Plant and soil nitrogen dynamics in mediterranean grasslands: a comparison of annual and perennial grasses.

The predominance of annual species in the rangelands of southwestern Spain is not due only to climatic factors but is also strongly influenced by grazing management. Manipulating the grazing system in an experimental plot gave a vegetation structure with patches of annual grasses (mainly Vulpia ssp. and Bromus hordeaceus) and patches of perennial grasses (mainly Phalaris aquatica). This vegetation change allowed us to test the hypothesis that life-cycle differences between annual and perennial grasses affect soil nitrogen availability and plant uptake. Nitrogen availability, measured by in situ incubation, and nitrogen uptake were measured through the growing period (October to June). Amounts of in situ mineralized nitrogen over the whole growth phase were more important for soils supporting perennials (37 ppm) than for soils supporting annuals (27 ppm). The difference between the mineral nitrogen produced in situ and the mineral nitrogen accumulated during the same time in the soil allowed an estimation of the maximum mineral nitrogen quantity which can be taken up by the vegetation during each incubation period. The quantities accumulated over the year were 47 and 38 ppm (or 103 and 83 kg/ha) for soils supporting perennials and annuals respectively. For the same period, amounts of nitrogen immobilized in biomass production were 90 and 70 kg/ha for perennials and annuals respectively. During the autumn, a large proportion of mineral nitrogen was leached from soils supporting annual plants which had only just commenced germination. By contrast, the ability to use mineral nitrogen as soon as autumn rains occurred gave a competitive advantage to the perennial species, but only if they were protected from grazing during this period. The higher mineralization and use of this nitrogen reserve by perennials indicates that they made more efficient use of nitrogen resources than annuals, and validate the initial hypothesis.