Simulated effects of precipitation and nitrogen on Serengeti grassland productivity

Abstract

In the Serengeti National Park, Tanzania, precipitation and soil nitrogen vary greatly between northwestern tallgrass areas and southeastern shortgrass areas, with the tallgrass having higher total precipitation and lower soil fertility. We used a model of grassland productivity, carbon/nitrogen cycling, and abiotic factors to test the hypothesis that tallgrass productivity is limited primarily by nitrogen availability while shortgrass productivity is limited by water. Under observed grazing intensities and ungrazed conditions, precipitation exerted primary control over grassland productivity for both regions, with differences in soil texture mediating soil water availability to the grasses. Mineral nitrogen availability interacted with water availability to influence productivity at precipitation levels ⩾ 130% of the mean. Nitrogen mineralization and precipitation were positively related for each grassland type, however, nitrification varied both between grassland types and between grazed and ungrazed conditions. Combined mineralization and nitrification could not maintain soil mineral nitrogen levels in the face of plant nitrogen uptake stimulated by increased precipitation, thus providing the mechanism by which nitrogen becomes a secondary limiting factor for both grasslands. Model experiments indicated that the pattern of primary limitation by precipitation and secondary limitation by nitrogen was robust to model assumptions concerning ungulate deposition of urine and dung nitrogen to the soil.