Effects of water stress on irradiance acclimation of leaf traits in almond trees

14 Pages·2012·1.11 MB·English

Tree Physiology 32, 450–463
doi:10.1093/treephys/tps016
Research paper
Effects of water stress on irradiance acclimation of leaf traits
in almond trees
D
o
w
n
Gregorio Egea1,2, María M. González-Real1,4, Alain Baille1, Pedro A. Nortes1,3, María R. Conesa1 lo
a
d
and Isabel Ruiz-Salleres1 ed
fro
m
1Área de Ingeniería Agroforestal, Universidad Politécnica de Cartagena, Paseo Alfonso XIII, 48, 30203 Cartagena, Spain; 2Present address: Área de Ingeniería Agroforestal, http
Universidad de Sevilla, Ctra. de Utrera, km 1, 41013 Seville, Spain; 3Present address: Departamento de Riego, Centro de Edafología y Biología Aplicada del Segura (CEBAS- s
CSIC), PO Box 164, E-30100 Espinardo, Murcia, Spain; 4Corresponding author ([email protected]) ://a
c
a
d
Received October 30, 2011; accepted February 9, 2012; published online March 22, 2012; handling Editor Ülo Niinemets e
m
ic
.o
u
Photosynthetic acclimation to highly variable local irradiance within the tree crown plays a primary role in determining tree p
.c
carbon uptake. This study explores the plasticity of leaf structural and physiological traits in response to the interactive o
m
effects of ontogeny, water stress and irradiance in adult almond trees that have been subjected to three water regimes (full /tre
e
irrigation, deficit irrigation and rain-fed) for a 3-year period (2006–08) in a semiarid climate. Leaf structural (dry mass per p
h
y
unit area, N and chlorophyll content) and photosynthetic (maximum net CO assimilation, A , maximum stomatal conduc- s
2 max /a
tance, gs,max, and mesophyll conductance, gm) traits and stem-to-leaf hydraulic conductance (Ks-l) were determined through- rtic
out the 2008 growing season in leaves of outer south-facing (S-leaves) and inner northwest-facing (NW-leaves) shoots. Leaf le-a
b
plasticity was quantified by means of an exposure adjustment coefficient (ε = 1-X /X ) for each trait (X) of S- and NW-leaves. s
NW S tra
Photosynthetic traits and K exhibited higher irradiance-elicited plasticity (higher ε) than structural traits in all treatments, c
s-l t/3
with the highest and lowest plasticity being observed in the fully irrigated and rain-fed trees, respectively. Our results sug- 2
/4
gest that water stress modulates the irradiance-elicited plasticity of almond leaves through changes in crown architecture. /4
5
0
Such changes lead to a more even distribution of within-crown irradiance, and hence of the photosynthetic capacity, as water /1
6
stress intensifies. Ontogeny drove seasonal changes only in the ε of area- and mass-based N content and mass-based chlo- 3
7
3
rophyll content, while no leaf age-dependent effect was observed on ε as regards the physiological traits. Our results also 6
1
indicate that the irradiance-elicited plasticity of Amax is mainly driven by changes in leaf dry mass per unit area, in gm and, by g
most likely, in the partitioning of the leaf N content. u
e
s
t o
Keywords: irradiance, leaf structural traits, mesophyll conductance, nitrogen-use efficiency, photosynthetic capacity, n 1
0
plasticity, Prunus dulcis, stem-to-leaf hydraulic conductance, stomatal conductance, water stress. A
p
ril 2
0
1
9
Introduction
2010) or water stress (Reich et al. 1989, Egea et al. 2011a).
Leaf phenology, structure and photosynthetic function define Fruit tree species also exhibit large seasonal variations of leaf
how much carbon a plant fixes during the growing season and, structural and photosynthetic traits, as observed in oaks
consequently, any change in either of these factors will sub- (Wilson et al. 2000) and in almond trees (Nortes et al. 2009).
stantially affect carbon gain. These factors are known to dis- Knowledge of the role that ontogeny, irradiance, water stress
play a wide degree of variability (i.e., plasticity) in response to and their interactions play in modulating leaf structure and
abiotic variables such as the amount of incident irradiance physiological function is therefore important for assessing the
(Valladares et al. 2000, Robakowski et al. 2003, Coste et al. productivity of fruit tree orchards (Mirás-Ávalos et al. 2011)
© The Author 2012. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Irradiance plasticity of almond leaf traits 451
and for characterizing the impact of agronomic practices (e.g., and architecture. In irrigated orchards, the amount of water
pruning strategy, deficit irrigation) or climate change scenarios supplied has a pronounced effect on almond crown leaf area
(IPCC 2007) on canopy carbon acquisition. and volume (Fereres et al. 1981, Klein et al. 2001, Egea et al.
A large body of work has established that the photosynthetic 2010) and therefore on light interception (Castel and Fereres
capacity of leaves acclimates to growth irradiance through 1982, Klein et al. 2001) and on crop yield (Hutmacher et al.
