Abstract

The preferred grain protein concentration (CP) of malting barley is 10.5-11.0%, but 9.5-11.5% is acceptable. It is a challenge for farmers to achieve this target with crops grown in heterogeneous fields and exposed to fluctuating weather conditions. There are also economic and environmental reasons to balance the supply of nutrients to plant requirements. This forms the basis for precision agriculture, where barley has received limited attention. The key factor for precision agriculture in malting barley is the ability to predict CP from early observations of the crop so as to control a second fertiliser application. This thesis investigates the possibility of predicting malting barley grain CP at an early stage of development and of using a second fertilisation application during growth for total nitrogen (N) adjustment. Three experiments were conducted. The first consisted of eleven field trials (1992-1994) and was used to compare broadcasting/harrowing and combi-drilling for applying full-rate fertiliser at sowing using two types of fertilisers; pure N and one also containing phosphorus (NP). The second experiment consisted of sixteen fertiliser field trials (2001-2003) and was used to examine the possibility of postponing the decision on total N. The third consisted of three evenly fertilised fields (2002-2004). In experiments 2 and 3, canopy reflectance was measured at developmental stages BBCH 32, 45 and 69. Soil macronutrients, organic matter and mechanical composition were analysed in all ex-periments. Malting barley yield was higher when fertiliser was combi-drilled into the soil and when NP fertiliser was used. Grain CP was predicted in the field (R2adj = 0.73) from soil electrical conductivity (SECa), the canopy reflection-based vegetation index (VI) TCARI/OSAVI estimated at BBCH 32 and the sum of daily maximum temperatures during anthesis and grain filling (STS). In the fertilisation trials, CP was predicted (R2adj = 0.83) by sowing day number and the VI TCARI evaluated together with solar angle at measurement. Grain yield was independent, and grain CP almost independent, of whether all fertiliser was applied at sowing or divided between sowing and BBCH 32.