Ecology of roadside plant communities

Roadside habitats nowadays contribute considerably to the amount of natural areas in the Netherlands. The ecological role of roadsides is recognized by the Dutch government and road authorities, and the interest in ecological management is growing. The study considers a cross-section of the more valuable plant communities occurring on Dutch road verges and investigates a large number of environmental factors. Aiming to provide an ecological framework that can be used as a reference for ecological management, the study presents an accurate synecological description of the studied syntaxa. An additional field experiment shows that large amounts of nutrients are lost from cuttings, already in the first few weeks after mowing. The effect of various hay removal delay times on the nutrient balance of the different communities is modelled. In most cases, cuttings should be removed within one week (or two at most) if the amounts of nutrients removed are to exceed the inputs through atmospheric deposition. Exploiting the unfertilized, semi-natural character of the sites and the wide range of conditions involved, the data are also used to detect and test general relationships. Investigating soil versus biomass relations it is shown that, over a wide environmental gradient, no simple relationship exists between the vegetation tissue concentration and soil availability of a nutrient (with the possible exception of K).Focussing on diversity, unimodal species richness relationships are identified for both maximum standing biomass and productivity, but these only explain a small part of the variation and are apparent only if soil factors are not considered. The "hump-shape" is more pronounced for standing biomass than for productivity, suggesting that competition for light is the key-factor. Particularly the numbers of rare and endangered species are curtailed strongly by high biomass values. Management affects species richness positively, but also this effect is apparent only if soil factors are not considered. To investigate the nature of Ellenberg indicator values, these are correlated to the measured soil and vegetation parameters. Soil pH is shown not to be adequately indicated by the so-called Ellenberg reaction values, but instead they properly reflect soil total calcium ( i.e. both exchangeable calcium and calcium in the form of carbonates). Ellenberg nitrogen values best indicate productivity. Ellenberg moisture values best indicate the average lowest soil moisture content, but they also provide an appropriate indication of the average annual groundwater level or alternatively the average spring groundwater level. Over the wide range of unfertilized conditions studied, in situ average annual N mineralization can be predicted well from a few directly measured soil parameters. The pool of mineral N just before the growing season is the best correlate. Together with moisture content and pH, this variable can explain 83.5% of the variation in annual N mineralization.Key-words:diversity, ecological management, indicator values, infrastructure, nitrogen mineralization, nutrient availability, productivity, standing biomass, vegetation.