Journal of Tropical Ecology (2006) 22:25–33.

Copyright © 2006 Cambridge University Press

doi:10.1017/S0266467405002750 Printed in the United Kingdom

Diversity and ecology of carrion- and fruit-feeding butterflies in Bornean

rain forest

K. C. Hamer∗1 , J. K. Hill†, S. Benedick‡, N. Mustaffa‡, V. K. Chey§ and M. Maryati‡

∗ School of Biology and Earth Biosphere Institute, University of Leeds LS2 9JT, UK

† Department of Biology, University of York, UK

‡ Institute for Tropical Biology and Conservation, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia
‡ Forest Research Centre, Sandakan, Sabah, Malaysia
(Accepted 20 June 2005)

Abstract: Tropical rain forests are well known as centres of insect diversity and much effort has focused on the role
of larval host plant specificity in generating and maintaining this diversity, but fewer studies have examined the
exploitation of different food resources by adults in this context. Tropical butterflies feed as adults on a wide range of
resources and we examined the diversity and ecology of species feeding on rotting fruit and carrion in a tropical lowland
rain forest in Sabah, Borneo. We found that species richness and diversity were significantly higher on carrion than
on fruit, and that this pattern was repeated at genus and family level. There was little similarity in species assemblages
on the two substrates and β-diversity between carrion and fruit comprised 33% of the total diversity of butterflies
feeding on decaying matter. β-diversity between canopy gap and shade microhabitats comprised 21% of total species
diversity on carrion but only 7% of the total on fruit, indicating greater functional diversity on carrion in terms of
light preferences. Captures were strongly male-biased on carrion but not on fruit, and recapture rates were much
lower on carrion than on fruit. Species from two subfamilies (Nymphalinae and Charaxinae) exploited both substrates
and for Charaxinae, there was evidence from adult flight morphology that species on carrion were capable of faster
more-powerful flight. These results support the notion of a distinctive carrion-feeding fauna comprising more mobile
species, which may use carrion to meet additional nitrogen requirements resulting from greater musculature. However
there was no relationship between flight morphology and substrate choice in the Nymphalinae, and carrion-feeding
may not have a unitary explanation.

Key Words: diversity partitioning, flight morphology, Lepidoptera, mud-puddling, tropical forest insects

INTRODUCTION the biology and ecology of different species, but few studies
have investigated tropical insects that feed as adults on
An important part of understanding species’ life histories
markedly different substrates (Blüthgen et al. 2003), and
is determining the selective pressures favouring the
so any differences among these species are only poorly
acquisition of different nutrients needed for reproduction
understood, and little is known about what determines
and survival (Boggs 1999, Karlsson 1995). This situation
substrate choice.
can be complex in holometabolous insects, where food
Tropical insects are highly diverse with many endemic
resources often differ markedly between larval and
species dependent on closed-canopy forest (Hamer et al.
adult stages. For instance, butterflies (Lepidoptera) are
2003, Sutton & Collins 1991). Much effort has focused
predominantly herbivorous as larvae whereas adults feed
on the potential role of larval host plant specificity in
on a much wider variety of food resources, especially
generating and maintaining this diversity (Fiedler 1998,
in the tropics where frequently used resources include
Megens et al. 2005, Novotny et al. 2002), but fewer
nectar, pollen and decaying organic matter (Adler 1982,
studies have examined the exploitation of different food
DeVries 1987). The exploitation of such different food
resources by adults in this context (Blüthgen et al. 2003,
resources is likely to be linked to widespread differences in
Erwin 1982). In temperate regions, adult Lepidoptera
most commonly obtain nutrients from nectar sources, and
1 Corresponding author. Email: k.c.hamer@leeds.ac.uk this behaviour has been well studied (May 1992, Scoble
26 K. C. HAMER ET AL.

