rlcosclei-ic-*.-, /
6 (3) : 47r-474(1999)
which commonindicesof sclerophyllybestreflect
differencesin leaf structure?'
philip k. groom
z, centre for ecosystemmanagement,school of natural sciences,edith cowan university, 100joondalupdrive'
joondalup 6027, australia, email: p.groom@cowan'edu'au
nyion b. lamont, school of environmental biology, curtin university of technology' gpo box ul98?' perth 6t[l' austra]ia'
abstract: when describing the sclerophyllous nature of leaves' two indices are most commonly cited: fibre:protein ratio
(fpr), better known as the f-o""f"r, ,i"rophylly index; -j l"ui -u,t per unit area-(lma)' or its inverse' specific leaf area
(sla). here, we assessthe jative i-port rrc" of th"." *o indi"", in accounting for changesin leaf structurr:' the primary
basis for variations in sclerophylly. fpr compares ,ou"*t (i'e., lignin and-cellulose lcrude fibre]) to oon-structural
(i.e., protein = protoplasm) i"li i#"arf, on the basis that in$easing sclerophylly is associated with a greater contribution
of crude fibre and smaller contribution of protein ,o ,o,"1 dty weigh't. ho*iuit,ioisin" the crude fibre content is just one
way of increasing sclerophylly, and a dccrease in the nitrogen content (i'e', protein)-does not conaibute dlectly to thc
impression of leaf hardness.vnlr" ppn lacks a clear -uio#"ur basis, it may provide a biochemical interpretation of scle-
rophylly. in contrast, r_r"re i, ,rr"
"-rs
product of leaf thickness and leaf density, two (often independent) attributes that are
linked to different
"o*pon"no
oi u leaf s anatomicavstructural attributes' wi show that fpr'and lma are often poorly
correlated and conclude that lma is a more useful measure of sclerophylly, especially when thiclsress and density are known'
i"y*irat, sclerophylly, leaf massper area,fibre:protein ratio' leaf structure'
r6sum6:deux indices sont souyentutilis6s pour caracteriserle degr€ de scldrophyllie.desfeuilles : le rapport fibres:prot.lines
(rfp), mieux
"o*r, "o*.
ilindice de scl6rophyllie de i-ovelesi, et la massi foliaire par unit6 de surface (mfs)' ou son
inverse la surface foliaire rpe"tftqr". nous 6valuons ici ta capacii6 ae ces deux.indicei ir expliquer les changementsci'l la
structure foliaire ir la u"r" rrier* des variations de sclerophyllie. tr rfp permet de comparer le matdriel structuraa (ligntnt: et
cellulose) au mat6riel non structural (prot6ines)ae u feuittl' en pr6sumant qu'une augmentationde la scl6rophyllie est asso-
ci6e ir une plus grande contribution des fibres
"t
un" ptu, rniut, cintribution o"s prot6ines d la masse totale sdche' cependa[t'
accroitre le contenu * nu.", ,r;"rt qulurr" d", fagoris d'augmenter.la scldrophyllie et une diminution du contenu en ^zote (elt
prot6ines) ne contribue p"".alr."t"ro"i.t e la r6sistance ae h iluille. bien que ti nfp ne se €f€re pas i une origine anatomique
claire, il peut fourn.u une rnterprdtation biochimique ae la scl6ronhlll*l.l-
":1::^t:
mfs est le produit de l'€paisseur de
la feuille et de sa densit6, deux caract6ristiques souvent ind6pendantes et associdesi des composantes diff6rentes de
l,anatomie et de la structure de la feuille. nous-d6montrons que te ng et le mfs sont souvent peu corr6l6s et concluons que
le mfs est une m"rur" ptu, otil" de la scl6rophyllie, surtout iorsque l'6paisseur etla densit6de la feuille sont connues'
mots-cl€s: scl6rophyllie, -"rr" i"ii"* p* utti't6 a" surface, rapport fibres : prot6ines, structure foliairc'
introduction
sclerophyllyis a termusedto describeleavesthatare
tough,stiff oi teattreryin texture(schimper,1903;seddon'
9fu). literally meaning
'hard-leaved',sclerophyllywas
thesemeasuresis that a changein their magnitude coincides
with a matchedchangein the level of sclerophylly' but how
accurateare theseindices in quantifying sclerophylly? how
well do they take into account variations in leaf structure'
the primary-basis for variations in sclerophylly? it is not the
intention bt ttris paper to review the background to all
degreeof sclerophylly. clearly, a good index
9f
a biological
coicept must accuiately reflect the variable, be suitable for
