Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Isotopic ecology of coyotes from scat and road kill carcasses: A complementary approach to feeding experiments

Isotopic ecology of coyotes from scat and road kill carcasses: A complementary approach to feeding experiments

  • Rachel E. B. Reid, 
  • Paul L. Koch


Scat is frequently used to study animal diets because it is easy to find and collect, but one concern is that gross fecal analysis (GFA) techniques exaggerate the importance of small-bodied prey to mammalian mesopredator diets. To capitalize on the benefits of scat, we suggest the analysis of scat carbon and nitrogen isotope values (δ13C and δ15N). This technique offers researchers a non-invasive method to gather short-term dietary information. We conducted three interrelated studies to validate the use of isotopic values from coyote scat: 1) we determined tissue-to-tissue apparent C and N isotope enrichment factors (ε13* and ε15*) for coyotes from road kill animals (n = 4); 2) we derived diet-to-scat isotope discrimination factors for coyotes; and 3) we used field collected coyote scats (n = 12) to compare estimates of coyote dietary proportions from stable isotope mixing models with estimates from two GFA techniques. Scat consistently had the lowest δ13C and δ15N values among the tissues sampled. We derived a diet-to-scat Δ13C value of -1.5‰ ± 1.6‰ and Δ15N value of 2.3‰ ± 1.3‰ for coyotes. Coyote scat δ13C and δ15N values adjusted for discrimination consistently plot within the isotopic mixing space created by known dietary items. In comparison with GFA results, we found that mixing model estimates of coyote dietary proportions de-emphasize the importance of small-bodied prey. Coyote scat δ13C and δ15N values therefore offer a relatively quick and non-invasive way to gain accurate dietary information.


Scat is ubiquitous and easy to sample and has therefore historically formed our perception of mammalian carnivore and omnivore dietary ecology [1]. The dissection, identification, and quantification of the material contained in scats is labor intensive and time consuming and can be hampered by observer bias [2] as well as discrepancies among diet quantification methods [1]. Gross fecal analysis (GFA) techniques are also likely to severely overestimate the importance of small diet items [36]. Stable isotope analyses of scat have the potential to provide a quick and possibly more accurate means of gaining dietary information from scat while also allowing for non-invasive isotopic investigation of short-term dietary shifts that might otherwise be masked by long-term averaging in other animal tissues [714].

Because scat is composed of a combination of undigested food, sloughed epithelial cells, and microbiota [15,16], there is concern that the poorly digested food in scat will disproportionally affect its isotopic values. Scat has therefore largely been excluded from controlled feeding studies, which are often used to determine species- and tissue-specific discrimination factors (the difference in isotopic composition between a tissue, such as hair, and an animal’s diet), which are essential for the accurate determination of dietary sources. Scat is, however, an ideal substrate; it is abundant, turns over rapidly, and can be collected opportunistically without needing to capture the target animal. Both feeding studies [9,1720] and field tests [7,12,13,21,22] support the notion that scat stable isotope values do reflect ingested diet for many organisms. Most of these studies have focused on mammalian herbivores rather than omnivores or carnivores. To our knowledge, the only non-herbivorous mammals to have their scats characterized isotopically are bears [12], mountain gorillas [7], chimpanzees [23], bats [24], and two species of big cats [25].

Here, we determine isotopic differences among coyote tissues (hereafter termed apparent enrichment, ε*, defined below) and ground truth scat carbon and nitrogen isotope values (δ13C and δ15N) as dietary proxies for coyotes (Canis latrans), one of the most abundant and ecologically-impactful carnivores in North and Central America. As long as the difference in timescale over which disparate tissues integrate diet is considered, tissue-to-tissue apparent enrichment factors are valuable tools that allow for comparison among studies that rely on different tissues. For example, bone collagen is often the tissue of choice for studies using historical and archaeological materials because it preserves relatively well over long timescales. Modern studies, on the other hand, often focus on more easily and ethically sampled tissues, such as hair, whiskers, or blood. We took a novel approach, choosing to characterize isotopic differences among coyote tissues by analyzing materials from individual road kill carcasses. One advantage to this approach is that it enables sampling of tissues that are impossible to non-invasively collect from living animals (e.g., bone collagen). It is also less resource intensive than a feeding study on animals in a controlled setting. The downsides to our approach are that we are limited to the small number of individual animals that were collectible during our study period and, more challengingly, we do not have direct measurements of the diets of road kill animals. Therefore, in order to derive diet-to-scat δ13C and δ15N discrimination factors for coyotes, we rely on previous results from controlled feeding studies on two other wild canid species: red foxes [26] and wolves [27]. We tested our derived diet-to-scat discrimination factors by applying them to fully dissected coyote scats (DNA-verified to species) and compared them with isotope values measured in known dietary items. We hypothesize that the coyote scat matrix—the material that binds the scat together—primarily contains material derived from the coyote itself (e.g., epithelial cells) and, once corrected for discrimination, scat stable isotope values will fall within the isotopic mixing space created by known dietary items (Fig 1). Finally, we compare quantitative estimates of coyote dietary proportions derived from Bayesian stable isotope mixing models with estimates of diet composition from GFA to determine whether an isotopic approach mitigates some of the known biases in GFA. We present C and N isotope data from multiple tissues sampled from four road kill coyote carcasses and from seasonally collected coyote scats that are DNA-verified to species.

Fig 1. Scat stable isotope sampling rationale.

Step 1 is to separate the coyote scat—the fine-grained material binding the scat together—from the clearly undigested scat components. In step 2, the undigested materials are identified to the finest taxonomical level possible. In step 3, we conduct stable isotope analyses of both the coyote (scat matrix) and its known diet (identified undigested material). After correcting scat values for diet-to-scat discrimination, we expect that they should fall within the mixing space created by known dietary items.

Materials and methods

Sample collection

We used road kill carcasses for the determination of coyote tissue-to-tissue enrichment factors. We collected road kill coyotes under California Fish and Game permit SC-11995 to R. Reid. All work with animals followed the guidelines of the American Society of Mammalogists [28] and was conducted with the approval of both the UC Santa Cruz Institutional Animal Care and Use Committee (permits Kochp1105 and Kochp1211) and the Office of Environmental Health and Safety. Specimens were collected fresh and stored frozen until dissection, thereby preventing tissue deterioration.

