Jump to main content.

Research Project Search
 Enter Search Term:
   
 NCER Advanced Search

Final Report: A Western Center for Estuarine Indicators Research which will Develop Indicators of Wetlands Ecosystem Health

EPA Grant Number: R828676
Center: Pacific Estuarine Ecosystem Indicator Research (PEEIR) Consortium
Center Director: Anderson, Susan L.
Title: A Western Center for Estuarine Indicators Research which will Develop Indicators of Wetlands Ecosystem Health
Investigators: Anderson, Susan L. , Allen, John , Bennett, Bill , Brooks, Andrew , Carr, Robert Scott , Cherr, Gary N. , Fujiwara, Masami , Green, Peter , Grosholz, Edwin , Hwang, Hyun-Min , Kendall, Bruce E. , Morgan, Steven , Murdoch, William W. , Nisbet, Roger M. , Ogle, Scott , Pawley, Anitra , Rose, Wendy , Stewart-Oaten, Allan , Swanson, Christina , Vorster, Peter
Institution: University of California - Davis , Pacific EcoRisk , The Bay Institute , U.S. Geological Survey , University of California - Santa Barbara , University of California - Santa Cruz
EPA Project Officer: Levinson, Barbara
Project Period: October 1, 2000 through September 30, 2004
Project Amount: $5,998,221
RFA: Environmental Indicators in the Estuarine Environment Research Program (2000)
Research Category: Ecological Indicators/Assessment/Restoration

Description:

Objective:

The goal of the Center was to develop new indicators of estuarine wetland health in marsh plants and animals so that corrective management of the nation’s wetlands can be implemented before it is too late, difficult, or prohibitively expensive. The project spanned all levels of the biological hierarchy from strands of DNA to populations to the landscape and involves ecotoxicologists, ecologists, biochemists, microbiologists, and remote sensing experts. Research was conducted at three sites in Tomales and San Francisco Bays in northern California and three sites in southern California across gradients of pollution impact from heavy metals, organic compounds, and excess nutrients.

PEEIR and the Resident Species Approach to Ecological Risk Assessment

Issue. Managers have long hoped for an integrated and practical approach to evaluate effects of toxic substances on resident estuarine species. Assessing toxicant responses in relation to the multiple stressors affecting our coast also is a significant goal, but it remains out of reach of current methodologies. For example, a “triad” approach, comprised of analytical chemistry, toxicity testing, and benthic surveys, currently is recommended for management of contaminated sediments, but this scheme is limited by poor linkage to the ecology of resident species, poor detection of chronic effects, and lack of consideration of multiple stressors. Nevertheless, this has been the best available management technique for more than a decade. A multitude of methods now exist to evaluate sublethal and chronic effects in estuarine species; yet, these are almost never utilized by managers. What are the barriers to implementation of updated techniques? Are the advantages worth the effort required to validate a complementary approach and develop management alternatives?

Major impediments include lack of consensus about: (1) what methods to use and how to integrate the results and interpret their meaning; and (2) where and when multiple lines of evidence are required to complete a viable management framework.

Major advantages include evaluation of: (1) effects in resident species, eliminating extrapolation from nonresident to resident species; (2) chronic and sublethal effects, which possibly are widespread but are not being evaluated; (3) long-term effects of contaminant mixtures; (4) multiple stressors; (5) multiple time scales or relevant spatial scales; and (6) linkage to effects on populations.

To realize these advantages, a new generation of ecological indicators is required. Managers need practical and effective techniques for a wide range of applications, from the management of contaminated sediments to more effective wetland restorations. A novel integrated science approach has to be employed that links the disciplines of ecology and toxicology in a rigorous manner. Toward that end, the Pacific Estuarine Ecosystem Indicator Research (PEEIR) consortium devised the Resident Species Portfolio (RSP) approach.

Approach and Rationale. PEEIR has developed a nested portfolio of indicators for resident salt marsh species, and our findings demonstrate the feasibility of ultimately creating such portfolios for any coastal or estuarine habitat type. Our RSPs are based on the principle that effects of stressors occur in individual organisms; yet, it often is populations and communities that managers care about the most.

PEEIR developed indicators and integration tools that: (1) refine our ability to identify stressors as well as characterize their effects on individual physiology, and (2) allow physiologically based indicators to be related to population and community phenomena. Indicators were integrated at multiple levels of biological organization (e.g., cellular, physiological, organismal, population, community) in carefully selected resident California salt marsh species (Figure 1).

We chose: (1) the mudsucker Gillichthys mirabilis and the shore crab Pachygrapsus crassipes as sentinel species because they are abundant, inhabit mud burrows, and remain in the same sites (~30 m range) for juvenile and adult phases; (2) the marsh plants pickleweed, Salicornia, and cordgrass, Spartina, because they comprise the majority of the plant biomass in the marsh, they are readily sampled, and they are an important conduit for food chain transfer of heavy metals; and (3) community- and/or ecosystem-level investigations in microbes using molecular techniques, evaluation of nutrient enrichment in the marsh food chain using stable isotopes, and estimation of changes in benthic and bird community diversity using parasites as a surrogate indicator, because each of these provided a snapshot of marsh condition that could not be obtained with standard survey techniques.

Exposures to toxic substances and/or nutrient enrichment were assessed in target tissues as well as in the analyses of sediment or water. Evaluation of exposure and effects in the same tissues was a critical component of our approach.

Figure 1. The PEEIR Conceptual Model is Based On Nested Indicators in Resident Salt March Species

Program design and sampling activities were thoroughly integrated, with more than 30 research laboratories quantifying indicator responses in each of these species synoptically. Fish indicators were measured by removing all necessary tissues from an individual fish, conducting the appropriate laboratory analyses, then finally integrating all data for each individual fish. Essentially, we would take the fish apart and put it back together again!

