| [ HOME | Forest Faunal Systems Home | Table of Contents | Gamma Home | The Finder | Glossary ] |
This chapter is very long so it has been split into several sections. You are in Part 3.
You may return to Part 1 or Part 2.
Tree Harvest Regulation
(See J.M. Gulgin. 1991. Uneven aged BDq regulation of Sierra Nevade mixed conifers, Western. Forestry September, 6(2):27-32 reprinted , J. Forestry 89(9):29-36. )The above concept of succession leading, at least in theory, to advanced-age forests can be used as a basis for combining and planning for many faunal and other forest benefits for people over time. It is useful to consider if or when this fits within conventional forest harvest operations.
In most forests, so-called timber demand or timber cut quotas drive the major forest activities. Generally commercial forest areas are cut in a sequence based on their benefit-to-cost ratios (although not on public forest lands). (A very small percentage of private forest land is harvested under the guidance of a university-educated forester.) The areas are cut (or simulated cuts are made and the area tallied) until the demand is met. When the quota is not achieved with cuts of areas with a positive B/C ratio, more acres are still cut. In some cases quota-based or volume-regulated cuts using the above method may be necessary (e.g., to achieve well-located browse for elk or to sustain a local industry) but, I think, rarely.
As acres are added to achieve the quota, these acres are added to what are called "lands suitable for logging." They may be unsuitable economically. Often terrain that makes costs high also makes erosion great, so that forest stands are unsuitable for hydrologic and ecologic reasons, but they can be positioned, administratively, as suitable. Rather than added, these should be subtracted from the timber base. This subtraction seems rational unless the "timber base" is viewed as dynamic, and that each area is continually evaluated in terms of species; species size, age, quality; wood demand; harvesting technology; access; topography; and knowledge of the ecosystem. When viewed as dynamic, an area may be in or out of the timber base depending on the many factors listed above (and others). A dynamic land designation system makes sense now that there is computer capability to analyze every acre of land annually.
Lands which may lose money when timber is harvested should be withdrawn from "the base.". This means that any allocation system no longer "works" on them. The system is constrained. They may be tallied as producing wildlife, water yields, etc., but they are no longer available to tally as part of the money-producing regulated wood farm. In fact, they are not included in the old-forest or "old growth" category. Later, trees may be cut (left or harvested) to achieve other objectives but the first-phase activity is to analyze for the potentially profitable timber base.
In some forest planning systems, it is difficult to tell whether an area is included on the timber base lands or is used to meet other system requirements. There is a need to treat all lands as equally subject to decision forces. Thus, based on economic, ecosystem, and other criteria, each area may participate in achieving citizen or owner objectives.
It needs to be said continually that it is rare, perhaps impossible, to achieve objectives. They may be sought, worked toward, approached but rarely reached. The perfect combination or action is sought, but humans must settle for best. Optimum is not perfect. It means best -subject to constraints. "Best feasible" does not equate with "best possible." There remain infeasible situations. There are situations in which there are only very bad solutions. Ranking very bad solutions can only produce very bad results. When there are no good solutions, usually the best action is to stop or delay until conditions change.
Some forests are said to be "area regulated," meaning that each year timber will be harvested from a relatively equal number of hectares of land. This is a straight forward management scheme. The ownership is mapped in approximately equal size units equivalent to how long it takes to grow a profit-making tree; they are cut regularly, replanted, and then X years later (the so-called rotation age) they are revisited and cut again. This is the logger's strategy. Mealy and Horn (1981) and Mealy et al. (1982) improved on the algorithm for interspersing harvests of approximately equal area. Roach (1974) developed harvest scheduling strategies and pointed to modern computer selection of an optimum strategy.
"Volume regulation" is the unperceptive mill-owner's strategy. It assures a constant flow of logs or pulpwood through a mill. The area cut over may differ because some sites may contain more wood than others. An attempt is made to stabilize the volume produced along with associated employment, mill operation, etc. The Sustained Yield Act requires on public land some semblance of a constant flow of wood. The on-going problem for all people interested in forests, both public and private, is that wood, logs, or pulp sticks are not the only products of the forests. So-called "yields" are not only cellulose but surface water, ground water, wildlife, recreation, viewscapes, ground wind cooling, erosion control, employment, regional business cashflow stability, financial speculation opportunities, and others. Some claim that the most desired sustained yield is profit, but then the debate centers on whether that must always be stable or increase for stockholders. Increasingly, wood is becoming a secondary product of forests. Perhaps it always has been. It is very clear that in many areas, land value, not the wood thereon, has been the corporate holding of interest. Land value generally increases faster than trees grow.
The above procedure with the successional curves can allow a complex set of forest yields to be sustained (if that means stabilized at some named rate, even zero). These yields can (and should) be defined as objectives to be met by the forests. The primary way to do so over time is to harvest timber, assure reforestation, and honor constraints on the system such as the set-asides. The private owner or public groups assign weights. Laws determine minimums and maximums, and the ecosystem with F13 adjustments does the rest. The results may provide less stability in the volume delivered to mills (but sufficient volume at lowest costs and more regional stability), assure citizen participation in public land use by them assigning their weights, and utilize the best of both area- and volume-regulation strategies.
The need for sophisticated silviculturists seems to have been replaced by a need for people who can mark a clear-cut area boundary. The successional procedure described herein requires sophisticated foresters or fauniculturists that can locate stands, identify parts of them, and fit harvest areas to the land to meet faunal, silvical, hydrologic, visual amenity and other citizen or client needs.
If forest land cannot produce the desired set of benefits at reasonable cost, then it needs to be left alone as a speculative investment or converted to a non-taxed public resource. Land publicly owned eliminates land value questions. There is no real need for the public to allow prolonged operation of a "total benefits mill" inefficiently any more than an economic system will allow a mill owner to put logs through a mill at higher costs than the boards coming out will bring in the market. There are too many areas, too many opportunities, too many ideas to force present forest managers and users into an exploitative, "take-wood-at-all-costs" attitude. Forests do need to be harvested (but not all of them, every stand at any time); forest stands need to be started, adjusted, and cut, but all done wisely and sensitively, using the complex messages of the land integrated into that band of information for people, the succession curve.
Not area- or volume-regulation but system regulation is needed. Linear programming was suggested by Hennes et al. 1971). It is the "profit," the total net benefits in all classes at the top of the system, that provide the basis for system regulation and sustenance, not the crude base properties of area and wood volume removal. Responsiveness to markets, demands, and costs are fully as important as responsiveness to area growth rates or outputs per year. The faunal system manager can operate rationally within a rational system. Rather than being viewed as single-minded, narrow, or even obstructionist, the faunal manager may participate as an active team member in a force that assures a set of long term benefits for his or her clients, private or public. Usually those benefits will be most greatly increased by work at the interface between forest and mill, between animal population and user group - not in the forest itself. There is danger in losing sight of what the faunal system manager should do. There are jobs for loggers, truckers, surveyors, researchers, etc. and all are important, but the managerial task is to see the system objectives, whatever they are, clearly, and to allocate the appropriate resources (labor, equipment, thoughtful hours, money, etc.) to achieving these. If the market value of a cant (a rectangular piece of wood from a log with 4 slabs removed) can be increased by one-half with two more passes through the saws, then where are the gains to be made in tree growth that will equal that? As what will a cherry log be most valuable, veneer or furniture blanks? There are parallel questions for the faunal system manager. What is the objective of the system - deer in the field; deer harvested; paid-for deer-observation hours? Perhaps the objective is net regional monetary income from a year-around deer-based recreational enterprise? Just as value is added by passing a cant through saws, value is added to a deer population by marketing an area for tourism, and developing hunting camp cooperatives, unique restaurants, winter feeding programs and sleigh rides, hiking trails, for-fee educational situations, and product sales. Producing deer is not the faunal system manager's objective anymore than producing cellulose is the objective for a forester. Producing benefits or financial yields from a valued product is the manager's quest. Why should a region in one state die with unemployed people and poverty when raw deer hides from that region are processed at great profit in an adjacent state? At least a small tanning operation could add some value within the region to the faunal system product. Skeptics can always find another "well, what about ... ?" question. Well, what about capital for a hide tanning operation? Can capital be found? Finding it is the manager's job as surely as finding fertilizer to get a good crop of game-bird seed. There are loans, stock options, and other creative ploys for gaining capital. Citizen-operated enterprises work, especially if making blue-chip returns on investments is not required, only the ability to stabilize a kindly, healthy, local community.
