
*FIELD EVIDENCE OF TREMATODE-INDUCED GIGANTISM IN HYDROBIA* SPP. (GASTROPODA: PROSOBRANCHIA)**
Alexander M. GORBUSHIN
WEB version of the paper published in J.Mar.Biol.Ass.U.K. (1997), 77, 785-800
Department of Invertebrate Zoology, Biology and Soil Faculty, St Petersburg State University, Universitetskaya nab., 7/9, St Petersburg, 199034, Russia; e-mail: gorbushin@yahoo.com
The study of growth rates in the prosobranch snails Hydrobia ulvae and H. ventrosa under field conditions showed that growth rate of snails infected by different trematode species is species-specific. Trematodes from the families Microphallidae and Heterophyidae cause gigantic growth whereas species from families Notocotylidae and Bunocotylidae have no effect on growth rate. This discrepancy is attributed to the different pathogenicity of the parasites. However, under experimental conditions with different host population densities the effect of infection by Bunocotyle progenetica effect varies from a tendency to stunt the growth (under high density) to a significant increase in growth rate (under low density). The effect of Himasthia sp. (Echinostomatidae) was shown to be population-specific. These findings agree with a previously reported hypothesis that the growth response of trematode-infected snails depends (among other things) on supplement of host-parasite system with food. Trematode infection caused parasitic castration in all studied host-parasite combinations. It is argued that the phenomenon of gigantism in infected snails is a consequence of the reduced sexual ability of the host. The correlation between snail growth rate and penis size of infected males was negative. The correlation between growth rate and penis size was not found in uninfected snails from the same age group.
Some of the factors that may regulate growth rate in infected snails are discussed from the energetic cost standpoint. A modified version of the previously reported 'energetic' concept explaining growth change following infestation is suggested. It is predicted that gigantism will generally occur in snails having intermediate longevity and infected with trematodes of low-pathogenicity.
INTRODUCTION
The change in snails growth pattern following infestation by the parthenogenetic generation of trematodes has been known since Wesenberg-Lund (1934) noted that infected snails attained a larger size than uninfected ones. The classical laboratory study by Rothschild & Rothschild (1939) on the growth of Hydrobia ulvae Pennant confirmed that the growth rate increased following infestation. This phenomenon, called 'gigantism', has become quite an enigma. The phase of initial collecting of phenomenological data was complete in the early 1980s. It is now clear that infestation by trematode parthenites may not only accelerate growth but also induce stunting effects on growth rate (Moose,1963; McClelland & Bourns, 1969; Sturrock & Sturrock, 1970; Meuleman, 1972; Sluiters et al, 1980; Wilson & Denison, 1980; Huxham et al., 1993). Two hypotheses have been produced to explain this phenomenon, the 'energetic' conception by Sousa (1983) and the 'adaptive' hypothesis by Minchella (1985). However, an adaptive hypothesis is often much easier to postulate than to demonstrate conclusively. As a result, Minchella's hypothesis has been criticized by Mouritsen & Jensen (1994). The conception of Sousa (1983) estimating the energetic budget of the host/parasite system and the directions of energy flow has been further developed by Fernandez & Esch (1991a).
Only four studies have investigated the influence of trematodes on snail growth using a field approach (Hughes & Answer, 1982; Sousa, 1983; Fernandez & Esch, 1991a; Huxham et al., 1995). All but one (Huxham et al., 1995) failed to show gigantism under field conditions. This fact made Fernandez & Esch (1991a) describe the phenomenon as a laboratory artifact. Huxham et al. (1995) found that all large specimens of H. ulvae in the Ytham Estuary, north-east Scotland were infected. However, this result can not be considered as conclusive evidence for gigantism as itself (see Sousa, 1983). Whether or not gigantism is a naturally occurring phenomenon has been speculated on by Fernandez & Esch (1991a).
The purpose of this study was to investigate the species-specific, population and region dependent effects of trematode parthenites on the growth pattern of H. ulvae and H. ventrosa Montagu, using a new method that allows an estimate of the growth rate of snails under natural conditions. A modified version of 'energetic' conception (Sousa, 1983; Fernandez & Esch, 1991a) is suggested.
MATERIALS AND METHODS
Collection of animals
The investigations were carried out in the Kandalaksha Bay of the White Sea (Russia), Limfjord (Denmark) and Wadden Sea (Germany). Three localities inhabited by Hydrobia spp. from the Kandalaksha Bay (the estuary of the Keret River, Chupa Inlet) were used for sampling: Suhaja Salma (Figure 1, Loc. 1); the south-eastern end of Malyi Gorelyi Island (Figure 1, Loc. 2); and a lagoon near by Levin Navolok (Figure 1, Loc. 3). In Denmark and Germany samples were taken in Lendrup Lagoon, Limfjord (Figure 1, Loc. 4) and near by Dangast (Figure 1, Loc. 5), respectively. Hydrobia ulvae and H. ventrosa were collected by sieving sediment through sieves retaining snails >0-3 mm.
Measurements and dissection
Only snails having one or more winter growth interruption lines on their shell (Gorbushin, 1993) were studied. The following measurements were taken from the snails' shell (Figure 2): shell diameter at the level of the last winter growth interruption line (do) (Figure 3) and shell angle gain (D j ) from the last winter interruption line to the rim of the shell aperture. All measurements were taken under the stereomicroscope using a standard ocular provided with a micrometer and an ocular angle meter. For angle growth measurement, the shell was vertically oriented so that the columelar axis was perpendicular to the plane of measurements. The error of the diameter measurement did not exceed 0-05 mm and the angle gain was measured with p /12 accuracy. After the measurements each snail was examined for trematode infection. Trematode species (Table 1) were identified according to Deblock (1980), Galaktionov (1985, 1991). During examination the sexual maturity of the individuals was determined.
Figure 2 The diagram of a Hydrobia shell measurements: do, shell diameter; D j, shell angle gain.

