Figure 3.

Computer‐based detection experiment. We used a citizen science game (a), based at the Natural History Museum in London, UK, where subjects searched for hidden crabs on a touch screen and detection times were measured. People were instructed to find crabs as quickly as possible from varied background types: mudflats (b), rock pools (c) and mussel beds (d). In citizen science experiment, crabs picked from mudflats, mussel beds and rock pools were presented against their own and other habitat types on touch screen. The bar plots illustrate which crabs are hardest to find (detection time, [e], in seconds to spot the crab from a background) and thus have the highest survival benefit hiding in three major tidal habitats (finding success, [f], as the proportion of successful clicks of particular crab type presented against different backgrounds). Receptor noise‐limited human vision model predicts that chormatic contrasts of all crabs were reasonably hard (i.e., <5 JNDs) to detect in the game (g), whereas luminance differences were larger and rendered some, except “mudflat crabs”, easier to find (h). Error bars show ± 2 SE