Ecology

Whole-genome sequencing of Zhoushan cattle and Wenling cattle populationsWe collected seven individuals of Zhoushan cattle (Fig. 1a, upper panel). We also collected nine individuals of Wenling cattle (Fig. 1a, lower panel). Wenling cattle have a prominent hump on the back, dewlap, and larger ears, suggesting that its genetic background is largely B. indicus (Fig. 1a, lower panel). We performed whole-genome sequencing of these samples. To resolve their phylogenetic positions and interrelationships within domesticated cattle, we combined our data of 16 cattle individuals with publicly-available whole-genome sequencing data of five individuals from the Angus breed, a typical B. taurus in Europe, and 33 individuals from nine breeds with genetic backgrounds similar to B. indicus3, giving a total of 54 individuals (Fig. 1b, c; Table S1). We performed read trimming and aligned the trimmed reads to the UOA_Brahman_1 assembly of the cattle genome11. This assembly represents the maternal haplotype of an F1 hybrid of Brahman cattle (dam) and Angus (sire)11. After variant calling and filtering, we identified 32,970,327 single-nucleotide polymorphisms (SNPs) and 3,331,322 small indels. Based on this genomic variant information, we conducted the population genomic analyses.Figure 1Phylogenetic analysis of Zhoushan cattle and other cattle breeds. (a) Gross appearance of Zhoushan (upper panel) and Wenling cattle (lower panel). Note that Zhoushan cattle have a dark black coat colour. The arrow indicates the curving horn of Zhoushan cattle. (b) Geographic map indicating the origins of Zhoushan (green dot) and Wenling (orange dot) cattle analysed in this study. We also examined other Chinese cattle (red dots) whose genome sequencing data were available. (c) Regional map around the Zhoushan islands. Wenling, Wannan, and Guangfeng are mainland regions close to the Zhoushan islands. (d) Neighbour-joining tree of the 54 domesticated cattle. The scale bar represents pairwise distances between different individuals. The maps were constructed by R38 and R packages of maps v3.3.0 (https://cran.r-project.org/web/packages/maps) and mapdata v2.3.0 (https://cran.r-project.org/web/packages/mapdata).Full size imageGenetic relationship between Zhoushan cattle and other domesticated cattleTo reveal the phylogenetic positions and interrelationships of Zhoushan and other domesticated cattle, we performed population genomic analyses on 54 cattle individuals. First, we calculated the pairwise evolutionary distance between individuals and generated a neighbour-joining (NJ) tree to reconstruct the phylogenetic relationships between individuals of Zhoushan and other domesticated cattle (Fig. 1d). In the NJ tree, cattle clustered consistently with their geographical location (Fig. 1d). Angus individuals formed a sister group to all other individuals, including Zhoushan cattle, Wenling cattle, and other B. indicus (Fig. 1d). The individuals of Zhoushan and Wenling cattle formed monophyletic groups and were sisters to each other (Fig. 1d). The cattle in Guangfeng formed another monophyletic group and were sisters to both Zhoushan and Wenling cattle (Fig. 1d). Cattle in Wannan, Ji’an, and Leiqiong formed a single group, sister to the cattle of Zhoushan, Wenling, and Guangfeng (Fig. 1d). Zhoushan, Wenling, Guangfeng, Wannan, and Ji’an are geographically close to each other (Fig. 1b, c). The cattle of Dianzhong and Wenshan, which are in the south part of China, were distant from them (Fig. 1d). Cattle in Pakistan and India were located near the root of the phylogenetic tree (Fig. 1d). The branch lengths of Zhoushan cattle were shorter than other B. indicus cattle, suggesting the reduced genetic diversity of Zhoushan cattle (Fig. 1d).To estimate the relatedness between Zhoushan and other domesticated cattle, we performed unsupervised clustering analysis with ADMIXTURE v1.3.0 software (https://dalexander.github.io/admixture/index.html)12. At K = 2, Angus cattle were distinct from all other cattle (Fig. 2a). At K = 3, Zhoushan and Wenling cattle were newly segregated from other cattle, suggesting that these two cattle breeds are genetically close to each other (Fig. 2a). The cattle of Guangfeng, Wannan, Ji’an, Leiqiong, and Wenshan had intermediate genetic structures between Zhoushan cattle and Dianzhong cattle (Fig. 2a). At K = 4, Zhoushan cattle and Wenling cattle were separated from each other (Fig. 2a).Figure 2Admixture and principal component analysis of Zhoushan cattle and other cattle breeds. (a) Admixture plot (K = 2, 3, 4) for the 54 cattle individuals. Each individual is shown as a vertical bar divided into K colours. (b) PCA plot showing the genetic structure of the 54 cattle individuals. The degree of explained variance is given in parentheses. Colours reflect the geographic regions of sampling in Fig. 1d. The cluster composed of cattle in Wenling, Guangfeng, Wannan, Ji’an, and Leiqiong is highlighted in the black dotted ellipse. (c) Estimate of the effective population sizes of Zhoushan (green) and Wenling (orange) cattle over the past 100 generations.Full size imageTo infer the population structure of cattle individuals analysed in this study, we conducted principal component analysis (PCA). The top three principal components accounted for 21.1% of the total variance (Fig. 2b). In the first component of PCA, Angus individuals were separated from all other cattle (Fig. 2b). Additionally, cattle of Wenling, Guangfeng, Wannan, Ji’an, and Leiqiong formed a cluster (dotted ellipse in Fig. 2b). In the second component of PCA, individuals of Zhoushan cattle were separated from all other cattle (Fig. 2b). In the third principal component, Wenling cattle individuals were separated from all other cattle (Fig. 2b).We estimated the trends of the effective population size of Zhoushan and Wenling cattle over the past 100 generations (Fig. 2c). Both populations showed decreasing trends of effective population sizes (Fig. 2c). The effective population size of Zhoushan cattle was estimated to be smaller than that of Wenling cattle, suggesting the effect of island isolation on the genetic diversity of Zhoushan cattle (Fig. 2c).Detection of candidate genes associated with dark black coat colour of Zhoushan cattleTo identify putative genes associated with the dark black coat colour of Zhoushan cattle, we searched genomic regions where the same mutations were shared between Zhoushan cattle and Angus cattle. To achieve this, we calculated the average fixation index (Fst) values in 40 kb windows with 10 kb steps (Fig. 3a). We identified four peaks of Fst at chromosomes 2, 4, 8, and 18 (Fig. 3a). Among these peaks, the highest peak of Fst was identified in the region from 51.05 to 51.35 Mbp on chromosome 18 (Fig. 3a, b). This region contains 18 genes (Fig. 3c). We searched for genes that have mutations altering the amino acid sequence and have been reported to be involved in the regulation of coat colour. Among these 18 genes, only the gene of melanocyte-stimulating hormone receptor (MC1R) is known to involved in the regulation of coat colour13,14,15. Therefore, we regarded MC1R as a strong candidate gene associated with the dark black coat colour of Zhoushan and Angus cattle (Fig. 3c). This gene is located in the region between 51,094,227 bp and 51,095,177 bp on chromosome 18. MC1R is expressed in the skin melanocyte and plays a crucial role in regulating animal coat colour formation16. Mutations of MC1R have been reported to be associated with black coat colour in some animals, such as cattle17, sheep16, pigs18, reindeer19, and geese20. In the protein-coding region of MC1R, we identified one missense mutation (c.583T  > C, p.F195L) and one synonymous mutation (c.663C  > T) (Figs. 3d, 4a). The missense mutation is located in the fifth transmembrane region of MC1R (Fig. 4b). All seven Zhoushan cattle were homozygous for the missense mutation (Figs. 3d, 4a). Four of five Angus individuals were homozygous for the missense mutation, and the remaining one was heterozygous for the missense mutation (Figs. 3d, 4a). Conversely, only 19% (8/42) and 33% (14/42) of B. indicus individuals were homozygous or heterozygous, respectively, for the missense mutation (Figs. 3d, 4a). The remaining 48% (20/42) of individuals of B. indicus were homozygous for the wild-type allele (Figs. 3d, 4a). We also found that the p.F195L mutation is also present in MC1R of Black Angus (accession number: ABX83563.1) in the NCBI Protein database (Fig. S1). Furthermore, we identified 15 upstream variants and three downstream variants in the intergenic regions between neighbouring genes (Table S2).Figure 3Genomic regions associated with dark black coat colour of Zhoushan cattle. (a) Manhattan plot for average Fst values in 40 kb windows with 10 kb steps between Zhoushan cattle plus Angus and other B. indicus. A region with an average Fst of more than 0.6 is coloured in green. The arrow indicates the highest peak. The x-axis represents chromosomal positions, and the y-axis represents the average Fst values. (b) Manhattan plot on chromosome 18 for average Fst values in 40 kb windows with 10 kb steps between Zhoushan cattle, Angus, and other B. indicus. (c) Regional plot around the MC1R gene. The genotype of each individual at each variant site is shown. The genotype homozygous for the reference allele is coloured grey. Heterozygous variants are coloured blue. The homozygous genotype for alternative alleles is coloured light blue. Note that homozygous genotypes for alternative alleles are enriched in Zhoushan and Angus cattle in this region. (d) Regional plot showing the mutations around MC1R gene.Full size imageFigure 4Secondary structure of MC1R and protein sequence alignment of MC1R orthologs. (a) Regional highlight of the c.583 T  > C mutation of MC1R. The genomic region from 51,094,590 to 51,094,598 bp on chromosome 18 is shown. Note that MC1R is located on the reverse strand. (b) Secondary structure of MC1R. MC1R is a seven-transmembrane receptor. The p.F195L mutation is located in the 5th transmembrane region and enclosed by the red circle. This figure is generated by using the Protter server application39. (c) Multiple sequence alignment of MC1R orthologs. The black rectangle highlights the 195th phenylalanine residues. The red rectangle encloses the p.F195L mutation in Zhoushan cattle. The cladogram of the species is shown to the left of the species name. The cladogram topology is derived from a previous study40.Full size imageTo characterise the missense mutation of MC1R (c.583T  > C, p.F195L) found in Zhoushan and Angus cattle, we estimated the degree of evolutionary conservation of the 195th phenylalanine of MC1R. We obtained various MC1R orthologs of vertebrates from eight eutherian mammals, two marsupial mammals, four reptiles, two birds, two amphibians, one lobe-finned fish, one polypterus fish, four teleost fish, and two cartilaginous fish (Table S3). We aligned these 26 sequences with MC1R of Zhoushan cattle and B. indicus (Fig. 4c). This analysis revealed that the 195th phenylalanine of MC1R is highly conserved among vertebrates (Fig. 4c).Furthermore, we verified whether any larger structural variants are spanning the MC1R region (chr18:51,058,185–51,148,307 bp) of Zhoushan cattle and Angus. If there are large structural variants in this region for these breeds, we should see regions where the read depth distributions are different among the groups. We assessed the integrated read depth distributions of Wenling cattle (n = 9), Zhoushan cattle (n = 7) and Angus (n = 5) (Fig. 5a). The read depth distribution was very similar among the three groups suggesting that there are not large structural variants spanning the MC1R region in these breeds (Fig. 5a). We also collected the sequence reads mapped to this region, and performed BreakDancer to detect structural variants21. However, no structural variants were detected in this region in any breeds. Moreover, we compared the reference genome sequence in MC1R region of the UOA_Brahman_1 assembly and that of the UOA_Angus_1 assembly11. The UOA_Brahman_1 assembly represents the maternal haplotype of an F1 hybrid of Brahman cattle (dam) and Angus (sire), and the UOA_Angus_1 assembly represents its paternal haplotype11. The results showed that the genome sequence in the MC1R region are highly preserved between these two assemblies (Fig. 5b).Figure 5Read depth distribution, genome alignment and admixture analysis of the MC1R region. (a) Read depth distributions in the MC1R region. The left panel shows the read depth distributions in the region from 51,058,185 to 51,148,307 bp on chromosome 18. The right panel shows the read depth distributions in the region from 51,090,618 to 51,099,796 bp on chromosome 18. For each breed, the sequencing reads were integrated. The first track represents read depth distribution in each breed, and the second track represents read alignments to the reference genome. For a given base position, if the base call in the sequencing read and the corresponding base in the reference genome are different, adenine is shown in green, thymine in red, guanine in orange, and cytosine in blue. (b) Dot plots showing the genome alignments of the MC1R regions of the UOA_Angus_1 assembly (chr18:49,477,288–49,566,766 bp) and the UOA_Brahman_1 assembly (chr18:51,058,185–51,148,307 bp). The left panel shows the genome alignment by minimap2 aligner and the right one shows the genome alignment by LASTZ aligner. The region corresponding to the MC1R gene body is highlighted in red. (c) Admixture analysis of the MC1R region. The SNPs located in the MC1R region (chr18:51,058,185–51,148,307 bp) were collected and subjected to admixture analysis. The order of the samples is the same as in Fig. 2a.Full size imageFinally, we deduced the origin of the MC1R haplotype in Zhoushan cattle. We collected the SNPs located in the MC1R region (chr18:51,058,185–51,148,307 bp) from all individuals and performed admixture analysis using these SNPs. The result showed that Zhoushan cattle and Angus shared highly similar genetic components (Fig. 5c). However, the other individuals of B. indicus showed genetic components that differed from both Zhoushan cattle and Angus (Fig. 5c). These results suggest that the MC1R haplotype in Zhoushan cattle is derived from B. taurus, even though the genome of Zhoushan cattle as a whole is that of B. indicus. More

