Jones CB (June, 1996) Temporal Division-of-Labor In A Primate: Age-Dependent Foraging Behavior. Neotropical Primates 4(2): 50-53. [with minor edits from the original]
Introduction
Division of labor based on age or size ["polyethism"] may reflect the reproductive condition of individuals in social groups. In 1967, West proposed the general hypothesis that hierarchical relations may be advantageous to both dominants and subordinates and that individuals of low rank may be inferior reproductives who benefit genetically from associations with and contributions to reproductively superior individuals. Since increasing age or size eventually entails decreasing reproductive value (Vx), several authors have noted that the display of social behavior, such as foraging behavior that benefits all members of a group, especially kin, should increase with age as the benefits from individual (selfish) reproduction decline (e.g., West-Eberhard 1975; Hrdy & Hrdy 1996). As individual reproductive value decreases, benefits (genetic or other) from assisting the reproduction of conspecifics (social behavior [cooperation, altruism as per W.D. Hamilton 1964]) may increase because costs (genetic or other [including, delayed costs or benefits] of social behavior) decrease with decreased benefits from individual reproduction. In order to test this hypothesis, I studied the relationship between adult female age, dominance rank, reproductive value, and social foraging behavior (food search and pursuit) for adult female mantled howler monkeys (Alouatta palliata Gray). [Author's note, 1/20/2020: "Temporal division-of-labor" (TDL) may, also, be termed, "age-polyethism" or "primitive" (totipotent) eusociality or Totipotent Eusociality (TE); see blogpost on General Mammalian Patterns, #28]
Subjects and Methods
During an extended period of study at Hacienda La Pacifica, Canas, Guanacaste, Costa Rica, I studied two marked, aged groups of mantled howler monkeys in two tropical dry forest habitats [Riparian, Group 5, and Deciduous, Group 12: see Jones 1980, Table 1]. For this species [and others of the genus], age and dominance rank are negatively correlated [Jones 1978, Jones 1980].
Social foraging was operationally defined as the behavioral series: feed-rest-move [at least 100 m]-feed, by a unit of more than three adults. These criteria were adopted in order to standardize measurement and to eliminate periods of food search within unusually large patches and by consort pairs. I identified which females in the primary study groups initiated foraging sequences and analyzed these observations by age.
My null hypothesis held that the frequency of foraging by females of any age class would be proportional to the total number of females who foraged in an age class. Two of the 15 [adult] females in one group [both young adults--Riparian Habitat Group 5] were never observed to direct foraging sequences and are excluded from analysis. Three [adult] females were aged on the basis of physical and behavioral traits other than tooth wear, and assignment to age classes for these females was made independent of the present analysis. Two of these females were observed from sub-adult through adult growth and classified as young adults; a third [adult] female, classified as middle-aged, was the mother of a sub-adult and a juvenile offspring, a highly unlikely combination for any other age class [see Glander 1980]. In my analysis of the second group [eight adult females--Deciduous Habitat Group 12], two young adult immigrant females were never observed to forage socially and were excluded from analysis. The pattern of results reported here would remain unaffected by alternative treatments of the raw data.
A monthly foraging rate for each forager was computed by dividing the frequency of foraging by the female's number of months resident in a group, a period of time varying from 10-14 months since some females emigrated during the study. These rates were compared with a female's age class, on the one hand, and dominance rank, on the other, to assess the relationship between the display of social foraging behavior and rank, and reproductive value [Vx, for her age class, (computed from) population data in Malmgren 1979, Table 23; equation after Wilson & Bossert 1971; c.f. Jones 1997] where relative contribution to future generations of an individual of a given age is quantified.
Results and Discussion
Table 1 [below--scroll down] presents the results of my analysis for the first group [Group 5] of foraging frequency as a function of female age, including, expected frequencies, and Chi Square. Computing "goodness of fit" led to an unequivocal rejection of the null hypothesis [P <- 0.001, X2= 107.64, df= 3]. Thus, old age and foraging frequency are significantly related. Young adult females initiate foraging significantly less than expected on the basis of their numbers [in Group 5: P <- 0.001], suggesting that such individuals are relatively "selfish" or are conserving time [T] and energy [E], possibly for reproduction and/or competition. Table 1 also shows that the middle-aged to old female foraged more than expected by chance [P <- 0.01], and this female succeeded the oldest and lowest-ranking female as the most frequent [social] forager when the old female emigrated in 1977 [personal observation*].
