Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Modelling of total soil carbon using readily available soil variables in temperate forest of Eastern Himalaya, Northeast India

Modelling of total soil carbon using readily available soil variables in temperate forest of... GEOLOGY, ECOLOGY, AND LANDSCAPES 2021, VOL. 5, NO. 3, 209–216 INWASCON https://doi.org/10.1080/24749508.2019.1706295 RESEARCH ARTICLE Modelling of total soil carbon using readily available soil variables in temperate forest of Eastern Himalaya, Northeast India a b a Gyati Yam , Om Prakash Tripathi and Debangshu Narayan Das a b Department of Zoology, Rajiv Gandhi University, Doimukh, India; Department of Forestry, North Eastern Regional Institute of Science and Technology, Nrjuli, India ABSTRACT ARTICLE HISTORY Received 12 October 2019 The study was carried out in temperate forest of Ziro valley, Arunachal Pradesh, Northeast India, Accepted 12 December 2019 during 2015–2016. Stratified random sampling was adopted for soil sample collection from three depths on a monthly basis in each of the permanent plots. Collected samples were analysed using KEYWORDS standard methodologies. The XLSTAT (ver.2019) was used for the partial least square regression and Microbial biomass carbon; modelling. The soil was acidic in nature having moderate soil moisture with very low temperature. PSLR; soil properties; VIPs; Soil texture varies from sandy loam to sandy clay loam in nature. Altogether, 16 readily available soil XLSTAT variables were used for identifying significant variables to be used for modelling the total soil carbon (TSC). Based on the results, five most contributing independent soil variables having high variable importance in prediction were used for modelling the TSC. The average annual TSC was recorded 5.33% for upper, 5.01% for middle, and 4.20% for lower soil surface, respectively. The developed depth-wise equations predict very close TSC with very low root-mean-square error (rmse) to observed values. Hence, the findings of the present study will be very much useful under limited data conditions to predict the TSC and also in inaccessible areas of temperate forest ecosystem. Introduction the ocean (Batjes, 2014). Soil and vegetation have very complex interrelation, and species composition has sig- Temperate forest being one of the major terrestrial nificant impacts on soil properties over a long period of biomes,soils arehighlyvariablebut generally fertilefor time (Bordoloi, Das, Yam, Pandey, & Tripathi, 2018). which soil quality is the key essential for reflecting the The concentration of elements in the soils is a good structure of vegetation (Gower, Landsberg, & Bisbee, indicator of their availability to plants. Nonetheless, for- 2003;Martinetal., 2001). Alternates between warm est soil plays a vital role in climate change mitigation by summers and markedly cool or cold winters which is restricting the direct release of carbon into the atmo- characterized by marked seasonality are the climate of sphere. Terakunpisut, Gajaseni, and Ruankawe (2007) temperate forests (Reich & Bolstad, 2001;Yam & stated that forests are the prominent sites to study climate Tripathi, 2016). This ecosystem plays a fundamental change in termsofnet carbon emissionsaswellas in role in the global carbon (C) cycle, regardless of the global storage capacity which is important for climate way that the biome C pool is the smallest among other regulation. As we know, the soil is particularly concerned biomes (Reich & Bolstad, 2001;Robinson, 2007). Bhatler with the cycling of nutrients which is highly governed by (2009) had reported that temperate forests store more the interactions with its surrounding and vegetation. But carbon than tropical forests. Different strategies for mea- now as a decisive aspect, the multifunctionality of soils is suring C reveal that temperate forests have presently net increasingly recognized in global land management Csinks (Martinetal., 2001). It was reported that 31% (Wiesmeier et al., 2019). and 41% of the soil organic carbon (SOC) pool are stored Carbon emission to the atmosphere is inevitably in 1–3 m depth in temperate evergreen and temperate increasing in many regions of the world due to varied deciduous forests, respectively (Jobbágy & Jackson, human activities. Various factors such as fossil fuel com- 2000). Thus, it was estimated that about 262 Pg C is bustion, deforestation, overpopulation, industrial emis- stored in temperate forest soils in the top 3 m depth. sion, land-use change, urbanization, etc., directly or Soil carbon is directly or indirectly proportional to indirectly emitting greenhouse gases causing global climate change and vice versa. Soil is considered to be the warming. But human being is considered to be the most highest sequester of atmospheric carbon as soil carbon. important one for deteriorating climate through various Potentially, the soil–vegetation carbon pool is much anthropogenic activities shaping their own kind of chan- more labile in the short term as compared with that of ged environment. The increase in atmospheric carbon the oceans although the soil carbon pool is smaller than dioxide (CO ) concentrations associated with the CONTACT Om Prakash Tripathi tripathiom7@gmail.com Department of Forestry, North Eastern Regional Institute of Science and Technology, Nrjuli 791109, India © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 210 G. YAM ET AL. progression of the anthropogenic impact on forest eco- systems from the molecule to the ecosystem level (Valladares, 2017). Sequestering of carbon serves as to compensate emissions from fossilfuelcombustionand other carbon-emitting activities while enhancing soil quality. An increase in the SOC pool is also essential to advancing global food security through maximising crop productivity (Lal, 2004). However, nutrient loss, soil ero- sion, soil conservation, desertification, organic matter decomposition, etc., may alter the soil dynamics. Soil carbon is also considered to be the largest pool of terres- trial carbon. Forest ecosystems store more than 80% of all terrestrial aboveground C and more than 70% of all soil organic C (Batjes, 2014; Jandl et al., 2007;Jobbágy & Jackson, 2000;Six et al., 2002). The soil carbon mainly stored in the form of SOC, microbial biomass carbon (MBC), and soil organic matter. Map 1. Study area (Ziro valley) in Arunachal Pradesh, India. Soil carbon is one of the important variables for deter- miningthefuturecarbonsink. Theimportanceofforest and soil C in mitigating and balancing the greenhouse above mean sea level (asl) during 2015–2016 in lower effect has been recognized, and an agreement was reached Subansiri district of Arunachal Pradesh, India (Map 1). under theKyotoProtocoltoinclude forest andsoilC Analysis of soil was done in Forest Ecology Laboratory of sequestration in the list of acceptable offsets (United North Eastern Regional Institute of Science and Nations Framework Convention on Climate Change Technology. It covers an area of 3460 km which is [UNFCCC], 1997). Lal (2005) had also stated that forest mostly mountainous terrain. The forests are rich in valu- soil carbon sequestration has a potential to decrease the able species of trees, shrubs, and different varieties of rate of enrichment of atmospheric CO . Many research- herbs. Cane and bamboo and exquisite floral treasures ers have enumerated the amount of carbon sequestered of wilder species may often be found in the shadowy in the standing trees (Brown & Gaston, 1995;Chave etal., recesses of the forest. The area mainly comprises sub- 2008;Houghtonetal., 2001; Malhi et al., 2006;Saatchi, tropical and alpine forests and has a variety of flora and Houghton, Dos Santos Alvala, Soares, & Yu, 2007;Yam& fauna, many of which are endangered. However, the Tripathi, 2015), but only a few studies have carried out on experiment was mainly carried out in the forests which soil carbon sequestration. Besides, for understanding the are considered as untouched and virgin natural tempe- soil dynamics and potential soil carbon sequestration, rate forest well protected by the “Apatani” people of this modelling of total soil carbon (TSC) may be an important place since time immemorial. The Ziro valley is a quaint attribute for the scientific researchers, stakeholders, and eastern Himalayan town located in the distant part of the governmental and non-governmental organisations in Eastern Himalayan region of India. Forest vegetation is managing terrestrial carbon and global carbon balance mostly dominated by Michelia champaca, Castanopsis in the future course of action. The main objective of this sp., Quercus sp., and Rhododendron sp. at woody layer study is to analyze the soil characteristics and soil carbon Cinnamomum verum, Mahonia nepalensis, Impatiens pool and to develop an appropriate equation for model- sp., and Berberis sp. at the shrub/sapling layer. Ground ling TSC in temperate forests of the Eastern Himalayan vegetation of forest was mainly composed of species like region, India. This will facilitate in predicting TSC by Begonia sp., Rubus ellipticus,and Houttuynia cordata simple mathematical equations using readily available (Yam & Tripathi, 2016). soil characteristics because sometimes it becomes very difficult for a researcher to go through experimenting all the forms of carbon. The findings of the present study will Methodology be very much useful under limited data conditions to Sites were selected in replicates along the altitudinal gra- predict the TSC and also in inaccessible areas of tempe- dients in temperate