This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. FOREST PRODUCTIVITY: LESSONS TO BE LEARNED Richard G. Cline Nelson S. Loftus ABSTRACT This paper is a discussion of our research progress in increasing forest productivity relative to that in cropland agriculture. Key differences seem to be due to the intense cooperation among various agronomic specialties, their ability to measure product yields over short periods of time, and the availability of data sets containing very specific crop / species, soil, climate, and fertility information. The lessons to be learned from field-crop production research include the value of ecological concepts, a focus on basic processes and controlling factors, the value of observation, the need to be active rather than reactive, and the advantages of working together. INTRODUCTION Forest productivity seems at times to be looked upon as a special, and somehow unrelated, aspect of agriculture. This view seems a bit narrow. Forestry, after all, is based on essentially the same biological processes as the rest of agriculture. This being true, what can we do to take advantage of the relationships? First, let us assume agricultural production (including forestry) is based primarily on our ability to retrieve a plant or plant-derived product from our environment. What we use it for after that is a related but secondary question. We could use it to feed ourselves, to feed animals that we subsequently eat, to produce fiber that we use, or to produce a recreational environment to be enjoyed. All of these things have value. We harvest them all in one form or another. For the sake of argument and comparison let us look at plant products that can be produced on the land and sold after harvest. Let us compare productivity, and our understanding of it, in cropland agriculture with that in forestry. We have a long history of production in both. Our ability to achieve production improvements, however, has been much more impressive in our croplands than in forestry. Why? The first general thought that comes to mind is historic. We have a rich history of intense cooperation among the various agronomic specialties in cropland agriculture. This has produced some truly impressive synergisms, particularly among the fields of crop science, soil science, Paper presented at the Symposium on Management and Productivity of Western-Montane Forest Soils, Boise, ID, April 10-12, 1990. Richard G. Cline and Nelson S. Loftus are Soil Scientists, Forest Environment Research Staff and Timber Management Research Staff, respectively, Forest Service, U.S. Department of Agriculture, Washington, DC. and genetics. The yield of wheat, for example, has more than quadrupled in the past 50 years. This cooperation is one of the main reasons for our rapid increase in crop production during the last half century. This kind of cooperation has often been nearly absent from the field of forestry. Why? Are the growth processes somehow different in forests than in field crops? That does not seem very reasonable. All plants operate pretty much the same way within certain limits. The same basic principles should apply if one allows for species and culture differences. COMPARING MEASUREMENT Maybe we can shed some light on this problem by comparing production measurement in these two areas of agriculture. Table 1 attempts to compare the process of production measurement in croplands with that of forests. While the comparison is not absolutely a true one for all cases, it seems correct in a general sense. The table suggests many more similarities than differences. This seems logical. The differences, while apparently few, are glaring, at least in practice. Table 1- A comparison of productivity measurement in cropland and forest agriculture Measurement considerations Cropland Forestland Measure Harvested product (grain) Harvested product (fiber) When measured Yearly Over the rotation What measured Mass (weight) Volume (when sold) Mass (weight) Volume (when sold) Measurement method Scales (grain elevator) Scales Volume measurement Volume estimates Measurement context Species, often variety specific Often not species specific Location specific (known soil, fertility levels) Not location, soil, or fertility-level specific Cost vs. return usually well known Cost VS. return often difficult to determine because of rotation length The harvested product seems like an obvious choice for a first view, and should provide a good basis for evaluation, providing good measures of the harvested product can be obtained. Similarities are much more obvious than differences here. The period of measurement is a bit more of a problem. Field crops are usually harvested during the same growing season they are sown, and reasonably reliable measurements can be obtained at that time. Forest products, trees for this example, are harvested over a rotation that can be 80 to 100+ years. At other times, field measurements and estimates with varying degrees of reliabili ty are used. Harvests can range from partial cuts through thinnings, where residual volumes are estimated, to clearcuts where all trees are removed and merchantable or total volumes are estimated. This is an important difference even though the units of measure might be quite similar-for example, tons of grain and tons of green wood. Another major difference is the measurement context. It seems the context of measurements in forestry is usually much less specific, making it more difficult to obtain useful information. A summary of this comparison of cropland and forest agriculture makes three points: Obviously