Combined Use of Carbon and Nitrogen Isotopes for Assessment of Soil Organic Matter Sources and Decomposition in a Typical Karst Area of Yunnan– Guizhou Plateau, Southwestern China

Soil organic matter (SOM) has substantial in�uence on geochemical cycle, soil stability and global climate change, however total organic carbon sequestration mechanisms in karst soil remain poorly understood. For this study we assess, total organic content (TOC), total nitrogen (TN), C/N ratio and isotopes of carbon and nitrogen in four soil pro�les over critical karst area to investigate organic matter source, mechanisms that in�uence fractionation and factors affecting SOM in Yunnan–Guizhou Plateau, Southwestern China. The results revealed that SOM comprised of mixed sources derived from both exogenous and endogenous materials. The soil pro�les indicate intense vertical variation in δ 13 C and δ 15 N with an increase in both isotopes in the upper layers, deceased in δ 13 C below 20 cm and irregular �uctuation in δ 15 N with depth. Mechanisms such as mineralization and selective preservation in�uence isotopic fractionation in the upper soil surface, while translocation, nitri�cation and denitri�cation dominated the subsoil layers. Variation in TOC, TN and stable carbon and nitrogen isotopes were in�uence by vegetation cover, topography, soil water and external contribution. Moreover, the decrease in TOC and TN with depth were due to downward translocation of dissolved organic carbon and nitrogen caused by monsoon climate. Our results revealed that combination of TOC, TN, C/N, δ 13 C and δ 15 N can be used as proxy to decipher SOM source, external in�uence and stability of karst soils. Furthermore, the intense change in δ 13 C and δ 15 N throughout the soil pro�les suggest that this karst soil is unstable which have implications for land management and carbon sequestration.


Introduction
In recent years, it has been recognized that sequestration of carbon and nitrogen dynamics in soil has impacted climate change (Durán et al. 2017;Laganière et al. 2010;Qiu et al. 2017).Soil is widely known to be a major reservoir of organic carbon and nitrogen sequestrations, which plays an integral role in reducing CO 2 in the atmosphere, hence regulating the Earth's carbon cycle and climate system (Reichstein et al. 2013).In addition, with increase in global population greater demand on soil for food security and human development will increase.Therefore, it is important to fully understand the mechanisms and processes affecting carbon and nitrogen, especially within karst environment as these ecosystems are most vulnerable to human threat and environmental change.Carbon and nitrogen are stable isotopes that occur naturally in the environment, and are often used by researchers to determine origin of soil organic matter (Peterson and Fry 1987) as well as processes that govern cycling (Balesdent and Mariotti 1996;Ehleringer et al. 2000;Gill et al. 1999;Poage and Feng 2004).The isotopic compositions in soils surface are inherited from present vegetation cover, usually δ 13 C values for higher terrestrial C3 and C4 plant matter are considered to between -23‰ to -30‰ and -9‰ to -17‰ respectively (Boutton 1991).
However, the fractionation by soil microorganisms during decomposition of soil organic matter (SOM) can complicates the contribution of C3 or C4 plants, and thus disturbs the accurate interpretation of SOM sources and estimation of its turnover (Krull and Skjemstad, 2003).Therefore, greater understanding of the 13 C fractionation in SOM decomposition is important to accurately interpret source.Biological processes are the major cause of variations in carbon and nitrogen isotope, which often recorded the largest isotopic fractionation.Most biological processes showed discrimination in isotopes, with preferential incorporation of the lighter isotopes (Menyailo et al. 2003), and leaving behind heavier isotopes in the substrate.
Basically, stable isotope (δ 13 C and δ 15 N) techniques and C/N ratios are robust tools that have been commonly used to determine sources, mixing and transformational activities in terrestrial, estuarine and coastal systems (Cai et al. 2015;Ramaswamy et al. 2008;Zhang et al. 2007).Carbon isotopes are widely used to explore the origins and the migratory principles of materials across several ecological environments, while nitrogen isotopes are extensively used in biological tracking.Therefore, both δ 13 C and δ 15 N isotopic composition along with C/N ratio can be used together to accurately determine the origin of organic matter (Cai et al. 2015;Middelburg and Herman 2007), processes and mechanisms in SOM decomposition.Since the magnitude and the direction of changes in SOM decomposition are still unclear in ecological fragile karst area, our study aims to used δ 13 C and δ 15 N along with C/N ratio to (1) investigate the source of soil organic matter, (2) identify the mechanisms that in uence C and N isotopic fractionation and (3) determine the factors affecting SOM decomposition over typical karst area of Yunnan-Guizhou Plateau, southwestern China.

Study area
The study area is located in Shilin County, Yunnan Province (24°30'-25°03'N, 103°10'-104°40'E) (Fig. 1a).The terrain is comprised of highlands in the east and low mountains in the west.The area experienced a typical subtropical climate with warm-humid conditions from May to October and cold-dry conditions from November to April.The original monsoonal climate divides the region into two distinct wet and dry seasons.The annual average temperature is 16.3 °C, with relatively large temperature difference between day and night (Yuan 1992).The annual average precipitation approximately ranged between 800-850 mm.Shilin County is located in the karst area of the eastern part of the country.The typical characteristic feature of the landform is karst Mountains, with few exposed limestone rocks and deposit of thin soil.The soil in the area is susceptible to erosion with sparsely stunt vegetation cover as means of protection and are characterized as red soil and limestone soil (Can et al. 2008;Huang 2009;Liu 2015).The dominant lithology in the area are mainly Limestone of Permian period, Limestone and Dolomite of Middle-Lower Cambrian and Slate, with Limestone of Sinian period (Fig 1b).

