Philippa Kaur

Darwin, C. On the various Contrivances by which British and Foreign Orchids are Fertilised by Insects, and on the good effects of Intercrossing. (John Murray, 1862).Darwin, C. The various Contrivances by which Orchids are Fertilised by Insects. Second edition, revised., (D. Appleton and Company, 1877).Sprengel, C. K. Das entdeckte Geheimnis der Natur im Bau und in der Befruchtung der Blumen. (Vieweg, 1793).Müller, H. Befruchtung der Blumen durch Insekten (Verlag Von Wilhelm Englemann, 1873).Book 

Google Scholar 
Riley, C. V. The yucca moth and yucca pollination. Rep. Missouri Botan. Garden 3, 99–159 (1892).Article 

Google Scholar 
Faegri, K. & Van Der Pijl, L. Principles of Pollination Ecology 3rd edn. (Pergamon, Berlin, 1979).
Google Scholar 
Fenster, C. B., Armbruster, W. S., Wilson, P., Dudash, M. R. & Thomson, J. D. Pollination syndromes and floral specialization. Annu. Rev. Ecol. Evol. Syst. 35, 375–403. https://doi.org/10.1146/annurev.ecolsys.34.011802.132347 (2004).Article 

Google Scholar 
Inouye, D. W. In The Biology of Nectaries (eds Elias, T. S. & Bentley, B. L.) 153–173 (Columbia University Press, 1983).
Google Scholar 
Irwin, R. E., Bronstein, J. L., Manson, J. S. & Richardson, L. Nectar robbing: ecological and evolutionary perspectives. Annu. Rev. Ecol. Evol. Syst. 41, 271–292. https://doi.org/10.1146/annurev.ecolsys.110308.120330 (2010).Article 

Google Scholar 
Irwin, R. E. & Maloof, J. E. Variation in nectar robbing over time, space, and species. Oecologia 133, 525–533. https://doi.org/10.1007/s00442-002-1060-z (2002).ADS 
Article 
PubMed 

Google Scholar 
Maloof, J. E. & Inouye, D. W. Are nectar robbers cheaters or mutualists?. Ecology 81, 2651–2661. https://doi.org/10.1890/0012-9658(2000)081[2651:ANRCOM]2.0.CO;2 (2000).Article 

Google Scholar 
Inouye, D. W. The terminology of floral larceny. Ecology 61, 1251–1253. https://doi.org/10.2307/1936841 (1980).Article 

Google Scholar 
Lyon, D. L. & Chadek, C. Exploitation of nectar resources by hummingbirds, bees (Bombus), and Diglossa baritula and Its role in the evolution of Penstemon kunthii. Condor 73, 246–248. https://doi.org/10.2307/1365847 (1971).Article 

Google Scholar 
Colwell, R. K., Betts, B. J., Bunnell, P., Carpenter, F. L. & Feinsinger, P. Competition for the nectar of Centropogon valerii by the hummingbird Colibri thalassinus and the flower-piercer Diglossa plumbea, and Its evolutionary implications. Condor 76, 447–452. https://doi.org/10.2307/1365817 (1974).Article 

Google Scholar 
Arizmendi, M. C., Dominguez, C. A. & Dirzo, R. The role of an avian nectar robber and of hummingbird pollinators in the reproduction of two plant species. Funct. Ecol. 10, 119–127. https://doi.org/10.2307/2390270 (1996).Article 

Google Scholar 
Arizmendi, M. C. Multiple ecological interactions: Nectar robbers and hummingbirds in a highland forest in Mexico. Can. J. Zool. 79, 997–1006. https://doi.org/10.1139/z01-066 (2001).Article 

Google Scholar 
Navarro, L. Pollination ecology and effect of nectar removal in Macleania bullata (Ericaceae)1. Biotropica 31, 618–625. https://doi.org/10.1111/j.1744-7429.1999.tb00410.x (1999).Article 

Google Scholar 
Traveset, A., Willson, M. F. & Sabag, C. Effect of nectar-robbing birds on fruit set of Fuchsia magellanica in Tierra Del Fuego: A disrupted mutualism. Funct. Ecol. 12, 459–464. https://doi.org/10.1046/j.1365-2435.1998.00212.x (1998).Article 

Google Scholar 
Skutch, A. F. Life histories of Central American birds. Families Fringillidae, Thraupidae Parulidae and Coerebidae. Pacific Coast Avifauna 31, 1–448 (1954).
Google Scholar 
Vuilleumier, F. Systematics and evolution in Diglossa (Aves, Coerebidae). Am. Mus. Novit. 2381, 1–44 (1969).
Google Scholar 
Graves, G. R. Pollination of a Tristerix mistletoe (Loranthaceae) by Diglossa (Aves: Thraupidae). Biotropica 14, 315–317. https://doi.org/10.2307/2388094 (1982).Article 

Google Scholar 
Hernández, H. M. & Toledo, V. M. The role of nectar robbers and pollinators in the reproduction of Erythrina leptorhiza. Ann. Mo. Bot. Gard. 66, 512–520. https://doi.org/10.2307/2398843 (1979).Article 

Google Scholar 
Neill, D. A. Trapliners in the trees: Hummingbird pollination of Erythrina Sect Erythrina (Leguminosae: Papilionoideae). Ann. Missouri Botan. Garden 74, 27–41. https://doi.org/10.2307/2399259 (1987).Article 

Google Scholar 
Hazlehurst, J. A. & Karubian, J. O. Nectar robbing impacts pollinator behavior but not plant reproduction. Oikos 125, 1668–1676. https://doi.org/10.1111/oik.03195 (2016).CAS 
Article 

Google Scholar 
Cuta-Pineda, J. A., Arias-Sosa, L. A. & Pelayo, R. C. The flowerpiercers interactions with a community of high Andean plants. Avian Res. 12, 22. https://doi.org/10.1186/s40657-021-00256-7 (2021).Article 

Google Scholar 
Askins, R. A., Karen, M. E. & Jeffrey, D. W. Flower destruction and nectar depletion by avian nectar robbers on a tropical tree, Cordia sebestena. J. Field Ornithol. 58, 345–349 (1987).
Google Scholar 
McDade, L. A. & Kinsman, S. The impact of floral parasitism in two Neotropical hummingbird-pollinated plant species. Evolution 34, 944–958. https://doi.org/10.2307/2408000 (1980).Article 
PubMed 

Google Scholar 
Ingels, J. Observations of the hummingbirds Orthorhynchus cristatus and Eulampis jugularis of Martinique (West Indies). Gerfaut 66, 129–132 (1976).
Google Scholar 
Feinsinger, P., Beach, J. H., Linhart, Y. B., Busby, W. H. & Murray, K. G. Disturbance, pollinator predictability, and pollination success among Costa Rican cloud forest plants. Ecology 68, 1294–1305. https://doi.org/10.2307/1939214 (1987).Article 

Google Scholar 
Kodric-Brown, A., Brown, J. H., Byers, G. S. & Gori, D. F. Organization of a tropical island community of hummingbirds and flowers. Ecology 65, 1358–1368. https://doi.org/10.2307/1939116 (1984).Article 

Google Scholar 
Lara, C. & Ornelas, J. F. Preferential nectar robbing of flowers with long corollas: Experimental studies of two hummingbird species visiting three plant species. Oecologia 128, 263–273. https://doi.org/10.1007/s004420100640 (2001).ADS 
Article 
PubMed 

Google Scholar 
Hazlehurst, J. A. & Karubian, J. O. Impacts of nectar robbing on the foraging ecology of a territorial hummingbird. Behav. Proc. 149, 27–34. https://doi.org/10.1016/j.beproc.2018.01.001 (2018).Article 

Google Scholar 
Boehm, M. A. Biting the hand that feeds you: Wedge-billed hummingbird is a nectar robber of a sicklebill-adapted Andean bellflower. Acta Amazon. 48, 146–150. https://doi.org/10.1590/1809-4392201703932 (2018).Article 

Google Scholar 
Igić, B., Nguyen, I. & Fenberg, P. B. Nectar robbing in the trainbearers (Lesbia, Trochilidae). PeerJ 8, e9561. https://doi.org/10.7717/peerj.9561 (2020).Article 

Google Scholar 
Lunardi, V. D. O., Silva, É. E., Silva, S. T. A. & Lunardi, D. G. Handroanthus impetiginosus (Bignoniaceae) as an important floral resource for synanthropic birds in the Brazilian semiarid. Oecol. Austr. https://doi.org/10.4257/oeco.2019.2301.12 (2019).Article 

Google Scholar 
Almeida, J. M., Missagia, C. C. C. & Alves, M. A. S. Effects of the availability of floral resources and neighboring plants on nectar robbery in a specialized pollination system. Curr. Zool. https://doi.org/10.1093/cz/zoab083 (2021).Article 

Google Scholar 
Rodríguez-Rodríguez, M. C. & Valido, A. Opportunistic nectar-feeding birds are effective pollinators of bird-flowers from Canary Islands: experimental evidence from Isoplexis canariensis (Scrophulariaceae). Am. J. Bot. 95, 1408–1415. https://doi.org/10.3732/ajb.0800055 (2008).Article 
PubMed 

Google Scholar 
Lohmann, L. G. Untangling the phylogeny of neotropical lianas (Bignonieae, Bignoniaceae). Am. J. Bot. 93, 304–318. https://doi.org/10.3732/ajb.93.2.304 (2006).CAS 
Article 
PubMed 

Google Scholar 
Olmstead, R. G., Zjhra, M. L., Lohmann, L. G., Grose, S. O. & Eckert, A. J. A molecular phylogeny and classification of Bignoniaceae. Am. J. Bot. 96, 1731–1743. https://doi.org/10.3732/ajb.0900004 (2009).CAS 
Article 
PubMed 

Google Scholar 
Lohmann, L. G. & Taylor, C. M. A new generic classification of tribe Bignonieae (Bignoniaceae). Ann. Mo. Bot. Gard. 99, 348–489. https://doi.org/10.3417/2003187 (2014).Article 

Google Scholar 
Gentry, A. H. Coevolutionary patterns in Central American bignoniaceae. Ann. Mo. Bot. Gard. 61, 728–759. https://doi.org/10.2307/2395026 (1974).Article 

Google Scholar 
Bertin, R. I. Floral biology, hummingbird pollination and fruit production of trumpet creeper (Campsis radicans, Bignoniaceae). Am. J. Bot. 69, 122–134. https://doi.org/10.2307/2442837 (1982).Article 

Google Scholar 
Bertin, R. I. Paternity and fruit production in trumpet creeper (Campsis radicans). Am. Nat. 119, 694–709. https://doi.org/10.1086/283943 (1982).Article 

Google Scholar 
Bertin, R. I. & Sullivan, M. Pollen interference and cryptic self-fertility in Campsis radicans. Am. J. Bot. 75, 1140–1147. https://doi.org/10.1002/j.1537-2197.1988.tb08827.x (1988).Article 

Google Scholar 
Bertin, R. I. Paternal success following mixed pollinations of Campsis radicans. Am. Midl. Nat. 124, 153–163. https://doi.org/10.2307/2426088 (1990).Article 

