CHRISTENSEN; TROPHIC RELATIONSHIPS OF SPARID FISHES 



allotment of points to each food organism as de- 

 scribed above and the mean value per fish is then 

 multiplied by the percentage of the total sample of 

 fish that contain that item. As can be seen, the CFI 

 combines the properties of both points and RI 

 methods and thus reduces to some extent the effect 

 of the problems discussed. 



The diet of these fish was also determined by the 

 occurrence method (Hynes 1950) as this indicates 

 the feeding preferences rather than the food's vol- 

 umetric value. This is determined as the percent- 

 age of fish in the sample analysed which contain 

 that particular food item. 



The dietary composition of D. sargus was 

 analysed for the winter (February- July) and 

 summer (August-December) periods, as it was 

 found to be seasonal. This was not done for the 

 other two species, as feeding seasonality is 

 synonymous in these fish with the change in diet 

 with age, as they exhibit discontinuous recruit- 

 ment. 



All skeletal material was cleared and stained 

 using the trypsin maceration, alizarin stain 

 method of Taylor ( 1967). The gut was dissected out 

 in the same specimens, drawn and measured, and 

 the gut length to standard length ratio (G/S) was 

 calculated as in Weatherly (1972). 



STUDY AREA 



The study area is situated about 3.2 km north of 

 Kleinemonde in the eastern Cape and is known 

 locally as Clayton's Rocks (Figure 1). The 

 shoreline consists of a gently shelving, sandy 

 beach with broken rocky areas of varying extent. 



26 30 



2645 



27 00 



27 15 



Figure l. — The study area in the Eastern Cape Province, South 

 Africa. Adapted from Topographical Chart 3326, Grahamstown. 



The smaller rocky outcrops are continually cov- 

 ered and uncovered by sand, as the beach is un- 

 stable and backed by large, shifting sand dunes 

 which move at right angles along the coast. The 

 rocky area under study is made up of sandstone 

 which strikes east-west and dips steeply south- 

 wards. This has resulted in the development of 

 gullies and pools partially sheltered from wave 

 action by ridges of resistant rock (Figure 2). The 

 maximum collection depth at the seaward edge of 

 the gullies was 3 m at low tide. 



Environmental conditions vary greatly and 

 salinities may fall to 25%o at low tide, which is 

 caused by freshwater seepage into the pools from 

 springs in the beach. During the day, at low tide, 

 surface water temperatures have been recorded 

 ranging from 26° (summer) to 15°C (winter) in the 

 intertidal and from 22° (summer) to 14°C (winter) 

 in the open sea. 



The other major fish species coexisting in the 

 study area are listed below with their general 

 biological characteristics, where known. 



ARIIDAE 



T achy sums feliceps: occurs singly, as a juvenile, 

 in crevices. 

 SPARIDAE 



Lithognathus lithognathus: occurs in small 



groups of juveniles. 

 Rhabdosargus holubi: juveniles, in small 



groups. 

 Sparodon durbanensis: juveniles, observed 

 from October to March, either singly or in 

 small groups. 

 CHEILODACTYLIDAE 

 Chirodactylus brachydactylus: as juveniles and 

 subadults, singly, mainly from June to 

 November. 

 MUGILIDAE 



Unidentified species: occur all year round as 

 juveniles. 

 CLINIDAE 

 Clinus cottoides: a purely intertidal species, 

 lives in weed, juveniles appeared about 

 June/July. 

 Clinus superciliosus: lives in weed, juveniles 

 only observed in November/December. 

 GOBIIDAE 



Caffrogobius caffer: intertidal species, juveniles 

 seen from June to November. 

 TETRAODONTIDAE 

 Amblyrhyncotes honckenii: singly or in small 

 groups. 



391 



