410 



76 OCEANOGRAPHY IN THE NEXT DECADE 



trolling the vertical flux of elements needs to be characterized. 

 The same is true for the evolution of the chemical properties of 

 sinking particles traveling through the complex food web in the 

 water column and on the seafloor. In the upper waters and ther- 

 mocline, large horizontal variations in primary productivity and 

 higher levels of food webs have been observed, but little is known 

 about the lateral variability of deep-ocean ecosystems. The ef- 

 fects of ecosystems on the chemistry of particles passing through 

 them are largely unknown. It is important to know whether there 

 is a simple proportionality between surface productivity and the 

 chemistry of the underlying water column or whether the particu- 

 lar faunal assemblages in the water column exercise a major role 

 in controlling element concentrations. 



The mechanisms of sediment interactions and diagenesis are 

 well studied for the major constituents, nutrients, and oxygen; 

 much more work is needed on the behavior of minor and trace 

 constituents. Intensive research on the behavior of trace elements 

 in the most reactive upper few meters of the ocean is also neces- 

 sary. Descriptions of composition changes need to be developed 

 for the various sediment types and environments through all phases 

 from initial burial to subduction. 



Reading the Record 



Complete decipherment of the proxy record contained in the 

 concentration distributions of trace elements and isotopes in sedi- 

 ments requires an understanding of the pathways of input and the 

 mechanisms of redistribution, removal, and transformation of the 

 elements studied. A given tracer may record a single aspect or 

 some combination of these factors. Ideally, multiple tracers should 

 be used as a check on internal consistency. Because only a small 

 number of tracers are now available, much development work is 

 required. 



It is known that the continental inputs of a number of poten- 

 tial tracers are changed markedly by human influence, which makes 

 estimates of their response to environmental changes difficult. 

 River inputs are often strongly mediated by coastal processes and 

 hence are sensitive to sea-level variations that change the size 

 and character of the coastal ocean. Comparative studies of shelf- 

 dommated systems (e.g., the Yangtze and Yukon rivers) with sys- 

 tems in which rivers discharge directly over deep waters (e.g., the 

 Columbia and Congo rivers) may be informative. 



The relative importance of physical and chemical redistribu- 