Each snail was examined for the presence of sexual products in the gonads (eggs in female and active spermatozoids in males). During dissection of H. ulvae males from Lendrup lagoon, the length of the penis was measured with 0-04 mm accuracy.
For the comparison between growth rates of infected and uninfected snails, only individuals with similar shell diameter at the level of the last winter line (do) were used due to obvious negative correlation between D0 and D j (Figure 5). For this purpose, pairs of snails with similar sex were chosen randomly by using computer algorithm. Statistical test of the differences between mean shell angle gains of infected and uninfected snails was carried out using two-tailed paired T-test. The number of compared pairs is presented in Table 2.
Cage experiment.
The growth rate of infected and uninfected H. ulvae and H. ventrosa was examined in 1991 during a cage experiment. The snails were kept in cages placed on the intertidal zone at the south-eastern end of Malyi Gorelyi Island supporting populations of both H. ulvae and H. ventrosa (Figure 1, Loc. 2). The tidal amplitude is -2 m. The salinity depends on tidal state and season, and ranged during the experiment between 3 and 15%. The overall Hydrobia density in 1991-1992 was 5000-7000 ind m-². The sediment was muddy and anoxic below 1.0-1.5 cm.
Experimental animals were collected by sieving surface sediment through sieves which retained snails larger than 0-3 mm. These were sorted according to species and sized in the laboratory. Equal sized snails (diameter of the last shell whorl) from the most numerous age group were taken for experiment. Hydrobia ventrosa and H. ulvae with shell diameters of 1-56 ±0-21 mm and 1-88 ±0-16 mm, respectively, were used. The snails of both species have survived one winter and their age at the start of the experiments was -8-10 months.
The three experimental cages were made of nylon nets with a mesh size of 0-3 mm taut on all sides of a wire frame shape with a vertical side of 6 cm and a square base with sides of 25,16,10 cm. In this way three snail densities were used: Dl, 3200; D2, 7800; and D3, 20,000 ind/m², i.e. each cage with base square of 1/16, 1/40 and 1/100 m²contained 100 H. ulvae and 100 H. ventrosa.
The cages were placed on the bottom of an intertidal pool (4-5 cm deep) in the mid-tidal zone. Each cage was pushed 1 cm into the sediment. Natural sediment was added through the roof of the cages, so that the sediment level inside the cages was equal to the surrounding sediment level. During the experiment the experimental set-up was inspected at regular time intervals through the transparent net walls. The experiment was initiated at the beginning of the snails' growth season (7 June) and finished 23 September, 1991. The line of winter growth interruption served as a natural mark for the start of the experiment. At the end of the experiment snails were collected from the cages and fixed in 70% ethanol, measured in the same manner as it is described above and examined for infection with trematodes. To measure the difference in growth rates a two-tailed paired T-test was used.
RESULTS
The parthenites of all seven trematode species (Table 1) were located in the snails' gonads and hepatopancreas. This type of localization leads to parasitic castration in both sexes. Microscopy of the gonad contents showed the absence of active spermatozoids in gonads of infected males in contrast to uninfected ones. Furthermore, infected males from Lendrup Lagoon exhibited a significantly (P<0-05) reduced penis size in comparison with uninfected ones (Figure 4). The number of eggs in the gonads of infected females varied from none to very few, and there was always less than in gonads of uninfected individuals. In the group of infected males from Lendrup lagoon there is a significant negative correlation between penis size and growth rate (r=-0-6, P<0-0002). In contrast the correlation between penis size and growth rate in the group of healthy males was not significant (r=-0-2, P>0-05).	Figure 4. Average penis size of infected and uninfected snails from Lendrup population: noinf, healthy males; all, all infected males (pooled data); mari, males infected with Maritrema subdollum; mic, Microphallus claviformes; cry, Cryptocotyle sp.; him, Himasthla sp.

