George Collier

The ecological determinants of reinforcement in the rat

Abstract The intake of a freely feeding animal can be characterized by four parameters: frequency of meals, duration of meals, rate or intensity of eating, and choice of items eaten. The values of these parameters characterize species and the ecological niche they occupy. Constraints on any one of these parameters leads to compensatory changes in other parameters over the feeding cycle. Any change in performance of the instrumental behavior ancillary to feeding which causes the values of these parameters to recover their optimum or privileged value, defined by their niche, will be strengthened. These changes in performance reflect what is usually meant by motivation and the process of reinforcement.

Schedule-induced wheel running

Abstract Rats reinforced for bar pressing on a variable interval schedule were allowed access to a running wheel. Wheel running displayed many of the characteristics of schedule-induced drinking. These data call into question the explanations of schedule-induced drinking based entirely upon thirst motivation, and suggest that it is only one of a number of post-reinforcement behaviors affected on an intermittent reinforcement schedule.

BODY WEIGHT LOSS AS A MEASURE OF MOTIVATION IN HUNGER AND THIRST

An animal deprived of food or water does two things: He loses weight and he becomes more active. Many studies have shown an intimate relation between these two classes of events (cf. Bolles, 1967). The premise of this paper is that further study of this relation may add to the knowledge of the underlying physiological substrates of hunger and thirst. Three hypotheses are examined: ( 1 ) Body weight loss as an active process involved in the defense of certain physiological parameters and, therefore, a basic measure of degree of depletion; (2) Body weight loss as a basic measure of motivation, as reflected in levels of instrumental or spontaneous activity; and (3) Depletion as related to motivation via the recovery growth function. The first question to be examined is how body weight (BW) is lost under conditions of restricted food or water intake. Both within and among species there is, in general, a linear relation between the log intake of water or food and log BW (Adolph, 1949; Kleiber, 1961; Richter & Brailey, 1929). In a fasting animal the same relation holds between his final (asymptotic) weight and the size of ration given. This suggests that for any given nutrient, the amount of the nutrient consumed is always proportional to the resulting BW, thus maintaining a constant ratio. If the nutrient is present in inadequate amounts, and none of the nutrientprocurring or -conserving responses of the animal is sufficient, it follows that BW must be lost in order to conserve the ratio of the nutrient consumed to BW. First, let us consider the case of water deprivation. The hypothesis is that the ratio of body water to lean body mass must be maintained (Collier, 1964; Collier 8~ Knarr, 1966; Collier & Levitsky, 1967). FIGURE 1 presents the weight-loss curves for two groups of rats. The animals represented by the upper curve rereceived 24 percent of their ad lib intake of water, whereas the lower curve animals received no water. TABLE 1 shows the percentage of water, determined by drying, in a selected set of organs from animals sacrificed at various times following the institution of the 24-percent ration. No significant differences as a function of time were observed. Similar results were obtained for the second group receiving no water (see also Wolf, 1958). Upon further examination of the curves in FIGURE 1, several questions arise: How is the rat losing weight, and what is the end point of the weight loss? Does BW stabilize on a limited ration and if so, at what value? Kleiber (1961 ) suggests that a fasting animal consumes himself at a rate proportional to the amount of active metabolic tissue left at any moment. If metabolic rate does not change greatlv during fasting, as suggested by Rixson and Stevenson (1957), one may predict from this assumption that log BW will decline linearly as a function of time (cf bottom curve of FIGURE I ). When some water or food is available, the rate of decline decreases as the BW approaches a value that can be supported by the amount of nutrient given. FIGURE 2 shows this relation between asymptotic weight and water ration for groups of rats given different proportions of their ad lib water intake. The form of this curve is the classic linear function relating log of food or water intake and log BW in ad lib-fed ani-

The effects of gonadectomy on the sex differences in dietary self-selection patterns and carcass compositions of rats.

Abstract The relationship of dietary self-selection patterns and carcass composition was studied in intact and gonadectomized male and female rats. Males consumed a significantly greater proportion of their diets as protein and had a greater proportion of their carcasses as fat than females. Gonadectomy increased the percentage protein selected by females to the level of intact males and altered the proportions of the carcass tissues to the levels of these males. The results were discussed in terms of the regulatory consequences of varying nutritional intake patterns and the role of female sex hormones in the regulation of body weight and composition.

Dietary self-selection by pregnant and lactating rats ☆

Abstract Protein-carbohydrate choice was studied in female rats through the course of the estrus cycle and during pregnancy and lactation. Both selecting and control diet fed animals grew at the same rates. Food intake was reduced during the estrus phase, but there were no changes in the dietary selection pattern. During both pregnancy and lactation protein intake was increased while carbohydrate intake was maintained at a level equal to nonimpregnated controls. These data were interpreted as supporting other studies showing that dietary self-selection follows the varying nutritional requirements of the organism.

Dietary self-selection in active and non-active rats.

Abstract Active rats given a choice of protein and carbohydrate chose a higher proportion of carbohydrates than did non-active animals. These results were interpreted as showing that active rats differ in their nutritional requirements and that this difference reflects changes in intermediary metabolism.

Effects of diet and deprivation on meal eating behavior in rats

Abstract The meal eating pattern of six male albino rats was observed during the feeding of Purina Chow, a calorically diluted diet, a calorically dense diet, and following 24 hr food deprivation. The total number of meals remained fairly constant throughout all conditions. The duration of a meal, on the other hand, was extremely sensitive to both changes in diet and to body weight loss following food deprivation.

The time window of feeding.

This paper contrasts the traditional depletion/repletion model of food intake with a longer-term perspective which focuses on function rather than mechanism. We review naturalistic observations as well as the economic relationships we have discovered in the laboratory by a cost/benefit analysis of feeding in a closed economy. We have manipulated feeding costs by means of operant methodology and have explored two classes of cost, the cost of initiating meals and the cost of consuming food. We conclude that when an animal can detect the cost/benefit structure of its habitat, its feeding behavior will tend to maximize benefits relative to cost in apparent anticipation of its nutritional requirements and environmental conditions. The time window over which these feeding decisions are integrated may be minutes, hours, months, or seasons depending upon the animals niche and current habitat. Feeding mechanisms based on momentary deficits and repletions are inadequate to explain these strategies, leaving the field of feeding mechanisms open for major discovery and revision.

Self-starvation: A problem of overriding the satiety signal?

Rats housed in either activity wheels or standard laboratory cages received access to food either ad lib or for one 60-min, two 30-min, or four 15-min periods per day. Imposition of restricted feeding schedules led to reductions in food intake and body weight which were greater for animals with access to activity wheels. Increases in activity reflected the percent of body weight loss, which varied directly with frequency of food access. Subsequent recovery of intake was facilitated by partitioning total feeding time into briefer but more frequent periods. In the most extreme frequency-of-access condition, animals with access to running wheels failed to recover from the reduction of intake incurred by imposition of the restricted feeding schedule, even though their total feeding time was the same as that of animals that did recover. These data indicate that self-starvation is not induced by activity per se but results from a general failure to recover intake which, in turn, results from a failure to override the satiety signal within a meal.

Dietary self-selection at cold temperatures ☆

Abstract Animals housed in the cold (2°C) were given the opportunity to select the components of their diets from protein and carbohydrae fractions. Housing in the cold was accompanied by depressed growth, which was not affected by giving the animals the opportunity to select their diets. Animals in the cold increased total caloric intake relative to controls at room temperature (22°C). Selecting animals housed in the cold chose a lower percentage of the diet as protein (and, therefore, a higher percentage as carbohydrate) than did controls. Since protein intake levels did not differ as a function of environmental temperature, the altered dietary selection pattern was due to an increase in the absolute amount of carbohydrates consumed. These data were interpreted as suggesting that the level of protein selected is determined by age factors and the level of carbohydrate selected is determined by energy requirements.