changes in structural and physiological traits, including, leaf 1994). Deficit irrigation (DI) applied during three consecutive
dry mass per unit area (M), leaf chlorophyll and nitrogen con- years (irrigating at 50% of crop water requirements) has been
a
tent and photosynthetic capacity (Rijkers et al. 2000, Valladares reported to reduce by ~35% the increase in crown volume
et al. 2000, Le Roux et al. 2001, Robakowski et al. 2003, over the same period with respect to fully irrigated (FI) trees,
Aranda et al. 2005, Warren et al. 2007, Coste et al. 2009, whereas the reduction is expected to be twice as large under
2010). The influence of within-crown irradiance gradients on rain-fed conditions (Egea et al. 2010). These water- stress-
leaf structural and photosynthetic traits has been described driven variations in crown leaf area and volume may affect the D
o
w
previously in several fruit tree species such as peach (Rosati spatial distribution of irradiance within the tree crown n
lo
et al. 1999, Le Roux et al. 2001, Walcroft et al. 2002) and (Lampinen et al. 2004) which, in turn, may modify the a
d
e
walnut trees (Le Roux et al. 1999, Frak et al. 2002). However, irradiance-elicited plasticity of almond leaf traits compared d
the mechanisms of photosynthetic light acclimation do not with FI trees. An important result reported in P. persica is that fro
m
seem to be universal, as they differed in these two species. the relationship between the absorbed quantum irradiance h
ttp
While acclimation resulted primarily from changes in M in wal- (PPFD ) and M is not linear (Le Roux et al. 2001), due to the s
nut trees (Le Roux et al. 1999), in peach trees accalimation insensiativity of aM to PPFD above a threshold value (Walcroft ://ac
a a a
d
resulted mainly from changes in M and leaf N partitioning (Le et al. 2002). This raises the question as to whether or not e
a m
Roux et al. 2001). water-stress-driven changes in crown architecture, and hence ic
.o
Water stress is known to influence the potential for photo- in within-crown irradiance distribution, are sufficient to drive u
p
synthesis acclimation to irradiance (Osmond 1983, Craven differences in irradiance-elicited acclimation of almond leaf .co
m
et al. 2010). The impact of irradiance × water stress interac- traits among irrigation regimes. To our knowledge, no study on /tre
tions on leaf traits has been assessed in studies of several this type of water stress × irradiance interaction has been ep
h
tree species. Quero et al. (2006) observed that the impact of reported for almond trees. Acquiring more knowledge on leaf ys
/a
drought on the photosynthetic rate was higher in two decidu- trait plasticity to both water stress and local irradiance is rtic
ous Quercus species grown under high irradiance, whereas important for understanding the adaptive mechanisms devel- le
-a
the influence of shade on the photosynthetic rate was more oped by drought-tolerant species in response to prolonged bs
pronounced when the water supply was not restricted. Guidi drought or to irrigation. It is also important for quantifying the trac
et al. (2008) also observed a higher impact of water stress on impact of water stress and local irradiance on canopy carbon t/32
/4
the photosynthetic rate and on stomatal conductance in sunlit acquisition. /4
5
than in shaded plants of Ligustrum vulgare L. Niinemets et al. Based on the above, the first issue of the study was to 0
/1
6
(1999) found that stomatal conductance under conditions of assess to what extent leaf structural and physiological traits of 3
7
severe water stress was lower in the upper (sunlit) than in the a drought-tolerant species like the almond tree acclimate to 36
1
lower (shaded) canopy in plants of Populus tremula L. and Tilia within-crown irradiance heterogeneity, with special focus on b
y
cordata Mill. In Prunus persica L. Batsch, Neri et al. (2003) the mechanisms involved in photosynthetic light acclimation. gu
e
reported that irradiance played an important role in the com- The second issue was to evaluate how water stress, through st o
petition for water among sun and shade leaves and that this changes in crown architecture and interactions with irradiance n 1
0
led to a reduction in the adaptive response of shade leaves to level and leaf ontogeny, modulate irradiance-elicited acclima- A
p
water stress. tion of almond leaf traits. Both issues are central to elucidating ril 2
The issue addressed in this paper deals with the plastic the role of physiological adjustments made by drought-tolerant 0
1
9
response of leaf traits to within-crown irradiance variability tree species to cope with water scarcity and optimize the
under several water stress regimes of a drought-tolerant fruit photo synthetic use of leaf resources across large gradients of
tree species, the almond tree (Prunus dulcis (Mill.) D. A. Webb). water availability and light. To this end, we characterized and
Almond trees survive in arid areas due to their tolerance to soil analyzed during the last year of a long-term experiment the
water restriction through osmotic adjustments, stomatal regu- seasonal trend of leaf structural and physiological attributes of
lation, leaf rolling and shedding, and increases in rooting depth outer south-facing (sun) and inner northwest-facing (shade)
(Castel and Fereres 1982, Catlin 1996, Torrecillas et al. 1996, fruit-bearing shoots (hereafter named S- and NW-leaves) in
Goldhamer and Viveros 2000, Lampinen et al. 2004, Matos almond tree crowns. The trees were subjected to three con-
et al. 2004). They also respond strongly to any supplementary trasting water regimes: full irrigation (FI), sustained deficit irri-
supply of water (irrigation), with drastic changes in tree volume gation (DI) and rain-fed (NI).