forest of Ziro valley, and 10 perma- rate forest ecosystem. nent plots of 50 m x 50 m were marked (Map 1). Prior to soil sample collection, the sample sites were cleared of Materials and methods living plants, plant litter, and surface rocks by taking care not to disturb the soil surface or sub-surface. Soils were Study site collected randomly from different points of each of the The present study was carried out in the temperate forest permanent plots in replicates considering three different of Ziro valley at an altitudinal zone of 2300–3000 m depths (0–10, 10–20, and 20–30 cm) for further GEOLOGY, ECOLOGY, AND LANDSCAPES 211 laboratory analysis. The composite soil samples were Table 1. Average of soil characteristics of different soil depths in the study area. prepared for each plot on a monthly basis for the period Upper Middle Lower of two consecutive years (2015–2016). Composite soil Soil variables surface surface surface samples for each depth were brought to the laboratory Total soil carbon (%) 5.33 ± 0.06 5.01 ± 0.09 4.20 ± 0.02 and proceeded with pretreatment by drying, homoge- Soil organic carbon (%) 4.99 ± 0.06 4.67 ± 0.09 3.84 ± 0.02 Soil inorganic carbon (%) 0.29 ± 0.01 0.30 ± 0.01 0.32 ± 0.01 nized by grinding and sieving (<2 mm), and then fol- Microbial biomass carbon (%) 0.05 ± 0.01 0.04 ± 0.01 0.03 ± 0.01 lowed by analysis of physico-chemical and Soil pH 4.88 ± 0.13 4.94 ± 0.11 4.52 ± 0.04 Soil moisture (%) 36.98 ± 2.63 36.49 ± 2.56 30.32 ± 1.85 microbiological characteristics of the soil. The SOC has Sand (%) 71.40 ± 0.32 72.32 ± 0.34 72.14 ± 0.46 been determined by following Walkley and Black (1934), Clay (%) 22.82 ± 0.11 23.00 ± 0.09 22.28 ± 0.18 Silt (%) 5.78 ± 0.28 4.68 ± 0.25 5.58 ± 0.28 and SOM was calculated by multiplying the SOC content Total nitrogen (%) 0.63 ± 0.05 0.65 ± 0.07 0.61 ± 0.05 by 1.724 assuming that soil organic matter contains 58% Total phosphorus (%) 0.20 ± 0.01 0.19 ± 0.01 0.19 ± 0.01 Total potassium (%) 0.23 ± 0.01 0.20 ± 0.01 0.20 ± 0.01 of SOC (Allen, Grimshaw, Parkinson, & Quarmby, Water-holding capacity (%) 54.38 ± 2.45 54.11 ± 1.52 51.15 ± 1.71 1974). Soil inorganic carbon (SIC) was estimated by a Soil temperature (°C) 2.21 ± 0.10 1.77 ± 0.13 2.18 ± 0.12 −3 Bulk density (g cm ) 0.58 ± 0.01 0.59 ± 0.01 0.61 ± 0.01 rapid titration method (Allen et al., 1974). MBC was Organic matter (%) 8.61 ± 0.10 8.04 ± 0.16 6.62 ± 0.03 estimated with field-moist soil by chloroform fumigation incubation method (Jenkinson & Powlson, 1976)asmod- ified slightly by Srivastava and Singh (1988). The TSC was depth. Average soil particle percentage of silt, clay, and determined by combining all the estimated soil carbon sand is found to be in 1:5:11 ratio, respectively, and soil form in the following manner, i.e., TSC stock = SOC + texture was sandy loam to sandy clay loam in nature. SIC + MBC. Total nitrogen was determined following the Soil moisture content was moderate to high and varied distillation method (Allen et al., 1974)and totalphos- between 21.53% and 57.34%. Water-holding capacity phorus using the molybdenum blue method (Allen et al., ranges between 42.78% and 68.57%. Bulk density var- 1974). Total potassium was determined by flame photo- ied from 0.52to0.61gcm . It was observed that the soil meter as outlined by Allen et al. (1974). Using copper cup moisture and water-holding capacity were decreasing having 7 cm internal diameter and 1.2 cm height, the with increase in soil depth. However, the reverse trend water-holding capacity was determined using Keen’sbox was observed with bulk density which increases with method (Allen et al., 1974). Soil pH was determined increase in soil depth. The total nitrogen (N) content in electronically by a digital pH meter in 1:2.5 suspension soil was recorded lowest in the month of July (0.27%) of fresh soil and distilled water (Anderson & Ingram, and highest during April (0.99%). Soil phosphorus (P) 1993). Soil moisture was determined by the gravimetric varied from 0.16% (June) to 0.26% (December), and method and soil temperature by soil thermometer. Bulk total potassium (K) recorded between 0.20% and density (D ) is the dry weight of a known volume of soil 0.37%. Further, no apparent pattern was observed for and was determined using the core method as described NPK along soil depths. SOC was recorded highest by Anderson and Ingram (1993). (5.48%) during June and lowest (3.71%) during Partial least square regression (PLSR) was performed January. The MBC was observed higher (0.07%) during using XLSTAT (2019 ver.) software for prediction and November and lower (0.01%) in the month of July. The data analysis. We have considered different forms of content of SOC and MBC decreases with increase in carbon present in the soil and available soil character- soil depth. In contrast, SIC was found to be increasing istics for modelling the TSC equations for different soil with increase in soil depth and recorded highest depths, i.e., upper, middle, and lower soil surface. For (0.37%) during July in 20–30 cm soil depth and lowest modelling, first, we have performed correlation matrix (0.25%) during August. Since the organic matter was and variable importance in the projections (VIPs) for considered as 1.724 of organic carbon, it followed the all the soil characteristics influencing TSC. Then, the similar trend to that of organic carbon. The average top five VIPs were selected and re-projected for model- total of all the soil characteristics is given in Table 1. ling and developing equations for each layer of soil. Modelling of total soil carbon Results For modelling, first, we have established an overall cor- Soil characteristics relation matrix between dependent and independent variables, and it was observed that the TSC was highly The depth-wise variation in soil characteristics was correlated with SOC followed by pH, clay, and silt. observed for all the studied soil variables. The soil However, it was found that the TSC was least dependent was found to be acidic in nature and pH ranges on other soil variables like sand particle, inorganic car- between 4.28 (November) and 5.60 (March). Soil tem- bon, and soil temperature (Table 2). Although the VIPs perature was recorded to be very low, i.e., 1.2°C during for inorganic carbon showed high value (Figure 1(a)), it January to 2.9°C in September. However, no specific was negatively correlated with total carbon in the pattern was observed in soil temperature across the soil 212 G. YAM ET AL. correlation matrix (Table 2). Accordingly, we have excluded the least contributing independent soil vari- ables and performed PLSR for all the different layers of the soil. Based on different statistical performance and error indices, overall PLSR predictive models show very good 2 2 predictivity (R = 0.99). It was found that Q =0.73 remained lower than 1 for all components; therefore, the quality of fit varied a lot depending on the soil parameters. VIP analysis reflects the relative impor- tance of each of the soil variables in the developed prediction models. Figure 1(a) represents the contribu- tion of different independent variables to that of depen- dent variables and was also found that coefficients are significantly different from zero (Figure 1(b)). Re-run- ning of PLSR of split data for other three different soil surfaces also showed good fitting model which repre- 2 2 2 sented R =0.99, R = 0.99, and R = 0.82 for upper surface, middle surface, and lower surface of the soil, respectively (Table 3). From Figure 2, it was observed that there is fluctuation in the VIPs when there is a decrease or an increase in the soil surface layer. In the upper and middle soil surface, the VIP analysis reflected highest for the organic carbon (Figure 2(a, b)). However, it was silt that represented the highest VIPs (Figure 2(c)). Figure 3 shows good fitting between observed and predicted TSC values as most of the prediction lies between 95% bound limit which reflects descent fitting from developed PLSR-based models. Hence, organic carbon can be considered as one of the most influencing parameters for predicting TSC in the soil depth between 0 and 20 cm. In another case, silt can be considered among the important variable in predicting the TSC content in 20–30 cm soil depth (Figure 2). However, variables such as bulk density, sand particle, soil temperature, and NPK having low VIP score below the threshold (VIP >0.8) were consid- ered to be less important and might be considered as good candidates for exclusion from the model (Figure 1 (a)). In all the three predicted models, the root-mean- square error shows very low value (0.016–0.131) which signifies that the predicted values are very close to observed values and are fit to the model as the distance between the predictions and the observations is low (Figure 3). The developed equations for modelling TSC in three different soil layers are given in Table 3. Discussion The study result showed that the soil is slightly acidic in nature. As soil water-holding capacity is primarily con- trolled by the soil texture and soil organic matter, the soil with sandy clay loam in the upper surface could have attributed to higher water-holding capacity in the upper surface than sandy loam in the lower surface. Hence, larger soil surface area in the upper surface has led to higher water-holding capacity and so the higher moisture Table 2. Correlation matrix for predicting total soil carbon and other variables. Variables OC IC MBC pH Moisture Sand Clay Silt TN TP TK WHC Soil temperature Bulk density TC OC 1.000 −0.422 0.272 0.414 0.354 −0.147 0.465 0.379 0.118 0.200 0.105 0.258 −0.140 0.027 0.998 IC 1.000 −0.318 −0.140 −0.105 0.127 −0.295 −0.152 −0.240 −0.079 −0.177 −0.125 −0.198 −0.180 −0.379 MBC 1.000 −0.175 −0.109 −0.386 −0.089 0.444 −0.180 −0.002 −0.045 0.012 0.147 0.139 0.288 pH 1.000 0.269 0.028 0.198 0.226 0.446 0.001 0.280 0.319 −0.109 0.083 0.407 Moisture 1.000 −0.033 0.134 0.080 0.107 0.124 0.360 0.014 0.132 −0.220 0.352 Sand 1.000 0.580 −0.196 −0.116 −0.015 0.028 0.033 −0.081 −0.025 −0.157 Clay 1.000 0.007 −0.050 0.197 0.086 0.035 −0.280 −0.129 0.449 Silt 1.000 0.151 0.226 0.066 0.296 0.014 0.029 0.393 Total nitrogen 1.000 0.026 −0.044 0.227 −0.126 0.033 0.109 Total phosphorus 1.000 −0.147 −0.039 0.066 0.003 0.198 Total potassium 1.000 0.038 0.276 0.046 0.096 WHC 1.000 −0.473 0.132 0.258 Soil temperature 1.000 −0.007 −0.147 Bulk density 1.000 0.023 TC: total carbon; OC: organic carbon; IC: inorganic carbon; MBC: microbial biomass carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium; WHC: water-holding capacity. GEOLOGY, ECOLOGY, AND LANDSCAPES 213 ab Figure 1. VIPs (95% confidence interval at 5% significant level) for overall soil characteristics. Table 3. Developed predictive models for estimating total soil carbon using Readily available soil characteristics(RASCs). Performance statistics Error statistics Soil depth Sl no. (cm) Developed equations (Eq) R SD MSE RMSE Eq. 1 0–10 TC = 1.146 + 1.038 × OC + 0.008 × pH − 0.001 × SM − 0.046 × Clay + 0.008 × Silt 0.993 0.021 0.001 0.016 Eq. 2 10–20 TC = 5.036 + 0.924 × OC − 0.125 × pH − 0.002 × SM − 0.155 × Clay − 0.019 ×Silt 0.990 0.041 0.001 0.031 Eq. 3 20–30 TC = 6.33 + 1.778 × OC − 0.761 × pH + 0.009 × SM − 0.266 × Clay + 0.202 × Silt 0.818 0.171 0.017 0.131 TC: total soil carbon; OC: organic carbon; SM: soil moisture. a. Upper surface b. Middle surface c. Lower surface Figure 2. VIPs (95% confidence interval at 5% significant level) for different developed models. Eq. 1 Eq. 2 Eq. 3 Figure 3. Predicted versus observed total soil carbon based on different equations (95% confidence interval at 5% significant level). percentage.Soilbulkdensity decreased with depth was in carbon stock could be due to fluctuation in soil bulk agreement with the results reported from the other stu- density. As it was very much addressed by several dies (Hajabbasi, Jalalian, & Karimzadeh, 1997;Sahani& researchers, bulk density has a direct influence on soil Behera, 2001). Temperatures recorded were very low, but carbon (Dar & Sundarapandian, 2013; Li, Wang, Endo, moisture content was moderate to high. It was also Zhao, & Kakubari, 2010). There was also a variation in observed that the soil carbon stock was decreasing with monthly carbon content estimated for two annual cycles increase in soil depth which was also reported by many with lower soil carbon density in the preceding year, researchers (Dar & Sundarapandian, 2013;Krishna, resulted in negative sequestration. This fluctuation in Varghese, & Mohamed, 2012;Vashum, 2016). Higher therangeofsoilcould be duetosurrounding floral carbon in the upper surface may be due to the rapid characteristics, decomposition of plant residues, root decomposition of forest litter by soil microbial organ- exudates, living or dead microorganisms, and soil biota, isms. Another reason behind the fluctuationinsoil which is often related to soil fertility. 214 G. YAM ET AL. Soil fertility generally improves physical (soil aera- But it is very important to observe monthly variation in tion, water retention, resistance to erodibility, etc.) and soil characteristics in order to obtain precise approximate biological properties (build-up of soil microorganisms, values for the prediction and modelling. For the authen- nutrients, etc.), which defines and enhances the produc- ticity of soil, one annual cycle is not enough to validate tive capacity of the soil. On the other hand, the land use the data so we have performed for two annual cycles, and and management system has a strong influence on soil accordingly, we have predictedthreemodels forcalculat- carbon too. Soils that are disturbed or undergoing a ing TSC. Many other models were also developed to transition because of vegetation manipulation have dif- estimate the variables of equations to demonstrate soil ferent organic matter dynamics than soils that are in carbon and other soil properties. These models were equilibrium with biological and environmental condi- produced by statistical methods based on various soil tions (Harrison, Post, & Richter, 1995). Combined properties and parameters as input variables. They are effects of physio-climatic characteristics of the site less difficult and inexpensively obtained than the direct could be another reason for temporal variation of soil measurements. Modeling studies suggest that forest soils carbon and so the soil organic matter in our study sites. are currently sequestering 30–50% of the estimated C Wiesmeier et al. (2019) have stated that vegetation type sink. However, modeled accumulation rates of soil C affects SOC storage by controlling both the input and have so far not been detected in nature (Jandl et al., decomposition of carbon. We have reported a decrease 2007;Liski et al., 2003). But the lack of suitable data in in MBC with increase in depth, whereas Chauhan, terms of soil carbon and other soil characteristics is still a Stewart, and Paul (1981) reported vice versa. Since we problem with respect to arriving at reliable estimates. have carried out our study in a temperate forest prob- But, there is a need for regional approaches to estimate ably, the microclimatic condition might have favoured the SOC storage capacity according to specificclimatic better microbial activity in the upper surface than the conditions and land use/management/vegetation charac- lower surface for enabling faster conversion of decom- teristics (Wiesmeier et al., 2019). There is a need for more posable litter into soil carbon. research on the potential for soil carbon in different parts Many researchers have also cited the importance of of the world for better integration of data to monitor soil microbial biomass (SMB) for playing a dual role in changing climatic conditions through soil carbon. SOM turnover, balancing SOM mineralization, and SOM stabilization processes. This indicates that SOC is Conclusion directly or indirectly associated with SMB. Therefore, the higher the function of SMB, the higher will be the con- Carbon emission to the atmosphere is inevitably increas- version of organic matter. In the present model, we have ing in many regions of the world due to human activities. shown that clay and silt playing another important role Sequestering of carbon helps off-set emissions from fossil with VIPs more than any other soil variables. This could fuel combustion and other carbon-emitting activities be attributed to the major SOM stabilization mechan- while enhancing soil quality. Temperate forest soil has isms; interaction of organic matter(OM) with mineral currently net carbon sink; hence, the present study has surfaces is regarded as quantitatively most important in a been carried out in temperate forest of Lower Subansiri wide range of soils (Sollins, Homann, & Caldwell, 1996; district. The study revealed depth-wise variation in soil von Lützow et al., 2006). This strong correlation of SOC characteristics having very low temperature and acidic in stocks with clay contents was also observed in numerous nature. It was sandy loam to sandy clay loam in nature studies at different spatial scales (Arrouays, Saby, Walter, having moderate-to-high soil moisture. The water-hold- Lemercier, & Schvartz, 2006; Hassink, 1997;Kaiser& ing capacity decreased with increase in soil depths. The Guggenberger, 2000; Zinn, Lal, Bigham, & Resck, soil was rich in soil nutrients mainly due to the quality 2007). Therefore, soil texture is probably one among and quantity of litter and decomposition. As the TSC was the other most promising factors to be used as an indi- highly correlated with SOC, pH, clay, and silt, these cator for SOC storage over a wider range of scales variables were used for modelling purposes. Based on (Wiesmeier et al., 2019). Temperate zones are considered the above variables, depth-wise equations were devel- as the regions of the world most uniformly and exten- oped using PLSR which resulted in good fitting with sively altered by humans and have had remarkable observed data. As the developed PLSR-based models impacts on global climatic change. In this study, we hadgood predictive capabilitybasedonidentified pre- have recorded an increase in potential sequestration of dictors, it is being endorsed that it might be incorporated temperate forest soil. Liski et al. (2003)have also into analyzing TSC in various land cover. Hence, the observed increased carbon sink in temperate forest. inclusion of the above variables as a predictor is being The comparison of soil carbon sequestration potential recommended for further studies. This will facilitate to on a yearly basis with the other study area is very difficult conservation biologists in predicting TSC by simple as the literatures are very much lacking in such studies; mathematical equations using readily available soil char- however, data on carbon stock densities have been acteristics in remote and inaccessible sites of the tempe- reported by many researchers in the other study sites. rate forest ecosystem and precise carbon budget. Further, GEOLOGY, ECOLOGY, AND LANDSCAPES 215 not much data are available on related study; hence, it Hajabbasi, M. A., Jalalian, A., & Karimzadeh, H. R. (1997). Deforestation