Ecology

Denlinger, D. L. Dormancy in tropical insects. Annu. Rev. Entomol. 31, 239–264. https://doi.org/10.1146/annurev.en.31.010186.001323 (1986).CAS 
Article 
PubMed 

Google Scholar 
Moreau, R. E. The breeding seasons of African birds—1. Land birds. Ibis 92, 223–267. https://doi.org/10.1111/j.1474-919X.1950.tb01750.x (1950).Article 

Google Scholar 
Fogden, M. P. L. Seasonality and population dynamics of equatorial forest birds in Sarawak. Ibis 114, 307–343. https://doi.org/10.1111/j.1474-919X.1972.tb00831.x (1972).Article 

Google Scholar 
Karr, J. R. Resource availability, and community diversity in tropical bird communities. Am. Nat. 110, 973–994. https://doi.org/10.1086/283121 (1976).Article 

Google Scholar 
Oppel, S. et al. The effects of rainfall on different components of seasonal fecundity in a tropical forest passerine. Ibis 155, 464–475. https://doi.org/10.1111/ibi.12052 (2013).Article 

Google Scholar 
Shaw, P. Rainfall, leafing phenology and sunrise time as potential Zeitgeber for the bimodal, dry season laying pattern of an African rain forest tit (Parus fasciiventer). J. Ornithol. 158, 263–275. https://doi.org/10.1007/s10336-016-1395-6 (2017).Article 

Google Scholar 
Withers, P. C. & Cooper, C. E. Metabolic depression: A historical perspective. In Aestivation: Molecular and Physiological Aspects, Progress in Molecular and Subcellular Biology (eds Navas, C. A. & Carvalho, J. E.) 1–23 (Springer-Verlag, 2010).
Google Scholar 
Fletcher, B. S., Pappas, S. & Kapatos, E. Changes in ovaries of olive fly (Dacus-oleae-(Gmelin)) during summer, and their relationship to temperature, humidity and fruit availability. Ecol. Entomol. 3, 99–107. https://doi.org/10.1111/j.1365-2311.1978.tb00908.x (1978).Article 

Google Scholar 
Braby, M. F. Reproductive seasonality in tropical satyrine butterflies—Strategies for the dry season. Ecol. Entomol. 20, 5–17. https://doi.org/10.1111/j.1365-2311.1995.tb00423.x (1995).Article 

Google Scholar 
Goehring, L. & Oberhauser, K. S. Effects of photoperiod, temperature, and host plant age on induction of reproductive diapause and development time in Danaus plexippus. Ecol. Entomol. 27, 674–685. https://doi.org/10.1046/j.1365-2311.2002.00454.x (2002).Article 

Google Scholar 
Valera, F., Casas-Crivillé, A. & Hoi, H. Interspecific parasite exchange in a mixed colony of birds. J. Parasitol. 89, 245–250. https://doi.org/10.1645/0022-3395(2003)089[0245:IPEIAM]2.0.CO;2 (2003).Article 
PubMed 

Google Scholar 
Grimaldi, D. The bird flies, genus Carnus: Species revision, generic relationships, and a fossil Meoneura in amber (Diptera: Carnidae). Am. Mus. Novit. 3190, 1–30 (1997).MathSciNet 

Google Scholar 
Valera, F., Casas-Crivillé, A. & Calero-Torralbo, M. A. Prolonged diapause in the ectoparasite Carnus hemapterus (Diptera: Cyclorrapha, Acalyptratae)—How frequent is it in parasites?. Parasitology 133, 179–186. https://doi.org/10.1017/S0031182006009899 (2006).CAS 
Article 
PubMed 

Google Scholar 
Sabrosky, C. W., Bennett, G. F. & Whitworth, T. L. Bird blow flies (Protocalliphora) in North America (Diptera: Calliphoridae), with notes on the Palearctic species. https://library.si.edu/digital-library/book/birdblowfliespro00sabr (Smithsonian Institute Press, 1989).Dodge, H. R. & Aitken, T. H. G. Philornis flies from Trinidad (Diptera: Muscidae). J. Kansas Entomol. Soc. 41, 134–154 (1968).
Google Scholar 
Couri, M. S. Notes and descriptions of Philornis flies (Diptera, Muscidae, Cyrtoneurinininae). Rev. Bras. Entomol. 28, 473–490 (1984).
Google Scholar 
Couri, M. S. Myiasis caused by obligatory parasites. Ia. Philornis Meinert (Muscidae). In Myiasis in Man and Animals in the Neotropical Region (eds Guimaraes, J. H. & Papavero, N.) 44–70 (Editora Pleiade, 1999).
Google Scholar 
Silvestri, L., Antoniazzi, L. R., Couri, M. S., Monje, L. D. & Beldomenico, P. M. First record of the avian ectoparasite Philornis downsi Dodge & Aitken, 1968 (Diptera: Muscidae) in Argentina. Syst. Parasitol. 80, 137–140. https://doi.org/10.1007/s11230-011-9314-y (2011).CAS 
Article 
PubMed 

Google Scholar 
Bulgarella, M. et al. Philornis downsi, an avian nest parasite invasive to the Galápagos Islands, in mainland Ecuador. Ann. Entomol. Soc. Am. 108, 242–250. https://doi.org/10.1093/aesa/sav026 (2015).Article 

Google Scholar 
Kleindorfer, S. & Dudaniec, R. Y. Host-parasite ecology, behavior and genetics: a review of the introduced fly parasite Philornis downsi and its Darwin finch hosts. BMC Zool. 1, 1. https://doi.org/10.1186/s40850-016-0003-9 (2016).Article 

Google Scholar 
Fessl, B., Heimpel, G. E. & Causton, C. E. Invasion of an avian nest parasite, Philornis downsi, to the Galapagos Islands: Colonization history, adaptations to novel ecosystems, and conservation challenges. In Disease Ecology: Social and Ecological Interactions in the Galapagos Islands (ed. Parker, P. G.) 213–266 (Springer, 2018). https://doi.org/10.1007/978-3-319-65909-1_9DO.Chapter 

Google Scholar 
McNew, S. M. & Clayton, D. H. Alien invasion: biology of Philornis flies highlighting Philornis downsi, an introduced parasite of Galapagos birds. Annu. Rev. Entomol. 63, 369–387. https://doi.org/10.1146/annurev-ento-020117-043103 (2018).CAS 
Article 
PubMed 

Google Scholar 
Anchundia, D. & Fessl, B. The conservation status of the Galapagos Martin, Progne modesta: Assessment of historical records and results of recent surveys. Bird Conserv. Int. 31, 129–138. https://doi.org/10.1017/S095927092000009X (2021).Article 

Google Scholar 
Coloma, A., Anchundia, D., Piedrahita, P., Pike, C. & Fessl, B. Observations on the nesting of the Galapagos Dove, Zenaida galapagoensis, in Galapagos, Ecuador. Galapagos Res. 69, 34–38 (2020).
Google Scholar 
Lack, D. Darwin’s Finches (Cambridge University Press, 1947).
Google Scholar 
Grant, P. R. Ecology and Evolution of Darwin’s Finches (Princeton University Press, 1986).
Google Scholar 
Bulgarella, M., Quiroga, M. A. & Heimpel, G. E. Additive negative effects of Philornis nest parasitism and small and declining Neotropical bird populations. Bird Conserv. Int. 29, 339–360. https://doi.org/10.1017/S0959270918000291 (2019).Article 

Google Scholar 
Causton, C. E. et al. Population dynamics of an invasive bird parasite, Philornis downsi (Diptera: Muscidae), in the Galapagos Islands. PLoS ONE 14(10), e0224125. https://doi.org/10.1371/journal.pone.0224125 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Hayes, E. J. & Wall, R. Age-grading adult insects: A review of techniques. Physiol. Entomol. 24, 1–10. https://doi.org/10.1046/j.1365-3032.1999.00104.x (1999).Article 

Google Scholar 
Lahuatte, P. F., Lincango, M. P., Heimpel, G. E. & Causton, C. E. Rearing larvae of the avian nest parasite, Philornis downsi (Diptera: Muscidae), on chicken blood-based diets. J. Insect Sci. 16, 84. https://doi.org/10.1093/jisesa/iew064 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Moon, R. D. & Krafsur, E. S. Pterin quantity and gonotrophic stage as indicators of age in Musca autumnalis (Diptera: Muscidae). J. Med. Entomol. 32, 673–684. https://doi.org/10.1093/jmedent/32.5.673 (1995).CAS 
Article 
PubMed 

Google Scholar 
Butler, S. M. et al. Characterization of age and cuticular hydrocarbon variation in mating pairs of house fly, Musca domestica, collected in the field. Med. Vet. Entomol. 23, 426–442. https://doi.org/10.1111/j.1365-2915.2009.00831.x (2009).CAS 
Article 
PubMed 

Google Scholar 
Mail, T. S. & Lehane, M. J. Characterisation of pigments in the head capsule of the adult stablefly Stomoxys calcitrans. Entomol. Exp. Appl. 46, 125–131. https://doi.org/10.1111/j.1570-7458.1988.tb01102.x (1988).Article 

Google Scholar 
Trueman, M. & D’Ozouville, N. Characterizing the Galapagos terrestrial climate in the face of global climate change. Galapagos Res. 67, 26–37 (2010).
Google Scholar 
Stramma, L. et al. Observed El Niño conditions in the eastern tropical Pacific in October 2015. Ocean Sci. 12, 861–873. https://doi.org/10.5194/os-12-861-2016 (2016).ADS 
Article 

Google Scholar 
Martin, N. J. et al. Seasonal and ENSO influences on the stable isotopic composition of Galapagos precipitation. J. Geophys. Res-Atmos. 123, 261–275. https://doi.org/10.1002/2017JD027380 (2018).ADS 
Article 

Google Scholar 
Grant, P. R., Grant, B. R., Keller, L. F. & Petren, K. Effects of El Niño events on Darwin’s finch productivity. Ecology 81, 2442–2457. https://doi.org/10.2307/177466 (2000).Article 

Google Scholar 
Sage, R. et al. Environmentally cued hatching in the bird parasite Philornis downsi (Diptera: Muscidae). Entomol. Exp. Appl. 166, 752–760. https://doi.org/10.1111/eea.12721 (2018).Article 

Google Scholar 
R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2021).Wood, S. N. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. Ser. B 73, 3–36. https://doi.org/10.1111/j.1467-9868.2010.00749.x (2011).MathSciNet 
Article 
MATH 

Google Scholar 
Lack, D. Breeding seasons in the Galapagos. Ibis 92, 268–278. https://doi.org/10.1111/j.1474-919X.1950.tb01751.x (1950).Article 

Google Scholar 
Grant, P. R. & Boag, P. T. Rainfall on the Galapagos and the demography of Darwin’s Finches. Auk 97, 227–244 (1980).Article 

Google Scholar 
Peck, S. B. Beetles of the Galápagos Islands, Ecuador: Evolution, Ecology, and Diversity (Insecta: Coleoptera) (NRC Research Press, 2006).
Google Scholar 
Grant, P. R. & Grant, B. R. The breeding and feeding characteristics of Darwin’s Finches on Isla Genovesa, Galapagos. Ecol. Monogr. 50, 381–410. https://doi.org/10.2307/2937257 (1980).Article 

Google Scholar 
Boag, P. T. & Grant, P. R. Darwin Finches (Geospiza) on Isla Daphne Major, Galapagos—Breeding and feeding ecology in a climatically variable environment. Ecol. Monogr. 54, 463–489. https://doi.org/10.2307/1942596 (1984).Article 

Google Scholar 
Schluter, D. Feeding correlates of breeding and social organization in two Galapagos Finches. Auk 101, 59–68. https://doi.org/10.1093/auk/101.1.59 (1984).Article 

Google Scholar 
Hau, M., Wikelski, M., Gwinner, H. & Gwinner, E. 2004 Timing of reproduction in a Darwin’s finch: Temporal opportunism under spatial constraints. Oikos 106, 489–500 (2004).Article 

Google Scholar 
Pike, C. L. et al. Behavior of the avian parasite Philornis downsi (Diptera: Muscidae) in and near host nests in the Galapagos Islands. J. Insect Behav. https://doi.org/10.1007/s10905-021-09789-7 (2021).Article 

Google Scholar 
Heleno, R. H., Olesen, J. M., Nogales, M., Vargas, P. & Traveset, A. Seed dispersal network in the Galapagos and the consequences of alien plant invasions. Proc. R. Soc. B. 280, 20122112. https://doi.org/10.1098/rspb.2012.2112 (2013).Article 
PubMed 
PubMed Central 