The entire world has responded to and been impacted by the COVID-19 pandemic. Humans have changed our activities and behaviors, illustrating that rapid societal change is possible. It is important to recognize that many of the root causes of this pandemic are the same as those that are worsening the global climate change and biodiversity crises. As we learn and adapt from this pandemic, opportunities for societal transformation that could change the world and the health of natural systems should not be missed. Vision is needed by our world leaders and those of influence now more than ever to rise from the pandemic years with pathways towards greater sustainability. We suggest seven strategies to maximize the COVID-19 pandemic as a pivot point for biological conservation (Fig. 1).Fig. 1: Seven strategies to maximize the COVID-19 pandemic as a pivot point for biological conservation.Societal transformation will promote a longer-term vision for both ecosystem and economic sustainability. Drawings were provided by Cerren Richards.Full size imageNew understanding gained through the pandemic can be incorporated into conservation plans moving forwards, which will take careful and insightful planning (Fig. 1(1)). This includes fine-tuning predictive models and conservation theory with greater skill and precision. For instance, confining humans to their residences at such large scales has underpinned estimates of the causal impact of reducing human activity on wildlife around the world11.Multiple disturbances and threats are increasing in frequency and intensity (e.g., pandemics, biodiversity loss, climate change). New methodologies with a multi-hazard risk perspective are required (Fig. 1(2)). We call for improvements to management models and prognostic tools to analyze and quantify vulnerabilities across ecological, social, and economic systems in future postpandemic scenarios, coupled with investments to build resilience in these diverse systems to multiple disturbances. Doing so will improve risk management before, during, and after disturbances, including those that overlap, and shift to a more preventative rather than reactive approach.Solutions need to be multisectorial and coordinated, rather than sacrificing one sector for another (Fig. 1(3)). Strategies can be designed and tested for decision-making to balance short-term gains versus investing in long-term transformations. This involves leveraging multidisciplinary knowledge, expertise, and resources toward a shared goal of producing better environmental and human well-being outcomes.Partnerships with local experts can support shared-conservation agendas to achieve both sustainable ecosystems and human well-being (Fig. 1(4)). Investing in local community experts and stewardship also has potential to build stronger local economies and long-term capacity. This requires development of the appropriate legislation and policies and adequate allocation of resources (especially funding) to support Indigenous Peoples and communities to participate and lead conservation efforts. For instance, support of local conservation efforts (e.g., expansion of Hawai’i’s Community Based Subsistence Fishing Areas) and inclusion of Indigenous management systems, are being collaboratively supported by Indigenous Peoples, local communities, governmental and non-governmental organizations, and scientists worldwide.Regions, which heavily and narrowly rely on funding from a single sector (such as international tourism) to support biodiversity conservation, are vulnerable to external shocks and require diversification. This is fundamental for economic resilience and protection against global crises such as pandemics (Fig. 1(5)). Diversification of local economies may offer viable alternatives to (over)exploitation or illegal and unregulated resource use.Strong links between environmental and human health have also come to light (“One Health”) that reinforce support of conservation programs and nature-based solutions18. This needs to be better reflected in policies, strategies, and action from global to local levels. Linking conservation of nature to human health may dampen economic drawdown and lead to strong human well-being and conservation outcomes (Fig. 1(6))Social, economic, and biological systems are intimately connected. We urge economists to engage with ecologists (and vice versa) in discussions about how ecosystem valuation can strengthen the relationship between sustainable development, nature, and society (Fig. 1(7)). More

1.Allen-Wardell, G. et al. The potential consequences of pollinator declines on the conservation of biodiversity and stability of food crop yields. Conserv. Biol. 12, 8–17 (1998).Article 

Google Scholar 
2.Wagner, D. L., Grames, E. M., Forister, M. L., Berenbaum, M. R. & Stopak, D. Insectdecline in the anthropocene: Death by a thousand cuts. Proc. Natl. Acad. Sci. 118, e2023989118. https://doi.org/10.1073/pnas.2023989118 (2021).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
3.Harrison, T. & Winfree, R. Urban drivers of plant–pollinator interactions. Funct. Ecol. 29, 879–888 (2015).Article 

Google Scholar 
4.Hall, D. M. et al. The city as a refuge for insect pollinators. Conserv. Biol. 31, 24–29 (2017).PubMed 
Article 
PubMed Central 

Google Scholar 
5.McFrederick, Q. S. & LeBuhn, G. Are urban parks refuges for bumble bees Bombus spp. (Hymenoptera: Apidae)?. Biol. Conserv. 129, 372–382 (2006).Article 

Google Scholar 
6.Wilson, C. J. & Jamieson, M. A. The effects of urbanization on bee communities dependson floral resource availability and bee functional traits. PLoS One 14, e025852. https://doi.org/10.1371/journal.pone.0225852 (2019).CAS 
Article 

Google Scholar 
7.Ives, C. D. et al. Cities are hotspots for threatened species. Glob. Ecol. Biogeogr. 25, 117–126 (2016).Article 

Google Scholar 
8.Tonietto, R., Fant, J., Ascher, J., Ellis, K. & Larkin, D. A comparison of bee communities of Chicago green roofs, parks and prairies. Landsc. Urban Plan. 103, 102–108 (2011).Article 

Google Scholar 
9.Threlfall, C. G. et al. The conservation value of urban green space habitats for Australian native bee communities. Biol. Conserv. 187, 240–248 (2015).Article 

Google Scholar 
10.Goddard, M. A., Dougill, A. J. & Benton, T. G. Scaling up from gardens: Biodiversity conservation in urban environments. Trends Ecol. Evol. 25, 90–98 (2010).PubMed 
Article 

Google Scholar 
11.Bartomeus, I. et al. Historical changes in Northeastern US bee pollinators related to shared ecological traits. Proc. Natl. Acad. Sci. U. S. A. 110, 4656–4660 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
12.Willmer, P. Pollination and Floral Ecology (Princeton University Press, Princeton, 2011).Book 

Google Scholar 
13.Danforth, B. N., Minckley, R. L. & Neff, J. L. The Solitary Bees (Princeton University Press, Princeton, 2019).Book 

Google Scholar 
14.Robertson, C. Heterotropic bees. Ecology 6, 412–436 (1925).Article 

Google Scholar 
15.Bascompte, J., Jordano, P., Melián, C. J. & Olesen, J. M. The nested assembly of plant-animal mutualistic networks. Proc. Natl. Acad. Sci. U. S. A. 100, 9383–9387 (2003).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
16.Memmott, J., Waser, N. M. & Price, M. V. Tolerance of pollination networks to species extinctions. Proc. R. Soc. B Biol. Sci. 271, 2605–2611 (2004).Article 

Google Scholar 
17.Tylianakis, J. M. & Coux, C. Tipping points in ecological networks. Trends Plant Sci. 19, 281–283 (2014).CAS 
PubMed 
Article 

Google Scholar 
18.Geslin, B., Gauzens, B., Thébault, E. & Dajoz, I. Plant pollinator networks along agradient of urbanisation. PLoS One 8, e63421. https://doi.org/10.1371/journal.pone.0063421 (2013).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
19.Baldock, K. C. R. et al. A systems approach reveals urban pollinator hotspots and conservation opportunities. Nat. Ecol. Evol. 3, 363–373 (2019).PubMed 
PubMed Central 
Article 

Google Scholar 
20.Kremen, C., M’Gonigle, L. K. & Ponisio, L. C. Pollinator community assembly tracks changes in floral resources as restored hedgerows mature in agricultural landscapes. Front. Ecol. Evol. 6, 170. https://doi.org/10.3389/fevo.2018.00170 (2018).Article 

Google Scholar 
21.Potts, S. G., Vulliamy, B., Dafni, A., Ne’eman, G. & Willmer, P. Linking bees and flowers: How do floral communities structure pollinator communities?. Ecology 84, 2628–2642 (2003).Article 

Google Scholar 
22.Cohen, H., Philpott, S. M., Liere, H., Lin, B. B. & Jha, S. The relationship between pollinator community and pollination services is mediated by floral abundance in urban landscapes. Urban Ecosyst. 24, 275–290 (2021).Article 

Google Scholar 
23.Menz, M. H. M. et al. Reconnecting plants and pollinators: Challenges in the restoration of pollination mutualisms. Trends Plant Sci. 16, 4–12 (2010).PubMed 
Article 
CAS 

Google Scholar 
24.M’Gonigle, L. K., Williams, N. M., Lonsdorf, E. & Kremen, C. A tool for selecting plants when restoring habitat for pollinators. Conserv. Lett. 10, 105–111 (2017).Article 

Google Scholar 
25.Köppler, M.-R. & Hitchmough, J. D. Ecology good, aut-ecology better; improving the sustainability of designed plantings. J. Landsc. Archit. 10, 82–91 (2015).Article 

Google Scholar 
26.Tabassum, S. et al. Using ecological knowledge for landscaping with plants in cities. Ecol. Eng. 158, 106049. https://doi.org/10.1016/j.ecoleng.2020.106049 (2020).Article 

Google Scholar 
27.Campbell, B., Khachatryan, H. & Rihn, A. Pollinator-friendly plants, reasons for and barriers to purchase. Am. Soc. Hortic. Sci. 27, 831–839 (2017).
Google Scholar 
28.Khachatryan, H. et al. Visual attention to eco-labels predicts consumer preferences for pollinator friendly plants. Sustainability 9, 1743. https://doi.org/10.3390/su9101743 (2017).Article 

Google Scholar 
29.Hitchmough, J. & Woudstra, J. The ecology of exotic herbaceous perennials grown in managed, native grassy vegetation in urban landscapes. Landsc. Urban Plan. 45, 107–121 (1999).Article 

Google Scholar 
30.Ault, J. Breeding and development of new ornamental plants from North American native taxa. Acta Hortic. 624, 37–42 (2003).Article 

Google Scholar 
31.Comba, L. et al. Garden flowers: Insect visits and the floral reward of horticulturally-modified variants. Ann. Bot. 83, 73–86 (1999).Article 

