Stable isotope analysis (SIA) is considered a helpful tool in a range of researches such as the study of climatic condition (Barnet et al. 2019), agricultural products, biological processes and biogeochemical cycles (Chaffee et al. 2007). In ecological studies, the measurement of stable isotopes in plants and animals is applied to the determination of animal feeding behavior, movement, and trophic position along food chains (Bouillon et al. 2011; Ben-David and Flaherty, 2012). Stable isotopes are safe (non-radioactive) and can be operated by humans. Even infants and pregnant women can be safely examined in medicine and nutrition studies. Naturally occurring stable isotopes are transferred from the physical environment to primary producers, as well as from a resource to its consumer, and emerge in hair, urine, feces, breath, and blood (Hagen, 1963; Rossi et al. 2018). Thus, they can be used to trace nutrient uptake in producers and consumers in both terrestrial and marine ecosystems (Madeira et al. 2019; Signa et al. 2019; Calizza et al. 2018). Many measurement techniques depend on natural differences in the way ‘light’ and ‘heavy’ isotopes react during metabolic processes through biological and chemical alterations. Other stable isotope techniques depend on adding trace amounts of compounds artificially enriched in the rare (heavy) isotope of the element of interest. These are called isotope tracer methods / techniques. About a century ago, Fredrick Soddy was the first to identify signs of the existence of isotopes (Wilkinson, 2018). Isotopes are classified into ‘Stable’ and ‘Unstable’ groups. The unstable isotopes, which are not the subject of this research, are radioactive. Here the question is ‘What are the stable isotopes?’. To answer this question, we should start by focusing on the atomic nucleus. Indeed, a different number of neutrons within the nucleus of the heavy and light stable isotopes leads to different atomic masses (Ben-David and Flaherty, 2012). Both light and heavy stable isotopes play a similar role in biological and chemical reactions, but with different response rates. The attractive forces and chemical bonds of the light stable isotope are weaker than the heavier isotope of an element. Thus, the lighter isotope reacts more quickly than the heavier one in both biological and chemical reactions. Even though oxygen (O), sulfur (S), and deuterium (D) are applied in some studies, nitrogen (N) and carbon (C) are the two main elements considered in the study of animal diet and food webs.
Nitrogen (14N, 15N)
Natural nitrogen includes two stable isotopes (14N, 15N). 14N is the most common isotope, while 15N is the rarest. Different nitrogen isotopes (14N and 15N) can be distinguished through thermal diffusion or chemical exchanges. Other isotopes of nitrogen can be found in nature, such as 12N, 13N, 16N, and 17N. However, these isotopes are radioactive. Living organisms through the ‘nitrogen cycle’ usually transform nitrogen. Microbes convert different nitrogen compounds (like ammonia, NH3+) to nitrates for green plants and algae (Finlay and Kendall, 2008). Animals get their required nitrogen by consuming other living organisms (Post, 2002). The measurement of the isotopic signature of nitrogen (δ15N) plays an important role in biochemical, industrial and ecological applications such as food preservation, quantification of ecological processes and feeding interactions among organisms, medical and biomedical research (Schellekens et al. 2011; Calizza et al. 2018; Signa et al. 2019), and climate studies (Dotsika and Diamantopoulos, 2019).
Carbon (12C, 13C, 14C)
One of the essential elements on earth is carbon, which forms the chemical basis of life. There are three natural isotopes of carbon, with atomic masses of 12, 13, and 14. 12C and 13C are stable and are used as tracers to understand nutrient cycling (Wang et al. 2019), food webs (Telsnig et al. 2019), and air-sea swapping of CO2 (Lynch-Stieglitz et al. 1995). Plants and phytoplankton have a preferential use of12C to convert sunlight and carbon dioxide into biomass. The ocean surface is separated from the deeper water. However, when plankton dies, it sinks and removes 12C from the surface (Flannery, 2006). The 14C, or radiocarbon, is unstable. It is produced in the atmosphere and absorbed by living organisms (Marra, 2019). Carbon signatures can be used in agricultural and climate studies, authentication of foodstuff, description of nutrient fluxes in ecosystems, and in the determination of the age of archaeological specimens (Zeuner, 1958; Aitken, 2013; Signa et al. 2019).
