Morales Lab

Ant-Publilia Mutualism

Population Dynamics

Image of Publilia concava
Publilia concava female with brood and eggs
(Click for larger image)
Image of salticid spider eating treehopper
Jumping spider feeding on P. concava adult
(Click for larger image)

The treehopper Publilia concava is a phloem feeding insect that excretes a sugary waste product that ants collect as a food resource. In return, ants protect treehoppers from predators but also appear to facilitate feeding by treehoppers; tended treehoppers show higher survivorship and are larger than untended treehoppers, even with predators excluded (Morales 2000a, Morales and Beal 2006). Benefit to treehoppers is density-dependent, with treehoppers in small aggregations benefiting more from ant-tending than treehoppers in large aggregations. This pattern of density-dependent benefit is driven primarily by the recruitment patterns of ants. In particular, ant colonies allocate a decelerating number of foragers as the number of treehoppers on a host-plant increases. At the scale of a single plant, this dynamic can be effectively modeled using a type II functional form to describe ant density as a function of treehopper density (Morales 2002). In general, this work supports consumer-resource models of mutualism as a viable theoretical framework for understanding conditionality in ant-homopteran, and other consumer-resource mutualisms.

Recently, my focus has shifted to understanding the spatial dynamics of mutualism, using the interaction between ants and treehoppers as a model system. Research in my lab is currently addressing two questions; first, what factors generate spatial structure in this system, and second, what are the consequences of this spatial structure for the dynamics of mutualism. An emerging picture is that the spatial distribution of treehoppers is generated passively by spatial variation in survival and actively via host-plant selection by females (Morales 2002, Morales and Beal 2006). Spatial variation in survival is driven by spatial variation at the community level. In particular, spatial variation in predator risk (top-down force) and host-plant quality (bottom-up force) are the primary determinants of treehopper success at local to landscape scales. Similarly, spatial variation in the composition of ant communities is the primary determinant of spatial variation in the outcome of the mutualism.

Ant-Treehopper Vibrational Signaling

Image of sonagram
Sonogram of Publilia concava alarm "chirp"
(Click for sound file)

Recent work by me and my lab has identified two vibrational signals produced by treehoppers. One of these signals, which we call a "chirp" is produced by both males and females, and apears to be a general distress signal. For example, chirps can be induced by squeezing treehoppers with tweezers and under natural conditions, treehoppers respond with chirps to predator encounters. We have tested the hypothesis that ants respond to treehopper signals using two native and one invasive species of Myrmica ants. Our results show that both of the native but not the invasive ant species showed an increase in foragers in response to chirp-signal playbacks. Both of two species for which data on the ratio of "foragers" to "patrollers" were collected (one native, one invasive) showed an increase in the number of patrollers in response to "chirp-signal" playbacks.

Flowering Phenology

I am currently collaborating with David Inouye from the University of Maryland to analyze flowering phenology data that he has collected from study plots at the Rocky Mountain Biological Laboratory over the past 25 years. One of the projects that we have worked on is the application of a mid-domain model to understand community-level patterns of flowering richness (Morales et al., 2004). The mid-domain effect is defined as a generated peak in richness at the mid-point of a bounded n-dimensional domain (click here for a simulation program of the mid-domain effect for PC or Mac). The mid-domain effect, which was originally conceived as an explanation for latitudinal or altitudinal gradients in diversity, has been receiving increasing attention in the literature. Our results suggest that the mid-domain effect applies at least as well to richness patterns across temporal domains - we found that a mid domain null-model matched observed patterns of flowering for 9 of the 12 year-site combinations that we analyzed.

Other work has focused on modeling the environmental factors that influence year-to-year variation in flowering success.