Discussion with Kevin for @gabrielgellner to think about:
Winder if you simply show a rapid pulse and it's omnivory response under passive and density dependent preference as a "purely bottom-up" driven dynamic omnivory case that can occur very naturally (eg passive no behavior even works).
And then do a no pulse under 3 K_ave and run a long run in each case and measure long term ave omnivory and top heaviness to show this is definitely driven to some extents by cascading top down influence.
Actually latter one just increase attack rates between C and R and P and C and so no bottom up signal at all. And for latter could do CV.
The problem with K is it is both bottom and top down so interaction strength cleaner.
Discussion with Kevin for @gabrielgellner to think about: