Fish Bulletin 176 - Lingcod

Table of Contents -- < Backward -- Forward >


Lingcod

Lingcod, a large carnivorous fish of the family Hexagrammidae, is an important and highly sought species in the northern and central California marine recreational fishery. In 1980-86, lingcod ranked among the top 10 most important species (as defined by our IRI) in all fishing modes except pier and dock (Figures 10, 13-17). Lingcod ranked first for spear, third for PRB, and fifth for CPFV and beach and bank fishing. Much of the contribution to IRI resulted from the percentage by weight of total catch. The desirability of lingcod is evidenced by its top rank for the spear mode, where prey species can be targeted. Anglers also commonly target lingcod by jigging large artificial lures, anchovies, or squid.

Lingcod range from Kodiak Island, Alaska to Point San Carlos, Baja California (Miller and Lea 1972). The area of greatest abundance is north of California (Phillips 1959). Commercial landings decrease from northern to southern California, with the sharpest decrease occurring south of Santa Barbara (Miller and Geibel 1973). Percentages of lingcod in the sampled recreational catch from 1980 to 1986 also indicate lingcod abundance drops sharply south of San Luis Obispo (Figure 42). No obvious changes in occurrence appeared during the 1982-83 ENSO. Mendocino/Sonoma had the highest annual average percent of total catch (4.3%) throughout Oregon and California.

In California, lingcod egg masses are deposited from November to early March in rocky habitats (Miller and Geibel 1973). Lingcod egg masses have been observed to depths of 97 m (O'Connell 1993). Water current flow is necessary for gas exchange and prevention of high mortality within the egg mass (Giorgi and Congleton 1984), and nests are typically located in areas of water movement. Male lingcod guard the egg mass from predators during incubation (Hart 1973). Removal of the male causes a high incidence of egg loss to predators (Low and Beamish 1978). Incubation takes about seven weeks in the Puget Sound area, and is probably briefer in California (Miller and Geibel 1973). Newly hatched larvae are 7-10 mm in length (Hart 1973).

After hatching, young lingcod are pelagic until they are about 70 mm long. Larvae and juveniles have been commonly found in the upper 30 m of the water column in the Gulf of the Farallones, with density decreasing with distance from shore (Adams et al. 1993). Phillips and Barraclough (1977) found that the pelagic stage lasts from about early March to early June in British Columbia waters, and that the young lingcod move progressively inshore during their pelagic existence. Timing of the pelagic stage appears to be about one month earlier in California (Adams 1986).

Lingcod lack a swim bladder, and after the pelagic stage they are typically found near the bottom. Newly settled juveniles about 70 mm long are found in sandy areas (Miller and Geibel 1973). At lengths of about 350 mm, they move onto rocky reef areas. Growth of lingcod is rapid in comparison to the rockfish. Lingcod from northern and central California reach 330 mm total length at age one and attain the present recreational fishery minimum size limit of 22 inches (560 mm) as 3-year olds (Miller and Geibel 1973). Adults can be found to depths of 420 m (Miller and Lea 1972).

Most tagging studies have found lingcod to be relatively nonmigratory. In a central California study, nearly all lingcod recovered were caught within about 5 km of the location where they were tagged (Miller and Geibel 1973). However, they reported that catch data indicate that at least a portion of the adult population occupying deeper habitats (over 90 m) annually migrates inshore to spawn. A portion of the adult male population appears to be residential in shallower areas and is available to recreational fisheries year around. Large gravid females become more frequent in the recreational catch in the fall months prior to spawning. An Oregon tagging study found no movement between inshore and offshore locations (Barss and Demory 1989). However, of 149 fish recaptured after being tagged near the San Juan Islands, 61 had moved from 8.1 to 50 km, and 13 had moved more than 50 km (Mathews and LaRiviere 1987). The general direction of movement was seaward. Nearly all those fish were tagged during March through May, which is during the December through May nesting period for the San Juan Islands (LaRiviere et al. 1981). Thus the movement may represent a tendency of some fish to return to more seaward habitats after spawning and nesting. Jagielo (1990) tagged lingcod within 3 miles of shore in March and April near Neah Bay, Washington; he found that 19% of the recaptured fish were recovered more than 5 miles from the tagging site, and the general direction of movement of those fish was seaward. If a tendency to migrate inshore to spawn in relatively shallow habitats exists, it may be related to a need for water circulation past the egg mass provided by surge and tidal currents. An important management question is whether offshore fish taken mainly by commercial fisheries are different stocks from inshore fish taken mainly by recreational fisheries.

