Tài liệu Global Change and the Function and Distribution of Wetlands ppt

The goal is to highlight the latest research from the world leaders researching climate change in wetlands, to disseminate research fi ndings on global change ecology, and to provide sou

of Wetlands

an important literature gap in the fi eld of global change as it relates to wetlands around the world The goal is to highlight the latest research from the world leaders researching climate change in wetlands, to disseminate research fi ndings on global change ecology, and to provide sound science

to the public for decision-making on wetland policy and stewardship Each volume will address a topic addressed by the annual symposium of the Society’s Global Change Ecology Section.

For further volumes:

http://www.springer.com/series/8905

Global Change and the Function and Distribution

of Wetlands

Beth A Middleton

National Wetlands Research Center

US Geological Survey

Lafayette, LA, USA

ISBN 978-94-007-4493-6 ISBN 978-94-007-4494-3 (eBook)

DOI 10.1007/978-94-007-4494-3

Springer Dordrecht Heidelberg New York London

Library of Congress Control Number: 2012942468

Chapters 2 and 4: © The U.S Government’s right to retain a non-exclusive, royalty-free licence in and

to any copyright is acknowledged 2012

© Springer Science+Business Media Dordrecht 2012

This work is subject to copyright All rights are reserved by the Publisher, whether the whole or part of the material is concerned, speci fi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on micro fi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed Exempted from this legal reservation are brief excerpts

in connection with reviews or scholarly analysis or material supplied speci fi cally for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work Duplication

of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer Permissions for use may be obtained through RightsLink at the Copyright Clearance Center Violations are liable to prosecution under the respective Copyright Law.

The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a speci fi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Part I Paleoecology and Climate Change

Insights from Paleohistory Illuminate Future Climate Change

Effects on Wetlands 3Ben A LePage, Bonnie F Jacobs, and Christopher J Williams

Part II Sea Level Rise and Coastal Wetlands

Response of Salt Marsh and Mangrove Wetlands to Changes

in Atmospheric CO 2 , Climate, and Sea Level 63Karen McKee, Kerrylee Rogers, and Neil Saintilan

Part III Atmospheric Emissions and Wetlands

Key Processes in CH 4 Dynamics in Wetlands and Possible Shifts

with Climate Change 99Hojeong Kang, Inyoung Jang, and Sunghyun Kim

Part IV Drought and Climate Change

The Effects of Climate-Change-Induced Drought

and Freshwater Wetlands 117

Beth A Middleton and Till Kleinebecker

Index 149

Paleoecology and Climate Change

B.A Middleton (ed.), Global Change and the Function and Distribution of Wetlands,

Global Change Ecology and Wetlands 1, DOI 10.1007/978-94-007-4494-3_1,

© Springer Science+Business Media Dordrecht 2012

Abstract Climate change could have profound impacts on world wetland environments, which can be better understood through the examination of ancient wetlands when the world was warmer These impacts may directly alter the critical role of wetlands in ecosystem function and human services Here we present a framework for the study of wetland fossils and deposits to understand the potential effects of future climate change on wetlands We review the methods and assump-tions associated with the use of plant macro- and microfossils to reconstruct ancient wetland ecosystems and their associated paleoenvironments We then present case studies of paleo-wetland ecosystems under global climate conditions that were very different from the present time Our case study of extinct Arctic forested-wetlands reveals insights about high-productivity wetlands that fl ourished in the highest lati-tudes during the ice-free global warmth of the Paleogene (ca 45 million years ago) and how these wetlands might have been instrumental in keeping the polar regions warm We then evaluate climate-induced changes in tropical wetlands by focusing

on the Pleistocene and Holocene (2.588 Myr ago to the present) of Africa These past

Roy M Huf fi ngton Department of Earth Sciences , Southern Methodist University ,

P.O Box 750395 , Dallas , TX 75275-0395 , USA

e-mail: bjacobs@smu.edu

C J Williams

Department of Earth and Environment , Franklin and Marshall College ,

P.O Box 3003 , Lancaster , PA 17604-3003 , USA

e-mail: chris.williams@fandm.edu

Climate Change Effects on Wetlands

Ben A LePage , Bonnie F Jacobs , and Christopher J Williams

ecosystems demonstrate that subtle changes in the global energy balance had signi fi cant impacts on global hydrology and climate, which ultimately determine the composition and function of wetland ecosystems Moreover, the history of these regions demonstrates the inter-connectedness of the low and high latitudes, and the global nature of the Earth’s hydrologic cycle Our case studies provide glimpses of wetland ecosystems, which expanded and ultimately declined under a suite of global climate conditions with which humanity has little if any experience Thus, these paleoecology studies paint a picture of future wetland function under projected global climate change

1 Introduction

Virtually every aspect of the planet Earth, especially climate, has changed over the last four billion years There is no reason to believe that these changes will cease, or more to the point, that we can stop such changes because they are now impacting our daily lives From a geological point of view, global climate change is inevitable, and we need to ask ourselves whether our efforts to curb such change is likely to have the desired mitigating effect? While the solution is complicated and certainly cannot be answered within the context of this chapter, our goal is to help put global climate change into a geological perspective with respect to wetlands

When Earth’s history is viewed in a geological context, we see a planet that has always been in a state of geologic and geomorphologic fl ux The Earth’s climate has changed considerably throughout geologic time and ironically, we live at one of the few times when global climate is cold, or what geologists call “icehouse conditions” For most of Earth’s history “hothouse or greenhouse conditions” prevailed, ice caps were absent, and the average global temperature was considerably warmer than at present The consensus among scientists is the anthropogenic input of greenhouse gases to the atmosphere, particularly carbon dioxide (CO 2 ), have triggered a phase of

global warming (Solomon et al 2007 ; Rosenzweig et al 2008 ) The pace and sity of future warming and the associated signi fi cant environmental changes are likely to be governed, in part, by anthropogenic greenhouse gas inputs

What then can the study of ancient wetland communities, some from millions of years ago, offer to understand better the effects of future climate change on wet-lands? It is important that we frame our discussion of wetland impacts in the context

of world wetland extent The current global wetland area is estimated to be mately 12.8 million square kilometers (km 2 ) or 8.6% of the total land area of the world (Schuyt and Brander 2004 ) In an ice-free world, the total wetland area could double in size to 25 million km 2 (18% of the total land area) if we assume that at least 50% of the area currently classi fi ed as ice (Greenland and Antarctica) and tundra would become wetland and the current wetland area of 12.8 million km 2 would be maintained This assumption seems reasonable judging from the geo-graphic extent and amount of Cenozoic-age (Fig 1 ; 65.5 to 2.588 million years old [Myr]) coals in northern and Arctic Canada, Iceland, Spitsbergen, Alaska, and Russia

approxi-Berriasian Valanginian Hauterivian Barremian Aptian Albian

Lower

Upper

Cenomanian Turonian Coniacian Santonian Campanian Maastrichtian Danian Selandian Thanetian Ypresian Lutetian Bartonian Priabonian Rupelian Chattian

Paleocene Eocene Oligocene Miocene Pliocene Pleistocene

Holocene

Tarentian Ionian Calabrian Gelasian Piacenzian Zanclean Messinian Tortonian Serravallian Langhian Burdigalian Aquitanian

Calibrated Age (Myr)

0.0117 0.130 0.781 1.806 2.588 3.600 5.332 7.246 11.608 13.82 15.97 20.43 23.03 28.4 33.9 37.2 40.4 48.6 55.8 58.7 61.1 65.5 70.6 83.5 85.8 88.6 93.6 99.6 112.0 125.0 130.0 133.9 142.2 145.5

Fig 1 Stratigraphic chart

showing the ages in millions

of years (Myr) of the geologic

periods and epochs The ages

follow those adopted by the

International Commission on

Stratigraphy ( 2010 )

These coal deposits indicate large areas of moderately productive wetlands extended from 50°N to the pole in the Northern Hemisphere throughout the Paleogene and Neogene (Bustin 1981 ; Bustin and Miall 1991 ; Kalkreuth et al 1993 ) Therefore, most of the 11.5 million km 2 of area currently classi fi ed as tundra may become wetland during future climate change so the 50% estimate of the conversion of tun-dra to wetlands is most likely an underestimate Nevertheless, global climate change will considerably increase the area of wetlands on the planet and these wetlands will undoubtedly have signi fi cant impacts on future climate change, carbon and nutrient cycling, and biodiversity

