ANIMAL BEHAVIOUR, 2002, 64, 629–643
doi:10.1006/anbe.2002.3082, available online at http://www.idealibrary.com on
The interplay between nutrient balancing and toxin dilution in
foraging by a generalist insect herbivore
M. S. SINGER*, E. A. BERNAYS† & Y. CARRIE
v RE‡
**Interdisciplinary Program in Insect Science, Center for Insect Science, University of Arizona
†Center for Insect Science, University of Arizona
‡Department of Entomology, University of Arizona
(Received 18 April 2001; initial acceptance 31 July 2001;
final acceptance 19 February 2002; MS. number: A9045R )
Food mixing by herbivores is thought to balance nutrient intake and possibly dilute secondary
metabolites characteristic of different host plant species. Most empirical work on insect herbivores has
focused on nutrient balancing in laboratory settings. In this study, we characterize food mixing
behaviour of the caterpillar Grammia geneura (Strecker) (Lepidoptera: Arctiidae) in nature and use the
observed patterns to design ecologically relevant experiments that reveal the relative importance of these
processes in food-switching behaviour. Our design involved both choice and no-choice experiments with
chemically defined diets in which primary nutrients and secondary metabolites were manipulated in
tandem. We analysed two stages in the process of food-switching behaviour: leaving food and accepting
new food. In nature, an individual’s rate of leaving host plants was positively associated with its
probability of rejecting plant species most recently eaten, but not related to its probability of accepting
different host plant species. Furthermore, an individual’s leaving rate was negatively related to its average
feeding bout duration. This relationship resulted partly from variation in the response of individuals to
nutrient imbalance and partly from shortened feeding bouts prior to switching, suggesting that a decline
in feeding excitation preceded searching for food that differed from that most recently eaten. Laboratory
experiments with synthetic diets indicated the importance of secondary metabolites in the decline in
feeding excitation prior to switching. Preference for new food depended strongly on secondary
metabolites in a manner consistent with toxin dilution. This is the first experimental evidence for the
process of toxin dilution in caterpillars, and for the combined influence of nutrients and secondary
metabolites on their foraging patterns in nature.

2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.
Optimal foraging theory argues that diet selection by
generalist herbivores should be greatly influenced by
limitations in food availability and quality (Stephens &
Krebs 1986). For generalist insect herbivores, small
resource demands relative to resource size generally
ensures sufficient food availability for individuals. Hence,
food quality is of utmost importance because it severely
limits individual performance, and presumably fitness.
The food quality of individual plants or plant parts varies
in acceptability or suitability due to physical differences
(e.g. leaf toughness), and variation in the composition of
primary nutrients (Slansky & Scriber 1985) and secondary
metabolic compounds (Bernays & Chapman 1994). For
Correspondence and present address: M. S. Singer, Department of
Entomology, University of Arizona, Tucson, AZ 85721, U.S.A. (email:
[email protected]). E. A. Bernays is at the Center for Insect
Science, University of Arizona, Tucson, AZ 85721, U.S.A.
0003–3472/02/$35.00/0

food mixing insects (i.e. those in which individuals eat
multiple plant species), host plant switching may function to balance nutrient intake on the one hand, while
minimizing toxicity from specific secondary metabolites
in host plants on the other.
The nutrient balance hypothesis proposes that food
mixing allows individuals to balance intake of different
nutrients (Pulliam 1975; Westoby 1978; Rapport 1980;
Waldbauer et al. 1984; Simpson et al. 1995). Physiological
and behavioural experiments clearly demonstrate that
food mixing can function to balance nutrient intake,
thereby enhancing growth and development (Waldbauer
& Friedman 1991; Simpson & Raubenheimer 1996).
Among insects, food selection on the basis of nutrient
balancing has been demonstrated in grasshoppers
(Simpson & Raubenheimer 1993), aphids (Abisgold et al.
1994) and caterpillars (Waldbauer et al. 1984; Simmonds
et al. 1992).
629
2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.
630
ANIMAL BEHAVIOUR, 64, 4
There is less experimental support for the toxin dilution hypothesis: food mixing allows individuals to avoid
ingesting toxic doses of the particular secondary metabolites characteristic of specific foods (Freeland & Janzen
1974; Freeland & Saladin 1989). Despite widespread evidence that secondary metabolites are often especially
toxic to polyphagous insects (Bernays & Chapman 1994;
Karban & Baldwin 1997), studies of herbivorous insects
have yet to claim any strong support for the toxin
dilution hypothesis. Indeed, it may be difficult to separate
herbivore responses to secondary metabolites from those
to other phytochemicals (including primary nutrients)
because interactions among such chemicals directly influence herbivore physiology (Slansky 1992; Hagele &
Rowell-Rahier 1999; Simpson & Raubenheimer 2001).
None the less, there have been few attempts to test
mechanistic predictions of these hypotheses with evidence from the detailed foraging behaviour of individuals
observed in nature and the laboratory. So far, these efforts
have been confined to grasshoppers (Bernays et al. 1992,
1994; Raubenheimer & Bernays 1993; Chambers et al.
1996) with one exception (Dethier 1988). The relatively
large amount of experimental work conducted exclusively in the laboratory has shown that both macronutrients and secondary metabolites alone and in
combination influence the feeding behaviour of foodmixing insects (reviewed in Bernays & Chapman 1994,
2000). However, because food-mixers may experience a
wide variety of combinations of nutrients and secondary
metabolites in natural host plants, field studies conducted in conjunction with simplified laboratory experiments are critical for understanding the relative roles of
nutrients and secondary metabolites in the mechanism
and function of foraging in its natural context. Here, we
attempt such a study with the food mixing caterpillar
Grammia geneura (Strecker) (Lepidoptera: Arctiidae). First,
we identify quantitative parameters of host plant switching by caterpillars in nature. Second, we examine
components of switching behaviour by caterpillars in
their natural habitat and those offered synthetic diets in
the laboratory. Third, we use laboratory experiments with
synthetic diets to determine the influence of primary
nutrients and secondary metabolites on the process of
switching.
We designed analyses of observations and experiments
to address the following behavioural predictions of the
nutrient balance and toxin dilution hypotheses. The
nutrient balance hypothesis predicts that food switching
should involve leaving nutritionally unbalanced food
and accepting new food that is nutritionally complementary to that recently eaten. By contrast, the toxin dilution
hypothesis predicts that switching should involve leaving
food after excessive exposure to toxic or deterrent effects
of particular secondary metabolites and accepting new
food lacking these secondary metabolites. We reasoned
that the avoidance of potential toxins by a food mixing
herbivore would be maladaptive if switching occurred in
response to intoxication (i.e. ‘passive response’ model
in Glendinning & Slansky 1995). Therefore, rather than
look at toxicity per se, we hypothesized that food mixing
herbivores would use secondary metabolites as signals of
particular food types (i.e. plant species in nature), and
that food switching would be initiated when such compounds deterred individuals from further feeding on a
particular food type (more akin to ‘active mechanisms’ in
Glendinning & Slansky 1995). The increased acceptance
of novel food types would result from the absence of
secondary metabolites recently experienced or possibly
the presence of different ones.
METHODS
Study System
Grammia geneura inhabits arid savanna and grassland of
the southwestern U.S.A., where its larvae graze on at least
80 species of (mostly herbaceous) flowering plant from
at least 50 taxonomically disparate families (Singer
2000). Many of its host genera (e.g. Astragalus, Plantago,
Euphorbia, various composites) are commonly avoided by
other generalist herbivores (e.g. livestock) because of
deterrence or toxicity. All insects in this study were (or
were recently descended from stock) collected in southeastern Arizona, where the larval stages coincide with
both winter/spring and summer rainy periods. Individual
caterpillars move among small, mostly annual herbs in
mixed-species patches (typically >10 spp.) in which
individual plants are often contiguous or nearly so.
Caterpillars of this species are most commonly found at
ground level and rarely climb more than approximately
1 m from the ground (M. Singer, personal observation).
Most feeding occurs diurnally, when larvae are most
active (M. Singer, personal observation). A particularly
unusual feature of the life history is that adult females
oviposit on the ground, rather than on host plants,
forcing larvae to perform all host plant selection (M.
