Millions of insects migrate across the Pyrenees: heavy transit and conflicting ecological roles

Although far less well-known to the general public than the migration of vertebrates such as birds, whales and fish, insect migration is also a frequent phenomenon in nature [1]. Despite the limitations imposed by their small size and the inherent difficulties in detecting insects in motion, the study of insect migration has undergone a true revolution as new techniques (e.g. entomological radar, modelling of the movement of air masses, isotopic analyses [2]) have now made it possible to quantify with precision the movements of insects and to trace their trajectories.

The most remarkable recent contributions to the study of insect migration concern movements taking place several hundreds of metres above ground level that are aided by tail winds (windborne displacement). Various studies have shown that insect migration is a common phenomenon involving the passage of millions of organisms and tons of nutrients (the so-called ‘bioflows’) between regions separated by hundreds of kilometres [3]. Long-distance movements are not performed by a random sample of species but, rather, are mainly performed by generalist and common species, capable of exploiting a great variety of resources, that are sometimes present in enormous abundances. This fact turns them into potential agricultural pests in many cases, so there is an obvious practical interest in detecting and understanding their movements [4].

Despite the importance of windborne movements and the use of radar in the study of insect migration, ground-based observations are key and will continue to be so in any attempt at fully understanding this phenomenon. The seminal work by Williams, one of the pioneers in the study of insect migration, was based on the collation of countless records by entomologists and amateur naturalists around the planet, mainly of butterflies migrating near ground level [5]. Although obtained in a much more anecdotal way, the observations made on 13 October 1950 by the famous evolutionary biologist David Lack and his wife, Elizabeth, at the Gavarnie or Bujaruelo Pass on the French–Spanish border in the Pyrenees highlighted the importance of mountain passes as channels for insect migration [6]. They observed a continuous passage through this high pass at 2273 m of large numbers of butterflies, dragonflies and other insects alongside many different species of migrating birds. These observations unequivocally indicated that insects, like birds, moved from the northern regions where they had been born to more southern wintering areas as the cold season approached. Subsequently, other entomologists have added observations of similar movements during the same period from other Pyrenean passes and the Alps, which together provide a more complete view of this phenomenon while indicating that the community of migratory insects is more extensive than first suggested by these observations [7].

All these studies show that an important part of insect migration takes place within the flight boundary layer or FBL (i.e. the lowermost layer of the atmosphere within which insects' self-powered flight speed exceeds the wind speed thereby allowing them to control their migratory direction). These movements are common, for example, in day-flying insects such as butterflies and dragonflies, but windborne insect migrants may also skip to the FBL to avoid the negative effects of occasional headwinds blowing at high elevation. This has been confirmed and carefully quantified in recent work published in this journal by Hawkes et al. [8], who revisited the same mountain pass where the Lacks had made their observations over 75 years ago, a true bottleneck for animal migration. Based on data from four consecutive autumns and combining three different ground-based sampling methods, Hawkes et al. offer a unique and comprehensive view of the insect community engaged in southward migratory flights across the Pyrenees. Large and relatively easy to identify insects (such as butterflies and dragonflies) were sampled by transect counts, while small insects such as flies were passively collected in a Malaise trap and subsequently identified in the laboratory. Estimates of the abundance of small migratory insects were obtained with a video camera that provided timed-counts in the sampling area.

During the four seasons in which the study was carried out, migratory insects of up to 20 families belonging chiefly to five orders were identified: Diptera (flies), Hymenoptera (bees and wasps), Hemiptera (bugs), Lepidoptera (butterflies and moths) and Odonata (dragonflies). The Diptera were by far the best represented order, since they accounted for up to 89% of all specimens, while Hymenoptera and Hemiptera contributed 6% and 5% of the specimens, respectively, and Lepidoptera and Odonata barely 1% each. The use of a two-way Malaise trap that passively captured insects flying predominantly in a southwards or northwards direction allowed these authors to establish two groups of species depending on their migratory strategies. On the one hand, some migrated mainly within the FBL, regardless of whether the wind direction was favourable (tailwind) or contrary (headwind) to the general southward migratory movement; on the other hand, some were mainly detected when the wind was blowing northwards, a situation which presumably forced insects to fly close to the ground to minimize the negative effects of a headwind on their southwards migration. The latter group corresponds to windborne migrants, which with a tailwind flew at high altitude to take advantage of the strong favourable wind. Some of the typical FBL migrants included the cabbage white and the clouded yellow butterflies (Pieris sp. and Colias crocea, respectively), various Diptera families (e.g. Chloropidae, Muscidae, Phoridae and Mycetophilidae) and solitary Hymenoptera and aphids. Among the typical windborne migrants, two families of Diptera, the hoverflies (Syrphidae) and the blowflies (Calliphoridae), and two butterflies, the red admiral (Vanessa atalanta) and the long-tailed blue (Lampides boeticus), were identified.

