Search

To understand the resilience of African savannas to global change, quantitative information on the long-term dynamics of vegetation is required. Past dynamics can be reconstructed with the REVEALS model, which requires pollen productivity estimates (PPE) that are calibrated using surface pollen and vegetation data. Here we calculated PPE values for five savanna taxa using the extended R-value (ERV) model and two pollen dispersal options: the Gaussian plume model (GPM) and the Lagrangian stochastic model (LSM). The ERV calculations failed to produce a reliable PPE for Poaceae. We therefore used Combretaceae as the reference taxon – although values obtained with Poaceae as the reference taxon are presented in the supplement. Our results indicate that Combretaceae is the taxon with the highest pollen productivity and Grewia the taxon with the lowest productivity. Acacia and Dichrostachys are intermediate pollen producers. We find no clear indication of whether the GPM PPEs or the LSM PPEs are more realistic, but the differences between these values confirmed that the pollen fall speed has a greater effect in the modelling of GPM than in the LSM. We also applied REVEALS to the pollen record of Lake Otjikoto (northern Namibia) and obtained the first quantitative reconstruction of the last 130 years of vegetation history in the region. Cover estimates for Poaceae indicate the predominance of a semi-open landscape throughout the 20th century, while cover values below 50% since the 21st century correspond to a thick savanna. This change in grass cover is associated with the spread of Vachellia, Senegalia and Grewia reflecting an encroached state.

Volcanic and sedimentary rocks of the Sinclair Supergroup occur in the Konkiep Terrane of Southern Namibia. Three volcanic and sedimentary cycles are recognised. In this work we describe and date volcanic rocks of the Barby Formation, a key unit in the Sinclair area. The coeval Spes Bona Syenite and the Tiras Granite Gneiss are also described and dated. The rock types in the Barby Formation are rhyolites, basaltic trachyandesites, trachybasalts and trachydacites as well as volcanoclastic rocks. The rocks are largely undeformed and partly altered by deuteric and contact metamorphic processes but not regionally metamorphosed. Our samples represent both the calc-alkaline and alkaline trends documented in previous work. U-Pb ion probe and laser ablation inductively coupled plasma (LA-ICP) multicollector mass spectrometer Lu-Hf microbeam analyses were made of zircon and baddeleyite grains from four samples. A felsic tuff sample from the base of the Barby Formation has a 207Pb/206Pb zircon age of 1214 ± 5 Ma (2σ). A rhomb porphyry sample from the top of an 8.5 km-thick stratigraphic section gives a 207Pb/206Pb baddeleyite age of 1217 ± 2 Ma. The Spes Bona Syenite which intrudes the top of the Barby Formation has a 207Pb/206Pb baddeleyite age of 1217 ± 3 Ma and an indistinguishable LA-ICP collision cell mass spectrometer Rb-Sr biotite isochron age of 1238 ± 20 Ma, showing that there was no >350°C regional metamorphic event. Multi-element diagrams for the calc-alkaline samples show a dominant signature of reworked crust which is superimposed on a possible subduction signature. However the alkaline samples contain clear subduction signatures which are not seen in the underlying 1.37 Ga Kumbis rhyolite. The Barby Formation samples and coeval Spes Bona Syenite have Lu-Hf crustal residence ages between 1682 and 1873 Ma, suggesting that both of these units formed from a mixture of juvenile mantle-derived and older crustal material. The Barby Formation is considered to have originated due to a subduction event which took place during the assembly of the Rodinia supercontinent. The duration of the Barby magmatic episode is constrained to a maximum 9 m.y. period between 1219 and 1210 Ma, and during this period the Konkiep Terrane was an active continental margin. The 1204 ± 9 Ma Tiras Granite Gneiss is slightly younger than the Barby Formation and intruded across the Lord Hills Shear Zone, which is the suture between the hardly metamorphosed Konkiep Terrane and the highly metamorphosed Grunau Terrane of the Namaqua-Natal Province. Its intrusion reflects the end of subduction-related volcanism, due to the collision of Namaqua terranes with the Konkiep Terrane.

