| A unit of Lasting Forests
evolving since March 30, 1999 |
|
A Total Forest Management Plan
|
|
|
The Trevey is a dynamic system. It is one that has changing or new data, new processors, or newly-edited text materials. More than this, it is designed to be a learning system. Such systems now exist and procedures are still being developed. Advances are likely to appear in the merger of the realms of "expert systems" and "artificial intelligence."
It seems reasonable that a system should be able to learn from and improve its performance over time. Under the best circumstances, learning is difficult. For a non-human computer-based system to learn, there is little experience. The parallel is in institutional learning.
Hilborn (1993) noted that resource agencies are victimized by great uncertainty. "Each decision is a gamble. " There are many situations in which there is no experience or not enough to quantify the likelihood of a change resulting from a decision. There is rarely a place where the lessons of the past are brought together and analyzed and disseminated through the system. Rarely are mistakes or failures analyzed, never recent ones. Mistakes are not discussed so they can be avoided or corrective, loss-reducing action taken.
The Trevey staff uses three kinds of institutional memory, namely files, reports, and the data base.
The Trevey staff seeks to include individual files by means of statistical analyses. Analyses are expedited; summary notes are included. Voluntary analytical assistance is provided.
Reports are included within the data base. Automatic indexing increases access to information.
Data bases are built and maintained.
The richest form of memory is stored in the cerebra of the staff and thowners and those experienced with the land unit. How to get the experience of from 10 to 40 years of work in people within the enterprise from a person into a useable form within The Trevey continues to be a challenge. Monitoring and evaluating can be costly, but within reason, these can result in learning and better responses in the future. At least some small level of monitoring can prevent reactive systems from being built and retained, those that are unable to look beyond the next crisis.
Experimental approaches can be taken (e.g., several things tried simultaneously to find out what works best) but these are costly, have delays, and rarely are situations available or possible to create that are really comparable and replicable in amount, space, or time.
We believe that institutions (and a planning system which is the potential core of each institution) can actually learn. At least by one definition of that term, this means to remember what has happened and assign some measure of goodness, or survivability, or success to the happening. Institutions do not learn because of high staff turnover, no one has time to master local skills, new staff are not trained by experienced staff, knowledge of retiring people is not captured, files are discarded and lost, reports are uncatalogued, things are forgotten, and computer tapes are lost or destroyed. There is rarely a "general staff " that stays together to become the institutional memory of the organization. These limitations of present systems are partially addressed by R* Guidance systems such as the following:
1. "Decision Document " are attempts to present major decision that are made and elements considered in a simple standard format.
2. "Decision Evaluations " are documents made in a similar form.
3. Short "Experience Reports " (about 5 pages) are filed. Theses are lessons learned and point to library information.
4. "Old-Timer Seminar Reports " are filed. These seminars are for long-term employees to reminisce about important lessons learned. They are for new staff to observe.
5. "Subsystem Histories ", long documents by a person appointed as historian, are maintained in the system.
6. In connection with this, each year an historical seminar is held to document lessons learned. During the seminar Experience Report authorship are assigned.
7. "Management Reports " (30-50 pages) and "Case Studies " describe programs, projects, and methods - details and history.
8. Retiring staff are debriefed and reports from these meetings are called "Retirement Debriefing. " These are reports made, often by several people, to capture as much knowledge as possible.
9. Beta estimation procedures are encouraged, i.e., the estimate of median values by asking for lowest values (a), highest values (c), then most likely value (b). These are used to estimate the median, M, which should be used in most decision making, namely
M = (a + 4b + c)/6.
New staff are encouraged (and trained) to utilize all of the above.
Knowing about things that do not work or fail does not prevent them being used again and again. This phenomenon needs study and special feedback developed. Change cannot be too rapid. Change with unpredictable returns can produce anomic behavior.
