Skip to the lab

EcoGEM Lab

Ecosystem

Atlantic Ocean

Food web

Population

Step 0 / 1,000
Graph options

Run advances to the next 100-step mark. Step advances once.

About this lab

Atlantic Ocean · Northwest Atlantic shelf

  1. Choose an ecosystem. Each has its own species and selected feeding relationships.
  2. Click a species. Turn it on or off, or change its diet. Arrows point from food to consumer.
  3. Run an experiment. Run advances to the next hundred-step mark; Step advances once. Changes pause the model so you can make a prediction.
  4. Follow the graph. Hover or use the slider to inspect a step. The data table shows the current step’s exact values.
  5. Compare. Keep a run, change one variable, then reset and run again. Faint dashed lines show the kept run from this ecosystem.

Reading the screen

Tap a species portrait to include it or choose its food. Highlighted choices are active. Arrows point from food to consumer. The scene illustrates abundance; organisms are not to scale.

Relative abundance · model units

How I made this model

I built this lab with Codex, OpenAI’s coding assistant, inspired by the original Annenberg Ecology Lab. Feeding relationships draw on published research. Codex selected the equations and numerical settings using established modelling ideas: resource-limited growth, competition, and feeding that levels off as food becomes abundant. These settings were not fitted to measurements of these species. The results illustrate possible patterns within this model; they do not predict real ecosystems.

What the model means

These are simplified webs, not complete accounts of each region. Species and diets vary by place, season, and life stage. Plant abundance represents food resources such as leaves, fruit, and seeds. Changes in the artwork illustrate abundance; depth and motion are decorative, not simulated animal behaviour. Use Motion to freeze the scene. Reduced-motion preferences are respected; motion stops offscreen, in background tabs, and while you inspect an earlier step.

Resource controls alter the producers’ carrying capacity. “Rainfall” and “nutrient supply” are teaching controls, not calibrated climate or chemical measurements. A model step is not a real day.

Experiments save separately for each ecosystem in this browser. Changing a preset starts a new run; your kept comparison stays. CSV and experiment downloads let you keep a separate record. Experiment data stays in your browser. The website counts page views without visitor cookies; it does not collect your experiment data. Use Teacher controls in the Menu to choose Guided or Independent mode. Guided mode adds optional pattern descriptions; both modes use the same equations. For guided questions, open the full EcoGEM learning experience, or read the teacher guide.

Ready

Original inspiration

Inspired by Ginger Booth’s Ecology Lab (2015).

Read the story, assumptions, sources, and APA credits

Teacher controls

Choose how much support to reveal. Changing support keeps the current experiment and its equations.

Guided adds a Patterns button and shades the graph interval it describes. Independent leaves interpretation to the learner.

Hints to include

Settings stay in this browser. For a class link, open Guided or Independent. The link sets the mode; hint choices remain local.

Patterns

What relationships might explain these changes? Increasing and decreasing describe net change, not a steady trend. Zero is not a stability judgement.

How to read these hints

These are descriptions of up to 50 recent steps, shortened after an intervention. Trend descriptions need at least 20 steps. “Changing little” means a range within 5% of the average. “Settling” means the later half varies less than half as much as the earlier half and spans less than 10% of the average. “Fluctuating” requires several reversals in five-step averages, not proof of a cycle. These thresholds are teaching choices, not validated measures of ecological stability. Zero abundance and excluded species are never classified as settled.

Abundance at step 0

SpeciesThis runKept run

Relative abundance in model units, not counts of real organisms.

Start a fresh run?