SEARCH

SEARCH BY CITATION

Keywords:

  • assimilated diet;
  • isotope discrimination;
  • nitrogen;
  • protein quality;
  • stable isotopes;
  • sulphur

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1. Accurately predicting isotope ratio discrimination is central to using mixing models to estimate assimilated diets of wild animals. This process is complicated when omnivores consume mixed diets because their discrimination is unlikely to be the weighted average of the various dietary constituents as occurs in current models.

2. We sought a basic understanding of how protein quality and quantity determine Δ15N and Δ34S in mammals and birds. Dietary protein is the primary source of both elements in many plants and animals. Low protein quality and high protein content have the potential to increase Δ15N by increasing protein turnover.

3. Protein quality, defined as the relative amount of the most limiting amino acid, accounted for 87–90% of the variation in Δ15N when mammals and birds consumed plant matter and mixed diets of plants and animals with protein of intermediate quality and quantity. However, foods containing relatively large amounts of high quality protein (i.e. vertebrate flesh) and foods with exceptionally low quality protein (e.g. lentils, Lens culinaris) had disparate nitrogen discriminations relative to what would be predicted from protein quality alone.

4. Supplementation of plant and animal diets with nitrogen-free carbohydrates and fats to reduce protein quantity did not reduce Δ15N in three plant-based diets fed to laboratory rats, but reduced Δ15N in two of three meat diets with >50% protein.

5. Δ34S was weakly correlated with Δ15N (R2 = 0·48) but was highly correlated with dietary δ34S (R2 = 0·89). Because methionine, a sulphur amino acid, was the most limiting amino acid in all diets, sulphur should be highly conserved as indicated by the lack of any change in Δ34S when diets were supplemented with carbohydrates and fat to both provide additional energy and reduce protein content.

6. Predictive equations incorporating both protein quality and quantity accounted for 81% of the variation in Δ15N and offer the opportunity to create more realistic mixing models to accurately estimate assimilated diets for omnivores.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Accurately predicting isotopic discrimination is central to estimating assimilated diets of wild animals when using stable isotopes (Martinez del Rio et al. 2009). While many studies have postulated or identified causes of variation in discrimination (Fantle et al. 1999; Roth & Hobson 2000; McCutchan et al. 2003; Pearson et al. 2003; Vanderklift & Ponsard 2003; Gaye-Siessegger et al. 2004, 2007; Robbins, Felicetti & Sponheimer 2005; Miron et al. 2006; Caut, Angulo & Courchamp 2008, 2009; Tsahar et al. 2008; Robbins, Felicetti & Florin 2010; Smith et al. 2010), none have proposed cause-effect biologically based models for accurately estimating unknown discriminations even though selection of discrimination values is the single most important assumption determining assimilated diet estimates. The lack of such models to accurately predict nitrogen, carbon, or sulphur discriminations, particularly for foods in mixed diets, may lead to unacceptable errors in estimating assimilated diets of ancestral humans and wild animals (Caut, Angulo & Courchamp 2008; Robbins, Felicetti & Florin 2010).

Current approaches to estimating unknown discriminations for foods consumed by free-ranging animals include: (i) feeding wild-collected foods to captive animals and directly measuring their discrimination, which is not always feasible and may rarely simulate field complexity; (ii) using a grand mean for all foods (e.g. 2·0–3·4‰ for nitrogen and 0‰ for sulphur), which ignores the three- to fourfold variation in Δ15N (e.g. c.−2 to 6‰) and Δ34S (e.g. c.−3 to 8‰); or (iii) using various regressions between dietary isotope values and discriminations that have been determined with captive wildlife consuming a wide range of foods, which describe very general relationships that may not be cause-effect (Peterson & Fry 1987; McCutchan et al. 2003; Vanderklift & Ponsard 2003; Robbins, Felicetti & Sponheimer 2005; Caut, Angulo & Courchamp 2009; Martinez del Rio et al. 2009; Robbins, Felicetti & Florin 2010).

Two major hypotheses have been proposed to explain much of the dietary-induced variation in Δ15N. The protein quantity hypothesis suggests that as dietary protein content (%) or intake (g day−1) increase, Δ15N will increase (Pearson et al. 2003; Martinez del Rio et al. 2009). The protein quality hypothesis suggests that as protein quality decreases, Δ15N will increase (Roth & Hobson 2000; Robbins, Felicetti & Sponheimer 2005; Robbins, Felicetti & Florin 2010). Both are based on the observation or hypothesis that as dietary protein intake or amino acid scavenging increase, nitrogen excretion will increase and lead to the preferential retention of 15N which will elevate the animal’s δ15N value relative to the diet.

Although Robbins, Felicetti & Sponheimer (2005) and Robbins, Felicetti & Florin (2010) found no support for the protein quantity hypothesis when plotting either nitrogen content or carbon : nitrogen ratios against Δ15N, such plots are confounded by lower protein, largely plant-based diets of poorer protein quality at one end of the regression and higher protein, largely animal-based diets of higher protein quality at the other. If both protein quality and quantity are important, Δ15N may be elevated when plant-based diets are consumed primarily because of their poorer protein quality and when animal-based diets are consumed primarily because of their higher protein content. Thus, we hypothesized that both protein quality and quantity may be important, but the relationships are more complex than either proposal alone suggests.

