Potassium application to the cover crop prior to cotton planting as a fertilization strategy in sandy soils

Potassium (K) is the most abundant plant cation and is required by cotton in amounts similar to nitrogen. It regulates stomatal function, CO₂ fixation, enzyme activation, nutrient transport, stress tolerance, and ultimately growth, fiber yield, and quality. In soils, K reaches roots largely by diffusion (72–96% of demand), with mass flow contributing when K in soil solution is high. Tropical sandy soils, dominated by 1:1 clays and Fe/Al oxides with low organic matter and clay, have low CEC and water retention, limiting K binding and increasing the risk of leaching; excess K can be leached below 1.0 m. Cover crops with vigorous, deep roots can reduce nutrient leaching. While N cycling by cover crops is well documented, fewer studies address K dynamics. Because K is largely in vacuoles/cytoplasm and not structurally bound, it is easily leached from decomposing residues and becomes available to the following crop. Grasses of the genus Urochloa can maintain soil cover during fallow and enhance nutrient cycling, reported for soybean and maize in clay soils. It is unknown if similar benefits occur for cotton in sandy soils where K leaching risk is high. Rapid K cycling by Urochloa could enable early K fertilization before planting, improving operational flexibility. Early K fertilization benefits cotton in medium/clay soils with high CEC are known, but data are lacking for sandy soils prone to K leaching, especially regarding fiber quality. Additionally, ruzigrass may improve soil physical structure via biopores and exude organic acids that complex Al and increase Ca, Mg, and K in CEC, potentially improving K use efficiency. The objective was to assess plant nutrition, fiber yield and quality of two cotton cultivars under different K fertilizer managements and the use of ruzigrass as a cover crop in a tropical sandy soil.

Prior research shows: (1) K is critical for cotton physiology and fiber formation; (2) in tropical sandy soils, low CEC and water holding elevate K in solution transiently and increase leaching risk, with excess K leached below 1 m; (3) cover crops, especially Urochloa spp., can reduce nutrient leaching and cycle K, as K is weakly bound in plant tissues and readily washed from residues, improving availability to subsequent crops; (4) in rotations (e.g., corn–brachiaria, soybean–maize) cover crops enhanced P and K balance in clay soils; (5) split K or slow-release K can improve fertilizer efficiency in sandy soils; (6) early K fertilization improved cotton yields in higher CEC soils, but evidence was lacking for sandy soils and fiber quality; (7) grasses can enhance K acquisition via root exudates mobilizing non-exchangeable K and extensive root systems; (8) crop and rainfall modulate K release from residues and temporal K dynamics. These findings motivate testing whether applying K to ruzigrass prior to cotton can safely replace split K in sandy soils while maintaining cotton yield and fiber quality.

Location and soil: Field experiments in 2016/17 and 2017/18 at Presidente Bernardes, São Paulo State, Brazil (22°07′32″ S, 51°23′20″ W; 475 m asl). Soil: Typic Rhodustult, sandy loam (USDA). Climate: Aw (tropical, dry winters, wet/hot summers). Pre-experiment soil sampling at 0–20 and 20–40 cm on May 05, 2016; key properties included low SOM, low CEC, acidic pH, and sandy texture (0–20 cm: pH 4.7, SOM 13.6 g dm⁻³, CEC 30.7 mmolc dm⁻³; sand 848 g kg⁻¹). Lime applied: 1400 kg ha⁻¹ (Sep 2016) and 2000 kg ha⁻¹ (Oct 2017). Weather (rainfall, temperatures, VPD) recorded (Figure 6). Design: 2 × 6 factorial (two cotton cultivars: FM 913GLT, FM 983GLT; six K fertilization schemes) in randomized complete blocks with five replicates. Treatments repeated on same plots both seasons. Cover crop and K fertilization schemes (Table 4):