1992). In the tropics, it is well known that members of test the hypothesis that carrion is exploited primarily by
certain groups such as Nymphalidae feed on rotting fruit species with rapid, powerful flight.
(DeVries & Walla 2001, Hamer et al. 2003, Schulze et al.
2001). In addition, some tropical species forage on carrion
(Austin & Riley 1995, Hall & Willmott 2000), although METHODS
it is not clear how widespread this phenomenon is, which
groups are involved, how distinct carrion-feeders are from Fieldwork took place at Danum Valley Field Centre,
those feeding on decaying fruit or to what extent carrion- Sabah (Malaysian Borneo; 5 ◦ N, 117 ◦ 50 E). The study
feeders contribute to overall species diversity. area lies within the Danum Valley Conservation Area
Among neotropical butterflies in the family Riodinidae, which is a 428-km2 area of lowland evergreen rain
carrion-feeders have different flight morphology from forest (details of site in Marsh & Greer 1992). Rainfall
other species, and may use carrion to provide sodium and averages 2700 mm y−1 , with some seasonal variation
other nutrients needed to maintain high metabolic rates but no pronounced dry season (Hamer et al. 2005).
during rapid flight (Hall & Willmott 2000). Butterflies We sampled fruit- and carrion-feeders with traps baited
with rapid, powerful flight are often associated with rain with either carrion (a mixture of prawn paste and fish)
forest gaps and edges (Hamer et al. 2003, Hill et al. or banana (see DeVries 1987 for details of trap design).
2001). Thus if carrion-feeders are rapid fliers, they may To examine the diversity of carrion- and fruit-feeders in
be particularly associated with canopy gaps, but data are relation to light regime, traps were hung 1–2 m from the
lacking to compare the microhabitats of carrion-feeders ground, from branches of trees > 50 m apart in forest
and other species. Moreover the relationship between gaps or under dense canopy. The forest gaps were formed
flight morphology and substrate choice has not been by natural treefalls, were at least 10 × 10 m in size and
evaluated for taxa other than Riodinidae and there are received full sun at midday. Average canopy openness
alternative explanations for carrion-feeding. For instance, (measured using a spherical densiometer; Lemmon 1957)
males may visit carrion to obtain sodium, or possibly other was 20% in gaps compared with < 1% in shade sites. Bait
nutrients, to transfer to females at mating (Pivnick & was placed in traps on the day prior to the first sampling
McNeil 1987, Sculley & Boggs 1996). However there day, and was left in the trap for the rest of the sampling
are relatively few data to evaluate the proportions of period. Fresh bait was added to each trap every day; this
males and females in different species that feed on ensured that all traps contained a mixture from fresh to
carrion, and further data are needed to address this well-rotted bait.
question. Traps were operated over a total of 70 d during
In addition to sex ratios, recapture rates on different three study periods; 18 September–14 October 2000,
substrates could also provide useful information about 22 August–12 October 2001, and 22 December 2001–
what determines substrate choice. For instance, if adults 3 January 2002. Between 24–40 traps were deployed
are utilizing both fruit and carrion as sources of metabolic during each period, with equal numbers of traps in gap
substrates to fuel continued activity, then we might expect and shade and baited with fruit and carrion, resulting
reasonably similar recapture rates on the two substrates, in a total of 2080 trap days. Traps were emptied daily
at least within a species. Conversely, if carrion is used between 15h00 and 17h00, and all trapped butterflies
mainly by newly emerged adults to provide nutrients prior were identified where possible in the field (following
to mating, we might expect much lower recapture rates Otsuka 1988, 1991), marked with a felt-tipped pen and
on carrion than on fruit. To our knowledge, however, no released. Some species (most Lycaenidae and Hesperiidae,
previous study has compared recapture rates on different plus genera Euthalia and Tanaecia) cannot be identified
substrates. reliably in the field and were collected and identified
In this paper, we examine the diversity of butterfly in the laboratory using keys and figures in Corbet &
species feeding on fruit and carrion in a tropical Pendlebury (1992) and Otsuka (1991); where necessary,
lowland rain forest in Sabah, Borneo. We compare the this included dissection of male genitalia. Mounted
diversity of species exploiting each substrate in gap and specimens and dissected genitalia are housed in the
shade microhabitats and we use additive partitioning Institute of Tropical Biology and Conservation, Universiti
(Lande 1996, Veech et al. 2002) to assess the relative Malaysia Sabah.
contributions of food resource and light regime to the
overall diversity of butterflies feeding on decaying matter.
We examine the utilization of carrion and fruit by species Species diversity on carrion and fruit
in different butterfly families and subfamilies and we
compare species on carrion and fruit in terms of sex In addition to comparing the numbers of species on
ratios and recapture rates. We assess the relationship each food type, we also examined species diversity
between flight morphology and substrate choice, and using Margalef’s index, following recommendations and
Carrion-feeding butterflies in rain forest 27

