Nitrogen Management in Malting Barley

There are many factors that affect yield and quality of malting barley. Our research measures the variation in yield and quality that California growers can manage with their nitrogen fertilizer (N) rate and timing. We measured yield and quality from several high performing malting varieties grown under a range of N fertilizer rates and application timings in 6 site-years across Northern California with varying yield potential. Simultaneously we recorded proximally sensed measurements from the plant-soil environment to better understand and quantify variability in yield and quality.

Initial results indicate that, similar to other research, grain protein yield increased as N rates increased at all sites. However, a majority of the variation in grain protein yield (81%) was due to effects other than N management. Yield varied by N timing as well as N rate, with pre-plant N applications having significantly lower yields than N applications that were split between pre-plant and tillering or applied all at tillering. Under higher rainfall conditions, timing differences were exaggerated since more of pre-plant N was likely lost. Grain protein concentrations were not affected by timing.

An ideal N rate and application timing would result in the highest possible yields while still meeting protein requirement (~9 – 10.5% protein for all malt craft brew). In our trials, upland, rainfed sites with low yield potential had higher grain protein. In order to reach acceptable protein levels, very little N could be applied in these environments. Lowland environments with higher yield potential and low starting soil N could utilize more N without reaching unacceptable grain protein concentrations.

Malt quality parameters, such as free amino nitrogen (FAN), diastatic power, protein modification (S/T), and beta glucan did not have strong relationships with fertilizer management, and were mainly controlled by the growing and malting environments.

Preliminary results from the proximally sensed measurements recorded in these trials show that NDVI at boot is highly saturated and not able to differentiate between high N treatments. NDVI from the Trimble Greenseeker explained 21% of variation in protein yield across environments at the boot stage. NDVI measurements from sUAS born multispectral cameras are on par with handheld NDVI measurements (R2 = 0.87). Other remote sensing indices such as NDRE, GNDVI, and VARI, often used with satellite data for agriculture models, are being explored. Initial results indicate that these indices do not saturate as easily as NDVI and thereby may capture more phenotypic variation.

drone image