Assessing carnivore diet by faecal samples and stomach contents: a case study with Alpine red foxes

Research on the feeding habits of mammalian carnivores relies mainly on the analysis of stomach contents and faecal samples, but the outcomes of these two methods have only been compared in a few studies, with contrasting conclusions. In an Alpine area of NW Italy, we analysed both fox faeces collected along standardised transects, and the stomach contents of road-killed individuals. Faecal analysis involved the identification of macroscopic fragments, the identification of earthworm chaetae, and the assessment of relative volumes using Kruuk and Parish’s technique. Use of both methods indicated that the diet of the red fox included mainly fruit and mammals, but quantitative differences emerged. Garbage, birds, and cultivated fruit were overrepresented in the stomach contents, while earthworms, mammals, and wild fruit prevailed in the faecal samples. Logistic Regression Analysis suggested that the method of analysis was the main factor in determining the occurrence of food items in fox diet. Nonetheless, evidence suggests that road-killed foxes may include a disproportionately high percentage of synanthropic individuals and therefore be biased towards anthropogenic food. Results suggest that by using Kruuk and Parish’s technique, the main limitation of faecal analysis, i.e. the inaccurate estimation of the relative volume of each food item, can be overcome.     

Kinship,inbreeding and fine‐scale spatial structure influence gut microbiota in a hindgut‐fermenting tortoise

Herbivorous vertebrates rely on complex communities of mutualistic gut bacteria to facilitate the digestion of celluloses and hemicelluloses. Gut microbes are often convergent based on diet and gut morphology across a phylogenetically diverse group of mammals. However, little is known about microbial communities of herbivorous hindgut‐fermenting reptiles. Here, we investigate how factors at the individual level might constrain the composition of gut microbes in an obligate herbivorous reptile. Using multiplexed 16S rRNA gene sequencing, we characterized the faecal microbial community of a population of gopher tortoises (Gopherus polyphemus) and examined how age, genetic diversity, spatial structure and kinship influence differences among individuals. We recovered phylotypes associated with known cellulolytic function, including candidate phylum Termite Group 3, suggesting their importance for gopher tortoise digestion. Although host genetic structure did not explain variation in microbial composition and community structure, we found that fine‐scale spatial structure, inbreeding, degree of relatedness and possibly ontogeny shaped patterns of diversity in faecal microbiomes of gopher tortoises. Our findings corroborate widespread convergence of faecal‐associated microbes based on gut morphology and diet and demonstrate the role of spatial and demographic structure in driving differentiation of gut microbiota in natural populations.     

Spatial patterns in brown bear Ursus arctos diet: the role of geographical and environmental factors

• 1 We reviewed worldwide spatial patterns in the food habits of the brown bear Ursus arctos in relation to geographical (latitude, longitude, altitude) and environmental (temperature, snow cover depth and duration, precipitation, primary productivity) variables.
• 2 We collected data from 28 studies on brown bear diet based on faecal analysis, covering the entire geographical range of this widely distributed large carnivore. We analysed separately four data sets based on different methods of diet assessment.
• 3 Temperature and snow conditions were the most important factors determining the composition of brown bear diet. Populations in locations with deeper snow cover, lower temperatures and lower productivity consumed significantly more vertebrates, fewer invertebrates and less mast. Trophic diversity was positively correlated with temperature, precipitation and productivity but negatively correlated with the duration of snow cover and snow depth. Brown bear populations from temperate forest biomes had the most diverse diet. In general, environmental factors were more explicative of diet than geographical variables.
• 4 Dietary spatial patterns were best revealed by the relative biomass and energy content methods of diet analysis, whereas the frequency of occurrence and relative biomass methods were most appropriate for investigating variation in trophic diversity.
• 5 Spatial variation in brown bear diet is the result of environmental conditions, especially climatic factors, which affect the nutritional and energetic requirements of brown bears as well as the local availability of food. The trade‐off between food availability on the one hand, and nutritional and energetic requirements on the other hand, determines brown bear foraging decisions. In hibernating species such as the brown bear, winter severity seems to play a role in determining foraging strategies. Large‐scale reviews of food habits should be based on several measures of diet composition, with special attention to those methods reflecting the energetic value of food.

     

Diet composition and foraging success in generalist predators: Are specialist individuals better foragers?

