Early cultivation and bioturbation cause high long-term soil erosion rates in tropical forests: OSL based evidence from Ghana Henrik Breuning-Madsen a, ⁎, Jeppe Ågård Kristensen a , Theodore W. Awadzi b , Andrew S. Murray c a Institute of Geosciences and Natural Resource Management, University of Copenhagen, Denmark b Department of Geography and Resource Development, University of Ghana, Legon, Ghana c Nordic Luminescence Laboratory, Aarhus University, Risoe Campus, Roskilde, Denmark abstract article info Article history: Received 4 April 2016 Received in revised form 26 October 2016 Accepted 7 December 2016 Available online xxxx Soil erosion in forests is often considered negligible or minimal. In some tropical forest ecosystems, however, the combination of continuous surface exposure of fine soil particles at the upper slope by soil macrofauna and epi- sodic precipitation can cause considerable downslope soil transport, and the introduction of slash and burn agri- culture is often found to speed up soil erosion even more. Here we, for the first time, use OSL dating to quantify the long-term soil erosion rate along a hillslope in the moist semi-deciduous forest zone of Ghana. During the last ~4–5000 years about 100 cm of sediment has accumulated at an approximately constant rate of ~0.13– 0.21 mm year -1 (1.95–3.15 t ha -1 year -1 ). This is about an order of magnitude higher than literature estimates from similar environments, primarily obtained using short-term experimental setups; this suggests that care should be taken when extrapolating short-term erosion estimates to centennial or millennial timescales. We fur- ther deduce that an apparent acceleration in erosion began ~ 4500 years ago. This may be due to the introduction of cultivation in the area, slightly earlier than previously suggested (2800–3500 years ago). In this case the ero- sion rate based on deposition below 100 cm (~0.05 mm year -1 ) can be taken to represent the (natural) erosion rate without human influence; this value is consistent with literature estimates. More retrospective studies of soil erosion rates are needed in order to confirm the validity of our methodology and results, as well as to test the re- liability of extrapolated short-term erosion estimates. © 2016 Elsevier B.V. All rights reserved. Keywords: Bioturbation Soil erosion Early cultivation OSL dating Tropical semi-deciduous forest Ghana 1. Introduction Soil erosion is a main cause of soil degradation around the world, not least in the humid tropics (Labriere et al., 2015). It is typically assumed to be very small or even non-existent under dense natural forest vege- tation (Garcia-Ruiz et al., 2015). However, in some tropical forest eco- systems the continuous exposure of readily erodible sediment to the surface by termite and ant activity (Awadzi et al., 2004; Lobry de Bruyn and Conacher, 1990; Nye, 1955) has long been believed to cause considerable erosion from upper slopes to foot slopes (e.g. Ahn, 1970; Nye, 1954). However, only relatively few estimates of long-term forest soil erosion rates exists, the great majority on short timescales (b 3 years, Garcia-Ruiz et al., 2015; Labriere et al., 2015). A recent soil erosion meta-analysis (Garcia-Ruiz et al., 2015) shows that there is a correlation between the estimated mean annual erosion rates and the duration of the studies, indicating that extrapolation of short-term ex- periments tend to overestimate the long-term erosion rates. This analy- sis also showed that wet tropical environments are underrepresented in soil erosion studies. According to Labriere et al. (2015) meta-analysis of soil erosion studies in the humid tropics, the majority of the soil erosion studies in forests last 3–5 years. Thus, there is a strong need for long- term estimates of soil erosion to validate the reliability of extrapolation of short-term results. Optically stimulated luminescence (OSL) dating is increasingly applied to study past landscape response to climate and land use change by dating foot-slope deposits (reviewed in Fuchs and Lang, 2009). It integrates soil erosion over hundreds to thousands of years, and can thus provide a real long-term estimate. Land use change due to human activity, primarily crop cultivation, is known to increase soil erosion substantially (e.g. Fuchs and Lang, 2009; Labriere et al. 2015). Unfortunately there is little firm evidence constraining the introduction of crop cultivation in West Africa. Adekola (2011) concludes that yam and oil palms were likely the first crops to be cultivated in forested areas of West Africa, as these are native to the region and still are some of the most important crops for domestic purposes. He suggests that the occurrence of stone axes from about 5000 years BCE indicates use of the oil palm as a crop. Nonetheless, most other botanical and archaeological evidence suggest that the major introduction of slash and burn cultivation was not until ~3– 4000 years ago (Adekola, 2011; Sowunmi, 1985, Stahl, 1993). Thus it is not clear when, or even if, human use of the landscape began to mod- ify erosion rates in tropical forest ecosystems in West Africa. The aim of this study is to quantify the deposition rate at a foot slope in the semi-deciduous tropical forest zone of Ghana. By dating sediment Catena 151 (2017) 130–136 ⁎ Corresponding author at: Oester Voldgade 10, DK 1350 Copenhagen K, Denmark. E-mail address: hbm@ign.ku.dk (H. Breuning-Madsen). http://dx.doi.org/10.1016/j.catena.2016.12.002 0341-8162/© 2016 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Catena journal homepage: www.elsevier.com/locate/catena