HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Search for Related Content PubMed Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Agricola Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Related Collections Turfgrass Management Water Stress Irrigation

Predicting Soil Water Content through Remote Sensing of Vegetative Characteristics in a Turfgrass System

^a Univ. of Florida, P.O. Box 110670, Gainesville, FL 32611
^b Dep. of Horticulture, Iowa State Univ., 106 Horticulture Hall, Ames, IA 50011. This journal paper of the Iowa Agric. and Home Econ. Exp. Stn., Ames, Iowa, Project No. 3601 was supported by Hatch Act and State of Iowa funds

^* Corresponding author (jkdk{at}ufl.edu).

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Scouting to determine soil water status throughout a golf course or large athletic field complex is quite time consuming and requires numerous observations to characterize variability across the site. The objective of this research was to evaluate the use of a ground-based remote sensing system to predict soil water content through partial least squares regression analysis of canopy reflectance data collected from perennial ryegrass (Lolium perenne L.) maintained at 12.7 mm and creeping bentgrass (Agrostis stolonifera L.) maintained at 6.3 mm during 2002 and 2003 on a Coland silty clay loam. Volumetric soil water at a 5 cm depth was measured by time domain reflectometry and was collected in conjunction with spectral radiance measurements obtained using a fiber optic spectrometer. Volumetric soil water content was best predicted with partial least squares regression analysis of creeping bentgrass canopy reflectance data with a maximum r² of 0.64 (P < 0.001) 1 d before development of drought stress symptoms. Similar results were observed for canopy reflectance data collected from perennial ryegrass plots, indicating that this technology and method of data analysis may be useful in the development of automated turfgrass irrigation management systems.

Abbreviations: ET, evapotranspiration • MLR, multiple linear regression • NIR, near-infrared • PC, principle component • PCA, principle component analysis • PLS, partial least squares • PRESS, predicted residual sum of squares • SEP, standard error of prediction • TDR, time domain reflectometry

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
PROPER SOIL WATER CONTENT is essential to maintaining the quality and playability of turfgrass areas. While rainfall in many parts of the United States is ample, it may not be evenly distributed, thus requiring the use of irrigation to prevent drought stress on turfgrass areas. Rising demand for available water resources increases the cost of available water. In addition, increased public awareness regarding the potential for environmental contamination has led turfgrass managers to seek to improve irrigation efficiency while limiting environmental risk related to nutrient and pesticide applications. As a result, it is increasingly important for turfgrass managers to carefully monitor the soil water status of their site and make adjustments in irrigation practices in response to water stress to make the best use of this natural resource. This is often accomplished through several methods, including soil-, meteorological-, and crop-based techniques. Soil-based techniques schedule irrigation events on the basis of the soil water status of the rooting zone, which is often determined by removing soil cores to estimate the soil water via the feel and appearance of the soil (USDA-NRCS, 1998). Irrigation rate and timing can also be determined through collecting meteorological data such as air temperature, vapor pressure, wind speed, and net radiation to estimate daily evapotranspiration (ET). Plant-based measurements that can be used to determine the water status of plants include leaf water potential, stomatal conductance, and canopy temperature (Brown et al., 2004; DaCosta and Huang, 2006). While these methods provide valuable information that can be used to determine the need for irrigation, they are time consuming and require numerous samples to characterize the variations across a site (Jackson, 1982).

Turfgrass managers have already increased their use of available management tools through the use of on-site weather stations and by scheduling irrigation application and amount based on calculated ET. However, these methods do not take into account the wide spatial variability in soil characteristics present in many turfgrass systems. As a result, changes are often made to entire zones of the irrigation system to account for localized problems. Irrigation systems with valve-in-head control allow for site-specific modifications to the irrigation rate and timing. By monitoring changes in water stress on a spatial basis through scouting or the use of remote sensing tools, managers can make modifications of irrigation at the individual sprinkler head level, targeting the applications of water to areas with the most need.

One limitation of traditional remote sensing systems like aerial photography and satellite imaging is that the shade from trees that is commonly cast across turfgrass areas results in altered spectral quality and intensity (Bell et al., 2000). Many handheld and ground-based remote sensing systems include ambient light sensors, integrated light emitting diodes, or other auxiliary light sources to reduce or eliminate the effect of fluctuations in ambient radiation on the data collected by the instruments (Bell et al., 2000).