Factors affecting individual diet specialization in generalist populations and the relationship between diet and foraging success remain poorly studied, particularly in terrestrial wide-ranging predators. We studied whether individual variations in diet in Montagu's harrier males (determined through a combination of direct foraging observations and pellet analysis) were associated with patterns of foraging habitat selection and foraging success of 12 radiotracked males during the breeding period. We found important differences in diet composition and breadth between individuals. Diet diversity was negatively related to hunting success: the most efficient individuals in terms of hunting success had the most specialized diet. This study also suggests an important role of individual foraging habitat selection in explaining individual diet, as the proportion of different prey types in the diet was associated with habitat composition within the home range, with higher proportion of those habitats that held higher abundances of their more frequent prey. This study thus provides evidence of individual diet specialization having a knock-on effect on foraging efficiency in a wide-ranging raptor and highlights the role of individual behaviour as a driving force of intra-population niche variation.     

Between‐individual variation drives the seasonal dynamics in the trophic niche of a Neotropical marsupial

The dynamics of population niches result from the variation in resource use within individuals and also from the variation between individuals. The prevalence of one mechanism or the other leads to competing hypotheses about the major mechanisms underlying the empirical observations of the contraction/expansion dynamics of the trophic niche in natural populations. In this study, we investigated how within‐ and between‐individual variation in resource use shapes the food niche dynamics of the woolly mouse opossum, Marmosa paraguayana (Didelphimorphia: Didelphidae), in a remnant of the highly seasonal Cerrado in south‐eastern Brazil. To do so, we analysed the faecal samples of live‐trapped individuals to determine their diets within the wet and dry seasons. In addition to a seasonal shift in the composition of the diet, the population trophic niche was significantly wider during the dry season than the wet season. This expansion resulted from larger between‐individual variation in the dry season that was not related to sex preferences, whereas the individual niche widths did not significantly increase from the wet to the dry seasons. Our findings add to the growing list of animal populations that show individual‐level variation in resource use. Furthermore, these results represent a pattern of individual‐level response to seasonal changes that is different from patterns reported for other organisms. We suggest that a pathway to build more realistic foraging models and produce more accurate predictions on population and community dynamics is to consider between‐individual variation and short‐term niche dynamics.     

Individual foraging specialisation in a social mammal: the European badger (Meles meles)

Individual specialisation has been identified in an increasing number of animal species and populations. However, in some groups, such as terrestrial mammals, it is difficult to disentangle individual niche variation from spatial variation in resource availability. In the present study, we investigate individual variation in the foraging niche of the European badger (Meles meles), a social carnivore that lives in a shared group territory, but forages predominantly alone. Using stable isotope analysis, we distinguish the extent to which foraging variation in badgers is determined by social and spatial constraints and by individual differences within groups. We found a tendency for individual badgers within groups to differ markedly and consistently in their isotope values, suggesting that individuals living with access to the same resources occupied distinctive foraging niches. Although sex had a significant effect on isotope values, substantial variation within groups occurred independently of age and sex. Individual differences were consistent over a period of several months and in some instances were highly consistent across the two years of the study, suggesting long-term individual foraging specialisations. Individual specialisation in foraging may, therefore, persist in populations of territorial species not solely as a result of spatial variation in resources, but also arising from individuals selecting differently from the same available resources. Although the exact cause of this behaviour is unknown, we suggest that specialisation may occur due to learning trade-offs which may limit individual niche widths. However, ecological factors at the group level, such as competition, may also influence the degree of specialisation.     

The influence of diet on foraging habitat models: a case study using nursing Antarctic fur seals

Predictable sources of food underpin lifetime reproductive output in long lived animals. The most important foraging areas of top marine predators are therefore likely to be related to environmental features that enhance productivity in predictable spatial and temporal patterns. Even so, although productive areas within the marine environment are distributed patchily in space and time, most studies assess the relationships between feeding activity and proximate, not long term, environmental characteristics. In addition, individuals within a population may exploit different prey types, and these are often associated with different hydrographic features. Until now, models attempting to associate core foraging areas (CFAs) of marine predators with the environmental characteristics of those areas have not considered the diet of individual animals, despite the influence this could have on these relationships. We used bathymetry and multi‐year (n=24) mean sea surface temperature and variability as predictors of CFAs of lactating Antarctic fur seals Arctocephalus gazella at Heard Island. The effect of prey types on the predictability of these models was explored by matching diet and foraging trip data of individual seals (n=40 seals, n=1 trip each). Differences in diet between seals were mirrored by their spatial behaviour. Foraging strategies differed both between and within groups of seals consuming different diets. Long‐term environmental parameters were useful for predicting the foraging activity of seals that consumed a single prey type with relatively specific habitat preferences, but not for those that consumed single or multiple prey types associated with more varied habitats. Ignoring individual variation in predator diet probably contributes to the poor performance of foraging habitat models. These findings highlight the importance of incorporating individual specialization in foraging behaviour into ecological models and management of predator populations.     

