Aggregated seed dispersal in a Neotropical coastal thicket vegetation: The role of microhabitat, dispersal syndrome and growth form

Abstract

Coastal thicket vegetation in Brazil (restinga) is characterized by three distinct microhabitats - thicket center, thicket edge, and open area - and the spatial pattern of seed dispersal varies among them. The objectives of this study were to differentiate microhabitats, considering dispersal syndromes and growth forms, and to determine the existence of seasonality in the seed dispersal for this ecosystem where the climate is seasonal. Seeds were monthly captured by seed traps during 24 months. They were counted, identified and dispersal syndromes and growth forms were determined. Number of seeds was low and most species were underrepresented in the seed rain. Abundance, richness, and diversity of seed rain were substantially higher in both thicket microhabitats than in the open area, probably due to the combined effects of seed production and seed retention in thickets, resulting in a non-random aggregated pattern of seed dispersal. Anemochory and zoochory were high in both thicket microhabitats, but in the open area anemochory was predominant. Herbs and shrubs contributed the greatest number of seeds and species in all three microhabitats. Our results showed a general aseasonal seed rain pattern at the community level, in contrast to the seasonal climate in this region. We conclude that seed dispersal face limitations in the plant community, natural regeneration potential is restricted, and degraded thicket vegetation will not be able to naturally regenerate the original biodiversity. Our findings considerably increase knowledge on the seed rain process at the community level and may contribute to the conservation and restoration in sandy coastal ecosystems.

Introduction

Coastal sandy ecosystems are composed of highly dynamic plant communities driven by several physical and biotic factors and responsible for protection against storms, waves and wind (Hesp, 2006; Maun, 2009). These ecosystems are subject to multiple harsh conditions (high sun exposure and evapotranspiration, extreme temperatures, frequent winds, and nutrient-poor sandy substrate), which favor aggregate distribution of plants in thickets, mainly in open-vegetation communities (Butterfield and Briggs, 2011; Scarano, 2009; Wright et al., 2017). Such characteristics create microenvironmental conditions that provide survival habitats for several flora and fauna species. Despite important environmental and socioeconomic functions that underpin ecosystem services, coastal plant communities are threatened by anthropogenic impacts, which results in habitat loss, lower biodiversity, and alteration of natural processes (Doody, 2013; Gómez-Zotano et al., 2017; Maun, 2009). Therefore, studies on the natural regeneration potential of coastal plant communities are an essential step in supporting restoration and biodiversity conservation initiatives.

The natural regeneration potential in a plant community greatly depends on seeds being dispersed to suitable sites for establishment or incorporated into the soil to become part of the seed bank (Chambers and MacMahon, 1994; Howe and Miriti, 2004). Both processes are the most important components of the regenerative niche (Grubb, 1977), especially in markedly heterogeneous habitats, such as those with two-phase vegetation structure because clumped dispersal in suitable microhabitats may benefit seedling establishment (Muller-Landau et al., 2002). In this sense, several studies on thicket communities in ecosystems worldwide have shown that seed dispersal occurs in a highly non-random aggregated pattern beneath or close to thickets, in relation to open areas (Caballero et al., 2008; Giladi et al., 2013; Pugnaire and Lázaro, 2000; Wenny, 2001). This may occur because thickets act as a “seed source” (greater seed production within thickets), providing “seed trapping” (accumulation of wind-dispersed seeds) or “seed discarding” (from frugivorous birds perching on branches) (Bullock and Moy, 2004; Caballero et al., 2008; Flores and Jurado, 2003; Giladi et al., 2013). By contrast, seed rain in open spaces between thickets is certainly distinct, due to sparse bird dispersal and predominance of wind-dispersed seeds (Cubiña and Aide, 2001; Kollmann, 2000; Teegalapalli et al., 2010). Besides, studies have shown that dispersal syndromes are related to plant growth forms, but the relationship varies depending on the habitats and it may influence the seed rain and the regeneration process (;Butler et al., 2007 Griz and Machado, 2001; Jara-Guerrero et al., 2011). Thus, dispersal syndromes and growth forms must be taken into consideration in studies on plant communities since they support important features of the functional diversity in ecosystems (Jara-Guerrero et al., 2011).