Google Scholar 
32.Garbuzov, M. & Ratnieks, F. L. W. Using the British National Collection of asters to compare the attractiveness of 228 varieties to flower-visiting insects. Environ. Entomol. 44, 638–646 (2015).PubMed 
Article 
PubMed Central 

Google Scholar 
33.Erickson, E. et al. More than meets the eye? The role of annual ornamental flowers in supporting pollinators. Environ. Entomol. 49, 178–188 (2020).CAS 
PubMed 
Article 

Google Scholar 
34.Garbuzov, M. & Ratnieks, F. L. W. W. Quantifying variation among garden plants in attractiveness to bees and other flower-visiting insects. Funct. Ecol. 28, 364–374 (2014).Article 

Google Scholar 
35.Russo, L., DeBarros, N., Yang, S., Shea, K. & Mortensen, D. Supporting crop pollinators with floral resources: Network-based phenological matching. Ecol. Evol. 3, 3125–3140 (2013).PubMed 
PubMed Central 
Article 

Google Scholar 
36.Thompson, J. D. How do visitation patterns vary among pollinators in relation to floral display and floral design in a generalist pollination system?. Oecologia 126, 386–394 (2001).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
37.Tuell, J. K., Fiedler, A. K., Landis, D. & Isaacs, R. Visitation by wild and managed bees (Hymenoptera: Apoidea) to eastern U.S. native plants for use in conservation programs. Environ. Entomol. 37, 707–718 (2008).PubMed 
Article 

Google Scholar 
38.Fowler, J. Specialist bees of the Northeast: Host plants and habitat conservation. Northeast. Nat. 23, 305–320 (2016).Article 

Google Scholar 
39.Jessica J. R. Catch the buzz-pollinator diversity, distribution, and phenology in Shenandoah National Park (Natural Resource Report. NPS/SHEN/NRR—2017/1441. National Park Service, 2017).40.Savoy-Burke, G. Woodland Bee Diversity in the Mid-Atlantic. (Master’s Thesis, University of Delaware, Newark DE, 2017).41.Fisher, R. M. Evolution and host specificity: Dichotomous invasion success of Psithyrus citrinus (Hymenoptera: Apidae), a bumblebee social parasite in colonies of its two hosts. Can. J. Zool. 63, 977–981 (1985).Article 

Google Scholar 
42.Packer, L., Genaro, J. & Sheffield, C. S. The bee genera of Eastern Canada. Can. J. Arthropod Identif. 3, 1–32 (2007).
Google Scholar 
43.Richardson, L. L., McFarland, K. P., Zahendra, S. & Hardy, S. Bumble bee (Bombus) distribution and diversity in Vermont, USA: A century of change. J. Insect Conserv. 23, 45–62 (2019).Article 

Google Scholar 
44.Domínguez-García, V. & Muñoz, M. A. Ranking species in mutualistic networks. Sci. Rep. 5, 8182. https://doi.org/10.1038/srep08182 (2015).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
45.Alarcón, R., Waser, N. M. & Ollerton, J. Year-to-year variation in the topology of a plant–pollinator interaction network. Oikos 117, 1796–1807 (2008).Article 

Google Scholar 
46.Dormann, C. F., Gruber, B. & Fruend, J. Introducing the bipartite package: Analysingecological networks. R News 8(2), 8–11 (2008).
Google Scholar 
47.Olesen, J. M., Bascompte, J., Dupont, Y. L. & Jordano, P. The modularity of pollination networks. Proc. Natl. Acad. Sci. 104, 19891–19896 (2007).ADS 
CAS 
PubMed 
PubMed Central 
MATH 
Article 

Google Scholar 
48.Wright, G. A. & Schiestl, F. P. The evolution of floral scent: The influence of olfactory learning by insect pollinators on the honest signalling of floral rewards. Funct. Ecol. 23, 841–851 (2009).Article 

Google Scholar 
49.Corbet, S. et al. Native or Exotic? Double or single? Evaluating plants for pollinator-friendly gardens. Ann. Bot. 87, 219–232 (2001).PubMed 
Article 
PubMed Central 

Google Scholar 
50.Campbell, D. R., Bischoff, M., Lord, J. M. & Robertson, A. W. Flower color influences insect visitation in alpine New Zealand. Ecology 91, 2638–2649 (2010).PubMed 
Article 

Google Scholar 
51.Harder, L. D. Morphology as a predictor of flower choice by bumble bees. Ecology 66, 198–210 (1985).Article 

Google Scholar 
52.Wilde, H. D., Gandhi, K. J. K. & Colson, G. State of the science and challenges of breeding landscape plants with ecological function. Hortic. Res. 2, 14069. https://doi.org/10.1038/hortres.2014.69 (2015).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
53.Knauer, A. C. & Schiestl, F. P. Bees use honest floral signals as indicators of reward when visiting flowers. Ecol. Lett. 18, 135–143 (2015).CAS 
PubMed 
Article 

Google Scholar 
54.Stearn, W. T. Nepeta mussinii and N. × Faassenii. J. R. Hortic. Soc. 75, 403–406 (1950).
Google Scholar 
55.Seitz, N., VanEngelsdorp, D. & Leonhardt, S. D. Are native and non-native pollinator friendly plants equally valuable for native wild bee communities?. Ecol. Evol. 10, 12838–12850 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
56.Kammerer, M., Tooker, J. F. & Grozinger, C. M. A long-term dataset on wild bee abundance in Mid-Atlantic United States. Sci. Data 7, 240. https://doi.org/10.1038/s41597-020-00577-0 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
57.Vaudo, A. D., Tooker, J. F., Grozinger, C. M. & Patch, H. M. Bee nutrition and floral resource restoration. Curr. Opin. Insect Sci. 10, 133–141 (2015).PubMed 
Article 

Google Scholar 
58.Salisbury, A. et al. Enhancing gardens as habitats for flower-visiting aerial insects (pollinators): Should we plant native or exotic species?. J. Appl. Ecol. 52, 1156–1164 (2015).CAS 
Article 

Google Scholar 
59.Mach, B. M. & Potter, D. A. Quantifying bee assemblages and attractiveness of flowering woody landscape plants for urban pollinator conservation. PLoS One 13, e0208428. https://doi.org/10.1371/journal.pone.0208428 (2018).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
60.Sponsler, D. B., Shump, D., Richardson, R. T. & Grozinger, C. M. Characterizing the floral resources of a North American metropolis using a honey bee foraging assay. Ecosphere 11, e03102. https://doi.org/10.1002/ecs2.3102 (2020).Article 

Google Scholar 
61.Rollings, R. & Goulson, D. Quantifying the attractiveness of garden flowers for pollinators. J. Insect Conserv. 23, 803–817 (2019).Article 

Google Scholar 
62.Blaauw, B. R. & Isaacs, R. Flower plantings increase wild bee abundance and the pollination services provided to a pollination-dependent crop. J. Appl. Ecol. 51, 890–898 (2014).Article 

Google Scholar 
63.Vrdoljak, S. M., Samways, M. J. & Simaika, J. P. Pollinator conservation at the local scale: Flower density, diversity and community structure increase flower visiting insect activity to mixed floral stands. J. Insect Conserv. 20, 711–721 (2016).Article 

Google Scholar 
64.Burkle, L. A. & Alarcon, R. The future of plant–pollinator diversity: Understanding interaction networks across time, space, and global change. Am. J. Bot. 98, 528–538 (2011).PubMed 
Article 

Google Scholar 
65.Roulston, T. H., Smith, S. A. & Brewster, A. L. A comparison of pan trap and intensive net sampling techniques for documenting bee (Hymenoptera: Apiformes) Fauna. J. Kansas Entomol. Soc. 80, 179–181 (2007).Article 

Google Scholar 
66.Baum, K. A. & Wallen, K. E. Potential bias in pan trapping as a function of floral abundance. J. Kansas Entomol. Soc. 84, 155–159 (2011).Article 

Google Scholar 
67.Robertson, A. W. & MacNair, M. R. The effects of floral display size on pollinator service to individual flowers of Myosotis and Mimulus. Oikos 72, 106–114 (1995).Article 

Google Scholar 
68.Bennett, A. B. & Lovell, S. Landscape and local site variables differentially influence pollinators and pollination services in urban agricultural sites. PLoS One 14, e0212034. https://doi.org/10.1371/journal.pone.0212034 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
69.Frankie, G. W. et al. Ecological patterns of bees and their host ornamental flowers in two Northern California cities. J. Kansas Entomol. Soc. 78, 227–246 (2005).Article 

Google Scholar 
70.Hamblin, A. L., Youngsteadt, E. & Frank, S. D. Wild bee abundance declines with urban warming, regardless of floral density. Urban Ecosyst. 21, 419–428 (2018).Article 

Google Scholar 
71.Wenzel, A., Grass, I., Belavadi, V. V. & Tscharntke, T. How urbanization is driving pollinator diversity and pollination—a systematic review. Biol. Conserv. 241, 108321. https://doi.org/10.1016/j.biocon.2019.108321 (2020).Article 

Google Scholar 
72.Potted herbaceous perennial plants sold. Census of Agriculture – 2014 census of horticultural specialties (USDA-NASS, 2014).73.Greenleaf, S. S., Williams, N. M., Winfree, R. & Kremen, C. Bee foraging ranges and their relationship to body size. Oecologia 153, 589–596 (2007).ADS 
PubMed 
Article 
PubMed Central 

Google Scholar 
74.Herrera, C. M. Daily patterns of pollinator activity, differential pollinating effectiveness, and floral resource availability, in a summer-flowering mediterranean shrub. Oikos 58, 277–288 (1990).Article 

Google Scholar 
75.Tuell, J. K. & Isaacs, R. Elevated pan traps to monitor bees in flowering crop canopies. Entomol. Exp. Appl. 131, 93–98 (2009).Article 

Google Scholar 
76.R Core Team. R: A language and environment for statistical computing. (R Foundation for Statistical Computing, Vienna, Austria, 2020)77.Lenth, R. emmeans: Estimated marginal means, aka least-squares means. R package version 1.5.3. (2020).78.Oksanen, J. et al. vegan: Community ecology package. R package version 2.5–7. (2020).79.Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, New York, 2016).MATH 
Book 

Google Scholar 
80.Lever, J. J., van Nes, E. H., Scheffer, M. & Bascompte, J. The sudden collapse of pollinator communities. Ecol. Lett. 17, 350–359 (2014).PubMed 
Article 

Google Scholar  More

1.Brooks, T. M. et al. Habitat loss and extinction in the hotspots of biodiversity. Conserv. Biol. 16, 909–923 (2002).Article 

Google Scholar 
2.Maschinski, J. et al. Sinking ships: Conservation options for endemic taxa threatened by sea level rise. Clim. Change 107, 147–167 (2011).ADS 
Article 

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

Google Scholar 
4.Heaney, L. R., Balete, D. S. & Rickart, E. A. Models of oceanic island biogeography: Changing perspectives on biodiversity dynamics in archipelagoes. Front. Biogeogr. 5, 249–257 (2013).Article 

Google Scholar 
5.Keppel, G., Lowe, A. J. & Possingham, H. P. Changing perspectives on the biogeography of the tropical South Pacific: Influences of dispersal, vicariance and extinction. J. Biogeogr. 36, 1035–1054 (2009).Article 

Google Scholar 
6.Laurance, W. F. Beyond Island biogeography theory. In The Theory of Island Biogeography Revisited (eds Losos, jB. & Ricklefs, R. E.) 214–237 (Princeton University Press, 2010).
Google Scholar 
7.Cheesman, L. E. Biogeographical significance of Aneityum Island, New Hebrides. Nature 180, 903–904 (1957).ADS 
Article 

Google Scholar 
8.Cox, B. T. M. & Burns, K. C. Convergent evolution of gigantism in the flora of an isolated archipelago. Evol. Ecol. 31, 741–752 (2017).Article 

