Sediment records of Lake Kariba and Lake Itezhi-Tezhi
Lake Kariba and Lake Itezhi‐Tezhi are two man‐made hydropower reservoirs located in the Zambezi River Basin in southern central Africa. Different studies have shown that these reservoirs have negative influence on downstream ecosystems. The goal of this study was to describe the biogeochemical properties of these reservoirs by analysis of sediment records, to get a better understanding of the impact of these reservoirs on downstream transport of carbon and nutrients. This study thereby focused on differences in the proportion of terrestrial and lacustrine organic matter, spatial variation in the reservoirs, historic changes and trends and the potential of the reservoirs to trap particles behind their walls.
Visual, physical and chemical analysis of samples from sediment cores collected along transects from the two main inflows to the dams revealed high spatial variation in the biogeochemistry of oth reservoirs. For most of the analyzed parameters, gradients from the inflows to the dam were found:
• Light beige colored se diments change to dark colored sediments.
• Sediments dominated by quartz minerals with particles of the coarse silt to fine sand fraction are replaced by sediments dominated by organic matter and clay minerals in the clay to coarse silt fraction.
• Organic carbon content increases from 2 – 3% to 3 – 5% (Kariba) and from 3 – 5% to 4 – 7%(Itezhi‐Tezhi), similar to an increase of total nitrogen and total phosphorous concentrations.
• Atomic C:N ratios decrease from 16 – 18 to 10 – 14 (Kariba) and 12 – 13 to 11 – 12 (Itezhi‐Tezhi).
• Carbon isotope ratios in Lake Kariba decrease from around ‐20‰ to values ranging from ‐27‰ to ‐24‰, while they do not change significantly in Lake Itezhi‐Tezhi and range from ‐25‰ to ‐22‰. Nitrogen isotope ratios do not change either in both reservoirs and range from 0‰ to 5‰.
Overall, these trends indicate the importance of terrestrial organic matter settling near the inflows, in opposition to a higher portion of autochthonous organic matter being deposited closer to the dams. In addition to spatial differences high temporal variation was detected, supposedly caused by the event‐driven character of the sedimentation regime in both systems. The cores collected far from the inflows show irregularly distributed layers with very different properties. These differences derive from the temporal influence of floods and autochthonous production. Flood layers contain a higher proportion of terrestrial organic matter, their organic matter content is lower and they consist of coarser material. The flood layers therefore resemble the sediments found near the inflows. The differences between layers are smaller in Lake Itezhi‐Tezhi because 2 the terrestrial influence on sediment properties is generally higher in this reservoir, due to its smaller size.
Some of the cores comprise two halves with pre‐dam sediments below sediments deposited after dam closure. The transition zone between the two halves features abrupt changes in bulk density, carbon, nitrogen and phosphorus concentrations, C:N and carbon isotope ratios. Hence, these changes could roughly be dated to the dam closures (~1960 for Kariba, ~1977 for Itezhi‐Tezhi). Although all cores could not be dated using radio nuclides, for cores containing pre‐dam sediments average sedimentation rates and mass accumulation rates of solid matter, carbon, nitrogen and phosphorus could be calculated. For the remaining cores, only minimum values could be stated. In summary, sedimentation rates in Lake Kariba decrease from more than 0.56 – 0.70 cm/yr to 0.19 – 0.50 cm/yr along transects from inflows to the dam. Sedimentation rates in Lake Itezhi-Tezhi range from 0.28 cm/yr (at the dam) to more than 1.9 cm/yr (minimum value close to the major inflow). The data suggests that in both reservoirs a significant proportion of material is deposited near the river mouths. Using these estimated sedimentation rates and the resulting mass accumulation rates, the accumulated mass of solid matter, carbon, nitrogen, and phosphorous were calculated. The resulting values could serve as sink terms in preliminary mass balances, which will help in quantifying the trapping potential of the reservoirs.