Laboratoire de Comportement Animal, CNRS URA 1291, INRA, 37380 Nouzilly, France.
Experience with chemosensory stimuli in the prenatal environment has first been shown to contribute to the chemosensory abilities of rat newborns. To assess the generality of olfactory engagement in adaptive responses around birth, perinates of other mammals were examined, including rabbits, sheep, and humans. When newborn (N) rabbits, lambs or humans were exposed to a paired-odor test between their amniotic fluid (AF) and their mother's colostrum (CO), they evinced no differential orientation to either stimulus. This undifferentiated response was interpreted in terms of chemosensory and/or motivational similarity between both perinatal substrates. This similarity hypothesis raised a set of predictions: 1) because of contemporaneous influence of dietary aromas, AF-CO similarity should be greatest close to parturition; 2) physiological changes in milk composition due to lactogenesis should be paralleled with the appearance of differentiated responses; 3) as no continuity can exist between AF and formula milks, artificially fed N should differentiate AF and formula odors; 4) N should differentiate milks from different lactating females; 5) fetuses should encode odors in the womb and N should use this information differentially; 6) N should prefer CO that olfactorily matches the odor that presumably dominated in their AF; finally, 7) if chemosensory continuity has adaptive value, its disruption should affect behavior in the postnatal niche. These predictions were empirically supported. Thus, mammalian N have the ability to selectively retrieve unique olfactory cues gained in their prenatal environment. Pre- and postnatal environments share sources of chemosensory invariance, so that N are preadapted to otherwise non-predictable features of the postnatal ecology.