Thus far, little use has been made of sulphur isotopes for estimating assimilated diet, although the consumer’s isotope value should reflect the dietary isotope value (Felicetti et al. 2003; McCutchan et al. 2003; Arneson & MacAvoy 2005). Δ15N and Δ34S may be related in that sulphur amino acids (methionine, cystine, cysteine and taurine) are important sources of organic sulphur (Arneson & MacAvoy 2005). If sulphur amino acids are important in determining protein quality, dietary sulphur amino acid content may be important in determining both Δ15N and Δ34S. Consequently, we sought a unified concept incorporating both protein quality and quantity that could be used to understand and accurately predict Δ15N, Δ34S and assimilated diets of omnivores.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Quantifying protein quantity and quality

While nitrogen or protein quantity (N × 6·25) has been measured in virtually all studies, protein quality has not. There are many measures of protein quality. Some are based on feeding studies (e.g. protein efficiency ratio, biological value, or net protein utilization) and others are based on how well the essential amino acid profile of a particular food matches a hypothetical perfect protein or the animal’s requirements (e.g. chemical score). The latter estimates are appealing in that amino acid profiles of many foods have been determined and very extensive effort has been made to define the amino acid requirements of domestic and laboratory animals [NRC (National Research Council) 1995].

The complete amino acid profiles of Chinook salmon (Oncorhynchus tshawytscha) and white-tailed deer (Odocoileus virginianus) fed to brown bears (Ursus arctos) and American black bears (Ursus americanus) (Hilderbrand et al. 1996; Felicetti et al. 2003) and various diets composed of corn, wheat, alfalfa, soybean meal, lentils, chicken meal, pork meat and bone meal and fish meal fed to laboratory rats (Robbins, Felicetti & Florin 2010; current study) were determined at the University of Missouri Agricultural Experiment Station Chemical Laboratories. Briefly, acid and alkaline hydrolysates were analysed using a high-performance liquid chromatographic amino acid analyzer. Additional amino acid profiles or protein contents of foods not reported by other investigators were estimated from the compilations of NRC (1994), Davis et al. (1994), American Casein Co. (Burlington, NJ, USA) and the USDA National Nutrient Database for Standard Reference, Agricultural Research Service (http://www.nal.usda.gov/fnic/foodcomp/search/) (see Table S1, Supporting information).

The basis for estimating protein quality was to express the concentration of each essential amino acid in the diet as a percent of the diet’s crude protein (N × 6·25) content. This relative concentration of each amino acid was compared with the estimated dietary requirement of that amino acid as a percent of the total protein requirement for growth by laboratory rats (Rattus rattus) (NRC 1995) to determine which amino acid might be most limiting. Amino acid requirements for growing laboratory rats were used as the standard for all animals because (i) the amino acid requirements for wild animals are almost entirely unknown; (ii) the current and previous study (Robbins, Felicetti & Florin 2010) used laboratory rats; and (iii) laboratory rats have not been heavily selected for meat, milk, or egg production as have many other domestic animals (e.g. livestock and poultry) and, therefore, may be a more appropriate comparison with wild animals.

Selecting nitrogen and sulphur discrimination values

Δ15N and Δ34S values for serum, plasma, whole blood, or red blood cells were sought for diets that covered the widest possible ranges of protein quality and quantity, had been fed long enough to ensure diet to animal equilibration and had been fed by multiple investigators or in various combinations to ensure that the reported isotope discriminations were reliable. Unfortunately, results on commercial rodent and poultry diets as well as several other diets could not be used because of the impossibility of estimating amino acid profiles. Similarly, feeding studies that used pelleted diets were excluded because of concern about protein damage (Robbins, Felicetti & Florin 2010), and studies that fed fungi, crustacea, or insects [e.g. mealworms (Tenebrio molitor)] were excluded because much of their nitrogen occurs as chitin (a nitrogen-containing carbohydrate) (Claridge et al. 1999; Pearson et al. 2003). For example, van Tets & Hulbert (1999) estimated that 69% of the nitrogen in mealworms occurred as non-protein chitin.

Testing the interaction between protein quantity, quality and Δ15N

Two approaches were used to test the interactions between protein quantity and quality in determining Δ15N. The first approach was an indirect test in which the relative amount of the most limiting essential amino acid was compared with the Δ15N for several diet–animal combinations used in the current and previous studies (Hobson & Clark 1992; Hilderbrand et al. 1996; Hobson et al. 1996; Ben-David & Schell 2001; Jenkins et al. 2001; Bearhop et al. 2002; Lesage, Hammill & Kavacs 2002; Felicetti et al. 2003; Sponheimer et al. 2003; Ogden, Hobson & Lank 2004; Arneson & MacAvoy 2005; Cherel, Hobson & Hassani 2005; Robbins, Felicetti & Sponheimer 2005; Podlesak & McWilliams 2006; Darr & Hewitt 2008; Tsahar et al. 2008; Robbins, Felicetti & Florin 2010) (see Table S2). If the protein quality hypothesis is valid, Δ15N should decrease as the concentration of the most limiting amino acid increases across diets. Similarly, if the protein quantity hypothesis is valid, Δ15N should increase above the relationship determined by protein quality alone once the most limiting amino acid is no longer the sole determinant of dietary protein turnover.

