PRIMARY PRODUCTIVHY m WATER RECIAMATION LAKES AT MICHIGAN STATE UNIVERSITY Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY DONALD WAYNE SCHLOESSER 1976 ABSTRACT PRIMARY PRODUCTIVITY IN WATER RECLAMATION LAKES AT MICHIGAN STATE UNIVERSITY By Donald Wayne Schloesser Primary productivity in the first lake in a series of four in a water reclamation facility was dominated by phytoplankton. Macrophytes plus their epiphytes, dominated primary productivity in the fourth lake. Using the diurnal oxygen curve method for obtaining estimates of net productivity and respiration, the plankton of the first lake on the average accounted for l00% of the ecosystem primary productivity and 75% of the ecosystem respiration. The remaining respiration occurred in the benthic community. In the fourth lake, an Elodea canadensis-epiphyte complex was responsible on the average for 75% of the ecosystem primary productivity and a like percentage of the ecosystem respiration. The remaining primary productivity ,occurred in the plankton. Primary productivity and respiration in the benthic community of this lake were too low to measure. Mean P/R ratios for the first and fourth lakes over the grow- ing season were 0.99 and l.07 respectively. The latter was obtained after correcting gross primary production and respiration in the macrophyte component for maximum internal lacunar oxygen changes that might have occurred. Repeated studies of the lakes are Donald Wayne Schloesser necessary to show whether increased seasonal autotrophy through the system is a consistent trend. Harvest of aquatic macrOphytes from water reclamation lakes may be a desirable management strategy. A P/R for the macrophytes can be estimated by the diurnal oxygen curve method used here to indicate the time of maximum biomass accrual. Experiments with different harvesting times and methods are needed to indicate whether multiple cuttings at maximum P/R or single cutting at max- imum seasonal standing crop would give the most desirable yield of macrophytes. PRIMARY PRODUCTIVITY IN WATER RECLAMATION LAKES AT MICHIGAN STATE UNIVERSITY' By Donald Wayne Schloesser A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1976 ACKNOWLEDGMENTS I would like to acknowledge and thank my major professor Dr. C. D. McNabb for his guiding support. To me, his greatest assets were patience and an intuitive understanding of a growing mind. The help of Drs. Eugene Roelofs and Thomas Dahr with my graduate program and in reviewing this thesis was appreciated. This work was supported in part by funds provided by the United States Environmental Protection Agency - Federal Water Quality Administration under Training Grant #T90033l awarded to the Department of Fisheries and Wildlife. 11 TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION DESCRIPTION OF STUDY SITE . METHODS . RESULTS . Abundance of Primary Producers Conditions of Light. . Primary Productivity and Respiration DISCUSSION . LITERATURE CITED . APPENDIX Page iv 16 16 19 38 46 51 Table A-4. LIST OF TABLES Selected characteristics of influent and effluent water in Lakes l and 4 expressed as means taken on survey dates during the interval 27 April to l7 October, l975 . . . . . . . Mean photoplankton chlorophyll-a concentrations in mg/m in 1.1 m of water during l975 . Weight determinations for the biomass of Elodea canadensis taken from enclosures within Lake 4 Daily surface radiation (kcal/mZ/day) on dates of study in Lakes 1 and 4 . . . . . . . Season l means of productivity and respiration in g 02/m /day for community types and ecosystems in Lakes l and 4 during l975 . . Phytoplankton organisms found in Lakes l and 4 Gross primary production in g C/mZ/day for ecosystem components in Lakes l and 4 . . . . . . . Percent transmittance of surface radiation at the 0.5 m and 1.0 m depths in Lakes l and 4 on dates of study . . . . . . . . . . . . . . . . Results of a modified two-way analysis of variance performed on gross primary production estimates within and between lakes iv Page 17 18 20 37 52 53 55 56 Figure LIST OF FIGURES Lakes of the water reclamation facility located on the Michigan State University campus . . . . . . . Plexiglass enclosure (A) used fori situ productivity and respiration measurements (B) . . . . The relationship between the extinction coefficients of light and chlorophyll-a concentrations found in Lake 1 Light transmittance occurring in the Open water and in a canopy of Elodea canadensis in Lake 4 Plankton (A) and whole ecosystem (B) productivity and respiration in Lake 1 in l975 . . Plankton (A), Elodea canadensis plus its epiphytes (B), and whole ecosystem (C) productivity and respiration in Lake 4 in l975 . . . . . . . . . . . Gross primary productivity (0) and respiration to) for the Elodea canadensis-epiphyte components in Lake 4 corrected for internal gas storage . . . Page 10 21 23 28 31 43 INTRODUCTION Over the last two decades ecological theory, emanating from consideration of energy flow (Odum, l956, 1957), has been concerned with metabolic patterns rather than structural composition. As one progresses up the organizational hierarchy of populations, eco- system subunits, and ecosystems, certain metabolic patterns become increasingly evident. This is because a large system is more apt to reflect homeostatic mechanisms than a smaller unit (Odum, 1953). The results of measurements become more consistent due to checks and balances that dampen biological oscillations along the hier- archical line. By monitoring the highest hierarchical level, the ecosystem, one should be able to classify systems by function, while considerable variability can exist in structural composition. Such an approach was described by Odum (l956) and expanded by Margalef (l968). A P/R ratio, where P = gross primary production and R = community respiration was used in their scheme. System types have been defined by their deviation of P/R from unity. An autotrophic system would exhibit a P/R > l; a heterotrophic system would exhibit a P/R < l. A third type, the mature undisturbed system, has been considered to be in steady-state having a P/R = l. Odum (1956) has shown P/R ratios to be an effective eco- system indicator in studies of the White River, exhibiting hetero- trophy in an area highly polluted with organic material, autotrophy in a nutrient rich recovery area, and steady-state in an unenriched area. From this comes a theory that the more organi- cally polluted a system, the lower the expected P/R ratio. This scheme appears ideal for classifying aquatic environments that have been exposed to man's activities. One purpose of this study was to determine whether the jn_§jtu_oxygen measurement methods of Odum could be used to detect differences in the P/R ratios of lakes in the Michigan State University water reclamation facility. A second purpose of this study was to partition ecosystems of the wastewater lakes into compartments by type of primary pro- ducer and to determine their influences on the total ecosystem P/R ratio. In this way ecological dominance by primary producer types could be determined. Where submersed vascular plants are a dominant type and their harvest is considered a desirable manage- ment strategy, Odum (l97l) suggests that their P/R ratio measured over the growing season can be used to set the best time for harvest. A third purpose of this study was to determine P/R for Elodea canadensis growing in a lake of the Michigan State Uni- versity system over the season so that recommendations regarding the time of harvest might be made. DESCRIPTION OF STUDY SITE This study was carried out on Lakes l and 4 of the Michigan State University Water Quality Management facility (Figure l). Built in 1973, this lake system receives effluent water from an East Lansing sewage treatment plant. The plant performs primary screening, activated sludge and partial tertiary treatment of domestic sewage. During this study approximately 45.6 million liters per day (ML/d) were processed at the plant with 43.5 ML/d discharged into the Red Cedar River. The remaining l.9 ML/d were pumped 7.25 kilometers and discharged into Lake 1. Of the 1.9 ML/d entering Lake l, 0.4 ML/d were taken for spray irrigation at a nearby site. Under gravitational flow, the remaining 1.5 ML/d were channeled through the lakes successively, with the effluent from (Lake 4 going into Herron Creek. The four lakes of the MSU system make up a total surface area of 16 hectares; they have a working depth of two meters. There are an additional 58 hectares equipped for spray irrigation. Three 0.4 hectare marshes are interconnected between Lakes 2 and 3. Lakes in the system are sealed with clay to prevent water loss. Relatively little detritus was present on the bottoms of the basins during this study. The water in Lakes 1 and 4 differed noticeably in appear- ance over the study interval. It was very turbid in Lake 1. Algae Figure l.--Lakes of the water reclamation facility located on the Michigan State University campus. Arrows indi- cate direction of flow. 0:06 occé .