Palaeobotany illuminates past environments by relating the fossilised species to the existing geological conditions. This has previously been done with fossilised leaves and spores but not with fossilised wood. The recovery of a significant quantity of wood from an area of Tertiary sediments in New South Wales, enabled the used of fossilised wood as a palaeoenvironmental tool. Tertiary sedimentary deposits of south eastern Australia are diverse lithologically, occupy distinct areas and are limited in vertical and horizontal extent. However, samples in museum collections together with samples from field work and descriptions of fossil wood from previous researchers enabled an analysis of the fossil wood. The geological and palaeontological aspects of the fossil wood were considered for each specimen. Only specimens with precise information concerning location and description of the sedimentary deposits in which the specimens were found were investigated. Lithology, sedimentary structures and the relationship with surrounding geological units were also considered. The samples were then classified and identified. It was possible to identify fossil wood to Family level by comparison with existing taxa. In the majority of cases, identification to species level was not possible due to the lack of detail in the specimen and because features such as colour cannot be used with fossilised specimens. With Australian fossilised wood, a systematic nomenclature based on structure observed within the palaeotaxa, would be more relevant. Comparisons of cell structures with previous work on palaeoenvironmental indicators was found to be possible. Fossil wood has two uses. Firstly, as a local environmental indicator, usually in conjunction with sedimentological data, assessing the rate and direction of water flow, types of depositional environments and localised floral assemblages. Secondly, as an indicator of regional climate. Within any one particular time period, comparisons between the cellular structures of wood found in different parts of south eastern Australia show gross changes in cell size, mean growth ring size and vessel size, which enabled generalisations about climate for each epoch in the Tertiary. Palaeoclimatic indicators from the wood concurred with previous climatic interpretations based on palynology and sedimentology. Cool conditions during the Palaeocene were clearly indicated by small cells and small growth rings which gradually increased throughout the remainder of the Tertiary. Several areas e.g., Dargo High Plains, where cold conditions existed in isolation could be clearly distinguished. This corresponds with the gradual northward movement of the Australian plate with consequent increasing temperatures on the mainland.

Dendroarchaeology is often discussed with reference to prehistoric contexts; however, it is equally relevant in historical contexts. Dendroarchaeological analysis independently dates the modem reconstruction of prehistoric Kinishba Pueblo with higher resolution than known documentary sources provide and illuminates reroofing practices. Kinishba is a large Mountain Mogollon pueblo on the Fort Apache Indian Reservation in east-central Arizona. Historically, the reconstructed Kinishba is attributed to the 1930s work of Byron Cummings. Nineteen dates from structural beams in nine rooms of the pueblo form a tight cluster with a strong terminal peak at 1950. This peak combined with beam and architectural attributes provides evidence for an undocumented, extensive reroofing episode. Thus, the target event was not the building of reconstructed rooms as originally expected, but replacement of the original reconstructed roofs. This study illustrates two important dendroarchaeological points. First, preconceived ideas, even if based on sound research, should not influence sample dating, which should be free of bias. Second, this study illustrates the importance of tree-ring dates in archaeology and the potential conflict between other types of data and tree-ring dates.

Cellulose is the most abundant and renewable biopolymer on earth. Generally, it is well known for its structure properties or its natural and industrial derivatives due to its biodegradability or easily controlled transformation. By contrast, non-biodegradable cellulose is relatively unusual and not intensively investigated. The discovery of well preserved Miocene fossil woods during a recent excavation from the Entre-Sambre-et- Meuse (ESEM) karsts (southern Belgium) gives a rare opportunity to investigate resistant cellulose. The wood specimens were examined using physicochemical and biochemical techniques in order to correlate the exceptional preservation of these fossilized remains after 15 million years to the non-biodegradability of their lignocellulose content. Structural and chemical changes were assessed by comparing the structural features of the fossil samples with those of their modern counterpart, Metasequoia. Solid state 13C nuclear magnetic resonance (NMR) and microscopic analysis showed good preservation of the cellulose structure in the fossil wood from the ESEM peat deposit. Moreover, there also appears to be a complete loss of hemicelluloses in the fossil wood structure maybe due to their branched structure and their lower molecular weight. According to several authors, the peatification, the initial biochemical stage of coalification, is characterized by a complete loss of hemicelluloses and a significant reduction in cellulose, which is completely degraded after a short span of geological time. However, the cellulose content, similar to that found in the living species Metasequoia, suggests that the ESEM fossil wood is at a very early stage of peatification. Compared to the lignin in modern Metasequoia, the lignin structure has undergone slight chemical alteration such as demethoxylation, cleavage of some b-O-4¢ linkages and alkylation of the resulting catechol-like structures. Furthermore, since uncondensed structures were degraded in the fossil woods, the condensed structures already present in the original lignin, became more dominant. Despite the substantial cellulose fraction available in the fossil tissue, degradation tests showed that the fossil cellulose could not be degraded by cellulases and microorganisms usually involved in the biodegradation of organic matter. This lack of bioavailability could be due to the structural features of the fossil biopolymers (cellulose, extractives and lignin) or the burial environment. In our work, we have analyzed the influence of the cristallinity, the size of crystallites, the type of crystal lattice and the ratio of two allomorphs I a/I b as well as the morphology of the cellulose microfibrilles on the cellulose digestibility. These cellulose structural features seemed to have no effect on cellulose biodegradability in the Miocene woods samples. On the other hand, the wood extractives (tannins, resin acids, terpenes,) may contribute to some resistance properties of the fossil cellulose but their action is not sufficient to explain the inhibition of the cellulose degradation. On the basis of our observations, we suggest that the presence of a modified lignin structure greatly influences cellulose biodegradability. Indeed, by altering the lignin structure with various delignification pretreatments and studying the effect of the resulting changes on enzymes efficiency, we showed that the disintegration of the condensed structures, linked to the drastic reduction of the uncondensed structures, could have significant impact on the major improvement of the cellulose bioavailability. In conclusion, the nature and/ or the proportion of intermolecular substructures could be the key of the cellulose protection. La cellulose est le biopolymère le plus abondant et le plus renouvelable sur terre. Généralement, il est reconnu pour ses propriétés structurales ou ses dérivés naturels et industriels, eux-mêmes caractérisés par leur biodégradabilité ou leur mise en uvre facilement contrôlée. Par contraste, une cellulose non-biodégradable est relativement inhabituelle et peu étudiée. La découverte de bois fossiles du Miocène extrêmement bien préservés au cours dexploitations récentes de karsts de lEntre-Sambre-et-Meuse (ESEM, Belgique) offre une rare opportunité détudier une cellulose résistante. Les spécimens de bois ont été examinés à laide de techniques physicochimiques et biochimiques en vue de corréler la préservation exceptionnelle de ces macrofossiles après 15 millions dannées à la non-biodégradabilité de leur contenu cellulosique. Les modifications structurales et chimiques ont été attribuées en comparant les caractéristiques des échantillons fossiles avec ceux de leur équivalent moderne, le Metasequoia. Les analyses microscopiques et par résonance magnétique nucléaire C13 de létat solide ont révélé une préservation importante de la structure cellulosique dans les bois fossiles des dépôts argilo-tourbeux de lESEM. De plus, la concentration en cellulose est proche de celle présente dans lespèce moderne, Metasequoia. Par contre, une perte complète des structures hémicellulosiques est constatée. Celle-ci pourrait être attribuée à leurs structures branchées et leurs faibles poids moléculaires. Ces observations suggèrent que les bois fossiles de lESEM se situent à un stade très précoce de la turbification. En effet, selon différents auteurs, la turbification, le stade biochimique initial de la coalification, est caractérisée par une perte complète des hémicelluloses et une réduction significative de la cellulose, qui est complètement dégradée après un très court intervalle de temps géologique. Par comparaison avec la lignine de lespèce moderne, la structure ligneuse fossile a subi de légères altérations chimiques telles que des déméthoxylations, le clivage de certains liens intermoléculaires (b-O-4¢ en majorité) et lalkylation des structures catéchols résultantes. En outre, les structures condensées déjà présentes dans la lignine dorigine, sont devenues plus dominantes par rapport aux structures non-condensées qui ont subi de fortes dégradations dans les bois fossiles. Malgré la fraction cellulosique substantiellement accessible dans les tissus fossiles végétaux, les tests de dégradation ont montré que la cellulose fossile ne pouvait être que faiblement dégradée par des cellulases et les microorganismes impliqués habituellement dans les processus de biodégradation de la matière organique. Ce manque de réactivité pourrait être lié aux caractéristiques structurales des biopolymères fossiles (cellulose, composés extractibles et lignine) ou à la matrice denfouissement. Limplication de celle-ci dans le processus de préservation de la cellulose a été écarté car les conditions environnementales présentes actuellement dans les karsts auraient du favoriser la biodégradabilité de ces bois. Au cours de nos recherches, nous avons analysé linfluence de la cristallinité, la taille des cristallites, le type de réseau cristallin et le rapport entre les allomorphes I a/I b ainsi que la morphologie des microfibrilles de cellulose sur la digestibilité de la cellulose. Ces paramètres structuraux associés à la cellulose ne semblent pas avoir deffet sur la biodégradabilité de la cellulose dans les échantillons de bois du Miocène. Dautre part, les composés extractibles (tannins, acides résiniques, terpènes,) peuvent contribuer à la capacité de résistance à la décomposition de la cellulose fossile mais leur action nest pas suffisante pour expliquer linhibition importante de la dégradation de la cellulose. Sur base de nos observations, nous suggérons que la présence dune structure ligneuse modifiée influence fortement la biodégradabilité de la cellulose. En effet, en altérant la structure ligneuse avec différents prétraitements de délignification et en étudiant limpact des transformations structurales sur lefficacité de la dégradation enzymatique, nous avons montré que la désagrégation des structures condensées, liée à la diminution drastique des structures non-condensées, pourrait avoir un impact significatif sur lamélioration majeure de la bioaccessibilité du substrat cellulosique. En conclusion, la nature et/ou la proportion des sous-unités intermoléculaires présentes dans la structure ligneuse pourrait être la clé de la protection exercée par la lignine.

