Cormorant pellets as a tool for the knowledge of parasite-intermediate host associations and nematode diversity in the environment

Most helminth parasites that occur in the marine environment have complex (indirect) life cycles. Among them, anisakid nematodes are important components in marine ecosystems (Rhode, 2005). Usually, the first step in the life cycle of the Anisakidae is an invertebrate (e.g. copepods) as first intermediate/paratenic host. Then, they frequently parasitize a fish as second intermediate/paratenic host, increasing the chances to reach their definitive host (DH), particularly piscivorous birds and mammals (Anderson, 2000; Rhode, 2005; Moravec, 2009). Therefore, the food web is the key to understanding the parasite community organization (Price, 1990; Anderson, 2000; Rhode, 2005).

The identification of host-parasite relationships and the role of each trophic item as intermediate, paratenic or definitive host are severely affected by difficulties associated in detecting endoparasites in each link of their life cycle. In such concern, the analysis of the stomach contents, pellets, and regurgitates of fish-eating birds, could be useful in detecting both larval parasites, and prey items that could act as intermediate/paratenic hosts in the environment. The Imperial cormorant or shag Phalacrocorax atriceps (King, 1828), and the Red-legged cormorant Phalacrocorax gaimardi (Lesson & Garnot, 1828), are two of the five cormorant species nesting along the Argentinean coast. Phalacrocorax atriceps is distributed from Punta León, Chubut, to the Beagle Channel, Tierra del Fuego on the Argentinean coast (Harrison, 1983; Frere et al., 2005). Phalacrocorax gaimardi nests from Bahía Sanguinetto to Monte León, Santa Cruz Province (Frere et al., 2005). They are top predators in the marine food chain of the Patagonian coast, including mainly fish, and also mollusks or crustaceans in their diet, constituting an excellent model for the analysis of parasite-host interactions in the marine environment.

With the aim of determining if pellets of Phalacrocoracidae are indicators of larvae-intermediate/paratenic host association, the objectives of this study were to identify Anisakidae larvae and prey items found in pellets of both cormorants P. atriceps and P. gaimardi; to estimate interactions among them suggesting larvae-prey associations, and to suggest gateways of those anisakid species that are commonly found parasitizing both fish-eating birds (e.g. Contracaecum spp.).

From 2006 to 2010, a total of 92 pellets of P. atriceps were collected between bird nests at random in the Punta León cormorant colony, Chubut province, Argentina (43°05’S; 64°30’W) (Fig. 1), in two consecutive breading seasons: 47 pellets from December 2006 to January 2007, and 45 from December 2007 to January 2008. Later, 86 pellets of P. gaimardi were collected at random from November 2009 to January 2010 at the colony of Isla Elena, Ría Deseado, Santa Cruz province (47°45’S; 65°56’W) (Fig. 1). These 86 pellets are a subsample of a larger data set, which were used to describe P. gaimardi diet by Morgenthaler et al. 2016. In all birds, sampling was performed during the chick-rearing period from hatching up to the appearance of true feathers since this is the period of maximum activity for the food search (Yorio et al., 1998; Svagelj & Quintana, 2007).

Fig. 1

Collected pellets were preserved in vials with 70 % ethanol, and once in the laboratory, they were disaggregated under a stereomicroscope. All hard prey remnants (e.g. otoliths, cephalopod beaks, etc), and nematode larvae were collected, cleared with lactophenol and studied under a light microscope (Garbin et al., 2007, 2008; 2011). Nematode identification was carried out following the appropriate taxonomic keys and bibliography (Hartwich, 1964, 1974; Fagerholm, 1990; Anderson, 2000). Parasite ecological indexes of prevalence (P), and mean intensity (MI) were calculated following Bush et al. (1997) only for larval stages. Prey items were identified by using reference collections, keys, and catalogues (Cousseau & Gru, 1982; Boschi et al., 1992; Gosztonyi & Kuba, 1996; Pineda et al., 1996; Volpedo & Echeverría, 2000). Prey occurrence was calculated for all items in pellets from both cormorant species.

A correspondence analysis (CA) was performed to evaluate the larvae-prey association (LPA) occurring in pellets from both fish-eating birds (Legendre & Legendre, 1998). This ordination technique allows the association of row (pellets) and column (species) frequencies in a contingency table. Prey item species with less than 5 % occurrence were excluded from the analysis since the rare taxa might introduce error and be placed at extreme ends of the first ordination axes relegating the major community trends to later axes (Gauch, 1982). We conducted all analyses with R 3.4.0 software (R Core Team 2017) using the vegan package (Ok-sanen et al., 2018).

All pellets were collected without causing disturbance at both cormorant colonies of Punta León, Chubut, and Isla Elena, Santa Cruz, Argentina, with required permissions of the Dirección de Flora y Fauna Silvestre, Chubut, and the Dirección de Fauna Silvestre y Areas Protegidas, Santa Cruz, Argentina, respectively.

