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.
Executive Summary THE U.S. NAVY'S EXTREMEEY-LOW-FREQUENCY (ELF) communications system consists of a transmitting facility in northern Wisconsin, near Clam Lake, and another facility in Michigan's Upper Peninsula, near Republic. In 1982, the Navy established an ecological monitoring program to determine whether electric and magnetic fields (EMFs) from the ELF communications system affected plant and animal populations or otherwise caused ecological changes in the areas surrounding the transmitting facilities. The ecological monitoring program comprised I! studies of wetlands, slime mold, birds, small vertebrates, litter decomposition and microflora, upland flora, aquatic ecosystems, pollinating insects, soil arthropods and earth- worms, and soil amebas. The Illinois Institute of Technology Research Insti- tute (ITTRI) provided management and engineering support for the program. Ecological studies were carried out by academic researchers. The monitoring program was completed in 1995. Each study investigated possible effects of EMFs by taking biologic measurements at a location (referred to as a treatment site) near one of the ELF transmitting facilities and at a corresponding control site where EMF strength was no more than one-tenth of the EMF strength at the treatment site. Most studies also addressed differences in factors other than EMFs from the ELF antenna, such as soil moisture. Measurements taken over time from pairs of treatment and control sites were analyzed statistically to test for significant differences between the sites.
2 EVALUATION OF ELF ECOLOGICAL MONITORING PROGRAM To summarize the I! final reports of biologic research: ITTR} reported that although some of the program's researchers believe that a few biologic changes might have occurred, all stated that there were no consistent, unequiv- ocal effects of ELF antenna operation on any of the variables they monitored. All concluded that the implications of their results do not indicate adverse ecological effects of significance due to the ELF facilities. At the request of the Navy, the National Research Council formed the Committee to Evaluate the U.S. Navy's Extremely Low Frequency Communi- cations System Ecological Monitoring Program in 1995 to evaluate independ- ently the program's objectives and design, data-collection methods, data analy- sis, and interpretations. The studies included in the ecological monitoring program were evaluated with standard criteria for all branches of scientific endeavor, including appropriateness and coherence of the hypotheses being tested, adequacy of experimental design to test these hypotheses, methods for data collection and analysis, and the soundness of conclusions drawn from investigators' observations. Throughout its review, the committee was aware of controversies sur- rounding scientific theories regarding the manner in which biologic systems might be affected by EMF exposure. However, it was beyond the committee's mandate to address more general questions of the plausibility of different theories of biologic effects or the plausibility of there being such effects at all. The Navy's studies were evaluated according to the criteria mentioned above- not for their ability to confirm or disprove specific theories of biologic effects. Also, the committee was not asked to determine whether EMFs in general (that is, from all sources) are a matter of concern. Rather, it reviewed the ecological monitoring program's assessment of the possible effects of the operation of the ELF communication system. COMMON ISSUES ARISING FROM EVALUATION OF INDIVIDUAL STUDIES The committee's evaluation of the 11 ecological studies in the Navy's ELF ecological monitoring program revealed several issues that were common to many or all of the studies. Those issues are summarized below. USE OF EXPOSURE DATA BY ECOLOGICAL MONITORING TEAMS Measurements of electric and magnetic fields in the vicinity of the study
EXECUTIVE SUMMARY 3 sites were made by UTR] personnel at regular intervals, and they provided this information to researchers. UTR! also provided operating records of the trans- mitting facilities. Such records were needed because the facilities were not operated continuously, even during years when they were fully operational, so study sites were not consistently exposed to ELF EMFs. At some times, dif- ferent portions of the antennas were turned on and off several times each day, with varying modulations, frequencies, current intensities, and phase angles. In a number of studies, especially those with short-term response variables, it was important to know whether the transmitter was on or off during critical exposure periods. However, there is little evidence that the research teams, except the upland flora and wetlands teams, considered or were aware of this factor in evaluating the results of their experiments. STUDY-SITE SELECTION Site selection was not easy. So many conditions had to be