Research Context
Soils of Serengeti have been characterised at relatively broad scales (Jager, 1982; Scoon, 2018; Hengl et al., 2015) and are an important influence on vegetation cover and biodiversity (Timbuka and Kabigumila, 2006; Sinclair et al., 2008; Mbise, 2015). The disturbance ecology on the kopje rock outcrops themselves has not investigated in much detail. The paleoenvironmental and historical changes of plant cover, biodiversity, and ecological disturbance, such as fire, remains relatively understudied (Byrom et al., 2015a, Byrom et al., 2015b). Kopjes and rocky hill habitats represent <1% of the total area of Serengeti National Park, Tanzania (total area: 7.4 km2; Byrom et al., 2015a and b; Reed et al., 2009). This new study presents some paleofire metrics measured from a colluvial soil accumulated on Ngong Rock, Moru Kopjes, Serengeti National Park, Tanzania, and from a sediment core collected in Speke Gulf, Lake Victoria, just west of SNP (Courtney Mustaphi, 2019; Courtney Mustaphi et al., 2025a; Courtney Mustaphi et al., 2025b). Increasingly, analyses of soil stratigraphies are being used to study paleoecological changes in the tropics for locations that have limited numbers of permanent lakes and wetlands (Tomlinson, 1974; McWethy et al., 2016; Montade et al., 2018: Mekonnen et al., 2022; Courtney Mustaphi et al., 2025c).
Abstract
Black carbon is a paleofire proxy that has been measured from glacial ice, snow, soils and lake sediments, though relatively few comparisons have been made with other fire indicators in sedimentary geoarchives. Microscopic charcoal, quantified from palynological microscope slides and macroscopic charcoal, quantified from wet-sieved deposits, are the most commonly applied methods for paleofire interpretation of Quaternary sediments. This research explores the down-profile patterns across three paleofire proxies (refractory black carbon, microscopic and macroscopic charcoal) and potential paleofire interpretations from a sediment core dating to the last centuries from Speke Gulf, Lake Victoria, and a young soil profile from a kopje located in the surrounding watershed in Serengeti National Park, Tanzania. The results of three paleofire metrics show similar trends within each site, with a positive trend across all metrics and increasing variability with increased measurement values (heteroscedastic). Notably, refractory black carbon (rBC) concentrations are two orders of magnitude higher in lake sediment samples compared to soil samples. rBC is positively correlated with both microscopic and macroscopic charcoal values and the overall profile patterns down the sediment core are similar, with the exception of the rBC increases from 2.5 to 0 cm depth that may result from increased fossil fuel combustion. The Speke Gulf rBC measurements are in an intermediate range between those published from glacial ice and other lake sediments. New rBC records from different ecosystems and temporal scales will provide paleofire insights and potential to interpret source areas and depositional patterns. The exploration of soil archives offers the potential to exploit semi-arid ecosystems and archaeological sites that have no nearby traditional paleoenvironmental study site targets. (Courtney Mustaphi et al., 2025a)
Keywords
Figure 1. Inset map of eastern Africa and study area (red square) with sample collection locations (sediment core, yellow circle; soil pit, yellow square) and main river catchments that flow to Speke Gulf, Lake Victoria. Heuristic diagram summary of the nested but different theoretical spatial scales (horizontal axis) of paleofire proxy indicators (greyscale boxes) and associated pyrogenic particle sizes and measurement techniques (vertical axis; Conedera et al., 2009; adapted from Brugger et al., 2022; Kehrwald et al., 2016). Basemap: Google Earth, 2023. http://earth.google.com/web/. (Courtney Mustaphi et al., 2025a and see Courtney Mustaphi et al., 2025b).
Embedded Figure below: Get the KML file for Google Earth.
Figure 2. Paleofire proxy results (concentrations) of macroscopic (>100 μm) and microscopic charcoal (>10 μm) and refractory black carbon (rBC) down the lake sediment core and soil pit profile. Arranged from larger particle sizes (left) to smaller (right). The uppermost five samples are highlighted (blue) to refer to Figures 3 and 4. Samples extracted for gamma counts are shown in vertical grey boxes (left). CRS age model estimates (year AD) are shown as secondary y-axis for the sediment core SPK7 (see Figure S1). Gamma counts for the four soil samples of NGO1 are shown in Figure S2). (Courtney Mustaphi et al., 2025a).
Supplemental Figure S1. Gamma counts with 95% CI uncertainty estimates of 2-cm intervals of dried sediments and high resolution (contiguous 0.5-cm intervals) co-located paleofire proxies measured in the Speke Gulf (SPK7) sediment core: macroscopic charcoal, microscopic charcoal, refractory black carbon (rBC). The sediments have a wet sediment density to dry sediment weight ratio between 6–8 and were aggregated to 2-cm thicknesses for gamma counting (left side, plotted by sample top depth) (Courtney Mustaphi et al., 2025b). The 210Pb background was reached below 14–16 cm. Peak 137Cs was detected at 6–8 cm depth and represented the peak atomic bomb testing around the year 1963. The downcore 210Pb profile suggests continuous sediment accumulation within 95% CI. The average sediment accumulation rate for core is 11 yr cm-1.
Supplemental Figure S2. Gamma counts of 137Cs (black bars) with 95% CI uncertainty estimates (grey bars) of 4 samples (2.5-cm thickness) and paleofire proxies measured in the Ngong Rock (NGO1) soil pit: macroscopic charcoal, microscopic charcoal, refractory black carbon (rBC).