Soil pro le and sampling
The study sites were selected through indoor research and eld surveys, according to soil type, soil affected by rocky deserti cation and human interference.Four distinct soil pro les were chosen, namely: Shilin Scenic Area 1 (SL1), Shilin Scenic Area 2 (SL2), Xinacun (XNC) and Dapo (DP) (Fig. 1c).
Shilin Scenic Area 1 (SL1) soil pro le is located in the undeveloped section of Shilin near the Stone Forest scenic area at an elevation of 1770 m, with GPS coordinates of 24°48′N; 103°18′E (Fig. 1b).The sampling point is located at the top of the hill, with a relatively at terrain and the immediate surrounding strata is bare.However, the surrounding vegetation cover is high and mainly consists of shrubs and herbaceous plants.Fresh soil samples were collected at an interval of 5 cm from top to bottom of the pro le, numbered and labeled SL1-1 to SL1-10, respectively to a maximum depth of 50 cm (Fig. 1c).The dominant bedrock was Limestone of Permian period and sample was collected (SL1-0) at a depth of 100 cm.A total of 10 soil samples with soil density of 16.91 g/cm 3 were collected.Shilin Scenic Area 2 (SL2) soil pro le is located in the same general area below SL1 at altitude 1730 m, 24°48′N; 103°18′E (Fig. 1b).These soil pro les are naturally formed conical ridge on upland landscapes.The SL2 sampling point is located on the hillside steep slope.The surrounding area is comprised of bare rocks with relatively low vegetation cover consisting of predominantly shrubs.In this section the soil pro le is naturally developed and without any visible sign of human disturbance.Fresh soil samples were collected at an interval of 12 cm from top to bottom of the pro le.The samples numbered and labeled from SL2-1 to SL2-10, respectively to a maximum depth of 113 cm (Fig. 1c).Limestone bedrock sample was also collected (SL2-0) at 163 cm.A total of 10 soil samples with density of 17.10 g/cm 3 were collected.
The Xinacun (XNC) soil pro le is located at an elevation of 1720 m, with GPS coordinates of 24°49′6″N; 103°18′37″E (Fig. 1b), on a mountain slope in close distance to a road construction site.This sample point was a man-made vertical soil pro le established by an excavator.The upper surface of this pro le comprised of weeds, shrubs and tall trees with loose soil containing plant roots and black humus.No bedrock was present in this excavated soil pro le, however the underlying bedrock was found in close proximity and identi ed as Limestone and Dolomite of Middle-Lower Cambrian.Fresh soil samples were collected starting at intervals of 10, 20, 40 to 50 cm from top to bottom of the pro le, numbered and labeled XNC-1 to XNC-11, respectively to a maximum depth of 245 cm (Fig. 1c).A total of 11 soil samples were collected with a density of 19.30 g/cm 3 .
The Dapo (DP) soil pro le is located at an elevation of 1920 m, with GPS coordinates of 24°50′21″N; 103°23′26″E (Fig. 1b) in the lower gully region of the upland landscapes.The area comprised of thatched grass, with few shrubs and a small number of coniferous trees.The underlying bedrock of this pro le is Slate, Phyllite, Siltstone intercalated with Limestone of Sinian period.Fresh soil samples were collected at an interval of 10 cm from top to bottom of the pro le, numbered and labeled DP-1 to DP-10, respectively to a maximum depth of 115 cm (Fig. 1c).Shale bedrock sample (DP-0) was also collected at 145 cm.A total of 10 soil samples, with density of 20.70 g/cm 3 were collected.
Soil organic matter content in the soil pro le usually changed signi cantly from the soil surface to a depth of approximately 30 cm, and then slightly varied at greater depths below 30 cm (Han et al. 2017).
Each studied soil pro le has different depth vary in thickness ranging from 50-250 cm (Fig. 1c).In order to captured and represent these change soil sampling was done according to the genetic horizons of each pro le; however, comparison was made over equidistant.Hence, soil samples were collected at varying intervals (5-20 cm), however all the soil pro les were chosen with a thickness of 50 cm and over so as to compare the changes of TOC and TN contents and their stable isotope compositions with depth (Fig. 1c).

Soil preparation and analysis
All soil samples collected were naturally air dried at room temperature and impurities such as plant roots, gravel, and other debris were removed and soil then stored in a plastic bag.Soil samples were later ground to 200 mesh size and packaged in zip lock bags to determine Total organic carbon (TOC), Total nitrogen (TN), C/N ratio, nitrogen and carbon isotopes (δ 15 N and δ 13 C).
