Surveys of larval sealworm (Pseudoterranova decipiens) infection in various fish species sam- pled from Nova Scotian waters between 1988 and 1996, with an assessment of examination procedures

Between November 1988 and October 1996, >10,000 fish from the Breton Shelf, Sable Island Bank and the northeastern Gulf of Maine were examined for larval anisakines. Larval sealworm, Pseudoterranova decipiens, occurred in 30 of 39 species surveyed, including 8 new host records, Enchelyopus cimbrius , Lycodes reticulatus , Eumesogrammus praecisus , Lumpenus lumpretaeformis, Lumpenus maculatus , Cryptacanthodes maculatus , Artediellus atlanticusand Triglops murrayi. The parasite was most prevalent and abundant in mature demersal piscivores and benthic consumers. Sealworm densities (nr kg -1 host wt.), however, were greatest in small benthophagous fish including mature E. cimbrius, A. atlanticus, T. murrayiand Aspidophoroides monopterygius , and juvenile Hippoglossoides platessoides . ANOVA revealed that geographical disparities in sealworm prevalence and abundance were highly significant in 14 of 20 species tested, although significant disparities between samples from each of the three areas were evident only in H. platessoides . Almost invariably, infection parameters were greatest in fish from Sable Island Bank. ANOVA also indicated that sealworm prevalence and/or abundance increased significantly in Sable Island Bank populations of Gadus morhua , H. platessoides , and seven other species between 1985-1986 and 1989-1990. Routine examinations, in which host flesh was sliced and candled, proved as efficacious as digestion in warm (35° C) pepsin-HCl for detection of larval sealworm in the flesh of large frozen fish. Procedures employing fresh (iced) samples, digestion at ambient temperature and microscopy are recommended, however, for surveys of small benthic consumers. Many of the sealworm infecting the latter hosts are tiny (2 to 10 mm in length) nematodes, which escape detection by routine inspection, and may not survive in warm pepsin-HCl solution. McClelland, G. and Martell, D.J. 2001. Surveys of larval sealworm ( Pseudoterranova decipiens ) infection in various fish species sampled from Nova Scotian waters between 1988 and 1996, with an assessment of examination procedures. NAMMCO Sci. Publ. 3: 57-76. 57 NAMMCO Scientific Publications, Volume 3 larger prey are often discarded by seals, it is possible that exploitation of heavily infected mature piscivores is underestimated in surveys which rely on otoliths for identifying and ageing seal prey. The “seal prey” time series was designed to provide data for a predictive sealworm model proposed at a two-part sealworm workshop held in Halifax Nova Scotia in April 1987 and June 1988 (Bowen 1990). Rather than monitoring larval sealworm infections in a single indicator host, as is the case in the “sealworm index” time series (McClelland and Martell 2001, McClelland et al.2000), the parasite was monitored in numerous fish species, at a few specified locations. As a consequence of dwindling groundfish resources and changing research priorities, the project was abandoned before completion. Some of the data collected in the Gulf of St. Lawrence have been reported (Boily and Marcogliese 1995, Marcogliese 1995), but much of it remains unpublished. “Seal prey” surveys of fish from the ScotiaFundy region proved useful in revealing small benthic consumers of potential importance in sealworm transmission. The results of surveys of the Breton Shelf, Sable Island Bank, and the northeastern Gulf of Maine are documented here, with analyses of spatial variations in sealworm prevalence and abundance in 1989-1990, and changes in infection parameters in fish from Sable Island Bank since 1986-1987 (McClelland et al. 1990). The efficiencies of procedures used for detecting sealworm larvae, namely candling, dissection and digestion, are also assessed. MATERIALS AND METHODS Marine fish samples were collected primarily from Department of Fisheries and Oceans (DFO) research vessels, Alfred E. Needler and Lady Hammond during cruises dedicated to surveys of larval P. decipiens, and other parasites of marine fish and benthic macrofauna. Samples for an inventory of larval anisakines in seal prey (the “seal prey” survey) were taken from the southeastern Breton Shelf, Sable Island Bank and the northeastern Gulf of Maine between November 1988 and August 1993, but primarily from February 1989 to October 1990 (Fig. 2). Fish were caught in sets of 30 to60 min. duration, with a Western IIA otter trawl towed at 6.5 km h , and stored in the freezer. On return to the DFO Halifax Fisheries Research Laboratory, samples were allowed to thaw for 24 h at room temperature (15 to 20 o C) prior to inspection. After total length, weight and sex were recorded, visceral organs and mesenteries were inspected for larval anisakines. Each fish was then boned, and fillets, napes (hypaxial musculature enclosing the body cavity) and flesh which remained on the frame were examined for nematodes by slicing and mechanical destruction of tissues (Power 1961). As a rule, nematodes were identified by eye and counted, but inspection with a dissecting microscope was sometimes necessary in order to identify smaller worms and specimens damaged (e.g. cut into fragments) during boning and slicing of flesh. P. decipienssibling species B, recently designated as P. decipiens (sensu stricto ) (Paggi et al. 2000) is the only species of sealworm known to occur in the