THE EFFECTS OF THE AQUATIC ANGIOSPERM CERATOPHYLLUM DEMERSUM 1.. ON PBOSPHORUS DYNAMICS IN A LABORATORY SYSTEM That: {or flu Degree OI M. S. MICHIGAN STATE UNIVERSITY Russell J. Erickson 1976 JIIIIIIIIIWIIIIIIIIIIIIIIIIII ‘ I 55555 3 1293 01059 9268 ABSTRACT THE EFFECTS OF THE AQUATIC ANCIOSPERM CERATOPHYLLUM DEMERSUM L. ON PHOSPHORUS DYNAMICS IN A LABORATORY SYSTEM By Russell J. Erickson Theoretical considerations and results of investigations reported in the literature suggest that aquatic macrophytes have both direct and indirect effects on the dynamics of phosphorus in aquatic systems. Direct effects are defined as resulting from the uptake or release of phosPhorus by the plants and indirect effects are those resulting from effects of the macrophytes on other properties of the system, such as pH, which in turn affect the behavior of phosphorus. The integrated effect of all involved factors has never been directly measured on any system and cannot be inferred from the existing body of knowledge. In this study, the effects of the aquatic angiosperm Ceratophyllum demersum L. on the behavior of phosphorus in simple water/sediment systems in ZO-gallon aquaria were studied. Phosphorus distribution between water and sediment was found to be significantly affected both by the presence of this plant and by changes in photoperiod which caused the plants to either grow and accumulate phosphorus or to senesce and release phosphorus. Direct effects of g; demersum were clearly present and exerted such a large influence on phosphorus levels that indirect effects were obscured and could be only tentatively established with the use of a parameter, the weighted mean rate of phosphorus removal, which was designed to compensate for all or most of the direct effect of the plants. @7622? Russell J. Erickson To better establish the existence and nature of indirect effects of g; demersum, a model of phosphorus dynamics in the aquaria was developed and used to conduct simulations of the aquarium experiment, with data partly from the aquarium experiment and partly from a series of beaker experiments in which phosphorus removal from water was measured as a function of pH, alkalinity, hardness, and orthOphosphorus concentration in small water/sediment systems comparable to those in the aquaria. These simulations roughly duplicated the values of phosphorus observed in the aquarium experiment, but a significant lack of fit occurred. Sensitivity analysis of these simulations substantiated the large, direct effect of Q; demersum on phosphorus behavior, and also showed slight effects on phosphorus dynamics due to decreases in evaporation caused by plants and major effects due to plant-induced changes in pH. It was concluded that these indirect effects most likely were also present in the actual aquarium experiment, though probably to different extents and in the company of effects from other factors that could not be included in the model. Testing of the weighted mean rate of phosphorus removal in these simulations showed that this parameter had some shortcomings, most importantly being influenced by direct effects of plants, but that, if used judiciously, it could serve as a useful indicator of indirect effects of plants on phosphorus dynamics, at least in systems comparable to the ones described here. THE EFFECTS OF THE AQUATIC ANCIOSPERM CERATOPHYLLUM DEMERSUM L. ON PHOSPHORUS DYNAMICS IN A LABORATORY SYSTEM By Russell J. Erickson 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 express my gratitude to E. W. Roelofs for his guidance and friendship as my advisor during my years as both an undergraduate and graduate student at Michigan State and to C. D. McNabb and F. M. D'Itri for their time and help in reviewing my research and this thesis. I would also like to thank all the faculty and students that I have been privileged to be associated with at Michigan State, who have in no small way contributed to the thoughts expressed in this thesis. I would like to especially thank Chuckie for all the good times. The research discussed here was made possible through support from a National Science Foundation graduate fellowship and from training grant T900-331 of the Environmental Protection Agency. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . Forms of inorganic phosphorus in water and sediment . Effects of chemical and physical factors on Redox potential. . . . . . . pH . . . . . . . . . . . . . . . . . Dissolved organic matter . . Sediment properties. . . . . Multivalent cations. . . . . Temperature and turbulence . Indirect effects of aquatic macrophytes Redox potential. . . . . . . pH . . . . . . . . . . . . . Dissolved organic matter . . Sediment properties. . . . . Multivalent cations. . . . . Temperature and turbulence . Direct effects of aquatic macr0phytes Statement of problem and plan of research . MATERIALS AND METHODS. . . . . . . . . Aquarium experiment . . . . . . . Experimental apparatus . . . Plants, water, and sediment. Experimental procedure . Beaker experiment . . . . . . Treatments . . . . . . Experimental procedure . RESUDTS AND DISCUSSION . . . . . . . . on phosphorus phosphorus dynamics on phosphorus Aquarium experiment . . . . . . . . . . . Plant growth and phosphorus content. Hater budget . . . . . . . . . . . . Twenty-four hour sampling. . . . . . Temperature, dissolved oxygen, Hardness and alklinity . . . Phosphorus . . . . . . . . . Beaker experiment . . . . . . . . Regression analysis. . . . Modeling and simulation of aquarium experiment . Sensitivity analysis . . . . iii and d SUMMARY AND CONCLUSIONS 0 o o o o o o o o o o o o o c o o o o o o o o 69 APPENDIX A "' ANALY‘T ICAL TECHNIQUES o o o o o n o o o o o o o o I o o 71 APPENDIX B - SUMMARY OF REGRESSION EQUATIONS . . . . . . . . . . . . 73 LITERATURECITED..........................75 iv Table (4“ 10 Table 10 11 12 LIST OF TABLES Summary of water chemistry of pond 2 of the Michigan State University Water Quality Management Project for the summer of 197uI I I I I I I I I I I I I I I I I I I I I I I I I I I I I Water chemistry treatment levels for beaker experiment. . . . Summary of plant weights and phosphorus contents in aquarium exmriment I I I I I I I I I I I I I I I I I I I I I I I I I Volumes (liter) of aquaria, influents, effluents, and evaporation in equarium experiment. . . . . . . . . . . . . . Results of twenty-four hour sampling during first run of aquarium experiment (day 14 to day 15). . . . . . . . . . . . Mean values of water chemistry parameters during aquarium exmriment I I I I I I I I I I I I I I I I I I I I I I I I I I Total phosphorus budget for first run of aquarium experiment (12-hr. photoperiod, aquaria 1, 4, 5 with plants) . . . . . . Total phosphorus budget for second run of aquarium experiment (14-hr. photoperiod, aquaria 2, 4, 6 with plants) . . . . . . Total phosphorus budget for unmodified simulations of first run of aquarium experiment (12-hr. photoperiod, aquaria 1, 4, 5 With PJ-antS) I I I I I I I I I I I I I I I I I I I I I I I I Total phosphorus budget for unmodified simulations of second run of aquarium experiment (14-hr. photoperiod, aquaria 2, 4, 6 With PhntS)I I I I I I I I I I I I I I I I I I I I I I I I Total phosphorus budgets for modified simulations of aquarium 5 of first run.of aquarium experiment (12-hr. photoPeriod). . Total phosphorus budgets for modified simulations of aquarium 4 of second run of aquarium experiment (14—hr. photoperiod) . page 20 23 27 29 31 45 57 58 65 Figure 10 11 LIST OF FIGURES page Midday temperature vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). . . . . . . . . . . . 33 Midday dissolved oxygen concentration vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). . . . . . 34 Midday pH vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). . . . . . . . . . . . . . . . . . 35 Hardness vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4 , 6 With plantS) I I I I I I I I I I I I I I I I I I I I I I I 37 Alkalinity vs. time for all aquaria during experimental runs one (left, aquaria 1, a, 5 with plants) and two (right, aquaria 2, u, 6 with plants). . . . . . . . . . . . . . . . . . 