DispatchR117away from it. To understand thisseeming paradox, Hwang et al. [8]considered how such a behavior mighthave evolved. In nature, a seriousthreat for insect larvae are parasitoids,insects whose larvae feed from thebody of other insects [11]. Drosophilamelanogaster has such an enemy in theparasitoid wasp Leptopilina boulardi,whose females penetrate the larvaewith their ovipositor and lay their eggs

cerebral signature for pain perception and itsmodulation. Neuron 55, 377391.

2. Vosshall, L.B., and Stocker, R.F. (2007).Molecular architecture of smell and taste inDrosophila. Annu. Rev. Neurosci. 30,505533.

3. Kernan, M.J. (2007). Mechanotransduction andauditory transduction in Drosophila. PflugersArch. 454, 703720.

4. Tobin, D.M., and Bargmann, C.I. (2004).Invertebrate nociception: behaviors, neuronsand molecules. J. Neurobiol. 61, 161174.

5. Woolf, C.J., and Ma, Q. (2007). Nociceptorsnoxious stimulus detectors. Neuron 55,353364.behavior, whereas activation of theother classes of neurons caused anaccordion-like contraction of thelarvae. This experiment is not onlyfascinating because of the moderngenetic tricks that allowed Hwang et al.[8] to turn light into a harmful stimulus.The experiment clearly demonstratesthat activation of class IV neurons issufficient to cause the nocifensiverolling behavior, whereas the otherclasses ofmultidendritic neuronsmightserve different functions in the contextof coordinated locomotion.One observation was puzzling,

however: the larvae rolled more oftentowards the side from which thenoxious stimulus came rather thanBiodiversity: ClimatHabitat Loss WhSpecies?

Habitat loss and climate change both killatter is a potent threat. Worse, its victipresently threatened by habitat loss.

Stuart L. Pimm

The important questions about speciesextinctions in the face of global climatechange are clear. As geographicalranges shift, some may contract. Forsome, the envelope of climaticconditions that describes the speciespresent range may no longer exist. Thespecies will likely go extinct. Howmanyspecies might suffer this fate [1]?Vitally, because a species cannot goextinct twice, are the species thatglobal change dooms different fromthe ones that habitat losses haveexterminated or soon will?The problems come from predicting

future species ranges. The idea thatspecies ranges might be neatlyinside the larval body. Hwang et al. [8]showed that indeed Drosophila larvaecan defend themselves against suchwasp attacks by performing theirrolling behavior. In a kung-fu-likefashion, the larva wraps the waspssting around its body, flips theattacking wasp through the air andonto its back,which gives the larva timeto escape. This fascinating new studyby Hwang et al. [8] vividly illustratesthat animal behavior often can beunderstood only if the context of theanimals natural ecology is taken intoaccount.

References1. Tracey, I., and Mantyh, P.W. (2007). Thee Change orich Will Kill More

l off species. New studies show that thems will likely be mostly those not

constrained by temperature and otherclimatic variables is irresistiblyseductive to anyone capable of gainingaccess to databases on speciesranges and the output of climatemodels predicting future climatevariables, and who has the ability towrite the necessary computer code toconnect one to the other. Theassumptions to do this, however, aremany and important. Are suchmodelling exercises predicting futureranges simply pointless?Science has measured the

increasing atmospheric concentrationsof greenhouse gases with improvingprecision for decades. The physicalconsequences range from simpleincreases in global temperatures to6. Tracey, W.D., Jr., Wilson, R.I., Laurent, G., andBenzer, S. (2003). Painless, a Drosophila geneessential for nociception. Cell 113, 261273.

7. Song, W., Onishi, M., Jan, L.Y., and Jan, Y.N.(2007). Peripheral multidendritic sensoryneurons are necessary for rhythmic locomotionbehavior in Drosophila larvae. Proc. Natl. Acad.Sci. USA 104, 51995204.

8. Hwang, R.Y., Zhong, L., Xu, Y., Johnson, T.,Zhang, F., Deisseroth, K., and Tracey, W.D.(2007). Nociceptive neurons protect Drosophilalarvae from parasitoid wasps. Curr. Biol. 17,21052116.

9. Boyden, E.S., Zhang, F., Bamberg, E.,Nagel, G., and Deisseroth, K. (2005).Millisecond-timescale, genetically targetedoptical control of neural activity. Nat. Neurosci.8, 12631268.