waters surveyed here (Brattey and Davidson 1996). Efficiencies of routine examinations of the musculature of groundfish were tested using procedures described by McClelland et al. (2000). Atlantic cod ( Gadus morhua ) (n = 90), sea raven ( Hemitripterus americanus ) (n = 23), long horned sculpin ( Myoxocephalus octodecemspinosus ) (n = 30) and Canadian plaice (n = 163) were collected from the Scotian Shelf and southwestern Nova Scotia during groundfish sampling cruises in June and July 1991. The flesh of individual fish, together with nematodes identified and counted following routine inspection, were placed in a 4 L beaker containing 2 L of 1% HCl with 5 g of 1:10,000 pepsin L. After incubation for 2 to 3 h at 35 to 40° C with continuous stirring, the contents of each beaker were strained through a series of sieves ranging from 5.0 to 0.3 mm in mesh size. All nematodes recovered, including those severed during boning and slicing of the flesh, were identified and counted as above. To determine the frequencies of occurrence of smaller sealworm larvae (2 to10 mm in length) in demersal fish, samples of various benthic 59 NAMMCO Scientific Publications, Volume 3 INTRODUCTION As they lack direct economic significance, small benthophagous fish, which include the juveniles of commercially important demersal species as well as underutilised species, are often overlooked as potential reservoirs of Pseudoterranova decipiens larvae. Recent field and laboratory studies have shown, however, that they may be essential in the transmission of sealworm to larger, commercially exploited fish, if not directly to definitive (seal) hosts (McClelland 1995). In an earlier survey of Sable Island Bank on the central Scotian Shelf (McClelland et al. 1990), larval sealworm were found in 26 of 32 marine fish species. While it was most prevalent and abundant in large demersal piscivores, P. decipiens often occurred in greatest density (nr/unit host weight) in small (juvenile and mature) benthic consumers (Fig. 1). Laboratory experiments (McClelland 1995), in which P. decipiens larvae were transmitted to fish via benthic copepod and amphipod intermediaries, revealed that sealworm grew to a length of 8 mm in crustacean hosts, but larvae as small as 2 mm in length were infective to small fish. One of the conclusions drawn from this study was that the primary fish hosts of sealworm are probably small benthic consumers. It was also apparent that sealworm larvae recently transmitted to these hosts would be considerably smaller than those detected by candling procedures typically employed in sealworm surveys. Hence, in order to obtain accurate worm counts when surveying infections in smaller demersal fish, more refined examination procedures involving microscopic inspection of host tissue squashes or digested host tissues might be required. Given that fishes of diverse phylogeny are qually susceptible to sealworm infection in the laboratory, light infection or absence of infection in natural populations of certain host species is probably attributable to ecological, behavioural and physiological (e.g. host response) barriers to the transmission of the parasite (McClelland 1995). Surveys of the diets and ascaridoid infections of flatfishes inhabiting Sable Island Bank (Martell and McClelland 1994, 1995), for example, indicated that dispari ies in parameters of sealworm infection among sympatric flatfish species was largely related to the exploitation of different prey. Juvenile Canadian plaice ( Hippoglossoides platessoides ), which were heavily infected with sealworm, fed on benthic suprafauna, i.e. free swimming organisms (amphipods, mysids etc.) closely associated with bottom . Winter flounder (Pleuronectes americanus ), which was rarely infected with larval sealworm, consumed more sedentary infauna and attached epifauna. As suggested by analyses of their diets, seals may incur P. decipiensinfections primarily through consumption of smaller fish. Larger demersal fish, while often heavily infected, are seldom exploited by grey ( Halichoerus grypus ) (Benoit and Bowen 1990a, 1990b, Bowen et al.1993) or harbour seals ( Phoca vitulina) (Bowen and Harrison 1996) seals in Atlantic Canadian waters. Nevertheless, seals may accumulate the majority of their sealworm by consumption of relatively few, large, heavily infected fish (McClelland et al. 1990). Moreover, in light of anecdotal reports that heads of 58 Sealworms in the North Atlantic : Ecology and Population Dynamics Fig. 1. Microscopic view of larval sealworm in fish muscle tissue. The worms in the photo are about 3-5 cm in length. Photo: T. Jensen larger prey


I IN NT TR RO OD DU UC CT TI IO ON N
A s they lack direct economic significance, small benthophagous fish, which include the juveniles of commercially important demersal species as well as underutilised species, are often overlooked as potential reservoirs of Pseudoterranova decipiens larvae.Recent field and laboratory studies have shown, however, that they may be essential in the transmission of sealworm to larger, commercially exploited fish, if not directly to definitive (seal) hosts (McClelland 1995).In an earlier survey of Sable Island Bank on the central Scotian Shelf (McClelland et al. 1990), larval sealworm were found in 26 of 32 marine fish species.While it was most prevalent and abundant in large demersal piscivores, P. decipiens often occurred in greatest density (nr/unit host weight) in small (juvenile and mature) benthic consumers (Fig. 1).
