bleaching may ultimately help reef corals to survive the recurrent and increasingly severe warming events projected by current climate models of the next half-century 3 . Bleaching is an ecological gamble in that it sacrifices short-term benefits for long-term advantage. This counters conventional wisdom that bleaching is detrimental from all perspectives, and supports the role of symbionts as adaptive agents 10,11 . Andrew C. Baker Wildlife Conservation Society, Osborn Laboratories of Marine Science, New York Aquarium, Surf Avenue at West 8th Street, Brooklyn, New York 11224, USA e-mail: abaker@wcs.org 1. Brown, B. E. Coral Reefs 16, S129–S138 (1997). 2. Glynn, P. W. Coral Reefs 12, 1–17 (1993). 3. Hoegh-Guldberg, O. Mar. Freshwat. Res. 50, 839–866 (1999). 4. Rowan, R. & Powers, D. A. Science 251, 1348–1351 (1991). 5. Glynn, P. W. & Colgan, M. W. Am. Zool. 32, 707–718 (1992). 6. Rowan, R. & Knowlton, N. Proc. Natl Acad. Sci. USA 92, 2850–2853 (1995). 7. Rowan, R., Knowlton, N., Baker, A. & Jara, J. Nature 388, 265–269 (1997). 8. Baker, A. C. The Symbiosis Ecology of Reef-Building Corals. Thesis, Univ. Miami (1999). 9. Toller, W. W., Rowan, R. G. & Knowlton, N. Biol. Bull. Mar. Biol. Lab. W oods Hole (in the press). 10. Buddemeier, R. W. & Fautin, D. G. Bioscience 43, 320–326 (1993). 11. Rowan, R. J. Phycol. 34, 407–417 (1998). 12. Kinzie, R. A., Takayama, M., Santos, S. R. & Coffroth, M. A. Biol. Bull. Mar. Biol. Lab. W oods Hole 200, 51–58 (2001). 13. Buddemeier, R. W. & Smith, S. V. Am. Zool. 39, 1–9 (1999). 14.Wilkinson, C. et al. Ambio 28, 188–196 (1999). Population control African elephants and contraception P rotected from hunting and provided with access to water-holes during droughts, elephant numbers can dou- ble in a decade, severely damaging natural vegetation and the many species dependent upon it. Culling is an effective but contro- versial control strategy, so Fayrer-Hosken et al. 1 have assessed the efficacy of using immunocontraception through vaccina- tion, concluding that this could be a practi- cal way of controlling elephant numbers. However, an intervention feasible in repro- ductive physiology may not be a practical way to control a population. Fayrer-Hosken et al. have not considered calculations 2,3 that undermine the practicality of their method, nor alternative management strategies. Controlling elephants in Kruger National Park, South Africa, by immuno- contraception would necessitate treatment of 2,250 cows each year over an initial period of 11 years (ref. 3). Even if individual treatments were 100% effective, the costs would be likely to exceed the total manage- ment budget of the South African national parks. The best results of Fayrer-Hosken et al. involved two of ten elephants becoming pregnant, and that was after receiving two booster vaccinations. The effectiveness of this method may be less than claimed. Of the control group, 89% became pregnant within a year. This seems high, exaggerating the difference between treated and control groups. Data from 813 adult cows culled in Kruger National Park between 1979 and 1994 showed that 51% (range, 36–77%) were pregnant. This is to be expected: gestation lasts 22 months and the calving interval is 44 months (ref. 2), so about 50% of a sample of cows should be pregnant. Thus, on average, females go for 22 months with- out becoming pregnant. In a random sam- ple of females monitored for 12 months, only 55% (not 89%) should therefore become pregnant. Between 16 and 1,846 elephants of all age classes and both sexes were culled annu- ally in Kruger National Park from 1967 to 1994. We share the desire to reduce culling and have sought methods to do so. Remov- ing or sterilizing 250 subadult females each year should reduce population growth to zero 2,3 . Moreover, densities of greater than 0.37 elephants per square kilometre result in reduced population growth rates — probably due to reduced reproductive out- put by newly sexually matured females or to increased calving intervals 2 . Culling, as conducted, maintained densities at which population growth was near