Open Access

Understorey vegetation gradient in a Eucalyptus grandis plantation between a savanna and a semideciduous forest

  • Pavel Dodonov1Email author,
  • Danilo Muniz da Silva2 and
  • Natália Bianca Rosatti3
New Zealand Journal of Forestry Science201444:10

https://doi.org/10.1186/s40490-014-0010-y

Received: 22 March 2013

Accepted: 24 February 2014

Published: 18 June 2014

Abstract

Background

Plant community assemblage is influenced by many factors, including soil characteristics and the arrival of diaspores from surrounding areas. These factors may be especially important in transition areas, leading to spatial gradients in the plant community.

Methods

This study was performed in the understorey of an abandoned Eucalyptus grandis W. Hill ex Maiden plantation between a savanna and a forest, 490 m apart, in south-eastern Brazil. This study assessed whether the spatial variation in several variables related to the understorey’s structure and composition is best described by linear or non-linear (quadratic) models. The linear model would indicate a gradient between the two vegetation types, whereas the quadratic model would indicate a stronger effect of the plantation’s edges.

Results

There was a gradient in species composition between the two edges of the plantation. Mean vegetation height was greatest at the savanna edge and lowest in the centre of the plantation. The total number of individuals per plot and the phylogenetic diversity decreased with distance from the savanna edge. Different patterns were observed for different dispersal syndromes, with animal dispersal being more common at the savanna edge, wind dispersal in the centre of the plantation and self dispersal at the forest edge.

Conclusions

The greater number of individuals at the savanna edge may indicate that dispersal and arrival of diaspores are the most important factors influencing community structure and composition of the understorey of this abandoned E. grandis plantation, with most propagules coming from the savanna area. The smaller vegetation height in the centre of the fragment may also indicate older colonisation at the edges. Therefore, in addition to highlighting the recovery potential of undergrowth beneath abandoned Eucalyptus spp. plantations, these results show that this recovery is spatially heterogeneous and that dispersal plays a large role in it. This should be taken into account in restoration projects. The authors recommend careful consideration before removing regenerating Eucalyptus spp. trees as part of the site restoration. Instead the focus should be on the recovery potential of the undergrowth, with gradual removal of Eucalyptus trees, if necessary.

Keywords

CerradoCommunity compositionDispersalDiversityEcotoneRestorationStructure

Background

Reconstructing a natural community is one of the greatest challenges for restoration ecology, partly because community assemblage is determined by several distinct factors such as environmental filters, diaspore availability, competition between and within species, and species potential for colonisation (Webb et al. [2002]; Chase [2003]). Natural recolonisation may aid in community reconstruction by allowing communities to regenerate with species from neighbouring vegetation patches. However, a successful application of this process requires a better understanding of how species colonise or recolonise disturbed areas (Török et al. [2011]). Colonisation is limited mostly by three factors: 1) diaspore availability in the neighbouring species pool and the dispersal capacity of different species (Török et al. [2011]); 2) ecological traits that determine the survival and establishment of different species under different environmental conditions (Lebrija-Trejos et al. [2010]); and 3) biotic interactions, such as facilitation and competition, that determine the patterns of species co-occurrence (Cianciaruso et al. [2009]). This is also true for natural recolonisation of disturbed environments, in which resprouting from pre-existent underground structures may also play an important role (Durigan et al. [1998]). Therefore, knowledge of how these factors affect transition area is essential for successful restoration planning.

Transition areas between different vegetation types, for example between savanna and forest, are usually characterised by a gradual shift in environmental conditions and species composition (Furley et al. [1992]). In some areas, however, abrupt changes in soil and other environmental characteristics may result in a sharp edge between the different vegetation types (Ruggiero et al. [2002], Durigan and Ratter [2006]). In either case, both savanna and forest patches influence the transition area, and the influence is stronger closer to each patch (Cadenasso et al. [2003]). This is likely to result in a spatial gradient in species composition during the recolonisation of a transition area, as the occurrence and density of different species will vary with the distance from the source of diaspores (Wolters et al. [2005]). The environmental conditions found in the transition area also affect the resulting species composition, as species originating from different environments are adapted to different environmental constraints. For example, savanna species are adapted to decreased soil fertility and water availability (Ruggiero et al. [2002]) and to regular fires (Moreira [2000]; Miranda et al. [2002]), factors that may lead to the local extinction of forest species.

Ecotones between forest and savanna are common in the São Paulo state, south-eastern Brazil, where cerrado sensu stricto, a savanna formation that is part of the larger floristic domain known as the Cerrado, occurs alongside semideciduous forests (Ruggiero et al. [2002]; Durigan and Ratter [2006]). Transition areas in this region are often replaced by agricultural and silvicultural land uses. Silviculture, a large part of which uses exotic Eucalyptus spp. trees, is the fifth most frequent land use adjacent to cerrado fragments in the São Paulo state (Durigan et al. [2007]). One characteristic of these Eucalyptus spp. plantations is that they often contain an understorey composed of native species (Saporetti Jr et al. [2003]; Neri et al. [2005]). This understorey may be used as a resource and refuge by animal species and also function as stepping stones or corridors linking natural areas that would not be connected otherwise (Machado and Lamas [1996]; Lyra-Jorge et al. [2008]). In addition, the understorey can increase local biodiversity by offering regeneration sites for plant species from different vegetation types (Saporetti-Jr et al. [2003]).

This understorey may also be used as a starting point to restore native vegetation when the plantation is no longer used, which makes assessments of its composition important for restoration ecology. Other studies (e.g. Durigan et al. [1998], Jepson [2005], Neri et al. [2005]) have already shown the potential of natural regrowth in abandoned Eucalyptus spp. plantations; however, these studies have not discussed large-scale spatial variation in the regrowth’s structure and composition. The objective of this study was to test for linearity or non-linearity of the gradient in vegetation structure, composition and diversity of the understorey of an abandoned Eucalyptus grandis W. Hill ex Maiden plantation located between a cerrado stricto sensu and a semideciduous forest to provide insights for restoration. The following questions were addressed in this study: (1) Are there gradual changes in species composition, richness, diversity and phylogenetic diversity between the two areas? These gradual changes are expected because both proximity to natural areas and edges affects vegetation structure and species occurrence and distribution (Harper et al. [2005], Wolters et al. [2005]). (2) Are these patterns linear or non-linear pattern, with a trough or peak in the middle? As colonisation starts from the edges, greater density of individuals and mean vegetation height should be found closer to the edges. (3) Are there more pronounced gradients on wind- and self-dispersed species than on animal-dispersed species? If this area is used homogeneously by the local frugivorous fauna, animals could disperse seed evenly through the area. (4) Do abundant species show a gradient in their distribution between the two areas? Differences in ecological processes and species composition between the savanna and forest areas are likely to affect the distribution of the dominant species in the study area. This article reports the actual compositional and structural variability of the woody species in the study area, discusses its probable causes and whether the results matched expected composition. Conclusions have been drawn and recommendations made for the restoration of natural vegetation in similar areas.

Methods

Study area and sampling methods

This study was conducted in the understorey of a site 490 m wide and 490 m long in the north-eastern portion of the Federal University of São Carlos (21°57’50” S, 47°51’55” W, 815–890 m a.s.l.; Santos et al. [1999]), in São Paulo state, Brazil. This site had been used previously as a Eucalyptus grandis plantation and was last harvested approximately 15 years prior to the study. It had not been managed since then and E. grandis trees had regrown from stumps left after the last harvest. These trees were approximately 15 m tall at the time of the study, with a highly homogeneous tree cover throughout the study site. This site is located between two natural areas: a semideciduous forest to the north and a regenerating cerrado sensu stricto savanna to the south (Figure 1). The predominant vegetation of the cerrado was fully grown native trees but there were also a few, widely spaced Eucalyptus spp. trees that had regenerated for more than 20 years from an previous Eucalyptus spp. plantation. The regional climate is seasonal, with dry winters and wet summers. Soils are acidic and nutrient-poor Oxisols with high aluminium content (Santos et al. [1999]).
Figure 1

Study areas. A: Satellite image of the study site, located in the north-eastern portion of the Federal University of São Carlos (21°57’50” S, 47°51’55 W) between a semideciduous forest to the north and a savanna to the south (white lines represent the transects); B: the forest edge; C: the interior of the site; D: the cerrado edge.

