Abstract
We used multi-season occupancy analyses to model 2 fates of northern spotted owl territories in relation to habitat amount, habitat fragmentation, and the presence of barred owls in Washington State, USA, 1989-2005. Local colonization is the probability a territory unoccupied by a spotted owl in year i would be occupied in year i + 1, and local extinction is the probability a territory that was occupied by a spotted owl in year i would be unoccupied in year i + 1. We found a negative relationship between local extinction probability and amount of late-seral forest edge. We found a negative relationship between colonization probability and the number of late-seral forest patches (higher fragmentation), and a negative relationship between colonization probability and the amount of non-habitat within 600 m of a spotted owl territory center (Akaike weight = 0.59). The presence of barred owls was positively related to extinction probability and negatively related to detection probability of spotted owls. The negative relationship between presence of barred owls and detectability of spotted owls indicated that spotted owls could be modifying their calling behavior in the presence of barred owls. The positive relationship between barred owl detections and local extinction probability suggests that because of competition with barred owls, spotted owls are being displaced. Published 2014. This article is a U.S. Government work and is in the public domain in the USA.