Hazard Structure and Repair Kinetics: A 2x2 Classification that Resolves the Stochastic-Death / Individual-Tolerance Dilemma

The two variants of GUTS (General Unified Threshold model of Survival), stochastic death and individual tolerance, disagree about no-effect concentrations under time-variable exposure, and no standard test separates them. Each variant fixes two independent choices, which form a 2x2 classification: whether the hazard accumulates while damage exceeds a threshold (cumulative) or is set by the highest damage reached (peak-based), and whether repair is proportional or saturating. Calibration fit cannot choose: each structure fits by adjusting its own damage recovery, the cumulative one faster in twelve of fourteen compounds. For two identical pulses separated by complete recovery, the cumulative hazard predicts equal survival through both whatever the parameters, so a carry-over ratio (second-to-first survival) above one excludes it; in Gammarus pulex fourteen of fifteen ratios are at or above one, seven significantly. The peak-based hazard predicts the ratio's maximum, approached as intervals lengthen in five of six compounds. For organophosphates, whose recovery outlasts the published pulse intervals, the one test on measured damage favours the peak-based hazard: diazinon-exposed Daphnia magna stopped being immobilised once acetylcholinesterase, still inhibited, began to recover. The peak-based hazard is the appropriate default for toxicokinetic-toxicodynamic risk assessment of narcotics and uncouplers, and provisionally of organophosphates.