Discussion

Compliance with Strategic Goals

The selected silvicultural approach seem able to meet the strategic goals defined in the introduction.

The disturbance resilience of a stand is a function of its variety, with regard to species composition as well as with regard to development stages. The choice of an uneven-aged regime allows to modify the species distribution by selectively promoting or suppressing individual species on a shorter time horizon, since it is not necessary to wait for the end of a rotation. Growth of individual trees of a desired species can be accelerated by providing them with a competitive advantage during thinnings. The presence of different development stages helps in reducing the sensitivity related to a certain physiological age. Some parts of the stand have a better chance to survive an impact than others, and this reduces the probability of a total loss.

The sustainable productivity of a stand under an uneven-aged management regime that minimizes losses such as mortality or soil erosion is at least close to if not at a maximum, since the products are available on a permanent basis. There is not a large initial benefit (as resulting from timber sales in the case of a clearcut), but there is also no need of high investments with a net return to be expected no earlier than 40 years later (as incurred for a plantation that needs to be considered in a sustainable model). Even though it may be difficult to prove the productivity maximization for the individual case, it seems unlikely that there are significant reductions when compared to other management regimes under the assumption of sustainability.

The adaptive potential provided by uneven-aged stands is larger than the one of even-aged ones. As mentioned above, transitions can be realized faster and with fewer losses, since it is likely that some desired elements are already part of the stand, and these can the be promoted by suitable silvicultural concepts.

General Stand Development

As can be seen from the simulation results that represent flow values (e.g. growth, removal, to a lesser extent also mortality), the DBQ model used allows to reach a stand structure and development that is more or less stable over time. Due to the initial deficits and surpluses in number of trees for a certain diameter class, it is takes usually two interventions to come close to the steady state. This means that it is not likely that a single fuel treatment can transform an unbalanced stand into one that is able to maintain the desired properties itself. The objective of creating a biological automaton can not be achieved at once, an iterative approach is necessary. As a consequence, fuel treatments as outlined here can only be successful if they are part of a long-term management policy based on intensive silviculture. Even after a steady stand structure is available, regular interventions are required to keep the stand within the range of stability.

It can be argued that before the advent of euroamerican settlers, the stand maintained themselves in a fire-resilient state without human intervention. However, in this period, fire acted as a thinning agent, and the high fire frequencies recorded show that these presumably stable stand structures also depended on regular interventions - not by humans, but by nature. What happens to a stand in the absence of regular small disturbances is illustrated by boreal and subalpine forests: after a long period of stability with buildup of energetic imbalances, catastrophic corrections in the form of stand-killing fires are required to get to a (temporarily) balanced state.

Influence of Treatment Policy for Large Trees

An effect that was expected, although not in such a pronounced way, is the variation of the stand development depending on the intensity with which the large trees were thinned. The difference it makes whether or not this compartment is included in the silvicultural concept is best illustrated by the mortality curves. If all trees larger than 30 inches are left untouched, the mortality drastically increases once a certain holding capacity of the site is exceeded. With regard to the fire resilience, this is of concern because of the high rate of dead fuel accumulation (even though the stems of these trees are usually not very effective in combustion due to low surface-to-volume ratio). However, even more important in this context is the development of the upper canopy. Left to themselves, the large trees alone can in the long run exceed the target canopy closure of 40% and thus probably also the critical bulk fuel density even in the absence of any other fuel. In such a stand, large trees that have a high probability of surviving a ground fire propagated by smaller trees and shrubs will eventually be eliminated by a continuous crown fire they propagate within their own canopy layer.

Another interesting effect is that already low or moderate thinning intensities (leaving 90% or 80% of the large trees after the intervention) result in significantly lower mortality and decrease the risk of crown fire propagation in the upper canopy. It is by no means necessary to remove all or most of the large trees to maintain the stand fire-resilient. The main focus of fuel treatments is still in the lower fuel levels, particularly in the ladder fuels. As long as they are not reduced to a safe density, any additional spacing of large-tree crowns does not help to improve the perspective of the stand as a whole surviving a fire.

Biomass Flow

With regard to the biomass flow, two stages need to be distinguished. In a first short- to medium term phase, there is a large amount of biomass resulting from fuel treatments. The unmerchantable wood calculated by the model is in the range of 0.5 to 1 bdt/ha/year (average of first the 40 years). This does neither include branches nor any wood that is considered merchantable by FVS. For one intervention every 20 years, this corresponds to 10 to 20 bdt/ha (4 to 8 bdt/acre). In most cases, branches will be removed with the stems, otherwise the fuel treatments result in a temporary increase in fuel loads instead of a decrease. According to the allometric equations by Gholz et al. (1979), the branch biomass for both Ponderosa pine and Douglas fir is at least 20% of the stem biomass in the Northwest (cf. Figure 1). Knowing further that at most 20% of the total stem wood is unmerchantable, inclusion of branches will result in doubling or rather tripling of the biomass flow, i.e. 30 to 60 bdt/ha (12 to 24 bdt/acre). This corresponds with experience from real fuel treatments performed in the Sierra Nevada, where 60 green tons/acre (approx. 24 bdt/acre) have been reported (John Sheehan 1999, personal communication).

The high flow resulting for the first 40 years is due to the need to reduce current fuel loads, and it is not a sustainable volume. The simulations show typically a sharp decrease in removal of both merchantable and unmerchantable volume after the first two interventions. After this peak, the flow remains on a more or less steady level for the remainder of the simulation period, usually at around one third of the initial flow (0.15 to 0.35 bdt/ha/year stem wood only, 0.4 to 1 bdt/ha/year including branches). This lower number does in fact represent a sustainable amount of biomass that can be removed, i.e. what the biological automaton can produce under ceteris paribus conditions without deteriorating (except for a possible nutrient drain in the case of whole-tree removal).

Limitations

Whenever the simulation results are interpreted, it is crucial to consider that they are subject to a couple of limitations.

First, the input data used do not represent real stands, but inventory plots. It is possible that the characteristics of a plot are influenced by its particular setting (i.e. hydrologic regime, nutrition), and due to the loss of information incurred during translation of the data files, it is not possible to account for such abnormal conditions.

Second, the simulations are performed on a plot or 1-acre level, not on a stand or even a landscape level. A development calculated by the model may be realized for some particular areas within a stand or perhaps for a stand as a whole, but not for an area as large as a National Forest. Thus, extrapolation requires not only an aggregation of the results, but also their scaling.

Third, the simulations assume that there are no disturbances such as fires, insect outbreaks, droughts, storms etc. which interfere with the stand development. Again, such an undisturbed development may be possible for a single stand, but not for a forest landscape.

Fourth, it is assumed that FVS is a correct model for forecasting the development of the stands considered. Obviously, this is a very strong assumption, but since the local FVS variants were specifically calibrated to match growth patterns in the respective regions, it seems reasonable to use this tool in the absence of any other model that is proven to be more accurate.

Finally, the study was primarily aimed at estimating the biomass flow that could result from fuel treatments, not any other effects. Factors such as habitat conservation, water quality or socioeconomic development as a consequence of fuel treatments are not assessed.

References

Gholz, H. A. et al. (1979) Equations for estimating biomass and leaf area of plants in the Pacific Northwest. Forest Research Lab Research Paper 41. Oregon State University. Corvallis, OR.

[Next: Issues for future work]