Directory UMM :Data Elmu:jurnal:T:Tree Physiology:vol17.1997:

Tree Physiology 17, 767--775
© 1997 Heron Publishing----Victoria, Canada

Photosynthetic decline and pigment loss during autumn foliar
senescence in western larch (Larix occidentalis)
SELMA I. ROSENTHAL1,2 and EDITH L. CAMM3
1

Department of Botany, University of British Columbia, BC V6T 1Z4, Canada

2

Present address: Department of Plant Pathology, Physiology and Weed Science, Virginia Tech, Blacksburg, VA 24061-0330, USA

3

University College of the Fraser Valley, Abbotsford, BC V2S 4N2, Canada

Received May 15, 1996

Summary We measured needle pigment content and photosynthetic rates of 1-year-old western larch (Larix occidentalis

Nutt.) during autumn foliar senescence. Chlorophyll (Chl) and
carotenoid (xanthophyll + β-carotene) contents of needles
declined 11 and 17%, respectively, before CO2 assimilation rate
began to decline. Chlorophyll a/b ratio, Chl/carotenoid ratio,
photochemical efficiency (Fv/Fm), and photochemical quenching did not begin to decline until late in senescence. Internal
CO2/ambient CO2 did not change during needle yellowing. In
seedlings in warmed soil (average 3 °C above natural conditions), the decline in needle chlorophyll content was delayed
by 10 days and the decline in CO2 assimilation rate was delayed
by 5 days, compared with seedlings in soil at ambient temperature. In seedlings exposed to an extended 16-h photoperiod, the
decline in needle chlorophyll content was delayed by 32 days,
and the decline in CO2 assimilation rate was delayed by
21 days, compared with seedlings exposed to natural day
lengths. In addition to delaying the onset of needle senescence,
the treatments affected the sequence of events during senescence. Differences among treatment groups provide evidence
that the onset of pigment loss and photosynthetic decline and
the sequence of events during needle senescence are affected
by soil temperature and day length.
Keywords: chlorophyll fluorescence, gas exchange, leaf senescence, photoperiod, soil temperature.

Introduction

Autumn foliar color change in deciduous perennials is triggered both by leaf-specific maturation processes and environmental conditions. The precise effects of individual
environmental conditions on leaf senescence have proved difficult to identify because several factors, such as air temperature, day length and rainfall, often change at the same time. In
addition, some cellular components deteriorate during senescence as a direct result of the environment, whereas other
components are only indirectly affected by the environment.
Previous studies on autumn senescing tissue in deciduous trees
examined pigment degradation (Goodwin 1958, Sanger 1971),

chloroplast ultrastructure (Cunninghame et al. 1982), gas exchange (Matyssek et al. 1991), chloroplast efficiency (Adams
et al. 1990), respiratory enzymes (Dean et al. 1993), fluorescence emission spectra (Lang and Lichtenthaler 1991), and
amounts of plant hormones (Osborne 1973). However, in these
studies no attempt was made to relate the findings with
changes in autumn weather. Other studies relating weather
conditions to the timing of senescence were not accompanied
by detailed physiological measurements (Addicott and Lyon
1973, Kozlowski et al. 1991, Worrall 1993).
Environmental conditions affect the onset of leaf senescence
as well as the process of senescence, i.e., the sequence and rate
of changes (Smart 1994). One hypothesis is that autumn
weather and cumulative stress imposed on leaves trigger senescence-related genes that coordinate a preprogrammed set of
events. This leads to the prediction that the onset of autumn

foliar senescence will vary with environmental conditions, but
that senescence processes will be relatively insensitive to
autumn conditions. An alternative hypothesis is that leaf components are affected differently by autumn conditions, in
which case both the onset and process of senescence will vary
according to environmental conditions. For example, in a year
when air temperature begins to decline before irradiance declines, carbon reduction may decline before light harvesting
capacity. In contrast,when autumn temperatures remain high,
a series of phytochrome-mediated events could initiate senescence. The existence of different patterns of senescence has
been documented for several systems (Pell and Dann 1991),
but not for the interrelated events occurring during autumn
foliar senescence.
We collected data on the photosynthetic apparatus of 1-yearold western larch (Larix occidentalis Nutt.) to quantify the
influence of autumn weather on foliar senescence. We compared outdoor seedlings under natural conditions to outdoor
seedlings with soil heating or supplemental light to extend the
photoperiod. The experimental design allowed us to quantify
the effect of soil temperature and photoperiod regime on foliar
senescence. The environmental manipulations affected both
the onset and pattern of decline in photosynthetic function.

