Tip:
Highlight text to annotate it
X
One of the things that’s characteristic of my field of ecology that’s actually related
to research in the social sciences, is that in ecology if you’re doing field work —
like this dataset entails — you really are dealing with a very complex ecosystem.
There are many, many variables out there that you could measure, but you can only measure
a small subset of them — and even fewer of them can you actually control experimentally.
Let me open this up here. You might just refer to it as the caterpillar dataset, but it actually
involves interactions between caterpillars, the plants that caterpillars eat, and also
strange and morbidly fascinating insects called parasitoids that attack caterpillars which
I’ll tell you more about in a minute.
I first just want to start off by telling you about ecology. Ecology is a science of
how organisms interact with their natural environment. This is actually extremely wide-ranging,
and this can include many, many different kinds of things. You can study ecology at
a global scale; you can study it at an extremely local scale like all of the microorganisms
living in your mouth or something like that. Ecology is extremely broad in its scale of
both time and space, and these are just a couple of examples of the interactions going
on — a butterfly visiting a flower; extracting nectar from a flower; it may pollinating the
flower in the process. Here is a deer that’s browsing on some vegetation, and so it’s
an herbivore-plant interaction.
Now, ecology happens to be pretty important because it helps you to understand how the
natural world out there works at all of these different scales. What you can do with that
information is learn a lot about how you might want to — not only just understand the world,
but how you might want to manage the world. This is an increasingly important thing for
the society.
As you probably realize, humanity dominates the earth. You cannot say, “Well, let sort
of ecology just do its thing and we’ll live over here, you know?” We control, to a great
degree, how ecology works because of our influence through human population size, human activities.
We’re really at a point where we need to understand the world in order to manage it
sustainably, okay? So whether you’re interested in, you know, preserving a cute little monkey
in the rainforest or whether you’re interested in the ecological interactions behind food
production, all of these things depend on the basic science of ecology.
One big aspect of an ecosystem that’s important to understand is the food web. The food web
is just a group of organisms that are interacting in that ecosystem through what are called
“trophic” interactions. That is who eats whom.
There are various aspects of that you can measure. Now, this is from a ecology textbook
just showing you a food web in a desert ecosystem. What you can see is that it’s pretty complicated.
These are just different groups of organisms, or sometimes particular species as you see
up above. All of these arrows just show the food web interactions, the trophic interactions.
The arrowhead points to essentially the consumer in an interaction. Here we have insects that
are consuming plants, and the arrowhead is pointing to the insect. We have arachnids
and spiders and other things — scorpions — these are eating insects and so forth.
Then there are some double-headed arrows because things eat each other and so forth.
This is just to give you a sense of what reality is like, and then what kind of research problems
ecologists are up against here. I’m really just going to focus on one particular set
of interactions in food web. This is the focus of my research and the dataset that you have
at your disposal. This is going to be kind of down here on the food web — the plants,
the insects that eat the plants, and then these other strange insects that I’ll tell
you about called “parasitoids.”
First, I’ll just tell you about herbivory. This is the herbivores eating the plants interaction.
This is a key food web interaction. Everywhere you go — any ecosystem that you go —
you have plants and you have herbivores interacting. They form the base of most food webs, and
so this is very fundamental to ecology.
You wouldn’t necessarily know it from this last picture, but these also make up the bulk
of the biomass in food webs. If you look outside, the world is green more or less. The plant
biomass is huge. Every time you go up a trophic level, and to the next step there are insects,
arachnids and then on up to the top predator at the top — the golden eagle there —
you get less and less biomass as you go up, and so a lot of what’s happening in the
world concerns these interactions, most of the biomass.
That’s why these are really important. For one thing, there’s a lot of diversity in
the insects that are eating plants. They may up the majority of herbivores, and then the
plants themselves are very diverse as I’ll tell you in a minute.
