Histological changes in gills induced by residues of endosulfan

Aquatic Tu~~col~gy, 23 (I 992) 65-84 8 1992 Elsevier Science Publishers B.V. All rights reserved 0166-44SX/92/$5.00

65

AQTOX 005 18

istological changes in gills induced by residues of endosulfan Barbara Nowak National Key Centre for Research and Teaching in Aquaculture, University of Tasmania-Launceston, Launceston. Tasmania, Australia

(Received 15 October 1991; revision received 10 February 1992; accepted 11 February 1992)

Respiratory diffusion distance was used to quantify structural changes in gills of catfish exposed to an organochiorine insecticide endosulfan under laboratory conditions. Oedema with lifting of lameliar epithelium and by~~lasia of lamellar epithelium were observed in gills of all catfish containing residues of endosulfan. Both these changes resulted in a statistically significant increase of respiratory diffusion distance. The mean respiratory diffusion distance of catfish collected in the cotton growing area was larger than in two control populations, but this difference was not statistically significant. However, the beta probability of accepting a false null hypothesis of no effect was very large and the effect size index indicated a high degree of departure from the null hy~th~is of no difl,;ence in the respiratory diffusion distance between fish from different ~pulations.

Key words: Fish; Gill; Respiratory diffusion distance; Endosulfan; New South Wales INTRODUCTrON

The respiratory system provides the most extensive interface of a fish with the aquatic environment. The respiratory surface is covered only by a thin epithelium, which forms a barrier between the fish’s blood and the surrounding water (Eller, 1975). Hence, fish gills are constantly exposed to various external factors. A wide variety of structural changes in fish gills has been reported as a consequence of an exposure of fish to pollutants (for review see Mallatt, 1985; Evans, 1987). The respiratory diffusion distance is the distance separating blood lacuna in the lamella from the external medium. Lifting of epithelium or hyperplasia of epithelium results in an increase of the diffusion distance, thus affecting exchange of gases. Increased thickness of the epithelial layers has been reported to result from hyperplasia following experimental exposure to toxicants (Eller, 1975; Kumaraguru et al.,

Correspondettlnre10: B. Nowak, National Key Centre for Research and Teaching in Aquaculture, University of Tasmania-Launceston, PO Box 1214, Launceston, Tasmania 7250, Australia.

66

1982; Solangi and Overstreet, 1982; Woodward et al., 1983). It has been suggested that this increase has a protective function of separating blood from the toxicant. At the same time, however, the respiratory diffusion distance increases, thus disturbing fish physiology (Eller, 1975). Endosulfan (6,7,8,9,10,10-hexachloro-1 ,5,5a,6,9,9a-hexahydro-6,9-methano-2,4,3benzo-dioxathiepin-3-oxide) is an organochlorine insecticide (for review see MaierBode, 1968; Goebet et al., 1982). Technical endosulfan is approximately a 7:3 mixture of two stereoisomers; a- and ~-~ndosulfan. The main degradation products are endosulfan sulfate, endosulfan diol, endosuifan a-hydroxyether and endosulfan lactone. While endosulfan sulfate is as toxic to fish as (x- and &endosulfan, other metabolites are relatively nontoxic. ~orph~metric methods enable the effect of a pollutant on respiratory function of fish to be quantified and at the same time they are more sensitive than measurements obtained by physiological methods (Hughes, 1976). To assess changes induced by endosulfan residues in fish gills the respiratory diffusion distance was measured in the gills of the catfish, K&tinus ~~~2~~~2~s (Plotosidae), simultaneously with the residues of endosulfan. MATERIALSANDMETHODS

Adult catfish, Tut~kum tmlutm~, were collected by overnight gill netting. Fish used for laboratory experiments were collected from control areas (Ballina and Glenbawn Dam, NSW, Australia) where endosulfan had not been used. They were transported to Sydney (Australian in plastic bags tilled with water and oxygen (volume ratio 2:l) and placed in insulated containers. Ice was added to the containers to maintain the initial water temperature. Acriflavine (2-4 mg I-‘) and salt (0.5%) were added to the water during transport, For the investigation of the effect of endosulfan in the field. wild catfish were collected from cotton growing areas (Gwydir River, NSW, Australia), where endosulfan has been used, and from two control sites (Horton River and Pindari Dam), both at least 50 km upstream from the cotton growing area. The fish were caught during endosulfan application season (summer) by overnight gill netting. Gill filaments were dissected and fixed for Further processing immediately after capture.

The fish tanks were filled with Sydney mains water from which chlorine ramines were removed by fiftration through sand and activated-carbon lowed by holding for at least one week in 5000 1epoxy-lined concrete tanks. in the fish tanks was filtered through filter wool, activated-carbon and

and chloftfters folThe water clay pipes

67

contained in an external plastic filter (Sacem Marathon 700). The fish were acclimated to the laboratory conditions for at least two weeks before the experiments started. They were fed every second day ad lib&urn. The food (frozen chicken liver, live freshwater shrimp and earthworms) was free of endosulfan residues and had only traces of other organochlorines (DDEcO.01 mg kg-‘). Tanks were maintained under flow-through conditions at the rate of one tank a day. Dissolved oxygen and ammonia levels were monitored during the acclimation period and experiments. Dissolved oxygen was always above 75% saturation. No ammonia was detected. Two experiments were performed to evaluate the effect of an acute exposure to endosulfan. Endosulfan was applied as Thiodan in both experiments. The pesticide was applied once at the beginning of each experiment at a nominal concentration of 1.0 pg 1-l (measured mean concentration 0.99 pg I-‘, SE=0.04). The water flow and filtration were stopped for 24 h after application of the pesticide. Experiment 1 was performed as sta.tic 24 h exposure with a single application of endosulfan at the beginning of the experiment. There were two 50 1 tanks for each treatment, Two fish (fish size range 209429 mm) were ahocated to each tank using random selection procedures. The tanks were randomly assigned to control and treated. All fish were killed 24 h after the application. Experiment 2 investigated effects of a single exposure to endosulfan and time after exposure. There were three 250 1 tanks for each treatment. One fish (fish size range 380-532 mm} was allocated to each tank using random selection procedures. Confounding of the two factors (fish and tank) was unavoidable to prevent the effect of fish stress caused by aggression and territorial behaviour, experienced in the previous experiment. Three control fish and three treated fish were killed one day and twentyeight days after the exposure.

Gill filaments were collected from hve fish from the middle section of the second left gill arch for light microscopy. The time interval between removing the samples and placing them in fixatives was shorter than 10 s and there was no extra time with the dissection and fixation of the experimental versus control fish. The filaments were immediately fixed in ice-cold 1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 24 h (Saito and Tanaka. 1980). immersion was used as the fixation tecllnique. since samples of the gills from the same individuals were used for chemical analysis. This precluded the use of perfusion, which is the best fixation teclltliq~le. The samples were then rinsed three times with 6. i M cacodylate buffer and postfixed in 1% osmium tetroxide for I h. Tissue samples were dehydrated in a series of acetone. The filaments were placed in a mixture of 100% acetone:Spurr’s resin ( 1:1) and allowed to infiltrate overnight, The infiltration was completed by transferring them to fresh Spurr’s resin for 12 h. The following day the filaments were embedded in Spurr’s resin (Spurr, 1969). Thick sections (0.5-t .Opm) were cut with a glass knife using the ul-

tramicrotome (Ultracut E). The sections were attached to glass slides and stained with 0.5% toluidine blue in 1% borax. For endosul~n analysis the remains of the filaments were dissected immediateIy after the fish were killed. The tissue was kept frozen before analysis. Samples were analyzed individually from each fish. The tissue (3 g) was emulsified with trisodium citrate (3 g), disodium hydrogen orthophosphate (3 g) and mixed with sodium sulfate (I 5 g). The dry mixture was eluted through silicic acid (2.5 g) and alumina (3 g) with 100 ml of a mixture of di~h~oromethane, hexane and acetonitrile ~1:~:0.007). The eluate was chromatographed between standards on GLC (Shimadzu GC-SA, 0.53 mm x 30 m J&W DBS megabore column, carrier nitrogen, oven isothermal 195OC, inj/det 280°C) using EC detector. The method has a detection limit of I .Opg kg-’ for alpha and beta endosulfan and 2.5 ,ug kg-’ for endosul~n suffate giving mean recoveries of 86.2% for a-endosulfan, 83.2% for P-endosulfan and 88.8% for endosulfan sulfate (Nowak and Ahmad 1989). This detection limit was further lowered by concentration of sample to 1 ml with gentle nitrogen flow if traces of endosulfan were detected in uncon~entrated sample. The residues of total endosulfan were calculated as the sum of the residues of cr-endosulfan, p-endosulfan and endosuffan sulfate.

Meiistrreiwvivs The respiratory diffusion distance was considered as the shortest distance between the outer and inner membranes of tamellar epithelial cells at two randomly chosen points of each Iamellum. These measurements were performed using an image analyzer. All the measurements were carried out in the proximal part of lamellae (base and middle) from the middle section of each filament, since this part was shown to be more sensitive to certain toxicants, such as oil-shale retort water, than the distal section (Johnson, 1983). In experiment 1 and for the fish collected from wild populations, four lamellae were chosen from one filament from each fish using random selection procedures. Two replicate measurements of the respiratory diffusion distance were taken from each IamelIa. In experiment 2, four lame~~ae were chosen from each filament (two filaments from each fish), using random seiection procedures. Four replicate measurements of respiratory diffusion distance were taken using an image analyzer.

The experiments were designed for statistical analysis by ANOVA (Sokal and Rohlf, 1969; Underwood, 1981). In experiment 1, which investigated changes 24 h after pulse exposure to endosulfan, the effect of treatment on gill morphology was analyzed using four factor nested analysis of variance (ANOVA), with “treatment” as a fixed factor, “tank”, “fish” and “lamella” as random factors. Two groups of fish were compared; control and those treated with 1.Opg I-’ of endosulfan.

69

In experiment 2, which tested effects of treatment and time after treatment, a five factor mixed model ANOVA was used. Two fixed factors, time and treatment, both with two levels, were arranged orthogonally. “Fish” (three levels) was nested in each of these factors, whereas “filament” (two levels) was nested in “fish” and “‘lamel]a” (four levels) was nested in “filament”. The effect of endosulfan on the respiratory diffusion dist:lnce in the field was assessed by using three factor nested ANOVA. “Lamella” (;bur levels) was nested in “fish” (four levels) which in turn was nested in “population” (three levels - two control populations from Horton River and Pindari Dam and one population from Gwydir River surrounded by cotton growing areas where endosulfan had been used). The results were considered to be statistically significant if P < 0.05. If significant differences were found, the means were compared using Student-Newman-Keul test (SNK). Before ANOVA was performed homogeneity of variances was tested using Cochran’s test (Underwood, 1981). If the variances were heterogeneous the data were transformed and resubmitted to the test again. If despite transfo~ations the variances were heterogeneous, Kruskal-Wallis test was used followed by a posteriori nonparametric multiple comparison (Siegel and Castellan, 1988). If the lower factor was statistically nonsignificant (P r 0.25) the mean squares and degrees of freedom for this factor and residual were pooled and reanalyzed (Wine:. 1971). The variance components were estimated as the ~r~entage of the SS attributable to each level in sampling. The significance was estimated not only in the usual sense (probability of rejecting a true null hypothesis), but also, if appropriate, the pi&dbrlity of accepting a false null hypothesis was estimated. Beta probabilities of main effect were calculated for each nonsignifi~al~t result to determine if the null hypothesis of no effect could be confidently accepted. These calculations, using the observed values for means, error variance and degrees of freedom, were based on Cohen’s crffect size index and an alpha value of 0.05. Power of the tests was determined using tables for power ofFtest in ANOVA (Cohen, 1988). RESULTS

The morphology of gill filaments is similar among different fish species (Laurent and Dune], 1980). The basic structure of gills in control catfish was the same as that described for rainbow trout, ~~~~~~2~~~~~~tnykiss (Kendall and Dale, 1979). zebra~sh, ~~ffc~~~~~~zio rerio (Karlsson, 1983; Karlsson-Norrgren et al,, 1985), and bluegill sunfish, Lepornis nracrochirus (Richmonds and Dutta, 1989). The respiratory lamellae were covered by an epithelial layer, which was usually two cells thick. Internally the lamellar blood sinuses were lined and spanned by pillar cells. The filament

Fig. I. Gill filament from control catfish. F = filament, L = lamella, FE = filamental epithelium. LE = lamellar epithelium, RBC = red blood celts. Ear size SO$m.

between gill lameltae was covered by thick stratified epithelium. Mucous cells and chloride cells were scattered in the interlamellar epithelium. Equally spaced secondary Iamellae, intact ceiiular layers and no sign of fusion between neighboring Iamellae were considered characteristic for normal structure (Fig. 1). Slight epithelial lifting was observed in one of the controls. The respiratory diffusion distance was within the range for more sluggish fish species (Hughes, 1984). Even for gills of control fish there was a considerable variation in the quantitative diffusion characteristics. as has been previously observe@ an rainbow trout, ~~~~~~/~~~c/zz{s I~?~~~.~s {Hughes et al., 1979).

Oedema with lifting of Iamellar epithelium and hyperpl~sia of Iamellar epithelium (Fig. 2) were observed in the gills of all fish containing residues sf endosulfan (Table 1). Lamellar epithelium was often infiltrated by Ieucocytes. The mean respi~tory diffusion distance was increased in the fish containing endosulfan residues in their gills (Fig. 3). The treatment had a statistically significant effect on the respiratory diffusion distance (ANOVA, Table 2). “Fish” was a statistically significant factor, indicating a great variability among individuals. This could be at least partly due to confounding “fish” and “filament” factors by analyzing lamel-

71

Fig. 2. Gill filament from catfish containing endosulfan residues. L = lamella, CL = clavate lamella, FE = filamental epithelium, LE = lamellar epithelium. Bar size 50 ,um. (a) Lifting of epithelium 1 day after exposure to endosulfan; (b) epithelial hyperplasia. clavate lamella and lifting of epithelium 28 days after exposure to endosulfan.

TABLE I Residues of endosulfan and structural changes in gills of catfish after acute exposure to endosulfan (experiment I). Leucocyte infiltration

Exposure endosulfan cont. @g I-’ )

Endosulfan residues &g kg-‘)

Lamellar epithelium

alpha

beta

Lifting

Hyperplasia

0.1 0.1

5.0 0

0 0

0 0

+ -

+

+ -

0.1 0.1 I .o

0 3.0 6.0

0.5 0 3.0

0 4.0 11.0

+ + ++

+ + ++

+ + I

I .o 1.0

7.5 I .o

4.0 1.0

13.0 5.5

++ +++

++ +

+

I .o

1.5

I .o

10.0

+++

+

sulfate

Data for individual %h. Control fish not included. Degree of structural changes: +++ very common. ++ commo1I, + present, - absent.

12 c 10 -

Control

Treated Treatment

Fig. 3. Effect of acute exposure to endosulfan on respiratory diffusion distance (experiment I).

lae only from one filament from each fish. “Tank” as a factor was not statistically significant and when data were pooled the effect of treatment on respiratory diffusion distance was more pronounced (ANOVA, Table 2). There was a positive correlation between endosulfan residues in gills and the respiratory diffusion distance (IC0.8003, df=6, P < 0.05). Lamellar sinuses were slightly dilated i. most of the treated fish (Fig. 4) There was, however, no statistically significant difference in the diameter of lamellar sinuses

TABLE 2 Effect of 24 h exposure to I .O pug I ’ of endosulfan on respiratory diffusion distance Cum) in the gills of catfish (experiment 1). Source of variation Treatment Tank (treatment) Fish (tank (treatment)) Lamellae (fish (tank(t))) Error Pooled data (C = 0.6393) Treatment Fish (treatment) Lamellae (fishttreat)) Error

df 1.2

2,4 4.24 24,32

136 6,24 24,32

MS

F

r

%

437.65 1.35 26.18 5.96 5.55

323.85 0.05 4.40 I .07

0.00 0.95 0.001 0.42

50.55 0.3 I 12.1I i6.5.2 20.5 I

437.50 8.89 3.05 9.23

49.23 2.91 0.32

0.0004 0.03 0.997

Cochran’s test nonsignificant, C=O.2300 (k=2. df=31). Degrees of freedom for /;.

variation

73

Control

Treated Treatment

Fig. 4. Effect of acute exposure to endosuifan on diameter of lameliar blood sinuses (experiment 1).

between treated and control fish (ANOVA, Table 3), but the test was very weak (Table 4). The effect size index, however, indicated only slight degree of departure from the null hypothesis of no effect of endosu~fan treatment on the diameter of lamellar sinuses. Proliferation of chloride cells was present in two of the treated fish. Ail these changes were more pronounced in the catfish containing higher levels of endosulfan residues in their gills. Lifting of respiratory epithelium was the only exception, since it was more common in the fish containing lower residues of total endosulfan. The residues, however. consisted almost only of endosulfan sulfate whereas in other fish there was a considerable residue of alpha isomer (Table 1). Qualitative differences in the residues may be responsible for these different structural alterations in catfish gills.

TABLE 3 Effect of 24 h exposure to I .Opg I-’ of endosulfan on the diameter of lamellar blood sinuses in the gills of catfish (experiment 1f. Source of variation

df

MS

F

P

% variation

Treatment Tank (treatment) Fish (tank ~treatment~) Lamellae (fish(tank(t)~) Error

1.2 2,4 4,24 24,32

21.46 9.76 3.55 2.76

2.20 2.75 1.29

0.27 0.18 0.30 0.15

I I .86

1.48

1.86

Cochran’s test nonsignificant, C=O,5874 (k=2, df=31). Degrees of freedom for F.

10.79 7.85 3fid:l 32.89

74 TABLE 4 Effect size index and beta probability. Time

Treatment

Lamellar blood sinuses Experiment 1 Experiment 2 Respiratory diffusion distance Wild catfish

ES

B

ES

B

0.26 0.11

0.93 0.89

NA 0.02

NA 0.95

0.34

0.86

NA

NA

NA. effect not tested.

Effect

oftime uftertxposure

on I/W structure

ofgilis

Oedema with lifting of lamellar epithelium and hyperplasia of lameilar epithelium were observed in the gills of all fish containing residues of endosulfan. irrespective of the time after exposure (Table 5). Lamellar epithelium was often infiltrated by leucocytes. The mean respiratory diffusion distance in gills of the treated fish was increased more than twice that in the controls (Fig. 5). This effect was the same after 28 days as it was 1 day after the exposure. The variances for the respiratory diffusion distance were significantly heterogeneous despite different transformations. A nonparametric test showed statistically significant difference between treatments (Kruskal-Wallis test, P < 0.00001). There was a statistically significant difference in the respiratory diffusion distance between the control and treated fish 1 day after exposure, as well as

TABLE 5 Residues of endosulfan and structural changes in gills of catfish one day and twenty eight days after acute exposure to endosulfan (experiment 2). Days after exposure 1 I 1

2X 28 28

Residues of endosulfan (~8 kg-‘)

Lamellar epithclium

alpha

beta

sulfate

Liftit;g

Hyperplasia

50.0 76.0 159.5 185.0 191.0 206.0

i 7.0

0 22.0 0 40.0 36.0 43.0

++ +++ + -1. + +-r-r

+ ++i-

19.0 n 75.0 69.0 63.0

Lcucocyte infiltration

++ ++

4 + -F ++ 1--_-

++ + /-

Data for individual fish. C’zmtrol fish not included. Dcgrcc of changes: +++ very common, ++ common. + present, - absent.

75

4

2

0 C day 1

Tdayl

C day 28

T day 28

Treatment Fig. 5. Effect of time after exposure on respiratory diffusion distance (experiment 2. C = control, T = treated).

between control and treated 28 days after exposure (P < 0.05, a posteriori nonparametric test: Siege1 and Castellan, 1988). While there was no statistically significant difference between both control groups, the respiratory diffusion distance of the treated fish 1 day after exposure was significantly different from the distance of the treated fish 28 days after exposure. In contrast to the results in experiment 1, the mean diameter of lamellar blood sinuses was slightly decreased in treated fish compared to control (Fig. 6j. This difference was not statistically significant (Table 6). but the test was too weak to confirm

2 0 C day 1

Tday

1

C day 28

T day 28

Treatment I-lg. h. Effect

oftimc alicr

exposure on diamctcr of lamcllar blood sinuses (cxpcrimcnt 2. c‘ = control. T = wcalcd).

76

that treatment with endosulfan did not affect the diameter of lamellar blood sinuses (Table 4). The degree of departure from the null hypothesis was not very large as indicated by the effect size index. Structural changes in gills, such as epithelial hyperplasia and oedema with lifting of Iame]lar epithelium were more pronounced in the individuals in which infiltration of leucocytes was not very common (Tables 1 and 5). On the other hand, alterations in the gills of fish which showed high degree of leucocyte infiltration were not as severe as in the other fish. The respiratory diffusion distance was more related to the relative residues of endosulfan isomers than to the residues of total endosulfan in the gills (Fig. 7a,b), although this relationship was not significant. probably due to low numbers of rephcates. This indicates that structural response of the gills rn~: be isomer-specific. The effect of the composition of residues on the structural changes in gills should be further investigated. The respiratory diffusion distance was related to the size of fish, measured as length (Fig. 8). This relationship was better reflected for the control fish. Due to more variation and low number of replicates, the respiratory diffusion distance for treated fish did not show such a clear relationship. Sttwture ofgills in wildfish c’ollecteci in the cotton growing ureu unciin r.ontr’olsites

The mean respiratory diffusion distance was larger for the catfish collected from the cotton growing area than from control sites although this difrerence was not statistically significant (ANOVA, Table 7, Fig. 9). This difference WGSnot as drastic as in the experimental fish. Additionally, the power of the test was lo\?.er th;;n desired due to low number of replicates and high individual variability (Table 4). The effect size index indicated high degree of departure from the null hypothesis of no difference in the respiratory diffusion distance between fish from different populations. There TABLE 6 Effect of time after exposure to I .Opg I-’ of endosulfan on the diameter of lamellar blood sinuses in the gills of catfish (experiment 2). Source of variation

df

MS

F

P

% variation

Time Treatment Fish (treat, time) Filament (fish) Lamella (filament) Time x treatment Error

I,8 I,8 8,8 8,24 24.148 138

0.35 85.17 53.47 39.09 7.21 1.35 I.18

0.00 1.59 1.37 5.42 6.1 I 1.14

0.94 0.24 0.33 0.0006 0.00 0.32

0.03 7.54 37.86 27.67 1I .29 0.12 15.49

Cd-mm’s

tCStnonsignificant. CO.2696 (k=4, df=47). Degrees of freedom for F.

77 la)

13 9

9.75 i ?

3.25

I

1

0

I

I

I

116.7 233.3 Totalendosulfan(pgkg-')

0

(b’ 13

350

R=0.7169

12

+

11 +

T = 10 -0 e 9

:i,_ -0.1

0.3

0

0.4

pO;l(a + sulfaPep

Fig. 7. Relationship between residue of endosulfan in gills and respiratory diffusion distance (experiment 3). (a) Total residue of cndosulfan: (b) relative residue of endosulfan isomers.

was a statistically significant difference between individual fish, confirming this great individual variability (Table 5). The relationship between respiratory diffusion distance and length was evident for catfish from the Gwydir River and Pindari Dam (Fig. 10). No such relationship could be shown for catfish from the Horton River, at least partly due to smaller size range and low number of replicates. Epithelial lifting, dilation of blood sinuses, vascular congestion and epithelial rupture were present in gills of most fish collected in the field at both control and cotton-

78

-

control

R=O.8168

-

treated

R=O.5869

I

I

I

I

31.75

38.5

45.25

52

12 -

421 25

Length (cm) Fig. 8. Relationship between fish length and respiratory diffusion distance (experiment 2).

growing sites. Sampling conditions were not as optimal as during laboratory experiments. Some of the catfish were in the net for a few hours and it was difficult to avoid overcrowding in the transport container for the few hours between release from the net and dissection. All these factors could result in structural changes in fish gills. It must be emphasized, however, that these conditions were the same at all sampling sites, so any detected structural differences between the wild fish should be due to the effect of site and not sampling technique. DISCUSSION

The changes found in the catfish, Tund~nus tandarzus, after exposure to endosulfan were more typical of response to organic pollutants than other irritants (for review see Mallatt, 1985; Evans, 1987). TABLE 7 Effect of population on respiratory diffxsion distance. Source of variation

df

MS

Population Fish (population) Lamellae (fish) Error

2-9 9,36 36.48

% variatian

F

P

0.175

1.55

0.26

11.91

0.113 0.02 1 0.017

5.38 1.24

0.000 1 0.24

34.61 25.72 27.76

Data transformed as log (x+1). Cochran’s test nonsignificant for transformed df=Degrees of freedom for F.

data, GO.3934 (k=3,

79

5

z4 J s3 2 1 0 Horton River

Pindari Dam

Gwydir River

Population

Fig. 9. Respiratory diffusion distance in the gills of wild catfish (Horton River and Pindari Dam -~control populations. Gwydir River - population from cotton growing area).

Oedema with separation of lamellar epithelium from the underlying basement membrane is an acute lesion (Ferguson, 1989). It was present in most of the catfish containing residues of endosulfan. Epithelial lifting has been previously described in the freshwater fish, Channa gachua, exposed up to 32 days to various concentrations of endosulfan, ranging from 3.5 to 10 ,ug 1-l (Dalela et al., 1979); in rainbow trout, Oncorhynchus mykiss, surviving exposures to 7.5 to 17.5 pug 1-l of heptachlor for two weeks (Wood in Eller 1975); in goldfish, Curassius aurutus, exposed to 100 pg I-’ of mirex for 56 days (Van Valin et al., 1968); in tidewater silverside, Menidia beryllinu, and hogchoker, Trinectes macufutus, after 7 to 30 days exposure to 100 mg I-’ of crude oil or 50% of water soluble fraction (WSF) (Solangi and Overstreet, 1982j; in cainbow trout, Oncorhynchus mykiss, after both aqueous and dietary exposure for 20 days to sublethal levels of permethrin (Kumaraguru et al. 1982); in Atlr,.nt’ic salmon, Salrno sakur, exposed to tributyltin-treated nets for 4 days (Bruno and Ellis, 1388); and in bluegill sunfish, Lepomis nzarrochirus, exposed to 50 ,ug 1 ’ of malathion for up to 4 days (Richmonds and Dutta, 1989). Hyperplasia of lamellar epithelium with fusion of the neighboring lamellae, observed in the catfish containing residues of endosulfan, is a chronic response (Ferguson, 1989). It has been reported previously in goldfish, Curassius uumfus, exposed to 1.O mg 1-l of mirex (Van Valin et al., 1968); tidewater silverside, Menidia ber_dliw and hogchoker, Trinectes nzaculatus, exposed for 15 days to 100 mg I-’ of crude oil or 50% of WSF (Solangi and Overstreet, 1982); cutthroat trout, Oncorhynchus clarki, exposed to 39 pg 1-l of oil for 90 days (Woodward et al., 1983); and rainbow trout, Oncorhynchus mykiss, after 20 days of aqueous or dietary exposure to sublethal levels

8

*

-

Pfndarf

ham

R=O.8958

-

Gwydir

River

MO.

-

Horton

River

R=O.O425

8240

7

25

3Q

35

40

45

50

Length (cm)

Fig. 10. Relationship between fish length and respiratory diffusion distance (wild catfish).

of permethrin (Kumaraguru et al., 1982). Epithelial hyperplasia was the characteristic alteration in the gills of channel catfish, Ic~alurus punc~utus, and rainbow trout, OnCY.W~~_WZ~ZUS mykiss, after 16 h exposure to two coal-derived materials; one of which contained PAHs and the other nitrogen heterocycles (Stoker et al., 1985). Lamellar fusion was observed in the freshwater fish, Chnnu gaclzucr, exposed to 3.5 to 10.0 ,ug I-’ of endosulfan (Dalela et al,, 1979). The respiratory diffusion distance in gills of catfish was significantly increased after treatment with endosulfan. This was a consequence of either Iifting of lamellar epithelium, frequently encountered in the treated fish, or hyperplasia of epithelium. An increase in the respiratory diffusion distance has been previously reported in rainbow trout, Oncoriz_ynchus mykiss, exposed to 0.09% of oil-shale retort water for 60 days (Johnson, 1983) and in rainbow trout exposed for 21 days to 2.Opg I-’ ofchlorothalonil (Davies, 1984). In the former case it was mainly due to the swelling of lamellar lymphatic space, infiltration of wGte blood cells and hypertrophy of epithelial cell, whereas in the latter it was a consequence of hyperplasia of lamellar epithelium. Lifting of epithelium can often be caused by tissue processing techniques (Gardner, 1975). ~pithelial capillary separation and epithelial cell hypertrophy of the gill lamellae of rainbow trout, O~~~~~~/~~~~~/?~~S ~~z~k~ss,were described as postmortem artifacts, induced by fixation delays of more than 20 s and use of formahn for fixation instead of Bouin’s (Speare and Ferguson, 1989). In this study, the fixation delay for the filaments from catfish was shorter than 10 s. Single filaments were fixed by immersion in glutaraldehyde. No information is available on the effects of the use of glutaraldeHyde as a fixative. Nevertheless, since the filaments from all catfish were fixed in

81

exactly the same way, any changes due to fixation procedures should be present in both treated and control fish. Pathological changes in gills have been used in the field as indicators of pollution. Severe hyperplasia and moderate hypertrophy of epithelial cells, clubbing and fusion of the secondary lamellae and moderate to severe oedema in the secondary lamellae were present in the gills of channel catfish, Ictalurus punctatus, caged below wastewater treatment plant (Mitz and Giesy, 1985). In comparison the catfish caged above this plant had only slight hyperplasia and hypertrophy of the lamellar epithelium. Although there was a slight increase in the mean respiratory diffusion distance in the gills of catfish from the cotton growing area compared to controls, all the values fell within the range for control fish from laboratory experiments. This could be partly due to differences in sizes of individual fish, The mean length was smaller for the fish from cotton growing area than for the fish from control sites. To reduce individual variability fish of the same size should be collected, as size of an individual affects respiratory drft’usion distance (Hughes et al,, 1979). In the fic!rl, in contrast to the laboratory experiments, the time after exposure and exposure frequency was not necessarily equal for the catfish from cotton growing areas, further increasing natural variability. Higher number of replicate fish as well as more sampling sites would be desirable in the field study. Most of the pathological changes in fish gills described in the literature were not quantified thus reducing their value and preventing any statistical analysis. Respiratory diffusion distance was one of the few exceptions. The alterations in the structure of catfish gills in the laboratory experiments in this study seemed to be related not only to the residue level but also to qualitative composition of the residues. Further investigation is necessary to establish if residues of alpha and beta isomers induce different biological responses. The respiratory diffusion distance wa:: relatively consistent in the control fish from both experiments and in catfish collected from the fiefd. A;l&tionally, the increase of respiratory diffusion distance induced by laboratory exposure to endosulfan was identical in both experiments. Furthermore, respiratory diffusion distance in the gills of rainbow trout, Orz~o~lz~n~Bus rrykiss. was previously shown to be more sensitive to oil-shale retort water than either fish survival or growth (Johnson, 1983). Hence. this variable could be a useful indicator of pollution. ACKNOWLEDGEMENTS

This research was financially supported by the University of Sydney, State Pollution Control Commission, Hoechst Ltd. and International Federation of University Women. Experiments and chemical analysis were performed at the Centre for Environmental Toxicology at the University of Technology. Sydney. Histology and image analysis were done at Electron Micrascope Unit at the University of Sydney. 1 would like to thank Dr J.S. Langdon for his critical comments on the manuscript.

82 REFERENCES Bruno, D.W. and A.E. Ellis, 1988. Histopathoiogicai effects in Atlantic salmon, Saltno sakar L., attributed to the use of tributiitin antifoulant. Aquaculture 72, 1.5-20. Cohen, J., 1988. Statistical power analysis for the behavioral sciences. L. Eibraum Associates, Hillsdale, N.J. Da]e]a, R.C., M.C. Bhatnagar, K. Tyagia and S.R. Verma. 1979. Histological damage of gills in Chwna g&w after acute and subacute exposure to endosuifan and rogor. Mikroskopie 35,3Oi-307. Davies, P.E., 1984. Chiorothaionii. Its environmental fate, toxicology and metabolism in fish. PhD thesis, University of Tasmania, Hobart, Australia. Eiier, L.L., 1975. Gill lesions in freshwater teieosts. In: The pathology of fishes, edited by W.E. Ribelin and G. Migaki, The University of Wisconsin Press, 305-330. Evans, D.H.. 1987. The fish gill: site of action and model for toxic effects of environmental pollutants. Environ. Health Perspect. 71,47-58. Ferguson. H.W.. 1989. Systemic pathology of fish. Iowa Sta!e University Press, Ames Iowa Gardner, G.R.. 1975. Chemically induced lesions in estuarine and marine teieosts. Ii: The pathology of fishes, W.E. Ribeiin and G. Migaki eds.. The University of Wisconsin Press, 657.-693. Goebel, H., S. Gorbach. W. Knauf, R.H. Rimpau and H. Hoettenbach. 1982. Properties, effects. residues and analytics of the insecticide endosulfan, Residue Rev. 83, I - 174. Hughes. G.M., 1976. Polluted fish respiratory physiology. In: Effect of pollutants on aquatic organisms, edited by A.P.M. Lockwood. Cambridge University Press, 163-185. Hughes, G.M.. 1984. General anatomy of the gills. In: Fish physiology, edited by W.S. Hoar and D.J. Randall. Academic Press. vol. X. Part A, i-72. Hughes, G.M.. SF. Perry and V.M. Brown, 1979. A morphometric study of effects of nickel, chromium and cadmium on the secondary iameiiae of rainbow trout gills. Water Res. 13, 665 679. Johnson, R.D., 1983. The use of quantitative histopathoiogy in aquatic toxicology: correlations between gill structure and routine toxicity measures in rainbow trout. PhD thesis. University of Wyoming, USA. Karisson. L.. 1983. Gill morphology in the zebrafish. Brac/ydattio wriu (Hamilton-Buchanan), J. Fish Bioi. 23.5 I i-524. Karisson-Norrgren. L., P. Runn, C. Haux and I.. Foeriin. 1985. Cadmium induced changes in jii morphoiogy of zebrafish. Brdtwianio rerio (Hamilton-Buchanan). and rainbow trout, Saltno gairdtwri Richardson. J. Fish Bioi. 27, 81-95. Kendall, M.W. and J.E. Dale, 1979. Scannit?: and transmi&.m eiectron microscopic observations of rainbow trout Ckitno gairdncri) gill. J. Fish. Rcs. Board Can. 36, iO72- 1079. Kumaraguru. A.K.. F.W.H. Beamish and H.W’. Ferguson, 1982, Direct and circulatory paths of Permethrin (NRDC-143) causing histopathoiogicai changes in the gills of rainbow trout, Saltno gairclttrri, Richardson. J. Fish Bioi. 20. 87-91. Laurent, P. and S. Dune]. 1980. Morphology of gill epitheiia in fish. Am. J. Physioi. 238, Ri47-Ri59. Maier-Bode. H.. 1968. Properties, effect, residues and analytics ;f insecticide endosuifan. Residue Rev. 22, i--14. Mailat, J.. 1985. Fish gill structural changes induced by toxicants and other irritants: a statistical review. Can. J. Fish. Aquat. Sci. 42, 630.-648. Mitz S.V. and J.P. Giesy, 1985. Sewage efiiuent biomonitoring. I. Survival. growth and histopathoiogicai effects in channel catfish. Ecotox. Environ. Safety IO, 22-39. Now& R. and N. Ahmad, 1989. Residues in fish exposed to sublethal doses of ondosuifan and fish collected from cotton growing area. J. Environ. Sci. Health B24,97- 109. Richmonds. C. and H.M. Dutta. 1989. Histopathologicai changes induced by ma,:.tthion in the g;:!s of biuegiil I.epotttis tnurrdtirus. Bull. Environ. Cczram. Toxicol. 43, 123-i 30. Saito. Y. and Y. Tanaka. 1980. Giutaraidehyde fix;.rtion of fish tissues for electron r;lieroscopy. .I. Electron Microsc. 29, l-7.

83 Siegel, S. and N.J. Castellan, 1988. Nonparametric statistics for the behavioral sciences. McGraw-Hill Book Company, Singapore. Sokal, R.R. and F.J. Rohlf, 1969. Biometry. R. Emerson, D. Kennedy, G.W. Beadle and D.M. Whitaker eds., W.H. Freeman and Co, San Francisco, 776 pp. Solangi, M.A. and R.M. Overstreet, 1982. Histopathological changes in two estuarine fishes, Menidiu beryllina (Cope) and Trinectes omulatus (Bloch and Schneider) exposed to crude oil and its watersoluble fractions. J. Fish Dis. 5, 13-35. Speare, D.J. and H.W. Ferguson, 1989. Fixation artifacts in rainbow trout (S&to gairdneri) gills: a morphometric evaluation. Can. J. Aquat. Sci. 46, 780-785. Spurr, A.R., 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26,781-783. Stoker, P.W., J.R. Larsen, G.M. Booth and M.L. Lee, 1985. Pathology of gill and liver tissues from two genera of fishes exposed to two coal derived materials. J. Fish Biol. 27, 31-46. Underwood. A.J., 1981. Techniques of analysis of variance in experimental marine biology and ecology, Oceanogr. Mar. Biol. Ann. Rev. 19, 513-605. Van Valin, C.C., A.K. Andrews and L.L. Eller, 1968. Some effects of mirex on two warm-water fishes. Trans. Am. Fish. Sot. 97, 185-196. Winer, B.J., 1971. Statistical principles in experimental design. McGraw Hill Kogakusha. Tokyo. Woodward. D.F., R.G. Riley and C.E. Smith, 1983. Accumulation, sublethal effects, and safe concentration of a refined oil as evaluated with cutthroat trout. Arch. Environ. Contam. Toxicol. 12.455-464.