Scientific Investigations Report 2008–5026
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008–5026
Time series of daily median water quality conditions at the two sites in Agency Lake are shown in figure 19. Temperature at both sites was similar throughout the field season. Daily median dissolved oxygen concentrations were at and slightly less than 4 mg/L (a high-stress dissolved oxygen concentration threshold for Upper Klamath Lake suckers [Loftus, 2001]) at site AGN during several days in July and for a few days in August. By contrast, comparable daily median dissolved oxygen concentrations occurred for only a short period in mid-August at site AGS. Through June, both sites recorded a steady increase in pH, with daily median pH reaching around 10 for both sites in late June and early July. In mid-July, however, pH varied considerably at site AGN as compared to site AGS. This variability continued until September, after which pH values were relatively stable at both sites. Early season similarity between sites AGN and AGS was evident in dissolved oxygen concentrations as well. Dissolved oxygen concentration varied more at site AGN than at site AGS beginning in late June. Coincident with pH, dissolved oxygen concentrations became more comparable between the two sites from September through the end of the field season. A similar relation between sites AGN and AGS also was observed in specific conductance values, with site AGN generally showing more variability as compared to site AGS from mid-July through the end of the field season.
Data recorded at continuous water quality monitoring sites in Upper Klamath Lake were combined into a single dataset for each parameter to determine lakewide daily medians (fig. 20). Dissolved oxygen and pH values followed a similar trend of increasing and leveling off through June, then decreasing steadily through much of July. Dissolved oxygen concentration reached a minimum in the last half of July that lasted for nearly 2 weeks through the beginning of August, and pH reached a minimum in early August that lasted for only a few days. Both parameters simultaneously recovered relatively rapidly to predecline levels in August and maintained those levels with some fluctuation for the rest of the field season. Temperature gradually increased through mid-July, maintained a plateau from mid-July through mid-August (coinciding with the period of lowest dissolved oxygen and pH values), and gradually decreased for the remainder of the field season. Specific conductance had an early period of relative stability and then began a steady increase from mid-July, peaking coincidentally with the period of minimum dissolved oxygen and pH values. A sharp, short decrease to fairly steady specific conductance values for the remainder of the field season followed.
Water temperatures in Agency and Upper Klamath Lakes (figs. 19 and 20) closely followed patterns in air temperature (fig. 6). Because these lakes are shallow, their small water columns contain a smaller thermal mass, relative to deep lakes, and therefore heat up and cool down more readily along with changes in air temperature. This process is aided by diurnal mixing known to occur in these lakes (Wood and others, 2006).
Dissolved oxygen and pH values have a positive relation through photosynthetic activity and the carbonate buffering system of natural waters (Wetzel, 2001), which is evident in data from both lakes. Dissolved oxygen and pH values in Agency and Upper Klamath Lakes are related to bloom dynamics, but bloom dynamics in the two lakes are largely independent. The bloom patterns observed at sites AGN and AGS (fig. 11) are evident in the dissolved oxygen and pH data collected at those sites (fig. 19). The highly variable pH at site AGN through much of August corresponds to the variability observed in the AFA bloom at that site during this time. Because there was no late-season bloom in Agency Lake, dissolved oxygen concentrations remained largely undersaturated through the last half of August and September. Thus, pH values did not rebound to values seen earlier in the season, as they did during the late-season bloom in Upper Klamath Lake (fig. 20).
