https://orcid.org/0000-0001-7713-8348
Figures
Abstract
In this study, we applied multiple reaction monitoring (MRM)-profiling to explore the relative ion intensity of lipid classes in plasma samples from sea turtles in order to profile lipids relevant to sea turtle physiology and investigate how dynamic ocean environments affect these profiles. We collected plasma samples from foraging green (Chelonia mydas, n = 28) and hawksbill (Eretmochelys imbricata, n = 16) turtles live captured in North Pacific Costa Rica in 2017. From these samples, we identified 623 MRMs belonging to 10 lipid classes (sphingomyelin, phosphatidylcholine, free fatty acid, cholesteryl ester, phosphatidylserine, phosphatidylinositol, phosphatidylglycerol, phosphatidylethanolamine, ceramide, and triacylglyceride) and one metabolite group (acyl-carnitine) present in sea turtle plasma. The relative ion intensities of most lipids (80%) were consistent between species, across seasons, and were not correlated to body size or estimated sex. Of the differences we observed, the most pronounced was the differences in relative ion intensity between species. We identified 123 lipids that had species-specific relative ion intensities. While some of this variability is likely due to green and hawksbill turtles consuming different food items, we found indications of a phylogenetic component as well. Of these, we identified 47 lipids that varied by season, most belonging to the structural phospholipid classes. Overall, more lipids (n = 39) had higher relative ion intensity in the upwelling (colder) season compared to the non-upwelling season (n = 8). Further, we found more variability in hawksbill turtles than green turtles. Here, we provide the framework in which to apply future lipid profiling in the assessment of health, physiology, and behavior in endangered sea turtles.
Citation: Clyde-Brockway CE, Ferreira CR, Flaherty EA, Paladino FV (2021) Lipid profiling suggests species specificity and minimal seasonal variation in Pacific Green and Hawksbill Turtle plasma. PLoS ONE 16(7): e0253916. https://doi.org/10.1371/journal.pone.0253916
Editor: John A. B. Claydon, Ocean Frontier Institute, CANADA
Received: January 12, 2021; Accepted: June 16, 2021; Published: July 19, 2021
Copyright: © 2021 Clyde-Brockway et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data are stored within Purdue's repository system. https://purr.purdue.edu/publications/3752/1.
Funding: Funding for this project was provided by the Department of Forestry and Natural Resources at Purdue University, West Lafayette, IN, The Leatherback Trust, Fort Wayne, IN and Equipo Tora Carey, San Jose, CR (to CCB). This work was also supported by the USDA National Institute of Food and Agriculture, Hatch Project 1019737 (to EAF) and Jack W. Schrey Distinguished Professorship funds (to FVP). Funding sources had no role in study design, analysis or interpretation of data, or writing of final report.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Wildlife face diverse and complicated anthropogenic threats, in addition to naturally occurring environmental stressors. Anthropogenic threats to sea turtles are widely studied and clearly defined [1], but to adequately protect these endangered species, it is important to understand how natural environmental variability influences their behavior and physiology. Animal behavior and physiology are reflected in plasma metabolites, such as lipid profiles, which also contain information about an organism’s diet, environment, and health [2–8]. Generally, lipids can be divided into structural lipids, storage lipids, and signaling lipids. Structural lipids make up cell membranes with chain lengths and saturation determining membrane fluidity [9–11] and include the phospholipids, specifically sphingomyelins, phosphatidylcholines, phosphatidylserines, phosphatidylinositols, phosphatidylglycerols, and phosphatidylethanolamines. Storage lipids are mostly dietary lipids that are consumed and either oxidized as energy or stored for future energy use, primarily free fatty acids and triacylglycerides [12–14]. Signaling lipids trigger specific cell responses and are precursors for hormones or carrier lipids [15].
Because most sea turtles are threatened or endangered, research has emphasized the understanding of life history patterns to better guide conservation strategies. To date, lipid studies in sea turtles have focused on foraging ecology [6] and reproductive biology [16–18]. However, lipid dynamics can also provide information on other aspects of physiology and behavior, such as chemosensory communication [19–21], oxidative stress associated with migration and feeding [22–24], lung structure and function [25–27], and as indication of disease status [28–31] and aging [32]. Comprehensive lipid-profiling is a powerful exploratory tool that highlights latent aspects of animal ecology and health through characterizations of lipid quantities relative to the whole profile. Further, targeted assessment required to quantify lipid concentrations depends on knowledge of specific lipids, fatty acyl chain composition, and saturation. Therefore, our first objective was to screen lipids and metabolites present in sea turtle plasma, using the Multiple Reaction Monitoring (MRM)-profiling method [33–36]. We present lipid profiles in foraging non-reproductive green (Chelonia mydas, Linnaeus, 1758) and hawksbill (Eretmochelys imbricata, Linnaeus, 1766) sea turtles from Costa Rica, including specific chain lengths and saturation levels during normal El Niño conditions [37]. Subsequently, we quantified how relative ion intensity of lipid profiles varied between species of sea turtles and across seasons to determine if phylogeny, diet, size, or temperature-influenced lipid profiles. Understanding how turtles relate to their dynamic environments will allow for specific conservation efforts better tailored to sea turtle’s needs.
Methods
Ethics statement
This study was carried out in accordance with institutional, federal, and international guidelines and permits. All data were collected under the protocol approved by Purdue Animal Care and use Committee (protocol #1510001309). Permissions to work with endangered sea turtles in Costa Rica were granted under permits from Guanacaste Conservation Areas (ACG) of the Ministry of Environment of Costa Rica (ACG-PI-PC-019, R-ACG-057-2016). Finally, samples were approved for importation into the United States by USFWS and CITES (CITES permit 17US06369C/9).
Study area
We conducted this study during 2017 in Matapalito Bay (10.93°N; -85.79°W) and Salinas Bay (11.04°N; -85.70°W) in North Pacific Costa Rica. This area represents one of three major upwelling areas along the Central American coast [38]. Costa Rican upwelling is especially strong because of the interaction between trade winds and Costa Rica Dome water patterns [38–40]. Therefore, sea turtles inhabiting these waters are exposed to unusually cold-water temperatures (< 20°C at 10 m) during the upwelling season (UP; November–March), and warmer temperatures (~28°C at 10 m) during the non-upwelling season (NUP; April–October). The transition time between seasons is not definite and varies between years. Additionally, the Eastern Pacific is exposed to multiannual (~4 yr) water cycles known as the El Niño Southern Oscillation (ENSO). During El Niño years, water temperatures are warmer than usual (decreased primary productivity) while La Niña years are characterized by colder than normal temperature and increased upwelling [37]. In this study, we collected samples during normal El Niño conditions (https://ggweather.com/enso/oni.htm).
