of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  363 Non-Invasive Method to Monitor Plant Water Potential of an Olive Orchard Using Visible and Near Infrared Spectroscopy Analysis C. Poblete-Echeverría a  and S. Ortega-Farías Centro de Investigación y Transferencia en Riego y Agroclimatología (CITRA) Universidad de Talca Chile G.A. Lobos, S. Romero, L. Ahumada and A. Escobar Centro de Mejoramiento Genético y Fenómica Vegetal (CMGFV) Universidad de Talca Chile S. Fuentes The University of Melbourne Melbourne School of Land and Environment Victoria 3010 Australia Keywords: spectrometer, olive, wavelength, climate change, pressure chamber Abstract Plant water status is an important factor to be quickly and accurately assessed temporally and spatially to maintain yield and quality standards in a changing environment. The pressure chamber is the most common method to measure midday stem water potential (MSWP), which is used as an accurate indicator of plants water status. However, the use of the latter method has the disadvantages of being manual, destructive, slow and impractical to have a good spatial representation of plant water status of a whole olive orchard. The objective of this research was to test an indirect method to estimate MSWP using visible and near infrared (NIR) spectroscopy analysis. This method is non-destructive and quick to implement in the field having the extra advantage of assessing the whole tree with a single measurement using short-range remote sensing. In this study, MSWP was measured using a pressure chamber (PMS Instrument Co. USA) to characterize water status of olive trees in a commercial drip-irrigated ‘Arbequina’ orchard (Pencahue, Region del Maule, Chile). Canopy reflectance was measured between 350 and 2500 nm wavelength region with a spectrometer FieldSpect3 (Analytical Spectral Device, ASD Co., USA) from 2.5 m distance. The spectroscopy analysis was done using partial least squares (PLS) regression by the Unscramble software. The validation process showed a root mean square error (RMSE) equal to 0.88 MPa and coefficient of determination (R  2 ) equal to 0.75. These results showed that the non-invasive method, using spectral reflectance data, could obtain a surrogate of stem water potential of olive trees, providing the possibility to screen more samples under field conditions and to automate the data gathering and analysis once a calibration curve is developed. INTRODUCTION Under a climate change scenario, plant water status monitoring has become a critical factor for maintaining yield and quality standards in an international competitive market. The pressure chamber is the most common method to measure plant water  potential, which is used as an accurate indicator of olive water status (Fernandez et al., 2006; Martin-Vertedor et al., 2011). In this sense, leaf water potential measurements are considered an accurate and reliable technique to characterize crop water stress, which has a Proc. VII th  IS on Olive Growing Eds.: F. Vita Serman et al. Acta Hort. 1057, ISHS 2014  364  been commonly used for irrigation scheduling (e.g., Alchanatis et al., 2010). However, leaf water status within a canopy is variable because it depends on the transpiration rate that a particular leaf has at the moment of measurement (Hsiao, 1990). As a solution to this problem, measurements of midday stem water potential (MSWP) were developed to integrate the whole-plant water status (Hsiao, 1990). Nevertheless, traditional methods for measuring MSWP require destructive sampling and pretreatment for up to one or two hours. So, the use of pressure chamber for measuring MSWP results unpractical for  precision agriculture, since a high number of measurements are required to have a good spatial representation of the whole olive orchard (Jones et al., 2004).  