Mg

NLAD

a

ARRA

KCLPPPS

1

patouTct

0h

Environmental and Experimental Botany 102 (2014) 1–10

Contents lists available at ScienceDirect

Environmental and Experimental Botany

jo ur nal homep ag e: www.elsev ier .com/ locate /envexpbot

orphological and physiological acclimations of coffee seedlings torowth over a range of fixed or changing light supplies

elson F. Rodríguez-López, Samuel C.V. Martins, Paulo C. Cavatte, Paulo E.M. Silva,eandro E. Morais, Lucas F. Pereira, Josimar V. Reis, Rodrigo T. Ávila, Alice G. Godoy,lyne O. Lavinski, Fabio M. DaMatta ∗

epartamento de Biologia Vegetal, Universidade Federal de Vic osa, 36570-000 Vic osa, MG, Brazil

r t i c l e i n f o

rticle history:eceived 4 June 2013eceived in revised form 19 January 2014ccepted 20 January 2014

eywords:offea arabicaight acclimationhenotypic plasticityhotosynthesishotoinhibitionun/shade tolerance

a b s t r a c t

Acclimations to low and high light levels are located in opposite extremes. We examined how the coffeeplant (Coffea arabica), which is considered to be shade-demanding although it performs well under full sunexposure, can acclimate to a range of light regimens. We hypothesised that the growth and physiologicalperformance depends on the total amount of light received by the plant per day and on the temporalorder of diurnal variations in the light supply. The biomass accumulation and allocation and the keyphotosynthetic traits of pot-grown coffee seedlings were examined over a range of light treatmentsas follows: plants grown entirely under 100%, 40% or 10% sunlight; plants grown at either 40% or 10%sunlight from sunrise to midday and then submitted to full sunlight until sunset; and plants grown underfull sunlight from sunrise to midday and then submitted to either 40% or 10% sunlight throughout theafternoon. The total biomass increased linearly with the increasing total light supply; however, plantsreceiving high amounts of light in the morning grew more than those receiving high amounts of light inthe afternoon. Extensive morphological changes (e.g., specific leaf areas and leaf area ratios) primarilyresponsive to total light rather than diurnal light fluctuations or light quality, at a given light supply, werenoted. In contrast, changes in the photosynthetic performance per unit leaf area among the treatments

were narrower and apparently unrelated to carbohydrate accumulation or photoinhibition. Overall, wefound a poor leaf-level physiological plasticity of traits to light. We suggest that adjustments in leafnumber and leaf area, coupled with whole-plant physiological adjustments, could largely account forthe differences in the biomass amongst treatments. We also suggest that coffee could be classified as ashade-tolerant species as its traits do not match the requirements of either a classic shade-avoiding or aclassic shade-demanding species.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

Although light is a crucial environmental resource that driveshotosynthesis and ultimately influences plant growth, both lownd high sunlight can limit plant performance. On the one hand,he growth and survival can be compromised by low sunlight;n the other hand, desiccation, heat and light stress can occurnder high sunlight conditions (Valladares and Niinemets, 2008).

o cope with these stresses, plants have evolved a number of well-haracterised biochemical, physiological and structural changes athe leaf and whole-plant levels that enable them to adjust to a

∗ Corresponding author. Tel.: +55 31 3899 1291; fax: +55 31 3899 2580.E-mail address: fdamatta@ufv.br (F.M. DaMatta).

098-8472/$ – see front matter © 2014 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.envexpbot.2014.01.008

particular set of light conditions (Boardman, 1977; Walters, 2005;Lusk et al., 2008; Niinemets, 2010). Many studies of the plant lightresponse have aimed at unveiling the morphological and physiolog-ical mechanisms as well as the ecological implications of toleranceto extremes, such as tolerance to either sun or shade (Boardman,1977). In these studies, although the plants may be constrained todifferent extents within their whole-plant allocation pattern, theytend to show universal plastic responses, such as increasing massallocation to the shoots rather than to the roots when growing inthe shade (Poorter et al., 2012). Irrespective, noticeably fewer stud-ies have been conducted considering that the timing and duration

of fluctuations of photosynthetically active radiation (PAR) can dra-matically vary throughout plant canopies (Niinemets, 2007, 2010).In this context, the patterns of biomass accumulation and alloca-tion in plants that are subjected to temporal variations in the PAR

2 tal an

a1

titeebtrtadtorts

tfswhpDhlasttm(ecdngdtsoas

aophmdrbtBcalnaebti

N.F. Rodríguez-López et al. / Environmen

vailability throughout the day have barely been explored (Ackerly,997).

The effects of the variable light environments are well charac-erised in the context of bursts of direct sunlight (i.e., sunflecks),n which the duration and frequency of light patches largely drivehe carbon assimilation and growth (Way and Pearcy, 2012). How-ver, very few studies (e.g., Wayne and Bazzaz, 1993; Watlingt al., 1997) have compared plants grown under fluctuating lightut receiving the same total daily PAR. Some evidence suggestshat the whole plant carbon gain may be affected by the tempo-al scale of diurnal fluctuations in light environments, even whenhe total amount of PAR received is kept constant (Ackerly, 1997,nd references therein), although this response may be speciesependent (Watling et al., 1997). To the best of our knowledge,hese few studies have been undertaken in the field of plant ecol-gy and forest biology, and virtually nothing has been exploredegarding the trends in the growth and physiological responses toemporal scales of diurnal variations in light availability in croppecies.

Coffee is one of the most important commodities in the interna-ional agricultural trade. It evolved in the understory of the Africanorest and is traditionally considered to be a shade-demandingpecies. In early plantations, Arabica coffee (Coffea arabica) bushesere planted under taller shade trees to simulate their naturalabitat. However, modern coffee cultivars can grow well and evenroduce greater yields in the sun than in the shade (DaMatta, 2004;aMatta et al., 2010), which suggests that coffee plants shouldave sufficient plasticity to acclimate themselves to contrasting

ight environments (Matos et al., 2009; Cavatte et al., 2012). Inny case, even in un-shaded plantations, the coffee plant may beubjected to pronounced temporal scales of diurnal variations inhe PAR supply, which are frequently found in hilly zone planta-ions. Depending on the relative crop position on a hill, the plant

ay be mostly shaded during either the morning or afternoonwith terrain exposure facing west and east, respectively). We havempirically observed that, when shaded in the morning under fieldonditions, the growth (and production) of coffee plants may beepressed relative to plants that receive a more continuous diur-al light supply; we have noted an inverse relationship in plantsrown under full sunlight in the morning and subjected to someegree of shade in the afternoon (personal observation). However,o the best of our knowledge, how the coffee plants adjust them-elves, both structurally and physiologically, to cope with this typef temporal variation in the PAR supply to improve their fitnessnd hence biomass production have never been examined in thispecies.

