Accumulation of organic carbon and its association with macro-aggregates during 100 years of oasis formation

Catena 172 (2019) 770–780

Contents lists available at ScienceDirect

Catena journal homepage: www.elsevier.com/locate/catena

Accumulation of organic carbon and its association with macro-aggregates during 100 years of oasis formation

T

⁎

Chenhua Lia, Yan Lia, , Jiangbo Xieb, Yan Liua, Yugang Wanga, Xuecan Liua a b

State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi 830011, China State Key Laboratory of Subtropical Silviculture, Zhejiang A & F University, Hangzhou 311300, China

A R T I C LE I N FO

A B S T R A C T

Keywords: Oasis formation Soil profile Soil organic carbon (SOC) Nitrogen (N) Macro-aggregate

The maintenance and accumulation of soil organic carbon (SOC) is critical to the agricultural sustainability and environmental stability in oasis-desert belts. The objective of this study was to determine the effects of oasis formation on SOC and aggregate structure, as well as the linkage between them, throughout a 0–200 cm soil profile. The investigation was conducted in five oases at the northern foot of the Tianshan Mountains in Central Asia. Oasis farmlands reclaimed 3, 5, 10, 20, 50, and > 100 years ago were compared with the desert pairs they originated from. The SOC content significantly increased throughout the whole profile after 20 years of reclamation, despite a loss in the first 10 years of reclamation. The values reached maxima at 50 years of reclamation, which increased by 67–135% compared to initial values. The macro-aggregate (diameter > 0.25 mm) fraction with high carbon (C) concentration significantly increased throughout the soil profile after reclamation, and showed the greatest variation during oasis formation, compared with the other aggregate fractions (0.25–0.053 and < 0.053 mm). These changes were significantly correlated with increases in soil nutrients and microbial biomass and decreases in soil pH and salt during oasis formation. In conclusion, the oasis formation enhanced SOC accumulation not only in topsoil but also in deep soil, and soil aggregate structure was improved by increased macro-aggregates. The formation of macro-aggregates and the increase in their associated C had significant correlations with SOC accumulation. Fertilization, especially inorganic nitrogen, very likely promoted the SOC accumulation and soil aggregation in concert with annual input of crop residues into the originally poor desert soils during oasis formation.

1. Introduction Soil organic carbon (SOC) accumulation is considered to be closely associated with soil quality, agricultural productivity and atmospheric carbon dioxide concentration (Croft et al., 2012). Therefore, SOC sequestration has a profound significance to ecosystem stability and agricultural sustainability (Lal, 2004; Ratnayake et al., 2017). Soil aggregates, as component of soil structure, play a key role in a number of soil processes, and are of particular importance for soil carbon (C) sequestration, given their physical protection of SOC from microbial attack (Kravchenko et al., 2011; Ananyeva et al., 2013; Lehmann and Kleber, 2015). Dynamics of macro-aggregates may be especially good indicators of potential C responses to land use changes and management (Madari et al., 2005). Moreover, macro-aggregates were found to contain the greatest quantities of organic C (Castro Filho et al., 2002).

As a result, they are thought to play a vital role in SOC stabilization especially in agricultural soils (Elliott, 1986; Pulleman and Marinissen, 2004). In arid areas there has been an increasing need for large-scale cultivation in desert regions to feed human populations. In China, although the oasis area accounts for only approximately 5% of the dry area (Han, 2001), the conversion of dry land into oasis can be of great significance to the ecosystem services due to agricultural production (Zhang et al., 2017). Large amounts of irrigation and fertilizer have to be applied to soils for crop growth and yield in dry lands. Moreover, these applications influence not only the topsoil where water and fertilizer are directly applied, but also the deep soil layers due to leached substances and altered root systems (Li et al., 2013). This can subsequently change the SOC profile. Our previous study on cotton fields revealed that after > 10 years of desert reclamation, SOC significantly

Abbreviations: SOC, soil organic carbon; C, carbon ⁎ Corresponding author at: Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, No 818 South Beijing Road, Urumqi, 830011, Xinjiang, China. E-mail address: [email protected] (Y. Li). https://doi.org/10.1016/j.catena.2018.09.044 Received 29 October 2017; Received in revised form 9 September 2018; Accepted 22 September 2018 0341-8162/ © 2018 Published by Elsevier B.V.

Catena 172 (2019) 770–780

C. Li et al.

(Chen et al., 2003). The zonal soil is Aridisol, mainly under the Calciorthid great group (USDA Soil Taxonomy) with low soil organic matter (SOM) (0.2–1.5%) and high salt (18–91 g kg−1) and pH (8.1–9.7). The basic features for the five oases are shown in Table 1. The five oases had similar mean annual temperature, dominant vegetation (Table 1), and soil texture (silt loam) (Supplementary Table 1). As a result, conventional and similar soil management has been carried out in these oasis farmlands. The applied fertilizer was chemical fertilizer with more nitrogen (N) and phosphorus (P) and less potassium (K), due to a general lack of N and P in this soil (Chen et al., 2008; Lai et al., 2014). On average, the application rates for N, P, and K were 222, 60 and 15 kg ha−1 year−1 in the form of urea, calcium superphosphate, and potassium sulfate, respectively. Flooding irrigation is common practice in these oases, with the amount of irrigation at 600 mm year−1. The total amount of irrigation is divided into 5–8 irrigation events each year, depending on crop species. After harvesting, a short stubble height (< 15 cm) were left in fields, which were subsequently plowed to a soil depth of 20 cm each year. The amount of crop straw crushed and returned to fields was approximately 2.1 million tonnes annually for the five oases, and its equivalent of C were 0.9 million tonnes (Li and Wang, 2016; Zhang and Pang, 2016). The local crops are mainly wheat, cotton, corn, beet, and tomato. According to our previous investigation, local farmers usually determine the crop species planted in the next year based on income in the current year. This meant that the longer reclamation lasts, the more complex is the history of planting crops. Therefore, the exact history of cropping system is not known and the effect of cropping system is ignored after 20 years of reclamation for all five oases.

accumulated in deep layers and the rate of C accumulation exceeded the losses in superficial soil layers (Li et al., 2010). Another field experiment on cultivation with winter wheat in arid lands showed that after 20 years of cultivation, all treatments including inorganic fertilizer alone and combined with straw resulted in increased SOC content in the topsoil, but only the treatment with straw increased SOC content in deep soil (Li et al., 2013). Luske and van der Kamp (2009) found that in 30 years of organic agriculture, reclaimed desert soils could sequester C and increase C stock from 3.9 to 28.8–31.8 t C ha−1. These studies suggested that desert reclamation had C sequestration potential. However, these previous studies were conducted mainly on oasis soils with a monoculture crop or specific management after a given time period of reclamation. There are no data on the temporal dynamics of C sequestration potential of oases, and it is still unclear how C accumulation in a desert environment is affected by cropping systems and management practices such as inorganic fertilization. This study focused on irrigated farmland in arid regions. Water is the key in this process of oasis formation and development (Zhang et al., 2012); runoff from mountains is used to irrigate farmland and create the oases. Since the 1950s, large areas of desert have been converted into farmland between the southern edge of the Gurbantonggut Desert and the northern foot of Tianshan Mountains in Central Asia (Wang et al., 2002), resulting in extensive distribution of oases in these areas. Usually, land nearest to the mountains (thus nearest to the runoff) was cultivated first, with newer cultivated land closest to desert. Hence, farmlands in this region include old oases reclaimed hundreds of years ago and new oases cultivated since the 1950s. Reclamation is usually conducted in only part of the areas where water is available; as a result, every pieces of farmland is usually adjacent to the native desert it originates from. We selected five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) located between the southern Gurbantonggut Desert and the northern Tianshan Mountains. In these oases, an investigation was carried out on farmlands reclaimed 3, 5, 10, 20, 50 or > 100 years ago and their desert pairs to compare the changes (0–200 cm) in SOC and aggregate structure. The objective was to determine the influence of oasis age on dynamics of SOC and aggregates and seeks answers to the following questions:

