3

Analysis of the Program

BIRDS

Allocation of Program Effort

The panel has analyzed the allocation of Environmental Studies Program effort among marine regions and categories of study. The eight marine regions are those listed in the section on benthic studies. The studies are divided into seven broad categories, as follows:

  • Colony inventory. Location and mapping of breeding colonies and estimates of numbers of birds of each species in each colony.

  • At-sea inventory. Surveys and mapping of distributions of birds at sea, usually based on standardized transect methods.

  • Shoreline habitats. Surveys of habitats and their use by birds in nearshore, intertidal, and supratidal zones, including studies of distribution, abundance, and feeding of birds in these zones during the breeding season and on migration.

  • Colony processes. Studies conducted at breeding colonies of phenology, reproductive success, trophic processes, recruitment and age structure, and changes in numbers.

  • At-sea processes. Studies relating the distribution of birds at sea to physical and biological features of the ocean environment, including prey numbers and distribution.

  • Exposure and effects. Studies of factors that control direct and indirect exposure of marine birds to oil and physiological and reproductive effects of exposure.

  • Modeling. Integrative models of the consequences of oil spills, including estimates of contact with bird populations and consequent mortality, physiological, and population changes.

MMS maintains records of the allocation of contracts by administrative region and by lease-sale area, but the Ecology Panel chose to examine allocation of project effort and funds by marine region and by category of ecological study. The allocation developed by the panel was based on a list of contracts provided by MMS. For projects conducted in more than one marine region or in more than one year or covering more than one category of study, the panel allocated project effort according to the following rules:

  • One project-year was assigned to each region for every year for which the project was funded.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology 3 Analysis of the Program BIRDS Allocation of Program Effort The panel has analyzed the allocation of Environmental Studies Program effort among marine regions and categories of study. The eight marine regions are those listed in the section on benthic studies. The studies are divided into seven broad categories, as follows: Colony inventory. Location and mapping of breeding colonies and estimates of numbers of birds of each species in each colony. At-sea inventory. Surveys and mapping of distributions of birds at sea, usually based on standardized transect methods. Shoreline habitats. Surveys of habitats and their use by birds in nearshore, intertidal, and supratidal zones, including studies of distribution, abundance, and feeding of birds in these zones during the breeding season and on migration. Colony processes. Studies conducted at breeding colonies of phenology, reproductive success, trophic processes, recruitment and age structure, and changes in numbers. At-sea processes. Studies relating the distribution of birds at sea to physical and biological features of the ocean environment, including prey numbers and distribution. Exposure and effects. Studies of factors that control direct and indirect exposure of marine birds to oil and physiological and reproductive effects of exposure. Modeling. Integrative models of the consequences of oil spills, including estimates of contact with bird populations and consequent mortality, physiological, and population changes. MMS maintains records of the allocation of contracts by administrative region and by lease-sale area, but the Ecology Panel chose to examine allocation of project effort and funds by marine region and by category of ecological study. The allocation developed by the panel was based on a list of contracts provided by MMS. For projects conducted in more than one marine region or in more than one year or covering more than one category of study, the panel allocated project effort according to the following rules: One project-year was assigned to each region for every year for which the project was funded.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology For projects involving more than one group of animals (birds, pinnipeds, cetaceans, or turtles), the panel assumed that project funds were assigned equally to each group. The panel assumed that project funds were assigned equally to each region and to each category of study listed in the project title or in titles of reports or publications. It is recognized that these assumptions are imprecise, but they should give a reasonable approximation of the actual allocation of program efforts. An analysis by the panel of MMS's bird studies shows that the regional allocation of bird-related effort in the MMS program has been extremely uneven. Almost all the program effort was allocated to waters off Alaska (85% of project-years and 67% of program funds) or to waters off California (12% of project-years and 29% of program funds). The large amount of program effort in Alaska was not unreasonable, in view of the large numbers of marine birds that breed there, the relatively poor initial knowledge of these birds, the drilling activity in Alaskan waters in the early years of the ESP, and the high costs of investigation there. A large amount of effort was expended in the Beaufort Sea region, which has a small number of pelagic birds breeding there, but in fact the effort was spent mainly on investigation of shoreline and lagoon habitats, where migratory birds and waterfowl are concentrated and are potentially vulnerable to OCS activities. The allocation in California of 29% of all the program funds allocated to birds appears out of proportion to the small numbers of marine birds that breed in California, but it is not; much of the work was conducted on the numerous migratory and wintering species and on effects studies that could be generalized to other regions. Essentially no MMS studies of birds have been conducted in the Atlantic regions, although some studies have been conducted by other agencies (e.g., Erwin, 1979 (FWS); Korschgen, 1979 (FWS); Powers, 1983 (NMFS)). Only about 2% of the program's resources were allocated to studies in the Gulf of Mexico. The amounts of effort and funding allocated to the Gulf of Mexico and the Atlantic regions are much smaller than would have seemed appropriate, given the magnitude of existing or proposed OCS activity. The lack of follow-up studies of birds in areas with oil production, i.e., in the Gulf of Mexico and southern California, has made it impossible to learn the long-term effects of OCS production on birds. The allocation of program effort among the various categories of study appears reasonable. About 31% of program funds was allocated to studies on breeding colonies, including surveys, censuses, and studies of productivity and diet. About 39% of program funds was allocated to studies at sea, mainly on distributional surveys and diet studies during the breeding season. About 17% of program funds was spent on shoreline habitat studies, mainly in Alaska, and about 11% on studies of exposure to and effects of oil. Only 3% of program funds (three projects) was spent on constructing models of effects of oil spills. Assessment of Results The ESP provided a major impetus for the study of marine birds in the waters of the United States. It sponsored major surveys of the distribution of seabird colonies and the distribution of birds at sea. Through those efforts, our knowledge of the distribution and abundance of breeding North American seabirds has been greatly increased, particularly in Alaska, for which only sketchy information was previously available. ESP studies of the pelagic distribution and abundance of birds were, with a few exceptions, the first systematic attempts to study the pelagic biology of seabirds in U.S. waters, and they have provided the foundation of our knowledge of the marine biology of seabirds and the data base needed by investigators to pursue more narrowly defined projects, supported by other agencies.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology In the pelagic realm, the early ESP efforts led to techniques that are now standard in surveying, data recording, and oceanographic and remote sensing. The studies of foraging distribution and trophic webs greatly increased our general knowledge of seabird pelagic ecology and provided data on which later research would be based. In contrast, the colony-based studies did not result in the same degree of advancement of a field of study as did the pelagic work. Useful modeling exercises were carried out by the ESP, but it is not clear how lasting a contribution they will make, because of the specialized nature of the models and the apparent lack of follow-through in generalizing, validating, and applying them. Finally, in toxicology, the ESP has funded the development of excellent new analytical techniques for detecting and identifying hydrocarbon residues and of innovative field studies of the effects of hydrocarbon pollutants. The physiological studies, although narrowly directed at the effects of oil, have increased our awareness of the sensitivity of some native species and of the importance of working with native species in the wild, rather than relying solely on laboratory tests with standard animal models. The following evaluation provides a more critical analysis of the strengths and weaknesses of the ESP relative to its mandate for environmental impact assessment. In the broad overview, the program supported a wide variety of research endeavors that added greatly to our knowledge of marine birds in a very short period. Most of the knowledge is available in the “gray” literature, and a fairly large amount is becoming available in the refereed literature. The latter body of published material is strong evidence of the accomplishments of the ESP effort. Selection of Species and Locations for Study The panel was not provided with any documentation as to the methods or rationale for the selection of bird species or sites for study. Although general planning meetings were held for both the Southern California Bight region (Lavenbin and Earle, 1975) and the Alaska region (NMFS, 1975), they were too general to yield carefully designed plans (see also Bartonek and Nettleship, 1979). The selection of species appears to have been largely at the discretion of the various investigators and to reflect a combination of factors, such as ease of study and scientific interest, as well as potential vulnerability to oil. This mix of criteria seems reasonable; in most regions where there was a substantial ESP effort, the variety of species chosen provided a reasonably good coverage of the most abundant species, the species most vulnerable to oiling, and easily studied species that might be good indicators of local environmental change. In the pelagic studies, little apparent effort was made to target selected species for study. An exception was the study of shearwaters (Puffinus spp.) in Alaska (Guzman and Myers, 1982). Generally, individual investigators singled out particular species for more thorough analysis, e.g., brown pelican, Pelicanus occidentalis, in California (Briggs et al., 1981b, 1983a). For the most part, studies were aimed at the investigation of given areas and encompassed all species present. That was usually appropriate, in that the integrated study of marine ecosystems is more likely, as a first cut, to provide a broad understanding of the multiplicity of factors that influence the pelagic distribution and abundance of birds than would the study of individual species. The allocation of effort in the study of the breeding distribution of birds was on a colony or area basis, rather than on a species basis. Surface-nesting and cliff-nesting species received the best coverage, and burrow-nesting (particularly nocturnal) species received less complete coverage. Nocturnal species are very difficult to count, and the sizes of their populations remain

