j. Plant Physiol. Vol. 148. pp.

483-493 (1996)

Non-Destructive Determination of Chlorophyll Content of

Leaves of a Green and an Aurea Mutant of Tobacco by
Reflectance Measurements

1 2 1
HARTMUT K. LICHTENTHALER , ANATOLY GITELSON , and MICHAEL LANG
1 Botanisches Institut, Lehrstuhl II, University of Karlsruhe, Kaiserstr. 12, D-76128 Karlsruhe, Germany
2 The Remote Sensing Laboratory, J. Blaustein Institute for Desert Research, Ben Gurion University of the Negev, Sede
Boker, Campus 84933, Israel

Received July 12, 1995 . Accepted November 30, 1995

Summary

Reflectance spectra from 400 to 800 nm with a spectral resolution of 1nm and the content of photosyn-
thetic pigments were acquired for leaves of an aurea mutant and a green tobacco (Nicotiana tabacum L.)
covering a ran~e of chlorophyll a content from 7 to 30.81lg cm -2 and total chlorophyll a+ b level from 8.4
to 41.4Ilgcm- leaf area. At a chlorophyll a content of more than 7llgcm-2 leaf area the reflectance near
670 to 680 nm was not sensitive to a variation in chlorophyll content due to saturation of the relationship
«absorption vs. chlorophyll content». The wavelength position of the red reflectance minimum (Rmin) in
the main red absorption bands of in vivo chlorophyll a proved to be independent of the chlorophyll con-
tent of leaves and thus is not suitable for chlorophyll determination. Maximum sensitivity of leaf reflec-
tance to variation in chlorophyll content was, however, found in a wide spectral range from 530 to
630 nm and near 700 nm. The wavelength of rhe red edge position of the reflectance spectrum (inflection
point IP) in the range of 694 to 706 nm correlated very closely with the leaves' chlorophyll content in a
curvilinear manner. The Ip is also closely correlated in a linear manner with the reflectance signals at 550
and 700 nm. Reflectances near 700 nm and 550 nm were found to be sensitive indicators of the red edge
position and the chlorophyll content of leaves as well. The ratios of reflectances in the near infra-red range
of the spectrum (above 750 nm) to that at 700 nm and at 550 nm, RN1R/R700 (or R750/R700) and RN1RI
R550 (or R750/R550), were directly proportional (r2>0.93) to the leaves' chlorophyll content. These two
novel indices for chlorophyll determination (R750/R700 and R750/R550) allowed the estimation of leaf
chlorophyll content with an error of less than 2.1Ilgcm-2. The two new vegetation indices, R750/R700
and R750/R550, are linearily correlated to the chlorophyll content. They exhibit a more than 6 times wi-
der dynamic range than the widely used Normalized Difference Vegetation Index NOV!. The new vegeta-
tion index ratios are very sensitive to changes in chlorophyll content at low, medium and high chlorophyll
content which is not the case for the NOV!. This suggests the application of the new indices R750/R700
and R750/R550 in the near and remote sensing of the chlorophyll content of terrestrial vegetation.
Key words: Chlorophyll content, inflection point IF, red edge, reflectance spectra, red reflectance minimum
(Rmin), remote sensing ofchlorophyll content, new vegetation index.

Abbreviations: Chi = chlorophyll; Chi a and b = chlorophyll a and b; c = carotenes; IP = position of the
inflection point at the red edge of the reflectance spectrum; LHCPs = Light-harvesting chlorophyll-pro-
teins of the photosynthetic photosystem II; NOV! = normalized difference vegetation index; NIR = near
infra-red range of the spectrum; red edge = rise in the reflectance spectrum of leaves between 680 to
750 nm; Rmin = red reflectance minimum near 680 nm caused by the main red absorption bands of in
vivo chlorophyll a; R550 and R700 = reflectance signals at 550 and 700 nm, respectively; x+ c = total leaf
carotenoids; x = xanthophylls.

