These results support earlier findings that social behavior and proximity among female white-tailed deer are affected by timing of parturition, whereby social behavior and interactions decrease starting at parturition [9, 10, 11]. Despite similar findings with other studies that used observational or static spatial overlap indices, our results provide more quantitative metrics to how similar movement is between pairs of female white-tailed deer. Since most early studies used static spatial overlap metrics to assess changes in social behavior and herd dynamics, we examined the correlation between home range overlap and measures of dynamic interaction. The positive correlation between percentage of home range overlap and global DI supports the hypothesis that static spatial overlap is an approximate method to track changes in spatial behavior among individuals. However, static measures of interaction are not reliable indices of relatedness, social group membership [2], actual contact, or cohesive movement [26]. Our study also had the added benefit of knowing the exact timing of parturition given the use of rVITs; most early studies relied on visual observations or predicted parturition from estimated conception dates [11, 44].
The consistent and gradual decline in mean local di following parturition was unexpected given past literature indicating that social interactions among female herd members virtually cease to exist until the fawn becomes more mobile [9, 21]. We expected population mean local di values to reflect a more abrupt disruption of female social dynamics after parturition to reduce predation risk effects, which is partially mitigated by solitude seeking and hiding of the newborn in smaller areas with fewer deer [10, 45]. Another consideration for the gradual decline in social interactions may be driven by methodological reasons, at least partially. We centered local di values for each interaction pairing on the hour for the first of 2 parturition events, so the more gradual decline in mean local di may reflect a short period where socially agonistic behaviors are only present in 1 member of each pairing because the other individual has yet to give birth. Despite agonistic behaviors from the parturient female, other collared females may exhibit spatiotemporal use of similar ranges; therefore, we still would document a potential interaction. It also is likely that individuals may interact during foraging bouts while the newborn remains hidden. Another study corroborates this hypothesis in that movement distance of female elk (Cervus canadensis) showed a similar pattern of decline after parturition, indicating that movement of the dam is still persistent albeit at a lower rate [44], so we expect female-female interactions to decline in relation to reduced movement.
Alternatively, the prolonged decline in mean local di values may accurately reflect a reduction in social tolerance and group cohesion which may be less immediate than we had hypothesized. If the energetic demands of lactation are driving the disruption of social dynamics, and those energetic demands accrue constantly throughout the summer, a steadier decline in social tolerance would be expected. Furthermore, declining social tolerance at the population level may be continuously offset as neonates die and dams without fawns return to prepartum levels of sociality [10]. While our sample size of individual females was limited, all dams in our study had at least 1 neonate survive beyond 1 week and 7 of 10 had at least 1 neonate survive to 2 weeks, suggesting the movement behavior of dams following the loss of their neonates was not a major factor in the prolonged decline of mean local di we observed.
Previous research reported the reduced sociability in captive females during lactation persisted for up to 12-weeks postpartum [10]. Ozoga et al. [9] similarly reported reduced levels of social group affiliation for up to 3 months postpartum in a captive population, but claimed territorial behavior ceased after 4 weeks. While our monitoring period only extended ~ 25 days postpartum, we observed a noticeable inflection point at ~ 14.8 days, where values of population mean local di began to increase following a period of steady decline. Although behaviors associated with the hider strategy among neonates may persist for up to 60 days, a transition to the neonate following the dam is not a discrete behavioral event [46]. Neonates become more agile with age and will begin to flee from predators at approximately 2 weeks old [47, 48, 49], and 2 weeks has also been proposed as the time needed to establish the dam-neonate bond [21]. Although the energetic demand of lactation persists well into the summer, the increasing mobility and agility of the neonates progressively reduces the risk of predation when traveling with the dam. Such a transition may allow a shift in the behaviors of the dam, whereas the benefit of shared group vigilance may begin to gradually outweigh the benefit of social isolation and predator avoidance. Similarly, Nixon et al. [50] reported a general increase in tolerance of other deer among dams 2–3 weeks postpartum in Illinois. Nixon et al. [50] also noted the reformation of female social groups progressed as a function of relatedness, such that dams began to tolerate social interactions with closely related females more readily than distantly or unrelated individuals following the 2–3-week postpartum period. Our observation of increasing mean local di values between 15–25 days postpartum may reflect a period of social reintegration among females, which are likely related [2, 3].
Additionally, agonistic social cues from adult females have been proposed as a proximate motivating factor for spring dispersal among juveniles of both sexes [19]. The ~ 15-day period of decreasing local di values observed corresponds directly to the peak of spring dispersal movements among juveniles in both Sussex County [51] and the broader region [52]. A relatively shorter period of heightened social aggression among related females may reduce the probability a juvenile female will disperse during the spring, which would in turn facilitate the maintenance of matrilineal groups.
A necessary caveat to our research is the limited geographic scope and sample size of our population. Because there is often a disparity in the social and spatial behaviors demonstrated by captive white-tailed deer [45, 53, 54], our objective required a free-ranging population. A sufficient sample size of interaction pairings of parturient females with precisely known parturition times would be logistically difficult over a large geographic area, and these spatial limitations were necessary given our research objective. Additionally, white-tailed deer demonstrate a great degree of behavioral plasticity across their range. Social isolation and agnostic behaviors are likely more prominent in populations with relatively concise conception and parturition periods. When parturition occurs over a relatively long period, as is the case in regions of the southeastern United States [45], periods of social agonism and aggression may be less pronounced. Our results, and the results of future research into deer social dynamics, should be interpreted within the context of regional variability in habitat and population demographics.