One thing that I really focus on in my research, and it’s also in this dataset that you have,
is the dietary breadth of the insect herbivores. Most species of insects that eat plants have
very specialized diets, and it’s been estimated — though we don’t know this because there
are so many out there that we have to estimate it — that about 90 percent of insect species
that eat plants feed on a very small set of related plant species.
That’s really different from the herbivores you might be familiar with like cows or human
beings which eat a lot of plants. Like we’re real generalists; we eat a lot of different
kinds of plants. Cows eat even more different kinds of plants than we do, and goats eat
even more than them. Most insects actually just eat a very restricted set of plants.
You might be familiar with something like a Monarch butterfly caterpillar. A lot of
people have come across this in their youth at some point, and these caterpillars only
eat milkweed plants, you know? They’re very specialized.
What’s interesting about caterpillars in temperate forests — like those around here
— is that actually many of them are much more general in their diets than you would
find say in a tropical rainforest. That’s actually kind of useful, because we can look
at herbivores and their interactions across a bunch of different plants that they eat.
Okay, so why does this dietary specificity or the diet breadth of caterpillars and other
herbivores matter ecologically? Well, really the strength and the structure of food web
interactions depends on this, and so it turns out to be extremely consequential for how
a food web works — what the structure of a food web is, what the dynamics are, et cetera.
You have these caterpillars, but they’re just specifically eating this one type of
oat here, like these oat-eating caterpillars are. They might do tremendous damage if a
whole population is just chewing down that oat. But if they’re kind of distributed
amongst a bunch of different tree species in the environment, then each tree species
might not be as affected. The strength of the interaction between the herbivores and
the plants is going to vary depending on the diet presence.
Similarly, caterpillars are food for many other things, including a lot of songbirds
— particularly in the forests around here. As we’ll see for carnivores eating herbivores,
the diet breadth of the herbivores might matter a lot as well.
What I’m going to show you is that if you look at the known diversity of major groups
of organisms on planet earth, you will see that more than half of them are insects. You
also see that this is plants over here, the maroon pie piece. Plants make up a large portion
of the species on earth as well. These groups of organisms that we’re talking about here
that are interacting here are extremely diverse. The only groups that are likely to be more
diverse — when we actually figure out ways of quantifying their diversity — are things
like bacteria and possibly microscopic fungi.
Insects and plants represent most of the diversity on the planet as far as we know now. If you
look at core dates, that’s the group that we belong to. This is the vertebrate animals,
which is only about 3 percent of the species, and so there is not that many species of vertebrate,
but many, many species of insects and plants.
In particular, from what we know now, there are about 265,000 species of plants — probably
a lot more to be discovered, particularly in the tropics. Herbivorous insects we know
that there are at least 400,000 species, but probably a lot more — probably more than
a million. Parasitoids, which again I haven’t defined yet for you, but I will in a minute,
but there are about 60,000 species known — but again, probably many more species.
These are the different groups of organisms that are interacting in the research that
we’ve done in this dataset that we’ve made available to you. They’re important
in terms of their evolutionary diversity and make up a large fraction of life forms on
earth.
So now we’re really going to focus a bit more on these interactions, these particular
groups of organisms. This is a tritrophic food web, and so this just means that there
are these three different tropic levels — the plants, the herbivores and the parasitoids.
Here is what some of them look like, just to give you a sense of what some of the caterpillars
around here look like. I’m not going to really tell you much about them, but we happen
to know quite a bit about each species of caterpillars — at least to have enough of
a sense of what kinds of characteristics they have related to these interactions. They are
pretty interesting. I find them kind of full of surprises. You know, you go out to the
forest and you find a new caterpillar and it looks really weird, or maybe suddenly different
from another one. There is a lot to look at. If you’re the kind of person that sort of
likes to just, you know, see diversity in the world, caterpillars are a pretty good
place to look.
Of course, they all eat plants. You may or may not know that. Just about all caterpillars
in the world eat plants. There are actually a few that don’t; that eat other insects,
which is kind of interesting, but for the most part they eat plants. All the ones in
this dataset eat plants.
