Symposium coverage, part 1: Delta and longfin smelt: Is extinction inevitable?

Cousins smelt symposium header fullPart 1 of coverage of symposium covers what we know and don’t know about the Delta and longfin smelt

Is extinction inevitable for the Delta and longfin smelt?  Despite extensive efforts and expense investigating potential causes of their decline, their numbers have reached an all-time low, particularly for the Delta smelt.  These species are listed as threatened through state and federal endangered species regulations, which means that water supplies are directly affected by their presence near the pumps and river flows are regulated based on their life cycle and known distribution.

As the species near extinction, research and management priorities must keep pace.  What can we do to preserve the species and avoid extinction?  This day-long symposium featured species status updates, discussed possible causal links to their decline, and explored how recovery planning, restoration, and decision-making efforts can intersect to sustainably manage the Delta and longfin smelt.

The symposium was held on March 29, 2016, and was presented by the Delta Science Program and the UC Davis Coastal and Marine Sciences Institute.  The format of the program featured individual presentations by scientists of their work and discoveries on the Delta and longfin smelt, and then wrapped up with a lively panel discussion.  The symposium will be covered in four parts.  Today, presenters discuss what we know and don’t know about the Delta and longfin smelt.  Tomorrow, presenters will discuss casual links to the species decline.  On Thursday, we’ll take a look at the future and what can be done moving forward to protect and restore the species.  And on Friday, coverage will wrap up with a panel discussion moderated by Dr. Peter Moyle.

(Not on daily email service?  Sign up here, and you won’t miss a post.)

Presentations featured in today’s post

Click on the link to jump to presentation or simply scroll to read each in sequential order:


DR. TED SOMMER: The ecology and management of Delta smelt: Insights into the most important non-game fish in the US

A look at what Dr. Sommer calls ‘the economically most important species in the country not named salmon’

Ted SommerBIO: Dr. Ted Sommer received his PhD from University of California at Davis, where he studied under noted fisheries biologist Dr. Peter Moyle.  Dr. Sommer is currently Lead Scientist for the California Department of Water Resources.  For the past 25 years his work has focused on native fishes, with studies on Delta Smelt, salmon biology, floodplain ecology, food webs, and hydrology.  Dr. Sommer has published more than 50 research articles in peer-reviewed scientific publications. 

Dr. Sommer began by noting that on the Pacific Coast, salmon are king; they are an iconic species and important from an economic standpoint and as a sport fishery. “In the Bay Delta, salmon are also a really important fish, but I would argue that Delta smelt are at least as important in terms of the economic and management significance of the species, and perhaps even more importantly, I would argue that they are the economically most important species in the country not named salmon.”

So how is that possible? You need to understand a little bit about Delta smelt life history, Dr. Sommer said, presenting a simplified life cycle adapted from Fred Feyrer at the USGS.

The Delta smelt is an annual fish, so everything happens really fast. It also occurs over a relatively small geographic area.  In fall and summer, Delta smelt are located in the low salinity zone for the most part, Suisun Bay and the North Delta where they feeds and mature;  that changes in winter, when we start to get freshets that tend to create an upstream shift in the distribution of the fish.  The fish spawn, they rear, and the fish are in the Delta, and gradually move downstream.”

Sommer smelt_Page_04 Sommer smelt_Page_05

The Delta smelt isn’t a sportfish, so how did the Delta smelt become such a big deal? “The Delta is the heart of our water distribution system in California,” explained Dr. Sommer.  “We have the Central Valley project and the State Water Project located in the south Delta.  We also have thousands of agricultural diversions in the area, and this is the water supply for 25 million people in the US, about 8% of the population.  It supplies a multi-billion dollar agricultural industry and also fuels a lot of our economy.  This really came to a head during the drought when we saw just how important our water supply became.”

Sommer smelt_Page_06The issue for Delta smelt is twofold,” he said. “If you look at the distribution map, it overlaps with the location of these large water diversions, so you’ve got the overlap.  The second point is the large water projects and the smaller water diversions affect the salinity field for Delta smelt, so we have a big setup for conflict here.”

Sommer smelt_Page_07Dr. Sommer then turned to the abundance trends, presenting some graphs showing results from Delta smelt monitoring surveys. “At the bottom, I show the IEP summer and fall surveys for juveniles and sub-adults, and what we see is there is a long-term decline in the species.  This led to federal listing and subsequent state listing in the early 1990s.  And as part of all this, we’ve had some pretty substantial biological opinions that have placed restrictions on our water delivery system, and also the status of Delta smelt has been a big driver for the State Water Resources Control Board’s water quality control plan, things like the Bay Delta Accord.  So the status of Delta smelt is a key driver of a lot of our operations.”

Sommer smelt_Page_09Since the Delta smelt was listed, other surveys for other life stages have been added in order to gain better insight, such as the Spring Kodiak Trawl and the 20mm index. “Honestly, the picture isn’t any prettier,” said Dr. Sommer.  “Starting around 2000, we had an even more substantial decline in several of the abundance indices; we recognize that as the Pelagic Organism Decline. It was quite a substantial decline, and things are even uglier now.”

It’s really come to a head during a drought and that’s why we’re here today. Starting in the beginning of the drought and in the subsequent years, we’ve kind of had one record low index after the other …  perhaps the most disturbing was the summer index.  It was the first zero index that we got for the species, so certainly a big red flag here.”

Sommer smelt_Page_10Not surprisingly, this has caused a lot of angst; it’s been the focus of media and public attention, it’s been in the national news, there have been lawsuits and even high-profile protests, he said. The issue was at a head at the time of the symposium as due to recent rains, outflow was high yet there were restrictions on exports, because, as Dr. Sommer put it, “the Delta smelt were exactly in the wrong place, and a lot of folks are concerned about all the water being ‘wasted’ going out the Golden Gate bridge.”

The upside of this is that it has spawned a lot of research on Delta smelt. “Around the time of the listing of Delta smelt, we knew almost nothing about the species – basically where it occurred and not a whole lot else, so we’ve seen a steady increase in publications, and right now in California, it’s probably the most studied native fish not named salmon.

Dr. Sommer noted that if you who don’t have read all 70+ papers on the Delta smelt, the IEP recently put together a synthesis report based around a conceptual model for Delta smelt. “Despite the fact that Delta smelt are an annual fish with a seemingly simple life cycle, things are really, really complicated,” he emphasized.  “We have this complex interaction between landscape factors, environmental drivers, and habitat conditions that all work to affect different life stages.  There is a whole lot of information behind all this.”

He then gave four ‘cool’ facts about the Delta smelt:

Sommer smelt_Page_131. The name: Delta smelt.The Delta smelt does indeed only occur here.  A lot folks assume from the name it only occurs from the Delta, and that is the core of its distribution, but what we’ve learned over the years is that the distribution a fair bit broader.  Susuin Bay is a really key part of the habitat, the Napa River is also a key part of the range of Delta smelt, and in higher flow periods, so is San Pablo Bay.  Seasonally when we have higher flows, we see smelt move periodically a fair ways upstream. You can catch them as far upstream as places like Knights Landing, so perhaps ‘Bay Delta smelt’ … ?”


