factor importance in glmm with a two-way ANOVA structure

Hi

I'm running a glmm with negative binomial family, because my dataset
has many zeros, yet, they are not structural zeros.

The model design is a 2-way ANOVA: habitat_type * season

In a glm, I can assess the relative importance of each factor
through the deviances (null and residual deviances of the model, and
partial deviances accounted by each factor).

The analogous way in a LMM would be the Sum of Squares.

But, how can I assess the performance (or variance accounted) by
different factors in a glmm?

I've been searching a lot in internet but I haven't found anything
clear about it.

I hope you can provide some advice on it.

Thanks!

Jp