Integrated crop-livestock system in tropical Brazil: Toward a sustainable production system

Agriculture, Ecosystems and Environment 190 (2014) 70–79

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Integrated crop-livestock system in tropical Brazil: Toward a sustainable production system Julio C. Salton a,∗ , Fabio M. Mercante a,b , Michely Tomazi a , Josileia A. Zanatta c , Germani Concenc¸o a , Wiliam M. Silva a , Marciana Retore a a

Embrapa Western Agriculture, Dourados, MS, Brazil CNPq Researcher, Brazil c Embrapa Forestry, Colombo, PR, Brazil b

a r t i c l e

i n f o

Article history: Received 4 March 2013 Received in revised form 30 September 2013 Accepted 30 September 2013 Available online 24 October 2013 Keywords: Soybean No-tillage Soil quality Savanna Pasture Brachiaria Crop rotation

a b s t r a c t Performance of soil management systems was initiated in 1995 in a field experiment in Dourados, MS, Brazil, with the following systems: CS – conventional tillage; NTS – no-tillage; ICLS – integrated croplivestock with soybean (Glycine max (L.) Merr.) and pasture under no-till, rotating every two years, and PP – permanent pasture. Pastures (Brachiaria decumbens) were grazed by heifers with stocking rate adjusted to constant supply of forage. The hypothesis was that rotation of crops and pastures would be more efficient and present beneficial effects to the environment. More complex and diversified production systems may exhibit synergism between components to result in better soil physical structure, greater efficiency in use of nutrients by plants, greater accumulation of labile fractions of soil organic matter, greater diversity and biological activity in soil, and lower occurrence of nematodes and weeds. Better soil conditions in ICLS allowed greater resilience; over the years of assessment soybean and pasture yields were less affected by drought and frost. The ICLS was very efficient, accumulating soil C and reducing emissions of greenhouse gases. Soil quality was improved in integrated systems with larger number of components and greater interaction between these components (ICLS) compared to simple systems. Based on soil attributes, we affirmed in this long-term study that the ICLS system is agronomically and environmentally efficient and sustainable. © 2013 Elsevier B.V. All rights reserved.

1. Introduction To intensify the production of food, fiber, and energy, production systems are constantly being reformulated to increase production efficiency, protect the surrounding environment, and/or promote ecological recovery. With the expansion of soybean cultivation, the degradation of large areas due to cattle grazing, and low livestock productivity (especially during the winter) in tropical Brazil, agricultural systems that integrate grain production and livestock grazing could be advantageous to both farmers and the environment. Integrated crop-livestock systems could make it possible to reconcile the recovery of pasture productivity with greater crop stability (Sulc and Tracy, 2007; Carvalho et al., 2010). Integrated crop-livestock systems have been adopted in several regions of Brazil. Due to the unique climatic, economic, and structural characteristics of each region, however, these production systems are arranged in various ways that differ in the species sequence, implementation details and rotation phases between

∗ Corresponding author. Tel.: +55 6734169748; fax: +55 6734169721. E-mail addresses: [email protected], [email protected] (J.C. Salton). 0167-8809/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.agee.2013.09.023

crop and livestock farming (Salton et al., 2008; Balbinot et al., 2009; Carvalho et al., 2010). In general, the adoption of integrated systems is beneficial by reducing pasture degradation (Kluthcouski and Stone, 2003). The benefits of integrated systems include increased soil fertility due to the accumulation of organic matter (Salton et al., 2010), improved nutrient cycling (Flores et al., 2008; Carvalho et al., 2010), increased fertilizer efficiency (Assmann et al., 2003), and better soil aggregation (Salton et al., 2008). Integrated systems also favor a more biologically active edaphic environment compared to other cropping systems (Silva et al., 2011). Similarly, crop fertilization improves both pasture productivity and livestock performance indexes (Carvalho et al., 2010). Rotation of crops with livestock can also help to break pest, disease, and weed cycles, thus reducing production costs, increasing economic and environmental outcomes (Lazzarotto et al., 2009; Martha Junior et al., 2011), and reducing the environmental risk posed by the proliferation of agrochemicals (Vilela et al., 2008). Accumulation of organic matter and improvements of soil chemical, physical, and biological properties suggest that integrated systems will reduce the environmental impact of agricultural production, mitigate greenhouse-gas (GHG) emissions, maintain or increase crop yields, improve water and nutrient usage

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

(Franchini et al., 2011), and enhance environmental resilience by promoting biological diversity and improving soil quality (Anghinoni, 2007). However, integrated systems must be designed differently across the large continental extent of Brazil. We compiled the results of a long-term experiment in Dourados, Mato Grosso do Sul, Brazil to assess the performance of simple and complex agricultural production strategies. Soil attributes and grain and beef production were monitored. Other agronomic and environmental features, such as weed dynamics and GHG emissions, were also monitored. We tested the hypothesis that a 2 year rotation of no-till crop and livestock production is more efficient and environmentally beneficial than less-complex systems, including both conventional and no-till systems. In this work, results of several reports during this long-term experiment were pooled through multivariate principal component analysis to infer about the efficiency and sustainability of integrated systems.

2. Materials and methods 2.1. The Dourados experiment A field experiment was initiated in 1995 at the Embrapa Western Agriculture station in the city of Dourados, Mato Grosso do Sul, Brazil (22◦ 16 56.08 S and 54◦ 48 17.17 W).The soil at this site is an Oxisol (kaolinitic with clay, silt, and sand contents of 630, 215 and 155 g kg−1 , respectively). The regional climate is classified as Cwa: a mesothermal, humid climate with hot summers and dry winters. The 25-ha experimental area was divided into plots where the following systems were used: a conventional system (CS) consisting of a soybean (Glycine max (L.) Merr.) monoculture followed by oats (Avena strigosa Schreb. or Avena sativa L.) under conventional soil tillage using a disk harrow every growing season; a no-till system (NTS) using a crop rotation including soybeans and corn (Zea mays L.) grown during the summer and wheat (Triticum aestivum L.), oat or turnip (Raphanus sativus L. var. oleiferus Metzg.) as cover crops, without soil tillage; an integrated crop-livestock system (ICLS) rotating every two years between crop species (soybean and oat) and pasture grass (Urochloa decumbens syn. Brachiaria decumbens Stapf) under no-till cultivation, with pastures grazed by heifers whose stocking rate was adjusted to ensure a constant forage supply of approximately 7% of body weight; and a permanent pasture (PP) consisting of B. decumbens grazed according to the same management strategy used in the ICLS. During experiment planning of this long-term trial, we envisioned the type of data to be collected. Since the area had homogenous soil and environmental conditions, we decided that large plots (2–4 ha per treatment), would make possible collection of all types of system data while allowing superposition of a geostatistical layer grid for use of geoprocessing tools. All soil and plant samples were collected using this grid of equidistant points spaced 30 m apart (Fig. 1). This design also facilitated mechanical operations like planting, harvesting and cattle management, while resembling management adopted in large, commercial production areas. NTS was divided into three subplots (NTSa, NTSb, NTSc) to account for different crops in the rotation. Every summer, two plots were planted with soybean, while the third was planted with corn; in winter, oat, wheat and turnip were used, according to the sequence oat/soybean/wheat/soybean/turnip/corn. ICLS was divided into two subplots (ICLSa, ICLSb) aiming to have cattle raising and grain crops in the same year with rotation between crop and livestock every two years. Soybean fertilization was based on replacing nutrient exportation in the grain of the crop, by applying 20 kg ha−1 of P2 O5 and 20 kg ha−1 of K2 O per ton of grain. For corn