We sampled seven sites, over 400 km of coast in northern and southern California (Figure 2). Sampling was conducted at fixed stations within sites at three tidal elevations. Field and laboratory experiments also were undertaken to elucidate key relationships among indicators and to further characterize specific exposures.

An important feature of the RSP approach is that variables needed to quantify the population-level significance of indicator responses via modeling investigations were considered carefully in the overall sampling design.

Figure 2. Aerial Photo Showing Locations of PEEIR Research Sites Along the California Coast

The integrative methodology employed in PEEIR could be used to make significant improvements in ecological risk assessment. Imagine a clinical study that neither controlled for confounding variables such as age or lifestyle and sex, nor measured all experimental variables in the same subject! Yet, this is exactly the type of design that is applied frequently to coastal assessments. They are most often a patchwork of indicators applied in various species. This makes it impossible to ascribe relative effects of multiple stressors statistically or to link biomarker responses to populations. To reiterate, key integrative concepts included the use of: (1) selected resident species for a specific habitat type to simplify the species selection process; ( 2) integrative sampling with most effects measured simultaneously in individual organisms to make feasible integrative statistics related to chronic effects and multiple stressors, as well as to make more humane use of fish tissue samples; (3) integrative statistics to discern the relative significance of responses to multiple stressors, to relate chronic effects to contaminant exposures, and to ascribe population level effect; (4) ecological research and modeling to elucidate and further validate the population-level significance of key responses; and (5) interplay between laboratory and field efforts to refine relationship of chronic effects to specific contaminant responses. Hence, this type of integrative approach could be a valuable new paradigm in ecological risk assessment that puts more “eco” into ecotoxicology.

Summary/Accomplishments (Outputs/Outcomes):

Fish

The mudsucker (Figure 3) proved to be an excellent sentinel species for evaluating effects of contaminants in California salt marshes. We developed an integrative statistical approach to expressing fish condition using a synthesis of biomarkers and various morphometric measurements. In addition, indicators with a more specific application were developed for use in charting the effects of endocrine disrupting compounds as well as contaminants that cause genetic damage and tumors. Responses along these two potential pathways of effect were developed as practical indicators (Figure 4).

Figure 3. The Longjawed Mudsucker, Gillichthys mirabilis, was a Valuable Sentinel Fish Species for Studying Salt Marsh Condition

Figure 4. Indicators of Reproductive Impairment in Fish Included Markers of Endocrine Disruption (Choriogenins, Ovotestis) as well as Indicators of Tumorigenesis and Apoptosis (Cell Death)

Additional findings include:

The development of all of these responses as indicators, and the testing of modeling assumptions, involved a combination of integrated laboratory and field studies to understand the physiological and ecological mechanisms responsible for changes in indicator values.

Crabs

Investigations using the shore crab, Pachygrapsus crassipes (Figure 5), resulted in the development of a simple and useful indicator. By quantifying crab embryo abnormalities along with contaminant levels in the embryos, the potential detrimental effects of contaminants at various wetland sites can be evaluated.

Figure 5. Assessing Crab Embryo Abnormalities in the Field is a Practical Indicator of Salt Marsh Condition

Embryo abnormalities can be assessed easily with a hand lens, and in our studies, the abnormalities were indicative of other reproductive effects in the same crabs.

The relationship between heavy metals in embryos and the frequency of embryo abnormalities was stronger when only the embryos in the outer portion of the clutch were analyzed. Of course, this is because only the outer portion is exposed directly to sediment. This observation highlighted the need to evaluate exposure and effects in the same tissue whenever possible.

In-depth research on crab movement and demographics confirmed that movement in juvenile and adult phases was quite limited, and crabs were an effective indicator of anthropogenic inputs of nutrients and other contaminants into marshes at small spatial scales.

Demography (population size, size structure, sex ratio) was a poor indicator of toxic effects on crabs, but possible detrimental effects of the invasive green crab, Carcinus maenus, were detected.

In contrast to the findings on fish, biomarker responses in crabs were not predictive of contaminant exposure; hence, integrative statistics focused on the relationship between detrimental reproductive effects and contaminant exposure. Shore crabs appear to be an excellent sentinel species for monitoring the condition of California coastal habitats, and we envision that this approach likely could be extended to east and gulf coast marshes using fiddler crabs (Uca pugilator, U. pugnax, U. minax) and grapsid crabs (Sesarma cinereum, S. reticulatum).

Plants

PEEIR research on salt marsh plants was significant because it resulted in indicators that can be applied at multiple spatial scales (Figure 6). Although plants make up the majority of the biomass of a marsh, indicators are rarely developed for this component of the ecosystem.

Figure 6. Metal-Laden Salt Exudates From Spartina, (Left) and Remote Sensing of Species Distribution (Stege Marsh, Right) Were Both Used To Create Novel Indicators Using Marsh Plants (© Michael Rigsby)

Remote sensing was used as a rapid and cost effective tool to characterize species distribution patterns among marshes. In addition, remote sensing and analytical chemistry on salt exudates from marsh plants were paired to develop a novel indicator of metals mobilization across a marsh landscape. Analysisof metals in salt exudates is a rapid and cost-effective technique.

Further methodologies for plants were developed in the laboratory, but continued investigations will be needed for field validation. For example, a new biochemical method was developed that quantifies the potential sequestration of toxic metals into food webs. The method involved electrophoretic analysis of phytochelatins, which are peptides that tightly sequester toxic metals in plants. Significant efforts were also made to link physiologic responses of marsh plants to spectral changes that could be observed by remote sensing (Rosso, et al., 2005).

Ecosystems

A variety of integrative measurements for assessing aspects of ecosystem health were considered. One line of investigation utilized the diversity of trematode parasites of snails to indicate diversity of other organisms in salt marsh ecosystems, including benthic invertebrates and birds. This method was validated in PEEIR field studies in Southern California marshes as well as in a marsh restoration. We determined that larval trematodes in snails may be used to monitor changes in biodiversity over time and among sites and that the technique is more cost-effective than standard field surveys of bird and invertebrate populations. The Southern California Wetland Recovery Project is recommending this technique for ongoing monitoring.

Microbial ecologists in PEEIR assessed whether changes in microbial community assemblages can be used as an indicator of contaminant exposure and effects. Two techniques were contrasted: terminal restriction fragment length polymorphisms (TRFLP) and phospholipid fatty acid profiling. The techniques revealed similar, but not identical, patterns in microbial assemblages between marshes. Multivariate statistical analyses revealed that microbial patterns were related to heavy metal rather than organic contaminants. These also may be best applied as indicators of long term adaptation to low level contamination, as they appear to show greater responses than common acute toxicity assessments or sediment chemistry (Cao, et al., 2006; Grant No. R828676C003).

An additional contribution of the microbiology team was a demonstration of how indicators of nitrogen cycling in wetlands can be used to guide management of nutrient inputs into wetlands and may contribute to wetland remediation efforts.

Stressor Characterization

At all phases of the PEEIR program, new methods of stressor characterization in marsh systems were identified. Some of these utilized the indicator species directly, and others simply were linked by field observations and laboratory and field experiments to the indicators. Nutrient inputs into marshes were evaluated using stable isotopes, and this provided an integrated indicator of nutrient enrichment. For many applications, the stable isotope technique is superior to the traditional approach of measuring dissolved nitrogen concentrations. Moreover, the PEEIR modeling team has used the field data on proportions of stable isotopes of nitrogen to better characterize trophic interactions and flow of nitrogen in marshes. Analytical chemistry of heavy metal and organic contaminants on sediment and tissues was extensive (Hwang, et al., 2006a, b) and involved innovative methods to characterize some chemical stressors as well as to rank the relative level of contamination among sites or provide linkage to specific effects.

In the course of the fish research, we field-tested a commercially available microplate assay for estrogenic potential to verify the presence of endocrine disrupting compounds in sediments. Linking the plate assay to fish effects was significant because this scheme can be used in the future to conduct bioassay-directed fractionation studies known as toxicity identification evaluations (TIE) to ascribe the relative potency of different chemical fractions. This would help to pinpoint which chemicals should be the target of any remediation activities.

Pathogen contamination in coastal sites is a serious problem that has resulted in beach closures that are costly to the economy and pose risks to human health.

Furthermore, wetland restoration activities can remobilize pathogens buried in sediment. As a portion of the microbe investigations, PEEIR researchers worked with the Southern California Coastal Water Research Project to compare several molecular techniques that might be used to characterize microbes in wastewater and, after further development, sediment samples as well. The TRFLP technique was among the most powerful in discriminating various sources of fecal contamination such as human, cow, dog, and gull (Figure 7). The next step is for researchers to improve the methods for applications to coastal sediment and water samples.

Figure 7. Principal Components (PC) Analysis of TRFLPs Clearly Discriminated Between Waste Sources Generated from Two Sewage Samples (S) and Feces from Cattle (C), Dog (D), Human (H) and Seagull (G)

Additional contributions on stressor characterization include: (1) remote sensing methodology to assess changes in marsh plane elevations attributable to climate change or direct habitat alteration; (2) a larval fish bioassay that improves on standard techniques by using growth rate as an endpoint (Rose, et al., 2005); and (3) a bioassay that uses a specific developmental abnormality in sea urchin embryos, called exogastrulation, to characterize exposures to polycyclic aromatic hydrocarbons (Pillai, et al., 2003).

Applications

Implementation of the RSP approach has three phases (Figure 8). The first phase is selection of resident species appropriate for the habitat type and major stressors in question. We found that two animals with different ecological niches, but some shared characteristics, together with one or two plant species provided an informative grouping for salt marsh habitats. However, selection of the number of species in a portfolio would depend on the nature of the problems under investigation. Species selection is informed by simple ecological models and practical considerations such as availability, ease of sampling, and reproductive strategy. Once species are selected, then possible indicators or measurements should be listed for further consideration.

Figure 8. The Three Phases of Resident Species Portfolio Development and Implementation

The second phase of RSP implementation involves the iterative development of a portfolio. This process is somewhat analogous to the development of an investment portfolio. Investment also is a complex process that can be simplified by applying selected principles to a step-wise decision making framework. The steps include:

STEP 1: Clarify the goal and timeframe of the investigation.

Just as an investment portfolio will vary dramatically for different goals (college? retirement?) any study of estuarine sites involves many choices. Is there a specific regulatory deadline or question? Is this a long term monitoring program with specific goals? Is this an assessment of resident species in decline?

STEP 2: Invoke concepts of risk and uncertainty.

Like financial investment, the goal of estuarine studies is to minimize uncertainty and maximize return. We want the most information possible for the dollar invested, with the least risk to the resources. Hence, managers must clearly characterize the potential resource risks and their magnitude (serious decline of a species? loss of major habitat or small section of marsh?) as well as the level of uncertainty regarding stresses on the resource. Once this is accomplished, the size and type of portfolio can be characterized.

STEP 3: Determine the appropriate level of diversification.

Typically, a financial portfolio will include a balance of diverse investments. In choosing indicators for RSPs, diversification also is an asset. In general, we distinguish between three types of indicators: (1) condition indicators, which provide a general indication of individual health (e.g., histopathology, reproductive output, growth); (2) diagnostic indicators, which give clues as to what stressors may be causing impaired condition (acetylcholinesterase enzymes, endocrine disruption); and (3) stressor indicators, which are used to quantify stressor levels (contaminants in sediment or tissues, nutrient levels, pathogens).