Tree Physiology Online at http://www.treephys.oxfordjournals.org
452 Egea et al.
Material and methods during the KF stage, when shoot and leaf growth had practi-
cally stopped (Egea et al. 2010). In order to reduce experimen-
Site description
tal errors, measurements were carried out early in the morning
The study was conducted in a 9-year-old almond orchard when diffuse radiation was dominant and the scattering of
(Prunus dulcis (Mill.) D. A. Webb cv. Marta, grafted on Mayor direct sunlight from the foliage was negligible. The measure-
rootstock), planted at a spacing of 6 m × 7 m and located at ment protocol for single trees is described in the LAI-2000
the Agricultural Experimental Station of the University of instruction manual (Li-Cor, 1992). The method requires deter-
Cartagena (37°35′N, 0°59′W). A schematic representation of mining path lengths and canopy volume, for which the FV2000
the stages of development of almond trees at the same loca- software (Li-Cor) was used. As the LAI-2000 sensor does not
tion can be found in Nortes et al. (2009). The measurement discriminate among foliar and non-foliar (i.e., branches) ele-
period covered the stages of rapid vegetative growth (VG, ments, the calibration equation derived by Egea (2008) for
from early April to early June) and kernel-filling (KF, from June this species was used to correct the LAI-2000 estimates of L . D
C o
w
to harvest at the end of August). The climate at the site is Measurements were performed on two leaf classes: leaves n
lo
Mediterranean, characterized by warm and dry summers and located at the south periphery of the tree crown (S-leaves) and a
d
e
mild winters. Mean values of annual temperature and precipita- leaves of northwest-facing inner shoots (NW-leaves) that were d
tion determined over 20 years at a nearby weather station are exposed to substantially lower irradiance than S-leaves. The fro
m
18.3 ± 0.9 °C and 355 ± 90 mm, respectively. In 2008, the incident photosynthetic photon flux density (PPFD) was mea- h
ttp
annual rainfall and reference crop evapotranspiration (ET), cal- sured by a lineal PPFD sensor (model EMS 7, PP-systems, s
culated using the FAO-Penman–Monteith equation (Alleno et al. Amesbury, MA, USA). Measurements were performed at mid- ://ac
a
d
1998), computed 446 and 1210 mm, respectively. The soil is day on four sunny days during the measurement period. e
m
deep, with a silt-clay-loam texture and available water-holding ic
.o
capacity of ~0.18 m m−1. Leaf structural traits u
p
In what follows, the subscripts m and a, respectively, refer to .co
m
Irrigation regimes mass- and area-based values. Leaf nitrogen content (N) and /tre
Three water treatments were applied for three consecutive relative chlorophyll content were measured fortnightly (from ep
h
years (2006–08): (i) FI (irrigated at 100% of standard crop April to August) at the two locations in fully expanded leaves ys
/a
evapotranspiration, ETc), (ii) sustained DI (irrigated at 50% ETc attached to fruit-bearing shoots. Samples were taken in three rtic
throughout the growing season) and (iii) NI treatment. They replicates per treatment and leaf class in the early morning and le
-a
frwoealplsoli wcacetader srai e pcdeo rm otrupetle attdemulyer inrntag na dntodhm e1 i2zlea tdsrte setgsar topiswetiric nargel pdlseicesaaigtseno. nwT hi(teh2 0stht0ur8ed)ey. idemarymc hem dsaiaasmtse pl(yleL wD. MLe) i gwahanesdd mtleoea aof sbuatrraeeinad tu(hLse)i n lgwe aeafr nef r eaasrleshao m mdaeesttsee rr(m L(FiLMnI)e-.3d L1 e0foa0rf bstract/32
/4
Irrigation water was supplied by means of a single drip line Leaf Area Meter, Li-Cor). /4
5
equipped with six pressure-compensated drippers (8 l h−1 per Leaf N content was determined after Kjeldahl digestion of 0
/1
6
dripper) per tree. The electrical conductivity of irrigation water leaves, which had been oven-dried at 80 °C for 48 h, weighed 3
7
was 1.2 dS m−1 with no salinity risk for this species. Trees were and carefully ground. For each leaf class, values of the leaf N 36
1
managed according to current commercial practices (i.e., a content were either reported relative to the total leaf area (N , b
a y
routine pesticide program was maintained, pruning was applied in gN m−2leaf) or as a proportion of leaf dry mass (Nm, in gN gue
manually in December and no weeds were allowed to grow in g−1DM). st o
the orchard). The annual amount of water applied to FI and DI The relative chlorophyll content was determined by a Minolta n 1
0
was 770 and 339 mm, respectively, for the 2008 growing sea- SPAD-505 chlorophyll meter (Spectrum Technologies, A
p
son and an average of 704 and 326 mm for the period Plainfield, IL, USA), provided with two LEDs emitting through a ril 2
2006–08. leaf sample at the red (~650 nm) and infrared (~940 nm) 0
1
9
wavelengths. The measurement in the red was associated with
Measurements
the absorption of chlorophyll, whereas that in the infrared was
Tree size used as a reference (Markwell et al. 1995). Three SPAD mea-
Measurements of trunk diameter and tree height were per- surements per leaf were taken between the base and the apex
formed in February 2008 by tape-measure and scaled poles, and subsequently averaged.
respectively. The relationship between SPAD readings and leaf chloro-
phyll content was derived from measurements taken in June
Crown leaf area and irradiance level and August. Leaf discs were collected from S- and NW-leaves
Crown leaf area (L ) was determined indirectly by means of an for which SPAD readings were previously recorded. After the
C
LAI-2000 Plant Canopy Analyzer (Li-Cor, Lincoln, NE, USA) leaf discs of each individual sample had been ground in 80%
Tree Physiology Volume 32, 2012
Irradiance plasticity of almond leaf traits 453
acetone, the total chlorophyll content on a fresh mass basis Stem-to-leaf hydraulic conductance
(Chl, mg g−1 ) was determined using a spectrophotometer Stem-to-leaf hydraulic conductance (K ) was estimated by the
FM s-l
(UV-1603 Shimadzu, Kyoto, Japan) at 652 nm, following the evaporative flux method (Brodribb and Holbrook 2003).
equation proposed by Hiscox and Israelstam (1979). SPAD Measurements were carried out during the KF stage on at least
readings and laboratory measurements of Chl fitted well for all two trees per replicate and on both S- and NW-leaf classes.
the treatments and leaf classes (C = 0.119 SPAD −1.95; Under steady-state conditions K was calculated according to
T s-l
R2 = 0.95). In the following, Chl is given either on an area basis Ohm’s law analogy:
(Chl , mg m−2 ) or on a leaf dry mass basis (Chl , mg g−1 ).