768

ROSENTHAL AND CAMM

Materials and methods
Experimental design
One-year-old western larch (Larix occidentalis) seedlings, obtained from Seedlot 5266, Thompson Okanagan Dry Seed
Planning Zone, BC (49° N, 119° W, elevation 1150 m) were
grown outdoors in 15-cm diameter pots from March until
August and watered regularly at the experimental site in Vancouver, BC (49° N, 123° W, elevation 90 m). In late August
(Julian date 240), potted trees were divided into three treatment groups: seedlings in natural conditions (control) (n = 26),
seedlings in warmed soil (n = 34), and seedlings receiving
extended photoperiod (n = 31). All of the seedlings were
placed in gravel-bottomed cold frames under a translucent
plastic roof. The frame for seedlings in warmed soil was heated
with cables under the gravel, which raised soil temperature an
average of 3 °C above ambient during the experiment. Soil
temperature probes were placed at a depth of 3 cm in the 15-cm
diameter pots. For seedlings receiving an extended day length,
a 16-h photoperiod was provided by means of supplemental
fluorescent lights (100 µmol m −2 s −1). A CR10 data logger

(Campbell Scientific, Edmonton, AB) recorded air temperature, soil temperature and relative humidity. Photosynthetic
photon flux (PPF) was measured with an LI-190SA quantum
sensor (Li-Cor, Inc., Lincoln, NE).
Our choice of a deciduous conifer allowed sequential harvesting of same age needles. On 37 days over a 15-week period
beginning on Julian day 247 (September 4), needle samples
(three per treatment) each consisting of one needle from a side
branch of each seedling were collected randomly from all
treatments between 0730 and 0930 h to insure three independent and representative samples from each treatment group
on each measurement day. Sampling was stopped when CO2
assimilation rates (see below) became negative. Because of a
shortage of material, we did not collect from seedlings in the
extended photoperiod treatment between days 320 and 338
(November 15 and December 3). Larix occidentalis is deciduous, though a few acropetal needles may remain green
throughout the winter. Because these wintergreen needles
(Richards and Bliss 1985) occur on the shoot tips, they were
not in our sampling area.
Measurements of pigment contents and photosynthetic rates
Chlorophyll and carotenoid (xanthophyll + β-carotene) contents were determined after acetone extraction (Lichtenthaler
and Wellburn 1983). Carbon assimilation rate at a PPF of
1000 µmol m −2 s −1 was measured at 20--25 °C with an ADCLCA2 (Analytical Development Company, Hoddesdon, UK)

infrared gas analyzer (IRGA) equipped with a narrow leaf
chamber. Leaf areas of individual needles were determined
from length and width measurements, and adjusted for needle
morphology (Vance and Running 1985). Stomatal conductance and internal CO2 concentrations were measured with the
IRGA.
Photosynthetic oxygen evolution was measured with a leaf
disc oxygen electrode (Hansatech, Kings Lynn, U.K.)
equipped with a red light emitting diode (660 nm). A disc with

a composite flat leaf surface area of 10 cm2 was made by
placing 30 needles side by side, trimming the apex and base
from the needles, and holding them together by four narrow
pieces of tape. The tape was placed at the edge of the leaf disc:
stomata occurred mostly on the adaxial leaf surface. To maintain chamber CO2 concentration, the mat in the leaf disc
chamber contained 1.0 M NaHCO3 adjusted to pH 9 with
1.0 M Na2CO3, and the chamber was flushed with 50 ml l −1
CO2 from a 1.0 M NaHCO3 solution (Walker 1987). Oxygen
measurements were corrected for respiration, which was measured both after 10 min in the dark and after 5--10 min of
light-saturating illumination. Respiration was assumed to remain constant as the actinic light increased.
Chlorophyll a fluorescence

Photochemical efficiency (Fv/Fm) and photochemical quenching (qQ, Kooten and Snell 1990) were measured with a pulse
modulated PAM 101/103 fluorometer (H. Walz, Effeltrich,
Germany) equipped with a Schott lamp (Schott, Mainz, Germany). Needles were dark-adapted for 10 min. The pulse
intensity to achieve close to saturation in a dark-adapted leaf
was determined by measuring Fv/Fm for green needles taken in
October from trees grown in cold frames at a PPF of 600 µmol
m −2 s −1 (n = 8 per light intensity) over a range of PPFs:
1300 µmol m −2 s −1 = 0.780 (0.023 SE); 2000 µmol m−2 s −1 =
0.781 (0.022 SE); 4000 µmol m−2 s −1 = 0.785 (0.020 SE); 7000
µmol m−2 s −1 = 0.782 (0.021 SE); and 13,000 µmol m−2 s −1 =
0.793 (0.013 SE). We used a saturating pulse of 2000 µmol
m −2 s −1, which was three to 10 times higher than the irradiance
that the plants were grown in. The method suggested by Markgraf and Berry (1990), which estimates the ‘‘true’’ height of a
saturating pulse, was used to correct for potential problems
caused by subsaturating pulses at higher actinic irradiances. At
low irradiances, L waves were visible (Larcher and Neuner
1989); however, calculating photochemical quenching with or
without L waves did not change the interpretation of the data.
Determination of break points and the onset of senescence
We determined the date of decline in the photosynthetic parameters (e.g., Chl content and carbon assimilation rate) as

previously described (Rosenthal and Camm 1996). Briefly, we
estimated a series of piecewise linear regressions (Neter et al.
1990, Wilkinson 1990) for each parameter as a function of
date, in which the slope of the first segment was constrained to
equal zero on the assumption that the parameter had not yet
begun to decline. The slope of the second segment was allowed
to differ from zero on the assumption that the parameter was
declining. The break point between the two segments was the
date at which the decline in the parameter is assumed to have
begun. Finally, separate piecewise linear regressions were estimated by imposing break points from the first measurement
day in August to the second to last measurement day in December. The date associated with the piecewise linear regression
having the lowest sum-of-squared residuals was selected as the
best estimate of the date at which a parameter began to decline.

