Palynological analysis of a late Holocene core from Santo Antônio da Patrulha, Rio Grande do Sul, Southern Brazil

EARTH SCIENCES

Palynological analysis of a late Holocene core from Santo Antônio da Patrulha, Rio Grande do Sul, Southern Brazil

Renato B. Macedo^I; Paulo A. Souza^I; Soraia G. Bauermann^II; Sérgio A.L. Bordignon^II

^ILaboratório de Palinologia, Departamento de Paleontologia e Estratigrafia, Instituto de Geociências Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, 91540-900 Porto Alegre, RS, Brasil

^IILaboratório de Palinologia, Universidade Luterana do Brasil, Av. Farroupilha, 8001, 92425-900 Canoas, RS, Brasil

^{Correspondence to}

ABSTRACT

A sedimentar core collected at Santo Antônio da Patrulha, Rio Grande do Sul State, southmost Brazil, was submitted to pollen analysis to provide the vegetational history of this region, and the paleoecological and paleoclimatic changes. A total of 98 taxa of palynomorphs was identified from 35 subsamples. Three radiocarbonic datings were obtained along a section of 115 cm depth, including the basal age of 4730 ± 50 yr BP. Pollen diagrams and cluster analysis were performed based on palynomorphs frequencies, demonstrating five distinct phases (SAP-I to SAP-V), which reflected different paleoecological conditions. The predominance of plants associated with grasslands in the phase SAP-I suggests warm and dry climate conditions. A gradual increasing of humidity conditions was observed mainly from the beginning of the phase SAP-III, when the vegetation set a mosaic of grasslands and Atlantic rainforest. Furthermore, the presence of some forest taxa ( Acacia-type, Daphnopsis racemosa, Erythrina-type and Parapiptadenia rigida-type), from the phase SAP-IV, is interpreted as an influence of the seasonal semideciduous forest in the study region. From the phase SAP-V (ca. 4000 yrs BP), the vegetation became similar to the modern one (extant Atlantic rainforest Biome), especially after 2000 yrs BP (calibrated age).

Key words: Palynology, Paleoecology, Paleoclimatology, Atlantic rainforest, late Holocene, Rio Grande do Sul State.

RESUMO

Um testemunho de sondagem coletado em Santo Antônio da Patrulha, Rio Grande do Sul, Brasil, foi submetido para análise polínica a fim de revelar a história vegetacional e mudanças paleoecológicas e paleoclimáticas. Um total de 98 táxons foi identificado a partir de 35 subamostras. Três datações radiocarbônicas foram obtidas ao longo de uma seção de 115 cm de profundidade, incluindo a idade basal de 4730 ± 50 anos AP. Diagramas polínicos e análises de agrupamentos foramrealizadas com base nas freqüências dos palinomorfos, demonstrando cinco fases distintas (SAP-I a SAP-V), as quais refletiram diferentes condições paleoecológicas. A predominância de plantas relacionadas à vegetação campestre na fase SAP-I sugere condições climáticas quentes e secas. Um gradual aumento nas condições de umidade foi observado principalmente no início da fase SAP-III, quando a vegetação conformou um mosaico de Campos e Floresta Atlântica. Além disso, a presença de certos táxons florestais (tipo- Acacia, Daphnopsis racemosa, tipo Erythrina e tipo Parapiptadenia rigida), a partir da fase SAP-IV, é interpretada como influência da Floresta Estacional Semidecidual na região de estudo. A partir da fase SAP-V (ca. 4000 anos AP) a vegetação tornou-se similar à moderna (atual Bioma da Floresta Atlântica), especialmente após 2000 anos AP (idade calibrada).

Palavras-chave: Palinologia, Paleoecologia, Paleoclimatologia, Floresta Atlântica, Holoceno tardio, Rio Grande do Sul.

INTRODUCTION

The vegetation of Rio Grande do Sul State (RS), Southern Brazil, is composed by grassland-forest mosaic as a result of paleoenvironmental changes, mainly characterized by climate variations during the Quaternary(Marchiori 2004). These grassland-forest mosaics were previously studied by pioneer naturalists at the end of the XIX century. Lindman (1906) observed that forests could expand over the grassland vegetation under drypaleoclimate conditions. Based on phytogeographicevidence, Rambo (1956, 1961) and Klein (1975) interpreted that the grasslands were the first plants constituents of the RS; therefore, advances of the forests is a recent answer to the humid climatic conditions.

The palynology is an important tool for the understanding of vegetation dynamics and paleoclimate reconstructions. Several studies based on pollen analysis have contributed in solving questions on the vegetacional succession during the late Quaternary in Southern Brazil, as summarized by Behling (2002), Lorscheitter (2003), De Oliveira et al. (2005) and Leal and Lorscheitter (2007). Palynological results showed that, during the latest Pleistocene, the landscape in these regions was characterized by the dominance of grasslands due to the cold and dry climate associated with glacial times. During the early to middle Holocene, the vegetation related to these grasslands prevailed in lowlands and highlands of the Southern region due to warm and dry climatic condictions after the glacial period. Pollen analysis of the late Holocene in Rio Grande do Sul Coastal Plain (RSCP) and adjacent areas showed that forests had expanded from 3500 yrs BP, especially after 2000 yrs BP, due to higher humidity.

