Terry, Great to hear from you! Because this problem obviously has wide interest (judging from the replies I received within one hour of sending out my intial plea for help), I am also responding through forest@nic.funet.fi. What you suggest (and guessed that we might already be doing) is correct, but it does not work on species for which reliable bark functions are unavailable, viz. the great majority of species in our natural and man-made forests in Australia and, I strongly suspect, the world. I am particularly interested in improving inventory in tropical forests using, for example, centroid sampling. For this to succeed, we need to do better in predicting bark thickness at the centroid position than using, as I presently do, one of the three generic functions of Grosenbaugh. My wish is to avoid using prediction functions altogether (and the possibility of inherent bias)! We should be developing a methodology which can be applied universally irrespective of species, location, whatever. An instrument that can measure bark thickness remotely at any position on the bole is what we need! Is it physically feasible? Who has a solution? Geoff Wood Geoff Wood, email: Geoff.Wood@anu.edu.au Forestry Dept Australian National University Canberra ACT 0200 Australia
Mail converted by
MHonArc 1.1.0