Methods of lipid-normalization for multi-tissue stable isotope analyses in tropical tuna

Lipid-correction models for δ13C values across small pelagic fishes (Clupeiformes) from the Atlantic Ocean

The interpretation of δ13C values in trophic ecology requires standardization of the lipid content of organisms estimated through their C:N ratio. To avoid time-consuming lipid extractions, the use of mathematical corrections has been developed for many years, and the conclusions generally point in the direction of species-specific adjustment of the models. This study aimed at defining the maximum taxonomic level required to obtain the best corrected δ13C values in small pelagic fish of the order Clupeiformes. δ13C values of six species were analyzed bulk and lipid-free, and were used to fit and validate linear and mass-balance models at different taxonomic levels. Despite a species effect combined with the C:N ratio effect, the corrected δ13C values produced by a global model for the Clupeiformes were as good as or better when compared to lipid-free samples than those produced by species-specific models, paving the way for possible generalization to other species in this order. At the order level, the linear model outperformed the mass-balance model.

Elevated Mercury Concentrations and Isotope Signatures (N, C, Hg) in Yellowfin Tuna (Thunnus albacares) from the Galápagos Marine Reserve and Waters off Ecuador

We examined how dietary factors recorded by C and N influence Hg uptake in 347 individuals of yellowfin tuna (Thunnus albacares), an important subsistence resource from the Galápagos Marine Reserve (Ecuador) and the Ecuadorian mainland coast in 2015-2016. We found no differences in total Hg (THg) measured in red muscle between the two regions and no seasonal differences, likely due to the age of the fish and slow elimination rates of Hg. Our THg concentrations are comparable to those of other studies in the Pacific (0.20-9.60 mg/kg wet wt), but a subset of individuals exhibited the highest Hg concentrations yet reported in yellowfin tuna. Mercury isotope values differed between Δ¹⁹⁹ Hg and δ202Hg in both regions (Δ¹⁹⁹ Hg = 2.86 ± 0.04‰ vs. Δ¹⁹⁹ Hg = 2.33 ± 0.07‰), likely related to shifting food webs and differing photochemical processing of Hg prior to entry into the food web. There were significantly lower values of both δ¹⁵ N and δ¹³ C in tuna from Galápagos Marine Reserve (δ¹⁵ N: 8.5-14.2‰, δ¹³ C: -18.5 to -16.1‰) compared with those from the Ecuadorian mainland coast (δ¹⁵ N: 8.3-14.4‰, δ¹³ C: -19.4 to -11.9‰), of which δ¹³ C values suggest spatially constrained movements of tuna. Results from the pooled analysis, without considering region, indicated that variations in δ¹³ C and δ¹⁵ N values tracked changes of Hg stable isotopes. Our data indicate that the individual tuna we used were resident fish of each region and were heavily influenced by upwellings related to the eastern Pacific oxygen minimum zone and the Humboldt Current System. The isotopes C, N, and Hg reflect foraging behavior mainly on epipelagic prey in shallow waters and that food web shifts drive Hg variations between these populations of tuna. Environ Toxicol Chem 2022;41:2732-2744. © 2022 SETAC.

Lipid-free tuna muscle samples are suitable for total mercury analysis

Tropical tunas are largely consumed worldwide, providing major nutritional benefits to humans, but also representing the main exposure to methylmercury, a potent neurotoxin that biomagnifies along food webs. The combination of ecological tracers (nitrogen and carbon stable isotopes, δ¹⁵N and δ¹³C) to mercury concentrations in tunas is scarce yet crucial to better characterize the influence of tuna foraging ecology on mercury exposure and bioaccumulation. Given the difficulties to get modern and historical tuna samples, analyses have to be done on available and unique samples. However, δ¹³C values are often analysed on lipid-free samples to avoid bias related to lipid content. While lipid extraction with non-polar solvents is known to have no effect on δ¹⁵N values, its impact on mercury concentrations is still unclear. We used white muscle tissues of three tropical tuna species to evaluate the efficiency and repeatability of different lipid extraction protocols commonly used in δ¹³C and δ¹⁵N analysis. Dichloromethane was more efficient than cyclohexane in extracting lipids in tuna muscle, while the automated method appeared more efficient but as repeatable as the manual method. Lipid extraction with dichloromethane had no effect on mercury concentrations. This may result from i) the affinity of methylmercury to proteins in tuna flesh, ii) the low lipid content in tropical tuna muscle samples, and iii) the non-polar nature of dichloromethane. Our study suggests that lipid-free samples, usually prepared for tropical tuna foraging ecology research, can be used equivalently to bulk samples to document in parallel mercury concentrations at a global scale.

Regional variations and drivers of mercury and selenium concentrations in Steller sea lions

Mercury (Hg) can be neurotoxic to mammals and impact reproduction, whereas selenium (Se) is an important antioxidant known to ameliorate some adverse effects of Hg. Total Hg concentrations ([THg]) were measured in lanugo (pelage grown in utero) of 812 Steller sea lion (Eumetopias jubatus) pups across Alaska and Russia to assess fetal exposure during late gestation. The molar ratio of total Se to THg (TSe:THg) was determined in whole blood collected from 291 pups. Stable isotope ratios of carbon and nitrogen were measured in sections of vibrissae (whiskers, n = 498) and in lanugo (n = 480) of pups grown during late gestation to track diet variations among adult females that can drive Hg and Se exposure during this critical fetal development period. Lanugo [THg] ranged from 1.4 to 73.7 μg/g dry weight with the lowest median [THg] in Southeast Alaska. Pups from the Western Aleutian Islands had higher median lanugo [THg] than pups from other metapopulations in Alaska. Over 25% of pups in the Western Aleutian Islands had [THg] above published risk thresholds (20 μg/g) for other mammals. Whole blood molar TSe:THg was significantly lower in the Western Aleutian Islands and in some parts of the Central Aleutian Islands with higher molar ratios found in the Eastern Aleutian Islands and Central Gulf of Alaska. This suggests a limitation on potential protective functions of Se in the western regions with the highest relative [THg]. The Central Aleutian Island pups with [THg] over 20 μg/g had higher δ¹⁵N ratios than pups with lower [THg] suggesting dams consuming higher trophic level prey is a key driver for Hg exposure. However, regional differences likely reflect variability in diet of the dam during gestation and in Hg food web dynamics between oceanic regimes east and west of key passes in the Aleutian Islands.

Legacy and emerging organic contaminants: Levels and profiles in top predator fish from the western Indian Ocean in relation to their trophic ecology

Tuna and billfish are large pelagic fish of ecological importance in open oceans. As top predators with a long lifespan, they are prone to exposure to various contaminants such as persistent organic pollutants (POPs) and contaminants of emerging concern. In this study, three pollutant families were investigated, including polychlorinated biphenyls (PCBs), organochlorinated pesticides (OCPs) and perfluoroalkyl substances (PFASs), including perfluorooctane sulfonate (PFOS) and perfluorocarboxylic acids (PFCAs). Contamination was investigated in individuals from three tropical tuna species, namely bigeye (Thunnus obesus), skipjack (Katsuwonus pelamis) and yellowfin (Thunnusalbacares) tunas and the billfish swordfish (Xiphias gladius), collected from various areas of the western Indian Ocean (WIO) in 2013-2014. Contamination levels and profiles were examined in fish muscle, together with biological parameters (fish length / age, sex, lipid content) and ecological tracers (carbon and nitrogen stable isotopes). POP levels were low in all species in comparison to other locations worldwide, revealing a low impact of anthropogenic organic contaminants in the WIO. A predominance of OCPs (especially DDTs) versus PCBs was highlighted in all species; PFASs were predominant over chlorinated POPs in tunas. Among the studied PFASs, long-chain PFCAs were found to prevail over PFOS in all species. Organic contaminant profiles differed across species according to their foraging habitat; swordfish and bigeye tuna, which both feed in deep oceanic layers, showed similarities in their contaminant profiles. Geographically, the distinct DDT profiles of fish from the Mozambique Channel suggested an exposure to different DDT sources, in line with regional use of this insecticide and coupled with an extended residence time of fish in the Channel. To our knowledge, the data presented here are among the first obtained for legacy and emerging organic contaminants in various species of large pelagic predators from the WIO.

