CONSERVATION OF SEEDS OF Myrcianthes pungens (Berg) Legr. IN DIFFERENT PACKAGING IN A CONTROLLED ENVIRONMENT

– Studying the physiological and biochemical behavior of Myrcianthes pungens seeds stored in diﬀ erent packages for up to ten months was the objective of this work. Seeds were collected in Toledo, Pato Bragado, and Marechal Cândido Rondon and stored in a cold and dry chamber (11 °C and 6,3% RH) for ten months. The germination test and the germination speed index (GSI), the seedling length (SL), the seedling dry matter mass (DMS), and the tetrazolium (ZT) test were performed and, for each evaluation, the water content was determined. Non-parametric analysis was used. During storage, the water content of the seeds in the plastic and glass containers varied little, but decreased in the paper packaging. The germination was maintained for ten months when the seeds were stored in plastic containers and for two months in the glass and paper ones. The variables GSI, SL, and DMS presented upwards and downwards variations during the storage period. The ZT revealed that, in the plastic packaging, the seeds remained viable for up to ten months, in glass, up to two months, and in paper, up to four months. Therefore, seeds of Myrcianthes pungens with high initial quality can be stored in a cold and dry chamber in plastic bags, maintaining viability and vigor for a period of ten months


INTRODUCTION
Myrcianthes pungens (Berg) Legr., popularly known as guabiju, belongs to the Myrtaceae family, and is now at risk of extinction (IUCN, 2018). The species is native and appears from the state of São Paulo to the north of Uruguay, reaching Bolivia, Paraguay, and Argentina (Raseira et al., 2004). The fruit of the species are globose berry type, containing one to two reniform seeds, which measure around 6 to 7 mm (Souza et al., 2011).
The fruit of guabiju is very appreciated by the native fauna and also by humans. The species can be used for decoration, urban aff orestation , and can be a source of income for families of farmers, considering that the fruit can be used in sweets, yogurts, juice, ice cream, etc. (Hossel et al., 2016). The wood is suitable for luxury joinery, the fl owers are melliferous, and the species is recommended for mixed plantings destined to the restoration of degraded areas of permanent preservation (Lorenzi, 1992).
Since its main means of propagation are the seeds, ex situ conservation should be a priority in seedling production programs for the various purposes of cultivation. However, guabiju presents limitations when it comes to storage (Hossel et al., 2016), because its seeds are classifi ed as recalcitrant, and seeds with this physiological behavior have high water content due to fruit ripeness and short longevity, even in conditions with high air humidity and low temperature .
Preservation of seed quality during storage requires the integration of several factors, such as temperature, relative humidity, type of packaging, and moisture content of the seeds, as well as previous knowledge of the physiological behavior of the seeds during storage, since some seeds require specifi c conditions (Gomes et al., 2013).
Seed conservation under given conditions has the basic function of maintaining the physical, physiological, and sanitary qualities, preserving the viability and vigor of the seeds (Marcos Filho, 2015). Therefore, seed quality assessment is very important for successful storage and seedling production. Vigor is one of the most important aspects of seed quality analysis, considering that the deterioration process is directly related to loss of vigor (Santos, 2018).
The execution of vigor tests tries to detect signifi cant diff erences in the physiological potential providing additional information to germination tests. Simultaneously, the results of these tests are expected to distinguish between high and low vigor lots. Also, the diff erences found can be related to seed behavior during storage and/or after sowing. The feasibility tests, in turn, aim at verifying if the seed is or is not able to germinate, by direct or indirect means (Marcos Filho, 2015).
The germination test has some limitations on its accomplishment, such as delays and possible modifi cations of the results due to the presence of fungi. Thus, it is necessary to use fast and reliable tests to estimate the viability of storage seeds (Cripa et al., 2014), such as the tetrazolium test, which is based on the activity of the dehydrogenase enzymes involved in the respiration process (Sousa et al., 2017).
