Sources of variation in splanchnic blood flow in steers

Data from six experiments were used to describe sources of variation in blood flow through portal-drained viscera and liver of 33 steers that were fed equal-sized meals every 2 h. The experiments were designed to create "steady-state" conditions under which response to various dietary or physiological treatments was assessed. Sums of squares for blood concentration of blood flow marker, venoarterial differences in blood flow marker, and blood flow were divided into variation attributable to steer; period (or time); the steer x period interaction; sampling days within steer and period; and replications (or samplings) within day, steer, and period. Steer was the largest single source of variation in arterial concentration of blood flow marker, accounting for from 42 to 80% of sums of squares among the experiments. However, replication within day, steer, and period accounted for more variation than steer in portal or hepatic blood flow in four of the six experiments. When balanced for treatment effects, steer accounted for 59%, period and the steer x period interaction accounted for 14%, and replication accounted for 27% of variation in portal blood flow. Corresponding percentages for hepatic blood flow were 45, 20, and 35%. I conclude that steer and replication within steer x period cells in a matrix of treatments are the two largest sources of variation and that there is more variation among samples on a given day within steer x period cells than among days in the same cells.