Helminth communities of loggerhead turtles (Caretta caretta) from Central and Western Mediterranean Sea: The importance of host's ontogeny

Parasitology International 59 (2010) 367–375

Contents lists available at ScienceDirect

Parasitology International j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p a r i n t

Helminth communities of loggerhead turtles (Caretta caretta) from Central and Western Mediterranean Sea: The importance of host's ontogeny Mario Santoro a, Francisco J. Badillo b, Simonetta Mattiucci a, Giuseppe Nascetti c, Flegra Bentivegna d, Gianni Insacco e, Andrea Travaglini d, Michela Paoletti a, John M. Kinsella f, Jesús Tomás b, Juan A. Raga b, Francisco J. Aznar b,⁎ a

Department of Public Health Science, Section of Parasitology, “Sapienza” University of Rome, P.le Aldo Moro, 00185 Rome, Italy Instituto Cavanilles de Biodiversidad y Biología Evolutiva, y Fundación General de la UV, Universitat de València, P.O. Box 22085, Valencia 46071, Spain c Department of Ecology and Sustainable Economic Development, Tuscia University, Viale dell'Università sn 01100 Viterbo, Italy d Stazione Zoologica Anton Dohrn, Naples, Italy e Centro Regionale Recupero Fauna Selvatica e Tartarughe Marine, S.W.F., Comiso (RG), Italy f Helm West Laboratory, 2108 Hilda Avenue, Missoula, MT 59801, USA b

a r t i c l e

i n f o

Article history: Received 17 September 2009 Received in revised form 17 April 2010 Accepted 30 April 2010 Available online 10 May 2010 Keywords: Caretta caretta Loggerhead sea turtle Helminth community Ontogeny Mediterranean Sea

a b s t r a c t We investigated the factors providing structure to the helminth communities of 182 loggerhead sea turtles, Caretta caretta, collected in 6 localities from Central and Western Mediterranean. Fifteen helminth taxa (10 digeneans, 4 nematodes and 1 acanthocephalan) were identified, of which 12 were specialist to marine turtles; very low numbers of immature individuals of 3 species typical from fish or cetaceans were also found. These observations confirm the hypothesis that phylogenetic factors restrict community composition to helminth species specific to marine turtles. There were significant community dissimilarities between turtles from different localities, the overall pattern being compatible with the hypothesis that parasite communities reflect the ontogenetic shift that juvenile loggerheads undergo from oceanic to neritic habitats. The smallest turtles at the putative oceanic, pelagic-feeding stage harboured only the 2 digenean species that were regionally the most frequent, i.e. Enodiotrema megachondrus and Calycodes anthos; the largest turtles at the putative neritic, bottom-feeding stage harboured 11 helminth taxa, including 3 nematode species that were rare or absent in turtles that fed partially on pelagic prey. Mean species richness per host was low (range: 1.60–1.89) and did not differ between localities. Variance ratio tests indicated independent colonization of each helminth species. Both features are expected in ectothermic and vagrant hosts living in the marine environment. © 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction The loggerhead sea turtle, Caretta caretta, belongs to a monophyletic group of chelonians containing 7 extant species that have adapted to live in the marine environment [1]. Loggerheads behave as carnivorous generalist feeders and forage principally on fish, crustaceans, molluscs, and other invertebrates, according to local availability [2–4]. However, loggerheads undergo an ontogenetic habitat shift and, therefore, they exploit broadly different trophic resources throughout their lives. After hatching, early juveniles disperse through the oceanic habitat where they prey upon epipelagic animals [2,5]. As larger juveniles, loggerheads gradually recruit to neritic habitats where they prey upon benthic animals [2];

⁎ Corresponding author. E-mail address: [email protected] (F.J. Aznar). 1383-5769/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.parint.2010.04.009

dietary data suggest the existence of an intermediate phase in which juveniles feed on both benthic and pelagic prey [3,4]. Late juveniles appear to exhibit fidelity to specific benthic foraging areas [6,7], but they have a great dispersal capacity and are able to move great distances [8–11]. Finally, when turtles become adult they return to natal beaches to reproduce, feeding on coastal benthic habitats [2]. The above ontogenetic and ecological factors should exert a strong influence upon the community structure of trophically-transmitted helminths of loggerheads. Aznar et al. [12] analyzed broad phylogenetic constraints on community structure and predicted that marine turtles as a whole should exchange few parasites with other marine vertebrates because marine turtles are ecologically isolated from other chelonians and have a peculiar physiology and diet. Thus, regardless of the locality, any marine turtle species would be expected to share parasite species only with other sympatric marine turtles [12]. The few available surveys from loggerheads support this

368

M. Santoro et al. / Parasitology International 59 (2010) 367–375

prediction, except for the limited occurrence of larval or immature helminths from fish and cetaceans [12–15]. The ontogenetic change in the diet of loggerheads might also influence the composition and abundance of their helminth communities. Valente et al. [15] reported an overall richer helminth fauna, but with significantly lower abundance at infracommunity level, in oceanic loggerheads off Madeira compared with neritic loggerheads from the Western Mediterranean [12]. Valente et al. [15] suggested that turtles from Madeira were exposed to a richer parasite fauna due to geographic factors, but the lower infection levels of individual turtles might be related to 2 key factors: (i) oceanic turtles are generally smaller than neritic turtles and, therefore, they consume infected prey at a lower rate; and (ii) perhaps more importantly, the oligrotrophic condition of the oceanic habitat coupled to the ‘dilution effect’ of pelagic conditions might limit the availability and encounter with intermediate hosts [15]. Two additional host factors, namely, ectothermy and dispersal behaviour during most of the life, have been suggested to contribute to depauperate and unpredictable infracommunities of trophicallytransmitted heminths in loggerheads [12]. Ectothermy would generally reduce opportunities for infection, whereas the dispersal capacity would reduce repeated recruitment with locally abundant helminth species. At present, there are limited data to confirm any of the above hypotheses [12–15]. In this paper, we analyse helminth communities of the loggerhead in as many as 6 localities from the Mediterranean Sea and address the following questions: (i) do the new data conform to the predictions made by Aznar et al. [12] concerning species' composition of helminth communities?; (ii) what is the role of geographical, host-related and habitat-related factors in providing infracommunity structure? Is there evidence for an influence of host's ontogeny on infracommunity structure?; and (iii) to what extent are infracommunities species-poor, and composed of independently colonizing species, regardless of the locality? 2. Materials and methods 2.1. Data collection A total of 182 loggerheads were obtained in 6 localities from the Western and Central Mediterranean during the period 1991–2008 (Table 1, Fig. 1). In the Western Mediterranean, a sample of 44 freshly dead turtles were found stranded along the coast of the Valencian Community, Spain (Fig. 1). At minimum, half of these turtles had died because of interactions with fishing activities [16]. A second sample consisted of 54 illegally captured turtles that were seized by the Spanish authorities in 1991. These turtles were captured by fishing trawls in the Spanish Mediterranean (probably off Balearic Islands), but information about the date(s) of capture is not available [3,12]. Basic infection data from these turtles were previously published [12], but are also included in the present comparative analysis. In the Central Mediterranean, loggerheads came from recovery centres in 3

