Subalpine fir populations were sampled at 3 elevations on southwest-facing Slopes on Klahhane Ridge, Blue Mountain, Mount Dana, and Dodger Point: the upper (treeline), middle, and lower extent of its elevation range in the Olympic Mountains, Washington. At least 25 trees were sampled from 20 m-wide band transects at each site, and dendrochronological techniques were used in processing tree ring samples. Monthly summaries of Washington State Division 4 climatic data for 1895-1990 were used in analyses (National Climatic Data Center Data) (see metadata set entitled: "Growth response of subalpine fir (abies lasiocarpa) to climate" for detailed descriptions on climate data study sites, sampling, and processing of tree ring samples). In order to examine the effects of extreme climatic events on growth, I first identified extreme climatic events that were potentially important to tree growth during the time period 1895-1990. Correlations between climate and mean growth chronologies indicate that 5 climatic variables are correlated with growth among many sites: winter precipitation (December-March precipitation), previous August temperature, November temperature, July temperature, and August temperature. The 5 years with highest and lowest values of these climatic variables were identified.
A comparison of growth in extreme climate years was performed by calculating a simple arithmetic mean for growth in the years with the 5 highest and lowest extreme climate values for each climatic variable (i.e., mean of the 5 indexed growth values from site chronologies in the 5 extreme climate years). The growth in extreme climate years was also summarized as the proportion of the 5 extreme climate years that resulted in a growth response that was 1 standard deviation (± 1 SD) from the mean of each site chronology. The proportion of years with faster or lower growth provides a more complete picture of growth response to extreme climate than can be provided with the mean growth response; the mean of the 5 mean indexed growth values from site chronologies is sensitive to non-responding growth years (e.g., growth can be ± 1 SD from the mean in 3 out of 5 years, but the mean of all 5 years is not substantially different from the average of the site chronology ). The proportion of years with ±1 SD growth is also summarized by different "site types" (i.e., wet, dry; high or low elevation sites) to determine patterns in growth in extreme climate years across the range of subalpine fir in the Olympic Mountains.
Climate in Extreme Growth Years
To further explore growth-climate relationships, I examined the climatic conditions that were present for the fastest and slowest growth years at each site. This approach uses the tree as an integrator of the many climatic factors that affect growth, and combinations of climate that result in fast and slow growth years are assumed to provide insights on the types of future climate that will have the most significant effects on growth. I selected the 5 fastest and 5 slowest growth values from mean indexed chronologies at each site for comparison. The 5 fastest and 5 slowest growth years were subjectively selected as the limit of extreme growth, because ring-widths in less extreme years (more average growth years) are usually less than 1 SD from the mean of the site chronology. Monthly temperature and precipitation variables, along with winter precipitation, were examined in each fast or slow growth year at each site.