The growth patterns of Hydrobia ulvae and H. ventrosa from the two populations are shown in Figure 5. The snails' growth rate correlates negatively with size at the start of the growth season. The smaller the shell diameter at the moment of winter line formation, the higher angle gain at the date of sampling. The growth rate response of infected snails depends on the trematode species infecting them. Notocotylus sp. does not increase the growth rate of infected individuals (Table 2). Likewise, there are no significant differences in the growth rates between snails infected with Bunocotyle progenetica and uninfected H. ulvae and H. ventrosa. In contrast, infestation by microphallid trematode species (Maritrema subdollum, Micropfullus claviformis and M. pirum) caused a significant increase in the growth rate of H. ulvae. Snails infected by M. pirum had the highest growth rate. It was possible to identify individuals infected by this species before dissection because, probably due to enhanced growth, the coloration of the periostracum on the last whorl was lighter than usual. The growth rate of H. ulvae individuals infected with Himasthia sp. differs significantly from the growth rate of uninfected ones. A similar effect on growth rates was produced by parthenites of Cryptocotyle sp. in both Hydrobia spp.	Figure 5. Growth rate (D j ) of H.ulvae and H.ventrosa as function of shell diameter (d0) on the moment of winter line formation; A - H.ulvae (1 - sexually immature, 2 - matured individuals); B - H.ventrosa

Table 2. Results of statistical analysis of differences in growth rate of infected by different species oftrematodes and uninfected Hydrobia ulvae and H. ventrosa
	H.ulvae		H.ulvae	H.ulvae	H.ulvae	H.ulvae	H.ventrosa	H.ventrosa	H.ventrosa	H.ventrosa
	Sukhaja		Sukhaja	Malyi	Dangast	Lendrup	Malyi	Malyi	Zalivnoe	Zalivnoe
	Bay		Bay	Gorelyi		Lagoon	Gorelyi	Gorelyi	Lagoon	Lagoon
	16.08.89		29.06.90	28.07.90	02.12.93	03.10.95	26.08.89	28.08.90	10.08.89	28.09.89
							=	=	=	=
Notocotylus sp.							(11)	(11)	(85)	(49)
							P>0-05	P>0-05	P>0-05	P>0-05
				=			=	=
Bunocotyle progenetica				(17)			(33)	(73)
				P>0-05			P>0-05	P>0-05
					=	+
Himasthla sp.					(21)	(10)
					P>0-05	P<0-005
	a =	b+	+	+		+			+	+
Cryptocotyle sp	(107)	(51)	(59)	(19)		(19)			(23)	(15)
	P>0-05	P<0-006	P<0-009	P<0-0004		P<0-0003			P<0-02	P<0-05
	+				+	+
Maritrema subdolum	(53)				(20)	(14)
	P<0-03				P<0-03	P<0-0004
	+		+			+
Microphallus claviformis	(13)		(9)			(12)
	P<0-04		P<0-05			P<0-001
	+
Microphallus pirum	(15)
	P<0-03
a, sexually immature individuals; b, mature individuals; =, growth rate of infected and uninfected snails do not differ; +, growth rate of infected snails is higher; number of compared pairs in brackets