We collected fresh coyote scats quarterly along a ~6 km coast-to-inland transect in 2011–2013 at Año Nuevo State Park and Reserve (San Mateo County, CA). As with the carcass samples, our scat collection followed all applicable institutional and national guidelines for the care and use of animals. Our field collection methods are described in full in [29]. Briefly, we cleared all scats from the transect one week prior to collecting all scats deposited in the intervening week. We recorded the scat locations with a GPS and stored the scats in the freezer in individually labeled Ziploc® bags. Scats deposited on the ground will experience weathering and decomposition at rates determined by both the local environment and the composition of the scat [30]. Previous research on wolf scats shows that scat DNA degrades more rapidly in samples that are in direct contact with soil, likely because they are kept wet, which may facilitate their invasion by decomposers (e.g., bacteria, fungi, insects) [31]. While plant δ13C and δ15N values have been shown to change over weeks to months during decomposition (e.g., [32,33]), given that we collected all scats within a week of their deposition, we argue that any potential changes in isotopic values associated with decomposition are not relevant to these samples.

Isotopic analyses

We sampled fur, muscle, bone, and feces from each animal carcass. We clipped a small tuft of fur from each animal's dorsum. Muscle tissue was consistently sampled from the right trapezius. We defleshed and removed a small chip of bone (~50 mg) for collagen extraction; if the carcass did not have easily sampled broken bones, we defaulted to a chip from the mandible. We sampled feces from what remained in the colon. Hair samples were rinsed with Milli-Q water, immersed in petroleum ether and sonicated for 15 minutes, rinsed again with Milli-Q water and dried in a 60°C oven overnight [34]. Our collagen extraction methods followed Brown et al. [35]. We decalcified bone fragments in 0.5N HCl for ~72 hours then followed with a 0.1N NaOH treatment for 24 hours to saponify fatty acids. We rinsed the samples five times in Milli-Q water, lipid extracted them in the same manner as the muscle tissue, rinsed them 5x with Milli-Q, and then freeze dried them overnight. Muscle samples were freeze-dried overnight and ground to a powder in an agate mortar and pestle. We lipid extracted muscle tissue and bone collagen by immersing the samples in 5 mL of petroleum ether [36] and sonicating for 15 min 2–3 times. Then, we rinsed 5 times in Milli-Q water before freeze drying the samples overnight. Muscle was re-homogenized and, like the collagen and hair samples, we weighed ~0.7 mg of material into 5 x 9-mm tin capsules. The atomic C/N ratio in the bone collagen samples fell between 2.8 and 3.4, which is within the range expected for modern collagen [37].

To prepare scat samples for isotopic analysis, we isolated the scat matrix—the material that binds the scat together—because we anticipated that this portion of the scat primarily contains material derived from the coyote itself (e.g., epithelial cells). By excluding poorly digested dietary components, we also prevent those items from having a disproportional effect on the scat stable isotope values. We extracted the scat matrix material by gently breaking apart oven-dried scats over a fine mesh sieve (0.420 mm). The matrix passes through the sieve while other scat components, such as fur, feathers, or bone are captured in the sieve. We then cleaned the matrix by placing it in filter paper cones and rinsing it first with Milli-Q water, then with 0.1N HCl to remove possible carbonate contaminants (e.g. bone fragments), and then again with Milli-Q. Because previous authors conjectured that rinsing with distilled or Milli-Q water following acidification may introduce bias in sample δ15N values [38,39], we characterized this possible bias by comparing 7 paired samples that were (a) rinsed following acidification or (b) not rinsed following acidification. We found no significant difference between the δ15N values of the two sample treatment groups (paired t-test: mean sample difference = -0.11‰, df = 6, p = 0.055; S1 Fig). After the scat samples were fully dry and homogenized, we weighed approximately 5 mg of scat matrix into 5 x 9-mm tin boats for isotopic analysis. Scats were identified to species with mitochondrial DNA [29].

To enable comparison between scat matrix and isotopic values of known diet components, we cleaned and analyzed a subset of identified scat components. We prepped hair and bone samples for isotopic analyses as described above. We prepped feather samples in the same way as hair. Arthropod, vegetation, and seed samples were repeatedly rinsed and sonicated in Milli-Q water (4x for 15 min), dried (60°C overnight), and then crushed with an agate mortar and pestle. We weighed ~0.7 mg of hair, feather, collagen, and arthropod samples into 5 x 9-mm tin boats. We divided vegetation samples into aliquots of ~0.4 mg for carbon isotopes and ~3 mg for nitrogen isotopes, and sealed them in 5 x 9-mm tin boats.

All samples were combusted via Dumas combustion using a Carlo Erba 1108 elemental analyzer and analyzed for δ13C and δ15N values on a ThermoFinnigan Delta Plus XP continuous-flow, isotope-ratio-monitoring mass spectrometer at the UC Santa Cruz Stable Isotope Laboratory. We report our results using δ notation, in which δHX = ((Rsample/Rstandard)– 1) x 1,000, where R is the ratio of the heavy isotope to the light isotope for element X. Carbon isotope values are reported relative to Vienna Pee Dee Belemnite (a marine carbonate) and nitrogen isotope values are reported relative to air, and the resulting value is expressed in parts per thousand (i.e., per mil, ‰). Sample isotopic values are corrected for size, drift and source stretching effects. The average analytical precision was <0.2‰ for both carbon and nitrogen, based on the SD of 41 replicates of an in-house standard (PUGel) and 20 replicates of a second in-house standard (Acetanilide). Atomic carbon and nitrogen elemental composition is estimated based on standards of known elemental composition (PUGel and Acetanilide) and precision of these known compounds is determined to better than 1%.

Often the offset, or fractionation, between two substances or tissues is expressed by Δ notation [40], in which ΔHXa-b = δHXa−δHXb. Though Δ values are relatively simple to calculate, they become less accurate with increasing differences between the δ values of the substances of interest [41,42]. Because scat isotope values have the potential to be quite different from other tissues, following Passey et al. [43] and Crowley et al. [41], for tissue-to-tissue differences we report the isotope enrichment values (ε) derived from the fractionation factor (α): αa-b = (δHXa + 1,000)/ (δHXb + 1,000) and εa-b = (αa-b− 1) x 1,000. Furthermore, we use the notation ε*, the apparent enrichment value, to denote that this is a non-equilibrium fractionation factor. It is important to note that the sign of enrichment depends on the tissue or substrate in the numerator when calculating α. Scat isotope values are not likely to be considerably different from diet values, however, so we follow the convention in the more recent literature and report “diet-to-scat” discrimination factors as ΔHX = δHXscat−δHXdiet (e.g., [25,44]).