The second approach was a direct test in which foods ranging in both protein quality and quantity were supplemented with additional energy to dilute the protein concentration and thereby reduce daily protein intake. The Δ15N and Δ34S of animals consuming the energy-supplemented diets should be less than the non-supplemented diets when protein quantity becomes important in determining discrimination. Specifically, we hypothesized that Δ15N values for plant-based diets would be less likely to decrease with energy dilution than animal-based diets.

Thus, the diets used included fish meal (Brevoortia tyrannus), chicken meal, pork meat and bone meal, soybean meal, lentils and wheat because they cover a wide range in both protein quantity and quality in both plants and animals. All feeds were purchased as single batches, finely ground and mixed thoroughly to ensure that composition and isotopic values were constant. Each diet was fed in the undiluted form followed immediately by the diluted form to the same 10 rats. The diluted diets were supplemented with nitrogen- and sulphur-free sucrose, starch and corn oil in the ratio of 5 : 2 : 2 : 1, such that the protein concentration was reduced by 50%. Further dilution was not attempted because of concern about creating nutritional deficiencies.

Ten male, Sprague–Dawley laboratory rats were used in all feeding trials. Each feeding trial lasted 21 days to ensure that plasma had equilibrated with the diet and followed the protocol of Robbins, Felicetti & Florin (2010). Blood samples were collected in heparinized tubes at the end of each feeding trial. Plasma was separated, frozen, and freeze-dried. All rats were fed ad libitum to promote positive energy and protein balance, weight gain, and therefore minimal tissue mobilization. Rats were weighed weekly.

Isotopic and statistical analyses

δ15N and δ34S values for diets and freeze-dried plasma were determined with a continuous flow isotope ratio mass spectrometer (Delta PlusXP; Thermo Finnigan, Bremen, Germany) at the Washington State University Stable Isotope Core Laboratory. Mean dietary isotope values were based on the analyses of five samples per diet. Nitrogen isotope ratios are reported as per mil (‰) relative to atmospheric nitrogen (δ15N). Sulphur isotope ratios are reported as per mil relative to Vienna Canon Diablo Troilite by assigning a value of −0·3‰ to IAEA S-1 silver sulphide. Laboratory reference standards (acetanilide and keratin for nitrogen and sulfanilimide, IAEA S-2, IAEA SO5, and IAEA S3 for sulphur) were interspersed throughout each analysis to ensure maintenance of calibration. Analytical errors (±1 SD) for the above standards were ≤0·1‰ for nitrogen and ≤0·4‰ for sulphur.

Linear and curvilinear least squares regressions were used to model all relationships (SAS 1998). Differences in slopes of regressions were tested using small sample t-tests (Kleinbaum & Kupper 1978). anova was used to test for differences in discrimination between diets. A P-value of <0·05 was considered significant. Means are reported with ±1 SD.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Nitrogen isotope ratio discrimination

Protein quality as defined by the relative methionine concentration accounted for 87–90% of the variation in Δ15N when animals consumed diets that ranged from 6·9% to 53·8% protein with methionine concentrations ranging from 1·3% to 2·6% (Fig. 1, see Tables S1 and S2). The inclusion of the other sulphur-containing amino acids that can partially substitute for methionine (i.e. cystine, cysteine and taurine) did not improve the regressions. The pattern of decreasing Δ15N with increasing protein quality occurred for laboratory rats consuming a wide range of single-item and mixed diets (Robbins, Felicetti & Florin 2010; current study), non-primate neonates consuming milk (Robbins 1993; Davis et al. 1994; Jenkins et al. 2001; Robbins, Felicetti & Sponheimer 2005), wild and domestic ruminants consuming alfalfa or alfalfa and corn (Jenkins et al. 2001; Sponheimer et al. 2003; Darr & Hewitt 2008), and yellow-vented bulbuls (Pycnonotus xanthopygos) and yellow-rumped warblers (Dendroica coronata) consuming mixed diets of casein and bananas (Tsahar et al. 2008) or casein, sugar and olive oil (Podlesak & McWilliams 2006).

image

Figure 1.  The relationship between dietary protein quality as defined by the limiting amino acid (methionine) and Δ15N for the plasma, serum or red blood cells of laboratory rats consuming various diets of corn, wheat, alfalfa, lentils, soybean meal, fish meal, pork meat and bone meal, chicken meal, and their mixtures (Robbins, Felicetti & Florin 2010; current study) and various wild birds and mammals consuming fish (Hobson & Clark 1992; Hilderbrand et al. 1996; Hobson et al. 1996; Ben-David & Schell 2001; Bearhop et al. 2002; Lesage, Hammill & Kavacs 2002; Felicetti et al. 2003; Cherel, Hobson & Hassani 2005), fish meal (Ogden, Hobson & Lank 2004; Arneson & MacAvoy 2005; Robbins, Felicetti & Florin 2010), quail (Hobson & Clark 1992), ungulates (Hilderbrand et al. 1996; Ben-David & Schell 2001; Bearhop et al. 2002), alfalfa or alfalfa and corn (Jenkins et al. 2001; Sponheimer et al. 2003; Darr & Hewitt 2008), non-primate milks (Davis et al. 1994; Jenkins et al. 2001; Robbins, Felicetti & Sponheimer 2005) and casein-supplemented diets (Podlesak & McWilliams 2006; Tsahar et al. 2008) (Table S2). Although results for vertebrate flesh with high quality protein (i.e. ungulates, fish meal, fish, chicken meal and quail) and lentils as the entire diet are plotted, they are not included in the regressions.

Download figure to PowerPoint

However, several diets had either higher or lower discriminations than predicted by the regression equations of Fig. 1. For example, lentils containing relatively low quality protein had a much lower Δ15N than predicted by the regressions. At the other extreme, high-protein meat diets containing relatively high quality protein (e.g. fish, fish meal, chicken meal, ungulates, and quail) fed to various mammals (Canis latrans, Halichoerus grypus, Mustela vison, Pagophilus groenlandicus, Phoca hispida, Phoca vitulina, U. americanus, U. arctos) and birds (Calidris alpine pacifica, Catharacta skua, Corvus brachyrhynchos, Falco peregrines, Larus delawarensis) had higher nitrogen discriminations than predicted from protein quality alone (Fig. 1, see Tables S1 and S2) (Hobson & Clark 1992; Hilderbrand et al. 1996; Hobson et al. 1996; Ben-David & Schell 2001; Bearhop et al. 2002; Lesage, Hammill & Kavacs 2002; Felicetti et al. 2003; Ogden, Hobson & Lank 2004; Arneson & MacAvoy 2005; Cherel, Hobson & Hassani 2005; Robbins, Felicetti & Florin 2010). Discriminations for five of six meat diets averaged 1·1 ± 0·4‰ higher (range = 0·5–1·6‰) (Fig. 1) than predicted from protein quality alone. The exception to this trend occurred when laboratory rats consumed pork meat and bone meal that contained relatively low quality protein (Fig. 1, see Tables S1 and S2). Its discrimination (5·0 ± 0·1‰) was similar to what would be predicted from the more general protein quality regressions of Fig. 1 (5·1–5·2‰).

The protein content of the plant and animal foods used in the protein dilution study ranged from 12·5% to 72·0% (Fig. 2, see Table S2); and protein quality in those foods was limited by the amino acid methionine, which ranged from 0·85% to 2·61% of the crude protein (see Tables S1 and S2). Average daily protein intake was reduced by 48·7 ± 3·5% when rats consumed the diluted diets relative to the undiluted diets. Rats gained weight on all plant-based diets with and without dilution (1·7 ± 1·2 g day−1, range = 0·7–3·7) and on five of six animal-based diets (1·3 ± 0·5 g day−1, range = 0·8–2·0). The exception was some rats lost weight on the undiluted pork meat and bone meal (−1·0 ± 1·5 g day−1), but they gained weight on diluted pork meat and bone meal (1·5 ± 0·35 g day−1). However, there was no difference in the nitrogen discrimination for rats that lost weight when consuming the pork meat and bone meal as compared with those that maintained or gained weight (t = 0·58, P = 0·59).

image

Figure 2.  The effect on nitrogen and sulphur discrimination when diets composed of fish meal, chicken meal, pork meat and bone meal, soybean meal, lentils and wheat were fed to laboratory rats with and without dilution. The diluted diets were created by supplementing each of the above foods with sucrose, starch and corn oil in the ratio of 5 : 2 : 2 : 1 to reduce protein concentration to half of that in the undiluted diet. The change in Δ15N and Δ34S is the difference between when laboratory rats were fed the diluted diet minus the undiluted diet.

Download figure to PowerPoint

Nitrogen discrimination did not decrease with energy dilution in any of the plant-based diets (lentils, F = 0·43, P = 0·52 and soybean meal F = 0·40, P = 0·54), although Δ15N slightly increased (0·17 ± 0·17‰) when wheat was diluted (F = 7·53, P = 0·01) (Figs 1 and 2). The mean difference in discrimination due to dilution for the three plant diets was 0·06 ± 0·13‰ and did not differ from 0 (t = 0·77, P = 0·52). Δ15N decreased in two of three meat diets (fish meal, = 91·4, < 0·0001 and pork meat and bone meal, = 34·9, < 0·001), but did not decrease when chicken meal was diluted (= 1·69, = 0·21) even though seven of the ten rats had lower discriminations when consuming the diluted diet (Figs 1 and 2).

