08071.1

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofitpublishers, academic institutions, research libraries, and research funders in the common goal of maximizing access tocritical research.

A physiologically explicit morphospace for tracheid-basedwater transport in modern and extinct seed plantsAuthor(s): Jonathan P. Wilson and Andrew H. KnollSource: Paleobiology, 36(2):335-355. 2010.Published By: The Paleontological SocietyDOI: http://dx.doi.org/10.1666/08071.1URL: http://www.bioone.org/doi/full/10.1666/08071.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in thebiological, ecological, and environmental sciences. BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies, associations,museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated contentindicates your acceptance of BioOnes Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercialuse. Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder.

A physiologically explicit morphospace for tracheid-based watertransport in modern and extinct seed plants

Jonathan P. Wilson* and Andrew H. Knoll

Abstract.We present a morphometric analysis of water transport cells within a physiologicallyexplicit three-dimensional space. Previous work has shown that cell length, diameter, and pitresistance govern the hydraulic resistance of individual conducting cells; thus, we use these threeparameters as axes for our morphospace. We compare living and extinct plants within this space toinvestigate how patterns of plant conductivity have changed over evolutionary time. Extinctconiferophytes fall within the range of living conifers, despite differences in tracheid-level anatomy.Living cycads, Ginkgo biloba, the Miocene fossil Ginkgo beckii, and extinct cycadeoids overlap with bothconifers and vesselless angiosperms. Three Paleozoic seed plants, however, occur in a portion of themorphospace that no living seed plant occupies. Lyginopteris, Callistophyton, and, especially, Medullosaevolved tracheids with high conductivities similar to those of some vessel-bearing angiosperms. Suchfossils indicate that extinct seed plants evolved a structural and functional diversity of xylemarchitectures broader, in some ways, than the range observable in living seed plants.

Jonathan P. Wilson*. Department of Earth and Planetary Sciences, Harvard University, Cambridge,Massachusetts 02138

Andrew H. Knoll. Department of Organismic and Evolutionary Biology, Harvard University, Cambridge,Massachusetts 02138

*Present address: Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena,California 91125. E-mail: [email protected]

Accepted: 5 October 2009

Introduction

Extant plants, particularly conifers andangiosperms, have evolved two distinct waysto achieve low-resistance water transport intheir vascular systems: the vessel and the torus-margo pit. Vessels are often wide (up to500 mm), long (.1 m), multicellular tubes thatallow water to be transported long distanceswithout flowing through a cell wall. They arecharacteristic of angiosperms, and occur, aswell, in a few ferns and gymnosperms. Thetorus-margo pit, a highly porous valve thatgreatly reduces the hydraulic resistance to flowbetween tracheids, is characteristic of conifersand Ginkgo. Do these strategies exhaust theways that seed plants can reduce flow resis-tance through the xylem, and if not, why dothey predominate among living plants?

To address these questions, we constructeda three-dimensional morphospace of xylemcells in which each axis has a discreteanatomical range, creating a space withinwhich every point has quantitative physiolog-ical meaning. This morphospace provides aframework for the structural and functional

comparison of living and extinct seed plants.We modeled fluid flow in xylem cells of 22living plants and six extinct taxa: the Paleo-zoic seed plants Lyginopteris, Medullosa, Callis-tophyton, and Cordaites, the Mesozoic seedplant Cycadeoidea, and the Miocene Ginkgobeckii. Our conclusion that fundamental dif-ferences exist in the occupation of conductiv-ity space between the Paleozoic and theRecent has important consequences for inter-preting the evolutionary and functional tra-jectories of seed plant wood.

Background

Living seed plants have converged upontwo dominant forms of wood anatomy, basedon differing ways of balancing structuralsupport and high-throughput water trans-port. Angiosperms exhibit a division of laborin the xylem, whereby most cells derivedfrom the vascular cambium fulfill one of twofunctions: structural support, in the form oflong, thin, highly lignified and nonconduct-ing cells called fibers, and transport, in theform of vessels, which are long, wide, multi-cellular conduits.

Paleobiology, 36(2), 2010, pp. 335355

2010 The Paleontological Society. All rights reserved. 0094-8373/10/36020009/$1.00

Conifers, on the other hand, rely on short,narrow tracheids for both water transport andstructural supportachieving increasedthroughput of water by means of highlyporous torus-margo pits. This allows conifertracheids to function effectively in structuralsupport, while attaining per-area xylem con-ductivities comparable to, or exceeding, thoseof some angiosperms when the two woodtypes have equal conduit diameter and length(Pittermann et al. 2005, 2006a,b; Jagels andVisscher 2006).

Broadly speaking, seed plants have pur-sued one or the other of these alternativestrategies throughout their evolutionary his-tory. Paleobotanists routinely distinguish be-tween the dense pycnoxylic wood character-istic of coniferophytes and wood composed oflarge conducting and abundant living cells,termed manoxylic, found in many angio-sperms and early seed plants (Esau 1977;Stewart and Rothwell 1993). Both manoxylicand pycnoxylic woods have evolved withinmultiple lineages (Namboodiri and Beck1968; Beck et al. 1982; Galtier and Meyer-Berthaud 2006). Understanding xylem evolu-tion through time, thus, requires functionalcomparisons, interpreted in the contexts ofstratigraphy and phylogeny.

We use morphospace analysis to facilitatethe functional interpretation of fossil xylemcells. Beginning with seminal work on coilingof nautiloids by Raup (1966, 1967), morpho-spaces have been used to characterize mor-phological variation within and among fossiltaxa. Two kinds of morphospaces have pre-dominated: theoretical spaces that take shapefrom general rules or equations of form, andempirical spaces, which are derived fromspecimen measurements and have the result-ing n-dimensions reduced by using datareduction techniques such as singular valuedecomposition (McGhee 1999). Each ap-proach has its limitations: theoretical mor-phospaces can be limited by the fidelity withwhich their initial rules or equations captureprocesses operating during morphologicalevolution, and statistically derived morpho-spaces lose or gain dimensionality accordingto the number of initial data points, makingthem inherently unstable. Furthermore, these

morphospaces have, to date, largely focusedon describing disparity and inferring para-meters such as development and ecology,rather than embedding function directlywithin the initial data set. Recent work onmultiple origins of leaves in land plants(Boyce and Knoll 2002) has shown thatdevelopment can be embedded directlywithin a morphological data set. We proposeto do the same with physiology.

Plants are an ideal set of organisms to test aphysiological approach to morphospace ana-lysis. First, basic plant life strategies varylittle across the clade. Save for a limiteddiversity of parasites and saprophytes, allliving plants conduct photosynthesis byusing water and carbon dioxide as substrates,and fossil plants worked the same way.Second, because they are stationary, plantsrely on chemical and biophysical means toadapt to their physical and biological envi-ronments. Thus, the ability of plants toconduct water and support above-groundtissues relates directly to the structure ofxylem and other cells in the stem. Finally,because they contain the decay-resistantbiopolymer lignin, plant tissues that conductwater and provide physical support have ahigh potential for preservation in fossils.

We focus on the functional evolution ofxylem cells for several reasons. First, they areabundant in the fossil record, with conduct-ing cells preserved in plants as old as theSilurian (Edwards 2003). By definition, alllineages of vascular plants inherited theirxylem cells from a common Paleozoic ances-tor; morphospace analysis allows us toexamine the functional consequences of evo-lutionary modification of the size, shape, andconductivity of these cells.

Second, wood cells have been a focus ofmuch research in plant biology, and so thequantitative physiological effects of woodanatomy are well studied (Zimmermann1983; Tyree and Ewers 1991). Wood structurehas a strong influence on plant hydraulics,which, in turn, controls how often stomataopen and, thus, indirectly limits the amountof carbon assimilation that can take place inleaves (Kramer and Boyer 1995; Taiz andZeiger 2002; Buckley 2005).

336 JONATHAN P. WILSON AND ANDREW H. KNOLL

Finally, there are well-known tradeoffsbetween structure and function in xylemcells, and well-described and -characterizedpatterns of hydraulic competition, particu-larly between angiosperms and conifers.

In this paper, we use a model of watertransport through wood cells to construct athree-dimensional morphospace of xylemcells where each axis has a discrete anatomi-cal range, creating a space within which everypoint is physiologically meaningful. A phys-iologically explicit morphospace allows us tointegrate phylogeny, function, and time in aneffort to shed light on questions of xylemevolution.

Methods

We adapted a model, developed by Sperryand Hacke (2004; Hacke et al. 2004) for single-cell hydraulics, to calculate resistance throughtracheids containing torus-margo or homo-geneous pits. We have applied this model tofossil plants in earlier work, and the methodsand results are described in detail therein(Wilson et al. 2008).

Model Description

According to the Ohms Law analogy, waterflow through plant vascular systems is analo-gous to the flow of current through an electricalcircuit (van den Honert 1948; Comstock andSperry 2000; Hacke et al. 2004; Sperry andHacke 2004). Both are driven by gradients: avapor pressure gradient between the interior ofa leaf and the atmosphere in the case of planttissues, and a voltage drop across a circuit inelectrical circuits. Resistance (Rtot) is oftenexpressed in terms of conductance (K), whichis the inverse of total resistance and therefore ameasure of flow that is independent ofpressure gradient (eq. 1). Conductivity (con-ductance per unit length) is often normalized toconduit cross-sectional area (Ksc; units: m

2/MPa*s; eq. 2) or wall-specific area (Ksp; units:m2/MPa*s; eq. 3).

