AN ON-FARM TEST OF PERENNIAL FORAGE GRASS VARIETIES UNDER MANAGEMENT INTENSIVE GRAZING

Perennial cool-season grasses have historically been bred and evaluate d strictly under hay managements with mechanical harvesting. Forage yi eld and persistence data collected under such circumstances may have l ittle value in choosing cultivars for management intensive grazing (MI G) systems. The objectives of this study were to begin developing a da tabase of cool-season grass cultivars for MIG and a protocol for expan sion of the database: Ninety-one grass varieties were planted in rando mized complete block designs in 1990 on three dairy farms in southern Wisconsin (Fayette silt loam and Dubuque loam-both fine-silty, mixed, mesic Typic Hapludalfs). Each study was rotationally grazed five or si x times in 1991 and 1992. Compressed pasture heights (bulk density) we re measured on each plot immediately before and after each grazing eve nt and converted to estimates of available forage using a linear regre ssion calibration. Apparent intake of each plot was computed as the di fference between pre- and post-grazing estimates of available forage. Across all varieties, available forage ranged from 1.2 to 1.7 tons/acr e, apparent intake ranged from 0.45 to 0.82 tons/acre, and ground cove r (fall 1992) ranged from 18 to 93%, with significant differences obse rved among species and within several species. Reed canarygrass (Phala ris arundinacea L.) had markedly greater available forage and apparent intake than the other very hardy species. Creeping foxtail (Alopecuri s arundinaceus Poir.) had very high apparent intake (0.70 tons/acre) a nd Kentucky bluegrass (Pea pratensis L.) had very low apparent intake (0.57 tons/acre) although their available forage differed by only 0.05 tons/acre. Timothy (Phleum pratense L.) varieties were highly variabl e, while smooth bromegrass (Bromus inermis Leyss.) varieties did not s how marked differences. Many orchardgrass (Dactylis glomerata L.) vari eties had extremely high available forage and apparent intake, always higher in apparent intake than tall fescue (Festuca arundinacea Schreb .) for the same level of available forage. Similarly, perennial ryegra ss (Lolium perenne L.) had higher apparent intake than meadow fescue ( F. pratensis Huds.) for the same level of available forage. Intermedia te and Italian ryegrasses (L. multiforum Lam.), festulolium (Festuloli um braunii K.A.), and 'Matua' prairie brome [Bromus unioloides (Willd. ) H.B.K] were not well adapted to the combination of MIG and harsh, re latively snow-free winters. The study provided the beginning of a data base that will be extremely useful in developing credible recommendati ons of perennial grasses for cool-season pastures.