ATTACKS OF CYCLOPOID ACANTHOCYCLOPS ROBUSTUS (SARS) ON NEWLY HATCHED CYPRINIDS

Waldemar G. Piasecki

ABSTRACT

Two experiments on newly hatched, 1-d-old, immobile and non-feeding larvae of carp (Cyprinus carpio Linne) and crucian carp (Carassius carassius (Linne)) were performed. Each experiment involved 3 treatments: WA (larvae + Acanthocyclops robustus); WAD (larvae + A.robustus + additional food for the cyclopoids consisting of Daphnia magna, rotifers, and ciliates); and W (larvae only; control). The carp larvae experiment took 5 days, while the crucian carp one proceeded for 6 days. During the experiments, the number of larvae in each treatment, their condition, degree of damage by A.robustus attacks as well as behaviour of both the larvae and cyclopoids were checked. The experiments demonstrated that A.robustus did attack the larvae, the attacks resulting in a considerable mortality. The mortality rate depended on the cyclopoid density as well as on the presence and quality of an alternative food supply. The attacks were most frequently effected by males and

Key words: copepods, fish, predation.

INTRODUCTION

Feeding by predatory fauna can be a source of a considerable mortality among eggs and larvae of numerous fish species. Fish eggs and larvae are frequently attacked by invertebrates, both benthic and pelagic [3, 8, 13, 14]. Among the latter, most of the references are made to the copepod crustaceans, particularly the predatory cyclopoids [2, 3, 5, 13, 14]. Fierce attacks on live fish larvae by the cyclopoids were observed in experimental carp cultures kept by the Department of Hydrobiology in special ponds filled with treated water from No. 1 retention basin of the POLICE Chemical Works [11, 12]. The attacks resulted in a total extermination of larvae <10 mm in length. The small, mass-occurring cyclopoid Acanthocylops robustus (Sars) was identified as the culprit.

The aim of this study was to confirm the attacks by A.robustus on fish larvae, to assess the scale of the predation, and to find out if the cyclopoid could affect the size of fish populations.

MATERIALS AND METHODS

The behaviour of A.robustus with respect to fish larvae was observed in two experiments, each involving a different cyprinid species: carp (Cyprinus carpio Linné) and crucian carp (Carassius carassius (Linné)).

In both experiments, the 1-d-old, non-feeding and immobile larvae of carp and crucian carp measuring, on the average, 6.26 and 5.04 mm, respectively, were used. The experiments were run in evaporation vessels filled with filtered and aerated water brought from those water bodies which the fish had been earlier collected from. The carp larvae were obtained from a spawning pond belonging to the Dzwonowo fish farm, while the crucian carp larvae were obtained as a result of natural spawning effected at the Department of Hydrobiology (the spawners had been kept by the Department of Aquaculture in the Dolna Odra power station heated water canal).

The experimental design involved 3 treatments applied to each experiment, each treatment being run in 10 replicates. The treatments were as follows:

Treatment 1 (WA): larvae + different developmental stages of A.robustus (equal proportions of mature females and males as well as copepodites V and IV);

Treatment 2 (WAD): fish larvae + different stages of A.robustus (as above) + additional cyclopoid food

Treatment 3 (W): fish larvae only (control).

The carp and crucian carp experiments differed in the duration, sample sizes, cyclopoid stocking density, and additional food sources for cyclopoids. The differences are summarised in Table 1.

Throughout the experiment, water temperature was measured daily. Daily checks on the number of larvae and their condition were made as well. Observations on the incidence of the attacks and the behaviour of the larvae were made. The mortality among the larvae was recorded. In both cases the experiment was terminated when the larvae began actively "hunting" for A.robustus (after 5 and 6 days in the case of carp and crucian carp, respectively). The A.robustus predation was photographed (Plates 1-9) and videotaped. Photographs were taken with the Multi Scan software (Computer Scanning Systems). The final results are presented in Table 2.

