Influence of cultivation and fertilization on total organic carbon and carbon fractions in soils from the Loess Plateau of China

Soil & Tillage Research 77 (2004) 59–68

Influence of cultivation and fertilization on total organic carbon and carbon fractions in soils from the Loess Plateau of China Tianyun Wu a,d , Jeff J. Schoenau b,∗ , Fengmin Li a,∗ , Peiyuan Qian b , Sukhdev S. Malhi c , Yuanchun Shi a , Fuli Xu e a State Key Laboratory of Arid Agroecology, Lanzhou University, Lanzhou 730000, PR China Department of Soil Science, University of Saskatchewan, 51 Campus Drive, Saskatoon, Sask., Canada S7N 5A8 c Nutrient Cycling Research Station, Agriculture and Agri-Food Canada, Melfort, Sask., Canada S0E 1A0 d Soil Science and Fertilizer Institute, Gansu Academy of Agricultural Sciences, Lanzhou 730070, PR China College of Resources and Environment, North-Western University of Agriculture and Forestry Technology, Yanling 712100, PR China b

e

Received 27 December 2001; received in revised form 10 October 2003; accepted 27 October 2003

Abstract To evaluate the degradation of soil quality and find ways to maintain soil fertility on the Loess Plateau of China, the effects of cultivation time on total organic carbon (TOC), light fraction of organic carbon (LFOC), and microbial biomass carbon (MB-C) in two soil chronosequences comprised of Huangmian (Calcaric Cambisols, FAO) and Huihe (Haplic Greyxems, FAO) soils were investigated. The effects of fertilization on the TOC and its fractions were also studied using samples from a long-term experiment on Heilu soil (Calcic Kastanozems, FAO). Upon cultivation, Huangmian soil (0–20 cm) lost 77% of TOC within 5 years, at a reduction rate of 2.15 Mg C ha−1 per year. The Huihe soil (0–20 cm) lost 70% of TOC at a rate of 0.96–1.06 Mg C ha−1 per year over 42 years. In the Huangmian soil, water and tillage erosion are likely the main reasons for organic carbon decline, while organic matter decomposition and water erosion appear to be dominant factors in the Huihe soil. The LFOC decreased by 73 and 90% for the Huangmian and Huihe soil for the corresponding period. Changes in microbial biomass carbon (MB-C) showed the same trend as TOC and LFOC. The results of the long-term experiment on the Heilu soil indicated that manure alone and manure plus nitrogen and phosphorus fertilizer treatments restored TOC and MB-C to the level of the native sod, indicating the importance of manure addition in maintaining soil fertility over the long term (20 years). The straw return plus nitrogen and phosphorus fertilizer treatment had a significantly higher TOC than nitrogen plus phosphorus fertilizer alone. Organic matter additions in the form of manure or straw, either alone or in combination with chemical fertilizers, appears to be more effective in maintaining or restoring organic matter in Heilu soil on the Loess Plateau than chemical fertilizer alone. © 2003 Elsevier B.V. All rights reserved. Keywords: Soil organic carbon; Light fraction of organic carbon; Microbial biomass carbon; Cultivation; Fertilization; Soil erosion; Loess Plateau of China

1. Introduction ∗ Corresponding authors. Tel.: +1-306-966-6844 (Schoenau)/+86-931-8912848 (Li); fax: +1-306-966-6881 (Schoenau)/+86-931-8912846 (Li). E-mail addresses: [email protected] (J.J. Schoenau), [email protected] (F. Li).

Soil organic matter (SOM) contributes to nutrient supply, improvement of soil physical properties, and protection from erosion (Stevenson, 1994) and thus a positive correlation between its content and soil

0167-1987/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.still.2003.10.002

60

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

quality is assumed (Janzen et al., 1997). The SOM, however, is easily lost upon cultivation by accelerated decomposition to CO2 as well as erosion. The content of SOM is determined by different environmental factors and managements. Among them, climate and topography are dominant on a large scale (Potter et al., 1998; Stevenson, 1994). Tillage and rotation are the major management factors causing changes in SOM content (Campbell et al., 1996). For example, the content of SOM may decrease sharply after undisturbed land is cultivated (Carter et al., 1998), and the longer the cultivation period, generally the lower the content of SOM (Dalal and Mayer, 1986, 1987). Fertilization can constrain the decrease of SOM to some extent (Odell et al., 1984) as can the reduction or elimination of erosion (Gregorich et al., 1998). The Loess Plateau of China, where agriculture began in the paleolithic period along the valley of the Yellow River and its tributaries, has a land area of 56,000 km2 . The main soils of the Loess Plateau developed from parent material of calcareous loess with a feature of low SOM content, on average around 10 g C kg−1 (Soil Survey Office in Gansu Province of China, 1996). Undoubtedly, the dry climate and sparse vegetation are mainly responsible for the low SOM (Janzen et al., 1997, 1998). However, a long period of cultivation and severe erosion on the Loess Plateau are likely other potential causes of low SOM. Soil erosion in this region has been well documented (Scientific Survey Office of Loess Plateau in CAS, 1991; Lal, 2002) and evidence of severe erosion problems in Loess Plateau soils has aroused criticism of the existing agricultural system and its sustainability, even though total agricultural output has been doubled or even tripled due to new crop varieties adopted and increased chemical fertilizer input in the past decades (Gansu Statistic Bureau, 2002). How-

