Study site and experiment design

The experiment site was situated at China Agricultural University's Qu-Zhou experimental station (115°01' E, 36°52' N) in North China Plain. The climate in this experimental station is warm, semi-humid and has rainy summers and dry, cold winters. The mean annual temperature is 13.2 °C and ranges from a minimum of -2.9 °C in January to a maximum of 26.8 °C in July. The mean annual precipitation is 542.7 mm, of which 60% occurs from July to September, and the annual non-frost period is 201 d. The soil at study site is an Aquic Cambisol according to the FAO.

The on-going long-term fertilizer experiment was designed with four treatments, laid out in a randomized complete block design, with each plot 4 × 8 m in size. The field included the following four treatments: effective microorganism compost treatment (EM, 15 t ha¹); traditional compost treatment (TC, 15 t ha¹); N and P fertilizer treatment (CF, 265.5 kg N ha^-1 and 90 kg P₂O₅ ha^-1), and control, unfertilized treatment. The traditional compost was 60% crop straw, 30% livestock dung, 5% cottonseed-pressed trash, and 5% bran (a mean nutrient content: 100.5 kg N ha^-1, 36 kg P₂O₅ ha^-1, and 196.2 kg K₂O ha^-1). Every 50 kg of EM compost consisted of 50 kg traditional compost appended with 200 mL concentrated EM; 1 mL of EM concentrate contained a minimum of 10⁵ viable organisms of Streptomyces albus, Propionibacterium freudenreichii, Streptococcus lactis, Aspergillus oryzae, Mucor hiemalis, Saccharomyces cerevisiae, and Candida utilis, in addition to an unspecified number of Lactobacillus sp., Rhodopseudomonas sp., and Streptomyces griseus. Nitrogen fertilizer used was 48% ammonium bicarbonate (17% N) and 52% urea (46% N); P fertilizer used was calcium super-phosphate (12% P₂O₅).

Compost or chemical fertilizer twice were applied as basal fertilizers annually, just before the planting of summer maize and winter wheat, and the amount of fertilizer was uniform in maize and wheat crop. All compost and chemical fertilizer were evenly spread into the soil surface by hand and immediately incorporated into the plough layer by tillage before sowing maize or wheat. Tillage was done to 20 cm depth by plough and followed by harrow. The fertilized and unfertilized plots were the same type of tilth. Neither compost nor chemical fertilizer was applied during maize and wheat growth stages. The fertilizer types, rates and times of application were typical for this region. Experimental field were cultivated with maize (Zea mays L.) in summer and with wheat (Triticum aestivum L.) in winter at the beginning of 1993, which is the typical planting mode in this region. The maize was sowed with direct seeding in June and harvested in October, followed by wheat, which was cultivated in October and harvested in June next year. The aboveground crop was reaped, and no straw was returned to incorporate into soil. However, maize or wheat stubble and root was incorporated into the soil with plow before planting wheat or maize. Besides fertilizer of application, all the other agronomic practice was identical. The total amount of C, N, P, and K input to the soil in all the four treatments through fertilization every year were showed in Table 1.

Table 1 Total amount of C, N, P, and K input to the soil through fertilization in all treatments every year. 

Control: Unfertilized treatment; CF: chemical fertilizer treatment; TC: traditional compost treatment; EM: effective microorganism compost treatment.

Soil sampling and analysis

Initial soil samples were collected in 1993 prior to the start of the experiment. After 11 yr field experiments, soil samples were collected from the 0-20 cm soil layer at each experiment plots, using a 2.5 cm diameter soil auger before maize planting on 8 June 2004. Each soil sample consisted of 15 cores (2.5 cm diameter), which were mixed to form a composite sample. The soil samples were stored in insulated and tied plastic bags and transported to the laboratory.

Soil bulk density was determined using the core volume and dry soil weight. Soil sub-samples were air-dried for 14 d at room temperature, sieved through a 1 mm screen, mixed, and sub-samples were used to analyze for available N, available P, available K and soil pH. The air-dried sub-samples were ground to pass through a 0.25 mm sieve to determine SOC and total N content. The potassium dichromate external heating method was applied to determine SOC content. The semi-micro Kjeldahl method and the alkaline-hydrolyzable diffusion method were applied to determine total N and available N content. Soil available P was extracted with 0.5 mol NaHCO₃ L^-1 (soil:solution 1:20) and measured with the Olsen method. Soil available K was extracted with 1 mol NH₄Ac L^-1 (soil:solution 1:10) and measured with the flame photometry method. Soil pH was measured in 0.01 mol CaCl₂ L^-1 slurry (soil:solution 1:2.5) using a glass electrode.

