A thermocline barrier to sedimentation in a small lake in the southeastern US.

Abstract

We suspended sediment traps at 2 m, 4 m, and 6 m near the deepest point (8. 7 m) in a 30 ha eutrophic lake in Georgia, U.S.A., for 13 consecutive months. Aside from one depth collection (2 m in August), the inorganic fraction was always greater than the organic fraction, by a ratio of 7:2 for the year. Inorganic flux reflected rainfall and peaked in December and March, while virtually all of the organic flux took place in March and April. During winter turnover (October through March), sedimentation rates increased linearly with depth. However during the summer, when the lake was stratified, sedimentation rates in traps below the thermocline were essentially constant and less than fluxes in the uppermost trap. The data suggest, therefore, that the thermocline acts as a barrier to sedimentation, and allows microscopic heterotrophs to consume detritus that has paused during its descent. As a consequence, nutrient loss to the hypolimnion is reduced, epilimnetic recycling is increased, and inorganic sediments seem to be redirected to the lake's stream outflow rather than deposited on the bottom.

Key words: metalimnetic plate, nepheloid layer, sedimentation, sediment traps, detritivores

Introduction

Recent investigations have revealed that settling material often collects in the vicinity of thermal discontinuities (e.g. MacIntyre et al., 1995), due to low turbulence and eddy diffusion (Megard et al., 1997) as well as low particle density. This accumulation of detritus forms a turbid layer which is often called a midwater nepheloid layer, and may concentrate food for aquatic detritivores of various taxa (Sanders et al., 1989; Bennett et al., 1990; Pace 1982; McDonough et al., 1986), and alter the sedimentation pattern in the lake.

Nepheloid layers generally occur on the bottom of lakes, but may also occur at the surface or at midwater depths; we refer to midwater nepheloid layers as "metalimnetic plates." While many large lakes contain nepheloid layers, these generally occur on the bottom and are attributed to river input or sediment resuspension, as in Lake Superior (Halfman and Johnson, 1989) and Lake Ontario (Sandilands and Mudroch, 1983). Surface nepheloid layers are often associated with thermal bars (i.e. areas where two water masses of differing temperature meet), and have been reported in Lake Michigan (Chambers and Eadie, 1981) and Lake Superior (Halfman and Johnson, 1989). Few midwater nepheloid layers have been reported, but may also reflect river input (e.g. Halfman and Scholz, 1993). We collected data to determine if the thermocline in a small freshwater system might influence sediment flux; if so, it suggests that nutrients may be more effectively recycled in the epilimnion than has been suspected previously.

Most lakes in the southeastern U.S. become stratified in the spring, developing two layers based on temperature; the epilimnion is the upper, warmer, well-illuminated zone that overlies the deeper, cooler, and darker hypolimnion which, in the case of most eutrophic lakes, is rapidly depleted of oxygen. The boundary between these two zones is the thermocline, defined as the depth of the maximum temperature gradient. The zone of decreasing temperatures in midwater, which includes the thermocline, is the metalimnion.

Lake Oglethorpe, Georgia (33[degrees]52, 12"N, 83[degrees]13'49"W) has been the focus of much limnological research, recently summarized by Porter et al. (1996). The lake is eutrophic, with an area of 30 ha and a maximum depth of 8.7 m. During summer stratification, bacterial production is high, the plankton consists of cyanobacteria and large algae, and grazers include protozoans (flagellated and ciliated), rotifers, cyclopoid copepods, and Chaoborus (an insect larva: Porter et al., 1996). The metalimnetic plate is well developed, and includes elevated concentrations of heterotrophs and chlorophyll-a. In the winter, the lake is isothermal due to atmospheric cooling; consequently, the lake is unstratified and mixes through all depths (it "turns over"). During this time, nutrients are recycled from the deepwater and sediments. The winter plankton include fewer bacteria and cyanobacteria, and the grazers are largely represented by crustaceans (especially the copepod Diaptomus: Pace and Orcutt, 1981).

Materials and Methods

We suspended sediment traps at a site in Lake Oglethorpe that was 8.5 m deep, attaching the traps to the line of an automatic-rewind clothesline reel which was anchored with steel weights; the free end was attached to a buoy. We attached three "chandeliers" to this line, each consisting of a 50 cm PVC disk with eight PVC tubes (49 mm diameter, 350 mm length) rising vertically from the disk (Fig. 1). The bottom end of each tube had been threaded to fit 49-mm-diameter jars. A drain spout was added to each tube approximately 1 cm above the disk to drain most of the water in the tubes during jar removal. The chandeliers were held in a horizontal position by harnesses and deployed so that the trap openings were 2 m, 4 m, and 6 m below the surface of the lake. Initial deployments included traps with no preservative, with saturated saline solution, and with salt-saturated Lugol's solution.

