ERA (Evidence for Resilient Agriculture) Feed Library
Domain:
agricultureRecord type:
dataset
Creator:
MulSte
Nab
NabEditor:
BioBioAll
Publisher:
Har
Host:
The ERA Feed Library is an analysis-ready reference dataset for livestock nutrition, covering more than 1,200 feed ingredients spanning forages, grains, by-products, legumes, shrubs, tree leaves, and grasses, including multiple processing forms.
It consolidates manually extracted data from over 700 livestock nutrition studies across Africa and harmonizes ingredient naming using the AOM classification.
To address missing values and standardize nutrient information, the dataset integrates and cross-references values from the ILRI African Feeds Database and Feedipedia. All nutritional variables are converted into canonical units (g/kg DM, MJ/kg DM), and a transparent multi-step inference procedure fills data gaps.
The dataset includes: - Harmonized ingredient names with nutritional values - Mapping to Feedipedia and ILRI feed codes - Country-level summary means for each ingredient
This resource provides a comprehensive, high-quality feed library suitable for livestock nutrition modelling, Tier-2 methane emission estimation, ration analysis, and agricultural research across African production systems.
Methodology:The ERA Feed Library compiles livestock nutrition evidence from the ERA database into a unified, analysis-ready, ingredient-level resource. Feed names from more than 700 African livestock studies are harmonized to a consistent vocabulary, resolving synonyms and inconsistent naming, harmonizing processes and adding scientific name linkage.
All nutrient and energy values are standardized into canonical units (g/kg DM and MJ/kg DM). When values are missing, a structured completion workflow fills gaps using two external reference sources (Feedipedia and the ILRI African Feeds Database), followed by inference from ingredient-level means. Each step preserves full provenance, ensuring that every value is traceable to raw data, external sources, or statistical inference.
To address missing values and standardize nutrient information, the dataset integrates and cross-references values from the ILRI African Feeds Database and Feedipedia. All nutritional variables are converted into canonical units (g/kg DM, MJ/kg DM), and a transparent multi-step inference procedure fills data gaps.
The dataset includes: - Harmonized ingredient names with nutritional values - Mapping to Feedipedia and ILRI feed codes - Country-level summary means for each ingredient
This resource provides a comprehensive, high-quality feed library suitable for livestock nutrition modelling, Tier-2 methane emission estimation, ration analysis, and agricultural research across African production systems.
Methodology:The ERA Feed Library compiles livestock nutrition evidence from the ERA database into a unified, analysis-ready, ingredient-level resource. Feed names from more than 700 African livestock studies are harmonized to a consistent vocabulary, resolving synonyms and inconsistent naming, harmonizing processes and adding scientific name linkage.
All nutrient and energy values are standardized into canonical units (g/kg DM and MJ/kg DM). When values are missing, a structured completion workflow fills gaps using two external reference sources (Feedipedia and the ILRI African Feeds Database), followed by inference from ingredient-level means. Each step preserves full provenance, ensuring that every value is traceable to raw data, external sources, or statistical inference.