I hope absence makes the heart grow fonder. After a long hiatus, I’m back with another research update. And, I have to say, it might be the most interesting (and hopefully most influential) project I worked on for my Ph.D.
But, first, let me set the stage for those of you who may be new to this blog. I work on brook trout ecology in the Loyalsock Creek watershed in Pennsylvania. The watershed is mostly forested, making it a great home for my Ph.D. research on brook trout population response to climate change. I’ve previously reported on other findings from my research showing that interbreeding between wild and hatchery fish was fairly minimal throughout the watershed and also some preliminary telemetry results showing that some brook trout seem to move from the small tributaries into the mainstem river after spawning season.
But, the question always remained, what happens to fish that get into the mainstem? There are a lot of predators in the mainstem and water temperatures in the summer far exceed brook trout thermal tolerance. So, many speculated that fish that got into the mainstem probably died within a few months. Moreover, our telemetry observations only found that a handful of fish seemed to have this migratory behavior. So, even if they did survive, could their behavior really drive any sort of population-level response?
The answer is a resounding yes.
I’m basing that response on a study we (myself, my advisor, and very importantly a collaborator in the statistics department) just completed that looks at genetic connectivity of brook trout populations across the Loyalsock Creek watershed. As you may recall from previous posts, maintaining and increasing connectivity among populations is one of the most important management tools we have for increasing population persistence and resiliency to future disturbance. And, we can measure the degree of connectivity between two populations by measuring the degree of genetic similarity. This isn’t so hard- we take a little fin clip from a bunch of individuals in each population of interest, from the fin we identify the genes present in each population, and then using some computer software we estimate the degree of connectivity.
Knowing if two populations are genetically dissimilar- and thus disconnected- is great, but it doesn’t necessarily explain why those populations are isolated from one another. Sometimes it’s easy. If there is a large waterfall that separates two populations, then it’s reasonable to assume that few individuals are moving back and forth between those populations and therefor connectivity is low. Other times it’s not so clear. There could be a hidden barrier (perhaps a road crossing with bad fish passage or an area with a steep slope), or it could be that our assumptions about what limits fish movement (and thus population connectivity) are wrong. That last point is important, because if we don’t know what we are looking for then we will never be able to identify and fix areas of stream that are reducing population connectivity or conserve areas that are important movement corridors.
So, we used some really fancy models (hence the phone-a-friend to the stats department) to essentially determine how various features in the watershed either resist or increase gene flow. We call this a riverscape genetics study- essentially seeing what features of the riverscape (which is like a landscape, only for streams and rivers) are responsible for producing the observed patterns in genetic connectivity. And, remember, individual fish are just bundles of genes, so this analysis is a proxy for determining which features of the watershed increased and decrease fish movement.
To run the analysis, we identify a bunch of variables we think could influence gene flow, and then let the model tell us whether there is actually a high probability that gene flow is influenced by each variable. So, we thought about it and decided to include 12 variables. This included some of the usual suspects like stream slope, road crossing density, and large barriers (like waterfalls), as well as some more unusual variables like distance to mainstem Loyalsock and a few things that essentially measure the location of a stream within the watershed. After it was all said and done, we found support for just four variables that influence gene flow in Loyalsock Creek, including:
Why am I so excited bout this study? First, for any fish biologists reading this post, the model we used is new, and I’m hoping it provides a framework for future riverscape genetics analyses (so, contact me for details!). Second, and most importantly, it definitively shows that the mainstem is not only brook trout habitat but may be some of the most important brook trout habitat in the watershed. Because larger rivers are thermally unsuitable for coldwater fishes during summer and don’t have large resident trout populations, they generally don’t receive the same conservation status as small tributaries. However, these rivers are critical migration corridors that are responsible for increasing population connectivity.
This study also gives some insights into how future disturbance could influence brook trout population connectivity. With climate change we are generally expecting increased floods and droughts- both of which will change stream flow patterns and could limit the ability of brook trout to move through the mainstem. This is particularly true given that there is only a small window of time where thermal conditions are suitable for brook trout to use the mainstem, and so disruption of flow for even a short period could have large effects on trout populations. Additionally, human disturbances that alter flow patterns, either through direct water withdrawals or watershed disturbances that result in a lowering of the water table, could influence flow patterns in larger rivers as well as increase the periodicity of flow in intermittent stream channels. So, if we want to maintain future brook trout population connectivity, we probably need to start thinking beyond just removal of physical barriers and conservation of natural stream flow patterns.
Finally, a word of caution. This study was conducted in Loyalsock Creek and, while some of the findings likely do translate to other watersheds, I would expect the results to change depending on the location. For example, as a largely undeveloped watershed, variables like road crossings and watershed development were not important for explaining population connectivity. These features undoubtedly influence brook trout populations, they are just uncommon in Loyalsock. But, I’m looking forward to this model being applied elsewhere and seeing how the results change across watersheds.
This one goes out to all my readers down south.
Northeast brook trout populations are what I would call “typical.” Fish move around some, but not a lot. Populations are not too genetically diverse, but there’s enough there for evolution to work with. No one stream contains a ton of fish, but we aren’t typically concerned that a population could be extirpated next year.
Drive a few hundred miles south and we’re telling a different story. There, brook trout populations are living life on the edge- literally and metaphorically. Southern-edge populations often live in streams that are hot, lack abundant food sources, and are threatened by barriers and an abundance of nonnative species. Compared with populations in the northeast, southern populations are also must older because they were never frozen out by the glaciers. But, with age comes genetic wear and tear- the older a population gets the more likely it is to have lost some genetic diversity due to random chance and catastrophic events that cause large population declines (floods, disease, etc.). Put all this together- the lack of connectivity, the low population sizes, and the limited genetic diversity- and southern-edge brook trout seem destined for population collapse.
