Responses of marsh algal communities to controlled nitrogen and phosphorus enrichment L. Gordon Goldsborough Department of Botany, Brandon University Winnipeg, Manitoba, Canada E-mail: ggoldsb@umanitoba.ca

Introduction

It is well known that nutrient enrichment of lakes, especially by nitrogen and phosphorus, contributes to their accelerated eutrophication. Specifically, increased nutrient levels cause increased total primary production, often by bloom-forming cyanobacteria. This observation has arisen largely from studies of deep aquatic ecosystems where phytoplankton are the predominant primary producers and pelagic invertebrates and fish are the primary consumers. There is comparatively little information on the responses of shallower, littoral-dominated ecosystems such as wetlands to nutrient enrichment, so the extent to which the above generalization applies to such systems is unknown.

There are several ways in which N and P levels in a wetland may increase. It is widely recognized that wetlands are maintained by periodic water level fluctuations (Good et al. 1978). During periods of climatically-induced drought or intentional drawdown, inorganic nutrients are generated via decomposition of organic materials in exposed sediments. These nutrients are subsequently released into the water column following reflooding. In addition, the feces of geese and ducks contribute significantly to the nutrient budget of lakes (Manny et al. 1994). It is likely that such additions are at least as important in wetlands, which are often important resting, feeding and breeding grounds for waterfowl (Batt et al. 1989). Benthivorous fish such as carp (Cyprinus carpio), introduced to wetlands from adjoining lakes and inflowing rivers, can increase sediment resuspension (Meijer et al. 1990), thereby enhancing efflux of sediment-bound nutrients. Finally, nutrients in fertilizers and sewage may represent significant inputs to wetlands receiving drainage from agricultural, domestic and industrial sources.

Using a series of large, littoral enclosures constructed originally for use in toxicological studies in the Delta Marsh (Goldsborough 1991), studies were initiated in 1993 to examine the responses by benthic algae (periphyton and epiplon) and phytoplankton to controlled additions of N and P. The basis for the experiment was an hypothesis that marsh plants and their epiphytes outcompete phytoplankton for nutrients under unperturbed conditions. When nutrients are added to the water column, phytoplankton may compete effectively. Thus, it was predicted that pulsed additions of N and P would cause a shift in the composition of algal primary producers, from an epiphyte-dominated system similar to that occurring naturally in the marsh to one in which phytoplankton were abundant.

Methods

Each experimental enclosure consisted of a wooden frame, 5 m by 5 m in size, with a 40 cm wide walkway around it. The frame was supported slightly above the water surface by foam blocks fastened under the walkway. A translucent plastic curtain was attached to the inner side of the frame. It extended through the water column and into the sediments. The bottom end of the curtain was fastened to metal rods embedded in the bottom, thereby preventing the direct movement of water between the inside of the enclosure and the surrounding marsh. The total enclosed water volume was about 15,000 L. Four enclosures fastened together were deployed in the center of the Blind Channel in mid-May, 1993.

Minnow traps were placed into each enclosure immediately after curtain installation to remove fish. The objective was to exclude fish as a potential consumer and, therefore, to reduce the number of variables in the experiment. However, fecund females laid numerous eggs on the inner side of the curtain prior to their removal and fry were subsequently observed in all enclosures. Traps were checked daily for the duration of the experiment.

One hundred cylindrical acrylic rods (0.64 cm diameter, 90 cm length) were positioned vertically in each enclosure on 18 May, 1993 so the uppermost 60 cm was available for periphyton (epiphyton) colonization. Prior to placement, the substrata were notched with a small saw to provide two adjacent 2.5 cm segments (27.5 to 32.5 cm above the sediment interface when in situ) for carbon fixation measurements. These were bracketed by two 10 cm substratum segments that were retained for algal chlorophyll analysis. The remainder of the substratum was unused. Sampling began on 1 June (designated as "week 1") and continued at weekly intervals until 17 August (week 12). At each sampling time, six colonized substrata were selected from random positions within the enclosure and subsampled segments were returned to the laboratory for analysis. At the same time, three 1 L samples of surface water (< 10 cm depth) were collected at random positions from each enclosure for analysis of phytoplankton carbon fixation and chlorophyll. Beginning on 7 June, three samples of surface sediments (circa 1-2 cm) were collected from random positions in each enclosure using a vacuum aspirator apparatus.

Substratum segments used in measurements of algal carbon fixation were placed in clear glass tubes containing filtered water from the same enclosure as the sample. Radiolabeled bicarbonate solution (1 µCi/mL) was added to each tube. The tubes were then placed in a benchtop water bath that maintained a constant temperature (25°C) and irradiance (500 µE/m2/s) over an incubation time of 3-4 hours. Following incubation, segments were collected onto filters under vacuum, fumed over concentration HCl to liberate residual inorganic radiolabel and placed into liquid scintillation cocktail. Sample radioactivity was determined by liquid scintillation counting and used, along with the pH, temperature and alkalinity of the incubation medium, to calculate carbon fixation rate during the incubation period. Substratum segments collected for chlorophyll analysis were frozen for at least 24 hours then thawed and placed in 90% methanol. Pigments eluted into the solvent were measured spectrophotometrically and chlorophyll concentration was calculated using the formulae of Marker et al. (1980).

