1 Introduction

In a semi-arid region of central India in 1980, the main cropping sequence of the sorghum-based crop production system was sorghum-wheat. The area under sorghum has been steadily decreasing recently, with soybean taking its place. In the previous few decades, improved cultivars that were more receptive to mineral fertilizers increased yield and the use of macronutrients. However, the increase in the use of macronutrients was disproportional, as reflected in most cereal-based production systems with having negative K balances as reported by Pathariya et al. 2022 (soybean –wheat), Ojha et al. 2021 (rice-wheat) and (Dobermann et al. 1999). Despite removing K by crops in large quantities than its external replenishment (Wang et al. 2024), K fertilization application has long been neglected (Das et al 2021). This trend over the years has led to K mining from Vertisol and Inceptisol soils of maize-wheat, soybean-wheat (Das et al. 2021) and from Vertisol of sorghum-wheat (Jadhao et al. 2019) and rice-wheat (Singh et al. 2002) cropping system. Consequently, the majority of crop plants have been reported to have a K deficiency (Meena et al. 2014). In the absence of solution K and exchangeable K from external applications, the productivity of many production systems relies on nonexchangeable reserves (structural K and fixed K) (Sparks 1987). The nonexchangeable K has been defined as the fraction of soil K, found between adjacent tetrahedron layers in di and tri- octahedral micas, vermiculites and intergrade mica plus the K present at the edges of weathered and vermiculites.

Its importance in the dynamics of potassium and plant nutrition is well known (Sparks and Huang 1985). Vertisols or associated vertisols are considered to be high in available K as well as in reserve K because of the presence of K bearing minerals, including biotite and vermiculite (Paul et al. 2021), so application of K fertilizers has not typically been considered necessary (Singh and Wanjari 2014).

The fixation and release of potassium in soil are part of a dynamic and reversible process influencing potassium availability to plants, controlled primarily by the type and particle size of the primary and secondary minerals present (Goulding 1987) and composition of clay minerals in soils (Srinivasarao et al. 1998). Nonexchangeable K is released to the exchangeable form when levels of exchangeable and soil solution K are decreased by plant uptake and leaching (Sparks et. al, 1980). This leads to K mining from a native pool of K. Under conditions of soil K depletion, the ability of soil minerals to release K generally follows the sequence: trioctahedral micas >dioctahedral micas > K feldspars (Sparks and Huang 1985; Fanning et al. 1989; Huang 2005). The ability of soil to release K is a measure of the potentially available K in the soil and aids in understanding target soil behavior as well as figuring out how much K is needed in a particular soil for a particular crop. K release to the soil solution and its movement to the root zone should be fast enough to support plant development growth in order to maximize any crop's potential yield in a given situation.

Potassium availability in different forms and combinations depends upon the equilibrium and kinetics reaction between forms of soil K, soil moisture content, temperature and the concentration of divalent cations in solution and on the exchange phase (Sparks and Huang 1985). When crops are grown intensively, the readily available and exchangeable K is removed, which frequently causes the exchangeable K to be released from nonexchangeable forms.

The dynamic of K in soil depends upon the rate of application and mining of K from the system (Singh et al. 2009). Equilibrium exists among different forms that affect level of K in soil solution at particular time and availability of K in plants. However, this dynamic equilibrium is also affected when applied K is either taken up by plants or leached into the lower soil horizons or converted into an unavailable form. The distribution of different forms of K in soils is important in understanding their availability to growing crops. The best method for recommending fertilizer to crops is to have a thorough understanding of the dynamics of K in soil. The aspect of potassium fixation and its native availability in relation to mineralogical makeup of the different soils or size differences among the fractions have been investigated extensively by Martin and Sparks, (1983), Srinivasarao et al. (2006) in Inceptisols, Vertisols and Alfisols. Also, the current interest in soil K fertility has switched from simple measurements of exchangeable K or non exchangeable K to determination of the rate at which K is supplied from these forms during crop growth (Cook 1979).Yaduvanshi and Swarup, (2006), Rao (2015), Basak and Biswas, (2009) reported the effect of industrial fertilizer and manure application on K release from Alfisols in rice-wheat, finger millet and sudangrass cropping systems. The knowledge is aimed to help understand fixing potential and rate of nonexchangeable K release from soils in relation to soil properties and devising the K fertilizer recommendations for sustaining productivity.

Dynamics of K in different soils has been studied extensively by many workers under soybean –wheat (Pathariya et al. 2022), Finger millet-Hybrid maize (Rout et al. 2017), rice-wheat (Sharma et al. 2013, Mavi and Benbi 2008), maize-wheat (Kurbah and Dixit 2019). These studies highlighted dynamics of K indicating different K fractions and their availability in response to nutrient management comprising soybean-wheat, finger millet-wheat, rice-wheat and maize –wheat cropping systems in Vertisols, Inceptisols, acid alfisols. Nonetheless, there are currently no reports on the fixation potential and nonexchangeable K release rate to assess the soil's ability to supply K to plants in Vertisols in the semi-arid tropical regions of Central India under the sorghum-wheat sequence. The study examines the kinetics of K release and fixation characteristics and potential in relation to soil properties under a 29-years continuous sorghum-wheat cropping sequence in Vertisols.

