The consequences of local adaptation in stream ecosystems: ecological roles and species distributions
Ecological and evolutionary processes are often assumed to operate at different temporal scales. Most ecological research assumes that species characteristics are static over scales of ecological interest (e.g. days, years, or even decades). However, a considerable amount of research now suggests that ecological and evolutionary processes operate on similar scales and often interact with implications for the role organisms play in an ecosystem. In this presentation, Dr. Thomas will highlight research results that illustrate how local adaptation (a) influences the phenotypic traits and ecological roles of fish in Trinidadian streams and (b) leads to higher species turnover of aquatic insects along tropical elevation gradients relative to temperate latitudes.
The selective pressures on all species change as species are gained and loss and both invading and resident species can face intense selective pressures following community changes. As time since invasion progresses and phenotypes adjust to new selective regimes, trait-mediated interactions between organisms and their environment also change. We have been examining this process using Trinidadian guppies (Poecilia reticulata) and their native ecosystems on the island of Trinidad. This work builds upon extensive prior research demonstrating rapid phenotypic change in guppies in response to the presence or absence of predators. In this presentation, we summarize the results of a large collaborative effort to examine ecological consequences of these changes using a suite of approaches that vary in their complexity.
The second portion of this presentation will examine whether aquatic insect taxa have great species turnover along elevation gradients in Ecuador versus Colorado and whether differences are predicted by the thermal biology of resident species. In essence, this works tests Janzen's classic hypothesis regarding "Why mountain passes are higher in the tropics" and whether it holds for aquatic insect populations. Briefly, our results indicate that tropical species have reduced thermal tolerance, narrower altitudinal ranges and greater species turnover with altitude than their temperate counterparts.
icon search Searchable Transcript
Toggle between list and paragraph view.
[00:00:01.424]Well thanks for coming everybody
[00:00:02.387]I know these local seminars aren't nearly as exciting
[00:00:05.261]as someone coming from somewhere else.
[00:00:08.040]So, I appreciate you showin' up.
[00:00:10.350]I'll also sort of qualify this a little bit
[00:00:12.295]in that I was putting this all together
[00:00:13.768]and I realized that
[00:00:14.817]I was hoping to talk both about our work in Trinidad
[00:00:17.336]as well as what we've been doing in Colorado and Ecuador,
[00:00:20.433]but to do that well was really hard to do in an hour.
[00:00:23.252]So, I'm gonna spend a lot more time
[00:00:24.487]kinda flushing out what we've done in Trinidad
[00:00:26.916]it's a little more complete story at the moment.
[00:00:29.731]Then I'll go back to spring
[00:00:30.932]and do some more interesting things so I thought I'd
[00:00:32.974]bring everybody you you to speed,
[00:00:34.300]and that's gonna be a little bit of review
[00:00:36.683]for some of the folks that were out
[00:00:38.083]at the nature center a few years back
[00:00:40.288]when we had the Eco-Evo sort of workshop out there,
[00:00:44.153]but there's some new stuff, especially when it comes
[00:00:46.207]to the whole system modeling, so I thought that might be
[00:00:48.196]worth kind of walking through and then to the extent
[00:00:50.542]that I have some time, I'll introduce and talk a bit about
[00:00:53.959]what we've been working,
[00:00:55.671]a project I've been involved in in both Colorado
[00:00:58.664]and Ecuador, looking at species distribution in the role
[00:01:01.684]of temperature and physiological thermal properties
[00:01:07.646]in dictating species turnover across altitude,
[00:01:10.840]and some of the implications for that.
[00:01:12.020]So hopefully, I can get a little time
[00:01:13.255]to talk about that at the end.
[00:01:16.870]So it might be a little bit better to just call this
[00:01:19.008]what I've been up to when I'm not around.
[00:01:21.495]So in the first several years I was here,
[00:01:24.170]I wasn't here a lot.
[00:01:25.391]Some of you guys might remember.
[00:01:26.475]I been around a lot more since I been teaching so much
[00:01:28.751]in the last several years,
[00:01:30.352]but those first few years I was gone,
[00:01:31.657]and where I was was mostly in Trinidad.
[00:01:34.699]And so I'm gonna spend some time today kind of
[00:01:36.799]walking through that project,
[00:01:38.332]hit some of the result highlights,
[00:01:40.877]and that's a continuing effort, so I kind of want
[00:01:43.706]to get you oriented to what we're up to now
[00:01:45.725]and we're kind of just now, I think,
[00:01:48.602]getting to the really interesting part of that project.
[00:01:52.769]We'll see if we can't keep the money going.
[00:01:57.010]So on the year I was born, it turns out, Evelyn Hutchinson,
[00:02:01.257]who's a very well known,
[00:02:03.027]very famous limnologist at the time,
[00:02:05.250]very well known ecologist,
[00:02:08.128]sort of summarized I think the popular opinion at the time
[00:02:12.154]that ecology really represents the template
[00:02:14.062]upon which evolution takes place.
[00:02:17.569]Evolution is responding to the ecological conditions
[00:02:19.875]of the here and now.
[00:02:24.495]Around that same time, there was the beginning of
[00:02:28.282]discussions about whether or not that wasn't
[00:02:30.797]a one way interaction, whether or not there was
[00:02:33.543]feedbacks between both ecological
[00:02:35.643]and evolutionary processes.
[00:02:38.263]And Pimentel sort of summarized this
[00:02:39.767]in some of this early work when he,
[00:02:41.355]in his paper that I've shown here,
[00:02:44.328]when he said that density influences selection,
[00:02:46.541]selection influences genetic makeup,
[00:02:48.702]which in turn then influences density.
[00:02:51.778]And that kind of is a quick little summary
[00:02:53.236]of how we might think about the feedbacks between evolution
[00:02:56.673]and ecological processes.
[00:02:58.879]In much of that earlier viewpoint,
[00:03:01.384]and what we've now I think started
[00:03:03.104]to learn sort of changed that
[00:03:05.059]and maybe accept that second viewpoint
[00:03:07.025]is that evolution and ecological processes
[00:03:11.982]can occur on the same timescales.
[00:03:13.637]I think historically, when we think
[00:03:15.604]about that first viewpoint,
[00:03:17.517]we really thought about those things
[00:03:18.523]as operating in two different time scales,
[00:03:20.554]and thus the interaction would be weak.
[00:03:22.638]With ecological processes have been sort of,
[00:03:25.588]contemporary time, course of a (quickly mumbles)
[00:03:28.756]a couple of years,
[00:03:29.622]but evolutionary processes taking place
[00:03:31.210]over longer timescales and really not interacting
[00:03:34.006]at those shorter timescales,
[00:03:35.497]but there's a lot of evidence now
[00:03:37.211]and a variety of organisms that
[00:03:39.081]at least rapid local adaptation
[00:03:42.597]and changing the genetic population structures
[00:03:45.679]can happen really fast.
[00:03:48.651]So what I'm gonna do, I got involved with a project
[00:03:51.752]that was using one of those model organisms for study
[00:03:55.045]in this eco-evo interaction.
[00:03:57.870]And that's the Trinidadian guppy.
[00:03:59.986]And there's been work on Trinidadian guppies
[00:04:01.837]since the 60s, actually 50s,
[00:04:05.480]and then sort of picked up by Dave Reznick and others
[00:04:07.833]and so I got involved in a project
[00:04:09.538]that was lead by David Reznick at UC Riverside,
[00:04:12.468]it had a large team, and I show the membership
[00:04:14.704]a little bit later, there was investigators from Cornell,
[00:04:19.567]from Colorado State, from MC State,
[00:04:23.531]Georgia, UC Riverside, all of those.
[00:04:27.942]What we were trying to do is sorta link
[00:04:32.036]the evolutionary and life history characteristics
[00:04:35.197]of this fish with their ecological role they play in nature,
[00:04:39.605]and to see how those shift,
[00:04:41.279]and whether or not we can measure those
[00:04:42.922]over the course of our careers,
[00:04:45.462](chuckles) as it turns out.
[00:04:49.033]So just to get everybody a little bit oriented,
[00:04:50.591]just so you know where Trinidad is,
[00:04:52.194]this isn't all that important but for one reason.
[00:04:54.948]So, it's a Caribbean island
[00:04:57.102]and it is definitely Caribbean by culture,
[00:04:59.130]but it is not Caribbean by life history,
[00:05:03.349]or by natural history.
[00:05:04.973]It periodically connects to the mainland
[00:05:07.329]of South America and as such has
[00:05:09.011]a very South American flora and fauna.
[00:05:11.988]Downside of that is it has a lot of snakes.
[00:05:14.166]And a lot of poisonous snakes, right?
[00:05:16.192]So, it's located right off the coast of Venezuela
[00:05:18.427]and the mouth of the Orinoco River,
[00:05:21.840]as shown here, and our research focused
[00:05:24.142]on this northern mountain range.
[00:05:26.471]So, where's my mouse?
[00:05:27.582]Shown up here, you can't,
[00:05:28.807]I don't think you can see this little black circle.
[00:05:30.596]Most of our work is up in this range,
[00:05:32.237]we have a variety of survey streams
[00:05:33.771]throughout that whole range,
[00:05:35.065]as well as some focal sites that are located
[00:05:41.834]on the upper La Laja River.
[00:05:43.266]So, we've got upper La Laja, lower La Laja,
[00:05:45.129]the Caigual branch, the Taylor branch,
[00:05:47.039]which is sort of four focal streams
[00:05:48.606]that I'll talk more about as we get into this a little bit.
[00:05:51.157]That's where a lot of our whole-stream manipulations are.
[00:05:57.808]It's a little slow moving here.
[00:06:00.187]There we go.
[00:06:01.231]So, here's a picture of two different phenotypic conditions
[00:06:06.461]for this particular fish species,
[00:06:08.373]these are the same species of fish
[00:06:11.469]with individual shown, these are males,
[00:06:14.540]with individuals on the left all from one location,
[00:06:17.091]individuals on the right all from another.
[00:06:19.522]One's from a high predation environment,
[00:06:21.002]one's from a low predation environment.
[00:06:23.116]I'll let you take a second to make a guess
[00:06:25.193]as to which one you think those might be.
[00:06:29.114]So, hopefully most of you guys got this right.
[00:06:31.496]Consistent with life history theory,
[00:06:34.223]when we have low-predation risk,
[00:06:36.560]you tend to get larger body sizes, or colorful males,
[00:06:40.886]these have a whole variety of characteristics,
[00:06:42.694]they tend to have fewer young,
[00:06:44.181]there's a whole bunch of things
[00:06:45.014]that change between these two phenotypes.
[00:06:47.845]And these are located in fairly close proximity
[00:06:50.620]to one another, they're in the same basin.
[00:06:53.133]These particular populations,
[00:06:54.153]they're only bout five kilometers apart in the same river.
[00:06:57.628]But they're separated and isolated
[00:06:59.716]by waterfalls that cause predators to fall off.
[00:07:03.808]So, as you move from the bottom
[00:07:04.928]of these mountain ranges towards the top,
[00:07:07.338]you sorta serially lose predators as you go up.
[00:07:11.983]And eventually, about halfway up that mountain range,
[00:07:14.043]maybe not even quite,
[00:07:14.898]you lose most of the large predators
[00:07:17.008]and you're really left over with rivulus,
[00:07:19.315]this fish down here,
[00:07:21.081]which is mostly a competitor with guppies
[00:07:22.585]and just guppies, those are the two remaining species.
[00:07:25.540]rivulus goes all the way to the top.
[00:07:28.110]It's actually quite an interesting fish in its own right,
[00:07:29.934]there's been a large part of this project
[00:07:31.307]sorta focused on that,
[00:07:32.659]I won't get a chance to talk much about it,
[00:07:33.998]but this guy goes terrestrial.
[00:07:36.693]It's a really interesting scenario.
[00:07:40.478]So, here are a number of different traits
[00:07:42.051]that vary between these two phenotypes.
[00:07:44.711]As you can see, low predation has a later age
[00:07:47.605]of first reproduction, a slower growth rate,
[00:07:51.528]bigger maximum body size, et cetera.
[00:07:53.313]A whole suite of things
[00:07:54.146]that you'd predict based on predation risk.
[00:07:56.789]It's sorta stemmed right from much of that theory.
[00:08:00.082]Now, several of these things have the potential
[00:08:03.288]to change the ecological role these species play
[00:08:06.043]in their environment, or in their ecosystem.
[00:08:08.704]And to date, prior to the study,
[00:08:10.536]no one had looked at that really at all.
[00:08:12.968]What was the ecological consequences
[00:08:14.682]of these phenotypic variations.
[00:08:17.552]So, that was really one of the big,
[00:08:19.338]that's sorta the main focus of much of what we do.
[00:08:25.001]And we used a variety of different approaches
[00:08:27.665]to get at this issue.
[00:08:29.389]So, here's a very simple sorta diagram
[00:08:31.588]that shows some of the trade-offs that we make
[00:08:34.816]when we try to design experiments for precision
[00:08:38.588]versus realism and generality.
[00:08:42.041]So, if we tend to think of these things related
[00:08:44.136]as we move from precision to realism,
[00:08:46.338]we can take a variety of different experimental approaches
[00:08:48.731]to sort of span that gradient.
[00:08:51.455]And for us, that ranged
[00:08:52.876]from very highly-controlled lab experiments conducted
[00:08:56.515]back in the states, using common gardens,
[00:08:59.781]like the common gardens sorta setup I have shown here.
[00:09:04.070]On this side, this is back at Cornell,
[00:09:05.504]but we have a similar facility at Colorado State as well.
[00:09:08.589]We had a series of stream-side mesocosms
[00:09:11.353]that we built in Trinidad, shown here,
[00:09:14.129]and I'll show some of these results.
[00:09:15.375]This is what I largely focused on
[00:09:17.275]in the talk back at the Nature Center.
[00:09:19.227]I'll show some of the same results, a little bit modified.
[00:09:22.094]But what I really wanna focus today on
[00:09:23.538]are some of the data and the results that we got
[00:09:25.569]from working in natural streams,
[00:09:28.332]whether it be just doing pool manipulations
[00:09:31.102]during the dry season, sorta short term,
[00:09:34.537]small spatial extent experiments
[00:09:37.017]where we just manipulated fish communities
[00:09:38.881]within individual pools.
[00:09:41.663]So, to experiment on which we manipulated conditions
[00:09:44.082]at the scale of the whole stream reach,
[00:09:46.152]and conducted experiments
[00:09:47.532]and looked at standing stock changes at that reach scale.
[00:09:52.545]So, I'm not gonna talk much about the lab work.
[00:09:56.351]There's some really interesting papers
[00:09:57.687]that have come out of that, it's interesting work,
[00:09:59.718]I'm just not gonna have time today.
[00:10:01.371]But I'll focus instead on these second three.
[00:10:07.454]So, the stream-side channel mesocosms,
[00:10:11.429]I have those sorta shown here.
[00:10:12.756]I'm sorry, I don't have a little bit better picture.
[00:10:16.192]But these are constructed right next to a stream
[00:10:18.367]that's running behind these.
[00:10:19.970]It's a small little spring-fed system,
[00:10:21.632]it was the second spring that we divert water
[00:10:23.576]straight into our head tank,
[00:10:25.053]so streaming water from that head tank
[00:10:26.757]into each of these channels
[00:10:28.252]so we can get a pretty different flow.
[00:10:29.774]So, these are sorta nice
[00:10:30.697]'cause we have standardized flow conditions,
[00:10:32.809]we put in a standardized set of substrates,
[00:10:35.109]and we can manipulate fish composition,
[00:10:37.857]fish density, and a variety of other things.
[00:10:40.803]Light, all kinds of things.
[00:10:42.868]It's really a nice sort of stream-side experimental setup.
[00:10:46.572]And a lot of the work that I'm gonna show
[00:10:48.526]and talk about in the next few slides is conducted
[00:10:50.293]by a whole team of people,
[00:10:52.109]but really lead by Ron Bassar,
[00:10:53.706]who is now a new faculty member back in Massachusetts,
[00:10:57.954]and Rana El-Sabaawi, who was a post-ac
[00:11:00.425]with myself and Alex Flecker at Cornell,
[00:11:02.759]and is now a faculty member of the University of Victoria.
[00:11:08.952]So just to jump into some of those results.
[00:11:10.990]So, just a quick backup, the way these experiments work
[00:11:13.643]is that we set the system, we let it equilibrate
[00:11:16.906]for a couple of weeks to come in equilibrium
[00:11:19.279]with the particles that are flowing through it,
[00:11:21.439]build up a benthic standing stock of algae.
[00:11:25.333]We introduced aquatic insects
[00:11:27.285]but they also sorta colonized,
[00:11:28.622]so we just give it a few weeks to kinda come into stasis,
[00:11:32.473]and then introduced fish, or introduced
[00:11:34.953]whatever the experimental manipulation might be.
[00:11:37.317]So in this case, we have fish-less controls,
[00:11:39.246]that's the None over here,
[00:11:40.868]we have a low-density treatment
[00:11:44.497]of both high-predation guppies,
[00:11:46.568]the high-predation phenotype,
[00:11:47.706]and the low-predation phenotype.
[00:11:49.809]And then we also have a high-density treatment.
[00:11:52.137]So, we have both phenotype,
[00:11:54.632]we have fish effects, phenotype effects,
[00:11:56.297]and density effects.
[00:11:58.447]And you can see here in the Anova Table,
[00:11:59.968]that we have significant differences
[00:12:02.295]between where you have fish
[00:12:03.991]and where you don't have fish.
[00:12:04.991]You see this is a log scale here on algal standing stocks,
[00:12:07.784]so you see that's a considerable drop.
[00:12:09.886]I'll show you that visually here in a second.
[00:12:11.963]But we also saw a significant phenotypic effect here
[00:12:16.648]with significantly higher algal standing stocks
[00:12:20.403]where we had high-predation guppies
[00:12:22.294]and more suppression where we saw the low-predation guppies.
[00:12:26.117]And then we also saw an affected density,
[00:12:28.172]that effect was exaggerated at the higher density.
[00:12:30.721]But you need not necessarily see that statistical table
[00:12:33.492]to see some of these effects, you'll see the fish.
[00:12:36.349]So this is, hopefully you can see this better
[00:12:37.706]than I can see it on my screen here.
[00:12:40.065]But with guppies shown here on the left,
[00:12:43.918]no guppies here on the right.
[00:12:45.193]Hopefully you can see that
[00:12:46.026]this has got a pretty luxuriant green,
[00:12:47.754]sort of algal coating on the bottom,
[00:12:49.273]it's a pretty dramatic effect.
[00:12:50.714]Talking about orders of magnitude difference
[00:12:54.160]in the level of standing stock in the streams.
[00:12:57.965]So, it's not a small effect.
[00:13:01.153]If we look at the invertebrates we found in those systems,
[00:13:03.182]we see a somewhat reversed situation.
[00:13:07.474]So, there's a fish effect again.
[00:13:08.658]When you put these fish in,
[00:13:09.874]you see lower amounts of invertebrates.
[00:13:12.044]But rather than having the low-predation guppies
[00:13:17.589]suppress them more so than the high-predation, it's fluent.