The take of lingcod off California is subject to regulation by both federal and state governments. Since 1983 the PFMC has specified an annual acceptable biological catch (ABC) for Washington, Oregon, and California of 7000 MT, within which 500 MT, 1000 MT, and 400 MT are allocated to the Eureka, Monterey, and Conception INPFC areas, respectively (PFMC 1992) (Figure 1). Due to lack of knowledge of lingcod population dynamics, the ABCs are based on harvest levels rather than stock condition (PFMC 1982). For recreational fishing in California, federal and state regulations specify a five-fish bag limit and a 22-inch minimum total length. The bag limit was initiated in 1980 and the minimum length in 1981. Prior to 1980 the bag limit was 10 fish.

Catch

The estimated average annual number of lingcod landed in northern and central California recreational fisheries, excluding PRB fishing in San Francisco Bay, rose from 65,000 in the 1958-61 survey to 112,000 in the 1981-86 survey (Figure 6). Average annual weight landed rose 74%, from 232 MT to 403 MT. Average weight per fish did not change between the two surveys (3.6 kg).

Total 1981-86 average annual recreational lingcod catch in northern and central California, including skiff fishing in San Francisco Bay, was 113,000 fish weighing 405 MT (Table 8). Nearly all fish were landed by boat modes. The Del Norte/Humboldt, Mendocino/Sonoma, and Santa Cruz/Monterey districts made strong contributions to total landings. Mean weight per fish exhibited a north-south cline, from 4.6 kg in Del Norte/Humboldt to 2.6 kg in San Luis Obispo.

As discussed earlier, CPFV log data are not reliable for use as annual catch estimates but may indicate fishery trends. CPFV log data indicate that lingcod catch, percent of total catch, and catch per fishing day peaked in 1972, 1980, and 1989, and have been in slow oscillating decline since the early 1970s (Figures 43 and 44).

There is presently no closed season for recreational or commercial harvest of lingcod in California. Monthly trends in recreational and commercial landings show that lingcod are taken year around, with highest landings in June through October (Figure 45). Relatively calm sea conditions during those months are at least partial cause for the higher landings. Thirty-seven percent of the recreational landings are made during the 6 months of greatest nest guarding activity, November through April. Miller and Geibel (1973) found recreational catch per day generally peaked in the winter months, a time of relatively low fishing effort. They attributed the peak to increased availability of fish in nearshore areas.

Northern and central California commercial landings of lingcod rose from 606 MT in 1958-61 to 942 MT in 1981-86, an increase of 55% (Figure 5). Weight of commercial landings was about twice the weight of recreational landings in 1981-86. Long-term trends in commercial landings nearly parallel the CPFV log landings, showing a sharp increase between 1969 and 1972 and an oscillating decline since then (Figure 43). The correlation between the commercial and CPFV landing data indicates the two fisheries are fishing the same stocks, or if the stocks are different, that the same factors are affecting abundance and year-class strength.

The decline in CPFV catch since the early 1980s was probably caused partially by the five-fish bag limit that began in 1980 and the 22-inch minimum size limit that began in 1981. However the decline lasted well beyond those two years, and commercial landings not subject to those limits also declined.

The percentage of the commercial catch taken by hook and line and entangling net gear has risen since the early 1980s (Figure 46). Trawl fisheries now account for only about half the commercial catch.

Length-frequency Analysis

Mean length of recreationally caught lingcod during 1980-86 was significantly different in northern California (695 mm) from central California (626 mm) (p < 0.0001), and was significantly different in the CPFV catch (674 mm) from the PRB catch (653 mm) (p < 0.0001); the interaction between geographical area and mode was also significant (p < 0.0001; two-way ANOVA).

Comparison of the northern California and central California length-frequency distributions segregated by mode showed that the shapes of the distributions did not differ greatly, and that the difference in mean length between the CPFV and PRB modes was much greater in central California (70 mm) than in northern California (11 mm) (Figure 47).

Since historical length sampling was not uniform with respect to time, location, or different fishing modes, sufficient length data existed to analyze historical length-frequency trends only for the CPFV catch in central California. Mean length in 1959-64 (601 mm) was significantly different from 1966-72 (659 mm) (p < 0.0001) and was also significantly different from 1980-86 (663 mm) (p < 0.0001; one-way ANOVA followed by Scheffe test) (Figure 48). Because the 22-inch (560-mm) size limit began in 1981, we applied the same tests to the data excluding all lengths less than 560 mm. For lengths greater than or equal to 560 mm, mean length in 1959-64 (671 mm) was significantly different from 1966-72 (710 mm) (p < 0.0001) and was also significantly different from 1980-86 (684 mm) (p = 0.05); mean length in 1966-72 was significantly different from 1980-86 (p < 0.0001). The length-frequency differences were probably not caused as much by differences in take as by presence of strong year-classes, as discussed below.