Considering the risk of extinction of the species and its economic and ecological importance, in addition to its use for the purpose of forests restoration, it is important to store the seeds in order to conserve them for future use. Thus, the objective of the present study is to study the physiological and biochemical behavior of Myrcianthes pungens seeds stored in diff erent packages for a period of up to ten months. The harvesting was performed manually when the fruit pieces were dark purple and started falling spontaneously (Lorenzi, 1992). The globose berry fruits were manually sifted in a sieve under running water, separating the seed from the pericarp. To eliminate excess water, the seeds were dried in the shade for 4 days.

Pieces
The seeds from the three collection sites were stored in a cold and dry chamber (11 °C and 6.3% HR), separately in trifoliate kraft paper bags (600 mL), clear glass (600 mL preserved type), and clear polyethylene plastic bags (approximately 8 mm thick and 600 mL), with initial moisture, for 0, 2, 4, 6, 8, and 10 months. The seed germination rate (SGR), seedling length (SL), and vigor were determined for each storage period and type of packaging, seedling dry matter mass (DMS), and tetrazolium test, so that if seeds remained in the package they were discarded, not influencing the seeds that remained stored.
The germination test was done with 4 replicates of 25 seeds, seeded in vermiculite, moistened until saturated. The test was carried out in a germination chamber (BOD) at 25 °C (Santos et al., 2004) with photoperiod of 12 hours. Evaluations were performed daily until stabilization and the results were expressed as the percentage of normal seedlings. The normal seedling criterion was the presence of epicotyl, hypocotyl, and primary root development and the fi rst pair of defi nitive leaves. After stabilization, the germination rate (Maguire, 1962) was calculated and the seedling length and dry matter mass of the normal seedling were evaluated.
The tetrazolium test was performed with 4 replicates of 25 seeds. The seeds were conditioned between paper moistened for 24 hours in BOD at 30 °C, without tegument. Subsequently they were placed in a 0.75% tetrazolium solution for 24 hours at 30 °C in the absence of light. The seeds were evaluated individually, and the results expressed as the percentage of viable seeds, as predetermined by the authors for the species.
The design was completely randomized. Nonparamectric analysis was used, since the data did not present normality and homogeneity. First, each factor (storage period, packaging, and collection site) was analyzed separately using the Friedman test and then the comparison of averages, comparing the packages in each storage period and the storage periods for each package to a 5% probability of error. The program used is Sigmaplot 14.0.

RESULTS
During storage, it was observed that the highest percentages of water content were found for seeds stored in plastic and glass, diff ering signifi cantly from the paper bag storage. All three types of packages were used in all collection sites (Table 1). When analyzing the behavior of the containers throughout the storage packages, it was verifi ed that, for the plastic bag, there was reduction of the water content, behavior similar to the observed for glass. The paper bag, in turn, showed signifi cant decrease in water content. The variation of the water content throughout the storage units for seeds in plastic bags and glass varied by up to 6%. In contrast, the seeds in paper bags presented an average variation of 29.6% reduction.
Enter Table 1. The percentage of germination was maintained until the end of the study only in the seeds stored in plastic bags. The seeds stored in glass or paper bags germinated by the fi rst two months (Table 2). In the packages throughout the storage, it was observed that there was variation in germination, with increases and decreases, being the highest percentages 90% for Toledo and 91% for Pato Bragado and Marechal Cândido Rondon. For seeds stored in glass in Toledo and Marechal Cândido Rondon, at two months, the percentages were lower (16 and 3% respectively), and in Pato Bragado, the germination was 80%. In paper bags, the percentages were low in the three collection sites and, at two months, it was 7, 16, and 8% for Toledo, Pato Bragado, and Marechal Cândido Rondon, respectively. The rate of germination (SGR) in the seeds stored in diff erent packages was higher for seeds in plastic bags, diff ering signifi cantly from the other packages, a fact observed at all collection sites (Table 2). For seeds in plastic bags, a constant variation was observed throughout the storage. For the seeds in glass and paper, there was a decrease of the SGR with the reduction of germination, a relation that was not observed in plastic bags.
