even though the whole intertidal zone, inckiding 

 tlie banks where Spartina grows, was available to 

 them. Pits dug by green crabs were not found in the 

 banks where some Mya lived in the sediments and 

 among the plant roots. Apparently the crabs could 

 not burrow easily into these compact sediments, 

 like the gravel and cobljle sediments described by 

 Dow and Wallace (1952). Instead, the gi'een crabs 

 dug numerous pits in the clam flats. Large crabs 

 frequently contained Mya and other animals living 

 in the sediments that required some digging to 

 obtain, but few ate Sparfina. 



Most small crabs live<l near the banks. Small 

 arthropods, plants, small Mytihis, and gastropods, 

 which were frequent in and beneath the vegetation 

 on the banks, provided small crabs with a con- 

 venient source of food. Finger-sized holes in the 

 banks, mats of algae, and a thick growth of 

 Spartina provided likely protection from preda- 

 tion. Several observations indicated that small 

 crabs inliabited the banks but not the subtidal 

 zone. In 1954, I trapped and dredged green crabs 

 in the subtidal zone and searched the banks for 

 them. All sizes were taken in the banks (fig. 6), 

 but the traps and dredge caught many large, but 

 few small crabs. These observations held true 

 for samples collected during 1955 and 1956. In 

 all 3 years, 47 percent of the 1,323 crabs caught 

 in the intertidal zone and only 1 percent of the 

 2,G56 crabs caught in the subtidal zone were small. 



The near absence of small green crabs in the 

 subtidal zone could have been misleading if the 

 mesh size of the dredge (51 mm.) had permitted 

 them to escape. When this same dredge was used on 

 September 6 and 8, 1955, in Rhode Island, how- 

 ever, 108 of the 123 crabs caught were less than 

 10 mm. wide. These small crabs were picked out of 

 clusters of recently set mussels and also were en- 

 tangled in algae and eelgrass. Although this ex- 

 ample was outstanding, many other dredge 

 samples off Rhode Island included some small 

 crabs and entangling debris. Entangling debris, 

 then, partially closes the dredge meshes and aids 

 in the capture of very small crabs. Debris, in the 

 form of old Mya shells and Spartina. nearly filled 

 each drag made at Plum Island River, but only 

 one small green crab was caught. Similar debris 

 was occasionally taken at Lufkins Flat, along 



with 27 small crabs. I conclude that small green 

 crabs were scarce in the subtidal zone at Plum 

 Island River but perhaps were occasionally more 

 abundant at Lufkins Flat than the dredge samples 

 indicated. 



THE DIURNAL FEEDING HABITS 

 OF GREEN CRABS 



Green crabs are rarely seen in the unflooded in- 

 tertidal zone during the daytime; most migi'ate to 

 and from the intertidal zone with the tide (Dexter, 

 1947; Edw-ards, 1958), and others hide in bank 

 caves and beneath sods during low water. Both 

 diurnal and tidal activity rhythms were reported 

 for green crabs by Naylor (1958) ; peaks of activ- 

 ity were during dai-kness and high tide. 



Activity of green crabs in relation to tide and 

 time of day probably affected the results of stom- 

 ach analyses. The effects of diurnal activities on 

 feeding habits were studied by examining stomachs 

 of crabs dredged during two successive low tides. 

 At most collection dates in 1956, one sample was 

 taken just after sunrise and another la.te in the 

 afternoon, before sunset. The crabs collected had 

 had the opportunity to feed in the flooded inter- 

 tidal zone during the high tide preceding capture, 

 during the night or the day, respectively. The to- 

 tal number of stomachs per sample was used to 

 obtain the percentage frequency of occurrence of 

 food items and the average numbers ingested per 

 crab. A preference to feed during a particular time 

 of day is emphasized by this method if either the 

 morning or afternoon samples contain a prepon- 

 derance of crabs ^\ith empty stomachs. 



The morning samples had more crabs than the 

 afternoon samples (table 5). A chi-square test 

 indicated significant differences from an expected 

 1:1 ratio (Plum Island River chi-square=47.42, 

 d.f.=5, P< 0.01; Lufkins Flat chi-square = 65.72, 

 d.f. = 6, P<0.01). The number of crabs containing 

 food and the average number of counted food 

 items per crab both were greater in the morning 

 than in the afternoon (fig. 7). Data on crabs with 

 and without food were tested with chi-square 

 (Plum Island River chi-square=6.79, d.f. = l, P< 

 0.01; Lufkins Flat chi-square=29.28, d.f.= l, P< 

 0.01). 



192 



U.S. FISH AND WILDLIFE SERVICE 