This chapter is focused on insights that can be garnered from the past that help

us understand the impact of global climate change on wetlands Paleobotanical research can illuminate past climate and other environmental conditions through the plant macrofossil (leaves, seeds, fl owers, seed cones, wood) and palynomorph (pol-len and spores) records After the composition and relative abundances of species in the paleo fl ora are known, climate and paleoecology can be reconstructed based on comparisons with nearest living relatives and the morphological (the study of form and structure) attributes of fossil leaves Paleobotany can also be integrated with physical geological studies to understand better such physical processes as moun-tain building, relative sea-level change, and sediment transport, deposition, and ero-sion involved in development of the regional landscape through time The relatively new discipline of geochemistry is focused on the study of elements that were part of these ancient environments and ultimately incorporated into plant tissues When applied in a multidisciplinary framework, the tools employed by geologists, paleon-tologists, and geochemists to reconstruct past climate and environments provide a better understanding of how plant communities functioned in the past and how they could respond to changing climate and environment in the future

2 The Study of Fossil Plants and Ancient Environments

Most fossil plant assemblages are the remnants of ancient wetland communities, and by virtue of their topographically low position on the landscape, wetlands are the most likely communities to be preserved because low-lying areas are often

fl ooded or saturated with water In water or under saturated conditions, the soil and organic matter become acidic and low in oxygen (anaerobic), and these conditions restrict the saprotrophs (decomposers) that break down organic matter As a result, the rate of organic matter accumulation is greater than the rate of decomposition Therefore, the nature of the accumulated organic matter can then be used as a proxy

to represent the composition of the former wetland communities at the site Considerable insight into how ancient wetland communities responded to regional and global climate change can be gained from both temporal (time) and spatial (geographic) studies of their composition, structure, and function Paleobotanical and paleoecological studies are usually based on more fragmentary components of whole communities than their modern counterparts Fossil plant assemblages are

best viewed as snapshots in geological time that represent days to years (sometimes hundreds of years) of organic matter accumulation over varying spatial scales It is

rare to fi nd entire plant communities preserved in situ (in place) and in those

instances, the preserved plant species are generally herbaceous (Kidston and Lang

1917 ; Rothwell and Stockey 1991 ; Wing et al 1993 ; Stockey et al 1997 ) , or times woody (Francis 1987, 1988, 1991 ; Jacobs and Winkler 1992 ; Basinger 1991 ; Williams 2007 ; DiMichele and Gastaldo 2008 )

While we are cognizant of the fragmentary nature of the plant fossil record and the limitations that various plant parts provide for interpreting and reconstructing past and future environments and climate, fossil plant remains provide proxies from which reasonably robust paleoenvironmental interpretations can be made using sys-tematic assessments As such, we discuss the major groups of plant organs that are commonly recovered from sedimentary deposits and the types of interpretations that are possible based on recover of these fossil tissues Nevertheless, before we begin, it is important that the reader understand the concepts of space and time and the limitations that each imparts on interpreting the plant fossil record

3 Spatial and Temporal Resolution

When working with data generated from fossil materials, one needs to be aware of the spatial and temporal scales represented and the limitations that these data place

on paleoecological interpretations Bennington et al ( 2009 ) identi fi ed temporal and spatial components, which must be considered when working with fossils including time averaging and source area (related to transport distance), respectively Both trans-port distance and time averaging are addressed by the fi eld of taphonomy; the study of how organisms become fossils (i.e., their transition from the biosphere

to the lithosphere) Taphonomic studies provide a mechanistic understanding of the processes of transport, burial, and preservation, which are factors that may bias the paleoecological interpretation of a fossil deposit Depending on the nature of the deposit, plant fossil assemblages generally provide a good indication of the amount

of transport endured by the plant remains and these deposits can be classi fi ed as autochthonous, allochthonous, and/or parautochthonous Autochthonous deposits

are those where there has been no transport and the fossils are effectively buried in situ

These types of deposits provide the most complete record of the plant composition

in the immediate burial area Allochthonous assemblages are comprised of fossils that have been transported and buried up to a few kilometers from where they grew Parautochthonous remains were transported a smaller distance Nevertheless, from

a taphonomic standpoint, even autochthonous deposits are likely to possess a age of non-local parautochthonous and allochthonous plant elements

When sampling and interpreting fossil assemblages, it is important to consider the spatial scale with regard to each type of deposit Fossil plant assemblages pre-served in a particular stratum across a region represent snapshots in time of the dominant species and in some cases changes in the dominant species can be recognized

if the bedding plane within which the plants are contained is preserved laterally

If one were to examine the fossil plants at various locations within a single deposit there would likely be many similarities in plant composition within this stratum, which would then be a re fl ection of the dominant plant species for the time and region But depending on the distance between the sampling locations, subtle changes in the composition and relative abundance (dominance) of the vegetation would be expected throughout this local landscape These changes could be due to changes in soil conditions, aspect, micro-topography, or hydrology (Fig 2 ) For example, assuming that there were suf fi cient depositional environments within each zone (Fig 2 ), the aquatic zone would be biased towards species growing in the aquatic and riparian zones with some elements from bottomland forests or more rarely from the uplands Sampling in the bottomland forest would provide an excel-lent proxy of the species composition growing in this zone within this stratum Riparian and upland elements would be represented in low numbers, and aquatic species would not be expected Similarly, if we were to collect samples in the uplands, we would not likely encounter any aquatic, riparian, and bottomland forest elements Furthermore, lateral sampling along a single fossiliferous deposit can pro-vide paleoecological information about heterogeneity in species composition due to the biotic factors themselves

To test these well-accepted paleobotanical assumptions Burnham ( 1989, 1997 ) sampled the forest fl oor litter in a number of fl oodplain forest sub-environments in

a Mexican paratropical forest and Costa Rican dry forest A variety of sub-environments

in the same stratigraphic level was necessary to increase the accuracy of regional reconstructions (Burnham 1989 ) Moreover, certain sub-environments such as channel deposits consistently misrepresented the source fl ora Sample size was crucial for reliably reconstructing local and regional vegetation communities The leaf litter study in the dry forest indicated that 70% of the tree species per hectare were

Fig 2 The relationship between local topography and spatial changes in the vegetation The

macro- and microfossils collected in the fi eld across these vegetation types would be analyzed to determine species composition and relative abundance The sampling location and frequency determines the accuracy of vegetation and climate reconstruction for the local and regional areas

of the study

represented in the leaf collecting baskets, which were placed over the forest fl oor From these data, the dominant and co-dominant species could be determined (Burnham 1997 ) Studies such as these illustrate the importance of understanding the relationships between the ecology and dynamics of modern forested ecosys-tems, geomorphology, and taphonomy

The second component identi fi ed by Bennington et al ( 2009 ) is that of temporal mixing or so-called “time averaging”, whereby events that happened at different times appear to be synchronous in the geologic record (Kowalewski 1996 ) For example, a stratigraphic horizon could contain the remains of several generations of plant com-munities that were never contemporaries This situation is inherent to most sedimen-tary deposits, even if sediment accumulation is continuous Even with precise age controls, such as those provided by annual laminations (varves) or materials amenable

to radioisotope dating (e.g., 14 C, 210 Pb), it is sometimes dif fi cult to know exactly how much time is represented by a speci fi c stratigraphic interval at a locality

A hypothetical stratigraphic column can illustrate this point (Fig 3 ) If we assume that sediment and plant accumulation are continuous throughout the section and we

Fig 3 In sedimentology the relationship between time and sediment accumulation rates can be

illustrated using a hypothetical stratigraphic column The ages can be determined using 14 C or another radioactive isotope that has a half-life suitable for the geologic age of the deposits The sediment accumulation rates are calculated on the basis of the amount of sediment that accumu- lated during the time represented between the 14 C levels This illustrates the point that although sediment accumulation may have been constant, the rate of sediment accumulation can vary through time Single point accumulation rates are based on the use of a single age date Compared

to a stratigraphic section that has multiple age dates, the same stratigraphic section that is brated with one age date can over- or under-estimate the rate of sediment accumulation The arrows

cali-at 300 and 350 cm indiccali-ate the loccali-ation of a 50-cm thick sediment package thcali-at was deposited instantaneously, probably during a fl ood event

accumulation rates

have only one radiometric age of 20,200 years at the bottom of the section, then the average rate of sediment accumulation over the 4 m section would be 1 cm every

51 years Although this assumption is reasonable, the example illustrates that although sediment accumulation may have been continuous, the rate of accumula-tion can be variable Similarly, if only one radiometric age date of 18,600 years at a depth of 250 cm is available, then the sediment accumulation rate for the 250 cm thick sedimentary unit would be 1 cm every 74 years Again, assuming that only one radiometric age (9,060 or 5,280 years) was available, the sediment accumulation rates would be very different (45 and 106 years per centimeter) from the other radio-metric ages There are many instances where a sediment core or outcrop (also called

a geologic section) contains a limited amount of material suitable for radiometric dating (in this case 14 C) and it is only possible to obtain a single radiometric age In these cases, the sediment accumulation rate can only be calculated from the location where the sample was collected to the top of the core or section and the accumula-tion rate of the sediment located below the sample location is unknown

