£2£UM^if-Situ Water-Quality Sensing: Biological Sensors April 28-30, 1980 STEERING COMMITTEE NOAA U.S.G.S. Barbara S. Pijanowski, Chairman National Marine Pollution Program Office Luther E. Bivins Office of Ocean Engineering Dr. Andrew Robertson Great Lakes Research Environmental Laboratory Richard W. Paulson Water Resources Division EPA John D. Koutsandreas Office of Research and Development Dr. Norman L. Richards Gulf Breeze Laboratory Edwin D. Kennedy Environmental Science Information Center University of South Florida Dr. J. A. Llewellyn College of Engineering U.S.C.O.E. Magnavox Corporation Dr. Harlan L. McKim Water Resources Support Center Theodore Major Environmental Systems Division 2nd Interagency Workshop on In-Situ Water-Quality Sensing: Biological Sensors Pensacola Beach, Florida April 28-30, 1980 Sponsors Environmental Protection Agency National Oceanic and Atmospheric Administration U.S. Corps of Engineers U.S. Geological Survey Host National Marine Pollution Program Office National Oceanic and Atmospheric Administration Rockville, Maryland i a ! 2 C Digitized by the Internet Archive in 2012 with funding from LYRASIS Members and Sloan Foundation http://www.archive.org/details/2ndinteragencywo00inte FOREWORD The 2nd Interagency Workshop on In-Situ Water-Quality Sensing convened during the period April 28-30, 1980 in Pensacola Beach, Florida. As the contents of this volume make clear, the need for effective tools for qualifying and quantifying ocean pollution is rapidly becoming critical. The organizers of the workshop hope that this publication will contribute to development of better measurement technology and will receive wide acceptance and careful consideration. Organizing a conference of this kind is tediously complicated and acknowledgement is especially appropriate for the Steering Committee, responsible for the overall effort. Its members and professional associations are listed on the inside front cover. Mention should be made that Mr. Arthur Solman of the Water Research Centre of the Medmenham Laboratory at Buckinghamshire came from England to attend and kindly contributed an extemporaneous presentation of their work there. His remarks are included in this conference report. Acknowledgement is made to Edwin D. Kennedy, NOAA, a member of the Steering Committee who guided this document through the exacting process of editing and putting it all between covers. Finally, the effort in coordinating the conference arrangements and coping with all the myriad details usually attendant to organizing such a conference on-site were most corapentently handled by Geraldine Cripe of EPA' s Gulf Breeze Laboratory in Pensacola Beach; her contribution is highly appreciated. A list of the participants in the conference appears at the back of this volume. 111 A. F. (TONY) MFNTINK 1924 - 1980 CHIEF INSTRUMENTATION DEVELOPMENT BRANCH ENVIRONMENTAL MONITORING AND S?JPFORT LABORATORY ENVIRONMENTAL PROTECTION AGENCY IV DEDICATION This 2nd Interagency Workshop on In-Situ Water-Quality Sensing is dedicated in memory of the late A. F. (Tony) Mentink, who died suddenly in his EPA laboratory in Cincinnati, Ohio, on January 23, 1980. At the time of his death, Mr. Mentink was Chief of the Instrumentation Development Branch of the Environmental Monitoring and Support Laboratory of the Environmental Protection Agency. He helped pioneer the development and deployment of automated in-situ water-quality systems, first becoming involved in this activity in the early 1960s. This work included the development of specific pollutant electrodes as well as multiparameter analytical equipment. Mr. Mentink also worked in the field, directing the installation of equipment, which required a strong background in communications technology such as automatic telephony and satellite transmitters as well as the use of computers. Mr. Mentink was born on April 7, 1923 in Cincinnati and was graduated from the University of Cincinnati in 1952 with a degree in electrical engineering, where he also pursued graduate work. He was elected to the honorary society Eta Kappa Nu. Mr. Mentink began his professional career as a technical engineer with the General Electric Corporation where he was supervisor of Engineering Analysis in 1961 when he resigned to become a water-quality consultant. Between 1968 and 1972, he was employed as Chief of Instrumentation Development with the Analytical Quality Control Laboratory of the Methods Development and Quality Assurance Research Laboratory. During the period 1965-1976, he published 13 technical papers on the subject of water-quality alone. As a member of the Institute of Electrical and Electronic Engineers, he was active in a number of its professional groups, including the Professional Group on Electronics, Medicine, and Biology. In dedicating this conference to Mr. Mentink, the members of the steering committee look back to his active participation in the first such gathering in Las Vegas in 1978, when his contributions were manifold. As development goes forward in this field, his guidance, advice, and technical help will be acutely missed. Mr. Mentink is survived by his wife Jody and five daughters. CONTENTS Foreword iii Dedication v Summary 1 Integration Panel Report 3 Presentations Remarks of Ferris Webster, Assistant Administrator for Research & Development, National Oceanic and Atmospheric Administration 7 Biological Monitoring of Hazardous Wastes in Aquatic Systems 11 by V.W. Lambou and L.R. Williams Enzymes and Other Biochemical Indicators of Toxicant Effects in Fishes 19 by Dr. Dora R. May Passino Overview and Assessment of Biological Sensing 45 by John D. Koutsandreas New In-Situ Techniques in Biomonitoring 55 by Dr. Harlod E. Schlichting, Jr. Processing of Zooplankton Samples by Electronic Image Analysis 59 by Dr. H. Perry Jeffries, Dr. Kenneth Sherman, Ray Mauer, and Dr. A.D. Poularikas Computer Interfaced Sysyems for Biological Monitoring of Water Quality 81 by Dr. John D. Cairns, Jr. Algal Chlorophyll Measurements: Traditional and Untraditional Applications of a Biological Sensor 103 by Bruce L. Kimmel Metabolic Response of Fish to Stress 119 by Dr. Donald E. Wohlschlag Enzymes and Antibodies as Sensors of Toxic Substances 143 by Dr. Donald R. Sellers Bioprobes for the Monitoring of Toxic Compounds and Hazardous Materials in Water 161 by Dr. Shia S. Kuan and Dr. George G. Guilbalt vii Potentiometric Biosensors Based on Heterogeneous Catalytic Systems 167 by Dr. Mark E. Meyerhoff Utility of the Microtox Luminescent Bacterial Assay for the Rapid Assessment of Aquatic Pollution 173 by Dr. Anthony A. Bulich The Use of Algae as Biological Accumulators for Monitoring 185 Aquatic Pollutants by Dr. Stephen L. Friant In-Situ Water Quality Monitoring Using Trout in the U.K. 207 by Arthur Solman Working Panel Reports Working Panel 1: Biochemical Systems 209 Working Panel 2: Bioaccumulation 221 Working Panel 3: Organism Behavioral Responses 229 Working Panel 4: Traditional Biological Measurements 235 Participants 241 vni SUMMARY The 2nd Interagency Workshop on In-Situ Water-Quality Sensing: Biological Sensors was convened at the Holiday Inn at Pensacola Beach, Florida during the period April 28-30, 1980. The purpose of the workshop was to review the potential of biological mechanisms and organisms as tools for bioassay, measure- ment and detection of water-quality parameters, with a view toward the development of automated in-situ sensors. The meeting was sponsored by the National Oceanic and Atmospheric Administration, the Environmental Protection Agency, the U.S. Geological Survey, and the U.S. Army Corps of Engineers. Present capabilities for collection of discrete water-quality information are limited by the lack of adequate in-situ sensing systems. Although advanced measurement systems using satellite and aircraft platforms with remote sensing capabilities are available and information communication networks for the transmission of data over wide areas is commonplace, the usefulness of such systems is limited because real-time or near real-time ground truth verification cannot be obtained. Furthermore, because sensors which directly detect environmental changes of interest are not available, measurements must be carried out through labor-intensive sample collection in the field with attendant deleterious effects such as contamination and in-transit deterioration. The field sampling process must then be followed by tedious analytical procedures in the laboratory. All these factors decrease the quality assurance of the information generated. The first interagency workshop on automated in-situ sensing was held in February, 1978 in Las Vegas and was sponsored by the Environmental Protection Agency, the National Oceanic and Atmospheric Administration, and the Interagency Work Group on Satellite Data Collection Systems. That workshop concentrated on identification of parameters that must be measured for water-quality assessment and physical and chemical techniques that can be developed for their in-situ detection. This second workshop was planned as part of the continuing interagency effort to improve measurement technology with major emphasis on the biosensing and biological aspects. A total of 14 papers were presented at the workshop. They covered traditional biological measurements, organism responses as tools for detection, enzymes as tools for detection, and organisms as samplers and bioaccumulators. Paper presentations were followed by four working groups which met concurrently to discuss the state of the art as presented in the papers and formulate recommendations concerning development of these techniques for pollutant and water quality sensing. Working groups included (1) Biochemical Systems, chaired by Mr. Theodore Major of the Magnavox Corporation; (2) Bioaccumulation, chaired by Dr. Andrew Robertson of NOAA/GLERL; (3) Organism Behavioral Responses, chaired by Mr. Richard W. Paulson, of the U.S. Geological Survey; and (4) Traditional Biological Measurements, chaired by Mr. John B. Bushman of the U.S. Army Corps of Engineers. In assessing the contextual thrust of the papers offered, the panels evaluated potential capabilities and limitations of biological sensors for improving water-quality measurements; in addition, the groups attempted to identify existing and potential concepts for use in the development of automated in-situ sensors. The panels also considered project development costs and time, and ultimately offered recommendations in terms of priorities related to development planning, taking into consideration the limited resources available and the various needs of the agencies, industries, and universities concerned. Principal conclusions and recommendations are presented in the Integration Panel Report that follows. In general, the summaries offered by the four panels reflected a marked enthusiasm for the near-future development of practical biosensors for field use. Development of such devices should lead to more reliable and cost-effective water-quality data. It was also recognized that the U.K. and Europe hold a clear lead over the U.S. in this field of water-quality monitoring and a review of what has been accomplished by the Europeans would be an appropriate initial step in developing a meaningful program for the U.S. INTEGRATION PANEL REPORT Composed of the chairmen of the working groups and members of the steer- ing committee, the Integration Panel met at the conclusion of the workshop with the objective of bringing together the diverse work of the four sep- arate panels. There was unanimous agreement among the panel and workshop participants that the emerging technology of biosensing is an exciting new development area as well as a field with great promise in both near- and long-term time frames for providing valuable tools for environmental measurement and assessment in both fresh and marine waters. One general theme supported the entire Integration Panel discussion: the Federal government has a responsibility to support and foster develop- ment of environmental measurement systems, including those involving the use of novel, innovative schemes for providing reliable, cost effective and automated systems where appropriate. Basic research should be carried out at the Federal level in a coordinated fashion to benefit all users, including state and local governments who have neither the funds nor competence to undertake such programs. If reliable devices and measure- ment schemes can be developed, demonstrated and accepted, particularly by the regulatory agencies, a market sufficient to interest private industry will evolve. The panel also felt it important that the Federal government support development of well documented techniques and proced- ures as well as hardware, since many environmental measurements are not now and may never be possible through hardware systems alone. The concept of biosensing is not new. In its simplest form it was employed more than a hundred years ago when deep miners used canaries to detect toxic gases in that underground environment. Little attempt has been made since that time, however, to seriously evaluate general biologi- cal concepts for detection of chemical and environmental change. Biosensors in the form of complete organisms or colonies of organisms can be used as gross integrating detectors for adverse water quality con- ditions. Analyses of their tissues can provide an integrated measure of the nature of the water in which they have existed. Changes in behavior can indicate stress caused by environmental change. Such techniques have already been demonstrated and, with minimum development effort, some of these techniques could be standardized and automated for routine monitoring. They could be particularly useful as alarms and mechanisms to trigger more directed and sensitive sampling and analysis. Biosensors can also be developed from parts or components of organisms to perform as detectors of a more specific nature. The natural selectivity that is present in certain components of some organisms can be harnessed to perform as single function detectors for specific chemicals or classes of compounds. Near Term Development The panel found that several biosensing schemes were further developed than anticipated. Modest development efforts toward standardization and automation, where possible, could provide operational systems within a very short time. • Behavioral changes in fish or other whole organisms can be used as alarms and triggering mechanisms, thus initiating intelligent sampling and control actions. The use of fish to monitor industrial effluent is currently well established in Great Britain. Although similar techniques have been demonstrated in the U.S., they are not yet generally accepted. Other organisms could also be used. The behavioral responses of bivalves, for example, could be exploited to indicate changes in the environment in which they reside; bivalves would be particularly sensitive to changes occurring at the water-sediment interface. Some investigations have begun in this area on marine scallops; however, fresh water bivalves present a longer term development problem because methods for their laboratory rearing are not yet well developed. • Organisms can also be used as integrating samplers. It has long been known that many organisms preferentially accumulate various mater- ials in their tissues, a factor that has been responsible for considerable economic loss over time because of the inability to harvest shellfish from polluted estuaries. On the positive side, however, organisms such as mussels and oysters have been used to collect cumulative samples of heavy metals, radionuclides, and other toxic substances for trend moni- toring. The analysis of fresh water periphyton biomass has been used to provide gross information leading to location of pollutant "hot spots." With these illustrations in mind, it might be useful to propose a system that couples the triggering function of organism behavioral response with the bioaccumulating characteristics of other organisms. In such a system, when the behavior of the triggering organisms indicates a detrimental change in water quality, water and tissue samples can be automatically collected. Subsequent analyses would determine abnormal levels in the water sample at the time the organisms was stressed, and a cumulative record of previous water content to assess chronic low-level exposure could be determined from analysis of the bioaccumulators. Depending on the degree of automation, systems such as this could find almost immediate application in monitoring drinking water systems and waste water and storm sewer effluents. Indeed, as previously mentioned, alarm systems using fish behavioral responses are already in use in Europe. • The panel agreed that additional refinement of chlorophyll measure- ment techniques would be only minimally fruitful. Although the ease with which measurements can be made lends itself to improvement, the overall fluorescence technique and equipment currently available is generally acceptable for most applications. The most important improvement desired is for an increase in data collection rate to keep up with the demand for the ground truth necessary for remote sensing verification, and not for additional detector development. • Enzyme systems to detect and measure nitrate and nitrite have been developed, demonstrated, and await only the interest and attention of a private entrepreneur for reaching commercial production. Their particular attraction for immediate application would be in automated measurement systems for drinking water where sufficient markets should exist to make the necessary production engineering profitable. • Enzyme systems for the near real time detection of toxic materials or classes of toxic compounds are currently under investigation and, while some systems have already been identified for limited application, others should be identified within the next few years. At the present time such systems can be used as real time alarms; however, their use as quantitative measurement devices holds some possibility with longer term development efforts. • The phenomenon of bioluminescence also holds some promise as a gross indicator/alarm system for detection of toxic substances. Bioluminescence devices are being developed as replacement for LC50 laboratory tests and have not been evaluated for application as alarms. It is estimated however that such devices could have fairly low operational costs if used to automate manual systems. Although systems based on bioluminescence detectors would most likely be limited to detection levels higher than the more sensitive "fish alarm" systems, a cost analysis would reveal the areas of utility for such systems. Long-Term Development • Heterogeneous enzyme systems for detection and measurement of spe- cific compounds offer development potential for the much longer term. These systems respond to a wide range of compounds of interest, are longer lasting, and can be made highly sensitive. While the development will involve considerable effort, potential benefit in terms of providing a net array of low cost sensing tools is also considerable. Development would benefit from close association with the medical instrumentation field. e As long-term integrators, fish and mollusk blood chemistries were felt to offer considerable promise. • Automated observation of plankton characteristics for sensing environmental change was felt to be impractical at present; however, a longer term development program could provide dramatic payoff in the future. In addition to endorsing the continued investigation of biological concepts for use in development of environmental measurement systems, the panel made two specific recommendations for implementation within the Federal government. (1) In areas where agencies have similar measurement responsibilities and interests, interagency technical reviews covering development programs should be held. Although agencies may not necessarily be working on the same types of development, these reviews could stimulate interest and lead to greater awareness of new development possibilities and lead to possible cooperative efforts. To build the broadest support, these reviews should be held at the technical level. (2) Any future Federal monitoring plan or plans should contain clear statements regarding specific measurement requirements. Without these statements, neither government nor industry can assess market requirements; consequently, it is unlikely that needed measurement and monitoring devices and systems will be developed and available when necessary. Remarks of Ferris Webster, Assistant Administrator for Research and Development National Oceanic and Atmospheric Administration at the 2nd Interagency Workshop on In-Situ Water Quality Sensing: Biologicaal Sensors Pensacola Beach, Florida April 28, 1980 I am pleased to be with you today to deliver the keynote address at this work- shop. To my knowledge, this workshop is the first of its kind. I know of no previous organized attempts to bring together experts in biology and engineering to combine ideas for new approaches to measuring pollution in water. The subject of this workshop is important. Concern about pollution of the air we breathe and water we use is near the top of the list of problems with which our society must cope. Pollution is everywhere. Its nature is complex. We see the effects of pollution, but often do not understand the causes. We seek technological solutions to problems that are both technological and social. Only a few years ago, the pollutants we were concerned with were limited to a few heavy metals and DDT. Today, because of our increased awareness, we are faced with a wide variety of toxic materials and hazardous substances. Both NOAA and the Environmental Protection Agency are concerned about the quantities of these wastes that have been disposed of in sensitive wetlands and coastal areas, and about the fact that our ability to detect and clean-up these areas is lacking. Reliable information about the state of water quality is our primary need. This, in turn, calls for measurement systems and sensors. Understanding and controlling pollution, whether it be in air or water, requires observations of ecosystems and physical environments over long periods of time. In many cases, we must collect our measurements in environments alien to direct human observa- tion. Innovation and technology transfer are needed. By technology transfer, I mean the application of new knowledge from one institutional setting to another. For example, space technology has provided us with the capability to remotely observe other planets. These successes have been transferred to other fields, many far removed from the space environment. The examples of the application of new technologies in other fields can inspire us to seek such uses in the ocean. If we can now remmotely measure the compo- sition of the surfaace of Mars, we should be able to examine the composition of the seafloor and riverbed sediments without laboriously taking grab samples. If we can now measure turbulence levels in blood vessels, perhaps we can do the same in the ocean to understand pollutant resuspension. In particular, advanced technologies in other fields might be used to improve our ability to observe ocean pollution and to understand its effects. An obstacle to technology transfer often is the difficulty of communicating new knowledge in one field to another that might apply that new knowledge. Each field has its own jargon, its own journals, its own learned societies. Even in a single field, communications can be surprisingly difficult; across fields, they are generally accomplished only by determined people. Thus, this workshop is particularly important because it represents the first step in beginning a dialogue between disciplines that have traditionally been separated. Outside of the medical field, biologists and engineers have not often collabor- ated. In ocean pollution, we know that such collaboration is essential. This workshop, if successful, will add impetus to the development of the technology needed to understand and control ocean pollution. The Federal Plan Last December, 13 Federal agencies completed the first "Federal Plan for Ocean Pollution Research, Development, and Monitoring." This document was written in response to legislation passed by the Congress in May 1978. Under this law, NOAA is required to take the lead agency responsibility in preparing a plan for marine pollution studies. The plan must coordinate the programs of all Federal agencies involved in marine pollution studies. It must identify national needs and problems, establish priorities, describe existing Federal capabilities and their ability to meet the priorities, and recommend changes in the program. The recommended changes should respond to changes in goals, funding levels, pooling of efforts between Federal agencies, and elimination of duplication. I would like to mention some of the conclusions and recommendations of the plan that are particularly relevant to this workshop. A major section of the Plan was entitled "Tools for Controlling and Minimizing Ocean Pollution." In that section, it was noted that the development of measure- ment technology affects all agencies and their programs, and furthermore, is frequently the limiting factor in developing information for dealing with ocean pollution problems. It went on to note that "state-of-the-art measurement of virtually all chemical and biological factors is limited to laboratory techniques as opposed to in-situ measurements," and that "methods are labor intensive, expensive, and not easily automated or transferred to the field." It continued to explain that "development in this area is frequently under emphasized because it often requires long lead times and cannot produce immediate results." Unfor- tunately, in the competition for scarce research dollars, we sometimes sacrifice research directed toward longer-term benefits. Consequently, we make few devel- opmental breakthroughs in instrumentation and end up having the same chronic measurement problems year after year. The plan recommends that special emphasis be given to the development of measurement technology that will benefit all agencies. It identifies the need for development of automated laboratory equipment, automated sampling equipment, and in-situ and portable measurement systems. All these recommendations are reflected in topics that you will be addressing in this workshop. The Federal Plan focuses on marine pollution problems; however, we must keep in mind, that in the measurement area, marine and freshwater technologies have much in common. Many of the deficiencies that exist in the marine field are virtually the same as those in nonmarine areas. Consequently, improvements that result from the work you do here will benefit all the Federal agencies responsible for water pollution measurements and science in general. NOAA's Marine Pollution Programs I would like to take this opportunity, on EPA turf, to say a few words about NOAA's activities in marine pollution and, in particular, about our work in sensing technology. The Ocean Pollution Act calls for NOAA to establish a comprehensive, coordinated, and effective ocean pollution research and develop- ment and monitoring program. We have developed a broad marine pollution program throughout NOAA. Organizationally, in the Office of Research and Development, we have set up an Office of Marine Pollution Assessment, gathering together existing programs in environmental assessment, ocean dumping, and hazardous materials spill response. The Office of Fisheries has a special need for under- standing the effects of pollution on marine habitats and has a strong program in the Northeast. In addition, we are developing plans to establish an opera- tional marine pollution monitoring capability. The Environmental Data and Information Service is working on the special needs of marine pollution data management. In technology development, NOAA is creating an Office of Ocean Technology and Engineering Services. This activity will be working with the Office of Marine Pollution Assessment to develop an expanded program of measurement technology. We must respond to increasing needs for technology in support of ocean pollution research and monitoring. At the moment, we are addressing the development of selected sensors. A miniature dissolved-oxygen sensor has been developed which we hope will increase the reliability of dissolved oxygen measurement at a reduced cost. We are conducting research to determine whether aerospace image-processing, enhancement, and pattern-recognition technology can be applied to analyzing plankton samples. We are beginning research on the use of biodetection mechanisms such as enzymes and antibodies for selective sensing of toxic substances. A number of the researchers involved in these projects are here and will be describing this work. NOAA is planning regional workshops and studies to determine optimum strategies for monitoring stressed areas throughout the U.S. coastal regionsal Our agency, along with others, represented here, has a commitment to the stewardship of the ocean environment. The increasing demand to watch over large ocean and coastal areas will require improved and innovative measurement tools that can sense trace quantities of toxic substances in the sediment, water column, and fish tissue. The sensors will have to do all these things and yet be housed in instruments that can be operated by low-skill-level personnel. We cannot afford to use research scientists to conduct routine surveillance. The results of this workshop can potentially benefit NOAA's pollution programs. We, and other agencies, need the kinds of capabilities you will be discussing here this week. Summary Biological organisms may hold the key to new breakthroughs in sensor concepts. Nature through evolutionary time has sharpened our senses and those of other living organisms with whom we share our planet. It is possible that sensitive and selective sensory mechanisms may be used directly to detect substances of interest. Perhaps even their mechanisms can be duplicated for use as indirect detectors in other applications. I hope that during the next two days, ideas will be exchanged which will spark innovative approaches to sensor development. The four working panels that begin meeting tomorrow provide an opportunity to assess the state of biological research and technology and to recommend directions for future emphasis. I hope the personal interactions in this small meeting will stimulate and act as a catalyst to new thinking. I am encouraged by the purpose of this workshop. I will await the results with interest. We in the Federal agencies can use your contributions in planning programs in this important and potentially fruitful area of biosensing. 10 BIOLOGICAL MONITORING OF HAZARDOUS WASTES IN AQUATIC SYSTEMS by V. W. Lambou and L. R. Williams Environmental Monitoring Systems Laboratory U.S. Environmental Protection Agency Las Vegas, Nevada 89114 The Environmental Protection Agency (EPA) has initiated an aggressive program in the United States to discover, investigate, contain, and clean up uncontrolled (abandoned) hazardous waste sites that present an imminent hazard to human health or the environment. Under the authority of the Solid Waste Disposal Act, as amended by the Resource Conservation and Recovery Act of 1975 (and its amend- ments), the EPA is also responsible for the development, promulgation, and enforcement of regulations for the management and disposal of hazardous waste. These hazardous waste programs require measurement methods, monitoring and exposure assessment guidelines, field response, and a quality assurance program to provide accurate and reliable data for enforcement actions and for making proper management decisions to protect human health and the environment. One of the key elements and needs in the hazardous waste programs is biological monitoring. This paper provides an overview as to the needs for the application of biological monitoring in aquatic systems and its usefulness for monitoring hazardous wastes and their leachates. NEED FOR BIOLOGICAL MONITORING Biological monitoring is rapidly attaining the recognition it warrants as an invaluable component in comprehensive environmental monitoring programs. Indeed, for many years there has existed a bias toward the "hard numbers" generated by physical and chemical analysis and a reluctance to accept the intrinsically less precise values obtained with biological systems. The composition of monitoring staffs in the EPA and the predecessor agency, the Federal Water Pollution Control Administraation, has generally reflected the emphasis on physical and chemical measurement. The reason for relegating biological monitoring to a weak supportive role in monitoring programs of the past were many. Some were valid. As viewed from the standpoint of the analytical chemist, the state-of-the-art of biological monitoring is still quite primitive. The water-quality biologist attempts to unravel the complexities of a highly intricate and variable system in an effort to determine if observed changes in the biota are real and, if so, whether these changes can be attributed to manmade influences. It has been very difficult to effectively model many of the critical processes of biological systems. The biological responses of many test organisms are highly variable, even under rigidly controlled laboratory conditions, and responses determined under such conditions may or may not represent the responses of similar organisms in natural ecosystems. To effectively represent the environmental interactions of a single organism in a natural system is one of the greatest challenges to modelers and programmers today. 11 The use of major groups of organisms (e.g., aquatic diptera) as biological indicators has been hindered by lack of standardized and accepted classification schemes. While some significant efforts have been made to standardize taxonomies in recent years, much work remains to be done in the area. Lack of biologically trained personnel has been more a consequence of inappro- priate talent-mix selection - a product of the historical emphasis on physical/ chemical measurement - than any lack of well-qualified biologists. A balance of disciplines which reflects today's monitoring needs should be achieved. In many cases this means the addition (or substitution) of biological personnel with training and experience appropriate to the monitoring program requirements. Comprehensive biological data bases (e.g., The National Phytoplankton Data Base originally developed by EPA's National Eutrophication Survey) have been almost nonexistent. There has been limited opportunity, at best, to evaluate biological monitoring data against similar data obtained in the same manner by other groups. For example, the Chesapeake Bay Program, which has cost millions of dollars to execute, could have benefited greatly had the many years of excellent data - gathered by the collective organizations with research inter- ests in the Bay - been available to all of the grantees/participants in some standard and uniform format. The standardization of selected biological methods, from sample collection through data analysis, has done much to lend credibility to biological measure- ment techniques. However, biological monitoring is a dynamic and emerging field; new methods must be developed, tested, and evaluated and guidelines for their use issued. Biological measurement is being applied with great expecta- tion in a new area, i.e., hazardous waste identification, for which it has not previously been evaluated. However, until biological data can consistently stand the tests of litigation, biological monitoring cannot come into its own as a primary regulatory tool. The increasingly complex monitoring problems of today, as typified by haz- ardous waste sites monitoring, require sophisticated and broad-spectrum approaches for resolution. Accelerated development of methods, including bio- logical methods, has been initiated to meet these needs. Many of the standards and controls promulgated by EPA are based upon biological testing and response. This is appropriate, as the primary mission of EPA is, simply stated, to safe- guard human and environmental health; and often the first indications of im- pending problems are detected through impacts on the biota. Physical or chemical testing cannot predict the cumulative impacts of mixed pollution sources upon the biotic components of the environment. Synergistic or antagonistic effects can only be determined through biological measurements. Approaches that assess the "net" impact of "black-box" effluents or leachates upon biological community components will find increasing value as rapid screens or guides to subsequent analysis or control procedures. The use of biological monitoring in conjunction with physical-chemical sensing is particularly advantageous in a systems approach. A strictly physical-chemical monitoring network that would monitor all suspected pollutants continually and throughout potentially impacted areas is very difficult and expensive to operate and could still fail to detect unsuspected or low-level pollutants. And often we don't have a complete list of pollutants which may impact water bodies. 12 However, analysis of biological samples, collected periodically from strategi- cally located stations, could detect pollution-induced changes in the biota and provide an alert to hazardous conditions. Foreign materials introduced into an aquatic environment interact in a complex and often nonlinear manner with one another and with the numerous other factors inherent to the environment. Aquatic organisms and communities respond to the sum of the interactions of these environmental factors. Thus, biomonitoring is particularly well suited to detecting changes in ambient conditions caused by both suspected and unsuspected foreign materials, even though the actual cause- effect relationships may be too complex to readily evaluate. Unfortunately, the often important effects of interactions between such factors as temperature, photoperiod, and flow rate, as well as toxic components in the ecosystem, can rarely be demonstrated in laboratory studies. In-situ biological monitoring remains the only effective way to measure the resulting complex impacts upon physiological and behavioral processes at the species, population, or community level. An additional and very significant advantage of biological monitoring is that it provides a mechanism for the integration of conditions between sampling periods. Aquatic communities are affected by short-lived perturbations of the environment and these effects normally persist for the weeks or months required for the communities to recover. Biomonitoring can be used to detect short-term events that can have catastrophic effects and which chemical/physical monitoring is likely to miss. Biological monitoring is especially advantageous because it will detect the full spectrum of suspected and unsuspected impacts including manifestations of intermittent insults even through periodic sampling. Pollu- tion is fundamentally a biological problem. We monitor certain chemical-physical parameters primarily because we know or suspect that they directly or indirectly impact living organisms. Once biological monitoring has shown that there is a problem, intensive physical-chemical and biological monitoring could pinpoint both the "danger spots" and the causative agents and sources. As additional information becomes available concerning the relationships between specific pollutants and specific changes in the biota, the identif icaiton of the pollutants and their sources will become easier. Only the complementary use of biological and physical- chemical monitoring will make it feasible to detect impacts and their sources over entire watershed systems. Major legislation that establishes requirements for biological measurement, monitoring, and assessment include: 9 Clean Water Act © Toxic Substances Control Act (TOSCA) © Solid Waste Disposal Act, as amended by the Resource Conservation and Recovery Act (RCRA) © Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) © Safe Drinking Water Act © Clean Air Act 13 Biological monitoring and assessment activities are presently conducted in support of such major efforts as: • Evaluation of environmental impact statements © Ecological assessments of spills or discharges of oil and hazardous materials e Evaluation of the "Clean Lakes" restoration efforts 0 Exposure assessment of toxic substances, pesticides • Lake classification surveys © Drinking water source evalutions © Basin-wide planning e NPDES compliance monitoring © Pre- and post-construction evaluations of energy resource development Areas in which it is anticipated that biological measurements will have increasing application include: © Hazardous waste identification associated with active and uncontrolled waste disposal sites and industrial process streams © Toxic substance and pesticide registration efforts o Dredge spoil monitoring e Ocean dumping impact assessment • Acid-rain impact assessment © Investigations relating exposure assessment to environmental and human health risks APPROACHES TO BIOMONITORING In-Situ Bioassays Measurements of the toxicity of pollutants provide information on their impact upon living organisms. Usually, bioassays are conducted in the laboratory; however, they can be conducted in-situ in the field. The placement of caged animals, in an outfall or receiving stream, is a direct approach to the measurement of the impact of the discharge of hazardous substances to surface waters. In-situ bioassays are subject to vandalism, flooding, and a host of other uncontrollable factors including exposure of test organisms to unknown constituents and dosages for indefinite periods of time. While not impossible, it is difficult to determine long-term effects of pollutants in the field, as confinement of the test animals may alter their natural growth, life expectancy, and reproduction. The results of the in-situ bioassays are apparent when the caged organisms die; however, responses may be less pronounced, and often measurements of physiological or behavioral responses such as respiration rate, movement, reproductive success, and development of tumors are required to detect environmental effects. Chronic exposure to low-level concentrations may have no measurable effect on the test organisms themselves, but the effect of exposure may be manifested several generations later in the form of abnormalities resulting from such exposure. 14 Bioaccumulation Living organisms have the ability to accumulate substances from the surround- ing water medium and thus act as natural compositors and integrators of hazard- ous substances present in surface waters. Plants and animals accumulate hazardous substances in tissue through biocon- centration (uptake from the water) and biomagnif ication (uptake through the food chain). For example, the concentration of radionuclides in algae at 10,000 times background levels have been documented, and the impacts of DDT upon top-of-the-food-chain organisms such as the osprey and brown pelican are well-known. The uptake characteristics of organisms vary by individual species and trophic level. Ideally, receptors at several trophic levels should be included in bioconcentration studies. The most obvious advantage of bioaccumu- lation studies is that they provide detection of hazardous materials, often at levels below analytical limits, and suggest the potential hazard to various food chain components, including man. Thus, these analyses offer an efficient means for screening and identifying potentially hazardous substances in water before they pose a serious risk to human or environmental health. Indicator Organisms Attempts have been made to use the abundance, presence, or absence of taxa (species or genera) as stand-alone indicators of environmental conditions. These assume that the pollutant or pollutants will have an effect on the indi- cator organisms and that it is possible to predict the abundance, presence, or absence of the organism in its unimpacted aquatic habitat. This is often very difficult, especially with groupings above the species level. For example, we have found that phytoplankton genera thrive over such a broad range of environmental conditions that they cannot be used as indicator organisms. In addition, seasonal variations in individual community components can effectively mask changes resulting from pollutant impact. Only with adequate base-line biological data can meaningful inferences be made under such conditions. Community and Population Changes The determination of community and population changes (e.g. , species composi- tion, diversity and abundance; standing crop; mortality rates; and community productivity) have been much more successful in measuring the response of aquatic communities to pollution-induced stress. However, it is difficult to obtain reliable measurements of this type because of the need to sample highly variable populations where composition and structure are subject to all the perturbations of the environment, both natural and man-induced. For example, immature aquatic insects are highly sensitive to pollution-induced stress conditions, but they also show large natural seasonal variations due to charac- teristics of their life cycles and natural fluctuations in stream conditions. In addition, identification of many groups of these insects poses special problems, due largely to the lack of taxonomic information on their immature stages. In spite of the dif f f iculties involved in characterization of aquatic communities, be they fish, invertebrates or plant life, carefully planned and executed investigations are well worth the effort. They offer a fairly efficient 15 means of detecting pollution-induced stress once the natural community patterns have been fully described. Once again, however, the existence of "pre-impact" base-line data is invaluable in interpreting the nature and magnitude of the changes noted. Species Health The health of organisms can be used to determine if hazardous wastes are impacting the aquatic biota or are potentially harmful to man. Parameters which can be measured to determine species health include growth, reproduction, longevity, plumpness, tumor development, and phenotypes. Where possible, "control" measurements, e.g., from upstream-downstream studies or analogies with similar systems, should be obtained to determine the significance of the changes noted. SELECTION OF BIOMONITORING TECHNIQUES The factors listed below should be considered in selecting biomonitoring techniques. e Status of development of the technique (e.g. , testing and validation) • Sensitivity (response to low concentrations of contaminants) • Simplicity (how much training and experience required to conduct) • Reproducibility (intra- and interlaboratory precision) • Coverage (spectrum of contaminants to which technique is responsive) • Cost (set-up, as well as sample-by-sample) • Turnaround time (compatible with monitoring requirements) • Confidence (past acceptance and success in application) • Nature of response of the technique (relationship of response to human and/or environmental health) Status of Development While some of the biomonitoring techniques have been validated, many have not been evaluated and/or validated relative to specific program requirements, e.g., monitoring hazardous wastes. Some promising new techniques are still too developmental for serious consideration at this time although further evalua- tion and testing are strongly recommended. Sensitivity Particularly in characterizing complex waste mixtures, sensitivity to low concentrations of bioactive compounds is an important consideration in predict- ing the potential impact of that waste upon human and environmental health. Response to low concentrations of contaminants can often be enhanced through 16 increasing the contact time of the test in the field. The time- sensitivity "trade-off" is only one of several that must be made in determining the final composition of selected biomonitoring techniques. Simplicity It is expected that personnel with a wide range of training and capabilities will be conducting and interpreting the biomonitoring. The more simple and straightforward the techniques are, the more likely comparable and meaningful results will be obtained by the spectrum of users. Even relatively straight- forward procedures should be documented in sufficient detail to assure that data from various biomonitoring activities are comparable. Reproducibility Techniques which are less critical with respect to test conditions and yield similar results in the hands of a variety of operators are favored. In some cases, the "effective reproducibility" may be a function of the ambiguity of the test end point (e.g., subtle behavioral responses versus mortality), the clarity of test procedure documentation, or the skill of the operator in inter- preting monitoring results. In general, only techniques that have been shown, through comparison studies, to give relatively uniform responses should be considered for general use. Coverage In assembling a battery of techniques to determine the impact on waterbodies, it is vital that there be as broad coverage of the spectrum of potential hazard- ous substances as possible. In general, "specialty tests" - for very limited groups of contaminants - have little general monitoring application unless the contaminants are of high priority and not effectively addressed by more broad- spectrum techniques. Synergistic or antagonistic effects of hazardous substances in the aquatic environment should be reflected in the biomonitoring results. Cost If the costs associated with individual techniques can be kept to a reasonable level, monitoring groups will be encouraged to use biological testing as part of their programs. Turnaround Time The time required to conduct many biological monitoring programs has hindered their general acceptance for many applications. Techniques requiring long-term baseline measurements (months or years) will not be used as extensively as those where meaningful interpretations can be made in a relatively short time frame. 17 Confidence Somewhat less tangible, but nonetheless important, is the credibility gained by certain tried-and-true biomonitoring techniques. Often the confidence associated with performing "familiar" tests is reflected in the successful application of that technique. A sensitivity is developed to disturbances or interferences in the basic technique which often detects problems before they are reflected in bad data. Well established techniques also tend to carry the weight of precedence in enforcement proceedings. Nature of Response While the relationship of some biological tests to human health and welfare are rather clearly defined, other biological techniques bear more obscure relationships. Absence of direct pharmacological relevance does not rule out the consideration of specific techniques; however, strong empirical evidence is required to support such techniques. CONCLUSIONS Biological monitoring has had a relatively weak supportive role in monitoring programs of the past; however, it is rapidly attaining the recognition it warrants in a systems approach to monitoring. Physical and chemical testing cannot predict the cumulative impacts and/or the interactions of mixed pollution sources on environmental health. Biological monitoring is well-suited for this as the biota in aquatic systems responds to the sum of the interactions of environmental factors and integrates their effects over time. Biological monitoring is a dynamic and emerging field and new methods must be developed, tested, and evaluated to meet the challenging monitoring requirements associated with the identification and assessment of hazardous wastes. For selection of appropriate biomonitoring techniques, • The monitoring requirements of a specific program should be clearly determined, • The selection criteria provided should be reviewed and prioritized relative to those monitoring requirements identified, and • Finally, the selection criteria should be rigorously applied in determining those tests or measurements needed to meet program objectives Biological monitoring methods which are sensitive, simple to apply, repro- ducible, and economical in their application, and which provide answers to vital environmental questions in realistic time frames, represent a goal toward which our research and monitoring groups must continually strive. 