RESULTS

Quantitative data on A.robustus predation on fish larvae

The two experiments showed that, indeed, A.robustus was aggressive towards the fish larvae. This is evidenced by changes in the number of larvae (means of 10 replicates) with time (Fig. 1) No natural mortality was recorded in both (carp and crucian carp) controls (W): all the control larvae survived in good condition.

The situation was different in the remaining two treatments. In the carp WA (larvae + cyclopoids), each evaporating vessel containing 20 A.robustus individuals showed, after 5 days, a reduction in the number of larvae by 2.2 on the average. The maximum and minimum reductions in the number of larvae, caused by cyclopoid predation, were 5 and 1, respectively (Table 2). In WAD, in which the additional food supply consisted of Daphnia magna and rotifers, the average reduction after 5 days was 1.7 larvae, the maximum (recorded in one replicate only) being 4.

Still more convincing results were obtained in the crucian carp experiment. An increased number of cyclopoids added to the evaporating vessels (50 per vessel) resulted in higher losses among the larvae in WA. The mean and maximum daily reductions were 0.65 and 3 larvae, respectively. Throughout the experiment, the average reduction was 3.9 larvae; the maximum loss of 7 individuals was recorded in one treatment. In WAD, in which the alternative food source for cyclopoids consisted of rotifers and ciliates added daily to each replicate, no predation-caused losses among larvae were recorded. The only case of mortality was registered on the last day of the experiment: the dead larva had no traces of damage on the surface of the body, nor was it observed to have been attacked by the cyclopoids.

The statistical treatment (Student's t test; Microsoft Excel) showed the differences between the control (W) and WA to be significant (p<0.01) in both experiments. In addition, the two WAD treatments were found to be significantly different (p<0.01).

Predation of A.robustus on fish larvae: observations

As shown by continuous observation of larval and cyclopoid behaviour in the two experimentals, attacks on the larvae were initiated by mature males and copepodites. It was only rarely that a female would attack a larva; females usually arrived on the scene when the larva was already weakened, and participated in preying. Due to difficulties in observations resulting from high motility of both the larvae and the cyclopoids and the need to use relatively low magnification, it was not possible to precisely determine proportions between the attacking developmental stages. However, the approximate data from 55 observations showed the attack to have been initiated by:

males: in 18 (33%) cases;

copepodites IV: in 16 (29%) cases;

copepodites V: in 12 (22%) cases;

females: in 3 (5%) cases.

In 6 cases (11%), the attacking stage could not be identified with certainty (Fig. 2).

The cyclopoids attacked most often the larva's fins and anus (solid arrows in Fig. 3). Subsequently, expecially when the larva was already weakened, its head was attacked, the eyes, nostrils, opercula, and gills (dashed-line arrows in Fig. 3) being aimed at particularly frequently.

An A.robustus "invasion" can be divided into a number of phases:

1. Initially, sparse assaults targeting the parts on the body mentioned above, i.e., the fins and the anus, are observed. The larva moving rapidly, shakes the cyclopoids off; they stay on its body for a few seconds only.

2. The initial damages appear: lesions of fins, blood vessels, and the yolk sac in the younger larvae. The fish moves less rapidly, whereupon the cyclopoids (three or four at the same time) attach themselves to the larva's body for a longer time (several seconds). Attacks on the head, nostrils and, particularly, gills begin.

3. The larva weakens and is not able to shake off the attacking cyclopoids which grow in numbers (up to 10 individuals attacking at the same time); it moves sporadically and less energetically. The larval body shows serious lesions: eaten out fins and fragments of the body around the anus and on the head (gills); towards the end of this phase, the larva does not move and the heartbeat ceases.

4. The dead larva is voraciously scavenged on by numerous (up to 20) cyclopoids.

The entire process of attack and scavenging, described above, may take from 15 minutes to few hours, depending on how hungry the cylopoids are. Once they stop feeding, the larval skeleton remains, although it, too, may be consumed, particularly when it belonged to a younger larva. When this was the case, no skeletal remains could be detected in a vessel next day.

