Philippa Kaur

Study site and field data collectionPermanent 1 m2 photoquadrats were established on Heron Reef in 1962/63, using 9 mm diameter mild steel (rebar) pegs, which were replaced over time. From the 1990’s, replacement pegs were stainless steel for greater longevity. Four sites were established, the protected (south) crest, inner flat, exposed (north) crest and exposed pools. Co-ordinates for each site are presented in Table 1, the layout shown in Fig. 2, and sites have been well described previously5,6. At each census, a 1 m2 frame divided into a 5 × 5 grid using string was placed over the pegs, and the quadrat photographed from directly above at low tide. From 1963 until 2003, a 35 mm camera and colour slide film were used. The camera was attached to a tripod affixed to the 1 m2 frame, and captured around 2/3 of the quadrat. The frame (and camera) were then rotated 180 degrees to capture the remainder of the quadrat. After 2003, a hand-held digital camera was used, with the entire quadrat being captured in a single image. Concurrent with each census, mud maps of each quadrat were hand drawn in the field, and all colonies identified in situ by someone with expertise in coral taxonomy.Table 1 Coordinates of the study sites on Heron Island Reef (WGS84).Full size tableFig. 2Quadrat layouts for each of the four sites respectively, noting that the north crest and north ridge have been treated as a single north crest site in previous publications. Underlining indicates original 1962/63 quadrats. Other quadrats were added in or after 2008, as indicated in the text. Contiguous quadrats are pictured bordering each other. Spacing between separate quadrats or groups of quadrats is not shown to scale. Note that up until 2005, NRNW was known as NR. The acronyms in each quadrat represent its name.Full size imageAt the protected (south) crest, a set of six contiguous quadrats were established in 1963 in a 2 × 3 arrangement parallel to the waterline, and about 420 m southeast of the island. This site is exposed at low tide, and was photographed once all water had drained off it. Images of quadrats A, C & E (the shoreward row) from 1963 to 2012 have been fully processed, and the data have been through QA/QC. Data after 2012 exist as images only. These quadrats form the basis of previous analyses1,4,5,6 for this site. Photographs are available for quadrats B, D & F, but apart from 2003–2010, have not been processed. In 2010, an additional two quadrats were established either side of the original six, leading to a 2 × 5 arrangement. Again, only imagery is available for these additional quadrats.At the inner flat, two pairs of contiguous quadrats were established in 1962, 44 m apart, about 70 m south of the island. This site is covered by ~10 cm of water at low tide, so could only be photographed on a still day. Imagery for this site is only available to 2012, after which the marker stakes appear to have been removed in a cleanup of the area. Images for one quadrat in each pair have been processed, but have not been subject to full QA/QC.At the exposed (north) crest main site, a set of four contiguous quadrats was established about 1100 m northeast of the island in 1963. An additional single quadrat (north ridge) was established 326 m to the east. Images from 1963 to 2012 have been fully processed, and the data have been through QA/QC. Data after 2012 exist as images only. In 2005, the single north ridge quadrat was expanded to 4 m2, and in 2008, both subsites were expanded to six quadrats in a 2 × 3 arrangement. These additional quadrats have been digitised up to 2012, but have not been through full QA/QC.The exposed pools are two individual quadrats about 5 m apart about 30 m north of the eastern (north ridge) exposed crest site. These are on the edge of a natural pool, and range from ~5–50 cm deep at low tide, and so could only be photographed on a calm day. Imagery for this site is only available until 2005, after which the marker stakes could not be relocated. Images from 1963 to 1998 have been processed, but have not been through full QA/QC.Retrieval of coral composition data from the photoquadratsProcessing of the images involved scanning the colour slides to produce digital images, and then orthorectifying each image to a 1 m2 basemap in ArcGIS (ESRI Ltd). The corners of the frame, and the holes for the string grid, were used as control points for the orthorectification. For images that originated as colour slides, each half of the quadrat was individually orthorectified to the same basemap, producing a single image of the entire quadrat (see Fig. 3). While contiguous quadrats were orthorectified individually, they were done so against a basemap containing all quadrats in the group, meaning that the resulting images can be easily merged to create a single image of the group. The outlines of all visible coral colonies ( >~1 cm2), and other benthic organisms such as algae and clams, were then digitised in ArcGIS to create a single shapefile for each quadrat for each year. Each colony was represented as an individual feature within the shapefile, and was assigned a unique colony number and species based on the mud maps drawn in the field. Colony numbers were consistent across years, allowing individual colonies to be tracked over time. If a colony underwent fission, the original colony number was retained for each, with the addition of a unique identifier after a decimal point. For example, if colony 35 split in two, the resultant colonies were identified as 35.1 and 35.2. If 35.2 later split again, the resultant colonies were identified as 35.2.1 and 35.2.2. If the colony overlapped the edge of the quadrat, only the area within the quadrat was digitised, and a flag was applied to indicate that only part of the colony was included (edgestatus = 1 in the data). Upon completion of digitisation, ArcGIS was used to calculate the area and perimeter of all colonies. While multiple census were conducted in 1963, 1971 and 1983, only a single census in each year has been processed. There are currently no plans to undertake further digitisation or QA/QC of this data set.Fig. 3Example orthorectified and stitched (prior to 2001) images from the NCNE quadrat, showing the effects of a cyclone that removed all colonies in 1972, and slow recovery over subsequent decades.Full size image More

Keith, D. A. et al. Nature https://doi.org/10.1038/s41586-022-05318-4 (2022).Article 

Google Scholar 
Domesday Book, or, The Great Survey of England of William the Conqueror A.D. MLXXXVI (Ordnance Survey Office, 1862).McMahon, G. et al. Environ. Manage. 28, 293–316 (2001).PubMed 
Article 

Google Scholar 
Spalding, M. D. et al. BioScience 57, 573–583 (2007).Article 

Google Scholar 
Holdridge, L. R. Science 105, 367–368 (1947).PubMed 
Article 

Google Scholar 
Köppen, W. in Handbuch der Klimatologie (eds Köppen, W. & Geiger, G. C.) 1–44 (Gebrüder Borntraeger, 1936).
Google Scholar 
Whittaker, R. H. Communities and Ecosystems (Macmillan, 1975).
Google Scholar 
Keddy, P. A. Trends Ecol. Evol. 9, 231–234 (1994).PubMed 
Article 

Google Scholar 
United Nations. Convention on Biological Diversity (UN, 1992).
Google Scholar 
MacArthur, R. H. Geographical Ecology: Patterns in the Distribution of Species (Princeton Univ. Press, 1972).
Google Scholar 
Schoener, T. W. in Community Ecology (eds Diamond, J. D. & Case, T. ) 467–479 (Harper & Row, 1986).
Google Scholar 
Winemiller, K. O., Fitzgerald, D. B., Bower, L. M. & Pianka, E. R. Ecol. Lett. 18, 737–751 (2015).PubMed 
Article 

Google Scholar  More

Bardgett, R. D. & van der Putten, W. H. Belowground biodiversity and ecosystem functioning. Nature 515, 505–511 (2014).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Guerra, C. A. et al. Tracking, targeting, and conserving soil biodiversity. Science 371, 239–241 (2021).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Wall, D. H. et al. (eds) Soil Ecology and Ecosystem Services (Oxford University Press, 2012).Jansson, J. K. & Hofmockel, K. S. Soil microbiomes and climate change. Nat. Rev. Microbiol. 18, 35–46 (2020).CAS 
PubMed 
Article 

Google Scholar 
de Vries, F. T. et al. Soil food web properties explain ecosystem services across European land use systems. Proc. Natl Acad. Sci. USA 110, 14296–14301 (2013).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Adhikari, K. & Hartemink, A. E. Linking soils to ecosystem services—a global review. Geoderma 262, 101–111 (2016).ADS 
CAS 
Article 

Google Scholar 
Pereira, P., Bogunovic, I., Muñoz-Rojas, M. & Brevik, E. C. Soil ecosystem services, sustainability, valuation and management. Curr. Opin. Environ. Sci. Health 5, 7–13 (2018).Article 

Google Scholar 
Wall, D. H., Nielsen, U. N. & Six, J. Soil biodiversity and human health. Nature 528, 69–76 (2015).Delgado-Baquerizo, M. et al. The proportion of soil-borne pathogens increases with warming at the global scale. Nat. Clim. Chang. 10, 550–554 (2020).ADS 
Article 

Google Scholar 
Rillig, M. C. et al. The role of multiple global change factors in driving soil functions and microbial biodiversity. Science 366, 886–890 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Guerra, C. A. et al. Global vulnerability of soil ecosystems to erosion. Landsc. Ecol. 35, 823–842 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
Geisen, S., Wall, D. H. & van der Putten, W. H. Challenges and opportunities for soil biodiversity in the Anthropocene. Curr. Biol. 29, R1036–R1044 (2019).CAS 
PubMed 
Article 

Google Scholar 
Jung, M. et al. Areas of global importance for conserving terrestrial biodiversity, carbon and water. Nat. Ecol. Evol. 5, 1499–1509 (2021).PubMed 
Article 

Google Scholar 
Xu, H. et al. Ensuring effective implementation of the post-2020 global biodiversity targets. Nat. Ecol. Evol. 5, 411–418 (2021).PubMed 
Article 

Google Scholar 
Díaz, S. et al. (eds). Summary for Policymakers of the Global Assessment Report on Biodiversity and Ecosystem Services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES, 2019); https://zenodo.org/record/3553579#.YyhIsXbMK70Phillips, H. R. P. et al. Global distribution of earthworm diversity. Science 366, 480–485 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
van den Hoogen, J. et al. Soil nematode abundance and functional group composition at a global scale. Nature 572, 194–198 (2019).ADS 
PubMed 
Article 

Google Scholar 
Delgado-baquerizo, M. et al. A global atlas of the dominant bacteria found in soil. Science 325, 320–325 (2018).ADS 
Article 

Google Scholar 
Tedersoo, L. et al. Global diversity and geography of soil fungi. Science 346, 1256688 (2014).PubMed 
Article 

Google Scholar 
Xu, X., Thornton, P. E. & Post, W. M. A global analysis of soil microbial biomass carbon, nitrogen and phosphorus in terrestrial ecosystems: global soil microbial biomass C, N and P. Glob. Ecol. Biogeogr. 22, 737–749 (2013).Article 

