PeterW. Dunwiddie'
Quaternary ResearchCenter
University of \(/ashington
S e a t t l e ,w A 9 8 1 9 5
DichotomousKey to ConiferFoliage
in the Pacific Northwest
Abstract
Planr macroiossils offcr xn important meins fbr reconstr ctrn!! past lorcst composition in lhe Pacillc
Nofrhwcsr. Coniflif foiixge preservecl in sedirnents can bc iLleniific(l to species using charirclefs visible
undcr a alissecting microscope. A dichoromous kcy is prcscntcd thit describes these characters uscLl
in the idenrificetion of li)ssil conifer foliase irom lrees in westcm Washinqton
lntroduction
The forests in much of western Washington, Oregon, and British Coumbia are
presently dominated b)' conittrs (Franklin and l)yrness 1973).Nuncrous studies
havc traced the history of these fbrests since the retreat of Pleistoceneglaciers
(Hansen1!47; Huesser197J, 1977;Mathewesand Rouse1975;Tsukaclae/ r/. 1981;
Leopold et al. 1982). Most of this work has reliecl on fossil pollen preserved in
lowland lake and bog seclimentsto reconstruct changes in forest composition
within a radiocarbon dated time framework.
onc difficulty with this al]proach has bcen the inability to identit-v certein pollen
types to specics. Piceq, Abies, Pinus, and Cupressaceaeare sevefal taxa where
different species cannot bc distinguished on the basis ofpollen. Specific identifica
tions of pollen macleby somc authors have been considered unreliablc folkrwirrp;
more thorough analyscs(Mack 1971). Thefefore, recent rvork increasingly relies
on macrotbssil rcmains of plants preservecl ir association witlt tlte pollen to
distinguish species(Was,vlikowa1979, Birks ?nd Birks 1980). This technique has
levealedsome unexpected past speciesoccu.rences,such as the prcscnceof P/cea
eflgelmannii Parry and Abies lqsiacarpa Hook. (Nuit.) in the Puget Lo\.'land dur
ing thc late Pleistocene(Rarnosk,v1981).
Nlv rcccnt stuclieson Mt. Rainier have suggestcdthat past fofest col.nposition
can be reconstfticted using macrofossil asscmblagesof conifer foliage (Du1]widclie
1983). It was found that coriaceousconittr leavesare frequently well-presefvcd
in sediments tiom small ponds at most elevations. Furtl]ermore, since macrofossils
are usually identiflablc to species,past floristic changescould be detected even
wher], for exampie, scveral species of ,,l&/e.sgrew together.
In this papef, I present a dichotomous key to the m:rcroscopicfealuresof con
ittr foliage that is ffequently preserved in poncls in s,'esternBritish Columbia,
\Vashington, anal northern Oregon.
' P r e s e ^ t r c l d r e s s il . o s r F a r n . H u m m o c k P ( ) l l . l R d . . N a r t u c k e t . 1 1 A 0 2 5 5 4
Northwest Science,Vol. 59, No. 3, 1985
185
ldentificationof ConilerFoliage
Exccllent keys exist for idcntification of conifers in the pacit'ic Northwcst. based
on charactersevident on living material or on herbarium specimens(Hitchcock
and Cronquist 1973).However, thcse keys arc ofien unsuitable for distinguislting
fossil necdles, sincc critical choices i1t the key may rcly or.rcharacters unavaileble
to the paleobotanist-cone shape, size, leafarrangement on the branch, bark coktr,
etc. Evcn when kcv charactcrs describe foliage, thcy may be oflittle use with fossil
material. Frequentl]., macrofossils are brokcn or fragmented and important
features the shape of the leaf apex or base, fbr example mav bc missing. In quantitative studies, such as the Mt. Rainier work dcscribed above, it is critical tltat
all needles and fragments are identified, and not merely the occasional intact oDe.
A second body of literature relevant to macfofossil ident;fication includes
anatomicalstualiesof leaf structure. An excellent bibliograplt)' of this work, much
of which is in older and lesswell known public?tions, is presentcd b_yI)elcourt
et ql- (1979). These studies emphasizcmicroscopic cellular fcatures that are evi
clent in thin sections cut with a tnicrotome from moclern spccilnens.One limitx
tion of tltis approach with macrofossilmaterialis that poor preservatiol.r,at timcs,
makes resolution ofcritical finc cellular featuresdillicult or ir.npossibie.A sccond
drawback is the considerabletime and effort involved in prope ,v mounting and
sectioning material for study with a compound iight microscope. This becomes
prohibitive in studies where hundreds of needles are exanrincd.
