Resolving biology’s dark matter : species richness, spatiotemporal distribution, and community composition of a dark taxon

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Resolving biology’s dark matter : species richness, spatiotemporal distribution, and community composition of a dark taxon © The Author(s) 2024 Published version Hartop, Emily; Lee, Leshon; Srivathsan, Amrita; Jones, Mirkka; Peña-Aguilera, Pablo; Ovaskainen, Otso; Roslin, Tomas; Meier, Rudolf Hartop, E., Lee, L., Srivathsan, A., Jones, M., Peña-Aguilera, P., Ovaskainen, O., Roslin, T., & Meier, R. (2024). Resolving biology’s dark matter : species richness, spatiotemporal distribution, and community composition of a dark taxon. BMC Biology, 22, Article 215. https://doi.org/10.1186/s12915-024-02010-z 2024 Hartopetal. BMC Biology (2024) 22:215 https://doi.org/10.1186/s12915-024-02010-z RESEARCH Open Access © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. BMC Biology Resolving biology’s dark matter: species richness, spatiotemporal distribution, andcommunity composition ofadark taxon Emily Hartop1,2*, Leshon Lee3,4, Amrita Srivathsan3,5, Mirkka Jones6,7, Pablo Peña‑Aguilera8, Otso Ovaskainen6,9, Tomas Roslin6,8 and Rudolf Meier5,10* Abstract Background Zoology’s dark matter comprises hyperdiverse, poorly known taxa that are numerically dominant but largely unstudied, even in temperate regions where charismatic taxa are well understood. Dark taxa are every‑ where, but high diversity, abundance, and small size have historically stymied their study. We demonstrate how ento‑ mological dark matter can be elucidated using high‑throughput DNA barcoding (“megabarcoding”). We reveal the high abundance and diversity of scuttle flies (Diptera: Phoridae) in Sweden using 31,800 specimens from 37 sites across four seasonal periods. We investigate the number of scuttle fly species in Sweden and the environmental fac‑ tors driving community changes across time and space. Results Swedish scuttle fly diversity is much higher than previously known, with 549 putative specie) detected, com‑ pared to 374 previously recorded species. Hierarchical Modelling of Species Communities reveals that scuttle fly com‑ munities are highly structured by latitude and strongly driven by climatic factors. Large dissimilarities between sites and seasons are driven by turnover rather than nestedness. Climate change is predicted to significantly affect the 47% of species that show significant responses to mean annual temperature. Results were robust regardless of whether haplotype diversity or species‑proxies were used as response variables. Additionally, species‑level models of common taxa adequately predict overall species richness. Conclusions Understanding the bulk of the diversity around us is imperative during an era of biodiversity change. We show that dark insect taxa can be efficiently characterised and surveyed with megabarcoding. Undersam‑ pling of rare taxa and choice of operational taxonomic units do not alter the main ecological inferences, making it an opportune time to tackle zoology’s dark matter. Keywords Dark taxa, Megabarcoding, DNA barcoding, Biodiversity discovery, Hierarchical Modelling of Species Communities, Diptera, Phoridae *Correspondence: Emily Hartop emily.har[email protected] Rudolf Meier Rudolf.Meier@hu‑berlin.de Full list of author information is available at the end of the article Page 2 of 16 Hartopetal. BMC Biology (2024) 22:215 Background If we go on the way we have, the fault is our greed and if we are not willing to change, we will disappear from the face of the globe, to be replaced by the insect. Jacques Yves Cousteau. Biodiversity loss in the Anthropocene is driven by changes in climate, land use, but also species introductions [14, 19]. Such loss can result in the concomitant decline in ecosystem services vital to society [47, 49] and may ultimately disrupt global supply chains and food security [81]. Accurate monitoring of biodiversity is therefore a global priority [54]. A crucial first step to monitoring biodiversity is obtaining robust quantitative baseline data. Most biologists would consider this to be data on species