Three transects were placed between the cerrado and forest edges, with 150 m between adjacent transects (Figure 1). Seven 25 x 25 m plots were delimited in each transect, leaving 50 m between consecutive plots. The first and last plot of each transect were placed approximately 10 m from the cerrado and forest edges, respectively. The three transects were labelled as “a”, “b” and “c”, and the plots in each transect were numbered from 1, corresponding to the plot closest to the cerrado edge, to 7, corresponding to the plot closest to the forest edge. Therefore, plot a1 is the plot of transect “a” that is closest to the cerrado and plot c7 is the plot of transect “c” that is closest to the forest. The three transects were treated as replicates and used the distance between the centre of each plot and the cerrado edge (i.e. 22.5 m for plot 1 and 472.5 m for plot 7) as the explanatory variable. Within each plot, all woody non-vine plants with height greater than 50 cm were sampled, considering stems that were clearly separated at soil level as different individuals, and classified them into four height classes: < 1.4 m, 1.4 – 1.6 m, 1.61 – 3 m, and > 3 m. All individuals to were then identified species level by consulting specialised literature and a taxonomist (see Acknowledgments) and classified all species into dispersal syndromes (wind-, self- and animal-dispersed) based on literature reviews or on the information available for other species of the same genus. As the E. grandis trees had been planting without prior removal of the underground plant structures (pers. comm., Maria Inês S. Lima), native species in this area may have originated either by resprouting or from recruitment. However, it was not possible to differentiate between these two types in the field.

Data analysis

Ordination analysis of vegetation composition

General differences in species composition among the sampling plots were assessed by means of a non-metric multidimensional scaling (NMDS) analysis, performed on untransformed data with the algorithm developed by Taguchi & Oono ([2005]), using Past 2.10 sofrware (Hammer et al. [2001]). Two similarity indices were used to assess the presence-absence and abundance patterns: the Jaccard index, which takes into account only the presence of a given species, and the Horn index, which considers the relative abundance of all the species (Jost et al. [2011]).

Vegetation variables

The data were described with 23 response variables related to species diversity, phylogenetic diversity, vegetation structure, dispersal syndrome, and distribution of abundant species. For species diversity, two measures of species richness were used, namely the number of species and the Fisher’s alpha diversity index, as well as Shannon’s diversity index and the corresponding Pielou’s evenness index (Magurran [1988]), calculated in Past 2.10. The Fisher’s alpha diversity index was used because species richness per se usually increases with increasing number of individuals, whereas the Fisher’s alpha diversity index is a robust species richness estimate that is not influenced by biases in the number of individuals (Maurer and McGill [2011]). To use Fisher’s alpha diversity index, the species abundance distribution should not be different from the log-series (Maurer and McGill [2011]), which was the case for our data (p = 1, as calculated using the Past 2.10 software).

Phylogenetic diversity was assessed by means of two indices, phylogenetic diversity (PD) and phylogenetic diversity-rate (PD-rate). The Phylomatic module of the Phylocom software was used to construct a phylogenetic tree of all the species sampled (Webb and Donoghue [2005]; Webb et al. [2008]). The lengths of the branches were estimated based on the current Phylomatic tree (tree R20091110; Webb et al. [2008]) with APG III (Angiosperm Phylogeny Group [2009]) and Wikström ages (Wikström et al. [2001]). The root node and all the dated nodes were fixed, and the undated nodes were placed evenly between the dated nodes or between the dated nodes and the terminal nodes by means of the “bladj” algorithm in the Phylocom software (Webb et al. [2008]). The calculation of PD was done by summing the lengths of the branches of the phylogenetic tree obtained for each plot (Faith [1992]). The units used for PD were millions of years (My). PD-rate was calculated by dividing PD by the number of species in each plot. Whereas PD is a measure of the phylogenetic distance between species in a community, PD-rate permits to verify whether the variation in PD is due to the number of species present or to the phylogenetic differences between the species. A low PD-rate indicates that the PD value at a site is due mostly to the number of species present, whereas a high PD-rate indicates that the PD is a result of the phylogenetic distances between the species.

The total number of individuals, the average vegetation height, the standard deviation of height and the coefficient of variation (calculated by dividing the standard deviation by the mean height) were used to characterise the vegetation structure. To calculate average vegetation height, we considered that the height of each individual corresponds to the mid-value of its height class (4.5 m for the last class, as most tall trees were still under 6 m in height) and calculated the mean of these mid-values. We used the same mid-values to estimate standard deviation, which was used as a measure of structural variability within the plot. The number and proportion of wind-, self-, and animal-dispersed individuals and species within each plot were used to analyse the dispersal syndromes. We also analysed the distribution of the seven species with at least 90 individuals because patterns of abundant species may indicate important ecological processes.

Regressions

Both quadratic and linear regressions were examined to assess the existence and shape of the gradient with respect to the 23 response variables, in the software R 2.13 (R Core Development Team, [2012]). Distance from the cerrado edge was specified as the explanatory variable. We chose quadratic regressions because they fit with the expected pattern of more extreme values at both edges, with either a trough or a peak at intermediate distances. In contrast, a linear regression would show a linear increase in the response variable along each transect. We considered a quadratic regression significant only when its quadratic term (x2) was also significant; otherwise we considered the linear regression to be a better model. We considered a significance level of 0.05 for all tests.

Results

Vegetation composition

Non-metrical multidimensional scaling ordination was conducted using both the Jaccard and Horn distance indices, Figure 2A & B, respectively. No clear separation among sites was generated with the Jaccard distance index but was more obvious when the Horn distance index was used. Specifically, the three plots of each transect that were closer to the cerrado edge were placed more closely together on the ordination plot with the Horn distance index. This resulted in a trend in species composition that may be represented by two zones, up to 185 m (plot 3) from the cerrado edge and further (Figure 2B). These results indicate that, even though there is a change in species abundances along the gradient, complete species replacement does not occur. Furthermore, in the ordination diagrams, plots located along the same transect appear to be more similar to one another than plots of different transects, although this separation is not as clear as for edge distances (Figure 2).
Figure 2

Gradient in the floristic composition in the understorey of the study site. Non-metrical multi-dimensional scaling using: A: Jaccard distance; and B: Horn distance. Letters a, b and c represent the three transects, and numbers 1 to 7 represent plots at the following distances from the cerrado edge: (1) 10–35 m; (2) 85–110 m; (3)160-185 m; (4) 235–260 m; (5) 310–335 m; (6) 385–410 m; and (7) 460–485 m. Grey areas represent areas of the ordination plot that are closer to plot 1, 2 and 3 than to the other plots, showing a pattern in the analysis performed with Horn distance. Stress: A = 0.3223, B = 0.2489.

Vegetation variables

There were 2235 individuals, which corresponded to 1702 individuals per hectare. These individuals belonged to 123 species from 81 genera and 39 families (Table 1).
Table 1

Floristic composition of the study area

Species

N

Rank

Distances

Dispersal

Anacardiaceae

    Astronium graveolens Jacq.

21

24

1-7

Wind23

    Tapirira guianensis Aubl.

6

36

1-3,5

Animal15

Annonaceae

    

    Annona coriacea Mart.

18

26

2-5,7

Animal15

    Annona crassiflora Mart.

2

40

1,4

Animal15

    Duguetia furfuracea (A.St.-Hil.) Saff.

19

25

1-5

Animal15

    Guatteria nigrescens Mart.

1

41

7

Animal17

    Xylopia aromatica (Lam) Mart.

13

30

1-5,7

Animal1

    Xylopia frutescens Aubl.

11

32

5,7

Animal34

Aquifoliaceae

    Ilex cerasifolia Reissek

3

39

5-7

Animal31

Araliaceae

    Schefflera vinosa (Cham. & Schltdl.) Frodin & Fiaschi

204

1

1-7

Animal14

Asteraceae

    

    Baccharis dracunculifolia DC.

33

18

1-6

Wind9

    Chromolaena cf. laevigata (Lam.) R.M. King & H. Rob.

2

40

6

Wind32

    Eupatorium cf. megaphyllum Baker

1

41

7

Wind

    Gochnatia pulchra Cabrera

16

28

1,3,4

Wind28

    Piptocarpha axillaris (Less.) Baker

4

38

3,6,7

Wind16

    Piptocarpha macropoda (DC.) Baker

11

32

1,5-7

Wind31

    Piptocarpha rotundifolia (Less.) Baker

24

21

1-7

Wind15

    Piptocarpha sp. Hook. & Arn.