K~1

Rtot1

Ksc~Ltracheid

RtotAconduit2

Ksp~Ltracheid

RtotAwall3

Total resistance has two components, fromthe lumen and the end-wall, respectively (eq. 4):

Rtot~RlumenzRwall 4Flow through the lumen (eq. 4) is described by

the Hagen-Poiseuille equation for flow througha cylindrical tube (Lancashire and Ennos 2002).Flow is proportional to the viscosity of water(nH2O; units: Pa * s) and one-half the tracheidlength, the average distance a parcel of watercrosses during its traverse of a tracheid. Thisvalue is inversely proportional to the fourthpower of the conducting cells diameter. Thus, asmall change in diameter can have a dramaticeffect on the lumen resistance:

Rlumen~64 nH2O Ltracheid

p (Dtracheid)4

5

End-wall resistance is a function of thenumber, size, and porosity of pits. Pits aremodeled as resistors in parallel: an increase inthe number of pits reduces the resistance of anindividual cell (eq. 6). Water enters and leavescells by flowing through two sets of pits, so theresistance from pits is doubled in the model.

Rwall~2Rpit

Npits6

Pits have an aperture and a membrane (eq.7). Pit apertures (Fig. 1) have a diameter (Dap)and a thickness (tap) and are modeled as smallcylindrical tubes according to the Hagen-Poiseuille equation (eq. 8). Pit membranescontain pores through which water passesfrom one cell to another. Membranes aremodeled as thin sheets (Vogel 1994) contain-ing a number of pores (Npore) with a givendiameter (Dpore; eq. 9). These pores are oftenmodeled as straight channels, though inobservation with scanning electron micro-scopy they often appear tortuous and par-tially occluded by gel-like macromolecularcompounds, such as pectin or hemicellulose(Rabaey et al. 2006; Sano and Jansen 2006;Jansen et al. 2007; Choat et al. 2008). As in aprevious study (Wilson et al. 2008), we used aresistivity constant (a; unitless) to modify the

PHYSIOLOGICALLY EXPLICIT MORPHOSPACE 337

pit membranes porosity. The previous studyderived a values for homogeneous pit mem-branes and torus-margo pits, and those valuesare used here: 16 and 1, respectively.

Rpit~aRmembranez2Raperture 7

Raperture~128 tap nH2O

pDap4

z24 nH2O

Dap3

8

Rmembrane~24 nH2O

Npore Dpore3

9

Examination of these equations shows thatthe most important factors in determiningtracheid-level conductivity are, in decreasingorder of importance: tracheid diameter(Dtracheid), end-wall resistance (Rwall 5 2*Rpit/Npits), and tracheid length (L). We use thesethree terms as axes for our morphospace.

Pit Area Resistance (rp)

To normalize for the effect of increased pitarea, we calculated pit area resistance (rp) forour taxa by normalizing end-wall resistanceto pit membrane area (eq. 10):

rp~Rwall

2Apit

Npits

2

10

We plotted pit area resistances againsttracheid dimensions to look in detail at

torus-margo pits versus pits with homoge-neous pit membranes. Recent work hasshown that these two pit types differ sig-nificantly when pit resistance is normalized toarea; because of the large pores in the margo,conifers have pit area resistances up to twoorders of magnitude lower than those ofangiosperms (Pittermann et al. 2005,2006a,b). Consequently, there are significanteffects of pit membrane pore size on values ofrp. Taxa that contained scalariform pits(Tetracentron, Trochodendron, and Cycadeoidea)were excluded from this analysis, for reasonsdescribed below.

Materials and Measurements

We measured xylem cell lengths and diam-eters from fossil stem material using lightmicroscopy in tandem with the publicly avail-able image analysis software package ImageJ(found at http://rsbweb.nih.gov/ij/), and alsocompiled values reported in the literature(Bailey and Tupper 1918; Langdon 1920;Chrysler 1926; Andrews 1940; Delevoryas1955; Scott et al. 1962; Bannan 1965; Rothwell1975; Terrazas 1991; Hacke et al. 2007; Ryberget al. 2007). Many of these compiled values arespecies averages for dozens or hundreds oftracheids (Chrysler 1926; Scott et al. 1962;Bannan 1965). Tracheid diameters weremeasured from cross-sections (interior wall tointerior wall). Tracheid length is difficult tomeasure in fossil material because thin sectionsprovide an essentially two-dimensional viewof a three-dimensionally complex structure.This problem is particularly acute forspecimens with long tracheids; two Paleozoicplants, Medullosa and Sphenophyllum, containtracheids that exceed the 34 cm length ofstandard thin sections (Andrews 1940; Cichanand Taylor 1984; Cichan 1985, 1986). Most seedplants, however, have tracheids that aresubstantially shorter than thin sections, andwe measured individual tracheids for whichend-walls were clearly visible. Tracheidlengths are thus the limiting measurement inour data set, for both living and extinct plants.

For living conifer and Ginkgo biloba mate-rial, each data point represents a speciesaverage reported by Bannan (1965) and Baileyand Tupper (1918). For Cycadeoidea and

FIGURE 1. Model of water transport through xylem cellswith morphospace dimensions highlighted: tracheid dia-meter (Dtracheid), tracheid length (Ltracheid), and end-wallresistance (Rwall). Water enters a cell through a series ofapertures on the wall, called pits, flows through the lumen ofthe cell, and exits through another set of pits. Pit morphology,size, number, and porosity determine total end-wallresistance. Homogeneous pits are characteristic of angio-sperms and stem seed plants, and torus-margo pits are foundin conifers and Ginkgo. Modified from Choat et al. (2008).


Callistophyton, dimensions are averages ofseveral specimens from literature reports(Rothwell 1975; Ryberg et al. 2007). For Ginkgobeckii, dimensions are our measurements (n 530) made from type material (Scott et al.1962). For Cordaites sp. and Lyginopteris old-hamium, each point represents an average often tracheids from an individual specimen.For Medullosa, each point represents anaverage of three tracheids from a singlespecimen.

We calculated pit resistance from anatomi-cal measurements, including torus diameter,margo diameter, pit thickness, and fraction ofxylem cell wall occupied by pits. Pit mem-brane porosity is based on values estimatedfrom direct measurement or inferred fromwhole-plant measurements (Choat et al. 2003,2004, 2006). We constrained size of the poresin the pit membrane on the basis of recentwork showing that the proportion of resis-tance coming from pits is approximately 65%of the total hydraulic resistance through axylem cell (Sperry et al. 2005; Wheeler et al.2005). Our values for pore size and numberwere thus based on a pore diameter value thatwould satisfy this relationship. The fraction ofcell walls occupied by pits was measureddirectly from specimens through serial longi-tudinal section or taken from anatomicaldescriptions, figures, or other reports in theliterature (Greguss 1968; Greguss and Balkay1972; Pittermann et al. 2006a).

We measured tracheid dimensions andcalculated pit resistance and tracheid con-ductivity for six extinct seed plants, 12 generaof extant gymnosperms, and three species ofvesselless angiosperms (Table 1, Fig. 2). Ourfossil seed plants include the Carboniferousstem seed plants Medullosa, Lyginopteris, andCallistophyton, the Paleozoic stem conifero-phyte Cordaites, the Mesozoic gymnospermCycadeoidea, and a Miocene ginkgophyte,Ginkgo beckii. The 12 genera of extant gym-nosperms include three cycads (Cycas, Micro-cycas, Dioon), Ginkgo, and eight conifers fromthe families Cupressaceae (Cupressus, Juni-perus, Metasequoia, Sequoia, and Thuja) andPinaceae (Picea, Pinus, and Pseudotsuga). Forseveral conifer genera, we obtained values formultiple species (Table 1). We also included

measurements from three species of extantvesselless angiosperms in the families Win-teraceae and Trochodendraceae: Drimys win-teri, Trochodendron araloides, and Tetracentronsinense. All extant genera in our sample sethave fossil records that extend into theMesozoic, making them effective proxies forMesozoic and Cenozoic torus-margo bearingconifers as well as for vesselless angiosperms.Despite the uncertainties in reconstructingwhole-plant ecology and stature from fossils,this data set contains plants that are arbores-cent (e.g., Metasequoia, Ginkgo beckii, vessellessangiosperms, Cordaites), shrubs or small instature (e.g., Cycadeoidea, cycads), climbers orcreepers (e.g., Lyginopteris, Callistophyton), oruncertain (e.g., Medullosa).

Nine of our plant genera contain torus-margo pits (both Ginkgo species and eightconifer genera), eight genera contain homo-geneous pit membranes (Medullosa, Lyginop-teris, Callistophyton, Cordaites, Cycas, Microcy-cas, Dioon, and Drimys), and three containscalariform pits (Cycadeoidea, Tetracentron, andTrochodendron). We have previously modeledfluid flow through torus-margo and homog-eneous pits (Wilson et al. 2008) and havecreated a method to calculate pit resistance inscalariform pits to simulate fluid flow in thelast three taxa.

Scalariform Pits

Computation of resistance through scalari-form pits has remained a difficult problem.The elliptical shape of scalariform aperturescomplicates direct analytical solution of theHagen-Poiseuille equation (eqs. 5, 8), andboth the presence of bars in some scalariformmembranes (Carlquist 2001) and their un-known function have further frustrated ef-forts to simulate flow through scalariformpits. Others have modeled an analogousanatomical structure, simulating flow fromone vessel element to another (Ellerby andEnnos 1998), but the conducting cells in thesesimulations did not contain pit membranebetween their bars. Furthermore, it is unclearif any species of Cycadeoidea contained barswithin their pit membranes, although mostappear to lack prominent bars.


