Abstract. Canestraro, B. K. & R. A. Kersten. 2018. The slope does not influence understory community on a Brazilian Montane Atlantic Forest. Darwiniana, nueva serie 6(1): 5-23.

The understory is a component of great relevance to the flora and structure of tropical forests. Mountain environments, through physical factors, can show differences in the structure and composition of this stratum. We aimed to analyze whether there is a correlation between the slope and the understory plant community. A transition area between the Montane Dense Ombrophilous Forest and the Araucaria Forest in the Atlantic Forest in South of Brazil was analyzed. 63 plots of 1 m² were made, being equally subdivided into three slope classes (0-25%, 40-60%, and >80%). 93 species, 63 genera and 42 families were recorded. The richest families were Myrtaceae (15 species), followed by Lauraceae (7), Dryopteridaceae (6), and Rubiaceae (5). The most frequent life forms were arboreal (43 species), followed by herbaceous (24) and shrubby (11). 58 species were transient and 35 were resident. The richness and diversity indices showed no significant differences between the structure and composition of the slope classes. The ordination analysis could not distinguish any slope class grouping. This is possibly justified by the environmental homogeneity of the sampled area created by the density and regularity of the canopy, climatic stability, high organic matter and shallow soil.

Keywords. Araucaria Forest; Dense Ombrophilous Forest; floristics; lower stratum; phytosociology.

Resumen. Canestraro, B. K. & R. A. Kersten. 2018. La pendiente no influye en la comunidad del sotobosque del Bosque Atlántico Montano de Brasil. Darwiniana, nueva serie 6(1): 5-23.

El sotobosque es un componente de gran relevancia para la flora y estructura de los bosques tropicales. Los ambientes montañosos, a través de factores físicos, pueden reflejar diferencias en la estructura y composición de este estrato. El objetivo del estudio fue evaluar si existe una correlación entre la pendiente del terreno y la comunidad de plantas del sotobosque. Se analizó una zona de transición entre la Floresta Ombrófila densa y la Selva de Araucaria en el Bosque Atlántico en el sur de Brasil. Se delimitaron 63 parcelas de 1 m², subdivididas por igual entre tres clases de pendiente (0-25%, 40-60% y >80%). Se registraron 93 especies, 63 géneros y 42 familias. Las familias más ricas fueron Myrtaceae (15 especies), Lauraceae (7), Dryopteridaceae (6) y Rubiaceae (5). Las formas de vida más frecuentes fueron arbórea (43 especies), seguido de herbácea (24) y arbustiva (11). Se registraron 58 especies transitorias y 35 residentes. Los índices de riqueza y diversidad no mostraron diferencias significativas entre la estructura y la composición de las clases de pendientes. Los análisis de ordenamiento no distinguieron diferentes grupos para las clases de pendiente. Esto posiblemente se justifica por la homogeneidad ambiental del área muestreada generada por la densidad y regularidad del dosel, la estabilidad climática de la región, el alto contenido de materia orgánica y los suelos superficiales.

Palabras clave. Bosque Atlántico Montano; fitosociología; florística; Selva de Araucaria; sotobosque.

INTRODUCTION

The Dense Ombrophilous Forest, also known as Atlantic Rain Forest, is one of the main phytogeographic units of the Brazilian Atlantic Forest hotspot. It is located in the east portion of Brazil and of Paraná State from 0 to 1,650 m a.s.l., including the Serra do Mar mountain range. This forest type is directly influenced by the warm and humid air mass from the Atlantic Ocean and by the evenly distributed rainfall throughout the year, which provides a climate with thermal stability (Maack,1968; Leite, 2002; Roderjan et al., 2002). The Subtropical Atlantic Forest (or Araucaria Forest), integrating the Brazilian Atlantic Forest hotspot, is characterized by the coexistence of the tropical and temperate flora and by the components Araucaria angustifolia (Bertol.) Kuntze, Podocarpus lambertii Klotzsch ex Endl., Lauraceae and Myrtaceae. It is distributed on highlands (750 – 1,300 m a.s.l) westwards from the Serra do Mar (Maack,1968; Roderjan et al., 2002). The climate is always humid with low temperatures and frosts are common in winter (Leite, 2002).

Environmental factors are usually considered the main causes of the large diversity of plant communities, either for canopy and for understory (Pereira et al., 2007). Some of these factors are: edges (Pereira et al., 2007; Silva et al., 2008), light gradients (Poulsen, 1996; Záchia & Waechter, 2011; Enoki & Abe, 2004; Inácio & Jarenkow, 2008), soil moisture (Záchia & Waechter, 2011; Cardoso & Schiavini, 2002; Enoki & Abe, 2004), soil properties (Beatty, 1984; Roderjan, 1994; Záchia & Waechter, 2011), litter thickness (Beatty, 1984; Poulsen, 1996), temperature (Roderjan, 1994), and also topography (Poulsen & Balslev, 1991; Záchia & Waechter, 2011; Enoki & Abe, 2004; Pereira et al., 2007; Silva et al., 2008).

