The molecular physiology of increased egg
desiccation resistance during diapause in
the invasive mosquito, Aedes albopictus
Jennifer M. Urbanski1, Joshua B. Benoit2, M. Robert Michaud2,
David L. Denlinger2and Peter Armbruster1,*
1Department of Biology, Georgetown University, 37th and O Sts. NW , Washington, DC 20057, USA
2Department of Entomology, Ohio State University, 318 W 12th Avenue, Columbus, OH 43210, USA
Photoperiodic diapause is a crucial adaptation to seasonal environmental variation in a wide range of
arthropods, but relatively little is known regarding the molecular basis of this important trait. In temper-
ate populations of the mosquito Aedes albopictus, exposure to short-day (SD) lengths causes the female to
produce diapause eggs. Tropical populations do not undergo a photoperiodic diapause. We identified a
fatty acyl coA elongase transcript that is more abundant under SD versus long-day (LD) photoperiods
in mature oocyte tissue of replicate temperate, but not tropical, A. albopictus populations. Fatty acyl
CoA elongases are involved in the synthesis of long chain fatty acids (hydrocarbon precursors). Diapause
eggs from a temperate population had one-third more surface hydrocarbons and one-half the water loss
rates of non-diapause eggs. Eggs from a tropical population reared under SD and LD photoperiods did
not differ in surface hydrocarbon abundance or water loss rates. In both a temperate and tropical popu-
lation, composition of hydrocarbon chain lengths did not differ between eggs from SD versus LD
conditions. These results implicate the expression of fatty acyl coA elongase and changes in quantity,
but not composition, of egg surface hydrocarbons as important components of increased desiccation
resistance during diapause in A. albopictus.
Keywords: invasive species; photoperiodic diapause; desiccation resistance; Aedes albopictus
The Asian tiger mosquito, Aedes albopictus, is currently
the most invasive mosquito species in the world (Benedict
et al. 2007). Also of considerable public health concern,
this aggressive daytime biting mosquito is capable of effi-
ciently transmitting Chikungunya, dengue, West Nile and
a variety of native North American arboviruses (Turell
et al. 2001; Gratz 2004). In the last 30 years, A. albopictus
has rapidly spread from its native Asian range across the
world and is currently found in at least 28 countries on
every continent except Australia and Antarctica (Benedict
et al. 2007). This rapid spread has been accomplished by
the worldwide transport of containers such as tyres
and pots of ‘lucky bamboo’ that harbour eggs and/or
larvae (Hawley et al. 1987; Scholte et al. 2008). Recent
analyses indicate that there are few countries where
A. albopictus could not exist and thus further spread
and accompanying public health concern is likely
(Benedict et al. 2007).
The first breeding population of A. albopictus in the US
was discovered in Houston, TX in 1985 (Sprenger &
Wuithiranyagool 1986), where it was probably introduced
via a shipment of used automobile tyres from temperate
Japan (Hawley et al. 1987). Within two years A. albopictus
had spread rapidly throughout the US, extending as far
north as Illinois and as far east and south as Jacksonville,
FL (Moore 1999). This rapid range expansion was
probably facilitated by the intact photoperiodic diapause
response of the invading population (Hawley et al.
1987). In its native Asian range, A. albopictus occurs
across an unusually broad latitudinal range, including
temperate populations that undergo a photoperiodic
diapause and tropical populations that do not enter a
photoperiodic diapause (Hawley 1988). In temperate
(diapausing) populations, exposure of pupal and adult
females to short-day (SD) lengths induces a developmen-
tal arrest of pharate larvae inside the chorion of the egg
(Wang 1966; Mori et al. 1981). It has been known for
some time that diapause eggs of A. albopictus have
increased survivorship under desiccating and cold-stress
conditions relative to non-diapause eggs (Sota & Mogi
1992; Hanson & Craig 1994). However, the mechanistic
basis of this stress resistance has not previously been
Juliano & Lounibos (2005) showed that mosquitoes
that produce desiccation-resistant eggs were more likely
to become established in non-native habitats relative to
mosquito species that produce desiccation-susceptible
eggs, presumably because desiccation-resistant eggs are
more likely to survive long-distance transport. Desicca-
tion resistance in mosquitoes has primarily been studied
by examining survival under a range of relative humidity
conditions (Sota& Mogi 1992; Gray & Bradley
2005). However, in a more detailed mechanistic study,
Benoit & Denlinger (2007) showed that diapausing
adult females of Culex pipiens had substantially lower
* Author for correspondence (email@example.com).
We dedicate this paper to the memory of our co-author and
colleague, Dr Rob Michaud, who sadly died on January 5, 2010.
Electronic supplementary material is available at http://dx.doi.org/10.
1098/rspb.2010.0362 or via http://rspb.royalsocietypublishing.org.
Proc. R. Soc. B (2010) 277, 2683–2692
Published online 21 April 2010
Received 23 February 2010
Accepted 31 March 2010
This journal is q 2010 The Royal Society
water loss rates relative to non-diapausing females
owing to larger body size, decreased metabolism and an
approximately two-fold higher accumulation of cuticular
hydrocarbons. The molecular basis of increased desicca-
tion resistance in diapausing C. pipiens has yet to be
Herein we describe a fatty acyl coA elongase from
A. albopictus that is upregulated under SD versus long-
day (LD) photoperiods in mature (stage V) oocytes of
temperate, but not tropical, populations. Fatty acyl coA
elongases are involved in the formation of very long
chain lipids which are known to affect water loss in insects
(Blomquist et al. 1987). Using a comparative approach
and based on well-established functional considerations,
we link this expression pattern to an increase in hydro-
carbon quantity and a decrease in water loss rate in
diapause relative to non-diapause eggs. These results pro-
vide insight into the physiological processes contributing
to the rapid global spread of this invasive mosquito.
2. MATERIAL AND METHODS
(a) EST identification
We identified a 186bp putative fatty acyl coA elongase
expressed sequence tag (EST) as potentially upregulated
under SD versus LD photoperiod treatments using a ‘SD
minus LD’ suppressive subtractive hybridization (SSH)
library of cDNAs isolated from mature (stage V) oocyte
tissue. The SSH library was constructed using a laboratory
F12 population from Berlin, NJ which was collected and
reared as described below. A comprehensive description of
the suppressive subtractive hybridization results is described
in another paper (Urbanski et al. in press).
