Geometry of nutrition in field studies: an illustration using wild primates

Abstract and Figures

Nutritional geometry has shown the benefits of viewing nutrition in a multidimensional context, in which foraging is viewed as a process of balancing the intake and use of multiple nutrients. New insights into nutrient regulation have been generated in studies performed in a laboratory context, where accurate measures of amounts (e.g. eaten, converted to body mass, excreted) can be made and analysed using amounts-based nutritional geometry. In most field situations, however, proportional compositions (e.g. of foods, diets, faeces) are the only measures readily available, and in some cases are more relevant to the problem at hand. For this reason, a complementary geometric method was recently introduced for analysing multi-dimensional data on proportional compositions in nutritional studies, called the right-angled mixture triangle (RMT). We use literature data from field studies of primates to demonstrate how the RMT can provide insight into a variety of important concepts in nutritional ecology. We first compare the compositions of foods, using as an example primate milks collected in both the wild and the laboratory. We next compare the diets of different species of primates from the same habitat and of the same species (mountain gorillas) from two distinct forests. Subsequently, we model the relationships between the composition of gorilla diets in these two habitats and the foods that comprise these diets, showing how such analyses can provide evidence for active nutrient-specific regulation in a field context. We provide a framework to relate concepts developed in laboratory studies with field-based studies of nutrition.
Right-angled mixture triangle showing hypothetical data representing five foods (f1-f5) and the diets (i.e. combinations of foods, hollow symbols) eaten by various animals. Mixture composition: each point shows the protein (P), carbohydrate (C) and fat (F) composition of the mixture. In this model, P and C are represented on the X and Y axes, respectively. Since P, C and F sum to 100 %, stipulating values for P and C implicitly fixes the value for F. For example, since food f1 contains 20 % P and 10 % C, the value for F is 100 % − (20 + 10) % = 70 %; point f1 therefore has X:Y:Z coordinates of (20:10:70). By the same reasoning, the P:C:F coordinates for point f3 are (20:60:20). The third axis label is given in square brackets above the plot, and the value for the third variable is given in square brackets on the respective isolines for that variable. Comparing mixtures: compositional relationships between mixtures are defined by discrete vectors, or combinations of vectors: (1) points that fall on the same vertical contain the same concentration of P (e.g. f1 and f3 both have 20 % P); equivalently, points aligning on a horizontal have the same C concentration; (2) the radials projecting from the origin represent P:C ratio (e.g. f3 and f4 have the same P:C ratio, which is lower than that of f1 and f2); and (3) fat concentration is represented by diagonals with a slope of negative 1. The fat concentration represented by each diagonal is obtained by subtracting from 100 % the value where the diagonal intersects the P and C axes. Thus, both f1 and f4 contain fat at a concentration of 100 % − 30 % = 70 %, and f2 and f3 have a fat concentration of 100 % − 80 % = 20 %. Combining mixtures: the diets that can be composed by combining foods are constrained by the spatial relationship between the component foods. By mixing its intake from two foods an animal can compose a diet with composition that falls anywhere on the line connecting the points representing these foods, but nowhere off this line. For example, the diet represented by the square could be assembled by combining f2 and f5, but none of the diets represented by triangles or circles could be assembled in this way. Where three or more foods are eaten, the resulting diet is confined to the polygon formed by joining the component foods. Thus, all the circle diets could be achieved by combining foods f1-f3, as can all but one of the triangular diets (the one with composition closest to f5). To achieve this latter diet composition, the animal would need also to eat f5. Interpreting scatter of replicate points: the pattern of scatter of replicate points can contain significant biological information. For example, the scatter among the circle diets is tightly compressed around a P:C balance vector, but spread along this vector. This could indicate that the animals prioritize dietary P:C balance over fat content. Conversely, the triangular diets are aligned with a fat vector of 20 % but span a range of P:C balances, suggesting prioritization of fat
No caption available
No caption available
No caption available
Content may be subject to copyright.
1 3
DOI 10.1007/s00442-014-3142-0
Geometry of nutrition in field studies: an illustration using wild
David Raubenheimer · Gabriel E. Machovsky‑Capuska ·
Colin A. Chapman · Jessica M. Rothman
Received: 10 February 2014 / Accepted: 3 November 2014
© Springer-Verlag Berlin Heidelberg 2014
(RMT). We use literature data from field studies of pri-
mates to demonstrate how the RMT can provide insight
into a variety of important concepts in nutritional ecology.
We first compare the compositions of foods, using as an
example primate milks collected in both the wild and the
laboratory. We next compare the diets of different species
of primates from the same habitat and of the same spe-
cies (mountain gorillas) from two distinct forests. Subse-
quently, we model the relationships between the compo-
sition of gorilla diets in these two habitats and the foods
that comprise these diets, showing how such analyses can
provide evidence for active nutrient-specific regulation in
a field context. We provide a framework to relate concepts
developed in laboratory studies with field-based studies of
Keywords Nutritional ecology · Nutritional geometry ·
Mixture triangles · Primates · Gorillas
Understanding the relationships between nutrition and
behaviour, ecology, physiology, and demographic processes
of animals is a central aim in nutritional ecology (Parker
2003; Barboza et al. 2009; Raubenheimer et al. 2009, 2012;
Lambert 2010; DeGabriel et al. 2014). Many studies have
shown that this can best be achieved by disentangling the
discrete and interactive roles of different food compo-
nents (Westoby 1974; Dearing and Schall 1992; Simp-
son and Raubenheimer 1993; Bowen et al. 1995; Robbins
et al. 2007). A geometric framework was introduced for
this purpose, called the geometric framework for nutrition
(Raubenheimer and Simpson 1993; Simpson and Rauben-
heimer 1993). This framework defines the important facets
Abstract Nutritional geometry has shown the benefits of
viewing nutrition in a multidimensional context, in which
foraging is viewed as a process of balancing the intake
and use of multiple nutrients. New insights into nutrient
regulation have been generated in studies performed in a
laboratory context, where accurate measures of amounts
(e.g. eaten, converted to body mass, excreted) can be made
and analysed using amounts-based nutritional geometry.
In most field situations, however, proportional composi-
tions (e.g. of foods, diets, faeces) are the only measures
readily available, and in some cases are more relevant to
the problem at hand. For this reason, a complementary
geometric method was recently introduced for analysing
multi-dimensional data on proportional compositions in
nutritional studies, called the right-angled mixture triangle
Communicated by Joanna E. Lambert.
D. Raubenheimer (*) · G. E. Machovsky-Capuska
Faculty of Veterinary Science, The Charles Perkins Centre,
School of Biological Sciences, University of Sydney, Sydney,
C. A. Chapman
Department of Anthropology McGill School of Environment,
McGill University, Montreal, QC, Canada
C. A. Chapman
Wildlife Conservation Society, Bronx, New York, USA
J. M. Rothman
Department of Anthropology, Hunter College of the City
University of New York, New York, USA
J. M. Rothman
New York Consortium of Evolutionary Primatology,
New York, USA
1 3
of animal nutrition (e.g. foods, nutrient requirements, body
compositions, nutrient utilisation) in a cartesian space,
where each dimension represents a food component. Mod-
elling nutrition in this way enables the combined effects of
different food components to be quantified, and the various
levels of response by the animal (e.g. intake, growth, nutri-
ent absorption, performance) to be integrated within this
multi-dimensional context (Raubenheimer and Simpson
1997; Raubenheimer et al. 2009; Simpson and Raubenhe-
imer 2012).
An important feature of this geometric framework is
that the axes are scaled as amounts. Consequently, many
of the factors described in a nutrient space are represented
as time-integrated rates (e.g. intake over a given period,
growth within the same period), while other factors can
be represented either as amounts (e.g. nutrient content of a
stipulated quantity of food) or as proportions (e.g. the bal-
ance of nutrients X and Y within the food, or the balance of
nutrients required by the animal). Combining amounts and
proportions in this way has proved a powerful approach
for predicting an animal’s behavioural and physiologi-
cal responses to the nutritional environment (Simpson and
Raubenheimer 2012).
There are, however, many situations where models
of the proportional compositions of mixtures are prefer-
able to models of the absolute amounts of the constituents
(Raubenheimer 2011). First, in field work. the complex
sets of interacting variables and the logistical challenges of
collecting reliable data typically constrain the possibilities
for nutritional studies. For example, in the field, it is chal-
lenging to measure the daily intake of nutrients by an ani-
mal, but an estimate of the proportional composition of the
diet can be obtained using gut contents analysis (Hyslop
1980; Kamler and Pope 2001; Petry et al. 2007; Macho-
vsky-Capuska et al. 2011; Tait et al. 2014), regurgitations
(Schuckard et al. 2012; Tait et al. 2014), faecal analysis
(Klare et al. 2011; Giri et al. 2011; Panthi et al. 2012), bite
rates analysis (Shrader et al. 2006; Paddack et al. 2006) and
related methods. Second, field-based questions often relate
directly to proportions rather than absolute amounts, as is
the case where the nutritional compositions of different
foods or of foods versus non-foods are compared. Third, for
some purposes, the inclusion of amounts in the model will
introduce noise or surplus information, which is avoided in
an analysis of compositions. For example, a comparison of
the diets versus body composition of animals from different
trophic levels can be made using proportional compositions
(Fagan et al. 2002; Raubenheimer et al. 2007), whereas
analysis of absolute amounts of nutrients in, say, a predator
and its various prey species would introduce variance asso-
ciated with body size, thus complicating the model for lit-
tle benefit. Fourth, because proportional measures such as
food compositions are easy to obtain compared to measures
of absolute amounts (e.g. daily intakes), there exists a
wealth of data in the literature from which compositional
data can be extracted for comparative and meta-analyses.
For instance, Raubenheimer and Rothman (2013) were
able to examine the nutritional correlates of insectivory
in humans and other primates using published data on the
compositions of prey insects, when very few measures of
amounts of insects eaten were available.