Potassium in the soil undergoes reversible exchange reactions with different pools, including solution, exchangeable, non-exchangeable, and structural K. The first two pools represent available K for crop uptake. The release of K from the non-exchangeable pool governs the K availability to crops, making non-exchangeable reserves vital for cropping productivity in the absence of a solution and exchangeable pool (Sparks 1987). Additionally, K release into the available pool is influenced by the type, particle size, and composition of primary minerals (Goulding 1987) and clay minerals in the soil (Srinivasarao et al. 1998). It is essential to study research on K release from the non-exchangeable pool in order to comprehend soil K dynamics and plant nutrition (Sparks and Huang 1985).

Non-exchangeable K assumes a significant role in maintaining K supply to plants in soils with low exchangeable K (Ghorban 2007), or soils where exchangeable K is nearing minimum exchangeable K (Das et al. 2019; 2021), or soils having significant amounts of K-fixing minerals or soils undergoing intensive cultivation without adequate external K input (Islam et al. 2017; Das et al. 2021). Potassium is usually abundant in vertisols dominated by smectites clay minerals, either exchangeably or readily available (Gurav et al. 2018), resulting in naturally high available and reserved K pools (Paul et al. 2021). Though K-bearing minerals are far less effective than conventional fertilizers in terms of providing readily available K, the former can supply enough K over a long period of time to yield agronomic benefits, particularly in highly weathered soils prone to leaching (Basak and Sarkar 2017). Recent reports on cereal-based production systems (maize-wheat, soybean-wheat, sorghum-wheat, rice-wheat) indicate negative K balance and K mining in Vertisols and Inceptisols, primarily due to imbalanced primary nutrient use (Dobermann et al. 1999; Singh et al. 2002; Jadhao et al. 2019; Das et al. 2021), resulting in K deficiency in many crop plants (Meena et al. 2014). The current emphasis on soil K fertility is on understanding the rate at which these forms supply K during crop growth, rather than just measuring exchangeable or non-exchangeable K (Cook 1979; Yaduvanshi and Swarup 2006; Rao et al., 2014; Basak and Biswas, 2009).

Thus, in order to characterize K release in Vertisols and to develop the best fertilizer recommendations for sustainable crop yield, it is essential to understand the release kinetics. Many studies have examined K dynamics in a variety of soils under a range of cropping sequences, including rice-wheat (Sharma et al. 2013; Mavi and Benbi 2008), finger millet-hybrid maize (Rout et al., 2014), maize-wheat (Kurbah and Dixit 2019; Martin and Sparks 1983; Srinivasarao et al. 2006).

These studies have focused on the K dynamics in Vertisols, Inceptisols, and acid Alfisols, as well as how nutrient management practices influence various K fractions and availability. However, in the semi-arid, subtropical regions of Central India, the sorghum-wheat cropping sequence is still adopted in some areas, but no research has been conducted to investigate the potential and release kinetics of non-exchangeable K.

The main hypothesis of this study is that the majority of the added K from inorganic fertilizers is fixed in soil and eventually becomes available to crops. However, there is currently a lack of precise information on the timing and mechanism by which this fixed K is made available to crops, as well as their impact on soil K status, crop yield, and K dynamics in Vertisols under the sorghum-wheat cropping sequence.

In order to gain insight into the soil's ability to supply K to plants, this study intends to investigate the kinetics of K release, fixation characteristics, and potential regarding soil properties in relation to crop yield under a 29-year-old sorghum-wheat cropping sequence in Vertisols.

2 Materials and Methods

2.1 Site and Treatment Description

The All India Coordinated Research Project on Long Term Fertilizer Experiment (LTFE) at Research Farm, Department of Soil Science and Agricultural Chemistry, Dr. PDKV, Akola, MS, India, launched a long-term field experiment on sorghum-wheat cropping sequence in 1988–1989 (200 42’ 34.45” N latitude and 76059’ 53.16” E longitude). The soil of experimental site was deep, clayey black soil taxonomically classified as fine, smectitic, calcareous, hyperthermic family of Typic haplusterts, also known as Vertisols (Soil Survey Staff 2014). The soil has a high water-holding capacity and a low hydraulic conductivity.

The soils are low in organic carbon (4.60 g kg-1) and available N (120 kg ha-1), very low in available P (8.40 kg ha-1) and high in available K (358 kg ha-1). The soils are marginal in S (11.8 mg kg-1) and Zn (0.62 mg kg-1) status (Table 1).

Table 1 Some important initial physico-chemical properties of the soil underStudy at the start of the experiment (1988-89)
Full size table

The experiment comprised cattle manure and fertilizers at different rates (Table 2). The annual total rainfall at the site during the experimental period was 644.6 mm as against the normal of 642.0 mm which was 100.3 per cent of normal.

Table 2 Some important soil properties after harvest wheat after 29th cropping year
Full size table

3 Cropping History and Treatments

Rainy season (July to October) was used to grow sorghum (variety CSH-9) and the rabi season (November-February) was used to grow wheat after sorghum was harvested over a 29-year period, every year. Hence, the sequence is referred to as the "sorghum-wheat cropping sequence." Throughout the experiment, the sorghum variety remained the same. But over the course of nine years (1988–1997), the wheat variety (AKW 471) underwent changes, and from 1997 to 2016, the wheat variety (AKW 1071) was grown.