© 1996 by Guslav Fischel Vetlag, SlUllgall

484 HARTMUT K. LICHTENTHALER, ANATOLY GITELSON, and MICHAEL LANG

Introduction length positions of the IP of the red edge has, however, not
yet been determined. Chlorophyll a is bound in vivo in the
Terrestrial vegetation is exposed to various kinds of natural photosynthetically active thylakoids of chloroplasts to several
and anthropogenic stressors (Lichtenthaler, 1996), many of chlorophyll-carotenoids-proteins which possess differential
which act simultaneously and will reduce the chlorophyll Chi a absorption bands (Lichtenthaler et al., 1981). The dif-
content of plants by a photooxidative chlorophyll breakdown ferent photosynthetic pigment proteins can be separated by
at a long-term stress exposure. Remote determination of the polyacryl gel electrophoresis and yield the photosystem I pig-
chlorophyll content from near and far distance by non-de- ment-proteins CPla and CPI, the photosystem II pigment-
structive methods is therefore a good mean to detect stress protein CPa as well as the light-harvesting chlorophyll-pro-
conditions in plants (Lichtenthaler, 1989). By non-invasive teins LHCPs, which belong to the light harvesting complex
techniques, such as reflectance measurements, one can rou- LHCII of photosystem II. In the photosystem I-proteins,
tinely follow at the same plant or canopy the successive de- CPla and CPI, the Chi a absorption maximum lies at
cline of chlorophylls at continuous stress exposure, and also 678 nm, whereas in the photosytem II protein CPa and the
the regeneration of plants and their photosynthetic pigment Chi a/Chl b-proteins of the pigment antenna (the LHCPs)
apparatus when the stressors are removed. the absorption band of the majority of Chi a molecules is
The increase in leaf reflectance in the visible range of the found at 672 nm (Lichtenthaler et al., 1981). In the reaction
spectrum at a decrease of the chlorophyll content is very centers of both photosystems there exist, however, minor
much used in ecophysiology and stress research of plants and Chi a forms which absorb at longer wavelengths such as the
has been applied by many authors (Buschmann and Lichten- Chi a dimers P680 (peak at 680 nm) and P700 (peak at
thaler, 1988; Buschmann et al., 1991; Chappelle et al., 1992; 700 nm) (e.g. Butler and Hopkins, 1970; French et al., 1972;
Carter et al., 1993 and 1994; Horler et al., 1983; Lichtentha- Lawlor, 1995). One can assume that the relative proportions of
ler, 1989; Rock et al., 1986; Schmuck et al., 1987; Thomas these pigment-proteins in leaves should determine the wave-
and Gaussmann, 1987; Yoder and Waring, 1994). For this length position of the red reflectance minimum of leaves.
purpose different reflectance signals at particular wavelengths The reflectance spectra of the tobacco leaves, taken in the
in the visible and near infra-red (NIR) range of the spectrum visible and near infra-red range of the spectrum, were ana-
were determined, and special ratios of the reflectance signatu- lysed in order to find spectral bands with a maximum sensi-
res are formed which function as indicators of leaf area index, tivity to a variation in chlorophyll content and to possibly de-
canopy cover and chlorophyll content. A much applied index vice new vegetation in,dices for a better non-invasive remote
is the Normalized Difference Vegetation Index NDVI (Rouse chlorophyll determination. The research focused on the fol-
et al., 1974; Guyot, 1990; Baret et al., 1992), which is not lowing spectral features of tobacco leaves which appear to be
only sensed in near distance but also from airborne systems useful for a remote chlorophyll determination:
including satellites. The NDVI, bound to the reflectance sig- (1) high sensitivity of reflectance in the range near 550 nm to
nals at 680 nm (R680, Rred) in the red range of the spectrum variation in Chi a content (Thomas and Gaussman, 1977;
near the red absorption bands of chlorophyll a and in the Tanner and Eller, 1986; Buschmann and Nagel, 1993; Yoder
NIR range above 750 to 800 nm (R800), is defined as the ra- and Waring (1994); Gitelson and Merzlyak, 1994 a and b;
tio of (RNIR-Rred)/(RNIR+Rred) usually expressed as (R800- 1996; Gitelson et al., 1996 a, b) and to different kinds of
R680)/(R800+R680). The vegetation index NDVI is, how- plant stress (Carter et al., 1993 and 1994; Schmuck et al.,
ever, not suitable for chlorophyll determination. It is precise 1987);
only at fairly low chlorophyll levels, but is not sensitive to (2) high sensitivity of reflectance in quite narrow spectral
variations in chlorophyll content at medium and higher bands from 690 to 710 nm to variation in Chi a content (Gi-
chlorophyll levels of the vegetation. For this reason several at- telson and Merzlyak 1994 a, band 1996; Gitelson et aI.,
tempts have been made in recent years in order to increase 1996 a, b);
the accuracy of a reflectance-based chlorophyll determination (3) very close correlation between reflectances at 550 nm and
by inclusion of other or additional reflectance bands and ra- 700 nm (Chappelle et al., 1992; Gitelson and Merzlyak
tios (Buschmann and Lichtenthaler, 1988; Buschmann and 1994a and b; 1996);
Nagel, 1993; Buschmann et al., 1991; Gitelson and Merzlyak (4) very close correlation between reflectances near 500 nm
1994a, 1994b and 1996; Yoder and Waring, 1994). and 670 nm (Gitelson and Merzlyak 1996; Gitelson et aI.,
The shift of the inflection point IP at the red edge rise of 1996 a); and
the reflectance spectrum towards shorter wavelengths (blue (5) a close relation between the red edge position and reflec-
shift of IP) seems to be a good indicator of a decline in the tance near 700 nm (Gitelson et al., 1996 b).
chlorophyll content (Horler et al., 1983; Lichtenthaler and In addition, the wavelength position dependence of the
Buschmann, 1987; Buschmann and Lichtenthaler, 1988; Cur- inflection point IP of the red edge as well as that of the red
ran et al., 1991; Baret et al., 1992; Gitelson et aI., 1996 b) and reflectance dip (Rmin) in the chlorophyll a absorption bands
has also been used in the remote sensing of forest decline around 680 nm on the chlorophyll content was determined.
(Rock et al., 1986; Schmuck et al., 1987; Lichtenthaler, The aim of the investigation presented here was to develop a
1994). The red reflectance dip Rmin in the main absorption better non-destructive technique for the determination of to-
bands of in vivo chlorophyll a broadens with increasing tal chlorophyll content by re-investigating different reflectance
chlorophyll content. Whether the wavelength position of the signatures using a data set of reflectance spectra from leaves
reflectance minimum Rmin depends upon the chlorophyll with different chlorophyll content of an aurea mutant and a
content and whether it exhibits any relation to the wave- green variety of Nicotiana tabacum L.
 Chlorophyll determination via leaf reflectance 485

Material and Methods Table 1: Aurea tobacco Sulsu: Level of photosynthetic pigments, pig-
ment ratios, refleCtance ratios and position of the infleCtion point
Plants (IP) of the red edge and red refleCtance minimum (Rmin) in yellow-
ish-green to light-green leaves. The leaves are arranged by increasing
Eight to ten week old tobacco plants (Nicotiana tabacum L.), chlorophyll content. Pigment values are based on 3 determinations
green form su/su and aurea mutant Su/su (Schmidt, 1971; Lichten- (SO < 5 %) and are given in llg . cm -2 leaf area.
thaler et al., 1975; Santrucek et al., 1992), were cultivated on a
mineral-containing peat in the greenhouse of the Botanical Garden Leaf number 7 6 4 3 2
of the University of Karlsruhe under standard conditions (cf. Lang
Pigment content
et al., 1996).
Chlorophyll a 7.00 10.80 12.00 12.80 13.50 14.90 15.50
Chlorophyll b 1.37 2.27 2.79 3.24 3.51 4.52 4.94
Total chlorophylls a+ b 8.37 13.07 14.79 16.04 17.01 19.42 20.44
Determination ofpigment content Total carotenoids x + c 3.21 3.95 3.91 3.46 3.34 3.54 3.67
Pigment ratios
Photosynthetic pigments, chlorophylls and carotenoids, were de- alb 5.11 4.75 4.30 3.95 3.85 3.30 3.14
termined spectrophotometrically (UV-2001 PC, Shimadzu, Ouis- (a+b)/(x+c) 2.61 3.31 3.78 4.63 5.10 5.49 5.57
burg, Germany) from extracts in 100 % acetone using the redeter- Reflectance ratios
mined extinction coefficients and equations of Lichtenthaler (1987) R750/R550 1.39 1.74 1.82 1.83 1.85 1.96 2.10
which allow the simultaneous determination of Chi a and b and to- R750/R700 1.54 1.82 2.16 2.25 2.27 2.52 2.54
tal carotenoids (x+ c) in the same extract solution. Red edge
position ofIP (nm) 694.0 697.0 698.0 698.0 699.0 700.2 700.0
position of Rmin (nm) 678.4 678.4 678.4 678.7 677.5 678.7 678.6
Reflectance measurements
Reflectance speCtra from 350 to 900 nm were recorded from the
upper leaf side with a spectrophotometer (UV-2001PC, Shimadzu,
Ouisburg, Germany) applying an integrating sphere and BaS04 as
reference. The wavelength scanning speed was set to 100 nm/min. Table 2: Green tobacco sulsu: Level of photosynthetic pigments, pig-
For the determination of the refleCtance signals at 490, 550, 670, ment ratios, refleCtance ratios and position of the infleCtion point
680, 700 and 750 nm, which were correlated to each other or used (IP) of the red edge and red refleCtance minimum (Rmin) in green
for the calculation of refleCtance ratios (R750/R550 and R750/ leaves. The leaves are arranged by increasing chlorophyll content.
R700), a spectral band width (slit width) of 5 nm (position 05 in the Pigment values are based on 3 determinations (SO < 5 %) and are
UV-2101PC) was applied with the specification that the height was given in llg' cm- 2 leaf area.
half in order to reduce the stray light.
Leaf number 6 5 4 3 2

