3.1.7 Alternative7: Inshore Juvenile Cod Overview Based on the information presented on juvenile Atlantic cod and the inshore areas of the Gulf of Maine, coastal juvenile cod HAPC designation is recommended on the criteria of ecological function and sensitivity to induced environmental degradation. An HAPC designation for the nearshore Gulf of Maine could assist in the enhancement of Atlantic cod and other groundfish species. In 1999, the Council voted to approve this alternative and include it in the next appropriate fishery management plan vehicle. Since that time, the Habitat Plan Development Team has advised the Habitat Committee, based on the supporting information, that the Alternative should be expanded to include two options for public comment: Option A: 0-10 meters (MLLW) Option B: 0-20 meters (MLLW) The most practical approach for delineating an HAPC for settled age-0 cod is to circumscribe the reported center of distribution for this life stage throughout the range of the stock. The information available suggests that the HAPC should be from the low tide line to a depth of 10 m (33') MLLW (Option A) or alternatively from the low tide line to a depth of 20 m (33') MLLW (Option B) from eastern Maine to the Rhode Island/Connecticut border. This narrow depth range describes critical habitat from settlement through the first autumn of life and overlaps seasonal habitat of age-1 juvenile cod. It also bounds the critical nursery zone for early benthic stages of important juvenile habitat for some other groundfish. 3.1.7.1 Alternative 7A: Inshore Juvenile Cod (0-10m depth contour) Recent Research 58 During the late 1980's, Atlantic cod inhabiting the waters off southern Labrador and eastern Newfoundland (viz. northern cod stock) underwent a dramatic decline in biomass, with the result that the famous Grand Bank fishery collapsed in 1992. This led to a fishing moratorium as well as an economic and ecological disaster (Hutchings 1996; Myers, et al. 1996). It also triggered unprecedented research at federal fisheries laboratories and universities in eastern Canada under the auspices of the Government of Canada's Northern Cod Science Program and the government / industry-funded Ocean Production Enhancement Network program. In total, there were 58 study initiatives and projects covering a broad suite of research costing about $48 million from 1990-95 (Campbell 1997). The following description of research results from Canada and other countries deals with life history and behavioral ecology of newly settled juvenile cod, particularly postsettlement events relating to habitat that may ultimately affect recruitment strength. Studies have focused on laboratory observation experiments as well as field capture efforts utilizing active sampling equipment, SCUBA and submersible vehicles for in situ observations, and seabed classification techniques for acoustically classifying juvenile habitat. Nearly three dozen scientific papers relating to this subject have been published in recent years. The information is directly applicable to coastal nursery areas in the Gulf of Maine, the inference being that knowledge gained from such studies should be used for more risk-adverse habitat management. Juvenile Cod Community & Interactions - Research Results Pelagic Juvenile Settlement Post-larval pelagic juveniles are transported by prevailing currents to shallow waters off eastern Newfoundland beginning in May and may continue arriving in periodic pulses as late as December (Methven and Bajdik 1994; Grant and Brown 1998). Their length upon settlement is 25-45 mm (Pinsent and Methven 1997). In southwest Nova Scotia, pelagic juveniles arrive inshore slightly larger (˜40-50 mm) in May (Tupper and Boutilier 1995a and 1995b), whereas influxes of larvae begin earlier in Massachusetts waters two to three months after hatching (Bigelow and Schroeder 1953). On Georges Bank, cod settle out in July at 40-60 mm, and those reaching rough and cobble bottom may experience reduced predation risk. This particular habitat may be an important demographic bottleneck to benthic recruitment on Georges Bank (Lough et al. 1989). Pelagic juveniles exhibit no preference for habitat types at settlement, and they occupy rock reef, cobble, eelgrass (Zostera marina) beds, and sand bottom (Tupper and Boutilier 1995b). Tupper and Boutilier (1995b) assumed that settling also occurred on macroalgae habitat, as noted by Keats et al. (1987) off eastern Newfoundland, however, algal stands were scarce in the St. Margaret's Bay, Nova Scotia, study area due to subtidal grazing by sea urchins (Strongylocentrotus drobachiensis). Urchins commonly leave a partially 59 denuded or "barren" zone along nearshore (˜2-12 m below MLW) sections of the maritime