Wednesday, August 6, 2008 - 10:50 AM

COS 54-9: Ant protection and plant reproduction: A meta-analysis of mutualism strength and its ecological correlates

Matthew D. Trager1, Smriti Bhotika1, Jeffrey A. Hostetler1, Gilda V. Andrade2, Mariano A. Rodriguez-Cabal1, C. Seabird McKeon1, and Craig W. Osenberg1. (1) University of Florida, (2) Universidade Federal do Maranhão


The costs and benefits for each partner can vary greatly in most mutualistic interactions, but we know surprisingly little about the factors that drive such variation across systems. We conducted a meta-analysis to quantify the effect of ants on plant reproductive output in ant-plant protective mutualisms. Specifically, we evaluated the relationship between foliar herbivory and reproductive output, and we also explored the sources of heterogeneity among studies to determine the importance of environmental variables and the features of the ant-plant. We used a hierarchical mixed-effects model to test the effects of these variables on the response of plant reproductive output (or foliar herbivory) to the presence or absence of protective ants.  

Analysis of 59 ant and plant species pairs showed that ant protection increased plant reproductive output by 50% over plants on which ants were not present. Effects on reproduction and herbivory were positively correlated, as expected, but the foliar herbivory response was substantially greater than the reproduction response. Plant species that produced domatia (an adaptation for attracting and housing protective ants) received significantly greater benefits from ant presence than plant species lacking domatia. The location of extrafloral nectaries did not affect ant protection, nor did the presence of honeydew-producing Hemiptera tended by ants. Perennial plants benefited more from ant protection than annuals, but plant growth form, habitat type and precipitation did not affect the benefit plants gained from ant protection. Contrary to theoretical predictions, the number of species of potential ant partners did not affect reproductive output, suggesting that the benefits of ant protection did not decrease as the relationship became more diffuse. These results support some theories about the function, evolution and maintenance of mutualisms, but also show that some findings from individual systems may not be consistent across ant-plant protective relationships.