During the first 6 years of funding, we developed a foundation of research that includes long-term field experiments and measurements in watersheds, atmosphere, intertidal marshes, estuary and coastal ocean. Experiments and observations examine LTER core areas: primary production, organic matter accumulation, nutrient cycling, higher trophic levels, and ecosystem disturbance.
WATERSHED HYDROLOGIC CYCLE
Following substantial changes in land use in the Ipswich and Parker River watersheds during the 20th century (from >50% agriculture in 1900, to > 85% forest in 1950, to >35% urban in 2000) (Pontius and Schneider 2001), we were surprised to find that discharge was largely unchanged. We expected to see increases in discharge due to the increase in impervious surface and a projected decrease in evapotranspiration (ET) associated with land use change from forest to urban (Claessens et al. submitted). Instead we found that diversions and climate change decreased the percentage of rainfall exported from the watershed in rivers.
Analysis of historic climate data shows that precipitation and ET increased 19% and 25% (rates of 2.9 mm and 1.6 mm per year, respectively) between 1931 and 1998 (Claessens et al. submitted). We attribute the discrepancy between expected reduction in ET and actual increase in ET to climate change. We observed a downward trend in minimum temperature beginning in the 1950s as well as an increase in dewpoint temperature and a convergence of minimum and dewpoint temperatures. This pattern is an indication that the near-surface atmosphere has become more humid, a result of increased ET.
Diversion of water for municipal use (drinking water and sewage) has had the greatest impact on hydrologic budgets (Canfield et al. 1999). Diversions today are roughly 20% of annual streamflow. Historically populations outside watershed boundaries have withdrawn water from the Ipswich and Parker rivers for drinking water. More recently water is diverted as sewage to treatment plants outside watershed boundaries. As urban density increases, septic treatment decreases and sewage export increases.
Watershed hydrologic modifications have altered biogeochemical cycles. For instance, N export in sewage is now a major component of the watershed N cycle (Williams et al. submitted a,b). The increased importance of surface flow in urban vs. forested areas leads to reduced contact with “biogeochemically active sites” and a greater input of constituents to streams (Pellerin et al. in prep, Wollheim et al. submitted). We also find that N retention declines in urban areas due to the increased runoff associated with impervious surfaces (Wollheim et al. submitted.
Estuarine hydrodynamics and characteristic mixing times are the result of river runoff, location of river runoff, estuarine geometry and tidal mixing (Vallino and Hopkinson 1998). Total average annual river discharge from the 610-km2 watershed is 11 m3 s-1, ranging from
As a result of seasonal variations in discharge, there are strong salinity gradients within the estuary that vary over time. During high flow, salinity drops to 0 ppt in the upper 10 km while during low flow salinity can exceed 15 ppt at the very head of the estuary. Likewise there are strong gradients in water residence times along the length of the estuary that also vary in response to river discharge. During low flow (
Additional factors controlling estuarine hydrology and hydrodynamics are local precipitation (especially during summer when ET is high and river discharge is low) and sea level variation. Mean sea level exhibits strong lunar, seasonal and annual cycles and variability. The effect of sea level variation is to alter tidal excursion lengths and marsh flooding depth and frequency. Direct precipitation becomes an increasingly important factor controlling salinity distribution, especially on the intertidal marsh in periods of low river discharge and low mean sea level when marsh flooding is limited or absent.
The productivity of intertidal marsh plants is strongly related to variations in salinity during the growing season (Morris and Haskin 1990, Morris 2000, Morris et al. 2002). The uptake of nitrogen by salt marsh vegetation is highly inefficient, varies with salinity and MSL, and contributes to the leaky behavior of the saltmarsh nitrogen cycle. The salt gradient moves in response to changes in freshwater discharge and sea level, and this can affect the export and import of ammonium in marshes along the salinity gradient (Koop-Jakobsen and Giblin 2002). As salinity declines upriver, species composition changes (Odum 1988) and the plant community shift to one limited by phosphorus (Daoust and Morris 2003).
Increased sea level rise is likely to promote the migration of salt marshes upriver and to reduce the extent of tidal fresh and brackish marshes. We have observed substantial marsh disintegration over the past 50 years at the mouth of the estuary, due to a combination of lateral erosion and marsh ponding (Cavatorta et al. 2003), which we associate with the long-term increase in sea level at PIE and reduced sediment loads as a consequence of reforestation of the watershed following abandonment of agriculture in New England. Disintegration results in an increase in total length of the marsh-water interface (Johnston et al. 2003) where there is considerable drainage of marsh porewater. Thus, increased sea level rise is likely to increase porewater drainage which is a significant source of inorganic and organic nutrients for the planktonic sub-system (Wright et al. 1987, Raymond and Hopkinson 2003).
