Effect of Land Use Changes on Carbon Stock Dynamics in Major Land Use Sectors of Mizoram, Northeast India

Land use change activities have greatly affected the total ecosystem carbon stock (TECS) and also contribute to global change through emission of greenhouse gases. The present study assessed the change in vegetation biomass carbon stock (VBCS) and soil organic carbon stock (SOCS) following conversion in major land use sectors (agriculture, agroforestry, forest and plantation) in Mizoram, Northeast India. SOCS was the highest in agroforestry (50.85 Mg C ha) and the lowest in agriculture (33.99 Mg C ha). VBCS was the highest in plantation (131.66 Mg C ha) and the lowest in agriculture (7.44 Mg C ha). The highest positive TECS change rate was observed when agriculture was converted to plantation (6.61 Mg C ha∙yr), while negative rate of change in carbon stock was observed following the establishment of agriculture from other land use. A positive rate of change was observed in both VBCS and SOCS with TECS rate of 3.58 Mg C ha∙yr when agriculture got converted to agroforestry. The absolute carbon stock change rates were higher in VBCS than SOCS signifying the importance to maintain tree based vegetation cover.

Pine and Teak. A total of 38 sites distributed over 16 land use types were selected for the present study ( Figure 1). The age of the different land use types were recorded from the landholders and villagers.

Biomass Carbon Estimation
Permanent plots (250 × 250 m 2 ) were set up in each of the selected land use types following ISRO-GBP/NCP-VCP protocol [21], and four nested sampling quadrats were laid in the corners. The permanent plot sizes were reduced to 100 × 100 m 2 in land use types where land holding size was less than 2 ha. Nested quadrats size were of 0.1 ha (31.62 × 31.62 m 2 ) for estimating tree and deadwood; two quadrats of size 5 × 5 m 2 to estimate saplings/shrub/bamboo; and four quadrats of size 1 × 1 m 2 for herbs and standing litter ( Figure 2). All trees  greater than 30 cm girth over bark at breast height (GBH, 1.37 m above ground level) were identified and tagged, and were measured for GBH using a metal tape and height by use of Haga altimeter and measuring tape. All coarse deadwood biomass (>10 cm diameter) were recorded following fixed-area sampling (FAS) [22]. Collar diameter (5 cm above ground level) and height of all saplings and shrubs encountered were measured with digital vernier caliper and measuring tape respectively. Plant density (ha −1 ) and basal area (m 2 •ha −1 ) was calculated from the expanded values of each nested plots and averaged for each land use type. Above ground biomass (AGB) and below ground biomass (BGB) were estimated using appropriate allometric models for the trees/deadwood and saplings/shrubs ( Table 1). The specific wood density values of trees were adopted from the Global Wood Density Database [23]. Volume of deadwood was multiplied with determined values of its corresponding decay classes, viz. sound, partial and full decomposed considered as 0.45, 0.35 and 0.25 g•cm −3 respectively [24].
Bamboo biomass (Melocanna baccifera) was estimated by harvesting bamboo culms samples from different diameter classes and total bamboo biomass was estimated as the sum after multiplying dry biomass with the corresponding densities of its diameter class. The aboveground biomass of herbs and litter was estimated by harvest method [25]. All biomass measurements were converted to a per hectare basis (Mg ha −1 ). Table 1. Summary of allometric equations used in the study.

Total Carbon Stock Change Estimation
Biomass carbon stock density (Mg C ha −1 ) for each of the land use was estimated by using the default carbon content of 47% [37]. Above ground biomass (AGB)