The scarcity of freshwater is a growing concern around the world, especially in the developing world and is increasingly threatening human survival in less affluent parts of the world. In arid and semiarid regions, freshwater resources are imperative for the survival of populations as well as development activities in urban centres. Preceding assessments for potential solutions have underestimated the scarcity of water by failing to capture main trends of precipitation and the extent of drought periods. This paper aims to investigate rainfall trends for the City of Windhoek, Namibia. A century scale is examined by reconstructing historical precipitation records, and future trends are subsequently estimated and interpolated using Ordinary Kriging in ArcGIS 10.3. Kriging responds and handles spatial correlation much better for observed data points. The results show locations that receive high rainfall and potential periods for harvesting rainwater. The findings further show that the periods of high rainfall are shorter, and after each high rainfall stretch, a long period of lower rainfall follows immediately. Each low rainfall period is anticipated to last longer than the previous one. These scenarios are highly significant for urban managers in dryland cities such as Windhoek, as they demonstrate when water would need to be extra sparingly managed. This study highlights the need for better water management strategies, if cities in dry climates are to survive the anticipated future water crises.

The dynamics of Cu and Pb were studied in eight soil profiles (Luvisols, Chernozems), located at various distances from the tailing dams near a closed Cu-Pb-(Zn)-Ag mine at Kombat (Namibia) by a combination of isotope and elemental studies, leaching techniques and speciation modelling. Tailings, soils and bedrocks were analyzed for the bulk Cu and Pb concentrations, for their chemical forms using the sequential extraction procedure and for the isotope compositions of Cu (δ65Cu) and Pb (206Pb/207Pb, 208Pb/206Pb). In spite of the differences in the contamination of the studied profiles by tailing dust, both soil types exhibit characteristic trends in the behaviour of metals and distribution of their isotopes. While most Luvisols exhibit an increase in the metal content towards depth in the profile, Chernozems exhibit the opposite trend (decrease in metal content with depth) or similar contents in the whole profile. Luvisols were mostly characterized by changing δ65Cu values as a function of depth. The value of δ65Cu decreases with depth to the Bt horizon and then the δ65Cu value increases with increasing depth. Chernozems mostly yielded uniform Cu concentrations and isotope compositions in all the soil horizons, clearly contrasting with the isotopically lighter compositions of the carbonate bedrocks. Positive δ65Cu values in the tailing material affect the surface of the soil, especially soil sampled in its vicinity. In most Luvisols and Chernozems, the isotope composition of Pb exhibits similar patterns with a low 206Pb/207Pb ratio on the surface of the profile and with a slight increase to depth. The low 206Pb/207Pb values at the surface of the profiles are caused by dust from the tailings. The elemental and isotopic patterns of both contaminated and uncontaminated soil types reflect not only the distance from the tailing dam but also pedogenetic processes in the profiles and geochemical behaviour of both contaminants. The copper elemental and isotope patterns reflect the formation of more mobile species in the surface environment and movement of solutions and colloids and precipitation onto secondary minerals in Luvisols. Plant uptake of lighter isotopes in Chernozems can influence the Cu isotope patterns in these soil types. The limited movement of colloids and higher sorption capacity of Chernozems is manifested in the isotope compositions of Cu and Pb in their profiles.

Global Circulation Models (GCMs) are used to forecast climate change in Southern Africa, and the evidence shows that the region is going to warm up by up to 2° by the year 2050. Namibia is one of the driest countries in Southern Africa and is at a high risk of becoming much drier than current situation by 57%. Very few studies have been carried out in Southern Africa to show actual impacts of climate change. Practical applicability of GCMs at a local spatial scale remains limited due to the coarse nature of the models. Hence, improvement of the GCMs must begin with better understanding of the local microclimates and how they respond to regional circulation patterns. In many regions of Southern Africa, the lack of potential tools to access old climatic records precludes the estimation of climate trends beyond 100 years. In spite of these impediments, there are areas with excellent tree species such as Dichrostachys cinerea that are able to be used as climatic archives for specific time periods. In this chapter, the study shows that the combination of tree ring chronologies and precipitation records is a powerful methodology in climate modeling in the southern hemisphere and reveals nuances that show climate change. The evaluation of data from tree rings coupled with precipitation trends reveals signals that show that climate has indeed been changing over the past ten decades and will have a negative impact on livelihoods. These data can now be used in predictive models that can be used to project future scenarios and assist policy makers and planners to see how climate will evolve in the next 50–60 years.