Few studies have been reported in the literature on the use of remote sensing to detect turfgrass water stress (Fenstermaker-Shaulis et al., 1997; Horst et al., 1991; Jones et al., 1992; Kenna, 1995; Nutter et al., 1991). The majority of remote sensing literature has focused on detection of growth responses and plant stress in agricultural crops (Aase et al., 1986; Asrar et al., 1985; Deering et al., 1975; Eckardt et al., 1982; Kleman and Fagerlund, 1987; Wanjura and Hatfield, 1987), turf chlorophyll content (Howell, 1999), and turf injury and quality (Bell et al., 2000; Trenholm et al., 1999). Repeatedly, results of these studies indicate that remote sensing instruments can reliably detect plant stress by monitoring changes in reflectance in the vegetation. As soil water decreases, plants exhibit a decrease in tissue water content, which in turn influences their reflectance properties (Carter, 1994; Fenstermaker-Shaulis et al., 1997; Osborne et al., 2002). Leaf reflectance in the visible portion of the spectrum (400–700 nm) is relatively low due to increased absorption of energy by chlorophyll and is correlated (r² > 0.97) with concentration of leaf pigments (Horler et al., 1983). As plants become stressed, they exhibit decreased reflectance in the near-infrared (NIR) spectral region due to decreased cell layers and an increased reflectance in the red spectral region due to lower chlorophyll concentrations (Carter, 1993; Guyot, 1990). Regular monitoring of these changes in spectral reflectance may reliably identify changes in plant growth or physiological status that result from biotic and/or abiotic stresses (Carter, 1993; Carter and Miller, 1994).

Drought stress has been associated with a decline in turfgrass quality associated with reductions in root growth, leaf water potential, cell membrane stability, photosynthetic rate, photochemical efficiency, and carbohydrate accumulation (Carrow, 1996; Perdomo et al., 1996; Huang et al., 1998; Huang and Gao, 1999; Jiang and Huang, 2000). Irrigation systems are often used to prevent drought stress in turfgrass systems. Adjustments to irrigation frequency and duration are made to ensure that areas that typically show the first signs of drought stress receive adequate water. As a result, many of the surrounding areas receive more water than required to maintain optimum quality. A reasonable goal for a remote sensing system is to use it for predicting the soil water status on a site-specific basis based on canopy reflectance before drought stress symptoms develop. This would allow for site-specific irrigation applications through modification of the irrigation system.

Turfgrass managers are often limited to applying irrigation during the early morning hours because of intense use of the site by patrons. Effectively scheduling irrigation on a site-specific basis will require detecting those areas where the plants are beginning to experience drought stress, although they may lack visible symptoms. By identifying these areas before visible symptoms of drought stress develop, managers can schedule an irrigation event and minimize stress to the turfgrass plants.

Research often involves the use of controllable variables (factors) to explain or predict other variables (responses). When the factors used to explain the variation in the data are few in number, not highly collinear, and have a well-understood relationship to the responses, multiple linear regression (MLR) can be a good way to develop models for data analysis (Tobias, 1997). Much of the remote sensing research has focused on the use of multispectral systems that acquire images in a few broad (>50 nm) spectral bands that are well suited to analysis using methods such as MLR. In contrast, hyperspectral spectrometers offer the ability to collect a continuous spectrum in many narrow (<10 nm) spectral bands. Because of the narrow spectral bands, the factors used are measurements from the spectrum that can number in the hundreds or thousands and are likely to be highly collinear. When using MLR to analyze data that is highly collinear, it is easy to produce a model that fits the data from the sample set perfectly while having little use in predicting result from new samples (Tobias, 1997). When this occurs, the model is said "over-fit" the data set. These are cases in which there are many factors but only a small number of these factors account for most of the variation in the response. Principle component analysis (PCA) has been used as a method to reduce explanatory variables, after which a regression is run on the principle component (PC) scores (Hatcher, 1994). A limitation to the use of PCA for variable reduction is that the PC scores are chosen without consideration of the response variable.

Partial least squares (PLS) regression is a statistical method developed for data compression and construction of predictive models when there are a large number of highly collinear factors (Tobias, 1997; Hansen and Schjoerring, 2003). During the calculation of PLS, the X- and Y-scores are chosen so that the relationship between successive pairs of scores are as strong as possible, resulting in the selection of a few noncorrelated principle components, otherwise known as factors. The factors represent the relevant structural information present in the reflectance measurements collected from the plant canopy. The predicted residual sum of squares (PRESS) statistic is based on the residuals generated by this process. Through cross-validation, the smallest model that has a PRESS statistic insignificantly larger than the absolute minimum PRESS statistic is selected. As a result, a PLS regression is not based on a single or even a few frequencies, as would be the case with MLR or stepwise regression (Tobias, 1997).

