winter flounder in comparable tows than Northeast II, designed for lobster pot hauling and modified for 

 trawling. 



Another factor influencing CPUE was the distribution of effort among stations in the river (Appendix 

 XXIV). Effort was most extensive in the lower river channel and adjacent shallows in 1976-77 and 

 occurred mainly in the channel during 1978-80. Through 1980, it was believed that heavy concentrations 

 of algae and detritus found in the upper river greatly reduced the number of winter flounder found there. 

 However, in 198 1 many winter flounder were found in the upper river and sampling effort has predominantly 

 shifted there duriing the past 6 yr. Detrital loads probably also have altered the efficiency of the trawl and 

 may have affected CPUE. Quantitative comparisons indicated significantly longer tow times in the upper 

 river in 1984 than in 1983 or 1985-86. This was most likely due to the net filling with material as it was 

 towed, thereby increasing drag. Decreases in average tow time have been due to lesser amounts of detritus 

 as well as the purposeful avoidance of the worst towing areas in station 51. The practice of concentrating 

 effort where most winter flounder were present prior to 1983 also may have had some consequence; 

 increased medians may have resulted from limiting the number of tows with fewer fish. 



Although each measure of abundance has inherent errors, the reasons for the disparities between the 

 .loUy index and CPUE are unknown. The standardization of tows to a uniform distance by station in 

 1983 and subsequent years lessened the variability in CPUE as seen by the smaller confidence interval 

 about each median and greater correspondence between the median and mean. The Jolly indices for 

 1984-86 were less precise with relatively large confidence intervals, which could have accounted for some 

 of the differences observed in the trends. Because of the small number of fish marked in 1986, the Jolly 

 estimate for that year may have been inaccurate. Although there were indications that the population 

 probably declined from 1985, the decrease would have been proportionately less if the CPUE was more 

 reliable than the Jolly composite index. 



Some difi'erences between CPUE and the Jolly composite index could be related to changes in winter 

 flounder distribution over time. Possibly, most winter flounder were present only in the lower river during 

 the 1970s and despite low to moderate population levels, high CPUE were obtained because of their 

 concentration in the relatively smaller portion of the river that was sampled. Then, in 1980, for unknown 

 reasons, winter flounder began to use the upper river as well. Although more abundant, they would have 

 been less concentrated throughout the river and lower CPUE were obtained. Alternatively, winter flounder 



34 



