Patreze & Cordeiro : Nodulation , arbuscular mycorrhizal colonization and growth of some legumes native from Brazil

(Nodulation, arbuscular mycorrhizal colonization and growth of some legumes native from Brazil). The effects of rhizobial and mycorrhizal inoculation, nitrogen and phosphorus fertilization on nodulation, mycorrhizal colonization and initial growth were examined in Brazilian native plants, Enterolobium contortisiliquum (Vell. Conc.) Morong, Inga laurina (Sw.) Willd., Lonchocarpus muehlbergianus Hassl and Platypodium elegans Vogel. The experiment was carried out in a glasshouse using plastic bags filled with a mixture of sandy soil from riparian forest and vermiculite (2:1) amended with basal nutrients including NP, P and N and infected with rhizobia (r), mycorrhiza (m) or both (rm), amounting seven treatments: NP, P, P+r, P+rm, N, N+m and N+rm, with ten replications each. The plants were analyzed at 120 and 255 days after sowing. P deficiency negatively affected growth and nodulation of all species. Autochton arbuscular mycorrhizal fungi (AMF) colonized host roots and fungal inoculations did not enhance mycorrhizal colonization, which was also favored by added P. Nodulation was relatively higher in E. contortisiliquum and L. muehlbergianus, mainly in treatments containing P, and with rhizobial inoculation (P+r). Plants from these treatments developed better than others and, despite the rhizobia inoculated had no synergistic relationship with inoculated AMF, they also showed the best percentages of mycorrhizal colonization. Moreover, these two species showed highest rates of acetylene reduction and highest leghemoglobin content. These results suggest that E. contortisiliquum and L. muehlbergianus can have advantages for establishment in soils with low nitrogen levels.


Introduction
Dual inoculation with rhizobia and mycorrhizal fungi is currently being suggested as a possible solution to reforestation and amendment of soil fertility.The effectiveness of rhizobia-mycorrhiza-plant interactions varies with host species, rhizobial strain, fungus species and soil conditions.In contrast to taxonomically restricted N 2 -fixing rhizobia, arbuscular mycorrhizas are extremely widespread (Brundrett 2002), besides showing low specificity (Mosse 1975).
Inga laurina (Sw.)Willd. is generally found in riparian forests primarily in seasonally flooded areas (Romagnolo & Souza 2000), but it has also long been used as a shade tree and green manure in coffee and cacao plantations and has good potential as an agroforestry species (Tilki & Fisher 1998).It was listed as a nitrogen-fixing tree by Halliday & Nakao (1982) and under its currently accepted synonym I. fagifolia, (L.) Benth.was reported to be nodulated (Faria et al. 1987).However these studies did not include any information on nodule physiology.AMF colonization in I. laurina has not been reported previously, although other Inga species are known to form such symbioses.
Lonchocarpus muehlbergianus Hassl has been used in carpentry and as an ornamental species (Lorenzi 1992).The genus includes 130 species of shrubs or trees from tropical and subtropical America and one species in Africa.L. muehlbergianus is one of 21 species from this genus with known nodulation capacity (Sprent 2001).Faria et al. (1987) were the first to observe its nodulation.This species can be founded in riparian forest (Silva et al. 1992;Bernacci et al. 1998;Nilsson 1989).Frioni et al. (1999) observed colonization by AMF in L. nitidus.
Platypodium elegans Vogel is used for wood and as an ornamental, mainly along the streets and avenues (Lorenzi 1994).This species was reported as nodulated by Halliday & Nakao (1982).It was sampled in 23.9% of studies involving floristic survey of riparian forest of Brazil (Rodrigues & Nave 2000).Carneiro et al. (1998) characterized the occurrence of AMF of P. elegans in a nursery.
The aim of this work was to investigate the effects of inoculation with rhizobia and/or arbuscular mycorrhizal fungi on nodulation, mycorrhizal colonization and growth of E. contortisiliquum, I. laurina, L. muehlbergianus and P. elegans under greenhouse conditions and different fertilization treatments.

