Philippa Kaur

Siński, E. & Welc-Falęciak, R. Risk of Infections Transmitted by Ticks in Forest Ecosystems of Poland. (Zarządzanie Ochroną Przyrody w Lasach 6, 2012).Kantar Public. Zwierzęta w polskich domach, 2017. http://www.tnsglobal.pl/archiwumraportow/files/2017/05/K.021_Zwierzeta_domowe_O04a-17.pdf.Maia, C. et al. Bacterial and protozoal agents of feline vector-borne diseases in domestic and stray cats from southern Portugal. Parasit. Vectors. 7, 115. https://doi.org/10.1186/1756-3305-7-115 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
Maia, C. et al. Bacterial and protozoal agents of canine vector-borne diseases in the blood of domestic and stray dogs from southern Portugal. Parasit. Vectors. 23(8), 138. https://doi.org/10.1186/s13071-015-0759-8 (2015).Article 

Google Scholar 
Baturo, I.M. Parki narodowe i krajobrazowe, rezerwaty przyrody. (Departament Turystyki, Sportu, Promocji Urzędu Marszłkowskiego Województwa Małopolskiego, 2010).Dulias, R. & Hibszer, A. Województwo śląskie—przyroda, gospodarka, dziedzictwo kulturowe (Wydawnictwo Kubajak, 2004).
Google Scholar 
Siuda, K. Kleszcze Polski Acari Ixodida). Część II. Systematyka i Rozmieszczenie (Polskie Towarzystwo Parazytologiczne, 1993).
Google Scholar 
Nowak-Chmura, M. Fauna kleszczy (Ixodida) Europy Środkowej (Wydawnictwo Naukowe Uniwersytetu Pedagogicznego, 2013).
Google Scholar 
Rijpkema, S., Golubić, D., Molkenboer, M., Verbeek-De Kruif, N. & Schellekens, J. Identification of four genomic groups of Borrelia burgdorferi sensu lato in Ixodes ricinus ticks collected in a Lyme borreliosis endemic region of northern Croatia. Exp. Appl. Acarol. 20, 23–30 (1996).CAS 
Article 

Google Scholar 
Wójcik-Fatla, A., Szymańska, J., Wdowiak, L., Buczek, A. & Dutkiewicz, J. Coincidence of three pathogens (Borrelia burgdorferi sensu lato, Anaplasma phagocytophilum and Babesia microti) in Ixodes ricinus ticks in the Lublin makroregion. Ann. Agric. Environ. Med. 16(1), 151–158 (2009).PubMed 

Google Scholar 
Massung, R. F. et al. Nested PCR assay for detection of granulocytic ehrlichiae. J. Clin. Microbiol. 36(4), 1090–1095 (1998).CAS 
Article 

Google Scholar 
Persing, D. H. et al. Detection of Babesia microti by polymerase chain reaction. J. Clin. Microbiol. 30, 2097–2103 (1992).CAS 
Article 

Google Scholar 
Sroka, J., Szymańska, J. & Wójcik-Fatla, A. The occurence of Toxoplasma gondii and Borrelia burgdorferi sensu lato in Ixodes ricinus ticks from east Poland with the use of PCR. Ann. Agric. Environ. Med. 16(2), 313–319 (2009).PubMed 

Google Scholar 
Siuda, K., Nowak, M., Gierczak, M., Wierzbowska, I. & Faber, M. Kleszcze (Acari: Ixodida) pasożytujące na psach i kotach domowych w Polsce. Wiad. Parazytol. 53, 155 (2007).
Google Scholar 
Zajkowska, P. Ticks (Acari:Ixodida) attacking domestic dogs in the Malopolska voivodeship, Poland. In Arthropods: In the contemporary world (eds Buczek, A. & Błaszak, C. Z.) 87–99 (Koliber, 2015).Chapter 

Google Scholar 
Szymański, S. Przypadek masowego rozwoju kleszcza Rhipicephalus sanguineus (Latreile, 1806) w warszawskim mieszkaniu. Wiad. Parazytol. 25, 453–458 (1979).PubMed 

Google Scholar 
Król, N., Obiegala, A., Pfeffer, M., Lonc, E. & Kiewra, D. Detection of selected pathogens in ticks collected from cats and dogs in the Wrocław Agglomeration South-West Poland. Parasit. Vectors. 9(1), 351. https://doi.org/10.1186/s13071-016-1632-0 (2016).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Kocoń, A., Nowak-Chmura, M., Kłyś, M. & Siuda, K. Ticks (Acari: Ixodida) attacking domestic cats (Felis catus L.) in southern Poland. In Arthropods in Urban and Suburban Environments (eds Buczek, A. & Błaszak, C.) 51–61 (Koliber, 2017).
Google Scholar 
Roczeń-Karczmarz, M. et al. Comparison of the occurrence of tick-borne diseases in ticks collected from vegetation and animals in the same area. Med. Weter. 74(8), 484–488. https://doi.org/10.21521/mw.6107 (2018).Article 

Google Scholar 
Cuber, P., Asman, M., Solarz, K., Szilman, E. & Szilman, P. Pierwsze stwierdzenia obecności wybranych patogenów chorób transmisyjnych w kleszczach Ixodes ricinus (Acari: Ixididae) zebranych w okolicach zbiorników wodnych w Rogoźniku (województwo śląskie) in Stawonogi. Ekologiczne i patologiczne aspekty układu pasożyt – żywiciel (eds. Buczek, C. & Błaszak, Cz.). 155-164 (Akapit, Lublin, 2010).Asman, M. et al. The risk of exposure to Anaplasma phagocytophilum, Borrelia burgdorferi sensu lato, Babesia sp. and co-infections in Ixodes ricinus ticks on the territory of Niepołomice Forest (southern Poland). Ann. Parasitol. 59(1), 13–19 (2013).PubMed 

Google Scholar 
Pawełczyk, O. et al. The PCR detection of Anaplasma phagocytophilum, Babesia microti and Borrelia burgdorferi sensu lato in ticks and fleas collected from pets in the Będzin district area (Upper Silesia, Poland) – the preliminary studies in Stawonogi: zagrożenie zdrowia człowieka i zwierząt (eds. Buczek, C. & Błaszak, Cz.). 111–119 (Koliber, Lublin, 2014).Strzelczyk, J. K. et al. Prevalence of Borrelia burgdorferi sensu lato in Ixodes ricinus ticks collected from southern Poland. Acta Parasitol. 60(4), 666–674. https://doi.org/10.1515/ap-2015-0095 (2015).CAS 
Article 
PubMed 

Google Scholar 
Zygner, W. & Wędrychowicz, H. Occurrence of hard ticks in dogs from Warsaw area. Ann. Agric. Environ. Med. 13(2), 355–359 (2006).PubMed 

Google Scholar 
Kilar, P. Ticks attacking domestic dogs in the area of the Rymanów district, Subcarpathian province Poland. Wiad. Parazytol. 57(3), 189–1991 (2011).PubMed 

Google Scholar 
Claerebout, E. et al. Ticks and associated pathogens collected from dogs and cats in Belgium. Parasit. Vectors. 6, 183. https://doi.org/10.1186/1756-3305-6-183 (2013).Article 
PubMed 
PubMed Central 

Google Scholar 
Schreiber, C. et al. Pathogens in ticks collected from dogs in Berlin/Brandenburg, Germany. Parasit. Vectors. 7, 535. https://doi.org/10.1186/s13071-014-0535-1 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
Eichenberger, R. M., Deplazes, P. & Mathis, A. Ticks on dogs and cats: A pet owner-based survey in a rural town in northeastern Switzerland. Ticks Tick-borne Dis. 6, 267–271. https://doi.org/10.1016/j.ttbdis.2015.01.007 (2015).Article 
PubMed 

Google Scholar 
Michalski, M. M. Skład gatunkowy kleszczy psów (Acari: Ixodida) z terenu aglomeracji miejskiej w cyklu wieloletnim. Med. Weter. 73(11), 698–701 (2017).
Google Scholar 
Geurden, T. et al. Detection of tick-borne pathogens in ticks from dogs and cats in different European countries. Ticks. Tick. Borne. Dis. 9(6), 1431–1436. https://doi.org/10.1016/j.ttbdis.2018.06.013 (2018).Article 
PubMed 

Google Scholar 
Namina, A. et al. Tick-borne pathogens in ticks collected from dogs, Latvia, 2011–2016. BMC Vet. Res. 15, 398. https://doi.org/10.1186/s12917-019-2149-5 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Bhide, M., Travnicek, M., Curlik, J. & Stefancikova, A. The importance of dogs in eco-epidemiology of Lyme borreliosis: a review. Vet. Med. Czech 49(4), 135–142 (2004).Article 

Google Scholar 
Burgess, E. C. Experimentally induced infection of cats with Borrelia burgdorferi. Am. J. Vet. Res. 53, 1507–1511 (1992).CAS 
PubMed 

Google Scholar 
Skotarczak, B. & Wodecka, B. Identification of Borrelia burgdorferi genospecies inducing Lyme disease in dogs from western Poland. Acta Vet. Hung. 53(1), 13–21 (2005).Article 

Google Scholar 
Skotarczak, B. et al. Prevalence of DNA and antibodies to Borrelia burgdorferi sensu lato in dogs suspected of borreliosis. Ann. Agric. Environm. Med. 12(2), 199–205 (2005).CAS 

Google Scholar 
Adaszek, Ł, Winiarczyk, S., Kutrzeba, J., Puchalski, A. & Dębiak, P. Przypadki boreliozy u psow na Lubelszczyźnie. Życie Wet. 83, 311–313 (2008).
Google Scholar 
Hovius, K. E. Borreliosis. In Arthropod-borne Infectious Diseases of the Dog and Cat (eds Shaw, S. E. & Day, M. J.) 100–109 (Manson Publishing, 2005).Chapter 

Google Scholar 
Zygner, W., Jaros, S. & Wędrychowicz, H. Prevalence of Babesia canis, Borrelia afzelii, and Anaplasma phagocytophilum infection in hard ticks removed from dogs in Warsaw (central Poland). Vet. Parasitol. 153, 139–142. https://doi.org/10.1016/j.vetpar.2008.01.036 (2008).Article 
PubMed 

Google Scholar 
Welc-Falęciak, R., Rodo, A., Siński, E. & Bajer, A. Babesia canis and other tick-borne infections in dogs in Central Poland. Vet. Parasitol. 166(3–4), 191–198. https://doi.org/10.1016/j.vetpar.2009.09.038 (2009).Article 
PubMed 

Google Scholar 
Michalski, M. M., Kubiak, K., Szczotko, M., Chajęcka, M. & Dmitryjuk, M. Molecular Detection of Borrelia burgdorferi Sensu Lato and Anaplasma phagocytophilum in Ticks Collected from Dogs in Urban Areas of North-Eastern Poland. Pathogens. 9(6), 455. https://doi.org/10.3390/pathogens9060455 (2020).CAS 
Article 
PubMed Central 

Google Scholar 
Nijhof, A. M. et al. Ticks and associated pathogens collected from domestic animals in the Netherlands. Vector. Borne. Zoonot. Dis. 7, 585–595. https://doi.org/10.1089/vbz.2007.0130 (2007).Article 

Google Scholar 
Adaszek, Ł, Martinez, A. C. & Winiarczyk, S. The factors affecting the distribution of babesiosis in dogs in Poland. Vet. Parasitol. 181, 160–165. https://doi.org/10.1016/j.vetpar.2011.03.059 (2011).Article 
PubMed 

Google Scholar 
Adaszek, Ł, Łukaszewska, J., Winiarczyk, S. & Kunkel, M. Pierwszy przypadek babeszjozy u kota w Polsce. Życie Wet. 83(8), 668–670 (2008).
Google Scholar 
Kocoń, A. et al. Molecular detection of tick-borne pathogens in ticks collected from pets in selected mountainous areas of Tatra County (Tatra Mountains, Poland). Sci. Rep. 10, 15865. https://doi.org/10.1038/s41598-020-72981-w (2020).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Asman, M. et al. Detection of protozoans Babesia microti and Toxoplasma gondii and their co-existence in ticks (Acari: Ixodida) collected in Tarnogórski district (Upper Silesia, Poland). Ann. Agric. Environ. Med. 22(1), 80–83. https://doi.org/10.5604/12321966.1141373 (2015).Article 
PubMed 

Google Scholar 
Stensvold, C. R. et al. Babesia spp. and other pathogens in ticks recovered from domestic dogs in Denmark. Parasit. Vectors. 8(8), 262. https://doi.org/10.1186/s13071-015-0843-0 (2015).Article 
PubMed 
PubMed Central 

Google Scholar 
Abdullah, S., Helps, C., Tasker, S., Newbury, H. & Wall, R. Prevalence and distribution of Borrelia and Babesia species in ticks feeding on dogs in the UK. Med. Vet. Entomol. 32(1), 14–22. https://doi.org/10.1111/mve.12257 (2018).CAS 
Article 
PubMed 

Google Scholar 
Bjoersdorff, A., Svendenius, L., Owens, J. H. & Massung, R. F. Feline granulocytic ehrlichiosis– a report of a new clinical entity and characterisation of the infectious agent. J. Small. Anim. Pract. 40(1), 20–24. https://doi.org/10.1111/j.1748-5827.1999.tb03249 (1999).CAS 
Article 
PubMed 

