IntroductionTop

Soil aggregation withstands soil fertility by reducing erosion and improving soil aeration, as well water infiltration and retention. In addition, soil aggregation prevents soil organic matter (SOM) from being mineralized by physically reducing the accessibility of organic compounds for microorganisms, extracellular enzymes, and oxygen (Spohn & Giani, 2010).

The study of aggregates is one way to quantify whether management practices improve natural characteristics and agricultural capacity of soil. Soil aggregate and structure stability is the result of the interaction among many agents, such as the environment, management practices, crop, inherent soil features, and soil biological and non-biological processes (Bronick & Lal, 2005). The size of soil aggregate and stability are also used to characterize soil structure because those indicators are correlated with various soil functions, including carbon sequestration and gas exchange through physical protection of SOM (Rabot et al., 2018). Other studies show the stability index is higher under reduced tillage conditions than under conventional tillage (CT) conditions in spring and summer (Eynard et al., 2004). Land use and associated management practices, such as crop sequencing, fertilization, soil conditioning, drainage, and irrigation are the major and most direct ways of affecting soil structure and properties by their impact on destruction forces and aggregate-creating processes (Lehrsch et al., 2012). Wright et al. (2007) reported that the use of the minimum tillage (MT) system improves the content of soil organic carbon (SOC) and increases soil microbial population. In other regions of the world, no-tillage (NT) and MT practices have shown to improve soil quality and crop productivity (Norton et al., 2012). In most agricultural systems, crop residues are the main substrate for soil microorganisms (Govaerts et al., 2009), and their mineralization rate depends on the placement and existence of favorable conditions (Helgason et al., 2014).

The sustainability of tillage systems has received considerable attention in recent years. In addition, with an increase in the adoption of conservation-oriented tillage practices, they are considered more sustainable than CT. Soil management involving reduced tillage has often been reported to improve soil structure (Oyedele et al., 1999). MT operations for example, have positive influence on soil physical, chemical and biological properties as macro-pore structure, aggregate stability, nutrients availability, and the diversity of microbial populations while reducing soil disturbance (Heidari et al., 2016). Soil aggregates have been found to be more stable in reduced tillage than in conventional moldboard tillage (Pagliai et al., 2004). Although the effects of tillage operations on soil structure has been well documented, past research mainly focused on aggregate stability (Emami et al., 2012). In addition, soil aggregate distribution and stability measurements have been suggested as soil quality indices. However, aggregate stability is often measured in terms of a specific aggregate size class that is not a measurement of whole soil structure (Six et al., 2000).

Due to the importance of soil in Iran’s food production and its key role in natural ecosystems, it is necessary to pay more attention to soil management issues in achieving sustainable development. Therefore, it is necessary to study the effective features of the maintenance and enhancement of soil structure stability indices resulting from various management practices. In arid and semi-arid regions, such as Northern Khorasan Province (Iran), because of limited rainfall and the decomposition of SOM, the careful management of organic matter and soil properties, such as soil biota and physical indices, as well as conservation of water, are very important. In addition, due to the high rate of cultivation in this region, the conservation of the structure and properties of the soil is of high significance. Therefore, to promote the preservation of soil and water resources in agricultural systems in arid and semi-arid regions, conservation tillage operations (NT and MT) are necessary, in that they can contribute to avoiding soil degradation by compaction.

The key hypothesis was that in the three year-old tillage systems for a rotation use of three years (wheat, canola, and wheat), NT and MT practices cause shifts in the soil bacterial community as well as in the microbial biomass carbon to improve soil structure stability indices relative to CT practices. Based on this introduction, in this research the effects of different tillage methods on soil stability indices, microbial biomass carbon, and the number of bacteria were investigated.

Material and methodsTop

Experimental area

The present research was conducted in Northern Khorasan Province (Iran). This province has the geographic coordinates of (37°28′N, 57°19′E), is 1070 m above the sea level, and has a temperate warm semi-arid climate. The slope of the studied site was 0-2%. The region of the study has warm summers and relatively cold winters. Based on a 44-year statistical period, the average annual rainfall at the meteorological station of the region is reported to be 259.7 mm, with the mean annual temperature of 13.7 °C.

Experimental design and treatments

This study used a factorial experiment with a randomized complete block design and three replications. The experimental treatments of this study consisted of different soil tillage methods: (1) CT (moldboard plowing, disc, leveling, furrowing, and seed drilling), (2) MT (chisel plowing, furrowing, and seed drilling), and (3) NT (seed drilling), which were considered together as the first factor. In addition, the management of plant residues, as the second factor, included no residue, 40% of plant residues, and 70% of plant residues. In the direct drilling method (NT), tillage operations were not done before seeding, and cultivation was done with a one-time tillage operation. The area of each plot for the tillage systems (first factor) was 30 m in width × 50 m in length, and the area of each individual plot for the plant residues (second factor) was 10 m in width × 50 m in length. The total area of the experiment (27 plots) was 13500 m2. In this study, 9 treatments were used, which included CT + no plant residues (CT0), CT + 40% of plant residues (CT40), CT + 70% of plant residues (CT70), no tillage + 0% of plant residues (NT0), NT + 40% of plant residues (NT40), NT + 70% of crop residues (NT70), MT + 0% of plant residues (MT0), MT + 40% of plant residues (MT40), and MT + 70% of plant residues (MT70), with each treatment having had three replicates.

The sources of nitrogen, phosphorous and potash were urea, triple superphosphate and sulphate of po-tash, respectively. Based on the soil testing results, the amounts of the applied fertilizer for all plots were 230-145-40 NPK. In addition, trifluralin as a herbicide for canola and 2, 4-Dichlorophenoxy acetic acid (2, 4-D) as a herbicide for wheat were applied 2 L/ha and 1.5 L/ha, respectively.

This study explored the effects of the three year-old tillage systems of CT, NT, and MT with the plant residues of 0%, 40%, and 70% on soil structure stability indices, the number of bacteria and microbial biomass carbon for a rotation of three years wheat (Triticum aestivum L.), canola (Brassica napus L.), and wheat.

Soil sample collection and experimental materials

Surface soil layers (0-20 cm) were randomly sampled from 10 points in each treatment plot with an auger and thoroughly mixed into one composite sample. For this purpose, a total of 27 composite samples (9 treatments × 3 replicates = 27 samples) were collected in sterile plastic bags and divided into two parts, for physical and micro-bial analysis, respectively. To manage plant residues in different tillage treatments, some plots received any resi-dues (0%), whereas, in the other plots, the plant residues (40% or 70%) were incorporated into the first 10 cm of soil uniformly by tilling using a cultivator.

Measurement of soil physical and microbial properties

In the end, after the third year, the soil samples were collected from the soil surface (0-20 cm), and soil structu-ral indices, i.e. the whole soil stability index (WSSI) and normalized stability index (NSI), the number of bacteria, as well as microbial biomass carbon were measured. The properties of the initial soil are listed in Table 1. In this study, soil texture was determined based on the hydrome-ter method (Bouyoucos, 1936, 1962; Beverwijk, 1967). In addition, the soil saturated paste extract (ECe) and the pH of the saturated soil paste were determined based on the study of Smith & Doran (1996), and the amount of SOC, as an index of the SOM, was determined based on the study of Nelson & Sommers (1982).

WSSI, aggregate distribution, and water-stable aggre-gation (WSA) were measured in dry-sieved aggregates in four aggregate size classes (9.5–2, 2–1, 1–0.25, and 0.25–0.053 mm). The dry sieving practice consisted of placing soil atop a screen with the size equal to that of the largest aggregates in size, tapping the sides at least 50 times with the palm to make the soil pass through the screen, collec-ting the soil passing through the screen on a piece of kraft paper, and pouring it onto a screen equal to the smallest aggregates in size, followed by tapping. Each aggrega-te of the size classes was collected individually from the largest to the smallest ones. The weight of the aggregates in each size class was measured and used to calculate the proportion of dry-sieved the aggregates in each size class (Pai) to the whole soil (Eq. [1]).

The soil on top of the 9.5 mm screen and below the 0.053 mm screen was collected and weighed as part of the summed total weight (WT) (Nichols & Toro, 2011). Pai was achieved as follows:

where, WA is the weight of the total material in each size class i, WC represents the weight of the coarse material measured during wet sieving for size i, WO is the weight of the aggregates placed on the sieve before wet sieving for size i.

WSA was measured in four subsamples from each ag-gregate size class according to the modified Kemper & Rosenau (1986)´s method. In brief, the aggregates (4 g for 9.5–2 and 2–1 mm aggregates, 2 g for 1–0.25 mm aggre-gates, and 1 g for 0.25–0.053 mm aggregates) were placed onto screens 1/4 of the smallest size and capillary rewet-ted for 10 min. Stable aggregates were separated by me-chanical wet sieving for 5 min using a piece of apparatus introduced by Kemper & Rosenau (1986). The material collected on the sieve was washed gently into weighing boats, dried at 70 °C, and weighed. The coarse material, including sand, roots, and particulate organic matter, was removed by dispersing the aggregates in 0.5% sodium hexametaphosphate, which was shaken periodically over a 5-min period. Next, pressurized water and rubber police-men were used to push the disrupted aggregates through a screen matching the smallest aggregates in the size class. The material remaining on the screen was collected, dried at 70 °C, weighed, and subtracted from the weight of the aggregates collected after wet sieving. The formula used to calculate WSA for each size class (Nichols & Toro, 2011) is as follows:

In addition, the dry aggregate size distribution and WSA calculated above were included in the formula for the WSSI (Nichols & Toro, 2011) as follows:

where, n represents the number of the aggregate size clas-ses, i equals n and decreases by an increase of 1 from the largest to the smallest aggregate size classes, I is each size class (Nichols & Toro, 2011). In addition, the NSI is the measure of soil stability based on comparing the aggre-gate distribution before and after the disruption proposed by Six et al. (2000). Briefly, samples from each treatment were applied into three groups before wet sieving. In one group, air-dried soil samples were rewetted by capillary overnight before wet sieving. In another group, soil sam-ples quickly were wet-sieved without rewetting (slaked), and the final group was provided for a maximum disrup-tion to measure the coarse material (i.e. sand and organic matter). Maximum disruption used forced water to destroy aggregates and wash them through stacked sieves. The capillary–rewetted and slaked samples were wet-sie-ved for 2 min through three sieves (2000, 250, and 53 µm) individually by moving the material that went through the sieve onto the next smaller sieve. The water-stable aggregates on sieves in both the capillary-rewetted and slaked treatments were collected, dried at 70°C, weighed, and corrected for the coarse material (Six et al., 2000).

