SciELO - Scientific Electronic Library Online

 
vol.73 issue1Cracking in sweet cherries: A comprehensive review from a physiological, molecular, and genomic perspectiveGenetic variation within three populations of Phycella australis (Phil.) Ravenna from Biobío Region, Chile, evaluated using ISSR markers author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Chilean journal of agricultural research

On-line version ISSN 0718-5839

Chilean J. Agric. Res. vol.73 no.1 Chillán Mar. 2013

http://dx.doi.org/10.4067/S0718-58392013000100011 

SCIENTIFIC NOTE

Validation of a leaf area prediction model proposed for rose

 

Giancarlo Fascella1*, Salem Darwich2, and Youssef Rouphael2

1Consiglio per la Ricerca e la Sperimentazione in Agricoltura, Unità di Ricerca per il Recupero e la Valorizzazione delle Specie Floricole mediterranee (Research Unit for Mediterranean Flower Species), Palermo, Italy. "Corresponding author (fascella@excite.it).
2Lebanese University, Faculty of Agricultural Engineering and Veterinary Medicine, Dekwaneh-El Maten, Beirut, Lebanon.


Leaf area (LA) is a valuable key for evaluating plant growth, therefore accurate, simple, and nondestructive methods for LA determination are important for physiological and agronomic studies. A LA prediction model based on leaf length (L) and width (W) and developed under greenhouse on 14 cultivars of rose (Rosa hybr.*) was validated on a different cultivar of R. hybrida 'Red France' and on a wild rose species (Rosa sempervirens L.) grown under open-field conditions with two light environments: ambient and 50% shade. Comparisons between measured vs. calculated LA using the following model: LA (cm2) = 0.56 + 0.717 LW, showed a high degree of correlation (R2 > 0.95) and provided quantitative evidence of the validity of the LA prediction model. Calculated LA values were very close to the measured values, giving an underestimation of 3.5%, 4.2%, 1.1%, and an overestimation of 1.3% in the prediction for R. hybrida ambient light, R. hybrida 50% shade, R. sempervirens ambient light, R. sempervirens 50% shade, respectively. This model can provide accurate estimations of rose LA independently of the genetic materials and the growing conditions and can be adopted in many experimental comparisons without the use of any expensive instruments.

Key words: Leaf length, leaf width, Rosa hybr.*, Rosa sempervirens, light environments, regression analysis, model validation.


INTRODUCTION

Leaf area (LA) is associated with many agronomic and physiological processes including growth, photosynthesis, transpiration, light interception, and energy balance (Rouphael and Colla, 2004; 2005; Antunes et al., 2008; Kandiannan et al., 2009; Spann and Heerema, 2010; Rouphael et al., 2010a). Leaf area can be measured by destructive or nondestructive measurements. Many methods have been devised to facilitate the measurement of LA. However, these methods, including those of tracing, blueprinting, photographing, or using a conventional planimeter, require the excision of leaves from the plants. It is therefore not possible to make successive measurements of the same leaf. Plant canopy is also damaged, which might cause problems to other measurements or experiments. Therefore, a nondestructive method for measuring LA is required by agronomists and physiologists. Accurate, nondestructive measurements permit repeated sampling of the same leaves over time and exclude biological variation in destructive methods (De Swart et al., 2004). Especially when using unique plants, for example in genetically segregating populations, nondestructive measurements are of a great value. A modeling approach involving linear relationships between LA and one or more dimensions of the leaf (length and width) is an inexpensive, rapid, reliable, and nondestructive method for measuring LA and would be more advantageous than many of the methods mentioned above (Rouphael et al., 2007; Tsialtas et al., 2008).

Various models relating leaf length and width to area have been developed for fruit trees (Demirsoy et al., 2004; Serdar and Demirsoy, 2006; Cristofori et al., 2007; 2008; Mendoza-de Gyves et al., 2007; 2008; Fallovo et al., 2008) and vegetable crops (Schwarz and Klaring, 2001; De Swart et al., 2004; Salerno et al., 2005; Rouphael et al., 2006; Rivera et al., 2007; Kumar, 2009; Rouphael et al., 2010b), whereas the estimation of LA of ornamental plants by using mathematical relationships between LA and one or more dimensions of the leaf has received very little attention (Fascella et al., 2009; Rouphael et al., 2010c; Giuffrida et al., 2011).