We studied arid desert soils from Namibia (Rosh Pinah) that were contaminated with up to 7 mg kg−1 of thallium (Tl) via dust emitted from a local flotation tailing dam. Chemical extractions of waste and soil materials indicated that most of the Tl is strongly bound, in accordance with X-ray diffraction and X-ray absorption spectroscopy data that point to the predominant association of Tl with metal sulfides and phyllosilicates. The isotope fractionation factor e205Tl of the soil samples (from −0.4 to +3.8) shows a positive linear relationship (R2 = 0.62) with 1/Tl, indicative for the mixing of two major Tl pools, presumably anthropogenic Tl and geogenic Tl. The e205Tl value for the topmost soil samples (∼+3) closely matches the e205Tl value for post-flotation waste particles with a diameter of

The Tsumeb mining District (Namibia) is affected by prolonged exposure to multisource contamination related to the mining and smelting of copper ores. This study was performed to assess the extent of contamination of topsoil and grass shoots on the basis of the copper isotopic composition. Compared to an uncontaminated area (δ65Cusoil = −0.01 to +0.14‰; δ65Cugrass = −0.17 to +0.03‰), the soil and grass in the contaminated area are enriched in the heavy isotope (δ65Cusoil = +0.13 to +0.76‰ and δ65Cugrass = +0.15 to +0.64‰). The Δ65Cugrass-soil value exhibits small variations (+0.06 to −0.17) due to the variable degree of grass surface contamination, depending on the grass washing procedure used. The enrichment of soil and grass in the contaminated area in 65Cu reflects the Cu isotopic composition of old flotation tailings (δ65Cu = +0.95 to +1.1‰), slag (+0.11 to +0.55‰) and particles which originate from the smelting and flue gas cleaning processes (+0.15 to +0.49‰). The long history of mining and processing of copper ores in the Tsumeb area and the different isotopic composition of ore concentrates smelted in the local metallurgical plant do not allow assessment of whether the smelting process leads to isotope fractionation of copper. However, the isotopic composition of Cu in the sulfidic feed, slag and dust particles sampled over several hours showed that the dust collected from the dusthouse of the local smelter is isotopically lighter (δ65Cu = +0.15‰) relative to the isotopic composition of the charge (δ65Cu = +0.28 and +0.44‰). The results of this investigation demonstrate that the isotopic composition of Cu in soils and grasses can be used to assess the intensity of contamination in areas where Cu ores are mined and processed provided that the isotopic compositions of the ore processing products are different from the isotopic compositions of the uncontaminated soils and vegetation.

Understanding vegetation-modern pollen relationships is essential to provide confidence in fossil pollen reconstructions of long-term vegetation changes in savanna ecosystems. In this paper we compare the taxonomical composition and the diversity (Hill NO, N1, N2) of vegetation and modern pollen along precipitation and local grazing-intensity gradients in Namibian savannas. Modern pollen was extracted from surface soil samples collected from 5 x 5 m plots distributed along four 500 m gradients. Vegetation was surveyed in each plot. The results show a high correspondence between vegetation and pollen data in terms of composition. Precipitation and grazing explain a significant although low proportion of compositional change in the vegetation and pollen spectra. We identified pollen taxa as indicators of grazing pressure such as Limeum, Alternanthera, and particularly Tribulus. Correspondence between vegetation and pollen data in terms of taxa richness (NO) is limited, probably because of the influence of landscape heterogeneity and openness, as well as low pollen concentrations. In contrast, the effective numbers of common and dominant taxa (N1, N2) are consistent among the different datasets. We conclude that in spite of limitations, modern pollen assemblages can reflect changes in vegetation composition, richness and diversity patterns along precipitation and grazing gradients in savanna environments. (C) 2018 Elsevier B.V. All rights reserved.

