Water and sediment quality have been affected by urban development in the San Juan Bay estuary system, Puerto Rico. Bottom-sediment samples were collected and analyzed for nutrients, trace elements and organochlorine compounds at six sites distributed throughout the estuary system. Sediment layers were assigned to chronostratigraphic units using cesium-137 dating techniques. Temporal changes observed in concentrations of anthropogenic nutrients, trace elements, and organic compounds measured in the cores were consistent for all sites. This supports the usefulness of using cesium-137 as an indicator of time horizons in estuarine sediments in spite of active resuspension and bioturbation that rework the surficial sediments. Total phosphorus concentrations increased from approximately 150 micrograms per gram for sediments deposited at the beginning of the twentieth century to more than 800 micrograms per gram in recently deposited sediments. At the most contaminated site sampled in the estuary, lead increased from a background concentration of about 30 to 750 micrograms per gram in the most recent strata, mercury increased from 0.16 to 4.7 micrograms per gram, and PCB's increased from 12 to 380 micrograms per kilogram. DDT and its metabolites were below the detection level of 0.1 microgram per kilogram in the oldest sediments, increased to 62 micrograms per kilogram in bottom sediment representing the period 1950-74, and decreased to 26 micrograms per kilogram for the bottom sediment representing the period 1975-95.

Since 1980, water quality in the lagoons and coastal plain streams improved as construction activity in the watersheds draining into the San Juan Bay estuary system slowed and wastewater treatment plant discharges were diverted to deep ocean outfalls. However, in the late 1980’s, construction activities increased in the headwaters of the largest river in the San Juan Bay estuary system, the Río Piedras, reversing the generally improving trends documented in index water quality properties recorded as part of ongoing monitoring programs.

The economy and associated land use of Puerto Rico, including the San Juan metropolitan area that borders most of the estuary, was mostly agrarian prior to 1925 (Birdsey and Weaver 1987). Development of urban areas increased after 1925, and a rapid industrial transformation began in the late 1940’s and early 1950’s (Seguinot- Barbosa 1983). The changes in land use and increasing population pressures resulted in specific changes in water quality and sediment quality in the estuary (U.S. Environmental Protection Agency 1992). The changes have been recorded in sediments deposited in undisturbed areas of the estuary and in the database of water-quality observations compiled during the last quarter-century. These records document the improved water quality that has resulted from implementing pollution control measures established in the 1970’s. In addition, the sediments deposited before significant changes in land use serve as a proxy of background water quality conditions before significant development took place. Characterization of predevelopment conditions is necessary to establish realistic goals for conservation and management plans.

Coordination between the Puerto Rico Environmental Quality Board (PREQB) and the U.S. Environmental Protection Agency (USEPA) in 1993 succeeded in establishing a San Juan Bay Estuary Program (SJBEP) under the auspices of the National Estuary Program. The SJBEP is charged with creating a Comprehensive Conservation and Management Plan (CCMP) for the estuary system. The CCMP will include actions to restore estuarine functionality to as close to predevelopment conditions as possible while balancing the needs of continued development. Water quality in the estuary system began to be measured with regularity in the early 1970’s, well after the system had been severely degraded by decades of intense use by agriculture and industry. This paper details temporal and spatial changes observed in the sediments for chronostratigraphic units in addition to temporal changes in the water quality in the San Juan Bay estuary system from 1970 through 1995.

The San Juan Bay estuary watershed covers 240 km² and consists of limited uplands draining north across a broad, flat coastal plain (fig. 1). The discharge from multiple drainage basins is mixed with the sea water in the lagoons and canals that constitute the watershed of the estuary system. The subtidal part of the system consists of the Bahía de San Juan (also referred to as San Juan Bay), connected by a series of natural and dredged channels to four lagoons: Laguna del Condado, Laguna San José, Laguna La Torrecilla, and Laguna de Piñones. In total, the inundated areas, including San Juan Bay and lagoons, cover 25 km². Quaternary deposits, which cover 147 km², include marine and riverine terrace deposits, alluvium, beach and swamp deposits, blanket deposits, and artificial fill. Volcanic and intrusive rocks cover 55 km² and are limited to exposures in the uplands in the southernmost part of the watershed. Limestone formations cover 13 km². These deposits are located mostly in the western part of the watershed and are actively mined to provide raw material for the island’s construction industry.