Sommer smelt_Page_152. Food habits of Delta smelt:Little tiny fish, little small mouth, so it’s not surprising that this is a plankton eater, and a lot of the emphasis of the past 10-15 years has been on a couple zooplankton prey, Eurytemora and Pseudodiaptomus that occur in the low salinity zone.  But we found based on some recent diet studies that the fish may not be quite as picky an eater as folks have assumed.  Pelagic prey up in the water column or demersal prey more down towards the bottom, and they’ve evaluated things for a couple different seasons, spring and fall.  What we see is near the top, yes Eurytemora and Pseudodiaptomus are a really big deal in the spring and into fall as well, and other calanoid copepods can also be part of the diets, but if you look farther down, you see demersal prey, particularly in fall, can be important food source as well.  So there is a little more plasticity than what we might have given the fish credit for and this has implications for its habitat use and possibly for habitat restoration.”

Sommer smelt_Page_163. Delta smelt like turbidity:The other thing that’s kind of cool and unusual about Delta smelt is their love for silty, muddy water.  For a lot of the different other fish species that we care about, we really focus in on cold, clear water. For species like salmon, a lot of the water quality objectives are based on that.  What we’ve learned for Delta smelt is that they really prefer more turbid conditions.  The disturbing thing in all this is that we’ve had a long term reduction in our sediment supply to the estuary.  In the right hand side, is work from Dave Schoelhamer who has documented a long term decline in sediment being delivered to the estuary and even worse, he documented a step change that occurred around the time of the Pelagic Organism Decline.  The net effect of this is that the water in the Delta is clearing rapidly, and this is one of the main sources of habitat degradation to Delta smelt.  A lot of the sediment is being held behind dams upstream, and the Gold Rush sediments have made their way through.”

Dr. Sommer noted that they have been able to use this to their advantage to help manage entrainment of fish. “I mentioned bad things happen when fish move up into the central Delta near the water diversions.  Well, Delta smelt don’t go all the way if there’s clear water in the way.  They tend to avoid it, at least that’s what the pattern seems to indicate, and so we’ve added a variety of fixed telemetry water quality stations to tell us what the turbidity levels are like and we’ve started over the past couple years to do turbidity transects.  This information is subsequently passed on to our water project operators who do their best to manage inflows from the dams and also export levels to keep those turbid waters from penetrating too far, and this has helped to limit some of the entrainment diversion losses at the water projects.”

Sommer smelt_Page_184. Delta smelt habitat is not just pelagic.Everyone thinks the Delta smelt as a pelagic fish as they are present in a lot of the open water areas.  What we found in recent years is that shallow open water areas like Liberty Island have just as many smelt or at least they are as common as some of the core Delta stations, at least seasonally or so.”

Dr. Sommer said to stay tuned, as there are a lot of things in the pipeline, such as smelt cams, genetic work, gear efficiency, and other things.

We’ve learned a lot about Delta smelt. They are doing badly, it’s a high profile issue of national importance, however, the reason we’re here today is the future of Delta smelt is not looking particularly rosy and the extreme drought that we’re seeing right now is just kind of a small taste of what we may be experiencing in the future,” he concluded.

Click to go back to the top.

DR. JON ROSENFIELD: Longfin smelt wonder, “Will humans evolve a spine before we fishes go extinct?”

We know enough about the longfin smelt to manage for their recovery, he says

Jon RosenfieldBIO: Over the past 7 ½ years, while serving as Conservation Biologist for the The Bay Institute, Dr. Jon Rosenfield has worked to bring scientific information and science-based decision-making processes to the planning and regulatory processes that affect the San Francisco Bay’s estuary and the Central Valley Rivers that feed it.  He has published several papers on the behavior, ecology, and systematics of fishes, particularly imperiled species. Most recently, he co-authored a paper on longfin smelt population dynamics with the US Fish and Wildlife Services’ Matt Nobriga (lead author).

Dr. Jon Rosenfield began by saying that he’s not entirely happy with the question the seminar poses, ‘is extinction inevitable for Delta smelt and longfin smelt?’ because the answer is dependent on the time frame and the management approach. “Over a very long time frame, extinction is almost certain,” he said.  “We can ask the dinosaurs about their chances and what they did to prevent their extinction.  But over a time frame of one to fifty years, the timeframe that those of us in this room will be working on this issue, the question is really irrelevant to our management.  Extinction is not an option.  We have state and federal Clean Water Acts, state and federal Endangered Species Act, and the public trust doctrine, and so we’re not encouraged to be considering such esoteric questions about whether extinction is inevitable.  I frankly think the question distracts from our responsibility to protect these fish and the ecosystem that they depend on.”

But with regard to longfin smelt, the question is what do we know about longfin smelt, and the answer is that we certainly know enough to manage for their success,” he said.

Dr. Rosenfield said the data in his presentation come from two studies:

Rosenfield smelt_Page_06The San Francisco Bay study, which samples the entire estuary year round every month with two nets: an otter trawl that samples the bottom of the water column and a midwater trawl that samples the midwater of the water column. This program started in 1980.

Rosenfield smelt_Page_07The Fall Midwater Trawl, which samples every month from September to December, and uses a midwater trawl. This program started in 1967.  It samples only the northern estuary whereas the Bay Study samples the entire estuary.

Longfin smelt are semelparous (meaning they reproduce only once during their lifetime), they have a two-year lifecycle, they are migratory, and they are partially anadromous.

Rosenfield smelt_Page_08He presented a graph from a paper published in 2007 in which he and Randy Baxter studied the median cohort through its 24 month lifecycle.  He noted that the bars show the percent presence at sampling locations and the line shows the average catch per unit effort in each month of life for three studies: the Bay Study’s midwater trawl, the Bay Study’s otter trawl, and Suisun March Survey.

The basic pattern is similar; that abundance and distribution are high early in life; they sort of peak in that first winter of life, and in the first summer of life, abundance and distribution decline. It might normally mean that the fish are dying and certainly they are dying, but it’s not all mortality because catch per unit effort and percent distribution both increase in that second winter of life; after the second summer of life they sort of come back into the ecosystem.  This is part of the reason we know these fish are migratory; they partially leave the sampling zone in the summer and they all come back during the winter.  We also know this because we catch longfin smelt in sporadic sampling efforts outside of the Golden Gate.”

Rosenfield smelt_Page_09The larvae of longfin smelt are concentrated near X2 or the low salinity zone, and freshwater inflow to the estuary affects their distribution,” Dr. Rosenfield said, presenting a schematic showing the distribution of longfin smelt larvae in February and March. “In dry years, they are concentrated in the west Delta and Suisun Bay area; and in wetter years, the center of their distribution is more in the San Pablo Bay, Suisun Bay, and Central Bay region.”

Rosenfield smelt_Page_10That distribution early in life has implications later in life,” he said. “It has implications for the distribution of juveniles.  The size of the dots represents average catch of longfin smelt in 1983, which was a very wet year. In 1992, a drier year, almost all of the catches are upstream.”

Rosenfield smelt_Page_10Dr. Rosenfield pointed out that the distribution of the fish also has implications for entrainment. “The spawning adults appear to be tracking where the salinity zone is and they move upstream of that,” he said.  “If the salinity zone is far to the east, then the fish move far to the east and then they are susceptible to entrainment, their larvae and subsequent juveniles are susceptible to entrainment.   Several authors have noted is a negative relationship between Delta outflow, which is shown here in red bars, and longfin smelt entrainment which is the blue line.  There’s a strong negative correlation between these two, so in years when outflow is low, a lot of longfin smelt get entrained.  When outflows are high, almost no longfin smelt are entrained at the state and federal water projects.”

