Vegetation patterns and the influence of rainfall after long-term fire suppression on a woody community of a Brazilian savanna

Abstract

We evaluated the structural and floristic characteristics of a Brazilian savanna fragment occupied by cerradão (CD) and cerrado sensu stricto (CS) in response to the influence of rainfall and long-term fire suppression. We carried out floristic, phytosociological and remote sensing studies in a cerrado fragment located in Corumbataí (SP, Brazil) after 43 years of complete fire suppression. We surveyed 43 plots of 200 m² each (17 plots in CS and 26 plots in CD) and all individuals ≥ 0.32 cm diameter measured at 30 cm from the ground were included in the sample. We calculated phytosociological parameters for each species and classified them in three ecological groups, namely savanna, generalist and forest species. The remote sensing analysis used aerial photographs and satellite images from 1962 to 2019 (i.e. 59 years). The structural study of community revealed high predominance of forest and generalist species when compared to savanna species. Non-linear correlation between CD expansion rates and total rainfall within the study period indicated a positive influence of the rainfall (R² = 0.42). Thus, our analysis indicated a tendency of a continuous and fast expansion of CD over areas of CS in the long-term absence of fire combined with periods of heavy rain.

Key words
Competition; ecological succession; local extinctions; microclimate; remote sensing

INTRODUCTION

The Brazilian tropical savanna, also known as Cerrado, is composed by different vegetation types, such as cerrado sensu stricto (discontinuous tree cover with a herbaceous shade-intolerant vegetation) and cerradão (forest vegetation type) (Furley 1999, Ratter et al. 2006). The dynamics and structure of different vegetation physiognomies in Cerrado are mainly influenced by climatic and edaphic factors (Eiten 1972, Goodland & Pollard 1973, Puigdefabregasand & Pugnaire 1999, Wiegand et al. 2006, Meyer et al. 2007, Wright 2007). However, other factors play an important role in the vegetation establishment, namely fire occurrence, soil moisture, geomorphologic and topographic components (Solbrig 1996, Pinheiro et al. 2010, Silva & Batalha 2008, Assis et al. 2011, Silva et al. 2013a, Bueno et al. 2018).

Large-scale fire disturbances frequently affect extensive vegetation areas, favoring the establishment of fire-prone plant species and the occurrence of savanna formations with less woody cover (Coutinho 1990, Hutchings 2007). Conversely, fire suppression in plant savanna communities (Miranda et al. 2009) can influence the natural vegetation dynamics by increasing the prevalence of plant formations with greater coverage of woody plants (Coutinho 1990, Crawley 2007). Indeed, the long-term absence of fire can trigger ecological succession tending to an edaphoclimatic climax of greater phytomass (Pinheiro & Durigan 2009) frequently dominated by fire-sensitive forest species (van der Maarel 1988, Hoffmann et al. 2012, Veldman et al. 2015a).

The Cerrado biome suffered a large area reduction in the last decades, but ecologically important patches of native vegetation remain in the state of São Paulo (Siqueira & Durigan 2007, Durigan et al. 2007). Some of them have a history of long-term fire suppression, e.g. a savanna remnants located in the district of Emas, Pirassununga (Coutinho 1990) and Assis Ecological Station, located in Assis (Durigan & Ratter 2006, Pinheiro & Durigan 2009).

The existence of savanna remnants with long-term fire suppression give scope for studying the consequences of fire absence on the floristic composition and vegetation structure. Thus, we evaluated the current floristic and structural characteristics of a savanna remnant with ca. 40 years of fire suppression, located in the Corumbataí municipality (SP, Brazil). In this locality, vegetation descriptions dated from the 1960s and 1970s indicate the dominance of open-savanna environments (Camargo & Arens 1967, Piccolo et al. 1971). No information is available on the structural and floristic data in this period.

In addition to the wildfires, periods of intense rainfall in savanna-forest ecotones can greatly influence the expansion of forest formations (Bowman et al. 2001, Banfai & Bowman 2005). Therefore, good conditions for forest expansion over savanna areas involve the combination of rainy periods and fire suppression (Brook & Bowman 2006, Ondei et al. 2017).

We expected that forest and generalist species would benefit from long-term fire suppression and higher water availability (Pinheiro et al. 2010) when compared to savanna species. Therefore, we expected an increase in woody cover, a pattern observed in other regions of Brazil (Durigan & Ratter 2006, Hoffmann et al. 2012). We also expected an encroachment of cerradão over areas of cerrado sensu stricto during periods of heavy rain, i.e. changes in the structural and floristic composition of the plant community due to the predominance of non-savanna species.

MATERIALS AND METHODS

Study area

The survey occurred in a cerrado fragment belonging to Instituto de Biociências of UNESP, campus Rio Claro. The study area is named Reserva de Cerrado “Prof. Karl Arens” and is located on the steep slopes of the Corumbataí municipality (SP, Brazil) (22º14’21.68” S - 47º40’52.85” W and 22º14’41.71” S - 47º41’14.80” W), in the transition between Depressão Periférica Norte and Cuestas Basálticas (Troppmair 2000). The cerrado fragment in Corumbataí has 38.8 ha and has been isolated for many years from other natural plant formations by pastures and sugar cane plantations. The study area has been effectively protected against cattle entrance and wood removal since the land purchase by UNESP in 1962. According to some university employees, fires no longer occurred after the property was purchased.

The climate is tropical humid with a dry (April – September) and wet (October – March) season. According to Köppen, the climate is Cwa (Alvares et al. 2014). The soil was classified as Red-Yellow Latosol, sandy phase, according to the classification adopted by Tauk & Marco (1990). The area has groundwater located 30 m below the surface (Camargo & Arens 1967). The nearest remnants of native vegetation were composed by seasonal forests on the top of hills and slopes and located a few kilometers away from the study area.

