Solomon: Thanks for posting this fascinating question: which of gazillions of possible climatic variables ought to be used to model species distributions? Please let us know what find out.
In the meantime, and offhand, I'd look for measures of protracted -2C episodes in the few weeks after bud-burst, or critical steps in reproduction such as seed set or germination. Plus degree day sums, probably at several thresholds (+0, +2, +5C) --- these should be corrected for day length if you're dealing with a significant latitudinal gradient. I know these indices are species specific; that's life. They also require you to work with at least daily Wx data, alas. I'd also dig into David Inouye's work. Good luck Steve Cumming Canada Research Chair (Boreal Ecosystems Modelling) D\'epartement des sciences du bois et de la for\^et Universit\'e Laval ph:418.656.2131x2593 -----Original Message----- From: Solomon Dobrowski [mailto:[EMAIL PROTECTED] Sent: Monday, June 05, 2006 7:51 PM Subject: Climate variables and time Hello All. In examining the literature on climatic predictor variables for predicting the distribution of mountain vegetation species, I have come across a huge range of variability in the temporal characteristics of these variables. For example, for mean temperature, I have seen studies use mean annual, mean July, mean between April and October, Degree Days, and countless other permutations. The same holds true for min and max temperatures and to a lesser extent precipitation. Lets assume that I have data summarized on the daily time-step. What is the consensus (if there is one) on the optimal temporal integration to use for montane and sub-alpine species? Along the same lines, does anyone know of papers that address this topic. Thanks a bunch. Solomon Solomon Dobrowski Tahoe Environmental Research Center (TERC) John Muir Institute of the Environment University of California, Davis 530 754 9354
