contents of stomachs and total number of each prey 

 taxa were counted and weighed (mg wet weight). 

 Food Items were Identified to major taxonomlc 

 groupings, and occasionally to genus or species. An 

 Index of Relative Importance (Pinkas et al. 1971), 

 which factors percentage frequency, weight, and 

 numbers, was computed for each food item in 

 stomachs. 



Algal associations were observed during 43 sam- 

 pling trips In the seasonal period of lumpflsh pres- 

 ence In tldepools, 1979-86, principally in the three 

 study pools. Trips ranged from a low of 2 in Decem- 

 ber to 10 in June and July during these years. 

 Samples of fish for food analyses were collected 

 from July to November (principally July and August) 

 in 1981, 1984, and 1985. Food habits data were 

 pooled because sample sizes were insufficient for 

 monthly or yearly comparisons. Sampling trips 

 averaged 80 minutes and were made exclusively dur- 

 ing daylight ebb tides. 



Results 



Food 



Amphlpods were the principal Item in the diet of 

 150 juvenile C. lumpus examined that ranged In 

 length from 9 to 50 mm; peak length frequency was 

 15 mm. Copepods, isopods, cumaceans, and marine 

 mites (Acarina) were also numerically important 

 (Table 1). Two lumpflsh had consumed larval fishes: 

 one a smaller lumpflsh and the other an unidentified 

 species. Seven stomachs (4.7%) were empty. Am- 

 phipoda were also most Important in weight. The 



Index of Relative Importance (IRI) indicates Am- 

 phlpoda (IRI = 6,732) and Copepoda (IRI = 2,650) 

 and, to a lesser extent, Isopoda (IRI = 798) were 

 the most important Items in the diet. Other foods 

 were of only minor Importance. 



There was no significant difference In the diets 

 of lumpfish between locations along Schoodic Penin- 

 sula (Table 1; x^ = 16.93, 0.10 > P < 0.05), except 

 that Polychaeta were consumed in higher numbers 

 at Blueberry Pool. More Cumacea and Copepoda 

 were consumed by fish in West Side Pool, but this 

 trend at West Side Pool may be a reflection of fish 

 size and the presence of Z. marina— excellent 

 mlcrohabltat for Copepoda and Cumacea. These 

 organisms were significantly more Important in the 

 diet of fish <15 mm (x" = 32.0; P < 0.05), while 

 Amphlpoda, Isopoda, and Polychaeta were more 

 significant (x^ = 51.0; P < 0.01) In the diet of juve- 

 niles >15 mm TL than in the diet of smaller fish. 



Algal Associations 



Juvenile lumpfish observed were from 6 to 50 mm 

 long and included primarily fish of age 0, but also 

 age 1, as judged by length-frequency graphs and 

 other work of Cox and Anderson (1922) and Dabom 

 and Gregory (1983). One juvenile, 80 mm TL (ap- 

 parently from an older year class), was also col- 

 lected. From 328 observations of algal attachments 

 of lumpflsh during daylight hours, definite patterns 

 emerged. Zoster a marina and 18 species of marine 

 algae were identified from the three pools during 

 a survey in July 1986. Juvenile lumpflsh were also 

 found attached to an additional species, Rhodymenia 



Table 1.— Percent occurrence and percent of total weight of food items in tfie diet 

 of juvenile Cycloplerus lumpus, and foods by pool location and size (percent occur- 

 rence). West Pond data were not sufficient for inclusion in site comparisons, but are 

 used in length comparisons; n = 1 50 for all three pools. Seven stomachs were empty 

 (4.7%) and are not included in food data. Fish size range was 9-50 mm, with a length 

 frequency peak of 15 mm. 



'Other items, in order of decreasing percentage occurrence; Cladocera. fish eggs. Diptera, Littorina 

 spp (Gastropoda), Mytilus edulis, algae, fish larvae, Caprellidea, dinoflagellates. cypris larval stages, 

 megalops larval stages. 



234 



