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.
SUMMARY INTRODUCTION This report reviews current knowledge about man-made causes of changes in concentrations of stratospheric ozone and the effects of those changes. Recent reports of the National Research Council (NRC 1975, 1976a,b, 1978, 1979a,b) have treated the chemical and physical aspects of potential reductions of stratospheric ozone in detail. Part I of this report reviews recent develop- ments on that subject. Part II deals with the effects of reduction of stratospheric ozone on humans, other animals, and plants, independently of what might cause the reduction. CHEMISTRY AND PHYSICS OF OZONE REDUCTION s determined The abundance of ozone in the stratosphere i_ by a dynamic balance among processes that produce and destroy it and transport it to the troposphere. According to current understanding, the most important photochemical reactions regulating ozone involve molecular and atomic oxygen and various radicals containing nitrogen, hydrogen, and chlorine. All of these compounds have natural sources, but their concentrations in the stratosphere can be significantly altered by human activities. The human activities that have thus far been identified as potentially influencing stratospheric ozone are as follows: The release of gaseous chlorinated carbon compounds, mainly chlorofluorocarbons (CFCs) and methyl chloroform (CH3CC13). CFCs are used as foam-blowing 1
2 agents, as working fluids in refrigeration systems, and as propellants in aerosol sprays. Methyl chloroform is an industrial solvent. These gases decompose in the stratosphere providing a significant source of radicals that contain chlorine. · The release of nitrous oxide (N2O) from combustion and its enhanced release from soils and waters as a result of various agricultural and waste management practices. Nitrous oxide decomposes in the stratosphere, introducing radicals that contain nitrogen. · The direct input of nitrogen radicals to the stratosphere due to nitrogen oxides (NOx) in aircraft engine exhausts. · The increased abundance of carbon dioxide (CO2) in the atmosphere due to combustion of fossil fuels and deforestation. Increased carbon dioxide has a subtle influence, causing the temperature of the stratosphere to decrease, which leads to increased stratospheric ozone, and changing stratospheric concentrations of water vapor. Key Findings and Conclusions Over the past several years, research, driven by discrepancies between theory and observation, has led to considerable improvement in our understanding of the effects on stratospheric ozone of releases of CFCs and oxides of nitrogen. As a result, previous discrepancies between the estimates of models of stratospheric processes and observed concentrations of certain important species have been reduced. Important discrepancies still remain, however, which means that there are still uncertainties inherent in the results of modeling exercises. Current scientific understanding, expressed in both 1- and 2-dimensional models, indicates that if production of two CFCs, CF2C12 and CFC13, were to continue into the future at the rate prevalent in 1977, the steady state reduction in total global ozone, in the absence of other perturbations, could be between 5 percent and 9 percent. Comparable results from models prevalent in 1979 ranged from 15 percent to 18 percent. The differences between current findings and those reported in 1979 are attributed to refinements in values of important reaction rates. Also, as an example, if the atmospheric concentration of N2O were doubled in the absence of other perturbations, total ozone would be reduced by between 10 percent and 16 percent. Although
atmospheric concentrations of N2O appear to be increasing, we cannot reliably project the future course of N2O emissions. Steady state reductions in both these cases would be reached asymptotically in times on the order of a century, although the assumption of doub- ling N2O concentrations is unrealistic on such a time scale. The effects of perturbations by CFCs and N2O are not additive, so the estimates of effects of combined perturbations require investigation of specific cases. These results should be interpreted in light of the uncertainties and insufficiencies of the models and observations. For example, other chemicals released from human activities are understood to have the potential for affecting stratospheric ozone. Examples are methyl chloride (CH3C1), carbon tetrachloride (CC14), and particularly methyl chloroform. Observations of critical species need to be extended and confirmed by a number of measurements using independent techniques. Important assumptions in the models about rate constants, distributions of certain species, and the reactions