Article start

Contents lists available atScienceDirectBiological Conservationjournal decline of the entomofauna: A review of its driversFrancisco Sánchez-Bayoa,, Kris A.G. Wyckhuysb,c,daSchool of Life & Environmental Sciences, Sydney Institute of Agriculture, The University of Sydney, Eveleigh, NSW 2015, AustraliabSchool of Biological Sciences, University of Queensland, Brisbane, AustraliacChrysalis, Hanoi, Viet NamdInstitute of Plant Protection, China Academy of Agricultural Sciences, Beijing, ChinaARTICLE INFOKeywords:ExtinctionPollinatorsAquatic insectsAgriculturePesticidesEcosystem servicesGlobal change ecologyABSTRACTBiodiversity of insects is threatened worldwide. Here, we present a comprehensive review of 73 historical reportsof insect declines from across the globe, and systematically assess the underlying drivers. Our work revealsdramatic rates of decline that may lead to the extinction of 40% of the world's insect species over the next fewdecades. In terrestrial ecosystems, Lepidoptera, Hymenoptera and dung beetles (Coleoptera) appear to be thetaxa most affected, whereas four major aquatic taxa (Odonata, Plecoptera, Trichoptera and Ephemeroptera) havealready lost a considerable proportion of species. Affected insect groups not only include specialists that occupyparticular ecological niches, but also many common and generalist species. Concurrently, the abundance of asmall number of species is increasing; these are all adaptable, generalist species that are occupying the vacantniches left by the ones declining. Among aquatic insects, habitat and dietary generalists, and pollutant-tolerantspecies are replacing the large biodiversity losses experienced in waters within agricultural and urban settings.The main drivers of species declines appear to be in order of importance: i) habitat loss and conversion tointensive agriculture and urbanisation; ii) pollution, mainly that by synthetic pesticides and fertilisers; iii)biological factors, including pathogens and introduced species; and iv) climate change. The latter factor isparticularly important in tropical regions, but only affects a minority of species in colder climes and mountainsettings of temperate zones. A rethinking of current agricultural practices, in particular a serious reduction inpesticide usage and its substitution with more sustainable, ecologically-based practices, is urgently needed toslow or reverse current trends, allow the recovery of declining insect populations and safeguard the vital eco-system services they provide. In addition, effective remediation technologies should be applied to clean pollutedwaters in both agricultural and urban environments.1. IntroductionFor years, biologists and ecologists have been concerned about theworldwide reduction in biodiversity undergone by many terrestrial andaquatic vertebrates (Ceballos and Ehrlich, 2002Ceballos and Ehrlich, 2002;Pimm and Raven,2000; Wilson, 2002), yet scientists have only recently voiced similarconcerns about invertebrate taxa, particularly insects. Population de-clines imply not only less abundance but also a more restricted geo-graphical distribution of species, and represent thefirst step towardsextinction (Diamond, 1989). Much of the blame for biodiversity lossfalls on human activities such as hunting and habitat loss through de-forestation, agricultural expansion and intensification, industrialisationand urbanisation (Ceballos et al., 2017; Maxwell et al., 2016), whichjointly claimed a 3050% encroachment on natural ecosystems at theend of the 20th century (Vitousek et al., 1997).There is compelling evidence that agricultural intensification is themain driver of population declines in unrelated taxa such as birds, in-sectivorous mammals and insects. In rural landscapes across the globe,the steady removal of natural habitat elements (e.g. hedgerows),elimination of natural drainage systems and other landscape featurestogether with the recurrent use of chemical fertilisers and pesticidesnegatively affect overall biodiversity (Fuller et al., 1995;Newton, 2004;Tilman et al., 2001). Recent analyses point to the extensive usage ofpesticides as primary factor responsible for the decline of birds ingrasslands (Mineau and Whiteside, 2013) and aquatic organisms instreams (Beketov et al., 2013), with other factors contributing to oramplifying their effects to varying extent. Yet, we don't know whetherthe same factors explain the parallel entomological demise that we arewitnessing.In 2017, a 27-year long population monitoring study revealed a 12 September 2018; Received in revised form 23 January 2019; Accepted 25 January 2019Corresponding author.E-mail, Sánchez-Bayo).Biological Conservation 232 (2019) 8–27Available online 31 January 20190006-3207/ © 2019 Elsevier Ltd. All rights reserved.T
