Paul F. Woods, Hydrologist (Limnology)

U.S. Geological Survey

A recently completed limnological study of Payette Lake, Idaho, documented substantial hypolimnetic dissolved-oxygen deficits in the summer and autumn of 1995 and 1996. In contrast, the 20.5-square-kilometer lake was classified as oligotrophic during 1995-96 on the basis of concentrations of total phosphorus, total nitrogen, and chlorophyll-a.

A new instrumentation package used for post-study monitoring of Payette Lake during 1997 has yielded high-resolution water-column profiles useful for investigating the lake’s hypolimnetic dissolved-oxygen deficit in relation to oxygen demands from the decay of senescent phytoplankton. The new instrumentation collects data at a rate of 8 scans per second for the following variables: depth, temperature, conductivity, dissolved-oxygen concentration and saturation, pH, oxidation-reduction potential, light transmissivity, photosynthetically active radiation, and chlorophyll fluorescence.

The chlorophyll fluorescence and light transmissivity profiles recorded large chlorophyll peaks within the euphotic zone and metalimnion. More importantly, chlorophyll fluorescence peaks within the hypolimnion also were recorded; these represented remnants of euphotic-zone chlorophyll pulses that probably occurred between profiling dates. No instrumentation for in-vivo profiling of chlorophyll fluorescence within or beneath the euphotic zone was available for the 1995-96 study; thus, chlorophyll production in Payette Lake during 1995-96 was likely under-estimated. The underestimation of chlorophyll-a may explain the discrepancy between the lake’s oligotrophic level of chlorophyll-a and its eutrophic level of hypolimnetic dissolved-oxygen deficit.

Payette Lake is in Valley County, one of Idaho’s rural, mountainous areas with a thriving tourism/recreation industry (fig. 1). Concerns over water-quality degradation caused by lakeshore and watershed development led to a series of water-quality studies between the late 1960’s and the early 1980’s (Idaho Department of Health, 1970; U.S. Environmental Protection Agency, 1977; Falter and Mitchell, 1981; Falter, 1984). Evidence of cultural eutrophication and bacteriological contamination was strong enough to prompt construction of a sewage-collection system in the early 1980’s for the developed part of the lake’s shoreline. Water-quality concerns continued into the 1990’s as residential development and recreational use of Payette Lake continued to increase.

The Idaho State Legislature responded to these concerns by passing the Big Payette Lake Water Quality Act in 1993. The Act mandated formation of a water-quality council and development of a water-quality study of the lake and its watershed. The resulting study was a cooperatively funded effort by the Idaho Division of Environmental Quality, which studied the watershed, and the U.S. Geological Survey, which studied the lake. The purpose of the U.S. Geological Survey limnological study, conducted during water years 1995 and 1996, was to determine Payette Lake’s assimilative capacity for nutrients so that its susceptibility to cultural eutrophication could be assessed. Five major tasks were undertaken: (1) Assess physical, chemical, and biological characteristics of the limnetic and littoral zones of the lake; (2) quantify loads of water and nutrients into and out of the lake; (3) develop an empirical nutrient load/lake response model; (4) use the model to simulate the lake’s response to hypothetical alterations in nutrient loads; and (5) estimate the nutrient loads added to the lake by 1994 forest fires that burned one-half of the lake’s watershed. The results of the 1995-96 study are described in a report by Woods (1997).

The 1995-96 study of Payette Lake documented substantial hypolimnetic dissolved-oxygen deficits during the summer and autumn of both years. In contrast, the lake was classified as oligotrophic on the basis of concentrations of total phosphorus, total nitrogen, and chlorophyll-a. The lake’s propensity to develop hypolimnetic dissolved-oxygen deficits, to the point of anoxia, was ascribed to physical limnological factors coupled with a long-term accumulation of oxygen-demanding organic matter produced within the lake or delivered by its watershed.