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology poorly known. To the extent that nocturnal species are potentially very vulnerable to oil pollution, more attention should have been paid to assessing the location and size of their nesting populations. In the colony-based studies of reproductive and trophic ecology, many factors appear to have influenced which colonies were chosen initially and the selection of species for study. Large colonies, important centers of the population of a given species, and accessibility were considered important. In no region was there a detailed plan at the outset for the coverage of particular species or a particular set of “representative” colonies. However, in most colonies investigated, a mix of species was used, including surface or cliff nesters and species that dive in pursuit of prey (alcids) and forage at or near the surface, often while airborne (kittiwakes, gulls, and terns). Studies of storm petrels, fulmars, and nocturnally active alcids were few and narrowly distributed. Conspicuously absent from the ESP for marine birds are studies of the several endangered or threatened species that occur in lease areas or habitats adjacent to OCS areas. We found no studies focusing on the short-tailed albatross (Diomedea albatrus), cahow or Bermuda petrel (Pterodroma cahow), California brown pelican (Pelecanus occidentalis) (although aspects of pelican biology were addressed in several California studies), Aleutian Canada goose (Branta canadensis leucopareia), snowy plover (Charadrius alexandrinus), California least tern (Sterna antillarum browni), roseate tern (S. dougallii), or the various threatened or endangered salt-marsh-inhabiting species, including some of the subspecies of the clapper rail (Rallus longirostis) or the various salt marsh sparrows. If anything, the endangered marine bird species appear to have been avoided as subjects of study in the ESP. In some cases, attempts to direct studies toward endangered species would have been futile, because of the small numbers present in OCS waters (e.g., short-tailed albatross). In other cases, the failure to examine carefully the potential impacts of OCS development on endangered and threatened bird species potentially vulnerable to OCS activities was a serious oversight. The rationale for the selection of colony sites for process studies and at-sea areas for both inventory and process studies is unclear. For colony process studies, we know of no attempt at the outset to select a representative group of major colonies in each OCS region, let alone in each marine domain, for careful, long-term study. Most colonies in most regions were studied for short periods, 1-3 years, and sometimes for less than one full season. This dispersed, fragmented approach dissipated resources and led to a diversity of data, most of which are too incomplete to provide a clear picture of the degree of natural variation in the system. After the first year or two of the program in each region, it would have been appropriate to select, in cooperation with other concerned state and federal agencies (U.S. Fish and Wildlife Service, and various state fish and game departments), the one or two most important colonies in each ecological domain on which to focus long-term attention (for the duration of the OCS program). An extremely productive effort of that sort, which took place in the fall of 1984 in Alaska for studies in the Bering Sea, included not only federal and state scientists but others, as well. However, this and later efforts came almost 8 years or more too late for influencing the major ESP efforts in Alaska. We are unaware of any equivalent cooperative planning process that was used to assess progress and to develop recommendations for priorities for future seabird work in any other OCS region. The sites for pelagic inventories of marine birds and pelagic process studies were decided more by the dictates of other programs in charge of ship use than by needs for adequate coverage to discern where and why marine birds might congregate at sea. Relatively little work was devoted to studying the distribution of vulnerable species (e.g., alcids) foraging near major colonies; efforts at the Pribilof Islands (Hunt et al., 1981a), at Kodiak Island (Gould

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology et al., 1982), and in the Southern California Bight (Hunt et al., 1979) were exceptions. For other areas, MMS obtained few or no data on where birds from important colonies might be expected to concentrate their foraging. Such data are essential for developing an appropriate environmental impact statement (EIS) for the protection of critical foraging areas, and for the protection of the colonies from spilled oil. Birds are not distributed randomly but are aggregated with respect to physical structures. Two major instances of aggregation in predictable locations are in the vicinity of fronts and where currents impinge on bathymetry of high relief, causing prey to aggregate and become more available to birds. To the extent that such areas are near colonies or are along migration routes, it can be expected that large numbers of birds will be concentrated in small areas. MMS should focus on process studies within foraging range of colonies and in known major migration routes to determine where predictably large concentrations of birds will occur. Colony Inventories Inventories of seabird colonies are now available for most of the coasts of the coterminous United States. The inventories were sponsored by MMS in Alaska (Sowls et al., 1978) and California (Sowls et al., 1980) and by the Fish and Wildlife Service (FWS) in Washington and Oregon (Varoujean, 1978), the Gulf of Mexico (e.g., Keller et al., 1984), and the Atlantic coast (e.g., Erwin and Korschgen, 1979). Additional surveys have been sponsored by the U.S. Army Corps of Engineers (e.g., Buckley and McCaffrey, 1978; Chaney et al., 1978; Peters et al., 1978; Schreiber and Schreiber, 1978). Together, the atlases and reports based on those surveys are an excellent source of information on areas where breeding seabirds were concentrated and where densities of seabirds were low, at least at the times when the surveys were conducted. However, MMS was unable to furnish the panel with maps depicting distribution and abundance of breeding birds; the lack of such maps limits the ability to retrieve information on birds at risk in the event of a major OCS oil spill or for use in planning or running quantitative oil-spill risk models. Recently, however, a computer mapping and analysis system was developed by FWS and NMFS for analyzing the spatial and temporal distribution, abundance, and life history of breeding colonial seabirds on the west coast of North America from California to Alaska. The data base for this system includes information from MMS-sponsored inventories (FWS/NMFS, 1991). Although the panel considers the available knowledge of colony locations and sizes adequate for assessing the potential for damage from OCS activities (i.e., as the basis for estimating risk), the available data are inadequate as the basis for assessing change (including measurement of impacts of actual events). First, the methods for estimating population size within colonies have been crude at best. Many colonies were surveyed only from a boat or from a passing aircraft. The surveys established where large numbers of surface- and cliff-nesting species were, but gave little information about burrow-nesting, particularly nocturnally active species. Estimates of numbers in the largest “supercolonies” are particularly uncertain. Although marked plots have been established in a small number of breeding colonies in Alaska (mostly by FWS, independently of the ESP), the relationship of numbers within these plots to colony size and regional populations is poorly understood. Hourly, daily, seasonal, and annual changes in numbers within study plots and within colonies are relatively large and are difficult to control for (Wanless et al., 1982). Thus, baseline measures of sample or regional populations are imprecise. Second, most of the regional surveys have been conducted only once. Movements of birds between colonies would confound attempts to infer regional trends from

