Effluent discharged by pulp and paper mills (PPE) using bleaching agents, thermomechanical agents, or other methods to produce newsprint from wood, have been reported to cause eutrophication, influence pelagic primary production, and induce stress-related effects in benthic macrofauna especially fish living downstream (Lehtinen 1990; Munkittrick et al. 1991; 1992; 1994; McMaster et al. 1991; Culp et al. 2003; Sandström et al. 2005). Long-term impact of the discharge might also affect ecosystem health (Sandström et al. 2005). Recent studies have provided evidence that resin acids, produced from softwoods during the pulping process, especially dehydroabietic (DHAA) and abietic acids (AA), can accumulate in sediment several kilometers downstream from the outfall and are responsible for the adverse effects observed in fish (Fragoso et al. 1998; Billiard et al. 1999; Lehtinen et al. 1999; Pacheco and Santos 1999; Lappanen and Oikari 1999). Several methods have been devised to assess stress-related effects in fish exposed to pollutants (Adams 1990). Some studies have revealed that fish exposed to sublethal levels of pulp and paper mill effluent (PPE) resulted in fin necrosis impaired growth, skeletal abnormalities, morphological deformities, and molecular, biochemical, physiological, immunological, reproductive, and histopathlogical changes, ultimately affecting the population and community structure (Lehtinen et al. 1984; Bengsston et al. 1988; Couillard et al. 1988; McMaster et al. 1991; Munkittrick et al. 1991; 1997; Adams et al. 1992; Pesonen and Andersson 1992). Disease and changes in the levels of parasites have also occurred in fish inhabiting degraded habitats, including locations where PPE was discharged (Lindesjoo and Thulin 1990; Khan and Thulin 1991; Overstreet 1993). There is increasing evidence that changes in parasitism might be useful as an additional biomarker to assess stress-related effects in fish induced by pollutants (Khan and Thulin 1991; Axelsson and Norrgren 1991; Khan and Payne 1997; 2004; Marcogliese and Cone 1997; Khan 2003a; b; 2004).

The winter flounder (Pleuronectes americanus) is an ideal sentinel fish species for a biomonitoring program since it inhabits shallow-water marine and estuarine areas, is sedentary, and buries itself in sediment where some anthropogenic compounds tend to accumulate (Scott and Scott 1988). Some studies have reported that this species was sensitive to pollutants including PPE but sampling was limited to once in late spring and summer (Khan et al. 1994a;b; 1996). A previous study reported abnormalities in flounder inhabiting a site near to a pulp and paper mill discharging untreated effluent into a fjord at Corner Brook, Newfoundland (Khan et al. 1996). It was observed also that winter flounder inhabiting another marine inlet in Newfoundland where primary-treated effluent was discharged caused severe abnormalities including changes in the levels of two parasites (Khan 2006). However, the long-term effects of exposure of the fish to untreated PPE is unknown especially its influence on fish size distribution, body condition, histological changes, and gonadal development. In addition, the response of external and internal parasites of this fish species to chronic PPE has not been previously determined, especially since some were affected either positively or negatively following exposure (Lafferty 1997). Consequently, the objective of the present investigation was designed to assess the health of winter flounder living near the pulp and paper mill discharging untreated effluent by repetitive sampling each spring over a 5-year period on over-wintering fish prior to feeding, using several conventional bioindicator methods (Adams et al. 1992). In addition, this integrated and multidisciplinary study also examined the potential use of four parasites as an indicator of PPE contamination (Khan and Thulin 1991; Khan and Payne, 1997; Khan 2006).

Materials and Methods

Site Description

The pulp and paper mill is located at the upper extremity of a fjord, the Humber Arm, at Corner Brook, Newfoundland (Fig. 1). The inlet receives fresh water from the Humber River and as a result, brackish water occurs in the upper 10 m, with salinity at 2% and surface temperature from 0° to 15°C. The mill once produced newsprint by thermomechanical and sulphite-bleaching processes up to 1992 but the latter process was discontinued after this time. Wet debarking of logs, consisting of black spruce (Picea mariana) and balsam fir (Abies balsamea), has led to a buildup of bark and fibre at the bottom, up to 2 km from the outfall as the current flows towards the ocean on the south side of the fjord. The discharge was estimated at 3.3 × 107 m3 per year (Sea Consult Limited 1993; Fraikin et al. 1995). The untreated effluent was composed of suspended solids (79–93 mg/L) that were visible as a brown discoloration at the surface up to 2 km from the point of discharge on the southern side of the fjord. It also contained tannins and resin acids (27 mg/L), primarily dehydroabietic (600–800 μg/L) and abietic acids (600–1200 μg/L) (unpublished report, Environment Canada, St. John’s, Newfoundland, 1994). The inlet also was a repository for raw sewage and discharges from a gypsum plant, two oil storage depots, and a fish processing plant but all combined do not equate with that discharged from the paper mill. Sediment, about 2 km from the mill at Birchy Cove, where the fish were collected, consisted of dark mud mixed with bark and fibre and emitted a strong odour of hydrogen sulphide. The bottom was virtually devoid of macro-invertebrates with the exception of the periwinkle (Littorina littorea) and sparsely distributed strands of kelp (Fucus sp.) Three reference sites, all located on the northern side of the fjord, including Summerside (1993), Meadows (1994–1995) and Cox’s Cove (1996–1997), were used since adequate numbers of flounder were not always present during some years at the first- and second-mentioned sites.

