Chemistry 141 



A second correlation matrix (Table 4-22) constructed for the first five 

 pond sediments and the soils (Table 4-18) and seven additional soils 

 described in Brown et al. (in press) suggests similar relationships over the 

 entire Barrow IBP site. For this matrix two additional parameters were 

 available for regression: the resin-exchangeable phosphate, which is the 

 phosphorus exchangeable with the radiophosphate sorbed on an anion 

 exchange resin; and the extractable iron, which is the sum of iron 

 solubilized by ammonium fiuoride, plus the iron in the first sodium 

 hydroxide extraction plus the iron in the reductant-soluble extractions 

 (Chang and Jackson). In this matrix there were again strong correlations 

 between phosphorus and iron parameters, especially the extractable iron. 

 Extractable iron was better correlated with total phosphorus (r = 0.76**) 

 and reductant soluble phosphorus (r = 0.78**) than either inorganic 

 phosphorus (r = 0.68*) or organic phosphorus (r = 0.56); however, the 

 relationship between total phosphorus and total iron found earlier for all 

 sediment analyses (Figure 4-21) was not evident in these soils and 

 sediments. The earlier relationship may have reflected homogeneity within 

 a single drained lake basin. 



The various correlations with the resin-exchangeable phosphorus are 

 strong evidence that phosphate is bound to sediments by sorption and is 

 not precipitated as discrete aluminum, iron, or calcium phosphate 

 minerals. Resin-exchangeable phosphate is a measure of that portion of 

 inorganic phosphorus which is sorbed on soil or sediment particles but 

 which is still capable of desorption and interaction with the water phase. 

 The NH4F-P and first NaOH-P fractions are capable of exchange with 

 phosphate dissolved in the water or of interaction with the water phase 

 only in soils or sediments in which they make up sorbed phosphate pools. 

 In contrast, in soils or sediments in which they make up discrete mineral 

 phases there would be (1) no expected relationship between resin- 

 exchangeable phosphate and any Chang and Jackson phosphorus 

 parameter, and (2) no appreciable pool of resin-exchangeable phosphate. 

 In these ponds the resin-exchangeable phosphorus pool accounts for 70% 

 of the phosphorus in the NH4F-P plus the first NaOH-P fractions. In 

 addition there are highly significant correlations between the resin- 

 exchangeable-P and the NH4F-P fractions (r = 0.87**) and between the 

 resin-exchangeable-P and the NH4F-P plus the first NaOH-P fractions 

 (r = 0.82**). All of these indicate that there is no discrete mineral 

 formation. The actual regression between NH4F-P plus the first NaOH-P 

 fractions and the resin-exchangeable phosphorus indicates almost a 1 to 1 

 correspondence such that NH4F-P plus the first NaOH-P = (I.l) resin- 

 exchangeable-P -I- 24. 



The importance of sorption-binding in pond sediments, as suggested 

 by these correlations, agrees with the theoretical stability field calculations 

 for aluminum, iron, and calcium phosphate minerals given in Brown et al. 

 (in press) for the pH, Eh, and ranges of cation concentrations found in the 