formula in Magurran (1988). This index is largely abundance in gap versus shade traps (range from 0 to
independent of the underlying species abundance 1 where 0 = trapped only in shade, 1 = trapped only in
distribution (Magurran 1988). To examine whether or gaps). This index was highly correlated with a separate
not species diversity differed between food types, we index of light preference calculated from independent data
used bootstrap methods to calculate standard errors for (Appendix 1 in Hamer et al. 2003) for species with n ≥ 2
Margalef’s index (Sokal & Rohlf 1995) and compared (Pearson correlation using arcsine-transformed data;
between samples using pair-wise randomization tests r = 0.89, n = 30 species, P < 0.001) and so as with carrion
based on 10 000 re-samples of species abundance data, preference, we are confident that we could characterize
following Solow (1993). This test calculates the observed species’ light preferences in this way.
difference in diversity between samples (δ) and assesses All trapped butterflies were measured before being
the significance of this value from its position among released or collected for identification. The following
the ordered values of δ found by randomization. For a measurements were taken by JKH using Vernier calipers
two-tailed test, the estimated P-value is given by the to an accuracy of 0.1 mm; body length, thoracic length
proportion of re-samples with values of δ greater than and width, forewing length (wing base to apex) and
the observed value (Solow 1993). This is an exact test; breadth (minimum distance between tornus and costa).
the nominal significance level is equal to the actual rate These measurements were then used to calculate two
of false rejection of the null hypothesis. variables; thoracic volume (length × width2 ) and wing
To determine whether the greatest similarity was be- area (wing length × breadth). In butterflies, flight speed
tween species assemblages on the same food resources is positively correlated with thoracic mass (principally
or in the same microhabitat, we calculated Sørenson’s flight muscle mass; Chai & Srygley 1990, Dudley 1990,
quantitative similarity coefficient (CN ; accounting for rel- Marden 1987) and thoracic mass is directly proportional
ative abundance in addition to species presence/absence) to thoracic volume (Srygley & Chai 1990). In addition, the
for all individuals recorded on each combination of food ratio of wing area to thoracic mass is strongly negatively
resource and microhabitat (formula in Magurran 1988). related to flight speed and metabolic rate during flight
To examine the contributions of microhabitat and food activity (Dudley & Srygley 1994, Hall & Willmott 2000).
resource to the overall diversity of butterflies feeding on We thus compared carrion- and fruit-feeders in terms
decaying matter, we used additive partitioning (Crist et al. of both thoracic volume and the ratio of wing area
2003, Gering et al. 2003) in which the total diversity of to thoracic volume. Wing area is proportional to wing
species was partitioned into positive, additive components length squared whereas thoracic volume is proportional
within microhabitats and food resources (α1 ), between to thoracic length cubed, and so we used values of
microhabitats within food resources (β1 ) and between thoracic volume raised to the power of 2/3 to create
food resources (β2 ). This resulted in α and β diversity a dimensionless ratio (hereafter termed WA:TV ratio,
being expressed in the same units, in contrast to the following Hall & Willmott 2000). All variables were log10 -
more commonly used analyses of β diversity involving transformed for analysis. Mean values were calculated for
dimensionless metrics based on species dissimilarity or males and females of each species. Both thoracic volume
turnover between pairs of samples (Magurran 1988). and WA:TV ratio were allometrically related to body
length and to account for this relationship, data were
analysed by analysis of covariance (ANCOVA) with log
Ecology and morphology of carrion- and fruit-feeders body length as a covariate. All analyses were weighted by
sample size.
To compare the ecology of species exploiting fruit and
carrion, we determined each species’ resource preferences
in terms of their proportional abundance in fruit versus RESULTS
carrion traps (index of carrion preference ranged from 0
to 1 where 0 = trapped only on fruit, 1 = trapped only Species diversity
on carrion). To reduce sampling error, we included only
species where the total number of individuals sampled Excluding recaptures, we sampled 1095 individuals of
was ≥ 2, following Davis et al. (2001) and Ribera et al. 85 species, comprising 631 individuals of 45 species
(1998). All species sampled in large numbers had a on fruit plus 464 individuals of 61 species on carrion
very clear preference (> 90% of individuals) for one (Appendix 1). The greater species richness on carrion
substrate or other (see below) and so we are confident was not an artefact of variation in sampling efficiency:
that this sample size was sufficient to characterize species’ after accounting for the different numbers of individuals
substrate preferences reasonably well. In addition to on the two food-types, species diversity (Margalef’s index)
carrion preference, we also calculated an index of each was significantly higher on carrion than on fruit (Table 1;
species’ light preference in terms of their proportional randomization test; δ = 2.92, P < 0.01). Analysing the
28 K. C. HAMER ET AL.

Table 1. Species richness, abundance and diversity of butterflies feeding on fruit and carrion at gap and shade sites.

Carrion Fruit
Overall
Gap Shade Total Gap Shade Total Total
No. species 46 40 65 37 38 45 85
No. individuals 269 195 464 253 378 631 1095
Margalef’s index ± SE 8.04 ± 0.82 7.40 ± 0.78 9.77 ± 0.83 6.52 ± 0.65 6.23 ± 0.68 6.82 ± 0.56 12.00 ± 0.81

100 clear preference within each species for one substrate or

other. For instance, among those 19 species with >10
Proportion of total (%)

β2 individuals recorded, the mean percentage of individuals

80 on the substrate with greater abundance was 92.3%
(SD ± 9.7; Appendix 1). There was significant variation
60 β1 in resource use among families (one-way ANOVA
of carrion preference index, using arcsine-transformed
40 data, followed by Tukey’s post-hoc test; F3,57 = 7.5,
α1 P < 0.001): species of Hesperiidae and Lycaenidae fed
almost exclusively upon carrion whereas species of
20 Nymphalidae and the single species of Riodinidae sampled
made more equal use of the two substrates (Table 2).
0 Within the Nymphalidae, there was significant variation
No. species Margalef in food preference among sub-families (F3,41 = 6.3,
P = 0.001): species of Satyrinae or Morphinae fed almost
Figure 1. Partitioning of species richness and diversity into separate
components. α1 (black bars) = diversity within microhabitats and food
exclusively on fruit whereas species of Nymphalinae
resources; β1 (white bars) = diversity between microhabitats within food and Charaxinae were more mixed (Table 2). Variation
resources; β2 (shaded bars) = diversity between food resources. within and among families in the utilization of carrion was
not related to species’ associations with gap or shade sites
(Pearson correlation between carrion preference index
two substrates separately, there was no significant and light preference index, using arcsine-transformed
difference between gap and shade sites in the diversity data; r = 0.1, n = 61 species, P = 0.4).
of species on carrion (Table 1; δ = 0.64, P = 0.6) or on
fruit (δ = 0.29, P = 0.7).
In terms of species composition, the greatest similarity Sex ratios and recapture rates
was between species assemblages on carrion in gap and
shade (Appendix 1; Sørenson’s coefficient; CN = 0.62) In general, there was a strong male bias among species
and on fruit in gap and shade (CN = 0.63). In contrast, sampled on carrion (mean proportion of males = 0.91,
there was very little similarity in assemblages on different n = 44 species, SD = 0.23) but not on fruit, where there
food resources in gaps (CN = 0.18) or shade (CN = 0.22). were approximately equal numbers of each sex (mean
Additive partitioning of diversity data showed that β proportion of males = 0.56, n = 38 species, SD = 0.30).
diversity between food resources comprised 39% of total Moreover for those 21 species sampled on both food types,
species richness and 33% of total species diversity. β there was a significantly greater male bias on carrion
diversity between microhabitats comprised 30% of species
richness and 21% of species diversity on carrion, but only
Table 2. Carrion preference index of species in different families of
17% of species richness and 6.5% of species diversity on butterfly and different subfamilies of Nymphalidae. Means followed by
fruit (Figure 1). different letters are significantly different at the 5% level (families and
subfamilies tested separately). N is the number of species in each case.
Family Mean ± SE N
Food preferences of individual species Nymphalidae 0.37 ± 0.055a 45
Riodinidae 0.57 ab 1
Excluding species where only a single individual was Lycaenidae 0.99 ± 0.003b 9
Hesperiidae 0.88 ± 0.127b 6
trapped, we sampled 1071 individuals from 61 species, of Subfamily of Nymphalidae
which 624 individuals from 38 species were sampled on Satyrinae 0.15 ± 0.085a 16
fruit and 447 individuals from 44 species were on carrion Morphinae 0.06 ± 0.058ab 3
(Appendix 1). Twenty-one species (25% of the total) were Nymphalinae 0.48 ± 0.104bc 18
recorded on both substrates but there was usually a very Charaxinae 0.70 ± 0.127c 8
Carrion-feeding butterflies in rain forest 29

