American Journalof Botany 82(10): 1257-1263. 1995.
LEAF
OPTICAL
GRADIENT
PROPERTIES
ALONG A VERTICAL
IN A TROPICAL
CANOPY IN COSTA
RAIN FOREST
RICA1
LOURENS POORTER,2,5 STEVEN F. OBERBAUER,3 AND DAVID B. CLARK4
2Departmentof Forestry,WageningenAgriculturalUniversity,P.O. Box 342, 6700 AH Wageningen,
The Netherlandsand Departmentof Plant Ecology and EvolutionaryBiology,
P.O. Box 800.84, 3508 TB Utrecht,The Netherlands;
3Departmentof Biological Sciences, Florida InternationalUniversity,Miami, Florida 33199 and
Fairchild Tropical Garden, 11935 Old CutlerRoad, Miami Florida 33156; and
4David B. Clark, La Selva Biological Station,OrganizationforTropical Studies, Apartado 676,
2050 San Pedro de Montes de Oca, Costa Rica
Leaf optical properties(400-1,100 nm) were compared forfourspecies of rain foresttrees with crowns in understory,
mid-canopy,and canopy positions to test whetheroptical propertieschange with lightenvironment.The species tested
representa spectrumof regenerationpatternsrangingfromshade tolerantto lightdemanding.Overall, leafoptical properties
of the fourspecies were similar. Differencesin absorptancewere small, but statisticallysignificantamong the species and
positions along the canopy gradient.Species absorptancedifferences
correspondedsomewhatto shade tolerance;two of the
shade species showed higherabsorptance in lower lightenvironments,while the sun species showed the reversepattern.
Specificleaf mass (leaf weightper unit area) and chlorophyllcontentper unit leaf weightalso changed along the canopy
gradient.Specificleaf mass was positivelycorrelatedand chlorophyllper unit leaf weightwas negativelycorrelatedwith
of absorption,as representedby the absorptance per unit leaf
increasinglightenvironment.Consequently,the efficiency
weight,increased as lightlevel decreased, largelydue to changes in specificleaf mass. In contrast,efficiency
of absorption
per unit leaf chlorophyllwas relativelyconstantwithlightenvironmentforthe two species measured forchlorophyll.
Rain forestcanopy trees typicallyencounter a wide
rangeof microenvironments
duringtheirlifetimes.In the
understory,leaves of seedlings and saplings experience
low lightlevels, but highrelativehumidityand moderate
temperatures.In contrast,leaves in the canopy are exposed to high light levels, but also high temperatures,
wind speeds,and vapor pressuredeficits(Chiariello,1984;
Fetcher, Oberbauer, and Strain, 1985; Kira and Yoda,
1989). Of these factors,light presentsthe largestenvironmentalgradientand is thoughtto be an important,
and in some cases the most important,determinantof
establishment,growth,and survival(Fetcher,Oberbauer,
and Chazdon, 1994). In general only -1% of the photosynthetically
active radiation (PAR) reaches the forest
floor(Chazdon and Fetcher, 1984; Chazdon, 1988), and
that which does has an altered spectrumdue to the selective filtering
of the forestcanopy (Endler, 1993). Red
lightlevels are more stronglyreducedthan far-redlevels,
resultingin a red: far-redratio that may be 10% of that
in theopen (Lee, 1987; Endler,1993; Turnbulland Yates,
1993). In contrast,lightin the upper canopy is oftenwell
above thelightsaturationlevel of photosynthesisforcanopy species (Fetcheret al., 1987).
As lightis a limitingresource,treesmightbe expected
to maximize lightinterceptionin the most efficient
way,
I Manuscriptreceived17 November 1994; revisionaccepted28 March
1995.
The authorsthankthe staffof the OrganizationforTropical Studies
at La Selva Biological Station fortheirlogisticsupport,and Frans Bongers, Robin Chazdon, and David Lee for valuable comments on the
manuscript.This work was supported by a grant from Wageningen
AgriculturalUniversityand by National Science FoundationgrantBSR8918185.
5 Author forcorrespondence.
that is, at the lowest costs in termsof support,construction,and transpiration(Horn, 1971; Givnish, 1984). How
do canopy treescope withthis strongverticalgradientin
lightavailabilitywhile regulatingtemperaturesand transpirationallosses? Trees adjust theirmorphologicaland
physiological propertiesat differenthierarchicallevels,
e.g., at the whole tree level and at the leaf level. For
instance, trees with well-exposed crowns in the canopy
have a different
allometry(lower height/diameter
ratio;
Halle, Oldeman, and Tomlinson, 1978), crownarchitecture,and leafarrangement(Horn, 1971) comparedto trees
in the understory.Likewise, at the leaf level, exposed
leaves in thecanopy are thicker,more verticallyinclined,
and have a higherspecificleaf mass, lower chlorophyll
contentper unitweight,and a higherchlorophylla/bratio
compared to understoryleaves (Oberbauer and Strain,
1986; Givnish, 1987; Pearcy, 1987; Bongersand Popma,
1988; Fetcher,Oberbauer,and Chazdon, 1994). Most of
these leaf characteristicsshould affectleaf lightcapture,
but fewstudieshave directlyexamined lightabsorptance
in response to the canopy gradient.
