STUDYING TROPHIC ECOLOGY IN MARINE ECOSYSTEMS USING FATTY ACIDS: A PRIMER ON ANALYSIS AND INTERPRETATION

Fatty acids (FA) represent a large group of molecules that comprise the majority of lipids found in all organisms. Their great diversity, biochemical restrictions and, in some cases, unique origin among plants and animals has fostered a number of areas of research, ranging from assessment of animal nutrition and metabolism, to investigating trophic interactions and ecosystem structure. Over the past three decades, we have observed the use of FA develop from a potential tool for delineating food webs (Ackman and Eaton 1966) to a powerful technique for quantitative assessment of predator diets (Iverson et al. 2004). Studies that have compared the FA found in predator fat stores with those found in their prey have allowed both qualitative (e.g., Horgan and Barrett 1985; Smith et al. 1996; Raclot et al. 1998; Dahl et al. 2000, 2003; Falk-Petersen et al. 2004) and quantitative or semi-quantitative analyses of diet (Kirsch et al. 2000; Iverson et al. 2001b, 2004; Iverson and Springer 2002) and comparisons of the spatial and temporal scales of foraging (e.g., Iverson et al. 1997a, b; Smith et al. 1997; Walton et al. 2000). These studies are possible because the FA consumed by a monogastric (i.e., single-chambered stomach) predator are deposited into adipose tissue with little modification or in a predictable pattern, thus providing an integrated record of dietary intake over time. Previous studies, together with the recent development of quantitative FA signature analysis (QFASA; Iverson et al. 2004), have generated considerable interest in the use of FA to study the foraging ecology and diets of upper trophic level predators, such as marine mammals, bears and seabirds. Given that the use of FA as an ecological tool is a relatively new and rapidly developing field, it is inevitable that a variety of methods have been used both to chemically analyze FA and to interpret the resulting data. In both of these areas, we believe there is some confusion in the literature. Furthermore, many investigators from around the world frequently inquire about the requirements and limitations of using FA to address ecological questions. Given this broad interest, particularly among researchers studying marine mammals, we felt it was both necessary and timely to review appropriate methods for the chemical analysis and interpretation of FA in the context of ecological research. Although we focus on applications involving marine mammals and seabirds, most of the issues we discuss are pertinent to other marine and terrestrial predators and the methods should be broadly applicable to other taxa. FA can be used to study foraging ecology and food webs in three ways. First, by examining changes in FA distributions, or "signatures" (Iverson 1993) of the predator alone, we can ask qualitative questions about spatial or temporal variations in diets, both among and within individuals or populations (e.g., Iverson et al. 1997a, b; Walton et al. 2000; Beck et al. 2005). The second use of FA is also qualitative, but relies on the use of individual biomarkers or tracers—that is, the existence of a unique FA found in a predator that can be traced to a single origin or prey species (e.g., Pascal and Ackman 1975, reviewed in Iverson 1993). In principle, this opportunity will be relatively rare, as FA present in either marine or terrestrial environments are relatively ubiquitous within those environments. However, this biomarker approach can be extended; for instance, unusual levels of certain FA or of ratios among FA that can only be attributed to one or a few prey types may indicate their likely unimportance or dominance in the diet (reviewed in Dalsgaard et al. 2003). The final and most ambitious way in which to use FA is to quantitatively estimate diet from FA signatures of predator and prey. The QFASA method (Iverson et al. 2004) uses a statistical model to compute the most likely combination of prey FA signatures that comes closest to matching that observed in the predator, after accounting for predator FA metabolism by mathematically weighting individual FA. This procedure requires knowledge of the FA compositions of all important potential prey species, as well as sufficient within-species sampling to assess variability or overlap in signatures with ecological and demographic factors (e.g., Budge et al. 2002, Iverson et al. 2002). Each of these uses of FA can provide somewhat different, but valuable information on the foraging ecology of free-ranging predators that are difficult to observe or access over much of their spatial and temporal ranges, and for which existing methods, such as fecal or stomach contents analysis, may provide biased information. Nevertheless, all of the methods using FA require knowledge of the underlying biochemistry and the appropriate techniques to ensure both accurate chemical analysis of FA