Google Scholar 
Traveset, A., Chamorro, S., Olesen, J. M. & Heleno, R. Space, time and aliens: charting the dynamic structure of Galapagos pollination networks. AoB PLANTS 7, plv068. https://doi.org/10.1093/aobpla/plv068 (2015).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Ervin, S. Nesting behavior of the large-billed flycatcher on Isla Santa Cruz. Noticias de Galapagos 51, 17–20 (1992).
Google Scholar 
Lincango, P. et al. Interactions between the avian parasite, Philornis downsi (Diptera: Muscidae) and the Galapagos Flycatcher, Myiarchus magnirostris Gould (Passeriformes: Tyrannidae). J. Wildl. Dis. 51, 907–910. https://doi.org/10.7589/2015-01-025 (2015).CAS 
Article 
PubMed 

Google Scholar 
Dudaniec, R. Y., Gardner, M. G., Donnellan, S. & Kleindorfer, S. Genetic variation in the invasive parasite, Philornis downsi (Diptera: Muscidae) on the Galapagos archipelago. BMC Ecol. 8, 13. https://doi.org/10.1186/1472-6785-8-13 (2008).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Cimadom, A. et al. Darwin’s finches treat their feathers with a natural repellent. Sci. Rep. 6, 34559. https://doi.org/10.1038/srep34559 (2016).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Common, L. K., Dudaniec, R. Y., Colombelli-Negrel, D. & Kleindorfer, S. Taxonomic shifts in Philornis larval behavior and rapid changes in Philornis downsi Dodge & Aitken (Diptera: Muscidae), an invasive avian parasite on the Galapagos Islands. InTech Open https://doi.org/10.5772/intechopen.88854 (2019).Article 

Google Scholar 
Common, L. K., O’Connor, J. A., Dudaniec, R. Y., Peters, K. J. & Kleindorfer, S. Evidence for rapid downward fecundity selection in an ectoparasite (Philornis downsi) with earlier host mortality in Darwin’s finches. J. Evol. Biol. 33, 524–533. https://doi.org/10.1111/jeb.13588 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Dieckhoff, C., Theobald, J. C., Waeckers, F. L. & Heimpel, G. E. Egg load dynamics and the risk of egg and time limitation experienced by an aphid parasitoid in the field. Ecol. Evol. 4, 1739–1750. https://doi.org/10.1002/ece3.1023 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
Papaj, D. R. Ovarian dynamics and host use. Annu. Rev. Entomol. 45, 423–448. https://doi.org/10.1146/annurev.ento.45.1.423 (2000).CAS 
Article 
PubMed 

Google Scholar 
Barton Browne, L., van Gerwen, A. C. M. & Williams, K. L. Oocyte resorption during ovarian development in the blowfly Lucilia cuprina. J. Insect Physiol. 25, 147–153. https://doi.org/10.1016/0022-1910(79)90093-3 (1979).Article 

Google Scholar 
Venkatesh, K. & Morrison, P. E. Some aspects of oogenesis in the stable fly Stomoxys calcitrans (Diptera, Muscidae). J. Insect Physiol. 26, 711–715. https://doi.org/10.1016/0022-1910(80)90045-1 (1980).Article 

Google Scholar 
Spradbery, J. P. & Schweizer, G. Oosorption during ovarian development in the screw-worm fly, Chrysomya bezziana. Entomol. Exp. Appl. 30, 209–214. https://doi.org/10.1111/j.1570-7458.1981.tb03102.x (1981).Article 

Google Scholar 
Curry, R. L. & Grant, P. R. Demography of the cooperatively breeding Galapagos mockingbird, Nesomimus parvulus, in a climatically variable environment. J. Anim. Ecol. 58, 441–463. https://doi.org/10.2307/4841 (1989).Article 

Google Scholar 
Calero-Torralbo, M. A. & Valera, F. Synchronization of host-parasite cycles by means of diapause: Host influence and parasite response to involuntary host shifting. Parasitol. 135, 1343–1352. https://doi.org/10.1017/S0031182008004885 (2008).CAS 
Article 

Google Scholar 
Larimore, R. W. Synchrony of cliff swallow nesting and development of the tick Ixodes baergi. Southwest. Nat. 32, 121–126 (1987).Article 

Google Scholar 
Bulgarella, M. & Heimpel, G. E. Host range and community structure of bird parasites in the genus Philornis (Diptera: Muscidae) on the Island of Trinidad. Ecol. Evol. 5, 3695–3703. https://doi.org/10.1002/ece3.1621 (2015).Article 
PubMed 
PubMed Central 

Google Scholar 
Common, L. K. et al. Avian vampire fly (Philornis downsi) mortality differs across Darwin’s finch host species. Sci. Rep. 11, 15832. https://doi.org/10.1038/s41598-021-94996-7 (2021).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Mendonca, E. & Couri, M. S. New associations between Philornis Meinert (Diptera, Muscidae) and Thamnophilidae (Aves, Passeriformes). Revta. Bras. Zool. 16, 1223–1225 (1999).Article 

Google Scholar 
Di Giacomo, A. G. Aves de la Reserva El Bagual. Historia natural y paisaje de la Reserva El Bagual, provincia de Formosa, Argentina. Inventario de la fauna de vertebrados y de la flora vascular de un área del Chaco Húmedo (eds. Di Giacomo, A. G. & Krapovickas, S. F.). Temas de Naturaleza y Conservación 4, 203–465 (Aves Argentinas/AOP, 2005).Koop, J. A. H., Causton, C. E., Bulgarella, M., Cooper, E. & Heimpel, G. E. Population structure of a nest parasite of Darwin’s finches within its native and invasive ranges. Conserv. Genet. 22, 11–22. https://doi.org/10.1007/s10592-020-01315-0 (2021).Article 

Google Scholar 
Kawecki, T. J. & Ebert, D. Conceptual issues in local adaptation. Ecol. Lett. 7, 1225–1241. https://doi.org/10.1111/j.1461-0248.2004.00684.x (2004).Article 

Google Scholar  More

Dall, S. R. X. & Johnstone, R. A. Managing uncertainty: Information and insurance under the risk of starvation. Philos. Trans. R. Soc. Lond. B 357, 1519–1526 (2002).
Article 

Google Scholar 
Balogh, A. C. V., Gamberale-Stille, G. & Leimar, O. Learning and the mimicry spectrum: from quasi-Bates to super-Müller. Anim. Behav. 76, 1591–1599 (2008).Article 

Google Scholar 
Barnett, C. A., Bateson, M. & Rowe, C. Better the devil you know: Avian predators find variation in prey toxicity aversive. Biol. Lett. 10, 20140533 (2014).Article 

Google Scholar 
Ruxton, G. D., Allen, W. L., Sherratt, T. N. & Speed, M. P. Avoiding Attack: The Evolutionary Ecology of Crypsis, Aposematism, and Mimicry 2nd edn. (Oxford University Press, 2018).Book 

Google Scholar 
Sherratt, T. N. State-dependent risk-taking by predators in systems with defended prey. Oikos 103, 93–100 (2003).Article 

Google Scholar 
Sherratt, T. N., Speed, M. S. & Ruxton, G. D. Natural selection on unpalatable species imposed by state-dependent foraging behaviour. J. Theor. Biol. 228, 217–226 (2004).MathSciNet 
Article 
ADS 

Google Scholar 
Gamberale-Stille, G. & Guilford, T. Automimicry destabilizes aposematism: Predator sample-and-reject behaviour may provide a solution. Proc. R. Soc. Lond. B 271, 2621–2625 (2004).Article 

Google Scholar 
Skelhorn, J. & Rowe, C. Avian predators taste-reject aposematic prey on the basis of their chemical defence. Biol. Lett. 2, 348–350 (2006).Article 

Google Scholar 
Skelhorn, J. & Rowe, C. Automimic frequency influences the foraging decisions of avian predators on aposematic prey. Anim. Behav. 74, 1563–1572 (2007).Article 

Google Scholar 
Brower, J. V. Z. Experimental studies of mimicry. IV. The reactions of starlings to different proportions of models and mimics. Am. Nat. 94, 271–282 (1960).Article 

Google Scholar 
Huheey, J. E. Studies in warning coloration and mimicry VIII. Further evidence for a frequency-dependent model of predation. J. Herpetol. 14, 223–230 (1980).Avery, M. L. Application of mimicry theory to bird damage control. J. Wildl. Manag. 49, 1116–1121 (1985).Article 

Google Scholar 
Nonacs, P. Foraging in a dynamic mimicry complex. Am. Nat. 126, 165–180 (1985).Article 

Google Scholar 
Rowland, H. M., Ihalainen, E., Lindström, L., Mappes, J. & Speed, M. P. Co-mimics have a mutualistic relationship despite unequal defences. Nature 448, 64–67 (2007).CAS 
Article 
ADS 

Google Scholar 
Skelhorn, J. & Rowe, C. Predators’ toxin burdens influence their strategic decisions to eat toxic prey. Curr. Biol. 17, 1479–1483 (2007).CAS 
Article 

Google Scholar 
Jones, R. S., Davis, S. C. & Speed, M. P. Defence cheats can degrade protection of chemically defended prey. Ethology 119, 52–57 (2013).Article 

Google Scholar 
Guilford, T. “Go-slow” signalling and the problem of automimicry. J. Theor. Biol. 170, 311–316 (1994).Article 
ADS 

Google Scholar 
Skelhorn, J. & Rowe, C. Taste-rejection by predators and the evolution of unpalatability in prey. Behav. Ecol. Sociobiol. 60, 550–555 (2006).Article 

Google Scholar 
Chatelain, M., Halpin, C. G. & Rowe, C. Ambient temperature influences birds’ decisions to eat toxic prey. Anim. Behav. 86, 733–740 (2013).CAS 
Article 

Google Scholar 
Peel, M. C., Finlayson, B. L. & McMahon, T. A. Updated world map of the Köppen-Geiger climate classification. Hydrol. Earth Syst. Sci. 11, 1633–1644 (2007).Article 
ADS 

Google Scholar 
Yamazaki, Y., Pagani-Núñez, E., Sota, T. & Barnett C. R. A. The truth is in the detail: predators attack aposematic prey less intensely than other prey types. Biol. J. Linn. Soc. 131, 332–343 (2020).Valkonnen, J. K. et al. Variation in predator species abundance can cause variable selection pressure on warning signalling prey. Ecol. Evol. 2, 1971–1976 (2011).Article 

Google Scholar 
Nokelainen, O., Valkonen, J., Lindstedt, C. & Mappes, J. Changes in predator community structure shifts the efficacy of two warning signals in Arctiid moths. J. Anim. Ecol. 83, 598–605 (2014).Article 

Google Scholar 
Bibby, C. J., Burgess, N. D., Hill, D. A. &. Mustoe S. H. Bird Census Techniques (2nd Edition). (Academic Press, London, 2000).Tsujimoto, D., Lin, C.-H., Kurihara, N. & Barnett, C. R. A. Citizen science in the class-room: the consistency of student collected data and its value in ecological hypothesis testing. Ornithological Sci. 18, 39–47 (2019).Article 

Google Scholar 
Bates, D., Maechler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Software 67, 1–48 (2015).Article 

Google Scholar 
Rainey, C. Dealing with separation in logistic regression models. Polit. Anal. 24, 339–355 (2016).Article 

Google Scholar 
Nakagawa, S. & Schielzeth, H. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Meth. Ecol. Evol. 4, 133–142 (2012).Article 

Google Scholar 
Hothorn, T,. Bretz, F. & Westfall, P. Simultaneous Inference in General Parametric Models. Biometrical. J. 50, 346–363 (2008).Paradis, E. & Schliep, K. Ape 5.0: An environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35, 526–528 (2019).Barnett, C. R. A., Ringhofer, M. & Suzuki, T. N. Differences in predatory behavior among three bird species when attacking chemically defended and undefended prey. J. Ethol. 39, 29–37 (2021).Article 

Google Scholar 
Carroll, J. & Sherratt, T. N. A direct comparison of the effectiveness of two anti-predator strategies under field conditions. J. Zool. 291, 279–285 (2013).Article 

Google Scholar 
Krebs, C. J. Ecological Methodology (2nd Edition). (Benjamin/Cummings, Menlo Park, CA, 1999).Oksanen, J. vegan: Community Ecology Package. (2020).R Development Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. URL http://www.Rproject.org (2017).Marples, N. M., Speed, M. P. & Thomas, R. J. An individual-based profitability spectrum for understanding interactions between predators and their prey. Biol. J. Linn. Soc. 125, 1–13 (2018).Article 

Google Scholar 
Boyden, T. C. Butterfly palatability and mimicry: experiments with anolis lizards. Evolution 30, 73–81 (1976).Article 

Google Scholar 
Järvi, T., Sillén-Tullberg, B. & Wiklund, C. The cost of being aposematic. An experimental study of predation on larvae of Papilio machaon by the Great Tit Parus major. Oikos 36, 267–272 (1981).Wiklund, C. & Järvi, T. Survival of distasteful insects after being attacked by naïve birds: a reappraisal of aposematic coloration evolving through individual selection. Evolution 36, 998–1002 (1982).Article 

