AOBPreview originally published online on January 24, 2008
Annals of Botany 2008 101(8):1139-1151; doi:10.1093/aob/mcm300
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Using a 3-D Virtual Sunflower to Simulate Light Capture at Organ, Plant and Plot Levels: Contribution of Organ Interception, Impact of Heliotropism and Analysis of Genotypic Differences
1 CIRAD, UMR AMAP (botAnique et bioinforMatique de l'Architecture des Plantes), Montpellier, F-34398, France
2 INRA, UMR759, 2 place Viala, Montpellier, F-34060, France
3 Unidad Integrada (FCA – UNMdp / EEA – INTA) Balcarce, CC 276, 7620 Balcarce, Argentina
4 Montpellier SupAgro, UMR759, 2 place Viala, Montpellier, F-34060, France
* For correspondence. E-mail jeremie.lecoeur{at}supagro.inra.fr
Received: 23 February 2007 Returned for revision: 14 June 2007 Accepted: 25 October 2007 Published electronically: 24 January 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Background and Aims: Light interception is a critical factor in the production of biomass. The study presented here describes a method used to take account of architectural changes over time in sunflower and to estimate absorbed light at the organ level.
Methods: The amount of photosynthetically active radiation absorbed by a plant is estimated on a daily or hourly basis through precise characterization of the light environment and three-dimensional virtual plants built using AMAP software. Several treatments are performed over four experiments and on two genotypes to test the model, quantify the contribution of different organs to light interception and evaluate the impact of heliotropism.
Key Results: This approach is used to simulate the amount of light absorbed at organ and plant scales from crop emergence to maturity. Blades and capitula were the major contributors to light interception, whereas that by petioles and stem was negligible. Light regimen simulations showed that heliotropism decreased the cumulated light intercepted at the plant scale by close to 2·2 % over one day.
Conclusions: The approach is useful in characterizing the light environment of organs and the whole plant, especially for studies on heterogeneous canopies or for quantifying genotypic or environmental impacts on plant architecture, where conventional approaches are ineffective. This model paves the way to analyses of genotype–environment interactions and could help establish new selection criteria based on architectural improvement, enhancing plant light interception.
Key words: 3-D virtual plant, light interception, plant architecture, Helianthus annuus, sunflower, heliotropism, organ irradiance, radiative balance
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Intercepted light is an important component for plant energy balance and plant biomass accumulation (Monteith, 1977; Jones, 1992). Higher plants display a high degree of plasticity in their morphological responses to light (e.g. Slade and Hutchings, 1987; Granier and Tardieu, 1999). These complex interactions between plants and their light environment make it difficult to estimate the amount of light intercepted at the plant scale. This problem may be circumnavigated by considering the canopy as a continuous medium. The amount of intercepted radiation can then be estimated by the Beer–Lambert law using the leaf area index (LAI) and an extinction coefficient, k, specific for the cover (e.g. Varlet-Grancher et al., 1989). This approach is appropriate for a developed canopy and has been extensively used in crop management to estimate light absorption by the canopy or organs. However, this classical approach is not designed to study heterogeneous or sparse canopies such as crops in the early stages of growth (Chenu et al., 2007b), orchards or vineyards (Louarn et al., 2007), rosette plants (Chenu et al., 2005), or isolated plants used in research trials which may be grown in pots. Furthermore, this classical approach cannot be used for artificial environments such as greenhouses or growth chambers (Chenu et al., 2005) where artificial lighting, walls and supporting structures make the incident radiation heterogeneous. Finally, this classical approach is not designed to analyse the impact of architectural changes over time, especially for short time steps such as heliotropism.
An alternative approach is to consider the crop as a population of individual plants and calculate the light microclimate for individual plant organs. Over the past few decades, computer programs have been developed to model, simulate and display the three-dimensional (3-D) architecture of plants (e.g. Jaeger and de Reffye, 1992; Prusinkiewicz, 1998). 3-D virtual plants (Barczi et al., 2008) can be used to estimate plant–environment interactions (Chelle, 2005), for example in terms of radiative balance (e.g. Dauzat and Eroy, 1997; Chelle and Andrieu, 1998, 1999) or energy balance (Rapidel et al., 1999).
The aim of the present study was to couple and adapt the AMAPsim architectural model (Barczi et al., 2008) and MMR radiative calculation software (Dauzat and Eroy, 1997; Dauzat et al., 2001, 2008) to simulate the amount of radiation absorbed by a 3-D virtual sunflower plant at organ, plant and plot scales over time, from plant emergence until maturity. Organ development and expansion were estimated over thermal time from fitting performed on measured data. This approach was tested in a field experiment on two genotypes of contrasted architectures. Simulations were also performed under a variety of environmental conditions, in the field and in greenhouses, with different densities, shading treatments, and isolated and/or decapitulated plants to analyse the effects of various plant architectures on light interception. The model developed was also used to quantify the contribution made by the different organs to light interception and the impact of heliotropism on the light intercepted by 3-D virtual plants.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Greenhouse and field experiments were carried out to (1) parameterize the AMAPsim model used to simulate sunflower growth and architecture, (2) adapt the MMR model to greenhouse conditions and (3) test the AMAPsim and MMR combination in such situations. 3-D virtual plants were built on the basis of measurements made for each experiment x treatment combination. The light environment for each treatment was measured for use in MMR. A combination of these two models was used to simulate the light balance at organ level over time for plants of different architectures (different genotypes, light treatments, densities and decapitulated plants)
Experimental design and culture conditions
Sunflower (Helianthus annuus L., hybrid ‘Albena’) plants were grown in Montpellier, southern France (43°40'N, 3°50'E), either in greenhouses (E–W orientation) or in the field during five experiments conducted in 1998, 1999 and 2001 (Table 1).
|
In the greenhouse experiments, seeds were sown at a depth of 0·03 m in columns (0·14 m diameter and 0·65 m height) consisting of a 1: 1 mixture (v/v) of loamy soil and organic compost. Soil was maintained at water retention capacity by watering columns three times a day with a modified, one-tenth-strength Hoagland solution corrected with minor nutrients. Additional light was provided by a bank of sodium lamps, maintaining an average photoperiod of 14 h.
