The survivorship and reproductive success of animals can often benefit from traits that confer high rates of net energy intake (see Schoener 1971). This simple idea has stimulated one of the most active areas of behavioral ecology, foraging theory (see Stephens and Krebs 1986; Stephens 1990). One branch of foraging theory, known as "optimal foraging theory", examines how animals gather energy by testing the predictions of optimization models, models that make predictions about the ideal solutions to hypothetical problems. This approach can be useful because it provides testable, quantitative predictions. Yet whereas optimization models can be effective for examining adaptations, they are less effective for examining constraints (see Gray 1987; Dunning 1990). A complete understanding of foraging behavior rests on an understanding of the constraints to which foraging animals are subject. Therefore, although many animals have a general tendency to try to obtain high energetic payoffs, their net rate of energy gain is limited by their perceptual and cognitive abilities, by architectural and developmental constraints (Gould and Lewontin 1979), and by the expression of other behaviors that are related to reproductive success. In addition, optimal foraging models often assume that foragers possess a complete knowledge of their environment. This assumption is always violated in the real world; animals' foraging decisions are generally dependent on very limited amounts of information (McNamara and Houston 1980; Stephens and Krebs 1986; Kacelnik et al. 1987; Roche 1996).

 Cindy Langley, a National Institute of Mental Health postdoctoral associate, and William Timberlake, co-director of the Program in Animal Behavior, along with research assistants Joe Leffel and Rory Gont, examined the foraging behavior of Norway rats (Rattus norvegicus) with these ideas in mind. Specifically, they tested whether rats foraging in a radial-arm maze maximized their net energy intake, or if their behavior was influenced by the expression of other behavioral tendencies. Her results have important implications for foraging theory and for experimental design.

 Timberlake (1990, 1993) has emphasized the importance of designing experiments that examine the natural behavioral traits and behavioral hierarchies that species possess, not just responses that are produced by the contingencies of particular experimental procedures. Previous work by Timberlake and others (e.g., Timberlake 1983; Timberlake & White 1990) has indicated the existence of species-typical perceptual/motor organization in rats related to common maze equipment, such as straight alleys, and radial arm mazes. The most novel finding was that rats appeared to efficiently search mazes in the absence of any food. It was out of this quest to examine the assumptions of previous maze experiments, and to examine the assumption of energy maximization, that Langley et al.'s experiments were born.

 Langley et al.'s experimental protocol was simple. Whereas many radial arm maze studies have used mazes with completely enclosed sides or arms that were raised off of the ground, Langley et al. observed the foraging behavior of rats in a radial arm maze that rested on the ground so that the rats were free to leave the maze and walk around the room at any time.

 Langley et al. performed four experiments. In the first, they allowed 10 female albino rats to forage in a six-arm radial maze (Fig. 1). The maze was made of plywood and fiberboard, was painted grey, and had a hexagonal center. A clear piece of plexiglass that was half covered with strips of black paper sat above the center of the maze, providing "cover" for the rat. Each arm was 13 cm wide and 70 cm long. In addition, each arm had 3.8 cm high fiberboard walls (the rats could easily climb over these walls). At the end of each arm there was a 7.5 cm diameter ceramic cup; the experimenter placed a 45 mg food pellet in each ceramic cup prior to each experimental trial. In each experimental trial a single rat was placed in the center of the maze and the experimenter recorded the path of the rat's movement on a data sheet.

 A hypothesis based solely on obtaining the highest possible net rate of energy intake would predict that the rats would travel to the end of one arm, eat the pellet found there, and then travel in a circle across the floor to the other five ceramic cups. Langley et al.'s observations did not support this hypothesis. Initially, the rats spent a number of days of extensively exploring the entire room, focussing on the walls and not paying much attention to the maze. Then, instead of following the shortest path from food cup to food cup, the rats tended to follow the arms of the maze, and only occasionally walked across the floor between two arms. The rats traveled along "paths" (the arms of the maze) significantly more than they traveled along walls, and they traveled along paths significantly more than they traveled in the open. The number of instances of individual rats moving between two arms across the floor in individual experimental sessions averaged less than one.

 Therefore, the rats were not maximizing their net rate of energy intake. However, the interesting question is the following: Why? Potential hypotheses include the following: (1) the rats have a tendency to follow paths; (2) the rats have a tendency to seek the cover at the center of the maze; (3) the rats have an obligate central place foraging strategy wherein they seek a central area after foraging in a patch; and (4) the rats may have not moved across the floor to adjacent arms because their eyesight is so poor that they could not see the adjacent arms across the floor. Both hypotheses two and three could be called central place foraging strategies (see Krebs et al. 1983; Stephens and Krebs 1986); the key difference is that hypothesis two assumes that the rats return to the center because of the cover it provides whereas hypothesis three assumes that the central place strategy is obligate, whether cover is present or not.

 In Experiment II, the "arms" of the maze led from food cup to food cup, instead of from maze center to food cup, so the arms formed a hexagon (Fig. 2). The center of the maze was still present and it still had a cover. Rats in the beginning of individual trials were placed in the center of the maze as in experiment one.

