Mutualisms are a key component of ecological communities, yet we lack a mechanistic understanding of how they support ecosystem function. In defensive ant-plant mutualisms, plants provision fixed carbon to ant mutualists in exchange for protection against herbivores. Investment in the mutualism can be costly, as plants must allocate carbon to the mutualism instead of to growth, reproduction, or storage. Thus, carbon can serve as a common currency to identify how mutualisms affect ecosystem function. This becomes even more important when the mutualism is disrupted. In a defensive ant-plant mutualism in Kenya, Phiedole megacephala ants have invaded the savanna, occupying Acacia (Vachellia) drepanolobium trees and extirpating mutualistic native ant species. Invaded trees suffer from increased elephant damage, which results in lower aboveground tree biomass, yet it is less clear how the invasion affects tree carbon allocation. We sampled tree main stems and roots for nonstructural carbohydrates from trees in uninvaded, recently invaded, and long invaded savannas to determine how the invasion affected tree carbohydrate reserves. We also sampled new growth from invaded and uninvaded trees for radiocarbon (14C) to determine if trees were using older carbon reserves or newly assimilated carbon for new growth.
Results/Conclusions
Carbohydrate and radiocarbon data indicated that trees in invaded savannas were under stress that could not be explained by increased herbivory alone. Starch in both stems and roots was marginally higher in recently invaded trees than in uninvaded trees but was lowest in long invaded trees. In contrast, sucrose, which typically mediates a tree’s stress response, was highest in long invaded trees. In addition, radiocarbon data suggested that invaded trees used older carbon for new growth, whereas uninvaded trees used newly assimilated carbon. Together these results suggest that trees in long invaded areas must draw on previously stored carbon to grow because they are assimilating and storing less carbon than trees in uninvaded areas. Because these samples controlled for the level of herbivore damage, there appears to be an additional carbon cost of invasive ants whose cause remains uncertain. We conclude that the carbon dynamics of ant-defended plants change when the mutualism is disrupted, which may have implications for ecosystem carbon cycles by reducing carbon storage in savanna trees.