71

and wheat, an average of 100 kg ha−1 of N per cropping cycle was applied as urea; cover crops (oats and turnips), as well as pastures, were not fertilized. Lime was applied in 2001 at the soil surface with 2 Mg ha−1 of dolomitic limestone. Soybean seeds were inoculated with Bradyrhizobium spp. in every planting. 2.2. Soil physical and chemicals attributes A fixed grid of 242 equidistant points distributed across the experimental area was used for sample collection to measure the physical and chemical attributes of the soil over time. Mean values were calculated and presented with the corresponding standard error (Fig. 1). 2.2.1. Physical attributes Soil aggregation was determined for monoliths with 10 cm × 10 cm with 20 cm depth that were kept in the shade and manually disaggregated by observing their weak points. Subsequently, all soil samples were air-dried and passed through a 9.52-mm mesh sieve. Following the method of Salton et al. (2008), soil aggregates from each sample were classified by size after both dry and wet sieving. A series of sieves with 4.76-, 2.00-, 1.00-, 0.50-, 0.25-, 0.105-, and 0.053-mm openings were used. After agitation, soil aggregates remaining in each sieve were weighed. Values obtained from dry and wet sieving were used separately to calculate mean weight-diameter (MWD). The ratio of the wet MWD to the dry MWD is the aggregate stability index (ASI). 2.2.2. Chemicals attributes Every two years, the 0–5-, 5–15-, and 15–30-cm soil-depth levels were sampled for chemical analyses. Concentrations of exchangeable bases [calcium (Ca) and magnesium (Mg)], phosphorus (P), and potassium (K) were measured based on Silva (1999). Total, organic, and inorganic P concentrations were measured by the ignition method (Saunders and Williams, 1955), in which samples were incinerated at 550 ◦ C for 1 h to analyze total P and not incinerated to analyze inorganic P. Total and inorganic P were then determined photocolorimetrically at 725 nm, and organic P was obtained as the difference between these two values. 2.2.3. Soil organic matter contents and quality Soil organic matter (SOM) and its fractions were monitored over time in soil samples obtained from the sampling grid and depths described in Section 2.2.2. Total organic carbon in soil (TOC) was expressed in terms of its content (g kg−1 ) and storage (Mg ha−1 ) in each soil layer. These calculations took into account the soil density values [C content (%) × soil density (g cm−3 ) × layer thickness (cm)].Carbon concentration was quantifying by dry combustion(TOC Analyzer –VCPN SSM-5000A, Shimadzu Corp, Japan). The carbon-retention rate of soil was calculated from the total C storage at the beginning of the experiment and measured periodically until 10 years. The amount of C in the particulate organic carbon (POC) fraction was determined through physical granulometric fractioning according to the method of Cambardella and Elliott (1992). First, a 20-g soil sample was dispersed in 80 mL (NaPO3 )6 at 0.5% and sieved through a 0.053-mm mesh sieve. Next, the mass and C concentration of the residue were determined, yielding the carbon concentration (g kg−1 soil) of the POC fraction. For samples collected in 2004, densimetric physical fractioning was used to obtain the free-light fraction (FLF), occluded-light fraction (OLF), and heavy fraction (HF) of the SOM. This method employs ultrasound dispersion in sodium polytungstate [Na6 (W12 O40 H2 )·H2 O] dissolved in distilled water and adjusted to a density of 2.0 g cm−3 . These procedures were described by Boeni (2007), who also determined the amounts of organic matter present in different types

72

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

Fig. 1. Location of the Dourados experiment, highlighting the grid of sampling sites used along the years and the disposition of cropping systems in the physical model (CS: conventional system, NTS: no-till system, ICLS: integrated crop-livestock system, PP: permanent pasture).

of C using NMR 13 C spectroscopy: alkyl C (0–45 ppm), O-alkyl C (45–110 ppm), aromatic C (110–160 ppm), and carboxylic C (160–220 ppm). Each type of C was quantified by integrating the corresponding peaks in the spectrum.

2.3. Soil biological attributes 2.3.1. Soil microbiology Microbiological attributes were assessed at the end of 12 years of experimental duration when the crops were in full bloom during summer and winter. Five samples were collected from each system, and each sample comprised 21 subsamples from the 0–0.10-m soil layer. Soil microbial biomass carbon, was determined with the fumigation–extraction method as proposed by Vance et al. (1987), with an extraction-efficiency correction factor (Kec) of 0.33. Microbial activity was assessed by the CO2 -evolution from aerobic incubation during seven days. The metabolic quotient was defined by the ratio of respiration and microbial biomass, as given by Anderson and Domsch (1990). The nutritional-quality index of the organic matter was represented by the microbial quotient, defined as the ratio of the microbial-biomass C to the total organic C in the soil.

2.3.2. Invertebrate soil macrofauna The invertebrate macrofauna community in soil was sampled during the winter and summer growing seasons in two consecutive years (2002 and 2003). At five equidistant points (30 m apart), soil monoliths (25 cm × 25 cm with 30 cm depth) were collected and subdivided into 0–0.1-, 0.1–0.2-, and 0.2–0.3-m layers, as described in the Tropical Soil Biology and Fertility method (Anderson and Ingram, 1993; Bignell, 2009). Invertebrate soil macrofauna were manually captured and examined under a binocular magnifier, and major taxonomic groups were identified and counted. Soil

invertebrate macrofauna community was characterized in terms of its density (individuals per m2 ) and richness (number of taxa).

2.3.3. Weed dynamics The weed community was monitored and characterized in each experimental treatment at the end of 16 years of experimental management. To monitor weed infestation, four 0.25 m2 areas within each management system were sampled prior to crop planting (at the beginning of summer). All emergent seedlings were collected, identified, and sorted by species. Seedlings were then dried for subsequent dry-mass determination. Soil seed-bank richness was estimated at depths of 0–5, 5–10, and 10–15 cm in each of the four management systems. Five 0.048 m2 samples from each depth were collected in each treatment. Soil moisture of each sample was determined and then corrected to 70% of field capacity. Samples were placed in 2-L PVC pots and placed in a greenhouse. All emergent seedlings in each pot were identified and collected at 20-day intervals for a total of four collections. After each collection, the soil in each container was turned over. Relative frequency, relative dominance, and relative abundance of each species were calculated, and the importance value index (IVI) was determined. Treatments and depths within treatments were compared using Sørensen’s similarity coefficient to estimate the current degree of similarity between weed communities as a function of sampling depth.

2.3.4. Nematodes Soil nematode population was assessed eight years after establishment. The experimental sampling-point grid was used, and five composite samples (of 12 subsamples each) were collected from each management system. The flotation–sedimentation–sieving method was used to extract the nematodes, followed by clarification through centrifugation in a sucrose solution (Jenkins, 1964).

