A comparison of five methods that estimate meal criteria for cattle

III data collected for feeding behaviour analysis, feeding events are generally separated by many very short to very long intervals during which no feeding occurs. When feeding is clustered in bouts, a meal criterion (that is the longest ton-feeding interval accepted as part of a meal) must be estimated before events can be grouped into meals. Until recently, three methods that estimate quantitative meal criteria were available. These methods consist of fitting a 'broken-stick' (two straight intersecting lines, both with a negative slope) to the frequency distribution (method 1), the log(e)-transformed cumulative frequency distribution (the log-survivorship curve; method 2) or the log(e)-transformed frequency distribution (method 3) of intervals between events. Recently, new methods have been proposed that fit either two (method 4) or three (method 5) Gaussians to the frequency distribution of log,transformed interval length (log-normal models). We compare the estimates obtained with these five methods when applied to a data set of 79575 intervals between visits to food dispensers. These were recorded with 16 lactating cows during an average period of 156.6 (s.d. 51.5) days per cow. Meal criteria were estimated as 1.9, 6.0, 7.5, 32.4 and 49.1 min by methods I to 5, respectively. Estimated daily number of meals ranged from 5.7 to 12.1 per cow and estimated average meal size from 4.0 to 8.4 kg. The observed probabilities of cows initiating feeding in relation to time since feeding last showed best agreement with the predictions of the log-normal models. We conclude that the first three methods do not, while log-normal models do, have an adequate biological basis for a clear interpretation of the estimated meal criteria. Log-normal models are, therefore, the most promising for estimating meal criteria in cattle and probably iii other species as well.