(b) 50and 30RACE
To determine the complete cDNA sequence of the putative
SMART RACE cDNA Amplification kit (BD Biosciences,
CC-30) were used to PCR-amplify the 50and 30RACE pro-
ducts which were gel-purified using the Nucleo Trap
Nucleic Acid Purification kit (Clontech, Mountainview,
CA). The PCR products were then cloned using the
TOPO TA Cloning kit with pCR2.1-TOPO Vector and
One Shot Top10F0chemically competent cells (Invitrogen,
Carlsbad, CA). Plasmids were purified using the WizardPlus
Miniprep DNA purification system (Promega, Madison, WI)
and sequenced on an ABI 3100 Genetic Analyser using Big
Dye chemistry and M13 primers modified for sequencing
reactions were required to determine the sequence of
the 30end of the cDNA. Primers Fatty3F (50-CGACTGC
AACTACCCGAAAGCC-30) and Fatty4F (50-CTCCTTC
ACGCATACAAAGCGCAG-30) were employed for these
final steps. Contig assembly was performed with SEQUENCHER
v. 4.5 (Gene Codes Corporation, Ann Arbor, MI). The
inferred amino acid sequence was deduced by using ORF
FINDER (http://www.ncbi.nlm.nih.gov/projects/gorf/) and was
used in a protein Blast search to identify putative orthologues.
(c) Quantitative reverse transcriptase-PCR
Five laboratory colonies of A. albopictus were established for
qRT-PCR experiments. For three temperate populations
(Burlington, NJ (NJ1); Salem, NJ (NJ2); Manassas, VA
(VA)), we collected at least 300 individual larvae and pupae
from at least 15 tyres within a site. The mosquitoes were
transferred to the laboratory where they were reared for at
least three generations under near-optimal conditions at
218C and approximately 80 per cent RH as described in
Armbruster & Conn (2006). We also established laboratory
colonies for two tropical populations (Honolulu, HI (HI);
Kuala Lumpur, Malaysia (KL)) using at least 1,000 eggs col-
lected from oviposition traps and kindly provided by Mr
Pingjun Yan, Department of Health, Honolulu, HI, and Dr
Indra Vythilingam, Institute of Medical Research, Kuala
Lumpur, Malaysia, respectively. A. albopictus is native to
Kuala Lumpur and is thought to have invaded Hawaii
during the 1890s (Joyce 1961), most probably from a
location in the Indian Ocean (Mousson et al. 2005). Tropical
populations, which do not undergo a photoperiodic dia-
pause, were included to confirm that expression differences
in response to LD versus SD conditions were directly attribu-
table to diapause and not to ancillary effects of differential
photoperiod. The fact that we have used two geographically
disparate tropical populations that do not undergo photo-
independently over the last at least 100 years represents a
conservative test for parallel underlying patterns of gene
We reared two replicate cages (i.e. biological replicates)
for each of the three temperate populations and each of the
two tropical populations under both SD and LD photo-
periods. For each biological replicate, approximately 300
male and female pupae from each temperate population
(F3–F6laboratory generation) and from each tropical popu-
lation (F6–F10laboratory generation) were divided with half
exposed to SD photoperiods and half exposed to LD photo-
periods. Ten to 20-day-old adult females were blood fed to
repletion on a human host, and four days post-bloodmeal
females were collected four hours after ‘lights on’ and
stored at 2808C. Mature (stage V) oocytes, identified by
the appearance of a visible exochorion surface pattern, were
dissected directly into RNAlater (Ambion, Beverly, MA).
Although we included females from a 10-day range of
chronological age, this variation is unlikely to have a large
effect on the abundance of mature oocyte transcripts since
ovarian development is more strongly influenced by time
since blood meal than chronological age (Clements 1992).
RNA was extracted from mature oocyte tissue using TRI
Reagent (Sigma Aldrich, St Louis, MO) followed by an iso-
performed for each sample by adding 1 ml DNase per 5 mg
total RNA in 1? DNase reaction buffer to a volume of
100 ml. The mixture was incubated at 378C for 10 min, fol-
lowed by the addition of 1 ml of EDTA and a second
incubation at 758C for 10 min. DNase treatment was fol-
precipitation. The resulting RNA was then used to perform
qRT-PCR, comparing fatty acyl coA elongase expression of
SD photoperiod versus LD photoperiod treatments with a
Brilliant II SYBR Green 1-Step qRT-PCR Kit (Stratagene,
La Jolla, CA) on an Mx3000P qPCR machine (Stratagene,
La Jolla, CA). Ribosomal protein L34 (NCBI accession
no. AF144549) was used as an endogenous control in
2684J. M. Urbanski et al.Desiccation resistance during diapause
Proc. R. Soc. B (2010)
all qRT-PCR reactions. The primer sequences are as
follows: Fatty5F (50-CCCCGGACAAAGGATTGGC-30),
(50-GGGCTCGTCTACCACGTTTA-30). Three replicate
reactions (technical replicates) were performed for each
RNA sample. Each reaction consisted of 1 ml of 50 ng ml21
RNA as template and 3 ml each of the forward and reverse
primers at a concentration of 1mM. Each reaction consisted
of a 30 min reverse transcription step at 508C and 10 min at
958C, followed by 45 cycles of 30 s at 958C, 1 min at 638C
and 30 s at 728C. A final cycle consisting of 1 min at 958C,
30 s at 558C, and 30 s at 958C was performed to determine
PCR specificity from the dissociation curve. Cycle threshold
(CT) values were averaged across triplicate reactions and
used to determine fold change differences between SD and
LD treatments for each population replicate using the
2(2DDC(t))method (Livak & Schmittgen 2001).
To confirm that SD photoperiods resulted in the pro-
duction of diapause eggs in temperate but not tropical
populations, a separate cage of mosquitoes was reared in par-
allel under LD or SD conditions as described above for each
biological replicate. Eggs were collected and stimulated to
hatch as previously described (Armbruster & Conn 2006).
Per cent hatch was recorded for each replicate.