In order to understand the local flora and the participation of each environmental factor within the richness and structure of the community is of utmost importance to embrace distinct growth forms in phytosociological studies (Moro & Martins, 2011). In tropical forests, the lower stratum is an important element, both on diversity and ecology (Cestaro et al., 1986; Andrade, 1992; Kozera et al., 2009). Specific studies on the understory layer, such as the herbaceous and shrubby components, are significantly scarcer than those on the arboreal layer (Gentry & Dodson, 1987; Munhoz & Araújo, 2011). However, research on the non-woody understory component (herbs, shrubs and climbers) had increased substantially in recent years (Andrade, 1992; Silva et al., 2008; Polisel et al., 2014), mainly in the South of Brazil (Cestaro et al., 1986; Kozera, 2001; Müller & Waechter, 2001; Kozera & Rodrigues, 2005; Silva & Britez, 2005; Inácio & Jarenkow, 2008; Kozera et al., 2009; Blum, 2010; Záchia & Waechter, 2011). Fine-scale heterogeneity is still ignored in the study of community response to gradients and available data remain insufficient for the understanding of abundance, distribution, richness, and diversity of species.

Studies on the flora of the mountain regions and at different geographical scales become interesting due to the gradual phytophysiognomy change over the altitudinal gradient. Variations in topography result in differences in the physiognomy and structure of interior habitats (Grytnes & Vetaas, 2002; Carvalho et al., 2007; Silva et al., 2008). Considering a smaller scale, the microrelief can lead to a mosaic of soil and to a vegetation pattern in the understory (Beatty, 1984). Specific researches on the relationship between the topography and the understory community are very scarce (Poulsen, 1996; Záchia & Waechter, 2011). Some studies have analyzed the arboreal and shrubby community (Blum, 2006) and the herbaceous group (Blum, 2010) along the altitudinal gradient in a Mountain range in Paraná State; slope however was just mentioned on descriptions but never taken in consideration by researchers. On the other hand, the studies of Enoki & Abe (2004), Pereira et al. (2007), Silva et al. (2008) and Záchia & Waechter (2011) analyzed the topography as a variable within the ecosystem. Thus, there are no studies emphasizing the role of the microrelief on the lower stratum community in forests of the mountain regions of Brazil.

Therefore, the hypotheses of this study are: a) the understory community is influenced by microrelief and show distinct structure, composition and diversity on different slope classes and b) the spatial distributions of plant species are related to particular topographic features.

MATERIALS AND METHODS

The “Serra do Mar” region is the elevation of the crystalline complex which predominates rocky outcrops from Pre-Cambrian period and slopes strongly dissected by the drainage that forms different sizes of valleys (Maack, 1968). The relief is mountainous with steep slopes, rocky shallow soils (Entisoils Orthend, Histosoils and Inceptisoils - Soil Survey Staff, 1999), rich in organic matter and with heterogeneous litter thickness (Roderjan, 1994; Portes et al., 2001; Blum, 2010).

The study was conducted on the western hillside of Anhangava Mountain (1,432 m), the summit of the “Parque Estadual da Serra da Baitaca” on the “Serra do Mar” mountain range, Paraná State, Brazil. The studied area is restricted to the middle third hillside of the mountain and presents a dissected microrelief (Figs. 1 and 2). The vertices of the area are point A (25°23’18.9’’S; 49°00’36.8’’W), B (25°23’20.3’’S; 49°00’37.1’’W), C (25°23’20.5’’S; 49°00’37.7’’W) and D (25°23’18.8’’S; 49°00’39.6’’W), totaling 3,350 m² and varying from 1,086 m a.s.l. at point D to 1,108 m a.s.l. at point B (Fig. 1). The studied area has two convergent drainage canals, one of which is a long and deep seasonal waterway (section A1 and point D in Fig. 1, 2 and 3). Close to the area there is a perennial stream. The points A and B are placed in a 1.5 m wide path towards the mountain peak with closed canopy (Fig. 1). Longitudinal sections of relief profile can be seen in figure Fig. 3.

The climate is classified as Cfb (Koeppen’s classification) (Roderjan, 1994). The mean annual temperatures of the coldest month vary between -3°C and 18°C and in the hottest month from 10°C to 22°C, with high rates of evenly distributed annual rainfall (Roderjan, 1994).

The vegetation is classified as Montane Atlantic Rain Forest (Montane Dense Ombrophilous Forest) (Roderjan et al., 2002). However, considering the short distance and the altitude, the environment is directly influenced by the Araucaria Forest (which extends until 1,000 m), and therefore the area should be considered an ecotone (Roderjan, 1994; Portes et al., 2001). The area is in the interior of an extensive forest fragment, in advanced regeneration stage and has no evidence of recent deforestation or selective logging (Figs. 1 and 3).

The following plant life forms were considered components of the understory: herbs, shrubs, juvenile trees and climbers (not connected to the phorophyte). The species were classified into two ecological groups (residents and transients) according to their typical stratum of permanence. Juvenile individuals which have temporary permanence in the lower stratum and occupy higher stratum when mature (trees and climbers) were classified as transients. Otherwise, species that normally complete their cycle in this stratum (herbs or small shrubs) were classified as residents (Gilliam et al., 1994). Both groups are typical of the lower stratum, although the transients are not exclusive to it (Gilliam & Roberts, 2003).

