Rapid, accurate time estimation in zebrafish (Danio rerio)

• Zebrafish were trained in a Pavlovian conditioning paradigm.
• The conditioned stimulus (CS) was an illuminated pole. The unconditioned stimulus (US) was dry shrimps. The conditioned response (CR) was biting directed at the pole.
• The CS–US interval was manipulated to look at its effect on the latency of the CR.
• The CR latency displayed proportional timing (the CR latency was proportional to the CS–US interval) and the scalar property (the CR latency standard deviation was proportional to its mean).
• It took as few as 40 trials for the CR latency to adapt to the CS–US interval.

Zebrafish were tested in an appetitive Pavlovian delayed conditioning task. After an intertrial interval of k*T s (k = 11.25; T = 8, 16 or 32 s), a small, translucent vertical pole was illuminated (CS) for T s. Food was presented at T/2 s. Pole-biting response latencies from CS onset were a linear function of the food delay T/2, with slope approximating unity (proportional timing), and standard deviation proportional to latency (scalar timing). Response latencies tracked changes in food delays even when they changed every other day. These findings are significant because the zebrafish genome has recently been sequenced, opening the door to studies in the genetics of interval timing.