We hypothesised that physiological performance, accumulationnd biomass allocation patterns in coffee plants are dependent notnly on the total amount of PAR received but also on the tem-oral scale of diurnal light availability. In this context, we furtherypothesised that plants receiving greater amounts of PAR in theorning than in the afternoon, but receiving similar integrated

aily photon flux, are able to accumulate more biomass, and theseesponses might be associated with increased carbon gains in com-ination with higher stomatal conductance in the morning whenhe PAR is non-limiting (Araújo et al., 2008; DaMatta et al., 2008;atista et al., 2012). Finally, we also hypothesised that physiologi-al (e.g., traits associated with photoprotection and maintenance of

positive carbon balance) rather than morphological (e.g., specificeaf area and leaf area ratio) adjustments are the major determi-ants for acclimations to temporal scale of diurnal light availability,s previously observed in coffee under fixed light supplies (Matos

t al., 2009; Cavatte et al., 2012). To test these hypotheses, theiomass accumulation and allocation and the key photosyntheticraits of coffee plants were examined over a range of fixed or chang-ng light supplies.

d Experimental Botany 102 (2014) 1–10

2. Materials and methods

2.1. Plant material, growth conditions and experimental design

The experiments were conducted in Vic osa (20◦45′ S, 42◦54′ W;650 m in altitude) in southeastern Brazil. Uniform coffee seedlings(C. arabica L. cv. ‘Red Catuaí IAC 44’) were grown from seedsand transplanted after growing three leaf pairs (January 7, 2010)into 12 L pots containing a mixture of soil, sand and compostedmanure (4:1:1 v/v/v). After transplantation, the seedlings were ran-domly submitted to seven light treatments as follows: plants grownentirely under 100%, 40% or 10% sunlight (S-100, S-40 and S-10,respectively); plants grown every day at either 40% or 10% sun-light throughout the morning (from sunset to midday) and thensubmitted to full sunlight until sunrise (S-40/100 and S-10/100,respectively); and plants grown every day under full sunlight fromsunrise to midday and then submitted to either 40% or 10% sunlightthroughout the afternoon (S-100/40 and S-100/10, respectively).Every day, the shade shelter was removed or added at middayaccording to the treatment. All of these light treatment combina-tions were applied for 150 days. The shade shelters were providedby neutral density black nylon nettings. The light quality (blue/redand red/far red ratios) beneath the nettings as well as in the open(direct solar radiation) was checked in cloudless days at dawn, mid-day and dusk using a FieldSpec® spectroradiometer (ASD, Boulder,USA) following the recommendations of Smith (1982).

The plants were irrigated and fertilised as required, and noapparent restriction was observed in the root development at theend of the experiment. The pot positions were periodically ran-domised to minimise any variation within each light environment.The plants were distributed over a completely randomised designwith 10 replicate plants per treatment, and the experimental plotconsisted of one individual plant per pot. When the measurementswere made on a single leaf (for photosynthetic parameters), theyoungest, most fully expanded leaves from five individuals wereused.

Throughout the experiment, the maximum and minimum aver-age temperatures were 27.7 ◦C and 14.7 ◦C, respectively. Theaverage air temperature was 21.4 ± 0.8 ◦C, and the relative humid-ity was 79.2%. The PAR was measured using LI-190SA quantumsensors (Li-Cor; Lincoln, NE, USA) that were positioned 0.5 m abovethe plant canopies. All sensors were installed at the experimentalsite and connected to an LI-1400 data logger (Li-Cor) that acquireddata from the sensors every minute and stored them as 5-min aver-ages. The day length corresponded approximately to 13.2 h at thebeginning of the experiment and it was progressively shortenedto approximately 10.8 h at the end of the study. The climate at theexperimental site is classified as Cwa (mesothermal) by the Köppensystem, with a rainy, warm season (October–March) followed by adry, cool season (April–September).

2.2. Growth traits

At the end of the experiment, various morphological traits[heights, total leaf numbers per plant, numbers of plagiotropicbranches and stem diameters (5 cm above ground)] were recorded.Total leaf areas were estimated by putting the maximum leafwidths and lengths into the equations described by Antunes et al.(2008). In addition, the plants were harvested and separated intoorthotropic and plagiotropic branches, leaves and roots. The rootswere washed thoroughly with tap water over a 0.5 mm screenedsieve. The plant tissues were oven-dried at 70 ◦C for 72 h, after

which the dry weights of the leaves, branches and roots were mea-sured. Based on these data, the total biomasses, leaf mass fractions(LMF), orthotropic branch mass fractions (OMF), plagiotropic (lat-eral) branch mass fractions (PMF) and root mass fractions (RMF)

tal and

wdrF

2

(mp(dI(cst1aaauetsetmard

ietds(u0waPs(soepqtwJwMcttc

o1taa

N.F. Rodríguez-López et al. / Environmen

ere obtained. The specific leaf areas (SLA) were estimated byividing the total leaf areas by the leaf biomasses. The leaf areaatios (LAR) were calculated as the product of the LMF and SLA.urther details have been described elsewhere (Dias et al., 2007).

.3. Photosynthetic measurements

The net rate of carbon assimilation (A), stomatal conductancegs) and internal-to-ambient CO2 concentration ratio (Ci/Ca) were

easured in an open system under ambient temperature and CO2artial pressure using two cross-calibrated infrared gas analysersLI-6400, Li-Cor). The measurements were made in clear-sky daysuring two time periods: 800–1000 h and 1400–1600 h (solar time).

n these periods we selected leaves based on their relative positionsazimuthal and vertical orientations) so that for the prevailing PARonditions during the measurements the selected leaves were asimilar as possible regarding to the intercepted PAR. The values ofhe intercepted PAR, measured with a photometer/radiometer (Li-85, Li-Cor, Nebraska, USA), were, on average, 1015 ± 104, 509 ± 40nd 130 ± 9 �mol photons m−2 s−1 at the leaf level (for 100%, 40%nd 10% sunlight, respectively); these values were quite similars measured at both mid morning and mid afternoon. To improveniformity and avoid temporary cloud covers, measurements of gasxchanges were then conducted using an artificial PAR equivalento the averaged values of PAR (mentioned above) intercepted by theampled leaves in their natural angles for each light treatment atach time point. After clamping the leaf tissue in the leaf chamber,he rates of gas exchange were typically settled within 3–4 min. The

easurements were repeated three times (two at the end of Aprilnd one in mid-May) so that the gas exchange parameters for eacheplicate were computed as the average values for the measuredays.

