Linkagesbetween chemical Section 3 and physicalresponses and biologicalpopulations

Limnol. Oceanogr., 41(5), 1996, 1041-1051 0 1996, by the American Society of Limnology and Oceanography, Inc. Annual cycle of autotrophic and heterotrophic production in a small, monomictic Piedmont (Lake Oglethorpe): Analog for the effects of climatic warming on dimictic Karen G. Porter and Patricia A. Saunders Institute of Ecology, University of Georgia, Athens 30602 Kurt A. Haberyan Department of Biology, Northwest Missouri State University, Maryville 64468 Alexander E. Macubbin Department of Experimental Therapeutics, Roswell Park Cancer Institute, Buffalo, New York 14263 Timothy R. Jacobsen Cumberland County College, Vinelands, New Jersey 08360 Robert E. Hodson School of Marine Sciences, University of Georgia, Athens 30602

Abstract Lake Oglethorpe, Georgia, is a small, monomictic impoundment in the southeastern U.S. Piedmont region. Seasonal differences in ecological processes correspond to the two major stratification patterns and result in seasonal differences in pelagic community structure,. The winter mixed period (November-March) provides light and temperatures suitable for active nutrient regeneration and uptake, population growth of edible phytoplankton and crustacean zooplankton, and low bacterial abundance and production. Monthly estimates of daily primary productivity are in the mesoeutrophic range (40-l ,420 mg C m-2 d-l); however, mean annual primary productivity is in the eutrophic range (500 mg C m-2 yr-I) because of greater activity in winter and an extended “growing season.” The stratified period (April-October) is characterized by rapid development of hypolimnetic anoxia, high bacterial production, cyanobacteria and large algae, heterotrophic flagellates, ciliates, rotifers, small cyclopoid copepods, and Chaoborus. Heterotrophic activity, HPLC chlo- rophyll a, and grazer activity were highest at the epilimnetic-metalimnetic boundary. Bacterial abundances (1.5-32 x 1O6 cells ml- I) increased in the during the stratified period. Long-term (1978-l 994) warming within the water column (0.06 1-O. 10°C yr-I) follows midwinter (January) air temperature trends. Using Lake Oglethorpe as an analog, we predict as consequences of climatic warming in northern dimictic lakes: increases in nutrient cycling rates, winter primary and secondary productivity, summer heterotrophic production, and microbial components of the food web,

. Acknowledgments In addition to the authors, Adrieen Debiase, Yvette Feig, Hal King, Doug Leeper, Bob McDonough, Judy Meyer, Michael Pace, John Orcutt, Amy Parker, Bob Sanders, and Susan Bennett Wilde provided major assistance in data collection, analysis and interpretation. John Chick and Phil Dixon aided in statistical analysis of the long-term thermal data, and Rob Reinert, David Bruce, and Mike Van Den Avyle provided insights about fish assemblages.Thanks to the Helfmeyers for accessto Lake Oglethorpe from their property. This research was supported indirectly by several grants from the National Science Foundation. Contribution 49 of the Lake Oglethorpe Limnological Association. 1041 1042 Porter et al.