composition and appropriate interpretation of the resulting data. The aim of this review is to provide investigators with a guide that describes the suite of requirements and tools necessary to conduct ecological studies using FA. We concentrate on issues related to the use of FA for studying trophic interactions or diet in upper trophic level marine predators, although we also touch upon other ways in which to use FA in ecological studies. We begin by providing a description of basic lipid structures and their biochemistry as these relate to diet and biosynthesis. We then discuss appropriate sampling, storage and chemical analysis, and highlight common and potential problems. We also describe the basic gas chromatographic (GC) instrumentation and software necessary for analysis, and briefly discuss some statistical methods currently used to analyze and interpret the resulting data. Lastly, we discuss taxon-specific issues in the use of FA. Lipids are a heterogeneous group of compounds that share the common property of being insoluble in water but soluble in organic solvents such as chloroform, hydrocarbons and alcohols. The most abundant lipids are those that contain FA as part of their structure. FA most commonly consist of an even-numbered, straight carbon chain, containing 14 to 24 carbons and zero to six double bonds, with a methyl terminus (CH3) at one end and an acid (carboxyl) terminus at the other (Fig. 1). Nevertheless, FA in certain food sources can be as short as 4–6 carbons (e.g., milks of ruminant mammals) and 8–10 carbons (e.g., milks of rodents, lagomorphs and primates; some seed lipids) (Iverson and Oftedal 1992). However, in marine and terrestrial food webs involving upper trophic level carnivores, the shortest FA that is generally present above trace amounts is 12 carbons in length, but this still usually accounts for ≤0.1% of total FA. Extremely short branched-chain FA are found in structural and acoustic tissues of odontocetes, but these are not related to diet (see below). At the other extreme, FA with >24 carbon atoms and >6 double bonds also exist, but only in trace amounts as intermediates in most organisms (Voss et al. 1991). Thus, FA of 14 carbons with no double bonds to 24 carbons with 6 double bonds represent the range in length and unsaturation that is relevant in the study of most predators. Example of FA structure and nomenclature. (A) The structure of 18:3n-3 with all atoms indicated; and (B) the same FA structure with carbon and hydrogen atoms assumed. FA are named using shorthand notation of A:Bn-X, where A indicates the number of carbon atoms, B is the number of double bonds and X indicates the position of the first double bond relative to the terminal methyl group (i.e., 18:3n-3). With this system, it is assumed that each double bond is separated by a methylene group (CH2 group). Some investigators use the slightly shorter notation of A:BωX (i.e., 18:3ω3), which has the same meaning but is simply an older format. This notation has developed because of metabolic relationships; once introduced into a FA, organisms have very limited ability to alter the position of the double bond closest to the terminal methyl group. A less common notation, that follows the recommendations of the International Union of Pure and Applied Chemistry (IUPAC), gives the position of double bonds relative to the acid group (COOH) and uses the Δ symbol (i.e., 18:3Δ9,12,15). This system has the advantage of using the same naming conventions as FA desaturase enzymes; however, the n-x system is more useful in tracking families of FA with similar structures. FA are most commonly named using a shorthand notation of A:Bn-x, where A indicates the number of carbon atoms, B is the number of double bonds and x indicates the position of the first double bond relative to the terminal methyl group (Fig. 1A). With this naming system, it is assumed that all FA are methylene-interrupted (i.e., each double bond is separated by a -CH2- group), which is by far their most common form. However, unusual FA do occur with double bonds that are non-methylene interrupted (NMI). Their presence in the food chain is due to synthesis by marine invertebrates, such as bivalve molluscs (reviewed in Joseph 1982). FA with zero and one double bond are known as saturated (SFA) and monounsaturated FA (MUFA), respectively. FA with two or more double bonds are known as polyunsaturated FA (PUFA). Lastly, some FA contain a methyl branch on the second or third carbon atom closest to the terminal methyl group. A methyl branch at the second carbon is indicated by prefacing the FA name with an "i" representing "iso"; "ai" (anti-iso) indicates a methyl branch at the third carbon (e.g., i-15:0, ai-17:0). FA rarely exist in free form and are usually esterified to a glycerol backbone. Any lipid containing an esterified FA is called an acyl lipid. Triacylglycerols (TAG) represent the common form of storage lipids and make up