effects on soil physical and chemical prop- requires further in-depth research to come up with more erties, Lordegan, Iran. Plant and Soil, 190(2), 301–308. appropriate models in extreme habitats. Findings can Harrison, K. G., Post, W. M., & Richter, D. D. (1995). Soil also play a significant role in spatial mapping of carbon carbon turnover in a recovering temperate forest. Global pool and sequestration for mitigating climate change. Biogeochemical Cycles, 9(4), 449–454. Hassink, J. (1997). The capacity of soils to preserve organic C and N by their association with clay and silt particles. Acknowledgments Plant and Soil, 191(1), 77–87. Houghton, J. T., Ding, Y. D. J. G., Griggs, D. J., Noguer, M., The authors are thankful to the local people (Apatani) for van der Linden, P. J., Dai, X., & Johnson, C. A. (2001). their helping hand and support to carry out the study. Yam Climate change 2001: The scientific basis. The Press is also thankful to Director, NERIST, for permission to carry Syndicate of the University of Cambridge, Cambridge. out soil analysis and University Grant Commission, New Jandl, R., Lindner, M., Vesterdal, L., Bauwens, B., Baritz, R., Delhi, for Dr D. S. Kothari award. Hagedorn, F., & Byrne, K. A. (2007). How strongly can forest management influence soil carbon sequestration? Geoderma, 137(3–4), 253–268. Disclosure statement Jenkinson, D. S., & Powlson, D. S. (1976). The effect of biocidal treatments on metabolism in soils: V. A method of measur- No potential conflict of interest was reported by the authors. ing soil biomass. Soil Biology and Biochemistry, 8,208–213. Jobbágy, E. G., & Jackson, R. B. (2000). The vertical distri- bution of soil organic carbon and its relation to climate References and vegetation. Ecological Applications, 10(2), 423–436. Kaiser, K., & Guggenberger, G. (2000). The role of DOM Allen, S. E., Grimshaw, H. M., Parkinson, J. A., & Quarmby, sorption to mineral surfaces in the preservation of organic C. (1974). Chemical analysis of ecological materials. matter in soils. Organic Geochemistry, 31(7–8), 711–725. Oxford, England:Blackwell Scientific Publications. Krishna, M. P., Varghese, R., & Mohamed, A. A. (2012). Anderson, J. M., & Ingram, J. S. I. (1993). Tropical soil Depth wise variation of microbial load in the soils of biology and fertility. A handbook of methods (2nd ed). midland region of Kerala: A function of important soil Oxfordshire, Wallingford, UK: CAB, International. physicochemical characteristics and nutrients. Indian Arrouays, D., Saby, N., Walter, C., Lemercier, B., & Journal of Education and Information Management, 1 Schvartz, C. (2006). Relationships between particle-size (3), 126–129. distribution and organic carbon in French arable topsoils. Lal, R. (2004). Soil carbon sequestration impacts on global Soil Use and Management, 22(1), 48–51. climate change and food security. Science, 304(5677), Batjes, N. H. (2014). Total carbon and nitrogen in the soils of 1623–1627. the world. European Journal of Soil Science, 65(1), 10–21. Lal, R. (2005). Forest soils and carbon sequestration. Forest Bhatler, R. A. (2009). Temperate forests store more carbon Ecology and Management, 220(1–3), 242–258. than tropical forests, finds study. https://news.mongabay. Li, P., Wang, Q., Endo, T., Zhao, X., & Kakubari, Y. (2010). com/2009/07 Soil organic carbon stock is closely related to above- Bordoloi, R., Das, B., Yam, G., Pandey, P. K., & Tripathi, O. ground vegetation properties in cold-temperate mountai- P. (2018). Modeling of water holding capacity using read- nous forests. Geoderma, 154(3–4), 407–415. ily available soil characteristics. Agricultural Research,1–9. Liski, J.,Korotkov,A. V., Prins,C.F.,Karjalainen, T.,Victor, D. DOI: 10.1007/s40003-018-0376-9 G.,&Kauppi,P.E.(2003). Increasedcarbonsinkintempe- Brown, S., & Gaston, G. (1995). Use of forest inventories rate and boreal forests. Climatic Change, 61(1–2), 89–99. and geographic information systems to estimate biomass Malhi, Y., Wood, D., Baker, T. R., Wright, J., Phillips, O. L., density of tropical forests: Application to tropical Africa. Cochrane, T., . . . Higuchi, N. (2006). The regional variation In:Fitzgerald J.F., Braatz B.V., Brown S., Isichei A.O., of aboveground live biomass in old-growth Amazonian Odada E.O., Scholes R.J. (eds) African greenhouse gas forests. Global Change Biology, 12(7), 1107–1138. emission inventories and mitigation options: Forestry, Martin, P. H., Nabuurs, G. J., Aubinet, M., Karjalainen, T., land-use change, and agriculture (pp. 51–62). Dordrecht: Vine, E. L., Kinsman, J., & Heath, L. S. (2001). Carbon Springer. sinks in temperate forests. Annual Review of Energy and Chauhan, B. S., Stewart, J. W. B., & Paul, E. A. (1981). Effect the Environment, 26(1), 435–465. of labile inorganic phosphate status and organic carbon Reich, P. B., & Bolstad, P. (2001). Productivity of evergreen additions on the microbial uptake of phosphorus in soils. and deciduous temperate forests (pp. 245–283). Terrestrial Canadian Journal of Soil Science, 61(2), 373–385. global productivity. Academic, San Diego. Chave, J., Olivier, J., Bongers, F., Châtelet, P., Forget, P. M., Robinson, D. (2007). Implications of a large global root van der Meer, P., & Charles-Dominique, P. (2008). biomass for carbon sink estimates and for soil carbon Above-ground biomass and productivity in a rain forest dynamics. Proceedings of the Royal Society B: Biological of eastern South America. Journal of Tropical Ecology, 24 Sciences, 274(1626), 2753–2759. (4), 355–366. Saatchi, S. S., Houghton, R. A., Dos Santos Alvala, R. C., Dar, J. A., & Sundarapandian, S. M. (2013). Soil organic Soares, J. V., & Yu, Y. (2007). Distribution of above- carbon stock assessment in two temperate forest types of ground live biomass in the Amazon basin. Global Western Himalaya of Jammu and Kashmir. India Forest Change Biology, 13(4), 816–837. Research Institute, 3(114), 2. Sahani, U., & Behera, N. (2001). Impact of deforestation on Gower, S. T., Landsberg, J. J., & Bisbee, R. L. (2003). Forest soil physicochemical characteristics, microbial biomass biomes of theworld.InR. A.Young &R.L.Giese (Eds.), and microbial activity of tropical soil. Land Degradation Forest ecosystem science and management. PP: 57-74. & Development, 12(2), 93–105. Hoboken: Wiley. 216 G. YAM ET AL. Six, J.,Callewaert,P.,Lenders,S., DeGryze, S., Morris,S.J., von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, Gregorich, E. G., . . . Paustian, K. (2002). Measuring and E., Guggenberger, G., Marschner, B., & Flessa, H. (2006). understanding carbon storage in afforested soils by physical Stabilization of organic matter in temperate soils: fractionation. Soil Science Society of American Journal, 66(6), Mechanisms and their relevance under different soil con- 1981–1987. ditions—A review. European Journal of Soil Science, 57 Sollins, P., Homann, P., & Caldwell, B. A. (1996). (4), 426–445. Stabilization and destabilization of soil organic matter: Walkley, A., & Black, I. A. (1934). An examination of the Mechanisms and controls. Geoderma, 74(1–2), 65–105. Degtjareff method for determining organic carbon in Srivastava, S. C., & Singh, J. S. (1988). Carbon and phos- soils: Effect of variations in digestion conditions and of phorus in the soil biomass of some tropical soils of India. inorganic soil constituents. Soil Science, 63, 251–263. Soil Biology and Biochemistry, 20(5), 743–747. Wiesmeier, M., Urbanski, L., Hobley, E., Lang, B., von Terakunpisut, J., Gajaseni, N., & Ruankawe, N. (2007). Luetzow, M., Marin-Spiotta, E., . . . Wollschlaeger, U. Carbon sequestration potential in aboveground biomass (2019). Soil organic carbon storage as a key function of of Thong Pha Phum National Forest, Thailand. Applied soils-a review of drivers and indicators at various scales. Ecology and Environmental Research, 5(2), 93–102. Geoderma, 333, 149–162. United Nations Framework Convention on Climate Change Yam, G., & Tripathi, O. P. (2015). Floristic diversity, com- (UNFCC). (1997). Kyoto Protocol, Kyoto (pp. 19). munity characteristics and above biomass of woody spe- UNFCC, United Kingdom cies in subtropical forests of Ziro valley, Arunachal Valladares, F. (2017). A mechanistic view of the capacity of Pradesh. Indian Forester, 141(12), 1269–1277. forests to cope with climate change. In Managing forest Yam, G., & Tripathi, O. P. (2016). Tree diversity and com- ecosystems: The challenge of climate change (pp. 15–40). munity characteristics in Talle Wildlife Sanctuary, Cham: Springer. Kluwer Academic Publishers. Arunachal Pradesh, Eastern Himalaya, India. Journal of Netherlands. Asia-Pacific Biodiversity, 9(2), 160–165. Vashum, K. T. (2016). Soil organic carbon sequestration Zinn, Y. L., Lal, R., Bigham, J. M., & Resck, D. V. (2007). potential of primary and secondary forests in Northeast Edaphic controls on soil organic carbon retention in the India. Proceedings of the International Academy of Brazilian Cerrado: Texture and mineralogy. Soil Science Ecology and Environmental Sciences, 6(3), 67. Society of America Journal, 71(4), 1204–1214. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Geology Ecology and Landscapes Taylor & Francis

Modelling of total soil carbon using readily available soil variables in temperate forest of Eastern Himalaya, Northeast India

Download PDF
8 pages

Loading next page...
 
/lp/taylor-francis/modelling-of-total-soil-carbon-using-readily-available-soil-variables-UHP0030ign

References (52)

Publisher
Taylor & Francis
Copyright
© 2021 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON).