6

36

3-5

Wind35

    Vernonanthura cf puberula (Less.) H. Rob.

1

41

2

Wind33

    Vernonanthura phosphorica (Vell.) H. Rob.

5

37

1-3

Wind14

    Vernonia rubriramea Mart. ex DC.

12

31

1-4,6,7

Wind28

Bignoniaceae

    Handroanthus chrysotrichus(Mart. ex DC.) Mattos.

2

40

1,3

Wind31

    Handroanthus ochraceus (Cham.) Mattos

4

38

2,3,5

Wind15

    Jacaranda caroba (Vell.) DC.

24

21

3-7

Wind28

    Tecoma stans (L.) Juss. ex Kunth

23

22

1-7

Wind18

    Zeyheria montana Mart.

3

39

3

Wind28

    Zeyheria tuberculosa (Vell.) Bureau ex Verl.

1

41

6

Wind31

Burseraceae

    Protium heptaphyllum (Aubl.) Marchand

2

40

5

Animal26

Caryocaraceae

    Caryocar brasiliense A.St.-Hil.

10

33

2-5

Animal15

Celastraceae

    Tontelea micrantha (Mart. ex Schult.) A.C. Sm.

1

41

3

Animal24

Connaraceae

    Connarus suberosus Planch.

9

34

2,3

Animal15

    Rourea induta Planch.

5

37

3,5

Animal15

Ebenacae

    Diospyros hispida A.DC.

26

20

1-7

Animal15

Erythroxylaceae

    Erythroxylum cf. deciduum A.St.-Hil.

5

37

2,4

Animal24

    Erythroxylum cuneifolium (Mart.) O.E. Schulz

9

34

2,3,5-7

Animal28

    Erythroxylum suberosum A.St.-Hil.

4

38

1,2,4

Animal15

Euphorbiaceae

    Alchornea triplinervia (Spreng.) Müll. Arg.

36

16

1.4-7

Animal11

    Sapium glandulosum (L.) Morong

9

34

1,3,4,7

Animal31

Fabaceae

    Acacia recurva Benth.

2

40

2

Self31

    Acosmium dasycarpum (Vogel) Yakovlev

1

41

2

Wind26

    Acosmium subelegans (Mohlenbr.) Yakovlev

10

33

2,5,6

Wind15

    Anadenanthera peregrina (L.) Speg.

1

41

2

Self29

    Andira inermis (Wright) DC.

3

39

5

Animal25

    Bauhinia rufa (Bong.) Steud.

51

13

1-7

Self15

    Bowdichia virgilioides Kunth

1

41

4

Wind15

    Chamaecrista flexuosa (L.) Greene

2

40

3

Self24

    Copaifera langsdorffii Desf.

74

8

1-7

Animal1

    Dalbergia miscolobium Benth.

5

37

2-4

Wind15

    Dimorphandra mollis Benth.

2

40

2

Animal15

    Machaerium acutifolium Vogel

7

35

1,4,7

Wind1

    Machaerium nyctitans (Vell.) Benth.

2

40

4,7

Wind4

    Machaerium stipitatum (DC.) Vogel

1

41

7

Wind31

    Pterogyne nitens Tul.

1

41

3

Wind27

    Senna rugosa (G. Don) H.S. Irwin & Barneby

2

40

4

Self28

    Senna splendida (Vogel) H.S. Irwin & Barneby

2

40

2,7

Self34

    Stryphnodendron adstringens (Mart.) Coville

2

40

4,5

Animal28

    Stryphnodendron obovatum Benth.

16

28

1-5,7

Self15

Lamiaceae

    Aegiphila lhotzkiana Cham.

22

23

1-7

Animal29

Lauraceae

    Endlicheria paniculata (Spreng.) J.F. Macbr.

1

41

1

Animal33

    Ocotea pulchella (Nees & Mart.) Mez

171

2

1-7

Animal8

Malpighiaceae

    Byrsonima intermedia A. Juss.

35

17

1-7

Animal15

Malvaceae

    Eriotheca gracilipes (K. Schum.) A. Robyns

17

27

2-4,6

Wind29

Melastomataceae

    Leandra lacunosa Cogn.

6

36

1,3,5

Animal14

    Miconia albicans (Sw.) Steud.

68

9

1-7

Animal15

    Miconia cf. ligustroides(DC.) Naudin

1

41

7

Animal15

    Miconia fallax DC.

17

27

1-4

Animal15

    Miconia ligustroides (DC.) Naudin

4

38

2,3

Animal15

    Miconia rubiginosa (Bonpl.) DC.

29

19

1-7

Animal12

    Miconia stenostachya DC.

4

38

1,2

Animal29

Meliaceae

    Cabralea canjerana (Vell.) Mart.

3

39

5,7

Animal30

    Cedrela fissilis Vell.

2

40

2,7

Wind9

Moraceae

    Brosimum gaudichaudii Trécul

9

34

1,5-7

Animal?1

    Ficus cf citrifolia Mill.

2

40

2,4

Animal31

Myristicaceae

    

    Virola sebifera Aubl.

14

29

2,4,5

Animal29

Myrtaceae

    Calyptranthes concinna DC.

5

37

5,6

Animal13

    Calyptranthes lucida Mart. ex DC.

2

40

5,6

Animal13

    Campomanesia adamantium (Cambess.) O. Berg

57

11

1-7

Animal15

    Campomanesia pubescens (Mart. ex DC.) O. Berg

2

40

3

Animal15

    Campomanesia sp. Ruiz & Pav.

2

40

5

Animal13

    Eugenia cf klotzschiana O. Berg

1

41

3

Animal13

    Eugenia dysenterica DC.

4

38

3,4

Animal1

    Eugenia obversa O. Berg

17

27

4-7

Animal13

    Eugenia punicifolia (Kunth) DC.

1

41

2

Animal15

    Eugenia pyriformis Cambess.

3

39

4,5

Animal13

    Myrcia bella Cambess.

57

11

1-7

Animal15

    Myrcia cf. tomentosa (Aubl.) DC.

6

36

1,2,5,6

Animal13

    Myrcia fallax DC.

37

15

1-7

Animal15

    Myrcia lingua (O. Berg) Mattos & D. Legrand

100

4

1-7

Animal13

    Myrcia tomentosa (Aubl.) DC.

2

40

1,3

Animal13

    Psidium grandifolium Mart. ex DC.

26

20

1-6

Animal24

    Psidium guajava L.

2

40

1,2

Animal13

    Psidium laruotteanum Cambess.

12

31

2-5,7

Animal24

Nyctagenaceae

    

    Guapira noxia (Netto) Lundell

14

29

1-6

Animal15

    Guapira opposita (Vell.) Reitz

4

38

1,4

Animal10

Ochnaceae

    Ouratea spectabilis (Mart. ex Engl.) Engl.

12

31

1-7

Animal15

Peraceae

    Pera glabrata (Schott) Poepp. ex Baill.

56

12

1-6

Animal15

Phyllantaceae

    Phyllanthus acuminatus Vahl

123

3

1-7

Self22

    Seguieria cf. americana L.

7

35

1,4-6

Wind6

Primulaceae

    Myrsine coriacea (Sw.) R.Br. ex Roem. & Schult.

90

7

1-7

Animal20

    Myrsine umbellata Mart.

93

6

1-7

Animal31

Proteaceae

    

    Roupala montana Aubl.

7

35

2,5,7

Self15

Rubiaceae

    Chomelia cf. obtusa Cham. & Schltdl.

1

41

1

Animal11

    Cordiera concolor (Cham.) Kuntze

1

41

4

Animal31

    Cordiera sessilis (Vell.) Kuntze

1

41

2

Animal28

    Palicourea rigida Kunth

3

39

2,5

Animal15

    Psychotria nuda (Cham. & Schltdl.) Wawra

5

37

1,2,6,7

Animal3

    Rudgea viburnoides (Cham.) Benth.

11

32

1,3,5-7

Animal29

    Tocoyena formosa (Cham. & Schltdl.) K. Schum.

2

40

1

Animal15

Rutaceae

    Zanthoxylum cf. fagara (L.) Sarg.

2

40

2

Animal5

    Zanthoxylum rhoifolium Lam.

