TROPHIC RELATIONSHIPS IN JUVENILES OF THREE SPECIES OF 

 SPARID FISHES IN THE SOUTH AFRICAN MARINE LITTORAL 



M. S. Christensen* 



ABSTRACT 



The feeding habits of three sparids, Diplodus sargus, D. cervinus, and Sarpa salpa, were studied. 

 Juveniles of these fishes occur commonly in the intertidal and immediately subtidal regions of 

 southeast Africa, while adults were only observed in these zones at high tide. Small juvenile D. sargus 

 feed largely on harpacticoid copepods, amphipods, and, in spring and early summer, cirripede nauplii, 

 chironomid larvae, and an unidentified trochophore larva. Larger individuals mainly take amphipods 

 and green algae. Successive size classes of D. cervinus feed mainly on harpacticoid copepods and 

 chironomid larvae, the shrimp Palaemon pacificus, amphipods, and then polychaetes. Sarpa salpa 

 ingest harpacticoids when small, diatoms and red algae as a large juvenile, and red and green algae as 

 an adult. Corresponding changes in gut length and dentition are reported for S. salpa. 



Marked ecological separation of the three species was observed. Small juveniles appear at different 

 times of the year and feed on different foods (dietary and temporal separation). Larger juveniles and 

 subadults have different diets or feed in separate parts of the littoral zone (behavioral, dietary, and 

 spatial separation). 



A brief review of methods of analyzing stomach contents is included and it is suggested that a 

 combination of points and ranking indices would be the most valuable. The method, here termed the 

 comparative feeding index, is described. 



The food and feeding relationships of fishes in the 

 intertidal zone of South Africa are poorly 

 documented and the results are largely qualita- 

 tive. The most important of these studies deal with 

 one or two species of the families Gobiidae (Pitt- 

 Kennedy^), Sparidae (Hutchings^), Cheilodac- 

 tylidae (Butler^), and Gobiesocidae (Stobbs^). 



Three species of sparids were investigated in the 

 present survey, Sarpa salpa (Linnaeus 1758), Di- 

 plodus sargus (Linnaeus 1758), and D. cervinus 

 (Lowe 1838). Barnard (1927), Smith (1965), and 

 Hutchings (see footnote 3) described S. salpa 

 (strepie) as being primarily a herbivore, whereas 

 Talbot ( 1954) found the fish to be omnivorous. The 



'J. L. B. Smith Institute of Ichthyology, Rhodes University, 

 P.O. Box 94, Grahamstown, 6140, South Africa. 



*Pitt-Kennedy, S. 1968. A preliminary investigation of feeding 

 in two gobies Coryphopterous caffer (Giinther) and C. nudiceps 

 (C. and V.) with not«s on their sexual maturity. Unpubl. honors 

 proj., 39 p. Zool. Dep., Univ. Cape Town, S. Afr. 



^Hutchings, L. 1968. A preliminary investigation into the 

 diets of two littoral teleosts, Sarpa salpa (Linnaeus) and 

 Pachymetopon blochii (Valenciennes), with notes on their biol- 

 ogy. Unpubl. honors proj., 25 p. Zool. Dep., Univ. Cape Tovm, S. 

 Afr. 



■•Butler, G. S. 1975. An investigation into the biology of two 

 inter and infratidal species of Cheilodactylidae (Pisces: Teleo- 

 stei). Unpubl. honors proj., 29 p. Zool. Dep., Rhodes Univ., 

 Graham-stown, S. Afr. 



'Stobbs, R. E. In preparation. Preliminary investigations into 

 the feeding behaviour and food preferences of Chorisochismus 

 dentex (Pisces: Gobiesocidae). 



Manuscript accepted August 1977. 

 FISHERY BULLETIN: VOL. 76, NO 2, 1978. 



latter study was made in an estuary, however, 

 where algae are generally less abundant than in 

 the intertidal region. Diplodus sargus (blacktail) 

 is described as being an omnivore (Biden 1954; 

 Talbot 1954; Smith 1965), as isD. cervinus (zebra). 

 Little, however, has been published on the food 

 and feeding habits of the juveniles of these three 

 species, although they are abundant in the inter- 

 tidal and immediately subtidal regions of this 

 coast. 



The objectives of this study have, therefore, 

 been to determine: 1) the diet of juveniles of these 

 three species; 2) how feeding changes with age and 

 season; 3) the degree of overlap between species, 

 possibly resulting in competition; 4) recruitment 

 times and approximate growth rates of the fish; 

 and 5) the relationship between dentition, gross 

 gut morphology, and diet. 



During the course of this study, existing 

 methods of fish feeding analysis were found to be 

 inadequate and a new technique is described 

 which overcomes some of the problems. 



MATERIALS AND METHODS 



Fish were collected from February to December 

 1975 at 2-wk intervals during spring tide, in spite 

 of the possibility of introducing biases, as diving 



389 



