Reach-scale sediment storage is rarely quantified in sediment budget studies, yet it has a considerable effect on the sediment delivery ratio at the basin scale, and on the accuracy of morphological methods of bedload estimation at the reach scale. Deployment of magnetic tracer particles allows accurate characterisation of sediment fluxes in gravel-bed rivers and provides the opportunity to quantify storage activity using reservoir theory. Activity was quantified at reach and sub-reach scales in two reaches of a small gravel-bed river and the possibility of quantifying the activity of smaller-scale sediment stores is explored. Reach-scale transit time functions were derived from the cumulative output of sediment against age or time. The shape of the functions in both reaches was determined by flood incidence and magnitude (hydraulics), sediment (tracer) availability, grain size and local morphology. Accurate transit time functions were difficult to determine due to reliance upon tracer output and the associated problems of tracer exhaustion over time and imperfect tracer recovery rates. In addition, if tracers of different ages are allowed to mix (for example where upstream input is possible), then the resultant transit time distributions (age) are not comparable with the timing of the hydraulic processes responsible unless the hydraulic conditions are constant. These results suggest that transit time is difficult to determine from tracer studies in gravel-bed rivers. Consequently, a refinement of transit time, the response time, is introduced and is defined as the time, after the initial input of tracer sediment, when cumulative tracer output exceeds the amount of tracer sediment remaining in storage. Whilst still based upon tracer output, the calculation of response time also utilises storage data and is expressed relative to time since the start of the study rather than age (it is therefore directly comparable with the incidence of hydraulic conditions). This provides a more informative measure of activity which is readily available from tracer data in gravel-bed rivers and allows evaluation of the importance of flood incidence and magnitude, grain size, morphology and reach characteristics/stability. Like transit time, response time estimates are hindered by mixing of sediment of different ages at the sub-reach scale. However, it does account for errors from less than 100% recovery and tracer exhaustion more effectively than transit time. Quantification of absolute and relative size effects in transport is also possible from response time.
