More stories

  • in

    Vapour pressure deficit determines critical thresholds for global coffee production under climate change

    Vega, F. E., Rosenquist, E. & Collins, W. Global project needed to tackle coffee crisis. Nature 425, 343 (2003).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Craparo, A. C. W., Van Asten, P. J. A., Läderach, P., Jassogne, L. T. P. & Grab, S. W. Coffea arabica yields decline in Tanzania due to climate change: global implications. Agric. For. Meteorol. 207, 1–10 (2015).ADS 
    Article 

    Google Scholar 
    Davis, A. P. et al. High extinction risk for wild coffee species and implications for coffee sector sustainability. Sci. Adv. 5, eaav3473 (2019).ADS 
    PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Davis, A. P., Gole, T. W., Baena, S. & Moat, J. The impact of climate change on indigenous arabica coffee (Coffea arabica): predicting future trends and identifying priorities. PLoS ONE 7, e47981 (2012).ADS 
    CAS 
    PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Davis, A. P., Mieulet, D., Moat, J., Sarmu, D. & Haggar, J. Arabica-like flavour in a heat-tolerant wild coffee species. Nat. Plants 7, 413–418 (2021).CAS 
    PubMed 
    Article 

    Google Scholar 
    Moat, J., Gole, T. W. & Davis, A. P. Least concern to endangered: applying climate change projections profoundly influences the extinction risk assessment for wild Arabica coffee. Global Change Biol. 25, 390–403 (2019).ADS 
    Article 

    Google Scholar 
    Moat, J. et al. Resilience potential of the Ethiopian coffee sector under climate change. Nat. Plants 3, 17081 (2017).PubMed 
    Article 

    Google Scholar 
    Kath, J. et al. Not so robust: Robusta coffee production is highly sensitive to temperature. Global Change Biol. 26, 3677–3688 (2020).ADS 
    Article 

    Google Scholar 
    Liu, L. et al. Soil moisture dominates dryness stress on ecosystem production globally. Nat. Commun. 11, 1–9 (2020).ADS 
    CAS 

    Google Scholar 
    Grossiord, C. et al. Plant responses to rising vapor pressure deficit. New Phytol. 226, 1550–1566 (2020).PubMed 
    Article 

    Google Scholar 
    IPCC Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds. Pörtner, H.-O. et al.) (Cambridge Univ. Press, 2022).Burke, M. et al. Higher temperatures increase suicide rates in the United States and Mexico. Nat. Clim. Change 8, 723–729 (2018).ADS 
    Article 

    Google Scholar 
    Burke, M., Hsiang, S. M. & Miguel, E. Global non-linear effect of temperature on economic production. Nature 527, 235–239 (2015).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Duffy, K. A. et al. How close are we to the temperature tipping point of the terrestrial biosphere? Sci. Adv. 7, eaay1052 (2021).ADS 
    CAS 
    PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Scheffer, M. et al. Early-warning signals for critical transitions. Nature 461, 53–59 (2009).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Schneider, S. H. Abrupt non-linear climate change, irreversibility and surprise. Global Environ. Change 14, 245–258 (2004).Article 

    Google Scholar 
    Lenton, T. M. Early warning of climate tipping points. Nat. Clim. Change 1, 201–209 (2011).ADS 
    Article 

    Google Scholar 
    Lenton, T. M. et al. Climate tipping points—too risky to bet against. Nature. 575, 592–595 (2019).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Lobell, D. B., Bänziger, M., Magorokosho, C. & Vivek, B. Nonlinear heat effects on African maize as evidenced by historical yield trials. Nat. Clim. Change 1, 42–45 (2011).ADS 
    Article 

    Google Scholar 
    Lobell, D. B., Deines, J. M. & Tommaso, S. D. Changes in the drought sensitivity of US maize yields. Nat. Food 1, 729–735 (2020).Article 

    Google Scholar 
    Lobell, D. B. et al. Greater sensitivity to drought accompanies maize yield increase in the US Midwest. Science 344, 516–519 (2014).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Rigden, A., Mueller, N., Holbrook, N., Pillai, N. & Huybers, P. Combined influence of soil moisture and atmospheric evaporative demand is important for accurately predicting US maize yields. Nat. Food 1, 127–133 (2020).Article 

    Google Scholar 
    Schlenker, W. & Roberts, M. J. Nonlinear temperature effects indicate severe damages to US crop yields under climate change. Proc. Natl Acad. Sci. USA 106, 15594–15598 (2009).ADS 
    CAS 
    PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    McDowell, N. G. et al. Mechanisms of woody-plant mortality under rising drought, CO2 and vapour pressure deficit. Nat. Rev. Earth Environ. 3, 294–308 (2022).ADS 
    CAS 
    Article 

    Google Scholar 
    Sinclair, T. R. et al. Limited-transpiration response to high vapor pressure deficit in crop species. Plant Sci. 260, 109–118 (2017).CAS 
    PubMed 
    Article 

    Google Scholar 
    López, J., Way, D. A. & Sadok, W. Systemic effects of rising atmospheric vapor pressure deficit on plant physiology and productivity. Global Change Biol. 27, 1704–1720 (2021).ADS 
    Article 

    Google Scholar 
    McDowell, N. G. & Allen, C. D. Darcy’s law predicts widespread forest mortality under climate warming. Nat. Clim. Change 5, 669–672 (2015).ADS 
    Article 

    Google Scholar 
    Abatzoglou, J. T., Dobrowski, S. Z., Parks, S. A. & Hegewisch, K. C. TerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958–2015. Sci. Data 5, 170191 (2018).PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    You, L., Wood, S., Wood-Sichra, U. & Wu, W. Generating global crop distribution maps: from census to grid. Agric. Syst. 127, 53–60 (2014).Article 

    Google Scholar 
    Fong, Y., Huang, Y., Gilbert, P. B. & Permar, S. R. chngpt: threshold regression model estimation and inference. BMC Bioinformatics 18, 1–7 (2017).Article 

    Google Scholar 
    Qin, Y. et al. Agricultural risks from changing snowmelt. Nat. Clim. Change 10, 459–465 (2020).ADS 
    Article 

    Google Scholar 
    Forster, P. M., Maycock, A. C., McKenna, C. M. & Smith, C. J. Latest climate models confirm need for urgent mitigation. Nat. Clim. Change 10, 7–10 (2020).ADS 
    Article 

    Google Scholar 
    Forster, P. M. et al. Projections of when temperature change will exceed 2 °C above pre-industrial levels. Nat. Clim. Change 10, 407–412 (2011).
    Google Scholar 
    Joshi, M., Hawkins, E., Sutton, R., Lowe, J. & Frame, D. Projections of when temperature change will exceed 2 °C above pre-industrial levels. Nat. Clim. Change 1, 407–412 (2011).ADS 
    Article 

    Google Scholar 
    IPCC, 2021: Summary for Policymakers. In Climate Change 2021: The Physical Science Basis (eds Masson-Delmotte, V. et al.) (Cambridge Univ. Press, in press).Lobell, D. B. et al. The critical role of extreme heat for maize production in the United States. Nat. Clim. Change 3, 497–501 (2013).ADS 

    Google Scholar 
    Sinclair, T. R., Hammer, G. L. & Van Oosterom, E. J. Potential yield and water-use efficiency benefits in sorghum from limited maximum transpiration rate. Funct. Plant Biol. 32, 945–952 (2005).PubMed 
    Article 

    Google Scholar 
    Martins, M. Q. et al. Protective response mechanisms to heat stress in interaction with high [CO2] conditions in Coffea spp. Front. Plant Sci. 7, 947 (2016).PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Rodrigues, W. P. et al. Long‐term elevated air [CO2] strengthens photosynthetic functioning and mitigates the impact of supra‐optimal temperatures in tropical Coffea arabica and C. canephora species. Global Change Biol. 22, 415–431 (2016).ADS 
    Article 

    Google Scholar 
    Ghini, R. et al. Coffee growth, pest and yield responses to free-air CO2 enrichment. Clim. Change 132, 307–320 (2015).ADS 
    Article 

    Google Scholar 
    Rakocevic, M. et al. The vegetative growth assists to reproductive responses of Arabic coffee trees in a long-term FACE experiment. Plant Growth Regul. 91, 305–316 (2020).CAS 
    Article 

    Google Scholar 
    Hammer, G. L. et al. Designing crops for adaptation to the drought and high‐temperature risks anticipated in future climates. Crop Sci. 60, 605–621 (2020).Article 

    Google Scholar 
    Gennari, P., Rosero-Moncayo, J. & Tubiello, F. N. The FAO contribution to monitoring SDGs for food and agriculture. Nat. Plants 5, 1196–1197 (2019).PubMed 
    Article 

    Google Scholar 
    Lesk, C., Rowhani, P. & Ramankutty, N. Influence of extreme weather disasters on global crop production. Nature 529, 84–87 (2016).ADS 
    CAS 
    PubMed 
    Article 

    Google Scholar 
    Ortiz-Bobea, A., Ault, T. R., Carrillo, C. M., Chambers, R. G. & Lobell, D. B. Anthropogenic climate change has slowed global agricultural productivity growth. Nat. Clim. Change 11, 306–312 (2021).ADS 
    Article 

    Google Scholar 
    Davis, A. P. et al. Hot coffee: the identity, climate profiles, agronomy, and beverage characteristics of Coffea racemosa and C. zanguebariae. Front. Sustain. Food Syst. 5, 740137 (2021).Article 

    Google Scholar 
    Sarmiento-Soler, A. et al. Disentangling effects of altitude and shade cover on coffee fruit dynamics and vegetative growth in smallholder coffee systems. Agric. Ecosyst. Environ. 326, 107786 (2022).CAS 
    Article 

    Google Scholar 
    Wood, S. N. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. B 73, 3–36 (2011).MathSciNet 
    MATH 
    Article 

    Google Scholar 
    Barton, K. MuMIn: multi-model inference. R-Forge http://r-forge.r-project.org/projects/mumin/ (2009).R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing https://www.r-project.org/ (2021).Harrison, X. A. et al. A brief introduction to mixed effects modelling and multi-model inference in ecology. PeerJ 6, e4794 (2018).PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Najafi, E., Devineni, N., Khanbilvardi, R. M. & Kogan, F. Understanding the changes in global crop yields through changes in climate and technology. Earths Future 6, 410–427 (2018).ADS 
    CAS 
    Article 

    Google Scholar 
    Ovalle-Rivera, O. et al. Assessing the accuracy and robustness of a process-based model for coffee agroforestry systems in Central America. Agrofor. Syst. 94, 2033–2051 (2020).Article 

    Google Scholar 
    Varma, S. & Simon, R. Bias in error estimation when using cross-validation for model selection. BMC Bioinformatics 7, 1–8 (2006).Article 

    Google Scholar 
    Yuan, W. et al. Increased atmospheric vapor pressure deficit reduces global vegetation growth. Sci. Adv. 5, eaax1396 (2019).ADS 
    CAS 
    PubMed 
    PubMed Central 
    Article 

    Google Scholar 
    Son, H. & Fong, Y. Fast grid search and bootstrap-based inference for continuous two-phase polynomial regression models. Environmetrics 32, e2664 (2021).MathSciNet 
    Article 

    Google Scholar 
    Wintgens, J. N. et al. Coffee: Growing, Processing, Sustainable Production. A Guidebook for Growers, Processors, Traders, and Researchers (Wiley, 2004). More

  • in

    Brain de novo transcriptome assembly of a toad species showing polymorphic anti-predatory behavior

    Sample collection and RNA preparationWe analyzed 6 adult yellow-bellied toad individuals representative of distinct behavioral profiles, i.e. prolonged unken-reflex display vs no unken-reflex display (thereafter referred as “ + ” and “-“, respectively). Behavioral profiles were scored as in Chiocchio et al.12: 3 toads showed prolonged unken-reflex (+), whereas the other 3 did not show unken-reflex (−), as reported in Table 1. Sampling procedures were approved by the Italian Ministry of Ecological Transition and the Italian National Institute for Environmental Protection and Research (ISPRA; permit number: 20824, 18-03-2020). After dissection, brain tissue was immediately stored in RNAprotect Tissue Reagent (Quiagen) until RNA extraction. RNA extractions were performed using the RNeasy Plus Kit (Quiagen), according to the manufacturer’ instructions. RNA quality and concentration were assessed by means of both a spectrophotometer and a Bioanalyzer (Agilent Cary60 UV-vis and Agilent 2100, respectively – Agilent Technologies, Santa Clara, USA).Table 1 Summary of the 6 libraries deposited in the Sequence Read Archive (SRA) of NCBI, in terms of number of raw and trimmed reads per sample.Full size tableLibrary preparation and sequencingLibrary preparation and RNA sequencing were performed by NOVOGENE (UK) COMPANY LIMITED using Illumina NovaSeq platform. Library construction was carried out using the NEBNext® Ultra ™ RNA Library Prep Kit for Illumina®, following manufacturer instructions. Briefly, after the quality control check, the mRNA sample was isolated from the total RNA by using magnetic beads made of oligos d(T)25 (i.e. polyA-tail mRNA enrichment). Subsequently, mRNA was randomly fragmented, and a cDNA synthesis step proceeded using random hexamers and the reverse transcriptase enzyme. Once the synthesis of the first chain has finished, the second chain was synthesized with the addition of the Illumina buffer, dNTPs, RNase H and polymerase I of E.coli, by means of the Nick translation method. Then, the resulting products went through purification, repair, A-tailing and adapter ligation. Fragments of the appropriate size were enriched by PCR, the indexed P5 and P7 primers were introduced, and the final products were purified. Finally, the Illumina Novaseq 6000 sequencing system was used to sequence the libraries, through a paired-end 150 bp (PE150) strategy. We obtained on average 52.7 million reads for each library. The sequencing data are available at the NCBI Sequence Read Archive (project ID PRJNA76401320).Pre-assembly processing stageA total of 316,329,573 pairs of reads was generated by Illumina sequencing. All of them went to a cleaning analytic step. The quality of the raw reads was assessed with the FastQC 0.11.5 tool (http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc), in order to estimate the RNAseq quality profiles. The quality estimators were generated for both the raw and trimmed data. The quality assessment metrics for trimmed data were aggregated across all samples into a single report for a summary visualization with MultiQC software tool21 v.1.9 (see Fig. 1). To remove low quality bases and adapter sequences, raw reads were also analyzed through a quality trimming step with Trimmomatic22, v.0.39 (setting the option SLIDINGWINDOW: 4: 15, MINLEN: 36, and HEADCROP: 13). All the unpaired reads were discarded. After the cleaning step and removal of low-quality reads, 297,354,405 clean reads (i.e. 94% of raw reads) were maintained for building the de novo transcriptome assembly (see Table 1).Fig. 1The cleaned reads from all samples were assessed with FastQC and visualized with MultiQC. (a) Read count distribution for mean sequence quality. (b) Mean quality scores distribution. (c) Read length distribution. (d) Per Sequence GC Content.Full size image
    De novo transcriptome assembly and quality assessmentAs there is no reference genome for B. pachypus, we performed a de novo transcriptome assembly procedure. The workflow of the bioinformatic pipelines is shown in Fig. 2. All the described bioinformatics analyses were performed on the high-performance computing systems provided by ELIXIR-IT HPC@CINECA23.Fig. 2Workflow of the bioinformatic pipeline, from raw input data to annotated contigs, for the de novo transcriptome assembly of B. pachypus.Full size imageTo construct an optimized de novo transcriptome, avoiding chimeric transcripts, we employed rnaSPAdes24, a tool for de novo transcriptome assembly from RNA-Seq data implemented in the SPAdes v.3.14.1 package. rnaSPAdes automatically detected two k-mer sizes, approximately one third and half of the maximal read length (the two detected k-mer sizes were 45 and 67 nucleotides, respectively). At this stage, a total of 1,118,671 assembled transcripts were generated by rnaSPAdes runs, with an average length of 689.41 bp and an N50 of 1474 bp (Table 2).Table 2 Similarity rate of newly assembled transcripts versus the de novo transcriptome of B. pachypus.Full size tableResults from the assembly procedures were validated through three independent validator algorithms implemented in BUSCO25 v.4.1.4, DETONATE26 v.1.11 and TransRate27 v.1.0.3. These tools generate several metrics used as a guide to evaluate error sources in the assembly process and provide evidence about the quality of the assembled transcriptome. Busco provides a quantitative measure of transcriptome quality and completeness, based on evolutionarily-informed expectations of gene content from the near-universal, ultra-conserved eukaryotic proteins (eukaryota_odb9) database. Detonate (DE novo TranscriptOme rNa-seq Assembly with or without the Truth Evaluation) is a reference-free evaluation method based on a novel probabilistic model that depends only on the assembly and the RNA-Seq reads used to construct it. Transrate generates standard metrics and remapping statistics. No reference protein sequences were used for the assessment with Transrate. The main metrics resulted from the assembly validators are shown in Table 2 (“Before CD-HIT-est” column). After this triple assessment validation step, the result of the assembly procedure become the input for the CD-HIT-est v.4.8.128 program, a hierarchical clustering tool used to avoid redundant transcripts and fragmented assemblies common in the process of de novo assembly, providing unique genes. CD-HIT-est was run using the default parameters, corresponding to a similarity of 95%. Subsequently, a second validation step was launched on the CD-HIT-est output file. To refine the final transcriptome dataset, a further hierarchical clustering step was performed by running CORSET v1.0629. Then, the output of CORSET was validated by BUSCO, and quality assessment was performed with HISAT230,31 by mapping the trimmed reads to the reference transcriptome (unigenes). Results from all validation steps are shown in Table 2 and discussed in the “Technical Validation” paragraph.Finally, the CORSET output was run on TransDecoder32,33, the current standard tool that identifies long open read frames (ORFs) in assembled transcripts, using default parameters. TransDecoder by default performs ORF prediction on both strands of assembled transcripts regardless of the sequenced library. It also ranks ORFs based on their completeness, and determines if the 5 ‘end is incomplete by looking for any length of AA codons upstream of a start codon (M) without a stop codon. We adopted the “Longest ORF” rule and selected the highest 5 AUG (relative to the inframe stop codon) as the translation start site.Transcriptome annotationWe employed different kinds of annotations for the de novo assembly. We introduced DIAMOND34, an open-source algorithm based on double indexing that is 20,000 times faster than BLASTX on short reads and has a similar degree of sensitivity. Like BLASTX, DIAMOND attempts to determine exhaustively all significant alignments for a given query. Most sequence comparison programs, including BLASTX, follow the seed-and-extend paradigm. In this two-phase approach, users search first for matches of seeds (short stretches of the query sequence) in the reference database, and this is followed by an ‘extend’ phase that aims to compute a full alignment. The following parameter settings were applied: DIAMOND-fast DIAMOND BLASTX-t 48 -k 250 -min-score 40; DIAMOND-sensitive: DIAMOND BLASTX -t 48 -k 250 -sensitive -min-score 40.Contigs were aligned with DIAMOND on Nr, SwissProt and TrEMBL to retrieve the corresponding best annotations. An annotation matrix was then generated by selecting the best hit for each database. Following the analysis of BLASTX against Nr, SwissProt and TremBL, we obtained respectively: 123,086 (64.57%), 77,736 (40.78%), 122,907 (64.48%) contigs. The results obtained following the analysis with BLASTP against Nr, SwissProt and TrEMBL were 96,321 (50.53%), 57,877 (30.36%) and 97,256 (51.02%) contigs respectively. All the information on the resulting datasets is resumed in Table 3.Table 3 Summary of homology annotation hits on the different databases queried in this study.Full size tableThe output obtained by the BLASTX annotation consisted in a total of 77391 sequences simultaneously mapped on the three queried databases (i.e., Nr, SwissProt and TrEMBL). The output obtained following the BLASTP annotation consisted in a total of 57704 sequences simultaneously mapped on the three databases. Venn diagrams are presented in Fig. 3, showing the redundancy of the annotations in the different databases for both DIAMOND BLASTX (Fig. 3a) and DIAMOND BLASTP (Fig. 3b). Furthermore, the ten most represented species and the ten hits of the gene product obtained respectively with BLASTX and BLASTP by mapping the transcripts against the reference database Nr are shown in Figs. 4 and 5. Since BLASTX translated nucleotide sequence searches against protein sequences the BLASTX results are more exhaustive than BLASTP results. Contigs were also processed with InterProScan35 to detect InterProScan signatures. The InterPro database (http://www.ebi.ac.uk/interpro/) integrates together predictive models or ‘signatures’ representing protein domains, families and functional sites from multiple, diverse source databases: Gene3D, PANTHER, Pfam, PIRSF, PRINTS, ProDom, PROSITE, SMART, SUPERFAMILY and TIGRFAMs. The obtained InterProScan results for all the unigenes are available on Figshare in the form of Tab Separated Values (tsv) file format, which includes the GO and KEGG annotated contigs, respectively.Fig. 3Venn diagrams for the number of contigs annotated with DIAMOND (BLASTX (a) and BLASTP (b) functions) against the three databases: Nr, SwissProt, TREMBL.Full size imageFig. 4Most represented species and gene product hits. Top 10 best species (a) and protein (b) hits present in the reference database (Nr, BLASTX).Full size imageFig. 5Most represented species and gene product hits. Top 10 best species (a) and protein (b) hits present in the reference database (Nr, BLASTP).Full size imageComparison with Bombina orientalis brain transcriptomeWe compared the brain de novo transcriptome of B. pachypus with the brain de novo transcriptome of B. orientalis, recently produced in the frame of a prey-catching conditioning experiment17,18. The B. orientalis transcriptome resource was downloaded from GEO archive of NCBI (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE171766). To make the datasets comparable, we first performed ORF prediction on B. orientalis trascriptome using Transdecoder, using default settings. Results from the B. orientalis trascriptome ORF prediction are available in Figshare at the following link https://doi.org/10.6084/m9.figshare.20319633/). We also applied the makedb function implemented in DIAMOND to create the protein database index. Then, we aligned the B. pachypus predicted coding sequences and proteins (query files) against the B. orientalis protein database (reference) using DIAMOND BLASTX and BLASTP, respectively. We obtained 167041 matches from BLASTX and 156248 matches for BLASTP. Results from the BLASTX and BLASTP comparisons, and the most matched proteins, are available on Figshare36 (link available in next paragraph). More

