Initial compatibility during a "Speed-Dating" test predicts postpairing affiliation in titi monkeys (Plecturocebus cupreus)

Abstract

Behavioral compatibility plays a critical role in shaping how potential mates interact with and evaluate each other and whether they choose to pursue a relationship. Compatibility is especially important for mate choice and relationship quality in pair-bonding species that form long-term attachments between mates. Although this process has been studied in humans and birds, relatively few studies have investigated it in non-human primates. In this study, we investigated whether pairing titi monkeys (Plecturocebus cupreus) based on initial compatibility increased postpairing affiliation between mates. Subjects were 12 unpaired adult titi monkeys (two cohorts of three males and three females). We determined each subject's initial interest in each opposite-sex potential mate in their cohort across a series of six 30-min interaction periods (i.e., "speed-dates"). To determine initial compatibility, we used the Social Relations Model to calculate relationship effects in initial interest (how much each subject uniquely preferred each potential mate beyond their own affiliative disposition and their partner's popularity). We then paired monkeys in a way that maximized net relationship effects between pairs, and measured longitudinal pair affiliation (Proximity, Contact, Tail Twining, and Combined Affiliation) for 6 months postpairing using daily scan-sample observations and monthly home-cage video recordings. Multilevel models showed that, on average, the six speed-dating pairs exhibited higher levels of Tail Twining (determined from scan-sample observations; β = 0.31) compared to a group of 13 age-matched colony pairs that were determined quasi-randomly without quantifying compatibility. The degree of initial compatibility within speed-dating pairs also predicted higher levels of Combined Affiliation (determined from video recordings) at earlier post-pairing time points, with the association peaking at 2 months postpairing (β = 0.57). These findings suggest that initial compatibility facilitates pair bonding in titi monkeys. We conclude by discussing how the speed-dating design can be used for colony management to inform pair-housing decisions.

Keywords: Social Relations Model; attachment theory; comparative animal model; initial attraction; pair bonding.

Figure 1.. Schematic of Social Relations Model

The figure shows a conceptual schematic of the Social Relations Model (SRM) and how it was used in this study to determine titi monkey’s initial compatibility towards different potential mates. The light-colored monkeys (on the left side of the arrows in each panel) represent a group of same-sex subjects, and the dark-colored monkeys (on the right side of the arrows in each panel) represent a group of opposite-sex potential mates. Monkey A represents a given focal subject, and monkey B represents a given focal target (a potential mate encountered during speed-dating). Black arrows represent components that pertain to each SRM effect. Actor effects (the left panel) represent monkey A’s general disposition to show interest towards all potential mates (including monkey B and others). Partner effects (the middle panel) represent how much interest monkey B received from all potential mates (including monkey A and others). Relationship effects (the right panel) represent how much monkey A uniquely preferred monkey B (the black arrow) above and beyond actor effects and partner effects (the grey arrows in this panel). We calculated these three SRM effects for time spent in the preference zone during speed-dating (see Kenny, 2019b; Kenny & La Voie, 1984) and used relationship effects as a measure of initial compatibility. The calculations are available in Supplementary Table S1.

Figure 2.. Schematic of Speed-Dating Testing Chamber

The figure shows a three-dimensional schematic of one of the speed-dating testing chambers. The black bars represent perches. The white line on the perches closest to the grated windows indicate the boundary of the preference zone (see Supplementary Figure S4). The numbers correspond to the following labels: 1 = The grated window that could be exposed or covered by a sliding metal door (the metal door is not pictured); 2 = A GoPro camera; 3 = A food bowl; 4 = A sliding door for catching and releasing the subject from the chamber. Abbreviations: in = inches.

Figure 3.. Speed-Dating Testing Schedule and Post-Pairing Timeline

In the first part of the study (Panel A), we used a speed-dating paradigm to assess initial compatibility between potential pair mates. The testing timeline included two days of habituation trials and six days of testing (with three speed-dating trials per day of testing, one with each opposite-sex potential mate in their cohort). After 2–10 weeks, subjects were paired with one the opposite-sex potential mates from their cohort based on initial compatibility (the day of pairing is represented by the grey diamond in each panel). In the second part of the study (Panels B and C), we measured longitudinal affiliation between each pair. For the first measure (Panel B), we used scan-sample observations and measured how often pairs were in Proximity, Contact, or Tail Twining (we also assessed Combined Affiliation) and averaged the daily percentage of observations that pairs were observed in each state across months 1–6 months post-pairing. For the second measure (Panel C), we scored 9 monthly home-cage video recordings per pair for the same affiliation states assessed in the scan-sample observations and determined the percentage of time that pairs spent in each affiliative state. For these measures, “month 0” was the first three days of pairing (filming days #1–3), and months 1–6 were a day at the end of each monthly interval (filming days #4–9).

Figure 4.. Titi Monkey Pairs Determined from Speed-Dating Show Higher Scan-Sample Tail Twining than Age-Matched Comparison Pairs

The graph shows that Speed-Dating Pairs (blue circles) exhibited higher levels of Tail Twining (determined from the scan-sample observations) than Age-Matched Colony Pairs (brown diamonds) across the first six months post-pairing (on the x-axis). Each data point represents the average daily percentage of observations that a pair was observed in each affiliation state (averaged across each monthly post-pairing interval; on the y-axis). The text labels show the effect size for the difference between groups (positive beta values indicate that speed-dating pairs exhibited higher pair affiliation than the age-matched colony comparison group); the bolded text for Tail Twining indicates that the comparison was significant (p < 0.05). The fit lines correspond to the average change in pair affiliation across time (which was significant for Contact and Combined Affiliation, indicating that, in both groups, these affiliation states tended to increase across time; see Table 4). The shading around each line represents the standard error of prediction.

Figure 5.. Degree of Initial Compatibility Predicts Video-Scored Combined Affiliation During the Initial Months of Pairing

The graph shows that the degree of initial compatibility that speed-dating subjects had for the mate they were ultimately paired with (i.e., relationship effects in time in the preference zone during speed-dating; on the x-axis) positively predicted increased levels of total Combined Affiliation (i.e., the total percentage of time that pairs spent in Proximity, in Contact, or Tail Twining on a given day of filming; on the y-axis) at earlier post-pairing time points (with each color and panel indicating a different monthly post-pairing filming occasion). The bolded text for Month 2 indicates that degree of initial compatibility was significantly associated with Combined Affiliation at this time point. The original model also showed a significant effect of month on Combined Affiliation (see Table 5), indicating that affiliation increased across time (as can be seen by how the data points increase across each panel).