600 



Fishery Bulletin 90(3), 1992 



.iL 



BEGULATEO MESH ABEA 

 SEASONAL REGULATED MESH ABEA 



Figure 1 



Georges Bank-Southern New England region of the North- 

 west Atlantic Ocean, showing geographic features and reg- 

 ulated mesh (5.5-inch) areas under the Northeast Multispecies 

 Fishery Management Plan of the New England Fishery 

 Management Council. 



terns is necessary. In addition to documenting geo- 

 graphic, seasonal, and age-specific aspects of distri- 

 bution, studies must examine mechanisms (such as 

 temperature, depth, spawning, or feeding behavior) 

 underlying these observed patterns. Numerous tag- 

 ging studies have been conducted for Atlantic cod in 

 the Northwest Atlantic: McKenzie (1934, 1956) and 

 McCracken (1956) described cod movements in Cana- 

 dian waters based on tagging experiments, and Smith 

 (1902), Schroeder (1928, 1930), and Wise (1958, 1962) 

 tagged cod from Woods Hole, Nantucket Shoals, and 

 Nova Scotia to New Jersey, respectively. More recent- 

 ly, cod distributions delineated by bottom-trawl survey 

 data from research vessels have been presented by 

 Scott (1988) for Canadian waters and Grosslein and 

 Azarovitz (1982) and Almeida et al. (1984) for U.S. 

 waters. None of these studies considered fish size in 

 the analyses. Overholtz (1984) found age-specific pat- 

 terns of distribution for another gadoid species, had- 

 dock, in the Georges Bank region. Wigley and Gabriel 

 (1991) and Bowman et al. (1987) examined distributions 

 of several juvenile fishes, including cod, using NEFSC 

 bottom-trawl survey data collected from Cape Hat- 

 teras, NC to Nova Scotia, Canada. 



In our study, NEFSC commercial landings data and 

 research-vessel survey data collected during 1982-86 



were analyzed to identify spatial and temporal patterns 

 as well as possible mechanisms associated with distribu- 

 tions of juvenile cod. The study period corresponds 

 roughly to the duration of the NEFMC's Interim Fish- 

 ery Management Plan for Atlantic Groundfish (31 

 March 1982 to 18 September 1986; NEFSC 1987). Dur- 

 ing this period, fishing practices were reasonably un- 

 changed, although (1) an increase in minimum mesh 

 size for the Georges Bank region from 5.125 inches 

 (130 mm) to 5.5 inches (140 mm) was implemented on 

 1 April 1983, and (2) the International Court of Justice 

 (ICJ) line dividing Georges Bank into U.S. and Cana- 

 dian portions (Fig. 1) was established in October 1984. 

 The study period also encompasses years of both strong 

 and weak recruitment as well as a 50% reduction in 

 spawning-stock biomass, events that allow evaluation 

 of resulting distributions over varying year-class 

 strengths and stock sizes. 



Materials and methods 



Distribution by temperature and depth 



Temperature and depth analyses were based upon 

 data collected for Atlantic cod during NEFSC spring 

 and autumn bottom-trawl surveys during 1982-86 in 

 the Georges Bank- Southern New England region in 

 depths of 9-366 m. The stratified-random survey design 

 and the standardization of survey gear and methodol- 

 ogy are described in detail by Grosslein (1969) and 

 Azarovitz (1981). Water-column temperature profiles, 

 including bottom temperature (recorded to 0.1 °C), 

 were obtained on approximately half the survey sta- 

 tions via expendable bathythermographs; depths (m) 

 were recorded using research-vessel electronic depth- 

 sounding equipment. All cod were measured (FL to 

 nearest cm) and a subset sampled for age, growth, and 

 maturity information. Otoliths were processed and age 

 determinations obtained according to procedures 

 described by Penttila (1988); maturity staging was 

 performed using classification criteria outlined by 

 Burnett et al. (1989). 



Estimates of mean temperature and depth (weighted 

 by number of fish in each tow), and associated stan- 

 dard errors and ranges, were calculated by age for each 

 season and tested for age-specific effects. Based upon 

 results from analyses of individual age-groups, ages 1 

 and 2 were combined as well as fish age 4 or greater. 

 These age-groups, as well as age-3 fish, were then re- 

 tested for age-group specific effects. Assumptions of 

 data normality were complicated by two factors. The 

 first is the inherent nature of survey catch data as 

 described by Pennington and Grosslein (1978), who 

 found that the two most likely models for the distribu- 

 tion of fish (i.e., heterogeneous Poisson and randomly- 