ISSN
2474-9508
DOI
10.1080/24749508.2019.1706295
Publisher site
See Article on Publisher Site

Abstract

GEOLOGY, ECOLOGY, AND LANDSCAPES 2021, VOL. 5, NO. 3, 209–216 INWASCON https://doi.org/10.1080/24749508.2019.1706295 RESEARCH ARTICLE Modelling of total soil carbon using readily available soil variables in temperate forest of Eastern Himalaya, Northeast India a b a Gyati Yam , Om Prakash Tripathi and Debangshu Narayan Das a b Department of Zoology, Rajiv Gandhi University, Doimukh, India; Department of Forestry, North Eastern Regional Institute of Science and Technology, Nrjuli, India ABSTRACT ARTICLE HISTORY Received 12 October 2019 The study was carried out in temperate forest of Ziro valley, Arunachal Pradesh, Northeast India, Accepted 12 December 2019 during 2015–2016. Stratified random sampling was adopted for soil sample collection from three depths on a monthly basis in each of the permanent plots. Collected samples were analysed using KEYWORDS standard methodologies. The XLSTAT (ver.2019) was used for the partial least square regression and Microbial biomass carbon; modelling. The soil was acidic in nature having moderate soil moisture with very low temperature. PSLR; soil properties; VIPs; Soil texture varies from sandy loam to sandy clay loam in nature. Altogether, 16 readily available soil XLSTAT variables were used for identifying significant variables to be used for modelling the total soil carbon (TSC). Based on the results, five most contributing independent soil variables having high variable importance in prediction were used for modelling the TSC. The average annual TSC was recorded 5.33% for upper, 5.01% for middle, and 4.20% for lower soil surface, respectively. The developed depth-wise equations predict very close TSC with very low root-mean-square error (rmse) to observed values. Hence, the findings of the present study will be very much useful under limited data conditions to predict the TSC and also in inaccessible areas of temperate forest ecosystem. Introduction the ocean (Batjes, 2014). Soil and vegetation have very complex interrelation, and species composition has sig- Temperate forest being one of the major terrestrial nificant impacts on soil properties over a long period of biomes,soils arehighlyvariablebut generally fertilefor time (Bordoloi, Das, Yam, Pandey, & Tripathi, 2018). which soil quality is the key essential for reflecting the The concentration of elements in the soils is a good structure of vegetation (Gower, Landsberg, & Bisbee, indicator of their availability to plants. Nonetheless, for- 2003;Martinetal., 2001). Alternates between warm est soil plays a vital role in climate change mitigation by summers and markedly cool or cold winters which is restricting the direct release of carbon into the atmo- characterized by marked seasonality are the climate of sphere. Terakunpisut, Gajaseni, and Ruankawe (2007) temperate forests (Reich & Bolstad, 2001;Yam & stated that forests are the prominent sites to study climate Tripathi, 2016). This ecosystem plays a fundamental change in termsofnet carbon emissionsaswellas in role in the global carbon (C) cycle, regardless of the global storage capacity which is important for climate way that the biome C pool is the smallest among other regulation. As we know, the soil is particularly concerned biomes (Reich & Bolstad, 2001;Robinson, 2007). Bhatler with the cycling of nutrients which is highly governed by (2009) had reported that temperate forests store more the interactions with its surrounding and vegetation. But carbon than tropical forests. Different strategies for mea- now as a decisive aspect, the multifunctionality of soils is suring C reveal that temperate forests have presently net increasingly recognized in global land management Csinks (Martinetal., 2001). It was reported that 31% (Wiesmeier et al., 2019). and 41% of the soil organic carbon (SOC) pool are stored Carbon emission to the atmosphere is inevitably in 1–3 m depth in temperate evergreen and temperate increasing in many regions of the world due to varied deciduous forests, respectively (Jobbágy & Jackson, human activities. Various factors such as fossil fuel com- 2000). Thus, it was estimated that about 262 Pg C is bustion, deforestation, overpopulation, industrial emis- stored in temperate forest soils in the top 3 m depth. sion, land-use change, urbanization, etc., directly or Soil carbon is directly or indirectly proportional to indirectly emitting greenhouse gases causing global climate change and vice versa. Soil is considered to be the warming. But human being is considered to be the most highest sequester of atmospheric carbon as soil carbon. important one for deteriorating climate through various Potentially, the soil–vegetation carbon pool is much anthropogenic activities shaping their own kind of chan- more labile in the short term as compared with that of ged environment. The increase in atmospheric carbon the oceans although the soil carbon pool is smaller than dioxide (CO ) concentrations associated with the CONTACT Om Prakash Tripathi tripathiom7@gmail.com Department of Forestry, North Eastern Regional Institute of Science and Technology, Nrjuli 791109, India © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 210 G. YAM ET AL. progression of the anthropogenic impact on forest eco- systems from the molecule to the ecosystem level (Valladares, 2017). Sequestering of carbon serves as to compensate emissions from fossilfuelcombustionand other carbon-emitting activities while enhancing soil quality. An increase in the SOC pool is also essential to advancing global food security through maximising crop productivity (Lal, 2004). However, nutrient loss, soil ero- sion, soil conservation, desertification, organic matter decomposition, etc., may alter the soil dynamics. Soil carbon is also considered to be the largest pool of terres- trial carbon. Forest ecosystems store more than 80% of all terrestrial aboveground C and more than 70% of all soil organic C (Batjes, 2014; Jandl et al., 2007;Jobbágy & Jackson, 2000;Six et al., 2002). The soil carbon mainly stored in the form of SOC, microbial biomass carbon (MBC), and soil organic matter. Map 1. Study area (Ziro valley) in Arunachal Pradesh, India. Soil carbon is one of the important variables for deter- miningthefuturecarbonsink. Theimportanceofforest and soil C in mitigating and balancing the greenhouse above mean sea level (asl) during 2015–2016 in lower effect has been recognized, and an agreement was reached Subansiri district of Arunachal Pradesh, India (Map 1). under theKyotoProtocoltoinclude forest andsoilC Analysis of soil was done in Forest Ecology Laboratory of sequestration in the list of acceptable offsets (United North Eastern Regional Institute of Science and Nations Framework Convention on Climate Change Technology. It covers an area of 3460 km which is [UNFCCC], 1997). Lal (2005) had also stated that forest mostly mountainous terrain. The forests are rich in valu- soil carbon sequestration has a potential to decrease the able species of trees, shrubs, and different varieties of rate of enrichment of atmospheric CO . Many research- herbs. Cane and bamboo and exquisite floral treasures ers have enumerated the amount of carbon sequestered of wilder species may often be found in the shadowy in the standing trees (Brown & Gaston, 1995;Chave etal., recesses of the forest. The area mainly comprises sub- 2008;Houghtonetal., 2001; Malhi et al., 2006;Saatchi, tropical and alpine forests and has a variety of flora and Houghton, Dos Santos Alvala, Soares, & Yu, 2007;Yam& fauna, many of which are endangered. However, the Tripathi, 2015), but only a few studies have carried out on experiment was mainly carried out in the forests which soil carbon sequestration. Besides, for understanding the are considered as untouched and virgin natural tempe- soil dynamics and potential soil carbon sequestration, rate forest well protected by the “Apatani” people of this modelling of total soil carbon (TSC) may be an important place since time immemorial. The Ziro valley is a quaint attribute for the scientific researchers, stakeholders, and eastern Himalayan town located in the distant part of the governmental and non-governmental organisations in Eastern Himalayan region of India. Forest vegetation is managing terrestrial carbon and global carbon balance mostly dominated by Michelia champaca, Castanopsis in the future course of action. The main objective of this sp., Quercus sp., and Rhododendron sp. at woody layer study is to analyze the soil characteristics and soil carbon Cinnamomum verum, Mahonia nepalensis, Impatiens pool and to develop an appropriate equation for model- sp., and Berberis sp. at the shrub/sapling layer. Ground ling TSC in temperate forests of the Eastern Himalayan vegetation of forest was mainly composed of species like region, India. This will facilitate in predicting TSC by Begonia sp., Rubus ellipticus,and Houttuynia cordata simple mathematical equations using readily available (Yam & Tripathi, 2016). soil characteristics because sometimes it becomes very difficult for a researcher to go through experimenting all the forms of carbon. The findings of the present study will Methodology be very much useful under limited data conditions to Sites were selected in replicates along the altitudinal gra- predict the TSC and also in inaccessible areas of tempe- dients in temperate forest of Ziro valley, and 10 perma- rate forest