ALBERT

Description: The effects of the past century's increase in atmospheric CO 2 concentration ([CO 2 ]) have been recorded in the stable carbon isotope composition ( 13 C) of the annual growth rings of trees. The isotope record frequently shows increases in photosynthetic CO 2 uptake relative to stomatal conductance, which estimates the CO 2 concentration gradient across the stomata ( c a – c i ). This variable, which is one control over the net photosynthetic rate, has been suggested as a homeostatic gas-exchange set point that is easy to estimate from 13 C and [CO 2 ]. However, in high-latitude conifer forests, the literature is mixed; some studies show increases in ( c a – c i ) and others show homeostasis. Here we present leaf and tree-ring 13 C data from a controlled experiment that tested factorial combinations of elevated [CO 2 ] (365 and 700 mol mol –1 ) and fertilization on mature Norway spruce ( Picea abies (L.) Karst.) trees in northern Sweden. We found first that the leaf carbon pool was contaminated by the current photosynthate in the older leaf cohorts. This is the reverse of the common observation that older photosynthate reserves can be used to produce new tissue; here the older tissue contains recent photosynthate. We found that the tree-ring data lack such contamination and in any case they better integrate over the canopy and the growing season than do leaves. In the second and third years of treatment, elevated [CO 2 ] alone increased ( c a – c i ) by 38%; when combined with fertilization, it increased ( c a – c i ) by 60%. The results of this study support the idea that annual rings provide a clearer isotopic signal than do foliage age-classes. The tree-ring data show that inferred ( c a – c i ) depends not only on [CO 2 ], but also on mineral-nutrient status. The differences in ( c a – c i ) are sufficiently large to account for the treatment-induced increase in wood-volume production in these stands.