Recent studies from temperate systems suggest variation in host plant quality (e.g. nutrients, defense compounds) affects herbivore population growth, ant-herbivore mutualisms, and predators of protected herbivores. Yet few investigations of bottom-up forces have been conducted in tropical systems, where strong predation pressure and highly protective ant-mutualists are well established. In coffee agroecosystems, Azteca instabilis ants protect the coffee green scale (Coccus viridis) from coccinellid predators (Azya orbigera), but little is known about whether nutrient addition affects coffee quality and insect interactions. I investigated whether differences in nutrient inputs to coffee plants affect coffee green scale growth, ant attendance, and predation by coccinellids with a manipulative experiment. At an organic coffee plantation in Chiapas, Mexico, I grew coffee seedlings (n=150) with three organic fertilizer treatments: high, medium, low (organic compost to sand ratio) and measured plant growth (height, leaf number, and leaf size) for two months. I infested each seedling with ~30 scales and excluded ants from half of the plants. I then located 25 A. instabilis colonies and placed 6 plants (1 per treatment) at each colony. Finally, I observed the effect of treatment on ant attendance, scale population size, and scale removal rates by coccinellids for two months.
Results/Conclusions

Initially, plants did not differ in terms of number of leaves, height, or leaf size. After two months, however, plants that received higher nutrient inputs had a higher number of leaves, were taller, and had faster leaf expansion rates than plants with lower nutrient levels. The highest fertilization treatment plants had higher numbers of scale insects and a higher number of Azteca-ants per scale insect than lower treatment plants. Generally, plants without ants had lower scale population sizes largely due to removal by coccinellid predators. Within the ant exclusion treatment, scale removal rate was highest in the highest fertilization treatment plants compared to the low treatment. These results indicate that nutrient inputs to coffee plants affect plant quality and insect food webs, but future research is needed to determine what aspect of plant quality (e.g. caffeine defense, leaf nitrogen levels, etc) mechanistically cause change in plants and in the insect interactions. This study is among the first to examine how bottom-up forces affect populations of potential pests, their mutualists, and their predators in coffee agroecosystems and adds to the under represented literature of bottom-up affects in tropical systems.