To determine the impacts of restoration on fish populations–particularly on Chinook salmon–scientists measured the opportunities salmon have to access and use habitat types. Opportunity was measured through the density and timing of salmon usage of the restored and reference habitats, determined through fyke trapping. Research began in 2004 and continued through 2014.
Measuring the opportunity for fish to access suitable habitat meant that scientists examined physical attributes of the river delta. These attributes were channel area, channel depth, water temperature, temporal variation in channel availability and landscape connectivity. Fyke trapping also allowed researchers to catch Chinook and chum salmon in each habitat location, which provides information on the number of fish present at different times of the year.
To monitor the physical characteristics, researchers from the Nisqually Indian Tribe and USGS used aerial imagery, physical measurements and inundation modeling. Coupled with the fyke trapping results, scientists were able to track increased opportunities due to broad-scale changes in available habitat on the Nisqually River Delta.
The studies primarily took place at two reference locations (Red Salmon Reference and Animal Slough) as well as at two restoration sites (Phase II, which was restored in 2006, and Madrone Slough, which was a part of the historic 2009 restoration).
Note: in the map below, Animal Slough is referred to as “Nisqually Reference”, Madrone Slough is referred to as “2009 Restored” and Phase II is referred to as “2006 Restored.”
Based on historic data and aerial images, researchers were able to develop a digital map of pre- and post-restoration channels across the river delta. Historically (before the 1800s), there were several sloughs and channels bringing tidally influenced water into the Nisqually Delta, and providing habitat for fish and birds. With the construction of the dikes, the number of channels was drastically reduced–McAllister Creek and the Nisqually River were among the only tidally influenced sources of water. As of 2011, however, the number of channels in the delta had greatly increased. (See Figure 1).
In addition, the researchers collected channel cross-section data annually between 2009 and 2014. Each channel sampled had two sampling sites: one located closer to the delta, and one further inland. This was to determine how quickly channels were eroding–thus becoming deeper–across the estuary.
Major channels, or those over 15 feet wide, showed drastic changes in channel area, length and edge when compared to 2005 conditions. As of 2011, major channel area had increased 42% delta-wide, and 580% within the restoration area. Major channel length increased 131% delta-wide and 579% within restored areas, while major channel edge increased 126% delta-wide and 490% in restored areas. In addition, minor channels (those under 15 feet wide) increased in area, length and edge by 214% delta-wide and 534% in restored areas.
With the reintroduction of tides, the restored channels grew deeper as sediment eroded away, while the reference sites remained relatively constant. This trends was particularly noticeable at sites on the west side of the Nisqually River; those on the east side experienced less change. At sites west of the river, changes in channel depth were related to two variables: restoration channel and offshore distance. In other words, seaward and midland channel cross sections grew deeper as time went on, while upstream channels remained more stable.
Figure 2 shows changes in channel elevation (depth) through time, from 2009 to 2014. It also compares elevations to those of the reference elevation. The sites west of the river, and those that were seaward, have shown the most erosion over time, and are approaching reference elevations. Graphs of individual channel cross sections are available on the “Geomorphology and Sedimentation” page.
To monitor changes in water quality, researchers installed Solinst LTC Data Loggers at 8 sites across the river delta from 2009-2014. These are marked as stars in Figure 1, above. The loggers collected water level (m), temperature (°C), and specific conductance (µS/cm) every 6-15 minutes. In addition, researchers collected surface and bottom salinity (ppt), temperature (°C), conductivity (PSU) and dissolved oxygen (mg/L) during each fyke trapping event. The researchers studied four variables:
»Time since restoration
»Distance from offshore (seaward vs. inland)
»Restoration status (restored or reference)
The scientists believed that these 4 variables would impact water quality parameters, like temperature and salinity, even when accounting for changing seasons.
As of 2014, water temperatures grew colder by about 0.157 °C per year across the delta. More specifically, seaward sites decreased the most while temperatures in reference sites stayed stable (Figure 1). On the other hand, inland restored sites, had warmer temperatures and displayed greater fluctuations during the year. The declines in temperature are most likely linked to deepening of restoration channels–deeper channels tend to have colder water temperatures. In addition, sites lacking emergent marsh vegetation displayed higher water temperatures because of a lack of shade. As channels continue to deepen and vegetation cover increases, water temperatures are expected to approach levels similar to reference sites.
