8. Tree demography along environmental gradients

First meeting 10 - 12 April 2006 held at Landcare Research New Zealand, Lincoln	PARTICIPANTS Organizers Peter Bellingham, Landcare Research NZ, and David Coomes, Cambridge University Richard Barker, University of Otago, Dunedin, NZ Charlie Canham, Institute of Ecosystem Studies, Millbrook, NY, USA Richard Duncan, Lincoln University and Landcare Research, Lincoln, NZ Bob Monserud, USDA Forest Service, Portland, OR, USA Sabrina Russo, University of Cambridge, UK

This working group is gathering together datasets where demography has been quantified in different parts of the range of a plant species, and especially, along physical gradients from one boundary to the other of a plant species, and preferably over long periods of time. The background to this is that demography is sometimes thought of as a property or trait of a species, that can be compared across different species, e.g. Silvertown et al. (1993). But this can not be right. Rather, demography is an outcome of plant traits operating in a given physical and biological environment. Demography must shift near the edges of ranges, with lambda falling below 1 as one passes the boundary. So therefore, the reason for being interested in demography is not as a means for comparing species directly, but rather as a means for understanding how the interplay of traits with environment translates into geographical distribution.

Given a collection of datasets where the changes in demography within species along gradients have been quantified, here are some significant questions that can be asked:

1. As one approaches the boundary of a distribution, is there consistency in which aspects of demography deteriorate so that the species can no longer sustain a population? For example, is it that seedlings fail to establish? Or that seed production declines and no longer outweighs seedling mortality? Or that adult mortality increases? Or that growth rates decline but mortality rates per time are unchanged, so that mortality per size-transition increases?

2. Given (as will surely be the case) that the answer to Q1 is not wholly consistent across all boundaries and all species, how are the answers patterned according to whether we are considering a physical-stress boundary versus a competitiveness boundary, or according to the life-history or other features of the plant species?

3. In recent years it has become standard to argue that rare and extreme events are very important for determining ranges and relative competitiveness of species. On the other hand, it remains quite possible that this is not really the case, and that the year-in-year-out carbon surplus is really the key. Where we have demographic studies that are long-term as well as distributed along gradients, what light do these shed on this question?

4. Considering what aspects of demography start to deteriorate most sharply near boundaries, what can be said about the kinds of traits that would be enlightening to consider, for predicting what sorts of species would be most successful in different places along a physical gradient?

The first meeting of the working group has set in train three activities:

1. Addressing challenges in estimating demographic rates from large-scale forest inventory data, often collected for other purposes
2. Improving methods for calculating demographic rates for trees given issues of censored data, irregular measurement periods, etc.
3. Evaluating changes of tree demographic rates of widespread species along temperature gradients.

No date has been set yet for reconvening the group, but parts of the group will meet during May-August 2006 to progress a paper combining these three activities..