Fig. 1. Sampling areas of loggerhead sea turtles, Caretta caretta, in the Central and Western Mediterranean. Ca: Campania; IC: Ionian coast of Calabria; IS: Ionian coast of Sicily; S(?): turtles seized by the Spanish authorities (see Materials and methods); TS: Tyrrhenian coast of Sicily; V: Valencia coast.

regions of Italy, namely, Campania, the Tyrrhenian and Ionian coasts of Sicily, and the Ionian coast of Calabria (Fig. 1). Turtles were found freshly dead, or were wounded or diseased and died in the recovery centre. The time between rescue and death of live turtles never exceed 12 h and, therefore, no influence of captivity on the helminth fauna was expected. All turtles were frozen until necropsy. After thawing, the curved carapace length notch-to-tip (CCL) was measured to the nearest cm. Turtles were categorized into the following age classes [17]: early juveniles (CCL ≤ 40 cm); late juveniles or sub-adults (CCL between 41 and 69 cm), and adults (CCL ≥ 70 cm). Except in Calabria, where early juveniles predominated, most turtles in all samples belonged to the group of late juveniles or sub-adults; a few putative adults appear in the samples from Ionian Sicily, Valencia and Campania (Table 1). The sex could not be determined in many individual turtles and was not included in statistical analyses. The heart, great vessels, urinary bladder, liver, oesophagus, stomach and intestine of each turtle were examined separately for helminths following the methods described by Greiner et al. [18] and Aznar et al. [12]. However, in the sample of seized turtles, only the gut, preserved in 10% formalin, was available for examination [12]. Worms were washed in saline and fixed in 70% ethanol. For identification, trematodes were stained with Mayer's acid carmine and mounted in Canada balsam, and nematodes and acanthocephalans were cleared in lactophenol. Voucher specimens are deposited in the U.S. National Parasite Collection Beltsville, Maryland (Accession numbers: 87591– 87595; 101622–101632; 101677–101678).

Table 1 Sampling features of loggerhead turtles, Caretta caretta, collected in 6 localities from the Western and Central Mediterranean. CCL is curved carapace length notch-to-tip; (n) is the number of turtles within a given category. Season: W, winter; Sp, spring; Su, summer; and A, autumn. Age class: EJ, early juveniles; LJ, late juveniles; and A, adults (see Materials and methods for details). Locality

Valencia (n = 44) Seizeda (n = 54) Calabria (n = 12) Ionian Sicily (n = 22) Tyrrenian Sicily (n = 18) Campania (n = 32) a

CCL (cm)

Season (n)

Mean ± S.D.

W

Sp

Su

A

Age class (n) EJ

LJ

A

55.7 ± 11.2 49.6 ± 9.0 38.2 ± 11.0 47.5 ± 13.1 45.1 ± 9.1 59.6 ± 11.4

7 – 1 8 6 5

7 – 2 4 4 20

24 – 8 4 6 6

6 – 1 6 2 1

3 10 8 6 6 3

36 44 4 15 12 24

5 0 0 1 0 5

These turtles were seized by the Spanish authorities in 1991 (see Materials and methods).

Year (n)

1991–1996 (9);1998–2001 (35) – 2007 (2); 2008 (10) 2006 (1); 2007 (10); 2008 (11) 2006 (1); 2007 (8); 2008 (9) 2007 (15); 2008 (17)

M. Santoro et al. / Parasitology International 59 (2010) 367–375

2.2. Identification to species level of Anisakis spp. larvae We sequenced a mtDNA 629-bp fragment of the cytochrome oxidase 2 (cox2) of 14 larvae of Anisakis Type I (sensu Berland [19]) collected from turtles in 3 localities (Table 2). The DNA extraction was performed using the CTAB method, and the PCR was carried out using primers 211 and 210, following Valentini et al. [20]. Sequences were compared with those from all Anisakis spp., which that are deposited in GenBank [20]: A. simplex (s.s.) (DQ116426), Anisakis pegreffii (DQ116428), A. simplex C (DQ116429), A. typica (DQ116427), A. ziphidarum (DQ116430), A. physeteris (DQ116432), A. brevispiculata (DQ116433), A. paggiae (DQ116434) and Anisakis sp. (DQ116431) (see Valentini et al. [20]). Maximum parsimony analysis was performed using MEGA 4 software program [21]. The reliability of the phylogenetic relationships was evaluated using nonparametric bootstrap analysis [22]; bootstrap values ≥ 60 were considered well supported [23]. 2.3. Overall faunal composition analysis We tested the prediction that only helminth species that are specialists of marine turtles will reproduce in loggerheads [12]. For each helminth species, the whole sample of worms was examined for the presence of gravid individuals. When they were found, we checked whether that helminth species had been reported in hosts other than marine turtles using an updated account of the references included in Aznar et al. [12], Santoro et al. [24], Badillo [25] and Santoro and Mattiucci [26]. If all worms in the sample were sexually immature, the infection with that species was considered as accidental. 2.4. Geographical comparison The geographical comparison was restricted to gastrointestinal helminths because in the sample of seized turtles from the Western Mediterranean only the gut was available for analysis (see above). Thus, a few non-gut helminths, including Plesiochorus cymbiformis and some individuals of A. pegreffii (see the Results), were excluded from calculations. A preliminary exploration of geographical signal was conducted without prejudging the locality of origin of the turtles [27]. The number of individuals of each helminth species from each infracommunity (that is, the community of helminths occurring in an individual turtle [28]) was square root-transformed, and the Bray– Curtis similarity coefficient was calculated between the individual turtles that harboured at least one parasite species. The resulting similarity matrix was represented in a two-dimensional plot using non-metric multidimensional scaling (NMDS) [27]. Inferential statistics on geographic patterns were carried out considering 6 localities, i.e., Valencia (n = 44), seized (n = 54), Campania (n = 32), Ionian Sicily (n = 22), Tyrrhenian Sicily (n = 18) and Calabria (n = 12) (Table 1). This subdivision is justified by the results of the NMDS, the distance between sampling localities and existence of geographical barriers (Fig. 1). The structure of infracommunities was compared between turtles from different localities with a nonparametric Analysis of Similarities (ANOSIM). The ANOSIM ranked similarities in the Bray–Curtis similarity matrix used for NMDS above. Under the null hypothesis, the ranks of similarities between and within localities should be the same on average [27]. This was evaluated with the statistic R, described in Clarke and Warwick [27]; values N 0 would indicate that turtles from the same locality are more similar to each other than to turtles from other localities [27]. The null hypothesis was built by calculating 20,000 R values under random permutation of turtles regardless of locality [27]. The overall comparison was followed by pair-wise comparisons between localities.