Study animals
In North Pacific Costa Rica, black and yellow morphotype green turtles (Chelonia mydas), collectively referred to as green turtles, and hawksbill turtles (Eretmochelys imbricata), inhabit coastal sites concurrently [41]. We live captured black (8 juvenile, 6 male, 3 female), yellow (9 juvenile, 2 female), and hawksbill (15 juvenile, 1 female) turtles using tangle nets [42] deployed in each bay once a month; however, we canceled the monthly sampling during dangerous weather. We opportunistically hand-caught turtles encountered when snorkeling. We brought turtles into the boat for processing, where we measured curved carapace length measured from the nuchal notch to the most posterior tip of the caudal peduncle (CCL; flexible measuring tape; ± 0.5 cm), and mass (Detecto 11S200HKG “S” hook hanging scale, Webb City, MO; ± 0.5 kg). To identify individuals and avoid pseudo-replication, we marked turtles with a unique passive integrated transponder (PIT) tag injected into the right shoulder (AVID2028 FriendChip, Norco, California, USA), and metal flipper tags (Style 681IC, National Band and Tag Company, Newport, KY, USA) in both hind flippers [43]. We grouped individuals into age/sex class by size, where juvenile green turtles had CCL ≤ 76 cm and adult turtles had a CCL > 76 cm, because the smallest turtle in this study with defining male characteristics was 77 cm CCL. This classification combined what other studies classified as sub-adult and adult classes [44, 45]. In hawksbill turtles, we classified juveniles as < 70 cm CCL and adults as > 71 cm CCL [46, 47].
Plasma collection and preparation
We collected a single blood sample (< 1 ml/kg) per individual using a 21 g non-heparinized needle from the cervical sinus [48], transferred the sample to a lithium-heparin tube, and stored it on ice until returned to the lab. Whole blood samples were centrifuged for 5 min at 1145 g (CAT No. 0151, Clay Adams Analytical Centrifuge, New York, NY). We isolated the plasma in a non-heparinized tube, and stored the plasma at -18°C for up to 1 year until we transported samples back to Purdue University. Samples were then subsequently stored at -20°C until analyses (within ~9 months).
Lipid analysis
For lipid identification, we used a recently published strategy, MRM-profiling mass spectrometry, which is a two-step process to identify lipid profiles [33–36]. The first step involved generating and analyzing a reference sample, for this step we combined 12.5 μL from 16 hawksbill turtles and 16 green turtles (randomly selected) to generate two 200 μL reference samples. We extracted lipids and fat-soluble metabolites from plasma samples following the chloroform-methanol Bligh and Dyer [49] lipid extraction procedure and stored extracted lipids at -80°C until analysis. For analysis, we resuspended dried samples in ACN+MeOH+300nM NH4Ac 3:6.65:0.35 (v/v).
We compiled a set of methods by combining m/z for all molecular ions based on the LipidMAPS online database (http://www.lipidmaps.org/) with expected ions resulting from the Prec or NL scans (1,412 lipids). We combined all potential MRMs into 11 methods (no more than 200 MRM per method). For each method, we used a capillary pump connected to the autosampler (G1377A) then connected to an Agilent 6410 QQQ mass spectrometer (Agilent Technologies, Santa Clara, CA, USA). One sample injection was performed to profile each lipid class. We processed the raw MRM mass spectrometry data using an in-house script and MRM transitions and exported the resulting ion intensity values to Microsoft Excel (v2016, Microsoft Corporation, Redmond, WA). We normalized the absolute ion intensity of lipids against the total ion current of the method for each turtle individually. We then selected any lipid that had a sample ion intensity higher than the blank, resulting in 623 baseline lipids and metabolites from 10 lipid classes and one metabolite class and combined these 623 lipids and metabolites into four methods.
This allowed us to identify and process only lipids and metabolites with ion intensities greater than that of the blank, suggesting significance to sea turtle physiology. The second step involved analyzing the individual turtle plasma samples for the identified baseline lipids. We used the same procedure except we modified the extraction procedure to work with 20 μL of plasma (due to sensitivity of the technique). MRM lipid-profiling provides information on diverse lipid classes (sphingomyelin, phosphatidylcholine, free fatty acid, cholesteryl ester, phosphatidylserine, phosphatidylinositol, phosphatidylglycerol, phosphatidylethanolamine, ceramide, and triacylglyceride), one metabolite group (acyl-carnitine), and aids in screening for shifts in lipid physiology that indicate environmental pressures, metabolic impairments, and resource allocation [14, 15, 28, 29, 33]. Even though the method proportionates tentative identification of the lipids at the species level, it allows a comprehensive lipid profiling and the understanding of the most important lipid classes and specific candidates related to a physiological or pathological condition. While we normalized each turtle profile by method in step one, here we normalized the absolute ion intensity of lipids and metabolites against the total ion current of the lipid class for each turtle. The resulting relative ion intensity of lipids, therefore, represented the proportion of total lipid class ion current for each turtle. For the acyl-carnitine metabolite class, we only included the biologically significant lipid fragment (85.1).
Statistical analysis
We conducted univariate and multivariate analyses on relative ion abundances using MetaboAnalyst 4.0 (http://www.metaboanalyst.ca, Rv3.4.3 [50, 51]). For each analysis, we normalized the data within MetaboAnalyst using the auto-scaling method (mean-centered and divided by the standard deviation of each variable). As an overview, we used principal component analyses and a partial least-squares discriminant analyses to evaluate clustering of turtles based on lipid profiles. Within MetaboAnalyst, we used the permutation test function with 1000 iterations and the cross-validation (CV) results to determine if the partial least-squares discriminant analysis (PLS-DA) model fit [52]. If our partial least-squares discriminant analysis permutations test was not significant or CV lacked predictive ability, we used our principal component analysis model. To identify differential lipids, we used Fold Change analysis (FC >2; principal component analyses data) or variable importance in the projection (VIP; partial least-squares discriminant analyses data) and univariate analyses (two-group data: t-test and fold change analysis; multi-group: one-way ANOVA with Fisher’s Least Significant Difference post hoc analysis), for each lipid or metabolite class separately. We classified differential lipids as those with VIP values > 1 or fold change threshold of 2, and significant univariate test results (p < 0.05). To investigate the effects of turtle size (curved carapace length and body mass), we conduced Pearson R correlations. We classified lipids into 5 categories based on correlation coefficient (r): strong negative correlation (-1.0 –-0.6), weak negative correlation (-0.59 –-0.4), no correlation (-0.39–0.39), weak positive correlation (0.4–0.59), and strong positive correlation (0.6–1.0).
We analyzed each lipid/metabolite class separately. First, we compared black turtles between bays and determined no significant difference across habitats. Therefore, for subsequent analyses we combined turtles from both bays. Next, we tested seasonal effects in black turtles (upwelling: November–March; non-upwelling: April–October). If we did not identify differential lipids between seasons, then we combined all black turtles and compared relative ion intensity for each lipid class between black and yellow morphotype green turtles and all green turtles across seasons. Assuming statistical similarity within all green turtles, we compared green turtles to hawksbill turtles and all individuals across seasons. When groups were not statistically similar, we used a one-way ANOVA to compare species and seasons simultaneously by dividing the data into 4 or 6 groups, depending on the analysis (by population and season). Further, we compared turtle size (curved carapace length and mass) between populations and across seasons and locations using Two-Way Analysis of Variance (ANOVA; black turtles), One-Way Analysis of Variance (ANOVA; yellow turtles) or Kruskal-Wallis tests (hawksbill turtles, population) in R statistical software (Version 3.4.4, [53]) and tested sex/age class differences in lipid relative ion intensity. In all analyses, we considered statistical significance as a threshold of p < 0.05. Because we were using relative ion intensity, higher values suggested more lipid present in plasma, however this was not comparable to concentrations. Further, we built our profiles by lipid class, therefore relative ion intensities could not be compared across lipid classes, only within classes.