(d) Hydrocarbon quantity
For analysis of hydrocarbon quantity the VA and KL popu-
lations were reared under SD and LD conditions as
described above. Because of the large number of eggs
required for these assays, five paired-replicate samples of
eggs (SD and LD) were obtained over the course of three
laboratory generations for both populations. For each popu-
lation, paired samples of SD and LD eggs were handled in
parallel, treated identically, and were 2–3 weeks old at the
time of analysis. However, the paired-replicate samples
from the temperate population were not all collected and
analysed simultaneously with the paired replicate samples
from the tropical population. Lipids distributed on the sur-
face of eggs were extracted by adding approximately 25 mg
of eggs from SD or LD conditions to 1 ml of hexane. The
mixture was gently agitated for 5 min and then filtered
through a glass pre-filter (Millipore). This process was
repeated twice before the filter was examined under a dissect-
ing microscope to verify that no eggs had been broken. The
resulting 2 ml of surface lipid extract was evaporated using
nitrogen gas and stored at 2708C. This extraction procedure
produces a sample that is over 99 per cent hydrocarbon as
(GC/MS, data not shown). Also, a fourth extraction per-
formed on both SD and LD eggs did not produce any
detectable lipids. Surface lipid extracts (hydrocarbons) were
quantified using a vanillin assay as described by Van
Handel (1985). The dried lipid extract was heated to
1008C with 200 ml concentrated sulphuric acid for 10 min,
and then 4.8 ml of vanillin solution (600 mg/500 ml 68%
phosphoric acid) was added. The final solution was mixed
in a vortex for 10 s and allowed to sit for 5 min until the
solution turned red. The optical density of the solution at
525 nm was measured in a spectrophotometer and compared
with a standard curve to determine the total mass of lipid in
the sample. Total lipid mass was then converted to a mg
lipid mg21wet egg mass to normalize differences in sample
(e) Hydrocarbon composition
To determine the chain length composition of hydro-
carbons on the surface of VA and KL eggs, three
replicate batches of eggs from SD and LD conditions
were collected from both populations and surface lipids
were extracted in hexane as described above except that
approximately 10 mg of eggs were used for each replicate
sample were dried under a gentle stream of nitrogen,
re-suspended in 100 ml of high-purity chloroform, and
placed in sampling vials with glass inserts and stored at
2708C until analysed by GC/MS.
For GC/MS analysis the hydrocarbon samples were hand-
injected (2 ml each) into a Thermo-Finnigan Trace GC/MS
instrument with a Restek 30 m fused silica column (I.D.
25 mm, 95% dimethyl siloxane, 5% diphenyl). The injector
temperature was set to 2208C and the oven was programmed
to heat from 1608C to 3008C, increasing 88C min21and
holding the 3008C temperature for an additional 5 min.
50 ml min21. The detector was set to detect mass units
from 50 amu to 450 amu, and peak areas were quantified
using XCALIBUR software that also drove the GC/MS unit
(Thermo, Inc.). This chromatographic programme allowed
sufficient chromatographic resolution to separate all hydro-
carbon peaks for quantification. Hydrocarbon peaks were
identified according to their spectral patterns (identified by
general chromatographic signature as well as comparison
with the NIST and Wiley Chemical libraries) and chain
length was established by comparison with authentic stan-
dards of heptacosane and hexacosane loaded into a
(f) Water loss rates
Three replicates of 15 eggs per replicate were collected on
successive days from VA and KL populations reared under
SD and LD conditions. All eggs were 10–14 days old at
the time of analysis. In insects, exposure to 0 per cent RH
permits the net transpiration rates (integumental plus respir-
atory water loss) to be determined since under these
conditions no water can be gained from the atmosphere
(Wharton 1985). This allows water loss to be measured
according to Wharton (1985):
where m0is the initial water mass, mtis the water mass at any
time t and k represents the rate of water loss (Wharton 1985).
Thus, the slope of a plot of ln (mt/m0) versus time is the net
transpiration rate (water loss rate) expressed as per cent per
Therefore, to measure water loss, the three-paired repli-
cate batches of 15 LD and 15 SD eggs from both the VA
and KL populations were initially weighed using an electro-
balance (CAHN 25, Ventron Co., Cerritos, CA), and then
moved into a sealed glass desiccator maintained at 20–
228C and 0 per cent RH, generated by solid CaSO4and ver-
ified with a hygrometer (Thomas Scientific, Philadelphia,
PA). Eggs were weighed individually every other day without
enclosure and returned to the above conditions within 2 min.
Eggs were dried until the mass was constant for five consecu-
tive days. The amount of water available for exchange (water
mass) was measured as the difference between the initial
mass and the dry mass.
orlnðmt=m0Þ ¼ ?kt;
Desiccation resistance during diapause
J. M. Urbanski et al.
Proc. R. Soc. B (2010)
(g) Statistical analyses
To analyse qRT-PCR gene expression data, we compared
fold change values (Livak & Schmittgen 2001) of temperate
(diapausing) populations to fold change values of tropical
sample Wilcoxon rank sum test. We note that employing a
two-tailed test is a conservative approach, because the tran-
script we tested using qRT-PCR was isolated from a ‘SD
minus LD’ SSH cNDA library, thus providing a priori evi-
dence that the transcript is differentially expressed as a
component of the diapause response. Although a one-tailed
test could therefore be justified in these analyses of gene
expression, we have chosen to report conservative probability
values based on a two-tailed test. Using a Wilcoxon rank sum
test, we also (i) compared the fold changes of the two tropical
populations (KL versus HI) and (ii) tested whether the
fold changes in the temperate populations or the tropical
populations were significantly different from 1 (i.e. no differ-
ential expression in response to SD versus LD photoperiod).
Surface hydrocarbon quantities of paired replicate samples of
SD and LD eggs from each population were compared for
each population using paired t-tests. As noted above, the
paired replicate samples from the temperate population
were not all collected and analysed simultaneously with the
paired replicate samples from the tropical population, so it
is not statistically valid to directly compare hydrocarbon
quantities between the temperate and tropical eggs. To com-
pare the composition of hydrocarbons from eggs produced by
females reared under SD and LD conditions from both tem-
perate and tropical populations, peak areas from GC/MS
output were quantified and converted to proportions and
analysed using NMDS (Non-Metric Multidimensional Scal-
ing), a robust ordination technique (Minchin 1987). We used
the proportion of each hydrocarbon to create a dissimilarity
matrix among the following treatments using the Bray–
Curtis dissimilarity coefficient (Faith et al. 1987): (i) LD
versus SD samples from the temperate population, (ii) LD
versus SD samples from the tropical population, and
(iii) LD and SD samples from the temperate population
versus LD and SD samples from the tropical population.
