What is the difference between main and total stochastic contributions?

The Main contribution is what you'd get if you fit a linear response surface . The Total contribution includes the interaction terms (e.g.  x₁*x₂  and x₁*x₁) as well.

The effect of a variable can be described in terms of its main or total effect. The main effect of a variable is computed as if it was the only variable in the system, while the total effect considers the interaction with other variables as well. The advantage of using the total effect is that the interaction terms, which can be significant, are included. By default, the total effects are displayed in the Sensitivities Plot in the LS-OPT Viewer.

For linear systems, the main and total effects are therefore the same. The second order effects must be computed, which increases computational costs considerably.