Chemistry 137 



increase the sorbed phosphorus pool and consequently DRP in the water 

 column. A detailed discussion of this whole problem is given in Prentki 

 (1976). 



The factors controlling this partition of phosphorus between water 

 and sediment have been investigated by correlation, isotherm, and 

 mapping techniques. Sediments used in these investigations were sliced 

 from cores or scraped from the sediment surface. Sediment sorption 

 isotherms and a one-point derivative phosphate sorption index (PSI), were 

 measured on wet sediments by procedures of Bache and Williams (1971). 

 The inorganic phosphorus extracted by oxalate was determined in 

 additional oven-dried (105°) subsamples of the same sediments utilized in 

 PSI measurements by use of the procedure of Williams et al. (1971c), 

 modified for smaller samples. This fraction is approximately equivalent to 

 the sum of NH4F-P through reductant-soluble-P in pond sediment (these 

 are the Chang and Jackson phosphorus fractions discussed earlier). 

 Organic phosphorus in sediment was usually measured as the difference 

 between the phosphate concentration in the oxalate extract and that in an 

 oxalate, extraction of an ashed subsample (550°C). A few organic 

 phosphorus determinations were also made by the difference between total 

 phosphorus and inorganic phosphorus in the Chang and Jackson 

 procedure described earlier and in Prentki (1976). The sediment inorganic 

 phosphorus that is resin-exchangeable was determined by equilibrating 

 sediments with an anion exchange resin and radiophosphate. 



Phosphate in water and in digests was determined by the methods 

 given earlier. Iron was measured by atomic absorption in many of the 

 extracts and digests. 



Phosphorus-Iron Correlations 



Phosphorus and iron in tundra sediments and soils show strong 

 correlation. Almost all the variation in total phosphorus content of the 

 sediment within or between ponds could be related to parallel changes in 

 sediment iron concentration. Thus, a regression (Figure 4-21) of all 

 measurements of total sediment phosphorus against sediment iron 

 concentrations, including those of samples collected in permafrost 

 underlying the ponds, resulted in a correlation coefficient of 0.81 and the 

 best fit line: P(Mg g~^) = (18) Fe (mg g" ')-l-297. The relationship between 

 iron and phosphorus was much stronger than that between either of these 

 and sediment organic content (as % organic weight) (r = 0.71 for 

 phosphorus and r = 0.47 for iron). Thus the correlation could not be 

 attributed to any coincidental association of both iron and phosphorus 

 with the organic or inorganic sediment fractions. Exclusion of the 

 sediment Mata from to 10 cm from the iron-phosphorus regression 

 improved the correlation and fit dramatically: P=:(22) Fe4-147, r = 0.91. 