met that per- fect matches between treatment sites and control sites were impossible. The program would have been improved by studying fewer variables at more and larger sites and by eliminating studies with poorly matched treatment and control sites. For example, in the soil arthropods and earthworms study, the dominant earthworm studied at the treatment site was not found at the control site. This experiment could never yield results that could clearly be attributed to ELF-EMF exposure, because there was no way of separating ELF-EMF effects from other factors without a control site at which the same earthworm was studied. ADEQUACY OF SITE REPLICATION In many of the studies, the main effect of interest, namely the effect due to the presence of ELF EMFs produced by the antennas, was pseudoreplicated (that is, not truly replicated) because there was only one study site for each level of exposure. The experimental data therefore only provide an estimate of the variance of responses studied within each site. They do not provide an estimate of the variance due to exposures across sites. The effects of the anten- nas on response variables are therefore confounded with the background ef- fects of the different soils, climate, and other characteristics of each site. A danger here is accepting the null hypothesis (that is, no effect of expo- sure to ELF EMFs) when it might be, in fact, false. In many of the studies, it is not possible to calculate the probability of such an error, because it de
4 EVALUATION OF ELF ECOLOGICAL MONITORING PROGRAM pends on an independent estimate of differences between treatment and con- trol, which requires replication of sites, not simply of plots within sites. The acceptance of all furler conclusions must proceed with these caveats in mind, but the caveats are generally not stated clearly anywhere in the reports. SPECIES SELECTION The diversity of species studied was considerable. It included representa- tives of most major taxonomic groups and types of organisms that have been reported to show effects of ELF-EMF exposure in previous laboratory or field studies. Aspects of concern include the lack of studies of amphibians and nonvas- cular plants. In the wetlands study, a moss population was found to increase significantly at the treatment site, but the response was not pursued, because the moss was not a target species; this is unfortunate because the finding might be an indicator that moss is especially sensitive to ELF-EMF effects. No studies focused on rare species or potentially endangered species. These omissions are problematic. Some rare species, such as predators and keystone species, can exert major effects on communities. It is not known whether rare populations at the edge of their range are more or less sensitive to bioassays of additional stresses than abundant species. RESPONSE VARIABLES Many short- and long-term response variables were measured, including characteristics of bird populations, soil microbiology and ecology, plant ecol- ogy, insect populations and behavior, water quality, fish ecology, and repro- duction. Such breadth is commendable for detecting potential effects at differ- ent ecological levels. However, the ability of the studies to detect possible ELF-EMF effects was generally weak. Antenna on-off activity was not related to measurements of short-term biologic responses. Possible small effects could have been difficult to detect, given the lack of statistical power of some studies and the occurrence of confounding variables. The term "small effects" is used in this report to refer to ecological effects whose magnitudes are not likely to exceed those expected from normal perturbations over the short term. A drought is one example of such normal perturbations. The few small possible effects that were found were often too readily dismissed in the study reports. The lack of knowledge of possible mechanisms in the scientific community and
EXECUTIVE SUMMARY 5 the lack of use of simulation models by the ELF researchers might have re- duced the researchers' ability to decide which organisms and response vari- ables are most likely to exhibit effects of the ELF antenna. STATISTICAE POWER The statistical power of an experimental design reflects the likelihood that an experiment will be able to detect the presence or absence of a phenomenon being studied. The committee noted lack of adequate statistical power as a problem that arose in more than one study. Studies with low statistical power would not be able to accept or reject the null hypothesis with sufficient conf~- dence. For example, in several components of the small vertebrates studies, the statistical power was dropped from 90% to 70%, thereby making it much less likely that effects, if any were present, would be detected. Some design changes led to substudies with statistical power of 30%. Such experiments offer no bounds on uncertainty, and it can reasonably be questioned whether such experiments should have been performed at