Gross fecal analysis

Following matrix extraction, we placed each scat in a bag made from nylon panty hose and washed it in a portable automatic washing machine (Haier Compact HLP21N; purchased expressly for this purpose) [1] without detergent to remove any residual matrix and to better separate the remaining components. Once dry, we placed the scat contents on a gridded sorting tray to estimate the percent by volume contributions of mammal, bird, reptile, invertebrate, and plant components to the nearest 5% [45]. While the fur in each scat was spread out, we sampled guard hairs from the center of each grid cell until we had examined 40–50 hairs and identified them to the finest taxonomic level possible through comparison with a guard hair reference collection housed at UC Santa Cruz and with published keys [4648]. We grouped identified hairs into four categories: marine mammals, small terrestrial mammals (≤ 1 kg), medium terrestrial mammals (> 1 kg, < 30 kg), and large terrestrial mammals (≥ 30 kg). To facilitate comparison with other studies we also calculated the frequency of occurrence of prey taxa as percentage of occurrence = (number of occurrences of prey type/total number of occurrence) x100 [49].

Data analysis

We used Stable Isotope Analysis in R (SIAR, package ‘simmr’) [50], a Bayesian stable isotope mixing model, to estimate the proportional contributions of various scat components to coyote diets. SIAR is capable of accounting for error in estimates of discrimination factors as well as for variations in the elemental concentrations of C and N in the food sources, which could otherwise bias model output [51]. We derived digestible [C] and [N] values for various coyote food sources through the USDA nutrient database as described by Koch and Phillips [52] (S1 Table). We converted stable isotope values measured in identifiable scat components (S1 Table) to values for muscle tissue, which makes up the bulk of assimilated diet, by applying published organism- and tissue-specific discrimination factors (S2 Table). Using these values, we ran the mixing models both for each scat individually as well as for all the scat samples collectively. Scats were adjusted for trophic discrimination by adding 1.5‰ ± 1.6‰ for δ13C values and subtracting 2.3‰ ± 1.3‰ for δ15N values (details regarding our arrival at these specific diet-to-scat discrimination factors are explained in the Results and Discussion).

The coyote dietary categories outlined above (bird, invertebrate, etc.) were not necessarily isotopically distinct or homogenous. For example, our subset of scat samples contained feathers from one bird feeding primarily on marine resources and from another feeding primarily on terrestrial resources. For the purposes of the mixing model, rather than combining these isotopically distinct birds into one source (“birds”), we kept them as separate source inputs (“marine bird” and “terrestrial bird”). To enable comparison among the mixing model predictions and those derived from the two GFA techniques, we recast the more numerous mixing model source categories into the eight organism-based dietary categories described above by combining the model predicted mean values. We performed all statistical analyses in R [53].


Apparent isotope enrichment factors

Carbon isotope values in coyote tissues increased consistently from scat, to muscle, to hair, to bone collagen in the four coyote carcasses we examined (Fig 2, Table 1). Nitrogen isotope values followed roughly the same trend, though the magnitude of change was somewhat less among tissues from a single individual. In addition, muscle δ15N values were equal to or slightly higher than hair values. Mean ε15* values between proteinaceous tissues were consistently small (≤0.3‰; Table 2), whereas mean ε13* values between proteinaceous tissues ranged from -1.0 to 2.3‰. Mean ε13* and ε15* values between collagen and scat (4.3 ± 2.3‰, 0.9 ± 1.9‰, respectively) as well as between hair and scat (4.1 ± 1.5‰, 0.9 ± 1.3‰, respectively) were higher than ε13* and ε15* values between proteinaceous tissues (Table 2).

Fig 2. δ13C and δ15N values measured in tissues of 4 road kill coyotes.

Symbol shades denote type of tissue sampled. Tissues from the same individual are connected by lines with different dash patterns.

Table 2. Apparent C and N isotope enrichment factors among sampled tissues of road kill coyote carcasses.

Given that we chose to work with road kill carcasses, we were not able to directly measure the diets of the individual animals we examined. Therefore, we derived diet-to-scat Δ13C and Δ15N values for coyotes by relying on our hair-to-scat enrichment factors combined with diet-to-hair discrimination factors determined for red foxes [26] and for wolves [27] (Fig 3). We calculated a diet-to-scat Δ13C value of -1.5 ± 1.6‰ and a diet-to-scat Δ15N value of 2.3 ± 1.3‰ for coyotes using red fox diet-to-hair discrimination factors from Roth and Hobson [26]. Using muscle-to-scat enrichment factors combined with Roth and Hobson’s [26] diet-to-muscle discrimination factors resulted in the same diet-to-scat discrimination factors for coyotes within error (Δ13C = -0.9 ± 0.8‰; Δ15N = 2.5 ± 1.3‰). If we instead relied on recently published diet-to-hair fractionation factors for wolves [27], we arrived at a diet-to-scat Δ13C value of 0.1 ± 1.6‰ and a diet-to-scat Δ15N value of 2.2 ± 1.3‰. We report the diet-to-scat Δ13C and Δ15N values for coyotes derived through red fox hair (Δ13C = -1.5 ± 1.6‰; Δ15N = 2.3 ± 1.3‰), as opposed to wolf hair, because these values most consistently place coyote scats within the mixing space provided by known dietary items. Furthermore, we argue that as omnivores, coyote digestive physiology is likely to be more like that of omnivorous red foxes than carnivorous wolves.

Fig 3. Illustration of derived diet-to-scat C and N isotope discrimination factors for coyotes.

The routes through diet-to-hair discrimination factors for red foxes (gray arrows) and through wolves (dashed gray arrows) are depicted. In both cases, the first step used our scat-to-hair ε13* and ε15* values (4.1 ± 1.5‰, 0.9 ± 1.3‰, respectively) to convert scat to hair; the dark gray oval surrounding the hair point depicts 1 SD around the enrichment factors. In step 2, we used published diet-to-hair enrichment factors for red foxes (C: 2.6 ± 0.4‰, N: 3.2 ± 0.3‰; [26]) and wolves (C: 4.25 ± 0.4‰, N: 3.1 ± 0.2‰; [27]) to convert hair to diet; gray error oval around the diet points depict the propagated standard deviation. Finally, in step 3, we calculated the values necessary to convert from diet to scat.