Singer, personal observation). Short-term observations
and analysis of plant fragments in the faeces of late
instars revealed that individual caterpillars eat multiple
host plant species (Singer 2000).
Pattern of Food Mixing
Field observations
To quantify host plant switching by individual caterpillars, we first observed a total of 43 caterpillars in nature at
several different times and locations. Observations of
5–180 min were made during both spring (near Oracle,
Pinal Co., Arizona, U.S.A., April 1995) and summer (near
Arivaca, Pima Co., Arizona, August 1993, 1995). These
data are included in only a single analysis here (see
below).
We used relatively long-term observations (7 h) in
nature to measure two key components of food switching: leaving host plants and accepting new plants. We
define switching as the following sequence of events: (1)
feeding on one food item, (2) walking away from it, (3)
contacting a second food item with mouthparts, and (4)
eating the second food. Leaving refers to steps 1–2, while
accepting new food involves steps 3–4. Food mixing refers
to switches between different plant species. We observed
11 final-instar larvae continuously for 7 h between 0900
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
and 1700 hours, when most larval activity occurs (Ash
Creek, Rincon Mtns, Pima Co., Arizona, 6 March–17 April
1996). The data from one individual was not included in
the analyses because it rarely fed. All observation days
were mostly or entirely clear and sunny. Air and ground
temperatures (in the shade) in foraging areas were
measured at several times of day during observations.
Each observer tracked a single, focal caterpillar throughout an observation session. Focal caterpillars were collected for rearing following the observation period to
make sure they were unparasitized, final-instar larvae.
For all observations, times at which various behaviours
initiated and terminated were recorded with digital
watches. Behaviour was classified in the following way:
‘feeding’ was defined as rhythmic movement of the head
while it was in contact with the plant surface (ingestion
of leaf tissue was usually visible); ‘walking’ was defined as
any locomotion exclusive of feeding; ‘resting’ was defined
as a lack of locomotion or feeding but included instantaneous movements like defecation and twitching;
‘tasting’ was defined as a brief period (<1 s) of contact
between a caterpillar’s mouthparts and plant tissue.
Tasting followed by walking was scored as a ‘rejection’.
Durations of feeding, resting and walking were measured
to the nearest second, while tastes and rejections were
recorded as instantaneous events. Feeding events were
composed of discrete bouts. These could not be grouped
as meals because there was no clear interbout criterion to
define meals (Simpson 1982), so bouts rather than meals
were used in all analyses. All plants contacted were
determined to species when possible (>95%). We provide
a list of all host plant species eaten by focal caterpillars
(Table 1).
Analysis of host plant mixing
We used linear regression to analyse the relationship
between critical parameters of switching behaviour in
nature: leaving rate (explanatory variable) and the probability of either rejecting a plant species most recently
eaten [p(Rs)] or accepting a plant species that differed
from that most recently eaten [p(Ad)] (response variables).
We calculated the leaving rate of individuals (per h) as the
number of times an insect left its host plant divided by
the total observation duration. To obtain values for the
response variables, we imposed several criteria on the
included data (7 h observations, N=10 insects). First, we
only considered contacts with plants during the walking
phase immediately following a feeding event of 60 s or
more to ensure that the plant previously eaten was
acceptable. Second, we excluded all contacts with grass
species. These are generally far less acceptable than forbs
to G. geneura (Singer & Stireman 2001). The probability of
rejecting plant species most recently eaten [p(Rs)] was
calculated for each individual as:
p(Rs)=Rs/Cs
where Rs is total number of rejections of plant species
most recently eaten, and Cs is the total number of
contacts with plant species most recently eaten. The
Table 1. Plant species eaten by caterpillars observed in nature
(1993–1996)
Plant species
Family
Amaranthus sp.
Ambrosia confertiflora
Bidens leptocephala
Cirsium neomexicanum
Erigeron sp.
Machaeranthera gracilis
Microseris linearifolia
Stephanomeria pauciflora
Amsinckia intermedia
Pectocarya platycarpa
Plagiobothrys arizonicus
Salsola tragus
Acalypha neomexicana
Euphorbia heterophylla
Euphorbia prostrata
Astragalus nothoxys
Lotus humistratus
Lupinus concinnus
Senna leptocarpa
Erodium cicutarium
Phacelia distans
Rhynchosida physocalyx
Sida abutiloides
Sphaeralcea angustifolia
Allionia incarnata
Oenothera primiveris
Plantago patagonica
Bromus rubens
Eriogonum polycladon
Eriogonum wrightii
Rumex sp.
Amaranthaceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Asteraceae
Boraginaceae
Boraginaceae
Boraginaceae
Chenopodiaceae
Euphorbiaceae
Euphorbiaceae
Euphorbiaceae
Fabaceae
Fabaceae
Fabaceae
Fabaceae
Geraniaceae
Hydrophyllaceae
Malvaceae
Malvaceae
Malvaceae
Nyctaginaceae
Onagraceae
Plantaginaceae
Poaceae
Polygonaceae
Polygonaceae
Polygonaceae
probability of accepting different forb species [p(Ad)] was
calculated for each individual as:
p(Ad)=Fd/Cd
where Fd is the total number of feeding events on different forb species, and Cd is the total number of contacts
with different forb species than that most recently eaten.
We analysed the association between leaving, switching and mixing rates (number per h), and average feeding
bout durations of individuals (mean of median bout
durations for each food) because we expected rejection of
food (manifested by locomotion) to result from negative
physiological feedbacks from food (manifested by shortened feeding bouts). We looked at this association among
individual insects because our observations suggested
that certain individuals left hosts frequently in association with brief feeding bouts, while others left hosts
infrequently and fed for relatively long periods at a time.
For field insects observed for 7 h, we used a simple linear
regression to analyse the relationship between leaving
rate, switching rate and mixing rate (response variables in
separate analyses), and average feeding bout duration of
individuals (explanatory variable in each analysis).
Upon finding that switching rates were most strongly
negatively associated with average feeding bout duration
(see Results), we tested the prediction that feeding bouts
631
632
ANIMAL BEHAVIOUR, 64, 4
Table 2. Recipes for artificial diets used in choice and no-choice
experiments
Diet
PC
P
C
Protein source
(% dry weight)
Casein (24.84)
Casein (34.77)
Casein (6.95)
Carbohydrate source
(% dry weight)
Sucrose (16.89)
Sucrose (6.95)
Sucrose (34.77)
All diets contained constant proportions of cellulose (53.60%),
essential micronutrients, minerals and vitamins: salt mix (2.38%),
linoleic acid (0.49%), cholesterol (0.49%), ascorbic acid (0.29%),
vitamin mix (0.06%) and choline chloride (0.74%) in a 3.2%
solution of agar and distilled water. Coumarin or citral (0.25%) was
also added.
associated with switching were shorter than other feeding
bouts. We considered feeding bouts immediately before
(bout 1) and after leaving (bout 2) for each field insect
that moved between individual plants (N=30 insects). We
used one-way analysis of variance (ANOVA) models to
analyse feeding bout durations associated with food
switching (bout 1 and bout 2 were response variables in
separate analyses), with the result of leaving (nonswitches
versus switches: either between different individual
plants of the same species or between different plant
species) as the explanatory variable. These analyses again
used the average bout duration of individuals (mean of
the median bout duration for each switch type; N=54
averaged feeding bouts).
Process of Food Mixing: Experiments with
Synthetic Food
Diets
For both the choice and no-choice tests, we prepared
synthetic diets of varied primary nutrient concentrations
(digestible carbohydrate and protein) and secondary
metabolites (citral and coumarin) in an otherwise constant, chemically defined substrate (Table 2). We chose
one suboptimal diet with a relatively low concentration
of protein relative to digestible carbohydrate, one diet
near the optimal balance of the nutrients according to
previous studies of other caterpillars in the superfamily
Noctuoidea (mean relative protein:carbohydrate (PC)
ratio=57:43, Simpson & Raubenheimer 1993), and
another suboptimal diet with a relatively high concentration of protein relative to digestible carbohydrate.
Specifically, the ratios were: 83:17 P:C, 60:40 P:C, and
17:83 P:C. We estimated that P:C ratios of host plant
foliage span a similar range, based on previously documented nutrient concentrations (Slansky & Scriber 1985).
