© 2002 by CRC Press LLC Species Sensitivity Distributions in Ecotoxicology © 2002 by CRC Press LLC Series Editor Michael C. Newman College of William and Mary Virginia Institute of Marine Science Gloucester Point, Virginia Environmental and Ecological Risk Assessment Published Titles Coastal and Estuarine Risk Assessment Edited by Michael C. Newman, Morris H. Roberts, Jr., and Robert C. Hale Risk Assessment with Time to Event Models Edited by Mark Crane, Michael C. Newman, Peter F. Chapman, and John Fenlon Species Sensitivity Distributions in Ecotoxicology Edited by Leo Posthuma, Glenn W. Suter II, and Theo P. Traas LEWIS PUBLISHERS A CRC Press Company Boca Raton London New York Washington, D.C. Edited by Leo Posthuma Glenn W. Suter II Theo P. Traas Species Sensitivity Distributions in Ecotoxicology This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior permission in writing from the publisher. The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC for such copying. Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. Visit the CRC Press Web site at www.crcpress.com © 2002 by CRC Press LLC St. Lucie Press is an imprint of CRC Press LLC No claim to original U.S. Government works International Standard Book Number 1-56670-578-9 Printed in the United States of America 1 2 3 4 5 6 7 8 9 0 Printed on acid-free paper Library of Congress Cataloging-in-Publication Data Catalog record is available from the Library of Congress © 2002 by CRC Press LLC Foreword Different species have different sensitivities to a chemical. This variation can be described with a statistical or empirical distribution function, and this yields a species sensitivity distribution (SSD). The idea to use SSDs in risk assessment originated almost simultaneously in Europe and in the United States. Scientists began to use these distributions for the derivation of environmental quality criteria, challenged by policy makers to make optimal use of single-species toxicity test data for chem- icals. This development coincided with the notion that risks cannot be completely eliminated but should be reduced to an acceptable low level. In 1990, the Organization for Economic Cooperation and Development (OECD) Hazard Assessment Advisory Body organized a workshop in Arlington, Virginia, to discuss these and other approaches for extrapolation of laboratory aquatic toxicity data to the real environment. The extrapolation workshop, together with other work- shops on the application of quantitative structure–activity relationships (QSARs) to estimate ecotoxicity data (Utrecht, the Netherlands) and effects assessment of chem- icals in sediment (Copenhagen, Denmark), formed the backbone of the OECD Guid- ance Document for Aquatic Effects Assessment , which was published in 1995. This guidance document is applied, for example, in the OECD existing chemicals program. As head of the OECD Environment, Health and Safety Division, which supported the transatlantic discussions on the use of SSDs in 1990, it is a great pleasure to see that this specific approach in ecotoxicology has been taken up by scientists and is still developing. The fact that it has become so well used in environmental manage- ment should not keep us from being critical and demanding about the scientific rationale and validity of the methods used. It is my firm belief that this book contributes to this goal and that it serves as an excellent stimulus to pursue the continued development of SSD-based risk assessment in ecotoxicology. Rob Visser Head, Environment, Health and Safety Division Organization for Economic Cooperation and Development © 2002 by CRC Press LLC Preface AIMS OF THE BOOK The aims of this book are many, but the most important ones are the following: • First, the concept that is the subject of the book, species sensitivity dis- tributions , is a practical method in ecological risk assessment and in decision-making processes. It is used in the derivation of environmental quality criteria and in ecological risk assessment of contaminated ecosys- tems. The question is, whether the past adoption of the concept has been a good decision, especially in view of the large investments in preventive and curative actions resulting from decisions based, fully or in part, on application of the concept. The editors, all working in governmental institutes, felt a sense of urgency in the air to summarize the state of the art of the concept, its scientific underpinning, its current uses, and its predictive accuracy, after approximately two decades of convergent evo- lution on two continents. Eventually, a review of the state of the art should promote better understanding of all issues relevant to the SSD concept and its applications. Therefore, the major aim is a better understanding of the science of ecological risk assessment concerning the use of a practically adopted method. • Second, the many relevant publications by academic, regulatory, and industrial scientists in North America and Europe have been scattered throughout the literature. Few papers have been published in the easily accessible scientific journals; many are in the “gray literature.” Further- more, most texts explain the issues in various, context-dependent lan- guages, with local jargon