Rosenfield smelt_Page_11Later life stages are euryhaline and diffuse, and again show seasonal migrations. In June and July for age 0 and age 1 fish, the peak densities are in the Central Bay, the channels (at least) of San Pablo Bay, and somewhat in Suisun Bay; in the early fall, October and November, the distribution has a landward or freshwater skew and most of the fish are found in the west Delta or Suisun Bay at both age 0 and age 1, which is in the second fall of life when they are preparing to spawn, he said.

Rosenfield smelt_Page_12He then presented a graphic of the life cycle of the longfin smelt. “The simple life cycle is that the adults are moving towards freshwater on the boundary of fresh and salty water habitats, they deposit their eggs largely in freshwater or perhaps slightly brackish water, then the larvae aggregate at that interface between fresh water and saltier water,” he said.  “Juveniles appear to have two strategies.  At least some are in the brackish water of the San Francisco estuary’s main embayments, some are in more marine environments or the central Bay, and there may be a back and forth where juveniles that wind up in the marine environment spend their second year in brackish environment and vice versa; we really don’t know much about this migratory behavior in and out of the Golden Gate.  But regardless of where they rear, they come back as adults, head towards freshwater where they spawn again.  When I say freshwater, I’m talking largely about the west Delta, but we also believe that these fish spawn in the Napa River, Coyote Creek, etc.  They are looking for that freshwater saltwater interface.”

Rosenfield smelt_Page_13He presented a slide of the longfin smelt abundance index from the Fall Midwater Trawl, noting that the data is from the 1980s onward; he explained that because longfin smelt have a two year life cycle, it’s cleaner to look at every other year as one year class, so the odd years have been broken out from the even years.

Either way, the population decline as represented by the Fall Midwater Trawl is horrific – greater than 99% since the 1980s,” he said.

Rosenfield smelt_Page_14He then presented a slide from the Bay Study’s midwater trawl, which samples adults, sub-adults, and juveniles. He noted that both year classes depicted here in blue and the otter trawl depicted in green.  “You see the same general pattern, and while it looks like the Fall Midwater Trawl decline is more dramatic and it is, it’s the difference between a greater than 99% decline, a 96% decline for the Bay otter trawl, a 98% for the Bay midwater trawl.  The fish have declined, there’s tons of evidence from independent sampling to show that.”

Rosenfield smelt_Page_15The final notable thing that we know about longfin smelt is their abundance shows a persistent significant correlation with freshwater flow to the estuary during the winter and spring months,” Dr. Rosenfield said, presenting a graph of the Fall Midwater Trawl Index from 1967 to 2014; the green bars show the Delta outflow in the March through May months.

The key thing to get from this diagram is that when Delta outflows are high, the longfin smelt population is high; when outflows decline, the longfin smelt population drops.  When outflows increase again, the longfin smelt population increases, etc.  This relationship has held over about 50 years of Fall Midwater Trawl sampling, and the flow relationship alone explains about 42% of the 50-year variation in longfin smelt abundance, which is pretty damn good for an ecological study.”

Although we know a lot about the longfin smelt, there still is the question of what’s driving their decline, how do we manage them and how do we restore them, he said. “Matt Nobriga at the US FWS and I took these major facts that I’ve outlined about longfin smelt and tried to determine whether the forces driving longfin smelt decline occur throughout their life cycle, or do they occur in particular parts of their life cycle and are forces in the different parts of the life cycle the same, or are they affected by different things as they migrate around,” he said.

Rosenfield smelt_Page_18So we tried to disaggregate the forces driving longfin smelt decline into two life stages using data from these same sampling programs,” he continued. “We used a stock recruit framework because fish make fish, despite all my graphs showing that flow makes fish, actually I’ll admit that fish make fish.  So it made sense to do this in a stock-recruit framework.  We parameterized our models with data from the Bay Study, the midwater trawl and the otter trawl data combined, and then we wanted to see whether we could recreate the Fall Midwater Trawl pattern of abundance, the Fall Midwater Trawl time series, using the parameters developed from this independent sampling program.”

We used a two life stage model, so how do age 2 fish produce age 0 fish and what’s the relationship between age 0 fish and the subsequent number of age 2 fish within a generation, looking first at how do spawners affect recruits,” he said. “There’s a spawner recruit relationship; we screened a number of variables to see how they might modify that relationship, and we found that freshwater flow was not surprisingly a strong modifier of the spawner-recruit relationship.  There’s also density dependence in this relationship. … We found small evidence for a step change in 1987 that would actually increase the spawner-recruit relationship, which is the opposite of our expectation but it was sort of marginally significant.  Freshwater flow is the one factor that we screened for that affects the relationship between spawner and recruits, and it hasn’t changed through time as far as we could tell.”

The second part of our spawner-recruit model then is how do recruiting fish, juveniles, survive to age 2 fish and what affects that? Here, freshwater flow was not an important modifier; there was density dependence, but none of these other physical variables had a significant effect.  We modeled several different ways in which survival rates could have changed over time, we modeled at as a continuous decline as there was some evidence for that, and we modeled it as a step change in 1989 which would correspond to the clam invasion two years later on the spawning population as there is some small evidence for that, and the best supported change was in 1991, but as you’ll see, even that step change has some questions associated with it.”

Rosenfield smelt_Page_19We were trying to develop parameters from Bay study data and see if we could then recreate the pattern seen in the Fall Midwater Trawl data,” Dr. Rosenfield said. “We had two statistically best models.  These models are indistinguishable from each other – one without a food web step change and one with a food web step change.  Here you see that we were able to recreate the Fall Midwater Trawl time series in the solid black line, using our statistical model, which the median projection is here in the dashed line.  Again, this model over here without a food web step change was similar or as good statistically as the one with the step change, so we don’t quite know about the food web impacts, although there is some decline in survival of juveniles.”

He then presented a conceptual model of how the fish migrate. “What affects juveniles is the spawning stock that you start with – not surprising,” he said.   “There is some density dependence involved there.  And the effect of freshwater flow on this relationship between adults and subsequent juveniles is very large, very significant and it hasn’t changed over time.  Flow affects adult to juvenile survivorship, and the number of juvenile recruits you start with does affect your number of adults, not so surprising there.  There’s density dependence in this relationship, and there’s declining survival between juveniles and adults over time which may be related to food availability, although we can’t quite tell when that problem arose or whether it’s just been declining over time sort of steadily.”

The thing to get out of this is that the part of the life cycle between adults and juveniles that happens in freshwater, and the one thing modifying our stock recruit relationship is flow; it’s the same factor for the past 35 or 50 years, depending upon which sampling program you use,” Dr. Rosenfield said.  “With the juvenile to adult survivorship, there’s some declining survival that’s not related to flow, but they key point is that this is happening largely outside of the freshwater environment downstream of the Delta, but in terms of things we can manage right now, we can manage the relationship between adults and juveniles by managing flow.”

Rosenfield smelt_Page_21Returning to the question, is extinction inevitable for Delta smelt and longfin smelt, Dr. Rosenfield said we’ve been here before with endangered species. “When I started working in the SF Bay estuary watershed, winter run Chinook salmon populations were 200, 300, 180 and if we asked the question, is extinction inevitable there, we probably would have been wrong, because within two generations, the populations had increased by two orders of magnitude.  Nature provides a great cost share when we do the right things to conserve species.  Populations want to survive and they grow exponentially.  And even for longfin smelt, at the end of the last drought, our Bay study midwater trawl index was down here in the 1000s, and coming out of the drought once wetter conditions returned, the population quickly increased by one order of magnitude or more in just a few generations.  So recovery is possible if we actually manage for recovery.”

