MORRIS: FEEDING HABITS OF BLACKSMITH 



An estimate of mean prey size was obtained by 

 dividing the total number of prey consumed into the 

 stomach weight for each fish collected. 



RESULTS 



Thirty species of fish were identified from the area 

 surrounding the discharge. Blacksmith were the 

 most abundant and frequently occurring fish (mean 

 estimate of abundance per transect >100 in- 

 dividuals, rank of the mean number per transect = 1 , 

 and frequency of occurrence per transect = 92.3). 

 Large schools arrived in the morning and began feed- 

 ing in the plume and outer plume. When feeding in 

 the outer plume, blacksmith would orient themselves 

 toward the plume, surrounding it, and feed on the 

 organisms that settled out of the rising effluent. 

 When in the plume, blacksmith were in constant 

 motion, being tossed about by the irregular flow, but 

 it was apparent from mouth action that these fish 

 were also feeding on suspended food items. 



The mean abundance per transect of blacksmith at 

 the Point for the quarterly sampling days in 1979 and 

 1980 was 148.4. They ranked first in mean number 

 per transect, with a mean frequency of occurrence of 

 86.2, and used the breakwater as their primary noc- 

 turnal sheltering site. 



There were no significant differences in either fish 

 length or fish weight, but there were significant dif- 

 ferences in stomach weight and stomach fullness be- 

 tween the two collection sites (Table 1). Fish 

 collected from the discharge had a greater amount of 

 food in their stomachs (an increase of 138%). 



Stomach fullness was not influenced by collection 

 time. The stomach weight and stomach fullness were 



TABLE 1. — Comparison of blacksmith, Chromis punctipinnis, collect- 

 ed from the discharge (thermal outfall) and the Point (Control Site), 

 King Harbor, Calif. 



'Note: The statistical package (SPSS) used was unable to compute P values lower 

 than 0.001 . Values below this number are represented as P < 0.001 . 



not significantly different between morning and 

 afternoon collections (£-test:£ = 1. 359, P= 0.181 and 

 t = 1.471, P = 0.147, respectively). Consequently, 

 the data collected from the discharge samples were 

 combined. 



The mean prey abundance, percent number, per- 

 cent volume, frequency of occurrence, and the 

 calculated IRI value of the 30 most abundant prey 

 items from each location are given in Table 2. A con- 

 tingency table analysis of the mean abundance 

 indicates thatthere was a significant difference in the 

 stomach contents between the two locations (G = 

 570.6, P <0.001, df = 17). The 10 most abundant 

 prey from each site (eliminating the smaller values) 

 are significantly different (G = 56 1 . 1 , P < 0.00 1 , df = 

 12). A comparison of the 10 highest IRI values from 

 each site are not significantly correlated (Kendall's 

 tau, t = 0.1868, P = 0.324, n = 14). A pictorial rep- 

 resentation of the IRI values is given in Figures 2 

 and 3. 



A comparison of the mean prey weight from each 

 sampling site revealed that blacksmith from the dis 

 charge ate larger prey than blacksmith from the Point 

 (discharge mean prey weight = 3.22 mg, SD = 4.01, 

 Point = 0.82 mg, SD = 0.81, t = 4.439, P <0.001). 



Temperatures from the discharge plume and base 

 were compared with surface and bottom tem- 

 peratures at the Point. The mean plume temperature 

 (26.3°C, SD = 3.3, n = 15) was significantly greater 

 (t-test: t = 5.69, P <0.001) than the mean surface 

 temperature from the Point (20.8°C, SD = 2.5, n = 

 30). Similarly, the mean base temperature (18.2°C, 

 SD= 2.4, n = 17) was significantly greater (t = 4.12, 

 P < 0.001) than the mean bottom temperature from 

 the Point (15.2° C, SD = 2.4, n = 30). 



The rank of the 10 most abundantly consumed prey 

 items was compared with the rank of the 10 most 

 abundant plankton items for both the discharge and 

 Point. There was no significant correlation for either 

 study site (discharge t = 0.01 10, P = 0.956, n = 14; 

 Point t= 0.2051, P= 0.329, n = 13). 



Between-site comparisons of the mean abundance 

 of six abundantly consumed prey items from both 

 stomach contents and plankton samples (Table 3) 

 show that two prey items, gammarids and 

 Polyophthalmus pictus, had a significantly higher 

 usage and availability at the discharge than the Point, 

 and that Calanus sp. and mysids had a higher usage 

 at the discharge but were not significantly more avail- 

 able. There was no significant difference in the usage 

 or availability of Oikopleura sp. between the Point 

 and discharge (although blacksmith from the Point 

 tended to eat a greater amount). 



The diets of blacksmith at the discharge and Point 



201 



