Breeding improvement of early growth in timothy (Phleum pratense L.) could help sustain high-quality vegetation in timothy-based grasslands. In this study, I investigated a useful selection method for early growth using 63 timothy genotypes, an orchard grass (Dactylis glomerata L.) cultivar (cv. Paykar), and a perennial ryegrass (Lolium perenne L.) cultivar (cv. Chinita) by field and laboratory tests in Kunneppu, Japan. This study employed the first crop dry matter yield in the established year as an early growth index. Early growth significantly differed among the genotypes and showed low to moderate narrow-sense heritability, which indicates that it is desirable to focus on line selection rather than individual selection. Multiple linear regression and canonical discriminant analyses revealed three useful indirect selection indices for early growth (plant vigor at establishment in field conditions, and thousand kernel weight and 7th day dry matter weight in Petri dishes in laboratory conditions). Predicted increases in timothy dry matter weight by multiple linear regression analyses using the above three traits also demonstrated that there is potential to achieve early growth similar to that of orchard grass, although it is difficult to secure as much early growth as perennial ryegrass.

The actual condition of winter damage (WD) for perennial ryegrass (Lolium perenne L.) was investigated in fields, in Tokachi region, located on the Pacific Ocean side of Japan, which is characterized by cold temperatures and scarce snow. Although WD has been frequent, the differences among different districts or years have been described based on the extent of differences and involved factors. Referral to the past literature concerning overwintering characteristics of temperate grasses allowed for the temporary establishment of the conditions of severe WD occurrence based on the relationship between the observed WD and meteorological elements. The relevance of the occurrence of severe WD events annually from 2001 to 2020 was explored at 105 major Automated Meteorological Data Acquisition System points in Hokkaido. Thirty points experienced severe WD in more than six of the 20 years in eastern Hokkaido, that is, the eastern part of the Pacific and Sea of Okhotsk sides. The risk of WD for these points was presumed to be high. Except for eastern Hokkaido, six points corresponded to a boundary of the Sea of Japan side to the western part of the Pacific side, and further at the western part of the Pacific side. In contrast, also in eastern Hokkaido, some points experienced few WD years. These findings suggested that it is important to evaluate the risk of WD in more detail. Furthermore, based on the relationship between the risk of WD and the climate during winter, these points could be classified broadly into three groups as follows: (I) points where normal wintering is difficult owing to scarce snow and cold temperature, (II) points where wintering is possible owing to warm temperature regardless of scarce snow, and (III) points where wintering is possible owing to steady snowfall from early winter or from the same time it gets cold, or owing to much snow, regardless of whether the temperature is warm or cold.