Google Scholar 
Pinheiro, C. E. G. & Campos, V. C. Do rufous-tailed jacamars (Galbula ruficauda) play with aposematic butterflies. Ornitol. Neotrop. 24, 1–3 (2013).
Google Scholar 
Halpin, C. G. & Rowe, C. The effect of distastefulness and conspicuous coloration on post-attack rejection behaviour of predators and survival of prey. Biol. J. Linn. Soc. 120, 236–244 (2017).
Google Scholar 
Sillén-Tullberg, B. Higher survival of an aposematic than of a cryptic form of a distasteful bug. Oecologia 67, 411–415 (1985).Article 
ADS 

Google Scholar 
Fisher, R. A. The Genetical Theory of Natural Selection (Clarenden Press, 1930).Book 

Google Scholar 
Chai, P. Field observations and feeding experiments on the responses of rufous-tailed jacamars butterflies in a tropical rainforest. Biol. J. Linn. Soc. 29, 161–189 (1986).Article 

Google Scholar 
Wang, L.-Y., Huang, W.-S., Tang, H.-C., Huang, L.-C. & Lin, C.-P. Too hard to swallow: A secret secondary defence of an aposematic insect. J. Exp. Biol. 221, jeb172486 (2018).PubMed 

Google Scholar 
Summers, K., Speed, M. P., Blount, J. D. & Stuckert, A. M. M. Are aposematic signals honest? A review. J. Evol. Biol. 28, 1583–1599 (2015).CAS 
Article 

Google Scholar 
Holen, Ø. H. Disentangling taste and toxicity in aposematic prey. Proc. R. Soc. B 280, 20122588 (2013).Article 

Google Scholar 
Speed, M. P. & Franks, D. W. Antagonistic evolution in an aposematic predator-prey system. Evolution 68, 2996–3007 (2014).Article 

Google Scholar  More

Bulgarelli, D., Schlaeppi, K., Spaepen, S., Ver Loren van Themaat, E. & Schulze-Lefert, P. Structure and functions of the bacterial microbiota of plants. Annu. Rev. Plant Biol. 64, 807–838 (2013).CAS 
PubMed 

Google Scholar 
Leach, J. E., Triplett, L. R., Argueso, C. T. & Trivedi, P. Communication in the phytobiome. Cell 169, 587–596 (2017).CAS 
PubMed 

Google Scholar 
Vorholt, J. A., Vogel, C., Carlstrom, C. I. & Muller, D. B. Establishing causality: opportunities of synthetic communities for plant microbiome research. Cell Host Microbe 22, 142–155 (2017).CAS 
PubMed 

Google Scholar 
Jiang, Y. et al. Plant cultivars imprint the rhizosphere bacterial community composition and association networks. Soil Biol. Biochem. 109, 145–155 (2017).CAS 

Google Scholar 
Garbeva, P., van Elsas, J. D. & van Veen, J. A. Rhizosphere microbial community and its response to plant species and soil history. Plant Soil 302, 19–32 (2008).CAS 

Google Scholar 
Li, Y. et al. Rhizobacterial communities of five co-occurring desert halophytes. PeerJ 6, e5508 (2018).PubMed 
PubMed Central 

Google Scholar 
Matthews, A., Pierce, S., Hipperson, H. & Raymond, B. Rhizobacterial community assembly patterns vary between crop species. Front. Microbiol. 10, 581 (2019).PubMed 
PubMed Central 

Google Scholar 
Perez-Jaramillo, J. E., Mendes, R. & Raaijmakers, J. M. Impact of plant domestication on rhizosphere microbiome assembly and functions. Plant Mol. Biol. 90, 635–644 (2016).CAS 
PubMed 

Google Scholar 
Xu, J. et al. The structure and function of the global citrus rhizosphere microbiome. Nat. Commun. 9, 4894 (2018).ADS 
PubMed 
PubMed Central 

Google Scholar 
Trivedi, P., Leach, J. E., Tringe, S. G., Sa, T. & Singh, B. K. Plant-microbiome interactions: from community assembly to plant health. Nat. Rev. Microbiol. 18, 607–621 (2020).CAS 
PubMed 

Google Scholar 
Howard, M. M., Munoz, C. A., Kao-Kniffin, J. & Kessler, A. Soil microbiomes from fallow fields have species-specific effects on crop growth and pest resistance. Front. Plant Sci. 11, 1171 (2020).PubMed 
PubMed Central 

Google Scholar 
Yan, Y., Kuramae, E. E., de Hollander, M., Klinkhamer, P. G. & van Veen, J. A. Functional traits dominate the diversity-related selection of bacterial communities in the rhizosphere. ISME J. 11, 56–66 (2017).PubMed 

Google Scholar 
Bakker, P. A., Berendsen, R. L., Doornbos, R. F., Wintermans, P. C. & Pieterse, C. M. The rhizosphere revisited: root microbiomics. Front. Plant Sci. 4, 165 (2013).PubMed 
PubMed Central 

Google Scholar 
Lundberg, D. S. et al. Defining the core Arabidopsis thaliana root microbiome. Nature 488, 86–90 (2012).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Busby, P. E. et al. Research priorities for harnessing plant microbiomes in sustainable agriculture. PLoS Biol. 15, e2001793 (2017).PubMed 
PubMed Central 

Google Scholar 
de Vries, F. T., Griffiths, R. I., Knight, C. G., Nicolitch, O. & Williams, A. Harnessing rhizosphere microbiomes for drought-resilient crop production. Science 368, 270–274 (2020).ADS 
PubMed 

Google Scholar 
Hamonts, K. et al. Field study reveals core plant microbiota and relative importance of their drivers. Environ. Microbiol. 20, 124–140 (2018).CAS 
PubMed 

Google Scholar 
Xu, Q. et al. Long-term chemical-only fertilization induces a diversity decline and deep selection on the soil bacteria. mSystems 5, e00337–20 (2020).CAS 
PubMed 
PubMed Central 

Google Scholar 
Richter, A., Schöning, I., Kahl, T., Bauhus, J. & Ruess, L. Regional environmental conditions shape microbial community structure stronger than local forest management intensity. Ecol. Manag. 409, 250–259 (2018).
Google Scholar 
Bais, H. P., Weir, T. L., Perry, L. G., Gilroy, S. & Vivanco, J. M. The role of root exudates in rhizosphere interactions with plants and other organisms. Annu. Rev. Plant Biol. 57, 233–266 (2006).CAS 
PubMed 

Google Scholar 
Wallenstein, M. D. Managing and manipulating the rhizosphere microbiome for plant health: a systems approach. Rhizosphere 3, 230–232 (2017).
Google Scholar 
Kuzyakov, Y. & Xu, X. Competition between roots and microorganisms for nitrogen: mechanisms and ecological relevance. N. Phytol. 198, 656–669 (2013).CAS 

Google Scholar 
Roller, B. R., Stoddard, S. F. & Schmidt, T. M. Exploiting rRNA operon copy number to investigate bacterial reproductive strategies. Nat. Microbiol. 1, 16160 (2016).CAS 
PubMed 
PubMed Central 

Google Scholar 
Wu, L. et al. Microbial functional trait of rRNA operon copy numbers increases with organic levels in anaerobic digesters. ISME J. 11, 2874–2878 (2017).CAS 
PubMed 
PubMed Central 

Google Scholar 
Nuccio, E. E. et al. Niche differentiation is spatially and temporally regulated in the rhizosphere. ISME J. 14, 999–1014 (2020).CAS 
PubMed 
PubMed Central 

Google Scholar 
Fan, K., Weisenhorn, P., Gilbert, J. A. & Chu, H. Wheat rhizosphere harbors a less complex and more stable microbial co-occurrence pattern than bulk soil. Soil Biol. Biochem. 125, 251–260 (2018).CAS 

Google Scholar 
Fan, K. et al. Rhizosphere-associated bacterial network structure and spatial distribution differ significantly from bulk soil in wheat crop fields. Soil Biol. Biochem. 113, 275–284 (2017).CAS 

Google Scholar 
Peiffer, J. A. et al. Diversity and heritability of the maize rhizosphere microbiome under field conditions. Proc. Natl Acad. Sci. USA 110, 6548–6553 (2013).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Baudoin, E., Benizri, E. & Guckert, A. Impact of artificial root exudates on the bacterial community structure in bulk soil and maize rhizosphere. Soil Biol. Biochem. 35, 1183–1192 (2003).CAS 

Google Scholar 
Kuzyakov, Y. & Razavi, B. S. Rhizosphere size and shape: temporal dynamics and spatial stationarity. Soil Biol. Biochem. 135, 343–360 (2019).CAS 

Google Scholar 
Ren, Y. et al. Functional compensation dominates the assembly of plant rhizospheric bacterial community. Soil Biol. Biochem. 150, 107968 (2020).CAS 

Google Scholar 
Chen, Y. et al. Organic amendments shift the phosphorus-correlated microbial co-occurrence pattern in the peanut rhizosphere network during long-term fertilization regimes. Appl. Soil Ecol. 124, 229–239 (2018).ADS 

Google Scholar 
Atulba, S. L. et al. Evaluation of rice root oxidizing potential using digital image analysis. J. Korean Soc. Appl. Bi 58, 463–471 (2015).CAS 

Google Scholar 
Schmidt, H., Eickhorst, T. & Tippkötter, R. Monitoring of root growth and redox conditions in paddy soil rhizotrons by redox electrodes and image analysis. Plant Soil 341, 221–232 (2011).CAS 

Google Scholar 
Pausch, J., Zhu, B., Kuzyakov, Y. & Cheng, W. Plant inter-species effects on rhizosphere priming of soil organic matter decomposition. Soil Biol. Biochem. 57, 91–99 (2013).CAS 

Google Scholar 
Finn, D., Kopittke, P. M., Dennis, P. G. & Dalal, R. C. Microbial energy and matter transformation in agricultural soils. Soil Biol. Biochem. 111, 176–192 (2017).CAS 

Google Scholar 
Jones, R. T. et al. A comprehensive survey of soil acidobacterial diversity using pyrosequencing and clone library analyses. ISME J. 3, 442–453 (2009).CAS 
PubMed 

Google Scholar 
Zhao, S. et al. Biogeographical distribution of bacterial communities in saline agricultural soil. Geoderma 361, 114095 (2020).ADS 
CAS 

Google Scholar 
Eiler, A., Heinrich, F. & Bertilsson, S. Coherent dynamics and association networks among lake bacterioplankton taxa. ISME J. 6, 330–342 (2012).CAS 
PubMed 

Google Scholar 
Zhou, J. et al. Generation of arbitrary two-point correlated directed networks with given modularity. Phys. Lett. A 374, 3129–3135 (2010).ADS 
CAS 
MATH 

Google Scholar 
Herron, P. M., Gage, D. J., Arango Pinedo, C., Haider, Z. K. & Cardon, Z. G. Better to light a candle than curse the darkness: illuminating spatial localization and temporal dynamics of rapid microbial growth in the rhizosphere. Front. Plant Sci. 4, 323 (2013).PubMed 
PubMed Central 

Google Scholar 
Blagodatskaya, E., Blagodatsky, S., Anderson, T. H. & Kuzyakov, Y. Microbial growth and carbon use efficiency in the rhizosphere and root-free soil. PLoS ONE 9, e93282 (2014).ADS 
PubMed 
PubMed Central 

Google Scholar 
Berendsen, R. L., Pieterse, C. M. & Bakker, P. A. The rhizosphere microbiome and plant health. Trends Plant Sci. 17, 478–486 (2012).CAS 
PubMed 

Google Scholar 
Mendes, L. W., Kuramae, E. E., Navarrete, A. A., van Veen, J. A. & Tsai, S. M. Taxonomical and functional microbial community selection in soybean rhizosphere. ISME J. 8, 1577–1587 (2014).CAS 
PubMed 
PubMed Central 

Google Scholar 
Hinsinger, P. Bioavailability of soil inorganic P in the rhizosphere as affected by root-induced chemical changes: a review. Plant Soil 237, 173–195 (2001).CAS 

Google Scholar 
Kuzyakov, Y. & Blagodatskaya, E. Microbial hotspots and hot moments in soil: Concept & review. Soil Biol. Biochem. 83, 184–199 (2015).CAS 

Google Scholar 
Loeppmann, S., Blagodatskaya, E., Pausch, J. & Kuzyakov, Y. Substrate quality affects kinetics and catalytic efficiency of exo-enzymes in rhizosphere and detritusphere. Soil Biol. Biochem. 92, 111–118 (2016).CAS 

Google Scholar 
Ma, X. et al. Spatial patterns of enzyme activities in the rhizosphere: Effects of root hairs and root radius. Soil Biol. Biochem. 118, 69–78 (2018).CAS 

Google Scholar 
Kroener, E., Zarebanadkouki, M., Kaestner, A. & Carminati, A. Nonequilibrium water dynamics in the rhizosphere: How mucilage affects water flow in soils. Water Resour. Res. 50, 6479–6495 (2014).ADS 

Google Scholar 
Carminati, A. Rhizosphere wettability decreases with root age: a problem or a strategy to increase water uptake of young roots? Front. Plant Sci. 4, 298 (2013).PubMed 
PubMed Central 

