Ecosystem Carbon Storage and Partitioning in Chato Afromontane Forest: Its Climate Change Mitigation and Economic Potential

Forests trap carbon dioxide (CO2) from the atmosphere, store in the form of carbon (C) and regulate climate change. In this study, C storage and climate change mitigation potential of Chato Afromontane forest was assessed from measurement of the major pools including the aboveground biomass, belowground biomass, dead tree biomass, plant litter and soil organic carbon (SOC). The result showed that biomass accumulation was comparatively larger for natural forest than plantations due to maturity, intactness and species diversity. The total C storage capacity of the forest ranged from 107.12 Mg ha-1 for acacia plantation to 453.21 Mg ha-1 for the intact natural forest. The mean C storage capacity by major pools ranged from 1.36 Mg ha-1 for the dead tree C to 157.95 Mg ha-1 for the aboveground C pool. The forest ecosystem accumulated a total of nearly 6371.30 Gg C in the vegetation plus soil to a depth of 60 cm. Conservation of the sacred forest will have an imperative implication to net positive C addition and regulation of climate change.


INTRODUCTION
Forests are land use systems with high tree population and store large quantities of C (Lal 2005). Forest ecosystem store more than 80% of all terrestrial aboveground C and more than 70% of all soil C (Batjes 1996). According to Batjes (1996) the pedologic and biotic pools together are called the terrestrial C pools, and they are estimated at 2860 Pg or 2860 Gt (1Gt = 1Pg = 1 billion metric tons). Terrestrial C is the C stored in terrestrial ecosystems as living or dead plant biomass (aboveground and belowground) and in the soil along with usually negligible quantities as animal biomass. The main C pools in tropical forest ecosystem are the living biomass of tree and understory vegetation, dead mass of litter and woody debris, and soil organic matter. The vegetation of tropical forests is a large and globally significant storage of C because tropical forests contain more C per unit area than any other land cover (Hairiah et al. 2011). The forest resources of Ethiopia store 2.76 billion tons of C (about 10 billion Mg of CO2) in the aboveground biomass (Yitebitu Moges et al. 2010). Forests can be both sources of atmospheric CO2 when disturbed by natural or human causes, and sinks, when vegetation and soil C accumulate after disturbance, depending on land management thus potentially accelerating or mitigating climate change (Lal 2004). The REDD + strategy, namely "reducing emissions from deforestation and forest degradation, and foster conservation, sustainable management of forests and enhancement of forest C stocks (through afforestation and regeneration) are keys to ensure net positive C addition that would then become a credit that could be sold in an international C market. However, the potential of C financing through REDD+ on forest C sequestration in tropical forests has not been systematically studied. The general allometric models developed by Pearson et al. (2005) and Chave et al. (2005) have been widely used, notably in the context of REDD+, and were recommended by the IPCC guidelines (IPCC 2006) for estimating C stocks in tropical forests. The general model developed by Chave et al. (2005) including tree height provided best biomass estimates specifically for moist tropical forests and reduce uncertainties as compared to other generic models (Ervan et al. 2013). Afromontane forests are among the most species-rich ecosystems on earth (Schmitt et al. 2010). The study was conducted in Chato Afromontane forest ecosystem, one of the largest sacred forests in Ethiopia comprising of untouched natural forest and tree plantations. Although not studied so far, the wide range of tree plantations and high endemic plant species in Chato forest makes it the most powerful C sinks in the tropics. Therefore, the study was designed mainly to estimate C storage capacity and CO2e sink of the forest ecosystem so as to unveil the climate change mitigation and economic prospective of the forest.

II.
MATERIALS AND METHODS 2.1. The study site Chato forest is situated between 9.62898256 to 9.810748292N and 36.90419252 to 37.06710714E (Fig. 1) in the western parts of Ethiopia with an elevation ranging from 1700 to 2350 m asl. It is found at about 30 km northwest of Shambu, the capital city of Horo Guduru Wollega Zone, Oromia Region. The forest was demarcated as National Forest Priority Areas (NFPA) and has been known by the name Chato-Sangi-Dangab forest in the country (EFAP 1994

Forest stratification and sampling techniques
Compartments or strata established during forest inventory by Horo Guduru Forest and Wildlife Enterprise, mainly based on forest stand type was used for biomass assessment. Besides, part of the forest that was not addressed during inventory by the Enterprise was stratified during the study. The area of each forest stand was tracked by using ground positioning system (GPS). In the stratum or forest stand, nested sample plots of 20 m x 20 m, 2 m x 2 m and 1 m x 1 m were randomly laid to measure the biomass of woody plants, herbaceous/saplings and litter biomasses, respectively. A total of 105 sample plots were taken for C stock inventory. Sample plots in the same stand, namely eucalyptus, acacia, Cupressus lusitanica, Juniperus procera, Gravellia robusta and natural forest were weighed to give average biomass and C stock for each stand type.

