Soil K forms and K budget in integrated crop-livestock systems in subtropical paddy fields

Highlights

• Exchangeable K forms were not sensitive to detect different amounts of K added.

• K addition greater than the export resulted in the fixation of K surplus by the soil.

• Rotation of rainfed crops with paddy rice increased the fixation of K by the soil.

• Non-exchangeable K was more useful to understanding the soil K dynamics.

• K fertilization in 2:1 claymineral soils should focus on the reposition of exported K.

Abstract

Diversified farming systems is a challenge in lowland soils around the world. In Southern Brazil, integrated crop-livestock systems (ICLS) have been used for grain crops and livestock production under no-till. However, their impacts on soil K dynamics are still poorly explored. The aim of this study was to quantify the K forms in the soil and the K budget after 66 months of different flooded rice production systems under no-till ICLS in a subtropical paddy field. Four production systems were established in 2013: (i) flooded rice monocropping under conventional tillage (CT) and winter fallow (R-CT); (ii) no-tillage (NT) flooded rice monocropping with ryegrass pasture in winter (R-ICLS); (iii) NT flooded rice and soybean rotation in the summer and ryegrass pasture in winter (RS-ICLS), and (iv) NT flooded rice, soybean, maize and stapf grass rotation in the summer period and ryegrass and clover pasture in the winter (RSM-ICLS). In 2018, soil samples were taken at the 0–10, 10–20 and 20–30 cm soil layers. Exchangeable potassium (K), non-exchangeable K, structural K, and total K were evaluated. Exchangeable K was not affected by the treatments. The higher the difference between inputs and outputs of K in the systems evaluated, the higher was the content of structural and total K in the soil. However, the K budget in the system was not sufficient to fully explain the variation in the non-exchangeable K content in the soil. Comparing the systems with the same frequency of flooded rice cultivation, the difference between inputs and outputs of K was about 2.3 times higher in the R-ICLS system (830 kg ha⁻¹ of K) compared to R-CT (366 kg ha⁻¹), which increase in 70 % the concentration of non-exchangeable K in the 0–30 cm soil layer (622 and 361 mg kg⁻¹ in the R-ICLS and R-CT, respectively). In the systems where flooded rice rotates with rainfed crops (RS-ICLS and RSM-ICLS), the drying cycles resulted in K fixation with higher energy, turning the surplus of K added via fertilization into structural forms. These results indicate that lowland soils cultivated with K fertilization rates greater than the export of K can result in a fixation of K, decreasing its availability in the short-medium term. Moreover, it strongly emphasizes that adequate K fertilization in lowland soils with 2:1 clayminerals should focus on the reposition of exported K in order to prevent temporary fixation of K by fixation on structural forms.

Introduction

Rice (Oryza sativa L.) is considered one of the most important grain crops in developing countries and it is responsible for feeding over 4 billion people in the world, with the prospect of increasing rice consumption by 25 % until 2030 (International Rice Research Institute IRRI, 2019). Rice is grown in an area of 161 million hectares worldwide (USDA, 2020), mainly in monocropping systems of flooded rice under conventional tillage (CT), with annual soil mobilization by plowing and harrowing combined with fallow in the off-season (Chauhan et al., 2017).

Soil disturbance has been pointed out as the main cause of soil degradation (Lal, 2015), leading to a decline in flooded rice yield over time (Bado et al., 2010). On the other hand, although still scarcely used by farmers, no-till (NT) has proved efficient in improving soil quality and flooded rice yield (Denardin et al., 2019). In this scenario, the crop diversification with paddy-upland crops rotation is sustainable way to increase soil fertility and rice yields (Hou et al., 2018; Yu et al., 2014). Moreover, alternating the grain crops (rice, soybeans and maize) with pastures for animal grazing in integrated crop-livestock system (ICLS) (Moraes et al., 2014), has been pointed out as a management strategy to booster soil fertility and nutrient use efficiency in subtropical paddy fields (Carlos et al., 2020).

It is well known that NT (Firmano et al., 2019) and crop rotations (Tiecher et al., 2017) increase the availability of potassium (K) in the soil. Although the dynamics of K in the soil in crop production systems has been widely explored (Vieira et al., 2016; Mortele et al., 2016, 2019; Gatiboni et al., 2019), there is still a lack of knowledge about the soil K dynamics in ICLS, especially considering paddy fields under different soil tillage systems and crop rotations.

Potassium in soil can be broadly divided into three main forms: (i) in small concentrations in the soil solution as a free ion (K⁺) (Rawat et al., 2016), (ii) as part of primary minerals highly resistant to weathering (structural form), and (iii) adsorbed to soil functional groups forming surface complexes with varying degrees of energy and availability for plants (Weil and Brady, 2017). The most available soil K form to plants can be extracted with chemical solutions such as Mehlich-1 and ammonium acetate, and it is generically called "exchangeable K" (Warncke and Brown, 1998). In the medium-long term, there is a continuous process of depletion of K retained with less energy by the soil, forcing the release of K adsorbed with greater energy, often present in the mineral structure (Fraga et al., 2009; Mortele et al., 2019). It means that part of the K present in the 2:1 claymineral interlayer can be released by diffusion into the soil solution. In order to estimate the quantity of K retained in this form, other extractors can be used, such as sodium tetraphenylborate. The amount of K extracted by this method is generically called “non-exchangeable K”.

The main soil type used for flooded rice cultivation in subtropical paddy fields is the Planosol derived from granite-gnaissitic and sedimentary rocks (Silva et al., 2019). These soils have high natural K reserves, mainly associated with the presence of K-feldspar and mica in the silt fraction, and the presence of 2:1 clayminerals (smectite, vermiculite and illite) in the clay fraction (Britzke et al., 2012). The non-exchangeable and total K content in Planosols can reach up to 1,750 and 39,500 mg kg⁻¹ respectively (Nachtigall and Vahl, 1991; Britzke et al., 2012). In these soils, when the amount of K supplied by fertilizers is lower than the uptake by plants, there is a release of K adsorbed in the siloxane cavities of 2:1 clayminerals (Golestanifard et al., 2020). On the other hand, when the addition of K is higher than the plant need, 2:1 clayminerals, such as vermiculite, can act as a K sink, adsorbing the excess K in their interlayer spaces (Mortele et al., 2019).

The combined use of crop rotation, ICLS and NT enhances K cycling, increasing K use efficiency (Assmann et al., 2017; Ferreira et al., 2009). In addition, in paddy fields, wetting and drying cycles (Weil and Brady, 2017) may result in fixation or release of K in soils with excess or deficiency of K, respectively (Scott and Smith, 1968; Olk et al., 1995; Golestanifard et al., 2020). Thus, the K sink or source behavior of the soil will be affected by the cultivation frequency of flooded rice and rainfed crops in summer (soybean, maize and pasture). Therefore, our hypothesis is that the K will accumulate differently in the soil with a traditional system based on flooded rice monocropping in summer and winter fallow with CT compared to a more diversified and conservationist system, such as ICLS under NT, especially when associated with paddy-upland crop rotation in summer season.

This study aimed to evaluate the forms of K in the soil and to quantify the K budget after 66 months of cultivation with different production systems of flooded rice, involving different soil tillage, animal grazing and crop rotation in a Brazilian subtropical paddy field.