Testing the effect of tidal channel excavation on vegetation development in tidal marsh restoration sites
In this effort, Skagit River System Cooperative is evaluating the effects of the Restoration of River Delta Channel Structure on Vegetation in the Skagit and Stillaguamish deltas. See also Tidal Channel Restoration Guidelines and Tidal Channel Reference Model.Previous monitoring of the zis a ba(Stillaguamish Tribe) and Fir Island Farm(WDFW) marsh restoration sites suggests extensive channel excavation to match reference conditions has dramatically accelerated vegetation development at zis-a-ba, while insufficient channel excavation has strongly inhibited vegetation colonization at FIF. This project test the hypothesis that poor channel development produces high sheet flow velocities and sheer stress over the marsh surface, thereby inhibiting seed recruitment and germination. This hypothesis will be tested through measurement of sheet flow velocities and seed retention on four restoration sites in the Skagit, Stillaguamish, and Snohomish deltas: zis-a-ba, Fir Island Farm, Leque Island, and Qwuloolt.
Problem Statement In tidal marsh restoration aimed at Chinook salmon recovery, successful vegetation establishment is critical to providing food web support (production of salmon prey) to juvenile salmon and other estuarine fish and wildlife. Healthy vegetation also enables tidal marsh resilience to sea level rise, erosion, and other physical disturbances by binding soils with extensive root systems and by facilitating suspended sediment deposition and marsh accretion when stems and leaves baffle and slow water currents. Additionally, successful vegetation establishment is a visible indication to the public that a restoration project is working, while bare ground or invasion by non-native weeds is a visible indication to the public of restoration failure. Visible failures can increase public distrust of and opposition to habitat restoration efforts. Thus, successful vegetation establishment in tidal marsh restoration projects is biologically and politically important.
Restoration ecology and management require predictive models that can set expectations of restoration actions and guide restoration design, monitoring, and adaptive management. The failure or success of vegetation establishment can only be evaluated in the context of an expectation, a prediction of what the restoration project should have achieved. If vegetation fails to meet expectations, then a remedy must be identified and implemented through adaptive management so that biological and social/political goals can be achieved. Monitoring that simply documents the status of a restoration project without evaluating that status in the context of expectations and goals, and without developing remedies when there are failures, is sub-optimal monitoring.
The Fir Island Farms (FIF) and Qwuloolt tidal marsh restoration sites, and to a lesser degree, Leque Island, are examples of problematic restoration that require diagnosis of the cause of the failure of native vegetation to fully and rapidly colonize the sites, and require elucidation of a remedy to the failure. The problems confronting these sites, failure of vegetation colonization, poor channel development, and an extensive clay hard pan (produced by antecedent agricultural activity [Kashirad et al. 1967; Raper et al. 1990; Chen and Tessier 1997]), are likely interrelated and are not be unique to these particular restoration sites. This constellation of problems has been anecdotally observed in other sites, e.g., the TNC-PSB restoration site near the mouth of Hat Slough in the Stillaguamish Delta, the Potter’s Slough restoration site in the tidal Willapa River, and the Elk River marsh dike breach near Westport. Thus, this may be a more widespread phenomenon than is currently recognized. Understanding the cause of vegetation failure on problematic restoration sites is key to developing a remedy and to improving tidal marsh restoration design and success throughout the region.
Hypothesis Statement Previous monitoring of the zis-a-ba and FIF marsh restoration sites suggests a positive correlation between the extent of tidal channel networks and colonization rates by native tidal marsh vegetation. The FIF restoration site has only one tidal channel outlet, much less total channel length and area than predicted by an allometric model developed from reference marshes (Hood 2019a, 2018, 2007), and low colonization by native vegetation. Six years after restoration, much of the site is either bare or colonized by relatively sparse patches of non-native Cotula coronopifolia. In contrast, the zis-a-ba restoration site has an extensive network of excavated tidal channels that is comparable in extent to those of reference marshes (Hood 2019b, 2018), and it was colonized extensively by the native bulrush, Bolboschoenus maritimus, by the end of the first year after site restoration, an extraordinarily rapid rate of vegetation development on a tidal marsh restoration site.
These contrasting conditions suggest the hypothesis that high channel development conveys most tidal energy through the channel network, resulting in quiescent conditions on the marsh surface that allow seed recruitment and germination, while poor channel development produces high sheet flow velocities and sheer stress over the marsh surface that inhibit seed recruitment and germination.
This hypothesis will be tested through observations of vegetation colonization and measurement of sheet flow velocities and seed retention on four restoration sites in the Skagit, Stillaguamish, and Snohomish deltas: zis-a-ba, FIF, Leque Island, and Qwuloolt. Measurement of sheet flow velocities and seed retention relative to the presence/absence of tidal channels will elucidate the possible mechanism responsible for observed problematic vegetation colonization.