Supplemental Figure S3. Scatterplots and Spearman’s rank correlation (rs) for the n=43 co-located paleofire proxies measured in the Speke Gulf (SPK7) sediment core and Ngong Rock soil pit (NGO1). Note that p-values are not presented here because of problems related to geologic time aggregation effects that lead to non-independence between samples in the stratigraphy.
References
Byrom AE, Nkwabi AJK, Metzger K, et al. (2015a) Anthropogenic stressors influence small mammal communities in tropical East African savanna at multiple spatial scales. Wildlife Research 42(2): 119–131.
Byrom AE, Ruscoe WA, Nkwabi AK, et al. (2015b) Small mammal diversity and population dynamics in the Greater Serengeti Ecosystem. In: Sinclair ARE, Metzger KL, Fryxell JM, et al. (eds) Serengeti IV: Sustaining Biodiversity in a Coupled Human–Natural System. Chicago: University of Chicago Press, pp.323–357.
Courtney Mustaphi CJ. 2019. Are colluvial soils on kopjes a potential geoarchive of palaeoenvironmental change in semi-arid regions? Twenty-Seventh Meeting of Swiss Sedimentologists SwissSed – University of Fribourg, 23 February 2019. pp. 22–23. [backup PDF]
Courtney Mustaphi CJ, Camara-Brugger SO, Chellman NJ, Muñoz S, Kariuki R, Damber M, Shoemaker A, Ekblom A, Munishi L, Lane P, Marchant R, Heiri O. 2025a. Multiple paleofire proxy metrics from tropical lake sediment and soil in the Greater Serengeti Ecosystem. The Holocene. [in press, CC BY 4.0 Open Access] Data: https://doi.org/10.7910/DVN/S4ILZT Supplemental information: https://journals.sagepub.com/doi/suppl/10.1177/09596836251340882/suppl_file/sj-docx-1-hol-10.1177_09596836251340882.docx
Courtney Mustaphi CJ, Camara-Brugger SO, Ekblom A, Munishi L, Kariuki R, Shoemaker A, Lane P, Marchant R. 2025b. Multiporate Poaceae pollen grains observed in the recent fossil record from the Greater Serengeti Ecosystem and Lake Victoria region. Review of Palaeobotany and Palynology 333: 105240. Data: https://doi.org/10.7910/DVN/JGZTIK
Courtney Mustaphi CJ, Platts PJ, Willcock S, Timberlake JR, Osborne J, Matimele H, Summer Osgood H, Muiruri V, Gehrels M, Bayliss J, Marchant R. 2025. Tree demographics and soil charcoal evidence of fire disturbances in an inaccessible forest atop the Mount Lico inselberg, Mozambique. Plants, People, Planet 7(1): 229–244. [CC BY 4.0 Open Access] Data: https://doi.org/10.7910/DVN/EZIWDT Video context: [YouTube]
Hengl, T., Heuvelink, G.B., Kempen, B., Leenaars, J.G., Walsh, M.G., Shepherd, K.D., Sila, A., MacMillan, R.A., de Jesus, J.M., Tamene, L. and Tondoh, J.E., 2015. Mapping soil properties of Africa at 250 m resolution: Random forests significantly improve current predictions. PloS one, 10(6), e0125814.
Jager, T. 1982. Soils of the Serengeti Woodlands, Tanzania. Agricultural Research Reports 912. Centre for Agricultural Publishing and Documentation Wageningen NL.
Mbise, FP. 2015. Factors affecting hyrax presence/absence and population size: A case study in Serengeti National Park, Tanzania. Unpublished master’s thesis. Norwegian University of Science and Technology.
McWethy, D.B., Neumann, F.H., Steinbruch, F., Ryan, C.M. and Valsecchi, V., 2016. Late Quaternary vegetation development and disturbance dynamics from a peatland on Mount Gorongosa, central Mozambique. Quaternary Science Reviews, 137, pp.221-233.
Mekonnen, B., Glaser, B., Zech, R., Zech, M., Schlütz, F., Bussert, R., Addis, A., Gil-Romera, G., Nemomissa, S., Bekele, T., & Bittner, L. (2022). Climate, vegetation and fire history during the past 18,000 years, recorded in high altitude lacustrine sediments on the Sanetti plateau, Bale Mountains (Ethiopia). Progress in Earth and Planetary Science, 9, 1–19.
Montade, V., Schüler, L., Hemp, A., Bremond, L., Salamanca Duarte, A. M., & Behling, H. (2018). Late Quaternary ecotone change between sub-alpine and montane forest zone on the leeward northern slope of Mt. Kilimanjaro. Journal of Vegetation Science, 29(3), 459–468.
Reed, D.N., Anderson, T.M., Dempewolf, J., Metzger, K. and Serneels, S., 2009. The spatial distribution of vegetation types in the Serengeti ecosystem: the influence of rainfall and topographic relief on vegetation patch characteristics. Journal of Biogeography, 36(4), pp.770-782.
Scoon RN. 2018. Geology of National Parks of Central/Southern Kenya and Northern Tanzania. Cham: Springer.
Sinclair ARE, Packer C, Mduma SAR, et al. (eds). 2008. Serengeti III: Human Impacts on Ecosystem Dynamics. Chicago, IL: University of Chicago Press.
Timbuka, C.D. and Kabigumila, J.D., 2006. Diversity and abundance of small mammals in the Serengeti Kopjes, Tanzania. Tanzania Journal of Science, 32(1), pp.1-12.
Tomlinson, R. W. (1974). Preliminary biogeographical studies on the Inyanga Mountains, Rhodesia. South African Geographical Journal, 56(1), 15–26.
**All Courtney-Mustaphi et al., publications listed above are data are Open Access CC BY 4.0 and with free access and reuse by the public, see here for requirements for use.