Laboratory experiments (McClelland 1995), in which P. decipiens larvae were transmitted to fish via benthic copepod and amphipod intermediaries, revealed that sealworm grew to a length of 8 mm in crustacean hosts, but larvae as small as 2 mm in length were infective to small fish.One of the conclusions drawn from this study was that the primary fish hosts of sealworm are probably small benthic consumers.It was also apparent that sealworm larvae recently transmitted to these hosts would be considerably smaller than those detected by candling procedures typically employed in seal-worm surveys.Hence, in order to obtain accurate worm counts when surveying infections in smaller demersal fish, more refined examination procedures involving microscopic inspection of host tissue squashes or digested host tissues might be required.
Given that fishes of diverse phylogeny are equally susceptible to sealworm infection in the laboratory, light infection or absence of infection in natural populations of certain host species is probably attributable to ecological, behavioural and physiological (e.g.host response) barriers to the transmission of the parasite (McClelland 1995).Surveys of the diets and ascaridoid infections of flatfishes inhabiting Sable Island Bank (Martell andMcClelland 1994, 1995), for example, indicated that disparities in parameters of sealworm infection among sympatric flatfish species was largely related to the exploitation of different prey.Juvenile Canadian plaice (Hippoglossoides platessoides), which were heavily infected with sealworm, fed on benthic suprafauna, i.e. free swimming organisms (amphipods, mysids etc.) closely associated with bottom.Winter flounder (Pleuronectes americanus), which was rarely infected with larval sealworm, consumed more sedentary infauna and attached epifauna.
As suggested by analyses of their diets, seals may incur P. decipiens infections primarily through consumption of smaller fish.Larger demersal fish, while often heavily infected, are seldom exploited by grey (Halichoerus grypus) (Benoit and Bowen 1990a, 1990b, Bowen et al. 1993) or harbour seals (Phoca vitulina) (Bowen and Harrison 1996) seals in Atlantic Canadian waters.Nevertheless, seals may accumulate the majority of their sealworm by consumption of relatively few, large, heavily infected fish (McClelland et al. 1990).Moreover, in light of anecdotal reports that heads of The worms in the photo are about 3-5 cm in length.Photo: T. Jensen larger prey are often discarded by seals, it is possible that exploitation of heavily infected mature piscivores is underestimated in surveys which rely on otoliths for identifying and ageing seal prey.
The "seal prey" time series was designed to provide data for a predictive sealworm model proposed at a two-part sealworm workshop held in Halifax Nova Scotia in April 1987 and June 1988 (Bowen 1990).Rather than monitoring larval sealworm infections in a single indicator host, as is the case in the "sealworm index" time series (McClelland andMartell 2001, McClelland et al. 2000), the parasite was monitored in numerous fish species, at a few specified locations.As a consequence of dwindling groundfish resources and changing research priorities, the project was abandoned before completion.Some of the data collected in the Gulf of St. Lawrence have been reported (Boily andMarcogliese 1995, Marcogliese 1995), but much of it remains unpublished."Seal prey" surveys of fish from the Scotia-Fundy region proved useful in revealing small benthic consumers of potential importance in sealworm transmission.The results of surveys of the Breton Shelf, Sable Island Bank, and the northeastern Gulf of Maine are documented here, with analyses of spatial variations in sealworm prevalence and abundance in 1989-1990, and changes in infection parameters in fish from Sable Island Bank since 1986-1987(McClelland et al. 1990)).The efficiencies of procedures used for detecting sealworm larvae, namely candling, dissection and digestion, are also assessed.