Land Use Characteristics
Density and basal area of the trees and saplings/shrubs/bamboo showed significant variation (p < 0.05) in different land use types ( Table 2). The highest tree density was encountered in arecanut plantation (1520 ha −1 ) and the lowest in grassland (110 ha −1 ). Saplings/shrubs/bamboo density was highest in bamboo forest (26,833 ha −1 ) followed by dense forest (7817 ha −1 ) and the lowest in oil palm plantation (1150 ha −1 ). No trees and woody species were encountered in the Wet Rice Cultivation land use system. Plant density (individual ha −1 ) variation in different land use types was greatly influenced by the intensity of anthropogenic interventions and management practices [39]. The presence of more number of trees in jhum fallow as compared to current jhum in the present study indicated a rapid recovery and high resilience of regenerating secondary forest, however tree growth was reportedly inhibited by shrub dominance [40].
Tree densities in old and young homegarden (513 and 408 ha −1 respectively) from the present study are higher in range with 239 -319 ha −1 in Kerala [41] and 220 -409 ha −1 in Philippines [42] which might be due to the maximum number of trees in lower girth classes (<90 cm dbh) and multi-strata canopy structure as reported from homegardens in Mizoram [43]. Low tree density (408 ha −1 ) observed in open forest is attributed to deforestation prevalent due to land use change conversions [44]. Low density of bamboo (Melocanna baccifera) in the present study with 26833 culms ha −1 might be due to overexploitation of mature culms more than 4 years old leaving mostly the immature culms [45]. Amongst  vegetation basal area with 16.0 m 2 •ha −1 followed by bamboo forest (14.3 m 2 •ha −1 ) and the lowest in arecanut and oil palm plantations (0.4 m 2 •ha −1 ).Basal area cover in different land use is dependent on species composition, tree size and growth pattern [46]. Despite low tree density, oil palm had the highest basal area (117.4 m 2 •ha −1 ) followed by coffee plantation comprising of matured trees (62.0 m 2 •ha −1 ) as a result of maximum number of trees in large circumference class. Whereas, the basal area in orange (7.7 m 2 •ha −1 ) and arecanut (17.6 m 2 •ha −1 ) plantations were low due to maximum distribution of stems in small size circumference class. Overall, average tree density and basal area was the highest in plantations with 745 ha −1 and 43.21 m 2 •ha −1 respectively showing significant differences with the other land uses. However, the Tukey HSD test indicate no significant differences in tree basal area cover among agriculture, agroforestry and forest (p > 0.05). Higher overall density and basal area in plantation compared to forest, agroforestry and agriculture indicate the efficiency of land use pattern where farmers follow intensive monoculture practices. However, land use conversion to plantation accompanies loss of biodiversity, vegetation structural changes, lower ground water table, etc. which further affects the ecosystem's carbon dynamics [47]. Soil bulk density (≤2 mm) and soil organic carbon (SOC) content also varied significantly (p < 0.05) amongst different land use types at various soil depths (Table 3). Agroforestry based land uses showed higher bulk density in all depths amongst the major land use sectors, however significantly (p < 0.05) different only in 0 -15 and 15 -30 cm. Soil bulk density was highest under old home garden (0 -15 and 15 -30 cm) and wet rice cultivation (30 -45 cm) and lowest at all depth classes in the pine plantations. The higher soil bulk density found in all soil layers in agroforestry and plantation compared to forest can be attributed to more soil compaction as a result of frequent cultivation