all miterial under study' and be relatively easy to measure'
various indices of sclerophylly have been proposed'
including leaf moisture content (loveless, 1961)' dry matter
content-(kalapos,1994),leaf thickness (mooney et al''
1982), leaf fraiture toughness(lucas & pereira' 1990)' and
anatomicalmeasurements,suchasrelative palisadethickness
(grubb, 1986). however, the fibre:protein ratio (better
k ro*n as the loveless sclerophylly index) and leaf mass
per area(lma), or its inverse specific leaf area (sla)' are
ly far the most widely used. the assumptionunderlying
as a purely descriptive term, requiring some
measurableindei for inter- or intraspecific comparisonsof
trec. 19984645; acc. 1999-01-06
2auths fq corespondene
of sclerophylly cited in the literature, but rather to
summarisethe featuresof the fibre:protein ratio (fpr) and
ivfe (ot sla) that underlietheir usageasmeasuresof
scleropirylly,andcommentontheirusefulnessin interpreting
thestruauialchangesassociatedwith sclerophylly'
fibre: protein ratio
this measureof sclerophyllywas proposedby
loveless(1961)afterhe comparedleaf chemistrydata
u"t*""n speciesdiffering in leaf texture'originally' he
thoughtthat crudefibre content(thatpart of dry matter
in*otiutein strongacidandalkali)wouldprovidethemost
usefulcomparativlmeasureof sclerophylly,whichheactu-
ally believedwas bestrepresentedby leaf dryness'
gnoov & leuot.rr:coupnr:ncin'dlcfsofsclerophylly
however, when comparinga selectionof
'mesophytic'
and
'sclerophyllous' leaves,lovelessfound that differences,in
fibre concentrationof the two leaf typeswere not great,and
could be partially explained by the fact that crude fibre per
dry weight doesnot take into accountother chemicalcom-
ponentsof the leaf. assuming that fibre was an important
component of sclerophylly, lovelessproposedthat any
estimate basedon fibre should be compared with a non-
structural component,either leaf fresh weight or somemea-
sure of leaf protoplasm.leaf fresh weight was dismissed
becauseit is unstable,dependingon weatherat time of
harvest and subsequenttreatment of the leaf. crude protein
content (calculatedas nitrogenconcentrationmultiplied by
6.25) was consideredto reflect the part of dry leaf tissues
that doesnot contain fibre. the ratio of crude hbre to crude
protein content(fpr) was then usedasthe standardmeasure
of sclerophyllyby loveless(1961; 1962).fpr has also
been used as a measureof leaf palatability to herbivores
(waterman et al., 1988;choong et al., 1992). however, the
concepts of palatability, digestibility and sclerophylly
should not be confused(turner, 1994a).
does fibre content adequatelyreflect differencesin leaf
properties as they affect sclerophylly? this might be true if
cell wall thickening and the presenceof sclerified tissues
was the only meansof increasingsclerophylly. but thereare
alternative ways of increasing sclerophylly that do not
necessarilyinvolve an increasein crude fibre or a decrease
in protein. this can be achieved by decreasingthe size of
cells without changing cell wall density or cytoplasmic
content,or by impregnatingexisting cells with extra cutin
or crystals,in place of cellulose or lignin. increasingthe
number or size of cells may changeleaf dimensions,inde-
pendentof their histochemistry.such anatomicalchanges
may be accompaniedby a reduction (groom, lamont &
markey, 1997)or increase(shipley, 1995)in leaf area,but
more certainly by increasesin leaf thicknessand/or density.
independentof areaor density, a thicker leaf will be tougher
than a thin one.
leaf mass per unit area
lma (previouslyknown asleaf specificmass)is calcu-
lated as leaf dry mass divided by projected leaf area. the_
inverse of this is sla. we are not aware of the history of
these indices,but mooney et al. (1977) used lma as an
index of sclerophylly over 20 years ago. both lma and
sla may be consideredthe products of leaf density and
thickness(witkowski & lamont, 1991),which are directly
related to structural properties of the leaf. lma has an
advantageover sla in that it is directly, rather than
inversely,relatedto sclerophylly.