Hydrogen (1H, 2H, 3H, 4H, 5H, 6H, 7H)
Hydrogen has two naturally stable isotopes, 1H and 2H. The 2H isotope is called deuterium (D), while 3H is known as tritium (T), which is radioactive. Four other hydrogen isotopes, 4H, 5H, 6H, and 7H, are highly unstable and have been synthesized in the laboratory by bombarding tritium and by fast-moving deuterium or tritium nuclei (Golovkov et al. 2003). Some applications of the hydrogen isotopes could be highlighted in the authentication of foodstuff, agricultural, ecological, geochemical studies, and medical applications (Finlay and Kendall, 2008; Boschetti et al. 2019).
Oxygen (16O, 17O, 18O)
Oxygen isotopes include three stable forms. The most abundant is 16O, while 17O and 18O are categorized as secondary stable isotopes. The 16O is mostly produced by massive stars composed only of hydrogen. 17O and 18O nucleosynthesis needs seed nuclei. The 17O is produced by hydrogen burning into helium in CNO (Carbon-Nitrogen-Oxygen) cycle, and 18O is made when the14N catches the4He nucleus (Meyer, 2005; Emsley, 2011). Oxygen isotopes can be used in the authentication of foodstuff, agricultural, ecological, geochemical, climate, and medical studies (Finlay and Kendall, 2008; Boschetti et al. 2019; Duffy et al. 2019).
Sulfur (32S, 33S, 34S, 36S)
Sulfur has twenty-four isotopes. Among these, 32S, 33S, 34S, and 36S are stable. Understanding acidic deposition in the forest ecosystems is the major application of sulfur isotopes (Campbell et al. 2006). The 34S values increase with pollution sources and gas emission, which makes sulfur a powerful detector (Mayer et al. 1993). The sulfur input to marine systems mainly arises from seawater sulfate (δ34S=21‰), whereas terrestrial inputs mainly depend on precipitations (δ34S=2-8‰) (Michener and Kaufman, 2008).
The abundance of stable isotopes in mineral and biological samples is measured as the heavy-to-light isotope ratio (R). For a given element (X), the isotopic signature of a sample (δ) is expressed as the per mill deviation (‰) from an international standard (Muccio and Jackson, 2009; Philp, 2015), according to the following equation:
δX= (RSample/RStandard-1) × 1000 (1)
The International Atomic Energy Agency (2004) and the National Institute of Standards and Technology (Muccio and Jackson, 2009) provided an accurate evaluation of the reference standard elements (Lynch-Stieglitz et al. 1995; Werner and Brand, 2001; Flannery, 2006; Brand et al. 2014). The ratios of 13C/12C, 15N/14N, 18O/16O, and 2H/1H have been used widely to measure stable isotopes in carbon, nitrogen, oxygen, and hydrogen (Muccio and Jackson, 2009). The analytical determination of δ values implies the use of few light gases such as CO2, CO, N2, O2, and SO2. Accordingly, this standard technology has been called Isotope Ratio Mass Spectrometry (IRMS) (Brand, 2004). This analytical method identifies the chemical substance by ionizing it, focusing the resulting ions into a beam, and by separating the light and heavy atoms according to their net electric charge (Finlay and Kendall, 2008). The classical method of analysis includes two gases that are stored in containers connected via capillaries to a switching unit, the changeover valve. An isotope ratio mass spectrometer uses one gas as its ion source, while other available gas flow to the waste vacuum line (Werner and Brand, 2001). Both gases are used and compared a few times and measured separately through the ion currents. The relative difference in the ratio of light and heavy ions is calculated according to an international relative isotope ratio scale (Paul et al. 2007). The instrument has six basic components (Figure 1) (Edmond de and Stroobant, 2013), which include: 1) a vacuum system; 2) an ion source; 3) a mechanism to concentrate ions into a narrower beam; 4) the speeding up of the beam; 5) a mass analyzer; 6) a detector. The material is initially present in the vacuum system, which produces the required low pressure to produce electrons and ions in the gas phase. Then, samples are transformed and concentrated into a narrower beam. (Brenna et al. 1997; Meier-Augenstein, 1999; Paul et al. 2007). Commonly, two connectors are used to introduce samples into isotope ratio mass spectrometry (IRMS): elemental analysers (EAIRMS) and gas chromatographers (GCIRMS). For many years, techniques such as gas chromatography (GC) and gas chromatography-mass spectrometer (GCMS) have been used to identify contamination sources (Philp, 2015). Similarly, the combined gas chromatography isotope ratio mass spectrometry (GCIRMS) technique can be used to determine individual compounds and soil contamination sources.
