Project Abstract

<p>&nbsp;                   <b>ABSTRACT&nbsp;</b></p><p>The increase of ultraviolet (UV) light levels in the northern hemisphere raises the spectre of possible problems for turfgrass plants due to long term exposure. Cool season turfgrasses which are susceptible to photoinhibition may suffer from a loss of productivity and growth, reducing their ability to sequester carbon. The effect on germination, plant responses in relation to UV absorbing compounds and how turfgrasses were evaluated for loss of quality with regards to linking new technology to visual ratings have had no or limited research. Enhancing the percentage of Kentucky bluegrass (Poa pratensis L.) seed germination and speed could benefit establishment of the grass in a greater range of environmental and geographical conditions.&nbsp;</p><p>Potential for the use of ultraviolet light to enhance Kentucky bluegrass seed germination exists through exposure to UV light. The effect of ultraviolet light may be lost with seed age. In altering wavelength exposure, there may be an opportunity to enhance the effect. In measuring turf quality, traditionally human measurement has been the standard method for both color and cover. Color, in particular, is thought to be controlled by pigment changes. In evaluating a total of 51 cultivars of tall fescue (Schedenorus phoenix Scop. Holub), perennial ryegrass (Lolium perenne L.) creeping bentgrass (Agrostis stolonifera L.) cv. ‘Penncross’ and ‘L-93’ it was found that nitrogen content is iii most crucial in color measurements. Reflective measurements did not correlate with nitrogen or chlorophyll content.&nbsp;</p><p>Extract measurements had stronger correlation with nitrogen content than pigmentation concentration. Current reflective measurement equipment may not be closely linked to visual rating of turfgrass color possibly due to variation in leaf surfaces. There is a difference in response to UV light among grass species. Creeping bentgrass ‘L-93’ produced increasing anthocyanin in response to UV light. The characterization of Cyanidin – 3 – O – glucoside was the first reported in the literature in creeping bentgrass. Carotenoids, zeaxanthin and β-carotene decrease in creeping bentgrass after exposure to UV-B and turfgrass quality and vegetative production in bentgrasses decreased to a greater extent than tall fescue or perennial ryegrass. All grasses have the ability to initially accumulate phenolic compounds and flavonoids in the tissue most exposed to light, although this still doesn’t mean a prevention in damage to photosynthetic machinery. Turfgrass recovery and maintenance of optimal photosynthetic rates will be crucial as breeders try to develop new cultivars that are adapted to higher levels of UV light. <br></p>

Project Overview

<p> <b></b></p> <b>INTRODUCTION&nbsp;</b><div>Turfgrass color is traditionally ranked on a 1-9 scale with a ranking of one equal to dead brown turf and nine equal to optimal green turf color in the eyes of a rater. However, this is a subjective measurement. In trying to develop a rapid method of measuring turfgrass responses to either stresses or management inputs, researchers have found success with using handheld reflectance systems in field situations to measure plant color. The measurement of plant disease damage and plant nutrient deficiencies have been successfully tested in creeping bentgrass (Agrostis stolonifera L.), and corn (Zea mays L.), (Nutter et al., 1993; Bausch and Duke, 1996). These systems provided highly correlated data and were shown to have positively correlated (R) values to visual ratings and even tended to be less biased (Nutter et al., 1993). There has been success using systems which are wavelength dependent to indicate biomass status and chlorophyll (680 – 780 nm) content (Filella and Penuelas, 1994). Wavelengths in the red&nbsp; range 680-780nm are considered the best estimator of leaf area index in (Capsicum annuum L.) and (Phaseolous vulgaris L.). Water stress can be measured using this technology also, however this was seen only after the stress was well developed (Nutter et al., 1993).</div><div><br></div><div> The digital color measurement systems in use have different functions. Digital imaging has been found to be extremely successful (r2 =0.99) in correlating Munsell color chips to nitrogen treatments in creeping bentgrass (Karcher and Richardson, 2003). The digital system used is also more precise when compared to traditional methods (Richardson et al., 2001). The digital analysis uses photography and software to develop ranking systems and has the ability to obtain large volumes of data from one picture, and separate out colors with excellent reproducibility. Colorimeters, which depend on a reflective light wavelength and intensity measurement have been used to separate creeping bentgrass cultivars and different nitrogen treatments (Landschoot and Mancino, 2000; Mangiafico and Guillard, 2005). The method used in our trials is offered as an alternative method for turfgrass color analysis which can use either handheld reflective measurements or spectral transmission benchtop measurements and link them to nitrogen measurement in turfgrass.