Objectives change, so even the basis for knowing "what works " or is pragmatically good can change. Objectives and their dynamics are discussed under the The Trevey section on the "Dynamics of Objectives. "
Research as Learning
Research on environmental areas is seen (by the staff of this planning system and others) as a truth-seeking process, a means to produce "finding " or "conclusions " that should be moved into decisions as rapidly as possible. Research builds a knowledge base produced inputs to decisions. Although differences in basic and applied research are debated, the debate is ignored: the need is for high quality information for improving difficult decisions, for finding "the optimum. "
We encourage use of the area for planned studies by scientists when such activities do not threaten the mission or security of thearea.
The plan has benefitted greatly from past research. Finding enhances the effectiveness of land management and managers' ability to meet various demands. Alig (1989), discussing difficulties of strategic research planning, noted that new demands, reduced funds, increased competitions, and a feeling that research results often do not support management activities as well as they might, "make a perfect matching of research findings with practitioner needs highly unlikely. " Improving the match seems possible.
Between- and among-area coordination of studies seems likely to produce efficiencies.
Research has provided information and conclusions that have become the basis for much of the analyses, decisions, and content of The Trevey. More than 20 different federal agencies conduct research on pieces of the environment. Research on the Everglades, for example, is being conducted by the Department of the Interior's U.S. Fish and Wildlife Service, the Environmental Protection Agency, the National Park Service, the Department of Commerce's National Oceanographic and Atmospheric Administration, the U.S. Department of Agriculture, and Army Corps of Engineers, and myriad of state and local agencies. Research is seen as the subsystem that provides inputs to decisions. It is of five major types: taxonomic and descriptive, parameter estimation, hypothesis testing, distribution development, and modeling. Research is only one, but an important, way to know anything. It has already contributed much. In the larger context of a dynamic system, many fruits of past work and investments are likely yet to be harvested. These returns can be used effectively in new models and expert systems.
Research needs to be continued in a strategic way on carefully-selected topics that will allow informed changes to be made in R very cost effectively. Results of past studies need to be used in the system; results need to be synthesized and used together. Data and photos (etc.) need to be protected. All results need to be brought to bear on decisions about the area. Preliminary and pre-publication reports are essential. Risks of losing data or reports are high and back-up and fail-safe efforts are essential needs.
Research is costly so strategic efforts are needed. A list of research topics is available but the over-arching concepts for conducting a research subsystem and selecting activities are as follows.
1. Probable identified use (potential use on the day a conclusion is reached).
2. Results likely to be used in influential decisions.
3. Well-designed sampling with
4. Sequential studies.
5. Use of Bayesian estimators.
6. Use of simulations to assure effects of conclusions.
7. Use of parameter estimates (high, low, and median).
8. Sensitivity analyses (5 and 6 and others).
9. Use of bracketing with Fibonacci search.
10. Use of median estimators (beta and Weibull distributions).
11. Interative data analyses (nightly or weekly prior to next sampling event) with projections of estimated change based on prior sampling.
12. Shared, non-duplicated, multiple-use equipment
The following is a list of needs. Individuals, corporations, foundations, and community groups may be encouraged to invest in these important research areas. The list is a basis for immediate planning. Subsequent public and staff inputs may change the list substantially. The list is of the needs perceived in plan development in 1999.
Atmosphere
Lithosphere
Hydrosphere
Biosphere
A comprehensive ecological simulation of the lands of the region is now feasible. As funds become available, efforts to develop such a simulator will be planned.
Ahlig (1989) said that the "long delays between problem identification and completion of research and the unpredictable nature of the research process make a perfect matching of research findings with practitioner needs highly unlikely."
A research strategy is needed, one that increases the chance for a match. Herein the objectives "drive" research. If a suggested project is not likely to provide knowledge for decisions that tend to improve R, then the proposed work is for someone else.
A comprehensive model (actual or anticipated) is the format for learning and for supportive studies. The model provides the places into which research finding will be placed.
A staff team suggests needs. Scientists seek to conduct cost effective studies. Assistance is gained to assure continuity of studies and learning. There are more needs than resources (all types). Allocating the resources is done based on cost-effective expected change in R.
Research is observed by invited reviewers who have a feedback role, not just one of monitoring.
See also notes on the web site and education.
| Quick Access to the Contents of LastingForests.com |
|---|
This Web site is maintained by R. H.
Giles, Jr.
Last revision January 17, 2000.