letters to nature

considerably depleted d13COC signatures obtained from sediments

deposited during glacial epochs5. However, the observed glacial
interglacial isotopic variations could instead reflect changes from a
C3-dominated (colder temperature) vegetation in the Mississippi
River drainage basin during glaciations to a more C4-dominated
vegetation in the Holocene epoch. This latter interpretation is
supported by the known temperature-dependent latitudinal distribution of plants in present-day North American grasslands21.
Application of these and other biological marker d13C data29, in
combination with compound-specific radiocarbon analysis14, to
downcore studies of marine sediments may help us to understand
the changes in vegetation (C3 versus C4) resulting from continental
climate change in North America and other areas of the
M
world22,2730.
Received 11 July; accepted 1 August 1997.
1. Hedges, J. I. Global biogeochemical cycles: Progress and problems. Mar. Chem. 39, 6793 (1992).
2. Goni, M. A. & Eglinton, T. I. Stable carbon isotopic analyses of lignin-derived CuO oxidation products
by isotope ratio monitoringGas ChromatographyMass spectrometry (irmGCMS). Org.
Geochem. 24, 601615 (1966).
3. Hedges, J. I. & Parker, P. L. Land-derived organic matter in surface sediments from the Gulf of Mexico.
Geochim. Cosmochim. Acta 40, 10191029 (1976).
4. Keil, R. G., Tsamakis, E., Bor Fuh, C., Giddings, J. C. & Hedges, J. I. Mineralogical and textural controls
on the organic composition of coastal marine sediments: hydrodynamic separation using SPLITTfractionation. Geochim. Cosmochim. Acta 58, 879893 (1994).
5. Jasper, J. P. & Gagosian, R. B. Glacial-interglacial climatically forced d13C variations in sedimentary
organic matter. Nature 342, 6062 (1989).
6. Muller, P. J. & Suess, E. Productivity, sedimentation rate, and sedimentary organic matter in the ocean.
I. Organic carbon preservation. Deep Sea Res. 26, 13471367 (1979).
7. Hedges, J. I. & Keil, R. G. Sedimentary organic matter preservation: an assessment and speculative
synthesis. Mar. Chem. 49, 81115 (1995).
8. Martin, J. H., Knauer, G. A., Karl, D. M. & Broenkow, W. W. VERTEX: Carbon cycling in the northeast
Pacific. Deep Sea Res. 34, 267285 (1987).
9. Van Andel, T. H. in Recent Sediments, Northwestern Gulf of Mexico (eds Shepard, F. P., Phleger, F. B. &
Van Andel, T. H.) 3455 (American Assoc. of Petrology and Geology, Tulsa, 1960).
10. Sims, P. L. & Coupland, R. T. Producers. in Grassland Ecosystems of the World: Analyses of Grasslands
and their Uses (ed. Coupland, R. T.) 4972 (Cambridge Univ. Press, 1979).
11. Deegan, L. A. et al. in Estuary Variability (ed. Wolfe, D. A.) 83100 (Academic, New York, 1986).
12. Ruttenberg, K. C. & Goni, M. A. Phosphorous distribution, C : N : P ratios, and d13COC in arctic,
temperate, and tropical coastal sediments: tools for characterizing bulk sedimentary organic matter.
Mar. Geol. 139, 123145 (1997).
13. Stuiver, M., Peason, G. W. & Braziunas, T. Radiocarbon age calibration of marine samples back to 9,00
cal yr. bp. Radiocarbon 28, 9801021 (1986).
14. Eglinton, T. I. et al. Variability in radiocarbon ages of individual organic compounds from marine
sediments. Science 277, 796799 (1997).