The pH was determined by using the potentiometric method (Wang et al. 2007) whereby a small portion of air-dried soil sample that were pass through a 1 mm sieve placed in a liquid to soil ratio of 2.5:1 (Hash HQ40D) and measured using a pH meter.The powder soil sample was pretreated with 1mol/L HCl, and the carbonate minerals were removed (Midwood and Boutton 1998).The inorganic N were removed using 2 mol L -1 potassium chloride (KCl) for 24 hours (Meng et al. 2005).The samples were washed with deionized water until the supernatant liquid pH value was neutral and then later dried at 60 °C.Approximately 100 mg of sieved dried soil sample was used for analysis of TOC, TN and stable isotopes contents.The TOC and TN contents were calibrated due to loss of carbonate and inorganic N respectively.The TOC and TN contents were determined by using an elemental analyzer (Elementar, Vario TOC cube, Germany) with a precision of C ± 0.1% and N ± 0.02%, monitored with standard samples.The stable carbon isotope ratio ( 13 C/ 12 C) and stable nitrogen isotope ratio ( 15 N/ 14 N) were determined by using the gas isotope ratio mass spectrometer (MAT-252, Germany).This testing was done by the State Key Laboratory of Environmental Geochemistry, Institute of Geochemistry.The measurements were normalized according to international standards material (δ 13 C VPDB : 45.6‰ ± 0.08‰; δ 15 N Air : 0.24‰ ± 0.13‰), and expressed as delta value (δ 13 C and δ 15 N) notation (‰) relative to Vienna Pee Dee Belemnite (VPDB) and atmospheric air, respectively, whereby: δ 13 C (‰) = [( 15 C/ 12 C) sample -( 13 C/ 12 C) standard] / ( 13 C/ 12 C) standard × 1000‰ and δ 15 N (‰) = [( 15 N/ 14 N) sample -( 15 N/ 14 N) standard] / ( 15 N/ 14 N) standard × 1000‰.Reproducibility was determined by replicate measurements that was better than 0.1‰ and 0.2‰ for δ 13 C and δ 15 N respectively.

Results
Table 1 illustrates depth distribution of pH, soil density, total organic carbon, total nitrogen, C/N ratios, and isotopic composition (δ 13 C and δ 15 N) of the soil pro les in Yunnan Plateau, Southwestern China.
3.1 Vertical distribution of soil pH, soil density, TOC, TN and C/N in soil pro les As illustrated in gure 2 the pH values of the studied soil pro les vary from 4.05 to 6.09, which are slightly acidic in nature, however DP soil pro le tend to be more acidic with average pH value of 4.23.There were slight uctuations in pH values with increasing soil depth, with the upper soil surface being more acidic than rest of the pro le except for SL1 and DP soil pro les (Fig. 2).
The soil density ranges from 1.13 g cm -3 to 2.30 g cm -3 with DP and SL2 recording the lowest and highest soil density respectively.Soil density varied with depth for the studied soil pro les with an overall general increase with depth (Fig. 2 and Table 2).
The TOC content within the studied soil pro les showed high variation with soil depth with a general decrease from upper soil surface to lower layers down to bedrock (Fig. 2 and Table 2).
Total Nitrogen (TN) tends to vary with soil depth for the studied soil pro les.Generally, there seem to be a decrease in TN with increasing soil depth with the upper soil surface having a higher TN value than the other lower soil pro les layer except for SL1 and XNC (Fig. 2 and Table 2).SL1 soil pro le is signi cantly higher in TN content in comparison to the other pro les.The surrounding high vegetation cover which mainly consists of shrubs and herbaceous plants as well as the rearing of animals may account for the relatively high TN content in SL1 soil pro le.
The studied soil pro les C/N values have high degree of variability, with values ranging from 1.5 (XNC) to 13.4 (SL1) with an average of 6.60 ± 4.02.The general trend indicates a gradual reduction in C/N values from the topsoil to the bottom of the soil pro le layer (Fig. 3 and Table 2).There is a signi cant difference in the upper topsoil C/N ratio of all studied soil pro les (>10), which may indicate that plant residues or animals input contributes to the relatively high C/N value in the upper soil surface.There are also slight differences, which may be caused by variation in hydrothermal conditions, soil formation characteristics and soil morphology in different section of the pro les (Wang et al. 2012;Zhu 2006).The degree of human disturbance or interference may also in uence the migration and transformation of nitrogen, resulting in C/N variation at different layers within the soil pro les.

The clay proportion and depth distribution of the soil pro les
Table 1 also shows the clay proportion depth distributions of the four studied soil pro les.All soil samples at various depths were dominated by silt and sand (average 82.37%), while clay fractions accounted for only 7.90 to 36.63% in an increasing order of XNC, SL2, SL1 and DP.The highest clay content generally recorded in the mid-section of the pro les, while the lowest value recorded in the upper surface except for XNC and DP soil pro les.

Isotopic composition
The δ 13 C and δ 15 N isotopic compositions of the four studied soil pro les from Yunnan Plateau, Southwestern China are presented in table 1 and gure 2. The overall δ 13 C and δ 15 N values ranged from -15.64‰ to -25.82‰ and from 2.77‰ to 14.36‰ respectively and varied with soil depth.The general trend indicates that δ 13 C values increased with depth for the initial 0-20 cm for the soil pro les except for DP soil pro le and the lower section of the pro les which indicate a decrease in δ 13 C values with depth (Fig. 2a, b, c and d) respectively.While δ 15 N values varied rapidly with depth with no obvious trend (Fig. 2).Additionally, δ 14 C radioisotope was used to determine the age of the soil pro les which ranged from 500 to 890 before present (BP), and 90 to 6537 BP for SL1 and SL2 respectively, no age determination data was available for XNC and DP soil pro les.

