Schedule-induced drinking as a function of interreinforcement interval in the rhesus monkey.

The team worked with four rhesus monkeys. Each monkey sat in a cage with a food cup and a water spout.

Food pellets arrived on fixed-interval schedules. The intervals were 30, 60, 120, 240, or 480 seconds. Water was always there. The researchers counted every lick over the study period.

Drinking peaked when food came every 120 seconds. Longer intervals made drinking drop off fast.

Two things drove the drop: monkeys started fewer drinks after each pellet, and each drink bout got shorter.

Ballard et al. (1975) saw the same upside-down U in rats, but they cut the percentage of pellets instead of stretching the interval. Both studies show less food access makes adjunctive drinking rise then fall.

Castilla et al. (2013) added hunger to the mix. Hungry rats drank even more at short intervals, proving the 120-s sweet spot moves when motivation changes.

Johnson et al. (1994) zoomed inside one 120-s cycle. They gave water probes at different moments. The closer the probe landed to the next pellet, the less the rat drank. This confirms the peak is not flat; timing within the cycle still matters.

If you run FI schedules in a classroom or animal lab, watch for extra drinking, pacing, or other adjunctive behaviors around the two-minute mark. That is when the schedule naturally makes them strongest. Move the interval shorter or longer, or add a brief water restriction, and the side behavior fades. Use this lever when self-injury or stereotypy crops up right after reinforcement.