a leaf m DM
= E
Leaf gas exchange Ks-l ∆Ψ
Leaf gas exchange was measured fortnightly from April to
August on S- and NW-leaves selected following the criteria where E is the leaf evaporative flux (mmol m−2 s−1) and ΔΨ is D
o
w
described for leaf structural traits. At least eight leaves per the water potential drop (MPa) across the stem–leaf pathway. n
lo
treatment and leaf class were measured. Maximum net CO E was measured with the CIRAS2 gas exchange system at a
2 d
e
assimilation (A ) and maximum stomatal conductance ambient PPFD. ΔΨ was obtained as the difference between d
(gs,max) were mmeaaxsured at PPFD ≈ 1500 µmol m−2 s−1, near stem water potential (Ψs) and leaf water potential (Ψl), both from
constant ambient CO concentration (C ≈ 380 µmol mol−1) determined with a Scholander-type pressure bomb (Model h
2 a ttp
aenxcdh alnegafe steysmtepmer aCtIuRrAe S2(T (lePafP ≈ S 3y0st e°Cm)s , wHiitthc hian , Hpoerrttafobrldes hgiraes, 3tio0n0 0w,e Sreo ial lsMoo uissteudre f oErq Ψuip dmeetenrt)m. iTnhaeti olne.a vΨes w faosr mE edaestuerremdin bay- s://ac
l s a
d
UK). The desired PPFD was provided using an internal red/blue applying the method of bagged (non-transpiring) leaves, e
m
LED light source (PC 069-1). C was controlled using the whose water potential is presumed to be in equilibrium with ic
a .o
CIRAS2 injection system and compressed CO cylinders. The the xylem water potential of the shoot proximal to the petiole. u
2 p
preset value of T was controlled by the CIRAS2 system using The leaves used for Ψ determination, enclosed in plastic bags .co
leaf s m
an external power source (12 V DC). In the following, area and and covered with silver foil for at least 2 h prior to the mea- /tre
dry mass-based values of Amax are denoted as Amax,a surements, were placed next to the leaves used for E and Ψl eph
(µmol m−2leaf s−1) and Amax,m (µmol g−1DM s−1), respectively. determinations. ys/a
Maximum photosynthetic nitrogen-use efficiency, PNUEmax, rtic
and intrinsic water use efficiency, WUE, were computed as Exposure adjustment coefficient (ε) le
i -a
rAemsapx,ea/cNtiav ely(.µmol g−1N s−1) and Amax,a/gs,max (µmol mol−1), Twhaes rcehsaproancsteer iozef da bleya ft htera eitx Xp otsou rseh oaodtj uesxtpmoesnutr ec oteof fiircriaednita n(εc)e, bstrac
similar to the shade adjustment coefficient used to characterize t/32
/4
Mesophyll conductance the plastic response to shade (Laisk et al. 2005), calculated as /4
5
Mesophyll conductance to CO diffusion (g , mol m−2 s−1) 0
2 m leaf /1
6
was determined from simultaneous measurements of leaf gas 3
exchange and chlorophyll fluorescence using the portable gas ε =1− XNW / XS 736
1
exchange system CIRAS2 and a FMS2 pulse-modulated b
y
fluorometer (Hansatech Instruments Ltd, Norfolk, UK). where XS and XNW are the values measured on the south and gue
Measurements were made during the KF stage on at least eight northwest-facing shoots, respectively. According to this equa- st o
leaves per treatment and leaf type (i.e., S- and NW-leaves). The tion, ε is positive when XNW is smaller than XS and negative n 1
0
determination of gm followed the ‘constant J method’ (Bongi and when XNW is higher than XS. Ap
L(2o0re1t1oa )1.989, Harley et al. 1992) as described in Egea et al. Statistical analysis ril 20
1
9
Linear regressions between leaf structural and physiological
Tree water status traits were compared by analysis of covariance to evaluate
Midday stem water potential (Ψ ) was measured fortnightly significant differences between slopes and non-zero inter-
md
between April and August with a Scholander-type pressure cepts due to treatment and/or leaf class effects. Differences
bomb (Model 3000, Soil Moisture Equipment, Santa Barbara, between water regimes and shoot exposure to irradiance
CA, USA) following the recommendations by Hsiao (1990). were analyzed by two-way analysis of variance. Post hoc pair-
Measurements were performed on at least four leaves per rep- wise comparison between all means was performed by
licate. For Ψ determination, selected leaves near the trunk Duncan’s multiple range test. All the analyses were carried out
md
were wrapped in small black polyethylene bags and covered using Statgraphics software (Statgraphics Plus for Windows
with silver foil at least 2 h before measurement. Version 4.1).
Tree Physiology Online at http://www.treephys.oxfordjournals.org
454 Egea et al.
Results Seasonal trend of leaf structural traits
Tree growth, irradiance level and water status Leaf dry mass per unit area (Ma = leaf thickness × leaf den-
sity) showed a sustained increase throughout the growing
Mean trunk diameter at the beginning of the measurement season, irrespective of the treatment and leaf class (Figure 2a
period (March 2008) was 19.9 ± 0.38 cm in FI, 17.9 ± 0.4 cm and b). The upward trend of M was concomitant with a signifi-
a
in DI and 16.1 ± 0.18 cm in NI trees. At the same period, can- cant increase in the leaf fresh mass per unit area (up to 20%
opy height was 3.98 ± 0.04 m in FI, 3.18 ± 0.05 m in DI and in FI) and in leaf dry-to-fresh mass ratio L /L (20–25%)
DM FM
2.94 ± 0.13 m in NI, whereas crown leaf area, determined at (data not shown), the latter considered as a surrogate of leaf
the end of the vegetative growth stage (June 2008), was density (Niinemets et al. 2005). Likewise, L /L was little
DM FM
122.9 ± 7.1 m2 tree−1 in FI, 72.3 ± 1.9 m2 tree−1 in DI and affected by local irradiance in all the treatments (P > 0.30),
43.1 ± 4.1 m2 tree−1 in NI. Midday measurements of incident suggesting that the higher M of S-leaves is due to greater leaf
a
quantum irradiance (PPFD) taken on S- and NW-leaves thickness. D
o
w
(Figure 1) were used to derive the exposure adjustment coef- The mass-based leaf nitrogen content (Nm) showed a down- nlo
ficients for PPFD (ε ) in FI, DI and NI trees. It should be ward trend irrespective of the treatment and leaf class a
PPFD d
e
noted that e is used here as a proxy of within-canopy light (Figure 2d and e), with a mean reduction of ~25–30% between d
environmentP PaFnDd that long-term measurements by quantum VG and KF stages (see Tables S1 and S2 available as fro