Previous descriptions of the vegetation in the study area by Camargo & Arens (1967) and Piccolo et al. (1971) indicate a dominance of open-savanna environments until the early 1970s. These authors did not provide detailed structural and floristic characteristics of the study area nor mentioned the occurrence of non-savanna species. The latter is a coherent information due to the predominance of heliophilous environments, where herbaceous-shrub species have been abundant, as described by these authors.

In our study period, the cerradão dominated almost the entire reserve “Prof. Karl Arens” and cerrado sensu stricto was restricted to a smaller area. We describe cerrado sensu stricto as a vegetation presenting 20 to 30% of woody cover, which varies from 3 to 8 m in height, and cerradão as a vegetation that has a forest aspect with a height ranging from 12 to 15 m (Furley 1999, Ratter et al. 2006). Cerradão vegetation predominantly occurs in deep, sandy, dystrophic, acid soils, and in regions where the annual rainfall is lower than 1600 mm (Eiten 1972, Coutinho 1990, Solbrig 1996, Pennington et al. 2006, Ratter et al. 2006). Representatives of the herbaceous community in the study area are Aristida pallens Cav. (Poaceae), Solanum aculeatissimum Jacq. (Solanaceae) and species of the genus Sida L. (Malvaceae) and Cyperus L. (Cyperaceae).

Experimental design

The structural and floristic data used in this study were obtained from a phytosociological survey concluded by us in 2005, after 43 years without fire in the study area. We installed 43 plots measuring 200 m² (10 m x 20 m) along a topo-edaphic gradient. The plots were positioned in a continuous way reaching 860 m in length and encompassed both vegetation physiognomies: cerrado sensu stricto located in the lower part of the area (17 plots) and cerradão located in the upper part (26 plots). The soil in the upper part of the area was mainly composed by clay and silt, while particles of fine and coarse sand were predominant in the lower part (see Pinheiro et al. 2010). We measured the diameters and heights of all tree-shrub individuals with diameter of ≥ 0.32 cm, measured at 30 cm from the ground.

When the specimens presented flowers and fruits, samples were collected and included in herbarium - Herbário Rioclarense (HRCB). The correct spelling of the scientific names, as well as the abbreviations of the authors’ names and the existence of synonymy were checked in floristic lists available in the website reflora.jbrj.gov.br - Programa Reflora.

We analyzed aerial and Landsat images to evaluate the expansion or retraction of each vegetation type. Thus, the visual interpretation of aerial photographs and Landsat images was performed using the hierarchical land use and land cover classification system (Anderson et al. 1976). That procedure classifies land use types at different levels of detail, i.e. relates each level of classification to the types of remote sensing data used. Thus, the caption of the Cerrado physiognomies map was assigned according to the assumptions made by Pereira et al. (1989) and Spínola et al. (2007), where the cerrado sensu stricto and cerradão physiognomies were identified based on color (tone), shape and texture of the objects distributed across the studied area.

Thus, the cerrado sensu stricto, observed in both aerial and satellite images, presented less intense hue and less roughness when compared to the cerradão, which has a more prominent color and texture. For the more accurate determination of each physiognomy in the Landsat Satellite images, we performed another process that resulted in vegetation index images. Vegetation indices have satisfactorily indicated the leaf phytomass by area (Pereira et al. 1989). Thus, the normalized vegetation index (NDVI) was used to delimit the two savanna physiognomies analyzed (Rouse et al. 1973). This index can be expressed by the following equation:

$$\text{NDVI = (NIR - R) / (NIR + R)}$$

where: NIR = near infrared (Band 4), R = red (Band 3). This index ranges from -1 to 1, depending on the vegetation density. Negative values represent bare soils, whereas dense forests will present values closer to 1 (Bitencourt et al. 1997). According to the same authors, cerrado sensu stricto physiognomies presented NDVI values between 0.2630 and 0.3813, while cerradão presented NDVI indices ranging from 0.4406 to 0.5589. Thus, the two Cerrado physiognomies could be differentiated and mapped.

The maps showing the areas occupied by cerradão and cerrado sensu stricto of the study area along 57 years were elaborated based on aerial photographs of 1962 (scale = 1:25000), 1972 (1:25.000), 1978 (1:35.000), 1988 (1:40.000), 1995 (1:25.000) and 2000 (1:30.000), by using a mirror stereoscope. The aerial photographs were taken by the BASE Aerofotogrametria e Projetos AS company, except those from 1972 and 1978, which were the responsibility of IBC/GERCA company. Images from the satellites LANDSAT for the years of 2005 (LT5), 2010 (LT5), 2015 (LC8) and 2019 (LC8) were also used in this study and were provided by United States Geological Survey (USGS) (Figure 1).

Figure 1
Spatio-temporal variation of cerradão (CD) and cerrado sensu stricto (CS) physiognomies in a cerrado fragment located in Corumbataí (State of São Paulo, Brazil) along 57 years (1962 to 2019) of fire suppression.

Data analyses

We obtained phytosociological parameters for each species by using the software FITOPAC 2.1 (Shepherd 2009), namely relative density, frequency and dominance. The Importance Value Index (IVI) resulted from the sum of these parameters (Mueller-Dombois & Ellenberg 1974). The values were obtained by following the equations below:

$$\text{DeR}=\text{n}i/\Sigma \text{n x 100}$$

$$\text{FeR}=\text{F}i/\Sigma \text{AF x 100}$$

$$\text{DoR}=\text{BA}i/\Sigma \text{BAt x 100}$$

where: DeR - relative density; ni - number of individuals sampled from species i; n - total number of individuals sampled; FeR - relative frequency; Fi - absolute frequency of species i; AF - absolute frequency of all sampled species; DoR - relative dominance; Bai - basal area of species i; BAt - total basal area; IVI = DeR + FeR + DoR (Damasceno-Junior & Pott 2011).