Faecal avoidance and the risk of infection by nematodes in a natural population of reindeer

We tested whether Svalbard reindeer (Rangifer tarandus platyrhynchus) minimise the risk of gastro- intestinal nematode infection by avoiding patches with a high density of faeces. This experiment was performed in preferred summer foraging habitat. The possibility that reindeer assess infection risk on the basis of faecal contamination levels across plant communities was determined by measuring the distribution of faeces in seven plant communities, and nematode developmental success in two plant communities with contrasting soil moisture content. We explored whether variation within individual reindeer in the levels of infection by gastro-intestinal nematodes was related to their diet. Reindeer avoided pastures where faecal contamination was increased, and thereby potentially reduced the risk of becoming infected by Trichostrongyle nematodes. Dung density was inversely related to soil moisture content, with high densities of faeces in dry plant communities and low densities in wet communities. However, nematode developmental success was positively related to soil moisture content, and was highest in the wetter sites. Thus, by avoiding dry areas with high dropping densities, reindeer would tend to feed in wetter areas where nematodes thrive. Therefore, dung density may be an unreliable predictor of the risk of infection. The absence of a strong relationship between an individual’s infection level and its diet might be due to the unpredictability of pasture infection level. Received: 8 July 1999 / Accepted: 20 January 2000     

Context-dependent specialisation drives temporal dynamics in intra- and inter-individual variation in foraging behaviour within a generalist bird population

Individual niche variation is common within animal populations, and has significant implications for a wide range of ecological and evolutionary processes. However, individual niche differences may also temporally vary as a result of behavioural plasticity. While it is well understood how niche variation is affected by changes in resource availability, comparatively little is known about the extent to which individual niche differences may vary within the annual cycle due to internal drivers. Here, we assess how time- and energy-constraints imposed by incubating and brood rearing affect inter- and intra-individual variation in the foraging behaviour of lesser black-backed gulls, a generalist seabird with strong individual niche variation. To this end, we compared daily foraging trips of 22 breeding and 23 non-breeding GPS-tracked adult gulls from two colonies in the Southern Bight of the North Sea over the course of the breeding season. We find that breeding birds, unlike non-breeding ones, did indeed alter their foraging behaviour during the breeding season. Both sexes reduced their searching effort by increasingly revisiting earlier foraging locations, allowing for shorter and more frequent foraging trips. Breeding females also showed pronounced shifts in their habitat use and strongly specialised on urbanised foraging habitats throughout the breeding season. Hence, while individual variation in habitat use remained largely consistent within non-breeders and in breeding males, individual variation among breeding females almost completely disappeared. Female lesser black-backed gulls are on average smaller, and therefore often outcompeted by males for the most profitable food sources. The temporal specialisation on spatially reliable anthropogenic food sources during breeding hence suggests a complex interplay between intrinsic competitive constraints, resource reliability and shifting time- and energy budges in shaping temporal dynamics in individual niche variation within our study population.     

Individual dietary specialization in a generalist predator: A stable isotope analysis of urban and rural red foxes

Some carnivores are known to survive well in urban habitats, yet the underlying behavioral tactics are poorly understood. One likely explanation for the success in urban habitats might be that carnivores are generalist consumers. However, urban populations of carnivores could as well consist of specialist feeders. Here, we compared the isotopic specialization of red foxes in urban and rural environments, using both a population and an individual level perspective. We measured stable isotope ratios in increments of red fox whiskers and potential food sources. Our results reveal that red foxes have a broad isotopic dietary niche and a large variation in resource use. Despite this large variation, we found significant differences between the variance of the urban and rural population for δ¹³C as well as δ¹⁵N values, suggesting a habitat‐specific foraging behavior. Although urban regions are more heterogeneous regarding land cover (based on the Shannon index) than rural regions, the dietary range of urban foxes was smaller compared with that of rural conspecifics. Moreover, the higher δ¹³C values and lower δ¹⁵N values of urban foxes suggest a relatively high input of anthropogenic food sources. The diet of most individuals remained largely constant over a longer period. The low intraindividual variability of urban and rural red foxes suggests a relatively constant proportion of food items consumed by individuals. Urban and rural foxes utilized a small proportion of the potentially available isotopic dietary niche as indicated by the low within‐individual variation compared to the between‐individual variation. We conclude that generalist fox populations consist of individual food specialists in urban and rural populations at least over those periods covered by our study.     

