Delta Independent Science Board member Dr. Joy Zedler shares her insights on the ups and downs of coastal wetland restoration and its implications for restoration in the Delta
Habitat restoration is a complex process, full of uncertainties and difficult to get right. In this brown bag seminar, Delta Independent Science Board member Dr. Joy Zedler shares her research and insights on restoring coastal wetlands and improving habitat for endangered species in San Diego Bay and salt marsh ecosystems in the Tijuana Estuary. Her work on evaluating efforts to mitigate impacts of freeway widening and flood control channel construction on endangered species in San Diego Bay, and her later work on restoring salt marshes at the Tijuana Estuary were all early examples of the application of adaptive management.
Dr. Joy Zedler is the Aldo Leopold Chair of Restoration Ecology at University of Wisconsin-Madison, wetland ecology and restoration, as well as a member of the Delta Independent Science Board since June of 2015.
“I created the Pacific Estuary Research Laboratory at San Diego State University, and we had a responsibility to do some monitoring and to figure out how to make things better at the same time,” Ms. Zedler said. “That morphed into what I began to call adaptive restoration, which is a more specialized case of adaptive management wherein one can create experiments that fill a restoration site and at the same time, you learn while restoring. It’s an approach that I think has tremendous potential for the restoration that’s going to happen here in the Delta.”
There are various schematics for adaptive management, but Dr. Zedler’s schematic is a simple one. “I use one where you plan, you implement, you assess, and you probably will find a problem; some midcourse corrections will be needed and then you go back, plan, and implement,” she said. “I think we all do it informally … For me it happens when I’m cooking. Whenever I’m trying to set up a nice meal and I don’t have the right ingredients, I become very adaptive. I find some other way to accomplish what I set out to do.”
San Diego Bay Restoration
The first step in adaptive management is to have a team involved in carrying out all of the steps, she said. “In the case of the San Diego Bay project, we had to have a decider, a person who plans and implements, and we had to have a monitor; those individuals had their own responsibilities,” she said.
For the San Diego Bay project, Cal Trans and the Army Corps of Engineers were responsible for the restoration; Cal Trans was widening Highway 5 and the Corps of Engineers was building a new flood channel. As the damagers, they had to mitigate for the loss of habitat for three federally endangered species, and Dr. Zedler’s group, PERL, was in charge of monitoring to see what was happening as a result of their implementation.
“The Fish and Wildlife Service had two roles as the decider: they set the criteria for what would be needed and they also were the judges as to whether or not what we told them met those criteria, so it was a very small, simplified adaptive management project at the very early days of that approach to restoration,” she said. “As you will see throughout my presentation, there were some ‘Overly Optimistic Predictions, which has the interesting acronym of ‘OOPS’.”
The mitigation at Sweetwater Marsh was mandated by the US Fish and Wildlife Service under the Endangered Species Act because of the widening of the highway. The damaged habitat involved three species: the salt marsh birdsbeak (a plant), the light footed clapper rail (a bird), and California least tern, which involved fish, she said.
So for the salt marsh birdsbeak, the Fish and Wildlife Service set criteria that it must be reintroduced and form five patches somewhere in the restored wetlands, and those patches of plants must be expanding each year for three years, she said. “Now if you’re a biologist, think about how you would try to put the expansion and effective reintroduction of a species into three little criteria like that,” she said. “It was painful and we learned a lot in attempting to use those criteria. This is a plant that about 15 centimeters tall and the flower itself is about 1 centimeter – it’s not very big. It’s a hemi-parasite which means it has to tap into the roots of a perennial host at some point in order to get through the summer and produce seeds. And producing seeds is essential because this is an annual plant, and it’s something that must get pollinated in order to produce seeds.”
The marshes they had to work with were partly remnant and partly restored or created. “It’s a little hard to tell in a messy urban situation like this,” she said. “San Diego Bay has had lots of fill-in … the Sweetwater Marsh was at one time all connected the shoreline, but as you can see, all kinds of fill have trapped those wetlands behind a lot of sandy soil.”