the returns for developing this kind of knowledge and understanding for our intensively managed croplands are much greater than they are for our forestlands. It is logical, then, to expect more effort in those areas where returns are greater. This is likely to continue in spite of increasing public attention on and demands for resource returns from forested wildland environments. We will need better data and firmer understanding of these environments to form the inferential foundations for future management decisions as these demands increase. WHAT CAN WE DO? It seems we have a challenge before us. We must overcome a limited, by comparison, availability of research data, probably research funding, expertise, and the continuity of that expertise in the future. We must be more efficient-use the data we have more efficiently and creatively and develop a more integrated understanding of our resource base. There are a number of specific things we can do. They are not new. They have been suggested before, but we probably have not made adequate use of them. Maybe, fuddy duddies that we are, we have trouble forcing ourselves out of the rut in which we happen to find ourselves. Here is a short list of some things to try, or try again: -We in forest science seem not to have brought the full range of our scientific weapons to bear on the productivity problem as has happened in cropland agriculture. We tend not to work together. The reasons are not important. We need to do it. The potential progress and benefits can be significant if the results in other areas of agriculture are any indication. - Use ecological concepts to establish reference points and a basis for data extrapolation and interpretation. We have tried this before, often with considerable success. Habitat type classifications and their growing utility are but one example. There are a variety of other possibilities. We also need to recognize, however, that classifications are all artificial constructs. They are tremendously valuable as a communication tool and as a method of extrapolating inference through inductive reasoning. The fact remains, however, that being artificial constructs of our momentary and ever-changing concepts, they need to be constantly challenged, tested, and revised. Their value is that they are capable of helping us organize our thoughts about complex landscape-based ecosystems, isolate the properties of those systems most useful to us, and facilitate our efforts to overcome our lack of adequate empirical data. -Our inability to measure our product reliably at short intervals hampers progress. This seems like a major problem. The long rotation length of forest crops makes it much more difficult to accumulate the largely empirical relational data sets critical to developing working mechanistic hypotheses. These hypotheses and the research that they generate provide the specific understanding needed to take advantage of the array of scientific synergisms seen in field-crop production. -Our lack of data sets containing sufficient species and growth environment (soil, climate, fertility) specificity is another problem. The problem seems different, but its effects are essentially the same as those related to long rotations. We do not have the empirical data so necessary to developing our concepts of mechanistic and process reI ation shi ps. -We need to focus on the basic processes and factors controlling productivity. This is actually a corollary of the previous point because the ecological concepts, if correctly conceived, will be related to those basic factors and processes. Our ability to correctly identify these things will be considerably enhanced by forming the alliances, previously mentioned, that have been so successful in cropland agriculture. Having increased our productive capacity for food, why should we ignore the potential for doing the same for forested environments, whether we are producing fiber, wildlife, or scenery? Research to determine productivity or response to treatment has historically depended on field-plot-based experiments. Results of this work have been used to infer the effects of management inputs on outputs given a defined set of resource properties. Our history in the more extensively managed forest environments has been less successful than that in croplands and more dependent on axiomatic information. Our efforts have been hindered by the difficulty of adequately controlling experiments in a highly variable wildland environment. This is not to suggest that we have no good research results or that there is no empirical data base to work from. We do have data, and good research has been done. - We need to relearn the value of observation. There is a strong tendency in the scientific literature to discount anything that does not have a number or a statistical procedure attached to it. Research should never be undertaken without adequate statistical preparation, but 2 regression, for example, does not prove cause and effect. Some of the most useful literature we have seen has been among the older articles where conclusions were more dependent on observation and where the logic connecting observation and conclusions seems better developed than is routine with our more sophisticated statistical procedures. Statistics are not a substitute for understanding what is observed of nature. properties we have been discussing. I would agree that the taxonomy has extrapolative value. It is also an excellent vehicle for communicating commonly understood ideas. It is, as are all classifications, a substantially artificial system