Sunday, December 29, 2013

Cutthroat genetics, part 2

     So, where we left off, our cutthroat researcher Kevin Rogers, along with others, put forth the theory that Green lineage fish represent the Colorado River cutthroats native to the Gunnison and Colorado, and Blue lineage fish are the Colorado River cutthroats native to the Yampa and White. When the new genetics information was combined with historical information, this was the theory that best explained what we were seeing.
     Now a couple of new questions arise. If this theory is true, is there any way to prove it? Also, if it is true, where does that leave the true Greenback - the fish native to the South Platte? Does it even exist  any more or is it actually extinct?
     Enter Jessica Metcalf again, and the Museum DNA project. As it turns out, a handful of early surveyors collected fish specimens from Colorado rivers and streams, spanning the period from 1856 to 1889. The specimens still exist, in places like the Smithsonian. For the first time in history, we have the technology to extract DNA from those specimens, perform the same AFLP analysis on the samples, and compare those results to today's populations. Metcalf published her findings in 2012:

Historical stocking data and 19th century DNA reveal human-induced changes to native diversity and distribution of cutthroat trout


Many species are threatened with extinction and efforts are underway worldwide to restore imperilled species to their native ranges. Restoration requires knowledge of species' historical diversity and distribution. For some species, many populations were extirpated or individuals moved beyond their native range before native diversity and distribution were documented, resulting in a lack of accurate information for establishing restoration goals. Moreover, traditional taxonomic assessments often failed to accurately capture phylogenetic diversity. We illustrate a general approach for estimating regional native diversity and distribution for cutthroat trout in the Southern Rocky Mountains. We assembled a large archive of historical records documenting human-mediated change in the distribution of cutthroat trout (Oncorhynchus clarkii) and combined these data with phylogenetic analysis of 19th century samples from museums collected prior to trout stocking activities and contemporary DNA samples. Our study of the trout in the Southern Rocky Mountains uncovered six divergent lineages, two of which went extinct, probably in the early 20th century. A third lineage, previously declared extinct, was discovered surviving in a single stream outside of its native range. Comparison of the historical and modern distributions with stocking records revealed that the current distribution of trout largely reflects intensive stocking early in the late 19th and early 20th century from two phylogenetically and geographically distinct sources. Our documentation of recent extinctions, undescribed lineages, errors in taxonomy and dramatic range changes induced by human movement of fish underscores the importance of the historical record when developing and implementing conservation plans for threatened and endangered species.

     So, there are a few big game-changers here. First, the sentence about six divergent lineages. Remember the old traditional model that I described at the beginning of the last post, about the three ranges of Greenback, Colorado River, and Rio Grande cutthroats? That is out the window now. Gone. It has been replaced by this:

     We have six historic lineages of cutthroats that are definitely distinguishable from each other genetically. Two of them are believed to be extinct - the fish native to the Arkansas River (historically referred to as the Yellowfin), and the fish native to the San Juan - which were never named. There is only one population in the state - alas, in the world - that matches the historic Greenbacks collected from the South Platte watershed. So in the span of a few years, we've gone from thinking (erroneously) that we had a good  number of Greenback populations up and down the east slope, to now knowing that we only have one population that resembles the fish that were there originally. This is a huge setback for Greenback recovery.
     Even before this study was complete, our cutthroat researcher Kevin Rogers had the foresight a few years ago to realize that there was something important about this single unique population, and had already been working with our hatchery managers to build a captive broodstock of these fish in our hatchery system. So we have these fish in captivity to guard against something catastrophic happening to the stream where they live, such as a large fire and/or extreme flooding. 
     We're hopeful that our Durango counterpart, Jim White, will manage to find a San Juan lineage population somewhere, but there aren't many stones still unturned in his part of the state.
    Notice that we're being careful not to call them subspecies. This whole issue is a perfect example of the quesiton of how exactly to define a subspecies. Geographic separation is the most commonly stated element defining subspecies. But are these subspecies or strains? Ultimately, these questions are up to the US Fish and Wildlife Service, because they are the entity that determines what populations warrant protection under the Endangered Species Act. The jury is still out on that but we are expecting some direction soon.
     Another question that arises, and one that certain sectors of the angling public definitely ask is, who cares? Why does this matter? Well, to some people it doesn't matter. But, if we've got populations of fish that are unique in the world and are still persisting in the drainages that they have lived in since the last ice age, isn't that valuable? If we are doing things that are leading to the extirpation of some or all of those populations, shouldn't we do everything we can to reverse those declines that we caused? For me, the answer to both of those questions is an emphatic "Yes."
     I'll wrap it up with a nice shot (taken by Kevin Rogers) of a fish from one of my native populations. This is a population Green lineage fish, in the Colorado River basin, that does not carry a Grand Mesa haplotype - which means that we have every reason to believe that this is the aboriginal fish, occupying this stream successfully since long before our time. There aren't very many of those populations, and everything I do in terms of cutthroat conservation from here going forward will radiate out from these fish in this stream.
     Thus ends the first year of this blog, and I'm happy with the level of success. We'll end up at right about 16,000 views for the year, which may not sound like a lot but I wasn't as consistent as I could have been so I'm happy with that. The whole purpose of this is to do a better job conveying information to the folks I work for (you) and so please be sure to let me know of any topics you'd like to see me cover, either in the comments or via email at Here's to a good 2014.