Google Scholar 
Djukic, I. et al. Early stage litter decomposition across biomes. Sci. Total Environ. 628–629, 1369–1394 (2018).Guerra, C. A. et al. Global projections of the soil microbiome in the Anthropocene. Glob. Ecol. Biogeogr. 30, 987–999 (2021).PubMed 
Article 

Google Scholar 
Cameron, E. K. et al. Global mismatches in aboveground and belowground biodiversity. Conserv. Biol. 33, 1187–1192 (2019).PubMed 
Article 

Google Scholar 
El Moujahid, L. et al. Effect of plant diversity on the diversity of soil organic compounds. PLoS One 12, e0170494 (2017).PubMed 
PubMed Central 
Article 

Google Scholar 
Guerra, C. A. et al. Blind spots in global soil biodiversity and ecosystem function research. Nat. Commun. 11, 3870 (2020).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Fierer, N. & Jackson, R. B. The diversity and biogeography of soil bacterial communities. Proc. Natl Acad. Sci. USA 103, 626–631 (2006).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Tedersoo, L. et al. Regional-scale in-depth analysis of soil fungal diversity reveals strong pH and plant species effects in Northern Europe. Front. Microbiol. 11, 1953 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
Popp, A. et al. Land-use futures in the shared socio-economic pathways. Glob. Environ. Change 42, 331–345 (2017).Article 

Google Scholar 
Dornelas, M. et al. Assemblage time series reveal biodiversity change but not systematic loss. Science 344, 296–299 (2014).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Egoh, B., Reyers, B., Rouget, M., Bode, M. & Richardson, D. M. Spatial congruence between biodiversity and ecosystem services in South Africa. Biol. Conserv. 142, 553–562 (2009).Article 

Google Scholar 
Jürgens, N. et al. The BIOTA Biodiversity Observatories in Africa—a standardized framework for large-scale environmental monitoring. Environ. Monit. Assess. 184, 655–678 (2012).PubMed 
Article 

Google Scholar 
Wyborn, C. & Evans, M. C. Conservation needs to break free from global priority mapping. Nat. Ecol. Evol. 5, 1322–1324 (2021).PubMed 
Article 

Google Scholar 
Hautier, Y. et al. Local loss and spatial homogenization of plant diversity reduce ecosystem multifunctionality. Nat. Ecol. Evol. 2, 50–56 (2018).PubMed 
Article 

Google Scholar 
Zhou, Z., Wang, C. & Luo, Y. Meta-analysis of the impacts of global change factors on soil microbial diversity and functionality. Nat. Commun. 11, 3072 (2020).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Eisenhauer, N., Schulz, W., Scheu, S. & Jousset, A. Niche dimensionality links biodiversity and invasibility of microbial communities. Funct. Ecol. 27, 282–288 (2013).Article 

Google Scholar 
Wagg, C., Bender, S. F., Widmer, F. & van der Heijden, M. G. A. Soil biodiversity and soil community composition determine ecosystem multifunctionality. Proc. Natl Acad. Sci. USA 111, 5266–5270 (2014).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Haines-Young, R. H. & Potschin, M. B. in Ecosystems Ecology: A New Synthesis (eds Raffaelli, D. G. & Frid, C. L. J.) Ch. 6 (2012).Smith, L. C. et al. Large‐scale drivers of relationships between soil microbial properties and organic carbon across Europe. Glob. Ecol. Biogeogr. 30, 2070–2083 (2021).Article 

Google Scholar 
Keesstra, S. et al. The superior effect of nature based solutions in land management for enhancing ecosystem services. Sci. Total Environ. 610-611, 997–1009 (2018).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Le Provost, G. et al. Contrasting responses of above- and belowground diversity to multiple components of land-use intensity. Nat. Commun. 12, 3918 (2021).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Tanneberger, F. et al. The power of nature‐based solutions: how peatlands can help us to achieve key EU sustainability objectives. Adv. Sustain. Syst. 5, 2000146 (2021).CAS 
Article 

Google Scholar 
Johnston, A. et al. Observed and predicted effects of climate change on species abundance in protected areas. Nat. Clim. Chang. 3, 1055–1061 (2013).ADS 
Article 

Google Scholar 
Hannah, L. et al. Protected area needs in a changing climate. Front. Ecol. Environ. 5, 131–138 (2007).Article 

Google Scholar 
Gallardo, B. et al. Protected areas offer refuge from invasive species spreading under climate change. Glob. Chang. Biol. 23, 5331–5343 (2017).ADS 
PubMed 
Article 

Google Scholar 
O’Neill, B. C. et al. The roads ahead: narratives for shared socioeconomic pathways describing world futures in the 21st century. Glob. Environ. Change 42, 169–180 (2017).Article 

Google Scholar 
Fedele, G., Donatti, C. I., Bornacelly, I. & Hole, D. G. Nature-dependent people: mapping human direct use of nature for basic needs across the tropics. Glob. Environ. Change 71, 102368 (2021).Visconti, P. et al. Protected area targets post-2020. Science 364, 239–241 (2019).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Allan, J. R. et al. The minimum land area requiring conservation attention to safeguard biodiversity. Science 376, 1094–1101 (2022).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Maestre, F. T. et al. Plant species richness and ecosystem multifunctionality in global drylands. Science 335, 214–218 (2012).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Delgado-Baquerizo, M. et al. Changes in belowground biodiversity during ecosystem development. Proc. Natl Acad. Sci. USA. 116, 6891–6896 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Mace, G. M. Whose conservation? Science 345, 1558–1560 (2014).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Amaral-Zettler, L. A., McCliment, E. A., Ducklow, H. W. & Huse, S. M. A method for studying protistan diversity using massively parallel sequencing of V9 hypervariable regions of small-subunit ribosomal RNA genes. PLoS One 4, e6372 (2009).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Stoeck, T. et al. Multiple marker parallel tag environmental DNA sequencing reveals a highly complex eukaryotic community in marine anoxic water. Mol. Ecol. 19, 21–31 (2010).CAS 
PubMed 
Article 

Google Scholar 
Ramirez, K. S. et al. Biogeographic patterns in below-ground diversity in New York City’s Central Park are similar to those observed globally. Proc. Biol. Sci. 281, 20141988 (2014).PubMed 
PubMed Central 

Google Scholar 
Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7, 335–336 (2010).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Edgar, R. C. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460–2461 (2010).CAS 
PubMed 
Article 

Google Scholar 
Edgar, R. C. & Flyvbjerg, H. Error filtering, pair assembly and error correction for next-generation sequencing reads. Bioinformatics 31, 3476–3482 (2015).CAS 
PubMed 
Article 

Google Scholar 
Edgar, R. C. UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. Preprint at bioRxiv https://doi.org/10.1101/081257 (2016).Tedersoo, L. et al. Towards understanding diversity, endemicity and global change vulnerability of soil fungi. Preprint at bioRxiv https://doi.org/10.1101/2022.03.17.484796 (2022).Quast, C. et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 41, D590–D596 (2013).CAS 
PubMed 
Article 

Google Scholar 
Delgado-Baquerizo, M. et al. Global homogenization of the structure and function in the soil microbiome of urban greenspaces. Sci. Adv. 7, eabg5809 (2021).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Phillips, H. R. P., Heintz-Buschart, A. & Eisenhauer, N. Putting soil invertebrate diversity on the map. Mol. Ecol. 29, 655–657 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
Xiong, W. et al. A global overview of the trophic structure within microbiomes across ecosystems. Environ. Int. 151, 106438 (2021).PubMed 
Article 

Google Scholar 
Drummond, A. J. et al. Evaluating a multigene environmental DNA approach for biodiversity assessment. Gigascience 4, 46 (2015).PubMed 
PubMed Central 
Article 

Google Scholar 
Oliverio, A. M., Gan, H., Wickings, K. & Fierer, N. A DNA metabarcoding approach to characterize soil arthropod communities. Soil Biol. Biochem. 125, 37–43 (2018).CAS 
Article 

Google Scholar 
Horton, D. J., Kershner, M. W. & Blackwood, C. B. Suitability of PCR primers for characterizing invertebrate communities from soil and leaf litter targeting metazoan 18S ribosomal or cytochrome oxidase I (COI) genes. Eur. J. Soil Biol. 80, 43–48 (2017).CAS 
Article 

Google Scholar 
Delgado-Baquerizo, M. et al. Multiple elements of soil biodiversity drive ecosystem functions across biomes. Nat. Ecol. Evol. 4, 210–220 (2020).PubMed 
Article 

Google Scholar 
Carter, M. R. & Gregorich, E. G. (eds) Soil Sampling and Methods of Analysis (CRC Press, 2007).Sparks, D. L. et al. (eds) Methods of Soil Analysis, Part 3: Chemical Methods (Wiley, 2020).Nguyen, N. H. et al. FUNGuild: an open annotation tool for parsing fungal community datasets by ecological guild. Fungal Ecol. 20, 241–248 (2016).Article 

Google Scholar 
Bell, C. W. et al. High-throughput fluorometric measurement of potential soil extracellular enzyme activities. J. Vis. Exp. 81, e50961 (2013).Wang, L. et al. Diversifying livestock promotes multidiversity and multifunctionality in managed grasslands. Proc. Natl Acad. Sci. USA. 116, 6187–6192 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Durán, J., Delgado-Baquerizo, M., Rodríguez, A., Covelo, F. & Gallardo, A. Ionic exchange membranes (IEMs): a good indicator of soil inorganic N production. Soil Biol. Biochem. 57, 964–968 (2013).Article 

Google Scholar 
Breiman, L. Random forests. Mach. Learn. 45, 5–32 (2001).MATH 
Article 

Google Scholar 
Friedman, J. H. Greedy function approximation: a gradient boosting machine. Ann. Stat. 29, 1189–1232 (2001).MathSciNet 
MATH 
Article 

Google Scholar 
Sharma, N. XGBoost. The Extreme Gradient Boosting for Mining Applications (GRIN Verlag, 2018).Chen, T. & Guestrin, C. XGBoost: a scalable tree boosting system. In Proc. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785–794 (Association for Computing Machinery, 2016).Wilson. ParBayesianOptimization: Parallel Bayesian Optimization of Hyperparameters. R version 1 https://CRAN.R-project.org/package=ParBayesianOptimization (2021).Hastie, T., Friedman, J. & Tibshirani, R. The Elements of Statistical Learning (Springer, 2001).Jackson, D. A. & Chen, Y. Robust principal component analysis and outlier detection with ecological data. Environmetrics 15, 129–139 (2004).Article 

Google Scholar 
Breiman, L. Bagging predictors. Mach. Learn. 24, 123–140 (1996).MATH 
Article 