Rosenfield smelt_Page_22To conclude then, I think the question for this symposium should be is extinction inevitable for Delta smelt and longfin smelt if we fail to use our best information to manage for success?” said Dr. Rosenfield.  “Then you could answer, that’s a stupid question, let’s get back to work, and let’s not try to manage without using our best available information.”

Click to go back to the top.

RANDY BAXTER: Slippery Slope: Documenting the Decline of Smelts in the San Francisco Estuary

A review of the monitoring surveys and what the data shows about the abundance of Delta and longfin smelt

Randy BaxterBIO: For over 27 years, Randall Baxter has worked for the California Department of Fish and Wildlife (DFW), first on the San Francisco Bay Study, a long-term fish and invertebrate monitoring project, then leading the Splittail Investigations Project and more recently as a supervisor for several long-term monitoring projects.  He has contributed as a member of the IEP Management Team, the Pelagic Organism Decline Management Team and most recently the Management Analysis and Synthesis Team (MAST).  He provides research-based knowledge to Department for the management Delta Smelt, Longfin Smelt and other estuarine species of concern.

Randall Baxter supervises several of the long-term fish monitoring programs for the Department of Fish and Wildlife. In this presentation, he discussed the information they’ve gathered regarding Delta smelt and longfin smelt.

He began with his take home messages:

The abundance of both smelt is in severe decline – over 90% for both, yet in recent years, both under good habitat conditions have rebounded and exhibited resilience,” he said.   “Shifts in distribution can influence abundance, and not capturing the entire distribution of the fish can also lead to an abundance index that’s biased a little low.  Zero abundance doesn’t equal zero population.  We’re out there taking a sub-sample of all the water that’s available, so we can’t capture – nor is it a goal to try and capture the entire population with our sampling efforts.  Many will survive out there even when we show a zero population abundance.”

Mr. Baxter said he would describe the monitoring surveys that are providing information on abundance and distribution, look at some recent patterns, and then identify some of the limitations in a couple of the sampling programs that hopefully put a little better spin on how many fish are out there.


There are four monitoring surveys conducted for Delta smelt, each targeting a different life stage:

The Spring Kodiak Trawl tracks adults in approximately the full range of Delta smelt, which is from western San Pablo Bay, upstream through Suisun Bay and somewhat into the marsh, and into the Delta.  The Spring Kodiak Trawl conducts a single surface tow at the stations that it samples.

The 20 mm survey is conducted in the late spring and targets larvae and small juvenile fish in the 15 to 30 mm range.  The 20 mm survey conducts three tows at a single location to help increase the volume of it’s sampling to better detect fish.

The Summer Townet targets juveniles and occurs in June through August.  The survey conducts three oblique tows in the water column to drive its data.

The Fall Midwater Trawl looks for sub-adults; it begins in September and ends in November, conducting a single oblique tow at each of the sampling locations.

Mr. Baxter noted that the way the sampling is important. “An oblique tow samples the water column,” he said.  “Delta smelt use the entire water column, but sometime early in the spring to early summer, they start moving towards the surface and by late summer and certainly into the fall, there’s a predominantly surface distribution.  With some of our gears, we’re not necessarily sampling the highest densities within the water column, because most of these gears follow the boat and by the time the gear gets near the surface, the boat has parted a lot of smelt and they’ve moved away.  I don’t have time to show you some evidence of this, but we do have some.”

Baxter 1_Page_05The abundance of Delta smelt has declined pretty significantly over time and in 2015, we’ve detected the lowest abundance in Delta smelt in all four surveys, so the decline has been pretty severe,” he said. “But when conditions have been optimal in the recent past, we’ve seen a circumstance where a relatively low adult abundance under good habitat conditions in 2011 has resulted in substantial increases in larvae, then juveniles, then sub-adults, so when the environmental conditions are suitable throughout the whole year, then finally we achieved our highest abundance index ever for the Kodiak trawl.  This abundance carried over into the larval stage but later in the summer of 2012, harsher conditions caused a pretty substantial decline.  So favorable conditions in terms of flow and modest water temperatures and food can lead to substantial rebounds in fish abundance or in Delta smelt abundance, and obviously harsh conditions can turn things around quickly.”

Mr. Baxter then talked about recent changes made to the two of the surveys for Delta smelt that were implemented in 2010 and 2011. He then presented a map of the Summer Townet sampling sites, noting that the historic sites are shown in black and the new sampling sites are shown in yellow.  “In 2011, we start sampling in the Cache Slough DWSC area and more recently have developed some abundance indices for the catches that were made in this area to combine with our historical index,” he said.  “We did similar sampling for Fall Midwater Trawl.  The graph shows the new some of the indices stacked on top of the historical indices, so the historical indices are in black and the distribution or the indices calculated in the same fashion from the Cache Slough DWSC area are in the pattern.”

Baxter 1_Page_06
Baxter 1_Page_07

Mr. Baxter noted that in the summertime, substantial portions of the Delta smelt population are rearing in the northern Delta and Cache Slough DWSC area. “The abundance declines along with the abundance in the core stations and in 2015, when the summer townet showed a 0 abundance index, there was still a blip of abundance that was detected in the northern Delta,” he said.  “By fall, the northern Delta has suffered through a summer’s worth of warming, and many of the Delta smelt that are up there migrate elsewhere, but we detected Delta smelt in the Fall Midwater Trawl between September and December in every year, although it’s not readily obvious in the abundance indices that I’ve placed here, so this represents one portion of the population distribution that has gone unsampled until recently by these two surveys.  Some of the previous surveys do cover this portion of the Delta smelt range.”


Mr. Baxter then turned to discuss the monitoring for longfin smelt. “The three monitoring surveys that we traditionally look to for abundance for longfin smelt are the Fall Midwater Trawl and the two Bay Study trawls,” he said.  “Fall Midwater Trawl includes sampling in San Pablo Bay as well as Suisun Bay and the Delta, and the Bay Study starts at the Delta confluence and samples throughout the estuary, and the reason that this is relevant is that longfin smelt don’t spend a lot of time in the upper estuary and they utilize the whole estuary and perhaps some of the local coasts.”

Baxter 1_Page_09One of the reasons why we don’t use indices from the larval sampling and early juvenile sampling is that the larval and juvenile sampling starts in eastern San Pablo Bay and Napa River area and Carquinez and samples only upstream of that,” he said. “Here I’m showing some of the historical Bay Study data for larvae, and I’m showing the larval relative distribution in a wet year and in a dry year.  Comparing that to the mean distribution of X2 and as you can see, as X2 moves down, the longfin larval distribution includes San Pablo Bay and a smattering of fishes all the way into South San Francisco Bay because of tidal dispersal.”

In contrast, in a dry year, they are upstream and within an area that’s sampled by these juvenile surveys, one result of this is if you were to use these surveys, you would see an opposite abundance outflow relationship to what we see for the age 1 and adult fish in that the higher the outflow, the lower the abundance because a greater fraction of the fish are being distributed into San Pablo Bay and downstream, out of the range of those larval sampling programs.”

Baxter 1_Page_10Mr. Baxter then presented a slide of the surveys for longfin smelt, noting that he is only showing the bottom 10% of the abundance indices. “You can still barely see recent abundance,” he said.  “Fall abundance in the Fall Midwater Trawl has been the lowest on record; that’s not the case for some of the Bay Study Surveys.”

