Larval survival (56%) after four days offeeding on leaves dusted with Btpollen wassignificantly lower than survival either onleaves dusted with untransformed pollen oron control leaves with no pollen (both100%, Pǃ0.008) (Fig.1a). Because therewas no mortality on leaves dusted withuntransformed pollen, all of the mortalityon leaves dusted with Btpollen seems to bedue to the effects of the Bttoxin.
There was a significant effect of cornpollen on monarch feeding behaviour(Pǃ0.0001) (Fig.1b). The mean cumula-tive proportion of leaves consumed perlarva was significantly lower on leaves dustedwith Btpollen (0.57DŽ0.14, Pǃ0.001) andon leaves dusted with untransformed pollen(1.12DŽ0.09, Pǃ0.007) compared withconsumption on control leaves withoutpollen (1.61DŽ0.09). The reduced rates oflarval feeding on pollen-dusted leavesmight represent a gustatory response of thishighly specific herbivore to the presence ofa ‘non-host’ stimulus. However, such aputative feeding deterrence alone could not explain the nearly twofold decrease in
214
ae100avral h75cran)om%(50 fo lav25ivruS01234Time (days)bn2oitpmus1.5noacv rfaale r1l eepvital0.5umuC01234Time (days)Figure 1Survival and leaf consumption of second-to third-instar monarch larvae on each of three milk-weed leaf treatments: leaves with no pollen (lightblue), leaves treated with untransformed corn pollen(green) and leaves dusted with pollen from Btcorn(dark blue). a,Mean (DŽs.e.m.) survival based on theproportion of larvae surviving in five replicates ofeach treatment. b,Mean (DŽs.e.m.) cumulative leafconsumption based on the total amount of leaf areaconsumed per larva in five replicates of each treat-ment. The amount of leaf area consumed per larvain each experimental unit was calculated for eachtime interval by dividing the amount of leaf areaconsumed in that interval by the number of larvaealive during the time interval. Cumulative consump-tion was calculated by summing the leaf area con-sumed per larva at each interval. Colours of linescorrespond to those of the bars in a.
consumption rate on leaves with Btpollencompared with leaves with untransformedpollen (Pǃ0.004).
The low consumption rates of larvae fedon leaves with Btpollen led to slowergrowth rates: the average weight of larvaethat survived to the end of the experimenton Bt-pollen leaves (0.16DŽ0.03g) was lessthan half the average final weight of larvaethat fed on leaves with no pollen(0.38DŽ0.02g, Pǃ0.0001).
These results have potentially profoundimplications for the conservation ofmonarch butterflies. Monarch larvae feedexclusively on milkweed leaves4; the com-mon milkweed, A. syriaca, is the primaryhost plant of monarch butterflies in thenorthern United States and southern Cana-da5. Milkweed frequently occurs in andaround the edges of corn fields, where it isfed on by monarch larvae6. Corn fields
© 1999 Macmillan Magazines Ltd
shed pollen for 8–10days between lateJune and mid-August, which is during thetime when monarch larvae are feeding7.Although the northern range of monarchsis vast, 50% of the summer monarch pop-ulation is concentrated within the mid-western United States, a region referred toas the ‘corn belt’because of the intensity offield corn production8. The large land areacovered by corn in this region suggests thata substantial portion of available milk-weeds may be within range of corn pollendeposition.With the amount of Btcorn planted inthe United States projected to increasemarkedly over the next few years9, it isimperative that we gather the data neces-sary to evaluate the risks associated withthis new agrotechnology and to comparethese risks with those posed by pesticidesand other pest-control tactics.John E. Losey, Linda S. Rayor, Maureen E. CarterDepartment of Entomology, Comstock Hall, Cornell University, Ithaca, New York 14853, USAe-mail: jel27@cornell.edu1.Ostlie, K. R., Hutchison, W. D. & Hellmich, R. L. Bt Corn andEuropean Corn Borer(NCR publ. 602, Univ. of Minnesota, St Paul, 1997).2.Fearing, P. L., Brown, D., Vlachos, D., Meghji, M. & Privalle, L.Mol. Breed.3,169–176 (1997).
3.Raynor, G. S., Ogden, E. C. & Hayes, J. V. Agron. J.64,420–427(1972).
4.Malcolm, S. B., Cockrell, B. J. & Brower, L. P. in Biology andConservation of the Monarch Butterfly(eds Malcolm, S. B. &Zalucki, M. P.) 253–267 (Natural History Museum of LosAngeles County, Los Angeles, 1993).