When populations get isolated, and when genetic diversity starts to drop, biologists often start questioning whether we should intervene. It wouldn’t be hard- we can simply do what brook trout used to be able to do themselves and move individuals between populations. We call this physical movement of individuals among populations translocating. With a little luck and a lot of research and planning, the translocated fish will spawn with the resident fish, and their offspring will have increased genetic diversity that contributes to the population for many generations after. This is exactly what we want because, as genetic diversity increases, we often see an increase in number and size of individuals in a population. Perhaps more importantly, we also see that genetically diverse populations are better able to survive disturbance events. Translocations must be a no brainer, right?
But, here’s the catch. The populations down south have been isolated for so long that many of them have evolved their own identity, and potentially might be on their own evolutionary trajectory. Translocations are only successful if the fish moving into a population have genetics that are similar to the resident population. Otherwise, the offspring may be genetically more diverse, but those genes may make fish poorly adapted for life in that environment. This is tough, because it can take many generations to realize that the translocated fish are having a negative effect, and by then it could be impossible to turn back the clock.
The genetic risks associated with translocations have been known for a long time. But, the south brings up another, more philosophical, dilemma. If we start translocating fish across multiple watersheds, we potentially erase all of those genetically unique populations. Do we really want to do that?
Seriously. That is my question to you. What is more important? A genetically distinct population that could collapse within the next 50 years? Or, a population that loses some of its uniqueness, but perhaps has more long-term stability?
Here’s the fun part- scientist haven’t decided the right answer. On the one hand, you have to balance the risks of translocation with the potential for population collapse due to low genetic diversity and isolation. But, who’s to say that the isolated, distinct population wouldn’t survive just fine on their own? Populations above waterfalls have existed for hundreds of years and they are doing just fine. On the other hand, what is the value of these genetically distinct populations? Are they locally adapted to those streams, and therefor possess unique genes that are worth conserving? Or, are they just one in the same with the neighboring populations?
We’ve definitely got some important decisions to make, and the right answer will surely vary across watersheds. But, perhaps the decision doesn’t need to be so black and white, either. For example, we probably don’t want to move fish with strong hatchery influence to watersheds that are comprised of completely wild genetics. This is particularly true given that fish stocked in the southeast are often descendent of the northeast (turns out…southeastern brook trout are hard to reproduce in captivity). But, what if we look at the watershed as a whole, identify the populations that have completely wild genetics, and only translocate to/from wild-only populations that are somewhat similar? Maybe this is a good compromise that would lead to moderate increases in genetic diversity while still maintaining some unique genes in the population.
Now is the time to be having these discussions. I recently sat down with the Trout Unlimited Southeastern Volunteer Coordinator and we chatted about how some states have very restrictive translocation policies making it difficult to do reintroductions or translocations. Good for those states, because they are probably also limiting the spread of hatchery genes into wild populations. There are still so many uncertainties that I think conservative approaches are probably the best choice right now for most streams. After all, most brook trout populations will be fine for the next few years while we take the time to do our due diligence and research needed to make the right decision.
But, pretty soon we do need to start having these discussions. Will you be ready to contribute?
This post was inspired by recent research published by Kasey Pregler and colleagues. I would encourage all to read the original manuscript found here.
Time flies when you’re working on a dissertation. I turn around and it’s suddenly the end of the semester and I haven’t posted any updates in over a month. Oops.
Truth be told, there hasn’t been a lot to update on recently. Preparing for outreach events had me spending more time photo shopping in Microsoft Paint than running data analysis in R. But, Ben and I put together some impressive posters, so I consider that time well spent. And, we showed them off this past week at an Earth Day event at the local elementary school.
I’ve also taken the show on the road, and have been visiting several chapters of Trout Unlimited to present the hatchery-wild brook trout interbreeding (or introgression) results that I discussed in a previous blog. At first I felt a little guilty about making introgression the main topic of my talk because I thought most everyone in attendance had already read my blog post. But, as science turns, after we submitted the manuscript for publication- and after we thought for sure we had covered all of our basis with analyses- reviewers unanimously wanted to see another analysis. Of course.
Truth be told, we were actually happy to do the requested analysis. Simply put- they wanted us to quantify how introgression varies across different habitats. Is introgression higher in sites with higher temperatures? Lower pH? What about introgression rates in small vs. big streams? Does distance to a stocking location influence introgression? These are all great questions because they can help us understand if there are site-level characteristics that make a site more likely to have higher rates of introgression. And, if fish are more likely to introgression in certain habitats, then we can use that information to potentially adjust our stocking protocols.
Unfortunately, while there was no doubt this analysis was worthwhile, we knew before we started crunching numbers that none of the results would be statistically significant. If you remember, less than 6% of all fish we tested showed signs of being introgressed. And, most fish at any of our 30 sample sites assigned to pure wild origin. Without the presence of introgressed fish in our study, we weren’t going to be able to find strong relationships between habitat and introgression. It’s like trying to quantify elephant habitat in Pennsylvania- if the elephants aren’t here, we can’t really quantify their habitat.
