Sowing rate and cultivar affect total biomass and grain yield of spring triticale (× Triticosecale Wittmack) grown in a Mediterranean-type environment

Abstract

Sowing rate and cultivar are basic management choices affecting biomass production, grain yield and forage quality of triticale, but their effect on this species has seldom been analysed through explanatory models such as grain yield=radiation intercepted (RI)×radiation use efficiency (RUE)×harvest index (HI). A 3-year field experiment was carried out in Sardinia (Italy) to explain the effect of five sowing rates (50, 100, 300, 500 and 700 germinable seeds per unit area) and of two cultivars of contrasting height on above-ground dry matter (AGDM) and grain production under both rainfed and irrigated conditions. The different sowing rates applied were effective in determining different plant numbers per unit area, although in 1994 the plant population density was lower than expected. Increasing sowing rates caused faster development, with a 1-week difference between extreme sowing rates in time to achieve double ridge and terminal spikelet. AGDM at anthesis was small or not influenced by sowing rate, due to the small differences in RI deriving from the greater LAI of the highest sowing rates (an average of 2.4 for 300, 500 and 700 versus 1.8 of 50 and 100). AGDM accumulated after anthesis increased as a consequence of increasing the sowing rate from 50 to 300, causing parallel differences in AGDM at maturity (from 1550 to 2177 g m −2 on average) and in grain yield (from 524 to 781 g m −2 on average). The greater AGDM accumulated after anthesis by 300 and 700 was due to a greater RI in 1996, but to a greater RUE in 1995, when a notable difference in sink size among sowing rates was detected. Sink size was also responsible for the higher HI observed at the highest population densities in two seasons out of three. The superiority of cultivar Rigel in AGDM production derived from its higher RUE (1.26 versus 1.04 g MJ −1 of Antares), associated with a lower specific leaf area and with a better light distribution within the canopy. The latter was a consequence of a lower radiation extinction coefficient before ear emergence (0.60 versus 0.70), and of less radiation intercepted by Rigel ears afterwards. A higher HI allowed the greater AGDM produced by Rigel to be translated into a higher grain yield in 2 years out of three, in spite of Rigel being taller. The framework used was effective in explaining the differences in AGDM and grain yield induced by the treatments applied. RUE was not affected by the different LAIs realised before anthesis, but was sensitive to variations in the sink capacity in the post-anthesis period and to differences in canopy structure between cultivars.