Plots of annual length-frequency data for boat modes from central California show four length-frequency modes of one or more dominant year-classes progressing through the fishery (Figure 49, modes A through D). Discerning length-frequency modes after 1980 is difficult due to small sample sizes. Ford-Walford analysis of the annual increments in length indicated by progression of the length-frequency peaks yields an age-length relationship similar to the age-length relationship for northern and central California males developed through surface reading of otoliths by Miller and Geibel (1973), except the Ford-Walford analysis failed to model the rapid growth occurring in the first year of life (Figure 50). Ford-Walford plot analysis applied to growth intervals of length-frequency peaks A through D produced a Von Bertalanffy growth equation of

l t+1 = 0.7941 + 152

where

L¥ = 741.

The similarity in slope between the two age-length relationships supports the Miller and Geibel (1973) relationship and also indicates length-frequency peaks A through D (Figure 49) actually do track growth of one or more dominant cohorts.

Using the Miller and Geibel (1973) age-length relationship for males (since males dominate recreational catches), we assigned approximate birth years for the dominant cohorts as follows (Figure 49): A-1956, B-1960, C-1968, and D-1977. The ecological implications of those dominant cohort birth years are discussed later.

The strong 1960 year-class affected the length-frequency analysis, contributing many relatively small fish in 1962-64, and many relatively large fish during 1966-72 (Figure 49); that strong year-class and the 1968 year-class are at least partially responsible for peak landings in the early 1970s and the apparent declines since then. Strong year-classes probably also caused the peaks in CPFV and commercial catch around 1980 and 1989. Since the early 1970s, the coincidence of strong year-classes with peaks in CPFV and commercial landings data (Figure 43) indicates that the CPFV and commercial landings data probably do reflect trends in stock abundance.

Length at Maturity

Knowledge of the length at which fish become sexually mature is important in evaluating the length-frequency composition of the catch, and also in evaluating a minimum size limit as exists for lingcod. The length at which lingcod mature has been found to vary within the species' geographic range, with males maturing at smaller lengths than females. British Columbia lingcod males and females examined in December through April matured at 520 mm and 768 mm, respectively (Wilby 1937). Fort Bragg lingcod males and females examined in October and November matured at 580 to 648 mm (Phillips 1959). Miller and Geibel (1973) examined 111 males and 180 females taken by CPFVs in the Monterey and Morro Bay areas in December through mid-February, and found that males matured at 390 to 590 mm, and females matured at 510 to 765 mm.

Richards et al. (1990) examined length and maturity relationships of lingcod from three areas in British Columbia; they found that males began to mature at 500 mm and were all mature at 700 mm, and females began to mature at 500 mm and were all mature at 750 mm. Length at maturity increased with increasing latitude for the three areas. They also found that the Miller and Geibel (1973) lingcod maturity data supported a conclusion that length at maturity increases with increasing latitude for both sexes coastwide.

The difference in length at maturity between central California and British Columbia is much larger for males than females (Figure 51). Length at maturity for Fort Bragg males (Phillips 1959) appears nearer to that of British Columbia males than to central California males. The central California data indicate that the 22-inch size limit allows for about 90% of the males but only about 30% of the females to mature before entering the recreational fishery. The Fort Bragg data indicate that the 22-inch size limit allows a low percentage of both males and females in northern California to mature before entering the recreational fishery.

The difference in male length at maturity between central California and northern California is surprisingly great relative to the difference between northern California and British Columbia. Since male lingcod play a critical nest-guarding role in reproduction and they are commonly taken in the recreational fishery, the difference is significant to fishery management.

Rockfish and Lingcod: Yearly Trends and Conclusions

An examination of yearly recreational trends in landings from 1981 to 1986 for northern and central California for rockfish, lingcod, and salmon suggests dynamic fisheries (Figure 52). In northern California, salmon and lingcod played dominant roles, and there was no discernible pattern of landings over the 6-year period among most of the species examined. In central California, salmon and lingcod play reduced roles, and there was a trend of decline in blue rockfish from 1982 to 1986 that was offset by increases among other species such as bocaccio, chilipepper, greenspotted rockfish, and salmon. Thus the fisheries, especially in central California, are in a state of flux and depend on availability of the desirable species.

It is difficult to argue, based on the data we examined, that any one stock of rockfish or lingcod is in immediate danger of collapse. The 1958 to 1986 recreational data had much variance and, except for the more common species such as blue rockfish and yellowtail rockfish, were too discontinuous over time to document interannual variation. The long-term average trends we were able to follow do not yet show the dramatic stock declines seen in some commercial fisheries such as those for abalone (Karpov and Tegner 1992) and California halibut (Barsky 1990).