Besides evaluating individual (univariate statistics) and a combination of lipids (multivariate statistics), the diversity of lipid structures present in the sea turtle plasma was relative to total fatty acyl chain length and unsaturation levels. For that, we divided each lipid class by carbon chain length (short, medium, long) and saturation level (0 = unsaturated, 1–2 = mono and di-unsaturated lipids, 3+ = polyunsaturated lipids) and presented the results as percent (%) of lipid class profile.
Results
We analyzed 44 plasma samples from sea turtles ranging from 38.5–92.0 cm curved carapace length and 6.0–88.0 kg body mass (Table 1). Due to scale malfunction, we recorded body mass of 39 turtles (out of 44 turtles; Table 1). Throughout 2017, we captured 26 turtles during the non-upwelling season and 19 turtles during the upwelling season (Table 1). Turtle length and body mass did not vary with season (F1,14 = 0.706, p = 0.415 and F1,12 = 0.158, p = 0.698, respectively) or location (F1,14 = 0.007, p = 0.934 and F1,12 = 0.000, p = 0.996, respectively) in black morphotype green sea turtles, or by season in yellow morphotype (F1,9 = 0.483, p = 0.505 and F1,5 = 0, p = 0.998, respectively) and hawksbill turtles (χ21 = 1.697, p = 0.193 and χ21 = 1.381, p = 0.24, respectively). However, the three populations were significantly different in curved carapace length (χ22 = 24.694, p < 0.001) and body mass (χ22 = 23.022, p < 0.001).
Lipid and metabolite profiles
We identified 623 baseline MRMs (referred as lipids or acyl-carnitines; all attributions are tentative) from 11 classes by using their relative ion intensities. These lipids and acyl-carnitines presented ion intensities higher than the blank sample in the plasma samples of the sea turtles (Table 2 and S1 Table). Among the structural lipids, most of the phosphatidylcholine and phosphatidylserine class profile consisted of short chain lipids (> 50%) while most of the phosphatidylethanolamine and phosphatidylinositol class profile consisted of long chain lipids (> 75%; Fig 1). The phosphatidylglycerol profile was evenly distributed across lipid lengths. Among the dietary, signaling, and fat storage lipids, short chain lipids comprised a higher percentage of total lipid profile in free fatty acids and triglycerides while most cholesteryl ester profile was comprised of medium to long lipid lengths (Fig 1). Saturation graphs revealed a higher percentage of unsaturated lipids in our phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol profiles with no difference in saturation across phosphatidylglycerol and phosphatidylserine profiles (Fig 2). In our dietary and storage lipids, most cholesteryl ester and triglyceride lipids were unsaturated while most free fatty acids were saturated (Fig 2). We did not include sphingomyelin, acyl-carnitine, or ceramide classes due to structure (S1 Table).
The x-axis represents chain length and was divided based on each profile separately. Bars and error bars represent mean ± standard deviation (%). We did not include sphingomyelin, acyl-carnitine, or ceramide classes due to structure. PC = phosphatidylserine (short = 30–36 carbons, medium = 38–42 carbons, long = 44–48 carbons); PE = phosphatidylethanolamine (short = 12–22 carbons, medium = 24–34 carbons, long = 36–42 carbons); PG = phosphatidylglycerol (short = 12–22 carbons, medium = 24–34 carbons, long = 36–44 carbons); PI = phosphatidylinositol (short = 12–22 carbons, medium = 24–34 carbons, long = 36–44 carbons); PS = phosphatidylserine (short = 14–24 carbons, medium = 26–36 carbons, long = 38–44 carbons); CE = cholesteryl ester (short = 12–15 carbons, medium = 16–19 carbons, long = 20–24 carbons); FFA = free fatty acid (short = 12–18 carbons, medium = 20–26 carbons, long = 28–34 carbons); TG = triacylglyceride (short = 48–50 carbons, medium = 52–54 carbons, long = 56–60 carbons).
Bars and error bars represent mean ± standard deviation (%). The x-axis represents saturation level where saturated indicates lipids 0 double bonds, unsaturated indicates lipids with 1 or 2 double bonds, and polyunsaturated indicates lipids with ≥3 double bonds. We did not include sphingomyelin, acyl-carnitine, or ceramide classes due to structure.
Most (500) baseline lipids had no relationship with turtle size. Of the 115 lipids that had significant correlation coefficients (r), we identified 64 that were negatively correlated to curved carapace length and 9 lipids that were negatively correlated to body mass (Table 2). Additionally, we identified 27 lipids that were positively correlated to curved carapace length and 15 that were positively correlated to body mass; 14 lipids were correlated to both curved carapace length and body mass (Table 2). We did not identify any correlation between acyl-carnitines (a metabolite class) and curved carapace length or body mass (Table 2). Results from our partial least-squares discriminant analyses revealed clear differentiation between green turtles and hawksbill turtles (Fig 3). When we included season (NUP = non-upwelling; UP = upwelling), we found slight deviation from clustering based on species (Fig 4). In the phosphatidylcholine class, which had 8 lipids that varied between seasons in hawksbills and 5 lipids that varies by season in green turtles (Table 4), both PLS-DA scores plots showed that the non-upwelling season had a larger cluster shift to the left. In the sphingomyelin class, which had 8 lipids that varied by season in hawksbill turtles and 5 lipids that varies by season in green turtles, the upwelling season had a larger spread in green turtles and similar cluster size but shift to right in hawksbill turtles (Fig 4).
Partial least-squares discriminant analysis (PLS-DA) component 1 (x-axis) and component 2 (y-axis) demonstrating species clusters as a result of plasma lipid profile in green turtles (CM, red triangle) and hawksbill turtles (EI, green plus sign). We live captured all turtles in North Pacific Costa Rica in 2017. Abbreviations: CE = cholesteryl ester; PC = phosphatidylcholine; PE = phosphatidylethanolamine; SM = sphingomyelin; PI = phosphatidylinositol; TG = triacylglyceride; Cer = ceramide; FFA = free fatty acids; PG = phosphatidylglycerol; PS = phosphatidylserine; Car = acyl-carnitine.
Partial least-squares discriminant analysis (PLS-DA) component 1 (x-axis) and component 2 (y-axis) demonstrating species clusters as a result of plasma lipid profile in green turtles (CM) and hawksbill turtles (EI) and season (NUP = non-upwelling; UP = upwelling). We live captured all turtles in North Pacific Costa Rica in 2017. Abbreviations: CE = cholesteryl ester; PC = phosphatidylcholine; PE = phosphatidylethanolamine; SM = sphingomyelin; PI = phosphatidylinositol; TG = triacylglyceride; Cer = ceramide; FFA = free fatty acids; PG = phosphatidylglycerol; PS = phosphatidylserine; Car = acyl-carnitine.
Differential lipids
We defined differential lipids as those that presented significantly different relative ion intensities between species or season (Table 2). Our univariate analyses and VIP/Fold-change scores identified 123 differential lipids (MRMs): 115 that demonstrated species specificity and 47 that varied by seasons (some lipids varied by species and season; Table 2). Habitat was not significant for any lipid or metabolite, therefore black turtles from both sites were combined in our analyses. Further, we did not identify any sex/age class differences in relative ion intensity. Species variability was lipid specific and hawksbill turtles had higher relative ion intensities in more lipids (Table 3). Additionally, we discovered that relative ion intensities of differential lipids were higher in the upwelling season compared to the non-upwelling season (Table 4 and S2 Table). While most lipids did not vary within green turtle morphotypes, 19 lipids in the phosphatidylcholine class varied between seasons in black morphotype green turtles or between black and yellow morphotype green turtles (S2 Table). One lipid in the phosphatidylglycerol class (PG 32:0) had a higher relative ion intensity in black morphotype green turtles compared to yellow morphotype green turtles. Finally, five sphingomyelin lipids had higher average relative ion intensities in the upwelling season in black morphotype green turtles (S2 Table). The acyl-carnitine and phosphatidylserine classes displayed no species or seasonal variability in relative ion intensity.