The frequency of height and diameter classes were presented in histograms to analyze the age structure of the community and recruitment patterns (Newton 2007), following the method used by Spiegel & Stephens (2009).

We have considered the concept fidelity to communities (Braun-Blanquet 1979) when analysing the structural and floristic changes in the study area. This concept represents the degree of fidelity of plant species to a particular vegetation physiognomy, directly influenced by the physical-chemical conditions of the habitat and by the species’ ability to get resources under competition. This concept is useful to discuss the temporal changes we have observed in our study (Grime 1979, van der Maarel 2005), since it highlights species that have similar performance in the ecosystem, based on a common set of biological attributes (van der Maarel 2005). We classified plant species in three ecological groups, namely forest, generalist and savanna species (Felfili et al. 2006, Andersen et al. 2007, Parr et al. 2012, Villard & Metzger 2014). Savanna and forest species are restricted to specific habitat conditions, depending on fire occurrences and light availability (Hoffmann et al. 2012, Pellegrini et al. 2016). Conversely, generalist species may grow under a wider range of conditions due to their high phenotypic plasticity (Maracahipes et al. 2018). We have assumed that species restricted to specific habitats are better indicators of environmental changes, since they are more susceptible to disturbance (Cáceres & Legendre 2009).

Our classification in three ecological groups was based on species habitats, available on the website of Flora do Brasil 2020 - Programa Reflora (reflora.jbrj.gov.br/reflora). Data validation and additional information were obtained from other sources (Rizzini 1971, Heringer et al. 1977, Leitão Filho 1992, Mendonça et al. 1998, Castro et al. 1999, Ratter et al. 2000, 2003, 2006, Durigan et al. 2004, IPJBRJ 2010a, b).

We calculated the area (in hectares) occupied by the two vegetation physiognomies along 57 years, and we plotted trend curves of cerrado sensu stricto retraction (exponential) and cerradão expansion (logarithmic) over the study period (Figure 2). We performed non-linear correlation analysis (Bates & Watts 1988, Mazucheli & Achcar 2002) between total rainfall (independent variable) and expansion rates of cerradão (dependent variable) in nine periods (1962-1971, 1972-1977, 1978-1987, 1988-1994, 1995-1999, 2000-2004, 2005-2009, 2010-2014, 2015-2019) – see Pinheiro et al. (2010) for more details. In our study, we assumed that the long-term fire suppression and rainfall were important factors to the encroachment of cerradão over areas of cerrado sensu stricto.

Figure 2
Variation of the area (hectares) occupied by cerradão and cerrado sensu stricto in a cerrado fragment in Corumbataí (State of São Paulo, Brazil) during 57 years of fire suppression.

RESULTS

Composition and structure of vegetation

Our phytosociological survey sampled 11,507 individuals from 103 different species and 43 families (generalists: 66 spp. and 7,951 individuals; savanna: 27 spp. and 456 individuals; forest: 9 spp. and 3,099 individuals) (^{Table SI} – Supplementary Material). The sum of the IVI species of the same ecological group highlighted their importance for the studied community: generalists (IVI = 223), forest species (49) and savanna (28). The five most numerous and important species were Daphnopsis fasciculata (IVI= 2,514 individuals), Amaioua guianensis (1,452), Copaifera langsdorffii (831), Miconia chartacea (769) and Ocotea pulchella (644). The former species presented greater relative density (21.85) and was classified as a forest species, while the other four species were generalists.

The same species described above represented about 50% of the total number of the sampled individuals, and presented high values for relative frequency, indicating species occurrence in several plots. Other species were also notable for the significant relative frequencies, e.g. Anadenanthera falcata, Myrsine umbellata, Schefflera vinosa, Siparuna guianensis and Virola sebifera. On the other hand, some species were abundant in few plots, e.g. Baccharis dracunculifolia, represented by 44 individuals occurring in four plots, with 40 of them sampled in a single plot of

cerrado sensu stricto. Among the savanna species, Q. grandiflora presented the highest IVI value, which is far below the values presented by the five most important forest and generalist species (Figure 3). These results highlight the level of importance of forest species and generalists when compared to savanna species in the study area.

Figure 3
Phytosociological parameters (relative density - DeR, relative frequency - FeR, relative dominance - DoR, importance value index - IVI) of the five most important species: Daphnopsis fasciculata, Amaioua guianensis, Copaifera langsdorffii, Pera glabrata, Ocotea pulchella. The savanna species with the highest value for IVI (Qualea grandiflora) is also shown for comparison.

More than half of the total sampled species, i.e. 58 species or 56.3% of the total species found in the survey, were represented by populations with up to 10 individuals. Most of the savanna species presented populations with few numbers of individuals, i.e. 21 species or 77.8% of the species belonging to this ecological group sampled, e.g., Aspidosperma tomentosum, Guapira noxia, Pouteria ramiflora and Stryphnodendron polyphyllum. These species accounted for only 7.05 of the total IVI, so that the sum of their population resulted in 83 individuals or 0.72% of the individuals sampled in the 43 plots. The most abundant families were Thymelaeaceae (2,514 individuals), Rubiaceae (1,875), Myrtaceae (1,215), Melastomataceae (1,103) and Fabaceae (873), which accounted for 66% of the total specimens sampled. Thymelaeaceae was more abundant due to the large number of individuals of D. fasciculata.