Documenting stress in a stock is inherently difficult for recreational fisheries. One problem is a lack of reliable yearly landing data such as is available for commercial fisheries. The survey data we relied on was beset by problems of small sample size, resulting in high annual variance in species' catch estimates (Miller and Gotshall 1965; Miller and Geibel 1973; Albin et al. 1993). The dynamic nature of recreational fisheries in shifting among species such as salmon, lingcod, bocaccio, and chilipepper as they become available, or as other desirable species decline, obscures catch trends. Increases in fishing power further mask evidence of stock decline (Gulland 1969; Ricker 1975). The shift to deep-water rockfish by PRB and CPFV modes as shallow-water rockfish declined is an example of an increase in recreational fishing power. Sonar fish finders have become cheaper, more sophisticated, and common on both CPFVs and PRBs. We have observed that fishing grounds previously inaccessible to PRBs have become accessible as vessel size and horsepower have increased. Boats now go further from port as near-port assemblages have been fished down (J. Mason, NMFS, pers. observ.).

Since 1958, average weight per fish declined for 12 of the 16 rockfishes we examined, including bocaccio, canary rockfish, vermilion rockfish, chilipepper, greenstriped rockfish, widow rockfish, starry rockfish, blue rockfish, black rockfish, brown rockfish, gopher rockfish, and olive rockfish. Among species also examined for length-frequency distribution only bocaccio, a fast growing species that reaches 50% sexual maturity in three to four years (Echeverria 1987), had a decline in size that could be attributed to an incoming strong year-class (1984). Blue rockfish, yellowtail rockfish, canary rockfish, and brown rockfish showed declines that may be related to environmental factors or fishing pressure.

Evidence of stress on lingcod and rockfish stocks was greatest in central California where sport effort was also greatest. A north-south cline of decreasing mean weight was evident for lingcod and most of the rockfishes (Tables 6 and 8). Length-frequency comparisons of lingcod, blue rockfish, yellowtail rockfish, canary rockfish, and brown rockfish showed that sizes taken were significantly larger in the north.

Our results suggest that, of the rockfishes, blue rockfish, yellowtail rockfish, canary rockfish, and brown rockfish are the main species requiring management attention. Blue rockfish and yellowtail rockfish are the most recreationally important rockfish in central and northern California. Annual average recreational catches of blue rockfish and yellowtail rockfish doubled between 1958-61 and 1981-86 to 416 and 350 MT respectively (Table 5), together representing 31% by weight of the 1981-86 recreational rockfish landings. Both species showed a loss of large adult sizes in central California during 1982-83 ENSO.

Since 1982 the central California blue rockfish catch showed a dramatic 88% decline from 500 MT in 1982 to 60 MT in 1986 (Figure 52). Whether the decline resulted more from fishing pressure or environmental factors could not be established from our study. Miller and Geibel (1973) warned that blue rockfish are localized as adults and could be fished out of an area. The difference in blue rockfish recruitment patterns we observed between central California and northern California suggests that separate stocks may exist that warrant separate management approaches. The large proportion of sexually immature fish in the central California catch is cause for added caution in managing this stock.

Yellowtail rockfish showed a significant slope of decline in length among both recreational and commercial fisheries that, when viewed in conjunction with the higher proportion of sexually immature fish taken by the recreational fisheries, suggests a need for management attention. Evidence suggests a California-Oregon substock. Fraidenburg (1980) and Tagart (1991) both reported north to south differences in yearly recruitment patterns as reflected in surviving dominant cohorts. Tagart (1991) interpreted the differences as suggesting a southern stock that corresponded to the Eureka/S. Columbia INPFC area. Our length-frequency analysis suggested similar recruitment patterns among fish taken in recreational fisheries in northern and central California. We also found that the 1982-83 ENSO and/or greater fishing pressure eliminated an older cohort in central California and not northern California. The result was that by 1986 most recreationally caught fish in the central area were sexually immature (Figures 31 and 32).

Canary rockfish and brown rockfish ranked only sixth and seventh in weight landed (Table 5) but also showed evidence of stress, requiring management attention. Canary rockfish declined in length and weight per fish between the 1958-61 and 1981-86 surveys as recreational landings doubled to 99 MT (Figure 40). A significant slope of decline in mean length is apparent in recent years for both commercial trawl and recreational catches in northern California (Figure 40). There was a high portion of sexually immature fish in the recreational catch (Figures 39 and 40). Canary rockfish are slow growing, taking seven to nine years to reach sexual maturity (Echeverria 1987). Adams (1992b) noted declines in commercial length and suggested the species needs additional monitoring in the future.

The annual average recreational catch of brown rockfish increased from 26 MT to 91 MT between 1958-61 and 1981-86 (Table 5). Brown rockfish showed the greatest drop in average weight per fish of all 16 rockfishes, declining from 1.14 to 0.58 kg/fish, a 49% decrease. Lengths declined to the point where, as with canary rockfish, most of the recreational catch in 1980-86 was sexually immature.