Hereafter, the first diet will be termed the protein-biased
diet (P), the second will be the balanced diet (PC), and the
third will be the carbohydrate-biased diet (C). The secondary metabolites were intended to represent signal
compounds (but not necessarily toxins) that the caterpillars would perceive as different, and might use to distinguish among host plant species in nature. Coumarin
(cou) and citral (cit) are volatile and likely to stimulate
olfactory receptors as well as gustatory receptors. Because
olfactory inputs are usually more specific than gustatory
sensation in insects (Chapman 1998), we thought the
caterpillars would be likely to discriminate between these
chemicals. The broad taxonomic range of plants used by
G. geneura suggested that coumarin and citral would be
ecologically relevant. We conservatively used relatively
low concentrations (0.25% dry weight of diet) of these
chemicals to minimize their possible toxicity or deterrence to inexperienced insects. Coumarin does stimulate
a gustatory deterrent cell at 5 mM in G. geneura (Bernays
& Chapman 2001); the effect of citral on the taste system
is unknown but appears not to be phagostimulatory from
our observations reported here. We offered the following
diets to insects in the behavioural experiments described
below: Pcou, Ccou, PCcou, Pcit, Ccit, PCcit.
Experimental conditions
Caterpillars (different sets of full siblings in each experiment) were reared until the second or third day of the
final stadium on a standard, wheat germ-based synthetic
diet (no added coumarin or citral) under controlled conditions of 28:25 C, 16:8 h light:dark cycle, then given
chemically defined food for a conditioning period
(defined below for each experiment) prior to observational experiments. We used this conditioning period
to allow time for postingestive effects of the conditioning
foods to influence foraging. About 1 h prior to observations, we replaced the conditioning food with fresh test
food. Cages were transparent, plastic boxes (1111
4 cm) with two screened, ventilation holes (1.5-cm
diameter) on opposite sides. The food pieces provided
(one or two) were approximately 222 cm in size.
During the observations in choice and no-choice tests
(always 0900–1500 hours), we continuously recorded the
caterpillars’ activities using The Observer software program (Noldus 1991). Specifically, we recorded periods of
feeding, walking and resting, as well as the instantaneous
behaviour of tasting, classified as in field observations.
During each of the observation sessions, two observers
each monitored nine individual insects in a room illuminated by fluorescent lights, and held at 28 C (1 C). All
treatments were represented during each session, but
were haphazardly arranged among the two observers,
who were blind to the treatment groups of individual
insects.
Choice experiment
To test our predictions of the nutrient balance and
toxin dilution hypotheses, we analysed the foodswitching behaviour of final-instar larvae continuously
observed over a 6-h period. Caterpillars were caged individually with two pieces of food placed approximately
4 cm apart (approximately the full length of a single
caterpillar), so that switching would cause caterpillars to
move between them. Insects were confined with their
conditioning food for 12 h at 25 C prior to observations.
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
In this experiment, the conditioning food choices and
test food choices for individual caterpillars were the same.
We assumed that foraging patterns would be established
during the conditioning period and that observations
would thus reveal stable patterns.
The experimental design involved six treatment
groups, each replicated 12 times (three in each observation session). Treatments included (1) Ccou/Pcit, (2)
PCcou/PCcit, (3) PCcou/PCcou, (4) PCcit/PCcit, (5) Pcou/
Ccou and (6) Pcit/Ccit. Although these treatments do not
represent the full range of diet combinations that are
possible, they are necessary and sufficient to test the
influence of nutrients, secondary metabolites and their
interaction on food switching. Additional combinations
of diet types (e.g. Ccit/Pcou) could not be tested simultaneously with those above because too few insects were
available. The use of synthetic diets with both manipulated nutrients and secondary metabolites mimics their
combined presence in natural host plants, allowing the
occurrence of possible nutrient–secondary metabolite
interactions. According to the nutrient balance hypothesis, insects should reject and leave food in the P versus C
treatments more readily than in the PC versus PC treatments. In addition, switches to foods with complementary nutrients (P to C, C to P) should result in increased
phagostimulation (measured as increased feeding bout
duration) for the second food over the first. The toxin
dilution hypothesis predicts that rejecting and leaving
should follow excessive exposure to either cou or cit.
Accordingly, switches to foods with different secondary
metabolites (cit to cou, cou to cit) should result in
increased feeding bout duration of the second food
relative to the first.
No-choice experiment
To test the predictions of both hypotheses regarding
the acceptability of foods following a longer period of
exposure to foods that are either nutritionally deficient or
potentially toxic due to secondary metabolites, we specifically addressed the readiness of insects to accept and eat
new foods. We used a C diet as the conditioning food
here because its nutritive bias probably reflects that of
host plant foliage in nature (i.e. nitrogen limited). Similar
responses to citral and coumarin foods in the choice test
(see Results) suggested that it was reasonable to choose
Ccit or Ccou as a conditioning food. We confined finalinstar larvae (as above) with the Ccit diet only for 20 h at
25 C. About 1 h prior to observations, we replaced the
partially consumed Ccit diet with a new piece of Ccit diet
to equalize food quality and availability. During the
observations, we monitored the first feeding bout on Ccit
(the conditioning food), then during the quiescent period
following this bout, we removed the Ccit food and
replaced it with the test food (Ccit again or a novel food).
We carefully placed the new food approximately 1–3 cm
in front of each caterpillar so that it would be likely to
encounter the food soon after activity resumed. By this
method we ensured that caterpillars encountered the new
food without disrupting their normal activity cycle.
Insects responded either by tasting the new food and
walking away (rejection) or by feeding (acceptance). In
two separate analyses, we used the proportion of insects
that accepted new food upon first contact, and the
feeding bout durations of those that accepted new food as
the respective response variables.
This experiment involved four treatment groups of new
foods: Ccit (i.e. same food=control), Ccou (different secondary metabolite), Pcit (complementary nutrient bias),
and Pcou (different secondary metabolite and complementary nutrient bias). Each treatment was replicated
17 times (except Ccou, N=19), and was represented during each observation session. As in the previous experiment, observers were blind to the treatment groups of
individual insects. According to the nutrient balance
hypothesis, Pcou and Pcit, the nutritionally complementary foods to Ccit, should be more acceptable than the
nutritionally similar foods, Ccou and Ccit. The toxin
dilution hypothesis, however, predicts that foods with
novel secondary metabolites, Ccou and Pcou, should
be more acceptable than their nutritionally identical
counterparts, Ccit and Pcit.
Analysis of experimental results
As with insects feeding on plants, we compared the
relationship between leaving rate and average feeding
bout duration among insects in different treatments
of the choice test. We used analysis of covariance
(ANCOVA) to explain interindividual variation in leaving
rate (response variable) with average feeding bout duration of individuals as the covariate, and diet treatment
as the explanatory variable. To qualitatively contrast the
nutrient balance and toxin dilution hypotheses, insects
offered P versus C foods were analysed separately from
those offered PC versus PC foods.
We investigated the possibility that feeding bouts
associated with switches were shorter than other feeding
bouts for two reasons. First, field insects had shortened
feeding bouts immediately prior to switching between
individual plants (see Results). In the choice experiment,
we examined whether this reduction in feeding was a
response to nutrients, secondary metabolites, or both.
Second, we specifically hypothesized that P versus C
treatments would show reduced feeding bouts in association with switches more than PC treatments because
there was a negative association between the leaving rate
and average feeding bout duration of individuals offered
P versus C foods in the choice test with synthetic foods
(see Results). To determine the influence of primary
nutrients and secondary metabolites on switching behaviour, we considered feeding bouts immediately before
(bout 1) and after leaving (bout 2) for each insect that
switched between food pieces in relation to its feeding
bouts that were not part of switches (bout 3). For these
analyses, we also specified that bout 3 feeds preceded (for
comparison with bout 1) or followed (for comparison
with bout 2) bouts of walking. We used two-way ANOVA
models (bout 1 minus bout 3 and bout 2 minus bout
3 were response variables in separate analyses), with
633
ANIMAL BEHAVIOUR, 64, 4
30
Temperature (°C)
nutrient content, ‘N’ (same: PC versus PC; or different: P
versus C), secondary metabolites, ‘T’, (same or different),
and their interaction as explanatory variables. These
analyses used the median feeding bout duration of
individuals for each food type.