Additional observations support the reliability of the above patterns. The oldest female in the second group [Group 12] foraged more frequently than any other adult female [P <- 0.001, X2= 17.29, df= 2; c.f. Jones 1998]. Similarly, the relationship between foraging rate and age class [Fig. 1] yields a significant positive correlation [rs= +0.629, P <- 0.05]. Related to this, the correlation between foraging rate and dominance rank [Fig. 2] is significant but negative [i.e., the higher the foraging rate, the lower the dominance rank, rs= -0.63, P <- 0.05]. Thus, the initiation of [social] foraging is significantly associated with female age and dominance rank.
It was hypothesized above that the expression of social behavior would increase with increasing age since reproductive value [Vx, Fig. 3] decreases with age and with it the benefits from selfish reproduction [i.e., adult females have less to lose and more to gain in fitness as they age]. Figure 3 shows the reproductive value curve for the population of mantled howler monkeys at Hacienda La Pacifica [after Jones 1997]. Comparing Fig. 3 with Figs. 1 and 2, consistent with expectation, a strong negative association appears to exist between reproductive value and rate of foraging. Reproductive value in the four adult age classes is negatively, and significantly correlated with social foraging rate/month [rs= -0.95, P <- 0.02]. These results support the view that increasing age or size eventually entails decreasing reproductive value and that the display of social behavior should increase with age as the benefits from individual [selfish] reproduction decline.
What features of the howlers' environment might favor temporal division-of-labor? On 52 occasions, I was able to record the specific resource upon which foraging sequences terminated. Forty-four [85%] of these sequences terminated on ephemeral food [i.e., fruit, flowers, or new leaves: see Jones 1996], while eight [15%] sequences terminated with feeding on mature leaves [P <- 0.001, X2= 49, df= 1]. Thus, the initiation of social foraging sequences appears to be associated with food, the local distribution of which is temporally uncertain; new leaves, flowers, and fruit. The old female initiated 21 of the 52 [foraging] bouts, 20 of these for ephemeral food [c.f. Jones 1998].
An old female's presumed experience with the mosaic of her home range might enhance her efficiency as a forager so that her foraging activity may yield an energetic and nutritional gain to other group members. Temporal uncertainty of preferred food resources [see Jones 1996] may favor individuals that are the beneficiaries of the foraging activity of others, particularly, kin, when reproductive value is low. Division-of-labor through differential social roles may be a function of relative reproductive value, and behavioral roles may be understood within the context of life history patterns. [------>across Social Mammals** & other Social Vertebrates**? across Social Animals**?].
Acknowledgments
I appreciate the comments of R.C. Lewontin, E.O. Wilson, M.J. West-Eberhard, I.S. Bernstein, W.C. Dilger, and K.E. Weber on an earlier draft of this note. I thank the W. Hagnauer family for permission to work on their property, Hacienda La Pacifica, and for logistic assistance. My gratitude to Norman J. Scott, Jr. [USFWS], for expert introductions to the conduct of fieldwork and for imparting a variety of skills, is immeasurable. The work was supported by grants from the National Fellowships Fund and the National Research Council.
Clara B. Jones, Institute of Animal Behavior, Rutgers University-Newark, 101 Warren Street, Newark, New Jersey 07102, U.S.A.
References
Glander KE [1980] Reproduction and population growth in free-ranging mantled howling monkeys. Am J Phys Anthropol 53: 25-36.
Hamilton WD [1964] The genetical theory of social behavior. J Theor Biol 7: 1-52.
Hrdy SB, Hrdy DB [1976] Hierarchical relations mong female hanuman langurs (Primates: Colobinae, Presbytis entellus]. Science 197: 913-915.
Jones CB [1978] Aspects of reproduction in the mantled howler monkey, Alouatta palliata Gray. Ph.D. Dissertation, Cornell University, Ithaca, NY.
Jones CB [1980] The functions of status in the mantled howler monkey, Alouatta palliata Gray: intraspecific competition for group membershi in a folivorous Neotropical primate. Primates 21: 389-405.
Jones CB [December, 1996] Predictability of plant food resources for mantled howler monkeys at Hacienda La Pacifica, Costa Rica: Glander's dissertation revisited. Neotropical Primates 4(4): 147-149.
Jones CB [1997] Life history patterns of howler monkeys in a time-varying environment. Bol. Primatol. Lat. 6(1): 1-8.
Jones CB [March-Dec, 1998] A broad-band contact call by female mantled howler monkeys: implications for heterogeneous conditions. Neotropical Primates 6(2): 38-40.
