Annual variations from 1877 to 1995 in tree-ring alpha-cellulose C-13/C-12
isotopic ratios for four subarctic Pinus sylvestris trees were determined,
and, in conjunction with a recent record of atmospheric (CO2)-C-13/(CO2)-C-
12 ratios, the historical pattern of photosynthetic isotope discrimination,
Delta(13)C, was evaluated. Year-to-year variability in Delta(13)C has been
as much as 1.5 parts per thousand with the period 1900-1920 showing an ext
ended period of unusually high photosynthetic discriminations. The summers
during these years were, on average, unusually cold. Since 1920 a long term
trend of increasing Delta(13)C of similar to 0.016 parts per thousand yr(-
1) is inferred. We compared measured Delta(13)C with those predicted on the
basis of the theoretical relationship between Delta(13)C and the ratio of
substomatal to ambient CO2 concentration, C-i/C-a using mechanistic equatio
ns for chloroplast biochemistry coupled with a stomatal conductance model.
Two variations of a nonlinear optimal-regulation stomatal conductance model
were compared. Although both models were based on the assumption that stom
ata serve to minimize the average transpiration rate for a given average ra
te of CO2 assimilation, one version of the model incorporated reductions in
stomatal conductance in response to recent increases in atmospheric CO2 co
ncentrations and the other did not. The CO2 sensitive stomatal model failed
to describe the long-term increase in C-13 discrimination, especially afte
r 1950. The insensitive model gave good agreement, suggesting that an obser
ved increase in subarctic Pinus sylvestris Delta(13)C since 1920 is attribu
table to recent increases in atmospheric CO2 concentrations with subsequent
increases in the ratio of substomatal to ambient CO2 concentrations. The m
odel was also capable of accounting for high frequency (year-to-year) varia
tions in Delta(13)C, these differences being attributable to year-to-year f
luctuations in the average leaf-to-air vapor pressure difference affecting
stomatal conductance and hence C-i/C-a.