• 0 K: No ruzigrass (RZ), no K.
• 116 K: No RZ; total 116 kg K ha⁻¹ to cotton, split 58 kg ha⁻¹ at 30 DAE and 58 kg ha⁻¹ at 45 DAE.
• 0 K + RZ: With ruzigrass; no K.
• 116 K on RZ: With ruzigrass; 116 kg K ha⁻¹ applied to RZ at vegetative stage (≈90 DAE of RZ).
• 58 K on RZ + 58 K on C: With ruzigrass; 58 kg K ha⁻¹ to RZ at vegetative stage + 58 kg K ha⁻¹ to cotton at 30 DAE.
• 116 K on C + RZ: With ruzigrass; 116 kg K ha⁻¹ to cotton split 58 kg ha⁻¹ at 30 DAE and 58 kg ha⁻¹ at 45 DAE. Cover crop management: Ruzigrass sown May 05, 2016 and June 06, 2017 at 14 kg ha⁻¹ pure viable seed. KCl used as K source to RZ in early September each year. RZ desiccated Nov 1, 2016 and Nov 6, 2017 with glyphosate (1.92 g ha⁻¹ a.i.). Cotton crop: Planted Dec 9, 2016 and Nov 23, 2017. Base fertilizer at planting: 56 kg P ha⁻¹ and 30 kg N ha⁻¹ as MAP. Sidedress N: 140 kg N ha⁻¹ split at 35 and 45 DAE (ammonium sulfate and urea). Boric acid foliar: 2.0 kg ha⁻¹ split into four weekly sprays from first flower. Crop protection per standard practice. Defoliation: 7 days pre-harvest with thidiazuron (60 g ha⁻¹ a.i.) + diuron (30 g ha⁻¹ a.i.). Harvest: handpicked at 141 DAE (2016/17) and 132 DAE (2017/18). Measurements: At full flowering, the 4th/5th leaf from the top sampled (10 plants per plot), oven-dried (65 °C) and analyzed for K by AAS. At harvest, plant height and node number measured (5 plants per plot). Yield, boll number and boll weight determined from all bolls in central 2 m of inner rows; subsamples ginned to determine gin turnout and fiber properties (strength, length, maturity, micronaire, short fiber content) using High Volume Instrument. Post-harvest soil (0–20 cm) sampled for exchangeable K (Raij method). Potassium use efficiency (KUE): Calculated as fiber yield increase per kg K applied. For treatments without RZ: KUE = (Y_116K – Y_0K)/116. For treatments with RZ: KUE = (Y_116K+RZ – Y_0K+RZ)/116. Expressed as kg fiber per kg K. Statistics: Data checked for homogeneity and normality, then subjected to three-way ANOVA (fixed factors: K management, cultivar, year) in RCBD with five reps. Since cultivars did not differ significantly, results averaged across cultivars. Means compared by Tukey’s test (P < 0.05). Pearson correlations computed among growth, yield components and fiber quality variables.

• Exchangeable soil K: Significant year × treatment interaction; overall higher in 2016/17 than 2017/18. By end of 2016/17, exchangeable K was higher in fertilized plots versus controls irrespective of ruzigrass; when K was applied to ruzigrass, soil exchangeable K was higher than control but lower than treatments without RZ. In 2017/18, K fertilization increased soil exchangeable K versus controls except when fertilizer was applied only to RZ. Exchangeable soil K correlated positively with yield (r = 0.42 in 2016/17; r = 0.53 in 2017/18; P < 0.05).
• Leaf K: In 2016/17, cotton after K-fertilized RZ had higher leaf K than unfertilized without RZ; in 2017/18, that treatment had lower leaf K than others. Leaf K was higher in 2017/18 across treatments, attributed to better rainfall distribution.
• Yield components and yield: Boll weight unaffected by K in 2016/17; in 2017/18 boll weight increased with K, particularly with RZ; even without K, boll weight was higher after RZ. Boll number: in 2016/17 increased with K only without RZ; in 2017/18, boll number was higher overall and responded to K broadly; with 0 K, boll number was higher after RZ. Gin turnout was lower in the absence of K. Yields were low and inconsistently affected by K in 2016/17 but were higher with K in 2017/18 regardless of RZ or K management. Without K, cotton yield was 16% higher when RZ preceded cotton.
• Plant growth and fiber properties (Table 1): In 2017/18 plants were taller (96.9 vs 81.7 cm), had more nodes (18.3 vs 12.1), and higher short fiber content (SFC 9.9% vs 7.0%) but lower fiber strength (29.9 vs 31.3 g tex⁻¹) and length (28.9 vs 29.6 mm) than in 2016/17. Plant height increased with K and was highest with 116 kg K ha⁻¹ without RZ. Node number was lower without K regardless of RZ. Fiber strength and length were not affected by K or RZ; SFC decreased when more K was available (lower SFC with K and/or RZ).
• Apparent K fertilizer use efficiency (KUE): In 2017/18 (the responsive year), KUE was lower when RZ was grown before cotton, irrespective of K application mode.
• Fiber quality: Micronaire showed year × treatment interaction, averaging lower in 2017/18. In both years, micronaire was lowest with no K and no RZ. Notably, growing RZ before cotton increased micronaire to a level comparable to applying 116 kg K ha⁻¹ without RZ. Fiber maturity responded to K in both years and was generally higher when K was applied to RZ or split between RZ and cotton.
• Correlations (Table 2): Vegetative traits and yield components were mostly positively correlated, though coefficients were generally < 0.60, except node and boll number with fiber yield (stronger associations). Fiber yield tended to be negatively correlated with most fiber quality parameters, with low but significant coefficients.

The study addressed whether applying K to a ruzigrass cover crop prior to cotton on a tropical sandy soil could sustain cotton nutrition, yield, and fiber quality without splitting K applications. Results indicate that K applied to ruzigrass was timely recycled to cotton, supporting similar yields and fiber quality to conventional split K directly to cotton. Benefits of ruzigrass included increased boll weight and number even without K fertilization, and a 16% yield increase at 0 K when RZ preceded cotton, reflecting enhanced K availability from cycling and improved soil–root conditions. Seasonal weather modulated responses: in 2016/17, water deficits around flowering and low temperatures during boll maturation limited K uptake (low leaf K) and yield; K release from RZ residues was likely delayed under dry conditions, explaining higher post-harvest soil K. In 2017/18, better rainfall improved K uptake (higher leaf K), plant growth, and yield. Splitting K did not improve yield or fiber quality when cotton followed RZ, likely because K near seeds at planting was absent (avoiding seedling damage) and because K taken up by RZ was released in time to meet cotton demand. Although apparent fertilizer K use efficiency was lower with RZ (due to additional K supplied via cycling and soil mobilization), cotton dependence on fertilizer decreased, indicating potential for optimizing or reducing K rates in systems including RZ. Fiber quality responses were favorable: SFC decreased with greater K availability and RZ; micronaire and maturity improved with K applied to or combined with RZ. Overall, applying all K to the cover crop before cotton appears agronomically sound in sandy soils when ruzigrass is used, mitigating leaching losses through biological capture and gradual release, and maintaining cotton performance.

Early application of K fertilizer to ruzigrass grown before cotton supports timely K recycling to the soil–plant system, ensuring adequate cotton K nutrition, yield, and fiber quality without the need to split K applications—a practice typically recommended for sandy soils without ruzigrass. By decreasing K loss risks, this strategy can, in the long term, allow for reductions in K fertilizer rates, lowering production costs and improving operational efficiency.

• The experiment was conducted at a single location over two seasons; findings may be site- and climate-specific.
• Seasonal weather strongly influenced K dynamics and crop response (e.g., drought and low temperatures in 2016/17 affected K uptake and yield), indicating dependence on rainfall for K release from residues.
• Soil type was a tropical sandy loam with low CEC; results may differ in finer-textured or higher-CEC soils.
• Apparent K use efficiency calculations are based on treatment contrasts and may not capture total system K dynamics (e.g., soil K pools, long-term depletion or accrual).
• While indications suggest potential to reduce K rates with ruzigrass, long-term impacts on soil K reserves were not assessed.