This paper presents interpretations about the vegetational succession of the Atlantic rainforest and seasonal semideciduous forest, as well as paleoclimates related to them, based on palynological data obtained from a core in Santo Antônio da Patrulha, RS, southmost Brazil. The results were compared with other palynological studies previously conducted in South Brazil.

STUDY AREA

The present study was made from a peat bog located in Santo Antônio da Patrulha municipality (coordinates 29º44'45"S, 50º32'56"W, high 37 m), far about 76 km from Porto Alegre and ca. 48 km from the AtlanticOcean. The access to this site from Porto Alegre isthrough the highway RS-474, followed by a secondary road (Figs. 1a-b). According to Fortes (1959), this area is part of the physiographic region of the Lower Northeast Slopes of Serra Geral, RS.

The peat bog sediments from Santo Antônio daPatrulha were deposited on sandstones of the Botucatu Formation, Jurassic/Cretaceous of Paraná Basin, which constitutes the basement of the analyzed core.

The presence of a grassland-forest mosaic characterizes the phytophisiognomy of the landscape. The forest comprises a mixture of floristic elements of Atlantic rainforest and seasonal semideciduous forest (Rambo 1956, Teixeira et al. 1986, Reitz et al. 1988, Leite and Klein 1990).

The climate of Southern Brazil is influenced bythe South Atlantic Anticyclone, a semi-permanent high pressure system that transports moist tropical air masses over the continent from easterly and north-easterly directions during the whole year. Disturbances are related to polar cold fronts, when it meets the tropical air masses and produces strong rainfall in Southern Brazil (Nimer 1989). This region is characterized by subtropical humid ( Cfa) in the Koppen classification (Moreno 1961), with regularly distributed rainfalls during the year and hot summers. The mean precipitation is around 1676.5 mm.a, and the average annual temperature is 19.8C; the average temperature of the hottest month is 24.4C, while the coolest months are 15.4C (data from Osório and Tramandaí meteorological station, IPAGRO 1979).

MATERIALS AND METHODS

The core was taken from the deepest portion near the center of the peat bog using a Russian corer sampler, whose maximum depth reached 115 cm. Sections of50 cm length were extruded in situ, wrapped in plastic film and aluminum foils. The section was transported to the Laboratory and stored in special conditions (ca. +4C) before sampling. Three sediment subsamples of 3 cm thickness were taken from the core and dated by the Accelerator Mass Spectrometry (AMS) in the CAIS Laboratory of the University of Georgia (USA). The calibration of radiocarbon datings was carried out afterCALPAL (Weninger et al. 2004). Ages were also calculated for each interpolated pollen subsample and pollen phase.

For pollen and charcoal analysis, 35 subsamples (1 cm3 volume) were taken at 3 cm intervals along the 115 cm cored. All subsamples were processed by standard pollen analytical methods (Faegri and Iversen1989), using HF, HCl, KOH, acetolysis, followed by filtering through a 250 m net. The slides were prepared in glycerol-jelly. Pollen preparation included the addition of exotic Lycopodium clavatum L. spores to determine pollen concentration (grains/cm3) and accumulation rates (grains/cm2/year), in agreement with Stockmarr (1971).

A minimum of 300 pollen grains were counted for each subsample. This pollen sum includes trees, shrubs and herbs. Aquatic, ferns and mosses spores, algae taxa, fungal spores and animal remains, as well as carbonized particles (5-150 m), were also counted, constituting a separated list expressed as the percentage of the total pollen sum. Calculations of concentration were also made for the carbonized particles (particles/cm3), as well as for accumulation rates (particles/cm2/year). Pollen identification was based on catalogues of published palynomorphs, as well as on pollen reference colections of Brazil Lutheran University (ULBRA). The word "type" was used when an accurate identification was impossible. Taxonomic descriptions and illustrations of the identified palynomorphs was an additional contribution (Macedo et al. 2009) and constitute the basis for the palynological analysis. Softwares TILIA, TGVIEW and CONISS version 2.0.2. (Grimm 1987) were used for plotting the pollen data, calculations and cluster analysis. The pollen diagrams include individual records of the taxa and records of the groups: grasslands, forest, indeterminate, aquatics, ferns, mosses, algae, fungi and concentration and influx of the pollen and charcoal particles, and a cluster analysis. The phases of the pollen record ( sensu Salgado-Labouriau 2007 p. 335-336) were established from changes in the pollen assemblages and from cluster analysis.

