| A unit of Lasting Forests
evolving since March 30, 1999 |
|
A Total Forest Management Plan
|
|
|
Robert H. Giles, Jr., PhD
A paper presented at the Conference on Education and Communication Applications in Natural Resource Management, Athens, GA, September 28,1995.
College of Forestry and Wildlife Resources
Virginia Polytechnic Institute and State University
Blacksburg, Virginia, USA 24061-0321
In order to teach well you must have a believable topic, one that" works" and one in which you, as teacher, personally believe. You must also know exactly what behaviors you wish your potential learners to display. Educating to change attitudes is passe'passuse the question remains: after their attaffixedanges, will students' behavior be that which I intend? An emphasis on behavioral change is not new, but it is rarely seen and rarely fully implemented. Behavioral change is the context for an alternative paradigm in natural resource management, one called the total land production system paradigm, or total.
This paradigm is not one for general discussion, an academic exercise, and an exploration of interesting options. It is the grounds for major change in the way land is used in a capitalistic, representative-government nation with a rapidly urbanizing population with extremely different economic and welfare conditions.
Teaching to achieve it will be difficult, but no more so than the difficulties we have faced for 50 years and that we still address in conferences such as this one. In this presentation, I briefly describe the elements of the paradigm and its educational dimensions. There are solutions here of natural resource issues. This is a modified "systems approach" (Von Bertalanffy 1968) 50 that is not new. The paradigm starts with context, then an emphasis on objectives (or goals).
Context
One aspect of the systems context is that we can view every land unit as unique. Every square meter is unique, at least every 30 x 30 meter Landsat pixel. The "watershed" is no longer relevant as a managerial unit. It over-aggregates information that we now have at very fine resolution. The land is Saolume 5 miles (3 km) above a map area and 5 miles below it. Phenomena within it are as much or more a function of things in contiguous map cells and nearby as they are of factors in the area. Migratory bird presence is a conspicuous example, just as is pollution of water above a stream-rich pixel. Nearby water influences nesting animals; shadow from nearby objects influence conditions on a site. What you measure in a quadrant may have littlifering on the processes and inputs in that square. Perhaps this is the singular meaning of "landscape ecology." Now, within geography information systems we can load, combine, and model phenomena within the cell as affected by presence or conditions outside of the cell. The educational implications are enormous in number and scope. We must work with our students to learn about all aspects of the land volume, then, aware that gaining "mastery" of the physical-biological system is impossible (thus irrational to try) we than seek specialists with one ancillary or sub-specialty. There will be a few who will become generalists. We need not stand in their way. The majority of students will struggle to gain a C-plus knowledge status in any discipline. The recommendation for one ancillary discipline is to assure there is functional awareness of (at least), ability to work with, the relations between very different fields. A specific implication is that we can stop classifying and teaching how to classify land because
We now have data and computer power to treat every land unit as unique. We no longer have to over-generalize for cartographic or other reasons.
Objective
The statement of objectives is in the coin of the day. "Money talks" is ore than an idle expression. I am of oftiew that all aspects of the land base can be expressed 5 "constraints." "Make as much money as you can," is the objective, then followed by constraints of "subject to not culling that stand of frees and staying a chain away from all streams." That statement of two constraints can be analyzed in terms of opportunity cost or foregone profit. The objective function is that of classic linear programming, namely, to maximize all values (convened to log (x + I) (Green 1979), all evaluated as maximum net present expected value, all values expressed as estimates, all with probability of occurrence, all over 150 years, and with analyses run four times with interest rates set as (I) high, (2) low, (3) most likely, and (4) at the Overton and Hunt (19) discounting procedure. This formulation of the objective sounds more complex than it is. It is a rough, conservative, parametric approach to well known financial analyses. It has many variables and can be very discriminating. It is one readily implemented on the PC (e.g., Lindo). This phase-one analysis of objectives gives four answers based on the above discounting procedures. They are likely to be the same, but several are different often enough to suggest the need for alternatives.
