20. Ru ¨hm, W. et al. The dosimetry system DS86 and the neutron discrepancy in Hiroshima—historical review, present status, and future options. Radiat. Environ. Biophys. 37, 293–310 (1998). 21. Arakatsu, B. Collection of Investigative Reports on Atomic Bomb Disaster 34–35 (Science Council of Japan, Tokyo, 1953). 22. Yamasaki, F. & Sugimoto, A. Collection of Investigative Reports on Atomic Bomb Disaster 18–19 (Science Council of Japan, Tokyo, 1953). 23. Gritzner, M. L. & Woolson, W. A. in US-Japan Joint Reassessment of Atomic Bomb Radiation Dosimetry in Hiroshima And Nagasaki-Final Report Vol. 2 (ed. Roesch, W. C.) 283–292(Radiation Effects Research Foundation, Hiroshima, 1987). 24. Barnes, I. L., Garfinkel, S. B. & Mann, W. B. Standardization of 63 Ni by efficiency tracing. Int. J. Appl. Radiat. Isot. 22, 781–783 (1971). 25. Marchetti, A. A. & Straume, T. A search for neutron reactions that may be useful for Hiroshima dose reconstruction. Appl. Radiat. Isotop. 47, 97–103 (1996). 26. Straume, T., Marchetti, A. A. & McAninch, J. E. New analytical capability may provide solution to the neutron dosimetry problem in Hiroshima. Radiat. Prot. Dosim. 67, 5–8 (1996). 27. Shibata, T. et al. A method to estimate the fast-neutron fluence for the Hiroshima atomic bomb. J. Phys. Soc. Jpn 63, 3546–3547 (1994). 28. Johnson, J. O. (ed.) A User’s Manual for MASH v.2.0–Monte Carlo Adjoint Shielding Code System Report ORNL/TM/11778/R2 (Oak Ridge National Laboratory, Oak Ridge, TN, 1999). 29. Rose,P. F. (ed.) ENDF/B-VI Summary Documentation, Cross-Section Evaluation Group Report BNL- NCS-17541 (Nuclear Data Center, Brookhaven National Laboratory, Upton,NY, 1991). 30. Feldman, G. J. & Cousins, R. D. Unified approach to the classical statistical analysis of small signals. Phys. Rev. D 57, 3873–3889 (1998). Acknowledgements We thank I. Proctor and A. M. Kellerer for their essential support of this project. We also thank S. Fujita and K. Shizuma for providing copper samples for this study. We thank the following organizations for supporting this work: the US Department of Energy, the US National Academy of Sciences, the European Commission, the German Federal Ministry of Environment, Nature Conservation and Nuclear Safety. Competing interests statement The authors declare that they have no competing financial interests. Correspondence and requests for materials should be addressed to T.S. (Straume1@aol.com). .............................................................. Niche lability in the evolution of a Caribbean lizard community Jonathan B. Losos 1 , Manuel Leal 2 *, Richard E. Glor 1 , Kevin de Queiroz 3 , Paul E. Hertz 4 , Lourdes Rodrı´guez Schettino 5 , Ada Chamizo Lara 5 , Todd R. Jackman 6 & Allan Larson 1 1 Department of Biology, Campus Box 1137, Washington University, St. Louis, Missouri 63130, USA 2 Department of Biological Sciences, Union College, Schenectady, New York 12308, USA 3 Division of Amphibians and Reptiles, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560, USA 4 Department of Biology, Barnard College, 3009 Broadway, New York 10027, USA 5 Instituto de Ecologı ´a y Sistema ´tica, CITGMA, Carretera de Varona km 3.5, Boyeros, La Habana 10800, Apartado Postal 8029, Cuba 6 Department of Biology, Villanova University, Villanova, Pennsylvania 19085, USA *Present address: Department of Biological Sciences, Vanderbilt University, VU Station B 351634 Nashville Tennesee 37235, USA ............................................................................................................................................................................. Niche conservatism—the tendency for closely related species to be ecologically similar—is widespread 1–3 . However, most studies compare closely related taxa that occur in allopatry 3 ; in sympatry, the stabilizing forces that promote niche conservatism 4,5 , and thus inhibit niche shifts, may be countered by natural selection favouring ecological divergence to minimize the intensity of interspecific interactions 6,7 . Consequently, the relative import- ance of niche conservatism versus niche divergence in determin- ing community structure has received little attention 7 . Here, we examine a tropical lizard community in which species have a long evolutionary history of ecological interaction. We find that evolutionary divergence overcomes niche conservatism: closely related species are no more ecologically similar than expected by random divergence and some distantly related species are ecolo- gically similar, leading to a community in which the relationship between ecological similarity and phylogenetic relatedness is very weak. Despite this lack of niche conservatism, the ecological structuring of the community has a phylogenetic component: niche complementarity only occurs among distantly related species, which suggests that the strength of ecological inter- actions among species may be related to phylogeny, but it is not necessarily the most closely related species that interact most strongly. Anolis lizards are a dominant component of Caribbean ecosys- tems (reviewed in refs 8 and 9) and are well suited for studies of the evolution of community structure because the species on individual islands have a long history of interaction and coevolution. For example, 55 of 58 species on Cuba are endemic (the remaining three have colonized other Caribbean islands from Cuba), and most are members of large clades that have diversified on Cuba 10 . Species on many islands attain extremely high densities 11,12 , and many species— differing in ecology, morphology, and behaviour—coexist locally 8 . Interactions among sympatric species can be strong 8,9,13,14 , usually as a result of interspecific competition, although intra-guild predation may sometimes be important 15 . We studied the community structure of anoles at Soroa, Bio- sphere Preserve Sierra del Rosario, in the Pinar del Rı ´o province of western Cuba. Eleven anole species occur sympatrically at Soroa, the highest anole diversity known from any island or continental site. Of these species, ten are either widely distributed in Cuba or are members of island-wide clades of ecologically similar species (for example, the Anolis equestris group, to which A. luteogularis belongs, occurs throughout Cuba and is composed of six primarily allopatric species similar in morphology and ecology). Because the clades of Cuban anoles to which the Soroa species belong are widespread and arose within a relatively short period in the distant past 10 (Fig. 1), the sympatric clades at Soroa have probably coexisted for a long time and over a large spatial scale. Thus, these Anolis species probably evolved in the presence of the same clades with which they currently coexist, a necessary prerequisite for community coevolution. We examined ecological relationships among these species to investigate whether the community exhibited nonrandom ecologi- cal or phylogenetic structure. We measured ecological variables relevant to the three resource axes that sympatric Anolis generally partition: structural habitat, thermal habitat, and prey size 16 . Prin- cipal components analysis reveals three significant axes of ecological differentiation (Table 1; results below are qualitatively unchanged if another, nearly significant, axis is also retained). Examination of the position of species in multivariate ecological space reveals both that niche use has not been conserved and that the community is nonrandomly structured (Fig. 2). The minimal extent of niche conservatism is indicated by the weak association between phylogenetic relationship and position in multivariate ecological space: phylogenetic similarity explains less than 4% of the variation in ecological similarity among species (Mantel test, P ¼ 0.11–0.30 depending on phylogenetic topology and mode of character evolution used in the analysis; all variables but one exhibit similarly low correlations with phylogenetic relationships (Table 1); P-values in Mantel tests based on 5,000 simulations). The molecular data strongly reject alternative phylo- genetic topologies in which ecologically similar species are grouped phylogenetically (see Supplementary Information). Although some closely related species differ little ecologically, many distantly related species are just as ecologically similar, and some closely related species are ecologically dissimilar (Fig. 2). Moreover, although members of the sagrei and porcatus clades form clusters in ecological space (Fig. 2; multivariate analysis of variance, MANOVA, Wilks’ l ¼ 0.013, F 12,10 ¼ 3.79, P ¼ 0.018), they are no more ecologically similar than would be expected for letters to nature NATURE | VOL 424 | 31 JULY 2003 | www.nature.com/nature 542 © 2003 Nature Publishing Group