Blending Seeded Bermudagrasses for the Transition Zone

J. M. Goatley, Jr., Turfgrass Specialist and Associate Professor, Crop and Soil Environmental Sciences Department, Virginia Tech, Blacksburg, VA 24061-0403; B. J. Horvath, Extension Turfgrass Pathologist and Assistant Professor, Department of Plant Pathology, Physiology, and Weed Science, Virginia Tech, Virginia Beach, VA 23455-3315; W. E. Thomason, Assistant Professor, K. L. Hensler, Research Faculty, and W. L. Askew, Turfgrass Program Manager, Crop and Soil Environmental Sciences Department, Virginia Tech, Blacksburg 24061-0403

Corresponding author: J. Michael Goatley, Jr. Goatley@vt.edu

Abstract

Seeded bermudagrass (Cynodon dactylon var. dactylon L.) use for turf continues to expand in the transition zone of the United States due to enhancements in cold tolerance and quality. However, superior cultivars such as ‘Riviera’ are often expensive or unavailable due to limited supplies. Could blending Riviera with lower quality, less expensive, and perhaps less adapted seeded cultivars such as ‘Arizona Common’ or ‘Wrangler’ lead to a predominant Riviera turf over time? Separate trials were initiated in 2004 and 2005 at Virginia Tech evaluating the establishment rate, spring green-up, and turf quality of blends of Riviera with Arizona Common or Wrangler at 0, 25, 50, or 75% by weight of pure live seed. As evidenced by enhanced spring green-up and higher visual quality ratings, Riviera dominance within the blends was observed as soon as the spring of the second growing season in both trials. Based on significant differences in mature seedhead height between standard plots of Riviera, Wrangler, and Arizona Common, mean seedhead height data for blends confirmed population shifts toward Riviera in October 2007. Arizona Common in particular offers the potential as a cost-saving seed diluent when blended with Riviera given Riviera’s inherent advantages in cold hardiness and tolerance to regular close clipping.

Introduction

A new generation of seeded bermudagrasses [C. dactylon var. dactylon (L.) Pers.] featuring enhanced cold tolerance and desirable turf quality characteristics became commercially available early in the 21st century (1,8,9,10,13). Their availability has resulted in the expansion of bermudagrass use for golf turf, sports turf, and home lawns throughout the transition zone of the United States (14,15,17). In particular, the field performance of ‘Riviera’ bermudagrass relative to other seeded and vegetative bermudagrasses has resulted in extensive use of this seeded variety throughout the transition zone (9,10,13).

The use of seed blends (defined as two or more cultivars of the same species) or mixtures (defined as two or more species) of cool-season turfgrasses is a standard recommendation of turfgrass textbooks as the genetic diversity expands performance potential of the establishment under different environmental and pest stresses (2); however this is not typically recommended in warm-season turf swards. Ultimately, the factor that makes blending acceptable in fine turf management is the achievement and maintenance of an acceptable level of visual uniformity within the polystand (18). While there are numerous seeded bermudagrass blends commercially available, there is little, if any, research on which to make research-based recommendations for using them, and in general, blends and/or mixtures of warm-season turfgrasses are discouraged due to non-uniformity issues (18). However, the diverse climatic conditions of the transition zone might work to the advantage of a turfgrass manager that blends an improved variety such as Riviera with lower quality, perhaps less adapted, and likely less expensive varieties such as Arizona Common or Wrangler. This research was conducted to evaluate turf establishment rates, spring green-up rates, visual turf quality, and population shifts over a two to three-year period in plots seeded with varying percentages by weight of Riviera combined with Arizona Common or Wrangler.

Seed Sources and Experimental Techniques

The cultivars Riviera (R), Wrangler (W), and Riata were obtained from Johnston Seed Co. (Enid, OK). Arizona Common was provided by Seeds West Inc. (Maricopa, AZ) in both 2004 and 2005. While both Riviera and Arizona Common have been used extensively in turfgrass systems, Wrangler was developed primarily for grazing or forage systems and is recommended in Virginia as a cold tolerant, perennial grazing or haying component within a season-long summer grazing program (5). Riata, a blend of approximately 60% Wrangler plus 40% Riviera by weight is also marketed by Johnston Seed Co. for grazing systems. The commercial availability of this blend and the documented persistence of Wrangler in Virginia’s climate were the basis for evaluating this blend.

