FIGURE 5-1 Hypothetical examples show how the impact of an oil spill and subsequent recovery can be assessed when the system under study undergoes natural variations (solid line). In (A), the system varies in time, but the long-term mean remains unchanged. In (B), there is a long-term decline in the state of the system (e.g., population size). Dashed lines indicate a “window” of normal variation about the mean (e.g., a 95 percent confidence interval). Operationally, “impact” occurs when the system is displaced outside this “window” (from Wiens, 1995, American Society for Testing and Materials).

1995; Spies et al., 1996; Peterson, 2001). Perhaps more difficult than detecting an effect is determining its significance (Boesch et al., 1987) (Figure 5-2). The spatial extent, persistence and recovery potential are all important, as is the perceived or monetary value of the affected resources. All else being equal, damage to a large area is more significant than damage to a small area of similar habitat. Damage to a small area that contains a highly valued resource can be of greater significance than damage to a much larger area devoid of valued resources. These issues are hotly contested after major pollution incidents.

DETERMINING EFFECTS IN A VARIABLE ENVIRONMENT

Oil can kill marine organisms, reduce their fitness through sublethal effects, and disrupt the structure and function of marine communities and ecosystems. While such effects have been unambiguously established in laboratory studies (Capuzzo, 1987; Moore et al., 1989) and after well-studied spills (Sanders et al., 1980; Burns et al., 1993; Peterson, 2001), determining the subtler long-term effects on populations, communities and ecosystems at low doses and in the presence of other contaminants poses significant scientific challenges. Multiple temporal and spatial variables make deciphering the effects extremely difficult, especially when considering the time and space scales at which marine populations and ecosystems change.

Marine ecosystems change naturally on a variety of time scales, ranging from hours to millennia, and on space scales ranging from meters to that of ocean basins. There are many causes of ecological change aside from oil pollution, including human disturbance, physical habitat alteration, other pollution, fishing, alteration of predation patterns, weather, and climate. Time scales at which oil affects the ocean range from days to years or even decades for some spills; chronic pollution occurs over years to decades. Oil spills affect the oceans at spatial scales of tens of square meters to thousands of square kilometers; chronic oil pollution can affect areas as small as a few square centimeters and as large as thousands of square kilometers.

Climatic changes can complicate the interpretation of contaminant impacts, especially if they have different effects on control and impact stations in an experimental design, or if a long time series of data is used to establish the “norm.” Considerable scientific attention has been directed to understanding how climatic forcing affects marine ecosystems and fisheries (Beamish, 1993; Hare and Francis, 1995; McFarlane et al., 2000). Climate change can be cyclical, e.g., the Southern Ocean Oscillation the Pacific Decadal Oscillation (Barnston and Livesy, 1999), the North Atlantic Oscillation (Trenbreth and Hurrell, 1994; Hare and Mantua, 2000), or can be secular e.g., gradual rise in upper ocean temperature.

The biological effects of oil pollution are often referred to as acute or chronic. Spills are commonly thought of as hav



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