Crop harvesting

Maize cob was harvested by hand on 8 October 2004. Cornstalk was reaped and removed from plots. Maize cob was dried by sun and grains were separated from cob using thresher. Maize grain yields were weighted after sun-drying and were recorded from a whole plot. Twenty maize cobs were used to evaluate agronomic character of corn. All the data were expressed on dry mass basis.

Calculation of soil C, N stock and sequestration

Soil organic C and total N stock were calculated by multiplying SOC or total N concentrations by bulk density and depth. The amount of sequestered organic C and total N in 0-20 cm soil depth in every treatment was estimated after subtracting the initial SOC and total N stocks.

Statistical analysis

All obtained data were subjected to ANOVA using the SPSS 11.5 software package (IBM, Armonk, New York, USA) and were used to evaluate differences between treatments. Difference obtained at P < 0.05 level was considered as significant using the least significant difference (LSD) test. Pearson linear correlation (two tailed) was used to evaluate the relationships between the parameters.

Soil organic C and total N concentration and stock

The content of SOC and total N through 11 yr field experiment (including control) was increased by 28.18%-112.54% and 37.29%-118.64% contrasted to the initial values, respectively. The highest contents of SOC and total N were found in the EM compost treatment and the lowest values were found in the control treatment after 11 yr experiments. Application of compost significantly (P < 0.05) increased SOC and total N contents in contrast to chemical fertilizer and control treatments. The content of SOC and total N due to 11 yr fertilizer application was increased by 6.57%-65.82% and 2.47%-59.26% in comparison to control, respectively (Table 2).

Soil organic C stock, sequestration and sequestration rate in two compost treatments were significantly (P < 0.05) higher than in the chemical fertilizer and control treatments. Soil total N stock, sequestration and sequestration rate in two compost treatments were significantly (P < 0.05) higher than in the chemical fertilizer and control treatments (Table 3).

Soil nutrients, soil pH and bulk density

The content of available N and available K through 11 yr field experiment (including control) were increased by 5.90%-73.29% and 11.43%-216.98% in contrast to the initial values, respectively. The content of available P through 11 yr field experiment in the control treatments were significantly (P < 0.05) decreased by 84.42% in contrast to the initial values. The highest contents of soil available N, available P and available K were observed in the EM compost treatment and the lowest values in the control treatment after 11 yr field experiment. The content of available N and available K in two compost treatments were significantly (P < 0.05) higher than in the chemical fertilizer and control treatments, and the content of available P in the fertilization treatments significantly (P < 0.05) higher than in the control treatment. The content of available P and available K in the EM compost treatment was significantly (P < 0.05) increased by 39.68% and 28.11% than in the traditional compost treatment, respectively (Table 2).

Soil pH and bulk density through 11 yr field experiment (including control) were decreased by 14.14%-18.47% and 2.65%-12.58% in comparison with the initial values, respectively. Soil pH in the fertilization treatment was significantly (P < 0.05) decreased by 1.59%-5.05% in contrast to control treatment. Soil pH in the fertilization treatments were significantly (P < 0.05) lower than in the control treatment, and were significantly (P < 0.05) lower in the compost treatments than in the chemical fertilizer treatment, and were significantly (P < 0.05) lower in the EM compost treatment than in the traditional compost treatment. Soil bulk density was significantly (P < 0.05) lower in the compost treatments than in the chemical fertilizer and control treatments (Table 2).

Table 2 Some soil physical-chemical properties in long-term fertilization experiment (1993-2004). 

Control: Unfertilized treatment; CF: chemical fertilizer treatment; TC: traditional compost treatment; EM: effective microorganism compost treatment.

Different letters indicate significant differences (P < 0.05) between treatments according to LSD multiple comparison.

Table 3 Soil organic C (SOC), total N stock and sequestration in long-term fertilization experiment after 11 yr of different fertilization treatments. 

Control: Unfertilized treatment; CF: chemical fertilizer treatment; TC: traditional compost treatment; EM: effective microorganism compost treatment.

Different letters indicate significant differences (P < 0.05) between treatments according to LSD multiple comparison.