38 Total phosphorus vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, a, 6 with plants). . . . . . . . . . . . . . . . . . 40 Dissolved phosphorus (left) and orthophosphorus (right) vs. time for all aquaria during experimental run two (aquaria 2, 4, 6 with plants) . . . . . . . . . . . . . . . . . . . . . . . 41 Response surfaces of the removal of orthophosphorus from water (DeltaP) vs. mean pH and mean orthophosphorus concentration at hardness = 250 mg/l CaCO and alkalinity = 150 mg/l Ca003 in beakers without (top) an with (bottom) sediments . . . . . . . 5o Unmodified simulations of hardness vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). . . . . . . . . . 54 Unmodified simulations of alkalinity vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). . . . . . . . . . 55 Unmodified simulations of total phosphorus vs. time for all aquaria during experimental runs one (left, aquaria 1, 4 5 with plants) and two (right, aquaria 2, 4, 6 with plants) . . . 56 vi Figure page 12 Modified simulations of total phosphorus vs. time for aquarium 5 of experimental run one (left) and aquarium 4 of experimental run two (right). . . . . . . . . . . . . . . . . . 63 vii INTRODUCTION Due to its role in limiting productivity, the distribution and dynamics of phosphorus in aquatic systems have been the subjects of much research. In particular, the exchange reactions of inorganic phosphorus between dissolved and solid states (e.g., precipitation, c0precipitation, adsorption) have received considerable attention since they account for a significant part of the phosphorus removed to the tropholytic zone of lakes and to sediments. Even more importantly, these reactions control the retention of phosphorus in sediments, including not only the inorganic phosphorus originally deposited in an inorganic form, but also the inorganic phosphorus released from decomposing, sedimented organic matter. Because of this, these exchange reactions exert a tremendous influence on the amount of phosphorus available to primary producers and thus affect the productivity of all components of aquatic systems. The following discussion will consider (1) the forms of inorganic phosphorus found in water and sediments (indicative of the exchange reactions occurring), (2) the chemical and physical factors which are known to affect the exchange reactions of inorganic phosphorus, (3) the indirect effects of aquatic macrophytes on phosphorus because of their effects on these chemical and physical factors, and (4) the direct effects of aquatic macrophytes on phosphorus dynamics through their uptake, retention, and release of phosphorus. 2 FORMS OF INORGANIC PHOSPHORUS IN WATER AND SEDIMENT Except where large inputs of synthetic inorganic phosphorus compounds occur, dissolved inorganic phosphorus exists almost exclusively as orthophosphate (H1P041‘3,i=0...3) and its various complexed forms, such as FeHPOL,+ (Stumm and Morgan, 1970). Orthophosphate is quite insoluble in water in the presence of a variety of cations and minerals and can be removed by adsorption to clays and metal hydroxides, especially ferric hydroxide (Stumm and Morgan, 1970), coprecipitation with calcium carbonate (Zicker gt 21., 1956; Otsuki and Wetzel, 1972), and precipitation as a phosphate mineral. Stumm and Morgan (1970) suggest that the phosphorus- containing minerals of significance in natural waters include various apatites ((Ca,H20)10(F,OH)2(P04,C03)6), brushite (CaHPoh-ZHZO), variscite (Alma-21120), strengite (FePOLVZHZO), and wavellite (A13(OH)3(P04)2). In addition, vivianite (Fe3(POh)2-8H20) has been reported in strongly reduced sediments (Mackereth, 1966; Rosenqvist, 1970). The formation of these phosphate minerals appears to be especially important in sediments for retaining phosphorus that was initially deposited in forms unstable under sediment conditions. For example, much of the apatite found in sediments is apparently formed via conversion of deposited calcium carbonate and orthophosphate released from organic matter decomposing in the sediment (Stumm and Morgan, 1970; Hilliams and Mayer, 1972). Similarly, vivianite would be formed from orthophosphate released from decaying organic matter and from ferrous iron derived from reduction of precipitated ferric iron (Williams and Mayer, 1972). Very few data exist regarding actual quantities of the above forms of solid inorganic phosphorus present in natural systems. The removal of orthophosphate from water and its retention in sediments have been 3 variously attributed to precipitation and coprecipitation with ferric iron (Einsele, 1936, 1938; Mortimer, 1941, 1942, 1971; Mackereth, 1966), manganese (Mackereth, 1966), and calcium (Hepher, 1958; Zicker gt al., 1958; Wentz and Lee, 1969), but these studies relied mainly on laboratory data, theoretical calculations, and correlations in field data, with little or no direct measurement of forms of phosphorus in and above sediments. The work by Mortimer (1941, 1942), though, did provide convincing evidence of the role of hydroxides.and hydrated oxides of ferric iron at the surface of oxidized sediments in removing orthophosphate from the water and preventing its escape from lower, reduced layers of sediments. More recent work on Lake Erie and Wisconsin lake sediments, using extraction techniques with some selectivity for forms of phosphorus, suggests that inorganic phosphorus is present in sediments associated with aluminum, iron, and calcium, with apatite and an amorphous hydrated iron oxide-orthophosphate complex being predominant (Williams gt al., 1971a, 1971b, 1971c; Shukla gt al., 1971; Li et al., 1972; Williams and Mayer, 1972). Additionally, data from Lake Erie suggest that, while surface sediments contain appreciable quantities of apatite, organic phosphorus, and sorbed (Al and Fe bound) orthophosphate, the values of the last two decline quickly with depth in the sediment, presumably in part due to conversion to apatite, which increases with depth (Williams and Mayer, 1972). Nelson (1967), using similar extraction techniques, reported the presence of strengite and variscite in quantities exceeding apatite in fresh-water sediments, but his results are questionable since these techniques do not actually distinguish these minerals from other forms of aluminum and iron bound phosphorus (Williams gt al., 1971a). As mentioned above, vivianite has also been found in some sediments, but its occurrence 1+ is infrequent and erratic, presumably due to the extremely low redox potentials required for its formation and to competing reactions (Rosenqvist, 1970). EFFECTS OF CHEMICAL AND PHYSICAL FACTORS ON PHOSPHORUS DYNAMICS The rate and extent of removal of dissolved inorganic phosphorus from water have been correlated with several chemical and physical factors, including redox potential, pH, dissolved organic matter, multivalent cations, sediment properties, temperature, and turbulence. In most cases, the effects of these factors are easily understandable when they are considered in relation to the forms of inorganic phosphorus discussed above. Redox Potential A large concentration of inorganic phosphorus is typically found in the hypolimnion of lakes with extreme clinograde oxygen curves. Einsele (1936, 1938) correlated this to the release of ferrous iron from the sediment, suggesting that phosphorus is held in oxidized sediments in association with ferric iron (FePOu, Fe(OH)3-P) and is released upon ,reduction of the ferric iron to ferrous iron, the phosphates and hydroxides of the latter being much more soluble. Subsequent work by Mortimer (1941, 1942, 1971) established more precisely the mechanisms involved in this phenomenon and is the basis for the following discussion. When oxygen is present in adequate quantities (1 mg/l 02 or more) in water overlying sediment, there exists a thin surface layer in the sediment ‘in which oxygen penetrates in amounts sufficient to maintain the redox potential (E?) at greater than 200 mv. This layer, the oxidized microzone, can vary from less than 1 mm to greater than 2 cm in thickness, depending on actual oxygen concentration and turbulence, and is characterized by 5 iron being almost exclusively in the ferric form, primarily as ferric hydroxide. Orthophosphate