|
|
|
| Author: | A.D. Brede |
| Keywords: | grass seed, mixtures, competition, Poa pratensis, smooth-stalked meadowgrass, population dynamics |
Abstract:
Kentucky bluegrass (Poa pratensis L.) cultivars are generally sown as blends of two or more cultivars.
This study was initiated to examine multi-way bluegrass blends on turfgrass quality (TQ) throughout the year at two US locations.
Five improved cultivars ('Award,' 'Liberator,' 'Chicago II,' 'Rambo,' 'NuBlue') with a range of documented TQ performance levels were used to study all possible blend combinations (monostands, 2-, 3-, 4-, and 5-way blends in equal proportions by seed weight). TQ ratings were taken monthly at duplicate trials in Maryland and Idaho over 2 and 3 years, respectively.
Stepwise regression showed higher R2 values in Idaho, probably resulting from consistent irrigation; the Maryland site was unirrigated.
Regression revealed a complicated cultivar effect, with all five cultivars participating in performance.
There were 10 instances of cultivar × cultivar interactions on TQ, a positive slope of which suggests that two cultivars acted synergistically to enhance quality.
A categorical ANOVA was used to explore criteria for ejecting a cultivar from a potential blend, and for testing the suitability of a monostand of a high performing cultivar in place of blending.
In 76 out of 221 possibilities, 'NuBlue' (a low performing cultivar) decreased TQ of a blend by its presence and increased it in 3 examples, all related to spring greenup.
A given blend candidate should be expected to contribute at least an equal number of positive and negative attributes to a blend, and thus 'NuBlue' would be rejected.
Out of 403 combinations, the 'Award' monostand outperformed 98 blend combinations and underperformed 9, suggesting that planting a top performing bluegrass monostand might be a valid agronomic practice.
|
Download Adobe Acrobat Reader (free software to read PDF files) |
|