Dynamics of detrital carbon pools following harvesting of a humid eastern Canadian balsam fir boreal forest

Highlights

• Carbon pools following harvest of a humid balsam fir boreal forest were studied.

• Models of carbon pools according to stand age were built from a chronosequence.

• Aboveground dead wood gets transferred to buried wood.

• Dead wood does not significantly participate to carbon of the fine soil fraction.

Abstract

Forest management strongly influences the carbon (C) budget of boreal forests and their potential to mitigating greenhouse gas emissions. A better quantification of the net changes of carbon pools with time since harvesting is necessary to guide the development of climate-friendly forest management practices. The objective of this study was to assess the evolution of forest C pools, with a special focus on detrital biomass, in an 80-year post-harvesting chronosequence consisting of 36 very homogenous stem-only harvested plots from a humid boreal balsam fir forest of eastern Canada. Dead wood C stocks comprised of snags, stumps, downed woody debris and buried wood averaged 37 Mg C ha⁻¹ and evolved according to an upward-facing «boomerang» shape pattern throughout the chronosequence (rapid decrease in the first years followed by a constant increase until the end of the time horizon). In contrast, soil C stocks (FH and mineral) averaged 156 Mg C ha⁻¹ and remain constant through time. Stand C sequestration increased rapidly in the early stages up to age 50 when it reached about 250 Mg C ha⁻¹, and then continued to accumulate at a slower rate. The temporal trends observed in C pools suggest that C originating from aboveground dead wood (snags, stumps, downed woody debris) is either leaving the system (respired or leached) or transferred into buried wood, and does not appear to influence the C stocks of the fine fraction of the organic and mineral soil horizons. However, the ultimate fate of dead wood C is still poorly understood and further research is needed in this field. We recommend fixing the length of harvest rotation at a minimum of 50 years for this ecosystem to allow the build-up of its dead wood capital, and to promote dead wood retention on site. We also recommend including buried wood in carbon inventories as this pool represents an important share of the detrital C stock in these humid boreal forests.

Introduction

Forests and the products they generate have a strong potential to mitigate climate change (Pan et al., 2011). Forests store great amounts of carbon (C) and provide renewable sources of wood products and energy that are considered as ecological alternatives to more carbon-intensive materials (Nabuurs et al., 2007). As a result, several initiatives are taken by both scientists and practitioners to implement climate-friendly forest management and wood production practices (Schulze et al., 2000, Millar et al., 2007, Malmsheimer et al., 2011).

Maximising the potential for climate change mitigation by the forest sector relies in good part on an understanding of the forest C cycle. However, many aspects of this cycle remain poorly understood, especially the decay dynamics of downed woody debris (DWD, i.e. laying dead wood) and C transfers to the soil or the atmosphere (Magnússon et al., 2016). Global vegetation (e.g. LPJ (Sitch et al., 2003), LM3V (Shevliakova et al., 2009)) or C dynamics models (e.g. Yasso (Rantakari et al., 2012), CBM-CFS3 (Kurz et al., 2009)) generally assume that a substantial portion of dead wood (i.e. snags, stumps, downed logs, large branches and dead roots) C is transferred to the soil organic matter reservoir (Cornwell et al., 2009). However, according to two syntheses (Nave et al., 2010, Thiffault et al., 2011a), the on-site retention of debris in the form of harvest residues (consisting of tree tops and branches) do not generally lead to significantly higher C content in the fine fraction of soil. Hence, it remains unclear to which extent dead wood contributes to soil stable organic matter formation and long-term soil C storage.

Since DWD constitute a large C pool (Laiho and Prescott, 2004, Russell et al., 2015), contribute to maintain soil fertility (Brais et al., 2006, Zhou et al., 2007, Angers et al., 2012) enhance ecosystem biodiversity (Harmon et al., 1986, Lassauce et al., 2011), may become more abundant (Cornwell et al., 2009, Hu et al., 2017) and have their decomposition rates affected by a changing climate (Olajuyigbe et al., 2012, Pietsch et al., 2014), characterizing their C dynamics is of crucial importance to provide adequate management guidance.

Worldwide, boreal forests (1135 Mha) are estimated to contain 32% of the total forest C stocks (Pan et al., 2011). They represent a net sink of 0.5 ± 0.1 petagrams (Pg) of C year⁻¹, which compensates for about one half of the annual anthropogenic C emissions (Pan et al., 2011). Based on the categories of the International Panel on Climate Change (IPCC), forest C pools consist of aboveground biomass (i.e. all living biomass above the soil), belowground biomass (i.e. live roots), litter (i.e. non-living biomass that is not in the dead wood pool), dead wood (above- and below-ground) and soil organic matter (Penman et al., 2003). Over two decades (1990–2008), Canadian boreal forests (270 Mha) have represented a net C sink of 11 teragrams (Tg) of C year⁻¹, but there is uncertainty about the perennity of this sink due to rising temperatures and accelerated frequency of natural disturbances.

The wood decay process is a major source of C emissions in biomes with great amounts of DWD (Wu et al., 2010), which is the case in boreal ecosystems, where DWD can account for between 10% (Stinson et al., 2011) to as much as 54% (Laiho and Prescott, 2004) of the total biomass at the forest stand level. The abundance of DWD is strongly influenced by stand age and generally follows a U-shaped pattern after disturbance, which is first explained by the degradation of the trees killed by the disturbance, followed by the gradual “recruitment”of dead trees as the new stand ages (Hély et al., 2000, Harmon et al., 2011, Russell et al., 2015).

Decomposition rates are highly variable, and in moist and cold stands presenting a thick moss layer, buried wood can be abundant and remain intact for decades or even centuries (Jacobs et al., 2015, Moroni et al., 2015). Post-decay C pathways are difficult to quantify and include one or more of the following: mineralization and emission of C towards the atmosphere, dissolution and leaching in the water system, incorporation in live organisms or stabilization in soil organic matter (Cornwell et al., 2009). The fate of C from the DWD will determine the amount of C that will ultimately be transferred to the soil. The lignin-rich carbon inherited from DWD could potentially be inferred by the concentration of lignin in FH horizons, with the presence of decomposing wood and the specific microclimate conditions in the forest floor promoting the accumulation/preservation of lignin or lignin-like compounds (Strukelj et al., 2013).

The general aim of this study was to study the dynamics of the main carbon pools, and especially those of detrital biomass and soil, along a post-harvest forest rotation occurring in a humid eastern Canadian balsam fir boreal forest. Using a chronosequence approach, we pursued the following two specific objectives: (1) to assess temporal changes in the size of all forest carbon pools as defined by the IPCC (Penman et al., 2003) following stem-only clearcut harvesting, and (2) to analyse the interactions between downed woody debris and other carbon pools to determine whether a pulse of debris will later lead to any observable increases in soil organic matter pools. We hypothesized that (1) dead wood is more abundant at the beginning and the end of the rotation, (2) the temporal pattern of the soil (FH and mineral horizons) C pool does not relate to the patterns of dead wood presence and abundance; however, the temporal pattern of the buried wood C pool follows that of dead wood and (3) the composition of the organic matter in the FH horizons does not show any chemical enrichment from decay of debris, that is, its lignin concentration will not vary significantly through time or in accordance with the dead wood accumulation curve.