Mycorrhizal dependency of mangaba tree under increasing phosphorus levels

The objective of this study was to evaluate the mycorrhizal dependency of mangaba tree (Hancornia speciosa) plantlets, under increasing levels of phosphorus fertilization. The experimental design was completely randomized in a 4×5 factorial arrangement with three mycorrhizal fungi inocula – Gigaspora margarita, Glomus etunicatum, or a pool of native mycorrhizal fungi (Acaulospora longula, Glomus clarum, Gigaspora albida, Paraglomus sp.) –, and a nonmycorrhizal control, in combination with five levels of phosphorus applied to the substrate: 0, 25, 50, 75, and 100 mg kg. After 180 days of growth, plantlets with inoculation of native mycorrhizal pool produced more shoot and root dry biomass and had higher shoot phosphorus content and accumulation. The noninoculated control showed the lowest values, independently of the phosphorus level. The highest relative mycorrhizal dependency occurred with the inoculation of native mycorrhizal fungi. Plants with mycorrhizal fungi did not respond to phosphorus addition above 50 mg kg. Mangaba tree is highly dependent on mycorrhiza, but the degree of dependency varies according to phosphorus levels and fungal inocula. In general, mangaba tree is more responsive to mycorrhizal fungi inoculation than to phosphorus addition.


Introduction
The mangaba tree (Hancornia speciosa Gomes) (Apocynaceae) is a native Brazilian fruit, which occurs naturally on low-fertility soils in the coastal areas and in the biome Cerrados (Aguiar Filho et al., 1998).Due to increasing deforestation and urbanization of coastal areas, this species has had its natural germplasm bank severely reduced.In Northeastern Brazil, an important pulp market has developed based on exotic tropical fruits, including mangaba.Propagation of elite mangaba trees is desirable to increase the productivity in orchards, but the use of seeds or vegetative methods is very laborious (Parente & Carmona, 1988).In addition, very few plants survive in the nursery and in the field, after transplanting, when the mortality may reach 60% (Sano & Fonseca, 2003).
It has been reported that dependency on mycorrhiza varies between and within plant species (Cavalcante et al., 2001;Zangaro et al., 2007) and those with shorter root length and greater diameter usually benefit more from mycorrhizal symbiosis (Zangaro et al., 2005).Morphological root traits, such as root geometry, rate of root growth, density and length of root hairs have often been used as indicators of mycorrhizal dependency (Zangaro et al., 2007).
Inoculation with arbuscular mycorrhiza (AM) seems to be an interesting tool to select elite mangaba genotypes and to help in the domestication and improvement of recalcitrant species, given that H. speciosa has shown to be highly responsive do AM, mainly in low-fertility soils (Costa et al., 2005).Thus, it is important to know the dependency of H. speciosa on AM, for seedlings production purposes and for the use of effective AM inocula during the nursery phase.Arbuscular mycorrhiza can reduce the need for fertilizers and anticipate the time for transplantation of plantlets produced in nurseries (Costa et al., 2003).In addition, more vigorous plantlets could be obtained, increasing the capacity to withstand the stress caused by transplantation from the nursery to the field.
The objective of this study was to evaluate the mycorrhizal dependency of mangaba tree (Hancornia speciosa) plantlets, under increasing levels of P fertilization.
Seeds from mature mangaba fruits were collected from native trees at coastal areas with "restinga" vegetation (called "tabuleiros" in the Brazilian Northeast), and were surface-sterilized with 2.5% (v/v) sodium hypochlorine for 20 min.Seeds were, then, rinsed in distilled sterile water and germinated in sterile Petri dishes with moistened filter paper in the dark at 28 o C. Five days after germination, seedlings were moved to asbestos trays containing autoclaved river sand, and were kept in greenhouse for more 17 days, when they were transplanted to the pots.
Polyethylene pots (5 dm 3 ) were filled with 4 kg of the mixture and received dolomitic lime (1.58 g per pot) in order to raise the pH (0.01 M CaCl 2 ) to 6, based on incubation method.Phosphorus was mixed into the soil as simple superphosphate (8.3% P; 18.8% Ca; 12.6% S) at four different rates corresponding to 25, 50, 75 or 100 mg kg -1 of P, besides a 0 P treatment.Afterwards, all pots received (mg kg -1 ): N (80 as NH 4 NO 3 ), K (150 as K 2 SO 4 ), Zn (2 as ZnSO 4 .7H 2 O), B (0.5 as H 3 BO 3 ) and Mo (0.1 as H 2 MoO 4 ).Nitrogen (NH 4 NO 3 ) was supplied again, as solution, with 20 mg per pot, for four times, during plant growth in greenhouse: at transplanting and 30, 60 and 90 days after; total additional N was 80 mg per pot.
Uniform plantlets with two pairs of true leaves, and with 10 cm high, were transplanted to the pots containing previously fertilized soil.