activities; however, similar bulk density values in agriculture with forest might be due to constant tillage practices adopted [48]. SOC concentration (%) decreased significantly (p < 0.05) with increasing soil depth in all land uses. The decreasing trend of SOC content with increasing soil depth, common in all mineral soils, is in agreement with earlier studies [36] [49]. This might be due to higher organic matter input and  Figure 3). Lower SOC content in plantation and agriculture as compared to forest and agroforestry could be due to less organic matter input and more soil disturbance resulting in high carbon mineralization rate as a result of cultivation [51] [52] [53]. Removal of biomass during harvesting and periodic tillage breaking up soil macro aggregates further reduces SOC content in agriculture and plantations [54]. The highest biomass (aboveground + belowground) was observed in coffee plantation (1065.44 Mg ha −1 ) followed by teak plantation (487.03 Mg ha −1 ), dense forest (341.38 Mg ha −1 ) and the lowest in wet rice cultivation (7.63 Mg ha −1 ). Amongst the forest, dense forest recorded the highest biomass with significant (p < 0.05) differences followed by open forest (178.54 Mg ha −1 ), bamboo forest (14.60 Mg ha −1 ) and grassland (39.22 Mg ha −1 ). The allometric models and sampling approach used could have substantial influences on the results of ground based biomass estimates, and thus locally developed and calibrated models have the potential to minimize this uncertainty in biomass carbon accounting [28] [55] [56]. However, in our present study, we used the most recent model keeping into account of similarities in climatic and ecological parameters.   (Table 5). SOC stock was highest in agroforestry (50.85 Mg C ha −1 ) and followed the trend: agroforestry > forest > plantation > agriculture (Table 5). Highest VBCS in plantations from other land uses signifies the great potential for biomass carbon sequestration through effective and proper management planning. The SOC stock up to 45 cm depth ranges from 16.00 (mango plantation) to 62.58 Mg C ha −1 (dense forest) and the total ecosystem carbon (TEC) stock which is the sum of VBC and SOC stocks ranged from 31.35 (jhum fallow) to 534.14 Mg C ha −1 (coffee plantation) ( Table 4). Differences in total ecosystem carbon (TEC) stock amongst land uses have been attributed to differences in VBCS and SOCS. The distributions of SOC proportion at different depths varied significantly in the different land use types ( Figure 4). Overall, an average of 46.58%, 31.33% and 22.09% of SOC stock were distributed within 0 -15, 15 -30 and 30 -45 cm soil depths respectively in all the land use systems. Location, soil type, tree species and plantation management system influencing soil bulk density and SOC content might be responsible for SOC stock differences between  [65]. SOC stocks distribution at various soil depths except 0 -15 cm across land use sectors indicated a significant decrease with increasing depth ( Figure 5). Understanding the SOC vertical pattern of different land uses will enhance our knowledge of carbon dynamics along a profile and its potential response to climate change [66]. SOC stock in agroforestry was significantly (p < 0.05) different only with agriculture and plantation. Agroforestry stored the highest SOC (50.85 Mg C ha −1 ) stock compared to forest and plantations and the lowest in agriculture for a soil depth of 0 -45 cm, which is comparable to findings from other studies where traditionally managed agroforestry systems have higher SOC storage than agricultural systems [17] [67] which may be due to differences in species diversity, composition and intensity of management practices leading to soil disruptions [68].

