Abstract. -The growth patterns 

 of Pacific whiting Merluccius pro- 

 d/uctus, also known as Pacific hake, 

 were examined for the period 1978- 

 88 using fishery-based estimates of 

 length-at-age. Mean length-at-age 

 and a delta method variance esti- 

 mate of mean length-at-age were 

 computed for geographic and tem- 

 poral strata in the U.S. fishery. 

 These calculations took into account 

 the two-phase sampling design used 

 to sample the catch. A factorial anal- 

 ysis of variance of length found sig- 

 nificant differences due to age, year, 

 region, sex, and time-period during 

 the fishery. Length increases with 

 age and season. Pacific whiting 

 found in the north are larger, and 

 females are larger than males. The 

 mean length-at-age began declining 

 in 1978, and reached a minimum in 

 1984. From 1984 to 1986, there was 

 a slight rebound in length-at-age, but 

 after 1986 length-at-age again 

 declined. To investigate the influence 

 of population density and environ- 

 mental covariates on annual growth, 

 a generalized form of the von Ber- 

 talanffy growth model was devel- 

 oped. Deviations from a baseline 

 model for sex-specific asymptotic 

 growth were significantly correlated 

 with changes in sea-surface temper- 

 ature and adult biomass. Regression 

 results indicate that a 0.5°C increase 

 in mean summer sea-surface temper- 

 ature would reduce annual growth 

 by 24% at age 1 and 12% at age 4. 

 In contrast, the effect of adult bio- 

 mass on annual growth becomes 

 greater with age. An increase of 

 200,000 metric tons (approximately 

 10% of the mean population biomass) 

 would reduce annual growth by 5% 

 at age 4 and by 10% at age 7. It is 

 proposed that the effect of popula- 

 tion density is greater for the older 

 Pacific whiting because their diet 

 has shifted from euphausiids, whose 

 abundance is closely coupled with en- 

 vironmental processes, towards fish 

 species with multiyear life cycles that 

 can be affected by intense Pacific 

 whiting predation. 



Detecting environmental 

 covariates of Pacific wliiting 

 Merluccius productus growth! using 

 a growtli-increment regression model 



Martin W. Dorn 



Alaska Fisheries Science Center, National Marine Fisheries Service. NOAA 

 7600 Sand Point Way NE. BIN CI 5700, Seattle, Washington 981 15-0070 



Manuscript accepted 18 February 1992. 

 Fishery Bulletin, U.S. 90:260-275(1992). 



This paper describes research on 

 the growth of Pacific whiting, also 

 known as Pacific hake, a gadoid spe- 

 cies that is an important component 

 of the California Current ecosystem 

 (Francis 1983). The coastal popula- 

 tion of Pacific whiting is currently 

 the target of a major fishery with an 

 average (1977-88) annual harvest of 

 147,000 metric tons (t) (Dorn and 

 Methot 1990). Adult Pacific whiting 

 migrate north in spring and summer, 

 feeding in the productive waters 

 along the continental shelf and slope 

 from northern California to Van- 

 couver Island, British Columbia. In 

 late autumn, Pacific whiting migrate 

 south to spawning areas from Point 

 Conception, California, to Baja Cali- 

 fornia (Bailey et al. 1982). The U.S. 

 fishery operates from April to No- 

 vember and in recent years has been 

 conducted primarily under joint- 

 venture arrangements, with U.S. 

 fishing boats delivering fish to pro- 

 cessing vessels from the Soviet 

 Union, Poland, Japan, and other 

 nations. The Canadian fishery for 

 Pacific whiting is conducted in sim- 

 ilar fashion, except that independent 

 fishing by the foreign fleet still ac- 

 counts for a significant portion of the 

 catch. 



Hollowed et al. (1988) observed 

 that the mean length-at-age of Pacific 

 whiting had declined in recent years, 

 and hypothesized that the disruption 

 of normal circulation and tempera- 

 ture patterns associated with the 



1983 El Nino may have been the 

 causative factor. The recniitment of 

 strong 1980 and 1984 year-classes 

 increased the population biomass 

 of Pacific whiting to a maximum in 

 1986. The decline in the length-at- 

 age could also have been a density- 

 dependent growth response to this 

 increase in population abundance. 

 Since the Pacific whiting resource is 

 managed by setting an annual quota 

 in tons based on a conversion using 

 weight-at-age from a projected yield 

 in numbers (Dorn and Methot 1990), 

 changes in growth must be taken 

 into account when making manage- 

 ment recommendations about the 

 resource. 



The objective of this paper is to 

 examine the pattern of growth vari- 

 ability displayed by the coastal Pacif- 

 ic whiting population, and, in partic- 

 ular, to determine whether environ- 

 mental covariates or fluctuations in 

 population density could account for 

 the recent changes in length-at-age. 

 Analysis of variance, while useful as 

 an exploratory technique to identify 

 the sources of variability in length- 

 at-age, is inadequate to describe 

 changes in asymptotic growth. The 

 nonlinear regression model pre- 

 sented in this paper is a simple, bio- 

 logically realistic model for ex^ploring 

 the environmental determinants of 

 asymptotic growth. Its potential util- 

 ity is not limited to the application 

 described in this paper, i.e., growth 

 of Pacific whiting. 



260 