---V t v 9:... b m 9.0.. . 3... v.9 395 59.22 829: a... m.» o . 09 O :Nm _ 9.01— N 9.0... Z 205 4 23589.... 283 E9; scum and an associated odor were noted for brief periods in hot weather. The water of Lake 4 was clear and had none of the unpleas- ant aesthetic properties found in Lake 1. Water analysis showed that Lake 4 had lower concentrations of important nutrients than Lake 1 (Table 1). Measurements made in this study showed oxygen was uniform in open water down to the water-sediment interface. Occasional low values were obtained at the water-sediment interface, but total depletion was not observed. Such lowering of oxygen concentrations at the interface is typically due to bacterial activity (Alsterberg, 1922; Edwards and Rolley, 1965). Percent oxygen saturation in the water columns of Lakes 1 and 4 ranged from 25% to 210% and 71% to 143% respectively. Over this study both lakes were aerobic. Water temperatures found in these lakes were typical of shallow systems in the temperate zone (Macan, 1958; Butler, 1963). Temperatures rose to approximately 10°C early in May, gradually reaching highs of 25 to 30°C by mid-summer. By mid—October water temperatures were again about 10°C. Maximum changes of 3°C in the entire mass of a lake over a ten-hour period were recorded. Solar and atmospheric heat influx and wind were related to temperature shifts that occurred. Average Secchi disk visibility in Lakes 1 and 4 were 1.3 and 2.0 meters respectively. Visibility in Lake 1 was divided into two distinct periods. During the 27 April to 15 July interval, mean Secchi transparency was 0.8 meters; in mid-summer visibility increased. In the interval 15 July to 17 October, the mean Secchi TABLE 1.--Selected characteristics of influent and effluent water in Lakes 1 and 4 expressed as means taken on survey dates during the interval 27 April to 17 October, 1975. Parameter Lake 1 Lake 4 Hardness mg/l-CaCO3a Influent 320 186 Effluent 313 145 Alkalinity mg/i-Caco3a Influent 149 82 Effluent 158 61 Nitrate mg/l-Na Influent 8.30 0.33 Effluent 6.50 0.17 Total Phosphorus mg/l-Pa Influent 1.04 0.80 Effluent 0.68 0.07 Kjeldahl Nitrogen mg/l-Na Influent 1.17 1.14 Effluent 3.39 0.64 Surface pH 8.6 9.9 Mean Surface Dissolved Oxygen mg/l 9.1 9.8 aHourly composite sample on survey dates analyzed by Technicon Solid Prep 111 Autoanalyzer. disk determination was 1.5 meters. Lake 4 visibility extended to the bottom of the basin throughout the season. Macrophytic vegetation was composed of species that had been seeded into Lakes 2, 3 and 4 during the fall of 1973 (McNabb gt_a1., l975). Potamogeton foliosus Raf., Elodea canadensis Michx. and Cladophora fracta (Dillw.) Keutzing were the abundant macrophytes during this study. In unseeded Lake 1, macrophyte growth was neg- ligible until mid-August; patches of E, foliosus (44 g/m2 ash-free dry weight) and Q, frggta (13 g/m2 ash free dry weight) were present thereafter. Lake 4 produced the largest macrophyte crop in 1975. _§. canadensis formed dense monotypic stands of vegetation (157 g/m2 ash-free dry weight). 3, foliosus formed thin patches of growth (10 g/m2 ash-free dry weight) around the outer margins of the E, canadensis stands. 9, fracta occurred in sparse amounts in Lake 4. Phytoplankton quantity and quality fluctuated dramatically over the summer of 1975 (McNabb gt al., 1975). In Lake 1, the total number of algal cells ranged from 1,500 to 70,000/m1 with a summer mean of 15,300 cells/ml. In Lake 4, total numbers ranged from 100 to 74,000/m1 with a summer mean of 12,400 cells/m1. Lake 1 was dominated by species of green algae and diatoms while Lake 4 was dominated by species of green and blue-green algae. Table A-1 gives a list of major species which were enumerated in water samples. METHODS This study took place during the first growing season in which the MSU water reclamation facility was operating. Combina- tions of plexiglass enclosures were used to partition the ecosystem so that benthic, planktonic and macrophytic production and respir- ation could be estimated. The nature of the enclosures is shown in Figure 2. They were made of 0.64 cm plexiglass and were 0.57 m by 0.57 m by 1.5 m in dimensions. For a set of measurements on primary production and community respiration, one enclosure was pushed into the sediments in an area free of aquatic vascular plants. In Lake 4, where Elodea canadensis existed over the entire study area, 3 m by 3 m sections were randomly cleared 6 