Systems of land use affecting nodulation and growth of tree legumes in different soils of the Brazilian semiarid area

The growth of tree legumes in degraded areas must be preceded by assessments of nodulation ability of naturally established rhizobia populations since such information contributes to defining the species which can be planted for recovering disturbed areas. The aim of this study was to evaluate the growth and natural nodulation of “sabiá” (Mimosa caesalpiniifolia Benth.) and leucaena (Leucaena leucocephala (Lam.) de Wit.) in soils of the Brazilian semiarid area under different systems of land use: native vegetation (locally called caatinga) and areas with different agricultural systems (a monocrop system and an intercropping with various species). For each species, a greenhouse experiment in randomized block design was realized, using soils of different types (Luvisol and Ultisol), with 4 replicates. The results evidentiate significant differences in the evaluated growth characteristics (height, leaflet number and shoot diameter) of M. caesalpiniifolia, that have displayed lower plant growth when cultivated in the Luvisol under conventional system. Plant growth, nodulation and total N accumulation in both seedling tree legumes increased in Ultisol under the different systems of land use. L. leucocephala showed higher potential of biological nitrogen fixation and nodulation effectiveness promoted by indigenous rhizobia.


INTRODUCTION
In tropical regions, the predominant agricultural production systems are based on the conversion of native forests to croplands, with cutting and burning of native vegetation, exploration and subsequent abandonment (fallow), before again clearing and burning cycle.In the semi-arid region of Brazil, the native vegetation (caatinga), which would cover an estimated area between 6.09 × 10 5 km 2 (Sampaio, 1995), is also a part of the shifting cultivation cycle, besides being the main form of native pasture for the extensive livestock farming in the region.Some of these uses over the centuries left the native vegetation degraded, with wide stretches in the process of desertification.Presently, there are less than 50% of the original vegetation (Menezes et al., 2012).The introduction of leguminous trees is considered one of the main practices that can be employed to recover degraded areas (Pereira and Rodrigues, 2012).The cultivation of these species promotes protection against soil erosion (Garba and Dalhatu, 2015) and improves the soil fertility by the addition of organic matter (Wu et al., 2016).However, the main characteristic of some legume species is their ability to establish symbiotic associations with bacteria that fix nitrogen (Moreira and Siqueira, 2006;Sprent, 2007) and the success of the integration of certain species depends on their nodulation with effective rhizobia.Thus, the use of tree legumes must be preceded by assessments of growth and nodulation ability of such species in association with the native rhizobial populations.This information will contribute to the definition of which species can be planted for the recovery of degraded areas.
The nodulation and the efficiency of biological nitrogen fixation process can be restricted by many conditions related to the plant, to microsymbiont and the climate and soil conditions.Of course, in the absence of native rhizobia populations, the symbiosis will not be established.Generally, microsymbionts populations are abundant in soil of the region that the legume species are native (Bala et al., 2003).But it may be that even in the presence of compatible rhizobia populations, the symbiosis is not efficient (Faye et al., 2007).The growth of rhizobia in free life in the soil and its ability to nodulate and fix nitrogen in symbiosis with the legume are sensitive to environmental conditions and can be dependent on soil quality.Different vegetation cover or managements affect the diversity of rhizobia (Jesus et al., 2005;Boakye et al., 2016), and may favor, differently, more or less efficient populations.
"Sabiá" (Mimosa caesalpiniifolia Benth.) is a small tree legume native from the Brazilian semiarid region which have great social and economic importance especially in function of the use in the production of fences, firewood and coal, and still be used for animal feed, due to the high nutritional value (Ribeiro et al., 2008;Costa Filho et al., 2013).Leucaena (Leucaena leucocephala (Lam.) de Wit) is a perennial tree legume distributed throughout the tropical region (Aquino, 2011), and have multiple uses as application in soil improvement, shading, windbreak and has been widespread in the tropics (Barreto et al., 2010).Leucaena and "sabiá" represent tree legumes with potential to be used in the recovery of degraded areas in da Silva et al. 3967 semiarid regions, especially due to their fast growth characteristics (Amaral et al., 2016), high biomass production (Moura et al., 2006) and ability to establish effective symbiotic associations with specific rhizobia strains (Reis Junior et al., 2010).Before these considerations, the aim of this study was to evaluate the growth and natural nodulation of "sabiá" and leucaena in soils that were originally covered by caatinga and currently are under different systems of land use.

