Forest Modelling
for Ecosystem Management,
Forest Certification, and Sustainable
Management Conference
Summary of the Discussion Session
1. How will we obtain the necessary information for ecosystem management, forest certification, and sustainable management?
2. What will be the future models and modelling efforts?
3. What are the likely needs for the far future?
|
"Forest Modelling: QUO VARDIS?" Download this presentation (PDF) |
|
|
"Discussion Presentation" Download this presentation (PDF)
|
|
|
"The Role of Models in Answering Questions
of Forest Sustainability" |
|
Summary of discussion
Synthesis Between Data and Understanding
Over the years,
considerable effort has gone into data collection. It is important to
synthesize what has been measured in the forest and what we have understood
thus far. Observation precedes understanding and sometimes leads to
hypothesis testing. We must also move from hypothesis testing to application
to addressing the changing objectives of sustainable resource management.
For example, empirical and process models are needed to address and
assess climate change and carbon sequestration.
Grounding Models in Reality
Models need to be based on real data. We need data to validate models especially when there are significant consequences to managing resources in a sustainable manner. While there is an increase in modelling and computation capacities, there has been little investment in acquiring quality data at various scales. Data quality is extremely important to models. Some of the modellers collect their own data, others use existing data or use a mixture of both. More needs to be done on monitoring and model validation.
In Forestry, we lack consistent techniques to track tree growth and mortality from regeneration to maturity. There is a need to improve modelling of early seral stages, mainly small tree modeling. Most models have high uncertainities in estimating the number and frequency of species, and how each species will behave in a given ecosystem. To understand the process, we need to know how species interact over time.
Challenge for Site Index
It is known that site index is a circular notion in forestry. In light of this, one of the participants forwarded a challenge to estimate site index from net primary production or from physiological processes. Since forests are the realization of various physiological processes and radiant energy interactions, examination of selected attributes and processes might provide a better site quality measure.
Need for Empirical Studies and Designed Experiments
There is a great need for empirical approaches, as experiments cannot substitute for empirical approaches that collect data and analyze retroactively. We do not need to wait for the whole rotation period to learn about the forest. We can observe and deduce future trends using empirical approaches. However, for certain objectives, we cannot detect trends without carrying out experiments.
There is a need to continue experiments, such as spacing and pruning trials, to understand the effects of various management interventions. Both the empirical and experimental approaches require rigorous scientific methods to achieve the objectives of sustainable resource management.
Models Versus Reality
According to Popper's theory (1963), all models are false. Instead of finding data to support the model, you look for data that falsifies it so that you can find the least useful model and move forward. However, if you do not have the data to falsify a hypothesis, you cannot consider the hypothesis to be false.
Scientists want testable hypotheses, but managers want solutions for the problems they face. Even though previous knowledge is based on ground data and field experience, as opposed to laboratory results, there should be no distinction between science and reality. Instead we need to harmonize them. While scientific research requires testable hypotheses and attempts to mimic reality through models, science also should address operational problems as well. However, the question of whether reality can be modelled remains unresolved.
When making decisions based on models, it is important to remember the risk of error in models. This is of particular importance to managers who legitimize their decisions based on models. Error can be caused by the distribution in variables, the lack of natural randomness or uncertainty, as well as the modeller's own biases.
Certification and Social License
It was noted that a change in paradigm for forest modelling is required to include environmental and social values in future modeling efforts.
Overview
After discussing many key subjects, the conference participants recognized that there are important gaps and uncertainities in our knowledge. These challenges were identified in four key areas: 1) regeneration and small tree modeling, 2) soil and underground processes, 3) public participation and public perception of forestry, and 4) using empirical and process models to assess climate change and carbon sequestration. The conference concluded by calling for future studies to address these and other challenges in forest modelling.
Reference:
Popper, K.R. 1963. Conjectures and Refutations: The Growth of Scientific Knowledge, Routledge & Kegan Paul, first edition.