Hostname: page-component-745bb68f8f-5r2nc Total loading time: 0 Render date: 2025-02-06T13:49:13.307Z Has data issue: false hasContentIssue false
  • English
  • Français

Ecological impacts of the microsporidian parasite Pleistophora mulleri on its freshwater amphipod host Gammarus duebeni celticus

Published online by Cambridge University Press:  10 June 2005

N. J. FIELDING ,
C. MacNEIL ,
N. ROBINSON ,
J. T. A. DICK ,
R. W. ELWOOD ,
R. S. TERRY ,
Z. RUIZ  and
A. M. DUNN
N. J. FIELDING
Affiliation:
School of Biology and Biochemistry, Queen's University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, N. Ireland, UK
C. MacNEIL
Affiliation:
School of Biology and Biochemistry, Queen's University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, N. Ireland, UK
N. ROBINSON
Affiliation:
School of Biology and Biochemistry, Queen's University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, N. Ireland, UK
J. T. A. DICK
Affiliation:
School of Biology and Biochemistry, Queen's University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, N. Ireland, UK
R. W. ELWOOD
Affiliation:
School of Biology and Biochemistry, Queen's University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, N. Ireland, UK
R. S. TERRY
Affiliation:
Centre for Biodiversity and Conservation, School of Biology, University of Leeds, Leeds LS2 9JT, UK
Z. RUIZ
Affiliation:
Department of Geography, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ, UK
A. M. DUNN
Affiliation:
Centre for Biodiversity and Conservation, School of Biology, University of Leeds, Leeds LS2 9JT, UK
Rights & Permissions [Opens in a new window]

Abstract

The microsporidian parasite, Pleistophora mulleri, infects the abdominal muscle of the freshwater amphipod Gammarus duebeni celticus. We recently showed that P. mulleri infection was associated with G. d. celticus hosts being more vulnerable to predation by the invasive amphipod Gammarus pulex. Parasitized G. d. celticus also had a reduced ability to prey upon other co-occurring amphipods. We suggested the parasite may have pervasive influences on host ecology and behaviour. Here, we examine the association between P. mulleri parasitism and parameters influencing individual host fitness, behaviour and interspecific interactions. We also investigate the relationship between parasite prevalence and host population structure in the field. In our G. d. celticus study population, P. mulleri prevalence was strongly seasonal, ranging from 8·5% in summer to 44·9% in winter. The relative abundance of hosts with the heaviest parasite burden increased during summer, which coincided with high host mortality, suggesting that parasitism may regulate host abundance to some degree. Females were more likely to be parasitized than males and parasitized males were paired with smaller females than unparasitized males. Parasitism was associated with reduction in the host's activity level and reduced both its predation on the isopod Asellus aquaticus and aggression towards precopula pairs of the invasive G. pulex. We discuss the pervasive influence of this parasite on the ecology of its host.

Keywords

AmphipodGammarusisopodmicrosporidianPleistophora mulleritrophic
Type
Research Article
Information
Parasitology , Volume 131 , Issue 3 , September 2005 , pp. 331 - 336
Copyright
© 2005 Cambridge University Press

INTRODUCTION

Parasites are increasingly recognized as potential influences on population and community structure (Price, Westoly and Rice, 1988; Minchella and Scott, 1991; Tompkins, White and Boots, 2003). For example, parasites regulate host populations (Prenter et al. 2004) and influence the outcome of interspecific interactions (Hanley, Vollmer and Case, 1995), including those between invading and native species (MacNeil et al. 2003a). However, while microsporidian parasites are ubiquitous in the animal kingdom (Weiss, 2001; Keeling and Fast, 2002), research has concentrated on parasite taxonomy, structure and pathogenicity, with their potential ecological impacts for natural populations being less well studied.

