Exploring patterns and mechanisms of interspecifi c and intraspecifi c
variation in body elemental composition of desert consumers
Ang é lica L. Gonz á lez , Jos é Miguel Fari ñ a , Adam D. Kay, Raquel Pinto and Pablo A. Marquet
A. L. Gonz á lez (firstname.lastname@example.org), J. M. Fari ñ a and P. A. Marquet, Center for Advanced Studies in Ecology and Biodiversity and Depto
de Ecolog í a, Pontifi cia Univ. Cat ó lica de Chile, Santiago. Chile. ALG and PAM also at: Inst. of Ecology and Biodiversity (IEB), Casilla 653,
Santiago, Chile. JMF also at: Dept of Ecology and Evolutionary Biology, Brown Univ., Providence, RI, USA. – A. D. Kay, Dept of Biology.
Univ. of St. Th omas. St. Paul, MN, USA. – R. Pinto, Univ. Arturo Prat, Iquique, Chile .
Key processes such as trophic interactions and nutrient cycling are often infl uenced by the element content of organ-
isms. Previous analyses have led to some preliminary understanding of the relative importance of evolutionary and
ecological factors determining animal stoichiometry. However, to date, the patterns and underlying mechanisms of
consumer stoichiometry at interspecifi c and intraspecifi c levels within natural ecosystems remain poorly investigated.
Here, we examine the association between phylogeny, trophic level, body size, and ontogeny and the elemental com-
position of 22 arthropod as well as two lizard species from the coastal zone of the Atacama Desert in Chile. We found
that, in general, whole-body P content was more variable than body N content both among and within species. Body
P content showed a signifi cant phylogenetic signal; however, phylogeny explained only 4% of the variation in body P
content across arthropod species. We also found a signifi cant association between trophic level and the element content
of arthropods, with carnivores having 15% greater N and 70% greater P contents than herbivores. Elemental scaling
relationships across species were only signifi cant for body P content, and even the P content scaling relationship was not
signifi cant after controlling for phylogeny. P content did decrease signifi cantly with body size within most arthropod
species, which may refl ect the size dependence of RNA content in invertebrates. In contrast, larger lizards had higher P
contents and lower N:P ratios than smaller lizards, which may be explained by size-associated diff erences in bone and
scale investments. Our results suggests that structural diff erences in material allocation, trophic level and phylogeny can
all contribute to variation in the stoichiometry of desert consumers, and they indicate that the elemental composition of
animals can be useful information for identifying broad-scale linkages between nutrient cycling and trophic interactions
in terrestrial food webs.
Oikos 120: 1247–1255, 2011
© 2011 Th e Authors. Oikos © 2011 Nordic Society Oikos
Subject Editor: Daniel Gruner. Accepted 19 December 2010
A central goal in ecological stoichiometry is to determine
how food web dynamics and rates of nutrient cycling are
constrained by the interaction between local nutrient avail-
ability and the requirements of multiple elements by organ-
isms (Sterner and Elser 2002, Moe et al. 2005). A fi rst step
in understanding these constraints is to identify patterns of
variation in the elemental composition of the organisms that
may infl uence ecosystem processes. Despite considerable
advances in understanding ecologically relevant variation in
the elemental composition of primary producers, studies on
consumer stoichiometry are still relatively rare- particularly
in terrestrial systems (Fagan et al. 2002, Sterner and Elser
2002, Schade et al. 2003, Woods et al. 2004, Kay et al. 2006,
Hamb ä ck et al. 2009, Mulder and Elser 2009). Identifying
patterns of stoichiometric variation in terrestrial consumers,
and determining their causes and consequences, should help
clarify how nutrient cycling aff ects and is aff ected by the
composition of terrestrial communities.
Previous studies on terrestrial consumers have shown
signifi cant phylogenetic and trophic level diff erences in
whole-body content of major elements (carbon (C), nitro-
gen (N), and phosphorus (P)) among arthropod taxa (Fagan
et al. 2002, Denno and Fagan 2003, Woods et al. 2004).
For instance, Fagan et al. (2002) found that recently derived
groups of terrestrial insects (Diptera and Lepidoptera) have
lower body N contents than more ancestral insect orders, and
insect predators have higher body N content than herbivore.
In addition, body P content has been shown to diff er among
trophic levels, although patterns of P content variation diff er
among systems (Woods et al. 2004, Martinson et al. 2008).
A central hypothesis in ecological stoichiometry theory is
the growth rate hypothesis (Sterner and Elser 2002), which
proposes that diff erences in whole-body P content among
invertebrate organisms results from diff erential allocation to
P-rich RNA needed for rapid growth. Consistent with this
hypothesis, faster-growing invertebrates tend to have higher
P-content than (otherwise similar) slower-growing organ-
isms (Sterner and Elser 2002). Th e growth rate hypothesis
also predicts size-related variation in invertebrate P content:
because small-bodied organisms tend to have faster growth
rates than larger organisms, body P content should decrease
with body size across species (Elser et al. 1996, Sterner and
Elser 2002, however see Gillooly et al. 2005). Tests of this
prediction have shown mixed results (Elser et al. 2000,
Woods et al. 2004, Gillooly et al. 2005, Bertram et al. 2008,
Martinson et al. 2008). For example, Cross et al. (2003) and
Woods et al. (2004) have reported that both N and P contents
in invertebrates decrease with body size, whereas Martinson
et al. (2008) found no comparable pattern in terrestrial detri-
tivores. Because of the inconsistencies, additional informa-
tion is needed to determine the mechanisms of size-related
variation in elemental stoichiometry of invertebrates.
