Thanks for your question, it’s a good observation. Another observation that I think helps to explain this is that the ‘pulses’ of mortality we see are in the smaller size classes. Consequently, these pulses are related to the pulses of regenerating trees that timber harvest produces.

Remember, in our model we assume that regeneration is only occurring in gaps produced by timber harvest. That means that new trees only show up in the stand after a harvest – and they only show up in the smallest size classes in these diagrams in the second timestep after harvest (in the first timestep they are still growing with DBH < 5cm). With new trees added to these size classes, competition for resources between trees is intense and inevitably many die. In the intervening timesteps mortality is lower because no new trees have been added in that timestep and competition is less fierce.

These ‘pulses’ of mortality may seem strange but we have to remember two things:
i) this is mortality is for the entire model timestep, which is one decade
ii) the assumption that regeneration only occurs in harvest-created gaps is what is driving the behaviour.

As with all modelling, the appropriateness of the model assumptions must be questioned. In a less-heavily managed forest our ‘regeneration only after harvest’ assumption might not be pertinent. But in the forests we’re simulating timber harvest is the dominant mechanisms of gap creation – other natural mechanisms like tree-fall in windstorms are reduced because ‘weaker’ larger trees that might be most susceptible are removed during harvest to improve ‘stand quality’.

Hope this all makes sense – any more questions/observations let me know!

- James