present in the overlying water is removed in large part by adsorption to this ferric hydroxide, as is orthophosphate released from decomposing organic matter deposited on the sediment. More importantly, the oxidized microzone prevents the passage of phosphorus to the water from lower layers of sediment, where large concentrations of dissolved phosphorus and ferrous iron exist due to reducing conditions. Virtually all upward diffusing orthophosphate will be precipitated or adsorbed' when it reaches this layer. While other solid inorganic forms of phosphorus certainly participate in this exchange between sediment and water, the correlation between the release of ferrous iron and phosphorus is so strong that iron must be considered the key factor in this process, at least where it is reasonably abundant. As oxygen drops to lower levels, the oxidized microzone becomes thinner and may eventually disappear, greatly reducing the ability of the sediment to remove phosphorus from the overlying water and allowing dissolved phosphorus to diffuse from the sediment in large quantities. Once in the water of the hypolimnion, some of this phosphorus may be transported into the trophogenic zone and result in increased primary productivity. Also, incrganic phosphorus levels will be higher at overturn, at least temporarily, and will likewise stimulate primary productivity. This is especially true if redox levels during stratification fall low enough to cause large scale reduction of sulfate to sulfide, which readily precipitates ferrous iron. After reoxygenation, this results in ferric iron levels too low to remove much of the phosphorus, at least initially. 21 Several investigations with sediments from a variety of locations 11“ (I) V. 3.8. m 6 have demonstrated a high degree of uniformity in the pH dependence of adsorption and desorption of phosphorus by sediments. Macpherson gt al. (1958) tested sediments and their ashes from eight lakes of widely varying productivity and, though the absolute rates of phosphorus release and uptake by these sediments varied as to lake type, they all showed the same general trends with pH. Maximum adsorption (minimum desorption) occurred at a slightly acidic (5-6) pH, adsorption decreasing at lower (3-5) and slightly higher (6-8) pH values and increasing again at somewhat basic (8-9) pH values. The same trends were observed by Gumerman (1970) for sediments from Lakes Superior and Erie, while Jitt (1959) only observed maximum adsorption at pH 5 and declining adsorption on either side of this value, without an upturn in adsorption at pH values greater than eight (apparently due to a washing out of the sediment of all but insoluble silt particles). Except for the adsorption at higher pH, the trends described above are remarkably close to those for ferric hydroxide and magnesium and aluminum silicates (Macpherson 23 al., 1958), suggesting that these or similar substances are responsible for orthophosphate removal by sediments at pH values below eight. The pH dependence of these substances is most likely due to their amphoteric nature, with the presence of sites suitable for phosphorus adsorption depending on the exchange of hydrogen and hydroxide ions with the substance (Stumm and Morgan, 1970). Also, the maximum adsorption at slightly acidic pH values corresponds well with the maximum mole fraction of dihydrogen phosphate (HZPou‘), a form that might be more amenable to adsorption to the minerals in sediment. Above pH 8, the increased adsorption is consistent with precipitation or coprecipitation of phosphorus with calcium (Stumm and Morgan, 1970). VE Nu: h’e DC He 7 The above studies dealt with sediments shaken with distilled water to which only potassium dihydrogen phosphate, hydrochloric acid, and sodium hydroxide were added for pH and phosphate concentration adjustments. Consequently, they have limited applicability to real situations and give no indication of the pH dependence of phosphorus-containing precipitates formed in the water overlying sediments. The number of protons on a phosphate ion is certainly pH-dependent and should have a marked influence on the precipitation and coprecipitation of the species mentioned above. Additionally, the precipitation of metal hydroxides and calcium carbonate, with which phosphate is coprecipitated, is definitely affected by pH. Otsuki and Wetzel (1972) showed the coprecipitation of phosphorus on calcium carbonate to be strongly pH-dependent, being greater at higher pH values, mostly due to increased calcium carbonate formation. Dissolved Organic Matter Concentrations of ionic iron and other metals are often found in water far exceeding equilibrium levels, which has led to the belief that they form complexes or chelates with dissolved or colloidal organic matter. Shapiro (1966) has shown ferric iron to form a colloidal association with naturally-occurring aliphatic, polyhydroxy, carboxylic acids (MW 200—400), possibly by chelation or complexation, but more likely by peptization. Regardless of the exact forms, this iron will remain in true or colloidal solution instead of being precipitated out, so that, especially if phosphate is bound to it (Golterman, 1967), phosphorus removal to the sediments is inhibited. Further evidence of the role of organic acids comes from the work of Harrison gt El- (1972), who demonstrated the solubilization of'FePOu, Ca3(P04)2, CaHPOQ, A12(P04)2, and Mg3(P04)2 by certain bacterial isolates 8 from lake sediments. They attributed this to chelation by organic acids secreted by the bacteria and showed that such acids were extremely effective in solubilizing these phosphate salts under sterile conditions. Sediment Properties As discussed above, the presence of ferric hydroxide in the surface layer of sediments has a pronounced effect on the ability of sediments to remove and retain phosphorus. In general, iron and aluminum hydroxides, oxides, and silicates are quite effective in adsorbing orthophosphate, so that the presence of any of these minerals would improve the ability of sediments to remove phosphorus. Work by Jitts (1959), Gorham and Swaine (1965), Livingstone and Boykin (1962), Li _e_t 3;. (1972), and Williams gt p1. (1971a, 1971b, 1971c) supports this statement. The effect of calcium carbonate in sediments appears less certain. Li 23.51- (1972) report that the calcium fraction of sediments does not appear to be related to the phosphorus exchange capacity, at least over a short time span, which is consistent with Otsuki and Wetzel (1972), who concluded that coprecipitation of phosphorus with calcium carbonate relies more on mechanical inclusion or the formation of a solid solution rather than strong surface adsorption. Zicker gt al.(1958), though, present data suggesting that calcium carbonate in sediments increases their ability to remove and retain orthophosphate, presumable due to surface adsorption, which had been demonstrated by Cole p_t_ 9;.(1953). Organic matter in sediments, though it represents a significant portion of the phosphorus removed into and retained by sediments, appears to have a negative impact on the ability of sediment to remove orthophosphate from water. Jitts (1959) demonstrated a decrease in phosphorus removal by sediments in proportion to their organic content. 9 He also found sediment particle size to increase with organic matter. He suggested that organic matter either competes with phosphorus for available sorption sites or physically masks the sites, in part by cementing sediment particles together. Work by Macpherson gt g}. (1958) also demonstrated an interference by organic matter with phOSphorus removal by sediments in that ashed sediments generally removed more phosphorus from water than whole sediments, even though ashing would have released additional inorganic phosphorus. Mpltivalent Cations As the above discussions suggest, multivalent cations (Fe+3, Mnfu, Caf+, Mg++, Al+3) are quite important in the exchange of orthophosphate between dissolved and solid states. It should then be expected that additions of such cations would enhance phosphorus removal, a principle that is recognized in lake rehabilitation and in waste treatment with the use of such substances as ferric chloride, alum, aluminate, and lime (Rohlich and Uttormark, 1972). Conversely, a lowering of cation levels may have a negative impact on phosphorus removal, even if, as in marl formation, some phosphorus is removed during the removal of the cations. Overall, it is not unreasonable to expect soft water lakes to tolerate less phosphorus loading than hard water lakes before significant productivity increases occur, assuming phosphorus is limiting production. Temperature Egg Turbulence Rates and extents of‘reactions are well known to depend on temperature, and reactions involving phosphorus exchange should not be exceptions to this, but studies regarding this are few. Gumerman (1970) found greater equilibrium concentrations of phosphorus at 22°C than at 4°C in sediments from Lakes Superior and Erie. Otsuki and Wetzel (1972) also found 10 increased temperature to increase phosphorus coprecipitation with calcium carbonate at pH values less than 9.5. This was attributed to increased calcium carbonate fermation rather than increased phosphorus inclusion in a given amount of precipitate. Due to the slow rates of diffusion in sediments, stirring of reduced sediments should increase phosphorus export. Also, in oxidized sediments, disruption of the oxidized microzone should allow at least a temporary loss of phosphorus and reduced materials from newly exposed reduced sediment layers before they are oxidized. Zicker 22.§l- (1958) found phosphorus release from bog lake muds to double with 'vigorous stirring.‘ His results, though, shed little light on the effect of natural turbulence on the release of phosphorus from lake sediments, especially in the littoral zone, where turbulence is highest and the sediments are in close proximity to primary producers that can, perhaps, use released phosphorus before it is returned to the sediments. INDIRECT EFFECTS OF AQUATIC MACROPHYTES ON PHOSPHORUS DYNAMICS Due to metabolic activities and, sometimes, just to their physical presence, aquatic macrophytes have been shown to affect the chemical and physical factors discussed above. Because of this, it can be inferred that aquatic macrophytes will exert some influence on the dynamics of inorganic phosphorus indirectly by their effect on such things as dissolved oxygen, pH, dissolved organic matter, sediment characteristics, multivalent cations, temperature, and water circulation. Redox Potential The oxygen produced by aquatic macrophytes during photosynthesis is, in the long run, nearly or completely balanced by oxygen consumed during respiration and decomposition. Because these processes are separated in 11 time and space, aquatic macrophytes can either cause oxygen depletion or accretion, depending on time and location, and thus either raise or lower the redox potential. More often than not, the depletion of oxygen by macrophytes will be more critical with respect to phosphorus dynamics since (1) the redox potential and its effects on inorganic phosphorus exchange only change markedly at very low oxygen levels and (2) macrophyte production of oxygen usually occurs in places where oxygen levels are not critically low, while consumption of oxygen resulting from macrophytes is often in areas already partly depleted of oxygen. Severe oxygen depletions (to less than 0.5 mg/l 02) have been reported in a stand of the emergent Glyceria aquatics (Dvorak, 1970), at the bottom of a stand of Elodea canadensis (Buscemi, 1958), at sediment level under a mat of Ceratophyllum demersum.(Goulder, 1969), and under a mat ofngphgg sp. and Hydrocharis sp. (StrasI-craba, 1965). In the last three of these studies, oxygen was found to be at or over saturation in the upper part of the plant beds, which was at most two meters above the anoxic zone. In Dvorak's study, oxygen was uniformly low throughoutthe plant bed, climbing markedly at the margins to levels of 6 mg/l 02 just outside the bed. In extreme situations, the decomposition of heavy growths of macrophytes can deplete oxygen in an entire pond or small lake, not only under ice cover, but also during calm, late summer days. Additionally, macrophytes can contribute to hypolimnetic oxygen depletion when decomposing macrophytes are are exported from the littoral to the profundal zone. The overall contribution of macrophytes to consumption of oxygen in hypolimnia is not well established, but it may be quite significant in small lakes, due to relatively larger littoral zones and to the fact that much of the decomposition of phytoplankton occurs in upper waters, while the more 12 resistant tissue of vascular plants would decompose to a greater extent at sediment level. 2! Associated with oxygen production and consumption are, respectively, a raising and lowering of pH, often of enough magnitude to significantly affect inorganic phosphorus dynamics. The data of'Dvorak (1970) on pH parallel his oxygen data, with pH dropping from approximately 8 just outside the bed of'Glyceria aquatics to about 6 within most of the bed. Straskraba (1965) observed a pH drop from 8.2 at the top to 7.0 at the bottom of a one meter water column under Naphgr sp. and Hydrocharis sp. Goulder (1969) found pH values to exceed 10 in a Ceratgphyllum demersum mat, with values dropping to below 8 just outside the mat. The pH values observed by Goulder are certainly high enough to result in marl precipitation if hardness and alkalinity are sufficiently high; in fact, he did observe drOps in alkalinity in the gquemersum beds, indicative of carbonate precipitation. In extreme cases, the elevation of pH by macrophyte photosynthesis can cause massive encrustation of marl on the plants and the deposition of carbonate rich sediments, presumable with significant phosphorus removal by COprecipitation. Dissolved Organic Matter Significant excretion of dissolved organic matter is known to occur in some aquatic macrophytes (Wetzel, 1969), but as yet the released compounds have not been characterized, so that the quantities of cation- chelating or peptizing organic acids released by these plants are unknown. In addition to this excretion while alive, macrophytes contribute dissolved organic matter to the water upon death and decomposition (Nichols and Keeney, 1973), both by release of cell contents and by bacterial 13 degradation of their tissues. Particularly important is the fact that the humic acids associated with iron in the work by Shapiro (1966) are, in general, products of microbial action on resistant types of organic matter, associated more with vascular plants than algae. Therefore, aquatic macrophytes may contribute to the dissolved organic matter associated with cation solubilization in amounts out of prOportion to their contribution to total production of organic matter in a basin. Sediment Properties Aquatic macrophytes can have a marked effect on both the mineral and organic content of sediments. Marl deposition, in large part due to macrophytes, has an impact on the nature of sediments and their phosphorus relations, as discussed above. In extreme cases, littoral sediments can be composed almost entirely of marl as a result of maor0phyte activity. The large, stable biomasses and resistant vascular tissue produced by macrophytes (as compared to phytoplankton) can cause significant increases in the organic content of sediments, not only in the plant beds, but also in other portions of a lake due to export of plant material by currents and turbulence. Ulehlova (1970) observed wide variations in the organic content of sediments (from less than 10% to greater than 90%) that he related to the types and abundance of macrophytes above the sediments. Multivalent Cations Due to high availability of calcium, magnesium, manganese, aluminum, and, usually, iron in relation to the physiological requirements of macrophytes for these elements (Boyd, 1969, 1970), aquatic macrophytes have relatively little direct impact on the levels of cations of importance to inorganic phosphorus exchange reactions. A major exception to this is the removal of massive amounts of cations by Sphagnum spp. mats, resulting in 14 the typically soft waters of bogs and bog lakes. Also, macrophyte— mediated production of marl can significantly reduce the concentrations of some cations. Temperature and Turbulence The physical presence and photosynthetic activity of macrophytes would logically be expected to alter light absorption and water circulation patterns, thus affecting both temperature and turbulence. Dvorak (1970) observed markedly lower temperatures in a bed of'Glyceria aquatica, presumably due to shading. Also, the low oxygen levels he reported are indicative of little surface turbulence and mixing. Straskraba (1965) found a temperature differential of 9.800 in a one-meter water column under Npphg; sp. and_Hydrocharis sp., indicating effects on both temperature and circulation, Goulder (1969) also observed marked stratification of chemical parameters in water too shallow (3 m) to support such stratification without the stabilizing effect of a‘g; demersum mat. DIRECT EFFECTS OF AQUATIC MACROPHYTES ON PHOSPHORUS DYNAMICS Phosphorus uptake by aquatic macrOphytes can occur from the water or sediments, or both (Sculthorpe, 1967). The impact of this uptake on water and sediment phosphorus levels can be great due to (1) large accumulations of biomass, (2) high physiological requirements for phosphorus in relation to its availability, and (3) assimilation of phosphorus in excess of immediate requirements, or so-called 'luxury consumption' (Boyd, 1969, 1970; Gerloff, 1969; Gossett and Norris, 1971; Kosek, 1971; McNabb, Tierney, and Kosek, 1972; Tierney, 1972; Wilson, 1970). Obviously, when phosPhorus is removed from the water by macrOphytes, it is unavailable for other primary producers and for removal into the sediment. Retention of phosphorus in macrophytes is not permanent, though, Old 9 LI 15 and all phosphorus taken up by macrophytes will eventually be released to the water in excreted matter and products of decomposition or be deposited in the sediments in an organic form. Nichols and Keeney (1973) demonstrated that, under oxidized conditions, phosphorus released from decomposing Myriophyllum exalbescens in laboratory sediment/water systems is rapidly removed into the sediment, though some is temporarily available for uptake by other primary producers. Uptake of phosphorus from water and its subsequent release by macrophytes will therefore markedly alter the distribution of phosphorus in time and space, though what total effect this has on primary productivity, removal of phosphorus to sediments, and the flux of phosphorus through aquatic systems is largely unknown. Uptake of phosphorus from sediments by aquatic macrophytes is, perhaps, of greater significance than uptake from water in that it constitutes a major reversal of the general tendency of phosphorus to be removed into the sediment under oxidized conditions. Besides providing nutrients necessary for growth of rooted macrophytes, this uptake can result in greater production by free-floating macrophytes and algae when the phosphorus is excreted into the water or released upon decomposition. In fact, certain estuarine macrophytes have been shown to, in effect, pump inorganic phosphorus from the sediments to the water (McRoy and Barsdate, 1970; MdRoy‘gp‘al., 1972; Reimold, 1972). The net phosPhorus pumped from sediment to water can even exceed the amount assimilated by the plants and thus has a marked influence on the phosphorus flux through the estuary and on the phosphorus available to other primary producers. In fresh-water, budgets of phosphorus for littoral and pelagic zones have suggested that export of phosphorus from littoral areas, presumably due in large part to macrophyte activity, helps maintain phytoplankton production in the pelagic waters 16 for a significant portion of the growing season (Wetzel, 1975). STATEMENT OF PROBLEM AND PLAN OF RESEARCH The following major points have thus far been made: (1) The exchange of inorganic phosphorus between solid and dissolved states is critical to the productivity of many aquatic systems. (2) Inorganic phosphorus exists as dissolved orthophosphate and a variety of solid forms, the relative importance of which are poorly known. (3) The removal of inorganic phosphorus from water and its deposition.and retention in sediments depend on a variety of chemical and physical factors, including redox potential, pH, dissolved organic matter, multivalent cations, sediment properties, temperature, and turbulence. (4) Aquatic macr0phytes can significantly alter the above factors and therefore can presumably affect phosphorus dynamics indirectly. (5) Aquatic macrophytes can directly affect phosphorus dynamics through uptake, retention, release, and deposition of phosphorus. This summary suggests an important and complex set of interactions between aquatic macrophytes and phosphorus distribution, but the present level of knowledge is insufficient to allow integration of all the discussed factors into a realistic scheme that will describe the overall effects of macrophytes on phosphorus cycling. In addition, few studies have even established empirically the net effect of a macrophyte on phosphorus, even under simple, restricted circumstances. The work to be reported here therefore attempted to (1) establish the integrated effect of the submerged angiosperm, Ceratophyllum demersum, on the distribution of phosphorus, through time, in small, controllable laboratory systems and (2) to determine the major mechanisms affecting phosphorus dynamics in these systems and how 9; demersum affects them. 17 lg; demersum was selected as the experimental plant because (1) it is known to significantly alter such water parameters as pH and dissolved oxygen, (2) it is an important component of many small lakes and ponds, (3) it is generally free floating in lab culture and therefore does not obtain an appreciable amount of nutrients directly from sediments, which makes analysis of phosphorus dynamics easier, and (4) it is amenable to laboratory rearing. With regard to the above goals, two experiments were devised. The first experiment (Aquarium experiment) consisted of monitoring phosphorus and other relevant factors in water/sediment systems established in twenty— gallon aquaria with metered flow-through of water. Treatments consisted of (1) the presence and absence of plants and (2) twelve- and fourteen- hour photoperiods, with three replicates each. The different photoperiods resulted in declining and increasing plant biomass, respectively, that were at least somewhat representative of stages in the annual cycle of this plant. The second experiment (Beaker Experiment) involved the determination of the rate of orthophosphate removal from water in one-liter beakers over twenty-feur hours. Treatments included (1) three levels of initial orthophosphate concentration, (2) three levels of initial hardness, (3) three levels of initial alkalinity, (a) five levels of initial pH, and (5) the presence and absence of sediment, in a factorial design with one replicate of most of the treatment combinations. From this experiment, a series of’regression equations for phosPhorus, alkalinity, and hardness removal were developed and used to simulate the behavior of these parameters in the aquarium experiment. Such a simulation allowed some conclusions to be drawn regarding how and why Ceratophyllum demersum affects the dynamics of phosphorus in these aquarium systems. V ~ fici‘ MATERIALS AND METHODS AQUARIUM EXPERIMENT Experimental Apparatus Experiments were conducted in a row of six 20—gallon aquaria (numbered 1 to 6) under a bank of twenty—four 40—watt Gro—lux fluorescent lights. At water level, light intensity varied from approximately 550 foot-candles in the end aquaria (1 and 6) to 650 foot-candles in the middle aquaria (3 and 4). Because of this gradient of light, aquaria were paired when treatments were assigned (i.e., if aquarium 5 received plants, aquarium 2 did not). Photoperiod was controlled with a timer switch through which all lights were circuited. For the initial experimental run (12-hr photoperiod), flow through the aquaria was controlled by a gravity/siphon system. Influent was delivered through glass and vinyl tubing from a head tank in which a constant water level was maintained by continual pumping