The example also illustrates that change in sediment accumulation rates are not identi fi ed by single age calibration points Multiple calibration points increase the accuracy for reconstructing the local vegetation community and physical setting, especially when interpretations require higher temporal resolution Moreover, the study of the sediments between radiometric dates provides constraints on the depo-sitional environment and questions such as basin stability (as it relates to tectonics), cyclicity/periodicity of the deposit, and the position of the sampling locations over the landscape can be determined In this example, four radiometric dates calibrate the section Between 20,200 and 18,600 years 150 cm of sediment accumulated over 1,600 years and between 18,600 and 9,060 years only 50 cm of sediment accu-mulated over 9,540 years From 9,060 to 5,280 years 150 cm of sediment accumu-lated over 3,780 years and the uppermost 50 cm of sediment accumulated between 5,280 years and the present Thus each centimeter of sediment between 20,200 and 16,060 years represents 11 years, between 18,600 and 9,060 years represents

191 years, between 9,060 and 5,280 years each centimeter represents 25 years, and between 5,280 years and the present each centimeter represents 106 years In this example, the sediment accumulation rates are highly variable The reconstruction of forest structure, composition, and dynamics would not be accurate if only the single point accumulations rates were used Use of any of the single point values alone would have either over- or under-estimated the time it took for the sediment to accu-mulate as well as the biological and physical processes represented during that interval of time The accumulation rates as based on the multiple point accumula-tion approach provide better estimates of the time it took for the sediment to accu-mulate within a depositional basin (Fig 3 )

Sediment accumulation rates are nothing more than averages that are based on modern processes and calibration points The concept of averaging the time taken for a package of sediment to accumulate is then applied to the vegetation preserved

in the sediment package Thus, using the example of the single radiometric age date

of 20,200 years (Fig 3 ), changes in the macro- and micro- fl ora throughout the 4 m section would be interpreted using a 51-year baseline with the assumption that deposition was continuous By virtue of the averaging process, instances of erosion

and periods of non-deposition are not considered unless the position of the erosional surface was obvious The assumption of continuous deposition and the sediment accumulation rate would no longer be valid At this point the radiometric age date could only be used to place the sediment package (up to the erosional surface) into

a chronostratigraphic framework (e.g., Epoch or Stage; see Fig 1 )

Occurrences of instantaneous deposition may further complicate the tion of sediment accumulation rates if such deposits go unrecognized in the sedi-mentary sequence Instantaneous deposits are those where a large thickness of sediment is deposited rapidly, perhaps in a matter of seconds to days These deposits are generally associated with major disturbances such as storms and mudslides as well as catastrophic events such as landslides and fl ooding induced failure of river-banks/levees In our hypothetical section (Fig 3 ), the arrows at 300 and 350 cm delineate an instantaneous deposit, which was 50-cm thick Deposits of this thick-ness are not uncommon during large fl ood events Although the deposit is bracketed

interpreta-by two radiometric age dates, the assumption of a uniform sediment accumulation rate between these dates is no longer valid The complication arises when instanta-neous deposits cannot be recognized based on sedimentological features Therefore,

if the instantaneous deposit was not differentiated from the surrounding tary deposits, then the 50-cm thick package, which was deposited over several days would be interpreted as having been deposited over 550 years In addition, analysis

sedimen-of the fossil fl ora sedimen-of this layer would lead to an inaccurate portrayal sedimen-of the actual local plant community composition, because the fl ood deposits might result in the concentration of reworked plant remains of different ages and from different loca-tions within the basin This is a good example of what paleobotanists call a time-averaged fl ora

The aspect of sediment accumulation rate is further complicated by stochastic accumulation rates, which are periods where there is no sediment accumulation, erosion, and/or a lack of geochronologic controls There are many more instances where the absolute age of a fossil assemblage is not known, but the composition of the fossils compares favorably other estimates using techniques that provide abso-lute age dates This practice is called relative age dating; however, the issue of time averaging with such deposits is magni fi ed because the entirety of the fossil deposit has only an approximation of its age and the amount of time represented in strati-graphic section is not known regardless of its thickness A centimeter of sediment could have accumulated over a period of seconds, minutes, decades, or hundreds of years or more Despite the inherent challenges presented by transport distance and time averaging, reasonably accurate reconstructions of ancient climates and envi-ronments can be made

4 Macrofossils

Plant macrofossils are organic remains of plants, which are generally large enough

to be seen without the aid of a microscope including leaves, seeds, fruits, wood, and seed and pollen cones (Figs 4–12 ) In most cases, these plant macrofossils were

preserved in fi ne-grained sediments such sandstones, siltstones, mudstones, and volcanic ash, which accumulated in small depressions, fl oodplains, lakes, swamps, and streams Depending on the type of deposit, the plant fossils are either autoch-thonous, allochthonous, or parautochthonous In all cases, each type of deposit provides information, which can be used to reconstruct the composition of the local and regional vegetation mosaic, and in some cases the environmental setting (e.g., regional climate, or local habitats including fl uvial, lacustrine, bottomland forest) In all cases, taphonomic processes determine the type and quality of plant preservation Understanding the taphonomy of a fossil plant assemblage is as important for reconstructing the ancient environment as it is for understanding spatial and temporal scales

Plants produce an indeterminate number of plant parts throughout their lives The shed parts have the potential to be preserved, but whether or not these are preserved depends on the manner (wind or water) and distance that the parts are transported, the energy conditions under which transport occurs, the suitability or potential for preservation, and burial conditions For example, most leaves or fl owers shed into high-energy environments such as fast fl owing streams are quickly destroyed The leaves of herbaceous species growing on a forest fl oor tend to decompose quickly and have poor preservation potential Plant parts that are woody

or resistant to abrasion such as nuts or woody seed cones can be preserved in energy fl uvial deposits; however, the distance of transport and abrasion encountered during transport will impact the quality of preservation Even woody debris can be destroyed if the transport distance is long and the abrasion encountered during transport is high or the burial conditions are not conducive for preservation (e.g., oxidizing setting)

Alternatively, plant parts preserved in low-energy environments such as wetlands provide a reasonably good archive of the species that grew in and near the wetland

If the rate of organic matter accumulation exceeds the rate of decomposition in such

a wetland environment, then a temporal component to the vegetation history of the wetland also might be preserved In many cases, the anoxic (oxygen poor) and acidic conditions associated with slow-moving to standing water limit the types of fungi and bacteria that decompose organic matter, thus providing ideal conditions for the preservation of plants The acidic conditions are due to organic acid accumulation

Fig 4–12 Middle Eocene age (45 Myr) macrofossils from Napartulik, Axel Heiberg Island,

Nunavut Canada Fig 4 Seed cones of the deciduous conifer Metasequoia occidentalis (dawn redwood) Scale bar = 3 cm Fig 5 Seed cone of the deciduous conifer Larix altoborealis (larch or tamarack) Scale bar = 1 cm Fig 6 Seed cone of Pinus sp (pine) Scale bar = 1 cm Fig 7 A fas- cicle of leaves of L altoborealis Scale bar = 1 cm Fig 8 Nyssa sp (tupelo) fruit Scale bar = 3 mm

Fig 9 Seed cones of Picea sverdrupii (spruce) buried in the channel sand deposits Scale

bar = 20 cm Fig 10 Seed cone of P sverdrupii Scale bar = 1 cm Fig 11 Leaves of Trochodendroides

sp ( t ), Ginkgo sp ( g ), and Nyssa sp ( n ) preserved in a mudstone block These trees grew in a

bot-tomland forest (Fig 20 ) and given the preservation quality of the leaves and the fi ne-grained nature

of the sediment, there was little transport of the leaves prior to burial Scale bar = 2 cm Fig 12 Leaf

of Quercus sp (white oak) in mudstone Scale bar = 2 cm

as organic matter decomposes These examples are over-simpli fi cations of the extremely complex processes associated with transport, burial, and preservation; however, these examples also demonstrate that many variables ultimately determine the type and manner of preservation Interested readers are encouraged to peruse the literature for more detailed information on taphonomy (Burnham 1989, 1990 ; Spicer

2000 ; Gastaldo, 1989, 1999; Gastaldo and Ferguson 1998 ; Gastaldo et al 1998 ; Gee and Gastaldo 2005 ; Burnham et al 2005 ; DiMichele and Gastaldo 2008 ; Vassio

et al 2008 )