Th e elemental composition of organisms can also vary
substantially due to diff erential allocation to other major
biomolecules (e.g. protein, lipids) or structural materials
(e.g. chitin, scales and bones) (Elser et al. 1996, Sterner and
George 2000, Bertram et al. 2008). In addition, individual
organisms often show changes in element content with onto-
genetic development (Sterner and Elser 2002, Mulder and
Bowden 2007). Variation in elemental composition across
ontogeny in invertebrates appears to be infl uenced by the
higher P demand associated with faster growth rates in lar-
val stages (Elser et al. 1996, Sterner and Elser 2002, Schade
et al. 2003). Furthermore, intraspecifi c variation in element
content may result from spatial and/or temporal changes in
the nutrient content of food resources (Persson et al. 2010).
Th ese deviations from strict homeostasis might account for
signifi cant elemental composition variation among individu-
als (Persson et al. 2010), and make diff erences among species
more diffi cult to detect.
Th e variation in empirical fi ndings across studies could be
partly explained by the fact that most studies have relied on
compiled metadata (Fagan et al. 2002, Hamb ä ck et al. 2009,
although see Woods et al. 2004). No previous studies have
examined mechanisms explaining both body N and P con-
tent of multiple invertebrate and vertebrate consumers at the
species level within a single system. Moreover, very few stud-
ies have examined the extent of interspecifi c and intraspecifi c
variation in elemental stoichiometry in natural ecosystems.
Variation in the elemental stoichiometry of organisms could
lead to important constraints in major life history traits and
in turn aff ect key ecological processes, such as trophic inter-
actions and nutrient cycling (Sterner and Elser 2002).
In this study, we examined variation in body C, N and
P content among and within species of desert consum-
ers from four well-delimited food webs from coastal areas
of the Atacama Desert, Chile. We focus on the dominant
arthropod species and on two lizard species: the only verte-
brate predators that occur in these ecosystems. Th ese natural
ecosystems that support suffi ciently simple and conspicuous
food webs provide us with a unique system for studying the
ecological stoichiometry of consumers in whole food webs.
Here we test general predictions that, across taxa, N and P
contents is higher (1) in predators than in herbivores, (2) in
more recently derived arthropod orders than in older orders,
and (3) in smaller-bodied taxa. We also test, within species,
whether (4) N and P contents are higher in early ontogenetic
stages than in adults, (5) N and P contents decrease with
body size in arthropods, and (6) N and P contents increase
with body size in vertebrates.
Material and methods
Arthropods and lizards were collected from four sites in the
coastal zone of the Atacama Desert, Chile. Th e area is charac-
terized by a relatively constant temperature (annual average
18.1 ° C, annual max. ? 20.9 ° C, annual min. ? 15.9 ° C), a
mean annual rainfall of 1.7 mm over the last century (Direc-
ci ó n Meteorol ó gica de Chile), and by the occurrence of fog
events mainly during the austral winter (June to August,
Cereceda et al. 2008). Fog originates from thick stratocumu-
lus banks below 1000 m which, when intercepted by isolated
mountaintops or steep coastal slopes of the Coastal Cordil-
lera, generates a fog immersion zone (Pinto et al. 2006,
Cereceda et al. 2008). Th e increased air moisture and depo-
sition in the fog zone is correlated with the development
of isolated vegetation ‘ islands ’ that spreads inland through
low elevation valleys or passes where fog is also frequent
(Pinto et al. 2006, Cereceda et al. 2008). Several terrestrial
bromeliad species of the genus Tillandsia inhabit southern
Per ú and northern Chile, which depend exclusively on fog
inputs as their primary water source (Pinto et al. 2006). Th e
most noticeable one T. landbeckii invades sandy soils cover-
ing vast areas and forming specialized communities called
‘ tillandsiales ’ . At each of the four T. landbeckii stand sites
we collected arthropods belonging to 22 species (i.e. Insecta
and Arachnida) and two lizard species. All individuals col-
lected were cooled on ice until and taken to the laboratory
for analysis. Arthropod ’ s guts were not removed prior to
chemical analyses due to their small size, therefore we kept
them at 4 ° C overnight to allow their gut content evacua-
tion (Evans-White et al. 2005). In contrast, we removed the
digestive tract of the lizards prior to analyses. Before chemi-
cal analyses the individuals were counted and identifi ed to
the level of genus or species. We also classifi ed arthropods by
trophic level using information from the literature and fi eld
observations. We collected larval and adult stages from true
fl ies ( Musca sp., Diptera), which we used to analyze the eff ect
of ontogenetic stage on body elemental content within this
Sample preparation and nutrient determination
We determined dry mass of individuals using an electronic
balance ( ? 0.1 μ g). We measured the phosphorus concentra-
tion in whole arthropods and whole lizards using potassium
persulfate and sulfuric acid digestion followed by ascorbate –
molybdate colorimetry. Prior to digestion, we gently crushed
individuals with a Tefl on-coated rod while in solution to
expose internal tissues to reagents. We performed all P analy-
ses using a fl ow solution autoanalyzer.
We determined the percent of N and C using a NC ana-
lyzer that involves complete combustion of samples. For
smaller arthropods ( ? 5 mg dry mass), we performed NC
analyses on whole individuals. For larger arthropods and
lizards, we analyzed NC levels in subsamples from dried
individuals that were fi rst crushed with a mortar and pestle.
We determined the percent recovery in P and NC assays by
comparison to bovine muscle standards from the Natl Inst.
of Standards and Technology (NIST 8414). We use the term
‘ P (or C or N) content ’ to describe P (or C or N) content as
a percent of dry body mass. Th e C:N:P ratios were calculated
in molar units. Th e analyzed number of arthropods and liz-
ards per species ranged from 1 to 42.
We tested the eff ect of phylogeny on body elemental content
(C, N, P and C:N ratios) and body size by calculating phy-
logenetic independent contrasts (PICs; Felsenstein 1985).
Th e PICs are widely used to correct statistically for poten-
tial non-independence of observations (e.g. among species)
due to their phylogenic relationships (Felsenstein 2008).