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

2.4. Grain and beef production Grain productivity was assessed in summer and winter by manually harvesting 6-m2 areas for grain sampling points selected from the point grid described in Section 2.2. Plant material was processed, and average yields were calculated. In 1999, weight gain of crossbreed (Nelore × Hereford), castrated calves between 12 and 19 months old was assessed. In 2011, weight gain of 17-month-old female calves (Charolais × Nelore) was assessed. Cattle grazed continuously in the two production systems planted with B. decumbens, with two replicate areas. Every 28 days, 0.5-m2 pasture samples were collected at six locations in each paddock to measure fodder mass and ensure the maintenance of a 7% forage supply (mass of forage per mass of animal). At each pasture cutting, animals were weighed to assess weight gain. 2.5. Greenhouse gases Emissions of CO2 , CH4 , and N2 O from soil were monitored biweekly during the soybean production cycle (December 2009 to April 2010) in the CS, NTS and ICLS plots. Air samples were obtained using rectangular static chambers with an area of 0.32 m2 and a volume of 96 L. The base of each air-sampling chamber, with a water-isolation space at the upper exterior edge, was inserted approximately 5 cm into the soil and served as a support for the top, which was coupled at the time of air sampling. Bases remained installed in the field throughout the assessment period. The top of each chamber was equipped with a thermometer and a manual valve to couple the syringes during air sampling. Internal 12-V fans were activated 30 s before each sampling to homogenize the air inside the chambers. Three sampling chambers were used in each treatment. Air was sampled at 9:00 am. The air within the chamber was sampled 0, 15, 30, and 45 min after the chamber was closed using 20-mL polypropylene syringes equipped with multi-position valves. Air samples were analyzed using gas chromatography no more than 36 h after collection. Gas fluxes from the soil were calculated using the equation F(NN2 O or C-CH4 ) = d[N2 O or CH4 ]/dt × M × P × V/R × T, where d[N2 O or CH4 ]/dt is the rate of change of the quantity of gas within the chamber over time; M is the molecular weight of the gas; V and T are the volume (L) and internal temperature (◦ K) of the chamber, respectively; P is the pressure inside the chamber, assumed to be 1 atm; and R is the universal ideal gas constant (0.0821 L atm K−1 mol−1 ). Variation in the gas concentration over time was used to calculate the slope or rate of change. Average daily N2 O and CH4 emissions were assumed from the average value of the three chambers installed in each treatment. Cumulative N2 O and CH4 emissions were obtained by integrating the area under each emission by time curve. Cumulative emissions were then subjected to analysis of variance, and differences in mean values were analyzed using Tukey’s test at a 5%. Net CO2 stored in the soil was calculated with respect to the 15-year average amount of carbon stored in the soil of the CS treatment (Zanatta and Salton, 2010). 2.6. Soil-quality indexes 2.6.1. Stratification ratio Stratification ratio (SR) of soil carbon was calculated as the ratio between the measured C content of the surface layer (0–5 cm) and that of the lower layer (5–15 cm), according to Franzluebbers (2002). 2.6.2. Carbon-management index Carbon-management index (CMI) was calculated according to Diekow et al. (2005) using the physical fractionation of SOM as an

73

alternative to the original method, which used KMnO4 to oxidize SOM. Values for TOC and its fractions were used to calculate organic carbon (OC) storage and SOM lability (ILab) sub-indexes relative to a native vegetation (NV), which was used as the reference condition. The CMI was calculated as the product of ILab and OC. 2.6.3. Ordination level The ordination level (OL) for each management system was given by the product of the C stocks and the average stableaggregate size (MWD) in the soil relative to the reference condition (NV) (Vezzani, 2001). 2.7. Multivariate analysis Principal component analysis (PCA) was applied to the data sets (except the soil-quality indexes) to reduce the dimensionality of the data sets and generate interpretable axes (Legendre and Legendre, 1998) for the combined assessment of all management systems. We conducted two PCA: one summarized data of physico-chemical soil attributes, greenhouse gases and beef and grain production, and another summarized data of biological soil attributes and soil carbon and organic matter fractions. Principal component analysis was performed using the Vegan package (Oksanen et al., 2006) in the R (R Development Core and Team, 2012) software platform. 3. Results and discussion Tables 1–6 present the primary results observed. Where available, corresponding data for native vegetation are shown as a reference. 3.1. Soil physical and chemical attributes 3.1.1. Chemical attributes During the 17 years of experimental management, including different fertilization and planting methods used in the grain-producing systems, soil chemical attributes were affected. Soil-fertility data [Ca + Mg, K, P, cation-exchange capacity (CEC), base saturation and aluminum saturation] did not display the same tendency to increase or decrease in concentration among treatments (Table 1). Decomposition of crop and animal residues, solubilization of mineral fertilizers, and quantity of crop residues following harvest and following grazing were the major factors leading to soil acidity (Flores et al., 2008). In conservation-oriented approaches, employing no-till cropping and pasture rotations, organic matter usually increased with time (Table 2). These changes in SOM can improve the efficiency of P-based fertilization and reduce the negative effects of soil acidity and aluminum toxicity (Carvalho et al., 2010). Such effects were observed in the ICLS treatment: although aluminum saturation was greater in ICLS (19.28 ± 1.37) than in CS (12.40 ± 1.98) or NTS (11.76 ± 1.09) and P levels were lower in ICLS (19.85 ± 1.06) than in other systems (CS: 25.17 ± 0.83 and NTS: 57.67 ± 2.99), the ICLS treatment yielded soybean production that was greater than or equal to that of CS and NTS (Table 4). This higher efficiency of ICLS system may be related to availability of P in organic form (Table 1). PP showed aluminum-saturation approximately two to three times lower than grain-producing systems. This effect may be related to reduced basic cations exportation via beef production, and absence of legume plants and nitrogen fertilizer that contribute to soil acidification (Vieira et al., 2008). The Ca + Mg contents, CEC, and base saturation were lower in CS than in NTS and ICLS (Table 1). This effect may have been related to increased nutrient deposition in the surface layer in no-till systems; in CS, tillage may have distributed these nutrients to greater depths. Moreover, the CEC of soils with high levels of iron oxide depends

74

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

Table 1 Physical and chemical soil attributes from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) in Dourados experiment. Attribute

Systems CS

Aggregation MWD (mm)

2.19 ± 0.085b

Aggregate stability index

0.72 ± 0.021

CEC, 0–5 cm, 2011a (cmolc dm−3 ) Basis saturation on CEC, 0–5 cm, 2011 (%) Al saturation on CEC, 15–30 cm, 2011 (%) P Mehlich-1, 0–5 cm, 2011 (mg dm−3 ) K, 0–5 cm, 2011 (cmolc dm−3 ) Ca + Mg, 0–5 cm, 2011 (cmolc dm−3 ) pH H2 O, 5–15 cm, 2011

12.82 ± 0.18 51.24 ± 1.93 12.40 ± 1.98 25.17 ± 0.83 0.84 ± 0.05 5.73 ± 0.27 5.57 ± 0.08

P organic, 0–5 cm, 2009 (mg kg−1 ) Relation P organic/P total, 5–15 cm, 2009

26.4 ± 5.03 0.55 ± 0.06

a b c

Ref. NTS

ICLS

3.18 ± 0.056

4.12 ± 0.064

0.77 ± 0.011 15.07 64.21 11.76 57.67 1.14 8.41 5.45

± ± ± ± ± ± ±

PP

0.91 ± 0.010

0.26 1.26 1.09 2.99 0.04 0.18 0.04

14.72 60.56 19.28 19.85 0.78 8.07 5.50

30.5 ± 7.67 0.63 ± 0.09

± ± ± ± ± ± ±

0.27 1.23 1.37 1.06 0.04 0.21 0.04

4.93 ± 0.056

Salton et al. (2010)

0.97 ± 0.008

Salton et al. (2008)

14.35 67.30 5.04 8.37 0.85 8.78 5.71

31.3 ± 2.05 0.76 ± 0.08

± ± ± ± ± ± ±

0.22 1.11 0.67 0.70 0.05 0.36 0.03

43.1 ± 2.66 0.87 ± 0.01

Salton et al.c

Venâncio et al. (2009)

Years indicate the data of soil sampling. Standard error deviation, n = 26, 82, 87 and 47, respectively for CS, NTS, ICLS and PP. Not published.