For each species the following characteristics were determined: physical habitat (terrestrial or epipetric), ecological habit (transient or resident) and the biological forms: herbaceous (erect, rhizomatous, creeping, rosulate, and caespitose), shrubby (erect, scandent and rosulate), arboreal, climber, and accidental herbaceous (rosulate), according to Mueller-Dombois & Ellenberg (1974). Accidental herbs are components of the lower stratum because they use the soil in some parts of their life cycle (which is commonly epiphytical) and can also be classified as facultative terrestrial herbs (Poulsen & Balslev, 1991).

Fertile individuals were collected, preserved and herborized according to usual methods and deposited in the Municipal Botanical Museum of Curitiba (MBM) herbarium. Specimens were identified to the lowest taxonomical level (when possible), using specialized literature, specialist assistance and consultancy from MBM’s collection. Indeterminate taxa were treated as distinct morpho-species. The scientific names, authors and classification system followed BFG (2015) and Prado et al. (2015).

In order to analyze the influence of the slope on the structure and composition of species, we delimited three classes of slope: 0-25 %, 40-60 % and > 80%. We did not perform analyses on slopes outside of these classes’ intervals. The slope rate was determined through a levelling rod, a wood stick and a level. Each plot comprised 1 m² of area (Munhoz & Araújo, 2011) and the measurements were within this area. As the plots had a fixed length (100cm), the height of the declivity (in cm) is equal to the percentage of the slope. We sampled 21 plots in each class of slope, with a total area of 63 m².

The multiple plots system (Mueller-Dombois & Ellemberg, 1974) was used. The 1m² quadrats were placed randomly and directly on the ground of the forest, following the relief topography and were positioned with at least 2 m between each other. Due to the constant shift on microrelief and the small length of the considered slopes it was not possible to follow any systematic sampling scheme. All individuals between 20 cm and 2 m height and with perimeter at breast height (1.3 m) less than 5 cm were included; for tree ferns the only criterion was stature. Woody individuals below 20 cm were treated as seedlings and ignored by the analysis due to their uncertain identification. The relative plant cover (%) was estimated by each species’ vertical projection of occupation including those individuals non-rooted inside the plot, but with cover projection inside of it.

The phytosociological parameters collected for each species were frequency, leaf cover and dominance. The importance value (IV) was calculated based on the relatives’ frequency and dominance (Mueller-Dombois & Ellenberg, 1974). The community diversity was calculated through the indices of Evenness (J), Shannon (H’) and Simpson (1/D and 1-D) (Magurran, 2011) and the richness was calculated by the Margalef (Dα) and Chao-1 indices (Chao, 1984). Those calculations were made using PAST 3.0-ORD (Hammer et al 2001) when software allowed and graphs with error bars were displayed. In order to compare diversities, we applied the diversity profiles method using the on-line version of the iNext R-package (Chao et al, 2016). To understand diversity through a single index is to project multidimensional complexity onto a one-dimensional ordinal scale. One way to surpass this problem is to quantify diversity by means of diversity profiles which is a curve showing the simultaneous values of a large collection of diversity indices. (Patil, 2014)

Considering species abundance on the plots, we performed an ordination analysis using Detrended Correspondence Analysis (DCA) downweighting rare species (Hill & Gauch, 1980). We used MRPP and PerMANOVA as an “a posteriori test” for all slopes classes, and for making the paired (two-by-two) comparisons (Anderson, 2005). Species with single occurrences (singletons) were excluded in both cases. The calculations were made using the PC-ORD program, version 6.0 (McCune & Mefford, 2011). We also compared floristic composition similarity using the Jaccard Index (IJ) (Mueller-Dombois & Ellenberg, 1974) and a modified version of this index, excluding the singletons.

RESULTS AND DISCUSSION

Floristic composition
We recorded 354 individuals belonging to 93 species, 63 genera and 42 families (Tab. 1). Three taxa were identified to family level (Myrtaceae) and four taxa to genus level due to their sterile state. From the total, 75 species, 48 genera and 33 families were Angiosperms and 18 species, 15 genera and 9 families were ferns. Lycophytes and Gymnosperms were not found. The richest families were Myrtaceae (15 species), Lauraceae (7), Dryopteridaceae (6) and Rubiaceae (5) and their sum represents 35% of the total richness (Tab. 1). Myrtaceae, Lauraceae and Rubiaceae were the most representative families on the Araucaria Forest (Kozera et al., 2006; Scheer & Blum 2011; Lima et al.,2012; Ribeiro et al., 2013) and also on a transitional area between Dense Ombrophilous Forest and Araucaria Forest (Reginato & Goldenberg, 2007).
The richest herbaceous family was Dryopteridaceae, as observed by Kozera & Rodrigues (2005), Kozera et al. (2006), Inácio & Jarenkow (2008), Paciencia (2008), Kozera et al. (2009), Blum (2010) and Lima et al. (2012). The most representative genera were Ocotea (6 species) and Eugenia (5), as observed by Ribeiro et al. (2013) in a fragment of Araucaria Forest.
The arboreal and shrubby strata were composed of taxa of both Araucaria and Dense Ombrophilous Forest (Roderjan et al., 2002; Blum, 2006; Scheer & Blum, 2011). While Cinnamodendron dinisii, Casearia decandra and Eugenia uniflora are typical of Araucaria Forest, Pouteria bullata, Myrsine umbellata, Guatteria australis and Cyathea phalerata represent the Dense Ombrophilous Forest type (Scheer & Blum, 2011).

