Background and aims Plant litter quality and water availability both control decomposition. The interaction of both parameters was never studied. We used a grassland site, where litter of contrasting quality, i.e. green litter (fresh leaves; high quality) and brown litter (dead leaves, which underwent senescence but which are still attached to the plant; low quality), is returned to soil. Green and brown litter were exposed in the field under regular weather and drought conditions. The objective of this study was to evaluate the effect of drought on the decomposition of both litter types. Methods We incubated green and brown litter of three different grassland species (Lolium perenne, Festuca arundinacea and Dactylis glomerata) alone or as litter mixture (1/3 of each of the three grassland species) in litterbags for 28 weeks. Drought conditions were simulated by rainfall exclusion. After incubation, litter residues were analysed for C and nitrogen (N) content and stable isotope composition. Additionally, we determined the response of the lignin and carbohydrate signatures to the contrasting conditions. Results C decomposition kinetics of green and brown litter under drought conditions could be explained by two pools of contrasting turnover times. Drought decreased leaf litter C and N decomposition by more than 50% compared to regular weather conditions, mainly by strongly decreasing the decomposition rate constants. The lowest C decomposition occurred for mixtures of litter from all three grassland species. Brown litter showed on average 15% higher reduction in carbon decomposition than green litter following drought. Lignin content remained similar for green and brown litter after drought and regular weather conditions, while sugar content remained similar in green litter and decreased by 18% for brown litter under drought conditions. Conclusions Our results showed different response of decomposition of litter with contrasting quality to drought. Low quality brown litter is likely to be more affected than high quality green litter. Thus, litter quality must be taken into account, when assessing the effect of drought on decomposition.

Summary — The dynamics of dry-matter accumulation and nitrogen uptake of sorghum and maize crops were compared over two successive years under dry and irrigated conditions. In non-limiting situations for growth, with irrigation and high nitrogen fertilization, maize growth was higher than that of sorghum because the leaf area index of maize developed earlier, leading to a larger quantity of intercepted radiation. The efficiency of transforming intercepted energy into aerial biomass varied little between the two species. In situations where nitrogen was limiting, the capacity of sorghum crops to take up nitrogen from the soil was always higher than that of maize. This feature was not linked to the greater planting density of the sorghum crop (50 cm between the rows) compared with the maize crop (75 cm), but seemed to be a characteristic of the species. A careful study of the evolution of the nitrogen nutrition level of the different crops throughout their growth showed that sorghum was capable of taking up much larger quantities of nitrogen from the soil than maize before its nitrogen nutrition became limiting. Drought reduced nitrogen availability in the soil for both crops. Maize was more sensitive to this restriction in nitrogen nutrition, and in addition to the direct effect of a water deficit on dry-matter accumulation in maize there was also this indirect effect of a nitrogen-nutrition deficit. All the results obtained in this series of experiments showed that the potential production of sorghum was lower than that of maize, but that its adaptation to limiting growth conditions, drought and limited nitrogen input should enable this species to grow in extensive cropping conditions. Moreover, the higher N uptake/water consumption of sorghum compared to maize indicates that the risk of nitrate leaching in the winter following a sorghum crop should be lower than after maize.

The influence of litter quality on plant litter decomposition rates is a crucial aspect of the soils C cycle. In grassland ecosystems, leaf litter, which is not removed either by herbivores or by mowing, returns to soil after the senescence process (brown litter). In grassland managed by mowing, another significant proportion of litter returns to the soil before senescence through harvesting losses (green litter). We hypothesized that changes in leaf tissue quality due to the senescence process would lead to contrasting decomposition dynamics of brown litter compared to green litter. Our conceptual approach included the monitoring of decomposition of green (fresh leaves) and brown litter (dead leaves, still attached to the plant) of three different grassland species (Lolium perenne, Festuca arundinacea and Dactylis glomerata) during a 1 year field incubation. After 0, 2, 4, 20 and 44 weeks, we retrieved the litterbags and analysed the remaining material for carbon and nitrogen content and stable isotope composition. Additionally, we determined the lignin content and composition by CuO oxidation and the non-cellulosic neutral carbohydrate content and composition after TFA hydrolysis. As expected, green litter, being higher in N and soluble compounds, while showing a lower C:N ratio and lower lignin contents compared to brown litter, was degraded at a higher rate. Carbon decomposition kinetics suggests that both leaf litter types consist of two pools with contrasting turnover times. The size of the active pool was related to the initial content of soluble plant litter compounds and the size of the recalcitrant pool was related to the lignin to N ratio of initial plant material. More lignin was lost from green litter compared to brown litter. P-coumaryl-type lignin units were decomposed at a higher rate than vanillyl and syringyl units. Total non cellulosic polysaccharide content showed little changes for both litter types. In contrast, the ratios of hexoses/pentoses (C6/C5) and desoxy sugars/pentoses (desoxy/C5) increased during decomposition of green litter only. This is an indication for an increasing contribution of microbial derived compounds being consistant with the higher decomposition rate of this material. Our results showed that grassland management (grazing versus mowing) could influence soil carbon sequestration through different proportions of green and brown litter returned to soil.

Soil microbial communities undergo rapid shifts following modifications in environmental conditions. Although microbial diversity changes may alter soil functioning, the in situ temporal dynamics of microbial diversity is poorly documented. Here, we investigated the response of fungal and bacterial diversity to wheat straw input in a 12-months field experiment and explored whether this response depended on the soil management history (grassland vs. cropland). Seasonal climatic fluctuations had no effect on the diversity of soil communities. Contrastingly fungi and bacteria responded strongly to wheat regardless of the soil history. After straw incorporation, diversity decreased due to the temporary dominance of a subset of copiotrophic populations. While fungi responded as quickly as bacteria, the resilience of fungal diversity lasted much longer, indicating that the relative involvement of each community might change as decomposition progressed. Soil history did not affect the response patterns, but determined the identity of some of the populations stimulated. Most strikingly, the bacteria Burkholderia, Lysobacter and fungi Rhizopus, Fusarium were selectively stimulated. Given the ecological importance of these microbial groups as decomposers and/or plant pathogens, such regulation of the composition of microbial successions by soil history may have important consequences in terms of soil carbon turnover and crop health.

No abstract