Terrigenous materials are being supplied to the world's oceans by rivers at a flux rate of about 1.5 to 2 x 1016 g/yr (Holland, 1981). Most of this material is trapped in estuaries or on continental shelves and only a minor portion reaches the deep sea. Eolian dust, perhaps only 5 percent of the total lithogenous flux (approximately 0.53 to 0.85 x 1015 g/yr, Prospero, 1981), dominates the mineral component of pelagic sediments in regions away from the influence of turbidites, equatorward of the effects of ice rafting and seaward of the influence of hemiplegic deposition (Windom, 1969, 1975; Rea et al., 1985). The eolian input estimate of Prospero (1981) when divided over the 360 x 1016 cm2 of the oceans and marginal seas results in an average dust flux value for the entire ocean of about 200 mg/cm2/kyr.
Dust is raised by storms from the arid and semi-arid surfaces of continents, elevated to the upper portion of the troposphere, transported global distances by the zonal winds and washed out of the atmosphere by rain. The long-term pattern of this sedimentary process is shown clearly in maps of the quartz content of surficial sediments of the sea floor (Leinen et al., 1986). Regions relatively enriched in quartz, which makes up roughly 10 to 20 percent of the total eolian load, extend downwind from the major deserts of the world—Gobi, Sahara, Arabia, Australia.
The short-term record of dust transport is seasonal. Spring storms are responsible for most of the annual transport. The best record of present-day eolian transport is that developed by Prospero for materials crossing the Atlantic from the Sahara. That record shows an order of magnitude variation in the amount of dust transported in any given year with maxima during the late spring or early summer. On a longer term basis the flux of dust crossing the North Atlantic shows a three- to fivefold increase during the height of the Sahelian droughts in 1973-1974 and 1983-1984, demonstrated times of significantly reduced rainfall (Nicholson, 1985; Middleton, 1985) in comparison to more normal years (Prospero and Nees, 1977, 1986). This sort of seasonality in dust transport with maxima in the late spring occurs wherever dust fluxes have been measured (Parrington et al., 1983; Uematsu et al., 1983; Merrill et al., 1989). Useful summaries of eolian processes and dust deposition have been given by Windom (1975), Prospero (1981), Rea et al., (1985), and Pye (1987).
The mass input of any sedimentary component to the sea floor can be quantified. The parameters necessary to determine the mass accumulation rate (MAR) in g/cm2/kyr, or flux, of dust are: the linear sedimentation rate (LSR) in cm/kyr, the dry bulk density (DBD) of the sediment in g/cm3, and the weight percent of the eolian component of the bulk sediment. Eolian flux values are the product of these three. The eolian component itself is isolated from the total sediment by a series of extractions that remove all other sedimentary components, leaving the minerals (Rea and Janecek, 1981). The amount of any volcanic ash remaining after extraction is estimated visually from smear slides and deducted from the total to give values for continentally derived material.
Sediment samples usually span one or two centimeters of any given core and so represent many hundreds to thousands of years of deposition. Furthermore, any initially discrete event or signal is smoothed by the bioturbation process that acts to homogenize the uppermost several centimeters of deep-sea sediments. Short-term climatic variability on time scales of decades to centuries is therefore smoothed and the records we present are representative of truly long-term climatic trends and events. Our working assumption, based largely on the data of Prospero on transport from Africa (Prospero, 1981; Prospero and Nees, 1986), is that the amount of dust transported to the oceans is a function of source area aridity; drier, less vegetated regions provide more dust to the atmosphere than do moist, well vegetated regions. This assumption is also consistent with information available for the pedology of the great loess-soil sequences in China (Kukla, 1987; Kukla and An, 1989) and the distal record of those sequences (Hovan et al., 1989). The supply of dust to the oceans is independent of the intensity of transporting winds (Chuey et al., 1987), that aspect of eolian deposition is recorded by the grain size variation of the dust (Rea et al., 1985; Pisias and Rea, 1988).
The input rate of dust to the pelagic oceans today (Holocene) ranges through three orders of magnitude, from low values of 1 or 2 mg/cm 2/kyr in the center of the South Pacific to more than 1000 mg/cm2/kyr just downwind from the North African and Asian source regions. There are perhaps 35 locations in all the oceans where flux values for clearly pelagic eolian minerals have been determined (Figure 8.1; Table 8.1), locations more than 500 of 1000 km offshore and thus beyond the effective range of hemipelagic deposition. All values calculated for near continent regions may include a significant hemipelagic component. When compared with information from nearby sediment traps or the data from island-based collectors, the dust flux values from sediment cores agree to within the combined errors of the two kinds of measurements (Rea et al., 1985).