We then tested for differences between treatments using
ANOSIM (Analysis of Similarity, Warwick et al. 1990) with
1000 permutations. We used analysis of variance (ANOVA)
to test for the effect of population, photoperiod and
rates, all with d.f. ¼ 1,175. The ANOVA was followed by
an a posteriori comparison of treatment
Bonferroni correction to control for experiment-wise error
(p , 0.05).
The full-length A. albopictus fatty acyl coA elongase cDNA
sequence (NCBI accession no. GQ168593) consists of a
372 bp 50UTR, a 1080 bp (359 amino acid) open reading
frame and a 495 bp 30UTR including the poly-A tail.
Amino acid residues 4–259 correspond to the highly con-
served ELO superfamily, which is involved in long chain
fatty acid elongation systems. A protein Blast indicated
96 per cent inferred amino acid identity with a fatty acyl
coA elongase in Aedes aegypti (Ribeiro et al. 2007), 86
per cent identity with AGAP004373-PA in Anopheles
gambiae, 61 per cent identity with elongation of very
long chain fatty acids protein 1 in C. pipiens quinquefascia-
tus and 66 per cent with CG31522 isoform B in
Drosophila melanogaster (see the electronic supplementary
(b) Quantitative reverse transcriptase-PCR
Fatty acyl coA elongase was upregulated under SD photo-
temperate but not tropical populations (figure 1). The
fold change values for temperate populations comparing
SD relative to LD expression ranged from 1.6 to 7 and
differed significantly (Z ¼ 2.1, p ¼ 0.036) from 1 (i.e.
no difference in expression in response to photoperiod).
The fold change values for the tropical populations
ranged from 0.56 to 1.7 and did not differ significantly
from 1 (Z ¼ 1.29, p ¼ 0.20). Fold changes of temperate
(diapausing) populations were significantly different
from fold changes in tropical (non-diapausing) popu-
lations (Wilcoxon signed-rank test, Z ¼ 2.04, p ¼ 0.04).
The fold changes of the two tropical populations were
not significantly different (Z ¼ 1.22, p ¼ 0.22).
(c) Hydrocarbon quantity
In the temperate population, there were 29 per cent more
surface hydrocarbons on diapause versus non-diapause
eggs (figure 2a; diapause mean ¼ 32.29+4.26 mg mg21
wet weight; non-diapause mean ¼ 22.93+4 mg mg21
wet weight; paired t-test, t ¼ 2.9, p ¼ 0.04). In the tropi-
cal population, there was no difference in egg surface
hydrocarbon quantity between LD and SD photoperiod
treatments (figure 2b; SD mean ¼ 48.53+6.1 mg mg21
wetweight;LDmean ¼ 47.5+3.46 mg mg21
weight; paired t-test, t ¼ 20.3, p ¼ 0.78).
(d) Hydrocarbon composition
The surface lipids of A. albopictus eggs featured 17 detect-
able major unsaturated hydrocarbons ranging in odd
numbers from 19 to 51 carbons in length (figure 3).
Minor hydrocarbons were a very small proportion of the
overall signal (less than 2%), and were therefore omitted.
NJ1 NJ2VA HI
Figure 1. Mean (+ s.e.) fold change of fatty acyl coA elongase
in mature oocytes of female A. albopictus exposed to short-
day relative to long-day photoperiods. Fold change values
differ significantly between temperate (NJ1, NJ2, VA) and
tropical (HI, KL) populations (Wilcoxon signed-rank test,
Z ¼ 2.038, p ¼ 0.04).
2686J. M. Urbanski et al.Desiccation resistance during diapause
Proc. R. Soc. B (2010)
Non-metric multidimensional scaling indicated no signifi-
cant difference in hydrocarbon composition between: (i)
diapause versus non-diapause eggs from the temperate
population (ANOSIM R ¼ 20.0741, p ¼ 0.49), (ii)
non-diapause eggs produced under SD versus LD
photoperiods from the tropical population (ANOSIM
R ¼ 0.0370, p ¼ 0.50), and (iii) diapause and non-
diapause eggs from the temperate population versus SD
and LD eggs from the tropical population (ANOSIM
R ¼ 20.0074, p ¼ 0.49).
(e) Water loss rates
The net transpiration (¼ water loss) rates for isolated eggs
were extremely low (figure 4). Diapause eggs from a tem-
perate (VA) population had water loss rates nearly half
those of non-diapause eggs (figure 4; diapause ¼ 1.32+
0.18% d21; non-diapause ¼ 2.25+0.2% d21). Results
of ANOVA indicated that water loss rates were affected
by population (F1,175¼ 147.94, p , 0.001), photoperiod
(F1,175¼ 134.24, p , 0.001) and population-by-photo-
posteriori comparison of mean water loss rates indicated
that the temperate (VA) diapause eggs had significantly
lower water rates (p , 0.05) than non-diapause eggs
from the temperate (VA) population and SD and LD
non-diapausing eggs from the tropical (KL) population,
all of which did not differ significantly (p . 0.05). Both
diapause and non-diapause eggs could tolerate loss of
25–27% of their water content, and thus, owing to their
differences in water loss rates, diapause eggs are able to
survive much longer. These survival rates were similar
to those found by Sota & Mogi (1992). Experiments indi-
cate that water vapour uptake is not used as a source to
replenish egg water stores, and thus eggs probably rely
on contact with free water to increase their water content
(data not shown).
p , 0.001).
The recurring arrival of the harsh winter conditions in
temperate zone habitats represents a fundamental chal-
lenge to the survival and reproduction of a wide variety
of insects. Many insects surmount this challenge by
means of photoperiodic diapause, the ability to assess
day length (photoperiod) as a token cue for initiating sea-
sonally appropriate developmental arrest (Tauber et al.
1986). Photoperiodic diapause thus provides an adaptive
mechanism for the temporal coordination of growth,
development and dormancy in a seasonal environment.
At the same time, it has become increasingly clear that
diapause does not represent a simple physiological shut-
down, but rather is a physiologically dynamic state with
unique patterns of gene expression at specific points
along the trajectory from initiation to maintenance and
termination of diapause (reviewed by Denlinger 2002).
Processes related to stress tolerance, such as cold and
desiccation resistance, appear to be particularly impor-
tant physiological components of the diapause response
(Yoder & Denlinger 1991a; Benoit & Denlinger 2007;
Rinehart et al. 2007).
The fatty acyl coA elongase we describe was isolated
from a ‘SD minus LD’ SSH library constructed using
mature (stage V) oocyte tissue of a temperate (diapaus-
ing) population. Full details of the SSH library are
described in another paper (Urbanski et al. in press).
