SILLIMAN: EXPERIMENTAL EXPLOITATION OF FISH POPULATIONS 



.30- 



< 



1.25- 

 o 



</) 



£■20- 



tiJ 



£.15- 



Q_ 



< 

 cr 



5 05- 



-I 

 a 



X 



1 1 1 1 I 



.10 .20 .30 



EXPLOITATION RATE PER 3- WEEKS, GUPPY 



.40 



Figure 15.-Yield isopleths from data supplied by the simulation 

 model. Numbers by isopleths indicate guppy yield plus swordtail 

 yield per 3-wk period, in grams. 



these isonleths reveals a ridge of high yields run- 

 ning roughly from m^ = 0.16, Wg = 0.08 to wi, = 

 0.33, ^2 = 0.24. Thus, the optimal e.xploitation rate 

 for the swordtail was always lower than that for 

 the guppy, in agreement with the previously 

 mentioned lower productive capacity of the 

 swordtail. Also, moving from the high yield ridge 

 toward either of the axes is moving toward lower 

 yields. To the extent that they can be generalized, 

 these two findings suggest that where two 

 populations are competing, fishing both will 

 produce greater sustainable catches than fishing 

 either alone, provided that fishing rates are ad- 

 justed to the relative productivity of each species. 

 The conclusion provides some support for the idea 

 that excessive fishing of the sardine alone led to 

 the catastrophic decline in catches in the Califor- 

 nia sardine-anchovy situation mentioned in the 

 introduction. It is recognized that in many real 

 fisheries, species are fished jointly by the same 

 gear. In this case it is not possible to adjust the 

 fishing rates separately, and results will be 

 different (yields for a joint F will be less than for 

 separate F's). 



The maximum sustainable two-species yield, as 

 indicated in Figure 15, is 4 g per 3-wk period, with 

 nil = 0.24, mg = 0.16. It is of interest to see how 

 efficiently food was used at these exploitation 

 rates. Since juvenile fish were returned to the 

 adult tanks from the nursery tanks, food placed in 

 the latter must be included. During exploitation, 

 all of the food in the adult tanks was eaten but not 

 in the nurseries. Therefore the food consumed was 



between 1.0 and 1.5 times the adult amount, since 

 food amounts for the nurseries were one-half 

 those for the adult tanks. Three times the weekly 

 totals of Table 1 gives 17.10 g for adult tanks only 

 and 25.65 g for adult plus nursery tanks (weights 

 of Artemia nauplii were negligible in data to two 

 decimal places). The 4 g per 3-wk yield therefore 

 represents food conversion efficiencies of between 

 0.16 and 0.23 



The above efficiencies may be compared with 

 those from the independent populations. The Fox 

 (1970) models fitted as described under "Deter- 

 mination of Constants" provided estimated 

 maximum 3-wk sustainable yields of 3.7 g for the 

 guppy and 2.9 g for the swordtail (the latter con- 

 verted from 4-wk to 3-wk basis). Food amounts 

 were the same as for the competing populations 

 and the comparable conversion efficiencies were 

 0.14 to 0.22 for the guppy and 0.11 to 0.17 for the 

 swordtail. The range for the guppy is in reason- 

 able agreement with the 0.20 reported by Silliman 

 and Gutsell (1958) and 0.23 by Silliman (1968). That 

 the guppy range is higher than the swordtail 

 range is in keeping with other findings of superior 

 guppy productivity reported above. Both ranges 

 are below that for the competing populations. If 

 significant, this difference suggests a slight gain 

 in efficiency of the competing populations over 

 either species growing alone. 



Because the above conclusions have been 

 derived from a mathematical model developed 

 from the Volterra (1928) and Fox (1970) models, it 

 is of value to refer to the work of Larkin (1963). He 

 too used the Volterra equations as a point of 

 departure. He applied his analyses only to 

 hypothetical data, but his conclusions are 

 nevertheless in general agreement with those 

 given above. It is of interest to note his statement: 

 "It is concluded that this formulation of in- 

 terspecific competition together with variations 

 should be applied to laboratory or natural situa- 

 tions to test its usefulness as a basis for predic- 

 tion." 



ACKNOWLEDGMENTS 



George F. Slusser contributed part of the initial 

 stocks of guppies and maintained the experimen- 

 tal populations during the first 20 mo of the 

 experiment. Subsequent maintenance was per- 

 formed successively by Christopher E. Mathews, 

 Martin G. Beam, Jimmy R. Chrnaoski, and Judy A. 

 Trauth. All of these persons are present or past 



887 



