Total Carbon Stock and Potential Carbon Sequestration Economic Value of Mukogodo Forest-Landscape Ecosystem in Drylands of Northern Kenya

Carbon sequestration is one of the important ecosystem services provided by forested landscapes. Dry forests have high potential for carbon storage. However, their potential to store and sequester carbon is poorly understood in Kenya. Moreover, past attempts to estimate carbon stock have ignored drylands ecosystem heterogeneity. This study assessed the potential of Mukogodo dryland forest-landscape in offsetting carbon dioxide through carbon sequestration and storage. Four carbon pools (above and below ground biomass, soil, dead wood and litter) were analyzed. A total of 51 (400 m) sample plots were established using stratified-random sampling technique to estimate biomass across six vegetation classes in three landscape types (forest reserve, ranches and conservancies) using nested-plot design. Above ground biomass was determined using generalized multispecies model with diameter at breast height, height and wood density as variables. Below ground, soil, litter and dead wood biomass; carbon stocks and carbon dioxide equivalents (CO2eq) were estimated using secondary information. The CO2eq was multiplied by current prices of carbon trade to compute carbon sequestration value. Mean ± SE of biomass and carbon was determined across vegetation and landscape types and mean differences tested by one-way Analysis of Variance. Mean biomass and carbon was about 79.15 ± 40.22 TB ha and 37.25 ± 18.89 TC ha respectively. Cumulative carbon stock was estimated at 682.08 TC ha; forest reserve (251.57 TC ha) had significantly high levels of carbon stocks How to cite this paper: Leley, N. C., Langat, D. K., Kisiwa, A. K., Maina, G. M., & Muga, M. O. (2022). Total Carbon Stock and Potential Carbon Sequestration Economic Value of Mukogodo Forest-Landscape Ecosystem in Drylands of Northern Kenya. Open Journal of Forestry, 12, 19-40. https://doi.org/10.4236/ojf.2022.121002 Received: October 3, 2021 Accepted: November 19, 2021 Published: November 22, 2021 Copyright © 2022 by author(s) and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access


Introduction
The main carbon pools on earth systems are atmosphere, terrestrial biosphere, ocean and Earth's crust (Hoover & Riddle, 2020). Terrestrial ecosystems (mainly forest, soil and wetland), are the major carbon pool components on earth's system (Beedlow et al., 2004;Lal et al., 2012;Xu et al., 2018) and largely contributes to the global carbon balance (IPCC, 2007;Hoover & Riddle, 2020). However, anthropogenic activities such as land-use change and combustion of biomass and fossil fuel are largely contributing to de-carbonization and accumulation of bio-spheric greenhouse gases (GHGs)- (Lal et al., 2012;Ciais et al., 2014;Friedlingstein et al., 2019). The accumulation of carbon dioxide (CO 2 ) and other GHGs in the upper atmosphere, has led to climate variability and associated stochastic events such as increases in the average global temperature, drought and flood events (Lal, 2004;Dabasso et al., 2014).
Since pre-industrial times, global CO 2 concentration in the atmosphere has increased by over 40% from about 277 parts per million (ppm) in 1750 to 407.38 ± 0.1 ppm in 2018 (Joos and Spahni, 2008;IPCC, 2013;Dlugokencky and Tans, 2018). Accordingly, warming from pre-industrial levels to the decade 2006-2015 was estimated to be 0.87˚C (IPCC, 2019) and reached approximately 1˚C above pre-industrial levels in 2017 (Allen et al., 2018). International efforts aim to limit the temperature increase to below 2˚C, preferably 1.5˚C above the pre-industrial level to reduce the risks and impacts of climate change (Gao et al., 2017;IPCC, 2018). According to the Nationally Determined Contribution synthesis report of 2021, to limit global warming to below 2˚C, CO 2 emissions need to decrease by about 25% from 2010 level by 2030 and reach net zero around 2070 (Chevallier, 2021). Consequently, climate mitigation strategies not only focus on reducing emissions of GHGs into the atmosphere but more on removing and stabilizing carbon concentration in the atmosphere (Gren & Aklilu, 2016).