provinces and Gulf of Maine. Age-0 Movements and Diel Feeding Shallow water depths (<5 m) and a strong attraction to features on most substratums, except sand, afford settled juveniles an environment conducive to growth and survival (Tupper and Boutilier 1995a; Grant and Brown 1998b). A similar attraction to structure, particularly eelgrass and kelp, by age-0 cod was found in shallow waters (< 10 m) along the Maine mid coast in 2000 (Lazzari et al. 2003). The shallowness appears to ecologically segregate the 0-group cod from older age-groups at least during daylight (Tupper and Boutilier 1995a; Fraser et al. 1996; Gotceitas et al. 1997; Grant and Brown 1998a and 1998b). Age-0 cod maintain a strict diurnal foraging cycle, school (or shoal) feeding on zooplankton in a tide-related pattern during the day, and remain near protective bottom habitat which they readily seek when threatened (Gotceitas and Brown 1993; Gotceitas et al. 1995; Grant and Brown 1998a). The mottled coloring of young juveniles effectively conceals them in a pebble-gravel environment (Lough et al. 1989; Gregory and Anderson 1997). In contrast, pelagic juveniles on Georges Bank maintain a nocturnal feeding pattern (Perry and Neilson 1988). Age-0 cod cease feeding in surface waters and disperse to the substratum at night (Grant and Brown 1998b) where they are less active to reduce interactions with potential predators (Grant and Brown 1998a). The diel change in vertical distribution and activity of 0-group cod coincides with a nocturnal shoreward movement and foraging by older (age-1-3) conspecifics (Bosgstad et al. 1994; Gotceitas et al. 1997; Grant and Brown 1998a). Intercohort cannibalism is common. The occurrence of age-0 cod in very shallow water (<1.2 m) at night (Methven and Bajkik 1994) has also been interpreted as possibly an evasive response to predation risk (Grant and Brown 1998a). Influence of Habitat Structure and Predation on Age-0 Demography Tupper and Boutilier (1995b) found that the spatial pattern of settlement was altered by post-settlement mortality in St. Margaret's Bay, Nova Scotia. Age-0 survival was positively correlated with an index of rugosity, a measure of actual bottom surface or complexity. Capture success by fish predators (in this case, three species of Cottidae during a diurnal field study) was inversely related to the index of rugosity. As a result, higher densities of age-0 cod were found in cobble and rock reef habitats than in eelgrass. However, the rugosity index could not account for the complexity of surface area that eelgrass offered. Higher survival in sites of cobble and rock reef was attributed to increased shelter that the more structurally complex habitats afforded coupled with decreased predator efficiency (Keats et al. 1987; Lough et al. 1989; Tupper and Boutilier 1995b; Gotceitas et al. 1997; Grant and Brown 1998a and 1998b). Tupper and Boutilier's (1995b) and Grant and Brown's (1998a) in situ studies confirmed earlier and subsequent laboratory experiments on substrate preference and predator 60 efficiency. Clearly, the presence of conspecifics may influence the distribution and food intake of age-0 cod in the wild. Both age-0 and 1 cod preferred finer grained substrate in absence of a predator, but when in the presence of an age 3 conspecific, young-of-theyear and age-1 either avoided the predator or selected the coarser substrate (cobble vs. gravel) where they hid in interstitial spaces (Gotceitas and Brown 1993; Gotceitas et al. 1995; Fraser et al. 1996). Age-0 avoided the yearling conspecific resulting in a significant increase in use of gravel and cobble confirming the level of habitat segregation noted in the wild (Fraser et al. 1996). Also, age-0 cod avoided kelp (Laminaria) except when exposed to an actively foraging predator and cobble was unavailable. In this situation, kelp significantly reduced predation risk (Gotceitas et al. 1995). Both field and laboratory studies indicate that the association with coarse substrates, when coupled with behavior patterns that reduce predation risk, give young cod competitive advantage in avoiding detection or capture. Eelgrass Habitat and Abiotic Factors The presence of eelgrass beds or meadows appears to be a very important factor influencing the distribution of age-0 cod throughout the Canadian Maritime Provinces (Tupper and Boutilier 1995a; Gotceitas et al. 1997; Grant and Brown 1998a, 1998b) and along the Maine coast (Lazzari et al. 2003). Grant and Brown (1998b) noted that cod were more highly concentrated in eelgrass beds with >65% submersed canopy coverage. Gotceitas et al. (1997) captured age-0 cod almost exclusively in eelgrass beds of Trinity Bay, Newfoundland, where their usage by 0-group cod was consistent spatially