In the field experiments, seeds were sown at a depth of 0·02 m in a deep sandy loam soil (fluvio-calcaric Cambissol) with 0·6 m between rows and 0·3 m between seeds in each row. Plants were grown in rows orientated N–S, with individual plots for each treatment ranging from 78 to 400 m2. Nitrogen fertilization was applied before sowing (80 kg ha–1). Soil water potential was monitored by means of tensiometers (DTE 1000 system, Nardeux, Saint-Avertin, France) and maintained by irrigation above –40 kPa in the top 0·5 m of soil, in a range for which neither stomatal conductance nor leaf growth was affected (Sadras and Milroy, 1996).
Four treatments were applied and two genotypes tested (Table 1). The ‘control’ treatment (Gh-Feb98-1, Gh-Nov98-1, Fd-May99-1 and Fd-May-01-1) plants were not subject to any shading or architectural manipulations. The second treatment (Fd-May98-2, Fd-May99-2) corresponded to isolated plants obtained by thinning after seedling emergence to avoid light competition between neighbouring plants. In the third treatment (Fd-May-99-31 and Fd-May-99-32), the plants in two plots were subject to shading that intercepted 75 % of the incident photosynthetically active radiation (PAR). The first plot (Fd-May-99-31) was shaded between 220 and 420 °Cd after plant emergence, i.e. the period corresponding to leaf development on the lower part of the main stem. The second plot (Fd-May-99-32) was shaded between 420 and 580 °Cd, during leaf establishment on the middle of the main stem. In the fourth treatment (Gh-Feb98-4, Gh-Nov98-4 and Fd-May98-4), the plants were decapitulated on appearance of the bud flower in order to suppress the reproductive sink and prevent the growing capitulum from shading the upper leaves. Finally, two different hybrids (‘Albena’: Fd-May01-1A and ‘Heliasol’: Fd-May01-1H) were grown under control conditions to analyse the effect of their contrasted architectures.
Environmental conditions and measurements
Environmental factors were measured in the middle of the trials. Incident, photosynthetically active photon flux was measured continuously above the canopy using a PAR quantum sensor (LI-190SB, LI-COR, Lincoln, NB, USA). Air temperature and relative humidity were measured 2 m above the soil surface using an HMP35 capacitive hygrometer (Vaisala Oy, Helsinki, Finland) placed in a ventilated shelter. Leaf temperature was measured using 0·4-mm-diameter copper–constantan thermocouples. Thermal time measurements for the different organs were measured during their development by means of five thermocouples placed in the apical bud and on the underside of the blade of both young and mature leaves on five plants used for non-destructive measurements in each treatment. Air and leaf temperature, incident photosynthetic photon flux density (PPFD) and relative humidity were measured every 10 s then averaged and stored every 600 s in a data logger (LTD-CR10 Wiring panel, Campbell Scientific, Shepshed, Leicestershire, UK). Over the 70 d following plant emergence, mean daily air temperature ranged from 19·3 to 22·7 °C for the different experimental situations, mean daily leaf temperature ranged from 18·8 to 22·4 °C and mean daily cumulated incident PAR from 8 to 55·8 mol m–2 d–1. Cumulated incident PAR over the plant cycle was three times higher in field summer experiments than in greenhouse winter experiments (Table 2).
|
Directional light fluxes were measured in the field and greenhouse using a ‘Turtle’ device composed of 16 or six PAR sensors facing different zenithal and azimuthal directions (Chenu et al., 2005). Because these measurements were not taken continually throughout the experiments, a procedure was developed to estimate directional light fluxes from incident light measured outside the greenhouse and artificial lighting by lamps. Artificial light was simply measured during the night using the Turtle device and applied over the duration of lighting. The natural light locally available within the greenhouse was related to the external directional radiation through an extinction coefficient accounting for light interception by greenhouse superstructures. Measurements were taken at the start of each experiment at soil level and 1 m high, on a 1 x 1-m grid.
The fraction of intercepted PAR (see eqn 3) was measured twice weekly in field experiment Fd-May01 both on both ‘Albena’ and ‘Heliasol’ using PCA sensors (LI-COR LAI-2000) positioned above and below the canopy.
Plant architecture and growth measurements
In all, eight to ten median plants were selected in each plot on the basis of stem diameter and plant height for non-destructive measurements. Destructive samples were taken from plants showing similar development and fitness.
Plant emergence was defined as the date on which the first leaf pair became visible (3–5 mm in length) between open cotyledons, i.e. around 150 °Cd (about 7 d) after seed sowing. Four to six plants were harvested three times a week from plant emergence to capitulum initiation to measure leaves enclosed in the apical bud. This was dissected under a microscope (Leica MZ75, Wetzlar, Germany) and all initiated phytomers (about 40 µm long) were counted. The areas of all leaves in the bud were measured using a camera coupled to an interactive image analysis system (Optimas V6·5, Media Cybernetics, Silver Spring, MD, USA).
Non-destructive measurements were made three times weekly from plant emergence to the time at which all organs ceased to expand (around 70 d after plant emergence). Plant height, stem diameter, internode and petiole length and diameter, and blade length and width (length > 1 mm) were measured with a ruler and digital callipers (0·1 mm) on the reference plants. Angles of phyllotaxy, petiole insertion on the stem, petiole and blade flexion and rotation were measured using a protractor (Pro 360, Travers, NY, USA). Blade shape was studied to ensure that there was no difference with position on the stem, and that blade area could be estimated from measurements of only length and width. A highly significant linear relationship (r2 > 0·99, n > 300, P < 0·001) was established in each experiment between the product of length x width and previously measured blade area.
Heliotropic movements were measured in the Fd-May-99-1 experiment every hour over the diurnal photophase using three cameras placed in a 3-D device. The plant development stage was just after initiation of the capitulum (which was 3 cm in diameter), when the stem had not yet completely elongated and had not completely stiffened through lignification (enabling torsion and bending movements), and when the cover had not yet closed. Organ displacements were then quantified (torsion and bending of stems, petioles and blades) by image analysis (Optimas V6·5, Media Cybernetics).
Developmental and growth equations
Daily thermal time was calculated by daily integration of leaf temperature minus a basis temperature of 4·8 °C (Granier and Tardieu, 1998). Daily thermal time was cumulated either from plant emergence (plant thermal time, PTT) or from organ initiation (organ thermal time, OTT). For a given organ, the calculation was made using (1) apex temperature from organ initiation to its emergence, (2) growing-zone temperature from organ emergence to the end of its expansion, and (3) mature leaf temperature thereafter. Plant thermal time was calculated as the average of the different organ temperatures.
Phytomer initiation rate (IR) was calculated from the number of phytomers initiated (N) and PTT:
|
| (1) |
|
| (2) |
Blade areas were calculated throughout their growth from length and width fits (eqn 2), as previously described.
3-D virtual plants of Helianthus annuus
Measurements of architecture, topology, morphology and phenology were used to construct a 3-D virtual sunflower using the AMAPsim program (Barczi et al., 2008). The field-grown sunflower genotype had a single reference axis that was constructed by considering four main classes of physiological age values (or ‘developmental steps’), namely the internodes, the petioles, the blades and the capitulum. Successive internodes on the main stem produced by the apical meristem were considered as the first steps in the model. They were numbered from the base to the top from 1 to 50, and were used to simulate growth of the stem (Barczi et al., 2008) up to 50 internodes. Each internode bore a petiole, and each petiole bore a blade (branching process). Petioles steps were numbered from 51 to 100 (branched on internodes 1 to 50), and blades from 101 to 150 (branched on petioles 51 to 100). At the top of the main stem, the capitulum was constructed using four organ types: the capitulum-receptacle (step 200), which also bore the capitulum-bracts (step 201), the ligularia-flowers (step 202) and the tubularia-flowers (step 203). The reference axis for sunflower was then the sequence of every possible step value sorted in growing order. Steps 151–199 were left undefined to allow for further applications if necessary. Each physiological age corresponded to a 3-D shape (Fig. 1), which was constructed using a 3-D polygonal shape modeller on the basis of observed organs. All shapes are normalized and a transform matrix was subsequently applied to extend these in length and width to the desired dimensions. Special care was taken with blade modelling to take account of blade curvatures, winglet elevation and the bending of its distal part.
|
Phytomer initiation and organ expansion over plant thermal time were parameterized in the AMAPsim program to compute the development and growth of each organ. 3-D virtual plants were built from (1) the values of the different parameters in the logistic equation (eqn 2) and previously fitted to data in order to generate smoothed organ growth and development, and from (2) distribution functions of observed data for blade angles and bending. Simulations were carried out from emergence to the end of the growth cycle as based on a daily time step and for every treatment x experiment. Phytomer initiation and organ growth and development were established from an average plant, meaning that all plants in the canopy possessed organs of the same size. By contrast, organ orientation and bending were estimated from a function of distribution based on measurements performed on five plants by treatments. Plants in the simulated canopy thus differed in terms of geometry.
An option was added to AMAPsim to calculate heliotropic movements in 3-D numerical plants on an hourly basis using measurements taken in the field (Fd-May99-1).
Virtual sunflower plot reconstruction
The LANDMAKER program (de Reffye et al., 1989; Lecoustre et al., 1992; Auclair et al., 2001) was used to reconstruct the canopy by positioning the plants in virtual plots in accordance with the planting design for each treatment x experiment. Virtual plants were positioned every 0·6 x 0·3 m with small random variations accounting for the observed deviations from theoretical positions. For plants grown in the canopy, a plot of 3·2 m2 was simulated compared with a plot of 16 m2 for isolated plants. Seedlings were azimuthally orientated at random. Plots were virtually duplicated in all directions by the MMR model such that radiative simulations were performed without border effects.
Light-balance calculation
The light climate in the virtual plots was simulated by the MMR (from the name of its three modules Mir, Musc and Radbal) model (Dauzat, 1994; Dauzat and Eroy, 1997; Dauzat et al., 2004) for short waves (400–700 nm). The program split the sky hemisphere into 46 sectors according to Den Dulk's ‘TURTLE’ model (den Dulk, 1989). The first module, Mir, deduces the interception of directional light by vegetation elements (elementary internodes, petioles, blades and capitulum) and soil from the number of pixels visible on images calculated from the central direction of each sector. The second module, Musc, calculates the multiple scattering of intercepted light and the resulting additional irradiation of vegetation and soil (Dauzat et al., 2008). These two modules provide a partial light balance for each turtle direction of illumination. The third module, Radbal, combines outputs of previous modules in order to obtain the complete light balance of the plot for the light conditions of the treatment considered. Radbal first calculates the direct and diffuse components of incident radiation and then distributes these components into the 46 sky sectors (Dauzat et al., 2001). Organ irradiance is finally obtained as the sum of partial irradiances by incident and scattered light received from each turtle sector. Downscaling of scattered light from canopy layer to organ supposes that interception of scattered light is proportional to organ surface area.
Calculation of radiative variables
The fraction of PAR intercepted by the canopy (
i) was estimated from the daily incident PAR above the canopy (PARincident) and the daily incident PAR at the level of the soil [PARincident (soil)]:
|
| (3) |
i depicts only the interception of incident radiation and does not account for scattered radiation. Conversely, the calculation of PAR absorbed by a plant organ includes its interception of incident PAR [PARi(organ)] plus the additional fraction of light scattered by vegetation and soil which is intercepted by the organ [PARs(organ)]. The daily PAR absorbed by an organ, PARa(organ), is thus given by:
|
| (4) |
and 
are, respectively, the reflection and the transmission coefficients determined for the PAR range. Reflectances of 0·15 and 0·18 were measured for the blades and the soil, respectively, using a spectroradiometer (Fieldspec, ASD Inc., Arvada, CO, USA). As proposed by Guyot (1990), the transmission and reflection coefficients were considered to be equal for the blades. Petioles, stem and capitulum (essentially composed of bracts during the period considered) were considered to have the same reflectance as blades, and zero transmittance (as proposed by Guilioni and Lhomme, 2006).
The fraction of absorbed PAR [a] was estimated as the sum of the PAR absorbed by all the individual organs of the plot, normalized by incident PAR and plot surface area [S(plot)]:
|
| (5) |
|
| (6) |
The study was conducted using 3-D virtual plants built with or without heliotropic movements (referred to below as ‘static’ and ‘heliotropic plants’). Static plants were constructed as previously described, i.e. with stem, petioles and blades in their average position. Stem, leaves and capitulum movements were mimicked on an hourly basis for ‘heliotropic plants’ to reproduce observed movements.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Final plant architecture: plant structure is stable while organ size varies
Non-decapitulated plants showed a stable number of phytomers, and this for all experiments and treatments (Table 2). ‘Heliasol’ genotype produced less phytomers than ‘Albena’, as previously reported (Lecoeur, 2001; Vear et al., 2002). Phyllotaxic angles were also stable for all plants under all experimental conditions with values of 90·6 ± 0·4° for the first four phytomers, 74·4 ± 0·3° for the fifth phytomer (transition) and 137·8 ± 0·6° for subsequent phytomers. Plants cultivated at a low density (greenhouse and isolated plants grown outdoors) reached a final height of 1·52 ± 0·02 m on average whereas those grown in the field at 5·6 plants m–2 were taller with final heights ranging from 1·98 ± 0·10 to 2·40 ± 0·07 m, resulting from longer internodes (Table 2). Stem diameter varied between the experimental treatments with values for isolated plants being two- or three-fold those in plants grown at a regular density.
Plant leaf area was the architectural variable most affected by the level of incident PAR (Table 2). The experiments conducted covered a broad range of incident PAR (from 8·0 to 55·8 mol m–2 d–1) and thus considered most of the conditions routinely encountered in sunflower cropping. Plant leaf area for ‘Albena’ ranged from 0·82 ± 0·08 m2 (Gh-Nov98-1) to 2·50 ± 0·24 m2 (Fd-May99-2). Removing the capitulum had a positive effect on both leaf area and stem diameter in greenhouse plants, but no effect was observed in isolated plants. Leaf area was slightly reduced when the plants were subject to early shading from 1·14 to 1·01 m2 (Fd-May99-31), although the effect was not significant. Late shading treatment (Fd-May99-32) had no effect.
3-D virtual plants: all experimental situations can be simulated by virtual plots
3-D virtual plants were simulated for each experiment and treatment as daily steps over the first 70 d after plant emergence, thus covering the entire period of organ expansion (Fig. 2A). Representations were constructed for plants grown under contrasted environmental conditions (Fig. 2B). The cumulated PAR absorbed by the plant over the 70 d following emergence varied from 50 mol of photons for greenhouse plants in the winter and 200–300 mol for field plants at a regular density, up to 2000 mol for isolated plants. The model was applied to the ‘Heliasol’ genotype (Fd-May01-1H), which produced fewer leaves than ‘Albena’, had a different leaf area distribution along the stem and a more flexible stem resulting in upper bending. A representation is given as an example for the Fd-May01-1H virtual plot (Fig. 3).
|
|
3-D virtual plants were constructed by fitting the measurements taken on all organs three times weekly. Linear fittings between observed and simulated blade areas gave a slope of 1·008 and r2 = 0·999, indicating that changes in individual blade areas with thermal time were satisfactorily rendered (Fig. 4, inset). The integration of organ expansion and development (phytomer initiation, blade and internode expansion) over time resulted in an accurate estimation of the different variables at the whole-plant level [e.g. total plant area (Fig. 4A) or plant height (Fig. 4B)] throughout the crop development period. The correlative coefficient of determination (r2) between observed and simulated values was elevated (greater than 0·85 in 80 % of tests) for all the architectural variables studied (e.g. internode and petiole lengths and diameters). Mean simulation errors, estimated by coefficients of variation of error (CVe), ranged from 3 to 8 %. Simulations carried out with only half of the measurements gave similar results, suggesting than less frequent measurements can be used to reconstruct dynamic 3-D virtual plants.
|
Light balance evaluation in 3-D virtual plants
Light balance was assessed by comparing the fraction of intercepted PAR (
i) simulated in virtual plots with the corresponding experimental measurements in the field (Fig. 5). The results are presented for both Fd-May01-1A and Fd-May01-1H, with the fraction of intercepted PAR expressed as a function either of thermal time (Fig. 5A, C) or LAI (Fig. 5B, D). The fraction of intercepted PAR was adequately simulated, with a coefficient of determination (r2) of around 0·990 and a CVe of less than 4 %. Despite the fact that this evaluation was made globally at the canopy level, the ability of the model to account for differences between genotypes and between different situations supports the choice of this approach to describe complex interactions between light and plant components.
|
Another evaluation was performed at plant maturity to test the model's accuracy at the organ level. The fraction of intercepted PAR was measured in canopy plants whose blades were cut such that only the stem, the petioles and capitulum remained. The fraction of PAR intercepted by these defoliated plants was 0·3 and did not significantly differ from the simulated data (0·33).
Light absorption at organ, plant and plot scales
The fraction of absorbed PAR (a) simulated at the plot scale showed a sigmoidal pattern over thermal time (Fig. 6). From emergence up to 200–300 °Cd, the fraction of absorbed PAR was fairly low then increased rapidly over the next 300 °Cd to reach a plateau of some 600 °Cd for greenhouse and field plants in the canopy (Fig. 6A, B) and 800 °Cd for isolated plants (Fig. 6C). Plateau values were around 0·9 in the greenhouse and field canopy experiments and around 0·08 for isolated plants. However, it was difficult to compare the fraction of absorbed PAR, at the whole-plant level, between isolated and canopy plants as this variable depends on plot size for isolated plants (eqn 5). In the canopy, when the fraction of absorbed PAR of the plot was expressed as a function of LAI, it increased almost linearly until LAI reached about 1·5, then increased in a curvilinear manner to a maximum of 0·9 (Fig. 6D). At equal LAI values, ‘Heliasol’ was more efficient than ‘Albena’ in intercepting or absorbing the incident light (Fig. 5). The Beer–Lambert law was used for calculations in 70-d-old plants when the canopy was homogeneous. Inverting the Beer–Lambert law using simulated light transmission and LAI on 3-D virtual plants gave an extinction coefficient k of 0·74 ± 0·14 in ‘Albena’ and 0·88 ± 0·16 in ‘Heliasol’. Simulation under a cloudy sky (13·7 mol m–2 d–1 of incident PAR, with 1 % direct radiation) increased the fraction of absorbed PAR by 2·9 % as compared with a clear sky (54·8 mol m–2 d–1 of PAR, with 52 % direct radiation), underlining the importance of radiative conditions on the fraction of intercepted PAR.