The AM inocula in pure culture (G. margarita or G. etunicatum) were multiplied for 120 days in pots grown with Panicum maximum Jacq., and consisted of 50 g of sandy soil containing approximately 1,650 spores of each AM fungus.Native AM inocula were taken from natural sites at "restinga" vegetation and were also multiplied in P. maximum as trap plant.After multiplication, the density of native spores was about 1,800 in 50 g of soil, with prevalence of A. longula, about four times higher than the other AM species (G.clarum, G. albida and Paraglumus sp.), which, in turn, occurred in almost the same proportion.The mycorrhizal inocula (50 g) were placed in a band approximately 5 cm below the soil surface, while the nonmycorrhizal control received 50 g of soil grown with P. maximum without AM fungi.
The experiment was conducted under greenhouse, at Universidade Federal de Alagoas.Noncontrolled temperature and relative humidity were recorded by a digital thermohygrometer; minimum and maximum temperature values ranged from 21.4 o C (night) to 37.9 o C (day); irradiance was 1,200 µmol m -2 s -1 ; relative humidity ranged from 60 to 90%; and the day length was 13 hours.The soil moisture was controlled by weighing and daily irrigation with distilled water, in order to maintain 80% of the water holding capacity.
Plants were harvested 180 days after transplanting, the shoots and roots were collected and the dry biomasses weighed, after drying at 70 o C for 48 hours.Relative mycorrhizal dependency (RMD) was calculated according to Plenchette et al. (1983): [(M-NM)/ M] x 100, in which: M is the total dry biomass of mycorrhizal plants; NM is the total dry biomass of nonmycorrhizal plants.The degrees of RMD were classified according to Habte & Manjunath (1991).Roots were carefully washed in tap water and cut into 1 cm long segments.A sample of 1 g of fresh root of each plant was clarified with KOH (Phillips & Hayman, 1970), and stained with acid fuchsin (Kormanik & McGraw, 1982).The percentage estimation of the mycorrhizal root colonization was based on the gridline intersection method (Giovannetti & Mosse, 1980).Phosphorus concentration in shoots was determined colorimetrically (Murphy & Riley, 1962) in nitric-perchloric acid digests.
Data were submitted to the analysis of variance, using the general linear models procedure of the SAS, followed by regression analysis (SAS Institute, 1996).Percentage values of mycorrhizal root colonization were transformed to arc sin (x) 0.5 prior to analysis.

Results and Discussion
Shoot and root dry biomasses increased due to mycorrhizal colonization at all P levels, mainly in plants with inoculation of native AM pool (NP) (Figure 1 A and B).The lowest dry biomasses for both shoot and roots were found in the nonmycorrhizal control plants (Co), irrespective to P levels.In plants infected with the native AM pool, maximum shoot and root biomasses were reached above 50 mg kg -1 of added P (around 65 mg kg -1 of P), the maximum point described by the quadratic regression adjustment, followed by a decrease thereafter.Considering the nonmycorrhizal plants, the increase of P levels in the soil increased only slightly the shoot and root dry biomasses, contrary to the observed for mycorrhizal plants.While plants infected with G. etunicatum and G. margarita responded at an intermediate level to P addition, the native mycorrhizal pool caused a more intense plant response to P. Previous report showed that increase of P levels in the soil suppressed mangaba plantlets response to Gigaspora albida, while G. etunicatum showed no effect at any P level ranging from 3 to 183 mg dm -3 (Costa et al., 2005).
In the present work, shoot dry weight increased up to 50 mg kg -1 , depending on the mycorrhizal inoculum.The behavior of root dry biomass was similar to the reported for shoots, except for G. etunicatum which increased root biomass to levels comparable to plants infected with the native mycorrhizal pool.This fact changed biomass allocation between shoots and roots: plants colonized by G. etunicatum showed a shoot to root ratio around 1, while in the other treatments, including control plants, the shoot to root ratio was about 2 (Figure 1 C).
The increase of soil P levels tended to increase the shoot to root ratio only in the control, whereas in the mycorrhizal treatments it remained more stable.Many researchers have shown benefits of inoculations with AM in perennial crops such as citrus (Nogueira & Cardoso, 2006), coffee (Tristão et al., 2006) and passion fruit (Cavalcante et al., 2001).In general, these effects are more evident when soil P level is deficient.Nevertheless, even with increasing P levels up to 100 mg kg -1 , in the present study, the plant growth response to AM was only slightly decreased.
