Abstract. Carbon dioxide concentrations in the atmosphere play an
integral role in many Earth system dynamics, including its influence on
global temperature. The past can provide insights into these dynamics, but
unfortunately reconstructing long-term trends of atmospheric carbon dioxide
(expressed in partial pressure; pCO2) remains a challenge in
paleoclimatology. One promising approach for reconstructing past pCO2
utilizes the isotopic fractionation associated with CO2 fixation during
photosynthesis into organic matter (εp). Previous studies have focused
primarily on testing estimates of εp derived from the δ13C
of species-specific alkenone compounds in laboratory cultures and mesocosm
experiments. Here, we analyze εp derived from the δ13C of
more general algal biomarkers, i.e., compounds derived from a multitude of
species from sites near a CO2 seep off the coast of Shikine Island
(Japan), a natural environment with CO2 concentrations ranging from
ambient (ca. 310 µatm) to elevated (ca. 770 µatm) pCO2. We observed
strong, consistent δ13C shifts in several algal biomarkers from
a variety of sample matrices over the steep CO2 gradient. Of the three
general algal biomarkers explored here, namely loliolide, phytol, and
cholesterol, εp positively correlates with pCO2, in agreement with
εp theory and previous culture studies. pCO2 reconstructed from the
εp of general algal biomarkers show the same trends throughout, as well
as the correct control values, but with lower absolute reconstructed values
than the measured values at the elevated pCO2 sites. Our results show
that naturally occurring CO2 seeps may provide useful testing grounds
for pCO2 proxies and that general algal biomarkers show promise for
reconstructing past pCO2.