TREE PHYSIOLOGY VOLUME 17, 1997

FOLIAR SENESCENCE IN WESTERN LARCH

Results
Environmental conditions and pigment loss

At the start of the experiment on Julian day 240 (August 27),
mean air temperature at needle harvest was 23 °C (Figure 1a)
and declined to a mean of 12 °C by Julian day 285 (October 11), when chlorophyll content (see below) began to decline. Soil temperature in the control treatment averaged 11.3
± 2.5 °C SD (n = 72) from September to December, with a
minimum of 5.9 °C and a maximum of 16.7 °C. In the warmed
soil treatment, soil temperature averaged 14.2 ± 2.1 °C SD
(n = 72, Figure 1b) with a minimum of 9.3 °C and a maximum
of 18.6 °C. For all seedlings, photosynthetic photon flux
(PPF), measured in the morning at needle harvest, declined
from 600 µmol m −2 s −1 at the start of the experiment to
200 µmol m −2 s −1 by Julian day 285 (Figure 2a). Day length
(Atmospheric Environmental Services, Vancouver, BC) declined from 13.5 to 8.5 h (Figure 2b). Day length was 11 h
when leaf chlorophyll content began to decline. Relative humidity increased from 60% in late summer and remained
consistently above 90% after Julian day 285 (Figure 2c).
Figures 3 and 4 show the time courses of chlorophyll and
carotenoid loss in seedlings in the three treatments. Among the
treatments, mean chlorophyll contents of green needles, meas-

Figure 1. (a) Air temperature (°C) at needle harvest for control seedlings
(r), seedlings with warmed soil (s), and seedlings receiving 16-h

photoperiods (q). (b) Minimum soil temperature for warmed-soil seedlings (dashed line) averaged 3 °C above soil temperatures for control
seedlings and seedlings receiving 16-h photoperiods (solid line). Maximum and minimum values for control seedlings were 16.6 and 5.9 °C,
respectively, and for warmed-soil seedlings, 18.6 and 9.3 °C.

769

ured before chlorophyll content began to decline, were not
significantly different (P > 0.10) (Table 1). The onset of the
decline in chlorophyll content, determined from piecewise
linear regressions, was Julian day 285 for control seedlings,
Julian day 295 for warmed-soil seedlings, and Julian day 317
for long-day seedlings. For each treatment group, the sum-ofsquared residual value for chlorophyll (Figure 5) had clearly
begun to rise above its minimum before the decline in CO2
assimilation rate began (see below). For example, among control seedlings, chlorophyll content declined 11%, from 276 to
245 mg Chl m −2, before the rate of CO2 assimilation began to
decline. We conclude, therefore, that the decline in chlorophyll
content preceded the decline in CO2 assimilation rate.

Figure 2. (a) Photosynthetic photon flux (PPF) (µmol m −2 s −1) measured at needle harvest (between 0730 and 0930 h) for seedlings
senescing outdoors (control, r), with warmed soil (s) and receiving

16-h photoperiods (q). (b) Photoperiod during the experiment. (c)
Relative humidity at needle harvest.

TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com

770

ROSENTHAL AND CAMM

Figure 3. Date of decline of chlorophyll and chlorophyll a/b ratio for
(a) control, (b) warmed-soil and (c) long-day seedlings. Each point is
the mean of three measurements. Dates of decline were determined by
a piecewise linear regression estimated for each measurement day (see
text and Figure 5). We defined the non-senescing (green) period by
constant Chl levels, early senescence by declining Chl contents but
constant Chl a/b ratio, and late senescence by declining Chl contents
and Chl a/b ratio.

Figure 4. Date of decline of carotenoid and Chl/carotenoid ratio for (a)
control, (b) warmed-soil and (c) long-day seedlings. Each point is the
mean of three measurements. Dates of decline were determined by a
piecewise linear regression estimated for each measurement day (see
text and Figure 5). See Figure 3 for definition of green, early and late
senescence.