Most small natural lakes in North America are glacial driven ecological patterns to predict the effects of climatic in origin and are classified as temperate and dimictic warming on dimictic lakes in north-temperate regions. (Hutchinson 1957). Studies of these north-temperate lakes are the basis for many limnological paradigms. Compa- rable small water bodies below the southern extent of Methods dimixis (36”N in the Midwest, 40”N in the Northwest, and 38”N in the Southeast, Hutchinson 1957) are gen- Study site-Lake Oglethorpe is in Oglethorpe County erally warm monomictic impoundments and have re- (33”52’12”N, 83’13’49”W) and has been a study site for ceived relatively little study. researchers at the nearby University of Georgia, Athens, Limnological comparisons have been made among since 1978. This small (AO = 30 ha) impoundment was tropical and temperate lakes but rarely include those at created in 197 1. It has a maximum depth of 8.7 m, a intermediate latitudes (e.g. Lewis 1987). Most reviews mean depth of 2.3 m, and a mean hydraulic residence considering lentic systems of southeastern North America time of 80 d (Raschke 1993). Compared with other small focus on natural lakes, which occur primarily on the impoundments in the southeastern Piedmont, Lake Ogle- Coastal Plain (Frey 1963; Crisman 1992), farm ponds thorpe is characterized as mesotrophic by the EPA be- (Menzel and Cooper 1992), and large reservoirs (Soballe cause of moderate epilimnetic nutrient (24 pg liter-* mean et al. 1992). A recent EPA study of small (~648 ha), total P) and chlorophyll a concentrations (10 pg liter-l stratifying impoundments in the Piedmont of North Car- mean Chl a) and a Carlson trophic state index of 53 during olina, South Carolina, and Georgia provides the first com- summer stratified period (Raschke 1994, pers. comm.). parative survey of these systems (Raschke 1993, 1994). The watershed consists of forests, farmlands, and private Lake Oglethorpe, our study site, was included in the EPA residences. The red-brown color associated with clay survey and is considered typical of small impoundments loading has been observed in the lake, but in recent years in the region with lakelike characteristics, being deep these events have been rare and have apparently declined enough to stratify and having a relatively long water res- as watershed development-a major cause of clay loading idence time. We define a small lake as any standing body in the Piedmont (Mulholland and Lenat 1992)-has lev- of water, natural or anthropogenic, that is lo-500 ha in eled off. The lake is used primarily for recreation, partic- surface area, deep enough to stratify, and with a long mean ularly fishing and swimming. The fish assemblage in both water residence time (> 2 months). 1977 and 1993 included bluegill (Lepomis machrochirus), Small monomictic impoundments are the appropriate green sunfish (Lepomis cyanellus), redbreast (Lepomis au- systems to compare and contrast with small dimictic lakes ritus), black crappie (Pomoxis nigromaculatus), golden of the north-temperate zone. Monomictic lakes may also shiner (Notemigonis crysoleucus), and largemouth bass be good analogs for the changes that could occur in dim- (Micropterus salmoides; Orcutt 1982; D. Bruce unpubl. ictic lakes due to climatic warming. In Canadian Shield data). lakes, regional warming trends have resulted in an in- crease in ice-free period and winter phytoplankton bio- Sampling and analyses -In 1979, monthly midday mass (Schindler et al. 1990). If warming trends continue, sampling was done at a permanent deep-water, foredam the thermal regimes of many small dimictic lakes may station at l-m depth intervals (down to 7 m). Temper- become more like those in warm monomictic systems ature and dissolved oxygen profiles were measured with (i.e. lacking a winter stratification period and with warmer an in situ thermistor and dissolved oxygen electrode (YSI water temperatures year-round). A better understanding Model 57). To look for possible long-term trends, we of limnological processes in warm monomictic lakes such examined temperature profiles collected at irregular in- as Lake Oglethorpe may improve predictions of the effects tervals over a 17-yr period (1978-l 994). Although a large of climatic warming in northern lakes. data set has been accumulated (N = 303 dates with tem- Our purpose is to provide a conceptual model of a perature profiles), profiles were not obtained at consistent monomictic lake in the southeastern Piedmont by de- time intervals. Therefore, we calculated detrended scribing ecological processes and associations over an an- monthly means (N = 108) for measurements of temper- nual cycle in Lake Oglethorpc, Georgia. In order to char- ature from each discrete depth (l-m intervals to 7-m acterize seasonal differences, we present a 13-month study depth) and the water-column average. Data were entered of physical and chemical patterns and of autotrophic and into general linear models procedure (SAS, version 6) heterotrophic processes during 1978-l 979. Our discus- using month and year as independent variables and month sion of physical and chemical patterns includes exami- as a class variable. Patterns in air temperature and rainfall nation of temperature and dissolved oxygen measure- data for the area (NCDC 1978-1994) were similarly as- ments spanning a 16.5-yr period. An annual cycle of or- sessed (SAS, version 6). Light profiles were measured at ganic sedimentation is provided for 1983-l 984. We sum- 0.25-m intervals with a LiCor LI-188B PAR irradiance marize studies of specific aspects of microbial production meter. Extinction coefficients were calculated according and consumption. These provide a general description of to Wetzel and Likens (1979). seasonal shifts in the relative importance of autotrophic Whole-water samples were collected at l-m depth in- and heterotrophic production to food-web structure that tervals with an g-liter PVC Van Dorn bottle. Subsamples correspond to the two major patterns of thermal regime: were processed and analyzed for concentrations of soluble mixis and stratification. Finally, we use these thermally reactive P (SRP) and total P (TP) (Menzel and Corwin Monomictic Lake Oglethorpe 1043

1965; Murphy and Riley 1962), ammonia N (Solorzano Temperature (“Cl 1969), and nitrate and nitrite (Strickland and Parsons 197 2). Dissolved inorganic and organic C (DIC and DOC) were measured by infrared gas analysis (Menzel and Vac- care 1964). HPLC chlorophyll a was determined (Jacob- sen 1982). Primary production was determined by the 14Cmethod (Schindler et al. 1972) using midmorning in situ incubations at l-m intervals from the surface to the 7-m depth. Lugol’s iodine-preserved phytoplankton were counted by the Utermohl method (Porter 1973). Samples for protozoan, rotifer, and crustacean zooplankton enu- meration were obtained on these same sample trips, and these methods are reported elsewhere (Orcutt and Pace 1984; Pace and Orcutt 1981; Pace 1982). Bacteria were enumerated by the DAPI direct count method (Porter and Feig 1980). Heterotrophic poten- tial - the maximal uptake and mineralization of [ 14C]glu- cose (Vmax)-was measured throughout 1’979 to assess patterns of bacterial production (Wright and Hobbie 1966; Vaccaro et al. 1977). Particulate adenosine triphosphate (ATP) concentrations were determined (Wiebe and Ban- Dissolved Oxygen (ppm) croft 197 5). In 1983-l 984, organic sedimentation was measured monthly with replicated (N = 8) sediment traps suspended at 2, 4, and 6 m in the water column. Traps consisted of 49-mm-diameter, 350-mm-long PVC tubes attached to the circumference of a 0.5-m-diameter frame. Saturated saline solution was used to exclude detritivores from screwtop powder jars threaded to the bottom of each tube. Total and ash-free dry weights were determined for contents of jars collected at monthly intervals.