Before sampling, it is necessary to define what kind of information is needed. This information usually depends on research objectives and the type of samples to be collected. Otherwise, the risk is to waste time and resources in collecting either wrong or not enough data. The sampling design is a tool that is utilized to infer how many data to collect, where, when, and how often they should be collected.
Samples of vascular plants should be collected in the field and separately kept cold or frozen until processing. Samples of non-vascular plants are divided into two sections: lichens and marine algae. Lichens samples need to be collected directly into paper bags and dried once in the laboratory (Eldridge et al. 2003). If no oven is available, they can be spread out in a warm and well-ventilated place in packets and stored upright in a box. The samples of marine algae may be partially dried in the sun, but the small ones should be placed between sheets of paper, and the large ones should be placed in a box for further drying. Afterward, specimens should be pressed and stored in a dry and warm place (Steinitz and Kurle, 2014).
As regards terrestrial animals, it is possible to collect a sample of muscle, skin, feathers, eggshell, egg albumen, fur hairs, bones, etc. The samples may reflect diets ingested months before sampling, e.g. during the moulting or laying phase, and they can be collected from live or dead animals. Bone growth rings and whole bone can reflect an annual diet trend, during the whole animal’s lifetime. If bone samples are collected, soft tissues should be removed, and bones should be rinsed to remove impurities. When dry, bone samples can be placed in a paper bag. Feather samples should be cleaned to remove residual dirt and oil using a chloroform-methanol solution (Paritte and Kelly, 2009). Also, inorganic calcium carbonate from eggshell samples should be removed through a process of acidification (Finlay and Kendall, 2008). As regards aquatic animals, small invertebrates can be collected by using kick nets, grabs, and litterbags, while large predators such as crabs, sea snails, and stomatopods, can be gathered by using traps baited with fresh fish flesh (Careddu et al. 2015). For bigger organisms, the muscle tissue often provides enough biomass to perform stable isotope analyses (Abrantes et al. 2013), while for small invertebrates such as amphipods and polychaetes, the whole body can be used (Ng et al. 2007). In order to study a consumer’s diet, sampling should also include any dietary item that has likely been accessed by the consumer. All animal samples should be kept frozen at -20 ˚C and then lyophilized or dried at 60 ˚C overnight and kept in dry conditions until analysis (Finlay and Kendall, 2008).
Before isotopic analysis, all samples should be reduced to a fine homogeneous powder with a ball mill (Rossi et al. 2018). Then, powder of animal tissues (0.20±0.05 dry-mg) and vegetal tissues (3.0±0.05 dry-mg) should be weighed into tin capsules and analysed with an isotope ratio mass spectrometer. Thus, based on δ13C and δ15N values of consumers and their potential food sources, the animal diet can be determined through the R software (R Core Team, 2013), and Bayesian stable isotope mixing model (Rossi et al. 2018). Specifically, Bayesian mixing models allow the estimation of the proportion of each resource in the consumer’s diet. The model requires three inputs: the isotopic signatures of the target consumer, the isotopic signatures of potential food sources, and the trophic enrichment factor, which represents the expected isotopic increase from a resource to its consumer due to metabolic processes (McCutchan et al. 2003; Careddu et al. 2015).