2.2. Soil sampling and analysis We selected the sampling sites of oasis farmlands with the time series of 3, 5, 10, 20, 50, and > 100 years of reclamation and their adjacent deserts distributed in each of the five oases. According to previous investigation, old oases reclaimed for over 50 years are usually located at the side of the river where water is most abundant, with stones found at 2.0 m soil depth. Given this, the sampling depth was set as −200 cm in this study. Soil samples were collected during June–October 2015 (after crop harvest) at the following depth intervals (cm): 0–20, 20–40, 40–60, 60–100, 100–150 and 150–200. Soil samples were taken vertically by auger to avoid disturbing the soil profiles. Five sampling points were chosen randomly in each farmland for each reclamation year for each oasis. Three composite samples were obtained from five different replicates in the same layer. Meanwhile, the plant cover (Table 1) was investigated in the adjacent desert of each farmland site, and desert soil samples were collected using the previously described method (Li et al., 2010). Some soil samples were stored in ziplock bags at 4 °C until microbial measurements were carried out within 2 weeks. The remaining soil samples were air-dried after removal of visible soil organisms and plant residues. Subsequently, they were passed through a 2-mm sieve for determination of soil properties or ground further to pass a 0.25-mm sieve for SOC determination, except for samples for aggregate analysis. Additional triplicate samples from the topsoil (0–20 cm) were taken using a cutting ring (volume

(1) Does oasis formation lead to consistent accumulation of SOC and nutrients throughout the soil profile? (2) Is there a positive relationship between aggregate size and SOC? 2. Materials and methods 2.1. Site description and experimental design We studied five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) mainly distributed in the sloping plain and the flat alluvial plain at the northern foot of the Tianshan Mountains in western China. These areas are typical temperate desert under a continental arid temperate climate. The desert in these regions is either covered mainly with halophyte vegetation dominated by shrubs and semi-shrubs (Table 1) or bare soil (Xu and Li, 2006). These plant communities are species-poor and most are single-layer structures with low coverage

Table 1 Basic characteristics of five oases used as sampling sites at the northern foot of the Tianshan Mountains, Central Asia. Oases

MAP (mm)

MAT (°C)

Altitude (m)

Longitude

Latitude

Plant cover (%)

Dominant vegetation

Manasi Shuimo Sangong Sigong Jimsar

173 190 164 163 212

6.9 6.8 6.9 6.9 7.0

334–706 465–635 452–581 481–557 563–653

84°42′–86°33′ 86°37′–88°58′ 87°50′–88°10′ 88°07′–88°32′ 88°30′–89°30′

43°27′–45°58′ 42°42′–44°08′ 44°05′–44°24′ 43°23′–44°14′ 43°00′–45°23′

25–36 30–40 31–41 31–41 33–42

Tamarix ramosissima; Haloxylon ammodendron; Reaumuria songonica; Kalidium caspicum

MAP: mean annual precipitation; MAT: mean annual temperature. 771

Catena 172 (2019) 770–780

C. Li et al.

100 g cm−3) to measure the soil bulk density in order to calculate SOC stock. The SOC (g C kg−1 soil) and soil total N (STN) (g N kg−1 soil) contents were measured using Total Organic Carbon/Total Nitrogen (TOC/ TN) analyzer (Multi C/N 3100, Analytik Jena, Germany). The SOC stock (t C ha−1) was calculated using the following equation: SOC = 0.1 × CS × h × s × ρ, where CS is soil organic carbon content (g C kg−1), h is the thickness of soil layer (cm), s is the unit area (1 ha), and ρ is soil bulk density (g cm−3). Microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) were measured using the fumigation-extraction method combined with the TOC/ TN analyzer. Soil total P (STP) (g P kg−1 soil) was determined by acid melted molybdenum, antimony, and scandium colorimetry (Murphy and Riley, 1962). Soil available N (SAN) (mg N kg−1 soil) and available P (SAP) (mg P kg−1 soil) were measured using the alkaline hydrolysis diffusion method (Greenfield, 2001) and the Olsen method (Olsen et al., 1954) respectively. The electrical conductivity (EC) and pH were measured using the conductivity method and potentiometry (at a soil/water ratio of 1:5) respectively (Vogel, 1994). The soil particle sizes (soil texture) were determined by Mastersizer 2000 (Malvern Instruments, Malvern, UK) and classified by the Udden-Wentworth scale standard. Soil aggregate size fractionation was determined using a wet sieving method (Elliott and Cambardella, 1991). Briefly, aggregates were separated through the upper and lower sieves (0.25 and 0.053 mm). The air-dried soil (100 g; < 10 mm) was spread on top of the 0.25-mm sieve and then submerged in a bucket of deionized water for 10 min, and aggregate separation was achieved by shaking samples for 15 min at 30 times min−1 using oscillation apparatus. Water-stable aggregates retained by each sieve were backwashed into preweighed containers after removal of visible plant residues, single sand or stone particles, oven dried at 50 °C for ≥16 h, and weighed. The amounts of true aggregates were obtained by subtracting the above plant residues and particles, which were also weighed. Soil that passed through the 0.053mm sieve (the smallest aggregate fraction) was not collected. The size of this fraction was determined by calculating the difference between whole soil and the sum of the two aggregate size fractions (> 0.25 and 0.25–0.053 mm). Subsamples of aggregate-size fractions were ground until they passed through a 0.25-mm sieve and were then analyzed for SOC and STN.