Results Temperature and oxygen -Temperature and oxygen profiles obtained in 1979 revealed two major thermal periods (Fig. 1). The entire water column at 4-l 2°C mixed during late fall, winter, and early spring (November- March). The summer stratified period (April-October) was characterized by epilimnetic temperatures of 26-28°C. Fig. 1. Temperature and oxygen distributions with depth in A slow erosion of the began in late August 1979. and continued until mixing was complete in November. Oxygen levels below saturation occurred in the lower wa- was observed in hypolimnetic waters (0. 10°C increase per ters during late winter (February-March), and a deoxy- year at 7 m; Fig. 2). Mean annual hypolimnetic temper- genated hypolimnion developed in April within 2-3 weeks ature was correlated with mean January air temperature of the onset of stratification. Oxygen maxima occurred at (P = 0.0405, R2 = 0.27). No positive trend was seen in the epilimnetic-metalimnetic boundary, resulting in a annual mean air temperature for the Athens area (NCDC positive heterograde oxygen curve in late summer. 1978-l 994), and there was no statistical relationship be- To look for possible long-term trends, we examined tween mean annual water temperatures and summer air temperature profiles collected from 1978 to 1994. Over temperatures. The trend of increasing temperature did the 17-yr study period, a warming trend was detected at not occur at a consistent rate from year to year. An in- all depths and for the water-column average temperature crease in water temperature and greater interannual vari- (P2 = 0.061-o. 10°C yr-l). The average rate of warming ability seemed to occur in the mid-to-late 1980s. generally increased with depth and reflected a warming Figure 3 illustrates the relationship between the shal- trend in winter air temperatures. Warming trends were lowest anoxic depth, defined as DO 5 1 mg liter-l, and generally weaker for upper water-column depths (l-3 m: time of year for all sampling dates with anoxia in 1978- ,& = 0.062-0.066”C yr-l; P = 0.0447-0.0803) and stron- 1994 (N = 222). When years for which we have early ger for lower depths (4-7 m: p2 = 0.086-O.lO”C yr-l; P spring data (N = 8) are compared, hypolimnetic oxygen = 0.0001-0.0090) and the water-column average (& = deficits consistently developed within a few weeks of 0.08 1°C yr-l; P = 0.0019). The greatest rate of warming stratification. Within 2-4 weeks oxygen was depleted from Porter et al.

B =: -2 0 E. 0 0ocoo~o go0 0 Cu amryo o”oc&3 08 O 0” Ooo 3 0 0 cm00 mzxt& 2 -4 (D&pm 0 “00zo 0 &oa, 0 .- -3 I 0 d 0 00 &I

COD 0 0 *Oa OCD 0ga 0000 an 00 0 o” 8 0

00 9 0 0 OCO a3 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 0 am .I .-LI.--I.~l-- I -1 1. I- I- l- Year FMAMJJASON D Fig. 2. Averageannual temperatures(mean * SE)for water- Month (1978-94) column average temperature (O), mean 7-m temperature 0, and mean January air temperature (A) from 1978 to 1994. Fig. 3. Extent of the anoxic zone on sampledates from 1978 through 1994. the stratified portion of the water column, resulting in an extensive anoxic zone below 2-3 m. These oxygen profiles concentrations (J. Meyer unpubl. data) increased from illustrate that destratification begins in late August to ear- - 1 pg N liter-l in September to 300 in February. In ly September, and that the water column does not become contrast to P, N concentrations did not increase during fully mixed until November. high rainfall in April 1979. After stratification, combined nitrate and nitrite concentrations declined to detection Rainfall- Seasonally, rainfall is distributed evenly limits throughout the water column: first in the hypolim- throughout the year in the southeastern U.S. (Wallace et nion (May-June), then in the (June-July), and al. 1992). Average monthly rainfall data from 1979 to finally in the metalimnion (July). 1994 (NCDC 1978-l 994, data not shown) indicates that Stratification led to a reduction of DIC in the epilim- this general pattern holds true for the Athens area: month- nion to a low of 0.95 mg C liter-l (Fig. 4), presumably ly rainfall averages ranged from 8.4 to 11.9 cm. Monthly due to the activity of photoautotrophs. Hypolimnetic re- rainfall was highly variable among years (C.V. = 44- lease by heterotrophs increased DIC concentrations to a 85%). Each month’s rainfall deviated by more than 50% maximum of 32.8 mg C liter-l in October. The pH of its long-term mean in 5-10 of the 17 yr studied. Peak changed very little during the year (6.88kO.04) in spite months showed no predictable pattern among years. In of large changes in C02, indicating a high noncarbonate 1979, maximum rainfall occurred in April (20.6 cm; mean buffering capacity of the water. = 8.4 cm, median = 7.1 cm for all April 1978-1994; DOC (J. Meyer unpubl. data) remained relatively con- NCDC 1978- 1994), and minimum rainfall occurred in stant in surface waters (3.49k 1.3 mg C liter-l) despite December (3.5 cm; mean = 8.8 cm, median = 7.7 cm for major fluctuations in microbial biomass and productivity all December 1978-1993; NCDC 1978-1994). throughout the year (see below). Diurnal fluctuations in DOC were also minimal. In the hypolimnion, DOC con- Water chemistry-SRP (Fig. 4) and TP (J. Meyer un- centrations were 4-6 mg C liter-l in May and increased publ. data) concentrations fluctuated considerably (l-l 44 to 19 by the end of stratification. pg SRP-P liter-l and 16-1,080 pg TP liter-l) over the annual cycle. SRP concentrations remained low through- Light, chlorophyll, phytoplankton, and photosynthe- out the winter mixed period. Concentrations of SRP and sis - Light extinction coefficients remained relatively con- TP were highest in midwater in April. These coincided stant (1 .O-3.8 m-l) throughout 1979, except during an with heavy spring rains, high sediment load, and high influx of fine clay sediments following heavy rains in April turbidity. During stratification, SRP levels decreased (10.0 m-l; Fig. 5). Secchi disk depths, when determined throughout the water column and remained low in the across the past 16.5 yr, were generally between 0.5 and hypolimnion; TP increased with depth. 1.5 with a maximum clarity of 3.5 m. Transparency was Concentrations of ammonia N increased with depth lowest during periods of heavy turbidity and red-brown after stratification and reached a maximum (1,400 pg N color due to clay loading via surface water runoff. The liter-l) at the end of stratification in October 1979 (Fig. frequency of clay-loading decreased in recent years, and 4). After complete mixing, ammonium concentrations water color was generally olive-green. Clarity was greatest declined at all depths, while combined nitrate and nitrite in early winter after fall algal blooms. Monomictic Lake Oglethorpe 1045

