SORTIE/BC for Black Spruce Stands in the Claybelt

Presented by Yves Claveau.

Study objective: Simulate long-term impact of partial cuts.

The objective originates from current Canadian harvesting paradigm of wishing to emulate natural disturbances. In the case of black spruce forests, people believe that adult mortality should be emulated by partial cutting. So we want to know, will we keep the same stand structure that is seen in old-growth stands?

Study region

The study site is the claybelt - ancient lakebed.

On well drained sites, we can observe the typical succession pattern. Shade intolerant species such as aspen and jack pine are typically replaced by black spruce. Black spruce can tolerate shade and have better regeneration capabilities when the moss thickness is high. Such species replacement can be observed when the fire interval is long enough. However, on wetter sites, we rather observe changes in stand structure from even-aged to uneven-aged stands as time since fire increase. In the case of older stands, the irregular structure is maintained over time. As in other ecosystem fire bring the stand into its initial stage. On site with poorer drainage, the paludification process can be observed. Paludification is the biomass accumulation caused by the increase in water availability and by the growth of sphagnum mosses. The increase in water availability is related to precipitation > than evapotranspiration. Over time, the moss layer lead to a decrease in site productivity through lower soil temperature, lower decomposition and mineralization rates and by lower root growth. Here fire can burn all of the peat layer or only part of it. As a results, stand dynamics following can be very different where the stand can start with an irregular structure when fire intensity is low.

Here is an example of a black spruce stand on a paludified site with its moss layer dominated by sphagnum species and the shrub layer dominated by ericaceous species.

Parameterization and model running

Species are chosen to reflect changing site conditions over time.

Changes in microsite initial conditions:

• Presence/absence of moss
• Drainage
• Fire intensity

Observe for the different initial conditions:

• Paludification
• Resource (light)
• Spatially explicit

Model changes needed:

• Growth - juvenile - taking into account light and local moss thickness
• Growth - adult - using competition, shading, moss thickness (similar to current NCI)

To answer the question, is the ecosystem driven by paludification?

Discussion of this talk

Question to the NZ folks: Are the NZ terraces similar to the situation Yves described? Yes, at least in some areas. No regeneration from seed though, and currently considered too complicated for the model.

Question: We have data for the effects of substrate on recruitment - is there a good correlation between moss cover and peat thickness? Is it easy to measure peat thickness? Yes. Is this a better predictor of growth than moss cover? Not yet known.

A good study to do - relationship of thickness of sphagnum and growth. It is too hard to simulate paludification itself, but you could study the effects of different levels of paludification. Currently Yves is unaware of any studies in this direction.

Foresters want to scarify sites to stop paludification and improve soil conditions - this would be something to model.

Question: Are there good models of paludification? There is a lot of work in climate change - people wanting to know of peatlands are source or sink of carbon - but otherwise don't think so.

Question: What nutrients are limiting? Unknown. But the idea is to use moss thickness as proxy for nutrient availability in order to have one parameter to influence growth.

Question: How do you predict moss thickness and its change over time? It's not easy - and don't know yet. And there might be interactions between sphagnum and canopy - which would be interesting to study.

Question: Would peat really affect the growth of adults? They would be rooted in the clay, right? And how does the acidification affect?

Any info about atmospheric pollutant deposition? Expected to be pretty low.

Discussion

Question: Anyone know of research on water limitation?

• Some people have looked at water table level
• Some data on BC trees in the edge of gaps
• It would be very useful to be able to study extreme events (moisture and freezing events) - some species are water-limited in certain areas, and in some areas frost presence is heterogeneous

It sounds like since there's the possibility of both static and dynamic resource templates in the model, the new design should cover what's needed for resources. Now the challenge is how to write the behaviors.

Question: Will you be able to change the resource grid at each timestep? Yes - for example, in exchangeable Ca, current work is on the dynamics of interactions with underlying parent material to look for non-additive effects.

When studying resources, it's important to know what effects are most important. The best models are simplest. So we should do sensitivity analysis.

Pseudo-species can be used when you don't have extensive-enough data to generalize your parameters. You can just model the extremes - "high-fertility-site birch".

Question: Can you capture temporal variability in the model? For example, seasonal changes in rainfall in the context of global warming with bigger peaks and troughs through time? You need settable-length timesteps, a way to describe the process through time, and a way to describe initial conditions. This is where the science comes in to make the descriptions.

Question: How would you do masting? Five-year timesteps smear out the effects. And it would be very interesting to study those effects. We'll use the settable-length timestep, and then a way is needed to trigger the masting.

We've always done a homogeneous physical environment - it would be interesting to be able to study temporal and spatial heterogeneity.

Windthrow - you average out over the timestep how many trees to kill. Would this approach work for studying hurricanes? The desire for hurricanes is to study effects of increasing frequency.

Defoliation events - how to model? We could make a generic behavior for that.

We could also make a generic behavior for pulsed events when certain conditions are met - like making an epidemic when firs get to a certain density. There are two kinds of pulsed events - certain model conditions are met, and user-set or stochastic.

Question: How can we capture a memory of conditions? For example, skid trails affect conditions for a long time. Functionally this is the same as the substrate's current memory conditions, but since this has large theoretical and practical implications, so it's important to develop this further.