 The rats in Experiment II (n = eight rats) tended to follow the arms between food cups rather then returning to the center of the maze after each cup visit. Rats in this experiment traveled along arms significantly more than they traveled along walls and significantly more than they traveled in the open. In addition, they visited adjacent arms more than 85% of the time, whereas in Experiment I the mean incidence of visits to adjacent arms was only 37%. The results from Experiment II rule out the hypothesis that the rats in Experiment I followed the arms because they were displaying an "obligate" central place foraging strategy or because they were displaying a preference for the cover provided by the center of the maze (a "facultative" central place foraging strategy).

 In Experiment III (n = 10 rats), the maze was arranged as in Experiment I with one important difference: no food was placed in the ceramic cups. Whereas in Experiments I and II the rats tended to visit all six food cups, in this experiment they did not (mean = 4.9 cup visits). This is not surprising; the cups were less important to the rats when they did not contain food. Langley et al. observed that rats in this experiment traveled in the open and along the walls more than in the other experiments. The proportion of wall-following was significantly higher than that of travel in the open over all days. The proportion of path-following was not significantly different than travel in the open over all the days. The proportion of wall-following was significantly higher than that of arm-following in the first block of trials, but not during the last block; this is because wall-following decreased significantly over days and arm-following increased significantly over days.

 In Experiment III the rats did not display central place foraging behavior and they did not follow the most energetically ideal route. But neither of these findings are at all surprising since no food was present in the cups; the rats may have adopted a "general search" pattern of behavior (Timberlake 1993) with which they were exploring the whole environment. The results of this experiment show that rats' behavior is influenced by the presence of food, but that there is some tendency to follow arms even in the absence of food (and the proportion of arm following even increased during Experiment III).

 To examine what the rats would do when arms were arranged both radially and circularly (Experiment IV; n = eight rats), the maze was arranged so that there were six radial arms radiating out from the center, but there were also arms connecting the food cups ( Fig. 3) as in Experiment II. The rats in this fourth experiment traveled along arms significantly more than they traveled along walls and significantly more than they traveled in the open. Their arm-following behavior did not consistently follow a "central place foraging pattern" or a "maximization of net rate of energy intake pattern", however. Instead, the rats tended to follow arms in both radial and circular patterns. This intermediate strategy was further reflected in the number of visits they displayed to adjacent arms: they visited adjacent arms _% of the time in Experiment I (radial maze), over 85% of the time in Experiment II (circle maze), and over 75% of the time in Experiment IV (radial/circle maze).

 This research project reminds us that selection does not act on individual traits or individual currencies, but on a combination of the outcome, in terms of inclusive fitness, of all of the individual's morphological, physiological, and behavioral traits. Many optimal foraging models assume that the currency of the net rate of energy intake is closely related to the ultimate currency of inclusive fitness. Yet this assumption is often violated. Foraging animals must make tradeoffs between the need for high energetic payoffs and other needs, including the need to avoid predators (Lima et al. 1985), the need to maintain territories (Davies and Houston 1981), the need to acquire limiting nutrients (Pierotti and Annett 1987), and the need to avoid toxins (see Stephens and Krebs 1986).

 Langley et al.'s experiments establish that there is more going on in the rats' foraging behavior than a simple quest for energy maximization. Rats clearly have a tendency to follow paths, at least as manifested by the wooden arms of a maze and the walls of the experimental room. The tendency to follow paths may be a result of selection for preference of cover or selection for using paths that aid in navigating through the environment. Mammals often use "exocentric" navigation (navigation based on environmental cues; Benhamou and Poucet 1996), and the arms of the maze may serve as useful cues that help keep the foraging rats orientated in space. The rats' tendency to follow paths may be a byproduct of some other, as yet unknown, adaptation, or it may be the product of a constraint of some kind. Further research is required, both on the Norway rat and on related species, to explore all of these possibilities.

 One of the hypotheses that remains to be falsified is the following: Do the rats avoid crossing the floor between arms in a radial maze because they cannot see the adjacent arms? This hypothesis could be tested with a six-arm "fan maze" in which the six arms have been compacted into 120 degrees of a circle. The distances between adjacent arms in such a maze would only be 40% as long as those in the radial arrangement. However, such an arrangement would also introduce the confound of decreasing the energetic cost of crossing between arms relative to following paths. An alternative, more critical, test would be to leave the maze arrangement as in Experiment I, but increase the visibility of the ends of the arms with visual landmarks (e.g., towers covered with broad black and white lines).

Projects such as Langley et al.'s, that investigate behavioral patterns along a variety of conceptual levels, offer us the ability to refine our assumptions about foraging behavior. Langley and Timberlake plan to conduct more experiments with "floor mazes" to further explore the path following behavior of rats. Experiments that would be particularly interesting to pursue include the following: (1) testing rats on a radial floor maze with rats that were trained with six cups in a circle on the floor where there are no arms and no maze center; (2) testing rats on a circle floor maze when the rats had been trained with cups but no arms or maze center; (3) testing the rats as in Experiments I - IV with no cover over the center of the maze.

 

LITERATURE CITED

 
Benhamou, S., and B. Poucet. 1996. A comparative analysis of spatial memory processes . Behav. Process. 35:113-126.

Davies, N. B., and A. I. Houston. 1981. Owners and satellites: the economics of territory defense in the pied wagtail, Motacilla alba. J. Anim. Ecol. 50:157-180.

Dunning, J. B. 1990. Meeting the assumptions of foraging models: an example using tests of avian patch choice. Stud. Avian Biol., 13, 462-470.

Gould, S. J., and R. C. Lewontin. 1979. The spandrels of San Marco and the panglossian paradigm: a critique of the adaptionist programmed. Proc. Roy. Soc. Lond. B. Biol. 205:581-598.

Gray, R. D. Faith and foraging: a critique of the "paradigm argument form design," Pp. 69-140 in A. C. Kamil, J. R. Krebs, and H. R. Pulliam, eds., Foraging Behavior. Plenum Press, New York.

Kacelnik, A., J. R. Krebs, J. R. and B. Enz. 1987. Foraging in a changing environment: an experiment with starlings (Sturnus vulgaris). Pp. 63-87 in M. L. Commons, A. Kacelnik, and S. J. Shettleworth, eds., Quantitative Analyses of Behavior, Vol. VI: Foraging. Hillsdale, New Jersey: Erlbaum.

Krebs, J. R., D. W. Stephens, and E. L. Charnov. 1983. Perspectives in optimal foraging. Pp. 165-216 in A. H. Brush and G. A. Clark, Jr., eds, Perspectives in Ornithology. Cambridge Univ. Press, New York.

Lima, S. L., T. J. Valone, and T. Caraco. 1985. Foraging efficiency-predation risk trade-off in the grey squirrel. Anim. Behav. 33:155-165.

McNamara, J. M. and A. I. Houston. 1980. The application of statistical decision theory to animal behavior. J. Theor. Biol., 85, 673-690.

Pierotti, R., and C. Annett. 1987. Reproductive consequences of dietary specialization and switching in an ecological generalist. Pp. 417-442 in A. C. Kamil, J. R. Krebs, and H. R. Pulliam, eds., Foraging Behavior. Plenum Press, New York.

Roche, J. R. 1996. Patch-leaving decisions in black-capped chickadees. Anim. Behav. 52: 289-298.

Schoener, T. W. 1971. Theory of feeding strategies. Ann. Rev. Ecol. System. 2:369-404.

Stephens, D. W. 1990. Foraging theory: up, down, and sideways. Stud. Avian Biol., 13, 444-454.

Stephens, D. W., and J. R. Krebs, J. R. Foraging Theory. Princeton, New Jersey: Princeton University Press.

Timberlake, W. 1983. Appetitive structure and straight alley running. Pp. 165-222 in R. L. Mellgren, ed., Animal Cognition and Behavior. North-Holland, New York.

Timberlake, W. 1990. Natural learning in laboratory paradigms. Pp. 31-54 in D. A. Dewsbury, ed., Contemporary Issues in Comparative Psychology. Sinauer, Sunderland, Mass.

Timberlake, W. 1993. Behavior systems and reinforcement: An integrative approach. J. Exper. Anal. Behav. 60: 105-128.

Timberlake, W., and W. White. Winning isn't everything: Rats need only food deprivation and not food reward to efficiently traverse a radial arm maze. Learn. Motiv. 21: 153-163.



_______________________________________________________________

 

WHO'S WHO IN THE PROGRAM IN ANIMAL BEHAVIOR: STEVE SCHOECH by Carolyn Pytte

 

STEPHAN SCHOECH came to Indiana University this fall from the University of Washington, Seattle, where he completed a Ph.D. with John Wingfield in 1995. His dissertation research, reproductive endocrinology of the Florida Scrub Jay, unites parental behavior, life history strategies, and hormonal mechanisms in answering proximal questions of evolutionary significance. Florida scrub jays are "cooperative breeders" in which some adults do not mate, but instead help their parents feed and raise additional offspring. While this behavior has been invoked as a classic example of kin selection, the underlying mechanisms upon which selection may operate to produce the behavior had been largely ignored until Steve undertook this research.

 Schoech's focus has been on investigating whether fundamental physiological differences in reproductive hormones and gonadal development may be correlated with the behavioral differences seen between breeding and non-breeding jays. Such a correlation offers crucial insights into the processes of evolution of alloparental behavior in this subspecies. His research parallels the interests of Drs. Ellen Ketterson and Val Nolan, with whom he is now working. Steve's interest in how hormones affect behaviors will compliment the Ketterson-Nolan research group's ongoing study of the effects of testosterone upon fitness in dark-eyed juncos.

 Steve put his talents to work right away; he conducted a seminar and laboratory session on techniques and applications of radioimmunoassays (a procedure to determine blood level titers of numerous hormones) in the RTG Fall seminar series called Methods in Behavioral and Neural Endocrinology Techniques. He will also co-lead a discussion on "Publishing Issues" in the RTG Seminar on Research Ethics. Steve will be spending his spring/summer field season at Mountain Lake Biological Station in western Virginia.