Estimation of biomass in different pools
The major biomass components or pools assessed include aboveground live biomass, dead tree biomass, below ground biomass, and litter biomass.

Aboveground live biomass (AGLB)
In each 20 m x 20 m sampling plot, diameter at breast height (DBH) and tree height (H) were measured for every live tree using caliper and hypsometer, respectively. Then, the aboveground biomass of live trees with DBH ≥ 5 cm was estimated by using general allometric equation recommended by Chave et al. (2005) for moist tropical forest stands as indicated hereunder: where AGTB is aboveground tree biomass (kg), ρ is wood specific gravity (g cm -3 ), D is tree DBH (cm), and H is tree height (m). Besides, live grasses, shrubs, herbs, saplings, and some tree seedlings from natural regeneration with a DBH < 5 cm (Pearson et al., 2005) were harvested in each 2 m x 2 m subplot located in every corner and center of the main plot (400 m 2 ) in the nest. In 4 m 2 subplot, total fresh weight of harvested plant material was measured, from which 500 g sample size was taken to the laboratory, ovendried at 85 0 C and reweighed to estimate the dry matter of aboveground grasses, shrubs, herbs and saplings biomass (AGHSB). Finally, aboveground live biomass was the sum of aboveground tree biomass and aboveground grasses, shrubs, herbs, and saplings biomass.

Belowground biomass (BGB)
Belowground biomass was estimated from aboveground biomass on the basis of root to shoot ratio of (0.24:1) recommended by Cairns et al. (1997) for moist tropical forests (woody and non-woody).

Dead tree biomass (DTB)
The dead tree biomass was estimated for standing and downed dead trees following (CSEMF 2011) equations.
Standing dead tree biomass (SDTB) was estimated by classifying the dead tree into three classes. The first class works for standing dead tree with small and large braches and twigs but without leaves. In this case, general allometric equation was used to estimate biomass and 2% was deducted due to absence of leaves. The second class works for standing dead tree with no twigs but only some large branches. In this case: The third class works for standing dead tree with bole (trunk) only. In this case: Dt are diameter at base of the tree and top of the stump, respectively. Downed dead tree biomass (DDTB) was determined from volume estimate as: , Db is diameter of the base of the dead wood (cm), Dt is diameter of the tip of the dead wood (cm), H is length of the dead wood (m).
For standing dead trees of case 2 and 3 and downed dead trees, sample wood density was estimated by floating method (by cutting a disk of wood) and drying until a constant mass was obtained. Hence, wood density was estimated by dividing dry weight of the disk by volume of the disk. Subsequently, standing dead tree biomass of class 2 and 3 and downed dead tree biomass was estimated by multiplying volume biomass by their wood densities. Biomass of standing dead tree under each case was summed up to give total SDTB. Finally, SDTB and DDTB were summed to provide DTB.

Litter biomass (LB)
The dry matter of litter and finer plant debris was collected from 1 m x 1 m plot in every four corners and center of the main 400 m 2 plot in the nest. In the 1 m 2 plot, litter was collected and total fresh weight was recorded, from which 250 g sample size was taken to the laboratory, oven-dried at 85 0 C and reweighed to estimate the dry matter.

Calculation of carbon stock from biomass
The amount of C stored in each pool (kg) was determined by multiplying the biomass of each pool (kg) to 0.50

Calculation of carbon storage capacity
Then, C storage capacity (Mg ha -1 ) was calculated by dividing the Cx stored in each pool and each subplot (kg) by area of the subplot (m 2 ) and multiplying with 10 as follows: where C storage capacity was estimated for each types of pools (i.e. AGB, BGB, DTB, and LB) expressed as Mg ha -1 and 10 is a conversion factor from kg m -2 to Mg ha -1 .

Estimation of soil organic carbon (SOC)
The SOC (Mg ha -1 ) to specific soil depth was estimated as: where OC is mg g -1 C concentration, d is soil thickness or depth i.e. 0-30 and 30-60 cm, ρb is bulk density of the soil (g cm -3 ) and CFU is correction factor for units (= 10 -1 ).

