor market demand, since flow in Juan de Fuca Strait, and this outflow, the meat yield is less, and there is a general lack of VICC, subsequently moves northward adjacent sufficiently large female crab. The size limit for to the coast of Vancouver Island. Crab mega- males, although not known to be based on any lopae are virtually absent in this current and are documented biological data, presumably allows concentrated on its seaward boundary (Jamieson successful mating to regularly occur since in and Phillips, 1988; Jamieson et al., 1989). It thus most locations many female crab caught during apparently acts as a barrier to the movement of the winter carry extruded, fertilised eggs (Jamie- plankton to shore (Thomson et al., 1989), and FIG. 1. Prevailing surface circulation off the British Columbia4Washington coast. A, winter. B, summer. Broken arrows indicate uncertain currents. Numbers give speeds (cms) (modified from Thomson, 1981). 370 MEMOIRS OF THE QUEENSLAND MUSEUM ALASKA major recruitment on the west coast of Vancou- 5 YS ver Island is thus only possible when appropriate meteorological events occur when megalopae ate present and ready to settle, namely May and June (Jamieson and Phillips, 1988). This does BRITISH COLUMBIA not always occur, and since 1983, major crab uw settlements at Tofino, British Columbia, were only observed in 1983 and 1989 (Smith and a a as 7.073075 77 79 BF 0S BS AF BD 9D Jamieson, 1989a: Jamieson, unpubl.). Little mt settlement occurred from 1985 to 1987, while WASHINGTON there was only limited settlement in 1988. South of Cape Flattery, located at the southern, Arnenacine : seaward end of Juan de Fuca Strait, the absence PUGET SOUND of a 8barrier9 current adjacent to the coast means that settlement is not physically impeded by nearshore currents, although years of excep- (L1A07N3D)I NGS Line G, June 1989 Abundance (per 10 square meters) Depth (m) 7000 600 44 C. mogleter CALIFORNIA 8eb [= Depth N 75w0a0 10 J F 8i 4,300 +k : eee, 8 / 4200 WS4 357 SSHy ABOt eQcl Mtem wBe e Ow DPN SPyS Ty 8oth 777R 4o0t BBkD HBAe BST hooO G nOlt DA - gate fe< e = = 100 SEASON or 0.01 4L 44 4 4L 4L_ oO FIG. 3. Dungeness crab landings by political jurisdic- Q 10 20 30 40 50 60 tion off western North America. Distance offshore (km) recent study by Jamieson and Thomson (un- publ.) indicates that megalopae are being con- centrated in an area of upwelling and/or seaward surface current movement of relatively low Line E, June 1989 velocity (1-5 cm sec9') between the VICC and the SBC (Fig. 4). Shoreward movement of sur- nba putenass (per 10 square meters) Depth (m) Bo face waters is associated with southerly winds 44 C, magiater (Fig. 5) but unless sufficiently intense or pro- = paptn /\_ }s00 longed, such movement is only to the seaward ws \ boundary of the VICC. The presence of the P/ fsS 400 VICC seems to direct most up-welled water off- _4 shore, while because of the shallow depth of the \ Ma (306 nearshore portion of the continental shelf, no \ Zz i \ a 7 200 shoreward movement beneath the VICC is possible. The only apparent opportunity for sub- O1eaE e \ \ V 44100 stantial movement to the shore is when the VICC temporarily breaks down because of cessation of 0.07 [ ie 4 " < 0 the surface outflow in Juan de Fuca Strait, which 10 20 30 40 50 60 Distance otfshore (km) is typically associated with sustained, strong southerly winds. This is accompanied by a rise in mean coastal sea level, enhanced wind and FIG. 4, Abundance of Dungeness crab megalopae in convective mixing of surface waters and the across-shelf transects in relation to depth. The north- cessation of upwelling 4 events similar to those flowing Vancouver Island Coastal Current is mostly which occur during and after the Fall Transition shoreward of 100 m depth, with the south-flowing (Thomson ef al., 1989). From a crab recruitment, Shelf-Break Current on the seaward side- RECRUITMENT PATTERNS OF DUNGENESS CRABS 371 a LV ij = SRITISH os055s//116 6 : a 5f | Neal COLUMBIA eee ® % E. * ak Vancouver 8 a 4 | {sland <ay & A a ene, hyyd s Hé. e¢ Tfy h3 F E MAP re TOFINOY Yo, INSERT Lace . Hm Js As anbebre SS a hs &, de o309/13 rv BARKLEY 0512/13 SOUND 0400/18 roy \6 rh ~~ -44 ~ 85 y =. 7, 2202/13 es \ \ yi} ~ 14 14° 100m St ey 8 y= 200m a 5/13 it ° oy rae ) S458 j % : @®> ¬ 8- j <| \ A \ \_- ae } ts ir ~L ¢ af= ' eZ Pa i 1 a lt } i j a \ t / \. ) 8 ny yi Ww : - 8 i) \o 20 30 Km ye i 200m a ¢ t ae" \, . 