Recently, a graphical approach, the right-angled mixture
triangle (RMT), was recommended as a complementary
geometric framework for problems in nutritional ecology
that involve primarily proportional data (Raubenheimer
2011). Like amounts-based nutritional geometry, RMT
provides a graphic model in which various facets of animal
nutrition can be represented and interrelated within a mul-
tidimensional context, but the axes represent proportions
of nutrients in mixtures (e.g. an animal’s diet) rather than
amounts. The exclusion of amounts in the model frees up a
dimension in RMT plots, enabling the relationships among
n components to be visualised in an n 1 dimensional
space. This property is particularly useful for represent-
ing three components in a regular two-dimensional plot,
because two-dimensional plots are intuitively accessible
and many problems in nutritional ecology concern three-
component mixtures. For example, numerous studies have
demonstrated the importance of the macronutrients pro-
tein, carbohydrate, and fat in the nutritional responses of
animals (Barboza et al. 2009; Simpson and Raubenheimer
2012; see also supplementary table in Raubenheimer et al.
2009), and the elements nitrogen, carbon, and phosphorus
have been identified in the science of ecological stoichiom-
etry as important drivers of ecosystem dynamics (Sterner
and Elser 2002). There are, however, also ways to repre-
sent more than three components in RMTs (Raubenheimer
Although RMT is technically different from amounts-
based nutritional geometry (henceforth ABNG), the two
approaches are complementary means for addressing simi-
lar questions under different circumstances. In contrast
with RMTs, however, the widespread application of ABNG
in laboratory studies has yielded a substantial body of
concepts around multidimensional analyses of nutritional
problems, the utility of which has been demonstrated in
a range of systems, questions and contexts (Simpson and
Raubenheimer 2012). For example, using this approach
Lee et al. (2008) and Solon-Biet et al. (2014) have dem-
onstrated that the life-extending effects of mild dietary
deprivation are not due to “caloric restriction” as widely
assumed, but rather specific effects of macronutrient ratios.
This has taken a number of years, partly because the data
suitable for amounts-based geometric analyses are seldom
found pre-existent in the literature, but require de-novo
studies designed for the purpose (although exceptions do
1 3
exist: Simpson and Raubenheimer 1997; Raubenheimer
and Simpson 1997; Lee et al. 2008).
Here, we illustrate the use of RMTs for addressing impor-
tant questions in field-based nutritional ecology, with the aim
of contributing to the development of a conceptual framework
for the study of animal nutrition in the wild. We are able to
do so by drawing on the conceptual foundation already devel-
oped in ABNG, and capitalising on the abundant proportions-
based literature data. Primates provide an excellent system
for our analyses, because a long history of field studies with
observational data of individuals has yielded abundant data
enabling us to illustrate how RMTs may be used to address a
range of significant questions in nutritional ecology. We also
illustrate how RMTs can be used as a synthetic and compara-
tive tool for integrating the wealth of published data on the
nutritional ecology of animals. Such integration is a power-
ful means for developing new insights and hypotheses, as has
recently been done in relation to the ways that appetite and
regulatory physiology interact with economics and global
change to generate obesity in humans and companion ani-
mals (Raubenheimer et al. 2014).
The right‑angled mixture triangle
Details of the history, derivation, and logic of the RMT are
provided by Raubenheimer (2011). In brief, RMTs provide
a means to represent mixtures (e.g. foods, diets, and animal
nutrient requirements) as points on a graph, and to extract
information from the geometric relationships among such
points. For the present purposes, there are four categories
of information that we wish to illustrate: compositional
representations, compositional comparisons, combinatorial
constraints, and patterns of scatter (Fig. 1). These catego-
ries of information provide the basis in the rest of the paper
for examples illustrating the application of RMT to specific
biological problems.
In RMTs, the composition of a mixture such as a food is
represented as an n-dimensional point in a space of n 1
dimensions. For example, if the macronutrients protein,
carbohydrate, and fat are the focus of the model, the mix-
ture is depicted as a 3-coordinate point in a two-dimen-
sional plot, where each coordinate gives the proportional
contribution of one of these components to the macronutri-
ent fraction of the mixture (Fig. 1). If a fourth component is
added to the model, for example fibre, then a 3-dimensional
plot can be used. Higher-dimensional mixtures can be rep-
resented in various ways (Raubenheimer 2011), but models
involving such data can be intractably complex and might
better be parsed into a series of three- or four-dimensional
Compositional differences between two mixtures (e.g.
foods) can be geometrically defined in terms of discrete
Fig. 1 Right-angled mixture triangle showing hypothetical data rep-
resenting five foods (f1f5) and the diets (i.e. combinations of foods,
hollow symbols) eaten by various animals. Mixture composition:
each point shows the protein (P), carbohydrate (C) and fat (F) com-
position of the mixture. In this model, P and C are represented on
the X and Y axes, respectively. Since P, C and F sum to 100 %, stipu-
lating values for P and C implicitly fixes the value for F. For exam-
ple, since food f1 contains 20 % P and 10 % C, the value for F is
100 % (20 + 10) % = 70 %; point f1 therefore has X:Y:Z coordi-
nates of (20:10:70). By the same reasoning, the P:C:F coordinates for
point f3 are (20:60:20). The third axis label is given in square brackets
above the plot, and the value for the third variable is given in square
brackets on the respective isolines for that variable. Comparing mix-
tures: compositional relationships between mixtures are defined by
discrete vectors, or combinations of vectors: (1) points that fall on
the same vertical contain the same concentration of P (e.g. f1 and f3
both have 20 % P); equivalently, points aligning on a horizontal have
the same C concentration; (2) the radials projecting from the origin
represent P:C ratio (e.g. f3 and f4 have the same P:C ratio, which is
lower than that of f1 and f2); and (3) fat concentration is represented
by diagonals with a slope of negative 1. The fat concentration repre-
sented by each diagonal is obtained by subtracting from 100 % the
value where the diagonal intersects the P and C axes. Thus, both f1
and f4 contain fat at a concentration of 100 % 30 % = 70 %, and f2
and f3 have a fat concentration of 100 % 80 % = 20 %. Combining
mixtures: the diets that can be composed by combining foods are con-
strained by the spatial relationship between the component foods. By
mixing its intake from two foods an animal can compose a diet with
composition that falls anywhere on the line connecting the points rep-
resenting these foods, but nowhere off this line. For example, the diet
represented by the square could be assembled by combining f2 and
f5, but none of the diets represented by triangles or circles could be
assembled in this way. Where three or more foods are eaten, the result-
ing diet is confined to the polygon formed by joining the component
foods. Thus, all the circle diets could be achieved by combining foods
f1–f3, as can all but one of the triangular diets (the one with composi-
tion closest to f5). To achieve this latter diet composition, the animal
would need also to eat f5. Interpreting scatter of replicate points: the
pattern of scatter of replicate points can contain significant biological
information. For example, the scatter among the circle diets is tightly
compressed around a P:C balance vector, but spread along this vector.
This could indicate that the animals prioritize dietary P:C balance over
fat content. Conversely, the triangular diets are aligned with a fat vec-
tor of 20 % but span a range of P:C balances, suggesting prioritization
of fat
1 3
quantitative vectors, or combinations of these; conversely,
similarities between mixtures can be recognized as shared
parameters within mixture space (Fig. 1). When two foods
are combined to form a diet, the set of possible diets is con-
strained to lie on the line connecting these foods. When
more than two foods are combined, the set of possible diets
falls within the area joining the points representing the
component foods (Fig. 1). Finally, the scatter of replicate
points in an RMT can provide important biological infor-
mation. Thus, if the compositions of the selected diets of
several replicate animals clustered more tightly along the
vector representing the balance of protein:carbohydrate
than the vector for dietary fat content, this might suggest
that the mechanisms regulating nutrient intake prioritize
protein to carbohydrate balance over fat intake (Fig. 1).
In the sections that follow, we analyse literature data to
show how the above principles can be used in field-based
primate studies to model key concepts in nutritional ecol-
ogy. Since many of these concepts have been developed in
ABNG, we also briefly explain how each concept is repre-
sented within that framework.
Foods: the composition of milk
Our first example concerns the multi-component compari-
son of the composition of foods. In ABNG, foods are rep-
resented in two ways. First, a stipulated measure of a par-
ticular food is represented by a point with coordinates that
give the amount of each of the focal nutrients. Second, the
more general property that pertains to any quantity of the
food, its nutrient balance, is given by the slope of the line
that joins the above-mentioned point with the origin. Such
lines representing the nutrient balance of a food are called
“nutritional rails”, reflecting the fact that as an animal eats
the food its nutritional state changes along a trajectory that
is coincident with the rail for that food.
As explained above, in RMTs the composition of foods
is depicted as an n-dimensional point in a space of n 1
dimensions. To illustrate the use of RMT in the analysis of
food compositions, we will use as an example the macronu-
trient content of primate milk. In our first analysis, we plot
the composition of field-collected samples of milk from
six species of primates (Fig. 2a). For two of these species,
common marmosets (Callithrix jacchus; Power et al. 2008)
and mountain gorillas (Gorilla beringei; Whittier et al.
2010), replicate samples are plotted, thus providing infor-
mation on the within-sample variability of the three nutri-
ents. For both gorillas and marmosets, the replicate samples
clustered more tightly along the negative-sloped diagonal
(representing protein concentration) and were more widely
spread along that vector suggesting that both species pro-
duce milk with a relatively fixed proportion of macronu-
trients contributed by protein, with higher intra-specific
variation in the fat:carbohydrate ratios. This has previously
been noted for common marmosets (Power et al. 2008) and
tufted capuchins (Milligan 2010; see also Raubenheimer
2011), and Fig. 2a shows that a similar pattern exists for
mountain gorillas. Raubenheimer (2011) suggested that
this pattern might reflect physiological regulation by the
mother to ensure that the suckling infant obtains a diet that
is balanced with respect to the protein:non-protein ratio, a
parameter that is commonly achieved through food selec-
tion and complementary feeding in weaned animals (Simp-
son and Raubenheimer 2012). Relatively high variability
in the carbohydrate:fat ratio likely reflects their substitut-
ability as sources of non-protein energy, as has been dem-
onstrated through the patterns of macronutrient selection
in several species including fish (Ruohonen et al. 2007),
domestic dogs (Hewson-Hughes et al. 2013), grizzly bears
(Erlenbach et al. 2014) and humans (Simpson and Rauben-
heimer 2005).