Because no comprehensive arthropod phylogeny is available
at species level for all our arthropods, we constructed a com-
posite tree including all 22 species of arthropods describing
the hypothesized evolutionary relationships among them,
using Mesquite 2.72 for phylogenetically based statistical
analyses ( ? http://mesquiteproject.org/mesquite/mesquite.
html ? ). To construct arthropod tree topology we followed
Regier et al. (2010) because this is the most complete and
updated order-level molecular phylogeny available. Below
order level, we used Flook et al. (2000) for Orthoptera; Hunt
et al. (2007) for Coleoptera; Han and Ro (2009) and Nihei
and Barros de Carvhalo (2007) for Diptera; phylogenetic
relationships within Lepidoptera were taken from Wahlberg
et al. (2005) and Mitchell et al. (2000). Th e phylogeny of
Arachnidae was drawn from Shultz (2007); Araneae from
Coddington and Levi (1991); and Acari phylogeny was
based on Dabert et al. (2010). For the few unresolved rela-
tionships within genus and species for which we had no
information, phylogenetic relationships were inferred from
current taxonomy (Felsenstein 1985, Garland et al. 2005).
We calculated independent contrasts for continuous traits
(i.e. N and P contents, C:N ratios and body size) using
the Phenotypic Diversity Analysis Program for Mesquite.
We assumed branch lengths were constant (equal to one)
as this method met assumptions of PICs analyses (Garland
et al. 1992). Contrasts were standardized and positivized
on the x-axis and linear regressions performed through the
origin (Garland et al. 1992). Contrasts were not calculated
for categorical data (i.e. trophic level) because trophic level
and phylogeny are closely correlated in arthropods in our
In order to test for a phylogenetic signal in stoichiom-
etry traits we used the Mantel test (B ö hning-Gaese and
Oberrath 1999). We created a dissimilarity matrix for stoi-
chiometric traits for each species pair by using Euclidean
distance and we also constructed a phylogenetic distance
matrix by counting the number of nodes that separate each
species pair (Fagan et al. 2002, Woods et al. 2004). Th e stoi-
chiometric dissimilarity matrices were regressed on the phy-
logenetic distance matrix and we tested the regression for
signi fi cance using the Mantel test. We tested the signi fi cance
of the t value from the Mantel test against a null distribu-
tion constructed by Monte Carlo randomizations, whereby
the phylogenetic matrix was held constant and species in the
stoichiometric matrix were reshuffl ed randomly (B ö hning-
Gaese and Oberrath 1999).
We performed a one-way analysis of covariance
(ANCOVA) to examine diff erences in C, N and P contents,
and C:N ratios among and within arthropod species, using
body size (dry weight in mg) as a covariate. Before analyses
consumer body size data were log transformed. In addition,
we performed ANCOVAs, with body size as the covariate to
test the eff ects of trophic level and ontogenetic stage on body
elemental content. All multiple comparisons between species
were performed using a Tukey HSD test. Signifi cance levels
were set at α ? 0.05 for all ANOVAs and ANCOVAs. We
used simple linear regression to determine the relationship
between body elemental content and log body size across
and within species. For these analyses, we included only
species with more than six individuals. All mean values are
shown with ? 1 SE. All statistical analyses were performed
with the statistical package R ver. 2.9.0 (R Development
Core Term 2009).
Stoichiometric variation among species
Th e analysis of body N and P content of 22 arthropod species
distributed across ten taxonomic orders (Araneae, Scorpi-
onoidea, Pseudoscorpionoidea, Acari, Solifugae, Orthoptera,
Coleoptera, Lepidoptera, Th ysanura and Diptera) showed a
wide range of variation (Fig. 1). Th e data revealed almost a
two-fold variation in body % N among arthropod species
(Fig. 1a), with a mean of 10.7 ? 0.09% N (F 21,288 ? 3.66,
p ? 0.001). Th e highest body % N was recorded for Arach-
nida: Acari species, with a mean value of 14.8 ? 1.50%
N, followed by some Insecta species such as Diptera and
Lepidoptera species. Nitrogen content was particularly low
for some Coleoptera (Tenebrionidae sp. 3 and sp. 5) with
a mean of 9.27 ? 0.30 and 9.13 ? 0.48, respectively. No
eff ect of phylogenetic relatedness in body N content of the
arthropod species was found (r ? – 0.069, p ? 0.832).
Body % P showed greater interspecifi c variation than body
% N, with a four-fold variation among species (F 18,265 ? 24.56,
p ? 0.001, Fig. 1b). Mean body % P across all arthropods
was 0.79 ? 0.08. Lepidoptera species (Nymphalidae and
Noctuidae) showed the highest body P content, with a mean
of 1.50 ? 0.30% P, whereas Coleoptera species (Anobiidae sp.
1 and Tenebrionidae sp. 1) and Orthoptera sp. 1 showed the
lowest body % P with a mean of 0.37 ? 0.02, 0.37 ? 0.05
and 0.36 ? 0.04% P, respectively. Results from the Mantel
test indicate that closely related species had signi fi cantly more
similar body P content than distantly related species (i.e.
there was a signifi cant phylogenetic signal for P (p ? 0.05)).
Phylogenetic relatedness explained 4.13% (p ? 0.05) of
the variance in body P content of the arthropod species.
Diff erences in body C content among species were small
but statistically signifi cant (F 21,288 ? 5.29, p ? 0.001), with
a coeffi cient of variation (CV ? SD/mean ? 5%). Mean
body % C content for all species was 42.45 ? 0.13, and
Diptera species had the lowest body % C content with a
mean of 39.65 ? 1.00 (Appendix 1 Fig. A1). Th e C:N stoi-
chiometry of arthropods diff ered signifi cantly among spe-
cies (F 21,288 ? 3.50, p ? 0.001) and showed a coeffi cient of
variation of 18%. Th e C:N ratios ranged from 3.9 ? 0.45
and also lower C:N (F 2,292 ? 27.52, p ? 0.001) and N:P
ratios (F 2,54 ? 7.86, p ? 0.001) for carnivorous arthro-
pods compared to herbivorous arthropods. Herbivores had
a mean of 9.84 ? 0.13% N and 0.50 ? 0.02% P whereas
carnivore mean was 11.36 ? 0.11% N and 0.85 ? 0.02% P.