and temporally. The eelgrass sites most sheltered to natural physical disturbance produced the highest catches; lower catches occurred at the shallowest and least saline sites (10.4- 19.5 ppt). Salinities were usually high (>25 ppt) at most Newfoundland study sites (Methven and Bajkik 1994). Age-0 tolerate much lower salinities as was observed in coastal waters of Wales and England where catches occurred from 20-31 ppt (Riley and Parnell 1984). Post-settlement cod may respond to environmental gradients in addition to substrate structure and salinity. For example, high water clarity may be important for feeding (Horne and Campana 1989). Strong tidal currents may be beneficial for concentrating food in seagrass beds (Tupper and Boutilier 1995b; Grant and Brown 1998a). Water temperatures coinciding with age-0 collections in St. Margaret's Bay, Nova Scotia, ranged between 4-9°C from May to July (Tupper and Boutilier 1995b) while July to September temperatures in age-0 habitat of Trinity Bay, Newfoundland, were 12-16°C with a year-round range of 1.7-17.0°C (Methven and Bajdik 1994). Water temperature might displace 0-group and yearlings to slightly deeper waters south of Newfoundland, however (Methven and Schneider 1998). Among-Habitat Variation in Age-0 Growth Growth of settled age-0 cod appears to be temperature dependent (Tupper and Boutilier 1995a). Growth was most rapid in eelgrass beds, which may positively effect overwinter 61 survival of demersal 0-group juveniles. Growth was slowest on sand bottoms; differences in growth between young inhabiting reef and cobble bottoms were not significant (Tupper and Boutilier 1995b). The growth advantage conferred by seagrass is related to the variety of microhabitats therein that support a diverse community of invertebrates (Orth et al 1984; Heck and Crowder 1991; Heck et al. 1995; Grant and Brown 1998a). Planktonic organisms may be passively concentrated by water currents and effectively retained within the eelgrass canopy. Also, invertebrates and fish may actively seek its confines even crossing predation-risky sand to reach isolated patches (Sogard 1992). Small planktonic crustaceans, but mostly copepods, are preyed upon by young cod (Keats and Steele 1992; Grant and Brown 1998a). When mouth gape size is large enough, at a length of 6 to 10 cm, cod transition to predominately benthic prey (Keats et al. 1987; Lomond et al. 1998) which they then consume at dusk and dawn (Grant and Brown 1998a). Among-Habitat Variation in Age-0 Survival Eelgrass provides age-0 cod protection from predators (Tupper and Boutilier 1995; Gotceitas et al. 1997; Grant and Brown 1998b). In a laboratory experiment, eelgrass significantly increased the time required for an age 3 cod to capture 0-group cod and decreased the number captured. With a predator present, young cod either hid in cobble or in eelgrass when stem density was >720 stems/m2. Time to capture was highest and total prey taken was lowest in combinations with cobble or vegetation of 1,000 stem/m2 (Gotceitas et al. 1997). Results demonstrated that high plant density and/or biomass, whether eelgrass or macrophytic algae (Isaksson et al. 1994), means reduced predation risk just as does use of certain substrates. Moreover, there may be a trade-off between nutritional gain and enhanced predation risk for age-0 cod utilizing eelgrass habitat (Tupper and Boutilier 1995). Mark-recapture experiments indicate age-0 cod remain very localized, not moving more than several hundred meters in both eelgrass and no-eelgrass habitats (Grant and Brown 1998b). Those that settled earliest and were largest at settlement grew faster and defended a larger territory than later/ smaller settlers (Tupper and Boutilier 1995a), thus a competitive advantage in growth and survival may exist for the earliest pulse of postlarval juveniles over those settling later when temperatures and day length are reduced (Tupper and Boutilier 1995b). Abundance in the seagrass sites of St. Margaret's Bay, Nova Scotia, was noted to decline after early June. This was attributed to predation rather than emigration because young were strongly site-attached and defended territory as they grew. Marked individuals were not found in areas surrounding the study site (Tupper and Boutilier 1995b). As the summer season advanced, a greater decline in abundance occurred in eelgrass beds and on sand than in structurally more complex reef and cobble habitat. Observing in situ young-of-the-year seeking shelter in rock crevices, empty scallop shells, and other debris within dense grass beds, Tupper and Boutilier (1995b) believed that cod out-grew eelgrass blades as suitable refuge. Unable to compete for nonexistent shelter on sand habitat, age-0 cod school for protection (Tupper and Boutilier 1995a); however, the population density in these areas reached zero by late June (Tupper and Boutilier 1995b). Predation by three Cottids - sea raven (Hemitripterus americanus), longhorn sculpin (Myoxocephalus octodecemspinosus), and grubby (Myoxocephalus aeneus) - was most successful on sand and least successful on cobble and rock reef. Age-0 cod schooling over sand bottoms have low site fidelity which is disadvantageous to survival (Grant and Brown 1998a). Young-of-the-year appear to lose site fidelity and disperse into deeper water during the December-January period (Tupper and Boutilier 1995a; Gregory and Anderson 1997) adopting winter behavior of reduced activity and food consumption (Brown et al. 1989). Still, some marked demersal juveniles remained localized in the shallowest (<1.2 m) sampling site in Trinity Bay, Newfoundland, throughout the winter even when ice was present (Grant and Brown 1998b). Age-0 and older juveniles are more adapted than adult cod to survive icy subzero water due to elevated plasma antifreeze levels in their blood (Goddard et al. 1992). Age-1 and Older Juvenile Habitat and Movements Age-1 juveniles are found during day and night in shallow inshore waters, including locations with moderate to high wave exposure (Keats 1990). Older juveniles are generally distributed farther away from shore than 0-group and 1-group cod and at depths >25 m. Age-1 associate to a greater degree with rocky substrate and fleshy macroalgae or bottom dominated by sea urchins and coralline algae (Keats et al. 1987; Keats 1990; Gotceitas et al. 1997). The association with a macroalgal canopy seems to be more one of refuge from predators than feeding purposes (Keats et al. 1987; Gotceitas et al. 1995; Gotceitas et al. 1997). They congregate in small groups near boulders and in large crevices. In Newfoundland bays, age-1 cod have been collected within a slightly narrower temperature range, 1-16°C, than demersal 0-group fish (-1.7-17°C) (Methven and Bajdik 1994). At dusk during summer and autumn seasons, age-1 and older juveniles move shoreward into warmer water feeding areas where the young-of-the-year cod are concentrated. The attracting stimulus appears to be the periodic influxes of early settled cod (Keats 1990; Clark and Green 1990; Methven and Bajkik 1994). Age-1 cod have usually been found feeding until dawn primarily on mysids and gammarid amphipods; however, when they become about three times larger than settled age-0 juveniles, they begin cannibalizing the demersal 0-group cod (Grant and Brown 1998a). By late fall, the earliest age-0 settlers may be large enough to begin intracohort cannibalism on the late settlers, as has been noted in waters of Iceland (Bogstad et al. 1994). When abundance of older juveniles is high, mortality may increase on young-of-the-year because of competition and predation from conspecifics (Grant and Brown 1998a). Age-1 cod have also been observed feeding on plankton after moving inshore in spring (Keats et al. 1987) as well as resting near bottom in shallow water at night (Keats and Steele 1992). In the latter situation, age-1 were not feeding and analysis of stomach contents indicated daytime foraging on planktonic crustaceans leading the authors to speculate that post-transitional feeding on benthic invertebrates might be patchy in space and time. Where, when, and to some extent what yearlings eat is likely related to trade-offs between predation risk and food availability. Juvenile cod may utilize the intertidal zone for feeding purposes although there is no mention of this in recent studies. Earlier, an underwater television camera mounted on a herding fence recorded 423 "young" Atlantic cod (no size given), and Atlantic tomcod (Microgadus tomcod), which were sometimes indistinguishable from cod, as well as six, 30-40 cm (age 3 to 4) cod moving up and down a beach, either with or against tidal current, during daytime between June and October in Passamaquoddy Bay, New Brunswick (Tyler 1971). Of eight fish species observed undertaking these movements, the cod/ tomcod combination ranked third, behind only winter flounder (Pseudopleuronectes amercianus) and Atlantic herring (Clupea harengus) in their use of the intertidal zone. Diel Differences in Abundance Keats (1990) found one- and two-year-olds 16 times more abundant at night than during the day while making SCUBA transects at a depth of 5-10 m MLW. Methven and Bajdik (1994) were able to seine age-1 cod throughout the year but only at night in a cove of Trinity Bay, Newfoundland, whereas age-1 were caught both day and night by Grant and Brown (1998a) in a different cove of the same embayment. An explanation for the difference in catch of yearling cod between the two studies