weeks in advance of measurements. This time interval proved satisfactory in that E, canadensis recolonization did not occur and Waters (1961) had shown it to be adequate for epipelic algal development. In Lake 4 where the submersed vascular plants were an important component of the ecosystem, a second enclosure was pushed into the sediments to surround a stand. A third enclosure, with one end sealed using plexiglass, was filled with water to pond-depth and placed upright in the water column with the sealed end down. The diurnal oxygen curve method of Odum and Hoskins (1957) and McConnell (1962) was used to measure rates of primary production and respiration within enclosures. The need to make corrections for 9 10 Figure 2.--P1exiglass enclosure (A) used for jg_situ productivity and respiration measurements (B). 11 12 surface gas exchanges was eliminated by covering the enclosed water with adjustable plexiglass lids. The particular technique was based on the light-dark bottle concept of Gaarder and Gran (1927). By measuring oxygen changes between dawn and dusk, it was possible to estimate net primary production (MP); the differences between oxygen concentrations at dusk and the following dawn were taken as esti- mates of respiration (R). A Yellow Springs Instrument Model 54 oxygen and temperature probes and meter were used in the measurements. Frequent membrane changes and Winkler standardizations (azide modification of APHA, AWWA and WPCF, 1971) assured accurate dissolved oxygen determina- tions; :_1% full scale deflection, the maximum being 1 0.2 mg/l. Measurements were taken at the surface, at depths of 0.5 m and 1.0 m, and at the water-sediment interface at depth 1.1 m. Oxygen concentrations were plotted against depth for each series of mea- surements. Dawn-dusk and dusk-dawn differences in oxygen concen- trations for enclosed water columns as a whole WEre obtained from these graphs using a polar compensating planimeter. Mean hourly net productivity and respiration rate were obtained by dividing the change in oxygen concentration for a light or dark interval by the number of hours in that interval. Hourly net productivity and respiration rates were added to obtain an estimate of gross primary productivity. Daily productivity was made equal to the hourly rates multiplied by the number of daylight hours. Daily respiration was made equal to the hourly respiration rate multi- plied by 24. Knowing the volume (390 liters) and areal dimensions 13 of the enclosures, the results were expressed as g 02/m2/day. For reference and in the interest of uniformity suggested by Vollenweider (1969), gross primary productivity was calculated as g C/mz/day and the results presented in the Appendix (Table A-2). These data were obtained by assuming a mean photosynthetic quotient of 1.2 (Westlake, 1963) and multiplying g 02/m2/day by 0.312 (Vollenweider, 1969). With the combination of enclosures and the methods employed, primary productivity and community respiration estimates of the eco- system were partitioned into planktonic, benthic and vascular plant- epiphyte components. Rates of the planktonic component were mea- sured directly in enclosures containing only the plankton. Rates in the benthic component were obtained by subtracting planktonic rates from rates measured in enclosures open to the bottom, but lacking macrophytes. Rates for the vascular plant-epiphyte com- ponent were obtained by subtracting planktonic and benthic rates from measurements in enclosUres open to the bottom and containing macrophytes. Sampling in these enclosures took place at regularly spaced intervals between 27 April and 17 October, 1975. Studies spanning 36-hour cycles were conducted 13 times in Lake 1 and 15 times in Lake 4. Only one of the lakes was surveyed on any particular date. Dates were not selected for particular conditions of weather. On each occasion a position was selected along the lake perimeter and the enclosures were set from a boat in 1.1 m of water. Wooden stakes were used to hold them steady in the water. Two replicates of each type of enclosure were used at any given time of measurement. 