TA

BL

E1.

Tax

am

easu

red

and

val

ues

use

dfo

rth

isan

aly

sis:

tax

a;d

iam

eter

;le

ng

th;

pit

typ

e;p

itd

ensi

ty;

mem

bra

ne

po

red

iam

eter

;re

fere

nce

s.

Aff

init

yG

enu

sS

pec

ies

Ag

eD

iam

eter

mic

ron

sL

eng

thm

mP

itty

pe

Fp

it1

frac

tio

nP

ore

dia

met

ern

mR

efer

ence

Ste

mg

rou

pse

edp

lan

tM

edu

llos

an

oei

Pen

nsy

lvan

ian

141.

724

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

ap

rim

aev

aP

enn

sylv

ania

n17

017

.5C

ircu

lar-

bo

rder

ed0.

440

Th

isst

ud

yS

tem

gro

up

seed

pla

nt

Med

ull

osa

ang

lica

Pen

nsy

lvan

ian

237

22C

ircu

lar-

bo

rder

ed0.

440

Th

isst

ud

yS

tem

gro

up

seed

pla

nt

Med

ull

osa

sp.

Pen

nsy

lvan

ian

151

25.4

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

asp

.P

enn

sylv

ania

n17

024

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

asp

.P

enn

sylv

ania

n12

123

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

asp

.P

enn

sylv

ania

n16

626

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

asp

.P

enn

sylv

ania

n18

427

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tM

edu

llos

asp

.P

enn

sylv

ania

n14

128

Cir

cula

r-b

ord

ered

0.4

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n10

86.

7C

ircu

lar-

bo

rder

ed0.

340

Th

isst

ud

yS

tem

gro

up

seed

pla

nt

Ly

gin

opte

ris

old

ham

ium

Pen

nsy

lvan

ian

80.3

5.1

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n92

.48

5.8

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n85

.93

4.3

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n90

.82.

94C

ircu

lar-

bo

rder

ed0.

340

Th

isst

ud

yS

tem

gro

up

seed

pla

nt

Ly

gin

opte

ris

old

ham

ium

Pen

nsy

lvan

ian

72.9

14.

36C

ircu

lar-

bo

rder

ed0.

340

Th

isst

ud

yS

tem

gro

up

seed

pla

nt

Ly

gin

opte

ris

old

ham

ium

Pen

nsy

lvan

ian

80.3

54

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n92

.48

4.14

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n94

.35

4.36

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pse

edp

lan

tL

yg

inop

teri

sol

dh

amiu

mP

enn

sylv

ania

n65

.62

2.68

Cir

cula

r-b

ord

ered

0.3

40T

his

stu

dy

Ste

mg

rou

pco

nif

erC

ord

aite

ssp

.P

enn

sylv

ania

n25

4.25

Cir

cula

r-b

ord

ered

0.2

40T

his

stu

dy

Ste

mg

rou

pco

nif

erC

ord

aite

ssp

.P

enn

sylv

ania

n30

3.3

Cir

cula

r-b

ord

ered

0.2

40T

his

stu

dy

Ste

mg

rou

pco

nif

erC

ord

aite

ssp

.P

enn

sylv

ania

n42

.44.

25C

ircu

lar-

bo

rder

ed0.

240

Th

isst

ud

yS

tem

gro

up

con

ifer

Cor

dai

tes

sp.

Pen

nsy

lvan

ian

18.5

52.

25C

ircu

lar-

bo

rder

ed0.

240

Th

isst

ud

yS

tem

gro

up

con

ifer

Cor

dai

tes

sp.

Pen

nsy

lvan

ian

30.2

15

Cir

cula

r-b

ord

ered

0.2

40T

his

stu

dy

Ste

mg

rou

pg

ym

no

sper

mC

alli

stop

hy

ton

por

oxyl

oid

esP

enn

sylv

ania

n10

010

Cir

cula

r-b

ord

ered,

7ro

ws:

0.3

40R

oth

wel

l19

75E

xti

nct

cyca

deo

idC

yca

deo

idea

spp

.L

ow

erC

reta

ceo

us

55.5

2.3

Sca

lari

form{

Rad

ial;

0.17

80G

enu

sav

erag

efr

om

Ry

ber

get

al.

2007

Ex

tin

ctg

ink

go

ph

yte

Gin

kgo

beck

iiM

ioce

ne

482.

9T

oru

s-m

arg

o2

row

s:0.

1540

0S

cott

etal

.19

62;

this

stu

dy

Ex

tan

tG

ink

go

Gin

kgo

bilo

baR

ecen

t40

2.8

To

rus-

mar

go

2ro

ws:

0.15

400

Bai

ley

and

Tu

pp

er19

18;

Sco

ttet

al.

1962

Ex

tan

tG

ink

go

Gin

kgo

bilo

baR

ecen

t80

7T

oru

s-m

arg

o0.

1540

0S

cott

etal

.19

62;

this

stu

dy

Ex

tan

tcy

cad

Cy

cas

thu

arsi

iR

ecen

t33

2.8

Cir

cula

r-b

ord

ered

*0.

1540

Ter

raza

s19

91;

rare

scal

arif

orm

Ex

tan

tcy

cad

Mic

rocy

cas

calo

com

aR

ecen

t45

6C

ircu

lar-

bo

rder

ed*

0.15

40C

hry

sler

1926

;h

alf

scal

arif

orm

Ex

tan

tcy

cad

Dio

onsp

inu

losu

mR

ecen

t32

6.8

Cir

cula

r-b

ord

ered

*0.

140

Bai

ley

and

Tu

pp

er19

18;

Lan

gd

on

1920

;G

reg

uss

1968

Ex

tan

tco

nif

erP

icea

gla

uca

Rec

ent

37.6

3.33

To

rus-

mar

go

8-20

per

trac

hei

d:

0.08

5{

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006


TA

BL

E1.

Co

nti

nu

ed.

Aff

init

yG

enu

sS

pec

ies

Ag

eD

iam

eter

mic

ron

sL

eng

thm

mP

itty

pe

Fp

it1

frac

tio

nP

ore

dia

met

ern

mR

efer

ence

Ex

tan

tco

nif

erP

icea

eng

elm

ann

iR

ecen

t35

.42.

9T

oru

s-m

arg

o0.

075

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

icea

eng

elm

ann

iR

ecen

t36

.72.

91T

oru

s-m

arg

o0.

075

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

icea

mar

ian

aR

ecen

t29

.73.

02T

oru

s-m

arg

o0.

071

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

icea

mar

ian

aR

ecen

t32

.53.

44T

oru

s-m

arg

o0.

071

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

icea

gla

uca

Rec

ent

37.3

3.01

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

stro

bus

Rec

ent

40.3

3.72

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

stro

bus

Rec

ent

414.

04T

oru

s-m

arg

o0.

085{

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

inu

sre

sin

osa

Rec

ent

37.4

3.36

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

pon

der

osa

Rec

ent

38.6

3.33

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

con

tort

aR

ecen

t33

.33.

01T

oru

s-m

arg

o0.

113

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erP

inu

sco

nto

rta

Rec

ent

35.4

2.9

To

rus-

mar

go

0.11

340

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

ban

ksia

na

Rec

ent

34.9

3.14

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pin

us

ban

ksia

na

Rec

ent

35.4

3.23

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Pse

ud

otsu

ga

men

zies

iiR

ecen

t40

.43.

39T

oru

s-m

arg

o0.

085{

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erC

up

ress

us

sarg

enti

iR

ecen

t29

.22.

66T

oru

s-m

arg

o0.

085{

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erC

up

ress

us

sarg

enti

iR

ecen

t31

.82.

91T

oru

s-m

arg

o0.

085{

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erS

equ

oia

sem

per

vire

ns

Rec

ent

43.7

4.12

To

rus-

mar

go

0.08

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erJu

nip

eru

sv

irg

inia

na

Rec

ent

25.1

1.98

To

rus-

mar

go

0.11

400

Ban

nan

1965

;F

pit

fro

mP

itte

rman

net

al.

2006

Ex

tan

tco

nif

erT

hu

jaoc

cid

enta

lis

Rec

ent

28.6

2.63

To

rus-

mar

go

0.08

5{40

0B

ann

an19

65;

Fp

itfr

om

Pit

term

ann

etal

.20

06E

xta

nt

con

ifer

Met

aseq

uoi

ag

lyp

tost

robo

ides

Rec

ent

332

To

rus-

mar

go

0.08

5{40

0Ja

gel

san

dV

issc

her

,20

06E

xta

nt

con

ifer

Met

aseq

uoi

ag

lyp

tost

robo

ides

Rec

ent

484.

5T

oru

s-m

arg

o0.

085{

400

Ban

nan

1965


We measured pit area in Cycadeoidea,Tetracentron, and Trochodendron tracheidsand simulated three scenarios to resolve thisproblem for Cycadeoidea. We calculated tra-cheid resistance and conductivity assumingpit membranes to be as resistive as conifers(rp 5 6 MPa * s * m

21), vesselless angiosperms(rp 5 16 MPa * s * m

21), or eudicots (rp 5336 MPa * s * m21). Total end-wall resistanceis inverted from equation 10. Average valuesfor pit area resistance are taken from theliterature (Sperry et al. 2005, 2007; Pittermannet al. 2006a; Hacke et al. 2007). For Tetracen-tron and Trochodendron, we used publishedvalues of pit area resistances and solved thesame equation (Hacke et al. 2007). Because weused pit area resistances to calculate totalend-wall resistance in Cycadeoidea, Tetracen-tron, and Trochodendron, we excluded thesegenera from our morphometric plot of rpversus diameter and length (Fig. 5).

