Structure and diversity of play behavior in juvenile Belding's ground squirrels

METHODS Video recording of behavior A total of 525 video recordings of play interactions of juvenile U. beldingi were made between July and August in 1998, 2004, 2016, 2017, 2019, 2024, and 2025 in a 25-hectare meadow at Tioga Pass (latitude 37.9, longitude 119.2, elevation 2950 meters) in Mono County, California, USA. Juveniles ordinarily emerge from the natal burrow in early to mid-July and engage in play during the two weeks following emergence from the natal burrow. Members of the field crew observed the study area throughout July to identify the natal burrows of litters as juveniles appeared above ground for the first time. After the natal burrow of a litter had been located, the field crew observed the litter for two weeks between 0730 and 1000 hours, the period of the day during which most play occurs among juveniles (Ryan et al., 2024). During observations, play interactions were recorded on video. In 2004, 2016, 2017, and 2019, juveniles were ear-tagged and dye-marked at their first emergence from the natal burrow as part of separate studies using methods described by Nunes et al. (1999). The field crew also observed adult U. beldingi in the study area throughout the day from 0730 to 1700 hours and video recorded five mating interactions between adults during the breeding period in 1998 and 31 aggressive interactions between adults in 2004, 2024, and 2025. Construction of ethograms During initial evaluation of videos to construct ethograms viewers watched videos at 0.25x speed and identified behaviors expressed repeatedly during juvenile play, adult mating, and aggressive interactions, Viewers compiled pictures and video clips of specific behaviors, and used the pictures and clips to prepare descriptions of behaviors in ethograms. The ethograms were compared to categorize specific juvenile play behaviors as sexual or aggressive based on their resemblance to adult behaviors. Juvenile play was frequently a modified version of adult behavior (Nunes et al. 1999). For example, play mounting involved contact between individuals similar to what is seen in adult mounting, but did not include intromission. Play fighting involved nuzzling targets on the body, such as the face, neck, shoulder, flank, or rump of the partner, rather than biting these targets. Some play behaviors were categorized as indeterminate. These behaviors 1) either showed similarities to behaviors present in both adult sexual aggressive interactions or 2) did not resemble any specific adult mating or aggressive behaviors. Evaluation of play interactions After ethograms were constructed, viewers watched videos of juvenile play at 1x speed to evaluate 1) the duration of play interactions, 2) the number of play bouts within interactions, and 3) the number of juveniles participating in interactions. We defined a play interaction as beginning when at least two squirrels came into proximity of one another, and acknowledged each other by facing off (standing on hind limbs and looking at each other), sniffing each other, or by one squirrel attacking another (see below). We defined a play interaction as ending when squirrels separated from each other by > 3 meters or ceased expressing play behaviors for longer than 10 seconds (Pasztor et al., 2001). We defined a play bout as a distinct occurrence of play behavior within a play interaction (Pasztor et al., 2001). Bouts were characterized by 1) two or more juveniles coming into contact with one another to initiate a specific play behavior, 2) the juveniles engaging in one or more play behaviors, and 3) the juveniles physically separating from each other to terminate a play behavior. The physical separation typically involved a squirrel breaking contact, moving a short distance away (> one body length) from its partner, and turning away. This separation was followed by the initiation of a new play bout or the termination of the play interaction. Play interactions were also classified based on the behaviors they included. Interactions were categorized as aggressive if they contained exclusively aggressive or aggressive and indeterminate behaviors, and as sexual if they contained exclusively sexual or sexual and indeterminate behaviors. Interactions were categorized as mixed if they contained both sexual and aggressive behaviors, with or without indeterminate behaviors. Interactions were classified as indeterminate if they contained only indeterminate behaviors. Diversity of behaviors within play interactions We assessed the diversity of behaviors expressed by juvenile