Ants, bees, crawly things: why would anyone want to study those? The first thing people ask me when I tell them I work on ants is: do you know how to kill them? In fact, I wouldn’t tell if I knew. Ants, and the other social insects, to me are one of the most fascinating things evolution has produced on earth. They represent a higher level of organization than most other animals, called “superorganisms” by some authors. What this means is essentially that they live in societies of their own with complex communication systems, division of labor, and built structures that can be thousands of times the size of the individual insects. How do social insects achieve this, when colonies are made up of individuals who have no template or overview of the whole society? The algorithms and methods used by ants and bees to achieve adapted collective behavior are the focus of my research.
For example, working with Lars Chittka I’ve discovered a communication system in bumble bees which is not dependent on two individuals meeting in person: one bee places information about the existence and quality of a newly discovered resource on a blackboard (a honeypot in the nest), where other bees can pick it up as needed. I’ve also discovered that a recruitment system doesn’t necessarily involve communicating where food is located; in bumble bees, it seems just the information on whether it is worth looking for food is sufficient. Our model calculations have shown that which communication system is most efficient depends strongly on the spatial distribution of resources.
This result is supported by one of my studies on honey bees: the well-known waggle dance, the honey bees’ way of communicating about food locations, is in fact only useful in some habitats, probably again depending on resource distribution. It is possible that most of the time, this elaborate dance makes no difference to foraging success at all.
I’ve also studied collective decision-making in rock ants, similar to those you find on pavements around the world. These little ants employ a voting-like procedure when deciding collectively which of several potential new homes to move to. My research with Nigel Franks showed that in fact the ants first make a decision about the urgency of the situation: if they have little time, only few ants contribute to the decision-making; if time is of no concern, a large number take part and vote on their favorite nest.
I therefore believe that social insects offer a wealth of methods, optimized by evolution, of solving organizational problems in complex systems that consist of (relatively) simple parts. This research is of great interest to computer scientists and engineers alike, who need such algorithms to design artificial distributed problem solving systems. However, I also think it shows that we should never underestimate an animal because it is small.
My main interests today are organisation of groups, mechanisms of coordination and task allocation and the role of communication in achieving coherent collective behaviour. I have worked with various species of social insects as model systems, studying foraging behaviour as well as decision-making and division of labour.
If you are interested in working with me in any of these areas as undergraduate or graduate student, or as a postdoc, feel free to send me an email at dornhaus@email.arizona.edu
More detail about previous areas of research below, or check my publications list.
in progress
- Trading speed against accuracy in collective decision-making and the effects of colony size on collective behaviour
The honey bees’ waggle dance is generally regarded as one of the most elaborate and fascinating communication systems in the animal kingdom. However, what kinds of selection pressures led to the evolution of such a system is not clear. We have shown that the benefits of this communication depend on the bees' habitat, and that under many circumstances the waggle dance provides no benefits at all. In particular in temperate habitats or in open, disturbed habitats the waggle dance might not confer any benefits in terms of increased foraging success. We tested the significance of the dance language to the foraging success of a honeybee hive under natural foraging conditions (Spain, India) and in a disturbed habitat (Germany). Using horizontal hives to prevent transmission of location information in recruitment, our results indicate that, although the dance behavior is important in stimulating bees to forage, the location information of the dance is not important enough to make a difference to the daily foraging success of a hive under food limiting as well as plentiful conditions in the temperate habitats tested (agricultural area in Germany or natural garrigue area in Spain), whereas colonies had significantly increased foraging success in India (dry forest). These results are important for reconstruction of the origins of the waggle dance and for considerations of flexibility in the evolution of communication systems in general.
- The significance of location communication in honeybee foraging
If one wants to study how and when recruitment was invented in the eusocial bees, bumblebees are the candidate group to look at. They are the next relatives to honeybees and stingless bees, some of which show complex recruitment behaviors. By comparing characteristics of these with behaviors found in bumblebees, one can draw conclusions about which of these evolved in what group. The bumble bee communication system is equivalent to the round dance in honey bees. We have shown that one forager can stimulate the entire foraging force of a colony to start foraging as well. Two pathways of information flow are used here: direct signals from the forager (a pheromone from a tergal gland and potentially acoustic signals from wing buzzing) and the honeypots in the nest as central information storage. Nest bees monitor honeypots and notice nectar influx. They also learn odour of profitable food sources from honeypots and prefer to forage from food sources with this odour. The more complex recruitment systems of honeybees and stingless bees may be derived from a system like the bumblebees', in which a colony can adjust its foraging effort to food availability and allocate its foragers to flower species that are currently rewarding with the mechanisms of alerting and odor learning.
- The bumble bee food alerting system