Discussion
Establishing metabolomic approaches to studying endangered species is a key step to assess adaptations and facilitate understanding of their physiology in dynamic environments [8, 31, 55, 56]. This has important ecological implications, particularly with changes in ocean temperature and nutrient variability that accompanies global climate change [57, 58]. Using a single plasma sample of 20μL and the MRM-profiling method, we were able to profile major lipid and metabolite classes that are relevant to sea turtle biology. Quantification of lipid and metabolite concentrations in plasma usually requires the use of larger amounts of samples and deuterated standards that are only effective if specific targeted lipids, chain lengths, and saturation levels are known. Therefore, using the data we present here (S1 Table), and data from Ahmadireskety et al. [8] researchers can compare and further investigate concentrations and patterns of lipid and metabolite expression within sea turtles. Further, due to the similarity between lipid species in green and hawksbill turtles (this study) and those in other studies of green, loggerhead (Caretta caretta), and leatherback sea turtles (Dermochelys coriacea), we predict that our data are applicable to other sea turtle species foraging in other environments throughout world [8, 59].
In our study, fewer than 20% of the lipids identified reflected species differences and/or seasonality in relative ion intensity in sea turtles foraging in Pacific Costa Rica. Among these, the most remarkable difference we observed in relative ion intensities of lipids was between species, suggesting variability is due to diet, genetics/phylogeny or age/sex, and proximity to initiation of migration (Table 3, Fig 3). In tuatara (Sphenodon punctatus [60, 61]), lizards (Amphibolurus nuchalis [62]), alligators (Alligator mississippiensis [63, 64]), frogs (Leptodactylus fallax [65]), fish [66], and birds [3, 5, 67] lipid compositions of diets influenced plasma lipid profiles. When foraging within the same habitat, green and hawksbill turtles have minimal overlap in diets [46, 68, 69], therefore diet could explain some of the variability in lipid profiles and partial least-squares discriminant analysis clusters. However, diet would not explain consistency in lipid relative ion intensity found between black and yellow morphotype green turtles. Studies suggest that black and yellow morphotype turtles tend to have distinct diets [46, 70], yet, our partial least-squares discriminant analysis and univariate analyses grouped them together.
Alternatively, in birds (Anatidae), phylogeny could be more important than dietary differences (yolk lipids) [71, 72]. While in fish (L. fabricii and L. maculatus), total lipid levels are genetically determined while habitat may determine lipid lengths and saturation levels [73]. Similarly, a study on great tits (P. major) suggests there is an interaction between diet and temperature effects on plasma lipid saturation level, and this effect is reflected in metabolic rate [2]. Therefore, lipid clustering observed in this study likely has a genetic or phylogenetic component similar to nesting green, loggerhead and leatherback turtles in Florida [8], but there may be an interaction effect we did not measure.
We identified size differences between species of turtle in this study; however, our correlation analyses suggest consistency of lipid profiles through age classes with smaller turtles potentially growing at a slightly faster rate than larger turtles (Table 2 [74, 75]). In loggerhead sea turtles (Caretta caretta), size related changes in cholesterol and triglycerides suggest shifts in growth rate and diet as turtles recruit from oceanic to neritic foraging habitats [74, 75]. Therefore, the triglycerides in this study that varied with curved carapace length may be an indication of turtles that recently underwent a shift in diet and growth rate. Mitochondrial phospholipids vary with age and growth in fish; however, our correlation results suggest no such relationship in our sea turtles [32]. We did not find sex/age class variability in our turtles despite our principal component analysis showing some clustering (S1 Fig) likely due to sample size or size classification method [44–47]. However, it is important to note that reproduction in reptiles can have dramatic effects on lipid profiles [12, 76, 77]. Some of the individual variability observed in our profiles could be caused by turtles approaching reproductive maturity and onset of migration, although lipid profiles do not indicate initiation of vitellogenesis [16, 17], and our turtles were small (Table 1). Therefore, we suggest that while the species variability in plasma lipids in sea turtles could be partially due to variability in diet, age, or sexual maturation, phylogeny is likely an important indicator of lipid profiles in sea turtles.
Our study identified lipid profile compositional differences between lipid classes that did and did not vary by season. Profiles of structural lipids that displayed seasonality (phosphatidylcholine, phosphatidylinositol, and phosphatidylethanolamine) favored unsaturation compared to saturated lipid species (Fig 2). While phosphatidylcholine profile had a higher proportion of shorter chain lipids, phosphatidylinositol and phosphatidylethanolamine profiles favored longer lipids (Fig 1). These structural lipid classes are related to membrane fluidity [10]. Seasonality in structural lipids suggest that our sea turtles might be displaying homeoviscous adaptation, similar to other amphibians and reptiles [78, 79]. Decreases in membrane rigidity are associated with increases in saturation level of lipids and decreases in lipid chain lengths [9, 13, 78], which maintains trans-lipid transport and metabolism in cold environments. In our study, 22 structural lipids had higher relative ion intensity in the upwelling season when water temperatures were colder (compared to 6 that had higher relative ion intensity in the non-upwelling season; Table 4). Of all lipids, we found more seasonality in hawksbill turtles than in green turtles, suggesting that hawksbill turtles are more affected by upwelling seasonal variability in nutrient availability and temperature than green turtles, potentially due to size differences between populations. Therefore, although sea turtles avoid areas with cold-water [80, 81], seasonal viability in metabolite and lipid profiles may have cold protective abilities that could permit cold-water foraging [56, 82] and minimized migration time [83].
In foraging studies, triacylglycerides, free fatty acids, cholesteryl ester, and acyl-carnitines are the classes used to predict foraging and metabolism [6]. We recorded lower relative ion intensities in triacylglycerides (n = 5), free fatty acids (n = 1) and cholesteryl esters (n = 2) during the non-upwelling season, suggesting slight decreases in foraging when upwelling decreases [6, 84, 85]. In addition, starvation in reptiles is accompanied by a shift from saturated to unsaturated free fatty acids [6, 84], while only one free fatty acid varied by season in this study (and only in hawksbill turtles), it was a highly unsaturated lipid (C22:6; Table 4). However, these conclusions are based on minimal observations (triacylglycerides = 4; free fatty acids = 1; cholesteryl ester = 2) and most of our data suggests consistency in foraging between seasons despite potentially dramatic water temperature changes.