Pigment content
Data analysis Chlorophyll a 21.70 22.50 27.80 27.30 29.00 30.80
All calculations including reflectance ratios as well as statistics Chlorophyll b 7.83 7.84 9.92 10.30 10.28 10.61
were performed with Excel (Microsoft, USA). Curve fitting was Total chlorophylls a + b 29.53 30.34 37.72 37.6 39.28 41.41
done with the help of Origin (MicroCal, USA). The normalized dif Total carotenoids x + c 5.47 5.67 6.77 8.03 8.16 7.09
ference vegetation index NOV! was calculated on the basis of the ref- Pigment ratios
lectance signatures at 680 and 800 nm as the ratio (R800-R680)/ alb 2.77 2.87 2.80 2.65 2.82 2.85
(R800+ R680) (Rouse et al., 1974; Lichtenthaler, 1994). (a+b)/(x+c) 5.40 5.33 5.57 4.68 4.81 5.84
The position of the inflection point IP of the red edge rise in the Refkctance ratios
refleCtance spectrum was determined as point of interseCtion with R750/R550 2.44 2.55 2.93 3.50 3.16 3.49
the zero line of the 2nd derivative of the refleCtance speCtrum, which R750/R700 2.82 2.99 3.53 4.10 3.86 4.17
was calculated applying the speCtrophotometer software of the UV- Red edge
2001 PC (Shimadzu). position oflP (nm) 702.0 702.0 703.5 704.5 705.6 706.0
The red reflectance minimum Rmin in the red absorption bands of position ofRmin (nm) 678.4 679.0 678.7 679.0 679.0 677.3
in vivo chlorophyll a was determined from the intersection with the
zero line in the 1st derivative of the reflectance speCtrum. It can also
be determined with the peak search program present in mOSt of the
commercial spectrophotometers.

green to light-green leaves of the aurea tobacco Su/su showed

high values for the ratio ChI alb, which declined with increas-
Results
ing chlorophyll content and age. The values of the ratio of
green to yellow pigments, chlorophylls/carotenoids (a+ b)1
Pigment content in the leaves
(x+ c) were low (2.61) in the young leaves of aurea tobacco,
In the leaves of aurea tobacco, which possesses low chloro- but increased with increasing age and chlorophyll content to
phyll levels and a retarded greening (Schindler et al., 1994), a value of 5.57 which is normal for green plant tissue. The six
the total chlorophyll content (ChI a+ b) ranged from 8.4 to green leaves of green tobacco, which shows a fast and normal
20.4 Ilg cm -2 leaf area. In the leaves of green tobacco the greening process response, exhibited values of the pigment ra-
chlorophyll amounts were higher and ranged from 29.5 to tios, Chi alb (2.65-2.87) and (a+b)/(x+c) (4.68-5.84) as
41.41lg cm -2 leaf area (Tables 1 and 2). The seven yellowish- usually found in green plant tissue (Tables 1 and 2).
486 HARTMUT K. LICHTENTHALER, ANATOLY GITELSON, and MICHAEL LANG

35
Spectral features ofreflectance spectra ofleaves
Nicotiana tabacum L.
30
The typical reflectance spectra of leaves of aurea and green r2 = 0.966
tobacco with different chlorophyll content (ChI a: aurea no. 5 25
= 12.0 llgcm- 2; aurea no. 4 = 15.511gcm- 2; green no. 5 = "#- 20
22.511g cm -2; green no. 3 = 27.311g cm -2) are shown in Fig.
0
1. They are characterised by a low reflectance in the blue LO 15
LO
spectral region between 400 to 500 nm, a reflectance max- c:: 10
imum in the green region near 550 nm, a reflectance mini-
mum near 680 nm followed by a steep rise in reflectance 5
(known as red edge) to the near infra-red region and a rela- 0
tively constant maximum reflectance above 750 nm (Fig. 1). 5 10 15 20 25 30 35
The reflectance above 750 nm is not influenced by the red ab- Chi a [1-19 cm- 2]
sorption bands of in vivo chlorophyll of leaves, and the varia-
tion of reflectance in this NIR range was relatively low for all 35
Nicotiana tabacum L.
13 leaves and ranged from 42 to 53 % reflectance. In fact, the 30
coefficient of reflectance variation R750 and R N1R (ratio of r2 = 0.974
25
standard deviation of reflectance to average reflectance value)
was less than 5 % for all 7 aurea and 6 green tobacco leaves ;? 20
e...
studied here.
g
LO
15
c:: 10
Reflectance in the blue region
5
The blue reflectance from 400 to 500 nm was characterised
by a fairly low signal (range from 5 to 9 % reflectance) in all 0
5 10 15 20 25 30 35 40 45
measured leaves (Fig. 1). Aurea tobacco leaves with a ChI a
Chi a+b [1-19 cm- 2]
content of 7 to 1211g cm -2 showed a tendency for a small re-
flectance dip near 440 nm probably due to the blue chloro- Fig. 2: Hyperbolic relationship (y = ax -b) between the green reflec-
phyll absorption band (Soret Band). This reflectance dip was tance R550 and chlorophyll a and total chlorophyll content (ChI a+
no longer seen at a leaf chlorophyll a content above 16 llg b) in green and aurea tobacco leaves.
cm- 2 .

tionship y = ax-b) to the ChI a and ChI a+ b content of the

Reflectance in the green region investigated leaves (Fig. 2). Aurea leaves with a ChI a content
The reflectance spectra showed maxima in the green region of 1211g cm -2 (aurea no. 5) exhibited a reflectance near 550
near 550 nm that were inversely correlated (hyperbolic rela- of about 30 % (Fig. 1). With increasing ChI a content the
550 nm reflectance decreased to less than 12 % in the green
tobacco leaf with a ChI a content of 27.311g cm -2 (green leaf
.,' ..... ':'.. , no. 3). The reflectance from 530 to 630 nm showed max-
50 Nicotiana tabacum L. .................................. imum sensitivity to ChI a and decreased to the same extent in
_-"1
:' .....
.... aurea no. 5 :''''' ~_ , 1 all wavelength positions of this range with increasing ChI a
.............. aurea no. 1 :l //"'- content from 7 to 30.811gcm- 2 • The curvilinear inverse rela-
;? 40 //:'
~ .•..•. green no .5 ::. tionship of ChI a and R550 exhibited a determination coeffi-
Q) :f: cient of r2 = 0.966 and total Chla+ b and R550 of r2 =0.974.
0 30 - - green no: 3 jfi
c
/;
;:.
~Q) \ .
-:-
::.