M MA AT TE ER RI IA AL LS S A AN ND D M ME ET TH HO OD DS S
Marine fish samples were collected primarily from Department of Fisheries and Oceans (DFO) research vessels, Alfred E. Needler and Lady Hammond during cruises dedicated to surveys of larval P. decipiens, and other parasites of marine fish and benthic macrofauna.Samples for an inventory of larval anisakines in seal prey (the "seal prey" survey) were taken from the southeastern Breton Shelf, Sable Island Bank and the northeastern Gulf of Maine between November 1988 andAugust 1993, but primarily from February 1989 to October 1990 (Fig. 2).Fish were caught in sets of 30 to60 min.duration, with a Western IIA otter trawl towed at 6.5 km h -1 , and stored in the freezer.On return to the DFO Halifax Fisheries Research Laboratory, samples were allowed to thaw for 24 h at room temperature (15 to 20 o C) prior to inspection.After total length, weight and sex were recorded, visceral organs and mesenteries were inspected for larval anisakines.Each fish was then boned, and fillets, napes (hypaxial musculature enclosing the body cavity) and flesh which remained on the frame were examined for nematodes by slicing and mechanical destruction of tissues (Power 1961).As a rule, nematodes were identified by eye and counted, but inspection with a dissecting microscope was sometimes necessary in order to identify smaller worms and specimens damaged (e.g.cut into fragments) during boning and slicing of flesh.P. decipiens sibling species B, recently designated as P. decipiens (sensu stricto) (Paggi et al. 2000) is the only species of sealworm known to occur in the waters surveyed here (Brattey and Davidson 1996).
Efficiencies of routine examinations of the musculature of groundfish were tested using procedures described by McClelland et al. (2000).Atlantic cod (Gadus morhua) (n = 90), sea raven (Hemitripterus americanus) (n = 23), long horned sculpin (Myoxocephalus octodecemspinosus) (n = 30) and Canadian plaice (n = 163) were collected from the Scotian Shelf and southwestern Nova Scotia during groundfish sampling cruises in June and July 1991.The flesh of individual fish, together with nematodes identified and counted following routine inspection, were placed in a 4 L beaker containing 2 L of 1% HCl with 5 g of 1:10,000 pepsin L -1 .After incubation for 2 to 3 h at 35 to 40°C with continuous stirring, the contents of each beaker were strained through a series of sieves ranging from 5.0 to 0.3 mm in mesh size.All nematodes recovered, including those severed during boning and slicing of the flesh, were identified and counted as above.
To determine the frequencies of occurrence of smaller sealworm larvae (2 to10 mm in length) in demersal fish, samples of various benthic consumers were collected from the northern and southern slopewaters of Sable Island Bank.They were stored on ice in portable coolers while at sea, and refrigerated at 0 to 2°C on return to the Halifax Lab.Within 2 to 5 days of capture, the fish were measured, weighed and gutted.The viscera were inspected for nematodes under low magnification with a 'Luxor' lamp.Bodies of fourbeard rockling (Enchelyopus cimbrius), Vahl's eelpout (Lycodes vahlii), hookear (Artediellus atlanticus) and mailed sculpin (Triglops murrayi), alligatorfish (Aspidophoroides monopterygius), spiny lumpsucker (Eumicrotremus spinosus), and juvenile plaice, collected from October 1991 to August 1993, were examined by mechanical destruction of the flesh under a "Luxor" lamp.Bodies of alligatorfish, sampled in August '93, and those of rockling, juvenile cod and haddock (Melanogrammus aeglefinus), mailed sculpin, lumpsucker, and juvenile plaice and halibut (Hippoglossus hippoglossus), sampled from May 1994 to October 1996, were placed, individually, according to size, in 0.4 to 4.0 L beakers containing 200 to 2000 ml of pepsin -HCl solution.After incubation at room temperature (18 to 22 o C) for 2 to 3 h with continuous stirring, the digests were inspected for nematodes by decanting into finger bowls in 50 to 100 ml aliquots, and examining with a dissecting microscope at medium to high power.All nematodes recovered, viable or necrotic, were identified (with the aid of a compound microscope for smaller worms), counted, rinsed in 0.9% saline, fixed in hot 5% glycerin in 70% ETOH, and cleared in glycerin.Lengths of sealworm were ascertained by placing cleared nematodes on a dissecting scope equipped with a drawing tube, and tracing their outlines onto a computer graphic tablet with an electronic pen.The images were fed directly to a Mackintosh Quadra 700 where they were converted to actual nematode length on a precalibrated template from NIH Image (version 1.61).
Quantitative parameters of infection such as prevalence (P), abundance (A), intensity (I), and density (D) are defined according to Margolis et al. (1982) bers and length ranges of strata varying with host species according to length structures of samples.Spatial and temporal variations in prevalence and abundance were analysed by two-way (location • length, or survey • length) ANOVA (SYSTAT) for each host species (McClelland et al. 1983a).For ANOVA of prevalence, individual infected fish were assigned the value "1", and each uninfected fish, the value "0" (Li 1964;Neter et al. 1985).To permit ANOVA of abundance, frequency distributions of sealworm counts, which were positively skewed to varying degrees, were brought closer to normality by a log (n + 1) transformation (Platt 1975).