ever, studies providing direct evidence of soil fertility and quality decline associated with cultivation in this region are few and the influence of management practices in restoring soil quality has received little attention. The first objective of this study was to assess the changes in soil organic carbon and its fractions as affected by cultivation of the dominant agricultural soils of the Loess Plateau. A second objective was to determine how management practices, especially organic matter amendment and fertilization, could alter the amounts and distribution of the soil carbon fractions, and thereby assist in maintaining or restoring soil productivity. 2. Material and methods 2.1. Sites and soils The soils used in the study are Huangmian soil (Calcaric Cambisols, FAO), Huihe soil (Haplic Greyxems, FAO), and Heilu soil (Calcic Kastanozems, FAO), which occupy areas of 160,080, 28,300 and 21,451 km2 , respectively, within the region. The basic properties of the soils are presented in Table 1. 2.1.1. Huangmian soil A chronosequence of Huangmian soil, which had been cultivated for 0, 5, 40 and 100 years up to the time of sampling, was sampled on sloping fields (<10% slope) on 14 May 2000. For each field, three composite samples were prepared by combining the cores from three squares. In each square of 70 m2 , 20 soil cores of 0–20 cm depth were taken using a grid sampling method and then mixed into one composite sample of about 1 kg (Crépin and Johnson, 1993). Before

Table 1 The properties of the uncultivated soilsa Soils

Huangmian Huihe Heilu

pH

8.3 8.3 8.4

CaCO3 (g kg−1 )

Total N (g kg−1 )

Bulk density (Mg m−3 )

130.8 99.4 113.8

0.68 1.68 0.76

0.98 1.08 1.20

Particle size distribution (g kg−1 ) Sand (>50 ␮m)

Silt (2–50 ␮m)

Clay (<2 ␮m)

72 98 39

656 634 651

257 254 298

a pH was determined in a soil/water suspension of 1:10; CaCO was analyzed by measuring volume of CO that was released from 3 2 the reaction of soil and HCl; total N was analyzed with Kjeldahl method.

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

sampling, five pits to a 2 m depth were excavated in each field of different cultivation years to examine the soil profile and ensure the same soil type was present. Similar topography, slope position, and management were used as criterion for selection, using a detailed soil map of the area. The native sod was located about ten meters from the soil cultivated for 5 years. The chronosequence is located near Tangjiabu in Dingxi County, Gansu Province, China, and the average annual precipitation at Tangjiabu is 410 mm, and average temperature is 6.4 ◦ C. As a result, only short and sparse vegetation covers hills and ravines in this region. Generally, 25–40 kg N ha−1 as urea or ammonium nitrate, 6.5–10.8 kg P ha−1 as calcium superphosphate and 0.75–1.50 Mg ha−1 of animal manure are applied annually in the sloping fields. The approximate nutrient content of applied manure was as follows: organic matter 25–45 g C kg−1 , total N 2.0–4.0 g N kg−1 , total P 1.8–5.2 g P kg−1 , available N 260–350 mg N kg−1 , soluble P 19.4–26.7 mg P kg−1 , soluble K 290–374 mg K kg−1 . About 0.75–1.00 Mg ha−1 of cereal grain or peas is harvested each year and straw is taken away with grain. Pea (Pisum L.)–spring wheat (Triticum aestivum L.)–flax (Linum L.) is the main rotation system in the area. Usually, farmers till the soil to a depth of about 20 cm with a plough in spring and autumn and hand cultivate to a depth of about 5 cm with a hoe for weeding 3–5 times during the growing season. 2.1.2. Huihe soil The Huihe soil, cultivated for 0, 4, 10, 20 and 42 years, was sampled from gently sloping fields (<5%) that was covered by forest previously in Beishui, Heshui County of Gansu Province. The soils were sampled on 25 May 2000. The same site selection and sampling method were employed as described for the Huangmian soil. The annual rainfall at this site is 590 mm, and the average temperature is 7.4 ◦ C. About 60% of rainfall comes in July–September. The Huihe soil was developed under forest. However, the original forest was destroyed in Beishui historically and the land was reforested centuries ago. In the last 100 years, farmers cleared the bush and grow 0.75–1.0 Mg ha−1 panic millet (Panicum italicum L.) or beans (Phaseolus L.), and annually apply 35–70 kg N ha−1 as urea or ammonium nitrate and 9.7–16.1 kg P ha−1 as superphos-