  • in

    Low functional vulnerability of fish assemblages to coral loss in Southwestern Atlantic marginal reefs

    Birkeland, C. Coral Reefs in the Anthropocene (Springer, 2015).Book 

    Google Scholar 
    Kleypas, J. A., Mcmanus, J. W. & Meñez, L. A. B. Environmental limits to coral reef development: Where do we draw the line?. Am. Zool. 39(1), 146–159. https://doi.org/10.1093/icb/39.1.146 (1999).Article 

    Google Scholar 
    Perry, C. T. & Larcombe, P. Marginal and non-reef-building coral environments. Coral Reefs 22, 427–432. https://doi.org/10.1007/s00338-003-0330-5 (2003).Article 

    Google Scholar 
    Wilkinson, C. R. Global and local threats to coral reef functioning and existence: review and predictions. Mar. Freshw. Res. 50, 867–878. https://doi.org/10.1071/mf99121 (1999).Article 

    Google Scholar 
    Mies, M. et al. South atlantic coral reefs are major global warming refugia and less susceptible to bleaching. Front. Mar. Sci. 7, 514. https://doi.org/10.3389/fmars.2020.00514 (2020).Article 

    Google Scholar 
    Leão, Z. M. A. N. et al. Brazilian coral reefsin a period of global change: A synthesis. Braz. J. Oceanogr. 64, 97–116. https://doi.org/10.1590/S1679-875920160916064sp2 (2016).Article 

    Google Scholar 
    Coker, D. J., Wilson, S. K. & Pratchett, M. S. Importance of live coral habitat for reef fishes. Rev. Fish Biol. Fish. 24, 89–126. https://doi.org/10.1007/s11160-013-9319-5 (2014).Article 

    Google Scholar 
    Alvarez-Filip, L., Gill, J. A. & Dulvy, N. K. Complex reef architecture supports more small-bodied fishes and longer food chains on Caribbean reefs. Ecosphere 2, 118. https://doi.org/10.1890/ES11-00185.1 (2011).Article 

    Google Scholar 
    Wilson, S. K., Graham, N. A. J., Pratchett, M. S., Jones, G. P. & Polunin, N. V. C. Multiple disturbances and the global degradation of coral reefs: Are reef fishes at risk or resilient?. Glob. Change Biol. 12, 2220–2234. https://doi.org/10.1111/j.1365-2486.2006.01252.x (2006).ADS 
    Article 