Ordination by Principal Coordinates Analysis (PCoA) was used for a synthetic view of community compositional changes (Orlóci et al. 2002). PCoA was based on pairwise Euclidean distances among subsamples (Legendre and Legendre 1998, Podani 2000). Distances were computed from squares root transforming pollen percentages in order to reduce the excessiveweight of dominant pollen taxa. Algae, spores, indeterminate palynomorphs and Cyperaceae taxa were not included on multivariate analysis, as well as subsamples that did not present pollen grains (phase SAP-II). The significance of ordination axes was evaluated by bootstrap according to Pillar (1999). The multivariate analysis PCoA was calculated using software MULTIV(Pillar 2008).

In order to assist the interpretation regarding the vegetation succession, a rapid survey of the flora was made near the point of collection of the core.

RESULTS

LITHOLOGY

The core retrieves a section of 115 cm of unconsolited sediments. Between 115 to 90 cm depth, sediments are mainly characterized by brown sandstones with decomposed organic matter. From 90 to 50 cm depth, sediments comprise dark mudstones/siltstones with organic material completely decomposed, including few root remains. Between 50 to 13 cm, sediments are constituted by brown-dark sandstones and a mixture of decomposed organic matter with abundant roots and plant remains. From 13 cm depth to the top of the core, the sediments are highy weathered, constituting the regolith, which was not palynologically analyzed.

RADIOCARBON DATING

The results of the AMS radiocarbon datings are presented in Table I. Based on these data, we consideredthe deposition of the studied core entirely positionedalong the late Holocene. However, sedimentation was not continuous. Three datings were obtained. The basal dating corresponds to 4730 50 14C yr BP (calibrated age 5461 yr BP), while the sample at 55 cm depth was dated as 4225 25 14C yr BP. It suggests a rapid rate of sedimentation between this interval (115 to 55 cm depth). The sample at 13 cm core depth indicates modern age (Anno Domini 1850). The results between 55 and 13 cm core depth suggest a slow sedimentation, envolving the vegetation history of the last 4200 yrs BP.

DESCRIPTION OF POLLEN RECORD

A total of 98 taxa of palynomorphs were retrieved and identified along the Santo Antônio da Patrulha core, including taxa of algae, Briophyte, Pteridophytes, Gymnosperms, Angiosperms, as well as fungi and animal remains. The pollen diagram (Figs. 2a-b-^c) shows the distribution of the palynomorphs, which were grouped according to their ecological affinities (habit and/or habitat), beyond pollen sum of different botanical groups (Fig. 3). Marked changes in the pollen assemblages, indicated by cluster analysis (CONISS), allow the establishment of five pollen phases: SAP-I, SAP- II, SAP-III, SAP-IV and SAP-V, which are described below (SAP is the abbreviation for Santo Antônio da Patrulha area) in an ascending stratigraphic order.

Phase SAP-I (5461-5443 cal yr BP)

This phase was characterized from the two basal subsamples (115-110 cm of depth), which are dominated by Cyperaceae (40-55%), Poaceae (46-28%), Baccharis-type (6-9%), Eryngium L., Spermacoce L. and other minor herbaceous components (Fig. 2b-3). The group of trees and shrubs presents lower percentages (1-4%), comprising mainly Ilex L. and Arecaceae. Other taxa are less than 1% (with Melastomataceae, Myrtaceae and Mimosa serie Lepidotae Benth., Fig. 2a). Percentages of pollen from aquatic plants ( Ludwigia L. and Utricularia L.) are also less than 1%. Ferns spores are scarce. Mosses spores are abundant in higher samples (27%), which are mainly represented by Phaeoceros laevis (L.) Prosk. (^{Fig. 2c}). Algae taxa show lower proportions. Concentration and accumulation rates of the charcoal particles are relatively high in this phase (Fig. 3).

Phase SAP-II (5443-5402 cal yr BP)

The interval at 110-105 cm depth (two subsamples) is characterized by the non-preservation of pollen grains. Few ferns and mosses spores, as well as algae taxa, were poorly preserved.

Phase SAP-III (5402-4955 cal yr BP)

This phase comprises 12 subsamples within the interval between 105-69 cm core depth. It is marked bythe abundance of herb pollens (52-95%), which are represented by Poaceae (31-53%), Cyperaceae (5-40%) and Baccharis-type (5-35%), as well as other less frequent taxa, like Eryngium, Cuphea carunculata Koehne, Cuphea Koehne, Vernonia-type, Spermacoce, Gnaphalium-type, Polygala L., Senecio-type, Holocheilus-type, Amaranthus/Chenopodiaceae-type and Elephantopus-type.Sums of forest taxa are better represented (between4-48%) due to the increasing of Ilex, which reaches up to 28%. Other forest taxa, such as Arecaceae, Myrtaceae, Melastomataceae, Alchornea Sw., Celtis L., Trema micrantha (L.) Blume., Anacardiaceae and Podocarpus L' Hér. ex Pers., also increased their proportions during this phase. Pollen grains from Araucaria angustifolia (Bertol.) Kuntze and Drymis brasiliensis Miers. occur respectively at 79-76 cm depth (Fig. 2a). Fern spores also increase in this phase, mainly Blechnum-type, Microgramma vacciniifolia-type and PolypodiumL. Spores from Dicksonia sellowiana Hook were found at 97 cm depth (^{Fig. 2c}). Moss spores from Phaeoceros laevis decrease during this phase. The concentration and accumulation rates of the charcoal particles and pollen grains were the highest ones in the core.