Google Scholar 
Lappin, M. R. et al. Molecular and serologic evidence of Anaplasma phagocytophilum infection in cats in North America. J. Am. Vet. Med. Assoc. 225(6), 893–896. https://doi.org/10.2460/javma.2004.225.893 (2004).Article 
PubMed 

Google Scholar 
Shaw, S. E. et al. Molecular evidence of tick-transmitted infections in dogs and cats in the United Kingdom. Vet. Rec. 157(21), 645–648. https://doi.org/10.1136/vr.157.21.645 (2005).CAS 
Article 
PubMed 

Google Scholar 
Tarello, W. Microscopic and clinical evidence for Anaplasma (Ehrlichia) phagocytophilum infection in Italian cats. Vet. Rec. 156(24), 772–774. https://doi.org/10.1136/vr.156.24.772 (2005).CAS 
Article 
PubMed 

Google Scholar 
Schaarschmidt-Kiener, D., Graf, F., von Loewenich, F. D. & Muller, W. Anaplasma phagocytophilum infection in a cat in Switzerland. Schweiz. Arch. Tierheilkd. 151(7), 336–341. https://doi.org/10.1024/0036-7281.151.7.336 (2009).CAS 
Article 
PubMed 

Google Scholar 
Heikkila, H. M., Bondarenko, A., Mihalkov, A., Pfister, K. & Spillmann, T. Anaplasma phagocytophilum infection in a domestic cat in Finland. Acta. Vet. Scand. 52(1), 62. https://doi.org/10.1186/1751-0147-52-62 (2010).Article 
PubMed 
PubMed Central 

Google Scholar 
Hamel, D., Bondarenko, A., Silaghi, C., Nolte, I. & Pfister, K. Seroprevalence and bacteremia of Anaplasma phagocytophilum in cats from Bavaria and Lower Saxony (Germany). Berl. Munch. Tierarztl. Wochenschr. 125(3–4), 163–167 (2012).PubMed 

Google Scholar 
Morgenthal, D. et al. Prevalence of haemotropic Mycoplasma spp., Bartonella spp. and Anaplasma phagocytophilum in cats in Berlin/Brandenburg (Northeast Germany). Berl. Munch Tierarztl. Wochenschr. 125(9–10), 418–427 (2012).PubMed 

Google Scholar 
Adaszek, Ł, Winiarczyk, S. & Łukaszewska, J. A first case of ehrlichiosis in a horse in Poland. Dtsch. Tierarztl. Wchschr. 116(9), 330–334 (2009).
Google Scholar 
Adaszek, Ł, Policht, K., Gorna, M., Kutrzuba, J. & Winiarczyk, S. Pierwszy w Polsce przypadek anaplazmozy (erlichiozy) granulocytarnej u kota. Życie Wet. 86, 132–135 (2011).
Google Scholar 
Adaszek, Ł, Kotowicz, W., Klimiuk, P., Gorna, M. & Winiarczyk, S. Ostry przebieg anaplazmozy granulocytarnej u psa—przypadek własny. Wet. w Praktyce 9, 59–62 (2011).
Google Scholar 
Adaszek, Ł et al. Three clinical cases of Anaplasma phagocytophilum infection in cats in Poland. J. Feline Med. Surg. 15, 333–337. https://doi.org/10.1177/1098612X12466552 (2013).Article 
PubMed 

Google Scholar 
Pusterla, N. et al. Seroprevalence of Ehrlichia canis and of canine granulocytic ehrlichia infection in dogs in Switzerland. J. Clin. Microbiol. 36, 3460–3462. https://doi.org/10.1128/JCM.36.12.3460-3462.1998 (1998).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Egenvall, A. et al. Detection of granulocytic Ehrlichia species DNA by PCR in persistently infected dogs. Vet. Rec. 146(7), 186–190. https://doi.org/10.1136/vr.146.7.186 (2000).CAS 
Article 
PubMed 

Google Scholar 
Shaw, S. E. et al. Review of exotic infectious dise-ases in small animals entering the United Kingdom from abroad diagnosed by PCR. Vet. Rec. 152(6), 176–177. https://doi.org/10.1136/vr.152.6.176 (2003).CAS 
Article 
PubMed 

Google Scholar 
Skotarczak, B., Adamska, M. & Supron, M. Blood DNA analysis for Ehrlichia (Anaplasma) phagocytophila and Babesia spp in dogs from Northern Poland. Acta Vet. Brno. 73, 347–351. https://doi.org/10.1136/vr.152.6.176 (2004).Article 

Google Scholar 
Adaszek, Ł. Wybrane Aspekty Epidemiologii Babeszjozy, Boreliozy i Erlichiozy Psów (Praca doktorska, 2007).
Google Scholar 
Kybicová, K. et al. Detection of Anaplasma phagocytophilum and Borrelia burgdorferi sensu lato in dogs in the Czech Republic. Vec. Born Zoon Dis. 9(6), 655–661. https://doi.org/10.1089/vbz.2008.0127 (2009).Article 

Google Scholar  More

Godfray, H. C. J. et al. Science 327, 812–818 (2010).ADS 
CAS 
Article 

Google Scholar 
Pimm, S. L. et al. Science 344, 1246752 (2014).CAS 
Article 

Google Scholar 
Chung, M. G. & Liu, J. Nat. Food https://doi.org/10.1038/s43016-022-00499-7 (2022).Myers, N., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A. & Kent, J. Nature 403, 853–858 (2000).ADS 
CAS 
Article 

Google Scholar 
A complex prairie ecosystem. National Park Service https://www.nps.gov/tapr/learn/nature/a-complex-prairie-ecosystem.htm (2022)Davalos, L. M. et al. Environ. Sci. Technol. 45, 1219–1227 (2011).ADS 
CAS 
Article 

Google Scholar 
Vijay, V., Pimm, S. L., Jenkins, C. N. & Smith, S. J. PLoS ONE 11, e0159668 (2016).Article 

Google Scholar 
Liu, J. et al. Ecol. Soc. 18, 26 (2013).CAS 
Article 

Google Scholar 
Liu, J. Consumption patterns and biodiversity. The Royal Society https://go.nature.com/3M19vup (2020).Xu, Z. et al. Nat. Sustain. 3, 964–971 (2020).Article 

Google Scholar 
Dou, Y., da Silva, R. F. B., Yang, H. & Liu, J. J. Geogr. Sci. 28, 1715–1732 (2018).Article 

Google Scholar  More

Spillover events, in which a pathogen that originates in animals jumps into people, have probably triggered every viral pandemic that’s occurred since the start of the twentieth century1. What’s more, an August 2021 analysis of disease outbreaks over the past four centuries indicates that the yearly probability of pandemics could increase several-fold in the coming decades, largely because of human-induced environmental changes2.Fortunately, for around US$20 billion per year, the likelihood of spillover could be greatly reduced3. This is the amount needed to halve global deforestation in hotspots for emerging infectious diseases; drastically curtail and regulate trade in wildlife; and greatly improve the ability to detect and control infectious diseases in farmed animals.That is a small investment compared with the millions of lives lost and trillions of dollars spent in the COVID-19 pandemic. The cost is also one-twentieth of the statistical value of the lives lost each year to viral diseases that have spilled over from animals since 1918 (see ‘Spillovers: a growing threat’), and less than one-tenth of the economic productivity erased per year1.

Source: Ref. 1

Yet many of the international efforts to better defend the world from future outbreaks, prompted by the COVID-19 pandemic, still fail to prioritize the prevention of spillover. Take, for example, the Independent Panel for Pandemic Preparedness and Response, established by the World Health Organization (WHO). The panel was convened in September 2020, in part to ensure that any future infectious-disease outbreak does not become another pandemic. In its 86-page report released last May, wildlife is mentioned twice; deforestation once.We urge the decision-makers currently developing three landmark international endeavours to make the prevention of spillover central to each.First, the G20 group of the world’s 20 largest economies provisionally agreed last month to create a global fund for pandemics. If realized, this could provide funding at levels that infectious-disease experts have been recommending for decades — around $5 per person per year globally (see go.nature.com/3yjitwx). Second, an agreement to improve global approaches to pandemics is under discussion by the World Health Assembly (WHA), the decision-making body of the WHO. Third, a draft framework for biodiversity conservation — the post-2020 global biodiversity framework — is being negotiated by parties to the Convention on Biological Diversity.Designed in the right way, these three international endeavours could foster a more proactive global approach to infectious diseases. This opportunity — to finally address the factors that drive major disease outbreaks, many of which also contribute to climate change and biodiversity loss — might not present itself again until the world faces another pandemic.Four actions The risk of spillover is greater when there are more opportunities for animals and humans to make contact, for instance in the trade of wildlife, in animal farming or when forests are cleared for mining, farming or roads. It is also more likely to happen under conditions that increase the likelihood of infected animals shedding viruses – when they are housed in cramped conditions, say, or not fed properly.Decades of research from epidemiology, ecology and genetics suggest that an effective global strategy to reduce the risk of spillover should focus on four actions1,3.First, tropical and subtropical forests must be protected. Various studies show that changes in the way land is used, particularly tropical and subtropical forests, might be the largest driver of emerging infectious diseases of zoonotic origin globally4. Wildlife that survives forest clearance or degradation tends to include species that can live alongside people, and that often host pathogens capable of infecting humans5. For example, in Bangladesh, bats that carry Nipah virus — which can kill 40–75% of people infected — now roost in areas of high human population density because their forest habitat has been almost entirely cleared6.Furthermore, the loss of forests is driving climate change. This could in itself aid spillover by pushing animals, such as bats, out of regions that have become inhospitable and into areas where many people live7.Yet forests can be protected even while agricultural productivity is increased — as long as there is enough political will and resources8. This was demonstrated by the 70% reduction in deforestation in the Amazon during 2004–12, largely through better monitoring, law enforcement and the provision of financial incentives to farmers. (Deforestation rates began increasing in 2013 due to changes in environmental legislation, and have risen sharply since 2019 during Jair Bolsonaro’s presidency.)Second, commercial markets and trade of live wild animals that pose a public-health risk must be banned or strictly regulated, both domestically and internationally.Doing this would be consistent with the call made by the WHO and other organizations in 2021 for countries to temporarily suspend the trade in live caught wild mammals, and to close sections of markets selling such animals. Several countries have already acted along these lines. In China, the trade and consumption of most terrestrial wildlife has been banned in response to COVID-19. Similarly, Gabon has prohibited the sale of certain mammal species as food in markets.

A worker in a crowded chicken farm in Anhui province, China.Credit: Jianan Yu/Reuters

Restrictions on urban and peri-urban commercial markets and trade must not infringe on the rights and needs of Indigenous peoples and local communities, who often rely on wildlife for food security, livelihoods and cultural practices. There are already different rules for hunting depending on the community in many countries, including Brazil, Canada and the United States.Third, biosecurity must be improved when dealing with farmed animals. Among other measures, this could be achieved through better veterinary care, enhanced surveillance for animal disease, improvements to feeding and housing animals, and quarantines to limit pathogen spread.Poor health among farmed animals increases their risk of becoming infected with pathogens — and of spreading them. And nearly 80% of livestock pathogens can infect multiple host species, including wildlife and humans9.Fourth, particularly in hotspots for the emergence of infectious diseases, people’s health and economic security should be improved.People in poor health — such as those who have malnutrition or uncontrolled HIV infection — can be more susceptible to zoonotic pathogens. And, particularly in immunosuppressed individuals such as these, pathogens can mutate before being passed on to others10.What’s more, some communities — especially those in rural areas — use natural resources to produce commodities or generate income in a way that brings them into contact with wildlife or wildlife by-products. In Bangladesh, for example, date palm sap, which is consumed as a drink in various forms, is often collected in pots attached to palm trees. These can become contaminated with bodily substances from bats. A 2016 investigation linked this practice to 14 Nipah virus infections in humans that caused 8 deaths11.Providing communities with both education and tools to reduce the risk of harm is crucial. Tools can be something as simple as pot covers to prevent contamination of date palm sap, in the case of the Bangladesh example.In fact, providing educational opportunities alongside health-care services and training in alternative livelihood skills, such as organic agriculture, can help both people and the environment. For instance, the non-governmental organization Health in Harmony in Portland, Oregon, has invested in community-designed interventions in Indonesian Borneo. During 2007–17, these contributed to a 90% reduction in the number of households that were reliant on illegal logging for their main livelihood. This, in turn, reduced local rainforest loss by 70%. Infant mortality also fell by 67% in the programme’s catchment area12.Systems-oriented interventions of this type need to be better understood, and the most effective ones scaled up.Wise investmentSuch strategies to prevent spillover would reduce our dependence on containment measures, such as human disease surveillance, contact tracing, lockdowns, vaccines and therapeutics. These interventions are crucial, but are often expensive and implemented too late — in short, they are insufficient when used alone to deal with emerging infectious diseases.The COVID-19 pandemic has exposed the real-world limitations of these reactive measures — particularly in an age of disinformation and rising populism. For example, despite the US federal government spending more than $3.7 trillion on its pandemic response as of the end of March, nearly one million people in the United States — or around one in 330 — have died from COVID-19 (see go.nature.com/39jtdfh and go.nature.com/38urqvc). Globally, between 15 million and 21 million lives are estimated to have been lost during the COVID-19 pandemic beyond what would be expected under non-pandemic conditions (known as excess deaths; see Nature https://doi.org/htd6; 2022). And a 2021 model indicates that, by 2025, $157 billion will have been spent on COVID-19 vaccines alone (see go.nature.com/3jqds76).