In detritivore species, N content was 10.34 ? 0.41%, which
was signifi cantly diff erent from carnivores (p ? 0.01), and P
content was 0.81 ? 0.08%, which was signifi cantly diff erent
from herbivores (p ? 0.001). Trophic level explained 19%
of the variation in body N and 41% of the variation in body
in Acari species to 6.1 in Lepidoptera species (Fig. 2). No
eff ect of phylogenetic relatedness in body C content and
C:N ratios of the arthropod species was found (r ? – 0.072,
p ? 0.729 and r ? 0.152, p ? 0.248, respectively).
Lizards showed an average of 9.75 ? 0.55% N and
4.56 ? 0.10% P in their bodies (Fig. 1). Th e Phrynosaura
reichei (Tropiduridae) showed signifi cantly higher body N
content than the Phyllodactylus gerrhopygus (Gekkonidae)
(F 1,22 ? 7.78, p ? 0.05), but they did not diff er in body
P content (F 1,26 ? 3.48, p ? 0.070). Th e lizard P. reichei
had signifi cantly lower body C:N ratios than P . gerrhopygus
(F 1,24 ? 11.88, p ? 0.01).
Th e comparison of the body element content among
trophic levels showed higher N (F 2,292 ? 30.9, p ? 0.001,
Fig. 3) and P contents (F 2,281 ? 97.64, p ? 0.001, Fig. 3),
Figure 1. Mean ( ? SE) nutrient content across species: (a) nitrogen
as a percentage of dry weight, and (b) phosphorus as a percentage
of dry weight. Diff erent letters above their bars were judged signifi -
cantly diff erent by ANCOVA. Tukey HSD was used for post hoc
analysis of diff erences among species. Letters inside parenthesis
after family names indicate the taxonomic order: (A) Acari, (Sc)
Scorpionoidea, (P) Pseudoscorpionoidea, (S) Solifugae, (Ar) Ara-
neae, (T) Th ysanura, (O) Orthoptera, (C) Coleoptera, (L) Lepi-
doptera, (D) Diptera and Sauria (Sa). Grey bars show vertebrate
species. ND indicates no available N % data for Tenebrionidae sp.
1 (C) and for Tephritidae (D). No available % P data for Sicariidae
sp. 2 (Ar). Th e species belonging to Noctuidae and Nymphalidae
(L) were not included in statistical analyses (n ? 1). Lizards were
analyzed independently from arthropods.
Figure 2. Mean ( ? SE ) C:N molar ratios across species. Diff erent
letters above their bars were judged signifi cantly diff erent by
ANCOVA. Tukey HSD was used for post hoc analysis of diff er-
ences among species. Letters inside parenthesis after family names
indicate the taxonomic order: (A) Acari, (Sc) Scorpionoidea, (P)
Pseudoscorpionoidea, (S) Solifugae, (Ar) Araneae, (T) Th ysanura,
(O) Orthoptera, (C) Coleoptera, (L) Lepidoptera, (D) Diptera and
Sauria (Sa). Grey bars show vertebrate species. Lizards were ana-
lyzed independently from arthropods.
Figure 3. Mean ( ? SE) body N (open circles) and P (closed circles)
content P for detritivores, herbivores and carnivores. Tukey
HSD was used for post hoc analysis of diff erences among trophic
Our results show that terrestrial consumers in desert food
webs diff er widely in their element content both within and
among species. Evolutionary and ecological factors appear
to explain diff erences among arthropod species. We found
a signifi cant eff ect of trophic level on the elemental stoichi-
ometry of arthropods, with carnivores having higher body
N and P content than herbivores. Ontogeny did explain
some, but not much, of the variation in the element content
of Diptera. Our fi ndings suggest that diff erences associated
with phylogenetic relateness, structural material allocation,
trophic level and body size contribute to extensive natural
inter- and-intraspecifi c variation in the elemental stoichiom-
etry of consumers in desert environments.
Stoichiometric variation among species
Body N content showed two-fold variation among spe-
cies, whereas body P content varied from two to fi ve-fold
among species. Th ese diff erences are not simple correlates of
phylogenetic relatedness. Within our dataset we found no
phylogenetic signal for N and C:N ratios. Th ese results are
consistent with those from a broad-scale literature survey by
Fagan et al. (2002), who found that arthropod N content
was not related to phylogeny at lower taxonomic levels. Our
results, along with those of Fagan et al. (2002), strongly sug-
gest a lack of fi ne-grained phylogenetic signal for arthropod
Stoichiometric variation within species
Diff erences in elemental content within species from diff er-
ent sites were small, but seven species from the whole-data
set (out of 13 species analyzed) showed high intraspecifi c
variability in body N or P content (Table 1). Th e body N
and P content of some Coleoptera and Arachnida species
diff ered signifi cantly (showing about two-fold to fi ve-fold
variation for each taxa). Body N, P content and C:P ratios
did not diff er between larval and adult stages for Diptera
(for N content, F 1,18 ? 0.968; p ? 0.338; for P content,
F 1,16 ? 0.629, p ? 0.439 and for C:P ratios F 1,7 ? 0.278,
p ? 0.614), in contrast to body C:N ratios that were much
higher for larval individuals than for adults (F 1,18 ? 18.30,
p ? 0.001, Appendix 1 Fig. A2).
Th ere were signifi cant allometric relationships for P con-
tent (but not N content) across all arthropods in the data
set (Table 2). However, the allometry for body P content
was not signifi cant after correcting for phylogeny (Table 2).
At the species level, all arthropods showed a negative body
P content allometry (Fig. 4a, Table 2), but this relation-
ship was signifi cant for only seven (of 13) species (Table 2).