|
Simulations indicated that before the capitulum developed (plateau on the graph), blade PAR absorption accounted for 80 % of incoming PAR for greenhouse and field plants in the canopy (Fig. 6A, B). After 800 °Cd, blade absorption decreased to around 60 %. This decrease corresponded to the increase in capitulum absorption, which reached its maximum at about 1200 °Cd. The capitulum intercepted a major proportion of incident light during the reproductive phase and reduced the light available for the blades. This effect was not observed in isolated plants, which received far more light laterally than plants in the canopy.
PAR absorption by stems and petioles was marginal and accounted for less than 5 % of the total simulated absorption in the whole plant. Their involvement in plant light absorption decreased after 600 °Cd due to increased shading by other plant organs. This weak contribution to light absorption clearly illustrates the minor contribution made by these organs to plant photosynthesis and carbon balance.
Simulation of leaf irradiance
Changes in relative leaf irradiance (RLI) over thermal time showed similar patterns in all the different experimental situations (Fig. 7). An initial decrease until 200 °Cd was due to the major increase in leaf area in small, young plants, whereas internode elongation was negligible. This resulted in elevated blade self-shading. RLI increased up to about 400 °Cd as rapid stem elongation moved the leaves away from each other, thus decreasing self-shading among the leaves. The RLI then decreased again as plant leaf area continued to increase until the end of expansion while stem elongation had ceased. The plateau value for RLI corresponded to the end of plant leaf expansion. The characteristic values for this general pattern were different between the experiments and treatments, illustrating the significance of an architectural approach in such a study.
|
Heliotropism does not improve light capture at the plot scale
Heliotropism movements concerned the azimuthal and zenithal directions of both the stem and the petioles, as well as blade bending and blade rotation over the mid-rib axis. The time course of the vertical projection of blade tip movements is presented in Fig. 8 for all leaves of a plant, as an illustration of the complexity of heliotropic movements. Leaves followed the sun's course with more pronounced movements for blades orientated in the south (e.g. leaves 7 and 9 in Fig. 8) than in the north (e.g. leaves 6 and 8). Leaves at the bottom of the stem tended to be more stable than those in the middle. The youngest leaves at the top (e.g. leaves 10–12 in Fig. 8) were not fully expanded and thus more erect, and their blade tips therefore showed less pronounced movements.
|
These complex movements were integrated in the model to evaluate their impact on light interception. PAR absorption by plant organs was estimated on an hourly basis for 3-week-old 3-D virtual plants simulated with (Fig. 9A) or without (Fig. 9B) heliotropic movements. At the time the simulation was made, the capitulum had just started its growth and its light interception was insignificant. Only minor differences were noted for PAR absorption by leaves (on a day with clear sky conditions) with and without these movements. A slight decrease in heliotropic plants was nevertheless observed at around 14:00 h (12:00 h solar time) related to an increase in mutual shading among the blades (Fig. 9C). The daily PAR absorbed by the blades was 6·89 mol d–1 per plant for static plants and 6·74 mol d–1 for heliotropic plants. Heliotropism thus induced barely a 2·2 % difference in blade light absorption.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Plant–environment interactions over time at the organ level and under natural and artificial conditions
Sunflower shows high PAR interception compared with other species (Jones, 1992), and this reached 90 % in the fully expanded canopy. The plateau for PAR interception was reached in the field at an LAI value of about 3·5 (Fig. 5), indicating that sunflower canopies are highly efficient light interceptors. Given their optical properties, 30 % of the PAR intercepted by blades is scattered, but most of this scattered fraction is subsequently re-intercepted by vegetation. This results in the canopy absorbing about 90 % of incident radiation (Fig. 6). These complex plant–environment interactions caused great variability in terms of RLI over time between the experiments and treatments (Fig. 7), making light interception difficult to predict.
Coupling an architectural model with a light model allowed us to estimate the light balance. To the best of our knowledge, the report herein is one of only a few describing and evaluating a model used to estimate crop light interception over time and at the organ level (e.g. Gautier et al., 2000). Simulated PAR interception was here successfully compared with field measurements over time for two varieties, and at the organ level in a specific experiment with blade removal. Evaluating simulations at the organ level would further validate the approach but field measurements of organ irradiation are possible for only a limited number of locations in a few plants.
The approach developed in this study could be used advantageously when investigating physiological and developmental responses induced by light components (in terms of quantity and quality), and a specific field of application is the comparison of these responses for genotypes with contrasted architectures (Chenu et al., 2007a). In dicotyledonous plants, leaf expansion is highly sensitive to the amount of light intercepted during the first exponential phase of this expansion (Granier and Tardieu, 1999; Chenu et al., 2005). For most leaves this phase occurs during the first quarter of the plant cycle, when the plants are small and the canopy is heterogeneous. The model developed here can be used to estimate the light intercepted during all phases of crop development, and especially during young stages when the underlying hypotheses of the Beer–Lambert law are not fulfilled. The first stages of plant development have a major impact on plant establishment and such a model could also help investigate other processes, the understanding of which relies on the estimation of light intercepted over this period (Chenu et al., 2007b).