In water year 1997, the U.S. Geological Survey began limnological monitoring of Payette Lake to provide trend information on trophic state variables (nutrients, secchi-disc transparency, and chlorophyll-a) and the hypolimnetic dissolved-oxygen deficit. The monitoring employed a new instrumentation package capable of high-resolution profiling of water-column variables. The purpose of this paper is to describe how the enhanced resolution obtained with the new instrumentation has been used to investigate the relation between the lake’s hypolimnetic dissolved-oxygen deficit and oxygen demands from the decay of organic matter produced in the euphotic zone by chlorophyll-based photosynthesis.

Payette Lake is a natural lake, formed by glacial activity, and is in the upper watershed of the North Fork Payette River. Outflow from the lake is regulated for irrigation purposes by a small dam completed in 1943; normal drawdown is 1.7 meters (m). The lake surface area and volume, excluding islands, are 20.5 square kilometers (km²) and 0.75 cubic kilometers (km³), respectively. Mean and maximum depths are 36.8 and 92.7 m, respectively, and shoreline length is about 36 kilometers (km). The principal tributary and outlet is the North Fork Payette River.

Payette Lake receives drainage from 373 km² of heavily forested, mountainous terrain. Elevations range from 1,520 m above sea level at the lake outlet to about 2,770 m. The geology is dominated by the Idaho batholith, which is characterized by crystalline igneous rocks. The dominant vegetation is subalpine fir, Engelmann spruce, and lodgepole pine. Major land uses are timber harvesting, recreation, residential development, and sheep grazing. Mean annual precipitation at the lake outlet is 660 millimeters (mm) and, at the watershed divide, is about 1,200 mm (U.S. Forest Service, 1995). Most precipitation is snow during October to May. Ice normally covers the lake from late December to late April.

The new instrumentation package, a Sealogger (Seabird Electronics, Inc., model SBE-25), collects water-column profile data at a rate of 8 scans per second for the following variables: pressure (depth), temperature, conductivity, dissolved-oxygen concentration and saturation, pH, oxidation-reduction potential, light transmissivity, photosynthetically active radiation, and chlorophyll fluorescence. At the recommended descent rate of 0.25 m per second, the Sealogger can obtain about 2,960 data points for each variable within a water-column profile extending the maximum depth of the lake. Either in real time or immediately after the profile is completed, the data can be downloaded to a computer, processed, and graphically plotted onsite.

The Sealogger was used to collect water-column profiles at limnetic stations 1, 3, and 4 (fig. 1) on the following dates in 1997: July 14-15, August 5 and 26, and September 16. The high-resolution, full-depth profiles yielded evidence of wide variations in the profiled water-column characteristics over depth increments of less than 1 m. Evaluation of the graphical plots provided better insight into the interaction of physical, chemical, and biological processes throughout the water column.

Of particular interest were the profiles of chlorophyll fluorescence and light transmissivity because they provided new information on the distribution of chlorophyll throughout the water column. Peaks in chlorophyll fluorescence often were inversely related to light transmissivity (fig. 2), which corroborated the presence of additional chlorophyll-bearing phytoplankton within the fluorescence peaks. Chlorophyll fluorescence within the euphotic zone varied widely and frequently changed substantially over depth increments of less than 1 m (fig. 3). Chlorophyll fluorescence tended to peak near the lower limit of the euphotic zone (figs. 3 and 4), which ranged from 2.5 to 4.5 m beneath the thermocline. More importantly, hypolimnetic peaks of chlorophyll fluorescence also were detected (figs. 5, 6, and 7); these represented remnants of euphotic-zone chlorophyll pulses that occurred between profiling dates.

The high-resolution profiles of chlorophyll fluorescence obtained in 1997 prompted a reevaluation of some of the results of the 1995-96 lake study. No instrumentation was available during that study for in-vivo profiling of chlorophyll fluorescence or light transmissivity within or beneath the euphotic zone. On the basis of the 1997 results, chlorophyll production in Payette Lake during 1995-96 was likely underestimated because the presence of metalimnetic and hypolimnetic chlorophyll layers was unknown.