[FIGURE 1 OMITTED]

Sediments were collected monthly for thirteen consecutive months beginning March 1983. During each visit we also measured the temperature-oxygen profile with a YSI temperature-oxygen meter. Jar contents were filtered onto Whatman GF/C filters, dried at 50[degrees]C for at least 48 hours, weighed on a Mettler H33AR balance ([+ or -]0.1 mg), ashed at 450[degrees]C for 4 hours, and weighed again.

Results and Discussion

We initially compared jars with no preservative to jars with saturated saline solution and others with salt-saturated Lugol's solution. Jars deployed with salt-saturated Lugol's solution tended to collect more material, largely in the form of poisoned migrators, so these collections were excluded from further analysis. Of the 39 chandelier collections (three depths for each of 13 months), five are represented by four replicate jars, two are represented by five replicate jars, six are represented by six replicate jars, and the remaining 30 are represented by seven replicate jars. For each chandelier, the standard deviation in the total flux is less than 0.71 g [m.sup.-2][d.sup.-1] (< 20% of the collected mass) except 6 m in March 1983 (s.d. = 1.85 of 16.16 g [m.sup.-2][d.sup.-1]) and 4 m in April 1983 (s.d. = 1.61 of 8.21 g [m.sup.-2][d.sup.-1]). Overall, half of the standard deviations are less than 0.30 g [m.sup.-2][d.sup.-1].

The lake was isothermal (temperature difference of < 2[degrees]C between depths 0.25 m apart) from September into April (Fig. 2). During this time sedimentation rates increased linearly with depth (Fig. 3). (We include April in these months because, although the lake was stratified during our visit on the 29th, it had been unstratified for most of the deployment period.) This increase with depth is expected, because flux at depth is a function of total overlying suspended particles (whether biogenic or inorganic).

[FIGURE 2 OMITTED]

The lake was stratified from May through August, during which time the thermocline depth increased from about 1.4 m to about 5.8 m. In contrast to winter, sedimentation rates during summer stratification were essentially independent of depth (Fig. 3, Table 1). Traps below the thermocline collected very similar masses regardless of depth. This suggests that there is little production and decomposition in the hypolimnion, and therefore that sediment flux below the thermocline reflected flux through the thermocline, rather than the thickness of the overlying water. Late in stratification (in August), fluxes were again linearly related to depth, probably because all traps were in, or immediately below, the epilimnion.

[FIGURE 3 OMITTED]

During much of the stratified period, flux into hypolimnetic traps was often less than flux in the uppermost trap, indicating the interception of material in the vicinity of the thermocline. Therefore, rather than sinking into the hypolimnion, epilimnetic detritus may be exported from the lake via its surface outflow. More importantly, Lake Oglethorpe often has a pronounced turbid layer (a "metalimnetic plate") lying just above the thermocline, which includes bacteria (McDonough et al., 1986, Porter et al., 1996) and heterotrophic micro flagellates (Bennett et al., 1990). Heterotrophic flagellates are the dominant grazers in Lake Oglethorpe (Sanders et al., 1989), and their distribution is strongly influenced by the distribution of their food (Pace 1982). Apparently these heterotrophs are taking advantage of the greater concentration of settling particles at the thermocline.

These observations have been supported by later studies. For example, Megard et al. (1997) report that zooplankton densities in western Lake Superior often peak at the thermocline. We suggest that such behavior by grazers not only conserves their swimming energy, but allows them to feed on settling matter that has concentrated at the thermocline. Hansson (1996) suggested that the thermocline acts as a barrier to vertical migration by algae (in this case, upward from the hypolimnion). MacIntyre et al. (1995) found a strong correlation between the vertical distribution of marine snow and density discontinuities off central California. Finally, Halfman and Scholz (1993) report a midwater turbid layer in Lake Malawi,Africa, linked to river inputs.

In conclusion, a thermocline can have significant effects on the distribution of settling matter and on biota. While nepheloid layers have been reported from large lakes worldwide (e.g. Chambers and Eadie, 1981; Halfman and Johnson, 1989; Halfman and Scholz, 1993), these are often benthic and composed of inorganic sediments from sediment resuspension or river input. Our study indicates that, at least in eutrophic lakes, thermoclines may create organic metalimnetic plates, and may greatly reduce organic flux to deeper waters. Consequently, it provides an environment suitable for detritivores, whose grazing probably returns important amounts of matter and energy to the planktonic food web.

Table 1. Average sedimentation rates for the limnological
seasons in Lake Oglethorpe. Entries are total monthly flux
as dry g [m.sup.-2] [d.sup.-1] and standard deviations.

                 Stratified season   Isothermal season
Trap depth (m)   (May-Sep)           (Oct-Apr)

     2.0         3.8 [+ or -] 0.4     6.0 [+ or -] 0.4
     4.0         3.9 [+ or -] 0.5     8.7 [+ or -] 0.8
     6.0         4.2 [+ or -] 0.4    11.5 [+ or -] 0.6

Acknowledgements

We are grateful to Yvette Feig, Bob McDonough, and Bob Sanders for field and lab assistance, and to Judy and Gene Helfmeyer for access to the lake. We are also appreciative of the comments and suggests made by Erika McPhee-Shaw on a previous version of the manuscript. This study was supported by National Science Foundation grants DEB 8005582 and DEB 8003254.

References

Bennett, S.J., R.W. Sanders, and K.G. Porter. 1990. Heterotrophic, autotrophic, and mixotrophic nanoflagellates: Seasonal abundances and bacterivory in a eutrophic lake. Limnol. Oceanogr. 35:1813-1832.

Chambers, R.L. and B.J. Eadie. 1981. Nepheloid and suspended particulate matter in south-eastern Lake Michigan. Sedimentology 28:439-447.

Halfman, B.M. and T.C. Johnson. 1989. Surface and benthic nepheloid layers in the western arm of Lake Superior, 1983. J. Great Lakes Res. 15:14-25.

Halfman, J.D. and C.A. Scholz. 1993. Suspended sediments in Lake Malawi, Africa: a reconnaisance study. J. Great Lakes Res. 19:499-511.

Hansson, L.-A. 1996. Algal recruitment from lake sediments in relation to grazing, sinking, and dominance patterns in the phytoplankton community. Limnol. Oceanogr. 41:1312-1323.

Megard, R.O., MM. Kuns, M.C. Whiteside, and J.A. Downing. 1997. Spatial distributions of zooplankton during coastal upwelling in western Lake Superior. Limnol. Oceanogr. 42:827-840.

MacIntyre, S., A.L. Alldredge, and C.C. Gotschalk. 1995. Accumulation of marine snow at density discontinuities in the water column. Limnol. Oceanogr. 40:449-468.

McDonough, R.J., R.W. Sanders, K.G. Porter, and D.L. Kirchman. 1986. Depth distribution of bacterial production in a stratified lake with an anoxic hypolinmion. Appl. Environm. Microbiol. 52:992-1000.

Pace, M.L. 1982. Planktonic ciliates: Their distribution, abundance, and relationship to micribial resources in a monomictic lake. Can. J. Fish. Aq. Sci. 39:1106-1116.

Pace, M.L. and J.D. Orcutt Jr. 1981. The relative importacne of protozoans, rotifers, and crustaceans in a freshwater zooplankton community. Limnol. Oceanogr. 26:822-830.

Porter, K.G., P.A. Saunders, K.A. Haberyan, A.E. Macubbin, T.R. Jacobsen, and R.E. Hodson. 1996. Annual cycle of autotrophic and heterotrophic production in a small, monomictic Piedmont lake (Lake Oglethorpe): Analog for the effects of climatic warming on dimictic lakes. Limnol. Oceanogr. 41:1041-1051.

Sanders, R.W., K.G. Porter, S.J. Bennett, and A.E. DeBiase. 1989. Seasonal patterns of bacterivory by flagellates, ciliates, rotifers, and cladocerans in a freshwater planktonic community. Limnol. Oceanogr. 34:673-687.

Kurt A. Haberyan (1), Department of Biology, Northwest Missouri State University, Maryville, MO 64468, USA. Email [e-mail].

Karen G. Porter, Institute of Ecology 00126, University of Georgia, Athens, GA 30602, USA. Email [e-mail]

(1) corresponding author