The experiment was laid in randomized block design with three replications. Plot sizes were 10 by 10 m (= 100 m2). For both seasons, there were seven treatments; however, FYM application was limited to sorghum. Treatments were absolute control (CK); 100% N (N); 100% of NP (NP); 100% of NPK; 150% of NPK; 100% of NPK + Farmyard manure at a rate of 5 t ha-1. The general recommended dose of chemical fertilizer was 100:50:40 kg N, P2O5, K2O and 120:60:60 kg N, P2O5, K2O ha-1 to sorghum and wheat, respectively. The total N, P and K content of FYM was 0.57%, 0.18% and 0.59%, respectively. FYM was applied on oven dry basis before sowing of sorghum, whereas, half dose of N and full dose of P and K was applied at the time of sowing to sorghum and remaining half dose of N was applied 30 days after sowing (DAS). For wheat crop, half dose of N and full dose of P and K was applied at the time of sowing, and remaining half dose of N was applied 21 (DAS). The 29th sorghum-wheat cropping cycle of the experiment during 2016-17 was studied in the present investigation. For sorghum, the recommended dose of fertilizer was 100:50:40 kg N, P2O5, K2O ha-1; for wheat, it was 120:60:60 kg N, P2O5, K2O ha-1. Farmyard manure (0.57 % N, 0.18 % P, and 0.59 % K) was added on oven dry basis before sowing of sorghum and wheat, whereas the half dose of N and full dose of P and K was applied at the time of sowing to sorghum and remaining half dose of N was applied 30 days after sowing (DAS). For wheat crop, half dose of N and full dose of P and K was applied at the time of sowing, and the remaining half dose of N was applied 21 (DAS).

4 Soil Characteristics used for the Study

Surface (0-0.20 m) samples were collected after harvest of wheat crop (29th cycle) in April 2017. The soil samples were air-dried, ground and passed through a 2- mm sieve. Basic soil properties were analyzed in the soil samples. Soil pH and electrical conductivity (EC) were determined in a 1:2.5 soil–water suspension as described by Jackson (1973). Available N was determined by alkaline-KMnO4 method (Subbaiah and Asija 1956). Organic carbon was determined by the wet digestion method according the procedure described by Walkley and Black (1934). Available P was determined by sodium bicarbonate (0.5 M NaHCO3, pH 8.5) extraction method (Olsen et al. 1954).

Soil samples (0-0.20 m and 0.20-0.40 m) were analyzed for forms of potassium viz; the water soluble K was estimated in 1:5 soil water suspensions after shaking it for an hour and K was estimated in the filtrate by flame photometer (Pratt 1965), the exchangeable K was determined by centrifugation and decantation procedure as described by Pratt (1965). A 50 ml centrifuge tube containing ten grams of soil sample was filled with 25 ml of ammonium acetate.

After that, the contents were shaken for ten minutes. The supernatant was decanted into a 100 ml volumetric flask. Three additional extractions were performed in the same manner. The combined extracts were diluted to 100 ml with ammonium acetate, thoroughly mixed, and potassium was determined using a flame photometer (Hanway and Heidel 1952), and boiling 1 N HNO3 extractableK (Wood and Turk 1941) was estimated by boiling 2.5 g soil sample in 25 mL of 1 N HNO3 in digestion tube andsuspensions were then filtrated through Whatman filter paper No. 1 and potassium was estimated in the filtrate by flame photometer (Pratt 1965). The data on various forms of K was used to calculate the correlation coefficient with yield and release characteristics.

5 Potassium Fixation Potential

Potassium fixation potential was determined by using 15 g of surface soil from various treatments mentioned above under long-term fertilizer experiment. The soils were air dried, milled, and passed through a 60 mesh sieve (0.25 mm). The soils were placed in 250 ml conical flasks and then saturated with calcium by equilibrating with 0.5 M CaCl2 (1: 5 soil: solution) 24 hrs. The soil was then washed with absolute alcohol and thereafter with de-ionized water to remove free chlorine and native exchangeable K (Martin and Sparks 1983). The samples were dried in an oven at 60 ± 2.0°C for 10 hours. Soil samples weighing 5 g each were divided into two parts to measure K fixation potential in duplicate. Duplicate samples of 5 g of CaCl2 saturated soil were placed in a polypropylene centrifuge tubes with 50 ml of 10, 20, 30, 40, and 50 mg L−1 K solution as KCl. The treated soils were equilibrated for 24 h at 30 0C (Conti et al. 2001). After 24 hours of equilibration, the soil suspension was centrifuged for 15 minutes at 2000 rpm (Rees et al. 2013), and fixed K was estimated in the clarified liquid phase using flame photometer. Added K that was not recovered in the solution phase was considered fixed. The K fixation potential was determined by the expression;

$${\text{K}}_{\text{fix}}\left(\text{mg }{\text{kg}}^{-1}\right)={\text{K}}_{\text{sol}}-{\text{K}}_{\text{eq}}$$
(1)
$${\text{K}}_{\text{fix}}\text{Potential}\left(\%\right)=\frac{{\text{K}}_{\text{sol}}-{\text{K}}_{\text{eq}}}{{\text{K}}_{\text{sol}}}\times 100$$
(2)

Where Ksol is the concentration of added K in the range of 10, 20, 30, 40 and 50 mg L-1,Keq is the K content in soil: KCl (1:5) suspension after equilibrating for 24 h.

6 Kinetics of K Release

The kinetics of K release from Ca-saturated soil previously used for K fixation potential was investigated. Samples free of soluble and exchangeable K were placed in duplicate (one gram each) into a 50-ml centrifuge tube and mixed with 20 ml of 0.01 M CaCl2 solution before being allowed to equilibrate at 25 0C for 0, 24, 48, 72, 96, 120, 144, 168, 192, 216, 240, 264, 288, 312, 336, and 360 hours. The suspension was shaken for 1 hour before each equilibration time, centrifuged for 10 minutes at 10,000 rpm, and the K in the supernatant was measured with a flame photometer (Singh et al. 2002). The contact process was repeated with a fresh 20 ml of 0.01 M CaCl2 solution (Jalali and Khanlari 2014). The release of nonexchangeable K with time was fitted by using the equations (Eq 3–6). In order to find out the possible mechanism of K release and distinctive release rate constants from soils in sorghum-wheat cropping sequence under long term manuring and fertilization, detailed description of these equations is very well elucidated by Martin and Sparks 1983 and Sparks 1989.

