p. 209 * pleometrosis = founding queens remain together after the first workers appear. * monogyny = 1 queen, polygyny = multiple queens, true polygyny = multiple egg-laying queens, oligoygyny = multiple queens that stay far away from each other within the nest. * There are many interesting questions related to queen number, such as its impact on the age structure and growth of colonies, how queens numbers are controlled, what leads one queen to dominate over others. We are focusing on factors that might encourage the formation and differential survival of a multiple-queen colony. * "Varation in the number of queens ... is widespread but not universal among the ants." (210) * Mature colonies of the genus Formicidae are generally monogynous. "Special ecological constraints" are the only thing that reverses this tendency. * Queens cost the colony a lot of energy * If colonies start off with single queens, the virgin queens can travel farther. If multiple queens are needed, the babies can't go as far. * Polygyny does arise in spite of its intrinsic disadvantages. * Increasing the number of queens is an efficient way to increase the worker population of the colony.
* Our model is based on Myrmecocystus mimicus and Solenopsis invicta. Mm. Workers raid nearby colonies after they are born, stealing the workers from those colonies and integrating them into their own colony. As a result, colonies started by 5 to 6 queens are favored. Larger colonies have a survival advantage. Even in colonies where supernumerary queens were killed, those colonies started by pleometrosis have larger growth rates. If the colony is not a certain size by winter, it will die. In southern Brazil, where Si. evolved, extremely large#of queen colonies form. It is a much more hostile habitat. There is no good opportunity for selective pressures to operate against large groups. Pleometrosis is less frequent.
*In our model, this "critical mass" of workers is simulated by the net logo neighbors primitive. A larger number of workers ensure that the queen will never be unprotected, and that the colony will survive. The necessity of having large numbers of workers increases when the danger of workers dying, and the danger of exposed queens dying is high.
*It is generally the case in ant colonies that single-queen colonies are favored unless there is a good reason to have a multiple-queen colony. In our model, the initial incidence of pleometrosis is random, and, in safer environments, single queen colonies will survive. Part of our intention was to create boundaries with a high (smooth) value of danger gradient. The queens that happened to be "planted" in safe areas would coalesce into a single colony
ACTORS
turtles and patches: an agent-based model. patch properties: danger turtle properties: 2 breeds, worker or queen.
INTERACTIONS
birth queens give birth to workers. movement: queens cannot move but workers can. death: this is the central interaction of the model. it is, on the simplest level, a turtle-patch interaction (give equation), but for queens, the danger of the patch they remain on is determined by the proximity of workers belonging to their colony. If a queen has workers or other queens from her own colony on all the patches around her, she has no chance of dying. VARIABLES
birth_rate
The number of new ants born per time step. Increasing birth_rate makes it harder for queens to die. This increases both a single-queen and multiple-queen colonie's chance of survival.
queens_born
This variable determines the number of queens initially scattered over the landscape. For more than 250 queens, the simulation becomes very slow. The notable characteristic of simulations with more than 250 queens is colonies with many queens -- 5 or more, become common. For low values of queens_born, multiple queen colonies are less likely.
danger_gradient and safety
The danger_gradient variable is extremely important to our model. A low danger gradient means ... If the danger gradient is raised, the landscape becomes smoother. Patches with similar danger levels are grouped together. Danger is meant to represent everything that affects ant mortality, including predators and quantity of food available. Safety does not change the danger gradient, but if does affect the magnitude of each ant's death probability. In simulations with a highter value of safety, ants are less likely to die.
adventure_affinity
A higher adventure affinity raises the maxiumum distance worker ants can travel in each time step. We expected that a higher adventure affinity would make the world more dangerous for the queens, as there is a higher probability that the worker ants surrounding each queen will go far enough away to leave her exposed. However, we were curious about the program's sensitivity to this parameter. Are there usually so many workers per patch that the distance they travel in each time step does not matter? Is this parameter only influential at certain times in the course of the simulation, such as when the population of a particular colony is low?
TESTING OF THE MODEL
We used Net Logo's Behavior Space feature to test multiple combinations of a range of values for each of the five variables.
** Data
CONCLUSIONS ** Interpretation of data...
** In general, ... Our model offers a possible explanation for the phenomenon of multiple queen colonies being advanatgeous in dangerous environments. By creating this model, we are offering a hypothesis (subject to future field and computer tests) that, a) spatiality alone may be able to explain the advantages of multiple queen colonies, whether or not it actually is the primary factor that spurred the development of pelometrosis, and b) Enviroments with low spatial heterogeneity in terms of danger are more dangerous. ????????
The next step for this particular program might be to allow raiding of nearby colonies. Workers could convert other workers to their own colony if they locally outnumbered those workers from another nest.
Another thing that could be added is competition among the queens. Our model does not show any disadvantages to multiple-queen colonies, which is highly unrealistic. Ants in nature will form multiple-queen colonies only if it is strictly necessitated by the environment.
A crucial next step in modeling this phenomenon is creation of a model that uses energy as its currency. A significant survival advantage created by pleotromeisis is that the queens can produce more initial workers each per brood. They get energy by eating each other's eggs. A model might have death probability relate to energy of the individual ants, which would in turn be related to the total number of workers in the colony.
There are many more questions to explore, such as, ...