Google Scholar 
Breiman, L., Friedman, J., Stone, C. J. & Olshen, R. A. Classification and Regression Trees (Routledge, 1984).Ord, J. K. & Getis, A. Local spatial autocorrelation statistics: distributional issues and an application. Geogr. Anal. 27, 286–306 (2010).Article 

Google Scholar 
Getis, A. & Ord, J. K. The analysis of spatial association by use of distance statistics. Geogr. Anal. 24, 189–206 (2010).Article 

Google Scholar 
Prasannakumar, V., Vijith, H., Charutha, R. & Geetha, N. Spatio-temporal clustering of road accidents: GIS based analysis and assessment. Procedia Soc. Behav. Sci. 21, 317–325 (2011).Article 

Google Scholar 
Lin, G. Comparing spatial clustering tests based on rare to common spatial events. Comput. Environ. Urban Syst. 28, 691–699 (2004).Article 

Google Scholar 
Araújo, M. B. et al. Standards for distribution models in biodiversity assessments. Sci. Adv. 5, eaat4858 (2019).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Rousseeuw, P. J. & van Zomeren, B. C. Unmasking multivariate outliers and leverage points. J. Am. Stat. Assoc. 85, 633–639 (1990).Article 

Google Scholar 
Hempel, S., Frieler, K., Warszawski, L., Schewe, J. & Piontek, F. A trend-preserving bias correction—the ISI-MIP approach. Earth Syst. Dyn. 4, 219–236 (2013).ADS 
Article 

Google Scholar 
Lawrence, D. M. et al. The Land Use Model Intercomparison Project (LUMIP) contribution to CMIP6: rationale and experimental design. Geosci. Model Dev. 9, 2973–2998 (2016).ADS 
Article 

Google Scholar 
Kim, H. et al. A protocol for an intercomparison of biodiversity and ecosystem services models using harmonized land-use and climate scenarios. Geosci. Model Dev. 11, 4537–4562 (2018).Dufresne, J.-L. et al. Climate change projections using the IPSL-CM5 Earth System Model: from CMIP3 to CMIP5. Clim. Dyn. 40, 2123–2165 (2013).Article 

Google Scholar 
Hurtt, G. C. et al. Harmonization of land-use scenarios for the period 1500–2100: 600 years of global gridded annual land-use transitions, wood harvest, and resulting secondary lands. Clim. Change 109, 117 (2011).ADS 
Article 

Google Scholar 
Hurtt, G. C. et al. Harmonization of global land use change and management for the period 850–2100 (LUH2) for CMIP6. Geosci. Model Dev. 13, 5425–5464 (2020).ADS 
CAS 
Article 

Google Scholar 
Riahi, K. et al. The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: an overview. Glob. Environ. Change 42, 153–168 (2017).Article 

Google Scholar 
O’Neill, B. C. et al. A new scenario framework for climate change research: the concept of shared socioeconomic pathways. Clim. Change 122, 387–400 (2014).ADS 
Article 

Google Scholar 
Newbold, T. et al. Global effects of land use on local terrestrial biodiversity. Nature 520, 45–50 (2015).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Powers, R. P. & Jetz, W. Global habitat loss and extinction risk of terrestrial vertebrates under future land-use-change scenarios. Nat. Clim. Chang. 9, 323–329 (2019).ADS 
Article 

Google Scholar  More

Barnosky, A. D. et al. Has the Earth’s sixth mass extinction already arrived? Nature 471, 51–57. https://doi.org/10.1038/nature09678 (2011).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Ceballos, G. et al. Accelerated human-induced species losses: Entering the sixth mass extinction. Sci. Adv. 1, 5. https://doi.org/10.1126/sciadv.1400253 (2015).Article 

Google Scholar 
Purvis, A. et al. IPBES global assessment on biodiversity and ecosystem services chapter 2.2 status and trends. Nature https://doi.org/10.5281/zenodo.5517457.svg (2019).Balvernara, P. et al. IPBES global assessment on biodiversity and ecosystem services chapter 2.2 status and trends. Drivers. Change https://doi.org/10.5281/zenodo.5517423 (2019).Carrol, C. & Noss, R. F. Rewilding in the face of climate change. Conserv. Biol. 35, 155–167. https://doi.org/10.1111/cobi.13531 (2020).Article 

Google Scholar 
Barr, S. L., Larson, B. M. H., Beechey, T. J. & Scott, D. J. Assessing climate change adaptation progress in Canada’s protected areas. Can. Geog. 65, 152–165. https://doi.org/10.1111/cag.12635 (2020).Article 

Google Scholar 
Convention on Biological Diversity. Aichi Target 11, Convention on Biological Diversity. https://www.cbd.int/aichi-targets/target/11. Accessed 14 May 2021.United Nations. Climate Change Pathways. https://unfccc.int/climate-action/marrakech-partnership/reporting-and-tracking/climate_action_pathways. Accessed 12 Sept 2022.Government of Canada. Canada’s nature legacy: Protecting our nature conservation/nature-legacy.html (2021).Coristine, L. E. et al. Informing Canada’s commitment to biodiversity conservation: A science-based framework to help guide protected areas designation through Target 1 and beyond. Facets 3, 531–562. https://doi.org/10.1139/facets-2017-0102 (2017).Article 

Google Scholar 
De Barros, A. E. et al. Identification of areas in Brazil that optimize areas that optimize conservation of forest carbon, Jaguars and Biodiversity. Conserv. Biol. 28, 580–593. https://doi.org/10.1111/cobi.12202 (2013).Article 
PubMed 

Google Scholar 
Jantz, P., Scott, S. & Laporte, N. Carbon stock corridors to mitigate climate change and promote biodiversity in the tropics. Nat. Clim. Change 4, 138–142. https://doi.org/10.1038/nclimate2105 (2014).ADS 
CAS 
Article 

Google Scholar 
Beaudrot, L. et al. Limited carbon and biodiversity co-benefits for tropical mammals and birds. Ecol. Appl. 26, 10998–11111. https://doi.org/10.1890/15-0935 (2016).Article 

Google Scholar 
Morelli, T. L. et al. Climate-change refugia: Biodiversity in a slow lane. Front. Ecol. Environ. 18, 228–234. https://doi.org/10.1002/fee.2189 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Stralberg, et al. Macrorefugia for North American trees ad songbirds: Climatic limiting factors and multi-scale topographic influences. Glob. Ecol. Biogeogr. 27, 690–703. https://doi.org/10.1111/geb.12731 (2018).Article 

Google Scholar 
Caroll, C. & Ray, J. C. Maximizing the effectiveness of national commitments to protected area expansion for conserving biodiversity and ecosystem carbon under climate change. Glob. Chang Biol. 27, 3395–3414. https://doi.org/10.1111/gcb.15645 (2020).Article 

Google Scholar 
Bradshaw, C. J., Warkentin, I. G. & Sodhi, N. S. Urgent preservation of boreal carbon stocks and biodiversity. Trends Ecol. Evol. 24, 541–548. https://doi.org/10.1016/j.tree.2009.03.019 (2009).Article 
PubMed 

Google Scholar 
Harris, L. I. et al. The essential carbon service provided by northern peatlands. Front. Ecol. Environ. 20, 222–230 (2022).Article 

Google Scholar 
Environment and Climate Change Canada. Canadian Environmental Sustainability Indicators: Canada’s conserved areas. environmental-indicators/conserved-areas.html (2020).Office of the Auditor General of Canada. Lessen learnt from 30 years of climate change challenges and opportunities. https://www.oag-bvg.gc.ca/internet/English/att__e_43948.html#hd3l (2020).Shea, T. et al. Canada’s Conservation Vision: A report of the National Advisory Panel. Government of Canada, 43 pp (2018).Environment and Climate Change Canada. Pan-Canadian Approach to transforming species at risk conservation in Canada. species-at-risk-conservation.html (2018).Bergerund, A. T. Caribou, wolves and man. Trends Ecol. Evol. 3, 68–72. https://doi.org/10.1016/0169-5347(88)90019-5 (1988).Article 

Google Scholar 
Vernier, L. A. et al. Effects of natural resource development on the terrestrial biodiversity of Canadian boreal forests. Environ. Rev. 22, 457–490. https://doi.org/10.1139/er-2013-0075 (2014).Article 

Google Scholar 
Wells, J. V., Dawson, N., Culver, N., Reid, F. A. & Slegers, S. M. The state of conservation in North America’s Borel Forest: Issues and opportunities. Front. For. Glob. Change 3, 90. https://doi.org/10.3389/ffgc.2020.00090/full (2020).Article 

Google Scholar 
COSEWIC. COSEWIC assessment and update status report on the woodland caribou Rangifer tarandus caribou in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. xi + 98 pp. (2002).COSEWIC. COSEWIC assessment and status report on the caribou Rangifer tarandus, Newfoundland population, Atlantic-Gaspésie population and Boreal population, in Canada. Committee on the Status of Endangered Wildlifein Canada. Ottawa. xxiii + 128 pp. (2014).Environment and Climate Change Canada. Amended Recovery Strategy for the Woodland Caribou (Rangifer tarandus caribou), Boreal Population, in Canada. Species at Risk Act Recovery Strategy Series. Environment and Climate Change Canada, Ottawa. xiii + 143pp. (2020).Environment and Climate Change Canada. Report on the Progress of Recovery Strategy Implementation for the Woodland Caribou (Rangifer tarandus caribou), Boreal population in Canada for the Period 2012–2017. Species at Risk Act Recovery Strategy Series. Environment and Climate Change Canada, Ottawa. ix + 94 (2017).Hebblewhite, M. Billion dollar boreal woodland caribou and the biodiversity impacts of the global oil and gas industry. Biol. Conserv. 206, 102–111. https://doi.org/10.1016/j.biocon.2016 (2017).Article 

Google Scholar 
Fortin, D., McLoughlin, P. D. & Hebblewhite, M. When the protection of a threatened species depends on the economy of a foreign nation. PLoS ONE 15, e0229555. https://doi.org/10.1371/journal.pone.0229555 (2020).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Drever, R. C. et al. Conservation through co-occurrence: Woodland caribou as a focal species for boreal biodiversity. Biol. Conserv. 232, 238–252. https://doi.org/10.1016/j.biocon.2019.01.026 (2019).Article 

Google Scholar 
Government of Canada. Pan-Canadian Framework on clean growth and climate change climatechange/pan-canadian-framework.html.Bradshaw, C. J. & Warkentin, I. G. Global estimates of boreal forest carbon stocks and flux. Glob. Planet Chang 128, 24–30. https://doi.org/10.1016/j.gloplacha.2015.02.004 (2015).ADS 
Article 