may be related to sampling techniques. The first study employed a 9 m seine pulled from a maximum depth of 1.2 m (no bridle). The second utilized a 30 m seine deployed by small boat 50 m from shore and pulled by towropes thereby encircling age-1 cod inhabiting a slightly greater depth range. Researchers studying young cod recognized that gear avoidance occurred during daylight, but avoidance was secondary to diel activity in explaining abundance differences between day and night catches for both age-0 and the older juveniles (Methven and Bajdik 1994; Gibson et al. 1996; Methven and Schneider 1998; Grant and Brown 1998b). Abundance of age-1 peaked in the shore zone from August-November and again in April-June period, but was much reduced in winter (temperature <0°C) indicating withdrawal to deeper habitat. The offshore movement by young cod was also reported in Passamaquoddy Bay, Bay of Fundy (MacDonald et al. 1984). 64 Juvenile Winter Habitat and Activity Juveniles inhabit progressively deeper water and associate with coarser substratum as they grow and mature, especially in winter (Keats et al. 1987). Age-1-4 cod were observed at 18 to 150 m from submersible vehicles during April (-1°C at 25-75m), Placenta Bay, Newfoundland (Gregory and Anderson 1997; Gregory et al. 1997). They found that 80% of two-to four-year-olds were associated with rock, boulders, and high bathymetric relief (cliffs) and often maintained fidelity to such features including crevices in rocks. They exhibited significant increases in swimming speed with increasing distance from structure. Yearling cod showed no such connection, 59% of those observed were primarily over gravel and low relief with the fish appearing to rely on cryptic patterns to remain undetected. Macroalgae was neither avoided or preferred by either group. Age-1 and ages 2-4 co-occurred laterally and vertically throughout the study area most abundantly at depths of 60-120 m. Juveniles did not appear to undertake a diel movement shoalward during the winter/ early spring season. However, onshore movements may be initiated during March and April after ice breakup and coincident with nearshore water temperature of ˜2-3°C. The same temperature prompts offshore movements in late autumn (Methven and Bajkik 1994). Sonic tagged age 3 cod (28-33 cm) rested almost exclusively in rocky areas at night during winter (Clark and Green 1990). Between June and September, however, individuals were active nocturnally and wide ranging (>3 km/day), moving daily between deep (30 m) cold water, where they were inactive in rocky areas, to shallow (<15 m) sandy substratum where they were active at night in relatively small feeding areas (˜540-2,580 m2). When the water column became isothermal in September, age 3 cod remained in the shallow water during daylight leading researchers to speculate that the switch from nocturnal to diurnal feeding might be an antipredator strategy, i.e., to avoid being cannibalized at night when adult cod are seasonally active in relatively shallow water. Other common predators of juvenile cod off Newfoundland are pollock (Pollachius virens) and shortfinned squid (Illex illecebrosus). Spatial Depth Gradient of Juveniles For three years following stock collapse, Methven and Schneider (1998) undertook extensive sampling of the Newfoundland coastal zone to a depth of 55 m and by a variety of gears. Finding consistent spatial and diel changes in catch across gears, they interpreted results as characteristic of cod distribution. Catch rate of age-0 cod was inversely related to depth each year, highest at night, and higher at 4-7 m, the center of 0-group distribution during autumn. There was a sharp decrease in catch rate at 20 m (Schneider et al. 1997). Demersal age-0 cod were found almost exclusively alongshore within the northeastern coastal bays of Newfoundland; yearlings extended further offshore and older juveniles were widely distributed on the continental shelf confirming an ontogenetic pattern of movement to deeper water with increasing size. Agedependent distribution was also obvious from trawl station catches on survey transects extending from the coast to hundreds of kilometers offshore (Dalley and Anderson 1997). When the stock was more robust, demersal age-0 cod were distributed more widely onto the shelf. The only coastal region of eastern Canada where the seasonal pattern of distribution for young cod appears to be different is the coastal portion of southern Gulf of St. Lawrence where water temperatures might be too warm during summer months (Hanson 1996). Fine scale distribution studies with trawls found that cod did not occupy water 2-12 m deep along shores of Prince Edward Island during summer. They were mostly absent from shallow