Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Watershed Sciences


William T. Helm


A major bioenvironmental effect of The United States Atomic Energy Commission's Underground Nuclear Testing Program on Amchitka Island, Alaska was the decimation of stream communities by drilling muds from the device emplacement holes. These muds reached two of these streams 4 months and 2 years before the Milrow and Cannikin tests, respectively. Recolonization of fishes and macroinvertebrates in these small streams (Clevenger and White Alice Creeks) began 1 month after spill abatement.

Complete recovery of Dolly Varden in Clevenger Creek occurred less than 5 years after the spill. Of the other native fishes, pink salmon spawned 1 1/2 years after the spill and threespine stickleback, which were formerly rare in the stream, remained absent after 5 years. Although the midge Chironomus riparius recovered completely in 2 1/2 years, populations of three of the nine sympatric macroinvertebrates were still below the pre-spill densities after 5 years. Annual production of juvenile Dolly Varden in Clevenger Creek 3 and 4 years after the spill (42.45 and 43.58 kg ha-1, dry weight) surpassed the pre-spill estimate (40.21 kg ha-1) because of the predominance of new year classes in the recovering population, and not because of increased growth rates. Annual production of the recovering midges in Clevenger Creek 3 years after the spill was about 1 1/2 times the pre-spill estimate (8.12 and 12.01 gm m-2, dry weight), because of high densities from reduced competition by sympatric macroinvertebrates and predation by Dolly Varden. Rapid expulsion of the pollutant by high runoffs and immigration of individuals from unaffected areas of the stream sped the initial recolonization. The increased turbidity and some bank sloughing caused by the Milrow event did not hinder this recovery.

The spill that decimated the communities of White Alice Creek, persisted for 1 1/2 years and slowed the recovery of fishes and macroinvertebrates in that stream. Physical damage and alteration of stream habitat by the Cannikin device further inhibited recolonization.

Macroinvertebrate species diversity of unperturbed Amchitka streams was low when compared to that of continental streams. The response of this index to seasonal variations in density invalidated its use as a sensitivity index to perturbations in these simply-structured island communities. The stability of these low diversity communities was not thoroughly evaluated, since their resistence was not thoroughly tested; the mud spills seemed toxic enough to have decimated most stream populations. However, resilience, or the recooperative process, appeared high as indicated by rapid recolonization of native species after abatement.