ii= 20 .,:'~
::.
Reflectance in the red region
Q)
.:Z, ·\,········~.··· . "
>. ..:.""..
L.
~ Reflectance spectra of tobacco leaves exhibited a reflec-
10 ....................:::?" " .... tance dip close to the red absorption bands of ChI a near 670
:'!..... ;;,-.;-:.: •••:;:;:: •••.;.~.:.:::'
to 680 nm. However, the variation in reflectance in this re-
o L....-&.._.....- & . . _.....-&........"---...L........"---...L..--'''---.... gion at 670 to 680 nm with respect to increasing levels of
400 500 600 700 800 900 ChI a or ChI a+ b content was relatively small. An increase in
wavelength [nm] ChI a from 7 to 22.511gcm -2 resulted only in a decrease from
approximately 10 % to 5 % in the percentage leaf reflectance
Fig. 1: Reflectance spectra of leaves with increasing chlorophyll con-
in this region, and then the reflectance remained nearly con-
tent of an aurea mutant (leaves no. 5 and no. 1) and a green variety
(leaves no. 5 and no. 3) of Nicotiana tabacum L. (see Tables 1 and stant when ChI a increased to more than 30 llg cm -2. In the
2). A minimum sensitivity to changes in chlorophyll content was blue region near 490 nm and in the red spectral region near
observed in the blue spectral range (400 to 490 nm) and near 670 nm the variation in reflectance with increasing chloro-
670 nm, whereas maximum sensitivity was found in a wide range of phyll a content was very small, and for a ChI a content above
the spectrum from 530 to 630 nm and near 700 nm. 1511gcm- 2 the relationship «reflectance vs. ChI-content» was
 Chlorophyll determinarion via leaf reflectance 487

saturated (data not shown). The similar behavior of the re- 30 Nicotiana tabacum L.
flectances in the blue range and near 670 nm (with small var-
iation and sensitivity to changing ChI a and ChI a+ b content) 25
r2 = 0.971
is documented by a very high correlation (r2 = 0.990) be-
~20
tween the reflectance near 670 nm, R670, and the reflectance ~
near 490 nm, R490, as shown in Fig. 3. 0 15
0
I'-
0:: 10 • green

Red edge ofleaf reflectance spectra and inflection point

a aurea
5
Another spectral range with high sensitivity to changes in
0
chlorophyll content was found in the reflectance near 700 nm 5 10 15 20 25 30 35
in the range of the inflection point of the red edge rise of the Chi a [lJg cm- 2l
reflectance spectrum (Fig. 1). As distinct as the green range
of the spectrum, a high spectral sensitivity to changes in the 35
ChI a and ChI a+ b level was found in a relatively narrow 30 Nicotiana tabacum L.
spectral band from 690 to 710 nm with a maximum sensitiv-
25 r2 = 0.967
ity at 700 nm. The coefficient of the reflectance variation in
this range (R700) was even higher as compared to that for ~20
~
the green reflectance R550 (data not shown). The relation-
ship between the reflectance R700 (Fig. 4), and also that of
g
I'-
15
R550 (Fig. 2), and the ChI a or ChI a+ b content was hyper- 0:: 10
bolic. In a similar way as for leaves of soybean (Chappelle et
5
al., 1992) as well as maple and chestnut (Gitelson and Merz-
Iyak, 1994a and bi 1996) a very close correlation (r2 = 0.978) 0
5 10 15 20 25 30 35 40 45
was found between the green and red reflectances, R550 and
2
R700 (Fig. 5). Chi a+b [lJg cm- l
The rise of reflectance in the near infra-red spectral range,
the red edge, can be characterised by the position of its inflec- Fig. 4: Correlation of the reflecrance at 700 nm (R700) versus rhe
tion point IP. The determination of the inflection point of the
chlorophyll a and a+ b content for 13 leaves of an aurea mutant and
green form of Nicotiana tabacum L. The lines indicate a hyperbolic
red edge is shown for a green tobacco leaf with medium ChI a relarionship berween chlorophyll content and R700. The square of
content (22.5Ilgcm-2) and for an aurea tobacco leaf with a rhe correlarion coefficient was 0.97.
low ChI a content (7llg cm -2) (Fig. 6). In the 2nd derivative
spectra the inflection point of the red edge of the aurea to-
bacco leaf (IP at 694 nm) was at shorrer wavelengths (blue tionship with a very high correlation coefficient (r2 = 0.973
shift) in comparison to the green tobacco leaf (IP at 702 nm) for ChI a and 0.966 for ChI a+ b).
(Fig. 6). The wavelength positions of the inflection point cor-
relate to the ChI a and ChI a+ b content in a curvilinear rela-
Position ofthe inflection point IP in relation to reflectance at
700nm
10 Nicotiana tabacum L. High sensitivity of the position of the IP to chlorophyll
r2 =0.990 content was observed in the spectral range from 690 to
705 nm (Figs. 1 and 7). The high correlation of ChI a content,
both with R700 and with the Ip, suggests a close relation be-
o tween these characteristics of the red edge of the reflectance
(J)

~
6
•
a
green
aurea
spectrum. In fact, a linear relation between R700 and the IP
position was found for green and aurea tobacco leaves with r2
= 0.956 (Fig. 8). Therefore, R700 measurements appear to be
a suitable parameter to obtain information on the relative po-
4 '--_~_..L...-_",--_-'-_~_-'-_...J
4 6 8 10 sition of the inflection point of the red edge and on the
R67Q (%1 leaves' chlorophyll content. The IP also correlates with the
R550 in a linear way (Fig. 8).
Fig. 3: Linear correlarion of the reflecrance ar 490 nm with that ar
670 nm (R490 versus R670) for all studied 13 leaves of an aurea mu-
tant and a green form of Nicotiana tabacum L. (chlorophyll a con- Position ofthe red reflectance minimum Rmin at the red edge
rent from 7 to 30.81lgcm -2. The solid line presents a linear relarion-
ship with the square of correlation coefficient r2 = 0.99. The wave- With increasing chlorophyll content the reflectance dip in
band near 670 nm lies in rhe range of rhe red absorption bands of the red chlorophyll absorption band near 680 nm became
chlorophyll a, whereas the waveband near 490 nm corresponds to broader (Figs. 1 and 6). The wavelength position of the mini-
joint absorprion bands of chlorophyll a, chlorophyll band carot- mum near 680 nm showed, however, similar values in the
enoids. range of 677.3 to 679 nm for the 13 leaves with different
488 HARTMUT K. LICHTENTHALER, ANATOLY GITELSON, and MICHAEL LANG

·30 a 706
Nicotiana tabacum L. Nicotiana tabacum L.
704
25 r = 0.978
2
r2 =0.973
702
~20
L E700
E-
0
0
15
n. 698 •a green
•a
r--
0:: 10 green aurea
696
aurea
5 694 a
0 692
10 15 20 25 30 35 0 5 10 15 20 25 30 35
R550 [%j Chi a [1J9 cm- 2j