all. RESPONSE TO REVIEW There was a great deal of variation in how TITR! and the researchers responded to external reviews. In many cases, the investigators were very responsive to reviewers' comments and critiques; that contributed greatly to their research and to the results of the monitoring program. In some cases, reviewers' comments were addressed through explanations of why suggestions were not heeded. In a few cases, suggestions were not followed, and research design, analytic techniques, or interpretations that needed improvement or cor- rection were left unattended. Some of the reviewers' comments were very cri- tical regarding statistical power, data archiving, and unwarranted dismissal of possible effects. Such criticisms should have been a clarion call to program managers that some studies had substantial problems. APPROPRIATENESS OF INTERPRETATION In several studies, modest but significant differences were observed between data collected at treatment sites and data from control sites. Research- ers conducting the studies concluded that five of these potential effects were
6 EVALUATION OF ELF ECOLOGICAL MONITORING PROGRAM due to factors other than the ELF antenna. Without attempting to judge whether any of those interpretations suggested a predisposition to a particular outcome, it is important to consider whether the conclusions were established with a credible scientific basis. in the course of this committee's review and discussion of the researchers' final reports, concerns arose about the scientific credibility of some of the conclusions. Differences between treatment sites and control sites that were dismissed by researchers and by TTTR! as not being clearly related to ELF exposure included the increase in bee overwintering mortality, the reduction in leaves per bee nest cell, accelerated litter decomposition, early eye-opening in mice, and depressed earthworm reproductive rates. The committee believes that some of those differences were dismissed too readily as alleged artifacts of environmental variations or experiment design. DIFFERENT METHODS FOR SIMIEAR ORGANISMS The broad range of studies in the program often resulted in more than one research team's examining possible ELF-EMF effects on similar organ- isms or processes, but with somewhat different protocols. The use of different methods is sometimes advantageous or necessary and does not necessarily negate conclusions from any one study. However, the use of different methods prevented the researchers from making valid cross-site comparisons and there- by impeded the realization of the full potential of integration across sites and organisms. There does not appear to have been much discussion of coordinat- ing methods before the experiments began. LACK OF INTEGRATION AMONG STUDIES AND SYNTHESIS OF INFORMATION Early research design should have been guided by recognition of interac- tions among ecosystem components and by encouragement of integration among studies with full development and application of appropriate statistical approaches. If there had been consistent integration and comparison of findings of the different studies, the overall research effort could have been improved with the same expenditure of resources. With better integration, there could have been more pursuit of promising results, the hallmark of good research. By failing to integrate the studies of different species and ecosystem processes,
EXECUTIVE SUMMARY 7 this large-scale effort largely surrendered the possibility of detecting small changes in interactions of components and gave up the major advantage of such large-scale research. DATA ARCHIVING There appears to have been no standard procedure for archiving data resulting from the ecological monitoring program. For some studies, original data were kept primarily in laboratory notebooks. As far as the present com- mittee is aware, no standard protocols were established for formatting, docu- menting, or reporting the data. Most of the data were never transferred to the program manager, and those data are not readily available. As program man- ager, UTRI should have been responsible for data archiving and planning for long-term availability of the information. IITRI,S ENGINEERING SUPPORT AND PROGRAM MANAGEMENT UTR} did a good job on the engineering aspects of the ecological moni- toring program in characterizing the spatial and temporal characteristics of the electric and magnetic fields. The instrumentation for ELF-EMF measurements appeared to be well designed, well calibrated, and properly used. ITTR} pro- vided ELF-EMF exposure information to the researchers for each study. In addition, ITTR] was responsive to requests from researchers for additional engineering support. However, in its review of individual study reports, as well as of the overall program of monitoring for possible effects of the ELF antennas, this committee discovered weaknesses in some aspects of I]:TR] management of the program. Three of these weaknesses appear to have been caused by lack of adequate oversight. First, [ITR} should have detected problems with the use of exposure data through annual research reports. In response, it should have provided more guidance for use of the exposure information and required that an EMF-exposure expert work closely with each study until the study leader understood the types of data available, data variability, and the best methods for applying the data. And there should have been greater involvement of an expert in broad-based applied statistics at the earliest phases in the design of this program's studies. Second, responsibility for the lack of archiving and of