Soluble reactive phosphorus @g liter -I)

E lo- .-a g 8- $ 0 6-

.-s 4- P

.V’ I N M A

1600 - y;TJ 1400 % ‘;FhlAM 0 1200

Ammonia (pg N liter -‘) E 1000 .-cn z 800 c z $ 600

1 I I I lIT-I NDJFMAMJJASONDJ Fig. 5. Annual variability in PAR light extinction coeffi- cients and areal primary productivity rates in 1979.

Standing stock concentrations of HPLC chlorophyll a (Fig. 6)-a measure of photosynthetically active pig- Dissolved Inorganic carbon (mg C liter -l ) ment-averaged 2 pg Chl a liter-l annually. Concentra- tions were higher in May through August, with a maxi- mum of 29 pg Chl a liter-l in the upper metalimnion during June. Both algal and bacterial chlorophyll were found below the euphotic zone during summer stratifi- cation (Jacobsen unpubl. data; see also Parker 1995). Phytoplankton showed distinct seasonal and depth dis- tribution patterns (Porter et al. unpubl data). Early in the winter mixed period, phytoplankton were distributed rel- atively evenly throughout the water column and the as- semblage was composed of small and colonial flagellates (chrysophytes and cryptomonads), diatoms, green algae, and large flagellates (dinoflagellates and euglenoids). In the early part of the stratified period, chrysophytes, cryp-

Fig. 4. Depth distributions of soluble reactive P, ammonia, and dissolved inorganic C concentrations in 1979. 1046 Porter et al.

Chlorophyll a (pg liter-l) tomonads, and green algae dominated the algal assem- blage, but filamentous cyanophytes and large (>20-pm GALD) algal flagellates (dinoflagellates and euglenoids) were the major components of the phytoplankton for most of summer. Large motile flagellates were found in the euphotic zone during stratification and cyanobacteria generally settled to the epilimnetic-metalimnetic bound- ary. These general patterns were also found in more recent studies of phytoplankton succession in Lake Oglethorpe (Parker 1995). Integrated photosynthesis expressed on an areal basis (Fig. 5) was highest from May to September (470 mg C m-2 d-l in June and 1,420 in August). The first peak in primary productivity occurred in May (900 mg C m-2 d- l), as stratification stabilized and while phytoplankton V,, (nmoles liter -1 h -l) biomass was still moderate. A second peak was observed in August (1,420 mg C m-2 d- ‘) as the epilimnion deep- ened and algal cells and nutrients that had accumulated in the metalimnion were dispersed into the cuphotic zone. Winter primary productivity remained moderate during October-February (loo-470 mg C m-2 d-l). Productivity rates were lowest in March and April (40-70 mg C m-2 d-l). Primary productivity rates did not correlate with HPLC chlorophyll a or cell concentrations (areal chlo- rophyll data not shown, but see Fig. 6). On most dates, maximum photosynthetic rates were at the surface, while maximum algal cell and active Chl a concentrations were at 2-3 m (Fig. 6). Annual primary production for 1979 was -500 g C m-2 yr-l.

Microbial abundance and production -Indicators of microbial biomass and activity showed distinct season- ality and patterns of vertical distribution during the strat- ified period. Concentrations of bacteria in the water col- umn ranged between 1.5 x lo6 and 3.2’~ lo7 cells ml-l, with an average of 94% small unattached coccoid cells (Porter and Feig 1980). Densities were lowest in winter and highest during summer stratification in the hypolim- nion (Fig. 6). Throughout the year, bacterial densities were lowest at the surface. During late fall and winter, bacteria were uniformly distributed with depth. During the stratified period, concentrations of bacteria were high- est immediately below (June-July) or just above (August- September) the interface between the epilimnion and the metalimnion. Particulate ATP (pg liter -l) Heterotrophic potential, the maximal uptake and min- eralization of [ 14C]glucose ( V,,,), was measured through- out 1979 as a relative indicator of bacterial production rates (Wright and Hobbie 1966). Values of V,,, were low and uniform throughout the water column during late fall and winter 1979. In April V,,, increased dramatically at all depths, peaking at 1 m. This coincided with peak SRP values and low light availability in that year (Fig. 4). During the stratified period, values were high in the lower

Fig. 6. Seasonal and depth variation in HPLC chlorophyll a, heterotrophic potential (V,,,), bacterial cell numbers, and

D J F M A M J ii S 0 N D particulate ATP in 1979. stable bounda nion was at 3 F hrlAMJJA SONDJFMA r-.r- 1 were as e

lationship between rainfall and sedimentation in the ten- tral deep-water basin.

ber-March) and stratified (Apri consistently from ye presence of an extensive anoxic hypolimnetie zone are autotrophic and het- 979, clay loading diminished predictable in the following

