Noelaerhabdaceae coccolithophores as recorders of ancient atmospheric CO₂
Understanding the relationship between carbon dioxide and global climate is critical for predicting the severity of future climate change resulting from ongoing anthropogenic CO₂ emissions. By the year 2100, atmospheric CO₂ concentrations may reach close to 1000 ppm. The impacts of such a massive pe...
Main Author: | |
---|---|
Format: | Thesis |
Language: | unknown |
Published: |
Columbia University
2020
|
Subjects: | |
Online Access: | https://dx.doi.org/10.7916/d8-6re0-d956 https://academiccommons.columbia.edu/doi/10.7916/d8-6re0-d956 |
Summary: | Understanding the relationship between carbon dioxide and global climate is critical for predicting the severity of future climate change resulting from ongoing anthropogenic CO₂ emissions. By the year 2100, atmospheric CO₂ concentrations may reach close to 1000 ppm. The impacts of such a massive perturbation to the carbon cycle will last for hundreds of thousands of years. Estimates for the magnitude of warming range between ~3 and ~12°C, depending upon the timescale considered and actual CO₂ trajectory. To better understand the likely impacts of the Anthropocene, we must examine how CO₂ and climate have varied in the past, particularly during warm periods in Earth history with CO₂ levels higher than modern. Studies of this nature require the use of geological archives to reconstruct past environmental conditions. The carbon isotope fractionation of marine algae recorded in alkenone biomarkers (εp) is one of the primary tools for reconstructing past atmospheric CO₂ variations. These molecules are produced by the Noelaerhabdaceae, a family of calcifying coccolithophorid algae that has flourished in the open ocean for tens of millions of years. Measurements of alkenone εp have offered critical insight into the evolution of Earth’s atmosphere. However, CO₂ reconstructions using alkenone εp in the Pleistocene seldom agree with the atmospheric CO₂ variations known from ice cores. The reasons for this incongruity are not well understood. Recent studies have provided a deeper understanding of the complex mechanisms affecting alkenone carbon isotope fractionation, suggesting that the conventional framework for interpreting these records in terms of CO₂ may require reconsideration. Quantitatively understanding the link between alkenone εp and CO₂ can provide new insights to the history of atmospheric CO₂ and global climate on Earth. Here I use an empirical approach to explore relationships between physiology, environment, and alkenone carbon isotope fractionation in order to quantify the influence of CO₂ on alkenone εp. I begin with both new data and a meta-analysis of previous controlled laboratory studies on one of the dominant extant alkenone-producing algae. These experiments identify and quantify the significant roles of CO₂, irradiance, and cell size on alkenone εp. I calibrate multiple linear regression models from these culture data and find that when irradiance is included, growth rate is an insignificant contributor to εp variability. I test this model in core-top sediments and in Pleistocene deep-sea sediment sequences when atmospheric CO₂ concentrations are known from ice-core gas bubble analysis. With paired measurements of coccolith size and alkenone carbon isotope ratios, I find that although there is no spatial relationship between εp and CO₂ in the modern ocean, alkenone εp can be modeled reasonably well in core-top sediments, suggesting that the culture-based model is faithfully capturing the sensitivity of εp to irradiance, CO₂ and cell size in the modern ocean. Using Pleistocene alkenone εp and coccolith size records, I identify locations best-suited for paleobarometry, finding that tropical sites with stable surface hydrography and low variability in incident irradiance are sites where CO₂ and cell size dominate the p signal during the late Pleistocene. I apply this model to a new deep-sea sediment record to reconstruct CO₂ changes over the last ~20 million years. This record exhibits a drop in atmospheric pCO₂ of ~600 µatm over this time period, and suggests that the global cooling and C4 grassland expansion in the Miocene were likely caused by pCO2 decline. Together, these results present a new way to understand alkenone εp variations through time and quantify past CO₂ changes. |
---|