Coupled modelling and sampling approaches to assess the impacts of human water management on land-sea carbon transfer.

Land-sea riverine carbon transfer (LSRCT) is one of the key processes in the global carbon cycle. Although natural factors (e.g. climate, soil) influence LSRCT, human water management strategies have also been identified as a critical component. However, few systematic approaches quantifying the contribution of coupled natural and anthropogenic factors on LSRCT have been published. This study presents an integrated framework coupling hydrological modeling, field sampling and stable isotope analysis for the quantitative assessment of the impact of human water management practices (e.g. irrigation, dam construction) on LSRCT under different hydrological conditions. By applying this approach to the case study of the Nandu River, China, we find that carbon (C) concentrations originating from different land-uses (e.g. forest, cropland) are relatively stable and outlet C variations are mainly dominated by controlled runoff volumes rather than by input C concentrations. These results indicate that human water management practices are responsible for a reduction of ∼60% of riverine C at seasonal timescales, with an even greater reduction during drought conditions. Annual C discharges have been significantly reduced (e.g. 77 ±â€¯5% in 2015 and 39 ±â€¯11% in 2016) due to changes in human water extraction coupled with climate variation. In addition, isotope analysis also shows that C fluxes influenced by human activities (e.g. agriculture, aquaculture) could contribute the dominant particulate organic carbon under typical climatic conditions, as well as drought conditions. This research demonstrates the substantial effect that human water management practices have on the seasonal and annual fluxes of LSRCT, especially in such small basins. This work also shows the applicability of this integrated approach, using multiple tools to quantify the contribution of coupled anthropogenic and natural factors on LSRCT, and the general framework is believed to be feasible with limited modifications for larger basins in future research.

Leaf anatomical traits determine the 18 O enrichment of leaf water in coastal halophytes.

Anatomical adaptations to high-salinity environments in mangrove leaves may be recorded in leaf water isotopes. Recent studies observed lower 18 O enrichment (ΔL ) of leaf water with respect to source water in three mangrove species relative to adjacent freshwater trees, but the factors that govern this phenomenon remain unclear. To resolve this issue, we investigated leaf traits and ΔL in 15 species of true mangrove plants, 14 species of adjacent freshwater trees, and 4 species of semi-mangrove plants at five study sites along south-eastern coast of China. Our results confirm that ΔL was generally 3-4‰ lower for mangrove species than for adjacent freshwater or semi-mangrove species. We hypothesized that higher leaf water content (LWC) and lower leaf stomatal density (LS) both played important roles in reducing ΔL in mangroves relative to nearby freshwater plants. Both differences acted to elongate effective leaf mixing length (L) in mangroves by about 200% and lower stomatal conductance by about 30%. Péclet models based on both LWC and LS could accurately predict ΔL . Our findings highlight the potential species-specific anatomical determinants of ΔL (or L), which has important implications for the interpretation of environmental information from metabolites produced by leaf water isotopes in palaeoclimate research.