Given the logistical challenges of collecting milk from
primates in the field, an important question is to what
extent the milk of captive animals resembles that of con-
specifics in the wild. To address this, Power et al. (2008)
considered samples from both wild common marmosets
(plotted as filled circles in Fig. 2a) and captive conspecif-
ics (hollow circles). The protein content of the milk from
captive and wild marmosets was relatively constant with
more variation in the balance of fat:carbohydrate. Further,
the analysis of Power et al. (2008) showed that the pro-
tein content of the milk from the two groups did not dif-
fer statistically, as is suggested in Fig. 2a by the alignment
Fig. 2 Macronutrient composition of mammalian milk (protein, fat
and carbohydrate, on a mass basis). a Replicate samples of milk from
wild and captive common marmosets (Callithrix jacchus, Power et al.
2008) had a similar proportion of protein (clustered along a negative
diagonal representing 20 % P), but varied in the fat:carbohydrate
ratio (was spread along the diagonal). Replicate samples of milk col-
lected from free-ranging mountain gorillas Gorilla beringei (Whittier
et al. 2010) also had a relatively constant protein content, but this was
lower (clustered along the diagonal representing 15 % P) than mar-
moset milk. Also plotted are single milk samples collected in the field
from four other primate species, three of which had similar protein
content to marmoset milk and the fourth had similar protein content
to gorilla milk, but a higher fat:carbohydrate ratio. b Comparison of
macronutrient content in the milks of apes, New World monkeys, Old
World monkeys and Strepsirrhines. Mean protein content was lowest
in apes (11.3 %, indicated by the red line), intermediate in Old World
monkeys (13.6 %, grey line) and highest in New World monkeys
(20.0 %, blue line). Strepsirrhines split into two clusters, one with
intermediate protein (16 %, dashed black line) and low fat, and the
other with higher protein (25 %, solid black line) and higher fat. The
two clusters of Strepsirrhines represent different dietary groups: the
former being herbivorous, and the latter including a substantial por-
tion of insects in the diet. Data from Hinde and Milligan (2011). c
Comparison of macronutrient content in the milks of primates with
a range of non-primate mammals. Primate milks generally have low
protein and fat, with high carbohydrate content compared with the
other mammals (colour figure online)
1 3
of the samples from the two groups along the same protein
isoline (negative diagonal). On the other hand, the captive
marmosets had a higher fat:carbohydrate ratio than wild
marmosets, as indicated by their displacement to the left
along the protein isoline (Fig. 2a), although there was sub-
stantial overlap. Overall, this suggests that milk taken from
1 3
captive marmosets might be representative of wild samples
with respect to the ratio of protein:non-protein energy, but
the balance of fat to carbohydrate is more context specific.
Despite the fact that the concentration of protein was
maintained relatively constant in the milk of both marmo-
sets and gorillas, there was also a marked difference in the
milk of these two species. The milk of gorillas clustered
along an isoline representing a lower protein concentra-
tion (displaced further from the origin, mean 14 %) than
marmosets (20 %) (t1,34 = 3.76, P < 0.001, independent
samples t test). Also plotted are single milk samples from
four other primates. Three of these had similar protein con-
tent to marmoset milk (Alouatta palliata, A. seniculus, and
Leontopithecus rosalia), and the fourth (Macaca sinica)
had similar protein content to mountain gorilla milk but a
higher fat:carbohydrate ratio.
Is the proportional protein content more generally a
dimension that distinguishes the milk of different primates?
Hinde and Milligan (2011) presented data which sug-
gest that this is the case. These data, plotted as an RMT
in Fig. 2b, show that across several species of apes, Old
World Monkeys and New World monkeys, protein was rel-
atively constant within groups compared with fat and car-
bohydrate, but differed between primate groups. Apes had
the lowest protein content (mean ± SE = 11.3 ± 1.27 %),
followed by Old World monkeys (13.6 ± 0.75 %) and
New World Monkeys (20.0 ± 0.72 %) (P < 0.0001, inde-
pendent samples Kruskal–Wallis test). A fourth group,
the Strepsirrhines, separated into two sub-groups, which
corresponded with dietary differences. Milk from the her-
bivorous Strepsirrhines had lower protein (16 ± 2.19 %)
and fat (13 ± 2.39 %) than that from species that include
a substantial proportion of insects (at least 27 %, National
Research Council 2003) in the diet (protein = 25 ± 1.76 %,
fat = 43 ± 2.38 %). These differences did not, however,
stand up to a more conservative phylogenetic analysis
(mean logit difference ±CI, protein = 0.454, 1.577 to
0.702; fat = 1.591, 3.090 to 0.140; phylogenetic mixed
model approach, as described by Hadfield and Nakagawa
2010). It would be worth exploring this question using a
larger sample size.
Finally, Fig. 2c compares the data for primate milk with
equivalent data for other mammals. The plot shows that
only the horse and the ass fell within the range for pri-
mates, whereas most other milks had higher proportional
protein and/or fat content than primates, clearly show-
ing that the milk of primates differs markedly from other
mammals in terms of its protein concentration. In addi-
tion, primates produce relatively dilute milks with lower
energy density than other mammals (Hinde and Milligan
2011); although not illustrated here, this can readily be
modelled using RMT (Raubenheimer 2011). Comparative
analysis suggests that such variation in the composition of
mammalian milk is due to a combination of phylogeny and
specific adaptations such as diet and life histories (Skibiel
et al. 2013).
Diets: Comparative nutrient intakes
RMTs also provide a means to visualise and compare the
relationships between the mixtures of foods eaten by ani-
mals and the resulting nutrient gains (i.e. animal diets). Fig-
ure 3, for example, presents the estimated annual intakes
of protein, non-structural carbohydrate and fibre in the
plant-derived component of the diets of six wild primates,
representing five species. These are chimpanzees (Pan trog-
lodytes), blue monkeys (Cercopithecus mitis), red-tailed
monkeys (Cercopithecus ascanius), and grey-cheeked
mangabey (Lophocebus albigena) in Kibale National Park,
Uganda, and mountain gorillas in Bwindi National Park,
Uganda, and Virunga National Park, Rwanda. Plotting the
diets of these populations in this way clearly illustrates
Fig. 3 Protein, non-structural carbohydrate (NSC) and neutral-deter-
gent fibre composition of the plant-derived component of the diets
of chimpanzees and three species of monkeys from Kibale National
Park, Uganda, compared with mountain gorillas from Bwindi and
Virunga (chimpanzees Pan troglodytes, green diamond; blue monkey
Cercopithecus mitis, blue circle; red-tailed monkey Cercopithecus
ascanius, blue pentagon; mangabey Lophocebus albigena, blue tri-
angle; Virunga gorillas, red elipse; Bwindi gorillas, red square).
Radials show the protein:NSC ratio, and the negative diagonals the
%NDF. The three different species of monkeys living in overlapping
home ranges had very similar dietary composition, as did the two
populations of mountain gorillas living in different habitats. Chim-
panzees, which overlap in habitat with the monkeys, had a lower
protein:carbohydrate ratio than the other species, but similar propor-
tional NDF intake as the monkeys (40 %). The diet of gorillas had the
highest protein:carbohydrate ratio, and also the highest concentration
of NDF (54 %) (colour figure online)
1 3
several interesting patterns in a single plot. First, the bal-
ance of protein, non-structural carbohydrates, and fibre in
the plant tissues eaten by the three monkey species was
remarkably similar (Conklin-Brittain et al. 1998). It is
interesting that the plant component of the diet of red-tailed
monkeys did not differ from the other two monkey species,
even though, in addition to plants, red-tailed monkeys also
include a significant proportion of high-protein insects in
their diet (Rode et al. 2006; Bryer et al. 2013). This sug-
gests that omnivory in this species is associated with a
higher protein target than the other monkey species, rather
than complementary feeding to achieve a similar nutritional
target (Raubenheimer and Jones 2006). Second, the fibre
content of the diets of chimpanzees and the three monkey
species was similar, but the proportion of fibre in the diets
of gorillas was higher. Third, the protein:non-structural
carbohydrate ratio in the diets of these primates increased
from chimpanzees to gorillas, with the diets of monkeys
being intermediate. Finally, the intakes of the two gorilla
populations converged in the nutrient space, despite the fact
that they lived in two geographically separate and botani-
cally very distinct forests (Rothman et al. 2007). This is
significant, for reasons that we explain next.
Role of nutritional regulation in diet selection
Amounts-based nutritional geometry has been used in
laboratory experiments to demonstrate that many animals
actively regulate their intake of different nutrients sepa-
rately to track an intake target. Known instances include
herbivores, omnivores, and predators, spanning inverte-
brates and vertebrates (Simpson and Raubenheimer 2012).
It is important in investigating this issue to determine the
extent to which the composition of the selected diet results
from active, homeostatic regulation of intake, or is simply a
passive consequence of the composition of available foods
(Raubenheimer et al. 2012).
There are several ways to do this. One approach is to
compare the intakes of two or more groups of similar ani-
mals that are provided with different nutritionally comple-
mentary food combinations. In this design, the animals in
the different experimental groups have to spread their feed-
ing among the foods differently to achieve the same nutri-
ent gain, because combining the foods in similar propor-
tions will result in different nutrient gains. Chambers et al.
(1995) took this approach in experiments on the African
migratory locusts (Locusta migratoria). There were four
groups of locusts, each of which was given a pair of syn-
thetic foods: one containing a protein:carbohydrate (P:C)
ratio of 1:2 and the other of 2:1. The treatments differed,
however, in the extent to which the foods were diluted
using indigestible cellulose: the macronutrient mixture
comprised either 42 % (dry weight) of both foods, 21 % of
both foods, or 21 % of one food and 42 % of the other. If
the locusts in the different experimental groups ate similar
amounts of the respective food pairings, then they would
end up with very different nutrient intakes. This was not the
case: the locusts spread their feeding across their respective
food pairings in such a way that the nutrient intake points
of the four groups converged tightly in the nutrient space.
These results demonstrate that locusts faced with variation
in the composition of available foods alter their feeding
behaviour to maintain a target macronutrient intake.