Materials and methods
Seeds of all species were surface sterilized.E. contortisiliquum and P. elegans seeds were scarified to break dormancy prior to planting.E. contortisiliquum seeds were smoothed and P. elegans seeds were left to soak in 50% sulphuric acid during 45 min and washed nine times prior to planting.I. laurina seeds were firstly surface sterilized and stored in trays with vermiculite and abscisic acid solution (10 -4 M) in the proportion 2:1 v/v in a cold chamber at 10±1 ºC and 85%±5% of relative humidity (Barbedo & Cicero 2000) during twenty-four days, since seeds of Inga species rapidly lose viability under normal environmental conditions.After that, the seeds of these species and L. muehlbergianus were germinated directly on substrate.
Seeds of all species, except I. laurina, were soaked in a turbid suspension (100 mL) of a mixture of two rhizobia strains for one hour before sowing.E. contortisiliquum seeds were inoculated with a mixture of strains IBRC-206 and 207, isolated from E. contortisiliquum nodules.L. muehlbergianus and P. elegans seeds were inoculated with IBRC-208 and 209 rhizobia strains isolated from nodules of L. muehlbergianus.Pre-germinated seedlings of I. laurina received 10 mL of a mixture of  strains isolated from Inga spp., directly on surface of pots, one week after sowing.To ensure rhizobia infection, seedlings of all species were reinoculated 30 days after sowing, with 10 mL of rhizobia suspension containing the same strains.
Aerial parts of this species were catalogued in the Bioscience Institute Herbarium -UNESP under the number HRCB-34330.
The plant height was measured at two weeks intervals from 30 days after sowing until the end of the experiment (255 days after sowing).Five plants per treatment were harvested 120 days after sowing and fresh nodules were sieved and counted those larger than 4 mm, between 2-4 mm and less than 2 mm of diameter.The other five plants per treatment were harvested 255 days after sowing, for leaf area measurement (CI-202 Area Meter, CID, Inc), dry weight of roots, stems and leaves, AMF colonization and nitrogen and phosphorus content of shoot tissues and substrate.
Nitrogenase activity and leghemoglobin content were evaluated in the nodules of two plants per treatment.Nitrogenase activity was measured by the acetylene reduction activity (ARA; Hardy et al. 1968) using the whole root system, and leghemoglobin content by the method of Becana et al. (1986) at 540 nm, using three samples of 0.5 g each.After that, fresh nodules were sieved to separate those larger than 4 mm, between 2 and 4 mm and less than 2 mm of diameter.They were then counted and dried.Nodule morphology was classified according to Sprent (2001).
Roots were stained (Philips & Hayman 1970) and percentage of AMF infected roots of two plants per treatment was estimated using the gridline intersects method (Giovannetti & Mosse 1980) under a stereomicroscope (40x) at 255 days after sowing.
Data were analyzed separately by ANOVA and means were compared by Duncan's test, at P ≤ 0.05, using the Statistica for Windows (StatSoft, Inc. 2000), except for P and N contents that were analyzed on combined samples (per treatment), AMF colonization, ARA and leghemoglobin content that were evaluated in two plants per treatment.

Results
Plants that received NP were the tallest (P<0.05) in relation to other treatments, mainly N treatments (phosphorous deficiency) (Fig. 1).E. contortisiliquum plants cultured on P and P+r treatments were as tall as the NP plants at 255 days (Fig. 1).Plants of this species receiving P+r treatment had highest leaf area.The dry mass of leaf, stem and root were also high, but did not differ of NP ones (Tab.1).For I. laurina, N, N+m and N+rm treated plants had leaf area similar to NP.For P. elegans, NP treatment differed significantly from the others in all growth parameters.P-fertilized treatments showed the best results of plant growth at 255 days in E. contortisiliquum and L. muehlbergianus (Fig. 1).Rhizobia inoculation in P-fertilized plants increased leaf area, leaf, stem and root dry mass of L. muehlbergianus at 255 days after sowing (Tab.1).