From the analyzed P. atriceps’s pellets, 39 different prey items were identified, belonging to four different animal taxa: fish, mollusks, crustaceans, and polychaetes (Table 1). From P. gaimardi’s pellets, 10 different prey items were identified, belonging to the same four different animal taxa (Table 2).

Third-stage (L3), fourth-stage larvae (L4), and adults of five Anisakidae genera were identified in pellets from both cormorant species: Contracaecum Railliet & Henry, 1912, in both bird species; Pseudoterranova Mozgovoi, 1951, Anisakis Dujardin, 1845, Terranova Leiper & Atkinson, 1914, only in pellets of P. atriceps; and Hysterothylacium Ward & Magath, 1917, only in pellets of P. gaimardi. Pseudoterranova spp. L3 showed the highest prevalence (P=65.13) followed by Anisakis spp. L3 (P=43.66), Contracaecum spp. L3 (P=24.3), and Terranova spp. L3 (P=18.21) in P. atriceps pellets. The highest mean intensity was detected for Pseudoterranova spp. L3 (MI=4.32), followed by Terranova spp. L3 (MI=2.54), Anisakis spp. L3 (MI=1.92), and Contracaecum spp. L3 (MI=1.73). Contracaecum spp. L3 had the highest intensity and Hysterothylacium spp. L3 was the most prevalent anisakid in pellets of P. gaimardi (P=60.34, MI=3.25, and P=55.17, MI=4.56 respectively).

Only anisakid L3 were included in the CA analysis. The overall inertia was relatively low (2.893 out of 26) but still significant (χ² = 7111.2, N = 2458, P = <0.0001), indicating a weak association between parasites and prey items of P. atriceps. Moreover, the first two CA axes accounted for 37.86 % of the data. On P. gaimardi, the overall inertia was relatively higher (2.98 out of 7) and significant (χ² = 882.04, N = 296, P = <0.0001), indicating a stronger association between parasites and prey items. The first two CA axes accounted for 46.31 % of the data.

The highest significant larvae-prey association (LPA) in pellets of P. atriceps was revealed for Contracaecum spp. L3 with Engraulis anchoita Hubbs & Marini, 1935, followed by Afroditidae, Odontesthes sp., Paralichthys sp., Gammaridae sp., and Ostracoda (Fig. 2). Pseudoterranova spp. L3 showed a close relationship with Polyonidae, Enteroctopus megalocyathus (Gould, 1852), and Octopus tehuelchus d’Orbigny, 1834. Anisakis spp. L3 significantly associated with Tegula sp., followed by Raneya brasiliensis (Kaup, 1856), and Percophis brasiliensis Quoy & Gaimard, 1825. Terranova spp. L3 strongly associated with Patagonotothen sp., Eunicidae, Triathalassothia argentina (Berg, 1897), and O. tehuelchus (Fig. 2).

Fig. 2

In P. gaimardi’s pellets, Contracaecum spp. L3 significantly associated with Sprattus fueguensis (Jenyns, 1842), followed by Ramnogaster arcuatta (Jenyns, 1842), Loligo gahi d’Orbigny, 1835, and Patagonotothen sp. Hysterothylacium spp. L3 associated with Nereididae, then with S. fueguensis, and Odontesthes sp. (Fig. 3).

Fig. 3

The Imperial shag in Punta León colony showed a preferably piscivorous diet with common prey items being R. brasiliensis, T. argentina and Ribeiroclinus eigenmanni (Jordan, 1888) (Malacalza et al., 1994; Punta et al., 2003). Fish such as S. fueguensis and R. arcuatta were also the most frequent items in the Red-legged cormorant diet (Morgenthaler et al., 2016). Recorded anisakid prevalences suggested that Pseudoterranova spp. L3. and Anisakis spp. L3 predominated in the environment over the other two anisakid genera in the Punta León sea area, Chubut coast, whereas Contracaecum spp. L3 and Hysterothylacium spp. L3 predominated in the environment of Puerto Deseado, Santa Cruz coast. Thus, this study showed that pellets might serve to indicate the general diversity of the environment.

Contracaecum is the only genus of Anisakidae in this study that matures to adults in birds. The most significant LPA found in P. atriceps pellets was that of Contracaecum spp. L3 with E. anchoita. Previously, Diaz (2006), and Garbin et al. (2007) had suggested that E. anchoita might act as a transmitter of Contracaecum pelagicum L3 to the Magellanic Penguin Spheniscus magellanicus (Forster, 1781) and P. atriceps. Later, Garbin et al. (2013) proved this presumption analyzing phylogenetically through molecular markers (mtDNA cox2, rrnS, ITS-1, ITS-2 genes) specimens of Contracaecum spp. L3 from E. anchoita, and C. pelagicum of both sympatric fish-eating birds P. atriceps and S. magellanicus from Península Valdés (Garbin et al., 2013), the same area where the present pellets were collected. From these results, we can verify that E. anchovy is the main gateway of Contracaecum spp. L3 into P. atriceps. Therefore, the analysis of pellets was efficient to suggest possible parasite-host associations.

