FISHERY BULLETIN": VOL. 70, NO. 4 



adult Diplodus which were analyzed. Fish scales 

 were a common constituent of stomachs of fish 

 in size classes between 21 and 70 mm in length. 

 The contribution of fish scales to the stomach 

 contents was at a maximum in 41 to 45 mm 

 fish (5%). The majority of the fish scales pre- 

 sent in stomachs are construed to have come 

 from cleaning activities since fish per se were 

 seldom encountered in the stomachs of any of 

 the Diplodus examined. 



As shown in Figure 2, the bulk of the diet of 

 juvenile D. holbrooki consists of free-living an- 

 imals and epiphytic algae. Cleaning activities 

 provide only a modest portion of the total food 

 ingested by fish in any of the size classes. In- 

 dividuals in size classes between 11 and 25 mm 

 are primarily planktivorous with veligers, cope- 

 pods, tunicate tadpole larvae, mysids, and small 

 shrimp accounting for 56 to 87% of the total 

 stomach contents. Fish over 25 mm become very 

 dependent on plant material in the form of 

 epiphytic algae. This plant material accounts 

 for 53 to 87% of the total stomach contents of 

 all size classes above 26 mm in length, including 

 adults. However, variable amounts of animal 

 material, especially sponge, copepods, shrimp, 

 and mysids, are also consumed. Stomach con- 

 tents of fish greater than 25 mm in length show 

 that Diplodus feeds heavily on plant and animal 

 material which grows attached to other subjects. 

 Even the crabs consumed by larger individuals 

 are primarily porcellanids (Petrolisthes sp.) 

 that live in association with submerged rock and 

 shell. 



Since there are no prior reports of D. hol- 

 brooki functioning as a cleaner, additional ju- 

 venile specimens were obtained from Dr. Carter 

 R. Gilbert. These specimens were collected in 

 1951 and 1957 near Cedar Key, Fla. Unfortu- 

 nately, the museum collections were short on 

 specimens in the 21 to 60 mm size range and 

 those that were provided did not have well-pre- 

 served stomach contents. All that can be stated 

 with certainty from these specimens is that ecto- 

 parasites did appear in stomachs of 5 of the 13 

 fish available in the size range of 26-35 mm. 

 Representatives of other juvenile size classes 

 were either not available or inadequately pre- 

 served for analysis. 



EFFICACY OF 

 SIEVE FRACTIONATION PROCEDURE 



The sieve fractionation procedure described 

 in Materials and Methods has contributed 

 greatly to our capacity to analyze quantitatively 

 the stomach contents of very small fish. This 

 procedure provides a convenient means of sep- 

 arating a heterogeneous mixture of minute food 

 items into fractions which are individually quite 

 homogeneous with respect to particle size. Table 

 2 presents the format used and the results ob- 

 tained following a sieve fractionation of stomach 

 contents of D. holbrooki belonging to the 11 to 

 15 mm size class. This table and the following- 

 discussion are provided to illustrate the efficacy 

 of this procedure. After identification of the 

 food items in each sieve fraction (Column A), 

 the determination of the portion of each fraction 

 attributable to a particular food item (Column 

 B) is eased considerably by the fact that all items 

 in the fraction are of comparable size. After 

 determining the dry weight of each fraction 

 (Column C), the contribution of individual food 

 items (Column D) is estimated by multiplying 

 the weight of the fraction by the value defining 

 the portion of the fraction attributable to that 

 food item. Finally, the total amounts and per- 

 centages of the entire stomach contents which 

 are attributable to the individual food items are 

 determined by simple compilation (Columns E 

 and F) . This procedure has now been employed 

 in this laboratory to analyze the stomach con- 

 tents of juvenile stages of 18 species of fishes. 



DISCUSSION 



Randall (1967) has described the Carangidae 

 as a family of carnivorous fish whose food habits 

 may be divided into three major categories: fish 

 feeders, plankton feeders, and mollusk feeders. 

 His classification was based primarily on feeding 

 habits of adult fish, and it was recognized that 

 not all of the species, including their juvenile 

 stages, fit perfectly into these categories. Ran- 

 dall analyzed stomach contents of Oligoplites 

 saurus (149 to 234 mm) from the West Indies 

 and found this carangid to be primarily a fish 

 feeder. Similar findings have been reported by 



1116 



