Biogeochemical Impacts of Dust on the Global Carbon Cycle - Mineral Dust: A Key Player in the Earth System - page 360

Fig. 14.2 Average global dust deposition flux based on Jickells et al. ( 2005 )

Bristow et al. ( 2010 ) report the composition and mineralogy of dust from the Bodélé
depression in the Sahara, a major dust source, and also conclude that this is an
important source of P to the Amazon. Swap et al. suggest that atmospheric P inputs
are equivalent to up to 3 % of the annual cycling of P by leaf fall in the Amazon.
Hence, on an annual basis, most of the P cycling is driven by internal recycling
within the ecosystem, but viewed from a different perspective, the atmosphere can
supply the entire soil P reservoir over timescales of 25,000 years (Okin et al.

2004 ). Phosphorus is usually considered as involatile and so is only lost from the
soil by runoff, but Mahowald et al. ( 2008 ) suggest that emissions associated with
biomass burning may now be of the same scale as atmospheric P deposition and
result in a net P loss from parts of the Amazon Basin and an increase in atmospheric
P deposition downwind of fires. As discussed by Okin et al. ( 2004 ) over timescales
of thousands of years, the climate in the Sahara desert, the source region for dust
over the Amazon, has changed dramatically; indeed, the Sahara desert was not truly
a desert 5,000 years ago. The terrestrial Amazon ecosystem may still be responding
to the increasing dust P input of the last few thousand years.
Pett-Ridge ( 2009 ) suggests that the atmospheric supply of phosphorus to the
montane wet tropical forests of Puerto Rico is higher than the input to the Amazon
and the soil P lower, suggesting that atmospheric P supply is quantitatively even
more important in the Puerto Rican system than in the Amazon. Overall Pett-Ridge
( 2009 ) suggests that atmospheric P supply is the main source of phosphorus in this
system which has a particularly high dust-associated P flux because it is directly
downwind of the Sahara desert. Similar nutrient-poor ancient soils to those in parts
of Puerto Rico and the Amazon underlie most of the tropical rainforests (Vitousek
and Sanford 1986 ; Augusto et al. 2013 ), and these authors present data which,
despite the uncertainties they note for it, do suggest that atmospheric deposition