Forest response to 1,000 years of drought variability in the Southwestern United States

Droughts in the early 1950s and early 2000s significantly accelerated tree mortality rates in the Southwestern United States. During the early 2000s, forest inventory data indicate that the proportion of dead piñon pine, ponderosa pine, and Douglas-fir trees doubled in the Southwest. The 2000s drought peaked in 2002 and was the most severe drought in at least 100 years. In 2011, precipitation, dew-point, and wind data indicate the intensity of the 2002 drought has been surpassed in a number of ways. Measurements of water potential in piñon pine trees in northern New Mexico indicate that, at present, trees have less access to soil moisture than in 2002 when widespread mortality occurred. How do these recent droughts compare to those of the last 1000 years? We used records of annual tree-ring widths from 309 populations of piñon pine, ponderosa pine, and Douglas-fir throughout the Southwestern United States to reconstruct a single record of regional drought stress from 1000-2005 AD. This record indicates that the last Southwestern drought similar in intensity to one in the early 2000s occurred in the late 1600s. Both of these droughts, however, paled in comparison to a mega-drought that occurred from 1575-1595. The emergence from this mega-drought, around 1600 AD, appears to mark a transition period from a time when droughts similar the early 2000s drought were much more common. Tree-age studies indicate a scarcity of Southwestern trees with rings extending back beyond the mega-drought of the late 1500s. This suggests that (1) the late-1500s mega-drought triggered a massive die-off of forests and/or (2) the higher frequency of drought events prior to the mega-drought sustained a much more sparse forest population than the one that has thrived from the 1600s through present. Given this apparent plasticity of Southwestern forests, a change in the forest population should be underway if higher temperatures contribute to forest drought stress. The Southwestern tree-ring record indicates that this is the case. During the 20th century, tree-ring widths correlated very positively with total winter precipitation and very negatively with spring-summer maximum temperature. This indicates that Southwestern forest growth is significantly impacted by both the amount of water delivered before the growing season and temperature during the growing season. We conclude that in the absence of a significant increase in winter precipitation, continued warming should lead to a more sparsely populated Southwestern forest population, similar to the one that appears to have existed during 1000-1600 AD.