PRODUCTION OF Colubrina glandulosa SEEDLINGS WITH DIFFERENT MYCORRHIZAL INOCULA

Colubrina glandulosa has potential for reforestation of disturbed areas. Seedlings inoculated with arbuscular mycorrhizal fungi (AMF) can be produced with biosolids from Sewage Treatment Plants. This study aimed to evaluate the response of Colubrina glandulosa seedlings to isolated or mixed inoculation of different AMF species on a biosolid-based substrate under greenhouse conditions. The experimental design was completely randomized, with 32 replicates (seedlings)/treatment and five treatments: (1) control or absence of inoculation (CT); (2) inoculation with Dentiscutata heterogama (DH); (3) inoculation with Gigaspora margarita (GM); (4) inoculation with Rhizophagus clarus (RC); (5) inoculation with the mixture of the three AMF species (MT). We evaluated height and collar diameter of the seedlings at 30, 60, 90, 120, and 140 days after the experiment installation. We evaluated shoot dry biomass (SDM), root dry biomass (RDM), and total dry biomass; SDM / RDM ratio; Dickson Quality Index; relative increments in height, collar diameter, and shoot biomass; nutrient content (N, P, K) in the shoots and roots; efficiency of absorption and utilization of nutrients; rate of root mycorrhizal colonization, at 140 days. In general, the highest values of growth and nutrition variables occurred in the MT treatment, which was recommended, in comparison with CT, DH, GM, and RC. Colonization rates were low (≤ 20 %), regardless of the mycorrhizal inoculum, due to the high levels of phosphorus in the biosolid.


INTRODUCTION
Colubrina glandulosa Perkins (popularly known as "saguaraji-vermelho" or "sobrasil") belongs to the Rhamnaceae family. This initial secondary tree species has rapid growth and can be planted in full sun, which promotes the formation of a microclimate favorable to the growth and establishment of forest species of more advanced successional groups, which require greater shading (SILVA et al., 2015). Additionally, Colubrina glandulosa has a high capacity of spatial occupation, from the regrowth of roots in Atlantic Forest areas that are under conditions of frequent fires (RODRIGUES et al., 2004). In the Amazon Forest biome, this species was among those that had the most satisfactory growth, in a plantation for reforestation of a degraded pasture area (GAMA et al., 2013). Thus, Colubrina glandulosa can be considered a key species in reforestation plantations of disturbed areas. Therefore, studies should be carried out focusing on the production of seedlings with higher quality, for the use in forest recovery programs. This context includes investigations about the inoculation of seedlings with arbuscular mycorrhizal fungi (AMF), which show higher quality when compared to non-inoculated seedlings (SCABORA et al., 2010;. Symbiosis between plants and AMF promotes benefits in growth and nutrition for the former, in addition to savings of more than 60 % in the application of phosphorus in the production of forest seedlings, compared to those not inoculated with AMF (ROCHA et al., 2006). AMF are also able to promote an increase in tolerance and/or resistance of terrestrial plants to water stress, salt stress and presence of heavy metals in soil solution, which are sequestered by glomalin, a protein secreted by these microorganisms that contributes fundamentally to the stability of soil structure (FOLLI-PEREIRA et al., 2012). However, the benefits promoted for plants by the mycorrhizal association depend on several factors, which include the availability of phosphorus in soil solution , the species of AMF involved (MARTIN et al., 2012) and species of plant host (SCABORA et al., 2010). Camara et al. (2017) evaluated the effect of single inoculation of Rhizophagus clarus Becker and Gerdemann (previously classified as Glomus clarum Nicolson and Schenck), Gigaspora margarita Becker and Hall, Dentiscutata heterogama (Nicol and Gerd) Walker and Sanders and mixed inoculation with the three AMF species on the production of Colubrina glandulosa seedlings in different substrate formulations, in a greenhouse. These authors concluded that the seedlings inoculated with mixed inoculum and produced on substrate consisting of 60 % of soil from A horizon, 20 % of bovine manure, 10 % of sand and 10 % of vermiculite had the best nutritional and growth conditions. The superiority of mixed inoculation of AMF was observed in other studies . However, this pattern does not occur in some situations, in which seedlings of tree species with higher quality can be obtained through inoculation with only one species of AMF, in comparison with the mixed inoculum (MACHINESKI et al., 2009). Also, in other situations, there may be no differences between the effect of mixed inoculum and the effect of single inoculation with a certain species of AMF on the seedlings produced (CARMO et al., 2016).