61

phate. No manure is applied to these soils. The same tillage and rotation system as well as field and residue management are practiced as on the Huangmian soil. 2.1.3. Heilu soil The Heilu soil sampled is part of a 20-year long-term (1979–1999) randomized complete block experiment. The soil is near Gaoping in Jinchuan County of Gansu Province and was sampled on 15 August 2000. The same sampling method was used as for the Huangmian soil. The 20-year treatments sampled were as follows: (1) Control, no fertilizer; (2) N, nitrogen fertilizer as urea or ammonium nitrate at the rate of 90 kg N ha−1 ; (3) M, manure at the rate of 75 Mg ha−1 ; (4) N + P, nitrogen fertilizer at the rate of 90 kg N ha−1 and phosphorus fertilizer as superphosphate at the rate of 32.3 kg P ha−1 ; (5) N + P + M, manure at the rate of 75 Mg ha−1 and nitrogen fertilizer at the rate of 90 kg N ha−1 plus phosphorus fertilizer at the rate of 32.3 kg P ha−1 ; (6) straw + N + P, returning wheat straw at the rate of 3.8 Mg ha−1 and nitrogen fertilizer at the rate of 90 kg N ha−1 plus phosphorus fertilizer at the rate of 32.3 kg P ha−1 . In treatments (5) and (6), 32.3 kg P ha−1 was applied every 2 years while in the other treatments, applications of fertilizers and manure were made every year. Straw was returned to treatment (6) but removed along with grain from other treatments. The treatments and yields are described in detail by Zhou and Ding (1999). An area of native sod in the same region was located and the soil was sampled. The native site was near a gulley and separated by steep cliffs from cropland. No charcoal was found in the profiles, establishing it as a natural sod. The nutrient content of applied manure was as follows: organic matter 20–25 g C kg−1 , total N 1.5–2.0 g N kg−1 , total P 0.8–2.5 g P kg−1 , available N 180–250 mg N kg−1 , soluble P 12.9–17.2 mg P kg−1 , soluble K 232–291 mg K kg−1 . A winter wheat (Triticum aestivum L.)–soybean (Glycine max L. Merrill)–maize (Zea mays L.) rotation system is practiced. The Heilu soil is located on tableland of the Loess Plateau, and the nature of the level land makes cultivation easy. Therefore, the site has an agricultural history longer than 2000 years. The average annual precipitation is 560 mm, and average temperature is 10.2 ◦ C in Jinchuan County.

62

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

2.2. Analytical methods 2.2.1. Total soil organic carbon (SOC) A 0.15–0.20 g soil sample, which was from a 3–5 g sub-sample ground with a ball-mill to pass a 100 mesh sieve, was combusted using a Leco CR-12 Carbon Analyzer set at 840 ◦ C, and the content of soil organic carbon was assessed directly by measurement of the CO2 using infrared detection (Wang and Anderson, 1998). The carbon mass was derived by multiplying content of SOC by depth and by soil bulk density. This method was also used to calculate carbon mass of light fraction. 2.2.2. Light fraction of organic carbon (LFOC) A sub-sample of 25 g soil, from a 0.5 kg sample ground to pass a 2 mm sieve, was weighed into a 250 ml centrifuge tube, then 50 ml of NaI with density of 1.70 g ml−1 was poured into the tube. The tube was stoppered and shaken at 200 rpm for 1 h. The wall of centrifuge tube and the stopper was then washed with 3–5 ml NaI and then was centrifuged for 20 min at a relative centrifugal force of 1000×g. The light fraction (LF), in suspension after centrifugation, was decanted into a vacuum filter unit with 0.45 ␮m nylon filter paper. NaI was collected for reuse and the LF on the paper was washed with 75 ml of 0.01 M CaCl2 , followed by at least 75 ml of distilled water. The LF was transferred with water into a vial and the excess water was evaporated for 24 h. The LF in the vial was dried at 50 ◦ C for 72 h and the weight of LF was obtained. The residual material in the centrifuge tube was extracted with NaI one more time and two aliquots of the LF were combined together for a sample (Gregorich and Ellert, 1993). The combined LF was ground to pass a 60-mesh sieve and combusted on the Leco CR-12 to determine the concentration of organic carbon in LF. 2.2.3. Microbial biomass carbon (MB-C) The MB-C was analyzed using the method described by Voroney et al. (1993). Two aliquots of 20 g of fresh soil were weighed into two 200 ml flasks and 40 ml of 0.5 M K2 SO4 was poured into each flask. The 1 ml of chloroform, which was purified two times, was added into one of the two mixtures. The flasks were stoppered and shaken at 200 rpm for 1 h. The filtrate was collected and bubbled with CO2 free air for 30 s after filtration. Then, 8 ml of the filtrate

was transferred to a 150 ml flask, and 0.075 g of HgO, 2 ml of 0.2 M K2 CrO7 , 10 ml of concentrated H2 SO4 and 5 ml of concentrated H3 PO4 was also added into the flask. The mixture was digested at 250 ◦ C for 30 min, and then was transferred into a 500 ml flask, titrated with 0.017 M FeSO4 using ferroin as an indicator. The test was replicated 3 times for each soil sample. MB-C was calculated by the difference in carbon content in fumigated and unfumigated soil: MB-C (mg kg−1 ) = (OCF − OCUF )/0.18 (Voroney et al., 1993). 2.2.4. 137 Cs About 120–150 g soil sample, ground to pass a 2 mm sieve, was weighed into a 150 ml Marinellis beaker. Then it was placed in a gamma ray spectrometer, described in detail by de Jong et al. (1982), to analyze 137 Cs in the soil. Soil loss by erosion was calculated by the method of de Jong et al. (1983): Cs loss (%) 0.95 137 Csuncultivated site − 137 Cscultivated site = 100 × 0.95137 Csuncultivated site (1)

137

Soil loss (Mg ha−1 ) = 137 Cs loss (%) × depth (cm) ×bulk density (Mg m−3 )

(2)

2.3. Statistical methods Statistical analysis was completed with SAS and an ANOVA was used to conduct the analysis of variance. Mean values are reported.