(mean proportion of males = 0.81, SD = 0.31) than on

Log [wing area/(thorax volume )]

1.9

2/3
fruit (mean = 0.60, SD = 0.34; paired t-test of arcsine-
transformed data; t20 = 2.46, P = 0.01). However, the 1.8
proportion of males also differed significantly between
families (two-way ANOVA using arcsine-transformed 1.7
data; F3,73 = 7.2, P < 0.001) with a significant interaction 1.6
between factors (F3,73 = 4.8, P < 0.01): the proportion of
males was much higher on carrion than on fruit in the 1.5
Nymphalidae but not in the Hesperiidae or Lycaenidae, 1.4
where nearly all individuals sampled on both food-types
were male. 1.3
There was a marked difference between recapture rates
1.2
on fruit and carrion (47.4% and 4.8% of individuals
marked respectively; χ12 = 65.0, P < 0.0001) and for 1.1
those 21 species recorded on both food types, the pro- 1.0 1.1 1.2 1.3 1.4 1.5 1.6
portion of recaptures on fruit (mean = 0.22, SD = 0.16) Log body length (mm)
was significantly higher than on carrion (mean = 0.05,
SD = 0.09; paired t-test; t20 = 3.6, P < 0.01). Figure 2. Wing area: thoracic volume ratio of carrion- and fruit-feeders
in relation to body length. Solid triangles, Charaxinae on carrion; open
triangles, Charaxinae on fruit; sold circles, Nymphalinae on carrion;
open circles, Nymphalinae on fruit. Carrion-feeders have significantly
Flight morphology of carrion- and fruit-feeders lower wing area: thoracic volume ratios than fruit-feeders within the
Charaxinae but not within the Nymphalinae.
Very few females were sampled on carrion, very few
species of Hesperiidae or Lycaenidae occurred on fruit and
within the Nymphalidae, very few species of Satyrinae or In this study, we recorded greater species richness and
Morphinae occurred on carrion (see above). Hence the diversity of butterflies on carrion than on fruit (Table 1),
following analysis of flight morphology is restricted to which suggests that carrion is an important food source
males of the subfamilies Nymphalinae and Charaxinae. for a wide variety of species. During the study, we
For this analysis, we characterized each species as caught a total of 1095 individuals from 85 species, which
primarily a carrion- or fruit-feeder, depending on which represents 70% of the total number of species, across all
bait attracted the most individuals. Three species (L. dirtea, families of butterfly, recorded by Willott et al. (2000) at
A. decora and E. monina) were sampled in equal numbers the same site over a 9-mo sampling period. There are
on each bait (Appendix 1) and so were excluded from the no previous data for carrion-feeders on Borneo but the
analysis. Separate analysis of each subfamily indicated number of species of fruit-feeders in this study (45) was
that species of Charaxinae associated with carrion had >80% of the total recorded at the study site over a 12-mo
significantly larger thoraces and lower WA:TV ratios period by Hamer et al. (2003). We are therefore confident
than those associated with fruit (ANCOVA; F1,5 = 23.7, that species feeding on carrion and fruit were adequately
P < 0.01 and F1,5 = 66.9, P < 0.001 respectively; sampled in this study.
Figure 2), indicating more rapid, agile flight. However In addition to supporting more species, carrion also
there was no difference between carrion-feeders and fruit- supported a greater range of species than fruit: species
feeders among species of Nymphalinae (thoracic volume, recorded on carrion were from 42 genera with 2.0
F1,12 = 0.6, P < 0.5; WA:TV ratio, F1,12 = 3.5, P < 0.09; species per genus on average, whereas species sampled
Figure 2). on fruit were from only 27 genera with 2.4 species per
genus on average (calculated from data in Appendix 1).
This pattern was repeated at family level: species in
DISCUSSION four families were sampled more frequently on carrion
than on fruit but, with the exception of a single
The importance of nectar and fruit as food sources species of Hesperiidae (Ancistroides nigrita), only species
for tropical butterflies is widely recognized, but the of Nymphalidae were ever sampled more frequently on
importance of carrion has often been overlooked in fruit than on carrion. These data indicate that carrion
ecological studies (DeVries et al. 1999, Fermon et al. 2005, was utilized by a taxonomically much more diverse range
Hamer et al. 2003). This omission is probably due in part to of species than was fruit. Although species richness was
the relative scarcity of carrion compared with other food higher on carrion in our study, the number of individuals
sources, and in part to the reluctance of field biologists to sampled was higher on fruit (Table 1), and the average
work with such malodorous baits (Hall & Willmott 2000). number of individuals per species was about twice as
30 K. C. HAMER ET AL.