&nbsp;</div><div><br></div><div> A majority of the color-based systems in use are dependent on a reflective measurement of turfgrass leaf surfaces. However, leaf surfaces vary in many ways due to environmental conditions and varietal differences. Tall fescue (Schedenorus phoenix (Scop.) Holub.) cultivars have a leaf thickness of between 0.31 and 0.36mm under drought stress and a drought free leaf width of 0.44 cm to 0.52 cm (Fu and Huang, 2004). &nbsp;&nbsp;</div><div><br></div><div> Nitrogen content causes variation in tall fescue leaf width from 0.31 cm to 0.75 cm (Rademacher and Nelson, 2001). Colorimeter systems do not have the capability to measure such narrow sizes creating the potential for inaccurate measurements due to different leaf age and different nitrogen concentrations. Upper and lower leaf surfaces also vary in thickness. The upper epidermal surface of Kentucky bluegrass can be 0.58 µm and 1.23 µm on the lower epidermal surface. This thickness variation in the leaf surface plays a role in light absorption, reflection and transmission (Cameron, 1970; Clark and Lister, 1975). Penetration and reflection of light varies by leaf thickness. Treatment by UV-B of alfalfa (Brassica campestris L.) resulted in an increase of leaf thickness by 45% - causing an increase in the amount of light back scattered (Bornman and Vogelmann, 1991). Variations across and within turfgrass species leaf thickness could influence reflective type measurements.&nbsp; Human eyesight uses rods and cones in the retina. The cones are sensitive to short, medium and long wavelengths and work to absorb opposing colors of red, green and blue (Brown and Wald, 1964; Roorda and Williams, 1999). The methods of color measurement by different handheld reflectance systems vary. The systems can use color rankings from the Munsell system (Escadafal et al., 1989), Hunter color systems or CIE Lab scales (Commission Internationale de l'éclairage), (Pattee et al., 1991). The use of the Munsell color system (Munsell, 1915) was designed to rank color by Albert Munsell (Nickerson, 1940). This is considered the initial color system, but it did not transfer consistently into digital reproduction and so more recent systems were developed such as CIELab and Hunter scales (Sharma and Trussell, 1997). These two systems attempt to replicate how the rods and cones work in the retina and are based on a fixed scale so as not to become device dependent and lacking reproducibility (Lemaire et al., 2005).&nbsp;</div><div><br></div><div> A handheld colorimeter system uses measurements in different color spaces and L* is considered a measure of lightness while a* measures green (-) or red, (+) and b* is a measure of blue (-) and yellow (+) colors on the wheel. Hue or the color attribute is also measureable in a degree system to fit the color wheel while chromaticity is used to measure the lightness (100 = white) or darkness (0 = black) of the color. This represents a similar system to human eyesight and can be converted into a 3-D image for digital analysis (American Society for Testing and Materials, 1989). The system emits a diffuse uniform light source over a set measuring area that is reflected and collected by an optical fiber cable which enables analysis of color (Minolta, Chroma Meter CR-300 Instruction manual). The light source can be manipulated to equal daylight conditions D65 or C which have similar light circumstances but D65 is preferred by CIE. There are other illuminant types which can replicate indoor lighting. The handheld systems are fast, simple and field usable (Shimada et al., 2004). Previous work investigated correlating colorimeter values to creeping bentgrass cultivars using the CIELab system and the Minolta CR-310 handheld colorimeter. The results indicated that the hue angle – the component of the system that describes the true color which is then further influenced by the lightness or chroma of the color – was the most consistent part of the data and correlated favorably with data collected from visual rankings (Landschoot and Mancino, 2000). The data, however, were not linked to any plant characteristic thus offering no ability to replicate the study. CIELab system data can&nbsp;&nbsp;be transferred to tristimulus based values such as XYZ values and so it offers a possible system that is transferrable between multiple pieces of equipment. This then allows for comparison of the same characteristics across grass types (Reinhard et al., 2001). Trials on creeping bentgrass using the normalized difference vegetation indices (NDVI) system compared to turfgrass pigmentation content indicated an influence on color ratings when nitrogen, magnesium and iron fertilization rates were varied. The variation in the color was only explained 36% of the time by chlorophyll pigmentation and that another factor or combination of factors was responsible for the result (Stiegler et al., 2005).