|
System-wide Monitoring Program
Synthesis of the Water Quality Data from 1995 to 2000 Chapter 2: Characterizing Trends in NERR SWMP Data Introduction AThe principal objectives of the NERR SWMP are to track short-term variability and long-term changes in a suite of water quality variables at representative estuarine ecosystems and coastal watersheds throughout the U.S. (Wenner and Geist 2001). Short-term variability (i.e., tidal and daily cycles) in water quality variables are presented in Chapter 5 of this report. This chapter focuses on characterizing seasonal, intra-seasonal and inter-annual variability in water quality parameters monitored at NERR SWMP sites. Diverse biogeography and disparities in watershed size, water body dimensions and physiography impede the ability to compare and contrast NERR SWMP sites. To facilitate comparison among sites, representative values (metrics) for each water quality parameter were calculated. Representative values were defined with the intention of characterizing the frequency that sites experience ecologically or physiologically significant extreme values for these parameters on a seasonal or annual basis. Statistical models used to test for differences between sites within each Reserve were based on deployment-level metrics; however, seasonal and annual trends were assessed. A recent synthesis of NERR SWMP data (Wenner et al. 2001) utilized multiple statistical models to test for differences between two sites within each Reserve. Because eight Reserves in this synthesis now include three replicate sites, and in an effort to improve the ability to statistically compare between Reserves for a given region, we used three-way Analysis of Variance (ANOVA) to test for differences among site, season and year. Methods Water Quality Metrics Two water quality variables, water temperature and DO (% sat), remained unchanged from the previous synthesis project. Metrics for these parameters were defined as the overall percent of time water temperature was <10°C or >25°C and the percent of time (average of the first 48-hours post-deployment) with hypoxia (<28% sat) and supersaturation (>120% sat) during summer (Jul-Sep). Mean water depth was defined as the mean depth measured by each YSI (from sensor to surface) plus the vertical relief (0.1-1m) between the sensor and the bottom sediment. Where YSI’s were mounted to floating objects or distance between the sensor and the bottom could not be determined, mean depth between mean high water (MHW) and mean low water (MLW), obtained from annual metadata or the Wenner et al. (2001) synthesis report, were used instead. Mean daily depth range was also included to represent mean daily tidal range. Overall mean salinity was again used as a metric, but was complemented by the inclusion of mean daily salinity range. Metrics for pH and turbidity (1999-2000 data only) were added to the current synthesis project and were defined as the overall percent of time pH is <7.0 or >8.0 and the overall percent of time that turbidity is >25 NTU, respectively. Seasonal and Intra-seasonal variability Descriptive examination of seasonal and intra-seasonal variability in water temperature, salinity, and precipitation was undertaken on a regional basis. Monthly precipitation data were obtained from the National Climatic Data Center (NCDC). Precipitation data for the nearest NCDC Weather Station were used; however, on occasion, data from multiple NCDC stations were used when data records were incomplete. Seventy percent of NCDC stations were located within 26 km of NERR SWMP sites. Monthly precipitation records from NCDC stations located 26-82 km away (mean = 40 km) were sometimes used because these stations represented the next closest station for which data for months with missing data were available. Two Reserves (NIW, NOC) provided partial precipitation data from stations closer to YSI sites than NCDC stations. Hypoxic Duration Duration of annual hypoxic events during the first 48-hours of each deployment were examined to characterize continuous hypoxic events at NERR SWMP sites between 1995-2000. Although deployments from all seasons were considered, only deployments with a full complement of records (96 observations) during the first 48-hours were analyzed. Hypoxic events were sorted into one of seven time classes (<4h, 4-8h, 8-12h, 12-16h, 16-20h, 20-24h, and >24h). To compensate for seasonal and inter-annual sampling variability among sites and regions, observations of hypoxic duration were extrapolated to a predicted annual frequency to facilitate comparisons on a standardized scale. Extrapolation was made using the following general formula: (n deployments) x (97 observations per deployment) ¸ (maximum annual observations). This approach was used by Wenner et al. (2001) to compare duration of hypoxia among NERR SWMP sites. Analysis of Variance Input data for ANOVAs were modified from last year for several parameters. Summary statistics for mean salinity, pH, and turbidity during the first 7-days of each deployment, rather than mean daily values, were used to decrease overall sample size to a level which would not de facto produce statistically significant results due to an excessively large n. Summary statistics for