15. Emerson, S. C. et al. Organic carbon in surface deep-sea sediments: C-14 concentration. Nature 329,
5154 (1987).
16. Hedges, J. I. & Mann, D. C. The characterization of plant tissues by their cupric oxide oxication
products. Geochim. Cosmochim. Acta 43, 18031807 (1979).
17. Opsahl, S. & Benner, R. Early diagenesis of vascular plant tissues: lignin and cutin decomposition and
biogeochemical implications. Geochim. Cosmochim. Acta 59, 48894904 (1995).
18. Goni, M. A., Nelson, B., Blanchette, R. A. & Hedges, J. H. Fungal degradation of wood lignins:
geochemical perspectives from CuO-derived phenolic dimers and monomers. Geochim. Cosmochim.
Acta 57, 39854002 (1993).
19. Hedges, J. I. et al. Compositions and fluxes of particulate organic material in the Amazon river.
Limnol. Oceanogr. 31, 717738 (1986).
20. Ziegler, F., Kogel, I. & Zech, W. Z. Alteration of gymnosperm and angiosperm lignin during
decomposition in forest humus layers. Z. Pflanzenernaehr. Bodenk. 149, 323331 (1986).
21. Teeri, J. A. & Stowe, L. G. Climatic patterns and the distribution of C4 grasses in North America.
Oecologia 23, 112 (1976).
22. Tieszen, L. L. & Archer, S. in Plant Biology of the Basin and Range (eds Osmond, C. S., Pitelka, L. F. &
Hidy, G. M.) 293321 (Ecological Studies Ser. Vol. 80, Springer, Heidelberg, 1990).
23. Parton, W. J., Schimel, D. S., Cole, C. V. & Ojima, D. S. Analysis of factors controlling soil organic
matter levels in Great Plains grasslands. Soil Sci. Soc. Am. J. 51, 11731179 (1987).
24. Chmura, G. L. Palynomorph distribution in marsh environments in the modern Mississippi Delta
Plain. Geol. Soc. Am. Bull. 106, 705714 (1994).
25. Prahl, F. G., Ertel, J. R., Goni, M. A., Sparrow, M. A. & Eversmeyer, B. Terrestrial organic carbon
contributions to sediments on the Washington margin. Geochim. Cosmochim. Acta 58, 30353048
(1994).
26. Goni, M. A. in Proceedings of the Ocean Drilling Program, Sci. Rep. 155 (eds Flood, R. D., Piper, D. J. W.
& Klaus, A.) 519530 (College Station, TX, 1997).
27. France-Lanord, C. & Derry, L. A. d13C of organic carbon in the Bengal Fan: source evolution and
transport of C3 and C4 plant carbon to marine sediments. Geochim. Cosmochim. Acta 58, 48094814
(1994).
28. Parton, W. J. et al. Observations and modeling of biomass and soil organic matter dynamics for the
grassland biome worldwide. Global Biochem. Cycles 7, 785809 (1993).
29. Bird, M. I. et al. Terrestrial vegetation change inferred from n-alkane d13C analysis in the marine
environment. Geochim. Cosmochim. Acta 59, 28532857 (1995).
30. Cerling, T. E., Wang, Y. & Quade, J. Expansion of C4 ecosystems as an indicator of global ecological
change in the late Miocene. Nature 361, 344345 (1993).
Acknowledgements. We thank J. W. Morse for the opportunity to participate on the cruise in Gulf of
Mexico; E. Tsamakis, K. Tholke, R. Ostermann, J. Irvine, N. Parmentier, L. Christman, J. Donahue and
E. Tappa for help with sample preparation and analyses; and G. Eglinton, B. Fry, J. Jasper, J. Hedges,
R. Benner, S. Smith and S. Montgomery for comments and discussions. This work was supported by the
US NSF.
Correspondence and requests for materials should be addressed to M.A.G. (e-mail: goni@epoch.geol.
sc.edu).