Fig. 1.
figure 1

Location of the pulp and paper mill at Birchy Cove (BC) and reference sites at Summerside (SS), Meadows (M), and Cox’s Cove (CC) in the Humber Arm (HA) and Middle Arms in western Newfoundland

Full size image

Fish Species

Winter flounder, prior to foraging activity, were captured between 1100 and 1500 hr by SCUBA divers using a dipnet and necropsied on site over a 3-day period during the last week of April and the first week of May, each year, 1993 to 1997. A total of 360 and 339 flounder were captured and examined near the mill and reference sites, respectively, over the 5-year period. The surface water temperature was 0–1°C in all five years. A record was made of external macroscopic lesions and parasites followed by the total body length, eviscerated body weight, and that of the liver and gonad. Samples of tissues, including the liver, spleen, gonad, kidneys, and gills, were preserved in 10% buffered (pH 7.6) formalin for histological assessment. The digestive tract was removed afterwards, frozen in dry ice, but thawed and examined later for metazoan parasites. These were enumerated and stained with borax carmine (digeneans) or cleared in Rubin’s fluid (acanthocephalans) for identification. Histological samples were processed by conventional methods and sections stained with hematoxylin and eosin. Additional sections of spleen were also stained with Perl’s Prussian blue for hemosiderin, its concentration estimated by digital image analysis and expressed as a percentage of the area scanned (Khan and Nag 1993). The number of macrophage centres containing hemosiderin in the liver and kidneys was estimated per mm2. This was also done for the number of clear cell foci in the liver. The stage in oogenesis and spermatogenesis was determined following the methods of Janssen et al. (1995) and Grier (1981).

Data Analysis

Condition (k) factor (eviscerated body weight/length3) and somatic indices (organ weight/eviscerated body weight) were compared for site significance by an analysis of covariance (ANCOVA). Means and standard errors of variables were calculated for each fish group. Data for the year 1994 compared fish length distribution between the two sites by a t-test. Chi-square and Fisher’s exact probability tests were used to compare prevalence (%) between sampling sites. The non-parametric test of Kruskal-Wallis compared parasitic mean abundance between samples from contrasting sites. Log transformation was used to convert parasite distributions to normality. Mean abundance, intensity, and prevalence follow the terminology proposed by Bush et al. (1997).

Results

Some significant differences were apparent between winter flounder sampled at Birchy Cove, down-current from the paper mill and the reference sites. There was an abnormal size distribution as a greater percentage of larger fish occurred near the mill than at the reference site (Fig. 2). Flounder smaller than 20 cm were seen during all five years of sampling only at the reference sites but these evaded capture. External lesions, mainly fin necrosis and less often ulcers, were noted only in fish captured near the mill (Fig. 3a). Differences of other biological variables were also observed between flounder taken from the contrasting sites. Condition (k) factor was significantly greater in reference than in mill samples in three or four years (Fig. 3b,c). Greater elevated levels of hepatosomatic indices (HSI) were noted in flounder captured near the mill than at the reference sites in three of the five sampling years (Fig. 3d,e). However, the gonadosomatic index (GSI) was different in only one of the 5-year periods (Fig. 3f,g). In contrast, delayed gonadal development occurred more often in both male and female flounder caught at Birchy Cove than at the reference sites (Fig. 3h,i). Examination of histological sections of the testes revealed that male flounder at the reference sites were spermiating whereas only spermatids were observed in the semeniferous tubules at Birchy Cove. Female fish at the reference sites were either spawning or in the pre-spawning stage as the germinal vesicle was located peripherally, in contrast to ovaries that, although enlarged, were in the previtellogenic stage at Birchy Cove. In 1997, all of the female fish sampled at Cox’s Cove showed normal ovarian development compared to 69% of the fish sampled at Birchy Cove. Some of the ovaries, 31%, displayed no evidence of development at Birchy Cove in contrast to none at Cox’s Cove. Moreover, about 20% were at the previtellogenic stage at Birchy Cove while none occurred at the reference site.

Fig. 2.
figure 2

Comparison of the total length (cm) groups of winter flounder sampled near a pulp and paper mill at Birchy Cove and a reference site at Meadows in spring 1994. Asterisks indicate significant differences (p ≤ 0.05) between comparison groups

Full size image
Fig. 3.
figure 3

Comparison of biological variables, mean and standard error (0 ± s.e), in winter flounder sampled at Birchy Cove (open circles) and reference sites (closed squares) over a 5-year period (1993–1997). Asterisks indicate significant differences (P ≤ 0.05), smaller for individual years and larger for all years. (a) epidermal lesions, (b) k-factor-male, (c) k-factor-female, (d) hepatosomatic index (HSI)-male, (e) HSI-female, (f) gonadosomatic index (GSI)-male, (g) GSI-female, (h) delayed gonadal development-male, (i) delayed gonadal development-female, (j) clear cell foci (CCF) in the liver, (k) hemosiderin in the spleen, and (l) Steringophorus furciger in the digestive tract

Full size image

Histopatogical changes occurred more often in the tissues of flounder taken at Birchy Cove than at the reference sites. These included clear cell foci in the liver and the concentration of hemosiderin deposits in macrophage centers (MCs) that were greater in fish living near the mill than at the reference sites (Fig. 3j,k). Additionally, the prevalence of MCs varied from 60–100% and 50–95% in the liver and kidneys, respectively, in samples taken at Birchy Cove whereas these values were 10–20% and 15–25%, respectively, in reference fish over the 5-year period. The highest values of the MCs, occurring in the liver and kidneys at the reference sites, were observed in 1993 in samples captured at Summerside. Lesions were apparent also in the secondary lamellae of the gills of fish captured near the mill than at the reference sites. Hyperplasia of the secondary lamellae at the distal extremities and at the bases, lamellar fusion, and telangiectasis were seen mainly in the Birchy Cove samples. Comparison of hyperplasia in the distal end of the secondary lamellae varied from 85–100% in samples from Birchy Cove in contrast to 10–15% in reference fish over the 5-year period. The highest values in the prevalence of these lesions occurred consistently in flounder sampled at Summerside.

There were also significant differences in the parasitic levels of flounder sampled from the two contrasting sites. The stomach of all fish sampled was devoid of food. An ectoparasite, metacercariae of a digenetic trematode, Cryptocotyle lingua, was more prevalent in samples taken from Birchy Cove than from the reference sites (Fig. 4). However, three endoparasites, two digeneans, Steringophorus furciger and Macvicarius soleae, and an acanthocephalon, Echinorhynchus gadi, were more abundant in reference fish than in samples taken near the paper mill (Fig. 4).

Fig. 4.
figure 4

a–c: Prevalence (%) of Cryptocotle lingua on the skin of flounder and mean abundance (0 ± s.e) of M. soleae and E. gadi, respectively, in the digestive tract of winter flounder sampled near a pulp and paper mill at Birchy Cove and at three reference sites from 1993 to 1997 inclusive. All values were significantly different (p ≤ 0.05) between the comparison groups

Full size image

Discussion

Four parasites, in the present study, were significantly different in flounder sampled near the paper mill than at the reference sites. Infestation with metacercariae of the ectoparasitic digenean, C. lingua, was observed primarily in samples captured near the mill whereas there were significantly fewer metazoan parasites in the digestive tract than in the reference samples. Recent field studies have provided similar evidence of differing parasitic levels in fish taken from different polluted habitats (Broeg et al. 1999; Steyermark et al. 1999; Khan 2003a; Khan and Payne 2004). Infestation with metacercariae in samples from Birchy Cove was probably associated with a decline of host defence since exposure to PPE reduced immune-dependent lymphocytic cells in winter flounder and also resistance to disease in juvenile coho salmon (Oncorhynchus kisutch) (Barker et al. 1994; Tierney et al. 2004) It is likely also that toxic components in the PPE were responsible for a lower mean abundance by voiding of the three metazoans in the digestive system of fish sampled near the mill following ingestion of sea water for osmoregulation. Alternatively, toxicity of the PPE to the free-swimming stages of S. furciger and M. soleae might have occurred following their release. It appears, then, that the difference in these parasitic levels was indicative of differing environmental conditions between the two comparison groups of fish and was useful as a valid bioindicator of stress (Khan and Thulin 1991; Khan and Payne 2004).