added. The secondary aim, necessary to understand the science, is to bring together open and gray literature, and to make the sources available in clear language in this book. • Third, by compilation and study of the available material and by review of past criticisms of the SSD concept and the solutions offered so far, a final aim becomes apparent. This aim is to suggest paths forward, to suggest solutions for the most relevant criticisms voiced in the past, and to break inertia in the evolution of the SSD concept itself. This should eventually lead to clear views regarding the advantages and limitations of the method for different applications. © 2002 by CRC Press LLC THE EVOLUTION OF EDITORIAL RISK The pursuit of these three aims began in 1998. At a conference in Bordeaux, organized by the European branch of the Society for Environmental Toxicology and Chemistry (SETAC), various Europeans working with the SSD concept were inspired by the local atmosphere to draft the raw outlines of a plan. After approximately 15 years of evolution on two continents, the need was felt to evaluate the SSD concept. The thought simmered for some time. It was brought to the Laboratory for Ecotoxicology at the Dutch National Institute for Public Health and the Environment (RIVM). At RIVM, Herman Eijsackers sowed the seed, and he and Hans Canton cared most for the undisturbed survival and growth of the young plant. In the next year, it grew into a formal RIVM project. RIVM employees were assigned to compile and evaluate the current state of the art, and to formulate ways forward. This was deemed a necessary task for RIVM, since many sites in the Netherlands are exposed at concentrations exceeding the Dutch Environmental Quality Criteria, and the project was expected to help answer the question: “What are the quantitative eco- logical risks of mixtures of chemical compound concentrations in the environment that exceed the Environmental Quality Criteria?” The efforts were supported by scientific advisory bodies of the RIVM. Soon, the RIVM project became an inter- national project, and the review plan reshaped into a book plan, with international editorship and contributions. The addition of a North American editor to this effort continued a connection that began at a 1990 OECD workshop on ecotoxicological extrapolation models (OECD, 1992). The most significant result of that workshop was the realization that a common approach was being used in the United States, the Netherlands, and Denmark to extrapolate from single species toxicity test results to biotic communi- ties. Because there was no name for that class of models, the Working Group B rapporteur coined the term species sensitivity distributions . That workshop contrib- uted to the subsequent expansion of the use of SSDs from the setting of regulatory criteria into the emerging field of ecological risk assessment. More to the point, it established the contacts and common interests among users of SSDs in North America and Europe that made this volume possible. ECOLOGICAL AND AUTHORSHIP RISKS The contributors to this book are specialists on risks, especially risks from chemical compounds in ecosystems. Especially they could have been reluctant to contribute to this book in view of various realistic risks associated with it. Nonetheless, they contributed of their own free will. What risks did authors and editors face? • First, they faced the risk that they would create a Gordian knot of risk concepts, definitions, and research results, when their goal was to unravel a knotty problem. If you try to imagine how to describe a Gordian knot, or a research plan to unravel it, you can guess how difficult that can be, especially when you want to do it in a scientific way. Where are the rope © 2002 by CRC Press LLC ends, and how do they causally connect? Those who contribute to a book on such a knotty problem might never be understood by readers or even by the other authors. • Second, there is the risk that the interpretation of the chosen risk definition (if any) would be strongly context dependent, yielding a hidden knot within a knot. In a scientific context, one can communicate about risks in a purely numerical context, without value judgments. In the societal context of risk-based decision making, however, risk has an aspect of value judgment. The contributors were aware of this extra complication, as they were recruited from those different contexts, so it was courageous to join. Thinkers and practitioners could have easily split, and two volumes rather than one volume could have resulted. • Third, there is the risk of interminable debate aroused by the published text, as a consequence of the preceding risks. The authors and editors could have chosen to keep the results of their debates among themselves, since the above risks were effectuated in their internal discussions. There might not have been a book at all. • Fourth, risks are associated with working on the border between science and policy. Scientists may develop methods that have policy implications, which may not be acceptable to policy makers or advocates for industry or the environment. Clearly, the assumption that SSDs are adequate mod- els of the environment is such a case, and work