278

Trade-off between parasitoid

resistance and larval
competitive ability in
Drosophila melanogaster
A. R. Kraaijeveld & H. C. J. Godfray
NERC Centre for Population Biology and Department of Biology,
Imperial College at Silwood Park, Ascot, Berkshire SL5 7PY, UK
.........................................................................................................................

The extent to which an organism is selected to invest in defences

against pathogens and parasites depends on the advantages that
ensue should infection occur, but also on the costs of maintaining
defences in the absence of infection. The presence of heritable
variation in resistance suggests that costs exist, but we know very
little about the nature or magnitude of these costs in natural
populations of animals1. A powerful technique for identifying
trade-offs between fitness components is the study of correlated
responses to artificial selection2,3. We have selected Drosophila
melanogaster for improved resistance against an endoparasitoid,
Asobara tabida. Endoparasitoids are insects whose larvae develop
internally within the body of other insects, eventually killing
them, although their hosts can sometimes survive attack by
mounting a cellular immune response46. We found that reduced
larval competitive ability in unparasitized D. melanogaster is a
correlated response to artificial selection for improved resistance
against A. tabida. The strength of selection for competitive ability
and parasitoid resistance is likely to vary temporally and spatially,
which may explain the observed heritable variation in resistance.
Parasitoids are very common natural enemies of many insects,
developing either externally or internally on their hosts7. Endoparasitoids frequently suspend development as eggs or first-instar
larvae while their hosts feed and grow in size, although during this
period they can be destroyed by encapsulation, a cellular immune
response. Parasitoids avoid host defences by hiding in tissue away
from circulating haemocytes, by avoidance of recognition through
molecular mimicry, or through debilitating the immune response
by toxins or symbiotic viruses6. Individual hosts often differ in their
ability to encapsulate parasitoids, and selection, quantitative genetic
and isofemale-line studies have shown this to have a genetic
basis811. The presence of additive genetic variation in encapsulation
ability suggests that there are costs to the possession of defences
against parasitoids, although these costs have yet to be identified in
any hostparasitoid interaction.
Asobara tabida (Hymenoptera, Braconidae) is a common larval
parasitoid of Drosophila species on fermenting substrates in Europe.
In central and northern Europe its main host is D. subobscura, in
which it is never encapsulated, although it is also found on
D. melanogaster at low frequencies12. In Mediterranean Europe D.
melanogaster is the main host. Our base stock population of
D. melanogaster was derived from 250 flies caught in the wild
near Leiden in the Netherlands and maintained in the laboratory
for three generations before the start of the experiment. The wasps
used in the selection experiments were a laboratory strain derived
originally from insects caught in Sospel, France, and cultured on D.
subobscura to prevent adaptation to D. melanogaster. The base stock
encapsulated only 5% of the eggs of the laboratory strain of A.
tabida, a figure typical of fly populations from northern Europe13.
The base stock was split into eight lines, four of which were
selected for increased encapsulation ability, the others acting as
controls. Each new generation of the selection lines was bred from
flies that had survived parasitoid attack (the encapsulated parasitoid
egg remains in the body of the fly and can be seen through the wall of

Nature Macmillan Publishers Ltd 1997

NATURE | VOL 389 | 18 SEPTEMBER 1997

letters to nature
the puparium). To minimize inbreeding, a population size of at least
100 was maintained in all lines. With the exception of exposure to
parasitoids, control and selection lines were cultured using identical
procedures. Larvae were exposed to parasitoids in the selection
treatment in conditions that lead to high rates of parasitism to avoid
selection on traits other than parasitoid resistance. The encapsulation ability of the flies was assessed by exposing 200 second-instar
larvae to parasitoid females for 2 h, after which the larvae were kept
at 20 8C for five days and then dissected to score capsule frequency.
The response to selection was rapid and similar across the four
lines (Fig. 1). The probability of encapsulation rose from 5% to a
maximum of ,60% in five generations. The maximum encapsulation ability is similar to that found in Mediterranean populations,
which are heavily attacked by A. tabida13. After selection, four pairs
of F1 lines were created by crossing each of the selected lines with a
different control line. In one cross of the pair, males derived from
selected lines and females from control lines, whereas in the other
cross the sexes were reversed. The mean encapsulation rate of the
offspring was 34% (standard error 4%), and was not influenced by
the origin of the female parent (t-test, P 0:9), excluding maternal
or cytoplasmic effects, and strongly suggesting that the genes
influencing resistance are not sex-linked.
We searched for correlated responses to selection by comparing a
range of fitness parameters in the control and selected lines. In a first
set of experiments we compared the performance of flies reared
under conditions of abundant food and very low competition. The
full details of the experiments will be reported elsewhere, but in
summary we found no significant differences in larval and pupal
survival, larval and pupal development time, adult longevity in the
absence of food, early female fecundity, adult size or fluctuating
asymmetry. Working with only eight lines in total we have limited
statistical power to detect subtle differences, but in no cases were
strong but insignificant trends detected.
In a second set of experiments we compared the performance of
control and selected flies under conditions of weak to strong
intraspecific competition (Fig. 2). Competitive ability was assessed
by rearing larvae with flies from a genetically marked tester
stock14,15 at four different levels of competition for food (see

Methods for experimental and statistical details). When competition was weak (0.4 ml and 0.2 ml larval food), survival was high
(,80%) and there was no difference between the two sets of lines
(P . 0:1). However, for more severe competition (0.1 ml), survival
was reduced and there was significantly greater mortality among
selected than control flies (P 0:011). This difference was also seen
at the most severe level of competition (0.05 ml), where survival was
much lower (,50%), although the statistical significance was less
(P 0:061) owing to the much greater within-line variability in
survival when resources were very scarce. At higher levels of
competition, flies are smaller, take longer to develop, and show
more fluctuating asymmetry, but we found no significant differences between the performances of the control and selected lines
(although there was a trend for selected flies to be smaller at the two
lower levels of competition).
Thus reduced survival in a high-competition environment is a
correlated response to selection for improved defence against
endoparasitoid attack in D. melanogaster. This suggests there is a
trade-off between defence against parasitoids and other components of fitness. We suspect that selected larvae allocate more
resources to the machinery of cellular encapsulation or to counteracting the wasps attempts to disable encapsulation, and hence are
less able to withstand very stressful conditions. Several other tradeoffs in the life-history strategies of D. melanogaster have been
identified only in circumstances when the fly is under stress2.
D. melanogaster populations in the wild feed in rotting fruit and
other fermenting substrates, and often experience a level of competition in the field comparable with those at which survival of the
selected larvae was reduced16.
We believe this to be the first demonstration of a cost of resistance
against parasitoids. For many species of hosts, parasitoid attack
rates vary both temporally and spatially, and the resulting fluctuating selection pressures, combined with costs to parasitoid defences,
may explain the additive genetic variance in defensive ability
observed in several host species810. The population dynamics of
hosts and their parasitoids are often coupled, and our findings
suggest that there may be complex joint evolutionary and population dynamic interactions. More generally, evidence for costs of

Figure 1 The frequency of encapsulation in control and selected flies, showing

Figure 2 The competitive ability of experimental flies (control lines, open

means and standard errors of the four selected and control lines. For logistical

symbols; selected lines, filled symbols) relative to a tester strain, showing means

reasons (assessment of encapsulation is very time consuming), control lines

and standard errors of the four selected or control lines at four levels of

were not assessed every generation. Differences were significant in generations

competition for larval food. The competition index and the statistical analysis are

5 (t-test, P , 0:01) and 8 (t-test, P , 0:01).

described in the Methods section.

NATURE | VOL 389 | 18 SEPTEMBER 1997

Nature Macmillan Publishers Ltd 1997

279

letters to nature
resistance against pathogens and parasites in animals is scarce, and
comes largely from studies on populations of farm or laboratory
animals with highly modified genetic backgrounds (reviewed in
ref. 1). Finally, because of the enormous amount known about
Drosophila genetics, we suggest that the interactions between
D. melanogaster and A. tabida may be a valuable model system for
the study of the evolution of resistance and of the genetic basis of
M
adaptation.
.........................................................................................................................