This panorama shows the relevance of investigating the effect of different AMF inocula, using the same substrate formulation, for the production of seedlings of Colubrina glandulosa, among other tree species that can be used in the reforestation of disturbed areas. In this context, it is possible to use sewage sludge from Sewage Treatment Plants (STP), which constitutes an organic material that is rich in nutrients called biosolid, after the stabilization process (GOMES et al., 2013;ABREU et al., 2019). Thus, the biosolid enables the production of quality seedlings for the reforestation of disturbed areas (TRIGUEIRO; GUERRINI, 2014), whose destination becomes more environmentally sustainable when compared to its disposal in landfills (GOMES et al., 2013).
The present study aims to evaluate the growth and nutritional status Colubrina glandulosa seedlings under inoculation with a mixture of different species of AMF (mixed inoculum) and inoculation with a single species of AMF, in a biosolid-based substrate, in a greenhouse. We tested the hypothesis that there is a difference in the response of Colubrina glandulosa seedlings to inoculation with the mixture of three species of AMF (Rhizophagus clarus, Gigaspora margarita and Dentiscutata heterogama), in comparison to inoculation with a single species of AMF and with the absence of mycorrhizal inoculation.

MATERIAL AND METHODS
Colubrina glandulosa seedlings were produced in a greenhouse of the Forest Institute of the Federal Rural University of Rio de Janeiro, Seropédica, RJ, Brazil. Containers with volumetric capacity of 980 mL were used. Each container was filled with approximately 1.0 kg of substrate consisting of a mixture of 20 % vermiculite and 80 % sewage sludge biosolid, which came from the Alegria Sewage Treatment Plant (STP), located in Rio de Janeiro, RJ. The chemical analysis of the biosolid, on a dry basis (%), indicated the following results: pH (H2O) = 5.45; N = 33.497 mg kg -1 ; Total P = 0.62 % or 6.20 g kg -1 ; Total K = 0.50 % or 5.0 g kg -1 ; Total Ca = 0.24 % or 2.40 g kg -1 ; Total Mg = 0.01 % or 0.1 g kg -1 ; Total Na = 0.16 % or 1.6 g kg -1 ; organic carbon = 28.7 % or 287.00 g kg -1 (ABREU et al., 2019). This material was moistened and autoclaved (temperature of 120 ºC, pressure of 1 atm, 1 hour) three times, with interval of 4 days between each autoclaving procedure .
Seeds of Colubrina glandulosa, which were donated by the Fernando Luiz Oliveira Capellão Forest Nursery, were disinfested with sodium hypochlorite (NaClO) at 2 % for five minutes. Then, dormancy was broken by immersion in concentrated sulfuric acid corresponding to twice its volume, in a glass beaker (BRANCALION et al., 2011). After 15 minutes, the beaker contents were stirred with a glass rod and, after 30 minutes, the seeds were separated and washed in a sieve under running water for 10 minutes. Then, they were placed outdoors on paper towels to drain surface water, on a laboratory bench. Subsequently, the seeds were subjected to a pregermination process in trays containing sand and inside a germination chamber (Biological Oxygen Demand), under constant light and temperature of 28 °C. After two days, the radicle had an average length of 4 mm. Five days after sowing, four pre-germinated seeds were inserted into a small orifice in the substrate of approximately 1.0 cm, in each container, and covered with a thin layer of the substrate itself. Approximately 15 days after this stage, thinning was performed, leaving only the most vigorous seedling in each container .