Altogether collectively, caterpillars probably represent the most important group of herbivores
in terrestrial ecosystems and so not in aquatic ecosystems, but on land.
Okay, here are the parasitoids that I was telling you about. Parasitoids are extremely
fascinating and also extremely creepy. They are essentially the reality of the monster
in the Alien movie. So if you’ve ever seen these movies, you know, this is a monster
that at one stage of its life lays its eggs inside a person. It’s all fictional, but
the little lizard-like thing hatches and busts out of the person killing it and then it turns
into this giant monster that eats people.
Okay, parasitoids actually do this, except they use insects as their host. There aren't
any parasitoids of people in reality, but if you’re a caterpillar, the world is a
very dangerous and scary place — in large part because of these parasitoids, which are
constantly out there trying to find you; lay eggs in or on you; and then
their
parasitoid larvae will develop inside you for the most part and try to complete their
growth, and then bust out and kill you. This is actually an extremely common interaction
in the world.
All insects that eat plants have at least one species of parasitoids that attack them
and make its living off of them. These are some examples from Connecticut Forest. Up
at the top left, this forest caterpillar actually has some parasitic wasp larvae — these are
actually ecto-parasitoids that just kind of hang out on the outside of the caterpillar
and slurp it slowly.
Then it also happens to have a couple of eggs from a fly like that fly shown up there in
the top middle. This caterpillar actually got attacked by multiple species of parasitoids
and did not make it. This is a fly. It sort of looks like a housefly, but it’s actually
called a tachinid fly. It’s part of a family of flies of about 10,000 species worldwide;
although, there are probably many, many more that we just haven’t described scientifically.
These flies are exclusively parasitoids as far as their life habits go, and many of them
attack caterpillars. These guys, they kind of look like houseflies, but they’re not
interested in much else other than caterpillars and basic life needs like mating and reproducing.
Over here you see a caterpillar that’s dead, and these are parasitoid flies, these tachinid
flies. They’ve kind of eaten it; they’ve come out of it, and then they’ve made these
pupas — these little tick-tack things. Those will turn into flies like that one.
These ones all along the bottom here, these are all caterpillars that have been killed
by parasitic wasps and so here is the cocoon of a parasitic wasp that’s sticking out
of this poor caterpillar. Here is another one. The wasp spun a cocoon and is dangling
down on this little thread, so this is very common. If you actually look around and you
look at enough caterpillars, you’ll eventually come across one that looks sort of like this
— dead from these things.
It turns out that in Connecticut Forest around here at least when we collect our data in
late spring or early summer, about a quarter of all of the caterpillars in the forest die
this gruesome death. This is a major mortality, and so that’s why it’s of interest in
part.
What we’re trying to do is understand these interactions in the context of these forest
ecosystems, which, [Name], who is a member of this class has done research with me and
he took these pictures. We have a bunch of tree species; a bunch of caterpillar species;
a bunch of parasitoid species, and we’re trying to quantify their interactions.
What I’m going to get to now is to tell you a little bit about — not in great detail
— but a little bit about how we collect the data and so forth. Then I’m going to
tell you a little bit about some previous published studies on this same topic, so that
you can get a sense of what kinds of research questions and findings people have already
published.
This study system, these are caterpillars in temperate deciduous forests in Northeastern
North America.
There are a bunch of caterpillars, many caterpillar species that just feed on plants within one
taxonomic family. You may not know much about the taxonomic families of plants, but that’s
okay. If you look at this dataset at some point, you will find that those taxonomic
families are listed in the dataset.
And then there are many general species of caterpillars, of plants from more than one
taxonomic family. These are things that eat oaks, maples and hickory or something like
that.