Both Chl a and Chl b contents began to decline at approximately the same time as total chlorophyll content. The decline
in total chlorophyll content in control, warmed-soil and longday seedlings began on Julian days 285, 295 and 317, respectively. The corresponding Julian days for the decline in Chl a
content were 285, 298 and 317, and for the decline in Chl b
content they were 275, 298 and 317, respectively. Two findings
need to be considered when interpreting these results. First, the
start of the decline in Chl a content and total chlorophyll
content differed, as indicated by the pronounced V-shaped
pattern to the plot of sum-of-squared residuals from the various
piecewise linear regressions. Second, the sum-of-squared residuals for Chl b content of control seedlings showed little
variation between Julian days 264 and 285. Thus, although the
global minimum in sum-of-squared residuals for Chl b content
occurred on Julian day 275, that value differed little from the
sum-of-squared residuals associated with Julian day 264 or
285. Accordingly, we see little evidence to suggest that Chl a,
Chl b or total chlorophyll start to decline on different dates.
The Chl a/b ratio started to decrease 3 weeks after the onset
of the decline in total chlorophyll content in control (Julian day
306) and warmed-soil seedlings (Julian day 314; Figure 3).
Based on the different onset dates of decline for chlorophyll

Table 1. Chlorophyll content and chlorophyll a/b ratio in Larix occidentalis needles at different stages of needle senescence (see Figure 3
for definition of green, early and late senescence). Values are means ±
95% confidence interval (sample size in parenthesis).
Green

Early
senescence1

Late
senescence2

Chlorophyll (mg m −2)
Control
276 ± 26 (43)
Warmed soil3 256 ± 16 (58)
Long day3
294 ± 18 (86)

205 ± 24 (24)
185 ± 25 (25)

68 ± 17 (12)
56 ± 28 (4)
119 ± 64 (14)

Chlorophyll a/b ratio
Control
3.17 ± 0.18 (43)
Warmed soil3 3.24 ± 0.11 (58)
Long day3
3.26 ± 0.09 (88)

3.1 ± 0.19 (29)
3.3 ± 0.29 (27)

2.19 ± 0.35 (15)
1.62 ± 0.78 (6)
2.06 ± 0.48 (15)

1

2
3

Chlorophyll content is different (P < 0.05) from that of green tissue,
whereas Chl a/b ratio is not different (P > 0.25) from that of green
tissue.
Both Chl content and Chl a/b ratio are significantly different (P <
0.05) from values for green tissue.
Warmed-soil and long-day treatment groups are not significantly
different (P > 0.10) from control seedlings.

TREE PHYSIOLOGY VOLUME 17, 1997

FOLIAR SENESCENCE IN WESTERN LARCH

771

Table 2. Carotenoid (xanthophylls (x) + carotenoids (c)) and chlorophyll/carotenoid ratio ((Chl a + Chl b)/(x + c)) in Larix occidentalis
needles at different stages of senescence (see Figure 3 for definition of
green, early and late senescence). Values are means ± 95% confidence
interval (sample size in parenthesis).
Green
Carotenoid (mg m −2)
Control
96.90 ± 10.66
(43)
Warmed soil3
86.01 ± 6.28
(58)
Long day3
100.86 ± 5.46
(88)
Chlorophyll/carotenoid ratio
Control
3.04 ± 0.41
(43)
Warmed soil3
3.22 ± 0.41
(58)
Long day3
2.90 ± 0.14
(85)
Figure 5. Determination of date of decline for chlorophyll contents,
CO2 assimilation (see Figure 6) and chlorophyll a/b ratio for control
seedlings based on piecewise linear regressions. Sum-of-squared residuals are plotted from separate piecewise linear regressions where
each regression corresponded to a different break point (see text). The
date associated with the minimum sum-of-squared residuals (arrows)
was chosen as the best estimate of the break point.

content and Chl a/b ratio, we defined early senescence as the
period when the chlorophyll content declined but the Chl a/b
ratio did not change (Figure 3). This stage occurred between
Julian days 285 and 306 (October 11--November 1) in control
seedlings and between Julian days 295 and 314 (October 21-November 9) in seedlings in the warmed soil treatment,
whereas in seedlings receiving an extended photoperiod, early
senescence did not begin until Julian day 317 (November 12).
Late senescence was defined as the time after the Chl a/b ratio
started to decline (Figure 3). Our definitions of early and late
senescence do not describe senescence in all tree species
(Bortlik et al. 1987, Adams et al. 1990, Dean et al. 1993) but
are useful in the interpretation of our data.
Dates for the start of the decline in total needle carotenoids
(Figure 4), which were the same as those for total chlorophyll,
were Julian days 285, 295 and 317 for control, warmed-soil
and long-day seedlings, respectively. Carotenoid content declined about 50% during autumn, from a mean over all seedlings of 94 mg m −2 in green tissue to 46 mg m −2 (Table 2) in
late senescing tissue. The Chl/carotenoid ratio also fell by
about 50% during the sampling period (Table 2), with the onset
of decline occurring on Julian days 298, 310 and 322, for
control, warmed-soil and long-day seedlings, respectively.

1
2

3

Early
senescence1

Late
senescence2

74.06 ± 5.43
(29)
65.90 ± 7.60
(27)

42.67 ± 5.83
(15)
38.67 ± 7.88
(6)
55.20 ± 16.20
(15)

2.76 ± 0.27
(29)
2.83 ± 0.32
(24)

1.53 ± 0.27
(12)
1.30 ± 0.39
(4)
1.84 ± 0.56
(14)

Carotenoid contents are less than (P < 0.001) those from green tissue
and Chl/Car ratios are not different (P > 0.10) from green tissue.
Both carotenoid contents and Chl/Car ratios are significantly different (P < 0.001) from green tissue.
Carotenoid contents for warmed-soil and long-day treatment groups
are not significantly different (P > 0.05) from control seedlings and
Chl/carotenoid ratios are not significantly different (P > 0.10) than
control seedlings.