Gross fecal analysis

We fully dissected a subset of 12 DNA-verified coyote scats, collected in two different seasons (spring and fall). Because coyote scats are not morphologically distinguishable from other mammalian mesopredator scats in the ecosystem, we previously found DNA-verification to species to be necessary [29]. We identified 25 different dietary components in these scats, including, for example, grass, black-tailed deer (Odocoileus hemionus), California sea lion (Zalophus californianus), and rabbits (Sylvilagus bachmani). The diet estimates derived from the different methods for quantifying coyote diet from scat are similar, but as expected, not identical (Fig 4). Collectively, terrestrial mammals had the highest frequency of occurrence in the scats (43%), with small mammals making up the bulk of that number (27%). These dietary sources were followed by various forms of vegetation (23%) and then marine mammals (12%). Birds, sand/gravel, invertebrates, and reptiles make up the remaining 22%. The percent by volume method similarly identified terrestrial mammals—represented in scat by both fur and bone—as the most frequently occurring coyote scat component (54 ± 13%). Of these, small terrestrial mammals were once again of greatest importance (24 ± 10%). Terrestrial mammals were followed in prevalence by marine mammals (22 ± 8%) and vegetation (13 ± 7%) and the remaining 13% is comprised of sand/gravel, invertebrates, birds, and reptiles. Once adjusted for trophic discrimination, scat matrix stable isotope values fell consistently within the isotope mixing space created by the known dietary components found in them, with two exceptions (samples 091011AN008 and 111411ANNU7) (Fig 5). The aggregated mixing model predictions diverged slightly from the GFA techniques; looking at the mean of the individual model predictions, we saw that terrestrial mammals still contributed the most to coyote diets (47 ± 13%), but small terrestrial mammals accounted for a smaller proportion that approaches significance (12 ± 7% vs. 24 ± 10% in percent by volume; paired t-test: df = 11, p = 0.057). Marine mammals also accounted for a higher percentage (28 ± 11%) than predicted by the two GFA methods (12% and 22 ± 8%), though not significantly higher (paired t-test: df = 11, p = 0.3).

Fig 4. Comparison of diet quantification methods.

Comparison of the proportional contributions of marine mammals, terrestrial mammals (small, medium and large), birds, reptiles, invertebrates, plants, fish, and non-food material (e.g., gravel/sand) to 12 DNA-verified coyote scats as identified by three methods: frequency of occurrence (white), percent by volume (light gray) and isotopic mixing models (dark gray). Error bars depict one standard error.

Fig 5. Scat isotope results.

Carbon and nitrogen isotope values (δ13C and δ15N) measured in twelve coyote scats (corrected for discrimination; open circles) from Año Nuevo State Park, CA, plotted in reference to isotope values measured in dietary components found in the scat.


Carbon isotope values

Although the carbon in animal tissues is supplied by diet, there are still significant differences in δ13C values among tissues [54]. With the exception of Coy4, we found collagen to have the highest δ13C values of the tissues we examined, similar to previous studies (e.g., [8,10,55]). Scat, on the other hand, consistently had the lowest δ13C values. There was considerable variability in the ε13* values between coyote tissues, with values for Coy1 and Coy4 deviating, sometimes in opposite directions, from those for Coy2 and Coy3. Some of this variability may be accounted for by the nutritional state of the individuals we sampled, though none of the carcasses sampled displayed any obvious outward signs of malnutrition or poor health. It has also been previously documented that carbon isotope discrimination among tissues can increase when the δ13C values of the dietary carbon sources are variable [56], suggesting that the variability we observe here could in part be driven by the diets these coyotes were consuming. Our sample size of 4 road kill individuals is not likely to be fully representative of the local wild population, however it is comparable to the number of individuals used in many controlled feeding studies [e.g., 17,24,44].

Our estimated diet-to-scat Δ13C value for coyotes (-1.5 ± 1.6‰) is of greater magnitude than the experimentally derived diet-to-scat discrimination reported for large mammalian herbivores (-0.8‰; [9]), but consistent in the direction of offset. A previous study found that herbivore scat contained more 13C enriched plant acid-detergent fibers than diet items and hypothesized a positive diet-to-feces C isotope discrimination factor [9]. It may be that the microfloral component of scat is the source of the low δ13C values, but Sponheimer et al. [9] did not observe an increase in scat δ13C following their removal from the sampled feces. Lipids are another known source of light carbon [54], but canids tend to use lipids quite efficiently. For example, Coffey et al. [57] found that fecal excretion of fat by healthy dogs varied from just 2 to 4% and was mostly composed of fatty acids. This could, however, vary significantly with the diet and nutritional state of the animal. Bulk diet-to-lipid discrimination can be quite large (e.g., gerbils = -3‰ [58]; striped skunks = -3.3‰ [59]), and given that scat δ13C values are generally lower than diet, but not as low as pure lipids, a small proportion of fatty acids in scat may be enough to account for low scat δ13C values. Tiger and snow leopard scat samples that were lipid-extracted prior to analysis were enriched in 13C relative to diet [25], strongly suggesting that lipids are responsible for the low δ13C values in carnivore scats, though future work on scat δ13C values is necessary to definitively resolve the source of the light carbon.

Nitrogen isotope values

Irrespective of tissue type, δ15N values increase with trophic level. This pattern is broadly attributed to the combined effects of (1) higher dietary protein intake with increasing trophic level and (2) the preferential excretion of 14N in urea, the main efflux of nitrogen in mammals, resulting in a body pool that is enriched in 15N relative to diet [60,61]. Modeling efforts suggest that nitrogen recycling is in part responsible for the 15N-enrichment in animal tissues [62]. Nitrogen metabolism is extraordinarily complex [63] and studies continue to demonstrate the intricate connection between isotopic ecology and nutritional ecology (e.g., [64]). A number of controlled feeding studies indicate the magnitude of the 15N trophic enrichment correlates with dietary protein content [15,6567], particularly in animals with high rates of N excretion relative to N assimilation. Other recent work has highlighted the importance of protein quality, suggesting that increasing dietary protein quality corresponds with reduced nitrogen discrimination [68]. In support of this idea, recent studies on fish and birds have shown that as the imbalance in amino acid composition between a consumer and its diet drops (i.e., as diet quality rises), amino acid trophic discrimination factors (which drive bulk diet-to-tissue fractionation) decrease [69,70].

