Net ammonification as influenced by plant diversity in experimental grasslands

Previous plant diversity experiments have mainly reported positive correlations between diversity and N mineralization. We tested whether this relationship can be explained by plant diversity-induced changes i) in the quantity or quality of organic matter or ii) in microclimatic conditions of central European grassland mixtures. We measured ex-situ net ammonification in a laboratory incubation of aboveground plant material and soil sampled in differently diverse plant mixtures. Secondly, in-situ net ammonification was assessed in a field incubation with mineralization cores containing standardized material in four treatments: soil only (control), and soil mixed with field-fresh plant tissue (grass, legume, or tall herb). We used 82 plots with varying species numbers (1, 2, 4, 8, 16, and 60) and numbers of functional groups (1–4; grasses, short herbs, tall herbs, and legumes). We determined the soil water content, total N concentrations of plant and soil, and NH₄–N release rates. In the ex-situ incubation under constant climatic conditions, functional group or plant species richness did not influence net ammonification rate constants (k) or the proportion of the organic N pool involved in ammonification (NH₄–N₀). The presence of legumes in plant mixtures significantly increased NH₄–N₀ and decreased k indicating elevated N leaching risks in legume-containing grassland mixtures. Mean in-situ net ammonification rates in the mineralization cores decreased in the following order: mixtures of soil with grasses (0.30 ± standard error 0.01 mg NH₄–N (g Nᵢₙᵢₜᵢₐₗ)⁻¹ d⁻¹) > tall herbs (0.25 ± 0.01) > legumes (0.22 ± 0.01) > control (0.07 ± 0.00). The type of incubated plant tissue also influenced the soil water content in the mineralization cores at the end of field incubation, likely because of different water retention capacities of the different plant tissue/soil mixtures. Significant plant functional group and species richness effects explained up to 13% of the variance of in-situ net ammonification rates. Because the effect of plant species richness disappeared if the type of incubated plant tissue and the soil water content were accounted for in a sequential ANCOVA, we infer that the soil water content was the main driver underlying the plant species richness effect.