Mr. Baxter reiterated the point that Dr. Rosenfield had made in his presentation that similar to Delta smelt, when longfin smelt encounter good environmental conditions, particularly higher than average winter spring outflows, the abundance is able to jump back. “We can see that in the last couple years of good outflow in the system, in 2006, and perhaps a little less so in 2007, based on Fall Midwater Trawl.  Based on Bay Study otter trawl, 2006 was a pretty substantial bump up and 2011 is there, as well.  So there is still a chance that populations are still potentially resilient.”

Baxter 1_Page_11Next, he showed what he said was evidence that longfin are using their habitat a little differently. “Here, I’m showing the distribution in black of the bottom salinities measured by San Francisco Bay Study in their sampling, and in red, the relative position of longfin smelt in that salinity habitat,” he said.  “The red shows Bay Study data for when longfin smelt were detected, and the overall pattern is for longfin to be distributing themselves farther downstream.  I’ve just plotted polynomial lines here to show the average through the points here, and as you can see, longfin smelt have a fairly broad distribution, but in about 2000, this distribution became narrower and started moving downstream.  I suggest that longfin are using the more marine aspects of the habitat and perhaps the open coast more than they have in the past.”


The abundance of both smelts is declining and multiple surveys for each substantiate the pattern.  Some surveys result in low detections, but Mr. Baxter is hopeful that smelts have not lost resilience.

Click to go back to the top.

DR. MANDI FINGER: Evaluation of genetic Ne of Delta Smelt provides hope for recovery

Genetically speaking, Delta smelt are recoverable, says Dr. Finger

Mandy FingerBIO: Dr. Amanda (Mandi) Finger is the Associate Director of the Genomic Variation Lab in the Department of Animal Science at UC Davis. She has researched the population genetics of Delta Smelt since 2011, and Longfin Smelt since 2014. Dr. Finger’s current smelt projects include directing the genetic management of the refuge population of Delta Smelt at the FCCL, collecting genomic data on historic Delta Smelt from the 1990’s to present day, examining population genetic structure of Longfin smelt from Alaska to California, and developing environmental DNA protocols for detecting Delta Smelt in the San Francisco Estuary.

Dr. Mandi Finger began by saying in her presentation, she would be talking about evaluating effective population size, and what is the known extent of genetic threats to Delta smelt at present; hopefully she will provide a little bit of a ray of hope.

Finger Slide 4First, she explained why genetic diversity is important and needs to be preserved.  “Natural selection acts on genetic variation in a population,” she said.  “Genetic variation allows adaptation to a changing environment, and clearly the Delta environment has been changing so we want to preserve as much genetic diversity as possible so that the Delta smelt can continue to adapt to the Delta.”

There are a few forces that act on the amount of diversity in a population.  Inbreeding and genetic drift will remove diversity, and immigration and mutation will add diversity, so these are competing forces in a population,” she said.  “Genetic diversity and conservation also have a very important link, and as such, it is recognized by the IUC and is one of three forms of biodiversity deserving conservation.”

Finger_DeltaSmeltNe_Page_05Dr. Finger presented a graph with population size and genetic variation plotted.  “Ideally, if you have a large population size with plenty of diversity and then you have a crash or decline, automatically you are reducing the amount of diversity in a population, and the population size can and probably will rebound more quickly than the overall diversity because it is restored through mutation or immigration.  In the case of Delta smelt, we probably have a single population so they can only be restored through mutation.”

She added that the duration and the severity of a bottleneck or decline affect the amount of diversity removed during such a decline or bottleneck, so you want to keep that decline to a minimum and you want to allow the population to rebound as quickly as possible.

Effective population size is a parameter that allows us to predict the loss of diversity using the relationship between added forces and removal forces of genetic forces in a population,” she said.  “So we can think about NE as being the size of an ideal population that loses heterozygosity due to these forces that remove diversity at the same rate as the real population that we’re concerned with.”

Finger_DeltaSmeltNe_Page_07Dr. Finger said that the ideal population is only hypothetical population and almost never exists; it has assumptions built in, such as random mating, infinite population size, and no migration, mutation or selection.  “Departures from this ideal population lower the effective population size of a real population, and there are aspects of life history that lower effective population size, such as unequal family sizes. An example is with elephant seals; males typically mate with a lot of females, so there are a lot of males that are not adding their genes to the next generation so the effective population size gets lowered.”

A population bottleneck is a sharp reduction in the size of a population due to environmental events such as earthquakes, floods, fires, disease, or droughts, or human activities, and that results in a reduction in diversity.  “If you have a population with a lot of diversity, goes through a bottleneck, and you don’t have as much diversity come out on the other side, so fluctuations in population size, demography, and life history aspects will have an effect on effective population size and thus the amount of drift over time.”

Dr. Finger said that if a real population loses diversity at the same rate as an ideal population of a hundred, the real population has an NE of 100, but that is almost never the case, even if there are 1000 individuals in the real population.  “NE is typically far lower the actual census size,” she said.

There’s a lot of controversy and debate over the NE to N ratio, she said.  There’s also an information bias, because rarely is the effective population size of a healthy population studied; most effective population size studies are looking at populations that aren’t doing very well.  “The average is about .1, so that means that if the effective population size is 1000, your census size is 10,000.  But this relationship is uncertain and population specific, and some highly fecund species typically have far lower NE to N ratios.”

Finger_DeltaSmeltNe_Page_10Dr. Finger said that NE is important because theory predicts that populations with an NE greater than 1000 will maintain almost 100% of their diversity over ten generations.

However, NE is arguably the most important and most difficult to evaluate directly.  There are many definitions of NE, many different time frames over which you are talking about NE, and lots of different estimators – you can estimate it demographically and genetically.  People spend their whole careers studying this parameter.”

Dr. Finger said for her independent analysis, she used two estimators: NELD which is measured from a single sample from one cohort, and NEV, which is variations in population size which is measured by taking two samples and measuring the intervening NE between those two time points.  “Just focus on the fact the NEV will recover and decline more quickly than NELD, but in large stable populations, they will be similar.”

For the study, about 2600 Delta smelt were sampled from 2011 to 2014 cohorts, with samples grouped into the cohorts in order to have a single generation for this estimation.  Twelve microsatellites that they developed in their lab were used; Dr. Finger then estimated the NE and NELD for each cohort.

Finger_DeltaSmeltNe_Page_14Dr. Finger then presented the results.  The upper 95% confidence intervals were all infinite, so for this slide and the next slide, she will be reporting the lowest 95% confidence interval, because it can be assumed that it’s the most conservative, lowest value for NE that we may be seeing in the smelt, she said.

She noted that the values change rapidly over time, which is just a product of the variation in the estimation methods.  “I don’t want you to look at this and think 2011 was great; we don’t have that 1:1 correlation yet, so consider this an index,” she said.  “The red line is indicating 1000, our minimum threshold and that’s what we really want to have our NEV above, and it turns out that it is above that, which is good news.”

Finger_DeltaSmeltNe_Page_15She then presented the results for NEV, reminding that this one is a little more changeable than NELD.  “Some of the 95% lowest confidence intervals were in fact below the 1000 threshold,” she said.