Methods

Competitive ability of selected and control lines. Competitive ability was

assessed by placing 15 second-instar larvae from the experimental lines (either

control or selected insects) with 15 second-instar larvae from the tester stock
(sparkling poliert, an eye-colour mutant) in agar-lined Petri dishes with
variable amounts of larval medium. Four food levels were used: 0.4, 0.2, 0.1 or
0.05 ml of larval medium (25 g live bakers yeast per 100 ml water), which
represent weak to strong competition regimes. We recorded the number of
experimental and tester flies that survived, the development time to pupation
and to adult eclosion, and female size and fluctuating asymmetry (wing
length). We performed 15 replicates of each combination of line and food level.
Data analysis. Survival data were analysed using both a robust competition
index (I) and a more sophisticated analysis using generalized linear modelling
techniques (II). For analysis (I) we calculated the competition index15,
loge=t 1, where e is the number of experimental and t is the number of
tester flies that survived in each replicate. The means of the competition index
for the eight lines were calculated and differences between selected and control
lines tested using the t-test with unequal variances. For analysis (II) the
untransformed survival data for the experimental flies were analysed in a
generalized linear model with binomial error variances, but using quasilikelihood estimation to account for overdispersion17. The numbers of tester
flies surviving was used as a covariate and significance assessed by nesting lines
within control or selection treatments (giving an F-statistic with one and six
degrees of freedom). Examination of the distribution of residuals supported the
choice of model. The results of the two statistical analyses were in broad
agreement (values in the text are from analysis (II)), with P for (I) given first:
0.4 ml, P . 0:1, P . 0:1; 0.2 ml, P . 0:1, P . 0:1; 0.1 ml, P 0:033,
P 0:011; 0.05 ml, P 0:066, P 0:061. The data for size and development
time were analysed using the appropriate general linear models equivalent to
analysis (II).
Received 25 March; accepted 8 July 1997.
1. Read, A. F. et al. in Ecology of Infectious Diseases in Natural Populations (eds Grenfell, B. T. & Dobson,
A. P.) 450477 (Cambridge Univ. Press, 1995).
2. Partridge, L. & Fowler, K. Direct and correlated responses to selection on age at reproduction in
Drosophila melanogaster. Evolution 46, 7691 (1992).
3. Rose, M. R. Laboratory evolution of postponed senescence in Drosophila melanogaster. Evolution 38,
10041010 (1984).
4. Salt, G. The Cellular Defence Reactions of Insects (Cambridge Univ. Press, 1970).
5. Gupta, A. P. in Comprehensive Insect Physiology, Biochemistry and Pharmocology, Vol. 3 (eds Kerkut, G.
A. & Gilbert, L. I.) 401451 (Pergamon, Oxford, 1985).
6. Strand, M. R. & Pech, L. L. Immunological basis for compatibility in parasitoid-host relationships.
Annu. Rev. Entomol. 40, 3156 (1995).
7. Godfray, H. C. J. Parasitoids, Behavioral and Evolutionary Ecology, (Princeton Univ. Press, NJ, 1994).
8. Carton, Y. & Bouletreau, M. Encapsulation ability of Drosophila melanogaster: a genetic analysis. Dev.
Comp. Immunol. 9, 211219 (1985).
9. Carton, Y. & Nappi, A. J. The Drosophila immune reaction and the parasitoid capacity to evade it:
genetic and coevolutionary aspects. Acta Oecol. 12, 89104 (1991).
10. Henter, H. J. & Via, S. The potential for coevolution in a host-parasitoid system. 1. Genetic variation
within an aphid population in susceptibility to a parasitic wasp. Evolution 49, 427438 (1995).
11. Hughes, K. & Sokolowski, M. B. Natural selection in the laboratory for a change in resistance by
Drosophila melanogaster to the parasitoid wasp Asobara tabida. J. Insect Behav. 9, 477491 (1996).
12. Kraaijeveld, A. R. & van der Wel, N. N. Geographic variation in reproductive success of the parasitoid
Asobara tabida in larvae of several Drosophila species. Ecol. Entomol. 19, 221229 (1994).
13. Kraaijeveld, A. R. & van Alphen, J. J. M. Geographical variation in encapsulation ability of Drosophila
melanogaster larvae and evidence for parasitoid-specific components. Evol. Ecol. 9, 1017 (1995).
14. Lewonton, R. C. The effects of population density and composition on viability in Drosophila
melanogaster. Evolution 9, 2741 (1955).
15. Santos, M., Fowler, K. & Partridge, L. On the use of tester stocks to predict the competitive ability of
genotypes. Heredity 69, 489495 (1992).
16. Atkinson, W. D. A field investigation of larval competition in domestic Drosophila. J. Anim. Ecol. 48,
91102 (1979).
17. McCullagh, P. & Nelder, J. A. Generalized Linear Models 2nd edn (Chapman & Hall, London, 1989).
Acknowledgements. We thank J. van Alphen, G. Boskamp, A. Burt, D. Ebert, J. Ellers, M. Fellowes, A.
James, A. Leroi, J. McCabe, A. Orr, L. Partridge, A. Read, M. Rees and J. Werren for help and advice.
Correspondence and requests for materials should be addressed to H.C.J.G. (e-mail: c.godfray@ic.ac.uk).