formation (Fig. 1 C-F). Although there were significant losses in the beginning of reclamation, these nutrient contents significantly increased throughout the profile after 20 years of cropping compared to the initial values (P < 0.05), and reached maxima after 50 years of cropping. Among them, soil N contents were highest at 50 years of reclamation and did not significantly decline after 100 years (Fig.1C, E). 3.2. SOC contents Desert reclamation had a very significant effect on organic C contents throughout the soil profile (P < 0.001; Table 2). In the first five years of reclamation, SOC contents decreased by 2.8–47.1% throughout the soil profile (Fig. 2A–C), and significantly decreased (by 22–47%) in the deep layer (60–200 cm) (P < 0.05; Fig. 2C). However, SOC contents increased significantly in topsoil and subsoil (0–60 cm) after 10 years of cropping (Fig. 2A, B) and throughout the whole profile (0–200 cm) after 20 years of cropping (P < 0.05). The oasis farmlands reclaimed for 50 years had the highest contents of SOC in the soil profile, which increased by 67–135% compared to initial values. In addition, SOC contents did not significantly decrease after 100 years of reclamation. The five oases presented similar trends with years of reclamation. For example, SOC contents throughout the profile reached their maxima in the 50th year (Fig. 3A–O); and there were significant SOC losses in the deep soil (60–200 cm) at the beginning of reclamation (Fig. 3K–O) (P < 0.05). The smallest increase of 33–90% throughout the profile was for Manasi (Fig. 3A, F and K), and the largest increase of 104–258% was for Jimsar (Fig. 3E, J and O). In the topsoil, after 50 years of reclamation, the increase in SOC contents in Manasi was from 5.9 ± 1.0 g kg−1 in the desert to 7.9 ± 1.2 g kg−1 in farmland (on average) (Fig. 3A); and the corresponding change in Jimsar was from 7.9 ± 1.3 to 16.1 ± 1.7 g kg−1 (Fig. 3E). The SOC stocks also exhibited similar trends (Table 3). During the beginning of oasis formation, the two oases had significant increases in SOC contents in topsoil (0–20 cm) and subsoil (20–60 cm) (P < 0.05), but the other three oases (Sangong, Sigong and Shuimo) experienced SOC losses (Fig. 3 B–D and G–I). Moreover, SOC contents significantly reduced throughout most of the profile during the first 10 years of reclamation in Sangong and Sigong (P < 0.05). The MBC contents significantly increased throughout the soil profile (0–200 cm) after reclamation (P < 0.05; Table 4), and the values reached maxima after 50 years of reclamation and stabilized afterward.

2.3. Data and statistical analysis Given that compaction after reclamation mainly occurs in the upper 60 cm of soil (Li et al., 2010), the total soil profile was divided into three layers: topsoil (0–20 cm), subsoil (20–60 cm) and the deep layer (60–200 cm), to make comparisons easier. Statistical analyses were carried out using the SPSS 11.5 package for Windows. Descriptive statistics were used to calculate averages and standard errors of the data. The data were normally distributed, as verified by test of normality and homogeneity of variance test. Under this premise, analysis of variance (ANOVA) and least significant difference (LSD, P < 0.05) tests were used to assess the significance of the effects of desert reclamation, reclamation years and soil depth on soil properties and aggregate structure. Testing of between-subjects effects was performed to evaluate the interaction effects of desert reclamation and oases.

3.3. Aggregate size distribution In the desert soil, the lowest proportion was the macro-aggregate fraction (> 0.25 mm) representing < 1% throughout the soil profile, which was mainly distributed in the topsoil (Fig. 4A). The whole profile mainly comprised two aggregate fractions: micro-aggregate (0.053–0.25 mm) and smallest aggregate (< 0.053 mm) fractions (Fig. 4B, C). The aggregate size distribution was dominated by the micro-aggregate fraction in topsoil and subsoil (0–60 cm) (Fig. 4B). However, in deep layer (60–200 cm), the smallest aggregate fraction had the highest proportion (70–80%) due to its increase with soil depth (Fig. 4C). Desert reclamation was also closely associated with changes in aggregate size distributions (P < 0.05; Table 2). Differences among oases had very significant effects on macro-aggregates (P < 0.001). Moreover, desert reclamation and differences among oases had a significant interaction effect on the macro-aggregate fraction (P < 0.05). Macroaggregate fractions significantly increased throughout the soil profile after reclamation (P < 0.05), and reached the highest proportion (1–17%) after 50 years (Figs. 4A and 5A). Jimsar showed the largest increases, and Manasi the least (Fig. 4A). The micro-aggregate distribution significantly decreased and the smallest aggregate fraction significantly increased (P < 0.05; Fig. 5B, C) throughout the soil

3. Results 3.1. Basic soil properties Desert reclamation significantly affected soil chemical properties and nutrients throughout the soil profile (0–200 cm) (P < 0.05), and the influence decreased with soil depth (Fig. 1). Compared to the initial values (i.e., desert soil), soil EC and pH significantly decreased throughout the soil profile for all farmlands (P < 0.05; Fig. 1A, B). Overall, STN, STP, and the available nutrients increased during oasis 772

Catena 172 (2019) 770–780

C. Li et al.

Fig. 1. Changes of selected soil properties [electrical conductivity (EC), pH, soil total nitrogen (STN), soil total phosphorus (STP), available N (SAN), and available P (SAP)] at 0–20, 20–60, and 60–200 cm depths during oasis formation with different reclamation years (3, 5, 10, 20, 50, and > 100), compared with desert soils. Means of reclamation time within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

Table 2 ANOVA results of aggregate size distribution, soil organic carbon (SOC) size fractions, microbial biomass carbon (MBC), and total SOC from five oases at the northern foot of the Tianshan Mountains. Aggregate size distribution > 0.25 mm Reclamation F value 21.457 Sig. 0.000*** Oases F value 10.051 Sig. 0.000*** Reclamation vs. oases F value 2.255 Sig. 0.005**

SOC size fractions

MBC

SOC

0.25–0.053 mm

< 0.053 mm

> 0.25 mm

0.25–0.053 mm

< 0.053 mm

4.766 0.004**

2.731 0.025*

13.476 0.000***

1.104 0.369

4.397 0.001**

3.759 0.003**

3.870 0.002***

5.466 0.005**

4.011 0.009**

5.384 0.001**

2.475 0.052

2.147 0.084

2.501 0.048*

2.327 0.054

0.253 1.000

0.257 1.000

0.880 0.626

0.249 1.000

0.201 1.000

2.523 0.039*

0.243 1.000

Statistical significance: *P < 0.05, **P < 0.01, and ***P < 0.001.

aggregate, micro-aggregate and smallest aggregate fractions accounted for 0.2–1.5, 10–15 and 82–93% of the total SOC, respectively; and the corresponding values in oasis soils were 1.2–13.1, 9–22 and 69–82% (Fig. 7A–C). The contribution of SOC content in the macro-aggregate fraction to total SOC significantly increased throughout the soil profile during oasis formation (P < 0.05; Fig. 7A). It reached maxima after 50 years (Fig. 7A), with the greatest and smallest increases in Jimsar and Manasi, respectively (Fig. 6A). Meanwhile, the C contribution in the micro-aggregate fraction increased for most of the reclamation years (P < 0.05; Fig. 7B). The SOC content in the smallest aggregate fraction clearly made up the highest percentage of total SOC in all oases (Fig. 7C). However, the proportional contribution of this C fraction to total SOC significantly decreased after reclamation (P < 0.05; Fig. 7C). The change trends of the above two C fractions contrasted with the amounts of their respective aggregate fractions.

profile after 5 years of reclamation. Among the three fractions, the macro-aggregate fraction showed the greatest increases and the most significant variations among reclamation years; the coefficients of variation during oasis formation were 80–100, 14–21, and 2–13% for the macro-aggregate, micro-aggregate, and smallest aggregate fractions, respectively (Fig. 5). 3.4. Aggregate-size associated carbon In all five oases, the highest concentration of SOC was in the macroaggregate fraction (3–146 g kg−1 aggregates) throughout the soil profile, intermediate in the smallest aggregate fraction (0.9–15.1 g kg−1 aggregates), and lowest in the micro-aggregate fraction (0.5–9.5 g kg−1 aggregates). The macro-aggregate associated SOC significantly increased throughout the soil profile following reclamation (P < 0.05; Fig. 6A), but the other two C fractions did not show this trend clearly (Fig. 6B, C). Moreover, in desert soils the SOC stored in the macro773