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology sample counts, but these movements have not been investigated for any species in the MMS program. Finally, counts are available only for breeding birds: the proportions of prebreeders and nonbreeders are unknown. In virtually no cases have surveys been accurate enough or counts been repeated enough to provide the means to detect any but the largest or most sustained changes in population size. The few exceptions to that generalization are long-term studies of individual species funded by agencies other than MMS (see Hunt, 1987; Nisbet, 1989). At-Sea Inventories Techniques for the at-sea survey of birds from both ships and aircraft were developed at the outset of the ESP (e.g., Briggs et al., 1981a). The newer surveys improved on many studies done elsewhere, in that they used a format that reduced (but did not eliminate) the problems of differential detectability of species and variations in visibility related to observers' height above the water and weather conditions. Aerial remote sensing of oceanographic characteristics coupled with observations of pelagic birds constituted an important method not only for describing bird distribution, but for testing hypotheses concerning the mechanisms responsible for the patterns observed (Briggs et al., 1987). The latter information is important for understanding whether the patterns observed are likely to be regular and frequent. The surveys of the distribution and abundance of marine birds at sea varied greatly from one marine region to another in spatial and seasonal coverage. No up-to-date maps showing the distribution of sampling effort by marine region and by season or maps of seabird distributions are available from MMS, although maps showing the areas covered by specific studies are available in the final reports of completed studies. Most parts of the continental shelf off Alaska received at least some coverage, particularly during the summer season. The Alaskan studies provide a good general overview of seasonal shifts in the distribution of birds and a preliminary indication of spatial patterns of abundance (Hunt et al., 1981b; Gould et al., 1982). Coverage of the coast and the waters off California has been excellent, and both seasonal and spatial patterns of distribution are known (Briggs et al., 1981a, 1983b, 1987). No MMS studies of pelagic distributions of birds off the Oregon and Washington coasts have been published, although a study of this region is to be completed in 1991. In the Gulf of Mexico, surveys of the distribution of marine birds have provided poor coverage of the area and only fair coverage of seasonal patterns of abundance. A pilot study (Fritts and Reynolds, 1981) provided aerial coverage of four small areas in August and December 1979, and a followup study covered these areas with a series of seven surveys between June 1980 and April 1981 (Fritts et al., 1983). In all, the study involved approximately 45 days of flying; about half the flights were at an altitude suitable for bird observations (91 m) and about half at a higher altitude (228 m), more suitable for spotting marine mammals. There appear to have been no MMS-sponsored surveys of marine bird distribution along the southern and central Atlantic coasts or for Georges Bank or the Gulf of Maine. However, much of this area has been surveyed by other agencies (e.g., Powers, 1983 (NMFS); Brown, 1986 (Canadian Wildlife Service)). Although sampling methods were standardized for MMS-sponsored studies on the Pacific coast and in Alaska, different methods were used in the surveys of the Gulf of Mexico and Atlantic waters. The lack of uniformity of data collection and storage prevents the construction of a nationwide data base, although some regional data bases are being maintained. To the panel 's knowledge, data from the Gulf of Mexico and Atlantic coast have not been incorporated into an MMS data base, from which they could be retrieved (e.g., for real-time response to an oil spill or for planning purposes, such as inclusion in a quantitative oil-spill risk model).

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology Examination of the available data on pelagic distributions of birds reveals that data on bird densities (transect counts) for relatively large areas consist of small numbers of counts with high among-count variances (e.g., Hunt et al., 1982). Even in the areas with the best coverage (Alaska and California), the data are useful only to provide a general picture of the patterns of spatial and seasonal abundance and hence to identify general areas in which bird populations might be at special risk. Because the samples are small and the variance-to-mean ratios (reflecting the patchiness of bird distributions) are high, the data are, for the most part, statistically inadequate to provide a reference against which to measure change. Shoreline Habitat Inventories MMS-funded surveys of inshore waters and intertidal and supertidal areas are limited in coverage and distribution. In Alaska, there have been extensive surveys of habitat availability (Arneson 1980), but only fragmentary studies of use of the habitats by birds. Parts of the Beaufort Sea coast have been well studied (e.g., Connors et al., 1981; Johnson, 1983), as have parts of Norton Sound (Drury et al., 1981), the Yukon Delta (Eldridge, 1987), the north coast of the Alaska Peninsula (Troy and Johnson, 1987), and some areas in the Gulf of Alaska where major migratory stopovers have been identified (reviewed in DeGange and Sanger, 1987). For the most part, however, use by birds of the coastal zone of Alaska remains poorly known because of extreme logistic difficulties. In the 48 states, MMS has provided partial support of beach and inshore bird surveys in the Southern California Bight and in a small portion of the Gulf of Mexico, and other agencies have sponsored beach surveys in central California. Most coastal areas remain unsurveyed. MMS has sponsored studies of the distribution of beach-cast carcasses of oiled birds in California (e.g., Stenzel et al., 1988) and on a small portion of the gulf coast of Texas, but few elsewhere. Recent studies of the proportion of oiled carcasses that actually reached the beach indicate a great deal of short-term variability in the likelihood that a carcass would reach the beach and be found (Page et al., 1982). Without extensive concurrent data on winds, currents, and carcass flotation, beached-bird surveys are of little value as measures of mortality and are of only modest value as a reference for future monitoring. Colony Process Studies Process studies conducted on colonies include measurements of phenology, diet, chick growth, and productivity; variations in these characteristics can be correlated with environmental factors, such as temperature and prey availability. Measurement of additional characteristics —such as survival, age at first breeding, recruitment, and dispersal —can then provide the basis for construction of models of population structure and of the relationship of population change to environmental factors. Because most seabirds are long-lived, total population size and colony numbers are less sensitive to short-term changes in the environment than are process measures, such as clutch size, chick growth rates, and productivity. Process studies are therefore essential, if effects of local or short-term environmental changes are to be detected. Colony process studies can provide information on the natural range of variation in system characteristics, an early warning of potential problems, and a means of detecting and predicting changes caused by environmental factors other than OCS activities. Properly executed, they can provide a historical record against which to measure change. They can also

