Ruzigrass affecting soil ‐ phosphorus availability

The objective of this work was to evaluate the effectiveness of ruzigrass (Urochloa ruziziensis) in enhancing soil‐P availability in areas fertilized with soluble or reactive rock phosphates. The area had been cropped for five years under no‐till, in a system involving soybean, triticale/black‐oat, and pearl millet. Previously to the five‐year cultivation period, corrective phosphorus fertilization was applied once on soil surface, at 0.0 and 80 kg ha‐1 P2O5, as triple superphosphate or Arad rock phosphate. After this five‐year period, plots received the same corrective P fertilization as before and ruzigrass was introduced to the cropping system in the stead of the other cover crops. Soil samples were taken (0–10 cm) after ruzigrass cultivation and subjected to soil‐P fractionation. Soybean was grown thereafter without P application to seed furrow. Phosphorus availability in plots with ruzigrass was compared to the ones with spontaneous vegetation for two years. Ruzigrass cultivation increased inorganic (resin‐extracted) and organic (NaHCO3) soil P, as well as P concentration in soybean leaves, regardless of the P source. However, soybean yield did not increase significantly due to ruzigrass introduction to the cropping system. Soil‐P availability did not differ between soluble and reactive P sources. Ruzigrass increases soil‐P availability, especially where corrective P fertilization is performed.


Introduction
In tropical and subtropical regions, most soils have low P availability and high adsorption capacity, making it necessary to repeatedly apply P fertilizers to soil in order to sustain high crop yields.Water-soluble P is rapidly transformed into P forms that are less or completely unavailable to plants.Soil organic P (Po) often accounts for 50-70% of the total P in soils, and it is mainly present as inositol pentaand hexaphosphates (Borie & Rubio, 2003).These organic forms must be mineralized to inorganic P (Pi), such as orthophosphate (ortho-P), to allow its uptake by plants.Therefore, soil-P fractionation is fundamental to understand P bioavailability in agriculture areas.
In 2006, the cropping system was changed to ruzigrass and fallow, instead of triticale or black oat and pearl millet.In April 2006, the area received another corrective phosphorus fertilization (3 rd application), with the same doses and sources as before, applied to the same plots.Triple superphosphate had 179 g kg -1 P, 92 g kg -1 Ca, and 10 g kg -1 S; and the reactive Arad phosphate had 139 g kg -1 P, 264 g kg -1 Ca, and 10 g kg -1 S. The fertilizers were broadcasted on soil surface; and ruzigrass was planted (without fertilizers) in half of the plots, at the seed rate of 30 kg ha -1 (32% of viable seeds).Half of the plots were left with spontaneous vegetation.Ruzigrass and the spontaneous vegetation were desiccated 215 days after emergence (DAE) using glyphosate at 2.88 kg ha -1 (a.e.).The impact of ruzigrass on the cropping system was studied in 2006-2007.
To estimate forage dry matter yield, plant residues were sampled at six random sites per plot, using a 0.25 m 2 (0.5x0.5 m) wooden frame, and dried in an air-forced oven at 60ºC, for 72 hours.Samples from the residues were weighed, and subsamples were analyzed for N, P, K, Ca and Mg concentrations.Nitrogen was determined by sulfuric acid digestion and steam distillation.P, K, Ca, and Mg were determined using a atomic absorption spectrometer AA-7000, (Shimadzu Scientific Instruments, Kyoto, Japan), after wet acid digestion.
In November 2006, six soil samples were randomly taken from each plot with an auger, at two depths (0-5 and 5-10 cm), and combined into a composite sample Accumulation of P in plant tissue may reduce its losses in soil by chemical fixation or occlusion.Cover crops are usually more efficient to absorb less labile P forms and, therefore, their introduction into cropping systems may improve P availability to plants, since ortho-P -which is readily available to them -is released back to the soil by the mineralization of those tissues (Pavinato & Rosolem, 2008).