In order to explain the release of K from the soils in 0-0.20 m depth among various treatments, different mathematical models (e.g., first-order, parabolic diffusion equation, zero-order, and Elovich) were tested for their suitability and best fit. The constants ‘a’ and ‘b’ of each model represent the intercept and the slope of the linear curves resulting from plotting the released K+ vs. time.

$$\text{First order reaction}:\text{In }\left(\text{K}0-\text{Kt}\right)=\text{a}-\text{bt}$$
(3)
$$\text{Parabolic diffusion}:{\text{K}}_{\text{t}}/{\text{K}}_{0}=\text{a}+ {\text{bt}}^{1/2}$$
(4)
$$\text{Zero order reaction}:\left({\text{K}}_{0}-{\text{K}}_{\text{t}}\right)=\text{a}-\text{bt}$$
(5)
$$\text{Elovich}:{\text{K}}_{\text{t}}=\text{a}+\text{b} \text{In}t$$
(6)

Where Kt is the cumulative amount of K released at time t (h); K0 is maximum released K (it was found from the equilibrium plots of K release versus time) in mg kg−1, a is the intercept and b is the constants (mg kg-1) indicates the release rate of soil K and t is time (h).

The goodness of fit of the data to these four mathematical models was tested by least square regression analysis. The standard error of estimates (SE) were calculated as per (Martin and Sparks 1983).

$$\text{SE}={\left[(\text{q}-{\text{q}}^{*}) 2/(\text{n}-2)\right]}^{1/2} .$$

7 Crop yield and plant analysis

Wheat crop was harvested at full maturity and grain yield was expressed based on 14% moisture content (Rao et al. 2014). Wheat yield was determined by discarding a 0.45 m border on all four sides. The concentrations of K in the grains and straw after oven drying were determined by acid digestion and analyzed using a flame photometer at 548 nm (Bhargava and Raghupati 2005).

8 Data Analysis

The data obtained for various soil properties were statistically analyzed using the ANOVA technique in a randomized block design. The 'F test' was used to determine the significance of the treatment differences and the least significant difference (LSD at P = 0.05). The data on various forms of K was used to calculate the correlation coefficient with grain yield and release characteristics. Simple correlation coefficients were developed to determine the relationship between wheat grain yield, various K fractions, and their release and fixing charcteristics at 0.20 and 0.20-0.40 m depths. All statistical analyses were carried out using SAS (version 9.0) and MINITAB (version 13.0) programs.

9 Results

9.1 Soil Properties after 29th Cropping Cycle

Soil organic carbon (SOC) increased with all fertilizer and manure treatments over the initial status of 4.60 g kg-1 in 1988-89, except for the control (3.79 g kg-1) (Table 2).

A significant increase in SOC stock was observed under 100% NPK + FYM, 150% NPK, and 100% NPK. The NPK + FYM treatment increased SOC to 7.87 g kg-1, which is 71% higher than the initial level, indicating significant SOC buildup and implying the need for a combination of organic manure and inorganic fertilizer to maintain soil organic carbon. The application of 150% NPK significantly increased SOC (6.64 g kg-1), which was slightly higher than 100% NPK.

The available nitrogen content of soil increased significantly from 120 kg ha-1 (initial status in 1988-89) to 324 kg ha-1 after wheat harvest (29th cropping cycle). In the current long-term experiment, continuous use of nitrogenous fertilizer over time gradually increased the status of available N over control values in all treatments; the maximum increase was observed under NPK + FYM (324 kg N ha-1), followed by 150% NPK (314 kg ha-1) under sorghum-wheat sequence.

The application of 100% NPK + FYM significantly increased available P followed by 150% NPK. The application of FYM had additive effect which may have also solubilized the native P in the soil by releasing various organic acids which had chelating effect reducing P fixation. Similar results were observed by Singh et al. 2008 in soybean-wheat cropping sequence of Vertisols. There was an improvement in available K with the application of 100% NPK + FYM and 150% NPK treatments. When compared with the initial value of available K, which was 358 kg ha-1, the build in the NPK + FYM treatment was significant, whereas, a depletion was observed in the control (167 kg ha-1), N, NP and 50% NPK treatment indicating that omitting K in crop nutrition resulted in maximum mining of its native reserve. Qiuyu et al. 2020 revealed that balance use of fertilizer significantly increased the soil quality index by mediating soil fertility which was characterized by a considerable improvement in all soil properties. The available S and Zn was increased significantly with the combined application of manures and fertilizers. The magnitude of increase was greater when using 100% NPK in conjunction with FYM.

10 Potassium Fixation Potential

Table 3 shows that the K fixing capacity varied between 63% in the 100% NPK+FYM treated plot and 77% in the control treatment at lower equilibrium solution K concentration (10 mg kg-1). At higher equilibrium solution K concentration (50 mg kg-1), it ranged from 64% in the 100% NPK + FYM treated plot and 84% in the control treatment.