In contrast, there was a signifi cant positive allometric scaling
of body N content for the Scorpionoidea species. Both lizard
species showed a signifi cant positive relationship between
P content and body size (Fig. 4b, Table 2). Th e body
N:P ratios showed a negative allometry for both species
(r 2 ? 0.59, p ? 0.05 for P. reichei and r 2 ? 0.453, p ? 0.01
for P. gerrhopygus ).
Table 1. Results of analysis of covariance of intra-specifi c variation in body N and P content of consumers across sites. Signifi cant differences
are in bold. ∗ p ? 0.05, ∗∗ p ? 0.01.
SpeciesSource of variation Range of variation (%)DF F-valuep-value
9.11 – 13.15
0.60 – 1.36
8.65 – 15.23
0.50 – 1.29
9.27 – 13.64
0.50 – 1.21
7.77 – 12.89
0.58 – 1.23
9.79 – 14.61
Ammotrechidae sp. 1 (S)
Ammotrechidae sp. 2 (S)
Sicariidae sp. 1 (Ar)
0.47 – 1.67
7.10 – 12.88
0.47 – 1.01
7.47 – 12.58
0.25 – 0.63
6.73 – 11.60
0.13 – 0.70
7.93 – 10.67
0.22 – 0.55
8.20 – 14.67
0.60 – 1.12
9.08 – 11.62
3.57 – 7.00
8.44 – 10.76
3.14 – 6.10
Tenebrionidae sp. 2 (C)
Tenebrionidae sp. 3 (C)
Anobiidae sp. 1 (C)
Order codes: (Sc) Scorpionoidea, (S) Solifugae, (Ar) Araneae, (T) Thysanura, (C) Coleoptera, (D) Diptera, and the two lizard species belonging
to (Sa) Sauria. ND indicates no data available for P and N analysis for Lycosidae sp. 1 and Sicariidae sp. 1, respectively.
Table 2. Scaling of body elemental content in arthropods and lizards with body size. The results show regressions on phylogenetic independent
contrasts (PICs) for whole arthropod data set and analogous results from non-phylogenetically controlled data for 15 arthropod species and
two lizard species. The r2-values, slope and p-values are indicated. Statistically signifi cant relationships are in bold (p < 0.05). Only species
with more than six individuals were used in regression analyses.
NP C:N ratios
Speciesr 2 slopep r 2 slopep r 2 slopep
Whole arthropod data set
Whole arthropod data set
Bothriuridae sp. 1 (Sc)
Ammotrechidae sp. 1 (S)
Ammotrechidae sp. 2 (S)
Sicariidae sp. 1 (Ar)
Salticidae sp. 1 (Ar)
Lycosidae sp. 1 (Ar)
Lepismatidae sp. 1 (T)
Ommexhixidae sp. 1 (O)
Tenebrionidae sp. 2 (C)
Tenebrionidae sp. 3 (C)
Tenebrionidae sp. 4 (C)
Tenebrionidae sp. 5 (C)
Anobiidae sp. 1 (C)
Anobiidae sp. 2 (C)
Muscidae sp. 1 (D)
Gekkonidae sp. 1 (Sa)
Tropiduridae sp. 1 (Sa)
Order codes: (Sc) Scorpionoidea, (S) Solifugae, (Ar) Araneae, (T) Thysanura, (O) Orthoptera, (C) Coleoptera, (D) Diptera and the two lizard
species belonging to (Sa) Sauria. ND indicates no data available for the analysis.
N content. Although the interpretation of phylogenetic
signal remains in dispute (Losos 2008, Revell et al. 2008),
one possible cause of this result is that life history traits
infl uencing body N content may evolve relatively fast as a
response to environmental conditions (Fagan et al. 2002).
Th e biochemical and morphological correlates of body N
(and C:N) diff erences among taxa are still poorly described.
Possible correlates include diff erential exoskeleton invest-
ment and cuticle composition (e.g. protein-to-chitin ratios)
(Sterner and Elser 2002).
We also found a small, but statistically signifi cant, phylo-
genetic signal for P. Th is result is consistent with a survey of
North American desert arthropods by Woods et al. (2004),
who found that recently derived orders (Lepidoptera and
Diptera) had higher body P content than other groups. Th e
relatively high body P content for Lepidoptera and Diptera
in our data set suggests that this may be a general pattern, at
least in dry ecosystems.
Invertebrate predator N content in our survey was higher
than that of herbivores and detritivores; this result is similar
to comparisons made in other systems (Fagan et al. 2002,
Matsumura et al. 2004, Kagata and Ohgushi 2007). Inter-
estingly, we also found that invertebrate predators have on
average higher body P content than herbivores and detriti-
vores. Th is fi nding contrasts with the results of Woods et al.
(2004), who did not fi nd diff erences in P content between
trophic levels (herbivores vs carnivores).
Contrary to general expectations on allometric relation-
ships in arthropod element content, we found no consistent
relationship between N content and body size of arthro-
pods when all taxa were pooled together. Phylogenetically
corrected and uncorrected analyses yielded diff erent results
for P content; only when phylogenetic relatedness was not
considered did we fi nd a signifi cant (negative) relationship
between P content and body size. Although our data set is
limited to 22 species, the weak association between whole-
body P content and body size is consistent with the analysis
of Gillooly et al. (2005), who modeled size-related changes
in the relative proportion of P-rich RNA (required for rapid
growth) versus P associated with other body pools such as
phospholipids that are invariant with body size.
To our knowledge, no other data about lizard nutrient
stoichiometry has been published. Summarizing available
data on larger vertebrates, Sterner and Elser (2002) reported
that, in birds and mammals, body N content ranged from
8 to 12%, P content from 1.5 to 4.5%, and N:P ratios from
5 to 15. Our data on lizards fall within these ranges, sug-
gesting a relatively high constancy in nutrient content across
Although our results show some associations between
body element content, phylogeny, trophic level, and body
size among species, the underlying basis of these associations
is still poorly understood. Th ere are two main categories of
explanation for these patterns: functional and economical
(Kay et al. 2005, Kay and Vrede 2008). Functional mecha-
nisms that have been off ered to explain species diff erences
in N and C:N ratios include diff erential exoskeleton invest-
ment and cuticle composition (e.g. protein to chitin ratios);
for example, fl ying insects may have reduced cuticle invest-
ment that trades off support and protection for increased
agility and decreased metabolic costs during fl ight (Fagan
et al. 2002, Lease and Wolf 2010). Similarly, the growth
rate hypothesis (which functionally links P content to
growth rate) has been off ered as an explanation for the high
(Kay et al. 2006), coupled with studies linking dietary
nutrition to the element content of organisms and trait
expression (Behmer 2009).
Stoichiometric variation within species
We found that both N and P content showed extensive
variation within individual arthropod species. Th is varia-
tion suggests that, although individual consumers might
maintain compositional homeostasis through selection
ingestion, assimilation, and excretion, compositional set
points may vary predictably with individual trait diff er-
ences. To date, studies examining within-taxa variation in
terrestrial consumers are still rare, however, some reports
exist on the variation of stoichiometry across development
stages and body sizes (Kay et al. 2006, Back et al. 2008,
Bertram et al. 2008). Although empirical studies suggest
that the element content of invertebrates is strongly associ-
ated with ontogeny (Sterner and Elser 2002), there is some
evidence that show no infl uence of development stage
on body element content of insects (Fagan et al. 2002).
Here we did not fi nd any signifi cant diff erence on N or
P content between ontogenetic stages; however, larvae
had signifi cant higher C contents and C:N ratios than
adult fl ies. Th e larval stage of holometabolus insects (e.g.
Diptera) is characterized by conversion of carbohydrates to
fat bodies, which supports the rapid growth of the larvae
and fuels the animal through the subsequent non-feeding
(pupal) period (Aguila et al. 2007). Th ese fi ndings sug-
gest that higher amount of larval-stored fat increase larval
C content and C:N ratios (with no changes in body N
content) in holometabolous insects, whereas hemimetabo-
lous insects do the opposite. Such life-history diff erences
between holometabolous and hemimetabolous insects will
likely play a main role in infl uencing ontogenetic diff er-
ences in insect stoichiometry and, thus should be consid-
ered in future studies.
At the intraspecifi c level we found that smaller inverte-
brates have higher P content. Current explanations for such
allometries focus on the functional link between growth rate
and body P content (Elser et al. 1996, Sterner and Elser
2002, Woods et al. 2004, Bertram et al. 2008). Th e allomet-
ric patterns in lizards are consistent with existing data show-
ing that body P content is an increasing function of body size
in most vertebrates (Dantas and Attayde 2007, Hendrixson
et al. 2007). In contrast, body N content and N:P ratios
have shown a negative, positive or no relationship with body
size in fi shes (Sterner and George 2000, Hendrixson et al.
2007, Torres and Vanni 2007). Th ese fi ndings suggest that
some fi shes and lizards may have less relative allocation of
N into muscles, but also the endoskeletal mass investment
seems to be variable across vertebrate species (Lease and Wolf
2010). Consequently, the scaling of structural support and
elemental stoichiometry in vertebrates exhibit slightly dif-
ferent allometric responses across taxonomic groups, which
could be partly driven by evolutionary history (Hendrixson
et al. 2007).
Overall, our study analyzed nutrient content variation in
both invertebrate and vertebrate species in desert ecosystems
within a well-delimited geographical area. To our knowl-
edge, our study is the fi rst to assess inter- and-intra specifi c
P contents of Lepidoptera and Diptera because these taxa
need high rates of growth and reproduction to effi ciently
exploit ephemeral food resources (Woods et al. 2004). Dif-
ferences in element content among consumers have also
been explained using economical mechanisms. For instance,
recent surveys suggest a major role of the elemental fac-
tors of the microhabitat (soil abiotics) in the distribution
of larger organisms (Mulder and Elser 2009). Moreover,
carnivores may have higher N content than herbivores
because they generally have higher assimilation effi ciencies
than herbivores (Lehman 1993, Sterner and Elser 2002),
and intraguild predation may allow predators to enhance
their N-intake by feeding on N-rich preys (Fagan et al.
2002, Denno and Fagan 2003). In addition, the signifi -
cant diff erence in P content between predators and herbi-
vores in our survey could be a consequence of the fact that
plant resources in our desert sites have particularly low P
content (Gonz á lez unpubl.). Clarifying the relative impor-
tance of functional and economic selection pressures on
body element composition is a major research challenge in
ecological stoichiometry (Kay et al. 2005). Progress toward
this end can be made with comparisons of element stoi-
chiometry, higher-order biochemistry, and functional traits
Figure 4. Allometric relationship of P content in arthropod and
lizard species: (a) Allometry of P content in arthropod species:
Ammotrechidae sp. 1 (S), Sicariidae sp. 1 (Ar), Salticidae
sp. 1 (Ar), Tenebrionidae sp. 2 (C),
(C) and Anobiidae sp. 2 (C) and (b) Allometry of P content in
Phyllodactylus gerrhopygus and
reichei . Note diff erence in scale of each fi gure. See Table 2 for statis-
Tenebrionidae sp. 3
Aguila, J. R. et al. 2007. Th e role of larval fat cells in adult
Drosophila melanogaster . – J. Exp. Biol. 210: 956 – 963.
Back, J. A. et al. 2008. Ontogenic diff erences in mayfl y stoichiom-
etry infl uence growth rates in response to phosphorus enrich-
ment. – Fundam. Appl. Limnol. 171: 233 – 240.
Behmer, S. T. 2009. Insect herbivore nutrient regulation. – Annu.
Rev. Entomol. 54: 165 – 187.
Bertram, S. M. et al. 2008. Extensive natural intraspecifi c variation
in stoichiometric (C : N : P) composition in two terrestrial
insect species. – J. Insect Sci. 8: 1 – 7.
