TWCA Program Protocol
The protocol for acceptance into the TWCA Program has been designed by select turf grass scientists from throughout the United States using rain out shelters to control drought stresses. These scientists use screening methods that exceed industry standards, determine the standards for each species, and determine the planting and evaluation criteria using digital image analysis. Management practices using multiple data sets from multiple sites or multiple years insure the integrity of the program.
Testing Procedures – Screening Methods
2.1 Plant Material Used:
Within any species tested, it will be imperative to compare experimental or commercial lines with existing standards for that species. In addition, once superior germplasm is identified for a particular species, a selection of those drought-tolerant lines would be included in future testing as standards.
2.1.1 Standards for Each Species
Initial standards for each species will be determined from preliminary results observed in initial drought studies conducted by NexGen Research and by selecting the top-ranked cultivar for overall turfgrass quality from the most recent National Turfgrass Evaluation Program. A minimum of four standards will be included in each trial.
2.2 Test Sites
As many as 5 test sites may be used in the overall program, to cover a geographic range that can accommodate warm-season and cool-season turfgrass species. Currently testing structures exist in Fayetteville, Arkansas (University of Arkansas), Virginia Beach, Virginia (Virginia Tech), and Albany, Oregon (NexGen Turf Research).
2.3 Planting and Establishment Methods
Studies will be conducted in approved structures that can restrict natural rainfall on the plot area during the drought stress period (Figure 1). Field testing may be conducted in regions where rainfall is normally not a factor during the drydown period. Entries in any experiment will be replicated three to four times in a randomized complete block design and established either from seed, sprigs, or sod in a minimum plot size of 1 x 1 m. Planting rates will be established for each species and reflect industry standards for that species. The plot area will be irrigated using overhead sprinklers as needed during establishment to promote germination and establishment and at a rate of 2.5 cm wk-1 thereafter to provide optimal growing conditions.
Following establishment, the experimental area will be maintained at an appropriate ht for each species and will be fertilized according to industry practices for the species. Plots will be maintained for a minimum of a single growing season prior to initiating drought stress. Preventative fungicides and/or insecticides will be applied to all plots prior to initiating drought stress to minimize confounding effects of fungal infection or insect infestation. Each drought stress simulation on a group of experimental entries will be replicated either in space (multiple locations) or in time (2 years) to validate the results. Management practices such as mowing ht, fertility, and pesticide applications, shall be detailed prior to the start of all studies and strictly followed at all sites.
2.4 Drought Simulation and Evaluation Methods
Prior to initiating drought stress, the experimental area will be saturated with 5 cm of irrigation per day for 3 consecutive days to eliminate any dry areas and produce uniformly wet conditions across all plots. Immediately thereafter, irrigation will be withheld to encourage drought stress symptoms.
2.4(a)Acute Drought Stress
The response of entries to acute drought stress will be evaluated a minimum of once weekly using digital image analysis techniques (Richardson et al., 2001) to quantify the percent green turf cover for each plot as drought becomes more severe (Figure 2). When all plots fall below a 25% green turf cover, the experimental area will be saturated with 5.0 cm of irrigation to initiate drought recovery. Thereafter, the experimental area will be irrigated weekly with 2.5 cm water and recovery of entries from drought evaluated weekly using digital image analysis until plots reached 100% green cover.
2.4(b) Chronis Drought Stress
Chronic stress will be monitored 3 times per week using the digital image analysis technique. Plots will be mowed 12-24 hours before digital images are taken of the turf plots. Turf plots that contain less than a pre-designated (40-60%) green cover will be given a � inch equivalent of water. The study should be conducted for a period of 8-12 weeks during environmental conditions that would be considered stressful for the species under evaluation. The total amount of water applied would be calculated for each cultivar at the end of the study and percent differences will be determined.
2.4(c) Evapotranspiration Rates
Test sites should record E.T. rates throughout the study. These E.T. rates may be utilized to estimate percent differences between cultivars in regards to water requirements. This is applicable when utilizing acute drought stress testing. (see Figures 3 and 4)
2.4(d) Statistical Analysis
The data will be fit to a Sigmoid variable slope model, [ green turf cover (%) = 100/(1+10^((Days50-DAI)*Slope)) ] where DAI = days after irrigation (ceased or initiated, for dry-down or green-up, respectively) and Days50 and Slope are estimated model parameters. Days50 is estimated to be the DAI when green turf cover = 50%. The Slope parameter defines how rapidly turf cover changes over time with larger positive or negative values representing steeper positive or negative slopes of the Sigmoid curve. A sum of squares reduction F-test will be used to determine if entries significantly affect green turf cover during drought stress and drought recovery (Motulsky and Christopoulos, 2003). An F-test comparing the sum of squares from a global model (all varieties share Days50 and Slope values) against the cumulative sum of squares from models where Days50 and Slope values will be determined separately for each variety. If the sum of squares is reduced significantly (P < 0.05) using separate parameter values, variety effects will be considered significant. Parameter estimates will be used to calculate confidence intervals (95%) for the number of DAI (or irrigation withheld) until each entry reaches 25, 50, and 75% green turf color (Motulsky and Christopoulos, 2003). At each turf coverage percentage (25, 50, and 75), entries will be considered significantly different if their confidence intervals did not overlap (Figure 3). Nonlinear regression analysis of the turf cover data will be performed using GraphPad Prism version 4.0 for Windows, (GraphPad Software, San Diego, CA).
Figure 4. This example shows the comparison of E.T. rates as seen on the confidence interval analysis in Figure 3. A 22% difference in evapotranspiration rate is seen between Geronimo (day 37) and Mallard (day 48).