Promiscuity and the Primate Immune System

I. M. Roitt, J. Brostoff, D. K. Male, Immunology (Gower Medical Publishing, London, ed. 5, 1998).

Alexander R. D., Annu. Rev. Ecol. Syst. 5, 325 (1974).

Freeland W. J., Biotropica 8, 12 (1976).

Møller A. P., Dufva R., Allander K., Adv. Study Behav. 22, 65 (1993).

Dobson A. P., Meagher M., Ecology 77, 1026 (1996).

Hausfater G., Meade B. J., Primates 23, 287 (1982).

K. K. Holmes et al., Eds., Sexually Transmitted Diseases (McGraw-Hill, New York, 1994).

Thrall P. H., Antonovics J., Bever J. D., Am. Nat. 149, 485 (1997).

S. B. Hrdy, P. L. Whitten, in Primate Societies, B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham, T. T. Stuhsaker, Eds. (Univ. of Chicago Press, Chicago, 1987), pp. 370–384.

C. P. van Schaik, M. A. van Noordwijk, C. L. Nunn, in Comparative Primate Socioecology, P. C. Lee, Ed. (Cambridge Univ. Press, Cambridge, 1999), pp. 204–240.

D. M. Taub, in The Macaques: Studies in Ecology, Behavior and Evolution, D. G. Lindburg, Ed. (van Nostrand, New York, 1980), pp. 287–344.

Analyses were based on independent contrasts (25, 26), with the exact method depending on the types of variables examined (27). For bivariate analysis of mixed categorical and quantitative variables (e.g., Fig. 1), we examined specific evolutionary transitions in the dependent variable using the BRUNCH algorithm in the CAIC computer program (27). For two quantitative variables (e.g., Table 1) and for all multivariate analyses [see (14)], we used the CRUNCH algorithm (27). The primate phylogeny was taken from (28). We found that logarithmic transformation of the data and equal branch lengths best met the assumptions of independent contrasts (29). We checked for and removed outliers in contrasts plots, as such outliers often have high leverage and may represent the influence of confounding variables (30). Because specific directional predictions were tested, we used one-tailed statistical tests, but the primary results are provided such that two-tailed probabilities can be calculated. The unpublished data sets and links to programs used to conduct the analyses are available at

International Species Information System, Physiological Reference Values [CD-ROM] (Minnesota Zoological Garden, Apple Valley, MN, 1999).

Substrate use: b = 0.04, F(1,20) = 0.86, P = 0.18; group size: b = −0.04, F(1,20) = 0.41, P = 0.26; number of mating partners: b = 0.06, F(1,20) = 4.21, P = 0.03. In multivariate analysis of substrate use and number of mating partners, we used a three-part ranked categorization of substrate use to match variation in the three-level mating partner categories. Substrate categories included: arboreal, semi-terrestrial in a wooded environment as intermediate, and maximally terrestrial in an open environment [updated from (31)]. Because no established methods exist for examining multiple categorical independent variables in contrasts analysis, and to increase sample sizes for these tests, we treated both variables as continuous in the CAIC computer program (27).

Clutton-Brock T. H., Harvey P. H., J Zool. 183, 1 (1977).

We investigated the relative roles of substrate use, mating partner number, and body mass as predictors of neutrophil counts. First, in a multivariate regression analysis of contrasts, both body mass and categorical measures of substrate use were statistically significant predictors of neutrophil counts [b = 0.17, F(1,35) = 9.69, P = 0.002 and b = 0.08, F(1,35) = 3.51, P = 0.03, respectively). Second, to determine if substrate use impacted the analysis of mating promiscuity, we excluded one contrast with a co-occurring shift in substrate use and mating partner number. Overall WBC counts, lymphocytes, and monocytes remained significantly associated with mating partner number (n = 8 contrasts, P = 0.01 to 0.04), whereas both neutrophils and body mass increased with transitions to greater terrestrial substrate use (n = 5 contrasts, t = 2.87, P = 0.02 and t = 2.81, P = 0.02, respectively). However, multivariate analysis of these variables using the CRUNCH algorithm (27) provided no significant results (n = 23 contrasts, P = 0.09 to 0.11). Finally, we tested whether the allometric relationship with neutrophils reflects an underlying life history correlate. In particular, a stronger immune system might be required in species with a longer life-span. However, longevity did not account for neutrophil counts when holding body mass constant in multiple regression [b = −0.23, F(1,34) = 1.82, P = 0.19]. No significant results were found in allometric analysis of other WBC types, although a negative slope for lymphocytes approached significance (b = −0.11, F(1,38) = 3.66, P = 0.06).

