GREEN FERTILIZATION WITH RESIDUES OF LEGUMINOUS TREES FOR CULTIVATING MAIZE IN DEGRADED SOIL

The objective of this study was to evaluate the effects of the addition of parts of leguminous trees on the growth and nutrition of maize (Zea mays L.), as well as on the chemical attributes of a degraded soil, 65 days after applying the residues. The experiment was conducted in pots, in a randomized block design with ten treatments and four replicates. The evaluated treatments were: T1 No residues of leguminous trees, T2 Leaves of Mimosa caesalpiniifolia, T3 Branches of Mimosa caesalpiniifolia, T4 Leaves + branches of Mimosa caesalpiniifolia, T5 Leaves of Mimosa hostilis, T6 Branches of Mimosa hostilis, T7 Leaves + branches of Mimosa hostilis, T8 Leaves of Gliricidia sepium, T9 Branches of Gliricidia sepium and T10 Leaves + branches of Gliricidia sepium. Pots were filled with soil from a degraded area and residues were added in the form of green mass after sowing the maize. Residues of leguminous trees positively influenced maize growth at 65 days after application and favored the accumulation of nitrogen, phosphorus and potassium in the shoots. Green fertilizers improved soil chemical attributes at 65 days after application, especially nitrateN (NO3 -N), ammonium-N (NH4 -N), total inorganic N (NO3 -N + NH4 -N) and K, demonstrating that these species are good options for recovering degraded areas in the semi-arid region of Ceará.


INTRODUCTION
In northeastern Brazil, deforestation associated with opening new areas for agricultural exploitation, firewood and charcoal production, and the lack of soil conservation practices have contributed to soil degradation.Soils of this region often have low fertility, which limits agricultural production (MUNDUS et al., 2008).Among family farms in the Sertão Inhamuns region, Ceará state, 90 and 50% of studied areas showed low organic matter (OM) and phosphorus (P) contents (SOUZA et al., 2015).Green fertilization with leguminous plants can be an option for agricultural systems in the Ceará state, because of the capacity of these species to add OM and significant quantities of nutrients, especially nitrogen (N).This can improve the chemical, physical and biological properties of soil, which in turn improves crop development (ANDRADE NETO et al., 2010;PRIMO et al., 2012;PEREIRA;SOARES;MIRANDA, 2016).In some cases, the amount of N supplied by these species can exceed 200 kg ha -1 (DINIZ et al., 2010).
The decomposition of the leguminous plants may occur in a relatively short time because of the low C:N relationship.Species such as Crotalaria spectabilis studied in cultivation in the Ceará presented half-life times(t 1/2 ) of 65, 53 and 54 days for N, P and K (PEREIRA; SOARES; MIRANDA, 2016).
There are many methods of using leguminous species to improve soil fertility and agricultural crops.In semi-arid regions, agroforestry systems with alley cultivation (alley cropping), which are the intercropping of leguminous shrub/tree species with the food crop of interest, can be one method, since periodic tree pruning results in green fertilization (PAULINO et al., 2011).
Implementing alleys as a form of land use in tropical regions is an option for managing degraded areas, especially sites where small farms prevail (SAMPAIO et al., 2015).There is still lack of information however on the nutrition of agricultural crops through green fertilization using leguminous trees in semi-arid regions, as well as on the best species to use and which parts of the plant best promote improvements in soil fertility.
This study was based on the hypotheses that one of the leguminous species evaluated will promote better development, greater accumulation and more efficiently recover macronutrients from the maize crop; residues from leguminous trees can improve soil chemical attributes at 65 days after application; and improvements in soil chemical attributes depend on the added plant part.
Given the above, this study aimed to evaluate the effect of adding residues from different parts of leguminous trees on the growth and nutrition of maize, as well as on the alterations of soil chemical attributes 65 days after applying the residues.

MATERIAL AND METHODS
The experiment was carried out from August to October 2012 in the facilities of a seedlings nursery located in Sobral municipality, Ceará State, Brazil, located at 3° 41' S and 40° 20' W, altitude of 69 m.The climate is BShw (hot semiarid) according to the Köppen classification system, with the rainy season lasting from January to June.The average annual temperature is 28 °C and average historical rainfall is 759 mm yr -1 (SOUZA et al., 2016).
Pots were filled with Luvisol (EMBRAPA, 2006), collected from the 0.0-0.3m soil layer in a visibly degraded area, absent vegetation and in the presence of laminar erosion, in the district of Jaibaras, Sobral-CE municipality (3º 43' 30" S, 40º 22' 30" W) in a region in the desertification nucleus of Irauçuba (CE).