We found that fatty acyl coA elongase transcripts were
more abundant in mature oocyte tissue under diapause
inducing SD conditions relative to diapause averting LD
conditions in replicate temperate (diapausing) but not
tropical (non-diapausing) populations (figure 1). Because
fatty acyl coA elongases encode proteins involved in the
synthesis of very long chain surface lipids that are
known to mediate desiccation resistance in a diverse
group of insects (Blomquist et al. 1987; Vaz et al. 1988;
Jua ´rez 1994, 2004; Yoder et al. 1995; Benoit & Denlinger
2007; Jua ´rez & Ferna ´ndez 2007), we hypothesized that
the increased transcript abundance of this gene might
be related to the previously documented increased survi-
val of diapause relative to non-diapause eggs under
desiccating conditions (Sota & Mogi 1992). Consistent
with this hypothesis, we found that in a temperate popu-
lation, diapause eggs had approximately 30 per cent more
surface lipids (more than 99% hydrocarbon) than non-
diapause eggs, but that eggs from a tropical population
reared under SD and LD conditions did not differ in sur-
face lipid quantities (figure 2). We also found that
diapause eggs from a temperate population had approxi-
mately one-half the water loss rate of non-diapause
(µg mg–1 wet weight)
Figure 2. Mean (+ s.e.) egg surface lipid (¼ hydrocarbon) quantities for a (a) temperate (VA) and (b) tropical (KL) population
of A. albopictus. Long-day (open bar) and short-day (filled bar) treatments differ significantly for the temperate (paired t-test,
t ¼ 2.90, p ¼ 0.04) but not tropical population (paired t-test, t ¼ 20.3, p ¼ 0.78).
Desiccation resistance during diapause
J. M. Urbanski et al.
Proc. R. Soc. B (2010)
eggs, but that eggs from tropical females reared under SD
and LD conditions did not differ in water loss rates
The vast majority of studies investigating the molecular
physiology of diapause consider a single population reared
on diapause-averting and diapause-inducing conditions.
Our experimental design leverages the rare opportunity
to compare the molecular physiology of temperate and
tropical populations from within the same species that
do and do not undergo a photoperiodic diapause response
(Hawley et al. 1987; Hawley 1988). This comparative
approach provides a particularly strong basis for establish-
ing that the physiological changes we describe in the
temperate (diapausing) populations are causal com-
ponents of the diapause response, rather than more
general responses to photoperiod per se. While a number
of studies have compared molecular aspects of photo-
periodic diapause between diapausing populations and
non-diapausing mutant or selected strains isolated in
the laboratory (Pavelka et al. 2003; Syrova et al. 2003;
Goto et al. 2006), we know of no other molecular physi-
ology studies that have explicitly compared naturally
occurring diapausing and non-diapausing populations.
Furthermore, because we used qRT-PCR to examine
fatty acyl coA elongase transcript abundance in replicate
temperate and tropical populations (figure 1), our results
2123 252729 3133 35
hydrocarbon chain length
3739 41 43 45474951
Figure 3. Composition of egg surface hydrocarbons from (a) temperate (VA) and (b) tropical (KL) populations of A. albopictus
produced under long-day (open bar) and short-day (filled bar) treatments. Bars represent mean (+ s.e.). Non-metric multi-
dimensional scaling indicated no significant difference in hydrocarbon composition between: (i) diapause versus non-
diapause eggs from the temperate population (ANOSIM R ¼ 20.0741, p ¼ 0.49), (ii) non-diapause eggs produced under LD
versus SD photoperiods from the tropical population (ANOSIM R ¼ 0.0370, p ¼ 0.50) and (iii) diapause and non-diapuase
eggs from the temperate population versus SD and LD eggs from the tropical population (ANOSIM R ¼ 20.0074, p ¼ 0.49).
2688 J. M. Urbanski et al.Desiccation resistance during diapause
Proc. R. Soc. B (2010)
further control for potential intraspecific variation in gene
expression (Whitehead & Crawford 2005) unrelated to
the diapause response of A. albopictus.
The fatty acyl coA elongase we describe in A. albopictus
contains the highly conserved ELO superfamily domain,
and the inferred amino acid sequence exhibits 96 per cent
identity to an A. aegypti fatty acyl coA elongase, identified
from salivary gland transcripts annotated by Ribeiro
et al. (2007). Fatty acid elongation in insects has been
studied in most detail in Musca domestica, Blatella germa-
nica and Triatoma infestans (Vaz et al. 1988; Jua ´rez 2004;
Jua ´rez & Ferna ´ndez 2007). The consensus view from
these studies is that fatty acyl coA elongases encode pro-
teins which are involved in the formation of long chain
fatty acids (Vaz et al. 1988; Jua ´rez 2004; Jua ´rez &
Ferna ´ndez 2007), which can then be converted by decar-
boxylation into hydrocarbon chains (Major & Blomquist
Surface lipids have been associated with desiccation
resistance in a wide variety of insects (Blomquist et al.
1987; Gibbs 1998) by functioning to decrease water
loss rates (Armold & Regnier 1975; Yoder & Denlinger
1991b; Yoder et al. 1992, 1995; Benoit & Denlinger
2007). Furthermore, previous studies have also found
Coudron & Nelson 1981; Kaneko & Katagiri 2004) and
in some cases also decreased water loss rates (Yoder &
Denlinger 1991b; Benoit & Denlinger 2007) specifically
associated with diapause. Increased production of epicu-
ticular lipids thus appears to be a common component
of both diapause and aestivation (dry season diapause)
in insects (Tauber et al. 1986). Previous studies on
A. gambiae (Goltsev et al. 2009) and several Aedes
(Telford 1957; Beckel 1958; Rezende et al. 2008) have
implicated the serosal cuticle to be important in deter-
mining desiccation resistance of eggs. While we cannot
rule out the possibility that the hexane extraction pro-
cedure we used may have removed some lipids from the
serosal cuticle inside the chorion, our results emphasize
that surface lipids on the outside of the chorion can
play an important role in determining the egg desiccation
resistance. A previous study has documented de novo
hydrocarbon synthesis in insect eggs prior to oviposition
(Jua ´rez 1994), supporting a direct role for the fatty acyl
coA elongase transcript we describe in mediating egg
desiccation resistance. However, hydrocarbons may also
be transported through the hemolymph to the oocytes
before oviposition and/or be synthesized during embryo-
logical development (Jua ´rez 1994). Future studies will
use RNA interference (RNAi) to ‘knock down’ transcript
levels at multiple developmental stages in order to more
precisely elucidate the mechanisms linking increased
fatty acyl coA elongase transcript abundance in mature
(stage V) oocytes to increased surface hydrocarbon
abundance of embryonated eggs.