diversity, abundance, and biomass for those taxa that contribute substantially to these quantitative metrics. However, most biodiversity studies cover only a few well-studied groups that are relatively easily identified and quantified (e.g. birds, mammals, amphibians, bees, and butterflies). Such charismatic taxa are then used as proxies for all taxa in a region [7, 8, 33, 56, 57, 61, 70, 79] instead of basing our understanding of global biodiversity on a broad and unbiased representation of biodiversity covering a wide range of traits and responses to environmental change [2, 21, 37]. A key component of global biodiversity are “dark taxa”, i.e. taxonomic groups for which less than 10% of the diversity is described and the species diversity is estimated to be upwards of 1000 [40]. Such poorly known groups do not just inhabit inaccessible realms like the deep sea [60] but also the terrestrial habitats in which we live. A recent study [66] revealed that 20 insect families (of which 10 belong to Diptera) account for > 50% of local species diversity. Alarmingly, the very same families suffer from extreme taxonomic neglect and are therefore poorly represented in biodiversity surveys. Identifying and tackling the diversity of these dark taxa with scalable techniques thus emerges as an urgent priority for biodiversity science. Large-scale studies of dark taxa have only become feasible in recent years due to advancements in sequencing technologies coupled with efficient single-specimen DNA barcoding workflows [16, 40, 51, 68, 69, 82] : “megabarcoding”). Such workflows allow large numbers of specimens to be processed and sorted to putative species (mOTUs), while providing exact specimen counts, and vouchers for subsequent morphological, taxonomic, and biological work. Scuttle flies in the family Phoridae (Diptera) have been considered the seventh most speciose and abundant insect family globally [66]. However, to date only ca. 4000 species have been described, although their actual diversity may be two orders of magnitude larger [68]. In addition to their extreme species richness, scuttle flies occupy a wide range of ecological niches, containing species that are herbivores and predators to scavengers, parasitoids, and parasites (reviewed by [20]). Nonetheless, previous ecological studies focusing on scuttle flies [23–30] have been of limited scope due to time-consuming morphological identification methods. It is essential that we assess taxa like these to ensure that ecological analyses reflect the bulk of biodiversity and represent a broad range of ecological niches. In this study, we use sorting with megabarcoding to generate the data for answering fundamental questions about this dark taxon. We ask how many species of scuttle flies occur in Sweden, how are they distributed across time and space, and what environmental variables drive their distribution. To test whether the choice of species and species delimitation method will affect the results, we carry out the same analyses using both mOTUs as species-proxies and haplotype diversity, and test whether rare species influence the overall inferences. We show how “dark taxon zoology” can quickly yield the answers urgently needed in the Anthropocene. Results Diversity We obtained a total of 31,739 COI barcodes belonging to 2697 haplotypes from scuttle fly samples from 37 sites (Fig.1, Additional file1: Fig. S1) and four seasonal time periods (Additional file1: Fig. S2) across Sweden (Additional file 1: Table S1). At the species threshold (1.7%) [40], we detected 549 mOTUs. Species accumulation curves revealed that scuttle fly species diversity was incompletely sampled overall and across sites, horticultural zones, and time periods (Additional file 1: Figs. S3, S4). Between 38 and 145 species were observed per site, with Chao1 richness estimates per site ranging from 83 to 244 species (Fig.2b, c). Regional richness estimates varied between nearly 400 species estimated in the southernmost coastal zone (zone 1), to fewer than 200 estimated for the alpine zone (Additional file1: Fig. S3). Midsummer and late-summer time periods were both characterised by richness