The vegetation height in forest ecosystems is spatially heterogeneous.
Gaps in the forest canopy are formed when trees fall, and the canopy
can recover by growth of small trees or extension of branches of
adjacent trees. In this talk I present the results of a study of the
spatial patterns which is based on stochastic lattice models.
In the first part of the talk I will investigate how well a simple
two-state model can describe the patterns of gaps. In the case of the
25.25 ha deciduous forest plot in the Ogawa Forest Reserve, we will
see that this model, which corresponds to the Ising model of
statistical mechanics, is insufficient to understand the patterns at
the local scale.
In the second part of the talk I will focus on multiple scales. I
compare the two-state model to a three-state model and a
“propagating-wave model” with a continuous range of states, by
considering the variance of the gap cover as a function of the scale,
and measuring the deviation of this function from a power law. The
two-state and three-state models are similar in this respect, and both
are close to forest data. The propagating-wave model shows a clear
deviation from the power law.
In the third part of the talk I will introduce a method to detect the
direction of propagating waves of disturbance and regeneration. This
allows us to distinguish models with a global direction (Shimagare
model), models with a local direction (propagating-wave, three-state
model), and models without direction (two-state model).
|