high on fruit as on carrion (14.0 and 7.6 respectively; separation of mud-puddling from carrion-feeding species
calculated from data in Table 1). Given the roughly even therefore appears to be based on nutrient requirements,
sex ratio on fruit and very strong male bias on carrion, not simply microhabitat. In addition, there are reports in
these data indicate that the number of males per species the literature of females mud-puddling and some species
was broadly similar on the two substrates, suggesting that of Lepidoptera in which only females puddle (Boggs &
there was probably no marked difference in population Jackson 1991), whereas carrion-feeders appear to be
sizes of fruit-feeders and carrion-feeders. almost exclusively male (Hall & Willmott 2000, this
There was little similarity in species assemblages on study). We therefore suggest that carrion-feeders and
the two substrates (Appendix 1) and β-diversity between mud-puddlers should be regarded as separate feeding
food-types comprised 33–39% of the total species richness guilds. Species that feed on other substrates such as
and diversity of butterflies feeding on decaying matter. faeces (e.g. Zeuxidia spp.) or sweat (e.g. Halpe spp.) should
Moreover the majority of species had a very clear probably be placed in one or other of the above guilds
preference for one substrate or other (see Results). These depending on the particular nutrients that they obtain in
data strongly suggest that there were distinct butterfly this way.
faunas feeding on carrion and fruit, with very few We found that species of Charaxinae feeding on carrion
species making extensive use of both food types. The had significantly larger thoraces and lower WA:TV
distinction between carrion- and fruit-feeding was not ratios than species feeding on fruit (Figure 2). Thus
simply a reflection of microhabitat requirements because our study provides some support for the hypothesis that
β-diversity between gap and shade contributed much less carrion is exploited by species with rapid powerful flight
to overall species diversity than did β-diversity between and high metabolic rates during flight activity (Hall
substrates (Figure 1) and there was no relationship & Willmott 2000). Within the Nymphalidae, carrion-
between species’ food preferences and associations with feeders were virtually absent from subfamilies Satyrinae
light. It is nonetheless possible that species recorded on and Morphinae (Table 2) and this provides further
carrion at ground-level could feed on fruit in the canopy, circumstantial support for the above hypothesis, because
although this seems unlikely: for instance, of 54 species species in these subfamilies generally have slower, less
of Nymphalidae sampled by Hill et al. (2001) using fruit- agile flight than species of Nymphalinae and Charaxinae
baited traps at the same study site, only four species (7%) (Hamer et al. 2003). The observation that virtually all
were sampled exclusively in the canopy and only one individuals feeding on carrion are males might also
of those (Charaxes solon) was sampled on carrion in this be explained by this hypothesis, because males have
study. greater flight activity than females (Boggs & Jackson
In addition to feeding on carrion, some tropical 1991, Downes 1973). Multi-species comparisons, such
butterflies feed from puddles, damp ground and other as those used in this study, can be confounded by
moist surfaces, a behaviour conventionally termed ‘mud- non-independence of data from closely related species
puddling’ (Adler & Pearson 1982, Beck et al. 1999). (Harvey & Pagel 1991). When phylogenies are well
Several authors have argued that carrion-feeding and known this problem can be avoided by analysis using
mud-puddling Lepidoptera represent a single feeding independent contrasts, but this was not possible in
guild, on the grounds that both substrates provide this study because the phylogeny of South-East Asian
adults with essential minerals (principally sodium) and butterflies is poorly resolved beyond the level of subfamily
that observed differences in the species present on (Corbet & Pendlebury 1992, Parsons 1999). The 23
carrion and at mud-puddling sites are due mainly to species of Nymphalinae and Charaxinae included in
microhabitat preferences (carrion-feeders in dense forest, the morphometric analysis in this paper were from
mud-puddlers at forest edges and other open sites) rather 19 genera and neither carrion-feeders nor fruit-feeders
than nutritional requirements (Boggs & Jackson 1991, were heavily weighted towards any particular genus
Downes 1973, Hall & Willmott 2000). However, we found (Appendix 1). Thus we are confident that our analyses
no relationship between light preference and substrate were not greatly confounded by pseudoreplication of
choice, and no evidence that carrion-feeders were data from closely related species, and that phylogenetic
associated with dense forest rather than gaps. Moreover analysis qualitatively would not alter our conclu-
in an experimental study of bait choice, species of sions.
Pieridae and Papilionidae (which we frequently observed Previous studies have suggested that individuals
mud-puddling at the study site but never recorded feeding on carrion are mainly young adults (Adler &
on carrion) were attracted mainly to sodium sources, Pearson 1982, Boggs & Jackson 1991) and in this study,
which they located using visual cues, whereas species there was a very low recapture rate on carrion but not
of Nymphalidae, Hesperiidae and some Lycaenidae were on fruit. Infrequent recaptures could result from greater
attracted more to protein sources, which they appeared dispersal among fast, powerful fliers (Hill et al. 2001), and
to locate using olfactory cues (Beck et al. 1999). The carrion-feeders might be expected to be more dispersive
Carrion-feeding butterflies in rain forest 31