But, because there are so few studies of introgression on stream trout, we knew that we could still use the results of the analysis to start building hypotheses about habitat variables that could matter. So, that’s what we did. We analyzed whether the probability of an individual being introgressed was related to eight “site-level” and three “watershed-level” habitat measures. The site-level variables includes measures of water quality and physical habitat that you would see if you were standing in the stream like pH, dissolved oxygen, and stream width (and we got lucky here, because all of that data was collected by Susquehanna University and the analysis would not have been possible without their willingness to share data). The watershed-level variables were those I measured back at the computer and were things like watershed area and landuse.
Like we expected, none of our models implicated any of our habitat variables as a smoking gun that increases introgression- all models showed a statistically insignificant relationship between introgression and the 11 habitat variables. But, a bit to our surprise, we did have a few variables that approached significance. Again, this was surprising because we had so little introgression in our data, and so even a variable that trends towards significance is worth a second look.
Probably not ironically, the variables that trended towards significant were also variables that have been suggested in other studies of lake-dwelling brook trout and stream brown trout as being modulators of introgression. It seems there is a consensus among studies that introgression rates are lower in larger streams, and at sites with low pH and higher adult brook trout densities. This makes some sense as higher wild trout densities increase competition (which likely decreases reproductive success of stocked fish) and larger streams have more stable flows that can help promote a large, healthy wild population. It’s suggested that pH might also be an important predictor of healthy wild trout populations, as macroinvertebrates densities are often higher in streams with higher pH.
So, at this point, these results can’t yet be used to really direct stocking efforts. They really just point to a need for more directed studies of introgression in stream trout. But, it does make you think. Most of the time we only stock streams without brook trout, or in streams with “marginal” wild populations. Those “marginal” populations, which tend to be smaller in size but existing in decent habitat, may be the populations that are most vulnerable to introgression. So, by trying to avoid wild-hatchery interactions and stocking in marginal populations, could we actually be increasing the probability of introgression? Again, we just need more data to tell.
I’m back! And, boy was my absence untimely. While I enjoyed soaking up the rays attending the annual meeting of the American Fisheries Society in Florida, I unfortunately missed the Pennsylvania Wild Trout Summit. The PA Fish and Boat Commission was quick to post presentations online, so I’ve been able to catch a few talks (including the one below by my advisor, Ty). But, I’ve also been reading some feedback from a few attendees and my takeaway is that the best talk wasn’t by a platform presenter- it was among members in the audience. One of the reasons I love studying trout is the passionate anglers and citizen scientists that are invested and devoted to wild trout conservation and restoration. There is no other angler base that is as informative and fun to interact with as you all, and I was sad to miss the opportunity.
My other observation is that there was some disappointment in what wasn’t discussed. Most notably, it seems a lot of people in attendance wanted to discuss the state’s trout stocking plans. I’m not surprised. Stocking is controversial and there will probably never be a stocking plan that makes everyone happy. But, I’m also encouraged. The public is trying to voice their opinions on this really complex problem, and, from what I’ve seen, seem to largely understand the delicate balance between the science of native fish conservation and the social dynamics of recreational fishing. It’s not an easy line to walk.
I’m also encouraged because it means there is interest in our current research beyond the scientific community. Our manuscript on native and hatchery fish interbreeding is nearing completion, and the results are getting closer to being released. Until then, I’ve been spending most of my days pouring over manuscripts published over the last 20+ years from other studies of hatchery-wild interbreeding and trying to summarize their findings. From this, I’ve already summarized the pros and cons to hatchery stocking, but I’ve left you in limbo the last two weeks. Overall, do hatcheries have more of a positive or negative effect on wild trout populations?
Before I answer that question, there are two caveats. First, I’m only discussing recreational stocking- or stocking done to temporarily increase population sizes to allow for increased angling opportunities. The potential pros and cons to conservation stocking are a bit different. Second, I am only focusing on the hard science. I’m not going to attempt to compare the social benefits of stocking with the impacts to native fish diversity. But, you should. Everyone should weigh the pros and cons and make their own informed decisions about stocking. It’s not my place to make the decision for you, but it is my job to present the science so that you can be informed. We know that stocking increases recreational opportunities and can be an economically profitable business, both of which valuable. Taking that into consideration, I have drawn a line in my mind where I think stocking is worthwhile and where it’s not. You need to find that line without someone telling you where they think you should put it.
So, after 20+ years of study, what do we know about the effect of hatchery stocking on wild trout populations?
So, where does that leave us? With a lot of uncertainty. Hatcheries can have negative effects on wild populations. But, not always. And, hatchery interbreeding can be high in stocked populations. But, not always. And, we know that there are long-term negative consequences of interbreeding. But, yet again, not always. We just don’t know.
Perhaps a more important question- where does that leave you in your thoughts on stocking?
I wrote last week of the two types of grad student vacations, conferences and field work. But, there’s another holiday that’s even rarer (at least for me) and merits even more celebration. I’m talking about your advisor’s vacation week, otherwise known as Grad Student Independence Week.
Truth be told, my advisor’s whereabouts don’t really influence my work ethic. For the time being, I’m working at my own self-defined pace (cross my fingers I can keep it that way). But, the closer we get to the beginning of the semester, the more sparse the office gets. With no one to pester during the day, why bother going in?
So, I didn’t. I slept in a little later (which for me is 6am), enjoyed coffee on my patio, and had one main goal: start working on the hatchery-wild hybridization manuscript. Data analysis is still on going, but at this point I know what the results are going to say. There’s no need to wait for the final numbers to crunch to start the long process of preparing the work for publication.