Maize grain yields and yield components

Maize plant height, spike length, spike width, line number per spike, grain number per line, grain number per spike, grain weight per spike, 1000-grains weight, and grain yields in the fertilization treatments were 16.38%-20.87%, 22.49%-52.99%, 8.37%-17.04%, 11.19%-20.48%, 27.83%-68.43%, 41.57%-102.14%, 79.42%-154.22%, 14.47%-30.09%, and 62.36%-163.49% increased compared with the control treatment. Spike length, grain number per spike, grain weight per spike, 1000-grains weight, and grain yields of maize in the EM compost treatment were significantly (P < 0.05) increased by 10.34%, 9.98%, 15.02%, 4.96%, 15.39% in contrast to the traditional compost treatment, respectively. Spike length, gain number per spike, gain weight per spike, 1000-grains weight, and grain yields of maize in the fertilization treatments were significantly (P < 0.05) higher than in the control treatment, and were significantly (P < 0.05) higher in two compost treatments than in the chemical fertilizer treatment. Plant height, spike width, line number per spike, and grain number per line in the fertilization treatments were significantly (P < 0.05) higher than in the control treatment, and spike width, grain number per line in two compost treatments were significantly (P < 0.05) higher than in the chemical fertilizer treatment (Table 4).

Table 4 Maize grain yields and yields components in long-term fertilization experiment in 2004. 

Control: Unfertilized treatment; CF: chemical fertilizer treatment; TC: traditional compost treatment; EM: effective microorganism compost treatment.

Different letters indicate significant differences (P < 0.05) between treatments according to LSD multiple comparison.

Correlation analysis

A significantly positive regression relationship could be established between SOC sequestration rate and grain yields of maize. Similarly, soil N sequestration rate was significantly positively correlated with grain yields of maize (Figures 1 and 2).

Figure 1 Regression relationship between soil organic C sequestration rate and grain yields of maize. 

Figure 2 Regression relationship between soil N sequestration rate and grain yields of maize. 

Soil organic C and total N

Soil organic C and total N content in two compost treatments were significantly increased than in the chemical fertilizer and control treatments and initial values, as was similar to the other study (^{Chu et al., 2007}). This manifested that increment of SOC and total N is more distinct in organic fertilizer treatments than in inorganic fertilizer treatments. This reason was that there had an addition of organic C incorporated into soil in organic fertilizer treatments. Soil organic C and total N contents were only slightly higher in chemical fertilizer treatment than in the control treatment, which was consistent with the previous reports (^{Su et al., 2006}). Nevertheless, ^{Zhong and Cai (2007)} in a paddy soil derived from quaternary red clay and ^{Mandal et al. (2007)} in Inceptisol of India found that SOC C and total N content were significantly higher in the chemical fertilizer treatment compared with unfertilized treatment. The possible reason was that there had great crop residues in the chemical fertilizer treatment than in control treatment.

Soil organic C and total N content were higher in control treatment than in initial values, and ^{Shen et al. (2007)} and ^{Bi et al. (2009)} reported similar results in double rice cropping systems or in rice-wheat agroecosystem of China. Similarly, ^{Kundu et al. (2007b)} also reported that SOC and total N content were significantly increased in the control treatment compared to the initial values in a sandy loam soil of the Indian Himalayas after 30 yr of cropping. Soil organic C contents marginally improved from the start of the experiment in the unfertilized control plots due to the addition of C from the crop stubbles and roots, higher humification rate, and lower decay rate (^{Kundu et al., 2007b}). ^{Kundu et al. (2007a)} reported that 26% of the added crop biomass C was annually humified from humus into soil and the increase in SOC concentrations.