Bottomland ( fl oodplain) and especially wetlands such as swamps, fens, bogs, and depressions can provide superb conditions (anoxic, acidic, and low energy) for deposition and preservation of plant remains The remains of ancient swamp and bottomland forest communities have been preserved worldwide (Heer 1868–1883 ; Dorf 1960 ; Smiley and Rember 1985 ; Christophel and Lys 1986 ; Christophel and Greenwood 1987 ; Wolfe and Wehr 1987, 1991 ; Basinger 1991 ; Schaarschmidt

1992 ; Mustoe 2001 ; Vassio et al 2008 ; Erdei et al 2001 ) Such well-preserved plant macrofossils provide tremendous opportunities for paleoecological and plant evolu-tionary research Macrofossils record not only an inventory of the plant species that grew in the area, but they may document signi fi cant changes in relative abundances and frequencies of species with shifts in climate, data that are important to our understanding of plant responses to current and future global climate change Fossil fl oras are commonly used to infer terrestrial paleoclimate One method is based on the climatic tolerances of the living forms; a method called the “nearest living relatives” approach The nearest living relative approach has been applied widely to interpret ancient climate and environments (e.g., MacGinitie 1941 ; Hickey

1977 ; Wing and DiMichele 1992 ) But the utility of the nearest living relative approach diminishes with the increasing age of the fossil remains That is, the fossil remains must be associated with a plausible living relative for the nearest living relative approach to be viable To use this approach, it must be assumed that the physiological requirements and climatic tolerances of the fossil representatives did not change appreciably through geologic time One more recent variant of the near-est living relative approach, the Coexistence Approach, is used to reconstruct the paleoclimate of the Cenozoic by fi nding the modern climate analog for several co-occurring genera in the paleo fl ora ( Mosbrugger and Utescher 1987 ) Another variant on this approach, Overlapping Distribution Analysis, also relies on the co-occurrence of a number of genera in the paleo fl ora and correlation with their modern climate analog (Tiffney 1994 ; Yang et al 2007a, b )

A widely used approach to estimate climatic paleotemperature is based on foliar physiognomy (Wolfe 1993 ; Wilf 1997 ) Nearly 100 years ago, Bailey and Sinnott

percentage of dicot species with leaves possessing entire margins Wolfe ( 1979 ) lished a linear regression of mean annual temperature versus the percentage of dicot species with entire margins for many modern forest communities and later improved the model by using a multivariate approach called Climate-Leaf Analysis Multi-variate Program (CLAMP) that includes 31 morphological characters (Wolfe 1993 )

estab-The foliar physiognomy approach has been used extensively for determining Late Mesozoic (99.6 to 65.5 Myr) and Cenozoic (65.5 to 2.588 Myr) paleotempera-tures Wilf ( 1997 ) later demonstrated that the paleotemperature signal is expressed primarily by the character of the leaf-margin alone and suggested using a univari-ate, rather than a multivariate approach Recently, some studies have demonstrated the value of a multivariate method using digitally manipulated and measured leaves to provide reliable (repeatable) measures of continuous, rather than cate-gorical variables such as tooth area and the ratio of tooth area:leaf perimeter

(Royer et al 2005 )

Generally, linear regressions and multivariate approaches for estimating past means of annual temperatures or mean annual ranges of temperatures have not been reliable for tropical paleo fl oras, most likely because the ecophysiology of plants with toothed leaves (non-entire margins) in the tropics differs from those growing in the temperate and boreal regions (Jacobs 1999, 2002; Burnham et al 2001 ) Nevertheless, rainfall amount is related to leaf area in modern plant communities, and this is a signi fi cant variable with regard to the estimation of past rainfall from fossil leaf assemblages, especially at low latitudes (Hall and Swaine 1981 ; Richards

1996 ; Wilf et al 1998 ; Jacobs 1999, 2002 )

Ancient atmospheric conditions such as the partial pressure of atmospheric CO 2

( p CO 2 ) can be estimated using fossil leaves Contemporary studies document that

p CO 2 is inversely correlated with the leaf stomatal indices of most vascular plant species (Woodward 1987 ; Woodward and Bazzaz 1988 ; Royer 2001, 2003 ; for

exceptions, see Haworth et al 2010 ) The stomatal index is the percentage of dermal cells in a given area that are recognized as guard cells, and stomata (open-ings) relative to non-stomatal epidermal cells The inverse relationship between

p CO 2 and stomatal index helps species to maximize the amount of carbon fi xed per

unit of water transpired (lost) When p CO 2 is high, the plant needs fewer leaf mata to sequester carbon, because the exchange can occur via simple diffusion When the p CO 2 is low more stomata are required The statistical relationship

sto-between stomatal index and p CO 2 for a particular species is calibrated using

her-barium samples and historical records of p CO 2

The inverse relationship of stomatal index and p CO 2 gives insight into the nature

of vegetation change and atmospheric composition over time By correlating the characteristics of a fossil assemblage (e.g., composition, structure, productivity)

with p CO 2 estimates over time, scientists can understand better the relationship

between species and the atmospheric composition Doria et al ( 2011 ) measured the

stomatal index of middle to late Eocene (42 to 37.2 Myr) leaves of Metasequoia

occidentalis (dawn redwood) from Northern Canada (ca 62°N paleolatitude)

Despite an estimated drop from 700 to 1,000 ppm to 450 ppm in atmospheric p CO 2 during the late middle Eocene, the composition of the vegetation did not change,

and high-latitude Metasequoia -dominated deciduous forests were not impacted by

rapid (10 4 years) changes in atmospheric p CO 2 These days, as the global CO 2 centration in the atmosphere continues to increase, an understanding of past vegeta-tion responses to changing CO 2 levels may help us predict how the vegetation will respond and sequester CO on a global scale

5 Palynology

Palynology is the study of plant spores and pollen grains (also called palynomorphs) (Fig 13 ) Pollen are the reproductive propagules of seed plants, while spores are reproductive units produced by the non-seed plants, which include algae, fungi, bacteria, mosses, hornworts, liverworts, lycopods, horsetails, whisk ferns, and ferns The cell walls of pollen and spores are composed of strongly bonded polymers, which make them extremely resistant to degradation in non-oxidizing environments, burial, and the process of preservation These cell walls are even resistant to the strong acids and bases, which are used to extract them from sedimen-tary rock Palynology has been the primary technique employed to document vegeta-tion response to past environmental change because of the resistance of pollens and spores to decay, and their ubiquity and abundance (Traverse 2008 )

Wetlands are excellent sources of pollen and spores and like macrofossils, palynomorph assemblages provide information useful in the reconstruction of past environments Palynomorphs are likely to disperse farther than plant macro-fossils because of their small size and thus more often provide environmental information at the regional, rather than at the local scale Nevertheless, the spatial resolution of the pollen fl ora is strongly in fl uenced by the size and nature of the depositional setting (e.g., lake versus bog) and the relevant source area (Sugita

where plant macrofossils are not, thereby providing another potential source of

Fig 13 Photomicrograph of a typical palynomorph preparation from a Holocene (~3,200 years)

peat near Lake Hovsgol, Mongolia Palynomorphs have been stained red with Safranin-O Note the

bisaccate pollen ( b ), pteridophyte spore ( s ), and scattered wood fi bers ( w ) (see Taddei et al 2011 for details)

data Palynomorphs and macrofossils can be used together if these are both present

to document shifts in the composition of vegetation due to biotic (biological) and abiotic (physical) processes

As is the case with most fossil plant remains, younger deposits can provide data

at biological scales of tens to hundreds to thousands of years For example, a typical sampling strategy for Quaternary (the last 2.588 Myr) lake deposits is to collect samples at roughly 100-year intervals ( Willis and Bennett 2001 ) If the deposits are less than 40,000 years old and contain plant remains (e.g., seeds, twigs, wood frag-ments), then the deposit may allow documentation of a series of radiocarbon ( 14 C) ages for the sediments, thereby permitting interpretation of palynological samples, which at high resolutions record biological succession and responses of vegetation

to climate changes in the context of absolute time Nevertheless, as is the case with all deposits where absolute age controls are not present, deposition is assumed to be continuous and the sediment accumulation rate to be constant

The addition of other sampling locations laterally within the same deposit vides the ability to assess the vegetation and changes at the local and/or regional landscape level By correlating 14 C ages throughout the section or some other dis-tinct feature preserved in the sediment (e.g., caliche, colored layers, and ash beds), the composition and structure of the vegetation can be interpreted in space and in time Such compositional differences can be interpreted in light of the geomorpho-logical (landscape) variation, environmental setting, or biological processes For

pro-example, Hayashi et al ( 2010 ) were able to show that the species growing around Lake Biwa, Japan, were strongly affected by long-term changes in seasonal tem-perature extremes (e.g., winter minima and summer maxima), which were driven by changes in solar insolation (measure of solar radiation energy expressed as watts per square meter (W m −2 ) received on a given surface area) over the last 150,000 years Jackson and Booth ( 2002 ) documented plant species migrations and the changing nature of community structure during the late Holocene at a resolution of 50 years within the context of millennial-scale climate change Analyses of this type are numerous and facilitate reconstruction of the local and regional vegetation, provid-ing scientists with an increased level of con fi dence in their reconstructions

