Development of feed composition tables using a statistical screening procedure

Authors: TRAN, HUYEN - University Of Nebraska; SCHLAGETER-TELLO, ANDRES - University Of Kentucky; CAPREZ, ADAM - University Of Nebraska; MILLER, PHILLIP - University Of Nebraska; Hall, Mary Beth; WEISS, WILLIAM - The Ohio State University; KONONOFF, PAUL - University Of Nebraska

Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/14/2019
Publication Date: 4/1/2020
Citation: Tran, H., Schlageter-Tello, A., Caprez, A., Miller, P.S., Hall, M., Weiss, W.P., Kononoff, P.J. 2020. Development of feed composition tables using a statistical screening procedure. Journal of Dairy Science. https://doi.org/10.3168/jds.2019-16702.
DOI: https://doi.org/10.3168/jds.2019-16702

Interpretive Summary: Although feed testing laboratories generate a large amount of feed composition data, managing, processing and reporting data that originate from multiple resources is a challenge. Feed composition data from 4 commercial laboratories in the United States were acquired and subjected to a statistical screening procedure, modified, and programmed to construct tables summarizing feed composition. The procedure was useful in identifying different maturities within populations of forage classes, differentiating between oilseeds and different oilseed meals, and identifying misclassified samples in feed samples with similar names. The resulting feed database will be useful in updating feed libraries used in diet formulation software.

Technical Abstract: Millions of feed composition records generated annually by testing laboratories are high-value assets that can be leveraged to benefit the animal nutrition community. However, managing, handling, and processing feed composition data that originate from multiple sources, lack standardized feed names, and contain outliers is challenging. Efficient methods that consolidate and screen such data are needed to develop feed composition databases with accurate means and standard deviations (SD). Considering the interest of the animal science community in data management and the importance of feed composition tables for dairy cattle industry, the objective of the current manuscript is to describe procedures used to construct feed composition tables from large datasets. A published statistical procedure designed to screen feed composition data was employed, modified, and programmed to operate using Python and SAS. The 2.76 million data received from 4 commercial feed testing laboratories were used to develop procedures and to construct tables summarizing feed composition. Briefly, feed names and nutrient analytes across laboratories were standardized, then erroneous data and duplicate samples were removed. Histogram, univariate, and principal component analyses were used to identify and remove outliers having key nutrients outside of the mean ± 3.5 SD. Clustering procedures identified subgroups of feeds within a feedstuff. Aside from the clustering step that was programmed in Python to automatically execute in SAS, all steps were programmed and automatically conducted using Python followed by a manual evaluation of the resulting Pearson correlation matrices and clusters. The input dataset contained 42, 94, 162, and 270 feeds from 4 laboratories and were comprised of 25 to 30 nutrients. The final database included 174 feeds and 1.48 million feed samples. The developed procedures effectively classified byproducts (e.g. distillers grains and solubles as low or high fat ), forages (e.g. legume or grass-legume mixture by maturity), and oilseeds vs. meal (e.g. soybeans as whole raw seeds, soybean meal expellers or solvent extracted) into distinct sub-populations. Results from these analyses suggest that the procedure can provide a robust tool to construct and update large feed datasets. This approach can also be used by commercial laboratories, feed manufacturers, animal producers, and other professionals to process feed composition datasets and update feed libraries.