We quantified distribution and behavior of adult spring chinook salmon (Oncorhynchus tshawytscha) related to patterns of stream temperature and physical habitat at channel-unit, reach-, and section-level spatial scales in a wilderness stream and a disturbed stream in the John Day River basin in northeastern Oregon. We investigated the effectiveness of thermal remote sensing for analyzing spatial patterns of stream temperature and assessed habitat selection by spring chinook salmon, evaluating whether thermal refugia might be responsible for the persistence of these stocks in rivers where water temperatures frequently exceed their upper tolerance levels (25ЊC) during spawning migration. By presenting stream temperature and the ecology of chinook salmon in a historical context, we could evaluate how changes in riverine habitat and thermal spatial structure, which can be caused by landuse practices, may influence distributional patterns of chinook salmon. Thermal remote sensing provided spatially continuous maps of stream temperature for reaches used by chinook salmon in the upper subbasins of the Middle Fork and North Fork John Day River. Electivity analysis and logistic regression were used to test for associations between the longitudinal distribution of salmon and cool-water areas and stream habitat characteristics. Chinook salmon were distributed nonuniformly in reaches throughout each stream. Salmon distribution and cool water temperature patterns were most strongly related at reach-level spatial scales in the warm stream, the Middle Fork (maximum likelihood ratio: P Ͻ 0.01), and most weakly related in the cold stream, the North Fork (P Ͼ 0.30). Pools were preferred by adult chinook salmon in both subbasins (Bonferroni confidence interval: P Յ 0.05); however, riffles were used proportionately more frequently in the North Fork than in the Middle Fork. Our observations of thermal refugia and their use by chinook salmon at multiple spatial scales reveal that, although heterogeneity in the longitudinal stream temperature profile may be viewed as an ecological warning sign, thermal patchiness in streams also should be recognized for its biological potential to provide habitat for species existing at the margin of their environmental tolerances.

No abstract

No abstract

No abstract

The geographical position of the island of New Guinea suggests that it may have been an early staging post in the Pleistocene settlement of Australia from the Indonesia-Indochina region. Previous data have not supported this, as archaeological sites 35,000 to 40,000 years old occur in southern Australia, whereas the earliest previously known in Papua New Guinea is 26,000 years old. We now report evidence that the north coast of Papua New Guinea was occupied at least 40,000 years ago. Sahuland, which is the greater land area of Australia and New Guinea plus their connecting continental shelf exposed as land when Pleistocene sea levels were lower than now, was occupied by humans in several widely separated areas at that time. A distinctive 'waisted axe' culture appears to have existed in New Guinea and probably in Australia in the Late Pleistocene, but antecedents are not yet known from east and southeast Asia. There is evidence for hafting of these tools at a date which is earlier than known elsewhere in the world.