These painstakingmicroscopic procedures vere found to be ururecessaryro
identify accurately the species of most cot.lifer foliage recovcred from poltcl
sediments in the North Cascades.The dichotomous key presenteclbcloq,. uses
macroscopicteaturesvisible with a dissectingmicroscopc at magnificationsof 7X
to 40x. It was prepared from observ:ttions made on fiesh and driecl foliage col
lected in westcrn Wasltington, in conjunction with several arratomicalstudies
( C o u l t e ra n c l R o s e 1 8 8 6 , D u r r e l l 1 9 1 6 , F u l t i n g 1 9 j 4 , H a r l o w 1 9 4 7 , L i u l 9 7 t ) .
Revisions of the ke)' were made on the basis of featur.esltoted ffom examina_
tion ofhundreds of macrofossil needles in the Mr. Rainier studies (Dunwiddie 1983).
Caretul study of macrofossils revealecl that several fe:rtures could be seen in pre_
served neecllesthat nornall]' are nor visible in unsectioned fresh mtterial. Sonle
important teaturesare the presence,size, end position of resin caltals.These are
frequently visible through the epidermis, which is often translucent in fossil
needles. Identification of spccies of ..lbies, Tsuga, pseudotsug.t, and pinus ltll rely
on these fe;rtrues.Thus the emphasis of the key is on fcatures visible in fossil
needles, although it can be uscd on fresh material as weli.
As $'ith other vegetativekeys, rhis kcy at tjmes deviatesfrom a strict sequencc
tnat separatesspcciesaccording to families, genera, ancl thcn species.f his is because
when certain features are missing in fossil material, the resemblancebetween
specics in differcnt genera ma_vbe stronger than betx,-eencongeneric species. f,he
key has becn constructed so that m?n1.species mav be identiticcl throuElh clifferent
Ieads.This ensuresthat a proper idcntification can be macle,even when an apex
or base may be missinEifrom a ncedle.
DicholomousKey to ConiferFoliage
1a Leavesscale-like,lcss than 5nn long; single or more often severalrnan\.pairs
oppositely attached.
Cupressiceae
186
Dun$.'iddie
1b Leavesncccllc like, entire needles rarely less than 5mm long; single (l]]ay be
Pinaceae
fasciclecl in 1'l/?rs).
LUpress?ceae
N{ost fbliagc rcmains :rttached ro small branches. In sediments, individual leaves
may b|eak off, but usually some fragments rcnain with several leavcs attached.
Leaf arrarlgenrent on these nrultiple leaved fragments is a useful character ir.l
distin€juishingthe two trcc genera.
1a Multipie-leavedfra€imentsclistinctlyflattened on all but tlte smallestbranchcs;
apex of dorsi-r'eltral lcavesoften obtuse, abruptly narrowing to apiculatetip;
lateral leavcs rounded on back, tapering gradually to apcx; tips of lateral and
Tbujtl plicqta I)onn.
dofsi-ventrai lcaf pairs close to same level.
1b Multiple-leaved fragmcnts rounaled, 4-angled, or only somewh2t flattened
(smallest branches may be indistinguishable ftom Thuja); dorsi-ventral and
lateralleavcssimila., gfacluallynarrowing to acute apex, often keeled on back;
tips of latcral and dorsi-ventralleaf pairs "t different levels (may be sub-equal
Cham.declparis nootkqtensis (D. Don) Spach
o n s m ; l l c s tb r r n c h c s )
Pinaccae
1a Needles triangular or semi-circular in cross-section, with sharp edges; tips upering to sharp point; basesnot greatly differentiatedfrom needle body; necdles
somelimcs attached to wood,v sllortshoot, tbrming fasciclesof 2, 3, or 5;neeclles
parallci sicled, linear. oftcn breaking cleanlv into short segmentsj intact neeclles
Pinus (p. 189)
rarely preservecl.
1b Neccllesvarioush' flattened or 3- or '1 sided, often with a medial groove on adax
ial surface, if triangular in cross-section,with rounded edges; tips notched.
rounded, or acute; basespctiolate or apetiolate,often with conspicuous ov?l
or circular attachment scar; necdles always single, parallel-sided and linear only
2r
rr'r Lq ri.r.