Dr. Finger then gave her interpretations of the data.  “My interpretation with this data is that Delta smelt are not immediately threatened by reduced evolutionary potential, but they probably are near that threshold,” she said.  “Genetic factors are probably not the main reason for Delta smelt decline, and for those of you who have been following what’s going on in the Delta, we all know that there are pretty serious threats that are environmental and now demographic.  Genetically speaking, Delta smelt are recoverable, but we don’t want to get to the point of no return.  We don’t want to enter the extinction vortex where we’re losing genetic diversity rapidly.”

She reiterated that the information is just one piece of the puzzle.  “It’s is heartening but I’m not saying that Delta smelt don’t have serious problems.”

Dr. Finger then discussed how not to use this information.  “This is a short-term data set.  The effective population size and diversity were likely far greater before the collapse, we all know they have declined, maybe they are somewhere else now, but the fact remains, I’m convinced and everyone’s convinced that they are in serious decline and that they have collapsed.  So even though the effective population size is probably over 1000 now, maybe it was far greater before.”

NE cannot be used to inform water operations in real time, she emphasized.  “Not only does it require a certain number of samples to actually genetically estimate as you need to get DNA from the tissue, it also requires genetic analysis, you need to have enough samples, and there is a lag time between demographic changes and changes in NE.  In the case of Delta smelt, it probably won’t be as great as some other long-lived fish, because they are an annual fish, so it actually could change fairly rapidly in the case of Delta smelt.”

Dr. Finger then gave her suggestions.  “Focus on maximizing abundance so we can maintain and prevent further loss of diversity; all the genetics in the world aren’t going to bring smelt back from having no habitat,” she said.  “Also consider alternative ways to monitor Delta smelt, such as the smelt cam and environmental DNA to determine if they near the pumps even though we didn’t catch one – that’s valuable information, even though it can’t estimate abundance.  And also coordinate with other datasets.  There are a lot of us here on campus that are studying Delta smelt and even using the same samples, so I think it’s probably worth combining these datasets and somehow making a value-added dataset.”

Finger_DeltaSmeltNe_Page_02The markers we used for Delta smelt didn’t have a lot of power so we are going to be collecting genomic data, which is far more powerful, and also using historic samples so we can really investigate these longer trends, because taking a sample of a population that’s been declining for 30 years isn’t going to tell us as much as we really want to know,” she said.  “Then there’s some little goodies that could be in genetic data, you just never know, so I’m going to be looking for a sex marker and determine if there’s a genetic basis for residency.”

And with that …

Click to go back to the top.

RANDY BAXTER: Drivers of Decline: Conceptual Models from the Management Analysis and Synthesis Team

An overview of the Delta smelt conceptual model

Randy BaxterBIO: For over 27 years, Randall Baxter has worked for the California Department of Fish and Wildlife (DFW), first on the San Francisco Bay Study, a long-term fish and invertebrate monitoring project, then leading the Splittail Investigations Project and more recently as a supervisor for several long-term monitoring projects.  He has contributed as a member of the IEP Management Team, the Pelagic Organism Decline Management Team and most recently the Management Analysis and Synthesis Team (MAST).  He provides research-based knowledge to Department for the management Delta Smelt, Longfin Smelt and other estuarine species of concern.

In this presentation, Randy Baxter provided a general overview of the Delta smelt life history conceptual model developed by the IEP Management, Analysis, and Synthesis Team (or MAST).

Baxter 2_Page_04Although the evolution of the MAST models goes back further, I’m going to begin with the Fall Los Salinity Habitat (FLSH) Adaptive Management Plan which provided the conceptual model that you see here,” he said.  “At the time, it accounted for all the factors in the fall low salinity zone analysis, in particular the idea that dynamic processes occurred within a static abiotic landscape within the Delta.  The review panel and the synthesis team itself, in looking over what we put together for the FLSH, identified some issues: the graphic was complicated and didn’t show the relationships that they wanted us to show, and it became apparent in the process that seasons other than the fall were important, and that the low salinity zone wasn’t the only place where important Delta smelt issues were happening.  In essence, they wanted a full life cycle conceptual model and we felt that that was appropriate too.”

So they set out to develop a full conceptual model that would fully explain the assumptions that went into it.  “The model itself must be able to both present the relationships and identify hypotheses that needed to be tested that are critical for the understanding,” he said.  “In order to do that, we developed a two stage approach.”

Baxter 2_Page_06In our presentation of the model, the first is this general summary of the Delta smelt life cycle model.  The inner portion has similarities to some of the POD species models.  In this circumstance, the environmental drivers packed within and across a stationary landscape, and the landscape attributes of bathymetry and channel configuration caused changes in these habitat attributes, and then the environmental drivers influence the habitat attributes, and the habitat attributes are the factors that Delta smelt are actually responding to.”

Baxter 2_Page_07This was broken up into four life stage quadrants that also were related to seasons of the year,” he said, acknowledging that the first stage conceptual model gets a little more complicated when some of the factors in each of the levels are displayed.  “We needed to develop some descriptions of the relationships between environmental drivers, habitat attributes, and Delta smelt responses, so we begin with an introductory section where we talk about background, about how we think the Bay Delta system works in general, and how these environmental drivers influence habitat attributes, and then go on to discuss the habitat attributes and Delta smelt responses in the context of hypotheses that describe how we think the Delta smelt responds to environmental conditions.”

There are some important caveats, he said.  “We didn’t always know the connections between some of these environmental drivers and habitat attributes.  We needed to make some assumptions, and in those cases, we tried to fully explain them.  Not all the possible hypotheses were tested or testable; in some cases, we didn’t have any data to conduct the test, even though we thought the hypothesis was important.  And in some cases, the processes might be important in some years and not in others.  Nonetheless, we included all of what we thought were relevant habitat attributes into our second stage models, or life stage models.”

Baxter 2_Page_11Mr. Baxter then presented the adult to larval life stage model that occurs in December through May, noting that there are some overlapping seasonal boundaries in the model. “In this circumstance, we’re indicating that food availability, toxicity, predation risk, entrainment risk, and water temperature are potential influences on survival, maturation, and fecundity of adult Delta smelt.  The red dots on this slide indicate that a hypothesis that were evaluated for the report.  The absence of a red dot indicates that there was not sufficient data to evaluate the hypotheses, and the dots on the arrow indicate that there’s a general discussion of the stage to stage survival here based on abundance data.  This will occur through all the Delta smelt life stages because we have good abundance information.

Baxter 2_Page_12He then presented the conceptual model for the egg and larval to juvenile transition period. “Here we were able to evaluate all the hypotheses,” he said.  “Food, predation risk, and water temperature were important factors.  Entrainment risk and transport direction are two sides of a hydrodynamic coin that cause death or dispersal of Delta smelt larvae.”

Baxter 2_Page_13He next presented the conceptual model for the summer season, juvenile to sub-adults.  “We were able to evaluate most of the habitat attribute hypotheses: food availability, temperature, and predation risk, and at this point, toxicity from harmful algal blooms picks up, because that’s been a factor in the estuary for the last dozen years or so.”

Baxter 2_Page_14Last, he presented the conceptual model for the fall, the sub-adult to adult stage.  “We were able to evaluate most of the hypotheses that we generated, they are similar to the previous hypotheses with the addition of size and location of the salinity zone added in.  Again the red dots indicate all the hypotheses that were testable and tested.”