Recently, we identified a new species of microsporidian, Pleistophora mulleri that infects the abdominal muscle of Gammarus duebeni celticus, the native freshwater amphipod crustacean in Ireland (Terry et al. 2003). The parasite is horizontally transmitted and specific to its G. d. celticus host, with laboratory studies indicating that it is not possible to transmit it to sympatric invasive amphipods Gammarus pulex, G. tigrinus and Crangonyx pseudogracilis (MacNeil et al. 2003b). Recently, we reported that P. mulleri infected G. d. celticus were more vulnerable to predation by G. pulex, which is threatening to replace the native G. d. celticus in many Irish rivers (MacNeil et al. 2003a). However, although predation is usually mutual between G. d. celticus and G. pulex, no study has considered how parasitism of G. d. celticus may alter its predatory impact on G. pulex, thereby further altering the biotic resistance of the native to species replacement. Therefore, we address this and investigate the association between infection by this microsporidian parasite and the ecology of its amphipod host G. d. celticus.

First, we describe temporal patterns in host abundance and test for links to parasite prevalence and burden. Then we investigate the association between parasitism with respect to gender and parameters linked to host fitness such as reproductive decisions and host activity. Gammarus species are strongly interactive, functioning as ‘keystone species’ influencing the population dynamics of other macroinvertebrate species (MacNeil, Dick and Elwood, 1997; Kelly et al. 2003). Therefore, we also examine the association between parasitism and the host's interactions with 2 co-occurring macroinvertebrates: a prey species, the isopod Asellus aquaticus (L.) and the invasive amphipod G. pulex.

MATERIALS AND METHODS

Patterns of parasitism in the field

Adult (males >25 mg, females >12 mg) G. d. celticus were sampled monthly from June 2000 to March 2002 along a 10 m stretch of the Downhill River (UK grid ref.: C753363). Each sample comprised 10 random 1-min kick-samples (0·9 mm mesh net), where the river-bed substrate is disturbed by kicking, thereby dislodging animals into the net. Animals were counted in the laboratory and P. mulleri infection was identified in live adult G. d. celticus using light microscopy, since infected abdominal (pleon) muscle tissue appears opaque (see MacNeil et al. 2003a,b). Individuals were classified as parasitized or unparasitized. Infection intensity was recorded as low burden (infection of one abdominal segment) or high burden (infection of entire pleon). Numbers of dead adult G. d. celticus were recorded for each sample, but parasite status could not be determined due to tissue degeneration.

To investigate parasite distribution in male and female hosts, in the November 2001 sample, all 185 unpaired adults were sexed and parasite status recorded (low and high burdens combined). Gammarus males guard females in precopula pairs for several days before mating occurs (Elwood, Gibson and Neil, 1987). Therefore, to investigate the impact of parasitism on mate choice, males and females from 29 precopula pairs were weighed, male parasite status determined and the male/female pair size ratios compared for parasitized and unparasitized males.

The association of parasitism with changes in host behaviour

In January 2002, Downhill River samples of G. d. celticus were collected and maintained in the laboratory in aerated stock tanks (36×50×15 cm deep) at 12 °C, 9[ratio ]15 h light[ratio ]dark with source water, flora and fauna. To explore the association between parasitism and host activity, unparasitized, low and high parasite burden G. d. celticus (n=30 each) were placed individually in circular containers (8 cm diameter) with lines drawn down the middle of the container bottom. Activity level was determined as the number of crosses of the line an individual made during 3 consecutive 2-min periods. Mean numbers of line crossings were examined in a two-factor ANOVA with respect to ‘parasite status’ and ‘time’ (repeated measure). Post-hoc tests were Fisher's Protected Least Significant Difference (FPLSD).

To investigate the association between this parasitism and host predation on A. aquaticus, in February 2001, G. d. celticus and A. aquaticus were collected from the Articlave River (UK grid ref.: C785338) and maintained as above. Plastic containers (6 cm diameter) were supplied with 50 ml of filtered Articlave water, substrate (1 cm hollow stone tubes), vegetation (Elodea canadensis, elm leaves) and pelleted catfish food. Over 10 days, cannibalism and predation were assessed twice daily in all intra- and interspecific dyadic combinations of adult and juvenile G. d. celticus and A. aquaticus (n=30 for each combination, except no juvenile G. d. celticus were found parasitized). Concurrently, mortality of single individuals (n=30 each case) was assessed twice daily over 10 days to allow confidence in concluding that cannibalistic/predatory interactions had occurred, as opposed to scavenging following death due to other causes.