if carrion is a scarcer and more ephemeral resource than LITERATURE CITED

fruit. However we recorded much lower recapture rates on
carrion than on fruit within species. Our data thus support ADLER, P. H. 1982. Soil and puddle-visiting habits of moths. Journal of
the notion that adults feed frequently and repeatedly on the Lepidopterists’ Society 36:161–173.
fruit but much less frequently on carrion after emergence. ADLER, P. H. & PEARSON, D. L. 1982. Why do male butterflies visit
This pattern is not compatible with the notion that adults mud puddles? Canadian Journal of Zoology 60:322–325.
use carrion as a source of nutrients to fuel continued rapid AUSTIN, G. T. & RILEY, T. J. 1995. Portable bait traps for the study of
flight (Hall & Willmott 2000). An alternative possibility butterflies. Tropical Lepidoptera 6:5–9.
is that differences among species in the exploitation BECK, J., MÜHLENBERG, E. & FIEDLER, K. 1999. Mud-puddling
of carrion reflect differences in the allocation of larval behaviour in tropical butterflies: in search of proteins or minerals?
food reserves during metamorphosis (Karlsson 1995). Oecologia 119:140–148.
During copulation, males transfer nitrogen to females via BLÜTHGEN, N., GEBAUER, G. & FIEDLER, K. 2003. Disentangling a
the spermatophore (Boggs & Gilbert 1979), and access rainforest food web using stable isotopes: dietary diversity in a species-
to nitrogen increases both male mating success and rich ant community. Oecologia 137:426–435.
female reproductive output (reviewed by Boggs 1999). BOGGS, C. L. 1999. A general model of the role of male-donated nutrients
The reason why males with large thoraces feed on in female insects’ reproduction. American Naturalist 136:598–617.
carrion may therefore be because males that allocate BOGGS, C. L. & GILBERT, L. E. 1979. Male contribution to egg production
a high proportion of larval protein resources to flight in butterflies: evidence for transfer of nutrients at mating. Science
muscle mass have insufficient remaining resources for 206:83–84.
spermatophore production and so need access to a BOGGS, C. L. & JACKSON, L. A. 1991. Mud puddling by butterflies is not
source of protein as adults, which they obtain from a simple matter. Ecological Entomology 16:123–127.
carrion. CHAI, P. & SRYGLEY, R. B. 1990. Predation and the flight, morphology,
We found no difference in the flight morphology of and temperature of neotropical rain-forest butterflies. American
carrion- and fruit-feeders within the Nymphalinae, which Naturalist 135:748–765.
suggests that interspecific variation in substrate choice CORBET, A. S. & PENDLEBURY, H. M. 1992. The butterflies of the Malay
within this subfamily is not related to flight muscle Peninsula. (Fourth edition). Malayan Nature Society, Kuala Lumpur.
mass or energy expenditure during flight. Further work 595 pp.
is now needed to compare carrion- and fruit-feeding CRIST, T. O., VEECH, J. A., GERING, J. C. & SUMMERVILLE, K. S.
species of Nymphalinae in terms of mating frequency 2003. Partitioning species diversity across landscapes and regions:
(Sculley & Boggs 1996), to test alternative hypotheses for a hierarchical analysis of α, β, and γ diversity. American Naturalist
carrion-feeding such as a relationship with the nutrient 162:734–743.
and mineral content of larval host-plants (Beck et al. DAVIS, A. J., HOLLOWAY, J. D., HUIJBREGTS, H., KRIKKEN, J., KIRK-
1999, Boggs & Jackson 1991), and to compare the SPRIGGS, A. H. & SUTTON, S. L. 2001. Dung beetles as indicators of
flight morphology of carrion-feeders with that of species change in the forests of northern Borneo. Journal of Applied Ecology
exploiting other resources such as nectar. 38:593–616.
DEVRIES, P. J. 1987. The butterflies of Costa Rica and their natural history.
Vol. 1. Papilionidae, Pieridae and Nymphalidae. Princeton University
Press, Princeton. 327 pp.
ACKNOWLEDGEMENTS DEVRIES, P. J. & WALLA, T. R. 2001. Species diversity and community
structure in neotropical fruit-feeding butterflies. Biological Journal of
We thank all the staff and scientists at Danum Valley the Linnean Society 74:1–15.
Field Centre, especially Glen Reynolds, for logistic support DEVRIES, P. J., WALLA, T. R. & GREENEY, H. F. 1999. Species diversity
and advice. We thank Yayasan Sabah (Forestry Division), in spatial and temporal dimensions of fruit-feeding butterflies from
the Danum Valley Management Committee, the State two Ecuadorian rainforests. Biological Journal of the Linnean Society
Secretary (Internal Affairs and Research Office), Sabah 68:333–353.
Chief Minister’s Department, and the Economic Planning DOWNES, J. A. 1973. Lepidoptera feeding at puddle-margins, dung and
Unit of the Prime Minister’s Department, Kuala Lumpur carrion. Journal of the Lepidopterists’ Society 27:89–99.
for permission to conduct research at Danum Valley. This DUDLEY, R. 1990. Biomechanics of flight in neotropical butterflies:
study is part of projects DV157 and DV158 of the Danum morphometrics and kinematics. Journal of Experimental Biology
Valley Rainforest Research and Training Programme 150:37–53.
and was supported by the British Government’s Darwin DUDLEY, R. & SRYGLEY, R. B. 1994. Flight physiology of neotropical
Initiative (DEFRA). This paper is based on research carried butterflies: allometry of airspeeds during natural free flight. Journal of
out whilst the authors were participants in the Royal Experimental Biology 191:125–139.
Society’s South East Asia Rainforest Research Programme ERWIN, T. L. 1982. Tropical forests: their richness in Coleoptera and
(programme publication number A/405). other arthropod species. Coleopterists Bulletin 36:74–75.
32 K. C. HAMER ET AL.