&nbsp; Extract measurements of chlorophyllous pigments using benchtop colorimeters can also give color measurements based on either transmittance or reflectance dimensions (Ahmed et al., 2000). Transmitted or reflected light beams are collected through a lens which is 8° from perpendicular to the sample surface. The beams are then integrated using a sphere and flash beams. The flash reduces the noise and allows for higher intensity light which normally reduces sensitivity of the pigments due to degradation when exposed to light for an extended period (Joilot and Joilot, 1984). Spectrophotometers have been used to test color differences between grasses extracts (Kavanagh et al., 1985). The transmitted light is then passed through a diffraction grating which separates and measures the wavelengths of light. These systems also possess the ability to measure in L*a*b* color space (Yudin et al., 1987). The use of extracts requires lab time however; this can be a downside to the use of benchtop color measurement due to cost and time while field measurements are easily and quickly carried out. </div><div><br></div><div> These systems offer opportunities to standardize color system management and take out subjectivity that can occur between trained and untrained raters of color (Gooding and Gamble, 1990; Landschoot and Mancino, 1997). There, however, needs to be a larger bank of color data compiled with these systems. There are limited data in relation to pigmentation influence on color rating and whether or not surface reflectance measurements or transmission type measurements are more appropriate for color analysis. The objective of this study was to discover how chlorophyll pigments influence color systems analysis and link pigment concentration to a particular trait within the color system to create a baseline of data for breeders and other end users. The study aims to determine what influences human color perception which could be related to stress responses during periods of pigment decline such as exposure to UV-B. It further aims to discover whether there is a correlation between visual rating and color system ratings.</div><div>&nbsp; <br></div><div><b>MATERIALS AND METHODS&nbsp;</b></div><div><b><i>Plant material&nbsp;</i></b></div><div>Two separate Turfgrass trials were run from March, 1st 2010 until June, 11th 2010, and June 15th 2010 until September 30th 2010 at The Ohio State University, Howlett Hall greenhouse, 680 Vernon Tharp St., Columbus OH, 43210. Six turfgrass varieties were seeded in a sand based rootzone, of 80/20 volume/volume sand / peat mix. There were 9 cultivars of perennial ryegrass, Kentucky bluegrass, creeping bentgrass, tall fescue, fine fescues, sheeps fescue (Festuca ovina L.), creeping red fescue (Festuca rubra L.), hard fescue (Fesutca brevapila Tracey.). Six cultivars of colonial bentgrass were evaluated. <br></div><div> The grasses were chosen based on their color rankings from national turfgrass evaluation program (NTEP) trial results for 2003 – 2007 for all grasses and based on commercial availability. Cultivars were randomly separated by the rankings of high / intermediate / low (Top 1/3, Middle 1/3 Lowest 1/3) of grasses on the NTEP evaluation results. Turfgrasses were watered with a mist system initially for 30 seconds every 10 minutes during germination (21 days). Once established, grasses were watered twice daily for 3 minutes. All turfgrass samples were treated with 20-20-20 liquid fertilizer (Scotts Co. Marysville, OH) at a rate of 4.06 kg ha-1 on a bi weekly basis. Each grass was then grown out to a height of 10 cm and maintained at this height for two months in a greenhouse with temperatures maintained at ~20° Celsius. <br></div><div><br></div><div> <b><i>Data collection&nbsp;</i></b></div><div>Turfgrass samples were placed in a completely randomized design in full sun conditions. A total of 44 untrained (not trained in evaluating turfgrass color) and trained (turfgrass researchers with training and experience in ranking turfgrass color) evaluators ranked grasses based on their color. Ranking was set with 1 = brown 6 = satisfactory and 9 = optimal dark green. Turfgrasses were irrigated twice on both days to prevent wilt and plant material was unranked for 1 hour periods post irrigation to prevent any impact of water remaining on leaf surfaces on rankings. On both occasions grasses were in full sun for the total period of time during ranking. A total