    Google Scholar 
    Sully, S., Burkepile, D. E., Donovan, M. K., Hodgson, G. & van Woesik, R. A global analysis of coral bleaching over the past two decades. Nat. Commun. 10, 1264. https://doi.org/10.1038/s41467-019-09238-2 (2019).ADS 
    CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Bellwood, D. R., Hughes, T. P., Folke, C. & Nystrom, M. Confronting the coral reef crisis. Nature 429, 827–833. https://doi.org/10.1038/nature02691 (2004).ADS 
    CAS 
    Article 
    PubMed 

    Google Scholar 
    Hughes, T. P. et al. climate change, human impacts, and the resilience of coral reefs. Science 301, 929–933. https://doi.org/10.1126/science.1085046 (2003).ADS 
    CAS 
    Article 
    PubMed 

    Google Scholar 
    Holbrook, N. J. et al. Keeping pace with marine heatwaves. Nat. Rev. Earth Environ. 1, 482–493. https://doi.org/10.1038/s43017-020-0068-4 (2020).ADS 
    Article 

    Google Scholar 
    Bleuel, J., Pennino, M. G. & Longo, G. O. Coral distribution and bleaching vulnerability areas in Southwestern Atlantic under ocean warming. Sci. Rep. 11, 12833. https://doi.org/10.1038/s41598-021-92202-2 (2021).ADS 
    CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Fontoura, L. et al. The macroecology of reef fish agonistic behaviour. Ecography 43, 1278–1290. https://doi.org/10.1111/ecog.05079 (2020).Article 

    Google Scholar 
    Inagaki, K. Y., Pennino, M. G., Floeter, S. R., Hay, M. E. & Longo, G. O. Trophic interactions will expand geographically but be less intense as oceans warm. Glob. Change Biol. 26, 6805–6812. https://doi.org/10.1111/gcb.15346 (2020).ADS 
    Article 

    Google Scholar 
    Longo, G. O., Hay, M. E., Ferreira, C. E. L. & Floeter, S. R. Trophic interactions across 61 degrees of latitude in the Western Atlantic. Glob. Ecol. Biogeogr. 28, 107–117. https://doi.org/10.1111/geb.12806 (2019).Article 

    Google Scholar 
    Pratchett, M. S. et al. Effects of climate-induced coral bleaching on coral-reef fishes: Ecological and economic consequences. Oceanogr. Mar. Biol. Annu. Rev. 46, 251–296. https://doi.org/10.1201/9781420065756.ch6 (2008).Article 

    Google Scholar 
    Graham, N. A. J. et al. Lag effects in the impacts of mass coral bleaching on coral reef fish, fisheries, and ecosystems. Conserv. Biol. 21, 1291–1300. https://doi.org/10.1111/j.1523-1739.2007.00754.x (2007).Article 
    PubMed 

    Google Scholar 
    Strona, G. et al. Global tropical reef fish richness could decline by around half if corals are lost. Proc. R. Soc. B 288, 20210274. https://doi.org/10.1098/rspb.2021.0274 (2021).CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    McClenachan, L. Extinction risk in reef fishes 199–207 (Cambridge University Press, 2015).
    Google Scholar 
    Power, M. E. et al. Challenges in the quest for keystones. Bioscience 46, 609–620. https://doi.org/10.2307/1312990 (1996).Article 

    Google Scholar 
    Pereira, P. H. C. et al. The influence of multiple factors upon reef fish abundance and species richness in a tropical coral complex. Ichthyol. Res. 61, 375–384. https://doi.org/10.1007/s10228-014-0409-8 (2014).Article 

    Google Scholar 
    Coni, E. O. C. et al. An evaluation of the use of branching fire-corals (Millepora spp.) as refuge by reef fish in the Abrolhos Bank, eastern Brazil. Environ. Biol. Fish. 96, 45–55. https://doi.org/10.1007/s10641-012-0021-6 (2013).Article 

    Google Scholar 
    Graham, N. A. J. et al. Extinction vulnerability of coral reef fishes. Ecol. Lett. 14, 341–348. https://doi.org/10.1111/j.1461-0248.2011.01592.x (2011).Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Cornwell, W. K., Schwilk, D. W. & Ackerly, D. D. A trait-based test for habitat filtering: convex hull volume. Ecology 87, 1465–1471. https://doi.org/10.1890/0012-9658(2006)87[1465:ATTFHF]2.0.CO;2 (2006).Article 
    PubMed 

    Google Scholar 
    Mouillot, D., Graham, N. A. J., Villéger, S., Mason, N. W. H. & Bellwood, D. R. A functional approach reveals community responses to disturbances. Trends Ecol. Evol. 28(3), 167–177. https://doi.org/10.1016/j.tree.2012.10.004 (2013).Article 
    PubMed 

    Google Scholar 
    Pimiento, C. et al. Functional diversity of marine megafauna in the Anthropocene. Sci. Adv. 6, 7650. https://doi.org/10.1126/sciadv.aay7650 (2020).ADS 
    Article 

    Google Scholar 
    Loiola, M. et al. Structure of marginal coral reef assemblages under different turbidity regime. Mar. Environ. Res. 147, 138–148. https://doi.org/10.1016/j.marenvres.2019.03.013 (2019).CAS 
    Article 
    PubMed 

    Google Scholar 
    Aued, A. W. et al. Large-scale patterns of benthic marine communities in the Brazilian Province. PLoS ONE 13, e0198452. https://doi.org/10.1371/journal.pone.0198452 (2018).CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Leão, Z. M. A. N., Kikuchi, R. K. P. & Testa, V. Corals and Coral Reefs of Brazil 9–52 (Elsevier Publisher, 2003).
    Google Scholar 
    Pinheiro, H. T. et al. South-western Atlantic reef fishes: Zoogeographical patterns and ecological drivers reveal a secondary biodiversity centre in the Atlantic Ocean. Divers. Distrib. 24, 951–965. https://doi.org/10.1111/ddi.12729 (2018).Article 

    Google Scholar 
    Floeter, S. R. et al. Atlantic reef fish biogeography and evolution. J. Biogeogr. 35, 22–47. https://doi.org/10.1111/j.1365-2699.2007.01790.x (2008).Article 