When I was an undergrad, I always thought that scientific publications were the works of brilliant scientists who wrote the equivalent of Shakespearian prose. I never thought I’d be smart enough to accomplish a similar feat. I actually still think that, except I’ve somehow been let into that elite crowd of published scientists seven times now. It still hard to believe I’ve reached the point in my career where I am the authority on a topic- someone out there is reading my manuscript and thinking I am the brilliant scientist. Crazy.
One thing I have learned along the way is that regardless of how smart you are, how great your research is, or how well you write, all manuscripts start in the same place. With a blank Word document that just stares at you. For me, it’s probably the single most intimidating and frustrating part of the publication process. Literally anything I put down “on paper” would represent an improvement over the blank page, but I just sit there for hours- staring, erasing, and getting more frustrated.
There’s all sorts of advice out there about how to be the best, most efficient writer- outline your ideas, write 30 minutes every day, discuss your paper beforehand, etc.- and I defy every single recommendation. That long, frustrating, fight with the blank page is just part of my process, and I need to work through before I can write something worth saving. And, the fight needs to be long and uninterrupted. Not a great task for tackling at the office where distractions are imminent, but a perfect job for celebrating my Grad Student Independence Week at home.
I actually only got one full day at home, but it was enough to win the battle and get a solid start on the manuscript. Time to save it, back it up, and not look at it for at least a few days. In the meantime, I go back to square one- read published manuscripts that I know are important for my study and that I will cite in my own publication to support why our study was needed and to add credibility to the results we found.
As I’ve said before, there aren’t a lot of studies on hatchery-wild interbreeding in brook trout. But, I did find one by Andrew Harbicht and colleagues (see below for a link to the manuscript) that looked at how the probability that hatchery trout will breed with wild trout changes depending on the environment. I’m still not releasing the result of our analysis, but studies like this are important regardless of what we find. Whether we find a high degree of interbreeding or not much at all, we need to know WHY we are getting that result. And, it makes sense that environmental conditions influence how much hatchery trout breed with their wild counterparts.
The study was conducted on several lakes in Algonquin Provincial Park in Ontario, Canada, of which some were never stocked with hatchery brook trout, and others had historic stocking that had been stopped 10+ years prior to their study. Immediately, you’ll notice there are some differences between their study and ours: we work on streams, and in areas that are currently being stocked with high densities of fish. Nevertheless, their results are important to keep in mind as we move forward. Most importantly, they found:
So, why is this study important for us? For starters, streams often support lower populations of brook trout than lakes, making us nervous that interbreeding may be more prevalent in streams than lakes- particularly, again, because stocking in our systems is frequent and on going. Our streams also have a wide range accessibility, pH, and other environmental variables (e.g., gradient and temperature) that influence population sizes and competition. Big picture, this study just shows us that introgression isn’t an all or nothing phenomena. Location matters a whole lot, and our results can’t be taken as the definitive response of trout to stocking.
But, all of this presumes that we are finding interbreeding. Which I’m not saying we are. I’m also not saying we aren’t. You’ll just have to stay tuned.
*Note: Content in this post is my own and may not reflect the opinion of the manuscripts' authors or the agencies they represent. I encourage you to read the manuscript, found here, so you can contribute to the discussion.
Last week I presented some preliminary genetics results using the image to the right, and used it as evidence that brook trout moving into Loyalsock Creek likely spawn outside of their home stream in later years. To put it another way, the fact that we see a mixing of genes at most locations (except for Mill Run), means that these populations were at least historically connected (and the telemetry data suggest at least some of them probably still are now).
Genetics data are such terrible guests for a blog. They come in unannounced, make a mess of the place, and don’t know when to leave. In this case, the mess they created was a lot of questions about what we expected this data to look like. While I’m excited that people want to know more, the answer just isn’t that easy.
For starters, my expectations of this dataset were zero. As you may recall, I hate genetics. My feelings towards the field are starting to evolve into a past-tense form of repulsion, but every day remains a struggle.
Those with a basic understanding of population genetics may have looked at that figure with doom and gloom and wondered why in the world those populations are so isolated. Separated by only a stone’s throw, we would expect all of those populations to be genetically very similar. This is particularly true when we think of land dwellers, which are generally more mobile than aquatic species. In fact, for many terrestrial species and larger-bodied fishes, we’d probably have to separate populations by many miles before we start seeing genetic isolation similar to the above diagram.
But, those that study headwater fishes may have found the results to be of no surprise. Most headwater fishes have very limited dispersal, owning to the fact that downstream habitats become increasingly wider and hotter with faster flows, making them less suitable environments for species that have evolved to live in tiny streams. There are also more predators downstream, so small bodies that are perfect for small streams quickly get eaten by larger fish like bass, pike, etc.
All this to say, interpreting genetics data is a not entirely straight forward. We can get numbers that tell us things about genetic diversity, population isolation, and anything else you might be interested in knowing. But, that’s not the full story, and those numbers are actually meaningless, and potentially dangerous, if used out of context. Our expectations for what the results should be really depend on the species, study location, historic stream use, stocking, etc., etc. It’s a bit of a detective game. So, for those interested in specific numbers describing diversity, FST, AR, NE, rxy, you’re out of luck. They’ll appear in a publication eventually, but only sparingly in this blog.
What I will do is compare, in broad terms, how our data stack up against other studies of brook trout. For starters, our genetics data cover far more than just the sites in that figure. We sampled 28 streams across the Loyalsock Creek watershed, making our study one of the largest scale studies of brook trout population genetics.