8 EVALUATION OF ELF ECOLOGICAL MONITORING PROGRAM planning for long-term availability of monitoring-program information appears to rest with ITTRI's management of the program. Third, ITTR] should have established a regular internal-review process to ensure that each study ade- quately addressed external criticism. Those problems appear to have originated in poor early planning by ITTRI or inadequate followup by ITTRI as problems arose during the duration of the program. ECOLOGICAL EFFECTS The Navy's ecological monitoring program reported no obvious adverse ecological effects, such as unusual changes in species populations or large- scale mortality of trees or other organisms, as a result of operation of the ELF communications system during the period of the monitoring program. The monitoring program also did not detect any small effects with well-defined consequences, such as decreases in reproductive fitness, that would be likely to result in major effects in the future. The present committee agrees with the general findings of the Navy's ecological monitoring program, within the limitations described in this report, that the researchers' observations provide no evidence of statistically signifi- cant, widespread, adverse effects of EMFs associated with the ELF antennas on bird populations, leaf-litter decomposition processes, upland flora, the movements of dragonfly larvae, the colonization of leaf litter, the movement of fish, soil arthropod populations, and soil ameba populations. However, some of the studies, as discussed in this report, had deficiencies that dimin- ished their capabilities to detect small effects. The committee recognizes that small effects on populations, mediated through modest changes in response variables, might slowly compound and only later become apparent. Numerous flaws in the ecological studies as designed, implemented, and interpreted-would have compromised detection of many possible small effects of the antenna operation. As discussed in Chap- ter 5, the individual ecological studies can be sorted into three categories: those which the committee judged to be acceptable with qualification, those which might be acceptable after more information is obtained or data are reanalyzed, and those which are unsalvageable because of serious flaws. RECOMMENDATION The complexity of assessing possible ecological impacts of ELF EMFs especially given the diversity of the ecosystems and the variability of their
EXECUTIVE SUMMARY 9 locations and their distances from the antennas, (and therefore the variability of exposures) made it extremely difficult to design and complete appropriate and comparable ecological monitoring studies. The following recommendation and suggested next steps reflect this committee's understanding of such diffi- culties but also indicate the committee's concern for bringing this ecological monitoring program to an appropriate and fruitful conclusion. DO NOT REPEAT THE FIEED STUDIES Dispite the weaknesses of the monitoring studies, the committee does not recommend that the field studies be repeated, because the extensive studies conducted to date have provided no evidence that exposure to ELF EMFs had obvious adverse ecological effects. Although caution must be used in drawing conclusions from results of most of the studies regarding possible small effects because of faulty study design or analysis, the committee considers it highly unlikely that repetition of the ecological monitoring studies undertaken in this program would produce any new findings about ecological effects of ELF EMFs. SUGGESTED NEXT STEPS REANALYZE THE EXPOSURE-ASSESSMENT DATA The ELF ecological monitoring studies were supplied with ELF-EMF data based on measurements made by TITR! engineering teams. The timing and location of the measurements differed among studies. They included mea- surements made only once a year at some study plots (as in the wetlands study), at the location of each individual of the response species of interest in others (as in the upland-flora study), and as a spatial gradient of exposure levels (as in the bird-nestling study). in addition to those different forms of available ELF-EMF exposure data, the study teams apparently were made aware of the variability in times and outputs of antenna operations. In some studies, the analyses and interpretations of ELF-EMF effects appear to have made use of the exposure data. However, some studies apparently used the information inappropriately, and others might not have fully recognized the importance of the vagaries of antenna operations and output. The committee suggests that the investigators from each ecological moni- toring study reassess their use of data on antenna operation and ELF-EMF data and, if they were used inappropriately, reanalyze the responses of selected