14 The units were positioned approximately 10 hours before water analy- ses were begun. A11 construction materials were non-growth inhibit- ing as described by Dyer and Richardson (1961). Percent transmittance of incident light was determined at each sampling with a submarine photometer connected to a low resis- tance galvometer (Fred Schueler, Waltham, Massachusetts). Trans- mittance, as compared to surface measurements, was determined at the 0.5 to 1.0 m depths. The quantity of incident solar radiation was determined by planimetry of curves from a recording Epply pyrheliometer. (Michigan State University's South Farm climatologi- cal station, approximately one kilometer north of the study area, maintained this pyrheliometer. Chlorophyll-a was selected as the quantitative measure for phytoplankton abundance. Equal volumes of water were taken from the two plankton stations, combined and immediately transported to the laboratory. After major zooplankters were screened (250 micron mesh), volumes of water were filtered through 0.45 :_.02 micron glass filters (Gilman Instrument Company, Metricel G.A.-6) at a pressure of 20 to 30 cm Hg. A 0.45 micron pore size has been shown to remove all significant phytoplankton (Lasker and Holmes, 1957). Samples were dark-stored at 4°C for 2 to 3 weeks before extraction was performed (Collins gt 11., 1975). Extraction was carried out under reduced light by manual grinding in an aqueous solution of reagent grade 90% acetone rendered basic with magnesium carbonate. After centrifugation, spectrophotometric procedures were carried out on a Bausch and Lomb Spectronic 20 to obtain an estimate of 15 chlorophyll-a. Calculations followed those of Talling and Driver (1963). Corrections for phaeophytins followed those of Lorenzen (1967). Knowing the volume and areal dimensions of the enclosures, mean values of three replicates were expressed as mg/m2 chlorophyll-a. E, canadensis was harvested from enclosures at the end of a period of measurement. While in the field, plants were handwashed trying not to brush off silt-like deposits found on the leaves and stems. Dry and ash weights were determined in the laboratory under desiccation at 105° and 550°C respectively. Ash-free dry weights (organic weights) were computed. Biomass was expressed as g/m2 ash- free dry weight. RESULTS Abundance of Primary Producers A one-way analysis of variance was performed on phytoplankton chlorophyll-a concentrations found in Lakes 1 and 4 (Table 2). Means were positively correlated with variances; the values were coded by adding one and making a logarithmic transformation (Snedecor and Cochran, 1967). Homogeneity of variances was then varified by Fmax tests. Lake 1 had significantly greater phytoplankton chlorophyll-a concentrations than Lake 4 (99.5% probability level). Concentra- tions in Lake 1 ranged from 2.7 to 52.1 mg/m2 with a seasonal mean of 17.4 mg/m2 chlorophyll-a. Phytoplankton chlorophyll-a in Lake 4 ranged from 0.0 to 5.9 mg/m2 with a seasonal mean of 2.5 mg/mz. In Lake 1 it was consistently high early in the season ( up to 25 June), and then varied widely. In Lake 4, concentrations were relatively high in early spring, then gradually decreased as the study pro- gressed. E, canadensis vegetation in Lake 4 was sufficient in early .spring to quickly shade the sediments upon which it was growing. Biomass within the enclosures ranged from 2.8 to 408.5 g/m2 ash- free dry weight with a mean of 220.6 g/m2 (Table 3). Percent ash of dry weight increased over the season; carbonate deposits were observed to increase concurrently on the adaxial plant surfaces. 16 TABLE 2.--Mean phytoplankton chlorophyll—a concentrations in 17 mg/m2 in 1.1 m of water. Date Lake 1 Date Lake 4 5/6/75 52.1 4/27 5.9 6/3 12.9 5/16 5.0 6/17 20.7 6/10 3.2 6/25 34.5 6/19 1.8 7/6 7.8 7/1 4.4 7/15 14.4 7/10 5.8 7/29 7.1 7/22 3.4 8/7 21.8 8/2 0.3 8/16 9.0 8/12 0.8 8/27 3.2 8/21 3.3 9/9 34.3 9/14 0 9/19 6.3 9/24 0.8 10/1 2.7 10/8 1.6 10/17 0.5 Seasonal mean 17.4 2.5 18 TABLE 3.--Weight determinations (g/mz) for the biomass of Elodea canadensis taken from enclosures within Lake 4. Date Percent Ash of Ash-free Dry Dry Weight WEIQDt 4/27/75 - 108-1 8.2 5/16 - 64.9 165.5 6/10 27.6 243.3 313.7 6/19 24.3 226.1 320.3 7/1 29.0 282.6 226.3 7/10 30.4 273.5 71.6 7/22 38.0 305.9 297.6 8/2 36.3 173.8 108.7 8/12 38.1 269.2 338.5 8/21 40.1 375.7 232.0 9/2 35.6 363.4 137.6 9/14 40.0 408.8 102.8 9/24 36.8 124.0 227.5 10/8 49.9 272.6 222.3 10/17 43.4 199.3 195.9 19 Conditions of Light The input of energy for study dates on Lake 1 varied less than for dates of measurement on Lake 4; seasonal coefficients of variation were 35% and 55% respectively (Table 4). Over the period of the study, enclosures in Lake 4 received less solar energy than those of Lake 1; means were 3401 kcal/mz/day and 4398 kcal/mZ/day respectively. In Lake 1, light attenuation was