Vegetation Biomass Carbon, SOC and Total Ecosystem Carbon Stock
The SOC: TEC stock ratio was highest for wet rice cultivation (92.75%) and the lowest was for mango plantation (16.57%) across the land use types (Table   4). SOC: TEC was the highest for agriculture (80.52%) followed by agroforestry (55.56%), and lowest in plantation (36.00%) being significantly (p < 0.05) different from other major land uses (Table 5) which indicates the direction of change in carbon storage in different pools. In the present study, the higher proportion of SOC with corresponding low floor biomass and vegetative cover makes the land use more prone to SOC losses through accelerated soil erosion [69].  implies the need for management practices to maintain the balance between plant biomass productivity and microbial decomposition for SOC stock following land use changes in the wake of impending climate change [70]. It is important to evaluate SOC storage potential and improve the biological cycle of ecosystems to maintain an equilibrium fixation and storage [71].

Carbon Stock Change Estimation
Land use changes showed positive and negative carbon stock change rates depending on land use type (Table 6). Positive rate of carbon stock change was observed in all the pools when current jhum was converted to teak, arecanut, young home garden and oil palm plantations. However, the conversion of current jhum to grassland witnessed a loss of SOC stocks (−0.22 Mg C ha −1 •yr −1 ) although the vegetation biomass carbon stock gained (0.34 Mg C ha −1 •yr −1 ). Land use conversions may result either a decrease [72] or an increase in rate of SOCS change [73]. The negative rate of change following conversions of dense and open forest to current jhum and grasslands observed in the present study is similar to reports from China [74].   agroforestry and forest to plantation; and agroforestry to forest (Table 7). Whereas, SOCS change rate were positive in conversion of land uses to agroforestry; agriculture and plantation to forest; and agriculture to plantation. A positive SOCS change rate of 0.33 Mg C ha −1 •yr −1 in the top 20 cm soil was reported following conversion of cropland to forest in China which is similar to the findings from the present study [77].   from tropical forest soils [78]. This implies the degradation impacts of slash and burn practices involved with shifting cultivation (jhum) in the tropics where land become scarce and leads to reduced fallow periods [79], thereby the natural nutrients recovery for crop production is not complete and the intensification of shifting cultivation on the same land makes it become unsustainable. In the present study, establishment of plantations following conversions from forest and homegardens observed a negative SOCS change rate similar to findings reported by other studies [80]. SOCS changes due to land use change are caused by changes in soil carbon inputs (litter quality and quantity) and outputs (alterations in decomposition processes) when one vegetation is replaced by the other [81]. VBCS change rate showed positive gain following conversion of land uses to plantation; agriculture and agroforestry to forest; and agriculture to agroforestry. Conversion of land uses to agriculture; forest and plantations to agroforestry; and forest to agroforestry had exhibited a negative VBCS change rate. This suggests that the tree-based systems have substantially enhanced the ecosystem carbon storage and aid to climate change mitigation/adaptation. In the present study, the absolute carbon stock change rates following land use change were higher in VBCS than SOCS, except for land use conversion from agroforestry to forest. This signifies the importance and vulnerability of vegetation biomass pool whose sequestration is greatly affected by land use management implications subject to changing climate and soil conditions. Therefore, land uses which are degraded physically, chemically and biologically following conversions need to be restored through tree-based systems. The study results indicate ecosystem carbon sequestration rates to be significantly high in plantations, however they have been often associated with environmental issues of biodiversity losses and disruption of ecological cycle [82]. Thus, the land use change management needs to focus and identify potential systems such as agroforestry systems which would preserve species, accumulate soil carbon and tree biomass in the longer run.

Conclusion
Diversified land use patterns in different sectors: agriculture, agroforestry, forest and plantations have been prevalent in Mizoram, Northeast India which is influenced by a combination of different reasons such as resource scarcity, market opportunities, policy interventions, increased vulnerability to resource access and change in attitudes. These land use changes in various forms affect the total ecosystem carbon storage whereby management practices involved induced great differences amongst carbon pools. The highest carbon stock was observed in plantations and the lowest in agriculture. Both biomass and soil carbon stocks were observed higher in tree-based land use systems compared to agriculture. The SOC stock proportionately contributed more in agriculture systems and less in plantations. Among all the managed plantations, coffee plantations have a dense canopy with large diameter shade trees and exhibited the highest carbon storage indicating the best management practices adopted. Absolute carbon stock change rate following conversions were maximum in agriculture with losses in all carbon pools. Conversions to agroforestry attract a positive change soil carbon pool; whereas conversions to plantations exhibited negative change in soil and a positive rate of change in biomass carbon with an overall gain in total ecosystem carbon stock. However, considering environmental management and conservation issues, the rampant conversion of land uses to plantation should not be encouraged. In all the land use changes, the rate of change is comparatively higher in biomass than soil carbon pools, which signifies that maximum gain/loss in total ecosystem carbon stock, can be achieved through the management perspectives to maintain vegetation type and cover in the system. Increase in vegetation and floor biomass will also eventually lead to soil carbon enrichment. Open forest and jhum fallows should be kept undisturbed and allowed to recover fully through natural and assisted regeneration to dense forest. Selective land use and adoption of scientific cultivation practices should be the efforts of policy makers in tune with climate and carbon mitigation challenges. Thus, agroforestry systems and plantations equipped with sustainable management practices could be adopted in large scale for restoration of degraded lands in Mizoram.