FERMON, H., WALTERT, M., VANE-WRIGHT, R. A. & MÜHLENBERG, MEGENS, H. J., DE JONG, R. & FIEDLER, K. 2005. Phylogenetic patterns
M. 2005. Forest use and vertical stratification in fruit-feeding but- in larval host plant and ant association of Indo-Australian Arhopalini
terflies of Sulawesi, Indonesia: impacts for conservation. Biodiversity butterflies (Lycaenidae: Theclinae). Biological Journal of the Linnean
and Conservation 14:333–350. Society 84:225–241.
FIEDLER, K. 1998. Diet breadth and host plant diversity of tropical- vs. NOVOTNY, V., BASSETT, Y., MILLER, S. E., WEIBLEN, G. D., BREMER,
temperate-zone herbivores: South-east Asian and west Palaearctic B., CIZEK, L. & DROZD, P. 2002. Low host specificity of herbivorous
butterflies as a case study. Ecological Entomology 23:285–297. insects in tropical forest. Nature 416:841–844.
GERING, J. C., CRIST, T. O. & VEECH, J. A. 2003. Additive partitioning OTSUKA, K. 1988. Butterflies of Borneo. Vol 1. Tobishima Corporation,
of species diversity across multiple spatial scales: implications for Tokyo. 61 pp.
regional conservation of biodiversity. Conservation Biology 17:488– OTSUKA, K. 1991. Butterflies of Borneo Vol 2. Tobishima Corporation,
499. Tokyo. 195 pp.
HALL, J. P. & WILLMOTT, K. R. 2000. Patterns of feeding behaviour in PARSONS, M. 1999. The butterflies of Papua New Guinea. Academic Press,
adult male riodinid butterflies and their relationship to morphology London. 736 pp.
and ecology. Biological Journal of the Linnean Society 69:1–23. PIVNICK, K. A. & MCNEIL, J. N. 1987. Puddling in butterflies:
HAMER, K. C., HILL, J. K., BENEDICK, S., MUSTAFFA, N., SHERRATT, sodium affects reproductive success in Thymelicus lineola. Physiological
T. N., MARYATI, M. & CHEY, V. K. 2003. Ecology of butterflies Entomology 12:461–472.
in natural and selectively-logged forests of northern Borneo: the RIBERA, I., DOLÉDEC, S., DOWNIE, I. S. & FOSTER, G. N. 1998. Effect
importance of habitat heterogeneity. Journal of Applied Ecology of land disturbance and stress on species traits of ground beetle
40:150–162. assemblages. Ecology 82:1112–1129.
HAMER, K. C., HILL, J. K., MUSTAFFA, N., BENEDICK, S., SHERRATT, SCHULZE, C. H., LINSENMAIR, K. E. & FIEDLER, K. 2001. Understorey
T. N., CHEY, V. K. & MARYATI, M. 2005. Temporal variation versus canopy: patterns of vertical stratification and diversity among
in abundance and diversity of butterflies in Bornean rain forests: Lepidoptera in a Bornean rain forest. Plant Ecology 153:133–152.
opposite impacts of logging recorded in different seasons. Journal of SCOBLE, M. J. 1992. The Lepidoptera: form, function and diversity. Oxford
Tropical Ecology 21:1–9. University Press, Oxford. 404 pp.
HARVEY, P. H. & PAGEL, M. D. 1991. The comparative method in SCULLEY, C. E. & BOGGS, C. L. 1996. Mating systems and sexual division
evolutionary biology. Oxford University Press, Oxford. 239 pp. of foraging effort affect puddling behaviour by butterflies. Ecological
HILL, J. K., HAMER, K. C., TANGAH, J. & DAWOOD, M. 2001. Ecology Entomology 21:193–197.
of tropical butterflies in rainforest gaps. Oecologia 128:294–302. SOKAL, R. R. & ROHLF, F. J. 1995. Biometry. (Third Edition). WH
KARLSSON, B. 1995. Resource allocation and mating systems in Freeman and Company, New York. 887 pp.
butterflies. Evolution 49:955–961. SOLOW, A. R. 1993. A simple test for change in community structure.
LANDE, R. 1996. Statistics and partitioning of species diversity, and Journal of Animal Ecology 62:191–193.
similarity among multiple communities. Oikos 76:5–13. SRYGLEY, R. B. & CHAI, P. 1990. Flight morphology of Neotropical
LEMMON, P. E. 1957. A new instrument for measuring forest overstory butterflies: palatability and distribution of mass to the thorax and
density. Journal of Forestry 55:667–669. abdomen. Oecologia 84:491–499.
MAGURRAN, A. E. 1988. Ecological diversity and its measurement. Croom SUTTON, S. L. & COLLINS, N. M. 1991. Insects and tropical forest
Helm, London, 160 pp. conservation. Pp.405–422 in Collins, N. M. & Thomas, J. A. (eds).
MARDEN, J. H. 1987. Maximum lift production during takeoff in flying The conservation of insects and their habitats. Academic Press, London.
animals. Journal of Experimental Biology 130:235–258. VEECH, J. A., SUMMERVILLE, K. S., CRIST, T. O. & GERING, J. C. 2002.
MARSH, C. W. & GREER, A. G. 1992. Forest land-use in Sabah, Malaysia: The additive partitioning of diversity: recent revival of an old idea.
an introduction to Danum Valley. Philosophical Transactions of the Oikos 99:3–9.
Royal Society London Series B 335:331–339. WILLOTT, S. J., LIM, D. C., COMPTON, S. G. & SUTTON, S. L. 2000.
MAY, P. G. 1992. Flower selection and the dynamics of lipid reserves in Effects of selective logging on the butterflies of a Bornean rainforest.
two nectivorous butterflies. Ecology 73:2181–2191. Conservation Biology 14:1055–1065.
Carrion-feeding butterflies in rain forest 33

