Materials and methods

We surveyed various crab species in central California as part of a broader study investigating the prey of sea otters to discover the intermediate host(s) of their intestinal parasites. Crabs were collected via hand nets and traps in Santa Cruz, Monterey and San Luis Obispo counties from municipal wharfs (Santa Cruz Warf, Monterey Bay Municipal Warf 2, Cayucas Pier and Port San Luis Pier) during the summer of 2019. Specimens were frozen to euthanize, and then dissected at the California Department of Fish and Wildlife office in Santa Cruz, California. Species investigated included Metacarcinus gracilis (n = 149), Metacarcinus magister (n = 4), C. productus (n = 64) and R. antennarium (n = 63). During these investigations, dissections of C. productus and R. antennarium from the Monterey Bay revealed infections by metacercarial cysts of an unknown trematode. The cysts had a thin brown melanized capsule, were spherical in shape and found within the haemocoel (just inside the carapace and body wall) and joints of the crabs. Metacercariae were manually extracted from their cysts and were examined morphologically and molecularly to determine the species.

For light microscopy, metacercariae were stained with acetic acid carmine, cleared in clove oil and mounted permanently in Canada balsam (fig. 1). Molecular characterization of the 28 s ribosomal RNA (rRNA) large ribosomal subunit gene was conducted with universal primers T16 (5′ GAG ACC GAT AGC GAA ACA AGT AC 3′) and T30 (5′ TGT TAG ACT CCT TGG TCC GTG 3′) (Harper & Saunders, Reference Harper and Saunders2001), sequenced via the Sanger method at the University of Otago, New Zealand. Characterization of the 18 s rRNA small ribosomal subunit gene used primers SB3a (5′ GGA GGG CAA GTC TGG TGC 3′) and A27a (CCA TAC AAA TGC CCC CGT CTG) and was sequenced via the Sanger method at the University of Alberta, Canada. Resulting sequences were identified with a nucleotide-sequence BLASTn search via the National Center for Biotechnology Information website (Altschu et al., Reference Altschu, Gish, Miller, Myers and Lipman1990; Madden, Reference Madden, McEntyre and Ostell2002). The 18 s sequences were edited with the software Mega X, and fragments aligned by MUSCLE with the program's default parameters (Kumar et al., Reference Kumar, Stecher, Li, Knyaz and Tamura2018). For the two species infected (C. productus and R. antennarium), we investigated the effect of host size (width of carapace), location of capture, sex and species on infection prevalence (binomial distribution, ‘logit’ function) and intensity (quasi-Poisson distribution, ‘log’ function) of the trematode, through Rstudio (n = 127) (Rstudio Team, 2015). Metacarcinus gracilis was not included in the model as they were not observed to be infected and would have led to zero inflation of the model.

Fig. 1. Metacercaria of Helicometrina sp. ex Cancer productus stained: (a) whole worm, with the genital pore highlighted (arrow); (b) posterior body showing immature testes (yellow) and ovary (grey). Scale bars: (a) 1 mm; (b) 500 μm.

Results and discussion

Morphological examination of the metacercariae identified the specimens as members of the genus Helicometrina due to the presence of nine testes and the genital pore located below the caecal bifurcation (fig. 1) (Cribb, Reference Cribb, Jones, Bray and Gibson2005). Metacercariae were, on average, 2190 μm long and 590 μm wide (see supplementary material), and the cysts were, on average, 731 μm in diameter. The 28S sequence from metacercariae retrieved from C. productus returned a closest match to Helicometrina nimia (ex Haemulon falvolineatum, French grunt, Mexico) with 97.97% identity (MK648305; see Pérez-Ponce de León & Hernández-Mena, Reference Pérez-Ponce de León and Hernández-Mena2019). The 18S sequence from metacercariae retrieved from C. productus returned a closest match with H. nimia, with 98.93% (KJ995999, ex Acanthistius pictus, brick sea bass, Chile; see Gonzalez, Reference Gonzalez2016). The 18 s sequence from R. antennarium also returned a closest match with H. nimia, with 98.64% (KY614306, ex Semicossyphus darwini, Galapagos sheephead wrasse, Chile; see Ñacari et al., Reference Ñacari, Sepulveda, Escribano, Bray and Oliva2018). In three partial 18 s sequences (two ex C. productus and one ex R. antennarium) with coverage at 307 bp, there were four haplotypes shared by the metacercarial sequences that were different from three exemplar sequences from South America (KJ995995, González et al., Reference González, Henríquez and López2013; KY614306, Ñacari et al., Reference Ñacari, Sepulveda, Escribano, Bray and Oliva2018; KF938641, Oliva et al., Reference Oliva, Valdivia, Chavez, Molina and Cárdenas2015). Previous reporting of 18 specimens of H. nimia from three different host species found nine variable sites in the 18 s region from a partial sequence of 372 bp, with an average pairwise difference of three (Oliva et al., Reference Oliva, Valdivia, Chavez, Molina and Cárdenas2015). We conclude our finding to likely be H. nimia based upon comparison to available sequences and morphological similarities. The sequences had some relation to Helicometrina labrisomi, but the specimens did not correspond to this species’ morphological description (Linton, Reference Linton1910).

