Conceptualising morally permissible risk imposition without quantified individual risks.

We frequently engage in activities that impose a risk of serious harm on innocent others in order to realise trivial benefits for ourselves or third parties. Many moral theories tie the evidence-relative permissibility of engaging in such activities to the size of the risk that an individual agent imposes. I argue that we should move away from such a reliance on quantified individual risks when conceptualising morally permissible risk imposition. Under most circumstances of interest, a conscientious reasoner will identify a gap between the factors they deem potentially relevant to the riskiness of an agent's behaviour, and the factors they are reasonably able to quantify. This then leads a conscientious reasoner to conclude that they cannot, in good faith, come up with a quantitative risk estimate that is genuinely tailored to the agent's particular situation. Based on this, I argue that principles of morally permissible risk imposition fail to provide us with practical guidance if they ask us to take into account our agent-specific risks in a quantified manner. I also argue that principles of permissible risk imposition which appeal to quantified individual risks implausibly imply that it is frequently indeterminate whether engaging in some risky activity is morally permissible. For both of these reasons, I contend that principles of morally permissible risk imposition should make no reference to quantified individual risks. They should instead acknowledge that any quantitative estimates that an agent might usefully be able to consider will likely not be tailored to the agent's idiosyncratic situation.

Greenhouse gas fluxes over managed grasslands in Central Europe.

Central European grasslands are characterized by a wide range of different management practices in close geographical proximity. Site-specific management strategies strongly affect the biosphere-atmosphere exchange of the three greenhouse gases (GHG) carbon dioxide (CO2 ), nitrous oxide (N2 O), and methane (CH4 ). The evaluation of environmental impacts at site level is challenging, because most in situ measurements focus on the quantification of CO2 exchange, while long-term N2 O and CH4 flux measurements at ecosystem scale remain scarce. Here, we synthesized ecosystem CO2 , N2 O, and CH4 fluxes from 14 managed grassland sites, quantified by eddy covariance or chamber techniques. We found that grasslands were on average a CO2 sink (-1,783 to -91 g CO2  m-2  year-1 ), but a N2 O source (18-638 g CO2 -eq. m-2  year-1 ), and either a CH4 sink or source (-9 to 488 g CO2 -eq. m-2  year-1 ). The net GHG balance (NGB) of nine sites where measurements of all three GHGs were available was found between -2,761 and -58 g CO2 -eq. m-2  year-1 , with N2 O and CH4 emissions offsetting concurrent CO2 uptake by on average 21 ± 6% across sites. The only positive NGB was found for one site during a restoration year with ploughing. The predictive power of soil parameters for N2 O and CH4 fluxes was generally low and varied considerably within years. However, after site-specific data normalization, we identified environmental conditions that indicated enhanced GHG source/sink activity ("sweet spots") and gave a good prediction of normalized overall fluxes across sites. The application of animal slurry to grasslands increased N2 O and CH4 emissions. The N2 O-N emission factor across sites was 1.8 ± 0.5%, but varied considerably at site level among the years (0.1%-8.6%). Although grassland management led to increased N2 O and CH4 emissions, the CO2 sink strength was generally the most dominant component of the annual GHG budget.

No shift to a deeper water uptake depth in response to summer drought of two lowland and sub-alpine C3-grasslands in Switzerland.

Temperate C3-grasslands are of high agricultural and ecological importance in Central Europe. Plant growth and consequently grassland yields depend strongly on water supply during the growing season, which is projected to change in the future. We therefore investigated the effect of summer drought on the water uptake of an intensively managed lowland and an extensively managed sub-alpine grassland in Switzerland. Summer drought was simulated by using transparent shelters. Standing above- and belowground biomass was sampled during three growing seasons. Soil and plant xylem waters were analyzed for oxygen (and hydrogen) stable isotope ratios, and the depths of plant water uptake were estimated by two different approaches: (1) linear interpolation method and (2) Bayesian calibrated mixing model. Relative to the control, aboveground biomass was reduced under drought conditions. In contrast to our expectations, lowland grassland plants subjected to summer drought were more likely (43-68%) to rely on water in the topsoil (0-10 cm), whereas control plants relied less on the topsoil (4-37%) and shifted to deeper soil layers (20-35 cm) during the drought period (29-48%). Sub-alpine grassland plants did not differ significantly in uptake depth between drought and control plots during the drought period. Both approaches yielded similar results and showed that the drought treatment in the two grasslands did not induce a shift to deeper uptake depths, but rather continued or shifted water uptake to even more shallower soil depths. These findings illustrate the importance of shallow soil depths for plant performance under drought conditions.