MATERIALS AND METHODS
Soil samples from the 0 to 20 cm superficial layer were collected in areas under diferent systems of land use in two municipalities, with different climate conditions and soil type (Table 1): Belo Jardim, in the Agreste zone, and Serra Talhada, in the Sertão zone.Each municipality area with native vegetation (caatinga) and areas with different agricultural systems (a monocrop system and an intercropping with various species) were selected.In both municipalities, the systems of land use were chosen in areas with the same soil type.
In the six areas, four plots with 10 x 10 m were established.In each plot, five sub-samples of soil were collected.The subsamples from the same plot were mixed to obtain composite samples to represent the treatments with different systems of land use.Physical and chemical analysis of soils samples (Tables 2 and 3) were realized following Embrapa (2011).Each treatment was sampled four times and correspond the four replicates used in the greenhouse experiment.The experiment was realized using a randomized block design with four replicates.To each of the two tree legume species, the experimental units consisted of three pots (one plant per pot) containing the soil samples collected in the different areas.Seeds of Mimosa caesalpiniifolia Benth.and Leucaena leucocephala (Lam.) de Wit. were surface disinfected in ethanol (70% v/v-3 min) and sodium hypochlorite (1% v/v-3 min), rinsed five times with sterile distilled water, rolled onto YMA plates to test for surface sterility and then were sown in the pots.The pots received destiled water until harvest (90 days after seed germination).Plants were harvested at 90 days after seed germination and determined the nodules number and the dry biomass of shoots and roots, after drying in an oven at 65°C, for 72 h.To determine the shoot biomass, the samples were passed in Willey type mill and subsequetly quantified the total N content, according to Embrapa (2011).
The results of plant height, stem diameter, number of leaflets, dry biomass of shoots, roots and nodules, nodules number and total N accumulation in shoots were submitted to analysis of variance by F *Corresponding author.E-mail: vinicius.agro2008.1@gmail.com.Tel: 55-81-33206237.Fax: 55-81-33206200.

RESULTS AND DISCUSSION
The occurrence of compatible microsymbionts populations was demonstrated by the natural nodulation of the tree legume species in both soils.The "sabiá" displayed a more abundant nodulation in the Ultisol; however, in the Luvisol, the nodules were greater.The plants did not show difference in the total biomass of nodules when comparing the two soils.The complexity of the different systems of land use did not influence natural nodulation of "sabiá".Leucaena produced more nodules when grown in the Ultisol under the conventional agriculture, and it may be observed that the nodules were small and the total biomass of nodules differed only in plants grown in the Luvisol submitted to the intercropping system (Table 4).
The soil samples collected in different systems of land use did not show limitations to nodulation with regards to soil acidity and the occurrence of exchangeable aluminum (Table 3).In spite of the higher mean annual precipitation in Belo Jardim than in Serra Talhada, the Agreste sites have higher water availability throughout the year than the Sertão site.Rainfall is concentrated mostly in three months (February to April) in the Sertão mesoregion (average of 73% of the annual rainfall) and the month of highest rainfall represents 41% of the annual rainfall.In contrast, in the Agreste mesoregion, rainfall is reasonably well distributed over five months, from March to July (67% of the annual rainfall), and the month of highest precipitation represents only 16% of the annual rainfall.In addition, there is a large interannual variation in total rainfall in the Sertão (Freitas et al., 2010), resulting in an decrease in survival of natural rhizobia populations established in Serra Talhada, which could explain the different nodulation observed for the two legumes grown in the two used soils.
The results of nodulation contrasted with data from other studies that compared the ability of two legume nodulation (Souza et al., 2007), since in the present study, there was a higher nodulation of leucaena.The species showed high specificity to the microsymbionts (Silva et al., 2009) and this fact may be related to native rhizobial populations prevailing in the areas and the ability of these native rhizobia to establish effective symbiotic associations.
Recent studies have shown that tree legumes have higher affinity for different proteobacteria groups and the "sabiá" is more associated with beta-rhizobia (Reis Junior et al., 2010;Martins et al., 2015), while leucaena tree legume nodulated predominantly with alpha -rhizobia (Wang et al., 1999).It is possible that different legumes and the simbionts partners have influenced the natural nodulation.However, probably is necessary for more refined studies regarding the ecology and diversity of nodulating rhizobia species in the Caatinga biome to explain the specific nodulating behavior among different legumes The height constituted an important variable in predicting the development of plants, being technically accepted as important measure for assessing the performance potential of the seedlings growth (Dutra et al., 2015).For tree legume species, the seedlings are ready to be planted in the field when they have the height between 15 and 30 cm (Paiva and Gomes, 2000).Thus, in this study, 30 days for all soils tested "sabiá" and leucaena are included in the above range.Therefore, according to this criterion, the seedlings of the two legumes would be suitable for transplantation to the field 30 days after sowing.
Seedlings of "sabiá" and leucaena developed normally with 100% of survival.The legumes seedlings grown in the Ultisol were higher than those grown in the Luvisol, especially when agricultural land use system was used (Figure 2).In soils derived from the land use system, "sabiá" seedlings reached height of 65.9 cm as compared to soil from Caatinga and agriculture (Ultisol) and lower height was observed in the Luvisol with the agriculture land use system.In leucaena grown in two soils, influence in height with the different land use system was not observed and varied from 68 to 90 cm (Figure 1).
The stem diameter is a very important variable in survival potential assessment studies and growth after planting (Souza et al., 2006;Cruz et al., 2012).In "sabiá", the stem diameter showed a similar response as obtained in plant height, and also promoted lowest development in the Luvisol submitted for agricultural land use system (Figure 2).In leucaena, the stem diameter was not influenced by the land use system, displaying an average diameter of 4.7 mm.However, the value of stem diameter reflects the actual standard quality of seedlings for transplanting to field conditions that depends on the species, the location sites and the techniques of production (Gomes et al., 2002).
The number of leaflets of the two species was influenced by land use system.For "sabiá", the lowest leaflets number was observed in the Luvisol under agriculture cultivation and the largest values obtaianed in the Ultisol under Caatinga land use system.For leucaena, there was a greater number of leaflets in the plants grown in the Luvisol submitted to the Caatinga land use system (Figure 3).
The dry biomass of shoots and roots of leucaena were quite similar in the different land use systems, with an average accumulation of 6.3 and 2.4 g for shoots and roots, respectively.For "sabiá", there was a lower biomass production of shoots in the Luvisol (Table 5).The highest values were obtained in agriculture land use system in plants grown in the Ultisol.When comparing the two legumes, it appears that leucaena showed a shoot biomass and root biomass higher than those found for "sabiá".These results were different when compared with those obtained by Souza et al. (2007) who found greater accumulation for "sabiá" biomass in shoots and roots when compared with leucaena tre legume.
The seedlings of leucaena accumulated more total N in the plant biomass.For the two legumes, the nutrient accumulation has been in greater quantity when used the Ultisol.In "sabiá", the best total N accumulation was observed in plants grown in soil under agriculture cultivation that present 36 mg plant -1 . When grown in soil of Belo Jardim under caatinga, leucaena showed the best total N accumulation that corresponded to 37.2 mg