hypoxia and supersaturation were defined as the percent of time during the first 48-hours post-deployment, the same definitions used by Wenner et al. (2001). Data fom 1996-2000 were used for dissolved oxygen, temperature and salinity; however, incomplete data between 1996-1998 prompted us to only use 1999-2000 data for pH and turbidity. The three-way ANOVA with interaction terms was first used to examine the data. If interaction terms were not significant at p < 0.05, they were removed from the model. Data were then tested for Normality (Shapiro-Wilk) and Heterogeneity of Variance (visual observation of residuals vs. predicted values) after placement in the three-way ANOVA model. To evaluate normality when using an ANOVA model, the appropriate test should be conducted for each of the defined groups (SAS 1987). Due to the complexity and difficulty of trying to run normality and homogeneity tests on each of the 16-60 groups, normality and homogeneity were tested on model residuals. If data were normal and homogeneous, values were not transformed for the model. If data were neither normal nor homogenous, then data were log-transformed (i.e., salinity, turbidity, pH) or arcsine-transformed (i.e., hypoxia, supersaturation) for the model. If data transformation still did not produce normal and homogenous data, then the data were ranked before being analyzed using the three-way ANOVA model. This ranking approach is a viable statistical alternative when assumptions are violated (SAS 1989). Results and Discussion Water Quality Metrics Water temperature <10°C was regularly experienced at monitoring sites in the Pacific Northwest and in the Mid-Atlantic and the Northeast regions (Appendix 10). Sites in these regions typically experienced water temperatures <10°C between 25-60% annually. Exceptions to this observation include sites at five Reserves (South Slough, Great Bay, Hudson River, Old Woman Creek, and Chesapeake Bay MD) that typically do not deploy or scale back deployment of YSI’s during the winter months. Conversely, water temperature >25°C was regularly experienced at monitoring sites in the Southeast, Gulf of Mexico, and Puerto Rico (Appendix 11). Sites in these regions typically experienced water temperatures >25°C between 25-60% annually, with sites in Puerto Rico experiencing water temperatures >25°C more than 90% of the year. Two Reserves in California (Tijuana River Estuary and Elkhorn Slough) did not experience either water temperature extreme combined more than 15% annually. Large inter-annual variation in percent of time with water temperature extremes was noted for numerous sites and may have been partially due to the amount of annual data collected at these sites (Appendix 1). Summertime (Jul-Sep) hypoxia during the first 48-hours post-deployment was highly variable among deployments for a given site between 1995-2000 (Appendix 12). Ten sites regularly experienced hypoxia more than 15% of the first 48-hours post-deployment. Four sites (ELKAP, ELKSM, TJROS, TJRTL) were located on the West Coast, two sites (CBMJB, DELPB) were located in the Mid-Atlantic, one site (SAPHD) was located in the Southeast, and all three sites at the Rookery Bay Reserve in the Gulf of Mexico. Summertime supersaturation during the first 48-hours post-deployment was also highly variable among deployments for a given site between 1995-2000 (Appendix 13). Seven sites regularly experienced supersaturation more than 15% of the first 48-hours post-deployment. Two sites (ELKAP, TJRTL) were located on the West Coast, two sites (WQBCB, WQBMP) were located in the Northeast, two sites (CBMPR, CBVGI) were located in the Mid-Atlantic, and one site (WKBWB) was located in the Gulf of Mexico. Mean water depth estimates provided here are slightly deeper than previously reported estimates, which reflected the mean depth of water above the sensor, rather than the mean depth of water of the water body where YSI’s were located (Wenner et al. 2001). Mean depth for all but two monitoring sites (PDBBY, SOSVI) on the West Coast was less than 2m (Appendix 14). Mean daily depth range at West Coast sites ranged from 0.1m (ELKNM) to 2.3-2.4m (PDBBY, SOSVI). Mean depth at sites in the Northeast ranged from < 2m (n=6), 2-4m (n=6), and >4m (n=2). Mean daily depth range at Northeast sites ranged from 0.1m for three non-tidal freshwater sites (OWCSU, OWCWM, and HUDSK) to 2.2-2.6m for sites at the Great Bay and Wells Reserves. Mean depth for most sites in the Mid-Atlantic, Southeast, and Gulf of Mexico was between 1-2m with depth at seven sites (MULB6, CBMPR, four SAP sites, and WKBFR) between 2-6m. Mean depth at four sites (DELPB, CBMJB, both JOB sites was less than 1m. Mean salinity at NERR SWMP sites ranged from 0-37 ppt and mean daily salinity range varied from 0-19 ppt (Appendix 15). Mean salinity at all but three sites (SOSSE, SOSWI, PDBJL) on the West Coast was >25 ppt. Mean daily salinity range at West Coast sites varied approximately 2-8 ppt for sites with mean salinity >25 ppt, and 17-19 ppt for three sites with mean salinity ~ 10 ppt. Mean salinity and mean daily salinity range at sites in the Northeast was highly variable. Mean salinity for all but one site (MULB6) in the Mid-Atlantic region was less than 20 ppt, whereas