In this paper we focus on leaf-levelchanges across the
canopy gradient.Here we test the hypothesisthat light
absorptance changes along the vertical gradient in the
forestfromunderstoryto canopy and examine whether
understoryleaves capture light in a more efficientway
than canopy leaves.
Given the lightincidenton a leaf,the absorptanceby
the leaf determinesthe amount of lightthat actuallybecomes available for photosynthesis.Generally, mature
greenleaves absorb 80-90% of the incidentPAR (Gates
et al., 1965; Ehleringer,1981; Lee and Graham, 1986).
As lightin the understoryis verylow, understoryleaves
should absorb more PAR thancanopy leaves, whichhave
to balance lightabsorption with requirementsfor tem-
1257
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
1258
AMERICAN JOURNAL OF BOTANY
peratureregulation(Givnish, 1984). Similarly,understory
leaves mightbe expectedto absorb more lightin the 700750 nm range than canopy leaves, as PAR is scarce in
the understory,and far-redlightis relativelymore available. Althoughquantum efficienciesare low in this spectralregion,thesewavelengthsmaycontributesignificantly
to overall carbon gain (Lee and Graham, 1986). Finally,
as carbon gain is low in the forestunderstory,understory
leaves are expectedto use theirresourcesmore efficiently
than canopy leaves, and capture more lightper unit dry
matteror per mole of chlorophyllinvested(Osborne and
Raven, 1986).
MATERIALS AND METHODS
This researchwas carriedout at theLa Selva Biological
Station,a researchfacilityoftheOrganizationforTropical
Studies, situated in the Atlantic lowlands of Costa Rica
(83059'W, 10026'N). The elevation rangesfrom30 to 100
m and the annual precipitationis 3,900 mm, with a
short dry period fromJanuaryto April (Sanford et al.,
1994). The foresthas been classified as a tropical wet
forest(Hartshorn,1983).
The species examined included two emergenttreespecies, Lecythisampla Miers (Lecythidaceae) and Dipteryx
panamensis (Pittier)Record & Mell. (Papilionaceae), and
two canopy species, Minquartia guianensis Aubl. (Olacaceae) and Simarouba amara Aubl. (Simaroubaceae).
All species are part of a long-termstudyon the demography and ecophysiologyof rain foresttrees (Clark and
Clark, 1987). They can be distinguishedinto two groups
withdistinctlifehistorypatterns(Clark and Clark, 1992).
Minquartia and Lecythis are shade-tolerantspecies of
which saplings can be found under low lightconditions
in mature-phaseforest(sensu Whitmore,1975). Dipteryx
and Simarouba on the otherhand, are shade tolerantand
thatstarttheir
high-light-demanding
species,respectively,
life cycle in the understory,but as juvenile tree size increases, they are increasinglyfound in more sunlitconditions in gap- and building-phaseforest.
wereconductedin
Measurementsof leafcharacteristics
Juneand Julyof 1991. Leaves used in measurementswere
collected fromtreeswithcrownsin eitherof threeheight
levels in the forestcanopy: the understory(1 2 m), the
mid-canopy(, 10 m), and canopy (>20 m). The understoryand mid-canopy trees were situated under closed
canopy, whereas the canopy treeswere exposed to direct
light at least during some part of the day. By using a
pruneror a rifle,healthy,fullyexpanded greenleaves were
collected just below the top of the tree in the case of
understoryand mid-canopytrees,and fromexposed sunlit
branches in the case of canopy trees.The youngestfully
matureleaves werecollectedfromtreesin theunderstory,
but because leaf production rates are low and leaf lifein some cases theseleaves
spans are longin theunderstory,
may have been older than leaves fromthe mid-canopy
and canopy. For the species withcompound leaves (Dipteryxand Simarouba), the largestleafletswere taken. In
total, 300 leaves were sampled: fourspecies, threeheight
levels in the canopy, fivetreesper heightlevel, and five
leaves per tree.
All individual treesfromwhich leaves were taken,except forfiveunderstorysaplingsof Dipteryx,werepartof
[Vol. 82
thelong-termstudyon treegrowthand censusedbymethods described in Clark and Clark (1992).
Collected leaves were stored under humid conditions
and analyzed for their spectralpropertieswithin24 hr.
For Dipteryx,Minquartia,and Lecythis,thepetioleswere
excised before measuringleaf area and not included in
the analysis. Leaf area was measured withan area meter
(LI-3 100, LI-COR Inc., Lincoln, NE). Of the fiveleaves
sampled per tree,one representativeleafwas analyzed for
its optical propertiesby means of a spectroradiometer
(LI- 1800,LI-COR Inc.) withan externalintegrating
sphere.
