DISEASES OF AQUATIC ORGANISMS 1 Published July 28 Dis. aquat. Org.

Infection dynamics of Lernaeocera lusci on sand goby minutus in the Oosterschelde (The Netherlands)

P. A. Van Damme, F. Ollevier

Laboratory for Ecology and Aquaculture, Zoological Institute, Naamsestraat 59, B-3000 Leuven, Belgium

ABSTRACT: Male and female sand gobies Pomatoschistus minutus (Pallas) from the Oosterschelde (The Netherlands) were examined for the crustacean mesoparasite Lernaeocera lusci (Bassett-Smith) from Apnl to July 1992. Prevalence increased steeply from 4 % in Apnl to 80% (females) and 57 % (males) in June. The peak abundance of recently attached infective stages (pennella larvae) on sand gobies was found in May. By June, about 70% of parasites were in the post-metamorphosis stage and possessed egg strings. Significant differences in abundance between males and females were found in June but not in May. This may be due to sex-specific differences in behaviour during spawning time. The majority of post-spawning probably d~edin July. Fish of the new year class which were cap- tured in July harboured no parasites.

KEYWORDS: Pomatoschistus minutus Lernaeocera lusci . Parasitism Oosterschelde

INTRODUCTION Mann (1964) and Petersen (1992) found that Lern- aeocera lusciwas responsible for a significant decrease Lernaeocera lusci (Bassett-Smith) is a crustacean in the condition factor of sand gobies. Moreover, both parasite which uses sole Solea solea as intermediate fat content of the liver and henloglobin content were host. Though bib Tnsopterus luscus is the typical de- lower in infected gobies as compared to non-infected finitive host, adult females can also be found attached individuals (Mann 1964). However, neither study pro- to many other (Boxshall 1974, vided a detailed account of the dynamics of the host Kabata 1979, Tirard 1991). L. minuta,which was previ- and parasite populations, though this may be essential ously distinguished from L. lusci on grounds of its small for a better understanding of the interaction between size (Kabata 1979), can be equalled to L. lusci because host and parasite individuals. For example, fish size of the host-dependency of parasite size (Van Damrne et and fish sex were not considered, both of which al. 1993). Sand goby Pomatoschistusminutus (Pallas), may have acted as confounding variables in the which was previously considered to be the typical host investigations. species of L. minuta, thus appears on the host List of In this study, we provide further details on the in- L. lusci. fection dynamics of Lernaeocera lusci during the High prevalences of Lernaeocera minuta (henceforth spring transmission wave in a tidal bay in the south- L. lusci) have been recorded on sand gobies along the ern North Sea in 1992. In a future paper the effect of German coast (86% by Mann 1964, 22% by Petersen L, lusci on hematological parameters will be pre- 1992) and along the Belgian coast (39% by Hamer- sented. lynck et al. 1989). In the Belgian as well as in the south- western Dutch coastal area, 2 transmission waves could be distinguished (Hamerlynck et al. 1989): a first MATERIAL AND METHODS wave occurred in late autumn when juvenile sand gobies were infected, and a second transmission wave Sand goby samples were collected in the western occurred in the following spring when mature, - part of the Oosterschelde, a tidal bay in the southwest ing sand gobies were infected. Netherlands. Surface water temperatures measured at

O Inter-Research 1994 Dis. aquat. Org. 19: 83-87. 1994

Table 1. Pomatoschistus minutus. Sampling dates, water temperature, number of male, female and juvenile sand gobies caught, sex ratios (ma1e:female) and mean total lengths (TL i SD) in the Oosterschelde in 1992. The year class to which the fish belonged is indicated in parentheses. F. ANOVA to test for differences in mean length between males and females (ns = not significant)

Sampling Water temp. Males (1991) Females (1991) Juveniles (1992) F Sex ratio date ("c) n TL n TL n TL M:F

6 Apr 28 Apr 15 May 26 May l1Jun 4 Jul

the sampling sites increased from 9 to 18 'C during the Table 2. Classification of adult female Lernaeocera lusci on sampling period (Table 1). Pomatoschistus minutus (after Van Damme Pc Harnerlynck 1992) On 6 April, 28 April, 26 May and 4 July the samples were taken with a commercial shrimp trawler using a Substage Definition 5 m beam trawl. The samples of 15 May and 11June were taken with a research vessel using a 3 m beam Pennella (PI) Straight body, no flexure trawl. AU fish collected (except for some of Pennella (P2) One point of flexure the 1992 year class caught on 4 July) belonged to the Immature (U) Two or three points of flexure 1991 year class. Due to the use of different sampling Genital region not swollen methods (e.g. differences in mesh size) the fish num- Mature pregravid (W) Genital region fully swollen bers do not reflect true densities. No external egg strings Mature gravid (X) External egg strings present On board, fish were anaesthesized in a benzocain X1 Immature eggs solution and preserved in formalin 8%. After about X2 Mature pigmented eggs 5 mo the total length (to the nearest mm) and total Y External egg strings partly or weight (to the nearest mg) were determined. The GSI completely spent (gonadosomatic index) of the female fish collected on Dead parasite (Z) Remains of holdfast embedded in 6 April, 15 May and 11 June was calculated as follows: host tissue