[00:13:20.971]So, whereas the algal standing stocks were suppressed most
[00:13:24.766]by the low-predation guppies,
[00:13:26.598]now we see the opposite, now we see more a greater magnitude
[00:13:30.791]of suppression of invertebrates
[00:13:32.197]by the high-predation guppies.
[00:13:34.124]So, that's the suggestion, that high-predation guppies
[00:13:36.293]are more focused on invertebrates,
[00:13:37.501]low-predation guppies, more focused on algae,
[00:13:40.428]and it also turns out
[00:13:41.261](softly talks off mic) more work on this.
[00:13:45.307]And if you look at the diets,
[00:13:46.798]you go into the guts and you dissect these fish
[00:13:49.235]and look at the guts at the end of the experiment,
[00:13:50.964]you see just what you would expect.
[00:13:52.428]You see a considerable amount,
[00:13:53.906]more invertebrates shown up here,
[00:13:55.094]much more in the guts of the high-predation fish
[00:13:58.198]than the low-predation fish,
[00:13:59.564]and we see just the opposite
[00:14:01.802]when we look at the diet times that we find in the guts,
[00:14:03.900]lots more in the low-predation,
[00:14:05.894]not that much in the high-predation.
[00:14:07.702]What's really interesting is the magnitude
[00:14:09.212]of this difference is identical for both.
[00:14:11.676]It's a threefold, just about a threefold change
[00:14:13.966]when we look at the logged invertebrate area for these
[00:14:18.885]and it's about the same down here,
[00:14:20.790]but in the opposite direction, which is kinda interesting.
[00:14:24.775]So, in this particular case,
[00:14:25.637]we saw a strong phenotype effect.
[00:14:27.406]We also did this with fish with different drainages,
[00:14:30.224]and that drainage actually also has a significant effect.
[00:14:34.851]So the take-home message from this part of it is that,
[00:14:37.645]and this is kind of important going forward
[00:14:39.174]to kinda think about what we might expect to see
[00:14:41.388]in the natural stream,
[00:14:42.904]is that HP guppies tend to eat more invertebrates,
[00:14:45.544]they tend to be invertebrate specialists,
[00:14:48.271]and LP guppies tend to eat more basal resources, algae,
[00:14:53.596]we also see them eating a lot
[00:14:54.838]of fine particular organic matter,
[00:14:57.352]so they also serve as detritivore as well.
[00:15:03.522]And one of the things that's nice
[00:15:04.753]about these mesocosms and in these particular studies
[00:15:07.063]is that you can parse the relative effects,
[00:15:10.656]how important is the phenotype effect relative
[00:15:13.484]to the light effect?
[00:15:14.394]We can do this for density in those previous experiments
[00:15:17.488]as we have have it for a variety of different variables,
[00:15:21.367]but this was the results from a paper
[00:15:23.468]that Rana had written, published back in 2015,
[00:15:27.811]that looked at the effect of light.
[00:15:29.241]So, we have high-light systems,
[00:15:30.752]and low-light systems, with fish, without,
[00:15:32.952]and then with the different phenotypes in both conditions.
[00:15:37.076]And in this case, what she's looking at is
[00:15:39.726]these are all variables of which we had significant effects
[00:15:44.233]of either phenotype or light.
[00:15:47.157]And if you look at that effect and parse it
[00:15:49.778]by what was driven by light
[00:15:51.310]versus what was driven by phenotype,
[00:15:53.961]that shows up here as the light bars,
[00:15:55.787]as the light contribution to that effect,
[00:15:58.172]the dark part is the phenotypic effect.
[00:16:00.873]And so, that varies quite a bit,
[00:16:03.571]depending on what you're looking at.
[00:16:05.003]So, if we look at some of these guppy traits, of course,
[00:16:07.881]the light is dominant in many of these things,
[00:16:10.332]offspring size is the phenotype effect.
[00:16:12.373]So we move over through this,
[00:16:13.955]we see that in the structural elements,
[00:16:16.222]that is the standing stocks of stuff
[00:16:18.014]at the end of the experiment in these mesocosms,
[00:16:21.518]that was really driven mostly by light.
[00:16:25.644]We had substantial effects of phenotype,
[00:16:28.496]but not nearly to the same extent
[00:16:30.267]that we saw with the light treatment.
[00:16:32.815]You look at the ecosystem processes,
[00:16:33.939]you start to see that shift a little bit.
[00:16:35.803]And this gives us a little insight
[00:16:36.996]to what we might see potentially in natural channels.
[00:16:40.482]We may not see big standing stock effects,
[00:16:42.126]but we may see process beings effected.
[00:16:45.133]So that's a little, a little (softly talks off mic)
[00:16:47.031]So, kind of a nice way to sorta look at these,
[00:16:49.291]but I think again here the take-home message,
[00:16:50.830]for me, out of this, is that these phenotypic variations,
[00:16:54.532]this effect of local adaptation, can be as significant
[00:16:58.011]as a major bottom-up driver like light.
[00:17:00.931]These are dark streams, we increase the light,
[00:17:03.584]in this particular experiment,
[00:17:06.200]I think the light went up by a factor of three,
[00:17:09.608]or four, I think.
[00:17:10.883]You know, a significant increase
[00:17:12.403]on light should have changed GPP
[00:17:14.201]and variety of things quite a bit,
[00:17:15.716]and yet we still see some of the phenotypic effects
[00:17:18.394]being of that same magnitude,
[00:17:21.355]which is really interesting.
[00:17:22.498]That was a real sorta eyeopener for me.
[00:17:26.357]And we see a very similar result
[00:17:27.628]when we look at the density effects relative to phenotype,
[00:17:30.120]and some of the other things,
[00:17:31.888]whether or not there's a predator there.
[00:17:34.700]So these are some of the references that show some of that.
[00:17:39.282]So this is a summary of what we found
[00:17:41.030]in those mesocosm experiments.
[00:17:42.965]And so here's our food web,
[00:17:44.232]this is basically what it looks like
[00:17:45.660]inside those mesocosms.
[00:17:48.328]All the solid lines are...
[00:17:51.821]Rates that we measured directly,
[00:17:53.767]dashed lines are ones we were not able
[00:17:56.804]to measure in those experiments.
[00:17:58.643]And the significant phenotypic effects are shown
[00:18:01.962]over here on the left,
[00:18:03.131]so we saw a significant effect
[00:18:04.089]on the algae and the invertebrates,
[00:18:05.480]that's what I showed in those data slides.
[00:18:07.268]But we also see that cascade
[00:18:08.493]in the gross primary production rates
[00:18:10.402]into leaf decomposition, into diet,
[00:18:13.869]which I showed as well,
[00:18:15.602]and also into fish excretion.
[00:18:17.901]So these fish excrete nutrients at different rates.
[00:18:21.537]And it also turns out they have different stoichiometries.
[00:18:25.585]The low-predation fish are more pee rich,
[00:18:28.048]therefore, excrete less pee.
[00:18:30.954]And you gotta think, a lot of these organisms
[00:18:34.854]have both direct and indirect effects.
[00:18:37.246]One of the things I'm not gonna have a chance
[00:18:38.437]to really talk about here for time's sake,
[00:18:40.205]but I'll just note it, is that there's a great paper
[00:18:42.271]by Ron in 2012 in which he parses direct
[00:18:45.941]and indirect effects using electronic exclosures
[00:18:49.603]inside the treatments.
[00:18:51.059]So inside each treatment,
[00:18:52.173]he has a little electronic fence,
[00:18:54.043]much like a cattle fence, but underwater,
[00:18:56.853]that excludes fish.
[00:18:58.827]But it receives the nutrients they excrete.
[00:19:02.382]And so he can parse the indirect effect
[00:19:04.292]of nutrient enrichment
[00:19:06.218]from the direct effect of consumption.
[00:19:08.972]And it turns out, these effects,
[00:19:10.467]in that case, it's opposite direction,
[00:19:13.103]and they're very similar in magnitude.
[00:19:15.136]In the end, the consumption one's up
[00:19:17.295]but the stimulation can be quite significant
[00:19:19.187]and I'll show that in a slide here in a minute
[00:19:21.569]when we look at some pool manipulations.
[00:19:24.339]So, a lot of the interesting ways
[00:19:26.044]to sorta look at these results.
[00:19:30.673]So the big question I always had in all of this
[00:19:33.760]is that I doubt, I'm just one of those people
[00:19:37.360]who doubts controlled experimentation. (laughing)
[00:19:42.145]I tend to study streams,
[00:19:43.706]streams are really dynamic,
[00:19:48.179]and yeah, you can do nice mesocosmic experiments
[00:19:50.350]and do studies in the lab,
[00:19:51.463]and they may mean absolutely nothing
[00:19:53.826]in our ability to predict what happens in nature.
[00:19:56.740]I see this a lot, actually.
[00:19:59.559]So, I was very curious to take the results
[00:20:01.065]that we found there and design work
[00:20:02.682]that we could do in natural streams
[00:20:03.969]to see if we see the same effects
[00:20:05.683]and compare whether or not if we do see those effects,
[00:20:07.590]are they a similar magnitude or not.
[00:20:09.819]And that's kinda what I'll spend a good chunk
[00:20:12.108]of the rest of the time talking about.
[00:20:14.444](softly talks off mic)
[00:20:18.839]So, as I just mentioned,
[00:20:22.182]I'll show you some nice slides, pretty pictures,
[00:20:23.943]but this place is not sunny very often.
[00:20:26.409]It has a dry season, but much of the year it's pretty rainy,
[00:20:28.923]and even in the dry season it can rain quite a bit.