We compared our calculated estimateswith empirical values of conductivity inliving conifers and relatively narrow andshort angiosperm vessels from Hacke et al.(2004) and Sperry et al. (2004). This allows usto compare vessels of the same size as ourtracheids directly with our analysis.

Results

Our results show three primary patterns.First, there is a very long tail to the distribu-tion of tracheid sizes. Mean tracheid diam-eters are large in Medullosa, Callistophyton,and Lyginopteris (121237, 100, and 65108 mm,respectively), while modern conifer tracheidsare small, never exceeding 40 mm in diameter.Callistophyton and Medullosa also have thelongest tracheids, reaching 10 mm and morethan 20 mm, respectively, and other seedplants have much shorter xylem cells, creat-ing a gap within the seed plant morphospace(Figs. 3, 4). Some tracheids from Ginkgo bilobacan reach dimensions of Lyginopteris andCallistophyton, but their occurrence is rare(Scott et al. 1962).

Second, most living gymnosperms occupya small portion of the morphospace, definedby narrow, short tracheids with moderatetotal pit resistances. Cycads, cyacadeoids,TA

BL

E1.

Co

nti

nu

ed.

Aff

init

yG

enu

sS

pec

ies

Ag

eD

iam

eter

mic

ron

sL

eng

thm

mP

itty

pe

Fp

it1

frac

tio

nP

ore

dia

met

ern

mR

efer

ence

Ex

tan

tv

esse

lles

san

gio

sper

mT

roch

oden

dron

aral

oid

esR

ecen

t12

.81

1.89

Sca

lari

form

0.18

Lar

ge:

200;

Sm

all:

20B

aile

yan

dT

ho

mp

son

,19

18;

Hac

ke

etal

.20

07E

xta

nt

ves

sell

ess

ang

iosp

erm

Tet

race

ntr

onsi

nen

seR

ecen

t21

.41.

97S

cala

rifo

rm0.

16L

arg

e:20

0;S

mal

l:20

Bai

ley

and

Th

om

pso

n,

1918

;H

ack

eet

al.

2007

Ex

tan

tv

esse

lles

san

gio

sper

mD

rim

ys

win

teri

Rec

ent

294.

3C

ircu

lar-

bo

rder

ed0.

077

40B

aile

yan

dT

up

per

1918

;C

arlq

uis

t19

88;

Hac

ke

etal

.20

07

*T

hes

etr

ach

eid

sar

ere

po

rted

toco

nta

inb

oth

scal

arif

orm

and

circ

ula

r-b

ord

ered

pit

s.{

Val

ue

isav

erag

efo

rco

nif

ers,

fro

mP

itte

rman

net

al.

2006

.{

See

M

eth

od

s.

1F

pit

frac

tio

nis

the

frac

tio

no

fw

all

area

cov

ered

by

pit

s.T

his

isth

era

tio

of

pit

mem

bra

ne

area

toto

tal

wal

lar

ea,c

alcu

late

db

yd

ivid

ing

the

pro

du

cto

fa

sin

gle

pit

sar

eaan

dth

en

um

ber

of

pit

sb

yth

eto

tal

wal

lar

eao

fa

trac

hei

d.


ginkgophytes, and conifers all overlap withinthis range, which is distinct from that occu-pied by the stem group seed plants in oursample set (Fig. 3). Third, the dimensions andpitting style of Cycadeoidea suggest, thoughthey do not prove, that its xylem may havefunctioned with pit area resistances similar tothose of the living vesselless angiospermsTetracentron and Trochodendron. Conifers havenotably low pit area resistances, and large-tracheid-bearing seed plants have high valuesof rp (Fig. 5). This is largely a function of pitmembrane pore size and pit area.

When these results are used to calculatesingle-tracheid conductivities, medullosantracheids occupy space that also containsshort, relatively narrow vessels (Hacke et al.2006; Sperry and Hacke 2004), whereas fossilconiferophytes and cycads fall near livingconifer tracheids (Figs. 6, 7). In the Paleozoic,apparently viny seed plants occupied theupper range of conductivity space (Fig. 8).Coniferophytes always occupy the lower

endwith or without torus-margo pits. Mod-ern angiosperms occupy and exceed the rangeof conductivity space that was formerlyoccupied by medullosans and other stemgroup seed plants.

FIGURE 3. Images of tracheids from three Paleozoic seedplants. A, Longitudinal section of Cordaites wood show-ing pits (scale bar, 20 mm). B, Macerated tracheid fromMedullosa (scale bar, 120 mm). C, Transverse section ofxylem from Lyginopteris oldhamium (scale bar, 80 mm).

FIGURE 2. A phylogenetic tree of taxa used in this study, based on Crane (1985), Doyle (2008), and Judd et al. (2007).Extinct lineages are in dark gray; living plants are in light gray. Relationships among fossil taxa are uncertain, but arenot critical to the conclusions of this study.


No living seed plant group occupies the spacethat Medullosa, Callistophyton, and Lyginopterisheld, with wide, long, and low-resistance tra-cheids. Today, the only plant that may fit in thisspace is the Costa Rican climbing fern Salpi-chlaena volubilis, reported to have tracheiddiameters in excess of 200 mm and lengths inexcess of 20 mm (Veres 1990). It is probable thatthis area of morphospace is limited to tropical,large-leaved, tracheid-bearing plants that do notrely on conducting cells for structural support(Wnuk and Pfefferkorn 1984; Mosbrugger 1990;DiMichele et al. 2006; Wilson et al. 2008).

Taphonomy obscures individual pit mem-brane porosity in most pre-Cenozoic fossilplants, with only rare exceptions (Beck andWight 1988), forcing us to rely on compara-tive biology for estimates of pore diameter.Although we can differentiate torus-margopits from circular bordered pits in fossilplants, we may never know exactly how largepores were in the pit membranes of ancientseed plants. As in other experimental and

theoretical work, pit membrane porosityrepresents the largest uncertainty in thisanalysis (Choat et al. 2003, 2008; Wilson etal. 2008). Our estimates of conductivity fallwithin the ranges of measurements fromliving conifers and angiosperms, however,adding confidence to our simulations (Fig. 6).

Low-resistance transport evolved at leastonce early in the seed plant clade. Stem seedplants, particularly ones found within thetropical Euramerican coal swamps, achievedlow-resistance transport by increasing bothsingle-tracheid dimensions and pit area.

Discussion

Interpretation of wood evolution requiresthat information from our morphospaceanalysis be integrated with phylogenetic andstratigraphic data. Molecular phylogeniessuggest that angiosperms are sister to allother living seed plants. Most major seedplant clades are extinct, however, and phylo-genetic analyses of the full range of seed plant

FIGURE 4. Separate two-dimensional views of our morphospace, split into total end-wall resistance versus diameterand total end-wall resistance versus length. Note the long tail in tracheid diameters that is absent in tracheid lengths.


diversity remain a challenge. For the pur-poses of discussion, we illustrate a treemodified from Doyle (2008) and others (Crane1985, Judd et al. 2007) (Fig. 2), but note thatour conclusions about xylem evolution do notdepend strongly on choice of phylogeny.

The Evolution of Pycnoxylic Wood

Pycnoxylic wood occurs in a number ofPaleozoic plants, including the progymno-sperm Archaeopteris (Beck 1960, 1962, 1970;Beck and Wight 1988), Cordaites, conifers,Glossopteris (Maheshwari 1972), and, perhaps,a greater diversity of clades recorded by formtaxa such as Dadoxylon (Erasmus 1976; Scott1902) and Zalesskioxylon (Feng et al. 2008).

At a minimum, pycnoxylic wood evolvedindependently in the progymnosperms; in thecommon ancestor of ginkgophytes, cordaites,and conifers (as the earliest known seedplants have manoyxlic wood much like that

of aneurophyte progymnosperms [Galtier1988; Galtier and Meyer-Berthaud 2006; Serbetand Rothwell 1992]); and in the clade repre-sented by glossopterids. The torus-margo pit-ting of extant conifers and Ginkgo evolvedtwice, independently in each of these clades.Such a conclusion is supported by phylogenies(Crane 1985) in which the plesiomorphiccharacter of ginkgophyte reproductive biologyplaces it below a node containing the repro-ductively similar cordaites and conifers (Florin1950, 1951; Rothwell 1982; Rothwell et al. 1997).Regardless of this relationship, however, theindependent origin of torus-margo pitting inGinkgo is mandated by the observation that allknown Paleozoic conifers have bordered pitswith simple pit membranes (Rothwell 1982;Stewart and Rothwell 1993; Taylor and Taylor1993; Rothwell et al. 1997). The earliest knownpycnoxylic wood with torus-margo pits occursin the Hettangian (Early Jurassic, ca. 210 Ma)

FIGURE 5. Separate two-dimensional views of pit area resistance (rp) versus length and rp versus diameter. Note thatMedullosa and Cordaites have relatively high pit area resistance values, whereas living conifers have low pit arearesistance values.


(M. Philippe personal communication, 2008).Insofar as stem ginkgophytes are well knownfrom the Permian and Triassic (Florin 1949;Anderson et al. 1998), neither stem conifers norstem ginkgophytes had torus-margo pits.