of water into the tank from a reservoir and overflow into a standpipe which returned water to the reservoir. Influent to each aquarium was restricted to approximately 100 ml/hr by a screw clamp at the exit of each delivery tube. Effluent from each aquarium was maintained by a siphon which delivered overflow to a two—gallon sampling bottle. An additional sampling bottle received water directly from the head tank for influent water chemistry determinations. Circulation in the aquaria resulted not only from flows caused by these influents and effluents, but also from air forced over the 19 aquaria at 5 to 10 fps by a large fan. This gravity influent system proved to be incapable of delivering sufficiently constant flows, resulting in poor estimates of influent volumes and in quite different volumes delivered to each aquarium. Therefore, for the second experimental run (14—hr photoperiod), this system was replaced with a variable-flow, seven—channel tubing pump, which delivered a constant and equal flow directly from the floor reservoir to each aquarium and to the influent sampling bottle, which then served for determination of influent volume as well as water chemistry. Plants, Water, and Sediment A large quantity of Ceratophyllum demersum was obtained from the waste stabilization ponds at Belding, Michigan on June 21, 1974. A portion of this was cleaned of filamentous algae and placed in two 2' x 8' x 1' holding tanks through which water was continually circulated, overflow draining into a reservoir in which it was aerated andpumped back to the tanks. The water in these tanks was from the same source as that used in the experimental runs in the aquaria and was renewed periodically. Light conditions were also made as near as possible to experimental conditions. All of the water used in the holding tanks and the aquaria was obtained from Pond 2 of the Michigan State University Water Quality Management Project, with the exception of the initial filling of tanks and aquaria for the first experimental run, when well-water from the Fisheries Research Laboratory was used in part (ca. 30%). Results of chemical analyses of this pond for the duration of the experiments were obtained from C. Annett of the Institute of Water Research (Table 1). The sediment used in all experiments was a dried, air-filtered, native kaolinite clay, obtained through a ceramics supplier. This clay was found (I) 20 Table 1. Summary of water chemistry of pond 2 of the Michigan.State University Water Quality Management Project for the summer of 1974. Date Alkalinity (mg/l Ca003) Hardness (mg/l CaC03) Dissolved Oxygen (mg/l 02) pH Calcium (mg/l Ca) Magnesium (mg/l Mg) Total Iron (mg/l Fe) Potassium (mg/l K) Sodium (mg/l Na) Chloride (mg/l Cl) Sulfate (mg/l sou) Ammonia (mg/l N) Nitrate (mg/l N) Nitrite (mg/l N) Total Phosphorus (mg/l P) Soluble Phosphorus (mg/l P) 6/17/74 175 273 10.6 8.75 110 17 1.0 5.9 79 95 76 0.26 0.32 0.09 0.72 0.60 7/8/74 186 234 8.5 8.20 93 18 1.1 6.4 76 100 74 0.31 0.32 0.07 0.70 0.69 8/1/74 197 10.1 0.28 0.23 8/21/74 173 0.14 0.04 0.01 0.38 0-37 nnr‘ u‘Jn‘ . Ge 5.. Ac ‘3 v. ”I U. s 21 to contain less than 0.5% organic matter, as determined by ashing oven— dried clay at 350°C for four hours. Testing of the settling of this clay in water showed that it formed a non-compacted sediment with a water content of approximately 42%, by weight. Experimental Procedure A mixture of 5000 g of dry clay and 4000 ml of water was sieved through a twenty meshrscreen and spread evenly on the bottom of each aquarium. This was allowed to settle at least twelve hours, resulting in about 200 ml of clear water overlying the sediment and in no size stratification of the clay as would occur if a greater volume of water had been used. Measured volumes of water were then siphoned or pumped into each aquarium as fast as possible without appreciably disturbing the sediment until the depth was sufficient for the effluent siphon to hold water. A large quantity of plants was then removed from the holding tanks, drained, and blotted to constant weight. A portion of these plants (400 g wet-weight for the initial run and 200 g wet-weight for the second run) was weighed and placed in a randomly selected member of each pair of aquaria. A fourth portion (ca. 100 g wet—weight) was also weighed and oven-dried (70°C for at least 72 hr) for dry-weight determination and phosphorus analysis. After the addition of plants, timed to occur at 'midday', water was added to each aquarium until the effluent siphons began to drip. A flow of approximately 100 ml/hr was then started in the influent system. Finally, temperature, pH, dissolved oxygen, hardness, alkalinity, total phosphorus, dissolved phosphorus (second run only), and orthophosphorus . (second run only) determinations were made on the water in each aquarium or on the stock of water added to the aquaria. Periodically throughout the runs, usually every two days, measurements 22 of pH, temperature, and dissolved oxygen were made in the aquaria and the reservoir at midday. At the same time, effluent and influent bottles were emptied and determinations of volume, alkalinity, hardness, total phosphorus, dissolved phosphorus (second run only), and orthophosphorus (second run only) were made. In the initial run only, influent rates were measured and adjusted daily. In the second run only, plant wet-weights were measured and part of the biomass was harvested for dry weight determinations and phosphorus analysis on day seven. Also in the second run only, influents and effluents were terminated on day twelve to allow independent estimates of evaporation rates from the aquaria. Additionally, samples were taken over a twenty-four hour period once during each experimental run. Both experimental runs were terminated on the fifteenth day, at which time determinations of hardness, alkalinity, and forms of phOSphorus were made on water drawn from each aquarium in additon to those determinations described above for intermediate days. Plants were removed, rinsed, drained and blotted, weighed, and oven-dried for dry-weight measurement and phosphorus analysis. Volumes within the aquaria were then determined and, finally, sediment samples were taken for moisture content and phosphorus analysis. BEAKER EXPERIMENT Treatments Treatments consisted of the presence or absence of sediments in the experimental beakers and of adjustments to pH, alkalinity, hardness, and phosphorus concentration in the water added to the beakers. This water was prepared from deionized distilled water by the addition of various compounds (Table 2) in amounts determined by the desired treatment levels (Table 2), the analysis of Pond 2 water (Table 1), and the results of the 23 Table 2. Water chemistry treatment levels for beaker experiment. I. Initial Alkalinity ‘nggl Alkalinity Compounds and concentrations used 1 ca. 0 Sufficient NaOH and HCl for pH adjustment 2 1.8 meq/l 0.2 meq HCl; 2.0 meq NaOH + NaHC03/l 3 3.6 meq/l 0.4 meq HCl; 4.0 meq NaOH + NaHCOB/l II. Initialng L213; .Pfi Ratios of normalities of HCl : NaHCO3 : NaOH 1 7.7 - 8.1 1 : 10 : 0 2 8.3 - 8.8 1 : 9 : 1 3 8.8 - 9.3 1 : 8 z 2 4 9.3 — 9.7 1 : 7 : 3 5 9.5 - 10.0 1 : 6 : 4 III. Initial Hardness Level Hardness Compounds and concentrations used 1 0.0 meq/l 5.0 meq NaCl/l; 1.6 meq Nazsou/l 2 3.1 meq/l 2.5 meq CaCl?/l; 0.8 meq MgSOu/l; 2.5 meq NaCl 1; 0.8 meq Nazsou/l; 0.27 meq Fe(N03)3/l 3 6.2 meq/l 5.0 meq CaCl /1; 1.6 meq MgSOu/l; 0.54 meq Fe(NO3)3/l IV. Initial Orthophosphorus Level (P053) Compound and concentration used 1 0.1 mg P/l 0.4393 mg 1012li 2 0.5 mg P/l 2.1966 mg KHZPOu/l 3 1.0 mg P/l 4.3932 mg KHzpou/l (D H11 (1’) $.. 