Density of cage population is counted as an average density between initial and final densities.
As expected, the increasing cage population density led to decreasing growth rates of snails (Figure 6). At densities close to natural density values (D2) the growth rates of infected and non-infected individuals did not differ (P>0-05). Infected snails from cages with low density (Dl) showed significantly (P<0-05) higher growth rates than uninfected snails. There is also a tendency to decreased growth rates among infected snails in comparison with uninfected snails under the high-density (D3) treatment but the difference is not significant.	Figure 6. Growth rate (D j ) of H.ventrosa as function of cage density: D1, 2600 ind/m2; D2, 7400 ind/m2; D3, 19,000 ind/m2.
DISCUSSION
Traditionally enhanced growth of snails infected by trematodes is connected with parasite castration - extirpation of the gonads by parasitizing parthenogenetic generations of trematodes. In the present study the parthenites of all species were found both in the hepatopancreas and gonad, and all induced castration. The mechanisms of this phenomenon are interesting.
In early works (Etges & Gress, 1965; Cheng & Lee, 1971; Stanislawski & Becker, 1979; Ishak et aL, 1975) the view is that degeneration of gonads and decreasing sexual products under infestation is the result of energy depletion of the host by the parasite. Other mechanisms may be related to the toxic effect of accumulated excretory materials (Rees, 1936) and pathological changes in the blood supply of gonads due to mechanical pressure (James, 1965; Wrigth, 1971). In addition to the toxic and pressure effects, rediae exert a great deal of mechanical damage by the active ingestion of tissues (Cheng, 1962,1963). Recent work indicates the possibility that the parthenites have the ability to regulate the amount of host gonadotropic hormones and inhibit ovulation in snails (Hordijk et al., 1991). Moreover, the mechanism of direct suppression of gonad activity by schistosomes which excrete the neurohormones with their own exometabolites has been described (Crews & Yoshino, 1990).
Evidence that infestation with trematodes severely decreases the sexual ability of Hydrobia ulvae has been described earlier (Rothschild, 1938; Mouritsen & Jensen, 1994). In this study a reduction in penis size in infected males has been shown. However, a more important fact is that a negative correlation between growth rate and penis size was found. A correlation between growth rate and penis size was not found in uninfected snails from the similar age group. It is postulated that the individuals studied in this population were sexually mature. The trematode infestation in this age cohort leads to enhanced growtH. Undoubtedly this is direct evidence for a relationship between castration and increased growth of infected snails. It is expected that at least four factors may regulate the level of parasitic castration of snails under natural conditions: (1) term of infection; (2) intensity of infection; (3) pathogenicity of parthenites; and (4) individual tolerance of snails.
As the study has demonstrated the phenomenon of enhanced growth of at least H. ulvae following the infestation by trematodes is widespread geographically. Other reasons but not regional ones are the cause of the phenomenon.
The effect of at least Cryptocotyle sp. parthenites on the growth of H. ulvae from Sukhaja Bay is age-specific. Snails that become infected before maturity grow at the same rate as uninfected. In contrast, sexually matured individuals show generally increased growth following infection.
These findings agree with Sousa's idea (1983) which originates from an earlier explanation of gigantism based on energetic allocation (Wilson & Denison, 1980) and states that the cessation of reproduction in mature snails as a result of castration leaves a large pool of energy that can be redirected into somatic growtH. Immature individuals, because much energy is invested in somatic growth, do not show gigantism since the energy freed by parasitic castration is too small.
However, not all host/parasite combinations in the current study demonstrate enhanced host growtH. The effect of trematodes on Hydrobia spp. growth is shown to be species-specific in that snails from the same population may be infected by different species. What is the reason for these differences?
Sousa (1983) supposed that costs of repair in case of infection by redial trematode species must be higher than in the case of sporocystoid ones. These differences relate to the different mechanisms of nutrition used by parthenites. Contrary to the sporocysts, rediae of some species may have a more destructive influence on the host tissues due to the attributed histiophagy. Many times I found undigested pieces of host tissue in the gut of the Notocotylus sp. rediae. Moreover, on occasions of very intense infection the presence of undigested cercariae in the gut of these rediae was observed. Evidently, this fact argues in favour of cannibalism and specifically 'predatory' capabilities of Notocotylus sp. rediae. Judging by the gut size, B. progenetica rediae also seem to be rather pathogenic. Moreover this species having an exotic monoxenic life cycle (A.M.G., unpublished data; A.A. Dobrovolsky, personal communication) should not be 'interested' in surviving the host for long. According to Sousa's hypothesis (Sousa, 1983), the energetic costs on regeneration in such host/parasite combinations should be very high, and increasing the host growth after castration seems problematical. However, rediae are not always active histiophages. For example, the mode of feeding of Cryptocotyle sp. redia is less destructive for the host. This type of redia has a short gut and, more probably, feeds on haemolympH. It is likely that the pathogenicity of this species, as well as species from the family Microphallidae (which all have sporocysts), is much less than that of B. progenetica and Notocotylus sp. Wrigth (1971) also noted different pathogenic levels of rediae and sporocysts. However, each case needs to be individually studied because pathogenicity varies significantly in different host/parasite combinations. Taking into account the positive growth response of Hydrobia spp. infested by Cryptocotyle sp. and Microphallidae these species seem to be less destructive than B. progenetica and Notocotylus sp.