In this study we investigated lipid profiles and relative ion intensities, as opposed to concentrations of specific lipids. This method does not allow us to specifically compare lipid concentrations between our study and the few other studies that have measured lipid concentrations. More specific targeted analysis, such as double bond position in glycerolipids fatty acyl chains [86, 87], is the next step to investigating lipid dynamics in sea turtles. Without specific lipid concentration, we were careful to limit our conclusions to trends in lipid expression and avoid definitive statements about overall amount of lipid present. Further, with the analytical strategy used here, we can only compare lipids within classes and cannot compare lipids across lipid classes. We have identified which lipids are present in sea turtle plasma and using this information, future analysis should build on this work and involve targeted concentrations or variability in genetic expression of specific lipids in sea turtles. Specifically, we suggest investigation of different conditions in sea turtles such as reproduction [12, 16], habitat [2, 79], captivity [61], foraging, body composition and health [31, 46, 88, 89]. Additionally, comparative studies between tissue types may provide insight into gross physiology [90]. Finally, sea turtles are exposed to unique environments as they dive and surface and lipid profiling may help scientists investigate dive capacity in sea turtles [26, 91]. In this study we included ~8 plasma samples with hemolysis, while Stacy et al. [92] suggests hemolysis can alter plasma concentrations of cellular components such as proteins and potassium, they did not quantify the effects of hemolysis on lipid concentrations. Physiologically, it is more likely that it is the lipid presence in blood that causes hemolysis [93]. If we removed samples with hemolysis, our lipid profiles may be biased. Because we processed all samples similarly, the lipid profiling should reflect the physiological state of the turtles. Future research should address whether hemolysis influences lipid profiles in sea turtles.
Conclusions
To address the limited literature on lipid profiles in the plasma of sea turtles, we report lipid profiling results including chain lengths and saturation levels of a diverse group of lipid classes (sphingomyelin, phosphatidylcholine, free fatty acids, cholesteryl esters, phosphatidylserine, phosphatidylinositol, phosphatidylglycerol, phosphatidylethanolamine, ceramides, and triacylglycerides), and one metabolite group (acyl-carnitines) to support future research into sea turtle lipidomics. We observed a general consistency in the composition of lipid profiles in sea turtles foraging in Pacific Costa Rica, with 20% of lipids measured having significantly different relative ion intensity between either season or species. Our results suggest that lipid profiles are determined partially by phylogeny, but may have a diet, maturation, or habitat component as well. Further, based on foraging lipids analyzed, we suggest that foraging is more consistent in green turtles throughout the year than hawksbills, with lower relative ion intensities present in the non-upwelling season. Finally, we suggest that lipid profiles indicated profile changes between seasons that are consistent with homeoviscous adaptation to changing water temperature. MRM-profiling is a powerful tool allowing researchers to evaluate wildlife foraging across habitats and over time, which has implications for both the future reproductive output of the population and ecosystem management [94, 95]. This study provides the baseline for extending MRM-profiling to sea turtles through a simple, fast, and minimally invasive technique that fosters future investigations of sea turtle health, physiology, and ecology.
Supporting information
S1 Fig. Partial least-squares discriminant analysis displaying age and sex clusters by plasma lipid profiles.
Partial least-squares discriminant analysis (PLS-DA) component 1 (x-axis) and component 2 (y-axis) demonstrating age class and sex clusters as a result of plasma lipid profile in green turtles (CM). We live captured all turtles in North Pacific Costa Rica in 2017. Abbreviations: CE = cholesteryl ester; PC = phosphatidylcholine; PE = phosphatidylethanolamine; SM = sphingomyelin; PI = phosphatidylinositol; TG = triacylglyceride; Cer = ceramide; FFA = free fatty acids; PG = phosphatidylglycerol; PS = phosphatidylserine; Car = acyl-carnitine.
https://doi.org/10.1371/journal.pone.0253916.s001
(TIF)
S1 Table. Tentative identification of lipids and metabolites.
Tentative identification of all lipids and metabolites present in sea turtle plasma (relative ion intensity higher than the blank). Lipids and metabolites with significantly different relative ion intensity between species or season are bolded. Abbreviations: CE = cholesteryl ester; PC = phosphatidylcholine; PE = phosphatidylethanolamine; SM = sphingomyelin; PI = phosphatidylinositol; TG = triacylglyceride; Cer = ceramide; FFA = free fatty acids; PG = phosphatidylglycerol; PS = phosphatidylserine; Car = acyl-carnitine.
https://doi.org/10.1371/journal.pone.0253916.s002
(XLSX)
S2 Table. Average relative ion intensity of seasonal lipids in black morphotype green turtles.
Average relative ion intensity of plasma lipids that varied by season (NUP = non-upwelling; UP = upwelling) in black morphotype green turtles, and plasma lipids that varied by morphotype in green turtles (black and yellow). PC = phosphatidylcholine; SM = sphingomyelin; PG = phosphatidylglycerol.
https://doi.org/10.1371/journal.pone.0253916.s003
(XLSX)
Acknowledgments
We would like to thank The Leatherback Trust and Equipo Tora Carey, their members and volunteers, for assistance with the fieldwork for this study; especially M. Heidemeyer, K. Mora, R. Mora, M. Mora, A. Lara, and M. Giry for assistance in Costa Rica. Additionally, we thank Dr. R. Goforth and Dr. J. Spotila for reviewing the manuscript.
References
- 1.
Lutcavage ME, Plotkin P, Witherington B, Lutz PL. Human impacts on sea turtle survival. In: Lutz PL, Musick JA, editors. The biology of sea turtles vol I. CRC Press; 1997. pp. 387–409.
- 2. Andersson MN, Nilsson J, Nilsson JÅ, Isaksson C. Diet and ambient temperature interact to shape plasma fatty acid composition, basal metabolic rate and oxidative stress in great tits. J Exp Biol. 2018; 221(24): 1–10. pmid:30361459
- 3. Dick MF, Guglielmo CG. Dietary polyunsaturated fatty acids influence flight muscle oxidative capacity but not endurance flight performance in a migratory songbird. Am J Physiol Regul Integr Comp Physiol. 2019; 316(4):R362–R375. pmid:30624975
- 4. González-Domínguez R, Sayago A, Fernández-Recamales Á. Direct infusion mass spectrometry for metabolomic phenotyping of diseases. Bioanalysis. 2017; 9(1): 131–148. pmid:27921460
- 5. Isaksson C, Andersson MN, Nord A, von Post M, Wang HL. Species-dependent effects of the urban environment on fatty acid composition and oxidative stress in birds. Front Ecol Evol. 2017; 5: 44.
- 6. Price ER, Jones TT, Wallace BP, Guglielmo CG. Serum triglycerides and ß-hydroxybutyrate predict feeding status in green turtles (Chelonia mydas): evaluating a single blood sample method for assessing feeding/fasting in reptiles. J Exp Mar Biol Ecol. 2013; 439: 176–180.
- 7. Meyer L, Chambers S, Gervais C, Pethybridge H, Beckmann C, Bruce B, et al. The use of muscle lipids and fatty acids to assess shark diet and condition. J Fish Biol. 2020 Oct 27. pmid:33111319
- 8. Ahmadireskety A, Aristizabal-Henao JJ, Marqueño A, Perrault JR, Stacy NI, Manire CA, et al. Nontargeted lipidomics in nesting females of three sea turtle species in Florida by ultra-high-pressure liquid chromatography–high-resolution tandem mass spectrometry (UHPLC–HRMS/MS) reveals distinct species-specific lipid signatures. Marine Biology. 2020 Sep;167(9):1–8.