The Gray Fossil Site in northeastern Tennessee preserves materials from a 5-million-year-old ecosystem, including wood from nearby trees. Trees provide a unique perspective into ecosystems as they are immobile during their lifetime. As such, tree rings convey information about precipitation and temperature for a given location. As southern Appalachia is one of North America’s most biodiverse regions, and modern climates are warming, understanding the warmer past climate of the Gray Fossil Site can help predict the region’s future. This study consists of three parts: conservation of wood remains, identification of taxonomic groups represented by the fossil wood, and the measured organic content of fossil wood from the Gray Fossil Site. When excavated, the wood is saturated due to a modern local high water table. A variety of drying methods from paleontology, archaeology, and dendrology were compared on saturated samples, from string wrapping to refrigeration, alcohol replacement to microwaving. The method least likely to cause warping and cracking is to wrap specimens in cotton string to slow drying. Microscopic examination of the wood reveals tree rings that have identifying features. Utilizing identification keys for tree rings, each specimen with visible rings can be identified. Taxa represented by fossil wood specimens are like those present in pre-modern forests. Finally, loss on ignition tests reveal that the Gray Fossil Site wood lacks extensive permineralization or mineral replacement. As the wood is 80-90% organic, alpha-cellulose can be extracted. The presence of alpha-cellulose, albeit stained with iron oxides, indicates that future stable isotope analyses are possible.

Two independent local tree-ring chronologies of oak in the Netherlands are described. Chronology I contains recent wood, wood from mills, and paintings. This chronology begins in A.D. 1973 and goes back to 1385 and is thought to be from inland areas of the Netherlands and the adjacent German area. Chronology 11 is built up from paintings and sculptures and ranges from A.D. 1623 to 1140; the origin of the wood is presumed to be from a coastal site in the Netherlands.

Huber identified in samples from the forests of central Europe features for characterising by their wood structure the two species of British oak. We have confirmed for recently felled oaks the suitability of his method of analysis and applied it to timbers from ancient buildings and to samples from sub-fossil oaks. The two species appear to have persisted in separate locations during the last ice age. Such analysis of the numerous Holocene oaks in Europe now being dated by den - drochronology offers the possibility of studying the separate migration of the species.

The Gray Fossil Site in northeastern Tennessee preserves materials from a 5-million-year-old ecosystem, including wood from nearby trees. When excavated, the wood is saturated due to a modern local high water table. Moisture in the wood prevents further dendroecological research, which would provide important, annual-scale climate information from tree rings visible in the wood. In order to analyze climate-sensitive wood variables, wood samples must be dried with minimal cracking prior to further research. To test the best method for drying wood samples, a variety of methods were studied. Cotton string, wrapped firmly around a sample, and a sandbox, comprised of a sample surrounded equally on all sides by sand within a five gallon container, were both be used to test the effects of minimizing expansion and contraction during drying. A vacuum oven, a microwave, and a refrigerator were used to monitor the rate at which the wood dries under different temperature conditions, and a control sample was dried in a fume hood as a comparison. An alcohol replacement test provided data on the rate of non-water evaporation. Drying methods were evaluated by measuring the drying speed of each sample and the degree of visible surface cracking. Of the methods tested, wrapping wood samples in cotton string at an even pressure, then allowing the sample to dry in a fume hood is the best practice for drying the wood from the Gray Fossil Site. The string resulted in the least cracking, and one of the shorter drying times without destroying the sample, as the vacuum oven and microwave tests did. This work not only provides a comparison of standard drying methods for saturated fossils of the non-wood varieties, but lays the groundwork for further studies examining the wood, tree rings, and climate at the Gray Fossil Site.