the majority of lipids found in adipose tissue and blubber. TAG consist of three FA molecules esterified to a glycerol backbone (Fig. 2A). Animals mobilize stored TAG if FA requirements are not met through the diet or, conversely, deposit TAG when dietary FA and energy intake exceed demands. Thus, the composition of TAG FA in major lipid storage depots is relatively dynamic and most readily influenced by diet. Lipid structures and FA derivatives produced using acidic transesterification. (A) A TAG consists of three FA moieties and a glycerol backbone. The reaction of TAG with MeOH in the presence of acid cleaves the FA from the backbone and forms the corresponding FAME; (B) WE, consisting of a FA esterified to a fatty alcohol, react with MeOH and acid in a similar manner to produce a FA and a fatty alcohol; and (C) A PL is similar to a TAG but contains a phosphate group attached to the third carbon of the glycerol backbone. The phosphate group may be bonded to a variety of derivatives; in this case, an amine group, choline, is attached. The structure illustrated is also an example of a specific type of phospholipid, the plasmalogen. Plasmalogens have an ether linkage to a long carbon chain containing a double bond adjacent to the ether functionality; this ether structure takes the place of one of the two FA and reacts with MeOH and acid to form an aldehyde that reacts further to produce DMA. Wax esters (WE) are a second type of lipid relevant to diet studies. WE consist of a FA esterified to a fatty alcohol (Fig. 2B). In theory, both the FA and the fatty alcohol can have any structure, but WE almost always contain saturated or monounsaturated fatty alcohols, unlike the FA portion which may have a variety of polyunsaturated structures. The functions of these WE are uncertain, but, like TAG, they are thought to be involved in energy storage in certain species of crustaceans, fish, and marine mammals. For example, copepods (e.g., Calanus spp.) store almost all their energy as WE (Sargent 1976), and some fish such as myctophids, barracudina (Notolepis rissoi), capelin (Mallotus villosus), and herring (Clupea harengus) store significant amounts of WE in their adipose tissues (Ackman et al. 1972; Ratnayake and Ackman 1979; Phleger et al. 1997, 1999). WE are also found in the melons and mandibular regions of echolocating toothed whales (Varanasi and Malins 1971, Litchfield et al. 1978) and comprise the majority of storage lipids in the blubber of Pacific beaked whales (Berardius bairdi) and sperm whales (Physeter macrocephalus; Litchfield et al. 1976, Koopman 2001). A third common acyl lipid class is the phospholipid (PL) group. PL normally consists of two FA esterified to a glycerol molecule that also contains a polar derivative of phosphatidic acid (Fig. 2C). PL are the structural components of all cell membranes. Because of the specialized functions of these lipids and membranes, organisms tend to conserve FA in PL, making this lipid class relatively robust to dietary changes in higher order predators and thus not informative as an indicator of diet. A special group of PL, the plasmalogens, contains a mono- or di-unsaturated carbon chain attached through a vinyl-ether linkage to the glycerol molecule in place of one of the FA (Fig. 2C). Vinyl-ether linkages are distinct from other ether linkages because of the presence of a double bond adjacent to the ether group. These lipids are thought to be involved in a variety of processes ranging from antioxidant functions to buoyancy control (Paltauf 1994) and are found throughout the marine food web. In particular, plasmalogens are common in bivalves (Joseph 1982) and in the gills of crustaceans (Chapelle and Benson 1986); smaller amounts are also found in many species of fish (Ackman 1998). Other classes of acyl lipids exist, including glycolipids, sphingolipids and ceramides, but these are relatively rare and contribute little to the overall FA composition of organisms. Harwood and Russel (1984) and Gunstone et al. (1986) provide comprehensive reviews of all lipid classes. A basic understanding of lipid biochemistry and metabolism is critical to the interpretation of FA patterns in predator lipid stores. The FA composition of predator lipid stores is the result of three metabolic sources: (1) unmodified dietary FA that are directly deposited in adipose tissue, (2) dietary FA that are modified at some point between absorption into the blood and deposition, and (3) endogenously derived FA arising from de novo synthesis in the predator. Thus, it is important to understand the potential relative input from each process in any given tissue. In monogastric predators such as marine mammals and seabirds, by far the greatest contributor to adipose tissue FA composition is direct deposition of FA from diet, so that the relative influence of diet on a predator's fat stores may often be readily apparent. However, there will also be contributions from the