A farmer in Myanmar gathers sap from a palm tree to make wine. Contamination of the collection pots with excretions from bats can spread diseases to humans.Credit: Wolfgang Kaehler/LightRocket via Getty

Preventing spillover also protects people, domesticated animals and wildlife in the places that can least afford harm — making it more equitable than containment. For example, almost 18 months since COVID-19 vaccines first became publicly available, only 21% of the total population of Africa has received at least one dose. In the United States and Canada, the figure is nearly 80% (see go.nature.com/3vrdpfo). Meanwhile, Pfizer’s total drug sales rose from $43 billion in 2020 to $72 billion in 2021, largely because of the company’s COVID-19 vaccine, the best-selling drug of 202113.Lastly, unlike containment measures, actions to prevent spillover also help to stop spillback, in which zoonotic pathogens move back from humans to animals and then jump again into people. Selection pressures can differ across species, making such jumps a potential source of new variants that can evade existing immunity. Some researchers have suggested that spillback was possibly responsible for the emergence of the Omicron variant of SARS-CoV-2 (see Nature 602, 26–28; 2022).Seize the dayOver the past year, the administration of US President Joe Biden and two international panels (one established in 2020 by the WHO and the other in 2021 by the G20) have released guidance on how to improve approaches to pandemics. All recommendations released so far acknowledge spillover as the predominant cause of emerging infectious diseases. None adequately discusses how that risk might be mitigated. Likewise, a PubMed search for the spike protein of SARS-CoV-2 yields thousands of papers, yet only a handful of studies investigate coronavirus dynamics in bats, from which SARS-CoV-2 is likely to have originated14.Spillover prevention is probably being overlooked for several reasons. Upstream animal and environmental sources of pathogens might be being neglected by biomedical researchers and their funders because they are part of complex systems — research into which does not tend to lead to tangible, profitable outputs. Also, most people working in public health and biomedical sciences have limited training in ecology, wildlife biology, conservation and anthropology.There is growing recognition of the importance of cross-sectoral collaboration, including soaring advocacy for the ‘One Health’ approach — an integrated view of health that recognizes links between the environment, animals and humans. But, in general, this has yet to translate into action to prevent pandemics.Another challenge is that it can take decades to realize the benefits of preventing spillover, instead of weeks or months for containment measures. Benefits can be harder to quantify for spillover prevention, no matter how much time passes, because, if measures are successful, no outbreak occurs. Prevention also runs counter to individual, societal and political tendencies to wait for a catastrophe before taking action.The global pandemic fund, the WHA pandemic agreement and the post-2020 global biodiversity framework all present fresh chances to shift this mindset and put in place a coordinated global effort to reduce the risk of spillover alongside crucial pandemic preparedness efforts.Global fund for pandemicsFirst and foremost, a global fund for pandemics will be key to ensuring that the wealth of evidence on spillover prevention is translated into action. Funding for spillover prevention should not be folded into existing conservation funds, nor draw on any other existing funding streams.Investments must be targeted to those regions and practices where the risk of spillover is greatest, from southeast Asia and Central Africa to the Amazon Basin and beyond. Actions to prevent spillover in these areas, particularly by reducing deforestation, would also help to mitigate climate change and reduce loss of biodiversity. But conservation is itself drastically underfunded. As an example, natural solutions (such as conservation, restoration and improved management of forests, wetlands and grasslands) represent more than one-third of the climate mitigation needed by 2030 to stabilize warming to well below 2 °C15. Yet these approaches receive less than 2% of global funds for climate mitigation16. (Energy systems receive more than half.)In short, the decision-makers backing the global fund for pandemics must not assume that existing funds are dealing with the threat of spillover — they are not. The loss of primary tropical forest was 12% higher in 2020 than in 2019, despite the economic downturn triggered by COVID-19. This underscores the continuing threat to forests.Funding must be sustained for decades to ensure that efforts to reduce the risk of spillover are in place long enough to yield results.WHA pandemic agreementIn 2020, the president of the European Council, Charles Michel, called for a treaty to enable a more coordinated global response to major epidemics and pandemics. Last year, more than 20 world leaders began echoing this call, and the WHA launched the negotiation of an agreement (potentially, a treaty or other international instrument) to “strengthen pandemic prevention, preparedness, and response” at the end of 2021.Such a multilateral agreement could help to ensure more-equitable international action around the transfer of scientific knowledge, medical supplies, vaccines and therapeutics. It could also address some of the constraints currently imposed on the WHO, and define more clearly the conditions under which governments must notify others of a potential disease threat. The COVID-19 pandemic exposed the shortcomings of the International Health Regulations on many of these fronts17. (This legal framework defines countries’ rights and obligations in the handling of public-health events and emergencies that could cross borders.)We urge negotiators to ensure that the four actions to prevent spillover outlined here are prioritized in the WHA pandemic agreement. For instance, it could require countries to create national action plans for pandemics that include reducing deforestation and closing or strictly regulating live wildlife markets. A reporting mechanism should also be developed to evaluate progress in implementing the agreement. This could build on experience from existing schemes, such as the WHO Joint External Evaluation process (used to assess countries’ capacities to handle public-health risks) and the verification regime of the Chemical Weapons Convention.Commitments to expand pathogen surveillance at interfaces between humans, domesticated animals and wildlife — from US mink farms and Asian wet markets to areas of high deforestation in South America — should also be wrapped into the WHA agreement. Surveillance will not prevent spillover, but it could enable earlier detection and better control of zoonotic outbreaks, and provide a better understanding of the conditions that cause them. Disease surveillance would improve simply through investing in clinical care for both people and animals in emerging infectious-disease hotspots.Convention on Biological DiversityWe are in the midst of the sixth mass extinction, and activities that drive the loss of biodiversity, such as deforestation, also contribute to the emergence of infectious disease. Meanwhile, epidemics and pandemics resulting from the exploitation of nature can lead to further conservation setbacks — because of economic damage from lost tourism and staff shortages affecting management of protected areas, among other factors18. Also, pathogens that infect people can be transmitted to other animals and decimate those populations. For instance, an Ebola outbreak in the Republic of Congo in 2002–03 is thought to have killed 5,000 gorillas19.Yet the global biodiversity framework currently being negotiated by the Convention on Biological Diversity fails to explicitly address the negative feedback cycle between environmental degradation, wildlife exploitation and the emergence of pathogens. The first draft made no mention of pandemics. Text about spillover prevention was proposed in March, but it has yet to be agreed on.Again, this omission stems largely from the siloing of disciplines and expertise. Just as the specialists relied on for the WHA pandemic agreement tend to be those in the health sector, those informing the Convention on Biological Diversity tend to be specialists in environmental science and conservation.The global biodiversity framework, scheduled to be agreed at the Conference of the Parties later this year, must strongly reflect the environment–health connection. This means explicitly including spillover prevention in any text relating to the exploitation of wildlife and nature’s contributions to people. Failing to connect these dots weakens the ability of the convention to achieve its own objectives around conservation and the sustainable use of resources.Preventive health careA reactive response to catastrophe need not be the norm. In many countries, preventive health care for chronic diseases is widely embraced because of its obvious health and economic benefits. For instance, dozens of colorectal cancer deaths are averted for every 1,000 people screened using colonoscopies or other methods20. A preventive approach does not detract from the importance of treating diseases when they occur.With all the stressors now being placed on the biosphere — and the negative implications this has for human health — leaders urgently need to apply this way of thinking to pandemics. More

Cockroach strainsAll cockroaches were maintained on rodent diet (Purina 5001, PMI Nutrition International, St. Louis, MO) and distilled water at 27 °C, ~40% RH, and a 12:12 h L:D cycle. The WT colony (Orlando Normal) was collected in Florida in 1947 and has served as a standard insecticide-susceptible strain. The GA colony (T-164) was collected in 1989, also in Florida, and shown to be aversive to glucose; continued artificial selection with glucose-containing toxic bait fixed the homozygous GA trait in this population (approximately 150 generations as of 2020).Generating recombinant lines and life history dataTo homogenize the genetic backgrounds of the WT and GA strains, two recombinant colonies were initiated in 2013 by crossing 10 pairs of WT♂ × GA♀ and 10 pairs of GA♂ × WT♀ (Fig. 3a). At the F8 generation (free bulk mating without selection), 400 cockroaches were tested in two-choice feeding assays (see below) that assessed their initial response to tastants, as described in previous studies11,26. The cockroaches were separated into glucose-accepting and glucose-rejecting groups by the rapid Acceptance-Rejection assay (described in Feeding Bioassays). These colonies were bred for three more generations, and 200 cockroaches from each group were assayed in the F11 generation and backcrossed to obtain homozygous glucose-accepting (aa) and glucose-averse (AA) lines. Similar results were obtained in both directions of the cross, confirming previous findings of no sex linkage of the GA trait27. These two lines were defined as WT_aa (homozygotes, glucose-accepting) and GA_AA (homozygotes, glucose-averse). To obtain heterozygous GA cockroaches, GA_Aa, a single intercross group was generated from crosses of 10 pairs of WT_aa♂ × GA_AA♀ and 10 pairs of GA_AA♂ × WT_aa♀.The GA trait follows Mendelian inheritance. Therefore, we used backcrosses, guided by two-choice feeding assays and feeding responses in Acceptance-rejection assays, to determine the homozygosity of WT and GA cockroaches. The cross of WT♂ × WT♀ produced homozygous F1 cockroaches showing maximal glucose-acceptance. The cross of GA♂ × GA♀ produced homozygous F1 cockroaches showing maximal glucose-aversion. The cross of WT × GA produced F1 heterozygotes with intermediate glucose-aversion. When the F1 heterozygotes were backcrossed with WT cockroaches, they produced F2 cockroaches with a 1:1 ratio of WT and GA phenotypes.The two-choice feeding assay assessed whether cockroaches accepted or rejected glucose (binary: yes-no). Insects were held for 24 h without water, or starved without food and water. Either 10 adults or 2 day-old first instar siblings (30–40) were placed in a Petri dish (either 90 mm or 60 mm diameter × 15 mm height). Each Petri dish contained two agar discs: one disc contained 1% agar and 1 mmol l−1 red food dye (Allura Red AC), and the second disc contained 1% agar, 0.5 mmol l−1 blue food dye (Erioglaucine disodium salt) and either 1000 mmol l−1 or 3000 mmol l−1 glucose. The assay duration was 2 h during the dark phase of the insects’ L:D cycle. After each assay, the color of the abdomen of each cockroach was visually inspected under a microscope to infer the genotype.We assessed whether the recombinant colonies had different traits from the parental WT and GA lines. We paired single newly eclosed females (day 0) with single 10–12 days-old males of the same line in a Petri dish (90 mm diameter, 15 mm height) with fresh distilled water in a 1.5 ml microcentrifuge tube and a pellet of rodent food, and monitored when they mated. When females formed egg cases, each gravid female was placed individually in a container (95 × 95 × 80 mm) with food and water until the eggs hatched. After removing the female, her offspring were monitored until adult emergence. We recorded the time to egg hatch, first appearance of each nymphal stage, first appearance of adults and the end of adult emergence. The first instar nymphs and adults in each cohort were counted to obtain measures of survivorship. Although there were significant differences in some of these parameters across all four strains, we found no significant differences between the two recombinant lines, except mating success, which was significantly lower in GA_AA♀ than WT_aa♀ (Supplementary Table 11).Mating bioassaysAll mating sequences were recorded using an infra-red-sensitive camera (Polestar II EQ610, Everfocus Electronics, New Taipei City, Taiwan) coupled to a data acquisition board and analyzed by searchable and frame-by-frame capable software (NV3000, AverMedia Information) at 27 °C, ~40% RH and a 12:12 h L:D cycle. For behavioral analysis, tested pairs were classified into two groups: mated (successful courtship) and not-mated (failed courtship). Four distinct behavioral events (Fig. 1c, Contact, Wing raising, Nuptial feeding, and Copulation) were analyzed using seven behavioral parameters as shown in Supplementary Table 2.We extracted behavioral data from successful courtship sequences, defined as courtship that led to Copulation. For failed courtship sequences, we extracted the behavioral data from the first courtship of both mated and not-mated groups, because most pairs in both groups failed to copulate in their first encounter, and there were no significant differences in behavioral parameters between the two groups.To assay female choice, we conducted two-choice mating assays (Fig. 1a). A single focal WT♀ or GA♀ and two males, one WT and one GA, were placed in a Petri dish (90 mm diameter, 15 mm height) with fresh distilled water in a 1.5 ml microcentrifuge tube and a pellet of rodent food (n = 25 WT♀ and 27 GA♀). To assay male choice, a single focal WT♂ or GA♂ was given a choice of two females, one WT♀ and one GA♀ (n = 27 WT♂ and 18 GA♂). Experiments were started using 0 day-old sexually unreceptive females and 10–12 days-old sexually mature males. Newly emerged (0 day-old) females were used to avoid the disruption of introducing a sexually mature female into the bioassay. B. germanica females become sexually receptive at 5–7 days of age, so the mating behavior of the focal insect was video-recorded for several days until they mated. Fertility of mated females was evaluated by the number of offspring produced. We assessed the gustatory phenotype of nymphs (either WT-type or GA-type) to determine which of the two adult cockroaches mated with the focal insect. Each gravid female was maintained individually in a container (95 × 95 × 80 mm) with food and water until the eggs hatched. Two day-old first instar nymphs were starved for one day without water and food, and then they were tested in Two-choice feeding assays using 1000 mmol l−1 glucose-containing agar with 0.5 mmol l−1 blue food dye vs. plain sugar-free agar with 1 mmol l−1 red food dye. If all the nymphs chose the glucose-containing agar, their parents were considered WT♂ and WT♀. When all the nymphs showed glucose-aversion, they were raised to the adult stage. Newly emerged adults were backcrossed with WT cockroaches, and their offspring were tested in the Two-choice assay. When the parents were both GA, 100% of the offspring exhibited glucose-aversion. When the parents were WT and GA, the offspring showed a 1:1 ratio of glucose-accepting and glucose-aversive behavior. Mate choice, mating success ratio and the number of offspring were analyzed statistically.We conducted no-choice mating assay using the WT and GA strains (Fig. 1b, d). A female and a male were placed in a Petri dish with fresh water and a piece of rodent food and video-recorded for 24 h. The females were 5–7 days-old and males were 10–12 days-old. Four treatment pairs were tested: WT♂ × WT♀ (n = 20, 18 and 14 pairs for 5, 6 and 7 day-old females, respectively); GA♂ × GA♀ (n = 23, 22 and 35 pairs); GA♂ × WT♀ (n = 21, 14 and 17 pairs); and WT♂ × GA♀ (n = 33, 19 and 15 pairs).To confirm that gustatory stimuli guide nuptial feeding, we artificially augmented the male nuptial secretion and assessed whether the duration of nuptial feeding and mating success of GA♀ were affected (Fig. 2c). Before starting the mating assay with 5 day-old GA♀, 10–12 days-old WT♂ were separated into three groups: A control group did not receive any augmentation; A water control group received distilled water with 1 mmol l−1 blue dye (+Blue); A fructose group received 3000 mmol l−1 fructose solution with blue dye (+Blue+Fru). Approximately 50 nl of the test solution was placed into the tergal gland reservoirs using a glass microcapillary. No-choice mating assays were carried out for 24 h. n = 20–25 pairs for each treatment.We evaluated the association of short nuptial feeding (Fig. 1c) and the GA trait we conducted no-choice mating assays using females from the recombinant lines (Fig. 3c). Before starting each mating assay with 4 day-old females from the WT, GA and recombinant lines (WT_aa, GA_AA and GA_Aa), the EC50 for glucose was obtained by the instantaneous Acceptance-Rejection assay using 0, 10, 30, 100, 300, 1000 and 3000 mmol l−1 glucose (WT♀ and WT_aa♀, non-starved; GA♀, GA_AA♀ and GA_Aa♀, 1-day starved). After the Acceptance-Rejection assay, GA_Aa♀ were separated into two groups according to their sensitivity for rejecting glucose; the GA_Aa_high sensitivity group rejected glucose at 100 and 300 mmol l−1, whereas the GA_Aa_low sensitivity group rejected glucose at 1000 and 3000 mmol l−1. We paired these females with 10–12 days-old WT♂ (n = 15 WT_aa♀, n = 20 GA_AA♀, n = 20 GA_Aa_high♀ and n = 17 GA_Aa_low♀).Feeding bioassayWe conducted two feeding assays: Acceptance-Rejection assay and Consumption assay. The Acceptance-Rejection assay assessed the instantaneous initial responses (binary: yes-no) of cockroaches to tastants, as previously described7,22,27. Briefly, acceptance means that the cockroach started drinking. Rejection means that the cockroach never initiated drinking. The percentage of positive responders was defined as the Number of insects accepting tastants/Total number of insects tested. The effective concentration (EC50) for each tastant was obtained from dose-response curves using this assay. The Consumption assay was previously described27. Briefly, we quantified the amount of test solution females ingested after they started drinking. Females were observed until they stopped drinking, and we considered this a single feeding bout.We used the Acceptance-Rejection assay and Consumption assay, respectively, to assess the sensitivity of 5 day-old WT♀ and GA♀ for accepting and consuming the WT♂ nuptial secretion (Fig. 2a, b). The secretion was diluted with HPLC-grade water to 0.001, 0.01, 0.03, 0.1, 0.3 and 1 male-equivalents/µl (n = 20 non-starved females each). The amount of nuptial secretion consumed was tested at 0.1 male-equivalents/µl in the Consumption assay (n = 10 each).The Acceptance-Rejection assay was used to calculate the effective concentration (EC50) of glucose for females in the WT, GA and recombinant lines (Fig. 3a, b). A glucose concentration series of 0.1, 1, 10, 100 and 1000 mmol l−1 was tested with one-day starved 4-day old females (n = 65 GA_Aa♀, n = 50 GA_AA♀ and n = 50 GA♀) and non-starved females (n = 50 WT_aa♀ and n = 16 WT♀).The effects of female saliva on feeding responses of 5 day-old WT♀ and GA♀ were tested using the Acceptance-Rejection assay (Fig. 4a). Freshly collected saliva of WT♀ and GA♀ was immediately used in experiments. Assays were prepared as follows: 3 µl of 200 mmol l−1 maltose or maltotriose were mixed with 3 µl of either HPLC-grade water or saliva of WT♀ or GA♀. The final concentration of each sugar was 100 mmol l−1 in a total volume of 6 µl. This concentration represented approximately the acceptance EC70 for WT♀ and GA♀27. Nuptial secretion (1 µl representing 10 male-equivalents) was mixed with 1 µl of either HPLC-grade water or saliva from WT♀ or GA♀, and 8 µl of HPLC-grade water was added to the mix. The final concentration of the nuptial secretion was 1 male-equivalent/µl in a total volume of 10 µl. This concentration also represented approximately the acceptance EC70 for WT♀ and GA♀ (Fig. 2a). The mix of saliva and either sugar or nuptial secretion was incubated for 300 s at 25 °C. Additionally, we tested the effect of only saliva in the Acceptance-Rejection assay. Either 1-day starved or non-starved females were tested with water only and then a 1:1 mixture of saliva and water. Saliva alone did not affect acceptance or rejection of stimuli. n = 20–33 females from each strain.To evaluate whether salivary enzymes are involved in the hydrolysis of oligosaccharides, the contribution of salivary glucosidases was tested using the glucosidase inhibitor acarbose in the Acceptance-Rejection assay (Fig. 4b), as previously described27. We first confirmed that the range of 0–125 mmol l−1 acarbose in HPLC-grade water did not disrupt the acceptance and rejection of tastants. Test solutions were prepared as follows: 2 µl of either HPLC-grade water or saliva of GA♀ was mixed with 1 µl of either 250 µmol l−1 of acarbose or HPLC-grade water, then the mixture was added to 1 µl of 400 mmol l−1 of either maltose or maltotriose solution. The total volume was 4 µl, with the final concentration of sugar being 100 mmol l−1. For assays with nuptial secretion, 1 µl of either HPLC-grade water or saliva from 5 day-old GA♀ was mixed with 0.5 µl of either 250 µmol l−1 of acarbose or HPLC-grade water. This mixture was added to 0.5 µl of 10 male-equivalents of nuptial secretion (i.e., 20 male-equivalents/µl). HPLC-grade water was added for a total volume of 10 µl and a final concentration of 1 male-equivalent/µl. The mix of saliva and either sugars or nuptial secretion was incubated for 5 min at 25 °C. All test solutions contained blue food dye. Test subjects were 5 day-old GA♀ and 20–25 females were tested in each assay.Nuptial secretion and saliva collectionsThe nuptial secretion of WT♂ was collected by the following method: Five 10–12 days-old males were placed in a container (95 × 95 × 80 mm) with 5 day-old GA♀. After the males displayed wing-raising courtship behavior toward the females, individual males were immediately decapitated and the nuptial secretion in their tergal gland reservoirs was drawn into a calibrated borosilicate glass capillary (76 × 1.5 mm) under the microscope. The nuptial secretions from 30 males were pooled in a capillary and stored at −20 °C until use. Saliva from 5 day-old WT♀ and GA♀ was collected by the following method: individual females were briefly anesthetized with carbon dioxide under the microscope and the side of the thorax was gently squeezed. A droplet of saliva that accumulated on the mouthparts was then collected into a microcapillary (10 µl, Kimble Glass). Fresh saliva was immediately used in experiments.GC-MS procedures for analysis of sugarsStandards of D-( + )-glucose (Sigma-Aldrich), D-( + )-maltose (Fisher Scientific) and maltotriose (Sigma-Aldrich) were diluted in HPLC-grade water (Fisher Scientific) at 10, 50, 100, 500 and 1000 ng/µl to generate calibration curves. Samples were vortexed for 20 s and a 10 μl aliquot of each sample was transferred to a Pyrex reaction vial containing a 10 μl solution of 5 ng/μl sorbitol (≥98%) in HPLC-grade water as internal standard and dried under a gentle flow of N2 for 20 min.Samples containing degradation products from nuptial secretions were prepared by adding 15 μl of HPLC-water to each sample in a 1.5 ml Eppendorf tube, vortexed for 30 s and centrifuged at 8000 rpm (5223 RCF) for 5 min to separate lipids from the water layer. The water phase was transferred to a reaction vial using a glass capillary. This procedure was repeated with the remaining lipid layer and the water layers were combined in the same reaction vial containing 10 μl of a solution of 5 ng/μl sorbitol and dried under N2 for 20 min.For derivatization of sugars and samples, each reaction vial received 12 μl of anhydrous pyridine under a constant N2 flow, then vortexed and incubated at 90 °C for 5 min. Three μl of N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA; Sigma-Aldrich) was added to each reaction vial and centrifuged at 1000 rpm (118 RCF) for 2 min. Vials were incubated in a heat block at 90 °C for 1.5 hr and vortexed every 10 min for the first 30 min of incubation.The total volume of sample was ~10 μl, and 1 μl was injected into the GC-MS (6890 GC coupled to a 5975 MS, Agilent Technologies, Palo Alto, CA). The inlet was operated in splitless mode (17.5 psi) at 290 °C. The GC was equipped with a DB-5 column (30 m, 0.25 mm, 0.25 μm, Agilent), and helium was used as the carrier gas at an average velocity of 50 cm/s. The oven temperature program started at 80 °C for 1 min, increased at 10 °C/min to 180 °C, then increased at 5 °C/min to 300 °C, and held for 10 min. The transfer line was set at 250 °C for 24 min, ramped at 5 °C/min to 300 °C and held until the end of program. The ion source operated at 70 eV and 230 °C, while the MS quadrupole was maintained at 200 °C. The MSD was operated in scan mode, starting after 9 min (solvent delay time) with a mass range of 33–650 AMU.For GC-MS data analysis, the sorbitol peak area was obtained from the extracted ion chromatograms with m/z = 205, the sorbitol base peak. The area of peaks of glucose, maltose and maltotriose were obtained from the extracted ion chromatograms using m/z = 204, the base peak of the three sugars. The most abundant peaks of each sugar were selected for quantification36, and these peaks did not coelute with other peaks. Then, the peak areas of the three sugars were divided by the area of the respective sorbitol peak in each sample to normalize the data and to correct technical variability during sample processing. This procedure was performed to obtain the calibration curves and quantification of sugars in our experiments.The results of sugar analysis using GC-MS are reported in Supplementary Figs. 1–4.Analysis of nuptial secretionsWe focused the GC-MS analysis on glucose, maltose and maltotriose in WT♂ nuptial secretion (Fig. 4c). To quantify the time-course of saliva-catalyzed hydrolysis of WT♂ nuptial secretion to glucose, 1 µl of GA♀ saliva was mixed with 1 µl of 10 male-equivalents/µl. We incubated the mixtures for 0, 5, 10 and 300 s at 25 °C, and added 4 µl of methanol to stop the enzyme activity (n = 5 each treatment). Each sample contained the nuptial secretions of 5 males to obtain enough detectable amount of sugars. For the statistical analysis, the amounts of sugars were divided by 5 to obtain the amount of sugars in 1 male (1 male-equivalent). These amounts were also used for generating Fig. 4c and Supplementary Table 9. In calculations of the concentration of the three sugars (mmol l−1), the mass and volume of the nuptial secretion were measured using 70–130 male-equivalents of undiluted secretion of each strain (n = 3). The mass and volume of the nuptial secretion/male, including both lipid and aqueous layers, were approximately 30–50 µg and 40–50 nl. Because it was difficult to separate the lipid layer from the water layer at this small scale, we roughly estimated that the tergal reservoirs of the four cockroach lines had 30 nl of aqueous layer that contained sugars.To quantify the time-course of saliva-catalyzed hydrolysis of maltose and maltotriose to glucose, 1 µl of GA♀ saliva was mixed with 1 µl of 200 mmol l−1 of either maltose or maltotriose (Fig. 4d, e). Incubation time points were 0, 5, 10 and 300 s at 25 °C and methanol was used to stop the enzyme activity. Controls without saliva were also prepared using HPLC-grade water instead of saliva and 300 s incubations. n = 5 for each treatment.PhotomicroscopyThe photographs of the tergal glands and mouthparts (Fig. 5) were obtained using an Olympus Digital camera attached to an Olympus CX41 microscope (Olympus America, Center Valley, PA).Statistics and reproducibilityThe sample size and number of replicates for each experiment are noted in the respective section describing the experimental details. In summary, the samples sizes were: Mating bioassays, n = 18–80; Feeding assays, n = 16–65; Sugar analysis, n = 5; Life history parameters, n  > 14. All statistical analyses were conducted in R Statistical Software (v4.1.0; R Core Team 2021) and JMP Pro 15.2 software (SAS Institute Inc., Carey, NC). For bioassay data and sugar analysis data, we calculated the means and standard errors, and we used the Chi-square test with Holm’s method for post hoc comparisons, t-test, and ANOVA followed by Tukey’s HSD test (all α = 0.05), as noted in each section describing the experimental details, results, and in Supplementary Tables 1–11.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More

