Final Report: Nonlinear Response of Pacific Northwest Estuaries to Changing Hydroclimatic Conditions: Flood Frequency, Recovery Time and ResilienceEPA Grant Number: R833015
Title: Nonlinear Response of Pacific Northwest Estuaries to Changing Hydroclimatic Conditions: Flood Frequency, Recovery Time and Resilience
Investigators: Wheatcroft, Robert A. , D'Andrea, Anthony F.
Institution: Oregon State University
EPA Project Officer: Hiscock, Michael
Project Period: July 1, 2006 through June 30, 2010
Project Amount: $620,182
RFA: Nonlinear Responses to Global Change in Linked Aquatic and Terrestrial Ecosystems and Effects of Multiple Factors on Terrestrial Ecosystems: A Joint Research Solicitation- EPA, DOE (2005) RFA Text | Recipients Lists
Research Category: Global Climate Change , Ecosystems , Climate Change
A warmer atmosphere holds more water vapor; hence a direct outcome of global climate change is an intensification of the hydrologic cycle. Rainfall intensity, which produces rapid runoff (i.e., floods), is on the rise and sediment yield from Pacific Northwest (PNW) basins has increased due to 20th century land use (e.g., timber harvesting). Consequently, estuarine sediment inputs have increased in magnitude and intensity, with important, but unknown ramifications for the health of, and ecosystem services provided by, PNW estuaries. It is difficult to evaluate the risk to estuarine ecosystems because the few studies to date have tracked only a handful of species and rarely track recovery from flood sedimentation events. Without a community-level understanding of the changes in biodiversity and functional redundancy in the intertidal benthic ecosystem, it is difficult to assess the risk of increasing flood sedimentation to Pacific Northwest estuaries.
Our research had four primary objectives:
1) Design and implement a manipulative field study that simulates different frequencies of flood sedimentation events (no, one, or two events in a single rainy season) to determine the ecological effects of flood sedimentation on intertidal benthic macroinvertebrate communities, the changes to the physical properties of the sediments that may impact the recovery of the benthos, and the susceptibility of sedimentation-stressed systems to increased invasion by non-indigenous species (NIS).
2) Use a combination of high-resolution benthic sampling and multivariate analyses of benthic community metrics to track the initial mortality and recovery of the benthic community. Particular emphasis was placed on identifying changes in functional biodiversity, documenting recovery times, tracking mortality and recovery of important functional groups, and delineating changes to populations of NIS relative to native species.
3) Collect and analyze sediment samples to track changes in important sediment properties (e.g., porosity, grain size, total organic carbon, oxygen penetration depth and profiles, phytopigment profiles, and benthic microalgal photosynthesis and biomass) that have direct or indirect effects on survival or habitat suitability to the benthic macroinvertebrate community.
4) Synthesize the datasets from this study to develop an empirical and theoretical framework for predicting the effects of flood sedimentation events on tideflat macrobenthic communities in PNW estuaries and how these changes impact ecologically and economically important biotic resources and ecosystem services.
The study was conducted between summer 2007 and late fall of 2008 on the extensive intertidal sand flats of Netarts Bay, Oregon (45°24’ N, 123° 56’ W). There, 12 randomly located 9-m2 study plots in the mid-intertidal were assigned to one of 4 treatments: pre‐tests, flood 1, flood 2 and control (3 replicates each). The main experiments began on 14 January 2008 when flood sedimentation events were simulated by depositing ~800 kg of fine-grained terrestrial sediment on each of the flood 1 and flood 2 plots (n = 6); 35 days later, a second flood bed was deposited on flood 2 plots (n = 3). Flood layers were consistently 2-4 cm thick. Replicate samples for macroinvertebrates and various ancillary parameters (listed above) were collected at 12 different sampling times (from days -2 to +324) following the initial flood deposition event. Sampling the main experiment required approximately 75 days of fieldwork by 3-5 personnel, whereas each deposition event required ~12 volunteers. Total person days of fieldwork is estimated to be > 350 days.
