Key pecking of 4 pigeons was studied under a two-component multiple schedule in which food deliveries were arranged according to a fixed and a variable interfood interval. The percentage of response-dependent food in each component was varied, first in ascending (0, 10, 30, 70 and 100%) and then in descending orders, in successive conditions. The change in response rates was positively related to the percentage of response-dependent food in each schedule component. Across conditions, positively accelerated and linear patterns of responding occurred consistently in the fixed and variable components, respectively. These results suggest that the response-food dependency determines response rates in periodic and aperiodic schedules, and that the temporal distribution of food determines response patterns independently of the response-food dependency. Running rates, but not postfood pauses, also were positively related to the percentage of dependent food in each condition, in both fixed and variable components. Thus, the relation between overall response rate and the percentage of dependent food was mediated by responding that occurred after postfood pausing. The findings together extend previous studies wherein the dependency was either always present or absent, and increase the generality of the effects of variations in the response-food dependency from aperiodic to periodic schedules.

Contemporary analyses of choice were implemented to analyze the acquisition and maintenance of response allocation in Lewis (LEW) and Fischer 344 (F344) rats. A concurrent-chains procedure varied the delay to the larger reinforcer (0.1, 5, 10, 20, 40, and 80 s). Delays were presented within sessions in ascending, descending, and random orders. Each condition lasted 105 days, and the entire data set was analyzed to obtain discounting functions for each block of 15 sessions and each food delivery across delay components. Both a hyperbolic-decay model and the generalized matching law described well the choices of LEW and F344 rats. Estimates of discounting rate and sensitivity to the immediacy of reinforcement correlated positively. The slope of the discounting function changed with presentation orders of the delays to the larger reinforcer. Extended training reduced differences between the LEW and F344 rats in discounting rates, sensitivity to the immediacy of reinforcement, and estimates of the area under the curve. We concluded that impulsive choice can change as a function of learning and is not a static property of behavior that is mainly determined by genetic and neurochemical mechanisms. Choosing impulsively may be an advantage for organisms searching for food in rapidly changing environments.

No abstract

Resistance of Temporally Controlled Behavior to Change Mirari Elcoro To extend research on the relation between temporal control and resistance to change, the temporal location of response-independent food was examined in a modified peak-interval procedure (PIP) with pigeons. Sequences of the PIP consisted of 2, 3, 4 or 5 FI 30-s trials followed by a 90-s peak trial, each separated by a 90-s blackout. In different conditions, response-independent food was delivered within the first, last, or a randomly selected FI trial of each sequence. Overall, quarter-life values, peak times, and peak rates showed the highest decrease in temporal control relative to baseline when disruption occurred in the last FI trial. Delay of disruption gradients, constructed from each of these measures, were similar in form and function to delay of reinforcement gradients.