The cyclopoids did not actively seek their prey. The initial attacks ensued after an accidental encounter rather than as a result of a purposeful search. When swimming, the cyclopoids encountered the immobile larva and attempted to burrow, with their mouth appendages, into protruding parts of the larval body. When the larva shook off the attacker and fled, no attempt to follow it was observed. The attacker did not try to repeat the attack. However, once the larva became weaker, the behaviour of A.robustus changed. During the second, and particularly during the third phase, the chase after a victim, when it tried to shake off the attackers and flee, was observed. The cyclopoids often attached themselves to a spot from which they had been earlier shaken off. During later stages of the attack, the feeding cyclopoids changed their position on the larva's body.

On day 5 and day 6 in the carp and crucian carp experiment, respectively, the larvae (measuring 7-8 mm) began to feed actively on the cyclopoids. Cyclopoid attacks were still observed, but no mortality ensued any more.

The entire process of attack and the behaviour of A.robustus during feeding can be viewed on a video film taped during the crucian carp experiment.

DISCUSSION

The experiment described above corroborated earlier signals on the youngest fish larvae being attacked by A.robustus [11, 12] and cast new light on the scale and mechanism of the phenomenon. The quantitative data in Table 2 allow to conclude that A.robustus did actually affect the survival rate of carp and crucian carp. The best evidence is provided by the statistically significant differences between the controls (W) and WA and WAD treatments. It should be mentioned here that, in order that no additional factors be introduced, the control larvae were not fed, and yet no losses were recorded among them. The cyclopoid stocking density, too, proved important for the extent of their effect on the fish. The increased density in the crucian carp experiment resulted in increased losses among the larvae. The scale of the reduction, however, was lower than expected. Earlier reports on A.robustus attacks on fish juveniles described a complete wipe-out of the larvae [11, 12]. Fritzsche and Taege [2], too, found much greater losses among the fish larvae as a result of predation by a related cyclopoid, A.vernalis. According to their data, the cyclopoid stocking density of 10-50 ind. dm^-3 caused, during 5 days, mortality as high as 80%, while at densities of 51-100 ind. dm^-3, 95% of the larvae (of Lepomis macrochirus) were eaten. The cyclopoid density higher than 250 ind. dm^-3 resulted in a total extermination of the newly hatched Hypophthalamichthys molitrix larvae. The A.robustus densities in this experiment were 40 and 100 ind. dm^-3 in the carp and crucian carp runs, respectively. The mortality obtained at a higher density was 70% at the maximum (in one replicate), and averaged 39%, i.e., much less than it could be expected. The reason could be sought, i.a., in biomass differences between the larvae and the cyclopoids. In spite of their very high energy demand resulting from their very high motility [7], the A.robustus cyclopoids may, at some point, become food-saturated. This is still more plausible as the cyclopoids used in the experiments originated from a well-fed culture. This argument is also supported by the fact that, in both experiments, the highest losses among the larvae were recorded during the initial three days of the experiment. Later on, the situation became more stabilised, as it is clearly illustrated in Fig. 1. When carrying out an experiment, a test population is usually strictly defined in terms of size and isolated from the outer world. Such a population is restricted in its ability of intercepting energy accumulated at a lower trophic level. It cannot be ruled out that this was the case during the two experiments. The situation is different in a natural water body, e.g., a retention pond, where a sufficiently large population of a predator is capable of exhausting the existing food supply at a much faster rate. This was most probably the case in the study by Szlauer and Szlauer [12] who tried to use zooplankton in the retention pond No. 1 as food for carp larvae. The population of A.robustus living there killed out the fish larvae completely. It should be added here that water temperature, too, is of importance as the cyclopoid energy demand grows with temperature and so does their feeding rate.