Physiological response of Swiss ecosystems to 2018 drought across plant types and elevation.

Using five eddy covariance flux sites (two forests and three grasslands), we investigated ecosystem physiological responses to the 2018 drought across elevational gradients in Switzerland. Flux measurements showed that at lower elevation sites (below 1000 m.a.s.l.; grassland and mixed forest) annual ecosystem productivity (GPP) declined by approximately 20% compared to the previous 2 years (2016 and 2017), which led to a reduced annual net ecosystem productivity (NEP). At the high elevation sites, however, GPP increased by approximately 14% and as a result NEP increased in the alpine and montane grasslands, but not in the subalpine coniferous forest. There, increased ecosystem respiration led to a reduced annual NEP, despite increased GPP and lengthening of the growing period. Among all ecosystems, the coniferous forest showed the most pronounced negative stomatal response to atmospheric dryness (i.e. vapour pressure deficit, VPD) that resulted in a decline in surface conductance and an increased water-use efficiency during drought. While increased temperature enhanced the water-use efficiency of both forests, de-coupling of GPP from evapotranspiration at the low-elevation grassland site negatively affected water-use efficiency due to non-stomatal reductions in photosynthesis. Our results show that hot droughts (such as in 2018) lead to different responses across plants types, and thus ecosystems. Particularly grasslands at lower elevations are the most vulnerable ecosystems to negative impacts of future drought in Switzerland. This article is part of the theme issue 'Impacts of the 2018 severe drought and heatwave in Europe: from site to continental scale'.

Effects of plant productivity and species richness on the drought response of soil respiration in temperate grasslands.

Soil respiration plays a crucial role in global carbon cycling. While the response of soil respiration to abiotic drivers like soil temperature and moisture is fairly well understood, less is known about the effects of biotic drivers, such as plant above- and belowground productivity or plant diversity, and their interactions with abiotic drivers on soil respiration. Thus, current predictions of soil respiration to summer droughts might miss relevant biological drivers and their interactions with abiotic drivers. Since drought events are expected to increase in Central Europe in the future, we simulated early summer drought using rainout shelters at 19 grassland sites, which differed in plant productivity and species richness in central Germany in 2002 and 2003. We tested the potentially interacting effects of drought with biotic drivers, i.e. annual above-ground productivity, species richness and root biomass, on the drought response of soil respiration in temperate grasslands. In both years, drought led to a significant reduction in soil respiration. The drought-induced reduction in soil respiration was largely driven by the reduction in above-ground productivity in response to drought. The extent of the drought response of soil respiration was dependent on the species richness level of the site and this interacting effect was explainable by the variation in root biomass (root biomass and species richness were positively correlated). Our findings highlight the importance of biotic drivers for the quantification of the drought response of soil respiration in grasslands.

The effect of physical back-diffusion of 13CO2 tracer on the coupling between photosynthesis and soil CO2 efflux in grassland.

Pulse labelling experiments provide a common tool to study short-term processes in the plant-soil system and investigate below-ground carbon allocation as well as the coupling of soil CO(2) efflux to photosynthesis. During the first hours after pulse labelling, the measured isotopic signal of soil CO(2) efflux is a combination of both physical tracer diffusion into and out of the soil as well as biological tracer release via root and microbial respiration. Neglecting physical back-diffusion can lead to misinterpretation regarding time lags between photosynthesis and soil CO(2) efflux in grassland or any ecosystem type where the above-ground plant parts cannot be labelled in gas-tight chambers separated from the soil. We studied the effects of physical (13)CO(2) tracer back-diffusion in pulse labelling experiments in grassland, focusing on the isotopic signature of soil CO(2) efflux. Having accounted for back-diffusion, the estimated time lag for first tracer appearance in soil CO(2) efflux changed from 0 to 1.81±0.56 h (mean±SD) and the time lag for maximum tracer appearance from 2.67±0.39 to 9.63±3.32 h (mean±SD). Thus, time lags were considerably longer when physical tracer diffusion was considered. Using these time lags after accounting for physical back-diffusion, high nocturnal soil CO(2) efflux rates could be related to daytime rates of gross primary productivity (R(2)=0.84). Moreover, pronounced diurnal patterns in the δ(13)C of soil CO(2) efflux were found during the decline of the tracer over 3 weeks. Possible mechanisms include diurnal changes in the relative contributions of autotrophic and heterotrophic soil respiration as well as their respective δ(13)C values. Thus, after accounting for physical back-diffusion, we were able to quantify biological time lags in the coupling of photosynthesis and soil CO(2) efflux in grassland at the diurnal time scale.