Abstract

This study aimed at understanding the dynamics of ecological processes and the use of secondary forests in Santa Catarina state (Brazil). The data base for these studies was formed through forest inventories carried out in the three forest types of the state. The results of this study demonstrate that the patterns of diversity are very similar among the three forest types; however, the species compositions among the types are quite different. A total of 343 woody species belonging to 73 families were found in the 24,000 m2 sampling area, revealing the potential role of secondary forest in the conservation of biodiversity at the landscape scale. As expected, a small set of pioneer species dominates young secondary forests with shade-tolerant species becoming structurally important after 30 years. The patterns of forest structure and species diversity observed in study largely conform to the postagricultural secondary succession observed for many tropical forests.

1. Introduction

The continuing loss of primary tropical forests makes secondary forests increasingly important for maintaining biodiversity across large forested landscapes [14], while also providing for environmental services and sustainable economic development [57]. Today, less than 25% of the original forest area in Santa Catarina state (Brazil) remains, mostly as small fragments of secondary forest in a mosaic intermixed with other land uses [8]. Most secondary forests in Santa Catarina and throughout much of the Mata Atlântica are privately held and result from fallow rotations in a dynamic agricultural system [9]. The recovery potential of such secondary forests is influenced by many interacting factors including the length and intensity of past land use, parcel size, soil conditions and landscape position, local climate regimes, proximity to forest seed sources, the presence or absence of seed banks, and dispersal agents, and interactions among regenerating species [1016].

The important roles that secondary forests can play in conservation and economic development are poorly understood in Brazil. Even well-intentioned Brazilian laws governing forest protection and use can be counter-productive to conservation. Recent laws that define a “forest” in terms of arbitrary structural characteristics have led small farmers to shorten fallow periods so that secondary forest succession does not proceed to the point where it is recognized as “forest”. This avoids federal/state restrictions on forest harvesting or land conversion [17] with the result that secondary forests which could contribute to biodiversity conservation while providing a range of nontimber forest products are now less likely to develop. The landscape increasingly is held only in the earliest stages of forest succession before being cropped again.

We focused on the successional dynamics of secondary forests in the youngest age classes that prevail across this landscape in order to suggest alternative definitions of “forest” that recognize their important values in conservation and economic development. In this paper we describe the regeneration patterns that occur in secondary forests of Santa Catarina on parcels having diverse cultivation histories and variable fallow periods. Anthropogenic disturbances such as forest clearing for agriculture are generally more intensive/severe than those arising from natural disturbances, but early stages of forest succession appear to be dominated in both instances by a small number of common woody species in well-defined stages typical of large gaps [16, 18, 19]. The specific objectives of this study were to examine the patterns of species diversity and abundance as well as the transitions during which woody species are added and removed during the early phases of secondary forest succession.

2. Methods

2.1. Study Areas

The study was carried out on 48 small farms in the counties of Anchieta, Garuva, Concórdia, Três Barras, Caçador and São Pedro de Alcântara, Santa Catarina, Brazil (Figure 1). These areas included three different vegetation formations common to the Mata Atlântica of Santa Catarina, seasonal deciduous forest, mixed ombrophylous forest and dense ombrophylous forest [20] recognized and protected by Brazilian Law no. 11.426/2006.

2.1.1. Seasonal Deciduous Forest (SDF)

This forest type occurs in the Uruguay River basin at elevations ranging from 200 to 600 meters with an annual precipitation of 1.800 mm/yr [2124]. It is characterized by a closed canopy dominated by Lauraceae and emergent deciduous species, mostly Fabaceae and a subcanopy dominated by Sorocea bonplandii, Gynnanthes concolor and Trichilia species; epiphytes are poorly represented in comparison with the dense ombrophylous forest [21].

2.1.2. Mixed Ombrophylous Forest (MOF)

This forest type is distributed across plateaus at approximately 500 meters elevation upwards on slopes rising to 1600 meters above sea-level with annual precipitation ranging from 1,600 to 2,100 mm [20]. This moist forest is characterized by a supercanopy dominated by Araucaria angustifolia, with subcanopy layers rich in species of Lauraceae, Myrtaceae, and Fabaceae [25, 26].

2.1.3. Dense Ombrophylous Forest (DOF)

This forest was formerly common in coastal regions at elevations below 500 meters. The forest is characterized by hot temperatures, heavy rainfall (annual precipitation ranges from 1,500 to 2,000 mm) and an evergreen canopy dominated by Lauraceae and Myrtaceae with an abundance of epiphytes and palms [18, 26].

