To: Oyster Pond Environmental Trust
From: Roxanne Marino, Ph.D. (staff scientist, The Ecosystems Center, MBL, Woods Hole, MA 02543), Robert Howarth, Ph.D. (David R. Atkinson Professor of Ecology & Environmental Biology, Cornell University, Ithaca, NY 14853), and Eric Davidson, Ph.D. (Woods Hole Research Center, Woods Hole, MA 02543)
Date: 22 August 2002
Over the past two months, we jointly supervised two undergraduate students in an NSF-funded training program designed to provide research experience for undergraduates. These students, Stacy Barron and Carolyn Weber, conducted two sets of experiments on nutrient limitation and salinity in Oyster Pond. Abstracts of these experiments will be published in the Biological Bulletin (copies attached). Here, we briefly describe the rationale for the design of these experiments, our interpretation of the results, and suggestions for further study to better understand how nutrient enrichment may affect water quality in Oyster Pond in the future. The students also conducted 3 transects along 4 stations in Oyster Pond, where they measured chlorophyll, inorganic nitrogen (N) and phosphorous (P) concentrations, and salinity at 3 depths, as well as secchi depth, light, and dissolved oxygen (the latter two were not done on all 3 dates due to technical problems). We are working with the students to finish analyzing these data, which we will then make available to OPET as well.
The background logic for the experiments is further developed in a series of papers and reports, including the report of the National Academy of Sciences' Committee on Causes and Management of Coastal Eutrophication (chaired by Howarth between 1998 and 2000: Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution, published in 2000 by the National Academy Press, Washington, DC). We have listed a few of these references at the end of this memo, and would be happy to provide them to anyone interested in seeing them.
Nutrient limitation in coastal marine ecosystems - some Background:
Nutrients are now generally recognized as the largest pollution problem for coastal rivers and bays, and in the United States, two-thirds of these ecosystems are moderately to severely degraded by nutrients. Coastal marine ecosystems vary in their sensitivity to nutrients, and many other factors interact with nutrients to control primary production and eutrophication, including light, mixing depths, residence times, and grazing by zooplankton and benthic filter-feeding animals. For many ecosystems, adding more nutrients will increase primary production, but the response can be amplified or damped by these other controlling factors.
For coastal marine ecosystems of moderate to high salinity (greater than about 10 parts per thousand, or o/oo), nitrogen is generally the nutrient most limiting to production. This is in sharp contrast to lakes of moderate to high productivity, where phosphorus is more generally limiting. Many scientists have pointed out this difference for more than 30 years, yet until quite recently, the management community had been slow to accept that coastal marine ecosystems are different from lakes, and nutrient management in the United States and Europe has focused largely on phosphorus. This has only recently changed, and now management strategies generally call for management of both nitrogen and phosphorus inputs to both coastal systems and lakes, but with a greater emphasis on controlling nitrogen inputs to coastal systems. Nitrogen is more mobile in the landscape and has more sources than does phosphorus (including an atmospheric source), and thus is more difficult to manage.
While many studies have demonstrated phosphorus limitation in freshwater lakes and ponds and nitrogen limitation in coastal ecosystems at salinities greater than 10-15 o/oo, there have been relatively few studies to determine whether nitrogen or phosphorus is more limiting in brackish coastal ecosystems with salinities in the range of Oyster Pond (2.3 o/oo during June-August 2002). Generally, nitrogen has been found to be more limiting to production at salinities greater than 3 to 6 o/oo in nutrient-enriched estuaries such as the Baltic Sea and the upper reaches of Chesapeake Bay and its tributaries, while phosphorus has been found to be more limiting at lower salinities. In Oyster Pond, two previous studies found that neither nitrogen or phosphorus additions affected phytoplankton growth (Caraco et al., 1987, Can. J. Fish. Aquat. Sci. 44: 473-476, and a Boston University Marine Program class project in 2001), but both of these experiments were conducted in October, and the Pond might react differently during summer months.