~ 0 EVALUATION OF ELF ECOLOGICAL MONITORING PROGRAM ecological variables. In addition, the results of several studies should be reana- lyzed so that firmer conclusions can be drawn. These are the studies labeled in Table 5-! as "Might Be Acceptable With More Tnformation or Analysis." The committee suggests that an organization that is independent of the Depart- ment of Defense and UTR} direct the reanalysis. The reassessment and reanal- ysis should be performed in close collaboration with biostatisticians familiar with this type of EMF-exposure assessment and engineers knowledgeable about field ELF-EMF exposure measurements. If reanalysis reveals statisti- cally significant or suggestive responses of ecological variables to ELF EMFs, these responses could be considered for further controlled study (as discussed below). The committee suggests reanalysis and possible controlled studies so that an opportunity to improve the understanding of ELF-EMF exposure and possible ecological responses is not lost. VARIABLES THAT TENDED TO SHOW MEASURABLE EFFECTS SHOULD BE SUBJECTED TO CONTROLLED LABORATORY STUDY The ELF ecological monitoring studies produced few results that tended to show effects of ELF-EMF exposure on selected ecological variables. One effect, the growth response of upland trees, might be an artifact of selective use of exposure data. in other cases, an effect might be a true measurable response, but the experimental design or the complexity of the surrounding ecosystem might have created an environment that made the findings suffi- ciently questionable to warrant further, more-controlled studies. Responses that perhaps could be tested under controlled laboratory condi- tions are the apparent increase in chlorophyIl-a in the aquatic-ecosystem study, the behavioral responses of bees and their overwintering mortality in the pollinating-insects study, and increased moss growth in the litter-decomposi- tion study. The chIorophyIl-a increase appeared to be an increase in cell den- sity rather than in chlorophyll per cell; this possibility could be tested and the ecological implications analyzed. A similar study might help in understanding whether alterations in bee behavior and mortality are repeatable and can be shown to be caused by ELF-EMF exposure or are artifacts of the less-con- trolled, more-complex study sites. The wetland study unexpectedly discovered more moss cover on decomposition bags closer to the antenna treatment sites than in intermediate treatment or background control sites. The increased moss cover caused problems in interpreting data on decomposition, but the variabil- ity in growth of moss should be considered for controlled investigation. The reanalysis of exposure assessments (as discussed above) might uncover addi
EXECUTIVE SUMMARY 11 tional suggestions of small but measurable responses of ecological attributes to ELF-EMF exposure. If it does, these responses could also receive further study under more-controlled conditions; such studies could be designed also to help to elucidate the mechanisms of response if an EMF effect is observed. Such information might guide researchers in deciding which organisms and response variables are most likely to exhibit effects, if any, of the ELF an- tenna. REANALYSIS OR LABORATORY STUDIES SHOULD BE REVIEWED INDEPENDENTLY Reanalysis of exposure assessments might or might not identify some effects of ELF-EMF exposure on ecological variables not previously observed, and laboratory tests might or might not confirm them. Reanalysis might also strengthen the credibility of the findings of some studies. The committee suggests that if reanalyses or laboratory studies are performed, the Navy should arrange for an independent evaluation by a few individuals to assess all of the findings resulting from the reanalyses. The individuals should include biostatisticians familiar with ELF-EMF exposure assessment and biologic expertise to determine what the weight of evidence indicates and the biologic or ecological implications of any substantiated treatment effects. A broader integration of all studies should be pursued through the use of quantitative methods designed for such purposes. Integration of related effects, although not statistically significant, can point to areas where additional study might be warranted. The results of the independent evaluation should be made publicly available. Such an independent final review would serve the Navy and the public in producing more-credible and improved findings of the monitoring program.