A simple model for rose (Rosa hybr.*) coming from 14 rose cultivars grown under greenhouse conditions has already been established (Rouphael et al., 2010c). However, the accuracy of the predictions is dependent on the variation of leaf shape between genotypes. Since leaf shape (length:width ratio) may vary among different genetic materials (Stoppani et al., 2003), and growing conditions (Rivera et al., 2007) we need a good model of nondestructive LA estimation to use in physiological study of rose independently of genotypes and growing conditions.

The aim of this study was to validate robustness of the model proposed by Rouphael et al. (2010c), for a R. hybrida 'Red France' and a wild rose species (Rosa sempervirens L.) in different growing environmental conditions (open-field with two light environments: ambient and 50% shading) from those on which it was developed.

MATERIAL AND METHODS

Rosa hybrida L. cv. Red France and a wild rose species (Rosa sempervirens L.) were grown under open-field conditions at the experimental farm of the Agricultural Research Council, Research Unit for Mediterranean Flower Species of Palermo (38°5' N, 13°30' E, 23 m a.s.l.), Italy, during the 2011 growing season. The local climate was characterized by mild and moderately rainy winters, and hot dry summers. The plants were grown into plastic pots containing 3 L of peat and perlite (1:1, v/v) in single rows with a plant density of 6.7 plants m-2. Fertigation was scheduled daily and applied with drip-irrigation system to ensure that water and nutrients were non-limiting. The nutrient solution had the following composition (mg L-1): 180 N, 50 P, 200 K, 120 Ca, 30 Mg, 1.3 Fe, 0.2 Cu, 0.2 Zn, 0.3 Mn, 0.2 B, and 0.03 Mo. The pH and the electrical conductivity were maintained at 6.0 and 2.0 dS m-1, respectively. Shading treatment of both species was achieved by suspending 1.8 m above the ground level a single layer of black polyethylene net which allowed a reduction of 50% of full sunlight. During the experiment mean max/min (July/December) temperatures 31.2/11.4 °C and 29.1/120 °C were recorded for plants cultivated at ambient light and under 50% shading, respectively. Average daily photosynthetic active radiation (PAR) during the experiment ranged from a minimum of 69.7 |imol m-2 s-1 in December to a maximum of 1786.5 |imol m-2 s-1 in July for plants grown at ambient light, and from 35.7 to 969.2 |imol m-2 s-1 for plants under 50% shading. The experiment was designed as a factorial combination of two rose species (R. hybrida L. and R. sempervirens L.) and two light environments (ambient light level and 50% shading). The four treatments were arranged in a randomized complete block design with three replicates per treatment. Each replicate consisted of 20 plants (a total of 120 plants per species). Leaves varied in size from large to small for each treatment and were selected randomly from different levels of the canopy and during different growth stages (vegetative stage, first appearance of flowers, and full blooming). A total of 600 whole leaves (about 150 leaves per treatment: R. hybrida ambient light level, R. hybrida 50% shading, R. sempervirens ambient light level, R. sempervirens 50% shading) were measured for LA, length (L) and width (W). Immediately after cutting, leaves were transported on ice to the laboratory. Leaf length was measured from lamina tip to the point of intersection of the lamina and the petiole, along the midrib of the lamina, while leaf width was measured from end-to-end between the widest lobes of the lamina perpendicular to the lamina mid-rib. L (cm), W (cm) and LA (cm2) of each leaf were measured and recorded using a WinDIAS image analysis system (Delta-T Devices, Cambridge, UK) calibrated to 0.01 cm2. The rose LA estimation model produced by Rouphael et al. (2010c) was

LA = 0.56 + 0.717 LW (1)

In the former study, LA estimation rose model was calibrated using 13 cultivars ('Vivaldi', 'Queen Elizabeth', 'Virgo', 'Velvet Star', 'Anna', 'New Dawn', 'Alba', 'Fairy', 'Iceberg', 'White Success', 'Kardinal', 'Rockstar', and 'Grand Gala') grown under greenhouse conditions during 2007 growing season, and the model was also validated using leaves from a different cultivar ('Dallas') during 2008 growing season.

For the validation of the LA prediction model proposed for rose, LA values obtained using the Rouphael et al. (2010c) model was plotted against observed LA measured by the area meter and the accuracy was based on the highest coefficient of determination (r2) and the lowest mean squared error (MSE). Slope and intercept of the model were tested to see if they were significantly different from the slope and intercept of the 1:1 correspondence line (Dent and Blackie, 1979). Regression analyses were conducted using the SigmaPlot 8.0 package (SigmaPlot, Richmond, California, USA).

