Decision-making of goats when exposed to choice feeding: Triggered by taste or smell?

Abstract

Animals recognise known forages. However, it is unclear whether recognition is mainly based on sensory characteristics linked to postingestive feedback or whether the latter is activated each time before a choice is made. Moreover, knowledge is scarce regarding the duration of the memory process and if the memory is primarily activated by smell or a combination of smell and taste. The hypothesis of this study was that in choice situations (1) the preference for one of two offered forages develops in the initial moments of feeding, and (2) this is led by recognition via smell linked to postingestive feedback. In order to study the potential relationships between initial feed intake (FI) (3 min) and short-term FI (3 h), a preference trial as a special form of feed choice experiments with Saanen-type wethers was undertaken. Forages comprised lucerne (Medicago sativa L., first cut), red clover (Trifolium pratense L., first cut) and Italian ryegrass (Lolium multiflorum Lam., second cut). Six different treatments of each forage type were ensiled such that silages of varying quality were produced. Each possible two-way combination of silages and standard hay (n = 21 combinations) was offered and complemented with video recordings of the first 3 min of feed consumption in the morning. Data were analysed via correlation and regression analysis using PROC CORR (Pearson correlation coefficients) and PROC REG of SAS for examination of the relationships between dry matter intake (DMI; g/30 min and g/3 h) and the duration (s) of FI in the first 3 min. The head was moved 0.8–3.0 times between both boxes until the decision was made, and no differences were observed in the frequency of head movements between the silage treatments. The decision did not seem to occur by chance: the initially selected lucerne silages were consumed in greater amounts than the initially non-selected silages within the first 3 min of feeding in 88.3% of the pairwise comparisons. With the red clover silages the proportion was 90.1%, and with the grass silages, 87.3%; 81.7%, 77.5% and 58.9% of the initially non-selected lucerne, red clover and Italian ryegrass silages respectively were not eaten at all (i.e. 0 s). A strong relationship was observed between DMI (g/30 min and g/3 h) and the duration of FI in the first 3 min, indicating a high predictive power of initial FI for the preference that develops within 3 h. Around 90% of the variation in DMI, both expressed as g/3 h and g/30 min, was explained by the duration of FI in the first 3 min of feeding. The results suggest that already in the very first minutes of feeding, recognition of the silages and decision-making processes take place based on the smell emitted by the silages.