Google Scholar 
Jennings, M. D. Gap analysis: Concept, methods, recent results. Land Ecol. 5, 15–20 (2010).
Google Scholar 
Environment and Climate Change Canada. Canadian Protected and Conserved Areas database. national-wildlife-areas/protected-conserved-areas-database (2019).DeLuca, T. H. & Boisvenue, C. Boreal forest soil carbon: Distribution function and modelling. Forestry 85, 161–184. https://doi.org/10.1093/forestry/cps003 (2012).Article 

Google Scholar 
Price, et al. Anticipating the consequences of climate change for Canada’s boreal forest ecosystems. Environ. Rev. 21, 322–365. https://doi.org/10.1139/er-2013-0042 (2013).Article 

Google Scholar 
Southee, F. M., Edwards, B. A., Chetkiewicz, C. B. & O’Connor, C. M. Freshwater conservation planning in the far north of Ontario, Canada: Identifying priority watersheds for conservation of fish biodiversity in an intact boreal landscape. Facets 6, 90–117. https://doi.org/10.1139/facets-2020-0015 (2021).Article 

Google Scholar 
Mitchell, M. G. E. et al. Identifying key ecosystem service providing areas to inform national-scale conservation planning. Environ. Res. Lett. 16, 014038. https://doi.org/10.1088/1748-9326/abc121 (2021).ADS 
CAS 
Article 

Google Scholar 
Labadie, G. P. D., McLoughlin, M. H. & Fortin, D. Insect-mediated apparent competition between mammals in a boreal food web. Proc. Natl. Acad. Sci. U S A. 118, e2022892118. https://doi.org/10.1073/pnas.2022892118 (2021).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Cameron, V. & Hargreaves, A. L. Spatial distribution and conservation hotspots of mammals in Canada. Facets 5, 692–703. https://doi.org/10.1139/facets-2020-0018 (2020).Article 

Google Scholar 
Ceballos, G. & Ehrlich, P. R. Global mammal distributions, biodiversity hotspots, and conservation. PNAS 103, 19374–19379. https://doi.org/10.1073/pnas.0609334103 (2016).ADS 
Article 

Google Scholar 
Anielski, M. & Wilson, S. Counting Canada’s natural capital: Assessing the real value of Canada’s boreal ecosystems. Ottawa, On: Canadian Boreal Initiative and Pembina Institute counting-canadas-natural-capital (2009).Kumaraswamy, S. & Udyakumar, M. Biodiversity banking: A strategic conservation mechanism. Biodiver. Conserv. 20, 1155–1165. https://doi.org/10.1007/s10531-011-0020-5 (2011).Article 

Google Scholar 
Garnett, S. T. et al. A spatial overview of the global importance of Indigenous lands for conservation. Nat. Sustain. 1, 369–374. https://doi.org/10.1038/s41893-018-0100-6 (2018).Article 

Google Scholar 
Godden, L. & Cowell, S. Conservation planning and Indigenous governance in Australia’s Indigenous Protected Areas. Restor. Ecol. 24, 692–697. https://doi.org/10.1111/rec.12394 (2016).Article 

Google Scholar 
Greg Brown, B. & Fagerholm, N. Empirical PPGIS/PGIS mapping of ecosystem services: A review and evaluation. Ecol. Ser. 13, 119–133. https://doi.org/10.1016/j.ecoser.2014.10.007 (2021).Article 

Google Scholar 
Martin, A. E., Neave, E., Kirby, P., Drever, C. R. & Johnson, C. A. Multi-objective optimization can balance trade-offs among boreal caribou, biodiversity, and climate change objectives when conservation hotspots do not overlap. Sci. Rep. 12, 11895. https://doi.org/10.1038/s41598-022-15274-8 (2022).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
COSEWIC. Canadian Wildlife Species at Risk. Committee on the Status of Endangered Wildlife in Canada (2018).Alberta Environment and Parks and Alberta Conservation Association. Status of the Arctic Grayling (Thymallus arcticus) in Alberta: Update 2015. Alberta Environment and Parks. Alberta Wildlife Status Report No. 57 (Update 2015). Edmonton, AB. 96 pp. (2015).Environment and Climate Change Canada (ECCC). 2016. Range map extents, species at risk, Canada. Government of Canada. Open Government Dataset. https://open.canada.ca/data/en/dataset/d00f8e8c-40c4-435a-b790-980339ce3121.Magurran, A. E. Measuring Biological Diversity 256 (Blackwell Publishing, 2004).
Google Scholar 
Caissy, P., Klemet-N’Guessan, S., Jackiw, R., Eckert, C. G. & Hargreaves, A. L. High conservation priority of range-edge plant populations not matched by habitat protection or research effort. Biol. Conserv. 249, 108732 (2020).Article 

Google Scholar 
Gaston, K. J. Rarity 201 (Chapman & Hall, 1994).Book 

Google Scholar 
Stralberg, D. Velocity-based macrorefugia for North American ecoregions. Zenodo. https://doi.org/10.5281/zenodo.2579337 (2019).Fuss, S. et al. Betting on negative emissions. Nat. Clim. Change 4, 850–853. https://doi.org/10.1038/nclimate2392 (2014).ADS 
CAS 
Article 

Google Scholar 
Chen, I., Hill, J. K., Ohlemüller, R. D. B. & Thomas, C. D. Rapid range shifts of species associated with high levels of climate warming. Science 333, 1024–1026. https://doi.org/10.1126/science.1206432 (2011).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Woodall, C. W. et al. An indicator of tree migration in forests of the eastern United States. For. Ecol. Manag. 257, 1434–1444 (2009).Article 

Google Scholar 
Iverson, L. R., Schwartz, M. W. & Prasad, A. M. How fast and far might tree species migrate in the eastern United States due to climate change? Glob. Ecol. Biogeogr. 13, 209–219 (2004).Article 

Google Scholar 
McLachlan, J. S., Hellmann, J. J. & Schwartz, M. W. A framework for debate of assisted migration in an era of climate change. Conserv. Biol. 21, 297–302 (2007).Article 

Google Scholar 
Sittaro, F., Paquette, A., Messier, C. & Nock, C. A. Tree range expansion in eastern North America fails to keep pace with climate warming at northern range limits. Glob. Change Biol. 23, 3292–3301. https://doi.org/10.1111/gcb.13622 (2017).ADS 
Article 

Google Scholar 
Ping, C. L. et al. Carbon stores and biogeochemical properties of soils under black spruce forest, Alaska. Soil Sci. Soc. Am. J. 74, 969–978. https://doi.org/10.2136/sssaj2009.0152 (2010).ADS 
CAS 
Article 

Google Scholar 
Hengl, T. et al. SoilGrids250m: Global soil information based on machine learning. PLoS ONE 12, e0169748 (2017).Article 

Google Scholar 
Chung, N. C., Miasojedow, B., Startek, M. & Gambin, A. Jaccard/Tanimoto similarity test and estimation methods for biological presence-absence data. BMC Bioinform. 29, 644. https://doi.org/10.1186/s12859-019-3118-5 (2019).Article 

Google Scholar 
Chung, N. C., Miasojedow, B., Startek, M. & Gambin A. Jaccard: Test Similarity Between Binary Data using Jaccard/Tanimoto Coefficients. R package version 0.1.0. https://CRAN.R-project.org/package=jaccard (2018). More

Characteristics of LVD dataset and assembled vOTUsThe land use virome dataset LVD was derived from 2.6 billion paired clean reads of sequences across 50 viromes of 25 samples with five types of land uses (Supplementary Data 2). A total of 6,442,065 contigs ( >1500 bp) were yielded, of which 764,466 (11.8%) contigs were identified as putative viral genomes through VIBRANT. Subsequently, putative false positive viral genomes were removed (see Methods section), and 27,951 and 48,936 bona fide viral genomes were retained from the 25 intracellular VLPs (iVLPs) and 25 extracellular VLPs (eVLPs) viromes, respectively. These genomes were clustered into 25,941 and 45,152 vOTUs for iVLPs and eVLPs viromes, respectively, in which the iVLPs and eVLPs viromes shared 11,467 (19.2%) vOTUs. Subsequently, they were merged and dereplicated, resulting in 59,626 vOTUs (Supplementary Data 3) for the following analysis. A total of 8112 (13.6%) vOTUs genomes were classified as complete, in which the median length of all and circular vOTUs were 25,183 bp and 45,511 bp, respectively (Supplementary Fig. 4).To explore the taxonomic affiliation of vOTUs in family and genus-level, a gene-sharing network consist of 59,626 vOTUs genomes from this study and 3502 reference phage genomes (from NCBI Viral RefSeq version 201) revealed 6009 VCs comprising of 37,224 vOTUs, of which 34,417 vOTUs were from LVD, besides 2794 singletons (2653 from LVD dataset), 16,056 outliers (15,833 from LVD) and 8492 overlaps (8061 from LVD) were detected (Supplementary Data 4). Of these, only 157 VCs contained genomes from both the RefSeq and LVD dataset (1864 viral genomes) (Supplementary Data 4). Most of VCs (1837, 30.4%) included only two members.At the family level, most of vOTUs were classified into Siphoviridae (712 by vConTACT2 and 29,671 (50.9%) by Demovir, tailed dsDNA), Podoviridae (610 by vConTACT2 and 9923 (17.6 %) by Demovir, tailed dsDNA), Myoviridae (485 by vConTACT2 and 5445 (9.9%) by Demovir, tailed dsDNA), Tectiviridae (50 by vConTACT2 and 10 (0.10%) by Demovir, non-tailed dsDNA) (Fig. 1). Besides, the Eukaryotic viruses Herpesviridae (159 by Demovir, 0.26%, dsDNA), Phycodnaviridae (120 (0.20%) by Demovir, dsDNA); the Virophage Family Lavidaviridae (15 (0.03%) by Demovir) were detected as well, but a majority of vOTUs were unclassified in genus-level.Fig. 1: The taxonomic assignment of LVD.Pie charts showing the affiliation of 56,870 vOTUs at family level assigned by script Demovir (a). and the affiliation of 1864 vOTUs at family level assigned by package vConTACT2 (b). Source data are provided in the Source Data file.Full size imageViral community structures differ across land use typesBray–Curtis dissimilarity of viral communities (median 0.9951) showed strong heterogeneity of viral communities among different sites (Fig. 2a). While, the Bray–Curtis dissimilarity (median: 0.5109) between paired viral communities of iVLPs and eVLPs from each site have a significant lower heterogeneity than inter-sites (Wilcox.test, p  0.05; Fig. 2b). Therefore, the paired iVLPs and eVLPs viromes from each site were merged for subsequently viral community analysis.Fig. 2: The macrodiversity of soil viral communities.a Boxplot showing Bray–Curtis dissimilarity of viral communities of intra-sites (between the corresponding community of iVLPs and eVLPs, n = 25) and inter-sites (between different sample sites, n = 300). The minima, maxima, center, bounds of box and whiskers in boxplots from bottom to top represented percentile 0, 10, 25, 50, 75, 90, and 100, respectively, the difference between different zones was tested using the two-sided Wilcox.test, ****p  More