waters (<20 m deep) in the Miramichi estuary and the contiguous Shediac Valley coastal shelf during any time of year. Yearlings and 2-year-olds, but not age-0 cod, were almost exclusively found in 15-35 m depths of the Gulf from June to early October before joining older age-groups in an extensive migration to deep (>100 m) offshore water for winter. The spatial depth gradient of juvenile cod from all other areas of eastern Canada seems consistent with published information from the Northeast Atlantic. The depth of highest age-0 cod abundance using a beam trawl off the British Isles was 6 m (Riley and Parnell 1984). Greatest density of age-1 cod sampled with gill nets off Greenland was <20 m (Hansen and Lehmann 1986; Hovgard and Nygaard 1990). Acoustic surveys off the Norwegian coast showed most juveniles at depths <35 m and highest densities of demersal 0-group cod very close to rocky shores where the research vessel could not survey (Olsen and Soldal 1989). Density-dependent Habitat Use and Mortality Contraction or expansion of geographic range with decreasing or increasing population size has been observed in a number of cod stocks including the Labrador-East Newfoundland complex and southern Gulf of St. Lawrence stock. In the latter region, the area occupied by age groups 3-8+ cod increased as abundance increased (Swain and Wade 1993). In comparison to the older cod, age 3 were more spatially restricted at low population size, their range expanded more slowly as abundance increased, and changes in relative density among parts of the Gulf were smaller between years of lowand high-abundance. Younger juveniles were thought to experience less severe competitive pressures for food or wider variation in habitat quality than the older agegroups. A behavioral theory applied to explain the pattern of geographic distribution is densitydependent habitat use. This hypothesis was applied to young cod in coastal habitats (Olsen and Soldal 1989) where catches of post-settlement juveniles showed a high degree of small-scale spatial consistency regardless of cohort size. In years of high yearclass abundance, density increases to an upper limit in the most suitable habitat and as the fitness of individuals occupying the prime sites declines due to intraspecific competition, diffusion to and use of suboptimal habitat expands. Accordingly, at low population size, individuals occupy habitat with high basic foraging and protective suitability. The theory was tested for the Labrador-East Newfoundland stock complex for which contraction has been confirmed for adult cod at low stock size (Taggart et al. 1994; Atkinson et al. 1997). Catches of age-groups 0-2 were analyzed from 1959-64 and 1992- 94 at a series of fixed sampling sites extending over 1,500 miles of Newfoundland coastline (Schneider et al. 1997). In years of low cohort size, contraction did not occur in coastal habitats, i.e., density of juvenile cod was independent of area within the occupied <20 m depth range. They noted that sampling sites with high densities in some years had low densities in years of high abundance, an observation inconsistent with spillover theory in good years. In support of density-dependent theory, high post-settlement densities of age-0 cod were found in eelgrass beds of Trinity Bay, Newfoundland, during 1994 and 1995, years of good and bad year-classes, respectively; however, a significant increase in abundance in less suitable no-eelgrass habitat was noted in 1994 when settlement strength was high (Grant and Brown 1998a). The high 1994 densities in less-utilized no-eelgrass habitat during a year of high abundance would be consistent with the hypothesis of densitydependent habitat use or selection. The researchers acknowledged that their observations were on a small temporal and spatial scale. Re-analysis of the fixed sampling site juvenile catch data from Newfoundland showed a stronger recruitment signal from a small number of sites visited frequently than the entire set of sites (Ings et al. 1997). The 1994 year class was ranked significantly stronger than the three previous year-classes following stock collapse in a broad-scale study (Anderson and Dalley 1997). On the other hand, there was no evidence of fewer settled 0-group juveniles anywhere along the coast in 1995 relative to the 1992-94 year-classes (Smedbol et al. 1998). For a number of cod stocks, variability in year class strength is usually determined in the larval stage and attenuated by density-dependent juvenile mortality (Myers and Cadigan 1993a). Biological processes that may result in density-dependent mortality would include: (1) competition for food with mortality resulting from increased predation or starvation; (2) intercohort cannibalism; (3) predators