The discussion here is partially about the context of any forest system. Perhaps I have expanded the context too far, away from the forest and into regional economics, the stockbroker's office, and even into the realm of whether cants or sawn wood can or should be shipped to Japan. In part, the issue is the manager's context. Some would say that they "do not feel comfortable" in such an expanded system. Managerial comfort is not that for which public or private clients pay tax-based or other salaries. They desire benefits and reduced risks. They desire returns on investments, not uneconomical production of wood, excessive deer, or low-quality experiences in days afield. Intuitively, they want whole system management. They have never had it, but they can imagine it. Individuals will complain when their personal weighted resource is not given the prominence expected, but they can be taught, and they can understand. The lead rule in working with the public in forest faunal systems is: "Never underestimate their intelligence, always underestimate their knowledge." The system and the approach being taken can be explained; it can work. It has worked in industry, the military, and elsewhere. It will work if seriously and persistently taken. It will work if there is ample feedback.
Biologists have long recognized that species are associated with each other. When you see A you can expect to see B, C, and D. On the other hand, they also recognize that in habitat type S of age x, you can usually find A, B, C, and D. What are the phenomena at work? These animals are not "friendly," obligate parasites or symbionts, innate competitors. These animals seek resources that occur in different amounts and change as stands age. What are "management indicator species " (National Forest Management Act, 36 CFR 219.19a)? Things that show the need for management? Conspicuous life groups that have many associates? Things that are sensitive to managerial action? Perhaps they are species linked tightly to one stage in one type? Once credited for the "featured species" concept (Giles 1962), (if so), I wish I could recall it. It was a way to suggest how to concentrate on maximizing conditions for certain game animals in a stand or compartment. In some areas a manager may feature species G, in another area, species H (based on user preferences, access, forest conditions (stand type, density, basal area, age, and expected harvest data in the rotation)). When I first discussed indicators, I did not have the total forest system in view; I did not know the relevance of succession curves. I was still a "game manager." All of that has changed. The featured species concept needs to be put on the shelf with other ideas that were good in their time. Managers get animals associated with stand type and age. They do not produce them. They can influence the shape of the succession curve a little but when it comes right down to the truth of stand management throughout a large forest, there is not much (if any) difference between forest rotations and budgets on lands where there are said to be featured species and forests where there are none. Perhaps some stands will be cut early to provide deer browse; perhaps others will have harvests delayed for adult wild turkeys; perhaps some inaccessible, high-cost advanced-age forest will be claimed as featuring black bear. The rational manager, applying feedback, seeks the significant difference. It can be replaced by an all-species concept (not just "many") embodied within system regulation.
A Temporary Alternative Regulation: Alpha
A way-station on the road to that above imagined grand state of affairs, an alternative that will satisfy legal and social requirements for biodiversity, an alternative that uses knowledge of associates and age-related faunal groups, and one that almost fits the widely-used area-regulation harvest strategy, is alpha regulation. This is a type-age and type-size harvest regulation that (1) assures the presence of all potential historical (200 years or more) forest types found within the forested area under management; (2) assures the presence of a reversed-J distribution within stands of 5-year age or equivalent size classes in stands in single-tree or group-selection harvest; (3) assures all stands are of reasonable size (e.g., equal to a greater than one hectare); (4) assures stands of equal type, if contiguous, must be greater than 5-years different in age; (5) may use any silvicultural system appropriate for a type but one that emphasizes cost-effective regeneration; (6) allows single-tree selection in stream-side zones (second order or greater streams) of 25 meters, but requires within-stream cross-current placement of large limbs of such trees or similar placement by felling of low value or non-merchantable tree boles; (7) allows harvest when a stand is demonstrably beyond "financial maturity" and has a positive present net value at the mill or market but may allow other harvests to achieve the desired age class distribution; (8) assures all snags are left, or if they are absent, at least one is created per stand; and (9) allocates 5 additional 5-year classes beyond the median age class for a type demonstrably shown as that of "financial maturity." The end result is a forest in continual, profitable production of wood but one with many small, even-aged stands of all types and all ages, many uneven-aged stands, and with one large area (sometimes entire areas) dedicated to older-aged trees. Streams are protected and rehabilitated without loss of timber value, clear cutting is allowed, an economic perspective is assured, and conditions of opportunity are assured for all plant and animal forms including contiguous areas into which animals (salamanders, etc.) may move when a stand is cut. The result will be an extremely diverse forest (many types, perhaps 20 or more 5-year age classes within each type or over 100 cutting units at minimum). Given set-asides such as wilderness areas, viewscapes, stream-side protection zones, and wetlands, and given lands unsuitable for harvest (for economic, geomorphological, or access reasons) the forest becomes a vibrant place for those who benefit from fiber as well as fauna. In part, alpha regulation feeds upon one underlying rationale of a biodiversity objective, namely to maximize future options. Today's foresters have already seen enough changes in technology, markets, laws, interest rates, and land values to know that there is future uncertainty about forest management. A forest of diverse types and ages, a mature tree reserve, and well-stocked and managed stands may not be a bad idea.
Alpha regulation includes the requirement that a stand be cut only if it achieves objectives to do so. Every cut has likely cost as well as influence on a set of objectives for many years and these include wildlife and profit and a sustained wood flow to the mill. This requirement makes many acres of land unsuitable for harvest cuts (due to type, quality, management costs, access, topography). It also, in effect, dedicates them to producing some of the many other benefits that people derive from forests other than wood. Unless these areas are properly accounted, i.e., their growth of wood not included in the growth = cut balance sheet, it is possible that an agency or company can show far more growth than cut and create a rosy picture of sustained production of biomass. If not included, corporate- or public-owned lands may be being overharvested. The growth on the productive acres may not match the harvests from these same acres. These situations are reason for economists to complain, but also faunal system managers may readily do so because they foretell boom-and-bust conditions for animals such as deer that are dependent on age-specific stand forage. They suggest excessive removal of older forests where so many faunal and floral forms are found. "Sustained" may connote the ups and downs of a mining camp that is not yet a ghost town, but for most people "sustained yield" suggests a considerable degree of stability. It is necessary that state, federal, and corporate forest inventories indicate the nature of the areas from which yield might be expected, then to regulate harvests on them accordingly. In some areas and agencies, timber harvest quotas for area foresters, derived from imprecise inventories of such lands and desires for forest-wide success stories and local taxes from wood sales, cause excessive harvests (and use inappropriate silvicultural harvest techniques). Both have been long-standing points of disagreement between foresters and faunal system managers. Cooperative work can be focused on inventorying; accounting growth; and calculating the allowable, annual objectives- achieving cut; not that based on a simple criterion of slowed tree growth. Cooperative work is needed in integrated pest damage management, e.g., control of deer damage to young trees (Chapter 11).
Cutting wood seems appropriate for anyone who can understand the end of a crosscut. Achieving a regulated forest for a century or two, one that achieves faunal and other objectives simultaneously requires abilities akin to those for playing three-dimensional chess. It is almost impossible to communicate the process. Most people assume it is simple. Inability to explain it easily produces a sense of hiding something and of frustration. Some revert to simple practice, somewhat suboptimal, but one that can be explained. Others try the practice and get criticism for not marketing "that really good stand over the ridge " or for harvesting "that beautiful grove where I first kissed my wife-to-be." The easy view is of the stand; the regulation is for the entire forest. In the faunal system, the stands are harvested or not, in a timely manner, to produce with high probability desired wildlife conditions over the planning period. That is done to achieve the stated desired benefits. This must be done at low cost. Due to budget constraints or some mental calculus, the manager may easily reach a state within a forest in which a perfect fit would occur with the objective if stand A was cut ... but the cost will be too great! The manager seeks the desired benefits but there is a cost component to the objective function. Together they must be satisfied over the long run. A delay may incur great long-term costs if the objective is to be achieved. Similarly, an infusion of "extra money" can result in excessive work that will unbalance the system for generations.