Samples of the plant residues that comprised the treatments were collected for chemical characterization by determining the contents of C, N, P, K, Ca and Mg (SILVA, 2009), whose results are presented in Table 1.
Each pot received 8.0 dm 3 of soil, measured with a 1.0-L cylinder.Based on the results of the chemical characterization analysis, the soil received 108.8 mg dm -3 of triple superphosphate, which corresponded to 90 kg ha -1 of P 2 O 5 (FERNANDES, 1993), using triple superphosphate as source.
After sowing the maize (Zea mays L.) variety 'BRS Gorutuba', the leguminous species residues were applied in pots in the form of green mass.The amount applied in each pot was equivalent to 73.0 g of dry matter, corresponding to 17,300 kg ha -1 , considering the mean dry biomass production (leaves + thin branches) of the three studied species in kg plant year -1 (BAKKE et al., 2007;FERREIRA et al., 2007;MARIN et al., 2006).The equivalence for dry matter was obtained through the water contents in the leaves and branches of each species.In the treatments composed of leaves + branches, the proportion was 50% for each plant part.To obtain the fraction "branches", branches with diameter ≤ 1.0 cm were selected and cut into pieces approximately 2.0 cm in length.The quantities of N, P, K, Ca and Mg contained in the residues applied in the pots are shown (Table 2).
Table 1.Chemical characterization of leguminous plant residues used in the study.*Branches with diameter ≤ 1.0 cm.Table 2. Amounts of nitrogen (N), phosphorus (P), potassium (K), calcium (Ca) and magnesium (Mg) from the residues of leguminous trees added to the soil.
Pots were irrigated daily using water from the municipal supply system of Sobral-CE, analysis showed the following chemical characteristics:pH = 7.0; EC = 0.22 dS m -1 ; Ca = 0.50; Mg = 0.75; K = 0.20; Na = 0.70; Cl -= 1.25; HCO 3 -= 1.0 (mmol c L -1 ).Initially, a water volume was applied to increase soil moisture to 80% of field capacity.Starting on the second day, the water depth to be applied was determined by weighing each pot and calculating the difference in mass relative to the previous day.
The seedlings were thinned 15 days after sowing (DAS), leaving the most vigorous plant in each pot.The plants were collected at 65 DAS, when 80% had already produced female inflorescence.Maize growth was evaluated by determining plant height (PH), number of leaves (NL), stem diameter (SD), root dry matter (RDM), shoot dry matter (SDM) and total dry matter (TDM).
PH, SD and NL were determined in the experimental area.To obtain SDM (stem, leaves and inflorescences) and RDM, plants were cut close to the soil, separated into roots and shoots and dried in an oven with air circulation and renewal at 60-65 ºC.TDM was obtained by adding RDM and SDM.
SDM samples were ground in a Wiley-type mill and used to determine the contents of N, P, K, Ca and Mg, following Silva (2009).Nutrient accumulation in maize plants was calculated by multiplying shoot dry matter by the contents of each nutrient.
Soil sampling was performed after plant collection in the 0.0-0.1 m layer, by collecting one sample for analyzing pH, EC, TOC, P, Na, K, Ca,  Mg, Al and (H + Al), and another for analyzing nitrate-N (NO 3 --N) and ammonium-N (NH 4 + -N), determined through steam distillation (DONAGEMA et al., 2011).Total inorganic N (NO 3 --N + NH 4 + -N) was obtained by the sum of the contents of NO 3 --N and NH 4 + -N.Recovery efficiency (RE) is the amount of nutrient accumulated per unit of nutrient applied, usually expressed in percentage, calculated by Equation 1: Where RE is the recovery efficiency (%), NAWF is nutrient accumulation with fertilization (kg), NAWoF is nutrient accumulation without fertilization (kg) and AAN is the amount of nutrient applied (kg).
The obtained data were subjected to the Kolmogorov-Smirnov test to verify normality and the F test for analysis of variance.When significant, means were compared by the Tukey test.All tests RE= NAWF -NAWoF AAN x 100 were done at a 0.05 probability level, using the statistical software SISVAR (FERREIRA, 2011).

RESULTS AND DISCUSSION
Results obtained for PH (Table 3) showed the addition of leaves + branches of G. sepium led to higher values compared to the other residues, except for the leaves of M. hostilis and leaves + branches of M. caesalpiniifolia.This can be explained by the fact that the "leaves" fraction has a low C:N ratio, especially leaves of G. sepium (Table 1), favoring faster decomposition and later supply of N to the crop.In addition, the mixture with branches may have favored a greater synchronism in the release of nutrients, since they have lower C:N ratio and thus undergo slower decomposition.These results corroborate Andrade Neto et al. ( 2010), who studied the effects of green fertilization on forage sorghum growth and yield and also observed a positive influence on plant height.Table 3. Mean values of plant height (PH), number of leaves per plant (NL), stem diameter (SD), root dry matter (RDM), shoot dry matter (SDM) and total dry matter (TDM) of maize plants at 65 DAS.
Means followed by the same letter in the column did not differ by the Tukey test at p < 0.05; **Significant at p < 0.01; *Significant at p < 0.05; ns Not significant at p < 0.05; (L) = Leaves, (B) = Branches and (L + B) = Leaves + Branches; CV = Coefficient of variation.
RDM was influenced by the application of leaves + branches of M. caesalpiniifolia, which differed only from the control (without residue application).All residues positively affected SDM and TDM, showing the potential of these species to promote maize development, since the dry matter is an important parameter for evaluating plant development.This effect of leguminous trees on the increase in maize biomass may result from a synchronized release of N in the soil, causing greater absorption by plants, since the residues were not incorporated, thus favoring a slower decomposition (SOUZA et al., 2016).
All residues applied to the soil positively influenced N accumulation (Table 4).This result is explained by the fact that these species have high N contents and release significant quantities to the soil, which favors absorption and accumulation by maize plants (DINIZ et al., 2010;PRIMO et al., 2012).Although leaves + branches of G. sepium had the highest absolute value, the effect of this residue differed statistically only from the addition of  1), the residue of G. sepium is a high quality material, having low contents of polyphenol and lignin (MUNDUS et al., 2008).
Means followed by the same letter in the column did not differ by the Tukey test at p < 0.05; **Significant at p < 0.01; *Significant at p < 0.05; ns Not significant at p < 0.05; (L) = Leaves, (B) = Branches and (L + B) = Leaves + Branches; CV = Coefficient of variation.
As for P accumulation, the residue applications to means that were statistically tied, but higher than the control.This indicates the residues caused P accumulation in the vegetative part of maize plants, regardless of species or plant part used.
Regarding K accumulation in the shoots of maize plants, the highest mean was obtained with the addition of G. sepium leaves + branches to the soil.This value differed statistically however only from the means caused by leaves + branches of M. caesalpiniifolia and the control, which had the lowest mean.Leaves + branches of G. sepium caused a 94.3% increase in K accumulation compared to the control.According to Ernani, Almeida and Santos (2007), K is washed from the organic material immediately after the death of the cells.Thus, when plant biomass is added to the soil, the response of the crops to K absorption can be faster in relation to the absorption of nutrients that depend on the decomposition and mineralization performed by microorganisms.
For Ca accumulation in the shoots of maize plants, only leaves + branches of G. sepium, leaves of G. sepium and leaves of M. hostilis resulted in higher means compared to the control, but did not differ statistically.For Mg accumulation, only maize plants that received leaves of G. sepium and leaves + branches of G. sepium were higher than the control.
Mimosa caesalpiniifolia branches promoted higher recovery efficiency (RE) of N in comparison to M. hostilis leaves and G. sepium leaves, with these being statistically equal to the other residues (Table 5).It is possible that the difference observed for M. caesalpiniifolia was due to this residue having the highest C:N ratio among those studied (Table 1), which favored slower decomposition of the material and therefore better use by plants.According to Fontanétti et al. (2006), crop efficiency in N recovery is associated with the synchronism between the release of N from green fertilizers and its absorption by plants.
For P, leaves + branches of M. caesalpiniifolia and leaves of M. hostilis promoted higher RE compared to branches of G. sepium and were statistically equal to the other treatments.In most treatments, the P recovery percentage was higher than that of N, which can be explained by the phosphate fertilization applied to the soil before planting.
From the observed data, it was possible to quantify the mean recovery rate of each leguminous species, as these values can contribute to fertilization management using green fertilizers in the maize crop.Thus, the following mean values are cited for each species (leaves, branches and leaves + branches), relative to N, P and K: M. caesalpiniifolia (7.5, 9.1 and 10.6%), M. hostilis (6.4,8.26 and 14.5%) and G. sepium (6.0, 5.9 and 9.2%).Total organic carbon (TOC) was positively affected by the residues of the leguminous plants and leaves of M. hostilis and 6.9 8.3 12.9 15.5 25.7 Table 5. Mean values of recovery efficiency (RE) of nitrogen (N), phosphorus (P), potassium (K), calcium (Ca) and magnesium (Mg) in the maize crop cultivated with addition of residues of leguminous trees.