Habitats and biological forms
Considering physical habitats, 85 species were exclusively terrestrial, four were exclusively epipetric and four used both substrates (Tab. 1). Regarding the biological forms, the arboreal type was represented by 43 species (46%), followed by herbaceous (including herbaceous/climber) with 24 species (25.8%), shrubby with 11 species (11.8%), arboreal/shrubby with 10 species (10.7%), exclusively climber with four species (4%) and only one accidental herbaceous epiphyte (1%) (Tab. 1). A similar proportion of biological forms was found by Gentry & Dodson (1987) in a tropical wet forest in Ecuador and by Andrade (1992), Kozera et al. (2006) and Lima et al. (2012) in the Atlantic Forest hotspot, excluding epiphytic species (Kozera et al., 2006; Lima et al., 2012). The low contribution of climbers was an evidence of a forest with a perennial and continuous canopy, as observed by Polisel et al., 2014 in an Araucaria Forest.
The most common habit of herbs was rhizomatous (11 species), followed by rosulate habit (7) (Tab. 1) as observed by Kozera & Rodrigues (2005) and Záchia & Waechter (2011) in different Atlantic Forest physiognomies. Among the obligatory herbs, 67% were composed of ferns and only 36% were angiosperms, in contrast to the proportion of ferns and flowering plants found by Kozera et al. (2009) and Blum (2010) on Dense Ombrophilous Forest and Záchia & Waechter (2011) in a coastal Atlantic Forest. Within the woody species, the erect biological form represented most records not only for trees (43), but also for shrubs (7) and trees/shrubs (10), as observed by Kozera & Rodrigues (2005) in a coastal Atlantic Forest.
Among the five species with the highest Importance Value (IV), four were herbs and sum 33.1% of the IV. Within the 20 species with highest IV, nine were herbs and sum 45.6% of the IV, whereas ten species are trees and shrubs and sum 25.5% of the IV (Tab. 1 and 2). It is possible to notice a gradual increase in the richness and in the IV of herbs from the flat (class 1) towards the slope (class 3) slope (Tab. 1 and 2). Záchia & Waechter (2011) observed this same pattern in a coastal forest.
Fifty-eight species (62%) were classified as transients and 35 (37%) as residents (Tab. 1). According to Kozera et al. (2006), in Ombrophilous Forests, the arboreal stratum (transient habit) is the most representative component of the community. Gomes-Westphalen et al. (2012) studying the Atlantic Forest understory, have also found a considerable amount of transient species (regenerative strata). However, among the 20 species with the highest IV, 13 were resident and seven were transient (Tab. 2), showing that the most significant species are resident, while the transient component is still rich, yet less expressive. The four most important species in class 3 (>80%) were resident (Tab. 2).