2a Ncedlesclelicatcand lir.Iear(lcsstltan l mm thick), sometimeswith a short,
hooked pcti(te; tips acute to obtusc; needles flattened, but usually with a
prominent to obscure ridge on the ab- and 2daxial sides; 2 stomatal ros's
l|ay be apparent on either side of the midrib on both ab- and ad2xial sru
faces;usually preservcd relativel,vintact (contrast wlth I'inlLs). I ctrlx (p.
189)
2b Needleslaeneralllrrobust, neither slender nor linear; other charactersvarious.
h u t l ] , , l c . , m D i r r e r Jl : i n / a / i . r 0 ) . l 8 t )
3a Ncccllessharpl,v:lcLltc,x'jtltout a medial groove o11lhe adaxial sur-face;
oftcn relatiYelystraight ancl pungent; base abruptl)' truncated at poil]t
Picea (.p. 19O)
tapered.
of attachnent, or only slig;htl_v
(if
from
cone
bearir.lg
branchcs);
3b Needles blunt, notchecl, or acutc derived
more or
surfacej
alwa-Ys
groove
less
evident
on
adaxial
m<xe or
medi:rl
bllt
not
truncated.
t*'isteal
or
straight,
lessdorsi-r.enually'flattened;base
1a
.1a llcsin canals 1, central near abaxiel surtace; needles narrowecl to a
slender petiole; tip obtusely rounded but not notched (if char?ctefs
Tsug.l (p. 19O)
not appzfent, go to 4D).
Dichototnolls Kev to Conifer Foliage
187
4b Resin canals 2, laterally arranged on opposite sides of central vascular
bundle; petiole little constricted, or slender in Pse4dotsuga; tip ob-Abies and Pseud.otsuga
tuse, acute, or notched.
Abies, Pseudotsuga,
and Tsuga
1a Needle tips notched; point of arrachmenr round and broad (if rip absem, go
_2a
to lb, p. 188)
2a Stomata present on ab and adaxial surface; resin canals remote from lateral
m a r g i n s rp e l i o l e l i l t l e c o n b t r i c r e d o
, ften noL twisted.
3a Needles frequently C-shaped, curving adaxially; resin canals large, medially located between ab- and ada-xialepidermises, but usually not evident near ne€dle base; adaxial stomata in 10-13 irregular rows forming
a broad band ll2 to 213 the widrh of needle surface, often absenr near
base, but present very close to tip; medial adaxial groove prominent;
abaxial stomata in 4,8 rows forming bands on either side of prominent
midrib; abaxial margins and midrib free of stomata, and often equal in
width to stomatal bands; abaxial epidermis usually concave berween
- Abi$ Iasiocarpa (Hook.) \urr.
midrib and margin.
3b Needles S or hockystick shaped; resin canals small, near abaxial epider
mis, about midway between midrib and margin; adaxial stomata in 2
distinct bands on either side of shallow medial groove, merging to form
a single band near fip; abaxial stomata in 4 bands of 3-5 rows each,
separated by midrib and resin canals; stomate-free margins relatively narrow; abaxial epidermis nearly flat between midrib and margin, hence
needle cross-section more or less triangular .Abies procerq. Rdndet
2b Stomata present only on abaxial surface (occasionally a few on adaxial tip);
resin canalssmall, close to abaxial epidermis and lateral margins; base usually
twisted, sometimes hooked or straight.
-Abies camabilis (Dougl.) Forbes arrd.Abies grandis (Dougl.) Forbes
(Needles ofthese 2 species are difficult to distinguish. Liu (1971) reports
that the hypodermis is more conspicuous and well developed in /.
amabilis than in l. grandis, but microtome ctoss sections would be
necessary to examine this feature.)
1b Needle tips obtuse, acute, or absent; point of attachment round or elliptical.-4a
4a Stomata present on ab and adaxial surface
5a Needles more or less straight, not C or S-shaped;petiole slender;
point of attachment small and elliptical.
5b Needles C or S-shaped;petiole broad; point of attachment round
and broad (needles with acute tips from fertile branches of Abies
I.tsiocarpa and A. procera may key out here
-see leads 3a and 3b, p. 188)
-6a
4b Stomata present only on abaxial surface.
6a Needles with slender petiole; poinr of attachment eliiprical,
longer than broad; tip rounded, not notched.
7a Needle length:width ratio less than 10; needle abruptly
contracted into petiole.
Tsuga beterophJ)lla (p. 190)
188
Dunwiddie
7b Needle length:width ratio greater than 10; needle gradu
ally tapering into Petiole.