Riviera was blended with Wrangler (R/W) or Arizona Common bermudagrass (R/C) at levels of 0, 25, 50, 75 or 100% by weight of pure live seed. These blends (as well as Riata) were seeded into a prepared seedbed that had previously been bermudagrass free at 24.5 or 49 kg pure live seed/ha. Trials were established on 8 July 2004 and 12 July 2005 at the Turfgrass Research Center of Virginia Tech, Blacksburg, VA (37°12'N, -80°24'W). Plots were established on a Groseclose silt loam (fine, kaolinitic, mesic typic Hapludult) with the 2004 site having a pH of 6.0 and P and K levels of 40 and 76 mg/kg and the 2005 site having a pH of 6.8 and P and K levels of 56 and 102 mg/kg. A granular 10-10-10 fertilizer was applied to the prepared seedbed at 49 kg/ha and lightly incorporated into the top 5 cm of the soil prior to planting. Seed for each treatment were mixed with 470 cc of dry masonry sand and applied to 0.9-m by 3-m plots using a gravity-fed drop spreader. The seed were lightly raked into the surface with a leaf rake and in order to minimize the potential for seed movement between plots, the planting bed was covered with seed germination cloth (Forestry Suppliers Inc., Jackson, MS) until germination was complete. Irrigation was applied as necessary for the first four weeks to maintain a moist surface conducive to germination, and after establishment, the irrigation was supplied only to maintain active growth during the growing season.

The plots were mowed with a reel mower (clippings returned) at 1.9-cm beginning approximately 4 weeks after establishment and mowing continued for at least two times per week throughout the growing season (mid May through early October) until the first week of August 2007 at which time the grasses were allowed to flower for approximately 8 weeks. Quinclorac (3,7-dichloro-8-quinolinecarboxylic acid) was applied at 49 kg active ingredient/ha on 22 July 2004 in the 2004 trial and on 26 July 2005 in trial 2 for postemergent crabgrass (Digitaria sp.) control during grow-in. Plot integrity was maintained by applying a 3% glyphosate [N-(phosphonomethyl) glycine] solution to all plot boundaries at least twice per growing season with a wick applicator (Southern States Cooperative Inc., Richmond, VA). Oxadiazon [2-tert-butyl-4-(2,4. dichloro- 5-isopropoxyphenyl)-Δ2-1,3,4-oxadiazolin-one] was applied at 122.5 kg active ingredient/ha for preemergent crabgrass control the first spring following each trial’s establishment (2 May 2005 for trial 1 and 26 April 2006 for trial 2). No other herbicides were applied for trial duration. Nitrogen was applied at 49 kg/ha as 45-0-0 during the first weeks of June, July, and August for each trial in each year.

Visual percentage ground cover ratings (0 to 100) were periodically recorded during the establishment year. Visual turf quality ratings considering density, color, and uniformity were made using a 1 to 9 scale (where 1 = unacceptable, 5 = acceptable, 9 = superior) across all active growing months for the duration of the trials. Visual percentage green-up (0 to 100) for each plot was estimated during each spring transition period.

Mowing ceased in August 2007 on all plots in both trials in order to provide approximately 8 weeks of potential reproductive growth. Based on visual observations, we hypothesized that differences in mature seedhead (spikes completely developed and expanded) heights observed in the 100% R, C, and W plots might provide a quantifiable means to evaluate population shifts within the blended plots. On 18 and 19 October, the heights (cm) of mature seedheads in the 100% R, C, and W plots were measured for samples of 2, 4, 8, 16, 32, 64, and 128 mature seedheads randomly selected while traversing the plots in a diagonal fashion from top to bottom and back until the desired number of mature seedhead heights were collected. An ANOVA and subsequent plot of these seedhead height measurements indicated that the sample variance in mature seedhead heights within the 100% R, C, and W plots stabilized between 32 and 64 random samples. Based on these findings, the heights of 64 randomly selected mature seedheads were measured in all plots in both trials on 22 and 23 of October 2007 for data analyses.

