Foodweb Kerplunk
Help

Help is available on several topics:


Launch Game | Top Page


How to Play: The Least You Need to Know
Gameplay is by playing "cards", at the bottom left of the screen. You play a card by clicking on it. What card(s) you can play now are highlighted in purple. The leftmost card you're stuck with--when it's highlighted, you must play it. These are the problems that happen to you during your tenure as councilmember. After playing that, you play your choice of the 3 other cards. Then the next mandatory card is highlighted. We stacked the decks.

The cards do things to the foodweb/ecosystem. There are three display areas to track your progress. The big one shows the foodweb structure--who eats whom. Clicking on its boxes highlights the species name and relationships. The bar chart to the right shows population levels. Below that is game time and score. Your efforts are scored on ecosystem health, money, and popularity.

After you play each card, Nature takes its turn and the foodweb progresses. After a number of cards are played (19 in the intro deck), the game ends, rating your performance against all three possible goals:

  1. Win re-election (moderately good stewardship).
  2. Beat the target high score.
  3. Beat the target low score.

Hints:

Topics
What's the Point?
This is an educational game. The point is to have fun and learn something along the way.

Foodwebs are a bit of a bugbear. They're so complicated--this eats that, which causes a side effect there, etc. Just look at that display, all those interrelationships. Hopeless, huh?

Not really. There is rhyme and reason to how it all hangs together. Life on Earth has managed to hang in there for billions of years, very rarely in real trouble. Most land and water on Earth is easily self-organized into foodwebs in the absence of torture. It's just difficult to express how it all hangs together. So, here's a game to play. In the game, you can play to preserve the foodweb, or play at dismantling it. Either way, you'll see what makes it tick.

Why did we do it? Foodweb Kerplunk is a prototype answer to a question. The question was, How do we reach adults with educational content? And hold their attention, without classroom or tuition or graded assignments? We teach Environmental Studies. We think adults are motivated by our subject--we all care about the quality of our environment, the future of our planet. This game is envisioned as the first piece of an online mini-course, Mortal Stakes: Populations in the 21st Century. We hope to see you again with future installments.

Topics
Help Along the Way
Help is available inside the game via the "?" icons on individual cards and menus. The "?" buttons in the "New Game" menu explain each game setup. The "?" button on the "Intro" game setup for a foodweb explains the ecosystem. (For example, Chaparral Intro Deck, from the "?" button by "Chaparral Intro" on the New Game menu, explains the structure of the Chaparral foodweb.)

Topics
Display & Options
Here is the Foodweb Kerplunk game screen, followed by explanations and options:

Cards
Gameplay is by playing "cards", at the bottom left of the screen. You play a card by clicking on it. What card(s) you can play now are highlighted in purple. The leftmost card you're stuck with--when it's highlighted, you must play it. After playing that, you play your choice of the 3 other cards. Then the next mandatory card is highlighted.

Each card has a title, a short description of what it does, and some icons at the bottom. The three bubble icons indicate the general effect (positive or negative) of the card on the three scores: ecosystem, money, and popularity. The question mark icon brings up a longer description of the card.

The card decks for a game are almost (but not completely) stacked in order. Most cards appear in the same order each time you replay the game. There are some exceptions, where playing one card or some event inserts a card semi-randomly into the remaining deck. For example, if you play a "Wallaby" card, a "Wallaby's Mate" card is inserted later in the same group of cards deck (mandatory or optional cards.) If the Lizards go extinct, a "Stock" card to add Lizards is added later in the deck. Etc.

Topics

Decks
The New Game button brings up a menu of Game Decks to choose from. At this writing, Kerplunk only has one foodweb to explore, the Chaparral, though more may be added sometime. The first five decks are "story" decks--the cards seriously limit your options and have a variety of effects and side effects. If you find this frustrating, or just want to manipulate the foodweb directly, please try the sixth deck, Mechanic.

The deck help files are:

 Topics
Control Buttons
There are six control buttons at the lower right:

Undo: Returns to the game state before the last mandatory card was played.

Restart: Using same deck (game setup), return to the beginning.

New Game: Menu of decks to choose from. Note that each deck has a "?" button next to it. That brings up help describing what that particular deck does. Help on the "Intro" deck describes the foodweb for a series of decks.

Options: Brings up a menu:

Help: Brings up this file.

Quit: Exits game and brings up the Credits.

Topics

Time and Score
Just above the control buttons is the Time and Score display. In the Chaparral Intro deck, game time progresses 3 steps per card played, then 10 steps after the "End Game" card. Each step is a simulator step, nature taking its course and species responding to their latest situation. (Steps per card can be reset via the Options button.)