10

33

1-3,6

Animal10

Salicaceae

    

    Casearia silvestris Sw.

12

31

2-6

Animal15

Siparunaceae

    Siparuna guianensis Aubl.

29

19

1-4,6,7

Animal4

Solanaceae

    

    Solanum lycocarpum A. St.-Hil.

48

14

1,2,4,7

Animal2

    Solanum mauritianum Scop.

58

10

1-7

Animal7

    Solanum paniculatum L.

95

5

1,2,4-7

Animal14

    Solanum pseudoquina A. St.-Hil.

1

41

5

Animal30

    Solanum reflexum Schrank

1

41

2

Animal35

    Solanum viarum Dunal

14

29

2,4,6,7

Animal19

Styracaceae

    Styrax ferrugineus Nees & Mart.

10

33

1,2,4,5

Animal15

Thymelaeaceae

    Daphnopsis fasciculata (Meisn.) Nevling

6

36

1,3,4,6,7

Animal21

Urticaceae

    

    Cecropia pachystachya Trécul

1

41

5

Animal23

Vochysiaceae

    Qualea cf. multiflora Mart.

1

41

3

Wind15

    Vochysia tucanorum Mart.

5

37

1,2

Wind31

Species are shown with the corresponding total number of sampled individuals (N), abundance rank (from the most to the least abundant), the plots at which the species was found (from 1 – closest to the cerrado edge to 7 – closest to the forest edge), dispersal syndrome (wind-, self- or animal-dispersed), and reference for the dispersal syndrome as superscripts. Families, authors, synonyms and accepted names were checked on The Plant List ([2010]) or Tropicos (Missouri Botanical Garden [2012]) databases with the Plantminer software (Carvalho et al. [2010]).

1 Almeida et al. [1998]; 2 Batalha and Mantovani [2000]; 3 Borgo [2010]; 4 Campos et al. [2009]; 5 Candiani [2006]; 6 Catharino et al. [2006]; 7 Coelho et al. [2010]; 8 Francisco and Galetti [2002]; 9 Frenedoso [2004]; 10 Galetti et al. [2011]; 11 Giehl et al. [2007]; 12 Goldenberg and Shepherd [1998]; 13 Gressler et al. [2006]; 14 Ishara and Maimoni-Rodella [2011]; 15 Jardim and Batalha [2009]; 16 Liebsch et al. [2009]; 17 Lobão and Mello-Silva [2007]; 18 Grau et al. [1997];; 19 Paise and Vieira [2005]; 20 Pascotto [2007]; 21 Polisel and Franco [2010]; 22 Ressel et al. [2004]; 23 Reys et al. [2005]; 24 Silva et al. [2009]; 25 Spina et al. [2001]; 26 Stefanello et al. [2009]; 27 Takahasi and Fina [2004]; 28 Tannus et al. [2006]; 29 Weiser and Godoy [2001]; 30 Wiesbauer et al. [2008]; 31 Yamamoto et al. [2007]; 31 Yamamoto et al. [2007] (for Myrsine umbellata); 32 Ye et al. [2004] (for Cedrela odorata); 33 Zipparro et al. [2005]; 34 Moura et al. [2011]; 35 based on species from the same genus.

Number of species and Fisher’s alpha diversity index varied from 24 to 44 among the plots. Shannon’s diversity index varied between 2.8 and 3.4 natural logarithm digits (nats) per individual. Pielou’s evenness index varied between 0.46 and 0.88. The PD varied between 1958 and 2995, PD-rate varied between 68 and 82, and the number of individuals per plot varied from 40 to 199. Most of the plants encountered were small (75.7% of individuals in the first height class, 9.6% in the second, 11.9% in the third and 2.8% in the fourth). The most abundant dispersal syndrome was animal dispersal, with 1776 individuals and 83 species, followed by wind dispersal, with 253 individuals and 31 species, and self dispersal, with 206 individuals and 9 species (Table 1).

The most abundant species were Schefflera vinosa (Cham. & Schltdl.) Frodin & Fiaschi (Araliaceae, N = 204), Ocotea pulchella (Nees & Mart.) Mez (Lauraceae, N = 171), Phyllanthus acuminatus Vahl (Phyllanthaceae, N = 123), Myrcia lingua (O. Berg.) Mattos & D. Legrand (Myrtaceae, N = 100), Solanum paniculatum L. (Solanaceae, N = 95), Myrsine umbellata Mart and Myrsine coriacea (Sw.) R.Br. ex Roem. & Schult. (Primulaceae, N = 93 and 90 respectively). The most abundant families were Myrtaceae, Solanaceae, Araliaceae, Fabaceae, and Primulaceae, with 336, 217, 204, 185 and 183 individuals respectively, and the most species-rich families were Fabaceae, Myrtaceae, Asteraceae, Melastomataceae, and Rubiaceae, with 19, 18, 11, 7 and 7 species respectively (Table 1).

Analysis of vegetation variables using quadratic and linear regressions

A single outlier, located at 322.5 m from the cerrado edge was removed from the analysis of variables related to vegetation height because this plot was partially occupied by the invasive fern Pteridium esculentum arachnoideum. The presence of this fern may have affected the size structure of the plot by reducing the number of plants in the smaller size classes. Ten out of the 23 response variables tested exhibited significant quadratic relationships with edge distance, with R2 ranging from 0.29 to 0.55 (Table 2), except for the distribution of Schefflera vinosa, which had an R2 of 0.78. Shannon’s H index, Fisher’s alpha diversity index, and the proportion of wind-dispersed individuals had peak values in the middle of the study site, whereas total number of individuals, mean height, height SD, and the distributions of S. vinosa, Myrsine umbellata and Myrsine coriacea all had a trough in the middle plots (Figures 3 and 4). Shannon’s evenness, PD-rate, height CV, and proportion of self-dispersed individuals increased from the cerrado to the forest edge, and species richness, PD, number of individuals, number of animal-dispersed individuals, and the distributions of S. vinosa, M. umbellata and M. coriacea, Phyllantus acuminatus and Ocotea pulchella all decreased from the cerrado to the forest edge (Figures 3 and 4). The R2 of the linear regressions ranged from 0.19 to 0.36. Only five variables, namely the proportion of wind-dispersed, self-dispersed and animal-dispersed species and the abundances of Myrcia lingua and Solanum paniculatum, did not present any significant distance-driven relationships (Figure 4).
Table 2

Coefficient of determination (R 2 ) of linear and quadratic regressions between the response variables and distance from cerrado edge along the transects

Response variable

Linear R2

Quadratic R2

Number of species

0.36**

0.38

Fisher’s alpha diversity index

< 0.01

0.28*

Shannon H index

0.01

0.35**

Pielou Evenness index

0.20*

0.41*

PD

0.23*

0.25

PD rate

0.41**

0.43

Number of individuals

0.37**

0.55*

Mean height

0.03

0.41**

SD of height

0.18

0.42*

CV of height

0.23*

0.35

Wind-dispersed individuals

0.01

0.28*

Self-dispersed individuals

0.21*

0.22

Animal-dispersed individuals

0.21*

0.35

Wind-dispersed species

<0.01

<0.01

Self-dispersed species

<0.01

0.02

Animal-dispersed species

<0.01

0.01

Abundance of Schefflera vinosa

0.68***

0.78*

Abundance of Ocotea pulchella

0.19*

0.27

Abundance of Phyllanthus acuminatus

0.04

0.33*

Abundance of Myrcia lingua

0.17

0.17

Abundance of Solanum paniculatum

0.09

0.09

Abundance of Myrsine coriacea

0.19*

0.43*

Abundance of Myrsine umbellata

0.33**

0.48*

For the variables related to vegetation height, we removed a single outlier, located at 322.5 m from the cerrado edge. We chose to remove this outlier because this plot was partially occupied by the invasive fern Pteridium esculentum arachnoideum, which may have affected the size structure of the plot by reducing the number of plants in the smaller size classes. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

Quadratic and linear regressions between the diversity or structural response variables and distance from cerrado edge, represented as the mid-point of each plot, along three transects in the understorey of the study site. Each plot presents two lines, one for quadratic regression, and other for linear regression. Significant regressions are presented with a continuous line, non-significant ones with an interrupted line. Please note that some of variables presented significance for both quadratic and linear regression. A: number of species (i.e. species richness), B: Fisher’s alpha diversity index, C: Shannon’s H index, D: Pielou’s evenness index, E: Phylogenetic diversity (PD), F: Phylogenetic diversity rate, G: total number of individuals, H: mean vegetation height, I: vegetation height standard deviation, J: vegetation height coefficient of variation. For the variables related to vegetation height, we removed a single outlier, located at 322.5 m from the cerrado edge and indicated by a *.