In Upper Klamath Lake, the coincidence of the seasonal maximum in lakewide median specific conductance with the seasonal minimums in lakewide median dissolved oxygen and pH values (fig. 20) also is coincident with maximum concentrations of orthophosphate and total ammonia (fig. 10). An inverse correlation also was observed between chlorophyll a and both orthophosphate and total ammonia (fig. 14). These observations suggest that the ionic nutrients released by decaying algal cells influence the conductance of the water in Upper Klamath Lake. This situation may have occurred at site AGS in late July, when a sharp decrease in dissolved oxygen and pH values was coincident with a sharp increase in specific conductance (fig. 19). However, a similar increase in specific conductance was not recorded at site AGS when dissolved oxygen and pH values quickly decreased with the decline in bloom conditions there in mid-August. Nutrient release and uptake by the variable bloom at site AGN may have contributed to variability in specific conductance, particularly through July and early August. However, the inverse relation of specific conductance with dissolved oxygen and pH values was not as strong during this period at site AGN as in Upper Klamath Lake. The available data indicate that concentrations of other ions were more influential in patterns of specific conductance in Agency Lake. Higher variability in specific conductance at site AGN through the end of the field season could be the result of influent waters from the Wood River and Wood River wetland, to which site AGN is in close proximity relative to site AGS. The specific conductance in the Wood River wetland is dominated by chloride ions (K.D. Carpenter, U.S. Geological Survey, unpub. data, 2008). Similar variability in specific conductance was observed at sites WMR and FBS (sites in the proximity of the Williamson River and west-shore wetlands, respectively) relative to other sites in Upper Klamath Lake.
Periods of very low dissolved oxygen concentrations in Upper Klamath Lake influence endangered sucker movements and have been implicated in die-off events of these species in Upper Klamath Lake (B.J. Adams and others, U.S. Geological Survey, unpub. data, 2005). To study the dynamics of dissolved oxygen in Upper Klamath Lake, time-series graphs of daily median dissolved oxygen percent saturation at each of the water quality monitoring locations (fig. 2) were examined (fig. 21). The graphs are ordered along the primarily clockwise current pattern in Upper Klamath Lake during the summer (fig. 7), beginning with the lower sonde at the middle of trench site (MDT-L). Periods of missing data at site MDT-L resulted from problems positioning the sonde at this site. Data for sites MDL and RPT are not shown but were similar to data from sites MPT and NBI. The graphs indicate that dissolved oxygen dynamics in the lake operated at two time scales in 2005: a long scale (months), and a short scale (days).
In the longer time scale, the dissolved oxygen patterns for sites in the north and west, principally sites MDT-L, MDT-U, EPT-L, EPT-U, SHB, BLB, MDN-L, MDN-U, and EHB, (fig. 21) follow a pattern similar to the lakewide median dissolved oxygen percent saturation (fig. 20). This pattern of gradual decline to substantially lower dissolved oxygen concentrations during the mid-season bloom decline and a recovery to predecline concentrations after the bloom recovered was not as strong at the other sites. This distinction was caused in large part by the bathymetry of the lake: sites MDT-L and EPT-L are in the trench portion of Upper Klamath Lake, where much of the water column lies below the photic zone. This characteristic provides more potential for oxygen consuming processes, such as sediment oxygen demand, algal respiration, and bacterial activity, to deplete dissolved oxygen, contributing to the gradual decline in dissolved oxygen concentration observed at these sites. The upper monitors at these deep sites also recorded this gradual seasonal decline in dissolved oxygen concentration because of frequent mixing within the water column, which is described later in this report. At most of the Upper Klamath Lake sites in the east and south, principally sites WMR, MPT, NBI, and HDB, the seasonal decline in dissolved oxygen was largely dampened (fig. 21), reflecting the smaller aphotic zone and lower potential for oxygen consumption in these shallower areas. Sites SHB, MDN-L, and BLB, although of similar depth to sites in the eastern portion of the lake (table 1), recorded a more pronounced seasonal decline in dissolved oxygen. This may be because sites SHB, MDN-L, and BLB receive water from the trench under the dominant flow pattern in the lake (fig. 7). The decline in dissolved oxygen at these sites is not as severe as that in the trench, however, because of the decreased potential for dissolved oxygen consumption in their shallower water columns.
Variability occurring on the order of days can be seen in the daily median dissolved oxygen percent saturation graphs in figure 21. Events of extremely low dissolved oxygen percent saturation operating at these shorter time scales may indicate movement through the lake of masses of water that are low in dissolved oxygen, with the conditions observed first at one location and then at subsequent sites as water moves along dominant flow paths in the lake (fig. 7). The condition of the mass of water when it reaches the next station along the flow path provides information about the processes that occurred as the water traveled between monitoring locations.
The darkly shaded sections of the graphs in figure 21 highlight an LDOE that was first observed at site MDT-L. From August 19 through 22, daily median dissolved oxygen saturation declined to nearly 5 percent at site MDT-L. Conditions were not as severe near the top of the water column (at site MDT-U) during this time. From August 20 through August 24, a subsequent LDOE occurred at site EPT-L, with daily median percent saturation of dissolved oxygen reaching almost zero. Low dissolved oxygen concentrations were distributed throughout more of the water column at this site. A subsequent decline in dissolved oxygen concentration occurred at site SHB from August 21 through 25. Percent saturation of dissolved oxygen went from supersaturated before the event (about 140 percent) to just undersaturated (about 80 percent) during the event at site SHB. The LDOE was less severe at SHB, demonstrating mixing of the low-dissolved-oxygen water once it enters the northern part of the lake. A subsequent dip in dissolved oxygen saturation was observed at site BLB from August 23 through 27, but there was no evidence of the LDOE at site MDN around this time, indicating that the LDOE was transported by higher velocities across the mouth of Shoalwater and Ball Bays but largely bypassed MDN, and was probably eroded by mixing before reaching FBS, as there was no evidence of the LDOE at that site. The LDOE likely originated in the trench, because dissolved oxygen saturation at sites immediately south of the trench (sites NBI, EHB, and HDB; fig. 2) did not decrease significantly immediately preceding the August 19 event at site MDT-L (fig. 21).
Stratification of lake waters is caused by density differences between the upper and lower parts of the water column. These density differences are primarily the result of water temperature differences, allowing stratification to be identified by differences in temperature between near-surface waters and near-bottom waters. The three sites (MDN, EPT, and MDT) equipped with a continuous water quality monitor at both 1 m from the surface (the “upper sonde”) and 1 m from the bottom (the “lower sonde”) provided the means to examine the dynamics of water column stratification in Upper Klamath Lake.
Vertical variability in water temperature, dissolved oxygen, and pH was calculated as the upper sonde value minus the lower sonde value for each hourly measurement at each site. Time series of these hourly vertical differences are compared with the hourly median wind speed at site MDN for a representative month of the 2005 field season (fig. 22). The diel (daily) pattern of vertical temperature variability observed in figure 22 is known to be typical of Upper Klamath Lake (Wood and others, 2006). This pattern appears to be driven, in part, by diel variations in wind speed. Typically, wind speeds are low enough to prevent water column mixing during the daylight hours and when skies are clear, maximizing solar heating of the water surface. Vertical temperature differences develop accordingly. During the evening, wind speeds increase, typically providing enough wind shear to induce water column mixing and erode stratification built up during the day. Even the deepest sites (EPT and MDT) undergo this diel buildup and breakdown of temperature vertical variability, although vertical temperature differences at site MDT (the deepest site) are consistently much greater than at site EPT. Diel cycles in the vertical variability of dissolved oxygen and pH also were observed, but vertical differences in these constituents were more likely to persist longer than 24 hours, especially at sites MDT and EPT. This likely results because AFA colonies, having the ability to regulate buoyancy, may congregate near the surface, maintaining vertical variability in dissolved oxygen and pH while temperature becomes more evenly distributed.
Periods of lighter wind result in less wind shear at the lake surface and less energy for mixing, allowing stratification to persist longer than one diel cycle. Stratification persisted, for example, at both trench sites from August 27 to August 29. During this period, the minimum vertical variability in temperature was about 1°C at both sites EPT and MDT. Winds rarely were greater than 5 m/s immediately preceding and during this time period. This period of lighter winds still provided enough energy to induce diel mixing at the shallower MDN site.
Water column stratification affects the dissolved oxygen dynamics. During periods of stratification, the upper water column can become decoupled from the lower water column, allowing oxygen consuming processes such as sediment and water column oxygen demand and algal respiration to deplete the lower part of the water column of oxygen, creating potentially unsuitable conditions for fish. Deeper waters in the trench, although typically undergoing a diel mixing of water, have more resistance to mixing than shallower waters because less of the water column is subject to solar heating, allowing for development of a greater temperature (density) gradient. Accordingly, deeper sites have a smaller proportion of water in the photic zone, which allows oxygen consumption processes to have a greater role in dissolved oxygen dynamics. Therefore, the lowest near-bottom dissolved oxygen concentrations during periods of stratification are likely to occur in the trench. This was observed in 2005 as well as in previous years (Wood and others, 2006). Extended periods of stratification may be an important source of chronic stress to fish, but they have not been identified as the most important cause of the extreme LDOEs that lead directly to fish die-offs (Wood and others, 2006).
