Whole-forest and understory measurements of carbon dioxide fluxes at a boreal black spruce forest

1. Introduction

Boreal forests represent 22% of global forest area, and more than half of North American forests. Being of small stature, they contain only 13% of global biomass carbon, but their peat-rich soils contain nearly half (43%) of global soil carbon. Recent climate warming has been pronounced in boreal North America, and precipitation has also increased. The potential vulnerability of boreal peatlands to changing climate represents a major uncertainty in predicting future carbon balances and atmospheric concentrations of CO₂.

2. Methods

2.1 Site Description

Our research started in 1994 at the BOREAS Northern Study Area Old Black Spruce site (55.88°N, 98.48°W), in northern Manitoba, Canada. The site is situated on the low-relief terrain of the Canadian Shield, near the northern edge of the boreal forest, in the zone of discontinuous permafrost. The forest stand is approximately 130 years old, and is a mosaic of forest, bog, and fen. Dense black spruce (10 m tall) dominates uplands with stunted spruce (1-6 m) and shrubs in poorly-drained bogs. Frequent fires limit accumulation of aboveground biomass in this type of forest. Our site currently has 85 Mg C ha^-1 aboveground, but belowground carbon stocks can reach 200 Mg C ha^-1 in poorly drained bogs.

2.2 Eddy Flux and Supporting Measurements

In 1994, we began automated measurements on a 30 m instrument tower of 4 Hz winds (in three dimensions) and CO₂ and H₂O concentrations. A suite of environmental measurements, including air and soil temperatures, precipitation, and radiation, are measured at 0.5 Hz. Eddy covariance fluxes are computed to derive 30-minute mean vertical fluxes of CO₂, H₂O, momentum, and heat through a horizontal plane at 30 m.

2.3 Modified Bowen Ratio and Supporting Measurements

During 2001-03 we added instrumentation to measure soil hydrologic parameters and their relationship with carbon fluxes at two sites, a Sphagnum bog and a forested upland. Soil moisture and temperature profiles, soil heat flux, and water table depth are being measured at these sites, which differ markedly in soil carbon profiles and hydrology. In 2003 we began measurements of CO₂ fluxes at 2 m via a modified Bowen ratio method.

3. Results & Discussion

3.1 Net Ecosystem Exchange

Whole-ecosystem fluxes of CO₂ as measured on the 30 m tower reveal ecosystem sensitivity to temperature, photosynthetically active radiation (PAR), soil moisture, depth of the water table, soil thaw, and the timing of seasonal variations. Photosynthesis commenced within a few weeks of above-zero daytime temperature, before snowmelt was complete. Onset of daily net uptake varied by a full month from mid-April to mid-May, reflecting the forest's ability to begin photosynthesis as soon as temperatures allow. Maximum daily exchange rates peaked at -2 g C m^-2 d^-1 in May and June in 2003, and decreased thereafter as respiration rates increased.

Fluxes of CO₂ from the modified Bowen ratio towers reveal widely differing understory carbon exchange patterns. In the forested upland, 2 m fluxes show loss of carbon throughout the growing season, peaking at +2.5 g C m^-2 d^-1 in August. In the Sphagnum bog, which is dominated by understory species such as Salix and Ledum groenlandicum, 2 m fluxes show carbon uptake from May through September. Uptake rates at the bog peak in July at -7 g C m^-2 d^-1, a rate more than double the whole-ecosystem rate of -2 g C m^-2 d^-1 observed at the 30 m tower.

3.2 Ecosystem Respiration

The seasonal cycle and climatic controls on ecosystem respiration are particularly important due to the large reservoir of soil organic matter at the site. The seasonal increase in respiration rates lags photosynthesis by several weeks overall. The delay is due to thawing of soils at 25-45 cm (Fig. 1, bottom), which provide a large reservoir of respirable organic matter. At the 30 m tower, high rates of respiration from mid-summer onward essentially cancel the photosynthesis, leaving whole-forest net carbon fixation near zero. Soil respiration rates, as measured by the 2 m towers, are much larger in the Sphagnum bog than in the forested upland, consistent with the relative abundance of soil organic carbon at the two study areas. Both sites exhibited a seasonal increase in respiration rates, peaking during the August soil temperature maximum (Fig. 1).

Superimposed on this seasonal forcing are fast responses from surface layers exposed to varying temperatures. Strong precipitation events strongly suppress deep soil respiration by raising the water table (Fig. 1), with the effect delayed about 2 days. These results demonstrate a strong functional dependence of peat decomposition rates on temperature and soil moisture.

3.3 Discussion

Our results suggest that hydrologic factors should be as important as temperature in determining the long-term fate of deep soil organic carbon. A linear model was used to determine the factors most responsible for controlling nighttime respiration at the bog site. Respiration was most sensitive to mid-depth soil temperatures, depth to water table, and soil moisture content, accounting for ~80% of the variance of 5-day means of nighttime CO₂ flux. Additional data will clarify these relationships further, but evidence from summer 2003 clearly shows that future precipitation regimes will be as important as temperature regimes in determining the long-term boreal forest carbon balance.