Search for: All records

Reconstructing past vegetation can elucidate the timing, climate forcings, and biotic mechanisms of ecosystem change. Plant macro- and microfossils are traditionally used to study past vegetation but suffer from production and taphonomic biases, such as underrepresentation of important herbaceous vegetation components. Geochemical proxies can fill this gap, but carbon isotopes (δ13C) in isolation are unable to distinguish between structurally different C3 habitats, such as forests and grasslands. Thus, new geochemical methods to identify grassy C3 ecosystems are necessary. We present n-alkane chain length distributions of 209 plant specimens from two Kenyan C3-dominated ecosystems, representing a wide range of plant functional types (PFTs). We find that C3 PACMAD grasses produce exceptionally high abundances of long chain C33 and C35 n-alkanes (ACL =32.7, mean C33 +C35 relative abundance =0.69), unlike other C3 PFTs which produce low abundances of C33 and C35 (ACL =28.9–30.3, mean C33 +C35 relative abundance =0.0–0.21). This finding highlights the importance of measuring and reporting the C35 n-alkane. Our data further demonstrate that n-alkane distributions can serve as a proxy for some African C3 PACMAD grasses, offering a new paleoecological tool for distinguishing C3 vegetation types. 

This dataset comprises plant wax n-alkane chain length concentrations and C25-C35 relative abundances of 209 plant specimens from two Kenyan C3-dominated ecosystems, representing a wide range of plant functional types (PFTs). Plant samples were collected in 2018 from Mount Kenya National Park (n=122) and Kakamega National Forest (n=87). At Mount Kenya National Park, samples were collected along an elevational transect (~2400 m to ~3600 m above sea level) from five different ecotones: lower montane forest, bamboo zone, upper montane forest, ericaceous shrubland belt and Afroalpine moorland. Kakamega National Forest is ~1600m above sea level and samples were collected from an open glade, forest path edge, and closed canopy forest. All plant samples were identified to family level, most to genus or species level. Information on collection habitat, photosynthetic pathway, and plant functional type are included. The goal of this dataset was to assess n-alkane distributions for chemotaxonomic signals. Sample analysis took place at Lamont-Doherty Earth Observatory and Harvard University between 2022-2024. n-Alkane data were quantified using a gas chromatograph mass selective detector (GC-MSD) and a flame ionization detector (FID), and response factor corrections were calculated and applied to measured n-alkane peak areas in order to calculate corrected concentrations. The odd n-alkane C25-C35 concentrations were relativized to sum to 1 for the final relative abundance data. For more detailed information, please consult the associated manuscript on the n-alkane distributions and their chemotaxonomic significance: Tweedy et al., 2025. 

Studies on the crustal structure of the Turkana Rift Zone (TRZ) in northern Kenya and southern Ethiopia began in the early 1980s. Initially driven by hydrocarbon exploration, these studies revealed that the rift zone comprises multiple fault-bounded basins ranging in age from the Eocene to the present. They also showed that the area hosts the intersection zone of the N-S trending basins of the Cenozoic East African Rift System (EARS) and the NW-SE-trending Mesozoic-Paleogene Central African Rift System (CARS). However, early seismic reflection and borehole data were mostly concentrated in the southern TRZ, resulting in limited subsurface data for its northern counterpart. This data gap has led to an incomplete understanding of the rift zone's regional crustal structure and how earlier CARS-related rifting influenced the development of the present-day EARS. Here, we leverage newly collected onshore and offshore subsurface industry datasets in the TRZ, spanning a 300 x 150 km region, to characterize the TRZ's crustal structure. We map several key subsurface horizons using a dense grid of 363 2-D seismic reflection profiles, which we tie to surface geology and borehole datasets. Mapping the acoustic basement produced new structure contour maps that provide high-resolution constraints on the TRZ’s crustal structure. Additionally, our isopach maps of key horizons show that strain migrated toward the modern rift axis, located along the center of Lake Turkana, following the widespread eruption of the Gombe Group basalt around 4 million years ago. Together, these results indicate that the area of maximum subsidence is collocated with the area transected by the CARS. Thus, we propose that these earlier episodes of rifting may have influenced the development and evolution of the modern EARS in the northern TRZ. These results provide crucial information for understanding tectonics in the context of hominin evolution and offer new insights into forming a divergent plate boundary. 

Reconstructions of eastern African vegetation and climate are critical for understanding primate and large mammal evolution in the Neogene. Insight into past ecological conditions can be gleaned from lipid biomarkers preserved in sedimentary archives, providing evidence for the role of habitats (e.g. open vs. closed vegetation) on evolutionary trait selection. A common paleoecological proxy is the 𝛿¹³C of n-alkanes, which integrates the distinct isotopic signatures of C3 and C4 vegetation. In typical modern tropical ecosystems, “woody” vegetation uses C3 photosynthesis while “grassy” vegetation uses C4 photosynthesis. Under these conditions, mixing models can then estimate the fraction of woody cover of a landscape. While the use of photosynthetic pathways to infer plant functional type (PFT) is powerful, this paradigm does not hold prior to the rise of C4 grasses at 10 Ma, leaving a gap in understanding of ecosystem structure in the early-mid Miocene. To address this issue, we investigate whether n-alkane chain length distributions (rather than 𝛿¹³C) hold information about plant functional type independent of photosynthetic pathway. Here, we present n-alkane chain length data from over 800 modern plant samples, representing a variety of different photosynthetic pathways, growth forms, habitats, and locations. This dataset comprises a significant literature review component, as well as over 400 new distributions generated in this study. We build upon our previous work using PCA and turn to non-linear methods – including both supervised neural network classifiers and unsupervised dimensionality reduction – to determine the potential of n-alkane distributions for PFT identification. Successful differentiation between woody and grassy PFTs using modern plant n-alkane chain lengths will provide a foundation for applying this tool to the geologic record. Our method will compliment well-established isotopic measurement practices while offering the novel ability to reconstruct vegetation structure in pure C3 ecosystems. This represents a particularly powerful tool for understanding ecological history prior to the rise of C4 grasses.