Social conditioned place preference in the captive ground squirrel (Ictidomys tridecemlineatus): Social reward as a natural phenotype

Abstract

Social behaviors of wild animals are often considered within an ultimate framework of adaptive benefits versus survival risks. By contrast, studies of laboratory animals more typically focus on affective aspects of behavioral decisions, whether a rodent derives a rewarding experience from social encounter, and how this experience might be initiated and maintained by neural circuits. Artificial selection and inbreeding have rendered laboratory animals more affiliative and less aggressive than their wild conspecifics, leaving open the possibility that social reward is an artifact of domestication. We compared social behaviors of wild and captive population of juvenile 13-lined ground squirrels (Ictidomys tridecemlineatus), the latter being 2nd- and 3rd-generation descendants of wild individuals. At an age corresponding to emergence from the burrow, postnatal day (PD) 38, captive squirrels engaged in vigorous social approach and play and these juvenile behaviors declined significantly by PD 56. Similarly, young wild squirrels expressed social proximity and play; affiliative interactions declined with summer's progression and were replaced by agonistic chasing behaviors. Social conditioned place preference testing (conditioned PDs 40-50) indicated that adolescent squirrels derived a rewarding experience from social reunion. Our results support the contention that undomesticated rodents have the capacity for social reward and more generally suggest the possibility that positive affective experiences may support group cohesion, social cooperation, and altruism in the wild.

Figure 1. Developmental timeline of experimental procedures and illustrated representation of social conditioned place preference

The top of the figure depicts the ages when post-weanling squirrels were tested for SI and conditioned and tested for social CPP, respectively. Grouped squirrels on the left side, in the bedding represented by gray dots, are undergoing social conditioning in a ‘corncob’ environment with tubes threaded within the interior. The isolated squirrel on the right side, in the bedding represented by diagonal lines, is undergoing conditioning alone in a ‘paper’ environment with tubes containing a smooth interior. The social CPP testing arena is illustrated at the bottom.

Figure 2. Behavioral differences between early and late adolescent ground squirrels in the laboratory

Labels on the abscissa indicate the age at testing and the sex of the interacting squirrels (test-stimulus combinations). (A) Play behavior of test squirrels diminished across adolescent development from PD38 to PD56 and (A') was most highly expressed in male-male pairs. (B) Social investigation was greater at PD38 versus PD56 and (B') was expressed more within female-female pairs. (C) Vigilance behavior was higher on PD56 versus PD38 and (C') did not differ between the sex-pairings of squirrels. (D) ‘Time budgets’ for squirrels during the 5-min test period demonstrating that a majority of time was spent engaged in behaviors that were not social in nature. N's = 4–7 squirrels per sex combination. All data are presented as the mean ± standard error. (*P<0.05 **P<0.001).

Figure 3. Social conditioned place preference in adolescent ground squirrels

Preference scores were generated by subtracting the amount of time each squirrel spent in the ‘paper’ environment from the amount of time in the ‘corncob’ environment. A positive preference score therefore indicates an individual spent more time in the corncob environment during the 30-min trial whereas a negative score indicates more time spent in the paper environment. N's = 6–18 squirrels per group. All data are presented as the mean ± standard error. (*P<0.05 **P<0.001).

Figure 4. Social behaviors of adolescent ground squirrels in the field

Expected weekly counts of social behaviors were generated assuming an equal distribution of the total social behaviors (n=228) over the 5-week observation period from late June and early August. (A) The number of social behaviors observed each week deviated from the expected values and generally decreased over the 5-week period. Social behaviors were categorized as approach, play, investigation, maintaining proximity or chasing (see Table 2), and total counts were tallied for each week. Observed behaviors in the field varied significantly from the expected values. (B) Approach, (C) play, (D) investigation and (E) proximity were predominantly observed during weeks 1–3 while (F) agonistic chasing became more frequent in weeks 3–5. Total minutes of observation: week 1 = 291 min, week 2 = 357 min, week 3 = 551 min, week 4 = 298 min, week 5 = 321 min. All data are presented as frequencies of occurrence.