[00:20:31.298]So, this is a pretty dynamic system
[00:20:33.927]and those kinds of flow variability, that--
[00:20:36.527](clears throat) Excuse me, that flow variability is absent
[00:20:39.243]in those mesocosms, we've taken that out of the picture.
[00:20:43.019]So, when we go into these natural systems
[00:20:44.795]at least we get to start that
[00:20:46.452]there's a lot of things that are different.
[00:20:47.938]One is the hydrology, these are flashing systems.
[00:20:51.136]So, this is the stage hydrograph for a few years
[00:20:54.648]down in Trinidad where we've got the wet season in white,
[00:20:57.228]the gray is representing the dry periods,
[00:21:00.092]and you can tell there's still a fair amount
[00:21:01.375]of rain in some of these dry periods,
[00:21:02.694]sometimes we get a pretty good recession.
[00:21:04.395]But these are very flashy systems, look at this site,
[00:21:06.527]these streams can go up a meter in about an hour
[00:21:10.274]if they get a good hard rain,
[00:21:11.702]to the point where you can't cross a stream
[00:21:12.787]and you get stranded there.
[00:21:16.289]So, the other things that sorta differ is
[00:21:18.496]not just the physical template,
[00:21:21.092]but also the complexity of the habitat results
[00:21:23.047]in a much more complicated, much more complex,
[00:21:26.140]biological community than we see in our mesocosms.
[00:21:28.733]Our mesocosms are largely dominated
[00:21:30.447]by a few invertebrates, mostly midges,
[00:21:32.574]they have very standardized substrate,
[00:21:34.391]rocks, gravels and fines.
[00:21:36.221]They don't have pools in riffles.
[00:21:39.288]There's a lot that's not there
[00:21:40.634]that we do have out in the natural system.
[00:21:42.565]So, there's a lot of reasons
[00:21:43.898]why we might not see those results perfectly translate
[00:21:47.059]to a natural stream.
[00:21:49.906]So, I'll start first by talking about some experiments
[00:21:52.705]that Troy Simon conducted down there,
[00:21:56.589]he was a grad student at University of Georgia,
[00:21:58.258]that I worked with he and Cathy Pringle
[00:21:59.871]on this particular work,
[00:22:02.115]and what he did is he did two different things.
[00:22:04.443]We went out and surveyed locations in which we had...
[00:22:09.733]We had natural pools that had existing,
[00:22:14.375]and then had upstream control areas that were guppy-free,
[00:22:17.210]just had rivulus,
[00:22:18.730]and then we had a whole series of other streams
[00:22:20.948]where we've done introductions of high-predation fish.
[00:22:24.043]So, we have, in this particular experiment,
[00:22:26.943]we had an introduction where we put high-predation fish
[00:22:29.982]just above the range of the low-predation fish
[00:22:33.867]in the Guanapo basin where we had our focused site.
[00:22:36.720]Extend the range about 100 to 200 meters.
[00:22:39.530]So in that location, you could go in and look
[00:22:41.996]at high-predation guppies in similar pools,
[00:22:44.766]and compare them to upstream controls where there's no fish.
[00:22:48.150]So, we kinda had a nice natural manipulation similar
[00:22:50.284]to what we had in the mesocosms.
[00:22:53.442]And then we also had these,
[00:22:54.997]so in this particular figure at the bottom,
[00:22:56.795]oh, wait, hold on, I'll step back.
[00:22:59.294]The next experiment he did,
[00:23:00.372]so, these are just without any manipulations,
[00:23:01.930]he didn't control density, he just went out,
[00:23:04.137]and these are naturally occurring,
[00:23:06.120]he sampled the standing stocks of insects and algae,
[00:23:08.909]looked at algal accrual, looked at those results.
[00:23:11.572]He also then went in and did manipulations.
[00:23:14.607]One of the nice things about working with these guppies
[00:23:16.943]is they're really easy to catch.
[00:23:18.739]All you have to do is sorta move your fingers,
[00:23:20.705]sorta noodle the bottom,
[00:23:23.968]I don't know, fishers people,
[00:23:24.920]you probably know what noodle is a little bit,
[00:23:26.465]and you sit there are sort of rough up some of the sediment
[00:23:28.706]and they'll just come right to you
[00:23:29.594]and you just scoop 'em up with a butterfly net.
[00:23:32.614]And so, in the mark and recapture aspect of the study
[00:23:35.063]I'll talk about in a little while,
[00:23:36.420]you know, you can recapture 95% of these pretty easily.
[00:23:40.333]It's a very nice system for that.
[00:23:42.168]The other advantage is you can take 'em all out.
[00:23:44.690]So, you can go out to these natural pools
[00:23:46.598]and you can manipulate densities, you can get rid of them,
[00:23:48.811]double the density, you can do anything you want.
[00:23:51.467]And they'll stay there, they have real high site fidelity.
[00:23:54.114]So what Troy did is he had these series of pools,
[00:23:57.350]he fished out all the fish, let 'em sit for a few weeks,
[00:24:00.411]and then he went back
[00:24:01.244]where he had three different treatments.
[00:24:02.911]So, he had a guppy removal
[00:24:04.822]where only thing in here is the rivulus,
[00:24:10.518]he had guppies with low density,
[00:24:12.437]and then he had guppies at high density.
[00:24:17.024]He also had a replicator where he had low-predation guppies
[00:24:20.369]and high-predation guppies.
[00:24:21.920]So, he basically had something that was very similar
[00:24:23.423]to what we had with the mesocosms.
[00:24:25.156]The other thing that he though is interesting,
[00:24:27.382]much like I mentioned briefly in the mesocosms,
[00:24:29.817]that he had these electronic exclosures
[00:24:31.783]in each one of these.
[00:24:34.257]And so, he'd have an electronic exclosure
[00:24:36.325]that was not electrified,
[00:24:37.846]it's just a metal grid on the bottom,
[00:24:40.199]just to make sure that that wasn't causing some effect
[00:24:44.390]on the behavior of the fish.
[00:24:47.346]Then he also had an electrified one that kept fish out.
[00:24:50.970]So these things basically electrified a grid,
[00:24:52.861]it causes a little current so the fish won't go in.
[00:24:57.163]But invertebrates do.
[00:24:58.502]You kinda have to fine-tune the voltage
[00:25:01.630]so that you keep the big things out
[00:25:02.996]and let the little things go back and forth.
[00:25:05.177]And by doing this,
[00:25:06.267]you get to tease apart the direct
[00:25:07.745]and indirect effects again.
[00:25:09.422]The direct effects of suppression,
[00:25:11.163]a little suppression by feeding,
[00:25:12.537]and algal stimulation by nutrient excretion.
[00:25:19.224]Alright, why isn't that going?
[00:25:27.834]This is not advancing, oops, sorry.
[00:25:32.141]There we go.
[00:25:35.684]Just a quick note here,
[00:25:37.148]you can see that these electronic exclosures,
[00:25:40.020]hopefully you can see this,
[00:25:41.274]you can see that they work fairly well.
[00:25:42.891]We've got this sorta nice green fuzz of algae around here,
[00:25:45.579]and you can see where the algae's been scoured all around.
[00:25:48.447]So these aren't minor effects either,
[00:25:49.543]much like what we saw in the mesocosms.
[00:25:51.852]And sure enough, when you go and you take a look at these,
[00:25:54.469]where we have rivulus only,
[00:25:56.247]this is the control situation,
[00:25:57.610]we have fairly equal amounts of algae,
[00:26:01.419]of chlorophyll light of standing stock here.
[00:26:03.641]Whether or not it's in the electronic exclosure or not,
[00:26:06.670]that's what these two color bars basically represent,
[00:26:09.445]where the shaded bars,
[00:26:11.181]the lighter colored bars are inside the exclosure,
[00:26:15.145]the electronic exclosure,
[00:26:17.086]and then the other ones are outside that.
[00:26:21.098]We had no cage effects in this particular experiment.
[00:26:23.831]And so you see, if you look at just where the fish were,
[00:26:26.519]when you have rivulus plus guppies,
[00:26:28.485]you see the suppression much like in mesocosms.
[00:26:30.579]If you double the density, you'll see a greater suppression.
[00:26:32.983]It's a pretty linear relationship that's quite nice.
[00:26:35.499]But inside those exclosures,
[00:26:37.581]we see increasing algal abundance.
[00:26:39.910]So, this decline here is a net effect
[00:26:43.282]of those direct and indirect effects.
[00:26:44.803]It's really, it could be and should be,
[00:26:47.310]actually, quite a bit less,
[00:26:48.511]by about the same amount of difference that you see here.
[00:26:51.923]And so, we see increases,
[00:26:53.055]and this is presumably due to the nutrient enrichment effect
[00:26:55.922]of those fish excreting nitrogen
[00:26:57.915]and phosphorous in the system.
[00:26:59.455]We know we get slightly detectable elevated concentrations.
[00:27:03.109]If you look and you calculate the excretion rates,
[00:27:05.390]it's a fairly...
[00:27:07.434]We can't account for it all in the concentration change,
[00:27:09.751]so you know it'd go somewhere.
[00:27:11.710]It's getting recycled back into the biology.
[00:27:14.760]So, kind of a cool way to look
[00:27:15.870]at these direct and indirect effects.
[00:27:20.676]Here on the left is the results
[00:27:22.710]of the surveys of natural pools without the manipulations.
[00:27:26.721]So, if you go and you look at low-predation sites,
[00:27:31.847]and compare, the variables here are all those,
[00:27:35.500]the control versus the exclusion,
[00:27:37.143]so where they're not there.
[00:27:39.310]Again, he had electronic exclosures,
[00:27:42.745]so, where he had fish in the same pool
[00:27:44.480]and where he didn't have fish.
