作者: David L. Garshelis
DOI: 10.1046/J.1523-1739.1997.96062.X
关键词:
摘要: Large numbers of sea otters ( Enhydra lutris ) died following the 1989 Exxon Valdez oil spill. Two previous studies estimated spill-related mortality, one from difference between counts before and after spill, other recovery rate tagged carcasses that had been released at sea. I used a derivative second approach, but revised it to account for (1) moribund hauled out on shore; (2) were collected sea; (3) differences in search effort different areas spill—considerations not addressed previously. The mortality estimation procedure presented here, applied Prince William Sound, Alaska, six input parameters, two with fixed values (the number (391) taken alive captivity (84)) four assigned range beach (60-90%), proportion recovered onshore (70-95%)), offshore deaths (20- 50%), (20-60%). Empirical estimates these param- eters available plausible each was delimited using data. Randomly-selected combinations within ranges produced sound 5 750, 5-95% quantiles ∪ 600-1000) lower than indicated by studies. These do diminish tragic nature instead highlight significance both process informational basis producing catastrophic loss. Estimates loss based faulty or inadequate data may mislead investigations population jeopardize public trust scientific assessments future catastrophes.