10. Schroll, C., Riemensperger, T., Bucher, D.,Ehmer, J., Voller, T., Erbguth, K., Gerber, B.,Hendel, T., Nagel, G., Buchner, E., et al.(2006). Light-induced activation of distinctmodulatory neurons triggers appetitive oraversive learning in Drosophila larvae. Curr.Biol. 16, 17411747.

11. Godfray, H.C. (2004). Parasitoids. Curr. Biol. 14,R456.

Julius-Maximilians-Universitat Wurzburg,Lehrstuhl fur Genetik und Neurobiologie,Theodor-Boveri-Institut, Biozentrum,Am Hubland, 97074 Wurzburg, Germany.E-mail: afiala@biozentrum.uni-wuerzburg.de

DOI: 10.1016/j.cub.2007.11.054many more complex, and sosometimes less certain, changes. Yet,temperature increases alone createineluctable biological complexities.Every species has to worry aboutfinding food, avoiding predators,disease, and other enemies, selectingthe nooks and crannies and otherphysical features that affordprotection, and so on down a long list.The temperature at which MotherNature sets her thermostat is only oneof many concerns. That said, manyplant species are limited by frosts andmuch solid natural history suggeststhat temperature and rainfall are keypredictors of species ranges. Roots[2] quantification of the large fractionof North American bird species thathave their northern winter rangescorresponding to particular averagewinter temperatures confirms thattemperature is a factor that eitherdirectly or indirectly affects manyspecies over large areas.Does a warming climate actually

change what species do? That isa tougher question. The importantanswer must again be a general onethat evaluates most species across

iretained almost all their forest cover. With global warming, this species will have to move

Current Biology Vol 18 No 3many ecosystems. Ecological effectsare so diverse and multi-factorialthat one can usually find a fewcherry-picked examples of almostanything. Comprehensive analyses byRoot et al. [3], Parmesan and Yohe [4],and Fitter and Fitter [5] examineda thousand or more individual studies.Not every study finds significantchanges, but these analyses founda huge preponderance of biologicalchanges in the expected direction. Themost common data are from onelocation over many years theflowering times of plants or the arrivaldates of bird migrants, for example.Importantly, there are studies acrossmany locations. They show speciesmoving to places previously too coldfor them. Great Britain, cold, damp,with depressingly few species, butinexplicably large numbers of amateurnaturalists to care for them, providesstudies with detail rarely duplicatedelsewhere [6].Whether species ranges shrink with

global warming is yet another issue.Weknow species best at their cold,northern limits, where they are

expanding. Increasingly compellingcompilations find that many completeranges are shrinking [7].Geneticalgorithms, suchasGARP [8],

can provide near-perfect descriptionsof present day species ranges. With noconstraint on thenumberofparameters,they would also readily describe theMona Lisas smile. Understanding isanothermatter. Fewstudies can test thepredictions of future ranges against realdata, for one needs spatially detailedand consistent observations overwidely separated times. Britishbirdwatchers are up to the task.As Araujo and Rahbek [8] report,

the models are not. Some 90% of the116 species modelled using theiractual 1970s distribution differedmarkedly in their actual and predictedranges in the 1990s. One of the moreobvious examples is the red-backedshrike, Lanius collurio, a bird Iremember fondly from my youthbecause it occurred in warm, openhabitats that afforded a chance to dryout my usually sodden campingequipment as I looked for it. As theclimate warmed it should have become

uphill to retain the same envelope of temperatures into the very much smaller areasshown in green. Figure generated using imagery from ESRIs ArcGIS Online data service(ESRI, Inc., Redlands, California). Species range map data provided by NatureServe incollaboration with Robert Ridgely, James Zook, The Nature Conservancys Migratory BirdProgram, Conservation Internationals CABS, World Wildlife Fund - US, and EnvironmentCanada - WILDSPACE.Figure 1. A perspective view of the coastal moin Brazil, looking towards the north-east.