[00:27:45.849]So, if you look at algal accrual rate
[00:27:47.303]and algal bio mass, like you'd expect,
[00:27:49.232]you see significant suppression
[00:27:51.288]of those in the low-predation guppy pools,
[00:27:54.385]no effect in the high-predation pools,
[00:27:57.523]you see a suppression of the invertebrate bio mass
[00:27:59.854]and the high-predation pools, you see some suppression,
[00:28:02.292]but it's not significant in the low-predation pool.
[00:28:04.320]So, this is almost exactly like
[00:28:05.943]what we saw in the mesocosms.
[00:28:08.574]If you go over and you look at...
[00:28:14.537]Oh, sorry, did I do this a little bit backwards?
[00:28:17.088]Yes, I did, these are from those exclosures
[00:28:20.294]that I was just talking about before
[00:28:21.588]where they had electronic exclosures
[00:28:24.252]in pools where they were manipulating the guppies, sorry.
[00:28:27.312]In the natural experiments shown here on the right,
[00:28:30.351]we've got native streams or low-predation guppies,
[00:28:32.469]and then these introduction streams,
[00:28:34.172]these are real streams of
[00:28:35.188]where we introduced high-predation guppies.
[00:28:37.209]And sure enough, you see algal suppressed
[00:28:39.885]between guppy-free versus
[00:28:41.445]where we had guppies and killifish.
[00:28:43.673]You see the suppression in the native streams
[00:28:45.566]where the low-predation guppies are,
[00:28:47.076]which actually gets stimulated a little bit
[00:28:48.382]in the high-predation.
[00:28:49.395]And the invertebrates, you see suppression
[00:28:51.968]where you had the high-predation fish,
[00:28:55.539]and in fact, you see actually a little increase--
[00:28:59.222](coughs) Excuse me, in those native streams,
[00:29:03.202]which is a little bit surprising,
[00:29:04.325]I didn't really expect to see that to go up.
[00:29:05.545]I would have expected that to go flat.
[00:29:07.021]But all in all, these sort of agree
[00:29:08.347]with what we saw in the mesocosms quite well.
[00:29:10.625]I mean, when we first got these results back,
[00:29:12.796]and when it was really kind of reaffirming
[00:29:15.674]that we're gonna be able to see this translate
[00:29:17.609]from what we saw in those stream-side channels
[00:29:19.451]to what we could actually go out and measure in the stream.
[00:29:21.975]In the back of my mind,
[00:29:22.988]I was always still a little concerned.
[00:29:25.182]Because when I go out to--
[00:29:26.667]While this was all happening,
[00:29:27.746]we had ongoing introductions of high-predation guppies
[00:29:30.320]into natural streams that we'd been monitoring,
[00:29:32.636]by this point, for a couple of years,
[00:29:35.233]and their populations had been growing up,
[00:29:37.034]and they about peaked by 2010.
[00:29:39.591]And when you walk out to those streams
[00:29:41.400]and you look above and below the waterfall
[00:29:43.442]where there are guppies and where there's not,
[00:29:44.953]you do not see obvious differences.
[00:29:48.568]You don't see a lot of algae upstream being seen
[00:29:52.403]and suppressed downstream.
[00:29:54.955]So, I always had that in the back of my mind
[00:29:58.107]as I was seeing these results.
[00:30:02.500]So, what happens in real streams?
[00:30:04.215]So, here's our four focal streams.
[00:30:06.416]We look at it in the whole reach scale
[00:30:09.200]and again, we've got two pairs,
[00:30:10.870]upper La Laja and lower La Laja, and Caigual,
[00:30:13.310]and Taylor both run parallel to each other,
[00:30:15.521]all four are fairly similar in size.
[00:30:19.510]Their drainage areas vary by just less than 10%,
[00:30:23.167]and so, we've got pretty nice paired systems.
[00:30:26.759]In each one of these, we had two reaches.
[00:30:31.131]So, this will pop up here in a sec.
[00:30:33.444]There we go.
[00:30:34.825]We had a control reach upstream of the upper most barrier.
[00:30:39.217]So, we have a fish barrier,
[00:30:41.020]and upstream control that is rivulus only,
[00:30:43.313]just that killifish.
[00:30:45.440]We have an introduction reach
[00:30:47.132]that's bounded by and upstream and a downstream waterfall
[00:30:50.001]where we put high-predation guppies.
[00:30:52.290]And then there's a downstream extraluminal reach
[00:30:54.353]that has a sort of mixture of high-predation guppies
[00:30:57.598]that are kinda coming down, interbreeding basically
[00:31:00.635]with low-predation guppies below it.
[00:31:03.776]And we have that in each site.
[00:31:05.181]Now, the difference between the blue and the green line here
[00:31:07.248]is in one of each of those pairs,
[00:31:09.128]one stream of each of those pairs,
[00:31:10.274]we went in and thinned the canopy
[00:31:13.033]to change the light condition.
[00:31:14.926]Because between high-predation
[00:31:16.028]and low-predation sites in nature,
[00:31:18.426]the high-predation sites are quite a bit bigger.
[00:31:20.428]They're wider, they have higher light,
[00:31:22.669]they have higher primary production.
[00:31:24.251]So, we wanted to know whether or not
[00:31:26.392]if you created that same condition,
[00:31:27.696]would you slow or change the local adaptation
[00:31:30.669]of those guppies when you put 'em
[00:31:32.069]in a low-predation environment,
[00:31:33.739]but it still had a lot of light,
[00:31:35.820]'cause these naturally have very, very little light.
[00:31:38.521]This is a dense three-story canopy,
[00:31:40.126]ambient light in most of these streams is maybe 5%
[00:31:43.956]in the middle of the day of ambiance,
[00:31:45.474]so they're really dark systems.
[00:31:47.446]So we went in, we cut the understory
[00:31:49.403]'course we didn't cut the overstory tree,
[00:31:51.644]we couldn't have done that,
[00:31:52.864]I wouldn't have done it if I could have done it.
[00:31:55.680]But we were able to clear out everything in the understory
[00:31:57.878]and increase light by about 50% to 60%,
[00:32:02.303]depending on which stream we're talking about.
[00:32:05.315]We did a little better job in the Caigual system
[00:32:08.210]versus the La Laja system
[00:32:09.563]because it had a lower canopy,
[00:32:11.705]it had been in production for cocoa,
[00:32:16.046]as well as fruit more recently,
[00:32:18.762]and so, the upper canopy wasn't quite as big.
[00:32:23.741]So this is sorta what these streams look like.
[00:32:25.672]This is our lower-barrier waterfall
[00:32:27.717]at our lower La Laja site.
[00:32:32.412]It's a terrible place to work.
[00:32:36.040]So, this just gives you some feel
[00:32:37.456]for what they'll look like.
[00:32:40.298]So, one of the things that's sorta important
[00:32:41.645]through all of this and what sort of the outer part
[00:32:44.216]or our research group has sort of been active in,
[00:32:46.682]and actually continues to be active in,
[00:32:48.374]is a very extensive mark and recapture
[00:32:50.092]looking at these guppy characteristics through time,
[00:32:52.661]tracking their local adaptation of phenotypic change.
[00:32:56.287]And this has been an absolutely massive undertaking.
[00:33:00.319]So again, as I've mentioned before,
[00:33:02.632]these fish, the nice thing about 'em
[00:33:05.669]is they're pretty easy to catch.
[00:33:07.262]So, when you go out and you do your mark and recapture,
[00:33:09.090]they have recapture efficiencies that, like I said before,
[00:33:11.862]range from about 93% to about 98%.
[00:33:14.413]They're getting pretty good,
[00:33:15.750]I think they stayed above 95 most of the time.
[00:33:17.650]So every month, they have a team of interns
[00:33:20.985]that goes out, collects all the fish
[00:33:23.355]in these studied reaches, puts 'em in Coke bottles,
[00:33:26.318]big plastic bottles, brings them back to the lab,
[00:33:29.399]puts them in other individual tanks
[00:33:31.592]for each pool they came out of, they take pictures,
[00:33:35.033]all new individuals get a unique elastomer tag, a tattoo,
[00:33:39.531]so they can track individuals.
[00:33:42.633]They take a genetic sample
[00:33:43.877]for doing some microsatellite work,
[00:33:47.190]and then put 'em back in the stream.
[00:33:50.369]So, they get the morphology of the fish,
[00:33:52.131]or the anatomy of the fish by photographs.
[00:33:54.795]So, lengths and shape and all these variety of things,
[00:33:56.958]and they can look at all these body changes
[00:33:58.860]and color changes through time.
[00:34:02.984]So, it's quite an effort, and this is still going on.
[00:34:06.497]They're now into their, they're just beginning the 11th year
[00:34:09.981]of doing this every month.
[00:34:12.305]It's an amazing pedigree
[00:34:13.825]that they've been able to develop from this.
[00:34:16.251]And this is the data through an early part of 2014.
[00:34:19.700]So, we had introductions in 2008,
[00:34:21.973]these are just the La Laja introductions.
[00:34:23.441]The other streams,
[00:34:24.371]they had the fish introduced a year later.
[00:34:26.435]And we conducted most of our manipulations
[00:34:28.189]in the La Laja sites
[00:34:29.539]'cause they're a little bit more pristine.
[00:34:31.165]So I'll sort of focus on that for now.
[00:34:33.314]So as you can see, there's a lot of variants in here.
[00:34:36.086]Notice here, I've got the blue now is the rainy season
[00:34:39.106]and the white is the dry season.
[00:34:41.237]I didn't make this, sorry for the confusion on that.
[00:34:43.092]But one of the things you can see,
[00:34:44.549]in upper La Laja, that's the one
[00:34:46.442]where we cleared the canopy in increased the light.
