Dr. Ishii - Mix N' Match Summary

At present, I believe there is not an immediate need to visulalize or organize all the datasets that are available at the various structural scales. Doing so at this stage, I think, would be mistaking the method for the objective. I think mixing and matching of datasets should be driven by a scientific need, rather than just a wish to archive.

I believe that it would be useful for datasets to be linked to each other so that each ecoloigst working at his or her respective spatial scales could extract relevant information from different datasets as the need arises. I think ecologists and biologists are mainly interested in understanding the processes rather than just documenting or archiving the data. For example, in my case, I work mainly with crown, branch-, and shoot-level structural data. I normally work with them separately to understand the processes that create the structure at each scale. After elucidating the processes, I look for factors at larger and smaller scales that may influence the process. At the crown-scale, I have found that branch growth, branch death and branch renewal by epicormic branching play important roles in development of crown structure from young to old trees of Douglas-fir. To understand what factors at larger and smaller spatial scales influence this process, I would be interested in getting light, crown position, and density information around the particular tree that I have analyzed the crown structure for. If this data could be extracted from the stand-level dataset, I would not necessarily need to look at the entire raw data for the stand, but rather local light, crown, and density information for certain trees. At the smaller scale, I analyzed the dynamics of the shoot populations within branches to show that after branches in old trees reach maximum size, distinct structural units (clusters of shoots) are born and die-back maintaining the foliage amount. The shoot-level data were organized in to these modular structural units, and into age-classes to show this. Again, I summarized the shoot-level data to make inferences to the branch level.

As illustrated in the example of my own work, I propose that, at this time, datasets be linked in such a way that needed information could be extracted at the request of the user. The databse manager need not distribute ALL the data to all users. For example, I do not need to see ALL the data at ALL the spatial scales for the 1000 year chronosequence. Rather, I would only need and would be interested in working with the crown-level measurements. After analyzing the crown data and elucidating important processes, I may request to the database manager that there be a link provided in the database, where I could calculate a local density (within a given radius of the tree in question) or local light environment for the trees that I have analyzed, because I think these are important factors that influence crown strucutural development.