In addition to the surface water discharge by streams to the estuary system, there are discharges from urban "flood control" drainage pumps. Combined sanitary and fluvial connections and untreated sewage discharges have direct impacts on the water quality and on the concentrations of pollutants scavenged from the water column and deposited in the sediments. In addition, Quebrada Blasina, which discharges into Laguna La Torrecilla, receives backwash from the Puerto Rico Aqueduct and Sewer Authority (PRASA) Sergio Cuevas Water Treatment Plant.

Trends in water and bottom-sediment quality were identified in bottom sediments assigned to chronostratigraphic units and in records of water quality measured at various sites throughout the estuary system.

Bottom sediments were cored and samples collected from six sites in December 1994: Bahía de San Juan (SJN600); Caño de Martín Peña (MPN500); Laguna Los Corozos (CRZ400); Laguna San José (SJS300); Laguna La Torrecilla (TRC200); and Laguna de Piñones (PNN100) (fig. 1). The sites were chosen in areas of fine-grained sediment deposits and away from dredged or filled areas. In addition, wherever possible, the coring sites were located in the central areas of water bodies to document contamination introduced from various sources around the perimeter of the water body and not just the influence of a local point source. To obtain enough material for radioisotope analysis, three cores from each site were recovered and homogenized in 5-cm intervals. The only exception was the Caño de Martín Peña site. Sedimentation rates were expected to be high there, and so only two cores were obtained and sampled in 10-cm intervals. The concentration of total phosphorus and the activity of the radioisotope Cs-137 was measured for each homogenized sediment sample collected in December 1994. Total phosphorus in the fine fraction (that portion passing through a 2-mm sieve) was determined using colorimetry techniques (method I-6600- 88 - Fishman and Friedman 1989).

Sedimentation rates at the six sites in the estuary system were estimated with Cs-137 techniques. Cs-137 is a by-product of atmospheric nuclear-bomb testing and has a half-life of 30 years. The radioisotope does not occur naturally and is distributed throughout the atmosphere after sub-aerial detonations. The radioisotope eventually settles out on the surface of the earth where it binds firmly to sediment particles. Fluvial processes transport the sediment and adsorbed Cs-137 through the drainage system to depositional areas (McHenry et al 1980). Strata in the sediments with high Cs-137 activity correlate to periods of increased activity of sub-aerial nuclear detonations. The most distinct Cs-137 time horizons, or peaks, usually identified are 1952, the first year of significant nuclear testing and subsequent fallout of Cs-137, and 1963, the most active year with frequent testing by both the United States and the former Soviet Union. The peaks may lag by a year or two after 1963 if there is any delay in the fluvial transport from the land surface to the depositional site (Crusius and Anderson 1995).

For all sites, except for the one in Caño de Martín Peña (MPN500), the average vertical accretion rates were estimated by assigning the year 1954 to the base of the lowest strata that had values of Cs-137 greater than the minimum reporting limit of 1 milliBecquerel per gram (mBq/g). This assumes that approximately 2 years passed after the initial widespread testing of thermonuclear weapons in 1952 (Crusius and Anderson, 1995) before significant quantities of Cs-137 began accumulating in the estuarine sediments. The Caño de Martín Peña site displayed significant activity of Cs-137 even in the deepest samples recovered during the December sampling. For this site, the base of the layer with the highest activity was assigned the year 1965, 2 years after 1963 (the year of maximum fallout). Sedimentation rates vary from year to year in response to varying amounts of runoff and changing sediment yields; however, for purposes of defining chronostratigraphic units, a constant sedimentation rate was assumed and the horizons representing the time intervals from 1925-49, 1950-74, and 1975-95 were defined. For trace elements and organic contaminants, a decadal scale sampling unit was used, so only longer-period fluctuations in precipitation and sediment yields would invalidate the assumption of constant sedimentation rate as yearly fluctuations would be averaged out. The average sedimentation rate was also used to convert depths to time periods for the results of the phosphorus analysis carried out every 5 cm for the core samples collected in December 1994. Assuming a constant sedimentation rate, the sediments sampled from 60 cm depth in Laguna de Piñones would have been deposited in the early 1700’s. Clearly, the assigned times of deposition are only approximate, but are still useful in appreciating general long term trends as shown in the phosphorus results presented later.

Sediment to be analyzed for trace elements and organic compounds sampled from cores from the six sites were homogenized by intervals representing the time periods 1925-49, 1950-74, and 1975-95. The longest core obtainable at the Caño de Martín Peña site during the July sampling was 185 cm, or 55 cm more than was possible in the December 1994 sampling from which the sedimentation rates were estimated. Therefore, the pre-1950 strata (178 to 185 cm) would represent 1948-49, not 1925-49 as was the case for all other cores. Each homogenized sample was analyzed for 17 organochlorine compounds including total PCB’s and PCN’s, and the trace elements As, Ba, Cd, Cr, Pb, Hg, and Se. Specific methods and discussion of data quality are included in Webb and Gómez-Gómez (1998).