related to the amount of chlorophyll-a found in the water; the product moment correlation coefficient (PMCC) was 0.72. There was an approximately linear relationship when the extinction coefficient (Vollenweider, 1955) was related to the logarithmic function of the chlorophyll-a con- centrations (Figure 3). A relationship such as this was not found in Lake 4; there the PMCC was 0.20. The percent transmittance of incident light in Lake 4 (cf. Table A-3) was greatly reduced by canopies of E, canadensis (Figure 4). Light transmittance into a canopy was regularly reduced to 1% of incident within the first 15 cm of vegetation. No differences in extinction occurred between the water above a canopy and the same stratum of water in areas that had been cleared of vegetation. Primary_Productivity and Respiration A two-way analysis of variance with replicates was performed on changes in the dissolved oxygen content of the water columns in enclosure types within Lakes 1 and 4 for daylight and darkness periods after the methods of Snedecor and Cochran (1967). Tukey's TABLE 4.--Dai1y surface radiation (kcal/mZ/day) on dates of study in Lakes 1 and 4. 20 Date Lake 1 Date Lake 4 5/6/75 3558 4/27/75 846 6/3 5166 5/16 4686 6/17 4023 6/10 6192 6/25 3920 6/19 4324 7/6 4648 7/1 5240 7/15 4806 7/10 3948 7/29 6659 7/22 5766 8/7 5713 8/2 1211 8/16 5205 8/12 4051 8/27 5341 8/21 611 9/9 5272 9/2 3883 9/19 1344 9/14 4328 10/1 1518 9/24 3179 10/8 1549 10/17 1094 Seasonal Means 4398 3401 Coefficient of Variation 35% 55% 21 Figure 3.--The relationship between the extinction coefficient of light and the chlorophyll-a concentrations found in Lake 1. Extinction Coefficient 22 l 1.0 Log“, Chlorophyl l-a (mg/l) 2.0 23 Figure 4.--Light transmittance occurring in the open water and in a canopy of Elodea canadensis in Lake 4. Depth (m) 24 °/o of Surface Light '09.. V I V 1 open water 1.0 - '0' h 10 E. canadensis c onopy at 0.5m 25 test for additivity was used to check block-treatment interaction of all analyses. In Lake 1, daylight oxygen changes within plank- tonic and benthic enclosures were significantly different (95% prob- ability level). The planktonic enclosures gained more oxygen in the daytime than the benthic enclosures. In Lake 4, these oxygen changes were not significantly different; however, E, canadensis enclosures had significantly greater oxygen gains during the day and greater oxygen loss at night than the planktonic and benthic enclosures (99% probability level). These results point to func- tionally different relationships between oxygen metabolizing com- ponents of the ecosystems of these lakes. Gross primary productivities in the planktonic and benthic enclosures of Lake 1 were not significantly different at the 95% level of confidence. This result was taken to mean that produc- tivity of benthic algae, if it occurred at all, was negligible relative to that of the phytoplankton. The procedure employed in this study to estimate productivity of the benthic algal community was to subtract daily gross production in the planktonic enclosures from daily gross production in the benthic enclosure. When this was done, values for the benthic estimate were positive 35% of the times, zero 4% of the times, and negative for the remaining per- centage of times. The mean for this value over the study period was -0.66 g 02/m2/day. The chi-square goodness of fit test (Snedecor and Cochran, 1967) showed that individual benthic esti- mates were normally distributed around a mean not significantly different from zero (95% probability level). If gross primary 26 productivity occurred in the benthos, these results suggest that on the average it was small and that the experimental technique used was not sufficiently precise to measure it. The benthic community of Lake 1 was taken to be essentially non-photosynthetic, consuming oxygen produced by the plankton at a measurable rate both day and night. Since daytime gains and nighttime losses in dissolved oxygen in the water columns of planktonic and benthic enclosures in Lake 4 were not different for the study dates, as cited above, the benthic component of the lake was taken to have a negligible impact on both gross primary production and respiration of the ecosystem. When . gross primary productivity of the benthos was calculated by the method of this study, the mean for the sampling dates was 0.23 g 02/m2/day. The chi-square goodness of fit test as presented above for Lake 1 benthic productivity was applied to the daily