"Sustained yield" of wood that must always be harvested at a constant financial loss can lead an individual or a region to bankruptcy. Surely this was not the intent of those seeking land renewal in the post- depression era of the 1930's by passage of laws with the sustained yield concept. Sustained yield gets more "press" than it is due. All of the forests of the U.S. or the world are not those of the U.S. Forest Service. Their excellent staff, research, and publications contribute to forestry around the world. Nevertheless, their actual area is 191 million acres out of about 686 million acres of forest land in the U.S. Most of the land with trees (some hardly "forests") is in non-industrial private ownership - unmanaged, high-graded, understocked, and far from being productive of timber over the long-run. Industrial lands are typically intensively managed, but these (as exemplified in Maine) are now showing signs of "butt-rot silviculture." Some tree-growing land, hardly recognizable as forests, have been affected by international competition, paper gluts, unstable mass housing starts, and by pressure to appear strong in the stock market and invulnerable to unfriendly corporation takeover. Such short-term financial emphasis is anathema in a long-term business. Trees are long term. A forest's yield can be large and stable when it is planned and well-managed. Investments may not be needed; costs of doing business are. Planting a tree for an enterprise is as fundamental as restocking a grocery store shelf. Unless it is done, the store cannot survive. The overall forestry operation has been off-the-shelf for many years and future generations will have to restock. (How very, very sad, especially for those of us with nieces, nephews, and grandchildren about whom we care! How sad for the people in those areas of the world where the forests have already been removed and they many never regain them.) Tree-culturists suggest that wildlife people over-generalize about silvicultural strategies; faunal system people suggest that "wildlife" is over generalized. Modern information overload affects everyone working in complex systems. A supervisor may request a 2-page summary; a wildlife manager may need 3 pages just to list the relevant factors and topics. The U.S. Forest Service is said to provide habitat for 3000 vertebrate species. Their names cannot be listed on 3 pages! The "bottom line" is often devoid of essential information in complex forestry faunal system situations. The situation is a design for conflict; large systems often have counter-intuitive solutions. There are new needs to communicate well; the real need is to devote more time to the task of communicating. Evidence over several centuries suggests that the quality of communication will not be improved very much. We have to afford the time to discuss, to gain computer assistance in learning about complex topics, and eventually regain trust in the land manager. Eventually we must come to the unpleasant margin where it is realized that everyone cannot understand everything, that every "opinion" is not as good as every other one, that while everyone has a right to an opinion, only informed ones are socially relevant and non-entropic, or that action can be delayed until all "the facts are in" or until everyone understands. It is at this margin where high quality forest faunal system management teeters. The premise is anathema to those who have the peculiar education that leads to incessant discussion without decision. It is anathema to the pooled ignorance of poorly planned and managed, faceless committees now seemingly in charge everywhere.
The forest faunal system manager may be able to assist in restoring silviculture to its full dimensions, energizing private land owners, and helping land managers achieve both sustained profits from wood harvests as well as many other sustained benefits from forests. One beauty of a systems approach is that we do not have to decide between wood-profit or other benefits. 
We may select both. The potential returns from diverse faunal resources of forests are annual; there is no reason to wait until trees are harvested to claim a singular return on an investment. One faunal system manager can very likely positively influence more forest land by providing a sophisticated financial and tax management analysis to private land owners than by providing a forest full of conventional foresters. There are nice words easily spoken about land ethics, but my view is that when forest management is computed for a family (not just one individual's retirement-expectancy), when current legal constraints are imposed, when currently available knowledge is used in forest manipulation, when no more than 2 percent in excess of bank-interest rate is demanded every year (a limited-greed criterion), when a reserve fund (insurance) is maintained for reforestation after fires, and when reasonable size tracts (over 100 stands) are operated as a unit (personal, cooperative or industrial) then we will have wonderful forests and their fauna. I hypothesize that there will not be a statistically significant difference in these forests and those managed by people said to have a land ethic.
I retreat quickly from the interest rate criterion in alpha regulation but suggest using Overton and Hunt's (1974) criterion, CAP5016, and making comparisons. Few investments in forests are rational when based on a pure financial analysis. Discounted at 7.25 percent, a forest worth $1 million at 200 years is worth only a buck and a nickel today. Holling (1978:118) noted that there are good formal grounds for not using the same discount rate for all projects. He observed that the "appropriate" rate for evaluations is a political or ethical question and that there is no technically correct answer to the proper rate. He suggested weighting - an expression of the difference in temporal preference for a condition now and one in the future. He said that people often select very different actions (perhaps 5 times higher) than would be indicated if they used conventional discount rates. Wigg and Boulton (1989) reported that his German Forestry instructor was asked about the discount rate used in managing forests 250 years old, much too old to be "economical." His reply was that discount rates were not used. He reasoned that over the last 100 years, the German mark had been reduced in value to zero three times but that his forests had never lost their value. His question was why anyone would compare the real value of a forest to the artificial value of a mark at any particular time.
Harvesting trees is a technique for managing faunal space. Harvesting trees is conceived as a major part of forestry. An alternative view is that forests can produce a set of benefits over time and that harvesting is just one tactic available to managers by which they may maximize those benefits. (See Kim 1973.) Noble (1973) expected that within a few years in the Eastern U.S., forests would be seen as " ... a by-product of forest management" used for the benefit of people, with the main product being environmental enhancement. Some people see logs or pulpwood as the primary forest product, wildlife as a by-product (Leopold 1933). Others see wildlife as a primary product from lands which just happen to have trees, and that trees can be harvested or merely felled as a means to produce succession curves and faunal spaces over time (and also non-negative monetary returns). In some areas (clearly not all), the industrial forester seeks to maximize wood quantity and/or quality from an area; wildlife may be a by-product. The public wildlife refuge manager seeks to maximize animals (6.3 million acres of forested lands are on Federal Refuges); wood harvested in doing so is viewed, at least by me, as a by-product. On U.S. Forest Service and other federal lands, "multiple use" has been declared law (Multiple-Use and Sustained Yield Act of 1960 (16 U.S.C. 528-P.L. 85-517)). This conspicuously vague idea has been around for a century (Cliff 1962). The phrase is more useful in political than in field situations. It has been hotly contested, elaborately discussed.
In the Multiple Use-Sustained Yield Act (PL 86-5517) there is the definition of multiple use: " ... harmonious and coordinated management of the various resources, each with the other, without impairment of the productivity of the land, with consideration being given to the relative values of the various resources, and not necessarily the combination of uses that will give the greatest dollar return or the greatest unit output."
This law, similar later ones, and policy such as "ecosystem management" have had profound effect on land management, especially as they have called for a re-examination and new exposition of objectives and their achievement. Reduced to a very dry bone, multiple-use management produces a continuing, optimum flow of goods and services from all the resources of an area. Multiple-use management requires a holistic view of the ecology of the land, human needs and values, and management strategy. The manager must consider all values of the natural resources and strive to integrate and coordinate the protection, development, and use of these resources into the most compatible and productive combination available. Multiple use is a concept of judicious, unified management of all resources in combination to best meet most of the needs of most of the people. Impairing the productivity of the land is resisted, but even impairment may be appropriate when it results in the greatest long-term benefits to people. Management decisions are based on values, but not only monetary values. Under the concept maximum production from all resources or from any resource is not required. It is recognized that a mule and a miner, or a picnicker and a power saw, cannot be put on the same ground at the same time; some uses cancel other uses. The concept is that in practice, no single resource will give maximum production, but the resource-composite, both spatially and temporally, will give optimum benefit and production.
When certain expressions are used too much, improperly, or in too many ways by too many people, they vaporize. Herfindahl (1961) discussed the difficulties of slogan-based management, particularly multiple-use. "Ecosystem management" is the slogan replacement in the late 1990's. Rational action toward achieving a multiple-use objective requires a precise definition, a clear statement of a quantifiable, calculatable objective.
Multiple-use is deceptively simple, thus the source of its political power. Ecosystem management, in an age in which school children know about ecology and ecosystems, is both simple of concept but known to be complicated. In the very phrase, the dimensions of economics, esthetics, energetics and enforcement have to be assumed to be of equal or greater importance. Because of differences in objectives (many of them hidden), differences in perceived constraints, differences in willingness to take risks, to include costs or not, and differences in predicted outcomes (based on different experiences or models), grueling discussions continue over the rightness of forest decisions. Many algorithms have been advanced for the best way to solve the mixed-benefit resource problem. Work with most of them show that they are strongly influenced by a few key factors (e.g., the length of planning period selected, the interest rate, the total area involved) not the esoteric or small-area phenomena. Work shows that many solutions are "correct" over a broad range of activities, i.e., they are not very sensitive to exact projects. Of course, certain additions of constraints and limits can have profound effects. As more life-groups are added to these models, the more narrow becomes the range, the more limited the solution space. It is not unexpected that for some areas when a multiple-use solution is attempted on a computer, the answer is that there is no possible solution. This is exactly the state of the stalemated committee! No action may be best. Otherwise shifts must be made in objectives or constraints lifted or greater risks taken. These are the exact equivalents in the computer, the committee, or the congress.
In a Council for Agricultural Science and Technology Report (1975), the difficulties in using the multiple-use concept in decision making were laid out as follows: They will be similar for ecosystem management and are:
The act of tree harvest is a decision over which faunal system managers need primary control. Its timing, type, and area are the factors to which the long-term faunal picture is most sensitive. Failing to influence these decisions, what else is to be done? Creative use of the F-13 rule (CAP157) is next most influential. A long list of potential actions can be made. Depending on how each action is sequenced and used, certain species may be increases or decreased or stabilized, depending on objectives. A partial list of actions, for example, is shown in Table 7.15. "Sustained-yield multiple-use management is market-oriented, an exercise in rationing determined by the economics of consumption. Multiresource management [being advocated by Behan] is land-oriented, an exercise in husbandry determined by the ecology of production. The
| Table 7.15. Potential faunal space management practices for forests, other than specifying timing and area of cut. Most of these are elements in the F-13 approach to managing succession. An Action List is also available. |
|
Set maximum size for clearcuts
Encourage a high interspersion index, even of short rotation species Encourage elongation of index, even of short rotation species Defer cuts of strategically located hardwoods Protect riparian vegetation Protect soft mast producers in timber stand improvement Stabilize roadsides Plant road beds and sides in preferred wildlife foods Encourage timing of cut or thinning to maximize distribution and growth of wildlife plant seeds Minimize disturbance from logging during nesting and breeding period Release fruit bearing trees Retain snags Retain select large-crown hard mast producing trees Use prescribed fire when conditions are favorable to litter-layer fauna Use prescribed fire to gain a desired understory Protect or reserve seeps; maintain shade for ponds Build teepee-shaped brush piles Encourage planting, clearing or thinning to achieve wide spacing, equivalent to spacing of trees at harvests Protect crop trees from wildlife damage Create trails |
difference is profound (Behan 1990:16). The systems approach being discussed in this book is about and for both of these, thus neither. It is a way to decide and act that is oriented to achieving a diverse set of human objectives that can be achieved in many ways. It is as grounded in economics as in ecology, as much in esthetics as in energetics. It is the operation of a system of constrained benefit production for today and the future, with feedback. The difference is profound.