Wednesday, December 18, 2013

Cutthroat genetics for the layperson

     I thought I'd give folks a rundown of the current state of affairs in the world of cutthroat trout genetics. It is a very confusing situation. I get asked by anglers a lot about where they can go to catch a cutthroat that is native or genetically pure. The answer to that question is quite complicated and involves a lot of history. The story that I'll give you has some generalities and glosses over some details but this is an endless rabbit-hole if you get bogged down in exploring every detail. You have to take this slowly, piece by piece, for it to make sense, and hopefully I can succeed in making it understandable.
     We have to start with the world as we knew it before 2007. Up to that point we understood that there were three subspecies of cutthroat trout native to Colorado: the Rio Grande, the Greenback, and the Colorado River. Their distributions are shown below:
    Credit for all the graphics in this post goes to Kevin Rogers.  In the map above, the green area is where the Greenbacks were native (east of the Continental Divide), and so on. This is a very clean and understandable distribution of subspecies, split out along drainage divides. It is important to keep in mind that especially with Greenbacks and Colorado River cutthroat, it is basically impossible to differentiate between the two in the field. Rio Grande are a little more different than the other two, but still there is a lot of visual overlap in the appearance of all three of these subspecies. So up to this point, the fish's LOCATION has been just as important as its appearance in identifying which subspecies it belongs to.
     Greenbacks were thought to be extinct until 1957, when an isolated population was found. Other remnant populations were found in 1965 and 1970. There were no stocking records in existence for the streams where these fish were found, and the best science at the time identified them as being Greenbacks, native to the Arkansas and South Platte. This led to several decades of conservation and reclamation work in the Akansas and South Platte river basins, to reestablish populations of these fish using progeny taken from these remnant populations that were found.
     So everything was moving along more or less swimmingly until 2007. A Geneticist from CU named Jessica Metcalf published a paper that rocked the world of cutthroat genetics. She used a new technique called AFLP - Associated Fragment Length Polymorphisms - to look at the genetics of our cutthroats in a higher level of detail than anyone ever had before. This is one very important thing to be aware of when trying to follow this progression: these genetic techniques are on the cutting edge of the technology that's available, so every decade or so there are new tools available that reveal things that we have never been able to see before. Here is the abstract from her paper:

Across the great divide: genetic forensics reveals misidentification of endangered cutthroat trout populations


Accurate assessment of species identity is fundamental for conservation biology. Using molecular markers from the mitochondrial and nuclear genomes, we discovered that many putatively native populations of greenback cutthroat trout (Oncorhynchus clarkii stomias) comprised another subspecies of cutthroat trout, Colorado River cutthroat trout (Oncorhynchus clarkii pleuriticus). The error can be explained by the introduction of Colorado River cutthroat trout throughout the native range of greenback cutthroat trout in the late 19th and early 20th centuries by fish stocking activities. Our results suggest greenback cutthroat trout within its native range is at a higher risk of extinction than ever before despite conservation activities spanning more than two decades.