- 9. Rawicz W, Olbrich KC, McIntosh T, Needham D, Evans E. Effect of chain length and unsaturation on elasticity of lipid bilayers. Biophys J. 2000; 79(1):328–339. pmid:10866959
- 10. Van Meer G, Voelker DR, Feigenson GW. Membrane lipids: where they are and how they behave. Nat Rev Mol Cell Biol. 2008; 9(2): 112–124. pmid:18216768
- 11. Hazel JR, Williams EE. The role of alterations in membrane lipid composition in enabling physiological adaptation of organisms to their physical environment. Prog Lipid Res. 1990; 29(3): 167–227. pmid:2131463
- 12. Price ER. The physiology of lipid storage and use in reptiles. Biol Rev. 2017; 92(3): 1406–1426. pmid:27348513
- 13. Price ER, Staples JF, Milligan CL, Guglielmo CG. Carnitine palmitoyl transferase activity and whole muscle oxidation rates vary with fatty acid substrate in avian flight muscles. J Comp Physiol B. 2011; 181(4): 565–573. pmid:21153644
- 14. Wenk MR. The emerging field of lipidomics. Nat Rev Drug Discov. 2005; 4(7): 594–610. pmid:16052242
- 15. Pereyra AS, Hasek LY, Harris KL, Berman AG, Damen FW, Goergen CJ, et al. Loss of cardiac carnitine palmitoyltransferase 2 results in rapamycin-resistant, acetylation-independent hypertrophy. J Biol Chem. 2017; 292(45): 18443–18456. pmid:28916721
- 16. Hamann M, Limpus C, Whittier J. Patterns of lipid storage and mobilisation in the female green sea turtle (Cheloniamydas). J Comp Physiol B. 2002; 172(6): 485–493. pmid:12192510
- 17. Kawazu I, Kino M, Yanagisawa M, Maeda K, Nakada K, Yamaguchi Y, et al. Signals of vitellogenesis and estrus in female hawksbill turtles. Zool Sci. 2015; 32(1): 114–118. pmid:25660704
- 18. Kwan D. Fat reserves and reproduction in the green turtle, Chelonia mydas. Wildl Res. 1994; 21(3): 257–265.
- 19. Alberts AC, Sharp TR, Werner DI, Weldon PJ. Seasonal variation of lipids in femoral gland secretions of male green iguanas (Iguana iguana). J Cheml Ecol. 1992; 18(5): 703–712.
- 20. Ibáñez A, Klein C, Quezada G, Krüger M, Brodesser S, Steinfartz S. Characterization of lipid structures in femoral secretions of Galápagos marine iguanas by shotgun lipidomics. Chemoecology. 2018; 28(1): 21–28.
- 21. Khannoon ER, El-Gendy A, Hardege JD. Scent marking pheromones in lizards: cholesterol and long chain alcohols elicit avoidance and aggression in male Acanthodactylus boskianus (Squamata: Lacertidae). Chemoecology. 2011; 21(3): 143–149.
- 22. Landys MM, Piersma T, Guglielmo CG, Jukema J, Ramenofsky M, Wingfield JC. Metabolic profile of long–distance migratory flight and stopover in a shorebird. Proc R Soc B: Biol Sci. 2005; 272(1560): 295–302.
- 23. Jenni-Eiermann S, Jenni L, Piersma T. Plasma metabolites reflect seasonally changing metabolic processes in a long-distance migrant shorebird (Calidris canutus). Zoology. 2002; 105(3): 239–246. pmid:16351872
- 24. Skrip MM, Bauchinger U, Goymann W, Fusani L, Cardinale M, Alan RR, et al. Migrating songbirds on stopover prepare for, and recover from, oxidative challenges posed by long‐distance flight. Ecol Evol. 2015; 5(15): 3198–3209. pmid:26355277
- 25. Daniels CB, Orgeig S, Smits AW, Miller JD. The influence of temperature, phylogeny, and lung structure on the lipid composition of reptilian pulmonary surfactant. Exp Lung Res. 1996; 22(3): 267–281. pmid:8792121
- 26. Gutierrez DB, Fahlman A, Gardner M, Kleinhenz D, Piscitelli M, Raverty S, et al. Phosphatidylcholine composition of pulmonary surfactant from terrestrial and marine diving mammals. Respir Physiol Neurobiol. 2015; 211: 29–36. pmid:25812797
- 27. Johnston SD, Daniels CB, Booth DT. Development of the pulmonary surfactant system in the green sea turtle, Chelonia mydas. Respir Physiol. 2001; 126(1): 75–84. pmid:11311312
- 28. Zhao YY, Cheng XL, Lin RC, Wei F. Lipidomics applications for disease biomarker discovery in mammal models. Biomark Med. 2015; 9(2): 153–168. pmid:25689902
- 29. Cordeiro FB, Ferreira CR, Sobreira TJ, Yannell KE, Jarmusch AK, Cedenho AP, et al. Multiple reaction monitoring (MRM)‐profiling for biomarker discovery applied to human polycystic ovarian syndrome. Rapid Commun Mass Spectrom. 2017; 31(17): 1462–1470. pmid:28656689
- 30. Deng K, Han P, Song W, Wang Z, Zhang F, Xie H, et al. Plasma metabolomic profiling distinguishes right-sided from left-sided colon cancer. Clin Chim Acta. 2018; 487: 357–362. pmid:30296444
- 31. Bembenek Bailey SA, Niemuth JN, McClellan-Green PD, Godfrey MH, Harms CA, Stoskopf MK. 1H-NMR metabolomic study of whole blood from hatchling loggerhead sea turtles (Caretta caretta) exposed to crude oil and/or Corexit. Royal Society open science. 2017 Nov 22;4(11):171433. pmid:29291126
- 32. Almaida-Pagan PF, Ortega-Sabater C, Lucas-Sanchez A, Gonzalez-Silvera D, Martinez-Nicolas A, de Lama MA, et al. Age-related changes in mitochondrial membrane composition of Nothobranchius furzeri.: comparison with a longer-living Nothobranchius species. Biogerontology. 2019; 20(1): 83–92. pmid:30306289
- 33. de Lima CB, Ferreira CR, Milazzotto MP, Sobreira TJ, Vireque AA, Cooks RG. Comprehensive lipid profiling of early stage oocytes and embryos by MRM profiling. J Mass Spectrom. 2018; 53(12): 1247–1252. pmid:30325087
- 34. Dipali SS, Ferreira CR, Zhou LT, Pritchard MT, Duncan FE. Histologic analysis and lipid profiling reveal reproductive age-associated changes in peri-ovarian adipose tissue. Reprod Biol Endocrinol. 2019; 17(1): 46–59. pmid:31189477
- 35. Harlow K, Ferreira CR, Sobreira TJ, Casey T, Stewart K. Lipidome profiles of postnatal day 2 vaginal swabs reflect fat composition of gilt’s postnatal diet. PloS one. 2019; 14(9): e0215186. pmid:31557164
- 36. O’Neil EV, Burns GW, Ferreira CR, Spencer TE. Characterization and regulation of extracellular vesicles in the lumen of the ovine uterus. Biol Reprod. 2020;102(5): 1020–1032. pmid:32055841
- 37.
Wang C, Deser C, Yu JY, DiNezio P, Clement A. El Niño and southern oscillation (ENSO): a review. In Glynn PW, Manzello DP, Enochs IC, editors. Coral reefs of the eastern tropical Pacific. Springer, Dordrecht; 2017. pp. 85–106.
- 38. Lavín MF, Fiedler PC, Amador JA, Ballance LT, Färber-Lorda J, Mestas-Nuñez AM. A review of eastern tropical Pacific oceanography: Summary. Prog Oceanogr. 2006; 69(2–4): 391–398.
