A preliminary study was undertaken to evaluate the water quality characteristics of urban runoff in Guam. The study was divided into three distinct phases. The first phase identified stormwater retention sites (sinkholes and ponding basins) in the northern half of the island. These were identified and delineated on a USGS 1:24,000 quadrangle topographic map and digitized into a geographic information system (GIS). The second phase of the study focused on the water quality characteristics (nutrients, major ions, and heavy metals) of urban runoff from several stormwater retention sites within the commercial and residential sectors of the community. Samples were manually collected from these sites on a limited number of occasions. In the third phase, a fully automated sequential sampling system was employed to collect runoff for analysis from a single storm drain that serviced the Palace Hotel in Tamuning. Samples from this large commercial complex were collected intermittently over a three-month period. The primary objective of this part of the program was to develop a fully automated model monitoring system that would: (a) provide for the integration of synoptic rainfall measurements with runoff samples collected over the course of runoff events, and (b) allow the autosampler to be interrogated from the laboratory to determine sampling status.

Runoff from the stormwater retention sites revealed substantial spatial and temporal variations in levels of nutrients (NO₃-N and o-PO₄-P), major ions (Ca, Mg and Na) and certain heavy metals (Ag, Cd, Cr, Cu, Fe, Mn, Ni, Pb and Zn). Several metals were found to be significantly associated with suspended solids. However, the ratios between "dissolved" and "particulate" metal fractions varied between metals both within and between sites. In contrast, the nutrients and major ions were almost exclusively confined to the dissolved fraction in the great majority of samples examined.

A comparative analysis with data previously reported for Guam's urban runoff indicated that nutrient and heavy metal concentrations had not increased significantly over the last two decades. In fact, Cd and Pb levels appeared to have diminished somewhat, possibly reflecting current restrictions in the commercial use of these elements. Overall, there were very few exceedences of current surface and drinking water quality standards. The significance of the ponding basin water quality data is discussed with reference to published findings elsewhere in the world and includes short notes on relevant aspects related to human and environmental health. Levels were generally considered low by world standards despite the occasional indications of mild nutrient enrichment at certain sites. The maximum NO₃-N and reactive o-PO₄-P levels recorded were 3.67 mg/L (Mariana Terrace ponding basin) and 1.92 mg/L (Harmon Sink), respectively.

The water quality characteristics of runoff collected from the Palace Hotel storm drain was very different from that found in the retention ponds and was generally enriched in all detectable components. Especially noteworthy were the extraordinarily high levels of o-PO₄-P determined on several occasions (up to 482 mg/L). We determined that the autosampler was triggered by a number of storm and non-storm events. The latter were generally higher in NO₃-N, o-PO₄-P, Na, Mg, and Cu, and were suspected of being generated by water sprinklers operating within the hotel's landscaped areas. The major source of enrichment was concluded to be water soluble fertilizers inadvertently scattered on paths, roads and walkways adjacent to targeted lawns and gardens, and flushed into the storm drain in irrigation runoff. The possible connection between the "seaweed problem" in Tumon Bay and fertilizers used in maintaining landscaped areas of adjacent hotels is discussed with recommendations for future research.