Re-carbonization of the biosphere is important to reduce net anthropogenic carbon emissions through sequestration of CO 2 (Lal et al., 2012). Carbon sequestration is an important ecosystem service provided largely by terrestrial ecosystems. Estimates suggest that terrestrial ecosystems release about 10 to 20% of the total global CO 2 to the atmosphere due to land degradation, but also sequesters about 30% of CO 2 emissions from anthropogenic activities (Gibbs et al., 2007;Harris et al., 2012;Houghton et al., 2012;Houghton & Nassikas, 2018;Friedlingstein et al., 2019;IPCC, 2019). Accordingly, between 2009 and 2018, the terrestrial CO 2 sink increased to about 3.2 ± 0.7 GtC yr −1 down from 1.3 ± 0.4 GtC yr −1 in the 1960s (Friedlingstein et al., 2019). Despite the high potential of these ecosystems to sequester carbon, emissions from land-use changes and deforestation coupled with other carbon sources outweigh the carbon sink leading to the accumulation of greenhouse gases.
Forested landscapes are the largest carbon pool of the terrestrial ecosystem and integral in global carbon cycle (Pan et al., 2011;Abere et al., 2017;Zhao et al., 2019). The carbon pools in forest areas include; living biomass (above and below-ground biomass), dead organic matter (dead wood and litter) and soils (soil organic matter)- (Pan et al., 2011;Zhao et al., 2019;Hoover & Riddle, 2020). According to Pan et al. (2011), the carbon stock in the world's forests is estimated to be 861 ± 66 Pg C, with about 383 ± 30 Pg C (44%) in soil (1 m depth), 363 ± 28 Pg C (42%) in live biomass, 73 ± 6 Pg C (8%) in deadwood, and 43 ± 3 Pg C (5%) in litter. Out of 861 Pg C, about 471 Pg (~45%) of it is stored in tropical forests. The uptake of CO 2 from the atmosphere and storage within the forested ecosystem is one of the most practical and feasible way of reducing present and future emissions of CO 2 in the atmosphere (Trumper et al., 2008). It is also less costly since, it is natural based process and can be enhanced through restorative land-use and sustainable management (Lal et al., 2012).
Reducing carbon emissions is critical in combating climate change. Various carbon reduction mechanisms have been put in place by the United Nations Framework Convention on Climate Change (UNFCCC). These include Kyoto protocol, reducing emissions from deforestation and degradation (REDD+), the nationally determined contributions as provided for in the Paris agreement and the creation of carbon credit offset markets. Accordingly, International initiatives to offset and maintain greenhouse gases require an understanding of the existing and future potential of forest landscapes in carbon emissions and sequestration (Lal et al., 2012). Therefore, estimation of biomass (carbon stock) is pre-requisite to quantify the potential carbon sequestration in forests including the woodlands.
The estimates of biomass (and carbon) can be determined through field inventories only or a combination with various remote sensing approaches (Ubuy et al., 2018) using both direct and indirect methods. Direct methods use biomass models developed through destructive sampling of selected trees, while indirect methods involve the use of allometric volume equations, form factor and biomass expansion factors and or with wood basic density (Chave et al., 2014;Nja- (Vashum & Jayakumar, 2012) and may not be applicable in protected or threatened forests (Tetemke et al., 2019). Therefore, the use of allometric models is the commonly used approach (Chave et al., 2003;Ngomanda et al., 2014). Conversely, the accuracy of biomass estimated using allometric models is dependent on the appropriateness and applicability of the chosen model (Chave et al., 2014). The models can be species specific (species-site specific, species specific but from multiple sites) or general (multiple species from single site or multi-species from several sites)- (Henry et al., 2011).