From the island in the bottom centre to the p360 km. Forested areas are green, while cleaexcept along two mountain ranges one follothe field guide range, of T. atra, Sekerciogluelevations, shown here in translucent blue. T

R118st forests of (mostly) the State of Rio de Janeiro

oint (Cabo Frio) at the middle right is roughlyred land is buff-coloured. Little forest remainswing the coast, the other further inland. Withinet al. [11] chose areas within the present dayhis species presently lives in areas that havewidespread. In fact, it no longer breedsin Britain regularly.Perhaps one should not expect to

be able to predict future ranges.Studies of British insects show that thepatterns of northward range expansionare mechanistically very complex [9],with changes in habitats, food plants,and dispersal that conventionalwisdom might have easily expected tobe in the opposite directions to thoseobserved [6]. Studies in theNetherlands of pied flycatchers andtheir insect prey show equally complexpatterns varying from year to year andplace to place [10]. Ecological realitymay not respect computationalconvenience.Given these difficulties, how should

one respond to recent work bySekercioglu et al. [11] that predicts thefate of the worlds 8500 landbirdspecies. They predict that 400 to 550 ofthem will go extinct by 2100 fora warming estimate of 2.8C, witha further 2150 species at risk ofextinction. Perhaps their most alarmingprediction is that only 21% of thesespecies are presently on the watch-listof species at risk of extinction [12]. Thatlist already contains about 12% of theworld birds, ones mostly threatenedby habitat loss, but also a basket ofother factors.In the Americas, the coastal moist

tropical forests of coastal Brazil containthe greatest concentration ofthreatened bird species [13]. Thisbiodiversity hotspot contains morethan 200 endemic bird species. It haslost more than 90% of its natural forestcover, with the greatest losses in theaccessible lowlands. The ruggedhigher elevations still have extensiveforest cover (Figure 1).Sekercioglu et al. [11] compiled

species range maps and thentrimmed them. For a detailed view oftheir methods, consider the black andgold cotinga, Tijuca atra, that livesbetween 1200 and 2050 meters abovesea level in the regions cloud forests.Its field guide range consists of anarea that spans the two roughly parallelmountain ranges running from thebottom left of the figure to the topright and all the lowland areas inbetween. Sekercioglu et al.s [11] firststep was to trim the field guide rangesto the appropriate elevations. In thiscase, only a very small fraction of thispossible range is within the birdselevation limitsvisible in the figure assmall isolated patches of blue.

Many rare, narrowly limited specieswith small geographical ranges livein lowland tropical forests and arealready threatened by the extensivedeforestation there. For New Worldpasserine bird species,w1500 livemostly below 1000 meters. There are,however, anotherw500 that liveentirely above 1000 meters abovesea level [15]. In coastal Brazil, theyhave fared better than their lowlandcounterparts. Thus far, that is. Simplephysics determines that there is alwaysless area, the higher one climbs.Certainly, the climate models can

make the wrong predictions. The shrikeillustrates Sods law everything thatcan go wrong will and at the worstpossible moment. The bird did worse,not better than expected. Faced with

2. Root, T.L. (1988). Atlas of Wintering NorthAmerican Birds (Chicago, Illinois: University ofChicago Press).

3. Root, T.L., Price, J.T., Hall, K.R., Schneider, S.H.,Rosenzweig, C., and Pounds, A.J. (2003).Fingerprints of global warming on wild animalsand plants. Nature 421, 5760.

4. Parmesan, C., and Yohe, G. (2003). A globallycoherent fingerprint of climate change impactsacross natural systems. Nature 421, 3742.

5. Fitter, A.H., and Fitter, R.S.R. (2002). Rapidchanges in flowering time in British plants.Science 296, 16891691.

6. Pimm, S.L. (2001). Entrepreneurial insects.Nature 411, 521532.

7. Thomas, C.D., Franco, A.M.A., and Hill, J.K.(2006). Range retractions and extinction in theface of global warming. Trends Ecol. Evol. 21,415416.

8. Araujo, M.B., and Rahbek, C. (2006). How doesclimate change affect biodiversity? Science313, 13961397.

9. Thomas, C.D., Bodsworth, E.J., Wilson, R.J.,Simmons, A.D., Davies, Z.G., Musche, M., andConradt, L. (2001). Ecological and evolutionaryprocesses at expanding range margins. Nature411, 577581.