In a similar vein, the four predicted variables also impacted channel salinities. Overall, salinity decreased by about 0.098 ppt per year, and indicated that changes in salinity did occur through time but are site-specific. Reference sites had salinity levels 3 ppt greater than restored sites, and displayed greater fluctuations. As of 2014, seaward sites are about 1 ppt lower than reference sites, while inland sites are roughly 4 ppt lower than reference sites. As time goes on, salinity levels are expected to approach values similar to reference sites, as tidal influence has more influence that river outflow and as channel depth increases.
Channel morphology and improved water quality only represent increased opportunity for salmon species if the new habitat types are accessible to the fish. Researchers modeled the connectivity and inundation through bathymetric sonar, topographic (LiDAR) and water level data. Pre-restoration conditions were characterized by using post-restoration data, clipped to the boundaries of historic levees that prevented water circulation.
Inundation was measured at 1-cm intervals, to model inundation at mean lower low water (MLLW), mean low water (MLW), mean tide level (MTL), mean high water (MHW), and mean higher high water (MHHW). Channels that were connected with at least 0.4 meters of water were deemed as preferred juvenile Chinook salmon. After restoration occurred, the frequency of inundation–or the amount of time the Nisqually estuary supported accessible and suitable habitat for Chinook–increased from 30% to 75%. Additionally, all three modeled restoration sites exhibited increased connectivity to the Nisqually River at or above MTL. Only one pathway developed to the 2006 Restored (at MTL) and Phase I (MHW) sites. However, three pathways to the 2009 Restoration Site were available at MHHW.
Beach seining and fyke trap sampling at sites across the delta helped researchers validate hypotheses relating to habitat accessibility. By noting the type of fish at each site, researchers were able to determine if fish were actually using the increased habitats. Beach seining began pre-restoration, with data collected twice a month February-October 2003 to 2006. Results of these data are available on the ‘Fish: Baseline Data’ page of this website. Post-restoration sampling occurred monthly, February-October 2010 to 2014. In addition, fyke trapping occurred about monthly April-July 2003 to 2006, and 2010-2012. The data provided a proportional presence for natural origin Chinook, hatchery origin Chinook and chum salmon.
The sampling data suggests that natural-origin Chinook, hatchery-origin Chinook and natural-origin chum all began using the restored channels in the first year following restoration. In addition, natural-origin salmon were distributed more broadly across the Delta temporally, while abundance of hatchery-origin fish were likely linked to timing of release and residency time. However, proportional presence values for natural-origin Chinook indicated less consistent use of the restoration sites.
Estuaries are an important habitat type for a variety of plant and animal species: tidal influences spread sediment, nutrients, detritus and aquatic organisms, improving habitat for birds and fish. Estuaries that have been altered by humans often starve the ecosystem of the tidal influence, leading to a loss in tidal channels, sediment starvation, marsh erosion and loss of critical habitat. The Nisqually River supports several salmon populations, including a threatened Chinook stock. Chinook salmon are thought to be the most estuary dependent salmon species, and stand to benefit greatly from the Nisqually Delta Restoration Project.
By examining the increased habitat opportunities for use by Chinook and other salmon, researchers are better able to understand the complex processes that occur after a large-scale restoration project. Indeed, the data shows an enormous increase in major and minor tidal channels across the delta along with increased accessibility. Water quality is also changing as a result of deepening channels, proving more favorable for fish. In fact, salmon catch data shows smolts used the restored channels as soon as one year post-restoration.
Overall, this monitoring effort has shown that estuaries can quickly regain useable salmon habitat, despite long-term alterations to the landscape. The increased opportunities for salmon displayed by the Nisqually Delta Restoration Project demonstrate that large-scale restoration projects can improve Puget Sound estuary habitat, and may play a positive role in salmon recovery.
- Ellings, C.S., E. Grossman, M. Davis, S. Hodgson, K. Turner, I. Woo, J. Takekawa, G. Nakai. 2015. Post-Restoration Changes in Salmonid Opportunity [In Revision]. Olympia, Washington.