Table 3 Potassium fixation capacity in different treatments under long term manuring and fertilization to sorghum-wheat sequence after 29thcycle
Full size table

The results showed that, continuous sorghum-wheat cropping without manure or fertilizers (control), 100% N of NPK and 100% NP of NPK fixed significantly more K than other treatments. In contrast, the ability to fix K decreased with increased K application, balance nutrition and an integrated nutrient-supply approach.

This is analogous to the report of Taiwo et al (2018), who concluded that the ability of the studied sandy loam and loamy sand soils to fix K was reduced by the application of poultry manures. They found that as organic manure rates increased, K fixation (1.7 to 0.38%) decreased, with 100 g/5 kg (equivalent to 5 tons manure per hectare) soil being the optimum rate, and poultry manure having a greater effect on K fixing power. The higher K fertilizer dose and exchangeable K, lower wet K fixation capacities were obtained in different soil types and nutrient levels. The present study revealed that, decreased K fixing capacity and higher amount of K release are closely associated with use of balanced fertilizers and integrated nutrient management approach.

11 Release Kinetics of Potassium

The results showed that parabolic diffusion and first order reaction described the reaction rates fairly well, as evidenced by the lowest standard error (SE) values ranging from 0.066 to 0.072 and 0.34 to 0.41, respectively (Table 4 and 5).

Table 4 The release rate constants, coefficient of determination (r2) and standard error of the estimates used to describe release kinetics of K from various treatments to sorghum–wheat system after 29 years cultivation
Full size table
Table 5 The release rate constants, coefficient of determination (r2) and standard error of the estimates used to describe release kinetics of K from various treatments to sorghum–wheat system after 29 years cultivation
Full size table

The parabolic diffusion equation also adequately described the data, indicating diffusion- controlled exchange (Jalali 2005). Schneider et al. (2013) concluded that a first-order model accurately described the time series of exchangeable-K levels during the K-fixation-capacity experiments. However, a parabolic diffusion model was more effective at explaining the variance.

The non exchangeable-K release rate constant across different levels of recommended dose of fertilizers and conjoint use of fertilizers and manures as estimated by first-order reaction indicated that, the highest release rate coefficients (b) were recorded with the application of 100% NPK along with FYM (b= 13.10 x 10-3 mg kg-1). The parabolic diffusion also provided the best fit for the kinetic data. The highest release rate coefficients (b) of parabolic diffusion were found in NPK + FYM. The release rate coefficients were 55.44 x10-3 mg kg-1 indicating minor differences variations in release rate constants. The higher K release rate constant with 100% NPK and FYM can be attributed to increased K uptake by crops, which reduces the concentrations of K in soil solution near the root surfaces, resulting the release of K held on the external surfaces of the particles.

Singh et al. (2002) observed that in such cases, K is released more gradually from the interlayer. This is supported by the data graphically represented in Fig. 1, which show that the rate of K release with time is constant and that equilibrium is reached after a relatively short sequential extraction period. The pattern of K release in first order and parabolic diffusion was similar (based on the lowest SE and higher coefficient of determination), but there were quantitative differences in the release K rate, which could be attributed to differences in manure and fertilizer levels. The rate of K release decreased as interlayer K was gradually released due to the increasing distance for subsequent K diffusion.

Fig. 1
figure 1

Rate of K release at different time intervals of K extraction in the 0-0.20 m depth in incremental and suboptimal doses of chemical fertilizers along with farmyard manure. CK; absolute control, NPK; Nitrogen, phosphorus and potassium, FYM; farmyard manure

Full size image

12 Cumulative and Rate of K Release with Time

The rate of K release is shown in Fig. 1. In general, the rate of K release decreased over time and under suboptimal treatment conditions. Initially, the rate of K release ranged from 19.9 mg kg-1 in control treatment to 56.2 mg kg-1 in NPK + FYM. Whereas, at sequential extraction time of 360 h, the rate of K release ranged from 3.50 to 7.10 mg kg-1. K release was significantly reduced and reached equilibrium at various times throughout the extraction process. The first extraction resulted in a comparatively higher potassium release, which gradually decreased with subsequent extractions and reached equilibrium in K release at intervals of 192 to 360 hours. An attaining equilibrium in K release is associated with the water soluble K, which taken up directly by plants and is components of dynamics equilibrium; it also ensures availability and gradual supply of K to crops. Crop responses to gradual release and applied K, highlight the need to evaluate the release characteristics of non exchangeable K fractions of the soil.

The higher rate of K release initially could be attributed to the release of loosely held forms, while the slower phase could be attributed to the strongly held forms. Misskire et al. (2019) found that initial rapid rate of K release followed by later slow release rate which was related to the position of K had in exchange sites. In the present study, continuous sorghum-wheat cropping with or without fertilization caused soil to develop more readily accessible exchange sites, either through organic matter or at the edges of illitic minerals, resulting in a curvilinear to parabolic desorption curve (Fig. 3). The cumulative K release was higher with the integrated use of manures and fertilizers. The lowest cumulative K was recorded initially during first extraction period, thereafter K release was linearly incresaesed as extraction time increased (Fig. 2). The magnitude of cumulative K release was higher initially and then attain equilibrium in subsequent extraction. However, there were significant variations among different treatments. Among the treatments, plots of cumulative K released with time were steeper initially during 1 to 168 h extraction time and became slightly gradual afterwards.