Although the ecological roles of this insect community are diverse, a very large percentage are recognized as pollinators (almost 90% of the identified individuals). This figure inflates their beneficial role somewhat, given that some species are less efficient pollinators than others, especially for crops. Despite this, a large number of the recorded insects undoubtedly provide this basic ecosystem service for humans. Moreover, the possible beneficial effect goes further because a significant number of these species in the larval stage also behave as decomposers (more than 30% of identifications, including several families of flies) or as predators of pests (e.g. hoverflies, which represented around 20%), two other important ecosystem services [9]. Only 5% of the insects were classified as agricultural pests, although this figure may be a significant underestimate, as discussed below. In any case, these figures show that the migration of insects across the Pyrenees is not only remarkable from a purely theoretical point of view but can also have direct implications for our society, even more so because of the magnitude of the movement detected. Indeed, a very remarkable result of the study was the quantification of the bioflow, which was estimated at 17.1 million insects crossing the narrow Bujaruelo pass (about 30 m wide) each autumn, which amounts to 3683 insects per metre and minute at times of maximum migratory intensity (1 h after solar noon). With over 2.7 million individuals, the fly Oscinella frit was the most abundant insect. This compares with a much lower number of 40 000 individuals for the most common butterfly, the clouded yellow.

While this quantification is highly commendable and represents a milestone that has never been reached before, some comments are warranted. First, Hawkes et al. completely neglect another key component of the migratory entomofauna, namely nocturnal insects. Surprisingly, they do not even mention a complementary study in which they assessed with light traps nocturnal migration in the Bujaruelo pass and found a large number of migrating moths [10]. Importantly, the most abundant species were all known agricultural pests (e.g. Helicoverpa armigera, Mythimna vitellina, Noctua pronuba, Autographa gamma, Agrotis ipsilon and Agrotis segetum), so the fundamentally beneficial role of this insect community needs to be nuanced. In fact, the negative impact of these species makes it even more necessary to study their movements, which calls for the use of light trapping as an additional method to be regularly included in any future assessment of insect migration across the Pyrenees (figure 1a,b).

Figure 1.

(a) Convolvulus Hawkmoth (Agrius convolvuli) and Death’s Head Hawkmoth (Acherontia atropos) regularly use mountain passes in their migrations between central and southern Europe. These two moths were captured on Col de Bretolet (1923 m) near the border between Switzerland and France in the Alps. (b) A network of light traps operating in strategic mountain passes could provide highly valuable information on nocturnal insects migrating through the Pyrenees, including some important pest species. Photos by (a) Irmi Zwahlen and (b) Jordi Jubany; permission to reproduce photographs has been granted by the copyright holders.

Another important aspect is the timing of this migration. The study by Hawkes et al. focuses exclusively on autumn migration; however, to fully understand the migratory dynamics we also need spring data, the season in which many of the subsequently detected species colonize higher latitudes after migrating from southern Europe or northern Africa. Having systematic spring passage data would, among other things, increase our knowledge of the source area of migrants [11], deepen our understanding of the phenomenon of partial migration in insects [12] and assess the selective value of migration based on population balances between the two migration periods [13].

A final aspect to be considered is the need for some degree of replication to achieve good spatial coverage and good representativeness throughout the whole Pyrenean range. It is clear, however, that sampling high mountain passes that are not easily accessible and are often subjected to adverse weather conditions makes this goal very difficult to achieve. One possibility to make it more feasible would be to set up a citizen science network to participate in the field work, for example, in the counts of the larger and more easily identifiable insects and, perhaps, also to help with the collection of samples from a hypothetical network of Malaise traps located at a series of strategic mountain passes. This possibility should not be ruled out given the enormous success citizen science has had in monitoring butterfly and bird populations in Europe. The fact that migration is concentrated in a few weeks in the spring and autumn makes the possible participation of volunteers more viable. A network of these characteristics could not only provide valuable data for estimating insect bioflow across the whole Pyrenees but would also multiply the possibilities of obtaining information of ecological interest such as predictable predator–prey interactions between birds and insects co-occurring at high densities during migration. Surprisingly, this aspect has remained very little explored so far, despite the fact that it could have important ecological implications [14].

Despite these shortcomings, Hawkes et al. have produced a fascinating work that, in addition to being a very valuable contribution to the study of migratory insects, will enable us to investigate whether the changes affecting this insect community are a response to global change. A notable example is that of Oscinella frit, the most abundant species in the study, but which, surprisingly, does not appear in the catalogue of the Diptera of Spain, Andorra and Portugal published by specialists in 2000 (Santos Rojo 2023, personal communication). In fact, it was not until the current century that this species began to appear in abundance in Spain, where it was found to cause damage to cereal crops in certain regions. Interestingly, in a recent study in which 40 farms in southern Canada were monitored for a decade, the abundance of Diptera increased in the farms with greatest agricultural intensity, especially when there was an increase in cereal crops such as wheat [15]. We can speculate that the great abundance of O. frit in the samples of Hawkes et al. could respond to a similar phenomenon and relate to an intensification of cereal cultivation north of the Pyrenees. Another sign of change is the relative rarity of dragonflies in recent samplings at Bujaruelo, which contrasts with what the Lacks described as an ‘uncountable stream’ in 1950. In any case, there is no doubt that consolidating the monitoring of migratory insects across the Pyrenees would be enormously helpful in understanding how these organisms and the ecosystem services they provide and the damage they can cause are being affected by global change.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

This article has no additional data.

Declaration of AI use

I have not used AI-assisted technologies in creating this article.

Authors’ contributions

C.S.: conceptualization, writing—original draft.

Conflict of interest declaration

I declare I have no competing interests.

Funding

No funding has been received for this article.