The Radar Decline Is Behaviour

01The evidence

• Radar monitoring detected fewer insects above warming landscapes, but also strong year-to-year variation linked to climatic conditions, which is consistent with a behavioural as well as an abundance signal.NSF radar monitoring

02The tension

Radar is treated as an unbiased census of insect biomass because it samples a large volume. But it samples only the individuals that are flying, at the heights and times the radar observes, and flight behaviour is strongly weather-dependent. That confound is acknowledged and rarely quantified.

03The theory

A radar observes a flux: insects crossing a volume per unit time. The flux is the product of density and flight activity. If warming reduces flight activity, for instance by making days too hot or winds too strong, the flux falls even at constant density.

If behaviour contributes, then radar trends should diverge from ground surveys, and the divergence should correlate with weather variables rather than with abundance. That is a testable prediction using existing paired datasets.

The theory does not deny decline. It partitions the signal, and it predicts that the radar decline should be steeper than the ground decline in regions where flight conditions have worsened.

The applied consequence is that trend estimates from radar should be corrected using a flight-activity model, and that monitoring programmes should pair radar with ground sampling so the partition can be estimated.

04What this is not

05Predictions

• P1Radar and ground-survey trends should diverge systematically.
• P2The divergence should correlate with weather variables rather than with abundance.
• P3Correcting for flight activity should reduce the estimated decline in regions where flight conditions worsened.

06What would refute it

Radar and ground trends tracking perfectly across regions would leave the behavioural confound negligible.

07What it would change

It would change how radar trends are reported, and force paired monitoring designs.