The score has 3 components: environment, money, and popularity. The environment score is calculated directly from the current state of your foodweb: how much area it has left and which species are still alive. Species are unequally weighted in the score. Money and popularity debits and credits are directly controlled by the cards. In general, money is how you're spending your tax dollars/allowance, and popularity is how people are reacting to the job you're doing as ecosystem steward.

Each game deck has a target hi and lo score to beat. Higher and lower scores are possible. To win re-election in the Chaparral games, the minimum score is set to 65% of the way between lo and hi score targets. There is a Score History button on the Options menu.

The little triplets of bubble/arrow icons on cards and card response menus are rough indicators of what effect the action will have on the score.

Topics

Population Levels
At the right of the screen is a barchart of population levels for each species. Each colored bar is current scaled population. The attached thin uncolored bar shows the population last time you played a card. The vertical lines split the display into quarters.

Clicking on the bar or name for a species highlights the species in the Foodweb display as well, with its relationships. (Try clicking on Coyote.)

The population bars are scaled for esthetics, so that changes are visible for all species. More abundant species (and space) are usually "bigger" than less abundant ones. The difference in scales is vast. If the width of the chart is 100, the grasshoppers are plotted at 30,000:1. (That means full width of the chart would be about 3 million.) The predators and the nearly extinct California Gnatcatcher are scaled at 1:1, with most of the other birds scaled at 27:1. Actual populations can be found from the Foodweb Dump button on the Options menu, if desired.

Topics

Foodweb
This section describes, not the real foodweb (see the Chaparral Foodweb discussion), but how to read the display.

The foodweb display is the largest area on the screen, to the upper left. Each colored box is a species (ex. "Raccoon"), or resource ("Space" or "Seeds"), or clumped species ("Rodents" for mice, squirrels, rabbits, moles, etc.) The foodweb is drawn with the base of the foodweb (space in the rain and sun) at the top of the screen. Clicking on a box highlights its name on the population bar chart and its relationships on the foodweb. You can't do anything to the foodweb display except read it.

How the game's simulator works is left for another section. But briefly, a line between two boxes shows an eaten/eater relationship. All the boxes on one horizontal level eat the previous level. Given a choice, a species' level is set by the highest it eats on a foodchain. So for instance, if you click the word "Roadrunner" on the population plot, lines show what the bird eats and is eaten by. The highest this bird eats on the foodchain is Lizards. It also eats all bug types. So do the Lizards. So Roadrunner is both a predator and a competitor of Lizards. (This is not uncommon.) Roadrunner is also eaten by all the mesopredators, the group below it on the display. In reality, Coyotes probably also eat Roadrunner, but this foodweb is simplified, and it's only a part-time Coyote.

There are seven levels on the Chaparral foodweb, meaning the longest foodchain is seven steps long. Coyote eats a mesopredator, which eats Roadrunner, which eats Lizards, which eats bugs, which eat plants, which "eat" space (space, sun, rain). All energy conversions are inefficient, so food value is lost along the chain. (The Coyote probably couldn't survive eating only mesopredators.) Competitors have to share what's available.

Consider the plight of the endangered California Gnatcatcher. It isn't obvious it's endangered. Beebee guns, perhaps? It eats only Otherbug (gnats), in this simplified web. There's no particular shortage of gnats. Clicking on Otherbug, we find the Gnatcatcher's competitors are Lizards and Roadrunner. For the most part, their predators are the same. Why should there be more Lizards than Gnatcatchers? Lizards even have an extra predator--the Roadrunner.

The problem here is that Lizards and Roadrunners eat all bugs, not just Otherbug. Otherbug is impossible to wipe out, but sometimes it does run short. When it does, there are many more Lizards and Roadrunners than Gnatcatchers eating at the Otherbug trough, and the Gnatcatcher's share (because it is less numerous than the others) is forced inexorably downward.

Any card that depresses bugs (such as Drought) gives the bug-eaters a double whammy: their food supply is depressed, and their predators rely more heavily on eating them than bugs. Similarly for depressing the rodent population: the mesopredators eat bugs, birds, lizards, and rodents. If tasty rodents are depressed, they eat more birds and lizards.

The species innate rate of growth, R--determined by how fast they can reproduce--determines what happens between competitors when there's a food surplus. Here, the Roadrunner and Gnatcatcher are set to the same R. It's about 1/3 the R of Lizards. You'll find Roadrunner is almost as hard to keep alive as the Gnatcatcher, at least in this 250 acre fragment of suburban chaparral. (R is not shown anywhere on the display, but you can do a Foodweb Dump from the Options menu if you want to look.) So the Gnatcatcher loses when there's a shortage, and loses when there's a surplus.

Extinct species are grayed out on the foodweb.