shocking 76% decline inflying insect biomass at several of Germany'sprotected areas (Hallmann et al., 2017). This represents an average2.8% loss in insect biomass per year in habitats subject to rather lowlevels of human disturbance, which could either be undetectable orregarded statistically non-significant if measurements were carried outover shorter time frames. Worryingly, the study shows a steady de-clining trend over nearly three decades. A more recent study in rain-forests of Puerto Rico has reported biomass losses between 98% and78% for ground-foraging and canopy-dwelling arthropods over a 36-year period, with respective annual losses between 2.7% and 2.2%(Lister and Garcia, 2018). The latter authors showed parallel declines inbirds, frogs and lizards at the same areas as a result of invertebrate foodshortages. Both studies agree with the declining trend inflying insects(mainly Diptera) observed a decade earlier in parts of Southern Britain(Shortall et al., 2009). As insects comprise about two thirds of all ter-restrial species on Earth, the above trends confirm that the sixth majorextinction event is profoundly impacting life forms on our planet(Thomas et al., 2004Thomas et al., 2004).While the arthropod declines in tropical rainforests correlate wellwith climatic changes, the 12 different factors (e.g. increases in arableland, deforestation, global warming) that were thought to be re-sponsible for year-to-year drops in insect biomass in Germany barelyaccounted for ~20% of observed declines. Rather surprisingly, 80% ofobserved inter-annual variability in insect numbers was left un-explained (Hallmann et al., 2017). Although the authors did not assessthe effect of synthetic pesticides, they did point to them as a likelydriver of the pervasive losses in insect biomass.The above studies, however, are in line with previous reports onpopulation declines among numerous insect taxa (i.e. butterflies,ground beetles, ladybirds, dragonflies, stoneflies and wild bees) inEurope and North America over the past decades. It appears that insectdeclines are substantially greater than those observed in birds or plantsover the same time periods (Thomas et al., 2004), and this could triggerwide-ranging cascading effects within several of the world's ecosystems.This review summarises our current state of knowledge about insectdeclines, i.e., the changes in species richness (biodiversity) and popu-lation abundance through time, and points to the likely drivers of thelosses so that conservation strategies to mitigate or even reverse themmay be implemented. Previous reviews are partial in scope, restricted toindividual groups of insects (e.g. butterflies, carabids) in specific re-gions, but no study has put together a comprehensive review of allinsect taxa nor compared the localfindings among different parts of theworld.2. MethodologyWe aimed at compiling all long-term insect surveys conducted overthe past 40 years that are available through global peer-reviewed lit-erature databases. To that effect we performed a search on the onlineWeb of Science database using the keywords [insect*] AND [declin*]AND [survey], which resulted in a total of 653 publications. The ma-jority of these referred to Hymenoptera (55), Diptera (45), Coleoptera(44) and Lepidoptera (37) taxa, among which only a few dealt withlong-term surveys. Reports that focused on individual species, pestoutbreaks or invasive species were excluded. We selected surveys thatconsidered all species in a taxon (e.g. family or order) within large areas(i.e. a region, a country) or smaller areas surveyed intensively overperiods longer than 10years. Additional papers were obtained from theliterature references. Finally, only surveys that reported changes inquantitative data over time, either species richness or abundance, wereconsidered. Thus, this review covers 73 reports on entomofauna de-clines in various parts of the world (Fig.1) and examines their likelycauses (Table S1). Because the overwhelming majority of long-termsurveys have been conducted in developed countries, particularly in thenorthern hemisphere, this review is geographically biased and does notadequately cover trends in tropical regions, where information on in-sect biodiversity is either incomplete or lacking (Collen et al., 2008).The above literature records use accurate scientific data on speciesdistribution from museum specimens (56%), which are compared withlong-term survey data obtained decades later (72%), and sometimesrely upon citizen science data (8%). Because the latter data tend tooverestimate the diversity of insects due to over-reporting of rare spe-cies (Gardiner et al., 2012), the overall assessment of biodiversity canbe considered conservative.Conservation status of individual species follows the IUCN classifi-cation criteria (IUCN 2009): threatened species include vulnerable(> 30% decline), endangered (> 50%) and critically endangered(> 75% decline) species. Data on population abundance are more dif-ficult to obtain than geographical distribution records, but a few reportsquantified the extent of such declines for Lepidoptera, Hymenopteraand dung beetles (Coleoptera). An annual rate of decline (percentage ofspecies declining per year) was estimated for each taxon and region.A meta-analysis of the declines among the various taxa and regionswas performed, with groups compared using analysis of variance(ANOVA). Enumerated drivers of the declines -as indicated by the re-ports' authors- are tabulated and analysed, and further discussed withreference to experimental and other empirical data available in theliterature.Fig. 1.Geographic location of the 73 re-ports studied on the world map. Columnsshow the relative proportion of surveys foreach taxa as indicated by different colours inthe legend. Data for China and Queensland(Australia) refer to managed honey beesonly. (For interpretation of the references tocolour in thisfigure legend, the reader isreferred to the web version of this article.)F. Sánchez-Bayo, K.A.G. WyckhuysBiological Conservation 232 (2019) 8–279
3. Taxa declines3.1. LepidopteraButterflies and moths are valuable indicators of environmentalquality, considering their high degree of host-plant specialisation andvulnerability to habitat deterioration (Erhardt and Thomas, 1991).Given their presence in a broad range of habitats, the loss of Lepi-doptera may directly impact the delivery of key ecosystem services suchas pollination and natural pest control (Fox, 2013). Moths, which areabout 10 times more diverse than butterflies, constitute important preyitems of bats and help sustain population levels of myriad other in-sectivorous animals (Hahn et al., 2015; Vaughan, 2008;Wilson et al.,1999).Maes and van Dyck (2001)were thefirst to report drastic changes inbutterfly biodiversity in Flanders (Belgium) during the 20th century,including the extinction of 19 (out of 64) native species since 1834.Habitat loss resulted in a steady decline of 69% of the 45 extant species,while the extinction rate increased from 0.2 to 1.7 species/5-year since1950 as urbanisation and agricultural intensification expanded eight-fold (Maes and Van Dyck, 2001). A follow-up study in the Netherlandsfound that 11 of the 20 most common and widespread butterfly speciesdeclined both in distribution and abundance between 1992 and 2007,whereas populations of species associated with natural habitats innature reserves remained stable. Local populations ofLasiommata me-geraandGonepteryx rhamniare now endangered and two other species(Aglais ioandThymelicus lineola) are vulnerable (van Dyck et al., 2009).In parallel, the range of distribution of 733 species of day-flying mothswas recorded between 1980 and 2000: overall declines were observedin 85% of species, 38% of them experiencing > 75% reduction in theirarea (i.e. critically endangered), 34% being considered threatened and15% vulnerable (Groenendijk and van der Meulen, 2004). In particular,47 of the 55 species monitored in peat-bog marshes declined, six re-mained stable and only two (Plusia putnamiandDeltote bankiana) in-creased in range (Groenendijk and Ellis, 2011). The most affectedspecies are those with lower dispersal abilities and preference for oli-gotrophic habitats.Among the 269 species of macro-lepidopterans monitored for50 years at the Kullaberg Nature Reserve (Sweden), 45% were de-clining, 22 were coloniser species and 159 were no longer found in2004 (Franzén and Johannesson, 2007). Monophagous and oligopha-gous species using grass or herbs in wetlands were declining more thanthose feeding on deciduous trees or shrubs, confirming that dietaryspecialists bear the brunt of the declines. Species with a shortflight-period or those restricted to non-forest habitats were all associated witha high extinction risk. A comparison of historical records of 74 but-terflies in Finland showed how 60% of grassland species declined overthe past 50 years, whereas 86% of generalist species and 56% of thoseliving at forest edge ecotones increased in abundance (Kuussaari et al.,2007).Common traits of the 23 declining species are a reduced mobi-lity, oligotrophic habitat preference and seasonal migration behaviour.Another study on the populations of 306 species of noctuid moths inFinland over 19881997 reported the greatest declines for species withcomparatively small geographic range, whereas polyphagous mothswith longerflight periods and those that overwintered as adults had thewidest distributions (Mattila et al., 2006). By contrast, species thatoverwintered as either larval or pupal stages suffered the largest de-clines over that period.Similarfindings were reported for north-eastern Spain, where yearlymonitoring of 183 butterfly species over 19942014 permitted an in-depth