From all the collection sites, only seeds in plastic bags presented normal seedlings until the end of the storage period (Table 3). In the other packages, the normal seedlings occurred by the second month. Comparing the eff ects of the packages over time, there was variation in the formation of seedlings in plastic bags. In glass, there was an increase in the length of seedlings whereas, in paper, a decrease could be observed for all the collection sites.
For the average dry matter mass of normal seedlings, it was observed that in the period of two months of storage in glass, the highest average value occurred in the seeds at Toledo (150 mg). In Pato Bragado, the highest average occurred in seeds stored in paper bags (103 mg). And, in the seeds collected in  Marechal Cândido Rondon, there was no diff erence between the packages in this period, which were 68, 102, and 79 mg for plastic bags, glass, and paper bags, respectively. In the period of four to ten months, only the seeds stored in plastic bags germinated and generated normal seedlings and therefore reached the highest averages when compared to the seeds stored in glass and paper bags, which resulted in null values ( Table 3).
As for the eff ects of packaging along the storage, it was verifi ed that seeds in the glass and paper packages at Toledo did not diff er signifi cantly at zero and two months. However, seeds in glass at Pato Bragado and Marechal Cândido Rondon, and in paper bags at Marechal Cândido Rondon had a decrease in the mass of dry matter. Also, seedlings in paper packages at Pato Bragado presented increase of the mass of dry matter. Seedlings in plastic bags varied.
Regarding the feasibility of the tetrazolium test, it can be concluded that in the range of two to ten months in plastic bags, the number of viable seeds decreased and diff ered signifi cantly from those in other packages, with the highest percentages for all collection sites.
The seeds collected from Toledo and Pato Bragado and stored in glass remained viable until two months. Those of Marechal Cândido Rondon remained viable until four months, although with smaller percentages. Seeds collected at Toledo and Marechal Cândido Rondon and stored in paper bags remained viable until four months and those collected from Pato Bragado, up to six months (5% at 6 months, table 4). The seeds in glass and paper packages presented reduced percentages of viability, diff ering signifi cantly, whereas in plastic bags, variations occurred over time.

DISCUSSION
The water content of the seeds was similar to that found by Santos et al. (2004) (around 40%) and by Fior et al. (2010) (between 41.1 and 43.6%), except for seeds stored in paper bags. These authors have reported that high levels of water at the maturation stage are a characteristic of recalcitrant species.
In the storage of Myrcia glabra (O. Berg) and Myrcia palustris DC. in polyethylene at 5 °C for 150 days, it has been verifi ed that there was no signifi cant change in water content . In turn, Hossel et al. (2017), when storing seeds of Eugenia pyriformis (Cambess) in paper bags at 5 °C, have verifi ed loss of water content, as observed in the present study and by Hossel et al. (2016) with seeds of M. pungens. On other hand, seeds of Cordia trichotoma (Vellozo) Arrabida ex Steudel stored in glass have had little variation of the water content during twelve months (Gusatto, 2015).
A major factor in storage is packaging (Hossel et al., 2016). The packaging used in storage should help to slow down the deterioration process by maintaining the initial water content of the seeds stored in order to reduce respiration (Amaro et al., 2015). Also, the deterioration of seeds is directly linked to the characteristics of the containers where they are stored, depending on their infl uence on the water vapor exchange between the seeds and the atmosphere and the conditions of the storage environment (Cardoso et al., 2012).