Demonstrating macronutrient regulation to a target
intake by free-ranging animals in the wild is more chal-
lenging. However, the core principle of testing for con-
stant nutrient intake in the face of variation in food com-
position applies equally in the laboratory and field. Using
RMT, this would mean comparing the nutrient gains of
different groups of the animals when feeding on disparate
food combinations. If the foods are combined in different
proportions that result in a diet of similar nutrient balance,
then this suggests that diet selection is driven by nutrient-
specific regulation; i.e. the animals in different environ-
ments are regulating food intake so as to gain the required
balance of nutrients. It is, however, also possible that the
distribution of food compositions in the respective envi-
ronments is, by coincidence, such that diet selection using
criteria other than nutrient requirements (e.g. frequency-
dependent selection of foods) results in similar nutrient
gain across environments. This possibility can be addressed
by comparing the frequency of different foods in the diet
and environment.
An example of such an analysis to test for nutrient-
based food selection in the wild is given in Fig. 4a. The plot
shows the estimated annual intake of protein, non-struc-
tural carbohydrate, and fibre in the diets of mountain goril-
las in Virunga and Bwindi National Parks (the same data
as in Fig. 3). Also shown are the compositions of all the
foods that contributed 1 % or more of the diets by weight
(these cumulatively amounted to 90 and 96 % of the diets
of the Virunga and Bwindi populations, respectively; Roth-
man et al. 2007). This figure shows that the similar nutrient
intakes of the two populations were compiled from differ-
ent combinations of foods, an outcome which suggests that
the diet of these gorillas is determined by active nutrient
As noted above, it remains possible, however, that the
similar nutrient intakes of Bwindi and Virunga goril-
las were a passive consequence of the relative availabili-
ties of different foods in the two habitats. If this were the
case, then observed diet compositions of the two popula-
tions would correspond with the hollow square (Virunga)
and circle (Bwindi) in Fig. 4a. Visually, it appears that this
is not the case. To evaluate this statistically, we tested for
relationships between percentage contribution to the diet
1 3
of the foods and their relative availability in the respective
habitats (Fig. 4b). There was no significant relationship,
suggesting that the similar dietary compositions of gorillas
in Bwindi and Virunga were not a passive consequence of
food availability, but involved selection of foods in a pat-
tern that was disproportionate in relation to availability.
Further, Fig. 5a, b shows that the mechanism for achiev-
ing the observed nutrient gains differed between the pop-
ulations. In the Virunga gorillas, 36 and 22 % of the diet
were contributed from the first and second highest-ranking
foods in the diet, respectively. The top contributing food
had a protein:carbohydrate ratio that was lower than the
overall diet, while the second had protein:carbohydrate
ratio that was appreciably higher than the diet. This shows
that 58 % of the diet of the Virunga gorillas was constituted
by mixing two nutritionally complementary foods. In con-
trast, the relationships among the top-ranking foods that
comprised 58 % of the diet of Bwindi gorillas were very
different. In this case, six foods were involved, all of which
had a protein:carbohydrate ratio similar to the diet overall.
The diet of Bwindi gorillas was therefore composed largely
from foods that had a similar balance to the selected intake
point, with complementary mixing of disparate foods play-
ing a lesser role than for the Virunga gorillas.
Digestive efficiencies
The functional significance of nutrition derives from the
interaction of ingestion and post-ingestive processing of
foods. Amounts-based geometric analysis can investigate
this issue, by constructing multi-component nutrient budg-
ets in which the amounts of the focal nutrients ingested,
used, and excreted are plotted separately in the same nutri-
ent space (Raubenheimer et al. 2009). Measuring post-
ingestive processing of dietary components in this way is
more challenging for free-ranging wild animals, because of
the difficulties of estimating for an individual the amounts
of a food eaten and amounts of faeces produced from that
It is, however, possible using RMTs to estimate the rela-
tive digestive priorities of wild animals by comparing the
compositions of their foods and faeces (Raubenheimer
2011). To illustrate, Fig. 6 shows an example concerning
the fibre components (cellulose, hemicelluloses, and lignin)
in fruit and leaves eaten by juvenile, female and silverback
gorillas from Bwindi, and in the faeces associated with
fruit- and leaf-eating periods. The plot shows, firstly, that
there was no distinct separation of fruit and leaves in the
composition space, indicating that the fibre composition of
these two foods did not differ (as noted previously by Roth-
man et al. 2006). Secondly, the ratio of lignin:hemicellulose
was higher in the faeces than the foods, reflecting the
fact that lignin is undigestible for gorillas whereas
Fig. 4 a Protein, non-structural carbohydrate (NSC) and neutral-deter-
gent fibre ratios of the principal foods (those contributing >1 % to the
diet) and diet composition of two allopatric populations of mountain
gorillas, in Virunga and Bwindi National Parks, Uganda. Circles repre-
sent the foods (blue) and diet composition (red) of Bwindi gorillas, and
squares represent the foods (green) and diet (red) of Virunga gorillas.
The hollow circle and square represent the expected diet composition
of Bwindi and Virunga gorillas, respectively, if foods were eaten in pro-
portion to their availability. The line joining the outermost foods from
each site delineates the accessible space available to each gorilla pop-
ulation given its choice of foods. Despite the foods differing between
the sites, the composition of the diet ingested by the two populations of
gorillas was closely similar, but different from the expected diet if feed-
ing was proportional to availability. b Scatterplot of the relationship
between availability and percentage contribution of foods to the diets
of Bwindi (blue circles) and Virunga (green squares) gorillas. The lack
of positive correlation suggests that foods were not eaten in proportion
to their availability (Virunga: Spearman’s rho = 0.465, P = 0.094;
Bwindi: Spearman’s rho = 0.160, P = 0.682). Food availability within
the home ranges of Bwindi and Virunga gorillas was estimated by
Plumptre (1995) and by Ganas et al. (2004), respectively. Data from
Rothman et al. (2007) (colour figure online)
1 3
hemicellulose is partly digested (Van Soest 1994; Remis
2000; Remis and Dierenfeld 2004). Thirdly, hemicellulose
was extracted from leaves to a greater extent than fruits,
as shown by the higher lignin:hemicellulose ratio in leaf-
derived than fruit-derived faeces. Finally, the points for
foods aligned on a diagonal that is closer to the origin than
the points for faeces, showing that the faeces were depleted
of cellulose relative to the foods. Overall, this analysis
shows that hemicellulose is depleted in the faeces of goril-
las relative to lignin, and this depletion is more pronounced
when eating leaves than fruit, reflecting the fact that leaf
diets are more digestible than fruit diets, due to the large
seeds in fruits that are not digested (Rothman et al. 2008).
Cellulose, too, is depleted in the faeces compared with the
foods, but in this case there was no apparent difference
between fruits and leaves.
We have used published data on primates to illustrate
how RMTs can be used to address a range of questions in
nutritional ecology. These span from basic questions con-
cerning, for example, the variation in the composition of
primate milk, to questions of practical importance in the
conservation of biological diversity. For example, by bet-
ter understanding the ways in which animals balance their
nutrient intake needs, we are better informed about their
habitat needs. Bears on salmon streams are a good exam-
ple: while it had previously been thought that salmon intake
maximisation is the best strategy for bears, it has become
clear in recent studies that bears can minimise energetic
requirements by appropriately balancing their protein to
non-protein energy ratio (Erlenbach et al. 2014). Similarly,
a recent study showed that giant pandas migrate between
two habitats to balance their intake of calcium, phosphorus
and protein (Nie et al. 2014). In both cases, conservation
decisions would need to take into account the non-substi-
tutability of habitats on which these species rely to balance
their nutrition.
There are several reasons why nutritional geometry in
general, including RMT and ABNG, is a powerful tool
for addressing such questions. First, it provides a means
of conceptualising nutrition in more than one dimension,
thereby capturing both the independent and interactive
effects of nutrients on animals. Numerous studies have
shown that these interactive effects play a substantial role
in influencing animals––their behaviour, physiology, life
Fig. 5 Food and diet composition of Virunga (a) and Bwindi (b)
gorillas plotted separately (symbols and data as in Fig. 4a). Also
shown is the % contribution of each food to the diet, with the top-
ranking foods that jointly contributed approximately 60 % of
the diet highlighted in red boxes. The comparison shows that
58 % of the Virunga diet was comprised by two foods, one with a
protein:carbohydrate ratio that was considerably greater than,
and the other smaller than, that of the diet composition. This dem-
onstrates that the diet of Virunga gorillas was assembled to a
large extent through complementary feeding. By contrast, 60 %
of the diet of Bwindi gorillas was composed from 6 foods, all of
which had a protein:carbohydrate ratio that closely resembled
the diet composition. This suggests a stronger role in the selec-
tion by gorillas in Bwindi of foods that are balanced with respect to
protein:carbohydrate, with a minimal role for complementary feeding
(colour figure online)
1 3
history, ecology and evolution (Despland and Noseworthy
2006; Behmer and Joern 2008; Hawlena and Schmitz 2010;
Simpson and Raubenheimer 2012; Saravanan et al. 2012).
Second, the geometric space provides a device in which
salient components of the interaction of animal and envi-
ronment can be conceptually and quantitatively interrelated
in common, multi-dimensional, terms. Examples presented
here include foods, diets, and faecal composition, and oth-
ers have been discussed elsewhere (Raubenheimer et al.
2009; Simpson and Raubenheimer 2012). Third, nutritional
geometry is versatile, because different combinations of
axes (e.g. nutrients) and model components (e.g. foods,
diets) can be selected to address specific problems. Finally,
although not illustrated in the present paper, non-nutritional
variables (e.g. life-history responses to nutritional state)
can be incorporated into geometric models using response
surface methodology (Lee et al. 2008; Jensen et al. 2012;
Blumfield et al. 2012).
Amounts-based nutritional geometry has been applied
extensively in laboratory studies, yielding advances in,
among other fields, foraging theory (Raubenheimer et al.
2007, 2009), life-history theory (Lee et al. 2008; Maklakov
et al. 2009; Simpson and Raubenheimer 2010; Jensen et al.