369

In each locality, infection parameters were estimated following Bush et al. [28] and Rósza et al. [29]. Parameters are defined here for clarity. For a sample of turtles, ‘prevalence’ is the percentage of turtles infected with a particular helminth species. For each individual turtle, ‘abundance’ is the number of individuals of a particular helminth species, ‘total helminth abundance’ the number of individuals of all helminth species, and ‘species richness’ the number of helminth species that each turtle harbours. When mean values of these parameters are to be calculated in a sample of turtles, zero values (i.e., uninfected turtles) are included [28]. In contrast, ‘intensity’ is defined as ‘abundance,’ but here uninfected turtles are not considered [28]. The 95% confidence intervals (CIs) for prevalence was calculated with Sterne's exact method [30], and for mean values of abundance, total helminth abundance, species richness, and intensity, with the bias-corrected and accelerated bootstrap method using 20,000 replications [29]. Values of total helminth abundance and species richness were compared between localities using Kruskal–Wallis tests with post hoc comparisons [31]. The 6 samples of this study, and from 3 previous surveys, i.e., Manfredi et al. [14] in the Adriatic Sea, Sey [13] off the coast of Egypt, and Valente et al. [15] off Madeira Islands (NE Atlantic) were used to perform a group-average hierarchical cluster based on the Bray–Curtis similarity matrix of prevalence data [27]. A total of 5000 random permutations were used to look for statistical evidence of genuine clusters among samples (see Clarke and Warwick [27] for details). Since larger samples of turtles can be enriched with rare helminth species, the cluster analysis was repeated excluding species with prevalence b 8.3 %, which is the minimum value that could be obtained in the smallest sample (Calabria, n = 12; 1/12 = 0.083; see Table 5). 2.5. Intralocality analysis In each locality, the influence of season on the prevalence of each helminth species was examined with Fisher's exact tests. The influence of season on abundance, total helminth abundance, and species richness was examined with Kruskal–Wallis tests, and the influence of CCL, with Spearman correlation tests. Overall association between helminth species was investigated with a variance ratio test [32]. This test compares the observed variance in helminth species richness per turtle, with the variance expected under the null hypothesis that occurrence of each species is independent on the others; the theoretical expected value is 1 (see Schluter [32] for details). To create a null distribution, we fixed the observed value of species' occurrences and randomized the occurrence of each species among the turtles, assuming they were equiprobable; we repeated this process 20,000 times. We were particularly interested in variance ratios that were high relative to the null model because this would suggest that several helminth species use the same intermediate host and/or co-occur in the same type of microhabitats where the turtle forages [33]. We used the package Primer v.6 [27] to carry out the NMDS and ANOSIM analyses; the free software Quantitative Parasitology v.3 [34] to set 95 % CIs; the free software Ecosim v.7 [35] to perform the variance ratio test, and the statistical package SPSS v.15 for the remaining analyses. In multiple comparisons, p-values were also corrected by the sequential Bonferroni procedure [36]. 3. Results 3.1. Identification to species level of Anisakis spp. larvae Based on the 629 bp sequence of the cox2 gene, all larval specimens of Anisakis Type I from loggerhead turtles were clustered with specimens of A. pegreffii in a well supported (bootstrap value = 100) monophyletic clade (tree not shown). The sequence of

370 Table 2 Basic infection parameters of helminth species found in 182 loggerhead turtles, Caretta caretta, in 6 localities from the Western and Central Mediterranean. P: prevalence (percentage of infected turtles). MI: mean intensity (mean number of worms in infected turtles). Numbers in parentheses are the 95% confidence interval of parameters. Species

a

Seized (n = 54) P

MI

Calabria (n = 12) MI

P

MI

Sicily (Ionian) (n = 22)

Sicily (Tyrrhenian) (n = 18)

Campania (n = 32)

P

P

P

MI

MI

MI

79.5 (64.9–89.2) 95.3 (67.6–150.9) 96.3 (87.3–99.3) 131.4 (98.2–177.1) 75.0 (45.7–92.8) 7.0 (3.0–13.9) 90.9 (70.9–98.4) 38.5 (28.0–52.9)

66.7 (41.4–84.4) 48.4 (27.6–94.0) 25.0 (12.2–42.3)

45.5 (31.2–60.3) 23.2 (11.6–42.8) 2.3 (0.1–12.1) 22.0

44.4 (23.7–67.0) 5.6 (0.3–27.1)

–

–

46.3 (33.2–60.3) 5.6 (1.5–15.5)

18.7 (11.6–29.8) 58.3 (29.4–81.9) 1.7 (1.0–2.6) 50.0 (29.1–70.9) 10.1 (5.5–17.0) 1.0 – – 4.5 (0.2–22.2) 2.0

5.6 (1.5–15.5)

8.0 (1.0–12.7)

–

–

4.5 (0.2–22.2)

1.0

7.8 (3.0–15.1) 2.0

5.6 (0.3–27.1) 25.0

3.5 (2.0–4.8)

28.1 (15.2–46.2) 18.7 (7.6–34.4) 9.4 (2.6–24.7) 13.3 (1.0–20.7) –

–

4.5 (0.8–15.6)

2.0 (1.0–2.0)

–

–

–

–

–

–

–

–

4.5 (0.8–15.6)

4.5 (3.0–4.5)

–

–

–

–

–

–

–

–

–

25.0 (12.2–42.3) 54.5 (17.4–159.0)

9.4 (2.6–24.7) 11.3 (5.0–17.3) –

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

15.6 (6.4–32.6) 43.4 (1.4–149.0) 25.0 (12.2–42.3) 65.2 (1.9–162.6)

15.9 (7.6–29.4)

1.9 (1.0–3.0)

27.8 (17.4–41.6)

1.5 (1.1–2.0)

4.5 (0.2–22.2)

9.0

2.3 (0.1–12.1) 1.0 – –

– –

– –

– –

– –

– –

– –

– –

– –

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

6.8 (1.9–19.0) –

– 6 (1.0–9.7) –

9.1 (1.6–29.1) –

4.0 (1.0–4.0) –

Specialist of marine turtles according the references included in Aznar et al. [12], Santoro et al. [24], Badillo [25], and Santoro and Mattiucci [26].