Height and diameter averages of sampled individuals were 3.8 m (± 2.3 m) and 3.9 cm (± 4.4), respectively. In our survey, only 2.2% of the individuals presented height ≥ 10 m, and 1.3% presented diameter > 14 cm, once we have included woody individuals with a minimum height of 1.5 m in our samples. Most of the sampled individuals was concentrated in the two first classes of height and diameter: 53.3% and 94.62%, respectively (Figure 4). The contribution of individuals belonging to forest and generalist species to the detriment of savanna species was remarkable. These are important indicators of the competitive advantage of fire-sensitive species after four decades of fire suppression. In the study area, generalist and forest species accounted for 95% and 96.6% of the total individuals included in the first class of height and diameter, respectively. The last six classes of height comprise a little less than 4% of the sampled individuals, i.e. few individuals are taller than 10 m. All classes of height presented individuals, while four classes of diameter did not, specifically classes of diameter > 50cm (Figure 4). The generalist species Ocotea acutifolia (13 and 14 m) and C. langsdorffii (14.5 m), and forest species D. fasciculata (13.5 and 14 m) presented the tallest individuals. With many individuals, D. fasciculata also presented the specimens with thick trunks. Forest species Siphoneugena guilfoyleiana (individuals with 37.2, 38 and 39 cm), and the generalist species A. guianensis (39.5 cm) and C. langsdorffii (80.2 cm) also presented the highest values for trunk diameter. These results pointed out the degree of efficiency of this species in occupying the study area. As for the other forest species used as examples, the maximal values for height and diameter of the arboreal Calyptranthes clusiifolia and Rudgea sessilis were 9 m and 12.7 cm, and 7 m and 14.3 cm, respectively.

Figure 4
Number of sampled individuals (ordinates) in different classes of height (in meters) and diameter (in centimeters).

Changes in composition and structure of vegetation over 57 - years period

The sequence of images over a 57-year period showed an efficient expansion of cerradão in the study area after long-term fire suppression (Figure 1). The cerradão encompassed 4.4 ha of the remnant area in 1962 and 37.63 ha in 2019 (Figure 2), representing 97% of the study area The mean annual rates of cerradão expansion from the temporal series studied was 0.53 ha (± 0.4). This expansion, according to the non-linear regression analysis (Figure 5), presented a positive correlation with the total rainfall (R² = 0.42), suggesting the influence of water availability for the cerradão expansion over areas of cerrado sensu stricto when wildfires are absent.

Figure 5
Non-linear correlation between total precipitation (independent variable) and rates of expansion of the cerradão (dependent variable) in nine periods: 2015-2019 (A), 2005-2009 (B), 2010-2014 (C), 1988-1994 (D), 1962-1971 (E), 1995-1999 (F), 1972-1977 (G), 2000-2004 (H), 1978-1987 (I).

DISCUSSION

Composition and structure patterns

Generalist species accounted for 69.1% of the sampled individuals and played an important role in the vegetation cover increase along the period of fire suppression (Pinheiro et al. 2010). During the first stages of ecological succession, the high richness and abundance of generalist species may have facilitated the colonization by forest species. An explanation for the predominance of the generalists in our study is possibly related to the common phenotypic plasticity within species of this ecological group (Johnson et al. 1996), conferring them a certain advantage in the colonization of new areas (Agrawal 2001), and a greater resistance to environmental variation (Gratani 2014).

The predominance of forest individuals, represented by 26.9% of the sampled specimens indicates a positive effect of fire suppression to forest species, e.g. D. fasciculata that accounted for 21.8% of the sampled forest species. On the other hand, fire suppression has negative consequences for fire-prone species, i.e. savanna species. The latter should have occurred in greater number in the study area until the early 1960s (Pinheiro et al. 2010), decreasing over time due to the long-term fire suppression. Indeed, vegetation descriptions of the study area made by Camargo & Arens (1967) and Piccolo et al. (1971) show the dominance of open-savanna environments. The observed successional pattern has similarity with the Relay Floristics principle (Egler 1954), not only for the loss of species groups, but also for the possible influence of invasive forest species during the process (Pulsford et al. 2016).

As expected, we have observed an encroachment of cerradão over areas of cerrado sensu strict. Our observations corroborate with results related to the expansion of forest formations in other areas of Cerrado in Brazil, e.g. central West Brazil (Ratter 1992, Geiger et al. 2011) and in the world, e.g. tropical savannas (Murphy & Bowman 2012, Stevens et al. 2017). An explanation for the forest expansion during the long-term absence of fire in cerrado fragments considers the vegetation itself as a fire barrier during the ecological succession (Silva et al. 2013b, Rossatto & Rigobelo 2016). Nevertheless, we have also to consider that the forest expansion in savanna areas with long-term fire suppression will depend on other factors, e.g. edaphic nutrients, e.g. Mg, Ca and P (Hoffmann et al. 2009, Pinheiro et al. 2016).

When it comes to the reproduction of plants in the study area, the frequencies of height and diameter classes indicate that population renewal is occurring efficiently. Indeed, the number of individuals in the first classes of height (3,549 sampled individuals, or 31% of the total of the sampled individuals) and diameter (9,394 sampled individuals, or 81.6% of the total of the sampled individuals) indicates the great competitive advantage that forest and generalist species have presented during four decades of fire suppression. Both ecological groups accounted together for 95.1% (generalists - 67.7%; forest - 27.4%) of the individuals in the first class of height, while accounted for 96.6% (generalists - 67.1%; forest - 29.5%) of the individuals in the first class of diameter.

Factors influencing forest expansion

The results of our work provided important indications on the influence of rainfall for the expansion of forest formations through savanna areas in the long-term absence of fire. Likewise, we provided relevant information about the negative consequences for savanna vegetation under long-term fire suppression.