Ricklefs, R. E. A comprehensive framework for global patterns in biodiversity. Ecol. Lett. 7, 1–15 (2004).Article 

Google Scholar 
Dolson, S. J. et al. Diversity and phylogenetic community structure across elevation during climate change in a family of hyperdiverse neotropical beetles (Staphylinidae). Ecography 44, 740–752 (2021).Article 

Google Scholar 
Montaño-Centellas, F. A., McCain, C. & Loiselle, B. A. Using functional and phylogenetic diversity to infer avian community assembly along elevational gradients. Glob. Ecol. Biogeogr. 29, 232–245 (2020).Article 

Google Scholar 
Wiens, J. J. et al. Niche conservatism as an emerging principle in ecology and conservation biology. Ecol. Lett. 13, 1310–1324 (2010).PubMed 
Article 

Google Scholar 
Cavender-Bares, J., Kozak, K. H., Fine, P. V. A. & Kembel, S. W. The merging of community ecology and phylogenetic biology. Ecol. Lett. 12, 693–715 (2009).PubMed 
Article 

Google Scholar 
Mayfield, M. M. & Levine, J. M. Opposing effects of competitive exclusion on the phylogenetic structure of communities. Ecol. Lett. 13, 1085–1093 (2010).PubMed 
Article 

Google Scholar 
Webb, C. O., Ackerly, D. D., McPeek, M. A. & Donoghue, M. J. Phylogenies and community ecology. Annu. Rev. Ecol. Syst. 33, 475–505 (2002).Article 

Google Scholar 
Hubbell, S. P. The Unified Neutral Theory of Biodiversity and Biogeography (MPB-32) Vol. 32 (Princeton University Press, 2001).
Google Scholar 
Kraft, N. J. B., Cornwell, W. K., Webb, C. O. & Ackerly, D. D. Trait evolution, community assembly, and the phylogenetic structure of ecological communities. Am. Nat. 170, 271–283 (2007).PubMed 
Article 

Google Scholar 
Cadotte, M. W. & Tucker, C. M. Should environmental filtering be abandoned?. Trends Ecol. Evol. 32, 429–437 (2017).PubMed 
Article 

Google Scholar 
Mouchet, M. A. et al. Functional diversity measures: An overview of their redundancy and their ability to discriminate community assembly rules. Funct. Ecol. 24, 867–876 (2010).Article 

Google Scholar 
Graham, C. H. & Fine, P. V. A. Phylogenetic beta diversity: Linking ecological and evolutionary processes across space in time. Ecol. Lett. 11, 1265–1277 (2008).PubMed 
Article 

Google Scholar 
Qian, H., Jin, Y., Leprieur, F., Wang, X. & Deng, T. Geographic patterns and environmental correlates of taxonomic and phylogenetic beta diversity for large-scale angiosperm assemblages in China. Ecography 43, 1706–1716 (2020).Article 

Google Scholar 
Swenson, N. G. et al. Phylogenetic and functional alpha and beta diversity in temperate and tropical tree communities. Ecology 93, 112–125 (2012).Article 

Google Scholar 
Qian, H., Hao, Z. & Zhang, J. Phylogenetic structure and phylogenetic diversity of angiosperm assemblages in forests along an elevational gradient in Changbaishan, China. J. Plant Ecol. 7, 154–165 (2014).Article 

Google Scholar 
Chase, J. M. & Myers, J. A. Disentangling the importance of ecological niches from stochastic processes across scales. Philos. Trans. R. Soc. B Biol. Sci. 366, 2351–2363 (2011).Article 

Google Scholar 
Leibold, M. A., Economo, E. P. & Peres-Neto, P. Metacommunity phylogenetics: Separating the roles of environmental filters and historical biogeography. Ecol. Lett. 13, 1290–1299 (2010).PubMed 
Article 

Google Scholar 
Ricklefs, R. E. Evolutionary diversification and the origin of the diversity-environment relationship. Ecology 87, 3–13 (2006).Article 

Google Scholar 
Zhang, J. L. et al. Phylogenetic beta diversity in tropical forests: Implications for the roles of geographical and environmental distance. J. Syst. Evol. 51, 71–85 (2013).Article 

Google Scholar 
Baselga, A. The relationship between species replacement, dissimilarity derived from nestedness, and nestedness. Glob. Ecol. Biogeogr. 21, 1223–1232 (2012).Article 

Google Scholar 
Leprieur, F. et al. Quantifying phylogenetic beta diversity: Distinguishing between ‘true’ turnover of lineages and phylogenetic diversity gradients. PLoS ONE https://doi.org/10.1371/journal.pone.0042760 (2012).Article 
PubMed 
PubMed Central 

Google Scholar 
Bishop, T. R., Robertson, M. P., van Rensburg, B. J. & Parr, C. L. Contrasting species and functional beta diversity in montane ant assemblages. J. Biogeogr. 42, 1776–1786 (2015).PubMed 
PubMed Central 
Article 

Google Scholar 
Economo, E. P., Narula, N., Friedman, N. R., Weiser, M. D. & Guénard, B. Macroecology and macroevolution of the latitudinal diversity gradient in ants. Nat. Commun. 9, 1–8 (2018).CAS 
Article 

Google Scholar 
Lessard, J. P., Fordyce, J. A., Gotelli, N. J. & Sanders, N. J. Invasive ants alter the phylogenetic structure of ant communities. Ecology 90, 2664–2669 (2009).PubMed 
Article 

Google Scholar 
Liu, C., Dudley, K. L., Xu, Z. H. & Economo, E. P. Mountain metacommunities: climate and spatial connectivity shape ant diversity in a complex landscape. Ecography 41, 101–112 (2018).Article 

Google Scholar 
Smith, M. A., Hallwachs, W. & Janzen, D. H. Diversity and phylogenetic community structure of ants along a Costa Rican elevational gradient. Ecography 37, 720–731 (2014).Article 

Google Scholar 
Machac, A., Janda, M., Dunn, R. R. & Sanders, N. J. Elevational gradients in phylogenetic structure of ant communities reveal the interplay of biotic and abiotic constraints on diversity. Ecography 34, 364–371 (2011).Article 

Google Scholar 
Guo, Q. et al. Global variation in elevational diversity patterns. Sci. Rep. 3, 1 (2013).CAS 

Google Scholar 
Kluge, J., Kessler, M. & Dunn, R. R. What drives elevational patterns of diversity? A test of geometric constraints, climate and species pool effects for pteridophytes on an elevational gradient in Costa Rica. Glob. Ecol. Biogeogr. 15, 358–371 (2006).Article 

Google Scholar 
Sanders, N. J., Lessard, J. P., Fitzpatrick, M. C. & Dunn, R. R. Temperature, but not productivity or geometry, predicts elevational diversity gradients in ants across spatial grains. Glob. Ecol. Biogeogr. 16, 640–649 (2007).Article 

Google Scholar 
Malsch, A. K. F. et al. An analysis of declining ant species richness with increasing elevation at Mount Kinabalu, Sabah, Borneo. Asian Myrmecol. 2, 33–49 (2008).
Google Scholar 
Pérez-Toledo, G. R., Valenzuela-González, J. E., Moreno, C. E., Villalobos, F. & Silva, R. R. Patterns and drivers of leaf-litter ant diversity along a tropical elevational gradient in Mexico. J. Biogeogr. 48, 2515 (2021).Article 

Google Scholar 
Szewczyk, T. M. & McCain, C. M. A systematic review of global drivers of ant elevational diversity. PLoS ONE 11, e155040 (2016).Article 

Google Scholar 
McCain, C. M. & Grytnes, J.-A.A. Elevational gradients in species richness. In Encyclopedia of Life Sciences (ed. Wiley, J.) (Wiley, 2010). https://doi.org/10.1002/9780470015902.a0022548.Chapter 

Google Scholar 
Silva, R. R. & Brandão, C. R. F. Morphological patterns and community organization in leaf-litter ant assemblages. Ecol. Monogr. https://doi.org/10.1890/08-1298.1 (2010).Article 

Google Scholar 
Dunn, R. R. et al. Climatic drivers of hemispheric asymmetry in global patterns of ant species richness. Ecol. Lett. 12, 324–333 (2009).PubMed 
Article 

Google Scholar 
Warren, R. J. & Chick, L. Upward ant distribution shift corresponds with minimum, not maximum, temperature tolerance. Glob. Chang. Biol. 19, 2082–2088 (2013).ADS 
PubMed 
Article 

Google Scholar 
Cerdá, X. & Retana, J. Alternative strategies by thermophilic ants to cope with extreme heat: Individual versus colony level traits. Oikos 89, 155–163 (2000).Article 

Google Scholar 
Kadochová, Š & Frouz, J. Thermoregulation strategies in ants in comparison to other social insects, with a focus on red wood ants (Formica rufa group). F1000 Res. 2, 280 (2013).Article 

Google Scholar 
Moreau, C. S., Bell, C. D., Vila, R., Archibald, S. B. & Pierce, N. E. Phylogeny of the ants: diversification in the age of angiosperms. Science 312, 101–104 (2006).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Rabeling, C., Brown, J. M. & Verhaagh, M. Newly discovered sister lineage sheds light on early ant evolution. Proc. Natl. Acad. Sci. 105, 14913–14917 (2008).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Ward, P. S., Brady, S. G., Fisher, B. L. & Schultz, T. R. The evolution of myrmicine ants: Phylogeny and biogeography of a hyperdiverse ant clade (Hymenoptera: Formicidae). Syst. Entomol. 40, 61–81 (2015).Article 

Google Scholar 
Pie, M. R. The macroevolution of climatic niches and its role in ant diversification. Ecol. Entomol. 41, 301–307 (2016).Article 

Google Scholar 
Smith, M. R. Revision of the genus Stenamma Westwood in America north of Mexico (Hymenoptera, Formicidae). Am. Midl. Nat. 57, 133–174 (1957).Article 

Google Scholar 
Herbers, J. M. & Johnson, C. A. Social structure and winter survival in acorn ants. Oikos 116, 829–835 (2007).Article 

Google Scholar 
Kaspari, M. & Weiser, M. D. Ant activity along moisture gradients in a neotropical forest1. Biotropica 32, 703–711 (2006).Article 

Google Scholar 
Flores, O., Seoane, J., Hevia, V. & Azcárate, F. M. Spatial patterns of species richness and nestedness in ant assemblages along an elevational gradient in a Mediterranean mountain range. PLoS ONE 13, 1–16 (2018).
Google Scholar 
Almeida, R. P. S. et al. Induced drought strongly affects richness and composition of ground-dwelling ants in the eastern Amazon. BioRxiv (2020).Le Breton, J., Chazeau, J. & Jourdan, H. Immediate impacts of invasion by Wasmannia auropunctata (Hymenoptera: Formicidae) on native litter ant fauna in a New Caledonian rainforest. Austral Ecol. 28, 204–209 (2003).Article 

Google Scholar 
Vonshak, M., Dayan, T., Ionescu-Hirsh, A., Freidberg, A. & Hefetz, A. The little fire ant Wasmannia auropunctata: A new invasive species in the Middle East and its impact on the local arthropod fauna. Biol. Invasions 12, 1825–1837 (2010).Article 

Google Scholar 
Wheeler, W. M. Ants: Their Structure, Development and Behavior (Columbia University Press, 1910).
Google Scholar 
Cavender-Bares, J., Ackerly, D. D., Baum, D. A. & Bazzaz, F. A. Phylogenetic overdispersion in Floridian oak communities. Am. Nat. 163, 823–843 (2004).CAS 
PubMed 
Article 

Google Scholar 
Parr, C. L., Sinclair, B. J., Andersen, A. N., Gaston, K. J. & Chown, S. L. Constraint and competition in assemblages: A cross-continental and modeling approach for ants. Am. Nat. 165, 481–494 (2005).PubMed 
Article 