Google Scholar 
Bertin, R. I. Effects of pollination intensity in Campsis radicans. Am. J. Bot. 77, 178–187. https://doi.org/10.1002/j.1537-2197.1990.tb13544.x (1990).Article 
PubMed 

Google Scholar 
Bertin, R. I. & Peters, P. J. Paternal effects on offspring quality in Campsis radicans. Am. Nat. 140, 166–178. https://doi.org/10.1086/285408 (1992).Article 

Google Scholar 
Kartesz, J. T. Campsis radicans. Floristic Synthesis of North America, Version 1.0. Biota of North America Program (BONAP) http://bonap.net/MapGallery/County/Campsis%20radicans.png. (2015).Kolodziejska-Degorska, I. & Zych, M. Bees substitute birds in pollination of ornitogamous climber Campsis radicans [L.] Seem in Poland. Acta Soc. Botanicorum Poloniae 75, 79–85 (2006).Article 

Google Scholar 
Catesby, M. The Natural History of Carolina, Florida and the Bahama islands. Volume 1. (Published by the author, 1731).Audubon, J. J. Ornithological Biography Vol. 3, 638 (Adam and Charles Black, 1835).
Google Scholar 
Audubon, J. J. Ruby-throated Hummingbird, plate CCLIII, The Birds of America Vol. 3 (Havell, 1835).
Google Scholar 
Nuttall, T. Manual of the Ornithology of the United States and of Canada. The Land Birds (Hilliard and Brown, 1832).
Google Scholar 
Stiles, F. G. & Freeman, C. E. Patterns in floral nectar characteristics of some bird-visited plant species from Costa Rica. Biotropica 25, 191–205. https://doi.org/10.2307/2389183 (1993).Article 

Google Scholar 
Stiles, F. G. Ecology, flowering phenology, and hummingbird pollination of some Costa Rican Heliconia species. Ecology 56, 285–301. https://doi.org/10.2307/1934961 (1975).Article 

Google Scholar 
McDade, L. A. & Weeks, J. A. Nectar in hummingbird-pollinated Neotropical plants I: Patterns of production and variability in 12 species. Biotropica 36, 196–215. https://doi.org/10.1111/j.1744-7429.2004.tb00312.x (2004).Article 

Google Scholar 
Wunderle, J. M. Jr. Nectar robbing by Orchard Orioles. Chat 44, 107–108 (1980).
Google Scholar 
Tyler, W. M. in Life histories of North American blackbirds, orioles, tanagers, and allies. Order Passeriformes: Families Ploceidae, Icteridae, and Thraupidae. United States National Museum Bulletin 211 (ed Arthur Cleveland Bent) 247–270 (United States Government Printing Office, 1958).George, F. W. Baltimore Orioles destroying trumpet vine blossoms. Wilson Bull. 46, 64 (1934).
Google Scholar 
Ridgway, R. The birds of North and Middle America, Part 2. Bull. U.S. Natl. Mus. 50, 1–834 (1902).
Google Scholar 
Scharf, W. C. & Kren, J. In Birds of the World (ed. Poole, A. F.) (Cornell Lab of Ornithology, 2020).
Google Scholar 
Morton, E. S. Effective pollination of Erythrina fusca by the Orchard Oriole (Icterus spurius): Coevolved behavioral manipulation?. Ann. Mo. Bot. Gard. 66, 482–489. https://doi.org/10.2307/2398840 (1979).Article 

Google Scholar 
Dickey, D. R. & van Rossem, A. J. The birds of El Salvador. Field Mus. Publ. Zool. 23, 1–609 (1938).
Google Scholar 
Baumel, J. J., King, A. S., Breazile, J. E., Evans, H. E. & Vanden Berge, J. C. (eds). Handbook of Avian Anatomy: Nomina Anatomica Avium, Second Edition. Publications of the Nuttall Ornithological Club no. 23 (Nuttall Ornithological Club, 1993).Beecher, W. J. Adaptations for food-getting in the American blackbirds. Auk 68, 411–440. https://doi.org/10.2307/4080840 (1951).Article 

Google Scholar 
Zusi, R. The role of the depressor mandibulae muscle in kinesis of the avian skull. Proc. U.S. Natl. Mus. 123, 1–28 (1967).Article 

Google Scholar 
Remsen, J. V. Jr. & Robinson, S. K. A classification scheme for foraging behavior of birds in terrestrial habitats. Stud. Avian Biol. 13, 144–160 (1990).
Google Scholar 
Skutch, A. F. Orioles, Blackbirds, and Their Kin (University of Arizona Press, 1996).
Google Scholar 
Hansell, M. P. Bird nests and Construction Behaviour 294 (Cambridge University Press, 2000).Book 

Google Scholar 
Bent, A. C. Life histories of North American blackbirds, orioles, tanagers, and allies. Bull. U.S. Natl. Museum 211, 1–531 (1958).
Google Scholar 
Dennis, J. V. Observations on the orchard oriole in lower Mississippi Delta. Bird-Banding 19, 12–21. https://doi.org/10.2307/4509997 (1948).Article 

Google Scholar 
Wunderle, J. M. & Lodge, D. J. The effect of age and visual cues on floral patch use by bananaquits (Aves: Emberizidae). Anim. Behav. 36, 44–54. https://doi.org/10.1016/S0003-3472(88)80248-3 (1988).Article 

Google Scholar 
Edge, A. A. Characteristics of nectar production and standing crop in Campsis radicans (Bignoniaceae). MSc thesis. (East Tennessee State University, 2010).Galetto, L. Nectary structure and nectar characteristics in some Bignoniaceae. Plant Syst. Evol. 196, 99–121. https://doi.org/10.1007/BF00985338 (1995).Article 

Google Scholar 
Elias, T. S. & Gelband, H. Nectar: Its production and functions in trumpet creeper. Science 189, 289–291. https://doi.org/10.1126/science.189.4199.289 (1975).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Elias, T. S. & Gelband, H. Morphology and anatomy of floral and extrafloral nectaries in Campsis (Bignoniaceae). Am. J. Bot. 63, 1349–1353. https://doi.org/10.1002/j.1537-2197.1976.tb13220.x (1976).Article 

Google Scholar 
Hermans, M. & Rasson, J. P. A new Sobolev test for uniformity on the circle. Biometrika 72, 698–702. https://doi.org/10.2307/2336748 (1985).MathSciNet 
Article 
MATH 

Google Scholar 
Landler, L., Ruxton, G. D. & Malkemper, E. P. The Hermans-Rasson test as a powerful alternative to the Rayleigh test for circular statistics in biology. BMC Ecol. 19, 30. https://doi.org/10.1186/s12898-019-0246-8 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
RStudio Team. RStudio: Integrated Development for R. PBC, Boston, MA http://www.rstudio.com/. (RStudio 2020).Beecher, W. J. Convergent evolution in the American orioles. Wilson Bulletin 62, 50–86 (1950).
Google Scholar 
Wolf, L. L., Hainsworth, F. R. & Stiles, F. G. Energetics of foraging: Rate and efficiency of nectar extraction by hummingbirds. Science 176, 1351–1352. https://doi.org/10.1126/science.176.4041.1351 (1972).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Wolf, L. L., Hainsworth, F. R. & Gill, F. B. Foraging efficiencies and time budgets in nectar-feeding birds. Ecology 56, 117–128. https://doi.org/10.2307/1935304 (1975).Article 

Google Scholar 
Alcantara, S. & Lohmann, L. G. Evolution of floral morphology and pollination system in Bignonieae (Bignoniaceae). Am. J. Bot. 97, 782–796. https://doi.org/10.3732/ajb.0900182 (2010).Article 
PubMed 

Google Scholar 
Gentry, A. H. Bignoniaceae: Part II (Tribe Tecomeae). Flora Neotrop. 25, 1–370 (1992).
Google Scholar 
Grant, V. Historical development of ornithophily in the western North American flora. Proc. Natl. Acad. Sci. 91, 10407–10411. https://doi.org/10.1073/pnas.91.22.10407 (1994).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
James, R. L. Some hummingbird flowers east of the Mississippi. Castanea 13, 97–109 (1948).
Google Scholar 
Van Nest, B. N., Edge, A. A., Feathers, M. V., Worley, A. C. & Moore, D. Bees provide pollination service to Campsis radicans (Bignoniaceae), a primarily ornithophilous trumpet flowering vine. Ecol. Entomol. 46, 117–127. https://doi.org/10.1111/een.12947 (2021).Article 

Google Scholar 
Patuxent Wildlife Research Center. Orchard oriole Icterus spurius. BBS summer distribution map, 2011–2015 (relative abundance map). https://www.mbr-pwrc.usgs.gov/bbs/ra2015/ra2015_red_v3.shtml (accessed 7 March 2021) (2021). More