Malmgren LA [1979] Empirical population genetics of golden mantled howling monkeys (Alouatta palliata) in relation to population structure, social dynamics, and evolution. Ph.D. Dissertation, University of Connecticut, Storrs.
West MJ [1967] Foundress associations in polistine wasps: dominance hierarchies and the evolution of social behavior. Science 157: 1584-1585.
West-Eberhard MJ [1975] The evolution of social behavior by kin selection. Quart Rev Biol 50: 1-33.
Wilson EO, Bossert WH [1971] A primer of population biology. Sinauer Assoc., Stanford, CN.
* The last time I saw Group 5's D4 female, she was seated alone and immobile on a tree limb; her face impaled with quills of the prehensile-tailed porcupine [or, coendou: Coendou]. I never encountered this female again.
**EO Wilson's 2019 book, Genesis, advances the idea that many Mammals, including, humans, may be "eusocial." Where "temporal division-of-labor" is demonstrated, taxa can be classified, "primitively eusocial;" this classification would apply, also, to any other Vertebrates or, indeed, to any other Animals, where "age polyethism" is identified. If "tradeoffs" [e.g., energetic, reproductive, survivval] are most likely to be observed in "poor" conditions [e.g., heterogeneous regimes where "fitness" is compromised; recurrent drought, unpredictable food or water supply], "age polyethism" may evolve to minimize energetic costs in time and space. While females are expected to be most sensitive to energetic effects, males, also, may benefit, under some conditions, from age-dependent responses. Furthermore, there may be energetic [reproductive] benefits in coordinating many maturational [age-dependent] and developmental [age-dependent] milestones or markers with one another as genetic and physiological energy-savings tactics and strategies.
-----------------------------------------------------------------------------------------------
Table 1. Age class, estimated age in years, number of females in each age class (N), observed (O), and expected (E) frequencies of social foraging, and chi square (X2) for a test of the null hypothesis. In “Age Class” column, YA= Young Adult; M-a= Middle-aged Adult; M-a-O= Middle-age to Old Adult; O= Old Adult.
Introduction
Division of labor based on age or size ["polyethism"] may reflect the reproductive condition of individuals in social groups. In 1967, West proposed the general hypothesis that hierarchical relations may be advantageous to both dominants and subordinates and that individuals of low rank may be inferior reproductives who benefit genetically from associations with and contributions to reproductively superior individuals. Since increasing age or size eventually entails decreasing reproductive value (Vx), several authors have noted that the display of social behavior, such as foraging behavior that benefits all members of a group, especially kin, should increase with age as the benefits from individual (selfish) reproduction decline (e.g., West-Eberhard 1975; Hrdy & Hrdy 1996). As individual reproductive value decreases, benefits (genetic or other) from assisting the reproduction of conspecifics (social behavior [cooperation, altruism as per W.D. Hamilton 1964]) may increase because costs (genetic or other [including, delayed costs or benefits] of social behavior) decrease with decreased benefits from individual reproduction. In order to test this hypothesis, I studied the relationship between adult female age, dominance rank, reproductive value, and social foraging behavior (food search and pursuit) for adult female mantled howler monkeys (Alouatta palliata Gray). [Author's note, 1/20/2020: "Temporal division-of-labor" (TDL) may, also, be termed, "age-polyethism" or "primitive" (totipotent) eusociality or Totipotent Eusociality (TE); see blogpost on General Mammalian Patterns, #28]
Subjects and Methods
During an extended period of study at Hacienda La Pacifica, Canas, Guanacaste, Costa Rica, I studied two marked, aged groups of mantled howler monkeys in two tropical dry forest habitats [Riparian, Group 5, and Deciduous, Group 12: see Jones 1980, Table 1]. For this species [and others of the genus], age and dominance rank are negatively correlated [Jones 1978, Jones 1980].
Social foraging was operationally defined as the behavioral series: feed-rest-move [at least 100 m]-feed, by a unit of more than three adults. These criteria were adopted in order to standardize measurement and to eliminate periods of food search within unusually large patches and by consort pairs. I identified which females in the primary study groups initiated foraging sequences and analyzed these observations by age.