Spatial variations in general water quality in the estuary system were measured at 15 sites on March 15-16, 1995 (dry weather conditions), and again on May 17-18, 1995, following a storm event on May 16, 1995. The mean daily discharges measured at stream gages in the watershed during the storm event on May 16 were exceeded on only 30 days of the 1995 water year (October 1994 - September 1995). Variables measured at each site include water depth, color, and vertical profiles of temperature, specific conductance, pH, and dissolved oxygen. In addition, samples collected near the surface of the water column at each site were analyzed for turbidity, total suspended solids, nitrate, nitrite, ammonia, orthophosphate, fecal-coliform bacteria, chlorophyll ‘a’, and chlorophyll ‘b’. All water and sediment samples were collected using standard USGS field sampling techniques (Ward and Harr 1990; Britton and Greeson 1987). The results from the wet/dry period sampling represent two data points in time from which only limited conclusions on the present day processes can be derived. Therefore, the water-quality data collected at sites in the estuary system’s watershed for the years 1970 through present were retrieved from the National Water-Data Storage and Retrieval System (WATSTORE). The WATSTORE database is populated with data collected by USGS personnel in support of a wide range of missions. Data from 42 sites within the estuarine boundaries have been collected as part of the USGS-Commonwealth of Puerto Rico water-quality monitoring network or the National Stream Quality Accounting Network (NASQAN), both of which were established to provide data to describe areal variability and detection of changes or trends in water-quality characteristics over broad geographical areas. Methods and results of the monitoring effort are published in yearly Water-Resources Data Reports for Puerto Rico and the U.S. Virgin Islands. The most recent one presents the data for water year 1996 (Díaz et al 1996).

The Bahía de San Juan, Caño de Martín Peña, and Laguna San José are classified by the PREQB (1990) as class SC waters, coastal waters intended for uses where the human body may come into indirect (secondary) contact with the water (such as fishing and boating), and for use in propagation and preservation of desirable species. The Laguna La Torrecilla and Laguna de Piñones are classified as SB waters, coastal waters and estuarine waters intended for use where the human body may come into direct (primary) contact with the water (such as bathing or swimming), and for propagation and preservation of desirable species. The rivers are classified as SD waters, surface waters intended for use as a raw source of public water supply, propagation and preservation of desirable species, and primary and secondary contact recreation. The water bodies in the San Juan metropolitan area are impaired and only partially support their designated uses ((USACE 1978; PREQB 1984; USEPA 1992).

The estimated sedimentation rates were similar at all sites with the exception of Caño de Martín Peña. The lowest sedimentation rate was 0.24 cm/yr in an isolated area of a mangrove-bordered lagoon and the greatest, 3.9 cm/ yr in a sediment-filled, natural tidal channel. Sedimentation rates exceeding 10 cm/yr are common where terrestrial runoff enters the low-energy waters over deep dredged areas in Bahía de San Juan, Caño de Martín Peña, Laguna San José, and Laguna La Torrecilla. The peak activity of Cs-137 measured in the strata sampled at site SJN600 in the Bahía de San Juan was 3.6 mBq/g in the top 5 cm of the sediment core; Cs-137 activity was not detected above the analytical reporting limit of 1 mBq/g at depths below 15 cm. The Caño de Martín Peña site, MPN500, receives significant loads of sediment from the Río Puerto Nuevo in addition to numerous combined sewer outfalls located along its southern and northern banks. The peak Cs-137 activity measured in the strata sampled at the Caño de Martín Peña site was 31.2 mBq/g, in the sediments sampled from 110 to 120 cm, which was similar to activities measured in Lago Loíza, a nearby reservoir. Maximum Cs-137 activity in the sediments cored from Laguna Los Corozos (site CRZ400) was 11.9 mBq/g, in the strata between 5 and 10 cm. In southern Laguna San José (site SJS300) the maximum Cs-137 activity was 12.9 mBq/g, in the strata between 10 and 15 cm. Circulation is restricted in Laguna San José (and its northern embayment, Laguna Los Corozos) and loads that enter the lagoon from sources around its perimeter are likely to be deposited there. This explains why the sedimentation rates were calculated to be higher at the two remotely located sites in Laguna San José than they were for the site in the Bahía de San Juan. The remaining two core sites in Laguna La Torrecilla (TRC200) and Laguna de Piñones (PNN100) receive only minimal amounts of fine-grained terrigenous sediments. Maximum Cs-137 activities for these sites (4.5 mBq/g for TRC200 and 1.5 mBq/ g for PNN100) were measured in the top 5 cm of the sediment cores. In order of increasing values, the average sedimentation rates were determined to be 0.24 cm/yr (PNN100), 0.37 cm/yr (TRC200), 0.49 cm/yr (CRZ400), 0.61 cm/yr (SJS300), 0.61 cm/yr (SJN600), and 3.9 cm/yr (MPN500). Storm runoff deposits large amounts of sediment into Caño de Martín Peña resulting in the high sedimentation rates observed there.

The results of the bottom-sediment survey can be used to put in perspective the impact of urban development in the watershed and degradation of the environmental water quality of the estuary system. For brevity, the following discussion will focus on total phosphorus measured in the surface waters and underlying sediments. In the San Juan Bay estuary system, areas with elevated concentrations of total phosphorus experienced wide variations in oxygen saturation levels and contained elevated concentrations of chlorophyll, ammonia, and bacteria. The spatial variations in total phosphorus contained in the surficial sediments reflect the average concentrations of total phosphorus measured in the water column (fig. 4).

Pollution control measures have improved overall water quality in the Bahía de San Juan, Quebrada Blasina, Laguna La Torrecilla, and Laguna de Piñones. However, a general degradation in water quality in the Río Piedras, corresponding to intensive urbanization in the Río Piedras uplands, has also been measured.

The most significant changes in nutrient concentrations measured in samples collected from the water column during the past 20 years has been at Laguna San José, Laguna La Torrecilla, and Laguna de Piñones (fig. 4). The decline in nutrient concentrations at Laguna La Torrecilla and Laguna de Piñones can be related to the diversion in 1985 of wastewater discharged previously to Quebrada Blasina from the Vistamar, Carolina, and Round Hills sewage treatment plants to regional wastewater treatment facilities that now discharge to ocean outfalls. While concentrations of nutrients in the surface waters sampled in Laguna Los Corozos are declining, the Baldorioty de Castro storm-water pumping station can be observed to operate almost continuously, discharging contaminated waters into the area even during dry periods (PREQB, 1984).

The most contaminated area, Caño de Martin Peña, had elevated ammonia concentrations (2.3 mg/L as nitrogen) and counts of fecal-coliform bacteria on the order of 100,000 col./100 mL. At none of the other sampled sites did fecal coliform concentrations exceed 6,000 col./100 mL, nor did dissolved ammonia exceed 0.45 mg/L. Concentrations of dissolved oxygen in the hyper-eutrophic Laguna San José varied from anoxic to more than 200 percent saturation levels.

Sediment quality and water quality in the San Juan Bay Estuary system have degraded throughout this century in direct response to intensive land use by agriculture, industry, and the population. Anthropogenic effects on sediment quality are most evident for constituents used in the uplands of the watershed that are subsequently adsorbed onto sediments and transported into the tidal portions of the estuary. Among these are PCB’s (from electrical transformers, high-pressure lubricants, and in insulation of wires and cables), dieldrin (from application to sugar cane crops), Pb (from leaded gasoline and paints), and Hg (from electrical and mechanical components).

Additional sampling of bottom sediments in the estuary is needed to determine if improving water quality is resulting in cleaner sediments in the most recent surficial bottom layers. Trends of increasing contamination observed in the decadal scale units discussed in this study may have since reversed. Organochlorine compounds entered the commercial market during the 1940’s (PCB’s, dieldrin, and DDT and its metabolites). Their concentrations in surficial bottom sediments are expected to decrease in response to microbial degradation and dilution as their use has been discontinued except for limited use of PCB’s in sealed transformers. In addition, the contamination of surficial bottom sediment samples in Laguna del Condado should be evaluated because a thermoelectric power plant existed within the eastern shoreline of the lagoon until about 1960.

Analysis of data collected from 1970 to 1995 shows that since 1985, the quality of water in the San Juan Bay estuary system has improved. Since then, leaded gasoline has been phased out, and wastewater from local plants, previously discharged to inland waters, is now collected and routed to regional facilities for discharge to deep ocean outfalls. An exception is the upland area of the Río Piedras were the trends of generally improving water quality were reversed in the wake of intensive development of the area in the late 1980’s.

Birdsey, R.A. and L. Weaver. 1987. Forest area trends in Puerto Rico: U.S. Department of Agriculture, Forest Service, Southern Forest Experiment Station, Research Note SO-331, February 1987: 1-5.