Remarks on the residual color patterns in the Kitadani Freshwater BivalvesResidual color patterns in the form of visible pigmentation on fossil molluscan shells are generally uncommon2,3. In the Paleozoic to Mesozoic fossil records, the color patterns were limited to marine species3, which are preserved as black to dark-colored bands running on the shell surface as melanin pigments20,21. The black to dark-colored stripes on the shells of the Kitadani Freshwater Bivalves resemble the color patterns in some extant freshwater bivalves, suggesting that the dark bands are residual color patterns remaining as melanin pigments. Consequently, the Kitadani Freshwater Bivalves represents the oldest and second fossil record of residual color patterns among fossil freshwater bivalves.The residual color patterns of the Kitadani Freshwater Bivalves resemble the color patterns of extant freshwater bivalves in terms of width, number, and distribution of the colored bands. Both the Kitadani Freshwater Bivalves and extant freshwater bivalves examined in this study consist of two types of color patterns: stripes along the growth lines and radial rays tapered toward the umbo. Notably, the former pattern is similar among all the species examined, as it forms in the peripheries of prominent growth lines occurring periodically. In the latter pattern, however, the morphology and distribution of the bands are slightly different between the Kitadani Freshwater Bivalves and the extant species. The Kitadani Freshwater Bivalves exhibits relatively distinct and wide radial rays running roughly parallel to the lengths of the sculpture elements (radial plications and/or wrinkles), while the extant species bear obscure and fine radial rays running diagonally to the lengths of the sculpture elements. Nonetheless, the taxa with V-shaped sculpture elements (wrinkles, ribs or arranged nodules) lack or bear ambiguous radial rays, whether extant (e.g., Triplodon spp., Indochinella spp. and Tritogonia spp.)13,15,22 or extinct (†Trigonioides tetoriensis).Hypothesis I: phylogenetic constraintsThe resemblance of the color patterns between the Kitadani Freshwater Bivalves and the extant unionids possibly resulted from the phylogenetic constrains. Each of the three species of the Kitadani Freshwater Bivalves belongs to a separate family (†Trigonioides tetoriensis: †Trigonioididae, †Plicatounio naktongensis: †Plicatounionidae, and †Matsuomtoina matsumotoi: †Pseudohyriidae) in the order Trigoniida19. Trigoniida in turn, forms the subclass Palaeoheterodonta with Unionida23. This raises a possibility that the color patterns observed in the Kitadani Freshwater Bivalves and the extant unionids is inherited from their most recent common ancestor. In other words, these color patterns, stripes along the growth lines and radial rays tapered toward the umbo, may be the apomorphy for Palaeoheterodonta. In fact, some extant trigoniid species belonging to Neotrigonia exhibit color pattern similar to those in the Kitadani Freshwater Bivalves and extant unionids in this study (e.g. Neotrigonia margaritacea)1.Interestingly, the coloration of color patterns is quite different between unioniids (green to blue colorings) and trigoniids (red to yellow colorings), and the oldest known color patterns of the Palaeoheterodonta (Myophorella nodulosa, a marine species of Trigoniida from the Oxfordian of the Early Jurassic) appears different (concentric rows of patches)10 from those of the Kitadani Freshwater Bivalves or the extant unioniids. These observations suggest that colorations evolved independently, in contrast to the color patterns, between Trigoniida and Unionida, and that Trigoniida more diverse color patterns than Unionida did in the Palaeoheterodont evolutionary history. Although further examination of the fossil record for the residual colors and color patterns in Palaeoheterodonta is essential, it is plausible that the habitat differences may have caused such discrepancy in the colorations and color patterns between Trigoniida (mainly marine) and Unionida (freshwater) in spite of the phylogenetic constrains.Hypothesis II: convergent evolutionThe other possible interpretation of the color pattern similarity between the Kitadani Freshwater Bivalves and extant Unionida, is the convergent evolution. One potential factor that may have caused this convergent evolution of the color patterns is an adaptation to their habitats. In general, much of the convergent evolution in animals occurs through the morphological evolution in response to their habitats24. Similarly in mollusks, shell colors and their patterns are generally influenced by their habitats2,6,25. Considering marine mollusks, the shell colors and their patterns have great diversity due to varying habitat environments, especially in coral reeves that exhibit various colors and complex ecosystem2,6. Conversely, in the freshwater ecosystem, the environmental colors are relatively monotonous with rocks, sand, mud, and green algae8, and such habitat conditions are likely indifferent between the Mesozoic and Cenozoic. As a result, the freshwater bivalves retained simple and monotonous color patterns for adapting to such environments throughout their evolution.Another conceivable factor to explain the convergent evolution in the color patterns of the studied freshwater bivalves is the selection pressure by visual predators. In general, the shell colors and their patterns in bivalves act as camouflages against the predators2,7,8,26,27,28. Previous studies have demonstrated that extant freshwater bivalves are preyed upon by crayfish, fish, birds, reptiles, and mammals29,30. Because shell colors in freshwater bivalves tend to be greenish, such colors may be an adaptation against visual predators for blending into the freshwater sediments on which abundant greenish phytoplanktons occur2,8. Therefore, the evolutionary conservatism in color patterns of freshwater bivalves may result from camouflages into freshwater microenvironments, which has been advantageous against visual predators since the late Early Cretaceous.The above discussion assumes that the visual predators of freshwater bivalves remained similar for at least 120 million years. Which animals could have been potential threads to the Kitadani Freshwater Bivalves, and, in turn, the Early Cretaceous freshwater bivalves? Among the extant visual predators of the freshwater bivalves, those whose lineages were present in the Early Cretaceous include crustaceans (especially brachyuran decapoda31), fish, lizards, turtles, crocodiles, birds, and mammals. Among them, the fossil record of durophagous lizards and mammals can be traced back only to the Late Cretaceous32,33. Conversely, lines of fossil evidence suggest that some fish34,35, turtles36, and crocodiles35 fed on molluscan invertebrates during the Early Cretaceous, and the Kitadani Freshwater Bivalves indeed occurs with abundant lepisosteiform scales, testudinate shells and crocodile teeth. Additionally, at least one Early Cretaceous avian species with crustacean gut contents can be attributed to the durophagous diet37, and the Kitadani Formation has yielded avialan skeletal remains38, and footprints39,40. Therefore, fish, turtles, crocodiles, and birds are likely candidates for visual predators of the Early Cretaceous freshwater bivalves, and have remained so until present. Additionally, while crustaceans have not been identified in the Kitadani Formation, they flourished in the Early Cretaceous and their remains occur with the fossil freshwater bivalves of the time elsewhere31. Thus, crustaceans may have also played a role as visual predators of the freshwater bivalves since the Early Cretaceous.In addition to the crustaceans, fishes, turtles, crocodiles and birds, the visual predators of the Early Cretaceous freshwater bivalves likely include extinct lineages. For example, some pliosauroid plesiosaurs are suggested as being durophagous34, although the freshwater members of the group are considered endemic41 and less likely to be a major thread to the Early Cretaceous freshwater bivalves. Another extinct candidate is non-avian dinosaurs. Ornithischians are suggested to have possessed a dietary flexibility including the durophagy. For instance, well-preserved hadrosaurid coprolites from the Late Cretaceous of Montana, U.S.A. include sizeable crustaceans and mollusks, possibly suggesting that the Cretaceous freshwater mollusks were consumed by these herbivorous dinosaurs42. In addition, some basal ceratopsian psittacosaurids are hypothesized for the durophagy based on the predicted large bite force in the caudal portion of the toothrow43. Among saurischians, some oviraptorosaurian theropods are indicated to consume mollusks with hard shells based on their mandibular features44. While hadrosaurids, psittacosaurids, and oviraptorosaurians have not been identified in the Kitadani Formation, psittacosaurids, and oviraptorosaurians are common elsewhere in the Early Cretaceous of East Asia45,46, and hadrosauroid Koshisaurus is present in the formation47. Because dinosaurs occupied a niche of large terrestrial predators throughout the Mesozoic, they may have acted as one of major mollusk consumers in absence of large lizards and mammals in the Early Cretaceous ecosystem. Thus, the predation pressure by visual predators to the freshwater bivalves in the Early Cretaceous is likely similar to that in the present. Consequently, one of evolutionary adaptations of the freshwater bivalves against such pressure has remained to camouflage in the phytoplankton-rich sediments, leading to the long-term evolutionary conservatism of their color patterns. More

Rio Conservation and Sustainability Science Centre, Department of Geography and the Environment, Pontifical Catholic University, Rio de Janeiro, BrazilBernardo B. N. Strassburg, Alvaro Iribarrem, Carlos Leandro Cordeiro, Renato Crouzeilles, Catarina Jakovac, André Braga Junqueira, Eduardo Lacerda & Agnieszka E. LatawiecInternational Institute for Sustainability, Rio de Janeiro, BrazilBernardo B. N. Strassburg, Alvaro Iribarrem, Carlos Leandro Cordeiro, Renato Crouzeilles, Catarina Jakovac, André Braga Junqueira, Eduardo Lacerda, Agnieszka E. Latawiec, Robin L. Chazdon & Carlos Alberto de Mattos ScaramuzzaPrograma de Pós Graduacão em Ecologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, BrazilBernardo B. N. Strassburg, Renato Crouzeilles & Fabio R. ScaranoBotanical Garden Research Institute of Rio de Janeiro, Rio de Janeiro, BrazilBernardo B. N. StrassburgSchool of Biological Sciences, University of Queensland, St Lucia, Queensland, AustraliaHawthorne L. BeyerAgricultural Science Center, Federal University of Santa Catarina, Florianópolis, BrazilCatarina JakovacInstitut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona, Barcelona, SpainAndré Braga JunqueiraDepartment of Geography, Fluminense Federal University, Niterói, BrazilEduardo LacerdaDepartment of Production Engineering, Logistics and Applied Computer Science, Faculty of Production and Power Engineering, University of Agriculture in Kraków, Kraków, PolandAgnieszka E. LatawiecSchool of Environmental Sciences, University of East Anglia, Norwich, UKAgnieszka E. LatawiecDepartment of Zoology, University of Cambridge, Cambridge, UKAndrew Balmford, Stuart H. M. Butchart & Paul F. DonaldInternational Union for Conservation of Nature (IUCN), Gland, SwitzerlandThomas M. BrooksWorld Agroforestry Center (ICRAF), University of The Philippines, Los Baños, The PhilippinesThomas M. BrooksInstitute for Marine & Antarctic Studies, University of Tasmania, Hobart, Tasmania, AustraliaThomas M. BrooksBirdLife International, Cambridge, UKStuart H. M. Butchart & Paul F. DonaldDepartment of Ecology and Evolutionary Biology, University of Connecticut, Storrs, CT, USARobin L. ChazdonWorld Resources Institute, Global Restoration Initiative, Washington, DC, USARobin L. ChazdonTropical Forests and People Research Centre, University of the Sunshine Coast, Sippy Downs, Queensland, AustraliaRobin L. ChazdonInstitute of Social Ecology, University of Natural Resources and Life Sciences Vienna, Vienna, AustriaKarl-Heinz Erb & Christoph PlutzarDepartment of Forest Sciences, ‘Luiz de Queiroz’ College of Agriculture, University of São Paulo, Piracicaba, BrazilPedro BrancalionRSPB Centre for Conservation Science, Royal Society for the Protection of Birds, Edinburgh, UKGraeme Buchanan & Paul F. DonaldSecretariat of the Convention on Biological Diversity (SCBD), Montreal, Quebec, CanadaDavid CooperInstituto Multidisciplinario de Biología Vegetal, CONICET and Universidad Nacional de Córdoba, Córdoba, ArgentinaSandra DíazUnited Nations Environment Programme World Conservation Monitoring Centre, Cambridge, UKValerie Kapos & Lera MilesBiodiversity and Natural Resources (BNR) program, International Institute for Applied Systems Analysis (IIASA), Laxenburg, AustriaDavid Leclère, Michael Obersteiner & Piero ViscontiDivision of Conservation Biology, Vegetation Ecology and Landscape Ecology, University of Vienna, Vienna, AustriaChristoph PlutzarB.B.N.S. wrote the first version of the paper. All authors provided input into subsequent versions of the manuscript. More