3. Results and discussion 3.1. Changes in total soil organic carbon content of soils with cultivation Total soil organic carbon is reported to decrease upon cultivation, with a sharp decline at the beginning, followed by a slower decrease and eventually leveling off at a new lower equilibrium level (Janzen et al., 1997; Dalal and Mayer, 1986; Campbell, 1978). The Huihe soil tended to follow the typical pattern for SOC decline (Table 2) reported in other studies of

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68 Table 2 TOC, light fraction (LF), MB-C and Years of cultivation

Total soil organic C (Mg C ha−1 )

Huangmian soil 0 year 13.9 5 years 3.2 40 years 5.3 100 years 5.0 LSD0.05 Pr > F Huihe soil 0 year 4 years 10 years 20 years 42 years LSD0.05 Pr > F

2.3 0.001

137 Cs

63

in soils (0–20 cm) of chronosequences

Light fraction Dry matter (g kg−1 soil)

C content (g C kg−1 LF)

11.8 7.6 5.8 6.6

112.0 38.9 77.1 87.2

2.3 0.001

56.8 46.1 27.8 22.6 16.7

37.6 30.5 13.2 8.8 6.8

14.1 0.004

12.4 0.0007

19.3 0.0002 175.0 141.1 124.0 133.6 95.8 57.7 0.1134

LFOC (Mg C ha−1 ) 2.6 0.7 1.0 1.4

(18.6a ) (22.3) (18.3) (27.2)

0.3 0.0001 14.7 9.9 3.5 2.7 1.4

(26.0) (21.1) (12.4) (11.8) (8.5)

7.7 0.014

MB-C (mg C kg−1 )

654 241 329 243

(9.2b ) (18.3) (13.8) (10.5)

137 Cs

changes and soil loss

137 Cs

and loss (Bq m−2 ) 729.7 67.2 (90.3c ) 148.2 (78.7) 193.2 (72.2)

Soil loss (Mg ha−1 )

2190 1750 1570

149 0.011 776 711 638 342 229

(3.0) (3.5) (4.9) (3.5) (2.9)

1948.8 1764.5 1046.7 911.5 961.4

(4.7) (43.5) (50.8) (48.1)

110 930 1160 1040

178 0.021

a

The percentage of TOC comprised of LFOC. The percentage of TOC comprised of MB-C. c The percentage of 137 Cs lost. b

SOC losses with cultivation (Dalal and Mayer, 1986, 1987). The Huangmian soil has a lower total organic carbon (TOC) content under natural vegetation than Huihe soil, and SOC mass was decreased by 77% in this soil after 5 years of cultivation, with a decline rate of 2.15 Mg C ha−1 per year (Table 2). Severe water erosion following cultivation on this steeply sloping land and the decomposition of soil organic matter are likely the main causes of the sharp decline after only 5 years. Wang et al. (2001) reported a large decrease in SOC concentration caused by erosion on soils of the Loess Plateau of China and Li et al. (2001) found that SOC had a strong relationship with soil redistribution caused by water erosion on a sloping landscape of Calcaric Cambisols within same region of this study. In addition, tillage erosion and mixing of top and subsoil from ploughing may also be a factor as farmers in this area practice contour ploughing on slopes. As a result, a 2–3 m high steep cliff at the upper side and a 2–3 m wide terrace down the cliff is formed, and soil is always moved from the upper point to the lower by the plough even though contour tillage is practiced. It was reported this area is also one of the districts that has the most severe soil erosion by water, with an es-

timated 6050 Mg km−2 per year of surface soil loss by water erosion (Scientific Survey Office of Loess Plateau in CAS, 1991). The results of 137 Cs analysis indicated that soil erosion of the Huangmian soil was quite severe: 90% of top soil (0–20 cm) was lost within 5 years after native sod was broken, equivalent to 2190 Mg ha−1 soil loss (Table 2), presumably due to water and tillage erosion. After 40 years of cultivation, SOC increased slightly to 5.30 Mg C ha−1 and was nearly constant from 40 years to 100 years of cultivation. A possible reason for this trend is the effect of repeated application of manure in later years of cultivation. The amount of plant debris (>2 mm) in the 0–20 cm layer remained constant in the cultivated soils, indicating that recent plant residue inputs were similar in the soils of different cultivation years (Fig. 1). The percentage of soil loss by erosion was higher than that of carbon mass loss (Table 2), implying that manure application and cropping can compensate for some of the SOC loss due to erosion and decomposition, with the assumption that erosion evenly carried away soil particles with different organic carbon content and density. In fact, soil particles with higher organic carbon content

64

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68 0.6

6

0.5

5

0.4

4

0.3

3

0.2

2

0.1

1

0

Debris in Huihe (g kg-1 )