Google Scholar 
Holz, M., Zarebanadkouki, M., Kaestner, A., Kuzyakov, Y. & Carminati, A. Rhizodeposition under drought is controlled by root growth rate and rhizosphere water content. Plant Soil 423, 429–442 (2018).CAS 

Google Scholar 
Tripathi, B. M. et al. Trends in taxonomic and functional composition of soil microbiome along a precipitation gradient in Israel. Microb. Ecol. 74, 168–176 (2017).PubMed 

Google Scholar 
Harms, A., Brodersen, D. E., Mitarai, N. & Gerdes, K. Toxins, targets, and triggers: an overview of toxin-antitoxin biology. Mol. Cell 70, 768–784 (2018).CAS 
PubMed 

Google Scholar 
Kearns, P. J. & Shade, A. Trait-based patterns of microbial dynamics in dormancy potential and heterotrophic strategy: case studies of resource-based and post-press succession. ISME J. 12, 2575–2581 (2018).CAS 
PubMed 
PubMed Central 

Google Scholar 
Klappenbach, J. A., Dunbar, J. M. & Schmidt, T. M. rRNA operon copy number reflects ecological strategies of bacteria. Appl. Environ. Microb. 66, 1328–1333 (2000).ADS 
CAS 

Google Scholar 
Schoeps, R. et al. Land-use intensity rather than plant functional identity shapes bacterial and fungal rhizosphere communities. Front. Micro. 9, 2711 (2018).
Google Scholar 
Nemergut, D. R. et al. Decreases in average bacterial community rRNA operon copy number during succession. ISME J. 10, 1147–1156 (2016).CAS 
PubMed 

Google Scholar 
Cui, J. et al. Carbon and nitrogen recycling from microbial necromass to cope with C:N stoichiometric imbalance by priming. Soil Biol. Biochem. 142, 107720 (2020).CAS 

Google Scholar 
Blagodatskaya, E. V., Blagodatsky, S. A., Anderson, T. H. & Kuzyakov, Y. Priming effects in chernozem induced by glucose and N in relation to microbial growth strategies. Appl. Soil Ecol. 37, 95–105 (2007).
Google Scholar 
Lecomte, S. M. et al. Diversifying anaerobic respiration strategies to compete in the rhizosphere. Front. Environ. Sci. 6, 139 (2018).
Google Scholar 
Herz, K. et al. Drivers of intraspecific trait variation of grass and forb species in German meadows and pastures. J. Veg. Sci. 28, 705–716 (2017).
Google Scholar 
Ravenek, J. M. et al. Linking root traits and competitive success in grassland species. Plant Soil 407, 39–53 (2016).CAS 

Google Scholar 
Larsen, J., Jaramillo-López, P., Nájera-Rincon, M. & González-Esquivel, C. Biotic interactions in the rhizosphere in relation to plant and soil nutrient dynamics. J. Soil Sci. Plant Nutr. 15, 449–463 (2015).
Google Scholar 
Raaijmakers, J. M., Paulitz, T. C., Steinberg, C., Alabouvette, C. & Moënne-Loccoz, Y. The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil 321, 341–361 (2009).CAS 

Google Scholar 
Ma, H.-K. et al. Steering root microbiomes of a commercial horticultural crop with plant-soil feedbacks. Appl. Soil Ecol. 150, 103468 (2020).
Google Scholar 
Hannula, S. E. et al. Persistence of plant-mediated microbial soil legacy effects in soil and inside roots. Nat. Commun 12, 5686 (2021).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581–583 (2016).CAS 
PubMed 
PubMed Central 

Google Scholar 
Hill, T. C., Walsh, K. A., Harris, J. A. & Moffett, B. F. Using ecological diversity measures with bacterial communities. FEMS Microbiol. Ecol. 43, 1–11 (2003).CAS 
PubMed 

Google Scholar 
Lima-Mendez, G. et al. Determinants of community structure in the global plankton interactome. Science 348, 6237 (2015).
Google Scholar 
Noble, W. S. How does multiple testing correction work? Nat. Biotechnol. 27, 1135–1137 (2009).CAS 
PubMed 
PubMed Central 

Google Scholar 
Luo, F., Zhong, J., Yang, Y., Scheuermann, R. H. & Zhou, J. Application of random matrix theory to biological networks. Phys. Lett. A 357, 420–423 (2006).ADS 
CAS 

Google Scholar 
Shannon, P. et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498–2504 (2003).CAS 
PubMed 
PubMed Central 

Google Scholar 
Bastian, M., Heymann, S. & Jacomy, M. Gephi: an open source software for exploring and manipulating networks. ICWSM 8, 361–362 (2009).
Google Scholar 
Peng, G. S. & Wu, J. Optimal network topology for structural robustness based on natural connectivity. Phys. A 443, 212–220 (2016).MathSciNet 

Google Scholar 
Ruan, Y., Wang, T., Guo, S., Ling, N. & Shen, Q. Plant grafting shapes complexity and co-occurrence of rhizobacterial assemblages. Microb. Ecol. 80, 643–655 (2020).CAS 
PubMed 

Google Scholar 
Newman, M. E. Modularity and community structure in networks. Proc. Natl Acad. Sci. USA 103, 8577–8582 (2006).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Deng, Y. et al. Molecular ecological network analyses. BMC Bioinforma. 13, 113 (2012).
Google Scholar 
Guimerà, R. & Nunes Amaral, L. A. Functional cartography of complex metabolic networks. Nature 433, 895–900 (2005).ADS 
PubMed 
PubMed Central 

Google Scholar 
Olesen, J. M., Bascompte, J., Dupont, Y. L. & Jordano, P. The modularity of pollination networks. Proc. Natl Acad. Sci. USA 104, 19891 (2007).ADS 
CAS 
PubMed 
PubMed Central 
MATH 

Google Scholar 
Ling, N. et al. Insight into how organic amendments can shape the soil microbiome in long-term field experiments as revealed by network analysis. Soil Biol. Biochem. 99, 137–149 (2016).CAS 

Google Scholar 
Louca, S., Parfrey Laura, W. & Doebeli, M. Decoupling function and taxonomy in the global ocean microbiome. Science 353, 1272–1277 (2016).ADS 
CAS 
PubMed 

Google Scholar 
Douglas, G. M. et al. PICRUSt2 for prediction of metagenome functions. Nat. Biotechnol. 38, 685–688 (2020).CAS 
PubMed 
PubMed Central 

Google Scholar 
Lennon, J. T. & Jones, S. E. Microbial seed banks: the ecological and evolutionary implications of dormancy. Nat. Rev. Microbiol. 9, 119–130 (2011).CAS 
PubMed 

Google Scholar 
Hedges, L. V., Gurevitch, J. & Curtis, P. S. The meta-analysis of response ratios in experimental ecology. Ecology 80, 1150–1156 (1999).
Google Scholar 
Rosenberg, M. S., Adams, D. C. & Gurevitch, J. MetaWin: Statistical software for meta-analysis. Version 2.0. Sinauer (2000).Viechtbauer, W. Conducting meta-analyses in R with the metafor Package. J. Stat. Softw. 36, 1–48 (2010).
Google Scholar 
Egger, M., Smith, G. D., Schneider, M. & Minder, C. Bias in meta-analysis detected by a simple, graphical test. BMJ 315, 629 (1997).CAS 
PubMed 
PubMed Central 

Google Scholar 
Calcagno, V. & de Mazancourt, C. glmulti: an R package for easy automated model selection with (generalized) linear models. J. Stat. Softw. 34, 1–29 (2010).
Google Scholar 
Gloor, G. B., Macklaim, J. M., Pawlowsky-Glahn, V. & Egozcue, J. J. Microbiome datasets are compositional: and this is not optional. Front. Microbiol. 8, 2224 (2017).PubMed 
PubMed Central 

Google Scholar 
Edgar, R. C. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32, 1792–1797 (2004).CAS 
PubMed 
PubMed Central 

Google Scholar 
Letunic, I. & Bork, P. Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Res. 44, W242–W245 (2016).CAS 
PubMed 
PubMed Central 

