Field experiments were conducted to investigate the seed attributes, movements and fates of dispersal units, and seedling establishment of Chrysopogon aucheri and Cymbopogon jwarancusa in a representative grassland ecosystem in upland Balochistan, Pakistan. Cymbopogon jwarancusa had more filled and viable caryopses than Chrysopogon aucheri. Seeds (spikelets) of both species had similar morphological features. Chrysopogon aucheri had one dispersal unit, a triplet spikelet. Cymbopogon jwarancusa had four types of dispersal units: a paired spikelet, a partial raceme, an entire raceme, and a partial inflorescence comprised of two racemes. Paired spikelets and partial racemes of Cymbopogon jwarancusa had greater mean dispersal distances (94 and 101 cm) from the edge of the basal crown of marked plants to the ground surface than triplet spikelets of Chrysopogon aucheri (79 cm). Spikelets of Cymbopogon jwarancusa and Chrysopogon aucheri moved mean distances of 26 and 32 cm, respectively, on the ground surface before becoming trapped in a microhabitat. The mean angle of dispersal for both species was toward the northeast, according to the prevailing wind direction. An ant (Tica verona) was the only detected seed (spikelet) predator for Chrysopogon aucheri. Both species had a weakly persistent soil seed bank, with higher amounts of seeds found under plant canopies compared to open interspaces. The recruitment of Chrysopogon aucheri and Cymbopogon jwarancusa seedlings from the natural seed bank was monitored in seven different microhabitats under natural and above-normal precipitation regimes . Above-normal precipitation increased seedling recruitment for both species in all microhabitats. Cymbopogon jwarancusa had higher seedling densities than Chrysopogon auchfiri. Seedling survival and tiller development for both species were greatest in the gravel microhabitat in the natural precipitation treatment. Monsoon rains in late July enhanced emergence of both species from recently dispersed seeds but emerged seedlings did not survive to the end of the growing season. The field studies indicate that Cymbopogon jwarancusa has a greater regeneration potential than Chrysopogon aucheri in this grassland ecosystem in upland Balochistan. It may be difficult to increase the composition of Chrysopogon aucheri, the more desirable species in these grasslands, when using management techniques that rely on natural regeneration .

Context and objectives: The impacts of climate change on forests and the increasing demand for ecosystem services emphasize how important it is to adapt forests to changing and partially uncertain climatic and societal conditions. A challenge for active forest adaptation is to safeguard and manage regeneration under increasingly extreme environmental conditions while minimizing risks to the remaining forest stand. Options for such targeted regeneration management arise from fine adjusting gap-based silvicultural interventions. The overarching aim of this study was to analyze the establishment of beech regeneration (Fagus sylvatica L.) in gaps under the influence of climate change in order to derive recommendations for targeted regeneration management. Gap experiment: Methodologically, this study focused on the investigation of fenced gaps ranging from 28 to 550 m2 in size (n=24), felled during the winter of 2016/17, within four beech and oak (mixed) forest stands located in the Rostocker Heide in Northeast Germany. A detailed assessment of the environmental conditions within these gaps was conducted, along with an annual (2017/18/19) inventory of artificial and natural beech regeneration. Since this period included the extreme hotter drought events of 2018 and 2019, a connection between climate change, forest structure, and thus silvicultural interventions, and the resulting regeneration dynamics was established. Vitality, stature, and growth, as well as the precise position of each individual, were recorded as regeneration characteristics. In order to capture the environmental conditions to which each individual was exposed as differentiated as possible, the following data were collected: (1) Direct, diffuse, and total radiation (DIRSF, DIFFSF, and PHAR) were determined using hemispherical photography. (2) Water availability (BGR) was estimated based on inverted, modeled fine root biomass distributions of the remaining beech (BGRB) and oak (BGRO) trees. (3) Nutrient availability (SF) and groundwater influence (GW) were derived from forest site maps, and small-scale variations in terrain elevation were recorded. (4) Annual vegetation surveys were conducted. Substudy 1 - Survival probability: Survival is a crucial factor for the success of regeneration and can be substantially affected by extreme weather conditions during theestablishment phase. Hence, the goal of the first substudy was to better understand the survival of beech regeneration under the influence of severe hotter drought events within the context of the gap experiment and to derive silvicultural interventions to safeguard and manage the survival of beech. For this purpose, the potentially interactive effects of resource availabilities and (a-)biotic factors on the survival probability of artificially regenerated beech seedlings were systematically analyzed using logistic generalized linear mixed-effects models (GLMM). DIFFSF, BGRB, and SF exhibited positive individual as well as interactive effects on the probability of survival, indicating complementary resource use of beech seedlings. On the other hand, increased DIRSF led to a significantly reduced survival probability. This effect, attributed to water stress, was enhanced when BGRB was reduced and when DIRSF interacted with stronger GW influence and exposed micro-sites. The model predictions indicated that the survival probability of beech seedlings was substantially reduced in closed forests, small gaps, and northern compartments of larger gaps under drought conditions. Conversely, increased resource availabilities in larger gaps under limited exposure to direct radiation resulted in improved survival probability. Thus, creating gaps larger than 200 m2 and focusing regeneration efforts on suitable, e.g., southern, gap partitions, present options for adapting silvicultural interventions in the context of hotter drought periods. Substudy 2 – Relative height growth: Height growth, alongside survival, is another central factor determining the overall success of regeneration, and is significantly influenced by resource availabilities. The objective of the second substudy was to establish a sound relationship between silvicultural interventions and the height growth of beech seedlings under the influence of hotter drought within the context of the gap experiment. This allows for the identification of potential establishment bottlenecks and opportunities for managing beech regeneration. To achieve this, the potentially interactive effects of PHAR, BGR, and SF on the relative height growth of artificial and natural beech regeneration were analyzed using multiplicative non-linear mixed-effects models (NLME). By means of these interactive models, both the positive effects of PHAR, BGR and SF on height growth could be specified and complementary resource use patterns of beech regeneration could be confirmed. It was also found that a larger initial root collar diameter was associated with a significant increase in height growth. Natural regeneration outperformed the height growth of artificial regeneration under all conditions except for optimal resource availabilities. These findings highlight the importance of root development in relation to height growth and the advantage of natural regeneration under waterand nutrient limitations. Furthermore, BRGB had a stronger effect on the growth of regeneration compared to BRGO. Under the given site and weather conditions, both in closed forests and small gaps, relatively low height growth was observed. A recommendation for adapting silvicultural interventions to hotter drought periods is to create gaps larger than 200 m2. In these, and especially in central gap compartments, significantly increased height development was observed. Negative effects on height growth above a certain gap size were not found but should be considered. Conclusions: The presented results indicate an analogous decrease in shade tolerance of beech regeneration with the given water limitation. A high susceptibility of beech regeneration to drought stress was confirmed. The average environmental conditions in small gaps (below 100 m2) and closed forests did not allow for successful establishment of beech in the context of hotter drought periods. This calls into question both the regeneration potential in natural beech forests, given the small-scale disturbances prevalent here, and the suitability of corresponding close-to-nature regeneration methods. The optimal gap size for beech regeneration depends on site, climate, and weather conditions. Under the conditions covered by this study, it varied between 200 and 400 m2. The negative effect of poor site conditions and stronger GW influence on regeneration success could be partially compensated for in larger gaps, where, on average, light and water availability was increased. However, above a gap size of approx. 400 m2, increased DIRSF can lead to a decrease in establishment success. The models further allow for the spatial explicit prediction of beech regeneration establishment success within the studied gaps. While this method requires extensive input data, it substantially improves prediction accuracy, especially in the case of larger gaps. Based on these predictions, there is considerable optimization potential for the planning and implementation of regeneration methods. In the context of adapting forests to climate change, the methodological approach presented in this study is particularly relevant. It enables the precise analysis of regeneration processes of further tree species, the identification of potential establishment bottlenecks, and the derivation of silvicultural strategies. Prospectively, this allows for more precise fine-tuning of regeneration dynamics in species- and structurally-rich mixed forests under climate change conditions.