48° JUNE 1986 f ; a aal ? | \26" r26° 4 FIG. 5. Tracks of 9 drifters deployed along a transect off Tofino, British Columbia, on June, 1986. Marks along each track indicate the position of each drifter at noon on the day (the adjacent 2-digit number) indicated. Four-digit numbers refer to the hour the drifter was deployed and/or retrieved (from Jamieson et al., 1989). tional settlement are also typically sporadic. The dance over most of the continental shelf, some temporal pattern of settlement along the Ameri- data suggest (R.A. McConnaughey, Univ. can coast has shown an annual variability in Washington, Seattle, WA, pers. comm.) that in magnitude of about ten-fold, with no obvious years of major settlement off Washington, most similarity to that off Vancouver Island. Major larvae may never move far from shore. This, landings have occurred about every 9-10 y (Fig. coupled with that area9s extensive favourable 3) and causative mechanisms have been hy- habitat, increases larval survival at settlement. pothesised and discussed extensively (Jamieson, There have been no studies of relative larval 1986a; Methot, 1989; Botsford e7 al., 1989). A distribution and abundance off Alaska. Annual satisfactory explanation has yet to be fully estab- commercial landings (Fig. 3) do not show the lished. While larvae have been observed in abun- same pattern evident off British Columbia or 372 MEMOIRS OF THE QUEENSLAND MUSEUM Megalopal Day Depth Distribution further south, but this may be because, like the crab fishery in British Columbia (Jamieson, Georgia Strait, July 19, 1989 1985), regional landings are the composite of a Percent number of distinct regional fisheries (Koeneman, 1985; Kimker, 1985; Merritt, 1985; 70 N = 143 Eaton, 1985). Recruitment patterns in each re- 60 | gion may be largely disconnected, giving no l | clear pattern for the region as a whole. Larval settlement patterns in the GS4PS com- plex are different from those along the outer 207 coast. Megalopae settle in abundance mostly 10 f= | later in the year (August-September), are 4<44 " T a : 44 smaller in size, and show some minor, although 80-100 100-120 120-140 140-160 160-180 180-200 perhaps significant, morphological differences Depth (m) (DeBrosse et al., 1990). Recent studies of meg- Maximum depth = about 400 m alopal spatial and vertical distribution (Jamieson and Phillips, unpubl.) in Georgia Strait indicate Megalopal Day Depth Distribution that in July, megalopae are found in high abun- dance throughout the Strait but that their vertical Outer coast, May and June, 1987 distribution during daylight seems to differ from Percent that found off the outer coast (Fig. 6). In the 44_ N= 175 | Strait, megalopae appear to descend to depths of about 150 m while offshore, they are seldom found in quantity below about 40 m (Jamieson etal., 1989). While megalopae have been caught below 100 m on the outer coast (Jamieson, unpub.data), they have been late intermoult stage, suggesting they may have been in the process of settling. Dungeness crab megalopae have been caught as deep as 273 m in coastal 0-12 12-25 0-25 inlets in epibenthic sled tows (Jamieson and Depth (m) Sloan, 1985), a depth where they are unlikely to Maximum depth + >400 m survive as juveniles. This difference in megalopal diel migration behaviour has interesting consequences, partic- FIG. 6. Daytime depth distributions of megalopae off the outer coast (modified from Jamieson et al,, 1989) ularly if zoea from the two regions show similar and Georgia Strait. Data collected with Tucker trawls differences as well. Since summer water in 1987 and 1989, respectively. temperatures below about 50 m in most of the Strait are 7-8°C (Thomson, 1981) [temperatures Juan de Fuca Strait and the penetration of outer on the outer coast at 25-50 m depth are about coast water (Fig. 8), sometimes containing meg- 12-14°C (Thomson e¢ al., 1989)] and daylight in alopae (P. Dinnel and D. Armstrong, unpubl.), July is about twice as long as darkness, growth has been documented (Thomson, 1981; Thom- rate would be reduced, thereby probably extend- son etal., 1989). However, for reasons described ing the larval period and possibly resulting in later, this is mostly on the American side of the their smaller size at settlement. It also could Juan de Fuca Strait and relatively few outer coast result in stock isolation, since the surface water megalopae would seem likely to penetrate into of Juan de Fuca Strait, the main connection be- the Canadian waters of Georgia Strait. tween the GS4PS complex and the Pacific Ocean, flows predominantly seaward (Fig. 7) while water below about 80-100 m flows pre- JUVENILE SURVIVAL dominantly shoreward (Thomson, 1981). Geor- UPWELLING DOMAIN: Megalopae settle to the gia Strait megalopae, which may spend most of benthos and metamorphose to first instar juve- their time at depth, would thus tend to be retained nile crabs primarily between May and June along within the GS4PS complex while outer coast the coast from Northern California through megalopae, which are mostly near the surface, Washington State (Botsford er al., 1989). Most would generally be prevented from entering the studies of juvenile crab populations in this area