The experiment showed the fish larvae not to be a preferred food for A.robustus, as illustrated by the treatment involving an additional food supply. While some larval mortality did occur in the carp experiment in which the additional food consisted of D.magna and rotifers, no larval mortality due to cyclopoid predation occurred in the crucian carp WAD in which rotifers and ciliates made up the alternative food. It can thus be concluded that A.robustus become dangerous to fish larvae once the cyclopoid's other food resources have been exhausted. In addition, it is the rotifers and ciliates rather than larger plankters (such as D.magna) that are preferred as food. Those results are somewhat in agreement with the general opinion on food preferences of not only A.robustus, but the remaining cyclopoids as well. According to Zaret (quoted by [9]), they belong to the size-dependent predators (SDP) group, i.e., those animals the feeding of which depends on the size of the potential prey, the mouth opening size being of no importance. Another example described by Pijanowska [9] qualifies cyclopoids to hemiphages, i.e., those predators which suck the prey out or break it apart. According to various authors referred to by Pijanowska [9], the hemiphagous predators almost exclusively select small-sized prey. It is difficult to accept this view completely and to adopt the 1 mm limit set by Dodson (quoted by [9]) as the maximum size of prey of preference. On the other hand, it is unquestionable that A.robustus, when offered prey in the form of fish larvae and rotifers or protozoans, will choose the smaller prey.

As observed in this experiment, males and copepodites were the stages most often found on the larval bodies. Fritzsche and Taege [2] reported A.vernalis copepodites V (47.5%) and also copepodites III and IV (25%) and mature males (27.5%) to be most active in attacking the larvae. Thus the two sets of observations differ somewhat, as mature males (33%), copepodites IV (29%), and copepodites V (22%) were most active in this study. It can be observed that the attacking stages were relatively small-sized ones, i.e., males and larvae. Females are larger and frequently carry egg sacs, and are thus more conspicuous in the water; they usually stayed away during the attack.

Prey spotting and the manner of cyclopoid attack are interesting. Most authors [4, 6, 9] reported cyclopoids to use sight when searching for food. The most recent studies [1, 4] show chemo- and mechanoreception, the latter detecting the most minute turbulence produced by potential prey, to be involved as well. Thus this could partly explain why the cyclopoids most often chose to attack a moment when the larvae were moving very slowly or stopped for a while. Perhaps this is why the males were so active, as they have particularly sensitive receptors, normally used also to locate a female for mating [10]. In addition, the transformed male antennullae, used to grasp a female, facilitate grabbing a fish larva as well. At later stages of the attack, A.robustus made an impression of searching a prey. The larvae were seriously damaged at that time, with body fluids leaking out. Thus it can be suggested that the cyclopoid chemoreceptors were used then to locate the wounded larva, followed by a search for it and the attack.

To sum up, one cannot unequivocally contend that, under natural conditions, A.robustus is capable of seriously affecting populations of fish spawning in summer. It can only be suggested that the predation by A.robustus is not a key factor, as shown by the fact, evident in this experiment, that the cyclopoids are not interested in fish larvae as food when rotifers and ciliates are present, and these are abundant both quantitatively and qualitatively in natural waters. Moreover, there is a host of other factors affecting the egg and larval survival under natural conditions, so the pressure exerted by A.robustus alone cannot, in all likelihood, be a key element. It should also be borne in mind that other species, co-occurring with A. robustus in summer, are predacious as well and have been found to prey on fish larvae, too. A.robustus may be, however, important in fish cultures. This can be particularly relevant in spawning ponds in culture facilities of carp and other warm water fish species. Such ponds are shallow and heat up very rapidly, which creates ideal conditions for the development of A.robustus. Thus it is quite possible that, once in such a pond and undergoing a very fast development, A.robustus can become a dominant zooplankter and, once its preferred food supply in the form of rotifers and protozoans is exhausted, it will threaten the newly hatched larvae. For this reason, a particular attention should be paid to the composition of zooplankton in waters feeding culture facilities.