Our result showed that application of compost or chemical fertilizer including unfertilized treatment could increase SOC stock, and average the SOC stock was increased by 9.98 t ha^-1 (0-20 cm) in the 11 yr period, which indicated potential for organic C sequestration in upland soils (^{Fan et al., 2008}). This result was similar to the other long-term fertilizer experiment site reported by ^{Zhou et al. (2013)}, who found that SOC stock in all treatments showed increasing trends, including the control treatment in a semi-arid cropland of the Loess Plateau region. However, an average SOC loss of 5.82 t ha^-1 (0-20 cm) in inorganic fertilizers alone treatments was observed over the 23 yr period in the arid region of northwest China, but long-term application of manure alone or combined with inorganic fertilizers could evenly sequestrate SOC of 5.90 t ha^-1 in the 23 yr period (0-20 cm) (^{Su et al., 2006}). Soil organic C sequestration (0-20 cm) was 5.81 t ha^-1 in inorganic fertilizer treatment, but was 14.33-15.35 t ha^-1 in compost treatments through the 11 yr period in the present study. This manifested that application of compost have strongly promoted SOC accumulation because compost was addition C resources incorporated into soils (^{Chakraborty et al., 2011}). At the same time, crop biomass in compost treatments was higher than in inorganic fertilizer treatment; correspondingly, crop residue (including crop stubble and root biomass) also were higher in compost treatments than in inorganic fertilizer treatment (^{Kundu et al., 2007a}; ^{Gong et al., 2011}). The soil organic C sequestration rate (0-20 cm) was 0.40-1.40 t ha^-1 yr^-1 in this experiment (average 0.91 t ha^-1 yr^-1); however, soil organic C sequestration rate was only 0.16-0.42 t ha^-1 yr^-1 after application of farmyard manure in an arid region of northwest China (^{Su et al., 2006}). In contrast to the other long-term experiment sites, SOC sequestration rate was only 0.15-0.51 t ha^-1 yr^-1 (0-20 cm) in Belle Mina, Alabama, southeastern USA (^{Sainju et al., 2008}), was 0.04-0.16 t ha^-1 yr^-1 in an Inceptisol in south-eastern Norway (^{Holeplass et al., 2004}). Soil organic C sequestration was significantly positively correlated with grain yields of maize. Generally, high crop grain yields corresponded high crop residues (including roots and stubbles), and high crop residues were closely associated with high SOC storages.

Soil total N stock (0-20 cm) in all treatments was increased by 0.62-1.66 t ha^-1 in the 11 yr period. ^{Sainju et al. (2008)} reported that soil total N stock (0-20 cm) was increased by 0.15-0.49 t ha^-1 in the 10 yr period in southeastern USA. Soil total N sequestration rate was significantly higher in two compost treatments than in the chemical fertilizer and control treatments. Similarly, ^{Gami et al. (2009)} observed that soil total N stock was significantly higher in farmyard manure treatment than in NPK and unfertilized treatments in Nepal. ^{Gong et al. (2011)} revealed that soil organic N sequestration was significantly higher in organic fertilizer treatment than in inorganic fertilizer treatment and control in the North China Plain. However, an average soil total N loss of 0.25 t ha^-1 (0-20 cm) in inorganic fertilizers alone treatments was observed over the 27 yr period in the Loess Plateau region of northwest China, and soil total N were sequestrated when long-term application of manure alone or combined with inorganic fertilizers (^{Zhou et al., 2013}).

Soil nutrients, soil pH and bulk density

The content of available N in compost treatments was significantly higher than in the chemical fertilizer and control treatments, and there had nonsignificant difference between chemical fertilizer and control treatment, and ^{Bi et al. (2009)} reported consistent with results. Nevertheless, ^{Shen et al. (2007)} found that available N content in the manure and chemical fertilizer treatments was significantly higher than in the control treatment. The content of available N in compost treatments was significantly higher than initial values, and ^{Bi et al. (2009)} also observed similar result in Jinxian experimental site. The content of available P was significantly increased due to long-term fertilization compared with the control. Similarly, ^{Liu et al. (2010)} found that available P content in manure and chemical fertilizer treatments was significantly higher than in the control treatment. The content of available P in the EM compost treatment was significantly higher than in chemical fertilizer treatment and consistent result reported by ^{Chu et al. (2007)}. However, ^{Lee et al. (2009)} reported that available P content in NPK fertilizer treatment was significantly higher than in the compost treatment. Soil available P content was significantly decreased in control treatment compared with initial values. Similarly, ^{Kundu et al. (2007b)} and ^{Lv et al. (2011)} also reported that available P content in long-term unfertilized soil was decreased. However, ^{Singh et al. (2007)} observed that available P content in 8 yr unfertilized Vertisol in India was increased. The available K content was significantly higher in the compost treatment than in the chemical fertilizer and control treatment and initial values, and ^{Liu et al. (2010)} reported consistent with results. However, ^{Shen et al. (2007)} reported that available K content was significantly decreased in the manure and chemical fertilizer treatments including control treatment compared to initial values, and ^{He et al. (2012)} found that available K content in the chemical fertilizer and control treatments was lower than initial values. The concentration of available P and K was significantly higher in the EM compost treatment than in the traditional compost treatment, which revealed that the effect of improved soil fertility with EM compost was better compared with traditional compost. This was because there are many advantageous microbes in EM compost, which accelerated release of plant-available nutrients from soil.

Soil pH in fertilization treatments was significantly decreased than in control treatment (^{Chu et al., 2007}). Soil pH in two compost treatments was significantly lower than in the chemical fertilizer treatment, which was contrasted to the result reported by ^{Liu et al. (2010)}. Soil pH in all treatment was significantly decreased than initial values. Similarly, application of fertilizer decreased soil pH (^{Hao et al., 2008}). However, soil pH in the topsoil rose from 5.2 to 6.0 in the West African Sahel irrigated rice within 4 yr fertilization and remained stable thereafter (^{Haefele et al., 2004}). Application of the compost caused increases in soil pH, with a slight reduction noted for NPK treatment in Ireland (^{Courtney and Mullen, 2008}).

Soil bulk density was significantly lower in two compost treatments than in the control treatment, as was consistent with the other report (^{Gami et al., 2009}); however, it was contrasted to the result of ^{Nayak et al. (2012)}. ^{Rasool et al. (2007)} observed that soil bulk density was significantly decreased in organic fertilizer treatments, but there had nonsignificant difference between chemical fertilizer and control treatment in paddy field, as was consistent with our results. ^{Gong et al. (2009)} found that soil bulk density was significantly decreased due to long-term application of manure or chemical fertilizer, and soil bulk density in the manure treatment was significantly lower than in the chemical fertilizer treatment.

Maize yields

Maize plant height was increased due to long-term fertilization, which was consistent with the reported by ^{Hu and Qi (2013)}, who found that long-term application of organic or inorganic fertilizer enhanced wheat plant height. Maize grains per spike and thousand grains weight were significantly increased due to long-term soil amendments. ^{Haefele et al. (2002)} observed that long-term application fertilizer significantly increased grains per spike and thousand grains weight of rice ‘Jaya’. In contrast, ^{Surekha et al. (2010)} reported that thousand grains weight of rice was not influenced by fertilization. ^{Kato and Yamagishi (2011)} found that thousand grains weight of wheat in the manure treatment was lower than in inorganic fertilizer treatment, which was contrasted to our results. Maize grains per spike and thousand grains weight were significantly increased due to EM application in the present study. Similarly, hundred grains weight of vigna were significantly increased due to EM application in farmyard manure and crop residues amended soils (^{Javaid and Bajwa, 2011a}).

Maize grain yields were significantly increased due to long-term application of compost or chemical fertilizer. Similarly, maize grain yields in the chemical fertilizer treatment were significantly higher than in the control treatment (^{He et al., 2012}), and maize grain yields were significantly increased due to long-term application of farmyard manure and chemical fertilizer (^{Liu et al., 2010}). ^{Mugwira et al. (2002)} also found that application of manure significantly increased maize grain yields. Maize grain yields in two compost treatment were significantly higher than in the chemical fertilizer treatment, as was consistent with the result reported by ^{Rasool et al. (2008)}, who found that maize grain yields in the farmyard manure treatment were significantly higher than in the chemical fertilizer treatment. Similarly, average rice and wheat grain yields in the farmyard manure treatment were significantly higher than in the chemical fertilizer treatment in 20 yr fertilizer experiment of Nepal.

Maize grain yields in the EM compost treatment were significantly higher than in the traditional compost treatment. These results were similar to the previous studies. For example, wheat and rice grain yields were increased when EM combined with manure or chemical fertilizer was applied. Soybean and mung-bean grain yields were significantly increased due to EM application in farmyard manure amendment soils (^{Javaid and Bajwa, 2011b}). Grain yields of rice were enhanced by 46% due to EM application in the green manure amended soils. The vigna grain yields increased by 24%, 15%, and 84% when application of EM in combination with farmyard manure, crop residues or NPK fertilizer compared with corresponding treatments without EM application, respectively (^{Javaid and Bajwa, 2011a}). The higher crop yields in EM in combination with organic materials amended soils could be attributed largely to the activity of the introduced exotic beneficial microorganism, which stimulated the decomposition of organic materials and the release of nutrients for plant uptake (^{Javaid and Bajwa, 2011b}).