A poorly characterized problem involving a large resource risk should result in the development of a large portfolio with an emphasis on condition and stressor indicators. Tissues for diagnostic indicators should be collected, a subsample analyzed and the remainder archived. In contrast, if the problem is an assessment of a well-focused issue, then diagnostic and stressor indicators would be used in a careful and definitive analysis of what contaminants are causing problems in the condition of a species.

STEP 4: Apply iterative process to final program design.

Figure 8 (Implementing the PEEIR Portfolio) illustrates the simple iterative process that is used in undertaking an investigation. This process is easy to implement given two conditions. The first condition is that core participants in three fields must be identified and must be willing to work in an integrative manner. The three core groups include: ecologists, environmental chemists, and ecotoxicologists who work at multiple levels of biological organization (cellular, physiological, organismal, population, community). Secondly, a plan for archiving of tissues and sediments is required, usually necessitating long-term access to an ultracold freezer.

The third phase of RSP implementation is indicator integration and interpretation, and it must blend seamlessly with progress in the second phase. This phase involves indicator interpretation and population-level interpretations. Although it is tempting to prescribe one approach to statistical integration for each species in the portfolio, we have found that this is not productive primarily because the different species and indicators respond very differently to specific stressors. However, for fish indicators, sequential multivariate analyses represent an exciting possible approach for describing relationships among indicators and between indicators and stressors.

The approach also provides a pathway for relating indicator responses to populations. Population modeling creates value added in describing various mechanisms of population response and potential relationship between incremental changes in indicators and population endpoints such as size structure, reproductive success, and population growth rate.

The Portfolio Approach can be used in a variety of resource investigations including: cleanup and remediation of hazardous waste sites, wetland restorations, sediment quality management, evaluation of species declines, long-term monitoring of resources, total maximum daily load (TMDL) investigations, assessment of the relative success of regulatory activities, and evaluation of the relative effects of emerging compounds.

Comparison of the RSP approach to a “Sediment Quality Triad” framework indicates that the multiple lines of evidence provided by the Triad and RSP method together can strengthen ecological risk assessment and reduce uncertainties significantly). A comparison to the Index of Biotic Integrity (IBI) also is underway and has been coupled with a powerful new example, of several indicators combined, to create a scorecard that will improve management of San Francisco Bay.

Conclusions:

The RSP approach represents a practical framework that can be implemented in a stepwise process, but we caution that maintaining an integrated approach is essential for success. Our most well-developed and cost-effective indicators would be no more expensive to utilize in management than a typical chronic toxicity test. These include:

Additional indicators that are ready for use but require university/agency partnership or special techniques that are not yet widely available include:

Promising new methodologies that would require further field validation include:

Our Web site will be updated as new publications on these indicators and their applications emerge. Using the three-phased process outlined above, the RSP method can be implemented in a wide variety of management applications. PEEIR has demonstrated the potential of this concept for salt marsh ecosystems, but it can be applied to other habitat types as well.

A promising new paradigm in ecological risk assessment has been created with this novel blend of ecology and toxicology (Figure 9). Integrated science is a powerful tool needed to develop the next generation of indicators that address the problems of scale, the simultaneous effects of multiple stressors, and the true complexities of nature.

Figure 9. The Synergy Between Ecologists and Toxicologists Was One of the Keys to the Success of the PEEIR Program

What PEEIR Data Tell Us About Sediment Quality Objectives

Issue. Sediments have been recognized as the final repository for organic contaminants and trace metals. Toxic chemicals accumulated in sediments, however, can penetrate back into the ecosystem by dispersing and moving up food chains and can cause unanticipated adverse effects. To better assess the quality of sediment, it is desirable to have multiple lines of evidence. The combination of rich chemical and biological datasets is helpful in developing appropriate management strategies for complex ecosystems.

Approach and Rationale. To determine sediment quality, various types of data were integrated, including sedimentary chemical concentrations, tissue body residues, acute porewater toxicity, and benthic invertebrate community profiles. Sedimentary concentrations of toxic chemicals were converted into toxic potentials (mean ER-M quotients, mERMQ) and then were compared to field biological indicators ranging from cellular to community levels according to methods devised by PEEIR. To determine the bioavailability of sedimentary contaminants, the body burdens of organic contaminants and trace metals in fish, clams, and crabs were measured. Marsh plant exudates were also measured.

Findings and Impact. Sedimentary chemical concentrations and PEEIR indicators showed the impact of land use patterns on the quality of sediment (Table 1 and see Figure 10 for location of PEEIR sampling sites). Toxicity test data and invertebrate community data also showed responses, but these were less conclusive. Fish tissue body residue and plant exudates data indicated that sedimentary contaminants are bioavailable and possibly can affect human health and wildlife. Fish indicators uncovered evidence of endocrine disruption, apoptosis, and tumors. Crabs exhibited decreased reproductive performance.

Table 1. Toxic Impacts of Sedimentary Contaminants in Salt Marshes

Figure 10. PEEIR Sampling Sites Along the California Coast.

Additional findings include:

Applications. The applications of the research include the following:

Limitations of Multi-Metric Indices, Including the Index of Biotic Integrity (IBI)

Issue. Biological integrity was a goal of the 1972 Clean Water Act but was at first assumed to be best measured by physical and chemical variables. However, measuring these variables can be impractical and/or expensive, and related biological effects can depend on their interactions and on how they vary in time and space. Measurement of biological variables now plays a large role in environmental management. Nearly all entities with monitoring responsibilities (e.g., states, territories, and tribes) use biological indicators. Comparing these indicators among sites, or to standards based on reference sites (i.e., relatively undisturbed), should help guide assessment and remediation efforts.

To facilitate such comparisons, many entities combine indicators into an overall multimetric index to summarize biological condition and distinguish degraded sites. At first, PEEIR considered developing such indices for West Coast wetlands. However, such standard indices require enough undisturbed sites in a region to estimate natural distributions of the indicators. West Coast wetlands are few in number, spread over many degrees of latitude, and are all disturbed. A new basis was needed. This is described in the Executive Summary, and involves multiple indices, each justified by its contribution toward an explicit goal.