Google Scholar 
9.Hamilton, A. M., Klein, E. R. & Austin, C. C. Biogeographic breaks in Vanuatu, a nascent oceanic archipelago. Pac. Sci. 64, 149–159 (2010).Article 

Google Scholar 
10.Coleman, P. J. Geology of the Solomon and New Hebrides islands, as part of the Melanesian re-entrant, Southwest Pacific. Pac. Sci. 24, 289–314 (1970).
Google Scholar 
11.Valente, L. et al. A simple dynamic model explains the diversity of island birds worldwide. Nature 579, 92–96 (2020).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
12.Keppel, G., Buckley, Y. M. & Possingham, H. P. Drivers of lowland rain forest community assembly, species diversity and forest structure on islands in the tropical South Pacific. Ecology 98, 87–95 (2010).Article 

Google Scholar 
13.Cheng, L. Insects in marine environments. Marine Insects 1, 1–4 (1976).
Google Scholar 
14.Ballantyne, L. A. & Buck, E. Taxonomy and behavior of Luciola (Luciola) aphrogeneia, a new surf firefly from Papua New Guinea. Trans. Am. Entomol. Soc. 105, 117–137 (1979).
Google Scholar 
15.Doyen, J. T. Marine beetles (Coleoptera excluding Staphylinidae). In Marine Insects (ed. Cheng, L.) 497–519 (American Elsevier, 1976).16.Topp, W. & Ring, R. A. Adaptations of Coleoptera to the marine environment. II. Observations on rove beetles (Staphylinidae) from rocky shores. Can. J. Zool. 66, 2469–2474 (1988).Article 

Google Scholar 
17.Lloyd, J. E. Fireflies (Coleoptera: Lampyridae). In Encyclopedia of Entomology 429–1452 (Springer Dordrecht, 2008).18.McDermott, F. A. Photuris bethaniensis, a new Lampyrid firefly. Proc. U. S. Natl. Mus. 103, 35–37 (1953).Article 

Google Scholar 
19.Vaz, S. et al. On the intertidal firefly genus Micronaspis Green, 1948, with a new species and a phylogeny of Cratomorphini based on adult and larval traits (Coleoptera: Lampyridae). Zool. Anz. 292, 64–91 (2021).Article 

Google Scholar 
20.Ballantyne, L. A. & Lambkin, C. Systematics of Indo-Pacific fireflies with a redefinition of Australasian Atyphella Olliff, Madagascan Photurolociola Pic, and description of seven new genera from the Luciolinae (Coleoptera: Lampyridae). Zootaxa 1997, 1–188 (2009).Article 

Google Scholar 
21.Ballantyne, L. A. et al. The Luciolinae of SE Asia and the Australopacific region: A revisionary checklist (Coleoptera: Lampyridae) including description of three new genera and 13 new species. Zootaxa 4687, 1–174 (2019).Article 

Google Scholar 
22.Saxton, N. A., Powell, G. S., Martin, G. J. & Bybee, S. M. Two new species of coastal Atyphella Ollliff (Lampyridae: Luciolinae). Zootaxa 4722, 270–276 (2020).Article 

Google Scholar 
23.Gassner, P. et al. Marine Atlas. Maximizing Benefits for Vanuatu. https://grid.cld.bz/Marine-Atlas-Maximizing-Benefits-for-Vanuatu1/10/ (2019).24.Saxton, N. A., Powell, G. S., Serrano, S. J., Monson, A. K. & Bybee, S. M. Natural history and ecological niche modelling of coastal Atyphella Olliff larvae (Lampyridae: Luciolinae) in Vanuatu. J. Nat. Hist. 53, 2271–2280 (2019).Article 

Google Scholar 
25.Rhoden, C. M., Peterman, W. E. & Taylor, C. A. Maxent-directed field surveys identify new populations of narrowly endemic habitat specialists. PeerJ 5, e3632 (2017).PubMed 
PubMed Central 
Article 

Google Scholar 
26.Phillips, S. J., Anderson, R. P. & Schapire, R. E. Maximum entropy modeling of species geographic distributions. Ecol. Model. 190, 231–259 (2006).Article 

Google Scholar 
27.Phillips, S. J., Anderson, R. P., Dudík, M., Schapire, R. E. & Blair, M. E. Opening the black box: An open-source release of Maxent. Ecography 40, 887–893 (2017).Article 

Google Scholar 
28.Warren, D. L. & Seifert, S. N. Ecological niche modeling in Maxent: The importance of model complexity and the performance of model selection criteria. Ecol. Appl. 21, 335–342 (2011).Article 

Google Scholar 
29.Stas, M. et al. An evaluation of species distribution models to estimate tree diversity at genus level in a heterogeneous urban-rural landscape. Landsc. Urban Plan. 198, 103770 (2020).Article 

Google Scholar 
30.Hernandez, P. A., Graham, C. H., Master, L. L. & Albert, D. L. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography 29, 773–785 (2006).Article 

Google Scholar 
31.Hernandez, P. A. et al. Predicting species distributions in poorly-studied landscapes. Biodivers. Conserv. 17, 1353–1366 (2008).Article 

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

Google Scholar 
33.Phillips, S. J. et al. Sample selection bias and presence-only distribution models: Implications for background and pseudo-absence data. Ecol. Appl. 19, 181–197 (2009).PubMed 
Article 
PubMed Central 

Google Scholar 
34.Silva, D. P., Aguiar, A. G. & Simião-Ferreira, J. Assessing the distribution and conservation status of a long-horned beetle with species distribution models. J. Insect Conserv. 20, 611–620 (2016).Article 

Google Scholar 
35.Cardoso, P., Erwin, T. L., Borges, P. A. & New, T. R. The seven impediments in invertebrate conservation and how to overcome them. Biol. Conserv. 144, 2647–2655 (2011).Article 

Google Scholar 
36.Hortal, J. et al. Seven shortfalls that beset large-scale knowledge of biodiversity. Annu. Rev. Ecol. Evol. Syst. 46, 523–549 (2015).Article 

Google Scholar 
37.Lomolino, M. V. Conservation biogeography. In Frontiers of Biogeography: new directions in the geography of nature (eds. Lomolino, M. V. & Heaney, L. R.) 293–296 (Sinauer Associates, Sunderland, Massachusetts, 2004).38.Whittaker, R. J. et al. Conservation biogeography: Assessment and prospect. Divers. Distrib. 11, 3–23 (2005).Article 

Google Scholar 
39.Cui, S., Luo, X., Li, C., Hu, H. & Jiang, Z. Predicting the potential distribution of white-lipped deer using the MaxEnt model. Biodivers. Sci. 26, 171 (2018).Article 

Google Scholar 
40.Moreno, R., Zamora, R., Molina, J. R., Vasquez, A. & Herrera, M. Á. Predictive modeling of microhabitats for endemic birds in South Chilean temperate forests using Maximum entropy (Maxent). Ecol. Inform. 6, 364–370 (2011).Article 

Google Scholar 
41.Raman, S., Shameer, T. T., Sanil, R., Usha, P. & Kumar, S. Protrusive influence of climate change on the ecological niche of endemic brown mongoose (Herpestes fuscus fuscus): A MaxEnt approach from Western Ghats, India. Model. Earth Syst. Environ. 6, 1795–1806 (2020).Article 

Google Scholar 
42.Abdelaal, M., Fois, M., Fenu, G. & Bacchetta, G. Using MaxEnt modeling to predict the potential distribution of the endemic plant Rosa arabica Crép, Egypt. Ecol. Inform. 50, 68–75 (2019).Article 

Google Scholar 
43.Kumar, S. & Stohlgren, T. J. Maxent modeling for predicting suitable habitat for threatened and endangered tree Canacomyrica monticola in New Caledonia. J. Ecol. Nat. 1, 094–098 (2009).
Google Scholar 
44.Yang, X. Q., Kushwaha, S. P. S., Saran, S., Xu, J. & Roy, P. S. Maxent modeling for predicting the potential distribution of medicinal plant, Justicia adhatoda L. Lesser Himalayan foothills. Ecol. Eng. 51, 83–87 (2013).CAS 
Article 

Google Scholar 
45.New, T. R. Conserving narrow range endemic insects in the face of climate change: Options for some Australian butterflies. J. Insect Conserv. 12, 585–589 (2008).Article 

Google Scholar 
46.Booth, T. H., Nix, H. A., Busby, J. R. & Hutchinson, M. F. BIOCLIM: The first species distribution modelling package, its early applications and relevance to most current MAXENT studies. Divers. Distrib. 20, 1–9 (2014).Article 

Google Scholar 
47.Hijmans, R. J., Cameron, S. & Parra, J. WorldClim, Version 1.4 (University of California, 2005).
Google Scholar 
48.Hijmans, R. J. et al. DIVA-GIS. Version, 7.5. A Geographic Information System for the Analysis of Species Distribution Data. http://www.diva-gis.org (2012).49.Phillips, S. J., Dudik, M. & Schapire, R. E. A maximum entropy approach to species distribution modeling. In Proceedings of the 21st International Conference on Machine Learning 655–662 (2004).50.Merow, C., Smith, M. J. & Silander, J. A. Jr. A practical guide to MaxEnt for modeling species’ distributions: What it does, and why inputs and settings matter. Ecography 36, 1058–1069 (2013).Article 

Google Scholar 
51.Phillips, S. J. & Dudík, M. Modeling of species distributions with Maxent: New extensions and a comprehensive evaluation. Ecography 31, 161–175 (2008).Article 

Google Scholar 
52.QGIS Development Team. QGIS Geographic Information System. Open Source Geospatial Foundation Project (2020).53.Phillips, S. J. A brief tutorial on Maxent. AT&T Res. 190, 231–259 (2005).
Google Scholar 
54.RStudio Team RStudio: Integrated Development Environment for R. RStudio, PBC. http://www.rstudio.com/ (2020).55.Zizka, A., Antonelli, A. & Silvestro, D. Sampbias: Evaluating geographic sampling bias in biological collections. Ecography 44, 25–32 (2020).Article 

Google Scholar 
56.Almeida, M. C., Cortes, L. G. & De Marco Junior, P. New records and a niche model for the distribution of two Neotropical damselflies: Schistolobos boliviensis and Tuberculobasis inversa (Odonata: Coenagrionidae). Insect Conserv. Divers. 3, 252–256 (2010).Article 

Google Scholar 
57.De Siqueira, M. F., Durigan, G., de Marco Júnior, P. & Peterson, A. T. Something from nothing: Using landscape similarity and ecological niche modeling to find rare plant species. J. Nat. Conserv. 17, 25–32 (2009).Article 

Google Scholar 
58.Wisz, M. S. et al. Effects of sample size on the performance of species distribution models. Divers. Distrib. 14, 763–773 (2008).Article 

Google Scholar 
59.McCune, J. L. Species distribution models predict rare species occurrences despite significant effects of landscape context. J. Appl. Ecol. 53, 1871–1879 (2016).Article 

Google Scholar 
60.Rinnhofer, L. J. et al. Iterative species distribution modelling and ground validation in endemism research: An Alpine jumping bristletail example. Biodiversity 21, 2845–2863 (2012).
Google Scholar 
61.Peterman, W. E., Crawford, J. A. & Kuhns, A. R. Using species distribution and occupancy modeling to guide survey efforts and assess species status. J. Nat. Conserv. 21, 114–121 (2013).Article 

Google Scholar 
62.Searcy, C. A. & Shaffer, H. B. Field validation supports novel niche modeling strategies in a cryptic endangered amphibian. Ecography 37, 983–992 (2014).Article 