Different methods have been developed over the last decade to generate 3-D plants using either direct measurements using a 3-D digitizer (Sinoquet et al., 1998), statistical distributions of parameters describing organ shapes and positions (de Reffye et al., 1988; España et al., 1998) or the simulation of plant growth (Mech and Prusinkiewicz, 1996; Fournier and Andrieu, 1998, 1999; Chenu et al., 2005). The method presented here can be used to simulate plant growth over time and was designed to interact with ecophysiological modules concerning, for instance, leaf growth and development response to light microclimate. At the same time that 3-D virtual plant modelling was developed, several ‘3-D’ light models were also devised to simulate light interaction with 3-D virtual plants (e.g. Pearcy and Yang 1996; Sinoquet et al., 1998). Unlike former 1-D models, these models can be used to detail how radiation is distributed between plant organs. These 3-D radiative models possess a number of distinguishing features. In most cases, diffuse radiation is assumed to be isotropic across the sky (Hanan and Begue, 1995; Chelle and Andrieu, 1998; Sinoquet et al., 1998) whereas the anisotropy of sky brightness is accounted for in MMR (Dauzat et al., 2001) and recently in DSHP (Wang et al., 2006). This was effected in MMR by dividing the sky into sectors and considering each sector as a light source. This division of light sources meant that the model was readily adapted for greenhouse conditions. Simulating artificial environments is important in horticulture or for research experiments carried out in greenhouses or growth chambers. In such environments, the temporal and spatial pattern of incident light is particular to each treatment, depending on external structures and the artificial lighting provided. If the light climate specific to the considered environment is taken into account, the PAR absorbed may be calculated for all growing conditions (e.g. Chenu et al., 2005, in a growth chamber).
The ability of models to simulate light scattering is another important feature of radiative models. Radiosity (Chelle and Andrieu, 1998; Evers et al., 2005; Soler et al., 2003) and Ray-Tracing (Allen et al., 2005) can be used to simulate light scattering in plant stands but the calculations involved are tedious. In the present study, the canopy light balance was calculated every day throughout the crop development period and this reduced computation times. The MMR model was efficient in this respect.
Importance of leaf blades and capitulum in the amount of radiation absorbed by the whole plant
Most of the light is intercepted by the blades, with stem and petioles contributing less than 5 % to total light interception. However, the capitulum becomes an important contributor to plant light balance in late growth stages, contributing more than 25 % to whole-plant absorption at maturity (Fig. 6). This high level of light interception by the capitulum should be taken into account when addressing the energy and water balances of a sunflower canopy (Guilioni and Lhomme, 2006). For plants in the canopy, shade caused by the capitulum decreased leaf irradiance by more than 25 % and this led to a significant decrease in daily leaf photosynthetic activity. However, bracts may also contribute to photosynthesis, especially before the capitulum opens. This leads to a question about the contribution made by capitulum organs to seed growth inside the capitulum (Dosio et al., 2003). By contrast, the large amount of light absorbed by the capitulum increases the temperature of this organ and thus accelerates its development (Guilioni and Lhomme, 2006).
Heliotropism and absorbed radiation at the plant level
Heliotropism in sunflower was thought to increase the light intercepted by the plant (e.g. Herbert, 1983, 1989), mainly by movements of the capitulum. The present study focused not only on heliotropism of the capitulum but also on movements of the main stem and the leaves. Leaf heliotropism was mainly observed for young leaves in expansion at the top of the stem, when the capitulum had not yet developed. The effect on light interception cannot be readily evaluated through field measurements because plants cannot be frozen in a given position. As an alternative approach, as used here, ‘heliotropic plants’ were compared with ‘static plants’ in terms of light interception throughout the day. The simulations showed that heliotropism has almost no effect on the light balance over one day at the crop level. Under such conditions, the estimation of PAR absorbed (plant level, daily time step) indicated that the effects of heliotropism were less than usually believed. It also showed that simulating multiple and complex heliotropic movements on an hourly basis are unnecessarily frequent when estimating the light balance of a sunflower crop. However, it is worth mentioning that cumulative effects in plant development may be important. Furthermore, in the particular case of studies on individual leaf expansion or photosynthesis, it may be interesting to estimate the effects of heliotropism at the organ level as this may provide an advantage in terms of light absorbed by the youngest expanded leaves located at elevated positions on the stem. Heliotropism modifies daily changes in the mutual shading of leaves and reduces the mean angle between the blade and the sun beam, increasing leaf irradiance (Shell and Lang, 1976). As the photosynthetic activity of leaves depends on their physiological age, heliotropism may increase plant photosynthesis by increasing the contribution made by young leaves, which are photosynthetically more efficient than older leaves. Similar processes are assumed to affect the efficiency of a plant's radiation use in relation to the fraction of direct and diffuse radiation, with an increase in diffuse radiation reducing the self-shading of leaves (Sinclair and Shiraiwa, 1993). Such phenomena may partly explain the high biological efficiency reported for sunflower in comparison with other C3 plants (Sinclair and Muchow, 1999). The model developed here could help us to understand these processes better. Furthermore, the work performed on absorbed PAR could be extended to other microclimatic variables such as organ temperature by adding an energy balance (Guilioni and Lhomme, 2006). It would then be possible to estimate other elaborate physiological variables such as photosynthesis (Rapidel et al., 1999; Franck et al., 2005).