Figure 4

Quadratic and linear regressions between the dispersal mode and species abundance response variables and distance from the cerrado edge, represented as the mid-point of each plot, along three transects in the understorey of the study site. Each graph presents two lines, one for quadratic regression, and other for linear regression. Significant regressions are presented with a continuous line, non-significant ones with an interrupted line. Please note that some of variables presented significance for both quadratic and linear regression. A: proportion of wind-dispersed individuals, B: proportion of self-dispersed individuals, C: proportion of animal-dispersed individuals, D: proportion of wind-dispersed species, E: proportion of self-dispersed species, F: proportion of animal-dispersed species, G: distribution of Schefflera vinosa, H: distribution of Ocotea pulchella, I: distribution of Phyllanthus acuminatus, J: distribution of Myrcia lingua, K: distribution of Solanum paniculatum, L: distribution of Myrsine coriacea. The pattern observed for Myrsine umbellata was similar to that of Myrsine coriacea and is not shown.

Discussion

The primary result of this study was the large number of species (126) found in the understorey of the study site. Other surveys of understorey in cerrado areas containing Eucalyptus spp. trees have found much lower numbers (between 39 and 47 species) (Saporetti Jr et al. [2003], Neri et al. [2005]). This variation may be at least partially due to differences in methodology as these earlier studies only included individuals with at least 3 cm diameter at soil level. The density of individuals of native species (1700 ind.ha−1) in the current study was between two and five times smaller than observed in two nearby cerrado areas, where density ranged between 4000 and 8250 ind.ha−1 (Dantas and Batalha [2011], Gonçalves and Batalha [2011]). This difference in density is even more pronounced given that the sampling criterion used in the current study included individuals with less than 3 cm diameter. Even so, the density observed in our study is within the recommended density of pre-existent or planted trees for the restoration of cerrado areas, which ranges from 1000 up to 2000 ind.ha−1 (Durigan et al. [2003]), showing that land used previously as Eucalyptus plantations may have a large potential for natural regeneration and restoration. The potential for regeneration of cerrado vegetation has already been documented for areas previously used as pasture or as Eucalyptus plantation (Durigan et al. [1998], Jepson [2005], Neri et al. [2005]). For example, in one study in eastern Mato Grosso state, central Brazil, over 50% of once-cleared cerrado areas experienced regeneration during the 13-year study period, mostly without human intervention (passive restoration) (Jepson [2005]). In addition to the results of these studies, our study showed that the regenerating vegetation is a phylogenetically diverse community, showing that not only many species occurred but that these species belonged to phylogenetically distant groups.

The undergrowth vegetation was spatially heterogeneous, with a conspicuous gradient in the undergrowth between the cerrado and the semideciduous forest. Of the above-mentioned studies, only one (Neri et al. [2005]) explored the spatial variation in the regenerating vegetation, but on a much smaller scale than our study. As expected, the proximity to native vegetation of either cerrado or semideciduous forest influenced species composition in the transition area. It appears that the species composition of plots between 0 and 185 m from the cerrado edge was more similar to that of the cerrado than plots further away (Figure 2). Changes in relative abundances were more evident than changes in species composition, as indicated by the lack of an obvious pattern in the NMDS ordination plot based on the Jaccard Index, which is a presence-absence similarity index. The most abundant species, i.e. those that were better able to recolonise this altered area, were from distinct families at the cerrado and forest edges. Species in the family Meliaceae were more abundant at the forest edge and members of the Araliaceae more abundant at the cerrado edge. The distribution of some of the most abundant species, S. vinosa and O. pulchella, decreased from the cerrado to the forest edge. In fact, of the most abundant species, only M. lingua and S. paniculatum did not present a distance-related trend. The importance of the semideciduous forest in determining the composition of this area is emphasised by the presence of typical forest species (e.g. Cedrela fissilis and Cabralea canjerana) closer to the forest edge. Thus, even though most species occur throughout the entire area, the composition of the whole community follows the cerrado-forest gradient, probably as a result of a variety of dispersal and establishment characteristics (Wulf and Heiken [2008], Pouliot et al. [2012]) of cerrado and forest species. In disturbed areas, plasticity and order of establishment usually have more pronounced effects on species’ abundances than biotic interactions (Pouliot et al. [2012]). Competition among native species is expected to be low at the current study site due to the low density of native trees. Therefore, the spatial structure encountered suggests that dispersal and recruitment are the main determinants of the species composition in this and other altered areas.

Measures of species diversity, phylogenetic diversity, vegetation structure, dispersal syndromes and distribution of abundant species also formed spatial gradients. Notwithstanding the small sample size and large variability between plots, many quadratic regressions were significant, suggesting that the forest and cerrado areas do in fact influence the gradients in composition, structure and distributions of this understorey community. The observed gradients are not likely to be explained by resprouting, which is common for, e.g., S. vinosa and species of the genus Miconia (pers. obs.), among many others (Hoffmann and Solbrig [2003]), because resprouting would be expected to be more homogeneous throughout the study site. As large changes in soil properties are not expected at this spatial scale (Dantas and Batalha [2011]), the observed gradient in species diversity and other characteristics probably resulted from the distance to the source of diaspores, similarly to what we observed for species composition.

Phylogenetic diversity decreased from cerrado to the forest edge, mainly due to decreasing richness towards the forest edge. As speciation is not important at the time scale addressed here, the colonisation process alone turns out as the major factor in determining patterns of phylogenetic diversity. It is likely that more individuals successfully reached and established closer to the cerrado edge because of higher propagule pressure. Edge-interior gradients in species richness, density, and dispersal have also been found previously in the first 25 m from a cerrado edge with a Eucalyptus spp. plantation understorey elsewhere (Neri et al. [2005]). The current study indicates that such gradients may also occur over larger distances (up to 185 m) and also occur for phylogenetic diversity and distributions of abundant species. Furthermore, the PD-rate increased modestly from the cerrado to the forest edge, while both PD and species number decreased. This indicates that: 1) at cerrado edge, there were more species which were phylogenetically similar, so that each of them added little phylogenetic diversity to these sites; 2) conversely, there were fewer species at the forest edge, but the phylogenetic distances among these species were greater. For example, Cabralea canjerana and Cedrela fissilis (Meliaceae), species that are more common in forest than in cerrado, appeared only close to the forest edge and are from a different family than the most abundant species (Schefflera vinosa, O. pulchella, P. acuminatus, Myrcia lingua, Solanum paniculatum, Myrsine umbellata and M. coriaceae). Myrtaceae and Fabaceae were among the most abundant and species-rich families and were distributed more or less evenly along the transects.

Plants inside the study site were smaller than those at either edge. These differences were more pronounced than could be explained solely by increased light near the edges of the study site (Bowering et al. [2006]; Delgado et al. [2007]; Pohlman et al. [2007]). Time since establishment is a plausible alternative explanation. This explanation is also consistent with the results for height standard deviation and coefficient of variation, indicating that plant height in the interior of the site is more homogeneous and consequently that understorey plants had a shorter time to develop in the interior of the site than at the edges. In addition, allelopathic effects of Eucalyptus spp. could hinder development of trees, causing the trough pattern observed for height, as these effects would be stronger in the centre of the fragment (del Moral and Muller [1970]; May and Ash [1990]; Khan et al. [2008]). Overall density of all understorey species decreased linearly from the cerrado to the forest edge. This may be because cerrado species are more adapted than forest species to the nutrient-poor soils present at this study site (Haridasan [2000]; Ruggiero et al. [2002]) and so are more likely to colonise the study site than forest species. This also indicates that allelopathic effects of Eucalyptus grandis do not affect the establishment of new individuals in a spatially explicit manner. Thus, even though forest species are capable of being dispersed into the site, the establishment and survival of species dispersed from the cerrado would be greater, explaining the observed gradient. Other factors, such as differences in soil or shading, may also have played a role. However, their importance was probably smaller, as these characteristics appeared to be reasonably homogeneous in the area and the light levels were sufficient for the establishment and growth of the cerrado species.

