Introduction

Parasites can play important roles in biological invasions: invading species may bring with them parasites or diseases which may detrimentally affect native species (Ohtaka et al. 2005; Rushton et al. 2000), or may themselves acquire parasites from their new environment (Bauer et al. 2000; Krakau et al. 2006). Alternatively invading species may lose their parasites, potentially giving them an advantage over native species (Torchin et al. 2003; Torchin et al. 2001). Parasites have been shown to be important mediators of interspecific interactions (Hatcher et al. 2006): they may confer a competitive advantage to the host species (Yan et al. 1998), alter dominance relationships and predation hierarchies (MacNeil et al. 2003a), and may promote species exclusion or coexistence (MacNeil et al. 2003b; Prenter et al. 2004). By mediating native-invader interactions, parasites can play a key role in the outcome of a biological invasion (MacNeil et al. 2003a; MacNeil et al. 2003b; Prenter et al. 2004). For example, in Northern Ireland, the acanthocephalan parasite Echinorynchus truttae reduces the predatory impact of the invasive amphipod Gammarus pulex on the native G. duebeni celticus (MacNeil et al. 2003b).

The North American Signal Crayfish, Pacifastacus leniusculus (Dana), has become established throughout Britain as a result of escapes from farms (Holdich et al. 2004). The species is highly invasive and commonly leads to the displacement of Britain’s only native crayfish Austropotamobius pallipes (Lereboullet) (Bubb et al. 2006; Kemp et al. 2003) As a result, populations of A. pallipes are now concentrated in central and northern England (Souty-Grosset et al. 2006) where they are of global importance, representing the densest concentrations of the species within Europe (Holdich 2003). The mechanism by which A. pallipes is displaced varies between populations. In some cases, the native species is displaced through competitive interactions, (Bubb et al. 2006); however the exact mechanism by which this occurs is unclear. In many water courses in the south of England, extinction of A. pallipes has resulted from crayfish plague (Kemp et al. 2003). The invasive crayfish, P. leniusculus, commonly acts as a reservoir for Aphanomyces astaci (the causative agent of crayfish plague), which is fatal to the native species (Holdich 2003).

Also of interest are two further parasites. The microsporidian parasite Thelohania contejeani (Henneguy), infects Austropotamobius pallipes causing porcelain disease and is the most widely recorded parasitic infection of this species (Alderman and Polglase 1988). Whilst the pathology of T. contejeani is not as severe as that of crayfish plague it can be a serious threat within crayfish aquaculture (Edgerton et al. 2002) and may cause changes in the ecology of its host through changes in diet (Chartier and Chaisemartin 1983); however the consequences of infection by many pathogen groups in European freshwater crayfish are largely poorly understood (Edgerton et al. 2004). Microsporidia are widespread in crustacean hosts (Edgerton et al. 2002; Terry et al. 2004) and can cause significant mortality (Alderman and Polglase 1988). A second parasite, the protist Psorospermium haeckeli (Haeckel) infects crayfish and has recently been isolated from A. pallipes (Rogers et al. 2003) and Pacifastacus leniusculus (Dieguez-Uribeondo et al. 1993). The influence of these parasites upon native/invasive interactions in crayfish is unknown.

In the UK, Yorkshire is a stronghold for A. pallipes: although P. leniusculus is present within the county in substantial numbers, it has not yet displaced many native populations and mixed populations do exist (Peay and Rogers 1999). Here we investigate possible competitive interactions by comparing the sizes of native and invading individuals in single species versus mixed species populations. Secondly we use PCR screening and sequence analysis to compare parasite diversity in the native and invasive crayfish, focusing on microsporidian parasites.

Materials and methods

Animal collection and measurement

A total of seven A. pallipes populations, four P. leniusculus populations and three mixed species sites were surveyed between June and August 2005 (Table 1). Sites in the Wharfe catchment were similar to each other and were typified by boulders and smaller stones overlying gravel. Sites in the Dearne catchment (including Cawthorne Dike) were also similar to each other and were typified by boulders and small stones overlying deep silt. Sites were surveyed for crayfish using a standardised manual survey of selected refuges within a site (Peay 2003). Selection of similar sized refuges at each site ensures no size bias during collection (Peay 2003). For each crayfish individual captured we recorded the species, size (carapace length) and sex. In addition any signs of disease, breeding or moult were recorded: microsporidian infections when at high burden typically cause opacity of muscle tissues as a result of spore replication and muscle pathology (Alderman and Polglase 1988); Aphanomyces astaci can be identified by the appearance of brown melanisations on the exoskeleton of the infected animal (Alderman and Polglase 1988). Following assessment, crayfish were set aside to prevent duplication of records, until the population assessment of the site had been completed. All Austropotamobius pallipes were then released; P. leniusculus were stored at −20°C.

Table 1 Field sites sampled during the study
Full size table

Statistical analysis

Statistical analyses were conducted using R version 4.2.1. (www.R-project.org). Linear mixed effects models (LMM) were fitted to the size distribution data for each species separately using Maximum Likelihood fits. Size was used as the dependent variable with population (single vs. mixed species) and sex as fixed factors; site identity was included in the model as a random factor to control for any inter-site differences in size composition.

In order to determine whether parasite prevalence differed between sexes or sizes of P. leniusculus, a Generalized Linear Model (GLM) with binomial error distributions was fitted to the data. Microsporidian presence or absence was used as the dependent variable with size and sex as fixed factors.

Non-significant fixed factors were removed from the maximal models in a stepwise fashion until only factors significant at the 5% level remained.

Screening for microsporidian parasites

Fifty-two P. leniusculus from the field collection (Table 1) were screened for microsporidia (Table 2). As A. pallipes is classified as vulnerable (IUCN 2004) and protected under Schedule 5 of the Wildlife and Countryside Act (1981), we did not screen live animals collected from the field; however twelve dead A. pallipes obtained from sites detailed in Table 1 were screened for microsporidia. Sampling was carried out towards the end of the breeding season when most young have hatched and dispersed (Holdich 2003). However, one female P. leniusculus still had two eggs attached; as many microsporidia are vertically transmitted (Dunn and Smith 2001) we also screened these to test for the presence of vertically transmitted parasites.

Table 2 Results of PCR screen for microsporidian infection in P. leniusculus
Full size table

Crayfish tissue (approximately 0.25 g) was dissected from tail muscle between the third and fourth pleonites, being careful to avoid sampling gut tissue. Eggs from the single gravid female sampled were collected and homogenised. DNA was extracted using a chloroform extraction described by Doyle and Doyle (1987) with modifications described in McClymont et al. (2005).

PCR of the host cytochrome C oxidase 1 (CO1) gene was used to confirm the quality of the DNA extraction before PCR for microsporidian SSU rDNA was carried out. Primers used for detection of host DNA were LCO1490 and HCO2198, which amplify a fragment of the CO1 gene (Folmer et al. 1994). The CO1 PCR protocol was as described in McClymont et al. (2005). Positive controls containing DNA extracted from microsporidium infected crayfish muscle stored in ethanol and negative controls containing deionised water in place of DNA were included for each reaction; the total reaction volume was 25 μl.

Three primer sets were used for detection of microsporidian SSU rDNA. V1f (Vossbrinck and Woese 1986) and 1492r (Weiss et al. 1994) are specific for T. contejeani (Lom et al. 2001), whilst both V1f and 530r (Baker et al. 1995), and 18sf (Baker et al. 1995) and 964r (McClymont et al. 2005) are general microsporidian primers. The PCR reaction mixture and protocols are as described by McClymont et al. (2005); annealing temperatures and PCR product lengths are shown in Table 3.

Table 3 PCR annealing temperatures and approximate expected product length for primers used in parasite detection and for sequencing
Full size table

Positive controls containing DNA extracted from microsporidium infected crayfish muscle stored in ethanol and negative controls containing deionised water in place of DNA were included for each reaction; the total reaction volume was 25 μl for initial parasite detection. PCR protocols were all carried out on a Hybaid Omn-E Thermal Cycler (Hybaid Ltd, Waltham, Massachusetts, USA).

Sequencing and phylogenetic analysis of microsporidia

Different primer sets gave positive bands in different individuals suggesting the presence of more than one microsporidian parasite within P. leniusculus. Therefore additional primers were used in order to obtain longer sequences: these were 580r (Vossbrinck et al. 1993), Ha3Bf (Gatehouse and Malone 1998), HG4r (Gatehouse and Malone 1998), 350f (Weiss and Vossbrinck 1998), HG4f (Gatehouse and Malone 1998), 1342r (McClymont et al. 2005) and 350r (5′-CCAAGGACGGCAGCAGGCGCGAAA-3′), together with new primers Thelof (5′-TCGTAGTTCCGCGCAGTAAACTA-3′) and BACF (5′-ATATAGGAACAGATGATGGC-3′). Annealing temperatures for all primer combinations are given in Table 3. Where PCR products were to be sequenced the amounts of reagents in the reaction mixture were doubled to give a total reaction volume of 50 μl.

50 μl of each PCR product were electrophoresed through a 2% agarose TAE gel in standard TAE buffer, stained with ethidium bromide and visualised by UV light to ensure successful amplification of the PCR product. PCR products were excised from the gel and purified using a QIAQuick Gel Purification Kit (Qiagen, Crawley, UK) and were sequenced on an ABI 3130xl capillary sequencer at the University of Leeds.

The closest matching sequence to each sequence generated within this study was determined using the NCBI-BLAST database (Altschul et al. 1997) and a percentage sequence similarity calculated using the pairwise alignment function in BioEdit (Hall 2005).

Screening for Aphanomyces astaci and Psorospermium haeckeli

In addition, a subset of 15 Pacifastacus leniusculus and three Austropotamobius pallipes from the field collection were screened for the presence of Aphanomyces astaci and of Psorospermium haeckeli.

Tissue was dissected from the eye to screen for the presence of A. astaci as in the early stages of the infection mycelium are known to be present within the cornea (Vogt 1999). DNA extraction was performed and confirmed as described previously. Primers 525 and 640 were used to screen for A. astaci, with an expected product length of 115 bp (Oidtmann et al. 2004). The reaction mixture comprised 0.625 U of GoTaq Taq polymerase and 5 μl 5× GoTaq buffer (giving a final concentration of 1.5 mM MgCl2 per reaction) (Promega, Southampton, UK), 0.04 mM dNTPs, 10 pmol of each primer, 1 μl DNA and deionised water in a total reaction volume of 25 μl. No positive control material was available; a negative control containing deionised water in place of DNA was included for each PCR reaction. The PCR protocol is as described in Oidtmann et al. (2004).

To screen for Psorospermium haeckeli, tissue was dissected from the subepidermal connective tissue as high parasite burdens have been reported from this tissue type (Henttonen 1996). DNA extraction was performed and confirmed as described previously. Primers Pso-1 (Bangyeekhun et al. 2001) and ITS-4 (White et al. 1990) were used to screen for P. haeckeli with expected product lengths of 1300 or 1500 bp (Bangyeekhun et al. 2001). The reaction mixture comprised 1.25 U of GoTaq Taq polymerase, 5 μl 5× GoTaq buffer (Promega, Southampton, UK), 2 mM MgCl2, 0.08 mM dNTPs, 20 pmol of each primer, 1 μl of DNA and deionised water in a total reaction volume of 25 μl. No positive control was available; a negative control containing deionised water in place of DNA was included for each PCR. The PCR protocol is as described in Bangyeekhun et al. (2001).

Results

Sizes of animals in single and mixed populations

Austropotamobius pallipes

Following stepwise deletion of non-significant fixed effects from the Maximal model, population composition (single vs. mixed species) was the only significant term remaining in the Minimum Adequate Model (LMM, F 1,73, P = 0.025) indicating a significant difference in size composition of single and mixed species populations. The mean size of A. pallipes was 28.5 mm in single species populations and 22.5 mm in mixed populations (Fig. 1).

Fig. 1
figure 1

Size distributions of Austropotamobius pallipes in single species and mixed species populations. A. pallipes individuals in single species populations were significantly larger than those in mixed species populations (LMM, F 1,73, P = 0.025)

Full size image

Pacifastacus leniusculus

Following stepwise deletion of non-significant fixed effects, population composition (single vs. mixed species) was the only significant term remaining in the Minimum Adequate Model (LMM, F 1,72, P = 0.028). P. leniusculus individuals in mixed populations were significantly larger than their counterparts in single species populations with a mean size of 36.3 mm in single species populations and 46.0 mm in mixed species populations (Fig. 2). Very few juveniles were observed in the mixed species sites.

Fig. 2
figure 2

Size distributions of Pacifastacus leniusculus in single species and mixed species populations. P. leniusculus individuals in single species populations were significantly smaller than individuals in mixed species populations (LMM, F 1,72, P = 0.028)

Full size image

Microsporidian parasites

All 12 A. pallipes individuals tested showed clinical signs of microsporidian infection through an opacity of the abdominal musculature; these all tested positive for microsporidian infection through PCR screening. As we were only able to screen dead individuals from the field, we were unable to estimate the prevalence of microsporidian infection for this species.

The prevalence of microsporidian infection in P. leniusculus ranged from 0.26 to 0.75, with an overall prevalence across all populations of 0.38 (Table 2). Six of the 20 infected individuals showed clinical signs of infection through an opacity of the abdominal musculature; one of these was dead when collected. There was no significant difference between the frequency of infection of males versus females (GLM, P 47 = 0.181) and there was no significant difference in sizes of infected versus uninfected individuals (GLM, P 48 = 0.831).

Parasite sequences

We obtained multiple sequences from four distinct microsporidian parasite species (Table 4). Three of these parasites, Bacillidium sp. PLFB32, Microsporidium sp. PLWB7A and Vittaforma sp. PLDH3, had not previously been reported from crayfish hosts and represent novel microsporidian sequences; the fourth, Thelohania contejeani, despite having been previously recorded in crayfish, had not been sequenced from either of the two study species.

Table 4 Summary of microsporidian parasite diversity in A. pallipes and microsporidian diversity and prevalence P. leniusculus
Full size table

Forty-four sequences from 29 individuals were 98–100% identical to T. contejeani isolated from the crayfish Astacus fluviatilis in France (Lom et al. 2001). These sequences were obtained from 17 P. leniusculus and 12 Austropotamobius pallipes. We detected two strains of T. contejeani within each crayfish species, corresponding to strains TcC2 and TcC3 described by Lom et al. (2001). We found strain TcC2 in seven individuals: three A. pallipes and four P. leniusculus. We sequenced strain TcC3 from 18 individuals: eight A. pallipes and 10 P. leniusculus. Four samples were not sequenced across the variable region and so could belong to either strain. In three cases we sequenced both strains from the same host, twice in A. pallipes and once in P. leniusculus.

Two sequences isolated from one P. leniusculus had 97% sequence similarity to Bacillidium vesiculoformis, a species that has to date only been described from the oligochaete worm Nais simplex in Scotland. One sequence isolated from a P. leniusculus egg had 75% sequence similarity to B. vesiculoformis; the parent crayfish tested negative for microsporidian infection.

Two sequences isolated from a single P. leniusculus host had 95% sequence similarity to Microsporidium sp. CRANFA isolated from the amphipod crustacean Crangonyx floridanus in Florida (Galbreath 2005), and 93% sequence similarity to a Vittaforma-like parasite isolated from a human host (Sulaiman et al. 2003).

We found no clinical/visible signs of Aphanomyces astaci infection in any of the individuals sampled. No evidence was found for infection by either A. astaci or Psorospermium haeckeli in any of subset the individuals screened for these parasites by PCR.

Discussion

Competitive interactions

In mixed populations the size distributions of both species differ from those in single species populations. Austropotamobius pallipes tend to be smaller in mixed populations (Fig. 1) whereas Pacifastacus leniusculus tend to be larger (Fig. 2). Displacement mechanisms proposed in other native-invader crayfish systems include reproductive interference (Westman et al. 2002); competitive exclusion from refuges resulting in differential predation (Vorburger and Ribi 1999); and differential susceptibility to diseases (Alderman and Polglase 1988).

In Finland, where P. leniusculus displaces the native Astacus astacus, it is thought that reproductive interference by dominant P. leniusculus males results in the majority of A. astacus females producing only sterile eggs (Westman et al. 2002). Our data provide no support for this mechanism of displacement in our study system as smaller Austropotamobius pallipes were more common in mixed populations (Fig. 1); this is in direct contrast to the pattern of fewer small A. pallipes in mixed populations that would be predicted by reproductive interference (Westman et al. 2002).

Our data show large A. pallipes to be under-represented in mixed populations (Fig. 1), which may reflect competitive exclusion by the larger (Lowery 1988) and more dominant (Vorburger and Ribi 1999) invader from limited refuges (Bubb et al. 2006), since small P. leniusculus and large A. pallipes overlap in size (Figs. 1, 2). P. leniusculus has been shown to oust other crayfish species from refuges (Söderbäck 1995) which would leave larger A. pallipes more vulnerable to predation (Söderbäck 1994, after Söderbäck 1992) and result in the reduction of large A. pallipes in the mixed populations seen within our study.

The absence of juvenile P. leniusculus from mixed populations (Fig. 2) is interesting, and implies that A. pallipes may in fact be influencing the population structure of the invading species. The moulting of juvenile P. leniusculus is synchronized, resulting in reduced intraspecific cannibalism (referenced in Ahvenharju et al. 2005). However, interspecific predation by the native A. pallipes (Gil-Sánchez and Alba-Tercedor 2006) as well as other predators such as fish (Söderbäck 1992) may underpin the observed reduction in juvenile P. leniusculus in mixed populations.

Parasitism in native and invasive crayfish

Four species of microsporidia were detected in the invasive crayfish P. leniusculus. In contrast, only one microsporidian parasite was detected from A. pallipes although the sample size was small. The overall prevalence of microsporidian infection in P. leniusculus was 38% (Table 2). This prevalence is higher than previous reports of visible microsporidiosis in A. pallipes in Britain (9%, (Brown and Bowler 1977); 26% (Rogers et al. 2003); 30% (Evans and Edgerton 2002)), France (0–8%, Chartier and Chaisemartin 1983) and Spain (1%, Dieguez-Uribeondo et al. 1993), probably reflecting a higher detection efficiency by PCR.

The T. contejeani sequences we obtained were identical to those previously isolated from Astacus fluviatilis (Genbank accession numbers AF492593 and AF492594, Lom et al. 2001). This is, to our knowledge, the first molecular confirmation of T. contejeani infecting P. leniusculus, as well as the first report of the parasite in an invasive species in Europe. Whilst T. contejeani has previously been reported from Austropotamobius pallipes in the UK (Brown and Bowler 1977; Edgerton et al. 2002; Rogers et al. 2003), these reports were based on light microscopy and lack the ultrastructural or molecular information to confirm species identity (Dunn and Smith 2001). This is the first molecular confirmation of the presence of T. contejeani infecting A. pallipes.

The presence of T. contejeani in the invasive P. leniusculus leads to the question of how the parasite has come to infect this species. Firstly, P. leniusculus may have brought the parasite with it from its native range. T. contejeani has been reported from a number of crayfish hosts (Graham and France 1986; Quilter 1976), and there is a single report of T. contejeani from P. leniusculus in its native range in California (McGriff and Modin 1983); but identification is based on spore size, and molecular or ultrastructural confirmation is lacking. The pattern of infection in the current study leads us to suggest that it is more likely that P. leniusculus in the UK has acquired T. contejeani from the native host. T. contejeani was detected in A. pallipes only sites, in mixed sites and in sites where only P. leniusculus occurred. Furthermore, identical sequences were found in the native and invading species (Table 4). These data fit a pattern of transmission from the native A. pallipes to the invading species in mixed sites. Detailed studies of the fitness effects of T. contejeani and its mode of transmission within and between crayfish species are required.

In addition, our T. contejeani sequences had 98–100% sequence similarity to the unclassified microsporidium, Microsporidium sp. JES2002H, which was detected in three species of amphipod in France (Terry et al. 2004). We suggest that Microsporidium sp. JES2002H and T. contejeani may be the same species, although confirmation awaits ultrastructural analysis of Microsporidium sp. JES2002H.

In one P. leniusculus we found a microsporidium sequence with 97% sequence similarity to Bacillidium vesiculoformis, a parasite previously described from the oligochaete worm Nais simplex in Scotland, UK (Morris et al. 2005). The sequence similarity indicates that the parasite is likely to be in the same genus as B. vesiculoformis; however further molecular and morphological analysis would be required to confirm this. This is the first record of a Bacillidium spp. in crayfish and supports Morris et al’s (2005) suggestion that B. vesiculoformis is a generalist parasite.

The Vittaforma-like parasite sequenced from P. leniusculus had closest sequence similarity (95%) to an unidentified Microsporidium sp. CRANFA sequenced from Crangonyx floridanus in Florida (Galbreath 2005), which suggests that this may be a parasite originating in the native range of P. leniusculus.

We also sequenced a novel parasite from a P. leniusculus egg. This parasite was unlike other microsporidia and had the closest sequence similarity (75%) to B. vesiculoformis. The presence of the parasite in the egg suggests vertical transmission (Terry et al. 2004), widespread amongst microsporidian parasites (Dunn and Smith 2001). Muscle tissue from the mother tested negative for microsporidian infection, but ovarian tissue was not screened in this study.

We found no evidence of Aphanomyces astaci or Psorospermium haeckeli within our study populations (although our results should be treated with caution owing to the absence of positive control material). The absence of crayfish plague may explain the persistence of mixed species populations in Yorkshire and highlights the need for vigilance in preventing plague from spreading into these rivers.

In summary, our size distribution data are in accord with a pattern of competitive exclusion of Austropotamobius pallipes from refuges leading to differential predation. We provide the first molecular evidence for the presence of the microsporidian parasite T. contejeani in both A. pallipes and Pacifastacus leniusculus. We also detected three novel microsporidian sequences in P. leniusculus. This raises the question of the effects of these parasites on host fitness as well as their potential influence on native - invader interactions.