Isotopic discrimination in helminths infecting coral reef fishes depends on parasite group, habitat within host, and host stable isotope value

Stable isotopes of carbon and nitrogen characterize trophic relationships in predator–prey relationships, with clear differences between consumer and diet (discrimination factor Δ13C and Δ15N). However, parasite–host isotopic relationships remain unclear, with Δ13C and Δ15N remaining incompletely characterized, especially for helminths. In this study, we used stable isotopes to determine discrimination factors for 13 parasite–host pairings of helminths in coral reef fish. Differences in Δ15N values grouped according to parasite groups and habitat within the host with positive Δ15N values observed for trematodes and nematodes from the digestive tract and variable Δ15N values observed for cestodes and nematodes from the general cavity. Furthermore, Δ13C values showed more complex patterns with no effect of parasite group or habitat within host. A negative relationship was observed between Δ15N and host δ15N values among different host-parasite pairings as well as within 7 out of the 13 pairings, indicating that host metabolic processing affects host-parasite discrimination values. In contrast, no relationships were observed for Δ13C values. Our results indicate that parasite group, habitat within host, and host stable isotope value drive Δ15N of helminths in coral reef fish while their effect on Δ13C is more idiosyncratic. These results call for use of taxon- or species-specific and scaled framework for bulk stable isotopes in the trophic ecology of parasites.

and vary in Δ 13 C, while trematodes have been found to vary in both Δ 13 C and Δ 15 N 18,19 . However, there can also be considerable variation within these helminth groups 14,20 and within other coral reef parasites such as copepods, cymothoids, gnathiids, isopods, and monogeneans 2,21-23 . Distinct differences for trophic discrimination in parasitic relationships are potentially caused by the combined effects of unique feeding ecologies, often reduced metabolic capabilities of the parasitic taxa being investigated 24,25 , and host metabolic effects due to parasitism 26 . Feeding ecology varies depending on whether the parasite feeds upon host tissue exclusively (on host or within; tissue type dependent 18 ), is able to supplement with material from within the dietary tract as the host feeds or from the environment (e.g. prey items, detritus, mucus, or blood 27 ), or can directly uptake nutrients from host tissues 28 . In addition, it has been suggested that trophic discrimination factors of parasites may not be fixed but scale with the isotopic signature of their hosts, both within 29 and among parasite species 14 .
Despite multiple investigations, clear stable isotope discrimination patterns between major parasite groups have not emerged, making simple incorporation of parasites into food web studies using a single universal trophic discrimination factor impossible. This knowledge gap warrants further investigation into the drivers of discrimination factors in helminths. In this study, we examined both δ 13 C and δ 15 N values from whole tissue of 136 helminth parasite-host pairings from coral reef fish to determine the isotopic relationship between taxonomically distinct groups of helminth parasites and to investigate the effect of the habitat within the host and host isotopic value on helminth isotopic discrimination. We expected that parasite isotopic enrichments and variability versus their host will be larger in parasites located in the dietary tract as their diet may not only include host material while parasites in the body cavity that solely utilize host tissues will have less variability in their isotopic discrimination. In addition, we expect a negative scaling of parasite discrimination factors with the δ 13 C and δ 15 N values of their hosts.

