Mountain hemlock (Tsuga mertensiana [Bong.] Carr.) growth and cool-season precipitation in Crater Lake National Park, Oregon

Abstract

This study describes the development of a new network of mountain hemlock (Tsuga mertensiana [Bong.] Carr.) records at high-elevation sites in Crater Lake National Park (CRLA), Oregon, and uses these data to make inferences about past climate in the park during the last five and a half centuries. The seven hemlock chronologies, which are constructed from 53-80 tree-ring samples at each location, are highly synchronous over time, which suggests that tree growth across the Park is controlled by similar sets of environmental factors. The climate signal preserved in tree-ring-width series was estimated by comparing them against local monthly climate data, and that analysis showed an inverse relationship between tree growth and cool-season precipitation (previous November, previous December, January, March and April) and a more modest positive relationship with growing season temperature (April-August). The inverse association between cool-season precipitation and hemlock growth is opposite to the relationship displayed by most tree-ring records in western North America. Based on the observed association between hemlock growth and climate, I identify several periods during the past five centuries where snowpack in CRLA was persistently high or low. Tree growth was combined with observed anatomical anomalies, including locally absent rings, traumatic lenses, and light latewood, to estimate past hydroclimate conditions within the park. The sequence from A.D. 1809, 1810, and 1811 stands out over the last five centuries because of its low growth and high concentration of anatomical anomalies, which are interpreted as evidence for the early onset of winter in A.D. 1809 and an exceptionally short growing season in A.D. 1810.