Canopy analysis of oak savannas and woodlands

If you have been following this blog, you know that for the past several years we have been constructing a tree database for Pleasant Valley Conservancy. We are permanently marking, measuring, and identifying each tree, and recording its position with GPS. The data are being visualized and analyzed using ArcGIS, which I have been working on (and slowly learning) for the past two years.

I have been following the lead of the silviculture literature, especially the nice book on the ecology and silviculture of oaks by Paul Johnson et al. (a new 2nd edition is now out). Among other things, Johnson and his colleagues at the U.S. Forest Service have developed a quantitative method for characterizing an oak savanna. This requires an estimation of the basal area per acre and the tree density per acre for each stand. From these data, using nomographs they have constructed, one can calculate the percent crown cover for a site.

The table below contains a summary of my calculations for the savanna, remnant prairie, and oak woodland areas of Pleasant Valley Conservancy. You need to use the Management Map provided on my web site to see exactly where each unit is, but I have added information in the table that should help even without the map.

Constructing this table turned out to be an all day job (perfect for this really frigid weather), and I ended up with a huge Excel spreadsheet that had all the data organized by management unit. (We have over 2300 trees recorded, and we are not finished yet!) Once the data from this spreadsheet were compressed, I ended up with the compact table shown below.

According to Johnson and colleagues, estimating crown cover can be used to monitor changes in an oak savanna with time. When restoring a savanna, the crown cover measurement on a savanna that has grown into a closed canopy stand can provide a baseline for use while following the restoration process. Thus, the crown cover provides a measure of the "state" of the savanna.

My canopy measurements agree quite well with the more qualitative measurements of other savanna workers in Wisconsin.

Note that you do NOT have to measure each tree to carry out this analysis. One can use sampling methods that foresters and plant ecologists have developed. Procedures for doing this are well described in the silviculture literature, and there are numerous textbook-like descriptions on web sites.

However, we have lots of other uses for our tree database.

My apologies that the table is a JPEG image. It turns out that Google Blogger does not permit direct insert of tables. To get the version you see here, I had to develop the table in Microsoft Word, print it out, and then scan it. (It will become considerably larger and more readable if you click on it.)