Physiologists recognize two types of "waste" nitrogen lost by vertebrates: endogenous urinary nitrogen (EUN) and metabolic fecal nitrogen (MFN) [64]. EUN is primarily composed of nitrogenous waste products, such as ammonia, uric acid, urea, and creatinine [71]. MFN, on the other hand, is made up of non-absorbed digestive enzymes, intestinal cellular debris, and undigested bacteria and mucus [71]. It is difficult to allocate the contributions of these components to feces as a whole, though previous work suggests that most fecal nitrogen is derived from sloughed endogenous tissues and microbial cells for herbivores [72]. Unlike EUN, fecal δ15N values are consistently enriched in 15N relative to diet [13,15,17,21], like other tissues. Fecal δ15N values elevated above those of diet suggest that the bulk of fecal nitrogen is from the animal itself, rather than undigested food. This hypothesis is supported by an observational study of captive mammalian herbivores, in which the authors found that 60–80% of herbivore fecal nitrogen was endogenous [73]. Omnivores and carnivores, however, do not rely on N-poor plants for protein and their differing digestive physiology may result in differing proportions of endogenous nitrogen in feces. Previously reported diet-to-scat Δ15N values are quite variable (S3 Table), ranging from as little as 0.8‰ for cows fed a variety of diets [74] to 3.1‰ for sheep fed alfalfa hay cubes [75]. For six different small mammal species kept on an artificial laboratory diet, diet-to-scat Δ15N values varied from 1.4 to 2.6 [44]. These small mammal species (including mice, voles, and a chipmunk; listed in full in S3 Table) have potentially quite different and variable wild diets, however, so it is unlikely that the standard rodent chow used in that study matched the composition of wild diets for all six of the species [56]. Experimentally derived diet-to-scat discrimination factors for two big cats [25] are comparable in direction and magnitude to our results for coyotes. In light of the observed negative relationship between trophic discrimination factors and diet quality [69,70], we hypothesized that diet-to-scat Δ15N values would inversely scale with dietary nitrogen intake and therefore would be smallest in carnivores, intermediate in omnivores and greatest in herbivores. The published diet-to-scat Δ15N data do not, however, exhibit a significant pattern in terms of trophic level (S2 Fig), either in line with our hypothesis or in contradiction to it. Some of the spread in diet-to-scat Δ15N values within trophic groups could be derived from the fact that in many controlled feeding experiments, the animals are allowed insufficient time to come into equilibration with their diets or perhaps because their artificial diets are not of the same quality and/or composition as they are in the wild [56]. Nonetheless, future controlled feeding studies will likely be most effective for elucidating the relationship among dietary protein quantity, quality, and nitrogen discrimination factors.

Given that δ15N values increase with trophic level, we anticipated scat matrix δ15N values to positively correlate with the amount of protein (meat) predicted to be in the coyote diets by GFA and the isotopic mixing models. While true, the correlations between scat δ15N values and the proportion of meat in the diet are weak and not statistically significant (using three different measures of proportion of meat: frequency of animal occurrence [r = 0.18, p = 0.57], % volume animal consumed [r = 0.16, p = 0.62], and mean proportion animal consumed [r = 0.32, p = 0.31; S3 Fig). The same is true for black bear scats [12]. Scat C:N ratios could alternatively be an indicator of an animal’s degree of carnivory, as C:N ratios tend to be high in plants [76,77] and low in animals [78]; because even the fine-grained scat matrix contains food waste in addition to coyote-derived material, animals consuming a largely plant-based diet might produce scats with high C:N ratios and those consuming other animals would then produce scats with lower C:N ratios. We found that scat C:N ratios are significantly negatively correlated with the percent volume of animal material found in the scat (r = -0.71, p < 0.01) and also negatively correlated (though not significantly) with the mixing model predicted mean proportion of meat in the coyote diets (r = -0.36, p = 0.25; S3 Fig). This lack of a significant relationship lends further support to the idea that the bulk of the scat matrix is composed of material derived from the coyote itself rather than from its food.

Determining coyote diet from scat

Scats likely capture a short window of food consumption. The average gut retention time for coyotes is on the order of a few days [79], though the incorporation rate of the epithelial cells found in the scat matrix may be longer. Carbon isotope turnover in goat feces following a diet switch was evident in just 2–3 days, however it took 60 days for individuals to reach equilibrium with their diets [80]. At these timescales, stable isotope analyses of scats are particularly useful when working with an organism for which seasonal dietary shifts are important [7,12]. There are other tissues that turn over relatively rapidly (e.g., plasma, breath, ever-growing hair), but their sampling requires physical contact with the animal. Scats provide a non-invasive way to gather short-term dietary information. Furthermore, scats can be linked to individuals either through direct observation (as demonstrated by [7]) or potentially through nuclear DNA analyses [81,82], making it possible to non-invasively monitor individual dietary preferences over time.

The method used to quantify diets from scats is particularly important for omnivores, such as coyotes and foxes; the less uniform an animal’s diet, the larger the disagreement among different methods [1]. Frequency of occurrence tends to over emphasize the importance of small food items in our data—small mammals were overwhelmingly identified as the most important—this result is likely because there are more indigestible parts per unit biomass for small mammals than for larger ones [1,83]. The percent by volume method mitigates small mammal inflation to some degree, but the results are overall quite similar to frequency of occurrence. The stable isotope mixing model results, however, identify larger bodied organisms, such as marine mammals and deer, as more important dietary components. Diet estimates based on biomass calculations also similarly address this bias by placing greater emphasis on larger bodied organisms than frequency of occurrence estimates [1]. While there are fewer indigestible hard parts from large animals in the scat, a substantial proportion of assimilated diet is coming from these animals. It makes sense that mesocarnivores would largely avoid the bones of deer- to pinniped-sized mammals, with the possible exception of fawns and neonate pups, when plenty of more easily consumed and high nutrient soft tissues are available. The stable isotope mixing models also de-emphasize the importance of plant matter to coyote diet. Grass accounted for most of the vegetation identified in these scats. While grass is frequently found in coyote scat, there is no consensus on its role as a food resource; some researchers suggest incidental ingestion while coyotes are capturing prey [84], while others argue it may be a necessary source of vitamins [85]. Our stable isotope mixing model results suggest that grass is less important to assimilated diet, and therefore lends some support to the idea of incidental consumption or some other non-nutritional explanation.

The strength of the stable isotope approach is that scat matrix isotope values provide a faster and less size-biased quantitative estimate of assimilated diet than most methods that rely on quantifying purely undigested material. However, a potential pitfall rests on the degree of isotopic variation in the system; problems will arise if dietary sources are indistinguishable from one another in isotopic space. In the system examined here, a marine resource is one of the major dietary components. Marine systems tend to have much higher carbon and nitrogen isotope values than terrestrial systems. Other sources of variation could come from anthropogenic food sources, which are often C4 labeled, or a more diverse flora containing both C3 and C4 plants. Regardless, questions about resource use may need to be recast, as traditional organism-based dietary categories may not correspond well with isotopic categories. A well-characterized local isotopic baseline will be critical for the interpretation of scat stable isotope values.

We have shown that coyote scat matrix carbon and nitrogen isotope values can serve as proxies for coyote diets. Our derived diet-to-scat discrimination factors are the best approximation available for a wild, mammalian omnivore on a natural diet and better validated than simply applying diet-to-feces discrimination factors observed for herbivores, as done previously [21]. Coyote scat matrix δ13C and δ15N values consistently plot within the isotopic mixing space created by known dietary sources, suggesting that these discrimination factors are appropriate. Stable isotope mixing model estimates of dietary proportions are complementary, though not identical, to estimates from GFA. Given that the stable isotope mixing model estimates placed greater emphasis on larger-bodied prey items, just as previous authors have noted for biomass calculations, these data suggest that scat stable isotopes provide less size-biased estimates of diet than GFA (given sufficient variation in the ecosystem). Finally, the tissue-to-tissue apparent enrichment factors determined in this study can be applied to wild animals and fossil organisms and will facilitate comparison among isotopic studies performed on a variety of tissue types.