Phase SAP-IV (4955-4402 cal yr BP)

This phase comprises six subsamples, between 69-51 cm depth. In this phase, taxa related to grasslands continue to be dominant (91-93%). Pollen sum of forest taxa decreases (6-12%). However, at the beginning of this phase, there are the first records of Chrysophyllum L., Ephedra tweediana C.A. Meyer, Acacia-type, Parapiptadenia rigida-type, Erythrina-type, Daphnopsis racemosa and Sebastiania Spreng. Pollen fromaquatics plants, such as Ludwigia, is less than 1%. Ferns spores decrease in this phase. Spores of mosses reached the highest percentages (16-74%), which are mainlydominated by Phaeoceros laevis (12-72%), while Sphagnum (Dill.) Hedw and Anthoceros punctatus L. occur in lower proportions. Algae taxa were mainly represented by Pseudoschizaea rubina Rossignol ex Christopher (17%). The concentration and accumulation rates of the charcoal particles and pollen grains were lower than in the previous phase.

Phase SAP-V (4402-137 cal yr BP)

The highest stratigraphic phase comprises 13 subsamples within the interval between 51-13 cm depth. This phase is characterized by a significant increasing of pollen sum of forest taxa (21-37%) and by the decreasing of herb pollens (62-78%). In general, all taxa related to forest were better represented in this phase than in the previous one. On the other hand, Podocarpus decreases in frequency. In the transition between phases SAP-IV and SAP-V, there is the first record of Zea mays L. and Pinus L., which were found in mostly subsamples of this last phase. Aquatic taxa were also better represented in this phase, mainly by Ludwigia, which reached up to 12%. Spores of ferns increased their percentages (6-31%), while spores of mosses and algae taxa decreased them. Spores of fungi also increased ( Glomus Tus. & C. Tus., Gelasinospora calospora (Mouton) C. Moreau & M. Moreau and Gelasinospora reticulispora (Greis & Greis-Dengler) C. Moreau & M. Moreau, Gaeumannomyces cf. caricis-type, Athelia-type and Tetraploa aristata Berk. & Br.). The concentration of the charcoal particles and pollen grains were higher than in the previous phase, while accumulation rates of the charcoal particles and pollen grains were lower.

MULTIVARIATE ANALYSIS

The multivariate analysis (PCoA) shows a synthesis of the changes in the pollen composition since 5461 cal yr BP until present 137 cal yr BP (Anno Domini 1850). The two dimensional ordination diagram accounts for 61% of the total variation in the data set, with 53 taxa in 33 subsamples (Figs. 4a-b). Pollen composition dynamics depicted by the ordination analysis shows phases of random, non-directional changes and periods in which the changes were directional toward more abundant forest taxa. Directional jumps, characterizing transitions of phases, were most evident from 5386 to 5314 cal yr BP, 5206 to 5189 cal yr BP, and 4402 to 3991 cal yr BP. Multivariate analysis of phase SAP-I shows the dominance of grasslands up to 5386 cal yr BP, which are represented by Poaceae and Baccharis-type. At ca.5314 cal yr BP (phase SAP-III), ordination diagramssuggest important vegetational changes by the appearance of forest taxa, such as Myrtaceae, Ilex, Arecaceae and Podocarpus. After 5058 cal yr BP, within the SAP-III/IV transition, grassland taxa diversified, including Holocheilus-type, Trixis-type, Spermacoce, Senecio-type, Vernonia-type, Lamiaceae, Gomphrena L., Eryngium, Melastomataceae and Cuphea. The phase SAP-V is evidenced in the PCoA diagrams by the increasing and diversity of trees and shrubs components, such as Anacardiaceae, Sebastiania, Chrysophyllum, Alchornea, Trema micrantha, Myrsine L., Daphnopsis racemosa, Mimosa L., Celtis and Erytrina-type.

BOTANICAL DATA

The main components of the modern flora around the peat bog are listed in the Figure 5, according to their type, physiognomy, habit and habitat, and including the taxonomic basis for genera and species.

DISCUSSION

The results obtained from the pollen analysis of Santo Antônio da Patrulha core revealed changes in the vegetational composition during the last millennia on the Lower Slopes of Serra Geral, RS. These vegetational successions reflected paleoenvironmental variations, which are described below.