The 2002 summer experiments in Oyster Pond - design and rationale:
Barron and Weber worked with us this summer to conduct two types of experiments in Oyster Pond: 1) a set of experiments in which ambient water from the Pond was enriched with either nitrogen or phosphorus, and 2) a set of experiments in which salinity was varied between 0.2 and 5 o/oo. All of these experiments were relatively short in duration (5-10 days), and all were conducted in 2-liter bottles. The bottles were incubated either in the Pond at ~0.5 meters depth, or in a growth chamber at Pond temperatures and under lights which provided an average light intensity similar to that in the upper 0.5-1 meter of the Pond. We took several precautions to minimize problems frequently encountered in such experiments. For example, we added bicarbonate to buffer against large changes in pH and to prevent carbon dioxide depletion, a major problem with some nutrient limitation bottle experiments done in freshwater lakes. Nonetheless, such small-scale, short-term experiments should be interpreted with care, for the following reasons:
Short-term nutrient addition experiments done in bottles and so isolated from the ecosystem, do not fully represent nature because, for example, they do not allow for changes in plankton community structure in response to feedback to the nutrient additions. It is also extremely difficult if not impossible to scale nutrient additions in these experiments to actual system nutrient loadings and availabilities (if these are even known) and so allow prediction of dose responses, as nutrient cycles are complex and depend on external inputs as well as internal transformation processes. Hence, these type of nutrient limitation experiments are traditionally done using nutrients (N or P) added in excess of the ambient water concentrations, to allow for a detectable response over the time frame water can reasonably be held small bottles (days to 1-2 weeks maximum). If a phytoplankton growth response (as for example, increasing chlorophyll a concentrations) is observed in response to the elevated N or P levels, it is assumed that the phytoplankton are limited in the same way at the in-situ nutrient availability (i.e. if phytoplankon biomass increases in response to N and not P enrichment, then growth of that phytoplankton assemblage at ambient levels of N and P is interpreted to be limited by N).
Despite the limitations and constraints on experimental design, short-term bottle experiments have been an important tool that has been used by many investigators (including ourselves) to gain information on the relative importance of one nutrient versus another to primary producers in an aquatic system. The information gained from such experiments is most powerful in conjunction with studies of primary production and nutrient input, availability, and cycling at other time and spatial scales (mesocosms and limnocorrals, whole ecosystems, seasonal and annual field observations).
The 2002 summer experiments in Oyster Pond - Results:
We ran two separate experiments where ambient water was enriched with either nitrogen or phosphorus. The purpose was to determine which of these nutrients is more limiting to phytoplankton growth in Oyster Pond during the summer under present nutrient and phytoplankton conditions. Both experiments showed that nitrogen is more limiting. This suggests that in order to reverse problems of excessive phytoplankton growth from nutrient overenrichment in Oyster Pond, nitrogen inputs will likely need to be reduced. In one experiment, both N and P were added, resulting in a much greater increase in phytoplankton growth than that of N alone. This indicates that with added N, P can quickly become limiting to further growth.
In the salinity experiments, ambient water with phytoplankton from Oyster Pond was diluted with either deionized water or filtered seawater from Vineyard Sound to provide salinities of 0.2, 2.3 and 5.0 o/oo. In one of these experiments, both N and P were added at moderately high levels (approximately equivalent to the N and P added alone in the first set of experiments). The purpose of this experiment was to determine whether the phytoplankton present in Oyster Pond were adapted for best growth at the ambient salinity. In many low salinity estuaries, the phytoplankton species present are largely freshwater forms, and would presumably grow better as salinity is decreased. Surprisingly, salinity had little influence in our experiment, and the phytoplankton grew well at all three salinities tested. Overall, the phytoplankton community was little affected by salinity even in this very short-term experiment.
In another experiment with different salinities, the bottles were enriched only with phosphorus. The purpose of this experiment was to examine whether nitrogen-fixing cyanobacteria ("bluegreen algae") present in Oyster Pond would respond more or less strongly to different salinities. These organisms are able to take molecular N2 present in the atmosphere and dissolved in water and convert it into biologically available nitrogen (the process of nitrogen fixation), which has been shown to aggravate nutrient pollution problems in freshwaters. These cyanobacteria are common in lakes, but are very rare in coastal marine ecosystems at higher salinities (their relative absence in these higher salinity coastal systems is one reason that nitrogen is often limiting). We had predicted that the cyanobacteria would grow better and fix more nitrogen at the lowest salinity (0.2 o/oo). However, we found the greatest increase in cyanobacteria and heterocysts (the specialized cells where N fixation occurs) at the ambient salinity (2.3 o/oo), suggesting that they are well adapted to the ambient Pond salinity. We note that these data are preliminary, as only one replicate bottle for each salinity (end of the experiment) has been counted so far.