To investigate the association between P. mulleri infection and behavioural interactions between G. d. celticus and the invasive G. pulex, we examined how predation of newly moulted G. pulex females by G. d. celticus males may be influenced by parasitism of the latter. Interspecific predation between these species is strongly influenced by moult stage and sex, with newly moulted females being highly vulnerable to male predation. Mating also coincides with the moult, with females often being attacked and predated despite being paired in precopula with their mate (Dick, Montgomery and Elwood, 1993). In December 2001, G. pulex from the Lagan river (UK grid ref.: J299679) were maintained in aerated tanks of source water at 12 °C and a 9[ratio ]15 h light[ratio ]dark cycle. Precopula pairs were separated by placement on tissue paper until the males released their females, then each was placed into individual plastic containers (8 cm diameter) containing 150 ml of 50% Downhill and 50% Lagan water. Females were inspected twice daily and newly moulted individuals presented to male conspecifics and allowed to form precopula for 30 min. Then, a male G. d. celticus, either unparasitized, with low or with high parasite burden (n=30 each, matched for size by eye to reduce stress and weighed after experiment to confirm equivalency), was introduced to a pair. Activities of the 3 animals were then recorded for 20 min (see Dick, Elwood and Montgomery, 1995 for details of behaviours recorded), using the Observer Software package (Anon, 1995). Animals were then left together for a further 12 h and any deaths of the amphipods recorded. Data were analysed using one-factor ANOVAs and FPLSD tests.

RESULTS

Parasitism in the field

Live adult G. d. celticus abundance ranged from 90 to 425 animals per sample, with lowest abundance in spring/summer and highest in autumn/winter (Fig. 1A). There was a distinct die-off of adults in July/August of both years. Live abundance was negatively correlated with numbers of dead animals recovered in samples (Spearman Rank Correlation rs=−0·71, n=22, P<0·001; see also Fig. 1A).

Fig. 1. (A) Gammarus duebeni celticus abundance (live and dead animals) and (B) prevalence of parasitized G. d. celticus and frequency of high burden parasitized G. d. celticus in the Downhill River population over a 22-month period (live animals only).

Overall parasite prevalence (low and high burdens combined) ranged from 8·5% in summer to 44·9% in winter (Fig. 1B) and this was positively correlated with host live abundance (rs=0·44, n=22, P<0·05; see also Fig. 1A and B). However, considering the percentage of parasitized G. d. celticus that had high parasite burden, there was a negative correlation with G. d. celticus live abundance (rs=−0·82, n=22, P<0·001) and a positive correlation with numbers of dead (rs=0·74, n=22, P<0·001; see also Fig. 1A and B).

For unpaired adults, the sex ratio was 1[ratio ]2 male[ratio ]female, with a significantly lower frequency of males parasitized than females (27% vs 62%, χ2=43·9, D.F.=1, P<0·0001). Parasitized males carried relatively smaller females (t=3·7, D.F.=27, P<0·001; Fig. 2).

Fig. 2. Mean (+S.E.) ratio of male: female weight in precopula pairs with unparasitized and parasitized guarding males.

The association of parasitism with changes in host behaviour

Host activity was significantly lower with increasing levels of parasitism (F2,87=12·1, P<0·0001, all pairwise FPLSD comparisons P<0·001; Fig. 3).

Fig. 3. Mean (+S.E.) number of line crossings of unparasitized, low and high burden parasitized Gammarus duebeni celticus.

Mortality of single G. d. celticus and A. aquaticus was zero. We thus conclude that any deaths in dyadic combinations were due to intra- or interspecific interactions. No animals moulted during the experiment. There was no cannibalism in A. aquaticus and only juvenile G. d. celticus were cannibalized by adults (3/30 by unparasitized adults, 0/30 by parasitized adults). There was no A. aquaticus predation of G. d. celticus. However, unparasitized adult G. d. celticus killed adult A. aquaticus, albeit at a low frequency (3/30), whereas parasitized adult G. d. celticus did not kill any adult A. aquaticus. Predation of juvenile A. aquaticus by unparasitized adult G. d. celticus was severe and significantly more frequent than that by parasitized G. d. celticus (16/30 compared to only 2/30, P<0·0001).

There was no significant difference in weight between unparasitized, low and high parasite burden male G. d. celticus (F2,87=1·4, N.S.). Parasitized G. d. celticus made significantly fewer antennal contacts with precopula pairs of G. pulex than unparasitized individuals (F2,87=58·9, P<0·0001) and this decreased significantly with parasite load (FPLSD, all P<0·001). Parasitized G. d. celticus tugged significantly less at the female and made significantly fewer grab/attacks at the guarding male (F2,87=6·8, P<0·001, Fig. 4A and F2,87=8·7, P<0·0001, Fig. 4B), but there were no differences between G. d. celticus with low and high parasite burden. The frequencies of defensive kicks by guarding G. pulex males and females was significantly lower when parasitized G. d. celticus were present (F2,87=27·1, P<0·0001 and F2,87=4·8, P<0·01) and this decreased as parasite burden of attacking G. d. celticus increased for guarding males (all P<0·05) but not for guarding females.

Fig. 4. Mean (+S.E.) frequency of (A) tugs at the guarded female Gammarus pulex by the attacking male Gammarus duebeni celticus and (B) grab/attacks at the guarding male G. pulex by the attacking male G. d. celticus.

In 88 of the 90 observations, aggressive interactions were observed between the native male and the invader pair, although after a further 12 h, none of the female G. pulex had been predated by the G. d. celticus.

DISCUSSION

The G. d. celticus population monitored in this study showed a distinct cycle of abundance that could be attributed to many causes. That the parasite may be one causal factor, however, is indicated by the decrease in parasite prevalence as host numbers declined. That is, parasitized animals seem more likely to die. In addition, marked declines in the host population coincided with large proportions of the parasitized animals harbouring the heaviest burdens. Schrag and Rollinson (1994) found that the weakest individuals in a population were most likely to acquire parasites. However, MacNeil et al. (2003b) found that P. mulleri infection could be experimentally produced in G. d. celticus, with healthy animals being easily infected, so the evidence suggests infection occurs to any individual. Clearly, co-incidental rises in temperatures, change in food supply and predator activity may all influence host abundance, but it is reasonable to conclude that P. mulleri is partially involved in structuring its host population. Indeed, crangonid sand shrimps infected with Pleistophora crangoni experience a winter and spring fall in prevalence associated with the related deaths of adult hosts (Breed and Olson, 1977). Seasonality of microsporidian parasites has also been attributed to environmental stresses on fish hosts, such as elevated summer water temperature and associated decreases in dissolved oxygen concentration (Chen and Power, 1972; Cali and Takvorian, 1991) and to opportunities for horizontal transmission to intermediate hosts in mosquitoes (Andreadis, 2002).

The study population also had a strongly female biased sex ratio. Gammarus duebeni harbours vertically transmitted microsporidia that cause feminization of broods (Dunn and Smith, 2001) and this may be the case here. In addition, however, a higher proportion of females than males harboured P. mulleri. This may reflect a higher susceptibility of females to infection or a higher mortality of infected males. Parasitism was associated with changes in parameters that influence host fitness. For instance, males do appear to incur a fitness cost of parasitism, since they carried smaller and hence less fecund females in precopula than did unparasitized males (see also Elwood et al. 1987). This may be due to the direct pathogenicity of the parasite in the abdominal musculature (Terry et al. 2003), impairing the physical exertion of precopula pairing, reduced mate searching abilities or female choice. Parasitism was associated with a significant reduction in the activity of the host suggesting that the parasites may cause muscle damage that inhibits locomotion (Terry et al. 2003). Indeed, we found experimentally that as parasite burden increased, host activity levels decreased. Similarly, reduced mobility and partial paralysis has been observed in the prawn Penaeus indicus infected with cotton shrimp disease (Ramasamy, Jaykumar and Brennan, 2000). Muscle fibres became crystallized and were eventually destroyed and replaced by spores of the invading microsporidian, Thelohania sp.

Gammarus spp. are predators of many other macroinvertebrates (e.g. chironomids, mayflies and mysids; MacNeil et al. 1997) and as such can influence community structure (Kelly et al. 2003). Our data indicate that predation pressure by G. d. celticus on A. aquaticus is moderated by P. mulleri infection, with infected G. d. celticus less able to prey upon A. aquaticus juveniles. Reduced aggression and predation have also been reported in the predatory mite, Phytoseiulus persimilis, infected with the microsporidian, Microsporidium phytoseiuli (Bjornson and Keddie, 1999).

Populations of G. d. celticus are actively replaced by invading G. pulex, principally through differential mutual predation that appears to be mediated by parasitism (MacNeil et al. 2003a). The present study suggests that parasitism by P. mulleri is associated with a reduction in the capacity of G. d. celticus to predate guarded females of the invader G. pulex. This may again be a direct result of the effects of the parasite on the host abdominal musculature. Further, since populations of G. d. celticus show wide fluctuations in density that may be partially driven by the parasite, the invasion success of G. pulex may be facilitated through decreased biotic resistance from the native at certain times of year. In addition, since P. mulleri does not transmit to invading G. pulex (MacNeil et al. 2003a,b), the parasite clearly disadvantages the native over the invader and may indeed speed the replacement of the former. We thus encourage further studies of the pervasive influence of microsporidia in mediating interspecific interactions and hence broader features of community organization and trophic function.

The authors are grateful to G. Riddell for her assistance with fieldwork and laboratory work. Thanks also to C. Aughey and S. Crean. Thanks to two anonymous referees, whose comments greatly improved this manuscript. This project was supported by NERC grant GR3/12871.

References

REFERENCES

Andreadis, T. G. ( 2002). Epizootiology of Hyalinocysta chapmani (Microsporidia: Thelohaniidae) infections in field populations of Culiseta melanura (Diptera: Culicidae) and Orthocyclops modestus (Copepoda: Cyclopidae): a three-year investigation. Journal of Invertebrate Pathology 81, 114–121.CrossRefGoogle Scholar
Anonymous ( 1995). The Observer. Noldus Information Technology. Base Package for Windows, Version 3.0. Wageningen, The Netherlands.
Bjornson, S. and Keddie, B. A. ( 1999). Effects of Microsporidium phytoseiuli (Microsporidia) on the performance of the predatory mite, Phytoseiulus persimils (Acari: Phtytoseiidae). Biological Control 15, 153–161.CrossRefGoogle Scholar
Breed, G. M. and Olson, R. E. ( 1977). Biology of the microsporidian parasite Pleistophora crangoni n. sp. in three species of crangonid sand shrimps. Journal of Invertebrate Pathology 30, 387–405.Google Scholar
Cali, A. and Takvorian, P. M. ( 1991). The incidence of Glugea stephani (Protozoa: Microsporidia) in winter flounder, Pseudopleuronectes americanus, from the New York-New Jersey lower bay complex and factors influencing it. Canadian Journal of Zoology 69, 317–321.CrossRefGoogle Scholar
Chen, M. and Power, G. ( 1972). Infection of American smelt in Lake Ontario and Lake Erie with the microsporidian parasite Glugea hertwigi (Weissenberg). Canadian Journal of Zoology 50, 1183–1188.CrossRefGoogle Scholar
Dick, J. T. A., Elwood, R. W. and Montgomery, W. I. ( 1995). The behavioural basis of a species replacement: differential aggression and predation between the introduced Gammarus pulex and the native G. duebeni celticus (Amphipoda). Behavioural Ecology and Sociobiology 37, 393–398.CrossRefGoogle Scholar
Dick, J. T. A., Montgomery, W. I. and Elwood, R. W. ( 1993). Replacement of the indigenous amphipod Gammarus duebeni celticus by the introduced G. pulex: differential cannibalism and mutual predation. Journal of Animal Ecology 62, 79–88.Google Scholar
Dunn, A. M. and Smith, J. E. ( 2001). Microsporidian life cycles and diversity: the relationship between virulence and transmission. Microbes and Infection 3, 381–388.CrossRefGoogle Scholar
Elwood, R. W., Gibson, J. and Neil, S. ( 1987). The amorous Gammarus: size assortative mating in G. pulex. Animal Behaviour 35, 1–6.CrossRefGoogle Scholar
Hanley, K. A., Vollmer, D. M. and Case, T. J. ( 1995). The distribution and prevalence of helminths, coccidian and blood parasites in 2 competing species of gecko – implications for apparent competition. Oecologia 102, 220–229.CrossRefGoogle Scholar
Keeling, P. J. and Fast, N. M. ( 2002). Microsporidia: biology and evolution of highly reduced intracellular parasites. Annual Review of Microbiology 56, 93–116.CrossRefGoogle Scholar
Kelly, D., Dick, J. T. A., Montgomery, W. I. and MacNeil, C. ( 2003). Differences in composition of macroinvertebrate communities with invasive and native Gammarus spp. (Crustacea: Amphipoda). Freshwater Biology 48, 306–315.Google Scholar
MacNeil, C., Dick, J. T. A. and Elwood, R. W. ( 1997). The trophic ecology of freshwater Gammarus (Crustacea: Amphipoda). Biological Reviews of the Cambridge Philosophical Society 74, 375–395.Google Scholar
MacNeil, C., Dick, J. T. A., Hatcher, M. J., Fielding, N. J., Hume, K. D. and Dunn, A. M. ( 2003 b). Parasite transmission and cannibalism in an amphipod (Crustacea). International Journal for Parasitology 33, 795–798.Google Scholar
MacNeil, C., Dick, J. T. A., Hatcher, M. J., Terry, R. S., Smith, J. E. and Dunn, A. M. ( 2003 a). Parasites mediate mutual predation between native and invasive species. Proceedings of the Royal Society of London, B 270, 1309–1314.Google Scholar
Minchella, D. J. and Scott, M. E. ( 1991). Parasitism: a cryptic determinant of animal community structure. Trends in Ecology and Evolution 6, 250–254.CrossRefGoogle Scholar
Prenter, J., MacNeil, C., Dick, J. T. A. and Dunn, A. M. ( 2004). Roles of parasites in animal invasions. Trends in Ecology and Evolution 19, 385–390.CrossRefGoogle Scholar
Price, P. W., Westloy, M. and Rice, B. ( 1988). Parasite-mediated competition: some predictions and tests. American Naturalist 131, 544–555.CrossRefGoogle Scholar
Ramasamy, P., Jaykumar, R. and Brennan, G. P. ( 2000). Muscle degeneration associated with cotton shrimp disease of Penaeus indicus. Journal of Fish Disease 23, 77–81.CrossRefGoogle Scholar
Schrag, S. J. and Rollinson, D. ( 1994). Effects of Schistosoma haematobium infection on reproductive success and male out-crossing ability in the simultaneous hermaphrodite Bulinus truncatus (Gastropoda: Planorbidae). Parasitology 108, 27–34.CrossRefGoogle Scholar
Terry, R. S., MacNeil, C., Dick, J. T. A., Smith, J. E. and Dunn, A. M. ( 2003). Resolution of a taxonomic conundrum: an ultrastructural and molecular description of the life cycle of Pleistophora mulleri (Pfeiffer 1895; Georgevitch 1929). Journal of Eukaryotic Microbiology 50, 266–274.CrossRefGoogle Scholar
Tompkins, D. M., White, A. R. and Boots, M. ( 2003). Ecological replacement of native red squirrels by invasive greys driven by disease. Ecology Letters 6, 189–196.CrossRefGoogle Scholar
Weiss, L. M. ( 2001). Microsporidia: emerging pathogenic protists. Acta Tropica 78, 89–102.CrossRefGoogle Scholar
Figure 0

Fig. 1. (A) Gammarus duebeni celticus abundance (live and dead animals) and (B) prevalence of parasitized G. d. celticus and frequency of high burden parasitized G. d. celticus in the Downhill River population over a 22-month period (live animals only).

Figure 1

Fig. 2. Mean (+S.E.) ratio of male: female weight in precopula pairs with unparasitized and parasitized guarding males.

Figure 2

Fig. 3. Mean (+S.E.) number of line crossings of unparasitized, low and high burden parasitized Gammarus duebeni celticus.

Figure 3

Fig. 4. Mean (+S.E.) frequency of (A) tugs at the guarded female Gammarus pulex by the attacking male Gammarus duebeni celticus and (B) grab/attacks at the guarding male G. pulex by the attacking male G. d. celticus.