Bottom-up impacts on the economics of a keystone ant-plant protective mutualism in Kenya

Background/Question/Methods Mutualisms are foundational ecological interactions that can be seen as a trade in commodities between interactive partners. In ant-plant mutualisms, the hosts invest in food rewards and housing in exchange for ant protection against herbivores. Fluctuations in resource availability impact the energetic budgets of the mutualistic species, and thus likely also affect the level of investment placed by each partner in the mutualism. However, the importance of bottom-up forces structuring investment into protective mutualisms, and whether the partners can track and adjust their investment accordingly remains poorly understood. We examined how increases in resource availability impacted host fitness, and the economics of the Vachellia drepanolobium ant-plant mutualism in Kenya. This myrmecophyte provides four symbiotic ant species with extra floral nectar and housing in exchange for various degrees of defense against herbivores. The ants have different energetic needs, and the identity of the mutualistic partners strongly impacts their fitness and survival. (1) Do rapid increases in resource availability alter host investment into growth, reproduction and defense? (2) Do changes in investment on the mutualism by the host tree affect the defensive activity and efficiency of the symbiotic ant Crematogaster mimosae? Our experiment was conducted at Mpala Research Center in Kenya. We increased resource availability by watering and fertilizing 20 trees (50 mg of triple-super phosphate), using 20 unmanipulated trees as controls. We monitored longitudinally (26 months) tree growth, reproduction, and investment in the mutualism, along with changes in the defensive activity of the ant colonies, and the herbivory levels suffered by the trees.

Results/Conclusions: Watered and fertilized trees (WF) produced nearly double the average monthly growth of control trees (C=35.7±7.2, WF=67.8±8.7 cm, P<0.001). Increased resources also translated into higher number of inflorescences (C=8.0±2.7, WF=22.5±2.3, P<0.001), and fruits (C=9.2±2.97, WF±3.5=20.6, P=0.034). Further, WF trees invested disproportionately more in nectar (C=4.68±0.53, WF=6.8±0.5 nectaries/90cm, P<0.001), and nesting space/90cm (C=1.3±0.1, WF=2.2±0.2 new domatia; P=0.003). In turn, ant colonies inhabiting WF trees increased worker densities (C=28.7±1.8, WF=39.4±2.8, P=0.001 ants/90cm), and were faster to attack intruders (C=3.16±0.31, WF=2.15±0.16, P=0.052). Interestingly, despite harboring more aggressive colonies with higher worker densities, WF trees suffered 68% more damage by megaherbivores than the controls (P<0.001). Bottom-up effects can rapidly shift the level of investment placed by each partner on this ant-plant mutualism. The symbiotic ants were able to track and respond to increased investment by the host tree, but this did not translate into lower herbivore damage.