Recreational and commercial catches of lingcod rose by 74% and 55%, respectively, between the 1958-61 and 1981-86 surveys (Figure 5). The CPFV catch, percent of CPFV catch, CPFV catch per fishing day, and commercial catch have been in slow oscillating decline since the early 1970s, with the oscillations caused by strong year-classes (Figures 43 and 44). The average slope of decline has been about 2% per year for the commercial catch and 3% per year for the CPFV catch. Mean length in the central California CPFV catch, adjusted for the 22-inch size limit, decreased significantly between 1966-72 and 1980-86; the average decrease was 1.9 mm per year. The decreases in catch and mean length may partially be artifacts of the strong 1960 year-class, and the decreases in CPFV catch statistics since 1980 may partially be artifacts of bag and size limit changes. Adjusted mean length in the central California CPFV catch increased significantly between 1959-64 and 1980-86. Considering the above, we are uncertain if the decreases in catch statistics and mean length since the early 1970s have been caused by stress to the population and represent a trend that will continue, or if they are effects of ongoing natural variations. The rates of the apparent decreases do not portend imminent fishery collapse. In 1993, relatively large numbers of juvenile lingcod were seen in underwater surveys (D. VenTresca, CDFG, pers. comm.), so landings may rise again in the near future. However lingcod stocks are vulnerable to overfishing. Stocks have become depressed, probably due to overfishing, in Puget Sound, Washington and in the Strait of Georgia, British Columbia (Yamanaka and Richards 1993). The desirability of lingcod combined with the decline of salmon and nearshore gill net fisheries provide a high likelihood that additional fishing pressure will focus on lingcod stocks, and more specific and efficient fishing methods will develop. For example, nearshore commercial hook-and-line fishing for lingcod and rockfish has increased greatly in recent years.

Management Options for Rockfish and Lingcod

Several approaches to the management of the recreational fishery for rockfish have been recommended including reduced bag limits, size limits, seasonal closure, and rotating area closure (Miller and Geibel 1973; VenTresca 1991). Problems associated with recreational management of rockfish include 1) difficulty in identifying individual species and 2) mortality associated with swim bladder expansion and stomach eversion that precludes catch and release as a management tool for most species. Studies of blue rockfish (Miller et al. 1967), a shallow-water species, showed that it would not suffer undue mortalities from capture and release if a size limit of 254 mm (10 inches) were imposed. Miller and Geibel (1973) suggested such a size limit would protect the stock from excessive mortality of sexually immature individuals. Blue rockfish are also easy to distinguish from other species except black rockfish (Miller and Lea 1972). The size minimum could be applied to both species. Relatively few black rockfish are taken in central California, while in northern California almost all were larger than 254 mm in 1980-86 (Figure 34).

Another strategy, which would decrease recreational impact on rockfish stocks collectively, would be to reduce the current bag limit of 15 fish in combination of species. A 10-fish limit would reduce the number taken by 9% while a six-fish limit would reduce the number taken by 25%. A 10-fish limit would affect only 6% and a six-fish limit 14% of the anglers with catches larger than the proposed limits (Figure 53).

Seasonal closure is another alternative that could be applied individually or collectively to the rockfish complex. One or more months could be closed during the peak rockfish take period of June through October (Figure 23). A closure to all species of rockfish would be less palatable for the recreational fishery of central California where an alternate fishery for salmon is less available.

Rotating area closure is another management option that could be used to reduce impact on species such as blue rockfish and yellowtail rockfish that have been described as residential and subject to localized exploitation (Miller and Geibel 1973; Carlson 1986). Carlson and Haight (1972) described evidence of home site and homing by adult yellowtail rockfish. Carlson (1986) studied a school of yellowtail rockfish on a sunken ship over an 11-year period and found negligible recruitment as a residential cohort grew. Miller and Geibel (1973) found little movement among tagged adult blue rockfish. Matthews (1990) displaced tagged copper rockfish, quillback rockfish, and brown rockfish. The fish showed homing behavior, returning to areas of quality habitat. Miller and Geibel (1973) described rotating area closures as feasible if two to three areas were rotated for periods of three to five years for each port area. They argued that such closures could allow increased sizes for heavily exploited stocks to develop provided that such closures were enforceable and applied to both inshore recreational and commercial fisheries. Increased pressure on areas left open could be addressed by simultaneously imposing decreased bag limits, size limits, or seasonal closures.

An alternative to rotating closures is permanent closure of selected areas to both recreational and commercial fisheries (Davis 1989). Such areas could provide refuge for brood stocks of large residential adult rockfish and other nonmigratory species, which could provide recruitment for other areas of the coast. An alternative to recreational and commercial closure of specified areas would be commercial closure only. For example, selected nearshore areas could be closed to commercial take of rockfish.