m
sensors or hemispherical photography would have provided a Supplementary Data at Tree Physiology Online). Water regime h
ttp
more accurate quantification of the light gradients (Niinemets did not affect N over the growing season in any of the leaf s
et al. 1999, Rosati et al. 1999). ε remained relatively con- classes. Despite mthe opposite trends of M and N (Figure 2a, ://ac
PPFD a m a
d
stant during the growing season in all three treatments, with b and d, e), the area-based leaf N content (N = N ⋅ M) e
a m a m
mean seasonal values of 0.62, 0.50 and 0.25 in FI, DI and NI, depended on the growth stage in both S- and NW-leaves (see ic
.o
respectively (data not shown). These results indicate that the Tables S1 and S2 available as Supplementary Data at Tree u
p
horizontal within-crown irradiance gradients are substantially Physiology Online). In FI, N was quite constant throughout the .co
a m
smaller in the NI trees than in the other two treatments, and growing season, while in DI and NI it showed lower values dur- /tre
that incident PPFD on NW-leaves was, on average, 12 and ing the KF stage (see Tables S1 and S2 available as ep
h
37% higher in DI and NI trees, respectively, than in FI trees. Supplementary Data at Tree Physiology Online). ys
/a
The effects of long-term severe drought on plant water sta- No significant differences in mean mass-based leaf total rtic
tus were apparent from the beginning of the measurement chlorophyll content (Chl ) were observed between growth le
m -a
ptherroioudg.h oTuhte thFeI tgreroews inmga isnetaaisnoend (rmeleaatinv evlya luheig ho f Ψ–m1d. 1v MalPuae)s, sSt1a gaensd iSn2 a nayv aoilfa bthlee alesa Sf ucplapslseemse (nFtaigruyr eD a2tga aatn dTr ehe; sPehey sTiaoblolegys bstrac
while Ψmd gradually declined in DI and NI trees, reaching mini- Online). Severe water stress (i.e., NI trees) lowered the mean t/32/4
mum values of ~–2.2 and –3.1 MPa, respectively, by the end of values of Chlm and Chla in S- and NW-leaves. Unlike Chlm, the /45
the measurement period (see Figure S1 available as mean Chl significantly increased as the growing season pro- 0
a /1
6
Supplementary Data at Tree Physiology Online). gressed in both leaf classes (see Tables S1 and S2 available as 3
7
Supplementary Data at Tree Physiology Online). 36
1
b
y
Irradiance plasticity of leaf structural traits g
u
e
Local irradiance (shoot exposure) had a moderate impact on st o
Ma, as εMa was lower than 0.20 throughout the growing cycle n 10
(Table 1, Figure 2c). Although ε tended to be somewhat A
Ma p
lower in FI than in the deficit treatments (Figure 2c), no signifi- ril 2
cant effect of water stress (Table 1) or growth stage was 0
1
9
observed on mean ε (Table 1).
Ma
N tended to be higher in S- than in NW-leaves but mean
m
ε values were somewhat lower than those observed for ε
Nm Ma
(Table 1). ε was not affected by water stress but showed a
Nm
marked seasonal effect, reaching near zero values during the
KF stage (Table 1; Figure 2f). Due to the higher M and N
a m
observed in S- than in NW-leaves, mean N was higher in the
a
Figure 1. Seasonal trend of incident PPFD on S- (black circles) and former (see Tables S1 and S2 available as Supplementary Data
NW-leaves (bars). For S-leaves, mean incident PPFD values across the
at Tree Physiology Online), resulting in mean ε values for both
three irrigation regimes are presented. The error bars denote the stan-
Na
dard error of the mean. leaf classes of ~0.20 and 0.10 during the VG and KF stages,
Tree Physiology Volume 32, 2012
Irradiance plasticity of almond leaf traits 455
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
d
e
m
ic
.o
u
p
.c
o
m
/tre
e
p
h
y
s
Figure 2. Seasonal trends (a and b) of leaf dry mass per unit area (M), (d and e) mass-based leaf nitrogen content (N ) and (g and h) mass-based /a
total chlorophyll content (Chl ) in (a, d and g) S-leaves (left panels)a and in (b, e and h) NW-leaves (central panels). Smeasonal trends of the expo- rtic
m le
sure adjustment coefficient (ε) for Ma (c), Nm (f) and Chlm (i). Abbreviations: S- and NW-leaves were obtained from fruit-bearing shoots located at -a
the south and at the northwest sides, respectively, of the tree crown; FI, full irrigation; DI, sustained deficit irrigation; NI, non-irrigated. bs
tra
c
t/3
2
/4
/4
5
0
Table 1. Mean values of the exposure adjustment coefficient (ε) of various leaf structural and physiological traits for each treatment (FI: full irriga- /1
tion; DI: deficit irrigation; NI: rain-fed) and for each growth stage (VG: vegetative growth; KF: kernel-filling). Ma (g m−2): leaf dry mass per unit area; 637
N (g g−1) and N (g m−2): mass-based and area-based leaf nitrogen content, respectively; Chl (mg g−1) and Chl (mg m−2): mass-based and 3
m N a N m a 6
area-based leaf chlorophyll content, respectively; Amax,m (µmol g−1 s−1) and Amax,a (µmol m−2 s−1): mass-based and area-based maximum net CO2 1 b
assimilation rate, respectively; gs,max (mmol m−2 s−1): maximum stomatal conductance; PNUEmax (µmol gN−1 s−1): maximum photosynthetic nitrogen y g
use efficiency (A /N); WUE (µmol mol−1): intrinsic water use efficiency (A /g ). u
max,a a i max,a s,max e
s
Stage Treatment t o
ε ε ε ε ε ε ε ε ε ε n
(ST) (TM) Ma Nm Na Chlm Chla Amax,a Amax,m gs,max PNUEmax WUEi 1
0
A
VG p
FI 0.08 0.09 0.16 −0.04 0.04a 0.53a 0.49a 0.55a 0.44a −0.06 ril 2
0
DI 0.18 0.11 0.26 0.06 0.22b 0.36b 0.25b 0.26b 0.14b 0.10 1
9
NI 0.15 0.04 0.19 0.00 0.13ab 0.32b 0.20b 0.26b 0.10b 0.08
KF
FI 0.08 0.04 0.13 −0.02 0.06 0.52a 0.49a 0.48a 0.44a 0.06
DI 0.13 0.01 0.12 0.00 0.14 0.40b 0.25b 0.38a 0.25b −0.03
NI 0.09 0.00 0.07 −0.02 0.07 0.09c 0.03c 0.06b −0.01c 0.03
P-value
TM 0.052 0.141 0.292 0.141 0.002 < 0.001 < 0.001 0.001 < 0.001 0.777
ST 0.276 0.004 0.007 0.004 0.135 0.061 0.214 0.422 0.989 0.809
TMxST 0.689 0.459 0.382 0.459 0.270 0.014 0.177 0.139 0.306 0.308
Within each column and growth stage, mean values followed by different letters indicate significant differences at P < 0.05. Separation by Duncan’s
multiple range test at the 95% confidence level. Within the analysis of variance, P-values lower than 0.05 are in bold.