Only leaves withoutepiphylls were used in these measurementsand leaves were carefullycleaned with damp
tissue prior to measurement.Reflectanceand transmittance were determinedat 2-nm intervalsat a wavelength
range of 400-1,100 nm. Absorptancewas calculated accordingto the formula:
Absorptance = 1 - reflectance- transmittance. (1)
Total absorptance was calculated separately for the
wavelengthranges400-1,100 nm, 700-750 nm (farred),
and 400-700 nm (PAR). Absorptance values presented
here are absorptancesfromleaves exposed to an artificial
light source. To determine the actual absorptances by
leaves in the field,the measured absorptancesshould be
multipliedby the spectraldistributionof themean in situ
lightenvironment.However, the spectraldistributionof
lightin the forestis spatiallyand temporallyhighlyvariable, due to occurrence of sunflecks,the reflectanceof
lightbyclouds,theleafarea ofthesurroundingvegetation,
and the absorptancecharacteristicsofthe vegetation(Endler, 1993). For comparisons,we measured leaves under
equivalentfulllightconditionsas presentedbyan artificial
lightsource witha radiationspectrumreasonablysimilar
to that of sunlight.
For two of the species, Dipteryxand Lecythis,chlorophyllcontentwas measured forthe same leaves used for
absorptancemeasurements,i.e., one leafpertreesampled.
Chlorophyllwas extractedand measuredfollowingmethods of Porra, Thompson, and Kriedemann (1989). Afterwardsleaves were oven dried for 48 hr at 70 C and
weighed separately.Mean specificleaf mass per treewas
calculated on the basis of fiveleaves per tree.
Because leaf lightenvironmentcannot necessarilybe
inferredfromheightin thecanopy,we also compared leaf
spectralcharacteristicsto visual indexes of leaf lightenvironment.These measurementsinclude crown illumination index (based on Dawkins and Field, 1978) and
number of crowns above the tree.Tree growthhas been
shown to be stronglycorrelatedwith both of these measures (Clark and Clark, 1992). Crown illuminationindex
rangesfrom1 to 5, witha value of I indicatingno direct
light, 1.5, 2, 2.5 indicatinglow, medium or high lateral
light,3 indicatingsome overhead light,4 indicatingfull
overhead light but lateral light partiallyor completely
blocked,and a value of 5 indicatingan open canopy overhead withfulllateraldirectlight.For these comparisons,
lightindex measurementstaken in 1990, the year prior
to measurementof leaf optical properties,were used so
thatlightenvironmentpresentat thetimethe leaves were
developed was most closely approximated.
Statisticalanalysiswas carriedout withthe SPSS package (SPSS, 1990). A two-wayANOVA and a Tukey test
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
October 1995]
Absorptance
Transmittance
Reflectance
0.0
1.0
0.8
8
0.2
00cqC)0
C~~~~~~~~~~~~~
Y g
0.4
0.6-
S
0
t!
O S
a
t 4_ _ N O
* **
* **
*
^~~~~~f*
0.6
0.4
kn
ttn
0
>t
C
La
1259
POORTER ET AL. - LEAF OPTICAL PROPERTIES OF TROPICAL TREES
0.2-
0.0
400
0.8
500
600
700
800
900
.1
1000
0
1100
Wavelength
(nm)
Fig. 1. Reflectance
(dottedline),transmittance
(dashedline),and
leafof Dipteryx
absorptance(solid line) spectrafora representative
panamensis.
were used for discriminatingbetween species-, height-,
and interactioneffects.Data were logtransformed
ifthey
did not meet the assumptionof normalityor if variances
had to be stabilized. Absorptance values were arcsine
transformed
priorto analysisto meettheassumptionthat
data were symmetricallydistributedand not truncated.
Results were regardedsignificantifP values were <0.05.
RESULTS
Spectral properties-Leaves of the four tree species
showedessentiallythesame basic patternsin theirspectral
properties.Leaves were characterizedby low reflectance
and transmittancein the PAR range(e.g., Fig. 1), whereas
bothwerehighforthewavelengthrangefrom700 to 1,100
nm. The resultantwas a strongabsorptance in the PAR
range,withonlya small dip around 550 nm. Absorptance
declined sharplyabove 700 nm, and was very low for
higherwavelengths.
Absorptancesof leaves were very similar forthe four
species. Mean PAR absorptance ranged from 89.6 to
92.5%, and matchingcoefficientsof variation were also
very low, rangingfrom0.8% to 1.4% (Table 1). Despite
the small range in absorptances,a significantdifference
across species was found (P < 0.001).
There was also a significanteffectof canopy position
on leaf absorptance(P < 0.001); fortwo species, absorptance was slightlyhigherformid-canopyleaves than for
canopy leaves (cf. Fig. 2). This resultwas mostlydue to
an increased absorptance around 550 nm (Fig. 2). Although there was a heighteffecton the absorptance of
leaves in the 700-750 nm range,it was ratherequivocal
forthe different
species (Table 1).