Catena 172 (2019) 770–780

C. Li et al.

3.6. Correlation analysis Soil EC and pH were negatively correlated with total SOC (P < 0.05; Fig. 8A, B). The STN was most strongly related to total SOC (P < 0.001; Fig. 8C), and STP was positively correlated with total SOC (P < 0.01; Fig. 8D). The MBC content showed a significant and positive correlation with total SOC (P < 0.01; Table 7). Furthermore, total SOC was significantly related to the three aggregate fractions in the soil profile, especially the macro-aggregate fractions (Table 7). There were also significant correlations between total SOC and the associated C in the macro-aggregate and smallest aggregate fractions (P < 0.001). The MBC content had the most significant and positive correlation with macro-aggregate fractions and their associated C (P < 0.01; Table 7). In addition, STP and SAN, and especially STN, were related to the macro-aggregate fractions and SOC contents in macro-aggregate and smallest aggregate fractions (P < 0.01; Supplementary Table 3). There were positive correlations between SAP and macro-aggregates and the associated C (P < 0.05). 4. Discussion 4.1. SOC dynamics during oasis formation Our study showed significant SOC accumulation in soil profiles during oasis formation (Figs. 2 and 3). The STN had the most significant correlation with SOC content (Fig. 8). Soil N availability regulates microbial communities and the decomposition and formation of SOM, which have great impacts on C cycling (Zhou et al., 2017). After water, N is the second limiting factor for plant growth in arid areas (Huang et al., 2015). As a result, N fertilizer is the overwhelmingly dominant fertilizer in local agricultural systems (Wang et al., 2002). Fertilizer application (especially N) on the basis of irrigation forms high primary productivity, which would input significant amounts of C to desert soils by dense roots and crop residues. Crop residue incorporation has a positive effect on C sequestration (Lugato et al., 2006; Russell et al., 2009). In addition, the incorporation of crop residues into soils with low antecedent SOC levels frequently also results in increases in SOC (Kong et al., 2005). Desert soils are likely to have high potential for C sequestration due to a low baseline of C content of the system. Studies form Mexico (Werner et al., 2003) suggested that reclamation of indurated volcanic soil provokes an accumulation of SOC due to a low C baseline, but also results in a severe N-loss due to the increase of runoff and sediment transport, which consequently causes an increment of soil C/N ratio. However, in the present study, the soil C/N ratios decreased significantly as SOC content increased during oasis formation (Table 5). Because the oases are flat, the runoff effect of nutrients (e.g., N and P) is relatively weak. This present study showed a lower C/N ratio (9.0–11.3) throughout the soil profile after 20 years of reclamation (Table 5). This indicated that incorporation of N, a more humidified state and potentially larger degree of microbial origin (Six et al., 2001). In the study areas, the deep soil layer inevitably experiences a small quantity of leached substances (e.g. dissolved organic C lixiviated from fresh organic matter and dissolved nutrients) from above due to irrigation at the beginning of reclamation. Fresh SOM inputs increase microbial activity and then stimulate the extra-mineralization of preexistent SOC in deep soil layers (Fontaine et al., 2007). This was likely an important reason why SOC contents in deep soil layers of all five oases significantly decreased at the beginning of reclamation (Fig. 3). Our study also confirmed this point: MBC contents increased significantly (Table 4) and soil C/N and MBC/MBN ratios declined sharply in deep soil layers at the beginning of reclamation for all oases (Tables 5 and 6). With cultivation going on, crop planting history becomes complicated, which means that a variety of plant residues would be added into superficial soil layers. These accumulated substances would be transported downward with irrigation water and be absorbed in deep soil layers (Li et al., 2013) due to higher clay contents in deep soil

Fig. 2. Soil organic carbon (SOC) contents at 0–20, 20–60, and 60–200 cm depths during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100). 0: desert soils. Means of reclamation time within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

3.5. Carbon to nitrogen (C/N) and MBC/MBN ratios Soil C/N ratios decreased significantly at the beginning of reclamation, especially in the deep soil layer (P < 0.05; Table 5), and increased with reclamation years and tended to be stable after 20 years. However, the values throughout the soil profile were still significantly lower than the initial value of desert soils (P < 0.05). The MBC/MBN ratios also significantly decreased throughout the soil profile in the first 10 years of reclamation (P < 0.05; Table 6), especially in the deep layer. The values tended to stabilize after 10 years of reclamation, but were significantly less than the initial value (P < 0.05). In addition, soil C/N ratios declined significantly in the aggregate fractions (0.25–0.053 and < 0.053 mm) throughout the soil profile in the first 10 years of reclamation (P < 0.05; Supplementary Table 2). The biggest decreases were in the smallest fraction (< 0.053 mm) from the deep layer. After 10 years of reclamation, soil C/N ratios in all aggregate fractions increased significantly (P < 0.05), and tended to be stable thereafter. The values in the two fractions (0.25–0.053 and < 0.053 mm) were still significantly less than the initial values throughout the soil profile; whereas the C/N ratios in macro-aggregate fractions were significantly higher than initial values (P < 0.05; Supplementary Table 2).

774

Catena 172 (2019) 770–780

C. Li et al.

Fig. 3. Changes of soil organic carbon (SOC) at 0–20 cm (A–E), 20–60 cm (F–J), and 60–200 cm (K–O) depths during oasis formation with different reclamation years (3, 5, 10, 20, 50, and > 100), compared with desert soils, in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. Error bars represent the standard error of the mean (n = 3). Means of reclamation time followed by the same lowercase letters are not significantly different (P > 0.05).

(8–15) (Paul and Clark, 1996). Diverse substance input and the abovementioned microbial shift were likely to promote SOC accumulation without provoking an extra-mineralization of pre-existent soil C after 20 years of reclamation. Therefore, all five oases showed similar changes: SOC content significantly increased throughout the soil profile

layers with the increase of reclamation time (Supplementary Table 1). Meanwhile, the MBC/MBN ratio throughout the soil profile decreased from its initial value of 8.6–11.3 to 6.1–8.5 after 20 years of reclamation (Table 6), indicating significant shifts from fungal to bacterial dominance, given a narrower C/N ratio (4–7) in bacteria than in fungi

Table 3 Soil organic carbon stocks in the topsoil (0–20 cm) during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100) in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. SOC storage (t Cha−1)

Reclamation time (years) 0

Manasi Shuimo Sangong Sigong Jimsar

16.3 18.9 17.5 16.9 19.8

3 ± ± ± ± ±

0.3d 0.4f 0.3d 0.4c 0.5e

17.6 20.2 11.7 10.8 23.1

5 ± ± ± ± ±

0.4 cd 0.5e 0.2 g 0.3f 0.7d

18.0 22.8 13.9 12.8 26.0

10 ± ± ± ± ±

0.4c 0.6d 0.2f 0.2e 0.7c

18.6 25.3 15.7 16.1 28.4

20 ± ± ± ± ±

0.4bc 0.7c 0.3e 0.4d 1.0b

18.8 25.5 22.0 19.6 29.8

Values are means with standard errors of means (n = 3). Means of reclamation time followed by the same lowercase letters are not significantly different (P > 0.05). 775

