Delete whole cohort through starvation

[wtraylor]

Idea

Currently, cohorts are reduced through starvation mortality by a certain fraction. An alternative would be to kill the cohort completely with a probability corresponding to that starvation fraction.

Advantages:

• Reducing the number of cohorts could improve calculation time!
• Less chance for ”zombie” cohorts: those cohorts with extremely low density, where it is ambiguous whether they should just be deleted to make space for re-establishment.

Disadvantages:

• Model becomes stochastic rather than deterministic. It needs to be evaluated whether a single simulation is representative or whether a number of samples needs to be calculated.
• In particular with short-lived herbivores (with fewer cohorts), killing a whole cohort can have a lot of impact on the whole population dynamic.