Climate driversTo explore the impact of climate on forest resilience (see the following sections), we used monthly averaged total precipitation, 2-m air temperature, evapotranspiration deficit and surface solar radiation downwards acquired from the ERA5-Land reanalysis product at 0.1° spatial resolution for the 2000–2020 period (https://cds.climate.copernicus.eu/cdsapp#!/home). Evapotranspiration deficit was quantified as the total precipitation minus evapotranspiration. In this study, we referred to climate regions as defined by the Köppen–Geiger world map of climate classification51 (http://koeppen-geiger.vu-wien.ac.at/present.htm). The original 31 climatic zones were merged into major zones and only those characterized by vegetation cover were included in our study (tropical, arid, temperate and boreal; Extended Data Fig. 8).Vegetation dynamicsNDVI data acquired from the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument aboard the Terra satellite was used to derive changes in global vegetation for the period 2000–2020. We used cloud-free spatial composites provided at 16-day temporal resolution and 0.05° spatial resolution (MOD13C1 Version 6; https://lpdaac.usgs.gov/products/mod13c1v006/) and retained only pixels with good and marginal overall quality. The MODIS-derived NDVI dataset represents a state-of-the-art product of vegetation state whose retrieval algorithm is constantly improved52, and being derived from a unique platform and sensor, it is temporally and spatially consistent. Vegetation dynamics were analysed in terms of kNDVI, a nonlinear generalization of the NDVI based on ref. 22 and derived as follows:$$text{kNDVI=}tanh left({text{NDVI}}^{2}right)$$
(1)
kNDVI has recently been proposed as a strong proxy for ecosystem productivity that shows high correlations with both plot level measurements of primary productivity and satellite retrievals of sun-induced fluorescence22. In addition, kNDVI has been documented to be more closely related to primary productivity, to be resistant to saturation, bias and complex phenological cycles, and to show enhanced robustness to noise and stability across spatial and temporal scales compared to alternative products (for example, NDVI and near-infrared reflectance of vegetation). For these reasons, it has been retained in this study as the preferred metric to describe the state of the forest ecosystem.To obtain an accurate estimate of resilience indicators, vegetation time series need to be stationary without seasonal periodic patterns or long-term trends53. To this aim, vegetation anomalies were obtained from kNDVI data by first subtracting the multi-year 16-day sample mean and then removing linear trends from the resulting time series. Missing data, due for instance to snow cover affecting the retrieval of reflectance properties, have been gap-filled by climatological kNDVI values. The time series of kNDVI-based vegetation anomalies was used to derive resilience indicators and assess their spatial and temporal variations (see next sections).Interannual changes in vegetation were assessed in terms of growing-season-averaged kNDVI. To this end, a climatological growing season that spanned months with at least 75% of days in the greenness phase was derived from the Vegetation Index and Phenology satellite-based product54 (https://vip.arizona.edu/) and acquired for the 2000–2016 period at 0.05° spatial resolution. In addition, forest cover (FC) fraction was derived from the annual land-cover maps of the European Space Agency’s Climate Change Initiative (https://www.esa-landcover-cci.org/)55 over the 2000–2018 period at 300-m spatial resolution. FC was retrieved by summing the fraction of broadleaved deciduous, broadleaved evergreen, needle leaf deciduous and needle leaf evergreen forest. FC was resampled to 0.05° to match the kNDVI spatial resolution.Spatial patterns of slowness and its dependence on environmental factorsIn this study, we quantified the resilience of forest ecosystems—their ability to recover from external perturbations—by the use of the 1-lag TAC (refs. 3,4,5). Such an indicator was initially computed on the whole time series of vegetation anomalies (2000–2020) for forest pixels with less than 50% missing data in the original NDVI and FC greater than 0.05 and referred to in the text as long-term TAC. This analysis was used to assess the spatial patterns of the forest slowness mediated by environmental factors that affect plant growth rates and capacity to recover from perturbations. The long-term TAC was explored both in the geographic and climate space (Extended Data Fig. 1). In the climate space, long-term TAC was binned in a 50 × 50 grid as a function of average annual precipitation and temperature, both computed over the 2000–2020 period, using the average as an aggregation metric weighted by the areal extents of each record. We retained only bins with at least 50 records.To explore the potential drivers of long-term TAC, we developed an RF regression model23 and predicted the observed long-term TAC (response variable) based on a set of environmental features (predictors). The use of machine learning in general and of RF in particular, being nonparametric and nonlinear data-driven methods, does not require a priori assumptions about the functional form relating the key drivers and the response functions. The environmental variables include vegetation properties (FC and growing-season-averaged kNDVI) and climate variables (total precipitation, 2-m air temperature, evapotranspiration deficit and surface solar radiation downwards). Each of the climate variables was expressed in terms of average, coefficient of variation and 1-lag autocorrelation and resampled to 0.05° spatial resolution to match the spatial resolution of kNDVI. All environmental variables were computed annually and then averaged over time, except the autocorrelation that was computed directly for the whole period, analogously to the long-term TAC. This resulted in a set of 14 predictors representing the forest density, the background climate, the climate variability and its TAC in the observational period (Extended Data Table 1). The RF model was developed by splitting the observed long-term TAC into two separate samples: 60% of records were used for model calibration, and the remaining 40% were used to validate model performances in terms of coefficient of determination (R2), mean squared error and percentage bias (PBIAS). Each record refers to a 0.05° pixel. The RF implemented here uses 100 regression trees, whose depth and number of predictors to sample at each node were identified using Bayesian optimization. The general model formulation is as follows:$$text{TAC},=,fleft(Xright)+{varepsilon }_{{rm{f}}}$$
(2)
in which f is the RF regression model, X are the environmental predictors and εf are the residuals. We found that the model explains 87% of the spatial variance (R2) of the observed long-term TAC with a mean squared error of 0.007 and an average overestimation of 0.058 (PBIAS; Extended Data Fig. 2a). By definition, machine learning methods are not based on the mechanistic representation of the phenomena and therefore cannot provide direct information on the underlying processes influencing the system response to drivers. However, some model-agnostic methods can be applied to gain insights into the outputs of RF models. Here we used variable importance metrics to quantify and rank how individual environmental factors influence TAC (Extended Data Fig. 2b). Furthermore, using partial dependence plots derived from the machine learning algorithm RF, we explored the ecosystem response function (TAC) across gradients of vegetation and climate features (Supplementary Discussion 1 and Extended Data Fig. 2c–f).CSD indicatorsTo explore the temporal variation in forest resilience, we used CSD indicators, here quantified in terms of temporal changes in TAC retrieved for two consecutive and independent periods ranging from 2000 to 2010 and from 2011 to 2020, and assessed the significance of the change in the sampled mean aggregated for different climate regions through a two-sided t-test (Fig. 1c). This analysis was complemented by the computation of TAC on the annual scale over a 2-year lagged temporal window (3-year window size) to track the temporal changes in CSD. This resulted in a time series of TAC with an annual time step.We point out that temporal dynamics of annual TAC are driven by two processes: the changes in the resilience of the system that affect the velocity of the recovery from external perturbations and the confounding effects of the changes in autocorrelation of the climate drivers (Xac) that directly affect the autocorrelation of NDVI. Given the specific goals of this study, we factored out the second process from the total TAC signal to avoid that an increasing autocorrelation in the drivers would affect our analysis and conclusions about the resilience and the potential increase in instability56. For this purpose, we disentangled the temporal changes in TAC due to variations in autocorrelation in the climate drivers (({rm{TAC}}| {X}_{{rm{ac}}})) by adopting the space-for-time analogy and applied the RF model (f) at an annual time step (t) in a set of factorial simulations as follows:$${text{TAC}}^{t},{rm{| }},{X}_{{rm{ac}}}=fleft({X}^{t}right)-fleft({X}_{-{rm{ac}}}^{t},{X}_{{rm{ac}}}^{2000}right)$$
(3)
The first term on the right side of equation (3) is the RF model simulation obtained by accounting for the dynamics of all predictors, and the second term is the RF model simulation generated by considering all predictors dynamic except the factors of autocorrelation in climate that are kept constant to their first-year value (year 2000). For such runs, we used predictors computed on an annual scale over a 2-year lagged temporal window, consistently to the TAC time series. We found that the direct effects of autocorrelation in climate have led to a positive trend of TAC in dry zones (due to the increasing autocorrelation of the drivers in these regions) and to an opposite effect in temperate humid forests (Supplementary Fig. 3). To remove these confounding effects, the estimated term ({{rm{TAC}}}^{t}| {X}_{{rm{ac}}}) is factored out from the TACt by subtraction to derive an enhanced estimate of annual resilience that is independent of autocorrelation in climate (Extended Data Fig. 3).Long-term linear trends computed on the resulting enhanced TAC time series (δTAC) represent our reference CSD indicator used in this study to explore the changes in forest resilience. δTAC was quantified for each grid cell (Fig. 1a) and represented in the climate space following the methodology previously described (Fig. 1b). We then assessed the significance of the trends at bin level by applying a two-sided t-test for the sampled trend distributions within each bin. This significance test is independent from the structural temporal dependencies originating from the use of a 2-year lagged temporal window to compute the TAC time series.Following an analogous approach described in equation (3), we disentangled the effect of the variation in forest density, background climate and climate variability on temporal changes in TAC (Fig. 1d,e). We recognize that other environmental factors not explicitly accounted for in our RF model could play a role in modulating the temporal variations in TAC. However, given the comprehensiveness of the suite of predictors used in equation (2) (Extended Data Table 1), it seems plausible that residuals mostly reflect the intrinsic forest resilience, the component intimately connected to the short-term responses of forests to perturbations, which is not directly related to climate variability. Forest ecosystem evolutionary processes could also play a role, but longer time series would be required to reliably capture these dynamics. Furthermore, abrupt declines (ADs) in the vegetation state and following recoveries, similarly to those potentially originating from forest disturbances (for example, wildfires and insect outbreaks), could influence the TAC changes. However, such occurrences, being distributed across the globe throughout the whole period, are expected to only marginally affect the resulting trend in TAC time series.Sensitivity analysisTo assess the robustness of our results with respect to the modelling choices described above, we performed a series of sensitivity analyses for the difference in TAC retrieved for the two independent periods (2000–2010 and 2011–2020). To this aim, we tested their dependence on: the quality flag of the NDVI data used for the analyses (good, good and marginal); the gap-filling procedure tested on different periods (year and growing season); the inclusion or exclusion of forest areas affected by ADs; the threshold on the maximum percentage of missing NDVI data allowed at the pixel level (20%, 50% and 80%); the threshold on the minimum percentage of FC allowed at the pixel level (5%, 50% and 90%); and the pixel spatial resolution used for the analyses (0.05°, 0.25° and 1°). In addition, we tested the sensitivity of the trend in total TAC signal on the moving temporal window length used to calculate autocorrelation at lag 1. Results obtained for the different configurations were compared in terms of frequency distributions, separately for climate regions (Extended Data Fig. 4), and further explored in the climate space (Extended Data Figs. 5 and 6). Outcomes of the sensitivity analysis are discussed in Supplementary Discussion 2.Interplay between GPP and CSDResilience and GPP interact with each other through mutual causal links. On one hand, a reduction in forest resilience makes the system more sensitive to perturbations with potential consequent losses in GPP (ref. 26). On the other hand, a reduction in GPP may lead to a decline in resilience according to the carbon starvation hypothesis, and may be associated with increasing hydraulic failure46. To explore the link between forest resilience and primary productivity, we quantified the correlation between TAC and GPP. Estimates of GPP were derived from the FluxCom Model Tree Ensemble for the 2001–2019 period at 8-daily temporal resolution and 0.0833° spatial resolution and generated using ecosystem GPP fluxes from the FLUXNET network and MODIS remote sensing data as predictor variables36 (http://www.fluxcom.org/). Annual maps of GPP were quantified and resampled to 0.05° to match the temporal and spatial resolution of TAC time series. The Spearman rank correlation (ρ) was then computed between annual GPP and TAC over a 1° spatial moving window to better sample the empirical distribution of the two variables (Fig. 2d). The significance of ρ(GPP,TAC) was assessed over the climate space separately for each bin (Fig. 2e), similarly to the approach used to test the significance of δTAC. Furthermore, we explored the relationships between the trend in GPP (δGPP) and the trend in TAC (δTAC) by clustering the globe according to the directions of the long-term trajectories of the above-mentioned variables (Fig. 2f).Disentangling the impact of forest managementTo characterize TAC on different forest types and disentangle the potential effects originating from forest management, results were separately analysed for intact forests and managed forests. Intact forests were considered those forest pixels constituting the Intact Forest Landscapes57 dataset (https://intactforests.org/). Intact Forest Landscapes identifies the forest extents with no sign of significant human activity over the period 2000–2016 based on Landsat time series. The remaining forests pixels—not labelled as intact—were considered as managed forests (Extended Data Fig. 8). The resulting forest type map is consistent with those used for United Nations Framework Convention on Climate Change reporting58, although with more conservative estimates of intact forests in the boreal zone due to the masking based on FC and percentage of missing data applied in this study.We analysed the differences in long-term TAC (computed for the whole 2000–2020 period) between managed and intact forests by masking out the potential effect of climate background. To this aim, we compared the climate spaces generated separately for managed and intact forests by extracting only those bins that are covered by both forest classes. The resulting distributions—one for each forest class—have the same sample size, and each pair of elements shares the same climate background. Potential confounding environmental effects on average recovery rates are, therefore, minimized. We then applied a two-sided t-test for analysing the significance of the difference in the sampled means (Fig. 2a). An analogous approach was used to test the differences in δTAC and ρ(GPP,TAC) between managed and intact forests (Fig. 2b,c).Early-warning signals of abrupt forest declinesWhen forest ecosystems are subject to an extended and progressive degradation, the loss of resilience can lead to an AD (refs. 3,4,5). Such abrupt changes can trigger a regime shift (tipping point) depending on the capacity of the system to recover from the perturbations (Supplementary Methods 1 and 2). We investigated the potential of changes in TAC as early-warning signals of ADs in intact forests over the 2010–2020 period. To this aim, we quantified at the pixel level ADs as the events occurring on a certain year when the corresponding growing-season average kNDVI was more than n-times local standard deviation below the local mean. Local mean and standard deviation (σ) were computed over the 10-year antecedent temporal window (undisturbed) period and n varies between 1 and 6 with higher values reflecting more severe changes in the state of the system. For each pixel and for each fixed n value, we recorded only the first AD occurrence, thus imposing a univocal record for each abrupt change in the state of the system.We then explored whether the retrieved ADs were statistically associated with antecedent high values of δTAC. To avoid confusion with the attribution of causality, for each AD that occurred at time t (over the 2010–2020 period), we derived the δTAC over the temporal window 2000 − (t − 1). The resulting trend in δTAC is therefore antecedent and independent of the changes in vegetation associated with the AD. Then, for each pixel with an AD at time t, we also extracted randomly one of the undisturbed (with no AD) adjacent pixels and retrieved δTAC over the same temporal window. This analysis produced two distributions of δTAC associated with pixels with and without ADs (AD and no AD, respectively). The two distributions have the same size and each pair of elements shares similar background climate. We calculated the probability of occurrence of AD conditional on the trend in δTAC (({rm{AD}}| delta {rm{TAC}})) as the frequency of ADs for which (delta {rm{TAC}}left(mathrm{AD}right)| > delta {rm{TAC}}left(mathrm{no; AD}right)), and the significance of the difference in the two sampled means (AD and no AD) was evaluated through a two-sided t-test. Probability and significance were assessed for different climate regions and severity of ADs (Fig. 3a). High statistically significant probabilities suggest that the AD is following the drifting towards a critical resilience threshold plausibly associated with changes in environmental drivers.We complemented the aforementioned analyses by retrieving the tolerance and proximity to AD, hereafter determined for a 3σ severity. We first quantified the TAC that proceeded the occurrence of an AD and followed a progressive loss of resilience as captured by positive δTAC. This value, hereafter referred to as abrupt decline temporal autocorrelation (TACAD), reflects the TAC threshold over which we observed an abrupt change in the forest state (Fig. 3b). The tolerance to AD was quantified as the difference between the local TACAD and the TAC value averaged over the 2000–2009 period to characterize the pre-disturbance conditions. The tolerance metric was explored across a gradient of aridity index59 (Fig. 3c).TACAD can be directly retrieved only on those forest pixels that have already experienced an AD. As a considerable fraction of undisturbed forests could potentially be close to their critical TAC threshold, or even have already passed it, it is important to determine their TACAD. To this aim, we developed an RF regression model that expresses the TACAD as a function of the set X of environmental variables used in model f (equation (2)) but excluding the autocorrelation in climate drivers (Xreduced) already disentangled in the TAC signal. The general formulation is as follows:$${{rm{TAC}}}_{{rm{AD}}}=gleft({X}_{text{reduced}}right)+{varepsilon }_{{rm{g}}}$$
(4)
in which g is the RF regression model, Xreduced are the environmental predictors and εg are the residuals. Implementation, calibration and validation of g follow the same rationale described before for the f model. We found that the RF model explains 50% of the variance (R2) of the observed TACAD, with a mean squared error of 0.019 and an average underestimation of 0.86 (PBIAS).The RF model was then used to predict the TACAD over the whole domain of intact forests and served as input to quantify the proximity to AD of undisturbed forest pixels at the end of the observational period (year 2020). Here we defined the proximity metric as the difference between the value of TAC in 2020 and TACAD. Proximity takes negative or zero values when TACAD has already been reached (({{{rm{TAC}}}^{2020}ge {rm{TAC}}}_{{rm{AD}}})) and positive values when there are still margins before reaching the critical threshold (({{{rm{TAC}}}^{2020} < {rm{TAC}}}_{{rm{AD}}})). Together (delta {rm{TAC}} > 0) and ({{{rm{TAC}}}^{2020}ge {rm{TAC}}}_{{rm{AD}}}) therefore represent the most critical conditions, as they indicate that the critical resilience threshold for AD has already been reached and the ecosystem is continuing to lose its capacity to respond to external perturbations. We finally quantified the amount of GPP potentially exposed to such critical conditions by linearly extrapolating the GPP for the year 2020 (available GPP data stop in 2019) and overlaying it on the map of critical conditions (proximity to ({rm{AD}} < 0) and (delta {rm{TAC}} > 0)).Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this paper. More