switching to abundant year-classes; and (4) a circumscribed area of prime juvenile habitat with those settling surviving while others do not, resulting in a upper limit to the number of survivors regardless of egg/larval production. This mechanism could involve food limitation and/or increased predation risk outside a prime nursery area. It presumes mechanisms maintaining a relatively constant density such as territorial behavior or some other form of density-dependent habitat utilization. Notwithstanding the study by Schneider et al. (1997), many of the research results discovered and re-confirmed by scientists undertaking the studies summarized herein, describe or infer habitat mediated density-dependent mortality rates. These mechanisms systematically affect cod survival rates from the post-settlement pelagic stage well into the demersal juvenile stage. Annual variation in survival rates on these life stages may be more important in affecting year class size than survival in presettlement stages (Sissenwine 1984). This suggests that the nearshore bottom habitat may become a potential bottleneck to year-class size particularly in areas where the availability of the most suitable habitat might be low. Summary of Research In shallow (< 5 m) coastal areas of eastern Canada, pelagic juvenile cod settle onto various subtidal habitats in several periodic pulses beginning in May. Age -0 cod settle in shallow waters (< 10m) of the Maine mid coast beginning in April. Space use is highly localized and primarily focused on the need to acquire food and avoid predators. Relative to fulfilling both needs, activity periods, substrate choices, and interactions with members of same species and others are critical. Diurnal feeding in inter-cohort schools aids location of patchily distributed plankton and provides protection against predators. Site fidelity and nightly concealment in all habitats, except sand, minimizes interactions with cannibalistic age-1 cod that move shoalward at dusk to feed. The spatial pattern of age-0 cod distribution is altered by post-settlement mortality such that abundance among bottom habitats matches substratum complexity: cobble/gravel = rock reef > eelgrass = macroalgae > sand. Of bottom habitats studied, eelgrass confers a significant advantage in growth to age-0 cod. Significantly reduced predation risk also occurs if eelgrass stems are above a threshold density and/or they are associated with cobble bottom. Eelgrass meadows are highly utilized as nursery habitat both spatially and temporally through at least mid-summer. The transition to a demersal existence occurs at a length of 6-10 cm and is marked by a switch to benthic prey foraged at dawn and dusk. The distribution of age-0 cod in autumn is centered at depths of 4-7 m MLW with a sharp drop off at 20 m. In late autumn/ early winter, age-0 lose site fidelity and disperse to deeper water where they congregate primarily over gravel and low relief cover. Older juveniles inhabit progressively deeper water and associate with coarser, hardbottom features as they grow. Seasonal inshore movements are usually associated with nocturnal feeding. Age-1 cod, while co-existing in all but the shallowest depths with young-of-the-year, are many times more abundant in the shore zone at night than during the day apparently attracted there by the presence of periodic influxes of postlarval pelagic juvenile cod. Competitive advantage accrues to the largest and earliest settling juveniles especially those finding coarse substratum with vegetative cover. Those less favored must disperse from feeding patches more often thereby accepting a lower rate of food intake in order to avoid detection and capture. As Tupper and Boutilier (1995b) hypothesized: "one habitat might supply the population with a greater number of smaller recruits, each 68 with a somewhat lesser chance of survival, while another habitat supplies fewer, larger recruits, each with a relatively high chance of survival". The trade-offs between habitat use and frequencies of feeding in the face of predation risk are processes consistent with density-dependent habitat use and mortality. Although empirical evidence of density-dependent usage off Newfoundland is contradictory, stock size/recruitment may not yet be large enough following the northern cod stock collapse to induce significant density-dependent effects on a large spatial scale. Nevertheless, behavioral research details ways age-0 juveniles respond to spatial heterogeneity, the consequences for fitness through utilization of resources, and the intra-specific competitive effects which emphasizes the importance of habitat availability and quality in determining recruitment success. References INSERT REFERENCES HERE ???