Results
We examined the isotopic discrimination of 136 helminth parasite-host pairings including trematodes (n = 27), cestodes (n = 19), and nematodes (n = 90) from 4 host reef lagoon-associated fish species (from the families Lethrinidae, Nemipteridae, Siganidae and Synodontidae; Table 1, Fig. 1). One trematode species and five of the nematode species were sampled from the dietary tracts (DT) of host species while the remaining one cestode and 4 nematode species were samples from the general cavity (GC). Six out of the 13 Δ 15 N values for parasite-host pairs were negative (Supplementary Table 1), with positive relationships (1.05 to 1.58‰) predominately occurring in dietary tract associated parasites (Fig. 1). In 9 out of 13 cases, δ 15 N values were significantly different between parasites and hosts (Supplemental Table 1 and Fig. 2). For the three parasite pairs found in both Lethrinus genivittatus and Nemipterus furcosus (i.e. Allardia novacaledonica, Callamanus sp., and Pseudophyllidae), Δ 13 C and Δ 15 N were consistently the same between the parasite-host pairings within the same host, either positive or negative, except for carbon in A. novacaledonica (0.47‰ L. genivitattus versus -1.10‰ N. furcosus) (Supplemental Table 1, Fig. 2). For Δ 15 N values, we found significant differences between trematodes, cestodes and the two habitats (DT & GC) of nematodes within hosts (One-way ANOVA: F 3, 132 :26.9 p < 0.001; Fig. 3). Host δ 15 N was examined versus Δ 15 N and we found that a linear regression for all of the predatory host pairings (e.g. with the herbivore host pairing white nematode-Siganus lineatus removed) had a negative slope of − 1.2 (R 2 = 0.071; Fig. 4) with negative slopes observed within individual pairings that were different than 0 for 7 of the parasite-host pairings at α = 0.05 ( Fig. 4; Supplementary Table 2). Δ 13 C values of the pairings were generally negative with lower δ 13 C in the parasite than those of the host fish, except for Philometra sp./Saurida undosquamis (Δ 13 C − 0.06 to 1.19‰), white nematode/S. lineatus (Δ 13 C − 0.38 to 3.81‰) and A. novacaledonica / L. genivittatus (Δ 13 C 0.47 to 0.98‰, Figs. 2 and 3). The first two of these relationships occur for parasites in the digestive tract and the last one in the general cavity of the host www.nature.com/scientificreports/ fish. Among the 13 parasite-host pairings tested, significant differences between host and parasite δ 13 C were observed for 10 cases with a Δ 13 C range from − 0.73 to − 1.97‰. There was no clear Δ 13 C distinction observed between parasites found in the digestive tract versus those found in the general cavity. For Δ 13 C values, we found no significant differences between trematodes, cestodes and the two habitats (DT and GC) of nematodes within hosts (One-way ANOVA: F 3, 132 :1.4 p = 0.2; Fig. 3). Host δ 13 C was examined versus Δ 13 C and found that a linear regression for all of the pairings combined minus the herbivore pairing had a negative slope of − 0.56 (R 2 = 0.196, p < 0.001; Fig. 4), but that linear regressions for individual pairings indicated no significant difference from 0 for those relationships (Supplementary Table 2). Mean δ 13 C and δ 15 N values for the host species L. genivittus were − 14.04 ± 0.51‰ and 8.97 ± 0.52‰ (mean ± SD; Δ 13 C then Δ 15 N throughout), N. furcosus were − 14.16 ± 0.49‰ and 9.31 ± 0.50‰, S. undosquamis were − 16.25 ± 0.49‰ and 8.07 ± 1.09‰, and S. lineatus were − 15.75 ± 0.40‰ and 8.97 ± 0.59‰, respectively ( Fig. 1; Supplementary Table 1), with significant differences observed between species for both δ 13 C and δ 15 N values (One-way ANOVA: δ 13 C, F 3, 177 :52.6, p < 0.001; δ 15 N, F 3, 177 :22.8, p < 0.001). Body size did not influence δ 13 C and δ 15 N values for S. undosquamis and S. lineatus for the size range and number of individuals considered here (Pearson correlation, p > 0.05 in all cases). By contrast, fish size significantly influenced δ 15 N values for L. genivittus and N. furcosus (p < 0.001 for both species; Fig. 5) but not for δ 13 C values (p > 0.05). Calculated trophic levels ( Table 1) were similar for the whole populations of fish sampled in this study for L. genivittus and S. undosquamis (2.9 and 2.8, respectively), S. lineatus having the lowest, and N. furcosus having the highest trophic levels (2.5 and 3.0, respectively; One-way ANOVA: F 3, 137 :42.7, p < 0.001). Trophic levels were also higher for larger fish for both L. genivattus (2.6 and 2.9 for 11-15 cm and 18-21.5 cm individuals, respectively; One-way ANOVA: F 1, 39 :35.9, p < 0.001) and N. furcosus (2.8 and 3.0 for 12.1-15 cm and 20-25.3 cm individuals, respectively; Oneway ANOVA: F 1, 61 :27.6, p < 0.001).