The Phase SAP-I (5461-5443 cal yr BP) showed the predominance of plants associated with grasslands, mainly represented by Poaceae and Baccharis-type (Figs. 2b-3). Furthermore, few records of trees and shrubs, ferns spores and algae taxa, as well as high frequency of mosses spores (Phaeoceros laevis), suggest warm and dry climate conditions, in agreement with the interpretations of Behling (2002) to South and Southeast regions of Brazil. According to this author, low precipitations and a long annual dry season, probablyof three months, should have limited the expansion of trees taxa during the middle Holocene in SouthernBrazil. Perhaps these paleoclimates conditions are related to changes in the trajectory of Polar cold fronts to the Atlantic Ocean, similar to modern events of La Niña, which causes drought in Southern Brazil. Other palynological studies performed in RSCP and adjacent areas also showed similar results (Bauermann 2003, Macedo et al. 2007, Leal and Lorscheitter 2007).

Poor preserved palynomorphs occur in Phase SAP-II (5443-5402 cal yr BP), in low frequency, preventing vegetational and paleoclimatic interpretations to this interval. These results can be related to taphonomic limitations, though no sedimentological and lithological different evidences are observed. This limitation was also recorded by Bauermann (2003) from the interval between ca. 6000 cal yr BP to 3163 29 14C yr BP within a core recovered from Barrocadas, RSCP.

The Phase SAP-III (5402-4955 cal yr BP) revealed pollen changes in the grasslands vegetation clearly highlighted by the multivariate analysis of PCoA, taking into account the gradual increasing of Baccharis-type, Poaceae and other herbaceous components, such as Cuphea, Eryngium, Gomphrena, Holocheilus-type, Lamiaceae, Melastomataceae, Polygala, Senecio-type, Spermacoce, Trixis-type and Vernonia-type (Figs. 4a-b). This phase is also marked by the increase of pioneers taxa of the Atlantic rainforest, such as Alchornea, Anacardiaceae, Arecaceae, Cecropia Loefl., Celtis, Chrysophyllum, Ilex, Melastomataceae, Myrtaceae, Myrsine, Podocarpus, Sebastiania and Trema micrantha (Figs. 2a-3). Thesepollen records suggest the beginning of the grassland-forest mosaics. These Atlantic rainforest components could possibly be distributed in areas of higher humidity, such as borders of rivers, forming the gallery forest. Furthermore, they could have been dispersed in small populations under waterlogged soil, forming swamp forests. In this interval, the initial and gradual appearance of certain taxa of the Araucaria forest (Araucaria angustifolia and Drymis brasiliensis) was observed, as well as some ferns spores ( Dicksonia sellowiana) developed on valleys and areas associated with the Serra Geral Plateau (Behling et al. 2004, Leonhardt and Lorscheitter 2010). These components are no more observed in the study area. We interpret that they were transported and subsequently incorporated within the analyzed sediments.

The beginning of the expansion of the Atlanticrainforest reflects changes to climatic conditions with higher temperature and humidity. High frequencies in records of ferns spores (Blechnum-type, Cyatheaceae, Microgramma vacciniifolia-type and Polypodium), algae ( Zygnema C.A. Agardh) and fungi spores corroborate this humidity conditions (^{Figs. 2c}-3). According to Martin et al. (1993), the constant rainfalls during the last 5000 yrs BP is probably due to El Niño events,which caused strong storms in South Brazil.

The increasing in temperature and humidity also coincides with the last maximum transgressive marine event marked in the lowlands RSCP at ca. 5000 yrs BP (Villwock et al. 1986, Villwock and Tomazelli 1995).

These paleoenvironmental interpretations are also compatible with the records of human occupation associated with the Umbu tradition (pre-ceramic period) observed by Dias (2003) and Dias and Jacobus (2003) in the arqueologic site "RS-S-327" located about 2 km of this study area. In addition, mammalian fossils of small size, related to grassland vegetation ( Thylamys Grey, Clyomys sp. nov. and Dicolpomys fossor Winge) and others typical of forests (Gracilinanus microtarsus Wagner and Phyllomys sp.), were found in the archaeological site above mentioned at stratigraphical levels corresponding to this phase. It confirms the grassland-forest mosaics hypothesis (Rodrigues 2008) to this region.

An important theme to be reanalyzed concerns the rates of concentration and accumulation of carbonized particles, which are the highest in this phase. It was expected that, by the increasing of humidity conditions as reflex of the entrance of Atlantic rainforest elements, natural events of fire could be less expressive. Human origin to these highest rates of carbonized particles is discarded. In those times, arqueological record of fire seem to be restricted to rock shelter, related to hunt of human Umbu tradition (Dias 2003). These carbonized particles are relatively well preserved, demonstrating that this site of deposition received most contributions from the nearest area, with a strong influence of organic matter. It is reinforced by the lithological data and by the highest sedimentation rate of the core (Fig. 3).

