NITROGEN SUPPLY TO CORN FROM SUNN HEMP AND VELVET BEAN GREEN MANURES

Due to their nitrogen fixation potential, legumes represent an alternative for supplying nutrients, substituting or complementing mineral fertilization in cropping systems involving green manuring. The objective of this study was to evaluate the N balance in a soil-plant system involving green manures [sunn hemp (Crotalaria juncea L.) and velvet bean (Mucuna aterrima Piper & Tracy)], both labeled with N. They were incorporated into two soils of contrasting textural classes: a clayey Eutrudox and a sandy-clayey Paleudalf, both cultivated with corn. The research was carried out in a greenhouse, using pots containing 6 kg of air dried soil, to which the equivalent to 13 Mg ha dry matter of above-ground mass plus 2.7 or 2.2 Mg ha of velvet bean and sunn hemp roots were incorporated, respectively, with N labeling of either shoots or roots. One hundred days after emergence of the corn, the velvet bean residues provided higher accumulation of N in the soil, higher absorption by corn plants and accumulation in the shoot. The green manure decomposition was more intense in the medium-textured Paleudalf. The highest nitrogen losses were also observed in this soil.


INTRODUCTION
The incorporation of plant materials to soils, with the objective of maintaining or improving fertility for the subsequent crop is known as green manuring.Because of its nitrogen fixation potential, legumes represent an alternative for supplying nutrients, substituting or complementing mineral fertilization in cropping systems involving green manuring.According to Montojos & Gargantini (1963), this practice, known for centuries, causes changes in soil physical, chemical and biological characteristics, bringing benefits to Sci.Agric.(Piracicaba, Braz.), v.66, n.3, p.386-394, May/June 2009 the subsequent crop (Gallo et al., 1983;Gallo et al., 1986;Pereira et al., 1988;Kanthack et al., 1991 andTanaka et al., 1992) both in small-scale cropping systems and in larger commercial areas such as those grown with sugarcane (Mascarenhas et al., 1994;Ambrosano et al., 2005).
Residue incorporation studies of legumes using 15 N label indicate that 10 to 34% of the legume N can be recovered in the subsequent rye or wheat crop (Ladd et al., 1981b(Ladd et al., , 1983;;Azam et al., 1986;Muller & Sundman, 1988;Ta & Faris, 1990), 42% in rice (Silva, 1991), 24% recovery from Velvet bean by corn crop (Scivittaro et al., 2003), around 15% of N recovery from sunn hemp in corn plants in no-till system (Silva, 2005) and 5% of N recovery from sunn hemp by sugarcane (Ambrosano et al., 2005).Ladd et al. (1981b) incorporated Medicago littoralis into two soil types and observed that the mineralization rate of N of the legume residue was more intense in the less fertile soil.In this soil the N mineralization rates were around 10%, while in the most fertile soil around 5%, in samples collected after wheat seeding and after the flowering, respectively.The authors suggested that legume residue decomposition provided long-term supply of N for the subsequent crops, by not supplying the nutrient as an immediate source.
The objective of this study was to evaluate the utilization of N by the corn plant, from legumes parts (shoots and roots) incorporated into the soil, and also understand the recovery of the soil nitrogen process in the soil-plant system.

MATERIAL AND METHODS
A sandy-clayey Paleudalf (PVAd) and a clayey Eutrudox (LVef), collected from the 0-20 cm layer were used.The soil chemical characteristics, before and after the incorporation of nutrients were evaluated according to Raij et al. (2001) and are shown in Table 1.The study was carried out in pots containing 6 kg of air dried soil, corresponding to 5.3 and 4.7 kg for the of oven-dried soil (110°C ) for the Paleudalf and the Eutrudox, respectively.All calculations were based on oven-dried soil.
Eight days before incubation the soil of each pot was fertilized with 1 g per pot of P (triple superphosphate); 700 mg per pot of K (potassium chloride); 250 mg per pot of Mg (magnesium oxide); 20 mg per pot of Zn (zinc sulphate); 20 mg per pot of Mn (manganese sulphate); 6 mg per pot of Cu (copper sulphate); 2 mg per pot of B (boric acid).To raise the base saturation to 70% only the PVAd received 5.4 g per pot of a mixture 3:1 of CaCO 3 and MgCO 3 .A basal fertilization was made in all treatments to ensure full corn development following the suggestion of Bull & Cantarella (1993).Eight days after soil fertilization and incorporation new samples were collected and analyzed (Table 1).

Green manures labeling with 15 N
In the first experimental phase velvet bean and sunn hemp were sown and labeled with 15 N, as described by Ambrosano et al. (1997).Nitrogen (1.2 g) was applied as ammonium sulphate with 11.37 15 N atoms % excess, in three split-applications.The dried labeled material contained 3.177 e 4.337% 15 N atoms, for velvet bean and sunn hemp respectively.The green manures were cut 94 days after emergence, when plants still had not reached flowering stage.

Green manures incorporated into the soils
The incubation was performed with 39 g per pot of dry matter of the above-ground parts of sunn hemp and velvet bean, equivalent to 13 Mg ha -1 of dry matter.Roots were also added: 6.6 and 8.6 g per pot of root dry matter, corresponding to 2.8 and 2.2 Mg ha -1 , respectively, of the sunn hemp and velvet bean, following the proportions of above-and below-ground parts obtained in the pots, 16.7% and 21.3% respectively of sunn hemp and velvet bean, as described in Ambrosano et al. (1997).Soils were incubated for 36 days, before corn sowing, and monitored for nitrogen immobilization, which was not observed (Ambrosano et al., 2003).
The incubation was carried out in a greenhouse, whose climatological data is shown in Figure

Corn hybrid sowing
Ten seeds per pot of AG-303 corn hybrid were sown and soon after the emergence the plants were thinned to four plants per pot.Thirty days after sowing pots were fertilized with 238 mg P (KH 2 PO 4 ), 500 mg K (K 2 SO 4 ), and 125 mg Mg (magnesium sulphate).The fertilization was repeated after 30 days, except for Mg.Corn plants were sampled five times, at 20, 30, 40, 60 and 100 days after emergence (DAE).Plants were separated in above ground part and roots and the dry mass (DM) was determined at each sampling.The sampled materials were washed and oven dried at 60°C and ground in a rotary mill with stainless steel blades.

Nitrogen concentration and isotopic composition in plant tissues and soil
Nitrogen concentration in plant tissues was determined using the semi-micro Kjeldahl method, including the NO 3 --N, according to Bremner (1996).Samples for isotopic analysis were processed by the modified Dumas combustion method, according to Proksch (1969), and the isotopic composition was determined using a VARIAN MAT model CH4 mass spectrometer (Trivelin et al., 1973).
The percentage and amount of nitrogen in the plant derived from the labeled green manure part (Npdfgm; % and QNpdfgm; mg per pot) in the soil (Nsdfgm; % and QNsdfgm; mg per pot), and in the soil mineral N (Nsmdfgm; % and QNsmdfgm; mg kg -1 ) were calculated based on the isotopic results (atom %), using the following equations (Trivelin et al., 1994): where a and b are 15 N abundance values (atom % in excess) in the plant (corn) and in the green manure (above ground parts and roots), respectively; NA is the amount of accumulated nitrogen in the plant (mg per pot).Similarly for the soil N, a and b are 15 N abundance values (atom % in excess) in the soil and in the labeled green manure (above ground parts and roots) respectively; NA is total amount of nitrogen in the soil (mg per pot).For the soil mineral N, the procedures were also similar.Soil nitrogen was quantified in samples dried at 60ºC, which were subsampled, and dried at 110ºC, to later adjust them for dry weight.The soil total-N analyses were made by digestion-distillation method according to Buresh et al. (1982).The isotopic N analyses were performed in the extract obtained for total-N determination after wet drying (Rittenberg, 1946), modified according to IAEA (1976).
Mineral N (nitrate and ammonium) was determined after extraction with 2 mol L -1 KCl, according to Buresh et al. (1982).The distilled extracts were also used for 15 N determinations.Isotopic determinations of soil mineral N were possible only up to the 3 rd sampling (40 DAE) because of the low soil mineral N content afterwards.
The soil mineral nitrogen content was calculated in mg per pot by multiplying the value in mg kg -1 by the total mass of oven-dried soil contained in the pots.The total mineralized N was calculated for the sunn hemp and velvet bean sand, without green manure, by adding the soil mineral N to the treatments of whole plant total N.
The N balance in the soil-plant system was calculated considering the nitrogen use efficiency by the whole plant, as well as the N present in the soil and the mineral N. The losses were determined by the difference between the N input via green manure plus soil N, and that taken up by the plant.
Nitrogen use efficiency (UE) was calculated according to equations 3, 4 and 5: % plant UE = (QNapdfgm + QNrdfgm/ QNagm) × 100 (3) % soil UE = (QNsdfgm/ QNagm) × 100 (4)  In the N balance, the applied nutrient amounts were computed in relation to those derived from the green manure (mg per pot), and soil N recovered was the sum of N-mineral plus total N in the soil coming from the green manure.
A completely randomized design with a factorial arrangement of treatments was used.Treatments were the combinations among two legumes (green manures), two soils; two legume parts labelled with 15 N (shoots or roots); five sampling times and three replicates.The experiment had also an additional treatment, without green manuring, consisting of two soils, five sampling times and three replications.Treatment factor means: soil and plant parts were compared by F-test.The Tukey test was used for comparisons between means of treatment factors (velvet bean, sunn hemp and without green manure) and polynomial regression was used to evaluate the effect of sampling times.The Sanest computer program was used to run the analyses.