5.1. g + 3.1 24 aquarium experiment (see Results and Discussion). Proportions of calcium, magnesium, and ferric iron were not varied independently due to the lack of data on their individual levels in the aquarium experiments. Sulfate concentration, ionic strength, and dissolved oxygen were kept relatively constant, while sodium, chloride, potassium, and nitrate were varied as required by the levels of the other species and were assumed unimportant to the phosphorus exchange reactions. Combinations of the above water chemistry treatments were run with and without the presence of sediments in order to evaluate what portion of orthophosphate removal was due to precipitation or coprecipitation with substances in the water and what portion was due to adsorption by the sediment. Factorially arranged, these treatments comprise 270 combinations, of which only 165 were run, the omitted ones being almost entirely those with low alkalinity and hardness values. Also, 25 of the combinations were repeated to provide better estimates of variation. Experimental Procedure Eight series of experiments were conducted in a set of twenty-four 1000 ml beakers, to which treatment combinations were randomly assigned. Sediment was prepared as in the aquarium experiment, except that deionized distilled water was used, and 50 ml of this sediment were added to those beakers selected for treatment combinations calling for sediment. Appropriate amounts of phosphate and acid stock solutions were added to deionized distilled water which had been vigorously aerated for at least two hours and 860 ml of this solution were transferred to each beaker, taking care not to disturb the sediments. Other stock solutions were then added to each beaker in amounts appropriate for the water chemistry required in each beaker and for a final volume of 900 ml. Finally, pH and 25 temperature were measured in each beaker and all beakers were covered with watch glasses and placed in a shaded area. Initial values for hardness, alkalinity, and orthophosphorus were measured on duplicate beakers for each combinations of water chemistry treatments. After twenty-four hours, pH and temperature were again measured in each beaker and a water sample was with drawn from each beaker for determination of hardness, alkalinity, and orthophosphate concentration. RESULTS AND DISCUSSION AQUARIUM EXPERIMENT Plant Growth and Phosphorus Content Photoperiod had a marked influence on the growth of Ceratophyllum demersum during the experimental runs (Table 3), the shorter photoperiod resulting in a decline of 30% to 40% in dry weight, as opposed to an increase of 40% to 50% under the longer photoperiod. Growth of plants in the holding tanks showed a similar difference in rates due to photoperiod and further demonstrated that the density of the plants did not contribute to this difference. The slowdown of growth in the later part of the second experimental run (Table 3) was apparently due to nutrient limitation. Changes in plant phosphorus content were also quite different between experimental runs (Table 3), largely due to the different growth rates, but also attributable to differences in plant condition and ambient phosphorus concentration. During the first experimental run, phosphorus equal to as much as 80% of that originally in the plants was released due to declining biomass and tissue concentration, the latter probably a result of both poor condition of the plants and lower concentrations of phosphorus in the aquaria than in the holding tanks. On the other hand, the positive growth rates during the second experimental run resulted in increases in phosphorus contained in the plants, even though moderate decreases in tissue phosphorus concentration occurred due to declining ambient phosphorus levels. As discussed below, this uptake and release of phosphorus by'gL demersum had Table 3. Summary of plant weights and phosphorus contents in aquarium experiment. 27 First Experimental.Run (Aquaria 1, 4, 5 with plants, 12-hr photoperiod) Aquarium Wet Weight (g) - Day 0 Day 15 Dry Weight (g) - Day 0 Day 15 Phosphorus (mg) - Day 0 Day 15 Second Experimental.Run (Aquaria 2, 4, 6 with plants, 14-hr photoperiod) Aquarium Wet weight (g) — Day 0 Day 7 Preharvest Postharvest Day 15 Dry Weight (g) - Day 0 Day 7 Preharvest Postharvest Day 15 Phosphorus (mg) - Day 0 Day 7 Preharvest Postharvest Day 15 1 400.0 321.2 21.25 13.20 202.8 80.8 2 200.0 264.4 200.0 236.4 16.14 19.95 15.09 17.26 72.45 86.27 65.26 63.88 4 400.0 316.3 21.25 14.34 202.8 98.0 L; 200.0 294.2 200.0 236.4 16.14 22.17 15.07 17.81 72.45 102.51 69.69 65.64 5 400.0 329.5 21.25 12.17 202.8 78-5 6 200.0 273.8 200.0 203.8 16.14 20.11 14.69 16.17 72.45 87.93 64.23 68.06 28 a great impact on the phosphorus levels in sediment and water during both experimental runs. Water Budget Volumes of water in the aquaria at the start and end of each run and volumes of influents, effluents, and evaporation are summarized in Table 4. For the second experimental run, all quantities were measured directly except evaporation, which was calculated by difference and was found to be significantly lower in aquaria with plants. Conversion of light energy to chemical energy in photosynthesis bng; demersum would account for an insignificant portion of the observed differences in evaporation, so the physical presence of the plants is presumed here to be the cause of these differences, most likely through effects on water circulation and surface turbulence. The latter was much reduced in the aquaria with plants. Since evaporation was found to be lower in aquaria with the macrophyte, effluent volumes from those aquaria were necessarily greater than in aquaria without plants. Together, these differences would obviously result in proportionately greater export of materials in aquaria with plants and greater accumulation of materials in aquaria without plants, with consequences on phosphorus distribution that will be evident below. For the initial experimental run, influent volumes as well as evaporation were unknown due to the unreliable delivery system. Because of this, the evaporation rates from the second run were applied to the appropriate aquaria in the first run and influent rates were then computed by difference. This is a major assumption with potentially a significant impact on the conclusions drawn from the experimental results. Because of this, data analysis on the first experimental run was repeated with a wide range of evaporation rates to insure that any trends observed were real and Table 4. Volumes (liter) of aquaria, influents, effluents, and evaporation in aquarium experiment. 29 First Experimental.Run.(Aquaria 1, 4, 5 with plants, 12-hr photoperiod) Aquarium Initial Aquarium Volume Influent Volume Evaporation Effluent Volume Final Aquarium Volume 1 49.4 46.1 11.6 34-5 49.5 2 50.2 48.6 14.2 34.5 50.2 3 50.6 50.2 13.8 36.4 50.6 4 48.5 42.0 10.3 31.7 48.6 5 48.9 41.7 9.7 32.0 49.0 6 51.9 38.7 13.5 25.2 51-9 Second Experimental Run.(Aquaria 2, 4, 6 with plants, 14-hr photoperiod) Aquarium Initial.Aquarium Volume Influent Volume Evaporationa Effluent Volume Final Aquarium Volume 1 51.0 28.9 13.5 17.1 49.3 2 50.8 28.4 9.7 20.3 49.1 3 51.0 28.2 13.8 16.6 48.8 4 50.8 29.0 10.3 21.8 47.7 5 51.0 29.1 14.2 16.7 49.2 6 50.8 28.2 11.6 19.6 47.8 a Mean effect of plants on evaporation = -0.22 liter/day; significant at the 0.04 level, based on a paired t-test. 30 not artifacts of this assumption. Twenty7four Hour Sampling Representative values of samples taken over a twenty—four hour period during the first experimental run are listed in Table 5. Temperature, pH, and dissolved oxygen in the aquaria showed definite diel cycles, the magnitude of which was similar in aquaria with and without plants, so that any conclusions drawn on data at midday were valid for the entire day. Interestingly, values of these parameters taken at midday are quite close to the means over the twenty-four hour period, a fact which was of significance in the simulations discussed later. Alkalinity, hardness, and phosphorus in the effluent did not demonstrate any significant differences between light and dark periods, but this does not preclude cycling that might cancel itself out within these periods, as approximately occurs in the cases of pH and dissolved oxygen. However, additional measurements of orthophosphorus in the aquaria during the second experimental run did show that there was no significant deviation of orthophosphorus concentration over a light/dark cycle from the general trends observed in the run. Temperature, Dissolved Oxygen, and pH Midday values for temperature, dissolved oxygen, and pH are plotted for all aquaria in both experimental runs on Figures 1, 2, and 