Bruce Herbold: The Value of Wetlands to Fish

Bruce Herbold began by telling a personal story. “I have some friends in Austria, and the dearest of them, her father, when he retired, took the family’s fortune and retired out into the countryside where he built a very large house,” he said. “And after he’d been living out there for awhile, he died. So her inheritance was a completely inappropriately sized house in an absolutely awful location, and it was pretty thoroughly worthless.”

Science conference — Click here for more coverage of the Bay Delta Science Conference.

Last March, Mr. Herbold was one author of ten on a short paper that appeared in the online journal, and he’d heard that the paper had been misinterpreted as suggesting there was no value in aquatic habitat restoration to fish. So when he was asked to give this presentation, he readily agreed. “That’s so totally wrong,” he said. “So that’s where I want to go. Yes, there is value there, but its location, location, location – and what you build there.”

He then presented a historical map of California, noting that he would be drawing on some of Robin Grossinger’s work on the historical ecology of the estuary. Mr. Herbold noted that the lake in the picture no longer exists but was a massive part of the Pacific Flyway, as well as the wetlands around Suisun Bay and the Delta. “Notice the dendritic nature of the rivers,” he said. “They actually look like rivers – they look like a Delta. They look a Delta going the wrong way, but they do look like a Delta or two, and it was embedded in wetlands. People who came here were impressed. We talk about how low productivity this estuary is, but the people who came here had an entirely different view of it. That it was richly aquatically diverse, filled with fish of all sorts, and highly productive. The dendritic nature of those channels and the long residence time was a perfect cooker for productivity.”

Mr. Herbold said that some of the great work that Robin Grossinger did was to identify some of the location issues on a broad scale. “If you are in the north Delta, you’re dealing with a very floodplainish sort of river that used to have lots of pans and little lakes in it that would stay there for long periods of time; it could be perennial habitat for some fish,” he said. “As opposed to the San Joaquin, which was very much a braided, wandering river channel coming down and being fed much more by cold water because those mountains were higher, so it was colder water for a longer period of time. It was much more riverine throughout. They both come together in a very tidal system in the middle.”

The middle part is what we think of as the Delta because that’s the only part that remains, he said. “We no longer have the San Joaquin coming in bringing snowmelt, and we don’t’ have floodplain very much, especially not this year, but we still do have tides as the driver down there,” he said. “So you cannot expect to do some things in the north Delta that are perfectly reasonable to do in the south Delta, or would be reasonable if you had other conditions fixed.”

He also pointed out the vast acreages: “120,000 acres down south, 360,000 acres up north, and 300,000 acres in the middle,” he said. “On that scale, the grandiose plans of 65,000 acres of restoration become quite small.”

He then presented a slide of the wetlands in the Bay and the Delta, and noted that historically, the Delta was wetland. “Just comparing it with the Bay, we have maintained and restored some wetlands in the Bay, and the percentage of restoration in the Bay is going to be huge because there weren’t that much, relative to open water. It’s always been a Bay down there, whereas up in the Delta, it was a mud pie with a crust of water.”

“Now a lot of the rivers are diverted before they get to the Delta, or the rivers are used as water conveyance structures, and their value as habitat has greatly decreased,” he said. “All of them are connected so that as tides move in, there are no isolated little bits that can serve as cookers for productivity, so they all move around. With that, anything that goes wrong in one area is rapidly conveyed everywhere else, so if there is a catastrophe in one section, it’s a catastrophe for everywhere.”

Mr. Herbold said he has been involved in a project with other scientists to develop a tidal wetlands monitoring framework. “Since we don’t have a lot of tidal wetlands to study, it will be very important to see how they work as we build them,” he said. “If all the talk is about adaptive management, you have to be studying that. It also implies that you want to be able to compare what happens in one place with something that is in some way comparable with what you do somewhere else. So an overall framework that says what is important to monitor, how we’re going to monitor it, and in these areas it may be more important, and in these areas it may not apply at all – that kind of overall framework as we go into this era of restoration, I think is essential.”