RESULTS AND DISCUSSION
The Eutrudox presented higher fertility and higher buffering capacity than the Paleudalf, as it can well be seen by the small change in pH and the lower reduction in the H+Al value, eight days after liming and chemical fertilization (Table 1).Both soils with green manure had higher N concentrations than those without it.Regardless of soil type, whenever velvet bean was the incorporated source, the N contents were higher (Table 2), as shown by the higher N levels in velvet bean compared with sunn hemp (Table 3).Higher N accumulation in velvet bean can be explained by its larger N 2 symbiotic fixation under the conditions of this study.No differences were observed in the temperature and relative humidity during the experimental period that would justify differences in the decomposition process (Figure1).
Table 2 shows the mean data of % Nsdfgm as a function of the incorporated legume plant part for both soils.For the contribution of the legume nitrogen in the Paleudalf soil, the N supply from the velvet bean was higher.For velvet bean, the total of 35% had the contribution of 27.8% from the above ground partand only 7.2% from the roots, for the Paleudalf soil, which was higher than for the Eutrudox, with contributions of 11.0% and 2.8%, respectively.For sunn hemp the values were lower for the two soils.These results are consistent with the higher mineralization rate of the organic matter in soils with less clay, as shown above.The lower C/N ratio in the velvet bean above ground part as well as in the roots was lower than in sunn hemp (Table 3), favoring its higher mineralization in both soils (Table 2).
Table 2 -Total N concentrations in the Eutrudox and Paleudalf soils for the treatments and nitrogen percentage derived from the green manure (% Nsdfgm).
Means followed by the same lower case letter, in the column, and capital letter, in the rows are not different (Tukey test, p > 0.05).

Soil Velvet bean Sunn hemp
Without green manure crop   Differences were found between the means of total N concentrations of both soils, with Eutrudox having 1.259 mg kg -1 and Paleudalf 389 mg kg -1 , regardless of treatment, incorporation or not of the green manures, legume plant part, or sampling time.This was largely due to the textural differences between the two soils, which slows down the organic matter decay (Stevenson, 1994;Bayer et al., 2004).This is corroborated by the higher levels of organic matter (OM) in the Eutrudox, 30 g dm -3 of OM versus 10 g dm -3 in the Paleudalf.More than 90% of N of the surface layers of most soils is in the organic form (Stevenson, 1982).
The amounts of N in the above ground part of the corn plant that was derived from the two green manures, as a function of time can be observed in Table 4.No differences among treatments were observed until 40 DAE, however at 60 DAE the velvet bean became the major N supplier to the corn plant, probably due to its larger N concentration and lower C/N ratio (Table 2).The percentage of plant nitrogen derived from the green manure was practically constant during the experiment (Table 4).This fact is not usually observed when mineral fertilizer or another soluble N source is applied, where a greater proportion of the fertilizer nitrogen use occurs initially, with a reduction in these values in subsequent stages (Ambrosano, 1989;Calvache Ulloa, 1982;Ng Kee Kwong & Deville, 1994).
Despite the higher amounts of N accumulated in the corn plants of the Sunn hemp treatments, at the 60 and 100 DAE (Table 5) the amount of N from the green manures was higher for the velvet bean, at any correspondent date, indicating the total N supply potential of this plant (Table 4).
The reduction of the amounts accumulated at the end of the cycle followed the same pattern verified Table 4 -Nitrogen percentage in the corn above ground part derived from the green manure above ground part and roots in function of the time.
Means followed by the same lower case letter, in the rows, are not different (Tukey test, p > 0.05).+ DAE -Days after emergence of corn.ns = non significant.

Sampling times
Above ground part Roots Total Table 5 -Accumulated nitrogen in the corn above ground part with and without green manure (QN), and amount of N accumulated in the above ground part of the corn plant derived from the green manure (QNapdfgm) as a function of time.
Means followed by the same lower case letter, in the rows, in each QN or QNapdfgm, are not different (Tukey test, p > 0.05).+ DAE -Days after emergence of corn.*Significant at (p < 0.05).

Sampling times
Velvet bean QN
The occurrence of the nitrogen decline accumulation in the above ground part of the corn plant, at the end of the cycle (Table 5) was also reported by other authors (Flores, 1986;Villas Bôas, 1990), as well as for sugarcane (Ng Kee Kwong & Deville, 1994).The probable cause of this could be the fact that some N of the above ground part was redistributed to the root system, as well as lost to the atmosphere by volatilization through the leaves via transpiratory strem in either immature or senescing leaves.According to Farquhar et al. (1979), an NH 3 gaseous losses occurs through the leaves due to an increase in proteolysis during senescence.
The accumulation of N in corn roots increased with time.These results indicate that there are also losses due to shedding, root exudation and root death (Allison, 1966), and that losses of N from the above ground part are much more expressive, that the above ground part N can be translocated to the roots reflecting on the accumulation of N in roots (Figure 2).There was an increase in dry matter mass of the corn above ground part until 100 DAE.At this stage, the plants were starting to flower.The only difference observed was at the first date (20 DAE), when the control plants grew more than those with green manure (Table 6).The results suggest that, at this time there was no net immobilization of nitrogen, however, by adding corn as a test plant, the initial productivity was harmed by the treatments with green manure as observed in Table 6.After that initial period, the corn productivity was similar to the control treatment.Regarding corn roots, there was an increase in DM with time, and as observed for the aerial part without treatment effects beyond 20 DAE (Table 6).
Differences occurred between the total soil N amount and concentration that came from the green manure above ground part over time (Table 4).The above ground part contribution was on average 16.3%, compared with 3.5% for roots, indicating that the above ground part participation was four times larger than that of the roots, reflecting the greater amount added from the above ground part.Over time the variable described a quadratic curve with the largest contribution at 20 DAE, decreasing afterwards until 60 DAE and rising again at 100 DAE; the same was not verified in roots.These results are mainly due to the uptake of the N incorporated by the corn from the green manure (Table 6) in a quadratic form, opposite to what was described in Table 7, with the highest N uptake at 60 DAE and a decrease thereafter up to 100 DAE.A constant release of nitrogen from the roots was also observed.The soil mineral N (nitrate and ammonium) contents were higher until 40 DAE in the velvet bean treatment followed by the sunn hemp and the control (Table 8).Beyond 60 DAE the values were lower than 1.6 mg kg -1 , without differences among treatments (Table 8).For the sunn hemp there were time differences, with a decline in the green manure contribution to the soil mineral N.However, this did not happen with the velvet bean treatment, indicating that, besides the largest contribution of the green manure plants to the soil mineral N, there also was a greater mineralization, since the testplant continued absorbing the mineral N from the soil.
The Paleudalf had higher mineral N content than the Eutrudox, regardless of the soil sampling time (Table 9).Ladd et al. (1983) also found higher organic matter mineralization in low-N soils.The authors observed that 13% of the mineral nitrogen originated from the added plant residue, compared to 5% in the high-N soil.Jans-Hammermeister et al. (1993) observed N mineralization rates in two soils after adding pea shoot and verified that the higher rates were associated with soils having lower clay content.Similar results were also verified by Ladd et al. (1981a) and other authors cited by Bartholomew (1965).Ambrosano et al. (2003) worked with these soils in a mineralization study without plant development and verified that after seven days of incubation with Sunn hemp, velvet bean and bean residue, the total mineral N amounts resulting from the treatments were the same.
Regarding the accumulation of nitrogen in the corn shoot (mg per pot), there was a quadratic response with increments up to 60 DAE, decreasing thereafter (Table 5).The corn plant did not present growth differences among treatments up to 30 DAE, due to the low plant development and nitrogen uptake, but there was greater N accumulation in plants under the green manure treatments, after 40 DAE showing that plants started to take up more nitrogen from that  Table 7 -Amount of total soil nitrogen (mean of the two soils) from parts of the green manure labeled with 15 N (QNsdgm, mg per pot) and percentage of total soil N from parts of the green manure labelled with 15 N at the sampling times.

Figure 1 -
Figure 1-Temperature and relative humidity changes inside the greenhouse during the experimental period.

Figure 2 -
Figure 2 -Amount of N accumulated in corn roots derived from the green manure (QNrdfgm).

Table 1 -
Chemical characterization of the soils used in the experiment.

Table 3 -
Chemical characterization of the leguminous plants.
followed by the same lower case letter, in the rows, in each above ground part or roots, are not different (Tukey test, p > 0.05).

Table 6 -
Dry mass production of the corn above ground part and roots with and without green manure treatments, as a function of sampling time.

Table 8
Means followed by the same lower case letter, in the rows, are not different (Tukey test, p > 0.05).+ DAE -Days after emergence of corn.*Significant at (p < 0.05).