Some (and I believe most) of these problems of forest regulation and management can be addressed by the alpha regulation strategy. There are needs for set-asides (camping areas, natural areas or wildernesses, or visual quality zones), regulations on using areas, and separating potentially-conflicting users by time-of-use (i.e., seasons, even a morning-afternoon split). On public multiple-use lands, the public is to be served. There is no reason this must be a policy (no matter how poorly stated or unmeasurable) for all public lands or private lands. National Wildlife Refuges should maximize wildlife benefits to people; National Recreational Areas maximize for quality recreational opportunity. All cannot be maximized on a single tract; equal benefits from all will suboptimize. The solution is in achieving Q* as described above, it being achieved by Q with each q being constrained to some minimum or maximum productivity and there being set-aside or area constraints.
One of the beauties of seeking to achieve Q* is that the search provides a potential for creating a simulator from which, as an educational or deductive device, users can ask about the consequences of a change in the objective(s), or their valuation, or about the consequences of different actions (even if they have been demonstrated to be suboptimum). The ability to deal with uncertain value weights or valuations has already been discussed under Objectives (Chapter 4). If a purely monetary system for decision making is demanded (a demand at odds with public budget allocation processes in health, arts, military, and highway safety and at the personal level in mate selection or child protection, but must we quibble?), then opportunity costs can be used. (The "opportunity cost" is the worth of 5 years of squirrel hunting hours in a stand which is at least that of the value of that stand cut now and the money banked for 5 years minus the estimated value of the stand 5 years later. Relative values of other fauna to the squirrel or other tree-dependent creature can then be assigned.)
Protection of the land (under the Multiple Use Act and other legislation) can be assured by using appropriate areas under the succession curves. The shape of the curves may change, but the actual area under the curves should be little changed, if any, by the F13 techniques. Knowledge will always be limited (Chapter 3 and Chapter 6) but improvements can be increased as findings flow over time into improving the curves.
Irreversibility of use can be solved in some cases by set-asides and by physical moves (e.g., of an archaeological site threatened by reservoir waters) and with education. Some can be handled by mitigation, where the decline in the Q curve due to the resource loss must be replaced in like amount by any combination of the remaining uses. The value Q* remains as the objective; the creative powers of manager-citizen team are needed to achieve it, even though there may be a pending or actual loss.
Silvicultural Systems
Recognizing that trees may be harvested to produce extra needed human benefits from a mix of animals (associated with different forest stand ages or types of cellulose), and that there are harvesting policies or ways by which tree harvests over large areas are regulated, then it is possible to address how harvest and regeneration can be done. Foresters have devised "silvicultural systems" (Burns 1983). Although stating that " ... in a well-managed forest, harvest cutting can be done only in the context of a complete system of forest culture "; the emphasis is on systems that appear "biologically feasible" (without emphasis on economic, energetic, or esthetic feasibility). The systems names now connote a harvest strategy but leading silviculturists call for a system composed of establishing, harvesting, regenerating, improving tree quality or growth, and maintaining desired species of trees in a stand of suitable structure. Perhaps, some day, the other elements - economic, energetic, esthetic, and enforcement - will be included in the definition. The faunal system manager, like the forest manager, concentrates on how to create and manipulate a many-form photosynthetic array with a perpetually positive water budget that captures light energy, regulates its temperature, and that stores it effectively to produce a set of products, conditions, and services of maximum benefits at low costs for people and within constraints. Effective energy storage includes reducing losses such as from disease, insects, floods, and climatic events (Kimmins 1987 and Gordon 1983). A part of silviculture system work can be conducted by an orderly series of stand treatments such as use of prescribed fire or reseeding. Each treatment must be performed on schedule for the system to obtain specified objectives. There is widespread agreement that silviculture must be site specific but this is violated in using widespread simplistic prescriptions.
There are 4 major named systems:
All have variations and subheadings. It is difficult to find more than a few examples of areas that have been under the consistent influence of these systems (suggesting far more time is spent on professing than practicing these systems). Test: Locate on private land an excellent example(s) close at hand for a field demonstration. The practice tends to be dominated more by policy and demands for short-term monetary returns than by biology (stated by many silviculturists as their primary emphasis, thereby allowing discrimination between them and people working in the areas of forest economics or management).
The recognized systems are:
1. Selection System - mature and immature trees are marked (selected by a forester) and removed, leaving an uneven-aged stand, usually of mixed species, of increasing high quality. Regeneration is usually continuous and is from natural seed sources and vegetative structures. It is said to be difficult to apply in most forest types but is esthetically pleasing to many people and of great importance to more forms of fauna than any other silvicultural system. Filip (1977) discussed its difficulties, saying that its proper application required linking considerations of residual stocking, a maximum-tree-size objective (diameter limits and quality), and a reversed-J diameter distribution. The topics as well as linkages are difficult, a difficulty which perhaps explains debates on-site over whether an individual is practicing selection forestry or "high-grading." Uneven-age management is achieved through the above selection procedures and results in continuous tall-tree cover, orderly growth of desirable trees, a range of diameter classes, and the potential for sustained, even-flow yields of wood from the land.
Single tree selection, once widely practiced, is in disfavor, not because of biology but high labor costs and the near impossibility of removing crop trees with heavy labor-saving equipment now in use without damaging residual trees (e.g., Lamson et al. 1984). Draft animals, cable logging, and small tractors were once used successfully by careful loggers. Use of mules to remove timber has been revived on at least one Federal Wildlife Refuge. Use of mules to remove logs, in addition to its intended success, has become a visitor attraction and has created a new animal husbandry enterprise within the vicinity of the Refuge. Regeneration was virtually assured by seed from the residual trees adjacent to the "hole in the forest" left by the removal of single large trees or by the tree reproduction already "sitting in wait," almost with no stem growth (but abundant root growth) under the crop tree. Costs of removing individual trees are high and require special skills. Regeneration costs are low or zero. Sites are undisturbed, soil erosion minimum, wildlife maximized (for the advanced-stage-of succession forms; only slightly less for the early stages (all occupying the many small "holes in the forests" (now called "canopy gaps")).
 |
| Professor Leon Minckler (shown here in 1969) wrote a forest woodland management textbook and was a proponent of group selection. |
|---|
Group Selection is like single tree selection but groups of mature, low quality and cull (zero potential monetary value from lumber or round wood sale) trees are carefully selected for removal by a forester leaving a well spaced, defect-free, mixed-age stand not unlike the single-tree selection system. A major difference seems to be that there may be more shade intolerant tree species in the stand than in single-tree selection forests. The holes are larger (since 3 to 10 trees may be removed) and will have a diameter greater than the tree height, thus more sunlight, thus fewer shade-tolerant-only tree species. These areas (which may be up to 1 acre but are usually much less) may resemble small clearcuts after harvest but they are a function of the trees once present, the topography, the resulting shadows, and the perceived regeneration. The system is tree and stand oriented, not area oriented. The system favors fauna of the young stands and leads to a mixed-age forest with balanced age (or size) classes (the reversed-J distribution) in a mosaic of small contiguous groups throughout a stand. Unless careful, the forester may not get reproduction (due to competition from native plants and coppice and, in some areas, due to the deer or rabbits destroying seedlings). Group selection cuts may include thinning followed by burning to produce an understory useful to many animal species. 