Google Scholar 
63.Virzi, T., Lockwood, J. L., Lathrop, R. G., Grodsky, S. M. & Drake, D. Predicting American Oystercatcher (Haematopus palliatus) breeding distribution in an urbanized coastal ecosystem using maximum entropy modeling. Waterbirds 40, 104–122 (2017).Article 

Google Scholar 
64.Greaves, G. J., Mathieu, R. & Seddon, P. J. Predictive modelling and ground validation of the spatial distribution of the New Zealand long-tailed bat (Chalinolobus tuberculatus). Biol. Conserv. 132, 211–221 (2006).Article 

Google Scholar 
65.Raxworthy, C. J. et al. Predicting distributions of known and unknown reptile species in Madagascar. Nature 426, 837–841 (2003).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
66.Thorn, J. S., Nijman, V., Smith, D. & Nekaris, K. A. I. Ecological niche modelling as a technique for assessing threats and setting conservation priorities for Asian slow lorises (Primates: Nycticebus). Divers. Distrib. 15, 289–298 (2009).Article 

Google Scholar 
67.Faith, D. et al. Bridging the biodiversity data gaps: Recommendations to meet users’ data needs. Biodivers. Inform. 8, 41–58 (2013).Article 

Google Scholar 
68.Pyke, G. H. & Ehrlich, P. R. Biological collections and ecological/environmental research: A review, some observations and a look to the future. Biology 85, 247–266 (2010).
Google Scholar 
69.Boakes, E. H. et al. Distorted views of biodiversity: Spatial and temporal bias in species occurrence data. PLoS Biol. 8, e1000385 (2010).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
70.Kramer-Schadt, S. et al. The importance of correcting for sampling bias in MaxEnt species distribution models. Divers. Distrib. 19, 1366–1379 (2013).Article 

Google Scholar 
71.Fourcade, Y., Engler, J. O., Rödder, D. & Secondi, J. Mapping species distributions with MAXENT using a geographically biased sample of presence data: A performance assessment of methods for correcting sampling bias. PLoS ONE 9, e97122 (2014).ADS 
PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
72.National Integrated Coastal Management Framework. National Integrated Coastal Management Framework and Implementation Strategy for Vanuatu. https://extwprlegs1.fao.org/docs/pdf/van171039.pdf (2010).73.Department of Environmental and Protection and Conservation. Coastal Development. https://environment.gov.vu/images/EIA/EIA_G%20Coastal%20development.pdf (2017). More

1.Reynolds, M. P. et al. Improving global integration of crop research. Science 357, 359–360 (2017).CAS 
Article 

Google Scholar 
2.Braun, H., Atlin, G. & Payne, T. Multi-location testing as a tool to identify plant response to global climate change. in Climate Change and Crop Production (ed. Reynolds, M. P.) 115–138 (CABI, 2010).3.Ray, D. K., Ramankutty, N., Mueller, N. D., West, P. C. & Foley, J. A. Recent patterns of crop yield growth and stagnation. Nat. Commun. 3, 1293 (2012).Article 

Google Scholar 
4.Liu, B. et al. Similar estimates of temperature impacts on global wheat yield by three independent methods. Nat. Clim. Change 6, 1130–1136 (2016).Article 

Google Scholar 
5.Lobell, D. B. & Field, C. B. Global scale climate–crop yield relationships and the impacts of recent warming. Environ. Res. Lett. 2, 14–21 (2007).Article 

Google Scholar 
6.Asseng, S. et al. Rising temperatures reduce global wheat production. Nat. Clim. Change 5, 143–147 (2015).Article 

Google Scholar 
7.Crespo-Herrera, L. A. et al. Genetic yield gains in CIMMYT’s international elite spring wheat yield trials by modeling the genotype × environment interaction. Crop Sci. 57, 789–801 (2017).Article 

Google Scholar 
8.Tester, M. & Langridge, P. Breeding technologies to increase crop production in a changing world. Science 327, 818–822 (2010).CAS 
Article 

Google Scholar 
9.Rosegrant, M. W. & Cline, S. A. Global food security: challenges and polices. Science 302, 1917–1919 (2003).CAS 
Article 

Google Scholar 
10.Li, Y., Suontama, M., Burdon, R. D. & Dungey, H. S. Genotype by environment interactions in forest tree breeding: review of methodology and perspective on research and application. Tree Genet. Genomes 13, 60 (2017).Article 

Google Scholar 
11.Mishra, R. M. et al. Crossover interactions for grain yield in multienvironmental trials of winter wheat. Crop Sci. 46, 1291–1298 (2006).Article 

Google Scholar 
12.Allard, R. W. & Bradshaw, A. D. Implications of genotype–environmental interactions in applied plant breeding. Crop Sci. 4, 503–508 (1964).Article 

Google Scholar 
13.Reynolds, M. P., Hays, D. & Chapman, S. Breeding for adaptation to heat and drought stress. in Climate Change and Crop Production (ed. Reynolds, M. P.) 71–91 (CABI, 2010).14.Leon, N., Jannink, J., Edwards, J. W. & Kaeppler, S. M. Introduction to a special issue on genotype by environment interaction. Crop Sci. 56, 2081–2089 (2016).Article 

Google Scholar 
15.Reynolds, M. & Langridge, P. Physiological breeding. Curr. Opin. Plant Biol. 31, 162–171 (2016).Article 

Google Scholar 
16.Gourdji, S. M., Mathews, K. L., Reynolds, M., Crossa, J. & Lobell, D. B. An assessment of wheat yield sensitivity and breeding gains in hot environments. Proc. R. Soc. B. 2018, 20122190 (2012).
Google Scholar 
17.Pingali, P. L. Green revolution: impacts, limits, and the path ahead. Proc. Natl Acad. Sci. USA 109, 12302–12308 (2012).CAS 
Article 

Google Scholar 
18.Sharma, R. C. et al. Genetic gains for grain yield in CIMMYT spring bred wheat across international environment. Crop Sci. 52, 1522–1533 (2012).Article 

Google Scholar 
19.Boehm Jr, J. D., Ibba, M., Kiszonas, A. & Morris, C. F. End-use quality of CIMMYT-derived soft kernel durum wheat germplasm. II. Dough strength and pan bread quality. Crop Sci. 57, 1485–1498 (2017).Article 

Google Scholar 
20.Lillemo, M., van Ginkel, M., Trethowan, R. M., Hernandez, E. & Crossa, J. Differential adaptation of CIMMYT bread wheat to global high temperature environments. Crop Sci. 45, 2443–2453 (2005).Article 

Google Scholar 
21.Manes, Y. et al. Genetic yield gains of the CIMMYT international semi-arid wheat yield trials from 1994 to 2010. Crop Sci. 52, 1543–1552 (2012).Article 

Google Scholar 
22.You, L. et al. Spatial Production Allocation Model (SPAM) 2005 V3.2 International Food Policy Research Institute (IFPRI), International Institute fo Applied Systems Analysis (IIASA) (2017).23.Finlay, K. W. & Wilkinson, G. N. The analysis of adaptation in a plant-breeding programme. Aus. J. Agric. Res. 14, 742–754 (1963).Article 

Google Scholar 
24.De los Campos et al. A data-driven simulation platform to predict cultivars’ performance under uncertain weather conditions. Nat. Commun. 11, 4876 (2020).Article 

Google Scholar 
25.Lantican, M. A. et al. Impacts of International Wheat Improvement Research 1994–2014 (CIMMYT, 2016).26.Dreccer, M. F., Bonnett, D. & Lafarge, T. Plant breeding under a changing climate. in Encyclopedia of Sustainability Science and Technology (ed. Meyers, R. A.) 8013–8024 (Springer, 2012).27.Laiding, F., Drobek, T. & Meyer, U. Genotypic and environmental variability of yield for cultivars from 30 different crops in German official variety trials. Plant Breed. 127, 541–547 (2008).Article 

Google Scholar 
28.Allard, R. W. Principles of Plant Breeding 2nd edn (John Wiley & Sons, 1999).29.Kusmec, A., Srinivasan, S., Nettleton, D. & Schnable, P. S. Distinct genetic architectures for phenotype means and plasticities in Zea mays. Nat. Plants 3, 715–723 (2017).CAS 
Article 

Google Scholar 
30.Gauch, H. G. Statistical Analysis of Regional Yield Trials: AMMI Analysis of Factorial Designs (Elsevier, 1992). More

1.Lapola, D. M. et al. Proc. Natl Acad. Sci. USA 115, 11671–11679 (2018).CAS 
Article 

Google Scholar 
2.Phillips, O. L., Brienen, R. J. W. & the RAINFOR collaboration. Carbon Balance Manag. 12, 1 (2017)..3.Hubau, W. et al. Nature 579, 80–87 (2020).CAS 
Article 

Google Scholar 
4.Fleischer, K. et al. Nat. Geosci. 12, 736–741 (2019).CAS 
Article 

Google Scholar 
5.Huntingford, C. et al. Nat. Geosci. 6, 268–273 (2013).CAS 
Article 

Google Scholar 
6.Koch, A., Hubau, W. & Lewis, S. L. Earth’s Future 9, e2020EF001874 (2021).CAS 
Article 

Google Scholar 
7.Eyring, V. et al. Geosci. Model Dev. 9, 1937–1958 (2016).Article 

Google Scholar 
8.Climate Change and Land: an IPCC Special Report on Climate Change, Desertification, Land Degradation, Sustainable Land Management, Food Security, and Greenhouse Gas Fluxes in Terrestrial Ecosystems. (IPCC, 2019).9.Friend, A. D. et al. Proc. Natl Acad. Sci. USA 111, 3280–3285 (2014).CAS 
Article 

Google Scholar 
10.Pugh, T. A. M. et al. Biogeosciences https://doi.org/10.5194/bg-17-3961-2020 (2020).Article 

Google Scholar 
11.Hartmann, H., Adams, H. D., Anderegg, W. R. L., Jansen, S. & Zeppel, M. J. B. New Phytol. 205, 965–969 (2015).Article 

Google Scholar 
12.Scheiter, S., Langan, L. & Higgins, S. I. New Phytol. 198, 957–969 (2013).Article 

