Macaques are risk-averse in a freely moving foraging task

Abstract

Rhesus macaques (Macaca mulatta) appear to be robustly risk-seeking in computerized gambling tasks typically used for electrophysiology. This behavior distinguishes them from many other animals, which are risk-averse, albeit measured in more naturalistic contexts. We wondered whether macaques' risk preferences reflect their evolutionary history or derive from the less naturalistic elements of task design associated with the demands of physiological recording. We assessed macaques' risk attitudes in a task that is somewhat more naturalistic than many that have previously been used: subjects foraged at four feeding stations in a large enclosure. Patches (i.e., stations), provided either stochastically or non-stochastically depleting rewards. Subjects' patch residence times were longer at safe than at risky stations, indicating a preference for safe options. This preference was not attributable to a win-stay-lose-shift heuristic and reversed as the environmental richness increased. These findings highlight the lability of risk attitudes in macaques and support the hypothesis that the ecological validity of a task can influence the expression of risk preference.

Figure 1

Subjects were tested within a large wire mesh enclosure. Juice feeders, attached to the walls of the cage in each corner, provided experimental stimuli and rewards. Five barrels served as perches for subjects to sit on during experimental testing.

Figure 2

Cartoon depiction of the freely moving patch-leaving risk task design and structure. (A) Subjects choose between four possible patches, two safe and two risky. Risk preferences manifest in the allocation of patch entries between the two patch types. (B) Once in a patch subjects receive rewards according to the predefined reward schedule and must choose to either leave or remain in the patch. Risk preferences at this state are expressed as different patch residence times between risky and safe patches.

Figure 3

Gain function, rate as a function of residence time, for safe patches (blue line) and risky patches (red line). The black arrow denotes the abscissa point of the maximum intake rate, and thus the rate-maximizing strategy for both patch types. Due to the programed variation in reward amounts, the gain function for risky patches diverges slightly from the safe patch at long residence times. This divergence arises due to the limitation of reward amounts being bounded at 0 seconds of solenoid open time.

Figure 4

Timeline of the juice gambling task. Offers were presented asynchronously and signaled different gambles for water rewards. Offer stakes were represented by the rectangle’s color (gray, blue, green), while probability was indicated by the size of an overlapping red bar.

Figure 5

Histograms of recorded patch residence times for all subjects in safe (blue) and risky (red) patches for the standard environment condition. Residence time is indexed as the turn length or number of consecutive lever presses at a given patch before leaving (x-axis). The y-axis denotes the number of times a particular turn length occurred at the patch type. Solid lines indicate Gaussian fits to the observed leaving times. Patch residence times are significantly longer for safe than risky patches, indicating risk aversion.

Figure 6

Histograms of recorded patch residence times for all subjects in the rich environment version of task. Plots follow the same conventions as Fig. 3. Subjects resided longer in risky patches than safe patches when the entire reward schedule for all feeder types was increased while maintaining the same variance as used in the standard environment.

Figure 7

Plotted utility functions for two subjects who participated in both the feely moving patch task (lower panels) and a standard chaired economic task (upper panels). Dotted lines represent 95% CI. Two of the same macaques are risk-seeking in the standard task (convex utility curves), and risk-averse the freely moving patch task (concave utility curves).