NERRS banner

System-wide Monitoring Program

Synthesis of the Water Quality Data from 1995 to 2000
Executive Summary

A total of 4,135 datasonde deployments conducted at 55 monitoring sites during the first six years (1995-2000) of the NERR SWMP were analyzed in this report. Due to the progressive implementation of the SWMP at NERR sites, total deployments in 1995 were substantially lower than the total number of annual deployments for the other four years, which progressively increased with time. Deployments conducted at NERR sites in the Southeast and Mid-Atlantic regions collectively accounted for approximately half of the total deployments analyzed, primarily due to: (1) the initiation of the NERR SWMP at sites in these regions, and (2) climates more favorable to intensive year-round sampling than in other regions. In general, fewer and longer deployments occurred in winter and shorter, more frequent deployments were conducted in summer. This scenario was particularly evident for West Coast, Northeast/Great Lakes, and Mid-Atlantic Reserves.

Most NERR sites collected 70% or more of the maximum possible water temperature observations for a given season or year. Total observations for salinity, depth and pH were similar to the total number of observations for water temperature, but typically observations for these parameters were 3-7% less than for water temperature. Deletion of suspected anomalous data largely accounts for these discrepancies. Similarly, substantially fewer observations for the same period for turbidity and dissolved oxygen data sets largely resulted from the deletion of suspected anomalous data. With regard to dissolved oxygen, increased deployment duration was shown to increase the percent of time with hypoxic conditions (DO <28% sat) and decrease the percent of time with supersaturation (DO > 120% sat) at many sites. These discrepancies were presumed to result from increased bio-fouling as a result of increased deployment duration, which represents a bias for estimating hypoxia and supersaturation at these sites. Increased hypoxia with deployment duration was most evident at West Coast, Mid-Atlantic, and Gulf of Mexico and Caribbean NERRs. Decreased supersaturation with deployment duration was most noticeable at Southeast and Gulf of Mexico and Caribbean NERRs. These trends became progressively more noticeable between 1995 and 2000, coinciding with reduced rainfall and more drought-like conditions at many of these NERRs.

Improved statistical and analytical approaches were included in this report, with much emphasis placed on identifying seasonal and inter-annual variability of parameters. Three-way ANOVA models revealed several significant inter-annual differences for water quality parameters for many sites; however, inter-annual trends with respect to Reserve or region were not readily apparent. Three-way ANOVA models revealed several significant differences for water quality parameters among sites for each Reserve with respect to season. Water temperature, salinity, and hypoxia events were generally greatest in the summer or fall and lowest in the winter or spring, when water temperature and salinity were lowest and supersaturation was greatest. These findings, particularly for salinity, appeared to be related to seasonal precipitation and evapo-transpiration. Furthermore, these patterns were also observed in metabolic analyses, with maximum heterotrophic conditions generally occurring in the summer or fall and minimum heterotrophic or maximum autotrophic conditions typically occurring in winter or spring.

Periodicity analyses primarily focused on daily and tidal cycle influence on water quality, but also determined how the relative contribution of each of these cycles varied seasonally and inter-annually. Water temperature was primarily influenced by the daily solar radiation cycle; however, water temperature in these shallow estuarine systems could also be strongly influenced by low tide events. Water depth was more influenced by tide than by daily cycles; however, daily (i.e., afternoon winds) and seasonal (i.e., prevalent wind direction) solar radiation cycles also appeared to influence water depth periodicity. Furthermore, seasonal precipitation and evapo-transpiration patterns may have influenced water depth in the shallow water bodies monitored by the NERR SWMP. Dissolved oxygen was primarily influenced by daily solar radiation cycles; however, strong tidal influence was also observed for sites located closer to inlets where cooler, more saturated marine waters replenish warmer, oxygen-reduced waters twice daily. This phenomenon was particularly noticeable during summer months when hypoxia was most prevalent, particularly during daytime low tide events.

Four distinct multivariate analyses were used to explore and develop “natural” relationships between climate and land-use patterns within a watershed and subsequent water quality in the downstream monitored water body. A common grouping in two of the four analyses (principal component analysis and non-linear multidimensional scaling) was the primary division of NERRs based on water temperature, followed by a division in salinity. Sites with low salinity were associated with a high percentage of agricultural land in the watershed and high turbidity, whereas sites with high salinity were associated with a lower percentage of agricultural land in the watershed and lower turbidity. The geographic regional distinctions evident from the water temperature division for both principal component analyses and non-linear multi-dimensional scaling were supported by discriminant function analyses, which grouped NERR sites according to geographic region, with a “mis-classification” error of only 6%. Cluster analyses, used previously in the 1996-1998 NERR SWMP data (Wenner et al. 2001) also roughly grouped the NERRs based on geographic region and latitude, further emphasizing the importance of climate and biogeography in controlling water chemistry of shallow water, estuarine systems.

Analysis of associated changes in water quality during the passage of tropical systems revealed only one consistent trend, an abrupt decrease in water temperature prior to storm passage, with increasing cooling effects strongly related to increasing storm intensity. This cooling effect is widely reported for tropical systems in the open ocean where it has been related to upwelling effects (see Chapter 6); however, the occurrence of this phenomenon in shallow water systems, in some instances hundreds of miles away from the storm center, have received little mention in the literature. Short-term changes in salinity resulted in salinity spikes when storms approached from the sea or decreases when storms passed from the west after having made landfall. Long-term changes in salinity resulted from slow-moving systems depositing copious amounts of precipitation, particularly when the precipitation occurred throughout the watershed of water bodies monitored by NERRs. Altered salinity distributions and excessive runoff from these storms subsequently resulted in ecological disturbances in some of these estuaries, with the return to pre-storm conditions requiring several years to achieve (i.e., Pamlico Sound following the passage of Hurricanes Dennis and Floyd in 1999).