NOTE Kennelly: Discard rates for Stenotomus chrysops 



189 



discard did not exist in the region (i.e. 

 in addition to the identified depths in 

 areas 613 and 621). Figure 4 shows the 

 temporal pattern of scup discard in 

 those areas and depths identified from 

 Figures 2 and 3 and reveals a consis- 

 tent pattern of high discard for area 

 613 but not for area 621. However, 

 more observer coverage throughout the 

 year could have provided evidence of 

 other consistent peaks of scup discard 

 in one or more areas and depths, or 

 could have defined better the small 

 period of high scup discard in area 613. 

 For example, the data indicated that 

 November and December were key 

 months, but more sampling in Octo- 

 ber and January could have widened 

 this time frame. 



A second problem with the data is 

 that they were collected from many 

 different boats, with different nets, horse powers, tow 

 durations, etc. Although such problems are avoided 

 in fishery-independent surveys by using standard- 

 ized gears and sampling protocols, they are unavoid- 

 able when dealing with observer programs whose 

 objective is to survey normal fishing operations across 

 a variety of vessels, gear-types, etc. in order to de- 



tect fleet-wide patterns. Variation in observer data 

 is inherent in all such programs, and it is only by 

 doing properly designed and replicated, stratified, 

 randomized observer surveys that such problems can 

 be accounted for 



In most studies that have quantified bycatches, 

 species-specific spatial, and temporal variabilities in 