Methods ‘PVM development and vegetation assessment’ For each study site, the existence of a problematic vegetation status must be established before the search for a solution can begin. Vegetation colonization success or failure will be determined by comparing observed restoration site vegetation with predicted vegetation colonization patterns developed from a Predictive Vegetation Model (PVM). PVM development and comparative vegetation assessment has been previously described (Hood 2013, 2019a, b), so only a brief summary is related here. RTK-GPS (3 cm horizontal and vertical resolution) is used to collect point data in the restoration site and nearby reference site on grids with spacing of 25-30 m between points. Within a 1-m radius of each point, the dominant plant species (by canopy cover) is determined by visual estimation, along with the top three sub-dominant species, and the elevation of the point. If no plants are present, then the point is categorized as bare ground. Elevation-frequency distributions for each observed plant species are generated and applied to binned lidar data to generate a GIS map of predicted dominance probability for each species for the restoration site, i.e., a PVM. The reference marsh from which the PVM is generated is generally proximal to the treatment (restoration) site, so that both sites share similar salinity and tidal regimes. The observed restoration site vegetation is then compared to the reference-generated PVM to assess vegetation colonization success or failure. If differences are subtle, rather than obvious, then boot-strapping methods are used for statistical comparison of the PVM and restoration site observations (Hood 2013).
‘Sheet flow velocity’ The effects of tidal channel excavation will be assessed by direct measurement of sheet flow velocities on flood tides. Sites, or portions of sites, with well-developed channel networks are hypothesized to have most tidal energy directed to the channels and less contained in sheet flow over the marsh surface. Sites, or portions of sites, with poorly developed channel networks will have more tidal energy remaining as sheet flow. This energy will be assessed with a velocity meter by point measurements when water depths over the marsh surface are < 20 cm, i.e., when there is the greatest likely sheer stress over the marsh surface. This means that velocity measurements will be constrained to a narrow window of time during each flood tide, so that many site visits will be required to characterize spatial patterns of marsh surface water velocities.
‘Seed retention’ Seed retention plots will measure 20 cm x 20 cm, be marked at the corners with wire survey flags, and be located in area of high vs. low surficial velocity. Seeds of maritime bulrush (Bolboschoenus maritimus) will be collected in the field, sorted and counted, and then a constant number (100) will be placed at each plot. Plots will be revisited each day after initial seed deployment, and the remaining seeds counted until a significant fraction has disappeared or two weeks have passed. If seed erosion follows a negative exponential function, then the half-life of seed retention will be calculated for each plot.
References Chen Y. and S Tessier. 1997. Techniques to diagnose plow and disk pans. Canadian Agricultural Engineering 39: 143-147. Hood WG. 2019a. Fir Island Farms 3rd-year Post-Restoration Tidal Marsh Monitoring Report. Prepared for the Washington Department of Fish and Wildlife. Skagit River System Cooperative. LaConner, WA. Hood WG. 2019b. zis-a-ba 2nd-year Post-Restoration Tidal Marsh Monitoring Report. Prepared for the Stillaguamish Tribe Natural Resources Department. Skagit River System Cooperative. LaConner, WA. Hood WG. 2018. Applying tidal landform scaling to habitat restoration planning, design, and monitoring. Estuarine, Coastal and Shelf Science www.sciencedirect.com/science/article/abs/pii/S0272771417302500. Hood WG. 2013. Applying and testing a predictive vegetation model to management of the invasive cattail, Typha angustifolia, in an oligohaline tidal marsh reveals priority effects caused by non-stationarity. Wetlands Ecology and Management 21:229-242. Hood WG. 2007. Landscape allometry and prediction in estuarine ecology: linking landform scaling to ecological patterns and processes. Estuaries and Coasts 30:895-900. Hood WG. (in revision) Tidal channel network development rates for marsh restoration. Ecological Engineering. Kashirad A, JGA Fiskell, VW Carlisle, CE Hutton. 1967. Tillage pan characterization of selected coastal plain soils. Soil Science Society of America Journal 31:534-541. Raper, RL, LE Asmussen, JB Powell. 1990. Sensing hard pan depth with ground-penetrating radar. Transactions of the ASAE 33: 41-46.
Relationships
- related to: remlinger farm restoration
- related to: restoration
- related to: river delta;hydrodynamics;channel structure;salmon
- related to: skagit delta
- related to: skagit river system cooperative
- related to: stillaguamish delta
- related to: tidal channel reference model
- related to: vegetation
- related to: zis a ba restoration