Climate seasonality and dispersal syndromes may also drive patterns of seed rain in plant communities in coastal sandy ecosystems worldwide (Castley et al., 2001; Rodriguez et al., 2017; Staggemeier and Morellato, 2011). Overall, phenological studies at the community level suggested that the greatest number of zoochorous species ripen their fruits mostly during the rainy season, when suitable water and temperature to maturation are plentiful. In contrast, the dry season favor seed rain by anemochorous species (Jara-Guerrero et al., 2011; Traveset et al., 2014). However, despite the more extreme microenvironmental conditions, phenological studies in the Brazilian coastal sandy ecosystems revealed aseasonal fruiting activity during the year, regardless of the degree of vegetation seasonality (Mendoza et al., 2017; Staggemeier and Morellato, 2011).

Covering most of the ca. 5000 km of Brazilian coastline, coastal sandy ecosystems (known as Restinga) support great biodiversity in a mosaic of plant communities. Most of the local flora is composed of plant species that migrated from the neighboring Atlantic forest to the geologically younger sandy plains (Rizzini, 1979; Scarano, 2009). These ecologically plastic species were able to establish and survive in the more extreme and seasonal drought conditions in the restinga habitats (Scarano, 2002). Similar to other shrub thickets worldwide, those in restinga face biodiversity loss due to anthropogenic stressors (Rocha et al., 2007; Scarano, 2009; Zimmermann et al., 2017), thus resulting in depletion of important ecosystem services (Groot et al., 2010; Mantovani, 2003). Despite an urgent need to restore the restinga ecosystems as much as possible (Freitas et al., 2016; Guerrero and Rocha, 2010; Scherer-Widmer, 2001; Silva et al., 2012), little is known about the natural regeneration potential of these plant communities (Braz and Mattos, 2010; Marques et al., 2015). Some studies suggest that restinga has potential for natural regeneration after disturbance (Garbin et al., 2018; Ribeiro and Melo Jr, 2016; Salimon and Negrelle, 2001; Simoes and Marques, 2007), however, this potential varies according to the type of plant community and dominance of regeneration modes (i.e., by sprouting or from seed) (;Cirne and Scarano, 2001 Simoes and Marques, 2007; Braz et al., 2014; Marques et al., 2015; Garbin et al., 2018). To our knowledge, no studies to date on restinga thicket communities have simultaneously evaluated seed rain patterns in response to spatial and temporal (including seasonal) variation, as well as their influence on species diversity, dispersal syndrome, and growth form.

We monitored the seed rain falling in traps for two years, comparing three microhabitats of a thicket community in a restinga to answer the following questions: (1) How do seed rain composition, abundance, and richness differ overall among the three microhabitats: thicket center, thicket edge, and open area? (2) How do abundance and richness vary according to dispersal syndromes (anemochory, autochory, and zoochory) and growth forms (herb, shrub, tree, and climber) of the seed rain in the three distinct microhabitats? (3) In the face of seasonal climate in several places along the Brazilian coast (Hoeltgebaum et al., 2018; Marques et al., 2015; Medeiros et al., 2007) and seasonal pattern of seed rain in some species (Begnini and Castellani, 2013; Zimmermann et al., 2017), is there also a pattern of seasonality in the seed rain of a plant community? Based on the foregoing, we hypothesized that: (1) Seed rain abundance and richness are greater in microhabitats below thickets than in the open area of the restinga plant community. (2) Due to the perch effect and composition of the thickets mainly by fleshy-fruited shrub species, zoochory is the main dispersal syndrome and shrubs are the main growth form in both thicket microhabitats. Anemochory and herbs are predominant in the open area, due to the wind that circulates easily among the small herbaceous species. (3) Due to the seasonal drought conditions in restinga habitats and seasonal responses related to dispersal syndromes in the neighboring Atlantic forest, we predict a seasonal seed rain pattern at the community level. This article is part of a larger study in a restinga plant community that also includes the evaluation of the soil seed bank, the seedling emergence, and seed germination, aiming to understand the different processes related to natural seed regeneration and contribute crucial information for plant conservation and ecological restoration in coastal endangered ecosystems.