Sediment slurry samples were transferred at the laboratory into beakers whose sides were blackened to prevent light penetrance. They were allowed to stand in a dark drawer for 24 hours then the supernatant was carefully withdrawn. The wet sediment surface was covered by a piece of lens paper and the beakers were transported to the Field Station Met Station where they received natural irradiance for circa 18 hours. The lens papers were then removed from each beaker and suspended in a sample of filtered water from the same enclosure from which the sediment samples were collected. The samples were shaken vigorously to dislodge associated epipelic algae and subsamples of the resulting suspension were dispensed into clear glass tubes for carbon fixation rate measurements. Incubation and post-incubation sample preparation conditions were identical to those used for periphyton and phytoplankton samples.

Two enclosures were arbitrarily chosen to receive nutrient additions and the other two enclosures were unmanipulated controls. The first nutrient addition took place on 20 June. An aqueous solution of NaH2PO4 and NaNO3 in a concentration sufficient to yield an initial expected concentration of 1 mgP/L and 10 mgN/L was added to the surface of each treated enclosure. The solution was mixed thoroughly in the surface waters using a paddle. Water samples were collected from all enclosures prior to the addition and at daily intervals after treatment, and analyzed for ammonia and orthophosphate. Additional samples were collected at weekly intervals from all enclosures and submitted to the W. M. Ward Technical Services Lab (Winnipeg) for analysis of nitrate+nitrite, total phosphorus, total dissolved phosphorus and total Kjeldahl nitrogen. A second spike addition of the same N and P concentration was made to the same treated enclosures on 2 August, after which water samples were collected at 12 hours intervals and analyzed for orthophosphate and nitrate+nitrite. Methods of water analyses followed Stainton et al. (1977).

Results & Discussion

Nitrate-N and orthophosphate-P levels in untreated enclosures, like those in the surrounding Blind Channel, were at or below detection limits throughout the experiment. There was no evidence of nutrient leakage from treated to untreated enclosures. Twelve hours after the first nutrient addition, N and P levels in treated enclosures were 6.6 mg N/L and 0.56 mg P/L, respectively. As expected, nutrient levels in surface waters decreased rapidly with time. Ten days after the addition, N and P levels were similar in treated and control enclosures. Frequent water sampling done after the second nutrient addition confirmed this rapid dissipation rate. Samples from treated enclosures collected 2.5 hours after the addition contained 9.0 mgN/L and 0.94 mgP/L (Fig. 1). First-order dissipation half lives for N and P, calculated using data collected during the first 8 days after treatment, were 108 hours (4.5 days) for N and 62 hours (2.5 days) for P. By 8-9 days after treatment, N levels were slightly above detection (0.1 mg/L) while P levels remained above detection for the remainder of the experiment (Fig. 1). The disappearance of added nutrients was likely due to the combined effects of assimilation by algae and macrophytes, and chemical transformation in the water column and sediments. The half-lives were longer than anticipated on the basis of tracer experiments done in lake epilimnia (e.g., Rigler 1956), where total phytoplanktonic assimilation within hours was the norm.

Periphyton biomass increased in nutrient-enriched enclosures, as compared to untreated controls (Fig. 2a) and remained higher throughout the experiment. On the other hand, phytoplankton biomass was lower in treated enclosures (Fig. 2b). There was no difference in epipelon biomass between enclosures (Fig. 2c). There was no apparent effect of the two nutrient additions on carbon fixation rates of periphyton (Fig. 3a), phytoplankton (Fig. 3b) or epipelon (Fig. 3c). In part, these results are due to high variability between substrata from a given enclosure, and variability between replicate enclosures. This variability arises from spatial variation in the abundance and species composition of macrophytes. Where macrophyte development is profuse, shading and competition for nutrients results in lower algal biomass.

