The Potential Impact of Commercial Agriculture on Saharan Dust Flux

 Mulitza et al. (2010) analyzed sediment cores off the coast of West Africa to determine if commercial agriculture increases the rate in which dust blows out of the Sahara; the Sahel dust flux. Saharan dust travels on the trade winds and creates a mass called the Saharan Air Layer, which makes up the dominant proportion of the atmosphere in the region. Dust, like air, can travel by convection, a cycling process of energy transfer in which heated, dry air becomes less dense and rises, and then sinks once more as the air condenses in the cold, higher atmosphere away from the earth’s radiation. In the Sahara region, this convective cycle, called the Hadley Cell, can carry dust on dry, descending air masses for thousands of miles. Thus large concentrations of Saharan dust can be found as far as the Caribbean and even northwards to the Mediterranean and Northern Europe. Dust flux contributes to nutrient cycling by carrying fungi and other materials usuable by phytoplankton to different parts of the ocean. It can also cool the Sahara’s surface by reflecting light, and can influence the patterns of tropical cyclones by reducing convection rates. As the warm Saharan Air Layer laden with dust rises, it doesn’t cool as quickly as it otherwise would because of the insulation of the dust particles, so that the hot air gets lifted above the cooler, humid marine air and creates an inversion, a phenomenon in which the air closer to the earth is cooler than the air above it. Since monsoons and hurricanes depend on the energy released from humid, rising air, such inversions suppressing the convection of the marine layer also suppress such storms. Through examining the past 3,200 years of sedimentary deposit in core records, Mulitza et al. discovered that the levels of dust flux in the most recent 300 years far exceeds levels of prior years and could influence weather patterns and atmospheric compositions in years to come.–Elise Wanger

Mulitza, S., Heslop, D., Pittauerova, D., Fischer, H., Meyer, I., Stuut, J., Zabel , M., Mollenhauer, G.,  Collins, J., Kuhnert, H., Schulz, M., 2010. Increase in African dust flux at the onset of commercial agriculture in the Sahel region. Nature 466, 226–228.
Geologists usually attribute the varying levels of dust flux to the amount of precipitation: the more rain, the less easily dust will get picked up and blown around. However,  in the past 300 years, human activities may have influenced dust emissions more significantly, as the data of Mulitza et al. suggest. Mulitza et al. constructed the record of dust flux off the coast of West African for the past 3,200 years through analyses of multicore and gravity core samples collected at a marine site, GeoB9501, located on the northern flank of the Mauritania canyon and about 323 meters below water. They subsequently compared the data collected to those of another report in Barbados—in order to act as a control to account for the possible influence of sediment from the runoff by local rivers in the GeoB9501 core—and then to an oxygen isotope record from Lake Bosumtwi in Ghana: a large, natural lake in the Sahel region that would have similar historical precipitation rates as site GeoB9501. While the dust flux levels from site GeoB9501 3,200–300 years ago correlate with temporally equivalent dust levels collected in Barbados and the isotope ratios from the lake core sample, the dust levels from GeoB9501 make a sharp and unexpected increase from about 200 years ago to 2000 CE that are not accounted for in the two other records.
 Mulitza et al. took two core samples at site GeoB9501: a 0.42 meter-long multicore, which provided information from about 1915–2000, and a 5.32 meter-long gravity core, which records the past 3,200 years of sedimentation. While a multicore uses machine-tools to dig into the sediment and is more reliable, the gravity core uses the force of gravity to sink through the sediment and can go deeper. With each core, Mulitza et al. looked at the grain-size distribution, the chemical compositions, and the dry bulk density of each sample slice to determine the rates of dust flux. The grain-size distribution allowed them to determine where the sediment came from, since “95% of the particles delivered by the Senegal River are smaller than 10 mm” (Mulitza et al., 2010; 226) while coarse dust from the Sahel region tends to be around 200 mm. The two sites also differ chemically, with Senegal River suspension carrying vastly more aluminum and iron, while the Saharan dust (Sahel) has more silicon, potassium and titanium. The researchers could therefore determine the ratio of Saharan dust to fluvial runoff in their samples. These elemental concentration data (Al, Si, Ca, K, and Ti) were also compared with nine contemporary samples from the Sahara-Sahel dust corridor—an eolian channel running from Chad to Mauritania—and ten samples of Senegal River suspension. Through these three methods, Mulitza et al. could determine the ratio of eolian to fluvial sediment and the source of the sediments. Taking the overall bulk density further elucidated the record by comparing how much overall sediment built up in the respective time periods. If some natural disaster caused rivers to overflow or a storm blow more dust, the bulk density could account for it.