Usually changing the scale of a dataset means that you can use your results to answer new questions, questions that may be more appropriate for how we manage species. For example, genetics data are usually collected at the scale of a single stream or maybe a few streams in a small area. However, we don’t generally manage fish at this small of a scale. We generally make management decisions for entire watersheds. So, given that we now have genetics data for an entire watershed, it makes sense that we can now shed new light into the population genetics and management of brook trout at a watershed-level.
Wrong. But, don’t feel bad. I led you on.
The reason our dataset isn’t all that revolutionary, at least at the surface, is because brook trout populations are known to quickly isolate, even at fairly small spatial scales. Even side-by-side tributaries can be isolated from one another. So, if we see isolation at small scales, it’s not that surprising to see isolation at large scales. And, it’s no secret that that is what we’ve found in much of our dataset. Sites that are separated by 3+ miles are isolated from one another, and that’s pretty typical for brook trout.
That said, brook trout genetics are very diverse, particularly at sites separated by less than a mile. Sometimes they are genetically different, and other times they aren’t. To me, this is where things get interesting. If two sites are separated by the same distance and similar habitat, why do we sometimes see isolation and other times connectivity?
I don’t have an answer, but we’re hoping to explore this question with our dataset. And, we have some preliminary ideas. For example, we see that sites near a mainstem river system seem to have more connectivity than sites that are connected by a mid-reach run. What is it about mainstem rivers that makes them better for connecting fish populations? Or, to really wig your brain, what is it about fish living near a mainstem that makes them different (i.e., more mobile) than fish living higher in the headwaters?
That last question may seem a bit far-fetched, but brook trout are known for having a diverse range of life histories. Many of you may be familiar with “coasters,” brook trout living in Lake Superior that make long-distance migrations into the lake’s tributaries to spawn. There are also “salters” which spend a significant portion of their life in saltwater before returning to smaller freshwater tributaries to spawn.
Are fish moving between headwaters and mainstems (like we see in Loyalsock) a true life history variant? If so, how cool would that be? We’re a really long way from being able to say anything about behavioral variation in populations, but I will say that we aren’t the first study to document such dramatic differences in individual behavior. So, there’s support out there for the idea.
But, going back to the main question, how do our data compare to other brook trout population genetics studies? That’s an easy, albeit unsatisfactory, answer. Previous studies showed a lot of variation, and our study shows a lot of variation. So, Loyalsock Creek, as a whole, is not more or less isolated than we would expect given other studies. When we zoom in we see patterns were certain sites do seem oddly disconnected, and others more connected than we would have thought. And, seeing if we can explain that variation is going to be what makes our study so interesting.
I’m in Virginia this weekend where I’m currently watching snow accumulate and my little snow pig root around. The ruler is reading around six inches, which in central Virginia is enough to lock the grids and cause chaos. Guess I’ll have to lounge on the couch all day. Shucks.
Readers of my post from last week may recall mention of some genetics models my computer was slowly cranking out. Yep- those are still running. But it’s close to finishing and I think I can predict what the results are going to be. It’s potentially really interesting, but, I don’t want to report any findings, no matter how preliminary, until I’m confident they can hold at least a little weight. Plus, I’m meeting with some collaborators next week and I think they will recommend a few tweaks to the analysis that will help clarify some oddities and rule out other possibilities. But, soon. Promise.
In the meantime, I’ve started the early stages of preparing the publication that will formally report the results from this study. This process is always a little daunting. It usually starts with a blank Word document, a lot of distractions and procrastination, and many frustrations as I try to find the perfect words and the perfect topics to tell the perfect story for the data. I call this stage “unproductive circling,” and for me it can last for days. Then, I break down and just start typing. The resulting text is usually a horrible mess of incomplete thoughts, poor grammar, cuss words, and confusion. But, it stops my wheels from spinning so I can get traction and move forward. From there it’s all about refinement- reread, rewrite, reword. Over and over until the paper you were so excited about becomes utterly boring and dull. That’s when you know you’ve done your job correctly.
In case you're curious, this is an example output from the models that are running on my computer. I'll explain what these mean later, but for now, populations (numbered at the bottom) are more genetically distinct if they are shown as a solid block of color. So, for example population #3 is more genetically distinct than population #4, which COULD indicate that #3 is more isolated than #4.
For those unfamiliar with scientific publications, it starts with an introduction. Usually, it’s a couple pages in length explaining what the scientific community already knows and justifying why your research project was needed to plug an information gap. Many writers will disagree, but I really enjoy writing introductions. There are millions of research articles out there, and it’s your job to figure out why your lone study is still important to science. It’s a messy thought exercise that ends in a succinct story. It’s as close as scientific writing gets to poetry.
Much of this week was spent “circling.” I did all I could do to avoid writing, but I finally opened the blank Word document and stared. The problem I had was trying to decide what about my project was interesting. I know that sounds a little backwards. After all, shouldn’t you know your research is interesting before you do it? Yes, but that’s often not how research works. It’s only after it’s all done that you realize you got an unexpected result, or you collect data knowing it will be of interest to someone, but don’t know who that someone is until you’re nearly finished. So, I kept circling between two fundamental questions- is this study interesting because it’s brook trout, or is it interesting because it’s a genetics study of an aquatic organism?
While the answer is both, it turns out that the impact of the study is probably further reaching if I stretch my mind beyond brook trout. Sure, the study will be important for brook trout management. But, can it tell us more about genetics in general?
As I struggled with this question I do what I normally do- I started chatting with someone online. Luckily for me, my friend, Will, happened to be on. He studies genetics in many several wildlife species and knows far more about the topic that I ever hope to. He has been my go-to person during this process because he can give me insights into why something isn’t working or a next step. As we were talking he turns my attention back to a paper that has been sitting in my “to read” library for a long time.