Differences in the composition of surface hydrocar-
bons have also been documented as a component of the
diapause programme in some insects (Jurenka et al.
1998; Kaneko & Katagiri 2004). However, our results
indicate the diapause programme of A. albopictus involves
quantitative (figure 2) but not compositional (figure 3a)
changes in surface hydrocarbons. This conclusion is simi-
lar to results in several other insect systems (Yoder et al.
1995; Kaneko & Katagiri 2004). The range of hydro-
carbon chain lengths detected from the surface of both
temperate and tropical A. albopictus eggs (19–51 carbons,
figure 3) is greater than the range of cuticular hydrocar-
bons found in D. melanogaster (approx. 21–33 carbons,
Foley et al. 2007), but consistent with the diversity
found in other insects (15–55 carbons, Nelson &
It is important to note that the association between
surface lipid levels and water-loss rates appears to differ
between temperate versus tropical eggs. The water loss
rate of the non-diapause eggs from the temperate popu-
lation does not differ from the water loss rates of the
eggs from the tropical population (figure 4). However,
although a direct statistical comparison is not valid
because paired replicate samples of tropical and temper-
ate eggs were not all collected at the same time for the
quantitative hydrocarbon analysis (see §2), the tropical
eggs appear to have higher surface lipid levels than the
temperate eggs (figure 2). As noted above, differences in
the composition of hydrocarbons could in principle con-
tribute to the differences between surface hydrocarbon
abundance and water loss rates in temperate versus
time (days)time (days)
ln (mt /m0)
Figure 4. Water loss rate (slope of (ln (mt/m0)) versus time) of A. albopictus eggs from (a) temperate (VA) and (b) tropical (KL)
population produced under long-day (open circle) and short-day (filled circle) photoperiods. Water loss rates were affected by a
significant population-by-photoperiod interaction (F1,175¼ 126.95, p , 0.001). Mean water loss rates of temperate (VA) dia-
pause eggs were lower (a posteriori contrast with Bonferroni correction, p , 0.05) than non-diapause eggs from the temperate
(VA) population and SD and LD non-diapause eggs from the tropical (KL) population, which did not differ significantly
(a posteriori contrast with Bonferroni correction, p . 0.05).
Desiccation resistance during diapause
J. M. Urbanski et al.
Proc. R. Soc. B (2010)
tropical eggs. However, the compositional hydrocarbon
profiles of temperate and tropical populations do not
differ significantly and in fact are remarkably similar
(figure 3). These results emphasize that in addition to
surface hydrocarbons, there are a variety of potential
metabolic and structural differences between temperate
and tropical eggs that could affect water loss rates. For
example, higher metabolic rates of tropical embryos
(Hadley 1994), differences in osmolite concentration
(Benoit et al. 2009), and size (Benoit & Denlinger
2007) or structural properties (Woods 2005) could all
explain why eggs from tropical populations appear to
have higher quantities of surface lipids (figure 2) but
equivalent water loss rates (figure 4) relative to non-
diapause temperate eggs. We are currently investigating
a number of these factors.
Despite the caveats noted above, our results implicate
fatty acyl coA elongase transcript abundance and hydro-
carbon synthesis as important physiological components
of stress resistance during diapause in A. albopictus.
Based on the strong association between surface lipid pro-
duction and desiccation resistance in other insects
(Armold & Regnier 1975; Blomquist et al. 1987;
Yoder & Denlinger 1991b; Yoder et al. 1992, 1995;
Benoit & Denlinger 2007), we believe the fatty acyl coA
elongase we have characterized may be involved in mediat-
ing the desiccation resistance of non-diapause eggs in
A. albopictus. Furthermore, because the inferred FATTY
ACYL COA ELONGASE amino acid sequence is
highly conserved (see the electronic supplementary
material, S1) we hypothesize that our results may be
pertinentto elucidating the physiological
desiccation resistance of the eggs of other mosquito
vectors. For example, we hypothesize that a similar
pathway may mediate the desiccation resistance of eggs
in A. aegypti, even though A. aegypti does not undergo
a photoperiodic diapause.
In mosquitoes, egg desiccation resistance is a trait of
fundamental ecological importance that has been shown
to play a role in mediating species interactions (Juliano
et al. 2002) as well as contributing to the ability to
become established in non-native habitats (Juliano &
Lounibos 2005). Our current results emphasize that
because stress response physiology is a critical component
of the diapause response, studying the physiology of dia-
pause is likely to uncover fundamental pathways of
stress response physiology relevant to a wide variety of
ecological phenomena. Ultimately, elucidating the under-
lying physiological basis of stress response traits such as
egg desiccation resistance may help to develop novel
approaches to pest control by genetically or chemically
disrupting important stress resistance pathways.
We thank Mr Pingjun Yan and Dr Indra Vythilingam for
providing tropical eggs of A. albopictus and Gina Wimp for
assistance with statistical analysis. We also thank two
anonymous reviewers for helpful comments on previous
versions of this manuscript. This work was supported by
funds from Georgetown University to P. Armbruster, the
J. Urbanski and NIH R01 AI1058279 to D. L. Denlinger.
Armbruster, P. & Conn, J. E. 2006 Geographic variation of
larval growth in North American Aedes albopictus (Diptera:
Culicidae). Ann. Entomol. Soc. Am. 99, 1234–1243.
Armold, M. T. & Regnier, F. E. 1975 Developmental
study of cuticular hydrocarbons of Sarcophaga bullata.
J. Insect Physiol. 21, 1827–1833. (doi:10.1016/0022-
Beckel, W. E. 1958 Investigations of impermeability, dia-
pause, and hatching in the eggs of the mosquito Aedes
hexodontus. Can. J. Zool. 36, 541–554. (doi:10.1139/
Bell, R. A., Nelson, P. R., Borg, T. K. & Caldwell, D. L. 1975
Wax secretion innon-diapausing
pupae of the tobacco hornworm Manduca sexta. J. Insect
Physiol. 21, 1725–1799. (doi:10.1016/0022-1910(75)
Benedict, M. Q., Levine, R. S., Hawley, W. A. & Lounibos,
L. P. 2007 Spread of the tiger: Global risk of invasion by
the mosquito Aedes albopictus. Vector-borne Zoonotic Dis. 7,
Benoit, J. B. & Denlinger, D. L. 2007 Suppression of water
loss during adult diapause in the northern house mos-
quito Culex pipiens. J. Exp. Biol. 210, 217–226. (doi:10.
Benoit, J. B., Lopez-Martinez, G., Elnitsky, M. A., Lee Jr,
R. E. & Denlinger, D. L. 2009 Dehydration-induced
cross tolerance of Belgica antarctica larvae to cold and
heat is facilitated by trehalose accumulation. Comp. Bio-
chem. Physiol. A 152, 518–523. (doi:10.1016/j.cbpa.
Blomquist, G. J., Nelson, D. R. & de Renobales, M. 1987
Chemistry, biochemistry and physiology of insect cuticu-
lar lipids. Arch. Insect. Biochem. Physiol. 6, 227–265.
Clements, A. N. 1992 The biology of mosquitoes: development,
nutrition and reproduction. London, UK: Chapman and
Coudron, T. A. & Nelson, D. R. 1981 Characterization and
distribution of the hydrocarbons found in diapausing
pupae tissues of the tobacco hornworm, Manduca sexta
L. J. Lipid Res. 22, 103–112.
Denlinger, D. L. 2002 Regulation of diapause. Annu. Rev.
Faith, D. P., Minchin, P. R. & Belbin, L. 1987 Compositional
Foley, B., Chenoweth, S. F., Nuzhdin, S. V. & Blows, M. W.
2007 Natural genetic variation in cuticular hydrocarbon
expression in male and female Drosophila melanogaster.
Gibbs, A. G. 1998 Water-proofing properties of cuticular
lipids. Am. Zool. 38, 471–482. (doi:10.1093/icb/38.3.
Goltsev, Y., Rezende, G. L., Vranizan, K., Lanzaro, G.,
Valle, D. & Levine, M. 2009 Developmental and evol-
utionary basis for drought tolerance of the Anopheles
gambiae embryo. Dev. Biol. 330, 462–470. (doi:10.1016/
Goto, S. G., Han, B. & Denlinger, D. L. 2006 A nondiapaus-
ing variant of the flesh fly, Sarcophaga bullata, that shows
arrhythmic adult eclosion and elevated expression of two
circadian clock genes, period and timeless. J. Insect Phy-
siol. 52, 1213–1218. (doi:10.1016/j.jinsphys.2006.09.
Gratz, N. G. 2004 Critical review of the vector status of Aedes
albopictus. Med. Vet. Entomol. 18, 215–227. (doi:10.1111/
2690J. M. Urbanski et al. Desiccation resistance during diapause
Proc. R. Soc. B (2010)
Gray, E. M. & Bradley, T. J. 2005 Physiology of desiccation
resistance in Anopheles gambiae and Anopheles arabiensis.
Am. J. Trop. Med. Hyg. 73, 553–559.
Hadley, N. F. 1994 Water relations of terrestrial arthropods. San
Diego, CA: Academic Press.
Hanson, S. M. & Craig, G. B. 1994 Cold acclimation, dia-
pause, and geographic origin affect cold hardiness in
eggs of Aedes albopictus (Diptera: Culicidae). J. Med.
Entomol. 31, 192–201.
Hawley, W. A. 1988 The biology of Aedes albopictus. J. Am.
Mosq. Control Assoc. 4, 1–39.
Hawley, W. A., Reiter, P., Copeland, R. S., Pumpuni, C. B. &
Craig, G. B. 1987 Aedes albopictus in North America—
probable introduction in used tires from northern
Asia. Science 236, 1114–1116. (doi:10.1126/science.
Joyce, C. R. 1961 Potentialities for accidental establishment
of exotic mosquitoes in Hawaii. Proc. Hawaii Entomol.
Soc. 17, 407–413.
Jua ´rez, P. 1994 Hydrocarbon biosynthesis in Triatoma infes-
tans. Arch. Insect Biochem. Physiol. 25, 193–206. (doi:10.
Jua ´rez, M. P. 2004 Fatty acyl-CoA elongation in Blatella ger-
manica integumental microsomes. Arch. Insect Biochem.
Physiol. 56, 170–178. (doi:10.1002/arch.20007)
Jua ´rez, M. P. & Ferna ´ndez, G. C. 2007 Cuticular hydrocar-
bons of triatomines. Comp. Biochem. Physiol. A: Mol.
Integr. Physiol. 147, 711–730. (doi:10.1016/j.cbpa.2006.
Juliano, S. A. & Lounibos, L. P. 2005 Ecology of invasive
mosquitoes: effects on resident species and on human
health. Ecol. Lett. 8, 558–574. (doi:10.1111/j.1461-
Juliano, S. A., O’Meara, G. F., Morrill, J. R. & Cutwa, M.
M. 2002 Desiccation and thermal tolerance of eggs and
the coexistence of competing mosquitoes. Oecologia 130,
Jurenka, R. A., Holland, D. & Krafsur, E. S. 1998
Hydrocarbon profiles of diapausing and reproductive
adult face flies (Musca autumnalis). Arch. Insect Biochem.
Physiol. 37, 206–214. (doi:10.1002/(SICI)1520-6327
Kaneko, J. & Katagiri, C. 2004 Epicuticular wax of large and
small white butterflies, Pieris brassicae and Prapae cruci-
vora: qualitative and quantitative comparison between
diapause and non-diapause pupae. Naturwissenschaften
91, 320–323. (doi:10.1007/s00114-004-0535-7)
Livak, K. J. & Schmittgen, T. D. 2001 Analysis of relative
gene expression data using real-time quantitative PCR
and the 2(-Delta Delta C(T)) Method. Methods 25,
Major, M. A. & Blomquist, G. J. 1978 Biosynthesis of hydro-
carbons in insects. Decarboxylation of long chain acid to
n-alkanes in Periplaneta. Lipids 13, 323–328. (doi:10.
Minchin, P. R. 1987 Simulation of multidimensional com-
Vegetation 71, 145–156.
Moore, C. G. 1999 Aedes albopictus in the United States: cur-
rent status and prospects for further spread. J. Am. Mosq.
Control Assoc. 15, 221–227.
Mori, A., Oda, T. & Wada, Y. 1981 Studies on the egg dia-
pause and overwintering of Aedes albopictus in Nagasaki.
Trop. Med. 23, 79–90.
Mousson, L., Dauga, C., Garrigues, T., Schaffner, F.,
Vazeille, M. & Failoux, A. B. 2005 Phylogeography of
Aedes (Stegomyia) aegypti (L.) and Aedes (Stegomyia)
albopictus (Skuse) (Diptera: Culicidae) based on mito-
Nelson, D. R. & Blomquist, R. J. 1995 Insect waxes. In
Waxes: chemistry and molecular biology and functions (ed.
R. J. Hamilton), pp. 1–90. Dundee, UK: The Oily Press.
Pavelka, J., Shimada, K. & Kostal, V. 2003 TIMELESS: a
link between fly’s circadian and photoperiodic clocks?
Eur. J. Entomol. 100, 255–265.
Rezende, G. L., Martins, A. J., Gentile, C., Farnesi, L. C.,
Pelajo-Machado, M., Peixoto, A. A. & Valle, D. 2008
Embryonic desiccation resistance in Aedes aegypti: pre-
sumptive role of the chitinized serosal cuticle. BMC
Dev. Biol. 8, 82. (doi:10.1186/1471-213X-8-82)
Ribeiro, J. M. C., Arca, B., Lombardo, F., Calvo, E., Phan,
V. M., Chandra, P. K. & Wikel, S. K. 2007 An annotated
catalogue of salivary gland transcripts in the adult female
mosquito, Aedes aegypti. BMC Genomics 8, (doi:10.1186/
Rinehart, J. P., Li, A. L., Yocum, G. D., Robich, R. M.,
Hayward, S. A. L. & Denlinger, D. L. 2007 Up-regulation
of heat shock proteins is essential for cold survival
during insect diapause. Proc. Natl Acad. Sci. USA 104,
11130–11 137. (doi:10.1073/pnas.0703538104)
Scholte, E. J., Dijkstra, E., Blok, H., De Vries, A., Takken,
W., Hofhuis, A., Koopmans, M., De Boer, A. & Reusken,
C. B. E. M. 2008 Accidental importation of the mosquito
Aedes albopictus into the Netherlands: a survey of mos-
quito distribution and the presence of dengue virus.
Med. Vet. Entomol. 22, 352–358. (doi:10.1111/j.1365-
Sota, T. & Mogi, M. 1992 Survival time and resistance to
desiccation of diapause and nondiapause eggs of temper-
ate Aedes (Stegomyia) mosquitoes. Entomol. Exp. Appl. 63,
Sprenger, D. & Wuithiranyagool, T. 1986 The discovery and
distribution of Aedes albopictus in Harris County, Texas.
J. Am. Mosq. Control Assoc. 2, 217–219.
Syrova, Z., Dolezel, D., Saumann, I. & Hodkova, M. 2003
Photoperiodic regulation of diapause in linden bugs: are
period and clock genes involved? Cell. Mol. Life Sci. 60,
Tauber, M. J., Tauber, C. A. & Masaki, S. 1986 Seasonal
adaptations of insects. New York, NY: Oxford University
Telford, A. D. 1957 The pasture Aedes of central and northern
California. The egg stage: gross embryology and resistance
to desiccation. Ann. Entomol. Soc. Am. 56, 537–543.
Turell, M. J., O’Guinn, M. L., Dohm, D. J. & Jones, J. W.
2001 Vector competence of North American mosquitoes
(Diptera: Culicidae) for West Nile virus. J. Med. Entomol.
38, 130–134. (doi:10.1603/0022-2585-38.2.130)
Urbanski, J., Aruda, A. & Armbruster, P. In press. A tran-
scriptional element of the diapause program in the
Asian tiger mosquito, Aedes albopictus, identified by sup-
Van Handel, E. 1985 Rapid determination of total lipids in
mosquitoes. J. Am. Mosq. Control Assoc. 1, 302–304.
Vaz, A. H., Blomquist, G. J. & Reitz, R. C. 1988 Character-
ization of the fatty acyl elongation reactions involved in
hydrocarbon biosynthesis in the housefly, Musca domestica
L. Insect Biochem. 18, 177–184. (doi:10.1016/0020-
Wang, R. L. 1966 Observations on the influence of photo-
period on egg diapause in Aedes albopictus Skuse. Acta
Entomol. Sinica 15, 75–77.
Warwick, R. M., Clarke, K. R. & Suharnso 1990 A statistical
analysis of coral community responses to the 1982–1983
El Nin ˜o in the Thousand Islands, Indonesia. Coral Reefs 8,
Wharton, G. W. 1985 Water balance of insects. In Compre-
hensive insect physiology, biochemistry, and pharmacology,
Desiccation resistance during diapause
J. M. Urbanski et al.
Proc. R. Soc. B (2010)
vol. 4 (eds G. A. Kerkut & L. I. Gilbert), pp. 565–603. Download full-text
Oxford, UK: Pergamon Press.
Whitehead, A. & Crawford, D. L. 2005 Variation in
populations. Genome Biol. 6, R13. (doi:10.1186/gb-
Woods, H. A. 2005 Oxygen and water flux across eggshells of
Manduca sexta. J. Exp. Biol. 208, 1297–1308. (doi:10.
Yoder, J. A. & Denlinger, D. L. 1991a Water balance in flesh
fly pupae and water vapour absorption associated with
diapause. J. Exp. Biol. 157, 273–286.
Yoder, J. A. & Denlinger, D. L. 1991b A comparison of the
water balance characteristics of temperate and tropical
fly pupae. Physiol. Entomol. 16, 375–380. (doi:10.1111/
Yoder, J. A., Denlinger, D. L., Dennis, M. W. & Kolattukudy,
P. E. 1992 Enhancement of diapausing flesh fly puparia
with additional hydrocarbons and evidence for alkane bio-
Biochem. Mol. Biol. 22, 237–243. (doi:10.1016/0965-
Yoder, J. A., Blomquist, G. J. & Denlinger, D. L. 1995
Hydrocarbon profiles from puparia of diapausing and
nondiapausing flesh flies (Sarcophaga crassipalpis) reflect
quantitative rather than qualitative differences. Arch.
Insect Biochem. Physiol. 28, 377–385. (doi:10.1002/arch.
2692J. M. Urbanski et al. Desiccation resistance during diapause
Proc. R. Soc. B (2010)