Here's the article in a nutshell: the edges of tropical forest fragments are exposed to wind, more sunlight, and lower humidity and the modifies the vegetation in several ways. The canopy becomes more open because more trees blow down and the understory becomes thicker because there is more light. There are more dead leaves on the ground because - well that part I'm less sure about - but there is a dense leaf litter layer. The vegetation is the habitat and the ground-foraging birds I worked on are particularly sensitive to vegetation (why? - that's an interesting question!) so this modified vegetation in fragments may make them unusable to the study birds. To test this hypothesis, I sampled bird habitat (>20 vegetation variables) by looking for foraging birds and I compared these sites to random sites in the interior of fragments - the "best" the fragment has to offer. Ends up that we found differences between bird habitat and fragment habitat and, as expected, the smaller the fragment the larger the differences.
My favorite part of this article is that we didn't use p-values! We took a Bayesian approach and report the probabilities of our predictions being correct. The coding was a bit tedious (because I don't know crap about it) but I love the results.
We did all of our work at the Biological Dynamics of Forest Fragments Project, just north of Manaus Brazil. I had a ton of support and I have to really thank my advisor, Phil Stouffer, who took a chance on me in 1993. This is our fourth paper together (two in Conservation Biology, one in Biological Conservation and one in J. Field Ornithology). I had a ton of help in the field from Flecha. This was a terribly tedious project - we counted every plant taller than 2 m in a 16 m diameter circle. It became a game to call out the dbh of a tree before we measured it. He also brought me breakfast every day (the trade off was that he didn't have to come out to the field until 930... I was there at 530).
|Edge of a 100-ha forest fragment.|