To determine whether insects were most stimulated by
new food that either was nutritionally complementary or
contained a different secondary metabolite, we compared
the difference between feeding bouts 1 and 2 among
different switch types in the choice test. Evidence for
either hypothesis would be an increase in the duration of
bout 2 relative to bout 1 when switches were between
foods with complementary nutritive bias or different
secondary metabolites, respectively. The value of bout 2
minus bout 1 was the response variable in a two-way
ANOVA with N and T switch types (same as above) and
their interaction as the factors. The same feeding bout
was never counted more than once in determining an
individual’s median value of bout 2 minus bout 1. Thus
in certain cases (e.g. an individual switched in succession
causing bout 2 of the first switch to be bout 1 of the
second switch), the second occurrence of a feeding bout
was excluded from the analysis.
To assess the acceptability of different food types in the
choice experiment, we compared across food types
(PCcou, PCcit, Pcou, Pcit, Ccou, Ccit) the duration of only
those feeding bouts not associated with switches. In a
one-way ANOVA, the average feeding bout duration of
individuals (mean of median bout duration for each food
type) was the response variable.
To test the predictions of each hypothesis in the
no-choice test, we compared both the outcome of the first
encounter with test food (feed or reject) and, for those
individuals that fed, the duration of the first feeding bout
on each test food (Ccit, Ccou, Pcit, Pcou). We analysed
the outcome of the first encounter across all treatments
with a G test and Williams’s correction (Sokal & Rohlf
1995). We then used G tests on subsets of the data as
pairwise comparisons to evaluate specific hypotheses
(unplanned tests of the homogeneity of replicates tested
for goodness of fit, Sokal & Rohlf 1995). To analyse the
duration of the first and second feeding bouts on test
foods (response variables in separate analyses), we used a
two-way ANOVA with the same switch categories as in
the choice test, ‘N’ and ‘T’ and their interaction, as
factors. We included only feeding bouts lasting longer
than 1 s in the analysis of bout 1 and bout 2.
When necessary, the variables were transformed in all
analyses (log[X+1] for rates and bout durations; arcsine
square-root for proportions) to improve normality and
homogeneity of variance of the data (Zar 1984). We
report an estimate of power for nonsignificant statistical
tests. Such estimates were obtained using an alpha level
of 0.05 and assuming the effect size and variance estimated from sampling data were close to the true parameters (SAS Institute 2000). Under these assumptions, a
power value reflects the likelihood of detecting an effect
given the sample size that was used. All analyses were
performed with the software package JMP (SAS Institute
2000), except G tests, which followed Sokal & Rohlf
(1995).
(a)
Air
Ground
28
26
24
22
20
0800
30
1000
(b)
1200
1400
1600
Feeding bouts
Switches
25
Relative %
634
20
15
10
5
0
0800
1000 1200 1400
Time of day (hours)
1600
Figure 1. (a) Mean air (– –) and ground (– – – –) temperatures (in
the shade) at hourly intervals during 7-h observations of focal
caterpillars. Vertical bars are standard errors. (b) Relative percentages
of the total number of feeding bouts (– –) and switches (– – – –)
between host plant species that occurred each hour during
observation periods.
RESULTS
Food Mixing in Nature
In nature, G. geneura caterpillars spent most of their
time resting, often on the ground or on low-growing
plants. Unlike most caterpillar species, however, they
periodically showed great mobility, locomoting mostly
on the ground and on low, weedy plants. Paths of
movement were often tortuous within patches of plants,
but relatively straight across stretches of bare ground.
During periods of locomotion, caterpillars frequently
contacted plants or other debris with their mouthparts.
Sometimes bites and bouts of feeding followed these
contacts. In many cases, however, caterpillars moved on
after such encounters.
Each of the insects observed for 7 h switched both
between and within host species (median switching
rate=2.4, range 1.0–5.6 switches/h; leaving rate=2.5,
range 1.1–5.7 moves/h). The rates of both feeding and
switching were highest during the morning when ground
temperatures were at or below 25 C (Fig. 1). The rate of
switching among different host plant species (food mixing) was normally distributed with a mean (SE) of
1.30.3 switches between species per hour.
The relationship between individuals’ leaving rates and
their probabilities of subsequently rejecting the plant
species most recently eaten was positive and significant
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
6
(a)
Leaving rate (per h)
0.8
0.6
0.4
0.2
5
4
3
2
1
0
2
3
4
5
Leaving rate (per h)
50
100
150
200
50
100
150
200
6
Figure 2. Relationship between individuals’ leaving rates and the
probability after leaving of rejecting the plant species most recently
eaten ( ), and accepting plant species different from that most
recently eaten ( ) for individuals observed for 7 h in nature (N=10).
Two data points ( ) were at the same location for the relationship
between leaving rate and accepting different plant species.
(R2 =0.87, P<0.0001; Fig. 2), suggesting that leaving a
previously acceptable host plant species signifies its rejection. By contrast, there was no association between individuals’ leaving rates and their probabilities of accepting a
plant species differing from that most recently eaten
(R2 =0.17, P=0.24; Fig. 2). This indicates some degree of
selective feeding: choosing new food is not simply a
matter of accepting the first new plant encountered after
leaving. The basis for this selectivity is addressed with
experimental results reported below.
Leaving Rate and Feeding Bout Duration
For caterpillars observed for 7 h in nature (N=10), there
was a negative association between average feeding bout
duration and either leaving (R2 =0.61, P=0.008), switching (R2 =0.67, P=0.004), or mixing (R2 =0.56, P=0.01) rate
(Fig. 3). These relationships suggest that a behavioural
continuum exists under natural conditions (i.e. multiple
plant species available), with some individuals feeding for
relatively short periods and leaving relatively frequently
at one extreme, and individuals at the other extreme
feeding for long periods and remaining relatively
sedentary.
Insects offered a choice of synthetic diets in the laboratory showed the same negative association between leaving rate and average feeding bout duration when the two
foods were nutritionally unbalanced and complementary
(Table 3, Fig. 4a), but not when foods were nutritionally
identical and balanced (Table 3, Fig. 4b). However, insects
offered nutritionally identical and balanced foods containing different secondary metabolites (PCcit versus
PCcou) showed a weak negative association between
leaving rate and feeding bout duration in contrast to the
weak positive associations between these parameters for
insects offered completely identical foods (Fig. 4b). Thus,
the negative association between feeding bout duration
and leaving rate observed in a field population of cater-
6
(b)
5
Switch rate (per h)
0.0
1
4
3
2
1
0
6
(c)
5
Mixing rate (per h)
Probability after leaving
1.0
4
3
2
1
0
50
100
150
200
Mean feeding bout duration (s)
Figure 3. Relationship between the mean feeding bout durations
and leaving (a), switching (b) and mixing (c) rates of 10 individual
caterpillars feeding on plants in nature. Simple regression lines are
shown for illustrative purposes.
pillars was also present within a single, full sibling family
when caterpillars foraged in a nutritionally diverse
environment, but not when members of the same full
sibling family were offered only nutritionally balanced
food items.
Process of Switching
Analyses of feeding bout durations before and after
leaving provides additional information on the behavioural connection between feeding and switching. The
median feeding bout duration before leaving host plants
635
636
ANIMAL BEHAVIOUR, 64, 4
Table 3. Analysis of covariance to explain interindividual variation in the rate of leaving food in the choice test with
synthetic diets
Source of variation
df
SS
F
P (power)
P versus C treatments
Feeding bout duration
Diet
Feeding bout duration*diet
Error
1
2
2
30
0.594
0.016
0.012
1.075
16.59
0.22
0.17
0.0003
0.8050 (0.29)
0.8413 (0.07)
PC versus PC treatments
Feeding bout duration
Diet
Feeding bout duration*diet
Error
1
2
2
30
0.015
0.075
0.072
0.610
0.73
1.85
1.76
0.3994 (0.13)
0.0331
0.1889 (0.34)
P: 85% protein; C: 85% carbohydrate; PC: 60:40% protein:carbohydrate.
Adjusted R2 (P versus C diets=0.32; PC versus PC diets=0.21).
(field insects: all switching individuals pooled, N=30)
differed according to whether or not the plant subsequently eaten was the same individual, a different individual of the same species, or a different species (ANOVA:
F2,51 =6.42, P=0.003) (bout 1, Fig. 5). Specifically, shorter
feeding bouts preceded switches to different individual
plants of the same (same versus different individual of
same species contrast: F1,51 =9.19, P=0.004) or different
species (same versus different species contrast: F1,51 =8.97,
P=0.004). The median feeding bout following leaving
(bout 2, Fig. 5) did not differ according to switch type
(ANOVA: F2,52 =0.64, P=0.53, power=0.15). Therefore,
shortened feeding bout durations were associated with
the initiation of host plant switching in nature.
The responses of insects offered a choice of synthetic
foods show that the reduced feeding bout duration prior
to switching was associated with effects of secondary
metabolites but not primary nutrients (Table 4). The
mean duration of bout 1 was shorter than the comparable
mean nonswitching bout in treatments containing alternative secondary metabolites (‘different’), but not in
those containing only a single one (‘same’) (Fig. 6). This
was true regardless of the nutrient content of food choices
(P versus C, PC versus PC). Thus, insects responded
differently to food before switching in situations where
they had experienced a single secondary metabolite versus a combination of two during the conditioning period.
Comparisons analogous to those done for field data
(above) revealed no effects of nutrients or secondary
metabolites on the duration of the feeding bout on the
second food (bout 2) relative to nonswitching bouts (data
not shown).
Comparison of the difference between feeding bouts
after and before switches (bout 2bout 1) shows that
feeding bout duration on new food increased when secondary metabolites were different (Fig. 7b, Table 5) but
not when nutrient composition was complementary (Fig.
7a, Table 5). Feeding bout 2 increased relative to bout 1 in
switches between foods with different secondary metabolites (Fig. 7b; PC versus PC, N=8: 5 cou to cit switches;
3 cit to cou switches; P versus C, N=6: 3 Pcit to Ccou
switches; 3 Ccou to Pcit switches), but not in switches
between foods with the same secondary metabolites (PC
versus PC, N=13 and P versus C, N=21). However, feeding
bout 2 did not increase relative to bout 1 in switches
between nutritionally complementary foods (cit versus
cit or cou versus cou, N=21: 11 P to C switches; 10 C to P
switches; Pcit versus Ccou, N=6) compared to switches
between PC foods (Fig. 7a; cit versus cit or cou versus cou,
N=13; cit versus cou, N=6).
In an analysis of all feeding bouts not associated with
switches, bout durations differed across food types in the
choice test (ANOVA on median feeding bout duration of
individuals per food type: F5,88 =3.20, P=0.01; Fig. 8).
Longer feeding bouts on PC and P relative to C foods
indicate the protein-rich foods are the most phagostimulatory (PC versus P contrast, F1,88 =1.20, P=0.28,
power=0.19; PC versus C contrast, F1,88 =13.32, P=
0.0004; P versus C contrast, F1,88 =5.10, P=0.026). Similar
feeding bout durations on coumarin and citral foods of
each nutritive type indicate similar acceptabilities of
these two secondary metabolites (cou versus cit contrast,
F1,88 =0.040, P=0.84, power=0.05).
Responses of insects in the no-choice test confirm that
caterpillars switching to new foods most readily accepted
and increased feeding on those with secondary metabolites that differed from the ones recently experienced. The
duration of final, conditioning feeding bouts on Ccit did
not differ across treatments prior to tests (ANOVA:
F3,58 =0.77, P=0.52, power=0.21). However, the decision
to feed or reject test foods differed according to diet
treatment (G test: 23 =17.28, P<0.001). Further comparisons (Fig. 9) revealed differences for Ccit versus Pcit
(G4,1 =6.46, P<0.05) and Pcit versus Pcou (G4,1 =11.54,
P<0.05), but not for Ccit versus Ccou (G4,1 =1.53, NS) and
Ccou versus Pcou (G4,1 =0, NS). The first feeding bout on
new foods also differed greatly among treatments (Table
6, Fig. 10a). Predictions of the nutrient balance hypothesis alone were not supported: the first feeding bouts on C
foods were longer (opposite to predicted) than those on
P foods (Fig. 10a). By contrast, predictions of the toxin
dilution hypothesis were supported: the first feeding
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
400
Pcit versus Ccou
Pcou versus Ccou
Pcit versus Ccit
(a)
Mean feeding bout duration (s)
4
3
2
Leaving rate (per h)
1
0
4
100
200
300
400
PCcou versus PCcit
PCcou versus PCcou
PCcit versus PCcit
(b)
3
Bout 1
Bout 2
300
200
100
0
SP
NP
NS
Food 2 type relative to food 1
Figure 5. Mean durations of final-instar feeding bouts before (bout
1) and after (bout 2) leaving, categorized by the type of switch
made. The abbreviation ‘SP’ refers to leaving and returning to the
same individual plant, ‘NP’ refers to a switch to a different individual
plant of the same species, and ‘NS’ refers to a switch to a plant of a
different species. Vertical bars are standard errors.
2
1
0
100
200
300
400
500
Mean feeding bout duration (s)
Figure 4. Relationship between the mean feeding bout duration and
leaving rate of individual caterpillars feeding on synthetic diets in the
choice test (diet P: 83% protein; diet C: 83% carbohydrate; diet PC:
60:40% protein:carbohydrate; secondary metabolites: cou: coumarin; cit: citral). These parameters were negatively associated for
insects with access to nutritionally unbalanced, complementary food
types (P versus C foods: (a)), but showed no association for insects
offered two pieces of nutritionally balanced, identical foods (PC
versus PC foods: (b)). Simple regression lines are shown for
illustrative purposes. The continuous lines in (a) and (b) correspond
to the Pcit versus Ccou and PCcit versus PCcit choice situations,
respectively.
bouts on cou foods were longer than those on cit foods
(Fig. 10a). There was no significant interaction between
nutrient content and secondary metabolites (Table 6),
although the statistical power was relatively low (0.08).
The second feeding bout on new foods differed across
diet treatments as well (Table 6, Fig. 10b). The prediction
of the nutrient balance remained unsupported. However,
the effect of secondary metabolites was no longer consistent with the toxin dilution hypothesis. The appearance of
the data (Fig. 10b) suggests an interaction between nutrients and secondary metabolites, yet it was not statistically
significant (Table 6, power=0.33) .
DISCUSSION
Pattern of Food Mixing
The relatively long-term behavioural observations of
caterpillars in nature showed a positive association
between two important components of food switching:
(1) the rate of leaving food and (2) the probability of
subsequently rejecting the food type most recently eaten.
This association not only suggests that leaving food
generally signifies its rejection, but also a mechanistic
connection between leaving rate and the strength of
rejection. Individuals observed to leave host plants at the
highest rates nearly always rejected host plant species
most recently eaten upon subsequent encounter. The
type of mechanism responsible is not known, but could
include physiological (e.g. detoxification enzymatic
activity) or neurological (e.g. behavioural habituation)
processes differentially induced among individuals. That
is, differential induction of detoxification enzymes or
habituation would cause individuals to differ in their
tolerance of host species, with the least tolerant individuals frequently leaving plants and rejecting the same type
upon subsequent encounter. Alternatively, this pattern
could be explained by fixed differences (e.g. genetic
variation) among individuals in the tendency to mix
foods.
Conditional differences in foraging pattern imply that
the process of food mixing behaviour depends upon both
food characteristics and caterpillar experience. Field caterpillars (with a choice of host plants) and caterpillars in
the choice test given two different synthetic foods that
were nutritionally unbalanced showed a negative association between leaving rate and average feeding bout
duration (Figs 3, 4a). For field insects, this pattern is most
simply (but perhaps not only) explained by the finding
that shortened feeding bouts preceded food switching
(Fig. 5). So, individual caterpillars with high switching
(and leaving) rates would have relatively many shortened
feeding bouts, causing a reduction in average feeding
bout duration. Thus when food choices were chemically
heterogeneous, the process of food switching began with
shortened feeding bouts, signifying rejection of the first
food.
637
ANIMAL BEHAVIOUR, 64, 4
Table 4. Two-way analysis of variance showing the effects of primary nutrients (‘nutritive differences’ between food
choices), secondary metabolites (‘toxin differences’ between food choices) and their interaction on the difference
between the average feeding bout duration prior to switching (bout 1) and the average duration of feeding bouts
not associated with switching (bout 3) in the choice experiment
Source of variation
df
SS
F
P (power)
Nutritive differences
Toxin differences
Nutrients*toxins
Error
1
1
1
27
157.82
179 305.31
20 570.66
1 093 340.8
0.0039
4.43
0.51
0.9507 (0.05)
0.0448
0.4821 (0.11)
Adjusted R2 =0.09.
350
200
(a)
100
300
0
250
–100
200
–200
–300
Same
Different
Toxin
Figure 6. The mean difference between feeding bout 1 (bout
immediately before switching) and bout 3 (nonswitch bout), categorized by the type of switch made (same versus different toxin for
each nutritive choice: : P versus C; : PC versus PC). Diet category
abbreviations as in Fig. 4. Negative values indicate that the feeding
bout prior to switches is reduced. Vertical bars are standard errors.
Mean feeding bout duration (s)
Bout 1 – bout 3 (s)
638
150
100
350
(b)
300
250
200
For laboratory insects, however, this pattern cannot be
explained as such because feeding bouts associated with
switches were not shortened for insects in two of the P
versus C treatments. Rather, nutritive variation among
food choices provoked a variable response among the full
sibling individuals, ranging from leaving food relatively
often in association with generally shorter feeding bouts
to rarely leaving and taking longer feeds. This pattern
may be explained by several possibilities. First, individuals may vary in behavioural response to nutrient imbalance according to differences in nutritional requirements
(e.g. due to size or sex). Another possibility is that individuals vary behaviourally per se, perhaps as a bethedging strategy (discussed below) or because there are a
variety of equally successful ways to meet nutritional
requirements. Regardless of the ultimate reason, it is
likely that the more sedentary individuals compensated
for nutritionally unbalanced food by feeding in longer
bouts, thereby causing the negative association between
leaving rate and average feeding bout duration.
The degree to which individual caterpillars (including
those from the same full sibling family) varied in rates of
switching and leaving in response to a choice of food
types is striking. In nature, such behavioural variation
would tend to make some individuals relative host plant
specialists and others relative generalists at any given
150
100
Bout 1
Bout 2
Figure 7. Mean duration of feeding bouts 1 and 2 of switches in the
choice experiment with synthetic diets. Switches between foods that
contained identical, balanced ( : PC versus PC) or complementary
( : P versus C) nutrient bias address the nutrient balance hypothesis
(a). Switches between foods that contained the same ( ) or different
( ) secondary metabolites address the toxin dilution hypothesis (b).
Diet category abbreviations as in Fig. 4. Vertical bars are standard
errors.
time. At an evolutionary level, the variance among individuals in degree of food mixing is consistent with a
strategy of diversified bet hedging (Philippi & Seger 1989;
Carrière et al. 1995). Simply stated, this is a theoretically
distinctive form of risk spreading wherein a genotype
may maximize its geometric mean fitness by producing
phenotypically variable offspring, each phenotype suited
to a different environment. This strategy, like other forms
of bet hedging, is particularly successful when the
environment varies unpredictably over time, as is the case
for the herbaceous plant community in the habitat of
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
(a)
First test bout
250
300
200
200
150
100
100
0
PCcou
PCcit
Pcou
Pcit
Food type
Ccou
Ccit
Figure 8. Mean durations of nonswitch bouts on different food types
in the choice experiment. Vertical bars are standard errors.
Frequency of acceptance
1.0
a
a
a
Mean feeding bout duration (s)
Mean feeding bout duration (s)
400
50
0
Ccit
(b)
Ccou
Pcit
Pcou
Second test bout
250
200
150
0.8
100
b
0.5
50
0.2
0.0
0
Ccit
Ccou
Pcit
Test food
Pcou
Figure 9. Proportion of insects previously conditioned on Ccit that
accepted the test food upon first contact in the no-choice test.
Diet category abbreviations as in Fig. 4. Letters denote significant
differences derived from G tests (see text for details) .
G. geneura (Singer 2000). While the mechanistic basis for
this variation cannot be evaluated without studies of
behavioural repeatability or heritability, the variation
itself is at least partly due to different leaving rates of
individual caterpillars in response to nutritional variation. This demonstration of individual variation was the
only influence of primary nutrients on food-switching
behaviour.
Process of Switching
Laboratory experiments with synthetic diets revealed
how effects of secondary metabolites predominated
throughout the process of switching between different
food types, the essence of food mixing behaviour. Effects
of secondary metabolites but not primary nutrients
reduced feeding prior to switches, apparently inducing
caterpillars to leave food. Secondary metabolites further
determined the acceptability of food subsequently
encountered in a manner consistent with toxin dilution.
The reduced feeding bout duration immediately prior
to switching foods in both field and laboratory settings
Ccit
Ccou
Pcit
Test food
Pcou
Figure 10. Mean durations of the first (a) and second (b) feeding
bouts on test foods for insects previously conditioned on Ccit in the
no-choice experiment. Diet category abbreviations as in Fig. 4.
Vertical bars are standard errors.
indicates reduced feeding excitation (Simpson 1995). This
appears to be a response to secondary metabolites in the
present study. In the laboratory choice experiment, feeding bouts prior to switches were reduced when food
choices contained different secondary metabolites (as
in nature), consistent with toxin dilution operating via
several possible processes. First, insects confined to food
choices with a single secondary metabolite (unlike field
insects) would have ingested relatively greater amounts of
it than insects that switched between foods with different
secondary metabolites. As a result of increased ingestion,
the former could have relatively increased activity of
detoxification enzymes (Glendinning & Slansky 1995;
Snyder & Glendinning 1996). This metabolic increase
might curtail the potentially toxic accumulation of
ingested secondary metabolites, allowing the retention of
a relatively high level of feeding excitation. Indeed, there
was no reduction in feeding prior to switching for insects
with a choice between foods with the same secondary
metabolite in this experiment. Second, the presence of
alternative foods may be detected: feeding insects might
smell or become distracted by odours of different secondary metabolites from nearby food, curtail feeding, then
switch. If no alternative odours are detectable, feeding
excitation may remain constant until other processes
639
640
ANIMAL BEHAVIOUR, 64, 4
Table 5. Two-way analysis of variance showing the effects of primary nutrients (‘nutritive differences’ between food
choices), secondary metabolites (‘toxin differences’ between food choices) and their interaction on the difference
between the average duration of feeding bouts immediately prior to switching (bout 1) and the average duration
of feeding bouts immediately after switching (bout 2) in the choice experiment
Source of variation
df
SS
F
P (power)
Nutritive differences
Toxin differences
Nutrients*toxins
Error
1
1
1
27
9403.46
169 742.97
9538.52
1 080 652.1
0.24
4.24
0.24
0.6318 (0.08)
0.0492
0.6294 (0.08)
Adjusted R2 =0.07.
Table 6. Analysis of variance showing the effects of primary nutrients (‘nutritive differences’ between conditioning
and test foods), secondary metabolites (‘toxin differences’ between conditioning and test foods) and their
interaction on the duration of the first (bout 1) and second (bout 2) feeding bouts on test food in the no-choice
experiment
Source of variation
df
SS
F
P (power)
Bout 1
Nutritive differences
Toxin differences
Nutrients*toxins
Error
1
1
1
58
2.98
1.33
0.056
10.70
16.12
7.23
0.30
0.0002
0.0093
0.5854 (0.08)
Bout 2
Nutritive differences
Toxin differences
Nutrients*toxins
Error
1
1
1
52
3.25
0.63
0.79
20.06
8.75
1.69
2.14
0.0021
0.6986 (0.07)
0.1305 (0.33)
Adjusted R2 (bout 1=0.23, bout 2=0.13).
(e.g. intrinsic activity rhythm: Bernays & Singer 1998)
cause switching. Third, insects might habituate to particular secondary metabolites via a reduction in chemosensory sensitivity, possibly related to detoxification
activity (Glendinning & Slansky 1995). Fourth, insects
could have learned whether more than one food type
(based on secondary metabolites) was present, resulting
in neurally driven changes in response to foods. These
processes are not mutually exclusive.