estimates of around 450 species, while the late spring and offseason time periods showed lower species richness estimates of around 300 species (Additional file1: Fig. S4). Total species richness in Sweden was estimated at between 652 species (for Chao1, Fig. 2a) and 713 species (combined non-para- metric estimator (CNE) in [63]), suggesting that 100–160 species of scuttle flies remain to be sampled. Page 3 of 16 Hartopetal. BMC Biology (2024) 22:215 Ordinations ofcommunity composition NMDS plots revealed a clear distinction between scuttle fly communities in the southern (zones 1–3) and northern (zones 4–alpine) plant hardiness zones (Fig.1,Additional file 1: Fig. S5). ANOSIM supported significant differences between southern and northern zones at this threshold (R = 0.58, p = 0.001), and SIMPER revealed that north–south similarity was just 13.2%, as compared to similarities of 24.6% and 23.2%, respectively, among samples within the northern and southern zones (Additional file1: Tables S2–S3). The separation between zones was consistent across clustering thresholds ranging from haplotypes-as-such to a threshold of 1.7% sequence similarity, with stress values around 0.21 (Additional file1: Fig. S5 top row). Above a threshold of 3%, however, the patterns were increasingly blurred and the stress values were higher (0.25–0.27) (Additional file1: Fig. S5 bottom row). These patterns were also evident in ANOSIM analyses, where the sample statistic decreased from 0.54 for haplotypes to 0.27 for 5% mOTUs, indicating a decreasing relationship between scuttle fly community composition and plant hardiness zones at higher clustering thresholds (Additional file1: TableS2). Similarly, in SIMPER analyses, average similarity between zones increased from 20.2% for haplotypes to 41.4% for 5% mOTUs (Additional file1: TableS3). For Fig. 1 The location of the 37 study sites of the Swedish Insect Inventory Project colour‑coded according to the plant hardiness zones (odlingszoner, 1–8 and alpine) of the Swedish Horticultural Society (Riksförbundet Svensk Trädgård) (map used with permission) next to an NMDS plot of study samples colour‑coded with the same zones Page 4 of 16 Hartopetal. BMC Biology (2024) 22:215 species, scuttle fly communities were found to be distinct across most zones (R = 0.51, p = 0.001), with higher (average = 28.5%) similarity within zones than between (average = 19.2%) zones (Additional file1: Tables S2, 3). Samples from the late spring, midsummer, and latesummer time periods showed a clear progression along the ranked plant hardiness zones, while offseason samples appeared more randomly distributed (Additional file1: Fig. S6). Northern sites (IV–alpine) showed higher distinctness across seasons (R = 0.59) than when considering all sites (R = 0.27) (Additional file1: TableS4). The between-season similarity of all sites averaged 16.9% and within time period similarity averaged 23.7% (Additional file 1: Table S5, top). However, for northern sites only (IV–alpine), between time period similarity was 21.3% and within time Fig. 2 a Species accumulation and Chao1 estimate curves for the scuttle fly dataset across Sweden. Notably, current sampling is far from exhaustive, and numbers of singletons and doubletons in our dataset are high regardless of sample size, b species accumulation curves by sampling sites, colour‑coded by zone, and c Chao1 estimate curves by sampling sites, colour‑coded by zone. For a map of the zones using the same colour codes, see Fig. 1 Page 5 of 16 Hartopetal. BMC Biology (2024) 22:215 period similarity was 37.2% (Additional file1: TableS5, bottom). Hierarchical Modelling ofSpecies Communities (HMSC) Models of the four response matrices (species-occurrence, haplotypes-occurrence, species-abundance, haplotypesabundance) showed relatively good MCMC convergence with potential scale reduction factors of the models’ beta and omega parameters close to one (Additional file1: Fig. S7). Explained variance averaged c. 30% for the occurrences and c. 60% for the abundances of both species and haplotypes, but there were large differences in model fit among taxa (Additional file1: Fig. S8), haplotype