Short Communication
Pigeons with continuous access to a schedule of reinforcement that required 50 pecks for each delivery of food (of 4 s) maintained their free-feeding body weights, but pigeons on a schedule of 200 pecks consistently lost weight. These results indicate that the upper level of a pigeon's ability to compensate for increased work loads per food presentation is relatively low or that pigeons must maintain a feeding rate in excess of 4 s of food every 40 to 67 s in order to maintain responding.
Keywords:Feeding behaviour, Weight changes, Fixed-ratio schedule, Pigeons
Introduction
A detailed analysis of the feeding behaviour of pigeons is presented by Zeigler, Green and Lehrer (8), Zeigler, Green and Siegel (9) and Zeigler (7). They do not, however, provide any information on the effects that the behavioural cost of food has on feeding. Collier, Hirsch and Kanarek (2) suggest that animals in the natural setting and undeprived animals in a chronic testing situation adjust their feeding behaviour in response to environmental constraints so that an optimal balance is maintained between long term food input and the expenditure of time and energy. They present data indicating that as the number of responses required for a meal increase the number of meals decrease but the duration of those remaining meals increases in a compensatory manner. They also show that as the cost per single reinforcer or "bite" increases the response rate increases. This compensatory rate increase is thought to occur in order to maintain a relatively constant feeding rate.
Remarkable examples of animals maintaining optimal daily intake of food in the face of enormous work loads have been presented. Hirsch and Collier (4) demonstrate that Guinea Pigs (Cavia porcellus) will respond 10,000 times a day, and Collier, Hirsch and Hamlin (I) show that rats will respond over 60,000 times a day to obtain their daily ration. In both cases subjects had been maintained on chronic fixed ratio schedules (6). In this procedure the animal is permanently housed in the testing situation and is required to respond a fixed number of times on a freely available operandum to obtain each portion of its daily diet.
The present study is concerned with the long term efficacy of the compensatory feeding mechanism in pigeons. In particular, we determine whether pigeons which are maintained on a chronic FR5O schedule can maintain optimal food input rates for more than a few days when the experimental conditions require additional work for each bite. The work-load increase was sufficient to produce lower response rates in birds maintained at 80% body weight and run in short daily sessions (3). However, it was small enough to allow birds to maintain the same feeding rate if the compensatory response rate increase predicted to occur under chronic schedules did in fact occur. The study is a systematic replication, using pigeons, of the earlier studies by Collier, Hirsch and Hamlin (1) and Hirsch and Collier(4).
Methods
Four pigeons of undetermined sex were used. All had been exposed to various light cycles for two years while being maintained on a chronic FR50 schedule (6). The schedule had provided food for 4 s following 50 responses. Three Birmingham Rollers were approximately four years old at the start of the experiment, and one adult feral pigeon was of undetermined age.
All pigeons were housed in a battery of four contiguous open-wire cages throughout the experiment. Each cage was 34 cm high x 40 cm wide x 44 cm deep. The fronts were modified to accommodate aluminum response panels measuring 30 cm wide x 34 cm high. Water cups beside the response panels were kept full. A 5 cm diameter food aperture was medially located on the response panel 9 cm above the wire floor. A 2 cm diameter key that required 0.15 N force to operate was located 9 cm above the food aperture and 9 cm to the right. Each key was dimly lit at all times. The experiment was conducted in a windowless room. Lighting was provided by a 60 watt bulb in front of the cages and a 40 watt fluorescent tube immediately behind the cages. The chambers were continuously illuminated throughout the 242 days of the experiment (LD 24:0) in order to eliminate variations in the feeding rate attributable to the light cycle. Food hoppers and water cups were serviced daily between 16:00 and 24:00. The pigeons were also weighed at that time. They had continuous access to water cups and to the response panel. The experiment involved maintaining the birds under either a chronic FR5O schedule or a chronic FR200 schedule. Each schedule was in effect until feeding behaviour was stable or until one or more birds had fallen to about 70%body weight (labeled a Phase). The chronic FR50 schedule followed every fiftieth peck with access to the food hopper for 4 s throughout the 24 h day (Phases 2 and 4). The chronic FR200 schedule required 200 pecks for each food presentation (Phases 1, 3 and 5). The duration of each of the five phases is indicated in Table 1.
| TABLE 1 SUMMARY OF PHASES |
||
|---|---|---|
| Phase Number | Schedule | Number of Days |
| 1 | FR200 | 50 |
| 2 | FR50 | 27 |
| 3 | FR200 | 82 |
| 4 | FR50 | 48 |
| 5 | FR200 | 35 |
Results and Discussion
The results are presented in Figure 1. The five bars associated with each bird indicate their percentage of free feeding body weight at the termination of each phase. All birds steadily lost weight during the FR200 phases, and quickly regained weight during all FR50 phases to plateau at or near their normal body weight.
