| Abstract
| - Two-bottle preference tests have often been used to make inferences about gustatory thresholds. The validity of such inferences depends on the extent to which taste differences produce differential fluid consumptions from the bottles. If equal amounts are consumed despite perceived gustatory dissimilarity, inaccurate threshold estimates result. Modifications of common preference-test procedures, interacting with tastant properties and genetic variation, disclosed instances of such non-discrimination among inbred strains of mice. With an hedonically neutral to mildly unpalatable tastant (sodium cyclamate), taste-aversion conditioning yielded lower thresholds. With a palatable tastant (maltose), testing without prior conditioning usually indicated lower thresholds. Descending concentration series produced lower thresholds than ascending concentration series, for both palatable (sucrose) and neutral (cyclamate) tastants. Previous experience with the tastants elevated preference scores, yielding lower thresholds with palatable tastants (sucrose, maltose) and higher thresholds with an unpalatable tastant (sucrose octaacetate). Within-strain discrepancies between thresholds indicated by the different methods were often large (2-3 log molar concentration steps). Amongstrain differences of comparable magnitude, both in preferences and threshold estimates, were found for sucrose, maltose and cyclamate.
|