Benchmark regression resultsParallel trend testThe premise of using DID is that the treatment group and control group meet the assumptions of parallel trend, which means that before ETS is officially implemented, the evolution trend of ecologicalization efficiency of industry of the control group and the experimental group is consistent and does not show a systematic difference. This study uses a more rigorous empirical test in parallel trend test: if the interaction coefficient is not significant and is different from zero before the implementation of ETS; and if the interaction coefficient is significant and is different from zero after the implementation of ETS, it indicates that there is no significant difference in ecologicalization efficiency of industry between the control group and the experimental group before the implementation of ETS. Results are shown in Table 4: before ETS was officially implemented, the difference coefficient was not significant; after the official implementation of ETS in 2013, the difference coefficient was significant and not equal to 0, and the ecologicalization efficiency of industry was improved significantly, which met the parallel trend of the DID. Therefore, it is scientific and reasonable to evaluate the effectiveness of ETS with DID.Table 4 Parallel trend test.Full size tableDynamic effect analysisTo compare the conditions of the experimental group and the control group before and after the implementation of ETS, dynamic graphs are drawn in this study, as shown in Fig. 1, which shows the impact of ETS on the regional ecologicalization efficiency of industry. The vertical line represents a 95% confidence interval and the broken line shows the marginal effect of regional ecologicalization efficiency, which means that the confidence interval contains is 0 before ETS’s implementation, and the result is not significant. In contrast, after 2013, the effect of ETS became apparent, the marginal effect gradually increased and the results became significant, perhaps owing to the implementation of ETS.Figure 1Dynamic analysis diagram.Full size imageThe effect of ETS on ecologicalization efficiency of industryControlling time effect and fixed effect, this study collected the data of pilot and non-pilot provinces of ETS from 2007 to 2019 to analyze the impact of ETS on the regional ecologicalization efficiency of industry and regional heterogeneity. The results are shown in Table 5. According to the results in the first column, ETS has significantly promoted the regional ecologicalization efficiency of industry, and the national implementation of ETS has achieved remarkable results. Compared with the regions that are not ETS pilot areas, the ecologicalization efficiency of industry of pilot provinces and cities has increased by 35%. Results also show that ETS has different effects on the ecologicalization efficiency of industry in different regions. Specifically, ETS significantly promoted regional ecologicalization efficiency of industry in the eastern and central regions, and the efficiency in the eastern region was more significant than that of the central region. However, the impact of ETS on the regional ecologicalization efficiency of industry in the western region was negative which may result from the fact that compared to the central and western regions, the east region has better economic development, advanced technology, and lots of talents that can respond to the implementation of ETS, accelerate the upgrade of industries, and improve the utilization level of regional resources. There are many traditional industries in the central and western regions, and the development of scientific and technological levels as well as the resource utilization efficiency there are relatively slow. Besides, it is difficult for the central and western regions to adapt to ETS in a short-term of time leading to the failure of improving the regional ecologicalization efficiency of industry in a short time.Table 5 Influence of ETS on ecologicalization efficiency of industry.Full size tableRobustness testPropensity matching score—double difference method (PSM-DID)The assumption of homogeneity and randomness between the control group and the experimental group is the premise of using the DID model. However, due to the large economic and regional differences among provinces and cities, there may be systematic differences between the experimental group and the control group, which may cause deviations in the results. Therefore, the data after propensity score matching is used in this study, making the matched individuals have no other significant differences unless they have been treated or not. The dual difference is conducted again to avoid self-selection bias, and the robustness of the above results is verified according to the measurement results. Control variables were used to match characteristic variables, the experimental group was matched with the control group, and the Logit model was adopted to delete the samples that fail to meet the matching criteria. After the matching, there are 168 observation values. The regression results of PSM-DID model show that, ETS has positive effects on the regional ecologicalization of industry (0.460***), which again proves that the conclusion that ETS improves regional ecologicalization of industry efficiency is reliable. The results are shown in Table 6.Table 6 The result of the PSM-DID.Full size tableCounterfactual testTo verify the robustness of the results again, six provinces and cities are randomly selected as experimental groups for multiple tests to construct new dummy variables of ETS, and the DID model was used again to verify the credibility of the above results. Four random samples were conducted in this study, and the results are shown in Table 7. It can be seen that the results are not significant, which also reversely proves that ETS improves the regional ecologicalization efficiency of industry.Table 7 Counterfactual test results.Full size tableActing pattern analysis of ETS on the regional ecologicalization efficiency of industryFirst, ETS may improve the regional ecologicalization efficiency of industry through industrial structure optimization and upgrading. Promoting upgrading of the industrial structure is one of the important approaches of social and economic development during the 14th Five-Year Plan formulation and is the only way to promote low-carbon and sustainable development of modern national industries. The upgrading of the industrial structure has been promoted to the national strategic level, contributing to the healthy development of the national economy system. ETS bring costs and benefits to enterprises, forcing them to transform and upgrade, increase investment in environmental protection and use clean energy, and accelerate the pace of energy conservation and emission reduction31. Second, ETS may improve the regional ecologicalization efficiency of industry through the coordinated agglomeration of resources. Marshall’s theory of scale economy, Krugman’s theory of new economic geography, Weber’s theory of agglomeration economy, Coase’s transaction cost theory, and so on reflect the importance of resource aggregation of economic activities through cost-saving, resource sharing, and other ways to improve industrial input–output efficiency, enhance industrial competitiveness, increase regional comprehensive strength and strengthen the competitive advantage of regional industrial clusters32. The benefits generated by resource aggregation far exceed the sum of benefits generated by various industries in the decentralized state. Under the pressure of ETS, enterprises may alleviate the mismatch between labor and capital through the collaborative aggregation of industrial resources, aiming to improve economic benefits and regional resource allocation efficiency and promote regional ecologicalization efficiency of industry. Third, ETS may improve the regional ecologicalization efficiency of industry by supporting ecological optimization. The sustainable development of the ecological environment is closely related to emission reduction policy. To alleviate the bad effects on the ecology, environmental protection is more and more brought to the attention of society and government. Policies for ecological protection have been introduced to reduce pollution20. All regions take effective and targeted measures to control environmental pollution and optimize the investment structure in light of their actual conditions. The purpose of ecological optimization is to improve the regional environment and strengthen pollution control which is one of the important parts of China’s fiscal spending. The government must guide the market to carry out ecological protection and environmental governance according to ETS. Studies have found that a low-carbon pilot policy helps to enhance the level of regional pollution control, promote the harmonious development of regional economy and environment, and then improve the regional ecologicalization efficiency of industry.To explore the transmission mechanism of ETS on the regional ecologicalization of industry efficiency, Baron and Kenny (1986)’s mediating effect model was referred to explore and verify whether there exists a structural optimization upgrade effect, resource synergistic agglomeration effect, ecological optimization support effect when ETC promotes regional ecologicalization efficiency of industry. Table 8 shows the regression results of the influence mechanism of ETS on the regional ecologicalization efficiency of industry. This study refers to the definition and research of industrial optimization and upgrading by Wang Qunwei, Huang Xianglan, and others, and the proportion of tertiary industry added value accounting for industrial added value is selected to measure the effectiveness of industrial optimization and upgrading. For resource synergistic agglomeration effect, this study refers to the calculation methods of Cui Shuhui, Chen Jianjun et al. and adopts the collaborative aggregation index of manufacturing and producer services to measure the collaborative aggregation effect of resources, which effectively avoids the scale difference between different regions. It can be seen from the table that the implementation of ETS has significantly influenced the three effects proposed by this study: the optimization and upgrading effect of industrial structure, the synergistic aggregation effect of resources, and the support effect of ecological optimization. In addition, ETS has a positive and significant impact on the regional ecologicalization efficiency of industry. The results in Columns 3, 5, and 7 of the table show the industrial optimization and upgrading effect, resource synergistic aggregation effect, structural upgrading effect, and resource allocation effect generated in the process of low-carbon pilot policy operation can significantly promote regional ecologicalization efficiency of industry and have an obvious intermediary effect. The mediating effect produced by industrial structure optimization and upgrading is about 0.042, the mediating effect produced by resource synergy agglomeration is about 0.148, and the mediating effect produced by ecological optimization support is about 0.166. According to the Sobal test results, all of them have passed the test, indicating that the above results are reliable.Table 8 Mediating effect test results.Full size table More