Fig. 2
figure 2

Cumulative K release with time as influenced by different levels of fertilizer in combination with farm yard manure. CK; absolute control, NPK; Nitrogen, phosphorus and potassium, FYM; farmyard manure

Full size image

The data fit explaining the desorbed K to 0.1 M CaCl2 solution irrespective of treatments as described by parabolic diffusion and first order (Figs. 3 and 4). It indicates the data fit explaining the fit of Kt/K0 and the t0.5 for parabolic diffusion and a fit of In (K0-Kt) and t for different treatments indicating the performance of these two equations holds good and explained better for Vertisols of Central India. Almost linear relationship between Kt/K0 and In (K0-Kt) with time, t indicates a successful description of the mathematical models tested.

Fig. 3
figure 3

Parabolic diffusion kinetics of K released in various treatments. CK; absolute control, NPK; Nitrogen, phosphorus and potassium, FYM; farmyard manure

Full size image
Fig. 4
figure 4

First order kinetics of K released in various treatments. CK; absolute control, NPK; Nitrogen, phosphorus and potassium, FYM; farmyard manure

Full size image

13 Yield of Wheat after 29th Cycle

Wheat yield was significantly increased by applying 100% NPK with farmyard manure at a rate of 5 tonnes ha-1 (Fig. 5). Wheat yields in plots receiving 100% NPK with farmyard manure at a rate of 5 tonnes ha-1 were comparable to those receiving 150% NPK. The application of balanced fertilizer (100% NPK) increased wheat yield over unbalanced fertilizer application (NP & N), suboptimal fertilizer (50% NPK), and absolute control, but the increase in yield was less over combined application of manure and fertilizer, highlighting the importance of integrated nutrient management. Continuous application of N and P in the absence of K reduces wheat yield. The application of K significantly increased the yield.

Fig. 5
figure 5

Effect of different treatment on grain and straw yield of wheat (after 29th cycle). Bars accompanied by distinct letters indicate significant differences (p ≤ 0.05) according to Duncan’s multiple range test (DMRT), CK; absolute control, NPK; Nitrogen, phosphorus and potassium, FYM; farmyard manure

Full size image

14 Correlation among Yield, K Addition and Removal with Release Rate Constants

Figure 6 shows a significant correlation (p<0.05) between various K forms and 'b' (release rate constants) values and wheat yields. Among various K form, exchangeable K (Ex-K) had higher correlation coefficient (r=0.90**) and among release rate constants of different kinetic equations, parabolic diffusion release rate constants had higher correlation coefficients (r=0.910**) with yield of wheat, indicating strong and positive correlation between yield and potassium forms and their release pattern. The relationship between K addition over 29 years in a sorghum-wheat sequence with release rate coefficient was positive and significantly correlated as evidenced by the higher correlation coefficient (r=0.910**) (Fig. 7).

Fig. 6
figure 6

Correlation coefficients for different forms of K and release rate constants of various equations. WS-K: Water soluble K, Ex-K: Exchangeable K, NE-K: Non exchangeable K, FO: First order, PD: Parabolic diffusion, ZO: Zero order, E: Elovich. Positive correlations are indicated by a dark blue shade (significant at p≤.01) and faint blue shade (significant at p≤.01). Non-significant is indicated by white shades

Full size image
Fig. 7
figure 7

Relationship among K addition in 29 years of sorghum-wheat cropping with release rate coefficient (b) (h0.5) obtained for different treatments through parabolic diffusion equations (Kt / K0 = a + bt0.5)

Full size image

15 Discussion

15.1 Soil Properties

The build-up of SOC was significantly higher under integrated nutrient management practices that included the use of chemical fertilizers and organic manures compared to suboptimal (N, NP & CK) and relatively lower levels of the recommended dose of fertilizer (50% of NPK). The addition of organic matter through farmyard manure (FYM) and root biomass contributed to better root growth. The use of balanced fertilizer (NPK) also resulted in substantially higher organic C level compared to the unfertilized control, owing to the biomass contribution from roots, crop stubbles, and residues. The decomposition of these added biomass materials most likely increased the organic C content, increasing SOC stock, implying that fertilization can improve crop-derived carbon inputs and SOC storage. Gong et al. (2012) also found that applying chemical fertilizers resulted in higher maize biomass and total organic carbon (TOC) than the CK treatment, which used no fertilizers.

The plots treated with FYM in combination with chemical fertilizer (NPK + FYM) may have facilitated the mineralization of soil nitrogen (N) due to favorable soil conditions, resulting in increased N availability (Santhy et al. 1998; Kumar et al. 2008). The optimal use of fertilizers and manures was essential for maintaining N status in Vertisols under cereal-based cropping systems. Suresh et al. (1999) and Ravankar et al. (2005) reported similar results, showing that minerals combined with organic manure were promising for maintaining higher N level in subtropical paddy soil. In semiarid regions, excessive N use has resulted in significant build-up (Schmitt and Randall 1994) due to increased mineralizable soil N, resulting in the persistence of inorganic N (Powlson et al. 2010). While the control treatment had low available N levels, continuous sorghum-wheat cropping slightly increased or maintained the initial N status, most likely due to mineralization of soil organic matter (SOM) which contributed to the replenishment of available N. Cruz et al. (2009) observed that soils with low in available N explained lower specific denitrification and higher specific nitrogenase activities in –N+P soil growing wheat. There was no significant difference in total soil N value between NP, P, and N alone treatments. However, Kaur and Singh (2014) found that 13 years of continuous rice-wheat cropping with no fertilizer application was insufficient to maintain the original total soil N level. Maintaining available N levels in semi-arid areas requires regular addition of organics to maintain SOC and soil fertility.