Gonad weight GSI = Total body weight - Weight stomach content RESULTS Prior to statistical analysis GSI was arcsine trans- formed. All fish collected before July belonged to the 1991 The female parasites which were attached to the gill year class (l+). In July this year class had all but disap- arches or in the gill cavity were counted and removed peared and was replaced by fish belonging to the 1992 from the host. They were assigned to one of 8 develop- (O+) year class. The numbers of female and male fish mental stages, based on the staging system of Van caught and sex ratios are given in Table 1. Between Damme & Hamerlynck (1992) (Table 2). Prevalence 4 April and 26 May the proportion of females in the (percentage of fish infected), abundance (mean num- catches increased progressively. ber of parasites per fish), intensity (number of parasites In all months, females were smaller than males, but on 1 host individual) and mean intensity (mean num- these differences were not significant (ANOVA, p > ber of parasites per infected fish) were used according 0.05) (Table 1).Juveniles of year class 1992 which were to the recommendations of Margolis et al. (1982). The collected on 4 July were significantly smaller than the relationship between the dependent variable parasite l+ females (ANOVA; F = 166, p < 0.001) and the intensity (y) and independent variables fish length (X,) l+ males (ANOVA; F = 229, p < 0.001) of June 1992 and GSI (X*) of females on 13 May and 11 June was (Table 1). analysed by a multiple regression analysis: Prevalence and abundance of Lernaeocera luscion female and male sand gobies are shown in Fig. 1. Prevalence steeply increased between 28 April and where a is the intercept, and b,and b2 are regression 15 May and then gradually increased to maximal lev- coefficients. els of 77 % (females)and 53 % (males) in June. For both Van Damme & Ollevier: Infection dynamics of Lernaeocera lusc~ 85

Table 3. Lernaeocera lusci infecting Pomatoschistus minutus. Prevalence Spearman Rank Correlations (R) between fish length and par- (%l asite intensity for males and females in the Oosterschelde in spring 1992. ns: not significant; 'p < 0.05; "p < 0.01

Date Sex n R

28 Apr F 10 0.39 ns M 8 -0.17 ns 15 May F 8 1 0.25' M 3 8 0.49" v - 26 May F 6 1 0.41" ----m- 11 Jun F 4 5 0.19ns M 32 0.25ns

6 Abundance Table 4. Lernaeocera lusci infecting Pomatoschistus minutus. Relationship between parasite intensity (y),fish length (X,) and gonadosomatic index (x2) according to the multiple regression equation y = a + b,x, + b2x2 on 15 May and on 11 June in the Oosterschelde in 1992. ns: not significant; "'p < 0.001

Males

15 May -5.55 0.11"' -0.03 ns 0.2 11 Jun 0.03 ns -0.05 ns 0 l I I April May June Fig. 1 Lernaeocera lusci infecting Pomatoschstus minutus. Prevalence and abundance of parasites on female (0) and male (m)sand gobies P minutus in the Oosterschelde (The Netherlands) in 1992. The maximum number of parasites on individual males and females, respectively, is indicated for each sample

sexes there was a similar pattern for abundance: be- tween the beginning of April and the end of June abundance increased from 0.06 to 1.77 (females) and from 0 to 1.42 (males). There were no significant dif- ferences in abundance between males and females on 28 April (Kruskal-Wallis; H = 0.23; p > 0.05) and on 15 May (H= 0.27; p > 0.05). However, female sand gobies harboured significantly more parasites than males on l1 June (H= 4.03; p < 0.05). Fish with more than 4 par- 9 15 I I asites were found in May and June (Fig. 1). April May June Fish length and parasite intensity were positively Fig. 2. Lernaeocera lusci infecting Pomatoschistus mlnutus. correlated on 15 May and on 26 May for both males Mean gonadosomatic index (GSI) of female sand goby be- and females. On 28 April and 11 June the correlations tween April and June 1992 in the Oosterschelde. SD, minlma between these 2 variables were not significant (p > and maxima are indicated 0.05) (Table 3). The mean GSI for female gobies was significantly Overall, the multiple regression equations between lower on 11 June (0.05 * 0.05) than on 9 April (0.08 * parasite intensity, GSI and fish length were significant 0.03) and 15 May (0.09 + 0.05) (Tukey test, p < 0.05) on 15 May (p < 0.01) but were not significant on (Fig. 2). The correlations between fish length and CS1 11 June (p > 0.05). Neither fish length nor GSI con- were not significant on 15 May (n = 44, R = -0.24, p > tributed significantly in the prediction of parasite 0.05) and on 11 June (n = 37, R = 0.03, p > 0.05). intensity on 11 June (Table 4). 86 Dis aquat. Org. 19: 83-87, 1994