Tree Physiology Online at http://www.treephys.oxfordjournals.org
456 Egea et al.
respectively (Table 1). Similar to ε , ε was growth stage Supplementary Data at Tree Physiology Online). PNUE
Nm Na max
dependent but insensitive to the water regime (Table 1). showed a similar seasonal response to A and A
max,a max,m
Mean ε values approached zero in all the water regimes (Figure 3g and h; see Tables S1 and S2 available as
Chlm
and growth stages (Table 1, Figure 2i). ε tended to be Supplementary Data at Tree Physiology Online).
Chla
higher in the deficit treatments than in FI during the VG stage Unlike A and A , light-saturated stomatal conduc-
max,a max,m
(Table 1). As expected, the Chl /N ratio was not affected by tance (g ) did not appear to be growth stage-dependent in
m m s,max
local irradiance but was significantly reduced in S- and any of the leaf classes (Figure 3d and e). Water stress signifi-
NW-leaves of the NI treatment (~18 and 20%, respectively, cantly decreased g throughout the growing season. Water
s,max
during KF, P < 0.05) compared with the values in FI and DI stress also reduced g in both S- and NW-leaves (Figure 4a).
m
(data not shown). In S-leaves, g was ~35 and 81% lower in DI and NI, respec-
m
tively, than in FI, whereas the respective reductions in
Seasonal trend of leaf physiological traits NW-leaves were ~26 and 66%. D
o
w
Both area-based (Amax,a) and mass-based (Amax,m) light- Stem-to-leaf hydraulic conductance (Ks-l) determined during nlo
saturated net photosynthesis were stage-dependent, as their the KF stage was significantly lower in S-leaves of DI and NI a
d
e
mean values were significantly lower in the KF stage than in compared with FI, whereas K was unaffected by water stress d
the VG stage for the two leaf classes (Figure 3a and b; Tables in NW-leaves (Figure 4b). s-l fro
m
S1 and S2 available as Supplementary Data at Tree Physiology Intrinsic water use efficiency (WUE = A /g ) increased h
i max,a s,max ttp
Online). Water stress significantly decreased A and A with the severity of water stress in both S- and NW-leaves (see s
as the growing season progressed, with a mmax,oare mamrkaex,dm Tables S1 and S2 available as Supplementary Data at Tree ://ac
a
d
decrease in S-leaves (see Tables S1 and S2 available as Physiology Online). No significant differences in WUE were e
i m
ic
.o
u
p
.c
o
m
/tre
e
p
h
y
s
/a
rtic
le
-a
b
s
tra
c
t/3
2
/4
/4
5
0
/1
6
3
7
3
6
1
b
y
g
u
e
s
t o
n
1
0
A
p
ril 2
0
1
9
Figure 3. Seasonal trends (a and b) of maximum net CO assimilation rate on an area basis (A ), (d and e) maximum stomatal conductance
2 max,a
(g ) and (g and h) maximum nitrogen use efficiency (PNUE ) in (a, d and g) S-leaves (left panels) and in (b, e and h) NW-leaves (central
s,max max
panels). Seasonal trends of the exposure adjustment coefficient (ε) for A (c), g (f) and PNUE (i). Abbreviations: S- and NW-leaves were
max,a s,max max
obtained from fruit-bearing shoots located at the south and at the northwest sides, respectively, of the tree crown; FI, full irrigation; DI, sustained
deficit irrigation; NI, non-irrigated. The error bars denote the standard error of the mean.
Tree Physiology Volume 32, 2012
Irradiance plasticity of almond leaf traits 457
Figure 4. (a) Mesophyll conductance to CO diffusion (g ) determined over the kernel-filling stage (KF) in S- and NW-leaves of FI, DI and NI trees;
2 m
(b) stem-to-leaf hydraulic conductance (K ) determined over the kernel-filling stage (KF) in S- and NW-leaves of FI, DI and NI trees. Different D
s-l o
upper-case letters within the same irrigation treatment indicate significant differences at P < 0.05. Within each individual leaf class (i.e., S- or w
n
NW-leaves), different lower-case letters indicate significant differences at P < 0.05. lo
a
d
e
d
observed between the two leaf classes, irrespective of growth The differential response of leaf physiological and structural fro
m
stage and irrigation regime (Table 1). traits was clearly reflected by the relationship between season- h
ttp
Irradiance plasticity of leaf physiological traits laovcearla girera εd ivaanlcuee sg roafd tiheen tlse a(fF itgrauirtes 6an).d W εhPPilFeD ,ε t hoaf tt hise, tphhey sreiolalotigvie- s://ac
a
d
Local irradiance had a deep impact on both A and A cal traits (A , g , g and K ), expressed either on an area e
max,a max,m max s,max m s-l m
(Figure 3a and b), with S-leaves presenting significantly higher or mass basis, were positively related with ε (Figure 6a), ic
PPFD .o
wvaeluree ss torof nAgmlya xi,an flaunedn cAemda xb,my twhaante rN sWtr-elesasv, ewsi.t hε lAomwax,ae r amnde anε Avmaax,ml- tshheo wco arnreys cploenadr icnogr rvealalutieosn owfi tεh oεf lea (f Fsigtruurcet u6rba)l .traits did not up.co
PPFD m
ues in DI and NI than FI throughout the growing season (Figure /tre
3c; Table 1). No stage-dependent effect was observed on ep
Discussion h
ε and ε , but the interaction treatment × growth stage ys
Amax,a Amax,m /a
was significant for εAmax,a (Table 1). Local irradiance affected Plastic responses of leaf traits to local irradiance regime rtic
PNUE and g similarly to A and A (Table 1, Figure le
max s,max max,a max,m -a
3f), with DI and NI displaying lower mean ε and ε Structural leaf traits bs