RESULTS AND DISCUSSION

Leaf shape
Area of R. hybrida leaves under ambient light level and 50% shading ranged from 6.0 to 39.5 cm2, length from 3.3 to 8.1 cm, and width from 2.4 to 6.6 cm, whereas LA of R. sempervirens under ambient light level and 50% shade ranged from 1.3 to 9.1 cm2, length from 1.8 to 6.0 cm, and width from 0.9 to 2.4 cm (Table 1). One leaf-shape parameter is the L:W ratio. In the current study, the leaf shape (L:W ratio) varied between the two species. Rosa sempervirens under ambient light level and 50% shading had narrow leaves (L:W = 2.05 and 2.24, respectively), whereas R. hybrida under ambient light level and 50% shading had wider leaves (L:W = 1.24 and 1.34, respectively; Table 1). Moreover, leaf shape did not change with leaf size, since L:W ratio was constant with increasing LA, with a similar pattern between growing conditions.

Table 1. The leaf shape (length: width ratio), mean, minimum (min) and maximum (max) values for the leaf length, leaf width, and leaf area of Rosa hybrida and Rosa sempervirens under ambient light level and 50% shading.
*Standard error (SE) in parenthesis.
tCoefficient of determination (i2) and mean squared error (MSE in cm2) of the linear regression between leaf width (W) and leaf length (L).

Validation of the leaf area prediction model
Regression analysis demonstrated strong relationships (P < 0.001) between LA and midvein length (L), maximum leaf width (W), product of length and width (LW), square of length (L2), and square of width (W2) (data not shown). This is in agreement with previous studies (Cristofori et al., 2007; Mendoza de Gyves et al., 2007; Rouphael et al., 2007) on non-destructive model development for predicting LA using simple linear measurements. Moreover, models with a single measurement of L (L and L2) were less acceptable for estimating rose LA as a result of their lowest coefficient of determination (R2), higher MSE. An improvement was possible for single LA estimation when W2 was used as an independent variable. To find a model to predict single LA accurately for rose species in different growing environmental conditions, the product of L x W was used as the independent variable (data not shown). This is in agreement with previous findings (Cristofori et al., 2007; 2008; Fallovo et al., 2008; Mendoza de Gyves et al., 2007; 2008; Rouphael et al., 2010b; 2010c), who reported that the model containing the product L x W gave a better estimation than models based on either L or W alone. Therefore, neither L nor W can be dropped when estimating rose LA.

When rose LA values measured by digital planimeter were plotted against calculated LA, coefficients of determination in all treatments were equal to or greater than 0.95 (Figure 1). The MSE was relatively low in all treatments and ranged from 0.03 to 0.06 cm2 for R. sempervirens under ambient and 50% shade, respectively, while for R. hybrida cv. 'Red France' the MSE ranged from 0.42 to 0.62 cm2. Comparisons between measured vs. calculated LA using the Rouphael et al. (2010c) model (Equation [1]), showed a high degree of correlation and provided quantitative evidence of the validity of the area prediction model (Figure 1). The regression lines of the measured vs. calculated values were not significantly (P = 0.53 and 0.78 for R. sempervirens under ambient light level and 50% shade, respectively; P = 0.65 and 0.54 for R. hybrida 'Red France' under ambient light level and 50% shade, respectively) different from the 1:1 correspondence (Figure 1). Moreover, calculated LA values were very close to the measured values, giving an underestimation of 3.5%, 4.2%, 1.1% and an overestimation of 1.3% in the prediction for R. sempervirens under ambient light level, R. sempervirens 50% shade, R. hybrida under ambient light level, R. hybrida 50% shade, respectively (Figures 1A, 1B, 1C, 1D, respectively). The close relationship between calculated and measured LA values has been observed earlier on other ornamental species such as Euphorbia x lomi (Fascella et al., 2009), bedding plants (Giuffrida et al., 2011) and rose (Rouphael et al., 2010c), when the product L x W was used as independent variable. Besides, the underestimation recorded in the current experiment, especially for R. sempervirens, could be attributed to the genetic variation (e.g. smaller leaves).


Figure 1. Relationship between measured vs. calculated leaf area using the model (Equation [1]) for Rosa sempervirens ambient light (A), R. sempervirens 50% shade (B), R. hybrida 'Red France' ambient light (C), R. hybrida 'Red France' 50% shade (D). Solid lines represent linear regression lines of the model. Dotted lines represent the 1:1 relationship between measured and calculated values.

Validation of LA models is an important step for determining usefulness and accuracy of LA estimation methods. Several researchers have validated LA prediction models for other species, such as grapevine (Vitis vinifera L.; Tsialtas et al., 2008), avocado (Persea americana Mill.; Celik and Uzun, 2002), chestnut (Castanea dentata L.; Serdar and Demirsoy, 2006), peach (Prunus persica L.; Demirsoy et al., 2004), kiwifruit (Actinidia deliciosa (A. Chev.) C.F. Liang & A.R. Ferguson; Mendoza-de Gyves et al., 2007), medlar (Mespilus germanica L.; Mendoza-de Gyves et al., 2008), persimmon (Diospyros kaki L.f.; Cristofori et al., 2008), hazelnut (Corylus avellana L.; Cristofori et al., 2007), small fruits (Fallovo et al., 2008), zucchini squash (Cucurbita pepo L.; Rouphael et al., 2006) , sunflower (Helianthus annuus L.; Rouphael et al., 2007), and eggplant (Solanum melongena L.; Rivera et al., 2007). Our results indicated that the proposed model could predict rose LA accurately independently from species (R. hybrida and R. sempervirens) and growing conditions (ambient light level vs. 50% shade).

CONCLUSIONS

To summarize, the length-width model proposed by Rouphael et al. (2010c) can provide accurate estimations of rose leaf area regardless of the genetic materials and the growing conditions. Because leaf width and midvein length are dimensions that can be easily measured in the field, greenhouse and pod experiments, use of this model would enable researchers to make non-destructive measurements or repeated measurements on the same leaves. Such model can estimate accurately and in large amounts the leaf area of rose plants in many experimental comparisons without the use of any expensive instruments, e.g., a leaf area meter or digital camera with image measurement software.

 

LITERATURE CITED

Antunes, W.C., M.F. Pompelli, D.M. Carretero, and F.M. DaMatta. 2008. Allometric models for non-destructive leaf area estimation in coffee (Coffea arabica and Coffea canephora). Annals of Applied Biology 153:33-40.         [ Links ]

Celik, H., and S. Uzun. 2002. Validation of leaf area estimation models (Uzcelik-1) evaluated for some horticultural plants. Pakistan Journal of Botany 34:41-46.         [ Links ]

Cristofori, V., C. Fallovo, E. Mendoza-De Gyves, C.M. Rivera, C. Bignami, and Y. Rouphael. 2008. Non-destructive, analogue model for leaf area estimation in persimmon (Diospyros kaki L.f.) based on leaf length and width measurement. European Journal of Horticultural Science 73:216-221.         [ Links ]

Cristofori, V., Y. Rouphael, E. Mendoza-De Gyves, and C. Bigniami. 2007. A simple model for estimating leaf area of hazelnut from linear measurements. Scientia Horticulturae 113:221-225.         [ Links ]

Demirsoy, H., L. Demirsoy, S. Uzun, and B. Ersoy. 2004. Nondestructive leaf area estimation in peach. European Journal of Horticultural Science 69:144-146.         [ Links ]

Dent, J.B., and M.J. Blackie. 1979. Systems simulation in agriculture. Applied Science Publishers, London, UK.         [ Links ]

De Swart, E.A.M., R. Groenwold, H.J. Kanne, P Stam, L.F.M. Marcelis, and R.E. Voorrips. 2004. Non-destructive estimation of leaf area for different plant ages and accessions of Capsicum annuum L. Journal of Horticultural Science & Biotechnology 79:764-770.         [ Links ]

Fallovo, C., V. Cristofori, E. Mendoza-De Gyves, C.M. Rivera, S. Fanasca, C. Bignami, et al. 2008. Leaf area estimation model for small fruits from linear measurements. HortScience 43:2263-2267.         [ Links ]

Fascella, G., P. Maggiore, G. Zizzo, G. Colla, and Y. Rouphael. 2009. A simple and low-cost method for leaf area measurement in Euphorbia x lomi Thai hybrids. Advances in Horticultural Science 23:57-60.         [ Links ]

Giuffrida, F., Y. Rouphael, S. Toscano, D. Scuderi, D. Romano, C.M. Rivera, et al. 2011. A simple model for nondestructive leaf area estimation in bedding plants. Photosynthetica 495:380-388.         [ Links ]

Kandiannan, K., U. Parthasarathy, K.S. Krishnamurthy, C.K. Thankamani, and V. Srinivasan. 2009. Modelling individual leaf area of ginger (Zingiber officinale Roscoe) using leaf length and width. Scientia Horticulturae 120:532-537.         [ Links ]

Kumar, R. 2009. Calibration and validation of regression model for non-destructive leaf area estimation of saffron (Crocus sativus). Scientia Horticulturae 122:142-145.         [ Links ]

Mendoza-de Gyves, E., V. Cristofori, C. Fallovo, Y. Rouphael, and C. Bignami. 2008. Accurate and rapid technique for leaf area measurement in medlar (Mespilus germanica L.) Advances in Horticultural Science 22:223-226.         [ Links ]

Mendoza-de Gyves, E., Y. Rouphael, V. Cristofori, and F. Rosana Mira. 2007. A non-destructive, simple and accurate model for estimating the individual leaf area of kiwi (Actinidia deliciosa). Fruits 62:171-176.         [ Links ]

Rivera, C.M., Y. Rouphael, M. Cardarelli, and G. Colla. 2007. A simple and accurate equation for estimating individual leaf area of eggplant from linear measurements. European Journal of Horticultural Science 70:228-230.         [ Links ]

Rouphael, Y., M. Cardarelli, N. Ajouz, A. Marucci, and G. Colla. 2010a. Estimation of leaf number of eggplant using thermal time model. Journal of Food Agriculture and Environment 8:847-850.         [ Links ]

Rouphael, Y., and G. Colla. 2004. Modelling the transpiration of greenhouse zucchini crop grown under Mediterranean climate using the Penman-Monteith equation and its simplified version. Australian Journal of Agricultural Research 55:931-937.         [ Links ]

Rouphael, Y., and G. Colla. 2005. Radiation and water use efficiencies of greenhouse zucchini squash in relation to different climate parameters. European Journal of Agronomy 23:183-194.         [ Links ]

Rouphael, Y., G. Colla, S. Fanasca, and F. Karam. 2007. Leaf area estimation of sunflower leaves from simple linear measurements. Photosynthetica 45:306-308.         [ Links ]

Rouphael, Y., A.H. Mouneimne, A. Ismail, E. Mendoza-de Gyves, C.M. Rivera, and G. Colla. 2010c. Modeling individual leaf area of rose (Rosa hybrida L.) based on leaf length and width measurement. Photosynthetica 48:9-15.         [ Links ]

Rouphael, Y., A.H. Mouneimne, C.M. Rivera, M. Cardarelli, A. Marucci, and G. Colla. 2010b. Allometric models for non-destructive leaf area estimation in grafted and ungrafted watermelon (Citrillus lanatus Thunb.) Journal of Food Agriculture and Environment 8:161-165.         [ Links ]

Rouphael, Y., C.M. Rivera, M. Cardarelli, S. Fanasca, and G. Colla. 2006. Leaf area estimation from linear measurements in zucchini plants of different ages. Journal of Horticultural Science & Biotechnology 81:238-241.         [ Links ] Ç

Salerno, A., C.M. Rivera, Y. Rouphael, G. Colla, M. Cardarelli, F. Pierandrei, et al. 2005. Leaf area estimation of radish from linear measurements. Advances in Horticultural Science 19:213-215.         [ Links ]


Schwarz, D., and H.P. Klaring. 2001. Allometry to estimate leaf area of tomato. Journal of Plant Nutrition 24:1291-1309.         [ Links ]

Serdar, U., and H. Demirsoy. 2006. Non-destructive leaf area estimation in chestnut. Scientia Horticulturae 108:227-230.         [ Links ]

Spann, T.M., and R.J. Heerema. 2010. A simple method for nondestructive estimation of total shoot leaf area in tree fruit crops. Scientia Horticulturae 125:528-533.         [ Links ]

Stoppani, M.I., R. Wolf, N. Francescangeli, and H.R. Marti. 2003. A non-destructive and rapid method for estimating leaf area of broccoli. Advances in Horticultural Science 17:173-175.         [ Links ]

Tsialtas, J.T., S. Koundouras, and E. Zioziou. 2008. Leaf area estimation by simple measurements and evaluation of leaf area prediction models in cabernet-Sauvignon grapevine leaves. Photosynthetica 46:452-456.         [ Links ]


Received: 17 July 2012.
Accepted: 29 December 2012.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License