The timing of daily minimum dissolved oxygen concentrations, daily maximum pH, and daily maximum water temperature were related to the total water column depth at the site. To aid in describing these patterns, histograms displaying the timing of the daily extreme parameter value from representative shallow (HDB), mid-depth (BLB), and deep (EPT) sites are shown in figure 23 for each of these water quality parameters. These histograms display the normalized frequency of occurrence of the daily extreme water quality reading as a function of the hour of day. Shallow sites (HDB, AGS, WMR, NBI, and FBS) had an approximate full pool depth of 2.5 to 2.8 m (table 1). Middepth sites (AGN, SHB, RPT, BLB, and MPT) had an approximate full pool depth of 3.0 m to 3.7 m. Deep sites (MDN, MDL, EHB, EPT, and MDT) had an approximate full pool depth of greater than 4.2 m. Patterns in the timing of daily extreme parameter values appeared over a continuum of water column depth, so the categories of shallow, middepth, and deep are not necessarily delineated by fixed depth.
The daily minimum dissolved oxygen concentration at shallow sites tended to occur around 7:00 a.m. The daily maximum in pH and temperature occurred most frequently in the afternoon to early evening between 3:00 and 8:00 p.m. This pattern reflects the strong influence of the diel cycle of photosynthesis and respiration where the sonde is positioned, which is typically within the photic zone at these sites, and is consistent with findings from data collected in previous years (Wood and others, 2006).
At middepth sites, however, the period of day when daily minimum dissolved oxygen concentration occurred became less well defined. At deep sites, there was no distinct period of the day when daily minimum dissolved oxygen concentrations tended to occur. The time of day when daily maximum pH values were likely to occur broadened at middepth sites to include more of the early morning hours and some late morning hours. At the deep sites, no strong tendency for the timing of daily maximum pH to occur in a specific time window was observed. A similar progression from shallow to deep sites was observed for the timing of daily maximum temperature: a strong likelihood for daily maximum temperature to occur in the early evening at shallow sites, a broadening of this time window into the late evening at middepth sites, and the weakest tendency for the daily maximum temperature to occur during a distinct time window at the deep sites.
The tendency of daily extreme parameter values to occur throughout the day at deep sites is caused by the increased tendency for these sites to develop thermal (density) stratification and the position of the sonde in aphotic waters at deeper sites. Below the photic zone, the rate of consumption of dissolved oxygen by metabolic respiration and other oxygen-demanding processes within the water column is greater than the rate of oxygen production by photosynthesis. Near the bottom, sediment oxygen demand also consumes dissolved oxygen from the water. If the water column develops even a small amount of stratification, dissolved oxygen can reach minimum daily concentrations in the lower part of the water column at any time of day, even during daylight hours. This is because dissolved oxygen concentration below the photic zone will decrease until a mixing event erodes the thermal stratification and mixes the lower layer, which is relatively depleted of dissolved oxygen, with the upper layer, which contains a relatively higher concentration after being in the photic zone over the course of the day. At deeper sites, therefore, the minimum dissolved oxygen concentration does not always occur in the early morning just before dawn, but has more potential to occur throughout the day, depending on when the wind picks up and mixes the water column. Water column stratification and sonde position in the water column affects the timing of daily maximum pH and temperature in a similar manner.
These patterns are similar to those seen in previous years (Wood and others, 2006). However, a continuum of characteristics from shallow to deep sites was more apparent in 2005 because of the wider distribution of sites around the lake and the greater range of site depths.
U.S. Department of the Interior | U.S. Geological Survey
URL: http://pubs.usgs.gov/sir/2008/5026
Page Contact Information: Publications Team
Page Last Modified: Thursday, 10-Jan-2013 18:44:23 EST
