Correction: 13 July 2007 www.sciencemag.org/cgi/content/full/317/5835/214/DC1

Supporting Online Material for Extraordinary Flux in Sex Ratio Sylvain Charlat,* Emily A. Hornett, James H. Fullard, Neil Davies, George K. Roderick, Nina Wedell, Gregory D. D. Hurst *To whom correspondence should be addressed. E-mail: [email protected] Published 13 July 2007, Science 317, 214 (2007) DOI: 10.1126/science.1143369

This PDF file includes: Materials and Methods Fig. S1 Table S1 References

Correction (13 July 2007): The previous version of Fig. S1 was incorrect. This revised supplement includes the correct Fig. S1.

Supporting online material Extraordinary flux in sex-ratio Sylvain Charlat1,2*, Emily A. Hornett1 +, James H. Fullard4, Neil Davies2,, George K. Roderick3, Nina Wedell5 & Gregory D. D. Hurst1 + 1

. Department of Biology, University College London, 4 Stephenson Way, London NW1 2HE, UK 2 . Gump South Pacific Research Station, University of California Berkeley, BP 244 Maharepa, 98728 Moorea, French Polynesia 3 . Environmental Science (ESPM), University of California, Berkeley, California 947203114, USA 4 . Department of Biology, University of Toronto at Mississauga, 3359 Mississauga Rd., N., Mississauga, Ontario Canada L5L 1C6 5 . School of Biosciences, University of Exeter, Cornwall Campus, Tremough, Penryn TR10 9EZ, UK *To whom correspondence should be addressed: E-mail: [email protected] + Current address: School of Biological Sciences, University of Liverpool, Crown Street, Liverpool L69 7ZB, UK Material and Methods Sampling and field sex-ratio Informal observations of field sex-ratio were made in May and June 2005 when JF traveled to Samoa. Butterflies were observed at three locations in Upolu (Apia: 13°50'S / 171°46'W; Lalomanu: 14°02'S / 171°26'W; Sa'anapu: 14°00'S / 171°53'W) and four location in Savaii (Salelologa: 13°44'S / 172°13'W; Fagamalo: 13°27'S / 172°21'W; Safotu: 13°27'S / 172°24'W; Samata'itai: 13°37'S / 172°41'W). Approximately 100 individuals were observed in total on each island. Systematic sampling was performed in April 2006. All males and females observed were collected, over the course of a single day at each of the following sites: Apia and Saleilua (14°01'S / 171°42'W) (in Upolu), Salelologa and Sagone (13°41'S / 172°38'W) (in Savaii). DNA extractions, PCR and sequencing DNA extractions and PCR procedures were as described (S1). Strain identity was initially confirmed by obtaining wsp sequences from 9 individuals (two females and two males from Upolu, one female and four males from Savaii) that were found to be identical to the wBol1 wsp sequence (genbank AB094382). PCR products were sequenced with primer 81F, after amplification with the 81F/691R primer pair (S2). We further typed the Samoan infection with the recently developed Wolbachia Multi Locus Strain Typing system (S3). The five MLST loci were sequenced on both strands following (S3) from four Samoan individuals (2 males from Savai, one male and one female from Upolu) as well as from five individuals from populations where the male-killing phenotype is still expressed (three females from Moorea and two females from Tahiti). No variation among locations was found. Sequences can be accessed in Genbank under accession numbers EF589952 to EF589956. The wBol1 strain received the ID number 40 and the sequence type 125 in the Wolbachia MLST database (http://pubmlst.org/wolbachia/).

Introgression experiments To test the hypothesis that the absent or reduced male-killing phenotype observed in Upolu and Savaii was due to the evolution of suppressor genes in the host, we introgressed cytoplasmic lineages from these two islands with nuclear genes from the island of Moorea, where host suppressor genes have never been observed and malekilling occurs (S1, S4). If absent or reduced male-killing in Samoan lines is caused by the presence of a recently evolved nuclear suppressor, we hypothesized that the Samoan Wolbachia should reveal male-killing ability on a non-suppressor Moorean genetic background. We mated a series of randomly sampled F1 Samoan females to uninfected Moorean males, and observed egg hatch rate. As expected, a clear shift of the distribution toward hatch rates values closer to 50% was observed (Fig. S1). Overall, F1 Samoan females crossed to Moorean males produced eggs with lower egg hatch rate (i.e. greater male death) than their pure bred Samoan parents (Mann-Whitney U test; Upolu: N1= 25, N2= 13, W = 574.5, p<0.01; Savaii: N1=21, N2=9, W=369.5, p<0.05). We then tested if repeated introgression of Moorea nuclear genes (devoid of suppressor) would result in the restoration of full male-killing, that is, the occurrence of all female broods. Two lines from the above were crossed to Moorea males over three generations. Full male-killing was retrieved in both cases, as predicted from the hypothesis that male-killing was suppressed by H. bolina nuclear genes in Samoa (Table S1). In one matriline, four all female broods were observed in G3 (with 6, 7, 17 and 36 adult females, respectively, and no males), together with 2 female-biased broods (respective number of females and males: 27 f / 6 m, 23 f / 12 m). In the second matriline, two all female broods were observed in G3 (composed of 7 and 17 adult females, respectively, and no males), together with one female-biased brood (producing 14 females and 2 males). Such heterogeneity among sister broods [also observed with the suppressor from South-East Asia (S5)], is expected from Mendelian segregation of a dominant single locus trait. Indeed, G2 mothers are expected to carry either one or zero copy of the suppressor allele, resulting in a mixture of female-biased and all female broods in G3, consistent with our observations. Statistical procedures We used contingency tests to detect variation of sex-ratio among broods obtained from females collected at a given location. When no heterogeneity was detected, broods were pooled within locations before locations were compared using contingency tests. At the highest level where no heterogeneity was observed (e.g. Upolu island), deviation of sexratio from 1/1 was tested using chi-square tests on pooled sex ratio data. We used MannWhitney U tests to compare egg hatch rates before and after initiation of the introgression experiment. All analyses were performed in R (S6). Supporting Tables and Figures Table S1. Sex-ratio (proportion males) of two Samoan matrilines on introgression of Moorean nuclear genes. Parental data is from Samoa x Samoa cross. G1 is the sex ratio produced on crossing daughters of these lines to Moorea males, G2 and G3 further

backcrosses of the lineage to Moorean males. n represents the number of progeny produced from a cross. Supporting References and Notes S1. S. Charlat et al., Current Biology 17, 273–277 (2007). S2. W. Zhou, F. Rousset, S. O'Neill, Proc Biol Sci 265, 509 (Mar 22, 1998). S3. L. Baldo et al., Appl Environ Microbiol 72, 7098 (Nov, 2006) S4. S. Charlat et al., Mol Ecol 14, 3525 (Oct, 2005). S5. E. A. Hornett et al., PLoS Biol 4, e283 (2006). S6. R Development Core Team (2005). R: A Language and Environment for Statistical Computing, 2.2.1 Edition (Vienna, Austria: R Foundation for Statistical Computing).

Table S1 Parental Matriline SR A

n

0.70 10

G1 SR

G2 N

0.35 17

SR

G3 n

0.19 21

SR

n

0.00 7 0.00 6 0.00 36 0.00 17 0.18 33 0.34 35

B

0.47 15

0.39 18

0.38 13

0.00 7 0.00 17 0.13 16

Figure S1. Distribution of egg hatch rates produced by Samoan wild caught females (in light gray) and daughters from these crosses mated to Moorean males (in dark gray).