Evaluating ex situ rates of carbon dioxide flux from northern Borneo peat swamp soils

Abstract

This study quantified CO₂ emissions from tropical peat swamp soils in Brunei Darussalam. At each site, soil was collected from areas of intact and degraded peat and CO₂ flux, and total organic content were measured ex situ. Soil organic content (~20–99%) was not significantly different between intact and degraded forest samples. CO₂ flux was higher for intact forest samples than degraded forest samples (~1.0 vs. ~0.6 μmol CO₂ m⁻² s⁻¹, respectively) but did not differ among forest locations. From our laboratory experiments, we estimated a potential emissions of ~10–20 t CO₂ ha⁻¹ y⁻¹ which is in the lower range of values reported for other tropical peat swamps. However, our results are likely affected by unmeasured variation in root respiration and the lability of resident carbon. Overall, these findings provide experimental evidence to support that clearance of tropical peat swamp forests can increase CO₂ emissions due to faster rates of decomposition.

1. Introduction

Although tropical peatlands cover only ~300,000–500,000 km² (Page et al., 2011), they are an important carbon store (~2% of global soil carbon; Sabine et al., 2004). Southeast Asia (SEA) contains a large area of tropical peatlands (248,000 km²), storing up to 68.5 Gt of soil carbon or ~77% of the tropical peat carbon pool (Page et al., 2011). However, between 1990 and 2007, 51,000 km² of peat swamp forests in Peninsular Malaysia, Sumatra, and Borneo have been deforested and drained (Miettinen & Liew, 2010). The lowering of groundwater levels and increase in soil temperatures (Sano et al., 2010) have been found to increase CO₂ emissions from soil respiration (Hooijer et al., 2012). Decomposition of these drained peatlands release 355–855 Mt of CO₂ annually (Hooijer et al., 2010), contributing to global carbon emissions.

2. Objective

Recognizing the carbon losses from drained peat suggests that land clearance incurs a high “carbon debt” or “carbon payback time” (Danielsen et al., 2009). A small number of studies report higher soil CO₂ flux from drainage-affected peat swamp forests than deforested burnt peatlands (Jauhiainen et al., 2008), as well as oil and sago palm plantation sites (Melling et al., 2005), which provided evidence to support this proposition. Our goal was to provide additional estimates of CO₂ emissions associated with decomposition from degraded and intact peatland using soils collected in Northern Borneo.

3. Methods

Soil samples (3× per site—10 cm diameter, 10 cm depth) were collected from degraded and intact peat at four sites in July 2014 (Figure 1; Jaafar et al., 2017). Degraded sites had modified drainage and fire damage. Cores were sealed individually and transported to King’s College London. CO₂ fluxes per core were measured using the LI-COR 6400-09. An airtight seal using a collar (10 cm diameter and 4.5 cm height) with 2.5 cm between the chamber bottom and peat was used. To avoid CO₂ build-up, 400 ppm and delta of 5 ppm were used; efflux was measured from the mean of five cycles (LI-COR, 2011). A subsample of peat was used for organic content. Two-way ANOVA was used to test for location and forest condition effects on CO₂ flux. Mann–Whitney was used for differences in organic content between forest conditions.

Figure 1. Location of peat sampling sites in the Belait District of Brunei Darussalam.

4. Results

Median CO₂ flux of intact peat (0.977 ± 0.167 μmol CO₂ m⁻² s⁻¹) was greater than degraded peat (0.565 ± 0.085 μmol CO₂ m⁻² s⁻¹; Figure 2). Among the four locations, Anduki had the lowest flux and Badas had the highest (~0.25 vs. ~0.8 μmol CO₂ m⁻² s⁻¹; Figure 2). Forest condition was a significant effect for CO₂ flux (p = .022), but not location (p > .05) and there was no forest condition–location interaction (p > .05). Organic content (20–99%) was not significantly different between forest conditions (p > 0.05). However, Anduki sites had lower organic content and showed a more marked difference between intact and degraded samples. This may be due to Anduki being more similar to secondary forest than the other peat swamp locations (Jaafar et al., 2017). Regardless, CO₂ flux observed between forest types could not be attributed to differences in soil organic content (Figure 3).

Figure 2. Boxplots of peat CO₂ flux by (a) forest condition (n = 12 for each condition) and (b) location (n = 6 for each location). Circles represent outliers. *p < .05; NS, not significant.

Figure 3. Scatterplot of peat CO₂ flux against soil organic content. Different symbols are used to represent the location (circle: Anduki; inverted triangle: Badas; square: Kuala Balai; diamond: Rasau) and condition (filled: intact; empty: degraded) of sampling sites.

5. Discussion

We found lower emissions for degraded site compared to drainage-affected peat, oil palm, and Sago plantations (Melling et al., 2005) and from Kalimantan forests (Jauhiainen et al., 2008). These differences are expected because only emissions from peat and no other sources were measured in our lab-based estimates. Organic content for all sites were similar despite flux differences and this suggests more labile carbon at intact sites. One feature of our lab-based protocol is root respiration is eliminated and fluxes represent only decomposition. Nonetheless, ex situ measurement will vary from in situ values (Wang et al., 2005) and future field-based studies are needed.

6. Conclusions

These results provide evidence of higher CO₂ emissions from intact than degraded forests. We caution that this should not be interpreted as evidence suggesting intact peat swamp forests are greater contributors to CO₂ emissions than degraded lands. Rather, we suggest our findings provides empirical evidence to support the proposition that clearance of undisturbed peat swamp forests will incur a higher carbon debt for biofuel production than conversion of degraded forest sites (Danielsen et al., 2009), due to higher CO₂ emissions from peat decomposition.