The general multi species-site models are appropriate for extensive forested landscapes with large number of different species (Chave et al., 2005) as is the case in this study. Drylands occupy about 45.4% to 47.2% of the world's total land area (Lal, 2004;Lal, 2019). The dryland ecosystem contributes significantly to land-based carbon sink and negative feedback to global carbon cycle given its expansiveness (Lal, 2019) and stores about one third of the global carbon stock (Trumper et al., 2008). The drylands of Kenya cover about 80% of the total landmass in the Country (Githae & Mutiga, 2021). These drylands are mostly utilized for pastoral systems. The potential of rangelands and dryland forests to store carbon is well documented globally (Lal, 2004;IPCC, 2007;Trumper et al., 2008;Lal, 2019) and is influenced by its response to communal grazing effects (Perez-Quezada et al., 2011). However, the potential of these pastoral ecosystems to sequester and store carbon is poorly understood in Kenya. Moreover, attempts to estimate carbon stock in such ecosystems have not considered heterogeneity of these landscapes.
This study was undertaken in the heterogeneous pastoral environments of Mukogodo forest landscape, in the drylands of Northern Kenya. The objectives of the study were to assess the potential of Mukogodo forest landscape to store and sequester carbon, and equivalent economic value of carbon sequestration. The study covered the woodlands (ranches and conservancies) and dry forest (Mukogodo forest reserve). The study sites were classified into three landscape types (forest reserve, conservancy and ranch) and six vegetation categories (closed forest, open forest, grassland, shrubland, bare land and Opuntia dominated areas) depending on vegetation life forms and canopy cover to capture landscape variability in drylands of Northern Kenya. The study applied the commonly used tree variables (Diameter at Breast Height-DBH, total tree height and wood density)- (Chave et al., 2014) to estimate above ground biomass carbon and analyzed four main carbon pools of forest landscapes (living biomass, soil, deadwood and litter). The purpose of this study therefore, was to understand the capacity of dryland ecosystems to offset carbon and combat climate change by estimating the carbon stock and carbon sequestration potential and worth. The findings can contribute to the development of conservation policies for these fragile ecosystems as carbon sinks and for understanding the potential for carbon credits and associated economic benefits to the society in future.  1994;Kagombe et al., 2006;Kagombe and Owuor, 2007). The landscape also ex-  The landscape is found in agro-climatic zone V, (semi-arid), thus an ecologically sensitive ecosystem (World Bank, 1993;KFS, 2008). It is characterized by rugged terrains with hilly masses of between 10% and 40% slope (Muchiri and Gachathi, 2006;KFS, 2008). The elevation ranges between 1600 to 2100 m. The mean annual rainfall ranges between 400 and 600 mm. The rainfall distribution is bimodal with long rains in March-April and short rains in October-December.
The Mukogodo landscape is inhabited by the Laikipia Maasai and the indigenous hunter-gatherer community Yaaku. The two Isiolo conservancies are mainly occupied by Samburu and Turkana in Oldonyiro conservancy, Turkana, Somali, Borana, and Samburu in Leparua conservancy. The main economic activity in the landscape is semi-sedentary pastoralism where cattle, sheep, goats and camels are kept in communal grazing lands (Ng'ethe et al., 1997;M'mboroki et al., 2018). Eco-tourism and nature-based enterprises are emerging economic activities in the landscape. Despite the significant importance of the Mukogodo ecosystem to the communities' livelihoods, forests and lands are threatened by degradation. The driving factors of degradation include forest fires, deforestation, charcoal production, and grazing pressure, which have reduced forest cover over the years (Webala et al., 2006;M'mboroki et al., 2018). For instance, between 1984 and 2014, the forested landscape in Mukogodo reduced by 3071 ha (24%)- (M'mboroki et al., 2018). This increases the pressure on existing landscape to provide vital ecosystem services such as carbon storage and sequestration.

Sampling and Data Collection
Primary data for estimating above ground biomass and carbon stock were collected through vegetation assessment in accordance with National forest inventory sampling framework (Hyvönen et al., 2016;Ndambiri et al., 2020). Multi-stage stratified random sampling technique was used to collect vegetation data. The landscape was first stratified into three types: forest reserve, group ranches and conservancies. In each of the landscape class, the vegetation was where Y is AGB (kg/ha), d is Diameter at Breast Height (cm), h is the total tree height (m) while ρ is the wood density (g·cm −3 ).