As our human population has grown, fishing technology has rapidly advanced and demands on fishery resources have rapidly increased. Management actions have frequently been too little and too late to stem fishery declines. Permanent closure of selected areas may become one of the few reliable management strategies for protecting fishery stocks (particularly rockfish that are late-maturing, long-lived, and have low natural mortality) for use by future generations.

Trends in lingcod catch should be closely monitored. If the downward trend continues, actions to reduce harvest should be taken. Since lingcod and rockfish are taken by both the recreational and commercial fisheries, management strategies that reduce overall fishing effort, such as area or seasonal closures that prohibit all fishing, would reduce harvest of lingcod. Options that might apply only to lingcod include decreasing the bag limit, increasing the minimum size limit, and prohibiting take of lingcod during the egg incubation season.

Decreasing the bag limit for lingcod would decrease recreational harvest. Decreasing the bag limit to three fish would reduce recreational harvest by 11% and decrease the catch of 10% of the anglers; a two-fish limit would reduce harvest by 21% and reduce the catch of 17% of the anglers (Figure 54).

Increasing the minimum size limit for lingcod would decrease take, and may provide additional opportunity for fish to reproduce before being caught, resulting in a higher quality fishery of larger fish. Miller and Geibel (1973) suggested a limit of 24 inches (610 mm), with an interim limit of 22 inches (560 mm) gradually progressing to 24 inches (610 mm) to minimize economic impacts to recreational fisheries. The present 22-inch limit could be increased to the 24-inch recommendation of Miller and Geibel (1973). Adams (1986) estimated that increasing the size at recruitment increased yield of lingcod biomass only under conditions of high fishing mortality (F = 0.7), but it increased yield of eggs under both high (F = 0.7) and low (F = 0.2) fishing mortality. Any effective program to increase size at recruitment should apply to commercial as well as recreational fisheries. One relatively simple and logical action would be to apply the recreational minimum size limit to commercial hook-and-line fisheries.

Prohibiting take of lingcod during the egg incubation season would decrease overall take and may also avoid take of nest-guarding males and decrease egg mortality. The aggressive behavior of nest-guarding males makes them especially susceptible to spear fishing and may also make them more susceptible to lures or bait. The State of Alaska has closed some nearshore areas to lingcod harvest from January through May to protect nest-guarding males from hook-and-line fisheries (V. O'Connell, Alaska DFG, pers. comm.). In northern and central California, closure of lingcod harvest during the egg incubation season (November-April) could be accomplished without affecting the period of major harvest (Figure 45).

A minimum size limit requires unhooking and releasing any undersize fish caught. Bag limits and seasonal closures also require release of fish if other species are being targeted after the limit has been reached or the season closed. Those management strategies require high survival of hooked-and-released fish. Survival of hooked-and-released lingcod is not known, but has been presumed high since they lack a swim bladder and have relatively tough mouth tissues. The use of bait, which the fish frequently swallow, probably results in lower survival than the use of artificial lures.

Areas for Future Work

Future research should be directed towards both the principal species taken and also those undergoing stress. Types of research needed include identifying units of stock by region, assessing stock size, exploring feasibility of rotating area closures, and identifying the impacts of other fisheries. Species that should be considered include blue rockfish, yellowtail rockfish, canary rockfish, brown rockfish, and lingcod. Stock size assessment models such as stock synthesis or cohort analysis could be developed for both blue rockfish and yellowtail rockfish. A precursor to such assessments should include age- structure analysis of recruitment patterns by area in central and northern California to identify more clearly what areas constitute isolated stocks. Our blue rockfish length-frequency analysis found little modal dominance from San Mateo County north, and similar modal progressions from Santa Cruz County south. Age data are needed, especially in the north, where modal dominance may be obscured by larger and perhaps older fish that have yet to be fished down.

Stocks of yellowtail rockfish off California are not clearly understood (Tagart 1991). Age analyses have been applied exclusively to trawl catch, not to recreational catch which comprises the majority of landings south of Sonoma County. Recruitment patterns could be used to determine whether the same unit stock is being exploited by both fisheries throughout California (Tagart 1991). The same types of questions applied to yellowtail rockfish and blue rockfish should be addressed for canary rockfish and brown rockfish. Unfortunately, recreational fisheries for those species may not be large or important enough to warrant the level of research effort needed for stock assessment.

Future investigations of lingcod should focus on whether lingcod stocks are in decline and on what management measures may be necessary. Such work should include estimation of current recreational catch, further analysis and evaluation of total commercial and recreational catch, and age-structured stock assessment. Since population levels and harvest are apparently driven by strong year-classes, monitoring year-class strength at an early age could be beneficial to management. Tagging studies and spawning-nest surveys using submersibles or remote cameras would help define inshore and offshore stocks. A study of the survival of fish that are hooked and released would help evaluate management alternatives involving catch and release of lingcod. A study of the impact of fishing on nest-guarding males and resultant egg mortality would help determine the value of a seasonal closure. Additional study of lingcod length at sexual maturity in different northern and central California locations would clarify length-maturity relationships and the value of the present recreational size limit and alternatives.