Community Similarity and Diversity
Pseuderanthemum riedelianum was the most important species with the highest IV within all three slope classes, due to its high frequency and dominance (Fig. 4). Parodiolyra micrantha and Ctenitis paranaensis featured between the six highest IVs; however, the species appeared in distinct orders within the slope classes (Fig. 4). Another species of Ctenitis (C. distans) also played an important role in the structure of an Araucaria Forest fragment, as seen by Rigon et al. (2011). Pseuderanthemum riedelianum and Parodiolyra micrantha have also exhibited relevant participation in the Submontane Dense Ombrophilous Forest communities in the study of Kozera & Rodrigues (2005). Poulsen & Balslev’s study (1991), on the understory herbs of the Amazon Forest, showed that a Poaceae (Pariana radiciflora) had the highest cover value, indicating that grasses are important components of the lower stratum, as well as Parodiolyra micrantha’s performance in this study. The tree Rudgea jasminoides was very important in this study (Tab. 2) and also presented a high IV in different Montane Araucaria Forest remnants as analyzed by Ribeiro et al. (2013). Were recorded 47 singletons species (50%) (found only once on the survey) and 10 doubletons. Rubiaceae was the most frequent family in the lower stratum for both Araucaria Forest and Dense Ombrophilous Forest, agreeing with a forestry survey by Roderjan et al. (2002).
In steep slopes (>80% slope), the fern Megalastrum abundans accomplished the highest dominance (15%) (Tab. 2), possibly due to its large fronds, which increase cover value. The ferns were among the species with the highest cover values (Tab. 2), due to their compound and large leaves, when compared with most of angiosperms leaves. The class > 80% of slope showed high leaf cover index by herbs and especially by ferns, probably due to its short and narrow roots that establish well on slopes, as observed by Záchia & Waechter (2011) (Tab. 2).
In general, the most frequent families were Acanthaceae (54 from 63 plots), followed by Rubiaceae (45), Dryopteridaceae (42) and Poaceae (31). Of the total, the five most important species corresponded to 38% of the Importance Value whereas 80% of the IV was achieved by the 29 main species. There were species exclusive to each class; however their IVs were not sufficiently relevant to show significant differences between the communities.
The richness was very similar among the slope classes, in spite of the fact that the Chao-1 indicated highest richness for class 3 (>80%) (Tab. 3, Fig. 5). In contrast, Silva et al. (2008) found distinct richness rates comparing three understory topographical sites. The dominances were also similar (Tab. 3). The number of sampled individuals was higher in class 40-60% (129) and lower in class 0-25% (105) (Tab. 3). The leaf cover index was lower in class >80% (31) and higher in class 40-60% (41.7) (Tab. 3).
The diversity indices showed structural similarity between the classes of slope (Tab. 3, Fig. 5). The other analyzed indices were not capable of showing differences between the community classes, however the diversity was higher in class >80%, due to its higher richness and evenness and lower dominance (Tab. 3). The Simpson’s index average (1-D) was 0.95 (increasing scale from 0 to 1), emphasizing the high diversity of the sampled area, despite the similarities between the classes (Tab. 3, Fig. 5). Záchia & Waechter (2011) found higher diversity in steep slopes due to the greater richness and evenness of the community in a costal environment. Silva et al. (2008), studying the edge vs. the interior understory community, reported homogeneous diversity indices in different topographical interior sites, however the lower slopes were more diverse, followed by the terraces.
The diversity profiles also indicated that the slope communities were very similar on species richness (q=0), Shannon Diversity index (q=1) and Simpson Diversity Index (q=2) (Fig. 6).
The ordination analysis (DCA) indicated that the plots were very similar despite the slope classes and that the groups were completely overlapped (Fig. 7). The low A-value on MRPP (A = 0,011) indicates that heterogeneity within groups equals expectation by chance and  PerManova indicates that only 2.7% of variance is explained by slope classes variable.
The plots on class 0-25% (red triangles in Fig. 7) presented higher heterogeneity, whereas the plots on class 40-60% (green circles in Fig. 7) were more homogeneous. The analysis showed the presence of several generalist and broadly distributed species within the plots, in other words, the sampled area was considered statistically homogeneous. The great distance between the points indicates distinct structural and floristic communities.
Paciencia (2008) analyzed the ferns’ and lycophytes’ community in altitudinal gradients in the “Serra do Mar” Mountains in Paraná. This study showed, by another correspondence analysis (NMDS), that the community was influenced by altitudinal variations among the gradients; however those localities with low altitudinal variations had similar compositions and structures.
Low floristic diference was also observed among slope classes. The Jaccard Indexes were high (0-25% - 40-60% = 0.37; 0-25% - >80% = 0.35; 40-60% - >80% = 0.39), but lower than expected due to the high number of singletons (50%). When those were excluded, the indexes were much higher (0-25% - 40-60% = 0.64; 0-25% - >80% = 0.60; 40-60% - >80% = 0.65).

Slope and abiotic factors
The relief of the Montane Dense Ombrophilous Forest involves steep slopes, contributing to shallow soils with little effective depth (Roderjan, 1994; Roderjan et al., 2002; Blum, 2006). The studied area shows a thick layer of litter, especially in flat areas, which can be related to the low temperatures of the region that slow down the organic matter decomposition in forest environments (Beatty, 1984; Blum, 2006). The relief heterogeneity creates distinct microenvironments for the understory community (Whitmore, 1990 apud Blum 2010); however, the small size of our sampled area did not capture this feature. Forest shallow and poor developed soils (“Entisols”) and high organic matter combined with homogeneous shading and a regular 22 m-high canopy can provide undifferentiation within the understory community (Beatty, 1984; Roderjan, 1994; Portes et al., 2001; Scheer & Blum, 2011).
The effect of this homogeneous and dense canopy (Figs. 1 and 2) is understory protection against high solar radiation, wind and air masses exposure and rain impact, preventing abrupt variation in soil humidity and temperature (Roderjan, 1994), thus contributing to less niche diversification and herb biodiversity (Blum, 2010; Polisel et al., 2014). In addition, the understory environment and the light availability strongly affect the herbaceous layer (Beatty, 1984; Poulsen & Balslev,1991; Gilliam & Roberts, 2003; Silva et al., 2008; Blum 2010) and the regenerative stratum (Enoki & Abe, 2004). The inconspicuous difference among diversity indices are evidence of this environmental homogeneity (Tab. 3) (Pereira et al., 2007; Inácio & Jarenkow, 2008; Silva et al., 2008). The meaningless variation in the herbaceous layer structure is a result of the perennial canopy of the forest, since the low solar incidence in the understory doesn’t cause differentiation on the structure and composition of the lower stratum (Enoki & Abe, 2004; Inácio & Jarenkow, 2008; Blum, 2010). Furthermore, the invariable climatic factors such as air humidity, temperature and rainfall are responsible for the climatic stability of the Anhangava Mountain region as analyzed by Roderjan (1994).
However, Poulsen & Balslev (1991) found correlation between the relief and the composition of the understory herbaceous layer in a tropical forest and Záchia & Waechter (2011) successfully associated the understory community and the topography in a coastal Atlantic Forest in South of Brazil; both studies analyzed small scale topography (variations between 8 to 19 m). Záchia & Waechter (2011) evaluated different variables as luminosity and soil composition and humidity, thus reflecting distinct environmental conditions and plant communities. Silva et al. (2008) demonstrated the edge effect is more responsible for the diversification of the composition and structure of the understory community than the topographical gradient in an area of 100m elevation range. Nonetheless, Pereira et al. (2007) analyzed both edge effect and topography related to tree community and presented a correlation between tree distribution and topography, probably due to its influence on soil fertility and its properties and ground-water regime. Pereira et al. (2007) studied a mountainous area between 800 and 1,000 m.a.s.l. Daws et al. (2002) stated that the environmental heterogeneity resulting from microtopography variation could explain the high species richness in a tropical forest in Panama, by providing distinct hydrologically defined niches. These authors analyzed the correlation of the microtopographic position (at a small scale) and the water regime of a river in plateaus and slopes. However, in Anhangava sampled area the plots were not influenced by water bodies.
Although some authors had found that the distribution of understory species was influenced mainly by microtopography (Beatty, 1984) and that variations in topography result in diversification in the structure of interior habitats (Carvalho et al., 2007; Silva et al., 2008) we found no difference among the slope classes. Assuming these findings, the present study refutes the proposed hypotheses, probably due to the specific environmental characteristics found in the studied region on Anhangava Mountain.