This conclusion appears to have broad generality and is not particular to the wetlands studied by PEEIR. Appreciating this led us to re-evaluate the rationale behind the use of a single multimetric index to characterize biological integrity.

Approach and Rationale. We focused on the best-known standard index, the IBI. A large literature in journals, books, and agency reports generally supports its use and applies it to one or more situations, usually freshwater stream sites but also wetlands or other habitat types. From these articles we extracted a list of claims and supporting arguments made for IBI. We also made a list of criticisms from the few papers and related indices that criticize IBI and also listed responses made to these critics by IBI proponents. We took care not to miss the virtues of IBI or of indices in general or merely to repeat earlier criticisms. It is significant that the critiques, though seeming sensible and from respected authors, have had little effect on its use. Proponents might have invested too much in IBI to give it up, but a good critique must strive for a coherent argument against the strongest case for IBI.

IBI calculation involves choosing a set of biological indicator variables, such as numbers of species in various classes (like natives, insectivores, or “intolerant of degradation”). Their values are awarded numerical scores, based on comparison with the values at undisturbed sites. Metrics that approximate what biologists would expect at minimally disturbed sites are assigned a score of 5; those that deviate somewhat from such sites receive a score of 3; those that deviate strongly are scored 1. IBI is their sum.

Understanding the condition of a marsh ecosystem requires assessment of many aspects of this complex environment. A critical evaluation of how indicators are combined is essential. Figure 11 is an aerial photo of Carpinteria Salt Marsh, which, although relatively undisturbed, is subject to multiple stressors including habitat modification, excess nutrient inputs, and urban runoff.

Figure 11. An Aerial Photo of Carpinteria Salt Marsh. PEEIR sampling stations are indicated by white dots.

Findings and Impact. The value added by combining the indicator values into a single number is claimed to be: (1) more precise than single indicators because it combines them; (2) quantitative; (3) an effective way to communicate with nonspecialists; (4) analogous to other useful indices; (5) able to summarize large datasets to reveal patterns in space and time; (6) a synthesis of ecological theory and professional judgment; (7) flexible; (8) sensitive to many types of degradation; and (9) justified by its successful use in many studies.

Our main criticism is that combining indicators of different features or problems into a single number obscures information needed by managers, scientists, and the public. Claim 1 is weak at best. Sites really needing assessment will have intermediate IBI scores that could result from thousands of possible combinations of indicator values. For example, in a study undertaken by PEEIR researchers that used 12 indicators, there were 8074 ways of obtaining a value of 48 for the IBI. These combinations carry different messages. Equal IBIs are not all equal for decisionmaking because unequal ones do not always imply that the site with the lower score is more degraded. Hence, claim 2 is false. Previous critiques of IBI usually list this criticism with several others, some of them arguable. Perhaps the potential significance of this primary criticism of IBI has been lost amidst a raft of other concerns.

We identified ways that an index combining diverse measurements can give useful information. It can be deduced or derived from mechanistic models or statistical methods like regression or factor analysis, or constructed intuitively from considerations of mechanisms, importance, etc. All are part of PEEIR’s approach. The key is that it estimates or predicts an independently defined (possibly implicitly, as in factor analysis) target of interest.

IBI is not useless. It can identify extreme sites where almost all indicators are low, and because the indicator values still exist, calculating the IBI may seem harmless. However, we argue that it is likely to be actively harmful, especially in view of claim 3 above.

In summary, if all indicators indicate the same problem, the methods for combining them should have known properties and achieve clear objectives, preferably quantitative; if they indicate different problems, then combining them obscures their messages.

Applications. PEEIR advocates the development of indicators that are linked by key properties and clear objectives. We are working with several agencies to promote the use of RSPs based, in part, on statistical indicators for use in ecological risk assessment. In addition, we collaborated with The Bay Institute on publication of the San Francisco Bay Scorecard. The individual scores in the SF Bay Scorecard represent a situation where quantities to be aggregated relate to previously (independently) determined regulatory concerns—they would become meaningless if someone attempted to calculate a “GPA.”

Value Added to Indicators through Mechanistic Modeling

Issue. The effects of stressors are typically characterized by effects on individual organisms. PEEIR has been studying many indicators that refine our ability to identify stressors and characterize their physiological effects. This was achieved through a carefully designed mix of observations, experiments, and mathematical models focusing on resident organisms. These resident species are common organisms in many West Coast marshes but are not necessarily species of concern to managers. Related observations on the resident species to other species in different environments requires understanding the biological and ecological mechanisms responsible for changes in indicator values.

Approach and Rationale. PEEIR used dynamic energy budget (DEB) models for this purpose. DEB models describe the acquisition by organisms of energy from food and its utilization for growth, maintenance, and reproduction, as shown in Figure 12.

Figure 12. DEB Model Schematic

Modifications of the models can be used to describe the flow of elemental matter and bioaccumulation of contaminants. With additional information on mortality, DEB models are a key component of a population model.

DEB models are process-based and offer an attractive compromise between the competing requirements of generality (a model should be applicable to many combinations of species and stressors) and testability (predictions should be based on verified mechanisms, not speculation). By determining how model parameters varied among species and environments, and how the changes co-varied with biomarkers, we sought to relate biomarkers more rigorously to physiological processes, thereby adding value to these biomarker measurements.

Deterministic DEB models incorporating contaminant effects have been used by previous workers, notably S.A.L.M. Kooijman in the Netherlands, to interpret standardized toxicity tests; however, the PEEIR studies required new theory to describe growth and reproduction in fluctuating (stochastic) environments (Gurney, et al., 2003; Gurney and Nisbet, 2004; Fujiwara, et al., 2004), and new methods for estimating model parameters (Nisbet, et al., 2004; Fujiwara, et al., 2005).

