Environmental proxies of antigen exposure explain variation in immune investment better than indices of pace of life.

Investment in immune defences is predicted to covary with a variety of ecologically and evolutionarily relevant axes, with pace of life and environmental antigen exposure being two examples. These axes may themselves covary directly or inversely, and such relationships can lead to conflicting predictions regarding immune investment. If pace of life shapes immune investment then, following life history theory, slow-living, arid zone and tropical species should invest more in immunity than fast-living temperate species. Alternatively, if antigen exposure drives immune investment, then species in antigen-rich tropical and temperate environments are predicted to exhibit higher immune indices than species from antigen-poor arid locations. To test these contrasting predictions we investigated how variation in pace of life and antigen exposure influence immune investment in related lark species (Alaudidae) with differing life histories and predicted risks of exposure to environmental microbes and parasites. We used clutch size and total number of eggs laid per year as indicators of pace of life, and aridity, and the climatic variables that influence aridity, as correlates of antigen abundance. We quantified immune investment by measuring four indices of innate immunity. Pace of life explained little of the variation in immune investment, and only one immune measure correlated significantly with pace of life, but not in the predicted direction. Conversely, aridity, our proxy for environmental antigen exposure, was predictive of immune investment, and larks in more mesic environments had higher immune indices than those living in arid, low-risk locations. Our study suggests that abiotic environmental variables with strong ties to environmental antigen exposure can be important correlates of immunological variation.

Financial support came from the Schure-Beijerinck-Poppings Fonds (to NPCH and AH), BirdLife Netherlands (to BIT), NSF grant IBN 0212587 (to JBW), and VENI and VIDI grants from the Netherlands Organisation for Scientific Research (to KDM and BIT).

This is the accepted manuscript. The final publication is available at Springer via http://dx.doi.org/10.1007%2Fs00442-014-3136-y.