[00:34:49.106]And you can see those populations,
[00:34:50.791]especially in those first few years
[00:34:52.441]as they started that build up,
[00:34:54.291]really were quite a bit higher
[00:34:55.790]than we saw in our low-light system.
[00:34:58.304]And really happy to see that effect, and I'm like,
[00:35:00.337]"Ah, yeah, this is really working well, this is great."
[00:35:02.489]It turns out that actually starts
[00:35:03.798]to go away a little bit over time.
[00:35:06.158]In part that has to do, when we were out here in 2012,
[00:35:09.419]I wasn't going down there maintaining that
[00:35:11.546]as well as I had been before.
[00:35:13.964]So, we gotta make sure that this sorta convergence
[00:35:16.499]that we see out here isn't in fact due
[00:35:18.391]to lack of maintenance of that treatment.
[00:35:21.095]I've been gone through these three years,
[00:35:22.669]so I'm hoping that's not the case.
[00:35:24.103]I've been told that's not true, so it's a little unclear.
[00:35:26.554]But one of the things we see in this
[00:35:27.705]is that in the dry season,
[00:35:29.011]we almost always see the high-light system
[00:35:31.067]overshoot the low-light system.
[00:35:33.442]And we got about four to five months,
[00:35:36.114]or three to four months, really,
[00:35:37.700]of pretty good dry conditions.
[00:35:39.436]That's the window where biological
[00:35:41.482]activities sorta takes off,
[00:35:43.091]we get increasing standing stocks of everything,
[00:35:45.602]algae, invertebrates, and the fish sorta track that.
[00:35:48.745]It's a little hard to pick out there,
[00:35:51.259]but we've got some statistics that show that pretty clearly.
[00:35:59.010]So, sorta the main experiment
[00:36:00.747]that's sorta why I was actually involved in this project
[00:36:03.231]that we wanted to do is we wanted to do something
[00:36:06.411]that was much more integrative,
[00:36:10.569]a series of measures that really were
[00:36:12.188]at the ecosystem's scale to see
[00:36:13.860]whether or not these phenotypic effects really cascaded,
[00:36:16.313]really changed the way these systems operate.
[00:36:18.926]So, one of the things we were doing
[00:36:20.239]is measuring oxygen concentrations
[00:36:22.924]for different weeks to get metabolism rates,
[00:36:26.733]so post-primary production and then community respiration,
[00:36:30.496]you got a paper that's just being submitted
[00:36:32.341]right now on that.
[00:36:33.774]And sure enough, we see some stimulation
[00:36:36.755]of primary production where the guppies were put in,
[00:36:39.913]which is interesting.
[00:36:41.682]But if you think about it,
[00:36:42.548]we put in high-predation guppies,
[00:36:43.875]they are largely invertivors.
[00:36:45.585]So, it's not surprising you might have a cascade
[00:36:47.800]that actually increases algal activity.
[00:36:50.454]Even though in those mesocosms,
[00:36:52.036]we saw growth suppression.
[00:36:53.598]We're starting to see algal activity anyway
[00:36:56.919]actually stimulated where the guppies are.
[00:36:59.254]We did not see changes in algal standing stock
[00:37:01.753]that were statistically significant.
[00:37:03.530]We've seen a few significant changes
[00:37:05.477]in the invertebrate community, mostly in midge suppression,
[00:37:08.587]which isn't too surprising, (softly talks off mic)
[00:37:10.440]not so much in the larger bugs.
[00:37:14.321]So, one of the things we did
[00:37:16.084]to try to get this sorta course
[00:37:17.695]or a sorta functional assessment of the stream
[00:37:19.947]is to use N15 isotope additions,
[00:37:22.703]to inject an isotope into the system
[00:37:24.896]and then track where it goes.
[00:37:27.701]This is a technique that I've been sorta doing quite a bit,
[00:37:30.280]I did it through, kinda was involved in these experiments
[00:37:32.700]as a post-ac and that's what sorta got be involved here.
[00:37:35.908]So let's walk through real quick on sorta how these work.
[00:37:38.829]We basically drip, this is a tub
[00:37:40.743]of Rhodamine dye and N15 ammonium,
[00:37:43.750]and that gets pumped into the stream for 10 days.
[00:37:46.864]We kinda bring it up so we can kinda saturate the system,
[00:37:49.657]and then we turn it off and we can watch it clear.
[00:37:52.199]So, this is not how red the stream was,
[00:37:55.553]this was graduate students not knowing how to do Rhodamine,
[00:37:58.208]but it's a nice picture of Rhodamine, so I put it in there.
[00:38:00.272]So, this is our tracer, so we know what the water's doing
[00:38:02.600]and we can look at the separation of the isotope,
[00:38:05.316]the ammonium in the water from that conservative dye,
[00:38:10.314]and we can get the uptake rate of the ammonium.
[00:38:12.678]You can also get conversion to nitrate
[00:38:14.084]and all kinds of that stuff.
[00:38:16.233]So basically, what we have is a system in where,
[00:38:18.637]this is the isotope proportion,
[00:38:21.026]and here's our addition period shown here in the shading.
[00:38:23.911]So, the ammonium's just gonna bop up
[00:38:25.440]to some high enrichment, stay there for 10 days,
[00:38:28.073]and then come back down.
[00:38:29.612]I'm not sure why he's got this modeled for 15,
[00:38:31.537]but it was really for 10.
[00:38:33.904]You can see here's what algae are likely to do,
[00:38:36.181]they're gonna slowly enrich,
[00:38:38.096]depending on how long you go,
[00:38:39.391]they may plateau, they may not, and then they decline.
[00:38:42.158]Grazers are gonna track that but lag a little bit,
[00:38:45.167]and then of course predators will lag behind the grazers,
[00:38:47.561]that's the general idea.
[00:38:51.126]And from that, then you can construct
[00:38:53.355]these nitrogen flow diagrams, ultimately,
[00:38:56.017]when we get through the whole end of this,
[00:38:57.035]where you can track the fate of nitrogen
[00:38:59.342]from the water column into the specific pools
[00:39:02.523]of primary uptake compartments,
[00:39:04.659]algae, microbes associated
[00:39:06.307]with the different detrital pools and the like.
[00:39:09.329]And then those food resources into their consumers.
[00:39:12.731]So these are all aquatic insects,
[00:39:14.587]this is Phylloicus is a shredder that eats leaves.
[00:39:17.293]This is a crab, a general list of mayfly here,
[00:39:22.391]this is a couple of grazers and a (softly mumbles)
[00:39:26.034]And so, we can see how they change
[00:39:27.622]from one treatment to another,
[00:39:29.331]don't worry about the numbers,
[00:39:30.270]I'll show you that in a little bit in a second.
[00:39:32.708]So, the standard approach for analyzing is,
[00:39:34.682]again, load these up and we sample at different locations,
[00:39:38.132]and then we watch them clear
[00:39:39.636]after we turn the dripper off.
[00:39:41.395]And you can model that clearance,
[00:39:42.849]with just an exponential decay,
[00:39:44.212]especially for the primary uptake compartments,
[00:39:46.679]and you can calculate the turnover rate or the uptake rate
[00:39:49.933]which are equivalent as the standing stocks not changing,
[00:39:52.177]from the slope of that line.
[00:39:54.421]And you can go through that and do that
[00:39:56.689]for all the different compartments.
[00:39:58.593]Well, that's okay,
[00:40:00.209]and that's the traditional way to do this.
[00:40:02.194]And I'll talk in a little bit,
[00:40:03.509]but for us, that became problematic.
[00:40:05.455]I'll talk about that more in a minute.
[00:40:08.016]But using that sort of approach,
[00:40:12.127]we've been publishing using
[00:40:13.441]that technique for quite a while,
[00:40:15.129]this has been a focus of work for probably the last 15,
[00:40:18.716]at least 15 to 20 years now,
[00:40:20.414]where isotopes have become more available
[00:40:22.077]and the analysis becomes a little bit more affordable.
[00:40:24.200]And so, you'll notice here, if you look at this,
[00:40:26.799]here's the thin canopy no-guppy, thin canopy guppies,
[00:40:29.681]so this is the high-light system, upstream and down,
[00:40:32.198]and the low-light system.
[00:40:33.542]Notice there's not massive changes here.
[00:40:36.046]I mean, we don't have full rearranging of these.
[00:40:38.513]The size of these of course
[00:40:40.682]represent the magnitude of the flux,
[00:40:43.376]these gray lines over here
[00:40:45.298]where the leaves and FBOM are insignificant,
[00:40:47.549]they did not change between either the thin canopy
[00:40:50.632]and where we have guppies or between the light of...
[00:40:54.386]There's no significant effect.
[00:40:56.340]The open ones had positive significant light effects,
[00:40:59.173]and the dark ones had both light and guppy effect.
[00:41:02.513]So, even though these bars don't change a lot,
[00:41:04.765]for example, if we look here in the thin canopy,
[00:41:06.477]this was obviously wider than we see up here for Psephenus,
[00:41:09.535]you know, we get some shifting,
[00:41:10.696]a little bit more stimulation in the Epilithon.
[00:41:13.604]It seems to be funneling a little bit more
[00:41:14.984]up into these grazers than we saw
[00:41:18.109]when there's not guppies around.
[00:41:19.799]We see some similar changes over here on the low-light site.
[00:41:23.770]But these are fairly modest changes,
[00:41:25.862]and one of the things we really wanna do is be able
[00:41:28.009]to statistically test whether or not our light treatment
[00:41:31.215]and our guppy treatment that ultimately,
[00:41:33.121]the evolutionary change, the local adaptation,
[00:41:35.894]we need a technique that statistical rigors.
[00:41:40.778]It allows up to really kinda answer those hypotheses
[00:41:44.213]in that framework.
