Insect decline in the Anthropocene: Death by a thousand cuts

Although conservation efforts have historically focused attention on protecting rare, charismatic, and endangered species, the “insect apocalypse” presents a different challenge. In addition to the loss of rare taxa, many reports mention sweeping declines of formerly abundant insects [e.g., Warren et al. (29)], raising concerns about ecosystem function.

Insects comprise much of the animal biomass linking primary producers and consumers, as well as higher-level consumers in freshwater and terrestrial food webs. Situated at the nexus of many trophic links, many numerically abundant insects provide ecosystem services upon which humans depend: the pollination of fruits, vegetables, and nuts; the biological control of weeds, agricultural pests, disease vectors, and other organisms that compete with humans or threaten their quality of life; and the macrodecomposition of leaves and wood and removal of dung and carrion, which contribute to nutrient cycling, soil formation, and water purification. Clearly, severe insect declines can potentially have global ecological and economic consequences.

While there is much variation—across time, space, and taxonomic lineage—reported rates of annual decline in abundance frequently fall around 1 to 2% (e.g., refs. 12, 13, 17, 18, 30, and 31). Because these rates, based on abundance, are likely reflective of those for insect biomass [see Hallmann et al. (26)], there is ample cause for concern (i.e., that some terrestrial regions are experiencing faunal subtractions of 10% or more of their insects per decade). To what extent such declines translate into shifts or losses of ecosystem function has yet to be assessed.

Not all insects are declining. Four papers in this special issue note instances of insect lineages that have not changed or have increased in abundance (24, 29, 32, 33). Many moth species in Great Britain have demonstrably expanded in range or population size (34–36). Numerous temperate insects, presumably limited by winter temperatures, have increased in abundance and range, in response to warmer global temperatures (29, 32, 35; but see ref. 37). Anthropophilic and human-assisted taxa, which include many pollinators, such as the western honey bee (Apis mellifera) in North America, may well thrive due to their associations with humans. Increasing abundances of freshwater insects have been attributed to clean water legislation, in both Europe and North America (17, 18). In some places, native herbivores have flourished by utilizing nonnative plants as adult nectar sources or larval foodplants (38), and there are even instances where introduced plants have rescued imperiled species (39).

Abundant evidence demonstrates that the principal stressors—land-use change (especially deforestation), climate change, agriculture, introduced species, nitrification, and pollution—underlying insect declines are those also affecting other organisms. Locally and regionally, insects are challenged by additional stressors, such as insecticides, herbicides, urbanization, and light pollution. In areas of high human activity, where insect declines are most conspicuous, multiple stressors occur simultaneously. Considerable uncertainty remains about the relative importance of these stressors, their interactions, and the temporal and spatial variations in their intensity. Hallmann et al. (13), in their review of the dramatic losses of flying insects from the Krefeld region, noted that no simple cause had emerged and that “weather, land use, and [changed] habitat characteristics cannot explain this overall decline…” When asked about his group’s early findings of downward population trends in insects (12), Dirzo summed up his thinking by stating that the falling numbers were likely due to a “multiplicity of factors, most likely with habitat destruction, deforestation, fragmentation, urbanization, and agricultural conversion being among the leading factors” (40). His assessment seems to capture the essence of the problem: Insects are suffering from “death by a thousand cuts” (Fig. 1). Taking the domesticated honey bee as an example, its declines in the United States have been linked to (introduced) mites, viral infections, microsporidian parasites, poisoning by neonicotinoid and other pesticides, habitat loss, overuse of artificial foods to maintain hives, and inbreeding; and yet, after 14+ y of research it is still unclear which of these, a combination thereof, or as yet unidentified factors are most detrimental to bee health.

More than half of the contributions in this collection directly or indirectly advance knowledge of particular stressors. Climate change, habitat loss and degradation (especially of tropical forests), and agriculture emerged as the three most important stressors considered by our authors, with the first of these receiving the greatest attention in the symposium. Halsch et al. (41) review studies implicating climate effects as drivers of insect population changes and discuss the complexities of responses of species to climatic stress in montane habitats. Janzen and Hallwachs (27) point to a warming and increasingly erratic climate as the most probable stressor underlying the region-wide losses of moths and other insects that they are monitoring across a large and multiecosystem area of northwestern Costa Rica. Høye et al. (32) assess faunal changes in Greenland, far removed from direct human impacts, and report only a modest signal of climate impacts to the arthropod fauna there, with many lineages increasing in abundance. Schowalter et al. (24) present new data suggesting that insect responses to temperatures within Puerto Rico’s Luquillo Experimental Forest, a hurricane-mediated ecosystem, are driven principally by frequent storms and poststorm successional history rather than by global climate warming. Three contributions from north temperate and Arctic locations point to range expansions or local population increases concomitant with warming temperatures (29, 32, 33), with the last of these noting examples of range extensions accompanied by declines in local abundance.