3.1 (0.2–16.6)

1.0

3.1 (0.2–16.6) 23.0 6.3 (1.1–20.0) 11.5 (2.0–11.5)

–

–

3.1 (0.2–16.6)

1.0

M. Santoro et al. / Parasitology International 59 (2010) 367–375

Digenea Enoditrema megachondrusa Calycodes anthosa Rhytidodes gelatinosusa Pachypsolus irroratusa Pleurogonius trigonocephalusa Plesiochorus cymbiformisa Orchidasma amphiorchisa Adenogaster serialisa Styphlotrema solitariaa Hemiuroidea sp. Nematoda Kathlania lepturaa Sulcascaris sulcataa Cucullanus carettaea Anisakis pegreffii Acanthocephala Rhadinorhynchus pristis

Valencia (n = 44) P

M. Santoro et al. / Parasitology International 59 (2010) 367–375

mtDNA cox2 obtained from A. pegreffii from C. caretta is deposited in GenBank under the Accession number FJ940755.

3.2. Overall faunal composition Fifteen helminth taxa were identified in the overall sample of loggerheads (Table 2). Seven species occurred in the stomach and intestine (Enodiotrema megachondrus, Calycodes anthos, Rhytidodes gelatinosus, Pachypsolus irroratus, Orchidasma amphiorchis, Hemiuroidea sp. and Sulcascaris sulcata) and 6 were restricted to the intestine (Pleurogonius trigonocephalus, Adenogaster serialis, Styphlotrema solitaria, Kathlania leptura, Cucullanus carettae and Rhadinorhynchus pristis). Individuals of A. pegreffii were found in the lumen of the stomach and intestine, but were also found embedded in the wall of the stomach, the liver and abdominal mesenteries. Individuals of Plesiochorus cymbiformis were found only in the urinary bladder. We found gravid worms in 12 helminth species which are considered to be specialists of marine turtles (Table 2); these species reproduced in all individual turtles in which they appeared. However, individuals of 3 additional species only occurred either as larvae (A. pegreffii) or in a sexually immature stage (Hemiuroidea sp. and R. pristis) (Table 2). Numerically, these accidental species represented a very small proportion of the total number of helminth individuals: 22 out of 7098 worms (0.3%) in the sample of seized turtles; 31 out of 3867 worms (0.8%) in Valencia; 8 out 897 worms (0.9%) in Tyrrhenian Sicily and 24 out of 1492 worms (1.6%) in Campania (Table 2).

371

Table 3 Mean similarity values (Bray–Curtis similarity index) of gastrointestinal helmith infracommunities from loggerhead turtles, Caretta caretta, between (lower diagonal) and within (diagonal) 6 Mediterranean localities. Values of statistic R, with the associated nominal p-values in parentheses, of pair-wise ANOSIM tests are on the upper diagonal (see the text for details). Calabria Sicily (I)

Sicily (T)

Campania

Seized

Valencia

Calabria

58.6

Sicily (I)

43.9

0.382 (0.0003) 58.5

Sicily (T)

44.8

56.5

0.270 (0.003) 0.013 (0.369) 52.1

Campania 18.4

13.8

14.5

− 0.021 (0.578) 0.298 (0.0001) 0.101 (0.040) 12.5

Seized

36.8

52.3

50.3

12.0

0.294 (0.004) − 0.021 (0.630) − 0.019 (0.551) 0.496 (0.0001) 48.9

Valencia

34.1

51.2

49.1

10.6

49.9

0.381 (0.0008) 0.029 (0.290) 0.049 (0.271) 0.653 (0.0001) 0.005 (0.366) 50.8

3.3. Geographical comparison The NMDS scatterplot of gastrointestinal infracomunities from the whole sample of 182 turtles is shown in Fig. 2. The stress associated to the NMDS was low and acceptably represented the relationship of ranked similarities between turtles [27]. Two patterns were apparent: (1) there was a high dispersion of turtles from Campania compared to dispersion in turtles from other localities; and (2) turtles from Campania and, to a lesser extent from Calabria, tended to be segregated from those from the remaining localities, which were clustered together. These patterns were quantitatively confirmed by values of mean similarity (Table 3) and by ANOSIM. The global test was significant (R = 0.214, p b 0.0001), and pair-wise comparisons indicated that only turtles from Campania differed significantly from those from other localities except Calabria, and turtles from Calabria differed from those of the remaining localities (Table 3). The patterns obtained from the ANOSIM were interpreted in the light of differences in abundance and composition of infracommunities between localities. First, species richness per turtle did not differ between localities (Kruskal–Wallis test, χ2 = 7.4, 5 df, p = 0.189), but

Table 4 Features of gastrointestinal helminth communities from loggerhead turtles, Caretta caretta, from 6 Mediterranean localities. Species richness: number of helminth species per turtle; total helminth abundance: number of helminth individuals regardless of species (includes 0 values); percent E + C: percent of total helminth abundance that this is accounted by the combined abundance of the 2 dominant species Enodiotrema megachondrus (E) and Calycodes anthos (C). Values are expressed as means; in parentheses are 95% confidence intervals based on 20,000 bootstrap replications. A null distribution for the Schluter's V-ratio for independent colonization of species was obtained by 5000 permutations (see Materials and methods). P (O ≥ E) and P (O ≤ E) are the probabilities that the observed value is higher, or lower, respectively, than those obtained under the null hypothesis. Significantly higher or lower values would indicate positive and negative associations between species, respectively. Locality

Species richness

Total helminth abundance

Percent E+C

Schluter's test V-ratio P (O≥E) P (O≤E)

Valencia Seized Calabria Fig. 2. Non-metric multidimensional scaling (NMDS) ordination of gastrointestinal helminth communities of the loggerhead turtles, Caretta caretta, from 6 localities from the Western and Central Mediterranean. A. Whole sample. Empty squares: Campania; grey circles: Calabria; black circles: Tyrrhenian Sicily + Ionian Sicily + Valencia + Seized turtles. B. NMDS plot identifying each of the 4 localities not separated in A. Grey squares: Tyrrhenian Sicily; empty triangles: Ionian Sicily; spotted circles: Valencia; crosses: seized turtles.

1.89 (1.61–2.18) 101.8 (74.7–156.1) 95.9 1.13 (83.8–98.9) 1.83 (1.63–2.02) 131.4 (99.5–177.6) 95.5 1.06 (90.0–98.2) 1.60 (0.42–1.83) 7.5 (3.6–15.3) 100 1.25

Sicily 1.71 (1.43–1.90) (Ionian) Sicily 1.69 (1.31–1.85) (Tyrrhenian) Campania 1.85 (1.44–2.41)

43.1 (32.3–56.4)

98.0 1.73 (93.9–99.5) 51.6 (31.5–94.1) 94.7 2.11 (66.5–99.4) 57.4 (27.0–123.1) 36.9 0.92 (21.2–55.7)

0.463 0.607 0.700 0.300 1.000 0.697 0.942 0.089 0.998 0.133 0.366 0.703

372

M. Santoro et al. / Parasitology International 59 (2010) 367–375

Table 5 Prevalence (% infected turtles in the sample) of gastrointestinal helminths of loggerhead sea turtle, Caretta caretta, in 9 localities from the Mediterranean Sea and NE Atlantic. Locality

Digenea Enodiotrema megachondrus Calycodes anthos Rhytidodes gelatinosus Pachypsolus irroratus Pleurogonius trigonocephalus Orchidasma amphiorchis Adenogaster serialis Diaschistorchis pandus Styphlotrema solitaria Pyelosomum renicapite Hemiuroidea sp. Nematoda Sulcascaris sulcata Kathlania leptura Cucullanus carettae Anisakis sp. Anisakis pegreffii Acanthocephala Rhadinorhynchus pristis Bolbosoma sp. Cestoda Nybelinia sp. Trypanorhyncha sp. a

Valencia (n = 44)

Seized (n = 54)

Calabria (n = 12)

Sicily (I) (n = 22)

Sicily (T) (n = 18)

Campania (n = 32)

Adriatica (n = 14)

Egypta (n = 33)

Madeiraa (n = 54)

79.5 45.5 2.3 0.0 4.5 0.0 0.0 0.0 0.0 0.0 15.9

96.3 46.3 5.6 5.6 0.0 0.0 0.0 0.0 0.0 0.0 27.8

75.0 58.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

90.9 50.0 4.5 4.5 0.0 0.0 0.0 0.0 4.5 0.0 0.0

66.7 44.4 5.6 5.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0

25.0 28.1 9.4 0.0 9.4 25.0 3.1 0.0 0.0 0.0 0.0

21.4 14.3 28.6 14.3 28.6 21.4 0.0 0.0 0.0 0.0 0.0

3.0 12.1 51.5 0.0 30.0 3.0 3.0 12.1 0.0 0.0 0.0

24.6 1.8 3.5 0.0 0.0 0.0 0.0 0.0 0.0 1.8 0.0

0.0 2.3 0.0 0.0 6.8

0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 9.1

0.0 0.0 0.0 0.0 0.0

25.0 15.6 3.1 0.0 6.3

74.1 0.0 0.0 0.0 0.0

24.2 18.3 0.0 0.0 0.0

0.0 0.0 0.0 35.7 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

3.1 0.0

0.0 0.0

0.0 0.0

1.8 1.8

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 0.0

0.0 6.0

3.6 0.0

Data from literature: Adriatic Sea [14], Egypt [13], Madeira [15].

total helminth abundance did (Kruskal–Wallis test, χ2 = 38.3, 5 df, p b 0.001). A post hoc comparison (p b 0.05) indicated that total helminth abundance in turtles from Calabria was significantly lower than that from other localities except Campania, and total helminth abundance in turtles from Campania was lower than that from Valencia and the seized turtles (Table 4). With regard to faunal composition, only 2 species, E. megachondrus and C. anthos, were common to all localities, but they exhibited the highest prevalence and intensity in all localities except Campania (Table 2). The numerical importance of these species in infracommunities was overwhelming: except in Campania, the combined abundance of E. megachondrus and C. anthos represented from 94.7% to 100% (in Calabria) of total helminth abundance (Table 4). In Campania, nematodes were particularly abundant (Table 2) and accounted for a substantial portion of total helminth abundance per turtle (32.7% on average). In summary, turtles from Calabria and Campania exhibited the most apparent differences in total helminth abundance and/or composition with the other samples, but resembled each other in having fewer worms of the two numerically most important species, E. megachondrus and C. anthos (Tables 2–4). Prevalence data from the 9 available surveys of gastrointestinal helminths from loggerheads in the Mediterranean Sea and NE Atlantic are shown in Table 5. Of the 19 parasite taxa recorded in these surveys, only E. megachondrus and C. anthos were common to all localities; acanthocephalans and cestodes were seldom found. The average-group hierarchical cluster based on prevalence data identified 3 significant grouping of samples: (1) Egypt + Campania + Adriatic; (2) Madeira; and (3) Calabria + Sicily (Ionian and Tyrrenian) + Valencia + seized turtles (Fig. 3). The same result was obtained after controlling for differences in host sample size (not shown). The mean CCL (one-way ANOVA, F(5, 181) = 11.12, p b 0.001; Table 1), and the number of turtles collected per season (Fisher's exact test, p b 0.001; Table 1) differed among localities. In addition, the sampling period differed among Spanish and Italian localities (Table 1). These sampling biases will be considered in the interpretation of putative geographical differences (see the Discussion).

3.4. Intralocality analysis In all localities, the number of turtles per season was independent of the year of capture (the minimum p-value observed for exact Chisquare tests was 0.080) and values of CCL did not differ between years or seasons (the minimum p-values observed for Kruskal–Wallis tests was 0.213). Therefore, we did not expect strong confounding effects between the relevant variables under study. In any locality, we found no significant differences between the prevalence of any helminth species and year or season (the minimum nominal p-value of Fisher's exact tests was 0.132). Likewise, there were no a significant relationships between CCL and the abundance of any helminth species with prevalence ≥10 %, nor with species

Fig. 3. Group-average hierarchical cluster analysis of helminth communities from 9 samples of loggerhead sea turtles, Caretta caretta, based on a Bray–Curtis resemblance matrix using prevalence data. Solid lines identify statistically significant clusters using 5000 permutations.