Mycorrhizal colonization was relatively low in all species (Tab.1).Percentage of AMF colonization of roots reached 24.72; 45.25; 39.48 and 15.35% in E. contortisiliquum, I. laurina, L. muehlbergianus and P. elegans, respectively (Tab.1).However, the highest values of AMF colonization occurred in P treatments inoculated with rhizobia, therefore non-inoculated with AMF in E. contortisiliquum and L. muehlbergianus.This indicates higher potential of infection by native fungi in the soil used as substrate and it becomes evident that the P fertilization is important for the  All nodulated species and nodules were identified as indeterminate (Corby 1988;Sprent 2001).We consider E. contortisiliquum and L. muehlbergianus as astragaloids (or caesalpinoids), I. laurina and P. elegans as mucunoid and aeschynomenoid respectively.All species, except P. elegans, were well nodulated and the nodules had a reddish internal colour (Fig. 2 D-F).Most part of nodules was between two and four mm in diameter (Fig. 2 and 3).Nodulation was suppressed by N-fertilizer and/or P deficiency (N, N+m and N+rm).The highest numbers and dry weight of nodules were observed in E. contortisiliquum, I. laurina and L. muehlbergianus plants grown in P, P+r and P+rm treatments (Fig. 3 and Tab. 2), showing the success of rhizobia inoculation in these plants.
Nitrogenase activity (ARA) and leghemoglobin content were higher in E. contortisiliquum and L. muehlbergianus than I. laurina and P. elegans (Tab.2).Due to the low numbers and small size of nodules from P. elegans, we were unable to obtain accurate data on nodule mass, nitrogenase activity and leghemoglobin content.

Discussion
Phosphorus nutrition appeared to be the major factor influencing growth and nodulation of these species.Low phosphorus availability is the limiting factor for plant growth in more than 60% of tropical soils and this affects particularly symbiotic legumes, which usually have higher P requirements than nonsymbiotic plants (Vadez et al. 1997), partly associated with their N 2 fixing capability (Frioni et al. 1999).Similar P effects have been reported in Leucaena leucocephala (Lam.) de Wit. by Brandon et al. (1997).In our experiments, E. contortisiliquum and L. muehlbergianus grew well in the presence of  .Average number of nodules per diameter size class (larger than 4 mm, between 2 and 4 mm and smaller than 2 mm) of Enterolobium contortisiliquum (Vell.Conc.)Morong, Inga laurina (Sw.)Willd.and Lonchocarpus muehlbergianus Hassl seedlings in response to treatments of mineral fertilization (NP, P, N) and inoculation with rhizobia (r), mycorrhiza (m) or both (rm) at 120 and 255 days of sowing.Total number of nodules followed by the same letter is not significantly different by ANOVA, followed by Duncan's test at P<0.05.£ = < 2; ¢ = 2 a 4; ¢ > 4. As demonstrated by Siqueira et al. (1998) in a study comprising 28 native species in woodland fragments in the state of Minas Gerais, Brazil, the response to treatments of P fertilization and mycorrhiza inoculation diminish with advance in succession.This agrees with our results, where I. laurina and P. elegans, late secondary or climax species (SMA, 2001), were less responsive to the P treatments than the pioneer trees E. contortisiliquum and L. muehlbergianus.In addition, under nutrient stress conditions mycorrhizal effect on initial seedlings growth is greater in small seeded species (Siqueira et al. 1998).
The increasing concentration of soluble phosphate in the soil normally decreases the percentage colonization of roots by AMF (Smith & Read 1997).In our experiment, the mycorrhizal colonization was relatively low, but the highest values were found in the treatments with higher P level.Moreover, mycorrhizal fungi colonized the seedlings of almost all treatments.This confirms that AMF are an integral part of the root system of almost all plants in the tropics, not only under natural conditions, but also in cultivation.Bereau et al. (1997) recorded 68 and 30% of endomycorrhizas in Inga sp. in two of 12 sites of the forest of French Guiana.Carneiro et al. (1998) characterized the occurrence of AMF of P. elegans in the greenhouse as an average of 20 to 49% of infected roots.Frioni et al. (1999) reported that E. contortisiliquum, I. uruguensis and L. nitidus had levels of colonization by AM fungi of 60, 47 and 69%, respectively, in the field.
