Carbon Stock Variation along altitudinal gradient of Forest in Galan District, Kellem

Background: Forest ecosystem plays a crucial role in the global carbon cycle; as such, mitigating high atmospheric concentrations of carbon dioxide and other greenhouse gases by naturally taking carbon from the atmosphere through photosynthesis. Verification and accounting of carbon stock in forest ecosystem have been renowned as a potential strategy to reduce and stabilize atmospheric concentrations of greenhouse gas. Forest sequesters and store more carbon than any other terrestrial ecosystem and it is an important natural break on climate change. It acts as a carbon reservoir by storing large amount of carbon in trees, undergrowth vegetation, forest floor and soil. Result: The mean carbon stock of each carbon pool was changed along altitudinal class of the study area. The largest mean above and below ground carbon stock was found in the second altitudinal class(1560.01-1643m) followed by first altitudinal class(1435-1560m) and the third altitudinal class (1643.01-1704m) of the study area. The largest mean dead tree and dead wood carbon stock was also stored in the first altitudinal class followed by the third and the second altitudinal class of the study area. The largest mean litter carbon stock was found in the first altitudinal class followed by the second and the third altitudinal class of the study area. The largest mean soil organic carbon was found in the third altitudinal Conclusions : The carbon stock variation along altitudinal gradients indicated that, altitude had no a statistically significant effect on any of the carbon pools except litter carbon of the study area at 95% of confidence interval.

deficient scientific quantitative data of carbon stock variation along altitudinal range of the forest as well as it filled the gap stated in the above researchers.

Description of the Study Area
The study area is located in Hawa Galan district, Kellem Wollega Zone, Oromia Region, Ethiopia. The forest covered 720 hectares (ha) of land and it is located at about 630km to the South west of the capital city of Ethiopia, Addis Ababa and about 22 km to the town of Dambi Dollo. Geographically, it is found between 8°42'32"N -8°42'34"N and 40°52'49"E -40°53'55"E which is shown in figure (a).

Figure a: Location of the study area
The altitudinal range of the study area also ranges from 1435 to 1704m above mean sea level. The mean maximum and minimum temperature of the study area was 30.84°C and 16.38°C respectively, and the mean annual rainfall was also 1,645mm, which is shown in figure (b).

Delineation of the study area
The boundaries of the study area were delineated by taking geographic coordinates through Geographical positioning system (GPS) at each turning point of the study area.

Stratification of the study area
Wacho forest was divided into stratum based on the relative homogenous unit of topography, which helps to form more or less homogeneous units of the forest to increase the efficiency and accuracy of the forest carbon accounting.

Determine size and shape of sample plot
About 400m 2 areas of square sample plot were employed for sampling of the study forest. Because square plot has a better probability to incorporate more of within plot heterogeneity and thus be more representative than the other shape of sample plots of the same area as indicated by Hairiah at al. [11].

Determine sampling techniques and sample size
Sample plots of 20m×20m were laid through systematic random sampling techniques to collect the required and relevant data of the study forest at every 200m difference between each sample plot and 300m difference between each transect line. Finally, a total of 73 sample plots were laid by using GPS instruments starting from the lower to the higher elevation of the forest figure (c).

Field Measurements Sampling and identification of trees and shrubs
According to Pearson et al. [20] all tree and shrub species having ≥ 5cm diameter at breast height (DBH) were measured from 400m 2 areas of sample plots using diameter tape, and the height of those trees was also measured by using hypsometer. Woody plants having multiple stems at 1.3 meter were considered as a single tree and the largest stem was taken, while woody plants forked below 1.3 meter was treated as a single individual as indicated by Pearson et al. [20,21]. Plant identification was done at the National Herbarium of the Addis Ababa University using published volumes of Flora of Ethiopia and Eritrea.

Sampling of dead trees and dead woods
The samples of dead tree and dead wood were collected by using the principles of Goslee et al. [9] as follows.
Standing dead trees in class one which characterized by the existence of branches and twigs only having ≥5cm DBH were measured from 400m 2 area of sample plots by using DBH tape and the height of those trees were measured by hypsometer. But the other types of standing dead trees with small and large branches only, trees with large branches only, and trees with bole only were measured its diameter at the base of the dead tree by using DBH tape and the height of those dead standing trees were measured by hypsometer. The lying dead woods, having ≥10cm diameter were divided into sections of roughly one meter and the exact length and diameter at the middle of each section was recorded.

Sampling of litters
According to the principles of Pearson et al. [20] litter samples were collected manually from each of the five 1m×1m areas of subplots, which located at the four corners and one at the center of the main plot.
About 100 grams of a composite sample was taken by mixing litter samples from each of the five sub plots of the main plot. Then it was placed in a plastic bag and labeled to which sample plot it belongs.
Then after about 100 grams of 73 labeled composite samples were taken to the laboratory of Debrezeit horticoop Agricultural research center and the litter samples were oven dried to a constant weight at 105°C for 12 hours and the carbon fraction of litter samples were determined in the laboratory using Walkley-Black Method, 1934.