50 ± ± ± ± ±

0.5bc 0.8bc 0.5c 0.6b 0.9b

21.2 35.6 30.9 27.6 42.3

> 100 ± ± ± ± ±

0.7a 2.0a 1.8a 1.1a 2.5a

19.1 27.7 25.7 26.4 42.0

± ± ± ± ±

0.5b 1.6b 1.6b 1.0a 2.3a

Catena 172 (2019) 770–780

C. Li et al.

Table 4 Microbial biomass carbon (MBC) at 0–20, 20–60, and 60–200 cm depths during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100) in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. MBC (mg C kg−1 soil)

0–20 cm Manasi Shuimo Sangong Sigong Jimsar 20–60 cm Manasi Shuimo Sangong Sigong Jimsar 60–200 cm Manasi Shuimo Sangong Sigong Jimsar

Reclamation time (years) 0

3

5

10

78 ± 6e 107 ± 9d 115 ± 16d 102 ± 12d 140 ± 16e

91 ± 5d 126 ± 8c 149 ± 8c 133 ± 7c 161 ± 10d

97 ± 5d 132 ± 8c 159 ± 9c 145 ± 7c 174 ± 9d

109 156 183 168 195

63 ± 4d 92 ± 3d 98 ± 5d 84 ± 8d 121 ± 8d

77 ± 5c 118 ± 5c 129 ± 5d 114 ± 6c 136 ± 6c

83 ± 5c 125 ± 7c 135 ± 5d 124 ± 6c 145 ± 6c

42 69 77 60 92

64 ± 4c 96 ± 2c 109 ± 3c 91 ± 4d 123 ± 5c

66 ± 5c 99 ± 6c 111 ± 5c 95 ± 5d 124 ± 4c

± ± ± ± ±

3d 2d 3d 4e 5d

20

± ± ± ± ±

6c 10b 13b 8b 10c

50

> 100

152 186 220 208 253

± ± ± ± ±

9b 12a 11a 11a 14b

177 206 239 228 290

± ± ± ± ±

13a 13a 18a 16a 22a

163 190 243 215 269

± ± ± ± ±

10ab 14a 14a 12a 12ab

99 ± 7b 145 ± 9b 153 ± 4c 141 ± 8b 173 ± 9b

115 168 171 160 201

± ± ± ± ±

8a 12a 11a 7a 9a

119 176 183 173 223

± ± ± ± ±

10a 8a 13a 11a 16a

117 172 178 165 212

± ± ± ± ±

7a 11a 10a 10a 11a

71 ± 4bc 114 ± 8b 123 ± 5b 106 ± 5c 139 ± 6b

78 ± 6ab 122 ± 9ab 136 ± 10ab 119 ± 6b 165 ± 10a

85 ± 6a 139 ± 11a 148 ± 13a 139 ± 12a 172 ± 11a

86 ± 7a 135 ± 8a 140 ± 6a 128 ± 8a 168 ± 12a

Values are means with standard errors of means (n = 3). Means of reclamation time within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

aggregation and the increase in macro-aggregates during oasis formation. The higher C/N ratios (11.6–14.8) associated with macro-aggregates (Supplementary Table 2) suggested that these fractions derived mainly from plants with considerable microbial degradation and immobilization of N (Six et al., 2001). Moreover, our work showed that macro-aggregate fractions (> 0.25 mm) contained the highest SOC concentration among all aggregate fractions, consistent with previous reports (Barthes et al., 2000; Fonte et al., 2014). Puget et al. (1995) observed greater organic C accumulation in macro-aggregates due to less decomposable SOM in these aggregates and direct contribution of SOM to the stability of macroaggregates resulting in C-rich macro-aggregates capable of withstanding slaking. Therefore, macro-aggregate formation is considered to be a mechanism of C sequestration (Mikha and Rice, 2004; Shi et al., 2017). In this study, desert reclamation significantly increased not only C content associated with macro-aggregates (Figs. 5 and 6), but also their C contribution to the total SOC during oasis formation (Fig. 7). Also, both macro-aggregates and associated C showed significant positive correlations with total SOC (Table 7). Therefore, their significant increases since reclamation were most directly associated with SOC

(including deep soil layers) after 20 years of reclamation. 4.2. Relationships between soil aggregates and SOC Oasis formation led to not only accumulation of SOC, but also significant increases in macro-aggregate fractions throughout the soil profile (Figs. 4 and 5). Water-stable macro-aggregates have positive effects on formation and stabilization of soil aggregate structure (Barthes et al., 2008). Our result indicated that soil structure was improved by increased macro-aggregates during oasis formation. This also confirmed previous findings that improved soil aggregation is accompanied by more SOC in agroecosystems (Barthes et al., 2000; Jiao et al., 2006; Fonte et al., 2014). The formation of macro-aggregates can be induced by fresh organic residue and the bonding function between the soil root system and hypha (Six et al., 1999; Bronick and Lal, 2005). In the present study, desert reclamation leads to the annual return of large amounts of crop residues and roots. Meanwhile, microbial biomass also significantly increased (Table 4) throughout the soil profile due to the improved soil environment (Fig. 1). These changes were likely to jointly drive soil

Fig. 4. Distributions of water-stable aggregates among different aggregatesize classes (> 0.25, 0.053–0.25, and < 0.053 mm) at 0–20, 20–60, and 60–200 cm depths as influenced by desert reclamation for 50 years in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. Desert: means for the corresponding desert soils from which the oasis was created. Error bars represent the standard error of the mean (n = 3). Means of the oases within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

776

Catena 172 (2019) 770–780

C. Li et al.

Fig. 5. Distributions of water-stable aggregates among different aggregate-size classes (> 0.25, 0.053–0.25, and < 0.053 mm) at 0–200 cm depth during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100). 0: desert soils. Means of reclamation time followed by the same lowercase letters are not significantly different (P > 0.05).

Fig. 7. Contribution proportion of soil organic carbon (SOC) in different aggregate-size fractions (> 0.25, 0.053–0.25, and < 0.053 mm) at 0–200 cm depth during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100). 0: desert soils. Means of reclamation time followed by the same lowercase letters are not significantly different (P > 0.05).