Fig. 5: Linear correlation of the reflectance R700 versus that of

R550. Means of aurea and ~reen tobacco with a chlorophyll a con- 706
Nicotiana tabacum L.
tent from 7 to 30.8 IJ.g cm- . The square of the correlation coeffi-
cient was r2 =0.978. The two reflectance signatures R700 and R550
704 r2 =0.966
were found to exhibit a maximum sensitivity to variations in chloro- 702
phyll content of leaves. E 700
E-
n. 698
a green
696
a aurea
Nicotiana tabacum L. 694 a
50
692
reflectance spectra 0 5 10 15 20 25 30 35 40 45
...... aurea no. 7 Chi a+b [1J9 cm-2 j
40
- - green no. 5
~
2..... Fig.7: Position of the inflection point (IP) of the red edge for all 13
leaves of the aurea mutant and green tobacco versus chlorophyll a
~ 30
c: and total chlorophyll content. The position of the inflection point
(ll
(IP) stringly correlated with the chlorophyll a and a+ b content in an
t> 20
Q) exponential manner (solid lines). Within the group of aurea and
ii=
~ green leaves it would be possible to draw a linear relationship, the
hyperbolic relationship, however, is the best fit for both groups.
10

0 chlorophyll content (Tables I and 2). Thus, the wavelength

0.15 position reflectance minimum Rmin did not show a relation-
2nd derivative ship to the Chi a content (Fig. 9) or to the position of the in-
0.10 ... aurea no. 7 flection point IP of the red edge (Fig. 10). This excludes
--green no. 5 Rmin as a mean for in vivo chlorophyll determination. The
fact, that Rmin does not change with increasing chlorophyll
0.05 content of leaves, is emphasized by the wavelength position
of Rmin being solely determined by the in vivo Chi a absorp-
tion bands of the different chlorophyll/carotenoid-proteins of
thylakoids. These absorption bands do not change with
nm increasing or decreasing Chi content, as shown here.

-0.10+--""T'"-"""'T--....---"T"'""-........-....,..----.-~ The reflectance ratios R750lR550 and R750lR700 in

400 500 600 700 800
comparison to the NDVI
wavelength [nm)
With respect to remote sensing of the Chi content of
Fig.6: (A) Reflectance spectra of leaves of an aurea mutant (dotted plants via non-invasive reflectance measurements one has,
line) and a green form (solid line) of Nicotiana tabacum L. Differ- however, to consider that the remote determination of the ex-
ences in the chlorophyll content between the two leaves can be act wavelength position of the inflection point IP is a rela-
monitored by the shift of the red edge as estimated via the 2nd de-
rivative of the reflectance spectrum. The position of the inflection
tively critical and difficult matter, and only accurate when
point IP (as intersection with the zero line) of the red edge in the high spectral resolution radiometry is used. A remote deter-
green tobacco leaf was near 702 nm, whereas that of the aurea leaf mination of the exact wavelength positions of IP is, however,
was shifted to a shorter wavelength at 694 nm (which is known as not required when one wants to obtain information on the
blue shift of the IP). chlorophyll content of leaves and canopies. It is sufficient to
 Chlorophyll determination via leaf reflectance 489

708
concentrate on those reflectance parameters which show great
sensitivity to changes in Chi a and Chi a+ b content, such as 706
• •
the reflectance signals R550 and R700. When the latter are
related to a «standard reflectance parametec», which shows a
704
Nicotiana tabacum L.
•
•
low or no sensitivity to changes in Chi content, such as
702 ••
R750, one can establish new reflectance ratios for the non-in- § 700 Q
0
vasive Chi determination. 9: 698 Q)

Due to the high sensitivity of the reflectance signals near

550 nm and 700 nm and the insensitivity of the NIR reflec-
696 •
0
green
aurea
0

694 0
692
668 670 672 674 676 678 680 682
708 Nicotiana tabacum L.
Rmin (nm)

704
r2 =0.938 Fig. 10: There is no correlation between the wavelength position of
the inflection point IP and that of the red reflectance minimum
E Rmin.
oS 700
9: • green
696 o aurea tance at 750 nm with respect to the chlorophyll content de-
scribed here and also found in chestnut and maple leaves (Gi-
o
telson and Merzlyak, 1994 b), we determined a very high cor-
5 10 15 20 25 30 35 relation between the two reflectance ratios R750/R550 and
R550 [%j R750/R700 and the Chi a content with r2 of 0.960 and
0.962, and the total Chi a+ b content of r2 0.936 and 0.927,
respectively (Fig. 11). Consequently, a linear relationship was
708 Nicotiana tabacum L. observed between the reflectance ratios R750/R550 and
, r2 = 0.956 R750/R700 with a high correlation coefficient of r2 = 0.984
704 • (Fig. 12).
The normalized difference vegetation index NOVI is often
E used for estimation of the chlorophyll content of leaves and
oS 700
c.. terrestrial vegetation as well as green biomass. The great dis-
advantage of the NOVI for chlorophyll determination is,
• green
696 however, that at medium and higher chlorophyll levels it does
o aurea
not show a good response or correlation to the chlorophyll
692 L-~..L...-~..L-~-'-~-'-_-'-~...L-~-l content. In fact, the NOVI values declined only from 0.81 to
o 5 10 15 20 25 35 0.75 when the Chla content decreased from 31 to IOllg cm- 2,
R700 [%j and the total Chi a+ b level dropped from 41 to I3llg cm- 2
(Fig. 13). This connection is based on the extremely low sen-
Fig. 8: Linear correlation of the position of the inflection point (IP)
sitivity of the R680 reflectance to changes in chlorophyll
with R550 nm and R700 nm for all 13 leaves of an aurea mutant
and a green tobacco variety. The correlation coefficient r2 was some- content at medium or high chlorophyll levels. In contrast, the
what higher for the correlation of the IP with R700 than with new reflectance ratios R750/R550 and R7501R700 estab-
R550. lished here showed a very high linear correlation to the Chi a
and total CW a+ b content for the whole range of chlorophyll
levels with r2 = 0.956 and 0.960, respectively (Fig. 13). In the
684 range of Chi a content from 7 to 30.8Ilgcm-2, the coefficient
Nicotiana tabacum L.
of NOVI variation was only 4.4 %, whereas for the ratios
682
R750/R550 and R750/R700 values of 28.6 % and 29.5 %,
680 respectively were achieved.
I 0 ocP Q;)
• • ••
•
678 0
.~ 676

•
Discussion
a::: 674 green
0 aurea
672 Reflectance of plant leaves exhibits signals in the visible
domain (VIS) ranging from 400 to 700 nm, in the near-infra-
670
5 10 15 20 25 30 35 red domain (NIR) from 700 to 1300 nm and in the middle-
infrared domain (MIR) from 1300 to 2500 nm (Guyot, 1990;
Chi a [l1g cm- 2j
Lichtenthaler, 1989 and 1994). The middle-infrared domain
Fig. 9: The wavelength position of the red reflectance minimum contains leaf reflectance minima caused by strong water
Rmin (in nm) is independent of the Chi a content of the 13 green absorption bands near 1450,1950 and 2500nm, but this range
and aurea tobacco leaves studied. is neither of interest nor suitable concerning the determina-
490 HARTMUT K. LICHTENTHALER, ANATOLY GITELSON, and MICHAEL LANG