5.Malcolm, S. B., Cockrell, B. J. & Brower, L. P. J. Chem. Ecol.15,819–853 (1989).
6.Yenish, J. P., Fry, T. A., Durgan, B. R. & Wyse, D. L. Weed Sci.45,44–53 (1997).
7.Brower, L. P. J. Exp. Biol.199,93–103 (1996).
8.Wassenaar, L. I. & Hobson, K. A. Proc. Natl Acad. Sci.USA95,15436–15439 (1998).
9.Andow, D. A. & Hutchison, W. D. in Now or Never: Serious NewPlans to Save a Natural Pest Control(eds Mellon, M. & Rissler, J.)19–65 (Union of Concerned Scientists, Cambridge,Massachusetts, 1998).
The mystery of female beauty
Yu and Shepard1have reported a preferencefor heavy women with high waist-to-hipratios (WHR) in a culturally isolated popu-lation in southeast Peru. Their findings areinteresting because a preference for lowWHR is widespread in westernized popula-tions2–5. However, we disagree with theirargument that cultural invariance is neces-sary for an adaptionist interpretation ofWHR preference.
WHR and waist circumference are posi-tively correlated with testosterone and neg-atively associated with oestrogen6. Womenwith low WHR have better health and fertil-ity than women with high WHR5. However,women in England and Texas with high
NATURE|VOL 399|20 MAY 1999|www.nature.com
scientific correspondence
a765No. of sons4321060708090100110120130Waist circumference (cm)mean age was 34.63DŽ7.23 s.d. years. Thewaist was measured at the narrowest por-tion between the ribs and the iliac crest,and hips were measured at the level of thegreatest protrusion of the buttocks. All sub-jects were standing and wore light outdoorclothing. Photocopies of children’s hands wereused to measure digit length (95 subjects,mean age 8.00DŽ1.43 s.d. years, were askedto place hands palm-down on the glassplaten). Digits were measured from thecrease proximal to the palm to the tip of babies than low-WHR women10). Thismight explain why traditional societies inwhich sons are valued over daughters mayprefer tubular women, which is consistentwith an adaptionist interpretation of thesepreferences.
J. T. Manning*, R. L. Trivers†, D. Singh‡,R.Thornhill§
*School of Biological Sciences,
University of Liverpool, Liverpool L69 3BX, UKe-mail: jtmann@liv.ac.uk
†Department of Anthropology, Rutgers University,New Brunswick, New Jersey 08901-1414, USAbthe digit9. Repeated measurements on the1.041.02photocopies and comparisons with X-rays)dof the hands indicated that there was a highn1.00ahconcordance with the photocopied mea- 0.98tfel(0.96surements. The mean D2/D4ratio for the oit0.94aright hand was 0.93DŽ0.03, and 0.94DŽ0.04r 40.92Dfor the left./0.902D0.88The waist circumference of mothers was0.86significantly and positively related to the0.840.550.60.650.70.750.80.850.90.951.0number of sons (Fig. 1a) but not to theMother’s WHRnumber of daughters. The relation betweena woman’s waist circumference and thenumber of her sons remained significantFigure 1Relations between female body shape andafter the effect of the age of the mother wasthe number of sons and testosterone levels in theirpartialled out (partial correlation, numberchildren. a,Number of sons and waist circumfer-of sons and waist circumference: rǃ0.18,ence in 141 Jamaican women; b,D2/D4ratio for thePǃ0.03; and age of mother: rǃ0.30,left hands of 95 Jamaican children and the WHR ofPǃ0.0005).their mothers. The lines are least-squares lines ofSimilar, but weaker, relationships werebest fit:a,yǃ0.023xǁ0.34; b,yǃǁ0.155x+1.059.present between hip circumference andWHR and the number of sons. Waist andWHR and thick waists tend to have morehip circumference and WHR were posi-sons7,8, and a preference for women withtively but not significantly related to thehigh WHR might result in selection forproportion of sons (Table 1). The Dincreased testosterone levels in children.ratio was significantly and negatively related2/D4Men in societies that prize sons over daugh-to WHR in the right and left hands; that is,ters and in which strength is an advantagewomen with high WHR tended to havemight therefore be expected to show a pref-children with low Derence for high WHR.Fig. 1b). As a low D2/D4ratios (Table 1 andWe present data from a rural Jamaicana high testosterone level in adults, this2/D4ratio is indicative ofpopulation showing that: (1) in agreementprobably means that women with highwith the data from England and Texas7,8,waist-to-hip ratios have children with highthere is a positive association between atestosterone.woman’s waist circumference and her num-Of course, our Jamaican sample is west-ber of sons; and (2) a high WHR in womenernized and we have no data on WHR pref-is associated with a marker for high testos-erences in adults. However, the positiveterone (a low ratio of the lengths of the sec-relation between waist circumferenceond and fourth digits, or Dand/or WHR and a tendency to producetheir children.2/D4; ref. 9) insons has now been demonstrated in threeOur Jamaican sample was drawn frompopulations and may well be universal.Southfield in the parish of St Elizabeth andThe high testosterone found in the chil-was part of a large long-term study of devel-dren of high-WHR women might relate toopmental stability (the Jamaican Symmetrymuscular strength (it is already known thatProject). We measured 141 women whosehigh-WHR women have heavier and tallerTable 1 Effects on offspring of WHC and WHR of Jamaican womenTraitNo. of sonsNo. ofProportionD2/D4D2/D4daughtersof sonsright handleft handrPrPrPrPrPWaist circumference0.210.010.020.790.170.052ǁ0.150.15ǁ0.180.08Hip circumference0.140.090.070.410.160.06ǁ0.020.87ǁ0.010.92WHR0.150.080.050.560.100.27ǁ0.200.048ǁ0.290.004The relations are shown between measurements of waist and hip circumference (WHC) and waist-to-hip ratio (WHR)for Jamaican women, and the proportion of sons (arcsine transformed, nǃ141 women) and the number of sons anddaughters (nǃ141 women) they have, and the D2/D4ratio in the left and right hands of their children (nǃ95 women).r, Coefficient of correlation; P, probability of wrongly rejecting the null hypothesis.NATURE|VOL 399|20 MAY 1999|www.nature.com
© 1999 Macmillan Magazines Ltd
‡Department of Psychology, University of Texas,Austin, Texas 78712, USA
§Department of Biology, University of New Mexico,Albuquerque, New Mexico 87131-1091, USA
1.Yu, D. W. & Shepard, G. H. Nature396,321–322 (1998).2.Singh, D. J. Pers. Soc. Psychol. 65,293–307 (1993).3.Singh, D. Hum. Nat.6,51–68 (1995).
4.
Furnham, A., Tan, T. & McManus, C. Pers. Indiv. Diff.22,539–549 (1997).
5.Henss, R. Pers. Indiv. Diff.19,479–488 (1995).
6.Evans, D. J., Hoffmann, R. G., Kalkhoff, R. K. & Kissebah, A. H.J. Clin. Endocrinol. Metab.57,304–310 (1983).
7.Manning, J. T., Anderton, R. & Washington, S. M. J. Hum. Evol.31,41–47 (1996).
8.Singh, D. & Zambarano, R. J. Hum. Biol.69,545–556 (1997).9.Manning, J. T., Scutt, D., Wilson, J. & Lewis-Jones, D. I. Hum.Reprod.13,3000–3004 (1998).
10.Brown, J. E. et al. Epidemiology7,62–66 (1996).
Evolutionary psychology suggests that awoman’s sexual attractiveness might bebased on cues of reproductive potential. Ithas been proposed that a major determi-nant of physical attractiveness is the ratiobetween her waist and hip measurements(the waist-to-hip ratio, or WHR): for exam-ple, a woman with a curvaceous body and aWHR of 0.7 is considered to be optimallyattractive1–3, presumably because this WHRis the result of a fat distribution that maxi-mizes reproductive potential4. It followsthat the preference for a curvaceous bodyshape in women should be universal amongmen and not be culturally based, becausenatural selection presumably favours cuesindicative of the most fertile body shape.Yu and Shepard have challenged thishypothesis5. They tested the preferences of aculturally isolated tribe of Peruvian Indians(the Matsigenka) by using a set of line-drawn figures of women who varied inapparent body-mass index (BMI) andWHR1. They claim that their results indi-cate a preference by this tribe for a tubularbody shape, rather than the curvaceousshape favoured in the United States5. How-ever, we believe that the conclusions of Yuand Shepard are undermined by a flawedassumption.
The drawings used by Yu and Shepardare arranged in three series5: underweight,normal and overweight. Within each series,the BMI of each of the four figures is sup-posed to be held constant, while the WHRis varied by narrowing the waist. However,we believe that this assumption is false
215
因篇幅问题不能全部显示，请点此查看更多更全内容