To be honest, there is nothing earth shattering about the paper. It describes ways organisms (and their genes) can distribute throughout streams. But, the paper is incredibly necessary because streams are a special type of habitat. Many genetics studies are conducted on terrestrial organisms that can disperse in any direction. But, aquatic organisms are restricted to streams (for at least part of their life) and so have very different controls on how, when, and where they can move.
In some ways, you’d think streams make genetics models easier. Movement can only be up and down. But, that’s not entirely true for organisms like salamanders and macroinvertebrates that can fly or walk on land. Even for fish, which really are restricted to in-stream movements, models easily get confusing when you meet a tributary. Now movement isn’t just up and down, but the fish can pick any number of tributaries to occupy.
Taken together, suffice to say streams quickly complicate gene models. But, this paper was helpful because it packages just about every possible way stream-dwelling organisms move into four models. (And, don’t get scared away by the term “model.” It this case a model doesn’t involved any complex math or equations, it’s just a possible scenario for how the world works.)
The four models included:
Imagine from Hughes et al. (full link to the paper below). In these images, imagine that there is a stream flowing north, and another flowing south. The dots represent populations where colors represent genetic similarity. Under the death valley model (A), there is no exchange of individuals between populations. For the stream hierarchy (B), organisms move within a waterrshed, but do not cross watershed boundaries (so genetics are similar for populations in the north flowing stream, but different between the north and south flowing stream). In the headwater model, organisms look for the next closest headwater, which is often in an adjacent watershed (so genes are shared among populations that are close in overland distance, regardless of watershed). And, in a widespread gene flow model, distance doesn't prevent movement and all populations are genetically similar.
Conceptualized models, such as those described above, may not completely explain a single dataset, but it does help quickly put a dataset into perspective. We know that trout shouldn’t conform to the widespread gene flow or headwater models. Both of those require overland dispersal or very, very far stream distance dispersal.
Theoretically, trout should conform to the stream hierarchy model. We know that trout move, and they are more likely to move within their watershed than outside. And, previous studies have found evidence to support that trout genetics follow other requirements of the of a stream hierarchy model.
But….(and here is why my study is special), most trout genetics studies are done at fairly small spatial scales. Usually 10 sites that are near one another and that are part of a larger watershed network. We studied 28 populations located throughout the Loyalsock Creek watershed. Near as I can tell, there isn’t a published study on brook trout that sampled as many fish from such a large distribution. So, hopefully, we can tease apart which genetics model most explains brook trout.
Why does this matter? From a practical view, it can help in management. Management is most often applied at a watershed scale, but that’s not often the scale scientists do genetic studies. So, our gene data, which is arguably one of the most powerful tools we have, is not informative to our management efforts- at least not without a lot of assumptions. But, if we know that brook trout genes, at a watershed scale, tend to conform to one of the aforementioned models, then suddenly we can make more informative guesses (notice I say guesses) about how trout are moving in watersheds where we haven’t done a robust genetics analysis.
From an ecological perspective, understanding which model brook trout conform to could be just the kind of interesting quirk ecologists dream about. From the telemetry data, we know brook trout move among tributaries. But, if they still conform to a death valley model, it could mean that they return back to their home tributaries. Brook trout, after all, are closely related to salmon. Or, it could be a mix of models, in which case we ask why some populations have adopted different behaviors that has then lead to distinctly different models of gene flow.
I have a hunch as to what is going on. But, only time will truly tell…
If you’re interested in reading the paper I described above, it can be found here.
Last week I broke down some of the nuts and bolts behind genetics studies, explaining what a microsatellite is and why they are useful for genetics studies. If that seemed confusing or uninteresting, stick with me! That information, while important and useful, isn’t required to understand what I’m doing. What is important is an appreciation that genetics can tell us a lot about how we should manage our fish populations, and for the next several posts I’ll be delving into some of the reasons genetics studies are so informative.
Let’s start with a thought exercise. You are a fish manager tasked with creating a conservation plan for a threatened fish species that is found in only one river in the United States. Populations of this species are currently isolated from one another by dams, and habitat is significantly reduced from historic conditions. Generally speaking, biologist cringe when they hear this, particularly at the phase ‘population isolation.’ Higher connectivity among populations generally means higher genetic diversity, which usually results in healthier, more stable populations (we’ll dive into those details later…for now just trust me). So, your gut reaction to this problem might seem easy - remove the dams, restore the habitat.
But, wait! Before you say ‘down with the dams!’, you need to consider nonnative species. Restoring connectivity for native fishes will also increase connectivity among populations of nonnative species, thereby improving the health of their populations as well. It will also likely allow nonnative species to invade new habitats were they could compete with, and potentially extirpate, populations of the threatened species you are fighting to restore.
So what do you do? Do you try to improve connectivity knowing it could increase the abundance of nonnative fish, but potentially improve health of native fish? Or, do you leave things as-is knowing native fish population may be declining, but at least they are protected from nonnative species? Before you make a decision, you may decide you need a better picture of how healthy native fish populations are.
This is a scenario often faced in fisheries management, and one that genetics can help resolve. And, this is exactly what scientists out west did with populations of Oregon chub, a fairly small minnow species that is only found in the Willamette River in Oregon (to read the full paper, click here). Construction of flood control dams and introduction of nonnative species, particularly sunfish and catfish, severally reduced populations and restricted them to isolated areas within the river. Through significant effort, populations of Oregon chub have been showing signs of improvement, and the species was even downgraded from endangered to threatened on the Endangered Species Act. But, populations size isn’t always indicative of long-term stability (as we’ll discuss below), and appropriate next steps for management were uncertain.
Enter genetics. Biologists collected tissue samples (in the form of a small portion of the caudal fin, which is not harmful and grows back) and ran analyses to determine how genetically different each population was from one another and how genetically stable each population was.
The result? Though most populations of Oregon chub were isolated from one another and were genetically different, most still maintained high genetic diversity. And, most populations still had a lot of adults reproducing, which indicates genetic diversity will likely remain high into the future. So, though genetics are not a definitive measure of fish health, the study did indicate that most populations were likely healthy and stable into the near future.
So, as a manager, you let out a sigh of relief. You dodged a huge bullet. Restoring population connectivity, which seems like it’s the right thing to do, would have probably been one of the worst things you could have done. In addition to the cost, benefits to Oregon chub would have been negated by the introduction of nonnative fish which would have reduced population sizes. And, you wouldn’t have known any of that if it wasn’t for genetics suggesting current populations were healthy.
The study also highlighted another important fact. Numbers can be deceiving, and a population that has a lot of individuals isn’t necessarily healthy. One of the largest populations of Oregon chub had the lowest genetic diversity. Why does this matter? Well, one of the recovery strategies for Oregon chub involves creating new populations by taking individuals from one population and putting them in an area of river with good habitat that is currently unoccupied (a practice we call “translocating”). Normally, you take individuals from strong, stable populations, and without genetics you assume population size is a proxy for health. But, that’s not always the case, and translocating individuals from an unhealthy population with low genetic diversity could be setting them up to fail in their new environment.
Genetics also helps identify the correct scale of management. For example, should each stream be managed differently. What about river? Or entire watersheds? Maybe the state of Pennsylvania? You want to manage the largest area possible because it’s cost effective and easier. But, threats affect fish differently at different locations. And, these affects are usually species-specific. So, there’s no easy answer as to what the scale of fisheries management should be, but genetics can help guide the choice.
But, I’ll save that topic for next time as it directly relates to trout management and some of the results we are finding in our studies.
After field work ended the question was ‘now what?’ I have a lot of datasets for several different projects, and I needed to figure out which to tackle first. But, despite exciting observations at the end of the telemetry season, I can’t start there. It’s a really messy dataset that’s going to require a lot of tinkering, input from collaborators, and there are some plans for more data collecting in the spring.
I ultimately decided to dust off a dataset from last year that was collected to determine the genetic structure of brook trout populations across the Loyalsock Creek watershed. But, I hesitated to start here. It’s no secret that I struggle with even the most fundamental concepts in genetics. I have a hard time understanding things I can’t visualize, and I managed to skirt my way around taking genetics classes as an undergraduate. But, the last few years I have grown to appreciate the questions you can answer about trout conservation through genetic studies. So, through a lot of hand holding from my friends who study genetics and our collaborators at the U.S. Fish and Wildlife Service, I am very slowly getting there.
One of the downfalls with genetics is that the topic can very quickly become technical and stray beyond the average person’s interest and understanding. But, that’s where my ignorance actually comes in handy. I don’t know enough to make the conversation technical! For the next few months I’ll be working on analyzing genetics data, and along the way I’ll try to breakdown all the concepts and terminology in a way that is both understandable and (hopefully) interesting.
For starters, let me go back. I am studying the population genetic structure of brook trout across the Loyalsock Creek watershed. What does that mean? Basically, I’m trying to see how genetically similar brook trout are from different locations around the watershed. We would expect that populations that are close to one another would be more similar than populations that are further away. This is because it’s more common for fish to move to, and then reproduce in, a neighboring stream than to make a long-distance movement to a stream many miles away. This is particularly true for species like brook trout which, compared to other species, don’t move very far. Trout also live in cold, headwater streams and warmer mainstem rivers act as barriers to movement, thereby further limiting exchange of individuals among populations.
There’s many ways to measure genetic diversity, but we are doing it by looking at differences in sections of DNA called microsatellites. To explain this concept a little further (mostly to myself...the visuals help), I often show the diagram below. Basically, every tissue in an organism’s body is made up of cells. Floating around inside the nucleus of those cells are chromosomes, and every chromosome contains thousands of genes. A gene is made of DNA, and DNA codes for proteins that ultimately produce all features of an organism. Put another way, genes are like instruction manuals, and DNA the step-by-step instructions for how to assemble, in this case, a trout.
The DNA inside genes is made up of base pairs (some of you may remember that there are four base pairs, adenine (A), cytosine (C), guanine (G), thymine (T)). While most of these base pairs code for specific proteins (for example, the base pair string of UCA codes for the protein ‘serine’), there are some sections of DNA that are “silent” and serve almost no biological function. One such case are sections of DNA known as microsatellites, which are sequences of 2-5 base pairs that are highly repetitive and do not code for a protein.
Microsatellites are powerful in population genetics studies for many reasons. First, because they do not code for a specific protein or trait, they are largely conserved in populations. This is important because if we were analyzing coding regions of DNA we wouldn’t know if absence of the DNA region was because of genetic isolation or because the environment was selecting against the trait that was being produce and therefor deleting it from the population.
The other reason microsatellites are useful is because they are prone to mutating. Again, because microsatellites do not have a functional purpose, these mutations are not harmful. But, these mutations work to give each population its own unique “signature” which we can track around the watershed as fish move around.