mean salinity for all but one site (NIWTA) in the Southeast region was greater than 20 ppt. Mean daily salinity range at sites in the Mid-Atlantic and Southeast regions was variable and ranged from 0-12 ppt. Mean salinity at sites in the Gulf of Mexico was £ 10ppt for two Reserves in the northern Gulf (WKB, APA) and 19-34 ppt for sites at the Rookery Bay Reserve in southwest FL. Minor to moderate daily salinity variation was observed for sites in the Gulf of Mexico. Mean salinity for sites at Jobos Bay, Puerto Rico, was >35 ppt with minor daily variation. Only NERR sites with mean salinity less than 15 ppt experienced pH values <7.0 at least 25% of the time between 1999-2000 (Appendix 16). Half of NERR sites (11 of 21) with mean salinity <15 ppt experienced pH values <7.0 at least 25% of the time. With the exception of SOSSE and SOSWI, other low salinity sites (PDBJL, WELHT, MULBA, DEL NERR, CBM NERR, and NIWTA) that experienced high frequency of pH <7.0 were typically located upstream with minor to moderate tidal influence. Values of pH <7.0 have been associated with reduced bivalve growth (Ringwood and Keppler 2002), increased fecal coliform survival (Solic and Krstulovic 1992), and increased toxicity of ammonia for certain fish species (Thurston et al. 1981). NERR sites with a wide salinity range experienced pH values >8.0 at least 25% of the time between 1999-2000 (Appendix 16); however, NERR sites with mean salinity >15 ppt (16 of 33) experienced pH values >8.0 more frequently than NERR sites with mean salinity <15 ppt (4 of 22). The effects of high pH are not well documented; however, pH >9.5 has been reported to increase phosphate uptake and contribute to algal blooms in tidal freshwater reaches of the Potomac River (Seitzinger 1991). Turbidity >25 NTU, the federal standard for high turbidity (www.epa.gov), was experienced more than 25% of the time in 1999-2000 at several NERR sites (Appendix 17). High turbidity was experienced at one West Coast site (PDBJL), the two Northeast freshwater Reserves (OWC and HUD), and about half of sites in other regions. Seasonal and Intra-seasonal variability Seasonal variation in water temperature was apparent for all regions; however, the extent of seasonal variation in water temperature varied among regions (Appendices 18-22). With the exception of PDBJL, mean water temperatures among West Coast sites varied £ 10°C between seasons. At NERRs in the Northeast and Mid-Atlantic regions, mean water temperature typically varied 10-20°C between seasons; however, seasonal temperatures were shifted by about 5°C between these regions, such that mean summer temperatures in the Northeast and Mid-Atlantic regions were 20°C and 25°C, respectively. Seasonal temperature variation at NERRs in the Southeast was typically 10-15°C and seasonal temperatures were approximately 5°C warmer with respect to NERRs in the Mid-Atlantic. Seasonal temperature variation at NERRs in the northern Gulf of Mexico (WKB, APA) were similar to seasonal variation observed for NERRs in the Southeast, but seasonal temperatures were also about 5°C warmer. Seasonal temperature variation at the RKB and JOB NERRs were <10°C and <5°C, respectively. Intra-seasonal variation was most pronounced during the first sampling season at sites. Seasonal salinity patterns between 1995-2000 were evident for most sites throughout the NERR SWMP. Most sites experienced the lowest mean salinity in winter/spring and the greatest mean salinity in summer/fall, regardless of geography (Appendices 23-27). Exceptions to this trend were observed at the Rookery Bay Reserve, where mean salinity in summer/fall was substantially lower than mean salinity in winter/spring. Seasonal variation in salinity was not well defined for seven sites (PDBBY, WQBMP, WELIN, MULB6, NOCMS, JOB09, and JOB10). Overall salinity at these sites was ³ 28 ppt and most were located in open water systems with minor to moderate daily salinity variation. Seasonal precipitation patterns were in phase with seasonal salinity patterns for sites on the West Coast and at the Rookery Bay NERR where maximum precipitation and minimum salinity occurred in winter and summer, respectively (Appendices 28 & 32). In contrast, seasonal precipitation for most of the other NERRs was six to nine months out of phase with salinity (Appendices 29-31). Seasonal precipitation was typically greatest in the summer or fall for NERRs in Puerto Rico, the northern Gulf of Mexico, the Southeast, and the Mid-Atlantic. At the WKB and APA NERRs, winter and summer were both wet seasons. Seasonal variability in precipitation was not discernable for NERRs in the Northeast. Intra-seasonal variability in precipitation between 1995-2000 was noted for many sites and was most pronounced in fall and winter for sites on the West Coast and in summer for sites on the East Coast. Substantially more precipitation was recorded in winters 1995 and 1998 at the TJR, ELK, and SOS NERRs and in fall 1996 at the ELK and SOS NERRs, than during other years (Appendix 28). At NERRs along the Eastern Seaboard, summers 1996 and 1999 were especially wet, as was winter 1998 at NERRs along the Eastern Seaboard and in the northern Gulf of Mexico (Appendices 29-32). Evapo-transpiration, as well as