Discussion
Δ 15 N varied inconsistently between and within taxa, with the most consistent result being elevated Δ 15 N (> 0‰) for dietary tract associated nematodes, likely associated with feeding on host dietary items in addition to tissue. Δ 13 C was consistently negative between parasite taxa and likely indicates increased reliance on fatty acids from the host to support tissue growth in reef fish-associated helminths. The varied relationships amongst and between taxa provide further evidence that parasite-host pairings are distinctly different than typical trophic relationships and warrant further investigation to adequately characterize parasite contributions to food webs.   18,19 ) indicating utilization of host-metabolized nitrogen derived from tissue in addition to nitrogen compounds derived from the host diet. The low Δ 15 N values (0.18 to 0.6‰) observed in this study demonstrate close association of the trematode with the host diet. Sole reliance on host tissue and therefore host-metabolized N would be expected to yield a "typical" trophic enrichment of ~ 3.4‰ expected between diet and consumer 8,11 , which is considerably larger than the observed Δ 15 N. Small Δ 15 N values likely reflect combined utilization of more 15 N enriched sources of nitrogen derived from host tissue as well as more 15 N depleted compounds either metabolized or directly assimilated. Negative Δ 15 N values have been previously observed for cestodes within fish hosts 15,16,18,19,27,30 and may be caused by direct utilization of relatively 15 N-depleted compounds from the host diet 16 or metabolically recycled 15 N-depleted amino acids produced by the gut microbial community 31 that both cestodes and trematodes are well-positioned to utilize while residing in the dietary tract.
Similarly, three of the general cavity associated nematodes displayed negative or neutral Δ 15 N values, indicating that direct uptake from the host without further metabolic processing by the parasite of nitrogen compounds is a likely pathway for N in this taxa. This uptake was not consistent across the taxa, with different species that target similar infection sites (e.g. dietary tract versus general cavity) displaying considerably different Δ 15   www.nature.com/scientificreports/ utilization of host-metabolized compounds from tissue or the taxa-dependent ability for nematodes to biosynthesize amino acids from nitrogenous compounds 25 .
Δ 13 C differences for parasite-host pairings. Δ 13 C values were predominately neutral or negative for ten of the parasite-host pairings examined, with positive Δ 13 C only observed for the A. novacaledonica-L. genivittatus, white nematode-S. lineatus, and Philometra sp.-S. undosquamis pairings. No change or depletion in δ 13 C does not agree with the expected 0.5-1‰ increase that is usually expected for trophic interactions 11 , but likely reflects reliance on lipids and fatty acids directly derived from the host or host diet to support helminth tissue growth. Platyhelminthes and some nematodes have been found to be incapable of de novo fatty acid synthesis and to have to rely on fatty acids derived from the host 25,32 due to incomplete metabolic pathways for lipid biosynthesis. Direct uptake of fatty acids and other lipids from the host would be expected to coincide with a www.nature.com/scientificreports/ minimum carbon fractionation as no further metabolic processing is required, thereby maintaining the relatively low δ 13 C values associated with lipids as they are incorporated into the parasite. The relatively uniform neutral or negative relationships for Δ 13 C values across the pairings located from both the dietary tract and the general cavity indicate that lipid carbon is likely utilized to support tissue growth beyond species closely associated with fatty tissues (e.g. blood and liver) 13,14 . This relationship should be examined further with methods that incorporate metabolic pathway techniques for target species beyond model organisms targeted for pathogenicity 24,25 . Pairings that have elevated Δ 13 C values likely reflect decreased utilization of host lipids and increased reliance on either host sugars or proteins processed through more complete metabolic pathways within the helminths to provide tissue carbon or potentially different δ 13 C compositions from different host tissues 18 .  