ecosystem. nent plots of 50 m x 50 m were marked (Map 1). Prior to soil sample collection, the sample sites were cleared of Materials and methods living plants, plant litter, and surface rocks by taking care not to disturb the soil surface or sub-surface. Soils were Study site collected randomly from different points of each of the The present study was carried out in the temperate forest permanent plots in replicates considering three different of Ziro valley at an altitudinal zone of 2300–3000 m depths (0–10, 10–20, and 20–30 cm) for further GEOLOGY, ECOLOGY, AND LANDSCAPES 211 laboratory analysis. The composite soil samples were Table 1. Average of soil characteristics of different soil depths in the study area. prepared for each plot on a monthly basis for the period Upper Middle Lower of two consecutive years (2015–2016). Composite soil Soil variables surface surface surface samples for each depth were brought to the laboratory Total soil carbon (%) 5.33 ± 0.06 5.01 ± 0.09 4.20 ± 0.02 and proceeded with pretreatment by drying, homoge- Soil organic carbon (%) 4.99 ± 0.06 4.67 ± 0.09 3.84 ± 0.02 Soil inorganic carbon (%) 0.29 ± 0.01 0.30 ± 0.01 0.32 ± 0.01 nized by grinding and sieving (<2 mm), and then fol- Microbial biomass carbon (%) 0.05 ± 0.01 0.04 ± 0.01 0.03 ± 0.01 lowed by analysis of physico-chemical and Soil pH 4.88 ± 0.13 4.94 ± 0.11 4.52 ± 0.04 Soil moisture (%) 36.98 ± 2.63 36.49 ± 2.56 30.32 ± 1.85 microbiological characteristics of the soil. The SOC has Sand (%) 71.40 ± 0.32 72.32 ± 0.34 72.14 ± 0.46 been determined by following Walkley and Black (1934), Clay (%) 22.82 ± 0.11 23.00 ± 0.09 22.28 ± 0.18 Silt (%) 5.78 ± 0.28 4.68 ± 0.25 5.58 ± 0.28 and SOM was calculated by multiplying the SOC content Total nitrogen (%) 0.63 ± 0.05 0.65 ± 0.07 0.61 ± 0.05 by 1.724 assuming that soil organic matter contains 58% Total phosphorus (%) 0.20 ± 0.01 0.19 ± 0.01 0.19 ± 0.01 Total potassium (%) 0.23 ± 0.01 0.20 ± 0.01 0.20 ± 0.01 of SOC (Allen, Grimshaw, Parkinson, & Quarmby, Water-holding capacity (%) 54.38 ± 2.45 54.11 ± 1.52 51.15 ± 1.71 1974). Soil inorganic carbon (SIC) was estimated by a Soil temperature (°C) 2.21 ± 0.10 1.77 ± 0.13 2.18 ± 0.12 −3 Bulk density (g cm ) 0.58 ± 0.01 0.59 ± 0.01 0.61 ± 0.01 rapid titration method (Allen et al., 1974). MBC was Organic matter (%) 8.61 ± 0.10 8.04 ± 0.16 6.62 ± 0.03 estimated with field-moist soil by chloroform fumigation incubation method (Jenkinson & Powlson, 1976)asmod- ified slightly by Srivastava and Singh (1988). The TSC was depth. Average soil particle percentage of silt, clay, and determined by combining all the estimated soil carbon sand is found to be in 1:5:11 ratio, respectively, and soil form in the following manner, i.e., TSC stock = SOC + texture was sandy loam to sandy clay loam in nature. SIC + MBC. Total nitrogen was determined following the Soil moisture content was moderate to high and varied distillation method (Allen et al., 1974)and totalphos- between 21.53% and 57.34%. Water-holding capacity phorus using the molybdenum blue method (Allen et al., ranges between 42.78% and 68.57%. Bulk density var- 1974). Total potassium was determined by flame photo- ied from 0.52to0.61gcm . It was observed that the soil meter as outlined by Allen et al. (1974). Using copper cup moisture and water-holding capacity were decreasing having 7 cm internal diameter and 1.2 cm height, the with increase in soil depth. However, the reverse trend water-holding capacity was determined using Keen’sbox was observed with bulk density which increases with method (Allen et al., 1974). Soil pH was determined increase in soil depth. The total nitrogen (N) content in electronically by a digital pH meter in 1:2.5 suspension soil was recorded lowest in the month of July (0.27%) of fresh soil and distilled water (Anderson & Ingram, and highest during April (0.99%). Soil phosphorus (P) 1993). Soil moisture was determined by the gravimetric varied from 0.16% (June) to 0.26% (December), and method and soil temperature by soil thermometer. Bulk total potassium (K) recorded between 0.20% and density (D ) is the dry weight of a known volume of soil 0.37%. Further, no apparent pattern was observed for and was determined using the core method as described NPK along soil depths. SOC was recorded highest by Anderson and Ingram (1993). (5.48%) during June and lowest (3.71%) during Partial least square regression (PLSR) was performed January. The MBC was observed higher (0.07%) during using XLSTAT (2019 ver.) software for prediction and November and lower (0.01%) in the month of July. The data analysis. We have considered different forms of content of SOC and MBC decreases with increase in carbon present in the soil and available soil character- soil depth. In contrast, SIC was found to be increasing istics for modelling the TSC equations for different soil with increase in soil depth and recorded highest depths, i.e., upper, middle, and lower soil surface. For (0.37%) during July in 20–30 cm soil depth and lowest modelling, first, we have performed correlation matrix (0.25%) during August. Since the organic matter was and variable importance in the projections (VIPs) for considered as 1.724 of organic carbon, it followed the all the soil characteristics influencing TSC. Then, the similar trend to that of organic carbon. The average top five VIPs were selected and re-projected for model- total of all the soil characteristics is given in Table 1. ling and developing equations for each layer of soil. Modelling of total soil carbon Results For modelling, first, we have established an overall cor- Soil characteristics relation matrix between dependent and independent variables, and it was observed that the TSC was highly The depth-wise variation in soil characteristics was correlated with SOC followed by pH, clay, and silt. observed for all the studied soil variables. The soil However, it was found that the TSC was least dependent was found to be acidic in nature and pH ranges on other soil variables like sand particle, inorganic car- between 4.28 (November) and 5.60 (March). Soil tem- bon, and soil temperature (Table 2). Although the VIPs perature was recorded to be very low, i.e., 1.2°C during for inorganic carbon showed high value (Figure 1(a)), it January to 2.9°C in September. However, no specific was negatively correlated with total carbon in the pattern was observed in soil temperature across the soil 212 G. YAM ET AL. correlation matrix (Table 2). Accordingly, we have excluded the least contributing independent soil vari- ables and performed PLSR for all the different layers of the soil. Based on different statistical performance and error indices, overall PLSR predictive models show very good 2 2 predictivity (R = 0.99). It was found that Q =0.73 remained lower than 1 for all components; therefore, the quality of fit varied a lot depending on the soil parameters. VIP analysis reflects the relative impor- tance of each of the soil variables in the developed prediction models. Figure 1(a) represents the contribu- tion of different independent variables to that of depen- dent variables and was also found that coefficients are significantly different from zero (Figure 1(b)). Re-run- ning of PLSR of split data for other three different soil surfaces also showed good fitting model which repre- 2 2 2 sented R =0.99, R = 0.99, and R = 0.82 for upper surface, middle surface, and lower surface of the soil, respectively (Table 3). From Figure 2, it was observed that there is fluctuation in the VIPs when there is a decrease or an increase in the soil surface layer. In the upper and middle soil surface, the VIP analysis reflected highest for the organic carbon (Figure 2(a, b)). However, it was silt that represented the highest VIPs (Figure 2(c)). Figure 3 shows good fitting between observed and predicted TSC values as most of the prediction lies between 95% bound limit which reflects descent fitting from developed PLSR-based models. Hence, organic carbon can be considered as one of the most influencing parameters for predicting TSC in the soil depth between 0 and 20 cm. In another case, silt can be considered among the important variable in predicting the TSC content in 20–30 cm soil depth (Figure 2). However, variables such as bulk density, sand particle, soil temperature, and NPK having low VIP score below the threshold (VIP >0.8) were consid- ered to be less important and might be considered as good candidates for exclusion from the model (Figure 1 (a)). In all the three predicted models, the root-mean- square error shows very low value (0.016–0.131) which signifies that the predicted values are very close to observed values and are fit to the model as the distance between the predictions and the observations is low (Figure 3). The developed equations for modelling TSC in three different soil layers are given in Table 3. Discussion The study result showed that the soil is slightly acidic in nature. As soil water-holding capacity is primarily con- trolled by the soil texture and soil organic matter, the soil with sandy clay loam in the upper surface could have attributed to higher water-holding capacity in the upper surface than sandy loam in the lower surface. Hence, larger soil surface area in the upper surface has led to higher water-holding capacity and so the higher moisture Table 2. Correlation matrix for predicting total soil carbon and other variables. Variables OC IC MBC pH Moisture Sand Clay Silt TN TP TK WHC Soil temperature