Each seed blend treatment was replicated 3 times in a randomized complete block design in each trial. Treatment means were calculated and analyses of variances (ANOVA) were performed on all data using SAS (SAS Institute Inc., Cary, NC). Based on the stepwise treatment design in varying seed percentages within R/W and R/C blends, single degree of freedom contrasts for determination of linear, quadratic, or cubic gradient trends in percent ground cover, percent spring green-up, and visual turf quality were performed using the General Linear Models procedure of SAS. Riata was excluded from the gradient trend analyses and means are presented only as a reference. Since there were significant differences in percent ground cover during establishment between seeding levels in both trials, these data were analyzed separately. For all other data, there were no differences in response to seeding rate and the data were combined. There was a significant measurement date by treatment interaction, so all data are presented for individual measurement dates. Mean separations of seedhead height measurements made in October 2007 were performed for all treatments (including Riata) by calculating an LSD value (P ≤ 0.05) using SAS.

Initial Establishment

As anticipated, initial plot coverage ratings were significantly higher for the 49 kg/ha seeding rate for either R/W or R/C blends in both trials (Table 1). There were no significant trends in coverage ratings at approximately 2 weeks after planting (WAP) for R/W blends for either seed rate or trial, but by approximately 6 WAP there were significant linear increases in plot coverage as percent of R increased in the seed blend. For R/C blends, higher percentages of Arizona Common in the blends tended to increase initial coverage at 2 WAP for either seed rate in both trials, although the differences were not significant in trial 1. Due to the slow germination and initial establishment of R alone the increase in initial coverage is a definite positive attribute of the blend rather than R seeded alone especially in light of the fact that R eventually dominates the sward. In addition to a significant linear response at 2 WAP in trial 2, a significant quadratic response was also observed as mean percent coverage declined as percent Riviera in the blend was ≥ 75%. There were no significant trends in R/C blends at 6 WAP in trial 1, but by 6 WAP in trial 2, coverage increased linearly as percent Riviera increased at either seeding level. The establishment vigor of R and C from seed are reported as very low (9,10) but both demonstrate excellent divot recovery characteristics as mature turfs (6). It has been previously observed by researchers in Mississippi that C had significantly higher germination rates than R(12) (Fig. 1). Even slight improvements in the seedling vigor of C are likely beneficial in enhancing the rate of coverage of R/C blends.

Fig. 1. Establishment comparison of a 25 kg/ha seeding rate of Arizona Common (left) and Riviera (right) at 2 weeks after planting on 8 July 2004 in trial 1.

Spring Green-up

Variety trials have consistently indicated Riviera to be one of the earliest greening bermudagrasses during spring transition (9,10,11,13), and this earlier green-up response has been utilized previously as an indicator of Riviera establishment when seeded into an existing stand of Arizona Common (12). Wrangler’s spring green-up characteristics were unknown, but it was anticipated that it too would green early given its development as a forage grass for northern transition zone locales (16). Spring green-up trends were variable for R/W blends at the earliest rating dates in each trial although where significant (18 April 2006 and 3 April 2007 in trial 1, and 3 April 2007 in trial 2), increases in Riviera content decreased spring green-up relative to the 100% Wrangler plots (Table 2). As transition progressed, there were significant positive linear or quadratic trends as percent Riviera increased in the blends in trial 1 at 10 May and 10 June 2005, and 8 May 2006. There were no additional significant trends in Trial 2 based on R/W blend composition. As expected for the R/C blends, the addition of Riviera was clearly expressed in significant linear and/or quadratic increases in spring green-up at all rating dates in either trial (Table 2, Fig. 2). That quadratic trends were recorded the first spring after establishment in both trials suggests that blends containing 25 to 75% Riviera were already expressing Riviera’s inherent advantages in cold tolerance, spring green-up, and tolerance to close clipping.

Fig. 2. Spring green-up of various percentage blends of Arizona Common (C), Riviera (R), and Wrangler (W) on 6 May 2005 in trial 1 established on 8 July 2004.