There was a series of eight discrete little islands with tidewater around them, she said. “That’s not very natural for a Southern California wetland, but that was what was designed before we were contracted to begin some monitoring work,” she said. “At one point, the highway department asked us to sow the seeds on one of these islands. It was a new island, so it was excavated around and it had some high salt marsh vegetation on it, but ‘OOPs’, there very few seeds produced by that population. There were plants, there were flowers, but there weren’t seeds. What’s going wrong? What predictions were overly optimistic there?”
As it turned out, research showed that the preferred pollinators are the ground nesting bee. “It’s not going to dig in the ground if it’s under tidewater, so this island out here isn’t going to provide habitat for the pollinators and seeds are not going to be produced,” she said. “The pollinators needed some place a little bit higher … that was a mid-course correction then because we felt that none of the islands was going to be suitable for this species to get its pollinators and be able to produce seeds, so we recommended moving the whole operation away from the island sites and sow the seeds on the mainland.”
Dr. Zedler said it’s important to note that some things that you ask of a restoration site might not be achievable and it may indeed be necessary instead to accommodate some species in somewhat degraded existing wetlands. “This is about the only salt marsh left in all of San Diego Bay as over 90% of the wetlands are filled and excavated, so it’s possible that historically it was just a limitation of seeds, because we planted it, and in a very short order we had 14,000 plants. Remember the criteria – at least five patches, a total of 100 plants, we felt ecstatic, and after three years, they were all expanding and the decider judged compliance because the criteria were met: there were at least five patches, over 100 plants, and the populations were growing.”
Dr. Zedler said she returned last year to Sweetwater Marsh to see if the population was self-sustaining, and after three years of drought, they saw fewer than 100 plants. “Next year I’ll go back again and we’ll see if we get a wet year and if their seed bank is able to perpetuate the species. But a lesson for the Delta would be long term monitoring as something that is called effective and reintroduced in three years might not be so healthy after about 20 years.”
For the light footed clapper rail, now called Ridgway’s Rail, the criteria was to create seven 2.5-acre home ranges that would have nesting sites. “If you know the biology of the species, could you create a set of criteria that would tell you if those criteria are met, decades into the future, you’ll still find this bird having nesting habitat?”
CalTrans contractors constructed and planted the marsh, and Dr. Zedler’s group started measuring the height of the plants because the clapper rail wants big tall plants to build their nests. “They pile up debris to get a platform that can float at high tide and weave the plant stems over the canopy for protection against aerial predators,” she explained.
But there were many things that went wrong with this site, she said. “Such as a part of katamarran hull was floating around and it would move with every tidal rise and it would smash the cordgrass,” she said. “That wasn’t the only problem, but something that we never would have thought of. Debris? It’s a plastic cup, or something, not katamrran hulls.”
Another overly optimistic prediction was that the soil they were creating the islands with would support cordgrass, she said. “Our field experiments showed that when you add nitrogen, you get tall cordgrass; when you don’t, it’s short, and not only is it short, it’s attacked by scale insects and a host of other problems,” she said. “The experimentation was very effective as it showed limitation by nitrogen, so the midcourse correction was to add nitrogen for a little bit longer. We thought this would work because you keep adding nitrogen, the roots and the rhizomes will hold that nitrogen and they will release it and recycle it, and below ground, a load of nitrogen will build up in the soil. But it was too sandy, and it never happened.”
“Our research pretty much indicated that this was a futile attempt to try to get sandy dredge spoils to produce tall cordgrass because the nitrogen was simply too mobile,” she said. “So the cordgrass was too short, and the Fish and Wildlife Service judged that after ten years from the initiation of the project, that they were out of compliance. They decided to require a different kind of a mitigation, something that was easier to achieve.”
“I think there’s another lesson here as well – that we will have a better system if we understand the soil, the microbiology, and the nutrient delivery,” Dr. Zedler said. “I went back to this site last summer and it’s worse than it was when we were sampling it in the 90s. It’s hasn’t built up that nitrogen load by itself and there’s a lot of erosion taking place in those lower elevations where the cordgrass is supposed to grow. I don’t think it’s hopeful that it will grow. I think we were right.”
“We recommended terminating project to Fish and Wildlife, essentially putting ourselves out of business,” she said. “There was a comment in the Delta Plan that there sometimes is distrust of scientists – that they just want to keep on monitoring because they have a business monitoring. In our case, we said no, even though we could go on monitoring it under Cal Trans payment, we would not because it’s futile. We convinced people that the science was strong enough and that we were honest about it.”
The third part of the project was to create tidal channels and smaller creeks with adequate fish for endangered California least terns. “In this case, fish weren’t valued so much for their own existence, but for the perpetuation of the California least tern, which is a migratory bird what nests in Southern California along the beaches,” she said. “The criteria were at least 75% of the native fish species at least 75% of their density found in natural channels. That was supposed to be allowed for three years, but the criteria were actually met in two years. Fish and Wildlife judged the project as in compliance in two years, and so compared to plants and compared to birds, if we can extrapolate from one example of each, fish were a breeze. I don’t know if that will carry over to the Delta.”
“So my thinking in the very early aspects of adaptive restoration is that we use the science to improve restoration,” Dr. Zedler said. “You can insert in a project an experiment here or there – we added an experiment with short-term nitrogen and that didn’t do it, so then we added another experiment which was a 5 year nitrogen addition experiment, and you use that knowledge while meanwhile you’re restoring the rest of the site in the usual, traditional way.”
It was a little different then what happened at Tijuana Estuary. “In Tijuana Estuary, we set up a special case of adaptive restoration where we deliberately set up the project in phases, we restored sites as experiments and covered as much of the site as we could, learning while restoring,” Dr. Zedler said. “For example, if you have a small site, you can test some alternative planting methods, or if you have larger site, you can take that knowledge and use it in the second phase, and test something like the need to install tidal creek networks.”
“In Tijuana Estuary is where I’ve worked the longest in Southern California,” she said. “It’s an interesting place where there’s no control over the watershed to speak of because it’s mostly in Mexico. It has a very large watershed; it’s not a very large estuary, but it does serve to teach us something about restoration.”
“I think learning while restoring is the easiest, and it starts with identifying what you need to know that you don’t know to restore a site,” she said. “There’s so much you don’t know, but it’s what you don’t know that you need to know that’s important in planning an adaptive restoration project. Secondly, set up the project in phases, where you can start small and grow based on what you started with in the first phase. The alternative approaches and larger field experiments become really exciting because you can manipulate land form, not just plantings.”
Always evaluate and apply that new knowledge, she advised. “Each time you have a phase, you are learning something and using it in the next phase, and then learning something else to carry all of that accumulated knowledge forward.”
The first phase was a little over an acre in size, but it was an opportunity to restore salt marsh and a tidal channel that went from a natural channel into an area of former sewage lagoons that had accumulated solid waste and had a lot of nutrients present, she said.
They tested the need to plant eight halophytes. It was an intertidal site, and cordgrass was planted on the edge. The experimental plots were 2 by 2 meters, and smoothed out and made as flat as possible; each plot was identical except for what was planted in them, she said.
“The flats themselves were planted with up to six of these eight native salt marsh plants, and they each have different above ground-below ground ratios, and different life histories,” she said. “We learned tons from this and we continued to learn from it for years since this experiment was installed in 1997. … What we learned was which of these species actually had to be planted. Some species you don’t need to plant; it’s what people prior to our work had been focusing on for planting and trying to get it to grow … wasting energy and wasting people’s time and if you just wait a year or two, it’s going to establish by itself from seeds from the nearest source of seeds.”
“The other five species, they don’t like to grow from seed very well and they don’t like to spread from seeds; they are more vegetative, they are perennials, and they all need to be planted,” she said. “We learned a lot that helped all future restoration projects in Southern California.”
They also planted plots that had one species at a time and then three species at a time, and then six species at a time to test a popular theory that the more species you plant, the more functioning you will find in the system, she said.
However, there was an unexpected algal bloom. “Remember, we were connecting to an old sewage lagoon,” said Dr. Zedler. “I think what happened here is that phosphorous was mobilized from the sediment, or maybe it was the nitrogen or the combination, so we had this massive algal bloom across our acre of plantings. That attracted coots, who came in to eat the algae; they weren’t eating our plants, they were just trampling on them. So we installed some chicken wire, and lo and behold, the coots respected the chicken wire – even though they could fly in over it, they didn’t … !”
