The response of grassland species to nitrate versus ammonium coincides with their pH optima
Dominance and occurrence of plant species change distinctly along soil reaction gradients; this has been attributed to different chemicals that can be either toxic or nutritional factors, varying with soil pH. The availability of different chemical forms of nitrogen (N) is one such factor, but the extent to which this can explain species distribution is yet unclear. We asked (1) is it possible to correlate the extent of species' physiological preference for ammonium vs nitrate to species occurrence along the soil reaction gradient; (2) how strong is the explanatory power of species occurrence along the soil reaction gradient for this preference, i.e. how much variation in species preference is explained by Ellenberg indicator values (EIV) for soil reaction; and (3) does inter-specific competition result in any shifts of this preference?
Dry sandy grasslands in Southern Germany, with soil reactions ranging from acidic to calcareous.
We carried out a common garden experiment and a greenhouse experiment with species from Central European sandy grasslands. In addition, we re-analysed previously published data. Availability of ammonium and nitrate was manipulated in the presence and absence of interspecific competition, while other factors including soil pH were kept constant. Growth responses were used to calculate species preferences concerning chemical N forms.
Species habitat niches for soil reaction, represented by their EIV, were found to strongly correlate with species preferences for certain N forms, with acidophilous species preferring ammonium and calciphilous species preferring nitrate. This was the case in all experiments and treatments without species interaction, but surprisingly not for species mixture treatments that might have involved additional soil biota with impact on N cycling.
High importance must be attributed to species' physiological preferences for certain N forms, because these impact on species occurrence along soil reaction gradients. Considering the known importance of other restrictions, e.g. aluminium toxicity or iron nutrition, it is likely that these factors, together with N form preference, act in concert. Our finding that species interaction impacts on such interrelations in an unexpected manner poses a future challenge to devise multi-factorial experiments on species occurrence along soil reaction gradients.