[00:41:45.592]And the error of this (quickly mumbles)
[00:41:48.273]really becomes a little bit problematic.
[00:41:51.822]So, I'll talk about that in a second.
[00:41:53.488]So again, just like we did in those mesocosms,
[00:41:55.447]we can parse these effects.
[00:41:56.791]So, this is again from Sarah's paper
[00:41:58.441]that came out last year.
[00:42:00.919]And so again, you can see that the guppy effect,
[00:42:03.142]the light bars here, is pretty much
[00:42:05.349]on par with the light effect,
[00:42:07.225]very similar to what Rana saw in her mesocosm experiment.
[00:42:10.723]And also very encouraging for us.
[00:42:12.134]It's really nice to sorta see these things cascade
[00:42:14.616]because that is not often what we see.
[00:42:16.768]So, really kinda nice to see
[00:42:18.404]that we see a very similar magnitude of effect
[00:42:20.805]between the phenotype and the light.
[00:42:25.645]So similar effects to what we saw in the mesocosms.
[00:42:29.133]Now, that standard approach
[00:42:30.146]that I just talked about is a little bit problematic.
[00:42:32.479]For all practical purposes,
[00:42:33.637]it really becomes an enormously sophisticated observation.
[00:42:38.756]Because we don't have great replication,
[00:42:40.508]we don't have good ways to put error around those,
[00:42:42.609]and as a result, we don't have good ways of running stats,
[00:42:45.237]on the whole system,
[00:42:46.674]when we compare the high light to low light,
[00:42:49.254]or compare guppy-free to where we have guppies,
[00:42:53.716]or one year to the next year.
[00:42:56.131]So, there's kind of some issues with that.
[00:43:00.610]So there's really no integrated analysis,
[00:43:03.030]no nice ability to kinda make
[00:43:04.710]those quantitative comparisons.
[00:43:06.609]And so, that's what we've been working on actually
[00:43:09.402]for the last five years.
[00:43:11.019]Kind of a bit slowly, we're making progress sorta steadily.
[00:43:14.179]And just got funding about a year and a half ago,
[00:43:16.801]a year ago, we have to get a little creative these days.
[00:43:19.514]We've have good luck within the stuff
[00:43:21.893]and I've made it to the full proposal stage a couple a times
[00:43:24.608]on this, and I'm not just quite there,
[00:43:26.996]I'll put it back in the spring.
[00:43:28.384]But we did get some money,
[00:43:29.591]in the meantime, from the Academy of Finland, of all places.
[00:43:32.876]Because one of our post-acs, Andres Lopez Sepulcre,
[00:43:35.516]is now a faculty member at University of Helsinki,
[00:43:39.308]and so, he and Rana, Sarah Collins and I got money
[00:43:42.100]about a year and a half ago to go back.
[00:43:44.140]Now that we're 10 years out,
[00:43:45.272]maybe continuing that mark and recapture,
[00:43:47.477]we know the phenotypic changes, now almost fully occurred,
[00:43:51.202]now we can go back and go look at what fish now,
[00:43:55.296]that should be more like low-predation fish,
[00:43:57.212]how they're affecting the system.
[00:44:00.594]So, what we need is a good way of comparing
[00:44:03.503]what we see in this coming year
[00:44:04.807]to what we saw back in 2010.
[00:44:08.487]Just like we need good reason, good ways,
[00:44:10.544]of comparing the two streams.
[00:44:12.221]And so, the new approach is to basically use a process model
[00:44:16.059]that uses a matrix sorta formulation.
[00:44:19.352]So again, we're doing a fairly simplified subset
[00:44:22.761]of the food web, and again,
[00:44:24.940]this is all based on the nitrogen flow,
[00:44:27.151]I'm not gonna go into the details of this.
[00:44:28.707]We can talk about it a little bit later.
[00:44:31.062]I may have to call the people
[00:44:32.365]if you have too complicated a question.
[00:44:35.508]But ultimately, what happens is so,
[00:44:38.457]this has been some work
[00:44:39.880]that Matthieu Bruneaux has been working on.
[00:44:41.883]He's a post-ac with our group right now,
[00:44:43.945]just an excellent programmer,
[00:44:45.880]and this is the simple food web
[00:44:47.669]that he has been working with.
[00:44:49.014]So, these color codes will stay the same,
[00:44:51.699]we've got the ammonium and nitrate Epilithon, Tricorythodes,
[00:44:56.630]these are consumers that we're modeling,
[00:44:58.645]these are all grazers,
[00:45:00.375]so we're looking at sorta the autotrophic line
[00:45:03.193]of the food web, all the way up
[00:45:04.868]to one predator here is Argia.
[00:45:06.473]We ultimately can take this to fish
[00:45:08.057]but the problem is they don't like us
[00:45:09.545]to paint a lot of fish,
[00:45:11.264]'cause we're doing the mark and recapture.
[00:45:12.547]They're awfully paranoid about us killing
[00:45:14.458]and sort of run the isotopes on all their fish.
[00:45:17.990]So, we're trying to work that out
[00:45:19.572]as we sort of speak.
[00:45:23.015]This is sorta what the data looks like.
[00:45:24.573]So here, we've got multiple transects,
[00:45:26.283]here's our ammonium, here's our different critters,
[00:45:28.710]here's our predator,
[00:45:30.738]and this is what the data looks like.
[00:45:32.300]This is the 15N data, here is the ammonium,
[00:45:34.393]we had three different transects at upstream,
[00:45:38.107]sorta midway down the reach,
[00:45:40.139]and then all the way at the bottom,
[00:45:41.458]you can see that ammonium is lost when you go through that.
[00:45:44.911]Ammonium only travels in these streams about,
[00:45:46.784]on average, about 60 meters before it's taken up.
[00:45:52.791]Some of that is taken up by biology,
[00:45:54.171]but some also gets converted to nitrate.
[00:45:56.667]There's a nitrification in here
[00:45:57.968]so we can track nitrate as well,
[00:45:59.140]and then we can track those things
[00:46:00.418]into these different compartments.
[00:46:03.554]And so, what we do is run
[00:46:04.569]this Markov Chain Monte Carlo stimulation.
[00:46:07.145]And so, in this particular case,
[00:46:08.496]we do this sorta simultaneously
[00:46:10.278]for all the different fluxes,
[00:46:12.026]all the different connections in the food web.
[00:46:14.431]So, what we have here is sorta a traces,
[00:46:16.693]this is 1000 iterations of that simulation,
[00:46:19.858]the different colors are different starting values
[00:46:22.809]to ensure that no matter what you start with,
[00:46:25.357]you converge on the same thing.
[00:46:28.585]And this is for uptake of ammonium to Epilithon,
[00:46:31.056]this is uptake of nitrate to Epilithon,
[00:46:32.800]so we get basically this distribution of value
[00:46:36.110]that we can use to place (softly talks off mic)
[00:46:39.556]so that allows us to produce something that looks like this.
[00:46:42.288]So here's the ultimate data for one of those releases,
[00:46:45.155]this is going back and modeling our 2010 releases
[00:46:47.649]that Sarah had done.
[00:46:49.424]And so, here's the Epilithon,
[00:46:51.711]and we've got different confidence intervals illustrated
[00:46:55.448]by the dark and the light shading.
[00:46:58.194]And so, here's the Epilithon,
[00:46:59.842]Tricorythodes, this is one of the grazers,
[00:47:01.664]Petrophila's another grazer, Psephenus's a water penny,
[00:47:05.074]it's another grazer, and then here's our predator.
[00:47:07.398]And so, you can see that this all
[00:47:09.029]sort of makes sense with one exception.
[00:47:11.746]This is something that we can talk about later
[00:47:13.672]when I have time to talk about it, obviously,
[00:47:15.447]is that notice that the Tricorythodes,
[00:47:18.333]this is a grazing insect,
[00:47:20.422]is more labeled than its food resource, the Epilithon.
[00:47:24.323]And that's because we are perfectly able
[00:47:27.798]to sample that material.
[00:47:31.344]When we sample Epilithon off a rock,
[00:47:33.075]that is the biofilm growing on the top surface of rock,
[00:47:35.043]not all of that, it's not a pure algal culture.
[00:47:39.483]Only some of that's algae,
[00:47:40.972]there's bacteria, there's detritus,
[00:47:42.428]and all the nitrogen that's in there
[00:47:44.169]gets integrated in that measurement.
[00:47:45.780]And it's very likely that some
[00:47:46.834]of these insects are selectively feeding
[00:47:48.298]or selectively assimilating.
[00:47:49.930]So, we see this over enrichment a lot.
[00:47:52.311]And there's some methods that have developed,
[00:47:54.034]we just had a paper come out in Ecology a little bit ago
[00:47:55.850]about how to deal with that.
[00:47:57.157]And that's what we did here.
[00:47:58.432]So, we've now gone through
[00:48:00.021]and kind of sorta solved that a little bit,
[00:48:01.644]and are in the process of modeling some
[00:48:03.950]of our other editions to sorta make
[00:48:05.414]these quantitative comparisons.
[00:48:08.313]Then in the end, you wind up getting a food web
[00:48:12.358]that has nitrogen fluxes like I've shown here.
[00:48:16.193]So that's where this is all heading.
[00:48:18.760]So that's kinda where we are,
[00:48:19.881]we've modeled the systems so that this is
[00:48:21.616]sorta the high-predation, in this case,
[00:48:23.400]it's the high-light system.
[00:48:25.818]So, we have this for the high-light and low-light,
[00:48:27.809]when the press is publishing the paper right now,
[00:48:29.510]it's publishing this model and making
[00:48:31.617]that initial 2010 comparison
[00:48:33.735]between no guppies and guppies,
[00:48:36.022]and high-light and low-light,
[00:48:37.220]much like Sarah did in comparing that.