strongly on the quantity of SOM. Thus, the lower SOM found in CS explains, at least in part, the lower CEC values observed in this treatment. 3.1.2. Physical attributes Table 1 shows the average size of water-stable aggregates, expressed as MWD for each management system. Systems with livestock grazing (ICLS and PP) had significantly greater MWD (ICLS: 4.12 mm and PP: 4.93 mm) compared to other systems (CS: 2.19 mm and NTS: 3.18 mm). Likewise, aggregate stability index was greater in the systems planted with B. decumbens. Larger and more stable aggregates are the result of increased biological activity, including growth of roots and fungal hyphae, and presence of residues from plants, insects, and other organisms. Under these conditions, more complex and variable structures are formed, including macro-aggregates. Land use management can alter incorporation of energy and carbon into the soil via plant biomass, including both above- and below-ground cultivated plants. Therefore, the soil exhibits different degrees of organization, forming aggregates that vary in shape, size, and stability (Six et al., 2004). Conversely, reduced flows of energy and carbon to the soil lead

to lower levels of organization, with simpler soil structure and predomination of micro-aggregates. Ultimately, soils with greater aggregation characteristics are considered of better quality than similar soils with weaker aggregation (Vezzani and Mielniczuk, 2009), mainly because TOC becomes physically protected in stable aggregates. 3.1.3. Soil organic matter content and quality Concentration of TOC, TOC storage, and POC stock were increased under grazing by livestock at the following order: CS < NTS < ICLS < PP, indicating that pasture contributed the most to the increase in SOM. The ICLS and PP showed the highest levels of carbon stocks in the layer of 0–20 cm, including against the reference condition (NV). Moreover, the portion of more labile C represented by POC was greater in systems with livestock, averaging 11% (ICLS) and 13% (PP) of TOC compared to 8% in CS. The POC fraction is usually reported as a highly sensitive indicator of changes due to and use and management practices (Cambardella and Elliott, 1992; Salvo et al., 2010; Salton et al., 2011). When considering only TOC, the stock in CS was similar to NV; however, the POC stock was lower in CS, similar in

Table 2 Quantity and quality of soil organic matter from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) and the native vegetation (NV) in Dourados experiment. Attribute

Systems

Ref.

CS

NTS

ICLS

PP

TOC, 0–5 cm, 2011 (g kg−1 )

15.89 ± 0.36a

19.01 ± 0.18

22.49 ± 0.59

27.06 ± 0.47

TOC, 0–5 cm, 2004 (g kg−1 ) TOC stocks, 0–20 cm, 2004 (Mg ha−1 ) POC stocks, 0–20 cm, 2004 (Mg ha−1 ) SOM lability, 0–20 cm (%)

18.94 ± 0.79 44.10 ± 2.80

19.01 ± 0.71 42.60 ± 1.93

23.04 ± 1.06 48.02 ± 1.14

26.75 ± 2.95 50.11 ± 3.73

25.83 ± 1.92 44.49 ± 2.42

3.60 ± 0.29

4.13 ± 0.46

5.20 ± 0.67

6.70 ± 1.20

4.95 ± 0.31

9.22 ± 0.32

C in the light free organic matter fraction, 0–5 cm (g kg−1 ) C in the light occluded organic matter fraction, 0–5 cm (g kg−1 ) C in the heavy organic matter fraction (g kg−1 ), 0–5 cm C O-alquil, 0–5 cm (%) Carbon retention rate after 10 years (1995–2004), 0–20 cm (Mg ha−1 year−1 ) a b c

NV –

10.76 ± 1.38

12.12 ± 1.55

15.60 ± 2.18

10.47 ± 0.39

–

1.15 ± 0.2b

1.28 ± 0.2

2.52 ± 0.6

2.70 ± 0.4

–

3.89 ± 0.1

6.04 ± 0.5

7.53 ± 0.4

3.60 ± 0.6

–

13.19 ± 0.6

17.16 ± 0.1

20.42 ± 1.6

19.53 ± 2.4

–

46.7 ± 0.2

50.2 ± 0.4

52.0 ± 1.1

−0.002

−0.169

Standard error deviation, n = 26, 82, 87 and 47, respectively for CS, NTS, ICLS and PP. Standard error deviation, n = 3. Not published.

0.440

0.909

Salton et al.c Adapted from Salton et al. (2011)

Boeni (2007)

– –

Salton et al. (2011)

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

75

Table 3 Soil biological attributes from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) and the native vegetation (NV) in Dourados experiment. Attribute

Systems

Ref.

CS

NTS

ICLS

PP

NV

Carbon on soil microbial biomass MBC (␮g C g−1 dry soil) Basal respiration (␮g C-CO2 g−1 soil day−1 ) Metabolic quotient – qCO2 (␮g C-CO2 ␮g−1 MBC h−1 ) Microbial quotient – qMIC – MBC TOC−1 (%)

320.4b† 14.1b 17.3a 2.0a

421.0ab 22.4ab 24.7a 2.1a

479.6ab 24.3a 23.5a 2.3a

542.0a 24.6a 22.7a 2.4a

965.1 53.2 22.4 2.4

Borges et al. (2009)

Density of soil invertebrate macrofauna (ind m−2 )

484 ± 106c†

913 ± 136b

1027 ± 114b

1715 ± 605b

3106 ± 388a

Silva et al. (2006)

◦

†

Macrofauna diversity (groups n )

9d

19b

15c

15c

21a

Silva et al. (2006)

Abundance of nematode Rotylenchulus reniformis (n◦ ind 300 cm−3 )

3424a‡

24b

4b

0b

0b

Sereia et al. (2007)

Area covered with weeds (%) Dry mass of the weedy community (g m−2 )

30 ± 10.4* 30 ± 1.66

25 ± 5.63 40 ± 2.29

20 ± 4.46 28 ± 1.62

5 ± 1.64 45 ± 1.30

– –

Concenc¸o et al. (2011b)

Soil covered by weed species after 20 days of soil disturbance % Total number of seedlings and plants of weed species after four soil disturbances spaced in 20 days (n◦ m−2 ) Total dry mass of weed species after four soil disturbances spaced in 20 days (g m−2 )

25 ± 1.75**

56 ± 3.75

8 ± 0.54

2 ± 0.14

–

Concenc¸o et al. (2011a)

400 ± 24.0

510 ± 35.2

120 ± 9.96

50 ± 3.80

–

18 ± 1.29

38 ± 2.85

5 ± 0.39

3 ± 0.23

–

† ‡ * **

Followed by the same letter do not differ by Tukey test (P < 0.05). Followed by the same letter do not differ by Duncan test (P < 0.01). Standard error deviation, n = 4. Standard error deviation, n = 5.

Table 4 Grain and beef productivity from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) in Dourados experiment. Attribute

Systems

Ref.

CS

NTS

ICLS

PP

Soybean yield in years with good rainfall distribution – 2006/07 (kg ha−1 ) Soybean yield in years with poor rainfall distribution – 2010/11 (kg ha−1 )

2984 ± 79.4a

3544 ± 49.0

3075 ± 43.4

–

1642 ± 86.8

2882 ± 46.8

2866 ± 70.3

–

Grazing from September–May/1999 – gain animal (kg an−1 day−1 ) Grazing from September–May/1999 – gain area (kg ha−1 )

– –

– –

0.814b 582.0

0.749 515.6

Machado et al. (2001)

Grazing from May to August/2011, only in pasture with severe frost – gain/animal (kg an−1 day−1 )

–

–

−0.047 ± 0.057c

−0.273 ± 0.005

Retore et al.d

a b c d

Salton et al.d

Standard error deviation, n = 26, 82, 87 and 47, respectively for CS, NTS, ICLS and PP. Medium values. Standard error deviation, n = 4. Not published.

NTS and greater in ICLS and PP systems compared to NV. From 2004 to 2011, TOC declined in the surface layer of soil in CS, but did not change significantly in other treatments (Table 2). Both soil preparation and smaller input of C from crop biomass residues probably contributed to POC loss, soil degradation and decrease of productive capacity. Separation of SOM by density for light free and light occluded fractions showed the same tendency as for POC, with greatest value when pastures were included in the system. The free light fraction

was greater in PP system, probably due to the continuous input of grass roots. This is a dynamic fraction, because it depends mostly on quality and quantity of organic residues (Boeni, 2007). Carbon concentration in the occluded organic fraction was 1.3 (NV) to 4.7 (NTS) times greater than in the free light fraction. Crop rotation and no till as in NTS maintained a level of occluded organic matter similar to the native condition (NV), while livestock grazing in ICLS and PP resulted in even greater values for this fraction. This denotes the possible effect of root systems to form stable aggregates (Table 1),

Table 5 Nitrous oxide and methane emissions, soil C sequestration, and net greenhouse gases emission in CO2 equivalence from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, in Dourados experiment. Attribute

Systems

Ref.