Although similar spatial and temporal data can be collected from peat, brown coal, lignite (coal), and lake deposits that are millions of years old, the temporal resolution is generally more dif fi cult to ascertain Contributing factors include inconsistent rates of sediment accumulation, periods of erosion or no sediment accumulation, and lack of suitable materials (e.g., single mineral crystals for U-Pb,

40 Ar/ 39 Ar) for geochronological or absolute age dating In most cases the samples from older deposits are collected at a much coarser resolution (due to sedimentary compaction) and the fl oral assemblage is clearly averaged over an interval of time Older deposits lacking absolute age controls are commonly correlated with depos-its, which have absolute age controls The age of fossil fl ora without absolute age control is then considered to be a relative age date While it is usually not possible

to obtain suitable resolution for processes such as succession at biological time scales (i.e., tens to hundreds of years) for sediments that are millions of years old, the local and regional patterns of vegetation change can still be interpreted in the context of climate and environmental change

From the standpoint of interpreting future climate change, the use of pollen and spores provides scientists with the greatest amount of data given that most sedimen-tary deposits contain pollen and spores Younger deposits have a better potential for interpreting vegetation change related to climate change effects The species pre-served in younger deposits can provide more accurate reconstructions of the cli-matic conditions than fossil species that are tens of millions years old Younger deposits are comprised of species that may not have evolved so that their physiolog-ical processes and climatic tolerances likely are similar to their living counterparts Nevertheless, used in combination with sedimentological and macrofossil analyses, pollen analyses are an even more powerful tool

6 Wood

Although fossil wood is often a component of fossil assemblages, it is an utilized source of information for reconstructing regional biodiversity, paleoenvi-ronment, and paleotemperature (Wheeler and Bass 1991, 1993 ; Wieman et al

contain in situ stumps and logs has reinvigorated the study of fossil wood and emphasized its importance to paleoecology In situ fossil forests that range in age

from the Holocene (11,700 years before AD 2,000) to the Carboniferous (359.2 to

299 Myr) provide a wealth of information including forest biodiversity, structure, biomass, productivity, environmental setting, paleoclimate, water-use ef fi ciency, and plant-fungal and plant-insect interactions (Figs 14–18; Jefferson 1982 ; Francis 1984, 1988, 1991 ; Creber and Chaloner 1985 ; Creber 1990 ; Taylor and Osborn 1992 ; Scott and Calder 1994 ; Pole 1999 ; Falcon-Lang and Cantrill 2000 ; Poole 2000 ; Labandeira et al 2001 ; Jagels and Day 2003 ; Williams et al 2003a,

2008 ; Akkemk et al 2009 )

One of the bene fi ts of working with well preserved in situ tree stumps and logs is

the amount and quality of the information preserved in the wood The stumps and stems generally provide suf fi cient information for genus-level identi fi cation, while the distribution of the stumps provides information on tree density and size-class distribution The logs provide information on tree size, taper, branching, vertical for-est structure, and stand dynamics The treetops provide a proxy of the live branches and foliage contained within the tree Collectively, these features provide details that can be used to reconstruct stand structure, tree height, stem volume, forest biomass,

and annual net primary productivity (Williams et al 2003a ) The methods used to calculate the values of these parameters are consistent with the well-known concepts

of modern, quantitative forest science (Whittaker and Woodwell 1968 ; Whittaker

et al 1975 ; Vann et al 1998 ; Arthur et al 2001 ; Williams et al 2003a )

Of these parameters forest biomass and annual net primary productivity are haps the most important for understanding the original climate and environmental conditions of the fossil species Forest biomass is the combined mass of the wood,

per-roots, and leaves, while the annual net primary productivity is the weight of wood, root, and leaves produced annually Both measurements are directly related to the amount of heat and water received by the vegetation (Whittaker 1975 ; Knapp and Smith 2001 ) Climate and carbon fl ux are closely coupled, and annual net primary productivity is directly related to the amount of energy (temperature) and water received (Whittaker et al 1975 ) Modern forests growing in colder or drier climates have considerably lower annual net primary productivity rates than those growing

Fig 14–18 Middle Eocene age (45 Myr) wood from Napartulik, Axel Heiberg Island, Nunavut

Canada Fig 14 In situ stump of Metasequoia occidentalis , which is approximately 60 cm in eter Fig 15 Excavated stem of M occidentalis from one of the fossil forest layers Fig 16

diam-Photograph of the upper portion of an M occidentalis stem that grew in the forest canopy Note the

meter stick in top right of the image for scale Fig 17 Photograph of a split M occidentalis stem

illustrating a buried branch This tree once produced branches basally, but as the forest canopy closed the light levels were reduced to the point where the tree could no longer sustain growth and self-pruned This information is useful for reconstructing forest tree canopy and tree life stage at the time of death (e.g., trees have branches on lower part of the trunk in younger stages) Scale

bar = 2 cm Fig 18 Photograph showing an approximately 3 m tall tree stem that once grew in a

bottomland forest and was buried during a major fl ood The entire center of the tree is hollow and

fi lled with sediment suggesting that the tree was hollow and probably dead at the time that it was buried Furthermore, it illustrates that under certain conditions large thicknesses of sediment can accumulate in a short period of time Note the 1 m long shovel for scale

in the wet tropical regions (e.g., 6.5 vs 29 Mg ha −1 ; Rodin et al 1975 ) Therefore, if the annual net primary productivity of modern and fossil forests can be determined, then the climatic conditions under which these forested wetlands grew can also be inferred (Woodward et al 1995 )

7 Geochemistry

Understanding the chemical composition of ancient atmospheres using try is important to reconstruct paleoenvironments Geochemistry is the study of the distribution of chemical elements and natural compounds on the Earth Geochemical approaches used in the study of plant fossils help determine the original chemical composition, deposition, burial, and thermal maturity of the fossil tissues, as well as the nature of chemical transformations in the paleoenvironment (van Bergen 1999 ) Studies aimed at better understanding the chemical processes associated with the preservation of plant fossils and the use of chemical techniques to free these fossils from rock can be traced back more than 150 years (Heer 1868–1883 ; Traverse

geochemis-2008) More recently, geochemical techniques using stable isotopes have been developed to determine paleoatmospheric conditions (Arens and Jahren 2000 ; Jahren and Sternberg 2008 ) Carbon stable isotopes in plant cellulose in peat have been utilized to reconstruct atmospheric CO 2 concentrations in the Quaternary

(2.588 Myr ago to the present) (White et al 1994 ) Others have utilized stable carbon, oxygen, and hydrogen isotopes of preserved plant tissues to infer shifts in

wetland hydrology across various time scales (Xie et al 2004 ; Yang et al 2005,

Csank et al 2011 ) Such geochemical techniques are often best utilized when paired with other proxies for paleoenvironmental reconstruction (Leng 2006 and papers

therein; Jones et al 2010 ; Markel et al 2010 )

8 Sedimentology

Sedimentary rocks are residues of older igneous (volcanic), metamorphic, and mentary rocks, which have been broken down by mechanical forces or weathering and transported by water, ice, wind, and/or gravity into a depositional basin (Fig 19 ) Understanding the processes associated with the transport and deposition of the rock particles and the manner in which the transported material accumulates pro-vides a wealth of information on depositional environment and climate For exam-ple, peat and coal accumulate in low-energy environments where water and vegetation are abundant and the rate of organic matter accumulation is generally greater than the rate of decomposition External factors such as subsidence (where the land surface becomes depressed or sinks) or faulting contribute to more rapid accumulation and formation of organic-rich deposits For this discussion, we focus

Fig 19 An idealized model depicting a watershed basin with a broad depositional continuum,

which ranges from high- to low-energy environments Plant remains from the uplands are

trans-ported into the fl oodplains/bottomlands ( F ) lakes, and swamps ( S ) by fl uvial systems, which have

varied water fl ow velocities and energy conditions Plant fossils that are preserved in the channel sands of the streams and mudstones of the fl oodplains/bottomlands are generally allochthonous and parautochthonous, although autochthonous remains are not uncommon The swamps were low-energy environments where transport was limited In most cases, the organic-rich deposits accumulated on the forest fl oor of these wetland ecosystems