4 Nicotiana tabacum L. 4 Nicotiana tabacum L.

--
0
l.O
l.O
a:
3 o 3
0
I'-
a:
r2 =0.962
.-
;; 2 ;; 2
l.O l.O
I'- I'-
a: a:

0 0
0 10 20 30 40 0 10 20 30 40
Chi a [~g cm-2 j Chi a [~g cm-2 j

4 Nicotiana tabacum L. 4 Nicotiana tabacum L.

0 3 o 3
l.O o
l.O
a: ~
-2 -2
~ ~
I'-
~ a:
Fig. 11: Linear correlarion of the new vege-
tation indices R750/R550 (A) and R7501
R700 (B) with the chlorophyll a and total
°0L-----,L10---2.... 0-~50
0--3:':0:----4.... °0L---1c':0---2L..0--3c':0---4L..0----JSO chlorophyll content for all 13 green and au-
Chi a+b [~g cm-21 Chi a+b [~g cm-2 j rea tobacco leaves studied.

1.0
3.5 Nicotiana tabacum L. Nicotiana tabacum L.
4 fee 0 - 0.8
r2 :: 0.984
3.0
a o
LO
LO :20:: 3 0.6
z
n::: 2.5
0
a 0.4 :5
LO 2.0 • green
I'- o NDVI
n::: o o aurea
R750/R550 0.2
1.5
o ~~.L-~.L--"-..L-~-':-~~-'--L-""""-' 0.0
1.5 2.0 2.5 3.0 3.5 4.0 4.5 o 5 10 15 20 25 30 35
R750 I R700 Chi a [IJQ cm-21

Fig. 12: Linear relationship between the two new vegetation indices Nicotiana tabacum L.
R750/R550 and R750/R700 as established with 13 leaves of differ- 4 0.8
ent chlorophyll content of a green and an aurea variety of tobacco.
o
0.6
z
o
tion of photosynthetic pigments. The NIR region includes 0.4 :5
the chlorophyll-dependent steep rise of reflectance (red edge)
o NDVI
as well as the infrared reflectance plateau which depends 0.2
upon the leaf structure as well as on the size of cells and aerial R750/R700
interspaces in the leaf mesophyll (Lichtenthaler, 1989; Gaus- 00~~L-~.L-4-.L..--"-..L-~-':-~-':-""""-' 0.0
man and Quisenberry, 1990). The VIS domain is character- 5 10 15 20 25 30 35
2
ised by low reflectance and low transmittance due to the ab- Chi a [IJQ cm- ]
sorption bands of chlorophylls and carotenoids in the blue
and red part of the electromagnetic spectrum. The reflectance Fig. 13: Comparison of the new vegetation indices R750/R550 and
maximum in the green region near 550 nm, which is due to R750/R700 with the old vegetation index NOV!. A. Comparison of
an absorption minimum of leaf pigments in this region, is re- R750/R550 and versus Chi a. The reflectance ratio R750/R550
shows a better correlation to the chlorophyll a content (r2 =0.96)
sponsible for the green colour of leaves together with leaf
than the vegetation index NOV! (r2 =0.66). B. Comparison of the
transmittance which exhibits similar spectral characteristics as variation of R750/R700 and NOV! with chlorophyll a content.
the reflectance (Lichtenthaler, 1994). Both new vegetation indices exhibited a better correlation to the
Remote sensing of terrestrial vegetation via reflectance chlorophyll a content than the old index NOV!. The new vegetation
spectroscopy led to the formation of ratios of reflectance indices R750/R700 and R750/R550 are much more sensitive to
bands (Howard, 1991). The ratio of infrared to red reflec- changes in chlorophyll content than the old index NOV!.
 Chlorophyll determination via leaf reflectance 491