Appendix 1. Butterfly species recorded on carrion and fruit at gap and shade sites. Data show the numbers of individuals in each case.
Carrion Fruit Carrion Fruit

Nymphalidae Gap Shade Gap Shade Total Nymphalidae Gap Shade Gap Shade Total
Satyrinae
1. Melanitis leda 0 0 2 3 5 52. Euthalia iapis 0 0 2 0 2
2. Elymnias esaca 2 0 0 0 2 51. E. monina 1 0 1 0 2
3. E. dara 1 0 0 0 1 42. Dophla evelina 0 0 15 14 29
4. Neorina lowii 1 1 33 58 93 43. Bassarona teuta 0 0 13 3 16
5. Lethe chandica 0 2 0 0 2 44. B. dunya 1 6 32 53 92
6. Mycalesis anapita 0 0 9 4 13 45. Lexias dirtea 1 0 0 1 2
7. M. patiana 0 0 4 4 8 46. Amnosia decora 0 1 0 1 2
8. M. fusca 0 0 0 1 1 47. Dichorragia nesimachus 0 0 2 3 5
9. M. kina 0 0 1 13 14 Charaxinae
10. M. janardana 0 0 1 1 2 53. Prothoe franck 0 9 17 24 50
11. M. dohertyi 0 0 5 15 20 54. Agatasa calydonia 0 2 3 1 6
12. M. mineus 0 0 1 1 2 55. Polyura athamas 2 0 0 0 2
13. M. horsfieldi 0 0 7 5 12 56. P. jalysus 2 1 0 0 3
14. M. orseis 4 4 19 25 52 57. P. schreiber 1 3 0 0 4
15. M. maianeas 0 0 5 27 32 58. Charaxes solon 1 1 0 0 2
16. M. oroatis 6 17 19 75 117 59. C. bernardus 17 18 9 1 45
17. Erites elegans 0 0 2 1 3 60. C. durnfordi 2 0 2 2 6
18. Ragadia makuta 2 0 16 16 34 Riodinidae
Morphinae 61. Zemeros emesoides 2 2 1 2 7
19. Faunis canens 0 0 0 1 1 Lycaenidae
20. F. stomphax 0 0 0 1 1 62. Curetis santana 2 1 0 0 3
21. Xanthotaenia busiris 0 1 0 0 1 63. C. tagalica 54 38 4 0 96
22. Amathusia phidippus 0 1 0 0 1 64. C. regula 6 2 0 0 8
23. Amathuxidia amythaon 0 0 0 1 1 65. Caleta elna 1 1 0 0 2
24. Zeuxidia aurelius 0 0 1 1 2 66. Megisba malaya 1 1 0 0 2
25. Z. amethystus 0 0 1 1 2 67. Nacaduba beroe 2 3 0 0 5
26. Z. doubledayi 0 0 0 1 1 68. N. kurava 1 0 0 0 1
27. Discophora necho 0 1 4 1 6 69. N. sanaya 0 1 0 0 1
Nymphalinae 70. N. hermus 0 1 0 0 1
28. Ariadne isaeus 1 0 0 0 1 71. N. angusta 1 0 0 0 1
30. Cupha erymanthis 4 2 0 0 6 72. Prosotas nora 0 1 0 0 1
29. Cirrochroa emalea 65 35 0 3 103 73. P. aluta 4 2 0 0 6
31. Vindula dejone 2 1 0 0 3 74. Cheritra freja 39 15 1 0 55
32. Paduca fasciata 15 4 8 1 28 75. Drupadia ravindra 1 0 0 0 1
33. Terinos clarissa 0 1 0 0 1 76. Remelana jangala 2 0 0 0 2
34. Kallima limborgi 0 2 2 10 14 Hesperiidae
35. Rhinopalpa polynice 1 1 6 0 8 77. Burara oedipodea 1 0 0 0 1
36. Chersonesia rahria 2 3 0 0 5 78. B. etelka 4 3 0 0 7
37. Athyma pravara 1 2 0 0 3 79. Hasora schoenherr 0 3 0 0 3
38. A. kanwa 1 0 0 0 1 80. Halpe zema 2 0 0 0 2
39. A. adunora 0 1 0 0 1 81. H. ormenes 4 0 0 0 4
40. A. nefte 1 0 0 0 1 82. H. pelethronix 1 0 0 0 2
41. Moduza procris 2 0 0 0 2 83. H. sikkima 1 0 0 0 1
49. Tanaecia pelea 0 0 1 0 1 84. Ancistroides nigrita 1 0 1 2 4
50. T. munda 0 0 1 0 1 85. Pithauria marsena 0 1 0 0 1
48. T. aruna 0 0 2 1 3