occurrence mean ± sd Tjur R2 = 0.30 ± 0.15 (range 0.05–0.82); haplotype abundance R2 = 0.61 ± 0.26 (range 0.00–1.00); species presence-absence Tjur R2 = 0.30 ± 0.14 (range 0.05–0.72); and species abundance R2 = 0.59 ± 0.26 (range 0.02–1.00). Sampling time period explained the largest fraction of variance, on average, in all four HMSC models (mean 10–11% in the occurrence models; mean 15–16% in the abundance models; Fig.3, Additional file1: Fig. S8). Mean annual temperature explained almost as much variance as sampling time period in the occurrence models (mean 7% and 8% in the species vs haplotype models), but clearly less than sampling season in the abundance models (mean 6% in the species models and 5% in the haplotype models) (Fig.3, Additional file1: Fig. S8). All response variables showed strong climatic (including temporal) community structure and there was also strong spatial structure in species and haplotype occurrences linked with the latitudinal temperature gradient across Sweden. Tree cover explained more variance in the abundance models (mean 6% for both species and haplotypes) than in the occurrence models (mean 1% for both species and haplotypes). Trapping effort also explained more variance on average in the abundance models (mean 7% and 8%) than in the occurrence models (mean 1.5% for both), as did the effect of having (vs not having) sequenced the full trap sample (2% mean for occurrence, 6% mean for abundance). While the occurrences of taxa and haplotypes showed statistically supported responses to all model covariates, their abundances showed responses less frequently, and responses with strong support were primarily with seasonal covariates (Fig.4). Most species and haplotypes with a statistically supported seasonal abundance trend peaked in the late spring. Compared to late spring, the midsummer fauna showed a reduction of 28% in species counts and 16% in haplotype counts, respectively, with a further reduction of 24% in both species counts and haplotype counts towards the late summer. Taxa and haplotypes were also usually more prevalent in late spring than in mid- or especially late summer. However, a minority of species and haplotypes (9% and 7%, respectively) showed the reverse pattern, being more prevalent in late summer than in the spring. Offseason captures in the late fall to early winter were consistently low, and the number of offseason samples included in the models was the smallest (n = 20 samples). Nonetheless, 33% of the species modelled and 29% of the haplotypes modelled were occasionally detected in offseason samples. Taxon occurrences showed a mixture of positive (29% vs 30% for species and haplotypes, respectively) and negative (18% vs 23%) responses to the annual temperature gradient (“bio1”) across Sweden (Fig.4), reflecting the broad-scale compositional changes from southern towards northern Sweden seen in the NMDS ordinations. Where detected, the occurrence responses of taxa to forest or woodland cover were more often positive than negative (10% positive vs 4% negative for species; 8% vs 2% for haplotypes; Fig.4). Longer trapping periods did not consistently result in higher detection probabilities of taxa, presumably due to seasonal differences in trapping duration (4% positive vs 5% negative responses for species; 8% positive vs 4% negative for haplotypes; Fig.4). As expected, most taxa (81–82% in both the species and haplotype models) showed a negative occurrence response to the binary variable indicating whether all specimens in a sample were sequenced or not (“Full- Sample”) (Fig.4). The mean annual temperature gradient across Sweden was predicted to affect the prevalence of 38% of species, but does not appear to be a main driver of species abundance (Fig.4). The predicted effect of the temperature gradient on species prevalence during late summer was positive in 22% of taxa and negative in 16% of taxa (Additional file1: Fig. S9). Beyond spatial patterns explained by these climatic predictors, there was evidence of localised spatial autocorrelation in species and haplotype site occupancies at scales of less than 40km. Neither species nor haplotype abundances