Google Scholar 
Retana, J. & Cerdá, X. Patterns of diversity and composition of Mediterranean ground ant communities tracking spatial and temporal variability in the thermal environment. Oecologia 123, 436–444 (2000).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Hawkins, B. A. et al. Energy, water, and broad-scale geographic patterns of species richness. Ecology 84, 3105–3117 (2003).Article 

Google Scholar 
Graham, C. H., Parra, J. L., Rahbek, C. & McGuire, J. A. Phylogenetic structure in tropical hummingbird communities. Proc. Natl. Acad. Sci. 106, 19673–19678 (2009).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Camacho, G. P., Loss, A. C., Fisher, B. L. & Blaimer, B. B. Spatial phylogenomics of acrobat ants in Madagascar—Mountains function as cradles for recent diversity and endemism. J. Biogeogr. 1, 1706–1719. https://doi.org/10.1111/jbi.14107 (2021).Article 

Google Scholar 
Lobo, J. M. & Halffter, G. Biogeographical and ecological factors affecting the altitudinal variation of mountainous communities of coprophagous beetles (Coleoptera: Scarabaeoidea): A comparative study. Ann. Entomol. Soc. Am. 93, 115–126 (2000).Article 

Google Scholar 
Halffter, G., Favila, M. & Arellano, L. Spatial distribution of three groups of Coleoptera along an altitudinal transect in the Mexican Transition Zone and its biogeographical implications. Elytron 9, 1–10 (1995).
Google Scholar 
Blaimer, B. B. et al. Phylogenomic methods outperform traditional multi-locus approaches in resolving deep evolutionary history: a case study of formicine ants. BMC Evol. Biol. 15, 1–14 (2015).Article 

Google Scholar 
Longino, J. T., Branstetter, M. G. & Colwell, R. K. How ants drop out: ant abundance on tropical mountains. PLoS ONE 9, e104030 (2014).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Longino, J. T. & Branstetter, M. G. The truncated bell: An enigmatic but pervasive elevational diversity pattern in Middle American ants. Ecography 42, 272–283 (2019).Article 

Google Scholar 
Branstetter, M. G. Origin and diversification of the cryptic ant genus Stenamma Westwood (Hymenoptera: Formicidae), inferred from multilocus molecular data, biogeography and natural history. Syst. Entomol. 37, 478–496 (2012).Article 

Google Scholar 
Prebus, M. Insights into the evolution, biogeography and natural history of the acorn ants, genus Temnothorax Mayr (hymenoptera: Formicidae). BMC Evol. Biol. 17, 1–22 (2017).Article 

Google Scholar 
Kluge, J. & Kessler, M. Phylogenetic diversity, trait diversity and niches: Species assembly of ferns along a tropical elevational gradient. J. Biogeogr. 38, 394–405 (2011).Article 

Google Scholar 
Janzen, D. H. Why mountain passes are higher in the tropics. Am. Nat. 101, 233–249 (1967).Article 

Google Scholar 
Fernandes, G. W. et al. Cerrado to rupestrian grasslands: Patterns of species distribution and the forces shaping them along an altitudinal gradient. in Ecology and Conservation of Mountaintop Grasslands in Brazil 345–378 (2016). https://doi.org/10.1007/978-3-319-29808-5_15.Leibold, M. A. et al. The metacommunity concept: A framework for multi-scale community ecology. Ecol. Lett. 7, 601–613 (2004).Article 

Google Scholar 
Perrigo, A., Hoorn, C. & Antonelli, A. Why mountains matter for biodiversity. J. Biogeogr. 47, 315–325 (2020).Article 

Google Scholar 
Myers, N., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A. B. & Kent, J. Biodiversity hotspots for conservation priorities. Nature 403, 853–858 (2000).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Colwell, R. K., Brehm, G., Cardelus, C. L., Gilman, A. C. & Longino, J. T. Global warming, elevational range shifts and lowland biotic attrition in the wet tropics. Science 322, 258–261 (2008).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Moreau, C. S. & Bell, C. D. Testing the museum versus cradle tropical biological diversity hypothesis: Phylogeny, diversification, and ancestral biogeographic range evolution of the ants. Evolution 67, 2240–2257 (2013).PubMed 
Article 

Google Scholar 
Borowiec, M. L. Generic revision of the ant subfamily Dorylinae (Hymenoptera, Formicidae). Zookeys 1, 280 (2016).
Google Scholar 
Lapolla, J. S., Brady, S. G. & Shattuck, S. O. Phylogeny and taxonomy of the Prenolepis genus-group of ants (Hymenoptera: Formicidae). Syst. Entomol. 35, 118–131 (2010).Article 

Google Scholar 
Schmidt, C. A. & Shattuck, S. O. The higher classification of the ant subfamily Ponerinae (Hymenoptera: Formicidae), with a review of ponerine ecology and behavior. Zootaxa 3817, 1–242 (2014).CAS 
PubMed 
Article 

Google Scholar 
Revell, L. J. phytools: An R package for phylogenetic comparative biology (and other things). Methods Ecol. Evol. 3, 217–223 (2012).Article 

Google Scholar 
Arnan, X., Arcoverde, G. B., Pie, M. R., Ribeiro-Neto, J. D. & Leal, I. R. Increased anthropogenic disturbance and aridity reduce phylogenetic and functional diversity of ant communities in Caatinga dry forest. Sci. Total Environ. 631, 429–438 (2018).ADS 
PubMed 
Article 

Google Scholar 
Divieso, R., Silva, T. S. R. & Pie, M. R. Morphological evolution in the ant reproductive caste. BioRxiv https://doi.org/10.1101/2020.07.18.210302 (2020).Article 

Google Scholar 
Paradis, E. et al. Package ‘ape’. Anal. Phylogenet. Evol. 2, 1–10 (2019).
Google Scholar 
Faith, D. P. Conservation evaluation and phylogenetic diversity. Biol. Conserv. 61, 1–10 (1992).Article 

Google Scholar 
Tucker, C. M. et al. A guide to phylogenetic metrics for conservation, community ecology and macroecology. Biol. Rev. 92, 698–715 (2017).PubMed 
Article 

Google Scholar 
Webb, C. O. Exploring the phylogenetic structure of ecological communities: An example for rain forest trees. Am. Nat. 156, 145–155 (2000).PubMed 
Article 

Google Scholar 
Tucker, C. M. et al. Assessing the utility of conserving evolutionary history. Biol. Rev. 94, 1740–1760 (2019).PubMed 
Article 

Google Scholar 
Kembel, S. W. et al. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26, 1463–1464 (2010).CAS 
PubMed 
Article 

Google Scholar 
R Core Team. A language and environment for statistical computing. R Found. Stat. Comput. 2, https://www.R-project.org (2021).Baselga, A. & Orme, C. D. L. Betapart: An R package for the study of beta diversity. Methods Ecol. Evol. 3, 808–812 (2012).Article 

Google Scholar 
Dobrovolski, R., Melo, A. S., Cassemiro, F. A. S. & Diniz-Filho, J. A. F. Climatic history and dispersal ability explain the relative importance of turnover and nestedness components of beta diversity. Glob. Ecol. Biogeogr. 21, 191–197 (2012).Article 

Google Scholar 
Peixoto, F. P. et al. Geographical patterns of phylogenetic beta-diversity components in terrestrial mammals. Glob. Ecol. Biogeogr. 26, 573–583 (2017).Article 

Google Scholar 
Körner, C. The use of ‘altitude’ in ecological research. Trends Ecol. Evol. 22, 569–574 (2007).PubMed 
Article 

Google Scholar 
Sundqvist, M. K., Sanders, N. J. & Wardle, D. A. Community and ecosystem responses to elevational gradients: Processes, mechanisms, and insights for global change. Annu. Rev. Ecol. Evol. Syst. 44, 261–280 (2013).Article 

Google Scholar 
Cuervo-Robayo, A. P. et al. An update of high-resolution monthly climate surfaces for Mexico. Int. J. Climatol. 34, 2427–2437 (2014).Article 

Google Scholar 
Hijmans, R. J., Phillips, S., Leathwick, J., Elith, J. & Hijmans, M. R. J. Package ‘dismo’. Circles 9, 1–68 (2017).
Google Scholar 
Schwarz, G. Estimating the dimension of a model. Ann. Stat. 6, 461–464 (1978).MathSciNet 
MATH 
Article 

Google Scholar 
Guthery, F. S., Burnham, K. P. & Anderson, D. R. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach Vol. 67 (Springer, 2003).
Google Scholar 
Mazerolle, M. J. Improving data analysis in herpetology: Using Akaike’s information criterion (AIC) to assess the strength of biological hypotheses. Amphib. Reptil. 27, 169–180 (2006).Article 

Google Scholar 
Ferrier, S., Manion, G., Elith, J. & Richardson, K. Using generalized dissimilarity modelling to analyse and predict patterns of beta diversity in regional biodiversity assessment. Divers. Distrib. 13, 252–264 (2007).Article 

Google Scholar 
Fitzpatrick, M. C. et al. Environmental and historical imprints on beta diversity: Insights from variation in rates of species turnover along gradients. Proc. R. Soc. B Biol. Sci. 280, 20131201 (2013).Article 

Google Scholar 
Manion, G. et al. gdm: Generalized dissimilarity modeling. R Packag. version (2018).Wickham, H. Ggplot2. Wiley Interdiscip. Rev. Comput. Stat. 3, 180–185 (2011).Article 

Google Scholar  More

1500-year stable carbon and oxygen isotopes in larch tree-ring celluloseThe δ13Ccell (Fig. 1a, Fig. S2) and δ18Ocell (Fig. 1b, Fig. S3) records span 516–2016 CE, at annual resolution. The δ13Ccell timeseries shows mostly increasing trends during the first millennium of the Common Era (516–1120 CE), and similarly at the end of the last millennium (1720–2016 CE). The maximum δ13Ccell value occurs in 2016 CE (−19.6‰; + 3.2σ), while the minimum occurs in 686 CE (−24.7‰, −3.6σ) relative to the average for the period 516–2016 CE (−22.04‰) (Table S2, Fig. S2). The standard error (SE) for the whole analysed period is 0.02.Figure 1Annually resolved δ13Ccell (a) and δ18O cell (b) in Siberian larch tree-ring cellulose chronologies for the period from 516 to 2016 CE. Chronologies are smoothed by a 101-year Hamming window to highlight a centennial scale. The dotted and dashed lines indicate the number of trees analysed.Full size imageThe δ18Ocell timeseries (Fig. 1b, Fig. S3) showed two positive and one negative extreme over the past 1500 years, with the minimum value (19.9‰; −6.3σ), occurring in 536 CE, and maximum values (31.9‰; + 3.8σ and 32.2‰; + 4.4σ), occurring in 1266 and 2008 CE, respectively (Table S2, Fig. S3). The SE for the whole analysed period is 0.03. The δ18Ocell data has higher standard deviation (SD) (1.15) than δ13Ccell (0.75).Less than 1% of values in the δ18Ocell record are classified as extreme, with the standard deviation ≥  ± 3σ. The δ13Ccell and δ18Ocell records are significantly correlated (r = 0.1, p = 0.0001, n = 1500).Local climate signals preserved in δ13Ccell and δ18Ocell recordsWe used weather observations from the local Mugur-Aksy weather station (50°N, 90°E, 1850 m asl) (Table S1) to derive quantitative paleoclimatic reconstructions from our δ13Ccell and δ18Ocell timeseries. A multiple linear regression analysis revealed significant correlations between δ13Ccell and July precipitation (r = −0.58; p  More

Foster, G. & Rahmstorf, S. Global temperature evolution 1979–2010. Environ. Res. Lett. 6, 044022 (2011).ADS 
Article 

Google Scholar 
Cahill, N., Rahmstorf, S. & Parnell, A. C. Change points of global temperature. Environ. Res. Lett. 10, 084002 (2015).ADS 
Article 

Google Scholar 
Yan, X. H. et al. The global warming hiatus: Slowdown or redistribution?. Earth’s Future 4, 472–482 (2016).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Karl, T. R. et al. Possible artifacts of data biases in the recent global surface warming hiatus. Science 348, 5632 (2015).Article 

Google Scholar 
Cohen, J. L., Furtado, J. C., Barlow, M., Alexeev, V. A. & Cherry, J. E. Asymmetric seasonal temperature trends. Geophys. Res. Lett. 39, 04705. https://doi.org/10.1029/2011GL050582 (2012).ADS 
Article 

Google Scholar 
Pepin, N. C. & Lundquist, J. D. Temperature trends at high elevations: patterns across the globe. Geophys. Res. Lett. 35, 14 (2008).Article 

Google Scholar 
Rangwala, I. & Miller, J. R. Climate change in mountains: a review of elevation-dependent warming and its possible causes. Clim. Change 114, 527–547 (2012).ADS 
Article 

Google Scholar 
Wang, Q., Fan, X. & Wang, M. Recent warming amplification over high elevation regions across the globe. Clim. Dyn. 43, 87–101 (2014).CAS 
Article 

Google Scholar 
Fan, X., Wang, Q., Wang, M. & Jiménez, C. V. Warming amplification of minimum and maximum temperatures over high-elevation regions across the globe. PLoS ONE 10, e0140213 (2015).PubMed 
PubMed Central 
Article 