witkowski & lamont (1991) noted that variations in
leaf densitymay be the resultof differencesin: (i) thickness
and densityof the cuticle and cell walls, (ii) inclusionsin
the cells (starch grains, crystals), and (iil) extent and abun-
dance of air spaces,sclereids,fibre groups and vascular
bundles.variations in leaf thicknessmay be dueto variations
in leaf shape,number of layersand lengthof palisadecells,
width of rest of mesophyll, epidermis, and hypodermis, and
placementof veins.thus, the two componentsof lma rep-
resent different structural aspectsof the leaf and should be
472
independent,this is demonstratedin fig:urel: lma
were not significantly different within leaf age clas
two co-occurringbanksiaspecies,but the,"u*ns we€,i
ferent.b. petiolarishasvery thick leavesof relativdf i
density while b. baueri has thinner leaves of high d
leaf density increasedwith agein both specieswhile
ness remained static. thus, for these two species,
density, andthicknessarenot correlated.
bulk densityhasbeenshown to accountfor vari
lma for a number of species,with most studies
on grass species(garnier & laurent, 1994; van
& poorter, 1994). most studiesthat relate changesin
to leaf thicknessfailed to calculate density fronr the dad.:
(abrams, 1994).witkowksi & lamonr (1991) showedrhai,;
for hine shrub speciesinhabiting a range of soil typ"s, the ,
6. 350
9f 3oo
.r 250
e!
q
o
5 0.8
o
o
i
t 0.7
fr
0.6
200
650
550
450
350
250
i
0.9
0 1 2 3 4 5
leaf age(years)
figurel. variations in leaf mass per area (lma) and its components,
leaf thickness and density, in co-occurring banksiapetiolaris (open circles)
and b. bauei (closed circles) in one- to five-year-old leaves. bars are se
for l0 replicatcs (witkowski et al., 1992).
0.5
lma of r::: :pecieswas linearly correlatedwith density,
trrres;r\. -.:i,-:r,e:s.asrdonesrt$rrbo{ndens\\yandt\,rc\ness.
comparingthetwo indices
in:re arerwo issuesto be considered:(i) whetherfpr
anrj l\l{ (or their components)are alwayscorrelated,and
riir rf ure)'arenot, which is a more valid index of sclero_
ph1lii i omitting conelationsbetween subsetsof the data,
rnere \\ere no overall relationshipsbetweenfpr and sla
or leaf thicknessfor monocotsand dicots in southern
.\usrraiia (specht& rundel, 1990).for a much largerdata
sel from many partsof the world, there was no correlation
between sla and fpr or crude fibre, but there was
between sla and nitrogen (turner, l9g4b). for 42 tree
species,choonget al. (1992) showedthat fpr was corre-
lated with sla and densirybut not with thickness,and that
sla was correlatedwith nitrogen but not with fibre. we
conclude that fpr and lma arepoor indices of eachother,
especially when attempting to interpret the results in terms
of their components.
although loveless(1961)proposedthatproteincontent
provided an independentbasisfor comparing fibre contents,
this and subsequentwork have usually shown a (negative)
relationship betweenthe two (steubing & albardi, 1973;
specht & rundel, 1990). where they are starisricallyinde-
pendent, there is the problem that a high fpr could be due
to a high fibre concentration(directly relatedto sclerophylly)
or a low protein content(not directly relatedto scteroptrytty;.
fp.rmay, however, provide us with a biochemical interpie-
tation of sclerophylly. highly sclerophyllousfloras are usu-
ally associatedwith nutrient deficient soils, particularly
those low in phosphorusand nitrogen (beadle, 1966;
lamont, 1994).becausephosphorusand nitrogen are essen-
tial requirementsfor protein synthesis,carbon-rich metabo-
lites which may havebeenusedto form protein aretherefore
diverted to form other compounds, including cellulose and
lignin. as a result, fpr may be viewed asmore of a measure
of metabolic efficiency than of sclerophylly. in addition, a
nutritive interpretation is of little value if varying levels of
sclerophylly are due to differencesin lieht or water avail-
ability rather than nutrient availabilit! (oertli, lips &
agami, 1990;groom& lamont, 1997).
