High detectability costs select weak warning signals in chemically defended gregarious prey

The evolution of costly chemical defenses in cryptic prey is puzzling, as conspicuousness should be adaptive for defended prey by enhancing predator avoidance learning. Here, we explore three hypotheses that may promote weak visual signal strategies: (1) Low prey conspicuousness is favored in diverse predator communities; (2) Conspicuousness is less critical if color patterns or behavior allow defended prey to be distinguished from cryptic prey; and (3) in gregarious prey species, aggregation itself could act as a deterrent signal. We used chemically defended and gregarious European pine sawfly (Neodiprion sertifer) larvae, exhibiting green-black-grey coloration, as our study system. First, we manipulated the conspicuousness of the larvae by placing them either solitarily or in groups and marking them with either cryptic or conspicuous coloration, then testing their survival against predators in the field. Second, we analyzed how conspicuous the ventral side (which larvae reveal during ..., , # GENERAL INFORMATION

## 1. Paper Citation

Koskenpato K, Mikkonen T, Mappes J, Valkonen J & Lindstedt C. 2026. High detectability costs select weak warning signals in chemically defended gregarious prey. *Behavioral Ecology.*

## 2. Brief abstract

Why do some chemically defended prey use weak visual signals when conspicuous aposematic coloration seems more effective for predator avoidance? Using weakly conspicuous pine sawfly larvae as a study system, we found that increased conspicuousness under predation risk from natural predator communities reduces their survival. Our findings highlight the importance of understanding ecological mechanisms behind the evolution of weak visual signal strategies.

## 3. Originators

Katja Koskenpato

## 4. Contact information

Katja Koskenpato, Faculty of Agriculture and Forestry, Department of Forest Sciences, University of Helsinki, Helsinki, Finland, [[e-mail]](mailto:[e-mail])

## 5. Year of data collectio...,