Quantifying total carbon stock (TCS)
The total C stock in the nested plot expressed in (Mg ha -1 ) was calculated by adding C stored in all pools in each subplot in the nest according to the equation: where CAGLB, CBGB, CDTB, CLB, and SOC were C stored in the aboveground live biomass, belowground biomass, dead tree biomass, litter biomass, and in the soil in the subplots expressed in (Mg ha -1 ), respectively. The amount of C stored in each types of forest stand (Mg) was calculated as follows: where Cplot is the total C stored in each plots expressed in (Mg ha -1 ), nplot is the number of sample plots in the stand, Ast is area of each stand (ha). The total C stock in the whole forest was calculated as follows: where CT is total C stock (Mg) and Cst is the total C stock of each forest stand (Mg).

Estimation of equivalent CO2 sink
Finally, as 1 Mg of soil C = 3.67 Mg of CO2 sequestered (Craig et al., 2010), the equivalent CO2 sink (Mg) in Chato forest was estimated based on the total C stock as follows: Values in Gg can be obtained by dividing Mg of OC or CO2 by 1000.

Statistical data analysis
Descriptive statistics was used to summarize mean and coefficient of variation of measured parameters. Generalized biomass models developed for moist tropical forests were used to determine carbon stock of forests. Mean separation was carried out using least significant difference (LSD) at p< 0.05.

Impact of stand type and biomass component on biomass accumulation
Biomass accumulation in the forest ecosystem is usually influenced by kind of forest, type of pool, tree size class and density, species composition, forest age, and level of protection, all of which determine the C storage level of the forest. The study result shows that Chato natural forest had accumulated large volume of biomass than plantation forest for similar pools (Table 1). Total biomass accumulation, the sum of biomass stored in all components, was highest for the natural forest followed by plantations including eucalyptus species, Cupressus lusitanica, Juniperus procera, Gravellia robusta, and lowest for acacia species. Larger biomass in natural forest might be attributed to maturity, species diversity and good understory cover. The study result shows that the average biomass stored (Mg ha -1 ) in different biomass pools decreased in order AGB > BGB > LB > DTB for all types of forest stands. The quantity of biomass accumulated in the aboveground biomass pool was significantly different from other pools at (p< 0.05) indicating more biomass was accumulated in the aboveground pool. The mean biomass accumulated in Chato forest by biomass components ranged from 2.73 Mg ha -1 in the dead tree to 315.90 Mg ha -1 in the aboveground biomass pools. Canopy cover, basal area, and height of trees might be attributed to the larger proportion of biomass in the aboveground biomass pool. The average value of the aboveground biomass for natural forest in the present study (603.72 Mg ha -1 ) was higher than the findings of Brown and Lugo (1982) (Table 2) while the remaining pools were accumulated only 21.01% biomass. In all forest stands, the smallest biomass was recorded in the dead tree/wood compared with other pools. Low tree mortality and decomposition of deed woods might be the causes for low dead tree biomass. The aboveground shrubs and saplings biomass was highly variable with stand type than other pools as depicted by larger coefficient of variation (76.73%) ( Table 1). This could be due to differences in suitability of various forest stands for the understory growth and it was more vigorous in the natural forest than plantations.   The C storage capacity varies with type of pool. The AGC pool and SOC were significantly different from other pools and from each other at (p< 0.05). The DTC and LC were also significantly different from AGC and SOC but not significantly different from each other at (p< 0.05). The study shows that the mean C stock of the major pools in each forest stand decreased as AGC > SOC > BGC > LC > DTC; implying more C allocation in the aboveground pool (Fig. 2). Nearly 63.45% of C was stored in the aboveground pool followed by 19.79, 15.23, 0.99 and 0.55% in the soil, belowground, litter, and dead tree, respectively. This was in line with Zerihun Getu et al. (2012) report that tropical forests in their natural condition contain more aboveground C per unit area than any other land cover type. The average C stored in the aboveground pool for the natural forest was 301.86 Mg ha -1 while that of plantation forest was 129.17 Mg ha -1 . Combining the C sequestered in the natural forest and plantations, the mean aboveground C storage capacity of the Chato forest was 157.95 Mg ha -1 . The average aboveground C for natural forest in the present study (301.86 Mg ha -1 ) was larger than the average C in the aboveground biomass for tropical forests in Malaysia (149 Mg ha -1 ) but smaller than estimates in the Phillipines (406 Mg ha -1 ) reported by Tara (2012) and Lasco et al. (2006), respectively. The study result indicated that average aboveground C in the tree plantations was better in eucalyptus species (205.07 Mg ha -1 ) and Cupressus lusitanica (180.14 Mg ha -1 ) as they are relatively older than other tree plantations. The mean belowground C for the natural forest (72.45 Mg ha -1 ) was much higher than that tropical forest in Malaysia (27 Mg ha -1 ) (Tara 2012). The contributions of DTC and LC to the total C pool were minor which might be due to decomposition of dead wood over time leading to loss of C. The mean SOC storage potential to a depth of 60 cm in Chato forest was 49.25 Mg ha -1 , where natural forest and plantations stored averagely 71.04 Mg ha -1 and 44.90 Mg ha -1 , respectively. The average SOC for the natural forest in the present study was a little higher than SOC stock range of 58.3 to 63.9 Mg ha -1 reported by Solomon et al. (2002) for humid tropical forest in southeastern Ethiopia and that of plantations is at par with 44.2 Mg ha -1 reported by Thomas et al. (2015). Mulugeta Lemenih et al. (2005) also found SOC storage of 23.4 Mg ha -1 for Cupressus lusitanica plantation which is lower than our findings (53.16 Mg ha -1 ).
The amount of C stored in the soil was greatly affected by species richness, age, size and density of forest and the understory cover. In forests with high plant diversity, it is likely that they would have litters with different degrees of chemical resistance; creating the possibility of longer residence of C through slower decomposition of litters and build up of soil C. As C is generally a more variable parameter, coefficient of variability (CV) was high for most C pools investigated within different forest stands (Table 3). Relatively, C stock variation within stand type was highest in DTC pool (71.50%) and lowest in SOC pool (33.38%). This implies C in the vegetation is more variable than C in the soil.

