Helping behavior in prairie voles: A model of empathy and the importance of oxytocin

Abstract

Several studies suggest that rodents show empathic responses and helping behavior toward others. We examined whether prairie voles would help conspecifics who were soaked in water by opening a door to a safe area. Door-opening latency decreased as task sessions progressed. Female and male voles stayed close to the soaked voles' side at equal rates and opened the door with similar latencies. When the conspecific was not soaked in water, the door-opening latency did not decrease. This suggests that the distress of the conspecific is necessary for learning to open the door and that the door-opening performed by prairie voles corresponds to helping behavior. Additionally, we examined the helping behavior in prairie voles in which oxytocin receptors were genetically knocked out. Oxytocin receptor knockout voles demonstrated less learning of the door-opening behavior and less interest in soaked conspecifics. This suggests that oxytocin is important for the emergence of helping behavior.

Keywords: Evolutionary biology; rodent behavior.

Graphical abstract

Figure 1

Experimental apparatus and behavioral results in wild-type voles (A) Schematic diagram of the experimental apparatus. A helper vole could rescue a soaked vole from the distressed situation of being in water by opening the circular door. (B) A photo image during the door-opening task. An acrylic plate covered both the ground and pool areas, which reflected the video camera and the ceiling of the laboratory. The area indicated by the blue shade is defined as the soaked vole’s side. (C) In this experiment, all sex combinations of the helper and soaked voles were examined. (D) The mean latency of door-opening in the helper voles of all sex combinations (n = 7 for all the groups; post hoc test using Holm’s method (∗∗ ps < 0.01), following the effect of session in a three-way ANOVA; F(6, 144) = 23.35, p < 0.001). (E) The percentage of door-opening pairs in each sex combination group. (F) The total number of door-opening sessions (two-way ANOVA; the effect of sex combination; F(1, 24) = 4.32, ∗p = 0.049) and the number of consecutive door-opening sessions. (G) The percentage of time that the helper stayed on the soaked vole’s side. (H) The mean duration of huddling during the 2-min interaction period after opening the door. (I) The time male and female helper voles spent in the ground area during the habituation period. The dashed line indicates the expected value. The error bars indicate the SE of means.

Figure 2

Prairie voles show prolonged latency of door-opening behavior toward non-soaked cagemates (A–C) There were two groups of paired voles. In one group, one of the pair (a cagemate of the helper vole) was soaked in water (Distress, seven pairs of same-sex males). In the other group, the cagemate was not soaked in water (No Distress, 12 pairs of same-sex males). The mean latency (B) and the percentage (C) of door-opening of the helper voles in the Distress and No Distress groups (two-way ANOVA; the effect of group; F(1, 17) = 5.98, ∗∗p = 0.003). (D) The total number of door-opening sessions (Welch’s t-test; t(16.26) = 3.30, ∗∗p = 0.004) and the number of consecutive door-opening sessions (Welch’s t-test; t(16.83) = 2.98, ∗∗p = 0.009) in the Distress and No Distress groups. (E) The percentage of time that the helper voles stayed on the soaked vole’s side in the Distress and No Distress groups. (F) The mean duration of huddling during the 2-min interaction period after opening the door in the Distress and No Distress groups. The error bars indicate the SE of means.

Figure 3

Oxtr KO voles demonstrate less helping behavior (A) In this experiment, the door-opening behavior of the helper voles with knocked out oxytocin receptors (Oxtr KO) was compared with that of wild type (Oxtr +/+). (B) The mean latency of door-opening in the helper voles of wild type (n = 14; seven were male and another seven were female; all were paired with a same-sex soaked cagemate) and the Oxtr KO helper voles (n = 8; four were male and another four were female; all were paired with a same-sex cagemate). The significant difference was tested with a two-way ANOVA (the effect of group; F(1, 20) = 27.93, ∗∗p < 0.001). (C) The percentage of door-opening pairs in the wild-type and Oxtr KO voles. (D) The total number of door-opening sessions (Welch’s t-test; t(8.57) = 4.41, ∗∗p = 0.002) and the number of consecutive door-opening sessions (Welch’s t-test; t(12.35) = 4.74, ∗∗p < 0.001). (E) The percentage of time that the helper voles stayed on the soaked vole’s side (two-way ANOVA; the effect of group, F(1, 20) = 41.87, ∗∗p < 0.001). (F) The mean duration of huddling during the 2-min interaction period after opening the door (two-way ANOVA; the effect of group; F(1, 20) = 9.49, ∗∗p = 0.006). (G–I) The time spent in the ground area (G, Welch’s t-test; t(14.42) = 5.45, ∗∗p < 0.001), the number of entries into the pool area (H, Welch’s t-test; t(19.09) = 3.73, ∗∗p < 0.001), and the time spent in the pool area per entry (I) in the habituation period. The dashed line indicates the expected value. The error bars indicate the SE of means.