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology be useful in increasing our understanding of the environmental correlates and natural causes of changes in reproductive and trophic biology of selected species. Colony process studies are essential as sources of data for constructing predictive models of impacts of OCS activities. Colony studies are unlikely to provide the basis for attributing specific observed changes to the effects of OCS activities, but might be useful in providing alternative explanations when changes are observed. To be of greatest value, colony process studies need to be carried out over long terms and be incorporated into monitoring programs. Within the ESP, however, most colony process studies have lasted only 2-4 years. In Alaska, colonies of various species have been studied for longer periods with support from MMS in some years and from FWS in others (e.g., Cape Lisburne, Pribilof Islands, and Prince William Sound). In California, MMS supported one 3-year study in the Southern California Bight that was continued with funds from other sources. In central California, one 19-year study at the Farallon Islands supported by non-MMS funds is going on now (Ainley and Boekelheide, 1990). In the panel's opinion, the colony process studies conducted under the ESP yielded much less valuable results than could have been obtained with better allocation of the same funds. The lack of a coherent plan and of a commitment to long-term studies at strategic locations at the outset has resulted in a large number of studies of short duration, many of which cannot be interpreted or integrated into an overall understanding of colony processes. Many of the studies were terminated 5-10 years ago; in the absence of recent followup, it is not clear whether the early results can be used to predict or evaluate impacts of OCS activities. Properly planned and executed, colony process studies could have been invaluable in identifying short-term changes and in correlating them with environmental changes. Without continuing studies, it will be difficult or impossible to identify natural causes of observed changes and hence to rule out OCS activities as causes of these changes. The ESP made a good start on thorough process studies at several colonies; the program is deficient in that it did not follow up this good start by shifting efforts at some of the sites to a continuing monitoring program designed for the duration of OCS activities in the selected areas, as is required by OCSLA as amended in 1978. At-Sea Process Studies In the context of the ESP, the primary function of at-sea process studies is to aid in understanding the observed patterns of avian distribution and abundance. Such understanding, based on elucidation of underlying processes, should help in the prediction of the degree of aggregation of at-sea populations and prediction of the locations of aggregations in specific circumstances. Process studies should integrate information across trophic levels and should ultimately be grounded in the understanding of physical and biological processes in the ocean. Most at-sea process studies of marine birds conducted in the ESP are of only limited value in serving those functions and were not designed to do so. In Alaska, most at-sea process studies investigated food habits of birds, but did not integrate the resulting information with physical data or data on prey availability. However, in the California studies, information on physical processes and chlorophyll concentrations obtained with remote sensing was combined with shipboard collections of data on the foods available and taken by birds in areas of seabird abundance to provide a picture of why particular species concentrated where they did (Briggs et al., 1987). The panel knows of no MMS-funded studies of the effects of marine processes on seabird distribution and abundance other than those in Alaska and California. In nearshore and shoreline habitats, data were gathered on the foraging of seabirds and

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology shorebirds along the Beaufort Sea coast (Connors et al., 1981) and in the coastal regions of the Gulf of Alaska (Senner, 1979). Those studies provided useful information on feeding sites, foods, and fat deposition of migrating shorebirds and yielded an assessment of the importance of the studied habitats to the successful completion of migration by the species. Studies of the use of coastal lagoons by waterfowl also provided valuable information on the use of the specific lagoons for the successful migration of the species and populations that used them. The process studies in the intertidal zone have yielded data essential for assessing the vulnerability of areas used as staging areas by important populations of shorebirds and waterfowl. Effects Studies, Modeling, and Estimation of Risk The MMS program has put only a small percentage of its seabird program effort into investigations of the potential effects of OCS activities on birds. The studies that were funded included investigations of the incidence of oil ingestion by storm petrels (Boersma, 1986) and of reactions of birds to oil and incidence of oiling (Varoujean et al., 1983), field studies of the effects of ingested oil on reproduction in alcids and petrels (Fry, 1987), studies of recovery rates of two species of birds on the Pribilof Islands (Ford et al., 1982), and a series of risk-assessment models, primarily for the Southern California Bight (e.g., Ford, 1985). Studies of the effects of the IXTOC-I oil spill on the coast of Texas showed decreased use of beaches by shorebirds and fouling of plumage and accumulation of globs of tar on the feet of gulls, terns, and shorebirds, including threatened and endangered species (A. Amos, Texas A & M Marine Station, pers. commun., 1991). On the whole, those efforts appear to have been useful. The panel considers the opportunistic use of wild species of seabirds in the field for toxicity studies to be a marked improvement over laboratory studies of waterfowl or other model species—species that might differ substantially in susceptibility to toxic effects of oil from the wild species that are at risk. However, there appears to have been a lack of followup in efforts that might be most useful. Although excellent techniques for the detection and identification of hydrocarbon compounds in tissues were developed, we know of no continuing, regular monitoring for contamination of bird tissues in any OCS region currently producing oil or gas or in areas near major transportation corridors. Several attempts were made to delineate “risky” areas on the basis of observed seabird distributions and on the basis of measures of patchiness. They included useful innovations in risk characterization, but the approach has not been followed up with applications throughout the several OCS regions. In particular, the risk modeling conducted for the Southern California Bight (Ford, 1985), although promising as a tool for identifying and measuring possible outcomes of a hypothetical oil spill, does not appear to have been followed up or generalized to other OCS regions. Thus, an opportunity to develop generalized risk models has not been exploited adequately. Likewise, there seems to be little attention to the effects of chronic contamination of low magnitude or of chronic disturbance. Gaps in the Available Data The panel sought to identify data deficiencies that were of central concern with respect to the goals of the ESP. The minimal program of studies that should have been included in the ESP from the start can be summarized as follows.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology Colony Inventories There are no known gaps in the inventory of seabird colonies. But almost none of the inventory counts provides a statistically adequate historical record against which to measure changes in population size. At too few colonies has there been the repetition of counts between years, let alone within years, needed for statistical soundness. Many surveys are now over 10 years old. The panel is aware that MMS is beginning to update colony censuses. At-Sea Inventories Large areas of the OCS have received little or no MMS coverage. We have almost no information on the birds at risk in those areas. The lack of coverage in the Gulf of Mexico is perhaps the most glaring deficiency; up-to-date survey data are also needed for Georges Bank. In some areas that have received coverage, the seasonal distribution of counts is uneven. The lack of surveys in Alaska during winter is of concern. Although such surveys are logistically difficult, winter is a time when birds might be concentrated in coastal areas or near the ice edge and thus be particularly vulnerable to spilled oil. Coverage of most areas has been insufficient to detect major at-sea migration routes and staging areas. The passage of birds on migration is a transitory phenomenon during which major segments of the population of a species can be concentrated in small, traditionally used locations for brief periods. The locations and the timing of their use might be known to local residents, but are probably not known to MMS. Shoreline Habitat Inventories The inventories of the shoreline habitat and nearshore waters throughout most OCS regions are generally insufficient to let MMS assess the potential risk of impact or the actual loss of birds if an oil spill occurs; data gaps appear particularly severe for the coasts of the Gulf of Mexico, the middle and south Atlantic seaboard, Cape Cod, and the Gulf of Maine. Data are also apparently unavailable for the northern California, Oregon, and Washington coasts. Colony Process Studies Information on age-specific mortality, fecundity, and recruitment —critical for the construction of population models—is lacking. There are no data on dispersal, and there are no continuing long-term studies of reproductive ecology and its relationship to environmental change. At-Sea Process Studies With few exceptions, there is insufficient knowledge of the major foraging areas near large colonies and of the degree to which they are geographically fixed because of topographically driven marine processes. Information is lacking on the physical and biological processes that result in predictably greater availability of prey in particular locations. More knowledge of the processes that determine the location of major concentrations of