Google Scholar 
Pepin, N. et al. Elevation-dependent warming in mountain regions of the world. Nat. Clim. Chang. 5, 424 (2015).ADS 
Article 

Google Scholar 
Piccarreta, M., Lazzari, M. & Pasini, A. Trends in daily temperature extremes over the Basilicata region (southern Italy) from 1951 to 2010 in a Mediterranean climatic context. Int. J. Climatol. 35, 1964–1975 (2015).Article 

Google Scholar 
Gonzalez-Hidalgo, J. C., Peña-Angulo, D., Brunetti, M. & Cortesi, N. Recent trend in temperature evolution in Spanish mainland (1951–2010): from warming to hiatus. Int. J. Climatol. 36, 2405–2416 (2016).Article 

Google Scholar 
McCullough, I. M. et al. High and dry: high elevations disproportionately exposed to regional climate change in Mediterranean-climate landscapes. Landsc. Ecol. 31, 1063–1075 (2016).Article 

Google Scholar 
Sanz-Elorza, M., Dana, E. D., González, A. & Sobrino, E. Changes in the high-mountain vegetation of the central Iberian Peninsula as a probable sign of global warming. Ann. Bot. 92, 273–280 (2003).PubMed 
PubMed Central 
Article 

Google Scholar 
Peñuelas, J. & Boada, M. A global change induced biome shift in the Montseny mountains (NE Spain). Glob. Change Biol. 9, 131–140 (2003).ADS 
Article 

Google Scholar 
Linares, J. C. & Tíscar, P. A. Climate change impacts and vulnerability of the southern populations of Pinus nigra subsp. salzmannii. Tree Physiol. 30, 795–806 (2010).PubMed 
Article 

Google Scholar 
Giorgi, F., Hurrell, J. W., Marinucci, M. R. & Beniston, M. Elevation dependency of the surface climate change signal: a model study. J. Clim. 10, 288–296 (1997).ADS 
Article 

Google Scholar 
Palazzi, E., Mortarini, L., Terzago, S. & Von Hardenberg, J. Elevation-dependent warming in global climate model simulations at high spatial resolution. Clim. Dyn. 52, 2685–2702 (2019).Article 

Google Scholar 
Poyatos, R., Latron, J. & Llorens, P. Land use and land cover change after agricultural abandonment. Mt. Res. Dev. 23, 362–368 (2003).Article 

Google Scholar 
Mouillot, F., Ratte, J. P., Joffre, R., Mouillot, D. & Rambal, S. Long-term forest dynamic after land abandonment in a fire prone Mediterranean landscape (central Corsica, France). Landsc. Ecol. 20, 101–112 (2005).Article 

Google Scholar 
Zellweger, F. et al. Forest microclimate dynamics drive plant responses to warming. Science 368, 772–775 (2020).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Ríos-Cornejo, D., Penas, Á., Álvarez-Esteban, R. & Del Río, S. Links between teleconnection patterns and mean temperature in Spain. Theor. Appl. Climatol. 122, 1–18 (2015).ADS 
Article 

Google Scholar 
Nogués-Bravo, D., Araújo, M. B., Errea, M. P. & Martinez-Rica, J. P. Exposure of global mountain systems to climate warming during the 21st Century. Glob. Environ. Chang. 17, 420–428 (2007).Article 

Google Scholar 
Vicente-Serrano, S. M., Beguería, S., López-Moreno, J. I., El Kenawy, A. M. & Angulo-Martínez, M. Daily atmospheric circulation events and extreme precipitation risk in northeast Spain: Role of the North Atlantic Oscillation, the Western Mediterranean Oscillation, and the Mediterranean Oscillation. J. Geophys. Res. Atmos. 114, D8 (2009).Article 

Google Scholar 
Guzman-Morales, J. & Gershunov, A. Climate change suppresses Santa Ana winds of southern California and sharpens their seasonality. Geophys. Res. Lett. 46, 2772–2780. https://doi.org/10.1029/2018GL080261 (2019).ADS 
Article 

Google Scholar 
Yu, M. & Ruggieri, E. Change point analysis of global temperature records. Int. J. Climatol. 39, 3679–3688 (2019).Article 

Google Scholar 
Giorgi, F. Climate change hot-spots. Geophys. Res. Lett. 33, 08707. https://doi.org/10.1029/2006GL025734 (2006).ADS 
Article 

Google Scholar 
García, M. J. L. Recent warming in the Balearic Sea and Spanish Mediterranean coast: Towards an earlier and longer summer. Atmósfera 28, 149–160 (2015).Article 

Google Scholar 
Toreti, A., Desiato, F., Fioravanti, G. & Perconti, W. Seasonal temperatures over Italy and their relationship with low-frequency atmospheric circulation patterns. Clim. Change 99, 211–227 (2010).ADS 
Article 

Google Scholar 
Scorzini, A. R. & Leopardi, M. Precipitation and temperature trends over central Italy (Abruzzo Region): 1951–2012. Theor. Appl. Climatol. 135, 959–977 (2019).ADS 
Article 

Google Scholar 
Lee, X. et al. Observed increase in local cooling effect of deforestation at higher latitudes. Nature 479, 384–387 (2011).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Juang, J.-Y., Katul, G., Siqueira, M., Stoy, P. & Novick, K. Separating the effects of albedo from eco-physiological changes on surface temperature along a successional chronosequence in the southeastern United States. Geophys. Res. Lett. 34, 21408. https://doi.org/10.1029/2007.GL03129 (2007).ADS 
Article 

Google Scholar 
Boulant, N., Kunstler, G., Rambal, S. & Lepart, J. Seed supply, drought, and grazing determine spatio-temporal patterns of recruitment for native and introduced invasive pines in grasslands. Divers. Distrib. 14, 862–874 (2008).Article 

Google Scholar 
Améztegui, A. Land-use changes as major drivers of mountain pine (Pinus uncinata Ram.) expansion in the Pyrenees. Glob. Ecol. Biogeogr. 19, 632–641 (2010).
Google Scholar 
Rambal, S. Relations entre couverts végétaux des parcours et cycle de l’eau. In L’eau des troupeaux en alpages et sur parcours: une ressource à gérer, aménager, partager (ed. Lepart, J.) 25–37 (Association Française de Pastoralisme et Cardère éditeur, 2015).
Google Scholar 
Fonderflick, J., Lepart, J., Caplat, P., Debussche, M. & Marty, P. Managing agricultural change for biodiversity conservation in a Mediterranean upland. Biol. Conserv. 143, 737–746 (2010).Article 

Google Scholar 
Abadie, J. et al. Forest recovery since 1860 in a Mediterranean region: Drivers and implications for land use and land cover spatial distribution. Landsc. Ecol. 33, 289–305 (2018).Article 

Google Scholar 
Cervera, T., Pino, J., Marull, J., Padró, R. & Tello, E. Understanding the long-term dynamics of forest transition: From deforestation to afforestation in a Mediterranean landscape (Catalonia, 1868–2005). Land Use Policy 80, 318–331 (2019).Article 

Google Scholar 
Wolpert, F., Quintas-Soriano, C. & Plieninger, T. Exploring land-use histories of tree-crop landscapes: a cross-site comparison in the Mediterranean Basin. Sustain. Sci. 15, 1267–1283 (2020).Article 

Google Scholar 
Lasanta-Martínez, T., Vicente-Serrano, S. M. & Cuadrat-Prats, J. M. Mountain Mediterranean landscape evolution caused by the abandonment of traditional primary activities: A study of the Spanish Central Pyrenees. Appl. Geogr. 25, 47–65 (2005).Article 

Google Scholar 
Malandra, F., Vitali, A., Urbinati, C., Weisberg, P. J. & Garbarino, M. Patterns and drivers of forest landscape change in the Apennines range, Italy. Reg. Environ. Change 19, 1973–1985 (2019).Article 

Google Scholar 
Zhang, Q. et al. Reforestation and surface cooling in temperate zones: Mechanisms and implications. Glob. Change Biol. 26, 3384–3401 (2020).ADS 
Article 

Google Scholar 
Rambal, S., Lacaze, B. & Winkel, T. Testing an area-weighted model for albedo or surface temperature of mixed pixels in Mediterranean woodlands. Int. J. Remote Sens. 11, 1495–1499 (1990).Article 

Google Scholar 
Luyssaert, S. et al. Land management and land-cover change have impacts of similar magnitude on surface temperature. Nat. Clim. Change 4, 389–393. https://doi.org/10.1038/nclimate2196 (2014).ADS 
Article 

Google Scholar 
Novick, K. A. & Katul, G. G. The duality of reforestation impacts on surface and air temperature. J. Geophys. Res. Biogeosci. 125, e05543 (2020).Article 

Google Scholar 
Davy, R. & Esau, I. Differences in the efficacy of climate forcings explained by variations in atmospheric boundary layer depth. Nat. Commun. 7, 1–8 (2016).Article 

Google Scholar 
Serafin, S. et al. Exchange processes in the atmospheric boundary layer over mountainous terrain. Atmosphere 9, 102. https://doi.org/10.3390/atmos9030102 (2018).ADS 
CAS 
Article 

Google Scholar 
Perugini, L. et al. Biophysical effects on temperature and precipitation due to land cover change. Environ. Res. Lett. 12, 053002 (2017).ADS 
Article 

Google Scholar 
Visbeck, M. H., Hurrell, J. W., Polvani, L. & Cullen, H. M. The North Atlantic oscillation: Past, present, and future. Proc. Natl. Acad. Sci. 98, 12876–12877 (2001).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Hurrell, J. W. Decadal trends in the North Atlantic oscillation: Regional temperatures and precipitation. Science 269, 676–679 (1995).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Martín, P., Sabatés, A., Lloret, J. & Martin-Vide, J. Climate modulation of fish populations: the role of the Western Mediterranean Oscillation (WeMO) in sardine (Sardina pilchardus) and anchovy (Engraulis encrasicolus) production in the north-western Mediterranean. Clim. Change 110, 925–939 (2012).ADS 
Article 

Google Scholar 
Schwingshackl, C., Hirschi, M. & Seneviratne, S. I. Global contributions of incoming radiation and land surface conditions to maximum near surface air temperature variability and trend. Geophys. Res. Lett. 45, 5034–5044 (2018).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Philipona, R., Behrens, K. & Ruckstuhl, C. How declining aerosols and rising greenhouse gases forced rapid warming in Europe since the 1980s. Geophys. Res. Lett. 36, L02806. https://doi.org/10.1029/2008GL036350 (2009).ADS 
Article 

Google Scholar 
Schwarz, M., Folini, D., Yang, S., Allan, R. P. & Wild, M. Changes in atmospheric shortwave absorption as important driver of dimming and brightening. Nat. Geosci. 13, 110–115 (2020).ADS 
CAS 
Article 

Google Scholar 
Norris, J. R. & Wild, M. Trends in aerosol radiative effects over Europe inferred from observed cloud cover, solar “dimming”, and solar “brightening”. J. Geophys. Res. Atmos. 112, D08214. https://doi.org/10.1029/2006JD007794 (2007).ADS 
Article 

Google Scholar 
Mateos, D. et al. Quantifying the respective roles of aerosols and clouds in the strong brightening since the early 2000s over the Iberian Peninsula. J. Geophys. Res. Atmos. 119, 10–382 (2014).Article 

Google Scholar 
Sanchez-Lorenzo, A. et al. Reassessment and update of long-term trends in downward surface shortwave radiation over Europe (1939–2012). J. Geophys. Res. Atmos. 120, 9555–9569 (2015).ADS 
Article 

Google Scholar 
Kambezidis, H. D., Kaskaoutis, D. G., Kalliampakos, G. K., Rashki, A. & Wild, M. The solar dimming/brightening effect over the Mediterranean Basin in the period 1979–2012. J. Atmos. Solar Terr. Phys. 150, 31–46 (2016).ADS 
Article 

Google Scholar 
Chiacchio, M. & Wild, M. Influence of NAO and clouds on long-term seasonal variations of surface solar radiation in Europe. J. Geophys. Res. Atmos. 115, 0022. https://doi.org/10.1029/2009JD012182 (2010).Article 

Google Scholar 
Wild, M. Decadal changes in radiative fluxes at land and ocean surfaces and their relevance for global warming. Wiley Interdiscipl. Rev. Clim. Change 7, 91–107 (2016).Article 

Google Scholar 
Held, I. M. & Soden, B. J. Water vapor feedback and global warming. Annu. Rev. Energy Environ. 25, 441–475 (2000).Article 

Google Scholar 
Dessler, A. E. & Sherwood, S. C. A matter of humidity. Science 323, 1020–1021 (2009).CAS 
PubMed 
Article 

Google Scholar 
Ruckstuhl, C., Philipona, R., Morland, J. & Ohmura, A. Observed relationship between surface specific humidity, integrated water vapor, and longwave downward radiation at different altitudes. J. Geophys. Res. Atmos. 112(D03302), 2007. https://doi.org/10.1029/2006JD007850 (2007).Article 

Google Scholar 
Parras-Berrocal, I. M. et al. The climate change signal in the Mediterranean Sea in a regionally coupled atmosphere–ocean model. Ocean Sci. 16, 743–765. https://doi.org/10.5194/os-16-743-2020 (2020).ADS 
Article 

