Heterogeneity Analysis of Cucumber Canopy in the Solar Greenhouse

Available online at www. sciencedirect.comJournal of Integrative Agriculture。ScienceDirect2014, 13(12): 2645-2655December 2014RESEARCH ARTICLEHeterogeneity Analysis of Cucumber Canopy in the Solar GreenhouseQIAN Ting-ting', LU Sheng lian', ZHAO Chun-jiang'"2, GUO Xin-yu', WEN Wei-liang2 and DU jan-jun2' Shanghai Jiao Tong University, Shanghai 200240, P.R.China'Beiing Research Center for Information Technology in Agriculture, Bejing 100097, P.R. ChinaAbstractDetailed analysis of canopy structural heterogeneity is an essential step in conducting parameters for a canopy structuralmodel. This paper aims to analyze the structural heterogeneity of a cucumber (Cucumis sativus L.) canopy by means of analyzingleaf distribution in a greenhouse environment with natural sunlight and also to assess the effect of structural canopy heterogeneityon light interception and photosynthesis. Two experiments and four measurements were carried out in autumn 201 1 and spring2012. A static virtual three-dimensional (3D) canopy structure was reconstructed using a 3D digitizing method. The diurnalvariation of photosynthesis rate ved using CIRAS-2 photosynthesis system. The results showed that, leaf azimuthas tested with the Rayleigh-test was homogeneous at vine tip over stage but turned heterogeneous at fruit harvest stage. Aftereliminating the influence of the environment on the azimuth using the von Mises-Fisher method, the angle between two sccessiveleaves was 1449; at the same time, a rule for the azimuth ditribution in the canopy was established, stating that the azimuthdistribution in cucumber followed a law which was positive spin and anti-spin. Leaf elevation angle of south- oriented leaves wason average 13.8° higher than that of north- oriented leaves. The horizontal distribution of light interception and photosynthesisdiffered significantly between differently oriented leaves. East- and west-oriented leaves exhibited the highest photosyntheticrate. In conclusion, detailed analysis of canopy structural heterogeneity in this study indicated that leaf azimuth and elevationangle were heterogeneous in cucumber canopy and they should be explicitly described as they have a great impact both on lightdistribution and photosynthesis.Key words: heterogeneity, inclination, leaf azimuth, light intereption, photosynthesiscanopy architectural traits.INTRODUCTION .Architectural traits (such as the number, shape, in-clination and azimuth of organs) are genotype specific,The structure of the canopy exhibits a large heteroge-while at the same time highly dependent on the climaticneity, not least since it is composed of hundreds andconditions. These traits can be used to analyze canopythousands of plant parts which themselves diverge indynamics and heterogeneity (Sarlikioti et al.201 1a).shape, size, topology and orientation. Therefore, it isTo date, several studies were conducted to modeldificult to establish an accurate canopy structural modelorgan emergency (Cao and Tibbitts 1995; Wiechersby using parametric method without detailed analysisof et al. 201 1), leaf area accumulation (Yang et al.1990),Received 19 November, 2013 Accepted 27 February, 2014QIAN Ting-ting, E-mail:qiantinging19831205@ 126. com; Correspondence ZHAO Chun-jang, Tel: +86-10-515中国煤化Imacjnercita.g.cnYHCNMH G◎2014, CAAS. All rights reserved. Published by Elsevier Ltd.doi: 10.1016/S2095-3119(14)60776-02646QIAN Tingting et al.internode elongation (Kahlen and Stitzel 201 1) and re--to explore the relationship between leaf shape andlationship between leaf area and position (Kahlen et al.photosynthetic performance;2008). But studies of inclination and azimuth model--to better understand the mechanisms of canopying are rare: Four types of leaf inclination distributionheterogeneity;(Wit 1965) as well as an ellipsoidal model presented-to improve the methodology for the extraction ofby Campbell (1986) were used to describe canopymodel parameters for 3D structure.structure and also to calculate light interception in aone-dimensional model (Campbell 1990), however,RESULTSthere are obvious flaws when these models are usedfor 3D virtual plant construction; a constant phyllo-taxis angle of 2/5x360°=144° is used in the structuralvon Mises-Fisher distribution analysis of leafmodel (Kahlen 2006), however, measured data showedazimuthslight deviations from this constant value. Since leafinclination and azimuth has been shown to also afeet The azimuth of the first leaf from the bottom of acanopy light interception and yield (Toler et al. 1999;plant was set to be 0° and azimuth angles of all otherFalster and Westby 2003), they should be explicitly leaves at the same plant were normalized with respectdescribed (Sarlikioti et al. 2011b). To the best of ourto the first leaf. Two types of azimuth distribution lawsknowledge, these characteristics in cucumber canopyare shown in Fig. 1-A and B: positive spin and anti-spin.structure have not been previously investigated.All plants in the M1-M4 measurement were separatedPlant structure influences a large number of physiolog-into two groups according to the two spin distributions.ical responses, e.g, photosynthesis and its feedback withApproximately equal numbers of positive spin andgrowth, or carbon partitioning to vegetative parts andanti-spin distribution were found in the four canopies.fruits (Godin 2000). A key feature of functional-struc-Mean direction u。and concentration parameter k of eachtural plant models (FSPMs) is that the heterogeneitiescanopy were calculated.of canopy light interception and photosynthesis can beTo show the distribution more directly, the period-simulated at the organ level where the plants and theic sequence was transformed into a linear sequencecanopy are constructed spatially explicitly with geom-in a clockwise direction (Fig. 2). There was an ob-etry and topology (Chen et al. 2014). Many studiesvious linear relationship between μo and leaf rank.have been done by using FSPMs: Wang e1 al. (2006)That is, the angle between two subsequent leavesused a 3D model to simulate canopy light interception;is constant. The average phyllotactic angle wasSarlikioti et al. (2011a) developed a static tomato FSPM 143.9 in plants with positive spin and214.40 (-145.60)to explore the spatial distribution of light absorption andin plants with anti-spin (Table 1). The concentrationphotosynthesis in a tomato canopy. Buck Sorlin et al.parameter k of each leaf rank is shown in Fig. 3. The(2011) yet also included the modeling of harvest andhighest k value was found in M3 while the lowest k valuebending, typical interactions with the developing plantappeared in M2. The variance at the early stage wasstructure, and their effects on light distribution applied alower than that at the late stage; in spring it was lowersimilar approach for a semi- dynamic FSPM of cut-rose.than that in autumn; in lower leaf ranks it was lower thanThree dimensional canopy modeling provides a bigthat in upper ranks.step forward towards detailed structural and functionalanalysis; at the same time, detailed analyses by usingthis method can also provide powerful support for modelLeaf azimuth distribution in the canopyconstruction.The objective of this study was to analyze charac-In order to study the architectural characteristics ofteristics of canopy heterogeneity in cucumber by using leaf azimuth deviation in the cucumber canopy, thea 3D model combined with a light distribution modelleaf azimuth dis中国煤化工into eightwith the aims:orientation clasYHCNMH e[s for leaf◎2014, CAAS. All rights reserved. Published by Elsevier Ltd.Heterogeneity Analysis of Cucumber Canopy in the Solar Greenhouse2647A■Leaf azimuth✧M1 OM2 OM3 XM4.5000 -4000.6一 30\” 30002112000.100061 240y1201520Leaf rank8101I050184000 -B0302025目0-270-324+Fig.2 Changes in average cumulative azimuth angle (0) as a function6- 2407120of leaf rank of all plants in four measurements (M1, M2, M3 andM4). The average leaf number per plant in four measurements were8 (MI), 13 (M2), 8 (M3) and 21 (M4), respeeively. AIl plants in10~four measurements were separated into two roups: positive spinand anti-spin according to their azimuth distribution pattern (thecumulative degree was calculated by integrating the average μ。valueFig. 1 Measured azimuth distribution of cucumber leaves as a of each leaf rank in a clockwise direction, with the cumulative degreefunction of leaf rank. A distribution law was found in the canopy: ofthe frst rank being set to0). The average cumulative azimuth anglepositive spin (A) and anti-spin (B). Leaf rank (ordinate) is increeasing of positive spin group was shown in A while the average cumulativecentrifugally, points signify leaf azimuth for a given rank. To visualize azimuth angle of anti-spin group was shown in B.the spin, points were joined by lines according to the rank sequence.able 1 Average angle between two subsequent leaves of eachazimuth distribution in the eight classes of the four measurementmeasurements are shown in Fig. 4. The distributionMeasurementAnti- spinPositive spin216.84145.69frequency was nearly uniform in the MI and M3 M2208.87142.08treatments, and the P value of the Rayleigh test213.71141.56218. 19146.34in Table 2 provides a further confirmation of thisstatement (P>0.05).The azimuth classes IV and V were significantly一✧MI一0-M2-心-M3--*--M4more frequent than the classes I and VIII in M212and M4. Frequency class distributions were fit-ted using quadratic curves for each measurement8-64(Fig. 4). The coefficient of determination (R2) was .higher than 0.7 in both measurements M2 and M4(Table 2). Cucumber leaves reoriented to the southas leaf number increased. The frequency of classesIV and V increased greatly in winter compared to2468101214161820that in summer. The frequency of classes IV andV increased to 40.9% in M2 while 27.9% in M4. Thatis, the azimuth of most leaves at a later developmentalFig.3 Compariso中国煤化工: k for fourstage reoriented to the south (Fig.4).measurements, as a fuYHCNM HG◎2014. CAAS. All rights reserved. Published by Elsevier Ltd.2648QIAN Tingting et al.Leaf elevation angles distribution in the canopy bution classes of four measurements was analyzed.Fig. 6 shows that elevation angles in classes I and VIIDynamics of average leaf elevation was basically theare obviously larger than those in classes III -VI. Thesame in the four measurements. At the top of a plant,result of t-test (P<5%) shows that there is a significantleaves tended to be more erect towards the plant stem anddifference between the southern and the northerm azimuthdecreased following a negative exponential from top toclasses. The average elevation of leaves oriented to thebottom. Below a certain position, elevation angles stop south was nearly -20°. These leaves were perpendiculardecreasing but still exhibit some fluctuations (Fig.5-A).to the sun incident.The relationship between leaf elevation (L) and leafpoition (LP) was described using fumction () with a Diurnal variation in light interception and pho-coefficient of determination (R2) of 0.957 (Fig. 5-B):tosynthesisLFLI=agxexp(- a -)+a,(1)Light interception in the M4 measurement was simulatedWere,1,116.7,.1.25.a.