Shifts in plant community structure of a threatened sandy grassland over a 9-yr period under experimentally induced nutrient regimes: is there a lag phase?
Does nutrient addition to nutrient-poor pioneer grassland lead to altered successional pathways after a lag phase? Are there shifts in plant functional types with time after nutrient addition? Is phytodiversity negatively affected by a 9-yr nutrient addition? Is succession affected by local seed availability?
Upper Rhine valley, Germany.
A five-fold replicated randomized block design was used for addition of phosphorus, organic carbon, nitrogen (low and high dose) or combined applications of high-dose N with P (NP), potassium (NPK) or other essential nutrients (NPKM) for 9 yr. Seed limitation was assessed as local seed rain. Data were analysed by ordination (DCA) and linear mixed models.
DCA revealed two successional pathways: one typical for sandy grassland, and another on plots with high-dose N, which was accelerated and clearly separated from the typical one after a distinct ‘lag phase’ of about 5 yr. As a general trend, phytodiversity diminished on all plots during succession, but the decrease was significantly stronger on plots with high-dose N, which had higher turnover ratios. Habitat-typical species from pioneer stages (‘stress tolerators’ and ‘ruderals’) and Red Data species decreased with nutrient addition. There was an increase in cover of tall plants, geophytes and hemicryptophytes, ‘competitor/stress tolerator/ruderal’ strategists and competitive graminoids on plots with high-dose N. Above-ground phytomass production of phanerogams increased three-fold, but was significantly lower for cryptogams. Litter accumulation increased five-fold on high-dose N plots. In contrast, low-dose N and P plots only showed responses for legumes, predominantly facilitated by P. Seed rain consisted of autochthonous and allochthonous species, but most abundant species were non-target species.
Community structure, plant strategies and general successional trajectories of the studied system changed on plots with high-dose N after 5 yr. A threatened vegetation type was gradually replaced by one with competitive drivers of succession. These findings emphasize the importance of long-term observations for studying nutrient effects.