From the group of species enhancing host growth, Microphallus pirum increases the growth rate to the highest degree. This species has a dixenic life cycle with metacercariae encysting in the gonads and hepatopancreas of the same individual snail host. According to these peculiarities of life cycle and from an energetic cost standpoint, the H. ulvael M. pirum system has two additional methods for saving energy. Firstly, renunciation of cercarial emission. In terms of energy requirement, cercarial production is a very costly process which may be compared only with host reproduction (Bourns, 1974; Wilson & Denison, 1980). Secondly, regeneration of host tissues damaged by cercaria during travel from the site of infection to water-exposed surfaces of the snail's body is not energy dependent.
The growth pattern of snails infected by Himasthla sp. gives specifically interesting information. Indeed, in one of the localities studied the infected snails increased growth rate and they did not in the other one. What is the reason for such heterogeneity?
Probably the pathogenicity of this species has an intermediate character so that regeneration costs of the host may be higher than in the case of Cryptocotyle sp. infections, but lower than in both B. progenetica and Notocotylus sp. combinations. It is important that the pathogenicity of Himasthia sp. seems higher than Maritrema subdollum which in both localities has caused increased H. ulvae growth. Also, Lauckner (1980) noted that the Himasthia is a more pathogenic species than Cryptocotyle.
The growth response of parasitic castration in host/parasite combinations may depend on the pool of energy that snail and parasites have available to share. Fernandez & Esch (1991a), developing Sousa's hypothesis, suggested that environmental variables leading to different amounts of food resources in different localities should be kept in mind. It is quite possible that in the Lendrup Lagoon there is sufficient energy available for both snail and parasites. At Dangast the population of snails infected by Himasthia sp. may be more starved than those from Lendrup Lagoon and snails infected by M. subdollum from both localities.
The growth pattern of H. ventrosa infected by B. progenetica during the cage experiment also suggests that the amount of available food is an important factor for the host/ parasite energetic equilibrium. Hydrobia ventrosa infected by B. progenetica do not show enhanced growth in both populations supporting this parasite, but under artificial low density in experimental cages they increased their growth rate (presumably as a result of higher food availability). It is now clear that the growth of Hydrobia spp. is very sensitive to intra- and interspecific competition (Gorbushin, 1996). In light of this data the case, described by Fernandez & Esch (1991a) as anecdotal in part, where infected Lymnaea stagnalis during mark-recapture field experiments and under laboratory conditions had showed opposite growth patterns (McClelland & Bourns, 1969), doesn't seem as conflicting.
Today there are two contradicting hypotheses which try to explain the great diversity of snail growth responses to infestation by trematodes. Minchella (1985) states that gigantism in infected snails is a result of host adaptation to infestation. In brief, castrated individuals 'aspire' to outlast their infection in the future. Therefore they increase their body size in order to compensate for lost reproductive effort when they are at last free from parasites. On this base Minchella (1985) predicts that gigantism will occur in long-lived species which should have a higher chance of surviving until recovery. However, this hypothesis requires that the hosts are able to loose their infection for which there is only little evidence (Rothschild, 1942; Sousa, 1983; Femandez & Esch, 1991b). Moreover, the finding that the highest growth rate occurred in individuals infected by Microphallus pirum, conflicts with this hypothesis. It is difficult, if not impossible, to expect a mollusc to outlast the metacercarial infection.
Sousa's (1983) conception takes into account the following factors influencing the probability of change of snail growth in different host/parasite combinations: (1) pathogenicity of parthenites; (2) age of host; and (3) annual amount of energy invested in breeding. The latter is a function of longevity. The graphical presentation of Sousa's idea is shown on Figure 7. It is important that existing data on long-lived snails are in agreement with this conception. Actually, both studies of long-lived Cerithidea californica (Sousa, 1983) and Littorina littorea (Hughes & Answer, 1982) found no evidence that infected snails grew faster than uninfected ones. Even more, the growth of C. californica was stunted by some, but not all, trematode species.
Figure 7. Hypothetical model of probability of growth alteration in a system of sexually matured snail - trematode parthenites as a function of snail longevity and pathogenicity of parthenites (Sousa' (1983) conception); + - enhanced growth; - - stunted growth; 1 - combination of short-lived snail host and low pathogenic parthenites, probability to increase the growth rate is high; 2 - combination of long-lived snail host and low pathogenic parthenites, probability of changing the growth rate tends to zero; 3 - combination of long-lived snail host and parthenites having relatively higher pathogenicity, probability to stunt the growth is high.