The effective long-term fire suppression regime and the edaphic water retention capacity due to higher percentage of silt and clay, favored the expansion of the cerradão over areas of cerrado sensu stricto (Pinheiro et al. 2010). Accordingly, our findings on the structural and floristic data demonstrated an increase of 90.7% (in 2005) of cerradão over cerrado sensu stricto in little more than four decades (Figure 2). Similarly, Pinheiro & Durigan (2009) showed an increase of 38% of cerradão over cerrado sensu stricto in another cerrado fragment in São Paulo state under long-term fire suppression (44 years). Other important indicator of the cerradão’s expansion capacity in the study area is its expansion rate (0.53 ha year^-1), which is faster than the one found by Couto-Santos et al. (2014; 0.13 ha year^-1) in a forest-savanna mosaic in the Amazon region. In the latter study, the authors also found a positive influence between forest expansion and precipitation. The Amazon region presented higher annual average rainfall (1,657 mm - average for 98 years) than the Corumbataí region (1,341 mm - average for 30 years; data from Núcleo de Monitoramento Agroclimático - NURMA).

The fast expansion rate of the cerradão contributed to explain the significant differences between our findings and the first descriptions of the vegetation in the study area (Camargo & Arens 1967, Piccolo et al. 1971). Furthermore, Camargo & Arens (1967) mentioned the presence of some savanna species that we have not reported systematically or opportunistically in our survey, such as Anacardium nanum A.St.-Hil. (Anacardiaaceae), Kielmeyera corymbosa Mart. & Zucc. (Calophyllaceae) and Palicourea rigida Kunth (Rubiaceae). Therefore, the decline of the light intensity due to increased canopy cover along 57 years of fire suppression (Figure 1) may have hindered cerrado sensu stricto species (Pinheiro & Durigan 2009, 2012).

In our study, we demonstrated that the total rainfall have influenced the expansion of cerradão when fire was absent, explaining 42% of the dependent variable. Other potential independent variables are the water retention capacity of the soil (Pinheiro et al. 2010) and the ability of forest species to increase soil nutrient availability during plant succession in the absence of fire (Silva et al. 2013b).

The decrease of light intensity, imposed by the increase of woody cover in the study area, may have interfered in the growth of seedlings and reproduction of heliophile species. Thus, savanna species adapted to less dense savanna physiognomies must have passed through a setback during the cerradão expansion. One of the greatest adversities faced by savanna species in shaded environments is the continuous investment of more resources in the development of the root system when compared to the stalk structure (Franco 2002). As a result, the survival of heliophile species in environments with greater woody cover may be hampered (Durigan & Ratter 2006). Low light intensity compromises species reproduction, including those commonly found in cerradões, e.g. E. gracilipes and V. tucanorum (Barbosa et al. 1999, Ronquim et al. 2003). Thus, the increase of woody cover compromises the establishment of savanna species, which need greater light intensity, and benefits forest and generalist species.

The long-term absence of fire, combined with periods of heavy rain, were essential for the expansion of the cerradão and the decline of savanna-plant populations in the cerrado fragment in Corumbataí. By carrying on the fire suppression regime, the study area may be totally occupied by the cerradão in a few years, with local extinction of savanna species, as a result of the current environmental conditions. Our greatest concern is the negative effects of the long-term fire suppression on the savanna remnants and their biodiversity (Bond & Parr 2010, Veldman et al. 2015b).

ACKNOWLEDGMENTS

We thank our field work team, namely Cecílio de Toledo, the great ‘Zí’ (in memoriam), Murillo Lino Bution and Edson Simão for their valuable assistance. We are very grateful to the professors Antônio Furlan (in memoriam), Marco Antonio de Assis, Renata Giassi Udulutsch for their help in botanical identification. The first author of this article is grateful for the doctoral scholarship provided by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

