Empirical prediction of organic matter digestibility in grasses, legumes and their mixtures from dairy farms in Atlantic Spain

Abstract

This study characterized the chemical composition and organic matter digestibility (OMD) of herbage from commercial dairy farms in Galicia (NW Spain), a humid‐temperate Atlantic region, and developed empirical prediction equations for OMD based on routine wet‐chemistry traits. A total of 593 samples were collected over 7 years (2007–2013), including grasses ( n  = 235; e.g., Italian and perennial ryegrass, cocksfoot, fescue), legumes ( n  = 212; e.g., white/red/crimson/Persian clovers, serradella, alfalfa) and grass–legume mixtures ( n  = 146). OMD was determined by a two‐stage in vitro method (buffered rumen fluid and pepsin).

Across the full dataset, acid detergent fiber (ADF) and neutral detergent fiber (NDF) were the strongest predictors of OMD. The best‐performing overall equation was: OMD (%) = 106.30–0.82·ADF − 0.25·CP − 0.37·HCEL, achieving external validation r ²  = 0.69, standard error of prediction (SEP) = ±3.66%, ratio of performance to deviation (RPDev) = 1.80 and range error ratio (RERev) = 9.1. Group‐specific equations were as follows:

Grasses: OMD (%) = 79.72–0.55·NDF − 0.21·OM (external validation r ²  = 0.54; SEP = ±3.60%; RPDev ≈ 1.39–1.47; RERev ≈ 8.0–8.4);

Legumes: OMD (%) = 75.67–0.51·NDF + 0.28·CP + 1.66·WSC − 0.047·WSC ² (external validation r ²  = 0.75; SEP = ±3.43%);

Mixtures: OMD (%) = 88.12–0.70·ADF + 0.25·WSC (external validation r ²  = 0.43; SEP = ±3.93%; RPDev ≈ 1.16–1.32; RERev ≈ 6.0–6.8).

These results show that empirical equations based on routinely available chemical analyses can provide a rapid, cost‐effective tool to estimate herbage OMD for Atlantic dairy systems; however, transfer to contrasting climates should be preceded by local validation or recalibration.