 Sediment older than 500 years with plankton embedded could be carbon-dated, but a lot of the gravity core and all of the multicore samples contained younger sediments. Researchers then used 210Pb/137Cs dating to estimate the year the sediment was deposited. The levels of 210Pb and 137Cs could be detected through gamma spectroscopy, since both elements have predictable levels of radionuclear activity. When sub-atomic particles interact—such as during radioactive decay—they release energy in the form of gamma rays. These rays were measured with a common scintillation detector, (NaI(T1)) (Pittauerova et al. 2009). Scintillation detectors are composed of crystals that stimulate the gamma rays with a photon, and subsequently emit light at an intensity proportional to the energy released. This intensity (measured in electron volts) has a typical peak depending upon the source: in the case of 137Cs, about 662 keV. The cesium detected could have only come from a nuclear bomb testing in the area in 1945. Researchers calculated the cesium would have peaked in 1963 and therefore calculated the age of the core layer with 137Cs. In the analysis of the multicore,  “measurement errors [were] relatively high and [did] not provide fine resolution,” (Pittauerova et al. 2009; 460) but they did match up with the expected peak of 1963.
To determine the age with lead, researchers used the CRS (Constant Rate of Supply) model (Pittauerova et al. 2009). The model assumes a constant rate of 210Pb supply and uses the mean sediment deposit rate determined from a least-squares fit procedure. In other words, CRS 210Pb dating is contingent upon correctly estimating the initial Pb laid down at each core layer. Radioactive decay from the initial point then proceeds predictably, with the activity of 210Pb declining, from the point of being laid down. In this case, Mulitza et al. used scintillation detectors again to pick up on the radionuclear activity of 210Pb/137Cs, but looked at the internal conversion instead. The internal conversion is the radioactive decay process in which an electron gets emitted from the atom due to interaction with an excited nucleus. Rather than the photopeak, which is the characteristic energy emitted from the element and consistent, the internal conversion changes with time.
Once the age, chemical composition, and origin of each sediment sample was determined, Mulitza et al. looked for correlations between the various factors: the dust fraction, the terrigenous fraction greater than 10 mm, the dust deposition flux, the runoff levels from the Senegal River, the dust concentration, and the oxygen-18 isotope ratio from the core sample of Lake Bosumtwi. The dust fraction is simply the proportion of sediment in each season that comes from dust, as opposed to fluvial runoff, as measured by the concentrations of various elements (Al, Si, Ca, K, and Ti). Since fluvial runoff tends to carry fine-grained, smaller sediments, calculating the terrigenous fraction greater than 10 mm is another method of estimating the proportion of dust flux in the sediment samples. Taking the data from radioactive dating and the proportions of dust corresponding with the dates, the dust deposition flux is a calculation of the rate dust gets deposited, measured in volume per year. The oxygen isotope ratio indicates the predicted precipitation levels of the various time periods. Oxygen-18 and oxygen-16 both exist naturally in the environment, with oxygen-16 being the more prevalent. Given that temperature and wind patterns (and therefore rain patterns) are fairly unchanged, the same δ18O will initially exist in the precipitation entering the lake. Since oxygen-16 is significantly lighter than oxygen-18, however, oxygen-16 will more easily change phases from liquid to vapor. So if the weather is dry, more oxygen-16 will evaporate and leave more oxygen-18 behind. But if the weather is wet, oxygen-16 may be entering the lake in high enough rates and evaporating slowly enough to have a lower δ18O. Thus a high δ18O indicates a drier season and a low δ18O indicates a wetter season.
The terrigenous fraction greater than 10 mm and the dust fraction correlate fairly consistently throughout the 3,200 year period, as does the dust flux. The δ18O also correlates with the other data until about 1700, with the dust fraction and the terrigenous fraction >10 mm lower with lower δ18O, indicating less dust with wetter conditions. Runoff is also graphed, but from the multicore only, meaning that the data goes back to 1915, and changes drastically from season to season. It does clearly decline starting 1968 during the Sahel drought, which begins to increase again just before 2000. Mulitza et al. also graphed the dust flux levels from a core in Barbados starting in 1914, as further evidence that neither the levels of dust nor the data are confounded with fluvial runoff.  In dust fraction, terrigenous fraction >10 mm and dust flux, levels take a sharp increase starting around 1700 that significantly exceeds anything from the previous years, and are not correlated with the oxygen isotope ratio. Mulitza et al. conjecture that this spike in dust flux could be the result of commercial agriculture in the region. They also propose that the cooling of the surface through agricultural processes like irrigation could be reducing monsoonal rainfall and thus contributing to desertification. Given the effect of desertification on global weather patterns, air quality and soil properties (Karlsson et al. 2008, Ozer et al. 2007), if agricultural practices truly have such an instrumental role on dust flux, than continuing and expanding such practices could have a deep effect on the future climate. The research of Mulitza et al. is a pivotal step towards understanding the Saharan dust system and figuring out the potential consequences of mass farming on this often neglected ecological system.
Other sources:
Pittauerova, D., Mulitza, S., Hettwig, B., Chehade, W., Stuut, J., Mollenhauer, G., Fischer, H., 2009. Application of self-absorption correction method in gamma spectroscopy for 210Pb and 137Cs sediment chronology on the continental slope off of NW Africa. Radioprotection 44, 457–461.
Karlsson, L., Hernandez, F., Rodriguez, S., Lopez-Perez, M., Hernandez-Armas, J., Alonso-Perez, S., Cuevas, E., 2008. Using Cs-137 and K-40 to identify natural Saharan dust contributions to PM10 concentrations and air quality impairment in the Canary Islands. Atmospheric Environment 42, 7034–42.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s