2012), conservation (Raubenheimer and Simpson 2006),
causes of human obesity (Simpson et al. 2003; Simpson
and Raubenheimer 2005; Gosby et al. 2011; Raubenhe-
imer et al. 2014), and the design of feeds for agriculture
(Ruohonen et al. 2007) and companion animals (Hewson-
Hughes et al. 2011, 2012). The potential of nutritional
geometry for field studies with ecological applications is
receiving increasing attention (Raubenheimer et al. 2009,
2012; Simpson et al. 2010; Kearney et al. 2010, 2012;
Tait et al. 2014). A limitation, however, is that accurate
amounts-based data on animal foraging can be challenging
or impossible to collect in the field. To date, only three field
projects have succeeded in this respect: Peruvian spider
monkeys (Ateles chamek, Felton et al. 2009a, b), mountain
gorillas (Rothman et al. 2011) and chacma baboons (Papio
hamadryas ursinus; Johnson et al. 2013). For this reason,
Raubenheimer (2011) introduced the RMT as a propor-
tions-based modelling platform that is less demanding of
data and can be applied broadly in field studies. The exam-
ples presented here are intended to illustrate how concepts
developed in ABNG can be represented in RMT, including
foods, diets, nutritional regulation, and digestive efficien-
cies. Primates provide a fitting system, because the extent
to which they have been studied in the wild provides abun-
dant published data for our illustration.
This highlights a further strength of RMT, namely that
they can readily be applied to re-analyses of the substantial
proportions-based published data (e.g. of food composi-
tions), whereas there are fewer amounts-based data avail-
able in the literature (but see Simpson and Raubenheimer
1997 for rats, Raubenheimer and Simpson 1997 for chick-
ens, and Lee et al. 2008 for flies). This enables RMT to be
used in literature-based studies, as demonstrated by Rauben-
heimer and Rothman (2013) in their investigation of the
nutritional drivers of insectivory in humans and other pri-
mates. It is hoped that the present paper will both stimulate
synthetic studies of published data on animal foraging and
diet choice, and help to frame new studies that will contrib-
ute to the understanding of comparative nutritional ecology.
Acknowledgments We are grateful to Dr Alistair Senior for assis-
tance with the comparative analysis of primate milk compositions.
This research was partially funded by Faculty of Veterinary Science
Research Fund, The University of Sydney. D.R. is part-funded by
Gravida, The National Research Centre for Growth and Development,
New Zealand.
Conflict of interest The authors declare that they have no conflict
of interest.
Fig. 6 Use of RMT to investigate relative digestive efficiencies of
fibre components of dietary fruits (circles) and leaves (triangles) by
comparing the foods (symbols with black outline) with faeces (sym-
bols without black outline). Colours distinguish juveniles (red), adult
females (green) and silverback (blue) mountain gorillas. The data
show that, for adult female and silverback gorillas and both food
types, faeces were lower in hemicellulose (shift to the left) and cel-
lulose and higher in lignin relative to matched foods. The data also
suggest that the faeces derived from leaves (comparison of triangles
with and without black outline) were reduced in hemicellulose rela-
tive to lignin to a greater extent than faeces derived from fruits (com-
parison of circles with and without black outline). The same pattern
applies for leaves eaten by juveniles (comparison of red triangles
with and without black outline), whereas faeces derived from fruits
eaten by juveniles were enriched in both lignin and hemicellulose,
and depleted to a greater extent in cellulose (18 vs. 30 % for other
faecal samples) (comparison of red circles with and without black
outline) (colour figure online)
1 3
Barboza PS, Parker KL, Hume ID (2009) Integrative wildlife nutri-
tion. Springer, Berlin
Behmer ST, Joern A (2008) Coexisting generalist herbivores occupy
unique nutritional feeding niches. Proc Natl Acad Sci USA
Blumfield M, Hure A, Macdonald-Wicks LK, Smith R, Simpson SJ,
Raubenheimer D, Collins C (2012) The association between the
macronutrient content of maternal diet, adequacy of micronutri-
ents during pregnancy. Nutrients 4:1958–1976
Bowen SH, Lutz EV, Ahlgren MO (1995) Dietary protein and energy
as determinants of food quality: trophic strategies compared.
Ecology 76:899–907
Bryer MAH, Chapman CA, Rothman JM (2013) Diet and polyspe-
cific associations affect spatial patterns among redtail monkeys
(Cercopithecus ascanius). Behaviour 150:277–293
Chambers PG, Simpson SJ, Raubenheimer D (1995) Behavioural
mechanisms of nutrient balancing in Locusta migratoria. Anim
Behav 50:1513–1523
Conklin-Brittain NL, Wrangham RW, Hunt KD (1998) Dietary
response of chimpanzees and cercopithecines to seasonal variation
in fruit abundance. II: macronutrients. Int J Primatol 19:971–998
Dearing MD, Schall JJ (1992) Testing models of optimal diet assem-
bly by the generalist herbivorous lizard Cnemidophorus murinus.
Ecology 73:845–858
DeGabriel JL, Moore BD, Felton AM, Ganzhorn JU, Stolter C, Wallis
IR, Johnson CN, Foley WJ (2014) Translating nutritional ecology
from the laboratory to the field: milestones in linking plant chem-
istry to population regulation in mammalian browsers. Oikos
123:298–308. doi:10.1111/j.1600-0706.2013.00727.x
Despland E, Noseworthy M (2006) How well do specialist feeders
regulate nutrient intake? Evidence from a gregarious tree-feeding
caterpillar. J Exp Biol 209:1301–1309
Erlenbach JA, Rode KD, Raubenheimer D, Robbins CT (2014)
Macronutrient optimization and energy maximization determine
diets of brown bears. J Mammal 95:160–168
Fagan WF, Siemann E, Denno RF, Mitter C, Huberty AF, Woods HA,
Elser JJ (2002) Nitrogen in insects: implications for trophic com-
plexity and species diversification. Am Nat 160:784–802
Felton AM, Felton A, Raubenheimer D, Simpson SJ, Foley WJ, Wood
JT, Wallis IR, Lindenmayer DB (2009a) Protein content of diets
dictates the daily energy intake of a free-ranging primate. Behav
Ecol 20:685–690
Felton AM, Felton A, Wood JT, Foley WJ, Raubenheimer D, Wallis
IR, Lindenmayer DB (2009b) Nutritional ecology of Ateles cha-
mek in lowland Bolivia: how macronutrient balancing influences
food choices. Int J Primatol 30:675–696
Ganas J, Robbins MM, Nkurunungi JB, Kaplin BA, Mcneilage A
(2004) Dietary variability of mountain gorillas in Bwindi impen-
etrable national park, Uganda. Int J Primatol 25:1043–1072
Giri S, Aryal A, Koirala RK, Adhikari B, Raubenheimer D (2011)
Feeding ecology and distribution of Himalayan serow (Capricornis
thar) in Annapurna conservation area. Nepal World J Zool 6:80–85
Gosby AK, Conigrave AD, Lau NS, Iglesias MA, Hall RM, Jebb
SA, Brand-Miller JI, Caterson D, Raubenheimer D, Simpson SJ
(2011) Testing protein leverage in lean humans: a randomised
controlled experimental study. PLoS ONE 6:e25929
Hadfield JD, Nakagawa S (2010) General quantitative genetic meth-
ods for comparative biology: phylogenies, taxonomies and multi-
trait models for continuous and categorical characters. J Evol
Biol 23:494–508
Hawlena D, Schmitz OJ (2010) Physiological stress as a fundamental
mechanism linking predation to ecosystem functioning. Am Nat
Hewson-Hughes AK, Hewson-Hughes VL, Miller AT, Hall SR, Simp-
son SJ, Raubenheimer D (2011) Geometric analysis of macronu-
trient selection in the adult domestic cat, Felis catus. J Exp Biol
Hewson-Hughes AK, Hewson-Hughes VL, Colyer A, Miller AT, Hall
SR, Raubenheimer D, Simpson SJ (2012) Consistent propor-
tional macronutrient intake selected by adult domestic cats (Felis
catus), despite variations in dietary macronutrient and moisture
content of foods offered. J Comp Physiol B, pp 1–12
Hewson-Hughes AK, Hewson-Hughes VL, Colyer A, Miller AT,
McGrane SJ, Hall SR, Butterwick RF, Simpson S (2013) Geo-
metric analysis of macronutrient selection in breeds of the
domestic dog, Canis lupus familiaris. Behav Ecol 24:293–304
Hinde K, Milligan LA (2011) Primate milk: proximate mechanisms
and ultimate perspectives. Evol Anthr 20:9–23
Hyslop EJ (1980) Stomach contents analysis––a review of methods
and their application. J Fish Biol 17:411–429
Jensen K, Mayntz D, Toft S, Clissold FJ, Hunt J, Raubenheimer D,
Simpson SJ (2012) Optimal foraging for specific nutrients in
predatory beetles. Proc R Soc Lond B 279:2212–2218
Johnson CA, Raubenheimer D, Rothman JM, Clarke D, Swedell L
(2013) 30 Days in the life: daily nutrient balancing in a wild chacma
baboon. PLoS ONE 8:e70383. doi:10.1371/journal.pone.0070383
Kamler JF, Pope KL (2001) nonlethal methods of examining fish
stomach contents. Rev Fish Sci 9:1–11
Kearney M, Simpson SJ, Raubenheimer D, Helmuth B (2010) Model-
ling the ecological niche from functional traits. Philos Trans R
Soc Lond B 365:3469–3483
Kearney MR, Simpson SJ, Raubenheimer D, Kooijman SALM (2012)
Balancing heat, water and nutrients under environmental change:
a thermodynamic niche framework. Funct Ecol 4:950–966
Klare U, Kamler JF, Macdonald DW (2011) A comparison and cri-
tique of different scat-analysis methods for determining carnivore
diet. Mammal Rev 41:294–312
Lambert, JE (2010) Primate nutritional ecology: feeding biology
and diet at ecological and evolutionary scales. In: Campbell C,
Fuentes A, MacKinnon KC, Panger M, Bearder S (eds) Primates
in Perspective, 2nd edn. Oxford University Press, Oxford
Lee KP, Simpson SJ, Clissold FJ, Brooks R, Ballard JWO, Taylor PW,
Soran N, Raubenheimer D (2008) Lifespan and reproduction in
drosophila: new insights from nutritional geometry. Proc Natl
Acad Sci USA 105:2498–2503
Machovsky-Capuska GE, Dwyer SL, Alley MR, Stockin KA,
Raubenheimer D (2011) Evidence for fatal collisions and klep-
toparasitism while plunge diving in Gannets. Ibis 153:631–635
Maklakov AA, Hall MD, Simpson SJ, Dessmann J, Clissold FJ,
Zajitschek F, Lailvaux SP, Raubenheimer D, Bonduriansky R,
Brooks RC (2009) Sex differences in nutrient-dependent repro-
ductive ageing. Aging Cell 8:324–330
Milligan LA (2010) Milk composition of captive tufted capuchins
(Cebus apella). Am J Primatol 72:81–86
National Research Council (2003) Nutrient requirements of nonhu-
man primates, 2nd edn. National Academic Press, Washington
Nie Y, Zhang Z, Raubenheimer D, Elser JJ, Wei W, Wei F (2014)
Obligate herbivory in an ancestrally carnivorous lineage: the
giant panda and bamboo from the perspective of nutritional
geometry. Funct Ecol. doi:10.1111/1365-2435.12302
Paddack MJ, Cowen RK, Sponaugle S (2006) Grazing pressure of
herbivorous coral reef fishes on low coral-cover reefs. Coral
Reefs 25:461–472
Panthi S, Aryal A, Lord J, Adhikari B, Raubenheimer D (2012) Sum-
mer diet and habitat ecology of red panda (Ailurus fulgens ful-
gens) in Dhopatan hunting reserve. Nepal Zool Stud 51:701–709
Parker KL (2003) Advances in the nutritional ecology of cervids at
different scales. Ecoscience 10:395–411
1 3
Petry MV, Fonseca VSD, Scherer AL (2007) Analysis of stomach
contents from the black-browed albatross, Thalassarche melano-
phris, on the coast of Rio grande do sul, southern brazil. Polar
Biol 30:321–325
Plumptre AJ (1995) The chemical composition of montane plants and
its influence on the diet of large mammalian herbivores in the
Pare National des Volcans, Rwanda. J Zool 235:323–337
Power ML, Verona C, Ruiz-Miranda CE, Oftedal OT (2008) The
composition of milk from free-living common marmosets (Cal-
lithrix jacchus) in Brazil. Am J Primatol 70:78–83
Raubenheimer D (2011) Toward a quantitative nutritional ecology:
the right-angled mixture triangle. Ecol Monogr 81:407–427
Raubenheimer D, Jones SA (2006) Nutritional imbalance in an
extreme generalist omnivore: tolerance and recovery through
complementary food selection. Anim Behav 71:1253–1262
Raubenheimer D, Rothman JM (2013) The nutritional ecology of
entomophagy in humans and other primates. Annu Rev Entomol
Raubenheimer D, Simpson SJ (1993) The geometry of compensatory
feeding in the locust. Anim Behav 45:953–964
Raubenheimer D, Simpson SJ (1997) Integrative models of nutrient
balancing: application to insects and vertebrates. Nutr Res Rev
Raubenheimer D, Simpson SJ (2006) The challenge of supplementary
feeding: can geometric analysis help save the kakapo? Notornis
Raubenheimer D, Mayntz D, Simpson SJ, Toft S (2007) Nutrient-spe-
cific compensation following overwintering diapause in a gener-
alist predatory invertebrate: implications for intraguild predation.
Ecology 88:2598–2608
Raubenheimer D, Simpson SJ, Mayntz D (2009) Nutrition, ecology
and nutritional ecology: toward an integrated framework. Funct
Ecol 23:4–16
Raubenheimer D, Simpson SJ, Tait AH (2012) Match and mis-
match: conservation physiology, nutritional ecology and the
timescales of biological adaptation. Philos Trans R Soc Lond B
Raubenheimer D, Machovsky-Capuska GE, Gosby AK, Simpson S
(2014) The nutritional ecology of obesity: from humans to com-
panion animals. Br J Nutr. doi:10.1017/S0007114514002323
Remis MJ (2000) Initial studies on the contributions of body size and
gastrointestinal passage rates to dietary flexibility among gorillas.
Am J Phys Anthropol 112:171–180
Remis MJ, Dierenfeld ES (2004) Digesta passage, digestibility and
behavior in captive gorillas under two dietary regimens. Int J Pri-
matol 25:825–845
Robbins CT, Fortin JK, Rode KD, Farley SD, Shipley LA, Felicetti
LA (2007) Optimizing protein intake as a foraging strategy to
maximize mass gain in an omnivore. Oikos 116:1675–1682
Rode KD, Chapman CA, Mcdowell LR, Stickler C (2006) Nutritional
correlates of population density across habitats and logging inten-
sities in redtail monkeys (Cercopithecus ascanius). Biotropica
Rothman JM, Dierenfeld ES, Molina DO, Shaw AV, Hintz HF, Pell
AN (2006) Nutritional chemistry of foods eaten by gorillas in
Bwindi impenetrable national park, Uganda. Am J Primatol
Rothman JM, Plumptre AJ, Dierenfeld ES, Pell AN (2007) Nutritional
composition of the diet of the gorilla (Gorilla beringei): a com-
parison between two montane habitats. J Trop Ecol 23:673–682
Rothman JM, Dierenfeld ES, Hintz HF, Pell AN (2008) Nutritional
quality of gorilla diets: consequences of age, sex, and season.
Oecologia 155:111–122
Rothman JM, Raubenheimer D, Chapman CA (2011) Nutritional
geometry: gorillas prioritize non-protein energy while consuming
surplus protein. Biol Lett 7:847–849
Ruohonen K, Simpson SJ, Raubenheimer D (2007) A new approach
to diet optimisation: a re-analysis using European whitefish
(Coregonus lavaretus). Aquaculture 267:147–156
Saravanan S, Schrama JW, Figueiredo-Silva AC, Kaushik SJ, Verreth
JAJ, Geurden I (2012) Constraints on Energy Intake in Fish: the
Link between Diet Composition, Energy Metabolism, and Energy
Intake in Rainbow Trout. PLoS ONE 7:e34743
Schuckard R, Melville D, Cook W, Machovsky-Capuska GE (2012)
Diet of the Australasian gannet (Morus serrator) at Farewell Spit,
New Zealand. Notornis 59:66–70
Shrader AM, Owen-Smith N, Ogutu JO (2006) How a mega-grazer
copes with the dry season: food, nutrient intake rates by white
rhinoceros in the wild. Funct Ecol 20:376–384
Simpson SJ, Raubenheimer D (1993) A multi-level analysis of feed-
ing behaviour: the geometry of nutritional decisions. Philos Trans
R Soc Lond B 342:381–402
Simpson SJ, Raubenheimer D (1997) The geometric analysis of
macronutrient selection in the rat. Appetite 28:201–213
Simpson SJ, Raubenheimer D (2005) Obesity: the protein leverage
hypothesis. Obes Rev 6:133–142
Simpson SJ, Raubenheimer D (2010) The nutritional geometry of
aging. Springer, Berlin
Simpson SJ, Raubenheimer D (2012) The nature of nutrition: a unify-
ing framework from animal adaptation to human obesity. Prince-
ton University Press, Princeton
Simpson SJ, Batley R, Raubenheimer D (2003) Geometric analysis of
macronutrient intake in humans: the power of protein? Appetite
Simpson SJ, Raubenheimer D, Charleston MA, Clissold FJ (2010)
Modelling nutritional interactions: from individuals to communi-
ties. Trends Ecol Evol 25:53–60
Skibiel AL, Downing LM, Orr TJ, Hood WR (2013) The evolution
of the nutrient composition of mammalian milks. J Anim Ecol
Solon-Biet SM, Aisling CM, Ballard JWO, Ruohonen K, Wu LE,
Cogger VC, Warren A (2014) The ratio of macronutrients, not
caloric intake, dictates cardiometabolic health, aging, and longev-
ity in ad libitum-fed mice. Cell Metab 19:418–430
Sterner RW, Elser JJ (2002) Ecological stoichiometry: the biology of
elements from molecules to the biosphere. Princeton University
Press, Princeton
Tait A, Raubenheimer D, Stockin KA, Merriman M, Machovsky-
Capuska GE (2014) Nutritional geometry of gannets and the
challenges in field studies. Mar Biol 12:2791–2801. doi:10.1007/
Van Soest PJ (1994) Nutritional ecology of the ruminant. Cornell Uni-
versity Press, Ithaca
Westoby M (1974) An analysis of diet selection by large generalist
herbivores. Am Nat 108:290–304
Whittier CA, Milligan LA, Nutter FB, Cranfield MR, Power ML
(2010) Proximate composition of milk from free-ranging moun-
tain gorillas (Gorilla beringei beringei). Zool Biol 29:1–10
... Examples of diet balancing through HCF are taxonomically widespread, ranging from slime molds to insects, spiders, fish, rodents, domesticated cats, and humans (reviewed in Simpson and Raubenheimer, 2012). As discussed below, similar results have been reported from observational studies of primates in the wild, for example where populations eating different food combinations ingest a nutritionally similar diet (Rothman et al., 2007;Raubenheimer et al., 2015;Hou et al., 2020) and where individual animals achieve consistent nutrient intakes across foraging days from different food combinations (Johnson et al., 2013). ...
... That this is driven by active nutrient selection, rather than passively by the distribution of nutrients within the food environment, is suggested by evidence that macronutrient intakes remain constant when feeding on different combinations of foods. For example, the annual diets of mountain gorilla populations in Bwindi and Virunga national parks are nutritionally similar even though they subsist on different food combinations (Rothman et al., 2007;Raubenheimer et al., 2015). At the level of individuals, a female chacma baboon fed on different combinations of ca. ...
Full-text available
Animals require specific blends of nutrients that vary across the life course and with circumstances, e.g., health and activity levels. Underpinning and complicating these requirements is that individual traits may be optimised on different dietary compositions leading to nutrition-mediated trade-offs among outcomes. Additionally, the food environment may constrain which nutrient mixtures are achievable. Natural selection has equipped animals for solving such multidimensional, dynamic challenges of nutrition, but little is understood about the details and their theoretical and practical implications. We present an integrative framework, nutritional geometry, which models complex nutritional interactions in the context of multiple nutrients and across levels of biological organization (e.g., cellular, individual, population) and levels of analysis (e.g., mechanistic, developmental, ecological, evolutionary). The framework is generalizable across different situations and taxa. We illustrate this using examples spanning insects to primates and settings (laboratory, and the wild), and illustrate its relevance for human health.