Variations in water residence time along the estuary play a major role in phytoplankton bloom occurrence (Holmes et al. 2000), the relative importance of pelagic and benthic primary production (Hughes et al. 2000), and bacterial community structure (Crump et al. in press). Phytoplankton blooms only occur in the oligohaline part of the estuary during midsummer when river discharge is low and residence time is greater than a week. Blooms also occur offshore during spring and advect into the estuary. Shifts in bacterioplankton community composition along the salinity gradient are related to residence time and bacterial community doubling time in spring, summer and fall seasons. Freshwater and marine populations advected into the estuary represent a large fraction of the bacterioplankton community in all seasons. However, a unique estuarine community forms at intermediate salinities in summer and fall when bacteria doubling time is much shorter than water residence time (Crump et al. in press). The mid-estuary is a region of high heterotrophic activity, with O2 levels occasionally less than 50% of saturation. This reflects large inputs of organic matter substrates from adjacent intertidal marshes (Raymond and Hopkinson, 2003) that are linked via fluctuations in mean sea level and marsh flooding frequency (Wright et al. 1987). Organic matter inputs from marsh porewater drainage are likely to be 10x greater than those from the rivers.
Watershed discharge-related spatial and temporal patterns of salinity are important controls of benthic N dynamics (Giblin et al. in prep, Weston et al. in prep) and hence productivity of the overlying water. In spring, when river discharge is high, salinity in the porewaters is low and the majority of the N remineralized in sediments is either held on exchange surfaces, or lost via denitrification. Salinity increases during the summer as discharge decreases and NH4+ is displaced from the exchange complex leading to a large benthic flux of NH4+ to the water column. This flux appears to be the major source of N supporting the mid-summer phytoplankton bloom in the oligohaline portion of the estuary. In addition, N2 losses decrease during summer as both coupled and direct denitrification rates reach minimal values in spite of high N mineralization rates. N2 losses are suppressed because nitrifiers appear to be unable to adapt to the rapid seasonal change in salinity, shutting off coupled nitrification / denitrification (Mondrup 2000) and because during mid-summer the process of dissimilatory nitrate reduction to ammonium (DNRA) appears to out compete denitrification for NO3- from the water column and water column NO3- concentration is low. In autumn, when the porewater salinities decrease, denitrification rates increase and NH4+ fluxes decrease. Preliminary work in intertidal vegetated sediments has also shown the importance of salinity in modulating the release of NH4+ (Koop-Jakobsen 2003) and measurements of denitrification are underway.
ESTUARINE HIGHER TROPHIC LEVELS
We have also found variations in river discharge and sea level to play a large role in the production of higher trophic levels. Abundance of the dominant marsh fish, mummichog, is positively related to the amount of flooded marsh in creek watersheds (Komorow et al. 1999). Factors affecting habitat quality for fish include marsh area, marsh edge length, flooding frequency, depth and duration and salinity (Haas and Deegan, in prep.). Each of these factors is directly related to sea level and/or river discharge. The species present, as well as their abundance, are related to water residence time, which sets a template that determines the pathways and fate of nitrogen in estuarine systems (Hughes et al. 2000, Holmes et al. 2000, Tobias et al. 2003a, b, Hughes et al., in prep.). When residence time is long, phytoplankton dominate nitrogen uptake and the food web has well developed benthic and pelagic communities with strong benthic-pelagic coupling controlled by the animal community. When water residence time is short, benthic microalgae are the dominant primary producers and the principal food chain is benthic.
N export from watersheds is increasing as urbanization proceeds. The timing of N export is largely driven by variations in river discharge. Climate and land use change, as well as water diversions, are causing a greater percentage of river N export to occur earlier in the spring. Interestingly the timing of maximum N export from the watershed is also the time of minimum estuarine residence time and suggesting that the majority of N exported from watersheds is passed directly to the coastal ocean without significant estuarine processing. The greatest N processing occurs in midsummer, when residence time is longest, but when N imports are least. We presently do not know how much of the N during high flow is retained in the estuary and released later in the year. Nor do we know what the potential mechanisms of N retention might be during the high discharge period. Interestingly, the N exported from many watersheds during late winter/early spring fuels a major plankton bloom in coastal waters of the Gulf of Maine, some of which is tidally mixed into estuarine waters. Thus high algal biomass in the lower Plum Island estuary is not produced locally but externally by N not retained during estuarine passage and by oceanic N sources. The major effect of the watershed on this estuarine system appears to be the influence of freshwater discharge on the spatial and temporal pattern of water residence time. Residence time and estuarine geomorphology define the template upon which ecological processes and foodwebs operate.