Google Scholar  More

Ant experimentsAnt coloniesSeven colonies of the garden ant L. niger collected from the Soka University and a nearby park were used in this study (Extended Data Table S1). They were placed in plastic cases (35 × 25 × 6 cm). Water was provided ad libitum. They were fed a sucrose solution, and were starved for 2–5 days before the start of the experiment. The colonies were queen-less colonies with 200–700 workers. Aqueous sucrose solution was used as a food resource (bait) in the experiments. The laboratory room where the experiments were performed and the ant colonies were kept was maintained at a temperature of 25–27 °C and a humidity of 60–70%. Artificial lights were also installed in this room.ApparatusWe used an apparatus, called “the main apparatus,” with two paths from the nest to the feeding site (length: 30 cm, width: 2 cm, height: 12 cm for the outward path and 15 cm for the return path) (Fig. 1). This apparatus could separate the outward path (bridge) from the inward path (bridge). Here, the outward path refers to that taken by ants from the nest to the feeding site, whereas the inward path refers to the path taken by ants from the feeding site to the nest.Figure 1The main apparatus used in the three experiments (the main experiment and the comparison experiments 1 and 2). Nests are connected to the experimental apparatus by a slope. In the main experiment, on the outward path, there is ant traffic from the nest to the feeding site on a pheromone trail, and on the inward path, there is ant traffic from the feeding site to the nest on a pheromone trail. In the comparison experiment 1, only a pheromone trail is present on both the outward and inward paths. In the comparison experiment 2, no pheromone trail or ant traffic is present on both the outward and inward paths.Full size imageTwo important features of this apparatus were as follows: firstly, it allowed ants to only enter the outward path from the nest. A rat-guard structure at the end of the inward path prevented the ants on the outward path from entering the inward path (Extended Data Fig. S1A). Secondly, we installed a vertical structure at the end of the outward path (height: 4 cm). After climbing the vertical structure, ants were not allowed to return to the outward path (Extended Data Fig. S1B). Moreover, we installed partitions on the feeding site, which also prevented ants from returning to the outward path after reaching the feeding site (Extended Data Fig. S1C). After entering the feeding site, ants had to pass through a narrow gap (width: 0.5 cm) created by the partition. No visual cues were offered as the apparatus was surrounded on all four sides by plastic walls.In this experiment, we made another apparatus for a single ant (target ant), which would be joining the ant trail on the main bridges (Extended Data Fig. S2). This apparatus, called “the confluence device,” was a detachable device that could be connected at right angles to the outward and inward bridges of the main apparatus. To connect this device to the outward bridge, we made the confluence path of this device under the inward bridge of the main apparatus, since the outward bridge was lower than the inward bridge. Thus, we made a slope on the outward confluence path connected to the outward bridge of the main apparatus. Further, because placing the ants directly on the sidewalk sometimes caused them to fall off the sidewalk owing to panic, we constructed a free space and a wall (height: 5 cm) in the middle of the confluence device on which the ants were placed calmly. Owing to this modification, we could let each target ant calm down and then access the main bridge whenever they wanted to. The apparatus used in this experiment was made of white plastic plates.Pheromone trail with ant trafficThis main experiment was limited to once a day for each colony. A sucrose solution was dripped into the feeding site. Target ants, which were walking on a plastic case as foragers, had been moved from their nests to another case immediately before a trail of (nontarget) ants was formed. Thus, dozens of ants were moved in advance to the case to be used as target ants. Subsequently, a trail of (nontarget) ants was formed from the nest to the main apparatus. Considering that it took some time for the ants that had finished foraging and returned to the nest to recruit their mates, the ants were left for approximately 40 min to an hour until a permanent ant trail was formed. It was difficult to form an ant trail immediately after the start of the experiment since no foraging pheromones could be produced in the first foraging trip on the outward path and since experienced foraging ants may make foraging pheromones on the outward path2,21,22. The target ants were allowed to enter bridges of the main apparatus after the establishment of a permanent ant trail. At that time, trails of individual target ants were started. Target ants were allowed to join at right angles to the path on the apparatus, one by one from the confluence device. Individual target ants were allowed to enter the main apparatus at four different points: (1) Left-Left (LL), located at the left side of the center of the outward path. The outward path was on the left side, whereas the inward path was on the right side for the experimenter when seen from the nest. (2) Left–Right (LR), located at the right side of the center of the outward path. (3) and (4) Right-Left (RL) and Right-Right (RR), located at the left and right sides of the center of the inward path, respectively (Fig. 2). We had set these four points to check if target ants tended to turn their body to a certain direction when entering the main bridges, regardless of the movement direction of the other ants. A video camera (Panasonic, AVCHD 30fps) was used to record the migration of ants to the feeding site or nest. Videos were taken from above, and target ants were used only once.Figure 2Four joining points (LL, LR, RL, and RR) and the confluence device (joining device). The confluence (joining) device was connected at right angles to the center of the outward and inward bridges of the main apparatus. Here, the LR version is shown as an example.Full size imageThe goal lines were set at 15 cm from the center of the main paths. We checked the side (nest side or feeding site side) from which a target ant passed the goal line.Pheromone trail with no ant trafficThis comparison experiment 1 was limited to once a day for each colony. Dozens of ants were moved in advance to another case to be used as target ants in a similar manner to the main experiment. The (nontarget) ants were left for about 40 min to an hour until a permanent ant trail was formed. Subsequently, we removed all the ants from the device. Then, target ants were allowed to enter on the side path one by one. In this case, we left the bait in place to control this experiment under the same condition as the main experiment. As the pheromone trail was created on the outward path as well as on the inward path, it was the only decision-making factor for the ants to join at the main path (outward/inward paths). We checked the side (nest side or feeding site side) from which a target ant passed the goal line in a similar manner to the main experiment.No pheromone trail or ant trafficThis comparison experiment 2 was limited to once a day for each colony. Dozens of ants were moved in advance to another case to be used as target ants in a similar manner to the main experiment. This experiment was conducted to investigate ant behavior under the following two conditions: (1) no ant trails and (2) no pheromones trails. The bait was in place in the same manner. We checked the side (nest side or feeding site side) from which a target ant passed the goal line in a similar manner to the main experiment. After each trial (the target ant passed the goal line), we wiped the apparatus with ethanol solution before the next target ant was allowed to enter the main paths.AnalysisThe goal lines were set at 15 cm from the center of the main paths. We checked which goal side the target ants reached the goal line on each trial. A reverse run referred to the goal to the nest on the outward path and the goal to the feeding site on the inward path. A normal run referred to the goal to the feeding site on the outward path and the goal to the nest on the inward path.In some cases of the main experiment, foraging (nontarget) ants that could not reach the feeding site on their outward path or could not return to the nest on their inward path would be against the ant flows. On the outward path, we considered that the ants conducted a “reverse flow” if the position of their heads was on the nest side compared with the position of their stomach. If not, we defined that the ants conducted a “normal flow” (Extended Data Fig. S3). On the inward path, we defined that the ants conducted a “reverse flow” if the position of their head was on the feeding site side compared with the position of their stomach. If not, we defined that the ants conducted a “normal flow” (Extended Data Fig. S3). We focused on the target ants that came in contact with ants with normal flow. Therefore, if an ant with reverse flow was located within 10 cm of the target ant, that trial was excluded from the analysis.Furthermore, we also evaluated if target ants coming in contact with foraging (nontarget) ants immediately after entering the trail would affect the goal choice. Therefore, we conducted an analysis focusing on the contact using the data from the main experiment. We examined whether or not the target ant made contact with other foraging ants until it passed a point 2 cm from the center of the path. As already mentioned, if the target ant came in contact with another ant moving against the normal flow of the ant trail, this contact was excluded from the counts. Moreover, we also excluded cases in which the body of target ants was on a point 2 cm from the center of the path by visual evaluation. Thus, we examined the goal choice of target ants by focusing on whether or not they came in contact with other ants immediately after joining the main bridges.We also conducted a preliminary experiment using a single path apparatus to investigate bi-directional trail behaviour. Please see the Extended Data File S1.Model descriptionThe models were coded using the C programming language. The model description follows the Overview, Design concepts, and Details protocol23,24.PurposeThe purpose of the model was to examine the mechanistic understanding of our findings. We adopted an action of target agents obtained from our ant experiments and compared it with another action of target agents on a trail that was contrary to the fact. To be more precise, target agents were allowed to obey an alignment rule in which they tended to move in the same direction with other agents. We named the former model as the reverse-rule model and the latter model as the alignment-rule model. By doing so, we could find the significance of our findings from ant experiments.Entities, state variables, and scalesWe developed two different models (reverse-rule model and alignment-rule model) that included two types of entities: agents and cells. The agent has the state variable Navigational state, which has two values: Navigational state = {wandering, foraging}. The cell has the state variable Pheromone; this value represents the amount of pheromones in each cell. We used a 2D lattice field and set a straight bridge with 61 cells × 5 cell sizes. We also set goal lines at x-coordinate =  − 30 and 30. If the agents reached coordinates satisfying their x-coordinate =  − 30 or 30, they were removed from the system. If the agents reached y-axis boundaries, their movement direction was restricted. Each trial continued until the target agent reached one of the two goal lines. However, trials were forcibly finished if the target agent never reached any goal line by t = 500-time steps. In total, we conducted 1000 trials.Process overview and schedulingAt the beginning of each trial, an artificial target ant (Navigational state = wandering) was introduced at the center of an artificial simulation field. Foraging agents (Navigational state = foraging) were randomly distributed on the simulation field in advance.Agents on the simulation field selected one direction from two directions (+ x and − x) on each time step and updated their positions. Briefly, an agent at coordinate (x, y) selected one direction from two directions (+ x and − x) and updated its position with one of the three coordinates—(x − 1, y), (x − 1, y + 1), or (x − 1, y − 1)—if it selected the − x direction, or—(x + 1, y), (x + 1, y + 1), or (x + 1, y − 1)—if it selected the + x direction by scanning pheromones on these three coordinates. For example, if an agent at coordinate (0, 2) decided to move in + x direction at one time, the position of this agent was replaced with one of (1, 3), (1, 2) and (1, 1) from (0, 2) by scanning pheromones on these three coordinates. The target agent selected the − x/ + x direction with equal probability on each time step until it met the foragers. In contrast, foraging agents tended to decide to move in the − x direction on each time step with a high probability and therefore they tended to select the − x direction for position updating. Foraging agents deposited pheromones before leaving the current cell (see submodel entitled “Position updating” and submodel entitled “Pheromone updating”). In contrast, the target agents did not deposit pheromones.Using above submodels, artificial ants sometimes met other agents. If the target agent (Navigational state = wandering) met the foragers (Navigational state = foraging), the target agent tended to select one direction from two directions (+ x and − x) on each time step thereafter with a high probability, which was dependent on which direction the met foragers came from. More strictly, in the reverse-rule model, the target agent tended to move in an opposite direction from the foragers if it met the foragers coming from the opposite direction. On the contrary, the target agent in the alignment-rule model tended to move in the same direction with foragers if it met the foragers moving in the same direction (see submodel entitled “The interaction between the target agent and foragers”). For example, in the reverse-rule model, if the target agent at coordinate (x, y), whose previous coordinate was (x − 1, y), met the forager coming from the opposite direction, whose previous coordinate was (x + 1, y), the target agent decided to move in + x direction on each time step thereafter with a high probability until similar events occurred.Design conceptThe mean goal time was the emergent property of the model. Sensing was important as the agents scanned the pheromone concentrations. Stochasticity was used to determine in which direction the agent moved and to select one cell using the pheromone concentrations.InitializationWe set a single agent (target agent) on the coordinate (0, 2) and its Navigational state was set to wandering (Extended Data Fig. S5A). We also set N foraging agents on the bridge whose Navigational state was set to foraging. Therefore, N + 1 agents were on the test field at the beginning of each trial. A target agent was the agent k = 0, whereas foraging agents were agents k = 1, 2, …, N. These foragers were randomly distributed on the bridge. Thus, x(k) (in) {n |− 30 ≤ n ≤ 30, n is an integer} and y(k) (in) {n | 0 ≤ n ≤ 4, n is an integer} for k  > 0.Foraging agents were set to move in the -x direction (Direction(k) for k  > 0 = − x). On the other hand, the target agent randomly chose one direction from two directions at the beginning of each trial (Direction(0) was set to + x or − x with equal probability). Herein, Direction(k) can be − x or + x, which implies bias in the movement direction. The parameter prob(k) indicates the probability of moving in Direction(k). The target agent selected the − x/+ x direction with equal probability on each time step until it met the foragers. Therefore, the parameter prob was set to 0.50 for the target ant (prob(0) = 0.50), whereas prob was set to 0.80 for foraging agents (prob(k) = 0.80 for k  > 0). The amount of pheromones on each cell was set to 1 at the beginning of each trial (pheromone(x, y) = 1) and the pheromone evaporation rate q was set to 0.99.The model descriptions are explained using submodels. A Submodel: the interaction between the target agent and foragers causes differences between two rules (the reverse-rule model and the alignment-rule model).SubmodelsSubmodel: the interaction between the target agent and foragersThe parameters Direction(0) and prob(0) were replaced with new ones whenever the following events occurred.In the reverse-rule model, for any agent k (k  > 0),Herein, (xt(k), yt(k)) indicates the x–y-coordinate for the agent k at time t. Furthermore, (xt(0), yt(0)) = (xt(k), yt(k)) means that the target agent and the agent k occupy the same cell at time t while (xt(0) − xt−1(0)) × (xt(k) − xt−1(k)) =  − 1 indicates that the target agent meets the agent k came from the opposite direction. The target agent replaces Direction(0) with an opposite direction from the forager k (see Extended data Fig. S5B).In the alignment-rule model, for any agent k (k  > 0),(xt(0) − xt−1(0)) × (xt(k) − xt−1(k)) = 1 indicates that the target agent meets the agent k came from the same direction. The target agent replaces Direction(0) with a same direction with the forager k (See Extended Data Fig. S5B).In the reverse-rule model, these events are driven from the experimental observations of real ants. Target ants appear to move against the trail and seem to move straight by contacting those other nestmates that come from the opposite direction. Also, target ants seem to select the reverse goal even if physical contact with ant nestmates does not occur immediately after entering the bridge. So, regarding parameter replacements, we did not consider the position at which the target agent met another agent. Note that foraging agents did not change these parameters until the end of each trial. Further, Direction(0) can be replaced with − x from + x and vice versa whenever the target agent meets foragers that come from the opposite direction.In the alignment-rule model, the target agent tends to move in the same direction with other agents. This is contrary to the experimental observations of real ants.Submodel: position updatingFor all k agents (k = 0–N), the movement direction and position updates are shown as follows (Extended Data Fig. S5C);Here, rnt(k) indicates a random number. Thus, rnt(k) (in) [0.00, 1.00].Prob(0) for the target agent is initially set to 0.50. Therefore, the target agent selects one direction from the two (− x and + x) on each time step randomly before the condition described in submodels—the interaction between the target agent and foragers is satisfied. On the other hand, foraging agents select − x direction with a high probability (= Prob(k)) on each time step. After selecting one direction from two (− x and + x), agents scan three cells in the direction of movement. Using pheromone concentrations on those three cells, they update their positions.If agents reach coordinates satisfying their y-coordinate = 4 or 0, those agents update their position by selecting not three but two coordinates since they are located on the edges of the bridge.Submodel: pheromone updatingForaging agents (k  > 0) deposited pheromones on the current cell when leaving that cell.Then, pheromones are evaporated using the evaporation rate q.For each time iteration, these submodels operated in the following order.STEP 1: The interaction between the target agent and foragers.STEP 2: Position updating.STEP 3: Pheromone updating.AnalysisTo check the accuracy of our model, we counted which goal side the target agent entered the goal line from using the reverse-rule model by setting N = 9. If the target agent passed the goal line at x-coordinate =  − 30 (30), we considered that it reached the normal (reverse) goal. Note that trials in which the target agent never reached any goal lines by t = 500 were excluded from this analysis. Furthermore, to investigate the adaptability of the reverse run mechanism, we examined the time until the target agent reached the goal lines using the reverse-rule model and the alignment-rule model. Herein, we set two different conditions with respect to the number of foraging agents (N = 4 and 9). More