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology migrant and wintering birds is needed. At-sea processes responsible for concentrating birds have been best studied in the Bering Sea, central California, and Georges Bank, but are not well known for other regions. Shoreline and Nearshore Process Studies For many, if not most, of the known major staging areas, the energetic consequences and the resulting impact on bird populations of the partial or complete loss of what are thought to be critical habitats for both waterfowl and shorebirds are unknown. Effects There are insufficient data on the behavior of birds at oil slicks. Too few species have been studied, and the data on those studied are inadequate for predicting the proportion of birds likely to encounter oil in the event of a spill. There is little or no information on factors that control the rates at which oil will be ingested and on whether oil is likely to be ingested primarily with food or through other activities, such as preening. There is a lack of knowledge of the generality of the hemolytic and reproductive effects found by Fry (1987) in the full range of wild birds likely to be exposed to oil. It is not now possible to assess the full number of birds killed in a spill with accuracy. Data on the present prevalence of oil contamination on and in live birds in the wild are sparse. Spills that occur should be used as study opportunities to obtain these kinds of data. Models and Estimation of Risk The means are lacking to predict the number of birds at immediate risk because of any particular oil spill, other than one in the Southern California Bight, and even the risk-assessment model for the Southern California Bight lacks validation. It is not yet possible to relate episodic events, such as mass deaths in an oil spill, to population structure and consequent changes in populations. Recommendations for Future Research The panel identified the kinds of data needed if the goals of the ESP are to be met fully. Some of the required studies, however, are not feasible, given the length of time needed for their completion, the cost in relation to the information they could yield, or logistic problems that almost certainly would preclude success. The panel therefore recommends below the studies that are most urgently needed for addressing the central goals of MMS, but tempers its recommendations with considerations of feasibility and cost-effectiveness. Colony Inventories There needs to be a single, integrated data base in which all information on location, species composition, and population size is recorded and from which data can be retrieved for

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology hydrocarbon development. Given our current understanding of processes that control benthic community structure, separation of these impacts is difficult, if not impossible. Summary of Monitoring Design Requirements The need for broad-scale survey work has passed, and adequate general descriptions of the benthic environments of the U.S. continental shelf are in hand (Rabalais and Boesch, 1987). Future efforts must be directed toward designing sampling programs that define the magnitude of a perturbation that will have no detectable biological effect. The ESP has the ultimate responsibility to design monitoring approaches that incorporate statistical analyses of monitoring data, interpretation of results, coordination of ancillary monitoring and research efforts, and the publication of conclusions that have been subjected to appropriate and independent peer review. The ESP has the responsibility to assess not only the impacts of offshore exploration and development, but also the effects of transportation or other support activities on shoreline habitats. An environmental monitoring program that will improve understanding of the impacts on benthic processes of OCS activities on the U.S. continental shelf must consider: Characterizing the spatial and temporal extent of impacts of discharges and disturbance, including the physical and chemical processes that regulate the fate and transport of contaminants from OCS activities. Measuring direct and indirect biological effects and distinguishing them from natural variation in ecosystem structure and function. Assessing the vulnerability of different shelf habitats and predicting recovery rates. Developing predictive models to assess the long-term effects of OCS activities. As our knowledge of benthic systems increases, our ability to differentiate between anthropogenic impacts and natural variation should also increase. Given the present state of knowledge, however, our ability to predict impacts is minimal. Carney (1987) recommended that future studies focus on four major questions to resolve the differences between anthropogenic impacts and natural variation in benthic systems: Can impacts be related to processes that can be studied effectively in the field? If a faunal census approach is used, is something other than a species-by-species approach more informative and cost-effective? Can we substantially increase our understanding of the relationship between faunal and environmental spatial and temporal variation? What types of powerful and robust statistical models might be most applicable to long-term effects studies? Until those questions are answered, studies might be limited to a faunal census approach, but at the very least such studies should follow good statistical design, should be aimed at hypothesis-testing, and should use both multivariate and univariate techniques. Recent efforts of the Intergovernmental Oceanographic Commission (IOC)/Group of Experts on the Effects of Pollutants (GEEP) Workshop on Biological Effects of Pollutants held in Oslo in 1986 demonstrated the robustness of multivariate and univariate techniques as applied to higher

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology taxonomic units (i.e., family and higher) in resolving site-specific differences. The results suggest that more cost-effective approaches might be feasible in the analysis of anthropogenic impacts on macrofaunal communities and perhaps also on meiofaunal and microbial components (Bayne et al., 1988). Preliminary surveys or reconnaissance can provide substantial information on a specific study site and provide the basis for the design of further long-term studies. Such a survey should include an analysis of the temporal and spatial patterns of faunal and environmental variables, including broad-scale characterizations of the biological, chemical, physical, and geological features of the region. Later studies should focus on defining the spatial and temporal distributions of impacts by time-series analysis of biological and other environmental features. That approach is now used in the assessment of the Santa Maria Basin and should serve as the model for future ESP benthic studies. In the design of a sampling program for assessing the effects of offshore oil and gas operational discharges on benthic ecosystems, some effort should be used to determine the persistence and degradation rates of contaminants in sediments and the flux of these contaminants between sediments, interstitial waters, and biota. If the flux is substantial, it is important to relate contaminant content in sediments, interstitial waters, and biota to changes in benthic biomass, the structure of benthic communities, recruitment success of benthic populations, and trophic interactions. Finally, demersal fish and shellfish populations that migrate little or not at all should be used as models to relate contaminant concentrations in biota to the activity of biotransformation mechanisms, reproductive condition, and incidence of energetic abnormalities and pathology. As a means of addressing subtle chronic impacts, such a study design must integrate an understanding of the physics of bottom boundary layers, the chemistry of bioavailability of contaminants, and the biology of population structure and function, resource use, and recovery. FISHERIES MMS has sponsored many studies on fishery resources and some studies on the structure and function of ecosystems. However, the studies have provided relatively few quantitative assessments of the effects of OCS oil and gas activities on fishery resources, and assessments of the effects on ecosystems are almost completely lacking. The reason is not a lack of concern, but rather the complexity of the assessment problem. As noted in Chapter 2, efforts to understand the causes of population fluctuations in fisheries have met with limited success in all programs. Information Needed The most basic information that is needed is the spatial and temporal distribution of fishery resources and key ecosystem components. Life-history patterns should be understood well enough to permit one to identify critical habitat whose degradation could jeopardize a population. Spawning grounds, areas of concentration of eggs and larvae, nursery grounds, and migratory routes are particularly important. Characterizing the distribution of fishery resources and other ecosystem components is needed for assessing the relative importance of habitat. In some cases, critical habitat is remote from drilling, but can be vulnerable to onshore OCS-related activity, e.g., pipeline routings