tance was established as vegetation index in order to estimate The inflection point IP of the red edge also exhibited a
green productivity. Biomass productivity was evaluated ap- very good curvilinear correlation to the chlorophyll content
plying the normalized difference vegetation index NOVI of leaves. Since its blue shift to shorter wavelengths with de-
(Tucker et aI., 1980), which is based on the red (near creasing chlorophyll content is associated with a considerable
680 nm) and NIR reflectances. Recent studies (Buschmann increase in the reflectance yield, it is a very exact indicator of
and Nagel, 1993) demonstrated that there exists no clear cor- changes in the chlorophyll content. From complete leaf re-
relation between NOV! and Chi a content, except for very flectance spectra, which can be recorded by high resolution
low chlorophyll levels which are usually not found in leaves radiometers, the determination of the wavelength position of
and plants. This is due to the fact that the R680 reflectance the IP is possible. In case of remote sensing of the chlorophyll
exhibits only an extremely low sensitivity towards changes in content the determination of the IP is quite difficult, since it
the Chi a or total Chi a+ b content at levels higher than 5 to would require reflectance sensors of high spectral resolution,
71lg cm ~2 (cf. Fig. 1). Thus a decrease in chlorophyll a+ b which are not available in the existing and future satellite sys-
content from 41 to 131lg cm -2 leaf area only caused a de- tems. The present reflectance sensors provide the radiance in
crease in the NOV! values of 9 % (see Fig. 13). An approx- several quite wide bands in the visible and NIR-range of the
imate linear relationship and a larger sensitivity of the NOV! spectrum.
values to changing chlorophyll contents was found in aurea The efficiency of the new vegetation indices, as compared
tobacco leaves only below a Chi a and Chi a+ b content of 7 to the presently used NOVI, can be estimated using their
and 10 Ilg cm -2, respectively. Except for aurea mutants, such coefficients of variation. For the same Chi a range from 7 to
low chlorophyll levels are only found in highly damaged and more than 30 Ilg cm -2, the coefficients' ratio of the variation
stressed plants, e.g. during autumnal leaf senescence. of the ratios R750/R550 and R750/R700 to that for the
With respect to chlorophyll determination we examined NOV! were 6.5 and 6.7, respectively. It means that the newly
the new reflectance ratios R750/R550 and R750/R700, established indices are more than 6-fold more sensitive to var-
which are very accurate for an in vivo chlorophyll determina- iations in Chi a content than the NOV!.
tion: Since the reflectance R700 is a good indicator of the red
(i) The infrared to green reflectance ratio, R750/R550, is edge position and also of the chlorophyll content, as shown
composed of the relatively insensitive (with respect to chloro- here, a determination of the exact wavelength position of the
phyll content) infra-red signal R750 and the chlorophyll-con- IP is not required if one wants to remotely sense the chloro-
tent sensitive green reflectance signal R550. This is why the phyll content of terrestrial vegetation. In such cases it is
values of the ratio R750/R550 decrease a lot from 3.5 to 1.4 highly sufficient to measure the reflectance in a narrow band
(-60 %) with decreasing Chi a+ b content from 41 to near 700 nm or even in somewhat broader bands (near
8.4llg cm- 2 (-80 %)" exhibiting a variation coefficient of 550 nm and near 750 nm). From these reflectance signals the
28.6 %. For the two tobacco varieties (green form su/su and ratio R750/R700, and alternatively the ratio R750/R550 are
aurea mutant Su/su) we found an exact linear correlation be- formed, which then function as indicators of the chlorophyll
tween R750/R550 and Chi a content (r 2 = 0.956) (see Fig. content. The reflectance in the NIR-range does not necessar-
13). Our results thus confirm former studies of Buschmann ily have to be R750, it could also be R780 or R800, since the
and Nagel (1993 and 1995), who found best linear correla- reflectance in this NIR region is fairly constant over a longer
tion between chlorophyll content and the reflectance ratio wavelength range. By selecting the appropriate filters one
infra-red/green, defined as log (R800/R550). should, however, stick very closely to the reflectance bands
(ii) The second newly established reflectance ratio R7501 R550 and R700, which are very sensitive to changes in
R700 is based a) on the NIR reflectance signal R750 which is chlorophyll content.
fairly inert and changes very little in tobacco leaves of differ- The wavelength position of the red reflectance dip in the
ent age and development, and b) on the reflectance signal red chlorophyll a absorption bands, Rmin, was found to be
R700 which responds with a large amplitude and sensitivity constant at wavelengths between 677.3 and 679 nm. It is
towards changes in the Chi a and Chi a+ b content of leaves. therefore independent of the chlorophyll content of the
The NIR signal R750 thus serves as an internal reference leaves of green or aurea tobacco (see Fig. 9). This demon-
with respect to reflectance properties caused by leaf structure, strates that the different chlorophyll-proteins with Chi a
whereas the R700 indicates changes going on in the leaf forms showing maxima at various red wavelengths (672,
chlorophyll content. In fact, the newly established ratio 678 nm) (Lichtenthaler et al., 1981), and the minor Chi a
R750/R700 declined from values of 4.2 to 1.6 (-62 %) with forms P680 and P700 and others (e.g. French et al., 1972;
a decline in Chi a content of -80 %. The ratio R750/R700 Lawlor, 1993) are present in leaves, no matter whether leaves
possesses a high linear correlation to the Chi a content with r2 contain a high or low chlorophyll content. The leaves of the
= 0.960 (see Fig. 13). aurea tobacco show higher Chi alb values (Table 1) than those
Our results with differently pigmented leaves of green and of green tobacco (Table 2), which is an indicator that the au-
aurea tobacco are in full agreement with the results of Gitel- rea leaves possess lower amounts of LHCPs than leaves of
son and Merzlyak (1994 a and b), who found for maple and green tobacco. Although the relative proportions of LHCPs
chestnut leaves a linear correlation of the two ratios R7501 with respect to the other chlorophyll proteins in green and
R550 and R750/R705 with the Chi a content (r2 >0.97). aurea tobacco are somewhat different, this has no influence
This demonstrates the general validity and significance of the on the overall wavelength position of the red reflectance min-
two new ratios R750/R550 and R750/R700 as excellent indi- imum Rmin. Small differences in the Rmin from leaf to leaf
cators of the in vivo chlorophyll content of plants. or even within different points of a leaf may also be due to
492 HARTMUT K. L!CHTENTHALER, ANATOLY GITELSON, and MICHAEL LANG