thaεn o Ff IA (mTaaxb,al,e A 1m)a.x,m, gs,max and PNUEmax presePnNtUeEdm axa tight lingse,maaxr mVearttcicha l tchaen ovpeyr tgicraald igernatds ieinn tMs a iann di rNraad hiaanvcee b e(eNnii noebmseertsv eda ntdo tract/32/4
relationship with Ψmd (Figure 5). The four linear regressions Tenhunen 1997, Wilson et al. 2000, Robakowski et al. 2003, /45
showed similar (P > 0.05) slope, indicating that all have the Montpied et al. 2009), while those of Nm have been reported 0/1
6
same responsiveness to water stress. The intercepts were also to be fairly constant (Ellsworth and Reich 1993, Abrams and 3
7
similar, except for ε , which showed a lower value Mostoller 1995, Wilson et al. 2000, Rijkers et al. 2000, 36
PNUEmax 1
(P < 0.05). Walcroft et al. 2002, Montpied et al. 2009). b
y
Except in NI, gm was significantly lower in NW- than in In our experiment, leaves grown under high irradiance (i.e., gue
S-leaves (Figure 4a), with mean εgm reaching 0.62, 0.57 and S-leaves) also had higher Ma than those grown under low st o
0.30 in FI, DI and NI, respectively. Ks-l was also ~70% lower in irradiance (i.e., NW-leaves) in all the treatments, as reported n 1
0
NW- than in S-leaves of FI (Figure 4b), resulting in a mean ε for other fruit tree species (DeJong and Doyle 1985, A
Ks−1 p
of 0.70, 0.44 and −0.04 in FI, DI and NI, respectively. Weinbaum et al. 1989, Walcroft et al. 2002). However, mean ril 2
values of ε revealed that the irradiance-dependent plastic- 01
Comparative response of irradiance plasticity of Ma 9
ity of M was low in all the treatments (within 0.08–0.18)
a
structural and physiological traits
(Table 1). The L /L ratio (a surrogate of leaf density) was
DM FM
Season-average ε values of the physiological traits (A , quite insensitive to local irradiance in all the treatments (data
max,a
A , g , g and K ) showed a downward trend with the not shown), suggesting that leaf thickness was greater in
max,m s,max m s-l
intensity of water stress, whereas the season-average ε values S-leaves. Thicker leaves are likely to have more chloroplasts
of the structural traits (M , N and N) were quite similar per unit of leaf area (Rieger and Duemmel 1992, Evans and
a m a
among irrigation regimes and were substantially lower (within Poorter 2001, Terashima et al. 2006) and hence a higher
the range 0.01–0.18) than those of physiological traits (see photosynthetic capacity, as previously observed in P. persica
Figure S2 available as Supplementary Data at Tree Physiology (Syvertsen et al. 1995) and in this study (Figure 3a and b;
Online). Table 1).
Tree Physiology Online at http://www.treephys.oxfordjournals.org
458 Egea et al.
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
Figure 5. Relationship between the exposure adjustment coefficient (ε) for (a) area-based maximum net CO2 assimilation rate (Amax,a), (b) mass- dem
based maximum net CO2 assimilation rate (Amax,m), (c) maximum stomatal conductance (gs,max) and (d) maximum nitrogen use efficiency (PNUEmax) ic
with midday stem water potential (Ψ ) for FI, DI and NI trees. The error bars denote the standard error of the mean. .o
md u
p
.c
o
m
/tre
e
p
h
y
s
/a
rtic
le
-a
b
s
tra
c
t/3
2
/4
/4
5
0
/1
6
3
7
3
6
Figure 6. Relationship between mean seasonal values of the exposure adjustment coefficient for the incident PPFD (εPPFD) with the corresponding 1 b
mean seasonal ε values for (a) area- (Amax,a) and mass-based (Amax,a) maximum net CO2 assimilation rate, maximum stomatal conductance (gs,max), y g
stem-to-leaf hydraulic conductance (K ), mesophyll conductance (g ), and (b) leaf dry mass per unit area (M), area- (N) and mass-based (N ) u
s-l m a a m e
leaf nitrogen content, area- (Chla) and mass-based (Chlm) total chlorophyll content. The solid lines represent the 1 : 1 relationship. The error bars st o
denote ± SE. In the y-axis labels, ε denotes the exposure adjustment coefficient for leaf trait X. n
X 1
0
A
p
The low plasticity of Nm observed in response to local irradi- decrease in Nm over time may result from (i) the dilution of N in ril 2
ance (Table 1) agrees with the results reported by Niinemets the leaf blade because of the accumulation of non-N c ompounds, 0
1
9
(1995) and Iio et al. (2005) and by Montpied et al. (2009), or (ii) from N allocation to developing almond kernels, espe-
who found a gradient of <10% N across a beech canopy. Our cially over the kernel-filling stage (Egea et al. 2009). The fact
m
results also agreed with previous studies on other fruit tree spe- that N was strongly related to M and to the ratio L /L
m a DM FM
cies, which reported that N was marginally affected by local (results not shown), a surrogate of leaf density (Niinemets et al.
m
irradiance (Le Roux et al. 1999, Rosati et al. 1999, 2000, 2005) suggests that the age-dependent decline of N was
m
Walcroft et al. 2002). In contrast with the limited impact of irra- dominated by changes in leaf density due to the accumulation
diance on leaf N content, ontogeny significantly affected the of non-photosynthetic compounds, although a concomitant allo-
seasonal time-course of N (Figure 2d and e). A similar cation of N to developing kernels cannot be ruled out.