Mr. Baxter said that a number of data gaps were identified:

  • Contaminants and toxicity effects: “A complicated issue and lots more needs to be done. It’s a changing landscape.”
  • Entrainment and transport: “We know a lot, but the fine details are still waiting additional modeling.”
  • Predation: “Predation was not able to be evaluated in a number of steps. Although we know in general the diets of a number of the predacious species in the estuary, we don’t know their effects on Delta smelt or where those occur, and whether it’s more important in one life stage or another.”
  • Food: “For a lot of our surveys, when we catch fish, we also have plankton sampling and we catch plankton and we can associate the diet that we develop from those fish with prey field from the zooplankton sampling, but that’s not universal; there are a lot of samples of fish diet data where we don’t have food densities that are contemporaneous or in the same location.”
  • Harmful algal blooms: “They have been problematic in the past and are still occurring. We need certainly targeted quantitative sampling or at least to continue what sampling we’ve had.”
  • Quantitative life cycle model needed: “We’re all looking forward to helping populate a quantitative life cycle model and feel that that would be a big step in getting through some of the last details in understanding Delta smelt.”

Mr. Baxter concluded by noting that more information on the MAST model for Delta smelt is available at the IEP website here:  Baxter 2_Page_02 or by clicking here: Delta Smelt MAST Report

Click to go back to the top.

DR. NANN FANGUE: Why thinking mechanistically matters:  Teasing apart multiple and interacting abiotic stressors

Dr. Fangue dives into the temperature, salinity, and turbidity responses of Delta smelt

Nann FangueBIO: Dr. Nann Fangue completed a BSc in Marine Biology (1999), an MSc in Biology (2002), and a PhD in Zoology (2007) at the University of British Columbia.  She held an NSF postdoctoral fellowship before joining the faculty of Wildlife, Fish, and Conservation Biology at UC Davis in 2009.  She is currently an Associate Professor of ecological physiology. Studies in the Fangue Fish Conservation Physiology laboratory are largely focused on California native fish species.  We apply classic and contemporary physiological tools to understand the habitat requirements of native species.  We couple molecular, biochemical, and whole-organism measure of performance to elucidate connections between environment, physiology, and ecosystem function.  The Fangue lab is composed of a large and diverse research team (postdoctoral scholars, graduate and undergraduate students, and technical staff) with expertise in field collection, rearing, and physiological and behavioral studies of fishes.

Dr. Nann Fangue then gave a presentation on the studies that have been done on the Delta and longfin smelt with respect to temperature, salinity, and turbidity responses conducted over the last few years.  “I titled the talk, ‘Thinking mechanistically”, because I think it’s important to consider these mechanistic responses and the more we can understand those, the better we can understand and make predictions about what we think are likely outcomes for Delta smelt or longfin smelt,” she said.

Fangue smelt_Page_03Organisms cope with challenges in their environment through underlying physiological processes, or mechanisms, which influence higher level functions.  “These sorts of responses are things that contribute to abundance and distribution patterns,” she said.  “You can think about a continuum from abiotic or environmental factors that influence everything from the level of the gene up to whole organism physiology and behavioral responses, clear to ecology and biogeography, so when I say mechanism, that’s what I’m talking about.  It’s the responses at the level of the organism that influence the distribution and abundance patterns.”

Fangue smelt_Page_04It’s an important framework for giving us predictive power to forecast biological effects and species responses,” she said.  “It’s very hard to describe complex systems like we have in the Bay Delta if we don’t understand the fundamental mechanisms that underlie these things, much like how can you describe and predict outcomes related to climate change if you don’t understand how carbon cycles, for example.”

The prescription for preserving Delta smelt and longfin smelt for a long time has been flow and habitat, but there are a lot of things that are parts of flow and a lot of components of habitat.  “If we understand the key components of each of those things, I think we can do better in terms of how we address flow and habitat.

Dr. Fangue said she now try and show as much data as possible on the Delta and longfin smelt’s tolerance to turbidity, salinity, and temperature.  “We have a lot of options for how we might study the influence of the environment on physiological processes, so I’m going to show you some measurements at the level of the gene all the way to whole organism function and a few things in between,” she said.


She started with data for the Delta smelt with respect to turbidity.  Studies have shown that when testing turbidity values between 0 to 250 NTUs, juveniles are able to feed pretty constantly up to a value of about 120 NTUs; then feeding performance is reduced.  “That gives you a range of feeding success for juvenile that differs from what we see in late larval fishes, so 12 to 80 NTUs seem to be the optimal turbidity levels for larval fish, and feeding declines as you move both into clearer water and at higher levels of turbidity,” she said.

Fangue smelt_Page_09They have also done some turbidity by salinity crosses and a matrix of exposures.   She presented a slide showing some molecular markers that are indicative of salinity stress or a very generalized stress response with respect to salinity and turbidity.  “Salinity markers are things such as sodium, potassium, ATPAs, or HSP-70 which is a generalized stress response marker,” she said.  “Delta smelt respond to salinity and those markers respond, and what you see is that at 2 and 6 ppt, you get very little stress relative to as you move outside of that salinity range.  There is a similar pattern for sodium, potassium, ATPAs, and that would be what you would expect if these animals prefer lower salinities.  There isn’t an impact of turbidity on these molecular markers.”


Fangue smelt_Page_11

They have also studied the salinity tolerance of Delta smelt at a variety of life stages both at the acute and the chronic scale.  Dr. Fangue presented a graph of acute salinity tolerance, noting that the time of exposure is on the x axis, and the y axis is showing cumulative proportional survival; they have done this for adults and juveniles at 2 ppt, 18.5 ppt and 33.5 ppt or roughly sea water.

What you can see is that these animals survive really well at low and medium salinities, but we do get some mortality in some individuals when we take them up to the salinities on the scale of seawater, so survival is affected in the high exposure groups,” she said.  “Adults and juveniles have a similar pattern of response and there an interesting critical window early on where some animals are osmotically disturbed and don’t survive, but then many more or many of them do and recover.  That suggests that these animals are robust to pretty high salinity levels, but we know that salinity osmoregulation is a costly thing.  There could be sublethal impacts with respect to salinity stress, and that might be important when thinking about habitat association and the known associations of Delta smelt to lower salinity waters.”

Fangue smelt_Page_13Osmoregulation is the process of maintaining an internal balance of salt and water in a fish’s body.  Dr. Fangue noted that the mechanisms are different in freshwater and sea water.  “Our fresh water fish is going to be more salty than the freshwater environment, whereas the seawater fish will internally be less salty than the seawater environment,” she explained.  “What that means is a freshwater fish has to actively uptake ions across the gill in a freshwater scenario; in a salt water scenario, these animals have to actively excrete ions across the gill, so they have to pump ions in opposite ways depending on the salinity environment that they find themselves in.  The notable thing to remember is to that pumping ions is one of the most expensive things that we do energetically, so these can be very challenging osmotically for fishes.”

Estuaries are dynamic salinity environments, so fishes that live there, have to be able to in many situations express both of these types of phenotypes,” she added.

Fangue smelt_Page_17The question is do we think Delta smelt are euryhaline or stenohaline, because the data suggests that they’re pretty robust to quite a wide range of salinities, although we know that in nature, they seem to like much lower salinities and a much more narrow range.  So they looked at the underlying biological processes that are thought to be important and conferring physiological salinity tolerance, and is there any evidence that sublethal osmoregulation stress could contribute to this habitat limitation, despite these animals having this robust tolerance to high levels of salt.

