Additivity
The losses add up. Here, 20 + 30 deaths would leave 50 survivors. This is the simple-addition reference.
Interactive ecology
Run a fictional heatwave and fungicide experiment, explore synergy and antagonism, and follow what happens after exposure ends.
01 · Expected and observed effects
In freshwater, a heatwave is a period of unusually warm water. A fungicide is a chemical used to control fungi. Explore what could happen when these pressures occur together or in different orders.
Drag the fungicide, heatwave or both into the tank. You can also select an icon and then + to add it. Press Escape to cancel a selection.
100 animals, ready to start
Use the slider to predict how many of the 100 animals survive.
All numbers are invented for learning, not findings or predictions for a real species.
Each group starts with 100 animals. Survival is measured after seven days. An empty setup runs the unexposed control. Adding both icons exposes the group to both pressures for the same seven days. Heatwave conditions and fungicide concentration are held the same in the single and combined treatments. The icons do not adjust dose or temperature.
The single-exposure results alone cannot determine the combined result. Matching the count is a good guess for this example, not evidence that someone could predict a real experiment.
02 · Making sense of the result
In the seven-day example, the heatwave alone kills 20 of 100 animals. The fungicide alone kills 30. Should exposing animals to both kill 50? That is one expectation, but it is not the only one.
The losses add up. Here, 20 + 30 deaths would leave 50 survivors. This is the simple-addition reference.
The combined harm is greater than the chosen reference predicts. For survival, that means fewer animals remain alive.
The combined harm is smaller than the chosen reference predicts. There can still be substantial harm.
100 × 0.80 × 0.70 = 56
Multiply the two survival proportions. This assumes independent effects on survival and allows for overlapping losses. It uses the example’s 100% survival in the control.
Different processes can produce the same count. In a real study, choose a reference that fits the biological question and explain why. Why the reference matters ↗
Worked example, seven-day exposure
16 fewer animals survive than expected, so there are 16 more deaths.
60 animals died. The comparison uses survival at the same time point.
Run both stressors together above to connect this panel to your result. Until then, these are worked examples.
Before you move on
The two single-exposure results have not changed.
Warmer water can change how quickly a chemical enters and leaves an animal. Coping with heat and dealing with chemical damage can also draw on limited energy. If one exposure leaves less capacity to cope with the other, the combined effect could be stronger. These are possibilities to test, not explanations established by our fictional counts.
Temperature and chemical uptake ↗A published model of combined stress ↗The animals most vulnerable to one stressor may also be vulnerable to the other. If these vulnerabilities overlap more than the reference assumes, the combined harm can be smaller than expected. Other combinations may alter exposure or trigger protective responses. Antagonism therefore does not necessarily mean the stressors cancel each other out.
Susceptibility and reference models ↗A mayfly experiment with changing interactions ↗A label describes a comparison, not a cause. A real study needs independent replicate tanks and uncertainty estimates. To find out why a result occurred, we would also need measurements such as chemical concentrations inside animals, feeding or energy reserves. An interaction in survival need not be the same for growth, or remain the same after recovery.
What happens after exposure ends?Explore recovery and feedbacks03 · Lasting effects and feedbacks
The water can return to its usual temperature before an animal recovers. But a lingering effect and a feedback are different things. What would we need to measure to tell them apart?
Imagine that previously exposed snails still eat less than unexposed snails kept in the same cool water. Their earlier experience is affecting them now. That is a carry-over effect, but it does not tell us whether a feedback is involved.
The heatwave changes the animal's condition.
The heatwave has ended. Its effects may remain.
Depleted energy reserves or other changes may persist, even without a return loop.
Lasting effects do not prove a feedback. In the experiment above, following animals after exposure helps reveal lasting effects. It does not, by itself, reveal a feedback mechanism.
Negative feedback counteracts a change. More grazing can leave less food, which limits further intake. That is the opposing effect shown here.
Positive feedback reinforces a change. For example, more nitrogen-fixing plants can enrich the soil with nitrogen, supporting more plant growth and further enrichment. “Positive” and “negative” describe direction, not whether an outcome is good or bad. Delays and other loops can change the eventual outcome.
Read the published explanation of ecological feedbacksIn an experiment with marine snails, carbon and nutrient enrichment increased algal production and changed food quality. The snails increased consumption while maintaining growth, helping to buffer the extra algal production. This links an individual response to a wider ecological effect; it is a separate study, not a measurement of the freshwater loops drawn here. Ghedini & Connell (2016)
Feeding links an animal's internal state with the world around it: a meal can replenish energy reserves while reducing the food available for the next meal. When closed loops at different biological levels influence one another, they can be described as nested feedbacks. A diagram makes the possible connections visible, but each return pathway still needs evidence. The levels do not have fixed speeds, and a lasting effect is not, by itself, evidence of feedback. These are published background ideas and hypothetical examples, not the model or findings from our developing work.
04 · Further reading
The activities use invented outcomes and simplified causal diagrams. These published papers provide the background and examples of how the questions can be tested.
In our water-louse experiments, we tested different routes of fungicide exposure, combined fungicide with predator risk, and measured recovery after chemical and heat exposure.
Read about the predator experimentReference models make different assumptions about the expected combined effect. These papers explain how the choice affects our interpretation of stressor interactions.
Carry-over effects connect earlier conditions to later performance. Feedback research goes further by examining how a response changes something that acts back on the organism or population.