The modifi cations of the water content of the seeds in the present work can be explained by the fact that the cold chamber was opened frequently during some periods, causing oscillations of temperature and humidity of the environment. Thus, the seeds stored in paper bags quickly responded to the change in their  0  85aB  85aA  85aA  94aCD  94aA  94aA  97aAB  97aA  97aA  2  84aB  37bB  16cC  95aBC  82bB  20cB  88aC  5cB  20bB  4  83aB  0cC  17bB  90aD  0cC  7bC  70aC  2cC  6bC  6  93aA  0cC  2bD  95aBC  0cC  5bD  98aA  0bD  0bD  8  95aA  0bC  0bE  98aAB  0bC  0bE  90aBC  0bD  0bD  10  93aA  0bC  0bE  100aA  0bC  0bE  90aBC  0bD  0bD Lowercase letters in the row at each collection site and uppercase letter in the column in each package and place of collection do not diff er according to the Friedman's test (p>0,05). microclimate. On the other hand, the semipermeable and impermeable packages did not change as much, since they did not exchange the water content with the environment or in, the case of plastic, it happened in a less effi cient way.
In the case of germination, Fior et al. (2010) have verifi ed that seeds of M. pungens stored for up to eight months in polyethylene maintained percentages above 80% up to fi ve months and subsequently declined by more than 50%. Some batches have lost practically all viability; however, germination has been maintained until the end of storage. In the present study, germination percentages were higher in most evaluations, and at ten months the percentage was above 91% (Table 2) for seeds stored in plastic bags. It was also verifi ed that, during the storage, the seeds showed diff erent performances in the germination test, varying between increases and decreases in the percentages. That could be related to the viability of the seeds or even errors at the time of the test execution, factors that can aff ect the measurement of the percentage of germination, as observed in the work.
In contrast, seeds of Eugenia pyriformis Cambess stored in paper bags have had a reduction in germination power and a rate of germination as water content is reduced (Hossel et al., 2017). A similar result was observed in the present study, where the reduction of the water content of the seeds was harmful, since seeds with water content between 12.1 and 13.5% (Table 1) are not able to germinate (Table 2). On the other hand, the seeds stored in glass maintained the water content. Nevertheless, they did not germinate from four months onwards.
Seeds of Myrcia palustris DC. stored in polyethylene bags have had a reduction of SGR at 150 days storage . In the present study, evaluating the three collection sites, each one presented reduction of the SGR at diff erent moments. According to Gusatto (2015), recalcitrant seeds do not support storage, presenting a decrease in germination and vigor, which is represented by SGR.
According to Amaro et al. (2015), the reduction of seedling emergence may be a result of the deterioration of the seed when packaged in a permeable package, since the greater permeability of this package does not reduce the metabolism of the seeds to the desired levels for the storage, causing the greater consumption of reserves and, thus, maintaining vigor. In addition to reducing the germination speed, the developmental unevenness between seedlings is another symptom of decline in physiological quality.
The germination of seedlings of Punica granatum L. stored in polyethylene at 5 °C, at 30 days, have shown a reduction in the average length of both shoot and root (Monteiro, 2017). The same author verifi ed that, when storing the seeds in polyethylene, paper bags, and glass, there has been a signifi cant increase of the mass of dry matter at 30 days when compared to the control group. In the present work, when evaluating at two months, there was a signifi cant reduction of dry matter mass of seedlings. The author also verifi ed that, up to 120 days of storage, there has been a variation of the mass of dry matter, alternating increases and decreases of the value, as it was observed in the present work.
Seeds stored in a controlled environment tended to be conserved for a longer period, since with the use of reduced temperature there may be less water loss to the environment. Thus, one of the ways to maintain the germination power for short periods of storage is to use packages that maintain the initial water content at temperatures of 4 and 6 °C (Hossel et al., 2016).
Packaging and storage conditions should contribute to the maintenance of uniformity of the seed water content. When these are stored in permeable packages, the water content changes as the relative humidity changes, but if they are semipermeable, there is resistance to gaseous exchanges, but nothing that completely prevents the exchange of moisture. In impermeable packages, in turn, there is no infl uence of the external air humidity on the seeds (Oliveira-Bento et al., 2015).
For the conservation of the germinative power of the seeds, it is necessary to keep them in a dry and cold environment. The drier and colder, within certain limits, the greater the chances of prolonging the conservation of the seeds. In environments without humidity and temperature control, the moisture present in the air may be suffi cient to promote the resumption of embryo activities if oxygen and temperature are suffi cient. Breathing, together with the action of microorganisms, provoke the heating of stored seeds, which can drastically reduce their viability (Silva, 2015).
By verifying the results of viable seeds through the tetrazolium test (Table 4) and percentage of germination (Table 2), it was observed that the percentage of viability was always higher than the germination. In addition, the seeds stored in paper bags and collected in Toledo and Pato Bragado presented viability up to four months, but the germination occurred at two months or before. According to Grunennvaldt et al. (2014), this diff erence between the tests is due to the fact that the tetrazolium test is an indirect and rapid analysis, informing the seed quality in less than 24 hours, depending on the species, but it is not possible to obtain more information on the percentage of dormant seeds and on the contamination by pathogens.
The diff erences between the values of the tetrazolium test and other vigor tests may occur due to diff erences in sampling, techniques unsuitable for tests, presence of hard seeds, and mechanical damages in seeds (Pinto Junior, 2010). The author, when studying the storage of seeds of Jatropha curcas L. in permeable, semipermeable, and impermeable packages, also has found diff erences when comparing the values of the tetrazolium test with those obtained from the other tests.
In the tetrazolium test, when the seed is immersed in the tetrazolium solution, it diff uses through the tissues, causing a reduction reaction in the living cells, which results in a red, non-diff usible compound, triphenylformazan, indicating that there is respiratory activity in the mitochondria and, consequently, the viability of the tissue (live) (Oliveira et al., 2018). There is a clear separation of living tissue, which breathes, from dead tissue that does not stain, and consequently does not breathe (Carvalho, 2016).
From two months of storage on, the appearance of brown colored seeds after the application of the tetrazolium salt began. When investigating a possible presence of fungi, it was verifi ed that the seed had the Aspergillus niger and the Penicillium sp. fungi. This was mainly observed in seeds stored in paper bags and glass, even though it could be detected, in a lower incidence, in seeds stored in plastic. However, it is not possible to attribute this fact (brown coloration in the tetrazolium test) to the presence of fungi and that should be further investigated in new research. Therefore, the seeds that presented this coloring were classifi ed as non-viable.
As reported by Aguiar et al. (2012), fungi of the species Aspergillus and Penicillium can cause direct damage to the seeds and in germination, with increase of the rate of fatty acids and production of toxins. The action of these microorganisms, provided there are conditions of humidity and temperature, accelerates the rate of decay of the seeds during storage (Carvalho and Nakagawa, 2000). The authors have added that this is a major problem for recalcitrant seeds, since they need to keep high water content, which is a favorable condition for the development of these fungi. Thus, the authors suggested that a fungicide treatment of the seeds should be carried out prior to storage, after partial drying, followed by storage at the lowest possible temperature (tolerable by the species).
Another factor that could explain the darkening of the seeds is that, during deterioration, there is a decrease in soluble sugars and in the total sugar content, and an increase of the reducing sugar levels. Consequently, there is loss in the ability to use carbohydrates, aff ecting the mobilization, from the reserve tissues to the embryonic axis. The presence of reducing sugars may induce deterioration of protein components through Amadori and Maillard reactions, especially in dry seeds. Maillard's reactions comprise a series of non-enzyme-driven reactions. These Maillard complexes undergo modifi cations for secondary ketoamines known as Amadori molecular arrangements, which become more stable but chemically reversible (Marcos Filho, 2015).
It can be verifi ed in the results of the tests applied to the seeds collected in diff erent places that there are diff erences, even on seeds of the same species. According to Teles (2017), it is important to work with fruit and seeds from diff erent localities because it is possible to verify the phenotypic diff erences determined by the environmental variations, even for the same species. In each locality, the seeds are subject to variations in temperature, light hours per day, rainfall indexes, and other characteristics that end up emphasizing certain aspects of its genetic composition, that is, one environment may be suitable for the expression of certain characteristics that may not manifest in another.

CONCLUSIONS
Seeds of Myrcianthes pungens with high initial quality can be stored in a cold room and dried in