A model to compare genotypes
The level and performance of 3-D virtual plant representations allowed us to undertake comparative studies of different genotypes. The two hybrids studied showed a similar plant leaf area, but ‘Heliasol’ absorbed more light than ‘Albena’, illustrating the possible impact of organ arrangement in light interception. The last 30 years of sunflower breeding have led to the selection of germplasms with different architectures (Debaeke et al., 2004). Modern hybrids show better intercepted light efficiencies than older varieties and architectural variables were found to be highly heritable (Triboi et al., 2004), thus paving the way to enhanced sunflower breeding through architectural studies. In such a context, using the model to analyse the relative contribution of architectural traits in various genotypes would be useful to propose new selection criteria according to their pertinence in improving light interception.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
We wish to thank Thierry Lefeuvre and Guillaume Sinniger for data collection and analysis in 1998 and 1999, respectively. Hubert Maillé gave us substantial technical assistance in 2001. Benoit Suard helped us greatly in matters of daily technical assistance, environmental data acquisition and database construction. We would like to thank Philippe Naudin for taking care of plant irrigation and Olivier Turc for his advice and data analysis and calculations. This work is part of the project PROMOSOL PRODUCTIVITE I financed by OLEOSEM.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
-
Alados I, Foyo-Moreno I, Olmo FJ, Alados-Arboledas L. Improved estimation of diffuse photosynthetically active radiation using two spectral models. Agricultural and Forest Meteorology (2002) 111:1–12.[CrossRef][Web of Science]
Allen MT, Prusinkiewicz P, DeJong TM. Using L-systems for modeling source–sink interactions, architecture and physiology of growing trees: the L-PEACH model. New Phytologist (2005) 166:869–880.[CrossRef][Web of Science][Medline]
Auclair D, Barczi JF, Borne F, Etienne M. Assessing the visual impact of agroforestry management with landscape design software. Landscape Research (2001) 26:397–406.[CrossRef]
Barczi JF, Rey H, Caraglio Y, de Reffye P, Barthélémy D, Dong QX, Fourcaud T. AMAPsim: an integrative whole-plant architecture simulator based on botanical and environmental plasticity knowledge. Annals of Botany (2008) 101:1125–1138.
Chelle M. Phylloclimate or the climate perceived by individual plant organs: What is it? How to model it? What for? New Phytologist (2005) 166:781–790.[CrossRef][Web of Science][Medline]
Chelle M, Andrieu B. The nested radiosity model for the distribution of light within plant canopies. Ecological Modelling (1998) 111:75–91.[CrossRef][Web of Science]
Chelle M, Andrieu B. Radiative models for architectural modeling. Agronomie (1999) 19:225–240.[CrossRef][Web of Science]
Chenu K, Franck N, Dauzat J, Barczi JF, Rey H, Lecoeur J. Integrated responses of rosette organogenesis, morphogenesis and architecture to reduced incident light in Arabidopsis thaliana results in higher efficiency of light interception. Functional Plant Biology (2005) 32:1123–1134.[CrossRef][Web of Science]
Chenu K, Franck N, Lecoeur J. Simulations of virtual plants reveal a role for SERRATE in the response of leaf development to light in Arabidopsis thaliana. New Phytologist (2007) a 175:472–481.[CrossRef][Web of Science][Medline]
Chenu K, Fournier C, Andrieu B, Giauffret C. An architectural approach to investigate maize response to low temperature. In: Scale and complexity in plant systems research: gene–plant–crop relations—Spiertz JHJ, Struik PC, can Laar HH, eds. (2007) b. The Netherlands: Springer. 201–210.
Dauzat J. Radiative transfer simulation on computer models Elaeis guineensis. Oléagineux (1994) 49:81–90.[Web of Science]
Dauzat J, Eroy MN. Simulating light regime and intercrop yields in coconut based farming systems. European Journal of Agronomy (1997) 7:63–74.[CrossRef][Web of Science]
Dauzat J, Rapidel B, Berger A. Simulation of leaf transpiration and sap flow in virtual plants: description of the model and application to a coffee plantation in Costa Rica. Agricultural and Forest Meteorology (2001) 109:143–160.[CrossRef][Web of Science]
Dauzat J, Martin P, Luquet D, Clouvel P. Using virtual plants for analysing the light foraging efficiency of cotton crop at low density. Annals of Botany (2008) 101:1153–1166.
Debaeke P, Triboi AM, Vear F, Lecoeur J. Crop physiological determinants of yield in old and modern sunflower hybrids. (2004) Proceedings of the 16th International Sunflower Conference. Fargo, ND, USA. 267–273. In: ISA, ed.
den Dulk JA. The interpretation of remote sensing, a feasibility study (1989) The Netherlands: Wageningen University. PhD thesis.
Dosio GAA, Rey H, Lecoeur J, Izquierdo NG, Aguirrezabal LAN, Tardieu F, Turc O. A whole-plant analysis of the dynamics of expansion of individual leaves of two sunflower hybrids. Journal of Experimental Botany (2003) 54:2541–2552.
España M, Baret F, Chelle M, Aries F, Andrieu B. A dynamic model of maize 3D architecture: application to the parameterisation of the clumpiness of the canopy. Agronomie (1998) 18:609–626.[CrossRef][Web of Science]
Evers JB, Vos J. A 3D virtual plant-modelling study. In: Scale and complexity in plant systems research: gene–plant–crop relations—Spiertz JHJ, Struik PC, can Laar HH, eds. (2007) The Netherlands: Springer. 213–221.
Evers JB, Vos J, Fournier C, Andrieu B, Chelle M, Struik PC. Towards a generic architectural model of tillering in Gramineae, as exemplified by spring wheat (Triticum aestivum). New Phytologist (2005) 166:801–812.[CrossRef][Web of Science][Medline]
Fournier C, Andrieu B. A 3D architectural and process-based model of maize development. Annals of Botany (1998) 81:233–250.
Fournier C, Andrieu B. ADEL-maize: an L-system based model for the integration of growth processes from the organ to the canopy. Application to regulation of morphogenesis by light availability. Agronomie (1999) 19:313–327.[CrossRef][Web of Science]
Franck N, Vaast P, Génard M, Dauzat J. Soluble sugars mediate sink feedback down-regulation of leaf photosynthesis in field-grown Coffea arabica. Tree Physiology (2005) 26:517–525.[Web of Science]
Gautier H, Mech R, Prusinkiewicz P, Varlet-Grancher C. 3D architectural modelling of aerial photomorphogenesis in white clover (Trifolium repens L.) using L-systems. Annals of Botany (2000) 85:359–370.
Granier C, Tardieu F. Is thermal time adequate for expressing the effects of temperature on sunflower leaf development? Plant, Cell and Environment (1998) 21:695–703.[CrossRef]
Granier C, Tardieu F. Leaf expansion and cell division are affected by reducing absorbed light before but not after the decline in cell division rate in the sunflower leaf. Plant, Cell and Environment (1999) 22:1365–1376.[CrossRef]
Guilioni L, Lhomme JP. Modelling the daily course of capitulum temperature in a sunflower canopy. Agricultural and Forest Meteorology (2006) 138:258–272.[CrossRef][Web of Science]
Guyot G. Optical properties of vegetation canopies. In: Applications of remote sensing in agriculture.—Steven MD, Clarck JA, eds. (1990) London: Butterworths. 19–44.