The feasibility of rotating area closures should be investigated as a management alternative. A precursor to such an investigation would be establishment of reserves in presently exploited areas. Each reserve should consist of several miles of coastline at depths encompassing the ranges of residential species. Residential species such as blue rockfish, subadult brown rockfish, black-and-yellow rockfish, gopher rockfish, kelp rockfish, treefish, yellowtail rockfish, china rockfish, cabezon, kelp greenling, and nest-guarding male lingcod (Miller and Geibel 1973; Feder et al. 1974; Matthews 1989; Larson 1992) could be monitored in the newly established reserves and in adjacent areas. Such studies could involve intensive creel survey, tagging studies, and research collections. Initially, intensive surveys would be needed to establish baseline species composition, stock abundance, and length frequencies. Changes in size and CPUE could be monitored over time to establish species recovery rates.

Our findings, based indirectly on species composition, are that CPFVs and PRBs have shifted toward the utilization of deep-water rockfish stocks since the 1958-62 survey period. Location and depth-at-capture data are needed from future marine recreational fishery surveys to more effectively evaluate changes in fishing power and the effects of competing fisheries.

Of most immediate management importance is an assessment of the impacts on nearshore (within three miles) stocks of the commercial long-line and other "alternate gear" fisheries that in recent years have replaced set-net fisheries (Haseltine and Thornton 1990) and have expanded greatly into nearshore areas formerly subject mainly to sport fishing.

Implications of Dominant Cohorts for Rockfish and Lingcod

Five of the six rockfish species with sufficient length-frequency data for interannual comparisons showed clear evidence of modal progressions indicating dominant periods of recruitment. The modal progressions were evident in spite of limitations, such as differential growth rates between sexes, inherent in using length distributions to follow year-classes. The sixth species, chilipepper, lacked a clear length-frequency pattern in the data we examined, but had strong year-class dominance determined from aging studies (Rogers and Bence 1992). Lingcod was also found to follow a pattern of year-class dominance. Year-class dominance is probably common among rockfish and lingcod in central California.

Using Ford-Walford plot analysis for blue rockfish and published ages for lingcod, chilipepper, bocaccio, and yellowtail rockfish, we plotted approximate birth years for major cohorts discussed in the text (Figure 55). The only cohorts examined were for the period when length-frequency data were available.

ENSO events (Dayton and Tegner 1989; Norton and McLain 1994), sea level height, southward transport, and zooplankton volume (Chelton et al. 1982) were compared to the timing of the recruitment events (Figure 55). Chelton et al. (1982) provided a time series of zooplankton volume and southward transport in the California Current that was continuous from 1950 through 1969. Their time series is intermittent following 1969 when CalCOFI cruises were cut back. They reported that San Diego, Los Angeles, and San Francisco averages of low sea levels corresponded to above normal southward transport and zooplankton biomass. Sea level heights provide a longer time series for comparison of changes affecting the California Current. Extremely high sea levels correspond to the three most recent major ENSOs in 1941, 1958, and 1983 (Chelton et al. 1982; Hollowed and Wooster 1992). Norton (1987) and Hollowed and Wooster (1992) used combinations of such physical parameters to characterize "cold" or "warm" years that were related to recruitment events for 14 species of groundfish.

The limited number of species we examined, imprecise methods used to assign birth year, and short time series for some species preclude any strong inferences from our results. However some generalizations can be drawn and comparisons made to the results of Norton (1987) and Hollowed and Wooster (1992) .

Generally, except for bocaccio and chilipepper in 1984 and lingcod in 1960, we did not find major recruitment events during or immediately following the 1957-59 or 1982-83 ENSO periods. Hollowed and Wooster (1992) found synchrony of extremely weak or strong year-classes for various species of northeastern Pacific groundfish. Synchrony of extremely weak year-classes occurred in 1953, 1954, 1958, 1959, and 1983 supporting our result that recruitment failure is likely during major ENSO events. They also found synchronous strong recruitment events in 1961, 1970, 1977, and 1984 that were associated with "warm" periods following major ENSOs characterized by reduced southward transport in the California Current. Among the species we examined, only bocaccio and lingcod had major cohorts in 1977 and bocaccio and chilipepper in 1984.

Norton (1987) compared sizes of year-classes recruited into California commercial catches of widow rockfish and chilipepper to the ocean climate for the 1965 to 1980 period. Most widow rockfish were recruited during "warm" years, and most chilipepper during "cold" years. The strong 1975 year-class for chilipepper fits the pattern-a year of low sea levels, high zooplankton, and southward transport-while the 1984 "warm" year (Hollowed and Wooster 1992) does not (Figure 55).