were spatially autocorrelated, nor did we detect statistical support for temporal autocorrelation in any model. Residual correlations in the distributions of taxa were detected among sites and samples in both the species and especially the haplotype occurrence models (Additional file1: Fig. S10). Residual associations of taxa over time were also evident, but less frequently (Additional file1: Fig. S10). Residual associations between taxon/haplotype occurrences likely indicate that our models either lack or imperfectly represent some of the variables that structure the occurrences of scuttle fly taxa and haplotypes in space and time. No residual associations among taxa were evident among sites or over time in the species or haplotype Page 6 of 16 Hartopetal. BMC Biology (2024) 22:215 abundance models, and very few were detected among samples (Additional file1: Fig. S10). Hence, covariance in the abundances of taxa and haplotypes across occupied samples was well modelled by the environmental and other covariates in these models. The observed richness of the excluded rare vs modelled common species and haplotypes was strongly positively correlated (R = 0.59 for species and R = 0.80 for haplotypes). Compositional differences between samples in terms of the rare vs common species and haplotypes were Fig. 3 Summary of explained variance in the occurrences and abundances of scuttle fly haplotypes and species across samples (fractions within bars represent the average percentage of variance explained by each fixed or random effect in the models). All four models show strong climatic structure (in shades of brown) on scuttle fly communities, as captured by fixed variables describing sampling season and mean annual temperature and a random effect based on median sampling date. The spatial fraction of explained variance (lime green) reflects community structural differences among sites that were not captured by the fixed effect covariates. Differences in sampling effort, i.e. whether or not trapped flies were all sequenced or not and the number of field trapping days per sample, also affect the predictability of community structure (shades of orange). The abundance of species and haplotypes, and to a lesser extent their occurrence, was also strongly structured by habitat type as described by forest and woodland cover (blue). Finally, we included a categorical random effect representing sample identity (purple) Page 7 of 16 Hartopetal. BMC Biology (2024) 22:215 also positively correlated (for taxon presence-absence Mantel R = 0.31 for species and Mantel R = 0.64 for haplotypes; for taxon abundance Mantel R = 0.30 for species and Mantel R = 0.60 for haplotypes). Community dissimilarity Turnover accounted for the bulk of spatial, temporal, and spatiotemporal variation in community dissimilarity (Fig.5). In the spatial analyses, the mean turnover of species and haplotypes between sample pairs was 0.75 and 0.87, respectively, while the corresponding mean values of nestedness were 0.07 and 0.03. In temporal analyses, similarly, the mean values of turnover were 0.45 and 0.74 for species and haplotypes, respectively, and the corresponding means of nestedness were 0.13 vs 0.07. Finally, mean turnover values for the spatiotemporal analyses were 0.75 and 0.84 vs a mean nestedness of 0.07 and 0.04 for species and haplotypes, respectively. Regardless of the time period and operational taxonomic units used, we found a significant positive correlation between turnover and geographical distances, meaning that communities in closer proximity to each other are more similar in composition (Additional file 1: Fig. S11). The mean spatial turnover of species communities was highest in the summer time periods (midsummer: 0.82, late summer: 0.84) and lower in late spring and offseason (0.78 and 0.72, respectively). However, we did not find any significant correlation between temporal distances (difference in mean week) and turnover (see TableS6). Patterns of nestedness showed no detectable correlation with any of the distances explored (except for the grouped temporal distance) (Additional file1: TableS6). Within each time period, scuttle fly communities become more distinct from neighbouring communities from late spring to late summer (Additional file1: TableS6 and Figs. S11–13). Discussion Our study marks the first country-wide examination of a dark taxon’s diversity and distribution. It revealed more than 500 species of scuttle flies based on processing ca. 31,800 specimens. The ecological analyses suggest that climate change will have profound (and quickly apparent) effects on communities of scuttle flies that could serve as early indicators of future shifts in the environment. We demonstrate that armed with recent advancements in sequencing technologies, bioinformatics pipelines, and molecular barcoding workflows [40, 52, 68, 69, 75, 82], we are now able to resolve patterns of alpha diversity, spatial and temporal Fig. 4 Predicted occurrence and abundance responses of 193 scuttle fly haplotypes and 162 species to HMSC model covariates. Positive (red) and negative (blue) estimated responses with a posterior probability of 0.95 are illustrated. The three rows below the intercept illustrate the estimated effects of three levels of a categorical variable representing sampling time period (midsummer, late summer, offseason) relative to the baseline (late spring). The subsequent three rows represent responses to % forest or woodland cover and mean annual temperature (“bio1”) at sampling sites. The final two rows represent the effects of two sampling‑related differences among samples: the number of trapping days and whether all specimens in the trap sample were sequenced, and hence available for HMSC analysis, or not Page 8 of 16 Hartopetal. BMC Biology (2024) 22:215 turnover, and species communities of challenging dark taxa. We show that for the scuttle flies in Sweden, many species remain undiscovered, that local communities show major turnover in space and time, and that ecological patterns are largely robust to the finer details of species delimitation. Below, we will discuss each of these findings in further detail—and the importance of these patterns in dark taxon biology to biodiversity science. Diversity Our sample only scratched the surface of the scuttle fly fauna of Sweden (Fig.2, Additional file1: Figs. S3, S4). While the true Swedish fauna of scuttle flies is still hard to estimate, the 549 putative species found (and 652–713 predicted based on current sampling) greatly exceed the 374 previously documented from the country. Previous estimates for the scuttle fly fauna have ranged from 1100 to nearly 2000 species, suggesting Fig. 5 Pairwise community dissimilarity among scuttle fly species and haplotype samples. Following Baselga and Orme [5], we partitioned overall dissimilarity into its turnover and nestedness components and illustrate these as well as overall dissimilarity as a function of distance (standardised Euclidean distance calculated from sample site coordinates) Page 15 of 16 Hartopetal. BMC Biology (2024) 22:215 32. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. DNA primers for amplifica‑ tion of mitochondrial cytochrome c oxidase subunit I from diverse meta‑ zoan invertebrates. Mol Mar Biol Biotechnol. 1994;3:294–9. 33. Garda AA, Stein MG, Machado RB, Lion MB, Juncá FA, Napoli MF. Ecology, biogeography, and conservation of amphibians of the Caatinga. In: Silva JMC, Leal IR, Tabarelli M, editors. Caatinga. Cham: Springer; 2017. p. 133‑149. Available from: https:// doi. org/ 10. 1007/ 978‑3‑ 319‑ 68339‑3_5. 34. Geller J, Meyer C, Parker M, Hawk H. Redesign of PCR primers for mitochon‑ drial cytochrome c oxidase subunit I for marine invertebrates and applica‑ tion in all‑taxa biotic surveys. Mol Ecol Resour. 2013;13:851–61. 35. Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, Rubin DB. Bayesian data analysis. 3rd ed. London: CRC Press; 2014. 36. Gilliland HC. Hundreds of new and unusual insects discovered in the Amazon’s canopy. Available from: https:// www. natio nalge ograp hic. com/ magaz ine/ artic le/ hundr eds‑ of‑ new‑ and‑ unusu al‑ insec ts‑ disco vered‑ in‑ the‑ amazon‑ canopy‑ featu re. 2021. 37. Goodsell R, Tack A, Ronquist F, van Dijk L, Iwaszkiewicz‑Eggebrecht E, Miraldo A, et al. The rarity of Invertebrates prevents reliable application of IUCN Red‑List criteria. EcoEvoRxiv. 2024. Available from: https:// doi. org/ 10. 32942/ X23G71. 38. Gullefors B. Limes norrlandicus ‑ a natural biogeographical border for cad‑ disflies (Trichoptera) in Sweden. Ferrantia. 2008;55:61–5. 39. Hari V, Rakovec O, Markonis Y, Hanel M, Kumar R. Increased future occurrences of the exceptional 2018–2019 Central European drought under global warming. Sci Rep. 2020;10:12207. https:// doi. org/ 10. 1038/ s41598‑ 020‑ 68872‑9. 40. Hartop E, Srivathsan A, Ronquist F, Meier R. Towards large‑scale integrative taxonomy (LIT): resolving the data conundrum for dark taxa. Syst Biol. 2022. https:// doi. org/ 10. 1093/ sysbio/ syac0 33. 41. Hijmans R. raster: geographic data analysis and modeling. R package ver‑ sion 3.6–27, Available from: https:// rspat ial. org/ raster. 2024. 42. Karlsson D, Hartop E, Forshage M, Jaschhof M, Ronquist F. The Swedish Malaise Trap Project: A 15 Year Retrospective on a Countrywide Insect Inven‑ tory. Biodivers Data J. 2020;8:e47255. https:// doi. org/ 10. 3897/ BDJ.8. e47255. 43. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30:772–80. 44. Kitching RL, Bickel DJ, Boulter S. The evolutionary biology of flies. In: Guild analyses of dipteran assemblages: a rationale and investigation of seasonal‑ ity and stratification in selected rainforest faunas. New York: Columbia University Press; 2005. p. 388–415. 45. Leray M, Yang JY, Meyer CP, Mills SC, Agudelo N, Ranwez V, et al. A new versa‑ tile primer set targeting a short fragment of the mitochondrial COI region for metabarcoding metazoan diversity: application for characterizing coral reef fish gut contents. Front Zool. 2013;10:34. 46. Linnaeus C. Systema naturæ, sive regna tria naturæ systematice proposita per classes, ordines, genera, & species. 10th ed. Leiden: Lugduni Batavorum; 1758. 47. Lu Y, Yang Y, Sun B, Yuan J, Yu M, Stenseth NC, Bullock JM, Obersteiner M. Spa‑ tial variation in biodiversity loss across China under multiple environmental stressors. Sci Adv. 2020;6(47). https:// doi. org/ 10. 1126/ sciadv. abd09 52. 48. Lundström J, Schäfer M, Hesson J, Blomgren E, Lindström A, Wahlqvist P, Halling A, Hagelin A, Ahlm C, Evander M, Broman T, Forsman M, Persson Vinnersten T. The geographic distribution of mosquito species in Sweden. J Eur Mosq Control Assoc. 2013;31:35. 49. Mace GM, Barrett M, Burgess ND, Cornell SE, Freeman R, Grooten M, et al. Aiming higher to bend the curve of biodiversity loss. Nat Sustain. 2018;1:448–51. 50. McGlynn TP, Meineke EK, Bahlai CA, Li E, Hartop EA, Adams BJ, et al. Tem‑ perature accounts for the biodiversity of a hyperdiverse group of insects in urban Los Angeles. Proc R Soc B. 2019;286:20191020. 51. Meier R, Hartop E, Pylatiuk C, Srivathsan A. Towards holistic insect monitor‑ ing: species discovery, description, identification, and traits for all insects. Phil Trans R Soc B. 2024;379:20230120. https:// doi. org/ 10. 1098/ rstb. 2023‑ 0120. 52. Meier R, Shiyang K, Vaidya G, Ng PK. DNA barcoding and taxonomy in Diptera: a tale of high intraspecific variability and low identification success. Syst Biol. 2006;55:715–28. 53. Miller G. Linnaeus’s legacy carries on. Science. 2005;307:1038–9. 54. Naeem S, Chazdon R, Duffy JE, Prager C, Worm B. Biodiversity and human well‑being: an essential link for sustainable development. Proc R Soc B Biol Sci. 2016;283:20162091. 55. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan. Community ecology package. 2019;2:5–6. 56. Ollerton J. Pollinator diversity: distribution, ecological function, and conser‑ vation. Annu Rev Ecol Evol Syst. 2017;48:353–76. 57. Orr MC, Hughes AC, Chesters D, Pickering J, Zhu CD, Ascher JS. Global pat‑ terns and drivers of bee distribution. Curr Biol. 2020;30:843–8. 58. Ovaskainen O, Abrego N. Joint species distribution modelling. Cambridge: Cambridge University Press; 2020. p. 372. (Ecology, Biodiversity and Conser‑ vation). ISBN: 978‑1‑108‑71678‑9. eISBN: 9781108591720. 59. Ovaskainen O, Tikhonov G, Norberg A, Guillaume Blanchet F, Duan L, Dunson D, et al. How to make more out of community data? A conceptual framework and its implementation as models and software. Ecol Lett. 2017;20:561–76. 