Giorgi, A., Pentimalli, D., Giupponi, L. & Panseri, S. Quality traits of saffron (Crocus sativus L.) produced in the Italian Alps. Open Agric. 2(1), 52–57 (2017).Article 

Google Scholar 
Winterhalter, P. & Straubinger, M. Saffron—Renewed interest in an ancient spice. Food Rev. Intl. 16(1), 39–59 (2000).CAS 
Article 

Google Scholar 
Schmidt, M., Betti, G. & Hensel, A. Saffron in phytotherapy: Pharmacology and clinical uses. Wien Med. Wochenschr. 157, 315–319 (2007).PubMed 
Article 

Google Scholar 
Siddique, H. R., Fatma, H. & Khan, M. A. Medicinal properties of saffron with special reference to cancer—A review of preclinical studies. in Saffron: The Age-Old Panacea in a New Light (ed. Sarwat,
M. & Sumaiya, S.) 233–244 (Academic Press, 2020).Chapter 

Google Scholar 
Abdullaev, F. I. Cancer chemopreventive and tumoricidal properties of saffron (Crocus sativus L.). Exp. Biol. Med. 227(1), 20–25 (2002).CAS 
Article 

Google Scholar 
Kafi, M., Koocheki, A. & Rashed, M. H. Saffron (Crocus sativus): Production and Processing (Science Publishers, 2006).Book 

Google Scholar 
Mir, G.M. Saffron Agronomy in Kashmir (1992).Melnyk, J. P., Wang, S. & Marcone, M. F. Chemical and biological properties of the world’s most expensive spice: Saffron. Food Res. Int. 43(8), 1981–1989 (2010).CAS 
Article 

Google Scholar 
Menia, M. et al. Production technology of saffron for enhancing productivity. J. Pharmacognosy Phytochem. 7(1), 1033–1039 (2018).
Google Scholar 
Tanra, M. A., Dar, B. A. & Sing, S. Economic analysis of Production and Marketing of saffron in Jammu and Kashmir. J. Social Relevance Concern 5(10), 12–19 (2017).
Google Scholar 
Husaini, A. M., Hassan, B., Ghani, M. Y., Teixeira da Silva, J. A. & Kirmani, N. A. saffron (Crocus sativus Kashmirianus) cultivation in Kashmir: Practices and problems. Functional Plant Sci. Biotechnol. 4(2), 108–115 (2010).
Google Scholar 
Amirnia, R., Bayat, M. & Tajbakhsh, M. Effects of nano fertilizer application and maternal corm weight on flowering at some saffron (Crocus sativus L.) ecotypes. Turkish J. Field Crops. 19(2), 158–168 (2014).Article 

Google Scholar 
Kumar, R. et al. State of art of saffron (Crocus sativus L.) agronomy: A comprehensive review. Food Rev. Int. 25(1), 44–85 (2009).Article 

Google Scholar 
Dhar, A. K. Saffron breeding and agrotechnology. Status Rep. PAFAI J. 12, 18–22 (1990).
Google Scholar 
Ehsanzadeh, P., Yadollahi, A. A. & Maibodi, A. M. Productivity, growth and quality attributes of 10 Iranian saffron accessions under climatic conditions of Chahar-Mahal Bakhtiari, Central Iran. Int. Symp. Saffron Biol. Biotechnol. 650, 183–188 (2003).
Google Scholar 
Duke, J. A. Ecosystematic data on economic plants. Quart. J. Crude Drug Res. 17(3–4), 91–109 (1979).Article 

Google Scholar 
Kanth, R.H., Khanday, B.A. & Tabassum, S. Crop weather relationship for saffron production. Saffron Production in Jammu and Kashmir, Directorate of Extension Education. SKUAST-K, India 170–188 (2008).Shinde, D. A., Talib, A. R. & Gorantiwar, S. M. Composition and classification of some typical soils of saffron growing areas of Jammu and Kashmir. J. Indian Soc. Soil Sci. 32(3), 473–477 (1984).CAS 

Google Scholar 
Nazir, N. A., Khitrov, N. B. & Chizhikova, N. P. Statistical evaluation of soil properties which influence saffron growth in Kashmir. Eurasian Soil Sci. 28(4), 120–138 (1996).
Google Scholar 
Ganai, M. R., Wani, M. A. & Zargar, G. H. Characterization of saffron growing soils of Kashmir. Appl. Biol. Res. 2(1/2), 27–30 (2000).
Google Scholar 
Ganai, M.R.D. Nutrient status of saffron soils and their management. in Proceedings of Seminar-cum-Workshop on saffron (Crocus sativus) 51–54 (2001).Molina, R. V., Valero, M., Navarro, Y., Guardiola, J. L. & Garcia-Luis, A. Temperature effects on flower formation in saffron (Crocus sativus L.). ScientiaHorticulturae 103(3), 361–379 (2005).
Google Scholar 
Galavi, M., Soloki, M., Mousavi, S. R. & Ziyaie, M. Effect of planting depth and soil summer temperature control on growth and yield of saffron (Crocus sativus L.). Asian J. Plant Sci. 7(8), 747 (2008).Article 

Google Scholar 
Kamyabi, S., Habibi Nokhandan, M. & Rouhi, A. Effect of climatic factors affecting saffron using analytic hierarchy process (AHP); Case Study Roshtkhar Region, Iran. (2014).Gupta, R. K. Saffron status and cultivation in northwestern Himalayas. Vegetos 20(1), 1–7 (2007).
Google Scholar 
Qin, A. et al. Maxent modelling for predicting impacts of climate change on the potential distribution of Thuja sutchuenensis Franch., an extremely endangered conifer from southwestern China. Glob. Ecol. Conserv. 10, 139–146 (2017).Article 

Google Scholar 
Fielding, A. H. & Bell, J. F. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environ. Conserv. 24(1), 38–49 (1997).Article 

Google Scholar 
Swets, J. A. Measuring the accuracy of diagnostic systems. Science 240(4857), 1285–1293 (1988).ADS 
MathSciNet 
CAS 
PubMed 
MATH 
Article 

Google Scholar 
Muscarella, R. et al. ENMeval: An R package for conducting spatially independent evaluations and estimating optimal model complexity for Maxent ecological niche models. Methods Ecol. Evaluat. 5(11), 1198–1205 (2014).Article 

Google Scholar 
Hao, T., Elith, J., Arroita, G. G. & Monfort, J. J. L. A review of evidence about use and performance of species distribution modelling ensembles like BIOMOD. Divers. Distrib. 25(5), 839–852 (2019).Article 

Google Scholar 
Thuiller, W. BIOMOD—Optimizing predictions of species distributions and projecting potential future shifts under global change. Glob. Change Biol. 9, 1353–1362 (2003).ADS 
Article 

Google Scholar 
Mykhailenko, O., Desenko, V., Ivanauskas, L. & Georgiyants, V. Standard operating procedure of Ukrainian saffron cultivation according to with good agricultural and collection practices to assure quality and traceability. Ind. Crops Prod. 151, 112376. https://doi.org/10.1016/j.indcrop.2020.112376 (2020).CAS 
Article 