It is well known and often repeated that optimizations are aids to making decision, that selecting the algorithm to use is itself a risky decision. Phase 2 analysis is the computation of the B index to the overall system performance where B* is the desired state of the system when owner's benefits are minimally achieved. The manager seeks to minimize that absolute value between the two for the lowest cost (C), i.e., the objective is to maximize R (achieve R*) where
R = B - B* /C.
We have learned, painfupainfully optimization is a complex process with many different procedures to fit special cases. It means maximization (or minimization, or stabilization) subject to some constraints. Formulating the objective is difficult and expressing all of the constraints is also difficult. Once used, a common experience has been realizing that it is very easy to suboptimize. That the answer was unexpected, counter-intuitive, "a little off" the expected (or the way things have been done in the past, presumably because they were optimum). The difference is worth study for even slight suboptimization over 100 years, over hundreds of acres as typical in forestry, can "mount up." Key to the concept of benefits is that there are many different kinds and that this is a total profit maximization effort over an unusually long? The concept of R* is fundamental and this paradigm, as all do, hangs on the objective. It is a decision. It requires a risk (for it may be poor or wrong) but not deciding, which I perceive to be our present state, is equally or more risky, i.e., leading to failure.
Our objective, the desired benefits, is the summation over all populations of people, objectives, and time of the four major variables: demand, value of each unit of demand, our edemandd outcome or 1.0 - risk of failure, and substitutability of units of demand. We analyze the units of resource demand D (that can be achieved under normal conditions in the wild) for many groups of people P, for each objective i, over a long planning period (at least 50 years). Similarly we express the relative importance of each unit of demand (a value V) that changes with age and other factors). We then estimate the risk of failing to achieve the demand levels. This gives us, when combined, an "expected value," expected estimate of the number of weighted units needed to satisfysatesverse group of people over time. The concept can be simplified for the single landowner. The computations are run with financial as well as energetic units for comparison. Assumptions are that actions will be legal and constraints included to represent laws and policies where needed. An enforcement cost is included and a 95% effectiveness of enforcement assumed. Citizens may be involved in quantifying all of the elements of B* and suggesting ways to reduce the value of C.
The emphasis, is on total production, not knottiestat of trees and includes items listed in Table 1. The individual products often are economical operations and it seems naive to suggest them. When they are operated (1) within the context of the other products as joint production from land, and (2) as separate across-ownership enterprises, they can be profitable.
The first condition achieves economies of scale, stable employment and its associated economies, and allows year-yowlinge of equipment, space, and labor. The second condition is novel but not unexpected. It unifies operations, not on the farm or forest, but in the enterprise. The land in effect is rented to serve an enterprise consisting of work on many areas. A honey enterprise, for a simple example, has hives on many ownerships. Centralized marketing, expert advice, low cost supplies, etc. make the enterprise (not the marginally profitable hive owner) viable. In combination with other enterprises (e.g., cattle, ecotourism, medicinal plants, bird watching, timber) the combined enterprises and diverse, stable, supportive, share common resources, and achieve economies of scale and synergism well known by ecologists. These enterprises are akin to populations; the combined system is akin to an ecological community.
In puts
The system inputs are neither new nor unexpected. The pattern of using the inputs, however, is spatial. Every cell, every land unit has hundreds of factors available for it. We do ecological and economic modeling at the land pixel scale - the intent is largely to assure production; the parallel work is to reduce losses. The two are linked, for example, in planting the right tree species on the proper site at proper spacing to avoid moisture stress and thus mortality and associated, subsequent insect attack (H. I. Heikkenen, personal comm.).
Landscape ecology concepts are included in new nearness-to or contiguity map layers. (For example, a cell near another cell with water in it is probably as valuable to some plants and animals as actually having water present in a cell.) We load a cell with factors in the cell and also factors not observed but nearby - in three-dimensional space.
We never have enough data. We never will. We believe it irrational to act as if we will (Giles et al. 1993). We use available data, of course, but our sampling strategies involve use of alpha levels of 0.2 and tolerable error of 0.10. We use expert systems concepts and attempt to provide continual feedback. 3
| Quick Access to the Contents of LastingForests.com |
|---|
This Web site is maintained by R. H.
Giles, Jr.
Last revision January 17, 2000.