However, there are some uncertainties in the Sousa (1983) conception. Firstly, the absence of a good definition of longevity produces the conditions when authors, which have opposite conceptions, used, for instance, H. ulvae for exemplifying gigantism among short-lived snails (Sousa, 1983) and in contradiction among long-lived ones (Minchella, 1985). This almost anecdotal situation illustrates one of the methodological problems, which have researchers studying this subject. Our knowledge about longevity of most of the snail species is far from perfect. In this respect Hydrobia spp. are a well-studied species (Rothschild, 1941; Muus, 1967; Anderson, 1971; Chatfield, 1972; Fish & Fish, 1974; Siegismund, 1982). According to analysis of size/frequency distribution, longevity of this species varies within the limits of 1-5-2-5 y, depending on the particular conditions of the habitat. According to the analysis of the number and localization of winter lines on the shells (Gorbushin, 1993) the White Sea populations of H. ulvae may contain more than 15% individuals surviving three years, and some snails can even over winter four times. The large parts of the White Sea H. ventrosa populations can survive two winters. In Danish waters (Limfjord, Wadden Sea) longevity of H. ventrosa has the same upper limit. The upper limit for H. ulvae is three winters (A.M.G., unpublished data).
Secondly, not all snails which expected to be short-lived are able to enhance growth as a result of parasitic castration (see review by Sousa, 1983; Minchella, 1985). This became clearer as a result of the study by Fernandez & Esch (1991a) on the growth of Helisoma anceps using a mark-recapture protocol. The study failed to show the phenomenon of gigantism in infected individuals in this short-lived (1-5 y) species. This result made Fernandez & Esch (1991a) suggest that gigantism may be a laboratory artefact caused by providing the snails with ad libitum food. According to authors, snails in nature usually subsist under competitive conditions leaving no energy free for increased growth rate.
The present investigation shows that the phenomenon of gigantism exists in field populations of Hydrobia spp. However, the growth response of infected snails varies significantly according to several factors, some of which are already mentioned above. Here a development of Sousa's (1983) idea and a hypothetical model for estimating the probability of alterations of the snails growth rate in various host / parasite combinations is suggested.
According to Sousa (1983) one of the reasons for low energetic investment in annual reproduction of long-lived snails is that they allocate a large pool of energy to maintenance, thereby enhancing the likelihood of surviving to the next reproductive period. This means that long-lived snails should have effective mechanisms of repair. The short-lived snails invest extensively in reproduction and sacrifice the effectiveness of reparative mechanisms. It follows that parthenites of some trematode species having wide specificity could be relatively more pathogenic for short-lived than for long-lived molluscs. I presume that less effective reparative mechanisms demand more energy for struggling with infection. From the standpoint of the energetic costs the energy pool released as a result of parasitic castration in short-lived snails may be absorbed by the parasites and reinvested by the snail in a struggle for survival. It can be safely asserted that the snail hosts with intermediate longevity have the intermediate strategy: they should have rather efficient (energy saving) reparative mechanisms, but also invest relatively large energy resources in reproduction. In this case the energy pool released as a result of parasitic castration may be sufficient for both parasite needs and additional host growth.
Therefore, it appears likely that under natural conditions of moderate deficiency of food resources neither short-lived species (<1-5 y) (Figure 8A, point 1) nor long-lived ones (more than 3-4 y) (Figure 8A, point 2) will exhibit gigantism. In cases of high pathogenicity of parthenites they will decrease growth (Figure 8A, points 3 & 4). The enhanced growth may be found in combinations: snails with intermediate longevity (2-3 y) like H. ulvae and parthenites with low pathogenicity (Figure 8, point 5). Under conditions of severe deficiency of food resources and starvation the probability of decreased growth rate in such combinations is higher than in the previous case (Figure 8B). In contrast, abundant food will increase the probability that parasitism causes enhanced growth (Figure 8C). Figure 8. Hypothetical model of probability of growth alteration in a sexually matured snail - parthenites system as a function of snail longevity, pathenogenicity of parthenites and amount of food resources: (A) - natural conditions of moderate deficiency of food recourse; (B) - conditions of severe food deficiency and starvation; (C) - conditions of abundant food recourse; + , enhanced growth; -, stunted growth
It is necessary to comment that the 'energetic' concept explaining the alterations in growth rate of infected snails lacks basic explanations of the physiological mechanisms for distributing energy resources in the snail / parthenites system. This is major problem of the concept. Results obtained with the model Lymnaea stagnalis—Trichobilharzia ocelata (De Jong-Brink, 1995) have confirmed the hypothetical role of the neuroendocrine systems in physiological effects evoked by schistosomes in their snail host. The perspective of such research is quite clear.
Field investigations in this area are far from complete but more interesting data should be obtained in laboratory experiments using multifactor design. Of great interest would be the study of growth patterns of snails with different longevity but infected by similar parthenites having wide specificity. The growth study of one snail species infected by different trematodes having different life cycles and inducing different pathogenicity would also be very useful. In any case it would appear likely that the phenomenon of alteration in snail growth following the infestation by trematodes gives a good chance to use it as an experimental model to study the energetic budget of the host/parasite system.
I am indebted to Dr Ingrid Kroencke who made possible the sampling in the Wadden Sea and Dr Thomas Jensen who managed my work at Rombjerg Station. My sincere gratitude is due to Dr Andrej Dobrovolskij, Dr Kim Mouritsen, Dr Thomas Jensen and Dr Alia Kharazova for valuable discussions and kind help during the preparation of the manuscript. Comments and criticism from Dr Mark Huxham allowed me to complete the study. The research was partially supported by the Russian Fund of Basic Research (RFBR), grant no. N 96-04-48-965.

