Helser and Hayes: Quantitative management advice based on stock abundance indices 



291 



While over 400 species of finfish and invertebrates 

 have been caught in NEFSC's bottom trawl survey 

 during 1963-92, 62 species are caught consistently 

 nearly every year (Fig. 1). Roughly half of these 62 

 species have some economic importance and are 

 therefore assessed by various means (NEFSC, 1993; 

 Fig. 2). For the traditionally important groundfish 

 species, such as Atlantic cod, Gadus morhua, had- 

 dock, Melanogrammus aeglefinus, and yellowtail 

 flounder, adequate data are available to perform a 

 size or age-structured assessment (yield per recruit 

 or Virtual Population Analysis [VPAp. These types 



100 -, 

 90 - 

 80 -. 

 70 - 

 60 

 50 -| 

 40 

 30 

 20 

 10 

 



Spring survey 



■■Liiii.iiiii 



5 10 15 20 25 30 



Autumn survey 



5 10 



QLtaLian^QQciDQ 



15 20 25 30 



Number of years 



Figure 1 



Frequency of the number of years various species (includes 

 over 400 fish and invertebrates) occurring in the North- 

 east Fisheries Science Centers autumn bottom trawl sur- 

 vey since 1963 and spring bottom trawl survey since 1968. 

 Data included up to 1992. 



of assessments are usually accompanied by biologi- 

 cal reference points based on fishing mortality rates, 

 absolute stock abundance levels, or both, as indica- 

 tors for stock sizes below which long-term yield or 

 productivity may be jeopardized. While survey abun- 

 dance indices are important to "calibrate" results of 

 analytical models for these species and stocks, they 

 serve as the only source of abundance information 

 to assess the status of the majority of stocks in the 

 Northeast region (Fig. 2). Research survey index- 

 based assessments generally provide only qualita- 

 tive advice on the relative size of the stock and typi- 

 cally do not generate reference points commonly used 

 by fishery managers. 



In this paper, we extend the present procedures of 

 deriving fitted survey abundance indices to inferring 

 population status relative to an index-based refer- 

 ence point in a probabilistic framework through simu- 

 lation. We use Pennington's ( 1985 ) a priori integrated 

 moving average approach to derive fitted survey time 

 series and then characterize trends in population 

 abundance relative to an index-based reference point 

 defined to be the lower quartile (25 th percentile) of 

 the fitted time series. Our choice of the lower quartile 

 for a reference point was rather arbitrary. However, 

 the use of an interquartile ( such as the 25 th percentile) 

 computed from the data series over a range of years 

 with reasonably high (as well as low) population sizes 

 probably provides a reasonable reference point and 

 would serve as such even as the time series lengthens. 

 We then used a bootstrap procedure to characterize the 

 uncertainty in both the fitted index and the reference 



35 n 



a 20 



ST 15 



Yield per Age 

 None Index recruit structured 



Figure 2 



Frequency of assessment types performed on 62 species of 

 fish consistly caught each year in the Northeast Fisheries 

 Science Center's autumn and spring bottom trawl survey. 