Britton, L.J., and P.E. Greeson. eds. 1987. Methods for collection and analysis of aquatic biological and microbiological samples: U.S. Geological Survey Techniques of Water-Resources Investigations, book 5(A4): 363 p.

Calvesbert, R.J. 1970. Climate of Puerto Rico and U.S. Virgin Islands, in Climatography of the United States No. 60- 52; Climate of the States: Washington, D.C., U.S. Department of Commerce, Environmental Science Services Administration, Environmental Data Service: 29 p.

Cleveland, W.S. 1979. Robust locally weighted regression and smoothing of scatterplots: Journal of American Statistical Association 74(368): 829-836.

Crusius, John, and R.F.Anderson. 1995. Evaluating the mobility of ¹³⁷Cs, ^239+240Pu, and ²¹⁰Pb from their distributions in laminated lake sediments: Journal of Paleolimnology 13: 119-141.

Díaz, P.L., Aquino, Zaida, Figueroa-Alamo, Carlos, Vachier, R.J., and A.V. Sánchez. 1996. Water resources data, Puerto Rico and the U.S. Virgin Islands, water year 1996: U.S. Geological Survey Water-Data Report PR-96-1: 564 p.

Ellis, S.R. 1976. History of dredging and filling of lagoons in the San Juan area, Puerto Rico: U.S. Geological Survey Water-Resources Investigations Report 38-76: 25 p.

Fishman, M.J., and L.C. Friedman, eds. 1989. Methods for determination of inorganic substances in water and fluvial sediments: U.S. Geological Survey Techniques of Water-Resources Investigations Report. book 5, chap. A1: 545 p.

Gellis, A.C. 1991. Construction effects on sediment for two basins, Puerto Rico: Proceedings of the Fifth Federal Interagency Sedimentation Conference. Las Vegas, NV. Chap. 4: 72-78.

Gómez-Gómez, Fernando, Quiñones, Ferdinand, and S.R. Ellis. 1983. Hydrologic characteristics of lagoons at San Juan, Puerto Rico, during and October 1974 tidal cycle: U.S. Geological Survey Open-File Report 82-349: 34 p.

Marsh, S.P. 1992a. Analytical results for stream sediment and soil samples from the Commonwealth of Puerto Rico, Isla de Culebra, and Isla de Vieques: U.S. Geological Survey Open-File Report 92-353A: 8 p.

Marsh, S.P. 1992b. Analytical results for stream sediment and soil samples from the Commonwealth of Puerto Rico, Isla de Culebra, and Isla de Vieques: U.S. Geological Survey Open-File Report 92-353B: data on a 5.25 inch diskette.

McHenry, J.R., Ritchie, J.C., and C.M.Cooper. 1980. Rates of recent sedimentation in Lake Pepin: Water Resources Bulletin 16(6): 1,049-1,056.

Puerto Rico Environmental Quality Board. 1990. Puerto Rico water-quality standards regulation, as amended: 100 p.

Puerto Rico Environmental Quality Board. 1984. 208 North Metro Region Project: Planning activities developed through U.S. EPA Grant Number P002140-01-9: 500 p.

Puerto Rico Planning Board. 1976. The San Juan City Edges Project: Technical Report partially financed by the National Endowment for the Arts in Washington, D.C.: Grant Number A40-42-12B: 92 p.

Seguinot-Barbosa. José,. 1983. Coastal modification and land transformation in the San Juan Bay area, Puerto Rico: Department of Geography and Anthropology, Louisiana State University, Baton Roque, Ph.D. dissertation: 302 p.

U.S. Army Corps of Engineers. 1978. Special report on Martin Peña Canal: U.S. Army Engineer District, Jacksonville, Florida: 26 p.

U.S. Environmental Protection Agency. 1992. Characterization of use impairments of the U.S. Virgin Islands and Puerto Rico: U.S. Environmental Protection Agency, Marine and Wetlands Protection Branch, December 1992: 196 p., app. A-E.

Ward, J.R., and G.A. Harr. 1990. Methods for collection and processing of surface-water and bed-material samples for physical and chemical analysis: U.S. Geological Survey Open-File Report 90-140: 71 p.

Webb, R.M.T., and Fernando Gómez-Gómez. 1998. Synoptic survey of water quality and bottom sediments, San Juan Bay estuary system, Puerto Rico, December 1994-July 1995: U.S. Geological Survey Water-Resources Investigations Report 97-4144: 70 p.