Spatial and temporal heterogeneity of inorganic soil nitrogen in a savanna: The role of riparian zone structure, hydrologic seasonality, and parent material

Background/Question/Methods

The Shingwedzi River drainage of Kruger National Park, South Africa, is a heterogeneous landscape characterized by contrasts in water availability, vegetation structure, and geology. Water resources are distributed heterogeneously in time as the system experiences distinct wet and dry seasons and ephemeral rivers. Riparian zones along these rivers are characterized by denser, taller vegetation relative to adjacent uplands. Within riparian zones, vegetation structure varies from closed canopy communities with grass and shrub understories, to communities dominated by grass and shrubs with few small trees. Rivers traverse granite and basalt parent materials, with the full range of vegetation structures found on each. 

We established 40 research plots to investigate the functional consequences of this temporal and spatial heterogeneity. Plots were stratified by riparian and upland landscape positions on both granite and basalt parent materials. Riparian plots were further stratified by vegetation structure classifications – continuous canopy, discontinuous, and sparse. There were five replicate plots for each treatment. We used mixed bed ion exchange resin bags in the rooting zone as an integrative relative measure of the flux of available inorganic soil nitrogen over three seasons: the dry season, the onset of the wet season, and the peak wet season.

Results/Conclusions

We found significant differences in available inorganic N across all spatial and temporal contrasts. Riparian sites exhibited 3x the available nitrate (NO₃-N) flux and 1.5x the available ammonium (NH₄-N) flux of upland sites, demonstrating a strong effect of landscape position on biogeochemical activity. Vegetation structure among riparian sites also strongly affected available soil NO₃-N; continuously canopied sites exhibited 2x the available NO₃-N flux of sparsely canopied sites. Available NH₄-N flux showed a similar pattern but with a lower magnitude of effect. The highest available NO₃-N and NH₄-N fluxes occurred during the early wet season across all contrasts, suggesting that ‘first flush’ microbial and nutrient dynamics are important in this system. The dry season exhibited the lowest available NO₃-N fluxes and the wet season had intermediate levels. Available NH₄-N flux, in contrast, was lowest in the wet season, due perhaps to high rates of microbial nitrification. Basalt sites showed 2x the available NO₃-N flux of granite sites, demonstrating that riparian soil N availability can be strongly affected by parent material, despite the high local heterogeneity and origins of alluvial soils. This work demonstrates the importance of temporal and spatial contrasts in understanding ecosystem function across heterogeneous savanna landscapes.