Study subjects and measurement of skin physiological parametersWe analyzed skin microbiome and mycobiome from cheeks and foreheads of healthy younger (19–28 years old, Y-group) and older (60–63 years old, O-group) Korean women who were free from cutaneous disorders (Table 1 and Supplementary Table S1). All 61 subjects had been living in Seoul, Korea, for more than 3 years with normal skin conditions. We preferentially selected those who had sebum secretion greater than 30 arbitrary units and moisture greater than 50 arbitrary units in both groups. Among the measurements of moisture content, pH, sebum content, and transepidermal water loss (TEWL), only sebum and TEWL decreased significantly in the O-group compared to the Y-group in the cheeks (P = 2.25e−06, Wilcoxon rank-sum test; P = 0.019, Welch two-sample t test) and forehead (P = 1.33e−06, Wilcoxon rank-sum test; P = 0.003, Welch two-sample t test). Whereas no significant differences were found in the average values for moisture (cheeks: Y-group, 59.9; O-group, 56.6; forehead: Y-group, 61.1; O-group, 58.7) and pH (cheeks: Y-group, 6.0; O-group, 5.8; forehead: Y-group, 6.0; O-group, 5.6) between the two age groups.Table 1 Characteristics of subjects for aged related skin microbiome and mycobiome study.Full size tableComparisons in cheek and forehead microbiome and mycobiome between the two age groupsWe analyzed bacterial communities from 27 Y-group samples (cheeks, n = 13; forehead, n = 14) and 24 O-group samples (cheeks, n = 12; forehead, n = 12) and fungal communities from 28 Y-group samples (cheeks, n = 15; forehead, n = 13) and 32 O-group samples (cheeks, n = 16; forehead, n = 16), except for samples that were eliminated from the Illumina Mi-Seq sequencing due to low sequence reads (bacteria,  3. 0) (Fig. 4). Pathways belonging to the metabolism category were dominant in each age group. In the cheek of the Y-group, pathways involved in energy metabolism by bacteria, such as glycolysis/gluconeogenesis, citrate cycle, pentose phosphate pathway, fructose and mannose metabolism, galactose metabolism, d-alanine metabolism, and thiamine metabolism, were predominant, whereas in the cheek of the O-group, degradation-related pathways, such as fatty acid degradation, synthesis and degradation of ketone bodies, benzoate degradation, and chloroalkane and chloroalkene degradation, were predominant. In the forehead of the Y-group, glycolysis/gluconeogenesis, pentose phosphate pathway, fructose and mannose metabolism, galactose metabolism, d-glutamine and d-glutamate metabolism, d-alanine metabolism, and thiamine metabolism pathway were significantly more abundant, whereas in the forehead of the O-group, fatty acid degradation, synthesis and degradation of ketone bodies, valine/leucine and isoleucine degradation, and limonene/pinene degradation pathway were significantly more abundant.Figure 4Heat map for significantly different predicted functional pathways on (a) cheeks and (b) foreheads of Korean women by age based on LEfSe analysis (LDA score  > 3.0).Full size imageThe metabolism pathway for biotin, a water-soluble vitamin that is effective for skin health and essential for keratin production15, was more prevalent in the cheek and forehead of the Y-group. Interestingly, the metabolism pathway for lipoic acid, which is known to possess beneficial effects against skin aging and is used widely in cosmetic and dermatological products16,17, was significantly higher in the foreheads of the Y-group. We tracked the specific ASVs possessing these pathways, in both biotin metabolism and lipoic acid metabolism, Cutibacterium sp. (ASV2136 and ASV2130) and Staphylococcus sp. (ASV3008) were predicted to have the top three relative abundances in KOs. The relative abundances in biotin metabolism and lipoic acid metabolism of Cutibacterium sp. (ASV2136) were 24.9% and 26.1%, respectively. The relative abundances for each pathway for Staphylococcus sp. (ASV3008) were 10.2% and 18.7%, and for Cutibacterium sp. (ASV2130), they were 9.3% and 10.0%, respectively. We confirmed these two pathways in the genome of skin bacteria, C. acnes (Supplementary Fig. S2). These additional analyses support the reliability of the function in the skin environment of Cutibacterium. Interestingly, from the LEfSe result, Cutibacterium sp. (ASV2136) had a significantly higher abundance in the cheek and forehead microbiome of the Y-group. The pathway of biosynthesis of lipopolysaccharide, also known as bacterial endotoxins, showed higher abundance in the cheek and forehead microbiome of the O-group. The ASVs that contribute to inferring the LPS biosynthesis pathway were identified as Paraburkholderia sp. (ASV5030) and B. vesicularis (ASV4155). Also, pathways related to antibiotic biosynthesis (biosynthesis of vancomycin group antibiotics) and bacterial motility (bacterial chemotaxis and flagellar assembly; both belonging to the cellular processes category) were prominent in the cheek and forehead of the O-group. PICRUSt2 analysis implied that, regardless of skin site differences, the potential functions of the microbial community that compose the skin microbiome were similar according to age.Network analysis on cheek and forehead microbiome and mycobiomeWe performed SParse InversE Covariance estimation for Ecological Association Inference (SPIEC-EASI) analysis to evaluate the overall network of the skin microbes. The results of network density (D) on 81 cheek and 87 forehead ASVs, calculated using the ratio of the number of edges, showed higher network density in the skin microbiome of the Y-group (D = 0.015 and D = 0.001, in cheek and forehead, respectively) than the O-group (D = 0.007 and D = 0.007, respectively) (Fig. 5). To examine network correlation between bacteria and fungi, network density for Bacteria–Fungi (DBF) was calculated by the actual number of edges and a potential number of edges in a correlation ([bacterial nodes × fungal nodes]/2). We confirmed higher network density in the cheek of the Y-group (DBF = 0.008) than the O-group (DBF = 0) and edges of the major bacterial and fungal taxa, such as Staphylococcus sp. (ASV3008)—M. sympodialis (ASV500) and Roseomonas sp. (ASV4088)—M. restricta (ASV482), were observed in the cheek of the Y-group. In the forehead, edges of Methylobacterium sp. (ASV4314)—M. globosa (ASV454), Methylobacterium sp. (ASV4314)—Zygosaccharomyces rouxii (ASV208), and Venionella sp. (ASV3575)—M. sympodialis (ASV500) were observed in the Y-group, and edges of Cutibacterium sp. (ASV2107)—M. globosa (ASV461), Staphylococcus sp. (ASV3024)—M. arunalokei (ASV446), and Methylobacterium (ASV4314)—M. dermatis (ASV448) were observed in the O-group (DBF = 0.004). We found a network between bacteria and fungi with different kingdom levels in the skin microbiome, and especially, we confirmed that different genus or species level microbe was involved in the microbial network according to skin location and Y-, O-group.Figure 5Network analysis of the ASVs on (a) cheeks and (b) forehead of Korean women. Each node represents the ASV and the size of the node is based on relative abundance of each ASV. Color markings indicate the major taxa except for unidentified bacteria or fungi. Shapes represent the level of kingdom, Bacteria (bold) and Fungi (dotted line). The ASVs were selected for bacterial ASVs found in more than half of all samples on the cheeks and forehead, respectively, and for the fungal ASVs with a relative abundance of more than 0.1% in each of the cheeks and forehead samples. The D value is network density calculated using the ratio of the number of edges.Full size image More

Analytical targetsTwo species of undulation motion rays with different pectoral fin shapes bred in KAIYUKAN were analyzed: sharpnose stingray Dasyatis acutirostra and pitted stingray Dasyatis matsubarai (Fig. 1a,b). Blender 2.7925 was used to construct stingray models from pictures26,27 as accurately as possible; Blender is a free and open-source 3D creation suite used to make realistic characters for movies, etc. Detailed information on how to construct models using Blender is provided in our previous paper28. To focus on the effects of pectoral fin movements, we did not consider the body’s shape as in the previous studies12,29. The height and disk width (WD) of all models were set to 0.01 m and 0.44 m, respectively, considering the previous studies30,31. The disk length of each model was determined from WD, referring to the aspect ratio of the rays’ photographs26,27; the disk length (LD) of D. acutirostra and D. matsubarai are 0.348 m and 0.344 m, respectively.Figure 1Analytical targets and description of motion. (a) Analytical model of D. acutirostra. (b) Analytical model of D. matsubarai. (c) Description of motion, (d) the relationship between any two points on the surface before and after the deformation.Full size imageMotionThe motion was given to satisfy the following equations:$$z = left{ {begin{array}{*{20}l} {A;sin left( {omega left( {t – kTleft( {frac{{angl{text{e}}left( {x_{i} ,y_{i} } right) – 10^{{text{o}}} }}{{All;angl{text{e}}}} – theta } right)} right)h_{1} h_{2} } right.} hfill & {left( {10^{{text{o}}} le angl{text{e}}left( {x_{i} ,y_{i} } right) le 170^{{text{o}}} } right)} hfill \ {A;sin left( {omega left( {t – kTleft( {frac{{350^{{text{o}}} – left( {angl{text{e}}left( {x_{i} ,y_{i} } right) – 10^{{text{o}}} } right)}}{{All;angl{text{e}}}}} right)} right)h_{1} h_{2} } right.} hfill & {left( {190^{{text{o}}} le angl{text{e}}left( {x_{i} ,y_{i} } right) le 350^{{text{o}}} } right)} hfill \ end{array} } right.$$
(1)
$$begin{array}{c}{h}_{1}=a{r}_{i}^{3}+b{r}_{i}^{2}+c{r}_{i}end{array}$$
(2)
$$h_{2} = left{ {begin{array}{*{20}l} {dleft( {angleleft( {x_{i} ,y_{i} } right) – 10^{ circ } } right)^{2} + eleft( {angleleft( {x_{i} ,y_{i} } right) – 10^{ circ } } right)} hfill & {left( {10^{ circ } le angleleft( {x_{i} ,y_{i} } right) le 170^{ circ } } right)} hfill \ {dleft( {350^{ circ } – left( {angleleft( {x_{i} ,y_{i} } right) – 10^{ circ } } right)} right)^{2} + eleft( {350^{ circ } – left( {angleleft( {x_{i} ,y_{i} } right) – 10^{ circ } } right)} right)} hfill & {left( {190^{ circ } le angleleft( {x_{i} ,y_{i} } right) le 350^{ circ } } right)} hfill \ end{array} } right.$$
(3)
$$begin{array}{c}{left({r}_{i}-{r}_{i-1}right)}^{2}+{left({z}_{i}-{z}_{i-1}right)}^{2}={left({r}_{i}^{mathrm{^{prime}}}-{r}_{i-1}^{mathrm{^{prime}}}right)}^{2}+{left({z}_{i}^{mathrm{^{prime}}}-{z}_{i-1}^{mathrm{^{prime}}}right)}^{2}end{array}$$
(4)
$$begin{array}{c}angleleft({x}_{i},{y}_{i}right)= angleleft({x}_{i}^{^{prime}},{y}_{i}^{^{prime}}right).end{array}$$
(5)
Equation (1) represents the amount of movement of the model surface in the z-axis direction, where (A) is the amplitude of the pectoral fin tip, (omega) is the angular velocity, (t) is time, (k) is the wavenumber, (T) is the period, angle(({x}_{i},{y}_{i})) is the angle made by the line connecting the center of rotation and any point (({x}_{i},{y}_{i})) on the model surface with the x-axis, Allangle is the range where the motion is given (160°), and (theta) is the phase difference between the movements of the right and left pectoral fins (Fig. 1c). ({h}_{1}) is the weighting from the center to the radial direction: it is necessary to set the amplitude at the ray’s center to zero and increase the amplitude toward the pectoral fin tip (a = 119.786, b = -7.957, c = 0.498). ({h}_{2}) is the weighting in the circumferential direction: it is necessary to increase the amplitude from the anterior to the tip of the pectoral fin and decrease the amplitude from the tip of the pectoral fin to the posterior (d = − 1.563 × 10–4, e = 0.025). Equation (4) is the condition in which the distance between two neighboring points in the same radial direction is equal before and after the movement (Fig. 1d). (r) is the distance between the center of rotation and any point (({x}_{i},{y}_{i})), defined as (sqrt{{x}_{i}^{2}+{y}_{i}^{2}}). Equation (5) defines angle(({x}_{i},{y}_{i})) as being constant before and after the move (Fig. 1c). The variables after the move are marked with ‘. Variables used in the analysis are A = 0.089 m, T = 0.499, k = 1.270, and ω = 12.599 rad/s. Videos of the created motion from the front and the side are shown in the “Supplement” (Supplement Movies 3, 4).Analytical conditionsAnalysis cases were conducted with eight conditions: two types of pectoral fin shape (Fig. 1a,b) and four types of phase difference (0 (T), 0.25 (T), 0.5 (T), and 0.75 (T)). These conditions were set for investigating the effects of phase differences between left and right pectoral fin movements on swimming and how these effects vary with pectoral fin shape.Numerical methodsA CFD simulation of the ray models in the water flow was performed using OPENFOAM, an open-source finite volume method CFD toolbox32, to calculate the forces acting on the rays in each axial direction and the moment around each axis. The governing equations were the continuity equation and the three-dimensional incompressible Reynolds-averaged Navier–Stokes equation, expressed by:$$begin{array}{*{20}c} {nabla cdot u = 0} \ end{array}$$
(6)
$$begin{array}{*{20}c} {frac{partial u}{{partial {text{t}}}} + nabla cdot left( {uu} right) = – nabla p + nabla cdot left( {vnabla u} right) + nabla cdot left[ {nu left{ {left( {nabla u} right)^{T} – frac{1}{3}nabla cdot uI} right}} right], } \ end{array}$$
(7)
where (u) is the velocity vector, t is the time, p is the static pressure divided by the reference density, (nu) is the kinematic viscosity, and I is the unit tensor. The Reynolds number was defined regarding the previous studies3 as:$$begin{array}{c}{R}_{e}=frac{U{L}_{D}}{nu },end{array}$$
(8)
where (U) (/ms) is the given flow speed3 (1.5 × LD/ms), ({L}_{D}) (m) is the length of the ray models, and (nu) is the kinematic viscosity of water at 20 °C (1.0 × 10–6 m2/s). The Reynolds number in this study is 1.8 × 105; considering this, we used the k–ω shear stress turbulence model33,34. The k–ω shear stress turbulence model is a type of Reynolds-averaged Navier–Stokes equation (RANS) turbulence model that is widely used to calculate for the fish swimming flow35,36,37. The overset grid method was used in this study; it is a generic implementation of overset meshes. For both static and dynamic cases, cell-to-cell mapping between multiple, disconnected mesh regions is employed to generate a composite domain38,39. This method permits complex mesh motions and interactions without the penalties associated with deforming meshes. The process is described in detail by Noack40. The calculation volume was 5.4 WD in length, 5.4 WD in height, and 5.4 WD in width (Fig. 2a,b). A hexahedral volume mesh was created using the snappyHexMesh of OPENFOAM. The fluid region was divided into two parts: the overset region and the background region (Fig. 2a,b). The overset region moves and transforms to match the motion of the ray and was made with fine meshes around the analysis target and coarse meshes in the outlying areas; a one-layer boundary layer mesh was created around the analysis target. The overset region shape is an ellipsoid (Fig. 2a,b). The minimum mesh volume is 7.3 × 10–10 (m3), and the maximum mesh volume is 2.6 × 10–2 (m3). The total number of meshes was 9.0 × 105. At the outlet boundary, the average static relative pressure was set to 0 Pa. The surfaces of the fish model were formed into non-slip surfaces.Figure 2Meshes for CFD simulation and differences in force between different meshes. (a) Meshes at the coronal plane of the whole fluid region. (b) Frontal cross-section of the fluid region at the green line in (a). The red region is the overset region. (c,d) Comparison of the instantaneous drag coefficient and the moment coefficient around the y-axis of D. matsubarai between the coarse, fine, and dense mesh.Full size imageThe drag coefficient ({C}_{D}left(tright)), the lateral force coefficient ({C}_{l}left(tright)), the lift coefficient ({C}_{L}left(tright)), the moment coefficient around the x-axis ({C}_{mx}left(tright)), the moment coefficient around the y-axis ({C}_{my}left(tright)) and the moment coefficient around the z-axis ({C}_{mz}left(tright)) were calculated as:$$begin{array}{c}{C}_{D}left(tright)=frac{Dleft(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}{W}_{D}}end{array}$$
(9)
$$begin{array}{c}{C}_{l}left(tright)=frac{lleft(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}{W}_{D}}end{array}$$
(10)
$$begin{array}{c}{C}_{L}left(tright)=frac{Lleft(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}{W}_{D}}end{array}$$
(11)
$$begin{array}{c}{C}_{mx}left(tright)=frac{{M}_{psi }left(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}^{2}{W}_{D}}end{array}$$
(12)
$$begin{array}{c}{C}_{my}left(tright)=frac{{M}_{phi }left(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}^{2}{W}_{D}}end{array}$$
(13)
$$begin{array}{c}{c}_{mz}left(tright)=frac{{M}_{theta }left(tright)}{frac{1}{2}rho {U}^{2}{L}_{D}^{2}{W}_{D}},end{array}$$
(14)
where (Dleft(tright)) is the calculated drag, (lleft(tright)) is the calculated lateral force, (Lleft(tright)) is the calculated lift, ({M}_{psi }left(tright)) is the calculated moment around the x-axis, ({M}_{phi }left(tright)) is the calculated moment around the y-axis, ({M}_{theta }left(tright)) is the calculated moment around the z-axis, and (rho) (kg/m3) is the density of water at 20 °C (998 kg/m3). As shown in a previous study41. the propulsive efficiency (eta) is defined as the ratio of output power ({P}_{o}) to input power ({P}_{e}) which can be written as:$$begin{array}{c}{P}_{o}left(tright)=frac{1}{T}{int }_{0}^{T}Dleft(tright)Udtend{array}$$
(15)
$$begin{array}{c}{P}_{e}left(tright)=frac{1}{T}{int }_{0}^{T}left[Dleft(tright)dot{x}left(tright)+lleft(tright)dot{y}left(tright)+Lleft(tright)dot{z}left(tright)right]dtend{array}$$
(16)
$$begin{array}{c}eta =frac{{P}_{o}}{{P}_{e}}.end{array}$$
(17)
An in-house program calculated the forces acting on rays in each axial direction and the moment around each axis. The numerical method’s validity and reliability were verified by comparing previous experimental and numerical analytical studies of heaving and pitching on airfoil naca001341. A high degree of similarity to previous studies was confirmed; the mean difference in the propulsive efficiency from the previous study of analysis was 5%, and the difference from the previous study of the experiment was 9%. Detailed information such as mesh, length, and velocity, of this analysis method’s verification is provided in the “Supplement”.A grid sensitivity study was conducted using three meshes: coarse, fine, and dense. The coarse mesh has 8.1 × 105 elements, the fine mesh has 9.0 × 105 elements, and the dense mesh has 9.9 × 105 elements. The analysis was conducted using a condition with no phase difference of D. matsubarai. As shown in Fig. 2c,d, the drag coefficient and the moment coefficient around the y-axis are almost the same when the mesh is fine and when the mesh is dense. The mean drag and propulsive efficiency error of fine and coarse meshes are 2.7% and 3.5%, respectively. The fine mesh was used in all simulation cases considering accuracy. We used the same meshes for all cases. More

