USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008–5201

Back to Table of Contents

Water Quality Monitors

Agency Lake and Upper Klamath Lake Daily Water Quality Conditions

2006 Conditions

Data from all continuous water quality monitoring sites in the open-water areas of Upper Klamath Lake were combined to calculate lakewide daily medians for dissolved oxygen concentration, dissolved oxygen percent saturation, temperature, pH, and specific conductance (fig. 21). Unseasonably warm and sunny conditions in mid-May resulted in an early start to the AFA bloom. This early season bloom was visually confirmed at water quality monitoring sites in Upper Klamath Lake. Also, researchers conducting larval sucker drift sampling at the southern end of the lake noticed large amounts of AFA in their sampling equipment in late May, but relatively small amounts of AFA in early June (Corene Luton, U.S. Geological Survey, oral commun., 2006). Local maximums in daily median dissolved oxygen concentrations and pH coincided with these observations.

After the early season bloom, dissolved oxygen concentrations were constant through June as water temperatures increased, resulting in a slow, steady increase of dissolved oxygen percent saturation. Through June, pH also increased, leveling off in mid-July. Toward late July, dissolved oxygen and pH decreased with the decline of the bloom as water temperatures approached seasonal highs. Seasonal low dissolved oxygen concentrations in Upper Klamath Lake were coincident with seasonal high temperatures but recovered to predecline levels in August as water temperatures cooled. Because the 100 percent saturation concentration of dissolved oxygen decreases as temperature increases, some variation with temperature in the concentration of dissolved oxygen is expected even if the percent saturation remains nearly constant. In Upper Klamath Lake, however, the percent saturation varies with the concentration, and the large magnitude of fluctuations in percent saturation indicates that important sources and sinks of dissolved oxygen are responsible, rather than a simple change in gas solubility with temperature. The seasonal high in temperature and low in dissolved oxygen conditions near the end of July were coincident with a period of low chlorophyll a concentrations at most sampling sites in the lake (fig. 10) that marked a temporary decline in bloom conditions. The decreased dissolved oxygen concentrations during this period resulted from increased oxygen demand from cell senescence and a reduction in the photosynthetic production of oxygen. A midseason minimum in lakewide median pH conditions occurred on August 1, 4 days after the seasonal minimum in lakewide dissolved oxygen concentrations. A slight peak in lakewide median specific conductance also was measured at this time, but after a small decrease, specific conductance continued to increase through mid-August before leveling off for the rest of the season.

Time series of daily median water quality conditions at the two sites in Agency Lake are shown in figure 22. Water temperatures in Agency Lake closely paralleled those in Upper Klamath Lake, and the fluctuations in both lakes followed patterns in air temperature (figs. 7, 21, and 22), as expected in these relatively shallow lakes. Dissolved oxygen concentrations and pH peaked in mid-June in Agency Lake, whereas these conditions reached their peak almost a month later in Upper Klamath Lake. Seasonal low dissolved oxygen concentrations occurred in late July at both Agency Lake sites, but this decrease began earlier at site AGN. As in Upper Klamath Lake, seasonal high temperatures in Agency Lake coincided with seasonal low dissolved oxygen concentrations. The timing of these seasonal extremes was almost coincident between the two lakes; however, the lowest daily median dissolved oxygen concentrations (4.60 mg/L at site AGN on July 15 and 4.53 mg/L at site AGS on July 24) corresponded to a percent saturation of 62.83 percent at site AGN and 63.95 percent at site AGS. The lowest lakewide daily median dissolved oxygen concentration (4.21 mg/L on July 28) in Upper Klamath Lake corresponded to a percent saturation of 58.67 percent. Unlike Upper Klamath Lake, daily median pH and dissolved oxygen in Agency Lake did not recover to predecline levels after late July. These differences also were noted in the 2005 data (Hoilman and others, 2008), and are consistent with the lack of a late season AFA bloom in Agency Lake.

Comparison to 2005 Conditions

Time series graphs of Upper Klamath Lake daily median water quality conditions in figure 23 compare water quality conditions in 2006 to those in 2005. Only data from sites monitoring similar areas of the lake between the 2 years—FBS, WMR, MRM (MPT in 2005), RPT, HDB, EHB, and NBI; as well as the upper and lower monitoring sites at MDN, EPT, and MDT—were included in calculating these lakewide daily medians.