lma respondsonly to structural changesrather than to
both structural and chemical changes.it is much more eco-
logically useful that nitrogen is correlated with the measure
ofsclerophylly (seeabove)ratherthan being a componentof
it, so that a cornmon cause(low nutrients)is distinguished
from its morphological effect. at best, fpr representsthe
density componentof sclerophylly (as do the water/dry
matter approaches)but ignores leaf (essentiallymesophyll)
thickness as an essentialcomponent of it. for example, it is
generally agreedthat adult and needle leavesaremore scle-
rophyllous than conspecific seedlingandbroad leaves.their
lma values support this impression, but the reasonis that
adult and needle leaves are thicker rather than denserthan
conspecific seedling and broad leaves(groom, lamont &
kupsky, 1994;groom, lamont & markey, 1997).
the major drawback with lma is that its measuremenr
is basedon leaf massand area,and yet its interpretationas a
measureof sclerophylly requires a knowledge of the two
ecoscence,vor-.6(3),1999
other componentsof mass:thicknessanddensity(witkowski
& lamon\, \99\, turner, 1994b). conceptual\y, it assumes
that thesetwo pathwaysareof equal weight in influencing a
change in the degreeof sclerophylly. for example, the
{egr9e
of sclerophylly may be doubled either by dbubling
density or thickness (or a combination of the two). in thii
regard, it is no different from fpr. thickness is seldom
determined(althoughit is simply done with vernier calipers
or a spring-loadedscrew gauge)and density almost niver
(although it is simply calculatedaslma/thickness).
,
there may be practical problems with flattening out
broad leavesto obtain area,whereasericoid leavesare best
treated as cylinders and a correction factor applied
(witkowski & lamont, l99l). when a prominenr midrib is
involved, the lamina should be used for all measurements,
and judgement about leaf thickness may be needed when
scatteredprotruding veins are involved. when shrinkage or
drying is likely (e.g., immature or succulentleaves)it is
essential that thickness and area are measured on fresh
(turgid) material before adequatedrying. it is important that
the density of such leavesbe calculatedon the original
dimensionsto ensureit is not overestimated.in this regard,
thick leavesthat collapseon drying cannot be considered
very sclerophyllous,although the hydrostatic pressure
associatedwith their turgidity may give the impression of
reasonablehardness.
t-:---.^ri., r rr rlrcrieiatiy, r-ivr.r and its componenis are simple to
measure,may be measuredon an individual leaf basis for
all leaves,the methodsare less destructiveand expensive
than other techniques,and replication and subsamplingare
easier.in addition, density and thickness may also be con-
sidered as (charles-edwardset al., 1986) contributors to
photosyntheticpotential, as well as essentialindices of
palatability (waller & jones,1991).in the absenceof a
thorough anatomicalleaf analysis,we advocateleaf mass
per unit area,togetherwith a knowledge of its two compo-
nents, thicknessand density, as the best currently available
measureof sclerophylly.
literature cited
abrams,m. d., 1994.genotypicandphenotypicvariationas
stressadaptationsin temperatetreespecies:a reviewof several
casestudies.treephysiology,14:833-842.
beadle,n. c. w., 1966.soil phosphateandits role in molding
segmentsof the australianflora andvegetation,with special
referenceto xeromorphyandsclerophylly.ecology,42:992-
1007.
charles-edwards,d. a., d. doley& g. m. remmington.,1986.
modellingplantgrowthanddevelopment.academicpress,
sydney.
choong,m. f., p. w. lucas,j. s. y. ong,b. pereira,h. t. w.
tan & i. m. turner,1992.l,eaf fracturetoughnessandsclero-
phylly: theircorrelationsandecologicalimplications.new
phytotogist,121: 597-610.
gamier,e. & g. laurent,1994.l*af anatomy,specificmassand
watercontentin congenericannualandperennialgrassspecies.
newphytologist,128:725-736.
groom,p. k- & b. b. lamont,1997.xerophyticimplicationsof
increasedsclerophylly:interactionswith waterandlight in
hakeapsiiorrhyncha seedlings.newphytologist,i36: 23l -237.
