THE HISTORY OF Robyn J. Burnham2 and Alan Graham;NEOTROPICAL VEGETATION:NEW DEVELOPMENTS ANDSTATUS'ABSTRACT The isolation of South America from Central America and Africa during the Tertiary Period left a strong imprint onthe flora of the Neotropics. South American Eocene through Miocene fossil assemblages, both pollen and macrofossils,document a rich tropical flora on the continental margins, and represent some of the only data on pre-landbridgelowland taxa in South America. Lowland Miocene floras from Amazonia are remarkably similar in their high diversityto Amazonian floras today based on lists of dominant families. Recent geophysical data on the uplift of the northernAndes show a strong correlation between uplift and the development and diversification of montane forests in Colombiaand Venezuela. The emergence of a continuous landbridge at 3 Ma between Central and South America is welldocumented and is demonstrated by the arrivai of temperate elements in South American highlands and concurrentappearance of South American taxa in Central America. There is no evidence for displacement of lowland tropicalplants in South America by northern immigrants, which appears to stand in contrast to the published record formammals. The mix of taxa in extant Mexican tropical floras derived from tropical South America, tropical CentralAmerica, and from remnants of northern tropical Eocene floras is strong evidence for the impact that the landbridgethrough the Panamanian isthmus had on the neotropical flora. The early appearance of low-elevation savannas is inferredfrom an increase in grass pollen in the middle Pliocene of Panama; however, widespread savannas are not indicatedhy pollen data from the Central American region. Rather, beginning in the latest Miocene Epoch and continuing up tothe Quaternary, a mix of tropical rainforest and mixed tropical woodlands is suggested for the lowlands, based on pollenevidence. Accumulating data on temperature changes during the late Tertiary and Quaternary Periods points to low-latitude temperature fluctuations of up to 6�C. Proposais of accompanying widespread rainfall fluctuations are moreequivocal. Rainfall probably varied regionally, resulting in a mosaic of habitats controlled by river migration, sea levelfluctuations, local dryness, and local uplift. Zones postulated as refugia provide testable hypotheses using neoecologicaland paleoecological data. The paleoecological data to test these hypotheses are still limited taxonomically and spatially.It is important to stress that the effect of the isolation of South American neotropical floras has not been erased in the3 million years since their connection with Central America. New data from middle and late Miocene floras in SouthAmerica will be critical in determining the degree to which the composition of South American floras has beeninfluenced by immigration of plants from the better-known Central American area to the north. The neotropics extend geographically from theTropic of Cancer to the Tropic of Capricorn, in-cluding environments as diverse as dry desert orhumid rainforest. Mean annual temperatures rangefrom over 30�C to as low as 10�C. Elevations fromsea level up to well over 6000 m are included. Thiswide range of climate and topography has a pro-found effect on the composition and structure ofneotropical vegetation. The area also has a distinc-tive geographic outline. It is roughly hour-glassshaped, with the narrow isthmus of Panama com-posing the fragile, primarily lowland connection be-tween large northern and southern land masses. Inaddition, both the northern and southern neotropicslie contiguous to subtropical and temperate areasto the north and south: no barriers to immigrationfrom these extra-tropical sources are present. Thereare as many reasons for neotropic patterns of di-versity and distribution today as there are differentlandforms and climatic regimes. If these differencesare played out over the course of the late Mesozoicand Cenozoic, when data from fossil deposits caninform us about the sequence of changes that hastaken place, the history of neotropical vegetationtakes on a dizzying complexity. We approach this review of the vegetational his-tory of the Neotropics by outlining four importantevents (Fig. 1). We initially treat the two major landareas separately, because for much of their historythey were indeed isolated from one another. Then,we summarize the interactions that postdate theconnection and treat selected aspects of the Neo-' We are grateful for careful review of earlier versions of this manuscript by Paul A. Colinvaux, Peter R. Crane,David L. Dilcher, and two anonymous reviewers. We also extend our appreciation to Bruce D. Patterson and David S.Webb for discussion of and comments on patterns of mammalian diversification and migration. 2 Museum of Paleontology, University of Michigan, Ann Arbor, Michigan 48109-1079, U.S.A. :' Department of Biological Sciences, Kent State University, Kent, Ohio 44242, U.S.A.ANN. MISSOURI BOT. GARD. 86: 546-589. 1999.