This study was designed to evaluate nitrification in Guam's northern soils as a possible cause of high nitrate concentrations in the aquifer system. Samples of topsoils were collected near drinking water well sites (Dededo, Mogfog, and Finegayan) after periods of dry and wet weather for one year. Nitrate production was examined in the laboratory in experimental flasks that contained ca. 1g of soil in 100 mL of rainwater or other media. Average nitrate production rates for soils at the three well sites were 220±320, 180 ±240, and 290 ±310 nmol NO₃^2- g^-1 d^-1 ( ± 1 SD, n = ca. 44 each), respectively. Nitrification was stimulated by the addition of ammonium. These rates were sufficient to enrich the aquifer with ca. 200,000 times more than nitrate than has been routinely observed.

To see if nitrate-rich transition zone waters could enrich the aquifer, nitrification was studied in coastal sediments. However, in coastal sediments, net denitrification occurred within 1 to 2 days. Stable N isotope analyses (δ¹⁵N) indicated that both soil and sediment N were derived from natural microbial processes and were probably not contaminated with domestic sewage.

Because laboratory flask data were much too high, nitrification was studied in an experimental soil column (170 cm² x 0.7 cm long) constructed with topsoil and underlying carbonate from the Dededo site. An average of 146μM NO₃^2- was observed in leach water from this column, which is essentially the same as that observed in the aquifer waters. At calculated average aquifer recharge rates of 1.15 m/yr, nitrification rates deduced with a mathematical model were also realistic.

The large difference between the flask rates and the soil column rates implies that potential nitrification is extremely high, but essentially all of the NO₃^2- is lost from the soil waters via denitrification prior to percolation down to the aquifer. If this is true, then it is reasonable to assume that denitrification occurred due to deoxygenation in waterlogged soils. Therefore, nitrate-rich soil water that recharges the aquifer must leave the soil horizon not only devoid of O², but also enriched with NH₄⁺, reactive P, metals, and other substances that are more soluble under anoxia. However, because aquifer waters are not anoxic but are routinely undersaturated with respect to O² (ca. 60% of saturation), they must be reoxygenated en route from the soil to the aquifer. Then, these substances may precipitate (e.g., metal-P complexes as metal apatites) in the freshly oxygenated recharge waters. Therefore, the carbonate platform below the soil horizon but above the aquifer may now be enriched with potentially harmful substances. Subsequent downward migration closer to the well waters will largely depend upon (1) their differential solubilities in oxygenated waters, (2) the mass of deoxygenated water that leaves the soil horizon, and (3) the rate at which waters migrating through the platform are being oxygenated.