A systematic review of methodologies for studying behavioral imprinting

Abstract

Behavioral imprinting is a learning phenomenon by which animals acquire preferences for stimuli through perceptual exposure during critical periods, without substantial external reinforcement. Since being acknowledged in 1516 by Sir Thomas More in artificially incubated domestic chickens, imprinting has been reported in diverse species, across various sensory modalities, and during different life-history stages. Due to this diversity, imprinting research uses highly varied methodologies, with distinctive differences between the methods employed for different types of imprinting. We systematically review relevant literature, identifying and describing the range of methodologies used to study imprinting across taxa and modalities. After compiling a representative dataset of 192 behavioral imprinting-focused experiments, we categorize studies by imprinting sensory modality, focal species, ontogenetic stage addressed, and methods applied for both exposure and testing. The majority of studies in the sample focus on filial imprinting in precocial birds but nonfilial types, such as sexual or home range imprinting in altricial and non-avian species, are also present, albeit at far lower proportions. Filial imprinting is studied across different sensory modalities, mostly through applying artificial stimuli, but nonfilial imprinting studies mainly use live animals as stimuli, without isolating the relevant sensory modalities. Most studies of filial imprinting measure preference by spatial proximity, following response, or the suppression of a fear response, whereas most studies of sexual imprinting employ the attempt frequency of sexual behaviors. Finally, we analyze the relative frequencies of methodological approaches in each imprinting category, to highlight potential biases due to uneven research effort rather than intrinsic biology. The patterns and biases in imprinting methodologies that we uncover hamper attempts to establish whether different forms of imprinting share mechanistic foundations, including whether imprinting constitutes a biologically meaningful learning category.