Infection prevalence was 14% in C. productus with an average intensity of 11.7 ± 3.2 per infected crab, and 9.5% in R. antennarium with an average intensity of 14.3 ± 4.9 parasites per infected crab. Statistical analysis revealed a significant effect of carapace width on prevalence (P < 0.01) and intensity (P < 0.01), and sex (P = 0.01) on intensity, with females having greater intensity of infections (fig. 2). Location of capture, crab species and their interaction had no significant effect in the models and were sequentially removed to create a minimum adequate model. This is interesting as no crabs in San Luis Obispo were infected. The lack of significant effect of location on prevalence may be due to low prevalence and our limited sample size. Future research should expand into broader sampling to determine if infection does occur further south, and if there is a significant difference in infection prevalence.

Fig. 2. Effect of carapace width in Cancer productus and Romaleonantennarium on infection. The two species are pooled as the model showed no significant difference between them. (a) Boxplot of the effect of carapace width on prevalence (presence/absence) of Helicometrina sp., median and quartiles. (b) Scatterplot of the interaction of carapace width on intensity of infection (total parasites) by sex of the crab host (male, blue; female, red).

No metacercariae were discovered in any of the Dungeness (M. magister) or graceful rock crabs (M. gracilis) collected. The lack of infection in M. gracilis may be due to differences in habitat and behaviour; M. gracilis are much smaller than C. productus and R. antennarium, and while they co-occur in some habitat, they are also found in shallower, open sandy habitats or eelgrass beds and feed on different prey species (Orensanz & Gallucci, Reference Orensanz and Gallucci1988; Orensanz et al., Reference Orensanz, Parma, Armstrong, Armstrong and Wardrup1995). Since H. nimia is not host specific in their decapod hosts in South America (Leiva et al., Reference Leiva, George-Nascimento and Muñoz2015, Reference Leiva, López, González and Muñoz2017), this lack of infection may be due to differences in habitat selection. It should be noted that only four Dungeness crabs were collected, possibly due to the time of year or the difficulty of collection via hand net/trap. Therefore, we cannot comment on whether this species is host for this parasite. Future sampling should target M. magister as they are of great economic significance (CDFW, 2019).

Species of Helicometrina have previously been reported to utilize various decapod crustaceans as intermediate hosts. In South America, host families include Epialtidae, Porcellanidae and Xanthidae (Leiva et al., Reference Leiva, George-Nascimento and Muñoz2015, Reference Leiva, López, González and Muñoz2017). There exists one report of Helicometrina cf. nimia infection in a species of Cancridae (Romaleon polyodon) in Chile, although this report was from a single crab (Leiva et al., Reference Leiva, George-Nascimento and Muñoz2015). The definitive hosts of Helicometrina spp. are teleost fish. Studies have reported Helicometrina spp. in central and South America from fish species in the families Merlucciidae, Serranidae, Pingipedidae, Labrisomidae, Cheilodactylidae, Ophidiinae and Gobiesocidae (Gonzalez et al., Reference Gonzalez, Barrientos and Moreno2006; Muñoz & Olmos, Reference Muñoz and Olmos2008; Morales-Serna et al., Reference Morales-Serna, García-Vargas, Medina-Guerrero and Fajer-Ávila2017). In North America, a study found H. nimia tended to be a fish generalist (Holmes, Reference Holmes, Esch, Bush and Aho1990). Helicometrina nimia has been reported in shiner perch (Cymatogaster aggregata) (Arai et al., Reference Arai, Kabata and Noakes1988), and (as Helicometrina elongata) in perch (Embiotocidae), Hubbs (Blenniidae) and scorpionfish (Scorpaenidae) in Southern California (Montgomery, Reference Montgomery1957). There is one report of H. nimia in Monterey California, where our study was conducted, in various species of fish (Chapa, Reference Chapa1969).

This is the first report of a Helicometrina species in crabs of the family Cancridae, from Pacific North America. The last report of this parasite in Californian fish was over 30 years ago (Holmes, Reference Holmes, Esch, Bush and Aho1990). Nonetheless, H. nimia is a generalist in terms of its definitive host use (Holmes, Reference Holmes, Esch, Bush and Aho1990; Gonzalez et al., Reference Gonzalez, Barrientos and Moreno2006; Muñoz & Olmos, Reference Muñoz and Olmos2008; Morales-Serna et al., Reference Morales-Serna, García-Vargas, Medina-Guerrero and Fajer-Ávila2017) so it is likely that the parasite has remained unreported in the fish fauna until now. For a first intermediate host, other species of Helicometrina use gastropod molluscs (Leiva et al., Reference Leiva, George-Nascimento and Muñoz2015, Reference Leiva, López, González and Muñoz2017). In the Mediterranean, species of Opecoelidae were identified in marine snails and abalone, and the authors noted that the parasites, unlike other trematodes, can infect multiple gastropod intermediate host species (Jousson et al., Reference Jousson, Bartoli and Pawlowski1999; Leiva et al., Reference Leiva, López, González and Muñoz2017). Other trematodes of the order Plagiorchiida have been seen to use mussels of the family Mytilidae (Perumytilus purpuratus) (Muñoz et al., Reference Munoz, Lopez and Cardenas2012) and scallops (Argopecten purpuratus) (Oliva & Sanchez, Reference Oliva and Sanchez2005) as their first intermediate hosts in South America. If H. nimia also utilizes mussels as their first intermediate host like their relatives in South America this could be significant, considering the possible importance of crab predation on controlling mussel populations (Hull & Bourdeau, Reference Hull and Bourdeau2017). However, the first intermediate host remains unknown (Leiva et al., Reference Leiva, López, González and Muñoz2017). Future research should seek this last missing link in order to complete our knowledge of the life cycle of this parasite in California and gain a better understanding of its potential role and effects in the ecosystem, and on crab populations.

Decapod crabs tend to be prey items to teleost fish early in their development and before they attain a size large enough to avoid predation (except during moulting) (Carroll & Winn, Reference Carroll and Winn1989). We would expect that infection then increases in prevalence and intensity with size and age (e.g. over time), due to continued exposure to parasite larvae as well as reduced predation. Our analysis showed a significantly positive relationship between carapace width and both infection prevalence and intensity (fig. 2). Interestingly, infection was not seen in any crab with a carapace width smaller than 85 mm. There are a few possible explanations for this (it may be due to higher rates of predation on smaller crabs as a direct result of infection). Digenean trematodes commonly alter the behaviour of their gastropod intermediate hosts (Mouritsen & Poulin, Reference Mouritsen and Poulin2002), and their crustacean intermediate hosts (McCurdy et al., Reference McCurdy, Forbes and Boates1999; Hansen & Poulin, Reference Hansen and Poulin2005; Lagrue et al., Reference Lagrue, Kaldonski, Perrot-Minnot, Montreuil and Bollache2007; Lefèvre et al., Reference Lefèvre, Adamo, Biron, Missé, Hughes and Thomas2009). Paragonimus cf. westermani alters the behaviour of its decapod host (Eriocheir japonica) (Kotsyuba, Reference Kotsyuba2018), Microphallus turgidus alters the swimming behaviour of Grass shrimp (Palaemonetes pugio) (Gonzalez, Reference Gonzalez2016) and co-infection by the trematode Maritrema sp. and acanthocephalans in the body cavity of shore crabs has been correlated with altered serotonin levels (Poulin et al., Reference Poulin, Nichol and Latham2003). The lack of infection in crabs with a carapace smaller than around 85 mm may, thus, be indicative of high predation rates in small individuals that are infected. Alternatively, gill physiology and respiration behaviours can affect exposure to trematodes in crabs, which may explain the lack of observed infection in smaller crabs (Smith et al., Reference Smith, Ruiz and Reed2007). Though not all trematodes enter their crab host via the gills, some entering through percutaneous penetration at the leg (Gyoten, Reference Gyoten2000). As some of the metacercariae were found in the leg joints of some crabs, percutaneous penetration may be more likely. Differences in habitat selection by juvenile and adults of C. productus may also explain the lack of infection in smaller crabs, as adults are more likely to be found in open areas and migrate at night to shallower waters (Orensanz & Gallucci, Reference Orensanz and Gallucci1988). The full life cycle of H. nimia and how it is transmitted to crabs is unknown. Future investigations of the life cycle and effects of H. nimia are essential for our further understanding of the importance of this parasite in marine ecosystems.

Cancer crabs are an important part of marine benthic communities from intertidal to deep water through consumptive and non-consumptive effects (Fanjul et al., Reference Fanjul, Bazterrica, Escapa, Grela and Iribarne2011; Boudreau & Worm, Reference Boudreau and Worm2012; Dairain et al., Reference Dairain, Legeay and de Montaudouin2019), as prey items for sea otters and fish species (Carroll & Winn, Reference Carroll and Winn1989; Fujii et al., Reference Fujii, Ralls and Tinker2017), as secondary controllers of mussel populations (Hull & Bourdeau, Reference Hull and Bourdeau2017) and as non-native species control agents (Jensen et al., Reference Jensen, McDonald and Armstrong2007; Epelbaum et al., Reference Epelbaum, Pearce, Barker, Paulson and Therriault2009). Due to their increasing socio-economic importance, it is essential that research be conducted to investigate not only their little-known population dynamics but also their parasite communities as well (Fitzgerald et al., Reference Fitzgerald, Wilson and Lenihan2018, Reference Fitzgerald, Lenihan, Wilson, Culver and Potoski2019). We suggest specific research investigating the possible behavioural effects of infection, and how this might be affecting the role of cancrid species in Californian marine food webs.