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript. More

All of the statistical analyses were conducted using the R programming language version 4.0.5 within the RStudio IDE version 1.4.124,25. Data visualization and processing were performed with the ‘tidyverse’ collection, ‘foreach’ and ‘doParallel’ R packages26,27,28. Geographical Information System (GIS) operations on raster and vector files were conducted using the ‘sf’, ‘exactextractr’ and ‘raster’ R packages29,30,31.Data sources and pre-processingBiodiversity dataWe used the North American Breeding Bird Survey dataset as our source of biodiversity data due to its long temporal coverage and spatial extent14,32. The BBS is composed of bird species abundance records collected since 1966 from over 4,000 survey routes across the countries of Mexico, USA and Canada. For this study we focused solely on routes in the USA, due to their longer time dimension. Data collection follows public access roads that are 24.5 miles long (circa 39.2 km) using a point count protocol whereby routes are surveyed every half-mile (800 m) for a total of 50 stops. At each stop, observers stand for 3 min and record the species and the abundance of every bird seen or heard within 400 m of their location. The routes are surveyed by volunteers with experience in bird observation, and surveys are conducted from late April to July to capture the peak of the breeding season.We selected the years 2001 and 2016 as the two timepoints of our analysis. This 15-year timeframe corresponded to the longest possible timespan for which land cover data products were available at high spatial resolution18. Before analysis, we subset the BBS dataset by removing routes that had incomplete survey lengths (less than 50 point count stops, indicated by the RouteTypeDetailID field value being less than 2 in the extracted BBS dataset) or that were surveyed under adverse weather conditions such as high wind and rain (as indicated by the Run Protocol ID field being equal to 1), which could affect bird occurrence and detectability. Following this filtering process, the total number of BBS routes analysed was 960 (Extended Data Fig. 1).For higher precision when inferring the relationship between avian diversity and environmental variables, we subdivided each route into five segments of equal length, consisting of 10 count locations each. This approach was motivated by the need to more closely associate bird communities with the land cover composition in the area in which they are found. To minimize the spatial autocorrelation between adjacent segments and avoid overlaps in landscapes analysed, we filtered the data to keep only the first, third and fifth segment of each route. These segments therefore formed our sampling unit used in all analyses.We recognized that environmental conditions and stochastic trends in populations could introduce variability in biodiversity calculated from bird community data. We therefore extracted, for each segment and each species, the average population count across a 3-year period centred on our two timepoints (2000, 2001, 2002 and 2015, 2016, 2017)33. We then calculated the mean abundance of each species across these 3 years.The effect of observer experience34,35,36 was accounted for by sourcing the observer ID responsible for each route at each timepoint and including it as a random effect in the legacy model (see ‘Model development’ section). We also controlled for the time of day as it is plausible to expect visibility and avian species activity patterns to vary between early morning and later parts of the day. Time of day for each segment was calculated by averaging across the start and end time data entries associated with each route, and then including this as a covariate in both the legacy and equilibrium models (see ‘Model development’ section). However, we did not model detectability issues associated with traffic noise and disturbance for two reasons. First, all BBS surveys are conducted along public access roads with a vehicle, so the disturbance is expected to be similar across sites. Second, previous studies have found no clear evidence for noise being the main cause for reduced bird abundance near roads37.Following these procedures, our processed BBS dataset included entries of mean abundances of each species for a total of 2,880 segments, corresponding to segment 1, 3 and 5 of 960 routes (Extended Data Figs. 1 and 2). For each segment, at each timepoint we calculated different measures of alpha diversity following the Hill numbers framework38. We then selected to use the effective number of species at q = 1, calculated as the exponential of the Shannon–Wiener Index38. The effective number of species at q = 1 sits at the theoretical half-way point between the classic species richness measure that accounts only for the absolute number of species (q = 0) and the Berger-Parker dominance index (q = infinity), which instead only reflects the most common species. Thus, the effective number of species is a robust alternative to species richness, which does not take account of species rarity or detectability and can thus lead to biased biodiversity estimates16,17.Land cover and environmental dataLand cover data for the US for our focal years of 2001 and 2016 were sourced from the open-access NLCD CONUS products developed by the US Geological Survey (USGS)18,39. The NLCD products are high-resolution (30 m pixel dimensions) classified raster files covering the land area of the whole USA. This dataset provides us with the opportunity to look at finely gridded spatio-temporal changes in a landscape over a relatively long timeframe of 15 years, while utilizing data collected and analysed with the same methods (for example, land use classification algorithms).To reduce the number of potentially collinear explanatory variables included in our models, we aggregated the land cover variables provided by the NLCD dataset. We summarized these to five land cover categories: ‘urban’ (an aggregate of the Developed-Open Space (subclass 21), Developed-Low Intensity (22), Developed-Medium Intensity (23) and Developed-High Intensity classes); ‘forest’ (an aggregate of the Deciduous Forest (41), Evergreen Forest (42) and Mixed Forest (43) classes); ‘grassland’ (an aggregate of the Shrub (52), Grassland/Herbaceous (71) and Pasture/Hay (81) classes); ‘cropland’ (cultivated Crops (82) subclass) and ‘wetland’ (an aggregate of the Woody Wetland (90) and Herbaceous Wetland (95) classes). The Perennial Ice/Snow (12), Open Water (11) and Barren Land (31) classes were excluded from the analysis as they were very uncommon in our dataset. The distribution and total area of the land cover categories across the US are shown in Supplementary Figs. 1 and 2. Temperature data were sourced from the 30 arc-seconds gridded PRISM climate database19 and were extracted as the mean across May and June for each group of years from which bird abundances were taken.We first sampled the landscape surrounding each segment using a range of buffer shapes and sizes, and then selected the buffer type on the basis of the capacity of each buffer type to explain the response variable. The types of buffers that we explored were: a circular buffer around the centroid of the polygon defined by the vertices of each segment (4,000 m radius) and a series of three buffers around the segment line (500 m, 2,000 m and 4,000 m radius). The best fit was given by the smallest buffer size of 500 m, shown in Extended Data Fig. 2, which also coincides with the BBS protocol effective counting distance of 400 m and more closely reflects the size of bird territories14. Land cover variables were computed as a percentage of the total buffer area. Change in percentage points for each land cover type between the 2 years was computed by subtracting the values at the two timepoints. A change product is also provided by the USGS databases40, but it does not meet our needs because it considers land cover changes based on a ranking. Nonetheless, a comparison of urban land cover change between the timepoints showed a similar result (Supplementary Fig. 4). Land cover data were processed geospatially using the NAD 83 Conus Albers Coordinate Reference Systems projection, EPSG 5070.Model developmentTheoretical backgroundWe developed a statistical model that conceptualized extinction debts and colonization credits by combining the following two concepts: (1) the settled biodiversity of avian communities in a given landscape composition (that is, a system at equilibrium) and (2) the lagged response in the species diversity in a given landscape due to recent land cover changes (that is, a system moving to a new equilibrium). We reasoned that, given enough time, and with no further changes in land cover, the effective number of species at a given location would eventually equilibrate. The equilibrium distribution of the effective number of species emerges with the waning of the legacy effect of previous landscape compositions in encouraging or impeding the recruitment and survival of particular species. We did not model these ecological mechanisms directly, but instead expressed the equilibrium of the effective number of species, and the rate of approach to this equilibrium, as empirical functions of environmental covariates. It is important to keep in mind that during a finite time interval following environmental change, it is possible that our observations of effective number of species represent a system in a transitory state towards its new equilibrium. Yet, environmental changes may occur at rates that never allow the system to equilibrate. Although the equilibration processes are latent (that is, not amenable to direct observation), the combination of equilibrium and temporal legacy components into an integrated model, applied to a dataset with extensive environmental replication (due to spatial expansiveness), has allowed us to retrieve distributions for all relevant model parameters (see below).Model overviewThe observed effective number of species Rs,t at site s in year t for t = t1,t2 is modelled as a normally distributed variate with mean μs,t and standard deviation σ$$R_{s,t} approx mathrm{Normal}left( {mu _{s,t},sigma } right)$$
(1)
We assume that, under landscape change, the system is in a state of flux and that the data are from observations witnessing the transition between two (unattained) equilibria. The expected state of the system at any given point in time, μs,t, was formulated as a mixture of past and future equilibrium distributions (that is, a weighted average of the two distributions, where the weights are given by the complementary proportions ω and 1 − ω)$$mu _{s,t} = fleft( {x_{s,t_2};beta } right)omega left( {{Delta}x_{s,t_1,t_2};gamma } right) + fleft( {x_{s,t_1};beta } right)left( {1 – omega left( {{Delta}x_{s,t_1,t_2};gamma } right)} right)$$
(2)
Here, the function f describes the equilibrium distribution of the effective number of species as a function of the configuration of the local environment, captured in covariates xs,t. The weighting function ω depends on covariates ys,t derived from the difference in the local land cover between 2016 and 2001 (that is, it is a function of the land cover change that has taken place). The mixture weights ω and (1 − ω) determine the relative importance of the two equilibrium distributions (past or current). If ω = 1, the interpretation is that the new equilibrium distribution has been completely attained, and thus the current (2016) effective number of species is entirely explained by the current (2016) land cover. Conversely, if ω = 0, the current effective number of species is entirely explained by the past (2001) land cover. The vectors of parameters β and γ, presented in equation (2), are inferred from model fitting.We also augmented equation (2) with a function g of static covariates and random effects z that we expect to have an impact on the effective number of species. Thus, the model comprised equilibrium components, a temporal legacy component and static covariates:$$mu _{s,t} = fleft( {x_{s,t_2};beta } right)omega left( {{Delta}x_{s,t_1,t_2};gamma } right) + fleft( {x_{s,t_1};beta } right)left( {1 – omega left( {{Delta}x_{s,t_1,t_2};gamma } right)} right) + gleft( {z_s;alpha } right)$$
(3)