The Naukluft Mountains in the Namib Desert are a high rainfall-high discharge area. It sees increased stream-, spring-flow as well as waterfalls during the rainy season. The mountains are a major resource for additional recharge to the Namib and Nama aquifers that are adjacent to the mountains. This paper aimed to highlight the potential vulnerability of the aquifers that surround the Naukluft Mountain area; if the strategic importance of the Naukluft Karst Aquifer (NKA) for bulk water supply becomes necessary. Chloride Mass Balance Method (CMBM) was applied to estimate rainfall available for recharge as well as actual recharge thereof. This was applied using chloride concentration in precipitation, borehole and spring samples collected from the study area. Groundwater flow patterns were mapped from hydraulic head values. A 2D digital elevation model was developed using Arc-GIS. Results highlighted the influence of the NKA on regional groundwater flow. This paper found that groundwater flow was controlled by structural dip and elevation. Groundwater was observed to flow predominantly from the NKA to the south west towards the Namib Aquifer in two distinct flow patterns that separate at the center of the NKA. A distinct groundwater divide was defined between the two flow patterns. A minor flow pattern from the northern parts of the NKA to the north east towards the Nama Aquifer was validated. Due to the substantial water losses, the NKA is not a typical karst aquifer. While the project area receives an average rainfall of 170.36 mm/a, it was estimated that 1–14.24% (maximum 24.43 mm/a) rainfall was available for recharge to the NKA. Actual recharge to the NKA was estimated to be less than 1–18.21% (maximum 4.45 mm/a) reflecting the vast losses incurred by the NKA via discharge. This paper concluded that groundwater resources of the NKA were potentially finite. The possibility of developing the aquifer for bulk water supply would therefore drastically lower recharge to surrounding aquifers that sustain local populations because all received rainfall will be utilized to maximise recharge to the NKA instead of surrounding aquifers.

This work investigates Mesoproterozoic volcanic rocks of the Haiber Flats, Barby, Welverdiend and Kairab Formations, and associated intrusive rocks in the Konkiep Terrane . Various correlations have been made in the past and recent dating work yields contrasting ages. U–Pb dating of zircons is applied in combination with Sm–Nd and Lu–Hf isotopes to determine the age, assess correlation and ascertain the much-debated tectonic setting . Ion probe U–Pb zircon dating yields ages of 1218 ± 5 Ma, 1212 ± 4 Ma and 1212 ± 5 Ma (all errors 2σ) for three samples of the Haiber Flats Formation and 1219 ± 13 Ma, 1215 ± 19 Ma for the Barby Formation. The ages obtained for the Barby Formation agree with the 1217 ± 2 Ma and 1214 ± 5 Ma ages reported from previous investigations. Rocks mapped as the supposedly older Kairab rhyolite and felsic tuff gave younger ages of 1220 ± 4 Ma and 1222 ± 10 Ma, respectively, which are coeval with the Barby and Haiber Flats Formations. Two Gorab Gabbro samples, previously mapped as Kairab Formation, yield ages of 1341 ± 8 Ma and 1336 ± 8 Ma. These older ages are within error of two Welverdiend Formation samples which gave ages of 1344 ± 6 Ma and 1337 ± 9 Ma, and the formation was previously dated in the type area at 1327 ± 10 Ma. These ages confirm the reassignment of rocks originally mapped as tholeiitic Barby Formation into the Welverdiend Formation, north of a newly identified stratigraphic hiatus corresponding to the geophysically defined Wereldend Lineament . Sm–Nd and Lu–Hf isotopic signatures for the Barby, Haiber Flats and coeval Kairab samples indicate that they are co-magmatic and derived from partly depleted mantle contaminated by older crust. TDM model ages of these samples range from 2.5 to 1.5 Ga with a major 2.2 to 1.8 Ga group. This suggests that the older crustal component was Paleoproterozoic but included some Archean material. The Welverdiend Formation and Gorab Gabbros show similarity in age, Sm–Nd and Lu–Hf characteristics and were probably derived from partially depleted mantle, or from a well-homogenised mixture of depleted mantle and older crust. REE diagrams indicate enrichment in LREE (La–Gd) and a flat HREE pattern (Tb to Lu). On spidergrams, most samples are characterised by negative Nb–Ta anomalies and positive K and Pb anomalies which indicate a subduction-related setting (with positive Sr anomaly) or a reworked crustal origin. The similarity in ages, isotopic signatures and geochemistry confirms the correlation of the Barby and Haiber Flats Formation.