*** Data transformed to square root; Means followed by the same letter in the column did not differ by the Tukey test at p < 0.05; **Significant at p < 0.01; *Significant at p < 0.05; ns Not significant at p < 0.05; (L) = Leaves, (B) = Branches and (L + B) = Leaves + Branches; CV = Coefficient of variation.
leaves + branches of G. sepium resulted in means higher than those of the control, not differing from the other treatments (Table 6).This difference, in relation to the control, can be explained by the fact that these two species have lower C:N ratios (Table 1), favoring a higher speed in the process of decomposition of the residues.
The residues of the leguminous plants did not cause an increase in soil pH (Table 1) while the leaves / leaves + branches of M. hostilis caused a reduction.In this case, the organic matter added to the soil through the residues acted as source of acidity.A possible explanation is that, at 65 days after the addition, the residues were still decomposing, releasing H + and reducing the pH.On this topic, Silva and Mendonça (2007) report that adding organic matter will increase or decrease the soil pH, depending on the predominance of processes that consume or release H + .-N) in the first 10 cm of the soil after maize cultivation with addition of residues of leguminous trees.
Means followed by the same letter in the column did not differ by the Tukey test at p < 0.05; **Significant at p < 0.01; *Significant at p < 0.05; ns Not significant at p < 0.05; (L) = Leaves, (B) = Branches and (L + B) = Leaves + Branches; CV = Coefficient of variation.There was no statistical difference between treatments regarding the Al + content (Table 6), which decreased compared to the value observed in the characterization of the soil used for cultivation (2.0 mmol dm -3 ).According to Iwata et al. (2012), a reduction in Al +3 contents is associated with the effect of soil organic matter, which acts by complexing the free aluminum in solution.
For potential acidity (H + Al), there was a significant difference only when the soil received leaves + branches of M. hostilis (highest value) and branches of M. caesalpiniifolia (lowest value).Aluminum saturation (m) significantly increased with the application of M. hostilis leaves, thus this treatment contributed to the highest mean and statistically differed only from the application of branches of G. sepium and leaves + branches of G. sepium.
The NO 3 --N content in the soil was affected by the addition of leguminous species residues, except for the leaves of M. caesalpiniifolia, which had a mean statistically equal to that of the control (Table 6).The highest NO 3 --N content was observed with the application of leaves + branches of G. sepium; however, it did not differ statistically from leaves of G. sepium, branches of G. sepium and leaves + branches of M. hostilis.The increase in NO 3 --N content caused by the addition of leaves + branches of G. sepium in relation to the control was equal to 26.4%.
As for the NH 4 + -N in the soil, the leaves + branches of G. sepium resulted in a higher mean, which was higher than the treatments with leaves of M. hostilis, branches of M. hostilis, leaves + branches of M. hostilis, leaves of M. caesalpiniifolia, branches of M. caesalpiniifolia and the control, which showed the lowest content (Table 6).The increment in NH 4 + -N caused by the addition of leaves of G. sepium compared to the control was 15.9%.
The NO 3 --N + NH 4 + -N content (Table 6) follows the same tendency that occurred for NO 3 --N and NH 4 + -N.In general, the residues that contain leaves resulted in higher soil N contents, especially G. sepium.This can be explained by the fact that this species showed the lowest C:N ratio in both leaves and branches (Table 1), which favors faster residue decomposition.Thus, N is more rapidly released into the soil, especially in the first 10 cm, where the mineralization of the biomass of the leguminous plants is most intense.These results corroborate those of Beedy et al. (2010), who worked with G. sepium intercropped with maize in South Africa and observed a positive effect of its biomass on the contents of NO 3 --N and NH 4 + -N in the soil.There was predominance of N in the nitric form (NO 3 --N) and these data are consistent with what is expected for aerated soils.Usually, the superficial soil layer has favorable conditions for the nitrification process, in which the ammoniacal N is transformed to nitric N (MAIA et al., 2008).Thus, the residues of leguminous plants caused the release of N, i.e., there was no limitation to the nitrification process.
Regarding K content (Table 7) except for leaves of M. hostilis and leaves + branches of M. hostilis, the studied residues have great potential regarding the release of K in the soil and G. sepium is the best species for this purpose.This is explained by the fact that G. sepium has higher K contents comparied to the others, in both leaves and branches (Table 1).Santos et al. (2010) also report that the K contents in the plots fertilized with G. sepium were higher than that found in the control during the first year of maize cultivation.