The objectives of this research were to determine if canopy reflectance data can be used to predict the soil water content in a turfgrass system through the use of PLS regression and to determine the relationship between canopy reflectance data, soil water content, and visual quality of water-stressed turfgrass plants.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Two field studies were conducted on nine in-play fairways and a creeping bentgrass (Agrostis stolonifera L.) tee at Veenker Memorial Golf Course in Ames, IA. The purpose was to investigate the relationship between turf quality, volumetric soil water content, and multispectral remote sensing data collected from the turfgrass canopy on a perennial ryegrass (Lolium perenne L.) fairway and a creeping bentgrass tee.

Perennial Ryegrass Study
A 2-yr study was conducted on fairways established with a perennial ryegrass mix (35.7% ‘Divine’, 32.6% ‘Majesty’, 29.9% ‘Enchanted’, 0.01% other crop, and 1.73% inert matter) on a Coland silty clay loam (fine loamy mixed mesic Cumulic Haplaquoll) (Fig. 1 , Table 1 ). Treatments were organized into a randomized complete block design with three replications. Blocking was based on topographical location in relation to the creek bed that transects the course. Plots were 18.3 by 18.3 m in size and located within the perimeter established by four irrigation heads on each of nine separate fairways, with each fairway representing an independent experimental unit (plot). Data were collected from the central region of each plot, avoiding border areas and the potential for variation due to drift.

View larger version (9K): [in this window] [in a new window]   Figure 1. Soil water release curve for the Coland silty clay loam soil relating pressure head to the volumetric soil water content.

Fertilization, mowing, and irrigation were applied uniformly across all plots before initiation of irrigation treatments. All plots received nitrogen at a rate of 48.8 kg ha^–1 and potassium at a rate of 39.2 kg ha^–1 on or about 20 April, 15 July, and 15 October each year. In addition, plots were treated with a growth regulator, trinexapac-ethyl [4-(cyclopropyl-&alpha-hydroxy-methylene)-3,5-dioxocyclohexanecarboxylic acid ethyl ester] (Primo Maxx, Syngenta Crop Protection, Greensboro, NC), at a rate of 763 g ha^–1 every 28 d as part of the normal golf course maintenance program to reduce clipping production. Turf was mowed 3 d wk^–1 at a height of 12.7 mm with clippings returned to the site.

Creeping Bentgrass Study
Treatments were applied to a stand of ‘Penn A-4’ creeping bentgrass established on a recently disturbed Coland soil (fine loamy mixed mesic Cumulic Haplaquoll) which had been leveled before establishment by seed in Jul. 2001. Four irrigation treatments (0, 25, 50, and 75% ET) were applied nightly to four plots that were 9.1 x 9.1 m in size. Treatments continued until more than 50% by area of the 0% ET plots showed signs of wilt, at which point treatments were ceased and all plots were returned to 100% ET to prevent unwanted damage to the golf course. Data was collected from the central region of each plot, avoiding border areas and the potential for variation due to drift. All treatments were replicated in time after re-randomization and repetition of the study following a minimum of 5 d at 100% ET to re-establish uniform soil water conditions across plots. Irrigation treatments were applied from 25 June to 29 June 2002, 5 July to 9 July 2002, 20 July to 25 July 2002, 1 Aug. to 2 Aug. 2002, 22 July to 26 July 2003, 9 Aug. to 13 Aug. 2003 and 5 Sept. to 10 Sept. 2003.

Fertilization, mowing, and irrigation were applied uniformly across all plots before the initiation of irrigation treatments. All plots received nitrogen at a rate of 48.8 kg ha^–1 and potassium at a rate of 39.2 kg ha^–1 on or around 20 April, 15 July, and 15 Oct. each year. In addition, plots were treated with a growth regulator, Primo Maxx, at a rate of 763 g ha^–1 every 28 d as part of the normal golf course maintenance program to reduce clipping production. Turf was mowed 4 d wk^–1 at a height of 6.3 mm, and clippings were removed from the site.