The area of the city of Tsumeb in northern Namibia is strongly affected by gaseous emissions and by dust fallout from the local smelter. This is also reflected in increased concentrations of lead and arsenic in blood and urine of the residents. Consequently, modeling of the dispersion of dust and SO2 emissions from the smelter was used in this study to delineate the contaminated area and to assess the health risks. The modeling results were verified by ground-based geochemical survey of soil and grass in the area. The results of modeling revealed that the concentrations of SO2 in the Tsumeb town were relatively low, whereas the highest dust fallout concentrations were found around the Tsumeb smelter. The Tsumeb town residential area was less affected due to favorable landscape morphology between the smelter and the city (the Tsumeb Hills). The results of modeling of dust fallout and geochemical survey coincided very well. Since the anthropogenic contamination was bound only to the surface layer of soil, the local soils were sampled at two depth horizons: topsoil and the deeper soil horizon. This enabled us to distinguish between the anthropogenic contamination of soil surface from natural (geogenic) concentrations of studied metals in the deeper part of the soil profile. Concentrations of metals in grass correlated with the concentration of metals in topsoil. In contrast to a good conformity with the modeling of dust fallout from the smelter and geochemical survey, the results of modeling of SO2 contents in the air, and total sulfur content in soils were different. Differences can be explained by additional sources of contamination, as for example a sulfate-rich dust fallout from local tailings ponds and slag dumps that were not considered in the SO2 dispersion model. The results of the present investigation can be used by the mining companies in the management of air quality, assessment of the efficacy of applied remediation measures, and in reducing the impact of dust fallout on the local ecosystem. The Municipal Administration may use these results to plan further development of the city of Tsumeb, especially in terms of further expansion of housing construction.

The southern Central Zone (sCZ) of the Neoproterozoic to early Palaeozoic Damara orogenic belt hosts significant uranium deposits in Namibia. Rossing South-West, a leucogranite-hosted uranium deposit in the sCZ is partitioned into two sections; north-easterly and south-westerly sections with the latter (termed mining licence 120, i.e. ML120) forming the locus of this investigation. The deposit lies within the vicinity and to the west of the Welwitschia lineament which served as a conduit for mineralized fluids along with the Rossing and Husab deposits. Mineralogy and major and trace element geochemistry of the uraniferous leucogranites in ML120 suggest they are S-type granites derived from a sedimentary source as evidenced by the presence of igneous phases of garnet and cordierite together. Microscope and electron microprobe results from 133 powdered, drill core and hand samples prove that the deposits host similar primary uranium mineralization (chiefly uraninite which mainly exist as independent uranium minerals) and similar secondary U mineralization. At Rossing-SW U minerals include uraninite, coffinite, pitchblende, betafite, uranophane, betauranophane, carnotite, gummite and might partially exist as Th isomorphs. Mineralization at all locales is associated with post-tectonic D and E-type sheeted leucogranites containing quartz, alkali-feldspar and variable amounts of plagioclase producing rocks of variable compositions from tonalite to alkali-feldspar granite. Total reserves calculated at a cut-off grade of 100 ppm were reported as 171 Mt of ore at 376 ppm for Rossing and 205 Mt of ore at 497 ppm for Husab (probable). The Rossing-SW orebody is open-ended however total inferred resources stand at 304.4 Mt of ore at 235 ppm while measured resources stand at 8830t ore at circa 231 ppm. A comparison of the geological parameters (uranium mineralogy, leucogranite petrology and geochemistry) suggests similar ore composition. Consequently, ore from Rossing-SW's ML120 area could be processed at existing metallurgical plants (such as at Rossing or Husab) with a probable increase in U3O8 yield.

Dichrostachys cinerea and Senegalia mellifera trees are the most widespread invasive encroacher species in Namibia. Understanding their growth performance under arid conditions is key to understanding and managing the arid ecosystems of Namibia. To determine their utility in dendrochronology, we analyzed the formation of growth rings for two species from three sites in Namibia. We applied cross-dating techniques to 69 tree discs, the analysis of which confirmed the formation of annual growth rings. The rings are characterised by alternating layers of low and high vessel density and are separated by bands of marginal parenchyma. D. cinerea shows a well-defined pattern of growth rings, whereas the ring patterns of S. mellifera are more complex. Because aging trees tend to display narrower rings toward the outside of the stem, we applied both ring-width index and basal area increment methods in developing ring-width chronologies for the two species. Descriptive statistics for the chronologies indicate that the average ring chronologies are of good quality and that the trees are responding to common environmental signals. The basal area increment methodology produced better statistics than the ring-width index for these two species. Furthermore, statistical characteristics of the chronologies indicate that D. cinerea has a higher mean sensitivity (MS), expressed population signal (EPS), and signal-to-noise ratio (SNR) than S. mellifera. Thus, we find that both species are suitable for dendrochronological study, with D. cinerea being better than S. mellifera.