18 ENZYMES AND OTHER BIOCHEMICAL INDICATORS OF TOXICANT EFFECTS IN FISHES1 by Dr. Dora R. May Passino Great Lakes Fishery Laboratory U.S. Fish and Wildlife Service Ann Arbor, MI 48105 The major purpose of this workshop is to evaluate in-situ biological sensors of water quality. This paper will provide an overview of the subject of enzymes and other biochemical indicators of toxicant effects in aquatic organisms, especially fish. To protect endemic and economically important fish within a particular ecosystem, water-quality biologists must measure biochemical indicators in fish species that have sensitivities comparable to the fish in the ecosystem. Because I represent the Great Lakes Fishery Laboratory, I will emphasize the Laurent ian Great Lakes, an ecosystem containing valuable fishery resources. The Great Lakes The Laurentian Great Lakes, having 95,000 square miles of water surface and extending 2,000 miles, are the world's largest reservoir of fresh water. For more than a century, many of the largest and most successful cities and industries of the Midwest have depended on this water as a vital element in their growth. Billions of gallons of municipal, industrial, and agricultural wastes have been discharged into these lakes because of the convenience of doing so. The dumping of wastes and physical alteration of the shores and waterways has been at the expense of the aquatic ecosystem. The fish communities of each of the Great Lakes have undergone marked transformation, especially during the last half century. Populations of endemic climax fish like the lake trout (Salvelinus namaycush) , walleye (Stizostedion vitreum) , and coregonid species have been drastically reduced as a result of overexploita- tion; invasion of the Great Lakes by sea lampreys (Petromyzon marinus), alewives (Alosa pseudoharengus) , and rainbow smelt (Osmerus mordax) ; introduction of Pacific salmon (Oncorhynchus spp.); and degraded environmental quality. A problem of major concern in Great Lakes research is the apparent reproduc- tive failure of lake trout in the Great Lakes. Lake trout spawn and produce viable gametes, but no survival past the fry stage has been observed except in Lake Superior. Although lake trout once supported valuable commercial fisheries in each of the Great Lakes, the species was probably extinct by 1956 in all but Lake Superior, which retained a residual population. The decline and near extinction of the species are believed to have been caused by a combination of excessive exploitation and of heavy predation by the sea lamprey, which invaded the Great Lakes through man-made canals connecting the lakes with the St. Law- rence River and the sea. In spite of control of the sea lamprey and the plant- ing of -millions of lake trout in the upper three lakes (Superior, Michigan, Contribution 000 of the Great Lakes Fishery Laboratory, U.S. Fish and Wildlife Service, Ann Arbor, Michigan 48105 19 Huron) since the mid-1960s, natural reproduction of planted fish has been known to occur in only two locations. Young fry have been collected in grand Traverse Bay (Lake Michigan) and Marquette harbor (Lake Superior). After a decade of research, the roles of various biological and environmental factors, including contaminants, are not yet clearly understood. Over the past 10 years, several major contaminants, notably mercury, DDT, PCBs, dieldrin, and mirex, have posed a threat to the Great Lakes ecosystem, caused curtailments of commercial fishing, and resulted in warnings to sport fishermen against consumption of fish. Regulation of the sources has resulted in a general decline of most of these contaminants, although substantial residues remain in the sediments and aquatic biota. New contaminants continue to be identified in the Great Lakes; among the more recent are chlorinated benzenes, chlorinated styrenes, toxaphene-like compounds, and dioxin. Aquatic toxicological data are needed for use in hazard assessment schemes that are being developed to avoid future crisis situations like those with mercury, DDT, and PCBs. Great Lakes Fishery Laboratory The Great Lakes Fishery Laboratory, Ann Arbor, Michigan, is administered by the Division of Fishery Ecology Research of the U.S. Fish and Wildlife Service. The mission of the Laboratory is to develop the knowledge and the technical basis for assessing, protecting, enhancing, and rehabilitating the valuable fishery resources of the Great Lakes and their habitat. This mission is pursued by a staff of over a hundred. The Laboratory operates five major research vessels, which are used to survey Great Lakes fish stocks with trawls, gillnets, larval-fish tow nets, hydroacoustic fish-detection systems, and other gears, and sample their habitat for information about its limnological characteristics, including the presence of toxic substances. The Laboratory, together with State and Canadian resource agencies, is working to determine the distribution and abundance of important sport and food fishes, the effects of fishing and other mortality agents on the stocks of these fishes, and the size and potential of populations of forage fish like the alewife and smelt, on which important game and food fishes feed. The Laboratory also tries to project the effects on fishery resources of alterations of the habitats caused, by example, by pollution, water withdrawal for power-plant cooling systems, channel-maintenance dredging, and shoreline modification such as bulkheading and draining marshes. Contaminant research at the Great Lakes Fisher Laboratory, which is funded by the Service's Environmental Contaminants Evaluation (ECE) program, is divided into two projects - Contaminant Toxicology, and Contaminant Dynamics - comprising eight work units. (See appendix A.) (Other ECE research units outside this Laboratory are listed in appendixes B and C.) The Contaminant Dynamics project is involved in studies of contaminant uptake from water and sediments, trend monitoring, and, more recently, field studies on effects of contaminants. The Contaminant Toxicology project, of which I am the leader, was initially oriented toward mode-of-action studies of contaminant effects on enzymes. Although the emphasis of the project has changed toward whole-animal studies, my associates 20 and I will in the future measure enzymes and other biochemicals in aquatic organisms whenever this approach seems likely to provide useful information to managers of fishery resource and water quality. Enzymes General review. For about two decades, water quality biologists have been searching for sensitive indicators of sublethal effects of contaminants to attempt to understand the mode of action of toxicants, and to develop a basis for corrective action in cleaning up water bodies before the health of aquatic populations is seriously threatened. Enzymes are attractive as indicators because they are more easily quantified than are other indicators of sublethal contaminant effects, such as changes in behavior. Useful precedents have been set in clinical medicine in the successful diagnosis of disease and evaluation of exposure to industrial chemicals or drugs by analyses of such variables as enzymes, blood chemistry, and liver function. In the fishery field, however, enzyme research is in its infancy: only two fish enzymes have yet been proven to be useful indicators of toxic effects of specific contaminants or classes of contaminants. The first of these enzymes is acetylcholinesterase, which modulates the amounts of the neurotransmitter acetylcholine in the nervous system. The mode of action of malathion, parathion, and other widely used organophosphate insecticides is inhibition of this enzyme. Measurement of acetylcholinesterase in the brains of distressed fish collected in the field has been successfully used to diagnose organophosphate poisoning of fish in natural waters (Coppage et al., 1975). The second indicator enzyme is delta-amino levulinic acid dehydratase (ALA- D), which is in the pathway of heme synthesis. Fish exposed chronically to lead suffer from anemia, which is probably a result of the disturbance of hemoglobin synthesis by the blocking effect of lead on ALA-D. Inasmuch as Hodson et al. (1977) reported that near-lethal exposures of rainbow trout (Salmo gairdneri) to cadmium, copper, zinc, and mercury did not significantly inhibit erythrocyte ALA-D, the enzyme appears to be specific for lead. In addition, the concentrations of lead that inhibit ALA-D activity during short-term exposures of fish are similar to concentrations causing other sublethal symptoms such as black tail, spinal curvature, and erosion of the caudal area of rainbow trout after many months of exposure. Hence, ALA-D fits the criteria for a short-term enzyme indicator of long-term sublethal effects of a contaminant on fish. The Columbia National Fishery Research Laboratory of the Service will be measuring blood ALA-D in fish exposed environmentally to lead released from mine tailings in Missouri. Their results may provide field verification of the usefulness of this enzyme. An enzyme that shows promise as an indicator is adenosine triphosphatase (ATPase) , a widely distributed enzyme involved in the energy-requiring active transport of electrolytes across membranes. The pesticide DDT inhibits mitochondrial ATPase from several fish tissues. However, different types of ATPases from some tissues may be activated in response to toxicants. The ATPases are not specific indicators of DDT exposure - inhibition by toxaphene, PCBs, and pyrethrins has also bee*T~reported. 21 Another approach to using enzymes as indicators of toxicant exposure is measuring the enzymes that metabolize or detoxify contaminants. Generally, when animals are exposed to toxicants, the detoxifying enzymes are induced - that is, more enzyme is synthesized. A good example of how detoxifying enzymes can be used to monitor specific pollutants was given by Walton et al. (1978), who found that hepatic aryl hydrocarbon hydroxylase (AHH) in the cunner (Tautogolabrus adspersus), a marine fish living off New England, was a sensitive indicator of petroleum contamination. Oil concentrations of 1 to 2 mg/1 caused twofold to sixfold induction of AHH above normal levels. Observation of elevated AHH followed oil contamination by 1 or 2 days. Because induction of AHH decayed in less than 7 days after exposure ceased, it reflected the current state of oil contamination. Feeding of crude oil or the tissue of mussels contaminated with oil at a concentration of 500 mg/kg caused up to fivefold induction. In-vitro tests with trout showed that AHH was not induced by PCBs, or by representative organochlorine, organophosphate, or carbamate pesticides. Research at Great Lakes Fishery Laboratory. We did some in-situ enzyme work in the Contaminant Toxicology project a few years ago. After completing in-vitro experiments to determine effects of heavy metals, PCBs, and DDT on the uricolytic enzyme allantoinase in lake trout liver (Passino and Cotant, 1979) and in-vivo experiments with PCBs, we collected lake trout from two areas - one near Isle Royale, a clean area in northern Lake Superior, and one near Saugatuck, Michigan, a moderately contaminated area influenced by industrialization, typical of southeastern Lake Michigan. We found significantly (P_ < 0.01) greater enzyme activity in small than in large lake trout from Lake Michigan. (See table 1.) Enzyme activity was also higher in the smaller fish from Lake Superior and from stocks in the Laboratory, but not significantly so. Although the interpretation of data from field samples is complicated by many unknown variables operating on enzyme systems in fish, we demonstrated an inverse correlation between enzyme activity and length of lake trout. This clear trend in activity may have been caused in part either by the greater age of the longer fish, or by higher levels of contaminants in older fish - especially in Lake Michigan. (See table 2.) This example illustrates the difficulty in establishing a causal relationship between residue levels and observed enzyme activities in fish collected in the field. The only exceptions may arise after clear-cut demonstration in the laboratory of specificity of a particular contaminant or class of contaminants for a particular enzyme. Another piece of work related to enzymes was that of measuring the subcellular distribution of chemical contaminants in lake trout liver. Such determination holds promise as a means for predicting which metabolic pathways are most likely to be affected, because each subcellular fraction or organelle is specialized for particular functions. Although we investigated mercury, PCBs, and DDT plus DDE, only the data on mercury are relatively free of problems. These data were obtained by collecting large live lake trout in Lake Michigan near Saugatuck, bringing it live to the laboratory, killing it, and immediately removing the liver and separating the subcellular fractions by differential centrifugation. One set of subcellular fractions was examined by electron microscopy to quantify the organelles present and the other set was analyzed for total mercury by a combustion amalgamation technique. (In addition, we had earlier obtained data 22 from a frozen liver.) The organelles of interest are shown in figure 1. Mercury was primarily (80%) in the submicroscopic material (see figure 2) - that is, in material so small it could not be seen at 7,700 x magnification. Of the remainder, 5% was in the mitochondria, 5% in the rough endosplasmic reticulum, and 1 to 2% in some of the other organelles. What does this distribution mean in terms of fish metabolism and the health of the fish? In other vertebrates, such as the rat, a metal-binding protein, metallothionein, has been found in liver and kidney. This protein serves as a protective mechanism in animals, because the protein-bound metal is not available to interfere with metabolic processes. Therefore, I suggest that lake trout may be able to protect themselves to a large extent from the toxic effects of mercury by concentrating mercury in the submicroscopic material of the liver, where the mercury is probably bound to a metallothionein-like protein. However, since other recent investigations on rainbow trout and goldfish (Carassius auratus) (Olson et al. , 1978; Marafante, 1976) showed that no more than 50% of mercury was bound to metallothionein-like proteins, some interference of mercury with metabolic pathways in the cytosol is possible. These pathways include fatty acid synthesis, glycolysis, many reactions in gluconeogenesis, the hexose monophosphate shunt, and activation of amino acids for protein synthesis. Similar results were obtained for the distribution of PCBs and DDT plus DDE in frozen liver. Although 96 to 97% of these organochlorines were found in the soluble fraction, the mitochondria were ruptured during the freezing and thawing process; consequently, confirming work needs to be done with fresh liver. The logical follow-up research would be to determine if the levels of contaminants present are inhibitory to key enzymes in these fractions, especially those in the cytosol. We did not do the follow-up research because it would have entailed a mode-of-action study at the subcellular level. Because fishery and water quality managers need to have technical information that is directly useful in their decision-making process, we are now conducting whole-animal studies, with emphasis on effects on growth and reproduction. Changes in biochemical indi- cators in animals exposed to toxicants can be useful in predicting long-term effects on populations. Other Biochemical Indicators As part of the continuing research on the reproductive failure of lake trout in Lake Michigan, we have been involved in two interproject studies (with the Contaminant Dynamics project) to evaluate the toxicity of major organochlorine contaminants to lake trout eggs and fry, which are critical life stages. In one study (1977), 20,000 fry from Lake Michigan lake trout were exposed to simulated Lake Michigan levels of PCBs (Aroclor 1254) and DDE, and to levels 5 and 25 x higher, in food and water. Ambient levels were 10 ng/1 PCBs and 1.0 ng/1 DDE in water and 1.0 ug/g PCBs and 0.1 yg/g DDE in food. Cumulative mortality of the fry exposed to simulated Lake Michigan levels of PCBs and DDE for 6 months was 40.7 - nearly twice that of unexposed (control) fry - and mortality at the highest exposure level was 46.5%. We concluded that, although several factors have undoubtedly contributed to the reproductive failure of lake trout, 23 the levels of PCBs and DDE in Lake Michigan during the early and mid-1970s were high enough to significantly reduce survival of fry in the lake and thereby impede restoration of the lake trout population to self-sustainability. During the course of this study, two interesting trends were observed that warrant further investigation. One trend was toward a slight increase in growth rate of the exposed fry. A second trend was toward an alteration in glycolysis in exposed fry that had been exercised to exhaustion in a swimming apparatus. Further work needs to be done to measure oxygen consumption, muscle lactate, and muscle glycogen in fry before and after swimming to exhaustion to determine if their stamina is affected by exposure to organochlorine contaminants. In the second interproject study (1979), we exposed eggs and fry from Lake Michigan lake trout to simulated Lake Michigan levels of organochlorines. Nominal concentrations were 1 ng/1 PCBs (Aroclor 1254), 0.2 ng/1 DDE, 1 ng/1 dieldrin, and 0.9 ng/1 technical grade chlordane, and 10 x multiples of these concentrations. No differences in mortality were observed between exposed and unexposed eggs and fry through 180 days of exposure, to the "swim-up" stage. Measurements of lengths taken from the time of hatch almost until swim-up (age about 60 days) showed a trend, significant at the 10% level, for increased growth of fry exposed to the lower levels of organochlorines, by time of swim-up. We also used a biochemical indicator of instantaneous growth by measuring RNA- DNA ratios. A high RNA-DNA ratio is correlated with rapid protein synthesis and weight gain. At swim-up, the RNA-DNA-ratio of the control fry (3.85 + 0.11 (+ SE) was significantly lower (P_ < 0.10) than that of the fry receiving higher levels of the organochlorines (4.38 + 0.04). The RNA-DNA ratio for the low treatment fry was intermediate (4.08 + 0.15). Thus the trend toward increased growth of fry exposed to organochlorines was supported by the RNA-DNA data. Kearns and Atchison (1979) demonstrated that field-collected yellow perch (Perca flavescens) showing different growth rates, as measured by changes in weight or length, had parallel changes in the RNA-DNA ratios. In their study, the growth rate of yellow perch exposed environmentally to cadmium and zinc from an electroplating plant was slower than that of controls in a cleaner area upstream from the plant. However, contaminated fish grew fastest in late summer, and uncontaminated fish in midsummer. The authors cautioned against comparing growth rates of fish in two populations at a single sampling period. Measurement of RNA-DNA ratios shows promise as an in-situ indicator of contaminant effects. Other biochemical indicators that could be examined when investigating repro- ductive failure of fish populations are reproductive hormones. The importance of reproductive hormones in enabling fish to spawn and produce normal young is well known. Hence, alterations in reproductive hormone levels by exposure to contaminants is a serious threat to the productivity of populations. Sangalang and Freeman (1974) monitored testosterone and 11-ketotestosterone in brook trout (Salvelinus fontinalis) during low-level (1 yg/1) exposure to cadmium. The concentrations of these two androgens were different in the fish exposed to cadmium, suggesting impairment in the clearance and use of testosterone and 11-ketotestosterone. The impaired clearance appeared to be related to the 24 abnormal vasculature and hemorrhagic necrosis observed in the testes of some of the cadmium-treated fish. These results are significant because brook trout spawn in rivers and streams that are prone to industrial pollution. Further studies are needed of fish reproductive hormones, especially field verification by the measurement of reproductive hormone levels in environmentally exposed fish. Investigators at the Columbia National Fishery Research Laboratory, who used bone composition and strength as sublethal indicators of contaminant effects, found that toxaphene exposure of fathead minnows led to decreased collagen concentration, altered amino acid composition, increased calcium concentration, and decreased bone strength of the vertebrae (Mehrle and Mayer, 1975). These biochemical manifestations of the "broken back" syndrome have been recently observed in striped bass (Morone saxatilis) collected in tributaries of Chesapeake Bay. Striped bass with higher body burdens of organochlorines and metals have lower collagen and decreased bone strength. Other types of biochemical indicators are those associated with the response of fish to stress, contaminants being one type of stress. Response to stress involves both the primary effects (hormones) and secondary effects, as shown in figure 3. Steroid hormones that have been investigated relative to contaminant effects are corticosteroid (cortisol, corticosterone) and catecholamines (epinephrine, norepinephrine). These hormones are part of the adrenal-pituitary response of vertebrates to stress. In our earlier (1977) study on lake trout fry, we measured Cortisol by radio immunoassay. No differences were found between exposed and unexposed fry. However, the variability of Cortisol within treatments was high. The fry were too small for sampling blood and the Cortisol had to be extracted with methylene chloride from homogenized whole fry. Although Cortisol may have been elevated during the initial exposure of the fry, the fish may have become adapted to the low-level chronic exposure. Schreck and Lorz (1978) found that exposure to copper resulted in a marked, dose-dependent serum Cortisol elevation in coho salmon (Oncorhynchus kisutch) . However, treatment with cadmium did not elicit a Cortisol elevation, even in moribund fish. They speculated that some but not all stresses produce elevations of Cortisol, and that the mammalian concept of General Adaptation Syndrome may not completely apply to fish. In addition, Cortisol may not be useful to detect effects of continuous exposure to sublethal contaminants, since Cortisol is elevated primarily during the initial period of exposure to the stressful agent and may later return to normal levels. Secondary response to stress includes changes in blood glucose and lactate, plasma free fatty acids, electrolyte balance, liver glycogen, immunosuppression, etc. (Mazeaud et al. , 1977). The clinical approach to blood chemistry is successfully applied to human beings because normal baseline values for such properties as electrolytes, glucose, and hematocrit have been established. For fish, however, such baseline values have yet to be established, especially for wild populations, although some progress is being made. Since homeostasis is not so finely tuned in poikilothermous fish as in homothermous animals, the range f-or normal values is much broader for fish. Consequently, deviations from normal must be greater to be statistically significant. In addition, the physiology of fish is profoundly influenced by factors such as season, spawning condition, and migrations. 25 Research and Instrument Needs Research. The above statements point out some of the challenges in applying biochemical information in a way that is helpful to managers of fishery resources and water quality, especially those charged with establishing water-quality criteria. An observed change in a biochemical indicator is significant to a manager only if it can be shown to influence higher levels of organization, ultimately the population level. An observed effect must be recognizable as being caused by contaminants and not by natural or endogenous agents. Background data on normal values and expected variability are required. The biochemical indicator to be considered for field monitoring should not be altered by the stress caused in capturing the fish. Because of these and other requirements, blood chemistry indicators such as glucose, electrolytes, and hemoglobin are not now particularly very useful because of the difficulties of establishing normal values and the changes that occur during capture of the fish. Enzymes that are specific for particular contaminants appear to be the most promising, together with reproductive hormones that show a significant response to contaminants. More research is needed to find enzymes and hormones that show specificity toward particular contaminants. Most likely, no one biochemical indicator will suffice to diagnose the health of fish, but rather a combination of indicators. Practitioners of clinical medicine normally measure an array of physiological and biochemical indicators before diagnosing disease. Wedemeyer and Yasutake (1977) have published methods for clinical indicators useful in assessing the health of fish populations. Research is needed to integrate laboratory and field studies and to develop correlations between adverse effects in laboratory animals and adverse effects in wild populations. Laboratory studies provide high precision for predicting effects of individual contaminants under controlled conditions, but these studies fail to duplicate the many interacting environmental variables. Field studies address real environmental problems but fail to establish cause-and- effect relationships and are usually not sensitive enough to detect adverse effects before they reach crisis proportions. Although the emphasis of this workshop is on in-situ or field indicators, it must be remembered that an integrated approach with both field and laboratory studies is needed to interpret and understand the response of aquatic organisms to contaminants. Instruments. Although this paper is primarily an overview, a few equipment or chemical needs should be mentioned. First, much time could be saved by biologists if they did not need to construct their own diluters for bioassays. Ideally, such marketed diluters should be portable enough to fit into a mobile or floating laboratory. Second, in-situ analytical monitors of chemical concentrations are needed in bioassay systems. These monitors could be used either for the toxicant in the bioassay system or perhaps for serum enzymes in larger fish. Some automated enzyme systems are already available for clinical laboratories and might be adapted for use with fish. Monitors for toxicants in a bioassay system would not need to identify unknowns, but only quantify a known chemical. Lastly, some fish hormones look promising as indicators of contaminant effects. The most sensitive methods generally are radioimmunoassay. However, all the available kits for radioimmunoassay are based on human antigens. Marketing of kits with antibodies to fish antigens is needed. Other specific instrument or chemical needs are considered in other papers in this compilation. 26 References 1. Coppage, D. L. , E. Matthews, G. H. Cook, and J. Knight. Brain acetyl- cholinesterase inhibition in fish as a diagnosis of environmental poison- ing by malathion, 0,0-dimethyl S-( 1 ,2-dicarbethoxyethyl) phosphorodithi- oate. Pesticide Biochemistry and Physiology. 5, 536-542. 1975. 2. Gingerich, W. H. , and L. J. Weber. Assessment of Clinical Procedures to Evaluate Liver Intoxication in Fish. Environmental Protection Agency, Duluth, Minnesota. EPA-600/3-79-088. 1979. 3. Hodson, P. W. , B. R. Blunt, D. J. Spry, and K. Austen. Evalution of erytho- cyte <5-amino levulinic acid dehydratase activity as a short-term indicator in fish of a harmful exposure to lead. Journal of the Fisheries Research Board of Canada. 34, 501-508. 1977. 4. Kearns, P. K. , and G. J. Atchison. Effects of trace metals on growth of yellow perch (Perca f lavescens) as measured by RNA-DNA ratios. Environmental Biology of Fishes. 4, 383-387. 1979. 5. Lehninger, A. L. Biochemistry, 2nd ed. Worth, New York. 1975. 6. Marafante, E. Binding of mercury and zinc to cadmium-binding protein in liver and kidney of goldfish (Carassius auratus L.). Experientia. 32, 149-150. 1976. 7. Mazeaud, M. M. , F. Mazeaud, and E. M. Donaldson. Primary and secondary effects of stress in fish: some new data with a general review. Transactions of the American Fisheries Society. 106, 201-212. 1977 8. Mehrle, P. M. , and F. L. Mayer, Jr. Toxaphene effects on growth and bone composition of fathead minnows, Pimephales promelas. Journal of the Fisheries Research Board of Canada. 32, 593-598. 1975. 9. Olson, K. R. , K. S. Squibb, and R. J. Cousins. Tissue uptake, subcellular distribution, and metabolism of CHoHgCl and CH^ HgCl by rainbow trout, Salmo gairdneri. Journal of the Fisheries Research Board of Canada. 35 , 381-390. 1978. 10. Passino, D. R. M. , and C. A. Cotant. Allantoinase in lake trout (Salvelinus namaycush) : In vitro effects of PCBs, DDT, and metals. Comparative Bio- chemistry and Physiology. 62C, 71-75. 1979. 11. Sangalang, G. B. , and H. C. Freeman. Effects of sublethal cadmium on maturation and testosterone and 11-ketotestosterone production in vivo in brook trout. Biology of Reproduction. 11, 429-435. 1974. 12. Schreck, C. B. , and H. W. Lorz. Stress response of coho salmon (Oncorhynchus kisutch) elicited by cadmium and copper and potential use of Cortisol as an indicator of stress. Journal of the Fisheries Research Board of Canada. 35, 1124-1129. 1978. 27 13. Walton, D. G., W. R. Penrose, and J. M. Green. The petroleum-inducible mixed-function oxidase of cunner (Tautogolabrus adspersus Walbaum 1792): Some characteristics relevant to hydrocarbon monitoring. Journal of the Fisheries Research Board of Canada. 35, 1547-1552. 1978. 14. Wedemeyer, G. A., and W. T. Yasutake. Clinical methods for assessment of the effects of environmental stress on fish health. U.S. Fish and Wildlife Service Technical Paper No. 89, 1-18. 1977. 28 APPENDIX A. ENVIRONMENTAL CONTAMINANTS EVALUATION - CURRENT AND FUTURE WORK UNITS AT THE GREAT LAKES FISHERY LABORATORY Contaminant Toxicology Project 1. Effects of contaminants on the invertebrate forage base of Great Lakes fish populations. 2. Effects of toxic substances on metabolism of fry from Lake Michigan lake trout. 3. Toxicological evaluation of recently identified contaminants in the Great Lakes. Contaminant Dynamics Project 1. Surveillance of contaminants in Great Lakes fishes. 2. Effects of multiple-contaminant exposures of eggs and fry of Lake Michigan lake trout. 3. Seasonal variation of organic contaminants and lipids in Lake Michigan fishes. 4. Size-specific mortality of Great Lakes lake trout fry with different burdens of contaminants. 5. Survival of lake trout eggs and fry in waters of the Upper Great Lakes. 29 APPENDIX B. ENVIRONMENTAL CONTAMINANTS EVALUATION - CURRENT WORK UNITS AT THE COLUMBIA NATIONAL FISHERY RESEARCH LABORATORY Columbia National Fishery Research Laboratory, MO. 1. Acute toxicity of environmental contaminants to freshwater fish and fish food organisms. 2. Development of methods and experimental conditions for assessing the acute toxicity of toxic materials and contaminants, identified in monitoring programs, to fish and fish food organisms. 3. Influence of temperature, turbidity, and pH on the toxicity of insecticides and herbicides used on cotton and soybeans to fish and fish food organisms. 4. Susceptibility of fish and fish food organisms following sublethal exposures to major classes of environmental contaminants. 5. Analysis of the relationship between physical, chemical, and bio- logical properties of toxic chemicals for predictions of biological effects in fishery resources. 6. Influence of natural environmental factors on chronic toxicity and residue dynamics of hexachlorocyclopentadiene in the fathead minnow. 7. Impact of inorganic contaminants from mineral and energy development on aquatic plants and salmonid fisheries of the Western U.S. 8. Integration and correlation of toxicological parameters for compre- hensive hazard assessment of PCB replacements and herbicides. 9. Biological significance of aquatic contaminant residues identified by the National Trend Monitoring Program (e.g., phenolics and industrial intermediates) to fish food organisms, salmonids, and channel catfish. 10. Effects of herbicides used in hardwood forest and range management practices on the reproduction of daphnids and emergence of midges. 11. Development of methods and experimental conditions for chronic tox- icity studies to better evaluate the impact of toxic chemicals on fishery resources. 12. Collation and assessment of information, and formulation of research design and objectives, to investigate suspected and future contami- nants of critical aquatic resources. 13. Partial-chronic studies comparing intermittent and continuous exposures of rainbow trout to toxic pollutants. 30 14. Development of laboratory procedures for assessing the influence of contaminants on rates of detritus conversion in streams impacted by forest management practices and land-use changes. 15. The secondary impact of agricultural chemicals (dimilin, dylox, sevin) on warmwater fish production due to destruction of invertebrate forage. 16. Investigations of the effect of pentachlorophenol (herbicide, wood preservative) on the productivity of warmwater ecosystems, with emphasis on the limnological factors that contribute most to ecosystem stability. 17. The fate of pentachlorophenol (herbicide, wood preservative) in warmwater ecosystems: uptake, persistence, degradation, and effects on aquatic plants and fish food organisms. 18. Effects of herbicides used in and along irrigation and waterway canals on survival and development of rainbow trout eggs and fry. 19. Feasibility of "herding" fish with copper sulfate and other selected chemicals out of ponds and irrigation canals to be treated with toxic herbicides. 20. Effects of controlled-release formulations of aquatic herbicides used in irrigation canals on survival of rainbow trout and invertebrates. 21. Field evaluation of chemically "herding" fish under canal and reser- voir conditions. 22. Delayed toxic effects of sublethal exposure of salmonids to aquatic herbicides associated with irrigation systems. 23. Salinity tolerance of hybrid grass carp. 24. Toxicity and biological availability of contaminants associated with particulate matter from coal-fired power plants to yellow perch and fish food organisms. 25. Effects of drilling and of crude and refined petroleum products on walleye, yellow perch, and aquatic food organisms of the Northern Great Plains. 26. Impact of herbicides associated with no-till agriculture on the fishery of the Northern Great Plains. 27. Toxicologic screening of contaminants against naiads of stoneflies and fingerling cutthroat trout. 28. Avoidance of selected herbicides and Wyoming crude oil by cutthroat trout. 29. Toxicity of mixtures of range management chemicals to cutthroat trout. 31 30. Accumulation in cutthroat trout of different hydrocarbon fractions from a reconsituted crude oil and the resulting biological effects. 31. Accumulation in cutthroat trout and resulting biological effects from petroleum hydrocarbons lost to surface water during refinement of crude oil. 32. Effects of several candidate mosquito insecticides on survival, drift, and emergence of Pteronarcella badia under simulated use patterns of exposure. 33. Development of hazard assessment techniques for evaluating the sublethal impact of toxic chemicals on reproduction, growth, and adaptability in rainbow trout, brook trout, and channel catfish. 34. Application of biochemical, physiological, and histopathological techniques to field monitoring studies: development of biological indicator techniques for assessing the impact of agricultural, industrial, and energy-related chemicals on fishery resources. 35. Influence of industrial and agricultural chemicals on biochemical profiles of mixed function oxidases (MFO) and related enzymes in fish and invertebrates from the Great Lakes and Lower Mississippi Valley. 36. Impact of contaminants from mining and energy development on critical habitat for Northwestern and Alaskan salmonid and anadromous fish populations. 37. Microbial degration of toxic industrial chemicals in freshwater hydrosoil of streams, lakes, and reservoirs. 38. Development of laboratory techniques to study the effects of toxic industrial and agricultural chemical contamination on microbial geographical cycle in streams, lakes, and reservoirs. 39. Disturbance of aquatic geochemical cycles by inorganic contaminants resulting from energy development in western United States. 40. Identification and determination of the impact in fish of metabolic products or organophosphate esters, replacements for PCBs used in hydraulic fluids. 41. Interpret the identity and impact of phenolic degradation products stemming from wood preservatives and industrial processes in eastern and southeastern rivers: laboratory metabolism and toxicity studies on important lower Mississippi River fishes. 42. Methods development and validation required to analyze samples of chronically and acutely exposed fathead minnows, brook trout and fish food organisms: emphasis on phenols (drilling fluids and wood preservatives), organophosphate insecticides, and industrial chemicals. 32 43. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: methods for fractionation of heterogenous chemical complexes such as phenols, polynuclear aromatic hydrocarbons, and heterocyclic compounds in water, sediments, and fish. 44. Development of analytical methodology for measurement (in water, fish, and fish food organisms) of heterocyclic and other herbicides used in intensive agricultural practices. 45. Methodology survey and development for analysis and assessment of polynuclear aromatics and other toxic compounds in water, sediment, and fish in intermountain streams, related to energy development. 46. Development and application of innovative materials for separating trace organics isolated from aquatic organisms: emphasis on herbicides, phenols, and polynuclear aromatics. 47. Analytical residue support for laboratory toxicity studies designed to evaluate the impact of environmental contaminants, including pentachlorophenol (wood preservative), phosphate esters (PCB substi- tutes), phthalate esters (industrial plasticizers) , organophosphate insecticides, and carbamate pesticides, on aquatic resources. 48. Analytical residue support for agencies involved in the assessment and mitigation of pollution and fish and wildlife kills. 49. Trends in levels of contaminants in fisheries of the Nation's major lakes and streams: National Pesticide Monitoring Program. 50. Effects of samples handing and storage conditions on NPMP samples: Shelf life of selected inorganic and organic contaminant residues in fish. 51. Sampling and analysis of fish tissues for toxic substances. 52. The ecological effects of contaminants such as PCB replacements on stream communities as determined under controlled conditions in artificial streams. 53. Disturbance of aquatic geochemical cycles by PCB and PCB replacements in streams, lakes, and reservoirs. 54. Chronic toxicity of PCB and PCB replacements on fish and fish food organisms in the Upper Mississippi River. 55. Effect of PCB replacements on the behavior of Northwestern and Alaskan salmonids and anadromous fish. 56. Impact of PCB and PCB replacements on aquatic snails (e.g., Campeloma sp.). 57. Identification and assessment of metabolic degradation products of PCBs in Great Lake salmonids and panfish of the Upper Mississippi River. 33 58. Development of analytical methodology for toxicity evaluation of PCB substitutes: emphasis on organophosphate and phthalate esters and chlorinated benzene dielectrics. 59. Determine accumulation and distribution of structurally related ortho, ortho* -chlorine PCBs in biota and other aquatic compartments of the Upper Mississippi River and organisms from related laboratory studies for model development. 60. Identification and analysis of chlorinated dibenzofurans (highly toxic contaminants present in PCBs and chlorinated phenols) in fish collected from NPMP and field survey investigations. 61. Dynamic transport and deposition of polychlorinated biphenyls in fish and waterfowl habitat of the Upper Mississippi River. 62. Relative efficiency of important fish and diving duck food organisms to bioaccumulate PCBs, and their biological effects. 63. Biological availability of PCBs associated with the habitat of prime food organisms of fish and waterfowl of the Upper Mississippi River. 64. Physical-chemical aspects of bioconcentration, including aqueous partitioning, that affect the availability of PCBs. 65 . Analytical support for research in the Nation's watersheds to study impact of acid-metal precipitation, changing land-use practices, irrigation return waters, and energy development on aquatic resources. 66. Preliminary studies to evaluate amino acid levels and incidence of neoplasms as indicators of contaminant stress on aquatic biota. 67. Identification of detected but unknown toxic organic contamnants in biota from NPMP. 68. Collation and assessment of field information and formulation of research design to investigate suspected and future contaminants on critical aquatic resources. 69. Evaluation of occurrence of bilateral asymmetry in yellow perch as a new method for indicating aquatic contamination. 70. Sublethal effects of Sevin-4 oil and othe forest insecticides on natural populations of salmonids in Maine streams. 71. Impact of acid rain and metals from stack and auto emissions on brook trout, Atlantic salmon, and other fishes of the Northeast. 72. Impact of increased agricultural chemical usage on aquatic environ- ments following the removal of bottomland hardwood forests in the lower Mississippi Alluvial Rivers. 34 73. Effects of contaminated irrigation return flows on striped bass and other anadromous and resident fish popultions of the Lower San Joaquin and Sacramento rivers. 74. Effects of brine discharges from petroleum wells on faunal composi- tion of fish and fish food organisms in Oklahoma streams. 75. Effects of forest management herbicides on anadromous salmonids and fish food organisms in streams of the Pacific Northwest. 76. Effects of removal of riparian vegetation and increased sediment load on coho salmon and other anadromous salmonid populations in the Pacific Northwest. 77. The role of suspended sediment in determining bioavailability, toxicity, and longevity of some important agricultural herbicides in aquatic environments of the Missouri River System. 78. Effects of copper-diquat and other agricultural chemicals used in the control of aquatic weeds on the decline in apple snail populations of the Loxahatchee National Wildlife Refuge, Florida. 79. Occurrence of hepatomas in natural populations of fish and their relation to contaminant levels. 80. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: detection and identification of phenolic polynuclear aromatic and heterocyclic compounds by mass spectrometry. 81. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: expand methods for separation and analysis of unknown compounds by adaptation of highly efficient capillary chromatographic instruments. 82. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP Investigations: development of liquid chromatographic separation and detection methods for more water-soluble, relatively persistent pollutants not readily detected by gas chromatography. 83. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: uses of laboratory minicomputers for reduction of data from complex chromatographic separations and documentation of the occurrence of unknown and known contaminants. 84. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: develop analytical survey methods to measure (in fish, sediment, and water) inorganic pollutants from mine and particulate byproduct leachates derived from energy production. 35 85. Discovery of exotic and other unknown contaminants in biota from critical habitats and from NPMP investigations: development and implementation of sophisticated new mass spectrometry detection systems specific for phenols, and nitrogenous and other heterocyclic pollutants. 86. Biochemical analysis of intralaboratory samples from hazard assessment studies of toxic materials. 87. Biochemical analysis of samples from field monitoring studies for assessing contaminant-induced stresses on aquatic popultions. 36 APPENDIX C. ENVIRONMENTAL CONTAMINANTS EVALUATION - CURRENT WORK UNITS, AT SERVICE FACILITIES OTHER THAN THOSE AT ANN ARBOR, MI, AND COLUMBIA, MD Seattle National Fisheries Research Center, WA. 1. Impact of contaminants from mineral and energy development on critical habitat for Northwestern and Alaskan salmonid and anadromous fish populations. 2. Determination of critical habitat requirements of Alaskan and Arctic fishes under overwintering conditions and their tolerance to environ- mental alterations. 3. Preparation of a manual of physiological methods for developing improved HSI equations for the FWS aquatic Habitat Evaluation Procedure for project impact evaluation. 4. Ecological impact of forestry practices on anadromous fish populations in southeastern Alaska. 5. Environmental tolerance levels of gas supersaturation in marine and alpine ecosystems. Southeastern Fish Cultural Laboratory, Marion, AL. 1. Acute, chronic, and subchronic toxicity of organic contaminants to striped bass and other priority warmwater fishes and forage. Cooperative Fishery Research Units 1. Georgia - Scanning electron microscope observations of gills of weakfish (Cynoscion regalis) and spot. 2. Georgia - Field acclimatization and laboratory acclimation of bluegills (Lepomis machrochirus) to pH. 3. Georgia - Adenylate energy charge as an indicator of stress in fishes. 4. Georgia - Changes in gill-tissue morphology induced by exposure to an anionic detergent. 5. Maine - Effects of Sevin applied to forests on food niche and growth of brook trout fry. 6. - Maine — RNA-DNA ratios as indicators of growth effects from insecticide exposure. 37 7. Maine - The ecological effects of log salvaging from reservoirs. 8. Maine - The ecological effects of log driving in relation to navigable rivers. 9. Maine - Population studies of intertidal fishes of the coast of Maine. 10. Ohio - Organochlorine contaminant levels in edible and inedible portions of selected Lake Erie fishes. 11. Ohio - Bilateral asymmetry in freshwater fish as an index of environmental contamination. 12. Oregon - Imprinting in salmon: Odorant recognition, effects of pollutants and artificial cues in saltwater. 13. Oregon - Chronic turbidity stress in juvenile coho salmon and steelhead trout. 38 Table 1. — Allantoinase activity of individual samples of frozen livers from wild lake trout collected in Lake Michigan off Saugatuck and Lake Superior off Isle Royale in 1974, and from Laboratory stocks of lake trout (+ SE in parentheses). (From Passino and Cotant, 1979.) Source , and number of fish Total length (mm) Group Mean Allantoinase specific activity (unit/mg protein) Lake Michigan 19 >500 707 (14) 0.0701a (0.00755) 18 <500 310 (26) 0.140a (0.0109) Lake Superior 18 >500 592 (10) 0.0712 (0.0103) 32 <500 350 (11) 0.0749 (0.00869) Laboratory^ 3 >500 564 (6) 0.109 (0.0127) 30 <500 290 (14) 0.197 (0.0125) aDifference between means of allantoinase activity within fish length group is highly significant (P<0.01) by analysis of variance. "Fish maintained at Great Lakes Fishery Laboratory 39 Table 2. — Contaminant analysis of composite samples of livers and individual samples of fillets of lake trout collected off Saugatuck in Lake Michigan and off Isle Royale in Lake Superior in 1974. Total length (mm) Livers Fillets PCB's DDT+DDE Total Hg Total Hg Source, and number of fish Group Mean + SE (yg/g)a (yg/g)a (yg/g)ab i (yg/g)ac Lake Michigan 10 >500 717 ± 23 21.4 4.68 0.939 0.594 10 <500 210 ± 14 <0.6 0.141 0.479 0.0552 Lake Superior 10 >500 586 ± 14 4.55 1.86 0.690 0.581 8 <500 284 ± 4 <0.6 0.206 0.105 0.152 ayg/g wet weight. " Average value of three (fish <500 mm long) or four ( >500 mm) analyses per composite sample. c Average value of two analyses per fillet for specified number of fish. 40 Glycogen Granules Endoplasmic Reticulum Microbodies Golgi Complex Lysosomes Ribosomes Nucleus Mitochondria Plasma Membrane Cytosol Figure! 1. Liver cell of the rat 41 9 •§.■= IE o2 zj w E ■ £ ■ -9 CO CO Q_ U E E CD ^_ > CD ZJ > 2 2 LL LL m y co 9S op 03 3 O CO CD CL *a o CO g c-l_ Q. u DC EN L|O0 . CO O CD 5 E DC O CO DC LU DC i CO E£ CD 'q. CO 03 CJ Tl rti C > o _1_ a CO D CD O 3 z o 00 o o CO o IX) o o 00 o CM o o 8iun|0A Aq Bh }U80J8d _CD ~CD c CT3 CD o 0 a) x «- O •H £ S fr CQ -rH fa c •H ■P rt 4-) co 0 (0 §• C jf U a o u I 4-> c d) o a) 04 o •H +J CD M O •H E CQ (0 H ft a) 0 « ft C 3 42 STIMULI External and Internal ♦ BRAIN C» •HYPOTHALAMUS \ >' \ anterior pituitary acthI \ Primary Effects Secondary Effects INTERRENAL CELLS CORTICOSTEROIDS CHROMAFFIN CELLS CATECHOLAMINES "LT WHITE BLOOD CELLS \ LIVER GLYCOGENS ELECTROLYTE MUSCLE PROTEIN\ IMMUNOSUPPRESSION BALAN cEeA _t~l JL i_r BLOOD GLUCOSE / BLOOD LACTATE/ PLASMA FFA/\ J~l HEART RATE' GILL BLOOD FLOW^ WATER CONTENT Freshwater^ Seawaler^ diuresis/ drinking rate/ melanocytes\ Figure 3. Scheme incorporating current information on the interrelationship between primary and secondary effects of stress in fish. (From Mazeud et al. 1977.) 43 OVERVIEW AND ASSESSMENT OF BIOLOGICAL SENSING by John D. Koutsandreas Office of Monitoring and Technical Support Office of Research and Development U.S. Environmental Protection Agency Washington, D.C. 20460 INTRODUCTION Biological in-situ sensors used to monitor indigenous terrestrial and aquatic biota could replace some of the more time-consuming and costly sampling and laboratory analytical techniques. Selective bioindicators and bioaccumulators are being utilized to identify potential exposure hazards before significant damage to the environment or human health occurs. Biological monitors can be used to identify toxic substances and to assess total environmental exposure to hazardous materials which have been originated for various sites. Terres- trial and aquatic organisms can function as in-situ monitors and remove much of the requirement for time-consuming and costly conventional monitoring tech- niques. Essentially, six approaches to biomonitoring are being evaluated at EPA. 1. In-situ bioassys Water, land, air 2. Bioaccumulation Aquatic organisms, terrestrial organisms 3. Indicator organisms Aquatic organisms, terrestrial organisms 4. Community and population changes (Species composition and abundance, biomass, diversity, productivity, mortality rates, etc.) Aquatic organisms, terrestrial organisms 5. Species health Tumor development, phenotypes, growth, reproduction, longevity, and stress measurements 6. Integration of biological methods to determine release of toxic substances from hazardous waste operations for a. Air, b. Water, and c . Land . 