Nowadays, spectroscopy analysis has gained attention in agriculture due to the  possibility of measuring, via reflectance the portion of incident energy that is reflected by an object (He et al., 2005; Raun et al., 2001; Brown et al., 2006). Vegetative tissues reflect poorly in the visible spectrum, but reflect an important amount in the non-visible spectrum of near infrared (NIR) (Pinter et al., 2003). Spectroscopy analysis has been implemented to obtain vegetation indices that combine the reflectance values in two or more contrasting wavelengths. One of the most used in agriculture is the normalized vegetative index (NDVI), which is obtained from reflectance in NIR and RED light regions. Some other indices have been developed to determine water status in different tissues and plants (Pinter et al., 2003; Rodríguez-Pérez et al., 2007). Recent studies have shown that hyperspectral sensors can be helpful to estimate water status. For example in grapevines Santos and Kaye (2009) attempted to use NIR spectroscopy to assess leaf water potential in ‘Syrah’, ‘Merlot’ and ‘Cabernet Sauvignon’, Vila et al. (2010) estimated leaf water potential using spectral reflectance from 325 to 1075 nm in ‘Malbec’, De Bei et al. (2012) estimated midday stem water potential using near infrared spectroscopy in ‘Cabernet Sauvignon’, ‘Shiraz’ and ‘Chardonnay’. In addition, developments in spectroscopic instrumentation and software analysis have made it possible to rapidly perform multivariate analyses of data. In spectroscopy, calibration is a key process, which uses multivariate regression techniques to relate reflectance at selected wavelengths with reference or ground-truth values. Once calibrated, the advantages of spectroscopy analysis are the speed of the analysis, simplicity in sample  preparation, multiplicity of analysis, non-destructive and the non-requirement for the use of chemical reagents (Williams, 2001). In order to allow the analysis of a great number of variables simultaneously, several calibrations may be required (Vila et al., 2010). Among the calibrations, one of the most used is the partial least squares regression (PLS). PLS is a statistical method used to construct regression equations, especially to be applied when the number of independent variables is much bigger than the number of cases, and when these variables are not independent (Hoskuldsson, 1988; Abdi, 2007). The objective of this research was to develop an indirect method to estimate MSWP of an olive orchard using spectroscopy analysis with multivariate analysis (PLS). MATERIALS AND METHODS In this study, MSWP was measured using a pressure chamber (PMS Instrument Co. USA) to characterize water status of olive trees in a commercial drip-irrigated olive (‘Arbequina’) orchard located in the Pencahue valley, Region del Maule, Chile during the 2011-2012 season. MSWP measurements were made at midday (1200 to 1400 h, solar time) on two shoots for each tree chosen from the mid-upper part of the canopy. Canopy reflectance was measured between 350 and 2500 nm wavelength region with a spectrometer FieldSpect 3 (Analytical Spectral Device, ASD Co., USA) at 2.5 m distance. The spectrum of for each sample was considered as the average of three successive scans in the same canopy. Radiometric calibration was obtained using a standard white panel (Spectralon ® , Analytical Spectral Devices, Boulder, Colorado, USA). The software RS3 TM (Analytical Spectral Devices, Boulder, Colorado, USA) was used to control the spectrophotometer and to acquire the spectral data. The spectroscopy analysis was performed using partial least squares (PLS) regression using the  365 Unscramble ®  software (CAMO ASA, Oslo, Norway) (De Bei et al., 2012). In this study, three dates were used to calibrate the model and other three dates were used to carry out an independent validation process. The performance of the calibration model was evaluated using the root mean square error of calibration (RMSEc), standard error of calibration (SEc) and the determination coefficient (R  2 ). Finally, the prediction ability of the model was evaluated using root mean square error of prediction (RMSEp), standard error of prediction (SEp) and R  2 . RESULTS AND DISCUSSION Figure 1 shows an example of reflectance spectral data obtained during the 2011-2012 season from canopies of the olive trees for different irrigation regimes obtained. In general, olive canopy spectral reflectance was low in the visible light band (350-700 nm), high in the near infrared (NIR) band (700-1300 nm) and intermediate in the mid-infrared (MIR; 1300-2500 nm). Reflectance depends in part on the water stored in the leaf cells, especially for NIR and MIR (Gao, 1996; Ceccato et al., 2001).  