The likely underestimation of chlorophyll production in 1995-96 may partly explain the discrepancy between the lake’s low biological production and its substantial hypolimnetic dissolved-oxygen deficit. The annual geometric mean concentration of chlorophyll-a during 1995-96 was 1.3 micrograms per liter (m g/L) (Woods, 1997); this concentration is within the oligotrophic range on the basis of Ryding and Rast’s (1989) open-boundary classification system for trophic state. In contrast, Payette Lake’s average hypolimnetic dissolved-oxygen deficit was 600 milligrams per square meter per day [(mg/m²)/d] during 1995-96 (Woods, 1997); according to Hutchinson (1957), a hypolimnetic dissolved-oxygen deficit in excess of 550 (mg/m²)/d is within the eutrophic range.

The hypolimnetic dissolved-oxygen deficit calculated using the 1995-96 dissolved-oxygen data was much larger than that predicted by the nutrient load/lake response model applied to Payette Lake (Woods, 1997). That model (Walker, 1996) empirically relates the hypolimnetic dissolved-oxygen deficit to nutrient and chlorophyll-a concentrations within the epilimnion. The model would underestimate the hypolimnetic dissolved-oxygen deficit because the underestimation of chlorophyll-a production would, in turn, cause underestimation of the oxygen demand exerted by decomposition of senescent phytoplankton.

Many sampling designs are available for evaluating chlorophyll-a concentrations in lakes. Chlorophyll-a in Payette Lake was sampled during 1995-96 by compositing three volume-weighted, point-depth samples collected from the euphotic zone with an opaque VanDorn sampler. For example, if the euphotic-zone depth was 10 m (measured with a spherical quantum sensor), then point samples were obtained at depths of 1.6, 5.0, and 8.4 m and composited. On the basis of figures 2-4, this sampling design likely caused underestimation of chlorophyll-a concentrations because maximum concentrations were often deep in the euphotic zone or were in the metalimnion. During 1997, a composite sample of chlorophyll-a was obtained through a weighted piece of tubing (12.5-mm diameter) lowered over the depth of the euphotic zone. This sampling design accounted for the variation in chlorophyll-a within the euphotic zone; however, it would not have provided a means for detecting the maximum chlorophyll-a concentration beneath the euphotic zone, as illustrated in figures 5-7.

Deep-lying chlorophyll layers have been reported (Fee, 1976; Priscu and Goldman, 1983; Pick, et al, 1984; and Woods, 1992). In lakes sampled during these studies, as well as in Payette Lake during 1995-97 studies, the euphotic zone was deeper than the epilimnion. Under that condition, the phytoplankton circulating within the epilimnion remain exposed to amounts of photosynthetically active radiation sufficient for photosynthetic production of carbon in excess of respiratory demands.

The designation of a lake’s trophic state can be influenced by the effects of sampling design. Woods (1986) used a large data base of chlorophyll-a concentrations from a 2-year limnological study of Big Lake, in south-central Alaska, to illustrate the effect of three different sampling designs on designation of trophic state. Design I was used at Big Lake; chlorophyll-a concentrations were measured biweekly using a combination of in-vivo fluorometry and numerous discrete-depth samples. The range in concentration for the 296 samples was 0.05 to 46.5 ?g/L; the larger concentrations often were measured in samples obtained near the lower limit of the euphotic zone or within the upper metalimnion. Designs II and III are subsets of Design I. Design II consisted of three biweekly samples obtained from the upper, middle, and lower depths of the epilimnion. Design III consisted of a single biweekly sample obtained near the lake surface. On the basis of the open-boundary, trophic-state classification system of Ryding and Rast (1989), results from Designs II and III indicated an oligotrophic lake, whereas results from Design I indicated a mesotrophic lake. The larger chlorophyll-a concentrations obtained by using Design I were attributable to sampling of the deep-lying chlorophyll layers near the bottom of the euphotic zone.

Chlorophyll-a concentrations in Payette Lake were underestimated during a 1995-96 study because the sampling design used during the study did not provide the means to detect substantial chlorophyll-a layers beneath the euphotic zone. A revised sampling design will be used to quantify chlorophyll-a concentrations throughout the water column of Payette Lake. Full-depth profiles of chlorophyll fluorescence will be obtained with in-vivo fluorometry. On the basis of the profile data, discrete-depth samples will be collected for extractive analysis of chlorophyll-a to accurately quantify chlorophyll-a throughout the water column.