Checked against current official sources: 12 September 2026.
A female Aedes mosquito is looking for water.
Not because she is thirsty.
Because she is looking for somewhere to lay eggs.
That search is a biological behaviour.
Singapore turns it into a measurement system.
The Gravitrap works because surveillance becomes easier when the animal you are trying to count is persuaded to come to the measuring instrument by its own reproductive behaviour.
Quick Read
Singapore works partly because dengue control does not depend only on waiting for people to fall sick. It also watches the vector before the disease signal arrives.
NEA says it has deployed about 72,000 Gravitraps at public and landed housing estates across Singapore. Developed by NEA’s Environmental Health Institute, the trap is designed to attract and capture female Aedes adult mosquitoes that are searching for sites to lay their eggs. It also prevents mosquitoes from emerging from eggs laid in the trap.
When Gravitraps are distributed across an area, repeated captures become a picture of mosquito population conditions in that neighbourhood and its vicinity. NEA updates information on areas with relatively higher Aedes aegypti populations monthly and describes this information as a lead indicator that helps partners and stakeholders take proactive preventive action.
The deeper mechanism is:
female Aedes seeks an egg-laying site → Gravitrap presents an attractive site → mosquito enters and is captured → eggs laid in the trap do not produce a new escaping generation → field teams inspect and identify captures → capture counts are aggregated across place and time → higher local mosquito population becomes visible before human case counts alone would reveal it → Town Councils and other stakeholders can prioritise prevention and control work where surveillance says mosquito pressure is higher.
This article does not claim that a Gravitrap alone controls dengue, that every captured mosquito carries dengue virus, or that low Gravitrap counts mean zero risk. It isolates one mechanism: use mosquito reproductive behaviour to create a distributed, repeatable sensor for the vector population.
Wait, What? The Mosquito Chooses the Sensor?
In a sense, yes.
Most environmental monitoring requires people to go out and find the thing being measured.
A temperature sensor waits for heat.
A rain gauge waits for rain.
A Gravitrap waits for a mosquito that is actively searching.
That turns behaviour into sampling power.
the mosquito is not merely detected; it helps deliver itself to the detector.
Why Female Aedes Mosquitoes Matter
Female mosquitoes are the relevant surveillance receiver because females seek blood meals and egg-laying sites.
The trap therefore targets a behaviour closely connected to the mosquito’s reproductive cycle.
NEA notes that one female mosquito can lay hundreds of eggs in its short lifespan.
That means one successful adult female is not merely one insect.
She is potential future population.
The Trap Does Two Jobs at Once
The first job is surveillance.
Captured adults tell field teams that gravid female mosquitoes are active nearby.
The second job is local removal.
The captured female is prevented from biting other people, and mosquitoes are prevented from emerging from eggs deposited in the trap.
One physical object therefore combines:
- measurement;
- capture;
- local interruption of reproduction.
But the strategic value comes from the network of traps, not one trap standing alone.
One Trap Is an Observation; 72,000 Traps Become a Surveillance Network
A single trap can tell you something about one micro-location.
A distributed network can reveal spatial patterns.
Which estate is rising?
Which cluster remains low?
Where did counts change after intervention?
Where should field attention intensify?
This is why scale matters.
the sensor becomes more useful when many identical questions are asked across space with enough consistency to compare the answers.
The Gravitrap and Human Dengue Cases Own Different Clocks
A dengue case is important evidence.
It is also downstream evidence.
For a person to become a reported case, several events must already have occurred:
- infected mosquito exists;
- mosquito bites a susceptible person;
- infection develops;
- symptoms may appear;
- person seeks care;
- case enters surveillance.
Gravitrap data observes the vector population earlier in that causal chain.
That is why NEA calls it a lead indicator.
Lead Indicators Buy Decision Time
A lagging indicator says:
the outcome has already happened.
A lead indicator says:
conditions associated with future risk are becoming stronger.
The Gravitrap does not predict who will get dengue.
It provides earlier evidence that the local vector population may be high enough to justify stronger preventive attention.
The Gravitrap Is Not a Mosquito Repellent
Repellent tries to keep mosquitoes away from a person or surface.
The Gravitrap does the opposite.
It attracts the target insect to a controlled location.
Repellent:
change mosquito behaviour away from the receiver.
Gravitrap:
use mosquito behaviour to bring the mosquito into a surveillance-and-capture device.
The Gravitrap Is Not Project Wolbachia
Project Wolbachia works through a population-suppression strategy involving male mosquitoes carrying Wolbachia.
