Humans and insects decide in similar ways

Abstract

Behavioral ecologists assume that animals use a motivational mechanism for decisions such as action selection and time allocation, allowing the maximization of their fitness. They consider both the proximate and ultimate causes of behavior in order to understand this type of decision-making in animals. Experimental psychologists and neuroeconomists also study how agents make decisions but they consider the proximate causes of the behavior. In the case of patch-leaving, motivation-based decision-making remains simple speculation. In contrast to other animals, human beings can assess and evaluate their own motivation by an introspection process. It is then possible to study the declared motivation of humans during decision-making and discuss the mechanism used as well as its evolutionary significance. In this study, we combine both the proximate and ultimate causes of behavior for a better understanding of the human decision-making process. We show for the first time ever that human subjects use a motivational mechanism similar to small insects such as parasitoids and bumblebees to decide when to leave a patch. This result is relevant for behavioral ecologists as it supports the biological realism of this mechanism. Humans seem to use a motivational mechanism of decision making known to be adaptive to a heterogeneously distributed resource. As hypothesized by Hutchinson et al. and Wilke and Todd, our results are consistent with the evolutionary shaping of decision making because hominoids were hunters and gatherers on food patches for more than two million years. We discuss the plausibility of a neural basis for the motivation mechanism highlighted here, bridging the gap between behavioral ecology and neuroeconomy. Thus, both the motivational mechanism observed here and the neuroeconomy findings are most likely adaptations that were selected for during ancestral times.

Figure 1. The motivational mechanisms of decision-making in parasitoids and the bumblebee.

The insect enters a patch with an initial motivation to stay, which decreases linearly when no reward is found (in the case of parasitoids, hosts for laying eggs). Note that the motivation suddenly increases (incremental mechanism —) or decreases (decremental mechanism – –) when a reward is found (Δ). The decision to leave the patch occurs when the motivation falls below a given threshold (*). From Waage (1979).

Figure 2. Screenshots of the virtual foraging game.

Top: An overview of the virtual meadow with the spatial distribution of the patches. Middle: The chests when the player enters a patch. Bottom: The player opened a filled chest and found the resource. The French text on the bottom screenshot asks the subject to note his motivation to stay in the dome (from 0 to 9). At any given time, the player knows both the number of items found (left top corner of the screen) and the instantaneous yield (left bottom corner).

Figure 3. Relationship between the visual classification of the overall fit and goodness of fit of individual motivation, given by R ² in the adjusted model 5.

b: bad visual adjustment (strong discrepancy between the course of motivation and the fitted values); m: medium visual adjustment (partial consistency); g: good visual adjustment (total consistency, no systematic bias).

Figure 4. Relationship between the size of the increments and the delay since the last discovery of an item.

Each petal of the sunflower plot represents one overlapped point. The relationship (red dotted line, increment = 13.67627(1-exp(-0.06025delay))) is in accordance with Waage's model but the effect is weak and blurred by a large noise.

Figure 5. Motivational course during the foraging time of a player classified as “poor”, “medium” and “good”, using the motivation adjustment described in the text.

Figure 6. Relationship between Cox's model β and the term I/b from Waage's model.

The Y-axis refers to the values of the effect of opening a filled chest plotted against on the hazard rate for Cox's proportional hazard model. The X-axis is obtained from the motivational model.