60. Rabone M, Wiethase JH, Simon‑Lledó E, Emery AM, Jones DOB, Dahlgren TG, Bribiesca‑Contreras G, Wiklund H, Horton T, Glover AG. How many metazoan species live in the world’s largest mineral exploration region? Curr Biol. 2023;33(12):2383–2396.e5. https:// doi. org/ 10. 1016/j. cub. 2023. 04. 052. 61. Ricklefs RE. Evolutionary diversification and the origin of the diversity–envi‑ ronment relationship. Ecology. 2006;87(S3–S13). https:// doi. org/ 10. 1890/ 0012‑ 9658(2006) 87[3: EDATOO] 2.0. CO;2. 62. Riksförbundet Svensk Trädgård. Zonkartan. 2018. Available from: http:// www. tradg ard. org/ svensk_ tradg ard/ zonka rtan. html. Cited 2024 May 28. 63. Ronquist F, Forshage M, Häggqvist S, Karlsson D, Hovmöller R, Bergsten J, et al. Completing Linnaeus’s inventory of the Swedish insect fauna: only 5,000 species left? PLOS One. 2020;15:e0228561. 64. Skvarla M, Larson J, Fisher R, Dowling A. A review of terrestrial and canopy Malaise traps. Ann Entomol Soc Am. 2020;114. https:// doi. org/ 10. 1093/ aesa/ saaa0 44. 65. SLU Artdatabanken. Famij: Phoridae ‑ puckelflugor. 2021. Available from: https:// www. dynta xa. se/ Taxon/ Info/ 20013 26. Cited 2024 May 28. 66. Srivathsan A, Ang Y, Heraty JM, Hwang WS, Jusoh WFA, Kutty SN, et al. Con‑ vergence of dominance and neglect in flying insect diversity. Nat Ecol Evol. 2023;7(7):1012–21. 67. Srivathsan A, Baloğlu B, Wang W, et al. A MinION™‑based pipeline for fast and cost‑effective DNA barcoding. Mol Ecol Resour. 2018;18:1035–49. https:// doi. org/ 10. 1111/ 1755‑ 0998. 12890. 68. Srivathsan A, Hartop E, Puniamoorthy J, Lee WT, Kutty SN, Kurina O, et al. Rapid, large‑scale species discovery in hyperdiverse taxa using 1D MinION sequencing. BMC Biol. 2019;17:96. 69. Srivathsan A, Hartop EA, Puniamoorthy J, Lee WT, Kutty SN, Kurina O, et al. MinION barcodes: biodiversity discovery and identification by everyone, for everyone. BMC Biol. 2021;19:217. 70. Svenning JC, Borchsenius F, Bjorholm S, Balslev H. High tropical net diver‑ sification drives the New World latitudinal gradient in palm (Arecaceae) species richness. J Biogeogr. 2008;35:394–406. 71. Tikhonov G, Opedal ØH, Abrego N, Lehikoinen A, Jonge MMJ, Oksanen J, et al. Joint species distribution modelling with the R‑package HMSC. Meth‑ ods Ecol Evol. 2020;11:442–7. 72. Tjur T. Coefficients of determination in logistic regression models ‑ a new proposal: the coefficient of discrimination. Am Stat. 2009;63:366–72. 73. Townes H. A light‑weight Malaise trap. Entomol News. 1972;83:239–47. 74. Truett GE, Heeger P, Mynatt RL, Truett AA, Walker JA, Warman ML. Prepara‑ tion of PCR‑quality mouse genomic DNA with hot sodium hydroxide and tris (HotSHOT). Biotechniques. 2000;29:52–4. 75. Wang WY, Srivathsan A, Foo M, Yamane S, Meier R. Sorting specimen‑rich invertebrate samples with cost‑effective NGS barcodes: validating a reverse workflow for specimen processing. Mol Ecol Resour. 2018;18:490. 76. Warton DI, Blanchet FG, O’Hara RB, Ovaskainen O, Taskinen S, Walker SC, et al. So many variables: joint modeling in community ecology. Trends Ecol Evol. 2015;30:766–79. 77. Wellenreuther M, Larson KW, Svensson EI. Climatic niche divergence or conservatism? Environmental niches and range limits in ecologically similar damselflies. Ecology. 2012;93:1353–66. 78. Wickham H. ggplot2: Elegant Graphics for Data Analysis [Internet]. New York: Springer‑Verlag; 2016. Available from: https:// ggplo t2. tidyv erse. org. 79. Wiens JJ. Global patterns of diversification and species richness in amphib‑ ians. Am Nat. 2007;170:S86–106. 80. Wolda H. Insect seasonality: why? Annu Rev Ecol Syst. 1988;19(1):1–18. Page 16 of 16 Hartopetal. BMC Biology (2024) 22:215 81. World Economic Forum. The Global Risks Report 2020 [Internet]. 15th ed. Geneva: World Economic Forum; 2020. [cited 2024 Sep 17]. Available from: https:// www. wefor um. org/ repor ts/ the‑ global‑ risks‑ report‑ 2020. 82. Yeo D, Srivathsan A, Meier R. Longer is not always better: optimizing barcode length for large‑scale species discovery and identification. Syst Biol. 2020;0:1–16. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations.