Without a reprex. See the FAQ it's not clear if the dependent variable is continuous, binary or categorical. Interpretation would be different.
The vignette has extensive worked examples
vignette("DHARMa", package="DHARMa")`
that show iterative model development to address misspecification problems. General remarks set the stage for understanding residuals, recognizing over/underdispersion through the plots and using parametric and non-parametric tests, the special case of zero-inflation, heteroscedasticity, detecting missing predictors and incorrect functional assumptions, distance-dependence in the residuals, and the adjustments needed to analyze the cases of Poisson, proportional and binomial data.
While the diagnostics provided are very powerful, they also require clear understanding of the underlying statistical principals.