Plants infected with the native AM pool (NP) experienced a shoot growth depression at the higher P levels.Even in the less effective symbioses (Ge or Gm), plants tended to show a growth depression, as illustrated by the quadratic regressions adjustments.Under excessive P, the symbioses may be impaired and induce growth depression in plant likewise in citrus (Nogueira & Cardoso, 2006).Although growth depression in citrus had been verified only at 1,000 mg kg -1 of P, mangaba plantlets exhibited this phenomenon just above 50 mg kg -1 , showing that the mycorrhizal symbiosis in mangaba is much more sensitive to P than in citrus.Such depression is generally attributed to the cost of the symbiosis, mainly with regard to the drain of carbohydrates from the host in a condition that the benefit from P uptake, mediated by the fungal partner, is lower than the C cost (Graham, 2000).To our knowledge, this is the first report on growth depression in mycorrhizal mangaba plantlets under excessive P.
The relative micorrhizal dependency (RMD) of mangaba plantlets varied with the AM inocula and P levels, but was considered very high at all P levels and mycorrhizal inoculum types (Habte & Manjunath, 1991), reaching more than 75% of dependency on AM (Figure 2).The RMD of mangaba in the treatments with pure inocula ranged from 78 to 85%, while the RMD with native AM pool was far above, around 93%.When inoculation was with native fungal pool, the RMD of H. speciosa increased in comparison to the inoculation with sole mycorrhizal species, independently of the P level.
Phosphorus levels had little influence on the RMD, which indicates high mangaba dependence on AM, even under higher P levels in the soil.Mycorrhizal dependency varies with plant species (Zangaro et al., 2007), soil nutrient levels (Costa et al., 2005;Zangaro et al., 2007), soil type (Van der Reijden & Kuyper, 2001), as well as with the mycorrhizal species (Costa et al., 2003(Costa et al., , 2005;;Nogueira & Cardoso, 2007).The degree of RMD was weakly dependent on soil P level, but was strongly dependent on the mycorrhizal species.This observation is very important for selection of more effective AM inocula, in order to improve mangaba seedlings production in nursery.
Root colonization by G. margarita or G. etunicatum reached different levels and declined with increasing P levels (Figure 3).Conversely, P levels did not affect mycorrhizal colonization, when the inoculum was the native AM pool.It was observed that G. margarita and G. etunicatum presented their highest root colonization between 0 and 25 mg kg -1 of added P. However, native AM presented average root colonization of 77%, considering all P levels.The higher colonization and the absence of P levels effect, in the treatment with the

NP Ge Gm
Relative mycorrhizal dependency (%) Figure 2. Relative mycorrhizal dependency (RMD) of mangaba tree inoculated with AM, in a sterilized soil with increasing levels of phosphorus.Native pool, NP; Glomus etunicatum, Ge; Gigaspora margarita, Gm.RMD was calculated as follows: [(M-NM)/M] x 100, in which M is the total dry biomass of mycorrhizal plants; NM is the total dry biomass of nonmycorrhizal plants (Plenchette et al., 1983).
native AM pool, can be attributed to the higher diversity of AM in this treatment, containing species more tolerant to P fertilization levels.Costa et al. (2005) showed that the increase of P levels did not affect the amount of arbuscules in the roots of mangaba trees, when the soil was not sterilized and contained high diversity of native AM.The native AM are often more adapted to local edaphic conditions, and their competitiveness may differ according to soil conditions and soil-plant-endophyte association.In addition, the AM have different strategies for root colonization.For example, the Gigasporaceae use to have less root colonization but produce more external mycelium, in an opposite fashion to the generally observed for the Glomeraceae (Hart & Reader, 2002).
In fact, in the present work, despite the external mycelium had not been accessed, lower root colonization was observed for G. margarita in relation to G. etunicatum, as previously surveyed.
In general, P addition in the soil increased shoot P concentration and accumulation more expressively in the mycorrhizal plants than in the nonmycorrhizal control (Figure 4 A and B).Plants infected with G. margarita showed no significant differences in the shoot P concentration along the P levels.The greater shoot P concentration and accumulation, verified in mycorrhized mangaba compared to the nonmycorrhizal plants, can be explained by several mechanisms used by the mycorrhizal external mycelium for a better soil exploration (Nogueira & Cardoso, 2007) and modifications in host plant physiology (Graham, 2000).The values of P concentrations are slightly higher than the reported ones for mangaba tree (Costa et al., 2005), especially in the mycorrhizal plants at higher P levels.Woody species adapted to low-fertility soils, like mangaba, usually have low growth rates, and are often very conservative in their use of nutrients.These species have low rates of tissue turnover and high degrees of nutrient reallocation, which contributes to a higher overall P utilization efficiency (Zangaro et al., 2007).
This study demonstrated that mangaba is very dependent on mycorrhiza, regardless the inoculum source, indicating that mangaba tree could not survive in natural low-fertility environments without mycorrhizal symbiosis.In addition, not only the inoculation with sole purified mycorrhizal species, but also native AM pool can anticipate the plantlets, in nurseries of mangaba seedlings, which can be transplanted earlier to the field.The plant response to AM inoculation is always dependent on plant species, AM fungi and environment interactions, which should be previously tested in each case.