280

Responses of primary visual

cortical neurons to binocular
disparity without depth
perception
B. G. Cumming & A. J. Parker
University Laboratory of Physiology, Parks Road, Oxford OX1 3PT, UK
.........................................................................................................................

The identification of brain regions that are associated with the

conscious perception of visual stimuli is a major goal in
neuroscience1. Here we present a test of whether the signals on
neurons in cortical area V1 correspond directly to our conscious
perception of binocular stereoscopic depth. Depth perception
requires that image features on one retina are first matched
with appropriate features on the other retina. The mechanisms
that perform this matching can be examined by using random-dot
stereograms2, in which the left and right eyes view randomly
positioned but binocularly correlated dots. We exploit the fact
that anticorrelated random-dot stereograms (in which dots in one
eye are matched geometrically to dots of the opposite contrast in
the other eye) do not give rise to the perception of depth3 because
the matching process does not find a consistent solution. Anticorrelated random-dot stereograms contain binocular features
that could excite neurons that have not solved the correspondence
problem. We demonstrate that disparity-selective neurons in V1
signal the disparity of anticorrelated random-dot stereograms,
indicating that they do not unambiguously signal stereoscopic
depth. Hence single V1 neurons cannot account for the conscious
perception of stereopsis, although combining the outputs of many
V1 neurons could solve the matching problem. The accompanying
paper4 suggests an additional function for disparity signals from
V1: they may be important for the rapid involuntary control of
vergence eye movements (eye movements that bring the images on
the two foveae into register).
When the image of an object falls on different locations on the
two retinae, this binocular disparity gives rise to a sensation of depth
(stereopsis). This process requires that an image feature on one
retina be matched with an appropriate feature on the other retina,
even though there are inevitably similar features nearby offering
potential matches. This problem is particularly apparent in
random-dot stereograms (RDS), where there are many identical
dots in both images. Of course it is not necessary to consider false
matches in RDS on a dot-by-dot basisspatial filtering of the
image before matching can substantially reduce the number of false
matches5. However, this filtering alone is insufficient to solve the
correspondence problem and further computation is required to
eliminate false matches. Consideration of the overall pattern of
disparities can indicate which potential matches form a globally
consistent pattern, and many computer algorithms achieve this5,6.
Although single neurons in V1 have long been known to signal
the disparity of a stimulus7,8, their role in stereo matching is unclear.
If they are directly responsible for the conscious perception of
depth, they should respond only when matches within the receptive
field are registered as globally correct. If binocular neurons respond
equally well to false matches at appropriate disparities, then further
processing (presumably outside V1) is required to account for the
psychophysical ability to discard false matches.
Whether or not disparity-selective neurons respond only to
correct matches is an open question. It has been demonstrated
that some V1 complex cells display disparity selectivity to dynamic
RDSs and it was concluded that these neurons signal the correct
binocular matches among a multitude of false matches ... (global

Nature Macmillan Publishers Ltd 1997

NATURE | VOL 389 | 18 SEPTEMBER 1997