Effect of the covariates lake, sex, and treatment on survival.

The first three columns (“Proportional hazards “) provide model diagnostics of the employed Cox proportional-hazards regression, i.e. they test the proportional hazards assumption for the model fit. The non-significant p-values of the statistics indicate that the assumption is not violated. The fourth column (coef [δ]) provides the model estimates for the three coefficients and the fifth column the respective standard errors (SE). All tested covariates had a significant effect (p-values in the sixth column).