INTRODUCTION
Host specificity is an essential component of parasite life histories and can incorporate the number and phylogenetic diversity of hosts that a parasite uses at specific life stages, as well as geographic differences in hosts used (Poulin and Keeney, Reference Poulin and Keeney2008; Poulin et al. Reference Poulin, Krasnov and Mouillot2011). The extent of host specificity has direct bearing on parasite evolution and transmission dynamics as adding hosts can provide additional resources and colonization opportunities to the parasite while decreasing risks of extinction (Poulin and Keeney, Reference Poulin and Keeney2008; Dunn et al. Reference Dunn, Harris, Colwell, Koh and Sodhi2009; Poulin et al. Reference Poulin, Krasnov and Mouillot2011). The downside to relaxing host specificity is that a parasite can be less able to adapt to each individual host species and is exposed to increased competition from other parasite species (Poulin, Reference Poulin2007; Hayward, Reference Hayward2010). The ability of a parasite to exploit new hosts can be limited by physiological incompatibilities between parasites and potential hosts (Sapp and Loker, Reference Sapp and Loker2000; Locke et al. Reference Locke, McLaughlin and Marcogliese2010), possibly resulting from evolved parasite specialization (Antonovics et al. Reference Antonovics, Boots, Ebert, Koskella, Poss and Sadd2013), and/or ecological/geographic separation decreasing the chances of parasites encountering hosts (Combes, Reference Combes2001).
The use of molecular data to reveal previously undescribed cryptic species of parasites has allowed for a more accurate determination of host specificity in many parasite taxa, as exemplified by the digenetic trematodes (Platyhelminthes: Trematoda) (Blasco-Costa et al. Reference Blasco-Costa, Balbuena, Raga, Kostadinova and Olson2010; Detwiler et al. Reference Detwiler, Zajac, Minchella and Belden2012; Curran et al. Reference Curran, Tkach and Overstreet2013). Digeneans typically utilize three hosts, including a molluscan first intermediate host and vertebrate definitive host, and morphologically distinct stages to complete their life cycles (Gibson and Bray, Reference Gibson and Bray1994; Cribb et al. Reference Cribb, Bray, Olson and Littlewood2003). Larval trematodes typically possess few morphological traits for identification, making it difficult to accurately assess the number and identity of parasite species infecting intermediate hosts (Poulin, Reference Poulin2011). Investigations into the genetic diversity of larval trematodes within intermediate hosts commonly reveal large numbers of previously undescribed species, altering interpretations of the number of parasite species a host is infected by and patterns of host specificity of the parasites (Donald et al. Reference Donald, Kennedy, Poulin and Spencer2004; Miura et al. Reference Miura, Kuris, Torchin, Hechinger, Dunham and Chiba2005; Leung et al. Reference Leung, Keeney and Poulin2009b ; Detwiler et al. Reference Detwiler, Bos and Minchella2010; Locke et al. Reference Locke, McLaughlin and Marcogliese2010). The number of trematode genetic lineages within intermediate hosts must therefore be determined before assessments of host specificity can be made.
Our goals were to investigate the genetic diversity and host specificity of digeneans exploiting two congeneric gastropod hosts. Zeacumantus subcarinatus is a marine mud snail that is common in intertidal bays, rock pools and other coastal habitats throughout most of New Zealand (Morton and Miller, Reference Morton and Miller1968; Morton, Reference Morton2004). This species has been the subject of numerous parasitological investigations, and at least nine species of trematodes use Z. subcarinatus as a first intermediate host, including four ‘cryptic’ species of Acanthoparyphium echinostomes revealed using molecular data (Leung et al. Reference Leung, Donald, Keeney, Koehler, Peoples and Poulin2009a , Reference Leung, Keeney and Poulin b ). Echinostomes (Echinostomatidae) are a diverse group of digenetic trematodes characterized by a circumoral head collar consisting of one or two crowns of spines (Kostadinova, Reference Kostadinova, Jones, Bray and Gibson2005). Many species are morphologically similar, and some are associated with animal diseases (Detwiler et al. Reference Detwiler, Bos and Minchella2010) and host manipulation (Leung et al. Reference Leung, Keeney and Poulin2009b ). The Acanthoparyphium echinostomes were identified from the Otago region of the South Island of New Zealand and their geographic distribution throughout New Zealand is currently not known. Zeacumantus lutulentus is a second species of mud snail that is also relatively common in New Zealand. Zeacumantus lutulentus is found on mid-tidal mud flats throughout the North Island and northern portion of the South Island of New Zealand, and can be found co-distributed or upshore from Z. subcarinatus in the same locations (Morton, Reference Morton2004; Keeney et al. Reference Keeney, Szymaniak and Poulin2013). In contrast to Z. subcarinatus, the trematode assemblage utilizing this species as a first intermediate host is unknown. Since the trematodes utilizing Z. lutulentus have not been examined, the degree to which these two congeneric, and often sympatric, snails share trematode species, including echinostomes, is not known. As both species can be found in the same bays, they are likely exposed to similar trematodes in these areas, potentially increasing the likelihood that these phylogenetically related snails will share trematodes.