taking place need to be tested. Furthermore, three important discrepancies between models and observations remain to be resolved: More chlorine monoxide (C10) is observed at altitudes above 35 km than is predicted, the behavior of NOx in winter at high latitudes is unexplained, and concentrations of CFCS in the lower stratosphere are lower than the models suggest. We anticipate that research on these problems in the field, in the laboratory, and in theory currently under way, planned, and proposed will lead to continued improvement in understanding, resulting in further reduction of the remaining discrepancies between theory and observation. In particular, simultaneous measurement of the important chemical species as a function of altitude and latitude by various methods should prove critical to improving understanding during the next several years. Examination of the historical record of measurements of ozone does not reveal a significant trend in total ozone that can be ascribed to human activities. This observational result is consistent with those of current models, since no detectable trend would be expected on the basis of current theory. Because data on total global ozone cannot be analyzed to distinguish among causes of ozone changes, total ozone data alone cannot be relied upon for early detection of an anthropogenic change. Measurement of the spatial and
4 temporal distribution of critical trace species and ozone, together with theoretical modeling taking into account all the major influences on stratospheric ozone, offers promise of understanding the causes of ozone changes and the consequences of alternative actions in response. Recommendations 1. The national research program, including atmospheric observation, laboratory measurements, and theoretical modeling, should maintain a broad perspective with emphasis on areas of disagreement between theory and observation. Highest priority in research should be given to a coordinated program to understand the spatial and temporal distributions of important species, such as C10 and the hydroxyl radical (OH). 2. The global monitoring effort should include both ground-based and satellite observations of total ozone and concentrations of ozone above 35 km, where theory indicates the largest reductions might occur. Sound, satellite-based systems for stratospheric observations are essential. 3. Potential emissions of N2O, CO2, CH3CC13, and other relevant gases should be assessed and their consequences for stratospheric ozone evaluated. Models should be developed to describe the consequences for stratospheric ozone of future emissions of these gases. BIOLOGICAL EFFECTS OF INCREASED SOLAR ULTRAVIOLET RADIATION Stratospheric ozone acts as a shield to screen out much of the short-wavelength ultraviolet (W) in sunlight. Slight changes in this ozone layer may result in large changes in the amount of damaging UV striking the surface of the earth. Living creatures have adapted to the present level of UV and to its fluctuations from season to season and during the day. Part II of this report gives the current state of knowledge about the effects on biological systems of an increase in W resulting from a decrease in stratospheric ozone concentration. Each of the findings and conclusions summarized below has important implications for future research--either in efforts to decrease the uncertainty in concepts or in
5 efforts to increase quantitative knowledge. These research implications are spelled out in our list of major recommendations. Recent advances in knowledge since the last NRC report on the subject (NRC 1979a) have clarified our view of the problem but have also pointed out scientific areas not emphasized in earlier reports that confound the simple prediction of the effects of ozone depletion on biological systems. The unraveling of these difficulties will be accomplished only by a research effort directed by knowledgeable scientists, especially photobiologists. In many instances, we are still not sure of the scientific questions to be asked. Similar comments were made in earlier NRC reports (NRC 1975). The fact that they have not been acted on with any reasonable financial commitment accounts for a large part of our inability to make better predictions. It seems certain that more than 90 percent of skin cancer other than melanoma in the United States is associated with sunlight exposure and that the damaging we nabs are in the UV-B region (290 nm to 320 nm) of the spectrum. A decrease in ozone will be accompanied by a well-predicted increase in W -B. We estimate that there will be a 2 percent to 5 percent increase in basal cell skin cancer incidence per 1 percent decrease in stratospheric ozone. The increase in squamous cell skin cancer incidence