     This paper identified a number of populations on the east slope that appeared to be Colorado River cutthroat when looked at with the new analysis. Up until that point our understanding was that these populations were Greenbacks. So the stories started hitting the popular media that seemingly incompetent biologists had been stocking "the wrong fish". What this paper did was redefine what the "right fish" and "wrong fish" are, and again, you can't tell by looking at them in the field.
     So here is what the situation looked like after the 2007 study:

     On this chart, the blue populations are what were referred to as "Lineage CR" - presumably Colorado River cutthroats. The green populations are "Lineage GB" - presumably greenbacks. We're now calling them Blue Lineage and Green Lineage, for reasons I'll explain below. But it was a major blow to find out that some of these east-slope populations did not appear to be Greenbacks after all. Especially Apache Creek down south - that was one of the original populations "rediscovered" when it was believed that Greenbacks were extinct.
     Now you have to start considering what happened with fish stocking historically in Colorado. In the early 1900's, many millions of eggs were harvested from Trappers Lake and stocked all over the state. At that time, a cutthroat was a cutthroat and all fish were considered the same from Yellowstone south. Distinct subspecies of the fish had not been recognized.  In fact, during a period when the Trappers Lake cutthroat population appeared to be struggling, Yellowstone cutthroat were stocked into Trappers to bolster the population. Up until that point, Trappers was the largest intact remaining population of pure Colorado River cutthroats, but once those Yellowstones went in, that was no longer the case.
    Another problem with that period in history is that for a long time we weren't aware of much in the way of stocking records. There was a "dark age" of fish stocking history in Colorado from about 1900 or so to 1950. Fish were dumped all over the place and no records were kept - apparently. However, Chris Kennedy, one of our colleagues who works for the U.S. Fish and Wildlife Service, has done a ton of spectacular historical forensics work and uncovered many old stocking records that were not previously known. Some of them were in old ledger books buried and literally forgotten in the basement of the state archives. He has pieced together a stocking history that shows that at least 26 million fish were reared and stocked from Trappers Lake all over the state, just from the years 1914-1925 (see below). This is also new information to us.

     One of the things that the 2007 study didn't really highlight is the fact that one Lineage GB population was found west of the divide, in the Gunnison Basin. You can see that one in the figure above. The presence of the Lineage CR fish east of the divide was easily explained by the new historical stocking information available to us, but there was no historical explanation for east slope fish moving west - we are not aware of that taking place on any large scale historically.
     So, in the aftermath of this study, we continued collecting genetic samples and using AFLP analysis for as many of our cutthroat populations as we could. In particular we were looking for more occurrences of Lineage GB fish popping up west of the divide. Sure enough, we found a lot of them - as illustrated here:

     In a few years' time, we had identified over 60 populations on of Lineage GB fish on the west slope. Remember that these are the fish that Metcalf's 2007 paper was treating as Greenbacks, native to the east slope. If they were native to the east slope, why would we find so many populations of these fish on the west slope? Also, here's another interesting thing: we're finding these populations all over the west slope EXCEPT for in the Yampa and White River basins (and the San Juan but that's a different story). To this day we still have not identified a single Lineage GB population from either of those drainages - everything is Lineage CR. So this has become an ever-larger question.
     Enter Chris Kennedy again with his historic investigations. He and others have also found old records of a lot of fish being raised from lakes on the Grand Mesa. In fact, from 1899-1909, we now have records that at least 29 million fish were hatched from Grand Mesa lakes and stocked all over the state, including many places on the east slope.

     So now, a new theory emerges: what if the differences that we're seeing between Lineage GB and Lineage CR fish is a difference WITHIN Colorado River cutthroats - that is, Lineage GB fish are native to the Colorado and Gunnison watersheds, and Lineage CR are native to the Yampa and White. Our historical stocking records support this theory. And if we're defining Greenback cutthroat as the fish native to the Arkansas and South Platte, then if this theory is true, Lineage GB does NOT equal Greenback. That is why we have started calling Yampa/White cutthroats "Blue Lineage" and Colorado/Gunnison cutthroats "Green Lineage", to move away from the Greenback name.

    I'm about half way through this story, so I'm going to stop it here and pick it up next week. Please be sure to let me know if you have any questions, either in the comments or by emailing me at If something is unclear to you, I guarantee you it's unclear to other people as well and I should try to do a better job explaining.

To be continued . . . 

Wednesday, December 4, 2013

We saw this coming . . .