Feeding stimulation from new foods was consistent
with toxin dilution but not nutrient balancing. Switches
in the choice test showed that new foods with a different
secondary metabolite were more phagostimulatory than
foods with the same one (Fig. 7b). Evidence for this comes
from the increase in feeding duration from bout 1 to bout
2 when the second food contained a different secondary
metabolite, but no such increase when foods in bout 1
and 2 contained the same secondary metabolite. When
switches occurred between nutritionally complementary
foods, there was no sign of increased phagostimulation
by the second food relative to the first (Fig. 7a). This
experiment does not rule out the possibility of feeding
preference based on nutritive need occurring over a
longer time scale (e.g. >4 h as reported for Spodoptera
littoralis: Simpson et al. 1988). However, such a process
would not explain the high frequency of host plant
switching by G. geneura in nature.
The results of the no-choice experiment generally
matched the prediction that caterpillars exposed to food
containing one secondary metabolite would more readily
accept (Fig. 9) and initially eat more (Fig. 10a) food with
a new secondary metabolite relative to food with the
same secondary metabolite. For food acceptance, this
prediction was upheld only for insects tested with the
protein-biased foods. The effect may have been absent for
the carbohydrate-biased diet because of the extremely
high sucrose concentrations in the food, likely to overwhelm differences in chemosensory response to the
coumarin and citral. Sucrose is a potent phagostimulant
for G. geneura (Bernays et al. 2000). Because insects cannot taste protein directly, the secondary metabolites were
likely to have dominated the flavour of the protein-biased
foods, causing a large difference in the acceptability of
Pcit and Pcou. Initial feeding bout durations for those
insects that initiated feeding revealed no interaction
between primary nutrients and secondary metabolites,
consistent with other work showing no effect of interactions between dietary nutrient content and secondary
metabolites on deterrency (Glendinning & Slansky 1994).
Together, these results suggest that caterpillars in nature
are likely to reject, or feed only briefly, on host plants
containing secondary metabolites recently ingested in
large amounts. The relatively high probability of caterpillars rejecting host plant species most recently eaten
supports this idea (Fig. 2).
Changes in feeding duration between the first two
bouts on test food in the no-choice test suggest the
importance of postingestive feedbacks from secondary
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
metabolites and nutrients. The mean feeding durations
on Ccou and Ccit foods reversed in the second bout
relative to the first (Fig. 10a, b), indicating possible post
ingestive effects of secondary metabolites. Relatively long
bouts on Ccit may have been possible because detoxification enzyme activity induced over the previous 20 h
reduced the accumulation of citral, thus allowing
a maximal amount of compensatory feeding on the
protein-deficient food. The decrease in feeding on Ccou is
consistent with short-term sensitization to coumarin or
the accumulation of quantities that provided some negative postingestive feedback. The difference in bout duration between Pcit and Pcou foods was maintained in the
second bout, perhaps because many of the insects in the
P treatments fed so little (or not at all) upon first contact.
Therefore, these insects received little chemosensory or
postingestive information that could be used to modify
subsequent feeding. Some of the insects in the P treatments (especially those with the novel secondary compound), however, fed for longer periods and probably
gained positive, postingestive feedbacks from nutrients.
This explanation is suggested by the positive association
between the duration of feeding bouts 1 and 2 for individual caterpillars in the P treatments (Spearman rank
correlation: rS =0.38, N=33, P<0.03), but not in the C
treatments (rS = 0.03, N=36, NS).
Results of the no-choice test did not support the
mechanistic prediction that nutritionally complementary
foods should be most phagostimulatory following
exposure to nutritionally unbalanced food (Simpson
& Raubenheimer 1996). After 20 h of exposure to
carbohydrate-biased food, caterpillar feeding bout durations on new, protein-biased foods were considerably
shorter than those on new, carbohydrate-biased foods
(Fig. 10a). These feeding bouts were also much shorter
than those on the same protein-biased foods in the choice
test (Fig. 8). The comparison of nonswitch feeding bouts
across foods in the choice test (longer bouts on PC and P
than on C foods) dispel the possibility that final instars
rejected P foods in the no-choice test because their nutrient intake targets (sensu Simpson & Raubenheimer 1993)
were more carbohydrate biased than anticipated. Rather,
this difference probably indicates that feeding bout durations in the choice test show the ‘normal’ feeding of
undeprived, nutritionally homeostatic insects.
A possible problem in using synthetic diets like the
ones used here is their simplicity relative to natural host
plants. The surface concentrations of amino acids and
sugars in plants, which directly elicit gustatory responses
of insects, may not accurately represent the internal
concentrations of digestible protein and carbohydrate
(Bernays & Chapman 1994). While the lack of added free
amino acids in the synthetic diets used here may be
unrealistic, several important parameters of foraging
behaviour were similar between insect responses to these
diets and natural host plants. We do not think these diets
diminished our chance of detecting nutrient balancing,
as experiments with similar diets have demonstrated
feeding decisions based on nutrient balancing over similar time scales in grasshoppers (Simpson & Abisgold 1985;
Chambers et al. 1997) and in the noctuid caterpillars
Helicoverpa (=Heliothis) zea (Cohen et al. 1988; Friedman
et al. 1991) and Spodoptera littoralis (Simmonds et al.
1992). However, the low concentrations of single
secondary metabolites used in our synthetic diets
may have reduced our chances of detecting behaviour
consistent with toxin dilution, making this study a rather
conservative test of its relative importance.
We have presented experimental evidence for the interplay between processes that balance nutrient intake and
those that reduce ingestion of particular secondary
metabolites in food mixing behaviour, along with field
observations consistent with these experimental results.
Food mixing in herbivorous insects may frequently
involve both nutrient balancing and toxin dilution, as
theory generally predicts. It would be informative to
know if these processes vary in relative importance
among different food mixing species. One might predict,
for example, that species feeding on a mixture of grasses
(e.g. Locusta migratoria, Orthoptera) would not often
encounter the problem of toxicity from secondary
metabolites, which are typically not at noxious concentrations in grass tissue. However, they would face great
difficulty in ingesting an optimal nutrient balance due to
typically high carbohydrate:nutrient ratios in grass tissue.
Indeed L. migratoria routinely chooses a balanced nutrient
intake from various combinations of nutritionally unbalanced but complementary foods (Behmer et al. 2001).
By contrast, species like G. geneura that feed on forbs,
characteristically defended with potently deterrent or
toxic secondary metabolites, would be expected to face the
problem of toxicity to a great degree. Further work that
relates physiological processes and behaviour observed in
the laboratory to behavioural patterns in nature will be
necessary to determine more generally the relative roles of
nutrient balancing and toxin dilution in food mixing.
Acknowledgments
We thank R. F. Chapman for help in designing experiments and helping us run them blind, and R. A.
Abernathy, J. L. Spencer and E. M. Wintermute for assistance with observations. A. Telang kindly offered recipes
for synthetic diets. R. F. Chapman, N. Moran, A. Mira and
an anonymous referee critically reviewed earlier drafts of
this manuscript. We especially thank F. Slansky for a
constructively critical, thorough review of the manuscript. This work was supported by a fellowship to M.S.S.
from The Center for Insect Science, University of Arizona,
and by the NSF-funded Research Training Group in
the Analysis of Biological Diversification, University of
Arizona.
References
Abisgold, J. D., Simpson, S. J. & Douglas, A. E. 1994. Nutrient
regulation in the pea aphid Acyrthosiphon pisum: application
of a novel geometric framework to sugar and amino acid
consumption. Physiological Entomology, 19, 95–102.
Behmer, S. T., Raubenheimer, D. & Simpson, S. J. 2001.
Frequency-dependent food selection in locusts: a geometric
analysis of the role of nutrient balancing. Animal Behaviour, 61,
995–1005.
641
642
ANIMAL BEHAVIOUR, 64, 4
Bernays, E. A. & Chapman, R. F. 1994. Host-plant Selection by
Phytophagous Insects. New York: Chapman & Hall.
Bernays, E. A. & Chapman, R. F. 2000. Plant secondary compounds
and grasshoppers: beyond plant defenses. Journal of Chemical
Ecology, 26, 1773–1794.
Bernays, E. A. & Chapman, R. F. 2001. Taste cell responses in the
polyphagous arctiid, Grammia geneura: towards a general pattern
for caterpillars. Journal of Insect Physiology, 47, 1029–1043.
Bernays, E. A. & Singer, M. S. 1998. A rhythm underlying feeding
behaviour in a highly polyphagous caterpillar. Physiological
Entomology, 23, 295–302.