Nitrogen discriminations can be predicted (R2 = 0·81, F ≥ 68·4, < 0·0001, Fig. 3) across the breadth of dietary data by either one of two equations utilizing both protein quality and quantity:

  • image(eqn 1)
  • image(eqn 2)

where X is protein quality [eqn 1, methionine content as a per cent of total dietary protein (Fig. 1a) or eqn 2, the relative deficit of the most limiting amino acid as a per cent of the requirement for growth by laboratory rats (Fig. 1b)] and Z is dietary protein content (% of total dietary dry matter). The equations utilized all data of Table S2 and Fig. 1 with the exception of the values for lentils (see Discussion). Protein quality accounted for 75% of the variation (F ≥ 98·9, P < 0·0001) and protein quantity for 7% (F = 2·5, P = 0·12). More complex regressions, such as curvilinear regressions or linear and curvilinear regressions with thresholds for a protein quantity effect (e.g. ≥50%), produced similar overall predictive capabilities (R2 = 0·82–0·84, F ≥ 78·9, P < 0·0001) and estimates of the relative importance of protein quality (74–76%) and quantity (5–6%) (eqns 1 and 2).

image

Figure 3.  The relationship between the observed discriminations of the diets in Table S2 and their predicted discriminations when solving eqns 1 and 2 utilizing their respective protein qualities and quantities. Dashed line is the 1 : 1 relationship between the variables.

Download figure to PowerPoint

Sulphur isotope ratio discrimination

Sulphur amino acids accounted for 84 ± 20% of the dietary sulphur in corn, wheat, alfalfa, soybean meal, fish meal, chicken meal, and pork meat and bone meal. However, Δ34S was not highly correlated with Δ15N (Fig. 4). Δ34S did not change when sucrose, starch and corn oil were added to any of the six feeds relative to the undiluted diets (mean change in Δ34S values with dilution = 0·02 ± 0·13, F = 0·34–1·26, P = 0·08–0·77) (Fig. 2). Dietary δ34S accounted for 89% of the variation in Δ34S (see Table S3, Fig. 5). Regressions between various measures of sulphur amino acid content, including total sulphur amino acid content, methionine content, cystine and cysteine content, and methionine to cystine ratio, had lower correlation coefficients that ranged from 0·46 to 0·77.

image

Figure 4.  The relationship between nitrogen and sulphur discrimination for a range of foods fed to laboratory rats (Robbins, Felicetti & Florin 2010; current study) and grizzly bears (Felicetti et al. 2003).

Download figure to PowerPoint

image

Figure 5.  The relationship between dietary δ34S, plasma or serum δ34S, and Δ34S for diets fed to laboratory rats (Robbins, Felicetti & Florin 2010; current study) and grizzly bears (Felicetti et al. 2003) (Table S3).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Numerous animal and dietary factors have been proposed to affect nitrogen discrimination by specific tissues. The animal factors include intake rate, growth rate, metabolic rate, isotope routing, and type of nitrogen excretion (ureotelic or uricotelic), and the dietary factors include protein quality and quantity (Martinez del Rio et al. 2009; Kelly & Martinez del Rio 2010; Smith et al. 2010). The animal factors create concern when trying to estimate the assimilated diets of both ancient and extant animals because they are rarely known.

The current relationships (Fig. 1 and eqns 1 and 2), which were developed for mammals and birds that were either maintaining or gaining weight, suggest that most of the variation in nitrogen discriminations under these conditions is due to dietary protein quality and quantity (R2 = 0·81–0·90). As we hypothesized, Δ15N values for animals consuming plant-based diets of lower protein quality, even when containing relatively large amounts of protein (e.g. lentils and soybean meal), were not reduced by dietary energy dilution. Δ15N values for animals consuming two of three high protein meat diets were reduced when diluted with additional energy, although the reductions were relatively small in both the dilution feeding studies (Figs 1 and 2) and in eqns 1 and 2 when using data from this and other studies.

Nevertheless, the predictive power of the regressions based on these two variables exceeds that determined by the more common regressions between dietary δ15N and Δ15N, which explain from 0 to 67% (mean = 40 ± 32%) of the variation in birds and mammals (Caut, Angulo & Courchamp 2009; Robbins, Felicetti & Florin 2010). Furthermore, the use of laboratory rat nutrient standards for growth as a basis for comparing a wide variety of birds and mammals of various sizes, gastrointestinal tracts, and productivity suggests that dietary-induced metabolic relationships determining discrimination are quite conservative.

Lentils, soybean meal, and pork meat and bone meal were chosen as test foods because of their high protein content but generally low protein quality. Both of these characteristics should produce relatively high Δ15N values, with the lentil value being extremely high. For example, the expected Δ15N for lentils based on the regressions of Fig. 1 would have ranged from 7·4% to 9·8‰. However, the Δ15N for lentils (5·6 ± 0·2‰), soybean meal (5·7 ± 0·1‰), and pork meat and bone meal (5·0 ± 0·1‰) did not exceed 6·0‰.