The chlorophyll a fluorescence parameters were determinedn the same leaves (and time periods) that were used for the gasxchange measurements with a portable pulse amplitude modula-ion fluorometer (Mini PAM, Walz, Effeltrich, Germany). Followingark adaptation for 30 min using standard leaf clips, the leaf tis-ue was illuminated with a weak modulated measuring beam0.03 �mol m−2 s−1) to obtain the initial fluorescence (F0). A sat-rating white light pulse of 6000 �mol m−2 s−1 was applied for.8 s to ensure the maximum fluorescence emission (Fm) fromhich the variable-to-maximum fluorescence ratio was calculated

s Fv/Fm = [(Fm − F0)/Fm)]. The leaf tissue was exposed to actinicAR (intensities equal to those used for the gas exchange mea-urements) for 60 s to obtain the steady-state fluorescence yieldFs). A saturating white light pulse (6000 �mol m−2 s−1; 0.8 s) wasubsequently applied to achieve the maximum light-adapted flu-rescence (Fm

′). The light-adapted initial fluorescence (F0′) was

stimated according to Oxborough and Baker (1997). Using thesearameters the photochemical (qP) and non-photochemical (NPQ)uenching coefficients, the actual quantum yield of the PSII elec-ron transport (˚PSII) and apparent electron transport rate (ETR)ere calculated by using the equations described by Maxwell and

ohnson (2000). All of the above photosynthetic measurementsere repeated three times (two at the end of April and one in mid-ay) so that the photosynthetic parameters for each replicate were

omputed as the average values for the measured days. Within eachime period, measurements were carried out in blocks in an attempto minimise the effects of potential variations of the environmentalonditions.

The photosynthetic light-response curves (A/PAR) werebtained by increasing the PAR in ten steps from 0 to

500 �mol m−2 s−1 at 25 ◦C. The leaf tissues were initially exposedo a low ambient CO2 concentration (50 �mol mol−1 air for 5 min tollow for stomatal opening; the A/PAR curves were then producedt a CO2 concentration of 400 �mol mol−1 air. The dark respiration

Experimental Botany 102 (2014) 1–10 3

rates (Rd), light compensation point (LCP), light saturation point(LSP) and light-saturated A (Amax) were determined from thesecurves by fitting the data to a modified rectangular hyperbolemodel proposed by Ye (2007). The model was fitted to the datausing nonlinear regression analysis by applying iterative curvefitting (minimum least square difference) using the MicrosoftExcel Solver tool (Microsoft Corporation, Redmond, WA, USA).It was assumed that the LSP corresponded to the PAR neededto achieve 90% of the Amax (Ye, 2007). The responses of A tothe internal CO2 concentration (A/Ci curve) were determined at1000 �mol (photons) m−2 s−1 at 25 ◦C. The measurements beganat a CO2 concentration of 350 �mol mol−1 air, and once the steadystate was reached, the CO2 concentration was gradually lowered to50 �mol mol−1 air and increased stepwise to 1600 �mol CO2 mol−1

air. From these curves, the maximum rate of carboxylation (Vcmax)and maximum rate of carboxylation limited by electron transport(Jmax) were estimated by fitting the mechanistic model of CO2assimilation proposed by Farquhar et al. (1980) with modificationsby Sharkey (1985).

2.4. Biochemical assays

Leaf discs were collected in clear-sky days, prior to flash freezingin liquid nitrogen and subsequent storage at −80 ◦C until analy-sis. A 10 mg sample of ground tissue was added to pure methanol,and the mixture was incubated at 70 ◦C for 30 min. After centrifu-gation (13,000 × g, 5 min), the hexoses (glucose and fructose) andsucrose in the supernatant were quantified; the concentration ofstarch was determined from the methanol-insoluble pellet as pre-viously detailed (Praxedes et al., 2006; Ronchi et al., 2006). Thetotal Chl and carotenoids (Car) were extracted using 80% (v/v)aqueous acetone and quantified according to a procedure reportedby Lichthenthaler (1987). The total nitrogen (N) was estimated asdescribed by DaMatta et al. (1999).

2.5. Statistics

The data were statistically analysed following either a ran-domised block design (A, gs, Ci/Ca and Chl fluorescence parameters)or a completely randomised design (the remaining parame-ters) with five to 10 replicates. The assumptions of normalityand homoscedasticity were checked. The data were analysedby one-way ANOVA, and the means were examined using theNewman–Keuls test at P ≤ 0.05. All statistical analyses were per-formed using the SAEG software version 9.1 (SAEG, 2007).

3. Results

3.1. PAR treatments

There were relatively large fluctuations in the total daily inci-dent PAR over the course of the experiment; such fluctuationsconsiderably diminished from early April onwards, a fact probablyassociated with a decreasing rainfall pattern (and thus less cloudcover) despite the decreasing daylength (Fig. 1). Over the experi-mental period, the incident PAR values above the plant canopieswere, on average, 26.4, 10.5 and 2.9 mol m−2 d−1 for S-100, S-40and S-10 individuals, respectively, with values ranging from 13.5to 19.2 mol m−2 d−1 over the plant canopies plants that were sub-jected to changing PAR supplies (Fig. 1). As expected, the plants thatwere shaded only in the morning received the lowest proportionof PAR during this period, while the opposite was true for plants

shaded only in the afternoon (Fig. 1).

The shade shelters did not alter the quality of the transmittedlight compared to direct sunlight, as deduced by the unchang-ing blue/red and red/far red ratios (approximately 0.98 and 1.17,

4 N.F. Rodríguez-López et al. / Environmental and Experimental Botany 102 (2014) 1–10

Fig. 1. The time course of daily photosynthetically active radiation, PAR(mol m−2 d−1) over the coffee plants (upper) that were subjected to seven light treat-ments, defined as follows: plants grown entirely under 100%, 40% or 10% sunlight(S-100, S-40 and S-10, respectively), plants grown at either 40% or 10% sunlight untilmidday and then submitted to full sunlight until sunset (S-40/100 and S-10/100,respectively) and plants grown under full sunlight from sunrise to midday and thensubmitted to either 40% or 10% sunlight throughout the afternoon (S-100/40 andS-100/10, respectively). The experiment, conducted at Vic osa, southeastern Brazil,was initiated on January 7, 2010. The average (over 150 days) of total daily PAR andt(

rfb

3

P9ttibttSstni

Table 1Blue/red and red/far red ratios, as measured at full sunlight (100%) or beneath theshade nettings (40% or 10% full sunlight) at dawn (approximately 5 min before sun-rise: 05:20 h), midday and dusk (approximately 5 min after sunset: 18:40 h) fortypical clear sky days (06–08 January, 2014) in Vic osa, southeastern Brazil. Thestandard error did not exceed 3% of the mean value.

Time of the day Light level (%) Ratios

Blue/red Red/far red

Dawn 100 1.34 0.7640 1.37 0.7910 1.37 0.80

Midday 100 0.98 1.2940 0.98 1.3010 0.96 1.28

Dusk 100 1.49 0.8040 1.44 0.7910 1.38 0.78

Fig. 2. The relationship between the total biomass and photosynthetically activeradiation (PAR) of coffee plants subjected to a range of PAR treatments. In the insets,

he proportion of PAR from sunrise to midday (PARSR-MD) and from midday to sunsetPARMD-SS) over the coffee plants are also shown (lower). n = 150 ± SE.

espectively), as determined at midday. These ratios, although dif-ering at both dawn and dusk, were virtually unchanged by thelack nettings that were used (Table 1).

.2. Growth traits

The total biomass increased linearly with the increasing totalAR supply (r2 = 0.93, P < 0.001; Fig. 2) with numbers ranging from.4 g in S-10 plants to 53.8 g in their S-100 counterparts (Fig. 3). Theotal biomass correlated significantly with the PAR received duringhe morning (r2 = 0.69, P = 0.02) but not with the PAR received dur-ng the afternoon (P = 0.34) (Fig. 2). These results indicate that theiomass accumulation responded not only to the total PAR but alsoo the temporal scales of the PAR supply; these findings were addi-ionally highlighted by the larger (52%) biomass in S-100/10 than in-10/100 individuals despite the small differences in the total PARupply (Fig. 1), whereas the differences (18%) in biomass between

he S-100/40 and S-40/100 individuals did not reach statistical sig-ificance (Fig. 3). However, a greater (35%) biomass was observed

n S-100/40 compared to S-40 individuals (Fig. 3).

the relationship between the total biomass and PAR received from sunrise to midday(PARSR-MD) or from midday to sunset (PARMD-SS).

The morphological traits examined showed a significantresponse to the differences in the varying PAR regimes (P < 0.05,Fig. 3). The plant heights decreased significantly (18%) in the S-10plants compared to plants from the other treatments, which did notdiffer from one another. The highest leaf numbers were observedin S-100 individuals and decreased by 23% and 71% in the S-40and S-10 treatments, respectively. In plants that were subjected totemporal variations in the PAR availability, the leaf number in theS-40/100 and S-100/40 plants were similar, whereas the number inS-100/10 individuals was 20% greater than in the S-10/100 individ-uals (Fig. 3). Compared to the S-100 plants, decreases in the totalleaf area were observed in S-40 (21%) and S-10 plants (64%). Thetotal leaf areas were similar in S-40/100 and S-100/40 plants, butthey were lower (29%) in S-10/100 plants than in their S-100/10counterparts. The numbers of plagiotropic branches were remark-ably lower in S-10 individuals and did not differ among the plantsfrom the other treatments. The pattern of stem diameter changesin response to the light regimens was similar to that of the totalbiomass (Fig. 3).

The SLA increased significantly with the decreasing PAR for indi-viduals grown at a fixed PAR; for individuals grown under varying

PAR, the SLA from the S-40/100 and S-100/40 plants was similarto the S-100 plants (∼11 m2 kg−1), whereas for the S-10/100 andS-100/10 plants, the SLA was similar to that of the S-40 plants

N.F. Rodríguez-López et al. / Environmental and Experimental Botany 102 (2014) 1–10 5

Fig. 3. Total biomass, height, stem diameter, plagiotropic branch number, total leaf number per plant, total leaf area, specific leaf area (SLA) and leaf area ratio (LAR) in coffeep . MeaN

(b(

tb

lants subjected to seven light treatments (see further details in the legend of Fig. 1)ewman–Keuls). n = 10 ± SE.

∼13 m2 kg−1) (Fig. 3). The changes in the SLA were accompaniedy alterations in the LAR (Fig. 3), which ranged from 5.3 m2 kg−1

S-100) to 10.2 m2 kg−1 (S-10).In contrast to LMF, which was virtually unaltered regardless of

he light regimens, some significant alterations in the patterns ofiomass allocation among the treatments were noted (Fig. 4): (i)

ns followed by the same letter do not differ significantly from one another (P < 0.05,

the RMF was larger (approximately 20%) in the S-100 plants thanin S-40 and S-10 individuals (leading to an increased root-to-shoot

ratio; data not shown) but similar in the plants from the other PARtreatments; and (iii) the overall proportion of the stem biomassinvested into plagiotropic branches decreased (at the expense ofthe orthotropic branches) with a decreasing PAR supply (Fig. 4).

6 N.F. Rodríguez-López et al. / Environmental an

Fig. 4. The biomass allocation traits [root mass fraction (RMF), orthotropic branchmass fraction (OMF), plagiotropic branch mass fraction (PMF) and leaf mass fraction(lo

3

vpidtiawnCld

TP�(p(

TLwN

LMF)] in coffee plants subjected to seven light treatments (see further details in theegend of Fig. 1). Means followed by a same letter do not differ significantly fromne another (P < 0.05, Newman–Keuls). n = 10.

.3. Photosynthetic performance

The gas exchange parameters were assessed under the pre-ailing PAR availability during their measurements, thus moreroperly reflecting the differences in the actual CO2 fixation capac-

ty between treatments. In the morning, A did not significantlyiffer among individuals receiving 40% or 100% sunlight, whereashe absolute lowest A values were found in the S-10 and S-10/100ndividuals (Fig. 5). Changes in A were highly correlated with vari-tions in gs (r2 = 0.76, P < 0.01, n = 35). Overall, the largest gs valuesere observed in the S-100 individuals; the gs did not differ sig-

ificantly among the plants from the other groups. The highesti/Ca (0.69) was observed in the S-100 and S-10/100 plants, and the

owest values (0.59) were observed in the S-40 plants with interme-iate values in the other groups (Fig. 5). In the afternoon, the A and

able 2hotosynthetic variables derived from the net photosynthetic rates (A) and irradiance cmol photons m−2 s−1), light saturation point (LSP; �mol photons m−2 s−1) and dark respi

A/Ci) curves [maximum rate of carboxylation limited by electron transport (Jmax; �mol−

lants that were subjected to seven light treatments. See further details in the legend of FiP < 0.05, Newman–Keuls test). n = 5 ± SE.

Parameters Light treatments

S-100 S-40 S-10

Amax 6.1 ± 0.3b 4.7 ± 0.1cd 4.0 ± 0.3d

LCP 10.3 ± 0.5a 6.8 ± 1.0bc 3.0 ± 0.4d

LSP 421 ± 27a 295 ± 12b 287 ± 6b

Rd 0.78 ± 0.01a 0.49 ± 0.07ab 0.22 ± 0.04c

Jmax 87.8 ± 4.8a 91.0 ± 1.6a 69.0 ± 8.7b

Vcmax 54.5 ± 3.3a 49.7 ± 4.5a 50.3 ± 2.2a

able 3eaf concentrations of total chlorophylls (Chl; g kg−1 DM) and carotenoids (Car; g kg−1 DMere subjected to seven light treatments (see further details in the legend of Fig. 1). Meanewman–Keuls test). n = 5 ± SE.

Parameters Light treatments

S-100 S-40 S-10

Chl 10.0 ± 0.9b 8.7 ± 0.9b 14.4 ± 0.7a

Car 2.1 ± 0.1b 1.9 ± 0.2b 2.8 ± 0.1a

Chl/Car 4.7 ± 0.2a 4.5 ± 0.1a 4.7 ± 0.1a

Chl a/b 2.8 ± 0.1a 2.9 ± 0.1a 2.9 ± 0.1a

Chl/N 3.1 ± 0.3b 2.6 ± 0.3b 4.7 ± 0.3a

d Experimental Botany 102 (2014) 1–10

gs decreased markedly relative to their values in the morning; thesetraits, as well as the Ci/Ca, did not differ significantly regardless ofthe treatments (Fig. 5).