The above equation was selected due similarity in species harvested to develop the model with those sampled in this study and the similarity in eco-climatic conditions with Mukogodo landscape. Moreover, the model applied more than one parameter (DBH, height and wood density) which tend to give reliable results (Chave et al., 2005;Nam et al., 2016;Aabeyir et al., 2020).
Secondly, BGB was estimated as a fraction of the above ground biomass by multiplying with a shoot root ratio of 0.28 (MEF, 2019) and comparing with values from other studies (Cairns et al., 1997;MacDicken, 1997;IPCC, 2003;Marklund & Schoene, 2006;Mokany et al., 2006). Dead wood biomass was estimated based on dead-live ratios of 0.12 (Marklund & Schoene, 2006). Litter biomass was assumed to be 5% of the total biomass (Marklund & Schoene, 2006;Pan et al., 2011;Sun and Liu, 2020). The soil carbon was assumed to account for 32% of the biomass (Pan et al., 2011) based on assumption that both AGB and BGB account for 56% of the total carbon pool (Pan et al., 2011). The total biomass of Opuntia stricta was assumed to be 63.52 TB ha −1 based on proxy of mean productivity of Opuntia ficus-indica found in several studies (Nobel, 1995;Nefzaoui et al., 2014;Dubeux Jr. et al., 2015;Fouche & Coetzer, 2015;Iqbal et al., 2020). Respective carbon stock in each carbon pool was estimated by multiplying biomass by a coefficient of 0.475 (Raghubanshi, 1991;Singh and Chand, 2012).
This study assumed that soil carbon accounted for 100% of total carbon stocks for bare land in accordance with Solomon et al. (2018). Accordingly, the carbon stock of bare land was assumed to be the average of grassland and shrubland soil carbon, since most bare lands in the landscape are within the two vegetation types.  (Petersson et al., 2012). The CO 2eq was then multiplied by the current prices of carbon trade to obtain potential carbon sequestration economic value as per Equation (2): where Vc is the value of carbon sequestration (US$), E CO2eq is the estimated carbon dioxide equivalent, while ρ is the price of carbon (US$).
According to the World Bank (2020), Carbon prices ranged from less than US$1 T −1 CO 2eq to US$119 T −1 CO 2eq , with almost half of the covered emissions priced at less than US$10 T −1 CO 2eq . This study used a conservative value of 2 US$ T −1 CO 2eq , the global average carbon price provided by IMF (2019)  The bare land contributed marginally to the carbon pools through soil carbon pool only (Figure 3).  To effectively participate in carbon market, reliable estimation of total biomass carbon storage is essential (Weiskittel et al., 2015). The robust estimate is also critical for sustainable forest management decision making, for monitoring status of the forest and reporting carbon stock dynamics as required by Reducing Emissions from Deforestation and Forest Degradation (REDD+) mechanism (Ubuy et al., 2018). Further, the international negotiations on offsetting greenhouse gases require reliable current and potential estimates of forested areas to emit and sequester carbon (Pan et al., 2011). Therefore, valuing of the forest areas for their carbon storage potential may influence their protection through development of financial incentives for carbon storage.

The Value of Carbon Sequestration
Results from the study, indicate that, most of the carbon pools were contributed by living biomass carbon (~56%), followed by soil (~32%), deadwood and litter carbon by about 7% and 5% respectively. The litter and deadwood carbon were within the range reported by other studies (Tiessen et al., 1998;Pan et al., 2011). The high contribution of living biomass over soil to overall carbon were in agreement with Meena et al. (2019) who reported that living plant biomass contribute about 40% to 49% and Simegn et al., 2014 who reported about 57% of the total carbon from living biomass. Abere et al. (2017) and Atsbha et al. (2019) also reported findings that were within the range found in the study. This was, however contrary to the findings by other studies undertaken in nearly similar dry ecosystems (Dabasso et al., 2014;Solomon et al., 2018;Gebeyehu et al., 2019) who found soil to contribute the greatest carbon storage potential than the other carbon pools.