The formula for the calculation of the disruption level of a size class upon slaking (DLSi) is as follows:

where, Pio stands for the proportion of the total sample weight in size class I before disruption (i.e. rewetted), Pi represents the proportion of total sample weight in size class I after disruption (i.e. slaked), Sio is the proportion of sand with size I in the aggregates of size I (aggrega-te-sized sand) before disruption, and Si is the proportion of sand with size I in the aggregates of size I after dis-ruption. All proportions have been expressed on a soil weight basis (g fraction g-1 soil). Size classes used in this study included I = 1: < 53 µm, I = 2: 53-250 µm, I = 3: 250-2000 µm, and I = 4: > 2000 µm. The factor before the multiplication sign in Eq. [4] calculates the level of disruption caused by slaking and corrects the proportions of aggregate size classes for the aggrega-te-sized sand content. Besides, this factor ensures that only weight losses are used in calculating the index; in other words, this factor will be 0 if there is a weight gain. The factor after the multiplication sign normalizes the weight loss to the maximum value for that size class. The aggregate-sized sand content could be used for sand correction. The whole soil disruption level (DL) was cal-culated as follows:

DL is the weighted sum of disruption for each size class I because a weight loss in a smaller size class is more indicative of soil instability than a weight loss in a larger size class.

Weighting factors for disruption in different aggregate size classes were assigned arbitrarily by ranking the ag-gregate size classes from 1 to 4. The weighting factors are not based on arithmetic or geometric means because the use of mean indices will be questionable if the aggregate distribution is skewed (Stirk, 1958), which is often the case. The maximum disruption level for each size class I [DLSi (max)] was calculated by the following formula:

where, Pp is the primary sand particle content with the same size as the aggregates of the size class after the complete disruption of the whole soil. The whole soil DL (max) is calculated using Eq. [5] with DLSi replaced by DLSi (max). The NSI is then computed as follows:

The DL was divided by the DL (max) to normalize the DL for the maximum possible disruption based on the pri-mary sand particle-size distribution.

The percentage of aggregate destruction (PAD) is ano-ther suitable index used to appraise soil physical structu-re, which is based on dry-sieved aggregates' mean weight diameter (dry MWD) and wet-sieved aggregates' mean weight diameter (wet MWD), which was calculated (Ca-tes et al., 2016) as follows:

where, Md and Mw represent the weight of dry MWD >0.25 mm and the weight of wet MWD >0.25 mm, respectively.

To determine the soil microbiological activity, 500 g of the rhizosphere soil sample was collected from the plot in each replication in every sampling. After removing the roots, the samples were passed through a 5-mm sieve. In addition, after removing the residues, the samples were passed through another 2-mm sieve. The number of mi-crobial populations was determined using a fluorescence microscope (Li et al., 2002). For this purpose, 10 μL of the final solution was placed on a slide and spread uni-formly with a needle. The slide had a concentric circle with the specific surface area of 1 cm2. For the purpose of bacterial staining, 5 mL of an orange acridine solution was added to the slide. After drying the slides, the num-ber of bacteria, lightened by staining, was counted using a 500x magnification fluorescence microscope. In the end, the number of bacteria in each gram of the dry soil was calculated according to the following equation:

where BF represents the number of bacteria counted on the slide, DF is the dilution factor (150), and AF shows the area of a microscopic field on the slide.

To measure microbial biomass carbon, the fumiga-tion-extraction method (Sparling & West, 1988) was used. In this method, the soil samples were sterilized by chloroform and extracted using a potassium sulfate solu-tion. In the end, the organic carbon content of the extract was measured, with the microbial biomass carbon obtai-ned as follows:

where S represents organic carbon (mg) per 100 grams of the dry soil, and C is organic carbon (mg) per 100 grams of the non-fumigated dry soil (control).

Table 1. Physicochemical properties of the initial soil in the ex-periment

ECe: electrical conductivity of a saturated soil extract. OC: or-ganic carbon. TNV: total neutralizing value (percentage of equi-valent CaCO3).

Statistical analysis

The analysis of variance for different tillage methods and plant residues on soil stability indices, microbial bio-mass carbon, and the number of bacteria was performed following the ANOVA technique with the SPSS V.19.0 (SPSS Inc., Chicago, IL, USA, 2010). Comparisons be-tween statistical averages of the various treatments were made with the Tukey test (α ≤ 0.05). Computation and preparation of figures were done using Excel 2016.