45 Biological monitoring is neither a new concept nor a new activity. In fact, it was associated with the origins of current environmental concerns, that is, disappearance of desirable game fish from polluted rivers, accumulation of nuisance algae in eutrophied lakes, fish kills from pesticide runoff, and plant kills near smelters. Recently, however, biological monitoring has received relatively less attention. This is for several reasons. One is the striking technical advances in our capa- bility to detect and measure, by physical and chemical means, small amounts of contaminants in the environment. Another is the emphasis of the environmental protection laws enacted in the early 1970s on preventing the entry of pollutants into the environment without a need for direct proof of environmental damage. Today we see a renewed interest in biological monitoring, which is the only means by which we can directly measure the degree of harm to man or an ecosystem resulting from environmental contamination or, conversely, the benefits from environmental protection. The purpose of this paper is to examine biological monitoring in the context of current Agency needs - assessing the strengths and weaknesses of available biological monitoring techniques and attempting to define where opportunities do and do not lie for more effective and fruitful use of biological monitoring by the Agency. WHAT IS BIOLOGICAL MONITORING? The term "biological monitoring" can be applied to a rather large and diverse set of activities. It does not have a precise, generally accepted meaning; it means many things to many people. For this paper, it will be defined in terms of biological monitoring. This may be defined as the use of a living organism to measure, for a specific time and location, the presence or amount of a sub- stance in the environment or in man, or the effect of a substance on the environ- ment or man. The keys to this definition are fourfold. 1. Use of a living organism 2. Specificity of time and location (to distinguish "monitoring" from "bioassay," such as the laboratory use of a fish or rat to determine an exposure/effect relationship) 3. Presence, amount, or effect in or on the environment 4. Relationship to a substance (Biological monitoring could be used to measure the environmental effect of a perturbation other than one resulting from entry of substance into the environment, e.g., destruc- tion of fish habitat by stream channel modification. However, for this paper, the term is limited to effects resulting from a substance.) There are three general approaches that can be taken to biological monitoring. 46 1. Use of an organism in the environment This involves measurement with an organism in the environment to determine a presence, amount, or effect. The organism can be either naturally present or introduced. a. Naturally present or indigenous organism (so-called "passive" monitoring) Examples of this type of monitoring include human epidemiological studies, surveys of stream biota, and pesticides residue monitoring. Two general approaches exist for this kind of monitoring. (1) Limited-species monitoring, where data are collected only from a single or several species. In this case, the organism selected is 1) the species to be protected, e.g. salmon, or is assumed either to be 2) particularly sensitive to the effect being monitored (e.g. eggshell breakage in ospreys due to DDT) or 3) to accumulate the contaminant of concern (e.g. radioactive iodine in animal thyroid tissue). In the latter two instances, the organism would be called and "indicator" or "sentinel" species. (2) Community monitoring, wherein some feature of a biological community is measured such as its structure (e.g. which species are present in what numbers, species diversity, etc.) or its function (e.g. rates of energy flow, organic decomposition, biomass production, etc.). b. Introduced species (so-called "active" monitoring) In this case, the organism is intentionally introduced into the environ- ment to take advantage of its sensitivity to an effect of concern or its ability to accumulate a contaminant. The classic example is the miner's "canary in a cage." More appropriate, as an example, is the use by the Germans of plants to detect air-pollution effects or accumulate heavy metals from air. 2. Use of an organism in the laboratory to measure a sample from the environment Here the organism is kept in the laboratory, but the sample is collected from the environment. Examples include the use of fish bioassays to determine the toxicity of effluent samples and the recent attempts to use the Ames test to trace presumed carcinogens from industrial sources through the environment. It must be carefully noted here that "bioassays" are a biological monitoring technique only when they are applied to an environmental sample, and not when used to determine an exposure/effect relationship (e.g., LD50) for a specific chemical. 3. Use of a simulated environment ("microcosm") to measure either an effect or the movement and/or transformation of a contaminant "Microcosms" are intentionally created, controlled assemblages of species of varying complexity and size (currently ranging from a jar that could be held in one's hand to very large "bags" suspended in the ocean) that are assumed to 47 behave like natural ecosystems - at least for the functions being measured. Their unique advantage is their presumed capability to measure community (i.e., multispecies) phenomena under controlled, replicable conditions. ASSESSMENT OF CURRENT BIOLOGICAL MONITORING TECHNIQUES As mentioned above, there are many purposes for biological monitoring and many methods which have been used or are under development. The following is an attempt to describe the major types of biological monitoring and to assess their existing utility and scientific credibility and to identify what changes in biological monitoring capabilities are likely to occur in the near future. 1. Measurement of an effect using an organism in the environment, either indigenous or introduced a. Examples • Stream surveys which measure population size, species diversity, etc. • Use of plant species, either indigenous or introduced, to measure effects of gaseous air pollutants • Human epidemiological studies b. Types © Single or limited species that use indicator, sentinel, or sensitive species, either indigenous or introduced 0 Community or ecosystem structure or function c. Advantages The only way to measure directly a biological effect in the real environ- ment. Such methods are most closely related to the real world, thus providing the most meaningful of all biological monitoring data. d. Disadvantages • Biological populations normally fluctuate widely and are affected by a large number of both natural and anthropogenic environmental factors. Thus, they are a rather insensitive measure of environ- mental contamination, reflecting only rather gross and fairly obvious changes. To detect changes, vast amounts of data may be required. Moreover, measurable populations are usually slow to respond to per- turbation, often revealing change only when the damage is already extensive. • It can be very hard to define the cause of a biological change without considerable supporting data on exposures of environmental toxicants and other biological parameters. For example, it may be hard to 48 relate a decline in the size of a population to a pollutant, rather than to a change in another environmental factor such as a change in climate. Changes in community or ecosystem function can be measured, but the ecological significance of relatively small changes are not well understood. Certainly, the public and government officials have not been presented a persuasive case for the significance to human welfare of small changes in ecosystem functions, such as energy cycling, • Indicator or sentinel species present special problems. Their response can vary considerably with the age, sex, and genetic background of the organism. Therefore, the advantage of an indicator organism, namely collection of data on a single species rather than the whole system, can be illusory; considerable data may be needed from the system to interpret the data from the indicator species. In order to choose an effective indicator species, the chemical of concern must be known and, often, its behavior and the nature of its effect on the environ- ment as well. Assessment General biological surveys which measure population sizes or other aspects of community structure or functions are usually too data- intensive to be useful for purposes other than determining the gross state of health of an ecosystem. Cost-effective examples of the type of survey are hard to find. For most cases which have been examined, the major deficiency has been the lack of a clearly defined objective - either a testable hypothesis or a specific range-finding goal to guide data collection. Even when the survey has a clear objective, the amount of data required to reveal a measurable change of signifi- cance may be impracticably large. Use of biological effects data for general trend monitoring seems to be even more impractical at this time. Without enormous amounts of data most trend monitoring nets are too coarse to detect effects at a useful level or in a useful time frame. If biological effects are to be used for trend monitoring, this should be done only for very specific and carefully targeted purposes. The system must be very carefully designed and evaluated for cost-effectiveness ahead of time. The most promising and underutilized approach to effects-monitoring appears to be active monitoring, using introduced species. This approach offers a high degree of sensitivity, reliability, and precision of data. Since many of the variables associated with the organism itself are controlled, the effects measured can be more easily attributed to environmental factors. A certain amount of accompanying data on the ecosystem are necessary to interpret changes occurring in introduced species, but the amount of data may be considerably less than necessary for indigenous species. Use of an appropriately sensitive introduced organism could provide a real time detection or alarm system for environ- mental chemicals, as well as a relatively reliable means for measuring change in environmental stress. 49 Active monitoring or use of introduced species has not been widely used in this country. The Germans and Japanese appear to have developed a number of such systems. The basic scientific capability exists for this approach to monitoring and, with an intensive effort, such systems could be developed and validated. Most effects-monitoring has involved acute toxicological effects on a sentinel organism or actual changes in numbers within a population. However, several recent approaches offer promise of measuring effects prior to actual death of the organism. These are (1) measurement in the laboratory of sublethal changes, such as physiological responses or states, or the behavior of organisms taken from the environment and (2) use of plant mutation rates to detect the pressure of genetically active chemicals in the environment. 2. Measurement of the presence or amount of a contaminant in the environment or in an organism a. Examples ® Measurement of pesticide residues in fish, shellfish, birds, and human tissue © Monitoring of radionuclides in the human food chain (e.g., cow's milk) or in accumulator organisms or tissues (e.g., 131j in animal thyroid tissue) • Use of plants in Germany to indicate the presence of heavy metals in air • The U.S. Mussel Watch program © Measurement of metals (e.g. , lead) in human hair as an indication of exposure (e.g., use of Isotope Zeeman Atomic Absorption Spec- trometer) b. Types Measurement in a human food chain organism to indicate the existence and potential magnitude of human exposure to a chemical Measurement in an accumulator organism (which need not be in the food chain) to detect the presence of a chemical in the environ- ment, and possibly, to measure its environmental level Measurement in a sampler or integrator organism (i.e., one which does not bioaccumulate the chemical of concern, to detect the presence of a chemical in the environment, and possibly, to measure its environmental level) Measurement of the chemical within the affected organism as an indication of actual exposure 50 e Specimen banking, which could involve any of the above, but with storage of the sample for deferred analysis (This allows retro- spective analysis for chemicals which were not of concern at the time the sample was collected, but have become of concern subse- quently. ) c. Advantages • When bioaccumualtion occurs, this approach may be more sensitive than physical/chemical methods. Essentially, the organism is used as an efficient "trappng" system. • Since the actual presence of chemical in an organism is measured, these data provide greater significance as to biological relevance. In some cases, such data can provide a reasonable measure of either the potential exposure in the food chain or the actual amount of chemical reaching and being taken up by the organism. d. Disadvantages 9 One usually must know ahead of time which chemical to look for (except for specimen banking) in order to select the appropriate organism for sampling. This is because the uptake or accumulation may vary considerably according to the chemical or the organism involved. We are not able at this time to define sentinels for a broad spectrum of chemicals. • Data on indigenous species can be highly variable, due to different responses with age, sex, and genetic differences. • Trend monitoring nets are usually either too coarse or too costly. For example, market-basket surveys have not been especially effective in detecting agricultural chemicals in food, since most exposures of concern are due to spills, inadvertant misuse, or other intermittent or local incidents. However, trend monitoring can be useful for meas- uring gross environmental changes, such as the widespread decline in amounts of DDT present in fish or shellfish, or for highly focused, local surveys. • Methods for reliability sampling and analyzing chemicals in living tissue are not generally available. e. Assessment Use of living organisms to measure chemicals in the environment or actually within an organism of concern (e.g., human tissue) is currently feasible. The keys to success are (1) knowing the chemical of concern and enough about its behavior and effects to be able to select an appropriate organism for sampling, (2) controllng the variability within the organism itself (using a standard, introduced organism is a very promising way to achieve this), (3) allowing for the necessary, and sometimes tedious refinement and 51 validation of chemical analytic methods, and (4) clearly defining a real- istic objective, i.e. , selecting changes of concern that can be detected within the time and sensitivity required by a system of reasonable cost. If designed properly, such systems can be very effective, for specific chemicals, as chemical detectors or measures of potential or actual exposure, Use of such systems for measuring trends over a broad area or for many substances is most likely to be impractical except for rather limited purposes such as the measuring changes in general environmental levels of pervasive chemicals such as DDT or PCBs. The two most promising developments with this approach to biological monitoring appear to be (1) use of introduced species and (2) measurement of actual amounts of chemicals in human tissue as an indication or even quantitative measure of actual exposure. 3. Laboratory analysis of environmental samples a. Types • Use of toxicity or another general biological parameter as a perform- ance requirement for a pollution control system, e.g. , inclusion of toxicity as measured by fish bioassays as a performance requirement for wastewater treatment systems. • Use of a biological-effects parameter, such as a genetic effect, as an indication of contamination of the ambient environment or a means for tracing contributing sources. b. Advantages • Use of a single, general parameter can be considerably cheaper than chemical analysis since it does not require measurement of a possibly large number of chemicals. • This approach can detect effects of unknown chemicals as well as synergism or antagonism among several chemicals. • The data are highly precise, reliable, and sensitive. In addition, bioassay data have historical credibility for regulatory activities. c. Disadvantages • It can be very hard to relate cause and effect. If the toxicity performance limit is exceeded, further data are needed to define the causal agent. • Each technique measures only one stress, for instance acute toxicity. Several assays may be needed to cover the major stresses of concern. • Such an approach does not normally account for bioaccumulation. 52 d. Assessment General toxicity, such as measured by a fish bioassay, has considerable promise for use now as a regulatory tool. It can be reliably and con- veniently done, and can be considerably cheaper than chemical analysis. It has certain limitations: the causal factor is not defined, bioaccumu- lation is not considered, and reflects only the single stress of toxicity. However, it may provide a reasonable and practical surrogate for a broader spectrum of effects. 4. Measurement of fate and/or effect in laboratory ecosystems - microcosms Microcosms, or controlled ecosystems, is a term which applies a rather wide range of experimental approaches to reducing the enormous number of varia- bles in the real environment while preserving essential interactions among components of a real environmental system. The microcosm need not simulate the real environment in all aspects, but only those of concern for the experiment. Its advantage is the ability to simulate systems effects or processes and not just those of a single species. Microcosms vary from rather small, multispecies assemblages in a jar that could be held in one's hand to rather large attempts to capture and control segments of the environment, such as an artificial stream channel or the suspension of a massive, semienclosed bag in the ocean. Because microcosms vary so widely in nature, purpose, size, and cost, it is hard to speak generally of their current or future possibilities. None ap- pear to have reached an operational stage of utility; thus, for now, they remain research techniques. However, they offer considerable promise as a means for measuring both environmental effects and environmental processes, such as transport and fate. Their ability to simulate, usefully, the real environment needs to be determined as soon as possible. CONCLUSIONS As pointed out earlier, biological monitoring provides a nearly-unique advan- tage in the measurement of real effects on the living environment, which has led recently to a renewed interest in this technique. The search for real or meaningful data has led most past efforts at biological monitoring in this coun- try to concentrate on measurements using indigenous organisms. These efforts have not been notably cost-effective. Because of the large natural fluctuations in environmental systems and the large number of variables affecting these fluctuations, such approaches to biological monitoring, although the most mean- ingful, are imprecise, insensitive and very hard to interpret. One way to get around the complexity and variability of natural systems is to monitor so-called indicator species. However, the advantage of simplicity in using indicators can be illusory. Often the overall environmental system has to be understood in order to select an appropriate indicator organism or to inter- pret the data resulting from its use. Thus the direct monitoring of effects using indigenous organisms is severely limited in practicability and utility. This is not to say that such activities 53 are totally impractical or useful. They have a definite place for measuring gross environmental conditions or changes or for intensive, targeted (and usu- ally costly) local surveys. They appear not to be promising for widespread, routine use at this time. This is especially true for broad environmental-effects screening to identify problems otherwise missed or trends that allow us to anticipate problems. Because of the variability of natural systems, the data needed to detect effects of changes in environmental systems with a useful degree of resolution would be enormous, and probably not worth the cost of collecting. Are there any opportunities for more effective use of biological monitoring for effects? The major promise appears to be with the use of introduced organisms - so-callled active monitoring. This approach appears to be used effectively by the Japanese and Germans. Use of introduced organisms can be less appealing scientifically, since it is a way of getting around having to understand natural systems, and it has some disadvantages in interpreting the meaning of the data. Nonetheless, it offers real promise for detecting and measuring environmental effects of chemicals and should be aggressively analyzed and, if need be, developed for use in this country. Biological monitoring for the presence or level of a chemical in the environment offers great promise for specific chemicals. Its successful use in the past has always, apparently, been associated with a specific chemical of concern: a radio- nuclide, a pesticide, or a heavy metal transported by air. As described above, its advantages are (1) greater biological relevance, (2) greater sensitivity, and (3) ability to indicate actual or potential exposures. Techniques either are available, under development, or can be relatively easily developed. As the Agency moves towards regulation of specific chemicals, this biological monitoring technique should prove very useful and cost effective. Again, the key is to moni- tor for a specific, known chemical. The possibilties are numerous: a detector, an alarm, a measure of food-chain levels, a measure of actual exposure. The use of a general toxicity measure as a performance requirement for environ- mental control systems, such as industrial wastewater treatment, seems to offer sufficient promise that it is puzzling why it has not found broad favor. Reluc- tance appears to result from concern that a general toxicity measure, such as a fish bioassay, does not identify the causal agent and does not detect certain effects, such as carcinogenicity. Nonetheless, this approach seems to warrant more serious consideration, expecially if linked in the future with other bio- assay techniques, such as those for genetic effects. One anomaly became apparent during this analysis. So far, most environmental attention appears to have been given to monitoring of aquatic systems, except for some work on pesticides and on terrestrial effects of air pollution. It seems as we move more to a concern over human health effects from environmental contaminants and protection of our environmental life support system that rela- tively greater attention is warranted for terrestrial systems, including soils. Our efforts on biological monitoring should reflect this shift in emphasis, particularly for future monitoring of hazardous waste sites. 54 NEW IN-SITU TECHNIQUES IN BIOMONITORING by Dr. Harold E. Schlichting, Jr. BioControl Company, Inc. Port Sanilac, Michigan 48469 and The University of Minnesota Forestry and Biological Station Lake Itasca, Minnesota 56460 INTRODUCTION Continuous in-situ biological monitoring is essential to the understanding and improvement of water quality because these systems continually assess the responses of selected indicator organisms to the changes occurring in the water chemistry and physical parameters and are much less expensive than ecological surveys. DETERMINING WATER QUALITY Successful predictions of water quality changes can be determined with bio- monitors in-situ depending upon the selection of indicator organisms, membrane filter selectivity, and the measurement of pertinent physical and chemical data at the test sites. Biomass increase and the health of the organisms within the biomonitors in-situ over 3- to 30-day periods reflect the degree of eutro- phication and the presence of toxic compounds such as pesticides, herbicides, heavy metals, and radionuclides. Bioconcentration of these substances within test organisms such as Hydrodictyon, Pithophora, Rhizoclonium and cultures from the Unialgal Pollution Indicator Set (Carolina Biological Supply Company 15- 1550) indicates the toxicity of these substances to the test organisms in-situ and their relative abundance in the water. Species diversity studies using mixed populations can be conducted with bio- monitors in-situ. Biomonitors are also ideally suited for studying possible morphological and physiological changes in species under different environmental conditions in nature and in the laboratory. Equilibration time, i.e., when the water chemistry within the biomonitor approximates that of the surrounding water, requires 24 to 48 hours, varying with the water quality. Procedures for using biomonitors (sterilization, types of media, test orgnisms, equilibration time, and bioassay techniques have been previously described (Schlichting, 1976). DILUTE MEDIA DUPLICATES NATURE Trainor and Shubert (1973) suggested that algae to be used in bioassay pro- cedures should be grown in a dilute medium which more closely duplicates con- ditions existing in nature. Daily transfers of algal cells to fresh dilute media supported good growth in their laboratory studies. Growing algae in dilute media within a biomonitor situated in the precise water area under 55 investigation for the time period desired eliminates the need for daily trans- fers. The nutrients and metabolic products are recycled by diffusion processes and mixing occurs from chamber agitation by the currents as well as biological pumping within the biomonitor. The composition of most defined media generally does not resemble that of natural waters and can even bring about morphological changes in the organisms under study. It is recommended that test site water, sterile filtered with 0.2 urn porosity filters, be used to prepare dilute media for culturing test organisms and filling the biomonitors. Dilute media, more closely approximating the mean chemical composition of river water in parts per million, would be: carbonate 58.4, calcium 15, silicon 13.1, sulfate 11.2, chloride 7.8, sodium 6.3, magnesium 4.1, potassium 2.3, nitrate 1, and iron 0.67 for a total of 120 ppm (Livingston 1963). PREDICTIVE VALUES USING ALGAE AS INDICATORS Algal growth potential (AGP) is defined as the organic mass resulting from a 7-day incubation of a culture grown in a natural water substrate under standar- dized laboratory conditions and expressed as mg of dry organic mass per liter of sample (Wang et al. 1973). A similar definition could apply to the growth of microalgae in biominitors in-situ for the same unit of time under various diurnal and seasonal environmental conditions. Thus biomonitor studies of a body of water over one year's time would give a good indication of the AGP for that given body of water and should have high predictive value for discerning future water quality changes. By growing algae in-situ within a biomonitor, grazing is eliminated and the organisms can respond to all the physical and chemical factors at the test site. The degree of growth and health of the algae per unit time will indicate water qualtiy (Mason 1978). The comparison of the growth of Anabaena within biomonitors placed one foot below the surface in Lake Itasca and in a sewage lagoon showed that growth was nearly twice as great in the lake as in the lagoon over a two-week period. (See table 1.) With higher nutrient content, alkalinity and total hardness in the lagoon, a higher growth rate would be anticipated. However, a greater turbidity and a mat covering the lagoon sur- face reduced the available light actually retarding growth. The results of the study demonstrated the importance of recording both chemical and physical parameters to better understand the experimental results. On the basis of this study, it is recommended that the following additional data be recorded in Biomonitor studies: ammonia nitrogen, specific conductiv- ity, total dissolved solids, total suspended solids, nitrite nitrogen, light penetration, and meteorological conditions (Mason, 1978). STUDIES USING THE BIOMONITOR Biomonitors are suitable for many other types of studies. Radionuclide bio- concentration in the aquatic environment as well as the degree of toxicity of sewage effluent, heavy metals, pesticides, herbicides, growth regulators, and 56 Table 1. — Growth of Anabaena in situ over a two-week period Parameters Lake Itasca Sewage Lagoon Increase in biomass 298 mg 121 mg (Inoculum 82 mg) Temperature range 20.5-24.5°C 19.8-26.5°C Nitrogen range 0.02-0.04 mg/1 0.06-0.16 mg/1 Orthophosphate 0.25-0.30 mg/1 5.50-9.60 mg/1 pH 9.2 9.1 Alkalainity range 165-168 mg/1 CaC03 343-366 mg/1 CaC03 Total hardness 141-143 282-285 Turbidity range 3-5 10-27 agricultural fertilizers to the indicator organisms in strip mine ponds, lakes, and rivers can readily be studied. These instruments are also suitable for environmental impact studies leading to the reclamation of waterways for public use. Species diversity studies of mixed plankton populations utilizing the Palmer Index (James 1975) and studies of possible morphological and physio- logical changes in species under different environmental conditions can also be determined by biomonitor in-situ studies. The incubation of phytoplankton in situ in the sea or lakes, although inconvenient, should be done more frequently (Fogg 1965). The great variety of studies in which biomonitors may be utilized to bio- logically monitor "nature's variables" is limited only the ingenuity of the researcher. 57 REFERENCES Fogg, G. E. 1965. Algal Cultures and Phytoplankton Ecology. U. Wise. Press. Madison James, D.E. 1975. Algae: Pollution Indicators. Carolina Tips 38(15): 57-58 Livingston, D. A. 1963. A Geological Survey Professional Paper 440-G, U.S. Govt, Printing Office. Washington, D.C. Mason, C. P. 1978. Unialgal Growth of Anabaena and Dictyosphaerium in Lake Itasca. J. Minnesota Acad. Sci. 44(2): 12-13. Schlichting, Jr., H. E. 1976. Biomonitor Laboratory Exercises. Carolina Tips 39(4): 13-14. Trainor, F. R. and Shubert, L. E. 1973. Culturing Algae in Dilute Media. Phycologia 12(1-2): 35-39. Wang, W. , Sullivan, W. T. , and Evans, R. L. 1973. Report of Investigation 72. 111. State Water Survey Div. , Urbana, 111. 58 PROCESSING OF ZOOPLANKTON SAMPLES BY ELECTRONIC IMAGE ANALYSIS1 by Dr. H. Perry Jeffries Graduate School of Oceanography University of Rhode Island Kingston, Rhode Island 02881 Dr. Kenneth Sherman and Ray Mauer National Marine Fisheries Service, Northeast Fisheries Center Narragansett, Rhode Island 02882 and Dr. A. D. Poularikas, C. Katsinis, and I. Melas Department of Electrical Engineering University of Rhode Island Kingston, Rhode Island 02881 INTRODUCTION Methods of identifying and counting zooplankton in preserved samples have changed little since Johannes Muller first towed a fine-mesh net through the ocean more than 100 years ago. Instrumentation basically consists of a microscope and a ruled counting chamber. Because time of analysis for a single sample ranges from hours to days, backlogging is inevitable. We are attempting to modernize the procedure, using improved image formation devices and computerized pattern recognition techniques. Thus far, total counts and size-frequency distributions can now be made in minutes with a simple, processor-controlled vidicon system, but calibration remains a problem awaiting sharpened contrast by optical edge enhancement and spatial filtering. Major representatives of North Atlantic zooplankton can be accurately classified to group (e.g. copepods, fish eggs, fish larvae, cladocerans, chaetognaths, and euphausiids) by discriminate analysis of simple morphometric relations (length, width, perimeter, and area). Discriminate analysis results of 93 well defined hand-traced zooplankton silhouettes resulted in 100% successful identification. When the same analysis was performed on 240 images of real zooplankton, the results, although not perfect, were extremely encouraging. 1. A technological bottleneck — We are solely dependent on manual operations for sorting of preserved zooplankton samples, even though physical fractionation procedures exist. A range of electronic approaches has been considered (Sutro 1974, Almeida et al. 1977), and particle discriminators work in limited ways (Beers 1976, Fawell 1976, Uhlmann, et al. 1978). But the task is so arduous iThis investigation was supported by NOAA/DOE Grants 04-8-M01-108 and NA80AA-H-00023. 59 that few technicians can work continuously at counting, identifying, and meas- uring organisms under a microscope. Personnel turnover is understandably high, and training periods further reduce sample output. Consequently, effective average processing time (per technician and year) ranges up to four days per sample. The resulting bottleneck means that sampling designs meeting modern-day sta- tistical requirements are impossible. The MARMAP program alone has a capability for collecting 4,500 samples annually (Sherman 1978) which is probably near the world capacity for sample processing. Clearly then, a technological break- through is required to support basic research in plankton biology and to meet the requirements of improved fisheries forecasting. We are addressing the problem with tools that have been developed in widely disparate fields. The reuslts are encouraging, but they do show the need for engineering advances in selected areas. Pattern recognition — Among the several approaches applicable to analysis of samples collected with zooplankton nets, we believe that pattern recognition by combined optical and electronic means is the most promising. In the sense used here, pattern recognition consists of counting particles, extracting visual features from individual organisms (either alive or preserved), measuring these features automatically, classifying the organisms into ecologically useful and taxonomically meaningful categories, and finally giving vital statistics of length, area, perimeter, etc., for each group. Classification to species is the ideal, and our work on morphometries indicates that it may be easier than one would imagine, but, for present purposes, we have established the following groups as technologically amenable to feature extraction and classification: • Copepods (copepodite stages I-VI) • Nauplii (various) • Zoeae (Brachyura) • Zoeae (Macrura) • Mysids, cumaceans, euphausids a Amphipods • Fish eggs • Fish larvae and chaetognaths • Lamellibranch larvae • Polychaete larvae • Medusae • Particulate debris, organic aggregate From total count and size-frequency distributions within these groups, useful descriptors of ecosystem state can be derived (Steele and Frost 1977). APPROACHES TO PATTERN RECOGNITION The components of a pattern recognition system suitable for preserved zoo- plankton samples consists of (1) a low power microscope or enlarging lens, with provisions for transmitted, incident and dark field illumination; (2) vidicon and input image monitor; (3) analog-to-digital converter; (4) minicomputer; and (5) output image monitor and line printer. (See fig. 1.) Figure 2 shows our 60 image processing system in more detail. The minicomputer is programmed to extract and measure specific characteristics. These features best characterize a large, but dimensional and finite, pattern space that has been reduced by the system's transducer (optical-vidicon system). The next step is a classification function that makes decisions for assignment of the input object to one of the predetermined classifications. (See fig. 1, Fu 1970.) 1. Mathematics of pattern recognition — The mathematics of pattern recognition vary from comparatively simple "template-matches" stored in the computer's memory to complex scene analyses consisting of so many features and patterns that the analysis must be described in hierarchical order from simpler subpat- terns. The decision-theoretic approach, which discriminates objects according to statistical models, has been most commonly used. In a statistical sense, "... each pattern is considered as a vector in n-dimensional space. The goal of the recognition system is to define partitions in this space such that each region can be identified with a class of pattern." (Viglione 1970). In the future, capabilties of the syntactic approach to description of structural pattern should become popular, especially for identifying biological objects (Fu 1978). Chromosome analysis, for example, has been automated by way of syntactic grammars. Our work to date has been limited to the discriminate- function approach, and although the classification ability for planktonic animals is surprisingly good, the two approaches will undoubtedly be combined in a semiautomatic instrument for classification, enumeration, and sizing of planktonic organisms. 2. Image formation by electronic means — In figure 3, the computer-generated outline of an adult female Centropages typicus viewed directly by vidicon is compared with a manmade drawing traced from the input monitor and then resub- mitted for computerized imaging. Note that edge definition of what man sees and interprets is far sharper than the computer generated pattern of an actual specimen. Herein lies the problem, much of it arising from the simple biologi- cal fact that low contrast makes planktonic animals hard to see. This is cer- tainly adaptive for organisms swimming in nature, but the problem it causes for us may be solved by such refinements as a laser beam directed at individual specimens and spatial filtering of the coherent image at the frequency plane. A second problem shown in fig. 3 (lower) is the orientation of dead specimens. Because their density is only slightly greater than neutral, they fall in unsta- ble positions. The left specimen appeared to have an asymmetric urosome, but this was not so; rather it was a visual aberration due to the copepod lying at about 45° on its major (length) axis, balanced tenuously on its plepods, which in this case happened to be extended. As shown in fig. 4, computer-generated images from silhouettes can be manipu- lated in various ways. Length, perimeter, width area, etc., of up to 5 organisms per field to view have been obtained. Figure 5 (top) shows the computer- generated image of a preserved crab larva and its smoothed verion superimposed on it. The lower part of the same figure shows the same zooplankton with smoothed silhouette. A copepod with its smoothed version is also shown. The smoothed perimeters have considerably less number of points and thus the time it takes the computer to extract the necessary features is shortened considerably. 61 To complement the above electrooptical approach, we have also explored applications of silhouette photography (Ortner et al. 1979). Images shown in fig. 6 may be better for input to currently available vidicons than are the organisms themselves, assuming, that a well focused image can be achieved. An advantage here is that this lensless system of direct photography by electronic flash can be done at sea with live organisms. Alive, they orient in rather repeatable ways from one individual to the next, but in a preserved sample random position is a problem that must be solved, possibly with a two-camera system trained on organisms hydrodynamically oriented in a flowing, as opposed to static, observation medium. RECOGNITION OF ZOOPLANKTON 1. Silhouette Identification of Zooplankton — To establish a reference data base free of electronic-imaging problems, we photocopied 93 published plates of different species, chiefly from Fishes Identification by Zooplankton of the Conseil International pour lf explortion de la mer. These species represented 5 of the 12 categories listed above. The groups which were represented included copepods, chaetognaths, euphausids, fish larvae, and cladocera. The variables used in the classification were perimeter (S^), square root of area (S2)> the maximum length (S3), and the maximum width (S4). Since one purpose of this work was to produce a classification process independent of the actual size of the images, 6 "normalized" variables were produced from the four originals. These were perimeter perimeter Xj = square root of area X2 = maximum length perimeter square root of an area X3 = maximum width X4 = maximum length square root of an area maximum length X5 = maximum width X5 = maximum width Figure 7 shows a typical clustering of the silhouettes on a two-dimensional feature space X2 - X4. The numbers in the field of this diagram refer to the previously defined taxonomic groups, e.g. , 1 = copepods. Although the clusters appear distinct, the cluster algorithm did not give accurate results. Therefore, we ran a discriminate analysis test (SAS program) using 93 organisms as a training test with a test set of 15 observations. The results showed perfect score by identifying all 15 animals. However, it must be pointed out that the perimeters of the animals were well and clearly defined with minimum additional electronic noise disturbing body outlines. In addition, the several silhouettes from the same species were similar in shape although they were different in size and orientation. 2. Identification of plankton — Two-hundred and forty animals of six groups (group: 1 - copepods, 2 - zoea, 4 - euphausid and mysid, 6 - fish eggs, 7 - fish 62 larvae, and 8 - pteropods-bivalve) were digitized and their edges were extracted. From the perimeters of these observations, additional features were extracted, which were (a) seven invariant moment functions, each one being a function of several moments; (b) the ratio of widths at one-third distance from the two edges of their longest dimension; and (c) circle similarity (a feature to dif- ferentiate circular animals, e.g., eggs, from any other group). From the 15 features, we selected 11. These were the 6 features which they were used to discriminate the silhouettes, three invariant moment functions, the 1/3 distant ratio and the circle similarity feature. Based on these 11 features, we ran discriminant analysis using SAS programs. Figure 8 shows a two-dimensional characteristic distribution of zooplankton on the feature space X2 - X4. Figure 9 shows the results of discriminant analysis. The training sample in this case was consisted of all 240 observations (animals). The large misclassi- fication error for groups 6 and 8 is due to the extreme similarity of the observations. Otherwise, the classification success shown in fig. 9 is very encouraging. Figure 10 shows results of discriminant analysis when the training sample was consisted of 200 observations (animals) and one test set of 40 observations. Figure 11 shows results on electronically smoothed body outlines which indi- cate that the success of classification is sensitive to smoothing. This is understandable, since smoothing tends to round long shapes and tends to elongate the round shapes. Additional studies are needed to create a more efficient algorithm such that the smoothed animals retain their identifying detail charac- teristics. Figure 12 shows mean misclassif ication error. These results are based on six different samples from the original and six from the smoothed animals. Each sample had one training set of 200 observations and one test set of 40 observa- tions. The mean misclassif ication error of the smoothed animals was somewhat higher, as was anticipated from the results mentioned above. In addition to the discrimination results, the automated analysis will provide far more needed information than human operators may provide within any reasona- ble time and effort. Vital statistics of area, perimeter, maximum lengths, and maximum widths for each group will be easily obtained. Figure 13 (upper part) shows the distribution of (square root of area) /(maximum width) and the lower part of the same figure shows the distribution of (maximum length) /(maximum width) of the group 6. Figure 14 shows the distribution of the first invariant moment of group 6. In this initial study, we have not kept the information about the microscope magnification used for each animal and thus any distribution of a single feature would have been meaningless. However, in our future studies this information will be included and distributions of the desired variables will be provided. 63 MEASUREMENTS WITH BAUSCH AND LOMB QMS SYSTEM Lengths of 300 Calanus f inmarchicus, copepodite stages IV-VI, were measured once manually and twice with the Bausch and Lomb QMS system (Maurer 1978). The machine consistently measured 0.1 mm lower than a technician through both runs, but upon adjustment for systematic error, the manual and electronically determined distributions, compared graphically in fig. 15, became statistically identical (Kolmogorov-Smirnov test). Furthermore, repeatability of machine measurements was extraordinarily high (P 0.01), which is much better than would be expected by any one or two technicians. A comparison of the times required for counting and length measurement by manual and by electronic means was made "pitting" an experienced technician against the QMS system. Five samples were prepared, each consisting of 25-100 stage V-VI Calanus finmarchicus copepodites, selected to yield a unimodal size frequency that would readily reveal minor deviations. The technician used a stereo microscope at 12x magnification, calibrated in the usual way with a stage micrometer. Copepods for electronic analysis were stained for about 16 hours (overnight) in a dilute solution of methyl blue to enhance image contrast. A specially made template of elipses in the general range of copepod lengths was used for calibrating the QMS system; detrital particles were rejected by the operator dialing a minimum size criteria; samples with up to 500 individuals required counting 10 successive fields. Operator control was insured by an enhanced outline automatically drawn around each object detected for measurement. Electronic counting was very fast, taking less than 0.1 second for a sample of 100 organisms. Manual counting required 0.7 min and measurement 40.2 min. Total time for all operations - from sample preparation through counting, data summation and plotting a size-frequency distribution - was 53.5 min manually but only 11.9 min by the image analysis system, for a saving of 78%. Manual processing of a 500-organism sample required 264.2 min vs 72.1 min electronically for a 78% saving. Elimination of plotting the length distribution would reduce manual analysis by 36 min, but data summation is a necessary step and by hand it alone takes 21.0 min. Even with a more elaborate image analysis than the QMS system, time savings would probably not exceed 85% over manual. Sample preparation still required up to 25 min for a 500-organism sample (examined as a series of subsamples in a multi-welled chamber). With new devices for sample preconditioning and automatic scanning, we should be able to count, measure, and group 500 organisms in 5-6 min. With allowances made for downtime and systems upkeep, this translates to about 5,000 samples per year, which is near the requirements of large-scale projects. Perhaps more important is the simple fact that any ecological science draws its very substance from a continuing census of populations in nature. Although this seemed impossible for plankton biology a short while ago, the way to go is now clear. The authors thank Dr. L. Pakula, Department of Mathematics, the University of Rhode Island, for his valuable comments with regard to the interpretation of the statistical results. 64 References 1. Almeida, S. P., J. Kim-Tzong eu and P. F. Lai, 1977. "A real-time hybrid optical system for pattern recognition applications," IEEE Trans IM-26:312-316. 2. Beers, J. R. , 1976. "Particle discriminators in studies of zooplankton biomass," p. 69-73 _In Zooplankton fixation and preservation, UNESCO. 3. Fawell, J. K. , 1976. "Electronic measuring devices in the sorting of marine zooplankton," p. 201-206 _In Zooplankton fixation and preservation, UNESCO. 4. Fu, K. S. , 1970. "Statistical pattern recognition," p. 35-70 _In J. M. Mendel and K. S. Fu (eds.). "Adaptive learning and pattern recognition systems, theory and applications," Academic Press, N.Y. 444 pp. 5. , 1978. "Syntactic (linguistic) pattern recognition," p. 96-133 In K. S. Fu (ed.). "Digital pattern recognition," Springer-Verlag, N. Y. , 206 pp. 6. Maurer, R. , 1978. "Plankton counting and measurement with an electronic system." Unpublished ms, National Oceanic and Atmospheric Administration. 7. Ortner, P., S. R. Cummings, R. P. Aftring, and H. E. Edgerton, 1979. "Silhouette photography of ocean zooplankton," Nature 277:50-51. 8. Sherman, K. , 1978. "MARMAP, a fisheries ecosystem study in the NW Atlantic; fluctuations in icthyoplankton-zooplankton components and their potential for impact on the system." Symposium on the study of Main Factors which Govern Functioning of Marine Ecosystem, unpublished ms. 9. Steele, J. H. , and B. W. Frost, 1977. "The structure of plankton communi- ties," Phil. Trans. Royal Soc. London, Ser. B. Biol. Sci. 280:485-534. 10. Sutro, L. L. , 1974. "Study of means of automatically classifying plank- ton," MIT Sea Grant Program Rpt. No. MIT SG 74-11. 11. Uhlmann, D. , 0. Schlimpert and W. Uhlmann, 1978. "Automated phytoplankton analysis by a pattern recognition method," Int. Revue ges. Hydrobiol. 63:575-583. 12. Viglione, S. S. , 1970. "Applications of pattern recognition technology," p. 115-162. JLn J. M. Mendel and K. S. Fu (eds.). "Adaptive learning and pattern recognition systems, theory and applications," Academic Press, 444 pp. 65 OBJECT FINITE DIMENSIONAL BUT LARGE ^ INFINITE DIMENSIONAL TRANSDUCER DEDUCTION DUE TO TRANSDUCER OBJECT LIGHTING <5> REDUCTION PROCESS FINITE DIMENSIONAL BUT SMALL DECISION PATTERN SPACE FEATURE SPACE NUMBER OF CLASSES CLASSIFICATION SPACE CONCEPTUAL/ZED DISC STORAGE OPTICS CAMERA ANALOG TO 4 DIGITAL CONVERTER COMPUTER MONITOR (PROCESSED IMAGE) MONITOR LINE PRINTER TYPICAL COMPONENTS Fig. 1. Typical electronic recognition system. 66 TV MONITOR MICROS A/ COPE VERT CYNC VIDEO ASSEMBLY LINE SELECTOR LINE SWITCHES HOf? SYNC \J VIDEO OBJECT T&1GGER COMPUTER ARM INTER- FACE ARM TRANSIENT RECORD Eft DATA DATA DISC STORAGE «-> CONTROL CONTROL 4 i J TV MONITOR (PROCESSED IMAGE) i > TELETYPE ■» H ETIS PL AY Fig. 2. Details of experimental setup. 67 Computer generated image of an actual copepod. Computer generated images of silhouettes traced from vidicon monitor. Figure 3. Computer-generated images of real and silhouette plankton images. 6S • ,■!■» xvsr ENTER PICTURE NUMBER pause eeeeei arr.nm - 34 PIC.NN » 34 PERM. - 1883.0 AREA * .181.1 LENGTH m 678.2 n««ia» 53.1 Figure 4. Maxima lengths and maxima widths superimposed on computer-generated images. 69 Fig. 5. Smoothed and unsmoothed computer generated images. 70 crfJu p *' » »'* % m 0 „ 1 * • - ' r* * • * « • 1 'X> t N % - * .• % % , |* X . « • t. » . , m V * » i" *♦ • » • * • * • - • > -. • ' « m , Fig. 6. Silhouette photographs of plankton. 71 • 9 «„<%** C 0 00 0 D > c a ta u 2 a u 3 O O w 10 in in in in in in in m in in u w 01 w en in in in m in in in in in in in in « — in in n cm n CM CM cm m CM n CM n ro po s+ssss + ssss + ssss+ssss + ssss + ssss+ssss + ssss + ssss*- in a> in a • CO o a. CO X a> o u in D • 4-> o (0 in ro c • •H © 4-> o PO 4-> D o O JC »H •H CO in CM 4- o O P •H o M CM 0) • 4-> o to 3 win © in O V) © V) o in © x r-j © N in CM © rv in N. O CJ o> n r> n M CM CM IM M 72 c < *C "O "O "0 •O -Q 3 } ! O r — > r-- 2 W CI O CD <<0 CO ■o r-i ■^ rs r-j »H fs. rs »H rs *r rs rs rs is «r rs rs rs rs rs rs is ts ci cd n n ck >o n © is * X ♦ • • • • * * *-* ,- *• + n n n rn cm fi ! < ! N -!- S. i • i © I i ! C CD E •H CD ■o — i c i ^* •H • ^- c ! © o 4-> ^ j c CO i © »H -*- «r Ql 1 ® *-4 CO i fi — fl Cfc- ! * o 1 © Q) i c I •H r-> ! M CD — f! 4-> I • CO 1 '*"' D — r j { • ! © c_> CD 6) 73 NIMBER OF OBSERVATIONS & PERCENTS CLASSIFIED INTO GRNUM: From GRNUM 1 1 2 4 6 7 8 TOTAL 17 1 0 0 0 0 18 94.44 5.56 0.00 0.00 0.00 0.00 100.00 2 0 2 0 0 0 0 2 0.00 100; 00 0.00 0.00 0.00 0.00 100.00 4 0 0 5 0 0 0 5 0.00 0.00 100.00 0.00 0.00 0.00 100.00 6 2 0 0 7 0 0 9 22.22 0.00 0.00 77.78 0.00 0.00 100.00 1 0 0 0 0 4 0 4 0.00 0.00 0.00 0.00 100.00 0.00 100.00 8 0 0 0 1 0 1 2 0.00 0.00 0.00 50.00 0.00 50.00 100.00 Total 19 3 5 8 4 1 40 Percent 47.50 7.50 12.50 20.00 10.00 2.50 100.00 Priors 0.4350 0.0850 0.1150 0.2450 0.0850 0.0350 Fig. 9. Discriminant analysis results with training sample consisted of all 240 observations. 74 NUMBER OF OBSERVATIONS AND PERCENTS CLASSIFIED INTO GRNUM: From GRNUM 1 2 4 6 7 8 Total 1' 103 2 0 0 0 0 105 98.10 1.90 0.00 0.00 0.00 0.00 100.00 2 0 19 0 0 0 0 19 0.00 100.00 0.00 0.00 0.00 0.00 100.00 4 0 0 27 0 1 0 28 0.00 0.00 96.43 0.00 3.57 0.00 100.00 6 0 0 0 58 0 0 58 0.00 0.00 0.00 100.00 0.00 0.00 100.00 7 1 0 0 0 20 0 21 4.76 0.00 0.00 0.00 95.24 0.00 100.00 8 0 0 0 3 0 6 9 0.00 0.00 0.00 33.33 0.00 66.67 100.00 Total 104 21 27 61 21 6 240 Percent 43.33 8.75 11.25 25.42 8.75 2.50 100.00 Priors 0.4375 0.0792 0.1167 0.2417 0.0875 0.0375 Fig. 10. Discriminant analysis results with training sample consisted of 200 observations and a test set of 40 observations. 