Influence of sexual genotype on agonistic behaviors and sex steroid levels of phenotypic males and females in the Nile tilapia (Oreochromis niloticus)

Mechanisms of sex determination and differentiation are extremely labile in fish, as demonstrated by the numerous sex reversal experiments performed on teleosts. In Nile tilapia, sex reversal processes using exogenous sex steroids allow to produce individuals with atypical sexual genotypes that constitute major tools to investigate the mechanisms of sex determination and differentiation, from gonad differentiation to sexual differentiation of brain and behavior.
The aim of this study was to assess the influence of sexual genotype and the role of circulating sex steroids on the expression of agonistic behaviors in Nile tilapia breeders.
Observations were carried out on fights staged between one male (M) and one female (F) acclimatized in 250-L aquaria at 27°C. Agonistic behaviors were recorded (for 25 min) in five crosses (6 repetitions with different pairs): MXY×FXX (control), MXY×FXY, MXY×FYY, MXX×FXX and MYY×FXX. Quantified behaviors were: fin raising, throat swelling, chasing, lateral attack, frontal display, tail beating, mouth fighting and biting. For steroid assay, blood was sampled on 10 individuals of each phenotype/genotype combination. Testosterone (T), 17beta-estradiol (E2) and 11-ketotestosterone (11KT) were measured by radioimmunoassay.
Expression of aggressive behaviors was significantly higher in couples with a XY or YY female than in MXY×FXX. The mean durations (in % of total time) of threatening behavior expression were respectively 41 ± 3 and 8 ± 1 %; and the mean frequencies of attacking behaviors were 110 ± 7 and 25 ± 5 n h-1. Expression level of agonistic behaviors in MXY staged with FXY or FYY seems to be adjusted to the aggressiveness level of females. Aggressiveness level was low and similar in MXY×FXX, MXX×FXX and MYY×FXX crosses. When comparing males together in these 3 crosses, only MXX showed a slightly but significantly higher expression of aggressive behaviors.
Compared to normal MXY, MXX had significantly higher levels of circulating 11KT (16.0 ± 4.1 and 26.5 ± 4.2 ng mL-1 respectively), that could be related to their higher aggressiveness. However, no similar difference was reported between females. E2 concentrations were similar between males (mean: 4.0 ± 0.4 ng mL-1) and increased in females with the presence of Y chromosome(s) (FXX: 6.5 ± 1.4, FXY: 9.5 ± 1.7 and FYY: 14.1 ± 2.2 ng mL-1). These results raised the question of an involvement of E2 in the control of agonistic behaviors in females. No influence of the genotype was observed on T levels.
Our results suggest that the presence of a Y chromosome increases aggressiveness in females. However, since the same relationship between aggressiveness and the Y chromosome is not observed in males, in which the level of aggressiveness is paradoxically higher in XX, we can hypothesize that the differences in aggressiveness are not directly dependant on the genotype but on the sex reversal procedures which young fry were exposed to during their sexual differentiation. These hormonal treatments could have permanently modified the development of the brain and consequently influenced the behavior of adults independently to their genotype. The role of endogenous steroids in agonistic behaviors needs further clarification.