The location of a microsatellite on a gene is referred to as a ‘locus,’ and we analyze 12 different microsatellite loci to make inferences about population genetic diversity. Basically, we look across all loci to see how different individuals are within a population, and compare that to how different individuals are across all populations. Thankfully there are software programs to do this.
If you’re curious, below is what this data looks like in real life. Every row on that spreadsheet represents an individual, and the columns represent the ‘genotype,’ or genetic composition, for each microsatellite. In this case, the genotype represents the number of times the base pair sequences repeats. So, for example, the microsatellite loci B52, which is a base pair sequence of GCGT, is repeated 207 times in the first individual on that spreadsheet. You'll also see that there are two numbers for each loci because trout are diploid, meaning they get one copy of the gene from their mother and one copy from their father (just like humans). And, if you quickly glance down the spreadsheet, you see a lot of similarities because all of those fish are from the same population.
I’ve spent the last few days trying to run some summary statistics to describe the genetic diversity of all 28 samples sites and determine how similar each site is to one another. I’ll report some of those results in the coming weeks but, for now, don’t worry if all of that seemed confusing. Just remember- we study sections of DNA called microsatellites, and the more similar the microsatellites are the more genetically similar two fish, or two populations, are.
That’s not so bad, is it?
I’m not even going to try to hide it. Very little got accomplished this week in the field. We had a lot of struggles early with equipment malfunctions, and no sooner than we started making progress it started raining. That’s field work, and we’ll try again next week.
When I started this blog I intended it to be both a source of updates for my studies, but also a place where interesting research by other trout ecologists could be made more accessible. While I tackle a few big-picture questions in the Previous Research tab, hundreds of research articles are published every day. Unfortunately, these articles are often hard to access without a university affiliation, and sometimes even harder to understand without a dictionary and a lot of patience. The knowledge used to manage our natural resources shouldn’t be held captive in the hands of a select few. Science has to do better.
The unexpected lull in field work left me a little time to catch-up on the growing stack of articles I’ve saved over the past few months. So, there’s no better time to start realizing the next phase of this website and discuss some interesting research by my fellow ecologists. In this inaugural research blog I discuss a paper that addresses the question:
How High Can a Trout Jump?
First off, why do we care? In addition to the increasingly large number of man-made barriers to trout movement (e.g., bridges), mountain streams have many waterfalls of various shapes and sizes. Knowing whether a fish can swim upstream of a barrier has significant implications for management. If a barrier is passable, then the population is said to be “connected,” which is important because connected populations are more resistant to extirpation (a concept I explain more detail here). However, if a barrier is not passable, a population will become separated into two subpopulations, each with a higher risk of collapsing and being extirpated when there is a disturbance.
One way to determine whether a population is connected is to measure genetic relatedness within and among populations. Now, bear with me. I know a lot of people hear “genetics” and tune out thinking the words to follow are going to be impossibly complicated and uninteresting (I once, and sometimes still do, fall into that category). But, the concept is quite easy, and it’s fascinating that we can do this study in fish. When we check for genetic relatedness, we are interested to see how similar the genetic composition is of fish within and among populations. If the genetics are similar, then we know the population is connected.
Put another way, checking for genetic connectivity is basically the equivalent of determining how genetically related everyone is at a family reunion (within-population) and then comparing that family to families across the United States (among-population). You would expect that within-population genetic relatedness would be high because you’re measuring parents, siblings, and cousins that share similar genes. You would also expect that families that live near one another would be genetically different, but still somewhat similar because individuals can easily cross over (marry) into other families. However, we wouldn’t expect a family from Pennsylvania to be too closely related to one from California because not too many people move that far away.
The question is, how many states away do we start seeing families becoming distinctly different? To bring this back to trout, how tall does a barrier have to be before trout can no longer swim past it and there are two genetically distinct populations?
This is the question that Anne Timm and her colleagues addressed in a recent paper published a few months ago in Environmental Biology of Fishes. They measured genetic relatedness of populations upstream and downstream of waterfalls ranging in height from 5-200 feet. They found that 13-foot barriers were large enough to significantly reduce genetic relatedness between two populations, but population separated by smaller barriers still had some degree of connection. This means that brook trout are at least sometimes capable of moving upstream of barriers that are less than 13-feet tall. That’s a really tall jump!
That statement is qualified by “sometimes” because movement is highly dependent on stream flow, availability of pool habitat near the falls for resting, and slope of the fall. So, in reality, 13-feet is an extreme maximum, and it’s unlikely that trout are regularly swimming upstream of barriers that tall. In fact, other biologists have found maximum jumping height to be less than 5 feet for brook trout, which is probably a more realistic height for trout to regularly jump.
More connected populations also have more genetic diversity, meaning that there are more genes in the entire gene pool. Thinking back to humans, a more connected population might have genes for all the possible colors of hair, but disconnected populations might lose the genes for red hair. As predicted, Timm and others also found that genetic diversity of upstream populations decreases when falls exceed 13 feet tall. This is problematic because more genetically diverse populations are usually more likely to survive environmental change, so this suggests that populations upstream of large falls may be more prone to extirpation.
What can we do with this information? Of course we aren’t going to remove waterfalls or artificially connect population near falls. But, it does give a quick threshold for identifying impassable barriers, which could help locate disconnected populations that may be at a higher risk of extirpation. Once identified, populations can receive special consideration for receiving additional habitat protection or possibly stocking to supplement genetic diversity. But, stocking has its own problems, a topic we’ll have to save for another time.