both fish host species L. genivattus and N. furcosus increased with body size leading to higher trophic levels in larger fish, a pattern commonly observed for coral reef-associated invertivore/carnivore fish, while there was no corresponding increase or shift in δ 13 C throughout ontogeny. No change in δ 13 C indicates that both smaller and larger individuals are relying on similar sources of underlying carbon production 33 , and that the larger individuals are feeding on larger prey with an elevated trophic level, i.e. elevated δ 15 N values 34 . Additionally, a negative linear relationship was observed for both Δ 13 C and Δ 15 N values versus δ 13 C and δ 15 N values for predatory fish hosts were observed with significant negative slopes for nitrogen in 7 of the 13 parasite-host pairs ( Fig. 4; Supplementary Table 2), while no significant within-pairings relationships were found for carbon (Supplementary Table 2). A negative linear relationship for trophic discrimination factors with increased host carbon and nitrogen values has previously been observed for parasite-host, predator-prey and herbivore-plant relationships 12,14,35 and may be associated with dietary quality. In this study, there appeared to be an increased spread in the fractionation (Δ 15 N values) observed within the herbivore S. lineatus, with the lowest δ 15 N values occurring with the highest Δ 15 N values and a distinct grouping of individuals with lower Δ 15 N values corresponding with the highest δ 15 N values (Fig. 4). This wide range of values may represent the relative richness in diet, with strictly herbivorous individuals causing a shift in their parasites towards exclusive utilization of host tissues, while individuals that supplement with animal protein (more omnivorous) provide additional material within their diet for their parasites to supplement from. Increased protein quality in a predator's diet results in a smaller ‰ difference between the diet and consumer, i.e. a smaller trophic fractionation 36 . This relationship coincides with the trend of decreased Δ 15 N values being observed for increased trophic level www.nature.com/scientificreports/ predation within predators in this study (Fig. 4). In herbivores, larger trophic discrimination factors for nitrogen are often observed 37 , and supplementation with protein (omnivory) would be expected to generate a negative offset in Δ 15 N if the parasites are supplementing from dietary protein in addition to host tissues.

Conclusion
In conclusion, we found that positive discrimination for nitrogen occurred more often in dietary tract associated helminths, while neutral or negative discrimination occurred for helminths from within the general cavity. Increased discrimination in the dietary tract is likely due to a combination of the increased need for metabolism of food taken from the host's diet and the host tissue in comparison to the helminths in the general cavity that appear to make use of direct uptake pathways for host nitrogen compounds with minimal metabolic reworking.
No differences were observed between the two parasite habitats for discrimination of carbon. This study characterized discrimination factors for carbon and nitrogen within helminths living in coral reef fish and highlights the uncertainties that remain in adequately describing parasitic relationships within food webs. These uncertainties call for the development of a taxon-or species-specific and scaled framework for using bulk stable isotope analysis to study the trophic ecology of parasites. In addition, further work using metabolomics and compound specific stable isotope techniques is warranted to better characterize the underlying metabolic differences that are driving the differences observed for trophic discrimination factors between parasite and hosts.

Methods
Sampled areas and studied species. Individual 39 , but has been observed to feed predominately on algae and to supplement with minor consumption of invertebrates in Yaté 40 . Parasites were present in all fish that were examined and appear to be ubiquitous within the species examined in this study. All individuals caught were immediately placed in ice until further processing in the laboratory. Each fish was measured to the nearest 0.1 cm (total length) and a small piece of dorsal muscle of each fish was sampled and immediately frozen at − 20 °C for further stable isotope analyses. To extract the parasites, the general cavity was first examined to collect parasites embedded in or attached to fish tissues outside of the digestive tract. In a second step, the method presented in Justine et al. 41 was applied to flush and extract living parasites from within the digestive tract using a 9% saline solution that was then briefly brought to near boiling to fix the parasites prior to transfer to 95% ethanol. All helminth parasites (i.e. nematodes, cestodes and trematodes) having a sufficient biomass were collected and immediately frozen. A total of 54 L. genivittatus were caught with 36 exploitable parasite-fish pairings, 99 N. furcosus with 75 exploitable parasite-fish pairings, and respectively 7 S. undosquamis and 18 S. lineatus were exploitable as parasite-fish pairings ( Table 1). All animal experimentation met the ABS/ ASAB guidelines for ethical treatment of animals and sampling protocols were approved by the internal ethics committee for the Université de la Nouvelle-Calédonie.