The average annual

has any measurable impact on biological variables in the 1048 Porter et al. winter mixed period. In particular, measurements of both 198 1; Pace 1982). In addition, hypolimnetic SRP con- nutrient concentrations and primary production rates centrations declined throughout summer stratification suggest high activity in winter compared to colder lakes. while TP increased, which is different from observed SRP SRP is low throughout winter, indicating that available accumulations in north-temperate lakes with anoxic hy- P is rapidly incorporated into the biotic pool. Increases polimnion (Wetzel 1983; Caraco et al. 1991; Krambeck in nitrate and nitrite concentrations throughout the water et al. 1994). Preliminary in situ EPA dome studies in column in winter, and corresponding decreases in am- Lake Oglethorpe (R. Raschke unpubl. data) indicate P monium concentrations, indicate that nitrifying bacteria release by anoxic sediments during summer stratification. are active during the mixed period. From October to It is possible that P does not accumulate because it is February, areal primary productivity rates (loo-470 mg retained in the active heterotrophic biota of the hypolim- C rnB2 d-l, see Fig. 5) are about equal to those in hy- nion. As an indicator of high biological activity, this trend pereutrophic Wintergreen Lake (Wetzel 1983). However, suggeststhat the observed hypolimnetic increase in DOC during this same time period Chl a concentrations are represents an accumulation of refractory compounds. similar to those in oligotrophic Lawrence Lake (Fig. 6; Wetzel 1983). Production per unit algal biomass is there- Seasonal shi$ in the dominance of macro- vs. micro- fore high, and this is presumably a result of relatively high zooplankton -The pelagic community is dominated by water temperatures, light availability, physical mixing, crustacean macrozooplankton in winter and protozoan and also rapid turnover due to grazing (see below). Algal and metazoan microzooplankton in summer (Pace and productivity throughout winter may retain nutrients in Orcutt 198 1). These seasonal differences in zooplankton the biotic pool and reduce the potential magnitude of community structure reflect differences in the dominant spring algal blooms. Compared with those in northern size and quality of food available during mixis and strat- lakes (Wetzel 1983, table 15.9), daily rates of primary ification (Fig. 8). The algal groups characteristic of winter productivity in Lake Oglethorpe (100-l ,420 mg C m-2 months were primarily small, high food-quality species. d-l) were in the mesotrophic range, while annual pro- The discrepancy between moderate productivity rates and ductivity (500 g C m-2 yr-l) was in the eutrophic range. low algal biomass accumulation can be explained by the This was due to the extended growing season. grazing of crustacean zooplankton, especially cladoceran species, which accumulated biomass in winter (4-60 an- Summer heterotrophic production -Heterotrophic ac- imals liter-l in 1979; Orcutt and Pace 1984). At this time, tivity is important throughout the summer stratified pe- bacterial numbers are low, and bacteria and bacterial con- riod. Deoxygenation of the hypolimnion is rapid (2-4 sumers are minor supplements to algal resources for crus- weeks) after the onset of stratification. Seasonal patterns taceans (Pace et al. 1983). Cladocerans were inefficient as of heterotrophic production measured by glucose uptake direct consumers of bacteria, contributing < 1% to com- in 1979 (V,,,, Fig. 6) were confirmed in 1984-1985 with munity grazing (Sanders et al. 1989). Mixotrophic fla- tritiated thymidine and leucine tracers (McDonough et gellates are the major bacterial grazers during winter mixis al. 1986). Highest rates were correlated with high Chl a (Bennett et al. 1990). concentrations at the epilimneticmetalimnetic bound- After the major biomass peaks in fall and winter, cla- ary, which was not a region of high bacterial or protozoan doceran population densities exhibit a smaller density abundance (Fig. 6; Pace 1982) but was a region of high pulse in late spring (April) and are at low densities by protozoan and metazoan grazing activity (Sanders et al. early summer (Pace and Orcutt 198 1; P. Saunders unpubl. 1989; Bennett et al. 1990). Hence, bacterial production data). This is contrary to the pattern observed in northern and standing stock were not correlated. dimictic lakes, where the spring-summer bloom usually Bacterial production may be directly supported by high has the greatest abundances (Horne and Goldman 1994). concentrations of filamentous algae at the interface be- In June 1979, there was an abrupt decline in crustaceans, tween the aerobic epilimnion and the anaerobic hypolim- particularly Daphnia, and crustacean populations re- nion. DOC concentrations remained moderate and rel- mained low until October. The early summer decline in atively constant throughout the year in the mixed portion crustaceans cannot be attributed to high temperatures of the water column (3.5 + 1.3 mg C liter- ‘) but increased (Orcutt and Porter 1983, 1984). Shifts in food quality, in the hypolimnion after stratification. Organic sedimen- with a reduction of small edible algae and an increase in tation was highest at the epilimnetic-metalimnetic bacteria not suitable for growth and reproduction of boundary in midsummer. Decreasing organic sedimen- Daphnia are in part responsible for the declines (Pace et tation with depth, and also declining ATP concentrations al. 1983). Increases in blue-green algal filaments that are in the metalimnion and hypolimnion in summer, suggest energetically costly for cladocerans to reject (Porter and decomposition of organic matter within the water col- McDonough 1984) also occurred in late spring. A model umn, rather than in the sediments. In the anoxic zone, by Pace et al. (1984) predicted that Chaoborus can also bacteria may also receive carbon via grazing activity. have a significant impact on Daphnia numbers and life Heterotrophic microflagellates are abundant in the hy- histories in the early (April-May) part of the stratified polimnion during stratification and peaked at summer period. Whatever the cause of low crustacean numbers, densities of 102-lo4 cells ml-l in the metalimnion (Ben- one consequence is that the “clear-water phase” that oc- nett et al. 1990). Distinct assemblages of protozoa are curs in early summer in many northern dimictic lakes also found in the anoxic hypolimnion (Pace and Orcutt due to grazing by crustacean zooplankton populations Monomictic Lake Oglethorpe 1049