As Minckler said (1974, 1980, 1987), the resulting opening retains the integrity of the forest; it has the same smell, feel, and appearance of the forest. Just as does responsible clearcutting, it requires a good inventory and assures adequate growing stock within the forest. Sensitive to individual tree growth conditions, to tree quality, and interest in moving to upper economic diameters, group selection is likely to produce more profit per acre over the long run on harvestable sites than clearcutting. Regeneration costs are minimal due to planned natural regeneration from border mature trees. Dense sprouting has prevented regeneration in many clearcuts; large hardwood tree stumps that result from group selection rarely sprout (Minckler 1987:24). Suppressed and small pole-sized trees are removed; soft mast producing shrubs are retained. It is a system very suitable for previously high-graded forests. Trees taken were large; trees left were not necessarily deformed, only small. The method allows the maintenance of a diverse forest appropriate to extremely diverse and extreme conditions of mixed deciduous forests (ice storms, winds, drought, excessive rains, fires (always creating uneven aged stands), insect outbreaks, etc.). Road costs are comparable if intermediate cuts (cleaning and thinning) are carried out within the harvest areas before they reach their maturity and planned rotation age.
Guttenberg's (1976) analysis of a group selection system in a southern pine forest seemed to demonstrate that a break-even economic condition was achievable on understocked loblolly-shortleaf pine stands - while achieving a natural appearance in the forest. This condition is highly consistent with many landowner objectives, is feasible in roadside zones where scenery is important, and is beneficial to many game and other faunal species. Leak and Gottsacker (1985) described these same conditions and how they can be gotten in northern forests by harvest strategies to achieve an uneven-age forest.
Leak and Solomon (1975) as did Smith (1974) discussed how full stocking and wide spacing can achieve initial rapid growth and maximum yield, simultaneously with abundant big game forage. The effort is for maximum timber profit yields from a comparatively few large trees with superior characteristics.
2. Shelterwood System - The mature stand is removed in a series of cuts (within a period usually less than 20 years) leaving a partial forest canopy until the last cut. The final cut removes shelter from the reproduction and allows the new, even-aged stand to develop. Its use is with species having seedlings that need protection or need shade to hamper undesired competing vegetation. Costs are high because of multiple returns to the same site; damage to reproduction is likely (Tesch et al. 1986). It can be advantageous in managing for layer-related faunal species and usually produces abundant grasses and forbs and saves yards for northern deer herds (Leak and Tubbs 1983). Ground forms are protected from intense sunlight of clearcuts.
3.The Seed Tree System - A large area is cleared of all trees except a few, scattered, genetically superior trees of desired species. These are left as seed sources. These seed trees may be harvested later after regeneration is assured. The system applies mainly to conifers. It is widely practiced. In some areas there are laws that require seed trees to be left. Unfortunately the number, size, and quality of these trees often seems to be less than desirable.
Extensive clearcuts (with or without a few crop trees left) are the beginning of large succession curves.
The shape of the succession curve will be influenced by the trees left for the regeneration. The curve will start sooner, be more dense, and the mature stages of the stand will be reached sooner than in those areas without seed trees.
Deferment cutting (Smith et al. 1989) has been used in Europe. Its results resemble a seed-tree cut but the residual trees (12-15 per acre) are not cut when the regenerated stand becomes established. It is used mainly to improve stand esthetics but has benefits to some animal species. Providing roosts for predators, they are not beneficial to some other species.
4.Clearcutting - All trees in an area are cut to produce an even-aged stand. The area is large enough to be mappable (usually more than 10 ha) and is included, typically, under an area-regulation harvest strategy. It includes natural or mechanical seeding or planting and resprouting of trees. It is said to require boundaries that fit the landscape and to require cleanup of debris and unmerchantable stems (though these are both rarely seen). There are no reserved trees. Snags may be left in some areas. In small clearcut areas, vegetation richness remains the same although abundance increases (Crouch 1985). Faunal species richness usually changes drastically, some species lost, some gained. In some cases, "clearings" or non-merchantable cuts are made to encourage growth, deer browse, or esthetics (Della-Bianca 1975; Hoffman 1987).
Debatable topics remain among the questions of propriety of clearcutting. McGee (1984), for example, cited studies in which hardwoods were not regenerated after clearcutting. He found oak leaf development was delayed in the openings and that clearing subjected seedlings to hotter daytime temperatures, colder nighttime temperatures, and to greater danger of frost than in the original stand.
Cutting of northern hardwood forests sets in motion a variety of ecological effects related to the removal of living vegetation and to the disruption of the forest floor. Stream flow is increased, transpiration is reduced, concentrations of dissolved chemicals in stream water may be increased several fold, and erosion and transport of particulate matter may be accelerated... Other changes may also occur, such as an increase in soil temperature and moisture, an increase in the rate of decomposition, an increase in nutrification, a decrease in the amount of organic matter in the forest floor, a reduction in canopy absorption and reflection of solar energy, an increase of dissolved substances in soil solution, and a decrease in the pH of drainage waters. The sum of all these factors can have a major destabilizing effect on the landscape....Deforestation had a major impact on both the amount and relative proportions of water, dissolved substances, and particulate matter lost from the ecosystem... Because of the virtual elimination of transpiration, greatly increased amounts of liquid water were drained from the deforested ecosystem during the growing season. Moreover, concentrations of dissolved substances in this drainage water were increased due to (i) accelerated decomposition, nutrification, and mineralization, mostly of organic matter, in the forest flora...; and (ii) the absence of nutrient uptake by vegetation. Coupled with the increased availability of light and with high soil temperatures, these conditions of increased soil moisture and nutrients provided a high potential for rapid regrowth of vegetation (Likens et al. 1978:492).
They reported that even though clearcuts are generally revegetated in several years, the substantial increase in concentrations and net losses of nutrients in drainage waters suggest that nutrient uptake by these plants is far from sufficient to prevent nutrient losses. Additional nutrient losses from the site also occurred in the wood removed. Although nutrients may be replaced by ecological processes, they found this may require 80 years and thus expressed a real possibility of long-term ecosystem "degradation." (This word only has meaning in the context of a specific set of objectives.) Duffy and Meier (1992) suggested that clearcut forests may never achieve the same vegetation of remnant primary forests (an expected conclusion because of the extreme variety of conditions over time of forest establishment). Although financial considerations and increased mechanization have been cited as the reasons for more clearcutting being done, Barlow (1972) presented data for Northeastern hardwood forests showing no clear mandate. Clearcutting may be economical or not depending on tree sizes, volumes per acre, logging conditions, stumpage, equipment available, and the size and (especially) the leadership of the logging crew.
Clearcutting has been very controversial for many reasons. It has produced giant forage fields for ungulates (Patton 1976) (but forage supplies impossible to stabilize), quality habitat for early-age forest fauna (but evident loss for mature stages). Size of cuts has offended the visual preferences of some people. The associated images of erosion, riparian vegetation lost, fish habitat destroyed, and faunal species shifts have been equally offensive. Size and shape of areas clearcut have been problems. Claims about forage produced for deer have been compromised by behavioral studies showing deer use only a limited zone at the edge (e.g., 90 meters, Blymyer and Mosby 1977). Reduced costs of logging and area-management operations have loomed larger than arguments about the capability of the practice to achieve specific silvicultural and regeneration objectives or about its ability to achieve multiple resource benefits on public lands. Regeneration has not occurred consistently or efforts to achieve it have often been very costly and have exceeded budgets that should reasonably have been a part of any silvicultural system design.
 |
| Leon Minckler (1969) was an outspoken critic of clearcutting and strong proponent of group selection. |
Concern about how harvest strategy influences wildlife, particularly browsing deer, has existed since before 1949. Morriss (1954) began studies in the early 1950's on clear cuts and selection cutting. He found, as others subsequently, that useable browse was approximately equal on the areas, that sprouts are heavily used in the first years after a cut, that forage quickly grows out of reach of deer, that deer subsequently concentrate feeding on the edges of roads, the cleared area itself being almost impenetrable. He understood then the often-forgotten premise: There is no justification for producing more browse than deer can use (or more of any other resources than can be used) or more deer than those for which there is perceived demand.
Ripley and Campbell (1960) restudied Morriss' areas, found the selection cut "somewhat overly used." They reasoned that in the clearcut area where browse was abundant, deer may have favored sprouts, thus permitting seedlings to develop. They suggested thinning after the stand grew beyond the reach of deer to favor crop trees (an act recognized to be usually prohibitively expensive under current economic conditions).
Della-Bianca and Johnson (1965), in an almost unprecedented study of an area years later, removed all but crop trees in plots as suggested by Ripley and Campbell. In addition to browse gained from the cleaning, they found cleaning "...essential in providing fast-growing quality timber stands and sustained yield of browse."
Changing uses of land has been listed by many faunal system managers as the most important problem of wildlife on forest lands. This includes loss of vegetated riparian areas, wetlands, and old-growth but primarily losses which occur in connecting hardwoods to pines. The effects of losing or adding these connections, actual or hypothetical (simulated), can be handled using the succession curves described above.
