Using Adaptive Psychophysics to Identify the Neural Network Reset Time in Subsecond Interval Timing

State dependent network models of subsecond interval timing propose that duration is encoded in states of neuronal populations that need to reset prior to a novel timing operation in order to maintain optimal timing performance. Previous research has shown that the approximate boundary of this reset interval can be inferred by varying the interstimulus interval between two to-be-timed intervals. However, the estimated boundary of this reset interval is broad (250-500ms) and remains underspecified with implications for the characteristics of state dependent network dynamics subserving interval timing. Here we probed the interval specificity of this reset boundary by manipulating the interstimulus interval between standard and comparison intervals in two subsecond auditory duration discrimination tasks (100 and 200ms) and a control (pitch) discrimination task using adaptive psychophysics. We found that discrimination thresholds improved with the introduction of a 333ms interstimulus interval relative to a 250ms interstimulus interval in both duration discrimination tasks, but not in the control task. This effect corroborates previous findings of a breakpoint in the discrimination performance for subsecond stimulus interval pairs as a function of an incremental interstimulus delay but more precisely localizes the minimal interstimulus delay range. These results suggest that state dependent networks subserving subsecond timing require approximately 250-333ms for the network to reset in order to maintain optimal interval timing.