Michael, I wish you the best of luck with your thesis. It is hard for me to tell whether any of your questions are ironic or rhetorical in nature, so I am taking them all seriously. Forgive me if I miss a smile. >Often tree growth models postulate an optimal growth function that is >reduced by environmental conditions. The implicit underlying >assumption is that a tree living alone (without other trees in >the immediate vicinity) grows optimal, whereas a tree in a forest is >to some extent suppressed in its growth. Or with other words: A tree >doesn't "like" to live in a forest. Not knowing exactly which models you are referring to, it is difficult to reply meaningfully, but the assumption *may* be more along the lines of a tree growing optimally under certain environmental conditions, which it is forced to share with other trees (because it can't kill them and trees don't walk away from a fight). Thus it might be the conditions driving the model rather than the trees. Trees could be seen as a physical expression of the conditions. Thus optimality doesn't really matter, as there are no decisions to be made by the trees. >Why has evolution not developed single trees? e.g. trees with >chemical weapons effectively destroying any others trees in its >surrounding + far-flying seeds. There are many examples of inhibitory mechanisms being used in the ineraction of trees with other trees and other life-forms - e.g. chemical allelopathy. I have just referred to Kimmins (1987) Forest Ecology, and I am sure you will find it in many texts. >It seems to me that for a tree living in a forest has some >advantages. And these advantages should somehow be incorporated in a >tree and stand growth model as a positive effect of other >interacting trees. I think that in those models which incorporate these processes do not necessarily measure the processes outside the forest - therefore by default the environmental conditions are those you mention, and there is no modelling of the "improvement" of growing conditions created by the forest. Furthermore, insofar as every model I have seen has involved, if not outright fudge factors, then data-dependent parameter estimates, the 'optimality' of the growth conditions under the model are really just expressions of the environmental conditions at the time. I have never seen a model which paid strict attention to the 'optimal' behaviour because they're all (correctly, in my opinion) more interested in typical behaviour. However, any model which considers regeneration is, in a sense, incorporating the positive interactions you note. I assert (without proof) that the stage of growth at which a tree is most sensitive to its environment is during establishment, and then the shading, microclimate, wind protection, etc. will have the most effect. At the early stages of life there are definitely advantages to being around other trees, but as the tree ages, these lessen, and perhaps eventually become a 'disadvantage'. But the tree is already there. Unlike many other organisms, it can't go hunting for better conditions. The other consideration is that it may well be completely mistaken to model single trees, if they cannot be considered to be independent units. Perhaps a multi-stage model, which models forests, and then stands within forests, and then trees within stands, would be more appropriate. What do you think? Have you (has anyone) worked on or seen such an approach? I have seen some ideas correlated to this from Oscar Garcia, who I hope won't mind being singled out as a major individual in this area. Andrew Andrew Robinson Phone : +1 612 625 5765 (work) Biometrics Graduate Student : +1 612 649 0347 (home) Dept. of Forest Resources FAX : +1 612 625 5212 University of Minnesota, 115 Green Hall "Walk softly, 1530 North Cleveland Ave. and carry a big statistic." St. Paul, Mn., 55108-6112 USA No statements above may be inferred to have the necessary support of my employer.
Mail converted by
MHonArc 1.1.0