M. Santoro et al. / Parasitology International 59 (2010) 367–375

richness or total helminth abundance (the minimum nominal p-value of Spearman correlation was 0.07). This conclusion remained unchanged even if turtles from several samples were pooled to increase the power of tests. For instance, the Spearman correlation coefficient between CCL and total number of worms for the whole sample (n = 182) was rs = 0.038, with an associated p = 0.611. Results from the Schluter's variance ratio test are shown in Table 4. The overall pattern of species' co-occurrences did not depart from the hypothesis of independent colonization of helminth species in any locality. 4. Discussion Results from 5 new samples of loggerhead turtles from several Mediterranean localities confirm the hypothesis of Aznar et al. [12], i.e., helminth communities are composed of helminth specific to sea turtles except for the occurrence, in very low numbers, of immature forms of 3 additional taxa. R. pristis, and probably Hemiuroidea sp., are fish parasites. Adopting the encounter/compatibility paradigm of specificity [37], these taxa likely are recruited to turtles in low numbers through shared prey (i.e., a narrow encounter filter) and are unable to mature because loggerheads provide a nutritionally unsuitable environment and/or mount a strong defensive response (i.e., a closed compatibility filter). On the other hand, fish, squid and other invertebrates serve as intermediate or paratenic hosts, and marine mammals as definitive hosts, for species of Anisakis [38]. Given the stage of development (all worms were 3rd-stage larvae) and the location (many of them were found in the liver, as it is typically observed in fish), we interpret that loggerheads physiologically function as paratenic hosts [38], but ecologically represent a deadend. The remaining helminth species found in loggerheads exhibit a cosmopolitan distribution and are shared with at least the green turtle, Chelonia mydas [13,39–42]; the only exception is C. carettae, which has been reported just in loggerheads [41]. The species A. serialis, C. anthos, E. megachondrus, O. amphiorchis, P. irroratus, P. cymbiformis, and K. leptura have also been reported in olive ridley, Lepidochelys olivacea [43–45]; C. anthos in the leatherback, Dermochelys coriacea [46], and E. megachondrus, O. amphiorchis, P. irroratus, P. trigonocephalus, P. cymbiformis, R. gelatinosus, and S. solitaria in hawksbill, Eretmochelys imbricata [47]. Geographically, C. carettae, A. serialis and S. solitaria represent new locality records for the Central Mediterranean; in the Mediterranean Sea these 3 species were recorded only in hosts from the Egyptian coast [39,48]. Because of the opportunistic nature of turtle sampling, several factors could confound the geographical comparison of gastrointestinal helminth communities from loggerheads. Seasonal and CCL distributions differed among samples; also, sampling was not always carried out in the same periods. Although the exact influence of these factors is difficult to establish, the intralocality analysis suggests that season and turtle size, per se, do not seem to strongly affect the structure of infracommunities. In addition, the comparison of diversity and similarity of infracommunities indicates little geographical signal, i.e., infracommunities from turtles from closer localities did not exhibit, on average, a more similar structure than those from distant localities. For instance, infracommunities from localities as far as Valencia and Tyrrhenian Sicily resembled more each other than those from Tyrrhenian Sicily and Campania (Table 3). The hierarchical cluster analysis based on helminth prevalence of all available surveys from loggerheads also failed to reveal a clear geographical segregation (Fig. 3). Our results neither conform to any obvious grouping of turtle samples based on the spatial genetic structuring of populations or on defined barriers for juvenile dispersal in the study area (see Carreras et al. [49] and Revelles et al. [50] for details). Infracommunity analyses rather suggest that loggerhead samples can be divided into 3 groups that would reflect the influence, on the helminth fauna, of the

373

ontogenetic shift of loggerheads from oceanic to neritic habitats [11]. These sample groups are (i) turtles from the Ionian coast of Calabria, (ii) turtles from Valencia, Sicily, and the seized sample, and (iii) turtles from the coast of Campania. Turtles from Calabria, which were the smallest in size/age (Table 1), exhibited the least diverse helminth communities and the lowest infection levels. These turtles acquired only the 2 regionally frequent species, namely, E. megachondrus and C. anthos, and harboured, on average, just 7.5 worms per turtle (Table 4). When these data are compared with oceanic loggerheads off Madeiran waters [15], similarities are apparent. Infracommunities in the latter sample were dominated by E. megachondrus; total helminth abundance sample was also low (it can be estimated as ca. 4 worms/turtle, see Valente et al. [15]), and species richness at the component community level was just 2 when the potential bias of sample size is taken into account (Table 5). We hypothesize that turtles from Calabria were also in the oceanic stage and fed mostly on epipelagic prey [3–5]. The low infection levels in individual turtles could mainly be related to a ‘dilution effect’ of oceanic conditions for parasite transmission [15,51]. The second group of samples includes turtles of medium size which were probably being recruited to the neritic habitat. With respect to the sample of Valencia and the seized turtles, this is indeed confirmed by dietary data showing that these turtles had fed on both pelagic and benthic prey in neritic waters [3,52]. Helminth infracommunities of these turtles were composed by E. megachondrus and C. anthos plus other species acquired probably by feeding on pelagic (e.g. A. pegreffii, see Mattiucci and Nascetti [38]) and benthic prey. The strong similarity among helminth communities of turtles from all localities might be related to 2 non-exclusive factors. One obvious possibility is that localities are all ecologically very similar; alternatively, turtles in this stage of development could undergo long-distance movements across the region of study which would result in substantial population mixing. There is evidence that, in the Mediterranean, juvenile loggerheads usually move great distances (in the order of hundred km), often at a fast travel rate (over 1 km h− 1) [9–11]. In other words, if travel speed between localities is fast compared to the life span of parasites, local peculiarities in parasite diversity and transmission would be inconsequential. The large-sized turtles from Campania harboured the most diverse (11 species) and distinct gastrointestinal helminth fauna. Infracommunities were composed of comparatively lower numbers of E. megachondrus and C. anthos, and a more diverse and abundant fauna of nematodes specific to marine turtles. In addition, cluster analysis grouped turtles from Campania with those from the Adriatic Sea and Egypt (Fig. 3), reflecting the high diversity and compositional similarity in their helminth fauna (Table 5). The question that arises is whether these common peculiarities are geographical and/or ecological in nature. Some data are compatible with the latter possibility, i.e., that turtles from this third group would be the oldest and more coastal. First, turtles from Campania were the largest in size (Table 1), which probably indicates that they had recruited to benthic neritic areas; Sey [13] also indicated the turtles from Egypt had been captured likely along the coast, and the turtles analyzed by Manfredi et al. [14] were collected in a basin with very shallow waters, i.e., the Adriatic Sea (note that neither of the two latter studies reported the size of turtles). Second, and particularly instructive, the occurrence of the nematode S. sulcata in these samples (Table 5) strongly suggests that at least some of the turtles had fed upon benthic prey, probably in inshore waters. Larval stages of the nematode S. sulcata are associated to a wide range of benthic bivalves and gastropods, particularly in coastal habitats [40,41,53]. The hypothesis that changes in gastrointestinal helminth communities among samples of loggerheads would primarily reflect the ontogenetic habitat shift in loggerheads is admittedly preliminary. For instance, differences between samples might also reflect availability