CONCLUSIONS

The richness of young trees reveals the relevance of the transient component in the lower stratum and in the diversity composition of forest areas. Herbaceous individuals are less frequent, yet with great family number. Among the species with the highest Importance Value, most are resident, however the two richest families are exclusively transient.

Despite the inconspicuous higher diversity observed in the slope class >80%, the analyses of diversity, richness and DCA show the differences of structure and composition are not significative and there is no correlation between the slope and the understory plant community, thus refuting the two proposed hypotheses. Furthermore, less than 3% of the variance is explained by the slope (PerMANOVA test). This lack of correlation with the slope variable probably occurs due to the influence of the climatic and environmental homogeinity in a well-preserved fragment in the Atlantic Forest of South of Brazil.

ACKNOWLEDGMENTS

We thank Eduardo D. Lozano, Marcelo Brotto, Vanessa Ariati, Felipe Marinero and MBM’s staff for their assistance in identifying the specimens and Eduardo D. Lozano, Vanessa Ariati, Marcelo Silva and Isidoro Canestraro Filho for their help during field work. Special thanks go to Luciano Mysczak for map designing and geographical description and to Michael Stuart Rose for manuscript revision. We would like to thank PUCPR for providing a Specialization Scholarship Grant (Marcelino Champagnat Award) to the first author for her graduate studies at PUCPR.

INTRODUCTION
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
ACKNOWLEDGMENTS

BIBLIOGRAPHY

Anderson, M. J. 2005. PERMANOVA: a FORTRAN computer program for permutational multivariate analysis of variance. Department of Statistics, University of Auckland, New Zealand.

Andrade, P. M. de. 1992. Estrutura do estrato herbáceo de trechos da Reserva Biológica Mata do Jambreiro, Nova Lima, MG. Dissertação de Mestrado, Universidade Estadual de Campinas.

Beatty, S. W. 1984. Influence of microtopography and canopy species on spatial patterns of forest understory plants. Ecology 65: 1406-1419.

BFG - The Brazil Flora Group. 2015. Growing knowledge: an overview of Seed Plant diversity in Brazil. Rodriguésia 66(4): 1085-1113.

Blum, C. T. 2006. A Floresta Ombrófila Densa na Serra da Prata, Parque Nacional Saint-Hilaire/Lange, PR - caracterização florística, fitossociológica e ambiental de um gradientealtitudinal. Dissertação de Mestrado em Engenharia Florestal, Universidade Federal do Paraná.

Blum, C. T. 2010. Os componentes epifítico vascular e herbáceo terrícola da Floresta Ombrófila Densa ao longo de um gradiente altitudinal na Serra da Prata, Paraná. Tese de Doutorado em Engenharia Florestal, Universidade Federal do Paraná.

Cardoso, E. & I. Schiavini. 2002. Relação entre distribuição de espécies arbóreas e topografia em um gradiente florestal na Estação Ecológica do Panga (Uberlândia, MG). Revista Brasileira de Botânica 25: 277-289.

Carvalho, W. A. C.; A. T. Oliveira-Filho; M. A. L. Fontes & N. Curi. 2007. Variação espacial da estrutura da comunidade arbórea de um fragmento de floresta semidecídua em Piedade do Rio Grande, MG, Brasil. Revista Brasileira de Botânica 30(2): 315-335.

Cestaro, L. A.; J. L. Waechter & L. R. M. Baptista. 1986. Fitossociologia do estrato herbáceo da mata de araucária da Estação Ecológica de Aracuri, Esmeralda, RS. Hoehnea13: 59-72.

Chao, A. 1984. Non-parametric estimation of the number of classes in a population. Scandinavian Journal of Statistics 11: 265-270.