on the book could have been stopped by the employers of the authors or editors. • Fifth, publicizing controversial technical and conceptual issues may be unwelcome, because SSDs are firmly embedded in the regulatory practices of the United States, the Netherlands, and other nations. Regulators may not want to be told that the scientific foundations of their actions are still questionable or subject to change. • Sixth, confusion and conflict could have been almost invited by the editors by their wish to bring together two historical lines of SSD evolution (the North American and the European) in a single volume, each with its own context of adopted principles, terminology, and legislation. AUTHORSHIP RISKS IN PRACTICE The editors have seen some of these risks in practice. At the first public introduction of the SSD concept in Europe, it was the initiator of the plan for this book who, metaphorically, suggested killing the first messenger. In 1983, Bas Kooijman, from the Netherlands Organization for Applied Scientific Research (TNO), was asked by the Dutch Ministry of the Environment to help resolve the ethical question: “How much toxicity test data for how many species are needed to underpin adequate risk assessment based decisions?” As a result, an initial Dutch TNO report from 1985 and a well-known paper, in Water Research in 1987, were published on the risky subject of the derivation of hazardous concentrations for sensitive species. This evolved further when Nico van Straalen from the Vrije Universiteit Amsterdam was invited to give a thought-provoking introductory plenary lecture at a 1995 meeting © 2002 by CRC Press LLC of the Dutch Provisional Soil Protection Technical Committee (V-TCB). He began this lecture on SSD basics avant la lettre by stating that he felt as if he were putting his head on the guillotine, while the audience members were handed a rope to release the blade. The lecture was completed in full health, although the pertinent audience member said in a whisper that he would have liked to pull the rope. This illustrates the risks of the science policy debate on the SSD concept in a nutshell. POST-WRITING RISKS Despite these risks, the contributors have not been reluctant. They produced 22 chapters, and no authors left because of inability to describe their strand of the knot. The contributors also have been willing to project themselves into the role and context of their colleagues. The 22 chapters are thus in one book, not two. Although debates have been many, we hope scientific growth has resulted. On publication of this book, only the post-writing risks remain. There is a need of risk management here. The management of that risk is your task as reader, acting in your own professional environment after reading the book. To help you with this, we have done our best to present the science and applications to you in manageable portions, despite the double Gordian knot. We identified four sections: I. General Introduction and History of SSDs II. Scientific Principles and Characteristics of SSDs III. Applications of SSDs A. Derivation of Environmental Quality Criteria B. Ecological Risk Assessment IV. Evaluation and Outlook By arranging the chapters within these sections, the different focuses of the chapters are presented. We can help in managing the remaining risks only a bit further, by stating that our discussions profited first from clearly defining the word risk when it was used, second from clearly defining or recognizing the context of those involved in the debate, and third from clearly distinguishing the values obtained in risk calculations from value judgments. All scientific fields can be seen as Gordian knots. For the field of ecological risk assessment, we hope to have cut through some surface layers, and we hope to have freed thereby some useful lengths of rope. This book is the result of the risky effort of many people, who all hope that the field of ecological risk assessment benefits from their efforts. Leo Posthuma, Glenn W. Suter II, and Theo P. Traas © 2002 by CRC Press LLC Acknowledgments The editors wish to acknowledge the valuable contributions to this book by: • Olivier Klepper, for starting the process that evolved into this book; • The authors , who volunteered to contribute to this book with a chapter, and who adapted their chapters based on comments of anonymous peer reviewers, section editors, and editors, so as to optimize scientific quality within the chapters, and line of reasoning among chapters in the four sections and throughout the book; • The section editors , who helped to identify highly qualified potential peer reviewers, so that all chapters were read by reviewers representing two types, namely, those expected to be familiar with the environmental policy setting in the continent of the author and those almost completely unfa- miliar with that context; the latter helped remove unnecessary jargon; • The reviewers , who performed their peer-reviewing work with enthusi- asm, resulting in main-line comments and detailed suggestions on all chapters, which greatly improved the contents of the book. The reviewers are: Prof. Dr. Wim Admiraal Department of Aquatic Ecology and Ecotoxicology University of Amsterdam Amsterdam, the Netherlands Dr. Rolf Altenburger Centre for Environmental Research (UFZ-Umweltforschungszentrum) Leipzig, Germany Dr. Steve Bartell The Cadmus Group, Inc. Oak Ridge, Tennessee, USA Dr. Jacques J.M. Bedaux Institute of Ecological Science Vrije Universiteit Amsterdam, the Netherlands Prof. Dr. Hans Blanck Botanical Institute Göteborg University Göteborg, Sweden Dr. Kym Rouse Campbell The Cadmus Group, Inc. Oak Ridge, Tennessee, USA Dr. Rick D. Cardwell Parametrix, Inc. Kirkland, Washington, USA Dr. Gary A. Chapman Paladin Water Quality Consulting Corvallis, Oregon, USA Dr. Peter Chapman Jealott’s Hill Research Station Zeneca Agrochemicals Bracknell, United Kingdom [...]