design Sixteen lactating Anglo-Nubian goats (36.2 ± 0.7 kg of BW) were randomly assigned to four experimental groups.
The experimental design was a quadruplicated 4 × 4 Latin square, with four treatments, four periods and four goats per treatment within each period (resulting in 16 replicates per-treatment for the whole experiment).
The four experimental treatments were randomly assigned to the four groups in the first period.
The goats were housed individually in tie stalls with free access to water and fed on the experimental diets to meet their nutrient requirements according to NRC (2007) recommendations plus a 10% margin.