We reviewed all North American radiotelemetry studies of within-home-range habitat selection by northern goshawks (Accipiter gentilis) and all studies relating territory occupancy and productivity to landscape habitat features. Goshawks selected habitats in the home range with structural characteristics of mature to old-growth forests, such as large trees and high canopy closure. We documented selection of these structures in a number of forest types, including ponderosa pine (Pinus ponderosa), mixed conifer, lodgepole pine (P. contorta), temperate rainforest, upland hardwood, and boreal forest, suggesting that goshawks are selecting forests for their structure rather than for species composition. Goshawks did not select stands with the greatest prey abundance. Selection for natural openings, edges, and stand diversity was inconclusive. Habitat selection patterns suggest that current goshawk management plans in the western United States may be inadequate.

No abstract

The mode and tempo of forest compositional change during periods of rapid climate change, including the potential for the fire regime to produce nonlinear relationships between climate and vegetation, is a long‐standing theme of forest ecological research. In the old conifer forests of the coastal Pacific Northwest, fire disturbances are sufficiently rare that their relation to climate and their ecological effects are poorly understood. We used a 14 700‐year high‐resolution sediment record from Yahoo Lake on the Olympic Peninsula, Washington, USA, to examine vegetation (landscape vegetation from pollen and local vegetation from macrofossils) and fire (landscape fire from total charcoal and local fire from charcoal peaks) in conjunction with independent records of climate. We hypothesized that the successional stage of the local forest will exhibit alternate stable states over a range of fire activity, that species turnover will increase abruptly above a certain level of fire activity and that both responses would be more gradual at the landscape scale than the local scale. Supporting these hypotheses, at the local scale, we found strong evidence for alternate stable states of late vs. early successional communities and inertia of species turnover to changing fire activity. At the landscape scale, vegetation responded more gradually to changing fire activity. From 14 700 to 7000 years ago, high landscape vegetation turnover occurred along with high landscape fire activity, especially during the warm summers of the early Holocene. In several instances, local species turned over completely following fire events but several centuries after climate change. In contrast, during the last 7000 years, the local forest composition was dominated by late‐successional species with little species turnover, despite periods of moderate fire activity. We suggest that the relatively minor climate fluctuations of the past 7000 years were not sufficient to cause large‐scale species turnover after fire. The Yahoo Lake fire and vegetation record of the early Holocene provides a model for dramatic ecosystem change following an anticipated shift to warmer summer temperatures.

No abstract

No abstract