Factors affecting TOC and TN vertical distribution during SOM decomposition
Land use signi cantly in uenced the levels of the TOC and TN contents in soils, more so in the rst 0 -100 cm layer (Tuo et al. 2018) as well as soil bulk density (Chukwudi et al. 2018).Several studies reported that TOC and TN contents were much higher in cropland than in native forestlands (Li et al. 2010;Wang et al. 2010;Zhang et al. 2014), conversely however Tuo et al. (2018) and Liu et al. (2019) found that forestlands lands tend to have higher TOC and TN contents than other land use such as cropland, grassland and shrubland.Our studied soil pro les were located in forested area over karst which explain the relatively high TOC and TN contents in upper soil surface and the low carbon at depth.The decrease in TOC content below 20 cm depth for all studied soil pro les (Table 1; Fig. 2), were due to the relatively small amount of nutrients that are able to in ltrate into deeper soil.Similar decrease in TOC with depth pattern was reported by Liu et al. (2019) for different land use over karst area.However, for undisturbed natural forests, usually this decrease trend in TOC is largely determined by organic decomposition processes that result in mostly recalcitrant soil carbon at depth (Campbell et al. 2009;Paul et al. 1997).Similarly, other studies reported that the decrease in TOC with depth was due to the fact that few residues and biomass were introduced in deep soil layers (Jobbágy and Jackson 2000;Wiesmeier et al. 2012), and that the proportion of carbon matter in the fraction cycle decreased at deeper layers (Trumbore 2000).In addition, soil density and clay content seem to in uence TOC and TN concentration in this karst soil as both indicate negative correlation for soil pro les SL2, XNC and DP, while positive correlation for SL1 soil pro le (Fig. 2, 4 and 5).This negative correlation relationship has been exhibited in other studies (Njeru et al., 2017).This was due to the suppression of possibly the process of mineralization and nitri cation in soils with high soil density values (De Neve and Hofman, 2000).Furthermore, soil with low density has the ability to store greater amount of TOC and TN content, as they can be mobilized through porous spaces within the soil pro les.
The climate in particular rainfall may also affect TOC and TN vertical distribution.In this study, carbon loss is mostly affected by the monsoonal climatic condition as the seasonal high and low precipitation impacted on the vertical translocation of DOC and anaerobic oxidation especially during the dry period e.g., as indicated by the sharp change in δ 13 C pro le for all sites except for DP pro le (Fig. 2).The observed pattern of δ 13 C for the soil pro les can be explained by the decomposition of TOC which directly affects isotopic composition by kinetic fractionation and preferential substrate decomposition (Guillaume et al. 2015).For all the studied soil pro les except for DP, the δ 13 C which is the heavier isotopes became enriched with increasing depth to approximately 20 cm.This clearly indicates kinetic fractionation whereby soil microorganisms during SOM decomposition preferentially removed the lighter δ 12 C isotopes leaving behind substrate that is isotopically heavy or enriched.On the other hand, at greater depth below 20 cm and for DP pro le, the decrease in δ 13 C was due to the translocation of DOC from surface to deeper depth Kalbitz et al. 2005.The subsoil horizons with low carbon contents, may have DOM readily adsorbed to mineral surfaces, resulting in the reduction of δ 13 C in soil (Kalbitz et al. 2000;Kalbitz et al. 2005).Monsoonal climatic condition in uence both decomposition and translocation processes which explain the carbon gain in the upper surface and carbon loss in subsoil.Zhu and Liu (2006) found similar result where δ 13 C declined with depth in karst soils.
Topography and vegetation also affect TOC and TN distribution within the karst soil pro les.The SL1 pro le is the uppermost pro le with an elevation of 1770 m, which explain having the lowest surface TOC content as a result of limited vegetation cover and its susceptibility to erosion and leaching.Xiong et al. (2018) demonstrated that this studied karst area is severely being affected by physical erosion and chemical weathering.Conversely, the SL2 upper surface soil has signi cantly higher TOC content (2.11 g kg -1 ) than the other soil pro les (Fig. 2b).This is primarily due to its location being below SL1 at an elevation of 1730 m and so bene ted from erosion and deposition of SOM from SL1.The other pro les surface soil TOC content relatively high (0.9 g kg -1 ) and re ect the input of freshly decomposed matter from surrounding vegetation.The soil depth distribution trend of total organic carbon content is consistent with some other studies (Li et al. 2009;Liu et al. 2019;Wang et al. 2016;Zhu and Liu 2006) and is closely linked to the evolution of the soil pro le (Chen et al. 2005) as well as the amount of activity from soil microbes (Zhu and Liu 2006).Both the TOC and TN concentrations are in uenced by the erosion of carbon rich soil from upper surface layer (Don et al. 2011;Lal 2001) and by the subsequent degree of decomposition, depositional processes, delivery routes and amount of preservation (Tani et al. 2002).Guillaume et al. (2015) have indicated that the processes of erosion and decomposition can potentially contribute to the decrease of C content and stocks in soil.While erosion affect the concentration of TOC and TN, decomposition tends to affect the isotopic composition.