its maximum. Culls should be delayed for one year after counts exceed 0.37 elephants per square kilometre to allow density dependence to reduce numbers naturally 2 . Culls may still be necessary, but they would then be much less frequent and involve far fewer animals. Stuart L. Pimm*, Rudi J. van Aarde† *Center for Environmental Research and Conservation, Columbia University, 1200 Amsterdam Avenue, New York, New York 10027, USA e-mail: stuartpimm@aol.com †Conservation Ecology Research Unit, University of Pretoria, Pretoria 0002, South Africa 1. Fayrer-Hosken, R. A., Grobler, D., Van Altena, J. J., Bertschinger, H. J. & Kirkpatrick, J. F. Nature 407, 149 (2000). 2. van Aarde, R. J., Whyte, I. & Pimm, S. Anim. Conserv. 2, 287–294 (1999). 3. Whyte, I., van Aarde, R. J. & Pimm, S. Anim. Conserv. 1, 77–83 (1998). Fayrer-Hosken et al. reply –– Pimm and van Aarde question the feasibility of controlling elephant numbers by immunocontracep- tion, arguing that the sterilization or removal of 250 subadult cows each year is the answer to population growth. However, there are no known safe methods of sterilizing free-roaming African elephants. Moving 250 subadult females to another park is impractical as there are very few areas able to receive elephants from Kruger National Park without becoming confront- ed with an elephant overpopulation prob- lem of their own. This number of subadult cows cannot be moved without disrupting the social order within their herds. Keeping them in their herds would mean that (assuming a mean herd size of 12.4, as shown in our study, and an average of 3 subadult females per herd) Kruger National Park would have to move 1,033 elephants — an unrealistic and expensive proposition. Hence the only practical way to remove 250 subadult females would be to cull them, which Pimm and van Aarde agree is an unaccept- able solution. We have shown that immunocontracep- tion using porcine zona pellucida (pZP) works in the African elephant, although its long-term effectiveness in controlling pop- ulations is still being evaluated in South Africa. The cost and speed of field delivery have not been assessed for vaccinating large groups of elephants. However, contrary to the calculations of population modellers 1,2 , immunocontraception has worked in herds of wild horses and white-tailed deer 3 . Preserving these magnificent creatures and their genetic contribution for the future is a common goal. On the basis of a single administration of a multiple-release pZP vaccine that is being developed for use in horses (I. K. M. Liu, personal communi- cation), it should be possible to reduce the first three vaccinations used in our original study to a single dose and so minimize the stress, cost and labour of elephant immunocontraception. We therefore question Pimm and van Aarde’s criticism regarding the practicality of field immunocontraception for Kruger Park’s elephant herds. It is our judgement that preserving these animals through immunocontraception is a realistic strategy that would save elephants without having to kill them. R. A. Fayrer-Hosken*, D. Grobler, J. J. van Altena, H. J. Bertschinger, J. F. Kirkpatrick *College of Veterinary Medicine, University of Georgia, Athens, Georgia 30602-7385, USA e-mail: rfh@calc.vet.uga.edu 1. van Aarde, R. J., Whyte, I. & Pimm, S. Anim. Conserv. 2, 287–294 (1999). 2. Whyte, I., van Aarde, R. J. & Pimm, S. Anim. Conserv. 1, 77–83 (1998). 3. Kirkpatrick, J. F., Turner, J. W., Liu, I. K. M., Fayrer-Hosken, R. A. & Rutberg, A. T. Reprod. Fertil. Dev. 9, 105–110 (1997). brief communications 766 NATURE | VOL 411 | 14 JUNE 2001 | www.nature.com brief communications is intended to provide a forum for both brief, topical reports of general scientific interest and technical discussion of recently published material of particular interest to non-specialist readers. Priority will be given to contributions that have fewer than 500 words, 10 references and only one figure. Detailed guidelines are available on Nature’s website (www.nature.com) or on request from nature@nature.com © 2001 Macmillan Magazines Ltd