Title: The Río Grande de Arecibo Watershed II Characterization and Management of Non-Point Pollution Sources

Duration: Unknown

Project Summary Narrative
Biological growth in surface waters due to nutrient over-enrichment is a major source of water pollution in the United States. Sediments and bacteria are considered the main surface water quality impairments in Puerto Rico, because nutrient (especially total phosphorus (P)) concentrations do not usually exceed the current water quality standard for total P of 1000 micrograms/L. It has recently been suggested that the current water quality standard for P greatly underestimates the effects of nutrients on aquatic biomass growth and that an appropriate nutrient criteria for P should be significantly lower than the current value. Immediate evaluation of the nutrient dynamics in stream waters, assessment of sources, and linkage between sources and water-quality targets must be performed in order to improve or at least maintain Puerto Rico's surface water quality. In this project we will perform quantitative evaluation of nutrient, bacterial and sediment dynamics in stream waters draining from small sub-watersheds with well-defined land use within the Rio Grande de Arecibo (RGA) Watershed. Water samples will be collected at regular bi-weekly intervals and during storm events using automated water samplers. Sediment, nutrient and bacterial concentrations will be related to discharge using synoptic or continuous stream-flow measurements. Refinement of the current land-use classification will permit direct linkages between non-point sources located within the sub-watersheds and discharges, which will enable us to develop export coefficients for conditions in Puerto Rico. Management alternatives will be suggested that will enable remedial activities that will eventually result in improvement of water-quality to the RGA watershed.

Justification
Nutrient over-enrichment of surface waters is a major source of water pollution in the United States, where approximately half of the waters reported by the States to be impaired are attributed to excess nutrients and related to biological growth (Parry, 1998; Daniel et al., 1998). Similar problems occur in Puerto Rico although less quantitative data is available to adequately assess the problem. Vachier (1994) suggested that as of 1993, sediments and the presence of high fecal coliform and streptococcal bacteria were the main surface water quality impairments. A recent analysis of water quality in the Río Cibuco watershed by Horsley and Witten, Inc. (2002), found that fecal coliform bacterial counts exceed the water quality standard limit of 2,000 CFU/100mL in 43% of the water samples taken in the mainstem of the river during the years 1990 to 2000. Sotomayor et al. (2001) summarized trends in total phosphorus (TP) concentrations over an eight year period for major rivers in Puerto Rico and found that TP concentrations in the 25 to 75 interquartile ranged from 0.04 to 0.29 mg P/L, with mean and median values of 0.30 and 0.09 mg P/L, respectively. In this study, there were strong correlations between TP and the presence of fecal coliform bacteria, fecal streptococcal bacteria, and total Kjeldahl nitrogen. Ramos-Ginés (1997) quantified mean TP concentrations greater than 0.1 mg P/L during both low and high flow events entering a eutrophic Lake in Central Puerto Rico which were due to a combination of point and non-point sources.

The Puerto Rico Environmental Quality Board (EQB) has included 60 water-bodies (all but one corresponding to rivers) in the list of impaired waters for Puerto Rico (303(d) list) (JCA, 2002). The reduction contrasts with the 199 water bodies reported in 1998 by USEPA (USEPA, 2002) because inclusion was based solely on the suspected impact of point-sources of pollution. In the 2002 303(d) list, 66% of the river miles evaluated were negatively impacted with regards to one or more water quality parameters. Of the 102 watersheds in Puerto Rico, the Rio Grande de Arecibo (RGA) watershed is of prime importance and seven other waterbodies were included in the 2002 303(d) list (JCA, 2002). The RGA watershed is located in the north central part of Puerto Rico and has a catchment area of 41,440 ha (Figure 1). It includes five municipalities with a population of 981,103. The watershed has areas of exceptional natural value and has 427 river miles distributed among eight principal tributaries, three water reservoirs, and 31 intakes for potable drinking water (JCA, 1999). There are six point sources with National Pollutant Discharge Elimination System (NPDES) permits, and has 4,148 farms encompassing an area of 108,088 acres (USDA, 1998). Lago Dos Bocas within the RGA watershed has been classified as mesotrophic due to nutrients from point and nonpoint sources (USEPA, 2002). Through a newly inaugurated aqueduct system, it is expected to provide 70 MGD of potable water to nearly 1,000,000 residents in the San Juan metropolitan area.

Our research group has conducted a preliminary assessment of the water-quality status of lakes in Puerto Rico and potential contaminant sources in the RGA watershed and selected selected sub-watersheds (Martínez et al., 2002). Results suggest that current water quality standard for phosphorus (P) in Puerto Rico of 1000 micrograms/L (JCA, 1990), greatly underestimates the effects of nutrients on aquatic biomass growth and does not relate to numeric nutrient criteria developed in other ecoregions of the continental U.S. A value of 19.3 micrograms/L of total dissolved P could be used as an upper limit to assign pristine status to lakes in Puerto Rico, and a range of total dissolved P of 40 to 60 micrograms/L could be indicative of lake eutrophication. Within the RGA watershed, although none of the 33 stations in streamwaters evaluated exceeded the local PREQB water quality standard for total P (1000 micrograms P/L), nine of the stations exceeded 100 micrograms/L P in at least one of the sampling events (Figure 1). Our analyses revealed that there were no well-defined critical geographical zones of P contamination on this watershed.

The project also addressed two of the most critical components of a TMDL: source assessment, and linkage between water quality targets and sources. The purpose was to assess the contribution of a particular land-use to the overall nutrient and sediment loads of the watershed. The RGA watershed was divided into a series of subwatersheds based on landuse, topography and hydrology. Five subwatersheds were selected for further evaluation based on their landuse and accesibility. Land-use classification was updated using 1994 digital orthoquad photographs and the Watershed Modeling System (WMS ver 6.1 EMS-1) software (Figure 2, Table 1). Figure 3 shows an example of land-use evaluation for the Río Caonillas watershed which includes Río Saliente and Río Jauca. Ground truthing of all subwatersheds is ongoing and has been performed in cooperation with UPR-AES and USDA-NRCS personnel.

Five subwatersheds were monitored at bi-weekly intervals (using grab samples) for sediment and nutrient loads from June through August 2002. All the subwatersheds had mean TP concentrations well in excess of what is generally suggested in the scientific literature for eutrophication control. In addition, we found extremely high bacterial coliform counts, an additional indication of the generalized impairment of the surface waters of the Arecibo watershed. However, our results were based on a very limited monitoring exercise (7 sampling events), and thus, the evidence is much too preliminary to exert any conclusive remarks.

Specifically, the different tributaries must be sampled through various cycles of high flow and base flow conditions to detect variations between seasons and hydrologic conditions. Herein, we propose to continue our monitoring effort for an additional year to accomplish such objective. In addition, we will specifically identify the specific contributing sources (e.g., animal production facilities, unsewered communities, crop farms, etc) on each subbasin. Further detailed characterization will be obtained by using automated water samplers to evaluate nutrient concentrations and bacterial discharges during storm events. This information will be combined with instantaneous flow measurements to calculate an export coefficient for each contaminant-land use combination. The latter information would be essential for the eventual development of a TMDL for fecal coliforms and other potential contaminants for the whole watershed.