Non-stressed olive trees showed a higher reflectance in the whole spectrum compared with stressed trees. However, the general trend of spectral curves was consistent, with same shape of peaks and valleys. Vila et al. (2010) showed that, stressed grapevines presented lower reflectance values in NIR region than non-stressed plants. Dates used for calibration (1, 3 and 5) showed an average value of MSWP of -2.72 MPa with minimum and maximum values of -5.85 and -1.25 MPa, respectively. In the case of date set used for the independent validation the average value of MSWP was -3.0 MPa with minimum and maximum values of -6.45 and -1.40 MPa, respectively (Table 1). Also, Table 1 shows that the dataset used for validation process presented higher values of standard deviation, coefficient of variation and range, since the olive trees were under high stress level on date 4. The PLS regression model used in this study was configured with six factors (principle components). The statistical analysis showed that the calibrated model had RMSEc and SEc values equal to 0.48 and 0.49 MPa, respectively (Table 2). Figure 2 shows that the comparison points in calibration were distributed closely to the regression line with R  2  of 0.85. On the other hand, RMSEp and SEp obtained in the statistical analysis of validation process were 0.88 and 1.15 MPa, respectively. These results are in agreement with those obtained by De Bei et al. (2011) in the MSWP estimation of grapevines using NIR (750-1050 nm) region. The latter study found standard error in cross validation (SEcv) of 0.11 to 0.23 MPa for ‘Shiraz’ and ‘Cabernet Sauvignon’ cultivars. Finally, Figure 3 shows that the comparison between predicted and measured MSWP in the validation process. In this case, the comparison showed a lower R  2  of 0.75 and some scattered values near -4.0 to -5.0 MPa. CONCLUSIONS The results of this study indicated that a non-invasive method using spectral reflectance data could be an alternative method for monitoring midday stem water  potential of olive trees. The advantages of this new approach are speed, non-destructive and low cost of analysis, providing the possibility to screen more samples under field conditions. Nevertheless, this technique requires complex statistical analyses in order to obtain a calibrated model. ACKNOWLEDGEMENTS The research leading to this report was supported by the scientific projects FONDECYT POSTDOCTORAL (Nº 3100128), CONICYT (Nº 79090035), FONDEF (D10I1157) and the research program “Adaptation of Agriculture to Climate Change (A2C2)” of the Universidad de Talca.  366 Literature Cited Alchanatis, V., Chone, Y., Cohen, S., Moeller, M., Sprinstin, M., Meron, M., Tsipris, J., Saranga, Y. and Sela, E. 2009. Evaluation of different approaches for estimating and mapping crop water status in cotton with thermal imaging. Preci. Agri. 11:27-41. Brown, D.J., Shepherd, K.D., Walsh, M.G., Mays, M.D. and Reinsch, T.G. 2006. Global soil characterization with VNIR diffuse reflectance spectroscopy. Geoderma 132:273-290. Ceccato, P., Flasse, S., Tarantola, S., Jacquemoud, S. and Gregoire, J.M. 2001. Detecting vegetation leaf water content using reflectance in the optical domain. Rem. Sens. Envi. 77: 22-33. Choné, X.C., Van Leeuwen, C., Dubourdieu, D. and Gaudillère, J.P. 2001. Stem water 575 potential is a sensitive indicator of grape water status. Ann. Bot. 87:477-483. De Bei, R., Cozzolino, D., Sullivan, W., Cynkar, W., Fuentes, S., Dambergs, R., Pech, J. and Tyerman, S. 2011. Non-destructive measurement of grapevine water potential using near infrared spectroscopy. Aus. J. of Grape and Wine Res. 17(1):62-71. Fernandez, J.E., Diaz-Espejo, A., Infante, J.M., Duran, P., Palomo, M.J., Chamorro, V., Giron, I.F. and Villagarcia, L. 2006. Water relations and gas exchange in olive trees under regulated deficit irrigation and partial rootzone drying. Plant and Soil 284:273-291. Gao, B.C. 1996. NDWI-A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Envi. 58:257-266. He, Y., Zhang, Y., Pereira, A.G., Gómez, A.H. and Wang, J. 2005. Nondestructive determination of tomato fruit quality characteristics using Vis/NIR spectroscopy technique. Inter. J. of Inf. Tech. 11(11). Hoskuldsson, P. 1988. PLS regression methods. J. of Chem. 2:211-228. Hsiao, T.C. 1990. Measurement of plant water status. p.243-279. In: B.A. Stewart and D.C. Nielsen (eds.), Irrigation of Agricultural Crops, Special Publication No. 30. Amer. Soc. Agron., Madison, Wis., USA. Jones, C.L., Weckler, P.R., Maness, N.O., Stone, M.L. and Jayasekara, R.J. 2004. Estimating water stress in plants using hyperspectral sensing. Proceedings for the ASAE/CSAE. Annual Inter. Meeting, Paper 043065, p.2-11. Martín-Vertedor, A., Pérez Rodríguez, J., Prieto, L. and Fereres, E. 2011. Interactive responses to water deficits and crop load in olive ( Olea europaea  L., cv. Morisca). Growth and water relations. Agri. Water Manag. 