(Luecke 1990) has not been observed on Lake Oglethorpe. This cannot be explained by the presence of clays, since a clear-water phase has not been observed in recent years that lack significant clay loading (Parker 1995). Dominance of the zooplankton biomass by rotifers (2- 80 x 1O2 animals liter- l, Orcutt and Pace 1984) and pro- tozoans (1 04-1 O5cells liter- l; Pace and Orcutt 198 1; Pace 1982) during summer stratification was correlated with increases in bacteria as a resource base (Pace and Orcutt

198 1). Flagellated protists, the major bacterivores in the 7 HETEROTROffllC f I I lake (Sanders et al. 1989), were not significantly correlated FUGELLATES MIXOTROPHIC with bacterial densities (Bennett et al. 1990). The presence FIAGELIATES of high densities of rotifers, which were found to con- tribute 68-92% of community grazing on heterotrophic FALL- WINTER-SPRING flagellates (Sanders et al. 1994), may have led to the low correlation between bacteria and flagellates. Rotifers also fed on some algae, ciliates, and bacteria. Rotifers are therefore likely major links between bacterial production and higher trophic levels during summer stratification. High abundances of rotifers and protozoans seem to also occur in tropical and subtropical lakes (Serruya and Pol- lingher 1983; Bays and Crisman 1983) and may be a consequence of enhanced microbial production. Rotifers are fed upon by Chaoborus and larval fish (Porter 1996; Bruce unpubl. data; Saunders unpubl. data). It may be that the early planktivorous life stages of larval bluegill, which are produced throughout summer in Lake Oglethorpe, are supported significantly by microbially based production. Thus, microbial production may ul- timately be supporting a bass-bluegill fish assemblage SUMMER similar to that of many northern lakes. Rotifers can in- teract with Daphnia populations directly through com- Fig. 8. Food-web relationships during the fall-winter-spring mixed period and the summer stratified period in mesoeutroph- petition (Gilbert 1988) and indirectly by providing food ic, monomictic Lake Oglethorpe (after Porter 1996). During for high numbers of predatory Chaoborus (Neil1 1985; mixis, high-quality algal food supports abundant populations Porter 1996). In summer, therefore, microbially based ofcrustaceans, particularly the cladoceran Daphnia spp. In sum- production may have a bottom-up effect as a resource mer, microbially based production (protozoans, rotifers, small base, a lateral effect (Porter 1996) via interconsumer com- cyclopoid copepods, and Chaoborus punctipennis larvae) sup- petition, and a top-down effect by supporting predators ports planktivorous fish and their larvae. such as Chaoborus and larval fish. Extensive anoxia and the habitats of higher trophic con- phytoplankton. These and warmer winter temperatures sumers-The rapid development of anoxia below 2-3 m may enhance nutrient cycling and primary productivity following thermal stratification may be significant to the rates. Crustacean zooplankton populations would grow survival of pelagic fish, the invertebrate predator Chao- during the ice-free period. Nutrients would accumulate borus, and crustacean zooplankton. Extensive hypolim- in the biota, rather than in the lower waters as in ice- netic anoxia is probably an important habitat limitation covered systems, because of physical mixing, increased for both crustacean zooplankton and centrachid fish of winter productivity, and biomass. Therefore, nutrients Lake Oglethorpe. Early anoxia in the lake may also pro- may be less available to support a spring algal bloom. vide Chaoborus with an important refuge from fish pre- Bacterial production would increase and its consumers dation. This refuge and ample microbially based re- (flagellates, ciliates, and rotifers) would dominate zoo- sources may allow Chaoborus to reach the exceptionally plankton biomass in summer. high abundances, up to 10 liter-r and 20,000 m-2, found Heterotrophic production would be particularly im- in the lake throughout summer (Porter 1996). portant to zooplankton in summer if autotrophic pro- duction were primarily by large, inedible forms, such as Predictions for dimictic lakes-If climatic warming con- blue-green filaments, dinoflagellates, and euglenoids. An- tinues, the seasonal patterns observed in Lake Oglethorpe oxia in the hypo- and metalimnion would develop earlier could be used as predictors of possible ecological changes in the stratified period, which may restrict the habitats in north-temperate lakes. The loss or reduction of winter of both crustacean zooplankton and fish, while providing ice cover would increase light availability, and longer a refuge for the anoxia-tolerant predator Chaoborus. If periods of winter mixis would reduce sinking losses for stratification, anoxia, and the shift to a microzooplank- 1050 Porter et al. ton-dominated zooplankton community occurs before HORNE,A. J., AND C. R. GOLDMAN. 1994. Limnology, 3rd ed. water temperatures allow significant fish activity, crus- McGraw-Hill. tacean zooplankton may decline before fish populations HUTCHINSON,G. E. 1957. A treatise on limnology. V. 1. Wiley. can utilize them for growth of early life stages. On the JACOBSEN,T. R. 1982. Comparison of chlorophyll a mea- other hand, microzooplankton and Chaoborus may pro- surements by fluorometric, spectrophotometric and high pressure liquid Chromatographic methods in aquatic en- vide an alternative resource base for larval and adult fish, vironments. Ergeb. Limnol. 