Layer specific NIR calibration for predicting apparent metabolizable energy (AME) of grains

This project aimed to improve the accuracy of Near Infrared Reflectance (NIR) calibrations for predicting the apparent metabolisable energy (AME) of cereal grains for laying hens. Conducted between July 2022 and May 2025, the study measured AME values of 72 Australian-grown wheat, barley, triticale, sorghum, and maize samples. 

Grains were scanned using both FOSS and Bruker NIR instruments, with calibrations developed for whole and ground samples. A combined 413-sample calibration significantly improved prediction accuracy. For the FOSS instrument, AME could be reliably predicted on a dry matter basis, while Bruker calibrations achieved higher accuracy on an as-fed basis, comparable to current broiler calibrations. Results demonstrated that both instruments provided consistent accuracy across grain types and preparation methods. 

The project delivered robust calibrations, enabling egg farmers to estimate grain AME more precisely when formulating layer diets. This ensures cost-effective feed use, reduces the risk of energy over- or underestimation, and supports sustainability by minimising nutrient wastage.