The habitat surrounding the future location of the Lookout Slough Tidal Restoration Project, located in the Cache Slough complex. Photo by Florence Low / DWR

SCIENCE FEATURE: Structured Decision Making for Delta Smelt Habitat: Synthesizing Multiple Streams of IEP Data to Inform Management

Our native species face numerous threats, including climate change, habitat loss, and invasive species.  Therefore, effective management strategies must be implemented to avoid extinction and promote recovery.  However, making decisions about managing species can be challenging for many reasons, such as high uncertainty, limited data, and multiple stakeholders with different interests.

Structured decision-making (SDM) is a process that can aid in species management by providing a framework for decision-making that is transparent, inclusive, and evidence-based.  At the 2023 Interagency Ecological Program workshop, Dr. Brittany Davis and Dr. Rosemary Hartman with the Department of Water Resources gave a presentation on how structured decision-making was used to make decisions about summer-fall habitat actions for Delta smelt, a listed species, in 2022.

The summer-fall habitat action is part of the current regulatory framework for the long-term operations of the Central Valley Project and State Water Project.  It is a conservation mitigation action intended to improve key physical habitat attributes and access to food for Delta smelt during summer and fall periods when habitat and food might be limited.  It is thought that increasing access to good habitat could potentially improve the species’ growth, survival, and ultimate recruitment.

The goals of the summer-fall action are to maintain low salinity habitat in Suisun Marsh and Grizzly Bay, manage the low salinity zone overlap with turbid water and food, and establish a connected low salinity habitat from the Cache Slough complex through Suisun Marsh.

To meet those goals, a suite of management actions were proposed for habitat and food subsidy actions.  Habitat actions include Fall X2, reoperating the Suisun Marsh salinity control gates to maintain low salinity habitat in the marsh, and a 100,000 acre-foot block of water during summer and fall in some years.  Food web actions, which currently are experimental and optional, include the North Delta food subsidy action, the Sacramento Deepwater Ship Channel Action, and managed wetlands in Suisun.

Planning the summer-fall habitat action

The Delta Coordination Group (or DCG), consisting of federal, state, and public water agencies, plans the summer-fall habitat action.  It is an annual process; each year, the Group evaluates the water conditions, water year forecasts, and Delta smelt distribution status and conditions and recommends the summer-fall habitat action through a systematic, structured decision-making approach.  This approach includes developing the decision frame, desired objectives, and performance metrics that result in a completed consequence table for evaluating the trade-offs for the various actions.

In 2022, the decision was, what suite of actions (both habitat and food) should the DCG recommend for the next summer-fall habitat period (June through October), given the likely water year types?  The likely water year types were forecasted to be below normal or dry years, so that was what was evaluated.

In the dry year, the Group evaluated three alternatives, including a no-action and two potential North Delta food subsidy actions using agricultural return drainage in the Calusa basin drain that would be redirected through the Yolo Bypass to create a food subsidy downstream in Cache Slough complex.  For the below-normal year, they considered a no-action alternative, reoperation of the Suisun Marsh salinity control gates with two operational triggers, a North Delta food subsidy action with two options, and food subsidy actions by redirecting Sacramento River water through the bypass to the Cache Slough complex with multiple alternatives at varied intensities and durations.

The different options for food subsidy actions look at questions such as, do you create a big pulse that’s high intensity and blast it through in a couple of weeks, or do you do a longer duration with a low intensity which increases residency time?  Does that mean more productivity and food?  What if there is a summer Sacramento River action and a fall agricultural action to restore more natural conditions?

Objectives and performance metrics

Based on the objectives, DCG outlined the directional objectives for the SDM model, including smelt food, habitat, growth, and survival, as well as costs, and the potential unintended effects of actions on contaminants or other species such as salmon.  Performance metrics and measures were determined for each objective.

For example, for the objective of increasing Delta smelt food, the performance metric would measure zooplankton biomass.  For the objective of increasing habitat for the Delta smelt, a habitat suitability index is used that considers things such as temperature, turbidity, salinity, and speed.  To measure Delta smelt growth and survival, a bioenergetics model is used to determine how much they grow throughout the summer.

Another objective is to minimize resource costs – don’t want to spend too much water or too much dollars in the process.  It’s important to minimize contaminants; the North Delta Flow action is known to potentially increase contaminants; there isn’t a lot of data on that, so they asked experts what they thought would happen (also known as expert elicitation).  Similarly, they didn’t want to make things better for Delta smelt but worse for salmon, so they asked salmon experts what they thought.