B ö hning-Gaese, K. and Oberrath, R. 1999. Phylogenetic eff ects on
morphological, life-history, behavioural and ecological traits of
birds. – Evol. Ecol. Res. 1: 347 – 364.
Cereceda, P. et al. 2008. Th e spatial and temporal variability of fog
and its relation to fog oases in the Atacama Desert, Chile. –
Atmos. Res. 87: 312 – 323.
Coddington, J. A. and Levi, H. W. 1991. Systematics and evolution
of spiders (Araneae). – Annu. Rev. Ecol. Syst. 22: 565 – 592.
Cross, W. F. et al. 2003. Consumer – resource stoichiometry in
detritus-based streams. – Ecol. Lett. 6: 721 – 732.
Dabert, M. et al. 2010. Molecular phylogeny of acariform mites
(Acari, Arachnida): strong confl ict between phylogenetic signal
and long-branch attraction artifacts. – Mol. Phylogenet. Evol.
56: 222 – 241.
Dantas, M. C. and Attayde, J. L. 2007. Nitrogen and phosphorus
content of some temperate and tropical freshwater fi shes. –
J. Fish Biol. 70: 100 – 108.
Denno, R. F. and Fagan, W. F. 2003. Might nitrogen limitation
promote omnivory among carnivorous arthropods? – Ecology
84: 2522 – 2531.
Elser, J. J. et al. 1996. Organism size, life history and N:P stoichi-
ometry. – Bioscience 46: 674 – 684.
Elser, J. J. et al. 2000. Th e evolution of ecosystem processes: growth
rate and elemental stoichiometry of a key herbivore in temper-
ate and arctic habitats. – J. Evol. Biol. 13: 845 – 853.
Evans-White, M. A. et al. 2005. Taxonomic and regional patterns in
benthic macroinvertebrate elemental composition in streams. –
Freshwater Biol. 50: 1786 – 1799.
Fagan, W. F. et al. 2002. Nitrogen in insects: implications for
trophic complexity and species diversifi cation. – Am. Nat. 160:
784 – 802.
Felsenstein, J. 1985. Phylogenies and the comparative method. –
Am. Nat. 125: 1 – 15.
Felsenstein, J. 2008. Comparative methods with sampling error
and within-species variation: contrasts revisited and revised. –
Am. Nat. 171: 713 – 725.
Flook, P. K. et al. 2000. Molecular phylogenetic analysis of the
Pneumoroidea (Orthoptera, Caelifera): molecular data resolve
morphological character confl icts in the basal Acridomorpha. –
Mol. Phylogenet. Evol. 15: 345 – 354.
Garland, T. et al. 1992. Procedures for the analysis of comparative
data using phylogenetically independent contrasts. – Syst.
Biol. 41: 18 – 32.
Garland, T. et al. 2005. Phylogenetic approaches in comparative
physiology. – J. Exp. Biol. 208: 3015 – 3035.
Gillooly, J. F. et al. 2005. Th e metabolic basis of whole-organism
RNA and phosphorus content. – Proc. Natl Acad. Sci. USA
102: 11923 – 11927.
Hall, S. R. 2009. Stoichiometrically explicit food webs: feedbacks
between resource supply, elemental constraints and species
diversity. – Annu. Rev. Ecol. Evol. Syst. 40: 503 – 528.
Hamb ä ck, P. A. et al. 2009. Eff ects of body size, trophic mode
and larval habitat on Diptera stoichiometry: a regional com-
parison. – Oikos 118: 615 – 623.
Han, H. Y. and Ro, K. E. 2009. Molecular phylogeny of the fam-
ily Tephritidae (Insecta: Diptera): new insight from combined
consumer stoichiometry from terrestrial food webs extend-
ing beyond a single nutrient and invertebrate consumers.
Our results support previous fi ndings indicating that taxo-
nomic and ecological factors have an important infl uence
on the natural variation of consumer C:N:P stoichiometry.
Th is work sets the stage for studies on the functional and
economical causes of these diff erences, which in turn can be
used to identify broad connections between nutrient cycling
and the structure of consumer communities.
Implications for food web structure
and nutrient cycling
Much of the work in ecological stoichiometry lies in the
understanding of the patterns and mechanisms of organ-
ism elemental composition as the departing point for
identifying the occurrence and magnitude of consumer –
resource elemental imbalances. Th ese elemental imbalances
between consumers and their resources are particularly
important because impose stoichiometric constraints on
consumers, providing powerful mechanisms that shape
food web structure and dynamics (Sterner and Elser
2002, Hall 2009). Although substantial evidence has
shown that stoichiometry of herbivore – plant interaction
is strongly imbalanced and infl uence both herbivore fi t-
ness and species interactions (Hall 2009), stoichiometric
mismatches and their consequences for higher trophic
levels have received far less attention. Th ere is some evi-
dence, however, that suggest an increase to nutrient-richer
biomass as we progress from plants up through the food
web, revealing stoichiometric imbalances at predator –
prey interactions (Denno and Fagan 2003, Matsumura
et al. 2004). In fact, individual/population growth limi-
tation driven by stoichiometric imbalances are also a real
possibility for predators (Fagan et al. 2002). Th ese observa-
tions are particularly important in light of evidence that
suggest how nutrient limitation may contribute to the evo-
lution of intraguild predation and lead to changes in the
stability and complexity of food webs (Fagan et al 2002).
Furthermore, rates and magnitudes of nutrient recycling by
consumers are aff ected by the stoichiometry of the organ-
isms and by the elemental mismatch between them and
that of their resources (Sterner and Elser 2002). Th erefore,
our study calls for further research into the patterns and
mechanisms that explain how elemental stoichiometry var-
ies across taxa and ecosystems, as well as its infl uence on
food-web structure and ecosystem functioning.