Torus-margo pits have also been reportedfrom the Gnetales, in Ephedra and Gnetum(Carlquist 1996). Because of their uncertainphylogenetic placement (Burleigh and Math-ews 2004; Bateman et al. 2006; Doyle 2006,2008), a confirmed appearance of torus-margopits in this group could represent eitherretention of the plesiomorphic state fromcrown-group conifers, or another indepen-dent origin in addition to Ginkgo.

All conifers and other seed plants withpycnoxylic wood and simple pit membranes

are extinct; all extant seed plants with pyc-noxylic wood have torus-margo pits. Thisstrongly suggests that the torus-margo pitconferred selective advantage on its bearers,although we cannot rule out selection for othertraits that co-occurred with these modified pitmembranes. Our morphospace suggests thatthe selective advantage was not simply higherconductivity, because Paleozoic plants withconifer-like wood achieved conductivity similarto that of living conifers by increasing the totalarea of pitting on tracheid walls. More likely,advantage must be sought in the dual functionof pycnoxylic tracheids. Because they allowcomparable rates of water transport with fewerpits, torus-and-margo tracheids are mechani-cally stronger than those of pycnoxylic Paleo-

FIGURE 6. Specific conductivity (conductance normalized to length and wall area; eq. 3) versus length and diameter,split into two 2-d plots. Angiosperm and tracheid values are from Hacke et al. (2004) and Sperry et al. (2004).


zoic woods. The torus-margo structure alsoallows for a tighter seal against embolism, ifsuch events are rare; when subject to repeatedembolism events, the torus may become stuckagainst the aperture, permanently sealing thetracheid against fluid flow (Zimmermann 1983;Sperry and Tyree 1990; Sperry et al. 1994; Hackeet al. 2001b; Pittermann et al. 2005, 2006a,b).Perhaps the tracheids observed today in coni-fers and Ginkgo present the optimum balanceamong mechanical support, cavitation avoid-ance, and water conductivity in plants devel-opmentally committed to pycnoxylic wood.

Consequences of High-Conductivity,Tracheid-Based Xylem

High-conductivity xylem based on tra-cheids can only function at a physiologicalcost. The same features that allow medullosan

or callistophytalean tracheids to move largevolumes of water with low resistance alsomake them vulnerable to damage fromdrought and frost: wide, long, and highlypitted tracheids are extremely vulnerable tocavitation and embolism, whereas narrower,shorter, and less densely pitted tracheids arerelatively risk averse (Dixon 1909, 1914;Sperry and Tyree 1988, 1990; Tyree andSperry 1988, 1989; Tyree and Ewers 1991;Sperry et al. 1994; Sperry and Ikeda 1997;Tyree et al. 1999; Hacke et al. 2001a).Furthermore, the structure of medullosanxylem in particular, with extremely longtracheids anastomosing in a network, allowsembolism to propagate over long distanceswithin a stem (Loepfe et al. 2007). The highproportion of living cells in the xylem ofmanoxylic plants may reflect the physiologi-

FIGURE 7. A histogram of conductivity in Paleozoic seed plants versus extant gymnosperms. Coniferophytes fall nearthe left wall of both histograms. In the Paleozoic, Lyginopteris and Callistophyton fall at intermediate values ofconductivity, and Medullosa at the high end. In the lower plot, angiosperm vessels fall at high conductivity values andnormally exceed values shown here. Medullosa overlaps with small angiosperm vessels.


cal need to refill embolized tracheids (Wilsonet al. 2008).

The large-tracheid-based hydraulic struc-ture also commits plants to environmentswithout freezing or moisture stress. Duringfreezing, air is pushed out of solution; largercells will produce more air when frozen,leading to larger cavitation events. A vastliterature shows that large conduits are morevulnerable to cavitation than small conduits(Hammel 1967; Sucoff 1969; Sperry and Sulli-van 1992; Davis et al. 1999; Ball et al. 2002,2006). Ice has an extremely low water potentialand creates a hydraulic gradient that candehydrate nearby cells and damage cellmembranes (Steponkus 1984; Pearce 2001).Large tracheids, like vessels, are potentialsources of nucleation, especially if their diam-

eter exceeds 44 mm (Davis et al. 1999). Modernplants mitigate this stress by refilling cavitatedtracheids through root pressure or another,more cryptic, mechanism to generate positivepressure in the xylem, or they grow new woodevery year. The small amount of wood foundin the Paleozoic seed plants with large trac-heids rules out new wood formation, and thelack of hydathodes on their leaves arguesagainst root pressure. In the absence of amechanism to repair freeze-thaw cavitation, itis unlikely that plants with large tracheidscould survive a cold climate. It is telling thatthe most conspicuous group of angiospermsthat have reverted to tracheid-based transport,the Winteraceae, appear to have lost vessels inresponse to increased exposure to frost events(Feild et al. 2000, 2002).

FIGURE 8. A decision tree of how to decrease total xylem resistance. There are two ways to decrease resistance: byreducing resistance of the wall and/or the lumen. To reduce end-wall resistance, there are two methods: increasingindividual porosity, through the development of the torus-margo pit, or by adding pit area, which appears to becommon in Paleozoic seed plants. To reduce lumen resistance, plants have turned either to multicellularity, which isfound in angiosperm vessels, or to enlarging individual xylem cells, which is found in Medullosa and certain otherPaleozoic seed plants.


As noted above, moisture stress impartssimilar size constraints to water transporttissues but for different reasons: large tra-cheids have more pit area and are vulnerableto defects in pit membrane development,leading to rare large pores that can allowembolism to spread (Wheeler et al. 2005).Finally, the thickness-to-span ratio of tra-cheids from Lyginopteris, Callistophyton, andMedullosa would have reduced their ability toresist drought-induced embolism and implo-sion (Hacke and Sperry 2001; Hacke et al.2001a,b). Although these cells reduced abilityto resist strong tensions does not directlycorrespond to a reduced degree of structuralsupport, given the complex nature of biome-chanical support in vascular plants (DiMi-chele 1979; Cichan 1986; DiMichele andDeMaris 1987; Niklas 1992; Rowe et al. 1993;Carlquist 2001; Niklas and Spatz 2004; Roweand Speck 2004; DiMichele and Gastaldo2008), a stem composed of highly conductive,wide tracheids would not, by itself, contributesignificant structural support to a plant.

Arborescence and Conductivity.In light of apossible tradeoff between low stature andhigh conductivity, a key component of vege-tative success in angiosperms may lie in thevascular cambium, not strictly in high-throughput vessels, as other authors havesuggested (Bond 1989). Angiosperm vegeta-tive success may reflect the evolution of avascular system that can conduct water athigh volumes and simultaneously retain theability to grow tall. In this scenario, the keyinnovation is the invention of fiber-basedstructural support derived from the vascularcambium. The Paleozoic plants that occupiedthe portion of our morphospace that is nowangiosperm-dominated were likely scram-bling or climbing plants, rather than free-standing trees, limiting the ecological zonesthey could occupy. Angiosperms are notlimited in this manner and can achieve staturecomparable to conifers and Ginkgo, but withhigh-conductivity wood irrigating large leafareas. This capacity for structural evolutionrelies on a vascular cambium that caninitialize both vessels and fibers, and fibersare the component unexplored by otherclades. A recent analysis of Early Cretaceous

European wood supports this interpretation:earliest angiosperm wood contained rela-tively thin-walled fibers, leading to a low-diversity flora composed of rapidly growing,scrambling forms; the evolution of thick-walled fibers allowed for stiff woods andaccompanied the radiation of angiospermsinto many different environments (Philippe etal. 2008).

There have been further elaborations onthis form within the angiosperms: palmsachieve large stature without a vascularcambium by means of their primary thicken-ing meristem (Tomlinson 1990); many speciesof the dicot Gunnera support several meter-scale leaves on stems less than a meter highbecause of structural support conferred byleaf bases (Batham 1943; Wilkinson 2000). Thediversity of structural adaptations foundwithin plants that contain vessels and fibersfar exceeds that in plants with pycnoxylicwood (Mosbrugger 1990; Niklas 1997).

In contrast, once the vascular cambiumbegins to produce pycnoxylic wood withoutliving cells in association with each tracheid,it may become difficult to revert to a condi-tion in which cambial initials can expand tothe diameters and lengths required to maketracheids of medullosan scale. The limitednumber of living cells prevents tracheids fromincreasing their initial diameter and length bythe factor of ten needed to reach this size. It ispossible that pycnoxylic wood prohibitsmodes of life that were common among stemseed plants in the Paleozoic, because it isdifficult to supply leaves with the amount ofwater required to irrigate the lamina whilemaintaining small-diameter stems required toclimb or scramble. A vine with coniferophytetracheids would require a larger area ofxylem to support a given flow volume thanone that had medullosan tracheids or vessels.Ferns, the most common tracheid-bearingclimbing plants in the modern flora, tend torely on fewer, larger xylem cells to conductwater, rather than large cylinders of xylem(Veres 1990).

Given these differences in structural sup-port and conductivity, it is somewhat surpris-ing that coniferous and angiosperm forestsare almost equally abundant: large, relatively


low-diversity gymnosperm forests cover theboreal regions of the Northern Hemisphere,and high-diversity angiosperm forests dom-inate the equatorial and temperate regions.The functional flexibility conferred on angios-perms by a wide range of anatomical char-acters, including their vascular cambium,allows them to dominate a wide array ofenvironments, from tropical forests to deserts.Conifers and gymnosperms do not exhibitthis range of environmental tolerance, buttheir dominance of Siberian and Canadiancircumpolar forests illustrates their ecologicalsuccess in niche environments.