- 39. Stuhldreier I, Sánchez-Noguera C, Rixen T, Cortés J, Morales A, Wild C. Effects of seasonal upwelling on inorganic and organic matter dynamics in the water column of eastern Pacific coral reefs. PloS one. 2015; 10(11): e0142681. pmid:26560464
- 40. Broenkow WW. The distribution of nutrients in the Costa Rica Dome in the eastern tropical Pacific Ocean. Limnol Oceanogr. 1965; 10(1): 40–52.
- 41. Bowen BW, Nelson WS, Avise JC. A molecular phylogeny for marine turtles: trait mapping, rate assessment, and conservation relevance. Proc Natl Acad Sci. 1993; 90(12): 5574–5577. pmid:8516304
- 42. Heidemeyer M, Arauz-Vargas R, López-Agüero E. New foraging grounds for hawksbill (Eretmochelys imbricata) and green turtles (Chelonia mydas) along the northern Pacific coast of Costa Rica, Central America. Rev Biol Trop. 2014;62(4): 109–118.
- 43. Heidemeyer M, Delgado-Trejo C, Hart CE, Clyde-Brockway CE, Fonseca LG, Mora R. et al. Long-term in-water recaptures of adult black turtles (Chelonia mydas) provide implications for flipper tagging methods in the Eastern Pacific. Herpetol Rev. 2018; 49(4): 653–657.
- 44. Chaloupka M, Limpus C. Estimates of sex-and age-class-specific survival probabilities for a southern Great Barrier Reef green sea turtle population. Mar Biol. 2005; 146(6): 1251–1261.
- 45. Vander Zanden HB, Bjorndal KA, Mustin W, Ponciano JM, Bolten AB. Inherent variation in stable isotope values and discrimination factors in two life stages of green turtles. Physiol Biochem Zool. 2012; 85(5): 431–441. pmid:22902371
- 46. Bjorndal KA, Bolten AB. Hawksbill sea turtles in seagrass pastures: success in a peripheral habitat. Mar Biol. 2010; 157(1): 135–145.
- 47. Limpus CJ. The hawksbill turtle, Eretmochelys imbricata, in Queensland: population structure within a southern Great Barrier Reef feeding ground. Wildl Res. 1992; 19(4): 489–505.
- 48. Owens DW, Ruiz GJ. New methods of obtaining blood and cerebrospinal fluid from marine turtles. Herpetologica. 1980; 36(1): 17–20.
- 49. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol. 1959; 37(8): 911–917. pmid:13671378
- 50. Chong J, Xia J. MetaboAnalystR: an R package for flexible and reproducible analysis of metabolomics data. Bioinformatics. 2018; 34(24): 4313–4314. pmid:29955821
- 51. Chong J, Soufan O, Li C, Caraus I, Li S, Bourque G, et al. MetaboAnalyst 4.0: towards more transparent and integrative metabolomics analysis. Nucl Acids Res. 2018; 46: W486–W494. pmid:29762782
- 52. Westerhuis JA, Hoefsloot HC, Smit S, Vis DJ, Smilde AK, van Velzen EJ, et al. Assessment of PLSDA cross validation. Metabolomics. 2008; 4(1): 81–89.
- 53.
R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.
- 54. Li M, Butka E, Wang X. Comprehensive quantification of triacylglycerols in soybean seeds by electrospray ionization mass spectrometry with multiple neutral loss scans. Sci Rep. 2014; 4:6581. pmid:25301200
- 55. Schock TB, Keller JM, Rice M, Balazs GH, Bearden DW. Metabotyping of a protected non-model organism, green sea turtle (Chelonia mydas), using 1H NMR spectroscopy and optimized plasma methods for metabolomics. Current Metabolomics. 2013 Dec 1;1(4):279–90.
- 56. Niemuth JN, Harms CA, Macdonald JM, Stoskopf MK. NMR-based metabolomic profile of cold stun syndrome in loggerhead Caretta caretta, green Chelonia mydas and Kemp’s ridley Lepidochelys kempii sea turtles in North Carolina, USA. Wildlife Biology. 2020 Feb;2020(1).
- 57. Abraham JP, Baringer M, Bindoff NL, Boyer T, Cheng LJ, Church JA, et al. A review of global ocean temperature observations: Implications for ocean heat content estimates and climate change. Rev Geophys. 2013; 51(3): 450–483.
- 58. Cavicchioli R, Ripple WJ, Timmis KN, Azam F, Bakken LR, Baylis M, et al. Scientists’ warning to humanity: microorganisms and climate change. Nat Rev Microbiol. 2019; 17(9): 569–586. pmid:31213707
- 59. Guitart R, Silvestre AM, Guerrero X, Mateo R. Comparative study on the fatty acid composition of two marine vertebrates: striped dolphins and loggerhead turtles. Comp Biochem Physiol Part B: Biochem Mol Biol. 1999; 124(4): 439–443. pmid:10665372
- 60. Blair TA, Cree A, Skeaff CM. Plasma fatty acids, triacylglycerol and cholesterol of the tuatara (Sphenodon punctatus punctatus) from islands differing in the presence of rats and the abundance of seabirds. J Zool. 2000; 252(4): 463–472.
- 61. Cartland-Shaw LK, Cree A, Skeaff CM, Grimmond NM. Differences in dietary and plasma fatty acids between wild and captive populations of a rare reptile, the tuatara (Sphenodon punctatus). J Comp Physiol B. 1998; 168(8): 569–580.
- 62. Geiser F, Learmonth RP. Dietary fats, selected body temperature and tissue fatty acid composition of agamid lizards (Amphibolurus nuchalis). J Comp Physiol B. 1994; 164(1): 55–61.
- 63. Lance VA, Morici LA, Elsey RM, Lund ED, Place AR. Hyperlipidemia and reproductive failure in captive-reared alligators: vitamin E, vitamin A, plasma lipids, fatty acids, and steroid hormones. Comp Biochem Physiol Part B: Biochem Mol Biol. 2001; 128(2): 285–294.
- 64. Staton MA, Edwards HM Jr, Brisbin IL Jr, Joanen T, McNease L. Essential fatty acid nutrition of the American alligator (Alligator mississippiensis). J Nutr. 1990; 120(7): 674–685. pmid:2114472
- 65. Jayson S, Ferguson A, Goetz M, Routh A, Tapley B, Harding L, et al. Comparison of the nutritional content of the captive and wild diets of the critically endangered mountain chicken frog (Leptodactylus fallax) to improve its captive husbandry. Zoo Biol. 2018; 37(5): 332–346. pmid:30221785
- 66. Watanabe T. Lipid nutrition in fish. Comp Biochem Physiol Part B: Comp Biochem. 1982; 73(1): 3–15.
- 67. González-Medina E, Castillo-Guerrero JA, Herzka SZ, Fernández G. High quality diet improves lipid metabolic profile and breeding performance in the blue-footed booby, a long-lived seabird. PloS one. 2018; 13(2): e0193136. pmid:29462199
- 68. Hill MS. Spongivory on Caribbean reefs releases corals from competition with sponges. Oecologia. 1998; 117(1–2): 143–150. pmid:28308480
- 69. Stringell TB, Clerveaux WV, Godley BJ, Kent FE, Lewis ED, Marsh JE, et al. Taxonomic distinctness in the diet of two sympatric marine turtle species. Mar Ecol. 2016; 37(5): 1036–1049.
