A field experiment was conducted with outplantings of the native perennial shrub Senna mexicana var. chapmanii in a semi-natural area adjacent to native pine rockland habitat in southern Florida. The presence of ants and the availability of extrafloral nectar were manipulated in a stratified random design. Insect communities were monitored and recorded over a period of six months with a view to addressing three main questions. Do ants provide biotic defense against key herbivores on S. chapmanii? Is the presence of ants on S. chapmanii mediated by EFN? Finally, are there ecological costs associated with the presence of ants on S. chapmanii, such as a reduction in alternative predator or parasitoid numbers? Herbivores on S. chapmanii included immature stages of three pierid butterflies, and adult weevils. Eight species of ants were associated with the plants, and other predators included spiders, ladybugs, wasps, and hemipterans. Parasitic, haemolymph-sucking midges (Ceratopogonidae) and parasitoid flies were also associated with the caterpillar herbivores, and possibly the extrafloral nectaries of the plants. The presence of ants did not appear to influence oviposition by butterflies, as numbers of lepidopterans of all developmental stages did not differ among treatments. Significantly more late instar caterpillars, however, were observed on plants with ants excluded, indicating that ants remove small caterpillars from plants. Substantially more alternative predators (spiders, ladybugs, and wasps) were observed on plants with ants excluded. Rates of parasitization did not differ among the treatments, but there were substantially fewer caterpillars succumbing to virus among those collected from control plants.We provide a rare look at facultative ant-plant mutualisms in the context of the many other interactions with which they overlap. We conclude that ants provide some biotic defense against herbivores on S. chapmanii, and plants benefit overall from the presence of ants, despite negative impacts on non-ant predators.

1. Insect predators often aggregrate to patches of high prey density and use prey chemicals as cues for oviposition. If prey have mutualistic guardians such as ants, however, then these patches may be less suitable for predators. 2. Ants often tend aphids and defend them against predators such as ladybirds. Here, we show that ants can reduce ladybird performance by destroying eggs and physically attacking larvae and adults. 3. Unless ladybirds are able to defend against ant attacks they are likely to have adaptations to avoid ants. We show that Adalia bipunctata ladybirds not only move away from patches with Lasius niger ants, but also avoid laying eggs in these patches. Furthermore, ladybirds not only respond to ant presence, but also detect ant semiochemicals and alter oviposition strategy accordingly. 4. Ant semiochemicals may signal the extent of ant territories allowing aphid predators to effectively navigate a mosaic landscape of sub‐optimal patches in search of less well‐defended prey. Such avoidance probably benefits both ants and ladybirds, and the semiochemicals could be regarded as a means of cooperative communication between enemies. 5. Overall, ladybirds respond to a wide range of positive and negative oviposition cues that may trade‐off with each other and internal motivation to determine the overall oviposition strategy.

Extrafloral nectar is an inducible defense in S. chapmanii. Developmental variations in its production support the growth differentiation balance hypothesis, while within-plant variations and damage responses support optimal defense theory.

1. Extrafloral nectar (EFN) provides plants with indirect defence against herbivores by attracting predatory insects, predominantly ants. Decades of research have supported the role of EFN as an effective plant defence, dating back to Thomas Belt's description of ants on acacia in 1874. 2. Despite this extensive body of literature, knowledge of the ecological role of EFN has rarely been applied in the field of pest management. We review the existing literature on the use of EFN in agriculture and consider the obstacles that have hindered this transition. 3. Chief among these obstacles is the influence of ecological context on the outcome of EFN-mediated interactions. As such, we consider the options for various agricultural systems in the light of the growth habit of EFN-producing species, focusing first on orchard species and then on herbaceous crops. In each case, we highlight the benefits and difficulties of utilizing EFN as a pest management tool and of measuring its efficacy. 4. Synthesis and applications. We argue that it is time for a shift in extrafloral nectar (EFN) research towards applied settings and seek to address the question: How can a context-dependent and often inducible plant trait be utilized as a reliable tool in agricultural pest management? Breeding crops for increased EFN production, and intercropping with EFN-producing plants, can enhance assemblages of beneficial insects in many agricultural settings. Orchard systems, in particular, provide an ecological context in which the attraction of ants can contribute to cost-effective and sustainable pest management programmes over a broad geographic range.