Go to the ant, thou theorist

Are quantitative models useful only when they make quantitative
predictions? In ecology, the near impossibility of this goal
seems in conflict with long and deep influence of theory, which
often uses complex methods to arrive at simple qualitative results.
How can we evaluate or even justify work where the eventual product
of much mathematical modeling labor is "it goes up?" Inspired by
Fisher's dictum "make your theories elaborate" I argue that the
best modeling in ecology transcends these apparent paradoxes
by being simultaneously simple and complex, and emulating
the seemingly crude problem-solving methods of social insects.