lusci to definitive hosts. In both Hamerlynck et al. (1989) and the present study the maximum infection levels in late spring coincide with temperatures above 11 "C (Table 1)and with the occurrence of intermedi- ate hosts (soles) which harbour large numbers of adult pre-metamorphosis female L. lusci (Van Damme & Ollevier unpubl.). Because the majority of sand gobies 28 April (n=4) l5 May (n=12) die soon after spawning (Fonds 1973) there is a consid- erable reduction of parasite population size in July. However, in summer the role of definitive host is increasingly taken over by juvenile bib (Van Damme unpubl.). Along the German coast, highest infection levels were found in AugusWSeptember (Mann 1964) or October (Petersen 1992). Because neither of these authors reported the size or possible age of the exam- 26 May !n=46) l! June (n=42) ined fish it is difficuit to ma~ea comparison. it appears that the spring transmission wave did not occur in the Fig. 3. Lernaeocera lusci infecting Pomatosctusfusminutus. Population structure of parasites on sand gobies on different German coastal waters, perhaps due to differing trans- sampling dates in the Oosterschelde in 1992. The number of mission dynamics of infective pennella larvae within parasites is indicated. For explanation of Me cycle stages see this area. Possibly, there is no spring overlap in the Table 2 distribution of sand gobies and soles in the German Wadden Sea. More data on spatial distribution and temporal patterns in occurrence of sole in the area The population structure of Lernaeocera Iusci was should be gathered before this hypothesis can be characterised by a clear temporal pattern (Fig. 3). tested (Petersen 1992). Juvenile stages (P1 and P2) were mainly found on 28 The significant difference in infection levels of males April and on 15 May, whereas adult female parasites and females in late spring (1 1 June) may have resulted with egg strings were found mainly after 28 May. This from sex-specific behavioural patterns during spawn- indicates that transmission occurred mainly prior to the ing time. The males of Pomatoschistus minutus build latter dates. There was no clear temporal pattern in the nests under bivalve shells which they cover completely occurrence of XI, X2 and Y substages. Overall, the with sand. They guard the eggs of 1 or more females majority (84 %) of adult female parasites belonged to until hatching (Fonds 1973, Hesthagen 1977). During the X1 substage. egg guarding, male I! microps exhibit low feeding activities and hence reduced mobility (Magnhagen 1990). Fonds (1973), Magnhagen & Kvarnemo (1989) DISCUSSION and Hamerlynck & Cattrijsse (in press) suggested this may also be the case in P. minutus. Furthermore, After a short pelagic stage of approximately 1 mo, Hamerlynck & Cattrijsse (in press) found that males sand gobies start to live in close proximity to the bot- consume predominantly Pomatoschistus eggs, poly- tom (Fonds 1973), and become increasingly suscepti- chaetes, bivalves and Gastrosaccus, which is a benthic ble to infection by Lernaeocera lusci. In late autumn in mysid species. The hypothesis that males are often the Belgian and Dutch coastal area, a small proportion hiding in their nests (Hesthagen 1977, Magnhagen (* 10%) of this demersal sand goby population is in- 1990) may also explain why sometimes small numbers fected with L. Iusci (Hamerlynck et al. 1989). These lat- of males are caught in early summer (Swedmark 1958). ter authors recorded no further infection in winter and In the present study this was the case on 26 May when a second, dramatic, rise in infection level during 98 % (n = 62) of the gobies were females (Table 1). On spawning time (May/June). The same seasonal pattern the other hand, females roam about during the spawn- was apparently found in the present study: the few ma- ing period (Hesthagen 1977, Hamerlynck & Cattrijsse ture parasites which were found on gobies in the be- in press) with consequent higher probability of be- ginning of April had probably infected their hosts in coming infected by Lernaeocera lusci than the males. the previous autumn. By June ca 80% of the females These sex-specific behavioural differences probably and 50 % of the males were infected with L. lusci. explain the different probabilities of males and females Sufficiently high temperatures and spatial overlap accumulating L. lusci. of intermediate and definitive hosts are 2 important Transmission of pennella larvae to sand gobies in prerequisites for succesful transmission of Lernaeocera spring is restricted to a very short period. Indeed, P1 Van Damme & Ollevier: Infection dynamics of Lernaeocera lusci 87