16: 35-45. such as the bass-bluegill assemblage in Lake Oglethorpe KRAMBECK,H., D. ALBREWT, B. HICKEL, W. HOFMANN, AND and many northern lakes. H. ARZBACH. 1994. Limnology of the Plul3see, p. l-23. Potential responses of northern, dimictic lakes to cli- In J. Overbeck and R. J. Chrost [eds.], Microbial ecology matic warming would not require new associations or of Lake Plul3see. Springer. processes by this analog model. In general, conditions LEWIS,W. M., JR. 1987. Tropical limnology. Annu. Rev. Ecol. that prevail in northern lakes at the end of summer (Au- Syst. 18: 159-184. gust) would be expected to extend throughout more of LIKENS,G. E.,J.S. EATON,AND N.M. JOHNSON. 1985. Mirror summer in the event of continued climatic warming. Lake- physical and chemical characteristics: B. Physical and chemical environments, p. 89-107. In G. E. Likens Overall, we predict that climatic warming would tend to [ed.], An ecosystem approach to aquatic ecology. Springer. shift patterns in northern dimictic lakes toward those LUECKE,~., M.J. VAN-NI,J.J. MAGIWSON,J.F. KITCHELL,AND observed in southern monomictic lakes. This pattern P. T. JACOBSON. 1990. Seasonal regulation of Daphnia should be enhanced with increasing eutrophy. In partic- populations by planktivorous fish: Implications for the spring ular, autotrophic production and crustacean macrozoo- clear-water phase. Limnol. Oceanogr. 35: 17 18-l 733. plankton will become more important in winter, and the MCDONOUGH,R.J.,R. W. SANDERS,K.G. PORTER,AND D.L. period dominated by blue-green algae, heterotrophic pro- KIRCHMAN. 1986. Depth distribution of bacterial pro- duction, protozoans, rotifers, and other metazoan micro- duction in a stratified lake with an anoxic hypolimnion. zooplankton would become longer and more important Appl. Environ. Microbial. 52: 992-l 000. to the pelagic food web in summer. MENZEL, R. G., AND C. M. COOPER. 1992. Small impound- ments and ponds, p. 389-420. In C. T. Hackney et al. [eds.], Biodiversity of the southeastern United States: Aquatic communities. Wiley. References -, AND N. A. CORWIN. 1965. The measurement of total phosphorus in seawater based on the liberation of organi- BAYS, J. S., AND T. L. CRISMAN. 1983. Zooplankton and tro- cally bound fractions by persulfate oxidation. Limnol. phic state relationships in Florida lakes. Can. J. Fish. Aquat. Oceanogr. 10: 280-282. Sci. 40: 1813-1819. -, AND R. F. VACCARO. 1964. The measurement of dis- BENNETT, S. J., R. W. SANDERS,AND K. G. PORTER. 1990. solved organic and particulate carbon in seawater. Limnol. Heterotrophic, autotrophic, and mixotrophic nanoflagel- Oceanogr. 9: 138-142. lates: Seasonal abundances and bacterivory in a eutrophic MULHOLLAND,P. J., AND D. R. LENAT. 1992. Streams of the lake. Limnol. Oceanogr. 35: 182 l-l 832. southeastern Piedmont, Atlantic drainage, p. 193-232. In BUFUWOLDER,J. M. 1992. Phytoplankton and episodic sus- C. T. Hackney et al. [eds.], Biodiversity of the southeastern pended sediment loading: Phosphate partitioning and United States: Aquatic communities. Wiley. mechanisms for survival. Limnol. Oceanogr. 37: 974-988. MURPHY, J., AND J. P. RILEY. 1962. A modified single solution CARACO,N., J. J. COLE, AND G. E. LIKENS. 199 1. A cross- method for the determination of phosphate in natural wa- system study of phosphorus release from lake sediments, ters. Anal. Chem. Acta 27: 31-36. p. 241-258. In J. J. Cole et al. [eds.], Comparative analysis NCDC (NATIONALCLIMATEDATACENTER). 1978-1994. Local of ecosystems: Patterns, mechanisms, and theories. Spring- climatological data, Athens, GA. NOAA, NES-DIS. er. NEILL, W. E. 1985. The effects of herbivore competition upon CARPENTER,S. R., ANDOTHERS. 1993. Primary production and the dynamics of Chaoborus predation. Arch. Hydrobiol. its interactions with nutrients and light transmission, p. 21: 483-491. 189-209. In S. R. Carpenter and J. F. Kitchell [eds.], The ORCUTT, J. D., JR. 1982. Population dynamics, physiology trophic cascade in lakes. Cambridge. and the adaptative significance of diel vertical migration CRISMAN,T. 1992. Natural lakes of the southeastern United by zooplankton in Lake Oglethorpe, Georgia. Ph.D. thesis, States: Origin, structure and function, p. 475-538. In C. T. Univ. Georgia. Hackney et al. [eds.], Biodiversity of the southeastern Unit- AND M. L. PACE. 1984. Seasonal dynamics of rotifer ed States: Aquatic communities. Wiley. and crustacean zooplankton populations in a eutrophic CUKER, B. E., AND L. HUDSON,JR. 1992. Type of suspended monomictic lake with a note on rotifer sampling techniques. clay influences zooplankton response to phosphorus load- Hydrobiologia 119: 73-80. ing. Limnol. Oceanogr. 37: 566-576. AND K. G. PORTER. 1983. Diel vertical migration by FREY, D. G. 1963. Limnology in North America. Univ. Wis- zooplankton: Constant and fluctuating temperature effects consin. on life history paramenters of Daphnia. Limnol. Oceanogr. GILBERT,J. J. 1988. Supression ofrotifer populations by Daph- 28:720-730. nia: A review of the evidence, the mechanisms, and the -, AND -. 1984. The synergistic effects of temper- effects on zooplankton community structure. Limnol. ature and food concentration on life history parameters of Oceanogr. 33: 1286-l 303. Daphnia. Occologia 63: 300-306. HOLM-HANSEN,O.,ANDC. R. BOOTH. 1966. Themeasurement PACE, M. L. 1982. Planktonic ciliates: Their distribution, of adenosine triphosphate in the ocean and its ecological abundance, and relationship to microbial resources in a significance. Limnol. Oceanogr. 13: 507-5 14. monomictic lake. Can. J. Fish. Aquat. Sci. 39: 1106-l 116. Monomictic Lake Oglethorpe 1051