Google Scholar  More

Species-level dataSpecies range maps were derived from BirdLife International and NatureServe50 and the IUCN51. The threat data were from the IUCN Threats Classification Scheme (Version 3.2), which contains 11 primary threat classes and almost 50 subclasses52. In Red List assessments, assessors assign those threats that impact the species. For birds, the scope of the impact is also recorded categorically as the percentage of the species population that the threat impacts (unknown, negligible, 90%) and the severity, describing the scale of the impact on population declines: unknown, no decline, negligible declines, fluctuations, slow but significant declines (30%).Model development approachWe designed our analytical framework with three considerations in mind. First, the threat location information is limited: for each species, the data only describe whether a species is threatened by a given activity anywhere within its range (data on the timing, scope and severity of threats are available only for birds and are not spatially explicit). Second, we wanted to compare the spatial patterns of threat against independent data on spatial distributions of human activities. Third, for many activities, the relationship between human activity (for example, hunting or invasive species and diseases) and biodiversity response is poorly understood. We therefore chose not to incorporate known patterns of human activity as explanatory variables in our models.In the absence of global datasets on the spatial patterns of the impact probability of each threat, we used a simulation approach to develop our models and assess the ability of different model parameterizations to reproduce our simulated threat. This process had four steps (Extended Data Fig. 1).Simulated threat intensity mapsFirst, we simulated a continuous synthetic threat across sub-Saharan Africa. The concept behind this is that a credible model should be able to reproduce a ‘true’, synthetic threat pattern on the basis of information comparable to that available in the Red List. To test this, we generated a set of synthetic, continuous surfaces of threat intensity with different levels of spatial autocorrelation and random variation (Supplementary Fig. 1). This was achieved by taking a grid of 50 km × 50 km (2,500 km2) pixels across the Afrotropic biogeographic realm (i.e., sub-Saharan Africa). Threat intensity was modelled as a vector of random variables, Z, one for each pixel i, generated with a correlation structure given by the distance matrix between points weighted by a scalar value, r, indicating the degree of correlation (equations (1–3)). Four values of r were used: 1 × 10−6, which yields very strong autocorrelation; 1 × 10−4, which yields strong autocorrelation; 0.05, which yields moderate autocorrelation; and 0.3, which produces a low-correlation, localized pattern (Supplementary Fig. 1). The model included the following equations:$${mathbf{Z}}(r) = U^{mathrm{T}}{mathrm{Norm}}left( {n,0,1} right)$$
(1)
$$W = UU^{ast}$$
(2)
$$W = {mathrm{e}}^{left( { – rD} right)}$$
(3)
where r is a scalar determining the degree of spatial autocorrelation (as r decreases, the autocorrelation increases), D is the Euclidean distance matrix between each pair of pixels, W is the matrix of weights for the threat intensity, U and U* are the upper triangular factors of the Choleski decomposition of W and its conjugate transpose, UT is the transpose of U and n is the number of pixels.We chose the Afrotropic biogeographic realm (sub-Saharan Africa) as our geography within which to develop the modelling approach because it permitted more rapid iterations than a global-scale simulation while also retaining characteristics of importance for the model evaluation such as strong environmental gradients and heterogeneity in species richness. However, for the simulation, no information from the geography or overlapping species ranges was used, except the spatial configuration of the polygons. Thus, the use of the Afrotropic realm was purely to avoid generating thousands of complex geometries for the purpose of the simulation. Using a real geography and actual species ranges ensures that our simulation contains conditions that are observed in reality (for example, areas of high and low species richness also observed in the real world). We took the simulated threat maps generated through this process to be our ‘true’ likelihood of a randomly drawn species that occurs in that location being impacted by the synthetic threat (Supplementary Fig. 1).Simulating the red-listing processSecond, we wanted to simulate the red-listing process whereby experts evaluate whether a threat is impacting a species on the basis of the overall threat intensity within its range. For this, we used the range maps for all mammal species in Africa and assigned a binary threat classification (that is, affected or not affected) to each species on the basis of the values of the synthetic threat within each species’ range. We assumed that the binary assessment of threat for a species is based on whether the level of impact across a proportion of its range is judged as significant. This step was intended to replicate the real red-listing process, where assessors define threats that impact the species on the basis of an assessment of the information available on threatening mechanisms and species responses. In practice, this was done by overlaying the real range maps for mammals over the four simulated threat surfaces and assessing the intensity of synthetic threat within each species range map. We wanted to assign species impacts considering that species will be more likely to be impacted if a greater part of their range has a high threat intensity. Understanding how to set a threshold for what intensity would constitute sufficient threat to be assessed as affected is a complicated exercise. We thus tested three thresholds to capture different assumptions. These thresholds were chosen after discussion with leading experts on the red-listing process. More specifically, we calculated the 25th, 50th and 75th percentiles of threat intensity across pixels within the species range. We then used a stochastic test to convert these quantiles to binary threat class, C. For each species, we produced a set of ten draws from a uniform distribution bounded by 0 and 1. If over half of the draws were lower than the threat intensity quantile, the species was classified as threatened for that percentile.The above simulation assumes perfect knowledge of the threat intensities across the species range, which might not always be the case in the actual red-listing process. In real life, certain areas within species ranges are less well known for a suite of different reasons. To incorporate some uncertainty about the knowledge of the red-listing experts about the ‘true’ threat intensity, we constructed a layer to describe the spatial data uncertainty associated with the Red List. This aspect was intended to simulate the imperfect knowledge of the simulated ‘Red List assessors’. This layer was calculated as the proportion of species present in a given location that are categorized as Data Deficient—in other words, there is insufficient information known about the species to assess its extinction risk using the IUCN Red List Criteria (Extended Data Fig. 7). Then, when calculating the 25th, 50th and 75th percentiles of threat intensity across each range, we weighted this calculation by one minus the proportion of Data Deficient species, so that more uncertain places (those with a greater proportion of Data Deficient species) contributed less to the calculation than locations where knowledge was more certain. These were then converted to a binary threat class accounting for uncertainty in expert knowledge among the simulated ‘assessors’, CUncertain, using the same stochastic process described above for the calculation of C.This step produced, for each species, a threat classification analogous to the threat classification assigned by experts as part of the IUCN Red List process. Six sets of threat classifications were produced for each synthetic threat surface, on the basis of the 25th, 50th and 75th percentiles with perfect (C0.25, C0.5 and C0.75) or uncertain (CUncertain-0.25, CUncertain-0.5 and CUncertain-0.75) spatial knowledge.Model formulation and selectionThird, using all species polygons with assigned threat assessments from step 2 (that is, affected or not affected), we fitted nine candidate models and predicted the estimated probability of impact for each grid cell. Then, in a fourth step, we compared the predicted probabilities of impact produced in step 3 with the original synthetic threat maps created in step 1 to test the predictive ability of our models.The Red List threat assessment does not contain information on where in the range the impact occurs. Therefore, a species with a very small range provides higher spatial precision about the location of the impact, whereas a species with a large range may be impacted anywhere within a wide region. To address this lack of precision in the impact location, we took the area of each species range to serve as a proxy for the spatial certainty of the impact information. The certainty that a species was impacted or not impacted in a given cell depended on its range size, R. The models we evaluated therefore incorporated R in different ways (Supplementary Table 1).The models were fitted as a binomial regression with a logit link function. For each pixel, the model predicts the probability of impact, PTh—in other words, the probability that if you sampled a species at random from those that occur in that pixel, the species would be impacted by the activity being considered. To account for uncertainties in the simulation of the threat assessment process (thresholds for impact and perfect or imperfect knowledge), models were fitted to the six sets of threat codes (C0.25, C0.5, C0.75, CUncertain-0.25, CUncertain-0.5 and CUncertain-0.75), and the root mean squared error (RMSE) was calculated between PTh and the simulated threat intensity, Z(r), for each value of r. For each simulation, we ranked the different models according to their model fit as measured by the RMSE. We assessed these ranks across all simulations and sets of threat codes. We evaluated the models on the basis of the ranks of RMSE, across the threat code sets and threat intensity maps. Rank distributions for each model are shown in Extended Data Fig. 2, and the results from these models are shown in Supplementary Tables 1 and 2.All models were correlated (Pearson’s r2  > 0.5), albeit with some variation between model types and across the simulation parameters (Supplementary Fig. 2). However, some models had greater predictive accuracy when evaluated using the RMSE. The top four ranking models were, in order of decreasing summed rank, (1) inverse of cube root of range size as a weight, (2) inverse 2.5 root of range size as a weight, (3) inverse square root of range size as a weight and (4) inverse natural logarithm of range size as a weight. The fact that these four models showed good model fit suggests that the best model structure had a measure of range size as a weight but that the model was not particularly sensitive to the transformation of range size.The best-fitting model across the range of simulation parameters was an intercept-only logistic regression where the response variable was the binary threat code (1 = threatened, 0 = not threatened) for each species in the pixel and where the inverse cube root of the range size of each species was used as a weight. The model was concordant across the set of simulated datasets with a relationship that was predominantly linear with r2 between 0.47 and 0.7, depending on simulation parameters for Z(r) in 0.05, 10−4 and 10−6, centred around unity and with the RMSE ranging between 0.129 and 0.337 depending on simulation parameters (Supplementary Figs. 2 and 3). The choice of the inverse cube root range size weight was based on the performance of this against eight other model types (Supplementary Fig. 4 and Supplementary Table 1).We conducted a decomposition of variance in model performance using a binomial regression model, with RMSE as the dependent variable and model type, knowledge level and autocorrelation structure as the independent factorial variables. This showed that knowledge about the threats underlying each species range and how that threat information is used in the assessment explained the vast majority (94.7%) of the variation in RMSE outcomes (Supplementary Fig. 4).For birds, further information on the scope of the threat was available as an ordinal variable describing the fraction of range that the threat covers. We explored the use of scope in our models but concluded that to avoid arbitrary decisions about the scope of non-threatened species (where they are either not threatened anywhere or threatened in only a small part of their range), and for consistency with other taxonomic groups, we would model birds using the same model structure as used for mammals and amphibians (see the Supplementary Methods for further details).Mapping probability of impactOnce the best-performing model was identified using the simulated data, we then used this model on the actual Red List threat and range data to develop threat maps. This model produced threat maps for each taxonomic group (amphibians, birds and mammals) of the probability of impact, PTh, for each individual threat. For a given pixel, threat and taxonomic group, this estimates the probability that a randomly sampled species with a range overlapping with that pixel is being impacted by the threat, while taking into account spatial imprecision in the Red List data.Threat maps were generated using range map data and threat assessments from the IUCN Red List18. We intersected range maps for 22,898 extant terrestrial amphibians (n = 6,458), birds (n = 10,928; excluding the spatial areas within the range that are associated with ‘Passage’—where the species is known or thought very likely to occur regularly during relatively short periods of the year on migration) and mammals (n = 5,512; including those with uncertain ranges) with a global 50 km × 50 km (2,500 km2) resolution, equal-area grid for the terrestrial world. This provided, for each 50 km × 50 km pixel, a list of the species whose range overlapped it, along with the associated range size of each species. For each pixel and taxonomic group (amphibians, birds and mammals) independently, we then modelled the probability of impact, PTh,Activity (for example, PTh,Logging for logging, PTh,Agriculture for agriculture or PTh,Pollution for pollution), for each of the six threats: agriculture, hunting and trapping, logging, pollution, invasive species and diseases, and climate change. We focused on these as the six main threats as defined by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services4, but our methodological framework is flexible and could be expanded to other threats in the IUCN classification19. We used only taxonomic groups with a sufficiently high total number of species and where they have been comprehensively assessed so that potential biases associated with the groups of species prioritized by experts are avoided.Calculating uncertainties for the threat probabilityWe estimated a measure of uncertainty associated with our impact probability predictions using maps of the proportions of Data Deficient species in each cell within each taxonomic class (amphibians, birds or mammals) as a measure of knowledge certainty in that cell. The rationale for this approach is that places with more Data Deficient species with unknown threatened status should have greater uncertainty in the probability of impact. We therefore created greater variation in the data where there were more Data Deficient species. We used the knowledge-certainty map to probabilistically draw a sample of 100 threat codes for each species, on the basis of the median Data Deficiency across the species range. The random sample changed the species threat code with a probability related to the proportion of Data Deficient species within its range. If the median proportion of Data Deficient species was zero, then we assumed that there was a small probability (0.005) that the species could have been incorrectly coded. Where the median proportion was greater than zero, the probability increased linearly. So, for a species with 5% Data Deficient species within its range, the sample changed the species threat code with a probability close to 5%; if the median proportion was equal to 0.5, then the probability of the species being incorrectly assigned was equal to 0.5. We then fitted the impact probability model with each of the 100 species threat codes and generated a distribution of predicted threat probabilities in each grid cell, from which we took the 95% confidence intervals as the uncertainty estimates (Extended Data Figs. 8–10).Evaluating modelled threat patternsWe evaluated the spatial patterns of threat on the basis of the real Red List threat assessment data against empirical data in two independent ways. First, we compared the probability of impact from logging and agriculture combined within forested biomes (that is, corresponding to remotely detected forest loss, which we refer to as the probability of impact from forest loss, PTh,Forest-loss) with data on forest cover change10. Forest cover change was aggregated from their native 30 m × 30 m (900 m2) resolution pixels to our 50 km × 50 km resolution pixels using Google Earth Engine. For each 50 km × 50 km pixel, we calculated the total area lost between 2000 and 2013 and the area lost as a proportion of the area in 2000. We restricted our analysis to forested biomes: (1) tropical and subtropical moist broadleaf forests, (2) tropical and subtropical dry broadleaf forests, (3) tropical and subtropical coniferous forests, (4) temperate broadleaf and mixed forests, (5) temperate coniferous forests and (6) boreal forests/taiga, following the World Wildlife Fund’s ecoregions classification53. The relationship between forest loss and the probability of impact from forest loss as captured by agriculture and logging overall showed a significant positive correlation: PTh,Forest-loss increased with increasing forest cover loss (P  More