With regard to dispersal, the proportion of animal- and self-dispersed individuals, but not species, varied along the gradient. Animal dispersal was the predominant syndrome, as expected for cerrado vegetation (Batalha and Mantovani [2000]; Jardim and Batalha [2009]), and the proportion of animal-dispersed plants decreased from the cerrado to the forest edge. A decrease in the number of animal-dispersed plants from edge to interior of a Eucalyptus sp. plantation has also been observed in a smaller-scale study (Neri et al. [2005]), whereas a study in native cerrado vegetation showed an opposite pattern, with the number of animal-dispersed individuals being smallest at the edge (Jardim and Batalha [2009]). From these results, it seems that the fauna in the cerrado that is capable of dispersing seeds avoids both agricultural land (Jardim and Batalha [2009]) and the interior of ex-plantation areas (Neri et al. [2005] and present study). It is known that Eucalyptus plantations may be used as corridor and habitat by top predators such as Puma concolor and Chrysocyon brachiurus (Lyra-Jorge et al. [2010]). However, the use of such areas by seed-dispersing fauna in general seems to be reduced, indicating that this is a poor habitat for a large proportion of the fauna, especially birds. This is similar to what has been observed in a study of avifauna in fragments of Atlantic Forest and Eucalyptus sp. plantation, where the number of birds recorded in the plantation was less than half of that recorded in the forest fragments in spite of a well-developed undergrowth (Dário et al. [2002]). Thus, the use of study area as a dispersal corridor by animals is probably below the desired levels.

Conclusions

The area studied here was located between two different vegetation types and gradients were observed in both species composition and vegetation structure. The results from this study indicate that the presence of Eucalyptus grandis trees from an abandoned plantation does not preclude colonisation of native species from neighbouring cerrado and forest patches. The understorey composition resulting from the succession process was spatially heterogeneous, with a clear pattern in vegetation structure and composition. Thus, the main driving factor of heterogeneity is the distance to areas of native vegetation. Natural regeneration under Eucalyptus plantations may be a viable form of restoration given adequate management, and any removal of residual Eucalyptus trees should be done gradually to prevent damage to the existing understorey.

Abbreviations

CV: 

Coefficient of variation:

PD: 

Phylogenetic diversity:

SD: 

Standard deviation:

Declarations

Acknowledgements

We are very grateful: to Jorge Yoshio Tamashiro for identifying many of species; Eduardo Hettwer Giehl, Igor Aurélio, Anne Maguran and Karen Harper, as well as the editor and three anonymous reviewers, for suggestions on a previous version of this manuscript; CNPq (National Counsel of Technological and Scientific Development) for scholarship granted to PD; Fapesp (State of São Paulo Research Foundation) for scholarship granted to NBR; and Capes (Coordination for the Improvement of Higher Education Personnel) for scholarship granted to DMS.

Authors’ Affiliations

(1)
Department of Hidrobiology, Federal University of São Carlos
(2)
Department of Environmental Sciences, Federal University of São Carlos
(3)
Department of Botany, Federal University of São Carlos

References

  1. Almeida SP, Proença CEB, Sano SM, Ribeiro JF: Cerrado: espécies vegetais úteis. Embrapa - CPAC, Planaltina; 1998.Google Scholar
  2. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III Botanical Journal of the Linnean Society 2009, 161: 105–121. 10.1111/j.1095-8339.2009.00996.xGoogle Scholar
  3. Batalha MA, Mantovani W: Some phenological patterns of cerrado plant species at the Pé-de-Gigante reserve (Santa Rita do Passa Quatro, SP, Brazil): a comparison between the herbaceous and woody floras. Brazilian Journal of Biology 2000, 59: 1–16.Google Scholar
  4. Borgo M: A Floresta Atlântica do litoral norte do Paraná, Brasil: aspectos florísticos, estruturais e estoque de biomassa ao longo do processo sucessional. PhD thesis, Universidade Federal do Paraná, Curitiba; 2010.Google Scholar
  5. Bowering M, Lemay V, Marshall P: Effects of forest roads on the growth of adjacent lodgepole pine trees. Canadian Journal of Forest Research 2006, 36: 919–929. 10.1139/x05-300View ArticleGoogle Scholar
  6. Cadenasso ML, Pickett STA, Weathers KC, Jones CG: A framework for a theory of ecological boundaries. Bioscience 2003, 53: 750–758. 10.1641/0006-3568(2003)053[0750:AFFATO]2.0.CO;2View ArticleGoogle Scholar
  7. Campos EP, Vieira MF, Silva AF, Martins SV, Carmo FMS, Moura VM, Ribeiro ASS: Chuva de sementes em Floresta Estacional Semidecidual em Viçosa, MG, Brasil. Acta Botanica Brasilica 2009, 23: 451–458. 10.1590/S0102-33062009000200017View ArticleGoogle Scholar
  8. Candiani G: Regeneração natural em áreas anteriormente ocupadas por florestas de Eucalytus saligna Smith. no município de Caieiras (SP): subsídios para recuperação florestal, MSc dissertation, Instituto de Botânica, Secretaria do Meio Ambiente do Estado de São Paulo; 2006.Google Scholar
  9. Carvalho GH, Cianciaruso MV, Batalha MA: Plantminer: A web tool for checking and gathering plant species taxonomic information. Environmental Modelling & Software 2010, 25: 815–816. 10.1016/j.envsoft.2009.11.014View ArticleGoogle Scholar
  10. Catharino ELM, Bernacci LC, Franco GADC, Durigan G, Metzger JP: Aspectos da composição e diversidade do componente arbóreo das florestas da Reserva Florestal do Morro Grande, Cotia, SP. Biota Neotropica 2006,6(2):art2. 10.1590/S1676-06032006000200004View ArticleGoogle Scholar
  11. Chase JM: Community assembly: when should history matter? Oecologia 2003, 136: 489–498. 10.1007/s00442-003-1311-7View ArticlePubMedGoogle Scholar
  12. Coelho GC, Rigo MS, Libardoni JB, Oliveira R, Benvenuti-Ferreira G: Understory structure in two successional stages of a Semi-deciduous Seasonal Forest remnant of southern Brazil. Biota Neotropica 2010,11(3):art4.Google Scholar
  13. Dantas VL, Batalha MA: Vegetation structure: fine scale relationships with soil in a cerrado site. Flora 2011, 206: 341–346. 10.1016/j.flora.2010.11.003View ArticleGoogle Scholar
  14. Dário FR, De Vincenzo MCV, Almeida AF: Avifauna em fragmentos de mata atlântica. Ciencia Rural 2002, 32: 989–996. 10.1590/S0103-84782002000600012View ArticleGoogle Scholar
  15. del Moral R, Mullher CH: The allelopathic effects of Eucalyptus camaldulensis . American Midland Naturalist 1970, 83: 254–282. 10.2307/2424020View ArticleGoogle Scholar
  16. Delgado JD, Arroyo NL, Arévalo JS, Fernández-Palacios JM: Edge effects of roads on temperature, light, canopy cover, and canopy height in laurel and pine forests (Tenerife, Canary Islands). Landscape and Urban Planning 2007, 81: 328–340. 10.1016/j.landurbplan.2007.01.005View ArticleGoogle Scholar
  17. Durigan G, Ratter JA: Successional changes in cerrado and cerrado/forest ecotonal vegetation in western São Paulo state, Brazil, 1962–2000. Edinburgh Journal of Botany 2006, 63: 119–130. 10.1017/S0960428606000357View ArticleGoogle Scholar
  18. Durigan G, Contiere WA, Franco GADC, Garrido MAO: Indução do processo de regeneração da vegetação de cerrado em área de pastagem, Assis, SP. Acta Botanica Brasilica 1998, 12: 421–429. 10.1590/S0102-33061998000400011View ArticleGoogle Scholar
  19. Durigan G, Melo ACG, Max JCM, Bôas OV, Contiere WA: Manual para recuperação da vegetaçào de cerrado. Instituto Florestal, São Paulo; 2003.Google Scholar
  20. Durigan G, Siqueira MF, Franco GADC: Threats to the Cerrado remnants of the state of São Paulo, Brazil. Scientia Agricola 2007, 64: 355–363. 10.1590/S0103-90162007000400006View ArticleGoogle Scholar
  21. Faith D: Conservation evaluation and phylogenetic diversity. Biological Conservation 1992, 61: 1–10. 10.1016/0006-3207(92)91201-3View ArticleGoogle Scholar
  22. Francisco MR, Galetti M: Aves como potenciais dispersoras de sementes de Ocotea pulchella Mart. (Lauraceae) numa área de vegetação de cerrado do sudeste brasileiro. Revista Brasileira de Botânica 2002, 25: 11–17.Google Scholar
  23. Frenedozo RC: Plant reproductive phenology and dispersal patterns after natural regeneration in a limestone mining spoil banks. Brazilian Archives of Biology and Technology 2004, 47: 261–271. 10.1590/S1516-89132004000200014View ArticleGoogle Scholar
  24. Furley PA, Proctor J, Ratter JA: Nature and dynamics of forest-savanna boundaries. Chapman and Hall, London; 1992.Google Scholar
  25. Galetti M, Pizo MA, Morellato LPC: Diversity of functional traits of flesh fruits in a species-rich Atlantic rain forest. Biota Neotropica 2011, 11: art18. 10.1590/S1676-06032011000100019View ArticleGoogle Scholar
  26. Giehl ELH, Athayde EA, Budke JC, Gesing JPA, Eisinger SM, Canto-Dorow TS: Espectro e distribuição vertical das estratégias de dispersão de diásporos do componente arbóreo em uma floresta estacional no sul do Brasil. Acta Botanica Brasilica 2007, 21: 137–145. 10.1590/S0102-33062007000100013View ArticleGoogle Scholar
  27. Goldenberg R, Shepherd GJ: Studies on the reproductive biology of Melastomataceae in “cerrado” vegetation. Plant Systematics and Evolution 1998, 211: 13–29. 10.1007/BF00984909View ArticleGoogle Scholar
  28. Gonçalves CS, Batalha MA: Towards testing the “honeycomb rippling model” in cerrado. Brazilian Journal of Biology 2011, 71: 401–408.Google Scholar
  29. Grau HR, Arturi MF, Brown AD, Acenõlaza PG: Floristic and structural patterns along a chronosequence of secondary forest succession in Argentinean Subtropical Montane Forest. Forest Ecology and Management 1997, 95: 161–171. 10.1016/S0378-1127(97)00010-8View ArticleGoogle Scholar
  30. Gressler E, Pizo MA, Morellato LPC: Polinização e dispersão de sementes em Myrtaceae do Brasil. Revista Brasileira de Botânica 2006, 29: 509–530.Google Scholar
  31. Hammer Ø, Harper DAT, Ryan PD: PAST: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 2001, 4: 9. Available at http://palaeo-electronica.org/2001_1/past/issue1_01.htm. Accessed 24 Jan 2014Google Scholar
  32. Haridasan M: Nutrição mineral de plantas nativas do cerrado. Revista Brasileira de Fisiologia Vegetal 2000, 12: 54–64.Google Scholar
  33. Harper KA, Macdonald SE, Burton PK, Chen JQ, Brosofske KD, Saunders SC, Euskirchen ES, Roberts D, Jaiteh MS, Esseen PA: Edge influence on forest structure and composition in fragmented landscapes. Conservation Biology 2005, 19: 768–782. 10.1111/j.1523-1739.2005.00045.xView ArticleGoogle Scholar
  34. Hoffmann WA, Solbrig OT: The role of topkill in the differential response of savanna woody species to fire. Forest Ecology and Management 2003, 180: 273–286. 10.1016/S0378-1127(02)00566-2View ArticleGoogle Scholar
  35. Ishara KL, Maimoni-Rodella RCS: Pollination and dispersal systems in a cerrado remnant (Brazilian Savanna) in southeastern Brazil. Brazilian Archives of Biology and Technology 2011, 54: 629–642. 10.1590/S1516-89132011000300025View ArticleGoogle Scholar
  36. Jardim AVF, Batalha MA: Dispersal syndromes related to edge distance in cerrado sensu stricto fragments of Central-Western Brazil. Brazilian Archives of Biology and Technology 2009, 52: 1167–1177. 10.1590/S1516-89132009000500014View ArticleGoogle Scholar
  37. Jepson W: A disappearing biome? Reconsidering land-cover change in the Brazilian savanna. The Geographical Journal 2005, 171: 99–111. 10.1111/j.1475-4959.2005.00153.xView ArticleGoogle Scholar
  38. Jost L, Chao A, Chazdon RL: Compositional similarity and ß diversity. In Biological diversity: frontiers in measurement and assessment. Edited by: Magurran AE, McGill BJ. Oxford University Press, Oxford, Oxford; 2011:66–84.Google Scholar
  39. Khan MA, Hussain I, Khan EA: Suppressing effects of Eucalyptus camaldulensis L. on germination and seedling growth of six weeds. Pakistan Journal of Weed Science Research 2008, 14: 201–207.Google Scholar
  40. Lebrija-Trejos E, Perez-Garcia EA, Meave JA, Bongers F, Poorter L: Functional traits and environmental filtering drive community assembly in a species-rich tropical system. Ecology 2010, 91: 386–398. 10.1890/08-1449.1View ArticlePubMedGoogle Scholar
  41. Liebsch D, Mikich SB, Possette RFS, Ribas OS: Levantamento florístico e síndromes de dispersão em remanescentes de Floresta Ombrófila Mista na região centro-sul do estado do Paraná. Hoehnea 2009, 36: 233–248. 10.1590/S2236-89062009000200002View ArticleGoogle Scholar
  42. Lobão AQ, Mello-Silva R: Guatteria (Annonaceae) do estado do Rio de Janeiro, Brasil. Rodriguésia 2007, 58: 859–884.Google Scholar