Irrespective of the PAR treatments and the time of the day, nosigns of photoinhibition were observed as the Fv/Fm ratio remainedhigh (∼0.80). Overall, the qP and ˚PSII tended to decrease and theNPQ tended to increase with the increasing PAR availability in themorning (Fig. 5). The ETR was markedly lower in the S-10 and S-10/100 plants than in the other plants (Fig. 5). These trends werealso observed in the afternoon (Fig. 5). Notably, the highest absolutevalues of NPQ in the afternoon were observed in the plants thatwere transferred from low to high PAR at midday (S-40/100 andS-10/100 individuals).

The largest Amax values were observed in the S-100/40 andS-100/10 individuals, the lowest were in the S-40 and S-10 indi-viduals and the intermediate values were observed for the plantsfrom the other PAR treatments (Table 2). The largest LSP valueswere obtained in the plants that were subjected to full sunlight inthe morning compared to their counterparts that were submittedto some degree of shade. The LCP decreased with the decreas-ing PAR supply in plants grown at a fixed level of PAR, althoughthis trait did not vary consistently among the plants that weresubjected to the varying PAR supplies (Table 2). In general, thechanges in the LCP paralleled those in Rd (r2 = 0.79, P < 0.01, n = 35).The Vcmax was unresponsive to the PAR, whereas the Jmax wassignificantly lower in the S-10 plants than in the plants fromthe other PAR treatments, which did not differ from one another(Table 2).

3.4. Biochemical traits

Overall, there were some diurnal differences in hexoses (fruc-tose plus glucose) and starch among the treatments (Fig. 6),whereas sucrose pools were virtually unresponsive to the PAR regi-

mens regardless of the time of samplings. Importantly, the totalnon-structural carbohydrates (represented by the sum of hexoses,sucrose and starch pools) did not differ significantly amongst treat-ments, with the exception of measurements conducted at 1100 h,

urves [light-saturated A (Amax; �mol CO2 m−2 s−1), light compensation point (LCP;ration (Rd; �mol CO2 m−2 s−1)], as well as from the A and internal CO2 concentrationm−2 s−1) and maximum rate of carboxylation (Vcmax; �mol CO2 m−2 s−1)] in coffeeg. 1. Means followed by the same letter do not differ significantly from one another

S-40/100 S-100/40 S-10/100 S-100/10

5.4 ± 0.3bc 7.2 ± 0.3a 5.6 ± 0.3b 7.2 ± 0.2a

9.3 ± 0.5ab 9.4 ± 0.8ab 10.4 ± 1.3a 4.8 ± 0.6cd

288 ± 14b 376 ± 25a 282 ± 23b 415 ± 15a

0.53 ± 0.07ab 0.71 ± 0.12a 0.57 ± 0.11ab 0.41 ± 0.12ab

88.8 ± 1.8a 87.8 ± 3.3a 98.0 ± 1.9a 93.3 ± 1.9a

60.2 ± 3.0a 59.3 ± 6.4a 52.3 ± 0.5a 49.7 ± 1.4a

) with the ratios of Chl/Car, Chl a/b and Chl/N (mmol mol−1) in coffee plants thats followed by the same letter do not differ significantly from one another (P < 0.05,

S-40/100 S-100/40 S-10/100 S-100/10

9.7 ± 0.3b 9.9 ± 0.6b 10.9 ± 0.6b 10.7 ± 0.8b

2.0 ± 0.1b 2.2 ± 0.1b 2.3 ± 0.1b 2.3 ± 0.1b

4.8 ± 0.1a 4.6 ± 0.1a 4.7 ± 0.1a 4.6 ± 0.2a

2.9 ± 0.1a 3.0 ± 0.1a 2.9 ± 0.1a 2.9 ± 0.1a

2.8 ± 0.1b 2.9 ± 0.2b 3.2 ± 0.2b 3.2 ± 0.2b

N.F. Rodríguez-López et al. / Environmental and Experimental Botany 102 (2014) 1–10 7

Fig. 5. Gas exchange parameters [net CO2 assimilation rate (A), stomatal conductance (gs), internal-to-ambient CO2 concentration ratio (Ci/Ca; �mol CO2 mol−1 air)] andchlorophyll a fluorescence parameters [variable-to-maximum fluorescence ratio (Fv/Fm), photochemical quenching coefficient (qP), non-photochemical quenching coefficient(NPQ), quantum yield of PSII electron transport (˚PSII) and electron transport rate (ETR)] as measured at two time points (800–1000 h and 1400–1600 h) in coffee plantss in eaca

wsp

i

ubjected to seven light treatments (see further details in the legend of Fig. 1). Withnother (P < 0.05, Newman–Keuls test). n = 5 ± SE.

here the total non-structural carbohydrate concentrations were

ignificantly lower in the S-10 and S-10/100 individuals when com-ared with their S-100/10 counterparts (Fig. 6).

The concentrations of Chl and Car and the Chl/N ratio were largern the S-10 plants than in the other PAR treatments when assessed

h time point, means followed by the same letter do not differ significantly from one

before midday (Table 3). The Chl/Car and Chl a/b ratios were unre-

sponsive to the PAR supply (Table 3). No signs of Chl degradationwere detected regardless of plant transfer from low to high PARenvironments (and vice versa) as assessed at the end of the day(data not shown).

8 N.F. Rodríguez-López et al. / Environmental an

Fig. 6. The total leaf non-structural carbohydrate concentrations (g kg−1 DW), rep-resented by the sum of starch, hexoses (glucose + fructose) and sucrose pools asmeasured at three time points (0500, 1100 and 1700 h) in coffee plants subjected toseven different light treatments (see further details in the legend of Fig. 1). Meansfollowed by the same capital letter (total non-structural carbohydrates) or by theso

4

aenuseobtoaea

the positive correlation between A (measured at 0800–1000 h) and

ame lowercase letter (starch, hexoses and sucrose) do not differ significantly fromne another (P < 0.05, Newman–Keuls test). n = 5.

. Discussion

Natural shade involves both decreases in light quantity andlterations in light quality (Smith, 1982). Here, we were able toxamine the effects of light amounts (and temporal scale of diur-al light availability) on biomass accumulation and related traits bysing neutral black nettings which have previously been demon-trated to cause no significant effect on light quality (e.g., Shahakt al., 2004; Griffith and Sultan, 2005), a fact that was further corrob-rated in this study. Despite our light quality assessments havingeen confined to a few days, it should be emphasised that the spec-rum of the global radiation is remarkably constant irrespectivef the season and weather conditions at a given latitude as long

s the sun is more than 10◦ above the horizon (Smith, 1982). Asxpected, we found alterations in the light quality in both dawnnd dusk (spectra enriched in blue and far-red light), although

d Experimental Botany 102 (2014) 1–10

these alterations took place regardless of the nettings that wereused. Therefore, we contend that differences in the growth traitsin response to the light treatments that appeared in the long termwere largely unrelated to light quality.