Supporting information

S1 Fig. Test of the effects of rinsing following acidification.

δ15N values measured in split scat samples that were either rinsed (dark grey) or not rinsed (light grey) after acidification. The mean sample difference is -0.1‰, which is indistinguishable from instrumental error (± 0.1‰).


S2 Fig. Δ15N by diet type.

Boxplots depicting nitrogen discrimination factors for four different types of mammalian feeders: herbivores, omnivores, carnivores and insectivores.


S3 Fig. Correlations by dietary proxy.

Correlations between scat C:N ratio, δ15N and δ13C and frequency of animal material occurrence (left column), percent volume animal material (middle column) and mean proportion meat consumed (right column).


S1 Table. Mixing model input.

Isotope values and digestible [C] and [N] values measured in items identified in each scat sample for input into the mixing models. Food source δ13C and δ15N values are converted to coyote diet space using discrimination factors listed Table S2.


S2 Table. Coyote diet-space corrections.

Organism- and tissue-specific isotope discrimination factors applied to coyote food source δ13C and δ15N values before input into the SIAR mixing model.


S3 Table. Published diet-to-feces Δ15N values in mammals.



We thank R. Baldwin, J. Nye, M. Miller and N. Kronick for assistance with carcass dissections; D. Andreasen for assistance with stable isotope analyses; S. Shane, M. Brown, B. Cohen, N. Kronick, M. Miller, D. Perkins, and L. Reid for assisting in the field; S. Shane, D. Perkins and T. Lambert for locating coyote carcasses. We are additionally grateful to Año Nuevo State Park and Año Nuevo Reserve for allowing scat collection and to the rangers and reserve staff for providing logistical support. Thanks also to S. Kim for providing feedback on a previous draft of the manuscript and to two anonymous reviewers for their constructive criticism.

Author Contributions

  1. Conceptualization: REBR PLK.
  2. Formal analysis: REBR.
  3. Funding acquisition: REBR PLK.
  4. Investigation: REBR.
  5. Resources: PLK.
  6. Visualization: REBR.
  7. Writing – original draft: REBR.
  8. Writing – review & editing: REBR PLK.