Gigantopterids.The Permian gigantopter-ids, largely found as compression fossils inChina and the central United States, are well-known for their angiosperm-like vegetativeorgans, including large, entire-marginedleaves and putative vessel elements (Mamayet al. 1988; Li et al. 1996; Beck and Labandeira1998; Li and Taylor 1998; Glasspool et al.2004a,b). Although permineralized specimensare rare, and reproductive organs have neverbeen foundpreventing any taxonomic linkwith other seed plant groups vegetativefeatures of gigantopterids suggest that theymay resemble medullosans and angiospermsin functional space, rather than conifers.Gigantopterid xylem cells, including theirputative vessel elements, are remarkablywide (.300 mm) and have radial walls thatare covered by circular-bordered pits, muchlike the xylem cells of Medullosa (Li andTaylor 1998; Li et al. 1996). All of thesefeatures suggest a plant that lived in a tropicalever wet environment and was extremelyvulnerable to cavitation, again, much likeMedullosa (Wilson et al. 2008).

Environmental Context.It is informative toview the Late Paleozoic seed plants withhigh-conductivity xylem in their environmen-tal context. All of these plant fossils, fromLyginopteris to Gigantopteris, were preservedin swamps, and it is likely that these plantsinhabited swampy, ever wet environments.During the assembly of Pangaea, which beganin the Pennsylvanian and continued throughthe Permian, these basins were closed, theseenvironments were lost, and these plantgroups disappeared from the record. Because

the taphonomic window is closed, we cannotrule out the possibility that these plantsmoved to extrabasinal environments, buttheir absence from the palynological recordpoints toward extinction rather than migra-tion (Taylor et al. 2008).

Patterns of Reducing Hydraulic Resistance.At the broadest scale, then, it appears thatthere are two major ways to reduce resistancein xylem cells, by modifying the wall resis-tance and/or by modifying the lumen resis-tance (Fig. 8).

One way to decrease wall resistance is byincreasing the porosity of individual pitmembranes. Torus-margo pits allow individ-ual pores to be up to 1 mm in diameter(Lancashire and Ennos 2002; Pittermann et al.2005, 2006a; Sperry et al. 2006; Choat et al.2008). These large pores dramatically increasethe porosity of an individual tracheid, makingconifers and ginkgos competitive with an-giosperms (Zimmermann 1983; Pittermann etal. 2005, 2006a). The combination of pyc-noxylic wood and torus-margo pits allows forlow-resistance fluid flow and structural sup-port in the same tissues.

Another way to decrease tracheid wallresistance is by increasing pit area in the cell.Increasing pit area, either by adding more pitsor by enlarging individual pit membranes,will reduce the resistance from xylem cellwalls. Tracheids with more or larger pits arerare in living plants but widespread in extinctgroups, including Lyginopteris, Medullosa, Cal-listophyton, Cordaites, early conifers (Philippeand Bamford 2008), gigantopterids (Li andTaylor 1998), and the Mesozoic seed plantPentoxylon (Taylor and Taylor 1993). Thedrawbacks to increased pit area may haveincluded decreased structural support fromthinner cell walls (Hacke et al. 2001a; Pitter-mann et al. 2006b) and an increased prob-ability that embolisms would spread becauseof microfibril failure, leaving large pores inpit membranes (Choat et al. 2003; Hacke et al.2007).

The other major way to decrease hydraulicresistance is to minimize the componentassociated with conducting cell lumens. Thereare at least two ways to accomplish this:increasing individual cell dimensions or


combining individual cells into multicellularvessels.

If all else is held equal (e.g., length-to-diameter ratio, proportion of wall that con-tains pits), above 30 mm in diameter, theproportion of total resistance coming fromlumens is approximately one-third of totalresistance and decreases with further in-creases in diameter (Schulte et al. 1987;Hacke et al. 2004; Sperry and Hacke 2004;Sperry et al. 2005; Wheeler et al. 2005).However, larger tracheids provide larger flowrates, and a single tracheid with diameter rwill conduct eight times more water per unitpressure gradient than two tracheids withdiameter half the size (van den Honert 1948;Tyree and Ewers 1991; Sperry et al. 2005).Increasing length is effective if fusiforminitials are narrow or cannot be enlarged,but length increases do not take advantage ofthe Hagen-Poiseuille relationship and insteadincrease flow rate in a linear relationship(Comstock and Sperry 2000). Therefore, it ismore economical to increase the diameter ofthe conducting cell, even if it comes at theexpense of the number of conducting cells ina stems cross-sectional area.

Relying on a smaller number of large-diameter tracheids appears to be commonamong stem seed plants, including Lyginop-teris, Medullosa, and Callistophyton, and israrely found among living ferns, includingOphioglossum and Salpichlaena volubilis. It isnotable that Ginkgo biloba contains broadleaves in addition to a small number of largetracheids, whereas most conifers lack both(Scott et al. 1962).

Another way to decrease lumen resistanceis by linking xylem cells together by dissol-ving their end walls to create vessels, dramat-ically increasing the length of the flow path.Combining this with multiple types of tissuesderived from the vascular cambium, such assmall, thick-walled fibers for structural sup-port, permits the development of wide xylemcells that maximize conductivity. The combi-nation of vessels and fibers is characteristic ofthe vast majority of angiosperms, somemembers of the Gnetales (Carlquist 1996),and the late Paleozoic gigantopterids (Li andTaylor 1998). It is telling that the only genus

of living gymnosperms that contains vines,Gnetum, contains vessels mixed with narrowtracheids (Carlquist 1996; Feild and Balun2008).

Conclusions

Because their physiological performance isclosely tied to preservable anatomical struc-ture, fossil vascular plants are amenable tomorphometric analysis that tracks functionaldiversity through time. Tracheids in mostliving and extinct gymnosperms are function-ally similar, but the conducting cells of somestem-group seed plants are exceptional: thewide, long, and relatively low-resistance tra-cheids of Medullosa, Lyginopteris, and Callisto-phyton conducted water at resistances com-parable to those of moderately sized angio-sperm vessels.

Seed plants have used several differentphysiological strategies to achieve high-con-ductivity xylem, but only two are pursuedtoday: high-porosity torus-margo pits andmulticellular vessels. Functional canalizationinto these two strategies reflects the extinctionof non-angiosperm seed plants. Anatomicallydistinct plants capable of high throughputwater transport were common in tropical wetforests of the late Paleozoic but did notsurvive Permian climatic change.

Our analysis highlights two points thatbecome clear only when morphometric andphylogenetic data are combined. First, mostprevious studies have suggested that thewidespread distribution of torus-margo pitsis due to the increased fluid flow facilitatedby large pores within the membrane; how-ever, our analysis shows that stem conifer-ophytes had tracheids with comparable fluidflow ability, but reduced structural supportand cavitation resistance. It is likely that theevolution of torus-margo pits reflects abalance among fluid flow, safety, and sup-port, rather than simply increased fluid flow.

Also, our analysis suggests that the vegeta-tive success of angiosperms is not simply afunction of high-throughput vessels, becausesome Paleozoic seed plants, including Medul-losa, had comparably functioning xylem cells.The key innovation may have been a vascularcambium that permitted development of


thick-walled fibers for structural support, aswell as vessels, facilitating the evolution oftall trees containing low-resistance xylem ableto irrigate large leaf areas.

Acknowledgments

We thank N. M. Holbrook, F. Rockwell, andS. Costanza for helpful discussion, and twoanonymous reviewers for comments thatimproved this manuscript. J.P.W. is sup-ported in part by the NASA AstrobiologyInstitute and a Deland Award for BotanicalResearch from the Arnold Arboretum ofHarvard University.

Literature Cited

Anderson, J. M., H. M. Anderson, and A. R. I. Cruickshank. 1998.

Late Triassic ecosystems of the Molteno Lower Elliot biome of

southern Africa. Palaeontology 41:387421.

Andrews, H. N., Jr. 1940. On the stelar anatomy of the

pteridosperms with particular reference to the secondary

wood. Annals of the Missouri Botanical Garden 27:51118.

Bailey, I. W., and W. P. Thompson. 1918. Additional notes upon

the angiosperms Tetracentron, Trochodendron, and Drimys, in

which vessels are absent from the Wood. Annals of Botany (old

series) 32:503512.

Bailey, I. W., and W. W. Tupper. 1918. Size variation in tracheary

cells. I. A comparison between the secondary xylems of

vascular cryptogams, gymnosperms, and angiosperms. Pro-

ceedings of the American Academy of Arts and Sciences

54:149204.

Ball, M. C., J. Wolfe, M. Canny, M. Hofmann, A. B. Nicotra, and

D. Hughes. 2002. Space and time dependence of temperature

and freezing in evergreen leaves. Functional Plant Biology

29:12591272.

Ball, M. C., M. J. Canny, C. X. Huang, J. J. G. Egerton, and J. Wolfe.

2006. Freeze/thaw-induced embolism depends on nadir

temperature: the heterogeneous hydration hypothesis. Plant,

Cell and Environment 29:729745.

Bannan, M. W. 1965. Length tangential diameter and length/

width ratio of conifer tracheids. Canadian Journal of Botany

43:967984.

Bateman, R. M., J. Hilton, and P. J. Rudall. 2006. Morphological

and molecular phylogenetic context of the angiosperms:

contrasting the top-down and bottom-up approaches used

to infer the likely characteristics of the first flowers. Journal of

Experimental Botany 57:34713503.

Batham, E. 1943. Vascular anatomy of New Zealand species of

Gunnera. Transactions of the Royal Society of New Zealand

73:7.

