LB 29-250 - The effect of urbanization on a desert ant-plant defense mutualism

Some of the most influential drivers of ecological diversity are mutualistic relationships between different species, but our understanding of how city landscapes affect these relationships is limited. Many plant species engage in a mutualistic relationship with different ant species in which the plants secrete a sugary nectar from extrafloral nectaries to attract ants who consume the nectar. In exchange, the ants provide protection by killing and eating other insects on the plants, including potential herbivores that may harm the plant. The barrel cactus (Ferocactus wislizeni) can develop mutualistic relationships with different ant species, including non-native species. In this study, we investigated the ant species composition and quality of protection services towards cacti in urban Tucson, AZ and compared them to cacti found in rural desert habitats outside of Tucson, AZ. We used surrogate herbivores (ten, first-instar Manduca sexta larva) to quantify ant response rate measured as the speed of recruitment (maximum number of ants found on plant after all caterpillar larva were killed) as well as aggressiveness (the number of ants necessary to remove 50% of the caterpillar larva). We also recorded and investigated habitat factors (habitat size, substrate, connectivity, sunlight, and temperature) and cactus characteristics (height, diameter, fruit number and density). In the ten different ant species that were found in both the rural and the urban landscapes, we found that aggressiveness and recruitment were fairly similar, but that ants were slightly more aggressive in the urban landscape. Surprisingly we only found one invasive species (Brachymyrmex patagonicus) in our disturbed urban sites and only on 4% of our cacti. Two common native species (Solenopsis aurea and Crematogaster opuntiae) found in natural landscapes around Tucson were not found in the city. Native fire ants (Solenopsis xyloni) were the most common species and were strongly correlated with highly disturbed habitats, meaning areas that were measured by little to no patch connectivity, small patch size and surrounding anthropogenic structures. Although S. xyloni were the most aggressive species and were therefore the best defenders of herbivores, they also required the least amount of ants to remove the caterpillars. In addition to this, they did not show any floral tissue avoidance and thus are likely to drive off pollinators and herbivores. The other habitat features did not show any significant correlation with ant behavior or cactus features, as we would have expected from rural studies, which highlights the complicated nature of urban environments.