Visual Quality

Wrangler and Arizona Common have significantly coarser leaf texture and are less dense than Riviera resulting in potential concerns in turf uniformity. Blending Riviera with either Wrangler or Arizona Common resulted in almost immediate improvements in visual turf quality ratings as compared to 100% standard treatments of Wrangler or Arizona Common (Tables 3 to 6, Fig. 3). The rapid occurrence of significant quadratic trends as percent Riviera increased (no later than early summer of the second growing season in either trial) indicates that the blends had quality ratings comparable to the 100% Riviera standard. The most significant improvements in quality were realized for the first rating dates of each season when the early green-up and improved density of the Riviera were most pronounced. The 100% Wrangler standard achieved acceptable quality ratings (ratings ≥ 5) at 50 and 60% of the rating dates in trials 1 and 2 (Tables 3 and 4), respectively. Similarly, 100% Arizona Common plots had acceptable quality ratings at 50% of the rating dates for both trials 1 and 2 (Tables 5 and 6). However, the blending of as little as 25% Riviera by weight resulted in acceptable quality ratings at ≥ 90% of the recording dates for R/W blends and ≥ 75% of the recording dates for R/C blends. The commercially available Riata (60% Wrangler/40% Riviera by weight) provided acceptable ratings at ≥ 80% of the rating dates.

Fig. 3. Visual turf quality of various percentage blends of Arizona Common (C), Riviera (R), and Wrangler (W) on 26 August 2006 in trial 1 established on 8 July 2004.

Using Seedhead Heights to Quantify Population Shifts

After mowing ceased in August 2007, all plots were allowed to develop fully expanded seedheads for approximately two months. What was suspected to be obvious visible differences between mean seedhead heights of the standard 100% Riviera, Wrangler, and Arizona Common treatments (Fig. 4) was confirmed by an ANOVA of mean population seedhead heights of the 100% R, C, and W plots and mean separation of mature seedhead heights between all treatments (Table 7). All blends of R/W (including Riata) had statistically similar seedhead heights to 100% Riviera plots, but had significantly lower seedhead heights than 100% Wrangler plots. This suggests that Riviera was dominating the blend and corroborates the subjective visual quality ratings indicating the same. For R/C blends, all treatments had significantly lower seedhead height measurements than 100% Arizona Common, but the blends did have significantly higher heights than the Riviera-alone standard. All R/C blends had statistically similar seedhead heights. These data also indicate that Riviera was the predominant grass within any R/C blend at 39 months from trial 1 initiation and 27 months for trial 2. Seedhead height data analyses provided a reasonably quick and accurate technique to quantify population shifts towards R within blending treatments after a 2 to 3-year period under these experimental conditions. This technique should not be considered as an alternative to standard morphological and molecular techniques used to identify and/or distinguish bermudagrass cultivars, but was used here because of clear differences in our preliminary sample data. We are confident based on these data that seedhead height could be used in our study as a relative measure of the amount of R in each plot.

Fig. 4. Varying height of seedhead development for Arizona Common, Riviera, Wrangler, and Riata (top row, left to right) on 21 August 2007.

Conclusions

Visually acceptable turfgrass quality was achieved from seeded bermudagrass blends containing as little as 25% of the superior Riviera cultivar, with significant improvements in quality evidenced as early as the second year of the trial. The differences in cold tolerance between Riviera and Arizona Common (1) and the superior tolerance of Riviera as compared to Wrangler (and Arizona Common) to regular 1.9-cm mowing give credence to seeding blends of these cultivars in transition zone climates (13). Differential tolerances in regular mowing heights initiated after seeding had previously been shown to alter populations of Kentucky bluegrass (Poa pratensis L.) and perennial ryegrass (Lolium perenne L.) (3). Similar to our results, Kentucky bluegrass blends containing superior and low quality cultivars were dominated by the superior cultivars over a two to three year period, with the low quality variety serving as a seed diluent that reduced initial turf quality and never completely disappeared from the stand (4). In our research, the initial reduction in turf quality from blending was limited, especially for blends containing ≥ 50% Riviera. Given the superior performance potential of the better adapted and higher quality Riviera component of the blend in the transition zone, the inclusion of inferior seeded varieties as seed diluents in blends can significantly reduce establishment costs. A survey of regional seed vendors in the winter of 2008 provided the following average prices on a cost per unit weight of pure live seed: Riviera at $62/kg, Wrangler at $37/kg, Arizona Common at $9/kg, and Riata at $46/kg. Blends containing at least 50% Arizona Common can provide significant savings in seed costs and the desired Riviera ultimately dominates the stand. Given higher seed costs, blends of R/W or the commercially available Riata are likely not as attractive to turfgrass managers as R/C blends.

Further research is required to evaluate the performance and potential cost savings when blending cheaper, less cold tolerant bermudagrass varieties with those well adapted to the transition zone climate. Would inferior quality but faster establishing seeded varieties affect the potential to achieve a stand dominated by a superior variety? The highly variable transition zone climate affords opportunities with seeded bermudagrass blends that negate recommendations discouraging bermudagrass blends.