Net carbon sequestration rate
Plantation forest established in the study site some 31 years back had added nearly 175.93 Gg C to the forest ecosystem. Our result shows that the average C sequestration rate for the plantations of varying age was 8.65 Mg ha -1 yr -1 ; the quantity higher than the average value of 3.98 Mg ha -1 yr -1 for mixed plantation forest in China reported by Yuanqi et al. (2015). Previous research indicated that plantation forests are a cost-effective means of sequestering C (Adams et al. 1999). Among plantations, eucalyptus species and Cupressus lusitanica were relatively matured and thus, stored more C than other plantations (Table 3). Young forest holds less C, but it is sequestering additional C over time. An old forest may not be capturing significant quantity of net new C but can continue to hold large volumes of C in the form of biomass over long periods of time. In line with this, Lewis et al. (2009) indicated that old natural forests may not be C neutral but continue to be C sinks and observed a slow increasing tree C storage rate of 0.49 Mg ha -1 yr -1 in African tropical old growth forests. Generally, a study by Popo-ola et al. (2012) indicated that planting new forests, rehabilitating degraded forests and enriching existing forests contribute to mitigating climate change as these actions increase the rate and quantity of C sequestration in biomass.

Climate change mitigation and economic potential of the forest
In the entire forest ecosystem, a total of 6371.30 Gg C was stored in the vegetation plus soil (Table 4). However, owing to protection of existing forests and expansion of plantations in the study area, there is rather a net addition of C to the forest ecosystem. Perez et al. (1997) suggested that the additional C sequestered from afforestation and reforestation could offset even the C release from deforestation. As net gain is the main concern in climate change mitigation strategies, the jungle forests of Amazon cannot be qualified for REDD+ if there is no positive addition of C to the system. The C emitted to the atmosphere from industries and other anthropogenic activities need to be offset by removal of the C by vegetation and artificial means, if any. In our study, we recognized that the undisturbed Chato forest fulfills the key REDD+ strategic areas and would be eligible for the international C market. By continuing the current afforestation and forest management program, the forest will be a potential emission reduction center. Tree seedling plantations on bare lands around the forest need to be strengthened to add more C to the system. Community should be empowered to own the forest and protect from potential dangers. Local government authorities need to be transparent and strong enough to protect the forest from potential destruction by private firms and individuals who involve in timber and charcoal production, if any. Generally, organizations working on REDD+ projects need to be transparent enough to ensure sustainability of the Chato forest if they are really concerned with tackling global climate change through afforestation, reforestation and forest management.

IV. CONCLUSION
Forest type, density, maturity stage and status of protection affect level of biomass accumulation. We understood that the undisturbed and matured natural forest stored more biomass than plantations. Intactness and species diversity of Chato moist evergreen Afromontane forest makes it one of the largest C reservoirs in the tropics. Carbon allocation was by far larger for the aboveground pool than any other pools. Younger and fast growing plantations have better C sequestration rate than the old forest and ensures net positive C additions to the forest ecosystem. Large volume of annually trapped C by the vast channels of Chato forest makes it the most significant regulator of global climate change. Sustaining the afforestation and forest management programs will possibly ensure the viability of the forest for REDD+ projects. Lastly, more research is required to explore the untapped potential of the forest and give due attention to develop C models specific to the sacred Afromontane forests.