Debris in Huangmian (g kg-1 )

Huangmian Huihe

0 0

10

20

30

40

50

60

70

80

90

100

Cultivation year (yr)

Fig. 1. Plant debris (>2 mm) in 0–20 cm layers of the Huangmian and Huihe chronosequence.

have been reported to be eroded prior to those of lower carbon (Woods and Schuman, 1988; Avnimelech and McHenry, 1984). The Huihe soil was formed under forest, with the highest organic carbon mass (56.78 Mg C ha−1 ) in top soil (0–20 cm) under native vegetation found on the Loess Plateau (Soil Survey Office in Gansu Province of China, 1996). There was a 70% reduction in SOC mass (0–20 cm) after 42 years of cultivation with a decline rate of 1.06 Mg C ha−1 per year for the first 4 years and 0.96 Mg C ha−1 per year for 0–42 years (Table 2), which is comparable to the loss rates reported for a Riverview soil in Australia (Dalal and Mayer, 1986) but only half of that in Huangmian soil. According to the 137 Cs data, soil loss was 4.7% of top soil mass (0–20 cm) while the loss of TOC mass in top soil (0–20 cm) was 17.4% after 4 years of cultivation. However, soil losses were 43.5–50.8% and TOC losses were 51.0–70.1% in the fields of 10–42 years

cultivation, respectively, compared to the native sod. Therefore, decomposition of SOM in the early cultivation period is postulated as the main reason for the TOC depletion and the influence of soil erosion on TOC decline increased with cultivation time, similar to what has been reported for some Canadian soils (Gregorich et al., 1998). 3.2. Light fraction of soil organic carbon The light fraction is mainly comprised of recent debris of plant, soil animals and microorganisms in various stages of decomposition. The light fraction serves as a readily decomposable substrate for soil microorganisms and as a short-term reservoir of plant nutrients. As a result, light fraction is sensitive to the recent inputs of plant residue and thus can be a better indication of effects of soil management and cropping systems than total organic matter in soils (Gregorich

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

et al., 1994; Janzen et al., 1992; Christensen, 1992). The light fraction of carbon (LFOC) in the Huangmian soil accounted for 18.3–27.2% of TOC in soil (Table 2). The dynamics of LFOC were very similar to that of TOC in this soil: a sharp decline (73%) in the first 5 years followed by a small increase in successive intervals. Soil erosion and vegetation change are factors likely responsible for the sharp decline initially, followed by the effects of manure application for the subsequent stabilization of LFOC amounts (Table 2). A significant correlation between LFOC and plant debris (>2 mm) (LFOC = 0.893 × Debris + 2.801, R2 = 0.777∗∗ ) further supports the concept that vegetation change and plant residue input were factors associated with the fluctuation of LFOC in the Huangmian soil. The Huihe soil had more light fraction than the Huangmian soil under native vegetation (Table 2). The reason is likely related to different vegetation and climate (Janzen et al., 1997, 1998): the Huihe soil developed under forest in a sub-humid area, while the Huangmian soil was formed under sparse grass in a semi-arid region. This is also reflected in the different amounts of plant debris in the top soil of the two native soils (Fig. 1), with Huihe having much larger amounts of plant debris than Huangmian. A larger LFOC decline than the total SOC was observed in the Huihe soil: the LFOC was reduced by 33% after 4 years of cultivation and 90% after 42 years, demonstrating that the LFOC is more sensitive than total SOC to vegetation change, cultivation and soil erosion. More plant debris in the soil of 4-year cultivation than the native might be a result of the sampling process in which dead large roots became part of the debris fraction after cultivation and were included in the sample but they were deliberately excluded when sampling in the native Huihe soil. A positive correlation between LFOC and plant debris was also found in this soil: LFOC = 1.771 × Debris + 2.470, R2 = 0.508∗∗ . 3.3. Microbial biomass carbon MB-C is an important attribute of soil organic matter quality as it provides an indication of a soil’s ability to store and recycle nutrients and energy. As a measure of organic matter quality, it also serves as a sensitive indicator of change and future trends in organic matter level (Gregorich et al., 1994). MB-C is

65

Table 3 Correlation between MB-C and SOC fractions in the soil chronosequences SOC fraction