other two processes. Thus, predator FA will never exactly match those of its prey. An understanding of when and where these differences in FA arise allows the investigator to select those FA that are most relevant to the question being posed. Lipid digestion and products— As noted earlier, we confine our discussion of lipid metabolism in higher order vertebrate predators to monogastric (non-ruminant) animals because, in ruminants, the extensive foregut fermentation generally results in the complete degradation of ingested FA. We also address only specific issues needed to understand how predator FA relate to diet, as there are several excellent reviews of general lipid and FA metabolism (e.g., Harwood and Russell, 1984, Cook 1991, Gurr and Harwood 1991). In the simplest and most common scenario, a prey item contains only acyl lipids, such as TAG and PL. During digestion by the predator, ingested lipids are hydrolyzed (i.e., ester bonds are broken) to yield free FA (FFA), monoacylglycerols, and, in the case of PL, lysophospholipids (a glycerol backbone with a phosphatic acid derivative and only one FA). These products pass through the mucosal wall of the small intestine, reform into acyl lipids (i.e., TAG) and are transported by chylomicrons (i.e., the lipoproteins which carry FA derived from diet, consisting primarily of TAG) in the blood to tissues. However, short- and medium-chain FA < 14 C are an exception, as they are not incorporated into chylomicrons, but are transported to the liver where they are immediately oxidized (Brindley 1991, Papamandjaris et al. 1998). Thus, FA < 14 C that are present in predator lipids can only have arisen from de novo biosynthesis within that predator and, therefore, have no relation to diet. Once circulating chylomicron-TAG arrive at the tissue, they are again hydrolyzed. FA are then either metabolized for energy or re-esterified into TAG for storage. Some FA are also used in structural tissues or can be transformed into hormones and secondary messengers. Although all tissues contain FA in membrane structures or in small lipid depots, it is adipose tissue (or blubber in the case of marine mammals) that is the principle FA storage site and it is composed primarily of TAG. Adipose tissue TAG are also the principle form of energy storage in animals, as animals have only a very limited capacity to store carbohydrate. All FA consumed in excess of immediate requirements are deposited in this reservoir. As noted above, other than FA < 14 C, most FA are deposited in the TAG reservoir of adipose tissue in similar proportions to that consumed in the diet. However, there are notable exceptions (Iverson et al. 2004) that are relatively predictable given our current knowledge of FA metabolism, as discussed below. In the case of TAG digestion discussed above, esterified FA are consumed and esterified FA are A slightly more process with the of prey that contain such as the fatty in an ingested WE is it is first hydrolyzed to produce its FA and fatty in the then to the fatty alcohol to its corresponding FA (i.e., an number of C atoms with the same double bond This FA then the total ingested of FA for into TAG, and is transported in chylomicrons and Thus, if a predator stores its FA as TAG in adipose tissue, it will do so if it a diet in In in predators that have to use WE as an energy storage the process FA consumed in the diet are to the corresponding fatty alcohol, which is then incorporated into WE for storage. is not known where this but given the of WE in circulating it likely that it takes place at the site of WE in storage tissue. is from and that of FA to fatty is not The for the of FA to fatty is specific for are rarely present as and Thus, it is important to two First, the and storage of WE are in a predator WE or TAG has to do with it stores FA as WE or WE storage to be a (Sargent et al. FA stored in a predator that WE will not be of all FA ingested because some portion of the fatty alcohol of the storage WE has been derived from the of ingested FA. Thus, both FA and fatty alcohol components be to the composition of FA of WE is more to than that of TAG but, accurate of ingested and deposited WE and all other acyl lipids, we can the alcohol and FA composition of the diet as a single FA that the upon to its adipose tissue. is, a TAG or a WE is by the predator, the same is FA either ingested FA or oxidized fatty is the fatty chain that is pertinent to diet and, because fatty and FA are the of each is The vinyl-ether carbon in plasmalogens are also modified digestion to yield FA. In the intestine, of the vinyl-ether linkage a fatty aldehyde (Fig. that is oxidized to the corresponding FA. The with plasmalogens is further because, with methods of analysis, the vinyl-ether linkage is transformed into a that with the of some common FA. Because plasmalogens and other FA such as WE, to the predator's FA their levels in the prey be Studies that to their to analysis by all compounds that are not FA are often containing these important FA FA modification and With an understanding of the relatively direct of FA from to deposition in adipose tissue, the to in FA to diet is the to which FA may be modified this or de by the predator. modification within the carbon chain of FA may occur between and and mammals are of the carbon chain of dietary FA by two carbon and double bonds 1991). However, these are to and and are both and the of diets as of total from particularly those in fat contents of prey it is likely that most and a diet in the range of fat or containing levels of which to of For instance, an at fat and contain about of as Thus, FA that have been and are to make a significant to the adipose FA in predators such A more significant modification process may be the chain of some In both marine mammals and seabirds, deposition, relative to diet, of most of and are observed (Iverson and Springer 2002, Iverson et al. 2004). on the has that this is as in other to chain of those et al. The products of this chain are and et al. with deposition of these is in marine and fat studies on have that dietary is directly deposited modification and to arise from de novo synthesis et al. 2004). Thus, the levels of relative to diet likely arise from chain of and However, the predictable between the and allows to for their levels in fat depots of marine predators (Iverson et al. et al. novo synthesis of FA (i.e., biosynthesis of FA from or may also occur within predators, but, like and it is limited to and and or by animals do not or but mobilize it and it for when a predator is a diet containing or excess including all the FA (i.e., that are in marine there is no to In and mammals, de novo synthesis only when the animal is a as about of total and diet. In the the of de novo synthesis is this with from terrestrial and so we can that the same is likely in other marine predators. Thus, de novo synthesis in the predator is to be for proportions of any FA other than In in is often present at amounts than that found in diet (Iverson et al. 2004), again that de novo synthesis is to be an important of FA in animals a diet. of some individual FA has been in and Iverson et al. The of was among the predators in that was was Raclot has that is by of the FA. However, this as of other polar FA, such as and is not from studies in both and have no in overall FA composition with Iverson and further studies on will be the of on FA composition are to have significant on diet branched-chain and straight chain FA are present in trace amounts in the adipose tissue of vertebrate predators. of particularly those of carbons in length, are derived from (Sargent et al. Ackman and normally in prey and predator adipose tissue at levels of total FA (Iverson et al. Raclot et al. Budge et al. 2002, Iverson et al. 2002). Their levels in prey will on of and the of present in the is known about the metabolism of these compounds once they are but they are like FA and are deposited in the adipose tissue as however, these FA are not particularly of dietary FA intake because they may be in both the predator and and are generally only present in trace The adipose tissues of a few marine predator contain amounts of short branched-chain FA (e.g., and For example, certain families of and contain considerable levels of the chain FA In the blubber and of some species (e.g., this is the FA, accounting for up to of total FA et al. 2003). These short branched-chain FA are not derived from but are de novo from acid and 1982). In the they are to a in tissue in water environments et al. and in to in and (Varanasi and Malins 1971, Koopman et al. This de novo synthesis most likely takes place in the tissues in which these FA are as any significant found in the likely be et al. 2003). This de novo synthesis like the branched-chain FA, these short branched-chain FA are not of diet. as discussed above, if consumed in diet they be oxidized after Nevertheless, the of these molecules in the tissues of some of animals requires methods of analysis (see chain below). A number of including adipose tissue, and may contain information about diet, with the tissue type on the focus of the and the limitations of sampling a group of These tissues provide into diet over temporal Adipose tissue is composed of fat and may be metabolic (i.e., site of energy or structural (e.g., in and in the of the of in 1998). adipose tissue usually has a specific composition of lipids and other or for its and relatively little structural adipose tissue will little to dietary lipid only adipose tissue that as a fat energy such as or is useful in diet is a specialized type of adipose tissue found in marine mammals, of within an of 1998). has a number of including as a to the of and as an energy storage et al. Iverson 2002). Thus, blubber can have a structure on it is in an that primarily structural

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.