The Gray Fossil Site in northeastern Tennessee preserves materials from a 5-million-year-old ecosystem, including wood from nearby trees. This study consists of three parts: conservation of wood remains, identification of taxonomic groups represented by the fossil wood, and measuring the organic content of fossil wood from the Gray Fossil Site. When excavated, wood specimens from the site are saturated due to a high local water table. After testing seven different techniques to dry wood specimens, wrapping a specimen in string and allowing it to dry slowly was the method least likely to cause warping and cracking. Microscopic examination of wood cross sections reveal tree rings with distinct anatomical features, with implications for taxonomic identification. Tentatively identified taxa that are present at the Gray Fossil Site are similar to those present in pre-modern forests of northeastern Tennessee. Finally, loss on ignition tests indicate that the Gray Fossil Site wood lacks extensive permineralization or mineral replacement. The presence of alpha-cellulose, albeit stained with iron oxides, illustrates the potential for future stable isotope analyses.

Esta tese foi concebida para se integrar ao eixo nº 5 (Paleoclimas e Paleoambientes no final do Peistoceno e Holoceno na região de Lagoa Santa) da terceira versão do projeto temático Origens e Microevolução do Homem na América: uma abordagem paleoantropológica (FAPESP proc. 04/01321-6), onde a principal questão abordada é a existência de um grande período (de 8.000 a 2.000 anos A.P) de drástica redução de sepultamentos e vestígios arqueológicos na região de Lagoa Santa, MG. Para esse período, as pesquisas palinológicas no Brasil central e sudeste, embora existam em número razoável, possuem interpretações paleoambientais contraditórias. Bem caracterizado no Hemisfério Norte, ao redor de 6000 anos A.P., o evento climático conhecido como Ótimo Climático, período predominantemente mais quente e úmido, se estabeleceu devido a uma maior insolação terrestre, propiciando a expansão de coberturas vegetais de maior porte e umidade de diversos pontos do globo e é sustentado por diversos autores para muitas localidades brasileiras. Por outro lado, embora não existam dúvidas de que o nível do mar esteve acima do atual na costa brasileira, é consenso que alterações climáticas foram mais amenas e aparentemente mais complexas. Amparados por este raciocínio, outros estudos sugerem eventos de seca no período aproximado de 6000 a 4500 anos A.P. Neste contexto, objetivando verificar a provável diferença climática ocorrida durante o Holoceno médio é confirmada pelo estudo de lenhos fósseis de região e sustenta a hipótese de a redução de sepultamentos da região reflete um deslocamento humano por causas climáticas, utilizaram-se duas abordagens diferentes no estudo anatômico de lenhos fósseis datados do Holoceno médio e tardio: uma abordagem florística, feita através da identificação taxonômica das amostras fósseis e comparação com localidades atuais; e através de uma abordagem anatômico-ecológica do lenho de Myroxylon peruiferum L.f., que utilizou correlações existentes entre o clima e as características anatômicas pra predizer o clima pretérito da região com base nas características anatômicas dos fósseis. Para a primeira abordagem, foram também realizados um levantamento florístico-lenhoso da APA Carste de Lagoa Santa e estudos fitossociológicos nas diferentes fitofisionomias da região, que proveram informações adicionais para as interpretações paleoambientais. Como resultado, os estudos florísticos e fitossociológicos, além de grande valia para a interpretação das assembléias fósseis, mostraram uma riqueza expressiva de espécies na região. A identificação das assembléias fósseis do Holoceno médio e tardio mostrou muita similaridade entre as duas assembléias, sugerindo a ausência de mudanças climáticas significativas a ponto de alterar a florística da região. Ainda, essas assembléias fósseis são relacionadas fortemente com o mosaico de vegetações existentes na região, que aponta para a manutenção desta vegetação desde o Holoceno médio. Por outro lado, a abordagem anatômico-ecológica revelou uma maior sazonalidade climática durante o Holoceno médio. Embora pareçam conflitantes, estes resultados são interpretados como complementares, e indicam um período sensivelmente mais sazonal no Holoceno médio, porém insuficientes para propiciar alterações florísticas no mosaico vegetacional que existia na época e caracteriza hoje a região. Esses resultados, portanto, não apóiam a hipótese de alterações intensas no clima, na flora e nem na presença das vegetações que ali existiam desde 5.000 anos A.P. / This thesis is designed to integrate the shaft Nº. 5 (Paleoclimates and paleoenvironments in the Late Pleistocene and Holocene at Lagoa Santa region) of the third version of the Project Origins and Microevolution of man in América: an approach paleoanthropological (FAPESP proc. 04/01321-6), where the main issue addressed is the existence of a large period (8000-2000 years BP) of drastic reduction of burials and archaeological sites in Lagoa Santa, MG. For that period, palynological research in central and southeastern Brazil, althoug in reasonable numbers, have contradictory paleoenvironmental interpretations. Well characterized in the Northern Hemisphere, around 6000 yeras BP, the climatic event known as Climatic Optimum, predominantly warmer and wetter, was established due to greater solar insolation, favoring the expansion of humid vegetation cover various parts of the globe and is supported by several authors for many places in Brazil. Moreover, although there is no doubt that the sea level was above the current along the Brazilian coast, there is consensus that climate change was milder and seemingly more complex. In the same way, other studies suggest drought events in the approximate period 6000 to 4500 years B.P. In this context, to verify if the climatic differences likely ocurred during the middle Holocene is confirmed by the study of fóssil Wood from the region and supports the hypothesis that the reduction of burials in the region reflects a human migration causes climate, we used two different approaches in a anatomical study of fóssil Wood dated to the middle and late Holocene: 1.) a floristic approach, made through the taxonomic identification of fóssil specimens and comparison with current local florist, and 2.) through na approach of ecological Wood anatomy of Myroxylon peruiferum L.f., which correlation between climate and anatomical characteristics were used to predict the past climate of the region based on the same anatomical characteristics in fossils samples. For the first approach, were also carried out a floristica-woody survey of APA-Lagoa Santa Karst and phytosociological studies in different Forest types in the region, which provided additional information for paleoenvironmental interpretations. As a result, floristic and phytosociological studies, besides showing great importance for the interpretation of fóssil assemblages, revealed and expressives richness of species. The identification of fóssil assemblages from middle and late Holocene showed high similarity between both, suggesting the absence of enough climate change lo alter the flora in the region. Still, these fóssil assemblages are strongly related with the mosaico of vegetation in the region, pointin to the maintenance of vegetation since the middle Holocene. Furthermore, the ecological-anatomy approach revealed a greater climatic seasonality during the middle Holocene. Although these two approaches seem conflicting, these results are interpreted as complementary, and show a markedly more seasonal period in middle Holocene, but insufficient to provide floristic changes in the vegetation mosaic that existed nowadays and characterizes the region. These results therefore do not support the hypothesis of intense changes in climate, flora and even in the presence of vegetation that existed there since 5000 years BP.

Esta tese foi concebida para se integrar ao eixo nº 5 (Paleoclimas e Paleoambientes no final do Peistoceno e Holoceno na região de Lagoa Santa) da terceira versão do projeto temático Origens e Microevolução do Homem na América: uma abordagem paleoantropológica (FAPESP proc. 04/01321-6), onde a principal questão abordada é a existência de um grande período (de 8.000 a 2.000 anos A.P) de drástica redução de sepultamentos e vestígios arqueológicos na região de Lagoa Santa, MG. Para esse período, as pesquisas palinológicas no Brasil central e sudeste, embora existam em número razoável, possuem interpretações paleoambientais contraditórias. Bem caracterizado no Hemisfério Norte, ao redor de 6000 anos A.P., o evento climático conhecido como Ótimo Climático, período predominantemente mais quente e úmido, se estabeleceu devido a uma maior insolação terrestre, propiciando a expansão de coberturas vegetais de maior porte e umidade de diversos pontos do globo e é sustentado por diversos autores para muitas localidades brasileiras. Por outro lado, embora não existam dúvidas de que o nível do mar esteve acima do atual na costa brasileira, é consenso que alterações climáticas foram mais amenas e aparentemente mais complexas. Amparados por este raciocínio, outros estudos sugerem eventos de seca no período aproximado de 6000 a 4500 anos A.P. Neste contexto, objetivando verificar a provável diferença climática ocorrida durante o Holoceno médio é confirmada pelo estudo de lenhos fósseis de região e sustenta a hipótese de a redução de sepultamentos da região reflete um deslocamento humano por causas climáticas, utilizaram-se duas abordagens diferentes no estudo anatômico de lenhos fósseis datados do Holoceno médio e tardio: uma abordagem florística, feita através da identificação taxonômica das amostras fósseis e comparação com localidades atuais; e através de uma abordagem anatômico-ecológica