Afforestation of formerly open, water-limited habitats is a globally widespread form of land conversion, driven by the need for timber and non-timber forest products, and capturing atmospheric carbon. However, increased tree cover can adversely affect hydrological cycles by increasing evapotranspiration and reducing downslope water yield, threatening agricultural production and water-dependent habitats. Here, we assessed the effect of afforestation with the non-native Robinia pseudoacacia in dry sandy upland areas of Hungary on the water supply of adjacent wetlands, and experimentally tested whether forest thinning can restore the hydrologic regime. We found that little moisture seeped down to the subsoil of plantations due to increased precipitation interception and transpiration, compared to grassy uplands. The lack of deep infiltration precluded lateral moisture seepage across the upland-wetland ecotone during the growing season. Conversely, we detected no notable interruption of lateral flows from grassy upland areas. Forest thinning (approx. 40% of the basal area) had little positive effect, probably because the lower interception and transpiration were offset by the disruption of the moisture-saving microclimate of the understorey and the quick recovery of the canopy cover. We conclude that the afforestation of upland areas is an important contributor to wetland loss in the study region and partial solutions are of limited effectiveness. We recommend a landscape-wide reconsideration of forestry strategies on uplands and planning alternative land use types to support soil moisture infiltration. One possible approach is to restore the original open ecosystems, but supporting solar developments in these marginal areas is also a viable option.

Sporobolus cryptandrus, a North American C4 grass, is a recent invader in some European sandy grasslands, particularly in Central Europe (Hungary), where it severely threatens native plant communities. As allelopathy has been documented for other Sporobolus species, we tested whether litter from S. cryptandrus has a different effect on the germination and seedling emergence of native grassland species compared to native grass litter. The germination and seedling growth of nine native grassland species and S. cryptandrus were examined in three treatments: without litter (control), with native litter and with S. cryptandrus litter. We hypothesized that: (i) litter has an overall negative effect on seedling germination and establishment compared to no litter; (ii) Sporobolus litter has a significantly more negative effect compared to native grass litter on seedling germination and establishment; and that (iii) the effect of litter type is highly species-specific. Our results showed that the presence of litter did not negatively affect germination and establishment across species, in fact, seedling length even increased in the presence of litter. Contrary to our expectation, Sporobolus litter negatively affected the germination and seedling growth only in the case of Bromus tectorum. For most of the other species, the effects of native and Sporobolus litter were highly similar, suggesting that litter influenced seedlings primarily through physical rather than chemical mechanisms, such as allelopathy. However, the clear suppression of B. tectorum by Sporobolus litter also suggests a potential allelopathic effect. According to our findings, litter effects are species-specific. Moreover, the suppression of B. tectorum by Sporobolus litter could offer novel insights for managing B. tectorum, a problematic invader in North America. Future research should explore the long-term effects of the litter of Sporobolus species in guiding restoration actions and invasive species management.

We selected 15 sheep-grazed sand pastures along increasing grazing intensity to study fine scale biomass patterns of main fractions (green biomass, litter) and that of plant species and functional groups (life forms and social behaviour types). We classified them into five grazing intensity levels based on stocking rate, proximity to drinking and resting places, and the number of droppings and other tracks of grazing animals. We formulated three study questions: i) How does increasing intensity of sheep grazing affect the amount of green biomass, species richness and their relationship in sand pastures? ii) How does increasing intensity of sheep grazing affect the biomass of perennial and short-lived graminoids and forbs? iii) How does disturbance value (expressed in the biomass ratio of disturbance tolerant and ruderal species) change along the gradient of grazing intensity? A unimodal relationship between green biomass and species richness was detected. Along the grazing intensity gradient, we found an increasing trend for species richness; significant differences for green biomass (decreasing trend) and litter (decreasing trend), moreover for graminoids (decreasing trend), and short-lived forbs (increasing trend). We found an increasing amount of disturbance tolerant and ruderal species with increasing grazing intensity. We concluded that stocking rate and proximity to drinking and resting places jointly affected vegetation and created an uneven pattern for composition and amount of biomass in all grazing intensity levels. Our findings might be instructive for pastures in densely populated regions which are prone to the encroachment of disturbance-tolerant and ruderal species.

Global warming, elevated atmospheric CO 2 concentrations, and increased likeliness of extreme drought and wildfires in many regions will likely favour C4 grass species. We explored the effect of the encroachment of an invasive perennial C4 grass, Sporobolus cryptandrus on the composition of soil seed banks in dry sand grasslands in Central Europe. In five mass-locality sites of the species we assessed the composition and vertical segmentation of the soil seed bank in twelve 1-m 2 plots along an increasing cover of the invasive species. We found that the seed bank diversity and density decrease with the increasing sampling depth; the decrease of density is affected by the increasing Sporobolus cover. Neither the diversity nor the seed bank density of other species were affected by increasing Sporobolus cover but, affected by the sampling site. Most of the studied seed bank characteristics were affected by the sampling depth, but none of them were affected by the increasing cover of Sporobolus. Increasing cover of Sporobolus in the vegetation was associated with an increasing proportion of Sporobolus seeds in the seed bank, and we found viable seeds of the species in the soil even in plots with no Sporobolus cover. Given that a decrease in the precipitation of the summer months and an increase in the frequency of droughts is projected in this region, we expect that the accumulated massive seed bank of Sporobolus will facilitate the further rapid spread of the species.

Understanding the responses of vegetation to grazing exclusion along a salinity gradient is useful for the management of grasslands. We studied the responses of vegetation to grazing removal (ungrazed areas) in three semiarid regions with different soil salinity levels: non-saline, moderately saline and hyper-saline. The results showed that Bray-Curtis dissimilarity between grazed and ungrazed areas were highest in the non-saline region. In the hyper-saline region, the grazing removal had no significant effect on any of the diversity indices, while in the non-saline regions, vegetation changes were occurred by grazing exclusion at the highest magnitude. Generally, the magnitude of vegetation changes by livestock grazing exclusion was decreased along the soil salinity gradient. This research could help to understand how disturbances and stresses interactively influence grasslands dynamics in semiarid regions and to understand the effects of grazing on grassland dynamics and sustainability in deserts in the context of salinization.