Wildlife Influence on Silvicultural Decisions
Cutting some forests for some species of animals in some areas for some specific period is usually necessary. Forests that cannot produce nascent profit from simple receipts-minus-costs can be cut in some agencies or industries, justified by the wildlife benefits. Enough "creative budgeting" has been done by all hands to realize that when a public or corporate project cannot be justified in one way, it usually can be justified in others. On one hand, the faunal resource manager wants animals to have as much monetary value as possible. An example of an estimate that can be used is the replacement cost assigned by a court to a convicted poacher. High values justify high budgets and high mitigation costs when habitats are lost to dams, power plants, and transmission facilities. If over-valued, however, when wildlife is thrown onto the benefits side of a benefit-cost analysis, they can be used to justify almost any act. Harvesting grossly submarginal timber is one example. Wildlife can allow road and timber cuts to be made through otherwise unprofitable areas to get to "the good timber." Only when animals are linked to formal benefits as discussed under objectives, can the real value of an animal produced by the forest be established. When demand is zero, the worth of the next deer or squirrel produced is zero! Changes in the forest, in weights, and in demands over 40 years, may be estimated and these used to estimate the worth of animals produced or the net values foregone if the trees are cut.
Single-tree-species forests are no longer as important as they once were. Most woods now have some market values. "How do you grow a good crop of trees?" seems like a reasonable question but one that is incredibly complex. It is reasonable to turn to a silviculturist, but there remain complications such as questions about "in perpetuity?", costs, ability to sustain the management effort, expected losses, profits, etc. It really is unfair to blame silviculturists for failing to answer the questions. Historically their role has been to control forest establishment, composition, and growth. That's all (Hawley and Smith 1954 and Hawley 1921). Hawley and Smith (1954:10) clearly understood there was no need to (and only difficulties later) to separate silviculture from diverse forest management and said "The separation of the two fields is made for educational convenience and does not prevail in practice ... The two are interdependent." Behan (1986, 1990) said that silviculture for fish and wildlife is no more appropriate than silviculture for sawlogs and that what is needed is a sensitive, operational, multiresource, interdisciplinary approach to forest systems.
In the past, faunal system managers have lived as if they were at some economic margin and were scraping around for faunal product values to encourage protection, justify budgets, argue for replacement costs, or to state impacts. These extra values are rarely used in military or health budgeting; managers are rarely working at the margin. Increasingly, the faunal enterprise (Giles and Nielsen 1990) will allow pecuniary expression of wildlife value. The rigged money-game is not the only one in town and a system using relative public benefits, the values attached to succession curves, may provide a much-needed alternative.
Multiple-use, as a public policy, rarely addresses costs. (It is difficult enough to describe the intended benefits from the uses.) Many forest plans have failed in gaining approval or in being executed because costs were ignored or assumed constant, or because simplistic benefit-to-cost relations were not examined. Results have been that, in some cases, timber sale income was below the cost of current timber value and the cost of producing the sale. A cost, the investment in forest improvement, is justified when the estimated discounted present cost (poorly perceived due to subsidies, market imperfections, etc.) is less than the increase in the expected discounted present value of the combined products or benefits produced. (See Pearse 1969:566.) Combining costs is fairly easy. A road with its multiple effects has a relatively easily estimated cost but secondary costs (e.g., erosion, fish loss, etc.) need to be included. Combining market and non-market products is difficult. Costs will be approximations if they can be made at all. Schuster et al. (1984) studied how non-timber resources cut into timber sale profits. Most were affected, but quantities were not estimated and comparisons between real costs and output market values may be interesting. Estimating costs will be very difficult because they are almost impossible to associate with an end result. Many investments (e.g., planting a tree) serve many resources or uses in the multiple-use mix. High and low costs, however, can be estimated. It is consistent with the procedure to make approximations of the median values of non-market products for groups of people. There will always be people for whom X is worthless, others for which X is priceless. Simulations can provide useful insight into what strategies will maximize these best-estimate valuations - of both costs and benefits2.
By using computers it becomes possible to discard the four above-named silvicultural systems and to decide on the right location for any stand within the tetrahedronal space of Fig. 7.33. A stand with unique characteristics (inputs) is to be managed for a set of
 |
| Fig. 7.33. Silvicultural systems, though only 4 by name, occupy infinite descriptive positions within the four dimensions of a tetrahedron. |
objectives. There are well known constraints such as species-specific soil suitability, tendencies of shallow-rooted exposed trees to be blown over, frost pockets, and insect and disease relations that amount to certain failures of trees. There are stocking guides, thinning guides, diameter-cut rules for profit making. In some areas elaborate silvicultural guides have been developed (e.g., Marquis et al. 1984 and Trimbel et al. 1974); computer programs are being developed by TVA, U.S. Forest Service, and private groups. Industries have had sophisticated prescriptive systems for many years that include known costs of alternative regeneration strategies, labor, and transportation. From the public (and increasingly the private forest owner) there is demand for improved systems to provide many forest benefits - from fish to viewscapes. Complex computer systems are needed to help prescribe the specific regeneration, thinning, management, and harvest strategy for each stand. Stands in a skillfully designed "rotation," which will become an irregular planned harvest sequence when the procedures of the aggregate curves is used, can comprise such a forest.
Burns (1983) assembled silvicultural papers on 48 major timber types. The scope of this task for timber is suggestive of the future tasks ahead for faunal system managers, for they too must articulate the range of practices most likely to produce fauna in each stage of each major forest type. Frustrations in developing the present book, throughout, have been in generalizing, for I have known all along that there is a forester in a station at the end of a road in a Douglas fir region asking "what, specifically, shall I do?" There is also a consultant for an owner of some of the 22 million acres of poorly stocked timberland in the south-central U.S. (Birdsey 1986) - and she is asking "what, specifically, shall I do?" Elsewhere, another asks skeptically "What is the relevance of your years of research to forest wildlife in my developing country's parks?" Hope barely exceeds despair.
Endangered Species Spaces
Managing the spaces of threatened or endangered species is like that for other life groups ... only more so. The emphases needed are:
The primary population-related general rules to accompany the above 8 faunal space-related rules are:
Faunal Space Evaluation
 |
| Fig. 7.34. Fauna are observed in many areas and the associated factors, x, y, and z are observed. Plotted, the animals only occur in the "cloud," the space depicted. Only considering X and Z factors, the animals appear on one surface. High density of points (sometime considered the best habitat) may be at the center, but not necessarily. Population occurrence is often highly skewed due to thresholds and other limits. In some cases the density distribution of animals in this factor space is uniform because of behavioral factors such as developed territories. |
"This area is excellent for turkey poults" is a spatial evaluation, expressed goodness. ("Badness" is hardly ever evaluated, partially because it is impossible to prove a negative.) Quality and primeness have been discussed, leading to an understanding of evaluation. There has been great confusion in using "analysis" and "evaluation." For the manager, the desire is always for evaluation. "I only need a diversity index!" Implied is that it will be used and that higher or lower diversity is of interest because of some objective. There are scientists who assert they measure and characterize animals and their spaces "just because." This is analysis. The manager may or may not have use for their results. He or she may measure and characterize spaces but these results will end up (if properly done) as input to a decision about faunal space goodness.
It is clear that faunal space is extremely complicated, multi-faceted, interactive as well as additive, and changing over time, sometimes quite rapidly. Very few managers of undomesticated animal populations are single-species managers (although they may prefer to be and sometimes act as if they are). The habitat that is good for life group A may be bad for B. Human values or not, spaces are species- or life-group-specific. Managers of bobwhite quail, for example, can look at a stand and mentally integrate length of edge, brush piles, distance of brush piles from edges, density of grasses and forbs for feeding, types of seed producing plants, probability of early spring seed supply, density of nesting sites, density of feral dog and predator tracks, then will look up, change the scale of analyses and judge whether the area is large enough, near other quail areas, and perhaps assess the probability of raptor influence. All of these can be measured, scored, and somehow integrated into a judgement of whether the stand is good for quail. The same can be done for other systems. (Expert systems are highly relevant.) The results can be very interesting (i.e., satisfy curiosity, related to or confound recent observations, etc.) but they are often of little use because they are single-species analyses, have wide confidence bands, and rarely can be tied to: what one or two factors at known cost can I change quickly and within-budget to get a desired positive response in quail numbers to meet a stated demand? (This is not asking too much!?)
The manager is involved with making spaces "better." In some cases, this means preventing losses or reductions in areas (e.g., by dams) or then quality (e.g., from toxicants). Habitat analyses serve well for such purposes. They describe spaces to be lost (or gained, since a powerline, for example, increases spaces for some, decreases them for other life forms). Translated to animal density losses or gains, then net effects can be judged as good or bad as when analyzing the alternatives in impact statements (as required by the National Environmental Policy Act of 1969). That evaluation, however, bears little relation to whether a land use will change. It documents the loss. Occasionally a less-bad-to-faunal- resource-objectives alternative will be selected. The pattern remains: loss of quality space within which forest faunal resource objectives may be achieved.
