Behavioral discrimination of binary mixtures and their components: effects of mixture interactions on coding of stimulus intensity and quality

There is mounting evidence that mixture interactions resultin a physiological response that is different from that predictedfrom observed responses to individual mixture components. Mixtureinteractions that act to alter the neural coding of mixtureintensity (intensity mixture interactions) or quality (patternmixture interactions) may ultimately lead to dramatic differencesbetween the perceived intensities and qualities of a mixtureand its components. These perceptions could be expressed andobserved at the behavioral level. Toward examining this question,we have tested the ability of the Florida spiny lobster (Panulirusargus) to behaviorally discriminate between three odorant compounds[adenosine 5'-monophosphate (AMP), L-glutamate (Glu), and taurine(Tau)] and their binary mixtures through the use of a differentialaversive associative conditioning paradigm. Six groups of lobsterswere used, each being conditioned to avoid one of the singlecompounds or binary mixtures. Behavioral expression of intensitymixture interactions was evident. Preconditioning response magnitudesto binary mixtures were either less than those to their components(e.g. AMP + Glu) or less than predicted from responses to theircomponents (e.g. AMP + Tau). Behavioral expression of patternmixture interactions was also observed. Relationships betweenthe quality of each binary mixture and the qualities of themixture's components were determined from the results of analysisof variance and multidimensional scaling analysis. Analysesincorporated observed responses to all stimuli and ‘predicted’responses to the binary mixtures. Lobsters easily discriminatedbetween the qualities of AMP, Glu and Tau. The quality of themixture of AMP + Glu was different from either component aswell as from the predicted value for this mixture. The mixtureof AMP + Tau was intermediate between both components and wassimilar to the predicted value. The mixture of Glu + Tau, whilemore similar to Glu than to Tau, was different from the predictedvalue, and there was some indication that the Glu was actingto suppress the response to Tau. Behavioral results for AMP+ Tau, which suggest no pattern mixture interactions betweenthese compounds, are in accordance with results of recentlyconducted binding assays which indicate independent receptorsfor these compounds (Olson et al., 1992). Results, especiallyfor AMP + Glu and Glu + Tau, are consistent with results ofour electrophysiological analysis of the effects of patternmixture interactions on coding of stimulus quality and intensityby olfactory receptor cells (Derby et al., 1991a,b). This providesfurther evidence for the effects of peripherally initiated mixtureinteractions on the coding and perception of the quality ofodorant mixtures. ¹Present address: Departments of Psychology and Biology, GeorgiaState University, University Plaza, Atlanta, Georgia 30303,USA     

Behavioral resolution of quality of odorant mixtures by spiny lobsters: differential aversive conditioning of olfactory responses

A differential aversive associative conditioning paradigm wasused to demonstrate the ability of spiny lobsters to discriminatebehaviorally between four 41-component chemical mixtures. Thesemixtures were based on natural extracts of crab, mullet, oysterand shrimp. Previously, we demonstrated, using the same paradigm,that lobsters conditioned to avoid the shrimp mixture coulddiscriminate, to varying degrees, between the shrimp mixtureand the other three mixtures. The present study was performedto compare the perceived quality of all four mixtures. Thiswas accomplished by conditioning four groups of animals, eachgroup to a different mixture type, and then using multidimensionalscaling (MDS) analysis to compare simultaneously the similaritiesand dissimilarities among the four mixtures as inferred by passiveand active avoidance behaviors. These two types of behaviorsrevealed different aspects of the discrimination: greater differentiationof the conditioned mixture from the non-conditioned mixtureswas indicated by active avoidance behavior, while gradationsin the degree of discrimination between the conditioned mixtureand each of the non-conditioned mixtures were indicated by passiveavoidance behaviors. Overall, lobsters perceived crab and shnmpmixtures as being similar to one another and dissimilar to mulletand oyster mixtures, while they perceived mullet and oystermixtures to be dissimilar to one another and to the crab andshrimp mixtures. Comparison of results from MDS analyses ofbehavioral mixture discrimination, neural mixture discriminationand mixture compositions may be used both to provide an indicationof the type of neural coding used to make these behavioral discriminationsand to identify the components of the mixtures that are responsiblefor recognition of and differential response to the mixtures.