Introduction

Epiphytism enables the enhancement of biodiversity in forests, providing the occupation of the different layers, creating environments susceptible for supporting life not dependent solely or directly from the soil (Madison 1977; Benzing 1990; Waechter 1992), becoming responsible for a significant part of diversity that makes the tropical rainforests one of the most complex ecosystems in the biosphere (Park 2003). It is estimated that 9% of all vascular plants are epiphytes (Zotz 2016), found almost exclusively in tropical and subtropical forests, and may represent in some countries more than 25% of flora species.

The sensitivity of epiphytic flora to moisture, associated with dependence on arboreal substrate, makes it a good environmental indicator for both successional stages of forest ecosystems and natural environmental variations (Triana-Moreno et al. 2003; Zotz and Bader 2009). For this reason, the communities of vascular epiphytes have been used as bioindicators of climate change, pollution and damage to ecosystems (Richter 1991; Lugo and Scatena 1992; Barthlott et al. 2001). This sensitivity becomes important in the context of conservation, given that many of the epiphytic species have a high degree of specialization with their phorophyte, thus becoming vulnerable to extinction by habitat change, climate change and, in the case of phorophyte-species specialists, by the phenomenon of coextinction (Clavel et al. 2011; Colwell et al. 2012). Nevertheless, the resilience of tropical epiphytes to environmental changes is still poorly understood (Larrea and Werner 2010).

The standards of richness, diversity and coexistence of vascular epiphytes have been studied in multiple approaches. The preference for phorophyte, interspecific interactions, soil conditions, microclimate variation, niche partitioning and dispersal limitation (Benzing 1990; Burns and Zotz 2010; Boelter et al. 2014; Zotz 2016) has been shown to be prominent in explaining these patterns found. Some studies associate high diversity observed in neotropical forests with attributes such as the age of the phorophyte and its physical characteristics such as branching and bark roughness, favoring water retention. These attributes tend to harbor a greater abundance of lichens and bryophytes, which results in positive interactions with many taxa of vascular epiphytes (Callaway et al. 2001, 2002; Hietz 2005).

The altitudinal and topographic variations observed in tropical forests generate different environmental conditions, featuring different microclimates for the vascular epiphytic guild (Woods et al. 2015). This environmental heterogeneity generates luminosity gradients (Benzing 1990; Clark et al. 1998; Cardelús et al. 2009) as well as moisture gradients (Cardelús and Chazdon 2005; De Le Rosa-Manzano et al. 2014). However, structural and successional aspects of arboreal vegetation are also important in the structuring of the epiphytic community, as these may be strongly influenced by the composition of the local arboreal assembly (Hietz and Hietz-Seifert 1995; Burns and Zotz 2010).

Our hypothesis is that the variation of microhabitats affects the richness of the vascular epiphytic guild; thus, this study aimed to analyze the distribution of vascular epiphytes in different environments of a subtropical forest in southern Brazil.

Materials and methods

Study area

– The study was conducted between 2010 and 2012 at the Serra Furada State Park (PAESF—Parque Estadual da Serra Furada), an Conservation Unit, located in the municipalities of Grão-Pará and Orleans, in the state of Santa Catarina, Brazil. The climate according to Köppen is Cfa (Alvares et al. 2013). PAESF has a vegetation formation characterized as subtropical rainforest, involving montane and upper montane formations, and the present study is performed on montane typology, according to IBGE (2012), at 400–1000 m altitudes.

Sampling method

– To sample the vascular epiphytic flora, each tree (phorophyte) was considered a sample unit, defined by the method of quadrants (Cottam and Curtis 1956). Six areas were established, and for each one, five point-centered quarters were determined, separated 20 m apart, totaling 120 sample units. The areas A–F were allocated in order to include different microhabitats (Fig. 1), with area A located at an altitude of 480 m and area F at 660 m. Areas A and B suffered anthropic interference, especially due to grazing in its surroundings and easy access, and an abundance of climbing plants was found by Oliveira (2016), in middle stage of natural regeneration. The other areas (C–F) are in an advanced stage of natural regeneration. Areas C and D are located parallel to watercourses, near rivers with waterfalls, with humid environments due to evapotranspiration; E and F are less humid, away from streams.