Under the conditions of the present study, leaves + branches of G. sepium caused higher Ca content, compared with the control.In relation to the other leguminous plants however, it differed statistically only from the leaves of M. caesalpiniifolia.The superiority observed in the release of this nutrient through the application of leaves + branches of G. sepium is explained by the fact that this species presented a higher Ca content and a lower C:N ratio (Table 1), favoring the decomposition and mineralization of this cation.Low efficiency of leguminous trees to recycle Ca of the soil, compared with the control treatment, was also reported by Nascimento et al. (2003).
The soil Mg content was affected by the addition of the residues of leguminous plants and the highest means were obtained by adding leaves of G. sepium and leaves + branches of G. sepium, which were the only treatments that led to means higher than the control (Table 4).There was a 44% increase in soil Mg, comparing the means of the treatment with leaves of G. sepium (highest mean) and the control.The effect of green fertilizers on the increase of Mg in the soil was also observed by Nascimento et al. (2003) and Faria et al. (2007).
With regard to Na, higher contents were observed in the treatments with branches of M. hostilis, leaves + branches of G. sepium and the control.These three did not differ statistically and were superior only to the application of leaves of M. caesalpiniifolia.According to the characterization analysis of the soil used, Na content increased in all treatments, since the initial content was 4.2 mmol c dm -3 and such an increase is related to the irrigation water.Regarding the difference between treatments, the highest Na contents were observed in the soil that received residues containing branches and in the control.Hence, soils with little or no cover favor higher evaporation of the irrigation water, thus increasing the capillary rise of salts (particularly Na) to the surface of the pots.
For the parameter sum of bases (SB), the addition of branches of G. sepium led to the highest mean, which was statistically equal to the treatments with leaves + branches of G. sepium and leaves of G. sepium and higher than the others (Table 7).The highest SB associated with the residues of G. sepium is consistent with the contents of K, Ca and Mg caused by the residues of this leguminous plant.The use of leguminous species capable of producing large amounts of residues favors reduction in the leaching of cations and increase in CEC, which is promoted by proportional increments in Ca, Mg and K contents and consequently in the sum of bases of the soil (TESTA; TEIXEIRA; MIELNICZUK, 1992).
Regarding cation exchange capacity (T), only the residues of G. sepium and branches of M. hostilis resulted in means higher than those of the control (Table 7).As mentioned for the sum of bases, the results of T are consistent with the observed contents of K, Ca, Mg and H+Al.As to the effective cation exchange capacity (t), the residues of G. sepium were the only ones to cause means higher than those of the control.
The results for T and t follow the same tendency, evidencing that the residues of G. sepium added to the soil led to the highest values of cation exchange capacity (CEC) at 65 days after application.According to Iwata et al. (2012), the CEC of soils with higher SOM contents, as occurred in the soil that received leaves + branches of G. sepium compared with the control, is also influenced by the high reactive power of the SOM, which is directly related to its various organic radicals.
For base saturation (V), there was no significant difference between the soil that received leguminous plants residues and the control.In this regard, Nascimento et al. (2003) found that the leguminous species pigeon pea (Cajanus cajan (L) Millsp.),brown hemp (Crotalaria juncea, L.) and velvet bean (Stizolobium aterrimum Piper & Tracy), despite promoting higher absolute values, did not differ significantly from an area without cultivation with respect to the effect on V.It is worth highlighting that high V is not always associated with fertile soil, as in the case of soils that have high Na contents due to irrigation with lower-quality water.This ion is not beneficial to the plants, but occupies the exchange site and is taken into account in the sum of bases, thus contributing to higher values of V.

CONCLUSIONS
Residues of M. caesalpiniifolia, M. hostilis and G. sepium positively influenced the growth of maize plants and favored the accumulation of macronutrients in the shoots, especially N, P and K at 65 days after applying the residues.
The improvements in soil chemical attributes caused by residues of M. caesalpiniifolia, M. hostilis and G. sepium demonstrate that these species are good options for recovering degraded areas in the semi-arid region of the Ceará state.

REFERENCES
Table 7. Mean values of the contents of phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), sodium (Na), sum of bases (SB), cation exchange capacity (T), effective cation exchange capacity (t) and base saturation (V) in the first 10 cm of the soil after maize cultivation with addition of residues of leguminous trees.
Means followed by the same letter in the column do not differ by Tukey test at p < 0.05; **Significant at p < 0.01; *Significant at p < 0.05; nsNot significant at p < 0.05; (L) = Leaves, (B) = Branches and (L + B) = Leaves + branches; CV = Coefficient of variation.