Pseudotsuga menziesii (Mirbel) Franco
(These 2 species are normally teadily distinguished by the
n u m b e r o f r e s i n c a n a l s- l e a d 4 a . p . 1 8 8 ) .
6b Needles with broader petiole; point of attachment nearly
round; tip various. -Abies affiabilis, A. Grandis, and Pseudo
tsuga menziesii (See lead 2b, p. 188 regarding these Abies
species.Pseudotsuga is normally distinguished by its slender
petiolate base and rounded, unnotched tip. Abies tips arc
usually notched, or occasionally acutely pointed on needles
from fertile branches. If both tip and base are absent,
distinguishing these taxa is difficult, although Pseudotsuga
ne€dles tend to be slenderer than these,4&les species. Durrell
(1916) reports that Pseudotsuga has 1 vascular bundle,
w|'ercas Abies has 2. This feature would require microtome
sectioning to see.)
Pinus
I
Entire needles are tarely found in sediment. Occasionally the lower portions of
the fascicles are recovered still attached to woody shortshoots, or the shortshoots
alone may be found. Attachment scars are preserved on these shortshoots, such
that 2, 3, and 5 needle species can be readily distinguished fiom each other'
1a Ne€dles semi circular in cross section, adaxial sufface somedmes concave; convex abaxial surface with 9 10 (15) distinct, evenly spaced rows of stomata across
entire surface; stomata regularly spaced within each row as well; adaxial stomata
(rarely absent) in 6 8 (11) similar distinct rows; resin canals 2, medial, usually
not apparent through epidermis; minute marginal serfations may be present
-Plnus contortct Dougl.
entire length of needle.
lb Needles 3-angled in cross-section; abaxial stomata present or absent. -2a
2a Abaxial surface broadly convex, much longer than either adaxial surface;
other featuressimilar to P. contorta, although presence of adaxial ridge form
ing 2 adaxial surfaces, and overall larger size, are diagnostic.
-Pinus Ponderosa Dougl.
2b Abaxial surface not much longer than either adaxial surface, needles appear
ing nearly equiangular in cross section; resin canals nearer to abaxial epider
3a
mis. hence often visible in unsectioned material.
3a Stomata usu"lly present only on adaxial surface; when rarely present
on abaxial surface, then only near tip; margins minutely serrate up to
-Pinus monticola Dougl.
1/2 length of needle.
3b Stomata present on both abaxial and adaxial surfaces; 1 to several rows
of stomata present along most of abaxial surface; margins minutely serP inus a lb i cau I i s Engelrri,.
rate only near tip.
LArlX
Disdnguishing Z. occidentalis Nutt. frcm L. lyallii P^tl. is difficult based on needle
morphology alone. Carlson and Blake (1969) report 8 epithelial cells surrounding
Dichotomous Key to Conifer Foliage
189
resin canals in Z. occidentalis,5 in L. b)dllii. This feature would require microtome
sectioning to see.
Picea
la Needles more or less flattened; adaxial surface planar or slightly ridged, with
numerous rows of stomata; abaxial surface with prominent midrib, without
stomata, or with only 1 or 2 rows on either sid€ of the midrib; resin canals
absenr.
_picea sitcbensis (Bong.) Cafr.
lb Needles nearly equally 4-angled; stomata abundant in 4 gfooves between angled
edges of needles; I or 2 large resin canals usually present (needles usually must
be sectioned to see resin canals).
_pice.1 engelmannii p"rry
Tsuga
la Needles strongly flattened; stomata only on abaial surface; adaxial surface
usually with a conspicuous medial groove; petiole very sl€nd€r and abruptly
twisted; dp obruse.
- Tsugq beteropbllla tRal.t catg.
1b Needles usually only somewhat flattened; stomata on both ab ancl adaxial
sur_
faces; medial groove present or inconspicuous; petiole often more gradually
tapered, less slendef, and not as abruptly twisted as 1a; tip often more acute.
-Tsuga rtertensiana (Bong.) Carr.
Acknowledgements
Financial support for this study was provided by a doctoral dissertation grant from
the Nadonal Science Foundadon (DEB-91074p1), as well as by awards from the
Danforth Foundation, the Geological Sociery of America, the American Alpine Club,
and Sigma Xi.