Fig. 1
figure 1

Sample units location (A–F) in the Serra Furada State Park, southern Brazil

Full size image

The criteria for including the phorophyte in the sample were to have a diameter at breast height (DBH) ≥10 cm. The epiphytes were sampled in two regions of the phorophyte, stem and crown, following the zonation proposed by Johansson (1974). The occurrence of epiphytic species in the stem and crown was recorded, by binoculars and photographic camera, for the presence (1) or absence (0) of the species. To estimate the distribution of each epiphytic species, procedures used by Waechter (1992) were adopted, based on the occurrence of individual phorophyte species.

The identification of species was based on Hoehne (1942, 1945, 1949, 1953), Pabst and Dungs (1975, 1977) Reitz (1983), Tamashiro and Zickel (1991), Guimarães (1998), Coelho (2000), Wanderley and Martins (2007), Azeredo and Citadini-Zanette (2012) and specialists. For the botanical families of angiosperms, the classification system adopted was APG IV (2016) and Smith et al. (2006) for ferns. The collected source material was deposited in the Herbarium Pe. Dr. Raulino Reitz (CRI) of Universidade do Extremo Sul Catarinense (UNESC), Criciúma, Santa Catarina.

Data analysis

– To assess the relation between the taxa and microhabitats, samples were submitted to correspondence analysis (Kent and Coker 1992), that may be applied to any data table that is dimensionally homogeneous and only contains positive integers or zeros (Legendre and Legendre 2003). The matrices of species and families were constituted by the number of occurrences in the area; therefore, we adopted the exclusion of the species under three sampled individuals and families with less than three species, preventing accidental species. A category called “no species” was created for those phorophytes who did not have vascular epiphytes (Oliveira et al. 2013; Padilha et al. 2015). By previous successful analysis, Bromeliaceae family was separated, based on their nutrient absorption mechanisms: tank bromeliads and atmospheric bromeliads (Benzing et al. 1978). Venn diagrams were made for species and families, considering the category “no species” in the analysis.

Results

A total of 46 species, distributed in 12 families, were recorded. The most representative families were Bromeliaceae (23.9%), Orchidaceae (17.4%), Polypodiaceae (15.2%) and Cactaceae (10.9%) (Table 1).

Table 1 Frequency of families and species in each area (A–F) in a subtropical forest in southern Brazil
Full size table

The areas A, B showed lower richness of vascular epiphytes in comparison with the areas C, D and E, F (Fig. 2) and the correspondence analysis formed three groups, based on frequency of families and species in each area. It was possible to see that in the cluster, most species tended to approach areas C, D.

Fig. 2
figure 2

Ordination diagram produced by correspondence analysis of species (a) and distribution of botanical families in the sample areas (b), in subtropical forest in southern Brazil. The species are represented by their acronym (four letters for the genus and four for the specific epithet)

Full size image

The Venn diagram shows overlapping families and species (Fig. 2), and 42% of the species are common among all areas (Fig. 2a) and areas C, D had the greatest number of unique species (19%). Regarding families (Fig. 2b), 62% are common among all areas, and areas C, D and E, F showed the highest numbers of common families (31%). No family was exclusive in the areas.

Discussion

The number of species found in this study is consistent with the expected pattern for tropical and subtropical forests of southern Brazil (Kersten 2006; Blum et al. 2011), as well as the taxonomic distribution of species in a few families, mainly in Orchidaceae, Bromeliaceae and Polypodiaceae (Zotz and Bader 2009; Freitas et al. 2016). The lower richness presented in areas A-B can be attributed to the conditions of microhabitat (availability of water, light, humidity, temperature, substrate stability, mineral nutrition and toxicity) (Wagner et al. 2015) provided by phorophytes present in the area. This condition can be associated with the degree of conservation in the areas, as vascular epiphytes respond differently to microclimate variations within the forest, which is considered crucial to the pattern of distribution of vascular epiphytes (Wagner et al. 2015). Studies point to the pattern of epiphytic colonization related to low diametric distribution of phorophytes (Zotz and Bader 2009; Wagner et al. 2015), as found in the areas A, B for this study. As the phorophyte diameter increases, the colonization by epiphytes rises proportionally (Ribeiro 2009). According to our results, we accepted the hypothesis that the variation of microhabitats affects the richness of the vascular epiphytic guild.

