Aboveground carbon of community-managed Chirpine (Pinus roxburghii Sarg.) forests of Nepal based on stand types and geographic aspects

Study area

Community forests dominated by P. roxburghii in the Phalebas municipality of the Parbat district in western Nepal were selected for the study (Fig. 1). The study area lies between 28° 09′ 35″ to 28° 12′ 02″ North and 83° 39′ 07″ to 83° 42′ 37″ East over 800 m to 1,500 m above mean sea level. The study area has a sub-tropical climate with a minimum temperature of 2 °C in December–January and a maximum temperature of 37 °C in July–August and an annual precipitation of 2,500 mm, much of which falls from June until mid-September (DCC, 2017). The study area landform has moderate to precipitous slopes underlain by granite and gneiss rock types which form coarse to medium-textured, acidic and shallow soils. Accessible CFs with active forest management including weeding, cleaning, thinning, pruning, salvage cutting and enrichment planting were selected based on detailed information obtained from the constitutions and operational plans held at the District Forest Office (DFO) Parbat, and by the District Coordination Committee (DCC) Parbat (DCC, 2017; DFO, 2017). Sampling plots were established across eleven CFs of various sizes and consisting of both monospecific and mixed stands of P. roxburghii where the other species include Schima wallichii (DC.) Korth., Shorea robusta Roth, Alnus nepalensis D. Don, Diospyrous malabarica (Desr.) Kostel., Myrica esculenta Buch.-Ham. ex D. Don and Castanopsis indica (Roxb. ex Lindl.) A. DC. The CFs cover 161.63 ha and provide benefits to 6,160 people of 1,032 households (DFO, 2017). All forests of the study area have been conserved and managed by the local communities for the last four decades (since the late 1970s). Before that period, the stands were badly degraded due to overexploitation of resources.

Sampling design

A total of twenty-four sample plots were established using a stratified random sampling design (Fig. 1). Study plots were randomly selected using ArcGIS 10.3 (Environmental Systems Research Institute Inc., Redlands, CA, USA), uploaded to a global positioning system for spatial navigation and ground-truthed. The shapefile of the study area was obtained from DCC and DFO Parbat.

Pure stands of P. roxburghii were classified as ‘monospecific’ and stands of P. roxburghii associated with other species were classified as ‘mixed’. Based on stand geographical location, forests were further classified into four aspects: Northeast (NE), Southeast (SE), Southwest (SW) and Northwest (NW), resulting in three sample plots for each sub-class of stand type and aspect, giving a total of 24 plots (2 stand types x 4 aspects x 3 replicates, Fig. 1, Table S1).

This study was focused on assessing AGC pools, while belowground carbon pools (tree roots and soil organic matter (SOM)) were not sampled due to limited resources. Circular sample plots of 0.05 ha were established to sample all aboveground vegetation according to stem diameter classes. Vegetation such as shrubs, herbs and grasses with a diameter at breast height (DBH, 1.3 m) of less than one cm were destructively sampled from five 1 m² quadrats located at the center and at four opposite locations of a plot (Fig. 1). The ground layer in monospecific P. roxburghii forest was dominated by ferns and grasses with no tree seedlings regenerating. In mixed stands, regeneration of Schima-Castanopsis species was sparse with no regeneration of P. roxburghii. Samples of Schima-Castanopsis understory were taken from five nested plots, clipped, weighed and mixed together in-situ. A subsample of approximately 100 g from this mixture was taken to the laboratory and oven dried at 70 °C for 96 hours prior to re-weighing. The mass of understory vegetation was estimated on an oven dry basis. Trees with DBH above one cm but less than five cm (5 cm > DBH ≥ 1 cm; referred to as saplings) were measured from a 0.01 ha (5.64 m) subplot, while trees with DBH ≥ 5 cm were measured within a circular plot of 0.05 ha (12.61 m) (Fig. 1). All trees were identified to species level and measured for DBH (diameter tape); trees with DBH ≥ 5 cm were measured for height to the top of the canopy (Sunto-clinometer, linear tape) and for crown density (densiometer).

Deadwood (standing dead trees, stumps, coarse woody debris), and litter were not measured due to almost absence of these carbon pools in Nepal’s community-managed forests. CF users collect litter for cattle beds and deadwood for fuelwood, thus not allowing these pools to accumulate (Sunam & McCarthy, 2010) (Fig. 2). There is no specific study addressing the impact of litter and deadwood removal in our study sites. However, operational forest management plans (plans prepared by forest users in support of district forest officials) of the study area have clearly mentioned the provision of the amount of litter and deadwood to be collected.

Estimating aboveground carbon

Aboveground tree biomass (DBH ≥ 5 cm) was calculated as sum of stem, branch and foliage biomass. Stem biomass was calculated from stem volume multiplied by species wood density (Jackson, 1994). Stem volume was calculated from the stem volume equation (Eq. (1)), (Sharma & Pukkala, 1990). Branch and foliage biomasses were computed from the stem biomass using species-specific branch to stem biomass and foliage to stem biomass ratios taken from the Master plan for the forestry sector of Nepal (MFSC, 1989), (Table S2) (1) $$\text{ln }\left(\text{V}\right)=\text{a}+\text{bln\hspace{0.17em}}\left(\text{DBH}\right)+\text{cln}\left(\text{H}\right)$$ where: V, over bark stem volume (dm³); DBH, Diameter at breast height (cm); H, Total tree height (m); and a, b, c are species-specific parameters (Table S3).

Aboveground sapling biomass (5 cm > DBH ≥ 1 cm) was extracted from biomass tables for community-managed forests (Tamrakar, 2000). Total sapling biomass included leaf, branch and stem compartments. Understory biomass (UB) was calculated using destructive sampling (Eq. (2)).