The Gravitrap is a surveillance-and-capture device.
Wolbachia intervention:
alter reproductive outcomes across the mosquito population.
Gravitrap:
sample gravid females and reveal where population pressure is higher.
They can coexist because one measures and locally removes while another changes population dynamics at a different scale.
Sampling Has a Geography Problem
Mosquitoes are not spread uniformly.
One block can differ from another.
One side of an estate can differ from the other.
Vegetation, containers, building form, human activity and local breeding opportunities matter.
A surveillance network therefore has to sample enough places that local concentrations become visible without pretending every square metre is being observed continuously.
The trap count is a sample of the population, not the population itself.
Sampling Has a Time Problem Too
Mosquito abundance changes.
Rain changes breeding opportunities.
Temperature affects life cycles.
Control work changes local populations.
A one-time survey can become stale.
Repeated traps make trend detection possible.
surveillance becomes powerful when the same measurement returns often enough to distinguish a persistent change from one unusual day.
Identification Turns Capture into Species-Specific Intelligence
Not every mosquito creates the same dengue risk.
The surveillance value increases when field captures are identified correctly.
A trap therefore begins a data pipeline:
capture → retrieval → identification → count → place-and-time record → population map → operational decision.
The physical trap is only the first layer.
The intelligence comes from turning biological specimens into structured observations.
The Map Is Not the Mosquito Population
NEA publishes areas with relatively higher Aedes aegypti mosquito populations.
The colour on a map is a representation produced from surveillance data.
It is useful because it compresses thousands of trap observations into something residents and stakeholders can read quickly.
But the map has limits.
NEA explicitly warns that low risk does not mean no risk.
One female mosquito can still reproduce quickly.
The map should change attention, not create complacency.
The Receiver Is Not Only NEA
NEA says stakeholders including Town Councils can use the data to prioritise dengue prevention and control measures.
That makes the surveillance receiver distributed too.
- NEA receives national and local population evidence.
- Town Councils receive prioritisation signals.
- estate managers receive a reason to inspect specific environments more intensively.
- residents receive a warning that vigilance should increase.
The trap becomes useful only when its data changes behaviour somewhere else.
The Bottleneck Is the Return from Data to Ground Action
A surveillance system can collect perfect counts.
The area can still remain risky if nobody acts on the signal.
The important chain is not:
trap → database.
It is:
trap → data → interpretation → prioritisation → source reduction and control work → new trap data → comparison.
The World Return is the next measurement after the intervention.
Competing Explanation: Why Not Just Inspect Every Drain and Pot Constantly?
Source inspection remains essential.
But a city contains enormous numbers of potential breeding sites.
Surveillance helps allocate finite inspection effort.
If the vector population is rising in one locality, that evidence can justify stronger attention there.
The trap does not replace inspection.
It helps decide where inspection is most urgent.
Competing Explanation: Why Not Wait for Dengue Cases?
Because by then transmission has already occurred.
A vector-population signal can appear before the downstream clinical outcome.
Dengue cases remain crucial.
Gravitrap surveillance adds an earlier lens.
Model Limit: Trap Counts Are Not Perfect Population Counts
A trap measures mosquitoes that encounter and respond to the trap.
Weather, competing egg-laying sites, local habitat and trap condition can influence captures.
The result is a surveillance index, not a census of every mosquito in the estate.
This is why trends, spatial comparisons and repeated sampling matter more than treating one trap count as absolute truth.
Model Limit: Low Population Does Not Mean Zero Transmission Risk
Dengue transmission depends on more than mosquito abundance.
Virus circulation, human immunity, movement, biting patterns and local conditions matter.
NEA therefore cautions residents to remain vigilant even when an area is not highlighted as having high Aedes aegypti population.
What Breaks First?
- A trap is damaged, removed or inaccessible for servicing.
- Trap placement stops representing the surrounding area well.
- Field retrieval is delayed and observations lose temporal resolution.
- Mosquito identification is inconsistent.
- Data are collected but not converted into operational priority.
- Stakeholders treat a low reading as permission to stop prevention work.
- A local breeding source grows between surveillance cycles.
- Intervention occurs but no later measurement checks whether mosquito pressure actually fell.
The useful audit question is:
if Aedes population began rising in this estate now, would the distributed traps detect the change early enough, would the data be identified and mapped correctly, and would somebody with authority turn that signal into faster source reduction and control before human cases became the only evidence?