Bulk density TC OC 1.000 −0.422 0.272 0.414 0.354 −0.147 0.465 0.379 0.118 0.200 0.105 0.258 −0.140 0.027 0.998 IC 1.000 −0.318 −0.140 −0.105 0.127 −0.295 −0.152 −0.240 −0.079 −0.177 −0.125 −0.198 −0.180 −0.379 MBC 1.000 −0.175 −0.109 −0.386 −0.089 0.444 −0.180 −0.002 −0.045 0.012 0.147 0.139 0.288 pH 1.000 0.269 0.028 0.198 0.226 0.446 0.001 0.280 0.319 −0.109 0.083 0.407 Moisture 1.000 −0.033 0.134 0.080 0.107 0.124 0.360 0.014 0.132 −0.220 0.352 Sand 1.000 0.580 −0.196 −0.116 −0.015 0.028 0.033 −0.081 −0.025 −0.157 Clay 1.000 0.007 −0.050 0.197 0.086 0.035 −0.280 −0.129 0.449 Silt 1.000 0.151 0.226 0.066 0.296 0.014 0.029 0.393 Total nitrogen 1.000 0.026 −0.044 0.227 −0.126 0.033 0.109 Total phosphorus 1.000 −0.147 −0.039 0.066 0.003 0.198 Total potassium 1.000 0.038 0.276 0.046 0.096 WHC 1.000 −0.473 0.132 0.258 Soil temperature 1.000 −0.007 −0.147 Bulk density 1.000 0.023 TC: total carbon; OC: organic carbon; IC: inorganic carbon; MBC: microbial biomass carbon; TN: total nitrogen; TP: total phosphorus; TK: total potassium; WHC: water-holding capacity. GEOLOGY, ECOLOGY, AND LANDSCAPES 213 ab Figure 1. VIPs (95% confidence interval at 5% significant level) for overall soil characteristics. Table 3. Developed predictive models for estimating total soil carbon using Readily available soil characteristics(RASCs). Performance statistics Error statistics Soil depth Sl no. (cm) Developed equations (Eq) R SD MSE RMSE Eq. 1 0–10 TC = 1.146 + 1.038 × OC + 0.008 × pH − 0.001 × SM − 0.046 × Clay + 0.008 × Silt 0.993 0.021 0.001 0.016 Eq. 2 10–20 TC = 5.036 + 0.924 × OC − 0.125 × pH − 0.002 × SM − 0.155 × Clay − 0.019 ×Silt 0.990 0.041 0.001 0.031 Eq. 3 20–30 TC = 6.33 + 1.778 × OC − 0.761 × pH + 0.009 × SM − 0.266 × Clay + 0.202 × Silt 0.818 0.171 0.017 0.131 TC: total soil carbon; OC: organic carbon; SM: soil moisture. a. Upper surface b. Middle surface c. Lower surface Figure 2. VIPs (95% confidence interval at 5% significant level) for different developed models. Eq. 1 Eq. 2 Eq. 3 Figure 3. Predicted versus observed total soil carbon based on different equations (95% confidence interval at 5% significant level). percentage.Soilbulkdensity decreased with depth was in carbon stock could be due to fluctuation in soil bulk agreement with the results reported from the other stu- density. As it was very much addressed by several dies (Hajabbasi, Jalalian, & Karimzadeh, 1997;Sahani& researchers, bulk density has a direct influence on soil Behera, 2001). Temperatures recorded were very low, but carbon (Dar & Sundarapandian, 2013; Li, Wang, Endo, moisture content was moderate to high. It was also Zhao, & Kakubari, 2010). There was also a variation in observed that the soil carbon stock was decreasing with monthly carbon content estimated for two annual cycles increase in soil depth which was also reported by many with lower soil carbon density in the preceding year, researchers (Dar & Sundarapandian, 2013;Krishna, resulted in negative sequestration. This fluctuation in Varghese, & Mohamed, 2012;Vashum, 2016). Higher therangeofsoilcould be duetosurrounding floral carbon in the upper surface may be due to the rapid characteristics, decomposition of plant residues, root decomposition of forest litter by soil microbial organ- exudates, living or dead microorganisms, and soil biota, isms. Another reason behind the fluctuationinsoil which is often related to soil fertility. 214 G. YAM ET AL. Soil fertility generally improves physical (soil aera- But it is very important to observe monthly variation in tion, water retention, resistance to erodibility, etc.) and soil characteristics in order to obtain precise approximate biological properties (build-up of soil microorganisms, values for the prediction and modelling. For the authen- nutrients, etc.), which defines and enhances the produc- ticity of soil, one annual cycle is not enough to validate tive capacity of the soil. On the other hand, the land use the data so we have performed for two annual cycles, and and management system has a strong influence on soil accordingly, we have predictedthreemodels forcalculat- carbon too. Soils that are disturbed or undergoing a ing TSC. Many other models were also developed to transition because of vegetation manipulation have dif- estimate the variables of equations to demonstrate soil ferent organic matter dynamics than soils that are in carbon and other soil properties. These models were equilibrium with biological and environmental condi- produced by statistical methods based on various soil tions (Harrison, Post, & Richter, 1995). Combined properties and parameters as input variables. They are effects of physio-climatic characteristics of the site less difficult and inexpensively obtained than the direct could be another reason for temporal variation of soil measurements. Modeling studies suggest that forest soils carbon and so the soil organic matter in our study sites. are currently sequestering 30–50% of the estimated C Wiesmeier et al. (2019) have stated that vegetation type sink. However, modeled accumulation rates of soil C affects SOC storage by controlling both the input and have so far not been detected in nature (Jandl et al., decomposition of carbon. We have reported a decrease 2007;Liski et al., 2003). But the lack of suitable data in in MBC with increase in depth, whereas Chauhan, terms of soil carbon and other soil characteristics is still a Stewart, and Paul (1981) reported vice versa. Since we problem with respect to arriving at reliable estimates. have carried out our study in a temperate forest prob- But, there is a need for regional approaches to estimate ably, the microclimatic condition might have favoured the SOC storage capacity according to specificclimatic better microbial activity in the upper surface than the conditions and land use/management/vegetation charac- lower surface for enabling faster conversion of decom- teristics (Wiesmeier et al., 2019). There is a need for more posable litter into soil carbon. research on the potential for soil carbon in different parts Many researchers have also cited the importance of of the world for better integration of data to monitor soil microbial biomass (SMB) for playing a dual role in changing climatic conditions through soil carbon. SOM turnover, balancing SOM mineralization, and SOM stabilization processes. This indicates that SOC is Conclusion directly or indirectly associated with SMB. Therefore, the higher the function of SMB, the higher will be the con- Carbon emission to the atmosphere is inevitably increas- version of organic matter. In the present model, we have ing in many regions of the world due to human activities. shown that clay and silt playing another important role Sequestering of carbon helps off-set emissions from fossil with VIPs more than any other soil variables. This could fuel combustion and other carbon-emitting activities be attributed to the major SOM stabilization mechan- while enhancing soil quality. Temperate forest soil has isms; interaction of organic matter(OM) with mineral currently net carbon sink; hence, the present study has surfaces is regarded as quantitatively most important in a been carried out in temperate forest of Lower Subansiri wide range of soils (Sollins, Homann, & Caldwell, 1996; district. The study revealed depth-wise variation in soil von Lützow et al., 2006). This strong correlation of SOC characteristics having very low temperature and acidic in stocks with clay contents was also observed in numerous nature. It was sandy loam to sandy clay loam in nature studies at different spatial scales (Arrouays, Saby, Walter, having moderate-to-high soil moisture. The water-hold- Lemercier, & Schvartz, 2006; Hassink, 1997;Kaiser& ing capacity decreased with increase in soil depths. The Guggenberger, 2000; Zinn, Lal, Bigham, & Resck, soil was rich in soil nutrients mainly due to the quality 2007). Therefore, soil texture is probably one among and quantity of litter and decomposition. As the TSC was the other most promising factors to be used as an indi- highly correlated with SOC, pH, clay, and silt, these cator for SOC storage over a wider range of scales variables were used for modelling purposes. Based on (Wiesmeier et al., 2019). Temperate zones are considered the above variables, depth-wise equations were devel- as the regions of the world most uniformly and exten- oped using PLSR which resulted in good fitting with sively altered by humans and have had remarkable observed data. As the developed PLSR-based models impacts on global climatic change. In this study, we hadgood predictive capabilitybasedonidentified pre- have recorded an increase in potential sequestration of dictors, it is being endorsed that it might be incorporated temperate forest soil. Liski et al. (2003)have also into analyzing TSC in various land cover. Hence, the observed increased carbon sink in temperate forest. inclusion of the above variables as a predictor is being The comparison of soil carbon sequestration potential recommended for further studies. This will facilitate to on a yearly basis with the other study area is very difficult conservation biologists in predicting TSC by simple as the literatures are very much lacking in such studies; mathematical equations using readily available soil char- however, data on carbon stock densities have been acteristics in remote and inaccessible sites of the tempe- reported by many researchers in the other study sites. rate forest ecosystem