On 6 November 1979, five consecutive otter 

 trawls of 10-min duration each were taken. The 



3.7 m otter trawl was equipped with a cod end of 



1.8 cm square mesh. Immediately following re- 

 trieval of the catch, sanddabs were sorted and 

 placed in sealed plastic bags. Any L. vulgaris 

 wandering about the catch were also retained. 

 After the last trawl, four scuba divers entered 

 the water and collected sanddabs by hand net, 

 attracting fish with bait (sea urchin roe). Fish 

 were transferred to sealed plastic bags. Speci- 

 mens were collected in this manner for approxi- 

 mately 40 min. All fish and isopods collected by 

 both methods were sexed and measured within 

 the next day. 



Host sex, total length to the nearest 0.1 cm, and 

 number of parasitic isopods harbored were de- 

 termined. Isopod total length was measured to 

 the nearest 0.1 mm. Isopod sex was determined 

 by several criteria: 1) Differential allometric re- 

 lationship of width to length (Montalenti 1941), 

 2) presence of penes in males, 3) asymmetry of 

 females, i.e., body twisted to the right or left 

 (Brusca 1978), and 4) presence of oostegites in 

 gravid females. Manca and aegathoid stages (see 

 Brusca 1978) were lumped as juveniles. 



15 



10 



OTTER TRAWL 

 • N = 56 



>- 

 O 1 



z 



LU 



o 15 



LU 



^^ 



rip 



D Uninfested 

 ^ Infested 



6 7 8 9 10 12 13 14 



SCUBA 



L N = 73 



6 7 8 9 10 11 12 13 14 

 FISH LENGTH (cm) 



Results and Discussion 



Otter trawl catches consisted only of flatfish, 

 and sanddabs comprised most of the catch. Size- 

 frequency histograms for sanddab hosts collected 

 by otter trawls and scuba divers were not signifi- 

 cantly different (Fig. 1, Kolmogorov-Smirnov 

 test, P>0.05). Inspection of these histograms in- 

 dicates that divers are able to sample small fish 

 more efficiently. Consequently, important infor- 

 mation regarding the acquisition of isopod para- 

 sites by young fish may not be obtained when 

 sampling with otter trawls. A comparison of the 

 percent of fish infested with L. vulgaris reveals a 

 highly significant disparity between the two 

 sampling methods. In the trawl, 21 out of 56 hosts 

 (37.5%) were infested versus 54 out of 73 hosts 

 (73.9%) in the diver sample (chi-square = 17.449, 

 P<0.005). 



Size-frequency histograms for all isopods re- 

 covered by both collecting methods were signifi- 

 cantly different (Fig. 2, Kolmogorov-Smirnov 

 test, P<0.05). If only male isopods were consid- 

 ered, the difference was very significant (P< 

 0.01) while for female isopods there was no dif- 

 ference (P>0.05). Apparently male L. vulgaris 

 abandon hosts in otter trawls prior to retrieval of 



Figure 1. — Comparison of size-frequency histograms for Ci- 

 tharickthys spp. collected by otter trawl and divers utilizing 

 scuba. Size distributions of sanddab hosts were similar (Kol- 

 mogorov-Smirnov test, P>0.05) but the percentage of hosts 

 harboring parasitic isopods was significantly different (chi- 

 square test, P<0.005). 



the catch. This was especially evident for small 

 males (Fisher exact test, P = 0.0057; small males 

 <10.6 mm vs. large males >10.6 mm). Large 

 males may also leave their hosts since a majority 

 of large male isopods were unassociated with 

 hosts in the trawl sample (13 out of 17 males, see 

 Figure 2). In one trawl sample a pod of male iso- 

 pods was found in the cod end indicating that 

 these individuals did not have sufficient oppor- 

 tunity to escape. Female isopods have feeble 

 crawling abilities (Brusca 1978) and do not ap- 

 pear to abandon hosts in trawls. 



Relative parasite density in the diver sample 

 was significantly higher (1.2 isopods per fish) 

 than otter trawl samples (0.4 isopods per fish) 

 (£-test, P<0.001) as was mean parasite intensity 

 with 1.6 and 1.1 isopods per infested host, respec- 

 tively (f-test, P<0.001). 



In conclusion, otter trawls consistently under- 

 estimate prevalence, relative parasite density, 

 and mean parasite intensity of L. vulgaris popu- 



908 