The great ecological amplitude of Bromeliaceae in this study, demonstrated by the distribution of species in ordination axes (Fig. 3), shows the relation between the degree of conservation (successional stage) and the presence of epiphytes. Bromeliaceae stands out in the secondary vegetation, as it is associated with propagule dispersal strategies (Cascante-Marin et al. 2006) or physiological adaptations, especially in atmospheric-habit species (Hietz and Hietz-Seifert 1995; Hietz et al. 2006).

Fig. 3
figure 3

Venn diagrams with the overlap of vascular epiphytic species (a) and families (b) between the sampled areas in subtropical forest in southern Brazil

Full size image

Reyes-García et al. (2008) studying the differentiation of niches in epiphytic bromeliads, showed an association with areas that suffered some disturbance, justifying the pattern found in this study for the areas A, B and for heliophytic species: Tillandsia aeranthos, T. stricta, T. usneoides, Vriesea carinata, V. flammea and V. vagans (Reitz 1983).

The ecological amplitude observed in Microgramma squamulosa in southern Brazil has granted to this plant the ability to survive in a range of environmental conditions and achieve high abundance in secondary vegetation (Borgo and Silva 2003), showing its high ability to support microclimate variations, found also in open areas, as in this study, and fragment edges.

Thus, the microclimate homogeneity observed in areas of average natural regeneration stage, when compared to areas at an advanced stage, shows a direct function of low specific richness found in areas A, B. Cascante-Marin et al. (2006) when studying communities of epiphytic bromeliads in Costa Rica found that the structure of the community cannot be taken only by physiological and/or morphological attributes, and emphasizes the aspects limiting the dispersal in successional stages caused by seed dispersal and availability. In this perspective, the history of life, the differences affecting the production and the location of seed sources, as well as local wind, may lead to a non-random dispersal. This justifies, in the present study, the number of “no species” (n = 12) phorophytes in areas A, B, and according to Cascante-Marin et al. (2006), the change in diversity and composition of species is expected during forest succession, according to factors specific to the location, affecting seed dispersal and availability, added to the performance and survival of the individual. Geraldino et al. (2010) reiterates that forest disturbances result in significant impacts on the community of vascular epiphytes, which can reflect on the ecological processes where they participate.

The separation of areas C, D in the ordination diagram (Fig. 3) reinforces the hypothesis (Larrea and Werner 2010) that heterogeneous environments are essential in the coexistence of vascular epiphytes, resuming the discussion that species exhibit different coexistence strategies, which somehow contributes to the community structure, indicating the influence of deterministic factors (separation of niches by environmental gradients), and stochastic events (colonization processes) (Hiura 2001), in structuring communities without establishing the relative importance of each one (Leibold and McPeek 2006). Thus, it is assumed that the higher the microclimatic heterogeneity within the forest habitats, the greater will be the epiphytes coexistence, which supports the richness obtained of 40 and 34 species for C and D areas, respectively, and also by the high frequency values obtained (187 and 147), as well as 11 and 12 of the 13 families used in correspondence analysis (Fig. 3). The greatest richness in these environments (C, D) may be related because of the proximity to rivers, which have many waterfalls and obstacles, generating the sprinkling that increases the availability of water for these plants (Bonnet et al. 2010) or because it is an area with low human impact and then there are older trees (Woods et al. 2015).

On a local scale, as in this study, the differentiation of microclimates is seen as one of the main mechanisms in the structure of vascular epiphytes communities (Hietz et al. 2006; Woods et al. 2015; Dislich and Mantovani 2016). The presence of watercourses at the sampling site of areas C, D promotes an increase in moisture and consequently the differentiation of habitats, which offers milder conditions for the establishment of vascular epiphytes, considering that water is the main factor related to vegetative growth (Bonnet et al. 2010; Boelter et al. 2014). The influence of moisture as a structural strength of the community in areas C, D can be evidenced by the number of unique species, these associated with Hymenophyllaceae, Orchidaceae, Dryopteridaceae and Araceae, the latter extending to the areas E, F (Figs. 2, 3). The pronounced effect of moisture on the richness and preference for habitat can be seen in the reduction of occurrence of taxa from Hymenophyllaceae and Selaginellaceae toward the extremes of the studied toposequence (areas A, B and E, F), which constitute hydrophilic taxa and extremely sensitive to habitat disturbance (Barthlott et al. 2001).