At 140 days after transplanting the pre-germinated seeds, the seedlings began to show signs that the container was no longer suitable for their full development and, for this reason, the experiment was ended on that date. Shoot height (H) and collar diameter (CD) were evaluated at 30, 60, 90, 120 and 140 days after transplantation, with a graduated ruler and a digital caliper, respectively. CD was not evaluated at 90 days after transplantation due to technical problems in the device on that date. At 140 days after transplanting the pregerminated seeds, the seedlings were removed from the greenhouse and divided into shoots and root system. Subsamples with approximately 0.5 g of fine roots were separated from 10 randomly selected replicates and subjected to bleaching and staining (KOSKE; GEMMA, 1989;GRACE;STRIBLEY, 1991). Then, the percentage rate of root colonization by AMF (ROOTCOL) was evaluated by the grid-line intersect method (GIOVANNETTI; MOSSE, 1980). Shoots and root system were dried in a forced air circulation oven (65 ºC, 72 h) to obtain the values of shoot dry biomass (SDM) and root dry biomass (RDM) on a digital scale with two decimal places. These data were used to calculate total dry biomass (TDM), SDM/ RDM ratio, Dickson Quality Index (DQI) and relative increments in the biomass of shoots (RIB_S), roots (RIB_R) and total (RIB_T). DQI ranges from 0 to 1.0 and was estimated from the following equation: DQI = (TDM / (H / CD) + (SDM / RDM)). RIB was estimated considering the percentage increments in shoot dry biomass, root dry biomass and total dry biomass, which were promoted by each inoculation treatment (DH, GM, RC or MT), compared to the control (CT) with the following equation The same equation was used to evaluate the relative increments in shoot height (RIH) and collar diameter (RICD), which considered the H and CD values obtained at the last evaluation time (at 140 days after transplanting). After weighing, the materials were ground and subjected to sulfuric digestion to determine N content, by the modified Kjeldahl method, P content, by colorimetry, and K content, by flame photometry (TEDESCO et al., 1995). The product of dry mass and nutrient content data were used to calculate the contents of N, P and K in the shoots (N_S, P_S and K_S, respectively) and in the root system (N_R, P_R and K_R, respectively). Then, the absorption efficiency (ABE) and utilization efficiency (NUE) for each nutrient were calculated. ABE makes it possible to evaluate the influence of the treatments on the ability of the plant to extract N from the substrate, which was converted into dry biomass, and was estimated by the equation proposed by Swiader et al. (1994): = ( _ + _ )/( ) NUE, which makes it possible to assess how much biomass was produced as a function of N absorption, was estimated by the equation of Siddiqi and Glass (1981) . The same procedure was performed for P and K.
To evaluate the influence of the observation period (collection time) on H and CD, repeated-measures ANOVA was applied. All data obtained were subjected to analysis of variance (One-way ANOVA) and, when the assumption of homogeneity of variances was met by the Levene test, treatment means were compared by the parametric LSD test. Otherwise, the means were compared by the non-parametric Kruskal-Wallis test. All statistical analyses, carried out with version 8.0 of the STATISTICA program, considered p < 0.05.
Multivariate analyses were used to assist in the interpretation of the data obtained through version 2.17c of the PAleontological STatistics (PAST) program. From the dendrogram generated in the hierarchical cluster analysis by Bray-Curtis distance and based on the single linkage method, the objective was to identify possible similarities between treatments. Principal component analysis was performed in order to identify the variables that correlated with the treatments. Both of these analyses considered the mean data of all variables studied, except for RIH, RICD and RIB; in the case of the variables H and CD, the respective means calculated considering all evaluation times were used.

RESULTS
There was significant interaction between the collection times and the treatments for H and CD (Table 1). In the comparison between two consecutive times of H evaluation, significant differences were observed in the effect of treatments between the first and second times and between the fourth and fifth times. There were no significant differences in the increment of H, in the pairs formed by the second and third times and by the third and fourth times. FLORESTA, Curitiba, PR, v. 50, n. 4, p. 1731-1740  In the comparison between two consecutive times of CD evaluation, significant differences were observed in the effect of treatments between the first and second times and between the second and third times (Table 1). There were no significant differences in the increment of CD between the third and fourth times.
Based on the mean results, which were calculated from the data found at each of the different evaluation times for these variables, it was possible to identify significant differences between treatments in relation to their effect on H and CD. By comparing the first and last evaluation times, it was found that the MT treatment caused a greater increase in H (1868 %) in absolute terms, compared to the treatments CT, DH, GM and RC (1831 %, 1582 %, 1623 % and 1423 %, respectively).
In general, this pattern resulted from the superiority of the MT treatment, which significantly influenced higher H values in comparison with the treatments CT (at 30, 90 and 120 days after experiment installation), GM (at 30 and 120 days), DH (at 30, 120 and 140 days) and RC (at 120 and 140 days) ( Table 2). The mean value of H, which was calculated considering the five evaluation dates, was significantly higher in MT, compared to the other treatments, which did not show significant differences from one another.
