A review of the sentinel prey method as a way of quantifying invertebrate predation under field conditions

Abstract Sentinel prey can provide a direct, quantitative measure of predation under field conditions. Live sentinel prey provides more realistic data but rarely allows the partitioning of the total predation pressure; artificial prey is less natural but traces left by different predators are identifiable, making it suitable for comparative studies. We reviewed the available evidence of the use of both types of invertebrate sentinel prey. Fifty‐seven papers used real prey, usually measuring predation on a focal (often pest) species, with studies overwhelmingly from North America. The median predation was 25.8% d−1. Artificial sentinel prey (45 papers) were used in both temperate and tropical areas, placed more above ground than at ground level. The most commonly used artificial prey imitated a caterpillar. Up to 14 predator groups were identified, registering a median of 8.8% d−1 predation; half the studies reported only bird predation. Predation on real prey was higher than on artificial ones, but invertebrate predation was not higher than vertebrate predation. Invertertebrate but not vertebrate predation was negatively related to prey size. Predation near the Equator was not higher than at higher latitudes, nor in cultivated than noncultivated habitats. The use of sentinel prey is not yet standardised in terms of prey size, arrangement, exposure period or data reporting. Due to the simplicity and ease of use of the method, such standardisation may increase the usefulness of comparative studies, contributing to the understanding of the importance and level of predation in various habitats worldwide.