Hormonal correlates of being an innovative greylag goose, Anser anser

A number of studies have focused on the spread of foraging innovations within animal populations, but only rarely have individual dispositions of becoming innovative been considered. With two groups of individually marked, hand-reared greylag goslings we investigated hormonal and behavioural correlates related to the individual's ability to perform operant tasks. During individual tests 6 weeks after hatching, goslings were given small food containers covered by lids. In the following winter the sibling groups were tested in a social set-up at a food dispenser, which the geese could activate by pulling a flap. We analysed individual faecal samples collected at 2, 6 and 12 weeks of age, and also after the individual tests, for excreted corticosterone and testosterone metabolites by enzyme immunoassay. During the individual test, 18 of 23 individuals learned to remove the lids. These 18 birds excreted higher faecal corticosterone concentrations than their respective controls 2 weeks after hatching. At the food dispensers, only four males became food producers; all the others scrounged. These four were in the group of 18 that were successful in the individual test and again tended, although not significantly, to have higher faecal corticosterone 2 weeks after hatching than the scroungers. In one of the groups, excreted corticosterone increased and excreted testosterone decreased after the individual test. Goslings successfully removing lids at 6 weeks raised their faecal corticosterone to a significantly greater extent than the unsuccessful individuals. Our results suggest that becoming an innovator may be contingent upon individual coping styles. Copyright 2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.     

A New Way of Assessing Foraging Behaviour at the Individual Level Using Faeces Marking and Satellite Telemetry

Heterogeneity in foraging behaviour can profoundly influence ecological processes shaping populations. To scale-up from individual foraging behaviour to processes occurring at the population scale, one needs to sample foraging behaviour at the individual level, and over large temporal scales or during critical seasons known to influence life-history traits. We developed an innovative technique to monitor foraging behaviour at the individual level in secretive species, a technique that can be ultimately used to investigate the links between foraging behaviour and life-history traits. First, the technique used a novel approach, namely the combination of telemetry tracking and biomarking of faeces with food dyes to locate fresh signs of presence left by individuals equipped with GPS collars. Second, the technique is based on the simultaneous or successive sampling of life-history traits and individual foraging behaviour, using tracks with high probabilities of recovery of dyed faeces. We first describe our methodological approach, using a case study of a large herbivore, and then provide recommendations and guidelines for its use. Sampling single snow tracks of individuals equipped with a GPS collar was a reliable way to assess individual winter foraging behaviour in a white-tailed deer (Odocoileus virginianus Zimmermann) population. During that period, the probability of recovery of dyed faeces within the range of the collar precision was very high for single snow tracks of equipped deer (97%). Our approach is well suited to study individual foraging behaviour, and could ultimately be used to investigate the interplay between intra-population heterogeneity in foraging behaviour, life-history traits, and demographic processes.     

Variation in foraging behavior facilitates resource partitioning in a polymorphic cichlid, Herichthys minckleyi

We examined foraging behavior (microhabitat use and feeding behavior) in a trophically polymorphic cichlid fish, Herichthys minckleyi, to address several questions regarding resource partitioning in this threatened species. These include: (1) do morphotypes demonstrate different foraging behaviors? (2) do individuals within a morphotype vary in their foraging behavior (e.g. are some individuals specialists, only using a subset of available resources, while other are generalists)? (3) do foraging behaviors vary between isolated pools? (4) do foraging behaviors vary across seasons? We quantified microhabitat use and feeding behavior for over 100 individuals (of two morphotypes) feeding freely in two isolated pools (populations) and across two seasons (winter and summer). We found differences in foraging behavior between morphotypes and individual specializations within morphotypes; i.e. some individuals specialize on certain food resources by using a few feeding behaviors within a subset of microhabitats, whereas others employ a range feeding behaviors across many microhabitats. Foraging behavior also varied between pools and across seasons. This spatial and temporal variation in foraging behavior and resource use may serve to maintain this polymorphism, as the relative fitness of the each morph may vary over space and time.     