will be about double that. Where this range the value falls depends on which theory used to make the estimate and on the appropriate dosimetric data used. The predicted increases are appreciably greater at lower latitudes than at higher. Although the incidence of malignant melanoma increases with a decrease in latitude, the degree to which sunlight in is is responsible Is not apparent, and there are few data implicating UV-B as the only responsible wavelength region. Therefore it is not appropriate to make quantitative predictions about the increase in the incidence of this disease associated with a decrease in ozone. Some of the difficulty in making quantitative predictions about humans comes from uncertainties (even in simple cellular systems) about the effects of inter- =~ ; an c amine ~ i nays wavelengths in a broad band, such as an ~ ~ w~= ALVIN -_~ _~ ,, 7 _ in the ultraviolet of sunlight, in producing antagonistic or synergistic effects. Moreover, it has been learned only recently that rapid repair of sunlight damage to human skin takes place during irradiation. An appreciable fraction is photorepair mediated by visible light, and a
6 similar phenomenon seems to take place in anchovy populations. The quantitative magnitudes of such effects are not known. The effects of ozone depletion on other animals and plants in the biosphere are as important as the direct effects on human health. However, scientists are still not able to predict quantitative effects on crop plants or ecosystems. The details of our findings and recommendations are spelled out in Chapters 3, 4, and 5. Key findings and conclusions and major research recommendations have been extracted from the chapters and are listed below. Estimates are given, where possible, of how long the recommended research might take under ideal circumstances Key Findings and Conclusions Molecular and Cellular Studies (Chapter 3) 1. Deoxyribonucleic acid (DNA) is probably the primary target in animal cells for most deleterious effects of UV-B, especially effects involving mutagenesis and neoplastic transformation. Other targets of possible biological significance for UV-B effects include membranes, ribonucleic acid (RNA), and proteins. 2. The spectrum for absorption of energy by DNA for wavelengths in the W-B region and the spectra for biological damage to DNA as a function of wavelength (action spectra) are known. The absorption spectrum and the action spectra are similar but not identical, probably because long-wavelength light is absorbed in some components of this genetic material that are not effective in changing the structure of DNA. The action spectra in the UV-B region for affecting mammalian cells (killing, mutation, and neoplastic transformation) are similar to those for damaging DNA. 3. The formation of pyrimidine dimers (bonds between pyrimidine residues in one of the two strands of DNA that distort the normal DNA helical structure) appears to be the major injury to DNA from W-B irradiation. 4. There are major interactions between the effects of UV-A (320 nm to 400 nm) and those of W -B on DNA in cells. Some of these are antagonisms, whereby W -A effects significantly reduce or repair the W -B damage. Except for photoreactivation, which involves enzymic splitting of pyrimidine dimers back to normal single
7 residues mediated by W-A and visible light, these interactions are still poorly understood. 5. In excision repair, dimers are removed from one strand of a DNA double helix by enzymes that work in the dark, leaving the unaltered strand as a template for reconstitution of a new normal strand. Photoreactivation and excision repair of pyrimidine dimers occurs rapidly in human skin. Ecosystems and Their Components (Chapter 4) 6. Both W -A and W-B have been reported to be detrimental to plant growth and development and to a number of physiological processes of plants, when examined under non-field conditions. The adaptability of plant species appears to be sufficient, under current ambient levels of UV-B, to maintain food crop yields. The potential for further adaptation to predicted increases in ambient W-B is not known. 7. Ambient W -B at present levels or similar levels in the laboratory can damage sensitive aquatic organisms or stares in their lifecycles that occur at the water's _ surface. Natural populations of aquatic organisms have adapted to current W-B levels so as co Relax ~ ~. ~= reproduction potential. In the case of anchovy larvae, it has been demonstrated that photorepair of W -B damage is effective even at UV-B levels significantly higher - than those that would result from predicted ozone depletions. Photorepair may be a general adaptive mechanism of organisms evolving in the presence of W -B. Currently, there is no information from which to predict the magnitude of adverse effects of enhanced W-B on . . aquatic organisms. 