Figure 6. Carbon dioxide assimilation and Fv/Fm for seedlings under
(a) control outdoor conditions or (b) with warmed soil. Each point is
the mean of three measurements. Arrows show the dates of decline in
Chl and Chl a/b from Figure 3.

Gas exchange
Carbon dioxide assimilation per unit area began to decline
approximately 1 week after the onset of the decline in chlorophyll content (Figure 6). The decline in photosynthetic rate

began on Julian day 296 in control seedlings (Figure 6a), Julian
day 301 in seedlings in the warmed-soil treatment (Figure 6b),
and Julian day 317 in long-day seedlings. Mean stomatal

TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com

772

ROSENTHAL AND CAMM

Figure 7. Stomatal conductance (mmol m −2 s −1) in naturally senescing seedlings (r) on Julian day 303 (October 29), and in warmed-soil
(s) and long-day (q) seedlings on Julian day 305. Each point is the
mean of three measurements.

Figure 8. Oxygen evolution per mg Chl for control seedlings. Each
point is the mean of three measurements. Line drawn using LOWESS
smoothing function in Systat (Wilkinson 1990).

conductance (Figure 7) values for all seedlings decreased from
300 mmol m −2 s −1 when needles were green to less than
50 mmol m −2 s −1 on Julian day 303 (October 29). The similar
onset date of decline for all seedlings suggests that the date of
decline in stomatal conductance was not affected by photoperiod or a small increase in soil temperature. We observed no
significant increase in the ratio of intercellular to ambient CO2
concentration in late compared to early autumn needles (Table 3).
Oxygen evolution per mg Chl increased during early senescence and then declined during late senescence (Figure 8). The
light-saturated rate of oxygen evolution was lower in senescing
tissue than in green tissue (Figure 9). Dark respiration was
constant during autumn and had a mean of 0.5 µmol O2 m −2 s −1
when needles were pre-illuminated at 100 µmol m −2 s −1 and
then dark-adapted for 10 min (data not shown). When dark
respiration was measured immediately after illumination at
470 µmol m −2 s −1, the mean rate was 1.5 µmol O2 m −2 s −1,
which is consistent with reports (Walker 1992) that respiration
is greater after exposure to high irradiances.
Table 3. Internal/ambient CO2 concentration in Larix occidentalis
needles at different stages of senescence (see Figure 3 for definition of
green, early and late senescence). Ratios are means ± 95% confidence
interval (sample size in parenthesis).

Control
Warmed soil3
Long day3
1

2
3

Green

Early
senescence1

Late
senescence2

0.847 ± 0.032
(42)
0.849 ± 0.024
(55)
0.846 ± 0.014
(85)

0.868 ± 0.010
(26)
0.856 ± 0.008
(27)

0.892 ± 0.032
(15)
0.875 ± 0.093
(6)
0.883 ± 0.039
(14)

Early senescence needles not significantly different (P > 0.10) from
green needles.
Late senescence needles not significantly different (P > 0.05) from
green needles.
Warmed-soil and long-day treatment groups not significantly different (P > 0.10) from control needles.

Figure 9. (a) Oxygen evolution per leaf area (20 °C, 50 ml l −1 CO2,
660 nm light) and (b) photochemical quenching (qQ) of naturally
senescing leaves in different stages: green (filled circle), early senescence (shaded circle) and late senescence (open circle). Senescence
stages defined in Figure 3 and in text. Data were similar for seedlings
in warmed soil or extended photoperiod. Error bars are 95% confidence intervals; sample size per light intensity was 43 for green tissue,
29 for early senescing tissue and 15 for late senescing tissue.

Chlorophyll a fluorescence
We measured photochemical efficiency of photosystem II
(PSII) (Fv/Fm) in green and senescing tissues to determine if
there was a correlation between a decline in PSII and declining
CO2 assimilation rates. Photochemical efficiency of dark
adapted green tissue (Fv/Fm) was 0.773 ± 0.017 SD (n = 46),
0.779 ± 0.020 SD (n = 62) and 0.787 ± 0.023 SD (n = 99) in

TREE PHYSIOLOGY VOLUME 17, 1997

FOLIAR SENESCENCE IN WESTERN LARCH

control, warmed-soil and long-day needles, respectively, and
declined to 0.5 in senescent needles (Figure 6a). Photochemical efficiency (Fv/Fm) began to decline after the decline in CO2
assimilation rate in all seedlings. The decline in Fv/Fm began
at the same time in control and warmed-soil seedlings (Julian
days 308 and day 306, respectively, Figures 6a and 6b) but was
delayed in long-day seedlings (Julian day 322).
Photochemical quenching (qQ, Kooten and Snell 1990) indicates the proportion of oxidized QA, the quinone acceptor of
photosystem II. To plot qQ as a function of light intensity, data
were grouped into green, early senescing and late senescing
phases as defined by chlorophyll loss (Figure 3). For all seedlings, photochemical quenching was less in late senescing
tissue than in green tissue at irradiances greater than 25 µmol
m −2 s −1 (Figure 9b).