In the present analysis, Contracaecum spp. L3 also showed an association with Odontesthes sp. Carballo et al. (2011) reported Contracaecum L3 parasitizing the silversides Odontesthes smitti (Lahille, 1929), and Odontesthes nigricans (Richardson, 1848) in Península Valdés coast, Argentinean Sea. Present results reinforced the idea that Odonthestes sp. also should act as an intermediate/paratenic host of Contracaecum spp. L3 in the area. Phylogenetic molecular studies should be carried out to confirm this LPA hypothesis (Garbin et al. 2013).

Another significant LPA found was that between Contracaecum spp. L3 and the polychaete Aphrodita sp. Some records of Contracaecum parasitizing polychaetes exist but not in the Aphroditidae to date (Peoples, 2013).

Lower LPA were those between Contracaecum spp. L3 and Paralichthys sp., Gammaridae and Ostracoda. Incorvaia & Díaz de Astarloa (1998) also found Contracaecum L3 in Paralichthys orbignyanus (Valenciennes, 1839), and Paralichthys patagonicus Jordan, 1889, from the Argentine sea. Some authors have suggested gammarid amphipods and ostracods as intermediate/paratenic hosts of Contracaecum spp. larvae (Bartlett, 1996; Anderson, 2000; Moravec, 2009). Related to this, some ostracod species are prey items of E. anchoita and Paralichthys sp. (Capitanio et al., 2005; Ide et al., 2006). Therefore, it is possible to suggest that these arthropods are involved in some stage of the Contracaecum life cycle in the study area.

In this study, Pseudoterranova spp. L3 were the most abundant anisakid found showing the highest association with Polyonidae. McClelland et al. (1990) found Pseudoterranova L3 in polychaetes after they ingested copepods experimentally. Also, Martell & Mc-Clelland (1995) pointed out polychaetes as transmitters of Pseudoterranova L3. Associations of Pseudoterranova spp. L3 with Enteroctopus sp., and Octopus sp. are strange since there are no records of this anisakid parasitizing those cephalopods. Terranova spp. L3 strongly associated with Patagonotothen sp, Eunicidae, and T. argentina, and Octopus sp. However, there are no records of this anisakid genus parasitizing any of the latter prey items up to date. Despite Anisakis spp. L3 closely associated with Tegula sp., it is not possible to speculate a parasite-host association because there are no records of gastropods as intermediate/paratenic host of anisakid species. None of the mentioned genera are parasites of birds.

The analyses carried out on P. gaimardi’s pellets from the Ría Deseado showed the two highest LPA of Contracaecum spp. L3 with both sprat species S. fueguensis and R. arcuata. Other studies have confirmed the parasitism of Contracaecum in the genus Sprattus (Skrzypczak & Rolbiecki, 2015; Zuo et al., 2016). In addition, Contracaecum australeGarbin et al., 2011, was reported previously parasitizing P. gaimardi in the same area (Garbin et al. 2014). Therefore, both sprats S. fueguensis and R. arcuata might play a role as intermediate/paratenic hosts of C. australe, being the main gateway to P. gaimardi in the study area. As mentioned before, phylogenetic molecular studies should be carried out on these nematodes (Garbin et al. 2013).

Also, significant LPA between Hysterothylacium spp. L3 with Nereididae, S. fueguensis and Odontesthes sp. were observed in P. gaimardi’s pellets. No records of Hysterothylacium parasitizing this polychaete are available. However, there are some records on Hysterothylacium aduncum (Rudolphi, 1802) isolated from Sprattus sprattus (Linnaeus, 1758) in different geographical areas in the Northeastern Atlantic and Southern Baltic Sea (Klimpel et al. 2007; Skrzypczak & Rolbiecki, 2015). In addition, Odontesthes bonariensis (Valenciennes, 1835) is parasitized by Hysterothylacium sp. larvae from two Argentinean lagoons (Drago, 2012). Thus, it is possible to suggest S. fueguensis and Odontesthes sp. as first intermediate hosts of Hysterothylacium spp. L3 since the adult nematodes parasitize other teleost fish.

In this study, Contracaecum also showed LPA with the Patagonian squid L. gahi in P. gaimardi. Some cephalopods have been recorded to be parasitized by Contracaecum L3 as paratenic -transport- hosts (Shukhgalter & Nigmatullin, 2001; Salati et al., 2013). However, only Loligo forbesi (Steenstrup, 1856) was shown to be infected with Contacaecum L3 (Smith, 1984). Not only surveys on this squid are needed but also molecular studies on Contacaecum L3 must be carried out.

According to the present results it is possible to postulate pellets as good tools to indicate parasite-host associations between anisakids and the prey items which act as intermediate/paratenic hosts. These kind of studies also provide information about the feeding habits of both birds and mammals, and about the diversity of parasites circulating in the environment.