Do pigeons (Columba livia) use information about the absence of food appropriately? A further look into suboptimal choice.

Abstract

In the natural environment, when an animal encounters a stimulus that signals the absence of food-a 'bad-news' stimulus-it will most likely redirect its search to another patch or prey. Because the animal does not pay the opportunity cost of waiting in the presence of a bad-news stimulus, the properties of the stimulus (e.g., its duration and probability) may have little impact in the evolution of the decision processes deployed in these circumstances. Hence, in the laboratory, when animals are forced to experience a bad-news stimulus they seem to ignore its duration, even though they pay the cost of waiting. Under certain circumstances, this insensitivity to the opportunity cost can lead to suboptimal preferences, such as a preference for an option yielding a low rather than a high rate of reinforcement. In 2 experiments, we tested Vasconcelos, Monteiro, and Kacelnik's (2015) assumption that, if given the opportunity, animals will escape the bad-news stimulus. To predict when an escape response should occur, we incorporated ideas from the prey choice model into Vasconcelos et al. (2015) model and made 2 novel predictions. Namely, both longer intertrial intervals and longer durations of signals predicting food or no food should lead to higher proportions of escape responses. The results of 2 experiments with pigeons supported these predictions. (PsycINFO Database Record

Frequency and value both matter in the suboptimal choice procedure

“…This was the case irrespective of condition order. These results are inconsistent with predictions of both the signal-value hypothesis (e.g., Stagner et al, 2012) and the recent models inspired by optimal foraging theory (Fortes et al, 2017;Vasconcelos et al, 2018). Probe trials with the S+ terminal-link stimuli of the Sig-Both condition revealed strong preference for the stimulus correlated with the higher frequency of reinforcement.…”

“…This was the case irrespective of condition order. These results are inconsistent with predictions of both the signal‐value hypothesis (e.g., Stagner et al, ) and the recent models inspired by optimal foraging theory (Fortes et al, ; Vasconcelos et al, ).…”

“…That opportunity is not afforded in most laboratory paradigms and, therefore, what seems suboptimal in the laboratory is not likely suboptimal in the real world. A recent attempt to formally quantify this functional approach, called the reinforcement rate model (RRM), focuses on the rate of energy intake in the presence of informative (i.e., signaled) stimuli and noninformative (unsignaled) stimuli (Fortes et al, ). While the RRM has met with success (Fortes et al, ; Fortes et al, ; Vasconcelos et al, ), the present results provide a challenge to RRM, which predicts indifference in the Sig‐Both condition because both alternatives are informative and thus functionally equivalent.…”

“…A recent attempt to formally quantify this functional approach, called the reinforcement rate model (RRM), focuses on the rate of energy intake in the presence of informative (i.e., signaled) stimuli and noninformative (unsignaled) stimuli (Fortes et al, ). While the RRM has met with success (Fortes et al, ; Fortes et al, ; Vasconcelos et al, ), the present results provide a challenge to RRM, which predicts indifference in the Sig‐Both condition because both alternatives are informative and thus functionally equivalent. Therefore, like the signal‐value hypothesis, RRM predicts indifference in the Sig‐Both condition and strong suboptimal preference in the Sig‐Sub condition, neither of which occurred.…”

“…Another recent model, based on considerations of foraging behavior in the natural environment, similarly predicts that reinforcement frequency on a signaled alternative should have no effect on preference (Fortes, Machado, & Vasconcelos, 2017;Fortes, Pinto, Machado, & Vasconcelos, 2018;Fortes, Vasconcelos, & Machado, 2016;Vasconcelos, Machado, & Pandeirada, 2018). One formalization of this functional model, called the reinforcement rate model (RRM) focuses on the rate of energy intake in the presence of informative (i.e., signaled) stimuli and noninformative (unsignaled) stimuli (Fortes et al, 2017).…”