analysis of population trends and associated factors for 66 spe-cies. While 15 species had increased in abundance,five remain stableand 46 are declining (Melero et al., 2016). Although the extent ofspecies decline did not differ among seven habitat types selected, forestbutterflies appeared to be declining faster than those from other habi-tats due mainly to specialisation of the larval trophic stage and multi-voltinism.A comprehensive report on the status of 576 species of butterflies inEurope found that 71 were threatened and declined over a 25-yearperiod (van Swaay et al., 2006). The largest declines occurred amongspecialist butterflies of grassland biotopes (19% species), wetlands andbogs (15%) and woodlands/forests (14%), due to habitat conversioninto crops and the adoption of intensive agricultural practices, e.g.fertilisers and pesticides negatively affected 80% species. Some species(Lopinga achineandParnassius apollo) had declined due to afforestation,i.e. conversion of open woodland habitats to dense forests. Climatechange only affected a few endemic species adapted to mountainousbiotopes. A more recent assessment of 435 butterflies native to Europe(van Swaay et al., 2010) found that populations of 19% of species aredeclining, particularly in Mediterranean and eastern countries, 8.5%species are threatened, and three are critically-endangered, i.e.Pierisbrassicae wollastoni,Triphysa phryneandPseudochazara cingovskii. Bel-gium and the Netherlands are the European countries with the highestlosses of butterfly biodiversity (19 and 17 country-level extinctions,respectively), whereas Denmark and the U.K. have the least (4 speciesextinct each) (Maes and Van Dyck, 2001). One species (Aricia hya-cinthus) is considered extinct over the European continent.Butterflies appear to be declining faster in the United Kingdom, as74% of 46 non-migratory butterflies restricted their distribution over19701999(Warren et al., 2001). Using a comprehensive databasecompiled by amateur collectors and scientists over a 29-year period inthe entire British Isles, the authors showed that habitat specialists ex-perienced the largest reductions in distributional area. Specialist andsedentary species not showing changes in abundance over 25 years hadreduced their distribution on average by 15%. Other studies indicatethat 41 out of 54 common butterflies had been declining since the1970s, with 26% of species showing decreases over 40% of their range(Fox et al., 2006), while 13% of 10-km squares in the monitoring gridreported disappearance of butterfly species (Thomas et al., 2004). Al-though authors did not attempt to correlate the declines with specificdrivers, the following combination of factors was suggested: habitatfragmentation and/or destruction, intensification of agriculture, in-cluding the increased usage of chemical fertiliser and pesticides, andperhaps over-collectingalthough such practice has been greatly re-duced by more environmental awareness. To minimise biodiversitylosses among butterflies and moths, the UK Butterfly MonitoringScheme (UKBMS) was created, which compiles data on the abundanceand distribution of all species across the country since 1976. An initialanalysis of 50 species showed a largefluctuation in butterflies amongyears, with specialist species having declined by 34% nationally sincethe scheme was established; generalist species had declined in England(12%) but little (6%) or not at all in Scotland. Major declines occurredin forests and farmland regardless of the efforts to restore biodiversityfrom 2000 onwards (Brereton et al., 2011). A further analysis of 17widespread and resident species of butterflies between 1984 and 2012showed that abundance of all species decreased by 58% since the year2000, while 15 species exhibited population declines at average annualrates between0.8% and6.7% (Gilburn et al., 2015). Thus, 41% ofthe species studied are threatened. Increasing summer temperatureshad a marked positive effect on butterfly abundance, whereas none ofthe other climatic factors could explain the decrease in their popula-tions. By contrast, the steepest declines occurred in areas with highproportions of farmland treated with neonicotinoid insecticides; indicesfor the 15 declining species showed negative associations with neoni-cotinoid usage.Similar rates of decline were reported among 337 moth speciesthroughout England between 1968 and 2003: 222 showed decliningpopulations over the 35-year study period, with a median 10-yr popu-lation decrease of 12%, although decreases were larger in the south ofthe country (17%) compared to the north (5%) (Conrad et al., 2006).The larger declining trends in the south were associated with the degreeof agricultural intensification, as also recorded at Rothamstead betweenF. Sánchez-Bayo, K.A.G. WyckhuysBiological Conservation 232 (2019) 8–2710