Toxkoo VoL 30, No. 3, pp. Printed in Great Britain.
259-264, 1992.
aa1-0101/92 s5.00+ .00 C 1992 Pers mon Prey plc
SURVIVAL TIMES AND RESISTANCE TO SEA SNAKE (AIPYSURUS LAEVIS) VENOM BY FIVE SPECIES OF PREY FISH K . D . ZIMMERMAN,' HAROLD HEATWOLE2*
and
H . 1. DAMES'
'P .O. Box 1138, Umm al Quwain, United Arab Emirates; 'Department of Zoology, North Carolina State University, Raleigh, NC 27695-7617, U .S.A . ; and 'Department of Mathematics, University of New England, Armidale, N.S .W ., 2351, Australia (Received
3
home
1991 ;
accepted
31
October
1991)
K. D . ZIMMERMAN, H . HEATWOLE and H . I . DAVIES . resistance to sea snake (Aipysurus Idevis) venom by five
Survival times and species of prey fish . Toxicon 30, 259-264, 1992 .-The LD50 values and survival times of three pomacentrid species and two blennies were measured after being subjected to the venom of one of their predators, the olive sea snake, Aipysurus Idevis . The species differed significantly in the speed of their responses to the venom. At high venom doses, blennies had higher survival times than pomacentrids, and in the latter group Ddscyllus survived longer than did species of Chromis. In part, this may be related to differences between blennies and pomacentrids in degree of cutaneous respiration. Relative survival- time was influenced by venom dosage; ranking order of species' survival times was different at low doses than at high ones, and taxonomic correlations broke down. INTRODUCTION RESPONSE To venom can be expressed either as resistance or as survival time . These differ in their ecological consequences in the predator-prey relationship . Resistance, indicated in the present study by LD 50 values and probit lines, is reciprocal to toxicity . From the standpoint of the prey, resistance is the important concept as it is a measure of whether or not the prey will survive a particular envenomation . From the standpoint of the predator, toxicity is the main consideration; it indicates whether venom is strong enough, or administered in sufficient quantities, to subdue a particular animal . Survival time, on the other hand, does not necessarily define the ultimate outcome of an envenomation, but rather the speed at which that outcome is achieved. For example, two venoms might both produce the same percentage mortality eventually, but one may do so rapidly and the other slowly. Survival time of prey is important to the predator. The longer the survival time, the greater is the chance the prey will escape or cause damage through counterattack. This consideration may be less important to the prey . Ifit does not survive a given envenomation it is trivial whether the speed of demise is rapid or slow, or even whether it escapes during that period if it merely succumbs anyway after it is free. Author to whom correspondence should be addressed . TOX 30 :3-C
259
260
K. D. ZIMMERMAN et al.
In the see-sawing evolutionary race between predator and prey, selection will favour prey with high resistances, and predators with highly toxic, rapidly acting venom. Snake venom probably evolved within such a feeding context, rather than in a defensive one, and its deterrent role against enemies would seem to be a secondary consequence of its primary function of preventing prey escaping or counterattacking (HEATWOLE, 1987). Despite this, most studies of sea snake venom have focused upon either its toxicity to laboratory animals or its clinical effects upon humans (SUTHERLAND, 1983). As valuable as these studies are in other regards, they reveal little about the effectiveness of venom in the interaction of snakes and their prey and have little bearing either on the life of snakes in their native environment or on the resistance of prey to attack . The present paper is a step toward redressing that situation by examining the responses of five species of fish to the venom ofAipysurus laevis, a sea snake known to prey upon them (McCosKER, 1975 ; VORM and VoRis, 1983; Bums, unpublished Ph .D . thesis). MATERIALS AND METHODS The prey selected were three pomacentrids, Chromis nitida, Chromis atriwctorahs and Dascylhu armanus, and two blennies, lstibknnius meleagris and Isttbknnius edentuhas. All species were about the same size, with a total range in body wt of 5-10 g. They were collected at Heron and Wistari reefs on the Great Barrier Reef, Australia. The Chromis species were captured by anaesthetizing with quinaldine (GnisoN, 1967), D. armanus by corralling them in small coral heads, which were then brought to the surface forfish removal, and the Istiblavdus speciesby hand-collecting from under rocks and dead coral during low tides at night. Experimental observations were carried out in an aquarium divided into seven compartments, each 12 x 10 cm and with its own independently controlled air supply . Six compartments were used for observing fish, one per compartment . The seventh compartment contained a filtration system serving the entire aquarium. Venom was collected from Aipysurus laevis from the Swain Reefs, Great Barrier Reef, lyophilized and stored at about 0°C until required. Prior to injection it was dissolved in 0.9% physiological saline. Serial dilutions of venom were used until no mortality occurred among injected fish. A further requirement was that survivorship should be between 16 and 84% for at least four dosage levels . For each dilution, 12 fish of each species were injected with a venom solution in the epaxial muscle of the caudal peduncle using an Agla micrometer syringe. Doses were in the range of 0.001-15.5 mg dry wt of venom per kg fresh body wt of fish. The fish were observed for 48 hr, a sufficient time for action of the venom to be manifested (CHRISTENaN, 1968); times of deaths were recorded . Controls (uninjeeted fish) and saline controls (fish injected only with physiological saline) also were observed for 48 hr to detect any effect of handling procedures or holding facilities . Resistance was modelled using probit analysis (FINNEY, 1971) with log dose being used as the independent variable. LDw values and their fiducial confidence intervals were calculated and then back-transformed to the original scale. The adequacy of the model was assessed through a test of heterogeneity and the probit lines compared for equality of slope (parallelism). Parallel lines were further tested for equality of intercept. The logistic curve was used to model mean time to death against venom dose. Calculations for probit and logistic curves were carried out using the Maximum Likelihood Program (MLP) (Ross, 1980). RESULTS
Controls and saline controls maintained good health throughout the experiments. There were considerable differences in resistance to venom. Dascyllus aruaruts was the most resistant, showing a more gradually sloping mortality curve (Fig. 1) and a higher LD50 (Table 1) than the other species. Following in descending order of resistance were: I. edentulus, C. nitida, I. meleagris and C. atripectoralis . Thus, resistance is not strictly along taxonomic lines. The probit model appeared adequate in all cases (tests of heterogeneity were all nonsignificant; P > 0.05). Tests for parallelism (Table 2) showed there were no significant differences among the slopes of the probit lines for C. nitida, D. aruanus and I. meleagris,
Sea Snake Toxicity
26 1
100
Dose (MQikfl) FIG. 1. MORTALITY OF PREY SPECIES OF FM AT DU+FMNT LEVELS OF ENVENOYATION.
Each curve represents 10 computer-cakUlated points equally spaced between 0 and 90% mortality . Circles: D. aruanw, dots: C. nitida; Xs : C. atripectoraAr, triangles : I. nttkagris; squares : I. edeitulua.
or of the lines for I. meleagris, I. edentulus and C. atripectoralis. However, the probit lines for C. nitida (P < 0.01) and D . aruanus (P < 0.05) were not parallel with the line for C. atripectoralis. Only parallel lines were tested for differences in position, with only C. atripectoralis and I. meleagris appearing to have the same intercepts . These relationships are also evident from the shifts of the curves in Fig. 1 . Comparisons of 1.D ., values (Tables 1, 3) show significant differences between C. nitida and D . aruanus (P < 0.05) and between members of the following pairs (P < 0.01): C. nitida and I. meleagris, D . aruanus and I. meleagris, C. atripectoralis and I. edentulus and I. edentulus and I. meleagris. Only C. atripectoralis and I. meleagris exhibited no significant difference in LD5 0 values . The highest dose acted rapidly on all species, with the two Chromis species showing the quickest response and the two Istiblennius species reacting the slowest (Fig. 2). At lower doses, I. meleagris demonstrated a relative change in position and was the quickest to react to the venom. Of the three pomacentrids, D . aruanus was slowest to react at all doses, whereas C. atripectoralis (the least resistant) reacted most quickly. There were no significant differences in the slopes of the probit lines between C. nitida, C. atripectoralis and D. aruanus (P > 0.05). This is also evident from their similar rates of decrease in survival time with increasing venom dose (Fig. 2). Istiblennius meleagris and I. edentulus TAa1B 1. P~ ANALYam OF To ncm of Aipywua laevit vE~ To Fm
sp~ OF MW FM
Spceis Cbromis nitida Chromis atrlpectoralis Dawylba arwaw lat"nnbu mekagris Istiblemlus edentuba
ID» (mg kg- ')
Slope
5% Fiducial limits
0 .19 0 .09 0 .27 0.09
-11 .9 -9l .7 -7.8 -47 .6 -19.2
0 .24-0 .16 0 .09-0 .08 0 .38-0 .22 0 .13-0.08 0 .27-0 .22
025
262
K. D. ZIMMERMAN et al.
T>jem 2. Tests PoR tearnaoae«rrr, PARALLEL~ Am POmTION OF PR~ I~ Pm Chrondu nitida, Chromis atripectorali3, Dascy1hu aruanus, IstibLmrius neeL~ AND Istible~ edentuhu Species compared C. nitids C. atripectoralis D. aruanus I. nwLogris I. edentuhrs
Heterogeneity d.f.
,e
X2
Parallelism d.f.
11 .59
22
14 .81**
4
C. nitida D. aruamu I. meleagris I. edbetulus
9.64
17
6.09
3
C. nitids C. atripectoralis
6.14
9
10 .48**
1
C. nitids D. ar~
5.66
9
0.16
C. nitids I. meleagris
6.82
8
C. nitida I. edentulus
5.54
C. atripectoralis D . artumus
X2
Position
d.f.
40.18**
3
1
7.95**
1
1 .94
1
11 .91**
l
8
4.74*
1
3.42
10
9.00**
1
C. atripectoralis I. nieleagris
4.58
9
2.56
1
0.52
1
C. atripectoralis I. edentulus
3.30
9
1 .95
1
44 .09**
1
4.10
9
1 .35
1
26 .63**
1
I. edent~
2.82
9
3.50
1
0.08
1
I. nmeieagris I. edentuhrs
3.98
8
0.11
1
29 .34**
1
D . aruamAs
I. mekagris D . aruanus
d.f. degrees of freedom; X' chi-square . *P 0.05 NS
C. atripectoralis !. edentuha
20.80
< 0.01
nteleagris I. edattuha
10 .79
< 0.01
L
NS - not significant.
tion with taxonomic category . In order of decreasing survival time the ranking at high doses is: I. edentulus, I. meleagris, D. asanus, C. nitida and C. atripectorahs. The blennies survived longer than the pomacentrids, and among the latter Dascyllus survived longer than the Chromis species. However, relative survival times among species depended on dose. At some lower doses, for example, I. meleagris had the shortest, rather than the longest, survival time, and in general taxonomic correlations broke down . Differences among species in response to venom may reflect differences in habitat and life styles. Of the two Istiblennius species, I. edentulus, the most resistant one, often was
i0" 10
102 10
0.1 0.001
t
0.01
t
~a
0.1
1
t 10
Dose Ftm 2. FsrFecT of SEA wAn vErrnr nose oN mEAN Tun To DEATH 1N raev sPBcms of Ftsx. Solid lines indicate pomaeentrids; dotted lines indicate gobies. Other symbols as in Fig. 1.
264
K. D. ZIMMERMAN et al.
isolated at low tide under rocks and coral in small pools away from the water's edge . Blennies have Cuaneous respiration (GRAHAM, 1976) and low oxygen requirements dictated by a secretive life style (FRY, 1957; Moyu and CECH, 1982 ; HUGHES, 1984). Both of these features would allow for longer survival in the face of breakdown of the ventilatory system, one of the consequences of sea snake envenomation (ZIMMERMAN et al., 1990). This may be why blennies had longer survival times at high doses than did pomacentrids . Once gill ventilation ceased, however, they survived only a while longer then the others, and death was inevitable . Cutaneous respiration is insufficient, on its own, to maintain adequate oxygen levels indefinitely while in the water (Moyu and CECH, 1982), even though it may allow prolonged existence out of water (NORMAN, 1975). It may be that the mechanism for death differs at different doses; perhaps neurotoxic components are more important at high doses, with other, slower-acting toxins becoming operative at doses below the lethal limits of neurotoxins. Acknowledgements-We are grateful to the University of New England for support of the project, to the Heron Island Research Station for facilities and equipment and to Dr Neomm PotuN for critical review of the manuscript. REFERENCES GH;usTw=N, P. A. (1968) The venoms of Central and South African snakes. In : Venomous Animals and their Venoms, Vol. 1, pp. 437-461 (BucHtt., W., BucrLEY, E. E. and Dauram, V., Eds) . New York : Academic Press. Fa4NEY, D. J. (1971) Probit Analysis, 3rd Edn. Cambridge: Cambridge University Press. FRY, F. E. J. (1957) The aquatic respiration of fish. In : The Physiology ofFishes, pp. 1-64 (BaowN, M. E., Ed .) . New York: Academic Press. GmwN, R N. (1%7). The use of the anaesthetic Quinaldine in fish ecology. !. Anlm . Ecol. 36, 295-301. GtuHAK J. B. (1976) Respiratory adaptations of marine air breathing fishes . In : Respiration of Amphibious Vertebrates, pp. 165-187 (Huatms, G. M., Ed .) . New York : Academic Press. HEATwoiB, H. (1987) Sea Snakes. Sydney: The New South Wales University Press. Hucttms, G. M. (1984) General anatomy of gills. In : Fish Physiology, Vol. 10, pp. 1-72 (HoAa, W. S. and RAmAm, D. J., Eds). New York : Academic Press. McCossmre, J. E. (1975) Feeding behaviour of Indo-Australian Hydrophiidae . In: The Biology of Sea Snakes, pp . 217-232 (DurnoN, W. A., Ed.). Baltimore: University Park Press. MaYLA P. B. and Cixz JA, J. (1982) Fishes: an Introduction to Ichthyology. Englewood Cliffs, NJ : Prentice-Hall . NoamAN, J. R. (1975) A History of Fishes . London: Ernest Bann. Ross, G. J. S. (1980) Maximum Likelihood Program. Harpenden, Hertfordshire, Lawns Agricultural Trust, Rothamated Experimental Station. SurmaRLAND, S. K. (1983) Australian Animal Toxins. New York : Oxford University Press. Wan, H. K. and Wan, H. H. (1983) Feeding strategies in marine snakes : an analysis of evolutionary, morphological, behavioural and ecological relationships. Am. Zool. 23, 411-425. ZnoatmAN, K. D., GATES, G. R. and HEATwor.E, H. (1990) Effects of venom of the olive sea snake, Aipysurus laevis, on the behaviour and ventilation of three species of prey fish . Toxicon 28, 1469-1478.
259-264, 1992.
aa1-0101/92 s5.00+ .00 C 1992 Pers mon Prey plc
SURVIVAL TIMES AND RESISTANCE TO SEA SNAKE (AIPYSURUS LAEVIS) VENOM BY FIVE SPECIES OF PREY FISH K . D . ZIMMERMAN,' HAROLD HEATWOLE2*
and
H . 1. DAMES'
'P .O. Box 1138, Umm al Quwain, United Arab Emirates; 'Department of Zoology, North Carolina State University, Raleigh, NC 27695-7617, U .S.A . ; and 'Department of Mathematics, University of New England, Armidale, N.S .W ., 2351, Australia (Received
3
home
1991 ;
accepted
31
October
1991)
K. D . ZIMMERMAN, H . HEATWOLE and H . I . DAVIES . resistance to sea snake (Aipysurus Idevis) venom by five
Survival times and species of prey fish . Toxicon 30, 259-264, 1992 .-The LD50 values and survival times of three pomacentrid species and two blennies were measured after being subjected to the venom of one of their predators, the olive sea snake, Aipysurus Idevis . The species differed significantly in the speed of their responses to the venom. At high venom doses, blennies had higher survival times than pomacentrids, and in the latter group Ddscyllus survived longer than did species of Chromis. In part, this may be related to differences between blennies and pomacentrids in degree of cutaneous respiration. Relative survival- time was influenced by venom dosage; ranking order of species' survival times was different at low doses than at high ones, and taxonomic correlations broke down. INTRODUCTION RESPONSE To venom can be expressed either as resistance or as survival time . These differ in their ecological consequences in the predator-prey relationship . Resistance, indicated in the present study by LD 50 values and probit lines, is reciprocal to toxicity . From the standpoint of the prey, resistance is the important concept as it is a measure of whether or not the prey will survive a particular envenomation . From the standpoint of the predator, toxicity is the main consideration; it indicates whether venom is strong enough, or administered in sufficient quantities, to subdue a particular animal . Survival time, on the other hand, does not necessarily define the ultimate outcome of an envenomation, but rather the speed at which that outcome is achieved. For example, two venoms might both produce the same percentage mortality eventually, but one may do so rapidly and the other slowly. Survival time of prey is important to the predator. The longer the survival time, the greater is the chance the prey will escape or cause damage through counterattack. This consideration may be less important to the prey . Ifit does not survive a given envenomation it is trivial whether the speed of demise is rapid or slow, or even whether it escapes during that period if it merely succumbs anyway after it is free. Author to whom correspondence should be addressed . TOX 30 :3-C
259
260
K. D. ZIMMERMAN et al.
In the see-sawing evolutionary race between predator and prey, selection will favour prey with high resistances, and predators with highly toxic, rapidly acting venom. Snake venom probably evolved within such a feeding context, rather than in a defensive one, and its deterrent role against enemies would seem to be a secondary consequence of its primary function of preventing prey escaping or counterattacking (HEATWOLE, 1987). Despite this, most studies of sea snake venom have focused upon either its toxicity to laboratory animals or its clinical effects upon humans (SUTHERLAND, 1983). As valuable as these studies are in other regards, they reveal little about the effectiveness of venom in the interaction of snakes and their prey and have little bearing either on the life of snakes in their native environment or on the resistance of prey to attack . The present paper is a step toward redressing that situation by examining the responses of five species of fish to the venom ofAipysurus laevis, a sea snake known to prey upon them (McCosKER, 1975 ; VORM and VoRis, 1983; Bums, unpublished Ph .D . thesis). MATERIALS AND METHODS The prey selected were three pomacentrids, Chromis nitida, Chromis atriwctorahs and Dascylhu armanus, and two blennies, lstibknnius meleagris and Isttbknnius edentuhas. All species were about the same size, with a total range in body wt of 5-10 g. They were collected at Heron and Wistari reefs on the Great Barrier Reef, Australia. The Chromis species were captured by anaesthetizing with quinaldine (GnisoN, 1967), D. armanus by corralling them in small coral heads, which were then brought to the surface forfish removal, and the Istiblavdus speciesby hand-collecting from under rocks and dead coral during low tides at night. Experimental observations were carried out in an aquarium divided into seven compartments, each 12 x 10 cm and with its own independently controlled air supply . Six compartments were used for observing fish, one per compartment . The seventh compartment contained a filtration system serving the entire aquarium. Venom was collected from Aipysurus laevis from the Swain Reefs, Great Barrier Reef, lyophilized and stored at about 0°C until required. Prior to injection it was dissolved in 0.9% physiological saline. Serial dilutions of venom were used until no mortality occurred among injected fish. A further requirement was that survivorship should be between 16 and 84% for at least four dosage levels . For each dilution, 12 fish of each species were injected with a venom solution in the epaxial muscle of the caudal peduncle using an Agla micrometer syringe. Doses were in the range of 0.001-15.5 mg dry wt of venom per kg fresh body wt of fish. The fish were observed for 48 hr, a sufficient time for action of the venom to be manifested (CHRISTENaN, 1968); times of deaths were recorded . Controls (uninjeeted fish) and saline controls (fish injected only with physiological saline) also were observed for 48 hr to detect any effect of handling procedures or holding facilities . Resistance was modelled using probit analysis (FINNEY, 1971) with log dose being used as the independent variable. LDw values and their fiducial confidence intervals were calculated and then back-transformed to the original scale. The adequacy of the model was assessed through a test of heterogeneity and the probit lines compared for equality of slope (parallelism). Parallel lines were further tested for equality of intercept. The logistic curve was used to model mean time to death against venom dose. Calculations for probit and logistic curves were carried out using the Maximum Likelihood Program (MLP) (Ross, 1980). RESULTS
Controls and saline controls maintained good health throughout the experiments. There were considerable differences in resistance to venom. Dascyllus aruaruts was the most resistant, showing a more gradually sloping mortality curve (Fig. 1) and a higher LD50 (Table 1) than the other species. Following in descending order of resistance were: I. edentulus, C. nitida, I. meleagris and C. atripectoralis . Thus, resistance is not strictly along taxonomic lines. The probit model appeared adequate in all cases (tests of heterogeneity were all nonsignificant; P > 0.05). Tests for parallelism (Table 2) showed there were no significant differences among the slopes of the probit lines for C. nitida, D. aruanus and I. meleagris,
Sea Snake Toxicity
26 1
100
Dose (MQikfl) FIG. 1. MORTALITY OF PREY SPECIES OF FM AT DU+FMNT LEVELS OF ENVENOYATION.
Each curve represents 10 computer-cakUlated points equally spaced between 0 and 90% mortality . Circles: D. aruanw, dots: C. nitida; Xs : C. atripectoraAr, triangles : I. nttkagris; squares : I. edeitulua.
or of the lines for I. meleagris, I. edentulus and C. atripectoralis. However, the probit lines for C. nitida (P < 0.01) and D . aruanus (P < 0.05) were not parallel with the line for C. atripectoralis. Only parallel lines were tested for differences in position, with only C. atripectoralis and I. meleagris appearing to have the same intercepts . These relationships are also evident from the shifts of the curves in Fig. 1 . Comparisons of 1.D ., values (Tables 1, 3) show significant differences between C. nitida and D . aruanus (P < 0.05) and between members of the following pairs (P < 0.01): C. nitida and I. meleagris, D . aruanus and I. meleagris, C. atripectoralis and I. edentulus and I. edentulus and I. meleagris. Only C. atripectoralis and I. meleagris exhibited no significant difference in LD5 0 values . The highest dose acted rapidly on all species, with the two Chromis species showing the quickest response and the two Istiblennius species reacting the slowest (Fig. 2). At lower doses, I. meleagris demonstrated a relative change in position and was the quickest to react to the venom. Of the three pomacentrids, D . aruanus was slowest to react at all doses, whereas C. atripectoralis (the least resistant) reacted most quickly. There were no significant differences in the slopes of the probit lines between C. nitida, C. atripectoralis and D. aruanus (P > 0.05). This is also evident from their similar rates of decrease in survival time with increasing venom dose (Fig. 2). Istiblennius meleagris and I. edentulus TAa1B 1. P~ ANALYam OF To ncm of Aipywua laevit vE~ To Fm
sp~ OF MW FM
Spceis Cbromis nitida Chromis atrlpectoralis Dawylba arwaw lat"nnbu mekagris Istiblemlus edentuba
ID» (mg kg- ')
Slope
5% Fiducial limits
0 .19 0 .09 0 .27 0.09
-11 .9 -9l .7 -7.8 -47 .6 -19.2
0 .24-0 .16 0 .09-0 .08 0 .38-0 .22 0 .13-0.08 0 .27-0 .22
025
262
K. D. ZIMMERMAN et al.
T>jem 2. Tests PoR tearnaoae«rrr, PARALLEL~ Am POmTION OF PR~ I~ Pm Chrondu nitida, Chromis atripectorali3, Dascy1hu aruanus, IstibLmrius neeL~ AND Istible~ edentuhu Species compared C. nitids C. atripectoralis D. aruanus I. nwLogris I. edentuhrs
Heterogeneity d.f.
,e
X2
Parallelism d.f.
11 .59
22
14 .81**
4
C. nitida D. aruamu I. meleagris I. edbetulus
9.64
17
6.09
3
C. nitids C. atripectoralis
6.14
9
10 .48**
1
C. nitids D. ar~
5.66
9
0.16
C. nitids I. meleagris
6.82
8
C. nitida I. edentulus
5.54
C. atripectoralis D . artumus
X2
Position
d.f.
40.18**
3
1
7.95**
1
1 .94
1
11 .91**
l
8
4.74*
1
3.42
10
9.00**
1
C. atripectoralis I. nieleagris
4.58
9
2.56
1
0.52
1
C. atripectoralis I. edentulus
3.30
9
1 .95
1
44 .09**
1
4.10
9
1 .35
1
26 .63**
1
I. edent~
2.82
9
3.50
1
0.08
1
I. nmeieagris I. edentuhrs
3.98
8
0.11
1
29 .34**
1
D . aruamAs
I. mekagris D . aruanus
d.f. degrees of freedom; X' chi-square . *P 0.05 NS
C. atripectoralis !. edentuha
20.80
< 0.01
nteleagris I. edattuha
10 .79
< 0.01
L
NS - not significant.
tion with taxonomic category . In order of decreasing survival time the ranking at high doses is: I. edentulus, I. meleagris, D. asanus, C. nitida and C. atripectorahs. The blennies survived longer than the pomacentrids, and among the latter Dascyllus survived longer than the Chromis species. However, relative survival times among species depended on dose. At some lower doses, for example, I. meleagris had the shortest, rather than the longest, survival time, and in general taxonomic correlations broke down . Differences among species in response to venom may reflect differences in habitat and life styles. Of the two Istiblennius species, I. edentulus, the most resistant one, often was
i0" 10
102 10
0.1 0.001
t
0.01
t
~a
0.1
1
t 10
Dose Ftm 2. FsrFecT of SEA wAn vErrnr nose oN mEAN Tun To DEATH 1N raev sPBcms of Ftsx. Solid lines indicate pomaeentrids; dotted lines indicate gobies. Other symbols as in Fig. 1.
264
K. D. ZIMMERMAN et al.
isolated at low tide under rocks and coral in small pools away from the water's edge . Blennies have Cuaneous respiration (GRAHAM, 1976) and low oxygen requirements dictated by a secretive life style (FRY, 1957; Moyu and CECH, 1982 ; HUGHES, 1984). Both of these features would allow for longer survival in the face of breakdown of the ventilatory system, one of the consequences of sea snake envenomation (ZIMMERMAN et al., 1990). This may be why blennies had longer survival times at high doses than did pomacentrids . Once gill ventilation ceased, however, they survived only a while longer then the others, and death was inevitable . Cutaneous respiration is insufficient, on its own, to maintain adequate oxygen levels indefinitely while in the water (Moyu and CECH, 1982), even though it may allow prolonged existence out of water (NORMAN, 1975). It may be that the mechanism for death differs at different doses; perhaps neurotoxic components are more important at high doses, with other, slower-acting toxins becoming operative at doses below the lethal limits of neurotoxins. Acknowledgements-We are grateful to the University of New England for support of the project, to the Heron Island Research Station for facilities and equipment and to Dr Neomm PotuN for critical review of the manuscript. REFERENCES GH;usTw=N, P. A. (1968) The venoms of Central and South African snakes. In : Venomous Animals and their Venoms, Vol. 1, pp. 437-461 (BucHtt., W., BucrLEY, E. E. and Dauram, V., Eds) . New York : Academic Press. Fa4NEY, D. J. (1971) Probit Analysis, 3rd Edn. Cambridge: Cambridge University Press. FRY, F. E. J. (1957) The aquatic respiration of fish. In : The Physiology ofFishes, pp. 1-64 (BaowN, M. E., Ed .) . New York: Academic Press. GmwN, R N. (1%7). The use of the anaesthetic Quinaldine in fish ecology. !. Anlm . Ecol. 36, 295-301. GtuHAK J. B. (1976) Respiratory adaptations of marine air breathing fishes . In : Respiration of Amphibious Vertebrates, pp. 165-187 (Huatms, G. M., Ed .) . New York : Academic Press. HEATwoiB, H. (1987) Sea Snakes. Sydney: The New South Wales University Press. Hucttms, G. M. (1984) General anatomy of gills. In : Fish Physiology, Vol. 10, pp. 1-72 (HoAa, W. S. and RAmAm, D. J., Eds). New York : Academic Press. McCossmre, J. E. (1975) Feeding behaviour of Indo-Australian Hydrophiidae . In: The Biology of Sea Snakes, pp . 217-232 (DurnoN, W. A., Ed.). Baltimore: University Park Press. MaYLA P. B. and Cixz JA, J. (1982) Fishes: an Introduction to Ichthyology. Englewood Cliffs, NJ : Prentice-Hall . NoamAN, J. R. (1975) A History of Fishes . London: Ernest Bann. Ross, G. J. S. (1980) Maximum Likelihood Program. Harpenden, Hertfordshire, Lawns Agricultural Trust, Rothamated Experimental Station. SurmaRLAND, S. K. (1983) Australian Animal Toxins. New York : Oxford University Press. Wan, H. K. and Wan, H. H. (1983) Feeding strategies in marine snakes : an analysis of evolutionary, morphological, behavioural and ecological relationships. Am. Zool. 23, 411-425. ZnoatmAN, K. D., GATES, G. R. and HEATwor.E, H. (1990) Effects of venom of the olive sea snake, Aipysurus laevis, on the behaviour and ventilation of three species of prey fish . Toxicon 28, 1469-1478.