The Tidal Wetlands Monitoring Project Work Team has been a lot of fun, he said. “There’s a bunch of really smart young people that have been hired recently or coming from the university, and a bunch of us old people who are trying to rein them in, and it has been great fun. We’ve had about 40 people show up for the project work team and general meetings. I encourage anybody who is interested to come into it. It is also not confrontational except at the intellectual level and so it is fun.”

Mr. Herbold then said he would be sharing some of the conceptual models that have been developed so far, along with some of the monitoring emphases and priorities. They are based on the DRERIP models, and he acknowledged they could be a little off-putting. “They contain a lot of information in their visual representation, so here we’ve tried to go for clarity of visual representation and put a lot of the details in the text behind that,” he said.

The conceptual models all start at the bottom tier with the landscape attributes, he explained. “Where is the site, relative to the ocean and relative to the midline of the channel as those are two main ecoclines that will drive what you can expect to have on a site, and start to define that location, location, location. Then you look at transport – what’s moving the materials onto and off of that site, how can you expect salmon to get there, or what will sediment do, and so we have a separate model about transport.”

“Season is a particular soapbox of mine, and so how seasonal differences show up on your restored sites becomes important to my mind and I’m starting to convince others,” he said. “Then what was the land use before you inundated it, what diversions or outfalls that are affecting the water that is inundating it, all of those are landscape level attributes that will control what you can expect.”

Then there’s a second tier of local attributes, he said. “What’s flowing down from the watershed, so you get a terrestrial interface, and that’s bringing things in, including some local water, and affecting local water quality, so if you have a lot of agriculture in the area, you may be getting a lot of contaminants you may be worried about.”

And then there are the physical attributes of the site. “You have the aquatic interface, you have a position on the ecocline, and you have a transport model, but what’s coming on to the site, what’s driving the water quality there, what are the physical conditions driving productivity, and what’s the sediment coming, and that then gives you the physical nature of your site,” he said. “We expect the wetland to evolve. As it acquires sediment, as vegetation grows, we expect it to change. It’s not going to be ‘stir the water in and have instant wetland.’”

The physical attributes of the site will drive the biological attributes, he said. “We then have an aquatic vegetation model of what will grow there, and a food web model of what to expect as productivity that’s not aquatic vegetation. Then a sub-model about clams, because we’re all worried about clams, and all of that feeds into the two things that this project work team is focused on: salmon and smelt. And so we have two separate models for them.”

He then presented the transport model. “Where you are on the ecocline will determine whether your site is being driven more by tides or by riverine flow; diversion impacts may affect that riverine flow, and as you move up through these, you get a different meaning of the tiers,” he said. “How well your source site is linked in and what are the materials that you are concerned about, are you concerned about movement of fish or are you concerned about movement of contaminants, or are you concerned about movement of sediment – all those are things we need to have a handle on if we’re going to predict how much is going to go through the intervening material, that connection, and end up on your target site. That will all be determined by how well connected your target site is. You design a breach pattern if you want it to be easily invaded; you design a different breach pattern if you want it to be a more isolated pond. How big it is will of course determine how much likelihood a given particle will arrive there … so this starts identifying what you need to be monitoring, which is the focus of our team.”

He then presented the transport model in a different format, noting that here it is like the tiers of a layer cake where the gray layer is the bottom tier. “I want to draw your attention to the fact that for wetlands, one of the most interesting things is in the winter-spring, the smelt and smolts show up, but not much growth is going on. In the summer and fall, you have all that insulation and you have all this growth, all that’s happening then, so there’s a disconnect there between those two seasons, so how do you get that productivity onsite to be accessible to your smelts and smolts. We’ve tried to capture then what happens as you go from the seasons on those different tiers. Then that has links to the food model and the evolution model.”

He then put up a slide of the Tidal Wetland Restoration Evolution Model. “The same tiers are at the bottom,” he said. “We start identifying that we need to know that is in common amongst all the sites, and as you start identifying sediment supplies, what the water movements onto and off the site are, you start getting a picture of how much sediment applies and what kinds of aquatic vegetation grow there. That will determine then the rapidity of your peat accumulation and your sediment accumulation, and that will determine what kind of wetland you have.”