Tabela 2. Valores médios de altura (H) e diâmetro de colo (CD) de mudas de Colubrina glandulosa em diferentes tratamentos de inoculação com fungos micorrízicos (Trat) aos 30, 60, 90, 120 e 140 dias após o transplantio das sementes pré-germinadas, em condições de casa de vegetação. Table 2. Mean values of height (H) and collar diameter (CD) of Colubrina glandulosa seedlings at 30, 60, 90, 120 and 140 days after transplanting of pre-germinated seeds, in different treatments of arbuscular mycorrhizal inoculation (Treat) under greenhouse conditions. The MT treatment also influenced higher absolute values of CD (406 %) compared to the treatments CT, DH, GM and RC (349 %, 363 %, 355 % and 370 %, respectively). The mean values calculated considering the four evaluation dates revealed that the treatments MT and RC did not differ from each other, and promoted significantly higher values of CD, compared with the other treatments (Table 2).
Tabela 3. Tabela 1. Análise de variância (One-way ANOVA) para incremento relativo em altura da parte aérea (RIH), biomassa seca da parte aérea, raízes e total (SDM, RDM e TDM, respectivamente), razão SDM / RDM, Índice de Qualidade de Dickson (DQI), taxa percentual de colonização radicular (ROOTCOL), conteúdos na parte aérea de N (N_S) e P (P_S), eficiência de absorção de N (ABE_N) e eficiência de utilização de P (NUE_P) de mudas de Colubrina glandulosa em diferentes tratamentos de inoculação com fungos micorrízicos arbusculares aos 140 dias após o transplantio das sementes pré-germinadas (tempo), em condições de casa de vegetação. Table 3. Analysis of variance (One-way ANOVA) for relative increment in shoot height (RIH), shoot dry biomass, root dry biomass, and total dry biomass (SDM, RDM and TDM, respectively), SDM / RDM ratio, Dickson Quality Index (DQI), percentage rate of root colonization (ROOTCOL), shoot contents of N (N_S) and P (P_S), absorption efficiency of N (ABE_N), and utilization efficiency of P (NUE_P) of Colubrina glandulosa seedlings in different arbuscular mycorrhizal fungi inoculation treatments at 140 days after transplanting of pre-germinated seeds (time), under greenhouse conditions * . The MT treatment promoted significantly higher values of SDM and TDM of the seedlings, compared to the other treatments of AMF inoculation (Table 4). The treatments MT and RC, which did not differ from each other, influenced significantly higher values of DQI and ROOTCOL, compared to CT. Regarding the variables RIH and SDM/RDM ratio, there were no significant differences between MT, CT, DH and GM, whose values were significantly higher than those observed in the RC treatment. On the other hand, the RC treatment influenced significantly higher values of RDM, compared to CT, DH and GM.

Treat
The MT treatment led to significantly higher values of N_S, in comparison with GM and RC, and of P_S, in comparation with CT, DH and GM (Table 5). Significantly higher values of ABE_N were verified for seedlings in the MT treatment, compared to CT and RC treatments; of NUE_N in the comparison with DH; and of NUE_P, when compared with GM and RC (Table 6).
The hierarchical cluster dendrogram pointed to the formation of two large groups: one group was formed by the MT treatment alone, while the other was formed by CT, DH, GM and RC (Figure 1b). This second group, in turn, was subdivided into two subgroups: one that was represented alone by CT, and the other by the treatments DH, GM and RC. Therefore, this multivariate analysis demonstrated that there was a greater similarity between the treatments DH, GM and RC with regard to their effects on the growth and nutritional status of the seedlings, while the effects caused on the seedlings by the treatments CT and MT differed from each other, and in the comparison with the other treatments.

DISCUSSION
The absence of root colonization in non-inoculated seedlings showed that the control was effective. Root colonization rates of Colubrina glandulosa seedlings were considered low (≤ 20 %), regardless of the mycorrhizal inoculum employed (single species of AMF or the mixture of the three species). This result was due to the high total phosphorus content in the biosolid (ABREU et al., 2019), which probably promoted high availability of this nutrient for the seedlings. In general, under conditions of greater availability of phosphorus in the soil solution, the plant does not stimulate the formation of symbiosis with AMF , whose maintenance would constitute an energy burden without due benefit for it . Therefore, under these conditions, there are low rates of root colonization by AMF (SMITH et al., 2010). (56 %) . For Apuleia leiocarpa (Vogel) J. F. Macbr (garapa), the rate of mycorrhizal colonization in the root system of seedlings produced in substrate with addition of the highest doses of phosphorus (650 mg kg -1 ) was on the order of 8 %, whereas with the addition of the