Evaluating temperate species in the subtropics. 2. Perennial grasses

Abstract

A series of 3-year experiments, spanning the period 1993–2006, evaluated the performance of commercially available cultivars and breeders’ lines of perennial ryegrass (Lolium perenne), tall fescue (Festuca arundinacea), brome grass (Bromus spp.) and phalaris (Phalaris aquatica) and the herbs, chicory (Cichorium intybus) and plantain (Plantago lanceolata) under irrigation in a subtropical environment. Measurements included annual and total dry matter (DM) yields, seasonal yields, persistence and resistance to rust in pure, nitrogen-fertilised swards.
Summer-active tall fescue cultivars averaged around 50 (range 40–70) t/ha DM over 3 years, except at Mutdapilly where the average was 25 (range 11–34) t/ha DM. This was 30% more than perennial ryegrass cultivars (38 t/ha DM), with the prairie grasses (45 t/ha DM) intermediate between these extremes. About 45% of this total yield from all species occurred in the first year. The most productive cultivars were AU Triumph, Dovey, Quantum, Jesup and Jesup MaxP. Grasslands Matua and Atom produced the highest yields for brome grass and Tolosa, Fitzroy, Dobson, Bronsyn, Cannon and Quartet were the highest-yielding of the perennial ryegrasses, along with the hybrid ryegrass, Horizon. Tall fescue produced higher spring, summer and autumn yields than the perennial ryegrasses in this environment, with perennial ryegrasses superior in winter.
Differences in persistence were of even greater magnitude; tall fescue cultivars were around twice as persistent as the perennial ryegrasses, while the bromes were less persistent than the perennial ryegrasses under a cut-and-carry regime.
Incidence of crown rust was generally lower in tall fescue than in perennial ryegrass. However, the newer ryegrass cultivars and some experimental lines showed higher resistance than the older cultivars. Phalaris and the brome grasses were not infected by rust.
First-year yields explained 67% of the variation in 3-year total yields and 50% of the variation in persistence after 3 years. The possibility of using detailed sampling in the first year and persistence over 3 years as a method of predicting cultivar performance is discussed.