Redistribution and emission of forest carbon by planned burning in Eucalyptus obliqua (L. Hérit.) forest of south-eastern Australia

Highlights

• C emission from forest fires is poorly described for Australian forests.

• Loss of forest C to the atmosphere is low in planned fire.

• Low intensity planned fire emitted C is similar to redistributed C.

Abstract

We applied standard forest inventory and soil sampling techniques to estimate carbon stocks in aboveground forest components and belowground in soil and fine roots to 30 cm depth before, and immediately after, low intensity planned burning. Three areas within Eucalyptus obliqua forest that had not been burnt for more than 20 years were sampled and, based on the calculated mass of components prior to and immediately following burning, the redistribution of carbon within the forest and carbon emission from the forest to the atmosphere were estimated. Of the 165 Mg C ha⁻¹ in aboveground components, about 7.8 Mg C ha⁻¹ was redistributed through accession of tree components to the forest floor and deposition of incompletely combusted plant structures and litter on the soil surface. A further 6.7 Mg C ha⁻¹ was emitted to the atmosphere, mostly from the on-ground and near surface components of litter, coarse woody debris and understorey. Overall, these E. obliqua forests sequester around 244 Mg C ha⁻¹ to 30 cm soil depth (excluding woody roots) and planned burning transferred about 3% of this C to the atmosphere and a further 3% was transformed to charred and fragmented organic matter and deposited on the soil surface. Estimating long term C balance of burnt forests will require a much better understanding of the decay characteristics of these charred organic materials and good measures of post-fire aboveground carbon accumulation. The more intense a fire becomes the greater will be C emission from coarse woody debris and understorey, indicating the importance of quantifying these components in emission estimates for more intense planned burns and for wildfires.

Introduction

Planned burning to reduce forest fuels and the risk of high intensity uncontrolled fire is practiced by fire management agencies around the world, especially in seasonally dry forests (e.g. Agee and Skinner, 2005, McCaw, 2013). As a crucially important response to reducing the emergence of high impact mega-fires (Attiwill and Adams, 2013, Williams, 2013), agencies are interested in the impact of planned fire on other environmental values such as carbon storage capacity, water quality and biodiversity.

A warming trend and increasing occurrence of weather extremes may already be contributing to an increase in wildfires in Australia, especially in southern Eucalyptus forests. During the decade 2003–2012 over 40% or about 4 M ha of the total forest area in the state of Victoria, Australia, has been burnt through in wildfires and the control line burning associated with them. In response to what seems like an ever increasing wildfire threat, forest managers now seek to maintain an annual controlled burn area equivalent to 5% of the treatable forest area in the state of Victoria – an area of about 300,000 ha and a threefold increase on recent annual burn areas (VBRC, 2010a). Because forest fuel management is a key tool in reducing fire risk (Penman et al., 2007), it is increasingly important to understand the impact of planned fire on the forest carbon cycle and on C emission to the atmosphere.

To account for forest fire emissions, it is crucial to have an accurate estimation of all categories of fuels consumed during fire. Studies of fuel beds have centred on combustion of fine fuels such as litter and near-surface vegetation, as these contribute most to fire behaviour (Keeley, 2009). However, the contribution of other fuels such as coarse woody debris to C emission is unknown in southern Eucalyptus forests. This is a gap in knowledge of fire effects in Eucalyptus forests as international fuel consumption models, such as CONSUME 3.0 or FOFEM 6.0, clearly show the increasing consumption of woody fuels as fire intensity increases. For southern Eucalyptus forests of Australia, our understanding of the quantity of these fuel types and their consumption across a range of fire intensities is poor. Furthermore, our knowledge of forest biomass and carbon stocks is mostly centred on overstorey trees and soil (e.g. Grierson et al., 1992), with few if any comprehensive assessments that include small trees, shrubs and forest floor fine and coarse fuels, where fire impact on C stock is likely greatest.

To address this lack of detailed knowledge of forest carbon loses in southern temperate Eucalyptus forests of Australia we evaluated the effects of planned burning on forest C pools in Eucalyptus obliqua lowland forests of southern Victoria. The first objective of the study was to quantify the amount of carbon and its distribution among C pools. How planned burning alters the mass of C in each of these pools formed our second objective, and determining the degree of uncertainty in estimates of the different forest carbon pools was our third objective. Finally we aimed to use the change in mass of forest carbon pools to estimate emissions of the C-containing greenhouse gases carbon dioxide (CO₂), carbon monoxide (CO) and methane (CH₄) from planned burning in these forests.

Overall we aimed to establish baseline knowledge of the distribution of forest C in southern Eucalyptus forests and to identify ways to improve sampling methods to increase the statistical power of pre- and post-burn comparisons for important fire-impacted components. Ultimately we envisage this type of forest carbon and fire data informing landscape level models of the emissions impacts of fire over large and varied areas of forest.