We developed the IUCN Global Ecosystem Typology in the following sequence of steps: design criteria; hierarchical structure and definition of levels; generic ecosystem assembly model; top-down classification of the upper hierarchical levels; iterative circumscription of the units and ecosystem-specific adaptations of the assembly model; full description of the units; and map compilation. Some iteration proved necessary, as the description and review process sometimes revealed a need for circumscribing additional units.Design criteria and other typologiesUnder the auspices of the IUCN Commission on Ecosystem Management, we developed six design principles to guide the development of a typology that would meet the needs for global ecosystem reporting, risk assessment, natural capital accounting and ecosystem management: (1) representation of ecological processes and ecosystem functions; (2) representation of biota; (3) conceptual consistency throughout the biosphere; (4) scalable structure; (5) spatially explicit units; and (6) parsimony and utility (see Supplementary Table 1.1 and Supplementary Information, Appendix 1 for definitions and rationale).We assessed 23 existing ecological classifications with global coverage of terrestrial, freshwater, and/or marine environments against these principles to determine their fitness for IUCN’s purpose (Supplementary Information, Appendix 1). These include general classifications of land, water or bioclimate, as well as classifications of units that conform with the definition of ecosystems adopted in the United Nations Convention on Biological Diversity45 or an equivalent definition in the IUCN Red List of Ecosystems30. We reviewed documentation on methods of derivation, descriptions of classification units and maps to assess each classification against the six design principles (Supplementary Table 1.2 for details).Typology structure and ecosystem assemblyWe developed the structure of the Global Ecosystem Typology and the generic ecosystem assembly model at a workshop attended by 48 terrestrial, freshwater and marine ecosystem experts at Kings College London, UK, in May 2017. Participants agreed that a hierarchical structure would provide an effective framework for integrating ecological processes and functional properties (Supplementary Table 1.1, design principle 1), and biotic composition (principle 2) into the typology, while also meeting the requirement for scalability (principle 4). Although neither function nor composition were intended to take primacy within the typology, we reasoned that a hierarchy representing functional features in the upper levels is likely to support generalizations and predictions by leveraging evolutionary convergence13. By contrast, a typology reflecting compositional similarities in its upperlevels is less likely to be stable owing to dynamism of species assemblages and evolving knowledge on species taxonomy and distributions. Furthermore, representation of compositional relationships at a global scale would require many more units in upper levels, and possibly more hierarchical levels. Therefore, we concluded that a hierarchical structure recognizing compositional variants at lower levels within broad functionally based groupings at upper levels would be more parsimonious and robust (principle 6) than one representing composition at upper levels and functions at lower levels.Workshop participants initially agreed that three hierarchical levels for ecosystem function and three levels for biotic composition could be sufficient to represent global variation across the whole biosphere. Participants developed the concepts of these levels into formal definitions (Supplementary Table 3.1), which were reviewed and refined during the development process.To ensure conceptual consistency of the typology and its units throughout the biosphere (principle 3), we drew from community assembly theory to develop a generic model of ecosystem assembly. The traditional community assembly model incorporates three types of filters (dispersal, the abiotic environment and biotic interactions) that determine which biota from a larger pool of potential colonists can occupy and persist in an area13. We extended this model to ecosystems by: (1) defining three groups of abiotic filters (resources, ambient environment and disturbance regimes) and two groups of biotic filters (biotic interactions and human activity); (2) incorporating evolutionary processes that shape characteristic biotic properties of ecosystems over time; (3) defining the outcomes of filtering and evolution in terms of all ecosystem properties including both ecosystem-level functions and species-level traits, rather than only in terms of species traits and composition; and (4) incorporating interactions and feedbacks among filters and selection agents and ecosystem properties to elucidate hypotheses about processes that influence temporal and spatial variability in the properties of ecosystems and their component biota. In community assembly, only a small number of filters are likely to be important in any given habitat13. In keeping with this proposition, we used the generic model to identify biological and physical features that distinguish functionally different groups of ecosystems from one another by focusing on different ecological drivers that come to the fore in structuring their assembly and shaping their properties.Hierarchical levelsThe top level of classification (Fig. 2 and Extended Data Tables 1–4) defines five core realms of the biosphere based on contrasting media that reflect ecological processes and functional properties: terrestrial; freshwaters and inland saline waters (hereafter freshwater); marine; subterranean; and atmospheric. Biome gradient concepts25 highlight continuous variation in ecosystem properties, which is represented in the typology by transitional realms that mark the interfaces between the five core realms (for example, floodplains (terrestrial–freshwater), estuaries (freshwater–marine), and so on). In Supplementary Information, Appendix 3 (pages 3–16) and Supplementary Table 3.1, we describe the five core realms and review the hypothesized assembly filters and ecosystem properties that distinguish different groups within them. The atmospheric realm is included for comprehensive coverage, but we deferred resolution of its lower levels because its biota is poorly understood, sparse, itinerant and represented mainly by dispersive life stages46.Functional biomes (level 2) are components of the biosphere united by one or more major assembly processes that shape key ecosystem functions and ecological processes, irrespective of taxonomic identity (Supplementary Information, Appendix 3, page 17). Our interpretation aligns broadly with ‘functional biomes’ described elsewhere24,25,47, extended here to reflect dominant assembly filters and processes across all realms, rather than the more restricted basis of climate-vegetation relationships that traditionally underpin biome definition on land. Hence, the 25 functional biomes (Supplementary Information, Appendix 4, pages 52–186 and https://global-ecosystems.org/) include some ‘traditional’ terrestrial biomes47, as well as lentic and lotic freshwater systems, pelagic and benthic marine systems, and anthropogenic functional biomes assembled and usually maintained by human activity48.Level 3 of the typology defines 110 ecosystem functional groups described with illustrated profiles in Supplementary Information, Appendix 4 (pages 52–186) and at https://global-ecosystems.org/. These are key units for generalization and prediction, because they include ecosystem types with convergent ecosystem properties shaped by the dominance of a common set of drivers (Supplementary Information, Appendix 3, pages 17–19). Ecosystem functional groups are differentiated along environmental gradients that define spatial and temporal variation in ecological drivers (Figs. 2 and 3 and Supplementary Figs. 3.2 and 3.4). For example, depth gradients of light and nutrients differentiate functional groups in pelagic ocean waters (Fig. 3c and Extended Data Table 4), influencing assembly directly and indirectly through predation. Resource gradients defined by flow regimes (influenced by catchment precipitation and evapotranspiration) and water chemistry, modulated by environmental gradients in temperature and geomorphology, differentiate functional groups of freshwater ecosystems25 (Fig. 3b and Extended Data Table 3). Terrestrial functional groups are distinguished primarily by gradients in water and nutrient availability and by temperature and seasonality (Fig. 3a and Extended Data Table 1), which mediate uptake of those resources and regulate competitive dominance and productivity of autotrophs. Disturbance regimes, notably fire, are important global drivers in assembly of some terrestrial ecosystem functional groups49.Three lower levels of the typology distinguish functionally similar ecosystems based on biotic composition. Our focus in this paper is on global functional relationships of ecosystems represented in the upper three levels of the typology, but the lower levels (Supplementary Information, Appendix 3, pages 19 and 20) are crucial for representing the biota in the typology, and facilitate the scaling up of information from established local-scale typologies that support decisions where most conservation action takes place. These lower levels are being developed progressively through two contrasting approaches with different trade-offs, strengths and weaknesses. First, level 4 units (regional ecosystem subgroups) are ecoregional expressions of ecosystem functional groups developed from the top-down by subdivisions based on biogeographic boundaries (for example, in ref. 50) that serve as simple and accessible proxies for biodiversity patterns51. Second, level 5 units (global ecosystem types) are also regional expressions of ecosystem functional groups, but unlike level 4 units they are explicitly linked to local information sources by bottom-up aggregation52 and rationalization of level 6 units from established subglobal ecological classifications. Subglobal classifications, such as those for different countries (see examples for Chile and Myanmar in Supplementary Tables 3.3 and 3.4), are often developed independently of one another, and thus may involve inconsistencies in methods and thematic resolution of units (that is, broadly defined or finely split). Aggregation of level 6 units to broader units at level 5 based on compositional resemblance is necessary to address inconsistencies among different subglobal classifications and produce compositionally distinctive units suitable for global or regional synthesis.Integrating local classifications into the global typology, rather than replacing them, exploits considerable efforts and investments to produce existing classifications, already developed with local expertise, accuracy and precision. By placing national and regional ecosystems into a global context, this integration also promotes local ownership of information to support local action and decisions, which are critical to ecosystem conservation and management outcomes (Supplementary Information, Appendix 3, page 20). These benefits of bottom-up approaches come at the cost of inevitable inconsistencies among independently developed classifications from different regions, a limitation avoided in the top-down approach applied to level 4.Circumscribing upper-level unitsWe formed specialist working groups (terrestrial/subterranean, freshwater and marine) to develop descriptions of the units within the upper levels of the hierarchy, subdividing realms into functional biomes, and biomes into ecosystem functional groups. We used definitions of the hierarchical levels (Supplementary Table 3.1) and the conceptual model of ecosystem assembly (Fig. 1) to maintain consistency in defining the units at each level during iterative discussions within and between the working groups.Working groups agreed on preliminary lists of functional biomes and ecosystem functional groups by considering variation in major drivers along ecological gradients (Figs. 2 and 3 and Supplementary Figs. 3.2 and 3.4) based on published literature, direct experience and expertise of working group members, and consultation with colleagues in their respective research networks. After the workshop, working groups sought recent global reviews of the candidate units and recent case studies of exemplars to shape descriptions of the major groups of ecosystem drivers and properties for each unit. Circumscriptions and descriptions of the units were reviewed and revised iteratively to ensure clear distinctions among units, with a total of 206 reviews of descriptive profiles undertaken by 60 specialists, a mean of 2.4 reviews per profile (Supplementary Table 5.1). The working groups concurrently adapted the generic model of ecosystem assembly (Fig. 1) to represent working hypotheses on salient drivers and ecosystem properties for each ecosystem functional group.Incorporating human influenceVery few of the ecological typologies reviewed in Supplementary Information, Appendix 1 integrate anthropogenic ecosystems in their classificatory frameworks. Anthropogenic influences create challenges for ecosystem classification, as they may modify defining features of ecosystems to a degree that varies from negligible to major transformation across different locations and times. We addressed this problem by distinguishing transformative outcomes of human activity at levels 2 and 3 of the typology from lesser human influences that may be represented either at levels 5 and 6, or through measurements of ecosystem integrity or condition that reflect divergence from reference states arising from human activity.Anthropogenic ecosystems grouped within levels 2 and 3 were thus defined as those created and sustained by intensive human activities, or arising from extensive modification of natural ecosystems such that they function very differently. These activities are ultimately driven by socio-economic and cultural-spiritual processes that operate across local to global scales of human organization. In many agricultural and aquacultural systems and some others, cessation of those activities may lead to transformation into ecosystem types with qualitatively different properties and organizational processes (see refs. 53,54 for cropland and urban examples, respectively). Indices such as human appropriation of net primary productivity55, combined with land-use maps56, offer useful insights into the distribution of some anthropogenic ecosystems, but further development of indices is needed to adequately represent others, particularly in marine, and freshwater environments. Beyond land-use classification and mapping approaches (Supplementary Information, Appendix 1, page 6), a more comprehensive elaboration of the intensity of human influence underpinning the diverse range of anthropogenic ecosystems requires a multidimensional framework incorporating land-use inputs, outputs, their interactions, legacies of earlier activity and changes in system properties17.Where less intense human activities occur within non-anthropogenic ecosystem types, we focused descriptions on low-impact reference states. Therefore, human activities are not shown as drivers in the assembly models for non-anthropogenic ecosystem groups, even though they may have important influences on the contemporary ecosystem distribution. This approach enables the degree and nature of human influence to be described and measured against these reference states using assessment methods such as the Red List of Ecosystems protocol30, with appropriate data on ecosystem change.Indicative distribution mapsFinally, to produce spatially explicit representations of the units at level 3 of the typology (principle 5), we sought published global maps (sources in Supplementary Table 4.1) that were congruent with the concepts of respective ecosystem functional groups. Where several candidate maps were available, we selected maps with the closest conceptual alignment, finest spatial resolution, global coverage, most recent data and longest time series. The purpose of maps for our study was to visualize global distributions. Prior to applications of map data to spatial analysis, we recommend critical review of methods and validation outcomes reported in each data source to ensure fitness for purpose (Supplementary Information, Appendix 4).Extensive searches of published literature and data archives identified high-quality datasets for some ecosystem functional groups (for example, T1.3 Tropical–subtropical montane rainforests; MT1.4 Muddy shorelines; M1.5 Sea ice) and datasets that met some of these requirements for a number of other ecosystem functional groups (see Supplementary Table 4.1 for details). Where evaluations by authors or reviewers identified limitations in available maps, we used global environmental data layers and biogeographic regionalizations as masks to adjust source maps and improve their congruence to the concept of the relevant functional group (for example, F1.2 Permanent lowland rivers). For ecosystem functional groups with no specific global mapping, we used ecoregions50,57,58 as biogeographic templates to identify broad areas of occurrence. We consulted ecoregion descriptions, global and regional reviews, national and regional ecosystem maps, and applied in situ knowledge of participating experts to identify ecoregions that contain occurrences of the relevant ecosystem functional group (for example, T4.4 Temperate woodlands) (see Supplementary Table 4.1 for details). We mapped ecosystem functional groups as major occurrences where they dominated a landscape or seascape matrix and minor occurrences where they were present, but not dominant in landscape–seascape mosaics, or where dominance was uncertain. Although these two categories in combination communicate more information about ecosystem distribution than binary maps, simple spatial overlays using minor occurrences are likely to inflate spatial statistics. The maps are progressively upgraded in new versions of the typology as explicit spatial models are developed and new data sources become available (see ref. 27 for a current archive of spatial data).The classification and descriptive profiles, including maps, for each functional biome and ecosystem functional group underwent extensive consultation, and targeted peer review and revision through a series of four phases described in Supplementary Information, Appendix 5 (pages 2–4). The reviewer comments and revisions from targeted peer review are documented in Supplementary Table 5.1. In all, more than 100 ecosystem specialists have contributed to the development of v2.1 of the typology.LimitationsUneven knowledge of Earth’s biosphere has constrained the delimitation and description of units within the typology. There is a considerable research bias across the full range of Earth’s ecosystems, with few formal research studies evaluating the relative influence of different ecosystem drivers in many of the functional groups, and abiotic assembly filters generally receiving more attention than biotic and dispersal filters. This poses challenges for developing standardized models of assembly for each ecosystem functional group. The models therefore represent working hypotheses, for which available evidence varies from large bodies of published empirical evidence to informal knowledge of ecosystem experts and their extensive research networks. Large numbers of empirical studies exist for some forest functional groups, savannas, temperate heathlands in Mediterranean-type climates, coral reefs, rocky shores, kelp forests, trophic webs in pelagic waters, small permanent freshwater lakes, and others (see references in the respective profiles (Supplementary Information, Appendix 4)). For example, Bond49 reviewed empirical and modelling evidence on the assembly and function of tropical savannas that make up three ecosystem functional groups, showing that they have a large global biophysical envelope that overlaps with tropical dry forests, and that their distribution and dynamics within that envelope is strongly influenced by top-down regulation via biotic filters (large herbivores and their predators) and recurrent disturbance regimes (fires). Despite the development of this critical knowledge base, savannas suffer from an awareness disparity that hinders effective conservation and management59. In other ecosystems, our assembly models rely more heavily on inferences and generalizations of experts drawn from related ecosystems, are more sensitive to interpretations of participating experts, and await empirical testing and adjustment as understanding improves. Empirical tests could examine hypothesized variation in ecosystem properties along gradients within and between ecosystem functional groups and should return incremental improvements on group delineation and description of assembly processes.High-quality maps at suitable resolution are not yet available for the full set of ecosystem functional groups, which limits current readiness for global analysis. The maps most fit for global synthesis are based on remote sensing and environmental predictors that align closely to the concept of their ecosystem functional group, incorporate spatially explicit ground observations and have low rates of omission and commission errors, ‘high’ spatial resolution (that is, rasters of 1 km2 (30 arcsec) or better), and time series of changes. Sixty of the maps currently in our archive27 aligned directly or mostly with the concept of their corresponding ecosystem functional group, while the remainder were based on indirect spatial proxies, and most were derived from polygon data or rasters of 30 arcsec or finer (Supplementary Table 4.1). Maps for 81 functional groups were based either on known records, or on spatial data validated by quantitative assessments of accuracy or efficacy. Therefore, we suggest that maps currently available for 60–80 of the 110 functional groups are potentially suitable for global spatial analysis of ecosystem distributions. Although, a significant advance on broad proxies such as ecoregions, the maps currently available for ecosystem functional groups would benefit from expanded application of recent advances in remote sensing, environmental datasets, spatial modelling and cloud computing to redress inequalities in reliability and resolution. The most urgent priorities for this work are those identified in Supplementary Table 4.1 as relying on indirect proxies for alignment to concept, qualitative evaluation by experts and coarse resolution ( >1 km2) spatial data.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More

