Identification of spectral bands to detect nitrogen and phosphorus deficiencies in winter wheat
Abstract
Past research in winter wheat (Triticum aestivum L.) showed the potential of spectral indices to detect winter wheat phosphorus (P) status although no specific indices were developed. An experiment was conducted at Perkins, OK in 2007/2009 to identify single or combined spectral indices that can detect nitrogen (N) and phosphorus (P) as well as P independent of N deficiency in winter wheat. A randomized complete block design with three replications was employed. Treatments included twelve factorial combinations of three rates of P (0, 34 and 67 kg P ha-1) and four rates of N (0, 56,112,168 kg N ha-1). Four types of spectral radiance measurements were collected and these included a full bandwidth spectrometer (300nm to 1100nm), Greenseeker^TM sensor, New-experimental 4 band sensor, and digital pictures at four different winter wheat growth stages. Forage and grain yield were collected and measured. Forage and grain N and P as well as postharvest soil residual P contents were determined. Correlation analysis was used to test the relationship between spectral readings vs. forage and grain yield, and forage and grain N and P content of winter wheat. Similarly, stepwise regression procedure was used to select wavelengths and rations of wavelengths that can detect winter wheat N and P status. Analysis of variance (ANOVA) was employed to test the effect of N and P rates on several variables. Spectral reflectances at certain wavelengths were identified from spectrometer data and indices that can detect N and P status of winter wheat were developed. Spectral Phosphorus Indices: SPI1 and SPI2 were developed from 915 nm numerators to 455 nm denominators, and 865 nm numerators to 505 nm denominator wavelengths, each averaged from10 nm bandwidth, respectively were significantly correlated with winter wheat forage P status over the two-year study. Also, these indices were significant for forage N content. Reflectances at single wavelengths, each average from 10 nm band widths between 605 to 695 nm were detected forage P content at Feekes 10 in 2008 and Feekes 7 and 10 in 2009 while the reflectance at wavelengths from 455 to 715 nm and from 815 to 925 nm were consistently correlated with forage N content at the above mentioned growth stages. There was no index, except the promising result of picture index (R/G), that could detect winter wheat forage P content independent of forage N content using the above instruments in the two-year study. This was likely because 1) wavelengths that detect forage P content were found within the range of the wavelengths that can detect forage N status and, 2) Nitrogen rate affected crop biomass and resulted in forage P content dilution as the crop grows.
Collections
- OSU Theses [15752]