Table 2. Linear and nonlinear selection gradients.
| Traits |
γ
coefficients |
||||||
|---|---|---|---|---|---|---|---|
| Gonopodium | Body area | Iridescence | Orange | Sperm velocity | Sigmoid displays | Gonopodial thrusts | |
| Gonopodium | 0.352 (0.531) | 0.040 (0.869) | 0.944 (0.001) | −0.105 (0.605) | −0.248 (0.206) | 0.011 (0.961) | 0.242 (0.360) |
| Body area | 0.288 (0.569) | −0.103 (0.825) | 0.711 (0.068) | −0.167 (0.505) | −0.361 (0.129) | 0.190 (0.451) | |
| Iridescence | −1.826 (0.045) | −0.519 (0.128) | −0.371 (0.206) | 0.155 (0.409) | −0.149 (0.623) | ||
| Orange | −0.912 (0.137) | −0.387 (0.237) | −0.275 (0.128) | −0.167 (0.524) | |||
| Sperm velocity | −0.794 (0.100) | −0.045 (0.857) | 0.498 (0.148) | ||||
| Sigmoid displays | −0.322 (0.478) | −0.044 (0.846) | |||||
| Gonopodial thrusts | −1.486 (0.001) | ||||||
| β coefficients | 0.292 (0.033) | −0.096 (0.533) | −0.144 (0.359) | −0.026 (0.851) | −0.007 (0.959) | 0.003 (0.985) | 0.168 (0.230) |
β and γ coefficients obtained with multiple regressions. In the diagonal quadratic selection coefficients are shown and represent disruptive (+) or stabilizing (−) selection acting on trait. Above the diagonal correlational selection coefficients represent traits selected to be positively (+) or negatively (−) correlated. In parenthesis P values obtained by a multiple regression model in which male reproductive success (n=60) was the dependent variable and the standardized male traits and their quadratic terms were the independent variables44.
In bold significant coefficients.