These results are significant in the context of a conceptual model for algal communities in temperate wetlands proposed by Goldsborough and Robinson (1995; Fig. 4). By this model, most areas of the Delta Marsh are "open marshes" in which epiphytes colonizing the surfaces of submersed macrophytes are the predominant algae. This dominance is maintained, in part, by physical disturbance caused by wave action and carp, and invertebrate herbivory, all of which ensure that epiphyton biomass remains sufficiently low that it does not eventually eliminate macrophytes via shading (cf. Phillips et al. 1978). When this stable state is perturbed through nutrient enrichment, the system may evolve either into a "sheltered marsh" where mats of filamentous green algae (metaphyton) predominate, or to a "lake marsh" dominated by phytoplankton. In either case, increased nutrient loading initially leads to increased epiphyton biomass which, in turn, leads to increased shading and bacterial nectrophy of macrophytes. If macrophytes are reduced to the extent that their stabilizing influence on the sediments, and ultimately the water column, is lost, competitive stress on phytoplankton for nutrients may be alleviated, causing them to flourish. The resulting phytoplankton blooms can lead to the eventual elimination of macrophytes, thence to the stable "lake marsh" state. Such a progression has occurred in wetlands such as those in the prairie "pothole" region of western Manitoba, North Dakota and Minnesota that have undergone nutrient loading (Hanson and Butler 1994). Alternatively, if macrophytes remain in enriched areas of the wetland, providing substratum for the initiation of metaphyton, a "sheltered marsh" persists. The latter alternative occurred in experiments comprising the Marsh Ecology Research Program (MERP), which were conducted at the Delta Waterfowl and Wetlands Research Station in the 1980s. During MERP, flooding of drawn-down areas of the marsh were observed to produce massive floating meadows of metaphyton (Hosseini and van der Valk 1989).

Results of this study contradict the prediction that marsh phytoplankton would respond positively to nutrient enrichment. Instead, increased epiphyton biomass and the persistence of macrophytes in all enclosures may explain the net reduction in phytoplankton. This indicates that the systems within treated enclosures were progressing to the "sheltered marsh" state. The absence of conspicuous metaphyton, contrary to model expectation, may relate to the nature of the experimental treatment. Although N and P levels were high immediately following spike additions, their rapid dissipation precluded a high ambient nutrient environment needed to cause such a shift. It would be interesting to determine if different results could be obtained by adding the same cumulative nutrient load at more regular intervals. Such an experiment is planned for the summer of 1994.

Acknowledgements

This research was supported by NSERC Research Grants to L. G. Goldsborough and G. G. C. Robinson. Samples were collected by Colleen Flynn and Jennifer Barker as part of their Co-op Workterms in the Environmental Science Program at the University of Manitoba.

Literature Cited

Batt, B. D. J., Anderson, M. G., Anderson, C. D., and Caswell, F. D. 1989. The use of prairie potholes by North American ducks. In A. van der Valk (ed.) "Northern Prairie Wetlands", pp. 204-227. Iowa State Univ. Press, Ames, Iowa.

Goldsborough, L. G. 1991. Effects of hexazinone on water chemistry and periphyton biomass and productivity in large littoral enclosures. University Field Station (Delta Marsh) Annual Report 26: 50-62.

Goldsborough, L. G. and Robinson, G. G. C. 1995. Pattern in wetlands. In R. J. Stevenson, M. L. Bothwell and R. L. Lowe (eds.) "Benthic Algal Ecology in Freshwater Ecosystems", Chapter 4. (in review)

Good, R. E., Whigham, D. F. and Simpson, R. L. 1978. "Freshwater Wetlands: Ecological Processes and Management Potential", Academic Press, New York.

Hanson, M. A., and Butler, M. G. 1994. Responses to food web manipulation in a shallow waterfowl lake. Hydrobiologia 279/280: 457-466.

Hosseini, S. M., and van der Valk, A. G. 1989. The impact of prolonged, above-normal flooding on metaphyton in a freshwater marsh. In R. R. Sharitz and J. W. Gibbons (eds.). "Freshwater Wetlands and Wildlife", pp. 317-324. DOE Symposium Series No. 61, USDOE Office of Scientific and Technical Information, Oak Ridge, Tennessee.

Manny, B. A., Johnson, W. C., and Wetzel, R. G. 1994. Nutrient additions by waterfowl to lakes and reservoirs - predicting their effects on productivity and water quality. Hydrobiologia 280: 121-132.

Marker, A. F. H., Crowther, C. A., and Gunn, R. J. M. 1980. Methanol and acetone for estimating chlorophyll a and phaeopigments by spectrophotometry. Arch. Hydrobiol. Beih. 14: 52-69.

Meijer, M. A., deHaan, M. W., Breukelaar, A. W. and Buiteveld, H. 1990. Is reduction of the benthivorous fish an important cause of high transparency following biomanipulation in shallow lakes? Hydrobiologia 200/201: 303-315.

Phillips, G. L., Eminson, D. and Moss, B. 1978. A mechanism to account for macrophyte decline in progressively eutrophicated freshwaters. Aquat. Bot. 4: 103-126.

Rigler, F. H. 1956. A tracer study of the phosphorus cycle in lake water. Ecology 37: 550-562.

Stainton, M. P., Capel, M J. and Armstrong, F. A. J. 1977. "Chemical Analysis of Fresh Water", 2nd ed., Can. Fish. Mar. Serv. Misc. Spec. Publ. 25.