CS

NTS

ICLS

N2 O (kg N2 O-N ha−1 ) CH4 (kg CH4 -C ha−1 )

0.568a* −0.590a

0.523a −0.414a

0.486b −0.180b

Zanatta et al. (2011)

Soil C sequestration (kg C ha−1 ) Net emission (kg CO2 -Ceq ha−1 )

– 67

15† 78

−455 −395

Zanatta and Salton (2010)

† *

Relative to CS. Followed by the same letter do not differ by Tukey test (P < 0.05).

76

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

Table 6 Soil quality index from cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) and the native vegetation (NV) in Dourados experiment. Attribute

SR – stratification ratio CMI – carbon management index OL – ordination level

Systems

Ref.

CS

NTS

ICLS

PP

NV

1.13d* 71§

1.33cd 82

1.64bc 104

2.00a 136

1.70ab 100

85§

85

124

135

100

Salton (2009)

*

Values followed by the same letter do not differ by Tukey test (P < 0.05). § Indices obtained for the layer 0–20 cm from mean values f soil properties (TOC, POC, MWD).

which provide physical protection for the organic matter inside aggregates; this interaction results from the rotation of crop and pasture components in a system. The heavy organic fraction, recognized as the most stable one mainly due to strong protection mechanisms as clay interaction, represented the major portion of soil carbon. Only the PP system had values similar to those in NV; more significant losses occurred in NTS. The labile fraction of organic matter, expressed through SOM lability and C O alkyl (carbohydrates) values (Table 2) was also greater in ICLS and PP, than in NTS indicating greater energy flux in the soil system. Greater SOM lability was attributed to the presence of the forage, which adds greater amount of organic matter to the soil than cropping alone. Moreover, there is a continuous exudation of substances from grass roots to the soil during growth, which is stimulated by grazing (Carvalho et al., 2010). This exudate is composed mostly of polysaccharides, a C O alkyl source (Kögel-Knabner, 2002). Distinct carbon retention rates were observed compared to the native condition (NV); there was no change or a negative effect in CS and NTS, and a positive carbon retention rate in ICLS and PP (Table 2).

3.2. Soil biological attributes 3.2.1. Soil microbiology Soil microbial-biomass carbon (MBC) was 320.4–542.0 ␮g C g−1 dry soil among systems at the end of 12 years of management. The MBC value of NV was greater than of the other systems (Table 3). Soil conditions in NV are more favorable for microorganisms than those in cultivated areas for several reasons, including the absence of tillage, greater plant diversity, maintenance of fungal hyphae, and accumulation of leaf litter on the soil surface (Matsuoka et al., 2003). According to Baretta et al. (2005), the Cerrado biome tends to have lower MBC levels compared to denser vegetation types in similar regions. Within agricultural management systems MBC of PP was greater than CS and similar to ICLS and NTS (Table 3). Basal respiration (C-CO2 ) represents soil biological activity and is directly related to the availability of microbial biomass. The C-CO2 value of NV was greater than of other systems (PP > ICLS > NTS > CS), as shown in Table 3. These findings corroborate observations of other studies, all of which found that more-conservative systems yield better conditions for the development of soil microorganisms (Roscoe et al., 2006; Balota et al., 2003). Among the agricultural management systems, ICLS and PP had greater C CO2 than for CS, while NTS was intermediate (Table 3). Elevated metabolic quotient (qCO2 ) would imply greater carbon loss. Elevated microbial quotient (qMIC) would imply accumulation of soil carbon over time. These indexes did not differ significantly among management systems (Table 3).

3.2.2. Invertebrate soil macrofauna Density and taxonomic richness of the invertebrate macrofauna community in soil differed among management systems. Lowest values were observed in CS (Table 3), while ICLS was equivalent to that of NTS and PP and greater than that of CS. These findings confirm the high sensitivity of the soil invertebrate macrofauna to anthropogenic disturbance. These results may be related to the absence of soil tillage, greater deposition of plant residues, and intense renovation of root mass in pastures, which result in more root exudation, less variation in soil temperature and water content, and increased SOM levels, thus promoting a favorable edaphic–climatic environment (Sánchez and Crespo, 2004; Silva et al., 2008). Diversity of soil macroinvertebrates was favored not only by environmental conditions found in ICLS and NTS, but also by the diversity of plant species in these production systems. Soil macroinvertebrates perform numerous essential functions, including decomposition, nutrient cycling, SOM mineralization, soil-structure modification, atmospheric-composition regulation, and biological control of pests and diseases (Sánchez and Reyes, 2003). 3.2.3. Weed dynamics Weed community analysis showed that areas without pasture and grazing generally accumulated more weed mass than areas that were periodically or continuously grazed. The area of soil covered by weeds was 87% greater in CS compared to the average of other treatments (Table 3). Generally, areas that were continuously or periodically grazed by livestock had fewer weeds than areas where only grain crops were grown. Livestock grazing also affected seed germination; weed seedlings from treatments that included grazing took longer to germinate and emerge from the soil surface. Regardless of the presence or absence of tillage, crop-only systems exhibited larger areas of soil covered by weeds. Weed species composition was highly similar throughout the soil profile in crop-only treatments but diverged among sampled depths in systems with livestock (Table 3). Conservation-oriented land-use strategies help to reduce the severity of weed infestations in cultivated fields. Under ICLS, it may be necessary to control the occurrence of forage-plant seedlings in successive grain crops by avoiding the use of forage species with high tolerance to herbicides such as glyphosate. 3.2.4. Nematodes Occurrence of Rotylenchuls reniformis differed significantly among management systems, with a much larger population in CS (3424 individuals per 300 cm3 ) than in other systems (24 nematodes per 300 cm3 maximum) (Table 3). This finding suggests that more diversified systems limit the proliferation of this nematode. The small populations found in NTS and ICLS indicate that these systems did not provide favorable conditions for this nematode, perhaps due to the use of non-host crops in rotation with soybeans, such as B. decumbens in ICLS and corn in NTS. 3.3. Grain and beef production More complex and diversified systems can be expected to display emergent properties. Such properties appear only when sufficient energy and matter accumulate in the soil to produce a level of organization. Crop yield under adverse climatic conditions is usually a good indicator of these properties (Vezzani and Mielniczuk, 2009). This hypothesis was verified by production responses of soybean and beef during the experiment. In years with ample rainfall, soybean production was equal in the three grain-production systems (Table 4). In years with poor rainfall distribution, with water deficits, ICLS and NTS exhibited