Due to the low AMF colonization in our experiment, we could not observe the effects of the increase of P acquisition by plants associated with mycorrhizal fungi, as discussed by Colebatch et al. (2002).Treseder & Allen (2002) proposed a model where fungi are more efficient scavengers for soil nutrients than plant roots.Nevertheless the time of observation could be extended for some years, once the species studied are perennial trees and 255 days could be short time to observe the AMF infection, colonization and efficiency in the root system.
In our experiment nodulation was high, except for P. elegans, where average numbers of nodules were 4.33 and 0.66 for P and P+r treatment, respectively, at the second harvest (data not shown).According to The N 2 fixing capacity of E. contortisiliquum, I. laurina and L. muehlbergianus was confirmed by the ARA assay.Some values of ARA were reported in the literature for species belonging to the same genera: 2.800 (in µmol ethylene.g - fresh nodule.h - ) for I. uraguensis Hook & Arn.(Frioni et al. 1998) and 0.098, 0.18, 0.09, 0.06, 0.1 and 0.016 (in µmol ethylene nodule.h - ) for E. gummiferum (Mart.)Macb., E. mongollo Mart., I. capitata Desv., I. luschnathiana Benth., I. nuda Salzm.ex Benth., I. sessilis (Vell.Conc.)Martius, respectively (Faria et al. 1984).The units are different on each case: in g of nodules and per nodule.It is difficult to compare these data firstly because they are from different species and secondly because ARA can be influenced by season, soil temperature drought and photoperiod (Moro et al. 1992).Furthermore, acetylene reduction assay can only indicate relative differences in nitrogenase activity; the assay does not measure total nitrogenase activity because the assay itself causes a decline.Nevertheless it can be useful in measuring relative differences in N 2 fixation potential with low levels of technology (Vessey 1994), as used in our experiment to compare the fertilization and inoculation treatments or to compare species, as long as the nodules must have the same type of growth.
Although in our experiments nodulation in P. elegans was low, this species is known to fix nitrogen (Tab.2).Goi et al. (1984) detected acetylene reduction activity of 9.2 µmol C 2 H 4 per nodule.h - and Faria et al. (1984) obtained 180 µmol C 2 H 4 per nodule.h - in seedlings inoculated by a mixture of 222 strains of rhizobia.
The low nodulation in P. elegans may be associated with lack of specificity between host plant and rhizobia, since the strains inoculated were isolated from L. muehlbergianus.It could also be related to P deficiency, which can limit nodulation indirectly by reducing legume growth, rather than by direct action on the infection process (Bordeleau & Prévost 1994).Thus, the slow growth of P. elegans and the low values of AMF colonization and nodulation (number of nodules, ARA and leghemoglobin content) show that this species must be more studied in relation to mineral nutrition.
The extent of root colonization cannot be an indication of AMF's ability to enhance plant growth as observed by and according to Roldan-Fajardo (1994) and Marques et al. (2001).An effective mycorrhizal colonization may affect nodule weight, leghemoglobin content and nitrogen fixing activity.The growth and development of seedlings in the greenhouse conditions with the use of non-sterilized soil may require only rhizobial inoculation, as found here for E. contortisiliquum, P. elegans and L. muehlbergianus plants that can form mycorrhizas with soil fungi.However, better results might be obtained with the use of selected mycorrhizal fungi.Our results support the evaluation of the effects of dual inoculation of these legume tree species under field conditions, as a way of enhancing plant survival.
treatment when dependent on biological nitrogen fixation by rhizobia (no N added), since N content in shoot tissue was high.
AMF colonization occurred with both the inoculated and native fungi and it is the first register of AMF colonization in this species.
*Values followed by the same letter(s) in the columns are not statistically different at 5% level determined by Duncan test.