benthic estimates from Lake 4 with the same result as for Lake 1. The benthic community of Lake 4 was taken to be essentially non- photosynthetic for the study interval. The rate of respiratory oxygen use by the benthic community was also too small to measure. With these considerations in mind, a modified two-way analysis of variance was performed on the estimates of gross primary productivity within and between lakes (Table A-4). Negative esti- mates for the benthic components were included for purposes of this analysis. The procedure involved a classification with unequal numbers having proportional subclasses (Snedecor and Cochran, 1967). After logistic transformations, variances were found to be 27 homogeneous by Tukey's test. Between lake comparisons were per- formed even though measurements were taken on the lakes on different dates. Observed differences between lakes (e.g. radiation influx), as well as unobserved differences, introduced an error not incor- porated into the pooled residual used to test between lake differ- ences. Pooled data showed significant differences existing between :1 Lakes 1 and 4. Several LSD tests were conducted to determine where i differences between lakes existed. Results of all tests (LDS having the lowest significance level of 95%) indicated a hierarchy of primary productivity as: Lake 1 Phytoplankton = Lake 4 Macrophyte-Epiphyte Complex V Lake 4 Phytoplankton V Lake 1 Benthos = Lake 4 Benthos Estimates of gross primary production, community respira- tion, and net primary production for the-days of study are shown in Figures 5 and 6. In Lake 1, phytoplanktonic oxygen production exceeded daily planktonic respiration except for three dates, on these latter occasions, P/R was very close to 1 (Figure 5A). Gross productivity was relatively high from mid-June to early August, and again on dates in September. In general, respiration in the plank- tonic community was high when gross primary productivity was high, and was low when algal productivity was low. The mean P/R ratio in the plankton for the dates of study was 1.31. 28 Figure 5.--P1ankton (A) and whole ecosystem (B) productivity and respiration in Lake 1 in 1975. Differences between gross productivity (0) and respiration OS) are darkened to show net productivity. 29 to. .5 .8 E. (m to . k ¥ I .m._ n m\n_ cums. cotEoE .93... < flop/zw/Zo 5 N. 200 l to. .5 (m .R .5 .B 30 mmdwmk. cows. Eofimxmoow 2055 33.9.. m q- (I) flop/Zw/zo 5 N_ 31 Figure 6.--Plankton (A), Elodea canadensis plus its epiphytes (B), and whole ecosystem (C) productivity and respiration in Lake 4 in 1975. Differences between gross produc- tivity (o) and respiration (£0 are darkened to show net productivity. 32 _\O_ _\m de n mi :82 _\m 200 _\N _\0 _\m ‘/ coined v9.3 4 ADP/zw/zo o 33 200 .\.0. .8 _\_m .\w .Ao to . a/ a .. es... .. . ./ 2...... J . ....._ .. ,. _ 2””... 5: 4.. .. ,_ \«..”’a’”//’ . 00.. um}. cams. 3.293 + m.mcmuocoofl v 3.0.. m q (I) Kop/zw/zo 6 N. 34 2.5 .qu :0 .x.» .x... .8 <0 i < 4 E. 4 2:: 4 . . ___ , _. 4: , _____ : . . 4.. .4. ‘ .1 a... :. \< madamxa zoos. 603326 29.3 v9.01. 0 v a) KDp/Zw/Zo b N. 35 From the discussion of benthic oxygen metabolism given above, where gross productivity was taken to be zero, the seasonal P/R ratio for that community was zero. Figure 5B, considering the eco- system as a whole, shows that daily oxygen consumption regularly exceeded gross production on dates early in the growing season. From early August to late September, whole ecosystem enclosures tended to have P/R ratios close to, or greater than 1. Measured respiration rates for the benthos can be obtained from Figure 5 by subtracting the planktonic respiration rate from the respiration rate given for whole ecosystem enclosures. Instances of hetero- trophy and autotrophy balanced out over the study interval so that the mean P/R ratio for the ecosystem was very close to l (0.99). Figure 6 presents data in a similar manner for Lake 4. The P/R ratio of the benthic community was taken to be zero because of lack of gross primary productivity. Since respiration in the benthos was too small to measure, that effect is not included in the data of Figure 6. The plankton community surrounding the vascular plants of the lake tended to be heterotrophic; plankton respiration exceeded phytoplanktonic gross 02 productivity on 12 of 15 study dates (Figure 6A). A mean P/R ratio of 0.92 was observed for the plankton over the study interval as a whole. Figure 6B shows that the Elodea canadensis-epiphyte