The biomass offers the most direct measure of plant perfor-mance as a product of growth (Dawson et al., 2012); therefore,changes in biomass could be translated into the clearest indicatorof a plant’s ability to respond to and take advantage of vary-ing resource (light) availability. In agreement with our workinghypothesis, we herein demonstrated that the ability of the coffeeplant to accumulate biomass depended on the total PAR, thoughthe amount of the PAR that was received by the plants during themorning played a more significant role than that received duringthe afternoon in driving the biomass accumulation, a fact proba-bly related to improved gas exchanges throughout the morning.Furthermore, we also demonstrated that whole-plant physiolog-ical adjustments could largely account for the differences in thebiomass, particularly in the plants subjected to changing PAR sup-plies.

4.1. Morphology or physiology: which was the main driver toexplain the changes in biomass over a range of light regimens?

When comparing the S-100, S-40 and S-10 individuals, theimproved biomass accumulation in the high PAR values was accom-panied by an increased leaf number (and total leaf area), whichhas also been noted in both herbaceous (e.g., Cookson and Granier,2006) and woody (e.g., Farque et al., 2001) species. In our case,the increased leaf number appeared to be largely related to ear-lier plagiotropic branch development (higher PMF), which impliesthat there were more nodes for anchoring the larger number ofleaves that are produced under high PAR conditions. Neverthe-less, the marked decreases in the morphological parameter LARwith the increasing PAR imply that the photosynthetic return perunit of resources invested in the construction of photosynthetictissues is likely to rise with the increasing PAR. This sugges-tion is consistent with the parallelism between total biomass andA (measured at 0800–1000 h) (cf. Figs. 3 and 5). Therefore, itis tempting to propose that physiological adjustments (e.g., netassimilation rate (NAR)) would play a major role in explaining theobserved differences in biomass for the plants grown at a fixed PARregimen.

Both the S-10/100 and S-100/10 individuals received similartotal diurnal PAR supplies; however, the larger biomass of the S-100/10 plants could reflect both an increased leaf number (and totalleaf area) and improved A per unit leaf area, which, in turn, couldbe associated with the higher PAR in the morning when the envi-ronmental conditions are more conducive for higher gas exchangerates in coffee (Batista et al., 2012). However, the differences incarbohydrate pools between the S-10/100 and S-100/10 individ-uals were also minimal, possibly reflecting a consequence ratherthan a cause of the depressed growth in the S-10/100 plants. Irre-spective, we noted that the morphological traits such as SLA andLMF (and hence LAR) were virtually unresponsive to the temporalscales of PAR supply when comparing the S-10/100 with S-100/10individuals (also when comparing the S-40/100 individuals withtheir S-100/40 counterparts). Therefore, we contend that physio-logical adjustments, such as NAR, could largely be accounted forby the differences in growth we found not only at fixed but also atvarying PAR regimens. In line with this suggestion, and consideringthat NAR is often strongly associated with A (Poorter et al., 2013),

total biomass (r2 = 0.71, P = 0.017, n = 8) further highlights the roleplayed by NAR in determining the differences in biomass accumu-lation across the light treatments.

tal and

4

aiabcfelItttldfms(Twoetdre

ttbagat1sbIrttt(wgaeeshoiae2iv

pa(tp

N.F. Rodríguez-López et al. / Environmen

.2. Photosynthetic performance and acclimation to varying PAR

In the morning, the differences in the actual A were largelyssociated with stomatal factors, although the photochemical lim-tations likely played a central role in constraining A in the S-10nd S-10/100 individuals as determined by their low ETR andecause their PAR supplies were at a level below their LSP. Inontrast, the significantly higher LSP noted in the plants receivingull sunlight in the morning should reflect a greater potential fornhanced carbon gains during that period, which may be trans-ated into an improved NAR and consequential higher biomass.n the afternoon, the strong decreases in gs imposed large limi-ations on A; however, because Ci/Ca did not decrease relative toheir morning values, an additional diffusive (mesophyll) limita-ion or some impairment to the CO2 fixation at the chloroplastevel cannot be ruled out (Flexas et al., 2007). In any case, theecreases in A in the afternoon were apparently unrelated to theeedback regulation by carbohydrate accumulation, which changed

inimally across treatments, or the photoinhibition of photo-ynthesis, which is in agreement with previous studies in coffeeAraújo et al., 2008; DaMatta et al., 2008; Batista et al., 2012).he pronounced increases in the NPQ, which has been associatedith large zeaxanthin pools and a higher de-epoxidation state

f the xanthophyll cycle in coffee (Matos et al., 2009; Pompellit al., 2010) upon transfer from low to high PAR coupled withhe elevated Fv/Fm values, suggest that coffee plants possess well-eveloped photoprotection mechanisms to cope with excessiveadiation loads (Chaves et al., 2008; Moraes et al., 2010; Pompellit al., 2010).

Overall, we noted some differences in the Amax but not in eitherhe Vcmax or Jmax among treatments. The unchanging pattern ofhe Vcmax is intriguing, given that a considerable investment woulde required not only to build but also to maintain relatively highmounts of active Rubisco (Millard et al., 2007). These findings sug-est that a relatively large investment in the capacity for carbonssimilation is retained even when the light supply is limiting, orhat significant diffusive limitations may exist (the case for the S-0 plants as evidenced by their low Amax). These findings mightuggest a wasteful use of resources given the low realisable Ay the leaves at the low PAR supply, especially in the morning.

ndeed, only minor physiological acclimations were observed inesponse to the varying PAR supplies in this study. Such differen-ial acclimations were mainly evident when comparing the extremereatments, i.e., the S-10 individuals with the plants from the otherreatments. These results corroborate the findings of Matos et al.2009) who also noted considerable physiological plasticity onlyhen comparing leaves in the extremes of a wide natural light

radient (i.e., diurnal intercepted irradiance of leaves differing by factor of 25 between the deepest shade leaves and the morexposed leaves in the canopy of coffee trees growing under fullxposure). Here, we showed that the S-10 individuals displayedome attributes of shade-acclimated leaves, such as a low Rd andigh SLA, Chl and Car pools, as well as a high Chl/N ratio, with allf these traits concurring for the lower values of LCP of the S-10ndividuals; collectively, these traits may improve light harvestingnd use such that a positive carbon balance may be maintainedven under the low PAR conditions (Walters, 2005; Lusk et al.,008). In summary, our results indicate a poor leaf-level plastic-

ty of physiological traits to cope with the temporal scales of thearying PAR.