Aboveground tree biomass of P. roxburghii and Schima wallichii species was converted to carbon using species specific carbon content (Negi, Manhas & Chauhan, 2003). Specifically, for P. roxburghii wood carbon fraction was 0.4632 and leaves of 0.4346; for Schima wallichii wood carbon fraction was 0.4505 and leaves was 0.4352. For other species, saplings and understory vegetation, a default carbon fraction of 0.47 was applied (IPCC, 2006).

Stastistical analysis

Student’s t-test was used to assess the significance of stand types (mono/mixed) and a one-way analysis of variance was used to assess the significance of aspects (NE, NW, SE, SW) on forest carbon stock (Zar, 2010) by using Minitab 18 (Minitab Inc. State College, State College, PA, USA). Data distribution was tested (frequency distribution) by each variable of interest for each stand type and aspect for the normality of distribution. The distribution pattern of variable of interest (DBH, Height) by stand types and aspects seemed approximately symmetric, balanced and regular. The student t-test showed that data distribution was normal (p = 0.0001, Table S4). Unless otherwise specified, 5% level of significance was used in all analyses in this study.

Community forest stand composition

Pinus roxburghii is the dominant species (80%) at the study sites, followed by Schima wallichii (19%), Myrica esculenta (0.7%) and Castanopsis spp. (0.5%) (Table 1). Stand type and aspect were significant factors affecting tree DBH and height, with trees in monospecific stands of larger DBH and height compared to mixed stands. Stands facing north had 28% greater diameter and were 19% taller compared with south facing stands (Table 1). Mixed stands were dominated by trees with DBH up to 30 cm, and the monospecific stands contained more trees of larger diameter (Fig. 3A). Both mono and mixed stands had a similar number of sapling trees (Table 1).

Community forest aboveground carbon

Monospecific stands of P. roxburghii stored significantly more carbon than mixed-species stands, with most carbon in trees with DBH > 50 cm (Fig. 3B); these larger diameter trees were a minor component in mixed-species stands. Total standing AGC of tree species of the study area varied with species. P. roxburghii species dominated both mixed and monospecific stands and contained significantly greater carbon than other species (p = 0.022, Fig. 4). In mixed stands, Schima wallichii stored 5.9 Mg C ha⁻¹, followed by M. esculenta with 0.25 Mg C ha⁻¹ and Castanopsis spp. stored a negligible amount of AGC with 0.09 Mg C ha⁻¹ (Fig. 4). Similarly, stands on north-facing slopes stored significantly greater amount of AGC (p = 0.002) than stands on south-facing slopes, following a decreasing trend from NE (124.8 Mg C ha⁻¹) > NW (100.9 Mg C ha⁻¹) > SE (75.3 Mg C ha⁻¹) > SW (57.62 Mg C ha⁻¹) (Table 2). The carbon stock of the whole forest averaged 89.65 Mg C ha⁻¹ contributed 99.3% (88.98 Mg C ha⁻¹) by trees, 0.5% (0.46 Mg C ha⁻¹) by sapling and 0.2% (0.21 Mg C ha⁻¹) by understory vegetation.

Optimizing forest management to integrate carbon values with traditional CF values

The results of this study provide an inventory basis to develop policy that optimizes trade-offs between removed products such as fuelwood and litter, and the relatively new value of standing carbon stock in Nepal’s CFs. The broader range of species and tree sizes of mixed stands provides a wider range of products for community use (Smith & Scherr, 2003; Chhatre & Agrawal, 2009) and biodiversity. This is in contrast to monospecific stands that are likely more well-suited for management as a carbon stock compared with mixed stands (Shrestha et al., 2013). The results presented here show that P. roxburghii is well-suited for management as a carbon stock in CFs; furthermore it is a fast-growing and productive pioneer species that can reach over 50 m height and up to 1 m in DBH (Jackson, 1994). However, while it can be suggested that the monospecific stands are a preferred option for carbon-oriented forest management, mixed stands contribute more to local people’s livelihoods and promote biodiversity (Smith & Scherr, 2003; Chhatre & Agrawal, 2009). This study was focused on the carbon aspect of CF and did not consider biodiversity parameters such as species richness or plant species diversity. The observed regeneration of dominant overstory species in the mixed stands contrasted with no regeneration in monospecific stands, indicating the ongoing biodiversity value of mixed stands. Moreover, a study by Maraseni & Pandey (2014) observed that soils under mixed-species forests have higher amounts of SOM than that of single species dominated forests, indicating the importance of quantifying SOM in determining carbon storage benefits of CFs. Several studies indicated a positive effect of CF management for biodiversity compared to non-CF (Bowler et al., 2012; Luintel, Bluffstone & Scheller, 2018a). Carbon focused forestry practices also contravene the social and environmental safeguard principle (UNFCCC, 2011). Only forests with diverse species can provide multiple livelihood benefits to rural communities (Arnold et al., 2011) and such forests are likely to play an important role in the resilience of local rural communities in developing countries such as Nepal.

The management of forests to increase sequestered carbon is a proven efficient and economical method for off-setting GHG emissions from other areas of the economy and constraining CO₂ accumulation in the atmosphere (Pearson et al., 2014; Gren & Aklilu, 2016). The relatively high carbon storage in the CFs studied here indicates a potential to increase carbon storage in Nepal’s CFs with enhanced management (Pandey, Cockfield & Maraseni, 2014a; Luintel, Scheller & Bluffstone, 2018b). Increase in carbon stock can also bring benefits under the REDD⁺ scheme as demonstrated in a study by Maraseni et al. (2014) where incentives of REDD⁺ payments led to a reduction in the extraction of litter, grass and fodder materials from CFs, increasing both their productivity and sustainability.