The Phase SAP-IV (4955-4402 cal yr BP) is very similar to the previous one, with a well-representedgrasslands vegetation in the studied region. However,pioneer components of the Atlantic rainforest (Arecaceae, Ilex, Myrtaceae, Myrsine and Celtis) showed a slightly reduction in their frequencies, as also verifiedby other authors (Lorscheitter 2003, De Oliveira et al. 2005). This reduction is attributed to the marine transgression above mentioned (ca. 5000 yrs BP) that controlled the Atlantic rainforest expansion, mainly at the Rio Grande do Sul Northern Coastal Plain with indirect reflexes in the studied area.

The first pollen records of Acacia-type, Daphnopsis racemosa Griseb., Erythrina-type, and Parapiptadenia rigida-type are interpreted as seasonal semideciduous forest influence. In addition, probably Ephedra tweediana moved from west to east. According to Rambo (1954), this species migrated from the Andes regions and subsequently penetrated in the South of RS from Argentina. Currently E. tweediana is restricted in the RSCP, ca. 30 latitude.

The uppermost Phase SAP-V involves the last4402 cal yr BP, showing a significant expansion of theforest vegetation and constituting the grassland-forest mosaics similar to the modern ones. The increasingof biodiversity of the Atlantic rainforest and seasonal semideciduous forest in the region was synchronous to the marine regression after 4000 yrs BP in RSCP. After this event, processes of desalinization in soils allowed that certain elements replaced the RSCP region and adjacent areas, according to Lorscheitter (2003), DeOliveira et al. (2005) and Leal and Lorscheitter (2007). High percentages of algae taxa (Pseudoschizaea rubina Rossignol ex Christopher, Zygnema) and mosses spores (Sphagnum Dill. Hedw) suggest the permanence ofthese humidity conditions and intense rainfall in this phase. Ferns spores such as Blechnum-type, Cyatheaceae and Selaginella became well represented even as aquatic elements (mainly Ludwigia). Spores of fungi mycorrhizal increased in frequency and can be associated with the development of soils. The expansion of the Atlantic rainforest and seasonal semideciduous forest may have minimized the fire frequency, which is also evidenced by the reduction of rates of concentration and accumulation of carbonized particles. According to Pillar (2003) and Overbeck et al. (2005), fire in the grasslands vegetation did not reach the inner part of forest due to the lack of biomass to promote flammability. Finally, the uppermost subsamples (from the 0, 35 m to top) of the core showed the human influence by the marked presence of Zea mays L.

The dating performed at 55 cm depth reveals an interesting and exciting aspect about the geological evolution of the analyzed sedimentary interval. According to the obtained datings, the section between 115 and55 cm had a high sedimentar rate, of about 0,124 cm/ year, including a time of ca. 485 years. Otherwise, for the interval between 55 and 13 cm depth, the sedimentation rate is much lower, of ca. 0,01 cm/year. Considering the dating accuracy and the granulometry sustenance, this difference can be explained by changes in the shape of paleotopography. At the beginning, the point of sampling probably corresponded to a more irregular surface on the ground, such as minor depressions or a cavity. The upper section was deposited on a more regular and planar surface, with a lower sedimentation rate. Furthermore, the presence of Pinus at the base of Phase V (ca. 50 cm of depth) corresponds to a younger and modern feature concerning the vegetational occupation. This basal record suggests a variability in the sedimentar pattern, with the possibility of having hiatuses from the base, which were not revealed by the lithological descriptions.

CONCLUDING REMARKS

The palynological record obtained from a peat bog at Santo Antônio da Patrulha, RS, southmost Brazil, revelead five phases, named SAP-I to SAP-V. These phases reflect different ecological and environmental conditions, mainly related to the vegetational succession and the paleoclimate variations, as summarized below. The results reinforced the use of pollen and spores assemblages for Quaternary analysis.

(i)A total of 98 taxa of palynomorphs were identified from 35 subsamples recovered from the Santo Antônio da Patrulha core, whose basal radiocarbonic dating (at 115 cm of depth) revealed an age of 4730 50 yr BP (calibrated age 5461 yrs BP).

(ii)Among the five identified phases, only the SAP-II (two subsamples 110 to 105 cm core depth) did not revealed an amount of pollen and spores for paleoecological and paleoclimatic interpretations.

(iii)Taking into account the main tendencies of the vegetational record, it is possible to observe a gradual increasing of humidity, mainly from the beginning of the SAP-III, when the pioneers of the Atlantic rainforest components, fern spores and fungi, became significant. Furthermore, grasslands present more diversity of taxa.

(iv) Grassland components are dominant along the entire core. However from the SAP-V (ca. 4000 cal yrs BP) forest taxa, Atlantic rainforest and seasonal semideciduous forest, present a gradual and significant increasing.

(v) The records and frequency of grasslands and forest taxa from the SAP-III show the beginning of the vegetation as a mosaic, as interpreted by other authors. Mammalian fossils and arqueological data support this idea, especially correlated to the SAP-III.

(vi) The progressive establishment of these forest elements in the region suggests warm condition from ca. 4000 cal yrs BP to present, corroborating previous palynological data obtained in the RSCP and following the marine regression during the latest Holocene times.

(vii) The presence of some pollen taxa, such as Acacia-type, Daphnopsis racemosa, Erythrina-type and Parapiptadenia rigida-type, is interpreted herein as an influence of the seasonal semideciduous forest. Other authors recorded these taxa from late Holocene cores in southmost Brazil, but they seem to be included as Atlantic rainforest components.

ACKNOWLEDGMENTS

Radiocarbonic datings were supported in part by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) through projects conducted by Adriana Schmidt Dias (474630/2004-8) and Rodrigo Rodrigues Cancelli (140924/2008-6). The authors thank to Valério De Patta Pillar for discussions and revision of the multivariate analysis, and to the referees for theirimportant suggestions. This paper is part of the Master's dissertation of the first author, with grants from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

Manuscript received on June 19, 2009; accepted for publication on December 8, 2009

• BAUERMANN SG. 2003. Análises palinológicas e evolução paleovegetacional e paleoambiental das turfeiras de Barrocadas e Águas Claras, Planície Costeira do Rio Grande do Sul, Brasil. Tese de Doutorado, Instituto de Geociências, Universidade Federal do Rio Grande do Sul, RS, 137 p.
• BEHLING H. 2002. South and southeast Brazilian grasslands during Late Quaternary times: a synthesis. Palaeogeog Palaeoclim Palaeoecol 177: 19-27.
• BEHLING H, PILLAR VD AND BAUERMANN SG. 2004. Late Quaternary Araucaria forest, grassland (Campos), fire and climate dynamics, inferred from a high-resolution pollen record of Cambará do Sul in southern Brazil. Palaeogeog Palaeoclim Palaeoecol 203: 277-297.
• DE OLIVEIRA ET AL. 2005. Paleovegetação e paleoclimas do Quaternário do Brasil. In: SOUZA CRG ET AL. (Eds), Quaternário do Brasil, Ribeirão Preto: Holos, p. 52-74.
• DIAS AS. 2003. Sistemas de assentamento e estilo tecnológico: uma proposta interpretativa para a ocupação pré-colonial do alto vale do rio dos Sinos, Rio Grande do Sul. Tese de Doutorado, Instituto interdepartamental em Arqueologia, Universidade Federal de São Paulo, SP, 399 p.
• DIAS AS AND JACOBUS A. 2003. Quão antigo é o povoamento do Sul do Brasil? Cepa 27: 39-67.
• FAEGRI K AND IVERSEN J. 1989. Textbook of Pollen Analysis. 4th ed., J Wiley & Sons, New York, 328 p.
• FORTES AB. 1959. Geografia física do Rio Grande do Sul. Porto Alegre: Globo, 393 p.
• GRIMM EC. 1987. CONISS: A Fortran 77 program for stratigraphically constrained cluster analysis by the method of the incremental sum of squares. Comput Geosci 13: 13-35. Available in:<http://demeter.museum.state.il.us/pub/grimm>
• IPAGRO - INSTITUTO DE PESQUISAS AGRONÔMICAS.1979. Observações metereológicas no Estado do RioGrande do Sul. Porto Alegre: Secretaria da agricultura do Rio Grande do Sul, (Boletim Técnico) 3: 271.
• KLEIN RM. 1975. Southern Brazilian phytogeographic features and the probable influence of upper Quaternary climatic changes in the floristic distribution. Bol Paranaense Geocienc 33: 67-88.
• LEAL MG AND LORSCHEITTER ML. 2007. Plant sucession in a forest on the Lower Northeast Slopes of Serra Geral, Rio Grande do Sul, and Holocene palaeoenvironments, Southern Brazil. Acta Bot Bras 21: 1-10.
• LEGENDRE L AND LEGENDRE P. 1998. Numerical Ecology. 2nd ed., New York: Elsevier, 853 p.
• LEITE PF AND KLEIN RM. 1990. Vegetação. In: MESQUITA OV (Ed), Geografia do Brasil - Região Sul, Rio de Janeiro: IBGE 2: 113-150.
• LEONHARDT A AND LORSCHEITTER ML. 2010. The last 25.000 years in the Eastern Plateau of Southern Brazil according to Alpes de São Francisco record. J South Am Earth Sci 29: 454-463.
• LINDMAN CAM. 1906. A vegetação no Rio Grande do Sul. Porto Alegre: Universal, 356 p.
• LORSCHEITTER ML. 2003. Contribution to the Holocene history of Atlantic rainforest in the Rio Grande do Sul State, southern Brazil. Rev Mus Argent Cienc Nat 5(2): 261-271.
• MACEDO RB, CANCELLI RR, BAUERMANN SG, BORDIGNON SA DE L AND NEVES PCP DAS. 2007. Palinologia de níveis do Holoceno da Planície Costeira do Rio Grande do Sul (localidade de Passinhos), Brasil. Gaea 7: 68-74.
• MACEDO RB, SOUZA PA AND BAUERMANN SG. 2009. Catálogo de pólens, esporos e demais palinomorfos em sedimentos holocênicos em Santo Antônio da Patrulha, Rio Grande do Sul, Brasil. Iheringia 62(2): 43-78.
• MARCHIORI JNC. 2004. Fitogeografia do Rio Grande doSul, Campos Sulinos. Porto Alegre: EST, 110 p.
• MARTIN L, FOURNIER M, MOURUIART P, SIEFEDDINE A AND TURQ B. 1993. Southern oscillation signal in South American palaeoclimatic data of the last 7000 years. Quarter Res 39: 338-346.
• MORENO JA. 1961. Clima do Rio Grande do Sul. Porto Alegre: Secretaria da Agricultura do Rio Grande do Sul, 42 p.
• NIMER E. 1989. Climatologia do Brasil. Rio de Janeiro: IBGE, 421 p.
• ORLÓCI L, PILLAR VD, ANAND M AND BEHLING H. 2002. Some interesting characteristics of the vegetation process. Commun Ecol 3: 125-146.
• OVERBECK GE, MÜLLER SC, PILLAR VD AND PFADENHAUER J. 2005. Fine-scale post-fire dynamics in southern Brazilian subtropical grassland. J Veg Sci 16: 655-664.
• PILLAR VD. 1999. The bootstrapped ordination reexamined. J Veg Sci 10: 895-902.
• PILLAR VD. 2003. Dinâmica da expansão florestal em mosaicos de floresta e campo no Sul do Brasil. In: CLAUDINO-SALES V (Org), Ecossistemas Brasileiros: Manejo e Conservação, Fortaleza: Expressão Gráfica e Editora, p. 209-216.
• PILLAR VD. 2008. MULTIV, software for multivariate exploratory analysis, randomization testing and bootstrapresampling (for Macintosh and Windows OS). Porto Alegre. Available in: <http://ecoqua.ecologia.ufrgs.br>
• PODANI J. 2000. Introduction to the Exploration of Multivariate Biological Data. Leiden, The Netherlands: Backuys Publishers, 407 p.
• RAMBO B. 1954. Análise histórica da flora de Porto Alegre. An Bot Herb Barb Rod 6: 9-111.
• RAMBO B. 1956. A fisionomia do Rio Grande do Sul. Porto Alegre: Selbach, 456 p.
• RAMBO B. 1961. Migration routes of the South Brazilian rain forest. Pesq Ser Bot 12: 1-54.
• REITZ R, KLEIN RM AND REIS A. 1988. Projeto Madeira do Rio Grande do Sul. Porto Alegre: Corag, 525 p.
• RODRIGUES PH. 2008. Didelphimorphia, Chiroptera e Rodentia (Mammalia) do Holoceno do estado do Rio Grande do Sul, Brasil: Aspectos taxonômicos, paleoambientais e paleoclimáticos. Tese de Doutorado, Instituto de Geociências, Universidade Federal do Rio Grande do Sul, RS, 203 p.
• SALGADO-LABOURIAU ML. 2007. Critérios e técnicas para o Quaternário. São Paulo: Ed. Blücher, 387 p.
• STOCKMARR J. 1971. Tablets with spores used in absolute pollen analysis. Pollen Spores 13: 615-621.
• TEIXEIRA MB, COURA NETO AB, PASTORE U AND RANGEL FILHO ALR. 1986. Vegetação. In: PROJETO RADAMBRASIL, Levantamento de recursos naturais, Rio de Janeiro: IBGE 33: 541-620.
• VILLWOCK JA AND TOMAZELLI LJ. 1995. Geologia Costeira do Rio Grande do Sul. Notas Técnicas 8: 1-45.
• VILLWOCK JA, TOMAZELLI LJ, LOSS EL, DEHNHARDT EA, HORN NO, BACHI FA AND DEHNHARDT BA.1986. Geology of the Rio Grande do Sul coastal plain. Quat S Am and Antarct Penins 4: 79-97.
• WENINGER B, JÖRIS O AND DANZEGLOCKE U. 2004. CalPal - The Cologne radiocarbon CALibration and PALaeoclimate research package. Available in: <http://www.calpal.de>

Publication Dates

• Publication in this collection
  27 Aug 2010
• Date of issue
  Sept 2010

History

• Accepted
  08 Dec 2009
• Received
  19 June 2009