Mixing groundnut residues and rice straw to improve rice yield and N use efficiency

Abstract

Groundnut as a pre-rice crop is usually harvested 1–2 months before rice transplanting. During this lag phase much of N in groundnut residues could be lost due to rapid N mineralization. Mixing of abundantly available rice straw with groundnut residues may be a means for reducing N and improve subsequent crop yields. The objectives of this experiment were to investigate the effect of mixing groundnut residues and rice straw in different proportions on (a) growth and yield of succeeding rice, (b) groundnut residue N use efficiency and (c) N lost (¹⁵N balance) from the plant–soil system and fate of residue N in soil fractions. The experiment consisted of six treatments: (i) control (no residues), (ii) NPK (at recommended rate, 38 kg N ha⁻¹), (iii) groundnut residues 5 Mg ha⁻¹ (120 kg N ha⁻¹), (iv) rice straw 5 Mg ha⁻¹ (25 kg N ha⁻¹), (v) 1:0.5 mixed (groundnut residues 5 Mg: rice straw 2.5 Mg ha⁻¹), and (vi) 1:1 mixed (groundnut residues 5 Mg: rice straw 5 Mg ha⁻¹). After rice transplanting, samples of the lowland rice cultivar KDML 105 were periodically collected to determine growth and nutrient uptake. At final harvest, dry weight, nutrient contents and ¹⁵N recovery of labeled groundnut residues were evaluated.

Significant effects of residue management treatments on rice growth and nutrient uptake dynamics were observed from 30 days after transplanting onwards. At final harvest, the highest grain yield and N content of rice were obtained in the 1:1 (3.7 Mg ha⁻¹) and 1:0.5 mixed treatments followed by sole groundnut residues > rice straw > NPK > control (2.8 Mg ha⁻¹) treatments. There was a significant relationship between mineral N content in soil before transplanting and rice yield at harvest. Mixed residue treatments had significantly (p < 0.1) higher groundnut-derived ¹⁵N recoveries (13%) in rice than in the sole groundnut treatment (10%) but they were lower than the recovery of labeled mineral N fertilizer in the NPK treatment (29%). The highest ¹⁵N recoveries in all treatments were observed in the topsoil (0–15 cm) where over 80% of ¹⁵N was located. Systems N unaccounted for from groundnut residues (32%) were similar to fertilizer N losses (30%), but they were lowest in the 1:0.5 mixed treatment (15%). The highest δ¹³C value occurred in residue treatments and ¹³C was positively related to N uptake while negatively related to internal N use efficiency. The result from our study demonstrated that mixing groundnut residues and rice straw could delay N release during the pre-rice lag phase leading to an improved synchrony in N demand/supply and increased growth and yield of the succeeding rice and reduced N losses from the soil–plant system.

Introduction

Rice (Oryza sativa L.) in tropical Asia is mostly grown under lowland conditions with one to three crops per year, depending on rainfall or availability of irrigation water, and the use of modern short-duration varieties. The adoption of legumes into rice-based cropping systems offers opportunities to increase and sustain productivity and income of smallholders in Southeast Asia (Whitmore et al., 2000, Wijnhoud et al., 2003). Northeastern Thailand is an important area for rice production and a major source of high quality rice for export. The region has tropical savanna climate with two distinct seasons, the dry season from November to April and the rainy season from May to October. The rainy season has a slight bimodal character (peaks in May–June and August–September) (Wijnhoud et al., 2003).

Several studies have shown that some legumes can grow well before rice, producing large amounts of residues and fixing atmospheric N₂, leading to considerable increase in yields of succeeding rice crops (McDonagh et al., 1995, Toomsan et al., 1995, Toomsan et al., 2000). Groundnut (Arachis hypogea L.) is known to fix substantial quantities of N₂ from the atmosphere under favorable conditions in the tropics, i.e. 80–150 kg N ha⁻¹ in 90 days (Giller et al., 1987, McDonagh et al., 1993, Toomsan et al., 1995). Considerable benefits through residual N supplied to subsequent crops by groundnut and other legumes have been observed, when their residues were incorporated into the soil (Sisworo et al., 1990, McDonagh et al., 1993, Toomsan et al., 1995). For example McDonagh et al. (1995) showed that N accumulation in pre-rice legumes ranged from 75 to 102 kg N ha⁻¹ resulting in an average 2 Mg ha⁻¹ increase of rice yield compared to rice grown after a bare fallow control.

Groundnut is a valuable cash crop and preferred option for small-scale farmers in rice-based systems in the Northeast of Thailand. Groundnut as a pre-rice crop is usually grown in December to early January and harvested during April–May depending on cultivar. The following rice crop is planted when sufficient water is available to control water levels in the paddy field. Hence, under rainfed conditions, the time gap between groundnut harvest and rice transplanting could be as short as a few weeks to more than 2 months depending on the year. As it is difficult to store groundnut residues during this lag period farmers often leave residues at the depoding place and no attempt is made to return the residues back to the field. An alternative practice could be the incorporation of groundnut residues in the field after the harvest to recycle the accumulated nutrients contained in the residues. However, during this lag period large N losses can occur, reducing the benefit to the succeeding rice crop and enhance environmental pollution. This is because groundnut residues decompose fast and peak net N mineral accumulation in soils occurs at about 2–4 weeks after incorporation. Previous studies indicated that incorporation of groundnut residues 2 weeks before rice transplanting resulted in highest benefits for growth and yield of succeeding rice (McDonagh et al., 1995). However, this management option may not be synchronized with labor availability for residue transportation and because of lack of adequate storage facilities to keep groundnut residues during this time gap (Promsakha Na Sakonnakhon et al., 2005).

Various types of organic residues can be obtained locally in Northeast Thailand. For lowland (paddy) systems these include rice straw, some green manures and tree leaf litter from scattered trees in the lowlands. Rice straw, which is in some cases burned, is available for both upland and lowland systems. Azam et al. (1991) and Eagle et al. (2000) found that incorporation of wheat (Triticum aestivum L.) and rice residues initially had negative effects on rice yield, with N immobilization being one of the main causes (Rao and Mikkelsen, 1976, Eagle et al., 2000). Immobilization of fertilizer and crop residue N in soil is one of the most critical aspects affecting long-term fertility in rice (O. sativa L.) (Bird et al., 2001). However, plant-available N, yield, and N uptake have all been positively affected by rice straw incorporation in the long-term (Verma and Bhagat, 1992, Cassman et al., 1993, Bird et al., 2001).

Incorporation of widely available rice straw into soil can be a strong means for controlling soil N dynamics and reducing leaching of fertilizer N, because straw enhances microbial N immobilization due to its high C/N ratio (Shindo and Nishio, 2005). Straw incorporation results in increases in microbial biomass and N mineralization (Bacon, 1990, Singh, 1995, Eagle et al., 2000) and greater soil organic carbon (SOC) and total soil N levels (Cassman et al., 1996) in the long-term. Therefore, straw incorporation has the potential to significantly alter residue N dynamics, and hence, soil nutrient supply (Eagle et al., 2000). Additionally, our previous work has shown that adding rice straw led to reduced N₂O emissions from groundnut residues but at a trade-off of increased methane emissions (Kaewpradit et al., 2008).

Vityakon et al. (2000) mixed groundnut residues with rice straw and found that peak N mineralization of the mixture was delayed from 4 to 8 weeks after incorporation. However, there was no crop planted to verify if rice was able to effectively take advantage of the delayed N release in that experiment. Similarly, according to Schwendener et al. (2005), N released from high quality leaves was immobilized upon addition of low quality leaves and their mixture was able to retain N in the available pool. The interactions of litters from different species in ecosystems may also directly affect decomposition rates of individual leaf types in mixtures, and/or the structure and activity of decomposer community (Gartner and Cardon, 2004). Chemical composition or quality of organic residues has a major influence on their rates of decomposition and N release when added to the soils (Cadisch and Giller, 1997). It is not practically feasible to change the quality of any given plant material. However, mulch or litter quality has been manipulated by mixing high- and low-quality organic matter to reduce leaching losses, prolong nutrient availability, and synchronize nutrient release with crop demand (Sisworo et al., 1990, Handayanto et al., 1997, Schwendener et al., 2005). Stable isotopes provide improved insights into plant–soil nutrient cycling (¹⁵N) as well as into interactions between plant nutrient status and ecophysiological processes. Carbon isotope (¹³C) discrimination has been widely used to describe the relationship between photosynthetic demand and diffusive supply of CO₂ (Farquhar et al., 1989). There has been further evidence that there is also a close relationship between ¹³C discrimination and plant N nutrition as photosynthetic capacity is related to leaf N status and soil N supply (Hamerlynck et al., 2004, Pansak et al., 2007). As irrigated lowland rice does not experience major water stress, taking into account isotopic carbon discrimination might provide further insight into N limitations to rice plants under different residue treatments.

The present study consisted of three parts, i.e. a preceding groundnut crop, a pre-rice lag phase and a rice phase aiming to investigate N dynamics and rice performance upon addition of groundnut residues and rice straw mixtures in different proportions. The dynamics of soil mineral N during the lag phase between groundnut harvest to rice transplanting, and associated greenhouse gas emissions (N₂O, CH₄) to rice harvesting have been reported by Kaewpradit et al. (2008). The objectives of this study were, thus, to assess the impact of addition of groundnut residues and rice straw mixtures in different proportions on: (a) growth, nutrient uptake dynamics and yield of succeeding rice, (b) groundnut residue and mineral N fertilizer N use efficiencies and (c) N unaccounted for (¹⁵N balance) from the plant–soil system as well as fate of residue N in soil fractions.