Uncertainties in interpretation of isotope signals for estimation of fine root longevity: theoretical considerations

This paper examines uncertainties in the interpretation of isotope signals when estimating fine root longevity, particularly in forests. The isotope signals are depleted δ 13 C values from elevated CO 2 experiments and enriched Δ 14 C values from bomb 14 C in atmospheric CO 2 . For the CO 2 experiments, I explored the effects of six root mortality patterns (on–off, proportional, constant, normal, left skew, and right skew distributions), five levels of nonstructural carbohydrate (NSC) reserves, and increased root growth on root δ 13 C values after CO 2 fumigation. My analysis indicates that fitting a linear equation to δ 13 C data provides unbiased estimates of longevity only if root mortality follows an on–off model, without dilution of isotope signals by pretreatment NSC reserves, and under a steady state between growth and death. If root mortality follows the other patterns, the linear extrapolation considerably overestimates root longevity. In contrast, fitting an exponential equation to δ 13 C data underestimates longevity with all the mortality patterns except the proportional one. With either linear or exponential extrapolation, dilution of isotope signals by pretreatment NSC reserves could result in overestimation of root longevity by several‐fold. Root longevity is underestimated if elevated CO 2 stimulates fine root growth. For the bomb 14 C approach, I examined the effects of four mortality patterns (on–off, proportional, constant, and normal distribution) on root Δ 14 C values. For a given Δ 14 C value, the proportional pattern usually provides a shorter estimate of root longevity than the other patterns. Overall, we have to improve our understanding of root growth and mortality patterns and to measure NSC reserves in order to reduce uncertainties in estimated fine root longevity from isotope data.