m
response was previously reported for sun leaves of almond The plasticity of N to local irradiance was similar to that of
a
trees (Nortes et al. 2009, Egea et al. 2010). The observed N . Although moderate compared with other studies (e.g.,
m
Tree Physiology Volume 32, 2012
Irradiance plasticity of almond leaf traits 459
Warren et al. 2007), the observed differences in N among leaf (2010) for rainforest tree species, but which contradicts the
a
classes agreed with previously reported differences between findings of Rijkers et al. (2000)who found no significant effect
sun and shade leaves in P. persica (Lloyd et al. 1992). The of vertical irradiance gradients and tree height on PNUE, and of
finding that the total pool of leaf N (both area- and mass- Rosati et al. (1999) who found similar PNUE in nectarine
max
based) was marginally affected by water stress × irradiance (P. persica) leaves irrespective of irradiance and N resources.
interactions agrees with a previous work on Quercus suber L. Unlike A and PNUE , WUE plasticity to irradiance was
max max i
(Aranda et al. 2005). slight, irrespective of the growth stage and irrigation regime
Shade leaves generally have higher Chl than sun leaves (Table 1). Changes in WUE may occur if A and g
m i max,a s,max
(Poorter et al. 1995, Rijkers et al. 2000), but in almond leaves respond differently to irradiance, but in our study both vari-
Chl was not affected by local irradiance (Table 1). However, ables presented a similar degree of plasticity to local irradiance
m
Chl presented a degree of plasticity similar to M (Table 1), as (Table 1). These results contrast with previous findings (Hanba
a a
corroborated by the strong positive relationship found between et al. 1997, Aranda et al. 2007, Craven et al. 2010) that indi- D
o
w
both variables when pooling data of S- and NW-leaves (results cated a positive response of WUEi to increasing irradiance, irre- nlo
not shown). A larger chlorophyll content on an area basis in spective of water regime (Aranda et al. 2007). Irradiance-elicited a
d
e
sun than in shade leaves was also reported by Montpied et al. changes of WUE have been attributed to variations in leaf mor- d
(2009) and Demarez et al. (1999) in beech canopies, while phology (Arandai et al. 2007). Leaf morphological traits (e.g., fro
m
Lichtenthaler et al. (2007) and Iio et al. (2005) did not observe M) were little affected by irradiance in almond leaves (Table 1), h
a ttp
this type of response. which could explain the lack of response of WUE to local irradi- s
ance in this species. i ://ac
a
d
Physiological leaf traits e
m
The photosynthetic capacity of leaves is known to display a Do water stress regime and ontogeny affect irradiance ic
strong degree of plasticity in response to changes in irradiance plastic response? .ou
p
within the canopy through a large variety of structural and physi- The interaction irradiance × water regime was highly significant .co
m
ological changes. Leaf photosynthetic capacity, as described by (P < 0.001) for Amax, gs,max and PNUEmax (Table 1), as illustrated /tre
Amax (both area- and mass-based), was more plastic to the irra- by the strong linear relationships between ε values of these eph
diance regime than leaf structural traits (Table 1). These results traits and Ψmd (Figure 5). These results are in contrast to those ys/a
are consistent with Valladares et al. (2000), who found an of previous work reporting independent effects of shade and rtic
almost twofold higher plasticity in physiological than in morpho- water shortage on seedlings of different species (Sack and le
-a
logical traits of 16 species of tropical rainforest shrubs, but dis- Grubb 2002, Sack 2004, Aranda et al. 2005), but agree with bs
agree with the results reported by Coste et al. (2010) for 12 other work reporting the opposite response (Holmgren 2000, trac
rainforest tree species. It is generally accepted that area-based Quero et al. 2006). In our study, the quasi-linear relationship t/32
/4
Amax increases with irradiance (e.g., Rijkers et al. 2000, Warren found between the degree of plasticity to local irradiance of /45
et al. 2007, Coste et al. 2009), as confirmed by our results A , A , g and g and the magnitude of the relative 0
max,a max,m s,max m /1
6
(Figure 3a and b; Table 1), whereas mass-based A was either gradient of irradiance (ε , Figure 6) suggested that leaf trait 3
max PPFD 7
unaffected (Coste et al. 2010) or marginally increased (Rijkers adjustments of shade leaves are tightly and locally driven by 36
1
et al. 2000). In almond leaves, the plasticity of A to local the amount of irradiance received by the leaf, irrespective of b
max,m y
irradiance was high, although slightly lower than that of Amax,a the water regime. gue
(Table 1). As Amax,a is equal to the product Amax,m × Ma, the Except for leaf N (both area- and mass-based) and mass- st o
observed irradiance-elicited changes in Amax,a must have resulted based Chl, ontogeny did not significantly affect the irradiance- n 1
0
from variations in both Amax,m and Ma. Previous works in other elicited plasticity of most of the studied leaf traits (Table 1). Ap
tree species (e.g., Gulmon and Chu 1981, Niinemets et al. 1998, This result supports previous findings which stressed that the ril 2
Rijkers et al. 2000, Montpied et al. 2009), including fruit trees degree of irradiance-elicited plasticity was independent of 0
1
9
(Juglans regia L., Le Roux et al. 1999), demonstrated that photo- ontogeny in most of the leaf traits of two rain-forest tree spe-
synthetic irradiance acclimation was dominated by changes in cies (Coste et al. 2009).
M . However, comparison of these results with our findings indi-
a
cates that there is no universal rule for the irradiance acclimation
Processes involved in irradiance-elicited plasticity
of photosynthetic capacity, as also suggested by Le Roux et al.
(2001), who found that changes in M and shifts in leaf N parti- Changes in N partitioning
a
tioning were the main factors driving irradiance-induced changes It is widely accepted that the acclimation of leaves to photosyn-
in the leaf photosynthetic capacity in P. persica. thetic irradiance may result from changes (i) in N , (ii) in M
w a
Light-saturated PNUE (PNUE ) increased greatly with irra- and/or (iii) in leaf N partitioning (e.g., Robakowski et al. 2003).
max
diance (Table 1), a behavior that was reported by Coste et al. Our finding that strong variations in A within the almond
max
Tree Physiology Online at http://www.treephys.oxfordjournals.org