plant
-1 (Table 5).The low aboveground biomass and total N accumulation in "sabiá" grown in the Serra Talhada soil were a result of the low nodulation in such areas.Commonly, nitrogen is the nutrient required in high amount by plants and is the nutrient that limit productivity in semiarid regions (Freitas et al., 2015) and is important to promote efficient management to optimize the FBN process.

Conclusions
In the seedling stage, the growth, development, nodulation and accumulation of total N in "sabiá" and leucaena were favoured in the Ultisol.
Leucaena showed efficient nodulation and greater potential for N 2 fixation in symbiosis wtih natural rhizobia populations established in the systems of land use B A studied.

Figure 1 .
Figure 1.Height of "sabiá" (A, Mimosa caesalpiniifolia) and leucaena (B, Leucaena leucocephala) tree legumes 90 days after cropping, in soils collected in the different land use systems, in the Brazilian semiarid region.Means with the same letter are not different by the Tukey test (p < 0.05).

Figure 2 .
Figure 2. Shoot diameter of "sabiá" (A, Mimosa caesalpiniifolia) and leucaena (B, Leucaena leucocephala) tree legumes 90 days after cropping, in soils collected with different land use systems in the Brazilian semiarid region.Means with the same letter are not different by the Tukey test (p < 0.05).

Figure 3 .
Figure 3. Number of leaflets for "sabiá" (A, Mimosa caesalpiniifolia) and leucaena (B, Leucaena leucocephala) tree legumes 90 days after cropping, in soils collected in the different land use systems, in the Brazilian semiarid region.Means with the same letter are not different by the Tukey test (p < 0.05).

Table 1 .
General characteristics of the municipalities of Belo Jardim, in the Agreste mesoregion, and Serra Talhada, in the Sertão mesoregion, semiarid of Pernambuco State, Brazil.

Table 2 .
Physical analyses and textural classification of the used soils submitted to different systems of land use in the Brazilian semiarid region.

Table 3 .
Chemical analysis of the soils collected in the different land use system in the Brazilian semiarid region.

Table 4 .
Number of nodules and dry biomass of nodules in seedlings of "sabiá" (Mimosa caesalpiniifolia) and leucaena (Leucaena leucocephala) in soils colleced in different land use systems, in the Brazilian semiarid region.
Means with the same letter are not different by the Tukey test (p < 0.05).

Table 5 .
Dry biomass of shoots, dry biomass of roots and total N accumulation in shoot biomass in "sabiá" (Mimosa caesalpiniifolia) and leucaena (Leucaena leucocephala) tree legumes grown in soils collected in the different land use systems in the Brazilian semiarid region.Means with the same letter are not different by the Tukey test (p < 0.05).