Findings and Impact. The findings include the following:

Applications. DEB models have been shown to characterize quantitatively the responses of individual organisms to environmental stress. They also can be used as a component of population models. The work reported here establishes that indicators based on individual physiology can help in parameterizing models used to anticipate long-term population changes.

Models Relating Indicators of Physiological Condition to Population Responses

Issue. The most interesting questions regarding ecosystem health address features of populations and communities: Are certain species declining, is there a robust food web, is species diversity high, etc.? However, with the exception of diversity, such characteristics are difficult to quantify directly unless the changes are dramatic and fast. PEEIR has focused much of its attention on indicators of the physiology and performance of individuals. These measures can be obtained much more rapidly than assessments of a population’s viability over multiple years and decades, and they are more easily linked to specific stressors. But can stressors that impact individuals also predict important properties of populations and communities?

Approach and Rationale. Sensitivity analysis quantifies the extent to which stressor-induced changes in individual physiology and performance, as measured by individual indicators, leads to changes in population and community performance, as described by population endpoints. Proportional changes in the endpoints are associated with proportional changes in the individual indicators, often termed elasticity. Because most of the endpoints are difficult to quantify directly, we used mathematical models to investigate the relationship between the indicators and the endpoints.

For the resident fish indicator (longjaw mudsucker, Gillychthys mirabilis), we used “dynamic energy budget” models to link individual performance to growth and reproduction; these were coupled with mortality models that included size-specific predation and direct effects of stressors on individual survival (Figure 13). The various individual indicators are correlated with various parameters in the model; the endpoints were determined by the model. Thus, we can mathematically map the relationships between changes in the indicators and changes in the endpoints.

Figure 13. Schematic View of Estuarine Fish Life History. Growth depends on temperature, food availability, and sublethal stressor burden; mortality is caused by starvation, predation, and lethal stressors, and often is size dependent. Reproductive output is strongly size-dependent. Eggs and very young fish (larvae) may move between nearby marshes.

Population endpoints include:

Findings and Impact. We parameterized the growth model using the longjaw mudsucker, and we estimated survival from mark-recapture data on this species at our field sites. Fecundity was assumed to be limited by the number of burrows available to the fish. We investigated an effect of a toxicant that reduced assimilation efficiency and increased the maintenance cost. Because of the density dependence, population density did not change. However, the size structure was markedly changed, so that the mean individual size was lower (Figure 14). The result was that a 20 percent reduction in feeding and maintenance efficiency caused a 50 percent reduction in population biomass.

Figure 14. Stable Size Distribution of a Modeled Mudsucker Population Not Subject to Toxicants (left) and Subject to a Toxicant That Reduces Feeding Efficiency by 20% and Increases Maintenance Costs by 20% (right)

Applications. This approach will be useful to agencies seeking to understand the population consequences of observed changes in the condition of individuals.

The San Francisco Bay Index (Ecological Scorecard): A Tool for Summarizing Condition at Regional Scales

Issue. Large-scale regional restoration and management programs need measures of ecosystem condition, but only a handful of programs have developed compelling reporting tools to communicate results to the public. The large number of organizations working on Bay Delta issues has impeded the development of a cohesive suite of indicators. The Bay Institute developed the Bay-Delta Ecological Scorecard to provide a regional, landscape-level assessment of ecosystem condition and trends to facilitate communication with the public, managers, and decisionmakers.

Nearly 40 indicators are presented as eight multimetric indices that track the Bay’s environment (Habitat, Freshwater Inflow, Water Quality), its fish and wildlife (Food Web, Shellfish, Fish), our management of its resources (Stewardship), and its direct value to the people who use it (Fishable-Swimmable-Drinkable). The process, products, and lessons learned provide an important example for developing report cards in other regions. PEEIR has provided support for refining the Scorecard and preparation of manuscripts for the scientific literature.

Approach and Rationale. San Francisco Bay is a large estuary with problems reminiscent of many highly developed aquatic systems. Using existing data sources, we divided the system into discrete ecosystem attributes, sought to assess condition and trends for each of these attributes, and compared results to historical conditions as well as agency-supported restoration targets and standards. The indicator development process, cohesive conceptual model, and results illustrate how existing regional level datasets can be used to develop a comprehensive overview of ecosystem quality to facilitate public-level communication.

In October 2003, we first reported on these results using a highly aggregated set of eight indices with associated scores, grades, and short and long-term trends. The scorecard for 2005 is shown in Figure 15. For many indices, we also reported on subregional conditions in the bay (Suisun, San Pablo, Central, and South Bays). For an example, see Figure 16. This approach avoids some common pitfalls of the IBI approach that have been described by the PEEIR group.

Findings and Impact. The findings include the following:

Figure 15. Scorecard Indicating the Condition (Status) and Trends for Eight Indices for 2005

Figure 16. Not All Regions of SF Bay Are Faring Equally Well. Here, two subregions, Suisun and Central Bays, show examples of indicators that measure current ecological condition compared to 10-20 years earlier. China Camp and Stege Marsh are PEEIR study sites.

