2. Pages 6 - 10: while in the document you recognize the importance of 
peatlands (wetlands) I think that special reference to their incorporation 
into models in needed.  In several places you refer to 'forest 
biogeochemistry' and'forest carbon' models, but there is a major step 
required to get the carbon cycle of wetland into these models.  In the case 
of forests, soil moisture is important, but it is not the only or 
overriding controlling variable.  However, in wetland, defining the 
moisture regime is the primary determinant of whether the systems are large 
(and persistent) sinks for CO2 and/or large or small source of CH4.  This 
means that the hydrological component of the ecosystem model will make or 
break the success of modelling the C cycle in these systems.  How one does 
this in regional to continental scale C modelling is certainly a challenge 
- one that few are addressing and one that there seems to be no obvious answer.

3. Page 11 #3 "Enhance ... inventories ..." - this is a critical problem 
and tends not to get much attention (it is not exciting!), but to define 
the initial conditions this information is needed for most models.

4. Page 16 "The budget problem ..." : I think here it might be worth 
working in the difference of net ecosystem exchange (what all the processes 
add up to) and net biome exchange (processes plus disturbance).  This is a 
critical issue in determining over what time period one is considering the 
problem of 'resolving' the budget - no ecosystems that I know of a near an 
equilibrium so one has to place all budget considerations into the context 
of its temporal location in a disturbance/land use regime.

5. Continuing on the budget issue: in most of the work we do on C cycling 
there is the implicit assumption that we know how to relate contemporary 
fluxes to long-term C cycling - I am not sure we really know how to do 
this.  Why I raise this issue is in our peatland work, where we have a 
complete record of C accumulation over the last 6 to 8 K years, we have an 
apparent disparity between what we are measuring now and what the past 
record tells us the peatland have accumulated.  In our case this difference 
amounts to about 50 to 60 g C/m^2/yr.  This raises the question of whether 
we are observing some shift in C uptake in eastern NA peatlands (climate, N 
desposition, elevated CO2 ...) or is it we do not know how to connect the 
contemporary to the past?

6. Page 20: in our modelling efforts coupling the model Steve Frolking 
developed (PCARS) to CLASS and a local climate model the major problem we 
encounter is the precipitation input to the ecosystem.  Ecosystems are 
sensitive to precipitation yet we do not have much confidence in what 
climate or meteorological models give us - the meteorological models are 
certainly better, but coupling them to ecosystem models in a regional 
domain becomes computationally intensive.  How do we handle the issue of 
uncertainty in this major input?

7. Page 32: "Flux-Canada" is referred to - I believe the correct name is 
FLuxnet-Canada.

8. Page 39 and on many pages before this: one of the major issues that 
needs to be resolved on the measurement and modelling of terrestrial 
systems is our treatment of soil CO2 exchange.  We build reasonable 
complicated models of photosynthesis and plant respiration and yet we treat 
soils with a simple first order approximation of mass, modified by soil 
moisture and temperature.  From a series of experiments it is becoming 
clear that we need to look below the ground a lot more and develop 
techniques to measure and then model respiration that comes from the 
decomposition of SOM, root exudates, roots, and recent litter (both root 
and above ground).  This is especially important given the results of the 
coupled carbon climate models - in the Cox et al. all the action was in the 
soils, but how well represented were the soils?  I do not have answers to 
these questions but I think it is the weakest component in our 
understanding of terrestrial C cycling.  This speaks directly to the issue 
you raised in the document of slow and fast processes - the soils span both.

Hope these are useful.  I am sorry I can not attend the meeting - I think 
it will be very interesting.  If you need some Canadian input in the future 
do not hesitate to contact me - I may not be the best person but I might be 
able to make the necessary connections for you.

Nigel
_______________________

Nigel T. Roulet
Professor of Geography & Director of the Centre for Climate and Global 
Change Research
McGill University, 805 Sherbrooke St. West, Montreal PQ H3A 2K6, Canada

Telephone (514) 398-4945; Facsimile: (514) 398-7347
roulet@geog.mcgill.ca
http://www.geog.mcgill.ca/faculty/
http://www.mcgill.ca/ccgcr/english.html