Fig. 6. Distributions of soil organic carbon (SOC) contents among different aggregate size fractions (> 0.25, 0.053–0.25, and < 0.053 mm) at 0–20, 20–60 and 60–200 cm depths as influenced by desert reclamation for 50 years in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. Desert: means for the corresponding desert soils from which the oasis was created. Error bars represent the standard error of the mean (n = 3). Means of the oases within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

777

Catena 172 (2019) 770–780

C. Li et al.

Table 5 Soil carbon to nitrogen ratio (C/N) at 0–20, 20–60, and 60–200 cm depths during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100) in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. Soil C/N

Reclamation time (years) 0

0–20 cm Manasi Shuimo Sangong Sigong Jimsar 20–60 cm Manasi Shuimo Sangong Sigong Jimsar 60–200 cm Manasi Shuimo Sangong Sigong Jimsar

3

5

10

20

50

> 100

11.4 12.7 11.4 12.6 11.2

± ± ± ± ±

0.3a 0.5a 0.4a 0.5a 0.3a

8.7 ± 0.3d 10.1 ± 0.4c 7.2 ± 0.2d 8.5 ± 0.3d 8.9 ± 0.4c

8.9 ± 0.2d 10.3 ± 0.3bc 7.4 ± 0.2d 9.1 ± 0.2 cd 9.0 ± 0.3c

9.8 ± 0.2c 10.6 ± 0.3bc 9.0 ± 0.2c 9.5 ± 0.3c 9.4 ± 0.2c

10.5 ± 0.3b 10.9 ± 0.3b 9.9 ± 0.2b 10.5 ± 0.3b 9.9 ± 0.2b

10.6 11.2 10.1 10.9 10.1

± ± ± ± ±

0.3b 0.4b 0.2b 0.4b 0.3b

10.4 11.0 10.1 10.4 10.0

± ± ± ± ±

0.3b 0.3b 0.3b 0.3b 0.3b

11.2 12.1 12.0 12.8 12.3

± ± ± ± ±

0.4a 0.5a 0.5a 0.6a 0.5a

9.1 8.1 7.8 8.3 9.4

± ± ± ± ±

0.3c 0.3d 0.3d 0.3d 0.3c

9.2 8.3 8.1 8.5 9.5

± ± ± ± ±

0.3c 0.2d 0.2 cd 0.2d 0.2c

9.5 9.5 8.5 9.7 9.6

± ± ± ± ±

0.2c 0.3c 0.3c 0.3c 0.2c

10.1 ± 0.3b 10.7 ± 0.3b 9.9 ± 0.3b 10.8 ± 0.3b 10.2 ± 0.3b

10.4 11.0 10.2 11.3 10.5

± ± ± ± ±

0.2b 0.4b 0.3b 0.5b 0.4b

10.2 ± 0.2b 10.8 ± 0.4b 9.8 ± 0.3b 11.2 ± 0.4b 10.3 ± 0.3b

11.6 12.1 13.0 13.8 13.0

± ± ± ± ±

0.5a 0.5a 0.6a 0.7a 0.5a

7.4 7.3 7.2 7.8 8.2

± ± ± ± ±

0.3d 0.2e 0.3d 0.3e 0.3d

7.5 8.0 7.3 8.4 8.2

± ± ± ± ±

0.2d 0.3d 0.2d 0.2d 0.2d

8.3 8.9 7.9 9.3 8.9

± ± ± ± ±

0.2c 0.3c 0.2c 0.2c 0.3c

9.0 ± 0.2b 9.8 ± 0.3b 9.0 ± 0.3b 10.0 ± 0.3b 9.8 ± 0.4b

9.3 ± 0.3b 10.1 ± 0.4b 9.4 ± 0.3b 10.2 ± 0.4b 10.3 ± 0.4b

9.2 ± 0.3b 10.1 ± 0.3b 9.5 ± 0.3b 10.1 ± 0.3b 9.9 ± 0.3b

Values are means with standard errors of means (n = 3). Means of reclamation time within a depth followed by the same lowercase letters are not significantly different (P > 0.05).

smallest aggregate fraction to total SOC was significantly reduced at the same time (Fig. 7). Furthermore, the C/N ratio in the smallest aggregate fractions had the greatest reduction during the beginning of oasis formation and had the lowest value (7.0–9.6) among the three aggregate fractions in oasis farmlands (Supplementary Table 2), which meant faster SOM decomposition or mineralization in these fractions (Cong et al., 2015). These results suggested, in the study areas, that SOC in the smallest aggregate fraction was likely less stable and easily consumed compared with other aggregate fractions during reclamation, although it represented the dominant proportion.

accumulation during oasis formation. Such close associations between SOC and macro-aggregates were also confirmed by the different oases in the arid zone: Jimsar had the greatest increase in total SOC accompanied by the greatest increase in the macro-aggregate fraction, and Manasi had the smallest increases in both these components (Figs. 3, 4 and 6). This could be explained by differences in soil background characteristics among these oases resulting from different geographical locations (Table 1). In addition, SOC content in the smallest aggregate fraction (< 0.053 mm) contributed most to the total SOC in the study areas (Figs. 6 and 7), because this fraction accounted for a large proportion of soil mass (Fig. 4). The associated C fraction was also significantly correlated with total SOC (Table 7). As a result, SOC stored in the smallest aggregate fraction could largely determine the dynamics of total SOC during oasis formation. The amount of the smallest aggregate fractions increased after reclamation (Fig. 5); however, the C contribution in the

5. Conclusions Oasis formation, with reclamation years, revealed increases in SOC and nutrient contents compared with adjacent desert. Significant accumulation of SOC throughout the profile (0–200 cm) occurred after

Table 6 Ratios of microbial biomass carbon to microbial biomass nitrogen (MBC/MBN) at 0–20, 20–60, and 60–200 cm depths during oasis formation with different reclamation years (0, 3, 5, 10, 20, 50, and > 100) in five oases (Manasi, Shuimo, Sangong, Sigong, and Jimsar) north of the Tianshan Mountains. MBC/MBN

0–20 cm Manasi Shuimo Sangong Sigong Jimsar 20–60 cm Manasi Shuimo Sangong Sigong Jimsar 60–200 cm Manasi Shuimo Sangong Sigong Jimsar

Reclamation time (years) 0

3

5

10.0 ± 0.5a 10.9 ± 0.6a 9.9 ± 0.5a 11.3 ± 0.7a 10.2 ± 0.5a

9.4 ± 0.4ab 9.9 ± 0.5ab 8.5 ± 0.3ab 10.4 ± 0.4ab 9.8 ± 0.5a

9.2 9.6 8.1 9.7 9.5

± ± ± ± ±

0.4ab 0.3b 0.3bc 0.4bc 0.3a

9.0 9.3 7.7 9.3 9.2

± ± ± ± ±

0.4b 0.4b 0.3 cd 0.4c 0.4b

8.0 8.5 7.2 7.8 8.3

± ± ± ± ±

0.3c 0.3c 0.3de 0.3d 0.4c

7.8 8.4 7.3 8.1 8.0

± ± ± ± ±

0.3c 0.3c 0.3de 0.3d 0.3c

7.9 8.5 6.9 8.2 8.1

± ± ± ± ±

0.3c 0.2c 0.3e 0.3d 0.3c

9.2 ± 0.4a 9.9 ± 0.5a 9.8 ± 0.4a 10.2 ± 0.6a 9.9 ± 0.5a

8.6 6.4 6.0 6.4 9.3

± ± ± ± ±

0.4a 0.4d 0.2d 0.3c 0.5ab

7.7 6.5 6.4 6.7 8.8

± ± ± ± ±

0.3b 0.3d 0.2d 0.2c 0.4bc

7.6 7.2 6.9 7.0 8.4

± ± ± ± ±

0.3b 0.3c 0.2c 0.3c 0.3c

7.4 8.0 7.0 7.7 7.6

± ± ± ± ±

0.3b 0.3b 0.3bc 0.3b 0.3d

7.5 8.1 7.1 7.9 7.5

± ± ± ± ±

0.2b 0.4b 0.3bc 0.4b 0.2d

7.5 8.1 7.0 7.7 7.5

± ± ± ± ±

0.3b 0.4b 0.2b 0.3b 0.3d

8.6 9.1 8.7 9.4 9.0

5.0 5.4 5.5 5.6 5.6

± ± ± ± ±

0.4d 0.3d 0.3d 0.3d 0.3d

5.8 6.3 5.7 5.9 6.3

± ± ± ± ±

0.3c 0.2d 0.3 cd 0.2d 0.2c

7.0 6.8 6.1 6.7 6.9

± ± ± ± ±

0.3b 0.2c 0.3c 0.2c 0.3b

7.2 7.4 6.8 7.4 7.3

± ± ± ± ±

0.3b 0.3b 0.3b 0.3b 0.3b

7.4 7.8 7.0 7.6 7.4

± ± ± ± ±

0.4b 0.4b 0.3b 0.3b 0.4b

7.4 7.5 7.0 7.6 7.2

± ± ± ± ±

0.3b 0.4b 0.3b 0.4b 0.3b

± ± ± ± ±

0.4a 0.4a 0.6a 0.6a 0.4a

10

20

50

Values are means with standard errors of means (n = 3). Means of reclamation time within a depth followed by the same lowercase letters are not significantly different (P > 0.05). 778