Evaluation of non-lethal gut microbiome sampling methods in a passerine bird

Gut microbial communities play critical roles in the biological functions of their host, such as mediating nutrient absorption, digesting food components the host cannot, and offering protection against enteric pathogens. Extensive research on gut microbial communities has been conducted on mammals, including humans and rodents, but much less work has been done in birds. Furthermore, much of the research on host–microbe interactions make use of faecal samples and rectal/cloacal swabs as a proxy for intestinal samples, which can be difficult to obtain directly. However, little is known about the overlap between the microbial communities of the gut, faeces and swabs, which limits interpretability of results based on faecal samples and swabs. To address this gap in knowledge, we compared the microbiome from five sample types – proventriculus, small intestine, large intestine, cloacal swabs and faeces – across individual Zebra Finches Taeniopygia guttata housed in constant conditions with a standardized diet. We compared diversity and community composition through 16S rRNA gene sequencing. Our results show that microbial communities from both cloacal swabs and faeces were distinct from proventriculus and small intestinal samples, but generally indistinguishable from large intestinal samples, indicating that these non-lethal samples may be useful proxies for large intestinal bacterial communities. Gaining insight into non-invasive sampling techniques for passerines has implications for studies of gut microbial diversity and abundance in wild bird populations. Furthermore, reliable non-lethal sampling is necessary for experiments where repeated sampling is required.     

Food habits of three carnivore species (Viverricula indica, Herpestes urva, and Melogale moschata) in Fushan Forest, northern Taiwan

The food habits and degrees of dietary overlap of lesser oriental civet ( Viverricula indica ), crab-eating mongoose ( Herpestes urva ), and ferret badger ( Melogale moschata ) inhabiting the Fushan Forest, northern Taiwan, were studied using faecal analysis between February 1993 and June 1994. Laboratory analysis of 154 civet faeces, and 174 mongoose faeces showed that both species fed on a wide variety of food items, including mammals, birds, reptiles, amphibians, fish, crustaceans, insects, oligochaetes, gastropods, chilopods, arachnids, and plants. Insects, oligochaetes, plants, and mammals were the four most important food items in the civets' diet, whereas crustaceans, insects, amphibians, and reptiles were the four most important food items consumed by mongooses. Amphibians were the only vertebrates, together with invertebrates and plants, found in the 64 ferret badger faeces we analysed, and oligochaetes, insects, and amphibians were the most important food items consumed by ferret badgers. The diversity of diet was highest in the mongoose, followed by the civet, and was lowest in the ferret badger.
The degree of dietary overlap was greatest between the civet and the ferret badger, followed by that of the civet and mongoose. The mongoose and ferret badger had the lowest degree of dietary overlap. However, the degree of dietary overlap varied in different seasons. Invertebrates were the most important food source for the carnivores in Fushan Forest.     

Individual–resource networks reveal distinct fruit preferences of selective individuals from a generalist population of the Helmeted Manakin

Ecological studies traditionally assume that generalist populations are homogeneous in the use of food resources, but empirical evidence supports that intraspecific differences in morphology, physiology and behaviour affect foraging decisions and promote diet variation among individuals. Furthermore, the temporal availability of resources may shape the dynamics of population trophic niche, which ultimately depends on individual niches. In this study, we investigated the seasonal changes in individual-based networks between the Helmeted Manakin Antilophia galeata, a generalist frugivorous bird, and fruiting plants, following theoretical models of interindividual diet variation based on the Optimal Diet Theory. Selective individuals were the majority of the generalist population of the Helmeted Manakin. Our results suggest that the structure of the individual-resource networks varied seasonally. We found that modularity was higher than expected by chance in the wet season, when fruit availability was also higher. In the dry season, modules disappeared and the network became more nested. These findings are consistent with the Distinct Preference Model of diet variation. We suggest that downscaling ecological networks to the individual level may reveal emergent properties that, albeit existent, are not evident in species–resources networks.     

Foraging behaviour and diet of an ectothermic herbivore: Testudo horsfieldi

Herbivorous vertebrates of arid regions are frequently faced with inadequate food quality, quantity or both. The time and energy devoted to foraging is vital to balancing their energy budgets. For desert ectotherms, a low metabolism should be advantageous, reducing their total energy requirement, but extreme ambient temperatures can strongly constrain these animals' activity periods. We provide the first data on the activity budgets, foraging behaviour and diet of a highly abundant, desert-dwelling, herbivorous ectotherm, the steppe tortoise Testudo horsfieldi . Extreme climatic conditions of Central Asia limit steppe tortoise's activity to only three months per year. They remain inactive most of their "active season" (90%), and spend very little time foraging (<15 min per day). This suggests that steppe tortoises can satisfy their energy requirements with modest feeding efforts. Interestingly, steppe tortoises avoid feeding on grass species and feed mostly on plant species that are usually highly toxic to mammals. This result suggests that steppe tortoises and ungulates do not compete for food.     