Bartz, K. K. & Naiman, R. J. Effects of Salmon-Borne nutrients on riparian soils and vegetation in Southwest Alaska. Ecosystems 8, 529–545 (2005).Article 

Google Scholar 
Erskine, P. D. et al. Subantarctic Macquarie Island—a model ecosystem for studying animal-derived nitrogen sources using 15N natural abundance. Oecologia 117, 187–193 (1998).ADS 
PubMed 
Article 

Google Scholar 
Hocking, M. D. & Reimchen, T. E. Salmon species, density and watershed size predict magnitude of marine enrichment in riparian food webs. Oikos 118(9), 1307–1318 (2009).Article 

Google Scholar 
Hocking, M. D. & Reynolds, J. D. Impacts of salmon on riparian plant diversity. Science 331, 1609–1612 (2011).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Hocking, M. D., & Reimchen, T. E. Salmon-derived nitrogen in terrestrial invertebrates from coniferous forests of the Pacific Northwest. BMC Ecol. 2, 4. https://doi.org/10.1186/1472-6785-2-4 (2002).Bilby, R. E., Fransen, B. R. & Bisson, P. A. Incorporation of nitrogen and carbon from spawning coho salmon into the trophic system of small streams: Evidence from stable isotopes. Can. J. Fish Aquat. Sci. 53, 164–173 (1996).Article 

Google Scholar 
Talley, D. M. et al. Research challenges at the land–sea interface. Estuar. Coast. Shelf Sci. 58, 699–702 (2003).ADS 
Article 

Google Scholar 
Mizutani, H. & Wada, E. Nitrogen and carbon isotope ratios in seabird rookeries and their ecological implications. Ecology 69(2), 340–349 (1988).Article 

Google Scholar 
Rowe, J. A., Litton, C. M., Lepczyk, C. A. & Popp, B. N. Impacts of endangered seabirds on nutrient cycling in montane forest ecosystems of Hawai’i. Pac. Sci. 71(4), 495–509 (2017).Article 

Google Scholar 
Sanchez-Pinero, F. & Polis, G. A. Bottom-up dynamics of allochthonous input: Direct and indirect effects of seabirds on islands. Ecology 81(11), 3117–3132 (2000).Article 

Google Scholar 
Wait, D. A., Aubrey, D. P. & Anderson, W. B. Seabird guano influences on desert islands: Soil chemistry and herbaceous species richness and productivity. J. Arid Environ. 60, 681–695 (2005).ADS 
Article 

Google Scholar 
Stapp, P., Polis, G. A. & Pinero, F. S. Stable isotopes reveal strong marine and El Nino effects on island food webs. Nature 401, 467–469 (1999).ADS 
CAS 
Article 

Google Scholar 
Anderson, W. B., Wait, D. A. & Stapp, P. Resources from another place and time: Responses to pulses in a spatially subsidized system. Ecology 89(3), 660–670 (2008).PubMed 
Article 