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

smaller productivity losses compared to CS. In 2010/11, for example, CS yielded only 60% of crop production in other systems. The two systems that incorporated livestock differed in beef production, which was measured during the rainy season (September–May). ICLS performed better than PP, with gains of 8.7% in the productivity per animal and 12.9% in the productivity per area. During winter when the climate is dry and many frosts occurred such as in 2011, both systems had depressed weight gain. Weight gain was 5.8 times greater in PP than in ICLS during this extreme winter event (Table 4). These findings confirm the hypothesis that increased complexity leads to greater resilience of agricultural production systems. This conclusion is important for agriculture and livestock farming in tropical regions, where the climate is frequently harsh and erratic. 3.4. Greenhouse gases Table 5 shows GHG emissions during the soybean growing season of 2010/11. Cumulative N2 O flux was greater in CS and NTS than in ICLS. Conversely, greater CH4 absorption was observed in CS and NTS than in ICLS. During the sampling period, net emissions were positive in CS and NTS and negative in ICLS, primarily due to differences in soil sequestration that offset N2 O emissions. Cumulative seasonal N2 O emissions were equivalent to 265 kg CO2 ha−1 in CS and 227 kg CO2 ha−1 in ICLS. Reduced emission in ICLS was related to reduced intensity of N2 O emissions during the crop-maturation period due to later aged-leaf deposition and rainfall. Similar results were reported by Rochette et al. (2004), who attributed enhanced N2 O production to stimulated N mineralization following deposition of leaves. Leaf fall in our study coincided with a rainfall period, possibly promoting conditions for N2 O production. Soil in ICLS displayed a significantly smaller CH4 -absorption capacity (−6 kg CO2 -Ceq ha−1 ) than soil in NTS (−13 kg CO2 -Ceq ha−1 ) and CS (−19 kg CO2 -Ceq ha−1 ). N2 O emissions were only partially counterbalanced by CH4 absorption because the radiative forcing potential of CH4 (25) is about 12 times smaller than that of N2 O (298). Only 2–7% of the emitted N2 O was counterbalanced by CH4 absorption. This value is relatively low but compatible with values reported in the literature (Galbally et al., 2008; Chen et al., 2011; Schoenbach et al., 2012). The ICLS approach had reduced capacity for CH4 absorption, as also observed previously by Dong et al. (2000), Liu et al. (2007) and Schoenbach et al. (2012). CH4 absorption in ICLS was 68% lower than in other treatments, exceeding the reduction found by Chen et al. (2011) for temperate plains (30%). The negative impact of ICLS on CH4 absorption may have been due to increased N cycling in the system, in which the presence of animals during the pasture cycle, may have affected the abundance and activity of methanotrophs (Zhou et al., 2008) and possibly decreased air diffusion that could have impaired CH4 diffusion (Chen et al., 2011). Low cumulative N2 O emissions and high C-sequestration rates in ICLS resulted in enhanced GHG mitigation potential compared to CS and NTS. The overall balance of GHG emissions was mostly determined by each system’s potential to sequester carbon in the soil, similar to that reported by Schoenbach et al. (2012) for the temperate region and Piva (2010) for southern Brazil. Piva (2010) reported positive net emissions in ICLS and NTS. These differences have two possible explanations: (1) greater potential net primary production of Brachiaria pastures compared to ryegrass that is directly reflected in C-sequestration rates and (2) longer duration of the experiment (15 vs. 3.5 years), which is reflected in improved physical properties of the soil. Piva (2010) reported low C-sequestration rates for the 0–100-cm soil layer (30 kg C ha−1 y−1 ). Although ICLS may increase gas emissions during the crop cycle, the pasture phase of this system can help mitigate overall gas emissions through soil C sequestration. Compression of soil caused by animals, increased

77

N cycling, and particularly emission of enteric CH4 during the livestock phase are of concern, and these factors require further investigation. Currently, much attention is given to the efficiency of production systems in balancing soil C and reducing GHG emissions. The ICLS treatment was highly efficient in these respects, accumulating more soil C and reducing CO2 and N2 O emissions while maintaining productivity. 3.5. Soil-quality indexes Among the various indicators of soil quality, those related to the accumulation and storage of SOM are the most useful, because SOM is sensitive to management system and correlates with many soil physical, chemical, and biological attributes. Soil-carbon stratification ratio (SR) is based on the fact that management-related changes in soil attributes occur primarily in the surface layer. The soil–atmosphere interface plays a fundamental role in the functioning of production systems, controlling and partitioning water and gas flows between the soil and atmosphere. Values of SR greater than one indicate accumulation of TOC in the surface layer, and larger values correspond to better soil quality. Values less than one indicate loss of soil quality. In addition, soil CMI summarizes the effects of soil-management methods on the quantity and quality of SOM. The ordination level (OL) is based on the flow of energy and matter into the soil, the accumulation of which results in the formation of stable macro-aggregates, representing the level of organization of soil. Soil-quality indexes shown in Table 6 confirm the various individual soil properties described earlier, indicating that soil quality increases in the order CS < NTS < ICLS < PP. Soil tillage decreases and system complexity increases in the same order, with fodder grass and livestock present in ICLS and PP, resulting in greater synergy between the components of the soil system. The NTS approach did not maintain soil quality relative to the reference condition (NV) in spite of the use of corn as a summercrop rotation every three years and application of more fertilizers in NTS than in ICLS. The OL, which accounts for both carbon content and soil structure, was equal in CS and NTS. This result may be due to tillage for sowing, which may affect soil structure even though it is less intensive than conventional tillage. Conversely, the presence of perennial pasture grass over long periods in ICLS promoted the recovery, improvement, and maintenance of soil quality. In this system, pasture grass contributed to greater accumulation of plant biomass from both the aerial portions and the roots and to the formation of aggregates through the physical activity of roots. 3.6. Multivariate analysis To synthesize results and help to understand the effects of management systems, a principal component analysis (PCA) was performed using the variables and derived indexes shown in Tables 1–6. The first and second PCA axes accounted for more than 70% of data variability. The different crop systems influenced the distribution of soil attributes and production, as indicated by the different groups formed (Fig. 2A and B). The first PCA axis separated the systems into (a) grain production alone (CS and NTS) and (b) with livestock (ICLS and PP) (Fig. 2A). When data from NV was included (Fig. 2B) separation of management systems followed the same trend, but ICLS and PP were grouped with NV. Both PCA plots showed that many of the desirable effects, such as improved soil properties, lower GHG emission, weeds and nematodes control, tended to be more expressive in ICLS and PP. These results provide clear evidence that cropping systems that include livestock are more environmentally sustainable. The economic aspect of these cropping systems, however, needs to be further evaluated.

78

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79

Fig. 2. Principal component biplot of soil attributes and grain and beef production in the cropping systems (CS: conventional, NTS: no-till, ICLS: integrated crop-livestock, PP: pasture) and the native vegetation (NV) in Dourados experiment. Plot A – N2 O: N2 O (kg N2 O-N ha−1 ); SGR: soybean yield in years with good rainfall distribution – 2006/07 (kg ha−1 ); CH4 : CH4 (kg CH4 -C ha−1 ); NE: net emission (kg CO2 -Ceq ha−1 ); SBR: soybean yield in years with bad rainfall distribution – 2010/11 (kg ha−1 ); SCS: soil C sequestration (kg C ha−1 ); CEC: CEC, 0–5 cm, 2011 (cmolc dm−3 ); BS: basis saturation on CEC, 0–0.05 m, 2011 (%); Porg: P organic, 0–5 cm, 2009 (mg kg−1 ); Porg/Ptot: relation P organic/P total, 5–15 cm, 2009; MWD: aggregation (mm); beef winter: Grazing from May to August/2011, only in pasture with severe frost – gain/animal (kg an−1 day−1 ); beef summer: grazing from September to May/1999 – gain animal (kg an−1 day−1 ); beef/area: grazing from September to May/1999 – gain area (kg ha−1 ); ASI: aggregate stability index. Plot B – NEM: abundance of nematode Rotylenchulus reniformis (n◦ ind 300 cm−3 ); TNWP: total number of seedlings and plants of weed species after four soil disturbances spaced in 20 days (n◦ m−2 ); SCWS: Soil covered by weed species after 20 days of soil disturbance %; WWP: total dry mass of weed species after four soil disturbances spaced in 20 days (g m−2 ); qMIC: microbial quotient – MBC TOC−1 (%); C-CO2 : basal respiration (␮g C-CO2 g−1 soil day−1 ); MDi: macrofauna diversity (groups n◦ ); C-BMS: carbon on soil microbial biomass (␮g C g−1 ); DIM: density of soil invertebrate macrofauna (ind m−2 ); CLF: C in the light free organic matter fraction (g kg−1 ); qCO2 : metabolic quotient – qCO2 (␮g C-CO2 ␮g−1 MBC h−1 ); CHOM: C in the heavy organic matter fraction (g kg−1 ); TOC: TOC, 0–5 cm, 2011 (g kg−1 ); POC: POC, 0–20 cm (Mg ha−1 ); CLOM: C in the light occluded organic matter fraction (g kg−1 ); L: lability (%): TOC-S: TOC, stocks, 0–20 cm (Mg ha−1 ); CSEq: carbon retention rate after 10 years, 0–30 cm (Mg ha−1 year−1 ).