S S S

F

F F

In most cases the alluvial (loose and unconsolidated) sediment deposited in

fl oodplains and swamps is transported by streams and rivers ( fl uvial) and the mation contained in these sediments is important for several reasons First, alluvial sediments provide insight into the tectonic (study of the position and movement of the Earth’s plates) setting; pulses of alluvial sedimentation generally occur in response to tectonic uplift (i.e., upward movement of plates associated with moun-tain building) or sea-level change (Miall 1994 ) Water velocity and depth, particle sorting mechanisms, and depositional environment are determined by the textural (grain size and shape, composition, and maturity) and structural (strati fi cation, bedding surfaces, and internal character of the beds) characteristics of the deposits Finally, the analyses of stratigraphic sections facilitate the recognition of the depo-sitional environments under which the various sedimentary or lithologic (rock) units formed Such analyses also provide a record of shifting sediment transport and deposition, and environmental conditions through time within the basin The character of the sedimentary units (e.g., conglomerate, sandstone, siltstone, mudstone, and coal) within a depositional basin is a re fl ection of the physical set-ting under which the sediment was deposited (Fig 20 ) Physical changes such as a drop in the sea-level, mountain uplift, stream meandering, stream channel avulsion

(formation of a new stream channel and abandonment of the old channel), and the formation of crevasse splays (a sedimentary deposit that forms in response to a breach of a stream channel or levee) are common in foreland basins and account for the stratigraphic deposition patterns observed (Fig 20 ) At a fi ner scale, the textural and structural features of the preserved sediments provide information on water velocity, direction of transport, as well as local and regional changes in the depositional environment (e.g., Davies-Vollum and Wing 1998 ) For example, the uppermost conglomerate (sand, coarse gravel, and large rocks/boulders) deposit indicates a rapid shift from low energy to much higher energy conditions and in a foreland basin setting, a shift from fi ner- to coarse-grained sediment usually indi-cates a period of mountain uplift or major faulting (movement between plates) (Fig 20 )

9 Early Cenozoic of the High-Latitudes

Although scientists agree that current global climate change is real, understanding the complex mechanisms responsible for the outcomes of such unprecedented rapid warming is still in its infancy While considerable time and effort is being focused

on global environmental change and its effects on biodiversity in the tropical regions, the polar regions (particularly the Arctic) stand to suffer the greatest changes due to polar ampli fi cation of global temperature change (Holland and Bitz 2003 ; Peacock

et al 2011 ) Since 1980, the polar regions have experienced the most rapid warming

on the planet of about 1°C per decade (Anisimov et al 2007 ) , and global climate models predict a 7–8°C warming in the high latitudes over the next 90 years

(Solomon et al 2007 ) Enhanced polar warming is predicted to lead to enhanced global warming due to increased carbon inputs, degradation of the current biodiver-sity, and the restructuring of Arctic communities and ecosystems (Arctic Climate Impact Assessment 2005 ; Schuur et al 2008 ) Of the myriad of questions that have been posed, perhaps the most fundamental to the climate change discussion is understanding of the structure and function of Arctic wetlands under global hot-house climate conditions The lack of modern analogues (i.e., forested Arctic) as well as our lack of historic experience with the changes associated with a transition

Fig 20 An idealized stratigraphic column showing the typical stratigraphy associated with a

fore-land basin (e.g., sediment basins that formed east of the Rocky Mountains in Montana, Wyoming, and Colorado in response to the uplift or formation of the Rocky Mountains) Fluvial, bottomland/ fl oodplain, and swamp environments are generally present in foreland basins The sandy siltstone and siltstone/mudstone deposits are typical for bottomland/ fl oodplain environ- ments The coal layers are the metamorphosed remnants of the organic-rich accumulations present

in swamp/wetland settings The sandstone ( SS ) and pebble lag deposits are consistent with those

found in fl uvial environments The trough and planar cross-bedding, ripple marks, and pebble lag are structural features of sandstone deposits that provide insight into the deposition environment

from icehouse to hothouse conditions makes it dif fi cult to predict accurately ing changes at several levels Nevertheless, the plant fossil record from the Arctic provides insight into the complexities of how Arctic wetland ecosystems functioned before the last ice age

Throughout the early Cenozoic (65.5 to 33.9 Myr), circumpolar Taxodiaceae (redwood)-dominated wetland and bottomland forests occupied the high latitudes

of the Northern Hemisphere (Heer 1868–1883 ; LePage et al 2005 ; LePage 2007 ) For one such well-studied middle Eocene (48.6 to 40.4 Myr; Eberle and Storer

1999 ; Harrison et al 1999 ) fossil forest at Napartulik (Inuk name for “place of trees”), Axel Heiberg Island, Nunavut, Canada (Fig 21 ), the structure, biomass,

Fig 21 Map showing the location of Napartulik (80°N) on Axel Heiberg Island, Nunavut, Canada

80 N

Lake Hazen

Meighen

Island

Napartulik

Axel Heiberg Island

Eureka Fosheim Peninsula

species composition, productivity, and stand development history have been mined (e.g., Basinger 1991 ; Francis 1991 ; Greenwood and Basinger 1993, 1994 ; McIver and Basinger 1999 ; LePage 2001, 2003a ; Jagels et al 2001, 2005 ; Williams

deter-et al 2003a, b ) This fossil forest grew at a paleolatitude of 78°N (Irving and Wynne

1991 ) , and was part of a much broader regional vegetation mosaic of ecological communities including upland, fl uvial, bottomland, and swamp forests The sandstone deposits are the remnants of ancient fl uvial systems that bisected the bottomland and swamp forests (Fig 22 ) Abrasion-resistant organs such as seed cones, nuts, and pieces of wood of the plants are commonly preserved in sand (Figs 8–10) The siltstone and mudstone deposits are the forest fl oors of the bottomland forests Given that transport in the fl oodplain was limited, a wider diversity of leaves, seeds, and fruits was preserved in the fl oodplain and bottomland forest deposits (Fig 23 ) The remains of the swamp-forest communities are the most spectacular and include

in situ stumps and logs and autochthonous leaf litter mats (Figs 4–10, 14–18 , 24 , 25 ) Based on the fl oristic composition (Table 1 ), Vann et al ( 2003 ) suggest the Napartulik

fl ora most closely resembled that of a modern Pennsylvanian fl ora

Detailed measurements of the dimensions of the fossil tree stems, stumps, branch stubs, treetops, annual ring widths, and height helped to determine species composi-tion (Williams et al 2003a, b ) Wood anatomy indicated that the swamp forests

were composed nearly entirely of Metasequoia The diameters and parabolic shape

Fig 22 A view of one of the numerous channel sand deposits that cross-cut the fossil forests and

fl oodplain/bottomland sediments at Napartulik These high-energy deposits provide information

on water velocity and depth, sorting mechanisms, depositional environment, and structural teristics of the depositional setting

Fig 23 Laterally extensive bottomland forest deposits showing vertical accretion of the fl oodplains

was occurring along with changes in the regional landscape Although there is no evidence of a complete turnover in the bottomland forest composition at Napartulik, the local environmental changes would have likely had an impact on the frequency and abundance of the local fl ora, providing the region with a fl oral mosaic

Fig 24 View of the fossil forests at Napartulik The dark bands are layers of coal and peat that are

the remains of swamp forest communities and include extraordinarily preserved (mummi fi ed)

plant macrofossils In situ stumps and stems are also preserved in many of these fossil layers

of the stumps and stems indicated that the best preserved forests were relatively of the same age and canopy height (40 ± 3 m) The fossil stems were generally free of protruding branches and the uppermost 9 m of the trees had branches with foliage The aboveground biomass and annual net primary productivity were 684 and 5.5 Mg ha −1 (wood plus foliage), respectively This was a very-high biomass forest, which was similar in stature and total biomass to the modern old-growth forests of the Paci fi c Northwest (Grier and Logan 1977 ; Gholz 1982 ) and coastal Cordillera

forests of southern Chile (Johnson et al 2000 ; Battles et al 2002 ) , whereas the annual net primary productivity was similar to that seen in modern cool-temperate deciduous forests (Johnson and Lindberg 1992 )

Clearly climate in the polar regions was favorable for forest growth These early Cenozoic Arctic forests probably represent the maximum forest productivity and biomass achievable near the poles, given the size of the fossil trees and geographic extent of the redwood-dominated late Paleocene and middle Eocene fossil forests (58.7 to 55.8 Myr and 48.6 to 40.4 Myr), respectively (Heer 1868–1883 ; Nathorst