Data Collection and Analysis
Data collection began at the initiation of treatments during each study period and continued daily until signs of wilt became evident on the plots receiving no irrigation (0% ET). Plots were evaluated for visual quality on the basis of color, shoot density, and uniformity of stand, where 1 = no live grass, 6 = minimally acceptable, and 9 = dark-green, dense, uniform grass. In addition, plots were evaluated for development of visible drought stress symptoms on a scale of 1 to 5, where 1 = no signs of drought stress, 3 = visible symptoms of wilt on at least 50% of the plot area, and 5 = permanent wilt. All visual data were collected immediately before collection of soil water readings and canopy reflectance data from each plot.

Remotely sensed data was collected daily during the application of irrigation treatments with a field portable fiber optic spectrometer fitted with 30-degree field of view optics (Model S2000, Ocean Optics, Inc., Winter Park, FL). To reduce variability resulting from cloud cover and solar zenith angle, the tip of the fiber was mounted inside a rectangular plastic hood that extended down to the turf canopy, blocking out the ambient solar radiation. Auxiliary lighting was provided by two 12-V halogen lights that emitted an irradiance of 2250 µmol m^–2 s^–1. Radiance values were expressed as percentage spectral reflectance after standardizing the spectrometer with a white standard. The spectrometer has a nominal spectral range from 400 to 1050 nm with approximately 0.3-nm nominal bandwidth. Thus, for each measurement, the spectrometer program automatically collects 2500 data points covering the calibrated spectral range. Approximately 65 measurements were collected from each plot every day during the study, with each measurement calculated as an average of eight consecutive scans that were taken on a 0.1-s interval. Recalibration of the instrument with the white standard was conducted immediately before collecting measurements from each plot to minimize electrical noise and data variability. To simplify data analysis and interpretation, a linear interpolation routine was used to estimate values at a 1-nm interval before analysis of the reflectance data. Canopy reflectance was measured daily between 1400 and 1800 h central standard time. The volumetric water content for each plot was calculated as an average from 20 random locations within each plot through time domain reflectometry (TDR) done in conjunction with collection of the remotely sensed data using 5 cm long probes inserted vertically at a 90 degree angle to the soil surface. Volumetric water content was measured with a three-rod probe of 5-cm length automated with TDR (TDR-100, Campbell Scientific). The probe was constructed and calibrated as described by Heimovaara (1993).

An analysis of variance was performed to test treatment effect on soil water status and visual quality. Mean separation was by Waller–Duncan k-ratio t test at the 0.05 probability level. Analyses were performed using SAS (SAS Institute, 2004). Leaf spectral reflectance data sets were analyzed relative to the soil water content using PLS regression as performed by SAS (SAS Institute, 2004). Equations were validated through split-sample cross-validation. For the cross-validation, 10% of the observations were left out for prediction at a time, and the number of factors that minimized the PRESS statistic was chosen (Fig. 2 ). This process was repeated so that every observation was used exactly once for cross-validation.

View larger version (10K): [in this window] [in a new window]   Figure 2. Illustration of the standard error of prediction vs. the number of factors in the partial least squares regression model.

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

The PLS model for creeping bentgrass plots 1 d before the development of visual drought stress symptoms could account for 99.9% of the variability in spectral reflectance data and 99.6% of the variability in soil water content with seven factors (Table 4). Similarly, the PLS model for perennial ryegrass plots 1 d before the development of visual drought stress symptoms could account for 99.9% of the variability in the spectral reflectance data and 97.5% of the variability in soil water content with four PLS factors. The PLS models for the perennial ryegrass areas were not as sensitive to changes in soil water content as indicated by higher PRESS statistics and lower percentage variation explained in the dependent variable. It should be noted that the minimizing number of factors (the number of factors needed to minimize PRESS) selected was not significantly different from the smallest number of factors needed to predict the variability in Y.

The loading curves from the PLS regression of perennial ryegrass spectral data 1 d before development of visual drought stress symptoms indicate that the 750- to 1100-nm range of the spectrum exhibited a higher degree of loading than the range between 400 and 750 nm (Fig. 3 ). The second factor showed separation between the visible and NIR bands. The third and fourth factors showed high degrees of loading around the far-red (690–710 nm) and red-edge (710–780 nm) regions. This was an expected result since the chlorophyll absorption bands and red-edge region are known to indicate crop stress (Carter, 1993; Carter, 1994). In addition, the third and fourth factors showed a high degree of loading in the region mid-infrared region (1100–1150 nm). This region is similar to that which was reported by Sims and Gamon (2003) to be effective for use in prediction of canopy water content (1150–1260 nm) in forest canopies. Similar patterns of loading were observed in PLS regression analysis of the creeping bentgrass data.