  43. Lyra-Jorge MC, Ciocheti G, Pivello VR: Carnivore mammals in a fragmented landscape in northeast of São Paulo State, Brazil. Biodiversity and Conservation 2008, 17: 1573–1580. 10.1007/s10531-008-9366-8View ArticleGoogle Scholar
  44. Lyra-Jorge MC, Ribeiro MC, Ciocheti G, Tambosi LR, Pivello VR: Influence of multi-scale landscape structure on the ocurrence of carnivorous mammals in a human-modified savanna, Brazil. European Journal of Wildlife Research 2010, 56: 359–368. 10.1007/s10344-009-0324-xView ArticleGoogle Scholar
  45. Machado RB, Lamas IR: Avifauna associada a um reflorestamento de eucalipto no município de Antônio Dias, Minas gerais. Ararajuba 1996, 4: 15–22.Google Scholar
  46. Magurran AE: Biological diversity and its measurement. Princeton University Press, Princeton; 1988.View ArticleGoogle Scholar
  47. Maurer BA, Mcgill BJ: Measurement of species diversity. In Biological diversity: frontiers in measurement and assessment. Edited by: Magurran AE, McGill BJ. Oxford University Press, Oxford; 2011:55–65.Google Scholar
  48. May FE, Ash JE: An assessment of the allelopathic potential of Eucalyptus. Australian Journal of Botany 1990, 38: 245–254. 10.1071/BT9900245View ArticleGoogle Scholar
  49. Miranda HS, Bustamante MMC, Miranda AC: The fire factor. In The Cerrados of Brazil: ecology and natural history of a neotropical savanna. Edited by: Oliveira PS, Marquis RJ. Columbia University Press, New York; 2002:51–68.Google Scholar
  50. Missouri Botanical Garden. (2012). Tropicos database. . Acessed 11 June 2012., [http://www.tropicos.org/]
  51. Moreira AG: Effects of fire protection on savanna structure in Central Brazil. Journal of Biogeography 2000, 27: 1021–1029. 10.1046/j.1365-2699.2000.00422.xView ArticleGoogle Scholar
  52. Moura FBP, Duarte JMM, Lyra-Lemos RP: Floristic composition and dispersal syndromes at an urban remnant from the Atlantic forest in Brazilian northeast. Acta Scientiarum Biological Sciences 2011, 33: 471–478.Google Scholar
  53. Neri AV, Campos EP, Duarte TG, Meira Neto JAA, Silva AF, Valente GE: Regeneração de espécies nativas lenhosas sob plantio de Eucalyptus em área de cerrado na Floresta Nacional de Paraopeba, MG, Brasil. Acta Botanica Brasilica 2005, 19: 369–376. 10.1590/S0102-33062005000200020View ArticleGoogle Scholar
  54. Paise G, Viera EM: Produção de frutos e distribuição espacial de angiospermas com frutos zoocóricos em uma Floresta Ombrófila Mista no Rio Grande do Sul, Brasil. Revista Brasileira de Botânica 2005, 28: 615–625.Google Scholar
  55. Pascotto MC: Rapanea ferruginea (Ruiz and Pav.) Mez. (Myrsinaceae) como uma importante fonte alimentar para as aves em uma mata galeria no interior do estado de São Paulo. Revista Brasileira de Zoologia 2007, 24: 735–741. 10.1590/S0101-81752007000300026View ArticleGoogle Scholar
  56. Pohlman CL, Turton SM, Goosem M: Edge effects of linear canopy openings on tropical rain forest understorey microclimate. Biotropica 2007, 39: 62–71. 10.1111/j.1744-7429.2006.00238.xView ArticleGoogle Scholar
  57. Polisel RT, Franco GADC: Comparação florística e estrutural entre dois trechos de Floresta Ombrófila Densa em diferentes estádios sucessionais, Juquitiba, SP, Brasil. Hoehnea 2010, 37: 691–718. 10.1590/S2236-89062010000400002View ArticleGoogle Scholar
  58. Pouliot R, Rochefort L, Karofeld E: Initiation of microtopography in re-vegetated cutover peatlands: evolution of plant species composition. Applied Vegetation Science 2012, 15: 369–382. 10.1111/j.1654-109X.2011.01164.xView ArticleGoogle Scholar
  59. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; 2012.Google Scholar
  60. Ressel K, Guilherme FAG, Schiavini I, Oliveira PE: Ecologia morfofuncional de plântulas de espécies arbóreas da Estação Ecológica do Panga, Uberlândia, Minas Gerais. Revista Brasileira de Botânica 2004, 27: 311–323.Google Scholar
  61. Reys P, Galetti M, Morellato LPC, Sabino J: Fenologia reprodutiva e disponibilidade de frutos de espécies arbóreas em mata ciliar no rio Formoso, Mato Grosso do Sul. Biota Neotropica 2005, 5: ShortComm2. 10.1590/S1676-06032005000300021View ArticleGoogle Scholar
  62. Ruggiero PGC, Batalha MA, Pivello VR, Meirelles ST: Soil–vegetation relationships in cerrado (Brazilian savanna) and semideciduous forest, southeasthern Brazil. Plant Ecology 2002, 160: 1–16. 10.1023/A:1015819219386View ArticleGoogle Scholar
  63. Santos JE, Paese A, Pires JSR: Unidades da paisagem (biótopos) do câmpus da UFSCar. EdUFSCar, São Carlos; 1999.Google Scholar
  64. Saporetti AW Jr, Meira Neto JAA, Almado R: Phytosociology of a cerrado understorey in a stand of Eucalyptus grandis W. Hill ex Maiden in Bom Despacho-MG. Revista Árvore 2003, 27: 905–910.Google Scholar
  65. Silva IA, Cianciaruso MV, Batalha MA: Dispersal modes and fruiting periods in hyperseasonal and seasonal savannas, central Brazil. Revista Brasileira de Botânica 2009, 32: 155–163.Google Scholar
  66. Spina AP, Ferreira WM, Leitao Filho HF: Floração, frutificação e síndromes de dispersão de uma comunidade de floresta de brejo na região de Campinas (SP). Acta botanica Brasilica 2001, 15: 349–368. 10.1590/S0102-33062001000300006View ArticleGoogle Scholar
  67. Stefanello D, Fernandes-Bulhão C, Martins V: Síndromes de dispersão de sementes em três trechos de vegetação ciliar (nascente, meio e foz) ao longo do rio Pindaíba, MT. Revista Árvore 2009, 33: 1051–1061. 10.1590/S0100-67622009000600008View ArticleGoogle Scholar
  68. Taguchi Y–H, Oono Y: Nonmetric multidimensional scaling as a data-mining tool: new algorithm and new targets. In Geometric structures of phase space in multidimensional chaos: applications to phemical reaction dynamics in complex systems. Edited by: Toda M, Komatsuzaki T, Konishi T, Berry RS, Rice SA. John Wiley & Sons, Hoboken; 2005.Google Scholar
  69. Takahasi A, Fina BG: Síndromes de dispersão de sementes de uma área do Morro de Paxixi, Aquidauana, MS, Brasil. Corumbá, MS, Brazil; 2004.Google Scholar
  70. Tannus JLS, Assis MA, Morellato LPC: Fenologia reprodutiva em campo sujo e campo úmido numa área de Cerrado no sudeste do Brasil, Itirapina – SP. Biota Neotropica 2006, 6: art8. 10.1590/S1676-06032006000300008View ArticleGoogle Scholar
  71. The Plant List. (2010). Version 1. . Acessed 11 June 2012., [http://www.theplantlist.org/]
  72. Török P, Kelemen A, Valkó O, Deák B, Lukács B, Tóthmérész B: Lucerne-dominated fields recover native grass diversity without intensive management actions. Journal of Applied Ecology 2011, 48: 257–264. 10.1111/j.1365-2664.2010.01903.xView ArticleGoogle Scholar
  73. Webb CO, Donoghue MJ: Phylomatic: tree assembly for applied phylogenetics. Molecular Ecology Notes 2005, 5: 181–183. 10.1111/j.1471-8286.2004.00829.xView ArticleGoogle Scholar
  74. Webb CO, Ackerly D, McPeek M: Phylogenies and community ecology. Annual Review of Ecology, Evolution and Systematics 2002, 33: 475–505. 10.1146/annurev.ecolsys.33.010802.150448View ArticleGoogle Scholar
  75. Webb CO, Ackerly DD, Kembel SW: Phylocom: software for the analysis of phylogenetic community structure and trait evolution. Bioinformatics 2008, 24: 2098–2100. 10.1093/bioinformatics/btn358View ArticlePubMedGoogle Scholar
  76. Weisbauer MB, Giehl ELH, Jarenkow JA: Padrões morfológicos de diásporos de árvores e arvoretas zoocóricas no Parque Estadual de Itapuã, RS, Brasil. Acta Botanica Brasilica 2008, 22: 425–435. 10.1590/S0102-33062008000200012View ArticleGoogle Scholar
  77. Weiser VL, Godoy SAP: Florística em um hectare de cerrado stricto sensu na ARIE – Cerrado Pé-de-Gigante, Santa Rita do Passa Quatro, SP. Acta Botanica Brasilica 2001, 15: 201–212. 10.1590/S0102-33062001000200007View ArticleGoogle Scholar
  78. Wikström N, Savolainen V, Chase MW: Evolution of the angiosperms: calibrating the family tree. Proceedings of the Royal Society B: Biological Sciences 2001, 268: 2211–2220. 10.1098/rspb.2001.1782PubMed CentralView ArticlePubMedGoogle Scholar
  79. Wolters M, Garbutt RA, Bakker JP: Plant colonisation after managed realignment: the relative importance of diaspore dispersal. Journal of Applied Ecology 2005, 42: 770–777. 10.1111/j.1365-2664.2005.01051.xView ArticleGoogle Scholar
  80. Wulf M, Heinken T: Colonization of recent coniferous versus deciduous forest stands by vascular plants at the local scale. Applied Vegetation Science 2008, 11: 307–316. 10.3170/2008-7-18432View ArticleGoogle Scholar
  81. Yamamoto LF, Kinoshita LS, Martins FR: Síndromes de polinização e dispersão em fragmentos da Floresta Estacional Semidecídua Montana, SP, Brasil. Acta Botanica Brasilica 2007, 21: 553–573. 10.1590/S0102-33062007000300005View ArticleGoogle Scholar
  82. Ye WH, Mu HP, Cao HL, Ge XJ: Genetic structure of the invasive Chromolaena odorata in China. Weed Research 2004, 44: 129–135. 10.1111/j.1365-3180.2004.00381.xView ArticleGoogle Scholar
  83. Zipparro VB, Guilherme FAG, Almeida-Scabbia RJ, Morellato LPC: Levantamento florístico de Floresta Atlântica no sul do estado de São Paulo, Parque Estadual Intervales, base Saibadela. Biota Neotropica 2005,5(1):Inv1. 10.1590/S1676-06032005000100015View ArticleGoogle Scholar

Copyright

© Dodonov et al.; licensee Springer 2014

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.