Sampling of soil organic carbon and bulk density
The samples of soil organic carbon (SOC) were collected by using the auger at a depth of 30cm from each of the five 1m×1m areas of subplots, which located at the four corners and one at the center of the main plot. About 100 grams of a composite sample were taken by mixing soil samples from each of the five sub plots of the main plot. Then it was placed in a plastic bag and labeled to which sample plot it belongs. Then after about 100 grams of 73 labeled composite samples were taken to the laboratory of Addis Ababa Agricultural research center. Then the field moist soil samples were dried to a constant weight in an oven at 105°C for 12 hours and the percentage of organic carbon was determined in the laboratory using Walkley-Black Method, 1934.
The bulk densities (BD) of the soil samples were also collected by using a core sampler at a depth of 30cm from each of the five 1m×1m areas of subplot pits, in which the samples of SOC were taken. The sub samples of BD were oven dried to a constant weight in an oven at 105°C for 24 hours to determine the oven dry weight of soil samples. The design of the main plot and subplot of the study samples is shown in figure (d).

Estimation of above ground tree biomass and carbon stock
The above ground biomass of trees and shrubs existed in the study area were calculated using the general allometric model of Chava et al. [3] as follows. The CO 2 equivalent sequestered in the aboveground biomass=AGC × 3.67……… (eq.3)

Estimation of below ground tree biomass and carbon stocks
Below ground biomass of trees and shrubs found in the study area was estimated by using root-shoot ratio factor of Mac Dicken [17]. According to Mac Dicken [17] and Pearson et al. [20], standard methods of estimating below ground biomass(BGB) and below ground carbon(BGC) can be obtained as 20% and 10% of above ground tree biomass respectively.
Where, BGB= below ground biomass BGC = carbon content of below ground biomass and 0.2 is the conversion factor (or root -shoot ratio), which is 20% of the above ground biomass.
The amount of CO 2 equivalent sequestrated in below ground biomass of the study area was calculated by multiplying BGC by the molecular mass ratio of carbon dioxide to Carbon (44/12) which is 3.67 as indicated by Pearson et al. [21].

Estimation of dead tree and dead wood biomass & carbon stock
The biomass of standing dead trees which characterized by the presence of branches and twigs and the absence of leaves was calculated using the appropriate equations of Chave et al. [3] as biomass estimation techniques of live trees, but about 6% of the biomass of leave was subtracted as it recommended by Pearson et al. [20]. The carbon stock of those standing dead trees was calculated by multiplying the standing dead tree biomass by 0.47, which is the default carbon fraction of Intergovernmental panel on climate change(IPCC) [14]).
The biomass of standing dead tree, which characterized by the presence of small and large branches only, the presence of large branches only, and trees having trunk or bole only was calculated using the volume of the cone as it recommended by Goslee et al. [9] VOL cone (cm 3 ) = The carbon stock of those dead trees was calculated by multiplying the dry biomass of dead tree with 0.47, which is the default carbon fraction of IPCC [14].
The biomass of lying dead wood was also calculated by using the volume and density of wood as recommended by Pearson et al. [20]. The carbon stock of those dead lying wood was calculated by multiplying the dry biomass of dead wood with 0.47, which is the default carbon fraction of IPCC [14].
The total carbon stock of dead tree and dead wood was calculated by summing up all carbon stock of dead trees and dead woods as follows.

Estimation of litter biomass and carbon stock
The litter biomass found in the study area was calculated by the formula of Pearson et al. [20] as follow.

Estimation of soil organic carbon (SOC)
The carbon stock density of soil organic carbon found in the study area was calculated using the volume and bulk density of soil as it recommended by Pearson et al. [21]. V = h × r 2 ………………………………………………………………… (eq.14) Where, V= volume of the soil in the core sampler (cm 3 ), h = the height of core sampler (cm), r = the radius of core sampler (cm). Moreover the bulk density of soil sample was calculated as follows.

Estimation of total carbon stock density
The total carbon stock density of the study area was calculated by using the equation of Subuied et al. [23] by summing the carbon stock densities of the individual carbon pools of the study area. variable (altitude) were processed and tasted by descriptive statistics and one way analysis of variance (ANOVA) at 95% of confidence interval. Descriptive statistics were used to summarize the mean carbon stock of each carbon pool of the study area, while one way ANOVA was used to determine the statistical significance difference of carbon stock along altitudinal gradient of the study area.

Results
There was a variation of mean carbon stock of each carbon pool along altitudinal gradient of Wacho forest, but it is not statistically significant at 95% of confidence interval except litter carbon stocks (table   1). The largest mean SOC was found in the third altitudinal class followed by the second and the first altitudinal class of the study area (table1). Because some part of the third altitudinal class of class of the study area was the place in which deposition of sediments due to soil erosion takes placed. Also, it might be due to soil type, soil depth, soil texture, tree cover and tree species, and degree of disturbance regime.
But the variation of the mean SOC stock distribution was not statistically significant along altitudinal class of the study area at α =0.05(F=0.84, P=0.920) as shown in (table 1). The mean SOC stock distribution of wacho forest was similar to the mean SOC stock distribution of Edgu forest, in which the largest mean SOC stock of Edgu forest was found in the upper altitudinal class followed by the middle and lower altitudinal class of the study area without statistically significant at α=0.05(F=1288,P= 0.311) [6].

Coclusion
The carbon stock of different carbon pools such as above ground carbon, below ground carbon, dead tree and dead wood carbon, litter carbon and soil organic carbon were varied within the study area due to the variation of environmental gradient. The upper altitudinal class of the study area was high in soil organic carbon, while the middle altitudinal class of the study area was high both in above and below ground carbon stock. The lower altitudinal class of the study area was also high both in dead tree and dead wood carbon stock and litter carbon stock. Even though altitudinal gradient was the factor that affects the carbon stock distributions of the study area, the carbon stock variations were not statistically significant at 95 % of confidence interval except litter carbon stock. So it was possible to conclude that; altitudinal gradient had not a statistically significant influence except litter carbon stock of Wacho forest.