Mean transmittancedifferedsignificantly
withposition
in the forestcanopy (P < 0.001), but reflectancedid not
(Table 1). For Lecythis,transmittancein the PAR range
was higherforleaves in the canopy than forunderstory
0
0)c~~~c
.c0C1
cc
e
C" o
\o, \O
2~~~~~~~~,
z~~~~~~~~1
Z
ZS
-R
?o
C) 0
YY cr C
ooco
H
.
4 0
r?qX
E.~~~~~~~~~~~0
0
c
F-4
?-t~~m
0~~~~~~~~~-r
00
C66oo6'a
b#~~~ cn
C13
'i- CC
0 Xt
0
o
=E
rz
1_4
C
CD
oo
o
-n
=
X
kn r
C
0 ES)
*
*
0
O-fccr
0
-
00
q000
40?
0
c-c
ON A71ONC.
0
1.c~~~~d
cl
~
Q~~cO000C
~
~~~00C(d~~~~
0)
c
0\co000
cl~~~~~~~~~~~~~~~~~~r
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
gcd
u
n
1260
[Vol. 82
AMERICAN JOURNAL OF BOTANY
. ..... Canopy
Midcanopy
Understory
Canopy-understory------
Midcanopy-understory-
Midcanopy-canopy
.
2. Mean specificleaf mass (SLM) of understory,mid-canopy,
and canopy leaves of four rain foresttree species. Symbols as in
Table 1.
TABLE
14w0~~~~~~~~~
Specificleaf mass (g/m2)
- 0.1
C
Q
} '.
0.750-
0.000
0~~~~~~~~~~~~~~~~~~~
500
600
700
800
900
1000
1100
Wavelength (nm)
Fig. 2. Mean absorptance (N =
from three differentheight positions
(solid line), the mid-canopy (dashed
The right y-axis indicates the mean
with differentcanopy position.
5) forDipteryxpanamensis leaves
in the forest canopy; the understory
line), and the canopy (dotted line).
difference spectrum among leaves
leaves. For Simarouba, which has very heavy leaves in
the canopy, the reversewas found.
Canopy
trends in SLM
and chlorophyll-The
specific
leaf mass increasedstronglywithheightin the forestcanopy for all species (P
<
0.001). Simarouba and Dipteryx
showed the largestincrease withcanopy height(Table 2)
of variationthan
as indicated by theirlargercoefficients
those of the more shade-tolerantspecies,Minquartia and
Lecythis (61% and 41% vs. 24% and 27%, respectively).
The two species analyzed forchlorophyllcontent,Dipin thechloteryxand Lecythis,did not differsignificantly
rophyllper unit leaf area or leaf dry weight(Table 3).
However, chlorophyllcharacteristicsresponded moderatelyto a position gradientin the canopy; chlorophylla!
b ratios increased with heightin the canopy (P < 0.05),
whereaschlorophyllcontentper unitweightdecreased (P
< 0.001). Total chlorophyllcontentper unitarea was not
forleaves along thecanopy gradient
different
significantly
(Table 3).
Correlations
with light environment-Within
Sa
Mg
La
All
30.Oa
48.6b
68.6c
49.1
19.0
41
36.9a
55.4b
135.2c
75.8
45.9
61
52. la
64.9a
73.8a
63.6
15.3
24
36.6a
36.9a
56.Ob
43.2
11.7
27
38.8
51.4
83.4
57.9
29.0
50
Height
Species
Interaction
-0.1
400
Understory
Mid-canopy
Canopy
All
SD
cv(%)
Dp
a species,
SLM correlatedwell with estimatesof the lightenvironment. For all species except Minquartia, SLM was positively correlated with crown illumination index (P <
0.05), and negativelycorrelatedwiththenumberofcrowns
above the tree (P < 0.05; Table 4). In the case of Minquartia, the range of crown illuminationenvironments
was smallerthan those of the otherspecies, so significant
correlationswere less likelyto be found.Also, estimates
of the light environmentwere unavailable for the five
whichprobsmallestDipteryxsaplingsin the understory,
ably experienced the lowest light conditions. Had they
been included,correlationsforDipteryxindividuals with
lightenvironmentlikelywould have been even stronger.
Correlationswith light environmentwere weaker for
PAR absorptance than for SLM. Only Dipteryx leaves
showed a significantnegative correlation between absorptance and crown illuminationindex, that is, leaves
in closed-canopyenvironmentsabsorbed more PAR than
leaves fromsunnyenvironments(Table 4). Lecythisleaves
showed a similar tendency,but the correlationwas only
close to significance(r5= -0.50, P = 0.056). Correlations
werestrongerbetweenabsorptanceand numberofcrowns
above (Table 4, Fig. 3). Absorptance of Dipteryxand
Lecythiswerepositivelycorrelatedwithnumberofcrowns
above, whereas Simarouba showed the opposite trend.
ofabsorptance-Efficiencyoflightcaptureby
Efficiency
leaves can be expressed as PAR absorptance per unit
biomass or per unit chlorophyllinvested. All species
showed similarresponsesin termsof efficiencies
of PAR
absorptance per unit dry weight; efficiencieswere 20100% higherforunderstoryleaves than forcanopy leaves
(P < 0.001, Fig, 4). Efficienciesof absorptance tended
also to be species-specific(P < 0.001). This resultwas
largelydue to interspecific
variationin SLM; species with
the lowest SLM also had the highestefficiency.