The continued toposequence separating areas E, F reveals an environment where the moisture factor ceases to be predominant, since the sample units, on average, are distanced 100 m from the watercourses. On the other hand, there was evidence of a significant increase in the mean diameter of the stems of phorophytes sampled for the areas E (42.0 cm) and F (31.3 cm) (Padilha et al. 2015). Woods et al. (2015) studying the rainforests canopies of Costa Rica noted that with increasing heterogeneity of microhabitat in the same phorophyte, there is an increase in richness and abundance of epiphytic species. The main reason for the heterogeneity is caused by the increasing height of the trees, in addition to the increase in diameter and the expansion and architecture of their canopies, where the ones presenting the highest number for these items had greater diversity of microhabitat and epiphytes (Woods and Dewalt 2013). This fact leads to the appearance of Araceae, Cactaceae, Gesneriaceae and Piperaceae. Larrea and Werner (2010), as they studied the response of vascular epiphytes to different intensities of land use, observed that Araceae was the only family that responded positively to the use of the landscape, not occurring in different areas of unmanaged forests. Barbosa (2005) points out in his study that Araceae family epiphytes have preference for certain characteristics of phorophytes such as larger diameters and rough rhytidome.

Woods and DeWalt (2013) showed preference in the colonization of habitats by species of Araceae, Gesneriaceae and Cactaceae in forests over 55 years old, which corroborates the data from this study and the hypothesis of the influence of height, diameter of phorophytes and water viability on the structure of an epiphytic community.

Faced with the different mechanisms that were observed acting in structuring communities, it is important to resume the discussion based on the data obtained, of which mechanisms are involved in vascular epiphytic communities, niche data (niche partitioning, regeneration niche) (sensu Silvertown 2004) or data related to dispersion (dispersal limitation, neutral model) (sensu Hubbell 2001). On the assumption that species occupy niches according to their ecophysiological characteristics, a prevalent issue is determining the environment characteristics, and consequently, the species that structure plant communities over time. It is assumed that species are not environmentally equivalent to justify the occurrence of complex patterns of vegetal diversity found in subtropical forests, as in this study, and therefore we believe the niche partitioning is one of the essential mechanisms in the structure of vascular epiphytic communities (Silvertown 2004; Silvertown et al. 2006; Dislich and Mantovani 2016).

It is shown that environmental and historical factors (Nakashizuka 2001) contribute to the structuring of epiphytic communities, especially (1) horizontal and vertical heterogeneity (branching and crown expansion, as well as height of the phorophyte) (Boelter et al. 2014; Woods et al. 2015; Dislich and Mantovani 2016), (2) disturbance regime (use of landscape, successional stages, anthropic disturbance) (Hietz et al. 2006; Larrea and Werner 2010; Woods and Dewalt 2013) (3) biotic interactions (Cascante-Marin et al. 2006; Clavel et al. 2011). Thus, our study shows that the differentiation of microhabitats caused by different successional stages and associated environmental variables, led to the coexistence of different species of vascular epiphytes. In this study, with the exception of biological interactions, these factors proved decisive in the coexistence of species in formations of subtropical forest in southern Brazil. The plants require the same environmental features: light, water, CO2, substrate and mineral nutrients; however, the ecological differences between vascular epiphytes are manifested in different ways to acquire these resources (Westoby et al. 2002).

We had evidence of processes based on the niche as a key mechanism in the structure of vascular epiphytes communities. The differentiation of niches is generated by microclimate gradients within the forest that may be associated with: the successional stage, in the case of areas with a history of land use by the heterogeneity of resources (e.g., water). Otherwise, the homogeneity of resources leads to the formation of structurally less complex communities, but no less important for conservation, since they keep the ability of evolution (successional replacement). According to our results, we accepted the hypothesis that the variation of microhabitats affects the richness of the vascular epiphytic guild.