and precise carbon budget. Further, GEOLOGY, ECOLOGY, AND LANDSCAPES 215 not much data are available on related study; hence, it Hajabbasi, M. A., Jalalian, A., & Karimzadeh, H. R. (1997). Deforestation effects on soil physical and chemical prop- requires further in-depth research to come up with more erties, Lordegan, Iran. Plant and Soil, 190(2), 301–308. appropriate models in extreme habitats. Findings can Harrison, K. G., Post, W. M., & Richter, D. D. (1995). Soil also play a significant role in spatial mapping of carbon carbon turnover in a recovering temperate forest. Global pool and sequestration for mitigating climate change. Biogeochemical Cycles, 9(4), 449–454. Hassink, J. (1997). The capacity of soils to preserve organic C and N by their association with clay and silt particles. Acknowledgments Plant and Soil, 191(1), 77–87. Houghton, J. T., Ding, Y. D. J. G., Griggs, D. J., Noguer, M., The authors are thankful to the local people (Apatani) for van der Linden, P. J., Dai, X., & Johnson, C. A. (2001). their helping hand and support to carry out the study. Yam Climate change 2001: The scientific basis. The Press is also thankful to Director, NERIST, for permission to carry Syndicate of the University of Cambridge, Cambridge. out soil analysis and University Grant Commission, New Jandl, R., Lindner, M., Vesterdal, L., Bauwens, B., Baritz, R., Delhi, for Dr D. S. Kothari award. Hagedorn, F., & Byrne, K. A. (2007). How strongly can forest management influence soil carbon sequestration? Geoderma, 137(3–4), 253–268. Disclosure statement Jenkinson, D. S., & Powlson, D. S. (1976). The effect of biocidal treatments on metabolism in soils: V. A method of measur- No potential conflict of interest was reported by the authors. ing soil biomass. Soil Biology and Biochemistry, 8,208–213. Jobbágy, E. G., & Jackson, R. B. (2000). The vertical distri- bution of soil organic carbon and its relation to climate References and vegetation. Ecological Applications, 10(2), 423–436. Kaiser, K., & Guggenberger, G. (2000). The role of DOM Allen, S. E., Grimshaw, H. M., Parkinson, J. A., & Quarmby, sorption to mineral surfaces in the preservation of organic C. (1974). Chemical analysis of ecological materials. matter in soils. Organic Geochemistry, 31(7–8), 711–725. Oxford, England:Blackwell Scientific Publications. Krishna, M. P., Varghese, R., & Mohamed, A. A. (2012). Anderson, J. M., & Ingram, J. S. I. (1993). Tropical soil Depth wise variation of microbial load in the soils of biology and fertility. A handbook of methods (2nd ed). midland region of Kerala: A function of important soil Oxfordshire, Wallingford, UK: CAB, International. physicochemical characteristics and nutrients. Indian Arrouays, D., Saby, N., Walter, C., Lemercier, B., & Journal of Education and Information Management, 1 Schvartz, C. (2006). Relationships between particle-size (3), 126–129. distribution and organic carbon in French arable topsoils. Lal, R. (2004). Soil carbon sequestration impacts on global Soil Use and Management, 22(1), 48–51. climate change and food security. Science, 304(5677), Batjes, N. H. (2014). Total carbon and nitrogen in the soils of 1623–1627. the world. European Journal of Soil Science, 65(1), 10–21. Lal, R. (2005). Forest soils and carbon sequestration. Forest Bhatler, R. A. (2009). Temperate forests store more carbon Ecology and Management, 220(1–3), 242–258. than tropical forests, finds study. https://news.mongabay. Li, P., Wang, Q., Endo, T., Zhao, X., & Kakubari, Y. (2010). com/2009/07 Soil organic carbon stock is closely related to above- Bordoloi, R., Das, B., Yam, G., Pandey, P. K., & Tripathi, O. ground vegetation properties in cold-temperate mountai- P. (2018). Modeling of water holding capacity using read- nous forests. Geoderma, 154(3–4), 407–415. ily available soil characteristics. Agricultural Research,1–9. Liski, J.,Korotkov,A. V., Prins,C.F.,Karjalainen, T.,Victor, D. DOI: 10.1007/s40003-018-0376-9 G.,&Kauppi,P.E.(2003). Increasedcarbonsinkintempe- Brown, S., & Gaston, G. (1995). Use of forest inventories rate and boreal forests. Climatic Change, 61(1–2), 89–99. and geographic information systems to estimate biomass Malhi, Y., Wood, D., Baker, T. R., Wright, J., Phillips, O. L., density of tropical forests: Application to tropical Africa. Cochrane, T., . . . Higuchi, N. (2006). The regional variation In:Fitzgerald J.F., Braatz B.V., Brown S., Isichei A.O., of aboveground live biomass in old-growth Amazonian Odada E.O., Scholes R.J. (eds) African greenhouse gas forests. Global Change Biology, 12(7), 1107–1138. emission inventories and mitigation options: Forestry, Martin, P. H., Nabuurs, G. J., Aubinet, M., Karjalainen, T., land-use change, and agriculture (pp. 51–62). Dordrecht: Vine, E. L., Kinsman, J., & Heath, L. S. (2001). Carbon Springer. sinks in temperate forests. Annual Review of Energy and Chauhan, B. S., Stewart, J. W. B., & Paul, E. A. (1981). Effect the Environment, 26(1), 435–465. of labile inorganic phosphate status and organic carbon Reich, P. B., & Bolstad, P. (2001). Productivity of evergreen additions on the microbial uptake of phosphorus in soils. and deciduous temperate forests (pp. 245–283). Terrestrial Canadian Journal of Soil Science, 61(2), 373–385. global productivity. Academic, San Diego. Chave, J., Olivier, J., Bongers, F., Châtelet, P., Forget, P. M., Robinson, D. (2007). Implications of a large global root van der Meer, P., & Charles-Dominique, P. (2008). biomass for carbon sink estimates and for soil carbon Above-ground biomass and productivity in a rain forest dynamics. Proceedings of the Royal Society B: Biological of eastern South America. Journal of Tropical Ecology, 24 Sciences, 274(1626), 2753–2759. (4), 355–366. Saatchi, S. S., Houghton, R. A., Dos Santos Alvala, R. C., Dar, J. A., & Sundarapandian, S. M. (2013). Soil organic Soares, J. V., & Yu, Y. (2007). Distribution of above- carbon stock assessment in two temperate forest types of ground live biomass in the Amazon basin. Global Western Himalaya of Jammu and Kashmir. India Forest Change Biology, 13(4), 816–837. Research Institute, 3(114), 2. Sahani, U., & Behera, N. (2001). Impact of deforestation on Gower, S. T., Landsberg, J. J., & Bisbee, R. L. (2003). Forest soil physicochemical characteristics, microbial biomass biomes of theworld.InR. A.Young &R.L.Giese (Eds.), and microbial activity of tropical soil. Land Degradation Forest ecosystem science and management. PP: 57-74. & Development, 12(2), 93–105. Hoboken: Wiley. 216 G. YAM ET AL. Six, J.,Callewaert,P.,Lenders,S., DeGryze, S., Morris,S.J., von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Matzner, Gregorich, E. G., . . . Paustian, K. (2002). Measuring and E., Guggenberger, G., Marschner, B., & Flessa, H. (2006). understanding carbon storage in afforested soils by physical Stabilization of organic matter in temperate soils: fractionation. Soil Science Society of American Journal, 66(6), Mechanisms and their relevance under different soil con- 1981–1987. ditions—A review. European Journal of Soil Science, 57 Sollins, P., Homann, P., & Caldwell, B. A. (1996). (4), 426–445. Stabilization and destabilization of soil organic matter: Walkley, A., & Black, I. A. (1934). An examination of the Mechanisms and controls. Geoderma, 74(1–2), 65–105. Degtjareff method for determining organic carbon in Srivastava, S. C., & Singh, J. S. (1988). Carbon and phos- soils: Effect of variations in digestion conditions and of phorus in the soil biomass of some tropical soils of India. inorganic soil constituents. Soil Science, 63, 251–263. Soil Biology and Biochemistry, 20(5), 743–747. Wiesmeier, M., Urbanski, L., Hobley, E., Lang, B., von Terakunpisut, J., Gajaseni, N., & Ruankawe, N. (2007). Luetzow, M., Marin-Spiotta, E., . . . Wollschlaeger, U. Carbon sequestration potential in aboveground biomass (2019). Soil organic carbon storage as a key function of of Thong Pha Phum National Forest, Thailand. Applied soils-a review of drivers and indicators at various scales. Ecology and Environmental Research, 5(2), 93–102. Geoderma, 333, 149–162. United Nations Framework Convention on Climate Change Yam, G., & Tripathi, O. P. (2015). Floristic diversity, com- (UNFCC). (1997). Kyoto Protocol, Kyoto (pp. 19). munity characteristics and above biomass of woody spe- UNFCC, United Kingdom cies in subtropical forests of Ziro valley, Arunachal Valladares, F. (2017). A mechanistic view of the capacity of Pradesh. Indian Forester, 141(12), 1269–1277. forests to cope with climate change. In Managing forest Yam, G., & Tripathi, O. P. (2016). Tree diversity and com- ecosystems: The challenge of climate change (pp. 15–40). munity characteristics in Talle Wildlife Sanctuary, Cham: Springer. Kluwer Academic Publishers. Arunachal Pradesh, Eastern Himalaya, India. Journal of Netherlands. Asia-Pacific Biodiversity, 9(2), 160–165. Vashum, K. T. (2016). Soil organic carbon sequestration Zinn, Y. L., Lal, R., Bigham, J. M., & Resck, D. V. (2007). potential of primary and secondary forests in Northeast Edaphic controls on soil organic carbon retention in the India. Proceedings of the International Academy of Brazilian Cerrado: Texture and mineralogy. Soil Science Ecology and Environmental Sciences, 6(3), 67. Society of America Journal, 71(4), 1204–1214.

Journal

Geology Ecology and Landscapes – Taylor & Francis

Published: Jul 3, 2021

Keywords: Microbial biomass carbon; PSLR; soil properties; VIPs; XLSTAT

There are no references for this article.