My null hypothesis held that the frequency of foraging by females of any age class would be proportional to the total number of females who foraged in an age class. Two of the 15 [adult] females in one group [both young adults--Riparian Habitat Group 5] were never observed to direct foraging sequences and are excluded from analysis. Three [adult] females were aged on the basis of physical and behavioral traits other than tooth wear, and assignment to age classes for these females was made independent of the present analysis. Two of these females were observed from sub-adult through adult growth and classified as young adults; a third [adult] female, classified as middle-aged, was the mother of a sub-adult and a juvenile offspring, a highly unlikely combination for any other age class [see Glander 1980]. In my analysis of the second group [eight adult females--Deciduous Habitat Group 12], two young adult immigrant females were never observed to forage socially and were excluded from analysis. The pattern of results reported here would remain unaffected by alternative treatments of the raw data.
A monthly foraging rate for each forager was computed by dividing the frequency of foraging by the female's number of months resident in a group, a period of time varying from 10-14 months since some females emigrated during the study. These rates were compared with a female's age class, on the one hand, and dominance rank, on the other, to assess the relationship between the display of social foraging behavior and rank, and reproductive value [Vx, for her age class, (computed from) population data in Malmgren 1979, Table 23; equation after Wilson & Bossert 1971; c.f. Jones 1997] where relative contribution to future generations of an individual of a given age is quantified.
Results and Discussion
Table 1 [below--scroll down] presents the results of my analysis for the first group [Group 5] of foraging frequency as a function of female age, including, expected frequencies, and Chi Square. Computing "goodness of fit" led to an unequivocal rejection of the null hypothesis [P <- 0.001, X2= 107.64, df= 3]. Thus, old age and foraging frequency are significantly related. Young adult females initiate foraging significantly less than expected on the basis of their numbers [in Group 5: P <- 0.001], suggesting that such individuals are relatively "selfish" or are conserving time [T] and energy [E], possibly for reproduction and/or competition. Table 1 also shows that the middle-aged to old female foraged more than expected by chance [P <- 0.01], and this female succeeded the oldest and lowest-ranking female as the most frequent [social] forager when the old female emigrated in 1977 [personal observation*].
Additional observations support the reliability of the above patterns. The oldest female in the second group [Group 12] foraged more frequently than any other adult female [P <- 0.001, X2= 17.29, df= 2; c.f. Jones 1998]. Similarly, the relationship between foraging rate and age class [Fig. 1] yields a significant positive correlation [rs= +0.629, P <- 0.05]. Related to this, the correlation between foraging rate and dominance rank [Fig. 2] is significant but negative [i.e., the higher the foraging rate, the lower the dominance rank, rs= -0.63, P <- 0.05]. Thus, the initiation of [social] foraging is significantly associated with female age and dominance rank.
It was hypothesized above that the expression of social behavior would increase with increasing age since reproductive value [Vx, Fig. 3] decreases with age and with it the benefits from selfish reproduction [i.e., adult females have less to lose and more to gain in fitness as they age]. Figure 3 shows the reproductive value curve for the population of mantled howler monkeys at Hacienda La Pacifica [after Jones 1997]. Comparing Fig. 3 with Figs. 1 and 2, consistent with expectation, a strong negative association appears to exist between reproductive value and rate of foraging. Reproductive value in the four adult age classes is negatively, and significantly correlated with social foraging rate/month [rs= -0.95, P <- 0.02]. These results support the view that increasing age or size eventually entails decreasing reproductive value and that the display of social behavior should increase with age as the benefits from individual [selfish] reproduction decline.
What features of the howlers' environment might favor temporal division-of-labor? On 52 occasions, I was able to record the specific resource upon which foraging sequences terminated. Forty-four [85%] of these sequences terminated on ephemeral food [i.e., fruit, flowers, or new leaves: see Jones 1996], while eight [15%] sequences terminated with feeding on mature leaves [P <- 0.001, X2= 49, df= 1]. Thus, the initiation of social foraging sequences appears to be associated with food, the local distribution of which is temporally uncertain; new leaves, flowers, and fruit. The old female initiated 21 of the 52 [foraging] bouts, 20 of these for ephemeral food [c.f. Jones 1998].
An old female's presumed experience with the mosaic of her home range might enhance her efficiency as a forager so that her foraging activity may yield an energetic and nutritional gain to other group members. Temporal uncertainty of preferred food resources [see Jones 1996] may favor individuals that are the beneficiaries of the foraging activity of others, particularly, kin, when reproductive value is low. Division-of-labor through differential social roles may be a function of relative reproductive value, and behavioral roles may be understood within the context of life history patterns. [------>across Social Mammals** & other Social Vertebrates**? across Social Animals**?].