    Google Scholar 
    Cord, I. et al. Brazilian marine biogeography: A multi-taxa approach for outlining sectorization. Mar. Biol. 169(5), 61. https://doi.org/10.1007/s00227-022-04045-8 (2022).Article 

    Google Scholar 
    Leal, I. C. S., Araújo, M. E. D., Cunha, S. R. D. & Pereira, P. H. C. The influence of fire-coral colony size and agonistic behaviour of territorial damselfish on associated coral reef fish communities. Mar. Environ. Res. 108, 45–54. https://doi.org/10.1016/j.marenvres.2015.04.009 (2015).CAS 
    Article 
    PubMed 

    Google Scholar 
    Kéry, M. & Royle, J. A. Applied hierarchical modeling in ecology: Analysis of distribution abundance and species richness in R and BUGS. In Prelude and Static Models Vol. 1 (eds Kéry, M. & Royle, J. A.) (Academic Press, 2016).MATH 

    Google Scholar 
    Hadj-Hammou, J., Mouillot, D. & Graham, N. A. J. Response and effect traits of coral reef fish. Front. Mar. Sci. https://doi.org/10.3389/fmars.2021.640619 (2021).Article 

    Google Scholar 
    McLean, M. et al. Trait similarity in reef fish faunas across the world’s oceans. PNAS 118(12), e2012318118. https://doi.org/10.1073/pnas.2012318118 (2021).CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Brandl, S. J. et al. Coral reef ecosystem functioning: eight core processes and the role of biodiversity. Front. Ecol. Environ. 17, 445–454. https://doi.org/10.1002/fee.2088 (2019).Article 

    Google Scholar 
    Eggertsen, L. et al. Seaweed beds support more juvenile reef fish than seagrass beds in a south-western Atlantic tropical seascape. Estuar. Coast. Shelf S. 196, 97–108. https://doi.org/10.1016/j.ecss.2017.06.041 (2017).ADS 
    Article 

    Google Scholar 
    Mouillot, D. et al. Functional over-redundancy and high functional vulnerability in global fish faunas on tropical reefs. PNAS 111, 13757–13762. https://doi.org/10.1073/pnas.1317625111 (2014).ADS 
    CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Briggs, J. C. Marine Zoogeography (McGraw-Hill, 1974).
    Google Scholar 
    Garcia, G. S., Dias, M. S. & Longo, G. O. Trade-off between number and length of remote videos for rapid assessments of reef fish assemblages. J. Fish Biol. 99(3), 896–904. https://doi.org/10.1111/jfb.14776 (2021).Article 
    PubMed 

    Google Scholar 
    Quimbayo, J. P. et al. Life-history traits, geographical range, and conservation aspects ofreef fishes from the Atlantic and Eastern Pacific. Ecology 102, e03298. https://doi.org/10.1002/ecy.3298 (2021).Article 
    PubMed 

    Google Scholar 
    Katsanevakis, S. et al. Monitoring marine populations and communities: methods dealing with imperfect detectability. Aquat. Biol. 16, 31–52. https://doi.org/10.3354/ab00426 (2012).Article 

    Google Scholar 
    Villéger, S., Mason, N. W. H. & Mouillot, D. New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology 89, 2290–2301. https://doi.org/10.1890/07-1206.1 (2008).Article 
    PubMed 

    Google Scholar 
    Maire, E., Grenouillet, G., Brosse, S. & Villéger, S. How many dimensions are needed to accurately assess functional diversity? A pragmatic approach for assessing the quality of functional spaces. Glob. Ecol. Biogeogr. 24, 728–740. https://doi.org/10.1111/geb.12299 (2015).Article 

    Google Scholar 
    R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2021)Kellner, K. jagsUI: A Wrapper Around ‘rjags’ to Streamline ‘JAGS’ Analyses. R package version 1.5.2. https://CRAN.R-project.org/package=jagsUI (2021)Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer-Verlag, 2016).Book 

    Google Scholar 
    Ferreira, C. E. L., Gonçalves, J. E. A. & Coutinho, R. Community structure of fishes and habitat complexity on a tropical rocky shore. Environ. Biol. Fish. 61, 353–369 (2001).Article 

    Google Scholar 
    Fulton, C. J. et al. Macroalgal meadow habitats support fish and fisheries in diverse tropical seascapes. Fish Fish. 21, 700–717. https://doi.org/10.1111/faf.12455 (2020).Article 

    Google Scholar 
    Ferreira, L. C. L. et al. Different responses of massive and branching corals to a major heatwave at the largest and richest reef complex in South Atlantic. Mar. Biol. 168, 54. https://doi.org/10.1007/s00227-021-03863-6 (2021).CAS 
    Article 

    Google Scholar 
    Lonzetti, B. C., Vieira, E. A. & Longo, G. O. Ocean warming can help zoanthids outcompete branching hydrocorals. Coral Reefs 41, 175–189. https://doi.org/10.1007/s00338-021-02212-9 (2022).Article 

    Google Scholar 
    Grillo, A. C., Candido, C. F., Giglio, V. J. & Longo, G. O. Unusual high coral cover in a Southwestern Atlantic subtropical reef. Mar. Biodivers. 51, 77. https://doi.org/10.1007/s12526-021-01221-9 (2021).Article 

    Google Scholar 
    Matheus, Z. et al. Benthic reef assemblages of the Fernando de Noronha Archipelago, tropical South-west Atlantic: Effects of depth, wave exposure and cross-shelf positioning. PLoS ONE 14(1), e0210664. https://doi.org/10.1371/journal.pone.0210664 (2019).CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Meirelles, P. M. et al. Baseline assessment of mesophotic reefs of the vitória-trindade seamount chain based on water quality, microbial diversity, benthic cover and fish biomass data. PLoS ONE 10(6), e0130084. https://doi.org/10.1371/journal.pone.0130084 (2015).CAS 
    Article 
    PubMed 
    PubMed Central 

    Google Scholar 
    Ferreira, C. E. L., Floeter, S. R., Gasparini, J. L., Ferreira, B. P. & Joyeux, J. C. Trophic structure patterns of Brazilian reef fishes: A latitudinal comparison. J. Biogeogr. 31, 1093–1106. https://doi.org/10.1111/j.1365-2699.2004.01044.x (2004).Article 