REFERENCES

Anderson, A., 1971. Intertidal activity, breeding and the floating habit of Hydrobia ulvae in the Ythan estuary. J. mar. biol. Ass. U.K., 51, P. 423-437.

Bourns, T.K.R., 1974. Carbohydrate and protein in Lymnaea stagnalis eggs and Trichobilharzia ocellata cercariae. J. Parasitol., 60, 1046-1047.

Chatfield, J.E., 1972. Studies on variation and life history in the prosobrach Hydrobia ulvae (Pennant). J. Conch. London, 27, 463-473.

Cheng, T.C., 1962. The effects of parasitism by larvae of Echinoparyphium Dietz (Trematoda: Echinostomatida) on the structure and glycogen deposition in the hepatopancreas of Helisoma trivolvis (Say). Am.Zool., 2 (4), 513.

Cheng, T.C., 1963. The effect of Echinoparyphium larvae on the structure of and glycogen deposition in the hepatopancreas of Helisoma trivolis and glycogenesis in the parasite larvae. Malacologia, 1, 291-303.

Cheng, T.C. & Lee, F.O., 1971. Glucose levels in the mollusc Biomphalaria glabrata, infected by Schistosoma mansoni. J. Invertebr. Pathol., 18, 395-399.

Crews, A.E. & Yoshino, T.P., 1990. Influence of larval schistosomes on polysaccharide synthesis in albumin glands of Biomphalaria glabrata. Parasitol., 101, 351-359.

De Jong-Brink M., 1995. How schistosomes profit from the stress responses they elicit in their hosts. Advances in parasitology, 35, Academic Press, London, 177-240.

Deblock, S., 1980. Inventaire des trematodes larvaires parasites des mollusques Hydrobia (Prosobranches) des cotes de France. Parassitologia; 22, 1-105.

Etges, F.J. & Gress, W., 1965. Effect of Schistosoma mansoni infection upon fecundity in Australorbis glabratus. J.Parasitol., 51, 751-760.

Fernandez, J. & Esch G.W. 1991a. Effect of parasitism on the growth rate of the pulmonate snail Helisoma anceps. J.Parasitol., 77, 937-944

Fernandez, J. &. Esch G.W, 1991b. Guild structure of larval trematodes in the snail Helisoma anceps: patterns and processes at the individual host level. J. Parasitol., 77, 528-539.

Fish, J.D. & Fish, S., 1974. The breeding cycle and growth of Hydrobia ulvae in the Dovey Estuary. J. mar. biol. Ass. U.K., 54, 685-697.

Gorbushin, A.M., 1993. Structure of winter growth interruption lines and mechanism of their formation on Hydrobia ulvae Pennant 1777 (Gastropoda, Prosobranchia) shell in the White Sea. Zoologichesky Zhurnal, 72, 29-34 [In Russian; English summary].

Gorbushin, A.M, 1996. Enigma of mud snails shell growth: asymmetrical competition or character displacement? Oikos. (in press).

Hordijk P.L., van Loenhout H., Ebberink R.H.M., de Jong Brink M. & Joosse J., 1991. Neuropeptide schistosomin inhibits hormonally induced ovulation in the freshwater snail Lymnaea stagnalis. J.Exp.Zool., 259, 268-271.

Huges R.N. & Answer P. 1982. Growth, spawning and trematode infection of Littorina littorea from an exposed shore in North Wales. J.Moll.Stud., 38, 321-330.

Huxham, M., Raffaelli D. & Pike A., 1993. The influence of Cryptocotyle lingua (Digenea) infections on the survival and fecundity of Littorina littorea (Gastropoda: Prosobranchia); an ecological approach. J. Exp. Mar. Biol. Ecol., 168, 223-238.

Huxham, M., Raffaelli D. & Pike A., 1995. The effect of larval trematodes on the growth and burrowing behaviour of Hydrobia ulvae (Gastropoda: Prosobranchia) in the Ytham estuary, north-east scotland. J. Exp. Mar. Biol. Ecol., 185, 1-17.

Ishak M.M., Mohamed A.M. & Sharaf A.A., 1975. Carbohydrate metabolism in infected and trematode-infected snails Biomphalaria alexandrina and Bulinus truncatus. Comparative Biochemistry and Physiology, 51, 499-505.

James, B.L., 1965. The effect of parasitism by larval Digenea on the digestive gland of the intertidal prosobranch Littorina saxatilis (Olivi) subsp. tenebrosa (Montagu). Parasitol., 55 (1), 93-115.

Lauckner, G., 1980, Diseases of mollusca: Gastropoda. In Diseases of marine animals, vol.1 (ed. O.Kinne) pp. 311-424, John Wiley and Sons, New York.

McClelland G. & Bourns T.K.R., 1969. Effect of Trichobilharzia ocellata on growth, reproduction and survival of Lymnaea stagnalis. Exp. Parasitol., 24, 137- 146.

