| Sound signals are often used by animals inhabiting in densely wooded environments because they are less affected by obstacles and can go over long distances.When individuals detect danger,they can send out vocal alarm signals,causing other individuals or social groups to raise vigilance and take preventive responses,which is an important survival strategy for animals to adapt to the environment.The golden snub-nosed monkey(Rhinopithecus roxellana)has a complex social system,they form obvious linear hierarchies among individuals or family units.With limited resources,severe competition for food were expected in this species.Previous research observations have shown that vocal communication is one of the crucial transit routes for information exchanges in golden snub-nosed monkeys.However,previous studies on this species only illustrate their vocal communication behavior in general,the types and complex functions of their vocal alarm signals and whether it is related to food competition strategies remains unknown.In this study,wild golden snub-nosed monkeys in the Guanyin Mountain Nature Reserve in the Qinling Mountains were selected as the study population.And a hexagonal camera array was set up at selected feeding grounds from October 2020 to July 2021 to collect data on the alarm behavior emitted by golden snub-nosed monkeys during feeding to investigate the functional properties of the alarm calls made by golden snub-nosed monkeys during food competition.Through comprehensive observations and systematic analyses,we attempt to elucidate the following questions:(1)What are the functions of alarm calls made by golden snub-nosed monkeys during competition for food resources?(2)Do individuals in the social groups use deceptive alarm signals as a strategy to increase access to food resources? These questions would promote further investigation on how individuals use deceptive sound signals in the competition for resources in this social primate,and would lead us towards a better understanding in the evolution of survival strategies of golden snub-nosed monkeys under their complex social systems.The results of the study show that:(1)The predator model was set up to reveal that golden snub-nosed monkeys have evolved a mixed alarm system.There is a functional reference system of alarm calls coding for different types of danger in the face of terrestrial and aerial predators.However,the alarm calls for different types of predators on land(leopards,snakes)do not encode information,which indicates an urgency reference system.This study suggests this alarm system may be related to the simplicity of the habitat in which golden snub-nosed monkeys live,the different predation pressures of aerial and terrestrial predators,and their body sizes compared to their predators.The behavioral responses to aerial predators are also consistent with the pursuit deterrence hypothesis of alarm calls.(2)The alarm call that this study focuses on is RRDA(resource-related deceptive alarm).A Poisson regression model to explore the factors influencing the frequency of RRDA in golden snub-nosed monkeys at each feeding session.The results showed that season,daily feeding sessions(morning or afternoon),visibility,average unit feed time,feeding unit,age-sex class,unit rank and unit size all significantly affected the number of alerts per session.The lack of food in winter,longer waits before feeding in the morning,and more energy consumption during low visibility(i.e.,rain and snow)will make monkeys hungrier and increase competition for food.The number of feeding units and the average time taken by a unit is also a reflection of the intensity of food competition.These factors are related to the levels of food competition intensity,and almost every RRDA could trigger an anti-predator response to some extent.As individuals flee around,the original pattern of foraging site occupation is disrupted,r resulting in a redistribution of food resources for all individuals in the feeding ground.(3)We build a GLM model to explore the factors influencing the success of each RRDA.The results showed that the time interval between the current RRDA and sex-age class significantly affect the success of each RRDA.RRDA is more likely to succeed when the signaler is subadult male and the time interval between two RRDAs is longer when the alarm is issued.(4)We build a GLM model to explore the factors influencing RRDA response rate.The model results showed that the orientation of the individual giving the alarm,the feeding session,the total time spent feeding,the video time when alarm was given and age-sex class significantly influenced the response rate per RRDA.RRDAs were more likely to be responded to when the alarming individual’s back was turned towards other individuals,the closer to the point at which feeding began,and when the alarming individual was an adult female.The success rate and response rate were found to be such that golden snub-nosed monkeys were unable to judge the reliability of alarms by context,and therefore the existence of an anti-deception strategy could not be demonstrated.(5)We explored the effect of pre-entry waiting time on RRDA through a weighted least square regression.Only the season significantly affects the frequency of RRDAs prior to unit entry.The incidence of RRDA is significantly higher in winter than in other seasons.Analysis suggests that the majority of those waiting to enter outside the feeding grounds are lower ranked units.The shortage of food in winter combined with the more pronounced social rank divisions in this season leading to greater intra-group food competition,resulting in more RRDAs in winter.(6)Analysis of the relationship between alarming individuals and responding individuals found that the response rate of individuals within the breeding unit was higher than that of individuals outside the unit.Since the kinship coefficient within the golden snub-nosed monkey was higher than that between the units,our results were consistent with the kin-selection hypothesis.This study suggests that the alarm calls made by golden snub-nosed monkeys in competition for food resources in the Qinling Mountains are consistent with a functional deception explanation,as well as with the kin selection hypothesis and the pursuit deterrence hypothesis.The results reveal the specific role of individual alarm calls in competition for food resources,and provide an informative model for exploring the function of acoustic signals for other wildlife. |