Stable isotope applications in animal nutrition
As stated before, SIA is considered a helpful tool to be used in many disciplines. Among these, stable isotope-based environmental studies have recently flourished. Stable isotope signatures can be used to measure environmental stressors by monitoring plant uptake of carbon dioxide (Zheng et al. 2019) and greenhouse gas emissions (International Atomic Energy Agency, 2004; Popa et al. 2014) as well as by tracing the source of water in catchments (Philp, 2015; Fiorentino et al. 2017; Barbieri, 2019) and organic and mineral compounds during biogeochemical processes (Finlay and Kendall, 2008) and cycles (Lichtfouse, 2000). The study of past climatic conditions is essential (Barnet et al. 2019; Jafari and Jafari, 2019) because it enables to modelling climate variability and make predictions of future conditions (Noorollahi et al. 2011). When dealing with animal nutrition, climate variability, ecological transitions, temporal and spatial scales, and individual choices can all affect variation and adaptation in the diet of organisms across trophic levels (Careddu et al. 2015; Bentivoglio et al. 2016; Calizza et al. 2018; Jafariand Jafari, 2019). Therefore, the study of temporal and spatial patterns of animal foraging through stable isotope analysis can provide useful information to predict future variations in feeding preferences according to climate change scenarios (Finlay and Kendall, 2008; Calizza et al. 2018; Rossi et al. 2019). In this perspective, oxygen isotope values can be used to indicate hotter and drier climate (18O enrichment) versus colder and wetter conditions (18O depletion). As an example, results published by Noorollahi et al. (2011) showed that increasing temperature has a positive correlation with rising δ18O values. Also, enrichment in15N has been reported as an indication of arid conditions (Pate and Anson, 2007). The δ18O and δ2H values have been shown to vary across geographic regions (Bowen and Revenaugh, 2003) or along environmental gradients (Lee et al. 2019), being thus useful to infer the geographic origin of samples.
Figure 2 Differences in δ13C (‰) and δ15N (‰) between food webs based on C3 and C4 plants (Schulting, 1998)
Figure 3 Differences in δ13C (‰) and δ15N (‰) values in Marine ecosystem (Schulting, 1998)
SIA has proven to be a beneficial tool for: (1) studying nutrient uptake by humans, nutrient body reserve and nutrient metabolism paths (Schoeller, 2002), (2) describing contaminant flows (Signa et al. 2019), trophic relationships and food web structures (Careddu et al. 2015; Rossi et al. 2015; Calizza et al. 2018; Signa et al. 2019), as well as nutrient status (Calizza et al. 2016), (3) examining animal movement and migration (Di Lascio et al. 2016; Cicala et al. 2019; Madeira et al. 2019), (4) clarifying patterns of resourceallocation (Stachowicz et al. 2007; Di Lascio et al. 2013), (5) identifying primary and secondary food sources (Komorita et al. 2014), and (6) detecting nutrient and mineral uptake by plants (Clewlow et al. 2019). The principal aspects of animal nutrition generally investigated through SIA are diet patterns and trophic position of organisms along food chains (Boecklen et al. 2011; Bentivoglio et al. 2016). As mentioned, stable isotope analysis enables the evaluation of the trophic position of organisms and populations in food webs. However, the isotopic values in the consumer alone do not trace its trophic position (Bentivoglio et al. 2016). Indeed, there is a stepwise increase between the corresponding isotope signature of food consumed and the consumers’ tissue. Such expected isotopic increase between consumers and resources is referred to as isotopic discrimination or trophic enrichment (McCutchan et al. 2003). As an example, when an animal, such as a cow or a sheep, eats a specific plant, it will express the plant isotopic value in its muscles, bones or teeth, but the heavy isotope will be relatively more retained in the consumer’s tissues than the light one (Careddu et al. 2015; Cassano et al. 2016; Reid and Koch, 2017). However, the plant energy and nutrition values also vary through growth stages (Jafari and Torbatinejad, 2015). Thus, it is essential to consider potential differences in environmental conditions and diet components when studying animal diet (Jafari and Torbatinejad, 2015). Isotopic fractionation is particularly marked for nitrogen, while the carbon and sulfur isotopic composition of consumers closely reflects that in their diet. Studies with cows, fish, and zooplankton show that animal’s