Google Scholar 
Reale, M. et al. The regional earth system model RegCM-ES: Evaluation of the Mediterranean climate and marine biogeochemistry. J. Adv. Model. Earth Syst. 12, e001812 (2020).Article 

Google Scholar 
Sen, P. K. Estimates of the regression coefficient based on Kendall’s tau. J. Am. Stat. Assoc. 63, 1379–1389 (1968).MathSciNet 
MATH 
Article 

Google Scholar 
Kelliher, F. M., Leuning, R. & Schulze, E. D. Evaporation and canopy characteristics of coniferous forests and grasslands. Oecologia 95, 153–163 (1993).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Linacre, E. T. Simpler empirical expression for actual evapotranspiration rates-a discussion. Agric. Meteorol. 11, 451–452 (1973).Article 

Google Scholar 
Jones, P. D., Jónsson, T. & Wheeler, D. Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and south-west Iceland. Int. J. Climatol. 17, 1433–1450 (1997).Article 

Google Scholar 
Palutikof, J. P. Analysis of Mediterranean climate data: measured and modelled. In Mediterranean Climate: Variability and Trends (ed. Bolle, H. J.) (Springer, 2003).
Google Scholar 
Martin-Vide, J. & Lopez-Bustins, J. A. The western Mediterranean oscillation and rainfall in the Iberian Peninsula. Int. J. Climatol. 26, 1455–1475 (2006).Article 

Google Scholar  More

We selected eleven major wheat production provinces of China for the study area, which comprise the largest winter wheat-sowing fraction: Henan, Shandong, Anhui, Jiangsu, Hebei, Hubei, Shanxi, Shaanxi, Sichuan, Xinjiang, and Gansu (Fig. 1). The wheat planting area is about 22 million ha in these provinces, accounting for more than 93% of the total wheat planting area. The total wheat production in these regions contributes more than 96% of the total wheat production in China, with more than 128 million tons in 201933.We developed a methodological framework for high-resolution AGB mapping. It mainly includes three parts: (1) Data acquisition and processing. (2) The WOFOST model parameterization and calibration. (3) Data assimilation (Fig. 1). Each part is explained in more detail below.Data acquisition and processingMeteorological dataChina Meteorological Forcing Dataset34,35 is used as weather driving data for the WOFOST model. The dataset is based on the internationally existing Princeton reanalysis data, Global Land Data Assimilation System data, Global Energy and Water Cycle Experiment-Surface Radiation Budget radiation data, and Tropical Rainfall Measuring Mission precipitation data. It is made by fusing the conventional meteorological observation data of the China Meteorological Administration. It includes seven elements: near-surface air temperature, air pressure, near-surface total humidity, wind speed, ground downward shortwave radiation, ground downward longwave radiation, and ground precipitation rate. The meteorological drive elements required for WOFOST are daily radiation, minimum temperature, maximum temperature, water vapor pressure, average wind speed, and precipitation. Details of these variables that participated in the WOFOST model can be referred to in Table S1.Soil characteristics measurements and phenology observationsSoil and phenology data were collected at 177 agricultural meteorological stations (AMS) from 2007 to 2015 (Fig. 1). Soil characteristics include soil moisture content at wilting points, field capacity, and saturation. To be consistent with the corresponding units in the crop model, the original data in weight water content was converted into volume water content through the corresponding soil bulk density measurements. Winter wheat phenology observations include the date of emergence (more than 50% of the wheat seedlings in the field show the first green leaves and reached about 2 cm), anthesis (the inner and outer glumes of the middle and upper florets of more than 50% of the wheat ears in the whole field are open, and the anthers loose powder), and maturity (more than 80% of the wheat grains turn yellow, the glumes and stems turn yellow, and only the upper first and second nodes are still slightly green). In most cases, the phenological stage “anthesis” is missing. The anthesis date was calculated by adding seven days to the observed heading date (when more than 50% of the wheat in the whole field exposes the tip of the ear from the sheath of the flag leaf).County-level yield statistics dataThe county-level yield data was collected from city statistical yearbooks of the study area from 2007 to 2015. Since most statistical yearbooks do not directly record per-unit yield data, the county-level yield was obtained by dividing the total yield and planting area. It is worth noting that all yields were calculated in units of metric kilograms per cultivated hectares (kg·ha−1).The winter wheat land cover dataWe used a winter wheat land cover product from a 1 km resolution dataset named ChinaCropArea1km36. This data was derived from GLASS leaf area index products and crop phenology from 2000 to 2015. This dataset is the base map of our data production.The MODIS LAI dataWe used the improved 8-days MODIS LAI products (i.e., 1 km) generated by Yuan et al.32 to assimilate the WOFOST model. The products applied the modified temporal-spatial filter and Savitzky-Golay filter to overcome the spatial-temporal discontinuity and inconsistence of raw MODIS LAI products, which makes them more applicable for the realm of land surface and climate modeling. The products can be accessed via the Land-Atmosphere Interaction Research Group website at Sun Yat-sen University (http://globalchange.bnu.edu.cn/research/lai).The WOFOST model parameterization and calibrationThe WOFOST model introductionThe WOFOST model was initially developed as a crop growth simulation model to evaluate the yield potential of various crops in tropical countries37. In this study, we chose the WOFOST model because the model reaches a trade-off of the complexity of the crop model and is suitable for large-scale simulations3. The WOFOST model is a typical crop growth model that explains crop growth based on underlying processes such as photosynthesis and respiration and their response to changing environmental conditions38. The WOFOST model estimates phenology, LAI, aboveground biomass, and storage organ biomass (i.e., grain yield) at a daily time step39 (Fig. 2).Fig. 2Schematic overview of the major processes implemented in WOFOST. The Astronomical module calculates day length, some variables relating to solar elevation, and the fraction of diffuse radiation.Full size imageZonal parameterizationWe first divided the study area covered by AMS into seamless Thiessen polygon zones. Each Thiessen polygon contains only a single AMS. These zones represent the whole areas where any location is closer to its associated AMS point than any other AMS point. Then, we assigned parameters to the entire zone based on the AMS data, including crop calendar (date of emergence) and soil water retention parameters (soil moisture content at wilting point, field capacity, and saturation). Besides, we also optimized two main crop parameters for controlling phenological development stages, namely TSUM1 (accumulated temperature required from emergence to anthesis) and TSUM2 (accumulated temperature required from anthesis to maturity), by minimizing the cost function of the observational and simulated date corresponding to anthesis and maturity.Parameter calibration within a single zoneWe implemented the calibration of parameters within every single zone, as illustrated in Fig. 3. We calculated the average statistical yield of each county within every single zone from 2007 to 2015, then ranked the counties in descending order and divided them into three groups, namely high, medium, and low-level yield counties, by the 33% quantile and 67% quantile of the average statistical yield. The three counties corresponding to 17% quantile, 50% quantile, and 83% quantile would be used for subsequent calibration and represent the corresponding three yield level groups. We used the statistical yields (converted to dry matter mass based on the standard moisture content of 12.5%) of the three counties for multiple years and a harvest index for each province to convert the county-level yield to AGB for calibration. The harvest index of each province was mainly estimated from the AMS’s dynamic growth records on the biomass composition of the dominant winter wheat varieties of the province and a published literature40. Besides, we collected the maximum LAI observations on all agrometeorological stations in all years in the study area, according to its histogram. We found that the histogram follows a normal distribution with a mean of 6.5 and a standard deviation of 1.5. Finally, we calibrated three sets of parameters corresponding to three yield level groups in each single zone according to the three selected counties.Fig. 3Flow chart of parameter calibration within a single zone.Full size imageWe designed a three-step calibration strategy for a specific yield level group. Firstly, as winter wheat varieties did not change significantly according to information recorded by agrometeorological stations from 2007 to 2015, we assumed the crop parameters of winter wheat remain unchanged every three years to combine three years of observational data to calibrate the parameters of the WOFOST model better. We maximized a log-likelihood function based on the maximum LAI statistics and every three-year county-level yield and AGB data mentioned to optimize selected crop parameters (see Table S2 in the Supplement Materials).The log-likelihood function was constructed as follows:$$log;{{rm{L}}}_{{rm{LAI}}}=-frac{1}{2}left[dlogleft(2pi right)+logleft(left|{Sigma }_{{rm{LAI}}}right|right)+{rm{MD}}{left({{bf{x}}}_{{rm{LAI}}};{mu }_{{rm{LAI}}},{Sigma }_{{rm{LAI}}}right)}^{2}right]$$
(1)
$$log;{{rm{L}}}_{{rm{TWSO}}}=-frac{1}{2}left[dlog(2pi )+logleft(left|{{boldsymbol{Sigma }}}_{{rm{TWSO}}}right|right)+{rm{MD}}{left({{bf{x}}}_{{rm{TWSO}}};{{boldsymbol{mu }}}_{{rm{TWSO}}},{{boldsymbol{Sigma }}}_{{rm{TWSO}}}right)}^{2}right]$$
(2)
$$log;{{rm{L}}}_{{rm{AGB}}}=-frac{1}{2}left[dlog(2pi )+logleft(left|{{boldsymbol{Sigma }}}_{{rm{AGB}}}right|right)+{rm{MD}}{left({{bf{x}}}_{{rm{AGB}}};{{boldsymbol{mu }}}_{{rm{AGB}}},{{boldsymbol{Sigma }}}_{{rm{AGB}}}right)}^{2}right]$$
(3)
$$log;{rm{L}}=log;{L}_{{rm{LAI}}}+log;{L}_{{rm{TWSO}}}+log;{L}_{{rm{AGB}}}$$
(4)
Where log L is the natural logarithm of the likelihood function, d is the dimension, that is, the number of years of joint calibration, which is set to 3 in this study xLAI is the vector composed of the maximum value of the 3-year LAI simulated by WOFOST, μLAI and ΣLAI are the mean vector and error covariance matrix of maximum LAI based on observation statistics. The annual maximum LAI was assumed to be independent, and the mean and standard deviation for each year was set the same as the result of Fig. 3. Similarly, xTWSO and xAGB are the yield vector and AGB vector at maturity of 3 years simulated by WOFOST, and μTWSO, μAGB are their corresponding county-level statistic vector, ΣTWSO and ΣAGB are their corresponding error covariance matrix. In this study, we assumed that the annual yield or AGB was independent, and their corresponding standard deviation was 10% of their statistical value. |Σ| is the determinant of Σ. The expression ({rm{MD}}{({bf{x}};{boldsymbol{mu }},{boldsymbol{Sigma }})}^{2}={({bf{x}}-{boldsymbol{mu }})}^{{rm{T}}}{{boldsymbol{Sigma }}}^{-1}({bf{x}}-{boldsymbol{mu }})), where MD is the Mahalanobis distance between the point x and the mean vector μ.Secondly, we optimized the inter-annual irrigation. We optimized two parameters every year: the critical value of soil moisture (denoted as SMc) and the amount of irrigation (denoted as V). When the soil moisture simulated by WOFOST is lower than SMc, an irrigation event will be triggered, and the irrigation amount is V cm. In this study, we combined three-year data for calibration with six parameters for optimization. The optimization strategy is the same as the previous step by maximizing the log-likelihood function. Finally, we fixed the optimized irrigation parameters and repeated the first step to calibrate the selected crop parameters and obtain the final optimal parameters.Data assimilationConsidering that MODIS LAI is relatively low compared to the actual LAI of winter wheat41, we select a weak-constraint cost function based on the least square of normalized observational and simulated LAI as shown in Eq. (5), which is assimilating the trend information of MODIS LAI into the crop growth model.$$J={sum }_{{rm{t}}=1}^{{rm{n}}}{left(frac{{{rm{LAI}}}_{{rm{MODIS}}}^{{rm{t}}}-{{rm{LAI}}}_{{rm{MODIS}}}^{min}}{{{rm{LAI}}}_{{rm{MODIS}}}^{max}-{{rm{LAI}}}_{{rm{MODIS}}}^{min}}-frac{{{rm{LAI}}}_{{rm{WOFOS}}}^{{rm{t}}}-{{rm{LAI}}}_{{rm{WOFOS}}}^{min}}{{{rm{LAI}}}_{{rm{WOFOS}}}^{max}-{{rm{LAI}}}_{{rm{WOFOS}}}^{min}}right)}^{2}$$
(5)
Where ({{rm{LAI}}}_{{rm{MODIS}}}^{{rm{t}}}) and .. are MODIS LAI and WOFOST simulated LAI of time t. ({{rm{LAI}}}_{{rm{MODIS}}}^{max}) and ({{rm{LAI}}}_{{rm{WOFOS}}}^{max}) are maximum of MODIS LAI and WOFOST simulated LAI. ({{rm{LAI}}}_{{rm{MODIS}}}^{min}) and ({{rm{LAI}}}_{{rm{WOFOS}}}^{min}) are minimum of MODIS LAI and WOFOST simulated LAI. J is the value of the cost function.We reinitialize the day of emergence (IDEM), the life span of leaves growing at 35 °C (SPAN), and thermal time from emergence to anthesis (TSUM1) in the WOFOST model on each 1 km winter wheat pixel according to cost function between WOFOST LAI and MODIS LAI. Besides, we applied the Subplex algorithm from the NLOPT library (https://github.com/stevengj/nlopt) for parameter optimization. More