374

M. Santoro et al. / Parasitology International 59 (2010) 367–375

of intermediate/paratenic hosts and differences in the complexity of local food webs [38,54,55]. A major problem is that nothing is known about the life cycle of any helminth species reported in this study, except for that of the accidental species and S. sulcata. One could just speculate, e.g., that E. megachondrus and C. anthos are either ecologically ubiquitous (i.e., they exploit both benthic and pelagic food webs through the use of many paratenic hosts) or have a predominantly pelagic cycle. On the other hand, the combined role of the rookery of origin (Atlantic vs. Mediterranean) and oceanographical barriers [49,50] should also be examined in order to investigate spatial differences in the helminth fauna of loggerheads in more detail. Interestingly, species richness in infracommunities did not differ significantly among samples regardless of sometimes obvious richness differences at the component community level (Tables 2,4). In addition, our data are compatible with the hypothesis of independent recruitment of parasite species, with no apparent influence of season or turtle size within each sample. These patterns would not be surprising if parasites recruit at very low rate and do not accumulate for long periods. Low recruitment rate would be favoured by (i) ectothermy, which imposes a low rate of food intake [3], (ii) and prey consumption over extensive areas, which would be linked to the vagrant behaviour of juvenile loggerheads [11]. On the other hand, individuals of A. pegreffii are the only parasites that are recruited as larvae and are expected to exhibit a long life span (see references in Mattiucci and Nascetti [38]). The life span of the digeneans and nematodes that reproduce in loggerheads is unknown but, according to information from allied taxa infecting fish, they would probably survive just for several months (see references in Rohde [55]). In conclusion, results from this study are based on the most extensive parasitological survey that has been carried out in C. caretta to date and (i) confirm the importance of phylogenetic constraints to understand the composition of gastrointestinal helminth communities; (ii) reveal consistent patterns of infracommunity structure across localities; (iii) more importantly, provide for the first time, an ecological hypothesis about the relationship between the ontogenetic habitat shift of the host and associated changes in the structure of gastrointestinal communities. Additional studies including population genetic analysis on juveniles, sub-adults and adults of loggerheads sampled from these and other Atlantic and Mediterranean locations, will allow to confirm this hypothesis. Acknowledgments We are grateful to the Centro Regionale Recupero Fauna Selvatica e Tartarughe Marine, (Comiso), Tartanet (Brancaleone), Centro di Recupero di Isola Capo Rizzuto (Crotone), and Anton Dohrn Zoological Station (Naples), for providing the turtle carcasses. Sampling of stranded turtles in Spain was made possible to the Service of Conservation of Biodiversity, Conselleria de Medi Ambient, Aigua, Territori i Habitatge, Generalitat Valenciana, Spain. The first author thanks Carmine Minichiello (Mozzo), Filomena De Martino, and Giuliana Dinaro and Antonio Intrigliolo for logistic support during the time of necropsies in Naples and Sicily, respectively. Annalisa Liotta and Domenico Piro helped with necropsies at the rescue centre of Brancaleone and Isola Capo Rizzuto. The support and help of members of the Marine Zoology Unit of the Cavanilles Institute of the University of Valencia, Spain, during the necropsies of turtles in Valencia, Spain, is greatly appreciated. References [1] Naro-Maciel E, Le M, Fitz Simmons NN, Amato G. Evolutionary relationships of marine turtles: a molecular phylogeny based on nuclear and mitochondrial genes. Mol Phylogenet Evol 2008;49:659–62. [2] Bjorndal KA. Foraging ecology and nutrition of sea turtles. In: Lutz PL, Musick JA, editors. The biology of sea turtles. Washington DC: CRC Press; 1997. p. 199–231.

[3] Tomás J, Aznar FJ, Raga JA. Feeding ecology of the loggerhead turtle Caretta caretta in the western Mediterranean. J Zool 2001;255:525–32. [4] Casale P, Abbate G, Freggi D, Conte N, Oliverio M, Argano R. Foraging ecology of loggerhead sea turtles Caretta caretta in the central Mediterranean Sea: evidence for a relaxed life history model. Mar Ecol Prog Ser 2008;372:265–76. [5] Bolten AB. Variation in sea turtle life history patterns: neritic vs. oceanic developmental stages. In: Lutz PL, Musick JA, Wyneken J, editors. The biology of sea turtles, Vol. 2. Boca Raton: CRC Press; 2003. p. 243–57. [6] Musick JA, Limpus CJ. Habitat utilization and migration in juvenile sea turtles. In: Lutz PL, Musick JA, editors. The biology of sea turtles. Washington DC: CRC Press; 1997. p. 137–63. [7] Bowen BW, Bass AL, Chow SM, Bostrom M, Bjorndal KA, Bolten AB, et al. Natal homing in juvenile loggerhead turtles (Caretta caretta). Mol Ecol 2004;13: 797–808. [8] Bentivegna F. Intra-Mediterranean migrations of loggerhead sea turtles (Caretta caretta) monitored by satellite telemetry. Mar Biol 2002;141:795–800. [9] Bentivegna F, Valentino F, Falco P, Zambianchi E, Hochscheid S. The relationship between loggerhead turtle (Caretta caretta) movement patterns and Mediterranean currents. Mar Biol 2007;151:1605–14. [10] Revelles M, Cardona L, Aguilar A, San Felix M, Fernandez G. Habitat use by immature loggerhead sea turtles in the Algerian Basin (western Mediterranean): swimming behaviour, seasonality and dispersal pattern. Mar Biol 2007;15: 1501–15. [11] Casale P, Freggi D, Basso R, Vallini C, Argano R. A model of area fidelity, nomadism, and distribution patterns of loggerhead sea turtles (Caretta caretta) in the Mediterranean Sea. Mar Biol 2007;152:1039–49. [12] Aznar FJ, Badillo FJ, Raga JA. Gastrointestinal helminths of loggerhead turtles (Caretta caretta) from the western Mediterranean: constraints on community structure. J Parasitol 1998;84:474–9. [13] Sey O. Examination of helminth parasites of marine turtles caught along the Egyptian coast. Acta Zool Acad Sci Hung 1977;23:387–94. [14] Manfredi MT, Piccolo G, Meotti C. Parasites of Italian sea turtle. II. Loggerhead turtles (Caretta caretta Linnaeus, 1758). Parassitologia 1996;40:305–8. [15] Valente AL, Delgado C, Moreira C, Ferreira S, Dellinger T. Pinheiro de Carvalho MAA, Costa G. Helminth component community of the loggerhead sea turtle, Caretta caretta, from Madeira Archipelago, Portugal. J Parasitol 2009;95:249–52. [16] Tomás J, Gozalbes P, Raga JA, Godley BJ. Bycatch of loggerhead sea turtles: insights from 14 years of stranding data. Endang Species Res 2008;5:161–9. [17] Bolten AB. Techniques for measuring sea turtles. In: Eckert KL, Bjorndal KA, AbreuGrobois FA, Donnelly M, editors. Research and management techniques for the conservation of sea turtles. Washington DC: IUCN/SSC Marine Turtle Specialist Group; 1999. p. 110–4. [18] Greiner EC, Forrester JJ, Jacobson ER. Helminths of mariculture-reared green turtles (Chelonia mydas) from Grand Cayman, British West Indies. Proc Helminthol Soc Wash 1980;47:142–4. [19] Berland B. Nematodes from some Norwegian marine fishes. Sarsia 1961;2:1–50. [20] Valentini A, Mattiucci S, Bondanelli P, Webb SC, Mignucci-Giannone A, ColomLlavina MM, et al. Genetic relationships among Anisakis species (Nematoda: Anisakidae) inferred from mitochondrial cox-2 sequences, and comparison with allozyme data. J Parasitol 2006;92:156–66. [21] Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 2007;24:1596–9. [22] Felsenstein J. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 1985;39:783–91. [23] Hills DM, Bull JJ. An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Syst Biol 1993;42:182–92. [24] Santoro M, Greiner EC, Morales JA, Rodríguez-Ortíz B. Digenetic trematode community in nesting green sea turtles (Chelonia mydas) from Tortuguero National Park, Costa Rica. J Parasitol 2006;92:1202–6. [25] Badillo FJ. Epizoitos y parasitos de la tortuga boba (Caretta caretta) en el Mediterraneo Occidental. Ph.D. Thesis, Universitat de Valencia, Valencia, Spain, 2007. [26] Santoro M, Mattiucci S. Sea turtle parasites. In: Wehrtmann IS, Cortés J, editors. Marine biodiversity of Costa Rica, Central America. Netherlands: Springer; 2008. p. 507–19. [27] Clarke KR, Warwick RM. Change in marine communities: an approach to statistical analysis and interpretation. 3rd Ed. Plymouth: PRIMER-E; 2001. [28] Bush AO, Lafferty KD, Lotz JM, Shostak AW. Parasitology meets ecology on its own terms: Margolis et al. revisited. J Parasitol 1997;83:575–83. [29] Rózsa L, Reiczigel J, Majoros G. Quantifying parasites in samples of hosts. J Parasitol 2000;86:228–32. [30] Reiczigel J. Confidence intervals for the binomial parameter: some new considerations. Stat Med 2003;22:611–21. [31] Conover WJ. Practical nonparametric statistics. 3rd Ed. New York: John Wiley & Sons; 1999. [32] Schluter D. A variance test for detecting species associations, with some example applications. Ecology 1984;65:998–1005. [33] Lotz JM, Bush AO, Font WF. Recruitment-driven spatially-discontinuous communities: a null model for transmitted patterns in target communities of helminths. J Parasitol 1995;81:12–24. [34] Reiczigel J, Rózsa L. Quantitative parasitology 3.0. Budapest; 2005. Available at http://www.zoologia.hu/qp/qp.html. [35] Gotelli NJ, Entsminger GL. EcoSim: null models software for ecology. Version 7. Jericho: Acquired Intelligence Inc. & Kesey-Bear; 2004. VT 05465. Available at http://garyentsminger.com/ecosim/index.htm. [36] Rice WR. Analyzing tables of statistical tests. Evolution 1989;43:223–5.

