182 



Fishery Bulletin 105(2) 



The primary objective of the present study was to in- 

 vestigate the possibility of using parasite markers to de- 

 lineate the western and eastern 4T plaice stocks, evident 

 in earlier mark-recapture experiments (Powles, 1965). 

 In light of potential mixing of southern Gulf and Cape 

 Breton Shelf plaice over-wintering in the Laurentian 

 Channel, parasite tags were also employed in an effort 

 to examine the discreteness of 4T and 4Vn stocks. 



Materials and methods 



Sampling of the host 



American plaice, 31 to 40 cm in total length, were col- 

 lected from two locations in the southern Gulf of St. 

 Lawrence (NAFO division 4T) and from three locations 

 on or near the Cape Breton Shelf (NAFO subdivision 

 4Vn) (Fig. 1). Samples from the southern Gulf were 

 collected from the Canadian Coast Guard Ship (CCGS) 

 Alfred Needier and the CCGS Teleost during a Fisher- 

 ies and Oceans Canada (DFO) demersal fish survey in 

 September of 2004, and plaice from the Cape Breton 

 Shelf were sampled from the Alfred Needier in May 2003 

 during a dedicated survey of parasites and diseases in 

 4Vn groundfish. Fish were caught with a Western IIA 

 otter trawl which was towed for 30 minutes at each sta- 

 tion. A total of 483 plaice were collected, including 137 

 plaice taken at depths of 43 to 88 m in the western 4T 

 division and 95 plaice sampled at depths of 49 to 70 m in 

 eastern 4T division. In the 4Vn survey, 128 plaice were 

 sampled from the eastern slope of the Smokey Channel, 

 46 along the eastern edge of the Cape Breton Shelf, and 

 77 in the Louisbourg Hole. "Channel," "Edge," and "Hole" 

 samples were taken at depths ranging from 77 to 124 

 m, 72 to 158 m, and 77 to 135 m, respectively. Samples 

 were frozen onboard in a walk-in freezer at -17°C and 

 transferred on landing to a -20° C walk-in freezer at the 

 Gulf Fisheries Centre (GFC), Moncton, New Brunswick, 

 where they were stored for future examination. 



Individual fish were thawed at room temperature 

 and measured lengthwise to the nearest centimeter. 

 External surfaces and gills were inspected by eye for 

 signs of trauma and other disease conditions. Viscera 

 were examined for endoparasitic helminths with a dis- 

 secting microscope, and fillets were removed and sliced 

 into thin sections (McClelland and Martell, 2001a) for 

 detection of larval digeneans and nematodes. All hel- 

 minth parasites were counted with the exception of 

 Stephanostomum baccatum metecercariae. Because the 

 latter were often too numerous and too widely distrib- 

 uted in host tissues to be counted within a reasonable 

 time frame, only their presence was noted. 



Statistical analysis 



Prevalence (?) and abundance (A) of individual parasite 

 species were calculated according to the methods of 

 Bush et al. (1997), with prevalence being the numbers 

 of infected fish in a sample divided by the total number 



of fish in a sample, expressed as a percentage, and 

 abundance being the total number of parasites recov- 

 ered from a sample divided by the total number of fish 

 in a sample. Individual fish were coded 1 (infected), or 

 (uninfected) (Li, 1964; Neter et al., 1985) for analyses 

 of prevalence, and a log in+l) transformation was used 

 to bring the distributions of parasite counts closer to 

 normality (Sokal and Rohlf, 1969; Piatt, 1975). Parasite 

 infection parameters that best distinguished between 

 sampling locations were selected by forward stepwise 

 discriminant function analysis (DFA). The Kappa sta- 

 tistic iK) was used to determine the improvement over 

 chance of DFA (Titus et al., 1984). Misclassification rates 

 of DFAs were calculated by employing cross validation 

 procedures described by Arthur and Albert (1993) and 

 Boje et al. (1997). All statistical procedures were per- 

 formed with Systat for windows, vers. 7.0 (SPSS Inc., 

 Chicago, IL). Because the survey was confined to plaice 

 in a narrow length range (31-40 cm TL) (McClelland 

 et al., 2000), effects of host size on parasite infection 

 parameters were not investigated. Plaice from the edge 

 of the Cape Breton Shelf were not included in the DFA 

 analysis because of the small sample size (« = 46) and 

 the fact that they were collected from two widely sepa- 

 rated groups of trawl sets (Fig. 1); the northern (n=21) 

 and southern {n=25) portions of the sample were more 

 distant from each other than were the samples from 

 Smokey Channel and Louisbourg Hole. 



Results 



Plaice sampled from the southern Gulf of St. Lawrence 

 (NAFO Division 4T) and the Cape Breton Shelf and 

 vicinity (NAFO subdivision 4Vn) ranged from 31 to 

 40 cm in total length (TL), but the great majority fell 

 within the lower half of this range. Mean length of 

 fish (±standard error) was 32.34 ±0.180 cm (n=137} for 

 plaice from the western 4T sample, and 32.31 ±0.22 cm 

 (/! = 95) for the eastern 4T sample. Mean lengths of 4Vn 

 plaice were 33.17 ±0.26 cm {n=128) and 32.88 ±0.23 cm 

 (n=n) for Smokey Channel and Louisbourg Hole fish, 

 respectively. 



Eleven species of endoparasitic helminths were identi- 

 fied during routine parasitological examinations of 483 

 plaice (Table 1), but the larval digeneans Otodistomum 

 sp., plerocercoid larvae of an unknown cestode species, 

 and the nematode Hysterothylaciuin aduncum were ob- 

 served too infrequently (<A'7c prevalence in all samples) 

 to be considered useful as biological tags (see MacKenzie 

 and Abaunza, 1998). Forward stepwise DFA of infection 

 parameters of the remaining species revealed that abun- 

 dances of the acanthocephalans Corynosoma strumosum 

 and Echinorhynchus gadi contributed significantly to 

 the classification of plaice from the southwestern and 

 southeastern Gulf of St. Lawrence (/if=0.57) and that 

 E. gadi abundance was the more significant variable 

 (Table 2). Cross-validation yielded an overall rate of 

 79% correct classification (Table 3), and the lowest rate 

 of misclassification (15%) was found in fish from the 