> 100

Catena 172 (2019) 770–780

C. Li et al.

Fig. 8. Correlations of soil electrical conductivity (EC), pH, total nitrogen (STN), and total phosphorus (STP) with soil organic carbon (SOC). Statistical significance: *P < 0.05, **P < 0.01 and ***P < 0.001. Table 7 Correlation coefficients (r) between aggregate size distribution, soil organic carbon (SOC) size fractions, microbial biomass carbon (MBC), and total SOC. Aggregate size distribution

Total SOC (r) 0–20 cm 20–60 cm 60–200 cm MBC (r) 0–20 cm 20–60 cm 60–200 cm

SOC size fractions

MBC

> 0.25 mm

0.25–0.053 mm

< 0.053 mm

> 0.25 mm

0.25–0.053 mm

< 0.053 mm

0.805*** 0.763** 0.714**

−0.402* −0.431* −0.487*

0.451* 0.394* 0.445*

0.816*** 0.828*** 0.842***

0.293 0.271 0.291

0.801*** 0.801*** 0.854***

0.801*** 0.871*** 0.852***

−0.633** −0.771** −0.443*

0.416* 0.329* 0.193

0.867*** 0.855*** 0.765**

0.292 0.281 0.213

0.654** 0.679** 0.711**

0.689** 0.778** 0.832***

Statistical significance: *P < 0.05, **P < 0.01, and ***P < 0.001.

20 years of reclamation, although there was SOC loss in the first 10 years. The SOC content reached maxima after 50 years of reclamation, and SOC stock increased from 17.9 to 31.5 t C ha−1 (on average) in the topsoil. Meanwhile, the macro-aggregates (> 0.25 mm) and the associated C significantly increased, and their C contribution to the total SOC also increased since reclamation. Our study revealed significant correlations between SOC content and macro-aggregates and the associated C. These changes were closely correlated with the increase in soil N contents during oasis formation. Desert reclamation decreased soil salt and pH and increased nutrients through fertilizer application and irrigation, which improved the soil environment, increased microbial biomass, and also led to inputs of more crop residues with diverse quality throughout the soil profile. This is very likely to be the major reason for the higher SOC and more macro-aggregates in

farmlands than in adjacent desert. This study emphasizes the high potential of C sequestration throughout the soil profile and the positive association between macro-aggregates and SOC accumulation during oasis formation.

Acknowledgments We thank Kang Jinhua and Zhang Hui for assistance with soil characterization and all staff at the Fukang Station of Desert Ecology. This work was supported by the National Natural Science Foundation of China (No. 41671114); ‘Western Light’ program of the Chinese Academy of Science (No. 2015-XBQN-A-06); and Key Research Program of Frontier Sciences of CAS (No. QYZDJ-SSW-DQC014). 779

Catena 172 (2019) 770–780

C. Li et al.

Appendix A. Supplementary data

Li, X.X., Wang, Z., 2016. Estimation and spatial and temporal distribution of biological resources from main crop straws in Xinjiang. J. HeiLongJiang Vocat. Inst. Ecol. Eng. 29, 3–4. Li, C.H., Li, Y., Tang, L.S., 2010. Soil organic carbon stock and carbon efflux in deep soils of desert and oasis. Environ. Earth Sci. 60, 549–557. Li, C.H., Li, Y., Tang, L.S., 2013. The effects of long-term fertilization on the accumulation of organic carbon in the deep soil profile of an oasis farmland. Plant Soil 369, 645–656. Lugato, E., Berti, A., Giardini, L., 2006. Soil organic carbon (SOC) dynamics with and without residue incorporation in relation to different nitrogen fertilization rates. Geoderma 135, 135–321. Luske, B., van der Kamp, J., 2009. Carbon sequestration potential of reclaimed desert soils in Egypt. In: Soil and More International Website. Louis Bolk Instituut & Soil and More International (Available:). http://www.soilandmore.nl/userfiles/ CarbonSequestrationPotentialReportFinal.pdf, Accessed date: 16 March 2011. Madari, B., Machado, P., Torres, E., de Andrade, A.G., Valencia, L.I.O., 2005. No tillage and crop rotation effects on soil aggregation and organic carbon in a Rhodic Ferralsol from southern Brazil. Soil Tillage Res. 80, 185–200. Mikha, M.M., Rice, C.W., 2004. Tillage and manure effects on soil and aggregate-associated carbon and nitrogen. Soil Sci. Soc. Am. J. 68, 809–816. Murphy, J., Riley, J,.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 31–36. Olsen, R., Cole, C.V., Watanabe, F.S., Dean, L.A., 1954. Estimation of Available Phosphorus in Soils by Extraction With Sodium Bicarbonate. USDA Circular No. 939 USDQ. Government printing office, Washington DC, pp. 1–19. Paul, E.A., Clark, F.E., 1996. Components of the soil biota. In: Paul, E.A., Clark, F.E. (Eds.), Soil microbiology and biochemistry. Academic Press, San Diego, pp. 71–107. Puget, P., Chenu, C., Balesdent, J., 1995. Total and young organic matter distributions in aggregates of silty cultivated soils. Eur. J. Soil Sci. 46, 449–459. Pulleman, M.M., Marinissen, J.C.Y., 2004. Physical protection of mineralizable C in aggregates from long term pasture and arable soil. Geoderma 120, 273–282. Ratnayake, R.R., Perera, B.M.A.C.A., Rajapaksha, R.P.S.K., Ekanayakea, E.M.H.G.S., Kumara, R.K.G.K., Gunaratne, H.M.A.C., 2017. Soil carbon sequestration and nutrient status of tropical rice based cropping systems: Rice-rice, rice-soya, rice-onion and rice-tobacco in Sri Lanka. Catena 150, 17–23. Russell, A.E., Cambardella, C.A., Laird, D.A., Jaynes, D.B., Meet, D.W., 2009. Nitrogen fertilizer effects on soil carbon balances in midwestern U.S. agricultural systems. Ecol. Appl. 19, 1102–1113. Shi, Z.L., Wang, J.X., Liang, H.X., Shi, H.P., Wei, B.M., Wang, Y.Q., 2017. Status and evolution of soil aggregates in apple orchards different in age in Weibei. Acta Pedol. Sin. 54, 387–399. Six, J., Elliott, E.T., Paustian, K., 1999. Aggregate and soil organic matter dynamics under conventional and no-tillage systems. Soil Sci. Soc. Am. J. 63, 1350–1358. Six, J., Guggenberger, G., Paustian, K., Haumaier, L., Elliott, E.T., Zech, W., 2001. Sources and composition of soil organic matter fractions between and within soil aggregates. Eur. J. Soil Sci. 52, 607–618. Vogel, A.W., 1994. Compatibility of soil analytical data: determinations of cation exchange capacity, organic carbon, soil reaction, bulk density and volume percent of water at selected pF values by different methods. In: Working Paper and Preprint 94/ 07. ISRIC, Wageningen. Wang, Z.Q., Li, S.G., Chen, X.J., 2002. Nutrient Circulation in the Farmland Ecosystems in the Desert Oasis. Science Press, Beijing. Werner, G., Gallardo, J.F., Christian, Prat, 2003. Rehabilitation of severely eroded and indurated volcanic ash soils in Mexico and Chile (REVOLSO). In: Geomorphic Hazards: Towards the Prevention of Disasters: Abstracts. Gand: IAG, 72–73. IAG Regional Geomorphology Conference: Geomorphic Hazards: Towards the Prevention of Disasters, Mexico (MEX). Xu, H., Li, Y., 2006. Water-use strategy of three central Asian desert shrubs and their responses to rain pulse events. Plant Soil 285, 5–17. Zhang, L.H., Pang, M.Z., 2016. Quantity estimation and utilization of main crop straws nutrient resources in Xinjiang. In: Phosphate & Compound Fertilizer. 9. pp. 47–50. Zhang, Q.Q., Xu, H.L., Li, Y., Fan, Z.L., Zhang, P., Yu, P.J., Ling, H.B., 2012. Oasis evolution and water resource utilization of a typical area in the inland river basin of an arid area: a case study of the Manas River valley. Environ. Earth Sci. 66, 683–692. Zhang, K.C., An, Z.S., Cai, D.W., Guo, Z.C., Xiao, J.H., 2017. Key role of desert-oasis transitional area in avoiding oasis land degradation from Aeolian desertification in Dunhuang, northwest China. Land Degrad. Dev. 28, 142–150. Zhou, Z.H., Wang, C.K., Zheng, M.H., Jiang, L.F., Luo, Y.Q., 2017. Patterns and mechanisms of responses by soil microbial communities to nitrogen addition. Soil Biol. Biochem. 115, 433–441.