And then we used eight pretty abundant tree species in this area to sample caterpillars
from, so we had to draw the line somewhere. Logistically, we couldn’t just sample caterpillars
from every tree species out there. We had to say that okay, eight seems like a reasonable
number. It’s going to give us enough kind of breadth to maybe see — maybe it will
be representative of the larger food web, but we have to impose limits just because
we’re the ones — me and my students are the ones — having to do this. We only have
so much time to do it, right? We don’t have a huge army. That’s more or less the scope
of this study.
In case you’re interested, these are the plant species. We call them host plant species
because we’re thinking about it from a caterpillar point of view, but there’s red maple, black
birch, chicory, beech, white oak, red oak, black cherry and witch hazel. If you’re
at all familiar with the forests around here and their natural history, these will be familiar
trees to you.
So the way that we actually collect the data is that we find individual trees. We try to
choose them haphazardly as much as we can. You’ll probably talk more later about kind
of random sampling and various aspects of sampling, but it really can’t be truly random,
because we’d have to basically assign numbers to all the trees in the forest and then randomly
choose those.
We really can’t do that, and so we have to do sort of like random collecting and identifying
individual trees, we sample caterpillars from just a subset of the trees. Typically, we’ll
choose a branch that we can access and we will either — in the past from Youth Data
we either visually surveyed the branch intensively so that we were sure that we got everything,
or we would actually knock the branch with a stick and caterpillars would fall off onto
a sheet underneath.
That’s a method that we’ve been using more recently for various reasons, but either
way, we think we’re getting caterpillars from a branch. This is our sample. What we
do is we collect all the caterpillars that we get from a tree branch and we bring them
back to our laboratory here at Wesleyan, and we continue growing these caterpillars on
the tree species that we found them on, and then they either grow up and turn into a pupa,
and then into a happy moth or a butterfly — or in a lot of cases they don’t make
it that far. They die from various causes, and one of the important causes is from parasitoids.
You’ve got your happy caterpillar there feeding on the plant, and the next day you
check it and it’s dead and there’s a parasitoid larvae. That happens, like I said, about a
quarter of the time, and so we measure the frequency and mortality from these parasitoids.
These are identified initially either as wasps or the tachinid flies that I told you about.
But because parasitoids themselves are very diverse and really people need a specialized
expertise to be able to identify them to the species level, we can’t really do that.
We have to send our specimens on to other people and so that’s going on. In the caterpillar
dataset there are not species identification for the parasitoids. You just say whether
it’s a tachinid fly or a wasp.
Lastly, over the course of this study the data that you have is over five field seasons,
from 2004 to 2008, and three different field sites all in Middlesex County, Connecticut.
One of the big questions in ecology and other fields is how local is an observed pattern.
Some of the kinds of things that you can do with this caterpillar dataset that have been
done before, but in other places — in other ecosystems — and we don’t want to just
say that oh, it’s been done here and therefore it applies to the world at large. We want
to know if patterns that we see somewhere else are also found here. If they are, they’re
general; that’s interesting. If they aren’t, then they’re more regionally specific; that’s
interesting. It’s important to know these things if we want to understand how the world
works.
I’ll first tell you a little bit about some findings from a similar study that I was actually
involved with when I was a graduate student at the University of Arizona. These are Arizona
caterpillar-parasitoid interactions. These are just some examples of what these things
look like. This is the food web that we put together between the caterpillars from Arizona
over here, and then these various parasitoids here. These are all actually tachinid flies
that we were dealing with.
As you can see, this is a pretty complex web of interactions. There are some very specific
interactions, and there is some much more general interactions — but the width of
these lines represent some measure of the strength of the interactions. That is how
abundant they were, et cetera.
There are a lot of things that we did with that data to make sense of it, but one of
the patterns that emerged that was a very important pattern and has been seen elsewhere,
was just this one, that the most abundant caterpillar species had the highest mortality
rates from parasitoid.
You can see this graph of the abundance, which in this case is a transformed measure of the
abundance of the caterpillars. In this case, we looked at the maximum parasitism rate for
each of these caterpillar species, and so each caterpillar species is represented by
a dot — a data point — and then this line that shows you that there is some kind of
a linear relationship there that we could uncover.
So you might want to know that the parasitism of Connecticut caterpillars depend on caterpillar
abundance. We can apply this same finding and say, “Okay, is this a general thing?”
and it’s pretty likely to be.
king at. This happened to be trees called willow and box elder. They then said, “Well,
here is the set of common caterpillars. Between these two tree species, do their parasitism
levels differ depending on the tree species?” Yes, they did.
Another thing about caterpillars is that they have very different appearances, and sometimes
these appearances give you clues as to how they interact with the plants that they’re
eating. Typically, when you see a caterpillar like a Monarch butterfly caterpillar, or something
else that looks like this, you think, ah-hah, this caterpillar is warning me through its
coloration that it is unpalatable; it is not good to eat. Not that you’d want to eat
a caterpillar, but a lot of other things would, like birds.
These things typically indicate chemical defenses, and many of these chemical defenses. Many
of these chemical defenses the caterpillars get from the plants that they eat. So like
a Monarch butterfly caterpillar eating a milkweed plant, we know that these milkweed plants
have heart poisons called cardenolides or cardiac glycosides. These caterpillars eat
those chemicals and they sequester them in high concentrations in their bodies, and this
makes them unpalatable to vertebrates like birds or mammals.
Similarly, caterpillars that are pretty fuzzy and brightly-colored like this are often pretty
well defended against predators like birds and mammals.
So just based on the appearance on the caterpillar species, you might hypothesize some characteristics
of how it interacts with its plants. It’s not something that it’s in the dataset that’s
available to you. The caterpillars are coded for how they look, and this could give you
clues as to what you might expect.
Basically, to tell you something of what you might expect with respect to parasitoids,
it was found in a tropical study in Costa Rica that the Costa Rican caterpillars, the
most chemically-toxic caterpillars experienced the highest mortality from parasitoids. That’s
shown by this bar here. These are caterpillars that are deterrent to generalist predators,
particularly ants.
What you can see in this particular comparison here is that the ones that don’t regurgitate
— that is basically they don’t throw up their gut contents when you mess with them
and they’re deterrent — these really get hammered by parasitoids. Whereas, when they
do regurgitate, that’s actually associated with much lower levels of mortality from parasitoids.
It’s not really clear what’s going on here, but the non-deterrant caterpillars,
in general, tend to have a different pattern.
The ones that don’t regurgitate in contrast with this over here, they have much lower
parasitism mortality, and then the ones that do regurgitate seem to have higher parasitism.
So you start to get these kinds of complicated interactions in the data. You’ll hear more
about what interactions mean later, but these complex patterns emerge.
You might want to know that Connecticut caterpillars with bold coloration or maybe hairy and spiny
exoskeletons, do they have more parasitism than those who don’t have these characteristics?
We can think about that.
Another thing related to this, and I’m kind of going to finish up with this, but the diet
breadth of insect herbivores — something that I was emphasizing a lot at the beginning
— we know that many dietary-specialized caterpillars and other herbivores are often
unpalatable to generalist predators, or they just might be hard to find and are not eaten
as much.
There are a couple of generalist predators out there — the paper wasp. These things
go after caterpillars and they feed them to the hives. They develop larvae in the hive
in the colony, and then here’s a spider that’s captured a grasshopper.
Like I said, the world is a dangerous place for a caterpillar. There are a lot of these
generalist predators — both invertebrates and vertebrates — out there. These studies
basically show this, and this particular study that doesn’t actually, I had data here,
but I guess it’s not going to show up, but it just shows that there are predation trials
with a bunch of caterpillars that make the same point that I just made earlier.
Oh, here we go; we’re pretty much done. Let me just stop here. This just brings up
another research question that you could look at in this data, but I’m going to let those
of you who are really interested in this kind of data start thinking about your own research
questions. I have a handout for you guys as well.