reflective of the thoughts of its framers at the time it was constructed. You might be aware that we are continuing to change it as we learn more about soils. Another point concerning the soil taxonomy seems important. It is, on the whole, a general system designed to reflect concepts of the whole population of soils as we presently understand them. This seems to be its main objective. We need that kind of an overall view to help us define and communicate ideas in common. Mapping problems and their associated local interpretations require a different objective, consequently a different classification design. It is basic to the logic of classification systems that, for optimum utility, the system design should change when its objectives change. I will provide two examples using taxonomy as a reference to make this point. The taxonomy uses 35 percent rock fragments as the class break between skeletal and nonskeletal families. Remember that this must be imposed downward on the series, taken as a taxonomic class. The field criteria for map unit design in a survey with which I am familiar placed the class break for map units near 20 percent. This suggests that the practical interpretive criteria did not fit the taxonomic criteria very well. The solution was to ignore the taxonomic criteria for survey purposes and handle the two taxonomic classes involved (20-35 percent and 35 to something over 55 percent) as similar soils. Another instance of map unit design at variance with taxonomic criteria relates to classes of Andepts (35 cm of andic, volcanic ash, material at the surface) and andic subgroups (18-35 cm of andic material). This material has some unique and valuable water-retention characteristics. These soils on steep south aspects seemed to need about 25 cm of this material to support certain types of productive timber stands and provide assurance of stand regeneration. The map unit design described Andepts and andic subgroups with ash caps greater than 25 em as similar soils in some map units and andic subgroups with typic subgroups (less than 18 cm) as similar soils in others. The same kinds of soils on steep north aspects in the same survey were not separated at the same boundary (25 em ash cap), because the water relations were not the same. The point of all this is that, with changing class separation objectives, class boundaries need to change. This is just fine for local interpretive purposes. The more general, formal taxonomies provide us with our conceptual and communication tools. They need the stability associated with consistently established class definitions. We need to recognize the value each has in our scheme of things and avoid making ourselves slaves to either. We build map units and interpretations from the soil properties and associated biological and physical responses we observe. We develop concepts and communicate using consistently defined sets of properties that we understand in common. -We need to resist the temptation to allow crisis to drive programs. Too often, we are driven by short-term objectives. This tendency is not going to go away. We will all succumb to the short-term imperative at times, but we must make sure, through it all, that we maintain (reestablish where necessary) the utility and value of long-term data. This is particularly important when the crop we grow is a long-term crop. If the long-term experiments of Rothamsted and other places are valuable for short-rotation species, think how valuable they must be for species with much longer rotations. - We need to take a lesson from our colleagues in cropland agriculture. We need to take advantage of the combined experience of the variety of specialty areas wi thin both forestry and agronomic sciences. We need to work together more than we do now. The discussion, to this point, has concentrated on commodity production by plants. Forests have much more to offer. Recreation opportunity, esthetics, and high-quality water are but three examples. The demand for these resources, and their value, seems to be increasing. This trend is expected to continue. We need to develop better productivity relationships for these resources because we are being asked to produce them. Our management systems need to provide reliable estimates of our ability to produce a complex array of values. Speakers answered questions after their presentations. Following are the questions and answers on this topic: Q. (referred by Glen Klock)-About 95 percent of the talks in this "Forest Soils Conference" have not used any reference to soil taxonomic units. Are we really using soils as currently defined in a taxonomic system, or do we get enough soil direction by characterizing such factors as soil compaction, soil erosion, nitrogen and carbon contents, and mycorrhizal conditions? If generalizations are used, how do we extrapolate data and communicate with others? Are we really preaching to the converted? A. (from Cline}-I will answer the last part of the question first because it is the easiest. Yes, we probably are preaching to the converted, especially at a conference like this attended mostly by resource professionals and most of those probably soil scientists. It is unlikely to be otherwise in this situation. The value of the conference of course, is not the preaching. It is the exchange of ideas and the publication we can use to inform others. Concerning the apparent lack of taxonomic class reference in papers presented, I would suggest that taxonomic classes are only useful if their specific interpretive value is known and well documented in relation to the soil 3