[00:48:38.934]And then of course now, we're gonna be doing it
[00:48:40.869]after local adaptation is carried out now in our 10th year.
[00:48:44.730]The newest introductions were in 2007,
[00:48:47.624]we'll be going back next spring,
[00:48:49.401]this coming spring, to repeat all of this.
[00:48:53.569]So, just to kinda summarize what did we do,
[00:48:56.481]so, we had our first year down there where had streams
[00:49:01.085]without guppies, just rivulus,
[00:49:03.810]low-light, no trimming,
[00:49:05.724]we'd gone out, we had standing stock sampled,
[00:49:07.883]all the background data.
[00:49:11.924]Then the canopy, went back in in a series of measurements
[00:49:16.575]through the dry season, at the end of the dry season,
[00:49:19.452]we introduced guppies.
[00:49:21.105]So, the next year, we had a system
[00:49:22.622]where we had high and low-light systems
[00:49:25.052]and we had guppy reaches and non-guppy reaches,
[00:49:28.773]this is where that project sorta ended.
[00:49:30.945]This is the condition that I was just sorta showing.
[00:49:34.454]What do those systems look like
[00:49:35.322]when we put high-predation phenotypes in there?
[00:49:37.845]So there've been some modest changes
[00:49:39.226]in those phenotypes and (softly talks off mic)
[00:49:42.201]What we're really interested in is,
[00:49:43.433]of course, what happens over time.
[00:49:45.551]This was always originally sort of conceived
[00:49:47.900]as about a 10 to 20-year project
[00:49:49.655]where we really track what this phenotype effect
[00:49:52.234]in the natural ecosystem look like.
[00:49:55.052]So, this is where we're gonna go back
[00:49:56.266]and do this spring.
[00:49:57.681]So, when I'm not around this spring,
[00:49:59.309](softly mumbles off mic)
[00:50:03.389]Alright, so I'm getting pretty close,
[00:50:05.380]I think, I'm just gonna take a couple of minutes
[00:50:08.589]just introduce one other thing.
[00:50:10.298]One thing I wanted to make sure I mention is, of course,
[00:50:12.358]I didn't do this by my own, by any means.
[00:50:14.482]This was a large team, we had 11 PI's,
[00:50:17.245]five post-acs, a whole slew of graduate students.
[00:50:20.974]And more importantly, we put through over 100 interns
[00:50:24.402]that went down and helped us with the study
[00:50:27.382]over the five-year period.
[00:50:28.847]This number now is, as some of the subsequent funding
[00:50:30.779]we've got, is closer to 200.
[00:50:32.672]It's been an amazing sorta program,
[00:50:35.069]they do three-month stints to do the mark and recapture,
[00:50:37.570]takes about three weeks, they get a week off,
[00:50:41.020]and they go back and do it all again.
[00:50:43.027]It's just been an amazing effort.
[00:50:48.003]So, in just real quick order,
[00:50:50.647]'cause I got about five minutes at the most,
[00:50:52.967]I just wanna try to say one of the other things
[00:50:54.922]I've been doing is kinda,
[00:50:57.660]when that ended, we got funded
[00:50:58.885]by the Dimensions of Biodiversity program
[00:51:00.702]to look at distributions of aquatic insects
[00:51:03.474]and how they vary between Colorado and Ecuador.
[00:51:07.100]In a way, what we're trying to do is test
[00:51:09.143]whether or not climate variability
[00:51:12.381]creates greater speciation,
[00:51:14.791]and greater species turn over with elevation in the tropics
[00:51:17.887]versus what we see in the temperate zone.
[00:51:20.191]And this was the hypothesis that Janzen,
[00:51:22.096]back in the 70's had put forth
[00:51:23.720]in his paper that sorta questioned
[00:51:25.861]whether or not mountain passes were higher in the tropics
[00:51:28.519]than they were in the temperate zone,
[00:51:30.294]which on the surface level, doesn't make any sense,
[00:51:33.089]height is height.
[00:51:34.278]But what he was meaning is that is the thermal barrier
[00:51:36.953]that it takes for a ectotherm
[00:51:39.809]to be able to get from one basin into the next basin
[00:51:42.602]is that pass, such a barrier,
[00:51:45.254]because the temperature's so far outside its range
[00:51:48.657]that it's acclimated to, that it's adapted to,
[00:51:50.828]that it can't do it.
[00:51:52.495]So that you affected the isolate populations
[00:51:54.496]as you cross mountain ridges.
[00:51:56.953]And as such, you get diversification
[00:51:58.607]across different valleys.
[00:52:00.075]You also get very little gene flow as you go up
[00:52:02.524]because species tend to become very thermally specialized,
[00:52:06.209]occupying very narrow thermal (mumbles off mic)
[00:52:11.149]And so, I'll just sorta jump to the conclusions on this
[00:52:13.313]is what we've been finding,
[00:52:14.893]and what our big finding right now is,
[00:52:16.770]is that sure enough, what we do see
[00:52:18.925]is we see greater species turnover
[00:52:20.960]with elevation in the tropics,
[00:52:22.968]greater species richness in both mayflies
[00:52:26.555]and caddisflies down there, that had not been shown.
[00:52:28.775]Aquatic insects had always been thought to be an anomaly.
[00:52:31.161]They've always been through to be more diverse
[00:52:32.923]in the temperate zone than in the tropics.
[00:52:34.529]Turns out, it's just a matter of...
[00:52:38.536]The taxonomy hadn't been that well worked out.
[00:52:41.032]The morphology doesn't vary that much,
[00:52:42.654]you really need to use genetic tools
[00:52:44.551]to really kinda see that picture.
[00:52:46.452]So when you use genetic barcoding,
[00:52:48.004]you see that pattern emerge really clearly
[00:52:50.225]for caddisflies and for mayflies,
[00:52:52.239]you don't see that happen for stoneflies.
[00:52:54.775]Stoneflies are more diverse here
[00:52:56.240]than they are in the Andes.
[00:52:57.619]So, that seems a little odd.
[00:52:59.803]Well it turns out, we also did a bunch
[00:53:01.138]of microsatellite work, we did rates
[00:53:03.411]to be able to quantify dispersal and gene flow,
[00:53:05.990]and rates of evolution.
[00:53:07.947]And what we winded up finding is
[00:53:10.690]that the evolutionary rates are always faster in the tropics
[00:53:13.977]than they are in the temperate zone, even for stoneflies.
[00:53:17.923]So that leads us to the question, well, how does that work?
[00:53:20.951]How can they be evolving more quickly,
[00:53:24.528]and yet be less diverse?
[00:53:26.772]Then of course, the answer is
[00:53:27.671]they got there much more recently.
[00:53:29.590]They're a recent invader of the tropics from North America.
[00:53:32.573]They simply haven't been there long enough.
[00:53:35.125]So, it's a really nice sorta complimentary sorta set
[00:53:38.202]of measurements that really were able to identify that.
[00:53:41.928]So, with that little sorta tease,
[00:53:43.635]I was hoping to talk more about that,
[00:53:44.981]but couldn't really do a good job at both.
[00:53:46.533]So, I'll just sorta leave it at that
[00:53:48.415]and thank you guys for coming.
[00:53:49.547]I know it's not easy to do at the end of the semester.
[00:53:52.536]So, thanks a lot.
[00:53:58.717]by things like kingfishers and small herons on the forest?
[00:54:02.597]Not in the low-predation sites,
[00:54:04.442]probably more so in the high-predation sites.
[00:54:06.324]The problem with the low-predation,
[00:54:07.737]there are bird predators, but those aren't the ones.
[00:54:10.571]There's a few other birds, I wish I could tell you
[00:54:13.795]what they are, I'm a terrible naturalist,
[00:54:15.249]but there is a little bit.
[00:54:16.384]The other predator is...
[00:54:19.443]So, frogs, to some degree,
[00:54:22.596]depending on where you're looking at,
[00:54:23.968]but for the fish, there are some small birds.
[00:54:29.153]The rivulus are weird because they flip across the ground
[00:54:31.902]to go try to find small pools
[00:54:33.523]where there's mosquito larvae,
[00:54:35.294]and you can find them 50 meters, from me to you,
[00:54:37.915]off a stream, just in the middle of a forest,
[00:54:40.128]just flipping across the ground.
[00:54:41.803]So, they get preyed on by a lot of things.
[00:54:43.700]The guppies are largely,
[00:54:45.483]their predators in the low-predation area
[00:54:47.441]are largely rivulus feeding on their young,
[00:54:50.759]they're live-bearers, so their young are vulnerable,
[00:54:56.430]and some small birds, but it's pretty minor.
[00:55:03.248][Moderator's Voice] Any other questions?
[00:55:08.836]End of the day.
[00:55:13.543]on the guppies...
[00:55:16.274]So, I guess what I'm driving at is--
[00:55:17.440]So, how long do they live or how often do they reproduce?
[00:55:20.861]about the mark-recapture stuff that you did,
[00:55:22.393]if they're going back once a month--
[00:55:23.359]Wanna say it's every three months,
[00:55:27.148]For reproduction cycle, for a given female,
[00:55:29.495]I think she'll have a clutch every three months,
[00:55:32.548]I think is about what it is.
[00:55:34.167]And then they live three to four years, it depends.
[00:55:42.181]You know, I can't remember how they quantify the number
[00:55:44.581]of generations that have already occurred,
[00:55:46.589]but I wanna say it's...
[00:55:50.931]It's in the double digits.
[00:55:59.872][Moderator's Voice] Any other questions?
[00:56:03.104][Moderator's Voice] Alright, so thanks.
Log in to post comments