feces are enriched in 15N versus the diet, but urinary nitrogen (both NH3+ and urea) is depleted in 15N. For example, cow urine can be 1‰ to 4‰ depleted in 15N versus diet, while feces (2‰), and milk and blood (4‰ both) are enriched in 15N. The ratio of sulfur isotopes (δ34S) varies substantially among salt marsh and marine primary producers from -9.6‰ to +12.9‰ (Currin et al. 1995). Thus, the δ34S can be used to identify resource pools in these ecosystems. Similarly, the δ34S values have been measured in transitional water ecosystems and marine ecotones to distinguish between marine and freshwater inputs (Peterson and Howarth, 1987; Currin et al. 1995; Martinetto et al. 2006; Finlay and Kendall, 2008). δ13C values in phytoplankton can vary markedly with latitude and longitude. Indeed, δ13C values may vary with temperature, location, and growth rates that can affect the carbon uptake rate by phytoplankton (Zheng et al. 2019). Significant differences in carbon isotopes between animals indicate that consumers rely on different food sources or that their respective food webs are based on primary producers characterised by different isotopic signatures (Michener and Kaufman, 2008). Differences in the processing of carbon, nitrogen, and sulfur isotopes by animals stand out even more clearly, when the whole food web is examined. In many food webs, nitrogen isotope values increase by 10‰ to 15‰ from basal resources to top predators due to 3‰ to 5‰ stepwise increase among subsequent trophic levels. The opposite effect – no change with increasing trophic level – is observed for sulphur (Saggar et al. 1981). The isotopic differences in consumer δ13C may arise also by the consumption of C3 or C4 plants. The C3 plants are linked with a wetter and colder climate, while C4 plants are related to more arid and warmer conditions. As a consequence of metabolic adaptation by plants to such different climatic conditions, C4 plants generally show markedly higher δ13C values (from -10‰ to -18‰) than C3 plants (from -22‰ to -30‰) (Philp, 2015) (Figure 4). In addition, C3 plants may also show lower δ15N values than C4 plants (Figure 4). DeNiro and Epstein (1981) mentioned that δ13C values in C3 plants averaged around -25.5‰, while values around -9.0‰ were reported for C4 plants (Figure 2). The δ13C value of meat was around -18.0‰, suggesting that the animal’s diet was composed by a mix of C3 and C4 plants (Deniro and Epstein, 1981). Differences in δ13C values have also been reported between terrestrial plants and aquatic algae (Rossi et al. 2010). The latter generally show higher δ15N values than the former (Figure 3) (Schulting, 1998). This mainly depends on differences between the sources of carbon used for primary production in the two systems (Schulting, 1998). As shown in Figure 4, the δ13C of terrestrial C3 primary producers is generally lower than that of marine producers (Vinagre et al. 2011). In addition, δ13C signatures in plants can be affected by plant development and water management (Schulting, 1998; Barbieri, 2019). Given such expected differences among marine and terrestrial ecosystems, isotopic differences among aquatic consumers can inform on the benthic or terrigenous origin of nutrient inputs at the base of coastal or littoral food webs. Furthermore, dissolved inorganic carbon in estuaries commonly derives from different sources, either CO2 from the atmosphere or the dissolution of carbonate with approximately zero per mill value of δ13C (Finlay and Kendall, 2008; Bouillon et al. 2011). Given the predominance of C3 metabolism in coastal and aquatic vegetation, the δ13C in aquatic consumers usually display values around -28‰. Chanton and Lewis (1999) showed that the δ13C values in estuaries are closely related to the soluble inorganic carbon and water salinity. At the community level, the range of δ13C values (Carbon Range) can provide a useful indication of the diversity of basal resources consumed by animals (Wilkinson, 2018) and both δ13Cand δ15N describe the niche space occupied by all the organisms. To move from the isotopic description of organisms to the quantification of trophic interactions within the food web, Phillips (2012) proposed the use of mixing model equations. By explicitly taking into account uncertainties in consumer and resource isotopic signatures, the development of Bayesian approaches has enabled a more robust description of trophic links between species (Careddu et al. 2015; Rossi et al. 2019). Some tissues, such as the dentine of teeth, hairs, and feathers are metabolically inert.
Figure 4 Trophic enrichment in stable carbon and nitrogen values from primary producers, to terrestrial herbivores and predators (circles and squares) and marine ecosystems (black triangles)
Differences in δ13C between food webs based on C3 (black square symbols) and C4 plants (black circles symbols), omnivore/carnivore (grey circles), carnivore (grey squares) are also shown (Reid and Koch, 2017; Schulting, 1998)
Therefore, the study of these tissues can inform on the isotopic signature of a consumer’s diet at the time of tissue deposition. If the rate of tissue deposition is known, these tissues can provide a timeline of the consumer’s dietary history (Layman et al. 2012). For example, Hobson and Sease (1998) recorded ontogenetic isotopic shifts in Steller sea lions from tooth annuli (Hobson and Sease, 1998). Newsome et al. (2009) documented temporal changes in resource use by the California sea otter Enhydra lutris nereis by using regular sections of whiskers (Newsome et al. 2009). In these cases, information on the inert tissue deposition processes is necessary. Indeed, the process can be continuous over time (e.g., for whiskers of some mammal species), or discontinuous (e.g., for feathers) (Layman et al. 2012). In addition, it must be considered that different tissues are characterised by different turnover rates, thus providing dietary information over different time scales. Therefore, turnover rate data in the distinct tissues are required to conclude the degree of dietary proficiency (Layman et al. 2012). For instance, in some vertebrates, blood plasma integrates the diet over days to weeks, whereas turnover in muscle tissue is on the scale of months (Dalerum and Angerbjörn, 2005; Phillips and Eldridge, 2006). Nowadays, SIA is utilized to address questions about human diets around the world, and it has been said that ‘we are what we eat’ (Deniro and Epstein, 1981). SIA provides quantitative data that complete floral, faunal, and other information about dietary habits of individuals. This passage through human metabolism is specifically valuable to the quantitative study of human nutrition (Cooper et al. 2019). Humans express different isotope signatures according to the consumption of C3 and C4 plants (Figure 2), terrestrial animal proteins like cow, sheep, and goat meat, or aquatic animal resources such as fish and shellfish (Figure 3) (Schulting, 1998). Interestingly, there are diverse plant groups in human nutrition that can be differentiated through δ13C values in human tissues (such as hair), including C3 plants such as wheat, barley, soy, potatoes, fruits, vegetables versus C4 plants such as corn, sorghum, millet, sugar cane. This difference is also reflected in animal-derived food products such as milk carbon signatures ranging from -14‰ (diet-based C4 plants) to -27‰ (diet-based C3 plants) (Petzke et al. 2005).
The present paper has highlighted the stable isotope concept, applications, measurement method, and its relationships with animal nutrition. The many examples cited allow us to conclude that the analysis of stable isotopes of nitrogen and carbon is a powerful tool for evaluating animal feeding choices and trophic position in food webs, as well as the trophic sources supporting aquatic and terrestrial consumers. In addition, coupled with isotopic Bayesian mixing models, stable isotopes are a valuable tool that can provide insights into the structure and the complexity of food webs, as well as into the pathways of nutrient and energy transfer among ecosystem compartments and trophic levels. Nevertheless, it must be noted that many of the ecological questions addressed through the analysis of stable isotopes are reliant on the assumption that source pools have distinguished isotope values. When sources cannot be distinguished, stable isotopes may have little performance in answering questions about trophic relationships. In this case, stable isotope analysis should be complemented with additional information, such as stomach content and/or feces analysis, as well as other data on feeding behaviour including direct observation of feeding preferences in the field. In any case, both source and consumer pools must be sampled on suitable spatial and temporal scales to provide reliable information on diet composition. Isotope signature differences in samples depend on the climate, the isotopic baseline of the food web the consumer is part of, organisms’ dietary habits, and body conditions. Therefore, all these aspects should be considered in isotopic studies in order to achieve accurate results. Besides its broad application in environmental and ecological studies, SIA is increasingly used in thestudy of human diet, and ithas the potential to resolve many ambiguities in nutritional and medical studies.
This work was supported by PNRA-2015/AZ1.01 (M.L. Costantini) and PNRA16_00291(L. Rossi).