In a compilation of 134 Δ15N values for various tissues from mammals, birds, crustacea, insects and fish (Vanderklift & Ponsard 2003), <4% of the values were above 5·5‰ and none exceeded 6‰. In a more recent compilation of 142 Δ15N values for mammals and birds (Caut, Angulo & Courchamp 2009), only four were above 6‰, although three of the four were incorrectly estimated from Felicetti et al. (2003) and actually ranged from 4·3‰ to 5·8‰. Thus, the aggregate of these observations suggests an upper limit to Δ15N of c. 6‰ for mammals and birds consuming foods that do not contain significant amounts of non-protein nitrogen. Therefore, Δ15N estimates produced by eqns 1 and 2 should be capped at a maximum of 6‰ unless a particular food–animal combination is known to produce a higher discrimination.

The regression between dietary δ34S and Δ34S has a higher correlation coefficient than those measured for similar carbon and nitrogen regressions and, therefore, may be all that is needed to estimate assimilated diet (Hilderbrand et al. 1996; Felicetti et al. 2003; McCutchan et al. 2003; Vanderklift & Ponsard 2003; Robbins, Felicetti & Florin 2010). We hypothesize that this high correlation coefficient occurred in this study because methionine was the primary, limiting, essential amino acid in all diets. Therefore, sulphur and the sulphur amino acids should be highly conserved during animal metabolism as demonstrated by the lack of any change in Δ34S during the dietary dilution study. The relatively low correlation coefficient (0·48) between Δ34S and Δ15N is similar to earlier results (0·44) for insects and fish (McCutchan et al. 2003), which suggests a more complex relationship between the two variables. Presumably, Δ15N reflects the metabolism of all amino acids and varies with both protein quality and quantity, whereas Δ34S reflects the metabolism of only sulphur amino acids. Therefore, the two variables are not directly related and the lower correlation coefficient should be expected.

If the above results linking protein quality, protein quantity, and Δ15N are confirmed or refined by further studies, estimating nitrogen discriminations for omnivores without detailed knowledge of the animal factors may not limit accurate estimates of assimilated diet. However, current limitations to this approach include the lack of: (i) a broad understanding of amino acid profiles in the wide range of foods consumed by wild animals and the time course of their metabolic interaction within the consumer that will determine if they are complementary or non-complementary; (ii) an understanding of how mixtures of protein and non-protein nitrogen (e.g. chitin) in insects, crustacea, and fungi determine nitrogen discrimination; and (iii) mixing models in which the discriminations of the individual foods vary from being independent, additive, and linear for foods consumed in metabolically distinct meals to dependent and curvilinear when foods with complementary amino acid profiles are consumed in metabolically mixed diets (DeGabriel, Foley & Wallis 2002; Robbins, Felicetti & Florin 2010) (Fig. 6).

image

Figure 6.  Illustration of assimilated diet estimates for an omnivore consuming a two-component diet (plants and animals) using either linear (no dietary interaction) or curvilinear (metabolically mixed diet with complementary amino acid profiles) solutions. The assumptions were that: (i) the plant component of the diet had a δ15N signature of −1·0‰, a protein quality of 1·4% methionine and a protein content of 24%, which gave a discrimination estimate of 5·0‰ (eqn 1); and (ii) the animal component had a δ15N signature of 4·0‰, a protein quality of 2·5% methionine and a protein content of 77%, which gave a discrimination estimate of 3·5‰. Intermediate discriminations for the metabolically mixed diets were determined by solving eqn 1 for various dietary mixtures. Because the discriminations at a given dietary mixture were lower when the two foods were consumed in a metabolically mixed diet than when there was no dietary interaction, the linear model underestimates the importance of animal matter and overestimates the importance of plant matter in the diet when the foods were consumed in a metabolically mixed diet. The maximum error in the assimilated diet estimates for each dietary component in this example was 14%.