1915 ; Schloemer-Jäger 1958 ; Koch 1963 ; Manum 1962 ; Schweitzer 1974 ; Wolfe

and Basinger 1999 ; Liu and Basinger 2000 ; LePage et al 2005 ; LePage 2007 ) Temperature, water availability, and light regime would be primary determinants

of lush plant growth in the Arctic under hothouse conditions, and these forests may

be of key importance to understanding the evolution and development of future

Fig 25 A lateral view of one of the fossil forest layers showing the in situ stumps, many of which

are up to 1 m in diameter

Table 1 Middle Eocene Napartulik Flora, Axel Heiberg Island, Nunavut, Canada

Macrofossils

Ginkgo sp Metasequoia occidentalis

Chamaecyparis eureka Kotyk Glyptostrobus europaeus

Abies sp Taiwania cf cryptomeroides

Larix altoborealis LePage et Basinger Platanus sp

Picea heibergii LePage Aceraceae

Picea nansenii LePage Acer sp

Picea palustris LePage Sparganiaceae

Pseudolarix wehrii Gooch Fabaceae

Pseudolarix amabilis (Nelson) Rehder unidenti fi ed

Tsuga swedaea LePage Cecidiphyllaceae

Carya veripites Wilson et Webster Caprifoliaceae

C viridi fl uminipites Wodehouse Diervilla sp

2007, 2009 ) , Kotyk et al ( 2003 ) , LePage et al ( 2005 ) , Jagels et al ( 2001, 2005 )

The question mark ( ? ) indicates the identi fi cation is tentative

Arctic wetland ecosystems Nearest living relative analyses of the Napartulik fl ora provide an estimate of mean annual temperature of 12–15°C, warm month mean of

>25°C, and a cold month mean of 0–4°C (Basinger et al 1994 ) These are tent with other estimates from Paleocene and Eocene fl oras from Ellesmere Island, Greenland, and Spitsbergen (McKenna 1980 ; Estes and Hutchison 1980 ; Axelrod

consis-1984 ; Koch 1963 ; Schweitzer 1980 ; Spicer and Parrish 1990 ; Boyd 1990 ) Climate estimates based on foliar physiognomy indicate a slightly colder paleoclimate com-pared to the nearest living relative approach For example, the physiognomic signa-ture of the Napartulik fl ora indicates a mean annual temperature of 9.3°C ± 2.0°C, a mean annual range of temperature of 13.8°C, and cold month mean temperatures of

−0.9°C ± 3.6°C (Basinger et al 1994 ; Greenwood and Wing 1995 ; Greenwood et al

2010 ) More recent stable isotope analyses of fossil animal remains provide rable early Eocene mean annual temperature and cold month mean temperature

compa-estimates of 8°C and 0–3.5°C, respectively (Eberle et al 2010 )

Whereas growing season temperatures may dictate rates of plant growth and productivity, winter temperatures and the high-latitude light regime may have been equally important in determining the success of certain species in the Arctic wet-lands During the Paleocene and Eocene, evergreen and semi-evergreen conifers

such as Cathaya (yin shan), Chamaecyparis (false cypress), Glyptostrobus (Chinese swamp cypress), Picea (spruce), Pinus (pine), Taiwania (Taiwania) , Thuja (cedar), and Tsuga (hemlock) grew at latitudes well above the Arctic Circle If we consider

the foliar respiration demands of these conifers, then constraints on winter ture must be considered (LePage 2003b ) If the winter temperatures were warm, then the respiratory carbon losses of overwintering foliage of evergreen taxa may have been substantial throughout the dark winter months In these conditions, these species neither had the capacity to store the amount of carbohydrates needed to survive a minimum of three months of complete winter darkness, nor to produce the carbohydrates they would have required via photosynthesis due to the lack of light (LePage 2003b ) Alternatively, under sub-freezing conditions, foliar respiration of the evergreen taxa would have “shut-down” metabolically throughout the dark win-ter months without depleting their limited stored carbohydrate reserves (LePage

sug-gests that the maintenance respiration burden may not have been as substantial as previously thought in overwintering polar evergreen taxa

The question of the polar light regime becomes intriguing when vegetation response to light quality is considered It is assumed that as Earth warms, vegetation zones will shift to the north (Arctic Climate Impact Assessment 2005 ) Although over-simpli fi ed, this assumption is probably correct, at least for most of the current vegetation zones As plants migrate northward they will be subject to changes in the duration, quantity, and quality of light Above the Arctic Circle there is a 3-month summer period during which the sun never drops below the horizon The plant fossil record indicates that despite a decrease in the amount of photosynthetically active radiation (PAR) and lower angle incident light as one moves farther to the north (e.g., Jagels and Day 2003 ; Vann et al 2003 ; Taulavuori et al 2010a ) , vegetative success was not precluded during the early Cenozoic

We could also ask whether the light regime imparted selective forces on the composition of the vegetation of the polar regions Plants respond to day-length, light quantity, and light quality, all of which are important for the successful timing

of germination, growth, fl owering, frost hardening, and dormancy These responses are particularly important in areas where the growing season is determined not only

by temperature, but also by the availability and quality of light Experiments ducted by Downs ( 1962 ) on the photocontrol of growth and dormancy in woody plants documented progressively greater growth and fresh weight of seedlings of

Picea glauca (white spruce), Larix kaempferi (Japanese larch), Pseudotsuga

men-ziesii (Douglas fi r), and Pinus radiata (Monterey pine) when exposed to different

light treatments including a 24-h photoperiod for up to 1 year Taulavuori et al

Kuhmo (64°N) and Ranua (66°N), Finland, that were transplanted 1°, 2°, and 3° latitude higher than their current distribution showed no detrimental effects on growth Nevertheless, of the seedlings tested by Downs ( 1962 ) Sequoia sempervi-

rens (coast redwood) did not respond similarly The redwood seedlings showed a

progressive increase in growth and fresh weight when exposed to 12-, 14-, and 16-h photoperiods, but then showed marked decreases in these parameters when exposed

to 20- and 24-h photoperiods Results such as these indicate some of the species currently living in the temperate and boreal zones may not be well adapted to the light regime above the Arctic Circle, which could act as a barrier to expansion or migration into the high latitudes

In an effort to understand better the early Cenozoic distribution of Metasequoia above the Arctic Circle, Vann et al ( 2003 ) performed gas exchange measurements

on living M glyptostroboides Their results indicated CO 2 uptake rates in response

to temperature were broad and maximum uptake values were observed between 15°C and 25°C, with an optimum temperature for photosynthesis of 20°C (Fig 26a ) The CO 2 uptake rate dropped rapidly above 25°C indicating a photo-inhibitory response, and damage to the photosynthetic apparatus was seen at temperatures above 40°C The CO 2 uptake rate was rapid at low light intensities Saturation was achieved at a light intensity of about 700 m E, which is less than one half of the amount of light currently received on a clear day (1,500–1,700 m E) in July at Napartulik (Fig 26b ; Vann et al 2003 ) It is worth noting that at 150 m E the leaves were able to maintain a 50% photosynthesis saturation rate, which means that even under the cloudy conditions experienced at Napartulik (300–500 m E), suf fi cient

light would have been available for Metasequoia to grow These experiments and

others (e.g., Jagels and Day 2003 ; Royer et al 2003 ) suggest that some plant species could thrive in a polar light regime Some of these species may be pre-adapted to such conditions, but responses such as frost hardiness and leaf senescence are still poorly understood

Two factors that would determine whether the vegetation of the polar regions will have a boreal or temperate character during climate change conditions is winter temperature and the geographic extent of wetland communities in the mid-

to high-latitudes of the Northern Hemisphere If the winter temperatures of the mid- to high-latitudes were around −20°C as suggested by global climate models

(Sloan and Pollard 1998 ) , then forests with boreal characteristics may develop there with low mean annual temperatures and short growing seasons The resulting wetlands and lowland forests would probably be similar in structure and possibly composition to the boreal ecosystems in current high-latitudes and tree-lines Alternatively, if the average winter temperature increases to 0°C with future cli-mate warming, then the growing season would be extended and more temperate communities could be expected to develop in the Arctic Nevertheless, the mecha-nisms warming the polar regions throughout the early Cenozoic are still poorly understood Forcing mechanisms such as increased ocean heat transport and up to

ten times of the present-day pre-industrial levels of p CO 2 have been proposed (Berner 1990 ; Otto-Bleisner and Upchurch 1997 ; Upchurch et al 1998, 1999 ; Pearson and Palmer 2000 ; Retallack 2001 ; Royer et al 2001 ) Warming the polar regions and Arctic Ocean using either mechanism creates positive feedbacks that impact climate models (Otto-Bleisner and Upchurch 1997; Upchurch et al 1998, 1999 ) Although Sloan and Rea ( 1995 ) have shown that six times of the modern (pre-industrial)

