Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
15 those categories contribute to the formation of a view that helps to put air-pollution stress in perspective with other stresses, provide a view of the current status of forests, and contribute to evidence that air pollution is affecting forests. For example, in 1986, an NRC committee reviewing acid deposition used the criteria of consistency, responsiveness, and mechanism to infer causal relationships among sulfur emissions, sulfate aerosol concentrations in the atmosphere, decrease in visibility, chemical makeup of rain, ant] acidification of surface waters in sensitive regions (NRC, 1986). The more nearly these criteria are met, the stronger the inference becomes. The criteria suggested are reasonably simple and comprehensive, and they are similar to those often used by epidemiologists to separate causes from chance associations in studying human diseases. In the use of biologic markers, documentation of change often is a critical first step in determining that a problem might exist. A marker might be specific enough to serve as a reliable indicator of system dysfunction, but not specific enough to support the inference of a causal relationship by itself. In that case, it can be used with appropriate information to test for more specific relationships with potential causal agents. The distinctions between general, integrative biologic markers of stress and markers that might be more suitable for discriminating among potential causes are discussed in the following sections. USING MARKERS IN SURVEYS AND EXPERIMENTAL STUDIES Three types of studies are being used in forest air-pollution research, as described by the papers in Part II: · Surveys of damage coupled with measurements of pollutants. · Controlled-exposure tests to determine dose-response relationships. · Experiments to determine mechanisms linking air pollution to specific effects. No approach alone is sufficient to establish cause-and-effect relationships for instances of forest damage observed in the field. Results of the three types of studies can be used together for assessing cause-and-effect relationships in the analysis of regional-scale changes in forests. Surveys of Stress and Damage In addition to reporting the number and species of trees showing visible injury, decreased growth, or mortality, damage surveys or monitoring programs also provide information on spatial or temporal trends in forest-canopy variables. Properly designed surveys can be used to determine the strength of an association between suspected causal agents and degree of damage in forests, determine the degree of spatial and temporal consistency, and define biologic and environmental gradients. Causality is supported if, for example, specific airborne chemicals (or other stress factors) are present consistently when or shortly before a change in condition or a measure of dysfunction in forests occurs. If a spatial or temporal gradient in the amount of injury or damage is observed along a corresponding gradient in exposure, causation is further supported. Surveys are strictly correlative and can be misleading if the suspected causal factor is not well measured and placed in temporal perspective. The strength of the inference is decreased if--as is oftmn true--the suspected cause is not specific to the observed effect. Because numerous natural and anthropogenic stresses occur
16 together, it is rare for forest damage or dysfunction to correlate well with only one potential agent or cause. Among the biologic markers reviewed in the workshop papers, the committee found the following to be especially appropriate for survey and monitoring work: visible symptoms (as discussed by Anderson, Barnard, Kape, Miller, and Schutt), changes in leaf morphology and histology (Miller and Schutt), tree-ring changes (Cook and Innes and Bondietti, et al.), nutrient balances (D.W. Johnson et al., Tingey, and Waring), stream chemistry (D.W. Johnson et al.), and indicator plants, including lichens (Weinstein and Laurence and Scott and Hutchinson). Some methods or techniques reviewed in the workshop appear to have consideral potential for use in survey and monitoring and deserve further research support: remote sensing of leaf morphology and biochemical changes (Rock et al.), study of resource allocation of carbon and other elements (Bondietti et al., Jones and Coleman, McLaughlin, Shortle, and Waring), and use of some metabolic characteristics (Norby, Richards, and Winner). Controlled-Exposure Studies Controlled-exposure tests range from simple comparisons of effects on otherwise healthy plants exposed to one suspected causal factor at a time to complex experiments in which plants are exposed to various combinations and amounts of suspected causal factors. Most controlled-exposure tests are performed with seedlings or small saplings of a single species in growth chambers, greenhouses, or open-top chambers. Experimental tests in which branches of mature trees are protected from exposure to pollutants or are exposed to known amounts of pollutants are becoming more common. The results of controlled-exposure tests are usually expressed as dose-response curves or as the minimal (threshold) exposure needed to induce symptoms of injury or dysfunction in the plant species under study. To satisfy the criterion of responsiveness, a quantitative relationship must be found between degree of exposure and extent of injury or dysfunction. Dose-response curves can be difficult to interpret, because the response of a plant to various air- pollutant exposures can be influenced by other biotic and abiotic environmental factors, including other kinds of stress. Controlled-exposure studies should be used to examine the interactions between air pollutants and other stresses. The more differences there are between the controlled environment and the field situation of interest, the more difficult it is to extrapolate from the experiment to the field. Most of the biologic markers mentioned above as appropriate for survey and monitoring studies should also provide valuable information if used in controlled- exposure studies; the results of the latter studies should improve understanding of the linkage between the effects measured by such markers and their causes. Additional markers reviewed in the workshop papers that the committee considered to be particularly appropriate for use in controlled-exposure studies are: lichens and other plant indicators (discussed by Scott and Hutchinson and Weinstein and Laurence) and certain secondary metabolites (Jones and Coleman). Other markers discussed during the workshop that showed promise for use in research to determine cause-and-effect relationships include aluminum mobilization (Bondietti et al. and Shortle) and leaf pigments (Heath and Richardson et alit Experiments to Determine Mechanism In establishing cause and effect, it is important to determine whether an observed change or dysfunction resulted from one or more suspected causal factors through a