Download figure to PowerPoint

This latter point means that discriminations may need to be predicted parameters in mixing models based on additional animal, dietary, and temporal inputs rather than the current a priori estimates. However, investigators working with piscivores, other carnivores, and herbivores may have a much easier task in estimating discriminations as many of these groups do not consume foods that vary as extensively in protein quality and quantity as do the diets consumed by some omnivores. Although there may be other diets with nitrogen discriminations that are outside the bounds of our current understanding, the equations developed in this study offer the opportunity to begin developing more complex and realistic mixing models for omnivores that more accurately estimate assimilated diets.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The project was approved by the Washington State University Institutional Animal Care and Use Committee (#03762) and funded by the Nutritional Ecology Research Endowment and the US Fish and Wildlife Service.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Arneson, L.S. & MacAvoy, S.E. (2005) Carbon, nitrogen, and sulfur diet-tissue discrimination in mouse tissues. Canadian Journal of Zoology, 83, 989995.
  • Bearhop, S., Waldron, S., Votier, S.C. & Furness, R.W. (2002) Factors that influence assimilation rates and fractionation of nitrogen and carbon stable isotopes in avian blood and feathers. Physiological and Biochemical Zoology, 75, 451458.
  • Ben-David, M. & Schell, D.M. (2001) Mixing models in analyses of diet using multiple stable isotopes: a response. Oecologia, 127, 180184.
  • Caut, S., Angulo, E. & Courchamp, F. (2008) Discrimination factors (Δ15N and Δ13C) in an omnivorous consumer: effect of diet isotopic ratio. Functional Ecology, 22, 255263.
  • Caut, S., Angulo, E. & Courchamp, F. (2009) Variation in discrimination factors (Δ15N and Δ13C): the effect of diet isotopic values and applications for diet reconstruction. Journal of Applied Ecology, 46, 443453.
  • Cherel, Y., Hobson, K.A. & Hassani, S. (2005) Isotopic discrimination between food and blood and feathers of captive penguins: implications for dietary studies in the wild. Physiological and Biochemical Zoology, 78, 106115.
  • Claridge, A.W., Trappe, J.M., Cork, S.J. & Claridge, D.L. (1999) Mycophagy by small mammals in the coniferous forests of North America: nutritional value of sporocarps of Rhizopogon vinicolor, a common hypogeous fungus. Journal of Comparative Physiology, 169B, 172178.
  • Darr, R.L. & Hewitt, D.G. (2008) Stable isotope trophic shifts in white-tailed deer. Journal of Wildlife Management, 72, 15251531.
  • Davis, T.A., Nguyen, H.V., Garcia-Bravo, R., Fiorotto, M.L., Jackson, E.M., Lewis, D.S., Lee, D.R. & Reeds, P.J. (1994) Amino acid composition of human milk is not unique. Journal of Nutrition, 124, 11261132.
  • DeGabriel, J., Foley, W.J. & Wallis, I.R. (2002) The effect of excesses and deficiencies in amino acids on the feeding behavior of the common brushtail possum. Journal of Comparative Physiology, 172B, 607617.
  • Fantle, M.S., Dittel, A.I., Schwalm, S.M., Epifanio, C.E. & Fogel, M.L. (1999) A food web analysis of the juvenile blue crab, Callinectes sapidus, using stable isotopes in whole animals and individual amino acids. Oecologia, 120, 416426.
  • Felicetti, L.A., Schwartz, C.C., Rye, R.O., Haroldson, M.A., Gunther, K.A., Phillips, D.L. & Robbins, C.T. (2003) Use of sulfur and nitrogen stable isotopes to determine the importance of whitebark pine nuts to Yellowstone grizzly bears. Canadian Journal of Zoology, 81, 763770.
  • Gaye-Siessegger, J., Focken, U., Muetzel, S., Abel, H. & Becker, K. (2004) Feeding level and metabolic rate affect δ13C and δ15N values in carp: implications for food web studies. Oecologia, 138, 175183.
  • Gaye-Siessegger, J., Focken, U., Abel, H. & Becker, K. (2007) Starvation and low feeding levels result in an enrichment of C-13 in lipids and N-15 in protein of Nile tilapia Oreochromis niloticus L. Journal of Fish Biology, 71, 90100.
  • Hilderbrand, G.V., Farley, S.D., Robbins, C.T., Hanley, T.A., Titus, K. & Servheen, C. (1996) Use of stable isotopes to determine diets of living and extinct bears. Canadian Journal of Zoology, 74, 20802088.
  • Hobson, K.A. & Clark, R.G. (1992) Assessing avian diets using stable isotopes II: factors influencing diet-tissue fractionation. The Condor, 94, 189197.
  • Hobson, K.A., Schell, D.M., Renouf, D. & Noseworthy, E. (1996) Stable carbon and nitrogen isotopic fractionation between diet and tissues of captive seals: implication for dietary reconstructions involving marine mammals. Canadian Journal of Fish and Aquatic Science, 52, 528533.
  • Jenkins, S.G., Partridge, S.T., Stephenson, T.R., Farley, S.D. & Robbins, C.T. (2001) Nitrogen and carbon isotope fractionation between mothers, neonates, and nursing offspring. Oecologia, 129, 336341.
  • Kelly, L.J. & Martinez del Rio, C. (2010) The fate of carbon in growing fish: an experimental study of isotopic routing. Physiological and Biochemical Zoology, 83, 473480.
  • Kleinbaum, D.G. & Kupper, L.L. (1978) Applied Regression Analysis and Other Multivariable Methods. Duxbury Press, Massachusetts.
