On the relationship between boreal forest browning and tree mortality: insights from Alaska

Abstract:

Long-term satellite measurements of vegetation productivity in high-latitude environments have revealed profound and widespread responses to climate warming. Although warmer and longer summers are causing the Arctic to “green”, many regions of boreal forest are showing the opposite response, particularly since the mid 1990s. This “browning” phenomenon was generally unexpected at the time of discovery, is not captured by global models, and may have profound consequences for the boreal biome. A number of studies have linked satellite-based browning trends to tree productivity through tree rings. However, our understanding of the environmental controls and ecosystem consequences of browning remains remarkably limited.

Here we examine to what extent browning patterns are related to a fundamental demographic process: tree mortality. We focus on a long-term inventory database in Alaska to characterize mortality events and trends from 1994 to 2014. These patterns were related to vegetation productivity indices from MODIS and the AVHRR-based GIMMS3g data set. We explore three central hypotheses: (1) mortality events are likely to be preceded by 5-10 year browning trends (“press stress”), (2) mortality events are likely to be preceded by distinct pulses of low productivity (“pulse stress”), and (3) long-term trends in mortality are related to long-term browning. Within our study region, which encompasses eastern Alaskan from the Pacific coastal mountains up through the interior, we find strong evidence for the first two hypotheses. The third is weakly supported, which may be a consequence of the episodic nature of mortality in the region. However, preliminary analyses in the southern Canadian boreal reveal a markedly stronger relationship between long-term mortality and browning. Taken together, our study suggests a robust correlation between satellite-based metrics of productivity and forest demography; one that has consequences for forest composition, carbon stocks, and early signs of a biome shift in boreal forests.