He then put up a slide of the Aquatic Vegetation Conceptual Model for Tidal Wetland Restoration. “The aquatic vegetation model is going to determine the evolution on site,” he said. “What you then have to look at, if you’re going to predict what aquatic vegetation is going to be then, then you need to know the substrate depth, you want to be looking at salinity, how much variable that salinity is, and all of that feeds into the aquatic vegetation that you see onsite. That then feeds into the food web model, the salmon smolt model, and of course as we just saw, the wetland evolution model.”

He then presented a slide of the food web model, noting that it’s the cleanest smooth food web model he’s ever seen. “It does contain within it the little clam model which is mostly about residence time in water and water depth because the deeper the water is, the less of it gets siphoned through the clams, and the longer it sits there, the more food it can grow. There’s a balance between those two things which might then allow you to produce something that’s not simply clam food.”

“This is the MAST smelt model, I’ve modified it a bit for clarity … but it’s interesting here that food, predation, temperature, entrainment, and toxicity are the things that affect the fish all year around. In the summer, we start worrying about harmful algal blooms, and as they start maturing, we worry about the size and location of the salinity zone, and all the time, we’re worried about the losses from other things and how much food they are getting in.”

“This structure of model I really want to sell because I think it is a good structural framework,” he said, noting that there are also salmon models and smelt models as well.

“So I’d like to take that transport model and put it into the context of how, if we want to build wetlands to support productivity in the low salinity zone, and in this context, it starts looking a little bit iffy,” said Mr. Herbold. “Let’s talk phytoplankton and zooplankton. Well, the longer it sits there, the more it will grow; the shorter it sits there, the less it will grow but the faster you move that off, so there’s a balance between productivity and movement off, which comes out to get that sort of increase and then it falls off when you no longer have water onsite long enough to grow anything.”

“If you make some assumptions about what that productivity is – some really generous assumptions, a large area of marsh that’s 2 meters deep that’s fairly productive, you assume the clams aren’t having any impact, and that nothing else is eating that productivity – it’s not moving downstream very much. You maybe are increasing phytoplankton in the channel nearby by 5%,” he said. “If you do the same opportunity with copepods, and you still make very generous assumptions about productivity, most of them would be eaten on site, but if we just assume they’re getting washed off, there’s maybe an increase of 3% on the adjacent deep channels, so that’s not a function that seems very useful.”

So what is useful? Mr. Herbold then briefly discussed three restoration projects: Lower Yolo Restoration, Prospect Island, and Dutch Slough, noting that they are all different.

Lower Yolo Ranch: “Lower Yolo Ranch was initially viewed as being a tidal system, but it’s up in the floodplain section, and it is in fact above the tidal reach right now,” he said. “It’s always had a pond in it. That pond was there in 1910; that pond is still there, although it’s now agricultural land and has been plowed under. You can still see the same structure, because it is formed by major processes of an alluvial flow coming down the hill and a floodplain flow going the other, and it causes a little pond to form there. If you dig that pond out as was originally proposed, it will back in, guaranteed. So that probably wasn’t a good way to design your site, so they have gone in and changed that site into something much more inline with the broad topography.”

Prospect Island: “Stuart and I were really involved in looking at Prospect Island, where we took various configurations of breaches to see what that did to residence time and what that would allow to grow. It was a really productive two days and we went a long ways in those two days on designing what would be the right thing to do with Prospect Island in this context of what are the functions you’re trying to control and what controls them.”

Dutch Slough: “Finally, Dutch Slough is down in that tidal area, and you have those tidal meandering channels an entirely different thing, so acre by acre, not so much. Location, location, location, yes, and it can do a lot for your fish.”

“I’d like to briefly say, because this is my soapbox, these models also allow you to start looking out for future changes,” Mr. Herbold said. “Sometimes, somebody going to start saying, maybe we will have zebra or quagga mussels in the Central Valley and maybe that will affect our ecosystems; we have a structure that would allow you to look at that. We’re also likely to have some more open water than we have now, and we might have even more open water than that, and we need to be planning for that future. And that is my end. Thank you.”