Validating fin tissue as a non-lethal proxy to liver and muscle tissue for stable isotope analysis of yellow perch (Perca flavescens)

Stable isotope ecology typically involves sacrificing the animal to obtain tissues. However, with threatened species or in long-term longitudinal studies, non-lethal sampling techniques should be used. The objectives of this study were to (1) determine if caudal fin tissue could be used as a non-lethal proxy to liver and muscle for stable isotope analysis, and (2) assess the effects of ethanol preservation on δ¹⁵N and δ¹³C in fin tissue of juvenile yellow perch Perca flavescens. The δ¹³C of caudal fin was not significantly different from liver (t₂₃ = -0.58; p = 0.57), and was more correlated with δ¹⁵N in liver (r² = 0.78) than muscle (r² = 0.56). Ethanol preservation enriched ¹⁵N and ¹³C for caudal fins, but by using our developed regression models, these changes in δ¹⁵N and δ¹³C can now be corrected. Overall, caudal fin tissue is a more reliable proxy to liver than muscle for δ¹⁵N and δ¹³C in yellow perch.

Effects of euthanasia methods on stable carbon (δ13 C value) and nitrogen (δ15 N value) isotopic compositions of fry and juvenile rainbow trout Oncorhynchus mykiss

RATIONALE

Carbon and nitrogen stable isotope analyses of fish tissues are now commonly used in ecological studies but mostly require the sacrifice of the animal. Ethical considerations recommend the use of anesthetics for tissue sampling. This study examines how anesthetics affect stable isotope ratios of fish compared with other euthanasia methods.

METHODS

Rainbow trout fry and juveniles were sacrificed using ice-freezing (as this common method used to kill fish does not affect natural isotopic ratios), electronarcosis or an overdose of chemical anesthetics (2-phenoxyethanol, benzocaine and clove oil). For fry, we sampled the whole animal whereas, for juveniles, white dorsal muscle, liver, red blood cells, plasma, external tegument and pectoral fin were sampled. Isotopic ratios and the elemental compositions of carbon and nitrogen were then measured.

RESULTS

The δ¹⁵ N values, and the C and N contents of all considered tissues as well as δ¹³ C values of muscle, liver, red blood cells and plasma, were not affected by the use of chemical anesthetics. Clove oil and to a lesser extent 2-phenoxyethanol and benzocaine decreased δ¹³ C values of whole fry and juvenile external tegument and pectoral fin. The use of electronarcosis drastically affects the δ¹³ C and δ¹⁵ N values of all fish tissues.

CONCLUSIONS

Anesthetics should be avoided for δ¹³ C analysis when tissues are in contact with the water containing the anesthetic. Ice-immersion has to be preferred when approved by guidelines. If not, benzocaine and 2-phenoxyethanol should be preferred over clove oil. Electronarcosis should not be used to kill fish until further investigations are performed.

Underestimation of chemical contamination in marine fish muscle tissue can be reduced by considering variable wet:dry weight ratios

Whether considered as a risk for human health or as ecological tracers, contaminants' concentrations measured in fish muscles are commonly expressed relative to wet or dry mass. Comparison of results required conversion factors (CF) but accurate values are scarce and case-specific. The present paper is aimed at investigating errors linked with the use of the theoretical value. Muscles dry and wet masses were measured in 15 fish species to determine the actual CF. Most CF were lower than the theoretical wet:dry ratio of 5 classically used, with variations at individual and species level. Muscle lipid content (inferred by C/N ratios) was a crucial factor explaining discrepancies, claiming for caution when working with lipid-rich species. The observed variability demonstrated that using the theoretical CF may be inaccurate, when actual CF largely differs from the theoretical value. Dedicated measurement is the better approach when accuracy is required.