8. From limited field experiments on terrestrial plants and laboratory experiments with captured or cultured aquatic organisms, it appears that different species of both plants and animals have different sensitivities to increases in UV-B above current levels. Changes in species compositions and abundances of organisms have been observed In s~mu~acea aquatic ecosystems subjected to enhanced W-B. Mathematical models show that in systems subject to large natural oscillations in the size of the Copulation, there are severe limitations on the minimum population density needed to maintain a species. However, the data currently available on food chains in the natural
8 ecosystem are not precise enough or complete enough to be used to predict population dynamics or the displacement of an individual species under current environmental conditions. It is doubtful therefore that a statistically significant causal relationship between increased W -B levels and food chain success can be predicted in the near future. 9. Only minor effects of increased W-B levels are predicted for animals used for human food. Direct Human Health Hazards (Chapter 5) 10. A reduction in the concentration of stratospheric ozone will not create new health hazards, but will · · . Increase existing ones. Effects Other Than Cancer 11. There is evidence that direct acute effects of UV on humans, such as sunburn (acute erythema) and corneal inflammation (photokeratitis), are linked more strongly to W -B than to W -A. 12. Acute erythema and photokeratitis can be predicted accurately for a given dose and spectrum of UV-B, since the action spectra, dose-response curves, and intensity-time reciprocity relationships are known. 13. Ultraviolet radiation affects many aspects of the immune system of animals and humans. Allergic contact dermatitis, skin graft rejection, tumor susceptibility, and function and viability of individual circulating and noncirculating cells of the immune system can be altered, primarily by W-B. Skin Cancer Other Than Melanoma 14. Data on the relative incidence rates of basal and squamous cell cancers in highly pigmented (black) versus lightly pigmented (white) persons indicate that more than 90 percent of skin cancers other than melanoma in U.S. whites are attributable to sunlight. 15. Molecular, cellular, and whole animal data all implicate UV-B as the major carcinogenic component of sunlight for skin cancers other than melanoma. The evidence is stronger for squamous than basal cell cancers because animals rarely get basal cell cancers. In
9 humans, basal cell cancers are virtually all related to sunlight. 16. Based on animal studies, UV-B is implicated not only as an initiator of carcinogenesis but also as a promoter (in the general sense and via indirect effects) of chemical carcinogenesis. With the current state of knowledge, it is not possible to assess the extent to which increasing exposures to chemicals would result in increases in skin cancers due to synergism, over and above any increase because of increased UV-B exposure alone. 17. A 1 percent reduction in the amount of strato- spheric ozone is predicted to give an approximate 2 percent increase in biologically effective UV-B. Epidemiological data suggest that a 2 percent increase in UV-B would give a 2 percent to 5 percent increase in basal cell skin cancers. For squamous cell skin cancers the increase would be about twice these values (4 percent to 10 percent). 18. The risk of developing skin cancers other than melanoma and the increased risk due to increased exposure to UV-B could be mitigated by individuals through changes in lifestyle that would reduce exposure. Melanoma 19. The incidence of skin melanoma appears to depend on latitude, an indication that sunlight is a contributing factor. Circumstantial evidence such as occupational differences and location of the cancers on the body suggests, however, that exposure to sunlight is only one of several factors. The association between sunlight and melanoma is not strong enough to make a prediction of increased incidence due to increased exposure to UV based on epidemiological data. 20. The only evidence that suggests UV-B causes melanoma in humans comes from studies of people with the inherited disease xeroderma pigmentosum. These people have a known defect in the mechanism that would repair UV-B damage to DNA, and they also have a very high incidence of skin cancers, including melanoma. 21. There are no reliable animal models for light-induced melanoma. The only models currently available are animals with chemically induced, preexisting pigmented lesions that can be made to look like melanoma after UV irradiation.
