Succession and disturbance in a successional gradient: impact of crucial abiotic and biotic factors

This thesis is focused on inland sand ecosystems, which are among the most threatened grassland types in Central Europe. These landscapes are highly fragmented and are inhabited by many endangered species in spatially isolated habitats. Recently, populations of the rare site-specific plant species decreased, the vegetation density increased and the typically open habitats became more and more ruderalised. As a reason, an enhanced nitrogen deposition and the abandonment of former extensively used grassland were discussed. Consequently, spontaneous succession can appear. To maintain community structure and population dynamics, different nature conservation management methods were introduced; in particular extensive sheep-grazing regimes became a suitable management tool in dry open habitats. The aim of this thesis was to elucidate the impact of crucial abiotic and biotic factors on succession in threatened inland sand ecosystems in a successional gradient. The main study site was the “Ehemaliger August-Euler-Flugplatz von Darmstadt“, a nature protection area in the northern Upper Rhine Valley of Germany. The characteristic plant communities of this site are considered highly threatened types in Central Europe according to the EU Habitats directive. The vegetation complex consists in the eastern part of early-successional stages of a pioneer Koelerion glaucae community, while in the western part mid-successional stages of Armerio-Festucetum trachyphyllae vegetation can be found. The substrates are primarily calcareous sands (Koelerion glaucae) with slight acidification of the topsoil (Armerio-Festucetum). The pre-grazing state in this system was characterised by a mowing regime during the military use and afterwards only by very extensive mowing and mulching. Sheep grazing was established in the area in the years 1999/2000. Nutrient increase, e.g. by atmospheric nitrogen deposition or transfer of soil particles, is a major threat for many types of endangered open vegetation. As a model system we used threatened sandy grassland, where a nutrient addition experiment in a five-fold replicated randomised block design was started in the year 2000. We analysed the whole data set for a nine-year period and ran extended approaches in the years 2007 to 2009. We added nitrogen in low dose (n) to simulate an enhanced atmospheric deposition, nitrogen in high dose (N), nitrogen in high dose in combination with different micro- or macronutrients (NP, NPK, NPKM), phosphorus (P), and a carbon source (C) to immobilise plant-available nitrogen to our experimental plots. Additionally, seed limitation was tested by assessing the local seed rain. The data set was analysed by means of detrended correspondence analysis (DCA) and linear mixed models. The DCA revealed two successional pathways: one line typical for threatened sandy grassland, and another one which was accelerated and clearly separated from the typical one. These separations were only detectable after a “lag phase” of about five years. As a general trend phytodiversity diminished on all plots, but the decrease was significantly stronger on plots with high-dose N, which have also shown higher turnover ratios. Site-specific species from pioneer stages (strategy type “stress-tolerators” and “ruderals”) and “Red Data” species decreased with nutrient addition. Additionally, we observed an increase in cover of tall plants, geo- and hemicryptophytes, “competitors/stress-tolerators/ruderals” and competitive grasses on plots with high-dose N. The above-ground phytomass production of phanerogams was up to 3-fold higher following N, NP, NPK or NPKM treatment, but it was significantly lower in the case of cryptogams. In contrast, low-dose nitrogen and phosphorus only had no impact on any species except legumes, which were facilitated predominantly by P. The extended approach has shown that most plants were responsive to nutrient addition and grew taller or set more fruits or seeds. In the local seed rain many autochthonous and allochthonous species were found. Anyhow, most abundant species were non-target species and even Calamagrostis epigejos caryopses were found. It could be shown by a field experiment that floristic structure, successional lines, phytodiversity and plant strategies of the studied system changed after a five-year period and a threatened vegetation type was gradually replaced by one with dominant and/or competitive species. These findings emphasise the importance of long-term observations for the study of effects caused by nutrients. I used a long-term field experiment lasting 10 years for further investigations concerning the effects of sheep grazing as a type of moderate disturbance. As “experiments by nature” happened during this time period, we were also able to study severe biotic (rabbit grazing) and abiotic (drought) disturbances, examining the recovery and resilience of the system after these severe disturbances. Within a six-fold replicated randomised split-plot design, 25-m² plots of Armerio-Festucetum trachyphyllae vegetation were studied from 2000-2009 in the northern upper Rhine valley (Germany). We analysed sheep-grazed and non-grazed plots (both with very low rabbit impact until the first half of 2005). In 2005 we established an additional plot type without rabbit and sheep grazing. Severe rabbit impact was studied in detail from 2006 to 2008 concerning flower resources, seed production, endozoochorous seed dispersal and phytomass extraction. Data were analysed by ordination methods and mixed linear models. Moderate disturbance by sheep grazing maintained species diversity in times of low rabbit densities. Severe drought in 2003 resulted in a marked decrease of species numbers, but the system recovered already in 2004. The subsequently analysed “high rabbit impact” resulted in decreases of species numbers, cover of bryophytes, flower quantities and seed content in faeces. Species numbers also declined on plots without rabbit grazing, but for different reasons. Shifts in phenological phases due to rabbit grazing were not observed, but the sum of flowers and number of flowering species decreased. After the decline of the rabbit population, species numbers slightly increased in 2009. Non-grazed plots showed an increase of litter. Despite severe disturbances by rabbits or drought, the community structure was relatively stable, especially in the case of non-ruderalised plots, and showed only slight changes. We developed a conceptual model showing the high potential of recovery and resilience even in the case of severe disturbances. Besides direct grazing effects like the creation of gaps by extensive livestock grazing, previous studies have shown benefits of seed dispersal (endo-, epizoochory), but little was known about post-dispersal processes such as trampling effects. We investigated the role of sheep trampling for the post-dispersal fate of dung-embedded seeds in sheep and rabbit dung and hypothesised a positive impact for the establishment of seedlings as a consequence of cracking the dung pellets. Sheep and rabbit dung was collected from threatened sandy grassland, and its seed potential was tested in a climate room. In a factorial field experiment we assessed the effects of trampling (using sheep-hoof replicas) and dung type on seedling emergence and fruiting success. For both dung types seedling emergence in the field was only 5 or 7 % of the potential without trampling but 18 or 14 % with trampling. Plots with trampled sheep or rabbit dung both showed significantly more seedlings (3.6- or 2.1-fold), more species (2.4- or 1.9-fold) and more fruiting seedlings (3.9- or 2.6-fold) and fruiting species (2.1- or 1.9-fold) compared to non-trampled dung plots. However, not only target species profited from trampling, but also non-target species. This is disadvantageous regarding nature conservation aims, but the proportion of target species was clearly increased by trampling. Sheep play a multifaceted role in dispersal processes: after endozoochoric transport they act as a sort of “gardener” not only for sheep-dispersed seeds, but also for those dispersed by rabbits