Phylogeny

Why has aphid/ant mutualism emerged in many species? Is ant attendance of aphids always beneficial to both species?



Photo by Jeremy Yoder
Photo by Jeremy Yoder

 

Ant attendence incurs some costs on aphids

Aphids show a range of associations with ants from nonattendance to obligate myrmecophily (ant attendance)(Stadler et al). In labratory experiments Symydobius oblongus (an obligate myrmecophile) performed better under the attendance of L. niger, while Aphis fabae cirsiiacanthoides (a non-mermecophile) performed worse than when unattended. However, Under field conditions, ant-attended colonies of both A. f. cirsiiacanthoides and S. oblongus achieved higher peak numbers and lasted longer than unattended colonies (Stadler et al). The costs of ant attendance to aphids include a restricted habitat, loss of nutrients through honeydew (though there is evidence that honeydew is primarily a waste product), and occasional predation from ants.

 

Aphid/ant mutualism has several origins within hymopteran evolution

Multiple emergences of a behavior in related species across a phylogenetic tree suggests that the behavior is highly favorable. Mutualism with ants is suspected to be a highly labile (open to change) trait within homopteran evolution (Shingleton et al). Shingleton et al. constructed a molecular phylogeny of 15 Chaitophorus Koch (Hemiptera: Aphidoidea) species, using mitochondrial cytochrome oxidase I and II sequences. Ant tending evolved, or was lost, at least five times during Chaitophorus evolution. Parametric bootstrapping (a method of statistical approximation) rejected the hypothesis of a single origin of ant tending in this genus (Shingleton et al).


Fig. 1. Log-likelihood ratio tests of monophyly of untended aphids LML is the log-likelihood score for the unconstrained (maximum likelihood) tree; L0 is the log likelihood score for the constrained (null hypothesis) tree; delta 0.95 indicates the critical value of the 95% confidence interval for the distribution of the log-likelihood ratio test scores (d star) simulated under the null-hypothesis; d obs is the log-likelihood ratio test score for the observed data and p is the probability of obtaining > or = d obs under the null-hypothesis of monophyly. (Shingleton et al).