Seasonal patterns in lakewide water quality dynamics were similar between the 2 years. The period with the lowest dissolved oxygen concentrations was late July during both years, and these conditions coincided with the highest lakewide average water temperatures during each year. In both years, a midseason minimum in pH occurred 3 to 4 days after the midseason minimum in dissolved oxygen. In 2005, the midseason minimum in pH corresponded with a maximum in specific conductance, although this correspondence between pH and specific conductance was not evident in 2006. Even with these similarities, notable differences in water quality conditions existed between the 2 years. Maximum lakewide temperatures were higher and minimum lakewide dissolved oxygen concentrations were lower in 2006. In addition, lakewide pH was higher before and after the bloom decline in 2006. Specific conductance was lower than in 2005 throughout most of 2006, partly because inflows were higher in spring 2006, and outflows at Link River were higher through June 2006 than during 2005 (Wood and others, 2008). Residence time and therefore concentration through evaporation was less in 2006 than in 2005, at least through the month of June.

In both years, the relation between daily median dissolved oxygen saturation and daily median temperature was different during the period of bloom decline than during the rest of the season (fig. 24). For purposes of this discussion, the period of bloom decline is defined in terms of dissolved oxygen saturation. In 2005 and 2006, the bloom began when the lakewide daily median of dissolved oxygen saturation decreased to less than 100 percent in July and ended when the lakewide daily median of dissolved oxygen saturation increased to greater than 100 percent in August. Spearman’s rank correlation coefficient analyses of dissolved oxygen and temperature were done separately for data collected during the bloom decline and the rest of the season for both years. In both years, a significant (p <0.0001) positive relation was noted between dissolved oxygen and temperature outside the period of bloom decline. During the 2005 period of bloom decline, no significant relation was observed and a strong inverse relation (r = –0.815) was observed in 2006. The lack of a significant relation in 2005 between temperature and dissolved oxygen during the bloom decline may have been partly due to the relatively narrow range of temperature measured in 2005 (a 2 degree range in 2005 in contrast to a 6 degree range in 2006).

In both years, dissolved oxygen saturation peaked at well over 100 percent saturation (nearly 140 percent in 2005 and about 130 percent in 2006), but the peak occurred at water temperatures well below the seasonal maximum (between about 16 and 22°C in 2005 and between about 18 and 23°C in 2006; fig. 24). In 2006, late season blooms characterized by greater than 130 percent saturation coincided with water temperatures between 13 and 15°C. In contrast, the highest water temperatures of the season in both years coincided with the lowest dissolved oxygen saturation values (about 70 percent in 2005 and about 50 percent in 2006). This relation between dissolved oxygen and temperature was noted in earlier years, particularly 2003 (Wood and others, 2006). In contrast to the observations for Upper Klamath Lake, other studies have shown that the optimum temperature for AFA can be as high as 29°C (Tsujimura and others, 2001). The fact that the highest temperatures are associated with bloom declines in Upper Klamath Lake is an indication that the effect of temperature on metabolic processes is secondary to other undetermined factors that precipitate a bloom decline.

Nearshore Water Quality Monitors

Prior to 2006, some water quality monitoring was done in the nearshore areas of the lake in support of juvenile sucker sampling, but no long-term records were obtained for an entire season that could be compared to the records from monitors in open-water. The five nearshore continuous water quality monitoring sites operated in 2006 (fig. 2) provided continuous records of water quality and were maintained identically to the open-water sites. This continuous monitoring of nearshore sites enables a comparison of water quality variables and their dynamics between the open-water and nearshore areas of Upper Klamath Lake. Nearshore site selection was largely based on consideration of areas known to be occupied by juvenile suckers.

The comparisons generally showed some differences between open-water and nearshore water quality dynamics, but these differences were largely due to the influence that data from monitors positioned in the lower part of the water column at deep sites had on the composite data set from open-water monitors. When data from the monitors positioned at the deep sites were removed, differences were not as great between open-water and nearshore areas. This is not surprising, given the differences in water quality dynamics between deep waters and shallow waters in Upper Klamath Lake (Wood and others, 2006; Hoilman and others, 2008). Because most of Upper Klamath Lake consists of shallow waters outside of the trench, comparisons of these areas of the open-water were made to nearshore water quality dynamics. These comparisons provided the most relevant answers to questions of similarity between open-water and nearshore areas of Upper Klamath Lake.

Water quality data for the nearshore monitors were combined in the same manner as open-water data to create graphs of daily median water quality conditions (fig. 25). Daily median temperature was higher in nearshore areas for most of the field season, whereas pH was generally lower or the same as open-water conditions. Seasonal patterns of, and relations between, dissolved oxygen, temperature, and pH in the open-water also were observed in nearshore water quality dynamics. Spearman’s rank correlation coefficients (table 5) supported these observations.

Back to Table of Contents