A forest project is proposed. An analysis is done. There are 37 bird species present now; there will be an estimated 37 species after the project. No influence!? Only 10 of the species are the same after the project (see Similarity Analysis). Is the project good? Is the habitat impacted (a word with negative or bad connotation)? Are the populations (animals used as the index to the habitat, or vice versa; a dangerous circle) impacted? Is the project good or bad? The habitat will be changed says the analyst; the habitat will be harmed says the evaluator. As described in Chapter 4, the analyses of the benefits - demand for creatures, extra creatures, all assigned value, expectancy, and potentially substitutable - are the dimensions for the valuation.
Many so-called habitat evaluations do not include likely changes, i.e., even though the future may not be predictable, at least inclusions of different age areas or stands may allow simple projections and deductions to be made. Knowing there are 500 individuals (+ or - 100) present today in area A is of little interest to most decision makers who know the project will be finished in 3 years and that its genuine positive effects will be felt in 15 years.
There has been atrocious waste of money on faunal impact studies, "habitat evaluations", because sample periods are too short (between-year variability is consistently great), sample sizes too small, variability among samples too great, no differences in importance of species specified (except that "endangered species" are usually judged to be infinitely important), and significant differences at excessively high confidence levels can rarely be determined. The money used in thousands of individual meaningless faunal "impact" studies could have paid, many times over, to provide for a comprehensive, nationwide biological survey or generic impact model (and yet can). There are always needs for field observations, for "grab samples," for quick studies, but elaborate field "scientific" studies will no longer fool many people or their courts, given that in the absence of "value criteria" the courts and other decision makers are going to work their magic on habitat analyses. Eventually, comparisons will be made between an unmanaged site to be developed (e.g., a dam (Giles and Dean 1990) and the post-developed site designed to create significantly more faunal (and other) benefits over the long run than existed or would have been experienced without the project.
Unless the faunal system manager actively incorporates public values into the concept of benefits as described in Chapter 4, there will be no need to pretend that evaluations are being rationally made (unless someone persists in discussing single-species management).
Reductionists rule, so it is reasonable to deal with single species, life-group or guild analyses (e.g., Table 7.16 and 7.17), the premise being that knowledge of what is good for each of them is needed. The analyses can follow those outlined above for succession. The best condition can be seen at the peak of the curve and it can be assigned a value of 1.0 or maximum value, the point at which conditions are best for the life group. Time is seen as important or integrative of other so-called environmental factors. The factors can thus be used to modify the slope of the curve. Environmental factors (f) have been called criteria, conditions, elements, and components. It is useful to see these, at least, as:
Pt+1 = f (F1,2,3 ... n)t Pt
the population P being a function of all these factors as well as itself in the immediate past.
Faunal spaces have been analyzed by every criterion imaginable. Partially this has been done in search of a single key factors. In other cases, key factors were assumed or known and managers have proceeded to quantify. Animals and plants both integrate many factors of the environment. When a relation between an animal or life group of interest (y) is found with another plant or animal (x), then most managers are happy with:
y = a + b x
| Table 7.16. Criteria for forming guilds among fish species (adapted from Terrell et al. 1982). |
|
| Table 7.17. One set of bird guild names (Droege and Sauer 1990). |
|
Grassland nesting
Wetland nesting Water fowl Woodland nesting Deciduous forest nesting Coniferous forest nesting Urban nesting Hunted Permanent resident Neotropical migrant Short-distance migrant Primary cavity nesting Secondary cavity nesting Open-cup nesting passerine Ground nesting Shrub nesting |
The managers can manipulate x or at least measure it and know the value of y (CAP110 or CAP71). Such models or equations are needed by the manager. They are fundamental tools for explaining, describing, and predicting faunal spaces. They are fundamental for evaluating costs because costs for every change in x can be related to y outputs.
The factors (F) in a list are often weighted by managers and expressed in terms of estimated relative importance or how sensitive the life group response is to each. Regression coefficients from the above analyses can be used for such weights but rarely are. The weighted lists can be used to compare areas or conditions grossly. The assumptions needed (e.g., that amount of use indicates preference or goodness) are enormous and usually erroneous because of changing life needs, changing availability of each factor, costs and risks to move, learned behaviors, barriers to movement, and territoriality. The analytical needs are those of expert systems or simply a tree-like decision structure that concludes that if F1 is present and F2 is present, or F10 is present, and F3 is less than Z, then...the habitat is "good." To argue ecosystem complexity and to treat it simplistically ... is simplistic.
Experts are important in faunal system work because of low funding for research and the difficulty of sampling over vast areas and the different conditions of the years. Evaluating spaces has been done in many studies. The 7 classes (Table 7.18) are useful for they can be easily understood, provide reasonable discrimination, and can be mapped. (Few people know what they would do with more fine-grained discrimination.) A rate of 0 to 10 can be used with all being compared to the highest quality assigned 10. Past efforts to use a scale of 0 to 100 have failed since most assigned values were made in units of 10. Assignment of values is varied due to definitions, past experience, unclear objectives (often species mixes), and familiarity with the specific area being evaluated. Schuster et al. (1985) described a Delphi approach to gaining more consistent evaluation.
| Table 7.18. A scale for grossly evaluating areas for fauna. |
|---|
|
I once thought I knew a forest well and white-footed mouse habitat best of all (Giles 1971). I placed traps where I "knew" I would trap mice. I selected the "best" sites. I could have done better if I had used random numbers to select my sites! That experience along with the observation that "wildlifers are the poorest hunters," the extreme variation observed in groups of experts doing evaluations, the tendency of experts to be over-confident, and the paucity of experts to do the field work over vast forested areas has led me to a strategy to separate field observations and valuation. This has been described in Teaford (1977) but originated with work with Rob Clements (formerly with the U.S. Forest Services) and was subsequently developed by Joe Coggin, Michael Fies, and Blair Jones of the Virginia Department of Game and Inland Fisheries. Part of the objective is to minimize training and to allow many people to participate in making observations of faunal spaces.
For example, if there are three categories of deer browse available and three classes of the amount of browse present that was utilized, the resulting 3 x 3 table produces 9 observable conditions - fair discrimination! Faunal system managers can independently assign a value to each of the 9 conditions. Which is the best of the 9 for deer? Assign it 10 or 100? Assign all other conditions relative values. The field forms are processed, the value matrix multiplied with each independent set of observations, and area represented by the sample multiplied. The results can be used to assign a quality - a browse goodness score to an area for deer. A relative goodness for deer or a needs-for-management computer map can be readily produced by conventional or computer methods.
Similar matrices can be developed for soft mast (soft fruits like grapes): a 3 x 3 matrix based on 1. none, 2. one species, and 3. two or more species with 1. zero, 2. sparse, or 3. abundant. These require the simplest, consistently-made observations possible but, in combination, provide mappable, useful discrimination. Weights can be changed to reflect 1 - species emphasis, 2 - research findings, and 3 - policy. Hypothesized changes can be produced given stand age. The observations are such that they can be easily and quickly done by people afield measuring trees. The observers are separated from the question of goodness (these are answered by experts independently creating the matrices) and they may be trained to concentrate on conditions and the physical realities of the site. Tendencies to bias observations are reduced because observers know data will be used in several ways with unknown weights and for several species, game and others.
Sometimes the manager needs to know the trend of forest conditions such as in grasses and forbs for all grazing forest animals. The usual questions about proportion of ground cover erosion, total biomass, and quality of biomass apply. In some areas, estimated proportions of the plant remaining will suggest future vigor but that may show up in pictures (including stereo pairs (Ratliffe and Westfall 1973)) taken from permanent points and from gross analyses of observations. Trend analyses require permanent records with pictures of each "condition" and permanent picture points. The rate of change over time (CAP110 and CAP86) is the statistic of interest. Total biomass may remain almost the same but plant species composition may shift, thus plant quality for the animals of interest. Lack of apparent change in photographs may lead uneducated people to the wrong conclusions about the trend.
The importance of deer throughout forests has led to many techniques for analyzing their habitat. Use rates determined from pellet-group counts (CAP07 and CAP40), sightings, radio telemetry, track counts, etc. are useful - more for habitat analyses than population estimation. Animal weight is often a good expression of habitat. This and other habitat- population phenomena are discussed by Kirkpatrick (1980). "Animals are integrators, plants are indicators" is a useful slogan in most situations.
Eventually, browse, forbs, and grasses have to be clipped to determine wet weight, dry weight, nutrients and energy. Once done (at high cost) then correlates can be established between stems counted, tree basal area, plot and transect work, and even observation of sites by trained staff. After observing the limits of the stem diameters to which animals will browse, clipping can be done to simulate browsing and thereby get the gross edible weight of plants. Some researchers point out that animals are selective in foraging. Consumption of annual production, not total biomass, will be measured for some species. In some areas, there are key plants, those well correlated with most of the others. Watching these plants will simplify the efforts. Rarely will one method be used. Comparisons of plants used by animals with these plants in animal exclosures can often be useful, but clipping inside an exclosure destroys its usefulness. They are costly!