1.Collins, J. P., & Crump, M. L. Extinction in Our Times: Global Amphibian Decline. (2009).2.Catenazzi, A. State of the world’s amphibians. Annu. Rev. Environ. Resour. 40, 91–119 (2015).Article 

Google Scholar 
3.González-del-Pliego, P. et al. Phylogenetic and trait-based prediction of extinction risk for data deficient amphibians. Curr. Biol. 29, 1557–1563 (2019).PubMed 
Article 
CAS 
PubMed Central 

Google Scholar 
4.Lips, K. R. et al. Emerging infectious disease and the loss of biodiversity in a Neotropical amphibian community. PNAS 102, 3165–3170 (2006).ADS 
Article 
CAS 

Google Scholar 
5.Lips, K. R., Diffendorfer, J., Mendelson, J. R. & Sears, M. W. Riding the wave: Reconciling the roles of disease and climate change in amphibian declines. PLoS Biol. 6, e72 (2008).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
6.Berger, L. et al. Chytridiomycosis causes amphibian mortality associated with population declines in the rain forests of Australia and Central America. PNAS 95, 9031–9036 (1998).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
7.Longcore, J. E., Pessier, A. P. & Nichols, D. K. Batrachochytrium dendrobatidis gen. et sp. nov., a chytrid pathogenic to amphibians. Mycologia 91, 219–227 (1999).Article 

Google Scholar 
8.Berger, L. et al. History and recent progress on chytridiomycosis in amphibians. Fungal Ecol. 19, 89–99 (2016).Article 

Google Scholar 
9.Martel, A. et al. Batrachochytrium salamandrivorans sp. nov. causes lethal chytridiomycosis in amphibians. PNAS 110, 15325–15329 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
10.Scheele, B. C. et al. Amphibian fungal panzootic causes catastrophic and ongoing loss of biodiversity. Science 363, 1459–1463 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
11.Lambert, M. R. et al. Comment on “Amphibian fungal panzootic causes catastrophic and ongoing loss of biodiversity”. Science https://doi.org/10.1126/science.aay1838 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
12.Scheele, B. C. et al. Response to Comment on “Amphibian fungal panzootic causes catastrophic and ongoing loss of biodiversity”. Science https://doi.org/10.1126/science.aay2905 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
13.Puschendorf, R. et al. Distribution models for the amphibian chytrid Batrachochytrium dendrobatidis in Costa Rica: Proposing climatic refuges as a conservation tool. Divers. Distrib. 15, 401–408 (2009).Article 

Google Scholar 
14.Zumbado-Ulate, H. et al. Endemic infection of Batrachochytrium dendrobatidis in Costa Rica: Implications for amphibian conservation at regional and species level. Diversity 11, 129 (2019).Article 

Google Scholar 
15.Crawford, A. J., Lips, K. R. & Bermingham, E. Epidemic disease decimates amphibian abundance, species diversity, and evolutionary history in the highlands of central Panama. PNAS 107, 13777–13782 (2010).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
16.Woodhams, D. C. et al. Chytridiomycosis and amphibian population declines continue to spread eastward in Panama. EcoHealth 5, 268–274 (2008).PubMed 
Article 
PubMed Central 

Google Scholar 
17.Catenazzi, A., Lehr, E., Rodriguez, L. & Vredenburg, V. Batrachochytrium dendrobatidis and the collapse of anuran species richness and abundance in the upper Manu National Park, southeastern Peru. Conserv. Biol. 25, 382–391 (2011).PubMed 
Article 
PubMed Central 

Google Scholar 
18.Voyles, J. et al. Pathogenesis of chytridiomycosis, a cause of catastrophic amphibian declines. Science 326, 582–585 (2009).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
19.James, T. et al. Disentangling host, pathogen, and environmental determinants of a recently emerged wildlife disease: Lessons from the first 15 years of amphibian chytridiomycosis research. Ecol. Evol. 5, 4079–4097 (2015).PubMed 
PubMed Central 
Article 

Google Scholar 
20.Soto-Azat, C. et al. Xenopus laevis and emerging amphibian pathogens in Chile. EcoHealth 13, 775–783 (2016).PubMed 
Article 
PubMed Central 

Google Scholar 
21.Ron, S. Predicting the distribution of the amphibian pathogen Batrachochytrium dendrobatidis in the new world. Biotropica 37, 209–221 (2005).Article 

Google Scholar 
22.Rödder, D., Kielgast, J. & Lötters, S. Future potential distribution of the emerging amphibian chytrid fungus under anthropogenic climate change. Dis. Aquat. Org. 92, 201–207 (2010).Article 

Google Scholar 
23.Murray, K. A. et al. Assessing spatial patterns of disease risk to biodiversity: Implications for the management of the amphibian pathogen, Batrachochytrium dendrobatidis. J. Appl. Ecol. 48, 163–173 (2011).Article 

Google Scholar 
24.Liu, X., Rohr, J. & Li, Y. Climate, vegetation, introduced hosts and trade shape a global wildlife pandemic. Proc. Biol. Sci. 280, 20122506 (2013).PubMed 
PubMed Central 

Google Scholar 
25.Olson, D. H. et al. Mapping the global emergence of Batrachochytrium dendrobatidis, the amphibian chytrid fungus. PLoS ONE 8, e56802 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
26.Penner, J. et al. West Africa—A safe haven for frogs? A sub-continental assessment of the chytrid fungus (Batrachochytrium dendrobatidis). PLoS ONE 8, e56236 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
27.Xie, G. Y., Olson, D. H. & Blaustein, A. R. Projecting the global distribution of the emerging amphibian fungal pathogen, Batrachochytrium dendrobatidis, based on IPCC climate futures. PLoS ONE 11, e0160746 (2016).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
28.Harris, R. N. et al. Skin microbes on frogs prevent morbidity and mortality caused by a lethal skin fungus. ISME J. 3, 818–824 (2009).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
29.Searle, C. L. et al. Differential host susceptibility to Batrachochytrium dendrobatidis, an emerging amphibian pathogen. Conserv. Biol. 25, 965–974 (2011).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
30.Farrer, R. et al. Multiple emergences of genetically diverse amphibian-infecting chytrids include a globalized hypervirulent recombinant lineage. PNAS 108, 18732–18736 (2011).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
31.Lips, K. R. Overview of chytrid emergence and impacts on amphibians. Philos. Trans. R. Soc. B. 371, 20150465 (2016).Article 

Google Scholar 
32.Bolom-Huet, R., Pineda, E., Díaz-Fleischer, F., Muñoz-Alonso, A. L. & Galindo-González, J. Known and estimated distribution in Mexico of Batrachochytrium dendrobatidis, a pathogenic fungus of amphibians. Biotropica 51, 731–746 (2019).Article 

Google Scholar 
33.Zumbado-Ulate, H., García-Rodríguez, A. & Searle, C. L. Species distribution models predict the geographic expansion of an enzootic amphibian pathogen. Biotropica 53, 221–231 (2021).Article 

Google Scholar 
34.Berger, L. et al. Effect of season and temperature on mortality in amphibians due to chytridiomycosis. Aust. Vet. J. 82, 434–439 (2004).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
35.Clare, F. C. et al. Climate forcing of an emerging pathogenic fungus across a montane multi-host community. Philos. Trans. R. Soc. B. 371, 20150454 (2016).Article 
CAS 

Google Scholar 
36.Bacigalupe, L. D., Soto-Azat, C., García-Vera, C., Barría-Oyarzo, I. & Rezende, E. L. Effects of amphibian phylogeny, climate and human impact on the occurrence of the amphibian-killing chytrid fungus. Glob. Change Biol. 23, 3543–3553 (2017).ADS 
Article 

Google Scholar 
37.Raffel, T., Michel, P., Sites, W. & Rohr, J. What drives chytrid infections in newt populations? Associations with substrate, temperature, and shade. EcoHealth 7, 526–536 (2010).PubMed 
Article 
PubMed Central 

Google Scholar 
38.Pounds, J. A. et al. Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439, 161–167 (2006).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
39.Hudson, M. et al. Reservoir frogs: Seasonality of Batrachochytrium dendrobatidis infection in robber frogs. PeerJ 7, e7021 (2019).PubMed 
PubMed Central 
Article 

Google Scholar 
40.Kriger, K. M. & Hero, J. M. Large-scale seasonal variation in the prevalence and severity of chytridiomycosis. J. Zool. 271, 352–359 (2007).
Google Scholar 
41.Longo, A. V., Burrowes, P. A. & Joglar, R. L. Seasonality of Batrachochytrium dendrobatidis infection in direct-developing frogs suggests a mechanism for persistence. Dis. Aquat. Org. 92, 253–260 (2010).Article 

Google Scholar 
42.Zumbado-Ulate, H., Bolaños, F., Gutiérrez-Espeleta, G. & Puschendorf, R. Extremely low prevalence of Batrachochytrium dendrobatidis in frog populations from Neotropical dry forest of Costa Rica supports the existence of a climatic refuge from disease. EcoHealth 11, 593–602 (2014).PubMed 
Article 
PubMed Central 

Google Scholar 
43.Bacigalupe, L. D. et al. The amphibian-killing fungus in a biodiversity hotspot: Identifying and validating high-risk areas and refugia. Ecosphere. 10, e02724 (2019).Article 

Google Scholar 
44.Flechas, S. V. et al. Current and predicted distribution of the pathogenic fungus Batrachochytrium dendrobatidis in Colombia, a hotspot of amphibian biodiversity. Biotropica 49, 685–694 (2017).Article 

Google Scholar 
45.Lampo, M. et al. Batrachochytrium dendrobatidis in Venezuela. Herpetol. Rev. 39, 449 (2008).
Google Scholar 
46.Valenzuela-Sánchez, A. et al. Genomic epidemiology of the emerging pathogen Batrachochytrium dendrobatidis from native and invasive amphibian species in Chile. Transbound. Emerg. Dis. 65, 309–314 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
47.O’Hanlon, S. et al. Recent Asian origin of chytrid fungi causing global amphibian declines. Science 360, 621–627 (2018).ADS 
PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
48.Soto-Azat, C. et al. The population decline and extinction of Darwin’s frogs. PLoS ONE 8, e66957 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
49.Soto-Azat, C. et al. ASG Chile leads update of the extinction risk of Chilean amphibians for the IUCN red list of threatened speciesTM. FrogLog 23, 6–7 (2015).
Google Scholar 
50.Mora, M. et al. High abundance of invasive African clawed frog Xenopus laevis in Chile: Challenges for their control and updated invasive distribution. Manag. Biol. Invasions. 10, 377–388 (2019).Article 

Google Scholar 
51.Solís, R., Penna, M., De la Riva, I., Fisher, M. & Bosch, J. Presence of Batrachochytrium dendrobatdis in anurans from the Andes highlands of northern Chile. Herpetol. J. 24, 55–59 (2015).
Google Scholar 
52.Soto-Azat, C. et al. Is Chytridiomycosis driving Darwin’s frogs to extinction?. PLoS ONE 8, e79862 (2013).ADS 
PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
53.Valenzuela-Sánchez, A. et al. Cryptic disease-induced mortality may cause host extinction in an apparently stable host–parasite system. Proc. Biol. Sci. 284, 20171176 (2017).PubMed 
PubMed Central 

Google Scholar 
54.Lips, K. R., Reeve, J. D. & Witters, L. R. Ecological traits predicting amphibian population declines in Central America. Conserv. Biol. 17, 1078–1088 (2003).Article 

Google Scholar 
55.Hero, J. M., Williams, S. E. & Magnusson, W. E. Ecological traits of declining amphibians in upland areas of eastern Australia. J. Zool. 267(3), 221–232 (2005).Article 

Google Scholar 
56.Kriger, K. M. & Hero, J. M. Altitudinal distribution of chytrid (Batrachochytrium dendrobatidis) infection in subtropical Australian frogs. Austral. Ecol. 33(8), 1022–1032 (2008).Article 

Google Scholar 
57.Skerratt, L. F. et al. Spread of chytridiomycosis has caused the rapid global decline and extinction of frogs. EcoHealth 4, 125 (2007).Article 

Google Scholar 
58.Langwig, K. et al. Context-dependent conservation responses to emerging wildlife diseases. Front. Ecol. Environ. 13, 195–202 (2015).Article 

Google Scholar 
59.Shaw, S. D. et al. The distribution and host range of Batrachochytrium dendrobatidis in New Zealand, 1930–2010. Ecology 94, 2108–2111 (2013).Article 

Google Scholar 
60.Ghirardi, R. et al. Endangered amphibians infected with the chytrid fungus Batrachochytrium dendrobatidis in austral temperate wetlands from Argentina. Herpetol. J. 24, 129–133 (2014).
Google Scholar 
61.Bielby, J., Cooper, N., Cunningham, A., Garner, T. & Purvis, A. Predicting susceptibility to future declines in the world’s frogs. Conserv. Lett. 1, 82–90 (2008).Article 

Google Scholar 
62.Barrionuevo, S. & Mangione, S. Chytridiomycosis in two species of Telmatobius (Anura: Leptodactylidae) from Argentina. Dis. Aquat. Org. 73, 171–174 (2006).Article 

Google Scholar 
63.Seimon, T. A. et al. Upward range extension of Andean anurans and chytridiomycosis to extreme elevations in response to tropical deglaciation. Glob. Change Biol. 13, 288–299 (2007).ADS 
Article 

Google Scholar 
64.Burrowes, P. A. & De la Riva, I. Unraveling the historical prevalence of the invasive chytrid fungus in the Bolivian Andes: Implications in recent amphibian declines. Biol. Invasions. 19, 1781–1794 (2017).Article 

Google Scholar 
65.Vredenburg, V. T., Knapp, R., Tunstall, T. & Briggs, C. Dynamics of an emerging disease drive large-scale amphibian population extinctions. PNAS 107, 9689–9694 (2010).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
66.Azat, C. et al. A flagship for Austral temperate forest conservation: an action plan for Darwin’s frogs bringing together key stakeholders. Oryx 55, 356–363 (2021). Article 

Google Scholar 
67.Pilliod, D. S. et al. Effects of amphibian chytrid fungus on individual survival probability in wild boreal toads. Conserv. Biol. 24, 1259–1267 (2010).PubMed 
Article 
PubMed Central 

Google Scholar 
68.Walker, S. F. et al. Factors driving pathogenicity vs. prevalence of amphibian panzootic chytridiomycosis in Iberia. Ecol. Lett. 13, 372–382 (2010).PubMed 
Article 
PubMed Central 

Google Scholar 
69.Kriger, K. M., Pereoglou, F. & Hero, J. M. Latitudinal variation in the prevalence and intensity of chytrid (Batrachochytrium dendrobatidis) infection in eastern Australia. Conserv. Biol. 21, 1280–1290 (2007).PubMed 
Article 
PubMed Central 

Google Scholar 
70.Petersen, C. E., Lovich, R. E., Phillips, C. A., Dreslik, M. J. & Lannoo, M. J. Prevalence and seasonality of the amphibian chytrid fungus Batrachochytrium dendrobatidis along widely separated longitudes across the United States. EcoHealth 13, 368–382 (2016).PubMed 
Article 
PubMed Central 

Google Scholar 
71.Thorpe, C. J. et al. Climate structuring of Batrachochytrium dendrobatidis infection in the threatened amphibians of the northern Western Ghats, India. R. Soc. Open Sci. 5, 180211 (2018).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
72.Sonn, J. M., Utz, R. M. & Richards-Zawacki, C. L. Effects of latitudinal, seasonal, and daily temperature variations on chytrid fungal infections in a North American frog. Ecosphere 10, e02892 (2019).Article 

Google Scholar 
73.Raffel, T. R., Halstead, N. T., McMahon, T. A., Davis, A. K. & Rohr, J. R. Temperature variability and moisture synergistically interact to exacerbate an epizootic disease. Proc. Biol. Sci. 282, 20142039 (2015).PubMed 
PubMed Central 

Google Scholar 
74.Woodhams, D. C. & Alford, R. A. Ecology of chytridiomycosis in rainforest stream frog assemblages of tropical Queensland. Conserv. Biol. 19, 1449–1459 (2005).Article 

Google Scholar 
75.Adams, M. J. et al. Using occupancy models to understand the distribution of an amphibian pathogen, Batrachochytrium dendrobatidis. Ecol. Appl. 20, 289–302 (2010).PubMed 
Article 
PubMed Central 

Google Scholar 
76.Fisher, M., Garner, T. & Walker, S. F. Global emergence of Batrachochytrium dendrobatidis and amphibian chytridiomycosis in space, time, and host. Annu. Rev. Microbiol. 63, 291–310 (2009).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
77.Schloegel, L. M. et al. Novel, panzootic and hybrid genotypes of amphibia chytridiomycosis associated with the bullfrog trade. Mol. Ecol. 21, 5162–5177 (2012).PubMed 
Article 
PubMed Central 

Google Scholar 
78.Wilson, E. A., Briggs, C. J. & Dudley, T. L. Invasive African clawed frogs in California: A reservoir for or predator against the chytrid fungus?. PLoS ONE 13, e0191537 (2018).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
79.Becker, C. G., Longo, A. V., Haddad, C. F. & Zamudio, K. R. Land cover and forest connectivity alter the interactions among host, pathogen and skin microbiome. Proc. Biol. Sci. 284, 20170582 (2017).PubMed 
PubMed Central 

Google Scholar 
80.McCoy, K. A. & Peralta, A. L. Pesticides could alter amphibian skin microbiomes and the effects of Batrachochytrium dendrobatidis. Front. Microbiol. 9, 748 (2018).PubMed 
PubMed Central 
Article 

Google Scholar 
81.Ellis, E. & Ramankutty, N. Putting people in the map: Anthropogenic biomes of the world. Front. Ecol. Environ. 6, 439–447 (2008).Article 

Google Scholar 
82.Rohr, J., Halstead, N. & Raffel, T. Modelling the future distribution of the amphibian chytrid fungus: The influence of climate and human-associated factors. J. Appl. Ecol. 48, 174–176 (2011).Article 

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

Google Scholar 
84.Echeverria, C., Coomes, D., Hall, M. & Newton, A. Spatially explicit models to analyze forest loss and fragmentation between 1976 and 2020 in southern Chile. Ecol. Model. 212, 439–449 (2008).Article 

Google Scholar 
85.Rodriguez, D., Becker, C., Pupin, C., Haddad, F. & Zamudio, K. Long-term endemism of two highly divergent lineages of the amphibian-killing fungus in the Atlantic Forest of Brazil. Mol. Ecol. 23, 774–787 (2014).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
86.Puschendorf, R., Hodgson, L., Alfors, R. A., Skerrat, L. F. & VanDerWal, J. Underestimated ranges and overlooked refuges from amphibian chytridiomycosis. Divers. Distrib. 19, 1313–1321 (2013).Article 

Google Scholar 
87.Scheele, B. C. et al. After the epidemic: Ongoing declines, stabilizations and recoveries in amphibians afflicted by chytridiomycosis. Biol. Conserv. 206, 37–46 (2017).Article 

Google Scholar 
88.Mendelson, J. R. III., Whitfield, S. M. & Sredl, M. J. A recovery engine strategy for amphibian conservation in the context of disease. Biol. Conserv. 236, 188–191 (2019).Article 

Google Scholar 
89.Van Rooij, P., Martel, A., Haesebrouck, F. & Pasmans, F. Amphibian chytridiomycosis: A review with focus on fungus-host interactions. Vet. Res. 46, 1–22 (2015).Article 
CAS 

Google Scholar 
90.Christie, M. R. & Searle, C. L. Evolutionary rescue in a host–pathogen system results in coexistence not clearance. Evol. Appl. 11, 681–693 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
91.Bletz, M. C. et al. Mitigating amphibian chytridiomycosis with bioaugmentation: Characteristics of effective probiotics and strategies for their selection and use. Ecol. Lett. 16, 807–820 (2013).PubMed 
Article 
PubMed Central 

Google Scholar 
92.Bosch, J. et al. Successful elimination of a lethal wildlife infectious disease in nature. Biol. Lett. 11, 20150874 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
93.Olson, D. M. et al. Terrestrial ecoregions of the world: A new map of life on earth. Bioscience 51, 933–938 (2001).Article 

Google Scholar 
94.Pellet, J. & Schmidt, B. R. Monitoring distributions using call surveys: Estimating site occupancy, detection probabilities and inferring absence. Biol. Conserv. 123, 27–35 (2005).Article 

Google Scholar 
95.Drechsler, A. & Bock, D. Ortmann’s funnel trap—A highly efficient tool for monitoring amphibian species. Herpetol. Notes. 3, 13–21 (2010).
Google Scholar 
96.Hudson, M. et al. Dynamics and genetics of a disease-driven species decline to near extinction: Lessons for conservation. Sci. Rep. 6, 1–13 (2016).Article 
CAS 

Google Scholar 
97.R Development Core Team. R: A Language and Environment for Statistical Computing, https://www.R-project.org/ (2019).98.Sanderson, E. W. et al. The human footprint and the last of the wild. Bioscience 52, 891–904 (2002).Article 

Google Scholar 
99.Center for International Earth Science Information Network (CIESIN). Gridded Population of the World, Version 3 (GPWv3). https://doi.org/10.7927/H4639MPP (2005).100.Booth, T. H., Nix, H. A., Busby, J. R. & Hutchinson, M. F. BIOCLIM: The first species distribution modelling package, its early applications and relevance to most current MAXENT studies. Divers. Distrib. 20, 1–9 (2014).Article 

Google Scholar 
101.Center for International Earth Science Information Network (CIESIN). Gridded Species Distribution: Global Amphibian Richness Grids. https://doi.org/10.7927/H4RR1W66 (2015).102.Fick, S. & Hijmans, R. WorldClim 2: New 1-km spatial resolution climate surfaces for global land areas. Int. J. Climatol. 37, 4302–4315 (2017).Article 

Google Scholar 
103.ASTER. ASTER global digital elevation model V003. https://doi.org/10.5067/ASTER/ASTGTM (2018).104.QGIS Development Team. QGIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org (2018).105.Zuur, A. F., Ieno, E. N. & Elphick, C. S. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 1, 3–14 (2010).Article 

Google Scholar 
106.Fox, J. Effect displays in R for generalised linear models. J. Stat. Softw. 8, 1–27 (2003).Article 

Google Scholar 
107.Carpenter, T. E. Methods to investigate spatial and temporal clustering in veterinary epidemiology. Prev. Vet. Med. 48, 303–320 (2001).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
108.Kulldorff, M. A spatial scan statistic. Commun. Stat-Theor. M. 26, 1481–1496 (1997).MathSciNet 
MATH 
Article 

Google Scholar 
109.Kulldorff, M. Information Management Services, Inc. SaTScanTM v.9.4.4: software for the spatial and space-time scan statistics. http://www.satscan.org (2009). More