When efficiency
was expressedas lightcaptureper unit
chlorophyll,however, another picture emerged. In that
were similarforleaves with
case, absorptanceefficiencies
different
heightsin the canopy (P > 0.05; Table 5).
DISCUSSION
Spectral properties-Mean PAR absorptances of the
four tropical tree species were as high as 91%, which
Mean chlorophyllcontentper unit leaf weightand leaf area,
and chlorophylla to chlorophyllb ratiosofunderstory,
mid-canopy,
and canopy leaves of two rain foresttree species. Symbols as in
Table 1.
TABLE 3.
Chlorophyll
(mg/g)
Understory
Mid-canopy
Canopy
All
SD
cv(%)
Height
Species
Interaction
Chlorophyll
(mg/m)
Chlorophylla/b
Dp
La
Dp
La
Dp
La
15.5a
12.4ab
9.2b
12.3
3.4
28
16.la
14.4ab
8.2b
12.9
4.3
33
458a
604a
595a
552
121
22
580a
527a
446a
517
92
18
2.37a
2.40a
2.70a
2.49
0.21
8
2.13a
2.24a
2.26a
2.21
0.20
9
ns
ns
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
ns
ns
*
*
ns
October 1995]
POORTER ET AL. -LEAF
1261
OPTICAL PROPERTIES OF TROPICAL TREES
and lightenvironmentforindividualsof fourrain foresttreespecies.
betweenleafcharacteristics
4. Spearman's rankcorrelationcoefficients
Parametersincluded in the analysisare specificleafmass (SLM), lightabsorptanceofleaves in the 400-700 nm range(ABS), crownillumination
index (CII), and the numberof crownsabove an individual (CA). The numberof replicatesis given in parentheses.
TABLE
ABS x CII
ns
-0.86** (10)
ABS x CA
SLM x CII
SLM x CA
closelyresemblesabsorptancevalues reportedfora range
ofotherrain forestspecies (Lee and Graham, 1986). Generally,PAR absorptancesforgreenleaves average - 85%,
but they may be stronglyreduced due to modifications
of the leaf surface,such as waxes, hairs,and salt bladders
thatincreasereflectance(Ehleringerand Werk,1984; Vogelmann, 1993). Leaf absorptancesof plant communities
tend to increase withhumidityof the habitat(Ehleringer
and Werk, 1984), which is consistentwith the high absorptancevalues found forthe species studied here.
variationwas foundin
Furthermore,littleintraspecific
theabsorptanceof leaves ofthestudyspecies withcanopy
position despite large differencesin leaf mass and presumably leaf thickness(Fetcher,Oberbauer, and Chazdon, 1994). For two of the species, PAR absorptances
wereslightlyhigherformid-canopyleaves thanforcanopy
in absorptanceweresmall,
leaves, but overall,differences
of variationwerelow. The
and correspondingcoefficients
same was found forabsorptance in the 700-750 nm region. Apparentlyit is advantageousfora leafto maximize
potentiallightabsorptance,whetherin the understoryor
canopy. Similarly,Lee and Graham (1986) did not find
in mean PAR absorptancesoflight-demanding
differences
ns
ns(14)
(13)
(15)
0.69** (15)
0.81***(15)
-0.79***(15)
ns(l4)
ns(l4)
ns (14)
-0.60* (13)
0.85***(14)
-0.88***(14)
0.86**(10)
0.67* (10)
-0.71* (10)
Lecythis
Minquartia
Simarouba
Dipteryx
species compared with shade-tolerantspecies fromrain
forest.
Both transmittanceand reflectancealso differedlittle
canopy positions. It has been sugforleaves of different
such as theforest
gestedthatin sunnymicroenvironments
canopy, increased reflectanceis beneficialforthe leaf,as
it reducesthe heat load and thustranspirationaland photosyntheticcosts (whenleaftemperaturesexceed thethermal optimum for photosynthesis;Givnish, 1984). This
argumentdoes not hold, however, if transmittancedecreases at the same time,resultingin an absorptancethat
is similarforunderstoryand canopy leaves, as was found
in the present study. Hollinger (1989) also found that
reflectanceby the adaxial leaf surfaceformontaneNothofagus stayedrelativelythe same withheightin the canthereflectanceofthepubescentabaxial
opy. Interestingly,
leaf surfaceincreased with heightin the canopy, which
he suggestedmightincrease the photosyntheticphoton
fluxdensityin the lower canopy by backscattering.
Canopy trends in SLM and chlorophyll-SLM increased with heightin the canopy forall fourspecies, a
findingconsistentwithresultsofpriorstudies(Jurik,1986;
E
CL05
0.85-
F
,
,
,--
,
,
, -
0.9
.
.