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology through coastal wetlands, gravel islands and causeways in arctic regions, and channelization. One MMS study (Turner and Cahoon, 1988) examined the reasons for the loss of wetlands and the increasing rate of wetland loss in Louisiana. The study produced estimates of direct and indirect impacts on wetlands, quantified the extent of those impacts that resulted from OCS activity, and addressed some key questions regarding the mechanisms of those impacts. Critical coastal habitat can also be vulnerable to oil spills or blowouts. Quantitative assessment of the effects of OCS activity requires information on the fates of contaminant, i.e., spatial and temporal distribution of contaminants and their degradation products. Overlap in the distributions of contaminants and ecosystem components determines the magnitude of ecosystem exposure. Research on the biological effects of exposure (specified by concentration and duration) is an essential element for analysis. Priority for studies of biological effects should be given to species that are potentially vulnerable (e.g., exposed to contaminants) and thought to be sensitive. Early life-history stages, such as eggs and larvae, are most likely to suffer lethal effects. The most important sublethal effects are the ones that can be related to demographic measures (e.g., reproductive output and growth rate) that govern the production and abundance of populations. Research might be necessary on the degree to which mobile organisms avoid and are likely to come into contact with oil and other material originating from OCS activity. For fishery resources, traditional models of fishery population dynamics can be expanded to account for other sources of stress, such as those resulting from OCS activity (e.g., Schaaf et al., 1987). That approach takes advantage of the large literature on fish population dynamics and response to fishing. Predicting the impacts of offshore off and gas activities requires an understanding of the responses of marine biota to both chronic small discharges of drilling fluids, formation waters, and other contaminants from platforms and accidental discharges of larger volumes from blowouts or oil spills. Biological effects of contaminant discharges on marine biota depend on: Bioavailability and persistence. The ability of an organism to accumulate and metabolize contaminants. The interference of contaminants with normal metabolic processes so as to alter an organism's chances for survival and reproduction in the environment. In considering the long-term effects of offshore oil and gas development activities, one should ascertain what biological effects might result in subtle ecological changes and impairment of fishery resources (Capuzzo, 1987). Rosenthal and Alderice (1976) concluded that the stages in the life cycle of marine fish (and most invertebrates) that are most sensitive to contaminant exposure are the development of gonadal tissue, early embryonic (pregastrulation) stages, the larval transition to exogenous food sources, and metamorphosis. Several investigators have shown that early embryonic and larval stages are more sensitive than later larval stages to petroleum hydrocarbons (see review in NRC, 1985; Capuzzo, 1987). Sublethal effects of petroleum hydrocarbons on early life-history stages include a wide range of developmental and energetic abnormalities (Kuhnhold, 1974; Linden, 1976; Smith and Cameron, 1979; Linden et al., 1980; Hawkes and Stehr, 1982; Capuzzo et al., 1984). Cod and pollack eggs collected after the Argo Merchant oil spill had a high incidence of cytological deterioration and abnormal differentiation (Longwell, 1977). Effects of petroleum hydrocarbons on reproductive and developmental processes include interference with hormone synthesis (Truscott et al., 1983), impairment of gonad development (McCain et al., 1978), transfer of hydrocarbons from gonads to early developmental stages (Koster and Van den Beggelaar, 1980),

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology and reduced hatching success (Kuhnhold, 1978; Sharp et al., 1979). Sharp et al. (1979) suggested that hydrocarbon exposure of embryonic and larval stages might result in the shunting of energy reserves away from critical differentiation and morphogenic processes to metabolic maintenance. Sublethal changes in energetics of adult organisms as a result of petroleum exposure can increase susceptibility to other environmental stresses, such as disease. Several studies have shown a direct correlation between hydrocarbon stress and increased incidence of histopathological conditions (McCain et al., 1978; Sindermann, 1982; Murchelano and Wolke, 1985). Populations of plaice (Pleuronectes platessa) collected from off-contaminated estuaries after a spill from the Amoco Cadiz had reduced growth rates and fecundity (Conan, 1982) and histopathological aberrations (Haensly et al., 1982; Stott et al., 1983). As illustrated in Figure 3-1, impairment of energetic and developmental processes occurs at much lower than acutely toxic concentrations. Polycyclic aromatic hydrocarbons and other lipid-soluble foreign compounds are metabolized by many marine vertebrates and invertebrates (Bend and James, 1978; Stegeman, 1981). Metabolism affects the disposition or increases the removal of the compounds; it can also transform them to potentially more toxic derivatives. Malins (1982) suggested that metabolic transformation could be linked to a wide range of cytological, structural, and developmental abnormalities. FIGURE 3-1 Comparison of lethal and sublethal effects of petroleum hydrocarbons on fish and invertebrates. Source: Vandermeullen and Capuzzo, 1983.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology Long-term effects of offshore oil and gas development can be both direct (through the loss of reproductive or recruitment success) and indirect (through the disruption in food-chain dynamics). Both types of impacts are difficult to assess, because of our lack of knowledge of natural variability in reproductive effort and recruitment and of flexibility in prey selection by commercially important species of fish and shellfish. Although MMS has used oil-spill models to predict the spatial and temporal interactions of spills with eggs and larvae, little attention has been given to predicting the impact of chronic exposure on reproductive or developmental success. MMS, through the NOAA/OCSEAP, supported numerous studies relevant to the fishery resources of the Alaska OCS, including toxicity studies related to early life-history stages of Alaska fish and invertebrate species (Rice et al, 1983); studies of sublethal effects of petroleum hydrocarbons on salmonids, king crab, and other species (Rice et al., 1983); studies of effects of hydrocarbons on cytological, cytogenetic, and histopathological disorders (e.g., McCain et al., 1978; Hawkes, 1980; Malins et al., 1982; Varanasi and Gmur, 1981); and studies of habitat and distribution of commercially important species (Rice et al., 1983). In addition, fishery models have been developed to assess the impact of OCS activities; these are discussed in the modeling section. Those studies have provided a fundamental understanding of the broad range of responses of fishery species to petroleum hydrocarbons. Techniques developed in them are suitable for application in monitoring the effects of OCS activities, especially the short-term effects of produced waters and the long-term effects of accidental petroleum discharges. Similar studies were conducted during the evaluation of the effects of the Exxon Valdez oil spill, but the results were not available to the panel for review. A recent workshop on fisheries oceanography in the Arctic (Meyer and Johnson, 1990) identified specific research needed for assessing fishery resources at risk from oil and gas activities in the Arctic, primarily in the Beaufort and Chukchi seas. Characteristics related to the distribution, habitat requirements, and reproductive biology of important species were identified as information gaps, especially during the winter months when habitats are covered by ice. ECOSYSTEM MODELING Mathematical models should play an important role in assessing the effects of OCS activity, although they must be used with caution. Models of the fates of contaminants—which require focused studies of physical and geochemical processes—should be developed to assess ecosystem effects of OCS activity. Models have been used to study the fates of contaminants from acute events, such as blowouts and oil spills. Biological Models Developed with MMS Funding Circulation Models Although this section is concerned primarily with fishery and ecosystem models, a brief discussion of circulation models is warranted, because the physical environment is the primary determinant of contaminant distribution. MMS has funded the development of several circulation models that range from regional-circulation models (e.g., of the West Florida Shelf) up to basin-scale models (e.g., the Gulf of Mexico). The models have been reviewed by the