Google Scholar 
Ellis, J. C. Marine birds on land: A review of plant biomass, species richness, and community composition in seabird colonies. Plant Ecol. 181(2), 227–241 (2005).Article 

Google Scholar 
Fukami, T. et al. Above- and below-ground impacts of introduced predators in seabird-dominated island ecosystems. Ecol. Lett. 9, 1299–1307 (2006).PubMed 
Article 

Google Scholar 
Wootton, J. T. Direct and indirect effects of nutrients on intertidal community structure: Variable consequences of seabird guano. J. Exp. Mar. Biol. Ecol. 151, 139–153 (1991).Article 

Google Scholar 
McCauley, D. J., et al., From wing to wing: the persistence of long ecological interaction chains in less-disturbed ecosystems. Sci. Rep. 2, 409. https://doi.org/10.1038/srep00409 (2012).Young, H. S., McCauley, D. J., Dunbar, R. B. & Dirzo, R. Plants cause ecosystem nutrient depletion via the interruption of bird-derived spatial subsidies. Proc. Natl. Acad. Sci. U.S.A. 107(5), 2072–2077 (2010).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Lindeboom, H. J. The nitrogen pathway in a Penguin rookery. Ecology 65(1), 269–277 (1984).CAS 
Article 

Google Scholar 
Mizutani, H., Kabaya, Y. & Wada, E. Ammonia volatilization and high 15N/14N ratio in a penguin rookery in Antarctica. Geochem. J. 19(6), 323–327 (1985).ADS 
CAS 
Article 

Google Scholar 
Anderson, W. B. & Polis, G. A. Nutrient fluxes from water to land: seabirds affect plant nutrient status on Gulf of California islands. Oecologia 118, 324–332 (1999).ADS 
PubMed 
Article 

Google Scholar 
Polis, G. A. & Hurd, S. D. Linking marine and terrestrial food webs: Allochthonous input from the ocean supports high secondary productivity on small islands and coastal land communities. Am. Nat. 147, 396–423 (1996).Article 

Google Scholar 
Goss, N. S. New and rare birds found breeding on the San Pedro Martir Isle. University of California Press 5, 240–244 (1888).
Google Scholar 
Velarde, E., et al., Nesting seabirds of the Gulf of California’s Offshore islands: Diversity, ecology and conservation. in Biodiversity, Ecosystems, and Conservation in Northern Mexico, Carton, J.-L. E., Ceballos, G., Felger, R. S. Eds. (Oxford University Press, 2005) pp. 452–470.Wilder, B. T., Felger, R. S. & Ezcurra, E. Controls of plant diversity and composition on a desert archipelago. PeerJ 7, e7286. https://doi.org/10.7717/peerj.7286 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Ellis, J., Fariña, J. & Witman, J. Nutrient transfer from sea to land: the case of gulls and cormorants in the Gulf of Maine. J. Anim. Ecol. 75, 565–574 (2006).PubMed 
Article 

Google Scholar 
Wilder, B. T., Felger, R. S. & Morales, H. R. Succulent plant diversity of the Sonoran Islands, Gulf of California Mexico. Haseltonia 2008(14), 127–160 (2008).Article 

Google Scholar 
Lucassen, F. et al. The stable isotope composition of nitrogen and carbon and elemental contents in modern and fossil seabird guano from Northern Chile—Marine sources and diagenetic effects. PLoS ONE 12(6), e0179440. https://doi.org/10.1371/journal.pone.0179440 (2017).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Robinson, D. δ15N as an integrator of the nitrogen cycle. Trends Ecol. Evol. 16(3), 153–162 (2001).CAS 
PubMed 
Article 

Google Scholar 
Szpak, P., Longstaffe, F. J., Millaire, J.-F. & White, C. D. Stable isotope biogeochemistry of seabird guano fertilization: Results from growth chamber studies with maize (Zea mays). PLoS ONE 7(3), e33741. https://doi.org/10.1371/journal.pone.0033741 (2012).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Ezcurra, E., et al. Natural History and Evolution of the World’s Deserts. Global Deserts Outlook. United Nations Environment Programme (UNEP), 1–26 (2006).Yetman, D. The Great Cacti: Ethnobotany and biogeography (University of Arizona Press, 2007).
Google Scholar 
Álvarez-Borrego, S. Physical oceanography. in A New Island Biogeography of the Sea of Cortés, Case, T. J., Cody, M. L., Ezcurra, E. Eds. (Oxford University Press, 2002), pp. 41–59.Douglas, R., Gonzalez-Yajimovich, O., Ledesma-Vazquez, J. & Staines-Urias, F. Climate forcing, primary production and the distribution of Holocene biogenic sediments in the Gulf of California. Quatern. Sci. Rev. 26, 115–129 (2007).ADS 
Article 

Google Scholar 
Urbán, J. Marine mammals of the Gulf of California: An overview of diversity and conservation status. in The Gulf of California: Biodiversity and conservation, R. C. Brusca, Ed. (The University of Arizona Press and the Arizona-Sonora Desert Museum, 2010), pp. 188–209.Hastings, P. A., Findley, L. T., & Van der Heiden, A. M. Fishes of the Gulf of California. in: Brusca, R. C., (eds) The Gulf of California: Biodiversity and conservation 96–118, The University of Arizona Press and the Arizona-Sonora Desert Museum (2010).
Google Scholar 
Polis, G. A., Hurd, S. D., Jackson, C. T. & Sanchez Piñero, F. El Niño effects on the dynamics and control of an Island ecosystem in the Gulf of California. Ecology 78, 1884–1897 (1997).
Google Scholar 
Wilder, B. T. & Felger, R. S. Dwarf giants, guano, and isolation: The flora and vegetation of San Pedro Mártir Island, Gulf of California, Mexico. Proc. San Diego Soc. Nat. Hist. 42, 1–24 (2010).
Google Scholar 
Medel-Narvaez, A., Leon Luz, J. L., Freaner-Martinez, F. & Molina-Freaner, F. Patterns of abundance and population structure of Pachycereus pringlei (Cactaceae), a columnar cactus of the Sonoran Desert. Plant Ecol. 187, 1–14 (2006).Article 

Google Scholar 
Felger, R.S., Wilder, B.T. in collaboration with Romero-Morales, H. Plant Life of a Desert Archipelago: Flora of the Sonoran Islands in the Gulf of California. Tucson, University of Arizona Press (2012).Wilkinson, C. E., Hocking, M. D. & Reimchen, T. E. Uptake of salmon-derived nitrogen by mosses and liverworts in Coastal British Columbia. Oikos 108, 85–98 (2005).CAS 
Article 

Google Scholar 
Barrett, K., Wait, D. A. & Anderson, W. B. Small island biogeography in the Gulf of California: Lizards, the subsidized island biogeography hypothesis, and the small island effect. J. Biogeogr. 30, 1575–1581 (2003).Article 

Google Scholar 
Young, H. S., McCauley, D. J. & Dirzo, R. Differential responses to guano fertilization among tropical tree species with varying functional traits. Am. J. Bot. 98, 207–214 (2011).PubMed 
Article 

Google Scholar 
Nobel, P. S. Environmental Biology of Agaves and Cacti. Cambridge University Press (2003).Ramirez, K. S., Craine, J. M. & Fierer, N. Consistent effects of nitrogen amendments on soil microbial communities and processes across biomes. Glob. Change Biol. 18(6), 1918–1927 (2012).ADS 
Article 

Google Scholar 
Craine, J. M. et al. Ecological interpretations of nitrogen isotope ratios of terrestrial plants and soils. Plant Soil 396, 1–26 (2015).CAS 
Article 

Google Scholar 
Schoeninger, M. J. & DeNiro, M. J. Nitrogen and carbon isotope composition of bone collagen from marine and terrestrial animals. Geochim. Cosmochim. Acta 48(4), 625–639 (1984).ADS 
CAS 
Article 

Google Scholar 
Amundson, R. et al. Global patterns of the isotopic composition of soil and plant nitrogen. Global Biogeochem. Cycles 17(1), 1031. https://doi.org/10.1029/2002GB001903 (2003).ADS 
CAS 
Article 

Google Scholar 
Kahmen, A., Wanek, W. & Buchmann, N. Foliar δ15N values characterize soil N cycling and reflect nitrate or ammonium preference of plants along a temperate grassland gradient. Oecologia 156, 861–870 (2008).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Bowen, T. Unknown Island: Seri Indians, Europeans, and San Esteban Island in the Gulf of California (University of New Mexico Press, 2000).
Google Scholar 
Evans, R. D. Physiological mechanisms influencing plant nitrogen isotope composition. Trends Plant Sci. 6(3), 121–126 (2001).CAS 
PubMed 
Article 

Google Scholar 
Dolby, G., Bennett, S. E. K., Lira-Noriega, A., Wilder, B. T. & Munguia-Vega, A. Assessing the geological and climatic forcing of biodiversity and evolution surrounding the Gulf of California. J. Southw. 57, 391–455 (2015).Article 

Google Scholar 
Case, T. J., Cody, M. L., & Ezcurra, E. A New Island Biogeography of the Sea of Cortés (Oxford University Press, 2002).Book 

Google Scholar 
Tershy, B. R. & Breese, D. The birds of San Pedro Mártir Island, Gulf of California Mexico. West. Birds 28, 96–107 (1997).
Google Scholar 
Tershy, B. R., Breese, D. & Croll, D. A. Human perturbations and conservation strategies for San Pedro Mártir Island, Islas de Golfo de California Reserve México. Environ. Conserv. 24, 261–270 (1997).Article 

Google Scholar 
Wilder, B. T. Historical biogeography of the Midriff Islands in the Gulf of California, Mexico. Dissertation. Riverside: UC, Riverside (2014).Post, D. M. et al. Getting to the fat of the matter: Models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia 152, 179–189 (2007).ADS 
PubMed 
Article 

Google Scholar 
Kiljunen, M. et al. A revised model for lipid-normalizing δ13C values from aquatic organisms, with implications for isotope mixing models. J. Appl. Ecol. 43, 1213–1222 (2006).CAS 
Article 

Google Scholar 
Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. B. lmerTest Package: Tests in linear mixed effects models. J. Stat. Softw. 82(13), 1–26. https://doi.org/10.18637/jss.v082.i13 (2017).Article 

Google Scholar 
R Core Team, R: A language and environment for statistical computing. https://www.R-project.org/ (R Foundation for Statistical Computing, Vienna, Austria, 2022). More