Applications. The applications of the research include the following:


Journal Articles: 34 Displayed | Download in RIS Format

Other center views: All 133 publications 35 publications in selected types All 34 journal articles

Type Citation Sub Project Document Sources
Journal Article Anderson SL, Cherr GN, Morgan SG, Vines CA, Higashi RM, Bennett WA, Rose WL, Brooks A, Nisbet RM. Integrating contaminant responses in indicator saltmarsh species [short communication]. Marine Environmental Research 2006;62(Suppl. 1):S317-S321. R828676 (Final)
R828676C002 (Final)
 not available
Journal Article Córdova-Kreylos AL, Cao Y, Green PG, Hwang HM, Kuivila K, LaMontagne MG, Van De Werfhorst LD, Holden PA, Scow KM. Diversity, composition, and geographical distribution of microbial communities in California salt marsh sediments. Applied and Environmental Microbiology 2006;72(5):3357-3366. R828676 (Final)
R828676C003 (Final)
 not available
Journal Article Fan TWM, Lane AN, Higashi RM. An electrophoretic profiling method for thiol-rich phytochelatins and metallothioneins. Phytochemical Analysis 2004;15(3):175-183. R828676C003 (2003)
R825960 (Final)
 not available
Journal Article Fan TWM, Lane AN, Chekmenev E, Wittebort RJ, Higashi RM. Synthesis and physico-chemical properties of peptides in soil humic substances. Journal of Peptide Research 2004;63(3):253-264. R828676C003 (2003)
 not available
Journal Article Field KG, Chern EC, Dick LK, Fuhrman J, Griffith J, Holden PA, LaMontagne MG, Le J, Olson B, Simonich MT. A comparative study of culture-independent, library-independent genotypic methods of fecal source tracking. Journal of Water and Health 2004;1(4):181-194. R828676 (Final)
R828676C003 (Final)
R827639 (Final)
 not available
Journal Article Fleming EJ, Mack EE, Green PG, Nelson DC. Mercury methylation from unexpected sources: molybdate-inhibited freshwater sediments and an iron-reducing bacterium. Applied and Environmental Microbiology 2006;72(1):457-464. R828676C003 (2004)
R829388C001 (2005)
  • Abstract from PubMed
  • Full-text: AEM Full Text
    Exit EPA Disclaimer
  • Other: AEM PDF
    Exit EPA Disclaimer
    • Journal Article Fujiwara M, Kendall BE, Nisbet RM. Growth autocorrelation and animal size variation. Ecology Letters 2004;7(2):106-113. R828676 (2003)
    R828676 (2004)
    R828676 (Final)
    		 not available
    Journal Article Fujiwara M, Kendall BE, Nisbet RM, Bennett WA. Analysis of size trajectory data using an energetic-based growth model. Ecology 2005;86(6):1441-1451. R828676 (Final)
    R828676C001 (2004)
    		 not available
    Journal Article Green PG, Fan TW-M, Higashi RM. Salt exudations from coastal wetland plants as a measure of metal mobilization from sediment. Environmental Toxicology and Chemistry (submitted, 2004). R828676C003 (2003)
    		 not available
    Journal Article Gurney WSC, Nisbet RM. Resource allocation, hyperphagia and compensatory growth. Bulletin of Mathematical Biology 2004;66(6):1731-1753. R828676 (2004)
    R828676 (Final)
    		 not available
    Journal Article Hechinger RF, Lafferty KD. Host diversity begets parasite diversity: bird final hosts and trematodes in snail intermediate hosts. Proceedings of the Royal Society B: Biological Sciences 2005;272(1567):1059-1066. R828676 (Final)
    R828676C001 (2004)
    R828676C003 (Final)
    		 not available
    Journal Article Huspeni TC, Lafferty KD. Using larval trematodes that parasitize snails to evaluate a saltmarsh restoration project. Ecological Applications 2004;14(3):795-804. R828676 (Final)
    R828676C001 (2002)
    R828676C003 (Final)
    		 not available
    Journal Article Hwang H-M, Green PG, Higashi RM, Young TM. Tidal salt marsh sediment in California, USA. Part 2: Occurrence and anthropogenic input of trace metals. Chemosphere 2006;64(11):1899-1909. R828676 (Final)
    		 not available
    Journal Article Lafferty KD, Holt RD. How should environmental stress affect the population dynamics of disease? Ecology Letters 2003;6(7):654-664. R828676C001 (2002)
    		 not available
    Journal Article Lafferty KD, Hechinger RF, Lorda J, Soler L. Trematodes associated with mangrove habitat in Puerto Rican salt marshes. Journal of Parasitology 2005;91(3):697-699. R828676C001 (2004)
    		 not available
    Journal Article Lafferty KD, Dunham EJ. Trematodes in snails near raccoon latrines suggest a final host role for this mammal in California Salt Marshes. Journal of Parasitology 2005;91(2):474-476. R828676C001 (2004)
    		 not available
    Journal Article Lafferty KD. Is disease increasing or decreasing, and does it impact or maintain biodiversity? Journal of Parasitology. R828676C001 (2002)
    		 not available
    Journal Article Lafferty KD, Porter JW, Ford SE. Are diseases increasing in the ocean?. Annual Review of Ecology Evolution and Systematics 2004;35():31-54 R828676C001 (Final)
    		 not available
    Journal Article LaMontagne MG, Holden PA. Comparison of free-living and particle-associated bacterial communities in a coastal lagoon. Microbial Ecology 2003;46(2):228-237. R828676 (Final)
    R828676C003 (2003)
    R828676C003 (Final)
    		 not available
    Journal Article LaMontagne MG, Leifer I, Bergmann S, Van De Werfhorst L, Holden PA. Bacterial diversity in marine hydrocarbon seep sediments. Environmental Microbiology 2004;6(8):799-808. R828676 (Final)
    R828676C003 (2003)
    R828676C003 (Final)
    		 not available
    Journal Article Li L, Ustin SL, Lay M. Application of multiple endmember spectral mixture analysis (MESMA) to AVIRIS imagery for coastal salt marsh mapping: a case study in China Camp, CA, USA. International Journal of Remote Sensing 2005;26(23):5193-5207. R828676 (Final)
    R828676C003 (Final)
    		 not available
    Journal Article Magalhaes C, Bano N, Wiebe WJ, Hollibaugh JT, et al. Comparison of ammonium oxidizing bacterial phylotypes and function between biofilms and sediments of the Douro River Estuary, Portugal. Environmental Microbiology (in review, 2005). R828676C001 (2004)
    		 not available
    Journal Article Morgan SG, Spilseth SA, Page HM, Brooks AJ, Grosholz ED. Spatial and temporal movement of the lined shore crab Pachygrapsus crassipes in salt marshes and its utility as an indicator of habitat condition. Marine Ecology Progress Series 2006;314:271-281. R828676 (Final)
    R828676C001 (Final)
    		 not available
    Journal Article Pillai MC, Vines CA, Wikramanayake AH, Cherr GN. Polycyclic aromatic hydrocarbons disrupt axial development in sea urchin embryos through a β-catenin dependent pathway. Toxicology 2003;186(1-2):93-108. R828676C002 (2003)
    		 not available
    Journal Article Rose WL, Hobbs JA, Nesbit RM, Green PG, Cherr GN, Anderson SL. Validation of otolith growth rate analysis using cadmium-exposed larval topsmelt (Atherinops affinis). Environmental Toxicology & Chemistry 2005;24(10):2612-2620. R828676 (Final)
    R828676C002 (2004)
    		 not available
    Journal Article Rosso PH, Ustin SL, Hastings A. Mapping marshland vegetation of San Francisco Bay, California, using hyperspectral data. International Journal of Remote Sensing 2005;26(23): 5169-5191. R828676 (Final)
    R828676C003 (Final)
    		 not available
    Journal Article Rosso PH, Pushnick JC, Lay M, Ustin SL. Reflectance properties and physiological responses of Salicornia virginica to heavy metal and petroleum contamination. Environmental Pollution 2005;137(2):241-252. R828676 (Final)
    		 not available
    Journal Article Rosso PH, Ustin SL, Hastings A. Use of lidar to study changes associated with Spartina invasion in San Francisco Bay marshes. Remote Sensing of Environment 2006;100(3):295-306. R828676 (Final)
    R828676C003 (Final)
    		 not available
    Journal Article Shaw J, Aguirre-Macedo L, Lafferty KD. An efficient strategy to estimate intensity and prevalence: sampling metacercariae in fishes. Journal of Parasitology 2005;91(3):515-521. R828676C001 (2004)
    		 not available
    Journal Article Spilseth SA, Morgan SG. Evaluation of internal elastomer tags for small, mature crabs. Crustaceana 2005;78(11):1383-1388. R828676 (Final)
    R828676C001 (2004)
    R828676C001 (Final)
    R825689C028 (Final)
    		 not available
    Journal Article Steets BM, Holden PA. A mechanistic model of runoff-associated fecal coliform fate and transport through a coastal lagoon. Water Research 2003;37(3):589-608. R828676 (Final)
    R828676C003 (2002)
    R828676C003 (Final)
    		 not available
    Journal Article Ward JR, Lafferty KD. The elusive baseline of marine disease: are diseases in ocean ecosystems increasing? PLoS Biology 2004;2(4):542-547. R828676C001 (2003)
    		 not available
    Journal Article Gurney WSC, Jones W, Veitch AR, Nisbet RM. Resource allocation, hyperphagia, and compensatory growth in juveniles. Ecology 2003;84(10):2777-2787. R828676 (Final)
    		 not available
    Journal Article Hwang H-M, Green PG, Young TM. Tidal salt marsh sediment in California, USA. Part 1: Occurrence and sources of organic contaminants. Chemosphere 2006;64(8):1383-1392. R828676 (Final)
    		 not available
    Supplemental Keywords:

    watersheds, estuaries, ecological effects, bioavailability, ecosystem indicators, aquatic, integrated assessment, ecological effects, ecosystem indicators, , estuarine research, aquatic ecology, environmental indicators, ecosystem assessment, biological markers, biomarker, biomarkers, ecological assessment, ecological exposure, ecosystem condition, ecosystem health, ecosystem indicators, ecosystem integrity, environmental consequences, environmental indicators, environmental stress, environmental stressor, environmental stressors, estuaries, estuarine ecosystems, fish, plant indicator, statistical evaluation, , Ecosystem Protection/Environmental Exposure & Risk, Geographic Area, Scientific Discipline, Waste, RFA, Ecosystem/Assessment/Indicators, exploratory research environmental biology, Aquatic Ecosystems & Estuarine Research, Bioavailability, Ecological Risk Assessment, Aquatic Ecosystem, Ecological Indicators, Ecological Effects - Environmental Exposure & Risk, Ecosystem Protection, Ecology and Ecosystems, State, biomarkers, ecosystem condition, water quality, California (CA), ecosystem indicators, ecosystem health, environmental indicators, environmental stressor, environmental consequences, wetlands, fish , statistical evaluation, GIS, plant indicator, ecological exposure, aquatic ecosystems, biological indicators, biological markers, ecosystem integrity, environmental stress, Western Center for Estuarine Research, ecological assessment, estuaries, estuarine ecosystems, biomarker
    Relevant Websites:

    http://www.bml.ucdavis.edu/peeir/ exit EPA

    Progress and Final Reports:
    2001 Progress Report
    2003 Progress Report
    2004 Progress Report
    Original Abstract

    Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
    R828676C000 Pacific Estuarine Ecosystem Indicator Research (PEEIR) Consortium: Administration and Integration Component
    R828676C001 Pacific Estuarine Ecosystem Indicator Research (PEEIR) Consortium: Ecosystem Indicators Component
    R828676C002 Pacific Estuarine Ecosystem Indicator Research (PEEIR) Consortium: Biological Responses to Contaminants Component: Biomarkers of Exposure, Effect, and Reproductive Impairment
    R828676C003 Pacific Estuarine Ecosystem Indicator Research (PEEIR) Consortium: Biogeochemistry and Bioavailability Component

    Top of page

    The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.

    Local Navigation

    Jump to main content.