... First, the nutrient balance model framed in nutritional geometry proposes that animals achieve an optimal balance of nutrients to meet their various requirements ('nutritional intake target' from now on) by mixing nutritionally complementary foods. In this model, foraging is primarily a process aimed at balancing multiple nutrients, not of energy acquisition per se (Raubenheimer et al., 2015;Simpson & Raubenheimer, 2012). Second, the energy maximisation model within optimal foraging theory (Stephens et al., 2006) states that animals are committed to maximise their daily energy intake ('energy intake target' from now on), in which they direct the search for food towards those items that are more energetically profitable. ...
According to diet‐regulation hypotheses, animals select food to regulate the intake of macronutrients or maximise energy feeding efficiency. Specifically, the nutrient balance model proposes that foraging is primarily a process of balancing multiple nutrients to achieve a nutritional intake target, while the energy maximisation model proposes that foraging aims to maximise energy. Here, we evaluate the adjustment of fruit diets (the fruit‐derived component of the diets) to nutritional and energy intake targets, characterizing the nutrient balance and energy maximisation strategies across fruit‐eating bird species with different fruit‐handling behaviours ("gulpers", which swallow whole fruits, and "mashers", which process the fruit in the beak) in subtropical Andean forests. Food‐handling behaviour determines the food intake rate and, consequently, influences animal efficiency to obtain nutrients and energy. We used extensive field data from the diet of fruit‐eating birds to test how species adjust their food intake. We used nutritional geometry to explore macronutrient balance and the effectiveness framework to explore energy‐acquisition effectiveness. Observed diets showed a good fit with predictions of a diet balanced in macronutrient proportions. With few exceptions, diets clustered near an optimal macronutrient mixture and did not differ from each other in terms of maximising energy intake. Moreover, when comparing our results with a random diet based on local fruit availability, birds tended to fit better to the nutritional target, and less to the energy target, than expected from a random diet. Fruit‐handling behaviour did not affect the ability of bird species to reach a nutritional target but it affected species energy acquisition, which was lower in mashers than in gulpers. This study explores for the first time different diet‐regulation strategies in wild fruit‐eating birds, and supports the argument that the diet reflects a specific regulation of macronutrients. Understanding why birds select fruits is a complex question requiring multiple considerations. The nutrient balance model explains the relevance of nutrient composition in the fruit selection by fruit‐eating birds, although it is still necessary to determine its relative importance with respect to other dietary drivers.
... Although much has been done to use field-based feeding observations to understand the nutrient targets of primates (Felton et al., 2009;Norconk & Conklin-Brittain, 2004;Raubenheimer et al., 2015;Rothman et al., 2011;Takahashi et al., 2019;Uwimbabazi et al., 2021;Wrangham et al., 1998), there is a striking absence of data on the effect that specific PSMs might play in either influencing food choice or modifying nutrient targets. In contrast to studies with marsupials (Marsh et al., 2015), rodents (Sorensen et al., 2005), lagomorphs (Bryant et al., 1983), and ungulates (Stolter et al., 2005), few studies on primates have focussed on specific, characterized secondary metabolites. ...
Full-text available
The role of plant secondary metabolites (PSMs) in shaping the feeding decisions, habitat suitability, and reproductive success of herbivorous mammals has been a major theme in ecology for decades. Although primatologists were among the first to test these ideas, studies of PSMs in the feeding ecology of non-human primates have lagged in recent years, leading to a recent call for primatologists to reconnect with phytochemists to advance our understanding of the primate nutrition. To further this case, we present a formal meta-analysis of diet choice in response to PSMs based on field studies on wild primates. Our analysis of 155 measurements of primate feeding response to PSMs is drawn from 53 studies across 43 primate species which focussed primarily on the effect of three classes of PSMs tannins, phenolics, and alkaloids. We found a small but significant effect of PSMs on the diet choice of wild primates, which was largely driven by the finding that colobine primates showed a moderate aversion to condensed tannins. Conversely, there was no evidence that PSMs had a significant deterrent effect on food choices of non-colobine primates when all were combined into a single group. Furthermore, within the colobine primates, no other PSMs influenced feeding choices and we found no evidence that foregut anatomy significantly affected food choice with respect to PSMs. We suggest that methodological improvements related to experimental approaches and the adoption of new techniques including metabolomics are needed to advance our understanding of primate diet choice. K E Y W O R D S meta-analysis, metabolomics, phylogeny, plant secondary metabolites, tannins Am J Primatol. 2022;e23397. |
... Large trees produce fruit, flowers, or leaves for several consecutive weeks and are not completely depleted after a foraging bout by a single group [24]. Moreover, because most primates travel hundreds or thousands of metres each day to feeding patches of different plant species in order to meet their nutritional goals [25], groups are unlikely to be able to monopolize a single resource for an entire day or fruiting period. As a result, these resources may engender intergroup contests in which both winners and losers reap foraging benefits. ...
The energetic costs and benefits of intergroup conflicts over feeding sites are widely hypothesized to be significant, but rarely quantified. In this study, we use short-term measures of energy gain and expenditure to test whether winning an intergroup encounter is associated with greater benefits, and losing with greater costs. We also test an alternative perspective, where groups fight for access to large food sources that are neither depletable nor consistently monopolizable: in this case, a group that has already fed on the resource and is willing to leave first (the loser) is supplanted by a newly arrived group (the winner). We evaluate energy balance and travel distance during and after encounters for six groups of red-tailed monkeys in Kibale National Park, Uganda. We find that winning groups experience substantial energetic benefits, but do so to recoup from earlier deficits. Losing groups, contrary to predictions, experience minimal energetic costs. Winners and losers are predictable based upon their use of the contested resource immediately before the encounter. The short-term payoffs associated with these stressful conflicts compensate for any associated costs and support the perception that between-group contests are an important feature of social life for species that engage in non-lethal conflicts. This article is part of the theme issue ‘Intergroup conflict across taxa’.
... This makes this an example of "hunting by expectation" (Hodges, 1981), a situation also known for nectar feeding with tamarins (Saguinus mystax and Saguinus fuscicollis : Garber, 1988). Thus, while it is important to consider the nutrient-balancing framework of primate foraging decision-making (Raubenheimer et al., 2009(Raubenheimer et al., , 2015, this and optimality are complimentary, because spatially aggregated foods must be found and then, once located, distinguished between for the highest nutrient reward with the lowest time/energy expenditure. ...
Full-text available
Fruit pulp is an easily handled energy source for many frugivorous species but generally has little protein. Accordingly, ripe-fruit specialist primate species with diets dominated by fruit pulp risk protein deficiency. While some species use leaf and flower buds, young leaves, and arthropods as an alternative protein supplement, highly frugivorous spider monkeys (Ateles spp.) use protein-rich young leaves and/or fig fruits. However, not all spider monkey populations have access to abundantly available figs. Comparing infestation frequencies of fruits on trees with those eaten by spider monkeys, we tested the hypothesis that, under such circumstances, spider monkeys preferentially choose those nonfig fruits with pulp infested by insect larvae (a highly protein-rich resource). We predicted that: (i) a large proportion of plant species eaten by Ateles would have insect larvae-infested fruits; and (ii) Ateles would actively select infested fruits. We tested these predictions with Ateles chamek and Ateles marginatus on the banks of the Tapajós River, Brazil. Across a 13-month sampling period, we recorded 27 plant species in the diet of the 2 Ateles species. Of these, 23 (85%) had larvae-infested fruits when sampled; 11 species (40%) had high levels of individual fruits infested (35-78%). We used Ivlev Values to quantify selectivity for infested/uninfested fruits in 20 plant species. Infested fruits were positively selected in 12 species (60%), while aversion to infested fruits occurred in 4 species (20%). This covert carnivory/faunivory in spider monkeys is a largely overlooked aspect of their feeding ecology. This situation would be nearly impossible to ascertain from behavioral observations alone, showing the value of integrated, multimethod approaches. The strategy used by Ateles spp. on the banks of the Tapajós highlights the flexibility of primate foraging choices and the importance of indirect source of protein to ripe-fruit specialist primates.
... Right-angled mixture triangles (RMTs) are two-dimensional representations of threecomponent mixtures (e.g., individual foods, observed diets) that provide a proportions-based approach rooted in nutritional geometry shown to be highly effective in illustrating field-based research (Raubenheimer, 2011;Raubenheimer et al., 2015). For example, three components in the standard two-dimensional Euclidean plot are represented by the X, Y , and Z axes, where Z is the implicit axis (= 100% − y-value − xvalue), consequently causing the focal axes (X and Y ) to be constrained along the line intersecting the point (100 − Z%) on each axis (Raubenheimer, 2011). ...
Full-text available
Flowers are ubiquitous in primate environments, yet their nutritional advantages are underexamined. Symphonia globulifera is a widely distributed tree exploited by a variety of animals in Africa and the Americas. We collected S. globulifera flower samples consumed by red-tailed monkeys ( Cercopithecus ascanius ) and compared them nutritionally to flower samples from other plant species in Kibale National Park, Uganda. Flowers were assayed for three fiber fractions (NDF, ADF, lignin), fat, crude protein, acid detergent insoluble nitrogen (ADIN), ash, and soluble sugars. We estimated available protein, total nonstructural carbohydrates (TNC), and metabolizable energy (ME). We calculated the mean and standard deviation for all nutrient categories and applied nutritional geometry to illustrate the balance among the energetic gains from available protein, fat, fiber, and TNC across flower species. Our results suggest that S. globulifera flowers provide an unusually high fat resource (14.82% ± 1.41%) relative to other flowers (1.38% ± 5.79%) and other foods exploited in the same habitat.
... Compared to a standard medium with a usual sucrose content of 5% (w/v), an HSD contains an elevated amount of sucrose, glucose or fructose, often with a total sugar content of approximately 20-30% (w/v) [126]. Based on the geometry of nutrition, this results in proportionally lower levels of proteins and fats [127]. ...