Acknowledgments

The researchers recognize the support of the Virginia Agricultural Experiment Station and thank Johnston Seed Company and Seeds West, Inc. for supplying all seed. Thanks to Mr. David McKissack for his assistance in data collection and plot maintenance.

Literature Cited

1. Anderson, J. A., Taliaferro, C. M., and Wu, Y. Q. 2007. Freeze tolerance of seed- and vegetatively-propagated bermudagrasses compared with standard cultivars. Online. Applied Turfgrass Science doi:10.1094/ATS-2007-0508-01-RS.

2. Beard, J. B. 1973. Turfgrass: Science and Culture. Prentice-Hall, Englewood Cliffs, NJ.

3. Brede, A. D., and Duich, J. M. 1984. Initial mowing of Kentucky bluegrass-perennial ryegrass seedling turf mixtures. Agron. J. 76:711-714.

4. Brede, A. D. 2004. Blending Kentucky bluegrass cultivars of different quality performance levels. Crop Sci. 44:561-566.

5. Fike, J. H., Teutsch, C. D., and Ward, D. L. 2005. Warm-season grass production responses to site and defoliation frequency. Online. Forage and Grazinglands doi:10.1094/FG-2005-0824-01-RS.

6. Karcher, D. E., Richardson, M. D., Landreth, J. W., and McCalla, J. H. 2005. Recovery of bermudagrass varieties from divot injury. Online. Applied Turfgrass Science doi:10.1094/ATS-2005-0117-01-RS.

7. Layton, J., Munshaw, G., Stewart, B., Baldwin, B., and Philley, H. 2007. Does seed coating affect bermudagrass germination? CD-ROM. Ann. Meet. Abstr. 168-4. ASA, CSSA, and SSSA, Madison, WI.

8. Martin, D. L., Wu, Y., Anderson, J. A., Anderson, M. P., Bell, G. E., and Walker, N. R. 2007. Bermudagrass cultivars with high quality and improved cold hardiness. Online. USGA Turfgr. Env. Res. 6(17):1-8.

9. Morris, K. N. 2002. National Turfgrass Evaluation Program. 1997. National Bermudagrass Test. Final Report NTEP No. 02-07. USDA, Beltsville, MD.

10. Morris, K. N. 2007. National turfgrass evaluation program, 2002 National bermudagrass test. Final Report NTEP No. 07-10. USDA, Beltsville, MD.

11. Munshaw, G. C., Ervin, E. H., Shang, C., Askew, S. D., Zhang, X., and Lemus, R. W. 2006. Influence of late-season iron, nitrogen, and seaweed extract on fall color retention. Crop Sci. 46:273-283.

12. Munshaw, G., Philley, H., Wells, W., and Stewart, B. 2007. No-till strategies for conversion of common to high-density seeded bermudagrass. CD-ROM. Ann. Meet. Abstr. 338-13. ASA, CSSA, and SSSA, Madison, WI.

13. Patton, A. J., Richardson, M. D., Karcher, D. E., Boyd, J. W., Reicher, Z. J., Fry, J. D., McElroy, J. S., and Munshaw, G. C. 2008. A guide to establishing seeded bermudagrass in the transition zone. Online. Applied Turfgrass Science doi:10.1094/ATS-2008-0122-01-MD.

14. Richardson, M. D., Karcher, D. E., Berger, P., and Boyd, J. W. 2004. Utilizing improved seeded bermudagrass on transition-zone sports fields. Acta Hort. 661:369-374.

15. Richardson, M. D., Karcher, D. E., Boyd, J. W., and McCalla, J. H. 2003. Managing the new seeded bermudagrasses. Golf Course Mgmt. 71:81-84.

16. Redfearn, D., Woods, B., and Taliaferro, C. 2007. Choosing, establishing and managing bermudagrass varieties in Oklahoma. Oklahoma Coop. Ext. Serv. PSS-2583. Oklahoma State Univ., Stillwater, OK.

17. Rogers, C. A. 2003. You’ve come a long way, Bermuda. Golf Course Mgmt. 71:91-95.

18. Turgeon, A. J. 1999. Turfgrass Management. 5th Edn., Prentice Hall, Upper Saddle River, NJ.