As case in point PP had lower production than ICLS, and full economic analyses are needed to understand implications.

4. Conclusions Currently, farmers and agricultural scientists are greatly concerned with the environmental impact of agriculture in addition to its productivity and sustainability. In this regard, although NTS and ICLS approaches exhibit similar productivity, the latter system is superior because it emits less GHG per production unit. Moreover, ICLS also performs better in terms of beef production due to the positive effects of rotation between crop and livestock production. Therefore, the findings of this study confirm the initial hypothesis and indicate that ICLS is an appropriate management system in the study area. This system performs better than the other systems studied in terms of environmental quality and productivity, guaranteeing the sustainability of agricultural activity. The ICLS design evaluated in the Dourados experiment (“Dourados System”) is recommended for implementation in tropical Brazil, because results of this experiment indicate its agronomical and environmental efficiency.

Acknowledgements This work was partially funded by Project iLPF/Bunge, CNPq (National Council for Scientific and Technological Development), Fundac¸ão Agrisus and Project Probio II/Embrapa.

References Anderson, J.M., Ingram, J. (Eds.), 1993. Tropical Soil Biology and Fertility. A Handbook of Methods. , second ed. CAB, Oxford, p. 221. 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. Anghinoni, I., 2007. Fertilidade do solo e seu manejo no sistema plantio direto. In: Novais, R.F., Alvarez, V.H., Barros, N.F., Fontes, R.L.F., Cantarutti, R.B., Neves, J.C.L. (Eds.), Fertilidade do Solo. Sociedade Brasileira de Ciência do Solo, Vic¸osa, pp. 873–928. Assmann, T.S., et al., 2003. Rendimento de milho em área de integrac¸ão lavourapecuária sob o sistema plantio direto em presenc¸a e ausência de trevo branco, pastejo e nitrogênio. Rev. Bras. Cienc. Solo 27, 675–683. Balbinot, A.A., de Moraes, A., da Veiga, M., Pelissari, A., Dieckow, J., 2009. Crop-livestock system: intensified use of agricultural lands. Cienc. Rural 39, 1925–1933. Balota, E.L., et al., 2003. Microbial biomass in soils under different tillage and crop rotation systems. Biol. Fertil. Soils 38, 15–20. Baretta, D., Santos, J.C.P., Figueiredo, S.R., Klauberg-Filho, O., 2005. Efeito do monocultivo de pinus e da queima do campo nativo em atributos biológicos do solo no Planalto Sul Catarinense. Rev. Bras. Ciênc. Solo 29, 715–724. Bignell, D.E., 2009. Towards a universal sampling protocol for soil biotas in the humid tropics. Pesq. Agropec. Bras. 44, 825–834. Boeni, M., 2007. Protec¸ão física da matéria orgânica em latossolos sob sistemas com pastagens na região do Cerrado Brasileiro, 136p. Tese (Doutorado) – Faculdade de Agronomia, Universidad Federal do Rio Grande do Sul, Porto Alegre, RS. Borges, C.D., Mercante, F.M., Silva, R.F., Salton, J.C., 2009. Impacto de sistemas agropecuários na qualidade do solo. In: 22. Congresso Brasileiro de Ciência do Solo. O solo e a produc¸ão de bioenergia: perspectivas e desafios. SBCS, Fortaleza, CD-ROM. Cambardella, C.A., Elliott, E.T., 1992. Particulate soil organic-matter changes across a grassland cultivation sequence. Soil Sci. Soc. Am. J. 56, 777–783. Carvalho, P.C.F., Anghinoni, I., Moraes, A., Souza, E.D., Sulc, R.M., Lang, C.R., Flores, J.P.C., Terra Lopes, M.L., Silva, J.L.S., Conte, O., Lima Wesp, C., Levien, R., Fontaneli, R.S., Bayer, C., 2010. Managing grazing animals to achieve nutrient cycling and

J.C. Salton et al. / Agriculture, Ecosystems and Environment 190 (2014) 70–79 soil improvement in no-till integrated systems. Nutr. Cycl. Agroecosyst. 88, 259–273. Chen, W., Wolf, B., Zheng, X., Yao, Z., Butterbach-Bahl, K., Brueggemann, N., Liu, C., Han, S., Han, X., 2011. Annual methane uptake by temperate semiarid steppes as regulated by stocking rates, aboveground plant biomass and topsoil air permeability. Global Change Biol. 17, 2803–2816. Concenc¸o, G., Salton, J.C., Brevilieri, R.C., Mendes, P.B., Secretti, M.L., 2011a. Soil seed bank of plant species as a function of long-term soil management and sampled depth. Planta Daninha 29, 725–736. Concenc¸o, G., Salton, J.C., Secretti, M.L., Mendes, P.B., Brevilieri, R.C., Galon, L., 2011b. Effect of long-term agricultural management systems on occurrence and composition of weed species. Planta Daninha 29, 515–522. Diekow, J., Mielniczuk, J., Knicker, H., Bayer, C., Dick, D.P., Kogel-Knabner, I., 2005. Carbon and nitrogen stocks in physical fractions of a subtropical Acrisol as influenced by long-term no-till cropping systems and N fertilization. Plant Soil 268, 319–328. Dong, Y.S., Zhang, S., Qi, Y.C., Chen, Z.Z., Geng, Y.B., 2000. Fluxes of CO2 N2 O and CH4 from a typical temperate grassland in Inner Mongolia and its daily variation. Chin. Sci. Bull. 45, 1590–1594. Flores, J.P.C., Cassol, L.C., Anghinoni, I., Carvalho, P.C.F., 2008. Atributos químicos do solo em func¸ão da aplicac¸ão superficial de calcário em sistema de integrac¸ão lavoura-pecuária submetido a pressões de pastejo em plantio direto. Rev. Bras. Ciênc. Solo 32, 2385–2396. Franchini, J.C., Costa, J.M.d., Debiasi, H., Torres, E., 2011. Importância da rotac¸ão de culturas para a produc¸ão agrícola sustentável no Paraná. Embrapa Soja, Londrina, pp. 52. Franzluebbers, A.J., 2002. Soil organic matter stratification ratio as an indicator of soil quality. Soil Till. Res. 66, 95–106. Galbally, I.E., et al., 2008. Soil–atmosphere trace gas exchange in semiarid and arid zones. J. Environ. Qual. 37, 599–607. Jenkins, W.R., 1964. A rapid centrifugal-flotation technique for separating nematodes from soil. Plant Dis. Rep. 48, 692–695. Kluthcouski, J., Stone, L.F., 2003. Desempenho de culturas anuais sobre palhada de braquiária. In: Kluthcouski, J., Stone, L.F., Aidar, H. (Eds.), Integrac¸ão lavourapecuária. Embrapa Arroz e Feijão, Santo Antônio de Goiás, pp. 499–522. Kögel-Knabner, I., 2002. The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil Biol. Biochem. 34, 139–162. Lazzarotto, J.J., dos Santos, M.L., de Lima, J.E., Moraes, A., 2009. Volatilidade dos retornos econômicos associados à integrac¸ão lavoura-pecuária no Estado do Paraná. Rev. Econ. Agron. 7, 259–283. Legendre, P., Legendre, L., 1998. Numerical Ecology. Elsevier Science, Amsterdam. Liu, C.Y., Holst, J., Bruggemann, N., Butterbach-Bahl, K., Yao, Z.S., Yue, J., Han, S.H., Han, X., Krummelbein, J., Horn, R., Zheng, X.H., 2007. Winter-grazing reduces methane uptake by soils of a typical semi-arid steppe in Inner Mongolia. Chin. Atmos. Environ. 41, 5948–5958. Machado, L.A.Z., Fabrício, A.C., Salton, J.C., 2001. Performance of steers in Brachiaria decumbens pastures, permanent and in rotation with soybean. In: 19. International Grassland Congress. Grassland Ecosystems: an Outlook into the 21st Century: Proceedings, FEALQ, Piracicaba, pp. 747–748. Martha Junior, G.B., Alves, E., Contini, E., 2011. Dimensão econômica de sistemas de integrac¸ão lavoura-pecuária. Pesq. Agropec. Bras. 46, 1117–1126. Matsuoka, M., Mendes, I.C., Loureiro, M.F., 2003. Biomassa microbiana e a atividade enzimática em solos sob vegetac¸ão nativa e sistemas agrícolas anuais e perenes na região de Primavera do Leste (MT). Rev. Bras. Ciênc. Solo 27, 425–433. Oksanen, J., Kindt, R., Legendre, P., O’hara, B., 2006. Community ecology package: Vegan. Version 1.8.1, Available at: http://cc.oulu.fi/jarioksa∼/ (accessed 9.10.10). Piva, J.T., 2010. Emissão de óxido nitroso e metano em um Latossolo Bruno submetido a sistemas de preparo e integrac¸ão lavoura-pecuária. Federal University of Paraná, Curitiba, pp. 57. R Development Core Team, 2012. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Rochette, P., Angers, D.A., Belanger, G., Chantigny, M.H., Prevost, D., Levesque, G., 2004. Emissions of N2 O from alfalfa and soybean crops in eastern Canada. Soil Sci. Soc. Am. J. 68, 93–506. Roscoe, R., Mercante, F.M., Mendes, I.C., Reis Júnior, F.B., Santos, J.C.F., Hungria, M., 2006. Biomassa microbiana do solo: frac¸ão mais ativa da matéria orgânica. In: Roscoe, R., Salton, F.M., Mercante, J.C. (Eds.), Dinâmica da matéria orgânica do solo em sistemas conservacionistas:modelagem matemática e métodos auxiliares. Embrapa Agropecuária Oeste, Dourados, pp. 163–198. Salton, J.C., 2009. Indicadores de qualidade do solo em sistemas de integrac¸ão lavoura-pecuária. In: 5. Congresso Brasileiro de Soja, Soja: fator de desenvolvimento do Cone Sul, Embrapa Soja, Londrina, PR.