     Ka-boom. Through the blowing snow in Grand County this morning if you listen closely enough, you can hear the low, dull sound of an implosion taking place. It's the sound of the Granby kokanee population crashing.
     We took our last kokanee eggs of the year on Monday, the 2nd. Our spawn season ended with a whimper. The total number of eggs we took from the Granby run was 357,425 out of 432 ripe females. My optimistic prediction for this year before the season was that we might get a half million. Granby needs to produce 1.2 million eggs just to sustain itself. Preferably, we'd like to use Granby eggs to stock other waters also, but that's obviously not happening any time soon. Below is the recent history of egg takes at Granby:

     Our first indication that this was going to happen was the trend in recent years of mysis densities in the reservoir. We sample mysis every year and estimate their density per square meter of water surface area. I like to look at the four-year rolling mysis density estimate because the average kokanee in Granby lives for four years. Here is that statistic, for the entire history of our mysis surveys at Granby:

     Mysis respond positively to high water years at Granby, and get knocked back during drought periods. The low point in 2004-2006 is the response to the Granby drawdown that took place during the drought of the early 2000's. When mysis densities are low, we get much better recruitment out of our kokanee stocking because there is less competition for the zooplankton that both mysis and kokanee eat. If you look at the "good" years of egg harvest, from 2004-2009, you can see that this recovery in the kokanee population was a response to that low-mysis cycle covering roughly the same time period. What we are seeing now is the fallout from the years of high mysis densities that peaked in 2010. 2013 was a lower-water year at Granby and the mysis density estimate was 280. We responded by loading Granby up with extra kokanee - we stocked 1,450,000 rather than the normal 1 million. So hopefully we got good recruitment out of those fish this past summer, but we won't see that in the spawning run until about 2015 or '16. What concerns me now is that the downward trend in this average appears to be flattening out, and for a real kokanee recovery to take place I think we need to see this statistic drop below 500 or less. If the snow keeps piling up, Granby will have another high-water year, mysis numbers will bounce back, and this trend will start heading upward again.
     Every year we run sonar surveys of our kokanee reservoirs to get an idea of the status of the kokanee populations and to predict what the spawning runs will look like. Our research section has a dedicated boat equipped with a scientific sonar rig for this purpose. It's basically a fish finder on steroids, with a lot more definition, giving you the ability to actually count fish and estimate their sizes. There is a lot of room for error in this estimate, but the trend definitely predicts what kind of kokanee run we're going to have. Here is the recent history of sonar surveys at Granby:
     The sonar surveys have been doing a good job of predicting the following year's egg take. So, the 2012 survey was showing roughly 1/3 the number of fish that were there in 2011. Lo and behold, our 2013 egg take was roughly 1/3 of our 2012 egg take. I don't have a 2013 number from this survey yet, but when I get it we will know if we've hit the bottom or if it will get even worse before it gets better. I'm betting on the latter.
     Now, the part that no one wants to talk about. We know that competition from mysis makes it very difficult for the kokanee population to persist. What about predation by lake trout? The recreational kokanee fishery is long gone from Granby. Serious kokanee anglers go elsewhere now.  However, we still need to maintain enough kokanee to provide a prey base for the lake trout, as well as survive to maturity to provide us with eggs. This is the irony of the misguided discussions you see on the discussion boards. The lake trout guys spin their arguments as some kind of kokanee-versus-lake trout struggle, when that is absolutely not the case at all. If you don't have a healthy kokanee population, what are trophy lake trout supposed to eat?
     As I've discussed before, I run a gillnet survey of the lake trout population every May. As expected, 2013 saw a significant decline in the body condition of the large lake trout that we captured. I handled some of the skinniest large lake trout that I ever have this year. Here is a data summary for the lake trout survey:
      I set each net for 6 hours, at the same locations each year. 2011 was the first year I surveyed the lake in this way. There were a couple big differences in 2013. The first was the number of large lake trout that I caught. This number was remarkably similar in 2011 and 2012 but took a big jump in 2013. I did have one net this year that caught nine fish over 24". I had not seen that before. However, even if you remove those nine fish, we still caught twice as many fish over 24" than in the previous two years. Does that mean we have a big increase in large lake trout? This netting survey does not have the degree of accuracy necessary for me to be able to say that. There is too much variation, thus room for error, in the catch rates. In addition, the lake was lowest in 2013, so this could possibly be a reflection of the fish being squeezed into a smaller volume of water. Are the numbers of large lake trout declining? No way. But I'm saying that based on professional judgement and opinion, not on statistical analysis.
     The most telling statistic here is the last two lines. Body condition for all lake trout captured remained the same as it has the past couple years. We know that lake trout smaller than 24" make a good living off the dense mysis population. However, when they grow beyond 24" they need to at least supplement their prey with a vertebrate food item, if not switch over completely. So the drop in average body condition of large fish is the big kicker. That is a large drop over a one-year period for a fish population that by nature changes slowly. And it is statistically significant. A body condition of 73.5 is a very poor fish.
     In an earlier post I had mentioned that it appeared to me that the variation in body condition was higher in 2013 among the larger fish. That is, we caught some really skinny ones, but there were also some fish still in decent body condition. The way to see if that is true is to look at the standard deviation in body condition for the samples from each year. So, for 2011 the standard deviation in this statistic for fish >24" was 8.4. In 2012 it was 8.3. In 2013 it was 8.6. So, my initial impression was wrong, and the variation in body condition was only very slightly greater in 2013 than in past years. We just had some ridiculously skinny fish. Here's an example:

     This fish is an embarrassment. This is a starving fish. I've seen plenty of anglers' pics of fish like this over the past year. These fish are telling us loud and clear that the predator population does not have an adequate prey base to support it. 
     I know that no one wants to kill large lake trout. If this is what the anglers want out of Lake Granby then we can continue down this path. But I want to make sure that everyone fully understands that choosing not to put pressure on the lake trout population at this point in time is a POLITICAL decision and there is not a single piece of biological evidence to support it.  One thing we can't do, is increase the numbers of rainbows being stocked to prop up the lake trout, thereby substituting a cheap prey base with one that is as much as ten times more expensive. In fact, the fishing for other species is likely to decline as the starving lakers turn to anything and everything to survive on.
     Another common misconception that pops up on the discussion boards frequently is that growth rates of lake trout in Granby are slow, period. End of story. The truth is that growth rates CHANGE based on predator-prey dynamics, and we have directly observed that change. Many people are familiar with the old tags that are still in some fish in Granby. Those tags were used in a growth study in the '90's, and this is where peoples' information about lake trout growth comes from. Here is the most important information from that study, which was conducted by Pat Martinez:

     This data is old, but it reflects a time at Granby that is very similar to our situation now. In 1998 the kokanee population had crashed completely and no eggs at all were taken at Granby. The kokanee had to be restarted with Blue Mesa fish.
     But here is the take-home message from that study: Lake trout growth rates CHANGE according to conditions in the reservoir. If there is no prey for the lake trout to eat, of course they are going to be slow. But they don't have to be slow as a rule.  Imagine a scenario with one, 24" lake trout in the whole lake and a huge kokanee population. Do you think that one fish would have an extremely slow growth rate? Of course not. Its growth rate would be through the roof. This is why I get frustrated when people say, "if you kill a 30-year-old lake trout, it's going to take 30 years to replace that fish." This is dead wrong because there is a 29-year-old lake trout waiting in the wings to take that 30-year-old's place. And one less large fish to compete with will increase that 29-year-old's growth rate. So it doesn't take 30 years to replace that fish. It takes one year. Could Granby's growth rates match what we see in Blue Mesa? No way - Granby simply isn't productive enough for that, and the mysis issue will always limit the potential. But could they be faster than they are now? Absolutely.
     Here's another thing to consider. In 1994 Wayne Hubert and a couple other guys published an article called "Interpreting Relative Weights of Lake Trout Stocks." In that article, they took data from 58 different lake trout populations throughout North America, across the range of the fish. The figure below is the most important part of the article:

     It's kind of hard to make sense of but bear with me. This is a cumulative frequency distribution. The X-axis, labeled Mean Wr, is the exact same equation that I use to rate body condition in the Granby fish. What this shows is that 100% of the populations studied had an average body condition below 135. That makes sense, because a relative weight of 135 is a very fat fish. About 65% of the populations had an average relative weight less than 100, and about 35% had an average greater than that. What I'm getting at, is look where Granby falls on this distribution - at the very bottom. Our fish are skinnier than 97% of the populations in North America. In fact, we may have the skinniest lake trout in the world.