component of the ecosystem switched from an autotrophic to a heterotrophic, then back to an autotrophic mode during the study. Comparison of Figures 6A and 68 reveals that rates of gross production and respiration in the vascular plant complex were regularly 2 to 3 36 times greater than the same rates in the plankton. Because of this, parallel pulses and recessions are evident when comparisons of oxygen metabolisms for the vascular plant-epiphyte component and the eco- system as a whole are made with Figures 68 and 6C. This dominance by the E, canadensis-epiphyte component is further demonstrated by no substantial differences of P/R ratios, these being 0.98 for E, canadensis and 1.00 for the whole ecosystem. Estimates for seasonal means of productivity and respira- tion for the dates of study are given in Table 5. One hundred per- cent of gross primary productivity and an estimated 75% of respir- ation in the Lake 1 ecosystem were accounted for by the plankton community. Mean gross primary production and respiration for the ecosystem of Lake 4 were found to be substantially higher than for Lake 1. 0n the average, 26% of the gross primary production in Lake 4 was attributable to the phytoplankton; 74% was attributable to the E, canadensis-eiphyyte complex. Respiration in these two components of the system was in the same ratio as gross production. 37 .mceoe mmmcu cwmpno o» now: mew: metamoFUcm umcucma use uwcoaxcmya seem mmzpe> .Amogpcmn mgu an uumemm on. c 0.». cw mwgzmopucm upgucmn can upcopxcmpa mo ucwpcou cmmxxo mewuiugmw: ace mswuxmc cw mmmcmcu :mmzumn umwxm go: new mwucmcmmwev achPewcmvm mmseummu .z—mcwugooum umpmsnvm mw:_m> Emumxmoum .muo» .om..- mm: cowgoacoga uwgpcmn am: “new comum>gmmno cwgzmmme mew use .0 we: mosucma cw cowuuzcoea mmoem page cowpasammm co om.p op umpumegou mameppmw mesh .meamopocm uwgucmn eoew mczmopucm owcopxcmpa mo mmop cmmxxo apwmu cwumpaopmo we cowpumgpnam an umcwmuno mm: mm.o to wzpm> some szpowuoauoga mo mamas chommmmii.m m.m<><><>< ><><><><><><>< ><><><><>< x 21 46 XXX ><><><><>< xxx x 17 45 1975). aWeekly samples from July to October 1975 (McNabb et al., 53 TABLE A-2.--Gross primary production in g C/mZ/day for ecosystem components in Lakes 1 and 4. Date Plankton Benthos Lake 1 5/6/75 0.67 0.28 1.03 0.57 6/3 0.35 -0.22 0.17 -0.02 6/17 0.83 -0.06 0.51 0.84 6/25 3.08 -l.38 3.10 -2.40 7/6 2.04 0.17 1.41 -0.29 7/15 3.61 -2.88 2.50 0.23 7/29 4.81 -l.1l 3.77 0.06 8/7 0.72 -0.30 0.58 -0.06 8/16 0.86 0.18 0.42 -0.12 8/27 1.49 -0.51 2.29 -0.63 9/9 3.55 -1.23 4.79 -0.30 9/19 1.29 0.89 1.82 0.00 10/1 0.14 0.48 0.13 -0.07 Mean 1.77 -O.30 Range 0.13 - 4.81 -2.88 - 0.89 TABLE A-2.--Continued. 54 E. canadensis- a Date T'epiphytes Plankton Benthos Lake 4 4/27/75 0.81 0.92 1.21 0.24 0.66 0.91 5/16 0.89 0.71 0.39 1.70 0.31 0.37 6/10 1.92 0.17 ' 0.02 3.24 0.98 0.82 6/19 1.47 0.82 0.87 2.94 0.58 1.19 7/1 3.21 0.55 0.93 2.77 0.87 0.63 7/10 1.90 0.19 0.16 0.40 0.03 0.40 7/22 2.11 0.87 1.24 3.31 0.78 0.72 8/2 0.39 0.78 0.53 1.19 0.48 0.62 8/12 1.29 0.50 1.39 1.75 0.64 0.96 8/21 0.16 0.44 0.62 1.07 1.06 0.78 9/2 4.10 0.57 0.75 1.86 0.55 0.61 9/14 2.64 1.03 0.57 1.35 0.65 0.59 9/24 1.23 0.39 0.35 1.45 0.27 0.11 10/8 0.69 0.24 1.22 0.32 10/17 0.22 0.00 0.42 0.01 Mean 1.60 0.55 0.68 Range 0.16 - 4.10 0.00 - 1.06 0.02 - 1.39 aColumn is Lake 4 benthic enclosure that was determined to be insignificantly different from Lake 4 plankton enclosure at the 95% probability level. 55 TABLE A-3.--Percent transmittance of surface radiation at the 0.5 m and 1.0 m depths in Lakes 1 and 4 on dates of study. Date 0.5m 1.0m Lake 1: 5/6/75 40 4 6/3 46 6 6/17 27 0 6/25 37 3 7/6 23 0 7/15 29 2 7/29 26 6 8/7 59 41 8/16 56 47 8/27 -- -- 9/9 70 60 9/19 88 85 10/1 -- -- Lake 4: 4/27 49 31 5/16 64 44 6/10 70 56 6/19 66 49 7/1 72 59 7/10 67 35 7/22 56 31 8/2 63 45 8/12 64 45 8/21 82 61 9/2 72 61 9/14 70 65 9/24 89 85 10/8 -- -- 10/19 -- -- .e axe. mwmcmvmcmu aw cu .e 0.6. couxcmpa «a 56 a .. axe. 6.50:66 w .m .. axe. eooxea.a w .aa mpmo.o momm.m mm Pmauwmmm umFooa ao.ma mm.m oeea.o cake.o P e. a 3. m> .m a .e mcomwgmasou nmpooa mmm~.o .m-.e cm .eu m> ea. Peacemoa «mom.o wmom.m N— cowpumemucu gm.Nm cc.“ AFNm.. “.mm.. . new m> ea a~m~.~ N. axea e 6.6. .moo.o e._..o mm ..m ms Fa. .eeewmom «moo.o oo...o N. eoeooeeope. em.ma cm.N~ omo~.o omo~.o . e.m as .a ommo.o NF mzmo . 6.6. ~o>m. .mwm . m: mm we moczom .mmxmp cmwzpmn can cmcuwz mwpoewpmm cowuozuoca acmswsa mmoem co umELomng muccwcc> mo mwmx_mcm xoziozp uwwmmuos a mo mupzmmmii.¢t< w.m<~ HICHI RRIES NINWNNIN