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology Physical Oceanography Panel, whose report (NRC, 1990a) should be consulted for details and evaluations of the models. The current distributions obtained with the circulation models are used in two ways. First, the simulated current distributions are used in conjunction with the Oil Spill Risk Assessment model to determine statistically the direction of transport of oil discharged into the marine environment and the probability that spilled oil will reach specified “targets” (usually segments of shoreline). The trajectories produced by the model can be used to assess the risk to a particular population (e.g., sea otters or birds), if the distribution and population dynamics of the population are known. Second, the simulated current distributions can be used directly in a coupled physical-biological model as variable coefficients to provide the physical forcing that transports constituents (e.g., biological constituents or contaminants) throughout a given region. Given the importance of currents in determining the distribution of constituents in the marine environment, the assumptions and procedures used to develop the circulation model must be carefully considered before the output of one of these models is used to assess biological questions. In particular, circulation models typically are formulated to represent space and time scales that are larger than the scales that are important in biological oceanographic and fishery problems. Population Risk Models MMS has supported attempts to construct models to investigate the effect of oil spills on organisms, such as marine birds and mammals. Reed et al. (1986) constructed a model to consider the effects of oil spills in the Bering Sea on the population dynamics of northern fur seals in different seasons. The available data were derived from observations of fur seal migration routes and various population measures, such as reproductive and mortality rates. The primary benefit of the study was in pointing to the need for additional data, if such modeling efforts were to be predictive for natural populations. The study of Siniff and Ralls (1988), also described earlier in this chapter, helped to predict the effects on the sea otter population of an oil spill along the California coast. Although the study was valuable in helping to predict recovery times for a population of sea otters after oil spills of different magnitudes and in different regions of the California coast, and it also demonstrated the need for additional information data on various population measures. Ford et al. (1982) used a modeling approach to consider the sensitivity of colony-breeding marine birds to oil spills within the birds' foraging area. Their model consisted of submodels for bird population demographics and foraging behavior. Sensitivity analyses of the population characteristics included in the submodels revealed that the model results were sensitive to values chosen for some of them. The study concluded that the realism of the modeling analysis was hampered by the lack of field observations that would make it possible to specify several of the critical model elements, a conclusion applicable to many modeling studies. Ecosystem and Oil-Flux Models As a result of MMS and MMS-related activities, mathematical models have been developed to investigate the biological impact of oil spills and to synthesize and integrate multidisciplinary data. One example is the BIOS (Biological Impact of an Oil Spill) ecosystem simulation model, developed by OCSEAP as part of its eastern Bering Sea environmental

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology impact study. It is a multicomponent model in which most parts of the marine ecosystem and the effects of pollution on organisms, such as fish, are represented in some fashion. The model was used primarily to estimate contaminant concentrations in some fish species and some benthic organisms that result from exposure to oil-contaminated water and sediments and the consumption of oil-contaminated food. Decontamination processes are also included in the calculation of oil-contaminant concentrations in a particular organism. In general, water-column processes (e.g., phytoplankton and zooplankton influences) are included only to the extent they provide food sources for the benthos for the transfer of oil to the sea floor. The model allows for spatial variability due to fish migration and provides an estimate of the amount of contaminated biomass that leaves the model region. The BIOS model and simulations performed with it are described in several technical reports (e.g., Gallagher, 1984; Gallagher and Pola, 1984; Swan, 1984). Although not funded by MMS, a complex ecosystem model was constructed for the Buccaneer Field study off the Louisiana coast (Fucik and Show, 1981). The model was supported by EPA and NOAA and was developed to aid in evaluating the effects of offshore oil and gas production on the marine environment. Like the BIOS model, the Buccaneer Field model is a multicomponent model. It consists of a system of coupled ordinary differential equations that describe the time-dependent behavior of several biological components and several chemical constituents. Each biological or chemical component is divided into several processes, each of which can be affected by processes in other compartments. The time-dependent model itself is coupled to a hydrodynamic model to obtain spatial distributions. The biological portion of the Buccaneer Field model consists of 10 components that describe the trophic dynamic relationships in the region of interest. The hydrocarbon portion comprises four chemical constituents—aliphatics, aromatics, branched and cyclic alkanes, and alkyl-substituted aromatics—that are divided on the basis of molecular weight and are acted on by six weathering processes. The Buccaneer Field model was used to investigate seasonal biomass trends and to describe the flow and storage of carbon between the various biological components. The chemical portion was used to estimate the percentage of total hydrocarbons lost because of weathering. Finally, the combined model was used to obtain average steady-state hydrocarbon concentrations in each biological compartment. Model results implied that the major flow of material in the Buccaneer Field system was through phytoplankton. The short residence time of phytoplankton in the Buccaneer Field region (due to advection), however, indicates that there is only a small probability that contaminated phytoplankton are being preyed on by other organisms in the Buccaneer Field. Nevertheless, contaminated particles might be transported to other locations and incorporated into the food webs there. Fishery Models MMS has sponsored the development of two fishery models to assess the effects of OCS activity quantitatively: the Georges Bank model (developed by Spaulding, et al., 1982) and the Bering Sea model (developed by Laevastu et al., 1985). MMS recently commissioned critical reviews of those models by Fletcher (1989) and Deriso (1989). The panel in general agrees with those reviews, whose major points are included in this section. The Georges Bank model was designed to assess the effects of an oil spill on fishery resources of Georges Bank. It considers only the direct effect of oil on pelagic early life-history stages. It is composed of four submodels: a hydrodynamic model, an oil-spill fate model, an