    Google Scholar 
    Fontoura, L. et al. Climate-driven shift in coral morphological structure predicts decline of juvenile reef fishes. Glob. Change Biol. 26, 557–567. https://doi.org/10.1111/gcb.14911 (2020).ADS 
    Article 

    Google Scholar 
    MacNeil, M. A. et al. Accounting for detectability in reef-fish biodiversity estimates. Mar. Ecol.-Prog. Ser. 367, 249–260. https://doi.org/10.3354/meps07580 (2008).ADS 
    Article 

    Google Scholar 
    Capitani, L., de Araujo, J. N., Vieira, E. A., Angelini, R. & Longo, G. O. Ocean warming will reduce standing biomass in a Tropical Western Atlantic reef ecosystem. Ecosystems 25, 843–857. https://doi.org/10.1007/s10021-021-00691-z (2022).Article 

    Google Scholar 
    Fogliarini, C. O., Longo, G. O., Francini-Filho, R. B., McClenachan, L. & Bender, M. G. Sailing into the past: Nautical charts reveal changes over 160 years in the largest reef complex in the South Atlantic Ocean. PECON 20(3), 231–239. https://doi.org/10.1007/10.1016/j.pecon.2022.05.003 (2022).Article 

    Google Scholar 
    Gasparini, J. L., Floeter, S. R., Ferreira, C. E. L. & Sazima, I. Marine ornamental trade in Brazil. Biodivers. Conserv. 14, 2883–2899. https://doi.org/10.1007/s10531-004-0222-1 (2005).Article 

    Google Scholar 
    Francini-Filho, R. B. et al. Brazil 163–198 (Springer, 2019).
    Google Scholar 
    Bellwood, D. R., Goatley, C. H. R. & Bellwood, O. The evolution of fishes and corals on reefs: Form, function and interdependence. Biol. Rev. 92, 878–901. https://doi.org/10.1111/brv.12259 (2017).Article 
    PubMed 

    Google Scholar 
    Nunes, L. T. et al. Ecology of Prognathodes obliquus, a butterflyfish endemic to mesophotic ecosystems of St. Peter and St. Paul’s Archipelago. Coral Reefs 38, 955–960. https://doi.org/10.1007/s00338-019-01822-8 (2019).ADS 
    Article 

    Google Scholar 
    Liedke, A. et al. Abundance, diet, foraging and nutritional condition of the banded butterflyfish (Chaetodon striatus) along the western Atlantic. Mar. Biol. 163, 6. https://doi.org/10.1007/s00227-015-2788-4 (2016).CAS 
    Article 

    Google Scholar  More

  • in

    Spatially structured eco-evolutionary dynamics in a host-pathogen interaction render isolated populations vulnerable to disease