Seal prey survey
From November 1988 to August 1993, a total of 9,523 fish belonging to 39 species were collected from the Breton Shelf, Sable Island Bank and the northeastern Gulf of Maine.Sable Island Bank samples included 4,847 fish from 32 species, while those from the Breton Shelf and Gulf of Maine were comprised of 2,122 fish from 24 species, and 2,554 fish from 22 species respectively.All fish from the Breton Shelf and Gulf of Maine, and the majority of those from Sable Island Bank were sampled between February 1989 and October 1990.Additional specimens of benthic consumers such as fourbeard rockling, Vahl's eelpout, snakeblenny (Lumpenus lumpretaeformis), wrymouth (Cryptacanthodes maculatus), hookear and mailed sculpin, alligatorfish and spiny lumpsucker were collected from Sable Island Bank between October 1991 and August 1993.Data from Canadian plaice collected during the latter period are not included in Table 1 or in analyses of spatial and temporal variation below.
When examined by routine procedures in which the musculature was sliced and candled, and the viscera inspected by eye, larval sealworm and whaleworm (Anisakis simplex) were found in the majority of fish species (30 and 26 of 39 host species respectively) inventoried.Larvae of a third anisakine species, Contracaecum osculatum, were found only in Atlantic cod and white hake (Urophycis tenuis) from the Breton Shelf and on Sable Island Bank.In this document, we report on but one of the three species, P. decipiens.
On Sable Island Bank, prevalence and abundance of larval sealworm were greatest (P=100%, A, ranging from 31.08 to 151.31) in large sea raven, cod and monkfish (Lophius americanus), but heavy infections (P ranging from 92 to 100%, A=12.54 to 17.56) were also recorded in mature ocean pout (Macrozoarces americanus) , longhorn sculpin, and Canadian plaice (Table 1).Mature sea raven, cusk (Brosme brosme), cod, monkfish and ocean pout were the most heavily infected hosts (P=96% to 100%, A=13.44 to 108.20) in the northeastern Gulf of Maine.On the Breton Shelf, where no monkfish, and only two sea raven and two ocean pout were collected, the parasite was most prevalent (P=100%) and abundant (A=20.00) in cod >70 cm in length.
The infection of greatest intensity (I=721) occurred in a 112 cm cod from the continental slope waters southeast of Sable Island.
When compared with their counterparts on the central Scotian Shelf, small benthic consumers from the Breton Shelf and Gulf of Maine were (relatively) lightly infected (D=2 to 42 kg -1 ).
Remarkably, the parasite was not found in hookear sculpin and alligatorfish from the Gulf of Maine.
Although sealworm larvae were confined primarily (>90%) to the fillets in the majority of fish surveyed, large numbers occupied body cavities and napes of large piscivores such as mature monkfish and various mature gadids and cottids.In large (> 50 cm in length) monkfish from Sable Island Bank (n=12), 132 (35%)       ANOVA revealed that geographical disparities in prevalence and/or abundance of sealworm were significant (P < 0.01) in 14 of 20 host species sampled in at least two of the three survey areas (Table 2).Levels of infection in samples from Sable Island Bank were significantly greater than those found in Breton Shelf samples in 10 of 14 species contrasted, and also exceeded those recorded in samples of 11 of 17 corresponding host species from the Gulf of Maine.Contrasts of Breton Shelf and Gulf of Maine samples, however, indicated that among 11 species common to both areas, only grey sole (Glyptocephalus cynoglossus) and plaice differed significantly in regard to parameters of sealworm infection.Invariably, spatial disparities in infection levels in four lightly infected hosts, silver hake (Merluccius bilinearis), wolffish (Anarhichas lupus), redfish (Sebastes fasciatus) and winter flounder, were not significant.Plaice was the only host species in which sealworm prevalence and abundance differed significantly in each of the three survey areas. of 1986of -87 (McClelland et al. 1990) and 1989-90 samples (herein) indicate that apparent increases in sealworm infection parameters in 8 of 10 host species from Sable Island Bank were statistically significant (Table 3).Both prevalence and abundance of the parasite had increased in monkfish, haddock, longhorn sculpin, grey sole and yellowtail flounder (Pleuronectes ferrugineus), while abundance alone had increased in cod, white hake and plaice.

Efficacy of examination procedures
Routine examinations (slicing and candling) for P. decipiens in the flesh of groundfish proved very efficient when tested against digestion procedures (Table 4).A total of 24 worms were recovered from cod < 30 cm in length by routine

67
NAMMCO Scientific Publications, Volume 3 In plaice < 30 cm in length, a total of 73 worms were found by routine inspection, and there was a net loss of a one nematode following digestion.While digestion procedures revealed infection in one additional fish in the 31-40 cm length range, the total number of nematodes found (and hence, abundance) declined from 458 following routine inspection to 440 after the flesh was digested and sieved.Some nema-todes severed during boning and slicing of the flesh did not survive digestion, and the extremities of many intact nematodes had also deteriorated.In plaice ≥41 cm in length, sealworm prevalence remained the same following digestion of the flesh, but there was a net loss of two worms; 62 larvae were found by mechanical inspection, and although digestion revealed a total of four previously undetected worms from three fish, six nematodes from two other fish were lost.
