Abstract:
* Leaf life span (LLS) has been intensively studied as a key functional trait, and it is thought to have evolved and acclimates so as to optimize carbon balance or nitrogen use. However, empirical studies have produced inconsistent results in support of the theoretical predictions of optimal LLS. * How rapidly daily carbon gain declines with leaf age is a critical parameter in the theories of optimal LLS, and it is often estimated from empirical data on the mean daily carbon gain of surviving leaves at each age class. We predict that such statistical approach should result in overestimation of daily carbon gain at the mean LLS, especially when LLS variation is large in the leaf cohort. * This prediction is supported by simple simulations; daily carbon gain linearly declines to zero at the death of each individual leaf within a cohort (Case 1), and daily carbon gain linearly declines to zero at the cohort mean LLS (t¯) (Case 2). In addition, variance in the initial carbon gain is considered, with the inverse relationship between the initial carbon gain and LLS but other assumptions are same as in the Case 1 (Case 3). Under the Cases 1 and 3, the mean daily carbon gain of surviving leaves at (t¯) is always positive, and it increases with increasing LLS variance within a cohort. Under the Case 2, the mean daily carbon gain of surviving leaves at (t¯) is zero regardless of variation in LLS, but this case is unrealistic as some leaves with negative carbon balance are assumed to survive for long-time when LLS variability is large. * Published data on multiple species demonstrate a positive relationship of photosynthetic capacity at (t¯) with LLS variability as predicted by our simulation under the Cases 1 and 3. This strongly suggests that the age-related decline of carbon gain may be underestimated in many previous studies that neglect within-cohort variation in LLS. * In conclusion, we call attention to the importance of LLS variations within a leaf cohort, which should be considered in empirical test of the theories of optimal LLS.
