Data for: A 2000-year sediment record reveals rapidly changing sedimentation and land use since the 1960s in the Upper Mara-Serengeti Ecosystem
The Mara River basin in East Africa is a trans-boundary river basin that highlights many of the development and conservation challenges in East Africa. The Mara River flows from its headwaters in the Mau Forest of Kenya through the northern portion of the Serengeti-Mara Ecosystem and into Lake Victoria in Tanzania, where it forms part of the headwaters of the Nile River basin. Changes in land use and landcover in the basin have raised concerns about the quality and quantity of water in the Mara River. We analyzed sediment cores from the Mara Wetland (near the river’s outlet into Lake Victoria) to evaluate how sedimentation rates and sources have changed historically, through a period marked by major changes in human and livestock population densities, land use, and disease epidemics (rinderpest). We collected sediment cores in August 2015 along a transect through the Mara Wetland from the upstream reaches to Lake Victoria. Slices of sediment cores collected from four distinct regions of the wetland were age-calibrated using radiocarbon and lead-210 dating methods. Core slices were then analyzed for sediment sources using a sediment fingerprinting approach, nitrogen and carbon stable isotope signatures, and mercury. Our results suggest that ecological conditions in the Mara River basin were fairly stable over paleoecological time scales (2000-1000 years before present), but there has been a period of rapid change in the basin over the last 250 years, particularly since the 1960s, likely due to anthropogenic factors. We also observed that downstream effects of landcover and land use change can be exacerbated by increasing occurrence of extreme rainfall events in the region. The Mara Wetland likely plays an important role in mitigating the impact of those factors on Lake Victoria. This work was the result of a partnership between Yale University, the Cary Institute of Ecosystem Studies, WWF-UK, WWF-Kenya and WWF-Tanzania.
Steps to reproduce
There are two zip files in this repository: 1. MainAnalysis.zip 2. SeasonalWetlandFingerprinting.zip In each of those zip folders is a R Markdown file that can be used to recreate all of the analyses used in this manuscript. Simply open the R Markdown file and follow the directions. There is also a corresponding HTML file that is the final run of each of those R Markdown files with all the analysis and figures present. I recommend just viewing the HTML file, but you can run everything again if you want by using the R Markdown files. The third file in this directory is the R file for running the Plum Bayesian age-depth model. This model had to be run separately from the R Markdown files. Run Plum first before running the R Markdown file in the MainAnalysis.zip folder. You'll need one of the files from the "Input" folder within the "MainAnalysis.zip" folder to get all the gamma counter data. Alternatively, you can run the R Markdown file using all the previous runs of Plum that are already present in that zip folder.