Impact of more persistent precipitation regimes on a temperate grassland

In the mid-latitudes, climate change is characterized by a shift towards more persistent precipitation regimes, i.e. longer periods of both drought and precipitation. The effect of such a shift on a grass-legume grassland was simulated in a 480-day field mesocosm experiment. Treatments of the gradient design differed in the length of alternating dry/wet periods of 1, 3, 6, 10, 15, 20, 30, and 60 days, either starting with a dry or a wet period, resulting in 16 different treatments. All treatments received the same total amount of water, i.e., all dry periods were alternated with wet periods of equivalent duration and all treatments finished having received an even number of periods. Each mesocosm was planted with 12 common grassland species varying in traits (grasses/forbs/legumes). 

Plant survival, diversity and aboveground biomass production were monitored regularly throughout the experiment. Microbial diversity was investigated after 60 and 120 days of experiment. Soil samples from the beginning and the end of the experiment were analyzed for root biomass, organic carbon, nitrogen, bulk density, soil water retention, pore size distribution, microbial biomass, basal respiration, nematodes, mesofauna and macrofauna. Hydrophobicity and infiltration was measured at the end of the experiment. We hypothesized that with more persistent weather, plant and soil biodiversity and functioning would become exponentially impaired and that this would occur at different tipping points for different ecosystem components.

Plant diversity decreased as expected with increased weather persistence, mostly due to loss of forbs and N-fixers, with a tipping point around a blocking duration of 20 days after the end of the first growing season (120 days). These initial diversity losses could be traced to the timing and intensity of the preceding dry periods. By the end of the experiment (480 days), species richness showed a more linear response pattern, suggesting disappearance of the initial timing & tipping point effects, yet this may in part be attributed to preclusion of non-native colonization throughout the experiment. Plant productivity first followed a similar but less steep decline, possibly because reduced water availability was partly compensated by greater nutrient supply rate observed after longer droughts. Later, productivity overall decreased especially in grasses and evened out possibly reflecting the cumulative nutrient depletion associated with previous harvests, nutrient leaching and/or microbial/plant immobilization or higher plant density. However, the negative response of the N-fixer/grass production ratio to weather persistence became even more pronounced. 

After 120 days, microbial biomass was affected negatively by the 60-day treatment with a tendency of fungal biomass and F:B ratio to peak under intermediate weather persistence. Bacterial alpha diversity reacted negatively to persistent weather after 60 days within dry start treatments with tipping point around 10 days, but the trend was opposite after 120 days in both wet and dry treatments. Fungal beta diversity (community dissimilarity) was also positively influenced by more persistent weather.

Furthermore we found that with longer drought, studied soils became increasingly hydrophobic and this reduced initial infiltration rates. We discuss how this effect could exacerbate drought stress as well as increase erosion risk in sloping grasslands.