10. Both, C., Bouwhul, S., Lessells, C.M., and

m

o

DispatchWith global warming, Sekerciogluet al. [11] predict that this speciesshould live 560 meters higher,corresponding to a 2.8C warming.Their model does not permit a futurerange to fall outside of the geographicalfield guide range. It does allowspecies to move upwards inside it.There are some areas where it could(theoretically) move uphill, to the placesshown in light green. Not everypopulation will be so lucky. Otherpopulations have no higher elevationsinto which to escape.T. atra lives in remote mountain

forests. It is common and has fared farbetter than lowland forest speciesbecause of its isolation. Yet, it is clearlya species at considerable risk fromclimate change. At even greater risk isits congener, the grey-winged cotinga,Tijuca condita. Sekercioglu et al. [11]did not model for it as it is one of theworlds rarest species and there wasinsufficient information about it [12].Living in forests at higher elevationsthan T. atra, it has even fewer places towhich to flee the heat.Understanding the details of

Sekercioglu et al. [11] is a dauntingtask. So, too, is understanding Jetzet al. [14], whose work covers similarground and draws broadly comparableconclusions. I had many moments ofwhy did they make that assumption?and what would be different if theyused my scenario rather than theirs?Moreover, these two species reach thepresent day tree line in many locations,so if the species are to move uphill,then their forest habitat must do sofirst and do so within the century inwhich the climate will warm. Thesequestions come on top of all theconcerns raised earlier about the entireclimate-envelope modelling process.Do these results have any credibility?My emphatic yes comes from

considering the governing dynamics.First, across a wide range of taxa, mostspecies have much smaller thanaverage geographical ranges. This isthe reverse of the Lake Wobegoneffect, where famously all children areabove average. Simply, there are a fewspecies with huge geographical rangesand they inflate the average. Second,Mother Nature is unkind. Small-rangedspecies are typically both locally rareand have narrow elevational limits [15].It gets worse. If NewWorld passerine

birds are typical, a quarter of all specieslive in mountains and so face movinguphill into inevitably smaller areas.

R119a promising future in Britain,persecution (from egg collectors) andthe birds small populations infragmentedhabitats likely doomed it.Ofcourse, massive habitat fragmentationand hunting are threats to many of theworlds small-ranged species.The ecological details may be

sublime, and the models of themworryingly simplistic, but theoverarching conclusion is chilling.Large numbers of species, thus-farlargely unaffected by human actions,are in danger of extinction from climatechange.

References1. Thomas, C.D., Cameron, A., Green, R.E.,

Bakkenes, M., Beaumont, L.J.,Collingham, Y.C., Erasmus, B.F.N., Ferreira deSiqueira, M., Grainger, A., Hannah, L., et al.(2004). Extinction risk from climate change.Nature 427, 145148.

Developmental BioIntercalation One S

Formation of the primitive streak, the estep during the early development of aintercalation and the planar cell polaritystep of gastrulation in the chick embry

Bertrand Benazerafand Olivier Pourquie*

Formation of the primitive streak is thefirst sign of gastrulation in amniotes,such as birds and mammals. Theprimitive streak is a stripe of cellsdefining the future midline of theVisser, M.E. (2006). Climate change andpopulation declines in a long-distancemigratory bird. Nature 441, 8183.

11. Sekercioglu, C.H., Schneider, S.H., Fay, J.P.,and Loarie, S.R. (2008). Climate change,elevational range shifts, and bird extinctions.Conserv. Biol., in press.

12. BirdLife International (2000). Threatened Birdsof the World (Cambridge, UK: Lynx Edicionsand BirdLife International).

13. Pimm, S.L., and Jenkins, C.N. (2005). Sustainingthe variety of life. Sci. Am. 293, 6673.

14. Jetz, W., Wilcove, D.S., and Dobson, A.P.(2007). Predicted impacts of climate changeand land-use change on the global biodiversityof birds. PLoS. Biol. 5, 12111219.

15. Manne, L.L., and Pimm, S.L. (2001). Beyondeight forms of rarity: which species arethreatened and which will be next? Anim.Conserv. 4, 221230.

Nicholas School of the Environment andEarth Sciences, Room A301 LSRC building,Box 90328, Duke University, Durham,North Carolina 27708, USA.E-mail: StuartPimm@aol.com

DOI: 10.1016/j.cub.2007.11.055

logy: Celltep beyond

quivalent of the blastopore, is a criticalniote embryos. Medio-lateral cell

pathway play a role during this earliest.

embryo and is considered to befunctionally equivalent to theamphibian blastopore. During amniotegastrulation, cells from the superficialepithelial layer, the epiblast, ingressventrally throughout the streak toform the two internal germ layers:the mesoderm and the endoderm.

Biodiversity: Climate Change or Habitat Loss Which Will Kill More Species?References