Discussion
The sequence of events during autumn foliar senescence in
western larch seedlings was similar to that in other deciduous
species (Sanger 1971, Lichtenthaler 1987, Adams et al. 1990)
in that chloroplasts remained efficient until the late stages of
senescence. In addition, the order of events----a decline in
chlorophyll content followed by declines in CO2 assimilation
rate, and, finally, Chl a/b ratio----is the same as that found in
western larch needles senescing in controlled environment
chambers at autumn temperatures and with a short photoperiod
(Rosenthal and Camm 1996). The similarity suggests that this
pattern of needle degradation is typical of a variety of autumnlike conditions. In the present study, the conditions at the time
chlorophyll content began to decline included: a decrease in
mean air temperature from 23 to 12 °C (Figure 1a), a decline
in irradiance to below 350 µmol m −2 s −1 (Figure 2a), a day
length of less than 11 h (Figure 2b), and a relative humidity
above 85% (Figure 2c).
The first measured parameter to decline was needle pigment
content (chlorophyll and carotenoids). During the early phase
of senescence when the amount of chlorophyll per leaf area
declined, chloroplast function did not decline. The stable Chl
a/b ratio during early senescence (Figure 3) suggests that Chl a
and Chl a/b-containing complexes were degraded at the same
rate. The constant Fv/Fm ratio indicates that the thylakoids
sampled by the fluorescence technique (i.e., those near the leaf
surface) remained functional. Furthermore, photosynthetic
rate per mg Chl increased during early senescence (Figure 8).
During late senescence, there was loss in function of nearly
all of the photosynthetic components. A decline in the Chl a/b
ratio and the Chl/carotenoid ratio ((Chl a + Chl b)/(xanthophyll + β-carotene)) suggested preferential loss of Chl a-containing proteins closely associated with the reaction centers
over loss of light harvesting proteins (Lichtenthaler 1987). The
subsequent decline in qQ, which was not apparent under light
limitation (Figure 9b), is consistent with a loss in PSII reaction
centers or a decrease in carboxylation, or both.
The ratio of internal to ambient CO2 concentration did not
increase significantly (Table 3) during late senescence suggesting that changes in photosynthetic function and carbon fixation

773

compensated for changes in stomatal conductance in yellow
leaves. A constant ratio of internal to ambient CO2 concentration during senescence has been reported in some species (e.g.,
Woolhouse and Jenkins 1983) but not in others (e.g., Makino
1984, Thomas 1991). The variations in stomatal conductance
during senescence that have been observed in some species
may be associated with stomatal patchiness (Thomas 1991)
and the history of the leaf (i.e., the age of the leaf or the
conditions inducing senescence, or both). In our study, the late
senescent tissue contained on average between 50 and 120 mg
Chl m −2 (Table 1) and had positive photosynthetic rates (Figure 6). Thus, these needles may contain less senescent tissue
than needles for which an increase in internal to ambient CO2
concentrations is reported.
The warmed-soil treatment delayed the onset of needle
senescence but, once senescence began, it progressed more
rapidly in warmed-soil seedlings than in control seedlings. For
example, the 3 °C increase in soil temperature (Figure 1b)
delayed needle yellowing by 10 days compared with control
seedlings (Figure 3). Other pigment-related parameters (initiation of decline in carotenoids, and in Chl/carotenoids and Chl
a/b ratios) were delayed by approximately the same period
(8--14 days). However, other processes were not delayed to the
same extent. For example, carbon assimilation and Fv/Fm began to decline in warmed-soil seedlings only slightly (2-5 days) later than in naturally senescing seedlings. As a result,
the interval between the onset of the decline in pigment content
and the onset of the decline in CO2 assimilation rate was
shorter for seedlings in warmed soil than for naturally senescing seedlings. Needle senescence in western larch seedlings in
controlled environment chambers showed a similar pattern:
CO2 assimilation rate began to decline approximately 1 week
after the onset of the decline in chlorophyll content in seedlings maintained at an air temperature of 15 °C, but the interval
was approximately 2 weeks when seedlings were maintained
at 8 °C (Rosenthal and Camm 1996).
The extended photoperiod treatment delayed the onset of
decline of most parameters. Delayed senescence has been
observed in deciduous trees growing near street lights (Olmsted 1951), though previous studies have not determined
whether the delay was the result of photoperiod extension or a
higher air temperature, or both. We have shown that extended
photoperiod alone can delay needle yellowing in larch. Furthermore, we have demonstrated that, during this extended
green period, the needles are functional, in contrast to some
stay-green mutants in which assimilation rates decline even
though the chlorophyll content is maintained (Smart 1994).
Rosenthal and Camm (1996) observed that senescing needles
of western larch maintained in a controlled environment
growth chamber set to a 16-h photoperiod and a temperature
of 8 °C showed a similar pattern of decline in CO2 assimilation
rate relative to the decline in chlorophyll content. Similarly, we
found that chlorophyll content and CO2 assimilation rate declined at the same time in autumn senescing needles of seedlings exposed to a 16-h photoperiod and outdoor temperatures.
The onset of a decline in stomatal conductance was not
affected by our treatments. It is possible that conditions not

TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com

774

ROSENTHAL AND CAMM

controlled for in this experiment might have triggered the
decline. For example, stomatal conductance in larch has been
shown to be correlated with irradiance (Benecke et al. 1981,
Matyssek and Schulze 1988) and to leaf-to-air vapor pressure
difference (Benecke et al. 1981, Sandford and Jarvis 1986,
Matyssek and Schulze 1988). Although a drop in soil temperature has been shown to lower stomatal conductance in other
conifers (e.g., Day et al. 1991), the 3 °C difference between the
soil temperature in our treatments was not large enough to
affect the onset of decline in stomatal conductance.
Our photoperiod treatment did not affect the onset of decline
in stomatal conductance. The extended photoperiod delayed
the onset of decline in assimilation rate and chlorophyll loss,
and thus might have been expected to also delay stomatal
conductance. However, the large changes in photosynthetic
photon flux density, air temperature and relative humidity
appear to have had a greater influence on the timing of stomatal
conductance changes than changes in assimilation rates and
Chl content.
We were able to study selected senescence-related events in
larch needles, even though such events are among the physiological and developmental changes experienced by the whole
plant. Development of the abscission layer (Neger and Fuchs
1915) and bud development induced by shortened days (Meijer and van der Veen 1957, Romberger 1963) may affect
translocation and assimilate partitioning. It is possible that our
environmental manipulations affected abscission layer development and bud formation thereby influencing the timing of
needle senescence indirectly as well as directly.
We were able to quantify the effects of changes in soil
temperature and photoperiod on the onset of different degradative events associated with foliar senescence in western larch.
Further, we found that because degradative events are differently affected by the changing environment, the course of
breakdown of the photosynthetic apparatus during senescence
can vary. We have demonstrated that the response of the plant
to yearly variation in environmental conditions can be balanced between assimilation and autumn shutdown.
Acknowledgments
This research was in part funded by a grant from the Natural Sciences
and Engineering Research Council of Canada and a GREAT award
from the Science Council of British Columbia and Weyerhaeuser
Canada. We also thank Drs. B. Green and I. Taylor for helpful discussions during preparation of this manuscript.
References
Adams, W.W., K. Winter, U. Schreiber and P. Schramel. 1990. Photosynthesis and chlorophyll fluorescence characteristics in relationship to changes in pigment and element composition of leaves of
Platanus occidentalis L. during autumnal senescence. Plant
Physiol. 93:1184--1190.
Addicott, F.T. and J.L. Lyon. 1973. Physiological ecology of abscission. In Shedding of Plant Parts. Ed. T.T. Kozlowski. Academic
Press, New York, pp 85--124.
Benecke, U., E.-D. Schulze, R. Matyssek and W.M. Havranek. 1981.
Environmental control of CO2-assimilation and leaf conductance in
Larix decidua Mill. Oecologia 50:54--61.