The C/N ratio in soil and litter may be an indicator of the rate of decomposition (Zhu and Liu 2006).For example, the lower the C/N value the higher the decomposition rate (Jafari et al. 2011).In comparison to other studies for e.g., Liu et al. (2019) found C/N values range from 8.22 to 10.52, while our study C/N ratio were >10 for all the upper surface soils, except for SL1 pro le (Table 2; Fig. 3).These ndings indicate that the upper soil surface for the studied soil pro les, have a low rate of decomposition and low accumulation.The C/N ratio re ects carbon input, therefore the relatively high C/N values in the upper surface indicate the high C content in the surrounding shrubs and herbs vegetation (Fig. 3).Furthermore, the results also indicate that SL1 average C/N values fall in the range of cropland (disturbed) i.e., >10, while the other soil pro les average C/N values re ect typical undisturbed native land (<10) according to Zhang et al. 2014.The soil C/N ratio showed strong positive correlation with TOC (r 2 =0.772, 0.822, 0.929 and 0.931) and TN contents (r 2 =0.694, 0.782) for SL1 and SL2, however weak correlation for XNC and DP soil pro les respectively (Fig. 5a & b).This was directly due to in uence of vegetation cover and water availability in the soil.
Stable carbon and nitrogen isotopes showed vertical variation in soils for this karst area with distinct characteristics.Factors such as land use, climate, topography and vegetation cover in uenced TOC and TN distribution during SOM decomposition in this karst area.

Source of Soil Organic Matter (SOM)
Soil is a direct sink for organic carbon and nitrogen with their stable isotopic signatures registered at any depth re ecting the dominant vegetation type.Higher terrestrial C3 plants have distinctly different δ 13 C value range from C4 plants as well as soil organic matter (Fig. 6).Many researchers used δ 13 C and δ 15 N isotopes along with C/N ratio to better identify the origins of soil organic matter (Cai et al. 2015).In using δ 13 C values, higher terrestrial C3 and C4 plant matter are considered to between -23‰ to -30‰ and -9‰ to -17‰ respectively (Boutton 1991), while that of soil organic matter ranges between -22‰ to -25‰ (Goni et al. 2003).Most non-xing nitrogen plants have δ 15 N values in the range of about +5‰±2‰, although plants with δ 15 N values less than -10‰ or greater than +10‰ are not unusual (Kendall et al. 2001) (Fig. 6).Our results indicated vertical variation of δ 13 C for each of the studied soil pro les (Table 1 and Fig. 2).The δ 13 C values for these studied soil pro les ranges between -15.64‰ to -25.82‰ with an average of -20.03‰ ± 3.02 which lies outside the range of both C3 plants (-26.5 to -28.0‰), and C4 plants (-12.5 to -14.0‰) according to O'leary (1988) and Tieszen (1994) respectively (Fig. 6).However, some values conversely fall in the range of C3 plants based on Boutton 1991 (-23 to -30‰) and outside the C4 plants range (-9 to -17‰).Some of these studied pro les for e.g., XNC and DP have larger range of δ 13 C values which fall in the C3 plant and soil organic matter range according to Goni et al. 2003 andThorp et al. 1998 (Fig. 6).The inconsistency in isotopic signature range was as a result of mixing of both C3 and C4 plants as well as the fractionation cause by in uence of soil microbes.In the case of SL1 the soil organic matter is strongly in uenced by other source (e.g., animal residues) especially since the δ 13 C values are relatively enriched (Fig. 6).The low value of δ 13 C in the upper surface of soil pro les implies that soil organic carbon mainly derived from decomposition of fresh plant matter, especially since surface horizons of soils re ect the C3 to C4 ratios of the present vegetation (Kelly et al. 1991).The δ 13 C values for soil pro les SL2, XNC and DP are closer to the C3 plant range while SL1 pro le is signi cantly different but closer to C4 plants (Fig. 6).The SL1 surrounding vegetation consist of shrubs and herbaceous plants which contributed to the δ 13 C values.The immediate surrounding vegetation of the other soil pro les comprised of shrubs, thatched grass, and coniferous trees that account for the high variation in isotopic signatures.Except for DP soil pro le that recorded the most depleted (-21.18‰)δ 13 C value in the upper soil surface than the other pro les (Fig. 2d).This is because DP pro le at the highest elevation would have the least contribution of soil deposition and accumulation from other landforms and so re ecting δ 13 C values close to the source.