Hanan NP, Begue A. A method to estimate instantaneous and daily intercepted photosynthetically active radiation using a hemispherical sensor. Agricultural and Forest Meteorology (1995) 74:155–168.[CrossRef][Web of Science]
Herbert TJ. The influence of axial rotation upon interception of solar radiation by plant leaves. Journal of Theoretical Biology (1983) 105:603–618.[CrossRef][Web of Science]
Herbert TJ. A model of daily leaf movement in relation to the radiation regime. Advances in Legume Biology (1989) 29:629–643.
Jaeger M, de Reffye P. Basic concepts of computer simulation of plant growth. The 1990 Mahabaleshwar Seminar on Modern Biology. Journal of Bioscience (1992) 17:275–291.[CrossRef]
Jones HG. Plants and microclimate. A quantitative approach to environmental plant physiology. (1992) Cambridge: Cambridge University Press.
Lecoeur J. Potentiels de rendement du tournesol: intérêt des modèles écophysiologiques pour le diagnostic agronomique. Les rencontres annuelles du CETIOM (2001) 6–10. Tournesol. CETIOM, Paris, France.
Lecoustre R, de Reffye P, Jaeger M, Dinouard P. Controlling the architectural geometry of a plant's growth. Application to the Begonia genus. In: Creating and animating the virtual world (1992) New York: Springer-Verlag. 199–214.
Louarn G, Guédon Y, Lecoeur J, Lebon E. Quantitative analysis of the phenotypic variability of shoot architecture in two grapevine cultivars (Vitis vinifera L.). Annals of Botany (2007) 99:425–437.
Mech R, Prusinkiewicz P. Visual models of plants interacting with their environment. Proceedings of SIFFRAPH '96 (1996) 397–410. ACM.
Monteith JL. Climate and efficiency of crop production in Britain. Philosophical Transactions of the Royal Society of London, Series B-Biological Sciences (1977) 281:277–294.[CrossRef]
Pearcy RW, Yang WM. A three-dimensional crown architecture model for assessment of light capture and carbon gain by understory plants. Oecologia (1996) 108:1–12.[CrossRef][Web of Science]
Prusinkiewicz P. Modeling of spatial structure and development of plants: a review. Scientia Horticulturae (1998) 74:113–149.[CrossRef]
Rapidel B, Dauzat J, Berger A. Application of a mock-up based transpiration and photosynthesis model on a coffee plantation in Costa Rica. In: Multi-strata agroforestry systems with perennial crops, International Symposium on Multi-Strata Agroforestry Systems with Perennial Crops—Jiménez F, Beer J, eds. (1999) Turrialba, Costa Rica: CATIE. 198–203.
de Reffye P, Edelin C, Françon J, Jaeger M, Puech C. Plant models faithful to botanical structure and development. In: Proceedings of SIGGRAPH '88, Atlanta, Georgia, 1–5 August 1988. Computer Graphics—Dill J, ed. (1988) 22:151–158.
de Reffye P, Lecoustre R, Edelin C, Dinouard P. Modelling plant growth and architecture. In: Cell to cell signalling: from experiments to theoretical models—Goldbeter A, ed. (1989) London: Academic Press. 237–246.
Sadras VO, Milroy SP. Soil–water thresholds for the responses of leaf expansion and gas exchange: a review. Field Crops Research (1996) 47:253–266.[CrossRef][Web of Science]
Sinclair TR, Muchow RC. Radiation use efficiency. Advances in Agronomy (1999) 65:215–263.[Web of Science]
Sinclair TR, Shiraiwa T. Soybean radiation-use efficiency as influenced by non-uniform specific leaf nitrogen distribution and diffuse radiation. Crop Science (1993) 33:808–812.
Sinoquet H, Thanisawanyangkura S, Mabrouk H, Kasemsap P. Characterization of the light environment in canopies using 3D digitising and image processing. Annals of Botany (1998) 82:203–212.
Shell GSG, Lang RG. Movements of sunflower leaves over a 24-h period. Agricultural Meteorology (1976) 16:161–170.[CrossRef][Web of Science]
Slade AJ, Hutchings MJ. The effects of light intensity on foraging in the clonal herb Glechomas hederacea. Journal of Ecology (1987) 75:639–650.[CrossRef][Web of Science]
Soler C, Sillion FX, Blaise F, de Reffye P. An efficient instantiation algorithm for simulating radiant energy transfer in plant models. ACM Transactions on Graphics (2003) 22:204–233.[CrossRef][Web of Science]
Triboi AM, Messaoud J, Debaeke P, Lecoeur J, Vear F. Heredity of sunflower leaf characters useable as yield predictors. (2004) Vol. 2. Proceedings 16th International Sunflower Conference: Fargo: ND. 517–523.
Varlet-Grancher C, Gosse G, Chartier M, Sinoquet H, Bonhomme M, Allirand JM. Mise au point: rayonnement solaire absorbé ou intercepté par un couvert végétal. Agronomie (1989) 9:419–439.[CrossRef][Web of Science]
Vear F, Bony H, Joubert G, Tourvieille de labrouhe D, Pauchet I, Pinocher X. The result of 30 years of sunflower breeding for France. OCL (2002) 10:66–73.
Wang XP, Guo Y, Li BG, Wang XY, Ma YT. Evaluating a three dimensional model of diffuse photosynthetically active radiation in maize canopies. International Journal of Biometeorology (2006) 50:349–357.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in Ann Bot:
- ContentSnapshots
Ann Bot 2008 101: NP.[Extract] [Full Text]
This article has been cited by other articles:
|
|
 |
|
  T. Fourcaud, X. Zhang, A. Stokes, H. Lambers, and C. Korner Plant Growth Modelling and Applications: The Increasing Importance of Plant Architecture in Growth Models Ann. Bot., May 1, 2008; 101(8): 1053 - 1063. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||