Bortlik, K., H. Gut and P. Matile. 1987. Yellowing and non-yellowing
trees: a comparison of protein and chlorophyll loss in senescent
leaves. Bot. Helv. 97:323--328.
Cunninghame, M.E., J.R. Hillman and B.G. Bowes. 1982. Ultrastructural changes in mesophyll cells of Larix decidua × kaempferi
during leaf maturation and senescence. Flora 172:161--172.
Day, T.A., S.A. Heckathorn and E.H. DeLucia. 1991. Limitation of
photosynthesis in Pinus taeda L. (loblolly pine) at low soil temperatures. Plant Physiol. 96:1246--1254.
Dean, M.A., C.A. Letner and J.H. Eley. 1993. Effect of autumn foliar
senescence on chlorophyll a:b ratio and respiratory enzymes of
Populus tremuloides. Bull. Torrey Bot. Club 120:269--274.
Goodwin, T.W. 1958. Studies in carotenogenesis. The changes in
carotenoid and chlorophyll pigments in the leaves of deciduous
trees during autumn necrosis. Biochem. J. 68:503--511.
Kooten, O. and J.F. Snell. 1990. The use of chlorophyll fluorescence
nomenclature in plant stress physiology. Photosynth. Res. 25:147-150.
Kozlowski, T.T., P.J. Kramer and S.G. Pallardy. 1991. Extent and
significance of shedding of plant parts. In The Physiological Ecology of Woody Plants. Ed. T.T. Kozlowski. Academic Press, New
York, pp 1--34.
Lang, M. and H.K. Lichtenthaler. 1991. Changes in the blue-green and
red fluorescence-emission spectra of beech leaves during the
autumnal chlorophyll breakdown. J. Plant Physiol. 138:550--553.
Larcher, W. and G. Neuner. 1989. Cold induced sudden reversible
lowering of in vivo chlorophyll fluorescence after saturating light
pulses. Plant Physiol. 89:740--742.
Lichtenthaler, H.K. 1987. Chlorophyll fluorescence signatures of
leaves during the autumnal chlorophyll breakdown. J. Plant
Physiol. 131:101--110.
Lichtenthaler, H.K. and A.R. Wellburn. 1983. Determination of total
carotenoids and chlorophylls a and b of leaf extracts in different
solvents. Biochem. Soc. Trans. 11:591--592.
Makino, A., T. Mae and K. Ohira. 1984. Changes in photosynthetic
capacity in rice leaves from emergence through senescence. Analysis from ribulose-1,5-bisphosphate carboxylase and leaf conductance. Plant Cell Physiol. 25:511--521.
Markgraf, T. and J.A. Berry. 1990. Measurement of photochemical
and non-photochemical quenching: correction for turnover of PS2
during steady-state photosynthesis. In Current Research in Photosynthesis. Vol. IV. Ed. M. Baltscheffsky. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 270--282.
Matyssek, R. and E.-D. Schulze. 1988. Carbon uptake and respiration
in above-ground parts of a Larix decidua × leptolepis tree. Trees
2:233--241.
Matyssek, R., M.S. Günthardt-Goerg, T. Keller and C. Scheidegger.
1991. Impairment of gas exchange and structure in birch leaves
(Betula pendula) caused by low ozone concentrations. Trees 5:5-13.
Meijer, G. and R. van der Veen. 1957. Wavelength dependence on
photoperiodic responses. Acta Bot. Neerl. 6:429--433.
Neger, F.W. and J. Fuchs. 1915. Untersuchungen über den Nadelfall
der Koniferen. Jahrb. Wissensch. Bot. 55:608--660.
Neter, J., W. Wasserman and M. Kutner. 1990. Applied linear statistical models, 3rd Edn. Richard D. Irwin, Inc., Homewood, IL,
pp 370--374.
Olmsted, C.E. 1951. Experiments on photoperiodism, dormancy and
leaf age and abscission in sugar maple. Bot. Gaz. 112:365--393.
Osborne, D.J. 1973. Internal factors regulating abscission. In Shedding of Plant Parts. Ed. T.T. Kozlowski. Academic Press, New York,
pp 125--147.

TREE PHYSIOLOGY VOLUME 17, 1997

FOLIAR SENESCENCE IN WESTERN LARCH
Pell, E.J. and M.S. Dann. 1991. Multiple stress-induced foliar senescence and implications for whole plant longevity. In Response of
Plants to Multiple Stresses. Eds. H.A. Mooney, W.E. Winner and
E.J. Pell. Academic Press, San Diego, CA, pp 189--204.
Richards, J.H. and L.C. Bliss. 1985. Winter water relations of a
deciduous timberline conifer, Larix lyalli. Oecologia 69:16--24.
Romberger, J.A. 1963. Meristems, growth and development in woody
plants. USDA Forest Service, Tech. Bull. No. 1293, Washington,
D.C., pp 46--47.
Rosenthal, S.I. and E.L. Camm. 1996. Effects of air temperature,
photoperiod, and leaf age on foliar senescence of western larch
(Larix occidentalis Nutt.) in environmentally controlled chambers.
Plant Cell Environ. 19:1057--1065.
Sandford, A.P. and P.G. Jarvis. 1986. Stomatal responses to humidity
in selected conifers. Tree Physiol. 2:89--103.
Sanger, J.E. 1971. Quantitative investigation of leaf pigments from
their inception in buds through autumn coloration to decomposition
in falling leaves. Ecology 52:1075--1089.
Smart, C.M. 1994. Tansley Review No. 64. Gene expression during
leaf senescence. New Phytol. 126:419--448.

775

Thomas, C., S.D. Davis and G. Tallman. 1991. Responses of stomata
of senescing and nonsenescing leaves of Nicotiana glauca to
changes in intercellular concentrations of leaf carbon dioxide. Plant
Cell Environ. 14:971--978.
Vance, N.C. and S.W. Running. 1985. Light reduction and moisture
stress: effects on growth and water relations of western larch seedlings. Can. J. For. Res. 15:72--77.
Walker, D.A. 1987. The use of the oxygen electrode and fluorescence
probes in simple measurements of photosynthesis. Oxygraphics
Ltd., University of Sheffield, Kings Lynn, U.K., pp 1--16.
Walker, D.A. 1992. Excited leaves. New Phytol. 121:325--345.
Wilkinson, L. 1990. Systat: The system for statistics. Systat, Inc.,
Evanston, IL, pp 304--305.
Woolhouse, H.W. and G.I. Jenkins. 1983. Physiological responses,
metabolic changes and regulation during leaf senescence. In
Growth and Functioning of Leaves. Eds. J.E. Dale and F.L.
Milthorpe. Cambridge, NY, pp 449--487.
Worrall, J. 1993. Temperature effects on bud-burst and leaf-fall in
subalpine larch. J. Sustainable For. 1:1--18.

TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com