In forested areas, the input of plant organic matter greatly affect the δ 15 N values in soil surface (Eissfeller et al. 2013).The δ 15 N values for plants are affected by the δ 15 N values of soil, resulting in a close relation between soils and plants nitrogen composition in the ecosystem (Nel et al. 2018).In addition, the in uence of atmospheric deposition can also affect δ 15 N values in soils, for instance Southwest China wet deposition for NO 3 -is 2‰ and for NH 4 + is -12‰ (Xiao and Liu 2002).In relation to δ 13 C, there are higher variation of δ 15 N values for the studied soil pro les with SL1 having a signi cantly higher average δ 15 N value than the other pro les (Table 1 and Fig. 2a).The δ 15 N values for SL1 soil pro le ranges between 10.43‰ to 14.85‰ and are outside the range of both C3 and C4 plants 1‰±4 or 5‰ and 3‰ ±4 or 5‰ respectively, (Fig. 6).The input of plant organic matter and atmospheric deposition cannot explain the high δ 15 N values in SL1 pro le.The presence of animal residue however could account for the relatively high δ 15 N value registered in SL1 soil pro le.The other soil pro les SL2, XNC and DP, δ 15 N values fall within the range of C3 and C4 plants, except for one point in XNC soil pro le (XNC-3 registered δ 15 N =13.22).The high δ 15 N value in XNC pro le maybe due to soil water migration and transformation of nitrogen as a result of microbial activities.Stewart et al. 2014 reported that soil water processes affected redox environment, which control nitrification and denitrification.In addition, based on other literature Fogg et al. 1998 reported that the average δ 15 N values of soil from sites under animal waste, fertilizer, soil organic matter and sewer septic sources ranged from (+10 to +25), (0 to +5), (-3 to +5), and (+7 to 15‰), respectively.Our results suggest that animal waste could likely be a potential source for SL1 pro le while the other pro les δ 15 N values fall in the range of soil organic matter with some amount of in uence from other sources.This high variability of δ 15 N values within a single source and the overlapping of values may prove challenging in the ability to accurately distinguish one source from another.The δ 15 N in the soil organic layer can indicate the relative rate of soil N cycling due to significant correlation between δ 15 N and mineralization and nitrification rate (Templer et al. 2007).The relatively high δ 15 N values in the studied soil pro les which uctuate with depth indicate microbial activities during mineralization, nitri cation and denitri cation.The enrichment of δ 15 N that occurred in the upper soil surface 0-20 cm of all the studied soil pro les were due to microbial activities present in karst soils.The high variability of δ 15 N in the subsoil was attributed to soil water processes and redox environment involving nitrification and denitrification (Stewart et al. 2014).The seasonal wet and dry conditions in this karst environment due to monsoonal climate facilitate intensive microbial activities in the upper surface as well as translocation of DON at depth resulting in high variability in δ 15 N.
In addition to δ 13 C and δ 15 N isotopic signatures, the C/N ratio can also be used to support the suggestion that the studied soil pro les are of mixed organic matter sourced.Studies have shown that the C/N ratio of higher terrestrial plants are usually >15 (Kendall et al. 2001) and that C/N ratio of soil organic matter is usually between 10 and 13 (Parton et al. 1987;Tiessen et al. 1984).The C/N ratio of SL1 soil pro le is between 10.3 and 13.4, with an average of 11.75±2.3(n=10) (Table 2), which suggest that the sources of organic matter are uniform and not mixed; in other words, the sources are in situ (indigenous) to the soil pro le.However, the other soil pro les showed high variation in C/N ratio in the upper soil surfaces which are signi cantly different (>10) from the lower layers (Table 2 and Fig. 3).This indicates that the sources of organic matter are mixed; re ecting in uence of both exogenous and indigenous sources.Based on our ndings it is evident that the source of organic matter within this karst area of Yunnan plateau, Southwestern China has mixed contribution of C3, C4 plants and microbes with animal residue in uencing SL1 pro le.
4.3 Mechanisms controlling δ 13 C and δ 15 N isotopes fractionation in karst soil pro les In ecological studies, difference in abundance of δ 13 C and δ 15 N isotopes are important in the evaluation of changes under varying conditions of SOM decomposition and land use.Overall, our results indicate that δ 13 C and δ 15 N values varied with soil depth with δ 15 N values showing greater irregular uctuation (Fig. 2).Meanwhile, the general trend observed for δ 13 C values was enrichment with increasing depth to approximately 20 cm, followed by depletion with increasing depth (Fig. 2a, b & c).Except, however for DP pro le which indicate a decrease in δ 13 C values with depth (Fig. 2d).This increased variation of δ 13 C with depth is a characteristic feature of forest soils (Chen et al. 2005;Powers and Schlesinger 2002a) and indicate microbial activities during SOM decomposition.Soils dominated by aerobic decomposition have a clear increase of δ 13 C values with depth.This is due to preferential release of the lighter δ 12 C isotopes during aerobic mineralization.The initial 0-20 cm increase supported the general trend expected for undisturbed soil where studies have shown that both isotopes become more enriched in the heavier isotopes with increasing soil depth (Broadbent et al. 1980;Campbell et al. 2018;Chen et al. 2005;Ehleringer et al. 2000;Garten et al. 2000;Gebauer and Schulze 1991;Ji-Suk and Hee-Myong 2018;Krull et al. 2006;Ledgard et al. 1984;Mariotti et al. 1980;Nadelhoffer and Fry 1988;Powers and Schlesinger 2002a,b;Steele et al. 1981).However, the increasing δ 13 C trend were only observed in the top surface layer for this karst soil.The increasing trend has been reported to be associated with a number of mechanisms which may include the Suess effect (i.e., decrease in δ 13 C in atmosphere due to combustion of fossil fuels), selective preservation, microbial turnover, microbial carbon mixing, and changes in plant community composition (Ehleringer et al. 2000;Krull and Skjemstad, 2003;Nadelhoffer and Fry 1988).In our study the observed enrichment trend of δ 13 C in surface soils were related to kinetic fractionation caused by microbes rather than Suess effect or changed in plant community composition.This is because Suess effect is likely to contribute to a vertical increase of δ 13 C throughout the pro le (Garten et al. 2000), and this was not observed in our pro les.