in which (fleft( {x_{s,t};beta } right)) are the equilibrium components for the two timepoints, (omega left( {Delta x_{s,t_1,t_2};gamma } right)) is the temporal legacy component, and (gleft( {z_s;alpha } right)) is the function that captures the static covariates and random effects, with α being the estimated static covariates parameter effects.Equilibrium componentsWe defined the equilibrium distribution of the effective number of species at a given timepoint as a function (fleft( {x_{s,t};beta } right)) of land cover. This function describes the expected effective number of species at location s, given sufficient time for the community to adapt to the given land cover composition. We now describe this function in more detail.The equilibrium component was formulated as a log-linear model comprising a total of I = 5 environmental covariates (the percentage cover of five landscape classes: urban, forest, grassland, wetland and cropland), using 2nd-order polynomial terms, captured by the coefficient j, to account for optima in effective number of species along each of the five environmental dimensions:$$fleft( {x_{s,t}} right) = {mathrm{exp}}left( {beta _0 + mathop {sum }limits_{i = 1}^{I = 5} mathop {sum }limits_{j = 1}^{J = 2} beta _{i,j}x_{i,s,t}^j} right)$$
(4)
In equation (4), the β parameters capture the effect of covariates on the equilibrium and are assumed to be the same for each environmental composition. A simplifying assumption necessary for the application of this model is that the effective number of species had equilibrated at the first timepoint. As data become available for more years in the future, the influence of this assumption on the model results will diminish and more accuracy will be achievable with multiple timepoints.To allow for conditionality in the effects of one land cover variable on the response of the effective number of species to another land cover variable, we extended this function with pairwise interaction terms k between all the linear terms for land cover variables and pairwise linear-quadratic terms, as follows:$$fleft( {x_{s,t}} right) = {mathrm{exp}}left( {beta _0 + mathop {sum }limits_{i = 1}^{I = 5} mathop {sum }limits_{j = 1}^{J = 2} beta _{0,i,j,}x_{i,s,t}^j + mathop {sum }limits_{i = 1}^{I = 4} mathop {sum }limits_{k = i + 1}^{K = 5} beta _{1,i,k}x_{i,s,t}x_{k,s,t}} right)$$
(5)
Temporal legacy componentThe main covariates, ({Delta}x_{i,s,t_1,t_2}), for the part of the model that captures the temporal legacy, (omega left( {{Delta}x_{s,t_1,t_2};gamma } right)), are derived from the change in land cover (({Delta}x_{i,s} = x_{i,s,t_2} – x_{i,s,t_1})) between the two timepoints$$x_{i,s,} = left{ {begin{array}{*{20}{l}} {x_{1,i,s} = left| {{Delta}x_{i,s}} right|,} hfill & {x_{2,i,s} = 0,} hfill & {{mathrm{if}},{Delta}x_{i,s} < 0} hfill \ {x_{1,i,s} = 0,} hfill & {x_{2,i,s} = {Delta}x_{i,s},} hfill & {{mathrm{otherwise}}} hfill end{array}} right.$$ (6) where ({Delta}x_{s,t,z}) is a vector of the ith environmental change variable (that is, urban, forest, grassland, wetland, cropland) at site s and for directionality z. The effect of these covariates on the mixture weights is given by:$$omega left( {{Delta}x_{s,t_1,t_2};gamma } right) = {mathrm{exp}}left( {mathop {sum }limits_{i = 1}^{I = 5} - gamma _{i,z}{Delta}x_{z,s,i}} right)$$ (7) This formulation weights the contribution that the environmental variables at the two timepoints have on the current effective number of species, as a function of the magnitude and directionality of change in each type of land cover covariate. The γ parameters, and subsequently the temporal legacy component, are allowed via the inclusion of the environmental change data ({Delta}x_{s,t,z}), to account for the distance between the land cover at the two timepoints, therefore quantifying how far the initial community would need to travel to reach equilibrium in 2016 as a function of the type, magnitude and directionality of change. It should be noted that our model, in equation (3), is only implicitly related to the speed with which the effective number of species reacts to environmental changes. Instead, it quantifies how much further it would still have to travel to reach the expected equilibrium associated with the current configuration of the landscape.Static covariatesAs described in model equation (3), we included a function of static covariates to which we can expect the effective number of species to respond without lags relating to the past landscape. We added a linear and quadratic fixed effect for temperature in 2016 to control for any difference in the effective number of species related to climatic characteristics and to allow for a parabolic relationship to be expressed (optima either at mean or extremes values). We also controlled for the heterogeneity of a landscape by including the effective number of land cover types, computed in the same way as the effective number of species, as a fixed effect40. A fixed effect for time of day, reflecting the time at which each segment was surveyed, was included to correct for differences in species detectability between early morning and later parts of the day41. An observer-level random effect was also added to control for variation between observers35,36 and partly account for between-route variation, given that we would expect observers who collect data from multiple routes to do so within a relatively small area. Spatial autocorrelation of the effective number of species was tested for all segments at once and by different radiuses for neighbour inclusion (500 m, 1,000 m, 5,000 m, 10,000 m, 100,000 m), using the Moran’s I statistic42. Spatial autocorrelation was not corrected for because Moran’s I was not significant at any spatial scale (P  > 0.05). Pseudo-replication between neighbouring segments was avoided by considering segments 1, 3 and 5, whose land cover buffers did not overlap (Extended Data Fig. 2).Model fittingThe model was fitted within a Bayesian framework using a Hamiltonian Markov chain Monte Carlo algorithm implemented in the STAN programming language43 version 4.3.0 and the ‘cmdstanr’ R package version 2.26.144.We ran 4 chains, sampling for 1,000 iterations with a burn-in period of 500 iterations each. These numbers of iterations were sufficient to achieve chain convergence. The STAN sampling was run on four parallel threads on a multi-core Intel i7 – 8750H processor with a maximum clock speed of 4.1 GHz.For the purposes of Bayesian inference, all slope parameters associated with the equilibrium component equation (5) and the static additive terms were assigned an unbiased prior (beta _{i,j} approx Nleft( {0,1} right),{mathrm{and}},z_s approx Nleft( {0,1} right),) where N is normal, with the aim of shrinking the parameter estimated towards 0 (that is, no covariate effect). A gamma distributed prior, with shape and rate 0.001, was assigned to the standard deviation of the random effect. For the following known and expected relationships, we also truncated the range of parameter values by bounding the upper or lower limits of the prior/posterior distributions. Intercept and standard deviation of the observer random effect were bounded below by 0. Linear effects for the environmental covariates and temperature were bounded below at 0, while their quadratic counterparts were bounded above at 0. Interaction terms were not limited. The temporal legacy component parameters were given a uniform (U) prior (gamma _isim Uleft( {0,1} right)), bounded between 0 and 1 to act as a weighting proportion between the present and the past. The upper bound on the gamma parameters to 1 does not bias us towards an increased contribution of the past land cover, but instead provides a more conservative approach.Model diagnostics were conducted by assessing chain convergence visually through trace plots, as well as statistically by employing the Gelman-Rubin test, which compares the estimated between-chain and within-chain variances45. Chain autocorrelation and the associated effective sample size were also monitored. In the case of low effective sample size, the chains were extended until the effective sample size exceeded a threshold value of 400. The marginal posterior distribution for each parameter was visualized via a density plot to check for multimodality.Model selection was conducted to inform choice of the size and shape of the land cover buffer around each sampled segment. We did so by comparing values of the Watanabe-Akaike Information Criterion leave-one-out (WAIC)-loo information criterion46 of four different models, each computed using land cover data calculated with two different buffer options of various sizes: a circular buffer around the centroid of the polygon defined by the vertices of each segment (4,000 m radius) and a series of buffers around the segment line (500 m, 2,000 m and 4,000 m radius). This approach was implemented through the ‘loo’ R package version 2.1, which provides an improvement on the original WAIC by including diagnostic measures around the point-wise log-likelihood value estimated around each sample draw47.Visualization of model predictionsA map of the USA (Fig. 1) was produced to represent the predicted extinction debts and colonization credits (that is, positive or negative distance in the effective number of species from the expected equilibria). The map was produced on a hexagonal grid at a spatial resolution of 10 km vertex-to-opposite-vertex, with each hexagon covering a total of 86 km2. Values of extinction debt and colonization credit were calculated by subtracting the predicted effective number of species produced by the model (equation 3) from the predicted effective number of species at equilibrium in 2016 (that is, when the legacy component equals 1). To correctly propagate and represent uncertainty in the extinction debts and colonization credits presented, this process was repeated 1,000 times for predictions originating from different draws from the posterior distribution. Uncertainty in the form of the geometric coefficient of variation, calculated as (root {2} of {{e^{left( {mathrm{log}left( {sigma + 1} right)^2} right)} – 1}}) where σ is the standard deviation, is mapped in Extended Data Fig. 4a. Extended Data Fig. 4 also includes a copy of Fig. 1 (Extended Data Fig. 4b) for reference, alongside upper (Extended Data Fig. 4c) and lower (Extended Data Fig. 4d) credible intervals.Over/underestimation values of biodiversity that could arise by neglecting debts and credits were computed as the difference between the effective numbers of species predicted by the equilibrium model and the legacy model, multiplied by 100 and then divided by the predicted effective number of species under the legacy model. This calculation results in a percentage measurement of the extent to which (in relative terms) the current effective number of species under- or overestimates the diversity that a given landscape can sustain at equilibrium.To further validate our predicted extinction debts and colonization credits, we compared the direction of the expected changes with the recorded difference in effective numbers of species between 2016 and 2019 (the latest year for which data are available). To do so, we sourced bird abundances from the North American BBS dataset14,32 for the year 2019 and conducted the same data processing as described above for the other two timepoints. We then conducted a Pearson correlation test to assess how well the observed change followed the model-predicted one. We are nevertheless aware that a 3-year timeframe is unlikely to be large enough for debts and credits to fully manifest.Plots were also generated to describe the behaviour of the mixture weight, ω (equation 7), which captures the contribution (weighting) of the landscape composition in determining the effective numbers of species at the two timepoints (Fig. 2 in the main text). Values of ω across the whole spectrum of plausible land cover change values (that is, from −100 to +100) were simulated by averaging over 10,000 draws from the posterior distribution of each γ parameter. Credible intervals were measured by taking the 95% range of the 10,000 draws.Explaining spatial variation in debts and creditsThe extinction debts and colonization credits predicted for the contiguous USA states were further modelled to identify which past land cover changes were the main drivers of the delayed biodiversity changes in USA bird communities. We considered the values of debts or credits associated with the 92,000 individual 86 km2 hexagons (Fig. 1) as a response variable. We then specified a Gaussian linear model including the magnitude of each land cover change as explanatory covariates. Positive and negative changes in each covariate were treated as separate linear components to differentiate their effects. The model was fitted to 1,000 sets of debts and credits, each originating from predictions based on independent draws from the posterior distribution. For each generalized linear model (GLM) fit, we then subsequently sampled each parameter distribution another 1,000 times and extracted the summarized parameter estimates from a total of 100,000 values. Model coefficients and their resulting uncertainty from the above process are presented in Fig. 4 and in more detail as part of Supplementary Table 3.Reporting SummaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More