High elevation montane and subalpine forests are responsible for the majority of terrestrial carbon uptake in the western United States. In these forests, variability in the timing and magnitude of precipitation, air temperature, and incoming radiation can influence both growing season length (GSL) and net ecosystem exchange (NEE), as well as how the linkage between the two varies along elevational and latitudinal gradients. In this study, we used eddy covariance to determine GSL and NEE at two montane sites in New Mexico and one subalpine forest site in Colorado. Since there is no agreed upon method to calculate GSL, we adopted a novel approach that aggregated the output of several different methods to produce a more robust estimate of GSL that includes uncertainty (σ).
Results/Conclusions
The subalpine site had the shortest GSL and lowest σ of all three sites (162 ± 9 days), whereas the lower montane forest had the longest GSL and greatest σ (289 ± 47 days). The upper montane site was intermediate in GSL length and σ (218 ± 22 days). At the montane sites, the interannual NEE variability was not correlated with GSL but was primarily determined by annual precipitation magnitude. For example, precipitation was extremely low in 2011, which resulted in the lower montane site acting as a carbon source. The annual subalpine NEE was significantly correlated with peak snow water equivalent (although this explained only about 50% of the variance), but was not correlated with GSL, contrary to previous studies at this site. Instead, daily NEE measurements were strongly influenced by incoming shortwave radiation, and to a lesser extent by total annual precipitation, indicating that increased cloudiness as a result of high summertime precipitation may have also served to limit carbon sequestration. These results constrain the effects of elevation and latitude on forest productivity, and suggest that the interaction between solar radiation and precipitation may be an important control on NEE from water-limited forest ecosystems.