In contrast soils that experience suppressed degradation due to anoxic conditions, coupled with DOC translocation usually indicate a decrease in δ 13 C values with depth.The lighter δ 13 C values is due to enrichment of recalcitrant organic substances during anaerobic mineralization, result in depletion of the heavier δ 13 C. Furthermore, subsoil horizons with low carbon contents, may have DOM readily adsorbed to mineral surfaces, resulting in the reduction of δ 13 C in soil (Kalbitz et al. 2000;2005).At greater soil depth, selective preservation result in a decrease of δ 13 C rather than an increase (Wynn et al. 2006) as seen in gure 2a, b and c, usually after 20 cm depth and DP pro le (Fig. 2d).This decrease is due to the preferential utilization of lighter δ 13 C from organic functional groups (e.g., simple carbohydrates) by microbial decomposers and selective preservation of plant lignin, which is depleted in 13 C (McCorkle et al.  2016).
Soil δ 15 N values related to the nutrient input, humi cation and nitrogen transformation as in uenced by land use and land cover change (Awiti et al. 2008).The δ 15 N values in this studied karst area forest soils (2.77 to 14.36‰) were signi cantly higher than those reported in other studies e.g., Liu et al. 2019 (2.9 to 4.3‰) indicating more rapid N cycling in these soils.The differences were due to in uence of animal residues, as well as the differences in forest vegetation foliar N content, which depend on species, precipitation and temperature (Craine et al. 2010).The general trend naturally for δ 15 N in forest soils was the presence of two pools; top surface soil with low δ 15 N values and subsoil with high δ 15 N values (Nadelhoffer and Fry 1988).The vertical distribution of δ 15 N in this karst area showed high uctuation which peak at depth for all the studied soil pro les (Fig. 2), except for XNC soil pro le which peak at 25 cm (Fig. 2b).This was due to the spatial dynamics of soil water processes and the in uenced on redox reaction associated with SOM nitri cation and denitri cation (Stewart et al. 2014).Water availability affect the cycling of carbon and nutrients and thereby regulating the growth and distribution of microbes.
Similarly, strong fluctuation in δ 15 N with depth were reported by Liu et al. 2019 in karst watershed.The high variability pattern of δ 15 N distribution in these soil pro les is associated with fractionations that occur during the process of mineralization, nitri cation and physical mixing.Similar ndings were reported by Nadelhoffer and Fry 1994.The peak of δ 15 N in the deeper soil pro le is common with other ndings (Mariotti et al. 1980;Nadelhoffer and Fry 1988;Steele et al. 1981).This is because, at depth, microbial excretions of ammonium and nitrate become less depleted in δ 15 N, while the humus content of nitrogen becomes enriched, and the uptake of nitrogen by plant roots lower the δ 15 N abundance in the microbial reserve hence the total soil becomes enriched (Nadelhoffer and Fry 1994).In addition, further δ 15 N enrichment with depth may be because of N loss during high precipitation input.The high complexity of soils containing several isotopically different forms of nitrogen (Ledgard et al. 1984;Tiessen et al. 1984) resulting in N irregular behaviour, hence the notable high variation in δ 15 N values with depth for the studied soil pro les.The δ 15 N isotopic composition in soils and groundwater may not only be in uenced by its source but also by microbial activities and physical processes such as ion exchange (Ostrom et al. 1998).
Our results in this karst environment illustrate a non-conservative behaviour in δ 15 N that is largely due to a combination of microbial activities and water availability, which affects redox reactions (Fig. 2a-d).The monsoonal climatic condition along with the topographical features of karst environment provide the uneven distribution of soil which result in the spatial and temporal disparity in soil water (Tokumoto et al. 2014).Essentially, the loss of nitrate through denitri cation result in enrichment of δ 15 N content in the remaining substrate, while during nitri cation process, the lighter δ 15 N isotope is preferentially incorporated into nitrate resulting in a decrease in δ 15 N (Ostrom et al. 1998).Therefore, the dominant mechanism affecting δ 15 N within karst area of Yunnan Plateau, Southwestern China are mineralization in the upper surface, while nitri cation and denitri cation in the subsoil.
4.4 Interactions among δ 13 C, δ 15 N isotopes with TOC, TN and C/N ratio Vegetation is the primary source of TOC and TN which can signi cantly affect the quantity and quality of soil organic matter (Podwojewski et al. 2011).Our results indicated that the soil pro les TN contents were positively correlated with the TOC contents (r 2 =0.982, 0.988) for SL1 and SL2, however weak correlation for XNC and DP respectively.The result implied that majority of nitrogen were of similar organic origin for SL1, SL2 and that soil water affected N in XNC and DP pro les (Fig. 7).The predominant sources of nitrogen are from animal residue in SL1 pro le, while leaf-litter, biological nitrogen xation, and organic matter dominate the other pro les (Bai et al. 2001).This strong positive correlation between TOC and TN contents has also been found in other studies e.g., (Diwediga et al. 2017;Gelaw et al. 2014;Liu et al. 2019;Liu and Wang, 2009;Xiong et al. 2018).
The relationships between soil particulate δ 13 C, δ 15 N and C/N ratios are shown in gure 6 and 8.The results indicate mixture of end member particulates from terrestrial C3, C4 plants and soil organic matter, with SL1 signi cantly different from the other pro les.The high δ 15 N values were due to the input of animal residue as it comprised of similar isotopic range as reported by Fogg et al. 1998. Typically, higher terrestrial vascular plants usually have C/N ratios higher than 15 (Meyers, 1994).Overall, for these studied soil pro les the C/N ratios ranged between 1.5 to 13.4 (mean 6.60 ± 4.02) indicating a mixture of terrestrial plants and other sources (Table 2; Fig. 8).A weak positive correlation was observed between the δ 13 C and C/N ratio for all the studied soil pro les, suggesting that to some extent both parameters are in uenced by the same factors such as source of organic matter and decomposition rate of SOM (Fig. 8a).