precipitation, appears to influence seasonal salinity at sites with discernable seasonal patterns. Seasonal variability in water temperature at sites on the West Coast and the Rookery Bay NERR was less pronounced than observed for other NERRs (Appendices 18 & 22); thus, seasonal variability in evapo-transpiration would also be expected to be less severe. At these sites, precipitation may be more important than evapo-transpiration in determining salinity distributions. Seasonal variability in water temperature at sites along the East Coast and northern Gulf of Mexico experience dramatic seasonal variation in water temperature (Appendices 19-22); thus, seasonal variability in evapo-transpiration should reflect this trend. At these sites, maximum precipitation and minimum salinity were out of phase, suggesting that direct precipitation input may not be as important in determining salinity distributions. At these sites, maximum evapo-transpiration in the summer and minimum evapo-transpiration in the winter provides an alternative explanation for seasonal salinity patterns. Seasonal salinity patterns at these sites may also be related to runoff of surface and/or groundwater in the spring. The effects of precipitation in determining salinity distributions appear to be most pronounced in the summer, during periods of maximum evapo-transpiration. During the summer, salinity at many sites was sustained at the annual maximum and, in some cases, the maximum observed between 1995-2000. During periods of maximum annual salinity, daily salinity variation was usually markedly less than observed during other seasons, and often, less than the mean daily variation for that season. This pattern likely reflects the effects of low precipitation and maximum evapo-transpiration during this time. During these periods, abrupt decreases in salinity are evident from annual scatter plots and may have been related to precipitation activity. Abrupt short-term decreases and sustained long-term decreases in salinity associated with precipitation events were documented during the passage of tropical systems in this report. Although several tropical systems had long-term decreasing effects on salinity (see Chapter 6), the immediate effect of these systems on salinity was to restore daily salinity variation patterns at these sites. This scenario was particularly evident for drought-stricken sites at the Great Bay NERR associated with the passage of Hurricane Floyd in 1999. Hypoxic duration Duration of hypoxic events during the first 48-hours post-deployment was examined for all 55 sites with data between 1995-2000. On average, 85% (range = 61-100%) of deployments at each site contained the full 48-hour record and were used in these analyses (Table 1). Of the deployments used in these analyses, 16% (range = 0-74%) of all deployments contained at least one hypoxic (DO <28% sat) event. The percent of deployments used in these analyses was similar among geographic regions; however, the percent of deployments with hypoxic events was substantially greater at West Coast and Gulf of Mexico/Puerto Rico sites (22-28%) than observed for the Mid-Atlantic (15%), Southeast (11%), and Northeast (6%) regions. A total of 1,564 hypoxic events were observed in the deployments examined (Table 2). Thirty-two percent of these events were observed at West Coast NERRs, down 8% from 1996-1998 (Wenner et al 2001). This finding loosely suggests that hypoxic events may have decreased in 1999-2000; however, 11% of West Coast deployments were not examined. Twelve percent of hypoxic events were observed at Northeast NERRs, the same as previously reported. Hypoxic events at Mid-Atlantic, Southeast, and Gulf of Mexico/Puerto Rico NERRs increased 2-4% (12-20% total) from 1996-1998 levels. Frequency of hypoxic duration for 1995-2000 data was similar to frequency of hypoxic duration in 1996-1998 (Wenner et al. 2001). Hypoxic events lasting less than 4 hours decreased by one percent and were compensated for by hypoxic events lasting 12-16 hours, which subsequently increased by one percent. Ninety-five percent of all hypoxic events lasted less than 12 hours, similar to 1996-1998 estimates (Wenner et al. 2001). Table 1. (see PDF) Annual sampling effort among regions, NERR SWMP 1995-2000. Table 2. Overview of deployments used to examine hypoxic duration.
Frequency of hypoxia >12 hours was similar among geographic regions (2-8%); however, frequency of hypoxia <12 hours was different among regions (Table 2). At NERRs in the Northeast and Southeast, hypoxic events <4 hours in duration accounted for 85-88% of total hypoxic events compared to 74-75% of total hypoxic events in the Gulf of Mexico/Puerto Rico and Mid-Atlantic regions and 63% of total hypoxic events at NERRs on the West Coast. Subsequently, greater percentages of hypoxic events lasting 4-8 hours in duration and 8-12 hours in duration were observed for NERRs on the West Coast, and greater percentages of hypoxic events lasting 4-8 hours, 8-12 hours, and >24 hours were observed for Mid-Atlantic and Gulf of Mexico/Puerto Rico NERRs. |
|||
Ocean and Coastal Resource Management | Ocean Service | NOAA | Department of Commerce
Privacy Policy | Disclaimer http://nerrs.noaa.gov/Monitoring/Synthesis5.html