The simulated flood deposits had an immediate and persistent impact on the sedimentary environment and its resident macroinvertebrate community. In particular, the grain size, porosity, organic carbon content and C:N ratio of the upper portion of the seabed in the flood plots differed substantially from those in the controls. Of particular note was the persistence of the flood deposits despite strong ebb tide flows (> 40 cm/s at 50 cm above the bed) that under normal circumstances would be sufficient to erode muddy sediment. That erosion did not occur was likely due to rapid dewatering of the deposits as they were subaerially exposed during each tidal cycle coupled with the absence of macroinfaunal bioturbation that typically increases water content and hence erodibility.
The persistence of the flood sediment meant that the biogeochemical characteristics of the seabed were substantially different in the flood plots compared to the controls. For example, C:N ratios were higher in the flood sediments until day 111 when it became evident that benthic microalgae (primarily pennate diatoms) had begun to recolonize the flood plots. The microalgal recolonization altered the oxygen profiles within the flood beds (greater oxygen penetration than previous flood oxygen profiles) and increased chlorophyll-a of the surficial sediments. This recovery of primary productivity only occurred when coarse ambient sediment that surrounded the flood plots had been transported over the muddy flood deposits. The observed substantive impacts on the benthic macroinvertebrate community were not surprising given the drastic physical and biogeochemical changes that occurred in the flood beds. The response to the disturbance, however, was variable and complex. For example, there was clear evidence for initial mortality for some of the community dominants (e.g., Leptochelia dubia, a tube-building cryptic tanaid crustacean & Eobrolgus chumashi, a phoxocephalid amphipod crustacean). Whereas in other cases, there was low abundance in the flood plots and controls during the wintertime, but preferential recolonization of the flood plots occurred during the late spring recruitment period (e.g., Mya arenaria (a bivalve mollusk) and spionid polychaete annelids). In contrast, other species (e.g., corophid amphipod crustaceans) exhibited decreased recruitment rates in the flood plots during the spring and summer (days 111-256). Ongoing analyses are being conducted to ascertain whether there was an increase in NIS due to the flood deposition events.
The idiosyncratic species-specific response that we observed was similar to findings of manipulative flood-deposition studies in New Zealand estuaries conducted by National Institute of Water and Atmospheric Research (NIWA) scientists. In those studies, a storm midway through the experiment dispersed the majority of flood sediment and facilitated recovery to pre-experiment abundances and species composition. In this study, an extraordinary winter storm occurred roughly six weeks before the start of our experiment. Although Netarts Bay is not affected by natural flood-driven depositional events because it lacks large rivers; high winds and currents do impact the bay, and these must have been significant given the > 85-knot winds recorded in the area. The macroinvertebrate community was therefore likely in a recovery state at the time of our experimental start. The last sampling point was unfortunately collected prior to any storms the following fall (there is considerable interannual variability in storm frequency and intensity in the PNW) and therefore the control and flood treatment communities, which no longer demonstrated significant differences in abundances, did not return to the original low species abundances observed at the start of our experiment. Although an idiosyncratic response is in some ways frustrating, the juxtaposition of so-called "dry storms" (i.e., the early December 2007 event) with flood-producing events is often observed in the PNW and will likely increase in frequency under future climate change.
Overall, this study met its stated objectives. First, we designed a statistically rigorous manipulative field experiment and implemented it over a 16-month period in a challenging and dynamic field setting -- the intertidal of Netarts Bay, Oregon. Second, we collected a large and comprehensive data set on organism abundance, species composition and functional groupings that document the macroinvertebrate (both native and non‐indigenous species) response to our simulated flood-deposition events. Third, a parallel set of data was collected that elucidated the physical and biogeochemical changes to the seabed in response to and following the flood deposition events. Fourth, we are working toward synthesizing the biological and abiotic data sets to develop an enhanced understanding of the response of Pacific Northwest estuaries to flood sedimentation events. These results are being (and will continue to be) disseminated through peer-reviewed publications.