View larger version (16K): [in this window] [in a new window]   Figure 3. Loadings for the partial least squares regression factors plotted against the spectral band wavelengths.

View larger version (20K): [in this window] [in a new window]   Figure 4. Predicted versus actual volumetric soil water content for all data collected in 2002 and 2003 1 d before and on the day of the onset of visual drought stress symptoms on perennial ryegrass (Lolium perenne L.) obtained using partial least-squares regression to relate spectral reflectance data in the visible–near-infrared wavelength range to the reference volumetric soil water content values. The graphed line represents a 1:1 relationship. SEP, standard error of prediction.

View larger version (19K): [in this window] [in a new window]   Figure 5. Predicted versus actual volumetric soil water content for all data collected in 2002 and 2003 1 d before and on the day of the onset of visual drought stress symptoms on creeping bentgrass (Agrostis stolonifera L.) obtained using partial least-squares regression to relate spectral reflectance data in the visible–near-infrared wavelength range to the reference volumetric soil water content values. The graphed line represents a 1:1 relationship. SEP, standard error of prediction.

	ACKNOWLEDGMENTS

 
We would like to thank The Toro Company and the Iowa Turfgrass Institute for financial support of this research. We would also like to extend thanks to John Newton, CGCS, for allowing this research project to be conducted on-site at Veenker Memorial Golf Course, Ames, IA.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication January 23, 2006.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 

