Controls over root lifespan are poorly understood. Hypotheses attempting to explain what controls and constrains root lifespan include the “Starch depletion hypothesis” (SDH) of Marshall and Waring, the “Resource optimization hypothesis” (ROH) and a new hypothesis proposed here, which is referred to as the “Metabolic activity hypothesis” (MAH). The Starch depletion hypothesis assumes that a finite amount of stored carbohydrates (starch) is deposited at root formation and that the rate the carbohydrates are depleted by root respiration determines the lifespan of the root. The Resource optimization hypothesis assumes that root lifespan is optimized to provide the greatest benefit in terms of water and nutrients for the least cost (usually measured in carbon) over the lifespan of the root or cluster of roots. The Metabolic activity hypothesis suggests that root lifespan is mainly governed by metabolic rate; roots with higher respiratory activity live shorter lives than those with lower respiratory activity. These hypotheses were tested in a common garden in Central Pennsylvania with temperate trees.
There is little evidence to support the starch depletion hypothesis. Results of 13C labeling of trees indicate C can be deposited in not only starch but also in structural materials well after the root is born. In addition, respiration rates indicate maintenance respiration far exceeds starch stores over the lifetime of a root. Long-term localized N fertilization led to increased root lifespan compared to unfertilized roots based on first quartile lifespans. These results provide support for the Resource optimization hypothesis, assuming that longevity was extended for the roots providing the most benefit for the least cost. Nonetheless, based on the negative correlation of root N concentration with root lifespan across species in this and in other studies, the MAH may still be supported under some circumstances. A better understanding of controls and constraints of root lifespan will allow better estimates of root turnover and allow models to be more predictive of shifts in root turnover with changes in species composition or the environment.