Response of shallow aquatic ecosystems to different nutrient loading levels = Respons van ondiepe aquatische oecosystemen op verschillende nutrientenbelastingnivo's

Eutrophication of surface waters leads to a decline of water quality, which becomes manifest as an impoverishment of the aquatic community. Insight into the effects of eutrophication on the structure and functioning of these communities and knowledge on underlying interactions is needed to quantify the required reduction of nutrient input.To investigate the effects of nutrient loading on the receiving water, it is important that environmental conditions that influence the response of a system to the level of nutrient input, are either controllable or measurable. This also holds for the initial conditions. These have to be identical to facilitate mutual comparison of different experimental - systems.In this research the effects of the level of nutrient loading are studied in experimental ditches. The emphasis on the internal cycling of phosphorus and the productivity of the system. with respect to scale, the ditches form a link between experimental set-ups in the laboratory and research on a full natural scale. In the first category the conditions can be well controlled, but extrapolation to a larger scale is often difficult, whereas in the second category the hydrology and environmental conditions cannot be controlled or measured with sufficient accuracy. Besides this, ditches also account for an important fraction of the Dutch surface waters.Eight ditches have been used, four of which have sand as bottom material. The other four have a clayish sediment. Initially, all sand ditches were dominated by benthic algae, and had virtually no macrophytes. The clay ditches were dominated by submersed aquatic macrophytes. The initial nutrient contents in both sediment types were low. For each sediment type four different levels of external nitrogen and phosphorus input were established, varying from as low as possible to extremely high. The levels are referred to as A (reference), B (second level), C (third level) and D (highest level).In the clay-ditches the development of the macrophyte communities after the start of the nutrient loading programme has been monitored (see chapter 3). This concerns the species composition and the relative areal coverage of individual species. Initially the species composition was virtually identical. The ditches were dominated by Characeae, with areal coverage percentages of over 90% in all four. Gradually, the Characeae were replaced by Elodea nuttalliiin all ditches, and the rate of the transition was positively related to the level of nutrient input. At the lowest level of nutrient input Elodeastill exhibited a vertical growth strategy at the end of the investigation period, whereas at the two intermediate levels the growth strategy of Elodea was horizontal. Here it formed a dense vegetation near the water surface with areal coverage percentages of almost 100% during the summer period. In the highest loaded ditch Elodea was in turn replaced by Lemna minor, which forms a dense bed on the water surface during the summer. Later on, a generally terrestrial species (Senecio congestus) showed up as a pleustophyte. The location of the production zone remained in the bottom part of the water column in the reference ditch. It shifted to the top part of the water column at the second and third trophic state, and to the water surface (and above) at the highest trophic state.In small systems the fluxes across the boundaries play a relatively large role. For gases the exchange between the air water interface is important. This exchange is determined by the mass transfer coefficient, which depends on environmental conditions. In stagnant waters wind plays an important role. For the study of primary production the exchange of oxygen and carbon dioxide are of importance. Analysis of diurnal dissolved oxygen curves can provide information on the productivity of the aquatic ecosystem. To be able to estimate primary production and oxygen consumption with sufficient accuracy, the reaeration term in the oxygen mass balance must be known.Chapter 4 describes experiments in which mass transfer across the air-water interface as a function of wind speed is measured with the use of a tracer gas. The obtained windreaeration relationship is subsequently implemented in a model to simulate the dissolved oxygen concentration in the reference sand ditch. The other model parameters, describing primary production and oxygen consumption, were measured. This resulted in a fairly good agreement between measured and simulated dissolved oxygen during a period of four days.Extrapolation to longer periods is not possible, as the values of the model parameters that describe primary production and oxygen consumption are not constant. To estimate primary productivity over longer periods, a parameter estimation routine was used to fit the oxygen model to measured daily oxygen curves. This was done for each individual day for a period of two years. Reaeration was again calculated from the windreaeration relationship. Gross production and oxygen consumption were subsequently estimated on a daily basis and as cumulatives over the two years period. Results are compared to the level of external nutrient loading.The application of the model and parameter estimation routine to the sand-ditches are described in chapter 5, and in chapter 6 it is applied to the clay-ditches. Due to frequent and longlasting anaerobia the method could not be applied to the highest loaded ditches (D) in both series.In the sand-ditches, dominated by benthic algae, both gross and net production were positively related to the level of external nutrient input. In the reference ditch, the accumulation of dry weight in the layer of benthic algae, as calculated from the oxygen mass balance and the stoichiometry of the photosynthesis reaction, showed good agreement with the measured increase. At the second and third level the measured accumulation was less than calculated.For application of the method to the ditches dominated by macrophytes the model was extended to a two layer model for those periods during which steep vertical gradients in the dissolved oxygen concentration occurred. Those vertical gradients occurred during the summer periods when Elodea nuttallii formed a dense vegetation near the water surface. Each layers is supposed to be ideally mixed. Due to the strong vertical attenuation of light, primary production is assumed to occur exclusively in the top layer. In the bottom layer only oxygen consumption, resulting from respiration, decomposition and sediment oxygen demand, occurs. The exchange between the top and bottom layer is driven by the vertical concentration gradient.The cumulative gross production is again positively related to the nutrient input. Cumulative net Production showes an irregular pattern. Initially it was the highest in ditch C, but during the second year the cumulative net production declined in this ditch due to an increased oxygen consumption.In the clay-ditches total gross production was smaller in the second than in the first year, whereas in the sand-ditches this was the other way round.In case of an enhanced phosphate loading to an oligotophic system, the sediment can take up a considerable part of the extra phosphate. The flux of phosphate from the water column to the sediment consists of the settling of particulate P, and a diffusive exchange of dissolved P between the water column and the interstitial water. The driving force for the diffusive exchange is a vertical concentration gradient. After enhanced phosphate input to the overlying water, the vertical concentration in the interstitial water is negative with depth, due to the sorption of P by sediment particles. The transport of ions along the concentration gradient is proportional to the effective dispersion coefficient D eff , which is the sum of molecular diffusion D mol , bioirrigation D bio , and a dispersion term D dis . The latter results from pressure gradients, induced by formation of waves by wind stress at the water surface, resulting in an oscillating horizontal flow. Due to the presence of particles this flow has a vertical component which becomes manifest as an additional dispersion term Ddis in the sediment. It is at maximum at the sedimentwater interface and attenuates with depth in the sediment. Ddis varies on a small time-scale due to variations in wave length and amplitude. In this study, and generally in shallow waters, Ddis at the surface boundary dominates in D eff , and can be one to two orders of magnitude larger than D mol . A sediment model that desribes the diffusive transport of dissolved phosphorus from the overlying to the interstitial water and subsequent sorption by the sediment particles (see chapter 7), was calibrated on a measured vertical concentration profile in the porewater, almost two years after the start of the nutrient loading program. THe results indicate that for this period Deff at the sediment-water interface was on average about 35 times D mol .The sorption of phosphate by the sediment particles contains a fast reversible step, which is adsorption onto the surface of the particles, and is followed by a slow step. The latter can be described as an internal diffusion of adsorbed P into the solid phase. This solid phase was modelled as a coating of metal(hydr)oxides (in this study the metal was mainly aluminum) surrounding the sand particles. Removal of the coating resulted in a virtually complete disappearance of the ability of the particles to sorb P. From experiments in which the coating was dissolved and subsequent (co) -precipitated with phosphate, it was estimated that the total sorption capacity of the whole coating was about 3.3 times the maximum surfacial adsorption capacity. Diffusion into the solid phase can be described analogous to diffusion in a fluid. The diffusion coefficient of phosphate in the Al- (hydr)oxide coating, experimentally estimated from the kinetics of phosphate sorption, was 2.7 10 -19m 2.day -1, which is between 10 -14and 10 -15times the molecular diffusion coefficient of phosphate in water. This shows that solid phase diffusion is a slow process. It has a time scale of several months. For the long term storage of phosphate in the sediment it is important, as ultimately ca. 70 % of the phosphate is internally stored in the metal-(hydr)oxide coating.After a phosphate dosage to the sand-ditches part of the added phosphate is rapidly taken up by the benthic algae. The uptake is in excess of the amount immediately needed for the assimilation of cell material. The kinetics of this phosphate uptake can be described well as a first order process (see chapter 8). The internal P-deficit, defined as the difference between the maximum and the actual internal P content of the cells, is the driving force. The first order uptake rate did not differ among three benthic populations, previously exposed to different levels of external nutrient loading (from the sand-ditches A, B and C). The maximum internal P concentrations P int.max . on the other hand were positively related to the trophic state, indicating mechanisms that regulate the phosphate storage capacity. The additional uptake capacity, i.e the initial P deficit, was smaller after exposure to higher levels of nutrient loading, but is somewhat extensible due to adaptation of P int.max . This has consequences for the long term stability of dominance by benthic algae. An upward movement of the production zone from the sediment top layer to the water column can occur when the nutrient availability in the water column is enhanced. This means that the uptake capacity of the benthic algae and sediment must have become insufficient to compensate for the external nutrient input.An increased productivity, and concomitantly an upward movement of the production zone, following enhanced nutrient loading, results in an enhancement of the carbon dioxide flux from the atmosphere. At high pH values the concentration of free C02 (aq) is generally lower than the equilibrium concentration following from its solubility. Due to chemical equilibrium reactions the concentration gradient in the boundary layer is steeper near the air-water interface, than it would be due to purely physical processes. This results in an enhanced C02 mass transfer across the air-water interface compared to that of a chemical inert gas. The acceleration factor depends on pH, thickness of the laminar boundary layer, total inorganic carbon and temperature. With a model that calculates the acceleration factor from these variables, the atmospheric carbon dioxide flux and its contribution to the net carbon assimilation have been estimated (see chapter g). The cumulative flux during the growth season increased with increasing levels of nutrient loading in the ditches dominated by benthic algae. From the ditches dominated by submersed macrophytes, the two ditches dominated by Elodea nuttallii (ditch B and C) had similar cumulative C-fluxes. These were estimated at 85 and 97 % of the cumulative net C assimilation during the same period, indicating the importance of atmospheric C02 for the carbon supply of these communities. In the ditch dominated by Characeae (ditch A), this fraction smaller.The fraction of the net C assimilation during the growth season that was covered by the atmospheric C-influx was substantially less in the ditches dominated by benthic algae than in the ditches with macrophytes.