This experiment demonstrated consistent weight loss associated with every imposition of an FR200 schedule. The inability of pigeons to intake food in the face of this seemingly modest additional requirement is consistent with expectations based on data collected from birds deprived to 80% of their normal weight and tested for short periods (3). However, it is not consistent with expectations based on data of Collier, Hirsch and Hamlin (1) and Hirsch and Collier (4). It is also puzzling in light of Murton's (5) report that pigeons in the wild spend nearly eight hours a day foraging when food is limited.
The pigeons used in this experiment required approximately 25 presentations of food each day to maintain their normal body weight. The FR200 schedule therefore only required about 28 min of responding at three responses a second or about 83 minutes at one response a second. These rates are not inordinately high for a pigeon and would not need to be sustained for very long periods in that pigeons normally spread their feeding over the daylight hours. If total daily work output is considered, 5,000 pecks is far less work than performed by rats (about 65,000 responses and 9.85 h a day) or Guinea Pigs (15,000 responses and 8.33 h a day). If work per feeding opportunity is considered, the present experiment required 200 responses per 4 s access to food while the schedule for rats required about 220 responses per 45 mg pellet and the schedule for Guinea Pigs required about 60 responses per 45 mg pellet. Even though no metric exists for comparing the feeding to cost functions in the pigeon, rat and Guinea Pig, it would seem reasonable to have expected pigeons to adapt to the additional work load. Four seconds of access to food for a pigeon provides a larger proportion of its daily food requirements than does a single pellet for the rat or Guinea Pig. In addition, there is no evidence that key pecking is more difficult than lever pressing. If anything, the opposite seems true (3).
To assure that experience with the chronic FR200 schedule did not play an important role in the ability of a bird to maintain itself on that schedule, the present design attempted to implement an FR200 schedule several times. There was no evidence that the FR200 schedule would be easier for experienced birds. Considering the inability of an FR200 schedule to maintain body weight in pigeons, it seems that either pigeons require a feeding rate in excess of 4 s of food every 40 to 67 s (the feeding rate obtained with 5 and 3 responses per s) to maintain feeding or that the rates controlled by schedules of reinforcement are not primarily the result of feeding strategies.
Acknowledgments
The authors gratefully acknowledge the contributions of Elizabeth Palya in all phases of this research. Requests for reprints should be sent to William L. Palya.
References
1. Collier, G., E. Hirsch and P. Hamlin. 1972. The ecological determinants of reinforcement in the rat. Physiology and Behaviour 9:705 -716.
2. Collier, G., E. Hirsch and R. Kanarek. 1977. The operant revisited. In: Handbook of Operant Behavior. Edited by W. K.. Honig and J. E .R. Staddon. Englewood Cliffs, N.J.: Prentice-Hall .
3. Ferster, C. B. and B.F. Skinner. 1957. Schedules of Reinforcement New York: Appleton-Century-Crofts.
4. Hirsch, E. and G. Collier. 1974. The ecological determinants of reinforcement in the guinea pig. Physiology and Behaviour 12:239-249.
5. Murton, R.K. 1965. The Wood Pigeon. London: Collins.
6. Palya, W.L., P. Hurst, J.J. Kenny, W.M. Gardner and G. Seale. 1983. Light cycle controlled feeding and weight changes in pigeons. Bird Behaviour 4:71-77.
7. Zeigler, H.P. 1976. Feeding behavior of the pigeon. Advances in the Study of Behavior 7: 285-3 89.
8. Zeigler, H.P., H.L. Green and R. Lehrer. 1971. Patterns of feeding behavior in the pigeon. Journal of Comparative and Physiological Psychology 76:468-477.
9 Zeigler, H.P., H.L. Green and J. Siegel. 1972. Food and water intake and weight regulation in the pigeon. Physiology and Behaviour 8:127-134.