Many of the butterfly declines in Europe appear to be directly linked to changes in agricultural practices, with the rate of losses accelerating after World War II, when family farms began to amalgamate into larger commercial operations, modern tractors and mechanized equipment were employed to accelerate industrialization of agriculture, insecticides became widely available, and synthetic fertilizers could be manufactured and applied in prodigious volume (1, 16, 23, 42). Since the 1990s, the computerization of farming has dramatically changed the nature of agriculture (e.g., software-driven machinery can yield numerous efficiencies in planting, chemical applications, and harvest), further disadvantaging small-scale operations and practices. Much modern agriculture has become incompatible with nature, with its effects in the tropics especially worrisome. There deforestation, principally for agricultural expansion, is progressing at alarming rates, with its effects on insects and other arthropods essentially unmeasured (1).

The chronic effects of nitrification, mostly from the combustion of fossils fuels and the manufacture of more than 200 million tons of reactive nitrogen (mostly for fertilizer) annually, is now widely recognized as a major global stressor of insect and plant diversity (34, 42–44). Similarly, there is mounting evidence that light pollution is driving local declines in suburban and urban locations (45, 46). Although not an emphasis of the 11 articles, urbanization is increasingly recognized as an important stressor (47, 48).

Essential time-series data on the rates, geographic scope, ecological aspects, and taxonomic nature of insect population trends are scant, relative to those for vertebrates. Most insect taxa are far more species-rich than vertebrate taxa; there are more than a million described species of insects and even the most modest estimates calculate that another 4.5 to 7 million remain unnamed (21, 22). Species-level taxonomy for many lineages is challenging and often effectively nonexistent, especially for tropical faunas (which represent the majority of insect diversity); identifying tiny species can involve microscopic study, genitalic dissection, and DNA analysis, requiring equipment and expertise far beyond that needed for the censusing of vertebrates or plants. As a consequence, essentially all historical datasets that could be used to study insect decline have been biased toward agriculturally important and showy, well-known, extratropical taxa with long histories of public interest. Even for butterflies, historical demographic data are limited, deriving primarily from a few historically wealthy countries with a relatively deep record of natural history study.

An especially challenging aspect of interpreting insect demographic data are their episodic generation-to-generation swings. Documenting a 1 to 2% rate of annual decline against a background of flux that may swing through two to three orders of magnitude in as many years often requires decades of census data, the collection of which can be labor-intensive and expensive. Additionally, because many insects are ecologically specialized and their home ranges minuscule, even small-scale spatial variation in environmental conditions (e.g., soil attributes or vegetation height) can add heterogeneity to census data (49, 50).

One way that researchers have dealt with the complexity of population-level stochasticity in insects is to aggregate data at higher taxonomic levels: For example, using total insect biomass as a proxy for biodiversity, or aggregating data across different sites. Studies that generalize across datasets, higher taxonomic categories, or ecological groups (e.g., refs. 17, 18, 51, and 52) provide much-needed perspectives relevant to ecological function as, for example, the amount of insect food available to nestlings (53) and other insectivores or the general health of a region’s pollinators. Although data aggregation and metaanalyses are required approaches for understanding global phenomena, by their nature, they often overlook species-level trends.

To make predictions about what lineages will be most threatened and to inform policymaking and other conservation actions, it is essential to pay attention to the details and assumptions that underlie aggregated, multitaxon, and multiregion analyses. Much resolution is lost when data for insects with different demographic features or ecological needs are combined, such as, for example, when counts for crop-infesting aphids are analyzed with those for rare, habitat-restricted tiger beetles. Insect biomass is not interchangeable across food webs or among major taxa any more than hummingbirds and jaguars can be summed: An abundance of leafhoppers would be of little value to a nesting pair of warblers, relative to an abundance of caterpillars. Much can be learned by disambiguating the demographic data harvested from cities and intensively managed croplands from those derived from fragile communities: For example, cloud forests, rain forests, Arctic-alpine systems, low-nutrient communities, island biotas, and others (Fig. 2). A common finding across many reports of insect decline is that the rates of decline for ecological specialists are much steeper than those for more generalized taxa (e.g., refs. 29, 33, 42, 54, and 55). Based on first principles, taxa representing higher trophic levels (e.g., insect parasitoids, and in turn their parasites) would be expected to be suffering some of the highest rates of decline, a matter where data are especially scant (but see refs. 27 and 56). Studies of divergent taxa or heterogeneous locations need to employ appropriate hierarchical analyses and exercise caution in accepting hypotheses of change or stasis, especially when statistical power is low.