  1. 1. Klare U, Kamler JF, Macdonald DW. A comparison and critique of different scat-analysis methods for determining carnivore diet. Mammal Rev. 2011;41: 294–312.
  2. 2. Spaulding R, Krausman PR, Ballard WB. Observer bias and analysis of gray wolf diets from scats. Wildlife Society Bulletin. 2000;28: 947–950.
  3. 3. Reynolds JC, Aebischer NJ. Comparison and quantification of carnivore diet by faecal analysis: a critique, with recommendations, based on a study of the fox Vulpes vulpes. Mammal Rev. 1991;21: 97–122.
  4. 4. Weaver JL, Hoffman SW. Differential detectability of rodents in coyote scats. Journal of Wildlife Management. 1979;43: 783–786.
  5. 5. Meriwether D, Johnson MK. Prey digestibility by coyotes. Journal of Mammalogy. 1980;61: 774–775.
  6. 6. Kelly BT, Garton E. Effects of prey size, meal size, meal composition, and daily frequency of feeding on the recovery of rodent remains from carnivore scats. Can J Zool. 1997;75: 1811–1817.
  7. 7. Blumenthal SA, Chritz KL, Rothman JM, Cerling TE. Detecting intra-annual dietary variability in wild mountain gorillas by stable isotope analysis of feces. Proceedings of the National Academy of Sciences. 2012;109: 21277–21282.
  8. 8. Kelly J. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Can J Zool. 2000;78: 1–27.
  9. 9. Sponheimer M, Robinson T, Ayliffe L, Passey B, Roeder B, Shipley L, et al. An experimental study of carbon-isotope fractionation between diet, hair, and feces of mammalian herbivores. Can J Zool. 2003;81: 871–876.
  10. 10. Koch PL. Isotopic study of the biology of modern and fossil vertebrates. In: Michener R, Lajtha K, editors. Stable Isotopes in Ecology and Environmental Science, 2nd edition. Blackwell Publishing, Boston; 2007. pp. 99–154.
  11. 11. Codron D, Codron J, Lee-Thorp JA, Sponheimer M, De Ruiter D, Sealy J, et al. Diets of savanna ungulates from stable carbon isotope composition of faeces. J Zool. 2007;273: 21–29.
  12. 12. Hatch KA, Roeder BL, Buckman RS, Gale BH, Bunnell ST, Eggett DL, et al. Isotopic and gross fecal analysis of American black bear scats. Ursus. 2011;22: 133–140.
  13. 13. Codron D, Codron J. Reliability of δ13C and δ15N in faeces for reconstructing savanna herbivore diet. Mammalian Biology. 2009;74: 36–48.
  14. 14. Botha MS, Stock WD. Stable isotope composition of faeces as an indicator of seasonal diet selection in wild herbivores in southern Africa: Research letter. South African Journal of Science. Sabinet Online; 2005;101: p. 371–374.
  15. 15. Sponheimer M, Robinson T, Ayliffe L, Roeder B, Hammer J, Passey B, et al. Nitrogen isotopes in mammalian herbivores: hair δ15N values from a controlled feeding study. Int J Osteoarchaeol. 2003;13: 80–87.
  16. 16. Putman RJ. Facts from faeces. Mammal Rev. Wiley Online Library; 1984;14: 79–97.
  17. 17. Sponheimer M, Robinson TF, Roeder BL, Passey BH, Ayliffe LK, Cerling TE, et al. An experimental study of nitrogen flux in llamas: is 14N preferentially excreted? J Archaeol Sci. 2003;30: 1649–1655.
  18. 18. Coates DB, Van Der Weide APA, Kerr JD. Changes in faecal δ13C in response to changing proportions of legume (C3) and grass (C4) in the diet of sheep and cattle. J Agric Sci. 1991;116: 287.
  19. 19. Wittmer MHOM, Auerswald K, Schönbach P, Schäufele R, Müller K, Yang H, et al. Do grazer hair and faeces reflect the carbon isotope composition of semi-arid C3/C4 grassland? Basic and Applied Ecology. 2010;11: 83–92.
  20. 20. Varo N, Amat JA. Differences in food assimilation between two coot species assessed with stable isotopes and particle size in faeces: linking physiology and conservation. Comp Biochem Physiol, Part A Mol Integr Physiol. 2008;149: 217–223. pmid:18222102
  21. 21. Codron J, Codron D, Lee-Thorp JA, Sponheimer M, De Ruiter D. Animal diets in the Waterberg based on stable isotopic composition of faeces. South African Journal of Wildlife Research. Sabinet Online; 2005;35: p. 43–52.
  22. 22. Codron D, Lee-Thorp JA, Sponheimer M, de Ruiter D, Codron J. Inter- and intrahabitat dietary variability of chacma baboons (Papio ursinus) in South African savannas based on fecal δ13C, δ15N, and %N. American Journal of Physical Anthropology. 2006;129: 204–214. pmid:16247809
  23. 23. Tsutaya T, Fujimori Y, Hayashi M, Yoneda M, Miyabe-Nishiwaki T. Carbon and nitrogen stable isotopic offsets between diet and hair/feces in captive chimpanzees. Rapid Commun Mass Spectrom. 2016.
  24. 24. Salvarina I, Yohannes E, Siemers BM, Koselj K. Advantages of using fecal samples for stable isotope analysis in bats: evidence from a triple isotopic experiment. Rapid Commun Mass Spectrom. 2013;27: 1945–1953. pmid:23939961
  25. 25. Montanari S, Amato G. Discrimination factors of carbon and nitrogen stable isotopes from diet to hair and scat in captive tigers (Panthera tigris) and snow leopards (Uncia uncia). Rapid Commun Mass Spectrom. 2015;29: 1062–1068. pmid:26044274
  26. 26. Roth J, Hobson K. Stable carbon and nitrogen isotopic fractionation between diet and tissue of captive red fox: implications for dietary reconstruction. Can J Zool. 2000;78: 848–852.
  27. 27. McLaren AAD, Crawshaw GJ, Patterson BR. Carbon and nitrogen discrimination factors of wolves and accuracy of diet inferences using stable isotope analysis. Wildlife Society Bulletin. 2015;39: 788–796.
  28. 28. Sikes RS, Gannon WL. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy. 2011;92: 235–253.
  29. 29. Reid REB. A morphometric modeling approach to distinguishing among bobcat, coyote and gray fox scats. Wildlife Biology. 2015;21: 254–262.
  30. 30. Godbois IA, Conner M, Leopold BD, Warren RJ. Effect of diet on mass loss of bobcat scat after exposure to field conditions. Wildlife Society Bulletin. 2005;33: 149–153.
  31. 31. Santini A, Lucchini V, Fabbri E, Randi E. Ageing and environmental factors affect PCR success in wolf (Canis lupus) excremental DNA samples. Mol Ecol Notes. 2007;7: 955–961.
  32. 32. Fourqurean JW, Schrlau JE. Changes in nutrient content and stable isotope ratios of C and N during decomposition of seagrasses and mangrove leaves along a nutrient availability gradient in Florida Bay, USA. Chemistry and Ecology. 2003;19: 373–390.
  33. 33. Wedin DA, Tieszen LL, Dewey B, Pastor J. Carbon isotope dynamics during grass decomposition and soil organic matter formation. Ecology. 1995;76: 1383–1392.
  34. 34. O'Connell TC, Hedges RE. Investigations into the effect of diet on modern human hair isotopic values. American Journal of Physical Anthropology. 1999;108: 409–425. pmid:10229386
  35. 35. Brown T, Nelson D, Vogel J, Southon J. Improved collagen extraction by modified longin method. Radiocarbon. 1988;30: 171–177.
  36. 36. Dobush GR, Ankney CD, Krementz DG. The effect of apparatus, extraction time, and solvent type on lipid extractions of snow geese. Can J Zool. 1985;63: 1917–1920.
  37. 37. Ambrose SH. Preparation and characterization of bone and tooth collagen for isotopic analysis. J Archaeol Sci. 1990;17: 431–451.
  38. 38. Jacob U, Mintenbeck K, Brey T, Knust R, Beyer K. Stable isotope food web studies: a case for standardized sample treatment. Mar Ecol-Prog Ser. 2005;287: 251–253.
  39. 39. Bosley KL, Wainright SC. Effects of preservatives and acidification on the stable isotope ratios (15N:14N, 13C:12C) of two species of marine animals. Can J Fish Aquat Sci. NRC Research Press; 1999;56: 2181–2185.
  40. 40. Martínez del Rio C, Wolf N, Carleton SA, Gannes LZ. Isotopic ecology ten years after a call for more laboratory experiments. 2009;84: 91–111.
  41. 41. Crowley BE, Carter ML, Karpanty SM, Zihlman AL, Koch PL, Dominy NJ. Stable carbon and nitrogen isotope enrichment in primate tissues. Oecologia. 2010;164: 611–626. pmid:20628886
  42. 42. Cerling TE, Harris JM. Carbon isotope fractionation between diet and bioapatite in ungulate mammals and implications for ecological and paleoecological studies. Oecologia. Springer; 1999;120: 347–363.
  43. 43. Passey BH, Robinson TF, Ayliffe LK, Cerling TE, Sponheimer M, Dearing MD, et al. Carbon isotope fractionation between diet, breath CO2, and bioapatite in different mammals. J Archaeol Sci. 2005;32: 1459–1470.
  44. 44. Hwang Y, Millar J, Longstaffe F. Do δ15N and δ13C values of feces reflect the isotopic composition of diets in small mammals? Can J Zool. 2007;85: 388–396.
  45. 45. McDonald JE Jr., Fuller TK. Effects of spring acorn availability on black bear diet, milk composition, and cub survival. Journal of Mammalogy. 2005;86: 1022–1028.
  46. 46. Mayer WV. The hair of California mammals with keys to the dorsal guard hairs of California mammals. American Midland Naturalist. 1952;48: 480–512.
  47. 47. Tumlison R. An annotated key to the dorsal guard hairs of Arkansas game mammals and furbearers. The Southwestern Naturalist. 1983;28: 315–323.
  48. 48. Debelica A, Thies ML. Atlas and key to the hair of terrestrial Texas mammals. Baker RJ, editor. Lubbock, TX: Museum of Texas Tech University; 2009. pp. 1–108.
  49. 49. Fedriani JM, Fuller TK, Sauvajot RM. Does availability of anthropogenic food enhance densities of omnivorous mammals? An example with coyotes in Southern California. Ecography. 2001;24: 325–331.
  50. 50. Parnell AC, Inger R, Bearhop S, Jackson AL, Rands S. Source partitioning using stable isotopes: coping with too much variation. PLoS ONE. 2010;5: e9672. pmid:20300637
  51. 51. Phillips D, Koch P. Incorporating concentration dependence in stable isotope mixing models. Oecologia. 2002;130: 114–125.
  52. 52. Koch PL, Phillips DL. Incorporating concentration dependence in stable isotope mixing models: a reply to Robbins, Hilderbrand and Farley (2002). Oecologia. 2002;133: 14–18. pmid:24599364
  53. 53. Team RC. R: A language and environment for statistical computing 2015 ed. Vienna, Austria: R Foundation for Statistical Computing.
  54. 54. DeNiro MJ, Epstein S. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta. 1978;42: 495–506.
  55. 55. Schoeninger MJ, DeNiro MJ. Nitrogen and carbon isotopic composition of bone collagen from marine and terrestrial animals. Geochimica et Cosmochimica Acta. 1984;48: 625–639.
  56. 56. Kurle CM, Koch PL, Tershy BR, Croll DA. The effects of sex, tissue type, and dietary components on stable isotope discrimination factors (Δ13C and Δ15N) in mammalian omnivores. Isotopes in Environmental and Health Studies. 2014;: 1–15.
  57. 57. Coffey RJ, Mann FC, Bollman JL. I. Fecal residue of fat, protein and carbohydrate in the normal dog. The American Journal of Digestive Diseases. Springer; 1940;7: 141–143.
  58. 58. Tieszen L, Boutton TW, G TK, Slade NA. Fractionation and turnover of stable carbon isotopes in animal tissues: implications for δ13C analysis of diet. Oecologia. 1983;57: 32–37. pmid:28310153
  59. 59. Hobson KA, Quirk TW. Effect of age and ration on diet-tissue isotopic (Δ13C, Δ15N) discrimination in striped skunks (Mephitis mephitis) †. Isotopes in Environmental and Health Studies. 2014;: 1–7.
  60. 60. Ambrose SH. Effects of diet, climate and physiology on nitrogen isotope abundances in terrestrial foodwebs. J Archaeol Sci. Elsevier; 1991;18: 293–317.
  61. 61. DeNiro MJ, Epstein S. Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta. 1981;45: 341–351.
  62. 62. Balter V, Simon L, Fouillet H, Lécuyer C. Box-modeling of 15N/14N in mammals. Oecologia. 2005;147: 212–222. pmid:16328553
  63. 63. Fuller MF, Reeds PJ. Nitrogen cycling in the gut. Annu Rev Nutr. 1998;18: 385–411. pmid:9706230
  64. 64. Martínez del Rio C, Wolf BO. Mass-balance models for animal isotopic ecology. In: Starck JM, Wang T, editors. Physiological and Ecological adaptations to Feeding in Vertebrates. Enfield, New Hampshire: Science Publishers; 2005. pp. 141–174.
  65. 65. Robinson TF, Roeder BL, Schaalje GB, Hammer JD, Burton S, Christensen M. Nitrogen balance and blood metabolites of alpaca (Lamapacos) fed three forages of different protein content. Small Rumin Res. 2005;58: 123–133.
  66. 66. Robbins CT, Felicetti LA, Sponheimer M. The effect of dietary protein quality on nitrogen isotope discrimination in mammals and birds. Oecologia. 2005;144: 534–540. pmid:15800751
  67. 67. McCutchan JH Jr, Lewis WM Jr, Kendall C, McGrath CC. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos. 2003;102: 378–390.
  68. 68. Robbins CT, Felicetti LA, Florin ST. The impact of protein quality on stable nitrogen isotope ratio discrimination and assimilated diet estimation. Oecologia. 2010;162: 571–579. pmid:19898979
  69. 69. McMahon KW, Polito MJ, Abel S, McCarthy MD, Thorrold SR. Carbon and nitrogen isotope fractionation of amino acids in an avian marine predator, the gentoo penguin (Pygoscelis papua). Ecol Evol. 2015;5: 1278–1290. pmid:25859333
  70. 70. McMahon KW, Thorrold SR, Elsdon TS, McCarthy MD. Trophic discrimination of nitrogen stable isotopes in amino acids varies with diet quality in a marine fish. Limnology and Oceanography. 2015;60: 1076–1087.
  71. 71. Robbins CT. Wildlife feeding and nutrition. San Diego, CA: Academic Press; 1993.
  72. 72. Van Soest PJ. Nutritional ecology of the ruminant. 2nd ed. Ithaca and London: Cornell University Press; 1994.
  73. 73. Schwarm A, Schweigert M, Ortmann S, Hummel J, Janssens GPJ, Streich WJ, et al. No easy solution for the fractionation of faecal nitrogen in captive wild herbivores: results of a pilot study. Journal of Animal Physiology and Animal Nutrition. 2009;93: 596–605. pmid:19178609
  74. 74. Sutoh M, Koyama T, Yoneyama T. Variations of natural 15N abundances in the tissues and digesta of domestic animals. Radioisotopes. 1987;36: 74–77. pmid:3575790
  75. 75. Sutoh M, Obara Y, Yoneyama T. The effects of feeding regimen and dietary sucrose supplementation on natural abundance of 15N in some components of ruminal fluid and plasma of sheep. J Anim Sci. 1993;71: 226–231. pmid:8454545
  76. 76. Meyers PA. Preservation of elemental and isotopic source identification of sedimentary organic matter. Chemical Geology. 1994;114: 289–302.
  77. 77. Prahl FG, Ertel JR, Goni MA, Sparrow MA, Eversmeyer B. Terrestrial organic carbon contributions to sediments on the Washington margin. Geochimica et Cosmochimica Acta. 1994;58: 3035–3048.
  78. 78. Sterner RW, Elser JJ. Stoichiometry and Homeostasis. Ecological Stoichiometry. Princeton, NJ: Princeton University Press; 2002. pp. 1–44.
  79. 79. Weaver JL. Refining the equation for interpreting prey occurrence in gray wolf scats. The Journal of Wildlife Management. 1993;57: 534–538.
  80. 80. Codron D, Codron J, Sponheimer M, Bernasconi SM, Clauss M. When animals are not quite what they eat: diet digestibility influences 13C-incorporation rates and apparent discrimination in a mixed-feeding herbivore. Can J Zool. 2011;89: 453–465.
  81. 81. Fedriani J, Kohn M. Genotyping faeces links individuals to their diet. Ecology Letters. 2001;4: 477–483.
  82. 82. Prugh LR, Arthur SM, Ritland CE. Use of faecal genotyping to determine individual diet. Wildlife Biology. 2008;14: 318–330.
  83. 83. Floyd TJ, Mech LD, Jordan PA. Relating wolf scat content to prey consumed. The Journal of Wildlife Management. 1978;42: 528–532.
  84. 84. Hawthorne VM. Coyote food habits in Sagehen Creek basin, northeastern California. California Fish and Game. 1972;58: 4–12.
  85. 85. Gier HT. Coyotes in Kansas. Manhattan: Agricultural Experiment Station, Kansas State College of Agriculture and Applied Science; 1968. pp. 1–120.