Since the early 1970s, following the discovery that nutrient fertilizers could enhance the rates of oil biodegradation in marine environments (Atlas and Bartha, 1972; 1973; Bartha and Atlas, 1976), bioremediation has been a potential treatment for mitigating marine oil spills (Atlas, 1977; 1995; Swannell et al., 1996). As discussed by Swannell and colleagues (1996), the Exxon Valdez oil spill (EVOS) has been the most studied case regarding the applicability of bioremediation. Bioremediation based on fertilizer addition was used in Prince William Sound (PWS) from 1989 through 1991 with a total of 107000 pounds of nitrogen applied in 2237 separate shoreline applications following physical recovery of oil by shoreline washing and skimming of oil floating on water. Figure 1 shows locations and amounts of nitrogen applied 1989–1990: amounts applied in 1991 were significantly less.
The application of bioremediation to EVOS followed extensive laboratory and field testing to demonstrate safety and efficacy. For example, tests conducted for 114 days in large microcosms (30.5 cm diameter, 91.4 cm deep) filled with oiled PWS sediments and using rising and falling water to simulate tidal movements showed that 25% of the initial oil mass was lost and that significant degradation of most types of hydrocarbons occurred (Bragg et al., 1992) (Fig. 2). Control microcosms poisoned to kill microbes showed no loss. These experiments showed that the rate of biodegradation slowed down once the more easily degradable components were depleted even when fertilizer was reapplied. It is important to understand that bioremediation does not increase the ultimate extent of hydrocarbon degradation, but only the rate of biodegradation while easily degradable hydrocarbons are present (Sugiura et al., 1997; Wang et al., 1998).
Field tests jointly conducted in PWS by the United States Environmental Protection Agency (USEPA), the Alaska Department of Environmental Conservation and Exxon (Prince et al., 1993) showed, with high statistical significance, that the rate of oil degradation was a function of the ratio of nitrogen/oil, the non-polar hydrocarbon fraction remaining, and time (Bragg et al., 1994). Bioremediation increased the rate of polycyclic-aromatic-hydrocarbon (PAH) degradation in the relatively undegraded oil found on the shorelines at this time by a factor of 2, and the degradation rate of total GC-detectable hydrocarbons by a factor of 5 relative to the controls (Bragg et al., 1994). Total PAH depletion was 44% of that present at the start of the test (Fig. 3).
It is important to recognize that biodegradation does not remove all of the hydrocarbons in crude oil – some compounds are recalcitrant to microbial attack such as higher-molecular-weight PAHs and the polars (NSO – molecules containing nitrogen, sulfur and oxygen) (Oudot et al., 1998; Wang et al., 1998). Further, if the biodegradation rate cannot be accelerated significantly – by a factor of at least 2 – it has been suggested that bioremediation is not worth considering (Zhu et al., 2001). As crude oil weathers the rate of biodegradation declines since hydrocarbon biodegradation follows first-order kinetics (e.g. Venosa et al., 1996; Wrenn et al., 2006), which means that as the concentration of degradable hydrocarbons decreases so does the potential applicability of bioremediation.
Bioremediation studies of shorelines with moderately to heavily weathered oil have not been reported. However, some studies indicate that once the easily degraded alkanes and lower-molecular-weight aromatics are removed, the rate of biodegradation of the remaining oil residues is no longer limited by nutrients, and the biodegradation rate naturally slows. In bioremediation tests conducted following the Prestige spill (Diez et al., 2005; Gallego et al., 2006), for example, the rate of biodegradation of all higher-molecular-weight components slowed dramatically at a point consistent with the disappearance of the lower-molecular-weight alkanes (Fig. 4). Similarly, bioremediation tests conducted by Venosa and colleagues (1996) on a simulated spill in Delaware Bay showed a decline in the rate of biodegradation once the alkanes were removed. Wrenn and colleagues (2006), in a series of microcosm tests, also found that the sensitivity of the oil mineralization rate to nutrient input decreased rapidly as the extent of oil degradation increased, and after about 2 weeks the rate of oil mineralization appeared to be independent of nutrient input. They concluded that there may be little value in maintaining a long-term supply of nutrients in contact with oil-contaminated sediments.
Many of the microorganisms capable of attacking higher-molecular-weight PAHs may operate by co-metabolism (Cerniglia, 1993; Kanaly and Bartha, 1999; van Herwijnen et al., 2003). When alkanes and lower-molecular-weight aromatic substrates have been removed from the oil residues, the enzymes needed to attack the higher-weight PAHs may not be induced, resulting in decreased rates of biodegradation.
Other rate-decreasing factors that may result from the removal of lower-molecular-weight compounds are the increase in viscosity of the remaining oil residue and the decrease in mass fraction of degradable hydrocarbons. These slow the diffusion of degradable hydrocarbons to the oil–water interface where they can be attacked by microbes. Uraizee and colleagues (1998) systematically altered the concentration of non-biodegradable heterocyclic polars and asphaltenes in crude oil and analysed its impact on transport of biodegradable hydrocarbons to microorganisms. As the polars content (and oil viscosity) increased, the maximum respiratory oxygen uptake decreased. Obviously, the weathering point at which the degradation of PAH is no longer nutrient limited will depend on the starting oil composition and the extent of prior weathering, so each case must be evaluated independently.
In addition, successful bioremediation in marine environments requires that water containing the applied nutrients must be able to contact the oil. Oil that is sequestered (buried under layers of sediment that impede free water flow) may not benefit from applied nutrients. In this case, as well as those where oxygen is limiting, tilling can be considered, but this raises the issue of physical disruption of the environment potentially causing greater environmental harm than benefit.
Since bioremediation was previously successful in PWS and since subsurface oil residues (SSOR) remain at some locations, there have been suggestions that bioremediation be applied again (Michel et al., 2006). In this study we examine the extent of weathering that had occurred by 2007, over the 18 years since the spill, and we estimate bioremediation potentials by examining the composition of the SSOR, its distribution by location and extent of depletion, and the current natural nutrient concentrations in pore water in its vicinity.