Density and abundance of plantlets
The non-emergence of seedlings in any control tray allowed us to conclude that there was no contamination on the trays. Seasonality influences the soil seed bank. The soil seed bank obtained in the rainy season presenting higher density and greater abundance of seedlings than the soil seed bank obtained in the dry season. In tropical forests, some species have seasonal dormancy to avoid adverse periods, such as those of lesser water availability in the environment (Garwood 1989). In different savanna physiognomies, the greatest seed dispersion occurs at the end of the dry season and beginning of the rainy season, and in dense savannas diaspores are dispersed in greater abundance in the dry season (Salazar et al. 2012). For many savanna species, the seed germination is associated with the rainy season; once during this period, water is not a limiting factor for establishing of new seedlings, existing different strategies, such as the presence of dormancy in the seeds, dispersion agents, and period of dispersion, which may provide the germination in this season (Oliveira 2008; Salazar et al. 2011; Escobar et al. 2018).
Higher temperatures that occurred in the months from October and January in which the experiment was conducted, may have influenced the greater abundance in the rainy season. Plant species have different temperature requirements for germination; some may germinate in a wide temperature range; however, others germinate only at low or high temperatures (Baskin and Baskin 2014). Thus, changes in temperature throughout the different seasons may inhibit or stimulate the germination of several species, controlling the exiting or permanence of many seeds in the soil seed bank. Several savanna species have their optimum temperatures for germination between 20 ºC and 30 ºC (Zaidan and Carreira 2008; Brancalion et al. 2010; Escobar et al. 2018).
Light influences the soil seed bank. The shading (70%) provided greater seedling emergence from the soil seed bank, contrary to our initial hypothesis and the standard in forest environments with shading of 50% (Scherer and Jarenkow 2006; Menezes et al. 2019) and 60% (Batista Neto 2005). The canopy of the study area reduces the light entrance into the community, favoring the shade-tolerant species. We highlight that the shady treatment condition is more similar to the forest environment where the material was collected.
When we analyzed seasonality and light jointly, we observed in the soil seed bank obtained during the rainy season and maintained in the shady treatment, significantly greater abundance than of the other treatments.
Floristic composition
Seasonality influences the floristic composition, differing in the dry and rainy seasons. The presence of some species only in a season suggests that such species form transient soil seed banks and, those found in both seasons, form persistent banks (Thompson et al. 1997).
Light affects the floristic composition of emerged seedlings in the soil seed bank. Shady treatment provided greater richness in species in the soil seed bank, suggesting that most of these species are shade tolerant. The species Baccharis dracunculifolia, Cecropia pachystachya, Plathymenia reticulata, Solanum mauritianum, Solanum paniculatum, Trema micrantha, and Xylopia aromatica, classified as shade intolerant (Durigan et al. 2004), emerged in the soil seed bank maintained in shady treatment. The germination process depends on other environmental conditions, such as water availability, temperature, and oxygen concentration, factors that may have caused the germination of such species in the shady treatment (Kerbauy, 2008; Brancalion et al. 2015). The presence of dormancy in the seeds of P. reticulata, Siparuna guianensis, and X. aromatica (Escobar et al. 2018), may have contributed to the lower abundance of these species.
We found a higher number of species than recorded in the soil seed bank of savanna physiognomies (cerrado denso and cerrado sensu stricto) (Salazar et al. 2011), and grassland physiognomies (campo cerrado and campo sujo) (Salazar et al. 2011; Machado et al. 2013), probably due to the exposure of the soil seed bank to different light conditions, which allowed the germination of shade-tolerant and light demanding species.
In other environments, the soil seed bank also presented the Asteraceae and Poaceae families among those with the highest species richness (Gasparino et al. 2006; Figueiredo et al. 2014; Moressi et al. 2014). Many species of these families have small seeds, which facilitate their incorporation into the soil seed bank. Comparing the soil seed bank with the above ground community, the families with the highest species richness were Vochysiaceae, Bignoniaceae (Francisco 2020); Rubiaceae, Myrtaceae, Lauraceae (Francisco 2020; Santos 2020), and Fabaceae (Weiser 2007; Francisco 2020; Santos 2020). From these, we found only Fabaceae and Rubiaceae; however, the species presented in the soil seed bank not always are the same in the above ground vegetation (Thompson and Grime 1979; Townsend et al. 2010).
Diversity and similarity
Seasonality and light affected the diversity of the soil seed bank. The dominance of certain species influences the results of diversity and similarity in the different treatments. The soil seed bank obtained during the dry season presented greater diversity than that obtained during the rainy season. This is due to the greater species richness and high evenness, demonstrated by the Simpson concentration index, recorded during the dry season. In the rainy season, some species presented abundance superior to the other species, leading to a lesser evenness and impacting the diversity value found in this season. The diversity found in the dry season was the closest to that found in different above ground community components in the sampled (H’ = 2.84 (Santos 2020); H’ = 3.253 (Francisco 2020).
In relation to light, we found greater diversity in the shady treatment because all attributes that compose diversity (abundance, species richness and evenness) were greater in the shady treatment than in the light treatment.
As we jointly analyzed the soil seed bank collected in different seasons and the light and shady treatments, we found greater diversity in the soil seed bank collected in the dry season and maintained in light. Despite the lower abundance value, the Simpson concentration index obtained for the soil seed bank collected in the dry season and maintained in the light was the highest, with evenness being the factor that most contributed to diversity. The soil seed bank obtained in the dry season and maintained in the shady also presented high diversity, with evenness and richness being the attributes that most contributed to diversity. The soil seed bank collected in the rainy season and maintained in the light and shady treatments presented the lowest diversity values.
The similarity among the different treatments and collection seasons was high. The greater similarity (68%) occurred between the soil seed banks of dry and rainy seasons maintained in the shady treatment. The soil seed bank collected in the different seasons share several species; however, there are also species that will only be present in the soil seed bank in a single season. Such data indicate that the soil seed bank is influenced by seasonality, with different species being recruited to the area according to environmental conditions, such as temperature, amount of light that reaches the soil seed bank, and water availability in the environment.
We conclude that seasonality and light influence the abundance, species richness, and diversity of the soil seed bank. In rainy seasons, as well as in environments of shady condition, the soil seed bank presents greater abundance and species richness. Diversity is higher in the dry season and in environments with light condition.