The higher availability of P may be attributed to the effect of organic matter, increasing soluble P in the soil through complex reactions with cations responsible for P fixation. Organic manure and inorganic fertilizers had a beneficial effect on increasing phosphate availability (Tolanur and Badanur, 2003; Bajpai et al., 2006). The results also suggest that the combined use of manures and fertilizers improves available P status, which is associated with reduced calcium carbonate (CaCO3) content in the soil, thereby reducing P fixation. Plots receiving only inorganic fertilizers had higher available P status, likely because of lower uptake than the externally applied P (Hegde, 1996) and Pathak et al. (2005). The higher build-up in available P with combined use of manures and fertilizer was due to the supply of P through FYM and reduction in P fixation, as well as mobilization and release of native P by organic acids produced during the decomposition of organic matter (Singh and Wanjari, 2007).

The higher availability of K with the combined use of manures and fertilizers, as compared to optimal (100% NPK) or super-optimal dose of NPK (150% GRDF), could be attributed to the addition of organic materials such as FYM, which supply additional nutrients N, P, K, and micronutrients to the soil. The higher available K in FYM-treated plots could be attributed to an increase in the cation exchange capacity (CEC) of soil, which is responsible for holding a greater amount of exchangeable K and aiding in the release of exchangeable K from non-exchangeable pools (Kher and Minhas, 1991). The interaction of organic matter with clay, besides direct potassium addition, may also be responsible for the improvement in available K of the soil. Sood et al. (2008) and Mandal et al. (1991) found that the build-up of potash in the soil was observed due to the application of organic manures and inorganic fertilizers, whereas, the availability of K in the soil decreased when fertilizer K and organics were not applied.

High K fixation was observed without added K fertilizers, under the sorghum-wheat sequence. This finding is consistent with Ogaard and Krogstad (2005), who reported an increase in fixation capacity of K in grass crops growing without added K fertilizers, ranging from 24-116 kg K ha-1 yr-1, which corresponds to 24-43% of the non-exchangeable K release based on soil K balances. Schneider et al. (2013) estimated greater soil fixation capacity (FC) from fixation experiments at two different K doses, concluded that the equilibrium K concentrations decreased and, conversely, the soil's K buffer power, β, and FC increased. However, Shakeri and Abtahi (2020) reported that the percentage of K fixation decreased as K concentrations increased. The decrease in K fixation with increasing levels of K fertilizers and NPK + FYM could be attributed to the fact that layered silicates have a limited fixation capacity, which can be increased or decreased by removing or adding cations to fixation sites. Furthermore, because of this limited capacity, increasing the K level would reduce the fixation percentage. Pratt and Goulben (1957) proposed a simplified explanation for this: when real fertilized soils are brought to the laboratory, they have a lower tendency to fix K than unfertilized soils. In the present study, K fixation was reduced in soils where K had been applied continuously for 29 years, whereas K fixation increased in soils with no manure or fertilizer application or a suboptimal K dose.

Among the various models, the first-order (FO) reaction provided a satisfactory explanation for the K release kinetics, as demonstrated by the higher coefficient of determination (R2) and the lowest standard error of estimate (SE). Misskire et al. (2019) also found that the power function equation showed the best fit for kinetic data. Rao (2015) successfully described K release using the FO and power equations, with rate constants ranging from 0.005 to 0.008 mg kg-1 h-1 in the FO model and 0.14 to 0.83 mg kg kg-1 h-1 in the power model for the soils. The application of NPK + FYM resulted in a higher K release rate across all sequential extraction periods. Numerous studies (Agyenim-Boateng et al. 2006; Srinivasarao et al. 1999) have reported that K release in the soil is influenced by biological activity, which is promoted by the application of organic manures. In the present experiment, the application of fertilizer K along with farmyard manure influenced K release parameters, as evidenced by the higher release rate constants. Singh et al. (2002) proposed that K concentrations in the soil solution near root surfaces decrease as K uptake increases, indicating that K is released from the external surfaces of soil particles. In the present study, the sorghum-wheat cropping system resulted in higher K uptake, temporarily declining K concentrations near the root zone, which likely influenced K release from interlayers.

The use of balanced fertilization in the sorghum-wheat sequence significantly improved the K release rate when compared to imbalanced fertilization. According to Uddin et al. (2011), incremental K doses resulted in more K release than suboptimal K doses, and the amount of K release increased with the applied K in soils. The control treatment showed a lower rate of K release during all sequential extraction periods. Similar results were reported by Shekhon and Rao (1985), who observed stepwise, constant, and cumulative K release in the range of 3.85 to 4.38, 0.03 to 0.08, and 3.85 to 10.10 meq 100 g-1, respectively, for shrink-swell soils of India. The rate of K release curve, represents different phases during the 360 hrs of K release studies. Based on the release rate (Jardin and Sparks, 1984), the curves are divided into two segments: the first part is associated with edge K, indicating rapid K release from the soil mica fraction and mass action exchange from the planar surface of the clay. The second part, or transition stage, occurs during the first 1000 hours of K release, crucial for replenishing labile soil K, and is mainly associated with wedge K and interlayer K (Benipal et al., 2006). Previous studies by Sparks et al. (1980), Rao et al. (1999), Ghosh and Singh (2001), and Jalali (2005) also reported two phases of K release. Brar et al. (2008) attributed the initial phase to the release of loosely held forms and the latter slower phase to strongly held forms of K.

In the present experiment under the sorghum-wheat sequence, the fertilizer schedule without K (control, 100% GRDF of N, and NP treated plots) significantly reduced K release parameters. This decrease in K release parameters is primarily due to continuous nutrient exploitation by the crops. The constant b for the first order (10.90 to 13.10 x 10-3 mg kg−1) and parabolic diffusion (50.79 to 55.44 × 10-3 mg kg−1), represent the release rate of K within the range previously reported by Misskire et al. (2019) of Vertisols of North Western Ethiopian Highlands and that by Srinivasarao et al. 1997 in Vertisols of Central India.

