MORPHOFUNCTIONAL ECOLOGY OF SEEDLINGS FOUND IN A TROPICAL RAINFOREST SEED BANK, IN THE NORTHEAST OF BRAZIL

The aim of this study was to evaluate the identification and floristic composition of seedlings of tree species, located in the soil seed bank of a remnant of the Dense Ombrophilous Forest of the Lowlands, in the Northeast of Brazil, from a morphofunctional ecology standpoint. A hundred and fifty two soil samples (0.2 m x 0.15 m x 0.05 m) were taken from a forest remnant, stored in polyethylene bags, identified and then taken to the Forest Nursery at the Federal Rural University of Pernambuco, where they were placed in polyethylene boxes under a 70% shade cloth and evaluated daily for a period of seven months. The seedlings emerged were photographed, separated into morphospecies, identified at the family and gender level and, when possible, at the species level. The density and absolute frequency, as well as the successional and morphological classification of the seedling cotyledons were estimated. In total, 1,577 seedlings of 31 morphospecies were cataloged. Melastomataceae was the family that presented the greatest richness, representing 48% of the species identified. The average density was 344.62 seeds.m, with higher densities for species of the family Melastomataceae, together with Maprounea guianensis and Cecropia palmata. There was a predominance of seedlings with epigeal and phanerocotylar germination, and with foliaceous cotyledons, corresponding to 93.56%. The seed bank was composed, predominantly, by pioneer species such as Cecropia palmata, Henriettea succosa, Miconia hypoleuca and Miconia tomentosa. Seedlings with epigeal germination, foliaceous and phanerocotylar cotyledons predominated for being easily established in more open areas.


INTRODUCTION
Morphofunctional ecology seeks to understand the functions of morphological structures that assist in the development and initial establishment of seedlings and in the complex succession process of plant communities (RESSEL et al., 2004;BOEGER, 2019). Thus, filling knowledge gaps about it can contribute to ecological studies. FLORESTA,Curitiba,PR,v. 51,n. 3,jul Studies have suggested that the variation in the characteristics of the cotyledons of the seedlings in terms of position, epigeal and hypogeal; texture or function, foliaceous or photosynthetic and nutrient reserve storage; and exposure, phanerocotylar and cryptocotylar may require different establishment strategies between species (RESSEL et al., 2004;GARWOOD, 2009;MENEZES et al., 2018).
In Brazil, more and more studies classifying the morphofunctional ecology of seedlings of tree species occurring in the Cerrado, in the Atlantic Forest, in the Amazon Forest and in the Caatinga have been carried out, showing a predominance of seedlings with epigeal germination and foliaceous and phanerocotylar cotyledons (RESSEL et al., 2004;CAMARGO et al., 2008;ALVES et al., 2016;MENEZES et al., 2018). These seedlings usually develop from small seeds, have rapid development and colonization and are found in open areas and in the early stages of succession (GOGOSZ et al., 2015;BOEGER, 2019).
Although the seedling stage represents one of the most critical phases in the life cycle of a plant (GARWOOD, 2009), it is in this initial phase that the morphological characteristics can assist in the knowledge of the regenerative processes of forests such as the seed and seedling bank.
The soil seed bank is formed by all viable seeds present in the soil and in the litterfall layer. They arrive through different dispersing agents and are incorporated into the soil, where they can remain viable for months and even years, until the emergence of favorable conditions to germinate, form the seedling bank and, later, take the place of the adult individual (SIMPSON; LECK; PARKER, 1989). Thus, its assessment allows us to conclude which species have initial potential for colonization in the area.
In Atlantic Forest areas, the seed bank species richness has been varied, considering that there may be a greater seeds inflow to compose the seed bank than an outflow to form the seedling bank (CORREIA; MARTINS, 2015;KUNZ;MARTINS, 2016;SILVA et al., 2018;SILVA et al., 2019). Thus, it is necessary to know the species autoecological aspects in the initial stages, since the heterogeneity existing in the environment and the plants responses to abiotic and biotic factors result in different ecological niches (GOGOSZ; BOEGER, 2019).
Therefore, the aim of this study was to evaluate the identification and floristic composition of seedlings of tree species, located in the soil seed bank of a remnant of the Dense Ombrophilous Forest of the Lowlands, in the Northeast of Brazil, from a morphofunctional ecology standpoint.