  • Lesage, V., Hammill, M.O. & Kavacs, K.M. (2002) Diet-tissue fractionation of stable carbon and nitrogen isotopes in phocid seals. Marine Mammal Science, 18, 182193.
  • Martinez del Rio, C., Wolf, N., Carleton, S.A. & Gannes, L.Z. (2009) Isotopic ecology ten years after a call for more laboratory experiments. Biological Reviews, 84, 91111.
  • McCutchan, J.H., Jr, Lewis, W.M., Jr, Kendall, C. & McGrath, C.C. (2003) Variation in trophic shift for stable isotopic ratios of carbon, nitrogen, and sulfur. Oikos, 102, 378390.
  • Miron, M.L.L., Herrera, M.L.G., Ramirez, P.N. & Hobson, K.A. (2006) Effect of diet quality on carbon and nitrogen turnover and isotopic discrimination in blood of a New World nectarivorous bat. Journal of Experimental Biology, 209, 541548.
  • NRC (National Research Council) (1994) Nutrient Requirements of Poultry. National Academy of Sciences, Washington.
  • NRC (National Research Council) (1995) Nutrient Requirements of Laboratory Animals. National Academy of Sciences, Washington.
  • Ogden, L.J.E., Hobson, K.A. & Lank, D.B. (2004) Blood isotopic (δ13C and δ15N) turnover and diet-tissue fractionation factors in captive dunlin (Calidris alpine pacifica). The Auk, 121, 170177.
  • Pearson, S.F., Levey, D.J., Greenberg, C.H. & Martinez del Rio, C. (2003) Effects of elemental composition on the incorporation of dietary nitrogen and carbon isotopic signatures in an omnivorous songbird. Oecologia, 135, 516523.
  • Peterson, B.J. & Fry, B. (1987) Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics, 18, 293320.
  • Podlesak, D.W. & McWilliams, S.R. (2006) Metabolic routing of dietary nutrients in birds: effects of diet quality and macronutrient composition revealed using stable isotopes. Physiological and Biochemical Zoology, 79, 534549.
  • Robbins, C.T. (1993) Wildlife Feeding and Nutrition. Academic Press, New York.
  • Robbins, C.T., Felicetti, L.A. & Florin, S.T. (2010) The impact of protein quality on stable nitrogen isotope ratio discrimination and assimilated diet estimation. Oecologia, 162, 571579.
  • Robbins, C.T., Felicetti, L.A. & Sponheimer, M. (2005) The effect of dietary protein quality on nitrogen isotope discrimination in mammals and birds. Oecologia, 144, 534540.
  • Roth, J.D. & Hobson, K.A. (2000) Stable carbon and nitrogen isotopic fractionation between diet and tissue of captive red fox: implications for dietary consideration. Canadian Journal of Zoology, 78, 848852.
  • SAS (1998) SAS/SAT User’s Guide, Version 6.12. SAS Institute, Cary.
  • Smith, J.J., Millar, J.S., Longstaffe, F.J. & Boonstra, R. (2010) The effect of metabolic rate on stable carbon and nitrogen isotope compositions in deer mice, Peromyscus maniculatus. Canadian Journal of Zoology, 88, 3642.
  • Sponheimer, M., Robinson, T., Ayliffe, L., Roeder, B., Hammer, J., Passey, B., West, A., Cerling, T., Dearing, D. & Ehleringer, J. (2003) Nitrogen isotopes in mammalian herbivores: hair d15N values from a controlled feeding study. International Journal of Osteoarchaeology, 13, 8087.
  • van Tets, I.G. & Hulbert, A.J. (1999) A comparison of the nitrogen requirements of the Eastern pygmy possum, Cercartetus nanus, on a pollen and on a mealworm diet. Physiological and Biochemical Zoology, 72, 127137.
  • Tsahar, E., Wolf, N., Izhaki, I., Arad, Z. & Martinez del Rio, C. (2008) Dietary protein influences the rate of 15N incorporation in blood cells and plasma of Yellow-vented bulbuls (Pycnonotus xanthopygos). Journal of Experimental Biology, 211, 459465.
  • Vanderklift, M.A. & Ponsard, S. (2003) Sources of variation in consumer-diet δ15N enrichment: a meta-analysis. Oecologia, 136, 169182.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Table••S1. The essential amino acid requirement for growth by laboratory rats (NRC 1995) relative to the same amino acids occurring in various foods used in the current and previous isotope studies (Hilderbrand et••al. 1996; Darr & Hewitt 2008; Tsahar et••al. 2008; Robbins et••al. 2010). Both the requirement and dietary amino acid profile are expressed as a per cent of total dietary protein. The subscripts in the diet columns indicate the deficit occurring for each amino acid in each food relative to the requirement.

Table••S2. Estimated or measured dietary protein characteristics and nitrogen discrimination of whole blood, plasma, serum, or red blood cells for the diets fed to a wide range of mammals and birds. Protein quality estimates are from Table••S1 and based on laboratory rat amino acid requirements for growth. The numbers following several mixed diets (e.g. 75•:•25 or 50•:•50) refer to the relative contribution of protein by the two ingredients (Robbins et al. 2010; current study). The ‘diluted’ diets refer to the addition of sucrose, starch and corn oil to reduce protein content by 50% while holding protein quality constant.

Table••S3. Sulphur isotope values for the diets and plasma of laboratory rats and grizzly bears and the corresponding discrimination (Felicetti et••al. 2003; Robbins et••al. 2010; current study).

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
FEC_1799_sm_ST1.pdf14KSupporting info item
FEC_1799_sm_ST2.pdf62KSupporting info item
FEC_1799_sm_ST3.pdf8KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.