3, respectively. Mean values of these and other parameters over the duration of each.run are listed in Table 6, along with the mean effects of plants on each parameter and the significance levels of these effects, based on a paired t—test. Of these parameters, temperature is the only one to fail to show any significant effects due to plants on any of the days or in the means. Consistent differences in temperature were only found 31 Table 5. Results of twenty-four hour sampling during first run of aquarium experiment (day 14 to day 15). Aquarium 1 2 3 4 5 6 pH — Midday 8.64 8.79 8.60 9.00 9.30 8.92 Lights Out 9.08 9.04 8.81 9.25 9.45 9.13 Midnight 8.55 8.75 8.51 8.83 9.09 9.06 Lights On 8.47 8.60 8.40 8.75 8.81 8.68 Midday 8.60 8.74 8.51 8.91 9.15 8.81 Dissolved — Midday 10.2 8.0 8.5 12.6 13.3 10.4 Oxygen (mg/l 02) Lights Out 17.5 13.4 10.6 16.9 18.6 14.5 Midnight 9.2 10.7 8.6 10.3 13.4 11.8 Lights On 6.8 8.1 6.6 7.0 9.8 8.6 Midday 10.4 10.7 8.5 12.1 11.8 11.0 (38 rature - Midday 22.8 22.8 22.8 23.0 23.0 22.8 Lights Out 23.4 23.7 23.8 24.0 23.7 23.4 Midnight 22.7 22.8 22.8 22.7 22.7 22.6 Lights On 22.6 22.6 22.7 22.6 22.6 22.5 Midday 23.4 23.6 23.7 23.7 23.6 23.4 Hardness - Light Period 280 162 188 282 269 167 (mg/l CaCOB) Dark Period 279 163 185 278 266 166 Alkalinity - Light Period 293 180 205 292 280 182 (mg/l Ca003) Dark Period 294 182 206 293 281 183 Total — Light Period .912 .058 .070 .723 .511 .055 Phosphorus (mg/1 P) Dark Period .858 .082 .087 .680 .481 .060 .pmmpup 000000 0 no women .poommm woman mo He>mH 00:000000000 mnma< 0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 mao>mn .2000 000.0 000.0 000.0 000.0 000.0 000.0 sam>oq 00000 000.0: 000.0: 00N.01 00- 00- 00.m+ 00.0+ 0.0+ mvomam 00 000000 200: 000.0+ 00+ 00+ 00.0- No.o+ 0.0 mpcsam mo 060000 seas 000.0 000.0 000.0 000 000 00.00 00.0 0.00 000.0 000 000 00.00 00.0 0.00 000.0 000.0 000.0 000 000 00.00 00.0 0.00 0 000.0 000.0 000.0 000 000 00.00 00.0 0.00 0 000.0 000.0 000.0 000 000 00.00 00.0 0.00 0 000.0 000.0 000.0 000 000 00.00 00.0 0.00 N 000.0 Am H\0sv monogamoraospso 000.0 Am H\0sv monogamonm 00>Homm00 000.0 Am H\0ev monogamonm 00009 000 000000 0\0s0 00000000 000 000000 0\0sv 0000000000 00.00 000 0\0sv semaxo 00>0omm00 00000: 00.0 00 00000: 0.00 Auov myopsnomsoa_000002 0 55000504 000000000020 00:00 .000000 :00: 0 .0 .0 00nssd0oma00 000002 00.0 00 00000: 0.00 Aoov auspaaoasae.0mss0= 0 ssHHdsd< AcoHHomoporm 0:100 .mpsmam £003 0 .0 .0 000050 use: .0 00909 33 24- 24— Ii 25 4 6 s\ 36\4 23_ 5’ \\5§ 4. 23— /z \2 IS %26 ?\I Ie/6\I532 8 2'2H2- 122?- In x 0 y. < c: E I ’ :2” 3 2I — 2 I /2/| l/\ ZOI— 20" 25 I \I 3/ .l ISL— I9I— J l L l l I l 1 I J O 3 . 6 9 I2 l5 0 3 6 9 l2 l5 TIME (days) Figure 1. Midday temperature vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). 4\ l8- l8- 6 4 2-——2s 4 I5 l5- .b DISSOLVED oxvceu (mg/I 02) I“ ”—5 l3 3 6 6— l l I l 1 O 3 6 9 l2 I5 0 3 TIME (days) OH- .9.— 'n' 6 Figure 2. Midday dissolved oxygen concentration vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). 35 4 I06 Io.oI- 4 6 2 9.5 9.5— 4 4/ 26 5 4/ 6 \ 4/ / \6 2 2 5 axes 2 4 35‘s 9.0 / 9.0f— \3 z 26 6\4 / ,0 2 6 g 4’// 5 ' 3 5 4/ 6N 3 I 5\ 4\45I 3 ELS Ilsr' I I \. / I '\' 236 S 35 I 8.0 8.0 _ l I I I I I J, i_J O 3 9 l2 l5 0 3 6 9 l2 I5 TIME (days) Figure 3. Midday pH vs. time for all aquaria during experimental runs one (left,)aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants . 36 between end and middle aquaria, attributable to the gradient of lighting, and between experimental runs, attributable to lower temperatures and humidities in the experimental room during the second run. Dissolved oxygen, on the other hand, showed significant effects due to g; demersum in both experimental runs. In the initial run, the effect varied through time, with aquaria containing plants showing higher dissolved oxygen levels during the first five days, lower levels from then to approximately day 12, and higher levels again after that. Because of this time dependence of the effect of the plants, mean dissolved oxygen levels did not show consistent differences between aquaria with and without plants. In the second run, however, dissolved oxygen levels in aquaria with plants were consistently higher for the entire duration of the run and had significantly higher means than the corresponding aquaria without plants. Comparing aquaria between runs, the ones without plants showed a high degree of similarity, considering the differences in initial dissolved oxygen concentration, whilethose with plants were markedly different. As would be expected since both parameters are largely a function of photosynthetic and respiratory activity, the trends of pH are quite close to those described above for dissolved oxygen and show comparable mean effects and significance levels. Hardness and Alkalinity Values for hardness and alkalinity for all aquaria in both experimental runs are plotted on Figures 4 and 5. Points at days 0 and 15 are as measured in the aquaria on those days, while intermediate points are those determined on effluent water, plotted at the midpoints of the periods over which effluent was collected. These parameters behaved quite similarly to each other and show comparable mean effects due to plants 37 300- 4~i soo- l—f>l4 4§Z§§7<5 §I /74 \5 4 I \‘\\5 + \5 I 275'- 275‘ §\l '76 6 g §\' 3 /' 5 3 2. I 3 \ 2; 5 3 a: \ C 3 ‘x‘ 226— 225- 2 3: 6 26\ 3 3\ 6 3 4 2 200- 200%,— 2\2 6 \\\‘6 6><2 4\4 ’- 6 \ 2 '75" '75— 4\6 I I L I I L I ! I j‘ 0 3 6 9 I2 l5 0 3 6 9 I2 I15 TlMEMays) Figure 4. Hardness vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). 300 r 275 - w ’5\ \ .7. 38 200 P o o O o 2230- O 5 ). 1". .2. ..I § 4225- I25 < 3 2 200* I00- 6 3 \ 2 2 \ 4 3 \ I75.L 6 73r- 4 6 2\ \ \6 4 I I I I %I I I I I I O 3 6 9 l2 l5 0 3 9 l2 l5 TIME (days) Figure 5. Alkalinity vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, u, 6 with plants ) . 39 (Table 6). In the first experimental run, aquaria with plants showed a slight increase in both hardness and alkalinity through day 10, presumably due to concentration by evaporation, followed by a slight decline which left final values at or near initial ones. Aquaria without plants, meanwhile, showed similar slight increases through day 5, at which time precipitous declines occurred, with values leveling out in the final days of the run. The declines in hardness and alkalinity were of similar magnitude and corresponded well to increasing pH, suggestive of carbonate precipitation. As would be expected, mean values for both hardness and alkalinity were significantly lower in aquaria with plants (Table 6). In the second experimental run, hardness and alkalinity in aquaria without plants behaved qualitatively similar to the first run, with major differences in the magnitude of the decline after day 5 being attributable to the different initial values of hardness and alkalinity. The aquaria with plants, however, behaved markedly different from both the aquaria without plants and the aquaria with plants in the first run, showing large, immediate declines in both hardness and alkalinity and lesser declines that continued throughout the run. Again this is best attributed to the rise in pH and suggests carbonate precipitation, deposits being observed on the plants starting on day 1. Hardness and alkalinity were significantly lower in aquaria with plants throughout this run and showed correspondingly lower means (Table 6), an effect opposite to that observed in the first run. Phosphorus Total phosphorus values for all aquaria in both experimental runs are plotted on Figure 6. Dissolved phosphorus and orthophosphorus values for the second experimental run are included in Figure 7. Even more so than 4O \ 0.4- TOTAL PHOSPHORUS (mg/l P) g» 010’ ¥ '- 0‘ u/ 0.2, 23 O Sh I l /7 0.2 - o.I - \ g 7: Q32 A‘EV’ L Hr—I‘s QI- l J___L L O 3 6 9 l2 l5 1.) 3 TlMEkloys) Figure 6. Total phosphorus vs. time for all aquaria during experimental runs one (left, aquaria 1, 4, 5 with plants) and two (right, aquaria 2, 4, 6 with plants). 41 Q?- 07— 0.6 - L, 0.6 - 3 if 3 05)- g 0.5- 35 I; :3 I g E (D ‘5 I 2 I o $ 0.4 " cIL 0.4 - o a, . I o “- 3: Q. :3 3 C3 > 5 I .l I— C>CL3T- czCL3- m o 9 35 o I 0.2- 0.2 - I 3 I Ql- QI- S 4 \74§>\