Selecting the factors to study is a difficult task. Many statistical techniques are available for separating those that seem influential. Managers typically try to study the factors to which their populations seem most sensitive. See Green (1979) and Ludwig and Reynolds (1988). Density of animals can be related to the conditional probability (frequency of) occurrences of the factors within the areas. An expected pattern result has led to the procedure being called pattern recognition or PATREC (Grubb 1988) after the name of a probabilistic procedure used in the medical profession. A site is observed. If the fact is present, the conditional probability is entered. If absent, a zero is entered, all are multiplied together (the probability of F1 and F2 and ...). This "probability of sample conditions" (Grubb 1988:3) is converted to the posterior probability, grounded in Bayes' theorem. The result is the probability that a tract being studied has a high (or low) density. As in other analytical techniques "goodness" for a species is implicit in most studies. By working on the factors that influence the index, then the probabilities may change. The method has great potential for use in comparing developments, managerial plans, and individual practices. The limitations are in the extreme conditions required - high density or low, occupied or not. The same conditional probabilities can be developed annually with succession curves discussed previously so that each point along the curve is the probability that conditions will be perfect (based on concepts of maxima, see the Appendix and Wittwer 1974). By using conditional probabilities (judgements assigned by experienced observers and experts along with field measurements and frequencies), I am convinced that major strides can be made in creating a powerful multispecies habitat analysis system, appropriately responsive to the variable, dynamic nature of faunal spaces and the equally variable, dynamic nature of the human demand and value system to which all analyses must be attached if evaluation is to be done.
Within the forest, many things are correlated; few things independent. Statistical techniques usually require independence, rarely available. Faunal system managers need to turn this realization into system control. Basal area (the area of the stump surface of all trees if they were cut 4.5 feet off the ground), easily measured by a so-called basal area prism, gives a measure that can be related to age, height, canopy closure, and many factors such as elk thermal cover (Dealy 1985). Canopy closure is related to understory biomass. Bulk density of soil, fairly readily measured, is strongly related to many forest phenomena. Proportion of clay is the key variable in soils work and, it too, quickly works its way into explanatory equations. Weight of adult male deer is another easily-gotten value (simply measure the heart girth and convert it to weight, CAP411) which integrates many area phenomena. Stand density (stems/ha) (CAP650) will increasingly be shown to have a strong role, along with basal area and presence of 1, 2, or 3 layers in describing faunal space.
To understand an animal and its faunal-space needs, it is necessary to have a mean weekly (at least) description of its nutritional (proximate analysis) and energy needs; the potential foods available; the foods probably selected from those available; the nutrients consumed and the energy lost. The research requirements are very great (52 weeks x 15 factors) and the odds are small that the studies will ever be done. Alternative strategies are needed. (See Chapter 6.)
Habitat Evaluation Procedure, HEP, (U.S. Fish and Wildlife Service 1976)
has become a well-studied procedure, has evolved, and is in popular use (Ellis et al. 1978, 1979; Terrell et al. 1982). HEP was similar to the work of Lobdell (1972) and Teaford (1977) who developed the concept of a production function, i.e., "suitability of some factor depending on the amount or magnitude of that factor" (Schamberger and Farmer (1978)). Each factor (a curve (often a set of broken lines) representing each) is weighted as in the above approaches. This is called the habitat suitability index (HSI). Ellis et al. (1978) found that the season in which a field analysis is made makes a significant difference in the suitability score computed. Roberts et al. (1985) reported the status of various species-specific literature reviews as potentially used in HEP. Wakeley and O'Neil (1988) reported on techniques to improve efficiency in HEP use, and provided a brief overview of it.
... HEP is an accounting system for quantifying and displaying habitat availability for fish and wildlife. HEP is based on Habitat Suitability Index (HSI) models that quantitatively describe the habitat requirements of a species or group of species. HSI models use measurements of appropriate variables to rate the habitat on a scale of 0 (unsuitable) to 1.0 (optimal). In a typical HEP study, a number of evaluation species are chosen for each cover type in the study area. Species may be chosen because of their ecological, recreational, or economic value; because they represent groups of species (i.e., guilds) having similar habitat needs; or because they represent important habitats in the study area. ... Advice on selecting evaluation species is given by Roberts and O'Neil (1985a).
...After cover types in the study area have been mapped and evaluation species selected, habitat variables contained in the species models are measured from maps, aerial photographs, or by on-site sampling. HSI values are then calculated, and the number of habitat units (HUs) is determined for each evaluation species. One HU is equivalent to 1 acre of optimal habitat2 ; therefore, the number of HUs for a species is calculated as the number of acres of available habitat times its suitability (HU = HSI x area). For species that use more than one cover type, an aggregate HSI is determined for the cover types used and multiplied by area to obtain HUs.
2Not necessarily my preference: Optimum faunal space is an n-dimensional volume providing opportunities for a faunal group population's survival and reproductive success over at least 200 years, a space devoid of excessive natural events that frequently produce catastrophic mortality or reproductive failure.
HUs available for each evaluation species are estimated for each of several target years ... over the life of the project; estimates of future habitat conditions are made for both 'with-project' and 'without-project' alternatives. Project impacts are then estimated by calculating the difference in average annual habitat units (AAHUs) for each species. Depending upon the size of the project, impacts may be estimated for various alternative project designs, with or without management plans to compensate for losses in habitat value. Development of mitigation plans involving trade-offs of one sort of habitat for another may involve the use of relative value indices (RVIs) that express the relative importance or priority of the evaluation species or their habitats. (U.S. Fish and Wildlife Service 1980).
Especially in the modified form, the procedure can be very effective. Given the large number of forest animals, the high cost of developing a species index booklet, the ponderous rate of development, and the continuing criticism or skepticism about the models, a less precise approach may be appropriate. The need is to realize the confidence requirements in each situation. The need is to develop a system that can mature if allowed. The need is to use the HSI models to simulate how the best decisions are now made, not to represent perfectly nature. Decisions will be made, immediately, and a perfect system for evaluating habitat will not be available - and perhaps never will. The subjective probability used in TVA's wildland resource allocation procedure (1972) and its succession curves (Giles 1970) are isomorphic with HEP's suitability indexes. The procedure has become very elaborate; user-friendly computers can now be used to eliminate the apparent complexity and programs have been made available. Nationwide suitability curves are not likely to produce consistent results. Abiotic factors, interspersion, etc. will eventually be used as modifiers of these curves and all will eventually become excellent full-scale, complex simulation models, not the present, largely-additive, largely point-in-time models. They will address the manager's real question: "If I do X, what will be the probable consequences (impacts, gains, etc.) to all of the fauna for which I am responsible over at least 50 years? What are the consequences compared to those on the area without my action, to another area, and to the area subjected to a high-intensity faunal management program?" Eventually change in potential human benefits related to species will be seen as the product being produced. This is closely related to abundance but since many species are usually managed, eventually relative species values to people and their likely change must be included.
Faunal Space Model Invalidity
How well does a mean (an average) represent a whole group of field observations? Probably not very well, but you get an idea based on the statistic called the standard deviation. With a mean (average) of 10 and standard deviation of 2, the forest observer will be fairly safe in saying that most observations (about two-thirds) in the future from the same population will be from 6 to 14. However, about one-third of the time they will be less than 6 or more than 14! Once a person calculates a mean, strange epistemological vapors invade the brain. If the value of the mean does not occur subsequently, then something seems to be wrong. The techniques used must be flawed, the model wrong!
Not anti-statistics, not anti-sampling, not anti-field observations, I must make these disclaimers before attempting to present a very complex concept, one that appears to contradict much analysis in faunal systems - both of populations and their spaces.
R = P-E-T +_ S
is a good model. It is consistent with what is known about the real world. R is runoff, P is precipitation, E is evaporation, T is transpiration, and S is storage or contributions from springs. Thinking through the system, analyzing the major components, and developing a parsimonious relationship seems like a reasonable activity. Throwing in numbers to such a relationship (CAP5040) may never produce a real-world match up, yet the model is good, it has served useful purposes (including those for thought, analytical framework, simplification, preparing a basis for hierarchical development, and providing a communication and teaching medium). Does it represent reality? All good models do not have to represent reality!
Holling (1978:95) observed that the central tenet of the modern scientific method is that hypotheses or models "can never be proven right; they can only be proven wrong. Thus, demands for validation are unsound. The task, is to state the objectives for the model or model building work very precisely, then to study how well they appear to have been achieved and where failures occurred. If reality is the criterion then invalidation the task ... setting the model at risk so as to suggest the limits of its credibility" (Holling 1978).
By using a concept of heuristic convergence, the light shown by all of the epistemological bases on the model, what we know can be ascertained. We can never really know anything except in some metaphysical sense.
| Quick Access to the Contents of LastingForests.com |
|---|
This Web site is maintained by R. H.
Giles, Jr.
Last revision January 17, 2000.