“Seven years after planting, you can see that the site has full cover,” she said. “You’ll notice a couple of things did really well – pickleweed and salt marsh daisy,” she said. “We expected to see all the species sustained at some level, but we learned a ton by seeing that in the marsh, even though we planted everything, not everything was going to persist in equal abundance; they were planted at about equal abundance in 1997, and they ended up very different.”
“This idea that you would have increased functioning or something like primary productivity with increased number species was true the third year, but it was not true after 11 years,” she said. “It was just random, and what took over was the most rambunctious plant, that’s the pickleweed. I suspect we could have predicted that, but here’s an experiment that demonstrates it.”
“I still think there’s some virtue in having some diverse plants in a site, because during some years, it may be that you have better conditions for some of the lesser species, so there still may be resilience, so if the function we’re really interested in resilience, we haven’t rejected this theory yet.”
“While the site was being adaptively restored, we learned that seven halophytes need to be planted in one way or another, seeds or plantings; only one would really come in by its own,” Dr. Zedler said. “We learned that species whose plantings increased initially increased functions, but they did not persist, so there’s a long term lesson there, which is to monitor long-term.”
So they brought what they learned forward to the second phase; they didn’t plant picklweed. The second phase was 20-acres, big enough to test the need to add topographic heterogeneity to the site. They added creeks and pools formed on their own, a benefit to both fishes and plants, she said.
They experienced some unanticipated problems with sediment, with one event bringing a lot of sediment down and filling large portions of the tidal creek, costing about $3 million to excavate. A sediment retention basin next to the order was filled in with the first flood, so that had to be redone, she said.
“The idea was to test the importance of tidal creek networks for fish and for plants,” said Dr. Zedler. “If you have site, can you leave it plain? Recall that I said that our first site, we made it so smooth so that everything was perfectly replicated for research purposes. Well that turned not to allow some of those species to persist. We learned through later experimentation that little pools allow one of the little annual plants to persist and not to be overtaken by the perennial pickleweed, so that tiny topography of 5 centimeters was critical for one species, and we didn’t’ have it because we were so excited about making everything perfect for research. You have to learn to be accommodating in the future.”
Not all unexpected events are necessarily bad. “Here is an example of an excavated tidal creek that received a lot of sediment from that 2004 extreme event, and one of the outcomes was that fish used these creeks to burrow and to feed, that was a good fish outcome.”
Another unexpected surprise was in the process of excavating, some remnants of pots that were judged to be Native American artifacts were found, so a complete anthropological survey had to be done. “They didn’t find any significant artifacts, but it did delay the project,” she said. “It’s the kind of thing you just don’t imagine is going to happen, and yet it did, so in the process of delaying, the site became a salt flat extremely hostile to our 5000 planted seedlings, and not including pickleweed, almost every single one of them died.”
“The experimental design had to be adaptively changed, let’s put it that way,” she said. “We had a few plants leftover so we completely redesigned our experimental planting based on the basis of what we can do with 180 plants. We decided to test proximity to creeks and tightness of clusters of plants and could they be far apart and do better, and so it was a lot of fun, but we too had to be adaptive.”
They got more pools than they had planned and more sediment than they expected; they attributed the formation of the pools due to their location where the fetch and turbulence were the greatest, she said. “But they turned out to be great for the California killish because they could now get from the big channel … and they can move onto the marsh plain. We call them oases for fish feeding. That was a happy outcome.”
“We did get a whole lot more sediment than we expected, and what was supposed to be nice replicated tidal creek networks became much fuzzier,” Dr. Zedler said. “The actual berm around the entire site breached … and a lot of this sediment just moved right into this site and filled in over a foot of sediment. We ended up with pickleweed habitat in front of cordgrass habitat, very strange.”
They did learn some things, Dr. Zedler said. “Tidal creeks enhance plant establishment, they enhance the fish use of creeks and the marsh plain, and they actually help allow some of that sediment that came in to flow out again, so there were measurable benefits. We had to adapt our monitoring to accommodate the changes that were occurring that were completely unexpected. We weren’t supposed to monitoring sediment deposition of this magnitude.”