MB-C = A + B (SOC fraction) A

B

R2

Pr > F

Huangmian

TOC LFOC

117.87 66.33

75.88 212.45

0.9227 0.7352

0.0001 0.0004

Huihe

TOC LFOC

139.21 363.13

25.86 27.38

0.7138 0.5392

0.0001 0.0018

Soil

determined by the quantity and quality of C input into a soil. Thus, it is influenced by agriculture practices such as tillage, cropping sequences and manuring, as well as by climatic conditions (Insam et al., 1989). Overall, MB-C of the two soils decreased with increase in cultivation time (Table 2). The proportion of total soil organic carbon comprised of MB-C for the Huangmian soil ranged from 9.2 to 18.3%, which was much higher than that reported by others (Insam et al., 1989; Anderson and Domsch, 1989) and the reason is unknown. The MB-C proportion for the Huihe was 2.9–4.9%. In both soils, the MB-C was strongly related to TOC and LFOC (Table 3). 3.4. Influence of fertilization on soil organic matter fractions It has been shown that application of chemical fertilizer and manure can increase SOM (Biederbeck et al., 1994; Campbell et al., 1991). Chemical fertilizer can increase shoot and root production of a crop, eventually increasing residue input into soil, while manure contains material that is a precursor to SOM and also provides available nutrients. This effect has been realized and used for nearly 4000 years in China, Japan, and Korea (Dormaar et al., 1988) to restore soil fertility and get a satisfactory yield. The results of the long-term experiment on Heilu soil of the Loess Plateau (Table 4) demonstrated the positive effect of manure amendment on SOM. The addition of 75 Mg manure ha−1 per year for 20 years produced the highest TOC and LFOC in 0–20 cm layer among the treatments, with TOC as high as the native sod. The TOC and LFOC of the N and N + P treatments were the lowest, even lower than the treatment of no

66

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

Table 4 The influence of fertilization on TOC, light fraction (LF), and MB-C in the Heilu soil (0–20 cm) Treatment

Native sod Control N M N+P M+N+P Straw + N + P LSD0.05 Pr > F a b

MB-C (mg C kg−1 )

Total soil organic C (Mg C ha−1 )

Light fraction Dry matter (g kg−1 soil)

C content (g C kg−1 LF)

LFOC (Mg C ha−1 )

25.9 19.5 15.9 26.0 15.9 24.0 18.8

9.8 5.5 4.9 8.2 6.0 10.3 6.8

185.7 142.4 150.7 186.5 117.2 159.8 129.5

5.2 2.1 2.0 3.8 1.6 3.9 2.2

2.9 0.001

1.9 0.0001

28.1 0.0006

(20.0a ) (10.9) (12.6) (14.7) (10.3) (16.2) (11.5)

0.8 0.0001

609 428 503 724 664 903 671

(5.8b ) (5.9) (8.6) (6.6) (10.0) (9.0) (8.9)

160 0.01

The percentage of TOC comprised of LFOC. The percentage of TOC comprised of MB-C.

fertilizer application (Control). Applying straw with nitrogen and phosphorus (straw + N + P) fertilizer resulted in higher SOC and LFOC than N + P alone, but SOC and LF for the treatments were still similar to that of the Control. The chemical fertilizer alone may have had the effect of enhancing the decomposition of SOM in this soil. This observation contradicts others observed on the Loess Plateau and the Canadian Prairie (Bremer et al., 1994; Campbell et al., 1991; Odell et al., 1984; Wang and Zhang, 1998). The addition of fertilizer may have stimulated microbial activity and enhanced decomposition. Support for this possibility is that the MB-C in soils that received chemical fertilizer was 17–110% higher than in the Control (Table 4). Higher MB-C suggests more intensive activity of microorganisms and thus more decomposition of soil organic matter (Anderson and Domsch, 1990). Insam et al. (1989) also found that MB-C was higher in the plots that received chemical fertilizer than the non-fertilized. The different manuring regimes had effects on soil biological quality reflected by the MB-C amount and the ratio of MB-C/TOC. Chemical nitrogen alone had no effect on MB-C whereas chemical nitrogen plus phosphorus significantly increased the MB-C compared to the Control. However, manure plus chemical fertilizer had larger effects than chemical fertilizer only. Returning straw plus fertilizer nitrogen and phosphorus produced a MB-C amount similar to the fertilizer nitrogen plus phosphorus. Manuring increased the ratio of MB-C/TOC compared to the native sod

indicating an improvement in soil biological quality (Gregorich et al., 1997). The long-term experiment on the Heilu soil is located in a level area of the Loess Plateau in which agriculture has been conducted for more than 2000 years. At this site, the TOC of the Control was 25% lower than the native sod and wind erosion rather than water erosion is likely the main erosional process responsible for the depletion. Manure application and return of straw residue appears effective in helping maintain the organic carbon of cultivated land in this soil. Applying straw and fertilizer together had a positive effect on organic carbon over fertilizer alone. Within the tableland area of the Loess Plateau, now about 1/3 of the winter wheat is harvested with combines because the cost to hire labor is becoming higher than to rent a machine. As a result, return of residue cover is expected to be more prevalent in the region which should help maintain or improve soil productivity on the Loess Plateau (Wu et al., 2003).

4. Conclusions The TOC, LFOC and MB-C of newly cultivated land on the Loess Plateau decreases sharply over a short period (5–42 years) on steeply sloping farmland (Huangmian and Huihe soil) where soil erosion is severe. However, TOC only decreased by about 25% on a flat area (Heilu soil) of the region, where water erosion is a less severe problem. The deterioration of soil

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

quality reflected in the large decrease in TOC and soil carbon fractions in the two main agricultural soils indicates the need for adoption of soil conservation practices such as reduced tillage in sloping farmland on the Loess Plateau to reduce soil erosion and organic matter loss. The results of this study indicate that reducing erosion along with the application of manure and crop residues, either alone or in combination with chemical fertilizer, should have a positive benefit in maintaining and restoring soil organic matter quantity and quality in this region of China.