Evidently, many encapsulated, necrotic sealworm in the flesh of sculpins were lost when incubated in pepsin-HCl at 35°C.Only 9 (56%) of 15, and 439 (72%) of 611 P. decipiens, detected by mechanical inspection of the flesh of longhorn sculpin and sea raven, respectively, were recovered after digestion.Digestion of small benthophagous hosts Digestion of host bodies at ambient temperature, followed by decanting and microscopic examination of the sediment proved far more efficacious for detecting sealworm in the flesh of small benthic consumers than routine dissections conducted with the naked eye or under low magnification with a "Luxor" lamp.Infection parameters revealed by dissection of fresh iced specimens (Table 5) were not dissimilar to those obtained through routine examination of frozen samples (Table 1).Aside from three (14%) of 21 larvae from alligatorfish, and one (0.5%) of 196 larvae from mailed sculpin, all sealworm detected by dissection of small demersal fish exceeded 10 mm in length, the smallest larvae being an 8.96 mm specimen from alligatorfish.
When digestion procedures were employed, on the other hand, 36 (24%) of 159 sealworm found in rockling, 34 (46%) of 73 in mailed sculpin, 72 (84%) of 86 in alligatorfish were <10.00 mm in length.In yearling plaice (<15.0 cm in length) and 0-group halibut, 31% (86 of 277) and 55% (6 of 11) of the larvae, respectively, were <10.00 mm long.Rockling yielded the smallest larva, a 2.14 mm specimen, but sealworm as small as 3.01-3.27mm in length were also detected in mailed sculpin, alligatorfish and juvenile plaice.As apparent from summaries of sealworm infection parameters (Table 5), digestion of rockling, mailed sculpin, alligatorfish, and plaice (>15.0 cm long) yielded far greater numbers of nematodes than dissection.Abundances of the parasite in digested samples were greater than those found in dissected samples by a factor of 16 in rockling, three in mailed sculpin, 19 in alligatorfish, and five in plaice; densities in digested samples exceeded those found in dissected samples by 4-to 17fold.
Only three (5% of 62) sealworm from juvenile cod, and 8 (4% of 196) from plaice >15 cm in length were <10 mm in length.Infection parameters in digested plaice 18-26 cm in length did not differ significantly from those found in dissected plaice or previously frozen plaice (Table 1) of similar size.Digestion of 32 lumpsuckers yielded only 11 worms, all of which exceeded 25 mm in length, while three sealworm, 9, 16 and 26 mm in length, were recovered from digests of four juvenile haddock, 14-16 cm in length.

D DI IS SC CU US SS SI IO ON N
The present survey confirms earlier observations (McClelland et al 1990) that, in Atlantic Canadian waters, larval sealworm is most prevalent and abundant in mature demersal piscivores (monkfish, Atlantic cod, cusk, and sea raven), and, to a lesser extent, in mature benthic consumers (ocean pout, longhorn sculpin and Canadian plaice).The survey further reveals, however, that sealworm densities are greatest, not only in the juveniles of sea raven, longhorn sculpin and plaice, but also in small, non-commercial benthophagous species, such as fourbeard rockling, Atlantic hookear and mailed sculpin, alligatorfish and spiny lumpsucker.Rockling, hookear sculpin and mailed sculpin are new host records for sealworm, and the parasite is also reported, for the first time, from arctic eelpout (Lycodes reticulatus), fourline snakeblenny (Eumesogrammus praecisus), snakeblenny, daubed shanny (Lumpenus maculatus) and wrymouth (see McDonald and Margolis 1995 for most recent list).
Analyses of spatial disparities in sealworm prevalence and abundance in the present study revealed that infection parameters in samples from Sable Island Bank were significantly greater than those found in Breton Shelf samples for 10 of 14 species contrasted, and also exceeded those recorded in samples of 11 of 17 corresponding host species from the Gulf of Maine.These results are consistent with the findings of earlier multispecies, and "sealworm index" surveys (McClelland and Martell 2001, McClelland et al. 1990, 2000), and clearly reflect the impact of the large Sable Island grey seal colony on P. decipiens infections in local groundfish populations.Grey seals probably have a marked influence on infections in groundfish from the Breton Shelf and northeastern Gulf of Maine groundfish as well, although harbour seals may play a significant role as definitive hosts in the latter region (McClelland et al. 2000).Along the southern Norwegian inshore, where grey seals are outnumbered or absent, heavy sealworm infections   (6.38-15.42)hippoglossus in groundfish are attributed to harbour seals (Aspholm et al. 1995, des Clers andAnderson 1995).In the past, this was also true for inshore areas of the southern Gulf of St. Lawrence, and the south shore of Nova Scotia (Scott and Martin 1959).
