Limited long-term effects of moderate livestock grazing and prescribed early fire on soil and root processes in Sudanian savanna-woodlands, West Africa

Savanna ecosystems are mainly shaped by climate, soil properties and ecological disturbances such as fire and grazing. The biophysical effects of fire and grazing depend both on their intensities and season of occurrence. The long-term effects of fire and grazing on key soil parameters are still largely unknown in West African savanna ecosystems; yet the knowledge of these effects is crucial to our understanding and sustainable management of savanna ecosystems. This study provides scientific evidence of soil responses to both fire and grazing management, and their interactions, based on field-data from two case-studies. The data are drawn from a long-term factorial experiment that has been established since 1992 in two State Forest reserves in Burkina Faso that mainly differ in their soil attributes (i.e. shallow or deep soil). Root biomass, soil CO2 efflux, organic matter content and C/N ratio, and soil water properties were assessed on plots subjected to two different treatments (i.e. moderate grazing and prescribed burning) over a 20year period. The results indicated that at the site with shallow soil, for both root density and organic matter content there was no significant difference between the studied soil depths (i.e. 0–5cm and 5–10cm) nor a significant effect of treatment (both p>0.05). In contrast, at the site with deep soil, root density increases with soil depth (p=0.011); it is 5 times higher at depths of between 5 and 10cm (3.26±1.24g/dm3) compared to the top 5cm (0.67±0.20g/dm3). The organic matter content marginally decreases (p=0.048) with soil depth. The C/N ratio is higher (p<0.05) under an annual early fire treatment compared to the grazing treatments and control, particularly at the deep soil site. In addition, the CO2 efflux varies significantly with the treatment at the two sites (p<0.004); grazed plots have higher respiration rates than control and burnt plots. Finally, water infiltration rates do not vary significantly with fire and grazing after 20-years of treatment. Our results provide an insight into the long-term effects of fire and grazing on soil water infiltration and soil CO2 fluxes in savanna ecosystems. We conclude that effects of fire and grazing on savanna soil processes are limited. We also highlight our contrasting results between the studied soil attributes across the two sites, providing evidence that depending on site differences, the effects of fire and grazing on soil chemical and water properties might change, suggesting caution when making extrapolation in the context of savanna land use planning and management.