Effects of agricultural management and manipulated plant species composition of permanent grassland on productivity and sward structure | Auswirkungen von landwirtschaftlichem Management und manipulierter Pflanzenartkomposition auf Produktivität und Narbenstruktur im Dauergrünland

In den letzten Jahren wurde der Zusammenhang von Biodiversität und

The relationship between biodiversity and ecosystem functions of grasslands has received increasing attention in recent years. Scientists have sought to disentangle this complex relationship by sown biodiversity experiments and observational studies. However, both methods have their advantages and shortcomings. Hence, by using a complementary approach, namely a removal experiment involving manipulation of established vegetation, we might gain new and interesting insight into the effects of biodiversity on productivity and other ecosystem functions. In this study, a removal experiment was set up in the Solling Uplands in 2008 where a permanent grass sward was manipulated by herbicides to obtain either pure grass swards (dicot reduced) or swards with relatively high amounts of forbs and legumes (moncot reduced), on top of untreated control swards. Species numbers were larger than 10/m2 and thus covered the upper part of the usually examined logarithmic scale. They were comparable to average spechies richness on European grassland. The manipulated swards were then examined under different environmental conditions, i.e. we created a gradient in management intensity, regulated by both the cutting regime and the nutrient supply. Three main aspects were investigated. At first, we were interested to see how the herbicide application changed the grass swards and if we could create an appropriate biodiversity gradient on our experimental area. The regrowth of swards after herbicide application was also monitored, focusing on plant functional traits (growth height, reproductive strategy) of colonizer plants. The second aspect was the productivity and quality of the manipulated swards under different management treatments. We hypothesized that a more diverse sward would yield more and higher quality forage than one with a lower diversity in terms of species richness, functional diversity (proportions of grasses, forbs and legumes) or biodiversity indices. Finally, we analyzed whether the functional diversity, the amounts of graminoid and dicot species influenced the sward structure in terms of leaf area index and vertical biomass distribution. Compared to the baseline data before treatment implementation, the combination of treatments, especially the herbicide application, led to a broad variety of swards with the sward type alone accounting for more than 25% of all variance in composition. The functional composition of the sward changed slightly, since mainly vegetatively reproducing species were able to expand after herbicide application. In grass rich swards these were mainly the smaller grasses, leading to a slightly lower canopy. We conclude that this kind of sward manipulation was successful because the swards differed significantly in species number, evenness and composition of functional groups. Neither the biodiversity nor the distribution of functional groups influenced the yield in this experiment. However, in fertilized plots, larger amounts of tall grasses and tall species altogether increased yields significantly. The swards rich in legumes and forbs tended to have a better quality than the control and grass-rich swards. Consequently, part of the forage quality parameters showed correlations with biodiversity indices since these increased with increasing forb content. Although the yields did not differ among the swards, the sward architecture showed different vertical patterns between moncot reduced and dicot reduced swards. The latter also had converted more of the fertilized nutrients into leaf area. Species richness did not influence sward structure, the main source of variation lay – apart from the management – in the grass:dicot ratio. Compared with artificial biodiversity experiments, in our experiment, biodiversity only had moderate effects on examined ecosystem functions probably due to the shorter and more natural range of biodiversity levels realized in our experiment.