The biomass of sardine (age one and older) off California and northern Baja California was estimated to have been about 463,000 mt (CV 44% and 95% confidence interval from 260,000 to 997,000 mt ) on July 1, 1996. That estimate was based on output from a stock assessment model called CANSAR.

For the first time, biomass estimates from CANSAR were compared to biomass estimates from two other stock assessment models (Virtual Population Analysis or VPA and an age structured model without catch at age data). Results from all three models were similar, although there was considerable uncertainty in biomass estimates for recent years.

Questions about stock structure and distribution were major sources of uncertainty in the sardine assessment. Recent data show increased sardine abundance in the Pacific Northwest off the Columbia River and as far north as British Columbia, but it was not possible to determine if those fish were part of the stock available to the California fishery and whether the biomass of sardine in northern areas was included in estimates from the models. Similar questions exist about sardine in Mexican waters south of Ensenada. These problems were confounded by the fact that survey and fishery data which were mostly collected within the Southern California Bight. Additional information is needed to resolve these questions because different modeling approaches gave similar results.

Estimates of sardine spawning biomass were available in 1986, 1987, 1988, 1994, 1995 and 1996 based on the Daily Egg Production Method (DEPM). Prior to 1994, spawning habitat was confined to waters south of Point Conception, where egg and adult data were collected. For those earlier years, survey coverage probably encompassed the geographic limit of spawning activity. Since 1994, spawning habitat has expanded as far north as the Pacific northwest. However, most adult fish data (spawning frequency, batch fecundity, mean female weight) were still from specimens collected off southern California or nearshore in the Monterey area. There is evidence of clines in growth rate and age composition based on latitude and distance from shore. Since female reproductive parameters vary with size and age, DEPM estimates in recent years may be biased due to the lack of adult fish samples from northern and offshore waters. Also, it is likely that surveys of spawning area since 1994 have not encompassed the northern limit of spawning habitat, which would result in underestimation of spawning biomass.

A harvest formula originally selected as the preferred option in the draft CPS-FMP was used by CDFG to calculate fishing quotas during 1996 and 1997. Using that formula and the 1996 estimate of biomass, a quota of 49,000 mt was established for the 1997 California fishery.

The California fishery for chub mackerel has declined precipitously since 1990. Statewide landings during 1996 continued the downward trend, and totaled only 10,300 mt. During the last few years, the principal cause of reduced catches has been low biomass and poor availability on the traditional fishing grounds in southern California waters. Cannery closures since 1993 may have also affected demand.

Several sources of information were available on the status of the chub mackerel stock, all of which suggest a decline in biomass compared to the late 1970's and 1980's. Landing statistics were available since 1978 for both the U.S. and Mexican fisheries, and both fleets show reduced catches during recent years. Catch rates for the CPFV fleet have also shown declining trends since the 1970's. Other fishery-independent data from aerial observations (spotter pilots) and plankton surveys (CalCOFI mackerel larvae samples) indicated lower abundance compared to the early 1980's.

A tuned virtual population analysis (VPA) model called ADEPT was used to estimate chub mackerel biomass. The model finds the best statistical fit between fishery-based, age-structured biomass estimates and other data from the aerial observations, plankton surveys, CPFV catch data, and a spawner-recruit relation.

Based on the estimated number of fish in each year class during the last quarter of 1995 (including the 1995 year class), and using certain assumptions for expected fishing mortality during the first half of 1996, chub mackerel biomass was projected to be 47,000 mt at the beginning of the 1996/97 fishing season, on July 1, 1996. There is a large degree of uncertainty in the 1996 biomass estimate because the 1995 year class (fish of age one) comprised most of the biomass.

The harvest formula for chub mackerel specifies that when biomass is between 18,000 and 136,000 mt, the season's quota shall be 30 percent of the total biomass in excess of 18,000 mt. Using that formula and the biomass projection for July 1, 1996, a commercial fishery quota of 8,700 mt was established for the 1996/97 California fishing season.

For several years, priority has been given to acquiring data from north of the Southern California Bight, and much of those data were used in the most recent sardine assessment. Since 1990, four sardine icthyoplankton surveys were conducted in waters north of Point Conception. Unfortunately, it was not possible to incorporate data from the two northernmost surveys in the most recent assessment. The 1995 CDFG survey was conducted during August, which proved to be too late in the season to detect sardine spawning that year. One of the 1994 NMFS surveys found sardine spawning off Oregon, but it was not determined whether those fish were part of the stock available to California fishermen, and so the findings were not used in the California assessment. Another sardine survey (including standard icthyoplankton sampling, CUFES, trawl sampling for adults, and a test of airborne LIDAR for school detection) was conducted by NMFS during the spring of 1997.

Fishery sampling was also extended to the north in recent years, in response to increased abundance of sardine. CDFG commenced sampling of the Monterey directed fishery in 1996, and the size-at-age data were used in the most recent sardine stock assessment. Two other promising sources of fishery data became available during 1996. The Pacific States Marine Fisheries Commission began funding a multi-year study to collect and analyze sardine and chub mackerel data from Pacific northwest whiting bycatch during 1996; and the Canadian Department of Fisheries and Oceans collected samples from an experimental sardine fishery in British Columbia during 1996 (D. Ware, Pacific Biological Station, Nanaimo, B.C., pers. comm.). Geographic stratification in CANSAR is presently limited to separate selectivity parameters for Mexican and California fisheries. When sufficient data become available from the new northern sources (including the Monterey fishery), further geographic stratification will be investigated.

Uncertainties in the sardine assessment emphasize the importance of recent efforts to obtain new data as sardine range has extended to the north. As additional data are obtained from these and other sources, assessment precision will improve.

Historically, chub mackerel spawning grounds have been centered off the coast of northern and central Baja California with juvenile and adult fish moving northward into the Southern California Bight, where they are targeted by the California directed fishery.

CDFG conducted an icthyoplankton survey in the summer of 1996 to determine the extent and intensity of chub mackerel spawning in waters off Baja California. Icthyoplankton and biomass data from 1951-1966, when CalCOFI plankton surveys routinely extended into Baja California waters, suggest that larval density in Baja California waters is correlated with spawning biomass. Data from the 1996 research cruise will be used in the 1997 chub mackerel stock assessment.