The K release data of parabolic diffusion plots for various treatments are depicted in Fig. 3. The treatment with 100% NPK + FYM resulted in a higher Kt/K0 value at various time intervals. The other treatments (control, N, NP, 50% NPK and 100% NPK) had lower Kt/K0 values but but showd similar trends of increasing at different sequential extractions. A linear fit to the parabolic diffusion model for all treatments indicates that release of non-exchangeable K is a diffusion-controlled process (Jardine and Sparks 1984). The K release was also fitted to the FO model, which was successfully described by Dhillon and Dhillon (1990, 1992) and Srinivasarao et al. (1999). The K release by the parabolic diffusion and first-order equations suggested the slow diffusion of K from illite interlayer positions and where K is strongly held in the soil. The variations in the constants a and b for the release of non-exchangeable K may be due to management interventions, particularly the optimal, suboptimal, and superoptimal levels of K applied since the inception of the experiment.

The yield of wheat was significantly higher with the combined use of fertilizer and organic manure. The INM treatment increased yield by 31% compared to NPK and 10% compared to 150% NPK. The yield gap was larger between INM and NPK but narrowed between INM and 150% NPK, indicating that incorporating 5 tonnes of FYM balanced 50% NPK by supplying significant amounts of major and micronutrients and acting as a buffer for yield changes, thereby maintaining soil fertility and sustainable crop productivity. Similar trends in yield were reported by Satisha et al. (2016) with the use of 100% NPK + FYM treatment. The application of 100% NPK + FYM recorded the highest grain yield of wheat, while the lowest yield was observed under absolute control. The increased wheat grain yield could be attributed to the supply of additional mineral nutrition through 100% NPK as well as primary, secondary, and micronutrients through FYM. Incremental levels of NPK were also beneficial in increasing yields when compared to suboptimal levels, which were associated with higher fertilizer NPK in these treatments. The response to applied sulfur was greater than that to zinc application, most likely due to the decrease in available sulfur in the soil with time. The response to applied sulfur can be expected because high analysis fertilizers with little or no sulfur are used in intensive cropping systems, resulting in increased sulfur removal from the soil (Ghosh 2020). The application of FYM alone failed to achieve sustainable yield, as seen in 100% NPK + FYM. This could be attributed to the low nutrient release from organic sources, which failed to meet the higher nutrient requirements of wheat and resulted in lower yields due to an inequity between plant nutrient demand and supply. As a result, using 100% NPK + FYM may be an effective way to maintain soil health and sustainable wheat yield. Given the scarcity of FYM, alternative organic inputs such as compost, crop residues, and green manuring should be considered. Continuous cereal-cereal cropping without the addition of P and K (100% NPK of N and 100% NPK of NP) significantly reduced the productivity of wheat over 29 years. The continuous use of only N and P resulted in potassium deficiency in the soil, as the crops absorbs large amount of K. This practice of skipping K from the regular fertilizer schedule is common in Vertisols, most likely because Vertisol is classified as a K-rich soil. However, such soils may be depleted after several years of intensive agricultural production (Singh et al. 2019). The exclusion of K significantly reduced the availability of K in the soil. In contrast, including K in the fertilizer schedule significantly increased wheat yield, indicating a positive crop response to applied K in Vertisols of Central India. Singh and Wanjari (2012) and Dwivedi et al. (2007) reported similar crop responses to applied potassium in Vertisols.

K release from the non-exchangeable pool depends on the K concentration in the soil solution. Regardless of treatment, the release rate of K increased in response to fertilizer K addition. The application of 100% NPK resulted in a K addition of 5070 kg ha-1 with a removal of 7690 kg ha-1, indicating that adsorption and desorption rates were mainly dependent on the soil K balance, and these processes are dynamic and depended on removal and addition of fertilizer K. Simonsson et al. (2007) reported that crop removal of K and K fertilizer addition rates may have a greater influence on actual K release rates than soil parent materials, as long as the soil contains enough available K for release. In the present experiment, the addition of fertilizer K increased soil available K, resulting in the release of K from the non-exchangeable pool.

16 Conclusions

The present study provides valuable insights into managing the rate of K release in shrink-swell Vertisols through tailored integrated nutrient management options. It is evident that continuous cropping in Vertisols without K addition leads to a decline in K status and may eventually result in yield failure. Modeling cumulative K release over a 29-year period revealed that K release rates in Vertisols can be estimated using first-order and parabolic equations, which were found to differe significantly among various management interventions. In the context of continuous sorghum-wheat cropping and K fertilization at the existing fertilizer K recommendations without the use of farmyard manure (FYM), there was evidence of native K depletion, resulting in a negative soil K balance. However, the addition of organic matter, such as FYM, effectively mitigated the negative K balance in the continuous sorghum-wheat cropping sequence. Therefore, it is recommended to either increase the existing inorganic fertilizer dose or adopt a balanced approach that includes inorganic fertilizers and FYM (NPK+FYM) to maintain soil K levels and ensure sustainable crop productivity. Furthermore, due to the scarcity of FYM, other alternative sources such as compost, value added compost, waste mica and the green manuring crops may replace FYM. More research is needed to combine these alternative resources with inorganic fertilizer to investigate potassium fixation, release, and dynamics in vertisols of Central India.