A survey of small mammals on eight islands in the Columbia and Willamette Rivers near Portland, Oregon was conducted in 1974. The islands were Govermnent , Sand, Lemon, McGuire and Sandy Islands in the Columbia River and East, Ross and Hardtack Islands in the Willamette River. The objectives of the study were to ascertain and compare the kinds, distribution and relative densitites of small mammals . A variety of traps was used to capture the animals including Museum Special Rodent Traps , back- break mouse traps, scissor and guillotine- type stab mole traps , Sherman traps , modified Young traps and pitfalls . Most traps were set in linear transects . Trapping was conducted with the objective of sampling the major habitats on each island. Observations of scats, tracks , burrows and dead animals were also noted. One species of small marmnal , Sorex vargrans , was found on all eight islands . Microtus was found on five islands and Peromyscus maniculatus on four islands . Two islands yielded Scapanus townsendii. Glaucomvs sabrinus was found on only one island. I hoped that evidence would be found to prove or disprove that current theories of island biogeography would apply to river islands. However, data in this study were inadequate to allow calculation of precise indices of population densities . Therefore, I was unable to determine if theories of island biogeography do apply to these river islands.

Over 115 miles of rock revetment have been constructed in the Willamette Basin to protect river banks. This study examines the effect on riparian mammals of maintaining these revetments by vegetation and debris clearing. Differences between cleared and uncleared revetments were measured using a canopy coverage method of vegetation analysis; mammals were studied by trapping, by direct observation, and by noting such signs as scats, tracks, etc.

This study is an investigation of the timing of extinction of late Pleistocene, large bodied mammalian herbivores (megafauna) and of the environment in which they lived. The demise of the megafauna near the end of the Pleistocene remains unexplained. Owing to potential human involvement in the extinctions, archaeologists have been particularly concerned to understand the causes for faunal losses. Our current lack of understanding of the timing and the causes of the extinctions in North America may result from a deficiency in understanding the histories of each individual species of extinct animal on a local level. Detailed regional chronologies of fauna are necessary for comparison with paleoenvironmental and archaeological data to help sort out causes for extinction. The Willamette Valley of western Oregon has long been noted for finds of megafauna, though records have not been synthesized since the early 20th century and these materials have remained largely unstudied. In this thesis, I first create a catalog of extinct megafauna recovered from the Willamette Valley. Next, using material from the northern valley, I employ AMS radiocarbon dating, stable isotope δ¹³C and δ¹⁵N analyses, and gross inferences based on the dietary specializations and habitat preferences of taxa in order to reconstruct environments and to develop a local chronology of events that is then arrayed against archaeological and paleoenvironmental data. The results of this study indicate that megafaunal populations in the northern Willamette Valley were contemporaneous with the earliest known human populations of the Pacific Northwest, as well as later populations associated with the Clovis Paleoindian Horizon. Consistent with the overkill hypothesis, radiocarbon ages span the length of the Clovis window, but no ages are younger than Clovis. Moreover, all radiocarbon ages are older than or contemporaneous to the onset of the Younger Dryas Stadial. No age ranges fall exclusively within the Younger Dryas. Comparison of megafaunal ages and paleoenvironmental records support the view that climate change contributed to local animal population declines. Prior to ~13,000 cal BP, the Willamette Valley was an open environment; herbivores mainly consumed C₃ vegetation. The timing of the loss of megafauna coincides with increased forested conditions as indicated by regional paleoenvironmental reconstruction. As the timing of megafaunal decline correlates with Clovis, the onset of the Younger Dryas, and increased forested conditions, it is not possible with the data currently available to distinguish the cause of extinction in the Willamette Valley. The age ranges of the fauna coupled with taphonomic and geologic context indicate that the fauna are autochthonous to the Willamette Valley; they do not represent ice rafted carcasses or isolated skeletal elements transported from elsewhere during late Pleistocene glacial outburst floods.