Year: 2014
Type: conference-paper
Abstract: Since the 1980’s NIR has been used in the Australian rice industry to enable mid-season crop nitrogen uptake to be determined (Batten et al. 1991). The nitrogen uptake value of the rice crop is used to guide crop nitrogen topdressing requirements at panicle initiation (PI) so maximum grain yield can be achieved at harvest. This process involves farmers sampling their crops, weighing the fresh samples, and then sending a microwave dried sub-sample into the lab for nitrogen (N) analysis by NIR. Many farmers find sampling their rice crops in water arduous, so options are being explored that don’t require physical sampling. Rice field maps of Normalised Difference Vegetation Index (NDVI) compiled from aerial or satellite collected data have been used in the rice industry for several years to determine sampling locations for the NIR Tissue Test. Some farmers have used the NDVI map alone to determine N topdressing rates but it is poorly correlated to crop nitrogen. We assessed a handheld “Greenseeker” instrument which measures NDVI. Results obtained from scanning and sampling 240 plots produced a correlation with plant N measured by Dumas combustion of r² = 0.23, PI dry matter r² = 0.79 and PI N uptake r² = 0.82. We also assessed a “Yara-N tester” which is a handheld meter that uses the 650 and 960 nm wavelengths to give a value that is proportional to total chlorophyll which is being used to estimate N in the leaves. The handheld meter was used to test 30 leaves from 120 plots of varying N content. The values obtained were correlated against Dumas combustion r² = 0.54 and PI N uptake r² = 0.075. Current research is investigating scanning the rice crop at panicle initiation with a hyperspectral scanner mounted 1m above the rice crops canopy. The aim is to determine how accurately rice crop nitrogen uptake can be measured remotely and which wavelengths are involved in the relationship. Over the past two seasons 240 plots of varying nitrogen status have been established each season. These plots have been scanned at PI with a Spectra Vista Corporation HR 1024 hyperspectral scanner mounted on a 4 wheel motorbike. Plant samples have been collected at the scanned site in each plot and plant N%, dry matter and nitrogen uptake have been measured. Preliminary investigations have been undertaken and are still underway determining the relationship between the spectra and plant measurements. If the prediction of rice N uptake is accurate enough, and the valuable wavelengths can be determined, they could be used in a filter instrument to scan rice fields. The N uptake map of a rice field could then be used to determine the crops nitrogen requirements on a spatial basis.
Authors: Dunn, Brian, Dehaan, Remy, Dunn, Tina S.
Language: English
Subfield: Ecology
Field: Environmental Science
Domain: Physical Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W2743631551