The present study examines the genetic diversity of echinostome trematodes in Z. subcarinatus and Z. lutulentus throughout the majority of their ranges in New Zealand. Our goals are to (1) quantify the genetic diversity and geographic distribution of echinostome trematodes infecting Z. subcarinatus and Z. lutulentus throughout New Zealand, (2) determine the levels of host specificity of echinostomes between these two snail hosts and (3) provide initial data to determine if geographic separation or host–parasite incompatibilities are responsible for observed host specificity patterns.
MATERIALS AND METHODS
A total of 45 sites were visited throughout New Zealand and 5590 Z. subcarinatus and 3298 Z. lutulentus were collected and examined for echinostomes between December 2006 and November 2007. Two to 849 (mean = 169±179 s.d.) Z. subcarinatus and 9–311 (mean = 143±73 s.d.) Z. lutulentus were examined from sites where they were recovered. Snails were kept in 2 L containers with water from the collection site until dissected. The maximum shell length of each snail was measured with vernier calipers to the nearest 0·1 mm (length data not used in this study), snails were dissected, and echinostome rediae were placed into 95% ethanol. We were not able to distinguish echinostome species prior to DNA sequencing. DNA was extracted from a single redia from each snail in 400 μL of 5% chelex containing 0·1 mg mL−1 proteinase K, incubated at 60 °C between 2 and 10 h, and heated to 100 °C for 8 min. An approximately 800 bp region of the COI gene was amplified using the primers JB3: 5′-TTTTTTGGGCATCCTGAGGTTTAT-3′ (Bowles et al. Reference Bowles, Blair and McManus1995) and CO1-R trema: 5′-CAACAAATCATGATGCAAAAGG-3′ (Miura et al. Reference Miura, Kuris, Torchin, Hechinger, Dunham and Chiba2005). Polymerase chain reaction (PCR) reactions (50 μL) contained 1·5 μL of DNA extraction, 200 μ m each dNTP, 1·5 mm MgCl2, 0·5 μ m each primer, 1× Taq buffer and 1·25 units GoTaq DNA polymerase (Promega). PCR amplification consisted of 2 min at 94 °C, followed by 35 cycles of 30 s at 94 °C, 30 s at 53 °C and 1 min at 72 °C and a final extension for 8 min at 72 °C. The ribosomal internal transcribed spacer region 1 (ITS1) was additionally amplified from a subset of trematodes representing all major COI clades. PCR conditions for the ITS1 region were as described for COI using 40 cycles with the primers BD1 5′-GTCGTAACAAGGTTTCCGTA-3′ and 4S 5′-TCTAGATGCGTTCGAARTGTCGATG-3′ (Bowles and McManus, Reference Bowles and McManus1993). COI and ITS1 PCR products were purified using QIAquick gel extraction and PCR purification kits (Qiagen), respectively. Purified PCR products were sequenced using PCR primers at the SUNY Upstate Medical University DNA core facility and the DNA Analysis Facility at Yale University.
DNA sequences were aligned using Clustal W (Thompson et al. Reference Thompson, Higgins and Gibson1994) as implemented in MEGA5 (Tamura et al. Reference Tamura, Peterson, Peterson, Stecher, Nei and Kumar2011). Bayesian phylogenetic analyses were conducted for COI and ITS1 separately using MrBayes 3·2 (Ronquist et al. Reference Ronquist, Teslenko, van der Mark, Ayres, Darling, Höhna, Larget, Liu, Suchard and Huelsenbeck2012). Both analyses sampled across GTR model space during the Markov chain Monte Carlo (MCMC) analysis by setting nst = mixed and rates = gamma. MCMC searches utilized three heated chains and were run for 5 000 000 (COI) and 1 000 000 (ITS1) generations with sampling every 100 generations. Maximum likelihood analyses were also conducted for COI and ITS1 separately using Garli v.2.01 (Zwickl, Reference Zwickl2006) with initial parameters set to run under the TPM2uf+I+G and TPM1uf models of sequence evolution for COI and ITS1, respectively, as selected by the Akaike information criterion (AIC) of jModelTest (Posada, Reference Posada2008). For each gene, 1000 bootstrap replicates were conducted and a consensus tree was produced using SumTrees (Sukumaran and Holder, Reference Sukumaran and Holder2010). Phylogenetic trees were visualized using FigTree v1.3.1 (Rambaut, Reference Rambaut2009). Outgroup sequences were Fasciola hepatica (GenBank accession # AF216697.1) and Fascioloides magna (GenBank accession # EF534998.1) for COI and F. hepatica (GenBank accession # JF708029.1) and Fasciola gigantica (GenBank accession # KC476171.1) for ITS1. Sequences from known Acanthoparyphium species recovered from Z. subcarinatus were included in phylogenetic analyses for species identification. These included Acanthoparyphium sp. A (GenBank accession #s FJ765457.1 for COI and FJ396155.1 for ITS1), sp. B (GenBank accession #s FJ765461.1 for COI and FJ396146.1 for ITS1), sp. C (GenBank accession #s FJ765463.1 for COI and FJ396153.1 for ITS1) and sp. D (GenBank accession #s FJ765466.1 for COI and FJ396157.1 for ITS1). Uncorrected p distances were calculated within and among major clades using MEGA5 with all insertion/deletions treated as single nucleotide differences for ITS1.
Genetic population structure of Acanthoparyphium sp. A and Himasthlinae gen. sp. (see Results) COI haplotypes among all sample sites and between pairs of sample sites was examined using analysis of molecular variance (AMOVA) with Arlequin 3·5 (Excoffier et al. Reference Excoffier, Laval and Schneider2005). Insufficient sample sizes for the other Acanthoparyphium species prohibited useful inferences. AMOVA analyses for both species incorporated Tamura and Nei's (Reference Tamura and Nei1993) model of sequence evolution and gamma-distributed substitution rate variation with shape parameter α = 0·01. This was the most appropriate model of sequence evolution available based on the AIC of jModelTest (Posada, Reference Posada2008).
RESULTS
Genetic diversity of echinostomes within snails
Snails infected with echinostomes were successfully collected from 30 sites, including 17 sites with infected Z. subcarinatus and 16 sites with infected Z. lutulentus (Table 1; Fig. 1). A total of 3863 Z. subcarinatus and 2361 Z. lutulentus were examined from these sites. Total echinostome prevalence within sites was 0·0–25% in Z. subcarinatus and 0·0–21·5% in Z. lutulentus (Table 1). As all echinostomes were not sequenced and the analysis of a single redia from each host precluded identification of co-infections, individual species prevalences could not be determined accurately. A 778 bp fragment of the COI gene was sequenced from 204 echinostomes, producing 124 different haplotypes (GenBank accession #s KJ956252–KJ956375).
Fig. 1. Sample sites where echinostomes were recovered. Site numbers 1–30 follow Table 1. Species recovered from each site are identified as Acanthoparyphium sp. A–E (A–E) and Himasthlinae gen. sp. (Hsp). Subscripts for Hsp represent host species: l = Zeacumantus lutulentus and s = Z. subcarinatus. Numbers represent how many sites in that region the species was recovered from (e.g. x3 = 3 sites). * Denotes that the GenBank sequence was recovered from a site, but the species was not recovered there in the present study.
Table 1. Sample sites where echinostomes were recovered with total prevalence and species identified. Data presented are the number of snails dissected from each site (n), the total prevalence of echinostomes and the echinostome species identified using COI. The number of each species recovered from individual sites is listed in parentheses after the species name. The sums of species identified within sites do not match total prevalences as not all echinostomes were sequenced for identification
Bayesian analysis utilizing the COI gene produced two highly diverged clades; a relatively shallow unidentified echinostome clade and a clade consisting of Acanthoparyphium species (Fig. 2). The unidentified echinostome's COI sequences were 18·9–22·1% divergent from those of the Acanthoparyphium species (Table 2). Based on cercariae, we have relatively few morphological characters to classify this species, but spine count (31) in the collar around the oral sucker associates it with the marine genera Curtuteria and Himasthla, both of which belong to the subfamily Himasthlinae (Yamaguti, Reference Yamaguti1975). The cercariae most closely resemble Himasthla spp. in having two pairs of five angle spines, and possessing cystogenous cells with cytoplasm filled with bacilliform granules. COI sequences of this species differ from Himasthla spp. by 19·0–19·5% (Osamu Miura, unpublished data). NCBI BLASTN searches (Altschul et al. Reference Altschul, Madden, Schäffer, Zhang, Zhang, Miller and Lipman1997) for COI returned a Curtuteria australis sequence as the most similar match with 82% maximum identity, but numerous species had maximum identities between 80 and 81%, overlapping with additional C. australis haplotypes. ITS1 comparisons were also inconclusive as the present species most closely matched several C. australis (ex: GenBank accession # FJ396159.1) with 92% maximum identity, several Acanthoparyphium species (ex: GenBank accession # FJ396157.1) with 88–89% maximum identity, and an Echinostomatidae sp. (GenBank accession # KC527820.1) with 88% maximum identity. Given the lack of both adult specimens and genetic similarities among existing echinostome sequences, it is prudent to maintain this species as an unclassified member of the subfamily Himasthlinae, and we refer to it throughout the manuscript as Himasthlinae gen. sp.
Fig. 2. Phylogenetic relationships of echinostomes based on COI sequences. Bayesian tree topology is shown with node support representing Bayesian support values/maximum likelihood bootstrap values. Acanthoparyphium species A–D GenBank sequences and representative sequences of Acanthoparyphium species E and Himasthlinae gen. sp. (Hsp) are labelled to provide species identification. The snail intermediate hosts from which major echinostome clades were recovered from are listed to the right of the tree.
Table 2. Ranges of p distances within (diagonal) and between Acanthoparyphium (Aca) species A–E and Himasthlinae gen. sp. (Hsp) for COI and ITS1
Members of the Himasthlinae gen. sp. clade (n = 107 trematodes) were recovered from both host species, but were more common in Z. lutulentus (n = 97 trematodes) than Z. subcarinatus (n = 10 trematodes). There was no evidence of lineage partitioning between host species and two identical haplotypes were recovered from both hosts. Within the Acanthoparyphium clade (n = 97 trematodes), five well-supported clades, representing the four previously identified species of Acanthoparyphium (spp. A–D), and an additional fifth species, referred to as species ‘E’ throughout the remainder of this manuscript, were recovered. None of the four species previously identified from Z. subcarinatus were recovered from Z. lutulentus while Acanthoparyphium species E was only recovered from Z. lutulentus. Relationships among the five Acanthoparyphyium species were not well resolved, with the exception of sp. C and D forming a clade (Fig. 2). With the exception of Acanthoparyphium species A and E forming a clade (poorly supported in the Bayesian tree), maximum likelihood analysis yielded a similar topology with slightly lower support for some of the relationships within the Acanthoparyphium clade and only Bayesian trees are presented with both Bayesian support and maximum likelihood bootstrap values (Fig. 2).
COI p distances were 0–2·3% within Acanthoparyphium species and 0–1·5% within the Himasthlinae gen. sp. In contrast, COI p distances were 9·3–14·8% between different Acanthoparyphium species and 18·9–22·1% between Acanthoparyphium species and Himasthlinae gen. sp. (Table 2).
An approximately (depending on insertion/deletions) 541 bp sequence including portions of the ITS1 region and 5·8S rRNA was analysed from 46 trematodes representing the major clades identified with COI, producing 11 different sequences (GenBank accession #s KJ956376–KJ956386). Overall, Bayesian results were consistent with COI results. Trematodes from the Himasthlinae gen. sp. COI clade (n = 16) all possessed an identical ITS1 sequence that was divergent from the remaining sequences (Fig. 3). The remaining sequences formed a well-supported Acanthoparyphium clade. All previous COI species identifications were supported with ITS1. Two different sequences from Acanthoparyphium species E found in Z. lutulentus (n = 6 trematodes) formed a clade with sp. A (0·87/86 support), and did not form their own clade due to the minimal sequence divergence of ITS1 between species A and E. Maximum likelihood analysis produced a similar topology, but did not recover individual clades for haplotypes comprising species C and D, which were not well supported in the Bayesian tree (Fig. 3).
Fig. 3. Phylogenetic relationships of echinostomes based on ITS1 sequences. Bayesian tree topology is shown with node support representing Bayesian support values/maximum likelihood bootstrap values. Acanthoparyphium species A–D GenBank sequences and representative sequences of Acanthoparyphium species E and Himasthlinae gen. sp. (Hsp) are labelled to provide species identification. The snail intermediate hosts from which major echinostome clades were recovered from are listed to the right of the tree.
ITS1 p distances were 0–1·1% within Acanthoparyphium species and 0% within Himasthlinae gen. sp. In contrast, ITS1 p distances were 0·4–2·6% between different Acanthoparyphium species and 9·1–10·2% between Acanthoparyphium species and Himasthlinae gen. sp. (Table 2).
Significant genetic differentiation was not detected over all populations (ΦST = 0·012, P = 0·422) or between any population pairs (P > 0·050) for Acanthoparyphium sp. A. For Himasthlinae gen. sp., genetic differentiation was not detected over all populations (ΦST = 0·022, P = 0·296), but was detected between 15 out of 171 population pairs (P < 0·050), all but one of which included site 2 (Waihi; 6 comparisons) or site 15 (Okiwi Bay; 9 comparisons) (Table 3). These results should be viewed as preliminary given that the small sample sizes in many sites for both species (Table 1) can decrease the likelihood of detecting existing genetic structure.
Table 3. Himasthlinae gen. sp. pairwise ΦST values with P<0·050. Site numbers are the same as in Table 1
Geographic distributions of species
Acanthoparyphium sp. A was found in Z. subcarinatus throughout the South Island and from site 1 on the North Island (Table 1, Fig. 1), sp. B was recovered from Z. subcarinatus at site 29 on the South Island, sp. C was found in Z. subcarinatus from three sites on the South Island, sp. D was found in Z. subcarinatus from sites 21 and 30 on the South Island and site 5 on the North Island, and sp. E was recovered from Z. lutulentus from sites 9, 10 and 16 on the Northern South Island and site 3 on the North Island (Table 1, Fig. 1). Himasthlinae gen. sp. was common throughout Z. lutulentus’ range (12 sites on the Northern South Island and sites 2, 4 and 6 on the North Island) and was found in Z. subcarinatus from five sites on the South Island (sites 7, 8, 21, 22 and 30) (Table 1, Fig. 1). Within Z. subcarinatus, this species was recovered from a single snail in each site, except for site 21 (Table 1).
DISCUSSION
This study is the first to examine the genetic diversity and distribution of echinostomes in New Zealand Zeacumantus species outside of the Otago region and the first to provide information regarding the specific trematode species parasitizing Z. lutulentus. We have identified five species of Acanthoparyphium utilizing Zeacumantus snails as first intermediate hosts throughout New Zealand. Four species (A–D) were previously identified from Z. subcarinatus in the Otago region (Leung et al. Reference Leung, Keeney and Poulin2009b ) and sampling throughout New Zealand did not increase that number. COI and ITS1 sequence divergences within and among these species in the present study support the presence of four species (Vilas et al. Reference Vilas, Criscione and Blouin2005) initially identified with 16S and ITS1 (Leung et al. Reference Leung, Keeney and Poulin2009b ). We recovered all of these species, except species B, outside of this area, with species A and D found in widespread locales. Given the relatively limited innate dispersal capability of free-living Acanthoparyphium cercariae, which can live 24–36 h outside of a host (Martorelli et al. Reference Martorelli, Poulin and Mouritsen2006), their dispersal is likely dictated by the vagility of their hosts (Blasco-Costa and Poulin, Reference Blasco-Costa and Poulin2013). As these species utilize cockles and nereid polychaetes (species B) for second intermediate hosts, and oystercatchers and potentially other shore birds for definitive hosts (Martorelli et al. Reference Martorelli, Poulin and Mouritsen2006; Leung et al. Reference Leung, Donald, Keeney, Koehler, Peoples and Poulin2009a ; Peoples et al. Reference Peoples, Randhawa and Poulin2012), it is likely that their species integrity is maintained over widespread distributions by the movements of bird definitive hosts (Blasco-Costa and Poulin, Reference Blasco-Costa and Poulin2013).
While only Acanthoparyphium sp. A was examined, the lack of detectable genetic structure among populations also supports the movement of Acanthoparyphium via definitive hosts. This pattern is in contrast to the strong levels of genetic structure observed among populations of Zeacumantus species throughout New Zealand (Keeney et al. Reference Keeney, Szymaniak and Poulin2013). Occasional dispersal of infected snails could also contribute to these geographic patterns as both Zeacumantus species have complex patterns of genetic structure throughout New Zealand, with evidence of occasional long-distance dispersal (Keeney et al. Reference Keeney, Szymaniak and Poulin2013). Given the small sample sizes for many sites, we do not want to overstate the lack of genetic structure observed in Acanthoparyphium sp. A. Acanthoparyphium species B appears to have the most restricted distribution of the four species infecting Z. subcarinatus. Given its low prevalence (we recovered one infected snail), it is possible that the geographic distribution of species B is more widespread than we documented. It is interesting to note that this is the only species identified as using a nereid polychaete, Perinereis sp., as a second intermediate host, although this could be the result of accidental infections with bivalves being the typical hosts (Leung et al. Reference Leung, Donald, Keeney, Koehler, Peoples and Poulin2009a ; Peoples et al. Reference Peoples, Randhawa and Poulin2012). How this could influence the species distribution is not clear if bird definitive hosts are also being utilized and polychaetes are widespread, but it could impact prevalence in snails if birds are more likely to be infected by species utilizing cockles as second intermediate hosts, such as species A, C and D (Leung et al. Reference Leung, Donald, Keeney, Koehler, Peoples and Poulin2009a ). A definitive identification of Acanthoparyphium species B's second intermediate host(s) is necessary before conclusive statements can be made regarding the mechanisms influencing its’ distribution.
An additional species of Acanthoparyphium (species E) was only recovered from Z. lutulentus. Sequence divergences and phylogenetic analyses revealed that this species falls within the Z. subcarinatus Acanthoparyphium species complex. COI sequences for this species are least divergent from Acanthoparyphium species A (Table 2) and Bayesian phylogenetic analyses group these species together, although this relationship is not well supported given the levels of divergence among species (Figs. 2 and 3). Given its location within the Z. subcarinatus Acanthoparyphium species complex, this species may have arisen via speciation following successful host switching to Z. lutulentus. Host switching may occur when a species that is ecologically and physiologically similar to the host is encountered (Poulin, Reference Poulin2007). For speciation to occur, gene flow would have to have been interrupted between trematodes utilizing the new host and those utilizing Z. subcarinatus. While we can only speculate, this could involve alternative definitive hosts feeding primarily on Z. lutulentus vs Z. subcarinatus, resulting in reproduction primarily among Acanthoparyphium infecting Z. lutulentus. Also, these species may utilize different second intermediate hosts which would increase the potential for differential definitive host use. This could have initially arisen by the slightly different ecological conditions the two Zeacumantus species prefer in some areas, as Z. lutulentus can be found upshore in more silty environments (Morton, Reference Morton2004) and Z. subcarinatus can often be found in sheltered rocky areas (Ozawa et al. Reference Ozawa, Köhler, Reid and Glaubrecht2009). Although thorough examinations have not been performed, different second intermediate hosts have been detected among Acanthoparyphium species A–D (Leung et al. Reference Leung, Keeney and Poulin2009b ). However it initially occurred, the strong host specificity for Zeacumantus observed among Acanthoparyphium species suggests that first intermediate host use could facilitate/maintain isolation with species evolving to infect Z. lutulentus potentially losing the ability to infect Z. subcarinatus (McCarthy, Reference McCarthy1990; Detwiler et al. Reference Detwiler, Bos and Minchella2010). While experimental studies would be necessary to definitively document strict host specificity (Poulin and Keeney, Reference Poulin and Keeney2008), the high number of each species of Zeacumantus examined provides support for the existence of strong host specificity for Zeacumantus hosts among New Zealand Acanthoparyphium species; not a single individual of Acanthoparyphium spp. A, C or D, was recovered from Z. lutulentus and no sp. E individuals were found in Z. subcarinatus.
A second species of echinostome, which we have referred to as Himasthlinae gen. sp. due to its morphological similarities to members of this subfamily and lack of genetic association to known genera, was common in Z. lutulentus. This species was found throughout Z. lutulentus’ range on the northern portion of the South Island and the North Island and occurred at relatively high prevalences in sites where it was recovered. This species was also found in Z. subcarinatus, making it the only species recovered from both snails. Although it was found throughout Z. subcarinatus’ range (Table 1, Fig. 1), it was less common in this host with 10 infections found in Z. subcarinatus vs 97 in Z. lutulentus. Genetically, all individuals are supported as a single species with maximum COI divergence of 1·5% and no differences detected among ITS1 sequences (Table 2) (Vilas et al. Reference Vilas, Criscione and Blouin2005). Two identical haplotypes were recovered from both snail species, revealing a lack of divergence between host species.
In general, digeneans exhibit the highest levels of host specificity for their molluscan first intermediate host (Pearson, Reference Pearson1972; Adamson and Caira, Reference Adamson and Caira1994), due to specific host-finding behaviours and/or immune response of non-host species (Sapp and Loker, Reference Sapp and Loker2000; Poulin, Reference Poulin2007). However, host specificity at this level can vary among even closely related echinostomes, revealing the importance of identifying cryptic species and host use before detailed studies of their biology and ecology can be conducted (Detwiler et al. Reference Detwiler, Bos and Minchella2010). We detected strong host specificity for snail first intermediate hosts among Acanthoparyphium species, some of which are widespread throughout New Zealand and were recovered from sites where both host species were present. Given that the relatively low prevalences of most species resulted in their recovery from only a few sites, we should not overstate that we have strong direct evidence that both hosts are exposed to the same echinostomes throughout New Zealand, as the majority of species A–D were recovered from sites without Z. lutulentus and the majority of species E were recovered from sites without Z. subcarinatus (Table 1). Species A provides the strongest evidence that host specificity persists despite exposure to both snail hosts. Therefore, host specificity appears to be driven by host–parasite incompatibilities. In addition, while Himasthlinae gen. sp. was rare within Z. subcarinatus, most specimens were recovered from sites without Z. lutulentus, further revealing that mobile and widespread avian definitive hosts are leading to exposure of both snails to similar trematodes.
As the only shared species, Himasthlinae gen. sp. provides a contrast to the pattern observed with Acanthoparyphium. Both clades are geographically widespread, with Acanthoparyphium species characterized by speciation and host specificity vs a lack of speciation and use of multiple hosts by Himasthlinae gen. sp. The utilization of multiple hosts at a particular parasite's life stage can occur when gene flow is maintained among parasite populations in both hosts following host speciation, host switching or the colonization of new host species (Brant and Orti, Reference Brant and Orti2003; Poulin, Reference Poulin2007). While taxonomic uncertainty does not allow us to determine precisely why Himasthlinae gen. sp. differs in its host specificity compared to the Acanthoparyphium species, multiple mechanisms could be at work. Infections of Z. subcarinatus could be accidental, as only 10 infections were recovered. However, given that no ‘accidental’ Acanthoparyphium infections were found, this would still suggest decreased fidelity or incomplete evolution of host specificity compared to Acanthoparyphium spp. Although more common in allopatric species, utilization of Z. subcarinatus could represent a host addition to take advantage of a commonly encountered alternative host as parasites can display reduced fitness in the novel host (Blair et al. Reference Blair, Davis and Wu2001; Donald et al. Reference Donald, Kennedy, Poulin and Spencer2004). Given the relatedness and potential physiological similarities of the hosts, there may have been weaker selection against adding it when sympatric, or it may have initially occurred in regions lacking Z. lutulentus, such as the Southern South Island. While high levels of gene flow would reduce the ability of parasites to adapt locally, we have preliminary evidence that some populations of this species may be more isolated. Overall, our findings suggest different patterns of host specificity between two related clades of echinostome trematodes, and highlight the need for further detailed studies of model systems before any generalizations can be made about the evolution of specialization in parasites.
ACKNOWLEDGEMENTS
We are extremely grateful to Isa Blasco-Costa for her tremendous help with trematode identification, and thank Kim Bryan-Walker, Janet Koprivnikar, Deanna Clement and Tommy Leung for their assistance in collecting snails from the South Island, and Katherine, Jack and Henry Keeney for their help in collecting snails on the North Island, and Tania King for her assistance with molecular lab work. We are also grateful to Sean Locke, Osamu Miura and Ryan Hechinger for providing unpublished sequence data for comparison, and to Hilary McManus for assistance with Garli.
FINANCIAL SUPPORT
Funding was provided by a University of Otago Research Grant, the Royal Society of New Zealand's Marsden Fund, Le Moyne College's Student Research Committee and Le Moyne College's Research and Development Committee.