Observations on ectoparasites responding positively and endoparasites responding negatively to environmental contaminants might be related to the level of toxicity especially along a spatial gradient. Siddall et al. (1993) noted that the prevalence of larval digeneans parasitizing the common whelk, Buccinum undatum, increased significantly with distance from a sewage-sludge dump site harboring toxic heavy metals. Another study reported that the mean abundance of two endoparasites, E. gadi and S. furciger, increased with distance from the outfall of untreated PPE whereas a myxozoan, Ceratomyxa acadiensis, infecting the gall bladder declined at the farthest point (Khan 1998). In a laboratory dose-response study simulating environmental contamination, exposure of parasitized winter flounder to gradient levels of sediment contaminated with either aromatic (PAH) or aliphatic hydrocarbons induced an increase in abundance of two ectoparasites, Trichodina spp. and Gyrodactylus pleuronecti, to threshold levels before declining at higher concentrations (Khan and Payne 2004). In contrast, the endoparasite, S. furciger, decreased progressively, most likely as a result of toxicity. Hemosiderosis, used as a bioindicator, increased progressively with PAH concentration. Consequently, relating changes in parasitism with one or more bioindicators will provide a more accurate assessment of the type of response to pollutants than parasitic levels alone.

The results from the present study also suggested a bias towards larger fish at the site near the pulp and paper mill than at reference locations. Moreover, body condition was significantly lower in fish near the mill than in reference samples except at Summerside where water characteristics were similar to Birchy Cove’s (Ledrew et al. 1988). Growth was reported previously to be affected in fish living downstream from a kraft pulp and paper mill (Munkittrick et al. 1991; 1992; 1994). Lake whitefish (Coregonus clupeaformis) and white sucker (Catostomus commersoni) showed evidence of delayed age to maturity but spoonhead sculpins (Cottus ricei) were older and larger near the mill than at the reference sites (Gibbons et al. 1998). It is likely that increased growth in flounder in the present study was associated with reduced reproductive output as observed in cunner (Tautogolabrus adspersus) inhabiting the same locality at Birchy Cove (Billiard and Khan 2003).

Liver enlargement, revealed as elevated HSI values, was also observed in winter flounder living near the paper mill and was most likely the result of exposure to toxic components of PPE and activation of the detoxifying enzyme system. A previous report on flounder taken from the same site showed elevated levels of ethoxyresorufin-O-deethylase (EROD) activity (Khan and Payne 2002a;b). Several studies have also noted that enlarged livers were associated with elevated levels of EROD activity caused by PPE (Lehtinen et al. 1990, McMaster et al. 1991; Lindström-Seppä and Oikari 1990; Gagnon et al. 1994; Parrot et al. 2000). However, petroleum hydrocarbons are also known to elicit a similar response of EROD activity in fish following exposure (Payne et al. 1987).

Delayed gonadal development was noted also more frequently in both sexes of flounder sampled at Birchy Cove than at reference sites although GSI values were similar in four of the five years. The latter occurred as the gonads of both sexes of flounder, especially females, at the reference sites were still enlarged with fluid after spawning and had not begun to regress to the interphasic condition. Gonadal disruption, including reduction of gonadal size and reduced fecundity, has been observed previously in fish and is associated with a decrease of sex steroids caused by phytosterols in wood (Munkittrick et al. 1991; McMaster et al. 1991; Parrot et al. 2000; Tremblay and Van der Kraak 1999). Additionally, failure of secondary sexual characters to develop and reduce fecundity were reported in fish exposed to PPE but this was not apparent in winter flounder (Munkittrick et. al. 1991; 1992; McMaster et al. 1991).

External lesions and histopathological changes in tissues were present in winter flounder captured near the paper mill in all of the five years. These observations were clearly indicative of stress-related conditions existing in the area and have been noted in previous studies in fish exposed to PPE, PCBS, and petroleum hydrocarbons (Lehtinen et al. 1984; Couillard et al. 1988; Bengtsson et al. 1988; Lehtinen 1990; Axelsson and Norrgren 1991; Khan et al. 1994a;b; Moore and Stegeman 1994; Khan 2003a,b). The presence of clear cell foci, considered as preneoplastic lesions, in the liver and macrophage centres in the spleen, liver, and kidney suggested that fish health was impaired (Myers et al. 1987; Moore and Stegeman 1994; Khan et al. 1994a). Other studies have also reported similar lesions in fish living in habitats affected by pollutants (Myers et al. 1987; Moore and Stegeman 1994; Moore et al. 2005).The results of the present and other studies provide evidence that disease manifestations in fish are valid indicators of environmental stress (Khan and Thulin 1991; Overstreet 1993).

The effects of both untreated and primary-treated effluent discharged by two pulp and paper mills using sulphite-bleaching methods on winter flounder in Newfoundland appear similar but distinct differences also occur. Fish sampled at both impacted sites near Port Harmon and Corner Brook inhabited bark- and fiber-laden sediment emitting an odor of hydrogen sulphide. Chronic exposure to the contaminated sediment resulted in significant differences in the prevalence of external and tissue lesions, disparity in fish sizes, condition factor, HSI, delayed gonadal development, and abundance of the endoparasite, S. furciger, compared to reference samples (Khan 2006; present study). However, the prevalence and/or abundance of three parasites, namely C. lingua in the skin and M. soleae and E. gadi in the digestive tract, were significantly different between fish samples at the impacted and reference sites in the Humber Arm but not apparent between comparable sites in St. George’s Bay (Khan 2006). It is likely that these differences were related to different ecological conditions that affected transmission of the parasites between the Humber Arm and St. George’s Bay.

Results from the present and previous studies on winter flounder sampled near the pulp and paper mill discharging untreated effluent conform to the criteria proposed by Adams (2005) to link a cause-and-effect relationship between environmental pollution and abnormalities in fish. These criteria were considered useful especially in the absence of specific contaminant identities and where multiple anthropogenic stressors occurred. The effects noted in flounder near the mill, using several bioindicators, were observed after repeated annual sampling over a 5-year period. Some of these changes have been reported also in fish living near the outfall of this and other paper mills in Canada and in Europe (Lehtinen et al. 1984; Munkittrick et al. 1991; McMaster et al. 1991; Billiard and Khan 2003; Sandström et al. 2005). Moreover, the effects in winter flounder were observed near to the mill and less often at the reference sites. A previous study at the same site, using multiple biomarkers and bioindicators, suggested that there was a dose-response relationship since a gradient effect was noted from the nearest to the most distant point from the mill (Khan and Payne 1997). Additionally, a laboratory study involving long-term exposure of winter flounder to sediment collected near the mill and at the site where most of the anomalies occurred, revealed multiple changes varying from fin necrosis, reduced k-factor, elevated HSIs, to mortality (Khan 1997). Except for the latter, the other changes were noted in flounder collected near the mill in the present study. Based on multiple bioindicators, there is evidence of a link between the discharge by the paper mill and the abnormalities observed in winter flounder.

In summary, an unusually higher prevalence of the ectoparasite, C. lingua, on the skin and gills and a paucity of three endoparasites, two digeneans, S. furciger and M. soleae, and an acanthocephalan, E. gadi, in the digestive tract of winter flounder captured near the pulp and paper mill than in samples taken at the reference sites is consistent with previous studies on parasites of fish as a useful bioindicator of habitat stress. It was reported that exposure of fish to environmental contaminants, both in laboratory and field studies, resulted in a significant increase externally of ciliates and a decline internally of metazoan parasites (Khan and Thulin 1991; Khan et al. 1994b; Khan 2003a;b; 2004; Khan and Payne 2004). Coincident with parasitic changes, flounder near the outfall from the paper mill were greater in total length and exhibited an enlarged liver but gonadal development was either retarded or delayed. This observation on length suggests that energy resources were allocated towards somatic rather than gonadal growth. Moreover, the consistently high prevalence of lesions externally and in organs such as the gills, liver, spleen, and kidneys was indicative of exposure to toxins. These results support the view that chronic exposure to contaminants impairs the immune system inducing radical changes in disease resistance resulting in greater susceptibility and eventually to disease including external parasites than in uncompromised fish. Consequently, the abnormalities observed in winter flounder living near the pulp and paper mill could ultimately affect population growth. The consequences involved in delayed spawning and survival of eggs and progeny of winter flounder living near the paper mill are unknown. However, Black et al. (1988) reported that PCBs acquired by eggs of female winter flounder were responsible for smaller than normal larval length. In view of these findings, future research should focus on egg production, hatch rate, growth, and larval survival to assess chronic exposure of flounder and other fish species to PPE. (Table 1)

Table 1. Number of winter flounder sampled for biological variables near a pulp and paper mill at Birchy Cove and reference sites at Summerside (SS), Meadows (MD), and Cox’s Cove (CC) in Newfoundland, 1993–1997 inclusive
Full size table