,
,
Lecythis |
o
*
,
0.8
0.0
.
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
,
0.0
0.5
U.
1.0
W
1.5
3
l
Minquartia |
U
--
2.0
2.5
3.0
3.5
4.0
Crownsabove
Fig. 3.
leaves.
Relationship between PAR absorptanceand numberof crownsabove the treeforDipteryx,
Simarouba,Minquartia,
and Lecythis
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
1262
[Vol. 82
AMERICAN JOURNAL OF BOTANY
*
Understory
U
Mid-canopy E
Canopy
5. Mean absorptance efficienciesper unit chlorophyllof understory,mid-canopy,and canopy leaves of two rain foresttree
species. Symbols as in Table 1.
TABLE
Absorptance400-700/chl (%/mg)
a
Dipteryx
0
Understory
Mid-canopy
Canopy
All
SD
cv (%)
E 3- 1L
0
a.~~~~~~~~~~~~
EL
0
U)
Height
Species
Interaction
b
0
Dipteryx
Simarouba Minquartia
Lecythis
(% PAR absorbed per unitbiomass)
Fig. 4. Absorptanceefficiencies
forDipteryx,Simarouba, Minquartia, and Lecythisleaves fromthree
different
positions in the canopy. Errorbars indicate standarderrorsof
the mean. Withineach species, means thatshare the same letterare not
different
at the 5% level. Data were arcsine transformed
significantly
prior to statisticalanalysis.
Oberbauer and Strain, 1986; Hollinger, 1989; Ellsworth
and Reich, 1993; Fetcher,Oberbauer,and Chazdon, 1994).
Increases in SLM were particularlystrongforthe lightdemandingspecies.A higherSLM oftenis associated with:
(1) thickerleaves, (2) greatermesophylldevelopment,and
(3) higher nitrogencontent per unit leaf area (Chabot,
Jurik,and Chabot, 1979; Mooney, 1985; Hollinger,1989;
Bongersand Popma, 1990). Greaterleaf thicknessmight
be a xeromorphicresponseto an environmentwitha low
relativehumidityand a highevaporativedemand. Greater
mesophylldevelopment,hence more photosyntheticapparatus and the nitrogenassociated with it, should lead
to a greaterper unit area photosyntheticrate in a lightsaturatedenvironment(Jurik,1986; Ellsworthand Reich,
1993). For understoryenvironmentson the otherhand,
leaves with a low SLM are advantageous, as theyhave a
higherproductivityper unit biomass invested (Bongers
and Popma, 1988).
Compared to SLM, the chlorophyllcontent per unit
biomass showed an opposite trendin that it was higher
for shaded understoryleaves than for exposed canopy
leaves. Opposite patternsin SLM and chlorophyllcontent
per unitbiomass lead ultimatelyto a chlorophyllcontent
per unit area that is similar for understoryand canopy
leaves.
Chlorophylla/b ratios tended to be lowest for leaves
as in SLM, the
fromthe forestunderstory.Interestingly,
morelightdemandingspecies ofthetwo species measured
forchlorophyll,Dipteryx,showed the greatestchangesin
chlorophylla/b ratios. Furthermore,chlorophylla/b ratios ofunderstoryleaves ofDipteryxweresimilarto those
of canopy leaves of the shade species, Lecythis. These
resultsare consistentwith the idea that chlorophylla/b
ratio changesare a responseto the far-red-enriched
spectral distributionof lightin the shade (Bjorkman, 1981).
Simarouba
0.196a
0.156a
0.156a
0.169
0.034
20
0.161a
0.173a
0.200a
0.178
0.030
17
ns
ns
*
Relations withlightenvironment-Comparingleaf parametersfor trees with different
heightpositions in the
canopy does not allow us to determineif the measured
parametersresponded to changes in tree ontogenyor to
an increase in light availability, because trees typically
encountermore sunlitconditionsas theygetolder. However, correlationof leaf parameterswithestimatesof the
lightenvironmentat least show that indeed, changes in
leaf parametersco-occur with changes in the lightenvironment.Overall, correlationsbetween leaf parameters
and lightenvironmentwere strongerwhen lightenvironmentwas assessed by numbersof crownsabove the tree,
instead of crown illumination index, possibly because
canopy illumination index and the number of crowns
above a treeweighdirectand diffuselightdifferently.
More
crowns above a tree representa highercumulative leaf
area index, whereas a highercanopy illuminationindex
representsa greateramount ofopen sky.These weightings
may be importantbecause increased leaf thicknessand
palisade layeringare thoughtto increase the path length
of directlight,but not diffuselight,throughthe leaf (Vogelmann, 1993).
Overall, although absorptance varied only slightlyin
responseto lightenvironment,theresultsforsome of the
species indicate thatleaves underlow lightconditionsdo
indeed absorb more lightthan leaves under sunlit conditions, giving some support for Givnish's (1984) hypothesisthatshade-adaptedunderstorytreesabsorb more
PAR than canopy trees.
Efficiencyof absorptance- Despite similar absorptances forcanopy and understoryleaves, the latterare far
moreefficient
in theirlightcaptureperunitbiomass. Such
a highefficiency
may give an advantage in an understory
environment,where carbon fixationrates are low and
where there is a premium on efficientuse of biomass.
Similarly,Lee and Graham (1986) found shade-tolerant
species to be more efficientthan sun species. How is it
that understoryleaves capturea similar amount of light
as canopy leaves, witha smallerinvestmentin biomass?
Lee and Graham (1986) concludedthatinherentspecies
differencesin anatomical featureswere the basis forthe
in absorptanceefficiency.
differences
Shade-tolerantspecies have a thinpalisade layerwithmoreequal dimensions
ofthecells,creatinga narrow,continuousand dense layer
of chloroplasts,that maximizes light capture efficiency
(Lee et al., 1990). In contrast,the more columnarcells of
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions
October 1995]
POORTER ET AL.-LEAF
OPTICAL PROPERTIES OF TROPICAL TREES
sun species allow lightto pass throughthecentralvacuoles
and reachchloroplastsin lowercell layers(Lee etal., 1990;
Vogelmann, 1993). Hence, the number of chloroplasts
exposed to saturatinglightlevels, and thereforephotosyntheticcapacity, is increased, but at the expense of a
reduced efficiency
of lightcapture.
anotherplausible
In addition to anatomical differences,
explanation is the higher chlorophyllcontent per unit
biomass of understoryleaves relative to canopy leaves.
Higherchlorophyllperunitbiomass, in combinationwith
a lower SLM, resultsin similar chlorophyllcontentsper
unit area, hence similar light-harvesting
capabilities,for
leaves fromdifferent
canopy positions.
LITERATURE
CITED
quantum fluxdensities.
0. 1981. Responses to different
BJORKMAN,
In 0. L. Lange, P. S. Nobel, C. B. Osmond, and H. Ziegler [eds.],
Physiological plant ecology. I. Encyclopedia of plant physiology,
Heidelberg.
57-107. Springer-Verlag,
BONGERS,F., AND J. POPMA. 1988. Is exposure-relatedvariation in
leaf characteristicsof tropical rain forestspecies adaptive? In M.
J. A. Werger,P. J. M. van der Aart, H. J. During, and J. T. A.
Verhoeven [eds.], Plant formand vegetationstructure,191-200.
SPB Academic Press, The Hague.
. 1990. Leaf characteristicsof the tropical rain
, AND
forestfloraof Los Tuxtlas, Mexico. Botanical Gazette 151: 354365.
CHABOT,B. F., T. W. JURIK,AND J. F. CHABOT. 1979. Influenceof
instantaneousand integratedlightfluxdensityon leafanatomyand
photosynthesis.AmericanJournalofBotany 66: 940-945.
CHAZDON, R. L. 1988. Sunflecksand theirimportanceto forestunderstoryplants. Advances in Ecological Research 18: 1-63.
1984. Photosyntheticlightenvironmentsin
, AND N. FETCHER.
a lowland tropicalrain forestin Costa Rica. JournalofEcology 72:
553-564.
1984. Leaf energybalance in the wet lowland tropics.
In E. Medina, H. A. Mooney, and C. Vazquez Yanes [eds.], Physiological ecology of plants in the wet tropics.Tasks forVegetation
Science 12, 55-89. Junk,The Hague.
CLARK,D. A., AND D. B. CLARK. 1987. Analisis de la regeneraci6nde
arboles del dosel en bosque muy humedo tropical: aspectos te6reticosy practicos.Revista Biologia Tropical 35(Suppl.): 41-54.
. 1992. Life historydiversityof canopy and emer, AN1'qDgenttreesin a Neotropical rain forest.Ecological Monographs 62:
315-344.
DAWKINS, H. C., AND D. R. B. FIELD. 1978. A long-termsurveillance
systemfor Britishwoodland vegetation.Departmentof Forestry,
Oxford University,Oxford.
J. R. 1981. Leaf absorptancesof Mohave and Sonoran
EHLERINGER,
desertplants. Oecologia 49: 366-370.
, AND K. S. WERK. 1984. Modificationsof solar-radiationabsorptionpatternsand implicationsforcarbon gain at the leaf level.
In T. J. Givnish [ed.], On the economy of plant formand function,
57-82. Cambridge UniversityPress, Cambridge.
1993. Canopy structureand
ELLSWORTH,D. S., AND P. B. REICH.
verticalpatternsof photosynthesisand related leaf traitsin a deciduous forest.Oecologia 96: 169-178.
ENDLER, J. A. 1993. The color of lightin forestsand its implications.
Ecological Monographs 63: 1-27.
1994. PhysioFETCHER, N., S. F. OBERBAUER, AND R. L. CHAZDON.
logical ecology of plants. In L. McDade, K. S. Bawa, H. Hespenheide, and G. S. Hartshorn [eds.], La Selva: ecology and natural
historyof a neotropicalrain forest,128-141. Universityof Chicago
Press, Chicago, IL.