This study revealed that both δ 13 C and δ 15 N of soil TOC and TN (Fig. 9a and b) re ect isotopic fractionation associated with litter fall decomposition, microbial activities, translocation and physical mixing processes.Forest soils are characterized by litter fall and root exudates that are constantly being decomposed and gradually mixed within the soil pro le (Acton et al. 2013).The older more decomposed SOM found lower in the soil pro le as a result of further decomposition, physical mixing, leaching and vertical mobilization (McCorkle et al. 2016;Nadelhoffer and Fry 1988).On the other hand, the newly decomposed or fresh C and N inputs found in the upper soil pro le (Wang et al. 2018).
Figure 10 shows the relationship between δ 13 C and δ 15 N for the soil pro les which indicate weak negative correlation.This correlation was interpreted as the mixing of isotopically distinctive carbon and nitrogen source (Huon et al. 2002;Meyers and Takemura 1997), as well as microbial activities.The SL1 soil pro le was signi cantly different from the other pro les, with intense enrichment in δ 13 C and δ 15 N values, suggesting a higher degree decomposition and in uence of animal residue (Fig. 2a and 7).Based on the C and N regression slope, the studied soil pro les can be placed in order of external in uence and soil stability with undisturbed forest soil as baseline, followed by DP, XNC, SL2 and SL1 (Fig. 10).This provides support that δ 13 C and δ 15 N isotopic assessments can be used as a monitoring tool to determine source, external in uence and soil stabilization.

Conclusions
This study provides insights into sources and distribution of soil organic matter, as well as the behaviour and mechanisms of carbon and nitrogen isotopes within karst soils in Yunnan Plateau, Southwestern China.The study indicated vertical variation in TOC and TN contents with soil depth which were largely in uence by climate, topography, vegetation cover and land use.
The joint end-member analysis of δ 13 C, δ 15 N and C/N ratios revealed organic matter of this karst soil have mixed sources which derived from both exogenous and endogenous materials.The external source being that of animal residues, and the internal source comprised of a mixture of terrestrial C3, C4 plants and soil microbes.In addition, the soil organic matter of these pro les was not dominated by any particular terrestrial plant group as they mostly fall outside the isotopic signature range and therefore showed characteristics of a mixing pool.We speculate that the ages of pro les are relatively young which were con rmed by δ 14 C radioisotope.
The carbon and nitrogen stable isotopes in these soil pro les revealed vertical variation and spatial differentiation, which re ected activities of SOM mineralization, translocation and leaching.We concluded that due to the variability and irregular pattern of δ 13 C and δ 15 N isotopes fractionation, that the forest soil C and N contents are unstable over this karst environment.The dominant mechanisms that in uenced δ 13 C variation within these soil pro les are mineralization in the upper surface while DOC translocation and selective preservation in the subsoil.Conversely, the high vertical variation in δ 15 N within these soil pro les were controlled by differences in the plant composition and fractionations that had occurred during the process of mineralization, nitri cation and DON translocation.
The analysis of results further revealed strong positive correlation among δ 13 C versus TOC, TN and C/N, while negative correlation among δ 15 N versus TOC, TN and C/N for all the studied soil pro les within this karst area in Yunnan Plateau, Southwestern China.The negative correlation between δ 13 C and δ 15 N has been interpreted as the mixing of isotopically distinctive carbon and nitrogen source as well as microbial activities and translocation.This was due to the fact that δ 13 C and δ 15 N become enriched in the upper surface during decomposition of SOM, with the preferentially incorporation of the lighter isotopes by soil microbes, leaving behind the heavier isotopes.Coupled with downward translocation of DOC and DON throughout the pro les result in a decrease in isotopes with depth.The ndings suggest that within karst environment combination of δ 13 C and δ 15 N isotopes along with other parameters can be used as a tracer proxy to determine soil organic matter source, external in uence and soil stabilization.

Figure 3 C
Figure 3

Figure 10 Relationship
Figure 10 DeclarationsZhang, C.,Liu, G. B., Xue, S., et al. 2013.Soil organic carbon and total nitrogen storage as affected by land use in a small watershed of the Loess Plateau, China.European Journal of Soil Biology, 54: 16-24.Zhang, J.,Wang Xiu-jun, Wang Jia-ping and Wang Wei-xia.2014.Carbon and Nitrogen Contents in Typical Plants and Soil Pro les in Yanqi Basin of Northwest China Journal of Integrative Agriculture Chen, F. R., Yang, Y. Q., et al. 2008.Distinguish of sources of organic matter in sediment in Pearl Estuary and adjacent water.Marine Environmental Science, 27: 447-451.Zhong, Z.,Chen, Z., Xu, Y., Ren, C., Yang, G., Han, X., Ren, G., and Feng, Y. 2018.Relationship between Soil Organic Carbon Stocks and Clay Content under Different Climatic Conditions in Central China.Forests, 9, Table showing average density, TOC, TN and C/N for the four studied soil pro les (SL1, SL2, XNC and DP) from karst area in Yunnan-Guizhou Plateau of southwestern China.