Primary-School Lens: The Mosquito Walks into the Measurement
Ask a child to imagine trying to count butterflies in a garden by chasing every butterfly.
Then imagine placing a safe observation station that attracts butterflies to one known place.
The child learns that behaviour can make measurement easier.
Secondary-School Lens: Sampling and Population Estimation
Place ten hypothetical traps across a map.
Give each trap a weekly count.
Ask students which neighbourhood appears to have higher mosquito pressure and why one week should not automatically define the long-term state.
The lesson connects ecology, sampling and time-series reasoning.
JC Lens: Leading Indicators and Decision Thresholds
At JC level, treat Gravitrap counts as a noisy leading indicator.
The decision problem is not simply to estimate population.
It is to decide when evidence is strong enough to justify intervention.
The engineering and public-health question becomes:
how should trap density, servicing frequency, species identification, spatial aggregation and action thresholds be designed so the surveillance system catches meaningful population growth early without overwhelming field teams with noise?
Thought Experiment: Perfect Trap, No Map
Every captured mosquito is counted accurately.
The counts are stored without location.
Measurement succeeds.
Operational geography fails.
Thought Experiment: Great Map, No Action
The highest-risk area is obvious.
No inspection or source reduction changes.
Intelligence succeeds.
Control fails.
Thought Experiment: Low Count, One Breeding Container
The neighbourhood map looks reassuring.
One neglected container produces a local burst of mosquitoes.
Population summary succeeds at its scale.
Local uncertainty remains.
Why Singapore Works Does Not Mean 72,000 Traps Eliminate Dengue
Mosquitoes reproduce quickly.
Breeding sites can appear anywhere water accumulates.
Virus transmission depends on human and mosquito movement.
Source reduction remains essential.
The serious claim is narrower:
NEA’s Gravitrap network converts the egg-laying behaviour of female Aedes mosquitoes into repeated local population observations, giving Singapore an earlier and more spatially specific signal for prioritising dengue prevention than waiting for human cases alone.
The Fifteen-Question Gravitrap Test
- Target: Which mosquito behaviour is the trap exploiting?
- Placement: Does the site represent the surrounding estate reasonably?
- Condition: Is the trap intact and operating as intended?
- Capture: Are gravid female Aedes mosquitoes being trapped rather than simply visiting?
- Emergence: Does the trap prevent mosquitoes from emerging from eggs laid inside?
- Servicing: Is retrieval frequent and consistent enough for useful trend data?
- Identification: Are mosquito species identified reliably?
- Location: Is every observation tied to the correct place?
- Time: Is every observation tied to the correct surveillance period?
- Trend: Is the system detecting change rather than reacting to one isolated count?
- Mapping: Are local differences made legible to decision-makers?
- Receiver: Do Town Councils and other stakeholders receive the signal?
- Action: Does higher population trigger stronger prevention or control?
- Public interpretation: Are residents reminded that low risk is not zero risk?
- World return: Do later trap counts show whether field intervention actually reduced mosquito pressure?
Frequently Asked Questions
How many Gravitraps has NEA deployed?
NEA’s current Aedes surveillance page says about 72,000 Gravitraps are deployed at public and landed housing estates around Singapore.
What do Gravitraps attract?
They are designed to attract female Aedes adult mosquitoes looking for sites to lay their eggs.
Do the traps only measure mosquitoes?
No. NEA states that captured female mosquitoes are prevented from biting other people, and the trap also prevents the emergence of mosquitoes from eggs laid within it.
Why does NEA publish mosquito population information?
NEA describes the information as a lead indicator that helps stakeholders take proactive measures in areas with relatively higher Aedes aegypti populations.
Does a low mosquito population area mean no dengue risk?
No. NEA explicitly warns that low risk does not mean no risk and that mosquito populations can expand quickly.
What is the main student lesson?
Good measurement can use the behaviour of the thing being measured. A biological instinct can become part of a city-scale sensing network.
Sources and Further Reading
- National Environment Agency — Surveillance of the Aedes aegypti mosquito population with Gravitraps.
- National Environment Agency — Dengue and Zika information.
Final Thought: The Mosquito Thinks It Found a Nursery
She is following an instinct older than cities.
Find water.
Lay eggs.
Continue the next generation.
The city does not need to persuade her to cooperate.
It needs only to understand the behaviour well enough to place a measuring device in its path.
That is why Singapore works, in another quiet way:
the city understands that sometimes the strongest sensor is not the one that searches hardest, but the one that knows what the thing being measured will come looking for next.