Acknowledgements This study is partly supported by NKBRSF Project G2000018603, International Foundation for Sciences (IFS, C/3313-2), and China Scholarship Council (990005). The efforts of Zhou Guangye and Ding Ningping, who work in Pinliang Agriculture Institute, are appreciated for supplying soil samples from the long-term experiment on Heilu soil. Dr. D.W. Anderson is appreciated for his advice on the selection of analytical methods. Our thanks also to Professor E. de Jong for his help in the analysis of 137 Cs.

References Anderson, T.H., Domsch, K.H., 1989. Ratios of microbial biomass carbon to total organic carbon in arable soils. Soil Biol. Biochem. 21, 471–479. Anderson, T.H., Domsch, K.H., 1990. Application of ecophysiological quotients (qCO2 and qD) on microbial biomasses from soils of different cropping histories. Soil Biol. Biochem. 22, 251–255. Avnimelech, Y., McHenry, J.R., 1984. Enrichment of transported sediments with organic carbon, nutrients and clay. Soil Sci. Soc. Am. J. 48, 259–266. Biederbeck, V.O., Janzen, H.H., Campbell, C.A., Zentner, R.P., 1994. Labile soil organic matter as influenced by cropping practices in an arid environment. Soil Biol. Biochem. 26, 1647– 1656. Bremer, E., Janzen, H.H., Johnston, A.M., 1994. Sensitivity of total, light fraction and mineralizable organic matter to management practices in a Lethbridge soil. Can. J. Soil Sci. 74, 131–138. Campbell, C.A., 1978. Soil organic carbon, nitrogen and fertility. In: Schnitzer, M., Khan, S.U. (Eds.), Soil Organic Matter, Developments in Soil Science, vol. 8. Elsevier, Amsterdam, pp. 173–271.

67

Campbell, C.A., Biederbeck, V.O., Zentner, R.P., LaFond, G.P., 1991. Effect of crop rotations and cultural practices on soil organic matter, microbial biomass and respiration in a thin Black Chernozem. Can. J. Soil Sci. 71, 363–376. Campbell, C.A., McConkey, B.G., Zeatner, R.P., Selles, F., Curtin, D., 1996. Long-term effects of tillage and crop rotations on soil organic C and total N in a clay soil in southwestern Saskatchewan. Can. J. Soil Sci. 76, 395–401. Carter, M.R., Gregorich, E.G., Angers, D.A., Donald, R.G., Bolinder, M.A., 1998. Organic C and N shortage and organic C fractions in adjacent cultivated and forested soils of eastern Canada. Soil Till. Res. 47, 253–261. Christensen, B.T., 1992. Physical fractionation of soil and organic matter in primary particle size and density separates. Adv. Soil Sci. 20, 2–90. Crépin, J., Johnson, R.L., 1993. Soil sampling for environmental assessment. In: Carter, M.R. (Ed.), Soil Sampling and Methods of Analysis. Canadian Society of Soil Science. Lewis Publishers, Division of CRC Press, Boca Raton, FL, pp. 5–24. Dalal, R.C., Mayer, R.J., 1986. Long-term trends in fertility of soil under continuous cultivation and cereal cropping in southern Queensland. IV. Loss of organic carbon from different density fractions. Aust. J. Soil Res. 24, 301–309. Dalal, R.C., Mayer, R.J., 1987. Long-term trends in fertility of soil under continuous cultivation and cereal cropping in southern Queensland. VI. Loss of total nitrogen from different particle size and density fractions. Aust. J. Soil Res. 25, 83–93. de Jong, E., Begg, C.B.M., Kachanoski, R.G., 1983. Estimates of soil erosion and deposition for some Saskatchewan soils. Can. J. Soil Sci. 63, 607–617. de Jong, E., Villar, H., Bettany, J., 1982. Preliminary investigations on the loss of 137 Cs to estimate erosion in Saskatchewan. Can. J. Soil Sci. 62, 673–683. Dormaar, J.F., Lindwall, C.W., Kozub, G.C., 1988. Effectiveness of manure and commercial fertilizer in restoring productivity of an artificially eroded Dark Brown Chernozemic soil under dryland conditions. Can. J. Soil Sci. 68, 669–679. Gansu Statistic Bureau, 2002. Gansu Rural Yearbook. China Statistics Press, Beijing, pp. 370–385. Gregorich, E.G., Ellert, B.H., 1993. Light fraction and macroorganic matter in mineral soils. In: Carter, M.R. (Ed.), Soil Sampling and Methods of Analysis. Canadian Society of Soil Science, Lewis Publishers, Division of CRC Press, Boca Raton, FL, pp. 397–405. Gregorich, E.G., Ellert, B.H., Gregorich, E.G., Carter, M.R., Monreal, C.M., Ellert, B.H., 1994. Towards a minimum data set to assess soil organic matter quality in agricultural soils. Can. J. Soil Sci. 74, 367–385. Gregorich, E.G., Carter, M.R., Doran, J.W., Pankhurst, C.E., Dwyer, L.M., 1997. Biological attributes of soil quality. In: Gregorich, E.G., Carter, M.R. (Eds.), Soil Quality for Crop Production and Ecosystem Health. Development in Soil Science, vol. 25. Elsevier, Amsterdam, pp. 81–113. Gregorich, E.G., Greer, K.J., Anderson, D.W., Liang, B.C., 1998. Carbon distribution and losses: erosion and deposition effects. Soil Till. Res. 47, 291–320.

