Climatic controls of ecohydrological responses in the highlands of northern Ethiopia.

Climate variability and recurrent droughts have a strong negative impact on agricultural production and hydrology in the highlands northern Ethiopia. Since the 1980s, numerous mitigation and land rehabilitation measures have been implemented by local and national authorities to reduce these impacts, are often poorly effective. As underlying reason may be that controlling relationships between climate and ecohydrology at medium-sized catchments (10-10,000km2) of semi-arid highlands are not well known. We investigated trends and relationships in precipitation, temperature, streamflow, and net primary productivity (NPP). The results were mixed, with both significant increasing and decreasing trends for temperature and streamflow. Precipitation time series did not show a significant trend for the majority of stations, both over the years and over each season, except for a few stations. A time series indicated a significant abrupt increase of NPP in annual, seasonal and monthly timescale. Cross-correlation and regression analysis indicate precipitation and maximum temperature were the dominant climatic variables in the Geba catchment for streamflow and NPP. In view of these results, also land use and land cover change over the past three decades was analysed as a possible factor of importance, as human intervention, may affect streamflow and NPP. Factors that mainly correlate with streamflow and NPP are precipitation and maximum temperature. Important interventions that appear beneficial for these responses are construction of micro-dams, soil and water conservation and ecological restoration measures. The awareness that interactions can be quite different in semi-arid and semi-humid regions, as well as in upstream and downstream areas, should be reflected in management aimed at sustainable water and land resources use.

Cadmium re-distribution from pod and root zones and accumulation by peanut (Arachis hypogaea L.).

Peanut (Arachis hypogaea L.) genotypes may differ greatly with regard to cadmium (Cd) accumulation, but the underlying mechanisms remain unclear. To determine the key factors that may contribute to Cd re-distribution and accumulation in peanut genotypes with different Cd accumulating patterns, a split-pot soil experiment was conducted with three common Chinese peanut cultivars (Fenghua-6, Huayu-20, and Huayu-23). The growth medium was separated into pod and root zones with varied Cd concentrations in each zone to determine the re-distribution of Cd after it is taken up via different routes. The peanut cultivars were divided into two groups based on Cd translocation efficiency as follows: (1) high internal Cd translocation efficiency cultivar (Fenghua-6) and (2) low internal Cd translocation efficiency cultivars (Huayu-20 and Huayu-23). Compared with Fenghua-6, low Cd translocation cultivars Huayu-20 and Huayu-23 showed higher biomass production, especially in stems and leaves, leading to dilution of metal concentrations. Results also showed that Cd concentration in roots increased significantly with increasing Cd concentrations in soils when Cd was applied in the root zone. However, there were no significant differences in the root Cd concentrations between different pod zone Cd treatments and the control, suggesting that root uptake, rather than pod uptake, is responsible for Cd accumulation in the roots of peanuts. Significant differences of Cd distribution were observed between pod and root zone Cd exposure treatments. The three peanut cultivars revealed higher kernel over total Cd fractions for pod than for root zone Cd exposure if only extra applied Cd was considered. This suggests that uptake through peg and pod shell might, at least partially, be responsible for the variation in Cd re-distribution and accumulation among peanut cultivars. Cd uptake by plants via two routes (i.e., via roots and via pegs and pods, respectively) and internal Cd translocation appear to be important mechanisms in determining Cd accumulation in the kernels of peanuts.

Heavy metal concentrations in soil and earthworms in a floodplain grassland.

We determined accumulated heavy metal concentrations (Cd, Pb, Cu, Zn) of earthworms in moderately contaminated floodplain soils. Both soil and mature earthworms were sampled before and after flooding and earthworm species were identified to understand species specific differences in bioconcentration. Accumulated metal concentrations in floodplain earthworms differed before and after flooding. Differences in uptake and elimination mechanisms, in food choice and living habitat of the different earthworm species and changes in speciation of the heavy metals are possible causes for this observation. Regression equations taken from literature, that relate metal accumulation by earthworms in floodplains as a function of metal concentration in soil, performed well when all species specific data were combined in an average accumulation, but did not address differences in accumulation between earthworm species.

Low-Resolution Modeling of Dense Drainage Networks in Confining Layers.

Groundwater-surface water (GW-SW) interaction in numerical groundwater flow models is generally simulated using a Cauchy boundary condition, which relates the flow between the surface water and the groundwater to the product of the head difference between the node and the surface water level, and a coefficient, often referred to as the "conductance." Previous studies have shown that in models with a low grid resolution, the resistance to GW-SW interaction below the surface water bed should often be accounted for in the parameterization of the conductance, in addition to the resistance across the surface water bed. Three conductance expressions that take this resistance into account were investigated: two that were presented by Mehl and Hill (2010) and the one that was presented by De Lange (1999). Their accuracy in low-resolution models regarding salt and water fluxes to a dense drainage network in a confined aquifer system was determined. For a wide range of hydrogeological conditions, the influence of (1) variable groundwater density; (2) vertical grid discretization; and (3) simulation of both ditches and tile drains in a single model cell was investigated. The results indicate that the conductance expression of De Lange (1999) should be used in similar hydrogeological conditions as considered in this paper, as it is better taking into account the resistance to flow below the surface water bed. For the cases that were considered, the influence of variable groundwater density and vertical grid discretization on the accuracy of the conductance expression of De Lange (1999) is small.

Infiltration and redistribution of LNAPL into unsaturated layered porous media.

Enhanced understanding of light non-aqueous phase liquid (LNAPL) infiltration into heterogeneous porous media is important for the effective design of remediation strategies. We used a 2-D experimental facility that allows for visual observation of LNAPL contours in order to study LNAPL redistribution in a layered porous medium. The layers are situated in the unsaturated zone near the watertable and they are inclined to be able to observe the effect of discontinuities in capillary forces and relative permeabilities. Two experiments were performed. The first experiment consisted of LNAPL infiltration into a fine sand matrix with a coarse sand layer, and the second experiment consisted of a coarse sand matrix and a fine sand layer. The numerical multi-phase flow model STOMP was validated with regard to the experimental results. This model is able to adequately reproduce the experimental LNAPL contours. Numerical sensitivity analysis was also performed. The capillarity contrast between sands was found to be the main controlling factor determining the final LNAPL distribution.

Transport and degradation of propylene glycol in the vadose zone: model development and sensitivity analysis.

Transport and degradation of de-icing chemical (containing propylene glycol, PG) in the vadose zone were studied with a lysimeter experiment and a model, in which transient water flow, kinetic degradation of PG and soil chemistry were combined. The lysimeter experiment indicated that aerobic as well as anaerobic degradation occurs in the vadose zone. Therefore, the model included both types of degradation, which was made possible by assuming advection-controlled (mobile) and diffusion-controlled (immobile) zones. In the mobile zone, oxygen can be transported by diffusion in the gas phase. The immobile zone is always water-saturated, and oxygen only diffuses slowly in the water phase. Therefore, the model is designed in a way that the redox potential can decrease when PG is degraded, and thus, anaerobic degradation can occur. In our model, manganese oxide (MnO2, which is present in the soil) and NO3- (applied to enhance biodegradation) can be used as electron acceptors for anaerobic degradation. The application of NO3- does not result in a lower leaching of PG nor in a slower depletion of MnO2. The thickness of the snowcover influences the leached fraction of PG, as with a high infiltration rate, transport is fast, there is less time for degradation and thus more PG will leach. The model showed that, in this soil, the effect of the water flow dominates over the effect of the degradation parameters on the leaching at a 1-m depth.