Strait. Sustained southerly winds can, though, have been descriptive portrayals of distribution temporarily stop the outflow of surface water in and abundance, growth and size-at-instar RECRUITMENT PATTERNS OF DUNGENESS CRABS 373 shallower than 40 m (Gotshall, 1978; Carrasco et al., 1985). Estimated density increases ap- proximately 20- to 50-fold as depth decreases from 41-70 m ta 16-40 m (Carrasco ef al, 1985: MeConnaughey and Armstrong, unpubl), Even inside of the 40 m isobath, density of 0+40+ DiEmaP TH (hHaere dg ees lronger crab fluctuates appreciably because of substrate lype. IL is highest on well sorted sand and lowest on gravel-cobble (Fig. 9), Trawl samples off Gialanes gerne suet iim) Grays Harbor, Washington, in [985 showed 0+ crab densities as high as 30,000 crab/ha on sand, compared to only about 200/ha on gravel. Post- FIG, 7. Average direction of water flow in cross-sec- settlement crab growth on the outer coast fram tion of Juan de Fuca Strait from Pillar Point (USA) June through September is typically relatively to Port Rentrew (Canada), Shaded is a seaward flow, slow, und young-of-the-year are only usually no shading is a landward flow (from Thomson, 1981). about third instar (13 mm CW) by September (Fig, 10; Gunderson et al., 1990). This Is due in (Cleaver, 1949; Poole, 1967). There have been large part to bottom water temperatures of less few quantitative surveys that provided estimates than 10°C in this upwelling system, As a con- of mortality or indices of year-class strength. but sequence of small size, crab mortality is high, those done indicate that considerable differences since they remain susceptible to many predators in survival occur within a region, depending on (Reilly, 1983) through the winter, Even numeri- both abiotic and biotic factors. It is important to cally Strong year-classes at settlement, such as briefly discuss these factors. in considering how that of 1985 (Gunderson e¢ al,, 1990), can have recruitment is mediated. their abundance depleted sufficiently through Although larvae may occur in abundance off- their first winter to become unexceptional il shore, juveniles for the most part seem to survive recruitment to the fishery, only in nearshore locations, in water generally Quantitative data of sufficient 0* 0+ y-class FIG. 8. Drawings based on infrared satellite images of sea-surface temperatures, showing a sequence of warm-water intrusion (heavy stippling) into Juan de Fuca Strait trom the Pacific Ocean in September, 1979. Intrusion was confined {o the southern half of the Strait and reached a maximum of 135 km from the entrance. Four days after the cessation of the causative southwest Winds, the seaward estuarine ¢irculalian was reestablished and the intrusion began ta he udvected our of the Strait (from Thomson, 1981). MEMOIRS OF THE QUEENSLAND MUSEUM 124 10wW Jettles 44 46 50N 124 20 12410Ww FIG. 9, Dungeness crab density (number ha9') offshore of Grays Harbor, WA, June 1985. survival over sequential years to construct in- (1984 year-class; Fig. 11) led to a record high dices of year-class strength and determine fishery in excess of 10,000 t in 1987/88 and spawner-recruit relationships are non-existent. 1988/89 (Fig, 3). Unique to the 1984 year-class Gotshall (1978) found no correlation between an was unusually rapid growth of the coastal cohort index of 0+ crab abundance and commercial (Armstrong and Gunderson, 1985), which re- landings 3.5 y later, based on cursory coastal and 50 estuarine surveys in northern California. How- ever, Warner (1987) measured 0+ crab density 45 in fall trawl surveys off northern California from 1972 through 1985 and noted that a 20-fold greater density for the 1972 year-class sub- sequently resulted in a near-record commercial fishery in 1976 (11,300 t). Tasto (1983) and Reilly (1983), who estimated the abundance of benthic juveniles and larvae in central Cal- ifornia, respectively, both observed relatively strong year-classes in 1975 and 1977, which were consistent with an increase in commercial CWMIAEDRATANH P ACE landings for California in 1978-80 (Methot, 1989; PMFC, 1989). The most comprehensive study of 0+ Dunge- 8 40 ness crab recruitment and survival has been done MONTH along the southern Washington coast between 1983 to 1989 (Armstrong and Gunderson, 1985; FIG. 10. Comparison of 0°+0+ Dungeness crab Gunderson et al., 1990), While they found that growth during their first summer inside an estuary (Grays Harbor, GH) and on the outer coast in near- high interannual variability in estimated abun- shore waters near the estuary (NS). Numbers indicate dance of 0+ crab did not equate well to the mean bottom water temperature (°C.); bar = 1 SE) strength of future fisheries, measures of 1* y-old (from Armstrong and Gunderson ,1985). crabs did. The high abundance of 1+ crab in 1985