ing short-term effects from high concentrations of petroleum. Chronic pollution, such as might occur from urban runoff into coastal embayments, may have continuous effects at low exposures. Not all oil pollution is clearly separable into these two categories. For example, exposure and effects are known to occur for long periods after some spills (Vandermeulen and Gordon, 1976; Sanders et al., 1980; Spies, 1987; Teal et al., 1992; Burns et al., 1993), and chronic exposures can be quite high, as is the case near petroleum seeps (Spies et al., 1980; Steurmer et al., 1982). The reader should bear this in mind during the ensuing discussion of the effects of acute and chronic exposure to oil. Additionally, this report generally focuses on the effects to benthic and wildlife populations, which were found to be most at risk from oil (Boesch et al., 1987).
It is within this complex multi-scale, spatial, and temporal environment that we are challenged to detect change caused by oil in the sea, and to assess the damage at the level of individuals, populations, communities, and ecosystems. Difficulty of detection increases with level of biological organization, with spatial and temporal scales of the affected system, and with the inherent variability of the system. Similarly, determination of complete recovery is complicated by this inherent variability.
The complex mosaic of change in the ocean has two aspects with regard to detecting the effects of oil pollution. First, it poses strategic challenges to determining the impact of oil through gathering observational data, as inevitably we make assumptions about the variability in the ecosystem and that variability can obscure large and continuing impacts. Second, the actual impact of the oil may be more complex than we realize if it interacts with spatially or temporally constrained phenomena.
In the closing decades of the twentieth century it was commonly held that the “balance of nature” has been severely altered by human actions. Consequently, much of our public policy was directed toward maintaining the status quo or returning ecosystems to a more pristine condition. While there is little doubt that human activities have had considerable impact in oceanic ecosystems, there has not been an equally widespread appreciation of how ecosystems change without human interference. The occurrence of several well-developed El Niños in the 1980s and 1990s made strong impacts on the public consciousness about longer-term cycles in the oceans. In Alaska, which has a strong resource-based economy, the rise and fall of salmon stocks in concert with the Pacific Decadal Oscillation (Beamish, 1993; Francis et al., 1998; Beamish et al., 1999) is now well known in the general population. Because public appreciation of ecosystem change seems to be following the growing scientific attention to long-term change in the oceans, the expectation that recovery of a polluted site will result in the return of an ecosystem to the state that it was in at the time of a pollution event is changing.
The observational framework for quantifying impacts involves determining differences based on sets of observations