Google Scholar 
Kothari, D., Thakur, M., Joshi, R., Kumar, A. & Kumar, R. Agro-climatic suitability evaluation for saffron production in areas of western Himalaya. Front. Plant Sci. 12, 657819. https://doi.org/10.3389/fpls.2021.657819 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
Mir, J. I., Ahmed, N., Wafai, A. H. & Qadri, R. A. Variability in stigma length and apocarotenoid content in Crocus sativus L. selections of Kashmir. J. Spices Aromatic Crops. 21(2), 169–171 (2012).
Google Scholar 
Nehvi, F. A. et al. New emerging trends on production technology of saffron. II Int. Symp. Saffron Biol. Technol. 739, 375–381 (2006).
Google Scholar 
Golmohammadi, F. Sustainable agriculture and rural development in Iran, Some modern issues in sustainable agriculture and rural development in Iran Germany, LAP LAMBERT Academic Publishing GmbH & Co. LAP Lambert Academic Publishing. Germany. ISBN-13, 978-3 (2012).Golmohammadi, F. Saffron and its importance, medical uses and economical export situation in Iran. in Oral Article Presented in: International Conference on Advances in Plant Sciences 14–18 (2012).Phillips, S. J., Anderson, R. P. & Schapire, R. E. Maximum entropy modelling of species geographic distributions. Ecol. Model. 190(3–4), 231–259 (2006).Article 

Google Scholar 
Elith, J. et al. Novel methods improve prediction of species distributions from occurrence data. Ecography 29(2), 129–151 (2006).Article 

Google Scholar 
Pearson, R. G., Raxworthy, C. J., Nakamura, M. & Townsend Peterson, A. Predicting species distributions from small numbers of occurrence records: A test case using cryptic geckos in Madagascar. J. Biogeogr. 34(1), 102–117 (2007).Article 

Google Scholar 
Wisz, M. S. et al. NCEAS Predicting species distributions working group. Effects of sample size on the performance of species distribution models. Diversity Distributions. 14(5), 763–773 (2008).Article 

Google Scholar 
Rebelo, H. & Jones, G. Ground validation of presence only modelling with rare species: A case study on Barbastella barbastellus (Chiroptera: Vespertilionidae). J. Appl. Ecol. 47(2), 410–420 (2010).Article 

Google Scholar 
Elith, J. & Leathwick, J. R. Species distribution models: Ecological explanation and prediction across space and time. Annu. Rev. Ecol. Evol. Syst. 40, 677–697 (2009).Article 

Google Scholar 
Palomera, S. et al. Mapping from heterogeneous biodiversity monitoring data sources. Biodiversity Conservation 21(11), 2927–2948 (2012).Article 

Google Scholar 
Garcia, K., Lasco, R., Ines, A., Lyon, B. & Pulhin, F. Predicting geographic distribution and habitat suitability due to climate change of selected threatened forest tree species in the Philippines. Appl. Geogr. 44, 12–22 (2013).Article 

Google Scholar 
Marcer, A., Sáez, L., Molowny-Horas, R., Pons, X. & Pino, J. Using species distribution modelling to disentangle realised versus potential distributions for rare species conservation. Biol. Cons. 166, 221–230 (2013).Article 

Google Scholar 
Phillips, S.J., Dudík, M. & Schapire, R.E. A maximum entropy approach to species distribution modelling. in Proceedings of the Twenty-First International Conference on Machine Learning 83 (2004).Baldwin, R. A. Use of maximum entropy modelling in wildlife research. Entropy 11(4), 854–866 (2009).ADS 
Article 

Google Scholar 
Izadpanah, F., Kalantari, S., Hassani, M. E., Naghavi, M. R. & Shokrpour, M. Variation in Saffron (Crocus sativus L.) accessions and Crocus wild species by RAPD analysis. Plant Syst. Evolut. 300, 1941–1944 (2014).Article 

Google Scholar 
Nemati, Z., Harpke, D., Gemicioglu, A., Kerndorff, H. & Blattner, F. R. Saffron (Crocus sativus) is an autotriploid that evolved in Attica (Greece) from wild Crocus cartwrightianus. Mol. Phylogenet. Evol. 136, 14–20 (2019).PubMed 
Article 

Google Scholar 
Proosdij, A. S. J. V., Sosef, M. S. M., Wieringa, J. J. & Raes, N. Minimum required number of specimen records to develop accurate species distribution models. Ecography 39, 542–552 (2016).Article 

Google Scholar  More

Brandes, N. & Linial, M. Giant viruses—big surprises. Viruses 11, 404 (2019).CAS 
Article 

Google Scholar 
Jover, L. F., Effler, T. C., Buchan, A., Wilhelm, S. W. & Weitz, J. S. The elemental composition of virus particles: implications for marine biogeochemical cycles. Nat. Rev. Microbiol. 12, 519–528 (2014).CAS 
Article 

Google Scholar 
Madsen, E. L. Microorganisms and their roles in fundamental biogeochemical cycles. Curr. opinion biotechnology 22, 456–464 (2011).CAS 
Article 

Google Scholar 
Gummow, B. Challenges posed by new and re-emerging infectious diseases in livestock production, wildlife and humans. Livest. Sci. 130, 41–46 (2010).CAS 
Article 

Google Scholar 
Becker, D. J., Streicker, D. G. & Altizer, S. Linking anthropogenic resources to wildlife–pathogen dynamics: a review and meta-analysis. Ecol. letters 18, 483–495 (2015).Article 

Google Scholar 
Woolhouse, M. E. & Gowtage-Sequeria, S. Host range and emerging and reemerging pathogens. Emerg. infectious diseases 11, 1842 (2005).Article 

Google Scholar 
Foster, R. et al. Pathogens co-transported with invasive non-native aquatic species: implications for risk analysis and legislation. NeoBiota 69, 79–102 (2021).Article 

Google Scholar 
Brasier, C. The biosecurity threat to the uk and global environment from international trade in plants. Plant Pathol. 57, 792–808 (2008).Article 

Google Scholar 
Ruiz, G. M. et al. Global spread of microorganisms by ships. Nature 408, 49–50 (2000).ADS 
CAS 
Article 

Google Scholar 
Essl, F. et al. Which taxa are alien? criteria, applications, and uncertainties. BioScience 68, 496–509 (2018).Article 

Google Scholar 
Blackburn, T. M., Bellard, C. & Ricciardi, A. Alien versus native species as drivers of recent extinctions. Front. Ecol. Environ. 17, 203–207 (2019).Article 

Google Scholar 
Hawkins, C. L. et al. Framework and guidelines for implementing the proposed iucn environmental impact classification for alien taxa (eicat). Divers. Distributions 21, 1360–1363 (2015).Article 

Google Scholar 
Corrales, X. et al. Advances and challenges in modelling the impacts of invasive alien species on aquatic ecosystems. Biol. Invasions 22, 907–934 (2020).Article 

Google Scholar 
Regulation, E. Regulation (eu) no 1143/2014 of the European Parliament and of the Council of 22 October 2014 on the prevention and management of the introduction and spread of invasive alien species. Off. J. Eur. Union 57, 35–55 (2014).
Google Scholar 
EU. Regulation (eu) 2016/2031 of the European Parliament of the Council of 26 October 2016 on protective measures against pests of plants, amending regulations (eu) 228/2013,(eu) 652/2014 and (eu) 1143/2014 and repealing council directives 69/464/eec, 74/647/eec, 93/85/eec, 98/57/ec, 2000/29/ec, 2006/91/ec and 2007/33/ec. Off. J. 317, 4–104 (2016).
Google Scholar 
Murtaugh, M. P. et al. The science behind one health: at the interface of humans, animals, and the environment. Tech. Rep. (2017).Ogden, N. H. et al. Emerging infectious diseases and biological invasions: a call for a one health collaboration in science and management. Royal Soc. open science 6, 181577 (2019).ADS 
Article 

Google Scholar 
Roy, H. E. et al. Alien pathogens on the horizon: Opportunities for predicting their threat to wildlife. Conserv. Lett. 10, 477–484 (2017).Article 

Google Scholar 
Ikner, L. A., Gerba, C. P. & Bright, K. R. Concentration and recovery of viruses from water: a comprehensive review. Food Environ. Virol. 4, 41–67 (2012).
Google Scholar 
Taylor, M. W. Introduction: A short history of virology. In Viruses and Man: A History of Interactions, 1–22 (Springer, 2014).Thakur, M. P., Van der Putten, W. H., Cobben, M. M., van Kleunen, M. & Geisen, S. Microbial invasions in terrestrial ecosystems. Nat. Rev. Microbiol. 17, 621–631 (2019).CAS 
Article 

Google Scholar 
Desprez-Loustau, M.-L. et al. The fungal dimension of biological invasions. Trends ecology & evolution 22, 472–480 (2007).Article 

Google Scholar 
Rivett, D. W. et al. Elevated success of multispecies bacterial invasions impacts community composition during ecological succession. Ecol. Lett. 21, 516–524 (2018).Article 

Google Scholar 
Dunn, A. M. & Hatcher, M. J. Parasites and biological invasions: parallels, interactions, and control. TRENDS Parasitol. 31, 189–199 (2015).Article 

Google Scholar 
Pyšek, P. et al. Macroecological framework for invasive aliens (mafia): disentangling large-scale context dependence in biological invasions. (2020).Hulme, P. E. et al. Blurring alien introduction pathways risks losing the focus on invasive species policy. Conserv. Lett. 10, 265–266 (2017).Article 

Google Scholar 
Gilroy, J. J., Avery, J. D. & Lockwood, J. L. Seeking international agreement on what it means to be “native”. Conserv. Lett. 10, 238–247 (2017).Article 

Google Scholar 
Webber, B. L. & Scott, J. K. Rapid global change: implications for defining natives and aliens. Glob. Ecol. Biogeogr. 21, 305–311 (2012).Article 

Google Scholar 
CBD Secretariat. Decision VI/23: Alien species that threaten ecosystems, habitats and species. Document UNEP/CBD/COP/6/23 (2002).World Health Organization. A brief guide to emerging infectious diseases and zoonoses. Tech. Rep. https://apps.who.int/iris/handle/10665/204722 (2014).Firrao, G. et al. Candidatus phytoplasma’, a taxon for the wall-less, non-helical prokaryotes that colonize plant phloem and insects. Int. J. Syst. Evol. Microbiol. 54, 1243–1255 (2004).CAS 
Article 

Google Scholar 
CBD. Pathways of introduction of invasive species, their prioritization and management (Secretariat of the Convention on Biological Diversity Montreal, 2014).OIE. Terrestrial Animal Health Code 2021 (OIE, 2021).Magliozzi, C. et al. bacteria and viruses traits and species-related factors. figshare https://doi.org/10.6084/m9.figshare.18550907.v2 (2022).Katsanevakis, S. et al. Implementing the European policies for alien species: networking, science, and partnership in a complex environment. Manag. Biol. Invasions 4, 3–6 (2013).Article 

Google Scholar 
Tsiamis, K. et al. The EASIN Editorial Board: quality assurance, exchange and sharing of alien species information in europe. Manag. Biol. invasions 7, 321–328 (2016).Article 

Google Scholar 
Wieczorek, J. et al. Darwin core: an evolving community-developed biodiversity data standard. PloS one 7, e29715 (2012).ADS 
CAS 
Article 