Acknowledgments
I appreciate the comments of R.C. Lewontin, E.O. Wilson, M.J. West-Eberhard, I.S. Bernstein, W.C. Dilger, and K.E. Weber on an earlier draft of this note. I thank the W. Hagnauer family for permission to work on their property, Hacienda La Pacifica, and for logistic assistance. My gratitude to Norman J. Scott, Jr. [USFWS], for expert introductions to the conduct of fieldwork and for imparting a variety of skills, is immeasurable. The work was supported by grants from the National Fellowships Fund and the National Research Council.
Clara B. Jones, Institute of Animal Behavior, Rutgers University-Newark, 101 Warren Street, Newark, New Jersey 07102, U.S.A.
References
Glander KE [1980] Reproduction and population growth in free-ranging mantled howling monkeys. Am J Phys Anthropol 53: 25-36.
Hamilton WD [1964] The genetical theory of social behavior. J Theor Biol 7: 1-52.
Hrdy SB, Hrdy DB [1976] Hierarchical relations mong female hanuman langurs (Primates: Colobinae, Presbytis entellus]. Science 197: 913-915.
Jones CB [1978] Aspects of reproduction in the mantled howler monkey, Alouatta palliata Gray. Ph.D. Dissertation, Cornell University, Ithaca, NY.
Jones CB [1980] The functions of status in the mantled howler monkey, Alouatta palliata Gray: intraspecific competition for group membershi in a folivorous Neotropical primate. Primates 21: 389-405.
Jones CB [December, 1996] Predictability of plant food resources for mantled howler monkeys at Hacienda La Pacifica, Costa Rica: Glander's dissertation revisited. Neotropical Primates 4(4): 147-149.
Jones CB [1997] Life history patterns of howler monkeys in a time-varying environment. Bol. Primatol. Lat. 6(1): 1-8.
Jones CB [March-Dec, 1998] A broad-band contact call by female mantled howler monkeys: implications for heterogeneous conditions. Neotropical Primates 6(2): 38-40.
Malmgren LA [1979] Empirical population genetics of golden mantled howling monkeys (Alouatta palliata) in relation to population structure, social dynamics, and evolution. Ph.D. Dissertation, University of Connecticut, Storrs.
West MJ [1967] Foundress associations in polistine wasps: dominance hierarchies and the evolution of social behavior. Science 157: 1584-1585.
West-Eberhard MJ [1975] The evolution of social behavior by kin selection. Quart Rev Biol 50: 1-33.
Wilson EO, Bossert WH [1971] A primer of population biology. Sinauer Assoc., Stanford, CN.
* The last time I saw Group 5's D4 female, she was seated alone and immobile on a tree limb; her face impaled with quills of the prehensile-tailed porcupine [or, coendou: Coendou]. I never encountered this female again.
**EO Wilson's 2019 book, Genesis, advances the idea that many Mammals, including, humans, may be "eusocial." Where "temporal division-of-labor" is demonstrated, taxa can be classified, "primitively eusocial;" this classification would apply, also, to any other Vertebrates or, indeed, to any other Animals, where "age polyethism" is identified. If "tradeoffs" [e.g., energetic, reproductive, survivval] are most likely to be observed in "poor" conditions [e.g., heterogeneous regimes where "fitness" is compromised; recurrent drought, unpredictable food or water supply], "age polyethism" may evolve to minimize energetic costs in time and space. While females are expected to be most sensitive to energetic effects, males, also, may benefit, under some conditions, from age-dependent responses. Furthermore, there may be energetic [reproductive] benefits in coordinating many maturational [age-dependent] and developmental [age-dependent] milestones or markers with one another as genetic and physiological energy-savings tactics and strategies.
-----------------------------------------------------------------------------------------------
Table 1. Age class, estimated age in years, number of females in each age class (N), observed (O), and expected (E) frequencies of social foraging, and chi square (X2) for a test of the null hypothesis. In “Age Class” column, YA= Young Adult; M-a= Middle-aged Adult; M-a-O= Middle-age to Old Adult; O= Old Adult.
| Age Class | N | O | E | (O-E)2/E (X2) |
| YA (5-7) | 5 | 15 | 42.4 | 17.71 |
| M-a (7-10) | 5 | 35 | 42.4 | 1.29 |
| M-a-O (10-15) | 1 | 18 | 8.1 | 12.11 |
| O (15+) | 1 | 33 | 8.1 | 76.54 |
| Total | 12 | 101 | 101 | 107.65 |