Bernays, E. A., Bright, K., Howard, J. J., Raubenheimer, D. &
Champagne, D. 1992. Variety is the spice of life: frequent
switching between foods in the polyphagous grasshopper
Taeniopoda eques Burmeister (Orthoptera: Acrididae). Animal
Behaviour, 44, 721–731.
Bernays, E. A., Bright, K. L., Gonzalez, N. & Angel, J. 1994. Dietary
mixing in a generalist herbivore: tests of two hypotheses. Ecology,
75, 1997–2006.
Bernays, E. A., Chapman, R. F. & Singer, M. S. 2000. Sensitivity to
chemically diverse phagostimulants in a single gustatory neuron
of a polyphagous caterpillar. Journal of Comparative Physiology A,
186, 13–19.
Carrière, Y., Roff, D. A. & Deland, J. P. 1995. The joint evolution of
diapause and insecticide resistance: a test of an optimality model.
Ecology, 76, 1497–1505.
Chambers, P., Sword, G., Angel, J. E., Behmer, S. & Bernays,
E. A. 1996. Foraging by generalist grasshoppers: two different
strategies. Animal Behaviour, 52, 155–165.
Chambers, P. G., Raubenheimer, D. & Simpson, S. J. 1997. The
selection of nutritionally balanced foods by Locusta migratoria:
the interaction between food nutrients and added flavours.
Physiological Entomology, 22, 199–206.
Chapman, R. F. 1998. The Insects: Structure and Function. 4th edn.
Cambridge: Cambridge University Press.
Cohen, R. W., Waldbauer, G. P. & Friedman, S. 1988. Natural diets
and self-selection: Heliothis zea larvae and maize. Entomologia
Experimentalis et Applicata, 46, 161–171.
Dethier, V. G. 1988. The feeding behaviour of a polyphagous
caterpillar (Diacrisia virginica) in its natural habitat. Canadian
Journal of Zoology, 66, 1280–1288.
Freeland, W. J. & Janzen, D. H. 1974. Strategies in herbivory by
mammals: the role of plant secondary compounds. American
Naturalist, 108, 269–289.
Freeland, W. J. & Saladin, L. R. 1989. Choice of mixed diets
by herbivores: the idiosyncratic effects of plant secondary
compounds. Biochemical Systematics and Ecology, 17, 493–497.
Friedman, S., Waldbauer, G. P., Eertmoed, J. E., Naeem, M. &
Ghent, A. W. 1991. Blood trehalose levels have a role in the
control of dietary self-selection by Heliothis zea larvae. Journal of
Insect Physiology, 37, 919–928.
Glendinning, J. I. & Slansky, F., Jr. 1994. Interactions of
allelochemicals with dietary constituents: effects on deterrency.
Physiological Entomology, 19, 173–186.
Glendinning, J. I. & Slansky, F., Jr. 1995. Consumption of a toxic
food by caterpillars increases with dietary exposure: support for a
role of induced detoxification enzymes. Journal of Comparative
Physiology A, 176, 337–345.
Hagele, B. F. & Rowell-Rahier, M. 1999. Dietary mixing in three
generalist herbivores: nutrient complementation or toxin dilution?
Oecologia, 119, 521–533.
Karban, R. & Baldwin, I. T. 1997. Induced Responses to Herbivory.
Chicago: University of Chicago Press.
Noldus, L. P. J. J. 1991. The Observer: a software system for
collection and analysis of observational data. Behavioural Research
Methods, Instruments and Computers, 23, 415–429.
Philippi, T. & Seger, J. 1989. Hedging one’s evolutionary bets,
revisited. Trends in Ecology and Evolutionary Biology, 4, 41–44.
Pulliam, H. R. 1975. Diet optimization with nutrient constraints.
American Naturalist, 109, 765–768.
Rapport, D. J. 1980. Optimal foraging for complementary
resources. American Naturalist, 116, 324–346.
Raubenheimer, D. & Bernays, E. A. 1993. Patterns of feeding in the
polyphagous grasshopper Taeniopoda eques: a field study. Animal
Behaviour, 45, 153–167.
SAS Institute. 2000. JMP Version 4. Cary, North Carolina: SAS
Institute.
Simmonds, M. S. J., Simpson, S. J. & Blaney, W. M. 1992. Dietary
selection behaviour in Spodoptera littoralis: the effects of conditioning diet and conditioning period on neural responsiveness
and selection behaviour. Journal of Experimental Biology, 162,
73–90.
Simpson, S. J. 1982. Patterns in feeding: a behavioural analysis using
Locusta migratoria nymphs. Physiological Entomology, 4, 325–
336.
Simpson, S. J. 1995. Regulation of a meal: chewing insects. In:
Regulatory Mechanisms in Insect Feeding (Ed. by R. F. Chapman &
G. de Boer), pp. 137–156. New York: Chapman & Hall.
Simpson, S. J. & Abisgold, J. D. 1985. Compensation by locusts
for changes in dietary nutrients: behavioural mechanisms.
Physiological Entomology, 10, 443–452.
Simpson, S. J. & Raubenheimer, D. 1993. A multi-level analysis
of feeding behaviour: the geometry of nutritional decisions.
Philosophical Transactions of the Royal Society of London, Series B,
342, 381–402.
Simpson, S. J. & Raubenheimer, D. 1996. Feeding behaviour,
sensory physiology and nutrient feedback: a unifying model.
Entomologia Experimentalis et Applicata, 80, 55–64.
Simpson, S. J. & Raubenheimer, D. 2001. The geometric analysis of
nutrient–allelochemical interactions: a case study using locusts.
Ecology, 82, 422–439.
Simpson, S. J., Simmonds, M. S. J. & Blaney, W. M. 1988.
A comparison of dietary selection behaviour in larval Locusta
migratoria and Spodoptera littoralis. Physiological Entomology, 13,
225–238.
Simpson, S. J., Raubenheimer, D. & Chambers, P. G. 1995. The
mechanisms of nutritional homeostasis. In: Regulatory Mechanisms
in Insect Feeding (Ed. by R. F. Chapman & G. de Boer), pp. 251–
278. New York: Chapman & Hall.
Singer, M. S. 2000. Ecological maintenance of food mixing in the
woolly bear caterpillar Grammia geneura (Strecker) (Lepidoptera:
Arctiidae). Ph.D. thesis, University of Arizona, Tucson.
Singer, M. S. & Stireman, J. O., III. 2001. How foraging tactics
determine host-plant use by a polyphagous caterpillar. Oecologia,
129, 98–105.
Slansky, F., Jr. 1992. Allelochemical-nutrient interactions in
herbivore nutritional ecology. In: Herbivores: Their Interactions with
Secondary Metabolites. Vol. 2 (Ed. by G. A. Rosenthal & M. R.
Berenbaum), pp. 135–174. New York: Academic Press.
Slansky, F., Jr & Scriber, J. M. 1985. Food consumption and
utilization. In: Comprehensive Insect Physiology, Biochemistry and
Pharmacology. Vol. 4 (Ed. by G. A. Kerkut & L. I. Gilbert),
pp. 87–163. Oxford: Pergamon.
Snyder, M. J. & Glendinning, J. I. 1996. Causal connection
between detoxification enzyme activity and consumption of a
toxic plant compound. Journal of Comparative Physiology A, 179,
255–261.
Sokal, R. R. & Rohlf, F. J. 1995. Biometry. New York: W. H. Freeman.
Stephens, D. W. & Krebs, J. R. 1986. Foraging Theory. Princeton,
New Jersey: Princeton University Press.
Waldbauer, G. P. & Friedman, S. 1991. Self-selection of optimal
diets by insects. Annual Review of Entomology, 36, 43–63.
SINGER ET AL.: NUTRIENT BALANCE AND TOXIN DILUTION
Waldbauer, G. P., Cohen, R. W. & Friedman, S. 1984. Selfselection of an optimal nutrient mix from defined diets by larvae of
the corn earworm, Heliothis zea (Boddie). Physiological Zoology,
57, 590–597.
Westoby, M. 1978. What are the biological bases of varied diets?
American Naturalist, 112, 627–631.
Zar, J. H. 1984. Biostatistical Analysis. 2nd edn. Englewood Cliffs,
New Jersey: Prentice Hall.
643