AND B. R. STRAIN. 1987. Efectosdel regimen
,G. ROJAS,
de luz sobre la fotosintesisy el crecimientoen plantulasde arboles
CHLARIELLO,N.
1263
de un bosque Iluvioso tropical de Costa Rica. Revista Biologia
Tropical 35(Suppl.): 97- 110.
, A-ND B. R. STRAIN. 1985. Vegetation effectson microclimate in lowland tropical forestin Costa Rica. International
Journalof Biometeorology29: 145-155.
GATES, D. M., H. J. KEEGAN, J. C. SCHLETER, AND V. R. WEIDNER.
1965. Spectral propertiesof plants.Applied Optics 4: 11-20.
GIVNISH,T. J. 1984. Leaf and canopy adaptations in tropicalforests.
In E. Medina, H. A. Mooney, and C. Vazquez Yanes [eds.], Physiological ecology of plants in the wet tropics.Tasks forVegetation
Science 12, 51-84. Junk,The Hague.
. 1987. Comparative studiesof leaf form:assessingthe relative
roles of selectivepressuresand phylogeneticconstraints.New Phytologist106: 131-160.
HALLE,F., R. A. A. OLDEMAN, AND P. B. TOMLINSON. 1978. Tropical
treesand forests.Springer-Verlag,
Berlin.
HARTSHORN, G. S. 1983. Plants. In D. H. Janzen [ed.], Costa Rican
naturalhistory, 18-157. Universityof Chicago Press,Chicago, IL.
HOLLINGER, D. Y. 1989. Canopy organizationand foliagephotosyntheticcapacity in a broad-leaved evergreenmontane forest.Functional Ecology 3: 53-62.
HORN, H. S. 1971. The adaptive geometryof tree form. Princeton
UniversityPress, Princeton,NJ.
JURIK,T. W. 1986. Temporal and spatialpatternsofspecificleafweight
in successional northernhardwood treespecies. AmericanJournal
of Botany 73: 1083-1092.
in microclimate.
KiRA, T., AND K. YODA.
1989. Verticalstratification
In H. Lieth and M. J. A. Werger[eds.], Tropical rain forestecosystems:biogeographicaland ecological studies, 57-72. Elsevier,
Amsterdam.
LEE, D. W. 1987. The spectraldistributionof radiation in two neotropicalforests.Biotropica 19: 161-166.
, R. A. BONE,S. L. TARSIS,AND D. STORCH. 1990. Correlates
of leaf optical propertiesin tropicalforestsun and extremeshade
plants.AmericanJournalof Botany 77: 370-380.
, AND R. GRAHAM. 1986. Leaf optical propertiesof rainforest
sun and extreme shade plants. American Journal of Botany 73:
1100-1108.
MoONEY, H. A. 1985. Photosynthesis.In M. J. Crawley [ed.], Plant
Ecology, 345-374. Blackwell Scientific,Oxford.
1986. Effectsof canopy position
OBERBAUER, S. F., AND B. R. STRAIN.
and irradiance on the leaf physiologyand morphologyof Pentaclethramacroloba (Mimosaceae). AmericanJournalofBotany 73:
409-416.
OSBORNE, B. A., AND J. A. RAVEN. 1986. Light absorptionby plants
and its implicationsforphotosynthesis.Biological Reviews61: 161.
PEARCY, R. W. 1987. Photosynthetic
gas exchange responsesof Australian tropical foresttrees in canopy, gap and understorymicroenvironments.Functional Ecology 1: 169-178.
PoRRA, R. J., W. A. THOMPSON, AND P. E. KRIEDEMANN.
1989. Determinationsof accurate extinctioncoefficientsand simultaneous
equations for assaying chlorophyllsa and b extractedwith four
different
solvents: verificationof the concentrationof chlorophyll
standards by atomic absorption spectroscopy.Biochimica BiophysicaActa 975: 384-394.
SANFoRD, R. L., P. PAABY, J. C. LUVALL, AND E. PHILLIPS.
1994. Climate, geomorphology,and aquatic systems.In L. McDade, K. S.
Bawa, H. Hespenheide,and G. S. Hartshom[eds.],La Selva: ecology
and naturalhistoryof a neotropicalrain forest,19-33. University
of Chicago Press, Chicago, IL.
SPSS. 1990. SPSS/PC+ V4.0 Base manual. SPSS INC, Chicago, IL.
TURNBULL, M., AND D. J. YATES.
1993. Seasonal variationin the red/
far-redratio and photon fluxdensityin an Australiansub-tropical
rainforest.Agriculturaland Forest Meteorology64: 111-127.
VOGELMANN, T. C. 1993. Plant tissue optics. Annual Review of Plant
Physiologyand Molecular Biology 44: 231-252.
WHITMORE, T. C. 1975. Tropical rain forestsof the fareast. Clarendon
Press, Oxford.
This content downloaded from 134.124.28.17 on Wed, 18 Dec 2013 19:37:49 PM
All use subject to JSTOR Terms and Conditions