Beck, A. L., and C. C. Labandeira. 1998. Early Permian insect

folivory on a gigantopterid-dominated riparian flora from

north-central Texas. Palaeogeography, Palaeoclimatology, Pa-

laeoecology 142:139173.

Beck, C. B. 1960. Connection between Archaeopteris and Callixylon.

Science 131:15241525.

. 1962. Reconstructions of Archaeopteris, and further

consideration of its phylogenetic position. American Journal

of Botany 49:373382.

. 1970. The appearance of gymnospermous structure.

Biological Reviews 45:379399.

Beck, C. B., and D. C. Wight. 1988. Progymnosperms. Pp. 185 in

C. Beck, ed. Origin and evolution of gymnosperms. Columbia

University Press, New York.

Beck, C., R. Schmid, and G. Rothwell. 1982. Stelar morphology

and the primary vascular system of seed plants. Botanical

Review 48:691815.

Bond, W. J. 1989. The tortoise and the hare: ecology of angiosperm

dominance and gymnosperm persistence. Biological Journal of

the Linnean Society 36:227249.

Boyce, C. K., and A. H. Knoll. 2002. Evolution of developmental

potential and the multiple independent origins of leaves in

Paleozoic vascular plants. Paleobiology 28:70100.

Buckley, T. N. 2005. The control of stomata by water balance.

New Phytologist 168:275291.

Burleigh, J. G., and S. Mathews. 2004. Phylogenetic signal in

nucleotide data from seed plants: implications for resolving the

seed plant tree of life. American Journal of Botany 91:1599

1613.

Carlquist, S. 1996. Wood, bark and stem anatomy of Gnetales: a

summary. International Journal of Plant Sciences 157:S58.

. 1988. Wood anatomy of Drimys s.s. (Winteraceae) Aliso

12:8195.

. 2001. Comparative wood anatomy. Springer, Berlin.

Choat, B., M. Ball, J. Luly, and J. Holtum. 2003. Pit membrane

porosity and water stress-induced cavitation in four co-existing

dry rainforest tree species. Plant Physiology 131:4148.

Choat, B., S. Jansen, M. A. Zwieniecki, E. Smets, and N. M.

Holbrook. 2004. Changes in pit membrane porosity due to

deflection and stretching: the role of vestured pits. Journal of

Experimental Botany 55:15691575.

Choat, B., T. W. Brodie, A. R. Cobb, M. A. Zwieniecki, and N. M.

Holbrook. 2006. Direct measurements of intervessel pit mem-

brane hydraulic resistance in two angiosperm tree species.

American Journal of Botany 93:9931000.

Choat, B., A. R. Cobb, and S. Jansen. 2008. Structure and function

of bordered pits: new discoveries and impacts on whole-plant

hydraulic function. New Phytologist 177:608626.

Chrysler, M. A. 1926. Vascular tissues of Microcycas calocoma.

Botanical Gazette 82:233252.

Cichan, M. A. 1985. Vascular cambium and wood development in

Carboniferous plants. 2. Sphenophyllum-Plurifoliatum William-

son and Scott (Sphenophyllales). Botanical Gazette 146:395403.

. 1986. Vascular cambium and wood development in

Carboniferous plants. 4. Seed plants. Botanical Gazette 147:227

235.

Cichan, M. A., and T. N. Taylor. 1984. A method for determining

tracheid lengths in petrified wood by analysis of cross-sections.

Annals of Botany 53:219226.

Comstock, J. P., and J. S. Sperry. 2000. Theoretical considerations

of optimal conduit length for water transport in vascular

plants. New Phytologist 148:195218.

Crane, P. R. 1985. Phylogenetic analysis of seed plants and the

origin of angiosperms. Annals of the Missouri Botanical

Garden 72:716793.

Davis, S. D., J. S. Sperry, and U. G. Hacke. 1999. The relationship

between xylem conduit diameter and cavitation caused by

freezing. American Journal of Botany 86:13671372.

Delevoryas, T. 1955. The Medullosae: structure and relationships.

Palaeontographica, Abteilung B 97:114167.

DiMichele, W. A. 1979. Arborescent lycopods of Pennsylvanian

age coals. Palaeontographica, Abteilung B 171:5777.

DiMichele, W. A., and P. J. DeMaris. 1987. Structure and

dynamics of a Pennsylvanian-age Lepidodendron forest;

colonizers of a disturbed swamp habitat in the Herrin (No. 6)

Coal of Illinois. Palaios 2:146157.

DiMichele, W. A., and R. A. Gastaldo. 2008. Plant paleoecology in

deep time. Annals of the Missouri Botanical Garden 95:144198.


DiMichele, W. A., T. L. Phillips, and H. W. Pfefferkorn. 2006.

Paleoecology of Late Paleozoic pteridosperms from tropical

Euramerica. Journal of the Torrey Botanical Society 133:83118.

Dixon, H. H. 1909. Transpiration and the ascent of sap.

Progressus Rei Botanicae 3:166.

. 1914. Transpiration and the ascent of sap in plants.

Macmillan, London.

Doyle, J. A. 2006. Seed ferns and the origin of angiosperms.

Journal of the Torrey Botanical Society 133:169209.

. 2008. Integrating molecular phylogenetic and paleobotan-

ical evidence on origin of the flower. International Journal of

Plant Sciences 169:816843.

Edwards, D. 2003. Xylem in early tracheophytes. Plant, Cell and

Environment 26:5772.

Ellerby, D., and A. Ennos. 1998. Resistances to fluid flow of model

xylem vessels with simple and scalariform perforation plates.

Journal of Experimental Botany 49:979985.

Erasmus, T. 1976. On the anatomy of Dadoxylon arberi Seward

with some remarks on the phylogenetical tendencies of its

tracheid pits. Palaeontologia Africana 19:127133.

Esau, K. 1977. Anatomy of seed plants. Wiley, New York.

Feild, T. S., and L. Balun. 2008. Xylem hydraulic and photo-

synthetic function of Gnetum (Gnetales) species from Papua

New Guinea. New Phytologist 177:665675.

Feild, T. S., M. A. Zwieniecki, and N. M. Holbrook. 2000.

Winteraceae evolution: an ecophysiological perspective. Annals

of the Missouri Botanical Garden 87:323334.

Feild, T. S., T. Brodribb, and M. Holbrook. 2002. Hardly a relict:

freezing and the evolution of vesselless wood in Winteraceae.

Evolution 56:464478.

Feng, Z., J. Wang, and G.-L. Shen. 2008. Zalesskioxylon xiaheyanense

sp. nov., a gymnospermous wood of the Stephanian (Late

Pennsylvanian) from Ningxia, northwestern China. Journal of

Asian Earth Sciences 33:219228.

Florin, R. 1949. The morphology of Trichopitys heteromorpha

Saporta, a seed plant of Paleozoic age, and the evolution of

the female flowers in the Ginkgoinae. Acta Horti Bergiani

15:79109.

. 1950. Upper Carboniferous and Lower Permian conifers.

Botanical Review 16:258282.

. 1951. Evolution in Cordaites and conifers. Acta Horti

Bergiani 15:285388.

Galtier, J. 1988. Morphology and phylogenetic relationships of

early pteridosperms. Pp. 135177 in C. Beck, ed. Origin and

evolution of gymnosperms. Columbia University Press, New

York.

Galtier, J., and B. Meyer-Berthaud. 2006. The diversification of

early arborescent seed ferns. Journal of the Torrey Botanical

Society 133:719.

Glasspool, I., J. Hilton, M. E. Collinson, and W. Shi-Jun. 2004a.

Defining the gigantopterid concept: a reinvestigation of

Gigantopteris (Megalopteris) nicotianaefolia Schenck and its

taxonomic implications. Palaeontology 47:13391361.

Glasspool, I. J., J. Hilton, M. E. Collinson, S.-J. Wang, and S. Li

Cheng. 2004b. Foliar physiognomy in Cathaysian gigantopter-

ids and the potential to track Palaeozoic climates using an

extinct plant group. Palaeogeography, Palaeoclimatology,

Palaeoecology 205:69110.

Greguss, P. 1968. Xylotomy of the living cycads. Akademiai

Kiado, Budapest, Hungary.

Greguss, P., and B. Balkay. 1972. Xylotomy of the living conifers.

Akademiai Kiado, Budapest, Hungary.

Hacke, U. G., and J. S. Sperry. 2001. Functional and ecological

xylem anatomy. Perspectives in Plant Ecology, Evolution and

Systematics 4:97115.

Hacke, U. G., J. S. Sperry, W. T. Pockman, S. D. Davis, and K. A.

McCulloch. 2001a. Trends in wood density and structure are

linked to prevention of xylem implosion by negative pressure.

Oecologia 126:457461.

Hacke, U. G., V. Stiller, J. S. Sperry, J. Pittermann, and K. A.

McCulloh. 2001b. Cavitation fatigue: embolism and refilling

cycles can weaken the cavitation resistance of xylem. Plant

Physiology 125:779786.

Hacke, U. G., J. S. Sperry, and J. Pittermann. 2004. Analysis of

circular bordered pit function. II. Gymnosperm tracheids with

torus-margo pit membranes. American Journal of Botany

91:386400.

Hacke, U. G., J. S. Sperry, J. K. Wheeler, and L. Castro. 2006.

Scaling of angiosperm xylem structure with safety and

efficiency. Tree Physiology 26:689701.

Hacke, U. G., J. S. Sperry, T. S. Feild, Y. Sano, E. H. Sikkema, and J.

