Monday, 28 February 2011

Total direct mortality of seabirds from the Exxon Valdez oil spill

Ford, RG | Bonnell, ML | Varoujean, DH | Page, GW | Carter, HR | Sharp, BE | Heinemann, D | Casey, JL
PROCEEDINGS OF THE EXXON VALDEZ OIL SPILL SYMPOSIUM. pp. 684-711. American Fisheries Society Symposium. Vol. 18. 

An estimate of total mortality of seabirds from the Exxon Valdez oil spill was derived from numbers of carcasses collected on shore. The processes leading to carcass collection were modeled to account for sinking of oiled birds at sea, removal from the beach by scavengers, and incomplete search effort. To determine rates of loss of oiled birds at sea, radio-tagged carcasses were tracked by aircraft and boat. Eight groups totaling 184 carcasses were tracked in spring and summer 1990; four groups were released in Prince William Sound and four in the Gulf of Alaska off Kodiak Island. Values of median persistence time for groups released in Prince William Sound ranged from 15 to 20 d, whereas those of releases in the Gulf of Alaska ranged from 7 to 18 d. Assuming that birds drifted passively with the oil, the rate function for loss of birds at sea was applied to probable point-of-oiling and time to landfall at a particular beach. Concentrations of birds subject to oiling were located with survey data gathered in the Outer Continental Shelf Environmental Assessment Program and by the U.S. Fish and Wildlife Service. Point of oiling and path of drift to each beach were determined with the National Oceanic and Atmospheric Administration (NOAA) HAZMAT oil trajectory model and NOAA overflights. Scavenging rates were estimated from field studies in Prince William Sound and on the outer coast of the Olympic Peninsula, Washington, in which oiled carcasses were monitored daily. In each study, it was found that about 50% of carcasses were removed by scavengers within 2 d and most of the remainder were taken within 5 d. To model beach search effort and its relation to scavenging rate, it was assumed that (1) carcasses were deposited with equal frequency on any nonreflective substrate, (2) that all carcasses on nonreflective but unsearched beaches were missed, and (3) that a fraction of those on searched beaches were lost as a result of scavenging. Estimates of proportion of carcasses lost because of scavenging were made considering the interval between deposition of oiled birds and initial search and the interval between successive searches of a given beach. Based on best estimates of model parameters, 375,000 birds were directly killed by the Exxon Valdez oil spill. Uncertainty in model parameters was examined by randomly selecting from the probability distribution of possible values. Among 1,250 model runs, the 5% lower bound and 95% upper bound were 300,000 and 645,000 birds, respectively. Uncertainty in parameters tended to skew model results toward higher estimates of total mortality.

Descriptors: Article Subject Terms Aves | biotelemetry | birds | carcasses | marine birds | mortality | mortality causes | oil spills | pollution effects | water pollution effects | Article Taxonomic Terms Aves | Article Geographic Terms INE, USA, Alaska, Alaska Gulf | INE, USA, Alaska, Alaska Gulf, Prince William Sound | USA, Alaska