-, ANDJ. D. ORCUTT,JR. 1981. The relative importance SCHINDLER,D. W., AND OTHERS. 1990. Effects of climatic of protozoans, rotifers, and crustaceans in a freshwater zoo- warming on lakes of the central boreal forest. Science 250: plankton community. Limnol. Oceanogr. 26: 822-830. 967-970. -, K. G. PORTER,AND Y. S. FEIG. 1983. Species- and -, R. V. SCHMIDT,AND R. A. REID. 1972. Acidification age-specific differences in bacterial resource utilization by and bubbling as an alternative to filtration in determining two co-occurring cladocerans. Ecology 64: 1145-l 156. phytoplankton production by the 14Cmethod. J. Fish. Res. - -,AND-. 1984. Life history variation with- Bd. Can. 29: 1627-1631. in ‘a parthenogenetic population of Daphnia parvula (Crus- SERRUYA,C., AND U. POLLINGHER. 1983. Lakes ofthe warm tacea: Cladocera). Oecologia 63: 43-5 1. belt. Cambridge. PARKER,A. K. 1995. Phytoplankton seasonal succession in a SOBALLE,D. M., B. L. KIMMEL, R. H. KENNEDY, AND R. F. warm-monomictic lake (Lake Oglethorpe, GA). M.S. thesis, GAUGUSH. 1992. Reservoirs, p. 421-476. In C. T. Hack- Univ. Georgia. ney et al. [eds.], Biodiversity of the southeastern United PORTER,K. G. 1973. Selective grazing and differential diges- States: Aquatic communities. Wiley. tion of algae by zooplankton. Nature 244: 179-l 80. SOL~RZANO,L. 1969. Determination of ammonia in natural 1996. Integrating the microbial loop and the classic waters by the phenolhypochlorite method. Limnol. Ocean- food chain into a realistic planktonic food web, p. 51-59. ogr. 14: 799-801. In G. Polis and K. Winemiller [eds.], Food webs: Integra- STRICKLAND,J. D. H., AND T. R. PARSONS.1972. A practical tion of patterns and dynamics. Chapman and Hall. handbook of seawater analysis, 2nd ed. Bull. Fish. Res. Bd. -, AND Y. S. FEIG. 1980. The use of DAPI for identifying Can. 167. and counting aquatic microflora. Limnol. Oceanogr. 25: VACCARO,R.F.,F. AUM,AND R.E. HODSON. 1977. Response 943-948. of natural marine bacterial populations to copper: Con- -, AND R. J. MCDONOUGH. 1984. The energetic cost of trolled ecosystem pollution experiment. Bull. Mar. Sci. 27: response to blue-green algal filaments by cladocerans. Lim- 17-22. nol. Oceanogr. 29: 365-369. RASCHKJZ,R. L. 1993. Guidelines for assessing and predicting WALLACE,J.B.,J.R. WEBSTER,ANDR.E. LOWE. 1992. High- eutrophication status of small southeastern Piedmont im- gradient streams of the Appalachians, p. 133-l 92. In C. T. poundments. U.S. EPA., Athens. Hackney et al. [eds.], Biodiversity of the southeastern Unit- . 1994. Phytoplankton bloom frequencies in a popu- ed States: Aquatic communities. Wiley. lation of small southeastern impoundments. Lake Reser- WETZEL,R. G. 1983. Limnology, 2nd ed. Saunders. voir Manage. 8: 205-2 10. -, AND G. E. LIKENS. 1979. Limnological analyses. Saun- SANDERS,R. W.,D. A. LEEPER,~. H. KING,AND K.G. PORTER. ders. 1994. Grazing by rotifers and crustacean zooplankton on WIEBE,W. J., AND K. BANCROFT. 1975. Use of adenylate en- nanoplanktonic protists. Hydrobiologia 288: 167-l 8 1. ergy charge ratios to measure growth rate of natural mi- -, K.G. PORTER,~. J. BENNETT,ANDA. DEBIASE. 1989. crobial communities. Proc. Natl. Acad. Sci. 72: 2 112-2 115. Seasonal patterns of bacterivory by flagellates, ciliates, ro- WRIGHT, R. T., AND J. E. HOBBIE. 1966. Use of glucose and tifers, and crustaceans in a freshwater planktonic com- acetate by bacteria and algae in aquatic ecosystems. Ecology munity. Limnol. Oceanogr. 34: 673-687. 47: 447-464.