Cover crops such as black oat (Avena sativa), velvet bean (Mucuna pruriens), common vetch (Vicia sativa), lupins (Lupinus albus), and the ones from Urochloa genus (Syn.Brachiaria) have been extensively studied as to their efficiency to cycle P.However, their P cycling potential in agriculture ecosystems are not fully understood.Moreover, the effects of cover crop residues on soil-P transformations and availability to succeeding crops are still unclear.
Ruzigrass (Urochloa ruziziensis) has been widely used in crop rotation and crop-livestock integrated systems in Brazil because of its good adaptation to low-fertility soils, high yield potential, good forage quality, and ready desiccation (Garcia et al., 2008).Furthermore, this tropical grass has been reported to increase P apparent recovery in cropping systems (Sousa et al., 2010).Increased P availability by cover crops has been observed under no-till because organic acids, stemming from organic matter breakdown, can compete with orthophosphate for adsorption sites (Pavinato et al., 2009).In addition, enhanced mineralization of organic P from the added plant residues increase soil available P (Erich et al., 2002).
Phosphorus pools in soil can be characterized by sequential chemical extraction procedures (Hedley et al., 1982).Ruzigrass efficiency in acquiring less soluble soil P and its P cycling potential are important information for managing P fertilization in soils with high P fixing capacity.If ruzigrass could enhance soil-P bioavailability, by uptaking less soluble forms of this nutrient and returning it to the soil by mineralization, it could enhance the agronomic efficiency of less soluble phosphates, such as the reactive Arad and Gafsa rock phosphates, which may be cheaper than soluble sources, but usually provides lower yields in the first cropping year (Horowitz & Meurer, 2004).
The objective of this study was to evaluate the effectiveness of ruzigrass (Urochloa ruziziensis) in enhancing soil-P availability in areas fertilized with soluble or reactive rock phosphates.per depth, for analysis.Soil pH was determined in 0.01 mol L -1 CaCl 2 at 1:2.5 soil/solution (w/v) ratio using a pH meter DM-22 (Digimed, São Paulo, SP, Brazil), and organic matter was determined by the Walkley-Black method, as described by Raij et al. (2001).
After ruzigrass desiccation (November 2006), soybean 'BRS 184' was mechanically planted over the standing cover crop residues, in rows 0.45 m apart, at a final average stand density of 320,000 plants per hectare.
No phosphate fertilizer was applied to seed furrows.After planting, 45 kg ha -1 K 2 O was broadcasted as potassium chloride to all plots.Soybean was harvested 128 days after plant emergence.At full flowering stage, 30 recently matured soybean leaves per plot were sampled -the third or fourth fully developed trifoliate from the top -, washed, dried at 60 o C for 48 hous, and ground for P analysis, as described for ruzigrass residues.
Soil P was fractioned according to Hedley et al. (1982), with modifications proposed by Condron et al. (1985).To estimate available P, 0.5 g soil was shaken in water suspension for 16 hours, on a horizontal shaker (end-over-end) with one strip of the anion exchange resin membrane Anionic Resin 204SZRA-88091668 (GE Water & Process Technologies, Trevose, PA, USA).Pi and Po fractions were extracted with 0.5 mol L -1 NaHCO 3 (pH 8.5); then, Pi (oxide-bound P) and Po fractions were extracted with 0.1 mol L -1 NaOH; following, calcium-bound P (Ca-P) was extracted with 1.0 mol L -1 HCl; and finally, Pi and Po fractions were extracted with 0.5 mol L -1 NaOH.The concentration of P in the extracts was determined by the ascorbic-reduction molybdate blue colorimetric method (Murphy & Riley, 1962).All samples were analyzed in triplicate.According to Hedley fractionation, the correspondent P fractions are: resin-Pi, readily inorganic P available to plants; 0.05 mol L -1 NaHCO 3 -Pi, available Pi to plants; 0.1 mol L -1 NaOH-Pi, oxide-bound P; and 1.0 mol L -1 HCl-Pi, bound Pi to Ca-phosphates and Pi which is occluded within sesquioxides.
