A SIMPLE METHOD FOR ESTIMATING THE FOOD CONSUMPTION 



OF FISH POPULATIONS FROM GROWTH DATA AND 



FOOD CONVERSION EXPERIMENTS 1 



Daniel Pauly 2 



ABSTRACT 



Experimental data on the gross food conversion efficiency of fishes (K l = growth increment/food in- 

 gested) are usually reduced to a model of the form K x = aW b ; it is shown that the model K-, = 1 - 

 (WIW ao ) 1 ' has a number of advantages over the traditional model. 



The new model can be used to compute the food consumption per unit biomass of an age-structured 

 fish population, by relying on the first derivative of the von Bertalanffy growth formula (VBGF) to ex- 

 press growth increments, and the identity of W^, in the VBGF and in the model expressing K x as a func- 

 tion of weight. 



Computed examples, using published growth and mortality parameters, and the results of food con- 

 version experiments were used to obtain consumption estimates in a carnivorous grouper (Epinephelus 

 guttatus) and an herbivorous angelfish (Holacanthus bermudensis). Results were shown to be most sen- 

 sitive to the parameter /J. Various applications of this simple model are discussed, particularly as a method 

 to estimate key inputs in J. J. Polovina's ECOPATH model. 



A multiple-regression extension of the basic model is presented which accounts for the impact of 

 factors other than body weight on values of K l and /?. This method is illustrated with an analysis of data 

 on dab (Limanda limanda). 



Estimating the quantity of food eaten during a cer- 

 tain period by a fish population from field data is 

 usually a difficult task and various sophisticated 

 methods developed for this purpose have data re- 

 quirements which can make their routine applica- 

 tion impossible (Beverton and Holt 1957; Ursin 

 1967; Daan 1973, 1983; Andersen 1982; Armstrong 

 et al. 1983; Rice et al. 1983; Stewart et al. 1983; 

 Pennington 1984; Majkowski and Hearns 1984). 

 Polovina (1984) recently presented a technique 

 for construction of ecosystem models which is 

 structured around a well-documented computer 

 program called ECOPATH (Polovina and Ow 3 ). In 

 situations where classical fishery data are sparse this 

 technique has the potential of becoming a standard 

 method for consolidating and examining the data 

 available on aquatic ecosystems. ECOPATH esti- 

 mates equilibrium biomass (B), annual production 



'Based on Chapter 3 of a "Habilitationschrift" presented in 

 December 1984 to the Dean of the Mathematics and Science Facul- 

 ty, Kiel University (Federal Republic of Germany) and titled "Zur 

 Biologie tropischer Nutztiere: eine Bestandsaufnahme von 

 Konzepten und Methoden." ICLARM Contribution No. 281. 



international Center for Living Aquatic Resources Manage- 

 ment, MCC P.O. Box 1501, Makati, Metro Manila, Philippines. 



3 Polovina, J. J., and M. D. Ow. 1983. ECOPATH: a user's 

 manual and program listings. Southwest Fish. Cent. Admin. Rep. 

 H 82-83. Southwest Fisheries Center Honolulu Laboratory, Na- 

 tional Marine Fisheries Service, NOAA, 2570 Dole Street, Hono- 

 lulu, HI 96822-2396. 



(P), and annual consumption (Q) for each group in 

 the model. ECOPATH requires a number of data 

 inputs for each group treated in the model and usual- 

 ly the most difficult to obtain is the average food 

 consumption per unit biomass (Q/B) of each group. 

 The present study derives a method to estimate Q/B 

 through a combination of experimental and field 

 data that are easily obtained. In the process, a model 

 is derived which will allow for more information to 

 be extracted from feeding experiments than has 

 hitherto been the case. 



MODEL FOR REDUCING 



EXPERIMENTAL DATA ON 



THE CONVERSION EFFICIENCY 



OF FISHES 



Usually laboratory or pond feeding experiments 

 lead to estimates of K x , the gross conversion ef- 

 ficiency, which are obtained, for short intervals, 

 from 



K x = growth increment/food ingested (1) 



(Ivlev 1939, 1966). 



Usually, K x declines with body size (other factors 

 affecting K x are discussed below) and it has become 



Manuscript accepted April 1986. 



FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 



827 



