Current challenge for milk recording in dairy sheep and goats: the simplification of milk sampling design for chemical composition and somatic cell counts of milk

In dairy small ruminants, costs of the standard A4 milk recording design, expressed in production margin per animal, are too high compared to that of dairy cattle. Thus it is a major obstacle to its use on a large scale. Therefore, different procedures of simplification both for quantitative (milk yield) and qualitative recording (milk composition for fat and protein contents and/or somatic cell count) have been proposed. Surveys carried out every two years in dairy sheep, in the framework of its ICAR working group, showed that the use of simplified designs for quantitative milk recording has been steadily increasing since 1990. Conversely, the use of simplified qualitative testing remains still limited. This paper describes a part-lactation sampling design (connected to the AC method for milk yield), which is the simplified qualitative recording method used in France in sheep, both for milk composition and for somatic cell count. The consequences of this simplified design on the loss of precision and the genetic parameters are shown. The favourable genetic trend for milk composition, estimated in the French dairy sheep breeds, shows that this design is reliable for EBVs, but needs cautious use for management purposes. The example of the sheep situation should be extended to goats to allow a larger development of milk recording in this species and particularly the recording of milk composition and somatic cell count.