Chao, A.; K. H. Ma, & T. C. Hsieh. 2016. iNEXT (iNterpolation and EXTrapolation) Online: Software for Interpolation and Extrapolation of Species Diversity. Program and User’s Guide published at http://chao.stat.nthu.edu.tw/wordpress/software_download/

Daws, M. I.; C. E. Mullins; D. F. Burslem; S. R. Paton & J. W. Dalling 2002. Topographic position affects the water regime in a semideciduous tropical forest in Panama. Plant and soil, 238(1), 79-89.

Enoki, T. & A. Abe. 2004. Saplings distribution in relation to topography and canopy openness in an evergreen broad-leaved forest. Plant Ecology 173: 283-291.

Gentry, A. H. & C. H. Dodson. 1987. Contribution of nontrees to species richness of a tropical rain forest. Biotropica 19(2):149-156.

Gilliam, F. S.; N. L. Turrill; S. D. Aulick; D. K. Evans & M. B. Adams.1994. Herbaceous layer and soil response to experimental acidification in a central Appalachian hardwood forest. Journal of Environmental Quality 23(4): 835-844.

Gilliam, F. S. & M. R. Roberts. 2003. The dynamic nature of the herbaceous layer: synthesis and future directions for research. In: F.S. Gilliam & M.R. Roberts. The herbaceous layer in forests of Eastern North America.Oxford University Press. New York.

Gomes-Westphalen, J. S.; A. C. Borges Lins-e-Silva & F. Soares de Araújo. 2012. Who is who in the understory: the contribution of resident and transitory groups of species to plant richness in forest assemblages. Revista de Biología Tropical 60(3), 1025-1040.

Grytnes, J. A. & O. R. Vetaas. 2002. Species richness and altitude: a comparison between null models and interpolated plant species richness along the Himalayan altitudinal gradient, Nepal. The American naturalist 159(3): 294-304.

Hill, M. 0. & H. G. Gauch. l980. Detrended correspondence analysis: an improved ordination technique. Vegetatio 42: 47-58.

Hammer, Ø. Harper, D.A.T., Ryan, P.D. 2001. PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4(1): 9pp

Inácio, C. D. & J. A. Jarenkow. 2008. Relações entre a estrutura da sinúsia herbácea terrícola e a cobertura do dossel em Floresta Estacional no Sul do Brasil. Revista Brasileira de Botânica 31 (1):41-51.

Kozera, C. 2001. Composição florística e estrutura fitossociológica do estrato herbáceo subarbustivo em duas áreas de Floresta Ombrófila Densa, Paraná, Brasil. Dissertação de Mestrado em Biologia Vegetal, Universidade Estadual de Campinas.

Kozera, C. & R. R. Rodrigues. 2005. Floresta Ombrófila Densa Submontana: florística e estrutura do estrato inferior. In: M. C. M. Marques & R. M. Britez (Org.). História natural e conservação da Ilha do Mel. Curitiba. Editora da Universidade Federal do Paraná.

Kozera, C.; V. A. de O. Dittrich & S. M. Silva. 2006. Composição florística da Floresta Ombrófila Mista Montana do Parque Municipal do Barigüi, Curitiba, PR. Floresta 36(1): 45-58.

Kozera, C.; R. R. Rodrigues & V. A. de O. Dittrich. 2009. Composição florística do sub-bosque de uma Floresta Ombrófila Densa Montana, Morretes, PR, Brasil. Floresta39 (2): 323-334.

Leite, P. F. 2002. Contribuição ao conhecimento fitoecológico do sul do Brasil. Ciência & Ambiente 24: 51-73.

Lima, R. A. F.; V. C. de Souza; V. A. de O. Dittrich & A. Salino. 2012. Composição, diversidade e distribuição geográfica de plantas vasculares de uma Floresta Ombrófila Densa Atlântica do Sudeste do Brasil. Biota Neotropica 12(1): 241-249.

Maack, R. 1968. Geografia física do Estado do Paraná. Curitiba, Banco do Desenvolvimento do Paraná.

Magurran, A. E. 2011. Medindo a diversidade biológica. Tradução Dana Moiana Vianna. Curitiba. UFPR.

McCune, B. & M. Mefford. 2011. PC-ORD - Multivariate Analysis of Ecological Data. Version 6.0. Gleneden Beach, MjM Software.

Moro, M. F. & F. R. Martins. 2011. Métodos de Amostragem do componente arbóreo-arbustivo. In: Felfili, J. M.; P. V. Eisenlohr; M. M. R. F. Melo; L. A. Andrade & J. A. A. Meira Neto (Eds). Fitossociologia no Brasil: métodos e estudos de caso. Viçosa. Editora UFV.

Mueller-Dombois, D. & H. Ellenberg. 1974. Aims and methods of vegetation ecology. New York. John Wiley & Sons.

Müller, S. C. & J. L. Waechter. 2001. Estrutura sinusialdos componentes herbáceo e arbustivo de uma floresta costeira subtropical. Revista Brasileira de Botânica 24(4): 395-406.

Munhoz, C. B. R. & G. M. Araújo. 2011. Métodos de amostragem do estrato herbáceo-subarbustivo. In: Felfili, J. M.; P. V. Eisenlohr; M. M. R. F. Melo; L. A. Andrade & J. A. A. Meira Neto (Eds). Fitossociologia no Brasil: métodos e estudos de caso. Viçosa. Editora UFV.