Management implications:
The experiments with nutrient enrichments of ambient water from Oyster Pond indicate that it may be important to control nitrogen inputs to the Pond, if the goal is to reduce phytoplankton production and associated eutrophication problems. As is now generally recommended for estuaries, it is probably also prudent to consider and control phosphorus inputs, especially as these inputs might change over time.
Would altering the salinity of the Pond have an affect on algal production and eutrophication? The 2002 summer experiments were too short in duration and do not encompass enough of the potential responses of the ecosystem to provide a clear answer to this question. For example, changing the salinity is likely to alter the net exchange of nutrients with bottom sediments, which may have a profound effect on eutrophication. Nonetheless, the experiment did show the presence of nitrogen-fixing cyanobacteria. These organisms can aggravate nutrient pollution problems by fixing nitrogen when it is in short supply and phosphorous is available, and potentially can counteract some of the benefit of reducing watershed inputs of nitrogen to an ecosystem. From other information, we think it likely that lowering the salinity of Oyster Pond would generally favor these nitrogen-fixing cyanobacteria (over a longer time period than examined in the short-term 2002 summer experiment), while increasing the salinity might suppress them. However, the response might depend on many factors, including changes in nutrient availabilities due to the response of sediments, and changes in the abundances and species composition of zooplankton and benthic filter-feeding animals.
Opportunities and needs for further research on nutrient pollution in Oyster Pond:
These experiments provided some interesting preliminary information on how Oyster Pond might respond in the future to nutrient inputs. We see two priority areas for further research on nutrient pollution in Oyster Pond: 1) research on characterizing the nutrient inputs and basic biological response variables, such as chlorophyll and primary production, over time; and 2) research on how Oyster Pond might respond to higher or lower salinities within the range that could be achieved by water flow changes at the weir.
Nutrient inputs to Oyster Pond are not well characterized at present. Most of the input of phosphorus and much of the input of nitrogen presumably come from septic systems and groundwater, but what is the time of transport between the septic systems and the Pond? Are the current nutrient loads to Oyster Pond in steady state, or is there a time delay so that the Pond is not yet experiencing the full effects of the current population density? Also, as with most coastal marine ecosystems, the nitrogen inputs from atmospheric deposition are very poorly known. These could be substantial, particularly if short-term transport of nitrogen oxides originating from traffic on Woods Hole Road is significant. Further, it would be useful to characterize the proportion of total nitrogen and phosphorous in the water seasonally that is in readily available, inorganic forms and in dissolved organic forms which are only partially available for plant growth.
To study how salinity changes might interact with nutrient pollution in Oyster Pond requires experiments of sufficient duration and size so as to capture changes in sediment processes and plankton community structure. We believe such experiments could be performed in "limnocorrals," or large plastic enclosures (usually between about 0.5 meter and 3 meters in diameter) within Oyster Pond over a scale of weeks to months. A critical question would be the response of the nitrogen-fixing cyanobacteria to the salinity change. Another key aspect would be how both the uptake and release of nitrogen and phosphorus from the sediments respond to variations in salinity. Limnocorral experiments could also be used to define a dose-response curve of nutrient inputs to phytoplankton production and eutrophication in Oyster Pond, if the experiments were appropriately scaled and of sufficient duration. Along with a seasonal study of phytoplankton and nutrient dynamics in the Pond, this type of experimental study would require larger scale and more sustained funding (i.e. a grant from NSF or some other agency), than some of the other suggested studies, which could be pieced together with smaller amounts of funding from local sources. We are interested in working with OPET and others to obtain funds to pursue some of these research questions in the future.
Selected References:
Marino, R. 2001. An Experimental Study of the Role of Phosphorus, Molybdenum, and Grazing as Interacting Controls on Planktonic Nitrogen Fixation. Ph.D. thesis, Cornell University, Ithaca, NY.
Marino, R., F. Chan, R.W. Howarth, M. Pace, and G. Likens. 2002. Ecological and biogeochemical interactions constrain planktonic nitrogen fixation in estuaries. Ecosystems, in press, October, 2002.
Howarth, R.W. and Marino, 1998. A mechanistic approach to understanding why so many estuaries and brackish waters are nitrogen limited. Pages 117-136 in Effects of Nitrogen in the Aquatic Environment, KVA Report 1998: 1, Kungl. Vetenskapsakademien, Stockholm (Swedish Royal Academy of Sciences).
Email OPET
Contact Webmaster