Harcourt A. H., Harvey P. H., Larson S. G., Short R. V., Nature 293, 55 (1981).

Harcourt A. H., Purvis A., Liles L., Funct. Ecol. 9, 468 (1995).

In males, overall WBC counts were highly associated with female counts in contrasts analysis [b = 0.64, F(1,38) = 24.07, P < 0.0001], as were particular WBC types (P < 0.05 in all tests). This makes sense in the case of sexually transmitted disease: when one sex experiences increased risk, then the other sex should experience a corresponding increase (32). Male overall WBC (t = 3.14, P = 0.007), neutrophils (t = 3.15, P = 0.007), and lymphocytes (t = 1.87, P < 0.05) increased significantly over evolutionary transitions in female promiscuity (monocytes, t = 1.61, P = 0.07). Patterns of male promiscuity are difficult to analyze because there is less detailed information on male partner number, and, within species, greater variance in male mating success (i.e., sexual selection) may weaken patterns across species. Thus, analyses of discrete transitions to increased partner number in males produced significant results for neutrophils (n = 4 contrasts, t = 2.66, P = 0.04) when using the same set of species, as in analyses of females. But other analyses were not significant, including those using a wider range of species to give more contrasts. In multivariate contrasts analysis, however, residual testes mass accounted for variation in male WBC in excess of that explained by female WBC counts [b = 0.08, F(1,20) = 4.86, P = 0.04). Although other explanations are possible, this relation is consistent with male mating promiscuity affecting basal WBC counts.

First, we compared human standard reference WBC counts (33) to values for nonhuman primates among the discrete mating categories. For overall WBC, neutrophils, and lymphyocytes, ranges for humans closely matched equivalent ranges in monogamous species, while human monocyte values best matched ranges for the intermediate category of “1+ mates” (see Fig. 1 caption). Second, we used the midpoint of human reference values in a hierarchical cluster analysis of the apes. We found that humans align most closely with the gorilla (Gorilla gorilla), a polygynous species with low sperm competition (17), and secondarily with a monogamous gibbon (Hylobates lar).

Lockhart A. B., Thrall P. H., Antonovics J., Biol. Rev. 71, 415 (1996).

Hart B. L., Neurosci. Biobehav. Rev. 14, 273 (1990).

Loehle C., Ecology 76, 326 (1995).

Antia R., Levin B. R., May R. M., Am. Nat. 144, 457 (1994).

Felsenstein J., Am. Nat. 125, 1 (1985).

P. H. Harvey, M. D. Pagel, The Comparative Method in Evolutionary Biology (Oxford Univ. Press, Oxford, 1991).

Purvis A., Rambaut A., Comput. Appl. Biosci. 11, 247 (1995).

Purvis A., Philos. Trans. R. Soc. London Ser. B 348, 405 (1995).

Garland T. J., Harvey P. H., Ives A. R., Syst. Biol. 4, 18 (1992).

Nunn C. L., Barton R., Am. Nat. 156, 519 (2000).

C. L. Nunn, C. P. van Schaik, in Reconstructing Behavior in the Fossil Record, J. M. Plavcan, R. F. Kay, W. L. Jungers, C. P. van Schaik, Eds. (Plenum, New York, in press).

Thrall P. H., Antonovics J., Dobson A. P., Proc. R. Soc. London Ser. B. 267, 1555 (2000).

W. Z. Borer, in Conn's Current Therapy, R. E. Rakel, Ed. (Saunders, Philadelphia, 2000), pp. 1205–1213.

Smith R. J., Jungers W. L., J. Hum. Evol. 32, 523 (1997).

Harvey P. H., Clutton-Brock T. H., Evolution 39, 559 (1985).

Goldizen A.W., Trends Ecol. Evol. 3, 36 (1988).

Ackerly D. D., Donoghue M. J., Am. Nat. 152, 767 (1998).

We thank S. Altizer, P. Bennett, C. Carbone, K. Jones, S. Patek, C. van Schaik, and K. Winkler for comments and discussion, C. Williams for calling our attention to the International Species Information System, and M. Carosi and M. Gerald for data on Cebus apella testes measurements. This research was supported in part by an NSF Postdoctoral Research Fellowship in Biological Informatics to C.N. and NIH Grant GM60766-01 to J.A.