Fig 26 Carbon dioxide exchange in leaves of living Metasequoia glyptostroboides in response to

various environmental parameters ( a ) Net photosynthetic response to temperature ( b ) Response

to light intensity measured as photosynthetically active radiation (PAR; m mol photons m −2 s −1 )

(Redrawn and modi fi ed after Vann et al 2003 )

45 40 35

25 30

15 20 10

5 0

p CO 2 levels during the Eocene (55.8 to 33.9 Myr) could raise the mean annual temperature in the polar regions to about 0°C, the global change models also indi-cate that these elevated CO 2 levels would raise the temperature in the tropics consid-erably Given that there is limited evidence (Huber 2008 ) in the tropics for such

elevated temperatures at this time, the idea of elevated p CO 2 alone does not account for the fact that temperate and boreal forests occupied the Arctic regions from the Cretaceous to the Pliocene

The global change models consistently estimate early Cenozoic mean annual temperatures in the Arctic to be approximately 10–20°C below the temperature estimates derived from fossil fl oras (Huber and Caballero 2011 and references

therein) Given that increased p CO 2 alone was dismissed as a viable climate forcing mechanism for the Arctic (Huber 2008 ) , other mechanisms such as methane (CH 4 ) were considered Methane is a potent greenhouse gas and exerts a strong in fl uence

on the atmosphere and global climate (Cao et al 1998 ) Interestingly, high rates of methane production during the early Cenozoic may have contributed to the forma-tion of polar stratospheric clouds, which would have warmed the troposphere and polar regions during the dark winter months without increasing temperature in the

tropics (Sloan et al 1992, 1999 ; Sloan and Pollard 1998 ) Modeling studies indicate that polar stratospheric clouds could increase surface temperatures in the Arctic during winter months by as much as 20°C (Sloan and Pollard 1998 ) Consideration

of polar stratospheric clouds as a viable forcing mechanism for warming the polar regions required an input of methane emissions of seven to ten times of pre-industrial levels for the global climate model to generate a scenario where the polar regions were warmed 10–20°C (Sloan et al 1999 ) To obtain the seven to ten times increase

in methane emissions Sloan et al ( 1999 ) suggested up to three times the area of wetlands that exist today would be needed Natural and manmade wetlands are important sources of methane, and our conservative estimate of doubling of the total wetland area in the future could provide the wetland area needed to generate the methane required to produce suf fi cient amounts needed to validate the notion that polar stratospheric clouds were responsible for keeping the Arctic warm during

the early Cenozoic (Cao et al 1998 )

The importance of this warming mechanism for winter warming of the polar regions in the future rests with suf fi cient wetland area to produce the amount of methane needed for polar stratospheric clouds to warm the Arctic, but also on the composition and quality of the organic matter (leaves) to support methanogenic bacteria The annual turnover of leaf biomass of the bottomland and swamp forests

in the mid- to high-latitude regions during the early Cenozoic would have been the primary source of carbon for methanogenesis and the chemistry of the foliage prob-ably played a key role in controlling methane production Lignin content, labile C content of the plant litter (cellulose, soluble carbohydrates), lignin:N ratio, and C:N ratio of leaf litters are good predictors of decomposability (Mellilo et al 1982 ; Ågren and Bosatta 1987, 1996 ; Ryan et al 1990 ; Berg and Tamm 1991 ; Bryant

et al 1998) Initial results from plantation-grown trees indicate that leaves of

Metasequoia are a better source of organic substrate than those of Larix (larch or

tamarack) (C Williams, unpublished)

Thus, if the total wetland area were to double in the future and the majority of these mid- to high-latitude wetlands were dominated by representatives of the Pinaceae (pine family) as is the case today in the boreal wetlands, it is unlikely that suf fi cient methane would be produced from the less decomposable litter Alternatively, if the vegetation growing in these wetlands were more temperate in character, the leaves would decompose more rapidly and methane production would increase In this scenario the increased methane production would contribute to the formation of polar stratospheric clouds, which would warm the polar regions and support more temperate vegetation In effect, the vegetation would be sustaining itself by driving climate in the high-latitudes

Therefore, if we are to understand how the polar regions and the rest of the world will respond to global warming in the coming decades, clearly we need to appreci-ate the complex inter-relationships and feedbacks among vegetation, microbes, bio-geochemical processes, and climates associated with the early Cenozoic wetlands at high latitudes

10 Pleistocene Holocene

The Pleistocene and Holocene Epochs are the most recent in geologic time, and comprise the Quaternary Period from 2.588 Myr ago to the present After debate by the International Stratigraphic Commission, the beginning of the Pleistocene now coincides with the base of the Quaternary ( Gibbard et al 2009 ) The Holocene

begins at the end of the Pleistocene (Walker et al 2009 ) , and currently there is debate regarding whether the Commission should establish the Anthropocene Epoch, to represent human impacts The age dates used to mark the geological boundaries in the of fi cial International Commission on Stratigraphy time scale are determined by the coincidence of signi fi cant biotic, stratigraphic, geochemical, and other physical, chemical, or biological changes at the proposed boundary The abil-ity to see the boundary in a sedimentary sequence as well as obtaining an absolute age date facilitates the process Although substantial glaciation began at the South

Pole at least 33 Myr ago (Zachos et al 2001 ) , the geological hallmarks of the Pleistocene re fl ect major fl uctuations in continental glaciers in both the Northern and Southern Hemispheres The Holocene represents the epoch after the last large-

scale continental glaciation (Walker et al 2009 )

The apex of the most recent glacial expansion, approximately 27,000 to 24,000 years ago is called the Last Glacial Maximum, and Northern Hemisphere glaciers extended south to about 40°N in North America, as far south as modern day

Ohio (Dyke et al 2002 ) Mountain or alpine glaciers existed still farther south in the Rocky Mountains and mountains of California The massive expanse of the conti-nental ice sheet that covered parts of the Northern Hemisphere had a tremendous impact on regional and global climate, and the relatively recent recession of the con-tinental ice sheets provides an opportunity to document more recent, but signi fi cant climate change in some detail Sediment and fossils preserved in cores from lakes

and oceans provide evidence of past biota, and ice cores taken from the remaining ice sheets from Antarctica and Greenland provide proxies of temperature, precipita-tion, and paleoatmospheric composition Even subtle climate and biotic changes can

be recorded spatially and temporally, documenting several thousand years of

paleo-ecological history over large areas (e.g., Williams et al 2004 ) Because modern humans evolved during the Pleistocene and developed sedentary civilizations during the Holocene, paleoecological records of climate change are relevant to our under-standing of the potential impacts of climate change on human populations

Geochronologies using climate and paleoecological proxies are critically tant, especially where the focus is on brief, but signi fi cant paleoclimatic events and their environmental impacts Where unaltered organic matter younger than about 40,000 years is preserved in wetland deposits, 14 C dating can give very precise time estimates Naturally occurring radioactive 14 C is generated by cosmic ray bombard-ment of carbon atoms in the upper atmosphere and has a half-life of 5,730 years The decay product ( 14 N) accumulates at this set rate as radioactive decay proceeds after a plant or animal dies Advancements in the last 30 years have enabled dating of tiny pieces of organic matter such as individual seeds using a particle accelerator com-bined with a mass spectrometer to measure individual 14 C atoms in a sample (Burr and Jull 2009 ) Samples older than 40,000 years or lacking organic matter must be dated using other radiogenic isotopes, or by using relative age-dating techniques Paleoecological information can be extracted from proxies other than wetland sedi-ments including tree rings and fossil soils (paleosols) Tree ring analyses provide cen-tury to millennial scale records of temperature and precipitation change, including amount and seasonality with a 1-year resolution Dendroclimatology is usually limited

impor-to tree species with ring-width sensitivity impor-to seasonal and yearly variations in rainfall and temperature, and to trees growing in regions with variable climate on an annual basis (i.e., not the low latitudes) Paleosols (ancient soils) can document information about paleoclimate thousands or even millions of years ago, via qualitative assessment, study of the soil’s chemical composition, and through isotope geochemical study of phyllosilicate minerals (minerals that contain silica and oxygen) preserved in the soils, but with less accuracy than other methods (Tabor and Yapp 2005 )

The following section provides case studies from wetlands across the African tropics, where Pleistocene and Holocene climate changes have had profound and sometimes differing impacts At a larger scale, the geologic and fossil records also document differing impacts of global climate change at low and high latitudes; thus, the expectation is that future climate change will result in both wetland expansion and contraction in various regions in Africa

11 The Tropics

The impact of global climate change on the tropics will be determined largely by the rate of warming and magnitude of change in seasonal precipitation, with the latter factor being the most important for the fate of tropical wetlands Climate across the