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.catena.2018.09.044. References Ananyeva, K., Wang, W., Smucker, A.J.M., Rivers, M.L., Kravchenko, A.N., 2013. Can intra-aggregate pore structures affect the aggregate's effectivess in protecting carbon? Soil Biol. Biochem. 57, 868–875. Barthes, B.G., Azontonde, A., Boil, B.Z., Prat, C., Roose, E., 2000. Field-scale run-off and erosion in relation to topsoil aggregate stability in three tropical regions (Benin, Cameroon, Mexico). Eur. J. Soil Sci. 51, 485–495. Barthes, B.G., Kouakoua, E., Larre-Larrouy, M.C., Razafimbelo, T.M., de Luca, E.F., Azontonde, A., Neves, C., de Freitas, P.L., Feller, C.L., 2008. Texture and sesquioxide effects on water-stable aggregates and organic matter in some tropical soils. Geoderma 143, 14–25. Bronick, C.J., Lal, R., 2005. Soil structure and management: a review. Geoderma 124, 3–22. Chen, P., Chu, Y., Gu, F.X., Zhang, Y.D., Pan, X.L., 2003. Spatial heterogeneity of vegetation and soil characteristics in oasisdesert ecotone. Chin. J. Appl. Ecol. 14, 904–908. Chen, S.H., Zhang, Y., Liu, J., Cui, S.S., Shen, J.P., 2008. Present situation, problems and countermeasure of applying fertilizer to cotton in Xinjiang. Xinjiang Agric. Sci. 45, 147–150. Cong, W.F., Hoffland, E., Li, L., Janssen, B.H., van der Werf, W., 2015. Intercropping affects the rate of decomposition of soil organic matter and root litter. Plant Soil 391, 399–411. Croft, H., Kuhn, N.J., Anderson, K., 2012. On the use of remote sensing techniques for monitoring spatio-temporal soil organic carbon dynamics in agricultural systems. Catena 94, 64–74. Elliott, E.T., 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native and cultivated soils. Soil Sci. Soc. Am. J. 50, 627–633. Elliott, E.T., Cambardella, C.A., 1991. Physical separation of soil organic matter. Agric. Ecosyst. Environ. 34, 407–419. Filho, C.C., Lourenço, A., Guimarães, M.D.F., Fonseca, I.C.B., 2002. Aggregate stability under different soil management systems in a red latosol in the state of Parana, Brazil. Soil Tillage Res. 65, 45–51. Fontaine, S., Barot, S., Barré, P., Bdioui, N., Mary, B., Rumpel, C., 2007. Stability of organic carbon in deep soil layers controlled by fresh carbon supply. Nature 450, 277–281. Fonte, S.J., Nesper, M., Hegglin, D., Velasquez, J.E., Ramirez, B., Rao, I.M., Bernasconi, S.M., Bünemann, E.K., Frossard, E., Oberson, A., 2014. Pasture degradation impacts soil phosphorus storage via changes to aggregate associated soil organic matter in highly weathered tropical soils. Soil Biol. Biochem. 68, 150–157. Greenfield, J.J., 2001. The origin and nature of organic nitrogen in soil as assessed by acidic and alkaline hydrolysis. Eur. J. Soil Biol. 52, 575–583. Han, D.L., 2001. Artificial Oases of Xinjiang. Chinese Environmental Sciences Press, Beijing. Huang, G., Li, Y., Su, Y.G., 2015. Divergent responses of soil microbial communities to water and nitrogen addition in a temperate desert. Geoderma 251-252, 55–64. Jiao, Y., Whalen, J.K., Hendershot, W.H., 2006. No-tillage and manure applications increase aggregation and improve nutrient retention in a sandy-loam soil. Geoderma 134, 24–33. Kong, A.Y.Y., Six, J., Bryant, D.C., Denison, R.F., van Kessel, C., 2005. The relationship between carbon input, aggregation, and soil organic carbon stabilization in sustainable cropping systems. Soil Sci. Soc. Am. J. 69, 1078–1085. Kravchenko, A.N., Wang, W., Worth, A., Smucker, A.J.M., Rivers, M.L., 2011. Fractal characterization of the intra–aggregate pore heterogeneity in macro-aggregates from contrasting land use and management treatments. Geophys. Res. Abstr. 13 (EGU2011–14155). Lai, B., Tang, M.Y., Chai, Z.P., Chen, B.L., Li, Q.J., Dong, J.H., Wang, F., Tian, C.Y., 2014. Investigation and evaluation of chemical fertilizer application situation of farmland in Xinjiang. Arid Zone Res. 31, 1024–1030. Lal, R., 2004. Soil carbon sequestration impacts on global climate change and food security. Science 304, 1623–1627. Lehmann, J., Kleber, M., 2015. The contentious nature of soil organic matter. Nature 528, 60–68.

780