68

T. Wu et al. / Soil & Tillage Research 77 (2004) 59–68

Insam, H., Parkinson, D., Domsch, K.H., 1989. Influence of macroclimate on soil microbial biomass. Soil Biol. Biochem. 21, 211–221. Janzen, H.H., Campbell, C.A., Brandt, S.A., Lafond, G.P., Townley-Smith, L., 1992. Light fraction organic matter in soils from long-term crop rotations. Soil Sci. Soc. Am. J. 56, 1799– 1806. Janzen, H.H., Campbell, C.A., Ellert, B.H., Bremer, E., 1997. Soil organic matter dynamics and their relationship to soil quality. In: Gregorich, E.G., Carter, M.R. (Eds.), Soil Quality for Crop Production and Ecosystem Health. Developments in Soil Science, vol. 25. Elsevier, Amsterdam, pp. 277–292. Janzen, H.H., Campbell, C.A., Izaurralde, R.C., Ellert, B.H., Juma, N., McGill, W.B., Zentner, R.P., 1998. Management effects on soil C storage on the Canadian prairies. Soil Till. Res. 47, 181–195. Lal, R., 2002. Soil carbon sequestration in China through agricultural intensification, and restoration of degraded and desertified ecosystems. Land Degrad. Develop. 13, 469–478. Li, Y., Lindstrom, M.J., Frielinghous, M., Bork, H.R., 2001. Quantifying the spatial patterns of soil redistribution and soil quality on two contrasting hillslopes. In: Stott, D.E., Mohtar, R.H., Steinhardt, G.C. (Eds.), Sustaining the Global Farm. Selected Papers from the 10th International Soil Conservation Organization Meeting, May 24–29, 1999. Purdue University and the USDA-ARS National Soil Erosion Research Laboratory, pp. 556–563. Odell, R.T., Melsted, S.W., Walker, W.M., 1984. Changes in organic carbon and nitrogen of Morrow plot soils under different treatments, 1904–1973. Soil Sci. 137, 160–171. Potter, K.N., Torbert, H.A., Jones, O.R., Jones, O.R., Matocha, J.E., Morrison, J.E., Unger Jr., P.W., 1998. Distribution and amount of soil organic C in long-term management systems in Texas. Soil Till. Res. 47, 309–321. Scientific Survey Office of Loess Plateau in CAS, 1991. Regional characteristics of erosion on Loess Plateau and reclamation. Science and Technology Press in China, Beijing, pp. 171–177 (in Chinese).

Soil Survey Office in Gansu Province of China, 1996. Soils in Gansu Province. Science and Technology Press in Gansu, Lanzhou, pp. 20–40 (in Chinese). Stevenson, F.J., 1994. Humus Chemistry: Genesis, Composition, Reaction. Wiley, New York, pp. 1–24. Voroney, R.P., Winter, J.P., Beyaert, R.P., 1993. Soil microbial biomass C and N. In: Carter, M.R. (Ed.), Soil Sampling and Method of Analysis. Canadian Society of Soil Science, Lewis Publishers, Division of CRC Press, Boca Raton, FL, pp. 277–286. Wang, D., Anderson, D.W., 1998. Direct measurement of organic carbon content in soils by the Leco CR-12 carbon analyzer. Commun. Soil Sci. Plant Anal. 29, 15–21. Wang, J., Fu, B.J., Qiu, Y., Chen, L.D., 2001. Soil nutrients in relation to land use and landscape position in the semi-arid small catchments on the Loess Plateau in China. J. Arid Environ. 48, 537–550. Wang, X.-D., Zhang, Y.-P., 1998. Variations and properties of soil humic acid fractions under different fertilization conditions. Acta Pedol. Sin. 35, 404–411 (in Chinese and with English abstract). Woods, L.E., Schuman, G.E., 1988. Cultivation and slop position effects on soil organic matter. Soil Sci. Soc. Am. J. 52, 1371– 1376. Wu, T., Li, F., Cao, X., Xiang, S., Li, Y., Li, L., 2003. Principles and techniques of soil fertility management in dryland area on the Loess Plateau within Gansu, China. In: Ji, L., Chen, G., Schnug, E., Hera, C., Hanklaus, S. (Eds.), Proceedings of the 12th International World Fertilizer Congress on Fertilization in the Third Millennium-Fertilizer, Food Security and Environment Protection, August 3–9, 2001, Beijing, China. Science and Technology Press of Liaoning, Shengyang, pp. 398– 414. Zhou, G.-Y., Ding, N.-P., 1999. Effect on soil fertility of combined application of organic and inorganic fertilizers in Heilu soil. In: Lin, S. (Ed.), Corpus of Soil Science and Plant Nutrition Research. Science and Technology Press in Shaanxi, Xian, pp. 613–618 (in Chinese and with English abstract).