79

Salton, J.C., Mielniczuk, J., Bayer, C., Boeni, M., Conceicao, P.C., Fabrício, A.C., Macedo, M.C.M., Broch, D.L., 2008. Agregac¸ão e estabilidade de agregados do solo em sistemas agropecuários em Mato Grosso do Sul. Rev. Bras. Ciênc. Solo 32, 11–21. Salton, J.C., Mielniczuk, J., Bayer, C., Fabrício, A.C., Macedo, M.C.M., Broch, D.L., 2011. Teor e dinâmica do carbono no solo em sistemas de integrac¸ão lavoura-pecuária. Pesq. Agropec. Bras. 46, 1349–1356. Salton, J.C., Costa, A.R., Zanatta, J.A., 2010. Aggregation of an Oxisol under management systems for 14 years in the Midwest of Brazil. In: 19th World Congress of Soil Science. Soil Solutions for a Changing World: Proceedings, IUSS/ASSSI, Brisbane, pp. 13–16. Salvo, L., Hernández, J., Ernst, O., 2010. Distribution of soil organic carbon in different size fractions, under pasture and crop rotations with conventional tillage and no-till systems. Soil Till. Res. 109, 116–122. Sánchez, S., Crespo, G., 2004. Comportamiento de la macrofauna del suelo en pastizales con gramíneas puras o intercaladas con leucaena. Pastos y Forrajes 27, 347–353. Sánchez, S., Reyes, F., 2003. Estudio de la macrofauna edáfica en una asociación de Morus alba y leguminosas arbóreas. Pastos y Forrajes 26, 315–320. Saunders, W.M.H., Williams, E.G., 1955. Observation on the determination of total organic phosphorus in soils. J. Soil Sci. 6, 254–267. Schoenbach, P., Wolf, B., Dickhoefer, U., Wiesmeier, M., Chen, W., Wan, H., Gierus, M., Butterbach-Bahl, K., Koegel-Knabner, I., Susenbeth, A., Zheng, X., Taube, F., 2012. Grazing effects on the greenhouse gas balance of a temperate steppe ecosystem. Nutr. Cycl. Agroecosyst. 93, 357–371. Sereia, A.F.R., Asmus, G.L., Fabricio, A.C., 2007. Influência de diferentes sistemas de produc¸ão sobre a populac¸ão de Rotylenchuls reniformis (Linford & Oliveira, 1940) no solo. Nematol. Bras. 31, 42–45. Silva, F.C. (Org.), 1999. Manual de análises químicas de solos, plantas e fertilizantes. Embrapa Comunicac¸ão para Transferência de Tecnologia, Brasília, DF. Silva, R.F., Mercante, F.M., Otsubo, A.A., 2008. Macrofauna invertebrada do solo em sistema integrado de produc¸ão agropecuária no Cerrado. Acta Sci. Agron. 30, 725–731. Silva, R.F., Aquino, A.M., Mercante, F.M., Guimarães, M.F., 2006. Macrofauna invertebrada do solo sob diferentes sistemas de produc¸ão em Latossolo da região do cerrado. Pesq. Agropec. Bras. 41, 697–704. Silva, R.F., Guimarães, M.F., Aquino, A.M., Mercante, F.M., 2011. Análise conjunta de atributos físicos e biológicos do solo sob sistemas de integrac¸ão lavourapecuária. Pesq. Agropec. Bras. 46, 1277–1283. Six, J., Bossuyt, H., Degryze, S., Denef, K., 2004. A history of research on the link between (micro)aggregates, soil biota, and soil organic matter dynamics. Soil Till. Res. 79, 7–31. Sulc, R.M., Tracy, B.F., 2007. Integrated crop-livestock systems in the U.S. corn belt. Agron. J. 99, 335–345. Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem. 19, 703–707. Venâncio, J.B., Zanatta, J.A., Salton, J.C., 2009. Distribuic¸ão das formas de fósforo após 15 anos da adoc¸ão de sistemas de manejo. In: Matéria Orgânica Ambiental e Sustentabilidade: resumos expandidos. 8. Encontro Brasileiro de Substâncias Húmicas, Embrapa Clima Temperado, Pelotas. Vezzani, F.M., 2001. Qualidade do Sistema Solo na Produc¸ão Agrícola. 184 f. Tese (Doutorado) – Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS. Vezzani, F.M., Mielniczuk, J., 2009. Uma visão sobre qualidade do solo. Rev. Bras. Ciênc. Solo 33, 743–755. Vieira, F.C.B., Bayer, C., Mielniczuk, J., Zanatta, J.A., Bissani, C.A., 2008. Long-term acidification of a Brazilian Acrisol as affected by no till cropping systems and nitrogen fertilizer. Aust. J. Soil Res. 46, 17–26. Vilela, L., Barcellos, A.O., Sousa, D.M.G., 2008. Benefícios da integrac¸ão entre lavoura e pecuária. Embrapa Cerrados, Planaltina. Zanatta, J.A., Salton, J., 2010. Soil carbon sequestration affected by no-tillage and integrated crop-livestock systems in Midwestern Brazil. In: 19th World Congress of Soil Science. Soil solutions for a Changing World: Proceedings, IUSS/ASSSI, Brisbane. Zanatta, J.A., Salton, J.C., Bayer, C., Tomazi, M., Colman, I., 2011. Emissões de gases de efeito estufa em sistemas de manejo de solo durante ciclo de produc¸ão da soja no Cerrado. In: 2. Reunião Paranaense de Ciência do Solo. UFPR, Curitiba, PR, pp. 24. Zhou, X.Q., Wang, Y.F., Huang, X.Z., Tian, J.Q., Hao, Y.B., 2008. Effect of grazing intensities on the activity and community structure of methane-oxidizing bacteria of grassland soil in Inner Mongolia. Nutr. Cycl. Agroecosyst. 80, 145–152.