Carbon accumulation and distribution in a boreal mixedwood
logged chronosequence near Wabowden, Manitoba

Jennifer Plaut
Bachelor's of Science in Environmental Science
December 2001

The boreal forest, along with the mid-latitude temperate forests and equatorial tropical forests, is one of the world's dominant forest ecosystem, occupying 11% of global land surface. The boreal forest is important in the context of the global carbon cycle because of the potential for long-term carbon storage. Although the forest itself is arguably less productive than other forest types worldwide, the short frost-free period, cold temperatures, and generally high soil moisture levels suppress soil microbe and detritivore activity. The turnover time for carbon that is input into the soil or forest floor can therefore be extremely long, and these reservoirs can accumulate large stores of carbon.

Whether the boreal forest, or a region thereof, acts as a net carbon sink or source depends on the relative magnitudes of terrestrial carbon uptake versus its release to the atmosphere. Carbon flux into forest ecosystems from the atmosphere occurs via gross photosynthesis by primary producers such as trees, shrubs, herbs, and mosses, and is manifest both above and below ground. Carbon returns to the atmosphere primarily through root respiration, but also through mineralization in the soil, gaseous emission, and burning.

Historically, wildfire and pest outbreaks were the dominant disturbance regimes affecting the global boreal forest. Both types of disturbance impact the partitioning of carbon stores among ecosystem pools such as aboveground biomass, woody debris, forest floor/organic soil, and mineral soil. Specifically, insects affect rates of carbon input into the detrital stores through tree damage and mortality, while fires both drastically change the labile carbon pools and add charcoal to the pool of "inert" carbon. The recent advent of commercial timber harvesting has its own suite of effects on carbon dynamics, though these have not been a primary focus of study in the boreal forest. In Manitoba specifically, commercial harvesting is so young that few mature, harvest-originated stands exist and in general there are relatively few data pertaining to harvested areas. Except for a few areas in southern Manitoba, the boreal forest there is all primary forest and has never been harvested before. While the area of land in Manitoba that is subject to timber harvest is miniscule compared to that burned by wildfires every year, the total harvested area in the Province will only increase. Quantification of the effects of human-mediated disturbance on boreal forest carbon distribution is essential to understanding the implications of changes in land management in this region. Understanding how carbon stores are distributed among ecosystem pools after commercial harvest will become more important as more land area is subject to logging disturbance, and allows greater resolution for age-structure based carbon cycle models.

To quantify the effects of forest management on carbon storage in the boreal forest, we estimated the carbon stores in three stands near Wabowden, Manitoba, Canada that were harvested 11, 18, and 30 years ago. We also sampled a 65-year old fire-originated stand and estimated its carbon stores for comparison. We developed site-specific allometric equations for the four most abundant tree species: trembling aspen, balsam poplar, black spruce, and jack pine. The carbon stores sampled directly in this study were: 1) aboveground biomass in trees and understory herbs, 2) coarse woody debris CWD), 3) forest floor and organic soil, and 4) mineral soil to a depth of 30 cm. Belowground biomass was accounted for indirectly as roots = 2.5cm in diameter were included in the forest floor samples but the coarse and fine root biomass was not incorporated into allometric equations.

Aboveground biomass at the harvested chronosequence increased with stand age at a faster rate than other boreal chronosequences of fire-originated black spruce and jack pine stands. This is primarily because regrowth at the harvested chronosequence consisted of faster-growing hardwood species such as aspen, paper birch, and balsam poplar. The fire-originated stands and the harvested stands were both stocked with conifers before their respective disturbance events, so harvesting appears to effect a substantial shift in species composition.

The coarse woody debris pool of carbon at the harvested stands was more similar in magnitude to jack pine stands in Saskatchewan than to black spruce stands to the north in Manitoba. Because wildfires vary so much in intensity, the amount of woody debris consumed or produced varies greatly and is not a wholly age-related function. Slash left on harvested sites, however, is determined by the wood end-product as well as environmental regulations and is therefore more predictable.

Forest floor carbon and carbon in the mineral soil (soil organic carbon) both decreased slightly with stand age in the harvested stands. In terms of overall magnitude, forest floor carbons stores were greater than those in more southern jack pine stands of comparable ages, though the mineral soil carbon data are less immediately comparable because of differences in sampling technique. The forest floor decrease following harvest has been documented repeatedly in temperate forests, and the explanations range from increased decomposition rates, to decreased litter inputs, to mixing of forest floor carbon into the mineral soil pool. The latter explanation is based on changes in logging technique over time from horses to tracked vehicles such as crawlers, to wheeled vehicles such as skidders. The decrease in forest floor carbon observed in the harvested stands in Manitoba is puzzling because the oldest stand is 30 years old. They were all harvested since the popularization of clear-cutting and the introduction of wheeled skidders, so the degree of forest floor/mineral soil carbon mixing should be relatively constant among the stands.

There is also no concurrent increase in carbon in the mineral soil, suggesting that perhaps other mechanisms are responsible. One proposed alternative is increased decomposition rates due to the effect of increased solar radiation on soil temperatures. The mineral soil carbon pool has the greatest turnover time compared to the other carbon pools, therefore the simultaneous decline in forest floor and mineral soil carbon warrants further scrutiny.

The harvested chronosequence differs from fire-originated stands primarily in the storage of aboveground carbon. In general, the aboveground biomass production rate is greater for the harvested sequence. Over the course of the chronosequence, the CWD and forest floor carbon pools decrease, but the increase in aboveground biomass more than compensates for this decrease. Still, CWD and forest floor carbon pools are greater than those in fire-originated stands, and contribute to overall elevated total ecosystem carbon values. It would seem, then, that harvesting increases ecosystem carbon storage outside of the living biomass pool. However, one issue that has not yet been discussed yet is the formation of charcoal by boreal forest fires. Charcoal is an inert form of carbon that may be integrated into the mineral soil but is not released to the atmosphere. The carbon in the CWD and forest floor of the harvested stands will either be mineralized or integrated into the mineral soil, but in either case its release to the atmospheric carbon pool is only a matter of turnover time.

A major finding of this study is that even un-harvested stands in this area of Manitoba (i.e. the 1935 stand) have different carbon distributions than stands in other areas of the boreal forest. Harvested stands behave differently still, and the amount of harvested land relative to land affected solely by wildfire is only going to increase. There are two main uses for the data contained in this study. The first is to constrain existing and future models of carbon cycling on a landscape level. Once the carbon distribution pattern of a given category of land has been determined (for instance, harvested stands 10-15 years in age in the boreal forest) the areal extent of that land type can be calculated and the data can be scaled-up. Secondly, if certain site preparation or harvesting techniques are found to cause minimal soil and forest floor disturbance, land managers may be encouraged to employ those techniques in the interest of carbon sequestration.