47t
gnoolt & leuor'r: cotwerinc indicesof sclerophylly
groom, p. k., b. b. lamont & l. kupsky, 1994.contrastingmor-
phology and eco-physiology of co-occurring broad and terete
leaves in hakea trifurcata (proteaceae).australian joumal of
botany,42:307-320.
groom, p. k., b. b. lamont & a. s. markey, 1997.influence of
leaf type and plant age on leaf structureand sclerophylly in
hakea (koteaceae). australian joumal of botany, 45: 827-838
grubb, p. j., 1986. sclerophylls,pachyphylls and pycnophylls:
the nature and significance of hard leaf surfaces. pages 137-
150 in b. e. juniper& t. r. e. southwood(ed.). insectsand
the plant surface.edward amold, london.
kalapos, t., 1994.leaf water potential:leaf water deficit rela-
tionship for ten speciesof a semiarid grasslandcommunity.
plantandsoil, 160:105-112.
lamont, b. 8., 1994. mineral nutrient relations in mediterranean
regions of california, chile, and australia. pages2ll-234 in
m. t. kalin arroyo, p. h. tndlet & m. d. fox (ed.). ecology
and biogeography of mediterraneanecosystemsin chile,
california, and australia. springer-verlag,new york.
loveless, a. r., 1961. a nutritional interpretation of sclerophylly
basedon differencesin the chemicalcomposition of sclerophyl-
lousandmesophyticleaves.annalsofbotany, 25: 168-184.
loveless, a. r., 1962. further evidenceto support a nutritional
interpretationof sclerophylly.annalsof botany,26: 551-561.
lucas, p. w. & b. pereira, 1990.estimation of the fracture tough-
nessof leaves.functionalfrology, 4:819-822.
mooney, h. a., j. kummerow, a. w. johnson,d. j. parsons,s.
keeley, a. hoffman, r. i. hays,j. giliberto & c. chu, 1977.
the producers: their resourcesand adaptive responses.pages
l-12 inh. a. mooney (ed.).convergentevolution in chile and
california: mediterranean-typeecosystems.dowden,
hutchinson and ross,stroudsberg.
mooney, h. a., j. kummerow, e. j. moll, g. orshan, m. c.
rutherford & j. e. m. sommerville, 1982.plant form and func-
tion in relation to nutrient gradients.pages55-76 ln j. a. day
(ed.). mineral nutrients in mediterranean ecosystems. south
african natural scienceprogamme,rep. 71, csir, pretoria.
oertli, j. j., s. h. lips & m. agami, 1990.the strengthof sclero-
phyllous cells to resist collapse due to negative turgor pressure.
acta oecologica,17; 281-289.
schimper,a. f. a., 1903.plant geographyupon a physiological
basis.(english translationby w. r. fisher). claredon press,
oxford.
seddon,g., 1974.xerophytes,xeromorphsand sclerophylls: the
history of some conceptsin ecology. biological joumal of the
linneansociety,6: 65-87.
shipley, 8., 1995. structured interspecific determinantsof specific
leaf areain 34 speciesof herbaceousangiosperms.functional
ecology,9:312-319.
specht,r- l. & p. w. rundel, 1990. sclerophylly and foliar nutri-
ent statusof mediterranean-climateplant communities in
southernaustralia. australian joumal of botany, 38:459-474,
steubing,l. & m. alberdi, 1973.the influence of phosphorusdefi-
ciency on the sclerophylly. oecologia plantarum,8:211-218.
turner, l. m., 1994a. sclerophylly: primarily protective?
functionalfrology, 8: 669-6'15.
turner, i. m., 1994b. a quantitative analysisof leaf form in woody
plants from the world's major broadleavedforest types. joumal
of biogeography,2l: 413-419.
van arendonk, j. j. c. m. & h. poorter, 1994.the chemical com-
position and anatomical structure of leavesof grassspeciesdif-
fering in relative growth rate. plant, cell and environment, 17:
963-970.
waller, d. a. & c. g. jones, 1991. measuring herbivory.
ecological entomology, 14: 479481
waterman,p. g., j. a. m. ross, e. l. bennett & a. g. d?rvies.
1988.a comparison of the floristics and leaf chemistry of the
treeflora in two malaysian rain forestsand the influence of leaf
chemistry on populations of colubine monkeys in the old
world. biological journal of the linnean society, 34: l-32.
witkowski, e. t. f. & b. b. lamont, 1991.leaf specific mass
confoundsleafdensity and thickness.oecologia, 88: 486-493.
witkowski, e. t. f., b. b. lamont, c. s. walton & s. radford,
1992.leaf demography, sclerophylly and ecophysiology of
two banksias with contrasting leaf life spans.australian
journalof botany: 4o:849-862.
,:j
:i
,tj
,;.t
=
;
::
: i