Plant species affect key ecosystem processes like nutrient cycling and overall ecosystem productivity through their litter. The outcome of litter effects is largely determined by its decomposability, which directly effects soil properties. If litter remains standing or unshed (i.e. marcescent), its final decomposability can be increased by photodegradation of recalcitrant structures (like lignin). If the litter is immediately shed, its decomposability largely depends on its original nutrient content. Moreover, plant species may affect soil also through other, more direct effects. It is however unknown whether marcescent and immediately shed litters affect soil, and by that plants, differently, whether direct effects of plants on soil interact with those of marcescent and shed litters, and whether these interactions are consistent under different soil conditions.

We set up a pot experiment, where we tested the effects of originally marcescent and shed litters (both added on the soil surface of the pots) on three grassland species (Bromus erectus, Filipendula vulgaris, and Plantago media) in contrasting soils from long-term stable ancient grassland and grassland restored on arable land 20 years before. We also tested how litter types and plant species affect soil chemical properties and microbial community (characterised by PLFA markers).

Marcescent litter contained a lower amount of nutrients, but still increased plant biomass more than shed litter, although only for F. vulgaris (likely due to mobilisation of soil nutrients).

The effect of litter on soil chemical properties and microbial community was low. These were largely affected by the plant species growing in the pot. The effect of these species on the microbial community was stronger in the undisturbed soil of ancient grasslands, while plant species affected mainly chemical properties in disturbed soil of restored grasslands. B. erectus slowed down the decomposition of both litter types in restored grassland soil.

The effect of marcescent litter on living plants was significant but species-specific and depended on soil conditions. Marcescence seems to have a stronger effect on plants in disturbed soil, which indicates its importance for recovery of the ecosystem after disturbance.

We set up a pot experiment, where we tested the effects of originally marcescent and shed litters (both added on the soil surface of the pots) on three grassland species (Bromus erectus, Filipendula vulgaris, and Plantago media) in contrasting soils from long-term stable ancient grassland and grassland restored on arable land 20 years before. We also tested how litter types and plant species affect soil chemical properties and microbial community (characterised by PLFA markers). Material for the experiment was sampled in grasslands of the White Carpathian Mts. SE of Czechia.