75 NUMBER OF OBSERVATIONS AND PERCENTS CLASSIFIED INTO GRNUM: From GRNUM 1 2 4 6 7 8 Total 1 17 1 0 0 0 0 18 94.44 5.56 0.00 0.00 0.00 0.00 100.00 2 0 2 0 0 0 0 2 0.00 100.00 0.00 0.00 0.00 0.00 100.00 4 0 0 4 0 1 0 5 0.00 0.00 80.00 0.00 20.00 0.00 100.00 6 1 0 0 8 0 0 9 11.11 0.00 0.00 88.89 0.00 0.00 100.00 7 0 0 0 0 4 0 4 0.00 0.00 0.00 0.00 100.00 0.00 100.00 8 0 0 0 2 0 0 2 0.00 0.00 0.00 100.00 0.00 0.00 100.00 Total 18 3 4 10 5 0 40 Percent 45.00 7.50 10.00 25.00 12.50 0.00 100.00 Priors 0.4350 0.0850 0.1150 0.2450 0.0850 0.0350 Fig. 11. Discriminant analysis results of smoothed plankton. 76 O O O O i?uMnir\iA • • • • VO HNlAN r-J CM CD i-H • Ol-h e O CD Wl^lTiH D co •H r-H z O Q O Q «OOiAO iA NQiTiiA CM CM CD >-\ • a. -H E O CD 10 -d" CO »H O co ■H HH s. o o o o « o o m o <3 CM O iA O CM CM CD — I • • Q. r-\ u e CD o CO CO * • • • o_ CD O •-( »H 4-> CD r-\ CM (D V) (D V) Z CD t-i CD (-> CD H Q. —t »— I— H- h- • E O •— 1 •— 1 "D "D CM CD CO <• iA O O) CO CO CD CD f-K CO •H i-H c c .c r. Z •H «H .»_> -U D> D> O O •H -H O O m cr o o o o O O CO CO Lu XiAOiAiA • • • • «— 1 rH O IA CM CD .— I m d —i E o CD Wn^HlT\ CO •H —1 z p p •H «H C C •H4J-H4J co co w f-i CD M CD 1— 1- >- h- HHT3"D CO CO CD CD c a r. n •H-H4J+I ODIO O •H »H O O O O CO CO 77 FREQUENCY 026.352215 2 + +____.f____+ + + +_s..-^+w — + + + 6 5 5 5 5 5 4 5.5 5 3 5 5 5 5 ? 5555 5 5 55 1 5 5 5 5 5 5.5 5 5 + ^ + + + 4 + + + + ■» 0.500 1.100 1.700 2.300 2.900 3.500 0.800 1.400 2.000 2.600 3.200 FREQUENCY 2244622141 6 6 5 6 4 6 6 6 6 3 6 6 6. 6 2 6 6 6 6 6 6 6 6 1 6 6 . - 4> 6 6 6-6 6 6 6 f + f + + + + + f + f_ 2.000 3.600 5.200 6.800 8.400 10.000 2.800 -.4. 400 ^S.000 7.600 9.200 Fig. 13. a) Distribution of (square root of area) /(maximum width), b) Distribution of (maximum length) /(maximum width). 78 FREQUENCY 10 23 14 A A 1 0 1 23 3 22 3 21 3 20 3 1? 3 18 3 17 3 16 3 15 3 14 3 3 13 3 3 12 3 3 11 3 3 10 3 3 3 9 3 3 3 8 3 3 3 7 3 3 3 6 3 3 3 5 3 3 3 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 — + + 4 4 + 4 f + + 1 + 0.155 0.165 0.175 0.185 0.195 0.160 0.170 0.180 0.190 T Fig. 14. Distribution of first invariant moment function. 79 en 60- o 9 so A Gl LlI 0. 40- c o y MANUAL ELECTRONIC O W GO 30 - 20- 0- 2.0 2.5 3.0 CEPHAL0TH0RAX LENGTH (mm) SAMPLE OF 300 Colonus finmorchicus MEASURED AT N 12 X Fig. 15. Error distribution of length measurement. 80 COMPUTER INTERFACED SYSTEMS FOR BIOLOGICAL MONITORING OF WATER QUALITY by John Cairns, Jr. Biology Department and University Center for Environmental Studies Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061 INTRODUCTION Effectiveness of pollution control historically has been assessed by chemical/ physical parameters. Yet the goal of pollution control is to protect living things. However, no equation exists that permits the precise conversion of the effects of chemical/physical alteration of the environment upon living things. No probe devised by man will measure toxicity or will measure the aggregate effect of multiple physical/chemical alterations upon aquatic biota. However, alterations of the physical/chemical quality of a receiving system rarely occur singly. An ordinary run-of-the-mill industrial waste discharge may typically contain 30 or more constituents. Furthermore, the concentrations of these con- stituents frequently vary markedly due to production and process changes that are cyclical. The problem is compounded further by the presence of many indus- trial and municipal discharges in an aquatic ecosystem and the fact that the natural environment itself rarely remains constant, adding yet another set of changing conditions. Abundant evidence exists that water hardness, temperature, dissolved oxygen concentration, pH, and a variety of other charcteristics may markedly alter the response of an organism to a chemical. The use of organisms as environmental condition sensors has been vastly underutilized by those charged with the re- sponsibility for environmental quality control. As a consequence of the failure to use living organisms as sensors, regulatory and waste treatment measures designed to protect the organisms are rarely as effective as they otherwise would be. It is worth noting, however, that in the absence of chemical/physical information the factors which cause the biological response will very likely be missed. It is, therefore, necessary to couple biological/chemical/physical data for effective environmental protection and quality control. No quality control system can operate effectively without a rapid feedback of information about the quality of the system being controlled. This monitoring of conditions is necessary in order to take immediate corrective measures when the quality falls below acceptable levels. This simple and seemingly obvious fact has been ignored almost entirely where environmental quality control prac- tices are concerned. One or two simple measurements such as BOD, pH, or temper- ature are taken occasionally, but practically never is the condition of the biota in the receiving system measured continuously as would be the case in almost any other quality control system. The reasons for this (other than the 81 failure of the biologists to develop methods which is no longer a valid reason for not engaging in biological monitoring) are (a) the cost of continual biological monitoring and (b) the long time customarily required for examination of biological material, and the analysis and reporting of the condition. In 1966, it was quite clear that biological monitoring as a routine and regu- lar practice would not be widely used even if federal and state regulatory agen- cies proposed it until such monitoring became cost effective and sufficiently rapid to enable corrective measures to be taken promptly. One of the earlier papers on this topic was Cairns et al. (1971). This paper described an "inplant" early warning system designed to detect deleterious materials before they left the plant site and entered the receiving system, and an "instream" method which permitted relatively rapid assessment of community condition of the biota in the receiving system. The inplant system involved a single species. The in- stream system involved a community of species. The inplant system was automated; the instream system was at that time not automated. Subsequently, it seemed desirable to provide a demonstration that most of the commonly used parameters to assess deleterious effects of pollution on aquatic biota could be automated with the use of simple electronic equipment. A number of categories was chosen for demonstration, that is, (a) physiological condition as measured by respira- tory (ventilatory) response, (b) changes in behavior, (c) changes in community structure as measured by diversity indices and indicator species ratios, and (d) changes in response to a rapidly changing wastewater in which both the quantity and quality fluctuated as it does in the "real world." This system would be totally unlike the steady state exposures common to practically all toxicity tests. EARLY WARNING SYSTEM An automated early warning system has been described in detail elsewhere (e.g., Cairns et al., 1973; Cairns, 1972; Cairns, 1975a, b) and will not be discussed extensively here. The system has also had site specific testing (e.g., van der Schalie et al., 1979; Gruber et al., 1978). Both salt and freshwater fish species may be used as sensors (Cairns et al. , 1980). The test chamber is shown in figure 1 and the wastewater testing schematic is shown in figure 2. The diurnal patterns and distress response are depicted in figure 3 and the decision diagram in figure 4. A comparison of this system with others has been made (Cairns and Gruber, 1980). AUTOMATED IDENTIFICATION The automated identification unit is a pattern recognition system not much dif- ferent operationally from the FBI fingerprint matching system. Detailed descrip- tions of the system may be found in Cairns et al. (1972 and 1977) and a check of its accuracy is presented in Cairns et al. (1979). Figure 5 shows how the structural pattern is recorded and figure 6 shows 12 optical spatial filters in a 2" x 2" glass photographic plate. Figure 7 shows a schematic of the system and figure 8 shows the correlation dots which appear when a pattern is "recognized." Figures 9 and 10 show how the information generated could be used to assess pol- lution. Each count would require about 40 hours of skilled technician time using traditional methods, but less than an hour would be required with automation. 82 AUTOMATION OF BEHAVIOR ANALYSIS The analysis of fish pereference and avoidance of certain physical and chemi- cal conditions is important when estimating the effect of an industrial waste mixing zone in a river or lake. Upstream migration could be blocked or impaired if fish were repelled by the material. Conversely, the fish could be killed if a waste attracted them into lethal concentrations of chemicals. The system was not designed to duplicate natural conditions but rather to get a preliminary esti- mate of preference, avoidance, and induction of aberrant behavior. Figure 11 (from Lubinski et al., 1977) provides a schematic of the system. Figure 12 illus- trates one type of evidence provided. Additional information may be obtained from Lubinski et al. (1978). SIMULATION OF VARIABLE EFFLUENT DISCHARGES Practically every labortatory toxicity test is designed to give a constant dose of a particular chemical to a particular test organism. While steady state concentration is often difficult to achieve, variability in toxicant concentration occurs accidentally rather than as a matter ef intent to do so. The Mount-Brungs serial diluters and other such devices were designed to enhance the probability of providing constant concentrations of test materials which might not easily remain at a constant concentration in nonrenewable batch test- ing systems. In the "real world," however, industrial discharges are rarely constant in quality, and often not even in quantity, for a substantial period of time. It would be astonishing if discharges remained constant for the most commonly used toxicity testing exposure period of 96 hours. Therefore, labora- tory tests with constant conditions are not particularly reflective of the types of dosing or exposure experienced by organisms in the "real world." The scientific method encourages keeping everything constant in an experiment except one variable (the one of particular interest in the study). Thus, the kind of constancy just described for laboratory testing is ideal if one wishes to get manuscripts accepted for publication in scholarly journals but does provide some difficulty when attempting to determine the relationship between this type of exposure and the exposure actually experienced by aquatic organisms. In addition to the reason just described, there are three additional reasons why such experiments are so rare, that is, (a) it is difficult to control delivery of the test material so that the concentration is reasonably constant, let alone controlling one deliberately designed to be variable; (b) it is difficult to preprogram such delivery in such a way that one can be assured that it actually will occur; and (c) it is difficult to record response of test organisms to highly variable conditions unless one has a continuously recording system. The system now under development for two years at this institution addresses all the aspects just mentioned. Although it is worth emphasizing that this is still an experimental system subject to the usual failures and uncertainties of any research activity, the evidence generated thus far indicates that the system may very well be suitable for testing the effects of variable waste loadings. A more detailed description of the apparatus and some of the test results may be found in Cairns and Thompson (in press), and a description of the statistical analysis used may be found in Thompson et al. (in press). 83 Figure 13 gives a schematic of the system used (from Cairns and Thompson, in press). Figure 14 shows a constant exposure and the variable response of the fish to it and figure 15 shows a pulsed exposure and the response of the fish to it. Note the recovery following the peak response. CONCLUSION Four types of computer interfaced systems designed to acquire biological response to pollutants are described. One (the early warning system) has been in development since 1966 and has been tested for over a year at one industrial site and for several years at another. The automated identification unit has been under development since 1970 and should shortly be ready for practical use if this type of information in this quantity is needed. The various forms of the behavior system have been under development since 1966 and have been tested in several practical situations. The unit designed to simulate fluctuating effluent discharges has been in development since 1978 and appears promising, although much additional work is needed before it can be routinely used for applied problem solving. The illustrations provided show that certain types of biological pollution assessment can be automated, and the information can be directly coupled with chemical /physical information. The minicomputer systems may be used not only to generate data but also for storage and analysis. These procedures substan- tially increase the amount of data that can be handled, reduce the time needed to analyze data, and vastly improve the opportunity for statistical evaluation of the data. Data storage and retrieval are also markably enhanced. Addition- ally, through the use of leased telephone lines, data can be transmitted to larger computers with more sophisticated capabilities should these be advan- tageous. If biological assessment is to enjoy widespread use, the old fashioned labor intensive measurements must be replaced in part, especially where routine meas- urements are concerned, by the more cost effective and time efficient computer interfaced systems. Skills of highly trained persons would be better used since these persons would be free of routine determinations and therefore able to take fuller advantage of their analytical skills. Additionally, a rapid response time permits equally rapid corrective action and, more important, enables one to make closer correlations between alterations of chemical/physical conditions and the biological responses they elicit. Since the U.S. Environ- mental Protection Agency and presumably other federal agencies charged with environment responsibilties have a vastly increased work load (Carter, 1980) without a concomitant increase in funds and personnel, it would appear advanta- geous to automate whenever possible, particularly since not only data genera- tion, but data analysis, storage, retrieval, and other activities now labor intensive can be handled by "shelf" electronics equipment. ACKNOWLEDGMENT This research was supported by a grant-in-aid from the Monsanto Fund. The author acknowledges Darla Donald for the editing and preparation of this manuscript and Betty Higginbotham for the final type copy. 84 LITERATURE CITED Cairns, Jr., Jr. 1972. "Rationalization of multiple use of rivers." Pages 421- 43 in R. T. Oglesby, C. A. Carlson, and J. A. McCann, eds. River Ecology and the Impact of Man. Academic Press, New York. Cairns, J., Jr. 1975a. "Quality control systems." Pages 588-612 In B. A. Whitten, ed. River Ecology. Blackwell's Scientific Publications Ltd., London. Cairns, J., Jr. 1975b. "Critical species, including man, within the biosphere." Naturwissenschaf ten 62(5) : 193-199. Cairns, J., Jr., K. L. Dickson, G. R. Lanza, S. P. Almeida, and D. Del Balzo. 1972. "Coherent optical spatial filtering of diatoms in water pollution monitor- ing." Archiv. Microbiol. 83:141-146. Cairns, Jr., Jr., K. L. Dickson, P. Pryfogle, S. P. Almeida, S. Case, J. M. Fournier, and H. Fugii. 1979. "Determining the accuracy of coherent optical identification of diatoms." Water Resour. Bull 15(6) : 1770-1775. Cairns, J., Jr., K. L. Dickson, and J. Slocomb. 1977. "The ABC's of diatom identification using laser holography." Hydrobiologia 54(1):7-16. Cairns, J., Jr., K. L. Dickson, R. E. Sparks, and W. T. Waller. 1970. "A preliminary report on rapid biological information systems for water pollution control." J. Water Polut. Control Fed. 42(5) :685-703. Cairns, J., Jr. and D. Gruber. 1979. "Coupling of mini- and microcomputers to biological early warning systems." BioScience 29( 11) :665-669. Cairns, J., Jr., and D. Gruber. 1980. "A comparison of methods and instrumenta- tion of biological early warning systems." Water Resour. Bull. 16(2) :261-266. Cairns, J., Jr., G. R. Lanza, R. E. Sparks, and W. T. Waller. 1973. "Develop- ing biological information systems for water quality management." Water Resour. Bull. 9(l):81-99. Cairns, J., Jr., and K. W. Thompson. In press. "A computer interfaced toxicity testing system for variable effluent loading." _In Second Symposium on Process Measurements for Environmental Assessment, A. D. Little, Inc., Boston, Mass. Cairns, J., Jr., K. W. Thompson, D. Landers, M. McKee, and A. C. Hendricks. 1980. "Suitability of some freshwater and marine fishes for use with a minicomputer interfaced biological monitoring system." Water Resour. Bull. 16(3). Carter, Luther, J. 1980. "Environmentalists seek new strategies." Science 208(4443) :477-478. Gruber, D., J. Cairns, Jr., K. L. Dickson, A. C. Hendricks, M. A. Cavell, J. D. Landers, Jr., W. R. Miller, III, and W. J. Showalter, Jr. 1978. "The construc- tion, development, and operation of a fish biological monitoring system." Contractor Report ARCSL-CR-78040, U.S. Army Armament Research and Development Command, Chemical Systems Laboratory, Aberdeen Proving Ground, Maryland. 85 Lubinski, K. L., J. Cairns, Jr., and K. L. Dickson. 1978. "Quantifying the effects of ammonia on the swimming behavior of bluegills. Pages 508-514 in D. D. Hemphill, ed. Trace Substances in Environmental Health, XII. University of Missouri, Columbia. Lubinski, K. L., K. L. Dickson, and J. Cairns, Jr. 1977. "Microprocessor- based interfaced converts video signals for object tracking." Computer Design Dec: 81-87. Patrick, R. , M. H. Hohn, and J. H. Wallace. 1954. "A new method for determining the pattern of the diatom flora." Notulae Naturae. 159:12. Thompson, K. W. , M. Deaton, R. V. Foutz, J. Cairns, Jr., and A. C. Hendricks. In press. "Application of a time series intervention analysis to describe the effects of an intermittent application of a toxicant on the respiratory behavior of the bluegill (Lepomis macrochirus). " Am. Soc. Ich. Herp. van der Schalie, W. H. , K. L. Dickson, G. F. Westlake, and J. Cairns, Jr. 1979. "Fish bioassay monitoring of waste effluents." Environ. Manage. 3(3) :217-235. 86 d o •H U P O CD rH 0 O s P /*-N T3 £ fl bJD 0 a •H Si X O -P d d CD 0 h P £ d ~ v_/ CO CD >» T5 h O o !h ■p P ed O rH CD •H iH P CD fl CD XI > -P •H s £ 0 h $H «H CD ,a CO B CD d bJD X Fh o ed ja p o CO CO CD •H H n 87 IN -PLANT MONITORING UNIT UPSTREAM WATER TO CENTRAL PROCESSOR DRAIN 2. An inplant monitoring system with control fish exposed to dilution water only and test fish exposed to a mixture of wastewater and dilution water. 88 60 40- 20- 0 DAY NIGHT DC cc 4(H 3. Dotted lines show confidence intervals, solid lines actual ventilatory response frequency. Note day-night differences 89 BIOMONITORING FACILITY BIOLOGICAL ASSESSMENT CHEM/PHYS ASSESSMENT CONTINUOUS a PERMANENT RECORDS / YES NO I ADDITIONAL TREATMENT REQUIRED ADVANCED CHEMICAL ANALYSIS — TYPE FOLLOW SPCC 8 BMP PLANS AS ON NPDES PERMIT DISCHARGE APPROVED IMPLEMENT EMERGENCY PROCEDURES WATER SAMPLE AUTOMATICALLY COLLECTED COMMUNICATE "EMERGENCY CONDITION" TYPE II I SUBJECTIVE DECISION 4. Decision-making diagram for handling "alarm" condition (from Cairns and Gruber, 1979 - reprinted with permission from BioScience) . 90 S£uj 1° O UJ u."> or o cr QCO HI w ZUJ <-* u. UJ or ujSj ft ou U.O or COqj UJu. £or LJ UJ or UJ X o o X o -Oo o z z z 2 ggss ^uj?o UJ< air ° CD v-» _ UJ r- Q |— — wo< if 0.5 _| ss UJ 0T Qj< 0. UJ a o -p ■p o w O o ^ UJ < H p - Q CO — CM — Q. 0. O w — z2co CO? UJ Q X I- Z° 5 UJ or U. coq < K o u S = CO z UJ I _J 0 - ItO UJ 00 J ^s UJ UJ I- 1- co >- CO I -J < z z -j ©5 <£ < O CO to i-o-Jtru. , IK3^o Q . 3Z< CL^Ct ZZ£K -3< o •H +-> £ 0) xi o u v c CD o a: uj CO < 93 8. Input and output planes for a correlation test of the diatom Heliopelta metii . The intensity of the correlation dots indicate an excellent match. Note the position of the dots relative to the input array. (Slide of arrayed diatoms courtesy of Drs . Ruth Patrick and Charles Reiner, Philadelphia Academy of Natural Sciences). 94 1 *t — c3 CVJ m CD U P i 00 CVJ <0 if) T5 CD -P 3 CVJ LU rH O 0 1 00 CVJ LU a 3 if) a * i > CVJ p •H ro if) a - <0 _l 3 ^ < S 10 0 O) 3 O rH <0 Q s - ^^ o • 00 > -P rH ~ ~ •H Q 73 P 00 2 1 ^^ SH O * O -H CD P> 3 Ph 1 P o S CVJ 3 0 CVJ 1 O 1 o "IVAH31NI d3d S3l03dS 95 en u UJ a. Q Z s o i 0 *H -P 03 T3 0 +-> 3 O ft s o «H >» +-> •H a 3 O a e^ o ^ -p io a © •H H aJ • «H Ctf o -p 0) O •H T5 CO G •H cd > 0 CD rH O CD cd +-> O •H £ • i S O CD CD -P T$ CO -H >> > CO •d fcdO-P G O •H ,Q rj CO -H -P rH G ft •h E O cd a co "H CO O rH o CO Jh CD -P -P Oj G •H Ph O -P O 3 o a s >H O o a h cd o •H X s rH T3 cd G -P cd •H hO - •H -p T5 O CD CO -r-j CD ,Q ■P O $h m 0 o G CD £1 coDP1 bJO 0 P> a •H a>< 0 rH rH T3 Cd 0 G -P ■P d O O b£X G rH •H Cd CO cd P> 3 0 -H £ W> 0 -H hCHH T3 cd O I • iH CO /-n O CO 0 +-> 0 cd -P CO -H cd r< >> G 13 0 rQ -H G CO T3 cd Tt U 0 0 0 - £ £ O CO -H O O Cl+JH O rH X •H Cd O -P =HH cd 0 G Cd T3 O cd CO -P a •H cd o fl +-> O P> cd cd 0 -h £ t3 CO bJ3 •H cd £ •p cd cd CS n 0 fl O 0 Cd rH Jn bJO 0 fl P O G rH S 0 t-H cd O •P •H G P> O G G T3 •H I CO 0 £ cd - • -H CO Si -P T5 0 O w G P> G cd cd 3 CO G a 0 0 •H -P 4-> T5 0 cd cd u a G G O cd -h -H O -P TS T3 O ?h ^ TJOO i * i L_» 1 1 aD LJ cr ZDT if) O Cl LJ LJ en ■* V > > > > > > » > > 1 »* o > LJ *»> 7 » » h- •• % y in i *' > > o Cl r>*i » ! »>; j > > * > > r*> > » \i»> > > i ^ i > J* »~ ^ iL »4aL CD !h CO o x CD bO c •H o o d -p -P -P •H d O X! -H 03 X •H O «H -P o o fl CD O CO •H O -P T3 •H CO H O ctf -p «H CD O r-i CO 3 0 CO fcdO ^3 O O -P (M 98 DECHLORINATED TAP WATER 22" C DILUENT WATER RESERVOIR r~~--\ I TOXICANT DELIVERY PUMP .-T..---M CD □ \l // if II II TOXICANT RESERVOIRS MIXING CHAMBERS WATER DELIVERY CHAMBERS ISOLATION ENCLOSURE AND FISH CHAMBERS SAND FILLED BOXES H^- *: \k STRIP CHART RECORDER U \- OSCILLOSCOPE TO IBM/370-158 13. Schematic of variable dosing apparatus. 99 10.0 9.0 ^\^ E 5 0 Jr** l6° FISH 7 TREATED 2I6 FISH 14 TREATED FISH 9 CONTROL TIME (HOURS) 14. Ventilatory response of fish to constant concentration. 100 9.0 8.0 V A> r- 7.0 1 1 1 216 FISH W TREATED 216 FISH 13 TREATED FISH 9 CONTROL TIME (HOURS) 15. Ventilatory response of fish to pulsed concentration. 101 ALGAL CHLOROPHYLL MEASUREMENTS: TRADITIONAL AND UNTRADITIONAL APPLICATIONS OF A BIOLOGICAL SENSOR Bruce L. Kimmel Biological Station and Department of Zoology University of Oklahoma Norman, Oklahoma 73439 INTRODUCTION Characteristics of a useful biological sensor for water quality monitoring must include (1) biological and ecological significance; (2) sensitivity to environmental events; (3) simplicity of measurement; and, preferably, (4) suitability for automated in-situ determination. Measurements of photosynthetic pigments, particularly chlorophyll a_, have long been recognized as valuable biological parameters for oceanographic and limnological investigations and water quality monitoring. Attributes which make chlorophyll determinations useful for these purposes include (1) the chlorophyll a^ molecule (fig. 1) which is the essential photocatalyst for the photosynthetic production of organic matter, and therefore, its measurement provides an indication of both algal biomass and potential production; (2) chlorophyll-bearing primary producers which form the ecological base of most marine and freshwater systems, and respond sensitively to environmental conditions by changes in cellular physio- logical status, productivity and/or biomass; and (3) algal photosynthetic pigments which can be extracted from mixed phytoplankton assemblages into organic solvents, and chlorophyll concentrations determined by spectrophoto- metry or fluorometric methods. However, the amount of chlorophyll present per unit algal biomass is not constant, but varies with light intensity, cellular physiological status and the taxonomic composition of the algal assemblage sampled. Thus, chlorophyll determinations for ecological or water quality purposes should be considered only semiquantitative estimates at best. That such a semiquantitative indicator has received continued scientific application is testimony to the biological and ecological pertinence of chlorophyll measurements. If scientific progress was contingent on the availability of perfect methodologies, there would be very little of it. Measurement of phytoplankton pigment levels, introduced by Harvey in 1934 (Vollenweider 1974), continues to be widely employed as an indicator of algal standing crop and potential primary production. Although considerable research effort has been invested in methodological modification, in-vitro pigment analyses (i.e., pigments extracted from the cells) have remained relatively unchanged during the past two to three decades (e.g., see the Report of the Committee on Chlorophyll from the international symposium, "Measurements of 103 Primary Production in the Sea," Bergen, Norway, September 1957: Rapp. Proc- Verb. , Cons. Internat. Explor. Mer 144, 1958). However, significant progress has been made in applying fluorometry to both in-vitro and in-vivo chlorophyll determinations. Fluorometric techniques extend the sensitivity of in-vitro measurements, while determinations of in-vitro chlorophyll fluorescence (i.e., fluorescence from intact living cells) provide not only an index of phytoplankton biomass, but can also serve as an indicator of the physiological status of naturally occurring phytoplankton assemblages (e.g., Samuelsson et al. 1978; Cullen and Renger 1979; Kimmel and White, in press). IN-VITRO CHLOROPHYLL METHODS Spectrophotometric Measurements — The spectrophotometry determination of extracted pigment levels continues to be widely applied as a measure of algal standing crop. A variety of organic solvents (e.g., acetone, methanol, anhy- drous ether, and dimethyl sulfoxide) has been employed in efforts to obtain sharper absorption peaks for the chlorophylls and accessory pigments (caroten- oids, xanthophylls, and phycobilins). Individual investigators and international working groups have recommended slightly differing empirically derived formulae for estimating chlorophyll a_, _b and c^ concentrations based on the determination of absorbance at several wavelengths, usually 630, 645, 665 and 750 nm (e.g., Richards and Thompson 1952; Parsons and Strickland 1963, UNESCO 1966; and Jeffrey and Humphrey 1975). These procedures were developed for the purpose of distinguishing the contributions of chlorophylls a^, _b, c^, and total carotenoids (which have overlapping absorption spectra) in a mixed pigment extract. Addi- tionally, much effort has been spent evaluating various combinations of concen- tration and storage procedures, determining pigment extraction efficiencies as a function of solvent and extraction method, and establishing corrections for the occurrence of phaeopigments (e.g., Humphrey 1963, Marker 1972, Jeffrey and Humphrey 1975, and Shoax and Luim 1976). Chlorophyll degradation products (phaeopigments) interfere with the spectro- photometric determination of chlorophyll concentrations by absorbing light in the same spectral region (663-665 nm) as chlorophyll a.. However, interference due to the presence of phaeopigments can be corrected for by determining the relative changes in absorbance (or fluorescence) induced by acidification of the sample (e.g., Yentsch and Menzel 1963; Holm-Hansen et al. 1965; Lorenzen 1967; and Moss 1967). Such a variety of protocols for chlorophyll measurement has been proposed that standardization of procedures to permit comparison of results has become of more immediate need than development of the "ideal" tech- nique. Most investigators in the U.S. now commonly employ the spectrophoto- metric and fluorometric methods described by Strickland and Parsons (1972) and Vollenweider (1974) for in-vitro determinations of chlorophyll concentrations. Fluorometric Measuremenmts — Photosynthetic pigment extracts are fluorescent; when excited by short wavelength radiation in the blue region (430-450 nm) , long wavelength radiation in the red region (650-680 nm) is emitted. An appropriate combination of excitation lamp, primary and secondary filters, and a red-sensitive photomultiplier permits fluorometric determination of extracted chlorophyll a_ with lOOx greater sensitivity than by spectrophotometric methods (Kiefer 1973a). Kalle (1951, cited in Krey 1958), Yentsch and Menzel (1963) and Holm-Hansen et al. (1965) were the first to take advantage of the sensitivity 104 and convenience of fluorometric methods for chlorophyll determinations in the field. Fluorometric techniques for chlorophyll and phaeopigment estimations are now used widely, particularly in studies of oligotrophic systems where the increased sensitivity provides maximum benefit. However, fluorometric measure- ments are in relative units and must be standardized against spectrophotometric determinations of "known" chlorophyll concentrations to provide quantitative chlorophyll estimates. IN-VIVO CHLOROPHYLL METHODS In-vivo Chlorophyll Fluorescence — The in-vitro chlorophyll techniques dis- cussed above necessarily involve (1) collection and filtration of discrete samples, (2) grinding and extraction of samples in an organic solvent, and (3) determination of the absorbance or fluorescence of each extracted sample. Although significantly less time consuming and laborious than estimation of algal biomass by cell enumeration procedures, in-vitro methods are not readily applied to continuous or in-situ water-quality monitoring. Measurement of in- vivo chlorophyll fluorescence avoids the concentration and extraction require- ments of in-vitro methods and thus makes continuous monitoring of algal chloro- phyll levels feasible. The characteristics of in-vivo chlorophyll fluorescence (maximum excitation at 436 nm, maximum emission at 685 nm) are only slightly different from those of chlorophyll in vitro. In-vivo chlorophyll fluorescence yield (fluorescence per unit extractable chlorophyll) is only about 10% that of chlorophyll dis- solved in acetone; however, the availability of highly sensitive photomultipli- ers in modern fluorometers minimizes sensitivity loss. Lorenzen (1966) first proposed a method for the continuous measurement of chlorophyll levels and demonstrated the correspondence of in-vivo chlorophyll fluorescence and extract- able chlorophyll for marine phytoplankton samples (fig. 2). Lorenzen' s method- ology (involving continuous pumping of unf iltered seawater through a standard fluorometer fitted with a large volume flow-through cuvette) has been widely applied in both marine and freshwater systems to the investigation of (1) hori- zontal chlorophyll distribution (e.g., Armstrong et al. 1967, Flemer 1969, and Caperon et al. 1971) and phytoplankton patchiness (e.g., Piatt et al. 1970 and Powell et al. 1975) and (2) vertical chlorophyll profiling (e.g., Berman and Rodhe 1971, Kiefer et al. 1972, and Berman 1972). Although continuous measurement of in-vivo chlorophyll fluorescence appears to provide an easily obtained real-time estimator of in-situ phytoplankton standing crop, the method is not without problems. The correspondence of in-vivo fluorescence to chlorophyll concentration is not constant but varies with (1) the recent light history of the cells, (2) algal physiological state as influenced by growth phase and available nutrients, (3) diel rhythms in photosynthetic activity, and (4) the taxonomic composition of the phytoplankton community sampled. Fluorescence yield (fluorescence / mg Chi a. ra~3) has been observed to vary 2-5x under laboratory conditions (Kiefer 1973b, Slovacek and Bannister 1973, and Heaney 1978) and up to lOx in natural phytoplankton assem- blages (Kiefer 1973c, Loftus and Seliger 1975, and Heaney 1978). Such varia- bility makes necessary frequent comparisons of in-vivo fluorescence readings with extracted chlorophyll measurements to permit accurate data interpretation. 105 DCMU-Enhanced In-Vivo Chlorophyll Fluorescence — Photosynthesis researchers have employed the herbicide DCMU (3-(3 ,4-dichlorophenyl)-l ,1-dimethylurea) in experimental investigations of the photochemical processes occurring in green plant cells for almost two decades. The maximal in-vivo fluorescence and lack of photosynthetic oxygen evolution observed when algae or other plant cells are exposed to light in the presence of DCMU has been attributed to DCMU-blocking of photosynthetic electron transport and the subsequent dissipation of much of the energy of excited chlorophyll molecules by fluorescence (Duysens and Sweers 1963 ).l Although the quantum yield of energy absorbed by the reaction centers of Photosystem II can be dissipated by several competing de-excitation processes (photochemistry, fluorescence from Photosystem II, spillover to Photosystem I, or radiationless transfer) a complementary relationship of photosynthetic activity and in-vivo chlorophyll fluorescence has often been observed (e.g., Butler 1966, Goedheer 1972, Papageorgiou 1975). Such correspondence suggests (1) a partial explanation for the variability of fluorescence yield observed for algal cultures and natural phytoplankton samples (i.e.; fluctuation in photo- synthetic activity, thus variable in-vivo fluorescence per unit chlorophyll), (2) the use of DCMU to reduce the variability of in-vivo chlorophyll measure- ments, and (3) application of the variable fluorescence response (i.e., the relative fluorescence increase induced by DCMU poisoning) as an indicator of algal physiological status. Slovacek and Hannan (1977) examined the applicability of DCMU-poisoning for maximizing the fluorescence yield of phytoplankton samples, and thereby reducing the variability of in-vivo chlorophyll fluorescence estimates of algal biomass. DCMU-poisoning of samples effectively reduced dependence of in-vivo chlorophyll fluorescence on environmental conditions and algal growth state, and decreased the variability of fluorescence yield (fig. 3). They concluded that DCMU- blocking of electron transport from Photosystem 11^ separates chlorophyll a_ from its physiological function, as does chlorophyll extraction into an organic solvent. However, other investigators have had more mixed results in applying DCMU-enhanced chlorophyll fluorescence to in-vivo chlorophyll measurements (e.g., Esaias 1978, McMurray 1978, Prezelin 1978, Frey 1979, and Slovacek 1978, 1979). Esaias (1978) observed little improvement in variability of fluorescence yield in a comparison of unpoisoned and DCMU-poisoned fluorescence measurements using parallel continuous-flow f luorometers. Slovacek (1978) noted similar difficulties, but attributed at least part of the variability to the continuous- flow system. White (1980) found that although DCMU-enhanced fluorescence provided increased analytical sensitivity at lower chlorophyll concentrations, the in-vivo fluorescence of both poisoned and unpoisoned samples was highly correlated with extracted chlorophyll ji and, thus, either measurement was equally suitable as a biomass indicator. 1-DCMU prevents reoxidation of the Q intermediate and thereby blocks noncyclic electron transport from Photosystem II and the oxidation of water (i.e., the Hill reaction). 2 At ambient temperatures, in vivo fluorescence emanates primarily from chloro- phyll a associated with Photosystem II; Photosystem I is only weakly fluorescent, 106 Samuelsson and Oquist (1977) and Heaney (1978) suggested that the variable nature of in-vivo chlorophyll fluorescence might be used to an advantage in the investigation of algal physiological state in situ. Halldal and Halldal (1973) had specifically indicated the potential usefulness of DCMU-enhanced in-vivo chlorophyll fluorescence for phytoplankton studies. By virtue of the approxi- mate complementarity of photosynthetic activity and in-vivo chlorophyll fluor- escence, an indication of the physiological status of a phytoplankton assemblage should be possible by comparison of the in-vivo fluorescence (unpoisoned) and the maximum (DCMU-enhanced) fluorescence. Samuelsson and Oquist (1977) and Samuelsson et al. (1978) reported the correspondence of DCMU-induced fluores- cence increases and photosynthetic l^C uptake in algal cultures, and further emphasized the potential applicability of the method to the investigation of natural phytoplankton communities. A direct technique for assessing the physiological status of naturally occur- ring phytoplankton assemblages in-situ would indeed be valuable for examining factors influencing phytoplankton productivity, biomass accumulation, spatial distribution, and bloom development and decline. Currently, it is necessary to make both primary productivity and extracted chlorophyll measurements to obtain data on phytoplankton physiological state or photosynthetic potential. Therefore, we conducted a series of laboratory and field evaluations of the DCMU-enhanced in-vivo fluorescence response as an index of phytoplankton physi- ological status. Although some of our results are being reported elsewhere (Kimmel and White, in prep.), they are not yet available in published form and therefore will be summarized here. We first tested the correspondence of fluorescence response and photosynthetic l^C fixation in laboratory light-deprivation experiments designed to simulate the gradual physiological decline experienced by phytoplankton cells introduced to aphotic depths (fig. 4). The fluorescence response index (FRI) proved to be directly related to photosynthetic carbon fixation in these experiments (fig. 5), thus indicating that the fluorescence response reflects changes in phyto- plankton photosynthetic potential, at least in the laboratory. 3 Subsequently, vertical profiles of in-vivo chlorophyll fluorescence (F-^), DCMU-enhanced flu- orescence (F^), and FRI in several Oklahoma reservoirs revealed differences in phytoplankton physiological status, apparently related to differing conditions of light and nutrient availability and vertical mixing patterns (figs. 6 and 7). In-situ sample displacement experiments conducted in conjunction with primary productivity measurements provided field evidence that FRI values accurately reflected phytoplankton photosynthetic potential (table 1).^ Flu- orescence response measurements recently conducted on near-surface samples taken along a well-established nutrient availability gradient in Lake Mead o Fluorescence response index = FRI = (Fi - F-) / F*; where F. = initial (unpoisoned) in-vivo fluorescence and F^ = DCMU-enhanced in-vivo fluorescence. FRI has a theoretical range of 0.0 (i.e., no photochemical electron transport) to 1.0 (i.e., all absorbed energy dissipated photochemically) , but usually ranges from 0.2 to 0.7. ^Samples obtained from several depths were incubated at a single depth to compare their photosynthetic potentials under similar light and temperature conditions. 107 Table 1. — Comparison of phytoplankton photosynthetic potential as reflected by photosynthetic carbon uptake (C uptake / F^) and DCMU-enhanced in-vivo chlorophyll fluorescence (as FRI) in Lake Texoma and Broken Bow Lake sample displacement experiments. Incubation depths were 1 m (520 uE m~2 sec ~1) and 3 m (300 uE m sec ) in Lake Texoma and Broken Bow Lake, respectively. F., which reflects algal chlorophyll, is expressed in relative units (counts sec~l) From Kimmel and White (in press). Sampling C Uptake C Uptake Depth (m) Fd (yg C l"1 hr _1) FRI F~j Lake Texoma, 26 Jul 1978: 1 226 46.86 0.51 0.21 3 215 43.25 0.50 0.20 8 202 44.66 0.50 0.22 Broken Bow Lake, 14 Aug 1978: 3 19 4.04 0.38 0.21 8 47 25.14 0.41 0.54 12 30 18.89 0.46 0.64 (Nevada-Arizona) ranged from 0.39 to 0.63 (as FRI) and were directly related to available nutrients and carbon fixation rates (Kimmel, unpublished data). Additionally, it appears that fluorescence response measurements may be appli- cable to algal bioassays. Preliminary tests show FRI determinations to be a sensitive and convenient means of assessing phytoplankton responses to experi- mental nutrient and silt additions. The same should hold true for toxicity bioassays, although we have not yet explored this potential application. Our results suggest that the chlorophyll fluorescence response technique provides a rapid and convenient method for obtaining information on the in-situ physiological status of naturally occurring phytoplankton assemblages, informa- tion otherwise obtainable only by measuring both primary productivity and extracted chlorophyll. However, in-vivo chlorophyll fluorescence is still a poorly understood phenomenon, and its application to field investigations and environmental monitoring is not free of uncertainty. Diel fluctuations in fluorescence response and photosynthetic activity (Prezelin and Sweeney 1977; and Prezelin and Ley, in press), the relationship of F^ and F^ to extractable chlorophyll levels (McMurray 1978, Esaias 1978, and White 1980) effects of algal community species composition (Kiefer 1973c, Heaney 1978, and White 1980) 108 and high light intensities (Vincent 1979, and G. P. Harris, pers. commun. ) on in-vivo f lueorescence and fluorescence response measurements all require further clarification. At present, DCMU-enhanced in-vivo fluorescence data must be interpreted cautiously and applied only as a relative indicator of algal biomass and physiological status. Our initial evaluation of the fluorescence response technique and its applica- tion in our limnological studies has produced a sense of cautious optimism in regard to future applications of the method. The simplicity and convenience of the fluorometric analysis, the ecologically integrative nature of the informa- tion obtained, and the potential of the technique for automated in-situ water quality monitoring comprise strong arguments for continued evaluation and devel- opment of in-vivo chlorophyll fluorescence response as a biological sensor. SUMMARY AND PROSPECTUS Chlorophyll a. is essential for the photochemical conversion of radiant energy to the potential energy represented by organic compounds. As such, estimates of chlorophyll levels present in marine and freshwater systems have long been em- ployed as a biological sensor of algal biomass and potential photosynthetic pro- duction. Spectrophotometric and fluorometric determinations of extracted chloro- phyll a. will continue to be widely used as indicators of phytoplankton standing crop; however, in-vivo fluorescence methods provide greater potential for in- situ monitoring purposes. Determination of horizontal and vertical phytoplankton distributions and auto- mated continuous monitoring of algal chlorophyll levels at stationary points are possible using continuous-flow fluorometry. Mapping of longitudinal chlor- ophyll gradients in near-surface water with continuous-flow systems provides a means of obtaining, on a frequent basis, the real-time ground-truth data required for evaluating the utility of satellite remote sensing for water quality monitoring. The combination of in-vivo chlorophyll fluorescence and DCMU-enhanced fluorescence determinations provides an indication of both algal biomass and physiological status. Thus, the fluorescence response technique represents (1) a significant increase in the information content of chlorophyll measurements and (2) a high potential for application to automated in-situ monitoring of algal communities. In-vivo chlorophyll fluorescence is a complex and poorly understood process; however, the sensitivity, convenience, and information content of fluorescence response measurements suggest that the applicability of the technique for water quality monitoring purposes should be thoroughly explored. ACKNOWLEDGEMENTS This review is a contribution from the University of Oklahoma Biological Sta- tion, Dr. L. G. Hill, Director. Research on Oklahoma reservoirs was supported by grants from the University of Oklahoma Research Council and the U.S. Depart- ment of Interior, Office of Water Research and Technology (A-088-OKLA) to the author. Lake Mead data were obtained while the author was a visiting research scientist at the Lake Mead Limnological Research Center, University of Nevada at Las Vegas, Dr. L. J. Paulson, Director. 109 REFERENCES Armstrong, F.A.J. , C.R. Stearns and J.D.H. Strickland. 1967. "The measurement of upwelling and subsequent biological processes by means of a Technicon autoanalyzer and associated equipment." Deep- Sea Res. 14: 381-389. Berman, T. 1972. "Profiles of chlorophyll concentrations by in vivo fluores- cence: Some limnological applications." Limnol. Oceanogr. 17: 616-618. Berman, T. , and W. Rodhe. 1971. "Distribution and migration of Peridinium in Lake Kinneret." Mitt. Int. Verein. Limnol. 19: 266-276. Butler, W.L. 1966. "Fluorescence yield in photosynthetic systems and its relation to electron transport," p. 49-73. _In D. Sanadi (ed.), Current Topics in Bioenergetics, v. 1, Academic Press. Caperon, J., S.A. Cattell and G. Krasnick. 1971. "Phytoplankton kinetics in a sub-tropical estuary: Eutrophication." Limnol. Oceanogr. 16: 599-607. Cullen, J.J. and E.H. Renger. 1979. "Continuous measurement of the DCMU- induced fluorescence response of natural phytoplankton populations." Mar. Biol. 53: 13-20. Devlin, R.M. , and A.V. Barkler. 1971. "Photosynthesis." Van Nostrand Reinhold, Duysens, L. , and H. Sweers. 1963. "Mechanism of two photochemical reactions in algae as studied by means of fluorescence," p. 353-372. _In Studies in Microalgae and Photosynthetic Bacteria, Jap. Soc. Plant Physiol., Univ. Tokyo Press. Esaias, W.E. 1978. "Use of DCMU and light exposure in continuous flow chloro- phyll fluorometry. " Abstr. , Amer. Soc. Limnol. Oceanogr. Mtg. , June 1978, Victoria, B.C. Flemer, D.A. 1969. "Continuous measurement of in vivo chlorophyll of a dino- flagellate bloom in Chesapeake Bay." Chesapeake Sci. 10: 99-103. Frey, B. 1979. "The ratio of DCMU-enhanced fluorescence to _in vivo fluores- cence as an indicator of the state of growth of natural phytoplankton populations." Abstr., Amer. Soc. Limnol. Oceanogr. Mtg., Jan. 1979, Corpus Christi, Tx. Goedheer, J.C. 1972. "Fluorescence in relation to photosynthesis." Ann. Rev. Plant Physiol. 23: 87-112. Halldal, P., and K. Halldal. 1973. "Phytoplankton, chlorophyll, and submarine light conditions in Kings Bay, Spitsbergen in July 1971." Norw. J. Bot. 20: 99-108. Heaney, S.I. 1978. "Some observations on the use of the in vivo fluorescence technique to determine chlorophyll-a in natural populations and cultures of freshwater phytoplankton." Freshwater Biol. 8: 115-126. 110 Holm-Hansen, 0., C.J. Lorenzen, R.W. Holmes and J.D.H. Strickland. 1965. "Fluorometric determination of chlorophyll." J. Cons. Perm. Int. Explor. Mer 30: 3-15. Humphrey, G.F. 1963. "Phytoplankton pigments in the Pacific Ocean," p. 123-141. In M.S. Doty (ed.), Proc. Conf. on Primary Productivity Measurement, Marine and Freshwater. U.S. Atomic Energy Comm. Rep. TID-7633. Jeffrey, S.W. 1974. "New spectrophotometric equations for determining chloro- phylls a, b, c\ and C2 in algae, phytoplankton and higher plants." CSIRO Marine Biochemistry Unit, p. 6-8. Jeffrey, S.W. , and G.F. Humphrey. 1975. "New spectrophotometric equations for determining chlorophylls _a, _b, cj and cj? in higher plants, algae and natural phytoplankton." Biochem. Physiol. Pflanzen 167: 191-194. Kiefer, D.A. 1973a. "The in vivo measurement of chlorophyll by fluorometry," p. 421-430. In L.H. Stevenson and R.R. Colwell (eds.), Estuarine Microbial Ecology, Belle W. Baruch Library in Marine Science 1, Univ. So. Carolina Press. Kiefer, D.A. 1973b. "Chlorophyll-a fluorescence in marine centric diatoms: Responses of chloroplasts to light and nutrient stress." Mar. Biol. 23: 39-46. Kiefer, D.A. 1973c. "Fluorescence properties of natural phytoplankton popu- lations." Mar. Biol. 22: 263-269. Kiefer, D.A. , 0. Holm-Hansen, C.R. Goldman, R.C. Richards and T. Berman. 1972. "Phytoplankton in Lake Tahoe: Deep-living populations." Limnol. Oceanogr. 17: 418-422. Kimmel, B.L., and M.M. White. In press. "DCMU-enhanced chlorophyll fluores- cence as an indicator of the physiological status of reservoir phytoplankton: An initial evaluation," p. 246-262. _In M. Lorenzen (ed.), Reservoir Phyto- plankton Interactions, Proc. Corps of Eng. Waterways Exper. Sta. Workshop, April 1979, Monterey, Ca. Krey, J. 1958. "Chemical methods of estimating standing crop of phytoplankton.' Rapp. Proc. -Verb. , Cons. Int. Perm. Explor. Mer. 144: 20-27. Loftus, M.E., and H.H. Seliger. 1975. "Some limitations of the in vivo fluor- escence technique." Chesapeake Sci. 16: 79-92. Lorenzen, C.J. 1966. "A method for the continuous measurement of _in vivo chlorophyll concentration." Deep-Sea Res. 13: 223-227. Lorenzen, C.J. 1967. "Determination of chlorophyll and phaeopigraents : Spectrophotometric equations." Limnol. Oceanogr. 12: 343-346. Marker, A.F.H. 1972. "The use of acetone and methanol in the estimation of chlorophyll in the presence of phaeophytin. "Freshwater Biol. 2: 361-385. Ill McMurray, G. 1978. "In vivo fluorescence studies in San Francisco Bay, California." Abstr. , Amer. Soc. Limnol. Oceanogr. Mtg. , June 1978, Victoria, B.C. Moss, B. 1967. "A spectrophotometry method for the estimation of percentage degradation of chlorophylls to pheo-pigments in extracts of algae." Limnol. Oceanogr. 12: 335-340. Papageorgiou, G. 1975. "Chlorophyll fluorescence: An intrinsic probe of photosynthesis," p. 317-371. In Govindjee (ed.), Bioenergetics of Photo- synthesis, Academic Press. Parsons, T.R. , and J.D.H. Strickland. 1963. "Discussion of spectrophotometric determination of marine plant pigments, with revised equations for ascer- taining chlorophylls and carotenoids." J. Mar. Res. 21: 155-163. Piatt, T. , L.M. Dickie and R.W. Trites. 1970. "Spatial heterogeneity of phyto- plankton in a near-shore environment." J. Fish. Res. Bd. Canada 27: 1453-1473 Powell, T.M. , and others. 1975. "Spatial scales of current speed and phyto- plankton biomass fluctuations in Lake Tahoe." Science 189: 1088-1090. Prezelin, B.B. 1978. "Relationship between photosynthetic capacity and fluor- escence in laboratory and field populations of phytoplankton." Abstr., Amer. Soc. Limnol. Oceanogr. Mtg., June 1978, Victoria, B.C. Prezelin, B.B., and A.C. Ley. In press. "Photosynthesis and chlorophyll a^ fluorescence rhythms of marine phytoplankton." Mar. Biol. Prezelin, B.B., and B.M. Sweeney. 1977. "Characterization of photosynthetic rhythms in marine dinoflagellates. II. Photosynthesis-irradiance curves and in vivo chlorophyll a^ fluorescence." Plant. Physiol. 60: 388-392. Richards, F.A. , and T.G. Thompson. 1952. "The estimation and characterization of plankton populations by pigment analysis. II. A spectrophotometric method for the estimation of plankton pigments." J. Mar. Res. 11: 156-172. Samuelsson, G. , and G. Oquist. 1977. "A method for studying photosynthetic capacities of unicellular algae based on in vivo chlorophyll fluorescence." Physiol. Plant. 40: 315-319. Samuelsson, G. , G. Oquist and P. Halldal. 1978. "The variable chlorophyll ji fluorescence as a measure of photosynthetic capacity in algae. Mitt. Int. Verin. Limnol. 21: 207-215. Shoaf, W.T., and B.W. Lium. 1976. "Improved extraction of chlorophyll ji and b from algae using dimethyl sulfoxide." Limnol. Oceanogr. 21: 926-928. Slovacek, R.E. 1978. "In vivo chlorophyll a_ fluorescence measurements." Abstr., Amer. Soc. Limnol. Oceanogr. Mtg., June 1978, Victoria, B.C. 112 Slovacek, R.E. 1979. "In vivo chlorophyll a_ fluorescence as an indicator of phytoplankton photosynthesis." Abstr. , Amer. Soc. Limnol. Oceanogr. Mtg. June 1979, Stony Brook, N.Y. Slovacek, R. E. , and T. Bannister. 1973. "NH4CI activation of the fluorescence yield in CO?-starved Chlorella pyrenoidosa. " Biochim. Biophys. Acta 325: 114-119. Slovacek, R.E., and P.J. Hannan. 