Full-text available
The model organism Drosophila melanogaster was increasingly applied in nutrition research in recent years. A range of methods are available for the phenotyping of D. melanogaster, which are outlined in the first part of this review. The methods include determinations of body weight, body composition, food intake, lifespan, locomotor activity, reproductive capacity and stress tolerance. In the second part, the practical application of the phenotyping of flies is demonstrated via a discussion of obese phenotypes in response to high-sugar diet (HSD) and high-fat diet (HFD) feeding. HSD feeding and HFD feeding are dietary interventions that lead to an increase in fat storage and affect carbohydrate-insulin homeostasis, lifespan, locomotor activity, reproductive capacity and stress tolerance. Furthermore, studies regarding the impacts of HSD and HFD on the transcriptome and metabolome of D. melanogaster are important for relating phenotypic changes to underlying molecular mechanisms. Overall, D. melanogaster was demonstrated to be a valuable model organism with which to examine the pathogeneses and underlying molecular mechanisms of common chronic metabolic diseases in a nutritional context.
Dietary protein and digestible carbohydrates are two key macronutrients for insect herbivores, but the amounts and ratios of these two macronutrients in plant vegetative tissues can be highly variable. Typically, insect herbivores regulate their protein-carbohydrate intake by feeding selectively on nutritionally complementary plant tissues, but this may not always be possible. Interestingly, lab experiments consistently demonstrate that performance – especially growth and survival – does not vary greatly when caterpillars and nymphal grasshoppers are reared on diets that differ in their protein-carbohydrate content. This suggests insect herbivores employ post-ingestive physiological mechanisms to compensate for variation in diet protein-carbohydrate profile. However, the molecular mechanisms that underlie this compensation are not well understood. Here we explore, for the first time in an insect herbivore, the transcriptional effects of two dietary factors: protein-to-carbohydrate ratio (p:c) and total macronutrient (p + c) content. Specifically, we reared Helicoverpa zea caterpillars on three diets that varied in diet p:c ratio and one diet that varied in total p + c concentration, all within an ecologically-relevant range. We observed two key findings. Caterpillars reared on diets with elevated total p + c content showed large differences in gene expression. In contrast, only small differences in gene expression were observed when caterpillars were reared on diets with different p:c ratios (spanning from protein-biased to carbohydrate-biased). The invariable expression of many metabolic genes across these variable diets suggests that H. zea caterpillars employ a strategy of constitutive expression to deal with protein-carbohydrate imbalances rather than diet-specific changes. This is further supported by two findings. First, few genes were uniquely associated with feeding on a protein- and carbohydrate-biased diet. Second, many differentially-expressed genes were shared across protein-biased, carbohydrate-biased, and concentrated diet treatments. Our study provides insights into the post-ingestive physiological mechanisms insect herbivores employ to regulate protein-carbohydrate intake. Most notably, it suggests that H. zea, and perhaps other generalist species, use similar post-ingestive mechanisms to deal with protein-carbohydrate imbalances – regardless of the direction of the imbalance.
The fundamental biological drivers of dietary intake are no different in humans than other species, from insects in laboratory studies to wild primates in natural ecologies. In this chapter, the authors show how research, initially on insects and subsequently many other species (from single cellular slime molds to apes in the wild), has suggested a new ecologically inspired approach for understanding the roles of biology, environment, and their interactions in driving the obesity epidemic, and potentially identifying solutions. They explain the theoretical foundations for the approach, illustrate its application to addressing relevant questions in some non‐human species, and show how it has been applied in studies of humans. Recent research suggests that the protein leverage hypothesis might provide a new approach for integrating with existing public health frameworks to understand how human biology interacts with transitioning food environments to generate epidemics of obesity and associated disease.
Full-text available
Food supply is one of the major drivers of animal behaviour and the gut microbiome is an important mediator between food supply and its effects on physiology. However, predicting the outcome of diet change on microbiome and consequences for the animal has proven extremely challenging. We propose this reflects processes occurring at different scales. Inadequate accounting for the multi-level complexity of nutrition (nutrients, foods, diets) obscures the diet influence on microbiome and subsequently animal. Here we present a detailed year-round, multi-level analysis of diet and microbiome changes in a wild population of a temperate primate, the rhesus macaque (Macaca mulatta). Total daily food and nutrient intake of six male and six female macaques was monitored in each the four seasons (total 120 days observations). For each individual we found significant variation in the microbiome between all four seasons. This response was more strongly correlated with changes in macronutrient intake than with food items and much of the response could be explained at the level of six ecological guilds – sets of taxa sharing similar responses to nutrient intake. We conclude that study of diet, microbiome and animal performance in ecology will more effectively identify patterns if diet is recorded at the level of nutrient intake. Although microbiome response to diet does show variation in species-level taxa in response to food items, there is greater commonality in response at the level of guilds. A goal for microbiome researchers should be to identify genes encoding microbial attributes that can define such guilds. This article is protected by copyright. All rights reserved
Full-text available
The diet of the Australasian gannet (Morus serrator) at Farewell Spit, New Zealand, was studied by the analysis of 70 regurgitations collected from the 1995 to 2001 breeding seasons. Surface schooling pilchard (Sardinops neopilchardus) was the main prey, followed by anchovy (Engraulis australis). The composition of the diet was similar in most seasons examined except in 1996 in which anchovy was the main prey item. Such a change in diet could be linked with a pilchard mass mortality in New Zealand in August 1995. The estimated annual prey consumption by birds at the Farewell Spit gannetry was 852 tonnes. Although annual catches of pilchard and anchovy by commercial fsheries in the area are still relatively small, an increase may interfere with prey availability, and in turn, increase competition between marine predators and infuence the breeding success. Our analyses of diet are consistent with previous studies showing that Australasian gannets as fexible foragers and they highlight their importance as bioindicators of fsh stocks in New Zealand.
Disparities in nutrient content (nitrogen and phosphorus) between herbivores and their plant resources have lately proven to have major consequences for herbivore success, consumer‐driven nutrient cycling, and the fate of primary production in ecosystems. Here we extend these findings by examining patterns of nutrient content between animals at higher trophic levels, specifically between insect herbivores and predators. Using a recently compiled database on insect nutrient content, we found that predators exhibit on average 15% greater nitrogen content than herbivores. This difference persists after accounting for variation from phylogeny and allometry. Among herbivorous insects, we also found evidence that recently derived lineages (e.g., herbivorous Diptera and Lepidoptera) have, on a relative basis, 15%–25% less body nitrogen than more ancient herbivore lineages (e.g., herbivorous Orthoptera and Hemiptera). We elaborate several testable hypotheses for the origin of differences in nitrogen content between trophic levels and among phylogenetic lineages. For example, interspecific variation in insect nitrogen content may be directly traceable to differences in dietary nitrogen (including dilution by gut contents), selected for directly in response to the differential scarcity of dietary nitrogen, or an indirect consequence of adaptation to different feeding habits. From some functional perspectives, the magnitude rather than the source of the interspecific differences in nitrogen content may be most critical. We conclude by discussing the implications of the observed patterns for both the trophic complexity of food webs and the evolutionary radiation of herbivorous insects.
Foraging deficiencies and supplementary feeding play critical roles in kakapo (Strigops habroptilus) breeding biology and conservation. We present a framework for the analysis of complex nutritional data (called the geometric framework - GF) which may contribute further understanding of the relationships between natural foods, supplementary feeding and kakapo reproduction. We outline the basic concepts of the approach, and illustrate its application using data for the protein, lipid and calcium content of a natural food (green fruits of rimu Dacrydium cupressinum) and a supplementary feed ("muesli"). We provide some pointers for the broader application of GF to the problem of kakapo supplementary feeding, and close with a brief review of a literature which suggests that calcium might be a key limiting factor in kakapo reproduction. We hypothesise that supplementary foods with low macronutrient:calcium ratios are likely to be most effective in supporting increased reproduction.
Nutrition has long been considered more the domain of medicine and agriculture than of the biological sciences, yet it touches and shapes all aspects of the natural world. The need for nutrients determines whether wild animals thrive, how populations evolve and decline, and how ecological communities are structured.The Nature of Nutritionis the first book to address nutrition's enormously complex role in biology, both at the level of individual organisms and in their broader ecological interactions. Stephen Simpson and David Raubenheimer provide a comprehensive theoretical approach to the analysis of nutrition--the Geometric Framework. They show how it can help us to understand the links between nutrition and the biology of individual animals, including the physiological mechanisms that determine the nutritional interactions of the animal with its environment, and the consequences of these interactions in terms of health, immune responses, and lifespan. Simpson and Raubenheimer explain how these effects translate into the collective behavior of groups and societies, and in turn influence food webs and the structure of ecosystems. Then they demonstrate how the Geometric Framework can be used to tackle issues in applied nutrition, such as the problem of optimizing diets for livestock and endangered species, and how it can also help to address the epidemic of human obesity and metabolic disease.
Strategies that cervid species use to meet nutritional requirements may vary depending on different spatial and temporal scales. The objective of this review is to provide an overview of contributions made to the field of nutritional ecology over approximately the last 10 y, specifically with respect to cervids. Mechanistic studies using captive animals in confined environments have established cause-and-effect relationships among foraging and digestive efficiencies, food availability and qualities, changes in body mass and condition, and reproduction. Studies using tamed free-ranging cervids in natural habitats have assessed the integrated effects of these reductionist components to validate animal-habitat interactions. At landscape levels, however, wild cervids are not as predictable because behaviours reflect trade-offs between many more variables. Geographic information systems allow for analyses of nutritional values of habitats as broadly indexed by remote sensing imagery, changes in the rates of animal movements (from global positioning satellite-collared animals) in relation to nutritional, energetic, and predation risk factors, and impacts of anthropogenic disturbances. Stable isotopes and fatty acids offer the potential to define spatial and temporal patterns of movements and diets. Further research is needed to determine whether stress hormones and/or nitrogen metabolites are able to provide links with animal condition and population variation. Our challenge is to understand nutritional ecology enough that we can manage for the flexibility in strategies that cervids with different trade-offs on different landscapes use to survive and reproduce.