Generally, major recruitment events for blue rockfish or yellowtail rockfish occurred during years of elevated southward transport and zooplankton abundance (1954, 1968, 1974, and 1975) (Figure 55). Exceptions occurred for blue rockfish in 1962, a year of elevated southward transport but not zooplankton. Exceptions for yellowtail rockfish occurred in 1976, a year of high southward transport, average zooplankton, and low sea level, and in 1980, a year of low southward transport.

Lingcod showed no pattern of year-class strength and oceanic conditions other than absence of recruitment during the two major ENSOs (Figure 55). The 1956 and 1968 year-classes were during years of high southward transport and zooplankton production, while 1960 and 1977 were both years of low southward transport.

Recruitment failure of yellowtail rockfish and blue rockfish during major ENSO events seems likely considering the reduced growth and increased mortality of both species during the 1982-83 ENSO. Direct impacts on gonadal development of yellowtail rockfish (Lenarz and Echeverria 1986) would reduce the number of larvae released. Direct mortality of adult blue rockfish described by Bodkin et al. (1987), or implied by our length-frequency analysis for both species, would directly affect the number of spawners. An additional source of mortality not addressed in our study is juvenile mortality in nearshore nursery areas. Warm water and decreased upwelling of nutrients associated with ENSO events directly reduced kelp canopy in southern California (Dayton and Tegner 1989) and would be expected to impact survival of juvenile blue rockfish that utilize those areas as nurseries.

Blue rockfish and yellowtail rockfish have epipelagic larvae and juveniles found in the California current (Miller and Geibel 1973; Ahlstrom et al. 1978; Lenarz et al. 1991). The larvae and juveniles of lingcod are also epipelagic with a strong nearshore distribution (Adams et al. 1993). Periods of southward transport associated with high zooplankton productivity (Chelton et al. 1982) may favor survival of larvae or presettled juveniles of blue rockfish, yellowtail rockfish, and lingcod.

Areas for Future Work

Additional studies are needed to assess environmental effects on rockfish and lingcod including 1) studies to establish population age structure; 2) studies to determine life stages contributing to establishment of strong year-classes (e.g. spawning success, larval or juvenile survival); and 3) the relative effects of ocean climate change by area of California.

To clearly identify periods of year-class success, aging based on age structures is needed for more rockfish and lingcod. Our study, which based age indirectly on size data and published age determinations from commercial trawl samples, is of limited value and cannot be applied to all species nor to all areas for the same species. Except for brown rockfish, all species we examined have differential growth rates for males and females (Miller and Geibel 1973; Echeverria 1987), which would obscure modal-progression analysis, especially at larger sizes where differences are most pronounced. In northern California, where blue rockfish stocks had not been fished down to the sizes seen in central California, the lack of discernible modes may in part reflect an accumulation of many year-classes of older fish. Age data derived from trawl fisheries may not be representative of inshore catches of all species. Trawl samples of yellowtail rockfish represented larger fish taken mostly in northern California, while recreational catches were of smaller fish taken throughout California.

Accurate age-structure analysis of adults would identify strong cohorts to match ongoing larval, juvenile, and adult assessments to identify the genesis of strong year-classes. Quantitative data from CalCOFI larval assessments and juvenile rockfish surveys, both offshore using trawl surveys (Lenarz et al. 1991; Adams 1992a) and nearshore in nursery areas, could be compared to results from adult age-structure analysis. Ongoing studies on the condition of adults reflecting gonadal condition (e.g. Lenarz and Echeverria (1986) on yellowtail rockfish) are needed to establish when year-class success or failure can be tied to spawning success.

Roughgarden et al. (1988) described a cline of diminishing nearshore Ekman-based upwelling from central California to Oregon that collapsed during the 1983 ENSO. The effects of the upwelling and ENSO events on growth could be more fully tested using adult age structures. A recent study by D. Woodbury (NMFS, unpublished data) on yellowtail rockfish showed that significant decreases in otolith deposition occurred during the 1982-83 ENSO. Similar effects were observed with otoliths of chilipepper (J. Mello, CDFG, pers. comm.) and canary rockfish (M. Yoklavich, NMFS, pers. comm.). Such work could logically be applied to blue rockfish. Blue rockfish are known to feed in response to periods of upwelling (Hobson and Chess 1988) and showed depressed growth during the 1982-83 ENSO (D. VenTresca, CDFG, pers. comm.). Otoliths from different locations along the California coast should be examined for evidence of clinal effects of the 1982-83 and 1992 ENSO events.


Table of Contents -- < Backward -- Forward >