Although mainstream media in the United States has paid little attention to the plight of pollinators (59), some policymakers at the federal level have taken note and made millions of dollars available for pollinator research and habitat improvement (63, 64) since the National Research Council’s (58) status report on pollinators, and the 2014 petition to list the monarch butterfly as a threatened species (65). The Agriculture Improvement Act of 2018 (66) is annually providing many millions of dollars for pollinator research and conservation across the United States.

Funding to support insect conservation research is growing. In June 2018, the European Union endorsed an initiative to protect pollinators across member states (67). Germany’s federal government pledged $118,500,000 for insect conservation, monitoring, and research in September 2019 (68). The Swedish government plans to spend $25 million on pollinator protection initiatives over the next 3 y (69). In the spring of 2020, the government of Costa Rica endorsed a $100 million-effort to inventory and sequence the DNA barcode region of every multicellular creature in the country over a decade, with funding to come from international sources. Much of the country’s biodiversity will be insects, with most of these to be captured by Malaise traps, as part of Costa Rica’s BioAlfa initiative (27).

The European Union and several other countries have passed legislation to restrict use of some pesticides. Several neonicotinoids, insecticides that are commonly applied to seed coats and then taken up by the growing tissues, have been banned from use on field crops because even a treated plant’s nectar and pollen can be toxic to nontarget insects that visit its flowers (70, 71). In January 2020, the Environmental Protection Agency released an interim decision on the use of neonicotinoids in the United States aimed at protecting pollinators (72). Another downside to the neonicotinoid class of systemic insecticides is their environmental persistence: Their half-lives can be up to 1,000 d in soils and more than a year in woody plants, and their water-solubility allows the pesticide to move into and accumulate in soil and lakes, creeks, and other water bodies (73).

Although recent positive actions have been taken by governments in Costa Rica, the European Union, Germany, Sweden, and the United States, changes in national leadership can easily reverse environmental initiatives. Under the Trump administration in the United States, multiple federal policy changes were made that are inimical to conserving biodiversity: mention of climate change was pulled from government websites; an Executive Order reduced the size of Grand Staircase-Escalante National Monument in Utah by a half-million acres; and the administration finalized plans to permit oil drilling in the Arctic National Wildlife Refuge.

The Xerces Society, the largest global group in the United States promoting insect conservation, has enjoyed considerable growth over the past two decades: from a staff of 4 in 2000 to 54 full-time employees in 2020. Their major priorities, while still rooted to the protection of rare and endangered species and fostering a greater appreciation for insects and other invertebrates, now include focused efforts to protect pollinators, make agriculture more diversity-friendly, reduce pesticide use and impacts, and take a more active role in developing policy for the protection of invertebrates. Their current mailings reach about 100,000 individuals, partners, and policymakers across the globe.

There are growing numbers of community (citizen) science and education initiatives to survey, conserve, and raise awareness of insects and their importance as pollinators, prey, nutrient recyclers, and focal organisms in science and technology, as well as art, literature, and other aspects of culture (28). Monarch Watch has 55,000 Canadian, Mexican, and US followers, many of whom are educators. The United Kingdom’s Big Butterfly Count enjoyed the participation of more than 111,600 community scientists in 2020 (74). Costa Rica’s BioAlfa is a tropical version (27), focused on having the entire country be appreciative of the socioeconomic opportunities offered by wild biodiversity. Such actions could play important roles in changing public attitudes toward insects and motivate efforts to protect them.

Social media groups with a focus on insects are flourishing. Facebook groups dedicated to insects are propagating worldwide, parsed by taxon, geographic region, and even life stage. Husbandry and pet keeping anchored to insects and other arthropods have become popular. The Caterpillar Identification of North America Facebook group presently has more than 13,100 members. Page views on the United States’ Moth Photographer Group’s website have gone from 1.8 million in 2010 to more than 14.7 million in 2019. iNaturalist, an online resource for taxonomical identification, currently boasts more than 13,600,000 insect posts since 2010, with verifiable records from almost every country in the world (Fig. 3).