Stable isotope preparation and analyses. Carbon and nitrogen stable isotope ratios (δ 13 C and δ 15 N) were determined for dorsal muscle tissue of all fishes collected. Fish muscle tissue is routinely utilized for stable isotope values for fish as it usually does not require lipid extraction prior to analysis 42 . Samples were freeze-dried and ground into a fine powder using a mortar and pestle. One milligram of powdered material was loaded into tin capsules and analyzed for each sample without prior treatment. This same procedure was used for parasites (whole animal) for samples that had sufficient dry mass (≥ 0.3 mg). 13 C/ 12 C and 15 N/ 14 N ratios were determined with a continuous-flow isotope ratio mass spectrometer (Thermo Scientific Delta V Advantage, Bremen, Germany) coupled to an elemental analyser (Thermo Scientific Flash EA1112, Bremen, Germany). The analytical precision was 0.1‰ for 13 C and 0.15‰ for 15 N, estimated using the internal standards leucine calibrated against 'Europa flour' and IAEA standards N1 and N2. Isotope ratios were expressed as δ notation (‰) differences from a standard reference material: where R is the corresponding ratio ( 13 C/ 12 C or 15 N/ 14 N) for both sample and reference standard and δX is the measured isotopic value in per mil (‰) relative to the international standard references are Vienna Pee Dee Belemnite (vPDB) for carbon and atmospheric N 2 for nitrogen.
Parasite-host discrimination factors were calculated using: where δX represents the isotopic value of carbon or nitrogen for each parasite-host tissue pairing examined.
Data analysis. The significance of differences in δ 13 C and δ 15 N between a fish and its parasite was tested with the Wilcoxon signed rank test when homogeneity of variances was not verified or paired samples t-tests (1) δX = R sample /R standard −1 × 10 3 ; where X is 13  www.nature.com/scientificreports/ when homogeneity of variances was verified, dependant on fish species. The relationships between fish size and isotopic values (δ 15 N or δ 13 C) were investigated with Pearson correlation coefficients. One-way analysis of variance (ANOVA) was used to determine significant differences between host and parasite δ 13 C and δ 15 N values and to explore the relationship between host size and trophic level. The relationship between host δ 13 C and δ 15 N values and Δ 13 C and Δ 15 N values were determined through linear regression followed with subsequent application of an F test of the modelled slope against a slope of 0. This was done among and within the host-parasite pairings. For the analysis among pairings we excluded samples from the herbivorous fish host S. lineatus (18 samples) due to their very different isotope values to avoid skewing the relationship due to explained outliers. The trophic level (TL) of fish individuals was calculated following the formulae of 10 : where λ is the trophic level of the source of organic matter, i.e. 1, : δ 15 N fish is the isotopic value of nitrogen for the considered fish, δ 15 N source is the isotopic value of the source of organic matter at the base of the food web, i.e. 3.59 for sedimentary organic matter 33 that concerns L. genivittatus, N. furcosus and S. undosquamis, all caught of sandy unvegetated bottom; and 2.12 for the most eaten algae by Siganus lineatus and Δ 15 N that is the trophic enrichment factor (TEF) between a food item and its consumer. Here, we adopted a value of 3.9‰ for S. lineatus 40,43 reflecting usually higher TEF for herbivores compared to the conventional 3.4‰ value 37 . For the three other species, we adopted a TEF of 3.0‰ because TEF are usually lower than the conventional value for carnivores 37,44 .