Better environmental data for BCI

[jarioksa]

We have now BCI.env data where almost the only useful piece of information are the UTM coordinates of sample plots. There are some other environmental variables, but these are constant or almost constant in all plots. These environmental data are based on Pyke et al (2001) J. Veg. Sci 12, 553-566. However, there is another paper from the same year with a more fine-grained environmental data by Harms et al. (2001) J. Ecol 89, 947-959. Actually, these papers have conflicing information: Pyke et al. report constant altitude, but Harms et al. have some 30 m altitude range. The Harms et al. data are given on 20x20 m subplots, and these should be translated to the 1 ha level. The data should be extracted from a map.

[jarioksa]

Data derived from Harms et al. (2001) was added in 45401c1. The variables added were the dominant (one) habitat type of each plot (ignoring streams), presence of streamside habitats in plots, and habitat type diversity of each plot. I found out that in cca the dominant habitat type gave very similar constraints as seven habitat type areas as separate variables (and one of these would be aliased as they sum up to 1 ha or 25 grid cells). I have the original grid data in BCI.env2 data in natto.