do lenho de Myroxylon peruiferum L.f., que utilizou correlações existentes entre o clima e as características anatômicas pra predizer o clima pretérito da região com base nas características anatômicas dos fósseis. Para a primeira abordagem, foram também realizados um levantamento florístico-lenhoso da APA Carste de Lagoa Santa e estudos fitossociológicos nas diferentes fitofisionomias da região, que proveram informações adicionais para as interpretações paleoambientais. Como resultado, os estudos florísticos e fitossociológicos, além de grande valia para a interpretação das assembléias fósseis, mostraram uma riqueza expressiva de espécies na região. A identificação das assembléias fósseis do Holoceno médio e tardio mostrou muita similaridade entre as duas assembléias, sugerindo a ausência de mudanças climáticas significativas a ponto de alterar a florística da região. Ainda, essas assembléias fósseis são relacionadas fortemente com o mosaico de vegetações existentes na região, que aponta para a manutenção desta vegetação desde o Holoceno médio. Por outro lado, a abordagem anatômico-ecológica revelou uma maior sazonalidade climática durante o Holoceno médio. Embora pareçam conflitantes, estes resultados são interpretados como complementares, e indicam um período sensivelmente mais sazonal no Holoceno médio, porém insuficientes para propiciar alterações florísticas no mosaico vegetacional que existia na época e caracteriza hoje a região. Esses resultados, portanto, não apóiam a hipótese de alterações intensas no clima, na flora e nem na presença das vegetações que ali existiam desde 5.000 anos A.P. / This thesis is designed to integrate the shaft Nº. 5 (Paleoclimates and paleoenvironments in the Late Pleistocene and Holocene at Lagoa Santa region) of the third version of the Project Origins and Microevolution of man in América: an approach paleoanthropological (FAPESP proc. 04/01321-6), where the main issue addressed is the existence of a large period (8000-2000 years BP) of drastic reduction of burials and archaeological sites in Lagoa Santa, MG. For that period, palynological research in central and southeastern Brazil, althoug in reasonable numbers, have contradictory paleoenvironmental interpretations. Well characterized in the Northern Hemisphere, around 6000 yeras BP, the climatic event known as Climatic Optimum, predominantly warmer and wetter, was established due to greater solar insolation, favoring the expansion of humid vegetation cover various parts of the globe and is supported by several authors for many places in Brazil. Moreover, although there is no doubt that the sea level was above the current along the Brazilian coast, there is consensus that climate change was milder and seemingly more complex. In the same way, other studies suggest drought events in the approximate period 6000 to 4500 years B.P. In this context, to verify if the climatic differences likely ocurred during the middle Holocene is confirmed by the study of fóssil Wood from the region and supports the hypothesis that the reduction of burials in the region reflects a human migration causes climate, we used two different approaches in a anatomical study of fóssil Wood dated to the middle and late Holocene: 1.) a floristic approach, made through the taxonomic identification of fóssil specimens and comparison with current local florist, and 2.) through na approach of ecological Wood anatomy of Myroxylon peruiferum L.f., which correlation between climate and anatomical characteristics were used to predict the past climate of the region based on the same anatomical characteristics in fossils samples. For the first approach, were also carried out a floristica-woody survey of APA-Lagoa Santa Karst and phytosociological studies in different Forest types in the region, which provided additional information for paleoenvironmental interpretations. As a result, floristic and phytosociological studies, besides showing great importance for the interpretation of fóssil assemblages, revealed and expressives richness of species. The identification of fóssil assemblages from middle and late Holocene showed high similarity between both, suggesting the absence of enough climate change lo alter the flora in the region. Still, these fóssil assemblages are strongly related with the mosaico of vegetation in the region, pointin to the maintenance of vegetation since the middle Holocene. Furthermore, the ecological-anatomy approach revealed a greater climatic seasonality during the middle Holocene. Although these two approaches seem conflicting, these results are interpreted as complementary, and show a markedly more seasonal period in middle Holocene, but insufficient to provide floristic changes in the vegetation mosaic that existed nowadays and characterizes the region. These results therefore do not support the hypothesis of intense changes in climate, flora and even in the presence of vegetation that existed there since 5000 years BP.