GUSHING: LINK BETWEEN SCIENCE AND MANAGEMENT 



may even rise if recruitment rises with reduced 

 fishing. 



It is Hkely that if fishing mortality is restrained 

 to the maximum of the yield per recruit curves of 

 most demersal fishes, recruitment would not be 

 very much impaired. However, the curve for the 

 herring is asymptotic and it was tacitly believed 

 that the herring could be fished very hard merely 

 because of the shape of the yield per recruit curve. 

 Another consequence was that the maximum for 

 the cod in the Barents Sea was overshot because in 

 international terms a high catch was needed at 

 the cost of a decreased catch per effort. The re- 

 cruitment to the Barents Sea stock was severely 

 reduced. 



The collapse of herring fisheries throughout the 

 northeast Atlantic and the failure of recruitment 

 to the Arcto-Norwegian cod stock have been attri- 

 buted to failure in the Commissions. These Com- 

 missions, however, are only as good as their 

 scientific advice, and the two great failures are 

 attributable to the unstated concept that fishing 

 could continue until recruitment was seen to fail. 

 Then failure had to be attributed either to natural 

 causes or to fishing, and stocks collapsed while the 

 scientists disputed the two possibilities. If re- 

 cruitment overfishing had been recognized as a 

 problem, perhaps collapse might have been 

 avoided. The solution to the problem of growth 

 overfishing inadvertently generated the problem 

 of recruitment overfishing. 



STOCK AND RECRUITMENT 



The dependence of recruitment on parent stock 

 was formulated by Ricker (1954, 1958), but the 

 variability of recruitment is very high and the 

 curve can only be fitted when decades of annual 

 data have been collected. Any decision on how 

 much fishing should be allowed is perhaps delayed 

 for statistical reasons and management cannot 

 proceed. On the other hand, if recruitment were 

 considered to be independent of parent stock, 

 management could start to take decisions more 

 quickly. With hindsight, the assumption made by 

 Thompson and Bell ( 1934) and Graham ( 1935) can 

 be justified because their work formed the 

 scientific basis of all international conservation in 

 the North Atlantic and North Pacific. However, 

 the danger of such a procedure is that any decline 

 in recruitment tends to be attributed to natural 

 causes rather than to fishing, and this step has 

 sometimes been taken without evidence. 



It is in the nature of the stock and recruitment 

 problem that there should be confusion about at- 

 tributing decline in recruitment to natural or to 

 man-made causes. However, Gushing and Harris 

 ( 1973) have devised a method of fitting the Ricker 

 curve which sets confidence limits to the position 

 of the curve itself; the standard deviation of re- 

 siduals sets limits to the variation of recruitment. 

 Then if stock is near that value which generates 

 maximum recruitment per unit stocK, any re- 

 cruitment that falls below the standard deviation 

 of the residuals has failed through natural causes. 

 If stock is low, such a distinction cannot be made 

 (because any recruitment below the standard 

 deviation of the residuals is zero) and failure is 

 attributed to natural or to man-made causes with- 

 out evidence either way; however, a prudent 

 manager might prefer to assume that recruitment 

 declined under the pressure of fishing and to take 

 appropriate action in the hope that the stock 

 would recover, as recently happened with the 

 British Columbia herring stock. 



The stocks that have failed because of this di- 

 lemma are the California sardine, the Japanese 

 sardine, and the northeast Atlantic herring; the 

 Arcto-Norwegian cod stock might fail for the same 

 reason. The failures of the first three were attrib- 

 uted to environmental factors on evidence that is 

 only circumstantial; more recent evidence sug- 

 gests that failure might have been due to fishing. 

 To pursue the argument further is sterile. The 

 scientific failure was the inability to make clear 

 the distinction between natural and man-made 

 causes. The failure in management was to delay 

 action until the distinction could be made, 

 whereas a prudent manager should have feared 

 the effect of fishing upon recruitment. 



In the stock and recruitment problem, long data 

 series are needed before any management deci- 

 sion can be taken. When the development of a 

 fishery proceeded slowly, this may not have mat- 

 tered because the maximum yield was attained by 

 gentle increments. Today, however, stocks are ex- 

 ploited rapidly, and there is the possibility that 

 recruitment will be diminished before the data are 

 available to describe the maximum sustained 

 yield. What is needed is an analytic model of the 

 stock and recruitment relationship on the lines of 

 that of the yield per recruit one, on the basis of 

 which decisions on management can be made 

 quickly without the laborious acquisition of long 

 series of data. 



861 