Once the DCG settled on the final performance metrics, the SDM technical team compiled a lot of abiotic and biological data and completed many complicated models that fed into each other.  While quantitative data was preferred, expert elicitations for contaminants and salmonid impacts were conducted in 2022.

Temperature and turbidity data came from IEP’s long-term monitoring data; zooplankton was from IEP monitoring data and integrated using a package developed by the zooplankton synthesis team.  Contaminant data came from a lot of special studies that were in the IEP’s work plan.  Expert elicitation leveraged the salmon and contaminant project work teams.

DWR provided the hydrodynamic modeling used to predict the salinity at Belden’s Landing for the different combinations of salinity control gate operations, as well as the flow in the Yolo bypass under the different scenarios proposed.  Once the salinity forecasts were determined, they used other models to look at the potential change in zooplankton.  DSM2 and Schism were used to calculate salinity in Suisun Marsh and the length of the North Delta food subsidy flow pulse for different combinations of water year type and action.

For habitat, the outputs of the DSM 2 models were then put into a three-dimensional hydrodynamic model and used turbidity and temperature data from similar years to forecast the amount of suitable Delta smelt habitat.  They then took the outputs of the habitat and food models and put that into yet another model to determine if the food was increased and salinity, turbidity, and temperature are changed, what will the likely change in Delta smelt throughout the summer be?

The consequence table

All of these metrics, along with expert elicitation on impacts of contaminants, resource costs, and impacts to salmonids, were put into an online application that allowed the decision makers to see how each scenario ranked.

The slide shows the consequence table for the three Suisun Marsh salinity control gate scenarios in a below-normal year:  no action and operation at two different salinity levels.  Each row is a performance metric, such as habitat suitability index or incremental growth in Delta smelt, that allows different actions to be compared.

The decision is made

The DCG management decision-making team then convened to look at the modeling and the consequences table.  There was much discussion and some arguing, but in the end, with the help of facilitators, the team was able to come up with a decision.

Since the decisions were made in March before it was known exactly what the water year would be, there were different actions depending on what hydrology developed.

“If it was a critical year, there’ll be no action as water quality in the Yolo bypass would likely not be good,” said Dr. Hartman.  “If it was dry, we would do the North Delta food subsidy action with agricultural water of long duration, low magnitude, and maximize opportunities for learning.  If it was below normal, we would do a North Delta food subsidy action with Sacramento River water with low long duration, low magnitude and operate the salinity control gates at four parts per 1000.”

“We were very proud of getting to this decision,” she continued.  “It’s a lot of work and a lot of trade-offs.  And then it was critically dry, so we couldn’t actually do an action.  But this did give us a chance to learn about the process, how to step through it as Group, and feel a lot more prepared for a year where we would have an action.”

Dr. Hartman said it is an adaptive management process.  Even though they did not do an action last year, they still monitored to see what would happen and reported on the outcomes.

The process starts again

The DCG is working through the structured decision-making process again.  They aren’t working through the salinity control gates as the operation is specified in the regulation, so they will operate them per the rules in the ITP.

“We still are doing the North Delta food subsidy action with structured decision-making, and we’ve added learning as a specific objective,” said Dr. Hartman.  “We’re also adding more special studies of stable isotopes on the Yolo Bypass to better learn from the action if we do it this year.  And we have a new study looking at managed wetlands in Suisun Marsh to hopefully use that as one of our actions in future years.”

QUESTION:  Since you have plans for how you will act based on the potential outcomes, what are you looking for from the stable isotope studies?  And then what actions might be taken?

“Some of the pilot data looks really promising for the stable isotopes, carbon signatures, and distinct gradient across the Yolo longitudinal axes, so we’re hoping that in an action, we can use this to determine the mechanisms of transport,” said Dr. Davis.  “If we’re able to trace upstream food subsidies moving downstream into the Cache complex or potentially in the diets of Delta smelt, then it’ll help us evaluate the efficacy of the action in future adaptive management and action as a whole – whether it’s Sacramento River action or ag action, how we’re moving water and food.”

“It’s that adaptive management loop because if we have some evidence thinking that a flow pulse to the bypass will increase food for smelt downstream,” said Dr. Hartman.  “But then there are other situations where it’s not as clear – it could just be a coincidence that we saw more phytoplankton downstream, so we want to get a better handle on it and using the stable isotopes to trace things down, I think will help.”

Print Friendly, PDF & Email