An experimental, behavioral, and chemical analysis of food limitations in mutualistic Crematogaster ant symbionts inhabiting Macaranga host plants

Obligate mutualistic plant-ants are often constrained by their plant partner's capacity to provide resources. However, despite this limitation, some ant partners actively reject potential prey items and instead drop them from the plant rather than consuming them, leaving the ants entirely reliant on host plant-provided food, including that provided indirectly by the symbiotic scale insects that ants tend inside the plants. This dependency potentially increases the efficiency of these ants in defending their host. We hypothesize that if this ant behavior was beneficial to the symbiosis, prey rejection by ants would be observed across multiple plant host species. We also hypothesize that plant-provided food items and symbiotic scale insects from other ant plants should be rejected. We address these hypotheses in the Crematogaster ant-Macaranga plant system, in which plants provide living space and food, while ants protect plants from herbivory. We observed food acceptance and rejection behavior across five ant species and three plant host species. Ants were offered three types of food: termites as a surrogate herbivore, symbiotic scale insects, and nutritious food bodies (FB) produced by different host plant species. The unique ant species living in M. winkleri was the most likely to reject food items not provided by the plant species, followed by ants in M. glandibracteolata, while ants in M. pearsonii accepted most items offered to them. Using stable isotopes, chemical cues, and proteomic analyses, we demonstrate that this behavior was not related to differences between plant species in nutritional quality or composition of FB. Isotopic signatures revealed that certain species are primary consumers but other ant species can be secondary consumers even where surrogate herbivores are rejected, although these values varied depending on the ant developmental stage and plant species. Macaranga pearsonii and M. glandibracteolata, the two most closely related plant species, had most similar surface chemical cues of FB. However, M. glandibracteolata had strongest differences in food body nutritional content, isotopic signatures, and protein composition from either of the other two plant species studied. Taken together we believe our results point toward potential host coercion of symbiont ants by plants in the genus Macaranga Thouars (Euphorbiaceae).

Food bodies in Cissus verticillata (Vitaceae): ontogenesis, structure and functional aspects

BACKGROUND AND AIMS

The distinction between pearl bodies (or pearl glands) and food bodies (FBs) is not clear; neither is our understanding of what these structures really represent. The present work examined the ontogenesis, structure, ultrastructure and histochemical aspects of the protuberances in Cissus verticillata, which have been described since the beginning of the 19th century as pearl glands or pearl bodies, in order to establish a relationship between their structure and function.

METHODS

Segments of stems and leaves in different stages of development were collected and fixed for study under light microscopy as well as electron transmission and scanning microscopy. Samples of FBs were subjected to chemical analysis using thin-layer chromatography.

KEY RESULTS

The FBs in C. verticillata are globose and attached to the plant by a short peduncle. These structures are present along the entire stem during primary growth, and on the inflorescence axis and the abaxial face of the leaves. The FBs were observed to be of mixed origin, with the participation of both the epidermis and the underlying parenchymatic cells. The epidermis is uniseriate with a thin cuticle, and the cells have dense cytoplasm and a large nucleus. The internal parenchymatic cells have thin walls; in the young structures these cells have dense cytoplasm with a predominance of mitochondria and plastids. In the mature FBs, the parenchymatic cells accumulate oils and soluble sugars; dictyosomes and rough endoplasmic reticulum predominate in the cytoplasm; the vacuoles are ample. Removal of the FBs appears to stimulate the formation of new ones, at the same place.

CONCLUSIONS

The vegetative vigour of the plant seems to influence the number of FBs produced, with more vigorous branches having greater densities of FBs. The results allow the conclusion that the structures traditionally designated pearl glands or pearl bodies in C. verticillata constitute FBs that can recruit large numbers of ants.