Forests cover almost a third of the earth's land surface, and have important feedbacks with regional and global climate through processes such as carbon exchange, evapotranspiration, and albedo. The effects of increasing temperatures on forest productivity have yet to be fully understood and will likely depend on latitude and forest type. Ecosystem responses may include competing influences from altered rates of photosynthesis, respiration, and/or decomposition, longer growing seasons, and nutrient availability effects. Water stress and drought effects may also become more important under increasing temperatures. It is extremely important to investigate not only average changes in climate, but also variability, since more frequent extreme events may occur, which can impact annual carbon sequestration.
Here we report observations from a standard open path eddy covariance carbon flux monitoring site at Haliburton Forest and Wildlife Reserve in central Ontario, which was established in 2009. Haliburton Forest is an uneven-aged mixed hardwood forest undergoing selection silviculture harvesting. Measurements are made from a 30 m tower with sensors located approximately 8 m above the forest canopy. These are complemented by observations of standard meteorological and environmental variables (air temperature, photosynthetically active radiation, soil moisture and temperature) as well as measurements of soil greenhouse gas fluxes. In the summer of 2011, CO2 and methane fluxes were also measured using a fast closed-path gas analyzer.
Of particular interest is a period of record high temperatures that coincided with leaf emergence at the beginning of the summer in 2010. Daily maximum temperatures at a weather station 25 km south of the research tower from May 25 to May 27 of that year were the highest observed on those dates in the twenty five year record, and were the hottest temperatures of the year. These extreme temperatures appeared to have deleterious effects on the emerging forest canopy. Using ongoing carbon flux measurements, as well as other supporting data, we show that this event had a lasting influence on carbon uptake through the entire growing season.
During the growing season (considered here as June 1 - September 30), over 70% of the net daily fluxes of carbon were emissions in 2010, while in 2011 that number was only 36% (see figure). Directly measured ecosystem respiration (night time fluxes, with u* > 0.2 m/s) during the growing seasons in 2010 and 2011 showed identical distributions, with means of 7.4 +/- 6.1 and 7.2 +/- 5.9 umol m-2s-1 respectively (see figure inset). This is evidence that the difference in net fluxes was driven mainly by changes in ecosystem photosynthesis, not respiration. Multi-year records of leaf area index and photosynthetic capacity indicate that not only was the canopy less full, but the leaves that were present also had diminished photosynthesis rates. On average, daily NEP in 2010 was negative (emissions) during the growing season (-131 +/- 217 mmol CO2 m-2d-1) whereas daily NEP in 2011 was on average positive (uptake) during the growing season (16.6 ± 259 mmol CO2 m-2d-1). Some growing season carbon flux data is also available for 2009 starting on August 14. Diurnal plots of NEP in 2009, 2010, and 2011 reveal similar peak carbon uptake at the same time of day for all three years, but afternoon NEP in 2010 is significantly reduced compared to 2009 and 2011, resulting in an earlier cross-over from net uptake to net emission compared to the others. Total net carbon fluxes were estimated for the summers of 2010 and 2011, although estimates were shown to be very sensitive to quality control methodology and gap-filling procedures. Overall, carbon sequestration in the summer of 2011 was 60 gC/m-2, whereas in 2010 a net carbon loss of 105 gC/m2 was calculated. Error analysis for these estimates is forthcoming.
Continued carbon flux measurements at Haliburton Forest in future years will reveal how significantly the summer of 2010 differed from normal conditions due to the extreme heat event at the beginning of the season. However, in light of the supporting data it is clear that it caused a significant change in the forest canopy and likely influenced carbon sequestration throughout the entire growing season.