Acknowledgements – We thank Carlos Gar í n, Mart í n Escobar,
Romina Villagr á n, Margarita Ru í z de Gamboa, Sebastian Armesto,
Vivian Jer é z, Marcos Ferru, Mois é s Aguilera, Mar í a Fernanda P é rez
and Sara Seidl for helping with fi eld and/or lab work. We are very
grateful to Andy Rominger, Marcia Kyle, Jim Elser and Spencer
Hall for their useful comments on earlier drafts of our manuscript.
Comments from Christian Mulder signifi cantly improved the
manuscript. Th is project was funded by FONDAP 1501-0001
(programs 3 and 4), CONICYT 24050045, FONDECYT
1040783/2004, ICMP05-002 and CONICYT PFB-023. Proce-
dures involving animals were approved by the ethics committee of
the Pontifi cia Universidad Cat ó lica de Chile and by the Chilean
Agriculture and Livestock Service (SAG).
Mulder, C. and Elser, J. J. 2009. Soil acidity, ecological stoichiom-
etry and allometric scaling in grassland food webs. – Global
Change Biol. 15: 2730 – 2738.
Mulder, K. and Bowden, W. B. 2007. Organismal stoichiometry
and the adaptive advantage of variable nutrient use and produc-
tion effi ciency in Daphnia . – Ecol. Model. 202: 427 – 440.
Nihei, S. S. and Barros de Carvalho, C. J. 2007. Phylogeny and
classifi cation of Muscini (Diptera, Muscidae). – Zool. J. Linn.
Soc. 149: 493 – 532.
Persson, J. et al. 2010. To be or not to be what you eat: regulation
of stoichiometric homeostasis among autotrophs and hetero-
trophs. – Oikos 119: 741 – 751.
Pinto, R. et al. 2006. Geographical distribution of Tillandsia lomas
in the Atacama Desert, northern Chile. – J. Arid Environ. 65:
543 – 552.
Regier, J. C. et al. 2010. Arthropod relationships revealed by phy-
logenomic analysis of nuclear protein-coding sequences. –
Nature 463: 1079 – 1083.
Revell, L. J. et al. 2008. Phylogenetic signal, evolutionary process
and rate. – Syst. Biol. 57: 591 – 601.
Schade, J. D. et al. 2003. Stoichiometric tracking of soil nutrients
by a desert insect herbivore. – Ecol. Lett. 6: 96 – 101.
Shultz, J. W. 2007. A phylogenetic analysis of the arachnid orders
based on morphological characters. – Zool. J. Linn. Soc. 150:
221 – 265.
Sterner, R. W. and George, N. B. 2000. Carbon, nitrogen, and
phosphorus stoichiometry of cyprinid fi shes. – Ecology 81:
127 – 140.
Sterner, R. W. and Elser, J. J. 2002. Ecological stoichiometry: the
biology of elements from molecules to the biosphere. – Prin-
ceton Univ. Press.
Torres, L. E. and Vanni, M. J. 2007. Stoichiometry of nutrient
excretion by fi sh: interspecifi c variation in a hypereutrophic
lake. – Oikos 116: 259 – 270.
Wahlberg, N. et al. 2005. Phylogenetic relationships and historical
biogeography of tribes and genera in the subfamily Nymphali-
nae (Lepidoptera: Nymphalidae). – Biol. J. Linn. Soc. 86:
227 – 251.
Woods, H. A. et al. 2004. Allometric and phylogenetic variation
in insect phosphorus content. – Funct. Ecol. 18: 103 – 109.
analysis of the mitochondrial 12S, 16S, and COII genes.
– Molecules Cells 27: 55 – 66.
Hendrixson, H. A. et al. 2007. Elemental stoichiometry of fresh-
water fi shes in relation to phylogeny, allometry and ecology.
– J. Fish Biol. 70: 121 – 140.
Hunt, T. et al. 2007. A comprehensive phylogeny of beetles reveals
the evolutionary origins of a superradiation. – Science 318:
1913 – 1916.
Kagata, H. and Ohgushi, T. 2007. Carbon – nitrogen stoichiometry
in the tritrophic food chain willow, leaf beetle, and predatory
ladybird beetle. – Ecol. Res. 22: 671 – 677.
Kay, A. D. and Vrede, T. 2008. Evolutionary and biochemical
aspects. – In: J ø rgensen, S. E. and Fath, B. D. (eds), Ecologi-
cal stoichiometry. Encyclopedia of Ecology. Elsevier, pp.
1472 – 1481.
Kay, A. D. et al. 2005. Toward a stoichiometric framework for
evolutionary biology. – Oikos 109: 6 – 17.
Kay, A. D. et al. 2006. Ant stoichiometry: elemental homeostasis
in stage-structured colonies. – Funct. Ecol. 20: 1037 – 1044.
Lease, H. M. and Wolf, B. O. 2010. Exoskeletal chitin scales iso-
metrically with body size in terrestrial insects. – J. Morphol.
271: 759 – 768.
Lehman, J. T. 1993. Effi ciencies of ingestion and assimilation by
an invertebrate predator using C-dual and P-dual isotope labe-
ling. – Limnol. Oceanogr. 38: 1550 – 1554.
Losos, J. B. 2008. Phylogenetic niche conservatism, phylogenetic sig-
nal and the relationship between phylogenetic relatedness and
ecological similarity among species. – Ecol. Lett. 11: 995 – 1003.
Martinson, H. M. et al. 2008. Detritivory: stoichiometry of a
neglected trophic level. – Ecol. Res. 23: 487 – 491.
Matsumura, M. et al. 2004. Does intraguild predation enhance
predator performance? A stoichiometric perspective. – Ecology
85: 2601 – 2615.
Mitchell, A. et al. 2000. More taxa or more characters revisited:
combining data from nuclear protein-encoding genes for phy-
logenetic analyses of Noctuoidea (Insecta : Lepidoptera). –
Syst. Biol. 49: 202 – 224.
Moe, S. J. et al. 2005. Recent advances in ecological stoichiometry:
insights for population and community ecology. – Oikos 109:
29 – 39.