U. beldingi in dyadic play interactions. For this analysis, we included only recordings of play interactions in which all behaviors expressed by juveniles could be clearly discerned throughout the recording (n = 267). Viewers evaluated recordings at 0.25x speed to ensure accuracy in identifying and evaluating behaviors. We used three different metrics to assess behavioral diversity within play: 1) the Shannon diversity index, 2) behavioral richness (# of behaviors expressed during interaction), and 3) the Pielou evenness index. These metrics have been adapted from community ecology to measure the diversity and structure of play interactions (Cordoni et al., 2022). The Shannon index for behavior can be calculated using Equation 1, where i as a specific behavior, ni as the number of times the behavior occurred during an interaction, and N as the total number of different behaviors observed in the interaction. Equation 1: Shannon Diversity Index: H′ = - Σ [(ni/N) * (ln ni/N)] In community ecology, the Pielou Index (J) is a measure of species evenness, evaluating the relative frequency of individuals in different species to assess how individuals are distributed across species. In this study, the Pielou index was used as a measure of how evenly behaviors were represented within play interactions. The Pielou’s Index can be calculated for behavior using Equation 2, where H′ is the observed value of the Shannon Index, and H′max is equal to ln (S), with S representing the total number of different behaviors observed in an interaction. Values in the Pielou Index range from zero to one, with values closer to one indicating that behaviors are more equally represented within an interaction (Cordoni et al., 2022). Equation 2: Pielou Index: J = H′/H′max Reciprocity within play interactions We evaluated the degree of reciprocity among juveniles in dyadic play interactions based on the extent to which partners took turns initiating new play bouts within interactions. For this analysis, we included only play interactions with more than one play bout and with both partners in view on the video recording during the entire interaction (n = 166). Interactions with only one play bout did not provide opportunities for juveniles to take turns initiating new bouts. Reciprocity was evaluated using a reciprocity index R (Equation 3) adapted from (Bauer and Smuts 2007; Ward et al. 2008; Stark et al. 2021; Ham et al. 2024). In the equation, A is the total number of play bouts initiated by juvenile A, and B is the number initiated by juvenile B. Juvenile A is the partner who initiated the first play bout during the interaction. Values of R in the reciprocity index range from zero to one, with zero representing a completely asymmetrical interaction (all play bouts initiated by the same partner) and one representing a perfectly symmetrical interaction and a high degree of reciprocity, with each partner initiating the same number of bouts. Equation 3: Reciprocity Index: R=1-|A - B|/(A+B) REFERENCES Bauer EB, Smuts BB. 2007. Cooperation and competition during dyadic play in domestic dogs, Canis familiaris. Anim Behav. 73:489–499. https://doi.org/10.1016/j.anbehav.2006.09.006. Cordoni G, Collarini E, Gioia M, Norscia I. 2022. Play fighting versus real fighting in piglets (Sus scrofa): Similar patterns, different structure. Behav Process. 203:104778. https://doi.org/10.1016/j.beproc.2022.104778. Ham JR, Pellis SM. 2024. Play partner preferences among groups of unfamiliar juvenile male rats. Sci Rep. 14:16056. https://doi.org/10.1038/s41598-024-66988-w. Nunes S, Muecke E-M, Anthony JA, Batterbee AS. 1999. Endocrine and energetic mediation of play behavior in free-living Belding’s ground squirrels. Horm Behav. 36:153–165. https://doi.org/10.1006/hbeh.1999.1538. Pasztor TJ, Smith LK, MacDonald NK, Michener GR, Pellis SM. 2001. Sexual and aggressive play fighting of sibling Richardson’s ground squirrels. Aggress Behav. 27:323–337. https://doi.org/10.1002/ab.1015. Ryan WJ, Kuan MB, Nunes S. 2024. Disruption of social play influences development of caution in juvenile ground squirrels. Front Ethol. 3. https://doi.org/10.3389/fetho.2024.1410334. Stark RAM, Ramkumar R, Pellis SM. 2021. Deficient play-derived experiences in juvenile Long Evans rats reared with a Fischer 344 partner: A deficiency shared by both sexes. Int J Comp Psychol. 34:5592. https://doi.org/10.46867/ijcp.2021.34.5592. Ward C, Bauer EB, Smuts BB. 2008. Partner preferences and asymmetries in social play among domestic dog, Canis lupus familiaris, littermates. Anim Behav. 76:1187–1199. https://doi.org/10.1016/j.anbehav.2008.06.004.