the fact that Rmin does not only reflect chlorophyll absorp- CARTER, G. A.: Responses of leaf spectral reflectance to plant stress.
tion. At low reflectance signals also the internal leaf optics American J. Bot. 80,239-243 (1993).
may show a larger influence in this reflectance range as well - Ratios of leaf reflectances in narrow wavebands as indicators of
as the relative contribution of the emitted red chlorophyll plant stress. International]. Remote Sens. 15,697-703 (1994).
fluorescence with a maximum near 684 to 690 nm (F684- CHAPPELLE, E. w., M. S. KiM, and J. E. McMURTREY: Ratio analy-
sis of reflectance spectra (RARS): An algorithm for the remote es-
F692) (Lichtenthaler and Buschmann, 1987; Buschmann and
timation of the concentrations of chlorophyll a, chlorophyll b,
Lichtenthaler, 1988; Buschmann et aI., 1994), which increa- and carotenoids in soybean leaves. Remote Sens. Environ. 39,
ses with decreasing chlorophyll content. Therefore our find- 239-247 (1992).
ings exclude the application of the wavelength position of the CURRAN, P. J., J. L. DUNGAN, B. A. MACLER, and S. E. PLUMMER:
reflectance parameter Rmin near 680 nm in the remote sens- The effect of a red leaf pigment on the relationship between red
ing of terrestrial vegetation. edge and chlorophyll concentration. Remote Sens. Environ. 35,
69-76 (1991).
FRENCH, c. S., Y. S. BROWN, and M. C. LAWRENCE: Four universal
Conclusion forms of chlorophyll a. Plant Physiol. 49,421-429 (1972).
GAUSMAN, H. W. and J. E. QUISENBERRY: Spectrophotometric de-
tection of plant leaf stress. In: KATTERMAN, F. (ed.): Environ-
The two new reflectance ratios R750/R550 and R7501 mental Injury to Plants, pp. 257-280. Academic Press, San
R700 are very suitable parameters for a rapid, non-destructive Diego (1990).
chlorophyll determination of leaves and plants to be used in GITELSON, A. and M. N. MERZLYAK: Spectral reflectance changes
stress physiology and remote sensing of vegetation. These two associated with autumn senescence of Aesculus hippocastanum L.
new vegetation indices have the great advantage that they and Acer platanoides L. leaves. Spectral features and relation to
match with the chlorophyll content of leaves and plants in a chlorophyll estimation. J. Plant Physiol. 143,286-292 (1994 a).
linear way. For this reason they are much better suited for re- - - Quantitative estimation of chlorophyll a using reflectance
mote sensing of chlorophyll content of terrestrial vegetation spectra: Experiments with autumn chestnut and maple leaves. J.
than the NOVI, which is only applicable at very low leaf Photochem. Photobiol. (B) 22,247-252 (1994 b).
chlorophyll levels, but insensitive and not suitable at a me- - - Signature analysis of leaf reflectance spectra: algorithm devel-
dium and higher chlorophyll content. opment for remote sensing of chlorophyll. J. Plant Physiol. 148,
495-500 (1996).
GITELSON, A., Y. KAUFMAN, and M. N. MERZLYAK: An atmospher-
Acknowledgements ically resistant «green» vegetation index (ARGl) for EOS-
MODIS. J. Remote Sensing of Environment. (1996a) in press.
We wish to thank Gabrielle Johnson for checking the English GITELSON, A., M. N. MERZLYAK, and H. K. LICHTENTHALER:
text. Detection of red edge position and chlorophyll content by reflec-
tance measurements near 700nm. J. Plant Physiol. 148,501-508
(1996 b).
References GUYOT, G.: Optical properties of vegetation canopies. In: STEVEN,
M. D. and J. A. CLARK (eds.): Applications of Remote Sensing in
BARET, F., S. JACQUEMOUD, and G. GUYOT: Modeled analysis of the Agriculture, pp. 19-27. Buuerworths Scientific Ltd., London
biophysical nature of spectral shift and comparison with informa- (1990).
tion content of broad bands. Remote Sens. Environ. 41, 133-142 HORLER, D. N., M. DOCKRAY, and J. BARBER: The red edge of plant
(1992). leaf reflectance. International J. Remote Sens. 4, 273-288
BUSCHMANN, C. and H. K. L!CHTENTHALER: Reflectance and (1983).
chlorophyll fluorescence signatures of leaves. In: LICHTEN- HOWARD, J. A.: Remote Sensing of Forest Resources, pp. 47 ff.
THALER, H. K. (ed.): Applications of Chlorophyll Fluorescence, Chapman & Hall, London (1991).
pp. 325-332. Kluwer Academic Publishers, Dordrecht (1988). LANG, M., H. K. LICHTENTHALER, M. SOWINSKA, F. HEISEL, J. A.
BUSCHMANN, C. and E. NAGEL: In vivo spectroscopy and internal MIEHE, and F. TOMASINI: Fluorescence imaging of water and
optics of leaves as basis for remote sensing of vegetation. Internat. temperature stress in plant leaves. J. Plant Physiol. 148,613-621
J. Remote Sens. 14,711-722 (1993). (1996).
- Reflexionsspektren von Blattern und Nadeln als Basis flir die LAWLOR, D. W.: Photosynthesis: Molecular, Physiological and Envi-
physiologische Beurteilung von Baumschaden. In: BITTLING- ronmental Processes, 2nd ed., Longman Scientific & Technical,
MAIER, L., W. REINHARDT, and D. SIEFERMANN-HARMS (eds.): Harlow Essex (1993).
Waldschaden im Schwarzwald, pp. 288-298. Ecomed, Lands- LrCHTENTHALER, H. K.: Chlorophylls and carotenoids: pigments of
berg (1995). photosynthetic biomembranes. Methods Enzymol. 148, 350-
BUSCHMANN, c., U. RiNDERLE, and H. K. L!CHTENTHALER: Detec- 382 (1987).
tion of stress in coniferous forest trees with the VIRAF spectro- - Possibilities for remote sensing of terrestrial vegetation by a com-
meter. IEEE Transactions on Geoscience and Remote Sensing 29, bination of reflectance and laser-induced chlorophyll fluores-
96-100 (1991). cence. In: Proceed. Internat. Geoscience and Remote Sensing
BUSCHMANN, c., E. NAGEL, K. SZAB6, and L. KOSCANYI: Spectro- Symposium, IGARSS'89, Vancouver, Vol. 3, pp. 1349-1354. Li-
meter for fast measurements of in vivo reflectance, absorptance brary of Congress, No. 89-84217 (1989).
and fluorescence in the visible and near-infrared. Remote Sens. - Spektroskopische Eigenschaften von Pflanzen und ihre Nutzung
Environ. 48, 18-24 (1994). zur Fernerkundung der Vegetation. Fridericiana 49, 25-45
BUTLER, W. L. and D. W. HOPKINS: Higher derivative analysis of (1994).
complex absorption spectra. Photochem. PhotobioI. 12, 439- Vegetation stress: an introduction to the stress concept in plants.
456 (1970). J. Plant Physiol. 148,4-14 (1996).
 Chlorophyll determination via leaf reflectance 493

LICHTENTHALER, H. K. and C. BUSCHMANN: Reflectance and SANTRUCEK, J., P. SIFFEL, M. LANG, H. K. LICHTENTIIALER, C.
chlorophyll fluorescence signatures of leaves. In: Proceedings of SCHINDLER, H. SYNKOVA, V. KONECNA, and K. SZABO: Phorosyn-
the Remote Sensing Symposium (IGARSS), Michigan, Vol. II, thetic activity and chlorophyll fluorescence parameters in aurea
pp. 1201-1206. The University of Michigan, Ann Arbor 1987. and green forms of Nicotiana tabacum. Photosynthetica 27, 529-
LICHTENTHALER, H. K., V STRAUB, and K. H. GRUMBACH: Une- 543 (1992).
qual formation of prenyl-Iipids in a plant tissue culture and in SCHINDLER, c., P. REITH, and H. K. LICHTENTHALER: Differential
leaves of Nicotiana tabacum L. Plant Science Leners 4, 61-65 levels of carotenoids and decrease of zeaxanthin cycle perform-
(1975). ance during leaf development in a green and an aurea variety of
cobacco. J. Plant Physiol. 143,500-507 (1994).
LICHTENTHALER, H. K., G. BURKARD, G. KUHN, and U. PRENZEL:
SCHMUCK, G., H. K. LICHTENTHALER, G. KRITIKOS, V AMANN, and
Light-induced accumulation and stability of chlorophylls and
B. N. ROCK: Comparison of terresrrial and airborne reflection
chlorophyll proteins during chloroplast development in radish
measurements of forest trees. In: Proceedings of the Remote
seedlings. Z. Naturforsch. 36c, 421-430 (1981).
Sensing Symposium (IGARSS), Michigan, Vol. II, pp. 1207-
ROCK, B. N., T. HOSHIZAKl, H. K. LICHTENTHALER, and G. 1212. The University of Michigan, Ann Arbor (1987).
SCHMUCK: Comparison of in situ spectral measurements of forest THOMAS, ]. R. and H. W. GAUSSMAN: Leaf reflectance vs. leaf
decline symptoms in Vermont (USA) and the Schwarzwald chlorophyll and carotenoid concentration for eight crops. Agron.
(ER.G.). In: Proceedings of the Remote Sensing Symposium J. 69,799-802 (1987).
(IGARSS), Zurich 1986, Vol. III, pp. 1667-1672. ESA Publica- TUCKER, C. J., B. N. HOLBEN, J. H. ELGIN, and J. E. McMURTREY:
tions Division, Noordwijk (1986). Relation of spectral data co grain yield variation. Photogrammet-
ROUSE, J. w., R. H. HAAS, J. A. SCHELL, D. W. DEERING, and J. C. ric Engineering and Remote Sensing 46,657-666 (1980).
HARLAN: Monitoring the vernal advancement of retrogradation YODER, B. ]. and R. H. WARING: The normalised difference vegeta-
of natural vegetation. NASA/GSFC, Type III, Final Report. rion index of small Douglas-fir canopies with varying chlorophyll
Greenbelt, MD, USA (1974). concentrations. Remote Sens. Environ. 49,81-91 (1994).