What have we learnt from carbon and water flux measurements following the recent mountain pine beetle epidemic in British Columbia?

The recent mountain pine beetle (Dendroctonus ponderosae) (MPB) outbreak in British Columbia (BC), which started in the late 1990's, caused the loss of approximately 710 million m³ of commercial timber in an area of about 18.1 million hectares.

Tree mortality due to MPB infestation is expected to significantly impact carbon (C) and water balances of forests. We made eddy-covariance (EC) measurements of CO₂ and water vapor fluxes over the past seven years in two unharvested stands in the BC interior, which were attacked by the beetle in 2003 (MPB-03) and 2006 (MPB-06), respectively. MPB-03 was an open-canopy lodgepole pine (Pinus contorta var. latifolia) stand with considerable secondary structure (trees and understory not killed by the beetle), whereas MPB-06 was a pure lodgepole pine stand with little secondary structure. Following the attack, these stands became annual C sources, i.e., the decrease in gross ecosystem photosynthesis (GEP) exceeded the decrease in ecosystem respiration (R_e). One year after the attack, in 2007, MPB-06 was a C source with an annual net ecosystem productivity (NEP) of -81 g C m^-2 yr^-1, resulting from a GEP of 440 g C m^-2 yr^-1 and an R_e of 521 g C m^-2 yr^-1. Four years after the attack, MPB-03 was still a C source with an NEP of -55 g C m^-2 yr^-1 (GEP = 430 g C m^-2 yr^-1 and R_e = 485 g C m^-2 yr^-1). The two stands, however, recovered faster than expected with MPB-06 and MPB-03 becoming C neutral 3 and 6 years, respectively, after attack due to increasing uptake of CO₂ by the remaining living trees and understory. Since then the two stands have been relatively weak C sinks with NEP values in 2012 (six and nine years after attack at MPB-06 and MPB-03) of 38 and 36 g C m^-2 yr^-1, respectively. Recovery was found to be much slower following clearcut harvesting with EC measurements in a clearcut in the same area indicating that it remained a growing season C source 10 years after harvesting.

Surprisingly, evapotranspiration (ET) in both unharvested stands changed little since the MPB attack with an annual average ET over the six years of 234 mm at MPB-06 and 284 mm at MPB-03. This indicates the important compensating effects of transpiration by the developing secondary structure and evaporation from the soil surface.

To study the effects of partial harvesting (the removal of all dead lodgepole pine trees), EC measurements were made between the fall of 2009 and 2013 in a mixed conifer stand that was attacked in 2005-06 and partially harvested in February and March 2009 (MPB-09). In the first three years after harvesting, this stand was an annual C source, but it has been steadily recovering and the C source weakened from an NEP of -121 g C m^-2 yr^-1 in the first year to an NEP of -52 g C m^-2 yr^-1 in the third year after harvesting. During the growing season of 2010, these measurements were supplemented with EC measurements in a nearby clearcut. This allowed the direct comparison of the effects of partial and clearcut harvesting on the C balance. MPB-09 was a small growing season C sink in 2010 (NEP = 9 g C m^-2), due to CO₂ uptake by the remaining trees and understory vegetation, whereas the clearcut remained a large C source (NEP = -103 g C m^-2). To fully investigate the effects of such management strategies on C and water balances following insect outbreaks, we are currently modifying the process-based model (3-PG), and evaluating the model using our flux measurements. In the model we will also consider changes in heterotrophic respiration due to decomposition of increasing numbers of fallen dead trees at the unharvested sites as well as wind throw at MPB-09.