Paciencia, M. L. B. 2008. Diversidade de pteridófitas em gradientes de altitude na Mata Atlântica do estado do Paraná, Brasil. Tese de Doutorado em Botânica, Universidade de São Paulo.

Patil, G. P. 2014. Diversity Profiles. Wiley StatsRef: Statistics Reference Online. John Wiley & Sons, Ltd.

Pereira, J. A. A.; A. T. Oliveira-Filho; J. P. Lemos-Filho. 2007. Environmental heterogeneity and disturbance by humans control much of the tree species diversity of Atlantic montane forest fragments in SE Brazil. Biodiversity and Conservation16: 1761-1784.

Polisel, R. T.; N. M. Ivanauskas; M. C. de Assis; G. J. Shepherd & K. Yamamoto. 2014. Structure of the understory community in four stretches of Araucaria forest in the state of São Paulo, Brazil. Acta Botanica Brasilica 28 (1): 86-101.

Portes, M. C.; G. O. F. Galvão & A. Koehler. 2001. Caracterização florística e estrutural de uma Floresta Ombrófila Densa Altomontana do morro Anhangava, Quatro Barras, PR. Floresta 31(1/2): 22‑31.

Poulsen, A. D. & H. Balslev. 1991. Abundance and cover of ground herbs in an Amazonia rain forest. Journal of Vegetation Science 2(3): 315-322.

Poulsen, A. D. 1996. Species richness and density of ground herbs within a plot of lowland rainforest in north-west Borneo. Journal of Tropical Ecology 12(2): 177-190.

Prado, J.; L. S. Sylvestre; P. H. Labiak; P. G. Windisch; A. Salino; I. C. L. Barros; R. Y. Hirai; T. E. Almeida; A. C. P. Santiago; M. A. Kieling-Rubio; A. F. N. Pereira; B. Øllgaard; C. G. Ramos; J. T. Mickel; V. A. O. Dittrich; C. M. Mynssen; P. B. Schwartsburd; J. P. S. Condack; J. B. S. Pereira & F. B. Matos. 2015. Diversity of ferns and lycophytes in Brazil. Rodriguésia 66: 1073-1083.

Reginato, M. & R. Goldenberg. 2007. Análise florística, estrutural e fitogeográfica da vegetação em região de transição entre as Florestas Ombrófilas Mista e Densa Montana, Piraquara, Paraná, Brasil. Hoehnea 34(3): 349-360.

Ribeiro, T. M.; N. M. Ivanauskas; S. V. Martins; R. T. Polisel; R. L. R. dos Santos & N. A. Miranda. 2013. Mixed rain forest in southeastern Brazil: tree species regeneration and floristic relationships in a remaining stretch of forest near the city of Itaberá, Brazil. Acta Botanica Brasilica 27(1), 71-86.

Rigon, J.; J. Cordeiro & D. A. Moraes. 2011. Composição e estrutura da sinusia herbácea em um remanescente de Floresta Ombrofila Mista em Guarapuava, PR, Brasil. Pesquisas Botânicas 62: 333-346.

Roderjan, C. V. 1994. O gradiente da Floresta Ombrófila Densa Altomontana no morro Anhangava, Quatro-Barras, PR. Aspectos climáticos, pedológicos e fitossociológicos. Tese de Doutorado em Engenharia Florestal, Universidade Federal do Paraná.

Roderjan, C, V.; F. Galvão; Y. S. Kuniyoshi & G. G. Hatschbach. 2002. As unidades fitogeográficas do estado do Paraná. Ciência & Ambiente 24: 75-92.

Scheer, M. B. & C. T. Blum. 2011. Arboreal diversity of the Atlantic Forest of Southern Brazil: from the beach ridges to the Paraná river. In: O. Grillo & G. Venora (eds.). The dynamical processes of biodiversity - Case studies of evolution and spatial distribution. InTech.

Silva, S. M. & R. M. de Britez. 2005. A vegetação da Planície Costeira. In: M. C. M. Marques & R. M. Britez (Org.). História natural e conservação da Ilha do Mel. Curitiba. Editora da Universidade Federal do Paraná.

Silva, A. G.; I. M. M. Sá-e-Silva; M. J. N. Rodal & A. C. B. Lins-e-Silva. 2008. Influence of edge and topography on canopy and sub-canopy structure of an Atlantic Forest fragment in Igarassu, Pernambuco State, Brazil. Bioremediation, Biodiversity and Bioavailability 2.1: 41-46.

Soil Survey Staff. 1999. Soil taxonomy: A basic system of soil classification for making and interpreting soil surveys. 2nd edition. Natural Resources Conservation Service. U.S. Department of Agriculture.

Whitmore, T. C. 1990. An introduction to the tropical rain forests. Clarendon Press, Oxford.

Záchia, R. A. & J. L. Waechter. 2011. Diferenciação espacial de comunidades herbáceo-arbustivas em florestas costeiras do Parque Nacional da Lagoa do Peixe, Rio Grande do Sul. Pesquisas, Botânica 62: 211-238.