1977. "In vivo fluorescence determinations of phytoplankton chlorophyll a_. " Limnol. Oceanogr. 22: 919-924. Strickland, J.D.H., and T.R. Parsons. 1972. "A practical handbook of seawater analysis," 2nd ed. , Fish. Res. Bd. Canada Bull. 167. UNESCO. 1966. "Determination of photosynthetic pigments in seawater." Monogr, on Oceanographic Methodology 1. Vincent, W.F. 1979. "Mechanisms of rapid photosynthetic adaptation in natural phytoplankton communities. I. Redistribution of excitation energy between Photosystems I and II." J. Phycol. 15: 429-434. Vollenweider , R.A. (ed.). 1974. "A manual on methods for measuring primary production in aquatic environments." IBP Handbook 12. Blackwell. White, M.M. 1980. "In vivo chlorophyll fluorescence as an estimator of chlorophyll a. in natural phytoplankton communities." Abstr., Amer. Soc. Limnol. Oceanogr. Mtg., Los Angeles, Ca. Yentsch, C. , and D. Menzel. 1963. "A method for the determination of phyto- plankton chlorophyll and phaeophytin by fluorescence." Deep-Sea Res. 10: 221-231. 113 CH, II /'"\ CM, H {CH.J- ^CHO K/\A •N N \ / Mg 11 j>-C,H, •A IV \ \ — H III /N 10 9 H,C H h-1- 5^— CH, V 1 ' O CH. COOCH, I c=o I o CH, i CH II H,C-CS H,C / CH, H.C-HC H»C, H,C-HC H>C, H,C-HC ^CH, OH, t ;Ch, 'CH, kCH, Figure 1. The chlorophyll a molecule 114 A: UJ UJ O UJ . x: P- (sdo) 30N30S3UOmd HOIHO OAIA Nl (lUd) X30NI (.-JM .-I 0 BTf) 3SNOdS3d '0S3aonid 3*\/idn noquvo \\\ i-. O O o 3DN3DS3b0mj to c c C u fc. 116 100 CO o. 3 80 o — o EXPER. I (y = l.45x+ 10.02; r*=0.98) 1 1 i 1 1 1 10 20 30 CARBON UPTAKE (fxq C h' hr-') Figure 5. Photosynthetic carbon fixation effect: on flourescence response index LAKE TEXOMA (OKLA.- TEXAS): 26 JULY 1978 % SURFACE LIGHT 0.1 I 10 100 UQ MICROBIAL C I-' 0 250 500 750 I I I I DISSOLVED 02 (mgl-) M9 P LITER"' mg C m-»hr-' ) 4 8 100 200 'N VIV0 CHL0R' 0 ■ i i i i FLUORESCENCE i_ 25 50 75 TEMPERATURE (°C) 0 10 20 UQ N LITER-1 (COUNTS SEC"1) FLUOR. RESPONSE INDEX 200 400 100 200 0.2 0.4 0.6 Figure 6. Difference in phytoplankton physiological status related to differing conditions of light and nutrient availability 117 BROKEN BOW LAKE (OKLA.)'. 14 AUG. 1978 % SURFACE LIGHT O.I I 10 100 I I I I ug MICROBIAL C |-' 0 100 200 300 I l i l DISSOLVED 02 (mg I-') IN VIVO mg C m-» hr-' 0 4 8 CHLOROPHYLL 0 4 8 12 i 1 1 1 FLUORESCENCE i 1 1 1 TEMPERATURE (°C) 0 10 20 (COUNTS SEC-') FLUOR. RESPONSE INDEX 25 50 0.2 0.4 0.6 1 I I l^_oO I .A— L Figure 7. Difference in phytoplankton physiological status related to difference condtions of vertical mixing patterns. 118 METABOLIC RESPONSE OF FISH TO STRESS1 by Dr. Donald E. Wohlschlag The University of Texas Marine Science Institute Port Aransas, Texas 78373 INTRODUCTION The fact that humans are primarily terrestrially oriented leaves the aquatic environments, especially marine water, out of sight and out of mind altogether too often in terms of recognizing effects of, and coping with, waste disposal. The lack of public awareness of effects of wastes on marine organisms is common- place. Indirect observations of stresses on marine biota are often recognized only after major damage has occurred — such as when a fishery fails. When a fishery failure, for example, has been recognized, it is difficult to decide whether the causative factors involved overfishing, pollution, naturally fluc- tuating fish populations, or naturally fluctuating environmental variables. Early warning and monitoring systems are needed. At the outset of a marine pollution study there is usually little definition of the extent of toxicity, although a facile assumption may be made that mobile organisms like fish would simply swim to less toxic areas. Unfortunately there is little direct evidence for this behavior except in laboratory studies; more likely, there is little evidence to indicate whether fishes have the sensory mechanisms to perceive anthropogenic materials in the first place and whether they would react solely, and negatively, to such materials in the second place. Many other favorable environmental factors like availability of food, spawning grounds, etc., might well mask unfavorable factors such as low levels of toxic materials. These types of organismal, population, or ecosystem response problems exist over and above the usual problems of pollution detection. Synthesizing existing information and providing techniques for the solution of specific problems of toxicity in open natural systems have been and remain elusive (NAS 1975). Historically, pollutants have first been identified after some environmental — usually biological — adversity had become obvious, and then the nature of the pollutant and its biotic effects would have been determined. The delay in assessing effects of pollutants on biological systems and the primary emphasis on identification of the pollutant has been the subject of some concern (NAS 1975, Brungs and Mount 1978, among others). To cope with these problems many authors and some agencies have suggested various schemes for biomonitoring, e.g. Cairns, et al. (1973a, b) or Cairns, et al. (1979). The magnitude of problems of estimating hazards has been empha- sized by Cairns (1980). With this type of emphasis, we can turn to a simple method of measuring what fishermen might perceive early on when fish are subjected to stresses. ^The University of Texas Marine Science Institute Contribution No. 000 119 Quite often fishermen recognize subtle adverse environmental effects on fishes long before acute effects are obvious and are generally recognized by various governmental agencies concerned with environmental quality. Why is this possi- ble? Why should not whatever aspects of the fish or the environment that the fisherman can perceive as stress be utilized in an analytical manner? The answers to these questions turn out to be rather straightforward. And the answers to how stress effects can be recognized can provide a rather elegant, quantitative method of bioassay by utilizing fishes as sensors of sublethal stress. Essentially the method is one of evaluating the decline with stress in metabolic output for sustained swimming performance, which was proposed by Fry (1947) and originally utilized by him and others chiefly to determine effects of thermal stresses (Fry, 1957, 1971). Emphasis throughout will be on sublethal stress effects. Acute stresses that produce mortality rates as high as those of 96 hr LC50 evaluations will not be discussed, for the reason that even at 1% of such high mortality rates, the natural mortality rates of many longer lived fish species would be excessive for population survival. The sublethal stress effects should be recognizable and measurable with precision at about 1% of the LC50 level or less. Techniques should be adaptable to open environmental systems and to endemic fishes as well as being adaptable to laboratory or treatment plant situations. Most importantly emphasis should be on a technique that does not necessarily depend on a priori knowledge of the chemical nature of toxicants. The measurement of subtle, sublethal stress effects should be sufficiently easy, yet sensitive enough so that true predict ivity is possible long before any acute damage would be evident. Measurements should be sensitive for assay purposes; they should be easy in the sense that complicated procedures requiring extensively trained personnel and costly equipment are avoided; they should be easy also in the sense that interpretation of results are straightforward and conceptually in line with pertinent real-world biotic and environmental features. PURPOSE The purpose of this presentation is to show by selected examples and general theoretical considerations how fish can be used to 1. Monitor consistency of water quality; 2. Establish optimal natural environmental conditions; 3. Assess and monitor normal, or sublethal departures from normal, conditions with respect to optimal natural environments; 4. Assess and monitor deleterious sublethal stress effects of toxic materials; and 5. Predict long term outcome of natural or anthropogenic stresses at sublethal levels. 120 With these aims in mind, the term "natural environment" may mean "ambient environment" if man-induced environmental modifications persist. WHY UTILIZE FISHES? The use of fishes for biological assays and for monitoring involves ecological considerations of both theoretical and practical importance. Examples of these considerations are noted below. 1. Fishes are widely known to the public at large as commercial and recreational resources. 2. Fishes are usually much better known taxonomically than other groups of organisms. 3. Among fishermen and biologists there is a considerable amount of "native wisdom" on the nature of fishes and their sensitivity to environmental variables. 4. There is a large literature on the highly sensitive reactions of fishes to natural and anthropogenic environmental factors, although few baselines have been established. 5. Fishes are useful comparatively in assessing environmental conditions because, more than almost any other taxonomic group, they are found in nearly all marine environments. 6. Throughout the world fishes tend to have rather similar physiological systems that can be compared among themselves with reference to adapta- tional propensities to specific environments; the ubiquitous distribution of marine fishes implies that they can be compared from one type of envi- ronmental regime to the next in terms of physiological characteristics. 7. Fishes in a given environment tend to have an ecological stability that assures survival over relatively long periods and at relatively constant numerical and biomass levels, compared to most other aquatic organisms. These levels can naturally vary usually less than, say, one order of magni- tude over periods of decades, whereas short-lived species, like micro- organisms at lower ecotrophic levels can naturally vary, say, ten orders of magnitude in a few hours or days in response to natural environmental changes. 8. Because most fish species are at the higher ecotrophic levels and tend to have relatively stable populations, their stabilizing and integrating effects on the overall natural ecosystems are most likely highly important. It is generally known — at least by hindsight — that all the above considera- tions are affected by both acute and chronic stresses on fish populations, whether these stresses are natural or anthropogenic. 121 BASIC THEORY Fry (1947) noted that the respiratory metablic scope for activity is a sen- sitive measure of well being in the case of natural variables like temperature. Metabolic scope is simply the difference between the aerobic metabolic rate at sustained swimming speeds and the rate at a maintenance, or standard, level (fig. 1). Respiratory metabolism is most commonly measured by determining the rate of consumption of dissolved oxygen by fish in a variety of ingeniously constructed chambers. Fry (1971) illustrates some of the more common types ot respiration chambers and discusses many of the techniques for the determination of respiratory rates of both standard and active levels. Determination of respiration rates at standard conditions involves many known pitfalls, most of which are avoidable with relatively simple precautions. These precautions are usually to assure that the rates are measured on fasted fish and are the lowest possible rates compatible with survival for several days. Fry (1957, 1971) and other authors have covered this subject extensively. The standard rate can not decrease, but it can increase if energy costs for internal regulation are required for ordinary natural stresses like temperature and sal- inity. The standard level may or may not increase with anthropogenic stresses, presumably depending whether or not the body "machinery" has any available adaptational mechanisms for the particular stress. The metabolic rate at maximum sustained swimming activity of a fish usually involves the oxygen consumption rate measurements for a speed of swimming that can be maintained without fatigue for long periods, say, 200 minutes or longer according to Beamish (1978) in a review of the entire subject of swimming capacity. Observations on Gulf coastal fishes reveal that with conditioning at about 1 length/ sec for about 12 hours and gradually increasing the swimming speed of a fish to its maximum in about 10 cm/sec increments each 5 min would produce highly reproducible results. Further if fish swam at even a few cm/sec over a maximum sustained rate, they would either tend to tire (and probably build up an "oxygen debt" with attendant anaerobic respiration) or they would initiate "burst-and-glide" swimming. However, after performing at the maximum rate for several hours, the same aerobic performance could be repeated, but not exceeded, with a short rest and no evidence of fatigue. The repeatability and the fact that the maximum sustained swimming performance has a sound basis in both physiological and hydrodynamic theory (Webb 1975, Webb 1978, and Magnuson 1978) make this measurement exceptionally useful. From figure 1, general theory of energetics and physiological performance, and practical experience, the logical possibilities of sublethal stress effects reduce the scope for activity and involve raising the standard level, or lower- ing the sustained active level, or both. Recently, Wohlschlag and Parker (ms) noted that diluted pharmaceutical industry wastes could slightly elevate active metabolic levels; although the swimming rates declined substantially, the standard metabolism increased and the fish showed incipient morbidity after about 2 days. This kind of elevated metabolism brought about by stimulatory effects of drugs, but not accompanied by locomotor activity, is possibly a kind of "spontaneous activity" (Fry 1957). 122 Spontaneous activity must be eliminated for the measurement of standard rates; but it ordinarily is trivial if fish will swim at maximum sustained rates. Several other higher levels of swimming activity and corresponding metabolic rates that should be mentioned are the shorter duration (20 sec - 200 min) higher prolonged rates and the very short duration (less than 20 sec) very high burst rates. Usually averaging less than the maximum sustained active rates are the routine rates that apply to foraging, schooling, migrating, station holding, and other swimming requirements. These other rates are defined and discussed in detail by Beamish (1978), of which only routine rates are important to the subject at hand inasmuch as they also tend to be depressed by stresses (Beamish 1964 and Wohlschlag and Cameron 1967). TECHNIQUES There is a large literature describing a variety of devices used for measuring the respiratory metabolic (oxygen consumption) rates of fishes at various stages of quiescence to extreme burst activities. A recent volume on the subject (Hoar and Randall, eds. 1978) has a number of chapters on technical and conceptual aspects of locomotion. Webb (1975) and Aleyev (1977) consider many hydromechani- cal and technical features. For this discussion, techniques used recently with Gulf coast fishes for the determination of metabolic scope will be emphasized (Wohlschlag and Wakeman 1978). Acclimation — Fish used for metabolic measurements should be acclimated at a constant temperature and salinity for at least two days with fasting at a range of 15-30°C and about 10-50 ppt salinity with no 2-day change much over a 5° C or 10 ppt salinity change from ambient to experimental conditions. At colder temperatures, longer periods are desirable. Fish used for activity measurements and for sublethal toxic material responses should be held for another two days under about 1 L/sec swimming acclimation or with the toxic material. (For toxic materials a 1% dilution of an LC50 level, if known, is a good rule-of- thumb initial concentrationl) Standard Metabolic Rates — Standard oxygen consumption rates can be determined directly by holding the fish in partially darkened and sealed containers that are protected from visual and other disturbances, starting with measured oxygen concentrations and continuing until the level would be no lower than about 70% saturation. Another method is to utilize appropriate sizes of tubes that are closed at each end, except for inlet and outlet tubes through which flows water whose oxygen contents and flow rates are constantly monitored. In this and the static system, the lowest rates consistently observed on a diel cycle correspond to the standard rates, but only if the fish are absolutely quiescent and show no evidence of spontaneous (nonlocomotory) activity. A third method is to plot oxygen consumption rates (usually per kg of fish weight) at various activities and calculate the resting rate at zero velocity. (See methodology below.) If the regression data show the usual scatter, a line drawn parallel to this regression through the lowermost points and extrapolated to zero velocity will usually be a better estimate of the standard rate. This method is known as the Brett (1964) method. 123 Routine Metabolic Rates — These rates can be measured in an annular chamber that is rotated at whatever velocity the fish will swim consistently for a period of an hour or so. Oxygen content of the chamber should be measured initially and about each 15 min thereafter, but the minimum level should be no lower than about 70% for most active species. This method follows that of Wohlschlag and Juliano (1959), Wohlschlag and Cameron (1967), Wolschlag, et al. (1968) and Wohlschlag and Cech (1970) in which the chamber is immersed in the large tank where the fish have been acclimated. The technique obviates the disturbing effects of transferring fish from tank to metabolism chamber. Swimming speeds are determined by counting the revolutions and multiplying by the approximate swimming track circumference of the chamber over a given time period. Usually, the actively swimming species will swim at a routine rate about equal to 1 L/sec, which tends to be equivalent to a common speed for schooling, foraging, etc., although by rotating the chamber at higher and higher speeds, some species may be induced to maintain much higher velocities. Com- parisons of routine rates with standard and active rates are in Wohlschlag and Wakeman (1978). Routine scope can also be utilized (Beamish 1964). Active Metabolic Rates — Metabolic rates at maximum sustained swimming veloci- ties (active rates) can be measured by a variety of respiratory chambers. Some of the more successful types are the rotating annular or toroidal types, the Brett swimming tunnels (Brett 1964), and the Blazka swimming tunnels (Blazka, et al. 1960). Fry (1971) illustrates some of these types of respirometers. For the studies described here a large Blazka-type chamber was constructed in such a way that it could be suspended in the water used for acclimation of the fish. Details are in Wohlschlag and Wakeman (1978) and in Wakeman and Wohlschlag (1978), from which the schematic figure 2 shows the essential working relation- ships. The suspension of the chamber on a trolley and the quick-disconnect feature for addition and removal of fish is a major feature. Direct measurement of swimming speeds is by utilizing a paddle type of electronic transducer speed- ometer inserted into the respiratory chamber access tubes. Indirect measurement of swimming speeds is by tachometer readings of the impeller shaft rotational speed which have been calibrated to direct measurements. Swimming speeds depend also on calibrations for water density that varies with temperature and salinity. Calculations — Because experience has indicated that the weight of the fishes contributes about 80% of the variability to oxygen consumption rate measurements when stepwise regression procedures are employed, it is usually not desirable simply to express metabolic rates as O2 consumed per unit weight per hr as is customary in many laboratory studies. Also, because it may be desirable to meas- ure metabolism over a range of body weights, swimming velocities, temperatures, salinities or other variables, the use of multiple regressions is indicated in the form where Y = a + bwXw + bvXv + . . . + b±X± + . . . , Y = expected rate of O2 consumption in logjo m8 02 /hr; a = constant 124 Xw, Xv and other terms are independent variables, respectively for logio weight in grams, for swimming rate in L (cm)/sec or in L^'^/sec whichever may be linear, etc. for other variables; and bw and bv are the respective partial regression coefficients. This procedure is described in most statistical manuals (e.g. Snedecor and Cochran 1967) or pretested library computer routines; it is the type of procedure used by Wohlschlag and Wakeman (1978) in which salinity Xs was used at higher orders. The procedure enables comparison at equivalent weight, activities, etc. providing extrapolation is not carried beyond the range of the observations. EXAMPLES OF STRESS EFFECTS ON METABOLIC SCOPE Temperature — Scope tends to be maximal at the optimal temperature, which for the spotted sea trout in figure 3 is probably around 23°C, although for many species scope may increase almost to the upper lethal level. Excellent reviews of temperature effects on fish physiology are in Brett (1970, 1971). Salinity — Scope for the spotted sea trout, a coastal and bay species, can be described over a salinity range by the data of Wohlschlag and Wakeman (1978). Quite obviously in figure 4 the greatest scope occurs around 20-25 ppt salinity which represents a salinity optimum for juvenile to adult fishes. Note the one point at 30, and perhaps 35, ppt salinity where the active metabolic rate is low; this rate is most likely caused by poor growth and partially starved fish. Note also that the routine metabolic rate (at a swimming speed of about 1 L/sec over the salinity range) is equal to the active rate at extremes of about 10 and 45 ppt salinity. This is the salinity range where small to large spotted seatrout can be caught in the south Texas area. Also the species has not yet been successfully cultured in freshwater situations as have some of the other sciaenids. In figure 5 , the optimum of the scope corresponds almost exactly with the maximum swimming rate. This observation suggests that direct observations of the maximum swimming rates could be utilized for monitoring purposes. Fishermen generally recognize the fact that healthy fish swim more vigorously than those under prolonged stress. Growth — After discovering that the active metabolic rates and the scopes were unusually low (Fig. 4) at 30 ppt salinity, it became obvious that the spotted seatrout for this portion of the experiment had been unusually slender. The fact that this species in mid to late summer in this area was not commercially marketable and the fact that the weight-length regressions of the experimental fish also indicated that they were considerably more emaciated as they became larger indicated their poor condition as in the figure 6, which shows the metabolic characteristics of the fish with a very poor condition compared to the average healthy fish at well fed and rapidly growing conditions. Actually under chronic stress, one of the first manifestations to be noticed in a fishery is that the so-called condition of the fish declines. This problem has been discussed by Cech and Wohlschlag (1975). 125 Red Tide Exposure — In one case of a spotted seatrout study, waters from Nueces Bay were transported to the Port Aransas Laboratory for comparison of routine metabolic performance. After this study had started there was evidence that a small red tide (Gonyaulax monilata) bloom had occurred in the Corpus Christi ship channel and some considerable amounts of this water, used as cooling water, had been discharged into Nueces Bay near the water collection site. As indicated in figure 7, the standard level increased in this water, so that if the active level, which was not measured then, was assumed to be at the same level as that for control fish (studied a short time later), the scope would be reduced as indicated. Unknown Chronic Pollution — Samples of water from four Galveston Bay stations were periodically transported to the Port Aransas Laboratory for the study of striped mullet routine metabolism during 1971-72 (Wohlschlag, et al. 1973). The results of these measurements compared to those made for fish in "clean" Port Aransas area water are presented in figure 8. In all cases, the routine levels were slightly depressed in the Galveston Bay waters, which in spite of proximity to the highly polluted Houston Ship Channel are among the more pro- ductive coastal waters of the northwestern Gulf of Mexico. Ocean Dumped Petrochemical Waste — Figure 9 shows how a low dilution of 0.2% v/v (about 1% of LC50 levels) of a well-weathered and deactivated waste from the biosludge ponds at the Shell Plant in Deer Park, Texas affected the meta- bolic scope of the red snapper at 20° and 28°C. Both liquid and sludge portions were evaluated at 20°C (Wohlschlag and Parker 1980, in press). The Gulf dumping program ceased before the procedures could be refined to utilize longer exposure periods at much lower concentrations. It would be desirable to investigate this problem with potential pollutants at levels below those considered as legally toxic by present standards. Ocean Dumped Pharmaceutical Waste — In figure 10 are some results of an experi- ment to ascertain the scope diminishing effects of partially decomposed compos- ite waste of the type dumped north of Puerto Rico (Wohlschlag and Parker ms). The waste held and transported with considerable delay at high summer tempera- tures was slightly more noxious in glass contaners than in polyethylene con- tainers. An additional series of experiments was conducted with the Puerto Rican com- posite and the individual pharmaceutical wastes at fairly low concentrations and exposures 2 hours or less, followed by a recovery period in clean water for several days (fig. 11). During the recovery period, average active metabolic rates and scope values were determined. In all cases the standard level tended to increase, and the swimming rate tended to decrease considerably, but the active metabolic rate in case of some of the individual pharmaceuticals was apparently stimulated to the extent that most of the fish showed signs of mor- bidity after about two days of the recovery period. This type of experiment needs further investigtation at lower exposure concentrations and times. Other- wise the experiments with longer times and still lower concentrations than those illustrated in figure 10 should be conducted for any persistent pollutant, whether or not its chemical nature is understood at the outset. 126 SUGGESTED PILOT MONITORING PROGRAM From the preceeding data it is clear that where there may be even very slight persistent stresses, the metabolic scope of fishes and their swimming perform- ances decline in proportion to the stress and all its associated chronic effects. Further, it is important to note that this technique has a very firm physiologi- cal basis and requires no a priori knowledge of the chemical nature of toxic materials or the biochemistry of toxicological effects on organisms. It would be possible in a monitoring operation to utilize more or less con- tinuous operations to measure swimming rates and general swimming performances both in flumes that contain clean water and in flumes to which suspected toxi- cants could be added for uniform concentrations. For metabolic rate determina- tions, there should be no problem to divert waters from the flumes to enclosed Blazka-type chambers for active metabolic rate determinations or to flow- through/static tubular chambers for standard metabolic rate determinations. (See the Cairns references.) All that would be required is for the Blazka chamber end dome opposite the dome that contains the impeller to be equipped with an inflow port and valve. By surrounding the Blazka and static chambers with the experimental or control waters, fish could be continuously acclimated and ready for experimentation at any time. It should be emphasized that such methods of holding the fish also permit the type of monitoring suggested by Cairns and his coworkers (Cairns, Sparks, and Waller 1973; Cairns, Dickson, and Lanza 1973; and Cairns and Gruber 1979) because much of the same apparatus can be used. A most useful feature of the equipment for continuous observations of swimming rates and the equipment for holding the fish during the standard metabolic rate measurements is that a large category of behavioral observations can also be made. Many behavioral observa- tions undoubtedly can be directly associated in quantitative terms with the degree of stress and depression of metabolic scope and swimming performance. Perhaps the most biologcally useful attribute of performance energetics is the pertinence of their measurements to the general bioenergetics of fish popu- lations. Recent references would include discussion by Webb (1978), Brett (1979), and Brett and Groves (1979). From these references, it will be clear that meta- bolic scope and swimming performance data used for biomonitoring or stress assessment all tie in with the generalizing concepts that bioenergetics embrace. ACKNOWLEDGMENTS The various aspects of this report have been based on research supported by the National Science Foundation, The Texas Water Quality Board, The Texas Water Development Board (now the Department of Water Resources), and the National Ocean Survey Ocean Dumping Program of the National Oceanic and Atmospheric Administration. Their administrative and financial support have been most important for the development and continuation of these studies. The aid of dozens of graduate students, technicians, and other colleagues in various phases of the research in both field and laboratory is gratefully ack- nowledged. The help in assembling and typing the report by Helen Garrett is most appreciated. 127 REFERENCES Aleyev, Yu. G. 1977. "Nekton." Dr. W. Junk, The Hague, 435 pp. Beamish, F.W.H. 1964. "Respiration of fishes with special emphasis on standard oxygen consumption." Canadian J. Zool. 42:177-188. Beamish, F.W.H. 1978. "Swimming capacity." pp 101-187. J_n: Fish Physiology, Vol. 7. "Locomotion." W.S. Hoar and D.J. Randall (eds.). Academic Press, New York, San Francisco, London. Blazka, P., M. Volf, and M. Cepela. 1960. "A new type of respirometer for the determination of metabolism of fish in the active state." Physiol. Bohemoslov, 9: 553-558. Brett, J.R. 1964. "The respiratory metabolism and swimming performance of young sockeye salmon." Journal of Fisheries Research Board of Canada, 21:1183-1226. Brett, J.R. 1970. "Temperature, Animals. Fishes," p. 516-616. _In: 0. Kinne (ed.), Marine Ecology, Vol. 1, Environmental Factors, Part 1, Wiley- Interscience, Ltd., London, New York, Sydney, and Toronto. Brett, J.R. 1971. "Energetic responses of salmon to temperature. A study of some thermal relations in the physiology and freshwater ecology of sockeye salmon" (Oncorhynchus nerka). Am. Zoologist 11:99-113. Brett, J.R. 1979. "Environmental factors and growth." pp. 599-675. _In: W.S. Hoar, D.J. Randall, and J.R. Brett, eds. "Fish Physiology," v. 8, Bioener- getics and Growth. Academic Press, New York, San Francisco, London. Brett, J.R. , and T.D.D. Groves. 1979. "Physiological energetics," pp. 279- 352. _In "Fish Physiology," Vol. 8, W.S. Hoar, D.J. Randall, and J.R. Brett, eds. Academic Press, New York, San Francisco, London. Brungs, W.A. , and D.I. Mount 1978. "Introduction to a discussion of the use of aquatic toxicity tests for evaluation of the effects of toxic substances," pp. 15-32. J_n: "Estimating the Hazard of Chemical Substances to Aquatic Life" (Cairns, J. Jr., K.L. Dickson, and A.W. Maki, eds.) American Society for Testing and Materials, STP 657, Philadelphia. Cairns, John, Jr. 1980. "Estimating hazard." Bioscience, 30: 101-107. Cairns, J., R.E. Sparks, and W.T. Waller. 1973. "A tentative proposal for a rapid in-plant biological monitoring system," pp. 127-147. "In: Biological Methods for the Assessment of Water Quality," J. Cairns, Jr. and K. L. Dick- son, eds. Am. Soc. for Testing and Materials, Special Tech. Publ. 528. Philadelphia. Cairns, J., K.L. Dickson, and G. Lanza. 1973. "Rapid biological monitoring systems for determining aquatic community structure in receiving systems." 128 pp. 148-163. _In: "Biological Methods for the Assessment of Water Quality," J. Cairns, Jr. and K.L. Dickson, eds. Am. Soc. for Testing and Materials, Special Tech. Publ. 528. Philadelphia. Cairns, J., Jr., and D. Gruber. 1979. "The coupling of mini- and micro- computers to biological early warning systems." Bioscience 29:665-669. Cairns, J., Jr., G.P. Patil, and W.E. Waters (eds.). 1979. "Environmental biomonitoring, assessment, prediction, and management — certain case studies and related quantitatiave issues." International Co-operative Publishing House, Burtonsville, Maryland. 468 pages. Cairns, John, Jr., and David Gruber. 1980. "A comparison of methods and instru- mentation of biological early warning systems." American Water Resources Bulletin, 16(2) :261-266. Cech, J. J. , Jr. and D.E. Wohlschlag. 1975. "Summer growth depression in the striped mullet," Mugil cephalus L. Contr. Mar. Sci. 19:91-100. Fry, F.E.J. 1947. "Effects of the environment on animal activity." Univ. Toronto Studies Biol., Ontario Fish. Res. Lab., 68:1-62. Fry, F.E.J. 1957. "The aquatic respiration of fish," pp. 1-63. Jin: M.E. Brown (ed.), The Physiology of Fishes. Acadmic Press, New York. Fry, F.E.J. 1971. "The effect of environmental factors on the physiology of fish," p. 1-98, J_n: W.S. Hoar and D.J. Randall (eds.) Fish Physiology, Vol. 6, Environmental Relations and Behavior, Academic Press, New York and London. Hoar, W.S., and D.J. Randall (eds.) 1978. "Fish Physiology," Vol. VII, Loco- motion. Academic Press, New York, San Francisco, London, xviii + 576 pp. Magnuson, J.J. 1978. "Locomotion by scombroid fishes: hydromechanics, morphol- ogy, and behavior." pp. 239-313." _In: "Fish Physiology," v. 7, W.S. Hoar and D.G. Randall, eds. Academic Press, New York, San Francisco, London. National Academy of Sciences. 1975. "Principles for Evaluating Chemicals in the Environment." Washington, D.C. Wakeman, J.M. , and D.E. Wohlschlag. 1978. "Salinity stress and swimming per- formance of spotted seatrout." Proc. Ann. Conf. S.E. Assoc. Fish & Wildlife agencies 31: 357-361 Webb, P.W. 1975. "Hydrodynamics and energetics of fish propulsion." Bull. Fish. Res. Bd. Canada 190. x + 158 pp. Webb, P.W. 1978a. "Hydrodynamics: nonscombroid fish," pp. 189-237. _In W.S. Hoar and D.G. Randall, eds. Fish Physiology, V. 7. Academic Press, New York, San Francisco, London. Webb, P.W. 1978b. "Partitioning of energy into metabolism and growth," pp. 184- 214. _In S.D. Gerking, ed. Ecology of Freshwater Fish Production. Wiley, New York. 129 Wohlschlag, D.E. 1973. "Respiratory metabolism of the striped mullet as an assay of low level stresses in Galveston Bay." pp. IV 1-77, C.H. Oppen- heimer, W.B. Brogden, and K.G. Gordon, eds. Toxicity Studies of Galveston Bay. Rept. Texas Water Qual. Bd. IAC (72-73) 183. Wohlschlag, D.E. 1977. "Analysis of freshwater inflow effects on metabolic stresses of South Texas bay and estuarine fishes: Continuation of exten- sion." Report to Texas Water Development Board, IAC (76-77)-1690. 106 pp. processed. Wohlschlag, D.E. and R.O. Juliano. 1959. "Seasonal changes in bluegill metabolism." Limnol. Oceanogr. 4:195-209. Wohlschlag, D.E. and J.N. Cameron. 1967. "Assessment of a low level stress on the respiratory metabolism of the pinfish (Lagodon rhomboides). Contr. Mar. Sci. 12: 160-171. Wohlschlag, D.E. , J.N. Cameron, and J.J. Cech. 1968. "Seasonal changes in the respiratory metabolism of the pinfish (Lagodon rhomboides) . " Contr. Mar. Sci. 13:89-104. Wohlschlag, D.E., and J.J. Cech. 1970. "Size of pinfish in relation to thermal stress response." Contr. Mar. Sci. 15:22-31. Wohlschlag, D.E. and John M. Wakeman. 1978. "Salinity stresses, metabolic responses and distribution of the coastal spotted seatrout, Cynoscion nebulosus. " Contr. Mar. Sci. 21:171-185. Wohlschlag, D.E. and F.R. Parker, Jr. 1980. "Metabolic sensitivity of fish to ocean dumping of industrial waste." In Press, B.H. Ketchum, D.R. Kester, and P.K. Park, eds. "Ocean Dumping of Industrial Wastes." Plenum Press. 130 ^ BURST <20 sec a a z UJ UJ Ql UJ a. X if) UJ e> >- z 200 min. (MAXIMU M , SUSTAINED) ROUTINE (FORAGING, STATION HOLDING, ETC.) STANDARD (COMPLETE REST ) Fig. 1. Relationships of energy levels and swimming speed levels. Note that scope can only diminish with stress because either the standard (minimum maintenance) level increases, or the active level decreases, or both. 131 Fig. 2. Schematic diagram of 207-liter modified Blazka respirometer. M — variable speed, constant torque 10 h.p. motor; B — flow linearizing baffles; ST — transparent acrylic plastic swimming tunnel; PS — posterior screen; CT — constant temperature water bath. Note the trolley system for suspending the respiration chamber, which separates at the gasket between the swimming tunnel and the impeller portions (Adapted from Wakeman and Wohlschlag 1978.) 132 8 00_ TEMPERATURE EFFECTS 3. 65 600 _ 2 00- L s« c -l I. I 2 f 2.68 L sec 0.9 4 - I 28°C 15° C C. nebulosus 20 ppt Fig. 3. Active, routine and standard metabolic levels, along with scope plots, at 2 8 and 15°C and at a salinity of 20 ppt for the spotted seatrout. The maximum sustained swimming speeds and the routine speeds are in lengths per second. After Wohlschlag (1977) . 133 10 20 30 SALINITY— PPT Fig. 4. Top panel — metabolic levels of the spotted seatrout at active (heavy points) , routine (triangles) . and standard (diamonds) conditions at 28 °C and a range of salinities. Lower panel — metabolic scope for maximum sustained activity over a range of salinities at 28°C from data above. Arrows indicate depressed metabolic levels of active fish in poor condition, as illustrated in Fig. 6. After Wohlschlag and Wakeman (1978) . 134 20. 8 . 16 >? 14 rz_ io. 50 SALINITY- PPT Fig. 5. Swimming characteristics of the spotted seatrout at 2 8°C and a range ot salinities. Swimming rate, X , expressed as square root of body length (cm) per second. Heavy points are averages of the maximum sustained (active) rates. Squares indicate the highest individual rates observed at each salinity. Arrow denotes fish in poor condition with depressed activity (see Fig. 6) . After Wohlschlag and Wakeman (1978) . 135 800. GROWTH- CONDITION EFFECTS 6 0 0 CM o O 2 00. •:■:•■:■:: •:•:•:■: ..-.;. .•:-:•:• :>:t '■:■:■:■ 1111 3.30 L s e c 0.99 3. I 6 L sec 0.83 NORMAL POOR C. nebu/osus 28° C 30 pp t Fig. 6. Growth-condition effects on spotted seatrout at 28°C and 30 ppt salinity in late summer. Metabolic levels shown at active, routine and standard rates with accompanying swimming speeds for active and routine rates. Note that the scope for normal fish is greatly reduced for those in poor condition. After Wohlschlag (1977) . 136 800- SUBLETHAL RED TIDE EFFECTS 600 _ *■ 400 i O o 2 0O_ ■:::'::: ■y 3. L 6 8 sec r~T I i 1.0 5 :<::' ;S>- ::::': ' :':". :•:•:•: X:' :>'•:•:> c :••:■:■:•:■:• > V- ■:■''■'; 1.16 NORMAL STRESSED C.nebulosus 2 8°C 2 5 pp t Fig. 7. Small red tide bloom effects on spotted seatrout metabolism at 28°C and 25 ppt salinity. Note that even if the active metabolic rate (which was not measured for the fish in waters exposed to bloom organisms) were not reduced, the scope is diminished because the standard rate of the exposed fish increases After Wohlschlag (1977) . 137 wmm i i I i ' i i i i i i i I ' i i i I i i i i I r i i >i CU •P .G •H 0> CM u> CM CM CM MMER mullet in (0 J-I cu 4-> fd rH 0 u id 4J c o •H CO CO E-t • CO J-I CU 4J fd rH fd 4-> i u fd 0 • J-I CU > •H 4J h- • & •H ■p 0 D CD rH O. CU fd ■H C •H r4 "~ U •I > •H CO CO 0 P CO CU > rH fd O CU 4-' CU P CU 0> CO CM 0 CU rH 0 4-» C0 CM to rH o a) fd CO •0 CD CO CU J-I Li •H CO z > 0 P cu -P 0) J4 CM 0) fX rH •H 4J (0 P CU 4-> CM 0. fd •H fd fx, 0 •H rH CO co j-i 0) 4-1 0 0 >1 4-> •H •H A fd fd CO CU •H 4-> «£> X 0 fd 0> (1) c rH CN 4-> CM CO a 0 4J CO 0) 0 J-I ■P c •H U •H X 0 4-> C 0 •H P 0 CM 0 5 a > t-\ u o fd 4-» 2 rH L J 3 fd rH CU > CU rH en 4-1 ^^ CM r 0 o fd CD G n CM a : J-i rH fd J4 0 •H rH 0 r-l rH fd >w N 0) U ro — co > £ i rH ■P 0 tr rH rCj *. Ol 0 •H CD TJ CO J4 4-J ri 4-» rH CO CM rH m fd 0 CD >i Xi fd & u rH rH 0 3 z a CO 4-> & C^ & S a E 0 H u fa •H I W CD CU n CO Fi CM - D td 4-> -M 0) o CO 0 CM r cu fd U ,cl c 0 J4 : 3 W l« \a ■P •H p, fe J_ o CM O d oo Cj> •H 138 CM O E 1000 H 800. 600 400 _ 200 - 20° C 28 ° C CONTROL SLUDG E LI QU I D CONTROL LIQUI D Fig. 9. Partially detoxified petrochemical sludge effects at a dilution of 0.2% v/v (about 1% of the LC5Q level) on the red snapper both at 20 °C with examples of separated sludge or of liquid portion only; at 2 8°C with liquid portion only. Scope in light shading; standard metabolism in dark shading. 139 800 E 600 400 200 0 L r"~~" — i WiM 2.6 0.5 0.5 MOmin 2 hr 2.9 2.6 IDENTITY : 0.0 2.0 3.1 (Control ) 3.2 3.3 3.4 3.5 3-6 Fig. 11. Exposure-recovery experiments with spotted seatrout to pharmaceutical wastes at dilutions sufficient to cause visible stress at tabulated exposure times and concentrations. Data based on experiments at 28°C and 35 ppt salinity with metabolic rates for an average recovery time of about two days in clean sea water. Maximum sustained swimming rates remained depressed in all cases after about two days . Standard metabolic rates in heavy shading; scope for activity in light shading. After Wohlschlag and Parker (MS) . 141 ENZYMES AND ANTIBODIES AS SENSORS OF TOXIC SUBSTANCES By Dr. Donald R. Sellers Midwest Research Institute 425 Volker Boulevard Kansas City, Missouri 64110 INTRODUCTION Biodetection, a new adjunct to conventional analytical chemical techniques, is based on the application of biochemical systems to the detection and monitor- ing of toxic substances in the environment. It is an exciting approach to many analytical problems because it offers sensitivity, selectivity, low cost, and the opportunity for real-time monitoring of air and water for subtoxic levels of pollutants. This article describes the rationale behind the use of these enzymes and antibodies, and shows how they can be used to solve real problems in monitoring and detection of enzyme inhibitors and antigens in air, water, and soil. Both insects and, unfortunately, humans are sensitive to poisoning by modern insecticides. The factor in both insects and humans that makes them sensitive to insecticides is an enzyme. Enzymes are used by nature to accelerate chemical changes in living things that ordinarily would occur too slowly. As biological catalysts, they are used over and over during metabolism and other chemical processes. The specific enzyme that is inhibited in this case, cholinesterase, is essential for proper transmission of impulses along nerve pathways. Blocking of this enzyme's function by reaction with pesticides interferes with proper nerve-impulse propagation and results in rigid muscular paralysis in the victim, whether man or insect. Of all the thousands of enzymes required for function of the human body, only one, cholinesterase, is markedly affected by organophosphate and carbamate pesticides. Furthermore, a lowered blood level of this enzyme is very strong evidence that the person being tested has been exposed to these kinds of chemi- cals. The inhibition of cholinesterase is thus extremely specific for only a certain class of compounds (pesticides) and, among these, only the organophos- phate and carbamate pesticides are effective as inhibitors. It has proved possible to remove this enzyme from animals, purify it, and incorporate it into instrumentation designed to detect and quantify pesticides in the environment. This combination of electronics and material from living tissue for the estimation of compounds in nature has been pioneered by the Life Sciences Section of Midwest Research Institute. We have named the process "biodetection." The biological part of the technique provides rapid sensitivity and specificity, while the electronics portion magnifies the signal that arises from the biological "sensor" into handy readout formats such as chart recordings or digital displays. Biodetection uses the remarkable specificity and sensitivity that are inherent in many biological systems. These systems have developed during aeons of 143 biological evolution into processes that have an extremely high degree of selectivity toward their target molecules. Enzymes generally react with only one compound or substrate to cause its chemical transformation into a different compound. Antibodies, a primary defense against infection in animals, are very specific in their reaction with antigens. Other biological systems are notable for their specificities such as chemotaxis (the attraction of an organism to a chemical) and even the senses of taste and smell. Some of these processes are the subjects of intense experimentation in the laboratories of Midwest Research Institute for possible application in biodetection. Others are under considera- tion for future study. USES OF CHOLINE STERASE Of all these sensitive systems, enzymes are currently the most attractive, in part because of the ease of handling of many enzymes. Also, more is known about enzymes and how they work than about the more complex biological processes. Of the thousand or so enzymes that have been isolated from biological tissues and characterized, cholinesterase is so far the most applicable for biodetection. One reason for this is that systems based on cholinesterase are sensitive to organophosphate and carbamate pesticides — the most widely used types of pesticides in the United States. The enzyme has several attributes which make it attractive for use in an instrument. For example, cholinesterase is easily immobilized or secured onto a solid support such as glass beads or polyurethane foam. Thus, the same enzyme moelcules can be exposed to a water sample and, if pesticide inhibitors are not present in great concentrations, exposed again and again without substantial losses of activity. Without immobilization to a solid support or some other enzyme retention technique, the enzyme, which is not inexpensive, would be washed away after analysis of each sample. Over long periods of testing, the immobilized, reusable enzyme system can result in significant financial savings compared to enzyme-wasting systems. Cholinesterase is also relatively stable over considerable ranges of pH and temperature. This factor results in increased adaptability and ease of handling of the enzyme. In addition, there are several ways to measure the activity of the enzyme. This makes the enzyme extremely versatile for incorporation into instruments of varying design and complexity. For example, in the presence of the appropriate substrate and color-producing reagents, cholinesterase activity can be measured spectrophotometrically. Wh^n another substrate is used, the activity can be estimated by the rate of acid production by the enzyme. For field applications, the enzyme can be immobilized on paper discs fabricated so that a change of color or a failure to change color will alert a worker to the presence of dangerous pesticide concentrations. Finally, electrometric techniques have proven to be invaluable for real time or instantaneous and continuous monitoring of natural and workplace environments. Each of these measurement techniques has been incorporated into pesticide detectors, Detection Principle — Enzymes are catalysts which increase the rates of specific reactions in living systems. Classically, the reaction of an enzyme with its substrate is given by E + S = N ES— > P + E , 144 where E is the enzyme, S is the substrate, ES is the enzyme-substrate complex, and P is the product. Enzyme is reformed during the reaction and functions as a catalyst. Formation of the product, under appropriate conditions proceeds linearly with time as in the "uninhibited" curve in figure 1. For an enzyme such as cholinesterase, reaction with a pesticide inhibitor (I) results in the formation of an enzyme-inhibitor complex (EI) that is not functional. E + I --> EI That is, the complex cannot make product. Biological Amplification — If 25% of the enzyme molecules are inhibited, only 75% will be available to make product, so product will be formed at 75% of the uninhibited rate. This is shown diagrammatically in figure 1. Since a certain constant number of new product molecules are formed each minute, the sensitivity of the system to pesticides can be increased by allowing longer incubation times. For example, let us imagine that each cholinesterase molecule processed about 84,000 substrate molecules per minute. Inhibition of only one enzyme molecule by one molecule of pesticide would result in the formation of 84,000 fewer product molecules during the first minute of incubation. Theoretically, if the incubation were 4 min rather than 1 min, the inhibition of one enzyme molecule by one pesticide molecule would prevent the formation of 336,000 product molecules, with a resultant apparent signal amplification of 1:336,000. This phenomenal increase is due to the reaction kinetics of the enzyme and is exclusive of electronic amplification. It has therefore been given the name "biological amplification." APPLICATIONS OF BIODETECTION USING CHOLINESTERASE Early interest in the possible use of cholinesterase as a sensor for inhib- itors in the environment came from the Department of Defense (D0D). Certain chemical agents called nerve gases are potent inhibitors of cholinesterase and can be lethal when encountered by humans. The DOD has maintained a continued interest in the detection of these compounds for the safety of personnel, not only in the field where chemical agents might have been deployed as a tactical weapon, but also in chemical demilitarization plants where stockpiles of nerve gases are being destroyed. The Army has supported research and development of instrumentation that ranges in sophistication from real-time monitors that feature preset alarm levels and time integration to simple paper discs which change color in the absence of inhibitors. Much of the technology developed during these programs is directly applicable in today's pesticide industry. Real-time monitors developed for chemical-agent demilitarization plants can be used, occasionally without major modification, in commercial pesticide-manufacturing plants. Many of the developments that were initiated during support from DOD contracts have been further refined with support form other nonmilitary agencies of the U.S. Government and private industry. Several of these areas of research have 145 resulted in byproducts or spinoffs of prototype instruments with a wide range of capabilities. Several of the more important of these advancements are described in the following sections. Real-Time Monitors — Many of the scenarios in which pesticides pose a potential health threat require real-time displays of pesticide levels. For example, in some areas in most pesticide manufacturing plants, accidental spillage can occur suddenly and with little visible indication. Workers in these areas need to be alerted as soon as possible. In another example, wastewater effluent from a pesticide plant can, at times, exceed pesticide levels considered safe by the Environmental Protection Agency (EPA). High pesticide concentrations here must be detected immediately so that the contaminated water can be shunted into reservoirs for further treatment to prevent dangerous pesticide levels from reaching wildlife and human drinking-water supplies. Several prototype real-time monitors have been constructed to meet particular needs of portions of government and industry. The Immobilized Enzyme Alarm (IEA) was developed as a high volume air monitor for the U.S. Army. It passed severe environmental chamber tests and maintained a sensitivity of 0.0001 ug agent per liter of air. The Immobilized Enzyme Monitor (IEM) (fig. 2) is a direct descendant of the IEA. The IEM was designed for real-time monitoring in pesticide plants during industrial hygiene studies, although it is now being used continuously to insure worker safety. In 1976, this instrument won an IR-100 Award, which is given each year to the 100 most significant new technical prod- ucts of the year by Industrial Research, Incorporated. The continuous Aqueous Monitor (CAM-1) was built during a program to detect low levels of organophos- phates and carbamates in water, and received an IR-100 award in 1972 (fig. 3). Each of these real-time monitors, although sampling from different media (air or water), functions by using very similar computer-controlled processes. How They Work — As mentioned above, immobilized cholinesterase has been used for the concentration and detection of inhibitors in air and water. In order to determine the relative activity of the enzyme product before and after exposure to the potential enzyme inhibitors, a synthetic substrate for the enzyme was chosen that would be easily detectable by electrochemical means. In this case, butyrylthiocholine iodide (BuSChI) was chosen because the unhydrolized substrate was stable in the presence of platinum gray electrodes to which a constant cur- rent was applied, whereas this same substrate after hydrolysis by cholinesterase was readily oxidized electrochemically. CH3 CH3 I. ChE L C3H7COSCH2CH2-KT - CH3 I- > C3H7COOH + HSCH2CH2— N — CH3 I- CH3 CH3 In this reaction, a thioester is cleaved by cholinesterase (ChE) to give thio- choline iodide. Under the influence of an electric current a part of it is converted to the disulfide. anode reaction 2 RSH *• RS - SR + 2H+ + 2e~ 146 Through the use of gel-entrapped butyryl cholinesterase on an open-pore polyurethane foam pad, it is possible to obtain good contact of the enzyme with the inhibitors or the substrate solution. The activity of the enzyme can be determined electrochemically by passing air and substrate through the enzyme pad, while 2 ua of current is applied to the electrodes in contact with the upper and and lower surfaces of the enzyme pad. If the enzyme is active, the substrate will be hydrolized as it passes down through the pad with the resulting presence of thiol at the lower platinum electrode (the anode) where it is sensed by the generation of a potential of ^ 250 mv. On the other hand, if the enzyme is inhib- ited by an organophosphate or carbamate, the thiol is not formed and a potential of ^ 500 mv is generated at the electrodes. Thus, by observing the changes in the cell potentials before and after exposure to enzyme inhibitors, it is possi- ble to observe whether none, part, or all of the enzyme has been inhibited by the materials brought into contact with it. A simple electronic circuit can be used to signal an alarm when the voltage rise exceeds a preset threshold voltage. In order to convert this electrochemical cell into the air or water monitor, it is necessary to establish a detection cycle and to add some components; these are shown in figure 4. The objective of the cycle is to introduce a time interval during which the enzyme pad is exposed to potential inhibitors (2 min) and an enzyme activity test period (1 min). Comparison of cell voltages from successive cycles is necessary to determine the presence or the absence of enzyme inhibitors. That is, increases in voltages from cycle to cycle are evi- dence for the presence of inhibitors; if the cycle-to-cycle increases are large, then the amount of inhibitor is also large. Before the activity of the enzyme pad is completely gone, it is necessary to add a fresh one. Figure 4 shows the fluid pathways in the monitoring system and the sequence of events making up the detection cycle. For 2 min of each 3-min detection cycle, water (with or without inhibitors) is pumped through the enzyme pad. During the last 1 min of the detection cycle, air is blown through the pad at the same time that the substrate solution is introduced and a constant current is applied for the last 40 sec of the 3-min cycle. At this time, a voltage is generated that is characteristic of the amount of enzyme activity present on the enzyme pad. By comparison of cell voltages from successive cycles it is possible to determine whether enzyme activity has been lost due to inhibitors in the sample. PASSIVE-DOSIMETRY The Biodetection Group at Midwest Research Institute has had success in measuring personnel exposures using passive dosimeters. These dosimeters are termed passive since they depend on air currents and diffusion to bring airborne pesticides into their collecting matrices. Positive displacement samplers continuously aspirate local air through the collector but have the disadvantages that they require a power source to run a pump and are usually relatively bulk. The passive dosimeter, on the other hand, can be as simple as a thin collector medium (similar to filter paper), enclosed in a plastic badge with a hole to allow access of pesticide to the collector, and a spring clip to facilitate attachment to the worker's clothing. 147 After a predetermined time of wearing the badge (usually 8 hr), the medium is extracted with solvent to remove adsorbed pesticides. The extract can then be analyzed spectrophotometrically for cholinesterase inhibitors. Spot Checker — A simple but effective device for detecting airborne pesticides in remote areas was developed and named the Spot Checker (fig. 5). It won an IR-100 Award in 1977. Each of the six orifices of the Spot Checker is covered with a filter impregnated with cholinesterase. During a single revolution of the hand crank, a preset volume (usually 25 ml) of air is drawn through each filter. If there are inhibitors in the air, they will combine with some of the immobilized enzyme and inactivate it. Subsequent addition of substrate and color reagent can be used as a test for remaining enzyme activity. Loss of activity indicates the presence of significant levels of pesticides in the air. Retention of activity shows that the air is relatively safe. Soil Analyses — Another biodetection technique which has been developed and used extensively at Midwest Research Institute deals with the quantification of pesticides in soil. In the first step of this procedure, the pesticide was extracted from the soil by blending the sample in the presence of a water- alcohol mixture. The extract was then centrifuged at very high speed (15,000 rpm) to separate particulate matter. The second step was a manual spectropho- tometry cholinesterase inhibition assay for actual quantification of the pesticide. DEVELOPMENTAL PROGRAM FOR OCEAN MONITORING It seems plausible that some of the toxic chemicals that are of concern environmentally would have an inhibitory effect on certain mammalian enzymes. Thus, it is possible that monitoring systems analogous to those described for cholinesterase could be designed and developed using other enzymes for the detection of other important pollutants. With this attractive prospect in mind, the National Oceanic and Atmospheric Administration is supporting work at Midwest Research Institute that has as its ultimate goal the development of ocean pollution monitors based on biochemical sensors. The initial work of this program has been the identification of enzymes that are likely candidates as biosensors. Computerized searches of the scientific literature produced a lengthy list of enzymes that were reported to be inhib- ited by four classes of compounds on EPA' s list of priority pollutants. These enzymes are shown in table 1. The numbers in parentheses under the name of the inhibitor class are the number of compounds in that class on the priority pol- lutant list. The numbers in the columns are the numbers of compounds in each class which were reported to have been tested for inhibition of the correspon- ding enzyme. Other criteria were also used in the initial selection of enzymes. The enzyme had to be commercially available in order to avoid costly and time consuming isolation and purification procedures. Additionally, the enzyme had to be easy to assay and require inexpensive assay reagents. From preliminary results given in table 2, it is clear that several of the enzymes are inhibited by specific compounds on the EPA priority pollutant list. 148 TABLE 1 ENZYME-INHIBITOR MATCHES FROM LITERATURE Pesticides Phenols Metals Phthalates Enzyme (16) (11) 5 (12) 6 (4) Lactic dehydrogenase 4 0 Aldolase 4 0 1 3 Hyaluronidase 0 0 1 0 Glucose-6-phosphate de- 0 0 0 3 hydrogenase . Malic dehydrogenase 1 6 0 0 a-Amylase . Pyruvate carboxylase a-Chymo trypsin , Glucose oxidase Histidase 12 0 0 0 5 0 0 0 0 3 0 0 0 0 7 0 0 0 6 0 Catalase 5 0 0 0 Alcohol dehydrogenase 3 4 0 Glyoxalate reductase 2 0 0 Glutathione-5-alkyltransf erase 3 0 0 ATP Phosphoribosyltranferase 1 0 0 Hexokinase 4 0 3 Pyruvate kinase 2 2 3 Acetate kinase 2 0 0 Creatinine phosphokinase 2 0 0 Fructose-6-phosphate kinase 2 0 3 8-Amylase 12 0 0 0 Cholines terase 3 0 0 0 ATPase 2 6 5 0 Carbonic anhydrase 1 0 7 0 Lipoxidase 0 1 0 0 Luciferase 0 1 5 0 Phosphorylase 0 3 0 0 Pepsin 0 1 0 0 Isocritric dehydrogenase 0 0 8 0 Glyceraldehyde-3-phosphate de- 0 0 1 0 hydrogenase Dopamine-p-hydroxylase 0 0 1 0 Phosphoglucomutase 0 0 1 0 Rhodanese 0 0 3 0 RNA Polymerase 0 0 1 0 Alkaline phosphatase 0 0 1 0 Ureas 0 0 5 0 Fructose-1 ,6-disphosphatase 0 0 1 0 Trypsin 0 0 2 0 Carboxypeptidase 0 0 6 0 149 TABLE 1 (concluded) Pesticii ies Phenols Metals Phthalates Enzyme (16) (11) 0 (12) 1 (4) Invertase 0 0 Ribonuclease A 0 0 5 0 Urokinase 0 0 3 0 Lysozyme 0 0 1 0 3:5' -Cyclic nucleotide phos 0 0 1 0 phodiesterase L-Asparaginase 0 0 5 0 Creatinase 0 0 2 0 a-Galactosidase 0 0 2 0 Glucose-6-phosphatase 0 0 1 0 Pepsin 0 0 1 0 Cytochrome c reductase 0 0 0 3 Phosphoglucose isomerase 0 0 0 3 Enolase 0 0 0 3 Enzyme chosen for initial laboratory work. One enzyme was chosen from each of the four classes of priority pollutants which have been iden- tified. Also chosen for further literature search to find other known inhibitors. Enzyme chosen for further literature search to find other known inhibitors 150 TABLE 2 PRELIMINARY RESULTS OF ENZYME INHIBITION TESTS Enzyme Inhibitor Class Results' Lactate dehydrogenase Organochlorine pesticides DDT, Toxaphene Lactate dehydrogenase Metals Ma late dehydrogenase Phenols a-Chymo trypsin Aldolase a-Amylase Hyaluronidase Glucose oxidase Histidase Glucose-6-phosphate dehydrogenase Catalase Urease Hexokinase Phenols Organochlorine pesticides Organochlorine pesticides Metals Organochlorine pesticides Metals Metals Phthalates Organochlorine pesticides Metals Alcohol dehydrogenase Metals (yeast) Alcohol dehydrogenase Metals (liver) Ag+, Cd+2, Cu+2, Zn+2, Se+\ Hg+2 4,6-Dinitro, 2,4,6- Trichloro, 4-Chloro-m- Cresol, Pentachloro Cu* Se+4, Hg+2, Cn", Fe+3 Nit2 Cd+2,C Sb+3,C Pb 2° Cu 2, Zn 2, As Organochlorine pesticides Aldrin, Chlordane, Heptachlor Epoxide, Endrin, DDT, DDD, DDE a Blank indicates not yet tested. Dash indicates none of the tested com- pounds gave significant inhibition. Compounds and ions listed gave significant inhibiton (10 4 M) . b Not an EPA priority pollutant list. c Possible activators. 151 These enzymes are good candidates for incorporation into continuous monitors. Work is underway to identify other enzymes that might be effective as biosensors, Future phases of this program will entail designing and fabricating real-time monitors based on these biosensors. ENZYME AMPLIFIED IMMUNOASSAY In addition to enzyme inhibition, antigen-antibody reactions are a sensitive and selective type of biochemical recognition reaction. Antibodies prepared in small animals to be sensitive to synthetic antigens have been used for several years for the quantitative analysis of drugs, metabolites, and hormones, most of which occur in extremely low concentrations. Most of these assays, however, rely on complex and bulky equipment, such as gamma and scintillation counters, to provide quantitation data about the antigen-antibody reaction. Work in our laboratories and elsewhere has shown that the detection of an antigen by an antibody can be signalled by a coupled enzyme system. The function of the enzyme is to amplify and provide a display, such as a color change, that is related to the amount of antigen present. We have initiated development of an enzyme-monitored immunoassay for the detection of 2 ,4-dinitrophenol (2,4-DNP), a toxic substance on EPA's list of priority pollutants. Although there are a number of possible variations in the procedure, the various steps shown in figure 6 explain how the most attractive approach will function in the presence and absence of the pollutant to be detected. Catalase is used to label the hapten, 2,4-DNP. Either nonlabeled 2,4-DNP to be assayed or catalase-labeled 2,4-DNP reacts with anti-DNP in solution. Then a fine, particulate, polymerized antibody to goat gamma globulin (the second antibody) contacts the solution and is separated out on a membrane or by a centrifugation step, depending on whether the test is being developed for field or laboratory use. The amount of catalase-labeled 2,4-DNP separated out with the second antibody is a function of the amount of unknown competing with this enzyme-hapten complex. Use of the substrate, hydrogen peroxide, and an oxygen electrode permits the quantitation of the unknown DNP. BIODETECTION FUTURE BRIGHT The use of biological mechanisms for the detection and quantification of compounds of interest is in its infancy. More and more frequently, articles are appearing in the analytical chemistry literature that describe new and practical uses of biodetection. Researchers are realizing that many biological systems have extraordinary chemical specificity and sensitivity. The challenge is to adapt these natural sensors to instrumentation and devices in ways that improve current methods of detection and analysis. The future for the Biodetection Group at MRI is exciting since new discover- ies in enzymology, immunology, and other areas of biochemistry are routinely revealing new biological mechanisms that may be applicable in biodetection. The future in this area of analytical chemistry looks very promising. 152 ACKNOWLEDGMENTS Much of the work described in this article has been supported financially by the U.S. Army, the U.S. Environmental Protection Agency, and the National Oceanic and Atmospheric Administration. The author is grateful for the guidance given by Mr. William B. Jacobs and Dr. Louis H. Goodson and for the tireless laboratory effort rendered by the technical staff of the Life Sciences Section. 153 BIOLOGICAL AMPLIFICATION 3 o u Q O a. Figure 1 - Biological Amplification. 154 Figure 2 - Immobilized Enzyme Monitor 155 ALARM TOTALIZER ALARM THRESHOLD SWITCH CURRENT LIGHT SUBSTRATE LIGHT WATER LIGHT HORN ALARM LIGHT POWER ON/OFF CLOCK HOLD RESET MANUAL CONTROL PANEL PAD CHANGE THRESHOLD SWITCH CELL VOLTAGE INDICATOR ALARM LIGHT CELL VOLTAGE RECORDER WATER PUMP MOTOR ELECTROCHEMICAL CELL-PAD CHANGER ASSEMBLY Figure 3 - Continuous Aqueous Monitor (CAM-1) 156 CO 1 1 4) I 2 CM *"" 4) «A "■■ a U >s • E u b 3 3 O) 3 t_ c o *E £ < c H- "O -o c c n _> o u CN __ 0) 13 0) *» «t a a a 13 E E J5 a. 3 3 3 CL O- a. CO < i! c o o E 2 3 «) S- o ••-> f- a; c «— o E b i— SZ C (O +-> •«- > c ■Mr- O a> s- C +■> (0 to A3 U 3 u c O) o ■•-> O w ID 3 157 Figure 5 - Spot Checker 158 Case I: DNP Present Case II: DNP Absent M DNP Unknown Anti-DNP A B Step 1: Add Anti-DNP Anti-DNP ^ Catalase-DNP | Step 2: Add Catalase-DNP Catalase-DNP E* uu Second Antibody Particles > Step 3: Add Second Antibody yy Second Antibody Particles + Soluble 11 Insoluble No DNP ^N bU > > Insoluble Step 4: Discard Insoluble Fraction- Add O2 Electrode and H2O2 Result: Result: Increase in O2 Means DNP Present No Increase in O2 Means DNP Absent Figure 6 - Diagram of Enzyme Immunoassay for DNP Detection (see next page for reagent code) 159 REAGENT CODE DNP, the "unknown" pollutant to be detected Anti-DNP, an antibody that binds DNP, made in rabbits W Anti-rabbit gamma globulin, a second antibody that binds the anti-DNP, made in goats, chemically treated to make it insoluble Catalase enzyme CO Catalase-DNP, covalently bonded reagent 160 BIOPROBES FOR THE MONITORING OF TOXIC COMPOUNDS AND HAZARDOUS MATERIALS IN WATER by Drs. Shia S. Kuan and George G. Guilbault Department of Chemistry University of New Orleans New Orleans, La 70122 INTRODUCTION The bioanalytical research group at the University of New Orleans is composed of biochemist and analytical chemists under the supervision of Dr. George G. Guilbault. The primary research interest has been recently devoted to the devel- opment of bioprobes for the analysis of substances of clinical, agricultural, and environmental importance. The so-called "bioprobe" is, in general, a device consisting of a sensor (i.e., an electrode) and bioactive materials (i.e., enzymes or antibodies) for highly selective analysis of certain compounds. During the past decade, we have successfully induced, isolated, and purified many enzymes. With the use of these enzymes, new bioprobes and analytical systems have been set up and fabricated for fast, simple, accurate, selective, and sensitive assay of a variety of compounds of interest. In this paper, only those analytical systems pertaining to pollutants present in water systems are described. ANALYTICAL SYSTEMS FOR NITRATE AND NITRITE It is well-known that the toxicity of nitrite is due mostly to the formation of carcinogenic N-Nitroso-compounds by the reaction between nitrite and secondary or tertiary amines in the animal tissues, and to the induction of methemoglobine- mia (caused by nitrite oxidation of hemoglobin from the ferrous to the fernic form). Though nitrate itself is not toxic, it can be converted to nitrite inside an animal body by microorganisms present in the digestive tract. Due to the increasing use of nitrate and nitrite in foods, fertilizers, detergents, and other fields of industry nowadays, the development of fast, simple, accurate, and economical methods for regulatory analysis of nitrate and nitrite in portable water is urgently needed. Though numerous methods for NOo and NO^ assay have been developed and reported elsewhere, most of these methods are either nonspe- cific and subject to interferences from diverse substances or very complicated to manipulate. The use of a bioprobe which combines speed, selectivity, and simplicity of operation for chemical analysis appears to be the best approach. The development of nitrate and nitrite bioprobes is based on the following reactions: nitrite N02~ + 6e~ + 8H+ reductase NH4+ + 2H20 (1) 161 The nitrite is selectively and quantitatively reduced to an ammonium ion by nitrite reductase in the presence of electron donors, for example, reduced methyl viologen NADH, and the ammonium ion generated is sensed by an ammonium ion selective electrode. T+ nitrate N03 + 2e + 2IT reductase N02 + H20 (2) As shown in equation 2, an enzyme, nitrate reductase, catalyzes the reduction of N0o~ to N02~ which is in turn reduced to ammonium ion by nitrite reductase. Source of Enzymes — Nitrate and nitrite reductases are generally found in certain microorganisms and plants and exist as constitutive enzymes in the living system. These enzymes can be induced as the growth conditions are altered. We have successfully induced, isolated, and purified severl nitrate reductases from E. coli and Chlorella vulgaris, and nitrite reductases from _E. coli, Azotobacter chroococcum and spinach (refs. 1, 2, 3, 4, and 5). These enzymes differ from one another in their stability, specificity, and catalytic activity and especially in the ability to use a lower redox potential electron donor. In general, inducible nitrate and nitrite reductases are not as stable as constitutive enzymes during purification, immobilization, and storage but have higher specific activities. Both stability and catalytic activity are important for the construction of a useful, sensitive, and fast response bioprobe. Thus, the use and selection of a proper enzyme for bioprobe depends mostly on the systems to be used for the analysis of the desired substrate. Electrodes a. Ammonia sensor - Several devices for sensing ammonium ion generated during enzymatic reaction have been developed in our lab and used for the construction of bioprobes as shown below. b. Nonactin membrane electrode - Details on the preparation and operation of this type electrode have been described (ref. 6) and will not be mentioned here. Nevertheless, we like to point out that this electrode is subject to some K*" and Na+ interferences, thus limiting its application for selective analysis of NH/ . c. Air Gap Electrode - To eliminate the cation interferences mentioned above, we used an air gap electrode to measure the amount of gas liberated from the enzyme - substrate reaction. An air gap electrode is composed of a combined pH electrode housed inside a polystylene cylinder and an air tight chamber. The gas evolved from the reaction mixture in the chamber diffuses to the elec- trode surface where it reacts with a wetting solution, thus causing a change of hydrogen ion concentration. The pH observed is proportional to the log con- centration of the gas and common interfering ions, e.g., K+ and Na+, do not interfere. d. Gas Selective Membrane Electrode - Ammonia selective gas membrane electrodes are now available commercially. For the detection of NH3, these electrodes require a pH which is much higher than the optimal pH for most of 162 the enzyme reactions. Therefore, it is not suitable for use as a bioprobe sen- sor. An ammonia gas sensing electrode which could detect the ammonium ion at lower pH's was then developed in this laboratory (ref. 7). The sensor was made by placing a piece of gas selective membrane over the surface of the radiometer pH electrode. A rubber plastic sealant and an "0" ring were used to fix the membrane and to prevent the leakage from the internal solution to the sample. This electrode gives a fast response time as well as a speedy recovery time and exhibits excellent performance in aqueous solutions without any interference originating from K+, Na+, or other cations. It is ideal for use in constructing a bioprobe for batch or continuous measurement of ammonia in solution. 3. Instrumental Setup - With the use of either soluble or immobilized enzymes, different systems have been set up for selective analysis of NOo and NOo"" in food as well as in water (refs. 2, 3, 4, 5, and 6). As described in these papers, soluble enzymes were employed for batch measurements and immobilized enzymes for applications in continuous flow systems. The ammonia generated was quant i- tated by an air-gap electrode. The only disadvantage in using this set up is that NaOH must be added to the chamber to liberate NH3 from the reaction mixture for better response. A very simple bioprobe could be developed using a gas mem- brane electrode as a sensor for convenient, sensitive, selective, and direct anal- ysis of NOo" and NO2"" if the isolation of more active and more stable nitrate and nitrite reductases could be achieved. Research along this line is now in progress. ANALYTICAL SYSTEMS FOR CYANIDE Cyanide has been one of the toxic chemicals discharged in industrial effluents. It is especially abundant around iron and steel industrial areas. Several years ago, biological detection and decontamination systems were developed in our lab- oratory to monitor and remove this highly toxic pollutant from the environment. Source of Enzymes - Two enzymes, rhodanese and 8-cyanoalanine synthase, have been isolated and used for this purpose (refs. 8 and 9). a. Rhodanese - This enzyme was isolated and purified from the liver of higher animals and catalyzes the following reaction: enzyme CN~ + S203 > CN" + SO3"" (3) Since the products are not toxic, this enzyme is excellent to be used for decon- tamination purpose to remove CN~ from water. b. 8 -Cyanoalanine synthase - This enzyme was induced from E. coli. 12, using a cultural medium containing 10"^ MKCN for growth. The purified enzyme catalyzes the injection of CN~ into the cysteine molecule with release of HS~: enzyme CN" + CH9-CH-C00H > CHo"CH-C00H + HS" /I I 2 I HS NH2 CN NH2 cysteine 8 -cyanoalanine 163 Because one of the end products, $-cyanoalinine, is a neurotoxin, this enzyme cannot be used for decontamination. However, the other end product, HS~, is very useful as an indicating ion for selective monitoring of the CN~ content in water. Sensors - Both cyanide and sulfide electrodes have been employed to monitor the cyanide. a. Cyanide Electrode - The electrode was prepared from a plexiglass tube onto which a cyanide sensitive membrane was fitted. A silver wire in 10 M AgNOo solution was used as the internal reference (ref. 8). Nevertheless, this electrode is subject to interference from thiocyanate and sulfite ion. b. Sulfide Electrode - The commercial sulfide electrode requires rather high pH for sensitive detection and is not proper for use as a bioprobe. A sulfide electrode which senses HS~ at lower pHs was made in the laboratory by gluing the compressed disc onto a glass tube using a silver billet electrode in 0.1M AgN03 solution as the inner reference system. This electrode is extremely selective when coupled with 3-cyanoalanine synthase, constituting a cyanide bioprobe (ref. 10). Instrumental setup - Both soluble and immobilized enzymes have been used in either batch or flow system for the measurement of cyanide ion in water system. The rhodanese bioprobe provides a sensitive and accurate means for CN~ analysis when compared with spectrophotometry procedures, since it requires less sample preparation and is easily adapted to an automated analyzer. Interferences from thiocyanate makes it less attractive for selective analysis. The $-cyanoalanine bioprobe has been successfully used for batch CN~ assay, as well as for continuous monitoring of CN~ removal from water system. This probe exhibits high selectivity; no other ion interferences. ANALYTICAL SYSTEMS FOR CARBAMATE PESTICIDES During the past decade, carbamate pesticides have been widely used in nature and, hence, pose an ecological problem after use. Several years ago, studies were conducted and biological systems developed for the monitoring and decon- tamination of these pollutants using microbial enzymes as catalysts. Source of Enzyme - An enzyme aryl acylamidase was isolated from pseudomonas striata that grew in mineral medium with acetanilide as the sole carbon and nitrogen sources for growth (ref. 11). The purified enzyme was found to be active toward the hydrolysis of phenyl or methyl carbamate pesticides. OH enzyme (\ + CO2 + CH3NH2 (5) 164 Sensor - Based on the air gap sensor principle, a simple electrode was developed to measure the methylamine generated from the reaction (ref. 12), Methylamine concentrations ranging from 10""-' - 10"^M can be easily detected. This electrode is selective and only slightly subject to dimethyl amine or ammonium ion interferences. It is ideal for the construction of a methyl carbamate bioprobe. REFERENCES 1. W. R. Hussein and G. G. Guilbault, Anal. Chim. Acta 72. 381 (1974). 2. C. H. Kiang, S. S. Kuan and G. G. Guilbault, Anal. Chim. Acta 80 209 (1975) 3. C. H. Kiang, S. S. Kuan and G. G. Guilbault, Anal. Chem. 50 1319 (1978) 4. Ibid 5_0 1323 (1978). 5. C. H. Kiang, Ph D. Dissertation, Univ. of New Orleans. 6. G. G. Guilbault, G. Nagy and S. S. Kuan, Anal. Chim. Acta 6_7 195 (1973) 7. K. W. Fung, S. S. Kuan, H. Y. Sung and G. G. Guilbault, Anal. Chem. !5J. 2319 (1979). 8. W. R. Hussein, L. H. VonStrop and G. G. Guilbault, Anal. Chim. Acta. 6_1^ 89 (1972) 9. S. S. Kuan, G. G. Guilbault and R. Cochran, Anal. Letters 3_ 593 (1970) 10. G, G, Guilbault, W. F. Gutknecht and S. S. Kuan, Anal. Biochem. 46_ 200 (1972) 11. K. P. Hsiung, S. S. Kuan and G. G. Guilbault, Biochem. Biophys. Res. Comm. 66 1225 (1975) 12. K. P. Hsiung, S. S. Kuan and G. G. Guilbault, Anal. Chim. Acta 84_ 15 (1976). 165 POTENTIOMETRIC BIOSENSORS BASED ON HETEROGENEOUS CATALYTIC SYSTEMS by Dr. Mark E. Meyerhoff Department of Chemistry The University of Michigan Ann Arbor, Mich. 48109 Potentiometric ion-selective membrane electrodes are now routinely employed in many biological and environmental analyses. The advantages of membrane electrode-based methods include low cost, simple instrumentation requirements, ability to do in-situ measurements (i.e., sample turbidity and color are not a problem) , and in-field capability. The range of substances measurable by these electrode systems can be expanded by coupling appropriate enzyme reagents to given electrodes to form selective enzyme electrodes (1). More recently, a new approach has been to use intact organelles and whole cells (i.e., heterogen- eous biocatalysts) in conjunction with membrane probes for the purpose of bio- chemical analysis (2-8). In this report, the work done to date in this new area will be summarized and some speculation as to the future analytical possibilities using these heterogeneous systems will be presented. In considering any biocatalytic process, a simplified expression of the reac- tion taking place can be summarized as biocatalyst A > B If product B is detectable by one of the many selective membrane electrodes available, it is a simple task to develop an electrode-based analytical method for substrate A. Previously, it has been assumed that purified enzymes were the most suitable reagents for such purposes, since the isolated enzymes should exhibit maximum biological activity and selectivity. However, it has recently been shown that in many instances intact cells or organelles can serve as the biocatalyst without loss of desired selectivity. This use of heterogeneous bio- catalytic reagents in conjunction with membrane electrodes is possible because products of extraneous catalytic pathways present are not sensed by the selective potentiometric detector. Moreover, in cases where biological activity is depen- dent on the integrity of cell structure, membrane probes offer the simplest means of incorporating these particulate biosystems into practical analytical methods. Table I lists the advantages of using such catalysts for analytical purposes. Table II summarizes the types of catalysts which may be used with membrane electrodes and the variety of analyses which have been or may in the future be performed. An example of the use of intact organelles for selective analysis can be shown in the measurement of thyroxine (T4) using an iodide selective membrane elec- trode and a rat-liver microsomal reagent (2). The microsomes catalytically deiodinate the T4 to T3 (triiodothyronine). The amount of iodide released can be directly monitored in the microsome suspension with the iodide probe. Figure 1 illustrates the analytical results and selectivity obtained using such a system. 167 Table I. — Analytical advantages of heterogeneous biocatalysts 1) No need for costly isolation and purification of active component - low cost 2) Necessary cof actors, etc. are already present 3) Stability of catalytic activity is maintained (i.e., catalyst is in natural environment). 4) Analytical multistep enzyme sequences are already present. Table II. Summary of heterogeneous biocatalysts used and types of analyses Biocatalyst Intact organelles, membranes Intact whole cells (bacterial, tissue) Analytes Amino acids Hormones Nutrients Antimicrobial agents Environmental toxins Inorganic ions It should be noted that attempts to extract the deiodinating activity from the microsomal fraction have been unsuccessful and, therefore, the only approach to using this bioactivity is to use the intact microsomal system. The application of immobilized microbial cells on electrode surfaces has given rise to "bacterial electrodes" sensitive to a variety of substances. These elec- trodes are characterized by prolonged lifetimes over previous isolated enzyme based electrodes for these same substrates. A glutamine sensor has been prepared by immobilizing the bacterial strain Sarcina flava on the surface of an ammonia gas sensing electrode (3). The electrode exhibits complete selectivity for L-glutamine over all other amino acids and has a useful lifetime of over 3 weeks. This is considerably longer than the 1-day reported for the isolated bacterial glutaminase used to prepare a glutamine enzyme electrode (9). Similarly, other bacterial electrodes have been prepared for aspartate (4) and arginine (5). A nitrate selective bacterial electrode has been developed (6), utilizing the bac- terium Azotobacter, which reduces nitrate to ammonia at the surface of an ammonia gas sensor. The electrode functions reproducibly over a wide concentration range and can be employed for nitrate determinations in environmental water samples, without interferences from other anions. An alternate means of preparing bioselective sensors may lie in the use of intact tissue cells. A porcine kidney slice has been used in conjunction with an ammonia gas probe to develop another glutamine selective electrode (8). Fig- ure 2 illustrates the typical response and selectivity of this biosensor. Once again, this electrode exhibits complete selectivity for glutamine over all other 168 amino acids and has a lifetime of over one month. Sodium azide is used to pre- serve the kidney cells from microbiological attack. This sensor has been used to directly determine glutamine in complex biological samples (10). It is evident that gas-sensing electrodes are quite popular as base probes in the development of these and other biosensors. The ammonia electrodes used in this work have been commercial models, which incorporate pH glass membranes as inner sensing elements. In an effort to reduce the size, cost, detection limits and fragile nature of current ammonia detectors, a new type of miniature ammonia gas sensor has been developed (11). Figure 3 schematically illustrates this new gas probe. The electrode employs a plastic ammonium ion sensitive membrane as the inner sensing element and will be used to prepare more simplified biosensors. This ammonia electrode can also be used for direct ammonia determinations in environmental water samples. In the future, electrode based heterogeneous biocatalytic assays will certainly be extended to determine other important amino acids, hormones, nutrients, etc. by using appropriate cellular substances. In addition, an even more exciting possibility will be the measurement of antimicrobial substances by determining their inhibitory effect on bacterial metabolic rates. With the proper choice of biological systems, it should be possible to develop simple and reliable electrode based bioassays for heavy metals, antibiotics, and other toxic substances. REFERENCES 1) G. G. Guilbault in "Methods in Enzymology," Vol. XLIV , K. Mosbach Ed., Academic Press, New York, N.Y. , 1976, p. 579. 2) M. E. Meyerhoff and G. A. Rechnitz, Anal. Lett., 1_2, 1339 (1979). 3) G. A. Rechnitz, T. L. Riechel, R. K. Kobos, and M. E. Meyerhoff, Science, 199, 140 (1978). 4) R. K. Kobos and G. A. Rechnitz, Anal. Lett., _10, 751 (1977). 5) G. A. Rechnitz, R. K. Kobos, S. V. Riechel and C. R. Gebauer, Anal. Chim. Acta, 94, 357 (1977). 6) R. K. Kobos, D. J. Rice and D. S. Fluornoy, Anal. Chem. , 5J_, 1122 (1979). 7) S. Suzuki and I. Karube, Anal. N.Y. Acad. Sci. , 326, 255 (1979). 8) G. A. Rechnitz, M. A. Arnold, and M. E. Meyerhoff, Nature, 2T%_, 466 (1979) 9) G. G. Guilbault and R. F. Shu, Anal. Chim. Acta, 56_, 333 (1971). 10) M. A. Arnold and G. A. Rechnitz, Anal. Chim. Acta., _11_3, 351 (1980). 11) M. E. Meyerhoff, "A New Miniature Potentiometric Ammonia Gas Sensor," paper presented at Pittsburgh Conference, Atlantic City, N. Y. , March, 1980. 169 1tf> CO h- (d e 0 CO 0 'H u •H s 0 •v3 O H -P u CD ^ r-t •* • CD o 'd c •H T5 o O o H IM CD 0 •!-• ro >i 4J . u -H CO «)-» > 4J C/) -r-UG JD 4-> CD o g 3 Q) O CO rH V< CD 3 W CO O) - T5 CD O C 2 fC CD 1 CD C CO -H C X 0 O & M W >, cd x: CO o o o 10 o CO C5 E-i CD U •H AUJ ' 3 170 140! 100!- > E 6 0;- 2 0 All other AMINO ACIDS L-GLUTAMINE 4 -log u ~ o i. .Cj 7 O Figure 2. Response and Selectivity of Tissue Based Glutamine Sensor. 171 (a) Ag/AgCI electrodes interna! buffer 0.01 M NH4CI -tygon tubing PVC membrane Q-ring gas membrane 3mm NH3 + H+^=±NH4"1 t NH __ Polymer membrane — buffer film 1 - gas permeable mem brane Figure 3 a) Schematic diagram of new ammonia sensor. b) Expanded view of sensing tip with response processes . 172 UTILITY OF THE MICROTOX LUMINESCENT BACTERIAL ASSAY FOR THE RAPID ASSESSMENT OF AQUATIC POLLUTION BY Dr. Anthony A. Bulich Beckman Instruments, Inc. 6200 El Camino Real Carlsbad, CA 92008 INTRODUCTION In 1979, the author described a new method for the quick assessment of toxicity in aquatic samples. *■ This bioassay, which utilized luminescent bacteria as the bioassay organism, displayed potential as a rapid, reliable, and inexpensive method for assessing the toxicity of water samples. The results outlined in this communication compare the luminescent bacterial assay (designated as Microtox) with conventional bioassays.* In addition, a more rapid and precise method for conducting the bacterial luminescent assay is described. The instrument is illustrated in figure 1. Description of Microtox Toxicity Analyzer — The Toxicity Analyzer is equipped with a rotary shutter built around a photomultiplier tube adjacent to a temperature controlled reaction chamber. This reaction chamber is provided with adjustable temperature control over the range of 10°C to 25 °C. The instrument is also equipped with a fifteen-well incubator chamber which was controlled within + 0.3°C of the selected reaction chamber temperature. The reaction chamber was controlled at 15 °C in all assays reported here. The photometer was equipped with an air pump and a special drying chamber. Air was pumped through a drying chamber for distribution to incubator and reac- tion chambers, preventing fogging of cuvettes and minimizing water condensation in the incubator wells. The photometer was connected to a 10-inch chart recorder and all values of light levels were obtained from the chart recordings. Reagents used for the assay — The bioluminescence assay required the use of 3 specially prepared reagents. 1. Freeze-dried luminescent bacteria — This preparation consisted of freeze dried Photobacterium phosphoreum NRRL B— 11177 (previously referred to as P. f ischeri (ref. 1). When reconstituted with one ml water, the viable bacterial population is approximately 10° per ml. This reagent was used for only 1 hour, after reconstitution or for two complete assays. 2. Reconstitution solution — This solution, used to hydrate the freeze-dried bacteria, was specially prepared distilled water, free of organic compounds. *Microtox is a trademark of Beckman Instruments, Inc., Fullerton, California 173 3. Diluent — This solution was used to dilute the reconstituted bacteria by adding 10 1 of bacterial suspension to 0.5 ml of diluent. The solution contained 2% NaCl which is necessary to osmotically protect the luminescent bacteria. Assay Procedure — All samples to be tested were adjusted to 2% + 0.2% NaCl by the addition of analytical grade NaCl. The samples were diluted using diluent solution such that the following concentrations were tested: 50, 32, 18, 10, and 5%. These diluted samples plus a clean water control (diluent only) were precooled to 15 °C in the incubator wells. The Microtox reagent was hydrated with 1 ml of reconstitution solution precooled to 3°C, and the bacteria transferred to testing cuvettes containing 0.5 ml of diluent equilibrated to 15 °C. Initial light measurements were made for each cell suspension. Aliquots (0.5 ml) of sample dilutions and the control were then added to the appropriate cell suspension cuvettes. After 5 minutes, final light measurements were made for each cuvette. The control was used to correct the sample for the time dependent drift in light output. Calculation of EC50 — Microtox data reduction was more precise when the gamma function (2) was used in place of percent light decrease which was used previ- ously (1). Gamma ( ) is the ratio of the amount of light lost to the amount of light remaining (figure 2). Thus, for the case in which a 50% light reduction is observed (EC50), =1. The procedure for calculating EC50 is as follows: A blank ratio (BR) was determined by measuring the light intensity of the control cuvette prior to addition of diluent (B0) and five minutes after diluent addition (B5). This ratio was applied to initial light intensity readings (I0) of the cell suspensions prior to sample addition. BR corrects I0 values for drift and other effects of diluent addition. This enabled the measurement of a true base line and thus the isolation of toxicity effects alone from the sample. B R = B _5_ c^o ~ !oBR The corrected initial light output (CIQ) and 5-minute light output values (I5) were used to calculate r^ for each sample dilution. corrected light lost _ c o - I c o , 5 light remaining _ 1^ 5 Values for T^ were plotted as a function of sample concentration on log-log graph paper. EC50 was determined by the intersection of a best fit line with r5 = 1.0. Procedure for Determination of Reproducibility — The reproducibility of this assay procedure was assessed by testing concentrations of sodium lauryl sulfate (SLS) (0.3, 1, 3 and 10 mg/liter). Tests were performed with three different 174 lots of Microtox reagent by three technicians on different instruments. Each technician performed nine assays with each reagent lot for a total of 81 determi- nations of EC5 0. Assay of Pure Compounds — Compounds were dissolved in the 2% NaCl diluent and appropriate dilutions prepared for testing. Those compounds which had limited water solubility (e.g., Aroclor 1242) were mixed with diluent for 2 days in a separatory funnel and the water phase removed for testing. It was assumed that this procedure saturated solutions. Literature values for concentrations of saturated solutions were used for the results reported. Assay of Complex Effluents — Complex effluents tested were obtained from vari- ous industrial and municipal sites. These samples were assayed using conven- tional fish and invertebrate assay procedures while concurrently assayed using the Microtox procedure. These comparative assays were performed by government, academic, and industrial laboratories. These laboratories have been evaluating the luminescent bacterial assay and have fish-testing capabilities. RESULTS Variation of Microtox Assay — All 81 determinations of EC50 were used to calcu- late a mean value of 1.57 mg/Liter. The data displayed a variance of 0.08 and a standard deviation of 0.28 mg/Liter. The coefficient of variation for all EC50s was 18.2%. Coefficient of variation within lots were between 6 and 10%. There were no significant differences between operators and instruments. Assay of Pure Compounds — Pure compounds which were assayed with Microtox are listed in table 1. Fish LC50 data were not experimentally determined but were obtained from the literature (4-11). In most cases a range was reported since the data were determined using several species and/or different testing periods. These data are presented graphically in figure 3. If the Microtox and fish data displayed equal sensitivities to the compound tested, all the data points would fall on the line (slope=unity) in the figure. Compounds which fell below the line were more sensitively detected by the fish bioassay, while those compounds which fell above the line were more sensitively detected by the Microtox method. Complex Effluents — Data in table 2 present the side-by-side comparisons of the two methods, used to test effluents from industrial and municipal plants. Sam- ples were coded because of the sensitive nature of the information. EC50 data reported as greater than 50% were estimated by extrapolation. These complex effluent data were plotted in figure 4 such that correlation could be estimated in the same manner as for pure compounds data. Table 2 contains more EC50 data than were plotted in figure 4, but those samples which gave no quantitative effect with both assay methods could not be presented graphically. DISCUSSION When comparing Microtox and fish sensitivities (figure 3), the trend was one of good general agreement between these two methods of bioassay. Defined toxi- cants were selected for an initial comparison of sensitivities since pure 175 compounds could be assayed quickly with Microtox and fish data were readily avail- able from the literature. The data suggest that 5-minute Microtox EC50 data are comparable with fish 24- to 96-hour data. These data also suggested the need for a direct experimental comparison using identical samples, particularly complex effluents. The data presented in table 2 and figure 4 show good agreement between EC50s for luminescent bacteria and LC50s for fish. It is important to note, however, that there was no rigorous attempt to account for differences in test conditions such as temperature, pH, water hardness, and other factors which might affect results. Some samples had coloration which may have caused nonspecific light decrease by light absorption. This is particularly true of reds and browns which absorb strongly in the 480-nm region, which is the peak of the biolumi- nescence. Sample number AA20 was dark brown in color and is the only sample suspected of a significant color interference. Samples AW24 (dark blue) and AA33 (light brown) showed no apparent effect due to color. Close inspection of the data in table 2 shows that 19 of the samples caused no effect or were slightly stimulatory with Microtox and caused no lethality with the fish bioassay. If these data could be plotted in figure 4, it would further support good correlation between 5-minute EC50 data and fish lethality data. The data in table 2 also suggested that invertebrate data for complex effluents did not correlate well with either fish or bacterial assay results. The modified assay procedure used to generate data in this report has shown several advantages. The procedure incorporates the simultaneous assay of a control as well as dilutions of sample. This permitted challenge of bioassay organisms, at the same time following reconstitution, and thus assured that the bacteria were physiologically identical when challenged with control-plus- sample dilutions. This improved assay procedure considerably shortened assay time. The entire assay including dilutions of the sample could be accomplished in 30 minutes from the time samples were collected. The reproducibility experiments using SLS represent a worst case since the test included different operators, instruments, and production lots of reagent. In addition, SLS represents a class of toxicant in which reaction rate is still rapid at 5 minutes. This fact made end-point precision difficult and thus entered into the reproducibility of the assay. It is felt, however, that a coefficient of variation of 18.2 percent is acceptable variability for a biological assay. CONCLUSIONS The Microtox bacterial luminescence bioassay method was evaluated, and the data procedure found to be a reliable method for determining acute toxicity for complex effluents. The 5-minute bioassay procedure was shown to agree favorably with 24- to 96-hour fish bioassays. An improved testing procedure was reported which allowed a complete determination of EC5 0s in 30 minutes. The Microtox system appears to be a convenient, reliable approach for rapidly determining acute toxicity of water samples. 176 REFERENCES 1. Bulich, Anthony A. "Aquatic Toxicology," ASTM STP 667, American Society for Testing Materials, Philadelphia, 1979. 2. Johnson, Frank H. , Eyring, H. , and B. J. Stover. "The Theory of Rate Process in Biology and Medicine." John Wiley and Son, New York, 1974. 3. McKim, J. M. , Anderson, R. L. Benoit, D. A., and Stokes, G. N. Journal of the Water Pollution Control Federation, Vol. 48, 1976, pp. 1544-1620. 4. Mattson, V. R. , and Walbridge, C. T. "Acute Toxicity of Selected Organic Compounds to Fathead Minnows." Ecological Research Series 600/3-76-097 U.S. Environmental Protection Agency, 1976. 5. Parrish, P. R. , et al. , "Chronic Toxicity of Methoxychlor, Malathion, and Carbofuran to Sheepshead Minnows." Ecological Research Series 600-3-77-059, U.S Environmental Protection Agency, 1977. 6. Tubb, R. A., ed., "Recent Advances in Fish Toxicology." Ecological Research Series 600/3-77-085, U.S. Environmental Protection Agency, 1977. 7. Weber, J. B. , Environmental Science and Technology, Vol. 11, 1977, pp. 756-761. 8. Brungs, W. A., Journal of the Water Pollution Control Federation, Vol. 49, pp. 1425. 9. Bond, R. G. , and CP Straub, eds., CRC Handbook of Environmental Control. Vol. 3. pp. 349-388. CRC Press, Boca Raton, Florida, 1973. 10. Brungs, W. A., Journal of the Water Pollution Control Federation, Vo. 50, 1978, pp. 1582. 11. Cairns, J., et al., "Effects of Temperature on Aquatic Organism Sensi- tivity to Selected Chemicals." Virginia Water Resources Center Bulletin, 106, Blackburgh, Virginia, VPI, 1978. 177 Figure 1. - Microtox Toxicity Analyzer 178 Toxicant added Czero time) 0% Lost (Relative Light Level) 100% Lost % A = V??,LOfffr X100 = ^b-x100 = -gxlOO = 40% Total Light a + b 100 r = ...! L!9i?t LOSt. =# = ^=0.667 Light Remaining b 60 %A = 7-^x100; T= ^ 1+r«.ww, - 100 -%A Note: A 50% effect is equal to a gamma of 1.0. %A — r Conversion %A r 1 10 5+053 -.111 15 --.176 20 25- -.333 30 35 --.538 40 45- -818 50 60 70 80 90 100% %A 0.01 0 .250 .429 •667 ■1.0 55--1.22 •1.5 65-1.86 2.33 75- -3.0 ■4.0 85 --5.67 9.0 95--19 99.-99 OO %A = % Decrease of Light Output Figure 2. Comparison of Gamma vs. I Decrease 179 RESPONSE TO PURIFIED TOXICANTS 100,000- 10,000- 1,000- X 100.0 8 o (0 il 10.0 - 1.0 - 0.1 - .01 — /% Ethanol Isopropanol •j f Urea N02. / - Phenol Xor • y Benzene. •SLS / p-Cresol. ™T / Malathion/ Zn» / - PCB1242/^U|| yfcp #NH3 - / #HCN /*HGII - 1 I I 1 I j^ j^ .01 0.1 1.0 10.0 100.0 1,000 10,000 Microtox EC50 (5 Min) 100,000 Figure 3. Comparison of Microtox pure compound EC50 data with fish LC50 data obtained from the literature. 180 0.01 COMPLEX EFFLUENTS (PERCENT V/V) 100 • • • • • • • •/ 10 - • ^ X y% (0 0) ■ /m - *■ w Fish LC50 b jf • 0.1 - • ^r • 0.01 J 1 _J i- 0.1 1.0 Microtox EC50 (5 Min) 10 100 Figure 4 Comparison of Microtox EC50 data with fish LC50 data using complex effluent samples. 181 TABLE 1 - COMPARISON OF MICROTOX EC50 AND FISH LC50 DATA USING PURE COMPOUNDS Toxicant mg/Liter TM Microtox Fish Assay 5 min. EC50 24-96 hr. LC50 Mercury II 0.065 0.01 0.9 Pentachlorophenate 0.5 0.21 0.6 Aroclor 1242 0.7 0.3 1.0 p-Cresol 1.5 3.5 - 19 Sodium Lauryl SO* 1.6 5 - 46 Ammonia (free) 2.0 0.068 8.2 Benzene 2.0 17 - 32 Zinc II 2.5 0.24 7.2 Malathion 3.0 0.07 19.5 Formaldehyde 3.0 18 - 185 Copper II 8.0 0.1 - 10.7 Cyanide (HCN) 8.5 0.1 0.44 Trinitrotoluene 20 26 Phenol 25 9 - 66 Chromium VI 70 29 - 133 Nitrite 420 19 - 230 1-Butanol 3300 1940 Isopropanol 42,000 4200 -11,130 Urea 24,000 12,000 Ethanol 31,000 13,500 values obtained from references 4-12 182 TABLE 2 - Same Sample Comparison of Complex Effluents5 LUMINESCENT BACTERIA1 SAMPLE CODE EC50C WP26 0.032 WP27 0.2 WP28 0.48 WP25 1.3 AA20 1.6 KP40 2.6 AM 7.3 3.2 KP42 3.5 KP43 4.3 AW21.1 7.0 AP11.1 11.5 KP41 13 AP11.2 15 AC 6 21 AW22 21.8 AC34 24 AC 5 25 AP16.3 40 AC23 48.5 AF36 60 AP16.1 62 AC 3 62 AC 1 68 AM 7.4 70 AM 7.2 74 AC43 82 AC 8 >100 AC44.2 >100 AM 2 >100 AC12.2 >100 AC12.1 >100 AW21.2 NEg AM 7.3 NE AC 4 NE AP16.2 NE AC45 NE AC37 NE AM 7.5 NE AM10 NE AW14 NE AC29 NE AC30 NE AH18 NE AW24 NE AF35 NE AA33 NE AC31 NE AC39 NE AB32 NE AC44.1 NE AC19 SS FISH INVERTEBRATES LC50C TYPEd LC50C TYPEd 0. 1 F96RT -- -- 0. 07 F96RT -- -- 0. 46 F96RT -- -- 0. 5 F96RT -- -- 21. 7 S24FM 100 S24D 2. 4 S96FM -- -- 3. 5 S24FM -- -- 3. 5 S96FM -- -- 29 S96FM -- -- 19 S24FM <5.6 S24D 48 S24FM 42 S24D 12. 2 S96FM Partialf -- 68 S24FM S24D 17. 9 S72FM >18 S72D 42 F96FM -- -- 13. 5 S24SM 21.5 S24MS 32 S24FM <56 S24D 66 S24FM -- -- 45 S96SM 35 F96MS NLe S24FM -- -- 52 S24FM -- -- NL S72FM 39.7 S72D NL S72FM 61.4 S72D 100 S24FM -- -- NL S24FM NL S24D 66 S24FM 80 S24D 21. 3 S24FM 27.2 S24D 74 S24IM -- -- NL S72FM 35 S72D NL S24FM Partial S24D NL S24FM -- -- 51 S24FM 100 S24D 56 S24FM 38 S24D 65 S24FM 78 S24D 75 S24FM Partial S24D 91 F96FM 23 F96D NL S24FM 42 S24D NL S24FM 80 S24D NL S24FM 100 S24D NL S24FM Partial S24D NL F96SM NL S96MS NL F96SM >56 F96MS NL S24FM NL S24D NL F96FM NL S48D NL F96B -- -- NL F96SM NL S96MS NL F96SM -- -- NL S96B -- -- NL F96SM NL F96MS NL S24B -- -- NL S24FM 81 S240 183 TABLE 2- Continued LUMINESCENT BACTERIAU EC50C FI SH INVERTEBRATES SAMPLE CODE LC50C NL NL NL NL TYPEd S24FM S24FM S24FM S24FM LC50C TYPEC AP16.4 AX13 AW15 AH 9 SS SS SS SS Partial S24D Partial S24D Partial S24D Partial S24D a Data furnished by independent evalua tors (listed in acknowledgments.) b 5 minute data c Percent Effluent causing 50% effect or 50% lethality. d Assay type: F=Flow thru; S=static; #=hours; RT=rainbow trout; FM=fathead minnow; B=bluegill; SM=sheepshead minnow; B=daphnia; MS=mysid shrimp e NL = no lethality f Partial = insufficient deaths for calculation of LC50 * NE = no significant effect SS = slight stimulation 184 THE USE OF ALGAE AS BIOLOGICAL ACCUMULATORS FOR MONITORING AQUATIC POLLUTANTS by Drs. Stephen L. Friant and John W. Sherman Academy of Natural Sciences of Philadelphia Division of Limnology and Ecology Philadelphia, Pa. 19104 INTRODUCTION Increasing awareness of and concern about the degradation of water has resulted in a growing body of knowledge concerning the aquatic environment in the natural state and stressed by pollution. The types, concentrations, and persistence of aquatic pollutants have been major research interests in the past decade. Kaiser (1971) estimated that as analytical detection limits increased from the ppm to the ppb range, the identification of trace chemical components could increase from 10-* to 10". While the accuracy of Kaiser's prediction cannot be determined, over 200 organics have been identified in drinking water alone. Sheldon and Hites (1979) identified over 100 organics present in the Delaware River with a significant number of the compounds originating from anthropogenic sources. A similar situation has been found for many other rivers in industrialized areas. The development and application of computerized gas chromatographs/mass spectrometers equipped with capillary columns permits the separation and subsequent confirmation of complex mixtures of trace organic pollutants. Concern about the effects of many heavy metals in the aquatic environment is clearly demonstrated by their inclusion in the Environmental Protection Agency Priority Pollutant List. Routine detection limits for metals have been extended through the use of the graphite furnace, atomic absorption spectroscopy, and plasma emission spectroscopy. As a result of these improved procedures, a rapidly expanding body of litera- ture describing levels of aquatic pollutants in both abiotic and biotic environ- ments has developed. Current research is focusing upon the determination of the biologically active forms of pollutants at the ug/1 level and evaluating syner- gistic and antagonistic interactions among pollutants within organisms. The levels at which pollutants are available to biota are not necessarily levels that can be detected analytically or the levels that theoretically would bioac- cumulate. A laboratory estimate of potential accumulation of aqueous organic pollutants is provided by the octanol-water partition coefficient (Chiou et al. , 1977). In the environment, however, there are many competing mechanisms or environmental influences that could either increase or decrease pollutant avail- ability (Hassett and Anderson, 1979). The levels, rates and persistence ulti- mately determine this availability. Biologically available levels of metals depend upon metal-complexation with inorganic and organic substances which may be present, precipitation or solid sorption (fig. 1). 185 Ingestion Macrophytes 45-285 ppm Attached Periphyton 28-98ppm wiacropnytes 4t>-z»t>ppm Bivalves 109-414 Figure 1. — The partitioning of zinc in the aquatic environment. Concern about the quality of surface waters and the origin of specific pollu- tants has promoted the development of remote samplers and monitors to obtain time-integrated water samples for analysis. While these units capably collect and monitor the discharge of industrial wastes, there would appear to be little application in monitoring certain aspects of industrial wastes, long-term general water quality, and especially the biological availability of pollutants. It is reported in the literature that analyses of a wide variety of parameters are possible automatically with in-situ monitors, when in fact, the analyses of only a few parameters by in-situ sensors have been proven reliable (Pijanowski, 1974). These include conductivity, temperature, depth, dissolved oxygen, and pH. Some instruments offer specific ion sensors but, because of the lack of specificity, these sensors are useful for only limited and specialized applica- tions. Each specific ion electrode must have a reference electrode and ampli- fier. Ingols and Craft (1975) point out that the choice of specific ion elec- trodes for monitoring river water quality will depend upon concentration range, and the ease and frequency of available service for the sensor. They felt that although specific ion electrodes are useful in the laboratory, field use is extremely limited. 186 Concurrent with the development of monitoring devices to measure general water quality and the presence of toxic substances has been the increased use of bio- monitoring, i.e., the use of organisms to determine the state of an aquatic environment (Cairns and Dickson, 1973). Weber (1973) divided biological moni- toring methods into three categories: "methods for studying the effects of pollutants on indigenous communities of aquatic organisms in receiving waters, methods employing captive organisms or controlled conditions to determine the toxicity or other biological effects of substances or conditions in effluent or receiving water and methods employing bioaccumulation or biomagnif ication to detect or monitor trends in the concentration of pollutants in surface waters." However, pollutant profiles obtained from biological organisms can be misleading due to the effects of extraneous variables on the physiology of the organisms used and because different organisms acquire different components of the total pollutant load, resulting in different profiles for each organism. For an organism to be useful for biomonitoring it must meet certain criteria. In a summary of these criteria Phillips (1977) stated the following. 1. The organism should be exposed to subacute levels of the pollutant. 2. The organism should have limited spatial mobility. 3. Significant abundance of the organism should exist in the study area. 4. The organism should have sufficient longevity to satisfy study objectives. 5. The organism must provide sufficient biomass for proper pollutant detection. 6. The orgnism should be adaptable to permit laboratory accumulation studies. 7. Monitoring program should be adaptable to both fresh and brackish water. 8. The organism should exhibit a pollutant concentration factor that permits detection. 9. A known relationship should exist between accumulation and exposure concentrations. 10. The organism should demonstrate a relationship between uptake and exposure concentration in the study area. The intent of this paper is to present an overview of possible biomonitoring uses of algae as aquatic pollutant bioaccumulators. A discussion of the factors that determine bioavailability of metals to bivalves and algae as well as the general relationships between exposure concentrations of metals and algal uptake is included. Original data from studies using algae growing on artificial substrates to monitor point-source trace metals and nonpoint-source organics are presented. 187 BIOLOGICAL AVAILABILITY OF METALS Estimating the biological availability of metals in water column samples by direct measurements presents difficulties because of many competiting reactions, including complexation, precipitation, and sorption. Also, frequent sampling is required to establish any water column trends. Luoma and Jenne (1976) attempted to isolate by chemical extraction procedures that fraction of cadmium, cobalt, silver, and zinc in sediments biologically available to deposit-feeding clams. They concluded that the type of extractant solution needed to estimate biologically available metals varied depending upon the metal and its form in the sediments. It is unlikely that any single analyti- cal procedure will describe that fraction of metals in sediments that is biologi- cally available. In contrast to water or sediment analysis, a direct indication of the biological availability of metals can be derived from an analysis of biological organisms, the single most important advantage of such an analysis, according to Phillips (1977). He extensively reviewed the use of biological organisms to monitor aquatic metal pollutants and pointed out that indicator organisms have been used by many re- searchers to monitor time-averaged concentrations of trace metals (Phillips, 1977). Pollution profiles obtained from the analysis of organisms will vary depending upon habitat and their interaction with the physico-chemical environment. Use of organisms without knowledge of mechanisms of interaction with the environment could result in an improper interpretation of the study results with respect to pollutant loading (Phillips, 1979). Metal accumulation by oysters has been reported to be not necessarily a func- tion of a particular source but also a function of location of the organisms in the estuary (Huggett et al., 1976). This may be due in part to the nonhomo- geneous nature of sediments, which may or may not reflect metal activity in the water column depending upon water column physico-chemical removal mechanisms to the sediments. Batley and Gardner (1978) point out that in the mixing zones of estuaries the most significant changes in metal speciation will occur. Greig and Wenzloff (1978) studied the uptake of cadmium, copper, silver, and zinc by the oyster Crassostrea virginica and found that only cadmium and silver were obtained directly from the water column. Copper was suggested to have been derived directly from seawater and/or food, suspended in sediment, or from sedi- ments directly. No relationship was found to exist between zinc concentrations and the environment. Greig (1979) showed that in ocean quahogs, surf clams, and oysters at least part of the total body burden of cadmium, copper, and silver is directly available from the water column. What fraction this represents of all environmental inputs was not reported. Phillips (1979) investigated the accumulation of cadmium, iron, lead and zinc by the mussel Mytilus edulis (L) in a field study. No relationship was found to exist between tissue metal concen- trations and water concentrations. Anderson (1977) studied the concentrations of cadmium, copper, lead, and zinc in six species of freshwater clams found in the Fox River. Considerable variation in metal concentrations was found to exist between taxa and between 188 individuals of the same taxon. A correlation was found between organism metal concentration and sediment metal concentrations. It was suggested that since the gills act as both a filtering mechanism for food and as a respiratory organ, metal could be sorbed from both water and substrate. It would appear from these and other studies that metals accumulated by bivalves represent an integration of metals from food, sediment, and the water column. The factoring of any particular mechanism of accumulation would apparently be very difficult because of the many interactions. Algae, on the other hand, present a different pollution profile of their environment. Metal accumulation by algae is apparently directly related to water-column metal concentrations and, for copper at least, this relationship exists due to free ion activity. Sunda and Guillard (1976) reported that bioavailability of copper to algae is a function of free ion activity rather than total metal present and have suggested that this relationship is probably not limited to copper. A similar copper-availability relationship was also reported by Gachter et al. (1973). The free ion relationship to availability is apparently not limited to algal uptake (Sunda et al. , 1978) but constitutes a fraction of the metal accumulated by other organisms. Metal activity is regulated by pH, ionic strength, particulate carbon, and chelating agents. As discussed by Cross and Sunda (1978), the energy of a chemical reaction is supplied by metal free-energy or metal activity. Biological availability is also dependent upon the molecular interaction between the metal and the organism, occurring at the contact site of the organism. Conway and Williams (1979) studied the uptake of cadmium by diatoms and found significant amounts of cadmium, at least initially, were present at the organic coating of the cell wall. It was suggested that ion exchange with a constituent of the polysaccharide coating could occur. This would suggest that metal competition at the reaction site is also a key factor in ultimately determining metal uptake. Kinkade and Erdman (1975) found that the rate and level of uptake for cadmium in a variety of freshwater organisms was affected when the water was varied from soft to hard. Similar results were reported by Guthrie and Cherry (1977) where different concentra- tions of heavy metals altered the bioaccumulation of all metals in a natural system. These investigations suggest that metal biological-availability is regulated by interactions that can decrease metal activity. The interaction of organic and inorganic complexes in the aquatic environment which reduces free-ion activity is difficult to measure and predict since these values will vary seasonally and from river to river and river to estuary. McCrady and Chapman (1975) found that copper added to natural river water always resulted in copper-ion concentrations that were less than theoretically calcu- lated concentrations. From the preceding discussion it is apparent that not only does analysis of biological organisms for metal content provide an estimate of biologically- available metal in the form of a recent history of exposure integrated over time but also reflects varying environmental factors including pH, ligands, competition of metals, and the natural background matrix. 189 ALGAE AS BIOACCUMULATORS OF METALS In the analysis of a pollutant from any matrix, i.e. water, tissue, or sedi- ment, the method of isolating the pollutant is a key consideration since it defines the compounds that will be isolated and their recovery, together with possible coextractives, may interfere with the analysis. Each isolation method has some degree of selectivity toward a specific compound, usually related to polarity, molecular weight, chemical speciation, etc. Similarly, biological systems can be thought of as a specific isolation method that defines the types of pollutants accumulated and the degree of accumulation. It is necessary, then, to understand the capabilities of bioaccumulation so that organisms may be properly used as an analytical tool in environmental monitoring. Freshwater and marine algae have been successfully used for water quality monitoring for years by studying organismal response to changes in the aquatic environment (Cairns and Dickson, 1973; Patrick, 1973). In addition, metal accumulation by algae has been used as a tool by researchers to monitor water- column trace metals resulting from point- and nonpoint-source pollution. This approach is possible because of the proportional relationship of metal bioaccu- mulation to exposure concentration (Phillips, 1979). Foster (1976) reports that this relationship should hold except for grossly polluted areas. Trollope and Evans (1976) point out that since no concentration procedures are required, metal monitoring using algae can avoid many of the difficulties associated with metal analysis in very dilute solutions. The collection and use of algae is simple and inexpensive and studies can be performed on a continuous or semi- continuous basis. The pollution profile obtained is a recent historical inte- gration or reflection of metals in the environment. Foster (1976) points out that this integration takes place over time intervals long enough to moderate the influence of fluctuations caused by short-term variations in water column metal loading. Recent literature has reported the use of algae as biomonitors and accumula- tors of metals including discussions of mechanisms of metal uptake and relation- ships that have been used to describe this uptake. Trollope and Evans (1976) profiled the concentration of copper, iron, lead, nickel, and zinc in freshwater algal blooms and found that differences in algal metal content were related to environmental effects. They found a linear relationship betwen zinc content of algae and that in the surrounding water, although no clear relationship was found to exist for the other metals. Biological accumulation factors ranged from 10^ to 10^. Martin and Broenkow (1975) and Windom et al. (1973) reported the use of oceanic plankton to monitor areas of pollution for cadmium and mercury, respectively. It was reported that mercury accumulation was independent of biological composition. In a similar study Burkett (1975) used freshwater algae to detect the presence of mercury. A knowledge of the relationship between metals and chemical speciation is extremely important in assessing the utility of a pollution profile obtained from algae. Skaar et al. (1974) reported that uptake of ionic nickel by a marine diatom was directly related to media concentration. Similar results 190 were reported for zinc in freshwater algal blooms (Trollope and Evans, 1976) and for natural communities of periphyton in a closed lotic microcosm (Cushing and Rose, 1970). The uptake of mercury by a freshwater diatom (Fujita and Hashizume, 1975) and lead for a marine diatom (Schulz-Baldes and Lewin, 1976) has been described by Freundlich adsorption isotherms; x/m = kCe^-'n, where x/m is the bioaccumulated concentration, Ce the equilibrium test concentration, and k and n are constants. The fit of the data for mercury and lead accumulation to the above equation suggests that at least a portion of a metal is accumulated by passive uptake (Filip and Lynn, 1972; Parry and Hayward , 1973; and Fujita et al. , 1976). Cushing and Rose (1970) showed that zinc uptake decreased with the addition of magnesium because of increased competition for adsorption binding sites. Organically complexed cadmium (Cossa, 1976) and iron (Goldberg, 1952) have been reported to be biologically unavailable to algae. Similar results have been reported for copper (Sunda, 1976), mercury (Fujita and Hashizume, 1975) and nickel (Skaar, 1974). The majority of the previous studies used unialgal cultures run in batch. The use of artificial substrates for periphyton-growth studies has presented an alternate approach for pollutant accumualtion studies (Patrick et al., 1975, and Johnson et al. , 1978). Periphyton colonizing artificial substrates have been used extensively as indicators of water quality (Patrick, 1973; Patrick et al., 1975; and Weitzel, 1979). A device, termed the Catherwood diatometer, permitting the continuous devel- opment of diatom communities on glass slides in natural and artificial settings (fig. 2) was developed and used by Patrick. Research employing the device in natural streams indicated that 75 to 80% of diatom species found in hand collec- tions would be found colonizing the glass substrates over a two-week exposure period. Benthic, epiphytic and planktonic diatoms are well represented on ex- posed slides. Use of the substrate enabled Patrick to quantify diatom community growth with respect to biomass and community structure (Patrick, Hohn, and Wallace, 1954). Variations within the structure of the communities were related to conditions such as stress caused by organic loading or toxicity in the aquatic environment (Patrick, Hohn, and Wallace, 1954, and Patrick and Strawbridge, 1963a; 1963b). Diatometer monitoring studies have been undertaken on rivers and estuaries across the United States; many of these ongoing programs have been continuous for over ten years, a few providing diatom growth records for up to 20 years. More recntly, Patrick (1967) reported on a system for studying natural com- munities of diatoms under laboratory conditions. Artificial substrates were placed in artificial stream beds located in greenhouses and exposed to natural stream water supplied by pumping systems. The test chambers allowed developing diatom communities to be subjected to controlled variables. Among others, completed experiments have involved amounts and ratios of nitrogen and phosphorus (Patrick, 1967), changes in pH and temperature (Patrick, Roberts, and Davis 1968), interrelationships between temperature and manganese concentrations (Patrick, Crum, and Coles, 1969), and variations in day length and temperature (Patrick, 1971). Patrick, Bott and Larson (1975) reported on the effects on diatom growth of vanadium, chromium, selenium, boron, nickel, and rubidium in varying concentra- tions. Tests employing substrates were again conducted in greenhouses during 191 Styrofoam Glass Slide Styrofoam Float Figure 2. — Catherwood Diatometer various seasons of the year. These experiments included analyses for accumula- ted test metals. Vanadium, chromium, selenite, and nickel were found to accumu- late in diatoms in proportion to test water concentrations. From these studies, artificial substrates would seem particularly suited for use in biomonitoring/ bioaccumulation studies, especially studies involving point-source discharge. BIOMONITORING CHEMICAL POLLUTANTS USING ALGAE ON ARTIFICIAL SUBSTRATES Trace Metals Study — A recent study was conducted on the Wateree River, South Carolina, to monitor trace metal levels in the area of a point-source discharge (Friant and Koerner, in press). Algal communities developing on artificial substrates were used as biomonitors in the river, which receives a secondarily treated effluent from a synthetic chemical manufacturing plant. Catherwood 1/2 mi above discharge (St Substrate sli replaced for Recovered sli cally identif and analyzed antimony, chr diatometers were placed at three stations: a reference station the discharge (Station 1), a station immediately downstream of the ation 2), and a station 1/2 mi below the discharge (Station 3). des were exposed in the river for two-week periods, removed and a total of 10 successive exposures from April to October 1977. des were air dried, acid digested with nitric acid, and biologi- ied. Duplicate slides were acid digested with hydrochloric acid by atomic absorption spectroscopy for biological accumulation of omium, copper, mercury, and zinc (Friant and Koerner, in press). Concentrations of biologically accumulated chromium, copper, and zinc for the three study stations are shown plotted for each exposure period in figures 3, 4, and 5. The corresponding box and whisker graphs (Tukey, 1977) illustrate 192 [/) 1 1 1 1 1 1 1 i i i i i i- 1 1 1 H h i i i i H 1- i - 1 1 -J— 4 /• r / / / / ,-"* •*** / .**•' / / # I *v X s . "•. X *X, f% s > O* i /C 1 .♦•' >v ' *■" > / • X / / / * / J r / •* / • *'' / r' < t i •- / •-... i i i Mill J 1 J _. 1 (0 CM CO (0 55 O) in 0> CN r- CN . 00 0) ^ ■ oo CN CM CO 00 CO o o o o o o a. UJ a UJ cd •H o c • +-> CO r^ co r-^ E /— >CJ> • rH 0 * u r* vL-' a; •P JO on e 0) c^n a CO CO «H OOP >s rH CO CO cd P S CO s O +-» CO H CO C MH -h 13 rH "3 O o T? p co CO /-> J-t - OOP J-t \ c/i 0) 00.O > 3- 3 -H ■H e r-. 0) p <0 0> K w> h 0) D G CD CD N h- t/) o 0. X O P H •H CD • o o • ** UD r- E w x: > '*., 4 — ^ ""*••. \ \ '"■•\ N V**'« \ \ > — \ ..V X ( > - W \'< pl - \ : \j k _ ..•••"/ .••"" / <" >1 > - "••. •' SK * < 7 1 - v / • x-/ \ i i i •4 * 1 1 1 1 1 1 i i i o o o o co CN CO +■• (0 CO CN CN CO CO CO 0) CD 1- CN CO CO CN CN lO o lO & Oi CN CN 00 o 0* CO uu -~. a CO uu 0) or D ^ to O 0. CO X LU h- ^ O "P T3 CD C c/: CO 43 00 +-> i— ( 3 S o O W ^ P G J= •H r-l 4-> >n cd cd cd oop s: rH CO aj •» P Cd E cd G O -H r.T3H M-l o o CO Si T3 O Cd cd cuu C X! •H CD J3 E 4- J-. CO 3 CD CD O ■P 4-i CO CD cd T3 p. * 4-> P /-> CO CD GfljD > ^ 3 -H DO CO CC v— ' Si CD CD CD (O P P C CD CD O E -m •iH o cd +-> -p ;s re cd P 'H CD +-» TD 4= C P CD E U O G G P -H O Sh U r- > T3 • P CD • CD > • Oh O • O CD . U rttO (w6/6rv) NOILVH1N30NO0 d3ddOO OOl CD rH 3 00 Uh 194 T — r i i i i i — r O O H o h H J L i i i i i i i O O o o (oiB/Bry) NOI1VU1N3DN00 ONIZ OOl CO no • I— 1 C t- cd cd r>» •H a» fH --7M-I 03 CO 4-> Cd 1 ^ CO E v-L/Xl +J E CO CD 4-> +J /— N CD o 00 CM T3 X CO C v— ' C50 *-< cd 3 (/) x: 4-> o rH O U CO JS c +-> o T™ u •H X G0 4-» f,5 CO cd S *-< rH +J (/) CO GO CO * cd «H lO 1— rH e*a o 05 ' d o 0) Jh co cd o u T~ lO a 0) ~ 0) T3 CD C CO 4-> 3 CN T™ •H to o CN _ . v^ e 0) C/3 CO 0) CD 4J 4-) cd •» CO CN - Ci 0) 4-> CD 00 CO Q o H3 CO > JO.H* or hi a r-\ 3 a o o 60 to CO \ CO GO U CD UJ 3. CD Jh CD 0) QL v_ / 4J CD D CD -P h- |s. co CO E cd O a C o s o 4-> CN _ CO X UJ •H cd CD •H _C CO r^- cd t3 +-> ?h CO CN _ .CN E C O »H if) CD CD 5-i U M-) V-> CN (0 - CN C 10 vO u u CD • > • O • CO CN _ CN e S^ ■-N, •H CD «tf m rsi u to IT) CD H GO •H tin 195 the distribution of values by range, quartiles, and median. Accumulated anti- mony and mercury were not found in sufficient quantity to warrant consideration in the study. To determine whether a station-related trend in metal accumula- tion existed, the time-series data were analyzed by a two-way analysis of variance. To account for the nonnormal distribution of values (apparent in the box and whisker graphs), a distribution-free version of the analysis of variance was used. Medians were used to express concentrations. Statistical analysis revealed that chromium was present at Station 2 in sig- nificant amounts (p = 0.01) (table 1). A highly significant accumulation of chromium existed at the discharge station, Station 2, and in somewhat lower con- centrations downriver at Station 3. Information concerning plant processes sup- plied by the chemistry plant personnel revealed that of the five metals analyzed only chromium was present in the process water and intermittently present in the treated effluent. This intermittent release of chromium may explain the large range of values for successive exposure periods determined for samples from Sta- tion 3 (fig. 3). Station 3 samples, although having higher values than the ref- erence station samples, were similar in range. Elevated chromium levels at this downriver station were probably due to river dilution of the discharge chromium. Analyses for other metals did not reveal any significant station-related accumulation (figs. 3 and 4). Chemical analyses of sediment samples from the study area failed to show any statistically significant accumulation of any metal. The lack of evidence for station-related accumulations of chromium in the sediments, while finding sig- nificant station differences in bioaccumulation, suggests that the bioaccumu- lated concentrations reflect the water column concentrations. These data appear to present a more realistic reflection of the chromium in the discharge than routine sediment analysis would provide. The chromium concentrations determined in the diatometer analyses represented the biologically available form isolated from natural algal communities (predominantly diatoms) colonizing the substrate. The consistent station differences seen in figure 3 suggest that any single exposure period of two weeks would have detected the presence of chromium at Station 2. Trace Organics Study — To test the feasibility of algal biomonitoring of trace organics, a study was carried out in the estuary portion of the Delaware River during the spring and summer of 1979. A series of 30 diatometers was placed from above the city of Philadelphia at the Torresdale Water Treament Plant Intake (river mile 110) to center city (river mile 100). The diatometers were grouped and slides within a group were pooled, thus providing three discrete samples representing areas of varying river pollutant loading (Sheldon and Hites, 1979). Exposed slides were removed and replaced every two weeks during the 10-week study period comprising a total of 5 exposure periods. Organics were extracted from samples using hexane and were profiled by gas chromatography using (1) electron capture detection for polychlorinated biphenyls and (2) flame ionization detection. The specific gas chromatographic conditions are listed in table 2. Biological sample preparation involved scraping the exposed slides followed by the sample preparation procedure shown in figure 6. 196 Table 1« — Medians of chromium, copper, zinc, and antimony concentrations in algae, based on 10 consecutive 2-week exposures of Catherwood Diatometers at three stations on the Wateree River, South Carolina; April through Sep- tember 1977. Analyses of variance on ranks for station differentiations. Med ians l yg/g) Analysis of Va riance Degrees of Sta. 1 Sta. l Sta. 3 Freedom F P Chromium 41 115 65 2/17 23.06 0.0001 Copper 51 77 51 2/17 2.86 0.0851 Zinc 45 45 42 2/17 0.51 0.6092 Antimony 1 1 1 2/17 0.14 0.8713 Table 2. — Gas chromatographic conditions for analysis of trace organics isolated from algae grown on artificial substrates. ECD 63Ni - Detector Column - 3% OV-101 Chromosorb 80/100 6 ft x 1/4' Isothermal 0 190°C Inlet - 200°C Detector - 350°C Carrier - Nitrogen Make-up - Argon/ 10% Methane FID - Detector Column - Capillary 30 meter OV-101 PTGC - 50-260°C at 7°C/min Initial hold - 3 min. Inlet - 200°C Detector - 225 °C Carrier - Nitrogen 197 Analytical Scheme Slides (Biological Material) Hexane Eluate Air dried Scrapped I I Soxhlet Extraction 6 hours 1 . Hexane or 2. Dichloromethane Silica Gel Chromatography Dichloromethane Eluate GC profile GC/MS Methanol Eluate Figure 6. Schema for isolation and analysis of trace organics in algae grown on artificial substrates. 198 A typical gas chromatogram of the electron-capture pollution profile for each of the three samples, all run under identical conditions, is shown in figure 7. Curve A is from the sample representing the area above the city in the vicinity of the water treatment plant ( -river mile 110). Curves B and C represent the areas around river mile 108 and 100, respectively, within a heavily industrial- ized area of the city (Sheldon and Hites, 1979). Curve D is a PCB standard, Aroclor 1254. When the standard and sample curves are compared, a high degree of similarity may be seen; there is obviously not an exact match, which suggests that the samples are mixtures of PCBs. Looking downriver (Curves A to C, fig. 7), a station-related pollution trend is apparent. The station above the city (Curve A) was generally lowest in PCBs throughout the study and overall higher levels may be seen in the curves rep- resenting heavily industrialized areas. It aplpears that both point and non- point sources are contributing PCBs to the river within the city while PCB contributions upriver of the city are somewhat less. The actual concentrations of PCBs for the complete study period are currently being determined. Flame ionization analysis for these compounds is shown on the chromatogram in figure 8. Again, a station-related trend may be seen moving downriver (Curve A to C); it is apparent that Curve B is considerably more complex than either of the other two curves. Curve B corresponds to the most heavily industrialized area along the entire stretch of the river in the study. Identification by GC- MS of some of the compounds present in this sample are shown in table 3. Table 3. — Organic compounds present in diatometer samples from = river mile 108 on the Delaware River. Gas chromatogram for these samples shown as Curve B, figure 8. Compounds confirmed in biomass Dichlorobenzenes Naphthalene Acenapthene Fluorene Anthracene Pyrene Benz ( a) anthracene Chrysene Compounds tentatively identified but not confirmed C3 Benzenes Methyl Naphthalenes Toluene Ethyl Benzene Xylenes 199 10 Time -mi nutes 15 Figure 7. Gas chromatograph of polychlorinated biphenyls from electron capture analysis of algae grown on artificial substrates in the Delaware River during August 1979. Curve A = samples collected at -ri mi 110, Curve B = samples collected at =ri mi 108, Curve C = samples collected at ^ri mi 100, Curve D = Aroclor 1254 used as standard. 200 15 Time Minutes Figure 8. Gas chromatograph of organics determined by flame ionization, capillary column analysis of algae grown on artificial substrates in the Delaware River during August 1979. Curve A = samples collected at -ri mi 110, Curve B = samples collected at -ri mi 108, Curve C = samples collected at -ri mi 100. 201 In addition to periphyton colonization, nonviable detritus and trapped sediment will also accumulate on the slides, although this did not appear to be a major problem in the above studies. The amount of nonbiological material has been found in other studies to range from slight to significant. Pollutant profiles obtained from areas of high detritus and suspended sediments will have a com- ponent of these materials. As a result, an estimate of direct biological accumulation by periphyton is more difficult to obtain in these areas. We are presently investigating the role of this material on the pollutant profile. SUMMARY AND CONCLUSIONS Investigations by other researchers and our current studies clearly demonstrate the potential value of using periphyton growth on artificial substrates as a monitoring tool for trace concentrations of heavy metals and organics in aqueous environments. This method offers both enhanced analytical detection and collec- tion of time-integrated data of biologically available pollutants. Applications include assessments of point-source discharges for biologically available pol- lutants, effects of dilution and identification of chemical sinks in flowing systems, and as exploratory techniques for characterizing pollutants resulting from both point- and nonpoint sources present over wide areas. Although not suitable for certain research and applied problems because of limitations involv- ing the mechanism of algal uptake and the complexity of chemical interactions in aqueous environments, this technique should be considered a powerful, cost- effective means of supplying considerable information both areally and through time, depending upon the nature of the problem and the sampling strategy. Important research directions suggested by investigations to date include studies of changes in periphyton communities exposed to varying matrices of metal and organics; determining correlations between amounts accumulated and dry ash-free weight, chlorophyll or lipid content; investigation of seasonal effects on biological availability of metals and organics; and the use of this method in estuarine and coastal environments. A research area of particular interest is the analysis for trace metals of unprocessed slides collected over the past 20 years as part of the ongoing diatometer program conducted by the Academy of Natural Sciences of Philadelphia. Slides are available from studies carried out on several major rivers in the United States; analyses of these slides could provide an historical perspective of these particular rivers. 202 REFERENCES Anderson, R. V. 1977. "Concentration of cadmium, copper, lead and zinc in six species of freshwater clams." Bull, olf Environ. Contam. Toxicol. (4):492-496. Batley, G. E. and Gardner D. 1978. "A study of copper, lead and cadmium specia- tion in some estuarine and coastal marine waters." Estuarine and Coastal Marine Science 7:59-70. Burkett, R. D. 1975. "Uptake and release of methylmercury-203 by Cladophora glomerata." J. Phycol. 11:55-59. Cairns, J., and K. L. Dickson, editors. 1973. "Biological methods for the assessment of water quality." American Society for Testing and Materials Special Publication 528. 256 pp. Cherry, D. S., and Guthrie R. K. 1977. "Toxic metals in surface waters from coal ash." Water Resources Bull. 13 (6): 1227-1236. Chiou, C. T. , V. H. Freed, D. W. Schmedding, and R. L. Kohnert. 1977. "Partition coefficient and bioaccumulation of selected organic chemicals." Environ. Sci. Tech. 11(5): 475-478. Conway, H. L., and S. C. Williams. 1979. "Sorption of Cadmium and its effect on growth and the utilization of inorganic carbon and phosphorus in two fresh- water diatoms." J. Fish. Res. Bd. Canada 36:579-586. Cossa, D. 1976. "Sorption du cadmium par une population de la diatomee Phaeodactylum tricornutum en culture." Mar. Biol. 34:163-167. Cross, F. A. and W. G. Sunda. 1978. "Relationship between bioavailability of trace metals and geochemical processes in estuaries." Pages 429-442 Jji Martin L. Wiley, ed. Estuarine Interactions. Academic Press, New York. 603 pp. Cushing, C. E., and F. L. Rose. 1970. "Cycling of zinc-65 by Columbia River periphyton in a close lotic mocrocosm. Limn. Oceanogr. 15(5) :762-767. Filip, S. D. and R. I. Lynn. 1972. "Mercury accumulation by freshwater algae Selenastrum capricornutum. " Chemosphere 6:251-254. Filip, D. S., T. Peters, D. A. Adams, and E. Joe Middlebrooks. 1979. Water Research 3(3) :305-313. Foster, P. 1976. "Concentrations and concentration factors of heavy metals in brown algae." Environ. Pollut. 10:45-53. Friant, S. L. and H. Koerner. In press. "Use of an in situ artificial substrate for biological accumulation and monitoring of aqueous trace metals - A pre- liminary field investigation." Water Research. Fujita, M. and K. Hashizume. 1975. "Status of uptake of mercury by the fresh- water diatom Synedra ulna." Water research 9:889-894. 203 , E. Takabatake, and K. Iwasaki. 1976. "Effects of light, magnesium and cyanide on accumulation of mercury by a freshwater diatom, Synedra» " Bull. Environ. Contam. Toxicol. 16(2) : 164-172. Gachter, R. , and K. Lum-Shue-Chan, and Y. K. Chan. 1973. "Complexing capacity of the nutrient medium and its relation to inhibition of algal photosynthesis by copper." Schweiz. z. Hydrol. 35:25 2-261. Greig, R. 1979. "Trace metal uptake by three species of mollusks." Bull. Environ. Contam. Toxicol. 22:643-647. Greig, R. A. and Wenzloff, D. R. 1978. "Metal accumulation and depuration by the American oyster, Crassostrea virginica. Bull. Environ. Contam. Toxicol. 20:499-504. Goldberg, E. D. 1952. "Iron assimilation by marine diatoms." Biol. Bull. 102::243-248. Hassett, J. P., and M. A. Anderson. 1979. "Association of hydrophobic organic compounds with dissolved organic matter in aquatic systems." Environ. Sci. Tech. 13(12) :1526-1529. Huggett, R. J., Bender, M. E. , and Slone, H. D. 1973. "Utilizing metal concen- tration relationships on the eastern oyster (Crassostrea virginica) to detect heavy metal pollution." Water Research 7:451-460. Ingols, R. S. and T. F. Craft. 1975. "Sensors for monitoring water quality." Pages 418-423 _in Water quality parameters. Special Technical Publication 573 American Society for Testing and Materials. Johnson, G. D. , A. W. Mcintosh, and G. J. Atchinson. 1978. "The use of peri- phyton as a monitor of trace metals in two contaminated Indiana lakes." Bull. Environ. Contam. 19:733-740. Kaiser, R. 1971. "Ultra trace analysis of volatile substances with GC methods in the sub ppm and ppb range down to 10"^'-'%." J. Chromatographic Sci. 9:227-238. Kinkade, M. L. and Erdman, H. E. 1975. "Influence of hardness components (CA+^ & Mg+2) in water on the uptake and concentration of cadmium in a simu- lated freshwater ecosystem." Environ. Res. 10:308-313. Luoma, S. N. and Jenne, E. A. 1976. "Estimating bioavailability of sediment- bound trace metals with chemical extractants. " Pages 343-351 ±n D. D. Hemp- hill, editor. Trace substances in environmental health. Univ. of Missouri, Columbia. Martin, J. H. , and W. W. Broenkow. 1975. "Cadmium in plankton: Elevated concentrations of Baja, Calif." Science 190:884-885. McCrady, J. K. and Chapman, G. A. 1979. "Determination of copper complexing capacity of natural river water, well water and artificially reconstituted water." Water Research 13:143-150. 204 Parry, G. D. R. , and J. Hayward. 1973. "The uptake of Zn65 by Dunaliella tertolecta Butcha.J." Mar. Biol. Assn. U.K. 53:915-922. Patrick, R. 1967. "The effect of invasion rate, species pool, and size of area on the structure of the diatom community." Proc. National Acad. Sciences 58(4):1335-1342. 1971. "The effects of increasing light and temperature on the structure of diatom communities." Limn. Oceanogr. 16(2) :405-421. 1973. "Use of algae, especially diatoms, in the assessment of water quality." Pages 76-95 in ASTM Special Technical Publication 528. , M. H. Hohn and J. H. Wallace. 1954. "A new method for detemrining the pattern of the diatom flora." Not. Nat. Acad. Nat. Sci. Philadelphia, No. 259. 12 pp. and D. Strawbridge. 1963a. "Methods of studying diatom populations. J. Water Poll. Contr. Fed. 25(2) : 15 1-161 . . 1963b. "Variation in the structure of natural diatom populations." American Nat. 97(892) :51-57. , N. A. Roberts, and B. Davis. 1968. "The effect of changes in pH on the structure of diatom communities." Not. Nat. Acad. Nat. Sci. Philadelphia, No. 416. 16 pp. , B. Crum, and J. Coles. 1969. "Temperature and manganese as deter- mining factors in the presence of diatom or blue-green algal floras in streams." Proc. National Acad. Sciences 64(2) :472-478. , T. Bott and R. Larson. 1975. "The role of trace elements in management of nuisance growths." Environmental Protection Technology Series, EPA-660/2-75-008. Corvallis, OR. Phillips, D. J. H. 1977. "The use of biological indicator organisms to monitor trace metal pollution in marine and estuarine environment s-A review." Environ. Pollut. 13:281-317. 1979. "Trace metals in the common mussel Mytilus edulis (L) and in the alga Fucus vesiculosus (L.) from the region of the sound (Oresund)." Environ. Pollut. 18:31-43. Pijanowski, B. S. 1974. "Comparative evaluation of in situ water quality sensors." Pages 95-107 _in Proceedings of Second Joint Conference on Sensing of Environmental Pollutants, December 1973. NOAA (National Oceanic and Atmospheric Administration). Schulz-Baldes , M. and R. A. Lewin. 1976. "Lead uptake in two marine phyto- plankton organisms." Biol. Bull. 150-118:127. Sheldon, L. S. , and R. A. Hites. 1978. "Organic compounds in the Delaware River." Environ. Sci. Tech. 12(10) : 1 188-1194. 295 Skaar, A., B. Rystad, and A. Jensen. 1974, "The uptake of n""-^ by the diatom Phaeodactylum trlcornutum. Physiol. Plant. 32:353-358. Sunda, W. G. and R. R. L. Guillard. 1976. "The relationship between cupric ion activity and the toxicity of copper to phytoplankton. " J. of Mar. Res. 34(4):511-529. , Engel, D. W. , and Thuotte, R. M. 1978. "Effect of chemical specia- tion on toxicity of cadmium to grass shrimp, Palaemonetes pugio: Importance of free cadmium" Environ. Sci. «Technol. 12(4) :409-413. Trollupe, D. R. and B. Evans. 1976. "Concentration of copper, iron, lead, nickel and zinc in freshwater algal blooms." Environ. Pollut. 11:109-116. Tukey, J. W. 1977. "Exploratory data analysis." Addison-Wesley , Reading, Mass. 688 pp. Weber, C. J. 1973. "Biological monitoring of the aquatic environment." Pages 46-60 in ASTM Special Technical Publication 528. Weitzel, P. L. , editor. 1979. "Methods and measurements of periphyton communi- ties: A review." Special Publication STP 690. American Society for Test- ing and Materials. 183 pp. Windom, H. , F. Taylor, and R. Stickney. 1973. "Mercury in North Atlantic plankton." J. Cons. int. Explor. Mer. 35(1): 18-21. 206 IN-SITU WATER QUALITY MONITORING USING TROUT IN THE U.K. by Arthur Solman* Instrumentation Engineering Water Research Centre Medmenham Laboratory Buckinghamshire SL7 2HD The surveillance of water quality is a problem in common to many countries and whilst there may be geophysical and organisational differences between them, it is possible to observe similar requirements in the approach to a solu- tion of that problem. The organisation of the water industry in the United Kingdom is such that the Water Research Centre acts as the central research agency for that industry. One aspect of in-situ biological monitoring receiv- ing increased research effort is that of direct river abstraction surveillance using fish. The technique employed uses noninvasive electrodes to detect bioelectric po- tentials generated by the fish and conducted through the water. Different sig- nal components can be identified: respiration, heart action (e.c.g.), coughing, and physical motion. These signals are amplified, filtered, and processed to give three sets of data: respiratory rate, heart rate, and a total integrated signal called activity. Several systems are in operation at water treatment plants being managed by nonspecialist staff in some instances. The population of trout are changed regularly from test to stock tanks and are kept in an area isolated from natural daylight and excessive noise. To maintain the fish at low physiological levels, the test fish are unfed. The supervision of the fish population and associated equipment has been reduced to a once weekly visit of no more than 2 hours. By using up to 6 test fish, it is possible to improve the confidence in the reliability of response to the presence of toxic substan- ces. It is not possible to produce a system in which the toxic substance can be identified nor quantified, but the strength of such a monitor lies in its ability to respond to a wide range of substances. Extensive toxicity tests have been conducted in the United Kingdom and Holland in an attempt to identify threshold levels to a variety of toxic substances and to seek out those that might affect man and not fish or vice versa. Typical results indicate the systems ability to detect toxic concentrations up to two orders of magnitude less than L.C. or L.D.50. The final system shows promise of acting as a 24- hour on-line monitor of toxicity at sublethal levels that may offer early warn- ing of the presence of a hazard both to the water intake protection situation and to the biological population of the environment itself. *Mr. Solman was the only non-American to attend the biosensing conference and, while without a formal paper for presentation, kindly consented to speak extemporaneously and answer questions on matters relating to water-quality management in the United Kingdom. Subsequently, he prepared this precis of his remarks. 207 Working Panel Report 1 Biochemical Systems Chairman: Mr. Theodore Major Manager, Environmental Systems Magnavox Corporation Fort Wayne, Indiana SUMMARY The Biochemical Systems Working Panel included representatives from government agencies, universities, and industry. The composite experience of the panel in research development and operation of automated in-situ water quality sensors and monitoring systems was 105 person-years. (The panel members are listed below.) The panel came away convinced that biosensors will fulfill a very substantial and complementary role in the overall water-quality monitoring mission for essentially three reasons. 1. Low cost, nonspecific pollution alarm systems are available for grab-sampler initiation. 2. Some low cost, highly specific pollutant sensors with great sensitivity are available for application (i.e., enzyme-based sensors). 3. Several new techniques are amenable to immediate prototype system fabrication and field test. The working panel opened with a round-table discussion of the problems encountered by the workers and organizations involved in water-quality monitoring programs. This was followed by detailed technical discussions of specific biosensing techniques. The output of the discussions included the four broad recommendations discussed below. RECOMMENDATION I The panel selected the sensing techniques outlined below for additional development and funding. A priority rating established by its members and a rating was assigned to each technique. A total of eight was recommended, No. 1 - Enzyme Probe Coupled to Ion-Selective Electrode 1. Applications - Detection of low concentrations of toxic compounds in drinking water, plant effluents. Alarm systems. 209 2. State of the art a. Status - research b. Detection capabilities/limitation - 10""6 lowest likely detection limit. Absolute selectivity. c. Expertise/sources Dr. Shia S. Kuan Department of Chemistry University of New Orleans New Orleans, LA 70122 d. Cost - $300,000 total to develop a field device- 3. Short-term research and development possibilities (less than 5 years) - Less than 5 years is more than enough for nitrate, nitrite detectors. 4. Long-term research and development possibilities (greater than 5 years) - See 3 above. 5. Remarks - Useful for monitoring drinking water. High priority. No. 2 - Immobilized Enzyme Sensors 1. Applications - Detection of low concentrations of toxic materials; surveillance; monitoring. 2. State of the art a. Status - Detectors for organophosphate and carbamate pesticide have been field tested and used routinely for hazard monitoring. b. Detection capabilities/limitations - Extremely selective; good sensitivity; must find appropriate enzyme system. c. Expertise/sources Dr. Donald R. Sellers Dr. George G. Guilbalt Midwest Research Institute Department of Chemistry 425 Volker Boulevard University of New Orleans Kansas City, Mo. 64110 New Orleans, La. 70122 d. Cost - $10,000 for cholinesterase inhibitor in any matrix; $100,000 for other enzymes/pollutants. 3. Short-term research and development possibilities (less than 5 years) - Short term. 210 4. Long-term research and development possibilities (greater than 5 years) - See 3 above. 5. Remarks - High priority. No. 3 - Automated Bioluminescence 1. Applications - Effluent monitoring; surveillance; enforcement; hazardous materials spills. 2. State of the art a. Status - Manual test in production. Automated test at R&D level, b. Detection capabilities/limitations - Comparable to fish bioassay results. Bacteria often more sensitive than fish. c. Expertise/sources Dr. Anthony A. Bulich Beckman Microbics 6200 El Camino Real Carlsbad, Ca. 92008 d. Cost - $250,000 3. Short-term research and development possibilities (less than 5 years) - Short term. 4. Long-term research and development possibilities (greater than 5 years) - Long-range R&D for chronic test. 5. Remarks - High priority. No. 4 - Enzyme Amplified Immuno Assay 1. Applications - Detection of low concentrations of target compounds. 2. Status of the art a. Status - Research and development b. Detection capabilities/limitations - Extremely sensitive and specific. Not a class detector. c. Expertise/sources Dr. Donald R. Sellers Midwest Research Institute 425 Volker Boulevard Kansas City, Mo. 64110 211 d. Cost - $250,000 per system 3. Short-term research and development possibilities (less than 5 years) - Short term. 4. Long-term research and development possibilities (greater than 5 years) - See 3 above. 5. Remarks - High priority. No. 5 - Heterogeneous Bioprobes 1. Applications - Detection of low concentrations of toxic compounds. Signal may be related to toxicity. 2. State of the art a. Status - Methods developed; feasibility demonstrated. Not yet applied to environmental problems. b. Detection capabilities/limitations - Dependent on toxic compound. Can measure enzyme at 10~^-'-M. Extreme specificity. c. Expertise/sources Dr. Mark E. Meyerhoff Dr. Gary Rechnitz University of Michigan University of Delaware Department of Chemistry Ann Arbor, Mich. 48109 Dr. Robert Kobos Virginia Commonwealth University d. Cost - $100,000 per year 3. Short term research and development possibilities (less than 5 years) - Less than 5 years for development of laboratory-based equipment. 4. Long-term research and development possibilities (greater than 5 years) - See 3 above. 5. Remarks - Medium to high priority. No. 6 - Fish Blood Chemistry 1. Applications - Evaluation of water quality. Measure effect on blood, liver, enzymes. Protection of particular species. Effluent toxicity. (Be alert for sublethal effects.) 2. State of the aart a. Status - Research b. Detection capabilities/limitations - Looking for sublethal effects, so probably parts per trillion. Currently manual method (as opposed to automated). 212 c. Expertise/sources Dr. Dora R. May Passino Dr. Linwood Smith Great Lakes Fisheries Laboratory College of Fisheries U.S. Fish and Wildlife Service University of Washington Ann Arbor, Mich. 48105 A. Calabrese National Marine Fisheries Service 3. Short-term research and development possibilities (greater than 5 years) - None 4. Long-term research and development possibilities (greater than 5 years) - Definitely long term. 5 . Remarks - High to medium priority. No. 7 - Bivalve Mollusk Blood Chemistries 1. Applications - Pollution and resource management 2. State of the art a. Status - Research b. Detection capabilities/limitations - Not necessary to put out field sensors; long lived; no seasonal changes. Catching organisms might change physiology. c. Expertise/sources Robert Dean C. E. Epitanio Dauphin Island Sea Lab College of Marine Studies Dauphin Island, Ala. 36528 University of Delaware 3. Short-term research and development possibilities (less than 5 years) - None 4. Long-term research and development possibilities (greater than 5 years) - Long term 5 . Remarks - Medium to high priority No. 8 - Microcalorimetry 1. Applications - Potential detector of toxic compounds. Based on measurement of the heat of metabolism of small organisms and changes in this heat as a result of stress. 213 2. State of the art a. Status - Research instrument. Measurements made. Likely to remain a laboratory research instrument but potential for in-situ instrument has not been considered. b. Detection capabilities/limitations - Not yet determined. c. Expertise/sources Edward Prosen National Bureau of Standards d. Cost - $100,000 identification of organism $1,000,000 development of prototype in-situ sensor 3. Short-term research and development possibilities (less than 5 years - None 4. Long-term research and development possibilities (greater than 5 years) - Complete development would require more than 5 years. 5. Remarks - Medium priority. Usefulness for biological studies should be further demonstrated before consideration is given to modifica- tion for in-situ monitoring. RECOMMENDATION II It was recommended that a Federal Plan for Water-Quality Research Development and Monitoring be generated by the responsible agencies. This plan would serve four purposes. 1 . Permit government program planners to allocate their resources in a systematic manner. 2. Permit industry to identify planned and budgeted programs and markets. (This is a mandatory step prior to allocation of internal research and development funds.) 3. Allow universities to allocate their resources to "real" water-quality monitoring problems. 4. Allow government agency researchers to integrate their activities with the Federal Plan. As an example, the Federal Plan for Ocean Pollution Research, Development, and Monitoring could form an integral element of the overall Federal Plan for Water Quality Research Development and Monitoring. (Note that integration of developmental programs was recommended by the first water-quality sensing workshop Automated In-Situ Water Quality Sensor Workshop, held in Las Vegas during the period February 14-16, 1978 and sponsored by the Environ- mental Protection Agency, the National Oceanic and Atmospheric Administration, and the Interagency Working Group on Satellite Data Collection Systems.) 214 Comments on the Federal Plan for Water-Quality Research Development and Monitoring The Federal Plan must be an interagency one that integrates the responsi- bilities of the several federal agencies involved, such as the Environmental Protection Agency, the National Oceanic and Atmospheric Administration, the U.S. Coast Guard, and the U.S. Geological Survey, the U.S. Army Corps of Engineers, and the Department of Agriculture, as well as regional state and local governments. While the overall plan should be comprehensive, its initiation should be accomplished on the basis of a demonstration project where the research, development, and field activities could be accomplished on a practical basis. The BLACK CREEK project sponsored by EPA in Allen County, Indiana, is an excellent example of such a project dealing with agri- culture pollution. The plan must provide for effective utilization of the data acquired. This provision would include five actions. 1. Input to polluters for immediate response Industries (stop the spilling and dumping) Transportation Municipal waste water treatment authorities 2. Input to pollution abatement organizations Contractors U.S. Coast Guard Municipal organizations 3. Public appraisal (alarms, trends) Drinking water Contact sports Fish consumption Ecological health of water bodies 4. Enforcement organizations EPA USCG State and municipal organizations 5. Legislators Local State Federal Data Distribution Standards must be set for data coding for computer-to-computer interaction. The NOAA/NESS data communications scheme for the GOES satellite must be incorporated into the system and could serve as the standard. The data rate is slow enough to be compatible with dial-up telephone systems similar to those used by the National Weather Service AFOS system. As an example, a 215 dial-up data distribution system installed in a river monitoring system could be used to warn municipal water treatment operators of impending slugs of pollutants coming downstream. Similarly, standards must be set for data accuracy (quality assurance). By setting such standards, the Water-Quality Data Distribution System would grow as dictated by priority requirements and demonstration projects. Monitoring System Automation The objective of automation is to save money and to acquire the warning data required to protect the public and ecologic health. Accordingly, automatic water-quality monitoring stations would impinge in four basic areas. 1. Drinking water processing plant inputs and outputs 2. Industrial plant outfalls and streams 3. Waterwater treatment plant outfalls 4. Remote inaccessible locations The Automatic Water Quality Monitoring stations should be equipped only to sense those pollutant parameters which will be required for immediate response. If this is not the case, the alarm sensor may trigger a sample acquisition for subsequent laboratory analysis and reporting. The Automatic Water Quality Monitoring stations may utilize three techniques to provide alarm sensors. 1. Biologic sensors (toxicity) Fish and mollusks - stress detection Bacteriums - luninescent stress 2. Traditional sensors Conductivity pH Dissolved Oxygen Temperature 3. Multiparameter analyzers (automated) Heavy metals - energy dispersive x-ray Pesticides (organic) - gas chromagraphs NOTE The primary application for multiparameter analyzers is in the inputs and outputs of municipal potable and waste water pro- cessing plants. 216 The working panel observed that much of this technology is already available but not generally utilized. The Ohio River Sanitary Commission (ORSANCO) is a leader in this respect; however, their annual equipment procurement budget is limited to approximately $100,000 per year. The enzyme sensors currently in early stages of development show great promise in providing both the specificity and the sensitivity required to perform in automatic alarm systems as well as in laboratory measurement systems. Manual Monitoring Applications Manual analysis (i.e. , in the laboratory) is particularly useful and applicable in the detection of chronic subtoxic level pollution. This procedure is necessary to predict water-quality trends and to establish the lower levels of permissible pollution concentration. Another very valuable product of laboratory analysis is the establishment of the sensitivity levels required of field instrumentation such as bioprobes and specific ion sensors. The tecniques for manual pollution trend monitoring may include the use of three bioaccumulators , that is, 1 . Algae 2. Bivalves (invertebrates) 3. Fish The monitoring process may provide for natural acquisition of the organ- isms in the monitoring site or for the caging and implantation of control organisms. The organisms are exposed for the desired period and returned to the laboratory for analysis using traditional instrumentation. RECOMMENDATION III A third recommendation concerned support by the federal government for the development of automatic in-situ water-quality monitoring systems. The key element of systems is the water-quality sensor or transducer. This support is necessary since a federal, state, or municipal agency is the primary customer for automated in-situ water-quality monitoring systems. These automated systems are required to 1. Save monitoring cost and manpower, 2. Provide near-real-time data as required for corrective action (alarms), 3. Provide a measure of pollutant volume rate of travel dispersion, etc., and 4. Trigger sample acquisition devices as required for enforcement pro- ceedings. RECOMMENDATION IV Lastly, it was recommended that federal, industrial, and university organizations be appraised that the research and development activity in the field of biosensing and ion-selection electrodes in Europe and 217 the U.K. is proceeding at an accelerating pace. (In the United States, the effort has been confined to a relatively small number of universities and research institutes.) Members Working Panel 1 - Biochemical Systems Mr. Ted Major, Chairman Magnavox Manager, Environmental Systems 1313 Production Road Fort Wayne, IN 46808 Dr. Donald R. Sellers Co-chairman Midwest Research Institute Life Sciences Department Kansas City, MO 64110 Dr. Anthony Bulich Beckman Microbics 6200 El Camino Real Carlsbad, CA 92008 Dr. Robert C. Dean Dauphin Island Sea Lab Dauphin Island, AL 36528 Miss Edith Gould NOAA National Marine Fisheries Service Milford, CT 06460 Mr. William B. Jacobs Midwest Research Institute Life Sciences Department Kansas City, MO 64110 Dr. William H. Kirchhoff National Bureau of Standards Office of Air and Water Management Washington, D.C. 20234 Dr. Shia S. Kuan Department of Chemistry University of New Orleans New Orleans, LA 70122 Dr. Mark E. Meyerhoff University of Michigan Department of Chemistry Ann Arbor, MI 48109 Dr. Dora R. May Passino Great Lakes Fishery Laboratory U.S. Fish & Wildlife Service Ann Arbor, MI 48105 Mr. Eugene M. Russin NOAA Ocean Technology & Engineer- ing Services Test & Evaluation Laboratory Rockville, MD 20852 218 Dr. Morton Smutz NOAA/OOE 6010 Executive Boulevard Rockville, MD 20852 219 Working Panel Report 2 Bioaccumulation Chairman: Dr. Andrew Robertson Head, Chemistry-Biology Group NOAA/GLERL Ann Arbor, Michigan Panel 2 was convened to discuss the use of aquatic organisms as accumu- lators or samplers of contaminants. The panel attempted to identify the most promising areas for development and use of such bioaccumulators. A summary of findings follows, particularly in terms of development. In addition, a brief outline of the bases used by the panel to arrive at these decisions is presented. (Panel members are listed below.) Potential Uses for Bioaccumulators As a first step in identifying the areas of most promise for developing bioaccumulation techniques, the panel identified the major uses that could be made of such measurements. This approach involved developing a list of the ways in which these measurements might aid in controlling water-quality degradation. Five such considerations are discussed below. 1 . As a screening technique. Organisms can be analyzed to determine their levels of bioaccumulation for a broad spectrum of contaminants. These analyses can aid in identifying which contaminants are found in a specific aquatic environment at substantial levels and those that warrant further study as potential problems. 2. As an early warning. Bioaccumulation may be used as a method to identify when contaminants are building to concentrations that may cause substantial deleterious environmental effects. 3. As a means to characterize the seasonal and geographical distribution of contaminants. Comparative measurements of bioaccumulation at various times and places may provide valuable information on the seasonal and geographical distributions of contaminants and thus provide assistance in understanding the dynamics and cycling of such substances. 4. As a regulatory aid. In many cases, regulatory agencies have set objectives or standards designating the concentrations that specific contaminants should not exceed in the environment. These are sometimes expressed as concentrations allowable in specific organisms, often fish. Thus, bioaccumulation measurements can give direct information on whether these control concentrations are being exceeded. Standards are also sometimes expressed as concentrations in a given volume of water. If the relation between concentration in the bioaccumulating organism to that in water is known, bioaccumulation measurements can be a convenient 221 and often less expensive way to determine whether such water concentrations are being exceeded. 5 . As indicators of sources of pollutants. Bioaccumulators may be placed at various locations throughout an area where excessive containment concentrations may pass. By comparing the bioaccumulation at such sites, it would often be possible to identify the containment source. This technique can be especially valuable when contaminant releases occur in locations or at times when conventional sampling would not detect them. General Categories of Bioaccumulation Techniques The panel concluded that the use of organisms as samplers or bioac- cumulators of contaminants can be usefully divided into three general categories. The first category includes the most straightforward techniques in which naturally occurring organisms are collected in their native environment and analyzed for their accumulated pollutants. In this category, nothing is known about the past history of the organisms analyzed, so interpretation of the results can be difficult and conclusions ambiguous. In the second category, organisms of known history are exposed in the laboratory to samples of the environment to be tested. After a set period of exposure in this laboratory setting, the organisms are analyzed to determine how much of the contaminant or contaminants of interest have been accumulated. The third category is similar to the second in that the experimenter places organisms of known history in contact with the suspected polluted environment. However, in this category the organisms are exposed in the natural environment rather than in the laboratory. After a specific period of time, the contaminant load is measured. A great deal of attention in pollution research has been given to the development of techniques for categories one and two. A number of methods are well established especially in category one. For the third category, however, while several specific techniques have been used, development of the approach has been so limited that its applicability is still largely unproven. The panel believes nevertheless that techniques from this category hold great promise for aiding water- quality studies and regulation. Consequently, they decided to limit their recommendations to identifying promising areas for development under this category only. General Points Before proceeding with recommendations, the panel developed a set of seven general points that apply to all the recommended techniques. 222 1. Techniques must be such that adequate material for analysis of the contaminant or contaminants of interest can be obtained. 2. Techniques should provide bioaccumulation concentrations that caa be related to ecosystem effects or studies providing this information should be carried out. 3. Values obtained by any technique should be relatable to environ- mental concentrations of the contaminant or studies to provide this information should be conducted. 4. As different organisms concentrate contaminants from different compartments of the environment, it is necessary to determine what compartment the organism in use is really sampling. 5. Techniques must be matched to the contaminant or contaminants of interest since different species bioaccumulate the same pollutant to very different degrees. 6. Similarly, the same organism bioaccumulates different contaminants to very different extents. 7. It is not enough to simply measure bioaccumulation in order to further water-quality management. Relating the measured bioaccumulation values to economic consequences is very important, as is the interpre- tation of the results in terms of regulatory needs. Specific Recommendations for Development of Bioaccumulation Techniques The panel concluded that there are two basic ways for placing organ- isms at a specific location in the environment and measuring the amount of contaminants they bioaccumulate. The more conventional method calls for putting determined numbers and types of organisms at a specific location and then retrieving them after a certain length of time and measuring their levels of bioaccumulation. The first three recommendations below relate to this subcategory. The second method is somewhat less conventional but holds substantial promise in the panel's view. This calls for placing a structure that will attract, or cause settling of, certain types of organisms in a designated location in the environment. After a certain length of time, the structure with the organisms that have settled and grown on it is retrieved and the bioaccumulations of the contaminant or contaminants of interest are determined. The final three recommendations apply to this second method. 223 Recommendations Application No» 1 - Biomonitoring Chambers 1. State of the art a. Status - Membrane biomonitoring chambers are in limited production; chambers with other types are in the research stage. b. Detection capabilities and limitations - These chambers have great applicability in determining bioaccumulation from the dissolved phase; they do, however, usually exclude particulate material. Fouling of the chambers may cause problems as well as the fact that the organisms them- selves restrict current flow and may alter other environmental conditions. c. Sources of expertise - Dr. H. Schlichting (BioControl Co.); Dr. McFetters, Montana State University; Dr. T. Smayda, University of Rhode Island 2. Short-term research and development possibilities (less than 3 years) - There is a need to reduce fouling problems, possibly by altering the confining material. Also, research should be continued on improving the design and materials of biomonitoring chambers and on defining the applicability of the technique. 3. Long-term research and development possibilities (greater than 5 years) - Long-term research and development is needed to automate water- quality measurements inside and outside the chambers and to relate bio- accumulation in organisms in the chambers to environmental effects. 4. Remarks - These chambers may be useful as a legally valid means to establish the source of a pollutant. There are a variety of promising areas for development of these chambers; thus, the panel rates them a high priority for further development. Application No. 2 - Biomonitoring by Sessile Filter Feeders 1. State of the art a. Status - There are several operational systems using sessile filter feeders to determine bioaccumulation (e.g., Mussel Watch). Validation of the meaning of the results from these systems is still required. b. Detection capabilities and limitations - Filter feeders bioaccumu- late particle-transported contaminants; they are much less likely to concen- trate contaminants from the dissolved phase. These organisms may have avoidance behavior for high concentrations of certain contaminants that would cause them not to accumulate the pollutant when it is present in such contentrations. 224 c. Sources of expertise - Scripps Institute of Oceanography (Dr. E. Goldberg); NERC Laboratory (Plymouth, England); NMFS Laboratory (Milford, Conn.) 3. Short-term research and development possibilities (less than 3 years) - Further research and development is required to determine what species to use for specific purposes and how to improve standardization of this technique. Further development is also needed to improve the retrieval system for certain applications. 4. Long-term research and development possibilities (greater than 5 years) - Long-term research is needed to improve understanding of the meaning of measured levels of bioaccumulation, especially with regard to effects on important resources. 5. Remarks - This method has high potential for combining measurements of bioaccumulation with determination of contaminant effects on the test organisms. The panel rates this technique as having high potential for further development. Application No. 3 - Biomonitoring by Confined Deposit Feeders 1. State of the art a. Status - The panel had no knowledge of any efforts to use confined deposit feeders as bioaccumulators. However, because sediments are often the ultimate sink for contaminants, development of such a system was considered very desirable. b. Detection capabilities and limitations - Such a system should have the capacity to monitor a broad spectrum of sedimentary contaminants. Possible difficulties include the effects of confinement on bioaccumula- tion, fouling interferences, and possible escape of the bioaccumulating organisms. c. Source of expertise - None known 2. Short-term research and development possibilities (less than 3 years) The first priority in the exploitation of this technique is to develop a prototype system. 3. Long-term research and development possibilities (greater than 5 years) - Long-term needs cannot be identified until development of this technique has been initiated. 4. Remarks - The panel felt that a technique to monitor pollutant bioaccumulation in the sediments is very badly needed, and so, even though the cost of development of a system to use confined deposit feeders may be quite high, they rated the development of this technique as having high priority. 225 Application No. 4 - Artificial Substrates for Freshwater Periphyton (Periphytometer) 1. State of the art a. Status - The biological aspects of colonization of artificial substrates is quite well understood. The instrument to perform this function is in the research state. (See the paper by Dr. Friant published elsewhere in this volume.) b. Detection capabilities and limitations - This technique will allow measurement of the bioaccumulation of a broad spectrum of contaminants. However, the technique is not suitable for strongly polar substances or those that are easily metabolized by the organisms involved. c. Source of expertise - Philadelphia Academy of Sciences (Dr. Friant) 2. Short-term research and development possibilities (less than 3 years) - Short-term research and development is needed on seasonal effects on biologi- cal availability of pollutants, on evolution of sediment contributons to contaminant accumulations on periphytometers, on establishing correlations between bioaccumulation and ash-free periphyton weight, and on how communities change on exposure to contaminants. 3. Long-term research and development possibilities (greater than 5 years) - Long-term research and development is needed to improve the design of the periphytometer and to automate the instrument for in-situ analysis of contami- nants. 4. Remarks - This technique would be especially valuable if biological measurements of diversity and species abundance were made on the same samples so that bioaccumulation values can be correlated with these biological properties, The panel rates this technique as having a high priority for further development. Application No. 5 - Artificial Substrates for Marine Periphyton 1. State of the art a. Status - The same instrument as used in the preceding freshwater application can be used in marine waters. However, it has received little use in these environments so far and may require some modification. b. Detection capabilities and limitations - Same as for freshwater application. c. Sources of expertise - Philadelphia Academy of Sciences 2. Short-term research and development possibilities (less than 3 years - Short-term research and development is needed to adapt this instrument for for marine environments and to develop the background to interpret the results obtained. 226 3. Long-term research and development possibilities (greater than 5 years) - Same as the freshwater instrument. 4. Remarks - The use of this instrument may present more difficulties in the marine environment than in freshwater due to heavier deposition of inorganic material and to settling of a greater variety or organisms. Both of these effects may make interpretation of the results more diffi- cult. Because of these potential difficulties, the panel rates this technique as having medium priority for further development. Application No. 6 - Artificial Substrate for Benthic Recolonization 1. State of the art a. Status - The study of benthic recolonization is quite well advanced, However, to the panel's knowledge, no artificial substrates have been devel- oped to take advantage of this for bioaccumulation studies. b. Detection capabilities and limitations - This technique would allow measurement of bioaccumulation by deposit feeders. One limitation is that retrieval may require divers. c. Sources of expertise - Virginia Institute of Marine Sciences, and the U.S. Environmental Protection Agency at Gulf Breeze, Florida. 2. Short-term research and development possibilities (less than 3 years) - Existing methods for studying recolonization need to be adapted for artifi- cial substrates to look at bioaccumulation. 3. Long-term research and development possibilties (greater than 5 years) - None identified 4. Remarks - This type of approach may be the most promising way to monitor bioaccumulation of pollutants in the deep sea, and would be parti- cularly useful for studying the recovery of communities and the relation of this to bioaccumulation. The panel rates this technique as having medium priority for further development. Members Working Panel 2 - Bioaccumulation Dr. Andrew Robertson, Chairman Dr. Richard E. Defenbaugh Head, Chemistry-Biology Group NOAA/GLERL Ann Arbor, Michigan New Orleans OCS Office Bureau of Land Management New Orleans, La. 227 Dr. Stephen L. Friant Dr. Raymond G. Merrill, Jr. Dept. of Limnology and Ecology Academy of Natural Sciences Philadelphia, Penna. Environmental Protection Agency Research Triangle Park, N.C. Dr. Harold E. Schlichting, Jr, BioControl Co., Inc. Port Sanilac, Mich. and University of Minnesota Biological Station Lake Itasco, Minn. Dr. Fred Thurberg National Marine Fisheries Milford, Conn. Dr. Harris H. White National Oceanic and Atmospheric Administration Rockville, Md. 228 Working Panel Report 3 Organism Behavioral Responses Chairman: Mr. Richard W. Paulson U.S. Geological Survey Water Resources Division Reston, Virginia The panel concluded that there were two objectives in studying orgnism behavioral response for application as in-situ sensors. (The names and associations of the panel members are listed below.) Objective No. 1 - To develop a technique to measure broad-scale variations in water quality. Objective No. 2 - To better understand the way organisms sense water quality in order to develop sensors that use the same technique. The panel also concluded that the following 9 criteria should be considered when organism behavioral response is used as an in-situ sensor monitoring mode. Criterion No. 1 - A real-world approach should be used when organism behavioral response is used in in-situ monitoring, i.e., unnecessary complexity is to be avoided. Criterion No. 2 - Common, easily available, endemic, and well-known species should be used where possible. Behavioral response of an organism is important when it deviates from the normal range, and it is imperative that the organism be a common one so that its normal response is understood. Criterion No. 3 - The organism should operate at ambient conditions of the water body under measurement in order to maximize the efficiency of station operation. Criterion No. 4 - The organism should not be so delicate that it cannot tolerate minor perturbations, nor so rugged that it does not respond to large perturbations. Criterion No. 5 - The organism should respond to water quality in a way that can be observed, measured, and monitored in a noninvasive way. Criterion No. 6 - The expected range of water quality should subject the organism to sublethal stress only. Criterion No. 7 - The organism should come to equilibrium in the test environment within a reasonable length of time, such as 24 hours. The org- anism population at a monitoring location should be managed so that part of the population should be in reserve, part coming to equilibrium, with part yet under test. 229 Criterion No. 8 - The organism should be used as a management tool in the sense that behavior outside the individual organism' s normal range should be used as a key for more detailed and conventional water sampling and analysis, Criterion No. 9 - Organisms will not be completely predictable in behavior and will respond to condition with some uncertainty, i.e., false alarms can be expected. Conclusions and Recommendations Fish probably offer the most immediate means of in-situ monitoring using behavioral response. It is appropriate to begin using fish in operational stations as is done in the United Kingdom and Europe. Behavior in invertebrates has not been as studied, as with fish, but there are a number of groups that offer possibilities as in-situ monitors. Crustaceans, including crabs, lobsters, crayfish, and shrimp have a number of behavioral patterns that could be adapted to monitoring, but would require a costlier and longer developmental program than for fish. In those groups that have been studied (e.g., crabs and lobsters), the responses to dissolved substances have indicated that invertebrate chemosensors are extremely sensitive and thus offer potential for in-situ monitoring. Research and development costs and time, though greater for invertebrates, may be warranted for some species because of their levels of sensitivity. Some molluscan behavioral responses, e.g., valve (shell) openings and closings (pumping behavior), have been used successfully as physiological monitors and are also quite sensitive to water quality changes. The valves can be hooked to hymographs or printers of some sort to facilitate automa- tion. The organisms feed themselves and require little care. Other bivalve species such as scallops employ swimming behavior to escape predators. This reaction can be tested by dropping starfish extract into the water. A large number of clam species are capable of leaping movements, whereby the foot is extended and curled under the body and then forcibly straightened. This movement propels the clam upward away from the bottom. Again, this can be elicited by dropping extract of predator species. Deviation of the swimming behavior may be an indication of the organism's ability to cope with water-quality variations. These behavioral responses indicate a very sensitive chemosensory mechanism which might ultimately be tapped for in-situ monitoring. The scope of organisms considered was limited by the small number of participants on the panel and their background and experience as biologists. Thus, this discourse should not be considered exhaustive. The panel' s recommendations follow. 230 Recommendat ions Application No» 1 - Fish Behavioral Response 1. State of the art a. Status - Instrumented automated monitoring systems exist to collect physiological and rheotactic data. Laboratory systems are in use in the U.S., while operational capabilities exist in the U.K., Holland, and West Germany. b. Detection capabilities and limitations - Capable of detecting toxins to levels 10^ lower than ^^5 0. Some substances toxic to fish are not toxic to man and the opposite is true as well. c. Sources and expertise U.S.A. Dr. John Cairnes University Center for Environmental Studies Virginia Polytechnic Institute Blacksburg, Va. Dr. Donald E. Wohlschlag Institute of Marine Science Port Aransas Marine Laboratory University of Texas Austin, Texas U.K. Holland (K.I.W.A.) South Africa W. Germany d. Costs and times Arthur Solman Water Research Centre Medmenham Laboratory Buckinghamshire Dr. Poels Morgan Kerren Aquatest and Zippe Monitor Developmental: Greater than $100,000 to develop improved sensor capability. Prototype: Greater than $20,000 Production: Certain European systems are presently available for $15,000 to $20,000. 231 2. Short-term research and development possibilities (less than 3 years) - New transducers to improve signal separation techniques and pattern recog- nition techniques need to be developed. This approach, however, should not inhibit operational use of this methodology in the short term. 3. Long-term research and development possibilities (greater than 5 years) - There is a need for development of flexible monitoring systems that can support a variety of organisms, a variety of sensors, ongoing testing, and a better data base of information for analysis. 4. Remarks - A continuing dialogue between workers active in this field and agencies responsible for environmental monitoring is definitely needed. Applicaation No. 2 - Invertebrates Such as Blue Crab (Callinectes Sapidus) 1. State of the aart a. Status - Experimental. Feeding behavior such as antenna flicking and gill bailing are being studied in detail. b. Detection capabilities and limitations - Blue crabs can respond to food materials at a level of 10"^ gram/liter. They respond with changes in the rate of antenna flicking and gill-bailing behavior. It may require some expertise to recognize and quantify behavioral response automatically. c. Sources of expertise W. H. Pearson, Batelle institute Bori 011a, NMFS, Sandy Hook d. Costs and times Developmental: On the order of 2 years and $100,000. Prototype: On the order of $50,000 Production: Difficult to assess 2. Short-term research and development possibilities (less than 3 years) - Much basic research is required but there is a good chance that responsive sensors could be developed within 3 years. 3. Long-term research and development possibilities (greater than 5 years) - Difficult to assess but automated systems to monitor invertebrates could be ready in 5 years. 4. Remarks - The risk is great as are the potential benefits. 232 Application No. 3 - Molluscs Such as Eastern Oyster (Hassastrea Virginica) 1. State of the art a. Status - Pumping behavior often is used as a physiological indi- cator, although not necessarily as a measure of toxics. b. Detection capabilities and limitations - Not well known beyond its ability to detect changes in temperature, salinity, and turbidity. c. Sources of expertise H. H. Haskes - Rutgers H. Hido - U. of Maine R. Hillman - Battelle d. Costs and times Developmental: $100,000 Prototype: 2 years on the order of $75,000 Production: Difficult to assess 2. Short-term research and development possibilities (less than 3 years) - Can probably be automated relatively easily. Research really required to assess reaction of organism to various toxics. 3. Long-term research and development possibilities (greater than 5 years) - Once the toxic reaction of the organism is known, the automation will be simple. 4. Remarks - With regard to freshwater molluscs, not much is known. On the other hand, a considerably larger body of knowledge is at hand regarding marine molluscs (those sold commercially). Application No. 4 - Zooplankton (diurnal movements with copepods and arrow worms, for example 1. State of the art a. Status - Much is known about vertical movements corresponding to day/night periods (photoresponse) . These vertical migrations might offer possibilities for study. 233 b. Detection capabilities and limitations - Limitations involve short life cycles. From species to species, there is a variability in behavior patterns. However, pure cultures are easy to raise. c. Sources of expertise C. I. Gibson S. Herman Battelle Lehigh University Marine Research Laboratory Bethlehem, Penna. Sequim, Washington d. Costs and times Developmental: Given $300,000, 3 to 4 years. Prototype: At this time, neither is specifiable. Production: At this time, neither is specifiable. 3. Short-term research and development possibilities (less than 3 years) - Short-term research could be aimed at defining migration patterns for representative species and assessing species ability to detect pollutants (and alter behavior). 4. Long-term research and development possibilities (greater than 5 years) - Longer term programs would probably be required to use this approach feasibly. 5 . Remarks - The probability of success using this technique is difficult to forecast. Members Working Panel 3 - Organism Behavioral Responses Richard W. Paulson U.S. Geological Survey Water Resources Division Reston, Virginia Dr. Robert E. Hillman Battelle Columbus Laboratory William F. Clapp Laboratories Duxbury, Massachusetts Mr. Arthur Solman Water Research Centre Medmenham Laboratory Buckinghamshire, England Dr. Donald E. Wohlschlag Institute of Marine Science Port Aransas Marine Laboratory University of Texas Austin, Texas 234 Working Panel Report 4 Traditional Biological Measurements Chairman: Dr. Harlan L. McKim Army Corps of Engineers Water Resources Support Center Fort Belvoir, Va. Cochairman: John Bushman Army Corps of Engineers Washington, D.C. CONCLUSIONS AND RECOMMENDATIONS Chlorophyll is a basic biological parameter whose measurement is useful for water quality monitoring purposes. However, further refinement of classical chlorophyll measurement techniques is probably not particularly fruitful. New techniques such as in-vivo chlorophyll fluorescence measurements have a high potential for automated in-situ water quality monitoring. Summary Many agencies have expressed a need to increase the volume of biological samples processed. Traditional biological techniques make this an extremely labor-intensive task. Increasing the speed of collecting and processing are of great concern. URI's research into the use of image processing and pattern recognition for zooplankton analysis represent a major step toward automation. VPI has been using optical image processing and pattern recognition to analyze diatoms. The panel recommends that for automated processing and analysis of biological samples: » Research into image processing and pattern recognition be continued. © Research into in-situ photographic teachniques be extended. o Long-term research into potential for telemetry of biological data be initiated. © Strategy of oceanographic and limnological sampling plans be prepared. • Encourage interaction with mapping of ocean productivity by remote (satellite) sensing techniques. 235 Recommendations for Traditional Biological Measurements Chlorophyll 1. State of the Art: a. Status - Chlorophyll measurements have been used for decades as estimates of algal standing crop (biomass) relative productivity, and trophic state. Chlorophyll pigment is closely coupled with photosynthesis, so the measurement is very useful for productivity — or eutrophication-related purposes. The most traditional method of measurement is spectrophotometric determination of absorbance on extracted samples (i.e., in vitro), although in-vitro measurements are also now used widely. In-vitro methods are rela- tively well standardized and commonly employed in oceanographic areas and limnological research. In-vivo (not extracted) chlorophyll fluorescence measurements are more recent, but are used widely to investigate vertical and horizontal distribution of phytoplankton. DCMV-enhanced in-vivo chloro- phyll fluorescence is a relatively new modification that significantly improves the information content of chlorophyll measurement by providing an indication of the physiological status of the phytoplankton assemblage sampled, in addition to the algae biomass present. b. Detection Capabilities /Limitations - Chlorophyll measurements are sensitive over a broad range of concentrations, particularly with fluorometry. Detection limits are not really a problem. Primary limitations stem from inherent variability of chlorophyll: plant carbon and fluorescence; chlorophyll relationships. Historical methodological problems with in-vitro methods are (1) differentiation of absorption (fluorescence emission peaks in complex pigment mixtures), (2) extraction efficiency, and (3) presence of phaeopigments. Phaeopigment concentrations can be estimated and corrected. In-vitro and in-vivo fluor- escence measurements must be related to spectrophotometric determinations to estimate chlorophyll concentrations. Chlorophyll measurement provides a good relative indication of algal biomass rather than an exact deter- mination. c. Sources of Expertise - Numerous chlorophyll methodologies are in common use in oceanographic/limnological research and water quality moni- toring. d. Costs and times - Traditional in-vitro methods can probably not be improved significantly. In-vivo fluorescence methods, particularly DCMV-enhanced fluorescence, have potential for further development. DCMV method requires more extensive evaluation and automation. Automation can probably be accomplished with off-the-shelf components (continuous-flow fluorometer, automatic injection system, timing devices, and data storage systems). The time required for evaluation of this method including proto- type development and field testing is about three years at a cost of approximately $100,000/yr. 236 2. Short-term research and development possibilities (less than 3 years) - Requirements for short-term R&D are to further evaluate the DCMV-method, including development of a prototype automated system and a field verification project. The entire program can be accomplished within a 3 year time frame. 3. Long-term research and development possibilities (greater than 5 years) - The long term goal would be to interface the system with other automated water quality monitoring systems (physical-chemical sensors, particle counters, image processing, auto analyzers, etc.). 4. Applications - The application of the system would be in continuous monitoring and/or surveying of (1) algal standing crop, (2) relative productivity/tropic state, (3) algal physiological condition, and (4) changes in the above due to nutrient/pollutant/toxicant loading. In addition, continuous flow fluorometry can provide the ground truth for chlorophyll measurement so that a comparison/evaluation of remotely-sensed (satellite) chlorophyll data can be made. 5 . Data Handling - Extent of data handling required depends on the apoplication. It can range from a simple chart recorder for survey work to a data logger/computer for continuous monitoring. Time series analyses is perhaps the best for large data sets. 6 . Summary a. Further refinement of classical chlorophyll measurement techniques is probably not required. b. In-vivo chlorophyll fluorescence measurements have high potential for automated water quality monitoring. c. DCMV-enhanced in-vivo chlorophyll fluorescence method provides information on the amount of algal biomass present and algal physiological condition. The method requires additional evaluation but it potentially can also be automated. Recommendations for Computerized Processing of Plankton Samples 1. State of the art a. Status: A developmental/engineering model has been constructed, programmed, and tested at the University of Rhode Island. Large scale sampling programs are in progress which require automated analyses. b. Detection Capabilities and Liimitations: Major taxonomic groups in the zooplankton have been measured, recognized, and counted (e.g., cope- pods, cladocera, chactognaths , fish eggs, and euphousids); more detailed classification appears possible upon solution of technical problems (imag- ing, specimen orientation). 237 c. Sources of Expertise: University of Rhode Island, NOAA/Narragansett - Poularikas , Jeff res, Oherman, Mouser NOAA/Miami - Ortner WHOI - Wiebe Scripps - Burg VPI - Cairns d. Costs and Times: Developmental: completed Prototype: 2-3 yeaars, 60-70K/unit (hardware costs) Production: no estimate at present 2. Short Term Research and Development Possibilities (less than 3 years): Further testing of algorithms; optimization of computer space and speed; refinement of imaging and specimen orientation techniques; field testing of prototype in other laboratories; and combining statistical and syntactic programming. 3. Long Term Research and Development Possibilities (greater than 5 years): o Incorporation of operation interaction - machine learning algorithms o Applications to phytoplankton o Photographic archiving of plankton samples o Telemetry of data — for making real-time decisions in the conduct of oceanographic sampling programs (vs. grid and transect approach) o Interaction with mapping of ocean productivity using remote sensing techniques o Fisheries optimization dynamics — determination of growth ratios (fish-scale image analysis), nutritional availabilities and requirements for sustained growth; and recruitment rates and success. 4. Remarks a. Applicability to analysis of ecosystem processes: The state and rate of change in relation to environmental variation, both natural and man-induced (e.g. plankton distribution around and in pollution fields arriving from wastewater discharges and petroleum spills) 238 b. Sensor need: A facility for analysis should be available at all levels, from local to international. c. Number of sensors: Regional pattern recognition centers should be established for high speed processing of plankton samples; the facilities would also operate on integrated data management system d. Likelihood of success: Identification, enumeration, and size- frequency distributions have been made by computerized means of pattern recognition; the need for engineering refinement in certain areas is now obvious. e. Cost effectiveness: A comparison of manual and automated means of zooplankton sample analysis (counting, size-frequency distri- bution) shared time saving of 60-80% by automated procedures. f. Skill level: A skilled operator capable of traditional zoo- plankton analysis is essential; also some facility with the management of mid-size computer programs would be desirable. A Bachelor of Science degree in natural science, electrical engineering, or computer science should be required, plus 6 months to 2 years practical experience. Members Working Panel 4 - Traditional Biological Measurements Dr. Harlan L. McKim, Chairman Army Corps of Engineers Water Resouraces Support Center Fort Belvoir, Va. John B. Bushman, Cochairman U.S. Army Corps of Engineers HQDA (DAEN-CWP-P) Washington, D.C. Dr. John L. Anderson U.S. Army Corps of Engineers MROED-HF Omaha, Nebb. Luther Bivins N0AA/00E Rockville, Md. Gary Burns EPA Bellevue, Wash. Dr. H. Perry Jeffries Graduate School of Oceanography University of Rhode Island Kingston, R.I. Dr. Bruce L. Kimmel Biological Station University of Oklahoma Kingston, Okla. 239 Dr. J. A. Llewellyn College of Engineering University of South Florida Tampa, Fla. Dr. Horace G. Loftin Naval Coastal System Center Panama City, Fla. Dr. Alex D. Poularikas Department of Electrical Engineering University of Rhode Island Kingston, R.I. George V. Shaefer Interocean Systems, Inc, San Diego, Calif. 240 PARTICIPANTS 2nd Interagency Workshop on In- Situ Water Quality Sensing; Biological Sensors Dr. John L. Anderson U.S. Army Corps of Engineers Omaha Office ATTN: MROED-HF 6014 USPO & Courthouse 215 North 17th Street Omaha, Nebraska 68102 8-864-4622 Luther Bivins N0AA/00E 6010 Executive Blvd. Rockville, Maryland 20852 8-443-8444 Dr. Anthony A. Bulich Principal Development Microbiologist Microbics Operations Beckman Instruments, Inc. 6200 El Caraino Real Carlsbad, California 92008 714-438-3515 Dr. Richard E. Defenbaugh New Orleans OCS Office Bureau of Land Management 500 Camp Street, Suite 841 New Orleans, Louisiana 70130 8-682-6541 Dr. Stephen L. Friant Department of Limnology and Ecology Academy of Natural Sciences (Philadelphia) Philadelphia, Pennsylvania 19104 215-299-1077 Miss Edith Gould Research Biochemist NOAA/NMFS, NEFC Milford Lab. Milford, Connecticut 06460 203-878-2459 Mr. Gary Burns Environmental Protection Specialist 15-158th Place, S.E. Bellevue, Washington 98008 206-746-9576 John B. Bushman U.S. Army Corps of Engineers HQDA-(DAEN-CWP-P) Washington, D.C. 20314 202-272-0131 Dr. John Cairns University Center for Environmental Studies Virginia Polytechnic Institute Blacksburg, Virginia 24061 703-961-5539 Dr. Robert E. Hillman Battelle, William F. Clapp Laboratories P.O. Drawer AH Duxbury, Massachusetts 02332 617-934-5682 William B. Jacobs Midwest Research Institute Kansas City, Missouri 64110 816-763-7600 Dr. H. P. Jeffries Prof, of Oceanography University of Rhode Island Kingston, Rhode Island 02881 401-792-6281 Dr. Robert C. Dean Dauphin Island Sea Laboratory Dauphin Island, Alabama 36528 Office: 205-661-8811 Home: 205-438-1690 Edwin D. Kennedy NOAA/EDIS (D-81) Rockwall Building Rockville, Maryland 8-443-8610 20852 241 Dr. Bruce L. Kimmel Asst. Director, Biological Station Asst. Professor, Dept. of Zoology Kingston, Oklahoma 73439 Biological Station University of Oklahoma Office: 405-564-2463 Home: 405-795-5966 Dr. William Kirchoff Office of Environmental Measurement National Bureau of Standards Washington, D.C. 20234 301-921-2711 Theodore Major Magnavox Corporation 1313 Production Road Fort Wayne, Indiana 46806 219-482-4411 Dr. Harlan L. McKim Army Corps of Engineers Water Resources Support Center WRSC-C Kingman Building Fort Belvoir, Virginia 202-325-0670 or 603-643-3200 John D. Koutsandreas Office of Monitoring & Technical Support (RD-680) Office of Research & Development U.S. Environmental Protection Agency 401 M Street, S.W. Washington, D.C. 20460 426-4477 Dr. Raymond G. Merrill, Jr. Research Chemist Environmental Protection Agency MD 62 IERL Research Triangle Park, North Carolina 27711 8-629-2557 Dr. Shia S. Kuan Department of Chemistry University of New Orleans New Orleans, Louisiana 70122 504-283-0323 Victor W. Lambou Advanced Monitoring Division Environmental Protection Agency Las Vegas, Nevada 89114 8-595-2695 Dr. Chung-Chiun Liu Case Institute of Technology Case Western Reserve University Cleveland, Ohio 44106 216-368-4122 Dr. J. A. Llewellyn College of Engineering University of South Florida Tampa, Florida 33620 813-974-2581 Dr. Mark E. Meyerhoff Assistant Professor Chemistry Department of Chemistry University of Michigan Ann Arbor, Michigan 48109 313-763-5916 Dr. Dora R. May Passino U.S. Fish and Wildlife Service Division of Fishery Ecology Research 1451 Green Road Ann Arbor, Michigan 48105 313-994-3331 Richard W. Paulson U.S. Geological Survey Water Resources Division MS 460 National Center Reston, Virginia 22090 8-928-6014 or 703-860-6104 Dr. Horace G. Loftin Code 793 Naval Coastal System Center Panama City, Florida 32407 904-234-4183 Mr. Alan Petroff American Digital Systems 2227 Drake Avenue Huntsville, Alabama 35805 205-883-9323 242 Peter Petroff Chairman of the Board American Digital Systems 2227 Drake Avenue Huntsville, Alabama 35805 205-883-9323 Barbara S. Pijanowski N0AA/NMPP0 6010 Executive Blvd. Rockville, Maryland. 20852 8-443-8817 Dr. Alex D. Poularikas Department of Electrical Engineering University of Rhode Island Kingston Rhode Island 02881 401-792-2514 Dr. Norman Richards Gulf Breeze Laboratory U.S. Environmental Protection Agency Sabine Island Gulf Breeze, Florida 32561 8-686-9228 Dr. Harold E. Schlichting, Jr. Biocontrol Co. , Inc. Port Sanilac, Michigan 48469 & University of Minnesota Biological Station Lake Itasco, Minnesota 56460 313-622-9980 Dr. Donald R. Sellers Senior Biochemist Life Sciences Department Midwest Research Institute Kansas City, Missouri 64110 816-753-7600 Groege V. Shaefer Interoceans Systems, Inc. 3540 Aers Court San Diego, California 92123 714-565-8400 Dr. Morton Smutz Office of Ocean Engineering National Oceanic and Atmospheric Administration Rockville, Maryland 20852 301-443-8385 Mr. Maurice E. Ringenbach NOAA/OMT 6001 Executive Blvd. Rockville, Maryland 20852 301-436-6914 Dr. Andrew Robertson NOAA/GLERL 2300 Washtenaw Avenue Ann Arbor, Michigan 48104 8-378-2281 Mr. Eugene M. Russin Chief, Sensor Test Branch Test and Evaluation Lab. , Code C651 National Oceanic and Atmospheric Admin. Rockville, Maryland 20852 202-426-9090 Arthur Solman Instrumentation Engineering Water Research Centre Medmenham Laboratory P.O. Box 16, Henley Road Medmenham, Mar low Buckinghamshire SL7 2HD Tel: 01-499-531 Dr. Fred Thurberg Research Physiologist National Marine Fisheries 212 Rogers Ave. Milford, Connecticut 06460 203-878-2459 Dr. Harris H. White NOAA/OA Code C22 Rockville, Maryland 2085 2 301-443-8242 243 \i Dr. Donald E. Wohlschlag Institute of Marine Science Port Aransas Marine Laboratory University of Texas Austin, Texas 78712 512-471-3434 or Port Aransas Marine Laboratory P.O. Box 1267 Port Aransas, Texas 78373 512-749-6743 INVITED GUESTS Dr. Ferris Webster Assistant Administrator for R&D National Oceanic and Atmospheric Administration Rockville, Maryland 20852 8-443-8344 Dr. Henry F. Enos EPA Director, Gulf Breeze Laboratory Gulf Breeze, Florida 32561 904-932-5311 Ms. Geraldine Cripe Gulf Breeze Laboratory Gulf Breeze, Florida 32561 904-932-5311 ~. . * u. S. GOVERNMENT PRINTING OFFICE : 1981 340-997/348 ^OF^ **rES o* ?' *E7^^™»™™bs A00007ia^0q53