2.2. Data Collection and Analysis

We inventoried a total of eighty 1 0 m × 1 0 m plots in the dense ombrophylous forest (Garuva and São Pedro de Alcântara sites), and forty 2 0 m × 1 0 m plots in each of the seasonal deciduous forest (Anchieta and Concórdia sites), and mixed ombrophylous forest (Três Barras and Caçador sites) types (a total of 160 plots covering 24,000 m2). Based on physiognomic features, together with land ownership information, we sited 160 plots distributed across four successional stages (after Klein [18]: shrubby stage (0–8 years), small tree stage (8–15 years), Arboreal Stage (15–30 years) and advanced arboreal stage (30–60 years)). All woody plants taller than 1.5 m were identified and measured for DBH (diameter at breast height = 1.3 m) and total height. Plant identifications were made with reference to the Angiosperm Phylogeny Group (APG) classification updated in APG II [27], and the expertise of Prof. Ademir Reis (Federal University of Santa Catarina and Curator of the Barbosa Rodriguez Herbarium) and Prof. Marcos Sobral (BHCB—Herbarium of the Botany Department at the Federal University of Minas Gerais).

Species richness (number of species), Shannon-Wiener’s index of diversity ( 𝐻 ), evenness ( 𝐽 ), Simpson’s diversity index ( 𝐷 ), “importance value” (IV) and absolute frequency were estimated using the FITOPAC 1 software package [28]. Rarefaction curves (Sobs) constructed using 100 randomized orders and the Chao1, Jackknife and Bootstrap estimators were derived using EstimateS 6.0 software [29]. Differences in values of stem density and basal area among stages were tested by one-way ANOVA with Tukey’s multiple comparison tests using the STATISTICA 6.0 software package [30].

3. Results

A total of 13,548 woody plants were identified on the 160 plots, representing 343 species from 73 families (the appendix). The most common families represented (number of species in parentheses) were the Myrtaceae (38), Fabaceae (33), Asteraceae (27), Lauraceae (27), Rubiaceae (20) and Melastomataceae (17). Forty-four species were common to all three forest types, representing around 13% of the species total, leading to a similarity estimate of about 10%.

The dense ombrophylous forest (DOF) type contained 66% (230) of all the species recorded for all forest types. Our sampling design placed more but smaller plots in the DOF formation and this likely introduced a slight bias towards finding more species in this forest type. When the differences in plot number and area sampled are accounted for (DOF2 in Table 2), we estimate that the dense ombrophylous forest type would still include 59% (204) of the species identified. The seasonal deciduous forest and mixed ombrophylous forest types were very similar in their estimated species richness but markedly less rich than the dense ombrophylous forest type.

The secondary succession process in all three forest types was characterized by an increase in (1) species richness, (2) evenness scores, and (3) Shannon-Wiener diversity index values. Only the Simpson’s diversity index values decreased over time during the formation of taxonomically more diverse and structurally more complex communities (Table 2).

Despite the relatively high species richness that our sampling revealed in these secondary forests, the rarefaction curves (Sobs) and other estimators of richness (Chao1, Jack2 and Bootstrap; Figure 2) suggest that our sampling only found about three-fourths (75% in SDF, 71% in MOF and 78% in DOF) of the maximum number of species expected to be present in these forest types.

Species composition for all three forest types changed with changes in the length of the preceding fallow period (Table 3). Although all three forests are marked by high species diversity, a few species dominate each stage of succession as demonstrated by the large Importance Values (Table 3), but this tendency decreases with increasing succession. As an example, the shrubby stage of all three forest types is dominated by Baccharis dracunculifolia and a few other species, while the small tree stage in the MOF and DOF types is dominated by Myrsine coriaceae and a few other species Nectandra lanceolata (SDF), Mimosa scabrela (MOF) and Miconia cinnamomifolia (DOF) dominated the Arboreal Stage and species such as Nectandra megapotamica (SDF), Ocotea puberula (MOF) and Hyeronima alchorneoides (DOF) dominated the advanced arboreal stage of each type.

During the early stages of secondary succession, species such as Trema micrantha (SDF) and Tibouchina trichona (DOF) had low absolute frequencies but they had high densities and/or dominance when they occurred. In the advanced arboreal stages typical understory species such as Cupania vernalis, Allophylus edulis (SDF and MOF) and Euterpe edulis (DOF) all had high frequencies of occurrence. Most species with large Importance Values were common or abundant in only one or two successional stages but were quickly replaced by other species during succession. However, a few persisted, for example, Tibouchina pulchra was common in three successional stages in the DOF.

4. Discussion

The sampling method used identified those species most common to the early successional stages of secondary forest regeneration in three forest types of the Mata Atlântica. The high number of species (230) found in the dense ombrophylous forest was not unexpected, as this forest type represents some 82% of all arboreal species found in Santa Catarina [26].

Diversity and especially species richness increased over time since abandonment, similar to what has been found in other studies of swidden fields, abandoned plantations and pasture [13, 3134].

The peak in the numbers of species found in Arboreal Stages of all forest types is consistent with the “intermediate disturbance hypothesis” [35], that during the early stages of secondary forest succession a few pioneer species dominate but are replaced over time by species better adapted to the increasingly competitive environments that develop. Species diversity is often greatest during mid-successional stages that contain both early and late successional species. In all three forest types studied here, the Arboreal Stage (15–30 years) had a richness that was greater than for any other stage in the SDF and MOF types and equal in richness to the advanced arboreal stage of the DOF type.

Species richness is only one component of diversity but because it gives the same weight to all species irrespective of their relative abundance it is strongly influenced by the number of rare species. Evenness, another component of diversity, is strongly influenced by the relative frequencies of dominant species [36, 37]. We use different measures of diversity to emphasize different diversity patterns that reflect different ecological processes.

Our Shannon-Wiener diversity estimates were similar to, or slightly higher than, those of other studies of comparably aged Mata Atlântica forests. For example, Oliveira [34] and Torezan [38] estimated 𝐻 = 2 . 5 1  nats/ind in five-year-old early successional stands; Pessoa et al. [39] estimated 𝐻 = 3 . 6 6  nats/ind in a thirty-year-old pole stand; and Oliveira [34] estimated 𝐻 = 3 . 3 3  nats/ind in a twenty-five-year-old stand and 𝐻 = 3 . 1 0  nats/ind in a fifty-year-old area stand. All these studies were conducted on sites with land use histories characterized by shifting cultivation.

The Shannon index is a common diversity estimate reported in the literature ranging from 3.26 to 4.36 for the mature forests of these types [4046]. The higher values reported here for all stages of our study in relation to values reported in the literature may be due to the sampling method and criteria used for plant inclusion. Also our use of more plots, especially noncontiguous plots, likely raised the levels of beta diversity.

Values of the Shannon index can be influenced by the number of species with intermediate values of relative abundance [47], and may introduce some variation in the estimates of species richness of communities [48]. As the value of Simpson’s 𝐷 has an inverse relationship with the indices of Shannon and evenness [49], the value of 𝐷 decreases along successional stages.

The floristic composition of secondary forest formations described in this study was very similar with respect to genera reported for other studies in the Mata Atlântica of southeastern Brazil (Table 4).

5. Conclusion

The floristic diversity observed for the forests we surveyed largely agree with patterns of post-agricultural secondary forest succession observed for many other neotropical forests. The chronosequence covers a relatively large range of successional ages (0–60 years) that results in large changes in species diversity and composition despite a limited geographical sampling. This result suggests that the patchy mosaic of secondary forests in Santa Catarina has a high potential for biodiversity conservation. Whether it can also provide for economic development in the way of nontimber forest products would depend upon the life histories of individual species and any rules instituted to guide their sustainable management.

The results of this study demonstrate that the patterns of secondary succession appear very similar among the three forest types with respect to changes in species richness and other measures. However, the species composition among the types, while quite similar in the earliest shrubby stage, diverges during succession, with the largest differences noted between the DOF and the other two types. As expected, and despite high species richness, a small set of “pioneer species” dominates the shrubby and small tree stages until about age 15. After that time, more shade-tolerant species increase during the Arboreal and advanced arboreal stages of forest succession, but only a small number become structurally important after 25–30 years.

A chronosequence approach only allows us to infer successional changes because we do not analyze the underlying processes mediating these changes on our sites (growth, mortality, and recruitment). Thus, we suggest that long-term permanent plots be established in this region, to improve our understanding of secondary forest dynamics, while also creating a framework for future comparative studies of the role of ecological processes and mechanisms in different successional stages.

Appendix

See Table 5.

Acknowledgments

The authors gratefully acknowledge the expert help of Dr. Marcos Sobral and Prof. Dr. Ademir Reis in the taxonomic identification of forest species. They thank the Núcleo de Pesquisas em Florestas Tropicais for the support. Financial support came from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brazil) and from the Fundação de Apoio À Pesquisa Científica e Tecnológica do Estado de Santa Catarina (FAPESC).