The variation in soil and living biomass carbon stock may have risen from the use of different biomass model, the application of empirical vs. secondary data and information in estimating carbon pools and methodological difference. Further, Zhao et al. (2019) indicated that the variation in data sources, estimation methods, scope of study area and environmental variables with different biotic and abiotic conditions and response to climate change may lead to significant variation in carbon storage estimates. Moreover, Keiluweit et al. (2015) re-Open Journal of Forestry ported that livestock grazing affect soil physico-chemical properties and nutrient cycling which result to soil organic carbon loss. The persistent grazing in Mukogodo forest landscape may have affected carbon storage potential of the soil carbon pool. The fact that majority of the carbon stock is stored in the living biomass suggests that any anthropogenic disturbances that might adversely affect the vegetation will have significant implication on carbon stock and sequestration potential of Mukogodo forest landscape.
The mean carbon stock from this study was slightly lower compared to those reported for nearly similar landscape in northern Kenya and Ethiopia (Dabasso et al., 2014;Gebeyehu et al., 2019), but was within the reported range in other dry forest-landscapes (Tiessen et al., 1998;Glenday, 2008;Simegn et al., 2014;Abere et al., 2017;Atsbha et al., 2019;Srinivas and Sundarapandian, 2019). The effects of vegetation and landscape type were significant on carbon stocks. The forest reserve stored most carbon within the landscape than the group ranches and conservancies. Furthermore, the high biomass carbon stock in closed forest than other vegetation types is in agreement with the findings of other studies in nearly similar ecosystems (Rajput et al., 2017;Solomon et al., 2017). In this study, the conversion of intact forest to open forest showed the potential carbon loss of about 73.37%, which is within the range reported by Wekesa et al. (2016). The observed variation in biomass carbon across the landscape and vegetation types may be due to variation in tree density, height, diameter size and low litter which facilitate decomposition of plant material for soil carbon formation. The large diameter, heights and density of trees in the forest reserve may have contributed to high carbon stock (Gibbs et al., 2007;Solomon et al., 2017;Dibaba et al., 2019;Srinivas and Sundarapandian, 2019) and their removal will impact largely on biomass dynamics in the landscape. The slightly high potential of conservancies to store carbon compared to the ranches is an indication that unsuitable land use practices such as intensive grazing have high potential of enhancing carbon emission and reducing the capacity of rangelands as carbon sink.
Sustainable management of forest areas and rehabilitation can enhance carbon stock. According to Mendelsohn et al. (2012), about 42% of carbon storage could be achieved through reduced deforestation, 3% from forest management, and estimated 27% from afforestation. Contrary, poor management coupled with deforestation and degradation can significantly reduce carbon storage (Dibaba et al., 2019). The existence of high carbon stock in the forest shows the potential of the area for climate change mitigation. The landscape should therefore be sustainably managed through reduction of deforestation and land degradation, promotion of sustainable landscape management to enhance in-situ carbon storage and carbon sequestration potential to mitigate effects of climate change and ensure continued provision of other ecosystem services.

Conclusion
This study estimated the biomass, total carbon stock, carbon sequestration po- The high carbon storage potential underscores the importance of the landscape as carbon sink and contribution to the global carbon cycle. Further, the large proportion of carbon in the landscape is stored in living biomass and closed forest, thus, a slight disturbance through deforestation and land use change may significantly reduce the carbon storage potential. The persistent exposure of group ranches to grazing had reduced their carbon storage potential by about 50.31% compared to the conservancies. The finding of this study will inform policy formulation on access of carbon funds through Clean Development and REDD+ mechanisms which will boost conservation and further enhance the carbon stocks.

Recommendation
Efforts should be enhanced to sustainably manage the landscape through restoration practices to reduce emissions associated with degradation and enhance carbon storage potential and flow of other ecosystem services. Continuous monitoring of carbon stock is also important to estimate net carbon storage and sequestration. To achieve this, the use of primary data in estimating carbon storage is highly recommended to give precise results. More research would be necessary to assess the impact of land use on carbon storage potential and feasibility of carbon credit investment in such pastoral ecosystems. Open Journal of Forestry