98:941-949.  Naes, T., Isaksson, T., Fearn, T. and Davies, A.M.C. 2002. A User-Friendly Guide to Multivariate Calibration and Classification. NIR Publications: Chichester, UK. Pinter, P.J., Hatfield, J.L., Schepers, J.S., Barnes, E.M., Moran, M.S., Daughtry, C.S.T. and Upchurch, D.R. 2003. Remote sensing for crop management. Photogrammetric Engineering and Remote Sensing 69(6):647-664. Raun, W., Solie, J.B., Johnson, G.V., Stone, M.L., Lukina, E.V., Thomason, W.E. and Schepers, J.S. 2001. In-season prediction of potential grain yield in winter wheat using canopy reflectance. Agronomy J. 93:131-138. Rodríguez-Pérez, J.R., Riaño, D., Carlisle, E., Ustin, S. and Smart, D.R. 2007. Evaluation of hyperspectral reflectance indexes to detect grapevine water status in vineyards. Amer. J. of En. and Vit. 58(3):302-317. Santos, A.O. and Kaye, O. 2009. Grapevine leaf water potential based upon near infrared spectroscopy. Sci. Agricola 66:287-292. Scholander, P.F., Bradstreet, E.D., Hemmingsen, E.A. and Hammel, H.T. 1965. Sap  pressure in vascular plants, negative hydrostatic pressure can be measured in plants. Science 148(3668):339-346. Vila, H., Hugalde, I. and Di Filippo, M. 2011. Estimation of leaf water potential by thermographic and spectral measurements in grapevine. R.I.A. 37(1):46-52. Williams, P.C. 2001. Implementation of near-infrared technology. p.145-169. In: P.C. Williams and K.H. Norris (eds.), Near Infrared Technology in the Agricultural and Food Industries. American Association of Cereal Chemist: St. Paul.  367 Tables Table 1. Summary of the descriptive statistical analysis of midday stem water potential (MSWP) measurements used in calibration and validation process. Data set Avg. S.D. C.V. Min. Max. Range Calibration -2.72 1.30 48.1 -5.85 -1.25 4.6 Validation -3.09 1.58 51.1 -6.45 -1.40 5.0 Avg. is the average (MPa); S.D. is the standard deviation (MPa); C.V. is the coefficient of variation (%); Min. minimum (MPa); Max. is the maximum (MPa). Table 2. Calibration and validation statistics of partial least squares regression. Data set Slope Intercept R  2  RMSE SE Calibration 0.85 -0.38 0.85 0.48 0.49 Validation 0.73 -0.41 0.70 0.88 1.15 R    2  is the determination coefficient (dimensionless); RMSE is the root means square error (MPa); SE is the standard error (MPa). Figures Fig. 1. Example of reflectance spectral data for different irrigation regimes from the olive canopy. Wavelengths with high atmospherics absorbance (from 1790 to 1940 nm) were not included in plot.         3        5        0        4        1        9        4        8        8        5        5        7        6        2        6        6        9        5        7        6        4        8        3        3        9        0        2        9        7        1        1        0        4        0        1        1        0        9        1        1        7        8        1        2        4        7        1        3        1        6        1        3        8        5        1        4        5        4        1        5        2        3        1        5        9        2        1        6        6        1        1        7        3        0        1        7        9        9        1        8        6        8        1        9        3        7        2        0        0        6        2        0        7        5        2        1        4        4        2        2        1        3        2        2        8        2        2        3        5        1        2        4        2        0    A   b  s  o   l  u   t  e  r  e   f   l  e  c   t  a  n  c  e   (  -   ) Spectral wavelength (nm) T0T1T2T3
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks