Sex reversal of pejerrey (Odontesthes bonariensis), a species with temperature-dependent sex determination,in a seminatural environment

This study examined the changes in sex ratios and sex reversal rates in pejerrey Odontesthes bonariensis that occur with the progression of the spawning season in a seminatural setting. Four groups of hatchery-produced pejerrey larvae were stocked in floating cages in La Salada de Monasterio lake (Pampas region), a natural habitat of this species, and reared from hatching beyond gonadal sex determination with minimum human interference. Cage 1 was stocked at the beginning of the spring spawning season and the other cages were stocked with monthly delays until cage 4 in early summer. The genotypic (amhy+, XY/YY; amhy−, XX) and phenotypic (testis, male; ovary, female) sex ratios and proportions of genotype/phenotype mismatched individuals were estimated and their relation to water temperature and daylength during the experiment was analysed by generalized linear modelling. Water temperature varied between 11 and 30.5°C, and daylength duration between 11 h 22 min and 14 h 35 min. Sex genotyping revealed nearly balanced sex ratios of XY/YY (46%–49.1%) and XX (50.9%–54%) fish in cages 2–4 whereas the genotypic sex ratio in cage 1 was clearly biased towards XY/YY fish (60.6%). Phenotypic males ranged from 42% to 54.4% in cages 1–3. Cage 4, in turn, had significantly more phenotypic males (66%). The percentage of XX males (phenotypic male/genotypic female) was 23.1% in cage 1, decreased to a minimum of 5.4% in cage 2 and gradually increased in cages 3 and 4 to a maximum of 40.7% in the latter. The percentages of XY/YY females (phenotypic female/genotypic male) were highest in cage 1 (30%) and decreased progressively in the other cages to a significantly lower value (4.3%) in cage 4. These results generally support the findings of laboratory studies on the effect of temperature on the sex determination of this species and also provide novel evidence of a XX genotype-specific masculinizing effect of short daylength.