Molecular sexing in Macquarie perch using PCR-RFLP
The sexing assay (PCR product bounded by F1-R1 primers digested withAse I at the Y-specific scaffold 633 allele T at base 93229) tested in two males and two females with known genotypes showed a single band in females, and three bands in males, as expected (Appendix C; Figure C2). A larger sample of 10 females and 10 males from Dartmouth/Yarra also showed a single band in females, but only two short cut fragments of Y in males, consistent with Y0 genotype (Figure C3). The lack of the uncut X-specific band in five of these males was unexpected, as their WGS-based genotypes showed them as heterozygotes for the cut-site. The same assay tested in known-sex Macquarie perch from King Parrot Creek/Holland’s Creek and Abercrombie populations yielded 2- or 3-band pattern in all males, but also in 20% of King Parrot Creek females and 60% of Abercrombie ones; the remaining females showed a single band, as expected (Figure C4). Because phenotypic females with a male-specific allele could potentially represent sex-reversed genotypic males (Baroiller, D’Cotta, & Saillant, 2009), we explored fertilization rates data for ten phenotypic Abercrombie females that yielded bands on the PCR-RFLP assay. Females with a male-specific allele (i.e. potentially feminized males) appear to have had eggs with lower fertilization rate (Supplementary Material 1). Of four females that produced a single female-like band, two produced eggs that were successfully fertilized in the hatchery, and two had eggs that failed to be fertilized. Of six females that produced two bands (male-like), four produced eggs that were not fertilized, one showed an egg-fertility rate of <1%, and one had eggs that were successfully fertilized. Further tests of the sexing assay in Macquarie perch of unknown sex from six additional populations yielded a single band in 34% of individuals, and two or three bands in 66% of individuals (Figure C5). Therefore, the PCR-RFLP sexing assay targeting Y-allele at 93299 is ~100% accurate only for the populations used in assay development, Dartmouth and Yarra (except it could missex ~2% of males as females, as one male out of 50 showed a female-like genotype; Fig. 1), but is not expected to work well in other populations of Macquarie perch due to presence of the Y-allele in phenotypic females.