Independently of the PAR treatments, the relatively low LSP cou-led with the low A suggests, at first glance, that coffee has a limited

bility to use the extra light (on a leaf area basis) at full exposureMatos et al., 2009). Nonetheless, to reconcile these results withhe findings that biomass depended strongly on the total PAR sup-ly, the LSP on a whole-plant level must be remarkably higher than

Experimental Botany 102 (2014) 1–10 9

that on a leaf-scale level. We therefore propose that, regardless ofthe PAR regimen, the coffee plant, even at the seedling stage, pos-sesses the ability to take advantage of full sunlight, and this abilityis likely to increase as the plant grows and its self-shading increasesaccordingly as the canopy becomes more closed.

4.3. Concluding remarks: could the coffee plant be considered ashade-demanding species?

When comparing the S-100, S-40 and S-10 individuals, thebehaviour of the morphological traits with the decreasing light sup-ply is clearly consistent with the shade avoidance acclimations (Yuand Ong, 2003; Franklin and Whitelam, 2005; Forster et al., 2011):increased LAR and SLA, increased allocation to orthotropic stems,delayed branching (less allocation to plagiotropic branches) anddecreased allocation to roots. Despite the plant height being lowerin the S-10 plants, its height-to-stem diameter ratio (the slender-ness index) was higher than in the plants from the other treatments(data not shown), once again highlighting a shading avoidancestrategy (Yagi, 2009) of the coffee seedlings. Furthermore, the cof-fee plant apparently is well protected against high irradiances andpotential photodamages. Considering this information togetherwith the fact that the biomass acted as a strong function of thelight supply, it follows that the coffee plant would behave as a clas-sic shade-avoiding species. However, all of these plastic responsesto the PAR supply were apparently independent on light quality,in contrast to what has been seen in true shade-avoiding speciesin which changes in blue light and red/far red ratios are known toplay important roles in morphogenetic processes such as leaf for-mation and leaf morphology (Franklin, 2008). Ultimately, changesin growth responses associated with diminished PAR levels in theabsence of spectral alterations suggest a behaviour more consis-tent with shade-tolerant than shade-avoiding species (Griffith andSultan, 2005). Furthermore, the poor response of physiologicaltraits to the PAR in both seedlings (this study) and adult trees(Araújo et al., 2008; Matos et al., 2009), especially those traitsinvolved with the maximisation of the photosynthetic carbon gain,characterises a weak leaf-level physiological plasticity that is typ-ical of shade-tolerant species (Valladares and Niinemets, 2008).Therefore, in line with the suggestion of Matos et al. (2009) basedon leaf-level traits, it is tempting to consider the modern coffeecultivars behave as shade-tolerant as their traits do not match therequirements of either a classic shade-avoiding or a classic shade-demanding species. These conclusions may help to reconcile someapparent dichotomous responses to the light supply, such as thestrong dependence of the biomass accumulation on the total PARavailability and the successful survival of coffee plants in the under-stories not only in the coffee’s native habitats but also in secondaryforests developed in abandoned areas previously cropped with cof-fee, as noted through empirical observations.

Acknowledgments

Scholarships were granted by the Ministry of Foreign Rela-tions (NFR-L), the National Council for Scientific and TechnologicalDevelopment, CNPq (SCVM, PCC, AOL) and the Brazilian FederalAgency for Support and Evaluation of Graduate Education (PEMS,LEM, AGG), all from the Brazilian Government, and are grate-fully acknowledged. Research fellowships granted by CNPq andFAPEMIG to FMD are also acknowledged.

References

Ackerly, D., 1997. Allocation, leaf display and growth in fluctuating light environ-ments. In: Bazzaz, F.A., Grace, J. (Eds.), Plant Resource Allocation. Academic Press,New York, pp. 231–264.

1 tal an

A

A

B

B

C

C

C

D

D

D

D

D

D

F

F

F

F

FF

G

L

L

0 N.F. Rodríguez-López et al. / Environmen

ntunes, W.C., Pompelli, M.F., Carretero, D.M., DaMatta, F.M., 2008. Allometric mod-els for non-destructive leaf area estimation in coffee (Coffea arabica and Coffeacanephora). Ann. Appl. Biol. 153, 33–40.

raújo, W.L., Dias, P.C., Moraes, G.A.B.K., Celin, E.F., Cunha, R.L., Barros, R.S., DaMatta,F.M., 2008. Limitations to photosynthesis in coffee leaves from different canopypositions. Plant Physiol. Biochem. 46, 884–890.

atista, K.D., Araújo, W.L., Antunes, W.C., Cavatte, P.C., Moraes, G.A.B.K., Martins,S.C.V., DaMatta, F.M., 2012. Photosynthetic limitations in coffee plants are chieflygoverned by diffusive factors. Trees 26, 459–468.

oardman, N.K., 1977. Comparative photosynthesis of sun and shade plants. Annu.Rev. Plant Physiol. 28, 355–377.

avatte, P.C., Oliveira, A.A.G., Morais, L.E., Martins, S.C.V., Sanglard, L.M.V.P., DaMatta,F.M., 2012. Could shading reduce the negative impacts of drought on coffee? Amorpho-physiological analysis. Physiol. Plant. 144, 111–122.

haves, A.R.M., Caten, A.T., Pinheiro, H.A., Ribeiro, A., DaMatta, F.M., 2008. Seasonalchanges in photoprotective mechanisms of leaves from shaded and unshadedfield-grown coffee (Coffea arabica L.) trees. Trees 22, 351–361.

ookson, S.J., Granier, C., 2006. A dynamic analysis of the shade-induced plasticityin Arabidopsis thaliana rosette leaf development reveals new components of theshade-adaptative response. Ann. Bot. 97, 443–452.

aMatta, F.M., Amaral, J.A.T., Rena, A.B., 1999. Growth periodicity in trees of Coffeaarabica L. in relation to nitrogen supply and nitrate reductase activity. Field CropsRes. 60, 223–229.

aMatta, F.M., 2004. Ecophysiological constraints on the production of shaded andunshaded coffee: a review. Field Crops Res. 86, 99–114.

aMatta, F.M., Cunha, R.L., Antunes, W.C., Martins, S.C.V., Araújo, W.L., Fernie, A.R.,Moraes, G.A.B.K., 2008. In field-grown coffee trees source-sink manipulationalters photosynthetic rates, independently of carbon metabolism, via alterationsin stomatal function. N. Phytol. 178, 348–357.

aMatta, F.M., Ronchi, C.P., Maestri, M., Barros, R.S., 2010. Coffee: environment andcrop physiology. In: DaMatta, F.M. (Ed.), Ecophysiology of Tropical Tree Crops.Nova Science Publishers, New York, pp. 181–216.

awson, W., Rohr, R.P., van Kleunen, M., Fischer, M., 2012. Alien plant species witha wider global distribution are better able to capitalize on increased resourceavailability. N. Phytol. 194, 859–867.

ias, P.C., Araújo, W.L., Moraes, G.A.B.K., Barros, R.S., DaMatta, F.M., 2007. Mor-phological and physiological responses of two coffee progenies to soil wateravailability. J. Plant Physiol. 164, 1639–1647.