of forty two surveys were collected between both experiments. Visual ratings by humans were carried out with permission of institutional review board (IRB project #2010E0291) on June 12th and again on September 22nd 2010. <br></div><div><br></div><div> Turfgrass samples were taken immediately following the visual rating and leaves were collected by area (40 mm-2 ) and on a mass basis (0.05 g) with three subsamples taken from each cultivar and placed in N,N-dimethylformamide (Sigma-Aldrich, St Louis, MO) to extract chlorophyll, (Moran and Porath, 1980). Samples were stored in dark conditions for a period of 48–72 hours after this. Extract was subjected to absorption analysis using a UV-Vis spectrophotometer (Shimadzu Scientific Instruments Inc., Columbia, MD) at wavelengths of 664, 647, 625 and 603 nm. Data were calculated using formulas from Moran, (1982) and chlorophyll totals, chlorophylls A and B were quantified on an area and mass basis. The premise for this was that the hand-held system measures directly on a set area on leaf tissue while benchtop systems require tissue extract for analysis. Correlating color with chlorophyll content based on the area and mass measurements were designed to fit this variation between color measurement systems. Freshly harvested samples were used for reflectance analysis. Turfgrass leaves were layered four times to create an optically dense stack (Mangiafico and Guillard, 2005), then three light measurements were taken on the top layer (Appendix A), using a Konica Minolta CRM 310 (Konica Minolta Sensing, Ramsey, NJ). The stacking and multiple readings were done to reduce variability due to turfgrass surface irregularity. The layers were then randomly re-ordered and measurements repeated. Samples were reconfigured three times for a total of nine color measurements per pot and thirty six measurements per cultivar. Data were taken for chroma, hue lightness (L*) red/green (±a*) yellow/blue (±b*). Prior to analysis the system was calibrated with a white plate provided by the manufacturer. <br></div><div><br></div><div> Tissue samples from each grass were taken and dried for 48 hrs at 60 °C prior to tissue analysis for percentage nitrogen content at this time (A&amp;L Eastern Laboratories, Richmond, VA).&nbsp;</div><div><br></div><div><b><i>&nbsp;Data analysis and experimental design&nbsp;</i></b></div><div>Four replications of each cultivar were used in two repeated trials. The pots were ranked by humans in a completely randomized design in an outdoor setting with full sun on turfgrasses throughout the day. During the second experiment a precipitation event forced the stoppage of ranking, this resumed once water was removed from leaf surface and cloud cover was gone. Data transformation of a*b* values is required to create the hue value which is in a degree format to fit onto the 3-D color circle associated with the system. The conversion formula is: degrees ((Atan2(b*,a*)) Following this conversion data was tested using robust regression ROBUSTREG analysis for data influence and outliers (Chen, 2002). Normality tests were carried out using the Kolmogorov – Smirnov test while skew and kurtosis were tested also. Data was then analyzed using multivariate principal components analysis with the R program (R Development core team, 2011; Jaisingh et al., 2006) to gain an in initial understanding on variable influence on color. The results of the multivariate analysis were then used to develop regression and correlation data using PROCREG and PROCCORR with SAS (SAS Institute, Cary, NC). Differences between cultivar color, nitrogen content and chlorophyll content were analyzed using PROCGLM. All grasses were analyzed separately and as a completely randomized design. <br></div><div> Color analysis of extract was carried out using the Hunter Colorquest XE system (Hunter Associates Laboratory Inc., Reston, VA). Turfgrasses were initially tested for spectral transmission properties to assess which volume of turfgrass was suitable to reduce high viscosity or high reflectance levels which could distort data. Turfgrass samples weighing 0.5 g were determined to be most consistent across all grasses used.&nbsp;</div><div>Turfgrass samples were then placed in 15 mL glass tubes with 10mL of N,Ndimethylformamide. N,N-dimethylformamaide was used to extract all color related to chlorophyll pigments due to its high efficiency (Stiegler et al., 2004). Samples were placed in reagent for 48hrs in darkness. Extract was then placed in a 20 mL cuvette with 10 mm path length. Each sample was then tested at illuminant settings of D65 and transmittance settings with data acquired for chroma, hue lightness (L*) green/red (±a*) yellow/blue (±b*). Prior to analysis, calibration was performed with a white plate provided by manufacturer and dimethylformamide was used as a blank. <br></div>