So they exposed fish to 2, 18, and 34 ppt salinities, and measured plasma osmolality.  “What you see here is from the baseline or from control through about 300 hours post exposure, you see an initial disturbance, in this case in plasma osmolality, or how salty the insides of these organisms are, and then you see a recovery; there is an elevated disturbance at 34 ppt but still a recovery, so these animals are adjusting their physiological capacity using mechanisms like ion excretion across the gill.”

Slide 18They have a lot of well developed transcriptomic resources for these species and they have a growing database on what the suite of genes that are changing through time in these organisms look like, she said.  She presented heat maps of the different genes, noting that the takeaway here is to show that the picture is very different depending on the salinity.  “As you increase salinity with these animals, their transcriptome is responding and in very complex ways.”

Fangue smelt_Page_19A simple way to look at it is just to look at the overall pattern of how much the transcriptomes are changing when these animals are exposed to different salinities.  “Here is our pre-experiment and our 0, 2, and 6 ppts, and this give you a sense for the trajectory or the overall kind of way that the transcriptomes are changing through time,” she said.  “The trajectories at the lower salinities are divergent when you compare them to conditions that are hyperosmotic, so 12 ppts and 18 ppt, so very different gene expression signal under these different conditions.”

Slide 20Dr. Fangue then displayed some data for a couple of genes that are important with respect to energy.  “These sublethal responses to salinity are energetically demanding, so you would predict that metabolic-related genes would be the ones that would be changing,” she said.  “What you can see are two genes, one involved in altered fat and glucose metabolism and the other involved in cellular growth, and what you see is that at 2 and 6 ppt here and here, those patterns are different than 12 and 18 ppt; these data are suggestive of the fact that when you start move these fish from salinities of 6 to higher levels, the pattern is really different and perhaps because these genes are involved in energy metabolism, they might be paying a sublethal cost to be able to do these things.”

To follow up on the data, they are studying whether they can measure a detectable metabolic cost under different salinities.  “You can do that using respirometry, and we’re starting to do some of that work to ask, can we detect an energetic signature that underlie these responses to salinity.”

So I showed you that these fish can osmoregulate across quite a range of salinities, but that coping with osmotic stress involves these large scale transcriptional changes in self-signaling and reestablishment of ionic balance,” Dr. Fangue said.  “It may be that altered metabolic in normal cell processes are suggestive of sublethal costs and those things might be very important to keep in mind.”


Slide 23She then discussed thermal tolerance in Delta smelt.  “We’ve made acute and chronic measurements of sensitivity across a large range of Delta smelt life stages, and what we tend to see is that as you move up in temperature, proportional survival declines, but it does so differentially among the different Delta smelt life stages,” she said.  “What we see is that juvenile fish have a higher tolerance than adults, than do post spawning adults, and if I had the larval data on here too you’d see that larval fish are actually more thermally tolerant than juveniles, so we see this stage-specific tolerance and thermal sensitivity in Delta smelt.”

Slide 25They then looked crudely at the environmental temperatures that these animals are found across, and mapped thermal sensitivity measures for late larval and juvenile and adult fishes.  “You can see that adults, for example, have a reasonable buffer between habitat temperatures that they experience and their thermal tolerance limits, but these things get a little closer for juvenile fish, and in fact, juveniles are more likely to feel in situ temperatures closer to their thermal tolerance limits.  There are some extreme events that do occur that are beyond these animal’s thermal tolerance limits at the moment.”

Slide 26There is a lot of data on gene expression profiles for thermal sensitivity in Delta smelt as well, and the story is the same, she said.  “As temperatures increase, as we move from left to right on the screen, you see a different pattern.  These animals are responding and the transcriptomic responses are complicated.  I wanted to use this to point out that gene expression profiles in some of these sublethal measures, if you understand them well, they can help you understand not only mechanisms but they can be leveraged to figure out where are these thresholds where the fish can compensate or use their plasticity to cope with environmental stressors and where do these things begin to fail; so what level of sublethal stress is too long or too much on a more persistent basis.  These sorts of things I think are going to be more increasingly important in integrating into restoration efforts.”

There is a nice dataset that compares Delta smelt and longfin smelt thermal sensitivity; it’s new data that’s just about to come out now, she said.  “We were able to compare the 50 days post-hatch thermal sensitivity of Delta and longfin smelt.  We measured thermal tolerance, we measured their metabolic responses, and we did some transcriptome sequencing, so I’m just going to hit a couple highlights for you.”

Slide 29She displayed a graph that plotted temperatures in the field for wild longfin smelt and wild Delta smelt collections.  “You can see that they are offset from one another,” she said.  “You tend to catch longfin smelt in cooler waters, and you can then associate that with the measures of thermal tolerance, in this case Delta smelt CT max and longfin smelt CT max that we determined in the lab.  What these data show you is that longfin smelt are more sensitive to elevated temperatures compared to Delta smelt.  It matches very nicely with their field distribution patterns.”

Fangue smelt_Page_30She presented a graph of the metabolic rate, noting that while it may look simple, it was a monumental effort to create it.  “Measuring metabolic rates in fishes is challenging enough, but doing it with Delta smelt and longfin smelt is hard to do,” she acknowledged.  “The picture is what a respirotometry chamber would look like – a little field vessel with a longfin smelt inside.  So all we’re doing in this experiments is we are sealing this vessel and we are measuring the fishes ability to draw down oxygen metabolically.”

She noted that oxygen consumption is shown on the x axis, and the two temperatures they tested at is shown on the y axis.  “This is the Delta smelt data and this is what you’d expect.  You’d expect as temperatures move up from 14 to 20 degrees; you would expect metabolic rates to also go up.  We don’t see that for longfin smelt; the interpretation is that these animals are facing some sort of limit, and presumably it might be some limit to their aerobic capacity or some mismatch between oxygen supply and oxygen demand.  When we looked at the transcriptome, there was a lot of evidence that suggests that that might be the case.  This is an interesting finding for us.


Dr. Fangue then gave her concluding thoughts and implications.  “We’re certainly dealing with a complex system and so I think that a mechanistic understanding and doing very rigorous science to tease apart what the true drivers are is really important.  I think focusing on what the key components of flow and habitat can really help us do more with less.  I think having lots of flow and lots of wonderful habitat is maybe a luxury;  I’m not sure moving forward we’ll be able to have as much of that as we might want, so trying to do more with less, if we understand what we need is only helpful.”

Fangue smelt_Page_32We still have little information with respect to multiple and interacting stressors and so keep in mind that stressors interact with one another – abiotic stressors, toxicants, all of those things – so you need to understand if these stressors are additive, if their synergistic, what those relationships are if you want to have predictive power moving forward.”

I told you all about plasticity or what can the animals we have today do with what they have, but I think it’s important to keep in mind that moving forward, we should start thinking about studying adaptive capacity, because there are some examples in fishes where thermal tolerance, for example, responds to environmental drivers and temperature on a relatively short time scale and that could be very helpful.  I think trying to get a handle on some of that would be beneficial.”

And I think with that …

Click to go back to the top.

For more from the Delta smelt symposium …

Click here to read more posts from the Delta smelt symposium.

Click here for all presentation videos from the Delta smelt symposium.

Coming tomorrow …

Presenters discuss the potential causal links to the Delta and longfin smelt declines.

Daily emailsSign up for daily email service and you’ll never miss a post!

Sign up for daily emails and get all the Notebook’s aggregated and original water news content delivered to your email box by 9AM. Breaking news alerts, too. Sign me up!

Print Friendly, PDF & Email