Using plant wax markers to estimate the diet composition of grazing Holstein dairy cows.

The objective of this study was to test whether diet selection of dairy cows under grazing conditions could be estimated using plant wax markers. Furthermore, differences between 2 cow strains and the effect of concentrate supplementation on plant species selection were investigated. The experiment was a study with a crossover design performed on an organic farm with 12 Swiss Holstein cows and 12 New Zealand Holstein cows. Both experimental periods consisted of a 21-d adaptation and a 7-d measurement period. All cows grazed full time in a rotational stocking system and received either no concentrate or 6 kg/d of a commercial cereal-grain mix. Representative herbage samples of each grazed paddock were taken and botanical composition of subsamples was manually determined. The average proportions of the plant species were 27.8% Lolium perenne, 6.1% Dactylis glomerata, 10.4% Trifolium repens, and 9.0% Taraxacum officinale. Other grass species were merged as "other grass" (38.2%) and other forb species as "other forbs" (8.5%). n-Alkanes, long-chain fatty acids, and long-chain alcohols (LCOH) were analyzed in the samples of plant species, concentrate, and feces from each cow. A linear discriminant analysis indicated that diet components were differentiated best with LCOH (96%) and worst with the combination of all marker groups together (12%). For each marker, the fecal marker recovery (FR) relative to dosed ytterbium was determined in 2 ways. Estimation of diet composition was performed with the software "EatWhat," and results were compared with botanical composition with the Aitchison distance. The results indicate that the diet composition of grazing dairy cows can be estimated using plant wax markers. Additionally, the calculation of FR led to mostly reliable results, yet this approach needs further validation. The most accurate estimation was achieved with the marker combination of n-alkanes and LCOH with a correction for FR. Less accurate estimations were achieved with long-chain fatty acids alone or in combination with n-alkanes. No difference relating to diet selection between the 2 cow strains was recorded, but supplemented cows apparently ingested higher proportions of T. repens than nonsupplemented cows. Awareness that supplementation influences selection behavior of grazing dairy cows may lead to adaptations in botanical composition of the pasture according to the demand of the animals.