Low water‐holding properties impose dry conditions on plants on serpentine soil. To test the hypothesis that leaf water relations are key plant characteristics to grow in serpentine soil, we compared these traits for four tree species (Quercus serrata, Clethra barbinervis, Magnolia obovata and Pieris japonica) growing in serpentine soil and brown forest (BF) soil. Despite a much lower soil moisture content, trees in the serpentine soil (S) plot showed similar predawn leaf water potential to trees in the BF plot in all species. Trees in the S plot showed higher drought tolerance or drought avoidance than trees in the BF plot. In the S plot, enhanced drought tolerance through osmoregulation was observed in Q. serrata and P. japonica, whereas decreased leaf capacitance was observed in C. barbinervis and Q. serrata. Decreased stomatal conductance, a typical way to avoid drought stress, was observed for serpentine C. barbinervis and M. obovata and midday leaf water potential was comparable between S and BF plots for C. barbinervis. Lamina dry mass per area was higher for C. barbinervis but lower for Q. serrata in the S plot. Wood density was higher in serpentine Q. serrata and M. obovata. On the other hand, all studied species showed lower leaf nitrogen concentration in the S plot. Our results suggest that (a) despite low water availability, the tress in the serpentine plot did not suffer drought stress, and (b) leaf water relations may be one of the traits that enabled trees to inhabit serpentine soil.

No abstract

CO2 efflux from the litter layer, litter heterotrophic respiration, is an important component of the forest carbon cycle. Litter heterotrophic respiration mediated by microorganisms varies in response to seasonal environmental changes, such as temperature and moisture. Here, we aimed to quantify seasonal variation in litter heterotrophic respiration and determine how the microbial biomass influences microbial activity and hence litter heterotrophic respiration in a warm temperate forest. We performed in situ high‐frequency measurements of litter heterotrophic respiration per unit area (R_area), which are able to capture CO2 pulses during rainfall, for over 2 years. Microbial activity, which is the CO2 efflux per unit weight (R_mass) considering the change in the amount of substrate, was calculated based on R_area. In parallel, we measured substrate‐induced respiration (SIR) each month, as an index of microbial biomass. We identified seasonal hysteresis in R_area, which was higher in spring (January to July) than in fall (August to December), despite the temperature being similar in both periods. Of interest, R_mass and SIR also showed similar seasonal hysteresis in relation to temperature. Additionally, potential microbial activity without the effect of temperature and moisture was positively related to SIR. This result indicates that seasonal hysteresis with temperature in microbial activity was driven by microbial biomass seasonality, and thus, it leads to seasonal hysteresis in litter heterotrophic respiration in relation to temperature. Our findings highlight the importance of not only the environmental factors and substrate depletion but also biotic factors for estimating heterotrophic respiration.

Serpentine ecosystems are characterized by soil with high heavy metal and low nutrient content, both of which are likely to influence the rate of leaf decomposition. Here, we report how leaf chemistry and the microbial community influence the leaf decomposition rate in a serpentine ecosystem. Fresh Clethra barbinervis and Quercus serrata leaves collected from serpentine and non‐serpentine sites in cool‐temperate forests were placed at both sites for several months, after which changes in leaf mass were determined. The N concentration in the initial Q. serrata leaves from the serpentine site was lower than in those from the non‐serpentine site. The Ni concentration in the initial C. barbinervis leaves from the serpentine site was higher than in those from the non‐serpentine site. The leaves from the serpentine site decomposed slower than those from non‐serpentine sites for both species. Although we expected that serpentine soil would have a negative influence on microbial decomposers, the Q. serrata leaves placed in the serpentine site decomposed faster than those placed in the non‐serpentine site, suggesting that the serpentine soil had no adverse effect on microbial decomposers. However, microbial respiration and biomass in an in vitro experiment were not high in the leaves decomposed on serpentine soil. Future work should reveal the mechanism behind this contradiction. Our results suggest that the change of leaf quality is key to understanding the leaf decomposition rate in a serpentine ecosystem.