Nathan Shelton and Selena Smith, University of Michigan
Carolyn Stroemberg, University of Washington
Grasslands make up ~40% of Earth's terrestrial vegetation and provide humans with our staple foods (e.g., corn, rice, wheat). Exactly how ongoing anthropogenic climate change will impact grassland ecosystems and agricultural crop productivity is a crucial question, and addressing it requires knowledge of how these important biomes were assembled in the first place. Based on plant macrofossils, co-evolved mammals, stable isotopic data, fossil soils, and phytolith (plant silica) data, the scientific community knows that grassland ecosystems dispersed at an intercontinental geographic scale sometime during the mid-Cenozoic. However, because different proxies suggest different scenarios for the emergence of grasslands both in terms of gross spatial and temporal patterns (e.g., North America vs. Africa), the details of this profound ecological transition are less certain. In most areas, only one paleo-vegetation proxy has been applied, so very little is understood about potential sampling and taphonomic biases in different records that could affect vegetation interpretation. In addition, potentially important environmental heterogeneity remains undocumented. To resolve these issues, the PIs will use an interdisciplinary, multi-proxy approach to reconstructing the evolution of grasslands in Montana over the past ~40 Ma that can be compared directly with a paleoclimatic reconstruction based on paleosols. Specifically, the PIs will construct a long-term, high-resolution record of paleovegetation using phytoliths, pedogenic carbonates, and organic matter collected from paleosols (i.e., C isotopes to assess photosynthetic pathways). They will focus on three critical climatic transitions that are broadly believed to have significantly impacted the presence and abundance of grasses in North America, namely the Eocene-Oligocene transition (EOT), the Oligocene-Miocene transition (OMT), and the Middle-Late Miocene transition (MLMT).
Aim: To reconstruct the vegetation history of Montana during the past 40 million years to examine how climatic change has impacted the spread of grassland ecosystems. Background Today, grassland ecosystems cover up to 40% of Earth’s land surface and provide critical habitats for a variety of animals, and ecosystem services to people. For example, most of our staple grains such as wheat, rice, and corn are grasses. Exactly how ongoing anthropogenic climate change will impact grassland ecosystems and agricultural crop productivity is a crucial question, and addressing it requires knowledge of how these important biomes were assembled in the first place. While grasses first evolved while dinosaurs still roamed the Earth, the formation of our modern, grass-dominated ecosystems occurred much later and seems to have involved several steps. A body of evidence shows that open-habitat grasses became ecologically dominant in most regions only during the last 25 million years ago, and it was not until 8-5 million years ago that tropical grasslands such as savannas, dominated by grasses using the C4 photosynthetic pathway spread at low- to mid latitudes. However, because different lines of evidence (e.g., fossil faunas vs. floras) suggest different scenarios for the emergence of grasslands, the details of this profound ecological transition are less certain. In most areas, only one line of evidence for paleo-vegetation has been applied, so very little is understood about how different types of data might bias the interpretation. To resolve these issues, the PIs compared several lines of evidence for paleovegetation to reconstruct the evolution of grasslands in Montana over the past ~40 Ma that and compared it directly with a paleoclimatic reconstruction based on fossil soils. Specifically, the PIs constructed a long-term, high-resolution record of paleovegetation using plant silica (phytoliths), soil carbonates, and organic matter collected from fossil soils (i.e., carbon isotopes to assess photosynthetic pathways). Scientific Merit and Outcomes: Our team undertook a many field excursions with teams from the University of Michigan and University of Washington, including senior personnel, graduate students, and undergraduate students. After collecting, collating, and analyzing over 1000 rock samples, we were able to draw a number of conclusions about the complex vegetation history of Montana: 1) Montana was relatively cool and dry since at least 14 million years ago, but was much warmer and wetter before 30 million years ago, and covered in (sub)tropical vegetation typified by palms and gingers. 2) C4 plants were present in Montana for at least the past 20 million years, substantially predating the "global" rise to dominance. 3) Warming or cooling temperatures did not drive the expansion and contraction of grassland vegetation; instead, the amount of precipitation was an important control. Broader Impacts and Outcomes Beyond the broader scientific impacts of the project for Earth Science, the project contributed to the education and professional development of a large number of graduate, undergraduate, and high-school students at the Universities of Michigan and Washington. Two PhD students at UM graduated and have moved on to post-doctoral fellowship positions and one UW student will complete her dissertation in 2015. Five of the undergraduate students are now in graduate school (one after earning a Fulbright Fellowship, another one an NSF GRFP), and four others are currently applying to graduate school. Most of the students who were trained come from groups underrepresented at the professional level in STEM fields. The project also involved making educational museum outreach exhibits at both UM and UW designed to engage K-16 students. PI Sheldon and graduate student Jennifer Cotton designed and disseminated pedagogical materials for teaching about sustainable agriculture via an article in the Journal of Geoscience Education. PI Strömberg and graduate student Harris featured the project in several UW Burke Museum outreach programs (e.g., Girls in Science, Dino Day). One of the research products was also a study that used climate change projections and insights from this project to understand which regions within the US will become less effective sinks for CO2 emissions and which will become more effective sinks. Products At the time of writing, ten papers have been published in scholarly journals, with others in review or in preparation for submission. Thirty abstracts have been submitted and talks or posters presented at national and international scientific conferences. Project websites and a lab blog have been published online; datasets for completed papers have been shared via journal data repositories and on UM’s project website, which will be maintained as a long-term archive of the scientific results.