LiteratureCited
Bamosky, C. W. 1981. A fecord oflate euaternary vegerarion from Davis Lake,
southern l,uget Lowtand,
W a s h i n g t o n . e u a a . R e s _t 6 1 2 2 t - 2 3 9
Birks, H- J. 8., and H. H_ Birks. 1980.
euarernary palaeoecology. Edward Arnolct, London.
Carlson, C. 8., and G. M. Blake. 1969. Hybridizarion of sresterriand subalpine
tarch. Monr. For. and
Cons. Exp. Sra_Bult. 37, Univ€rsiry of Montana, Missoula. pp. 1 li.
Coulter, J. M-, andJ. N. Rose. 1886. Synopsis ofNorrh American pinei, based
upon teai anatomy. Bot.
Gaz. 1\:256 262, 3O2-2O9.
Delcourt, P. A., O. W. Davis, and R. C. Brighr. 1979. Bjbtiogrrphy oi raxonomic
trrefature for rhe iden_
tification of fruits, seeds, and vegerative plant fragments.-OnNffM
681S, Oak Ridge Nationxl
Laborarory, Oak Ridge, Tennessee.
Dunwicldie, P. W. 1983. Holocene foresr dynamics on Mounr Rainier, Washingrcn.
Llniversiry of
Washingron, Seatde. Dissertarion.
Durrell, L. W. 1916. Nores on some North Americen conifers based on leafcharacters. proc.
Iowa Acaci.
Sc\.23:519 5a2.
Fr2nklin, J. F., and C. T. Dym€ss. i 973. Narural vegetarion of Oregon and \V{shington.
ceneral Tec-hnical
Reporr PNW-8, U.S. Foresr Servjce, \yashington, D.C.
Fulling, E. H. 1934. Identification, by leai structure-, of the sp€cies ofAbies
cultivxted in the United
S t a r e s . 8 u 1 1T. o r r . B o r . C l u b 6 1 ( 9 ) , 1 9 7 - 5 2 4 .
Hansen, H. P- 1947. Posrgiacial foresr succession, climare and chronology in
rhe pacific Norrhwcst.
Trans. Am. phil. Soc. 37:t ll0.
Harlow, W. M. 1931. Th€ idenrificarion of rhe pines of the Unired States,
native and introduced, by
needle srrlrcture. N.Y. Srate Colt For. ar Symcuse Univ., Tech. pub. No.
32. (Aull Vol. XX, N;.
2), 19 p.
19O
Dunwiddie
Heusser, C. J. 1973. Environmentel sequ€nce fotlowing the I'raser advance of the Juan de Fuca lobe,
washington. Quat. Res. 3:284'301r.
. 1977. Quaternxry palynology of the Pacific slope of washington. Quat. Res. 8:282-306.
Hitchcock, C. L., ancl A. Cronquist. 1973. Florz of the Pacific Northwest. Univefiity of lvashington
Press. Seattle.
Leopold, E. 8., R. Nickmann, J. L Hedges, andj. R. Ertel. i 9a2 - Pollen and lignin records of late Quatern a r y v e g e t a t i o n ,L a k e W a s h i n g t o r } .S c i e n c e2 1 8 ; 1 3 0 5 - 1 3 0 7 .
Liu, Tang Shui. I971 . The g€nus ,4rtes. Taipei, Taiwan, Mus. Quart. J. 24:lO5 172.
Mack, R. N. 1971 . Poll€n size variation in some west€rn North Amedcao pines as related to fossil pollen
identificetion. Northw. Sci. 45:251 269.
Mathewes, R. \{/., and G. E. Rouse. 1975. Palynology aod paleoecology of postglacial sediments from
tlre lower Fraser River Canyon of British Columbia. Cxn. J. E^rth Sci. l2:7a5'756.
Tsukeda, M., S. Sugita, and D. M. Hibbert. 1981. Paleoecology in the Pacific Northwest L Lale Quater
nary vegetarion and climate. Verh. Interna!. Verein. Limnol. 2l:73of1'7.
wasylikowa, K. 1979. Plant macrofossil analysis. 1t, B. E. Berglund (ed.) Palaeohydrological Changes
in the Tempemte Zone in the Last 15000 yea$. Subproiect B. Lake ancl Mire Environments
Vol. 2. Guideline for Specific Methods. Iflternetionxl Geological Correlation Programme Proj
e c t 1 5 8 .P p . 2 9 1 ' 3 1 J .
ReceiueqlI Marcb 1984
Accepted for publication
7 June 1981
D i c h o t o m o u . K ey t o C o n i f e r F o l i a g e
lo I