Response of Larval Transport to Changes in the Landscape at the Williamson River Delta


Animations accompanying this report show the simulation of the tracer representing the total density of suckers (combination of LRS, SNS/KLS, and unidentified suckers) during 70 days in May through July, 2006 through 2009. Results from these simulations include:


• The larvae leave the Williamson River channel in “pulses” at night, the manifestation of the nighttime-only drift along the entire length of the channel.

• In all years, the simulated larval aggregate tends to “hug” the eastern shoreline of Upper Klamath Lake, a manifestation of the clockwise circulation pattern that occurs under prevailing wind conditions from the west to northwest (fig. 4).

• All larvae in the simulation enter Upper Klamath Lake at the mouth of the Williamson River channel in 2006 and 2007, before the restoration at the Delta began. 

• A strong wind reversal occurred in 2006 between May 21 and 24 (fig. 4A). This reversal caused the simulated larval aggregate to be advected northwest from the mouth of the Williamson River. The prevailing clockwise circulation reversed briefly, and the aggregate was advected in a counter-clockwise direction as far as the western shoreline. This reversal also resulted in larvae entering Agency Lake through Agency Straits.

• In 2008, after Tulana was flooded, but Goose Bay was still closed off from the Williamson River channel, the aggregate spread across Tulana and entered both Agency Lake and Upper Klamath Lake through breaks in the levees and over the levees that were submerged during this time.

• In 2009, after Goose Bay was flooded, strong prevailing winds tended to advect the simulated larval aggregate through Goose Bay in preference to Tulana, causing the centroid of the aggregate to enter Upper Klamath Lake through breaks in the levees and over the submerged levees on the south side of Goose Bay, which resulted in many larvae entering Upper Klamath Lake south of the mouth of the old Williamson River channel.


The changes in larval dispersal between 2006 and 2009 also can be seen in the time series of simulated densities at fixed larval catch sites (figs. 5, 7, 8). Simulated densities at site 25981 in Upper Klamath Lake near the mouth of the Williamson River channel, as well as at sites A and B in Tulana near the Williamson River channel (fig. 1), are shown in figure 5. Simulated densities within Tulana are available only for 2008 and 2009, after the restoration of the Delta started and the levees around Tulana were breached. Also shown is the simulated density at a point within the Williamson River channel (WR in fig. 5) about 1 km downstream of the Modoc Point Road boundary. This is not a fixed site for larval catches, but model output at the site is included to provide the larval density values in the channel before larvae enter the Delta. The simulated density at site 25981 decreased as a fraction of the simulated density upstream between 2006 and 2009. Between 2006 and 2007, this decrease is attributable to the higher flows in 2006 (fig. 6), which both decreased the larval travel time to Upper Klamath Lake, thus reducing accrued mortality in the river prior to entrance into the lake, and increased the momentum of the river flow, which pushed the larval aggregate farther into the lake. A large additional decrease in the density at site 25981 relative to the density upstream in 2008 and 2009 is a consequence of flooding first Tulana in 2008 and then Goose Bay in 2009. As a result, most of the larvae dispersed through Tulana in 2008 and the combined area of Tulana and Goose Bay in 2009 before reaching site 25981. The difference in travel time through the Delta between 2008 and 2009 can be seen in the simulated densities at sites A and B. In 2008, the densities at these two sites were similar, whereas in 2009, when Williamson River flows were lower, the greater travel time resulted in more accrued mortality between these two sites, and the simulated density at site B was lower than at site A (fig. 5). 


The time series of the simulated daily maximum density at selected fixed sites along the shoreline in Upper Klamath Lake are shown in figure 7. Site U6 is located on the south side of the Williamson River channel near its mouth. To the east and south of the mouth along the shoreline are sites U5 and U4. To the west and north of the Williamson River mouth are sites 25976 and 25977 (fig. 1). The simulated densities at site U6 and the relation of those densities to the upstream densities in the Williamson River are similar to those at site 25981. Of note in figure 7 is the relation of simulated densities at sites to the northwest and southeast to densities at site U6. In 2006–08, there is a decrease in density from sites U6 to U5 to U4, in the prevailing direction of transport of the larval aggregate along the shoreline, reflecting the increasing travel time along this pathway, and therefore increasing dispersal and mortality along the pathway as well. Sites northwest of the Williamson River mouth are in the direction opposite to the prevailing transport; thus, in 2007, simulated densities at sites 25977 and 25976 were lower than densities at sites U6, U5, and U4. In 2006, simulated densities at site 25976 were comparable to densities at site U5, a consequence of the large wind reversal in 2006 that pushed the aggregate of larvae northward for a few days around May 21. 


After Tulana was flooded, larvae moving into the Delta could exit at two levee breaches along the Upper Klamath Lake shoreline, approximately opposite sites 25977 and 25976. These new pathways resulted in higher concentrations at these two sites relative to the Williamson River channel, than prior to the start of restoration. Of particular note is the change in the simulated densities at site U5 between 2008 and 2009. In 2009, after Goose Bay was flooded, the simulated densities at sites along the Upper Klamath Lake shoreline were highest at site U5, and lower both to the southeast at site U4, and to the northwest at sites U6, 25977, and 25976. The configuration of the landscape prior to 2008 forced virtually all the larvae to exit the Delta at the mouth of the Williamson River, which is manifested as proportionately higher densities at site U6 in those years. When Tulana was flooded, proportionately fewer larvae exited at the old mouth of the Williamson River as larvae were dispersed throughout Tulana and exited through breaches into Upper Klamath Lake and, to a lesser extent, Agency Lake. In 2009, after Goose Bay was flooded, most of the simulated larvae that exited the Delta to Upper Klamath Lake did not do so at the mouth, but through the open levees on the Goose Bay side of the Delta.


Four sites are located on the northern side of the Delta in Agency Lake—sites 25978, 25979, A1, and A2 (fig. 8). Site A1 was outside of the numerical grid used for 2006 and 2007 simulations. Few simulated larvae ended up in Agency Lake prior to the restoration of the Williamson River Delta, even though simulated densities were approximately an order of magnitude higher in 2006 than in 2007 because of the strong wind reversal in 2006 that carried the aggregate of larvae northward from the river mouth. In the 2008 configuration of the landscape, more larvae passed through Tulana and through breaches on the northern levees surrounding Tulana to end up in Agency Lake, even though the peak densities there were lower than in Upper Klamath Lake, even at sites located near the openings in the levees (sites 225978 and 25979). Generally lower densities in Agency Lake in 2009 are a consequence of larval dispersal over the combined area of Tulana and Goose Bay, and the preferential exiting of the larvae on the Upper Klamath Lake side of the Delta under prevailing wind conditions. 


One fixed site (U8, simulated densities not shown) is located on the western side of Upper Klamath Lake. Simulated densities this far from the source at the Williamson River were low; peak densities were 0.18, 0.15, 0.03, and 0.02 fish per cubic meter in 2006 to 2009, respectively. The simulations showed that densities at site U8 decreased markedly in 2008, after the flooding of Tulana, because the larval aggregate was forced farther out into the lake in 2006 and 2007, and more of the larvae were entrained in the clockwise circulation gyre and carried to the western side of the lake. 


First posted April 2, 2012