Google Scholar 
Darwin Core. Darwin Core quick reference guide. https://dwc.tdwg.org/terms/ (2018).R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org/ (2020).Wickham, H. ggplot2: Elegant Graphics for Data Analysis, https://ggplot2.tidyverse.org (Springer-Verlag New York, 2016).Schwarzl, T. ggBubbles: Mini Bubble Plots for Comparison of Discrete Data with ‘ggplot2’ R package version 0.1.4 (2019).Moon, K. R statistics and graphs for medical papers (Hannarae Seoul, 2015).Current, C. Invasive species compendium. Wallingford, UK: CAB Int. Available online: www.cabi.org/isc (accessed on 19 August 2020) (2011).Adams, M. J. & Antoniw, J. F. Dpvweb: An open access internet resource on plant viruses and virus diseases. Outlooks on Pest Manag. 16, 268 (2005).Article 

Google Scholar 
Adams, M. J. & Antoniw, J. F. Dpvweb: a comprehensive database of plant and fungal virus genes and genomes. Nucleic acids research 34, D382–D385 (2006).CAS 
Article 

Google Scholar 
Benson, D. A. et al. Genbank. Nucleic acids research 41, D36–D42 (2012).Article 

Google Scholar  More

Sakamoto, M. Primary production by phytoplankton community in some Japanese lakes and its dependence on lake depth. Archiv für Hydrobilogie. 62, 1–28 (1966).
Google Scholar 
Vollenweider, R. A. Scientific fundamentals of the eutrophication of lakes and flowing waters, with particular reference to nitrogen and phosphorus as factors in eutrophication (Organisation for Economic Co-operation and Development, 1968).
Google Scholar 
Edmondson, W. T. Phosphorus, nitrogen, and algae in Lake Washington after diversion of sewage. Science 169, 690–691 (1970).ADS 
CAS 
Article 

Google Scholar 
Dillon, P. J. & Rigler, F. H. The phosphorus-chlorophyll relationship in lakes. Limnol. Oceanogr. 19, 767–773 (1974).ADS 
CAS 
Article 

Google Scholar 
Jones, J. R. & Bachmann, R. W. Prediction of phosphorus and chlorophyll levels in lakes. J. Water Pollut. Control Feder. 48, 2176–2182 (1976).CAS 

Google Scholar 
Schindler, D. W. Evolution of phosphorus limitation in lakes. Science 195, 260–262 (1977).ADS 
CAS 
Article 

Google Scholar 
Filstrup, C. T. & Downing, J. A. Relationship of chlorophyll to phosphorus and nitrogen in nutrient-rich lakes. Inland Waters. 7, 385–400 (2017).CAS 
Article 

Google Scholar 
Schindler, D. W. Recent advances in the understanding and management of eutrophication. Limnol. Oceanogr. 51, 356–363 (2006).ADS 
Article 

Google Scholar 
Quinlan, R. et al. Relationships of total phosphorus and chlorophyll in lakes worldwide. Limnol. Oceanogr. 66, 392–404 (2020).ADS 
Article 

Google Scholar 
Yuan, L. L. & Jones, J. R. Rethinking phosphorus–chlorophyll relationships in lakes. Limnol. Oceanogr. 65, 1847–1857 (2020).ADS 
CAS 
Article 

Google Scholar 
Carlson, R. E. A trophic state index for lakes. Limnol. Oceanogr. 11, 361–369 (1977).ADS 
Article 

Google Scholar 
Neveux, J. et al. Comparison of chlorophyll and phaeopigment determinations by spectrophotometric, fluorometric, spectrofluorometric and HPLC methods. Mar. Microb. Food Webs 4, 217–238 (1990).
Google Scholar 
Lampert, W. & Sommer, U. Limnoecology (Oxford University, 2007).
Google Scholar 
Kovalevskaya, R. Z., Zhukava, H. A. & Adamovich, B. V. Modification of the method of spectrophotometric determination of chlorophyll a in the suspended matter of water bodies. J. Appl. Spectrosc. 87, 72–78 (2020).ADS 
CAS 
Article 

Google Scholar 
Søndergaard, M., Lauridsen, T. L., Johansson, L. S. & Jeppesen, E. Nitrogen or phosphorus limitation in lakes and its impact on phytoplankton biomass and submerged macrophyte cover. Hydrobiologia 795, 35–48 (2017).Article 

Google Scholar 
Søndergaard, M., Jensen, J. P., Jeppesen, E. & Møller. P. H. Seasonal dynamics in the concentrations and retention of phosphorus in shallow Danish lakes after reduced loading. Aquat. Ecosyst. Health Manag. 5(1), 19–29 (2002).Magumba, D., Atsushi, M., Michiko, T., Akira, K. & Masao, K. Relationships between Chlorophyll-a, phosphorus and nitrogen as fundamentals for controlling phytoplankton biomass in lakes. Environ. Control. Biol. 51(4), 179–185 (2013).CAS 
Article 

Google Scholar 
Smith, V. H. & Shapiro, J. Chlorophyll-phosphorus relations in individual lakes. Their importance to lake restoration strategies. Environ. Sci. Technol. 15(4), 444–451 (1981).Pothoven, S. A. & Vanderploeg, H. A. Seasonal patterns for Secchi depth, chlorophyll a, total phosphorus, and nutrient limitation differ between nearshore and offshore in Lake Michigan. J. Great Lakes Res. 46, 519–527 (2020).CAS 
Article 

Google Scholar 
Søndergaard, M. & Jeppesen, E. Lake Søbygaard, Denmark: phosphorus dynamics during the first 35 years after an external loading reduction. In: Internal Phosphorus Loading: Causes, Case Studies, and Management (ed. Steinman, A.D. & Spears, B. M.) 285–299 (J. Ross, Plantation, 2020).Guildford, S. J. & Hecky, R. E. Total nitrogen, total phosphorus, and nutrient limitation in lakes and oceans: Is there a common relationship?. Limnol. Oceanogr. 45, 1213–1223 (2000).ADS 
CAS 
Article 

Google Scholar 
Jones, J.R. et al. Nutrients, seston, and transparency of Missouri reservoirs and oxbow lakes. An analysis of regional limnology. Lake Reser. Manag. 24, 155–180 (2008).Pikovsky, A., Rosenblum, M. & Kurths, J. Synchronization. A universal concept in nonlinear sciences (Cambridge University, 2001).Book 

Google Scholar 
Kuramoto, Y. Chemical Oscillations, Waves, and Turbulence (Springer, 1984).Book 

Google Scholar 
Sazonov, A. V. et al. An investigation of the phase locking index for measuring of interdependency of cortical source signals recorded in the EEG. Biol. Cybern. 100, 129–146 (2009).Article 

Google Scholar 
Medvinsky, A. B. et al. Temperature as a factor affecting fluctuations and predictability of the abundance of lake bacterioplankton. Ecol. Complex. 32, 90–98 (2017).Article 

Google Scholar 
Zhukova, T. V. & Ostapenya, A. P. Estimation of efficiency of nature protection measures in water catchment area of the Naroch lakes. Natural Resources. 3, 68–73 (2000) ((in Russian)).
Google Scholar 
Burlakova, L. E., Karatayev, A. Y. & Padilla, D. K. Changes in the distribution and abundance of Dreissena polymorpha within lakes through time. Hydrobiologia 571, 133–146 (2006).Article 

Google Scholar 
Ostapenia, A. P. et al. Bentification of lake ecosystem: causes, mechanisms, possible consequences, prospect for future research. Trudy BGU. 7, 135–148 (2012) ((in Russian)).
Google Scholar 
Karatayev, A.Y., Burlakova, L.E. & Padilla, D.K. Impacts of Zebra Mussels on aquatic communities and their role as ecosystem engineers. In: Leppäkoski, E., Gollasch, S., Olenin, S. (eds) Invasive Aquatic Species of Europe. Distribution, Impacts and Management (Springer, Dordrecht, 2002).Adamovich, B. V. et al. The divergence of chlorophyll dynamics in the Naroch Lakes. Biophysics 60, 632–638 (2015).CAS 
Article 

Google Scholar 
Zhukova, T. V. et al. Long-term dynamics of suspended matter in Naroch Lakes: Trend or intervation. Inland Water Biol. 10, 250–257 (2017).Article 

Google Scholar 
Adamovich, B. V. et al. Eutrophication, oligotrophication, and benthiphication in Naroch Lakes: 40 years of monitoring. J. Siber. Federal Univ. Biol. 10, 379–394 (2017).Article 

Google Scholar 
Ostapenya A.P. et al. Ecological passport of Lake Myastro (EcoMir, Minsk, 1994) (in Russian).Kantz, H. & Schreiber, T. Nonlinear time series analysis (Cambridge University, 1997).MATH 

Google Scholar 
Kot, M. Elements of mathematical ecology (Cambridge University, 2001).Book 

Google Scholar 
Turchin, P. Complex population dynamics. A Theoretical/Empirical Synthesis (Princeton University, Princeton, 2003).MATH 

Google Scholar 
Cazelles, B. & Stone, L. Detection of imperfect population synchrony in an uncertain world. J. Anim. Ecol. 72, 953–968 (2003).Article 

Google Scholar 
Karatayev, A. Y., Burlakova, L. & Padilla, D. K. The effects of Dreissena polymorpha (Pallas) invasion on aquatic communities in Eastern Europe. J. Shellfish Res. 16, 187–203 (1997).
Google Scholar 
Lia, J. et al. Benthic invaders control the phosphorus cycle in the world’s largest freshwater ecosystem. PNAS 118(6), e2008223118. https://doi.org/10.1073/pnas.2008223118 (2021).CAS 
Article 

Google Scholar 
Mikheyeva, T. M. et al. The dynamics of freshwater phytoplankton stability in the Naroch Lakes (Belarus). Ecol. Ind. 81, 481–490 (2017).Article 

Google Scholar 
Harris, P. H. Phytoplankton ecology. Structure, functioning and flucttuation (Chapman & Hall, London, New York, 1986).Jeppesen, E., Jensen, J. P., Søndergaard, M. & Lauridsen, T. L. Response of fish and plankton to nutrient loading reduction in eight shallow Danish lakes with special emphasis on seasonal dynamics. Freshw. Biol. 50, 1616–1627 (2005).CAS 
Article 

Google Scholar 
Nezlin, N.P. & Li, B-L. Time-series analysis of remote-sensed chlorophyll and environmental factors in the Santa Monica–San Pedro Basin off Southern California. J. Mar. Syst. 39, 185–202 (2003).French, T. D. & Petticrew, E. L. Chlorophyll a seasonality in four shallow eutrophic lakes (northern British Columbia, Canada) and the critical roles of internal phosphorus loading and temperature. Hydrobiologia 575, 285–299 (2007).CAS 
Article 

Google Scholar 
SCOR-UNESCO Working Group no. 17. Determination of photosynthetic pigments in sea-water. Monographs on Oceanologic Methodology 9–18 (UNESSCO, Paris, 1966).Semenov, A. D. Guide on the chemical analysis of continental surface waters (Gidrometeoizdat, 1977) ((in Russian)).
Google Scholar 
Wetzel, R. G. & Likens, G. E. Limnological analysis (Springer, 2000).Book 

Google Scholar 
Steffen, M. & Bartz-Beielstein, T. imputeTS: time series missing value imputation in R. R J. 9(1), 207–218 (2017).Article 

Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, Vienna, 2020). More