... continents (North America and Europe) The section illustrates the events that have occurred at the interface of science and regulation, homologies and divergence in SSD-based methods, and the need to unite the existing theories and applications © 2002 by CRC Press LLC 1 General Introduction to Species Sensitivity Distributions Leo Posthuma, Theo P Traas, and Glenn W Suter II CONTENTS 1. 1 1. 2 1. 3 1. 4 1. 5... Distribution 0 .1 0 -3 -2 -1 0 1 2 3 4 5 Log Concentration (µg.l -1 ) EQC HC5 FIGURE 1. 1 The basic appearance of SSDs, expressed as a CDF The dots are input data The line is a fitted SSD Forward use (arrows from X → Y) yields the PAF as defined in Chapter 4, or similar estimates of risk as defined by other authors (see also Chapters 5, 15 , and 17 ) Inverse use (arrows from Y → X) yields an EQC at a certain cutoff... II CONTENTS 1. 1 1. 2 1. 3 1. 4 1. 5 Introduction Variability and Species Sensitivity SSD Basics SSD-Related Questions 1. 4 .1 Ecotoxicological Issues Regarding the Input Data 1. 4.2 Statistical Issues 1. 4.3 Issues Related to Ecological Interpretation of SSD Output 1. 4.3 .1 Environmental Quality Criteria 1. 4.3.2 Ecological Risk Assessment Aims of the Book Abstract — The species sensitivity distribution (SSD)... desirable trait since it encourages testing Although single-chemical criteria continue to be derived and revised, the methods have not been updated since 19 85 and few new criteria have been developed in © 2002 by CRC Press LLC the United States during the 19 90s relative to the 19 80s and late 19 70s This is in part because of increased emphasis on alternative methods First, during the 19 80s, subchronic... mostly been used in the derivation of EQCs, both in the United States and in Europe (see Figure 1. 1, inverse use) More recently, SSDs have been used as models of risk to ecological communities or ecosystems as illustrated in © 2002 by CRC Press LLC Figure 1. 1, forward use (Solomon et al., 19 96; Klepper and van de Meent, 19 97; Cardwell et al., 19 99; Steen et al., 19 99; Chapters 15 through 20) In either case,... American History of Species Sensitivity Distributions Glenn W Suter II Chapter 3 European History of Species Sensitivity Distributions Nico M van Straalen and Cornelis J van Leeuwen SECTION II Scientific Principles and Characteristics of SSDs Chapter 4 Theory of Ecological Risk Assessment Based on Species Sensitivity Distributions Nico M van Straalen Chapter 5 Normal Species Sensitivity Distributions and... Chapter 19 Mapping Risks of Heavy Metals to Birds and Mammals Using Species Sensitivity Distributions Theo P Traas, Robert Luttik, and Hans Mensink Chapter 20 Ecotoxicological Impacts in Life Cycle Assessment Mark A J Huijbregts, Dik van de Meent, Mark Goedkoop, and Renilde Spriensma SECTION IV Evaluation and Outlook Chapter 21 Issues and Practices in the Derivation and Use of Species Sensitivity Distributions. .. overinterpretation of the output For example, Blanck (19 84) has already pointed out that a method that is based on single -species toxicity data cannot account for ecosystem-level events such as ecological interactions, implying strict logical limits on the interpretation of risks calculated with the SSDs 1. 4 SSD-RELATED QUESTIONS A basic assumption in ecological risk assessment is that laboratory-generated... variation in species sensitivities to toxicant exposure, and on a specific method to address this variation © 2002 by CRC Press LLC Different ecologists and ecotoxicologists independently designed ecotoxicological assessment systems based on the variance in response among species (Klapow and Lewis, 19 79; Mount, 19 82; Blanck, 19 84; McLaughlin and Taylor, 19 85; U.S EPA, 19 85a; Kooijman, 19 87) Interspecies... arrows in the graphs indicate that the SSD concept can be used in a “forward” as well as “inverse” way (Van Straalen and Denneman, 19 89; Chapter 4) For the inverse use, such as the derivation of environmental quality criteria, a cutoff percentage p is chosen (to protect 1- p percent of species, Y-axis), and the desired “safe” 1 0.9 0.8 0.7 0.6 0.5 Risk Assessment 0.4 PAF 0.2 L(E)C50 or NOEC 0.3 Species Sensitivity . imprint of CRC Press LLC No claim to original U.S. Government works International Standard Book Number 1- 5 667 0-5 7 8-9 Printed in the United States of America 1 2 3 4 5 6 7 8 9 0 Printed on acid-free. W. Suter II Theo P. Traas Species Sensitivity Distributions in Ecotoxicology This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted. in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior permission in writing