Network analyses support the role of prey preferences in shaping resource use patterns within five animal populations

Individual variation is an inherent aspect of animal populations and understanding the mechanisms shaping resource use patterns within populations is crucial to comprehend how individuals partition resources. Theory predicts that differences in prey preferences among consumers and/or differences in the likelihood of adding new resources to their diets are key mechanisms underlying intrapopulation variation in resource use. We developed network models based on optimal diet theory that simulate how individuals consume resources under varying scenarios of individual variation in prey preferences and in the willingness of consuming alternate resources. We then investigated how the structure of individual–resource networks generated under each model compared to the structure of observed networks representing five classical examples of individual diet variation. Our results support the notion that, for the studied populations, individual variation in prey preferences is the major factor explaining patterns in individual–resource networks. In contrast, variation in the willingness of adding prey does not seem to play an important role in shaping patterns of resource use. Individual differences in prey preferences in the studied populations may be generated by complex behavioral rules related to cognitive constraints and experience. Our approach provides a pathway for mapping foraging models into network patterns, which may allow determining the possible mechanisms leading to variation in resource use within populations.     

Individual variation in resource use by opossums leading to nested fruit consumption

Despite recent findings on the ecological relevance of within population diet variation far less attention has been devoted to the role diet variation for ecological services. Seed dispersal is a key ecological service, affecting plant fitness and regeneration based on foraging by fruit‐eating vertebrates. Here we used a network approach, widely used to understand how seed‐dispersal is organized at the species level, to gain insights into the patterns that emerge at the individual‐level. We studied the individual fruit consumption behavior of a South American didelphid Didelphis albiventris, during the cool–dry and warm–wet seasons. In species–species networks the heterogeneity in specialization levels generates patterns such as nestedness and asymmetry. Because generalist populations may be comprised of specialized individuals, we hypo thesized that network structural properties, such as nestedness, should also emerge at the individual level. We detected variation in fruit consumption that was not related to resource availability, ontogenetic or sexual factors or sampling biases. Such variation resulted in the structural patterns often found in species–species seed‐dispersal networks: low connectance, a high degree of nestedness and the absence of modules. Moreover structure varied between the warm–wet and cool–dry seasons, presumably as a consequence of seasonal fluctuation in fruit availability. Our findings suggest individuals may differ in selectivity causing asymmetries in seed dispersal efficiency within the population. In this sense the realized dispersal would differ from the expected dispersal estimated from their average dispersal potential. Additionally the results suggest possible frequency‐dependent effects on seed dispersal that might affect individual plant performance and plant community composition.     

Reliability of δ13C and δ15N in faeces for reconstructing savanna herbivore diet

We tested the reliability of herbivore faecal δ¹³C and δ¹⁵N values for reconstructing diet through review of an extensive database derived from a 3-year study of ungulates in South Africa's Kruger National Park. Faeces are a useful material for stable isotope studies of diet because they record dietary turnover at very short time scales, and because sampling is non-invasive. However, the validity of faecal isotope proxies may be questioned because they represent only undigested food remains. Results from Kruger Park confirm that free-ranging browsers have faecal δ¹³C consistent with C₃ feeding, grazer faeces are C₄, and mixed-feeder faeces intermediate. Although the respective ranges do not overlap, there is significant variation in faecal δ¹³C of browsers and grazers (～2.0–4.0‰) across space and through time. We demonstrate that most (～70%) of this variation can be ascribed to corresponding patterns of variation in the δ¹³C of C₃ and C₄ plants, respectively, re-enforcing the fidelity of faecal isotope proxies for diet but highlighting a need for mixing models that control for variations in plant δ¹³C in order to achieve accurate diet reconstructions. Predictions for the effects of climate (rainfall) and ecophysiology on ¹⁵N-abundance variations in mammals do not persist in faeces. Rather, faecal δ¹⁵N tracks changes in plant δ¹⁵N, with further fractionation occurring primarily due to variations in dietary protein (reflected by %N). Controlling for these effects, we show that a dual-isotope multiple source mixing model (Isosource) can extend diet reconstructions for African savanna herbivores beyond simplified C₃/C₄ distinctions, although further understanding of variations in mammal δ¹⁵N are needed for greater confidence in this approach.