=-29.11. a, anda, are by combining the 3D srucural model with a light dis-coefficients from the data fitting. a, is the asymptote oftribution model. The leaves were distributed into fourthe curve, which means that below the seventh-youngest orientation groups in the horizontal and five layers in thephytomer, leaf elevation fluctuated around -29.11°.vertical. Diurnal light interception of leaves in differentIn order to find the underlying mechanism for this groups and layers was simulated with an hourly step size,fluctuation, average leaf elevation in the eight distri-0.9) at an earlydirectly lead to heterogeneity in light interception andstage when leaf number is less than eight. However, asone in photosynthesis. To investigate the horizontalleaf numberincreased, mutual shading of leaves alsoand vertical distribution of photosynthesis, leaves in the increased, and leaf distribution became non-uniform.canopy were separated into three layers in the verticalMore leaves avoid the northern region presumably in or-plane and four distribution groups in the horizontal plane.der to intercept m中国煤化工er, becauseTwelve leaves were selected from three positions (top,short day length arl|YHCNMHGconducive◎2014. CAAS. All rights reserved. Published by Elsevier Ltd.2650QIAN Tingting et al.North: East10001000-800-600-40200A Soub8:00 10:00 12:00 14:00 10.008:00 10:00 12:00 14:00 16.0012:00 14:00 16:00 18:0016:00 18:00Time (h)South;West ;员80(800600-:4008:00 10:00 12:00 14.00 608:008:00 1.00 1200 1400 16060Time (h) .Time (hFig. 7 Simulated light interception of leaves at dfferent orientation (azimuth) groups and layers. The canopy height was 1.8 m in total andthe canopy was distributed into five layers from bottom to top. Each layer contains 4-5 leaves with layer 5 being the topmost. In addition, thecanopy leaves were separated into four distribution groups with respect to the horizontal azimuth: North (316°-45°), East (46*-1359), South(136*-2259) and West (226*-315*). Diurnal variation of light interception of leaves was simulated between 8:00 and 17:00 using a time step ofone hour, on May 23, 2012.for crop growth and development, where a slow solar to the plant leaf arrangement rule, Richter (1978) indicatedelevation increases mutual shading between neighboring that the third leaf usually makes a random choice as toplants. This result indicated that plants will adapt theirwhich way the genetic spiral will be wound. Therefore,structure to the local environment in a sunlit greenhouse it is possible that the two distribution laws appeared in thewhich exhibits a particular light environment with a lightcanopy with equal probability (50%). It is thus necessarygradient decreasing from south to north. Kahlen (2006) to consider these distribution laws in the 3D reconstructionfound that cucumber leaves can reorient their azimuthof the cucumber canopy.angle to intercept more light. We considered that thisThe results of the leaf elevation analysis show that leafis the reason behind our finding that southerly orientedelevation decrease following a negative exponential lawleaves are more frequent in winter (M2 measurement) from top to bottom, while below the seventh-youngestthan in summer (M4 measurement).phytomer leaf elevation fluctuate around -29.110. TheDue to the influence of the local light environment the gradient in leaf elevation has a big effect on both lightangle between two subsequent leaves was not constant absorption and photosynthesis (Pearcy et al. 1990) andand it was difficult to obtain the angle between two sub-a similar trend as the result presented by Kahlen (2006)sequent leaves directly. However, when we used the von except that the fuctuation angle was nearly -69°. RossMises- Fisher equation to analyze the distribution law,(1981) and Yang et al. (1990) found that the average leafsome intrinsic characteristics were found, i.e, positive angle indices were038an4 Q 30 reenentively, translat-spin and anti-spin laws in the canopy and a constant leafing to -349 and中国煤化工grees with .angle (144°) between two successive leaves. With respect the results of thisIYHCNMHGdfference◎2014, CAAS. All rights reserved. Published by EIsevier Ltd.Heterogeneity Analysis of Cucumber Canopy in the Solar Greenhouse2651-✧North ---- East - 公-South - -X- WestTop layer,E50300X、国10-.100的8:0010:00 12:00 14:00 16:00 18:008:00 10:00 12:00 14:00 16:00 18:00Time (h)Time (h) .25 7600]Middle layer20-合400-治、5。o-th发@.ao8:(8:0010:00 12:0014:00 16:00 18:0025 ]Bottom layer600 7Botom layer0-500合4000