Energy allocation strategy of skipjack tuna Katsuwonus pelamis during their reproductive cycle

The lipid composition of somatic and reproductive tissues was determined for female skipjack tuna Katsuwonus pelamis caught in the western Indian Ocean between latitude 10° N and 20° S and longitude 40° and 70° E. The highest total lipid (TL) contents were in the liver and gonads, with white muscle levels approximately three-fold lower. Three lipid classes dominated: triacylglycerols (TAG), sterol esters and wax esters (SE-WE) and phospholipids (PL). Collectively, these accounted for between 70 and 80% of TLs. Changes in lipid concentrations were evaluated over the maturation cycle. Immature fish had the lowest gonad and liver TL levels; concentrations of TL, TAG, SE-WE and PL accumulated from immature to mature (spawning-capable) phase, reflecting sustained vitellogenic activity of the liver and a transfer of lipids to developing oocytes from the onset of vitellogenesis. Gonado-somatic and hepato-somatic indices were positively correlated with each other and positively related to TL in the gonads and liver. Fulton's condition index and lipid concentrations in muscle did not vary significantly over the maturation cycle; fat content in the main storage tissues was undepleted as the ovary developed. Hence, K. pelamis apparently supports reproduction directly from food intake over the breeding season. In the gonads, reserve lipids (SE-WE and TAG) and sterols were related to batch fecundity but this was not the case for somatic and hepatic tissues. These results suggest that K. pelamis utilizes an income breeding strategy.

Persistent Organic Pollutants in albacore tuna (Thunnus alalunga) from Reunion Island (Southwest Indian Ocean) and South Africa in relation to biological and trophic characteristics

The contamination of albacore tuna (Thunnus alalunga) by Persistent Organic Pollutants (POPs), namely polychlorinated biphenyls (PCBs) and dichlorodiphenyl-trichloroethane (DDT), was investigated in individuals collected from Reunion Island (RI) and South Africa's (SA) southern coastlines in 2013, in relation to biological parameters and feeding ecology. The results showed lower PCB and DDT concentrations than those previously reported in various tuna species worldwide. A predominance of DDTs over PCBs was revealed, reflecting continuing inputs of DDT. Tuna collected from SA exhibited higher contamination levels than those from RI, related to higher dietary inputs and higher total lipid content. Greater variability in contamination levels and profiles was identified in tuna from RI, explained by a higher diversity of prey and more individualistic foraging behaviour. PCB and DDT contamination levels and profiles varied significantly in tuna from the two investigated areas, probably reflecting exposure to different sources of contamination.

Fish tissue lipid-C:N relationships for correcting δ(13) C values and estimating lipid content in aquatic food-web studies

RATIONALE

Normalizing δ(13) C values of animal tissue for lipid content is necessary to accurately interpret food-web relationships from stable isotope analysis. To reduce the effort and expense associated with chemical extraction of lipids, various studies have tested arithmetic mass balance to mathematically normalize δ(13) C values for lipid content; however, the approach assumes that lipid content is related to the tissue C:N ratio.

METHODS

We evaluated two commonly used models for estimating tissue lipid content based on C:N ratio (a mass balance model and a stoichiometric model) by comparing model predictions to measure the lipid content of white muscle tissue. We then determined the effect of lipid model choice on δ(13) C values normalized using arithmetic mass balance. To do so, we used a collection of fish from Lake Superior spanning a wide range in lipid content (5% to 73% lipid).

RESULTS

We found that the lipid content was positively related to the bulk muscle tissue C:N ratio. The two different lipid models produced similar estimates of lipid content based on tissue C:N, within 6% for tissue C:N values <7. Normalizing δ(13) C values using an arithmetic mass-balance equation based on either model yielded similar results, with a small bias (<1‰) compared with results based on chemical extraction.

CONCLUSIONS

Among-species consistency in the relationship between fish muscle tissue C:N ratio and lipid content supports the application of arithmetic mass balance to normalize δ(13) C values for lipid content. The uncertainty associated with both lipid extraction quality and choice of model parameters constrains the achievable precision of normalized δ(13) C values to about ±1.0‰.