10 Major Research Recommendations The estimates following each recommendation of how long the research might take are educated guesses based on the experience of individual committee members. The estimates provide only a rough idea of how long the research might take under ideal circumstances. Molecular and Cellular Studies (Chapter 3) 1. An understanding is needed of why broad bands of UV (heterochromatic radiation) often do not act on DNA in viva and on in vitro cell systems as a simple sum of monochromatic wavelengths. (a) Studies of interactive effects between UV-A and W -B are fundamental to understanding the mechanisms of cancer induction by sunlight. Such studies r employing bacteria or cultured mammalian cells, would take about two to five years. (b) An understanding is needed of W -A-induced repair systems in bacteria, as a first step in understanding possible similar systems in higher organisms. This would take about two to five years. (c) Experiments should be conducted to determine the rate and extent of photoreactivation in humans in sunlight. Data are needed on how the level of dimers depends on the relative amounts of W -A and visible light compared with the amount of dimer-producing UV-B. These experiments would take about two to five years. 2. Data are needed on the rates of repair, in the dark and in laboratory light, of W -irradiated human skin cells as a function of UV dose. The differences, if any, between acute and chronic irradiations should be deter- mined. One might be able (with informed consent) to study individuals who are exposed to high levels of UV-B as part of phototherapy for psoriasis. The aim of such experiments would be to determine whether the kinetics of dark repair of damage from pyrimidine dimers in human skin show two components, a slow one and a fast one, as is true for human cells irradiated in vitro. The two components represent repair of DNA in different regions of the DNA strands. Equally important questions are, what other types of biologically important damages occur
11 in skin, what are their lifetimes, and are any of them persistent? These data could be obtained in about four or five years. Ecosystems and Their Components (Chapter 4) 3. Techniques must be developed for simulating changes in W -B under natural ambient conditions. Only in this way can dose-response relationships be obtained. If these techniques cannot be developed for studies at temperate latitudes, they might best be achieved in a low- latitude (subtropical), minimal-cloud-cover, multiuser facility, which would provide W -B radiation corresponding . . . -^ r=~ scone concentrations at more northern 1acz tudes. Priority should be given to screening representa ~__= ~ ~ _,-~ ~ Imp ~ ; ~-i By \?_ tlve species 01 1mpoLCant 1~" Flails ayes ~ v- Gus.._._,` ing possible adverse effects on crop productivity. Dosimetry and environmental regulation techniques must be developed to ensure optimum experimental conditions-- conditions equivalent to the higher latitude ambient field conditions of the plants being tested. Without strict attention to these control conditions, studies will have limited potential for extrapolation or prediction. It would take about three years to develop the facility and another three years to conduct the species screening experiments. 4. The effects of W dose on elements of aquatic food chains cannot be determined unless (a) the underwater spectral irradiances are integrated over the varying positions of organisms in water columns to obtain the exposures that simulate spectral intensities in the natural systems, and (b) damage to individuals can be related to population dynamics in the natural ecosystem. This would require an integrated research approach involving physical hydrography, physical optics, and organism physiology. It would take about five years to develop this approach and obtain results. Unless W-B studies are made as a part of an ecosystem study, effects on populations and interactions among populations cannot be predicted. (Testing for whole ecosystem effects is addressed in another NRC report, Testing for Effects of Chemicals ~ ~c~L~c~ (NRC 1981).) war made for anchovy larvae An attempt to incorporate such an integrated approach ~ . The interdisciplinary approach used in the anchovy study to assess W -B damage to food chains, together with the specific laboratory
12 measurements, should serve as a model for future research proposals. Direct Human Health Hazards (Chapter 5) 5. Studies (animal and human) should be conducted in the developing field of photoimmunology to determine the magnitude of UV effects on the human immune system, the effective wavelengths, and the dose-response relationship. Results may increase understanding of skin cancer mechanisms, other effects of UV on skin, and certain other diseases. five years. 6. Animal studies of UV-induced skin cancers other than melanoma are needed to understand interactions among parameters such as intermittent exposures, different wavelengths, dose rates, chemical carcinogens and promoters, and agents that modify cellular responses to irradiation. These studies would take about two to five These studies would take about two to years. 7. The Surveillance, Epidemiology, and End Results program of the National Cancer Institute routinely collects data on incidence of melanoma. The incidence of skin cancers other than melanoma should be surveyed every decade at a time coinciding with the population census, so as to determine trends in time. Only a few locations are necessary, but these should be the same as past survey locations. Data should be collected in a way that permits cohort as well as cross-sectional analysis. 8. Animal models for W - or light-induced melanomas are needed. They would allow studies of action spectra, dose-response curves, waveband interactions, and other parameters. It is not possible to predict how long it would take to develop such models. 9. To determine the association between W and melanoma, it would be useful to determine the incidence of the various subtypes of melanoma and their dependence on latitude. Although this will be difficult because the majority of melanomas are of the superficial spreading type, the methodology is available. Careful epidemio- logical studies that are based on reliable clinical and histological studies of subtypes of melanoma are needed.