There were great disparities in larval sealworm infection parameters in small benthic consumers from our three survey areas (Table 1).Hence, it would seem that these species might prove useful, like Canadian plaice (McClelland andMartell 2001, McClelland et al. 2000), as indicator hosts for monitoring geographical and temporal distributions of the parasite.However, while surveys of these smaller fish may provide a more accurate picture of local distributions of larval sealworm at specific sites, e.g.Sable Island Bank, they would be difficult to conduct on a larger scale.As evident from research cruises employing trawls equipped with liners, these hosts are neither as widely distributed nor as abundant as plaice, and, because of their small size, they are not caught with commercial gear.Further, a significant proportion of larval sealworm found in these smaller hosts are < 10 mm in length (below), and can be detected only through time consuming digestion or tissue squash procedures and microscopy.Plaice of the size (31 to 40 cm TL) monitored in our "index" surveys (McClelland andMartell 2001, McClelland et al. 2000) are seldom infected with smaller larvae, and can be examined efficiently by slicing and candling of the flesh.Notably, of 11 host species which were common to the three areas surveyed here, plaice was the only the only species having significant disparities in sealworm prevalence and abundance in each area.
Given the growth of the Sable Island grey seal population over the course of the decade (Zwanenberg and Bowen 1990), it was not surprising that parameters of sealworm infection in our 1989-1990 samples of groundfish from Sable Island Bank were significantly greater than those recorded during an earlier survey (McClelland et al. 1990).As shown in our more recent "sealworm index" surveys (McClelland and Martell 2001), however, sealworm abundance in 31-40 cm plaice from Sable Island Bank has been declining since 1990, despite the continued growth of the grey seal population.Although this fact was not brought out in the present document, falling infection levels were detected in small benthic consumers collected from 1991 to 1993 even though they were subjected to more rigorous examinations.Heavy sealworm infections could prove lethal to fish of this size, either directly, through damage to vital organs and tissues, or indirectly, by impairing the host's ability to forage and avoid predators (McClelland 1995).Possibly, the most heavily infected benthic consumers have become increasingly vulnerable to predation pressure as predator (seal) populations continue to grow, and groundfish stocks decline (Mohn and Bowen 1996).In support of this hypothesis, the tails of worm count distributions in plaice from Sable Island bank have become increasingly truncated in more recent "index" surveys (McClelland and Martell 2001).
As evident from the results of earlier surveys (Young 1972, Platt 1975, Wootten and Waddell 1977, McClelland et al. 1990), accurate estimates of sealworm levels in large piscivores, including mature monkfish, gadids and sculpins, can only be obtained by examining napes and body cavities, as well as fillets.The present study shows that nematodes occupying the body cavity and napes may represent a significant proportion, if not the majority of sealworm in large demersal piscivores, and also in tiny (<10 cm in length) benthic consumers such as mailed sculpin and lumpsucker.Power (1961) demonstrated that candling procedures, routinely conducted at fish plants for detection and removal of larval sealworm from cod fillets, were not very effective.Using "destructive" examinations in which fillets were cut into thin slices prior to candling, he found that the majority of sealworm in the fillets of larger cod escaped detection on the production line.An assessment of Power's slicing and candling technique (herein) reveals that this approach, used in our present and earlier surveys (McClelland et al. 1990, 2000, McClelland and Martell 2001), may be as efficacious as more laborious and time consuming digestion procedures.Indeed, some moribund nematodes from frozen samples, especially those damaged during boning or filleting, may be destroyed when incubated in warm pepsin-HCl solution.Unfortunately, since de-structive examination of both fillets and napes, along with inspection the visceral organs and body have not been universally employed in sealworm surveys, it is often difficult to make temporal and spatial comparisons with the results of other investigators (McClelland et al. 1990).One final factor that must be considered when assessing the accuracy of survey results is the possibility that some sealworm larvae may simply be too small to detect in the white flesh of groundfish with the unaided eye.Larval sealworm as small as 2 mm in length are infective to fish hosts in laboratory transmissions (McClelland 1995), and 4 to 5 mm sealworm have been found in juvenile Icelandic cod through microscopic examinations of tissue squashes (Pálsson MS 1979).However, the smallest larvae typically reported from surveys employing candling of sliced or whole fillets have been 14 to 15 mm in length (Scott andMartin 1957, Templeman et al. 1957).Present results reveal that failure to detect smaller nematodes is cause for concern, but perhaps, only when it pertains to data from small benthophagous hosts.Whereas the smallest nematodes detected by slicing the flesh of small benthic consumers were about 9 mm long, larvae as small as 2 mm in length were recovered by digestion of host bodies at ambient temperature and microscopic examination of the sediment.Twenty-four to 84 % of the larvae yielded by digestion of fourbeard rockling, mailed sculpin, alligatorfish, and 9 to 15 cm juvenile plaice were <10 mm in length.Sealworm of this size also occurred in juveniles of cod and other commercially exploited species including haddock (Melanogrammus aeglefinus), plaice >15 cm in length, and halibut (Hippoglossus hippoglossus), but were not nearly as numerous in these latter hosts.Only 4% to 5% of the larvae from 20 to 29 cm cod and 15 to 26 cm plaice were <10 mm in length.The especially low worm yields from dissections of alligatorfish (Table 5) are attributable mainly to the anatomy of the species.Alligatorfish are thin elongate fish with thick skin and relatively little flesh, and often weigh only a gram or two.Sealworm larvae, which average ca.7 mm in length in alligatorfish, were probably destroyed by the knife during dissection.
Larval sealworm recovered by digestion of tiny benthophagous fish (mature fourbeard rockling, hookear and mailed sculpin, alligatorfish and 9 to 15 cm juvenile plaice) had bi-to polymodal length distributions (Fig. 3).The first mode consisted primarily of larvae in the 2 to 10 mm length range, i.e. similar in size to P. decipiens larvae, naturally found in benthic crustaceans, and to those successfully transmitted to fish in the laboratory (McClelland 1995).This would seem to indicate that these hosts acquire P. decipiens larvae by feeding on invertebrate hosts during a particular season or seasons, rather than continuously throughout the year.Notably, Mysis mixta, which are naturally infected with larval sealworm on Sable Island Bank (Martell and McClelland 1995), are consumed by juvenile plaice most frequently during winter (Martell and McClelland 1994).As a probable consequence of exploitation of M. mixta in winter, there is a strong pulse of 2-10 mm sealworm larvae in spring samples of juvenile plaice from Sable Island Bank (McClelland 2000).
While not prominent in seal diets (Benoit and Bowen 1990a, 1990b, Bowen and Harrison 1996, Bowen et al 1993) small, seemingly inconsequential benthic consumers such as rockling, hookear and mailed sculpin, alligatorfish and juvenile plaice are frequently consumed by larger, economically important species such as cod (Scott and Scott 1988).Hence, they may be the source of heavy sealworm infections in commercially exploited fish, and play a significant, albeit indirect role in the transmission of P. decipiens to seals.Aspholm et al. (1995) speculate that bullrout (M.scorpius) (shorthorn sculpin in the northwest Atlantic) performs a similar function in Norwegian inshore waters.
In some instances, sealworm may be transmitted directly to seals by invertebrate hosts.Larvae >5 mm in length appeared to be capable of maturing in an in vitro system (McClelland and Ronald 1974).P. decipiens L4s as small as 8 mm in length have been found in stomachs of newly weaned harbour (Boulva and McLaren 1979) and grey seal pups (unpublished data) feeding on invertebrates.

Fig. 1 .
Fig. 1.Microscopic view of larval sealworm in fish muscle tissue.The worms in the photo are about 3-5 cm in length.Photo: T. Jensen

A
AC CK KN NO OW WL LE ED DG GM ME EN NT TS S The authors thank Dave Marcogliese (Environment Canada, Montreal) and Johanne Guerin (Department of Fisheries and Oceans Maurice-Lamontagne Institute, Mont-Joli, Quebec) for assisting in the collection of samples during research cruises in February 1990 and January 1995.Gratitude is also extended to Ione von Herbing (University of Maine) and Mark

Table 2 .
Larval sealworm (Pseudoterranova decipiens) infections in various groundfish species from the Breton Shelf (1), Sable Island Bank (2) and the northeastern Gulf of Maine (3); results of 2-way-ANOVAs of sealworm prevalence (P) and abundance (A) with host length and geographic origin.

Table 4 .
Prevalence (P), abundance (A), maximum intensity (Imax) and density (D) (no.kg-1 host round weight) of Pseudoterranova decipiens larvae found in frozen samples of Scotian Shelf and Gulf of Maine groundfish by routine examination employing slicing and candling of host flesh, and following digestion of flesh in pepsin-HCl solution at 35º C; groundfish samples were collected in June and July 1991.

Table 5
Pseudoterranova decipiens larvae recovered from fresh iced samples of small benthic consumers by dissection, and by digestion of host bodies in pepsin-HCl at ambient temperature; samples were collected from Sable Island Bank between October 1991 and October 1996.