Topics
Soundtrack
The soundtrack is explained in the Chaparral Intro Deck help file.

Topics
The Simulator: How It Works
Foodweb Kerplunk is a game user interface built on an ecological simulator. The simulator was written for the game. It's based on one of the simplest ecological models, called the Discrete Logistic Growth Equation:
Symbol Means
Nt population at time t
Nt-1 population at time t-1
R population growth rate
K carrying capacity
mostly determined by food supply

In English, this equation says that for a given habitat, there exists a carrying capacity, K, which is the right population level for this species. If the population rises above K, it will drop back down. If lower, it will grow. How fast the population grows or falls depends on how close it already is to K, and R, its growth rate. The population grows at its fastest, R, only when it is far below K. Near K, the population neither grows nor shrinks. (If R is high enough, the population oscillates, but we don't set R that high.)

R values used in Chaparral range from 0.15 for Scrub and Raccoon, to 1.7 for Otherbug. There are some R=0's as well, but more on that below. The ecologically savvy will notice that we don't really know what values to use for R. It isn't readily measured for one species, let alone a whole foodweb's worth of species.

So, that's how a population grows, "all else being equal." But K, the carrying capacity, is not a constant here. It's set by what's available to eat. The crux of this simulator is repeatedly finding K for each species.

In each timestep, the simulator does:

  1. Grow each population based on its last K and N (apply discrete logistic growth).
  2. Inflict damage to the eaten populations based on current eaters (decrease N).
  3. Allocate K between competitors based on remaining N.
  4. The N and K at this point are the ones reported on barchart/dump.
It takes a fair bit of gymnastics to determine which species get what damage and share of K, when the competitions are asymmetric. For this discussion, I'll just say all the populations are converted into biomass; nobody is allowed to prefer one food over another; biomass is scaled by food value and discounted by the inefficiencies of eating it; and at some point, the biomass turns back into a population headcount. Which basically means that whatever there's more of, gets eaten most.

There are a few more parameters used in the simulator than appear in the discrete logistic growth equation. Specifically:

Parameter Means
efficiency How useful this species' biomass is as food. It can be eaten at this% efficiency. Birds and mammals and lizards in the Chaparral are set to 85% efficiency, bugs to 50%, scrub to 1%, and grass and herbs to 1.5%.
scale The relative size of this creature. Used to turn biomass into headcount. Examples: the birds are 1.0, Otherbug 1000.0, Scrub 0.1, Coyote 0.0005 (because it's only a part-time resident.) This is expressed like a density (creatures per square meter), but it's describing the body size that's eating.
damage Usually 1.0 or close to it. After a species' eating biomass is determined, this is used to scale it back for some reason of restraint. Cat is restrained to doing .3 damage because people feed it. Cicada does .5 damage because they're not that troublesome to the tree roots they feed on.
dominant The idea here is that unequal competitors get disproportionate shares of resources. Non-dominant species take leftovers. The simulator uses this to set K for competing plants. The dominant plants share all available space K, but with damage < 1. So there is K left over for nondominant plants. In the Chaparral, Scrub and Grass are dominants, with damage rates of 0.2 and 0.6 respectively. Herbs grow around them, but cannot displace them. Dominance could be an issue with other species, but for the Chaparral, it's only used for the plants.

Two species in the Chaparral (Cat and Coyote) are "sponsored". This means that although they eat here, their numbers are independent of the K-setting dynamics, and their R is unused. For Cats, this is because their food and reproduction are more-or-less controlled by humans. For Coyotes, the problem is range. Typical Coyote ranges are huge compared to any suburban land fragment.

Seeds should be made by the plants, that submodel wasn't implemented here. Seeds are treated more-or-less like another kind of Space. They are increased or decreased by the cards, not by how the plants are doing. Seeds are kept separate because eating seeds is not equivalent to eating plants--it does the plants no harm.

Although the Chaparral isn't a bad model, I won't defend it as science. I have very little basis for setting the parameters. The reality check used to parameterize the foodweb was a comment in the paper describing this system (Crooks & Soule, 1999) giving proportions and implied numbers of the lizards, birds, and rodents the cats drag home. The paper also indicated the cat and mesopredator densities, the rare-visitor role of the coyote, and listed the bird species. The Chaparral model gets the numbers and proportions pretty close to what the paper implied. The paper didn't provide all the detail needed to structure this foodweb. That was fleshed out by surfing the net to find out what species to put in a chaparral, and what they generally eat. Then I pared my cartoon down to what seemed a minimum, omitting details like roadrunners eating snakes, etc., but leaving all the bird species because they formed the core of the story. What a gnatcatcher eats was supplied only by its name.

Topics | Credits