Pittermann. 2007. Water transport in vesselless angiosperms:

conducting efficiency and cavitation safety. International

Journal of Plant Sciences 168:11131126.

Hammel, H. T. 1967. Freezing of xylem sap without cavitation.

Plant Physiology 42:5566.

Jagels, R., and G. E. Visscher. 2006. A synchronous increase in

hydraulic conductive capacity and mechanical support in

conifers with relatively uniform xylem structure. American

Journal of Botany 93:179187.

Jansen, S., Y. Sano, B. Choat, D. Rabaey, F. Lens, and R. R. Dute.

2007. Pit membranes in tracheary elements of Rosaceae and

related families: new records of tori and pseudotori. American


Judd, W. S., C. S. Campbell, E. A. Kellogg, P. F. Stevens, and M. J.

Donoghue. 2007. Plant systematics: a phylogenetic approach.

Sinauer, Sunderland, Mass.

Kramer, P., and J. Boyer. 1995. Water relations of plants and soils.

Academic Press, San Diego.

Lancashire, J. R., and A. R. Ennos. 2002. Modelling the

hydrodynamic resistance of bordered pits. Journal of Experi-

mental Botany 53:14851493.

Langdon, L. M. 1920. Stem anatomy of Dioon spinulosum.

Botanical Gazette 70:110125.

Li, H., and D. W. Taylor. 1998. Aculeovinea yunguiensis Gen. et Sp.

Nov. (Gigantopteridales), a new taxon of gigantopterid stem

from the Upper Permian of Guizhou Province, China. Interna-

tional Journal of Plant Sciences 159:10231033.

Li, H., E. L. Taylor, and T. N. Taylor. 1996. Permian vessel

elements. Science 271:188189.

Loepfe, L., J. Martinez-Vilalta, J. Pinol, and M. Mencuccini. 2007.

The relevance of xylem network structure for plant hydraulic

efficiency and safety. Journal of Theoretical Biology 247:788

803.

Maheshwari, H. K. 1972. Permian wood from Antarctica and

revision of some Lower Gondwana wood taxa. Palaeontogra-

phica, Abteilung B 138:143.

Mamay, S. H., J. M. Miller, D. M. Rohr, and W. E. Stein Jr. 1988.

Foliar morphology and anatomy of the gigantopterid plant

Delnortea abbottiae, from the Lower Permian of West Texas.

American Journal of Botany 75:14091433.

McGhee, G. R. 1999. Theoretical morphology: the concept and its

applications. Columbia University Press, New York.

Mosbrugger, V. 1990. The tree habit in land plants: a functional

comparison of trunk constructions with a brief introduction

into the biomechanics of trees. Springer, Berlin.

Namboodiri, K. K., and C. B. Beck. 1968. A comparative study of

the primary vascular system of conifers. III. Stelar evolution in

gymnosperms. American Journal of Botany 55:464472.

Niklas, K. J. 1992. Plant biomechanics. University of Chicago

Press, Chicago.

. 1997. The evolutionary biology of plants. University of

Chicago Press, Chicago.

Niklas, K. J., and H. C. Spatz. 2004. Growth and hydraulic (not

mechanical) constraints govern the scaling of tree height and


mass. Proceedings of the National Academy of Sciences USA

101:1566115663.

Pearce, R. S. 2001. Plant freezing and damage. Annals of Botany

87:417424.

Philippe, M., and M. K. Bamford. 2008. A key to morphogenera

used for Mesozoic conifer-like woods. Review of Palaeobotany

and Palynology 148:184207.

Philippe, M., B. Gomez, V. Girard, C. Coiffard, V. Daviero-

Gomez, F. Thevenard, J.-P. Billon-Bruyat, M. Guiomar, J.-L.

Latil, J. Le Loeuff, D. Neraudeau, D. Olivero, and J. Schlogl.

2008. Woody or not woody? Evidence for early angiosperm

habit from the Early Cretaceous fossil wood record of Europe.

Palaeoworld 17:142152.

Pittermann, J., J. S. Sperry, U. G. Hacke, J. K. Wheeler, and E. H.

Sikkema. 2005. Torus-margo pits help conifers compete with

angiosperms. Science 310:19241924.

. 2006a. Inter-tracheid pitting and the hydraulic efficiency

of conifer wood: the role of tracheid allometry and cavitation

protection. American Journal of Botany 93:12651273.

Pittermann, J., J. S. Sperry, J. K. Wheeler, U. G. Hacke, and E. H.

Sikkema. 2006b. Mechanical reinforcement of tracheids com-

promises the hydraulic efficiency of conifer xylem. Plant, Cell

and Environment 29:16181628.

Rabaey, D., F. Lens, E. Smets, and S. Jansen. 2006. The

micromorphology of pit membranes in tracheary elements of

Ericales: new records of Tori or Pseudo-tori? Annals of Botany

98:943951.

Raup, D. M. 1966. Geometric analysis of shell coiling; general

problems. Journal of Paleontology 40:11781190.

. 1967. Geometric analysis of shell coiling: coiling in

ammonoids. Journal of Paleontology 41:4365.

Rothwell, G. W. 1975. The Callistophytaceae (Pteridospermop-

sida). I. Vegetative structures. Palaeontographica, Abteilung B

151:171196.

. 1982. New interpretations of the earliest conifers. Review

of Palaeobotany and Palynology 37:728.

Rothwell, G. W., G. Mapes, and M. R. H. 1997. Late Paleozoic

conifers of North America: structure, diversity and occurrences.

Review of Palaeobotany and Palynology 95:95113.

Rowe, N. P., and T. Speck. 2004. Hydraulics and mechanics of

plants: novelty, innovation, and evolution. Pp. 297326 in A. R.

Helmsley and I. Poole, eds. The evolution of plant physiology.

Elsevier, London.

Rowe, N. P., T. Speck, and J. Galtier. 1993. Biomechanical analysis

of a Paleozoic gymnosperm stem. Proceedings of the Royal

Society of London B 252:1928.

Ryberg, P. E., E. L. Taylor, and T. N. Taylor. 2007. Secondary

phloem anatomy of Cycadeoidea (Bennettitales). American


Sano, Y., and S. Jansen. 2006. Perforated pit membranes in

imperforate tracheary elements of some angiosperms. Annals

of Botany 97:10451053.

Schulte, P. J., A. C. Gibson, and P. S. Nobel. 1987. Xylem anatomy

and hydraulic conductance of Psilotum nudum. American


Scott, D. H. 1902. On the primary structure of certain Palaeozoic

stems with the Dadoxylon type of wood. Transactions of the

Royal Society of Edinburgh 40:331365.

Scott, R. A., E. S. Barghoorn, and U. Prakash. 1962. Wood of

Ginkgo in the Tertiary of western North America. American


Serbet, R., and G. W. Rothwell. 1992. Characterizing the most

primitive seed ferns. I. A reconstruction of Elkinsia polymorpha.

International Journal of Plant Sciences 153:602.

Sperry, J. S., and U. G. Hacke. 2004. Analysis of circular bordered

pit function. I. Angiosperm vessels with homogenous pit

membranes. American Journal of Botany 91:369385.

Sperry, J. S., and T. Ikeda. 1997. Xylem cavitation in roots and

stems of Douglas-fir and white fir. Tree Physiology 17:275280.

Sperry, J. S., and J. E. M. Sullivan. 1992. Xylem embolism in

response to freeze-thaw cycles and water stress in ring-porous,

diffuse-porous, and conifer species. Plant Physiology 100:605

613.

Sperry, J. S., and M. T. Tyree. 1988. Mechanism of water stress-

induced xylem embolism. Plant Physiology 88:581587.

. 1990. Water-stress-induced xylem embolism in 3 species

of conifers. Plant, Cell and Environment 13:427436.

Sperry, J. S., K. L. Nichols, J. E. M. Sullivan, and S. E. Eastlack.

1994. Xylem embolism in ring-porous, diffuse-porous, and

coniferous trees of northern Utah and interior Alaska. Ecology

75:17361752.

Sperry, J. S., U. G. Hacke, and J. K. Wheeler. 2005. Comparative

analysis of end wall resistivity in xylem conduits. Plant, Cell

and Environment 28:456465.

Sperry, J. S., U. G. Hacke, and J. Pittermann. 2006. Size and

function in conifer tracheids and angiosperm vessels. American

Journal of Botany 93(10):14901500.

Sperry, J. S., U. G. Hacke, T. S. Feild, Y. Sano, and E. H. Sikkema.

2007. Hydraulic consequences of vessel evolution in angio-

sperms. International Journal of Plant Sciences 168:11271139.

Steponkus, P. L. 1984. Role of the plasma membrane in freezing

injury and cold acclimation. Annual Review of Plant Physiol-

ogy 35:543584.

Stewart, W., and G. W. Rothwell. 1993. Paleobotany and the

evolution of plants. Cambridge University Press, Cambridge.

Sucoff, E. 1969. Freezing of conifer xylem sap and the cohesion-

tension theory. Physiologia Plantarum 22:424431.

Taiz, L., and E. Zeiger. 2002. Plant physiology. Sinauer, Sunder-

land, Mass.

Taylor, T. N., and E. L. Taylor. 1993. The biology and evolution of

fossil plants. Prentice-Hall, Upper Saddle River, N.J.

Taylor, T. N., E. L. Taylor, and M. Krings. 2008. Paleobotany: the

biology and evolution of fossil plants. Academic Press, San

Diego.

Terrazas, T. 1991. Origin and activity of successive cambia in

Cycas (Cycadales). American Journal of Botany 78:13351344.

Toml

08071.1

Documents