• Aase, J.K., J.P. Millard, and B.S. Brown. 1986. Spectral radiance estimates of leaf-area and leaf phytomass of small grains and native vegetation. IEEE Trans. Geosci. Remote Sens. 24 :685–692.[CrossRef]
• Asrar, G., E.T. Kanemasu, R.D. Jackson, and P.J. Pinter. 1985. Estimation of total above-ground phytomass production using remotely sensed data. Remote Sens. Environ. 17 :211–220.[CrossRef]
• Bell, G.E., T.K. Danneburger, and M.J. McMahon. 2000. Spectral irradiance available for turfgrass growth in sun and shade. Crop Sci. 40 :189–195.[Abstract/Free Full Text]
• Brown, C.A., D.A. Devitt, and R.L. Morris. 2004. Water use and physiological response of tall fescue turf to water deficit irrigation in an arid environment. HortScience 39 :388–393.[Abstract/Free Full Text]
• Carrow, R.N. 1996. Drought avoidance characteristics of diverse tall fescue cultivars. Crop Sci. 36 :371–377.[Abstract/Free Full Text]
• Carter, G.A. 1993. Responses of leaf spectral reflectance to plant stress. Am. J. Bot. 80 :239–243.[CrossRef][Web of Science]
• Carter, G.A. 1994. Ratios of leaf reflectances in narrow wavebands as indicators of plant stress. Int. J. Remote Sens. 15 :667–703.
• Carter, G.A., and R.L. Miller. 1994. Early detection of plant stress by digital imaging within narrow stress-sensitive wavebands. Remote Sens. Environ. 50 :295–302.[CrossRef]
• DaCosta, M., and B. Huang. 2006. Deficit irrigation effects on water use characteristics of bentgrass species. Crop Sci. 46 :1779–1786.[Abstract/Free Full Text]
• Deering, D.W., J.W. Rouse, R.H. Haas, and J.A. Schell. 1975. Measuring forage production of grazing units from Landsat MSS data. p. 1169–1178. In Proc. 10th Int. Symp. on Remote Sensing of the Environment. ERIM, Ann Arbor, MI.
• Eckardt, F.E., L. Heerfordt, H.M. Jorgensen, and P. Vaag. 1982. Photosynthetic production in Greenland as related to climate, plant cover, and grazing pressure. Photosynthetica 16 :71–100.[Web of Science]
• Fenstermaker-Shaulis, L.K., A. Leskys, and D.A. Devitt. 1997. Utilization of remotely sensed data to map and evaluate turfgrass stress associated with drought. J. Turfgrass Manage. 2 :65–81.
• Guyot, G. 1990. Optical properties of vegetation canopies. p. 19–43. In M.D. Steven and J.A. Clark (ed.) Applications of remote sensing in agriculture. Butterworths, London.
• Hansen, P.M., and J.K. Schjoerring. 2003. Reflectance measurement of canopy biomass and nitrogen status in wheat crops using normalized difference vegetation indices and partial least squares regression. Remote Sens. Environ. 86 :542–553.[CrossRef]
• Hatcher, L. 1994. A step-by-step approach to using SAS for factor analysis and structural equation modeling. SAS Inst., Cary, NC.
• Heimovaara, T.J. 1993. Design of triple-wire time domain reflectometry probes in practice and theory. Soil Sci. Soc. Am. J. 57 :1410–1417.[Abstract/Free Full Text]
• Huang, B., J.D. Fry, and B. Wang. 1998. Water relations and canopy characteristic of tall fescue cultivars during and after drought stress. HortScience 33 :837–840.[Abstract/Free Full Text]
• Huang, B., and H. Gao. 1999. Physiological responses of diverse tall fescue cultivars to drought stress. HortScience 34 :897–901.[Abstract/Free Full Text]
• Horler, D.N.H., M. Dockray, and J. Barber. 1983. The red edge of plant leaf reflectance. Int. J. Remote Sens. 4 :273–288.[CrossRef]
• Horst, G.L., J.P. Klingenberg, and T.P. Riordan. 1991. Spectroradiometry data performance among 35 buffalograss genotypes. p. 176. In 1991 Agronomy abstracts. ASA, Madison, WI.
• Howell, B.M. 1999. The effectiveness of sensor based technology to determine nitrogen deficiencies in turfgrasses. M.S. thesis, Oklahoma State Univ., Stillwater.
• Jackson, R.D. 1982. Canopy temperature and crop water stress. Adv. Irrig. 1 :43–85.
• Jiang, Y., and B. Huang. 2000. Effects of drought or heat stress alone and in combination on Kentucky bluegrass. Crop Sci. 40 :1358–1362.[Abstract/Free Full Text]
• Jones, K.J., L.T. Las, and H.D. Shew. 1992. Infrared image analysis of turf disease control experiments in North Carolina. Phytopathology 82 :1122.
• Kenna, M.P. 1995. Detecting turf stress with remote sensing. Grounds Maint. 10 :17–20.
• Kleman, J., and E. Fagerlund. 1987. Influence of different nitrogen and irrigation treatments on the spectral reflectance of barley. Remote Sens. Environ. 21 :1–14.[CrossRef]
• Nutter, F.W., Jr., M.L. Gleason, and N.E. Christians. 1991. Effect of visual, remote sensing, and image analysis assessment methods on intra-rater and inter-rater reliability estimates in the dollar spot-bentgrass pathosystem. Phytopathology 81 :1182.
• Osborne, S.L., J.S. Schepers, D.D. Francis, and M.R. Schlemmer. 2002. Detection of phosphorus and nitrogen deficiencies in corn using spectral radiance measurements. Agron. J. 94 :1215–1221.[Abstract/Free Full Text]
• Perdomo, P., J.A. Murphy, and G.A. Berkowitz. 1996. Physiological changes associated with performance of Kentucky bluegrass cultivars during summer stress. HortScience 31 :1182–1186.[Abstract/Free Full Text]
• SAS Institute. 2004. SAS/STAT 9.1 user's guide. SAS Inst., Cary, NC.
• Sims, D.A., and J.A. Gamon. 2003. Estimation of vegetation water content and photosynthetic tissue area from spectral reflectance: A comparison of indices based on liquid water and chlorophyll absorption features. Remote Sens. Environ. 84 :526–537.[CrossRef]
• Tobias, R.D. 1997. An introduction to partial least-squares regression. Available at SAS Institute http://ftp.sas.com/techsup/download/technote/ts509.pdf (verified 14 Jan. 2007).
• Trenholm, L.E., R.R. Duncan, and R.N. Carrow. 1999. Wear tolerance, shoot performance, and spectral reflectance of seashore paspalum and bermudagrass. Crop Sci. 39 :1147–1152.[Abstract/Free Full Text]
• USDA-NRCS. 1998. Estimating soil moisture by feel and appearance. Program Aid no. 1619. USDA-NRCS, Washington, DC.
• Wanjura, D.F., and J.L. Hatfield. 1987. Sensitivity of spectral vegetative indexes to crop biomass. Trans. ASAE 30 :810–816.[Web of Science]

This article has been cited by other articles:

				A. R. Johnsen, B. P. Horgan, B. S. Hulke, and V. Cline Evaluation of Remote Sensing to Measure Plant Stress in Creeping Bentgrass (Agrostis stolonifera L.) Fairways Crop Sci., October 22, 2009; 49(6): 2261 - 2274. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Search for Related Content PubMed Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Agricola Articles by Dettman-Kruse, J. K. Articles by Chaplin, M. H. Related Collections Turfgrass Management Water Stress Irrigation