For more than thirty years, stable isotopes have been used as tools to address a wide range of questions in ecology, from elucidating key aspects of physiology and nutrition to tracking the movement of animals and defining the structure of biological communities [1
]. The value of stable isotope-based tools has been demonstrated repeatedly throughout this period, leading to an exponential increase in their use: a Web of Knowledge search for the topic “stable isotope*” yields fewer than one hundred results from the 1970s and more than four thousand in each of the past three years. However, the successful use of stable isotope-based tools by ecologists requires careful consideration of the fundamental processes whereby stable isotopes are metabolized and incorporated into the animals and plants being studied [4
]. In particular, the assimilation of dietary nutrients into the organism [2
], the routing and fractionation of assimilated nutrients within the organism [6
], and the rate of isotopic incorporation or turnover [12
] all have substantial influence on the results and interpretation of isotope-based ecological studies. Failure to account for these factors can result in the misidentification of diet composition, incorrect estimates of the timing of movements and dietary shifts, and even the incorrect assessment of trophic position and community structure. Accordingly, considerable progress has been made in testing assumptions about assimilation, routing, and isotopic turnover in lab settings, a necessary step to ensure that those processes are properly accounted for in ecological studies [4
]. Nevertheless, in spite of the work estimating the turnover rates of tissues in a wide range of taxa [16
], the importance of turnover processes to whole-animal metabolism and isotopic signatures is often not explicitly recognized in studies using isotope-based tools. Therefore, the purpose of this review is to 1) illustrate how stable isotopes can be effectively used to estimate the turnover rates of animal tissues and key compounds within tissues, and 2) discuss the use and importance of isotopic turnover estimates for ecological and nutritional studies, with a particular focus on how metabolic physiology provides the foundation for these applications. The physiological mechanisms and ecological applications that we discuss here are broadly relevant to all stable isotope studies, including those in terrestrial, freshwater, and marine systems, regardless of the specific isotope(s) used. We also outline the potential future of isotopic turnover measurements and particular challenges that future studies may approach using this methodology.
2. How to Measure Turnover Rates with Stable Isotopes: from Elements to Molecules
Estimating the turnover of a compound or tissue typically requires labeling the molecules of interest at one time and then tracking the concentration of those labels over time as the molecules are excreted, degraded, or converted into other forms and replaced with unlabeled molecules (Figure 1
A). The isotopic turnover of a compound or tissue therefore involves labeling the constituent atoms of molecules of interest with traceable stable isotopes of the elements that make up the molecule. Theoretically, any element could be labeled and tracked, but for practical reasons most studies of isotopic turnover have focused on carbon, nitrogen, and hydrogen using the 13
N, and 2
H isotopes, respectively. Turnover rates have also been estimated for the 34
S and 18
O isotopes of Sulfur and Oxygen, respectively, but estimates for these elements have been far less common [17
The isotopic label used in a given study should be chosen to match the compounds of interest in that study (Figure 1
B). Carbon, as the defining constituent, can be used to label all organic molecules, making it the most relevant element for the majority of isotopic turnover studies. Hydrogen and oxygen are also applicable to most organic molecules found in animal tissues, and are also commonly used to measure the turnover of body water, which underlies the estimation of energy expenditure with doubly labeled water [20
] as well as body composition with deuterium [24
]. In contrast, nitrogen is restricted to amino and nucleic acids and sulfur is primarily found in the amino acids cysteine and methionine, making 15
N and 34
S most applicable to measuring the turnover of proteinaceous tissue. Thus, for studies focused on the turnover of bulk tissues, 13
C will typically be the most straightforward label available.
Most studies, however, are interested in the turnover of more specific tissue components (e.g., proteins and fats, or amino acids and fatty acids), which requires either labeling those specific components or separating them prior to stable isotope analysis. Separating tissue samples into their macromolecular components can be readily achieved by isolating lipids and carbohydrates from protein components via lipid extraction [27
] and cation exchange purification [30
], respectively. Compounds of similar macromolecular classes can then often be separated by gas-chromatography [11
]. Alternately, studies focused on protein turnover could label just that tissue component with 15
N- or 34
S-enriched amino acids. Even more specific labels can be created for any compound by positioning isotopically heavier atoms at specific positions within the molecule [36
], although these may also require purification before analysis. In general, highly specialized labels will be most useful when the study is measuring the turnover of a small set of very specific compounds, and appears less beneficial as the scope of the study widens. For many ecological applications, whole-tissue or macromolecular turnover will be sufficient, whereas the turnover of specific compounds may be more important for nutritional and pharmacological studies.
Once an isotopic label has been chosen, the frequency, or enrichment, of that label needs to be manipulated, so that the rate of change over time can be measured (Figure 1
C). Typically, this will involve either enriching or depleting the tissue(s) with the label and then reversing the enrichment/depletion to ensure that a large enough change occurs to be accurately and precisely measured. Large scale manipulations are usually most readily accomplished with complete shifts in diet. For example, tissue 13
C concentration can be enriched by feeding animals diets based on C4
plants, or depleted with diets based on C3
plants, or enriched with marine diets and depleted with terrestrial diets [9
]. Similarly, 15
N can be enriched by using animal protein in diets, and depleted by using plant protein [9
]. It may also be possible to manipulate tissue 2
H by sourcing diets from different locations along the geographic 2
H gradients [9
]. Drinking water can also be spiked with 2
H or 18
O to produce whole-animal enrichment with those labels. For such large-scale diet manipulations, equilibrating animals with the initial diet is ideal to ensure that all tissues have a consistent and predictable isotope value, but this may not always be possible for tissues and species with very slow turnover rates. For more specific labels, direct administration is often applied with a dose of labeled molecules into the digestive tract by gavage or directly into the bloodstream or tissue by injection.
Previous studies and reviews have discussed the collection of tissue samples and the analysis of isotopic turnover data in great detail, and are excellent resources for those designing experiments [19
]. Briefly, several particularly important considerations are the number and spacing of samples over time following manipulation of the isotopic label and the statistical model used to describe changes in isotopic enrichment. The goal of sampling is to precisely track changes in isotopic enrichment, so as many samples should be taken as possible. However, when the number of samples is limited, they should be concentrated in the period immediately following the diet shift or the administration of the label (Figure 1
D). This spacing, usually following a geometric pattern (e.g., 0, 1, 2, 4, 8, etc.), ensures that more samples are taken when changes in isotopic enrichment are occurring most rapidly. Changes in isotopic enrichment are typically described and turnover rates estimated using exponential decay models (Figure 1
F), with the most commonly used being first order kinetic models of the form:
is the isotopic enrichment at time t
is enrichment at equilibrium with the second diet in ‰, y0
is enrichment at the time of the diet shift in ‰, t
is the time since the diet shift, and τ is the mean retention time of the isotope, which can be replaced with λ, the kinetic rate constant equal to 1/τ. Nevertheless, other options are possible, most notably multi-compartment models [47
], and may be favored on either empirical grounds or for mechanistic reasons if it is known that multiple sources contribute to the isotopic makeup of a given tissue or pool of molecules. Another important consideration for studies focused on the turnover of specific compounds is the interconversion between different molecules. This interconversion can decouple isotopic labels from their original compounds and should be accounted for by either isolating the compounds of interest during analysis or by correcting the measured isotopic enrichment of samples for the rate of conversion to other forms [11
5. The Future of Isotopic Turnover
Thus far, we have described the process and importance of estimating turnover rates for ecological studies that make use of stable isotope-based techniques. Below we briefly outline some of the future opportunities and challenges that isotopic turnover studies may face, including developing a more detailed mechanistic understanding of isotopic turnover, integrating isotope clock information about resource use with the direct tracking of movements, and adapting turnover-based methods in nutrition for use on wildlife.
5.1. What Mechanisms Drive the Turnover of Isotopes?
In the sections above we discussed how understanding variation in turnover rates among tissues, individuals, and species allows for more exact and effective diet reconstructions and isotopic clocks. It follows that a more precise knowledge of the mechanisms that contribute to such variation can further refine these ecological methods. Developing mechanistic models of turnover will be a particularly important step towards understanding the allometry of turnover and improving predictions of turnover rates derived from empirical models in novel taxa, thereby broadening their application to ecological studies. As we have described, there is already evidence suggesting that protein synthesis and degradation, rather than metabolic rate, are key mechanisms driving the turnover of non-lipid tissue components. Thus, while more studies are needed to verify this relationship, another important follow-up step is to integrate the knowledge of the regulation of protein synthesis with turnover rates. Specific topics of interest include 1) the links between isotopic and cellular or organelle turnover, perhaps contributing to rhythmic cycles of isotopic turnover or differences among tissues, 2) the relationship between the synthesis of specific proteins and isotopic turnover, and 3) the influence of dietary availability and the cycling of amino acids among tissues on isotopic turnover. Moving forward, it will also be important to ask similar questions about the turnover of other tissue components, particularly lipids, which may have different relationships with metabolism or among tissues than those of protein.
5.2. How Can Turnover Be Integrated with Tracking Technologies?
In recent years there have been tremendous advances in the number and sophistication of tracking technologies available to ecologists [126
]. Since one of the major ecological applications of stable isotopes is for tracking the location and timing of animal movements, there is an excellent opportunity to cross-validate the results of these methods. A particularly important step is to test the precision of estimates of departure and arrival times estimated from stable isotope enrichment data with isotopic clocks. Moreover, by pairing movement histories with isotope data it may be possible to assess the effect of fasting and exercise on stable isotope signatures in a natural context. The combination of these effects with a greater mechanistic understanding of turnover could greatly refine stable isotope methods. Meanwhile, the ability to pair stable isotope data with movement data creates the possibility of more detailed analyses of resource use during large scale movements, such as diet selection at migratory stopover sites. These finer details are likely to be an important supplement to data collected from coarser tracking technologies, such as light-level geolocators, and may provide new insights into the motivations for animal movements.
5.3. How Can Turnover Inform Wildlife Nutrition Studies?
As we discussed above, methods that link isotopic turnover to diet requirements and nutritional state, or methods that infer past nutritional stress from isotopic signatures, have mostly been pursued in human contexts and will require some development before they can be reliably used to study the nutrition of wild animals. At a minimum, turnover rates of multiple tissues will need to be estimated and the isotopic signatures of nutritional stress will need to be established with greater confidence for a wider range of species. Subsequently, it will be necessary to empirically test the predictions for nutritional time series based on differences in turnover among tissues. As with ecological stable isotope methods, this process will be facilitated by a more detailed, mechanistic understanding of the drivers of turnover, which may elucidate any differences among tissues, individuals, and species. In addition to general mechanisms driving turnover, it will be important to clarify mechanisms specific to nutritional stress such as changes in routing and rates of de novo synthesis of tissue components. Investment in such challenges will further improve stable isotope-based assessments of nutritional state and history in ecological studies, thereby expanding our understanding of the motivations and success of animals.