A randomized complete block design was carried out with four replicates, in a 3x2 factorial arrangement, with three P initial treatments: no P; 80 kg ha -1 P 2 O 5 as triple superphosphate; and 80 kg ha -1 P 2 O 5 as reactive phosphate; with or without ruzigrass.Data for each soil depth were analyzed separately.Plots were 5.0x8.0 m, and blocks were set 9.0 m apart from each other to allow machine traffic.Results were subjected to statistical analyses using SAS for Windows 6.11, version 8.2 (SAS Institute, Cary, USA), through the GLM procedure.Means were compared by the LSD test, at 5% probability.

Results and Discussion
Treatments did no differ as to ruzigrass average dry matter yield (4,644 kg ha -1 ) and P content in plant tissue (2.1 g kg -1 ).The average contents of N, K, Ca, and Mg tissue were 12.7, 18.0, 6.3, and 4.3 g kg -1 , respectively, which also were not influenced by P fertilization treatments.Nutrient contents were within the adequate range reported by Malavolta et al. (1997).
Neither P treatments nor ruzigrass cultivation affected soil pH and organic matter contents (Table 1).However, phosphate broadcast on soil surface, after several years of no-till, generally increased available P, irrespective of P source, mostly in the 0-5 cm depth, except for 0.5 mol L -1 NaHCO 3 -Pi extractor (Table 2).Organic acid exudation by plant roots possibly affects P movement in the soil profile (Pavinato & Rosolem, 2008).In the present study, this movement Treatment (1)  pH SOM P-resin depended on P source and on the presence of ruzigrass (Table 2 and 3).
Ruzigrass increased resin-extractable P (available P) in the soil at both depths where P treatments were applied (Table 2).At the 0-5 cm depth, forage increased available P by 13%, when reactive rock phosphate (RRP) was used, and by 21% when soluble phosphate fertilizer was used.Higher effect of ruzigrass in available P was observed at the 5-10 cm soil depth, where the relative values increased 76 and 77%, respectively.Available P in the control plot was not affected by the cultivation of the cover crop.
Soil-P fractionation showed that ruzigrass had minimal effects on Pi as extracted by 0.1 mol L -1 NaOH, regardless of P fertilizer levels or sources (Table 2).It had a significant effect on HCl-Pi, which increased 40% at the 0-5 cm soil layer in plots with RRP, and decreased 34% at the 5-10 cm layer in plots with soluble fertilizers.The introduction of ruzigrass to the cropping system significantly increased extractable-Po (0.5 mol L -1 NaHCO 3 ) at the 0-5 cm soil layer, both in the control and in the treatment which received water-soluble P (Table 3).At the 5-10 cm soil layer, only plots which received a soluble P source showed a significant increase in extractable-Po due to ruzigrass cultivation.
No changes occurred in other soil-Po fractions due to ruzigrass cultivation, except for NaOH-Pi, a less labile form as compared with resin extracted P, which was increased at both the 0-5 and 5-10 cm soil layers (Table 2).
Ruzigrass cultivation significantly interacted with phosphate treatments (Table 3) as to available P, HCl-Pi, and 0.5 mol L -1 NaOH-Pi, at both soil depths.
Table 2. Mean contents (mg kg -1 ) from different fractions (1) of inorganic P (Pi), at different soil depths, as affected by surface broadcast application of phosphorus sources, and by Urochloa ruziziensis cultivation (2) .
Table 3. Mean contents (mg kg -1 ) from different fractions (1) of organic P (Po), at different soil depths, as affected by surface broadcast application of phosphorus sources, and by Urochloa ruziziensis cultivation (2) . (1)P fractions as in Hedley et al. (1982) fractionation: 0.5 mol L -1 NaHCO 3 -Po, available Po; and 0.1 mol L -1 NaOH-Po and 0.5 mol L -1 NaOH-Po, less available forms of PO. (2) Means followed by equal letters, lowercase between ruzigrass treatments and uppercase between P treatments, do not differ by LSD test, at 5% probability.
For Po fractions, the interaction occurred only for 0.5 mol L -1 NaHCO 3 extractable Po, at the 0-5 cm soil layer.
Phosphorus fertilization increased soybean yield and its P content in leaves.(Table 4).However, ruzigrass cultivation had no effect on soybean yield, but increased P content in leaves.In general, P contents were low compared with the adequate range (over 2.5 g kg -1 ), as reported by Rosolem & Boaretto (1989).Its deficiency was more severe and common in plots without fertilizer and ruzigrass.Ruzigrass and P sources did not interact as to soybean yields and leaf-P contents.Francisco et al. (2007) reported that P fertilization of increased finger millet (Eleusine coracana) biomass production, but had no effect on subsequent soybean yield.
The original soil P values (17.2 mg dm -3 of resin-P), down to 10 cm, was probably sufficient for ruzigrass growth, since there was no response of P fertilization as to nutrient contents on leaves and dry matter yield.Similarly, Horowitz & Meurer (2004) found no differences in the response of the genus Urochloa to natural and soluble P sources, as to biomass yield and plant-P concentration.This shows that the species is well adapted to soils with low natural fertility.
The observed buildup of available P in the topsoil due to ruzigrass cultivation agrees with the findings of Pavinato et al. (2009).This effect may result from the role of organic acids on P sorption, since ruzigrass can exude citrate or oxalate under low pH conditions (Louw-Gaume et al., 2010).Low molecular-weight organic acids, such as those, complex soil Al and compete for P exchanging sites, resulting in less P sorption in the soil (Pavinato & Rosolem, 2008).Low-P supply enhances the activities of phytases and root acid phosphatases in some grasses (Rao et al., 1999).
Soil P extracted with 0.1 mol L -1 NaOH was not affected by ruzigrass or P sources (Table 2).Gatiboni et al. (2007) observed no changes in soil-P fractions as affected by fertilization, whereas Silva et al. (2003) observed significant decreases in Fe/Al-P after growing Urochloa sp. in a pot experiment, due to high P uptake by grass.In pot experiments, plant roots are so densely distributed that nearly all soil in the pots are close to the rhizosphere, where pH is usually lower, and P availability may be higher.This may explain the effect observed by Silva et al. (2003).
Ruzigrass cultivation also affected Ca-Pi, extracted with 1 mol L -1 HCl, in the uppermost soil layer, when RRP was applied (Table 2).This result can be partially explained by the Ca level in Arad phosphate (37%), which is higher than that of triple superphosphate (13%).Moreover, some of the undissolved RRP might add to the extracted Ca-P, as evidenced by higher Ca-P in the 0-5 cm soil depth of plots.Similar results were obtained by Rodrigues et al. (2009), which observed increases in Ca-P after RRP application.The effect of ruzigrass in reducing Ca-P, in soluble phosphate plots, is likely related to the utilization of this P fraction by the cover crop.
The increase in 0.5 mol L -1 NaHCO 3 extractable Po (Table 3), due to ruzigrass cultivation, can be attributed to the exudation of some Po or from a more intense cycling of plant residues.Similar result was reported by Silva et al. (2003).Organic P constitutes a signifcant portion of total P, ranging from 15 to 80% in most soils (Stevenson, 1982), and it contributes substantially to plant-available P. NaHCO 3 extractable Po, although not directly absorbed by plant, is generally considered to be readily or potentially available to plants because of its low molecular weight and prone to readily mineralization (Hedley et al., 1982;Gatiboni et al., 2007).
The results found here indicate that U. ruziziensis cultivation as cover crop can enhance P availability in soils with high-P fixation capacity.Further research is required to identify which mechanisms are involved in assessing less labile P forms by the cover crop.
Table 4. Soybean yield and P content in leaves as affected by broadcast application of phosphorus sources and by Urochloa ruziziensis cultivation (1) .

Table 1 .
Selected chemical characteristics of the soil prior to the field trial (March 2006).