M. Santoro et al. / Parasitology International 59 (2010) 367–375 [37] Kuris AM, Goddard JHR, Torchin ME, Murphy N, Gurney R, Lafferty KD. An experimental evaluation of host specificity: the role of encounter and compatibility filters for a rhizocephalan parasite of crabs. Int J Parasitol 2007;37:539–45. [38] Mattiucci S, Nascetti G. Advances and trends in the molecular systematics of Anisakis nematodes, with implications for their evolutionary ecology and hostparasite co-evolutionary processes. Adv Parasit 2008;66:47–148. [39] Yamaguti S. Synopsis of digenetic trematodes of vertebrates. Tokyo: Keigaku Publishing Co; 1971. [40] Lichtenfels JR, Bier JW, Madden PA. Larval anisakid (Sulcascaris) nematodes from Atlantic molluscs with marine turtles as definitive hosts. Trans Amer Micros Soc 1978;97:199–207. [41] Lester RJ, Blair D, Heald D. Nematodes from scallops and turtles from Shark Bay, Western Australia. Aust J Mar Freshw Res 1980;31:713–7. [42] Bursey CR, Richardson EK, Richardson DJ. First North American records of Kathlania leptura and Tonaudia tonaudia (Nematoda: Kathlanidae), parasites of marine turtles. Comp Parasitol 2007;73:134–5. [43] Brooks DR, Frazier J. New host and locality for Kathlania leptura (Rudolphi) (Nematoda: Oxyurata: Kathlanidae). Proc Helminthol Soc Wash 1980;47:267–8. [44] Pérez-Ponce De León G, García-Prieto L, León-Régagnon V. Gastrointestinal digenetic trematodes of olive ridley's turtle (Lepidochelys olivacea) from Oaxaca, Mexico. Taxonomy and infracommunity structure. Proc Helminthol Soc Wash 1996;63:76–82. [45] Santoro M, Morales JA. Some digenetic trematodes of the olive ridley sea turtle, Lepidochelys olivacea (Testudines: Cheloniidae) in Costa Rica. Helminthologia 2007;44:25–8. [46] Threlfall W. Three species of Digenea from the Atlantic leatherback turtle (Dermochelys coriacea). Can J Zool 1979;57:1825–9.

375

[47] Dyer WG, Williams EH, Bunkley-Williams Jr L, Moore D. Some digeneans (Trematoda) of the Atlantic hawksbill turtle Eretmochelys imbricata imbricata (Testudines: Cheloniidae) from Puerto Rico. J Helminth Soc Wash 1995;62:13–7. [48] Baylis HA. Report on a collection of parasitic nematodes, mainly from Egypt. Part. 1. Ascaridae and Heterakidae. Parasitology 1923;15:1–13. [49] Carreras C, Pont S, Maffucci F, Pascual M, Barcelo A, Bentivegna F, et al. Genetic structuring of immature loggerhead sea turtles (Caretta caretta) in the Mediterranean Sea reflects water circulation patterns. Mar Biol 2006;149:1269–79. [50] Revelles M, Caminas JA, Cardona L, Parga M, Tomás J, Aguilar A, et al. Tagging reveals limited exchange of immature loggerhead sea turtles (Caretta caretta) between regions in the western Mediterranean. Sci Mar 2008;72:511–8. [51] Raga JA, Fernández M, Balbuena JA, Aznar FJ. Parasites. In: Würsig B, Perrin WF, Thewissen HGM, editors. The encyclopedia of marine mammals parasites. Academic Press-Elsevier; 2009. p. 821–30. [52] Maison E. L'écologie alimentaire des tortues caouannes (Caretta caretta Linnaeus, 1758) en Méditerranée occidentale et les menaces anthropogéniques associées. Rapport de Stage. Master 2 Océanographie, spécialité Biologie et Ecologie Marines. Centre d'Océanologie de Marseille, France, 2006. [53] Berry GN, Cannon LRG. The life history of Sulcascaris sulcata (Nematoda: Ascaridoidea), a parasite of marine molluscs and turtles. Int J Parasitol 1981;11: 43–54. [54] Esch GW, Fernández JC. A functional biology of parasitism. London: Chapman and Hall; 1993. [55] Rohde KC. Ecology of marine parasites: an introduction of marine parasitology. 2nd Ed. Wallingford: CAB International; 1993.