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology ichthyoplankton transport and fate model, and a fish-population model. The first two submodels have been reviewed earlier (NRC, 1990a). The Georges Bank model is applied to four traditionally important fish species: cod, haddock, sea herring, and yellowtail flounder. The ichthyoplankton transport and fate model treats pelagic eggs and larvae as passive drifters, subject to local horizontal currents averaged over the top 10 m of the water column. The currents are derived from the hydrodynamic model. The vertical distribution of ichthyoplankton, however, is known to be variable and frequently is often related to such physical features as thermoclines (e.g., Buckley and Lough, 1987). Ichthyoplankton might not be passive with respect to vertically averaged current fields; some species are known to have persistent concentrations in areas where average horizontal currents are strong (e.g., Georges Bank herring, Iles and Sinclair, 1982). The spatial and temporal distribution of the pelagic ichthyoplankton of Georges Bank has been sampled extensively by NMFS. Where the data are available, one can use empirical distributions of ichthyoplankton in oil-spill assessment models, instead of modeling ichthyoplankton transport. Where data are lacking, data on distribution should be collected to verify ichthyoplankton transport models. Where spawning has a wide temporal and spatial distribution (e.g., Georges Bank cod), verification will be difficult. The Georges Bank fish population submodel assumes a compensatory spawner-recruit function, although no empirical data support this assumption for cod. If the assumption is invalid, the impact of an oil spill will be underestimated. Because the commercially exploited fishery resources of Georges Bank have been depleted by overfishing (NMFS, 1988b), they might not be able to compensate for the additional stress of an oil spill. The maximal effect of an oil spill should have been assessed without the assumption of spawner-recruit function. The Bering Sea model is more ambitious (not necessarily more useful) than the Georges Bank model. It considers the effect of oil that settles to the bottom on demersal eggs and larvae of Pacific cod and red king crab, in addition to the effects of surface off on pelagic eggs and larvae of yellowfin sole and walleye pollack. It also models the tainting of juvenile and adult king crab, sockeye salmon, and 16 other fish species. The Bering Sea model is poorly documented, so it is difficult to evaluate. Many parameter values are drawn from unpublished sources that are not readily accessible. The initial conditions for juvenile and adult fish biomass are taken from an ecosystem model, “Dynumes” (Laevastu et al., 1985), which is also poorly documented and unverified. The Bering Sea model shares the two shortcomings of the Georges Bank model. It assumes passive drift of eggs and larvae according to vertically averaged horizontal currents; empirical data on the spatial and temporal distribution of eggs and larvae in this region are inadequate for use in impact assessment or model verification. For most simulations performed with the Bering Sea model, recruitment is assumed to be independent of spawners (having no functional relationship with them); this is equivalent to assuming strong compensation, which can result in an underestimate of the impact of an oil spill. A noncompensatory spawner-recruit relationship is used in a sensitivity analysis, but it is not used as the basis of the primary conclusions from the Bering Sea model. Drilling Discharges The Offshore Operators Committee (OOC) and Exxon Production Research Company developed a computer model to simulate the short-term fate of drilling mud discharged into most areas under a variety of environmental and discharge conditions. The model has its origins in the U.S. Army Corps of Engineers-Environmental Protection Agency Dredge Spoil

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology Model. It is used to predict the average concentrations of solids and soluble components in the discharged material in the region surrounding the discharge site and the concentration of the solids initially deposited on the sea floor. The OOC mud-discharge model was distributed to OOC member companies and federal and state agencies concerned with offshore-drilling discharge regulation in September 1983 (O 'Reilly et al., 1989). The model has been used extensively in predicting the impact of discharged material on the marine environment. The mud-discharge model assumes that drilling-mud discharges originate as a jet from a submerged pipe at an arbitrary orientation. The behavior of the discharged material after release is assumed to be characteristic of one of three phases: convective descent of the jet of material; dynamic collapse, which occurs when the descending plume either hits the bottom or arrives at a level of neutral buoyancy, when descent is retarded and horizontal spreading dominates; and long-term passive diffusion, which begins when transport and spreading of the plume is determined more by ambient currents and turbulence than by any dynamic character of its own (Brandsma et al., 1980). Each phase can be represented mathematically, given an appropriate set of assumptions. The equations that describe the first phase of the discharge, a negatively buoyant plume, are for conservation of mass, momentum, buoyancy, and solids. The conservation equations are in terms of distance along the plume. The mud-discharge model allows for up to 12 discrete classes of particles and a fluid fraction. Each particle class is described by the particle concentration, particle density, and particle settling velocity. The jet itself is described by its radius, its velocity along the jet axis, and its density. The second phase, dynamic collapse, is governed by the same conservation equations, but with the addition of an equation to describe the collapse of the elliptical cross section that is assumed to characterize the plume. In the final phase, the plume has become dynamically passive and is acted on by turbulent diffusion, advection, and settling of the solid particles. The long-term diffusion of the plume in this phase is handled with a LaGrangian scheme of diffusion. Detailed descriptions of the mud-discharge model are given by Brandsma et al. (1980) and Brandsma and Sauer (1983). The mud-discharge model was verified and calibrated with field studies of discharges into the environment. The most extensive was conducted from a production platform off California in January 1984. In general, the observed and predicted distributions of the particles originating from the discharge agreed. The field study and comparisons between the model and observed distributions are described by O'Reilly et al. (1989). The OOC mud-discharge model appears to be adequate for determining the average concentration of solids and particles immediately after discharge into the environment. It is not sufficient for determining the longer-term distribution of discharged material; it was not designed to provide more than short-term near-field distributions. Evaluation of MMS Modeling Efforts Aside from a major expenditure of resources to develop a fishery model, MMS has made few attempts to use models to consider oil and hydrocarbon effects on marine organisms. In general, the models that have been developed for the latter purpose consist of systems of coupled ordinary differential equations that consider the time-dependent effects of oil contaminants on various organisms. The parameterization of biological processes is based on empirical formulations; in some cases, the “educated guess” approach is used for unknown quantities that are difficult or impossible to estimate. That is not uncommon in biological models; however, the MMS models, such as the BIOS model, have a preponderance of

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology coefficients that are so determined. Consequently, conclusions or predictions derived from the models have little value. In some instances, the biological models have been coupled to hydrodynamic models to obtain spatial distributions. Again, the validity of the circulation model must be considered when assessing the reliability of the model predictions. MMS has used the models it has developed in two ways: to assess acute short-term effects and to assess chronic long-term effects. Acute effects can be addressed with the oil-spill trajectory models that have been developed for various regions. Those models give the statistical probability related to where the impact of an oil spill will be greatest. That information, when used in conjunction with models of bird or seal distributions, can be useful for short-term management decisions. The effectiveness of the approach is determined by the adequacy of the oil-spill model and the availability of good biological distribution data and measurements of population growth and mortality. The models presented by Ford et al. (1982) and Reed et al. (1987) are examples. As indicated in each of the models, however, values of critical population parameters are not available. In addition, the circulation models on which the population risk models depend are of questionable accuracy (NRC, 1990a). The assessment of chronic effects entails a different modeling approach. Models that treat couplings between ecosystem components are required, but measurements of the trophic couplings need to be available first. As discussed earlier, MMS has attempted to develop this type of model. The existing models, such as BIOS and Buccaneer Field Study, are complex, cumbersome, and dependent on little known or completely unknown coefficients. The type of data needed to verify their output is lacking. This field needs improvement, especially if MMS desires to focus on long-term effects of OCS activities.

OCR for page 41
Assessment of the U.S. Outer Continental Shelf Environmental Studies Program: II. Ecology This page in the original is blank.