REFERENCES

• AGRAWAL AA. 2001. Phenotypic plasticity in the interactions and evolution of species. Science 294: 321-326.
• ALVARES CA, STAPE JL, SENTELHAS PC, GONÇALVES JLM & SPAROVEK G. 2014. Köppen’s climate classification map for Brazil. Meteorol Z 22: 711-728.
• ANDERSEN AN, PARR CL, LOWE LM & MÜLLER WJ. 2007. Contrasting fire-related resilience of ecologically dominant ants in tropical savannas of northern Australia. Divers Distrib 13: 438-446.
• ANDERSON JR, HARDY EE, ROACH JT & WITMER RE. 1976. A land-use and cover classification system for use with remote sensor data. Washington: US Geological Survey, 28 p.
• ASSIS ACC, COELHO RM, SILVA PE & DURIGAN G. 2011. Water availability determines physiognomic gradient in an area of low-fertility soils under Cerrado vegetation. Plant Ecol 212: 1135-1147.
• BANFAI DS & BOWMAN DMJS. 2005. Dynamic of a savanna-forest mosaic in the Australian monsoon tropics inferred from stand structures and historical aerial photography. Aust J Bot 53: 185-194.
• BARBOSA AR, YAMAMOTO K & VÁLIO IFM. 1999. Effects of light and temperature on germination and early growth of Vochysia. tucanorum Mart., Vochysiaceae, in cerrado and forest soil under different radiation levels. Rev Bras Bot 22: 275-280.
• BATES DM & WATTS DG. 1988. Nonlinear regression analysis and its applications. New York: J Wiley & Songs, 365 p.
• BITENCOURT MD, MESQUITA JR HN, MANTOVANI W, BATALHA MAPL & PIVELLO VR. 1997. Identificação de fisionomias de cerrado com imagem índice de vegetação. In: Leite LL & Saito CH (Eds), Contribuição ao conhecimento Ecológico do Cerrado – Trabalhos Selecionados do 3º Congresso de Ecologia do Brasil. Brasília: Universidade de Brasília, p. 316-320.
• BOND WJ & PARR CL. 2010. Beyond the forest edge: ecology, diversity and conservation of the grassy biomes. Biol Conserv 143: 2395-2404.
• BOWMAN DMJS, WALSH A & MILNE DJ. 2001. Forest expansion and grassland contraction within a Eucalyptus savanna matrix between 1941 and 1994 at Litchfield National Park in the Australian monsoon tropics. Glob Ecol Biogeogr 10: 535-548.
• BRAUN-BLANQUET J. 1979. Fitosociología: bases para el estudio de las comunidades vegetales. Madrid: H. Blume Ediciones, 820 p.
• BROOK BW & BOWMAN DMJS. 2006. Postcards from the past: charting the landscape-scale conversion of tropical Australian savanna to closed forest during the 20th century. Landsc Ecol 21: 1253-1266.
• BUENO ML, DEXTER KG, PENNINGTON RT, PONTARA V, NEVES DM, RATTER JA & DE OLIVEIRA-FILHO AT. 2018. The environmental triangle of the Cerrado Domain: Ecological factors driving shifts in tree species composition between forests and savannas. J Ecol 106: 2109-2120.
• CÁCERES MD & LEGENDRE P. 2009. Associations between species and groups of sites: indices and statistical inference. Ecology 90: 3566-3574.
• CAMARGO PN & ARENS K. 1967. Observações sobre uma reserva de cerrado. Rev Agricult 42: 3-9.
• CASTRO AAJ, MARTINS FR, TAMASHIRO JY & SHEPHERD GJ. 1999. How rich is the flora of Brazilian cerradão. Ann Mo Bot Gard 86: 192-224.
• COUTINHO LM. 1990. Fire in the ecology of the Brazilian cerrado. In: Goldammer JG (Ed), Fire in the tropical biota. Berlin: Springer-Verlag, p. 81-105.
• COUTO-SANTOS FR, LUIZÃO FJ & CARNEIRO FILHO A. 2014. The influence of the conservation status and changes in the rainfall regime on forest-savanna mosaic dynamics in Northern Brazilian Amazonia. Acta Amaz 44: 197-206.
• CRAWLEY MJ. 2007. Life history and environment. In: Crawley M (Ed), Plant ecology. Malden: Blackwell Science, p. 73-131.
• DAMASCENO-JUNIOR GA & POTT A. 2011. Métodos de Amostragem em estudos fitossociológicos sugeridos para o Pantanal. In: Felfili JM, Eisenlohr, PV, Melo MMRF, Andrade LA & Meira Neto JAA (Eds), Fitossociologia no Brasil: métodos e estudos de casos. Viçosa: Editora UFV, p. 295-323.
• DURIGAN G, BAITELLO JB, FRANCO GADC & SIQUEIRA MF. 2004. Plantas do cerrado paulista: imagens de uma paisagem ameaçada. São Paulo: Instituto Florestal, 475 p.
• DURIGAN G & RATTER JA. 2006. Successional changes in cerrado and cerrado/forest ecotonal vegetation in Western São Paulo state, Brazil, 1962-2000. Edinb J Bot 63: 119-130.
• DURIGAN G, SIQUEIRA MFD & FRANCO GADC. 2007. Threats to the Cerrado remnants of the state of São Paulo, Brazil. Sci Agric 64: 355-363.
• EGLER FE. 1954. Vegetation science concepts I. Initial floristic composition, a factor in old-field vegetation development with 2 figs. Vegetatio 4: 412-417.
• EITEN G. 1972. The cerrado vegetation of Brazil. Bot Rev 38: 201-341.
• FELFILI JM, FELFILI MC, FAGG CW, REZENDE AV, NOGUEIRA PE & SILVA JÚNIOR MC. 2006. Phytogeography of cerrado sensu stricto and land system zoning in central Brazil. In: Pennington RT, Lewis GP & Pennington RT (Eds), Neotropical savannas and seasonally dry forests: plant diversity, biogeography, and conservation. Boca Raton: CRC Press, Taylor & Francis, p. 79-94.
• FRANCO AC. 2002. Ecophysiology of woody plants. In: Oliveira PS & Marquis RJ (Eds), The cerrado of Brazil: ecology and natural history of a Neotropical savanna. New York: Columbia University Press, p. 178-197.
• FURLEY PA. 1999. The nature and diversity of Neotropical savanna vegetation with particular reference to the Brazilian cerrados. Glob Ecol Biogeogr 8: 223-241.
• GEIGER EL, GOTSCH SG, DAMASCO G, HARIDASAN M, FRANCO AC & HOFFMANN WA. 2011. Distinct roles of savanna and forest tree species in regeneration under fire suppression in a Brazilian savanna. J Veg Sci 22: 312-321.
• GOODLAND R & POLLARD R. 1973. The Brazilian cerrado vegetation: a fertility gradient. J Ecol 61: 219-224.
• GRATANI L. 2014. Plant phenotypic plasticity in response to environmental factors. Adv Bot 2014: 1-17.
• GRIME JP. 1979. Plant strategies, vegetation processes, and ecosystem properties. Chichester: J Wiley & Sons, 417 p.
• HERINGER EP, BARROSO GM, RIZZO IA & RIZZINI CT. 1977. A flora do cerrado. In: Ferri MG (Coord), Simpósio sobre o cerrado: bases para a utilização agropecuária. São Paulo: EDUSP, p. 211-232.
• HOFFMANN WA, ADASME R, HARIDASAN M, CARVALHO MT, GEIGER EL, PEREIRA MA, GOTSCH SG & FRANCO AC. 2009. Tree topkill, not mortality, governs the dynamics of savanna–forest boundaries under frequent fire in central Brazil. Ecology 90: 1326-1337.
• HOFFMANN WA, GEIGER EL, GOTSCH SG, ROSSATTO DR, SILVA LCR, LAU OL, HARIDASAN M & FRANCO AC. 2012. Ecological thresholds at the savanna-forest boundary: how plant traits, resources and fire govern the distribution of tropical biomes. Ecol Lett 15: 759-768.
• HUTCHINGS MJ. 2007. The structure of plant populations. In: Crawley MJ (Ed), Plant ecology. Malden: Blackwell Science, p. 325-358.
• IPJBRJ. 2010a. Catálogo de plantas e fungos do Brasil, vol. 1. Rio de Janeiro: Instituto de Pesquisas Jardim Botânico do Rio de Janeiro, 875 p.
• IPJBRJ. 2010b. Catálogo de plantas e fungos do Brasil, vol. 2. Rio de Janeiro: Instituto de Pesquisas Jardim Botânico do Rio de Janeiro, 873 p.
• JOHNSON KH, VOGT KA, CLARK HJ, SCHMITZ OJ & VOGT DJ. 1996. Biodiversity and the productivity and stability of ecosystems. Trends Ecol Evol 11: 372-377.
• LEITÃO FILHO HF. 1992. A flora arbórea dos cerrados do estado de São Paulo. Hoehnea 19: 151-163.
• MARACAHIPES L, CARLUCCI MB, LENZA E, MARIMON BS, MARIMON JR BH, GUIMARÃES FA & CIANCIARUSO MV. 2018. How to live in contrasting habitats? Acquisitive and conservative strategies emerge at inter-and intraspecific levels in savanna and forest woody plants. Perspect Plant Ecol 34: 17-25.
• MAZUCHELI J & ACHCAR J. 2002. Algumas considerações em regressão não linear. Acta Sci 24: 1761-1770.
• MENDONÇA R, FELFILI JM, WALTER BMT, SILVA JR MC, REZENDE AV, FILGUEIRAS TS & NOGUEIRA PE. 1998. Flora vascular do cerrado. In: Sano SM & Almeida SP (Eds), Cerrado: ambiente e flora. Planaltina: EMBRAPA, p. 287-556.
• MEYER KM, WIEGAND K, WARD D & MOUSTAKAS A. 2007. The rhythm of savanna patch dynamics. J Ecol 95: 1306-1315.
• MIRANDA HS, SATO MN, NASCIMENTO NETO W & AIRES FS. 2009. Fires in the cerrado, the Brazilian savanna. In: Cochrane MA (Ed), Tropical fire ecology: climate change, land use, and ecosystem dynamics. Berlin: Springer, p. 427-450.
• MUELLER-DOMBOIS D & ELLENBERGH H. 1974. Aims and methods of vegetation ecology. New York: J Wiley & Sons, 547 p.
• MURPHY BP & BOWMAN DMJS. 2012. What controls the distribution of tropical forest and savanna? Ecol Lett 15: 748-758.
• NEWTON AC. 2007. Forest ecology and conservation: a handbook of techniques. Oxford: Oxford University Press, 454 p.
• ONDEI S, PRIOR LD, VIGILANTE T & BOWMAN DM. 2017. Fire and cattle disturbance affects vegetation structure and rain forest expansion into savanna in the Australian monsoon tropics. J. Biogeogr 44: 2331-2342.
• PARR CL, GRAY EF & BOND WJ. 2012. Cascading biodiversity and functional consequences of a global change–induced biome switch. Divers Distrib 18: 493-503.
• PELLEGRINI AFA, SOCOLAR J, ELSEN PR & GIAM X. 2016. Trade-offs between savanna woody plant diversity and carbon storage in the Brazilian Cerrado. Glob Chang Biol 22: 3373-3382.
• PENNINGTON RT, LEWIS GP & RATTER JA. 2006. An overview of the plant diversity, biogeography and conservation of Neotropical savannas and seasonally dry forests. In: Pennington RT, Lewis GP & Ratter JA (Eds), Neotropical savannas and seasonally dry forests: plant diversity, biogeography, and conservation. Boca Raton: CRC Press, Taylor & Francis, p. 1-29.
• PEREIRA MN, KURKDJIAN MLNO & FORESTI C. 1989. Cobertura e uso da terra através de sensoriamento remoto. São José dos Campos: INPE, 126 p.
• PICCOLO ALG, THAMAZINI LI, MASSA CS, CESAR O, PAGANO SN, MORAES JAPV & AMARAL H. 1971. Aspecto fitossociológico de uma reserva de cerrado. Rev Agricult 46: 81-92.
• PINHEIRO ES & DURIGAN G. 2009. Dinâmica espaço-temporal (1962-2006) das fitofisionomias em unidade de conservação do Cerrado no sudeste do Brasil. Rev Bras Bot 32: 441-454.
• PINHEIRO ES & DURIGAN G. 2012. Diferenças florísticas e estruturais entre fitofisionomias do cerrado em Assis, SP, Brasil. Rev Árvore 36: 181-193.
• PINHEIRO LFS, KOLB RM & ROSSATTO DR. 2016. Changes in irradiance and soil properties explain why typical non-arboreal savanna species disappear under tree encroachment. Aust J Bot 64: 333-341.
• PINHEIRO MHO, AZEVEDO TS & MONTEIRO R. 2010. Spatial-temporal distribution of fire-protected savanna physiognomies in Southeastern Brazil. An Acad Bras Cienc 82: 379-395.
• PUIGDEFABREGASAND J & PUGNAIRE FI. 1999. Plant Survival in Arid Environments. In: Pugnaire FI & Valladares FA (Eds), Handbook of Functional Plant Ecology. New York: Marcel Dekker, Inc., p. 381-405.
• PULSFORD SA, LINDENMAYER DB & DRISCOLL DA. 2016. A succession of theories: purging redundancy from disturbance theory. Biol Rev 91: 148-167.
• RATTER J. 1992. Transition between cerrado and forest vegetation in Brazil. In: Furley P, Proctor J & Ratter JA (Eds), Nature and dynamics of forest-savanna boundaries. London: Chapman & Hall, p. 417-430.
• RATTER JA, BRIDGEWATER S & RIBEIRO JF. 2003. Analysis of the floristic composition of Brazilian cerradão vegetation III: comparison of the woody vegetation of 376 areas. Edinb J Bot 60: 57-109.
• RATTER JA, BRIDGEWATER S & RIBEIRO JF. 2006. Biodiversity patterns of the woody vegetation of the Brazilian cerrado. In: Pennington RT, Lewis GP & Ratter JA (Eds), Neotropical savannas and seasonally dry forests: plant diversity, biogeography, and conservation. Boca Raton: CRC Press, Taylor & Francis, p. 31-66.
• RATTER JA, BRIDGEWATER S, RIBEIRO JF, DIAS TAB & SILVA MR. 2000. Estudo preliminar da distribuição das espécies lenhosas da fitofisionomia cerrado sentido restrito nos estados compreendidos pelo bioma cerrado. B Herb Ezechias Paulo Heringer 5: 5-43.
• RIZZINI CT. 1971. A flora do cerrado: análise florística das savanas centrais. In: Ferri MG (Coord), Simpósio sobre o cerrado. São Paulo: EDUSP, p. 107-153.
• RONQUIM CC, PRADO CHBA & PAULA NF. 2003. Growth and photosynthetic capacity in two woody species of cerrado vegetation under different radiation availability. Braz Arch Biol Technol 46: 243-252.
• ROSSATTO DR & RIGOBELO EC. 2016. Tree encroachment into savannas alters soil microbiological and chemical properties facilitating forest expansion. J Forest Res 27: 1047-1054.
• ROUSE JW, HAAS RH, SCHELL JA & DEERING DW. 1973. Monitoring vegetation systems in the great plains with ERTS. In: Earth resources technology satellite-1 Symposium, 3. Washington: NASA, Goddart Space Flight Center, p. 309-317.
• SHEPHERD GJ. 2009. FITOPAC 2.1 (versão preliminar). Campinas: Departamento de Biologia Vegetal, Universidade Estadual de Campinas, 5 p.
• SILVA DM & BATALHA MA. 2008. Soil-vegetation relationships in cerrados under different fire frequencies. Plant Soil 311: 87-96.
• SILVA DM, BATALHA MA & CIANCIARUSO MV. 2013a. Influence of fire history and soil properties on plant species richness and functional diversity in a Neotropical savanna. Acta Bot Bras 27: 490-497.
• SILVA LC, HOFFMANN WA, ROSSATTO DR, HARIDASAN M, FRANCO AC & HORWATH WR. 2013b. Can savannas become forests? A coupled analysis of nutrient stocks and fire thresholds in central Brazil. Plant Soil 373: 829-842.
• SIQUEIRA MF & DURIGAN G. 2007. Modelagem da distribuição geográfica de espécies lenhosas de cerrado no Estado de São Paulo. Rev Bras Bot 30: 233-243.
• SOLBRIG OT. 1996. The diversity of the savanna ecosystem. In: Solbrig OT, Medina E & Silva JF (Eds), Biodiversity and savanna ecosystem processes: a global perspective. Berlin: Springer, p. 1-27.
• SPIEGEL MR & STEPHENS LJ. 2009. Estatística. Porto Alegre: Artmed, 600 p.
• SPÍNOLA CM, BECHARA FC & BARRETTO KD. 2007. Uso de sensoriamento remoto na identificação de fitofisionomias do cerrado lato sensu. R Bras Bioci 5: 378-380.
• STEVENS N, LEHMANN CER, MURPHY BP & DURIGAN G. 2017. Savanna woody encroachment is widespread across three continents. Glob Chang Biol 23: 235-244.
• TAUK SM & MARCO RA. 1990. Atividade de solo sob cerrado tratado com doses cumulativas de vinhaça, no município de Corumbataí, SP. Naturalia 15: 1-10.
• TROPPMAIR H. 2000. Geossistemas e geossistemas paulistas. Rio Claro: IGCE, UNESP, 127 p.
• VAN DER MAAREL E. 1988. Vegetation dynamics: patterns in time and space. Vegetatio 77: 7-19.
• VAN DER MAAREL E. 2005. Vegetation ecology–an overview. In: Van Der Maarel E (Ed), Vegetation ecology. Malden: Blackwell Publishing, p. 1-51.
• VELDMAN JW ET AL. 2015a. Toward an old-growth concept for grasslands, savannas, and woodlands. Front Ecol Environ 13: 154-162.
• VELDMAN JW, OVERBECK GE, NEGREIROS D, MAHY G, LE STRADIC S, FERNANDES GW, DURIGAN G, BUISSON E, PUTZ FE & BOND WJ. 2015b. Where tree planting and forest expansion are bad for biodiversity and ecosystem services. BioScience 65: 1011-1018.
• VILLARD MA & METZGER JP. 2014. Beyond the fragmentation debate: a conceptual model to predict when habitat configuration really matters. J Appl Ecol 51: 309-318.
• WIEGAND K, SALTZ D & WARD D. 2006. A patch-dynamics approach to savanna dynamics and woody plant encroachment–insights from an arid savanna. Perspect Plant Ecol 7: 229-242.
• WRIGHT SJ. 2007. Plant Diversity in Tropical Forests. In: Pugnaire FI & Valladares F (Eds), Functional Plant Ecology. Boca Raton: CRC Press, Taylor & Francis Group, p. 351-367.

SUPPLEMENTARY MATERIAL

Table SI. Phytosociological parameters of the species sampled in the savannic community of Corumbataí (SP - Brazil). Notations: Individual number (IN); relative density (DeR); relative frequency (FeR); relative dominance (DoR); importance value index (IVI); ecological groups (EG); forest species (F); generalist species, (G); savanna species (S).

Publication Dates

• Publication in this collection
  16 Apr 2021
• Date of issue
  2021

History

• Received
  16 Nov 2019
• Accepted
  20 Apr 2020