Description of PTEsThe descriptive statistics of the contents of soil PTEs in the study area were shown in Table 1. From Table 1, the mean contents of As and Ni in the oil-affected soils exceeded their corresponding risk screening values33, which may damage the soil ecological environment and affect crop growth. Compared with the secondary standard of soil environmental quality34, the mean contents of As, Cu and Zn were all lower than their corresponding Grade II standard values, but the mean contents of Cd, Cr, Ni and Pb in the oil-affected soils were 1.07, 7.46, 7.14 and 1.36 times of their standard values. In contrast with the background value of Hubei province35, except Mn, the mean contents of As, Cd, Cr, Cu, Ni, Pb, Zn and Ba in the oil-affected soils all exceeded their background values. Meanwhile, the variation coefficient of Cr (1.41) was greater than 1. In general, the soil Cd concentration in the study area was higher than that around Gudao Town, a typical oil-producing region of the Shengli Oilfield in the Yellow River Delta, China12, and from Yellow River Delta, a traditional oil field in China9, but was lower than that around two crude oil flow stations in the Niger Delta, Nigeria36. The concentrations of other PTEs were higher than the corresponding element concentrations, detected in the soil around Gudao Town, a typical oil-producing region of the Shengli Oilfield in the Yellow River Delta, China12, from Yellow River Delta, a traditional oil field in China9, and around two crude oil flow stations in the Niger Delta, Nigeria36. The above analysis exhibited that PTEs in the oil-affected soils had a certain degree of accumulation and may be affected by human activities.Table 1 Statistical characteristics for potential toxic elements in in the study area (mg·kg−1).Full size tableLevels of PTEs enrichment and pollutionThe EF and PLI of soil PTEs in the study area were calculated to evaluate the pollution degree of soil PTEs. The calculation results of EF and PLI were shown in Fig. 2 and Table S4. From Fig. 2, the mean EF values of PTEs were showed as Pb  > Cr  > Ni  > As  > Cd  > Zn  > Cu  > Ba. The mean EFs of all PTEs were greater than 1. Among them, the average EF of Cu, Zn and Ba was between 1 and 2, which was slightly enriched. And As (2.18) and Cd (2.12) were moderately enriched. In particular, the average EF values of Cr, Ni and Pb were 14.23, 8.69 and 15.45, respectively, reaching a significant enrichment level, and all samples of Cr, Ni and Pb were at moderate or above enrichment, of which 10% of the Cr samples were extreme pollution, 85% of Cr samples, 95% of Ni and 5% of Pb (Table S4) were significantly enriched. These proved that these PTEs were generally enriched in the study area, especially Cr, Ni and Pb.Figure 2The map of enrichment factor and contamination factor of PTEs in the study area.Full size imageExcept Mn, the average CF values of other PTEs were all  > 1 (Fig. 2), indicating that the accumulation of Mn in the study area was relatively light, and there was no obvious Mn pollution. The CF values of all samples of As, Cr, Ni and Pb, 80% of Cd samples, 75% of Cu samples, 30% of Mn samples, 65% of Zn samples and 75% of Ba samples (Table S4) were higher than 1. And the mean CF values of Cr, Ni and Pb were 14.21, 7.58 and 12.73, respectively, certifying that the pollution of Cr, Ni and Pb in the study area was considerably serious. PLI was calculated based on the CF value of PTEs, and the results were shown in Fig. 2. The average value of PLI was 2.62, indicating that the soil PTEs in the study area were seriously polluted.Spatial distribution of soil PTEs in the study areaGeostatistical analysis was utilized to do ordinary Kriging interpolation of the PTEs in the study area, the results were shown in Fig. 3. As shown in Fig. 3, the spatial distribution of As, Cr, Ni, Zn and Ba was relatively consistent, and their hot spots were concentrated in the southeast, northwest, and central and eastern parts of the study area where oil wells were distributed. The spatial distribution of Cr and Ni exhibited that there were large-scale hotspots near the oil wells, and the content of Cr and Ni in these hotspots was much higher than second-level environmental quality standards of China, which proved that the content of soil Cr and Ni was significantly affected by the oil production activities of the oil production plant. There were crude oil leaks in B and C, and the contents of Zn and Ba in the vicinity of these two oil wells were relatively high, indicating that soil Zn and Ba in this area may be affected by the crude oil leakage, resulting in a certain degree of accumulation in the soil. The area with the second highest As content mainly resided in the middle of the study area. According to the survey, the herbicides were sprayed every year around the H oil well in the middle of the study area, indicating that the accumulation of As in the soil was not only related to oil extraction activities, but also to the use of pesticides (contains copper arsenate, sodium arsenate, etc.)10, 14. In addition, the hot spots of spatial distribution of Pb, Cd and Mn were concentrated in the southeast, and Cu was mainly concentrated in the southeast and midwest. As analyzed above, in addition to Mn, the PTEs Pb, Cd and Cu all have a certain degree of accumulation. And the investigation found that there were many petroleum machinery manufacturing plants in the central and eastern part of the study area, therefore, the accumulation of Pb, Cd and Cu in the soil may be related to factors such as petroleum extraction, crude oil leakage and machinery manufacturing. The above analysis indicated that the influence of human activities is evident on the distribution of soil PTEs3, 23.Figure 3spatial distribution map of soil PTEs in the study area.Full size imagePotential ecological risk assessmentThe potential ecological risk assessment model after adjusting the threshold was used to evaluate the PER of the oil production plant. The individual potential ecological risk of PTEs was shown in Table 2. From Table 2, the average ({E}_{r}^{i}) values of PTEs were Cr  > Pb  > Cd  > Ni  > As  > Cu  > Zn  > Mn. The average ({E}_{r}^{i}) values of Cr and Pb were 79.62 and 63.64, respectively, reaching a relatively high level of potential ecological risk; the average ({E}_{r}^{i}) values of Cd and Ni were 55.95 and 37.91, respectively, which were at medium potential ecological risk level; the average ({E}_{r}^{i}) values of other PTEs were all lower than 30, with minor potential ecological risk. Specifically, all samples of Cu, Mn and Zn were at slight potential ecological risk level; 5% of As samples, 80% of Cd, 85% of Cr, 80% of Ni and 100% of Pb (Table S5) were at medium and above potential ecological risk. In particular, the potential ecological risks of 35% of Cd samples, 10% of Cr samples, 5% of Ni samples and 80% of Pb samples (Table S5) were relatively high, 10% Cd samples reached high potential ecological risk level, and 10% Cr samples had extremely high potential ecological risk. In summary, Geostatistical analysis shows that the hotspot distribution of all PTEs in the study area is almost related to the distribution of oil wells. In addition, the hotspot distribution of PTEs may also be related to factors such as agricultural and industrial activities3. The average value of PER in the study area was 265.08, and the proportions of the three risk levels of medium, slightly high and high were 5%, 75% and 20%, respectively (Table S5). It proved that the study area was at a higher potential ecological risk. Among them, the PER values of samples A, B, D, E, F, G, H, I and J (Table 2) were all greater than 280, reaching fairly high ecological risk.Table 2 Single ecological risk index and potential ecological risk of soil PTEs in study area.Full size tableHuman health risk assessmentThe non-carcinogenic risk assessment of As, Cd, Cr, Cu, Mn, Ni, Pb, Zn and Ba in the soils of the study area was carried out, and the assessment results were shown in Table 3. The THI values of children and adults under the three exposure routes of soil PTEs in the study area were 7.31 and 1.03, respectively, and the THI values were all  > 1, which indicated that soil PTEs around the oil production plants posed significant non-carcinogenic health risks to children and adults. The non-carcinogenic hazardous quotient (HQ) of children and adults in Table 3 revealed that the HQ of all PTEs for adults under each exposure route was less than 1, while the HQ of Cr and Pb for children under the oral intake route was greater than 1, which were 4.91 and 1.17, respectively. For HQ with different exposure routes of the same PTE, each soil PTE presented the risk of oral ingestion  > oral and nasal inhalation risk  > skin contact risk. The result was in agreement with the reports14, 37. Therefore, oral intake was the main exposure route of non-carcinogenic risk, and oral intake of Cr and Pb caused serious non-carcinogenic risk to children. Statistical analysis of HI for soil PTEs in the study area showed that the HI values of PTEs for children were significantly higher than those of adults, and the HI values of PTEs in children and adults were all Cr  > Pb  >   > As  > Ni  > Mn  > Ba  > Cu  > Zn  > Cd. Among them, the HI values of all PTEs for adults were less than 1, indicating that the non-carcinogenic risks caused by a single PTE did not have a significant impact on adults; while the HI values of Cr and Pb for children were 4.93 and 1.17 greater than 1, indicating that they have caused serious non-carcinogenic risk to local children. In addition, the HI values of As and Ni for children and the HI values of As, Cr and Pb for adults were all greater than 0.1, which requires attention. In summary, children suffered from significant non-carcinogenic risk, and adults suffered from minor non-carcinogenic risk in the study area; soil Cr and Pb were the most important non-carcinogenic risk factors for children and adults in the study area.Table 3 Non-cancer and cancer risk assessment of adults and children under different exposure routes.Full size tableIn this study, soil As, Cd, Cr, Ni and Pb from the study area were assessed for carcinogenic risk, and the results were shown in Table 3. The TCRI of children and adults under the three exposure routes of these five PTEs were 9.44E−04 and 5.75E−04, respectively, indicating that soil PTEs around the oil production plants have caused serious carcinogenic risk to local children and adults. The CR values of children and adults showed that the CR values of Cr (6.33E−04) and Ni (2.64E−04) for children, and Cr (3.87E−04) and Ni (1.49E−04) for adults were all greater than 10–4. In addition, As, Cr and Cd all presented oral intake risk  > oronasal inhalation risk  > skin contact risk. In conclusion, Cr and Ni caused serious carcinogenic risk for children and adults in the study area, and oral intake was also the primary way of carcinogenic risk. The CRI statistics of adults and children exhibited that the CRI values of all PTEs were lower than those of children. The CRI values of the PTEs in adults and children under the three exposure routes were Cr  > Ni  >   > As  > Pb  >   > Cd. Among them, the CRI values of Cr and Ni in children and adults by oral intake were both greater than 10–4, showing a strong carcinogenic risk. It is noteworthy that the assessment based on total concentrations of PTEs in soil might overestimate potential health risks38. The above analysis revealed that both children and adults in the study area suffered from serious carcinogenic risks, and Cr and Ni were the chiefly carcinogenic risk factors. More