Similarly, a close attention should be paid to the zooplankton of heated, treated water bodies such as the POLICE Chemical Works retention pond. An attempt to rear fish fry in such a pond, or to use its water and zooplankton in the culture, may fail in the presence of A.robustus. For this reason, Szlauer and Szlauer [12] recommend that in order for such a culture to be successful, a fry weighing not less than about 1 g (measuring 3 - 4 cm) should be safely used. According to observations made during this study, even much smaller, 7-d-old larvae of carp and crucian carp, measuring 7-8 mm, were resistant to A.robustus attacks.

CONCLUSIONS

The Acanthocyclops robustus predation on fish larvae was confirmed in two laboratory experiments.

The predation intensity depends on the cyclopoid stocking density and on the presence of other plankters making up an alternative food source for A.robustus. When the cyclopoids could avail themselves of rotifers, ciliates, or cladocerans in addition to fish larvae, they were less active in attacking the latter or ceased the attacks altogether.

The most frequent attackers were mature males (33%), copepodites IV (29%), and copepodites V (22%). Females usually fed upon the already killed larvae.

The experiment allows to contend that due to their food preferences, A.robustus are not a major threat to the newly hatched fish larvae. Such a threat may, however, become important in culture ponds or in heated water offering good conditions for the development of A.robustus and where the species may become a zooplankton dominant.

REFERENCES

1. Dussart B.H., Defaye D., 1995, Copepoda. Introduction to the Copepoda. Guides to the identification of the Microinvertebrates of the Continental Waters of the World. No. 7, SPB Academic Publishing bv, Amsterdam.

2. Fritzsche S., Taege M., 1979, Schädigungen bei der Aufzucht von Fischbrut (Hypophtalmichthys molitrix, Cyprinus carpio und Coregonus albula) industriemäßigen Produktionsanlagen durch starkes Vorkommen von Ruderfußkrebsen (Copepoda). Zeitschrift für die Binnenfischerei der DDR, 26: 304-308.

3. Hartig J.H., Jude D.J., Evans M.S., 1982, Cyclopoid predation on Lake Michigan fish larvae, Can. J. Fish. Aquat. Sci., 39: 1563-1568.

4. Lampert W., Sommer U., 1996, Ecology of inland waters, PWN, Warszawa. (In Polish).

5. Mamcarz A., 1990, Conditions for growth of larval Coregonus peled (Gmelin) during pond culture, Acta Acad. Agricult. Techn. Olst., Protectio Aquorum et Piscatoria.

6. Mikulski J. St., 1982, Biology of inland waters, PWN, Warszawa. (In Polish).

7. Morris M.J., Kolhage K., Gust G., 1990, Mechanics and energetics of swimming in the small copepod Acanthocyclops robustus (Cyclopoida), Marine Biology (Berlin), 107: 83-92.

8. Opuszyński K., 1983, Fundamentals of fish biology, PWRiL, Warszawa. (In Polish).

9. Pijanowska J., 1985, Mechanisms employed by planktonic animals in defence from predators, Wiad. Ekol., 31 (2): 123-172.

10. Szlauer L., 1959, Observations on reproduction of the cyclopoid Cyclops kolensis Lilljeborg, Pol. Arch Hydrobiol., 5 (18) 2: 111-120. (In Polish).

11. Szlauer B., 1985, Hydrobiological characteristics of the Police Chemical Plant effluent reservoir, D.Sc.Dissertation, Wydaw. AR Szczec. Rozpr. (In Polish).

12. Szlauer B., Szlauer L., 1987, Testing the secondary treatment of effluents from the "Police" Chemical Plant in carp ponds, Pol. Arch. Hydrobiol., 34: 135-145.

13. Żuromska H., 1967, Assessment of mortality of eggs and larvae of roach (Rutilus rutilus L.) in lacustrine spawning, Rocz. Nauk Rol. 90-H-3: 539-556. (In Polish).

14. Żuromska H., 1967, Some causes of mortality of eggs and larvae of roach (Rutilus rutilus L.) in lacustrine spawning grounds, Rocz. Nauk Rol. 90-H-3: 557-579. (In Polish).

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed ‘Discussions’ in each series and hyperlinked to the article.