Meuleman E.A., 1972. Host-parasite interrelationships between the freshwater pulmonate Biomphalaria pfeifferi and the trematode Schistosoma mansoni. Neth. J.Zool., 22, 355-427.

Minchella, D.J. 1985. Host life-history variation in response to parasitism. Parasitol., 90, 205-216.

Moose, J.W., 1963. Growth inhibition of young Oncomelania nosophora exposed to Schistosoma japonicum. J.Parasitol., 49, 151-152.

Mouritsen K.M. & Jensen K.T., 1994. The enigma of gigantism: effect of larval trematodes on growth, fecundity, egestion and locomotion in Hydrobia ulvae (Pennant) (Gastropoda: Prosobranchia). J.Exp.Mar. Biol.Ecol., 181, 53-66.

Muus B.J. 1967 The fauna of Danish estuaries and lagoons. Distribution and ecology of dominating sp.-s in the shallow reaches of the mesogaline zone. Meddel. fra. Danmarks. fisk. havund., N.S., 5, 1-316.

Rees, W.J., 1936. The effect of parasitism by larval trematodes on the tissues of Littorina littorea Linne. Proc. zool. Soc.Lond., Year 1936, 357-368.

Rothschild, M., 1938. Further observations on the effect of trematode parasites on Peringia ulvae (Pennant 1777). Nov. Zool., 41 (2), 84-102.

Rothschild M., 1941. Observation on the growth and trematode infection of Peringia ulvae (Pennant, 1777 ) in a pool in the tamar saltings, Plymouth. Parasitology, 33, 406-415.

Rothschild M., 1942. A seven-year-old infection of Cryptocotyle lingua Creplin in the periwinkle Littorina littorea L. J.Parasitol., 28, 350.

Rothschild A. & Rothschild M. 1939. Some observations on the growth of Peringia ulvae (Pennant 1777) in the laboratory. Nov.Zool., 41, 240-247.

Siegismund H.R., 1982. Life cycle and production of Hydrobia ventrosa and H.neglecta (Mollusca: Prosobranchia). Mar. Ecol. Proc. Ser., 7, 75-82.

Sluiters, J.F., Brussaard-Wust C.M. & Meuleman E.A., 1980. The relationship between miracidial dose, production of cercariae, and reproductive activity of the host in the combination Trichobilharzia ocellata and Lymnaea stagnalis. Z.Parasitenkd., 63, 13-26.

Sousa W.P. 1983. Host life history and the effect of parasitic castration on growth a field study of Cerithidea californica Haldeman (Gastropoda: Prosobranchia) and its trematode parasites. J.Exp. Mar.Biol. Ecol., 73, 273-296.

Stanislawski E. & Becker W., 1979. Alterations of the free amino acid content in the haemolymph of Biomphalaria glabrata (Pulmonata) in starvation and after infection with Schistosoma mansoni (Trematoda). Comparat. Biochem. and Phisiol., 63, 477-482.

Sturrock B.M. & Sturrock R.F., 1970. Laboratory studies of the host-parasite relationship of Schistosoma mansoni and Biomphalaria glabrata from St. Lucia, West Indies. Ann. Trop. Med. Parasitol., 64, 357-363.

Wesenberg - Lund, C.J., 1934, Contribution to the development of the trematode digenea. Part II. The biology of the freshwater cercariae in Danish freshwaters. D.Kgl. Danske Vidensk. Selsk. Skrifter, 5, 1-223.

Wilson R.A. & Denison J., 1980. The parasitic castration and gigantism of Lymnaea truncatula infected with the larval stages of Fasciola hepatica. Z. Parasitenkd., 61, 109-119.

Wrigth, C.A., 1971. The pathogenesis of helmints in the Mollusca, Helminth, Abstracts, 35, 207-224.


Figure 3. Photographs of H.ulvae shells from (A) Lendrup lagoon (Limfjord); and (B) - Suhaja bay (White Sea). L, winter-grade lines. bar - 1 mm.

Table 1. Species of trematodes encountered during the study
Family	Species	Life cycle	Parthenites
Bunocotylidae	Bunocotyle progenetica	monoxenic	rediae
Notocotylidae	Notocotylus sp.	dixenic	rediae
Echinostomatidae	Himasthia sp.	trixenic	rediae
Heterophyidae	Cryptocotyle sp.	trixenic	rediae
Microphallidae	Microphallus claviformis	trixenic	sporocysts
	Microphallus pirum	dixenic	sporocysts
	Maritrema subdollum	trixenic	sporocysts

	D1	D2	D3
Density of cage population (ind/m²)	2600	7400	19000
Survival of H. ventrosa (%)	65	80	86
Rate of infestation by B. progenetica (%)	20	18	17
Number of compared pairs	12	14	14