ABSTRACT Many different types of insects make seasonal migrations over vast distances, typically from one broad geographical region to another, most often involving a latitudinal change in one direction in spring with a reversal of this direction in autumn. However, a small handful of these species instead migrate from an enormous geographical area to a highly specific destination they have never previously visited, a journey they make only once. Of these, only two – the diurnal monarch butterfly and the nocturnal bogong moth – are well studied. Even though these lepidopterans have sophisticated multisensory compass mechanisms to guide their long journeys, some studies question whether they are capable of navigating to their goal, or whether they just end up there more or less by chance, pushed by the prevailing winds (a ‘stochastic wind-borne’ transport mechanism). At the other extreme is the possibility that monarch butterflies and bogong moths are ‘true navigators’, with the ability to directly travel to their distant goal by using a ‘compass’ to guide them in their inherited migratory direction and a ‘map’ that continuously updates their current position. In this Review, we will argue that the evidence for stochastic wind-borne transport and true navigation is weak, and that the current weight of evidence supports ‘vector navigation’. We present a hypothesis that individual Lepidoptera use vector navigation to migrate to a distant goal by employing favourable winds and global compass mechanisms to choose their desired flight directions during consecutive journey segments (or vectors), each of different length and direction, and with each vector transition being initiated by innate recognition of local sensory cues. We further hypothesise that this recognition is passed on to coming generations via epigenetic memory.
Warrant et al. (2026) studied this question.