and P2 stages were predominantly found in 1 sample LITERATURE CITED (May), whereas the June samples contained mainly mature The most plausible explanation for Boxshall, G. A. (1974). lnfections with parasitic copepods in North Sea marine . J. mar. biol. Ass. U.K. 54: 355-372 this was re- Fonds, M. (1973). Sand aobies- in the Dutch Wadden Sea cently given: it was found that all infective stages of (Pomatoschistus, , Pisces). Neth. J. Sea Res. 6: Lernaeocera lusci leave sole in a very short period, 417-478 probably triggered by rising temperatures (Van Hamerl~nck, O., Cattrijsse, A (~npress). The food of Pomatoschlstus minutus (Pisces, Gobiidae) in the Belgian Damme & Ollevier unpubl.). Based on the present coastal waters: a comparison with the food of its potential study we can provide some estimates for the duration competitor P. lozanoi. J. Fish Biol. of the life cycle stages of L. lusci. The majority of the Hamerlynck. O., Geets, A., Van Damme, P. A. (1989). The par- parasites which infected sand gobies after 28 April asites of two sympatric gobies Pomatoschistus minutus and l? lozanoi in the Belgian coastal waters. In: Wouters, were mature by 11June, suggesting a duration of 1 to K.. Baert. K. (eds.) Comptes rendus du symposium 'Inver- 2 mo from initial infection to maturation. D. tebres de Belaiaue'd A KBIN. Brussels.. A 27-30 In the present study no evidence for parasite- Hesthagen, I. H. (1977). Migrations, breeding and growth in induced host mortality nor for a relationship between Pomatoschistus minutus (Pallas) (Pisces, Gobiidae) in parasite intensity and mating success of sand gobies Oslofjorden, Norway. Sarsla 63: 17-26 Kabata, Z. (1979). Parasitic Copepoda of British fishes. The was found. Hosts with more than 5 parasites were Ray Society, London found at all sampling dates in May and June (Fig. 11, Magnhagen, C. (1990). Reproduction under risk in suggesting- - that heavily infected hosts suffer no high the sand goby, Pomatoschistus minutus, and the black mortality during this sampling period. Hamerlynck et goby, Gobius niger the effect of age and longevity. Behav. Ecol. Sociobiol. 26: 331-335 al. (1989) suggested that Lernaeocera lusci may further Magnhagen. C., Kvarnemo, L (1989). Big is better. the Lmpor- exacerbate the lowered condition of post-spawning tance of size for re~roductive success in male Pomato- sand gobies and hence induce post-spawning host schlstus minutus (Pallas) (Pisces, Gobiidae). J. Fish Biol. mortalitv, but this could not be tested due to the small 35: 755-763 sample size in July. Furthermore, no significant c0ri-e- Man", H. (1964). Vorkommen, Verbreitung und ~chad- wirkung von Lernaeocera minuta T. Scott (Copepoda par- lation was found between the gonadosomatic index asitica). Veroff. Inst. Meeresforsch. Bremershafen 9: 79-83 and the parasite intensity of females on 15 May and Margolis, L.. ~sch,G. W., Holmes, J. C., Kurie, A. M., Schad, 11 June (Table 4). However, field sampling may be G. A. (1982). The use of ecological terms in parasitology. inadequate to find such a relationship. Probably, the J. Parasitol. 68: 131-133 Petersen, F. (1992). Nutritional condition and parasites of the female sand gobies are infected with L, lusci at a time sand goby Pomatoschistus minutus (Pallas) from the when they have already released 1 or more batches of German Wadden Sea. O~helia35: 197-207 eggs. The effect of these parasites on the reproductive Swedmark, M. (1958). Sur la variation geographique de success of the sand crobies,- if present, may therefore be Gobius minutus Pallas. 1. Biologie et croissance. Archs. obscured by many confounding variables Zool. exp. gen. 95,32-51 Tirard. C. (1991). Biodiversite et biogeographie evolutive dans les systemes h6tes-parasites: le modele Gadiformes (Teleosteens) - Copepodes et Monogenes. These de doc- torat, Universite de Montpellier Van Damme, P. A., Hamerlynck, 0. (1992). The infection dy- namics and dispersion pattern of Lernaeocera branchialis Acknowledgements. This research was supported by FKFO- on O+ whiting (Merlangius medangus L.) in the Ooster- project 2 0086.88. P.A.V.D. received a grant from the National schelde (SW Netherlands). J. Fish Biol. 41: 265-275 Science Foundation of Belgium (NFWO). N. Hoebeke is Van Damme, P. A., Maertens, D.. Arrumm. A., Hamerlynck, acknowledged for the use of the commercial vessel YE 75. 0.. Ollevier, F. (1993). The role of Callionymus lyra and C. C. Heip is acknowledged for the use of the RV 'Luctor' reticulatus in the life cycle of Lernaeocera lusci (Bassett- (NIOO). Snlith). J. Fish Biol. 42: 395-401

Responsible Subject Editor: W. Korting, Hannover, Germany Manuscript first received: October 29, 1993 Revised version accepted: April 20, 1994