Meta-analysis of the effect of pasture allowance on pasture intake, milk production, and grazing behavior of dairy cows grazing temperate grasslands

Daily pasture allowance (PA) is defined as the product of pregrazing pasture mass and offered area, and is the major grazing management factor determining pasture utilization per unit area and daily performance of grazing dairy cows. The objective of the present study was to perform a meta-analysis reviewing the effect of PA on pasture intake, milk production, milk composition, and grazing behavior of dairy cows. Experiments studying the effect of PA on pasture intake or milk production, which eventually included milk composition or grazing behavior data, or both, were selected to create a database. Papers were selected only if at least 2 PA were compared under the same experimental conditions, particularly the same pasture mass (i.e., where PA levels were only obtained through changes in daily offered area). The final database included 97 PA comparisons reported in 56 papers. For analytical purposes, the database was subdivided into 3 subsets that varied according to the estimation height (EH) at which PA was determined; that is, PA above ground level (PA0 subset), PA above 2.5 to 3.5cm (PA3 subset), and PA above 4 to 5cm (PA5 subset). Statistical analyses were conducted independently on the PA0, PA3, and PA5 subsets and on the whole database (global analysis) by using linear and nonlinear mixed-model procedures. The curves, either exponential, quadratic, or linear, describing the effects of PA on pasture intake, milk production, or grazing behavior of dairy cows are conceptually similar, whatever the EH. The equations describing these curves are, however, specific for each EH. Accordingly, from typical low to high PA, the increase in pasture intake (0.13 vs. 0.21 vs. 0.28kg/kg of PA), milk production (0.11 vs. 0.17 vs. 0.24kg/kg of PA), and milk solids production (0.008 vs. 0.010 vs. 0.013kg/kg of PA) per kilogram of increase in PA was lower for PA0 than for PA3, and for PA3 than for PA5. Grazing time increased from low to medium PA and did not vary from medium to high PA. Pasture intake rate seemed to increase from low to medium PA because of greater bite mass, whereas it increased from medium to high PA because of greater biting rate. The present meta-analysis demonstrated that the general relationship between PA and any dependent variable is quite strong and independent of EH. This suggests no specific relationship for some parts of the world or methodology approach, with a high portability of the global equations calculated here. These results are useful for improving grazing management and modeling on pasture-based dairy systems.