Application of n-alkanes as diet composition markers in grazing/browsing goats and sheep: effect of using different faecal recovery corrections and plant species grouping approaches

Abstract

The present study aimed to test the effect of applying different sets of n-alkane faecal recoveries and different plant species grouping approaches on the estimates of the diet selected by goats and sheep when grazing/browsing complex vegetation communities composed of mosaics of heather-gorse vegetation interspersed with patches of perennial ryegrass. The use of different sets of n-alkane faecal recoveries significantly affected the estimated proportions of most vegetation components in both animal species. The diet estimates were also significantly affected by the calculation approach: A1 (using all individual plant species); A2 (grouping plant species with similar n-alkane profile, with equal weighting within the group); and A3 (grouping plant species with similar n-alkane profile, with different weighting according to the botanical composition of the plot). There was a significant interaction between calculation approach (A) and faecal recovery sets (FR) on the resulting estimates of the Ericaceae species in the diet of goats and of all diet components in sheep. When using A1 approach, the estimated proportion of Agrostis capillaris in the diets was high and seemed to be overestimated, compared with its low availability (1.1% of total cover) in the field. However, the application of calculation approaches A2 and A3 gave lower dietary proportions of component D2 (Agrostis capillaris and Ulex gallii) and higher proportions of components D1 (Lolium perenne and Pseudarrhenatherum longifolium), D3 (Erica umbellata, Erica cinerea, Erica australis, Calluna vulgaris) and Erica arborea. These results indicate that the application of the n-alkane technique for estimation of the diet composition in field with complex vegetation conditions is sensitive to the use of faecal recovery corrections and to the grouping of possible dietary components.