Data collection area
Soil samples for seed bank evaluation were collected from a remnant of the Dense Ombrophilous Forest of the Lowlands (IBGE, 2012), located in the municipality of Sirinhaém, State of Pernambuco, Brazil. The remnant has an approximate area of 44 ha and is located on land belonging to Usina Trapiche S/A., under geographic coordinates 8°33'4.77"S and 35°8'51.72"W.
The region has a monsoon climate (Am) according to the Köppen classification (ALVARES et al., 2013), with the rainiest period between May and July, and the hottest and driest period from September to December, with rainfall and mean annual temperature (1980 to 2018) of 2,000 mm and 27 ºC, respectively (APAC, 2019).

Soil sample collection
The seed bank was evaluated using 152 soil samples removed with the aid of a gardening spade, a hoe and a frame used to delineate the area, measuring 20 cm x 15 cm x 5 cm (length, width and depth) totaling 4.56 m 2 .
For soil sample removal, 38 subplots of 1 m² were delimited inside the 38 plots, installed along three transects to evaluate the arboreal component of the study area (LIMA et al., 2019). The soil samples were taken 50 cm from the vertices of these subplots. Thus, each set of four samples was approximately 65 m away from the next set ( Figure 1). Leaves and dry branches not yet decomposed were removed, and the soil samples for the seed bank evaluation were considered soil and litterfall, depending on the plot, since there were plots in which the soil was fully exposed.
The collection was carried out in the third week of September 2018. The samples were stored in polyethylene bags, identified with the number of each plot; then, they were taken to the Forest Nursery of the Department of Forestry Science (DCFL) of the Federal Rural University of Pernambuco, (UFRPE), Recife-PE, where they were placed in polyethylene boxes, without the presence of substrate, with the same dimensions as the frame used to delimit the soil, 20 cm x 15 cm x 5 cm (length, width and depth), randomly arranged in the seedbed to ensure the same conditions for all samples and covered with a 70% shade cloth to facilitate germination, as proposed by Silva et al. (2019). Along with the soil samples, containers with washed and sterilized sand were randomly placed in an oven at 105ºC for two hours, to evaluate possible seed rain contamination. The samples were manually watered with the aid of a watering can, twice a day, except on rainy days. After 130 days of the experiment installation and after five days without the emergence of new seedlings, all the emerging seedlings were removed and some specimens of each morphospecies were transplanted into polyethylene bags, with a capacity of 1.0 kg, or containers with a capacity of 500g, so they could develop, and then a reliable identification could be made.
Melastomataceae seedlings were placed in small containers, with a capacity of 50 g or 100 g, and later transplanted into larger containers, until the development of morphological characteristics that could differentiate them at the species level. The substrate used in all containers was forest soil, removed from the same remnant, and medium vermiculite (ratio 2:1).
After removing all the seedlings, the soil was turned over, so that the seeds that might be at the bottom of the soil could be closer to the surface and facilitate their germination. The seed bank was evaluated daily for a period of seven months (end of September to end of April), using the seedling emergence method.
During soil collection and monthly visits to the study area in the following months, shrub and arboreal individuals were observed to verify the occurrence of flowering and fruiting. Botanical material was removed from individuals that presented flowers and fruits. This material was properly herborized, identified by comparison with exsiccates in the Sérgio Tavares Herbarium (HST), at the Department of Forestry Science of the Federal Rural University of Pernambuco, and by specialists.
When species bore fruit, the fruits were collected and processed and the seeds were put to germinate in trays containing vermiculite as substrate and then placed inside the greenhouse in the Forest Nursery. Thus, with seed germination and seedling growth, it was possible to identify the seedlings that emerged in the seed bank.

Seedling definition and data analysis
The concept of seedling is varied and there is no precise definition, since different authors can take into account physiological and ecological aspects. Thus, it is known that a seedling appears when a seed germinates, but its complete formation is not fully defined, causing some authors to consider the individual as a seedling, while depending on the seed reserves, if a portion of the biomass comes from the seed reserves, or even when it presents some functional structure originating from seed reserves (MELO et al., 2004;CAMARGO et al., 2008). However, under natural conditions, this differentiation is not possible due to the different morphological characteristics presented by the species and by the capacity to capture resources. Thus, in the present study, FLORESTA,Curitiba,PR,v. 51,n. 3,jul individuals were considered as seedlings after the emission and full expansion of the eophyll, as proposed in the works by Gogosz et al. (2015), Gogosz and Boeger (2019) and Silva et al. (2019).
The seedlings that emerged in the seed bank were enumerated, and the number, morphospecies, and morphology of the cotyledons were noted in a spreadsheet, and the seedlings were then photographed. Subsequently, they were separated into morphospecies, identified at family level, genus and, when possible, at species level with their scientific name and respective authors, and updated according to the Missouri Botanical Garden database, by the website www.tropicos.org, according to the Angiosperm Phylogeny Group IV (APG IV) classification system.
The emergence speed was obtained with the weekly count of the number of emerged seedlings. The following phytosociological parameters were estimated for each species identified in the seed bank: absolute density (AD), relative density (RD), absolute frequency (AF) and relative frequency (RF), described by Mueller-Dombois and Ellenberg (1974). The data were processed using the software Microsoft Excel for Windows™ 2016.
The seedlings identified were classified according to the successional classification following the methodology proposed by Gandolfi et al. (1995), who consider three main groups: pioneer species, early secondary and late secondary, in addition to unclassified species that do not fit into any of these three groups, either because of lack of studies or for being identified only as morphospecies. The successional classification was performed through observations of adult tree species identified in the field and in the study by Lima et al. (2019) developed in the study area and also by bibliographic research.
The functional morphology of seedling cotyledons was analyzed as proposed by Garwood (2009), who considers the hypocotyl elongation as epigeal (E) or hypogeal (H); the cotyledons exposure, as cryptocotylar (C) or phanerocotylar (P); and the texture, as foliaceous (F) and reserve (R), which function as photosynthesizers and nutrient reserve storages, respectively, resulting in seedlings with the following combinations PEF, PER, PHF, PHR, CER, CEF, CHR and CHF .

RESULTS
During the observation period, 1,577 seedlings belonging to 13 botanical families were recorded. Seeds germinated from a total of 31 morphospecies, 24 being identified at a specific level, 4 at a generic level and 3 could not be classified at any taxonomic level (Table 1), as they did not present morphological characteristics for comparison with adult individuals and possible identification. Melastomataceae was the family presenting the highest richness with 48% of the species identified, and 79% of the seedlings emerged.
The emergence of tree species started three weeks after the installation of the experiment with the species Trema micranta and Maprounea guianensis; later, in the fourth week, the species Cecropia palmata, Trema micrantha, Maprounea guianensis, Henriettea succosa, Miconia cf. pyrifolia and Miconia minutiflora emerged. The species Miconia spp. and Henriettea succosa were constant, emerging throughout the evaluation of the bank. The greatest seedling emergence occurred in the 18th, 25th, 26th and 29th week (Figure 2) In the successional classification, there was a predominance of pioneer and early secondary species with 35.48% each, followed by late secondary species with 6.45%, whereas 22.59% of the species were left unclassified because they were not identified at a specific level. Bowdichia virgilioides and Helicostylis tomentosa were the only late secondary species found in the present study.