In conclusion …
Dr. Zedler than gave her conclusions. “We have adaptive management wherein you could add experiments to sites that are being restored in traditional ways, which can become a little more informative,” said Dr. Zedler. “At the same time as restoration is taking place, you can use small sites to test smaller scale questions, like what to plant where and how, and larger sites for larger scale treatments such as topographic heterogeneity which is a real important thing to include in a restoration site.”
“I still think we need to have very ambitious goals, and we can expect to achieve some of them,” she said. “We can also say it the other way. Don’t expect to achieve more than a few of them, but I like to see the positive side. I think if we can achieve some of the goals, we will be far ahead, especially if we are building on the knowledge that we learned from things that aren’t the way we expect them to develop.”
“It’s hard to restore habitat for various species,” she said. “We had three kinds of rare species; it was hard to restore everything except the fish for the least tern. Restoring those populations can be harder than just restoring their habitat. Sometimes the weather can help if you can get rainfall; plan to plant before the rainfall, drought afterwards in the Mediterranean climate where the tides expose so much soil is a bit of a hindrance. Expect sedimentation and erosion to completely reconfigure topography, but somehow retain heterogeneity to support biodiversity and yes OOPS, erosion control berms don’t always hold.”
“Future phases of the work in the Tijuana Estuary is going to be up to 250 acres, and I think we’ll be able to accommodate a lot of experiments there,” she said. “Compared to the Delta, the opportunities are really quite small so I thought I would try to put it into scale. I’ve been talking to you about these miniscule little projects in San Diego Bay and Tijuana Estuary, and this is approximately the scale with the Delta … The actual scales are here at the base. Sweetwater Marsh, a little more over 1 km square, compared to over 2900 km square for the Delta.”
“A theme that was running through my slides was that we really need to anticipate when we are being overly optimistic,” she said. “In Madison, I try to do similar things. I try to experiment while creating or restoring wetlands. This is our stormwater research facility with stormwater coming in from an urban landscape, and it can slow down through four weirs equally into four absolutely identical engineered wetlands, and of course they turned out to be not identical at all, even though to the inch they were graded and treated the same.”
“Then OOPs, they weren’t supposed to be exporting nutrients; they were designed to trap nutrients, but they ended up exporting nutrients, and you can ask how did that happen? Well there was a best management practice around saying you always add 6 inches of topsoil to a wetland if you want to grow plants in it, and they had to grow plants to get thick vegetation to withstand the inflow of the water and hold the soil. Well it was completely wrong because that’s the only place that we can figure out where the nitrogen and phosphorous came from that was exported.”
Other things happened as well. “We had 27 species of plants that were to be planted, and one in three and nine species combinations, and none of them even germinated, so the site was either toxic or the seeds were mishandled by the contractors. We don’t actually know. I think it was the latter because we did some seed bank studies and nothing germinated in the lab either.”
“We had a big infestation of crown vetch which is used as a planting on highways, and it became 100% cover on these berms,” she said. “Where did that come from and how come it’s no place else in this 1200 acre arboretum in Madison? Well, they brought in a track vehicle that carefully made these desired topographic heterogeneity that we wanted in each of these swales, but the track vehicle was not cleaned before it came into the site, so it must have been a seed bank that it was carrying on its tracks, and it very effectively planted the entire berm, so if you want to plant a berm with whatever, put it on a tracks of a track vehicle.”
“And then the last thing, everybody knows what a weir is,” she said. “We asked for weirs so we could direct the water into these swales equally, and every single weir in and out leaked, and the water went around the edges, and one of the project proponents had the audacity to say, well ‘you didn’t tell us the weirs weren’t supposed to leak,’” she said. “So they all had to be augmented and that slowed up the project for two whole months during the growing season which is almost death on research because you have the students doing their work and they have to get their theses done. We did publish the work but it was certainly much more difficult.”
“I’m going to quit with this, and say given many overly optimistic predictions, it’s wise, really wise to invest in research alongside the restoration and have fewer OOPS than we have seen in the earlier days of restoration,” Dr. Zedler said. “My new book which is free online, Salt Marsh Secrets: who uncovered them and how? There are 21 chapters and stories of all the students and how they’ve done such a great job and where they are now. I’ve kept track of about 40 or so of them, and some of them are right here in the Bay Area.”