A Simulation of Optimal Foraging: The Nuts and Bolts Approach James D. Thomson FOR THE PAST several years, ecology classes at the University of Wisconsin-Madisonhave used a mechanical model to introduce students to the branch of optimal foraging theory that focuses on the benefits and costs of feeding on assemblages of prey that have various feeding and handlingcharacteristics. The success of this approach and its popularity with students has prompted this brief description. Additional details, including a copy of the exercise handout, are available from the author on request. At least one other optimal foraging laboratory has beer, published (Charnov et al., 1976), but it is a pencil-and- paper exercise. In our version the prey are real entities that are caught by the students under the constraint of a time limit. Most students respond enthusiasticallyto this challenge. The exercise has the further advantage of giving students an opportunityto compare the success of various predators in different environments. One disad- vantage is the ratherhigh initial investment in equipment, but the materials will last forever; and if the laboratoryis discontinued, the parts are readilyreusable. The Physical Model Holling (1959, 1965) derived some of the fundamental equations of predation rate using data taken while a blind- folded individualsearched a tabletop for sandpaper disks. Our exercise is an extension of this work. Our environ- ment is a plywood sheet about .75m x lm, but size is not critical.The sheet is studded with an assortment of pro- trudingmachine bolts, threads up, with a hex nut on each. We use 1/4-inch, 3/8-inch, and 7/16-inch diameter bolts to represent small, medium, and large prey. Blindfolded students search the board for prey; once the prey is dis- James D. Thomson (photo not available) received an A.B. degree in biology from the University of Chicago in 1972, and completed M.S. and Ph.D. degrees in zoology at the University of Wisconsin-Madison in 1975 and 1978. During 1979 and 1980 he was a postdoctoral fellow in the Department of Zoology at the University of Toronto. He is cur- rently an assistantprofessor in the Department of Ecology and Evolution at the State Universityof New York at Stony Brook, New York 11794. He has writtenseveral papers on pollination ecology that are in various stages of publication in such journals as Ecology, Oecologia, Journal of Animal Ecology, and American Naturalist. covered, the nut must be unscrewed. Removing the nut simulates the predator'shandling of the prey. The model has no provision for pursuitof prey; thus, the best biologi- cal analogs are found in predation similarto that of bark or foliage gleaning birds that do search for and handle prey but that rarelyhave to pursue it. Handlingtime is ad- justable by the number of turns required to remove the prey nut; we use a wooden spacer or a stack of washers to standardize this figurefor each prey type. Morphological specialization of predators is simulated by requiring the use of mechanic's tools for removal of the nuts. The "generalist" predator uses a set of three box- end wrenches of the appropriate sizes; this ensures approximately equal competence at handling all prey types. The other predator type is a "specialist" on medi- um-size nuts; s/he wields a correctly sized twelve-point socket wrench on a speeder handle. A speeder handle is simply a crank with a knob on one end and provision for attaching a socket at the other; it allows rapid turning of low torque nuts. The best size is 3/8-inch square drive. The speeder handle allows great proficiency with the me- dium prey, but the specialist must use a clumsy 8-or 10- inch adjustable "crescent" wrench to capture small or large prey. This limitation accords with the "jack-of-all- trades, ace-of-none" proposition that underlies most theo- rizingabout animal adaptation. To continue the analogy to birds, birds with highly specialized bills (such as the Crossbills,whose mandibles overlap) are probably very efficient at harvesting certain types of food (coniferous tree seeds in the case of the Crossbills), but much less effi- cient at handling nonspeciality items than a bird with a more generalized bill would be. The relative success of these predators, which could be viewed as a competitive balance, shifts with changes in the distribution and abun - dance of prey. For simplicity,all our prey are the same height and diameter-extra width is provided for the smaller sizes by drilled dowel spacers that also serve to fix handling time-so that all sizes are equally likely to be encountered. (The simulation can be made more com- plex and the analysis more interesting if the prey are differentially conspicuous.) The prey distribution on the boards may be varied in both overall density and proportional representation of 528 THE AMERICAN BIOLOGY TEACHER, VOLUME 42, No. 9, DECEMBER 1980 Downloaded from http://online.ucpress.edu/abt/article-pdf/42/9/528/38527/4447089.pdf by guest on 04 November 2021