- 70. Sampson L, Giraldo A, Payán LF, Amorocho DF, Ramos MA, Seminoff JA. Trophic ecology of green turtle Chelonia mydas juveniles in the Colombian Pacific. J Mar Biol Assoc UK. 2018; 98(7): 1817–1829.
- 71. Speake BK, Surai PF, Bortolotti GR. Fatty acid profiles of yolk lipids of five species of wild ducks (Anatidae) differing in dietary preference. J Zool. 2002; 257(4): 533–538.
- 72. Speake BK, Herbert JF, Thompson MB. Comparison of the fatty-acid compositions of prey items and yolks of Australian insectivorous scincid lizards. J Comp Physiol B. 2004; 174(5): 393–397. pmid:15085383
- 73. Murzina SA, Nefedova ZA, Falk-Petersen S, Ripatti PO, Ruokolainen TR, Pekkoeva SN, et al. Lipid status of the two high latitude fish species, Leptoclinus maculatus and Lumpenus fabricii. Int J Mol Sci. 2013; 14(4): 7048–7060. pmid:23535338
- 74. Stacy NI, Bjorndal KA, Perrault JR, Martins HR, Bolten AB. Blood analytes of oceanic-juvenile loggerhead sea turtles (Caretta caretta) from Azorean waters: reference intervals, size-relevant correlations and comparisons to neritic loggerheads from western Atlantic coastal waters. Conservation physiology. 2018;6(1):coy006. pmid:29479433
- 75. Casal AB, Camacho M, López‐Jurado LF, Juste C, Orós J. Comparative study of hematologic and plasma biochemical variables in Eastern Atlantic juvenile and adult nesting loggerhead sea turtles (Caretta caretta). Veterinary Clinical Pathology. 2009 Jun;38(2):213–8. pmid:19192261
- 76. Harris HS, Benson SR, Gilardi KV, Poppenga RH, Work TM, Dutton PH, et al. Comparative health assessment of western Pacific leatherback turtles (Dermochelys coriacea) foraging off the coast of California, 2005–2007. Journal of Wildlife Diseases. 2011 Apr;47(2):321–37. pmid:21441185
- 77. Deem SL, Norton TM, Mitchell M, Segars AL, Alleman AR, Cray C, et al. Comparison of blood values in foraging, nesting, and stranded loggerhead turtles (Caretta caretta) along the coast of Georgia, USA. Journal of Wildlife Diseases. 2009 Jan;45(1):41–56. pmid:19204334
- 78. Price ER, Sirsat TS, Sirsat SK, Kang G, Keereetaweep J, Aziz M, et al. Thermal acclimation in American alligators: effects of temperature regime on growth rate, mitochondrial function, and membrane composition. J Therm Biol. 2017; 68: 45–54. pmid:28689720
- 79. Mineo PM, Waldrup C, Berner NJ, Schaeffer PJ. Differential plasticity of membrane fatty acids in northern and southern populations of the eastern newt (Notophthalmus viridescens). J Comp Physiol B. 2019; 189(2): 249–260. pmid:30673816
- 80. Coles W, Musick JA. Satellite sea surface temperature analysis and correlation with sea turtle distribution off North Carolina. Copeia. 2000; 2000(2): 551–554.
- 81. Crear DP, Lawson DD, Seminoff JA, Eguchi T, LeRoux RA, Lowe CG. Seasonal shifts in the movement and distribution of green sea turtles Chelonia mydas in response to anthropogenically altered water temperatures. Mar Ecol Prog Ser. 2016; 548: 219–232.
- 82. Zepeda-Borja KM, Ortega-Ortiz CD, Torres-Orozco E, Olivos-Ortiz A. Spatial and temporal distribution of sea turtles related to sea surface temperature and chlorophyll-a in Mexican Central Pacific waters. Rev Biol Mar Oceanogr. 2017; 52(2): 375–385.
- 83. Hays GC, Dray M, Quaife T, Smyth TJ, Mironnet NC, Luschi P, et al. Movements of migrating green turtles in relation to AVHRR derived sea surface temperature. Int J Remote Sens. 2001; 22(8): 1403–1411.
- 84. McCue MD. Fatty acid analyses may provide insight into the progression of starvation among squamate reptiles. Comp Biochem Physiol Part A: Mol Integr Physiol. 2008; 151(2): 239–246. pmid:18657629
- 85. Williams TD, Guglielmo CG, Egeler O, Martyniuk CJ. Plasma lipid metabolites provide information on mass change over several days in captive Western Sandpipers. The Auk. 1999; 116(4): 994–1000.
- 86. Tang S, Fan L, Cheng H, Yan X. Incorporating Electro-Epoxidation into Electrospray Ionization Mass Spectrometry for Simultaneous Analysis of Negatively and Positively Charged Unsaturated Glycerophospholipids. Journal of the American Society for Mass Spectrometry. 2020 Nov 24. pmid:33232136
- 87. Tang S, Cheng H, Yan X. On‐Demand Electrochemical Epoxidation in Nano‐Electrospray Ionization Mass Spectrometry to Locate Carbon–Carbon Double Bonds. Angewandte Chemie International Edition. 2020 Jan 2;59(1):209–14. pmid:31639243
- 88. Bressan MJ, de Lima TG, de Melo LF, Rigoglio NN, Lopes EQ. Evaluation of the green turtles’(Chelonia mydas) body conditions found in the Juréia-Itatins Mosaic’s conservation units, South Coast of the São Paulo state. Brazilian Journal of Animal and Environmental Research. 2020 Aug 26;3(3):1871–84.
- 89. Bressan MJ, de Lima TG, de Melo LF, Rigoglio NN, Lopes EQ. Green turtle (chelonia mydas) feeding behavior characterization from dead specimens on Peruíbe beach mosaic Jureia-Itatins, South coast of São Paulo State, Southeast Brazil. Brazilian Journal of Animal and Environmental Research. 2020 Aug 17;3(3):1855–70.
- 90. Osthoff G, Hugo A, Govender D, Huchzermeyer F, Bouwman H. Comparison of the lipid composition of three adipose tissue types of male and female wild Nile crocodiles (Crocodylus niloticus). J Herpetol. 2014; 48(4): 525–531.
- 91. Daniels CB, Barr HA, Power JH, Nicholas TE. Body temperature alters the lipid composition of pulmonary surfactant in the lizard Ctenophorus nuchalis. Exp Lung Res. 1990; 16(5): 435–449. pmid:2226354
- 92. Stacy NI, Chabot RM, Innis CJ, Cray C, Fraser KM, Rigano KS, et al. Plasma chemistry in nesting leatherback sea turtles (Dermochelys coriacea) from Florida: understanding the importance of sample hemolysis effects on blood analytes. PloS one. 2019 Sep 10;14(9):e0222426. pmid:31504062
- 93. Dimeski G, Mollee P, Carter A. Increased lipid concentration is associated with increased hemolysis. Clinical chemistry. 2005 Dec 1;51(12):2425. pmid:16306120
- 94. Harrison XA, Blount JD, Inger R, Norris DR, Bearhop S. Carry‐over effects as drivers of fitness differences in animals. J Anim Ecol. 2011; 80(1): 4–18. pmid:20726924
- 95. Solow AR, Bjorndal KA, Bolten AB. Annual variation in nesting numbers of marine turtles: the effect of sea surface temperature on re‐migration intervals. Ecol Lett. 2002; 5(6): 742–746.