arque, L., Sinoquet, H., Colin, F., 2001. Canopy structure and light interception inQuercus petraea seedlings in relation to light regime and plant density. TreePhysiol. 21, 1257–1267.

arquhar, G.D., von Caemmerer, S., Berry, J.A., 1980. A biochemical model ofphotosynthesis CO2 assimilation in leaves of C3 plants species. Planta 149,78–90.

lexas, J., Díaz-Espejo, A., Berry, J.A., Cifre, J., Galmes, J., Kaldenhoff, R., Medrano, H.,Ribas-Carbó, M., 2007. Analysis of leakage in IRGA’s leaf chambers of open gasexchange systems: quantification and its effects in photosynthesis parameteri-zation. J. Exp. Bot. 58, 1533–1543.

orster, K.A., Ladd, B., Bonsert, S., 2011. Optimal allocation of resources in responseto shading and neighbours in the heteroblastic species, Acacia implexa. Ann. Bot.107, 219–228.

ranklin, K.A., 2008. Shade avoidance. N. Phytol. 179, 930–944.ranklin, K.A., Whitelam, G.C., 2005. Phytochromes and shade-avoidance responses

in plants. Ann. Bot. 96, 169–175.riffith, T.M., Sultan, S.E., 2005. Shade tolerance plasticity in response to neutral

vs green shade cues in Polygonum species of contrasting ecological breadth. N.Phytol. 166, 141–148.

ichthenthaler, H.K., 1987. Chlorophylls and carotenoids: pigments of photosyn-thetic biomembranes. Methods Enzymol. 148, 350–382.

usk, C., Reich, P.B., Montgomery, R.A., Ackerly, D.D., Cavender-Bares, J., 2008.Why are evergreen leaves so contrary about shade? Trends Ecol. Evol. 23,299–303.

d Experimental Botany 102 (2014) 1–10

Maxwell, K., Johnson, G.N., 2000. Chlorophyll fluorescence – a practical guide. J. Exp.Bot. 51, 659–668.

Matos, F.S., Wolfgramm, R., Gonc alves, F.V., Cavatte, P.C., Ventrella, M.C., DaMatta,F.M., 2009. Phenotypic plasticity in response to light in the coffee tree. Environ.Exp. Bot. 67, 421–427.

Millard, P., Sommerkorn, M., Grelet, G.-A., 2007. Environmental change and carbonlimitation in trees: a biochemical, ecophysiological and ecosystem appraisal. N.Phytol. 175, 11–28.

Moraes, G.A.B.K., Chaves, A.R.M., Martins, S.C.V., Barros, R.S., DaMatta, F.M., 2010.Why is it better to produce coffee seedlings in full sunlight than in the shade?A morphophysiological approach. Photosynthetica 48, 199–207.

Niinemets, Ü., 2007. Photosynthesis and resource distribution through plantcanopies. Plant Cell Environ. 30, 1052–1071.

Niinemets, Ü., 2010. A review of light interception in plant stands from leaf to canopyin different plant functional types and in species with varying shade tolerance.Ecol. Res. 25, 693–714.

Oxborough, K., Baker, N.R., 1997. Resolving chlorophyll a fluorescence images ofphotosynthetic efficiency into photochemical and non-photochemical compo-nents – calculation of qP and Fv

′/Fm′ without measuring F0

′ . Photosynth. Res. 54,135–142.

Pompelli, M.F., Martins, S.C.V., Antunes, W.C., Chaves, A.R.M., DaMatta, F.M., 2010.Photosynthesis and photoprotection in coffee leaves is affected by nitrogen andlight availabilities in winter conditions. J. Plant Physiol. 167, 1052–1060.

Poorter, H., Niklas, K.J., Reich, P.B., Oleksyn, J., Poot, P., Mommer, L., 2012. Biomassallocation to leaves stems and roots: meta-analyses of interspecific variationand environmental control. N. Phytol. 193, 30–50.

Poorter, H., Anten, N., Marcelis, L.F.M., 2013. Physiological mechanisms in plantgrowth models: do we need a supra-cellular systems biology approach? PlantCell Environ. 35, http://dx.doi.org/10.1111/pce.12123.

Praxedes, S.C., DaMatta, F.M., Loureiro, M.E., Ferrão, M.A.G., Cordeiro, A.T.,2006. Effects of long-term soil drought on photosynthesis and carbohydratemetabolism in mature robusta coffee (Coffea canephora Pierre var. kouillou)leaves. Environ. Exp. Bot. 56, 263–273.

Ronchi, C.P., DaMatta, F.M., Batista, K.D., Moraes, G.A.B.K., Loureiro, M.E., Ducatti, C.,2006. Growth and photosynthetic down-regulation in Coffea arabica in responseto restricting root volume. Funct. Plant Biol. 33, 1013–1023.

SAEG, 2007. Sistema de Análises Estatísticas e Genéticas: versão 9.1. Fundac ãoArthur Bernardes, Vic osa.

Shahak, Y., Gussakovsky, E.E., Cohen, Y., Lurie, S., Stern, R., Kfir, S., Naor, A., Atz-mon, I., Doron, I., Greenblat-Avron, Y., 2004. ColorNets: a new approach for lightmanipulation in fruit trees. Acta Hort. 636, 609–616.

Sharkey, T.D., 1985. Photosynthesis in intact leaves of C3 plants: physics, physiologyand rate limitations. Bot. Rev. 78, 71–75.

Smith, H., 1982. Light quality, photoperception, and plant strategy. Annu. Rev. PlantPhysiol. 33, 481–518.

Valladares, F., Niinemets, Ü., 2008. Shade tolerance, a key plant feature of complexnature and consequences. Annu. Rev. Ecol. Evol. Syst. 39, 237–257.

Walters, R.G., 2005. Towards an understanding of photosynthetic acclimation. J. Exp.Bot. 56, 435–447.

Watling, J.R., Ball, M.C., Woodrow, I.E., 1997. The utilization of lightflecks for growthin four Australian rain-forest species. Funct. Biol. 11, 231–239.

Way, D.A., Pearcy, R.W., 2012. Sunflecks in trees and forests: from photosyntheticphysiology to global change biology. Tree Physiol. 32, 1066–1081.

Wayne, P., Bazzaz, F.A., 1993. Birch seedling responses to daily time courses of lightin experimental forest gaps and shadehouses. Ecology 74, 1500–1515.

Yagi, T., 2009. Ontogenetic strategy shift in sapling architecture of Fagus crenata inthe dense understorey vegetation of canopy gaps created by selective cutting.

Can. J. Forest Res. 39, 1186–1196.

Ye, Z., 2007. A new model for relationship between irradiance and the rate of pho-tosynthesis in Oryza sativa. Photosynthetica 45, 637–640.

Yu, H., Ong, B.L., 2003. Effect of radiation quality on growth and photosynthesis ofAcacia mangium seedlings. Photosynthetica 41, 349–355.