    The pathosystemPlantago lanceolata L. is a perennial monoecious ribwort plantain that reproduces both clonally via the production side rosettes, and sexually via wind pollination. Seeds drop close to the mother plant and usually form a long-term seed bank47. Podospharea plantaginis (Castagne; U. Braun and S. Takamatsu) (Erysiphales, Ascomycota) is an obligate biotrophic powdery mildew that infects only P. lanceolata and requires living host tissue through its life cycle48. It completes its life cycle as localized lesions on host leaves, only the haustorial feeding roots penetrating the leaf tissue to feed nutrients from its host. Infection causes significant stress for host plant and may increase the host mortality31. The interaction between P. lanceolata and P. plantaginis is strain-specific, whereby the same host genotype may be susceptible to some pathogen genotypes while being resistant to others49. The putative resistance mechanism includes two steps. First, resistance occurs when the host plant first recognizes the attacking pathogen and blocks its growth. When the first step fails and infection takes place, the host may mitigate infection development. Both resistance traits vary among host genotypes49.Approximately 4000 P. lanceolata populations form a network covering an area of 50 × 70 km in the Åland Islands, SW of Finland. Disease incidence (0/1) in these populations has been recorded systematically every year in early September since 2001 by approximately 40 field assistants, who record the occurrence of the fungus P. plantaginis in the local P. lanceolata populations30. At this time, disease symptoms are conspicuous as infected plants are covered by white mycelia and conidia. The coverage (m2) of P. lanceolata in the meadows was recorded between 2001 and 2008 and is used as an estimate of host population size. In the field survey two technicians estimate Plantago population size by visually estimating how much ground/other vegetation P. lanceolata foliage covers (m2) in each meadow. The proportion of P. lanceolata plants in each population suffering from drought is also estimated annually in the survey. Data on average rainfall (mm) in July and August was estimated separately for each population using detailed radar-measured rainfall (obtained by Finnish Meteorological Institute) and it was available for years 2001–2008.Host population connectivity (SH)27 for each local population i was computed with the formula that takes into account the area of host coverage (m²) of all host populations surveyed, denoted with (Aj), and their spatial location compared to other host populations. We assume that the distribution of dispersal distances from a location are described by negative exponential distribution. Under this assumption, the following formula (1) quantifies for a focal population i, the effect of all other host populations, taking into account their population sizes and how strongly they are connected through immigration to it:$${S}_{i}^{H}=mathop{sum}limits_{jne i}{{{{{rm{e}}}}}}^{{-alpha d}_{{ij}}}sqrt{{A}_{j}}.$$
    (1)
    here, dij is the Euclidian distance between populations i and j and 1/α equals the mean dispersal distance, which was set to be two kilometres based on results from a previous study16.The annual survey data has demonstrated that P. plantaginis infects annually 2–16% of all host populations and persists as a highly dynamic metapopulation through extinctions and re-colonizations of local populations16. The number of host populations has remained relatively stable over the study period49. The first visible symptoms of P. plantaginis infection appear in late June as white-greyish lesions consisting of mycelium supporting the dispersal spores (conidia) that are carried by wind to the same or new host plants. Six to eight clonally produced generations follow one another in rapid succession, often leading to local epidemic with substantial proportion of the infected hosts by late summer within the host local population. Podosphaera plantaginis produces resting structures, chasmothecia, that appear towards the end of growing season in August–September31. Between 20% and 90% of the local pathogen populations go extinct during the winter, and thus the recolonization events play an important role in the persistence of the pathogen regionally16.Inoculation assay: Effect of connectivity and disease history on phenotypic disease resistanceHost and pathogen material for the experimentTo examine whether the diversity and level of resistance vary among host populations depending on their degree of connectivity (SH) and disease history, we selected 20 P. lanceolata populations for an inoculation assay. These populations occur in different locations in the host network, and were selected based on their connectivity values (S H of selected populations was 37–110 in isolated and 237–336 in highly connected category, Fig. 1). We did not include host populations in the intermediate connectivity category that was used in the population dynamic analyses in the inoculation assay due to logistic constraints. Podosphaera plantaginis is an obligate biotrophic pathogen that requires living host tissue throughout its life cycle, and obtaining sufficient inoculum for experiments is extremely time and space consuming. In both isolated and highly connected categories, half of the populations (IDs 193, 260, 311, 313, 337, 507, 1821, 1999, 2818 and 5206) were healthy during the study years 2001–2014, while half of the populations (IDs 271, 294, 309, 321, 490, 609, 1553, 1556, 1676 and 1847) were infected by P. plantaginis for several years during the same period. We collected P. lanceolata seeds from randomly selected ten individual plants around the patch area from each host population in August 2014.To acquire inoculum for the assay, we collected the pathogen strains as infected leaves, one leaf from ten plant individuals from four additional host populations (IDs 3301, 4684, 1784, and 3108) in August 2014. None of the pathogen populations were same as the sampled host populations and hence, the strains used in the assay all represent allopatric combinations. Both host and pathogen populations selected for the study were separated by at least two kilometres. The collected leaves supporting infection were placed in Petri dishes on moist filter paper and stored at room temperature until later use.Seeds from ten mother plants from each population were sown in 2:1 mixture of potting soil and sand, and grown in greenhouse conditions at 20 ± 2 °C (day) and 16 ± 2 °C (night) with 16:8 L:D photoperiod. Due to the low germination rate of collected seeds, population 260 (isolated and healthy population) was excluded from the study. Seedlings of ten different mother plants were randomly selected among the germinated plants for each population (n = 190), and grown in individual pots until the plants were eight weeks old.The pathogen strains were purified through three cycles of single colony inoculations and maintained on live, susceptible leaves on Petri dishes in a growth chamber 20 ± 2 °C with 16:8 L:D photoperiod. Every two weeks, the strains were transferred to fresh P. lanceolata leaves. Purified powdery mildew strains (M1–M4), one representing each allopatric population (3301, 4684, 1784 and 3108), were used for the inoculation assay. To produce enough sporulating fungal material, repeated cycles of inoculations were performed before the assay.Inoculation assay quantifying host resistance phenotypesIn order to study how the phenotypic resistance of hosts varies depending on population connectivity and infection history, we scored the resistance of 190 host genotypes, ten individuals from each study populations (n = 19), in an inoculation assay. Here, one detached leaf from each plant was exposed to a single pathogen strain (M1–M4) by brushing spores gently with a fine paintbrush onto the leaf. Leaves were placed on moist filter paper in Petri dishes and kept in a growth chamber at 20 ± 2 with a 16/8D photoperiod. All the inoculations were repeated on two individual Petri plates, leading to 760 host genotype—pathogen genotype combinations and a total of 1520 inoculations (19 populations * 10 plant genotypes * 4 pathogen strains * 2 replicates). We then observed and scored the pathogen infection on day 12 post inoculation, under dissecting microscope. The resulting plant phenotypic response was scored as 0 = susceptible (infection) when mycelium and conidia were observed on the leaf surface, and as 1 = resistance (no infection), when no developing lesions could be detected under a dissecting microscope. A genotype was defined resistant only if both inoculated replicates showed similar response (1), and susceptible if one or both replicates became infected (0).Statistical analysesBayesian spatio-temporal INLA model of the changes in host population sizeTo study how the pathogen infection influences on host population growth, we analyzed the relative change in host population size (m2) (defined as population size (t) − population size (t−1))/population size (t−1)) between consecutive years utilizing data from 2001 to 2008 in response to pathogen presence-absence status at t−1 (Supplementary Table 2). To assess whether this depends on host population connectivity, we estimated the separate effects of pathogen presence/absence in the previous year for connectivity categories—high-, low, and intermediate—that were based on the 0.2 and 0.8 quantiles of the host-connectivity values (Fig. 1A and Supplementary Figs. 1, 2). This allowed us to directly assess and compare the effect of the pathogen on host population growth in the extreme categories between isolated and highly connected host populations which were represented in the sampling for the inoculation study (Fig. 2).As covariates, we included the proportion (0–100%) of dry host plants measured each year within each local population as well as data on the amount of rainfall at the summer months (June, July, and August) obtained from the satellite images, as these were suggested be relevant for this pathosystem in an earlier analysis16. Observations where the change in host population size, or the host population coverage had absolute values larger than their 0.99 quantiles in the whole data, were regarded as outliers and omitted from the analysis. Before the analyses, all the continuous covariates were scaled and centred, and the categorical variables were transformed into binary variables.The relative changes in local host population size between consecutive years was analyzed by a Bayesian spatio-temporal statistical model that simultaneously considers the effects of a set of biologically meaningful predictors. The linear predictor thus consists of two parts (2,3):$$beta {X}_{t}+{z}_{t}{A}_{t}$$
    (2)
    where (beta) represents the correlation coefficients corresponding to the effects of environmental covariates, ({z}_{t}) corresponds to the spatiotemporal random effect, and ({X}_{t}) and ({A}_{t}) project these to the observation locations. For ({z}_{t}) we assume that the observations from a location in consecutive time points (t−1) and t are described by 1st order autoregressive process:$${z}_{t}=varphi {z}_{t-1}+{w}_{t}$$
    (3)
    where ({w}_{t}) corresponds to spatially structured zero-mean random noise, for which a Matern covariance function is assumed. Statistical inference then targets jointly the covariate effects (beta), the temporal autocorrelation (varphi), and the hyperparameters describing the spatial autocorrelation in wt. From these the overall variance, as well as spatial range—a distance after which spatial autocorrelation ceases to be significant—can be inferred (Supplementary Fig. 3). For more detailed description of the structure of the statistical model and how to do efficient inference with it using R-INLA, we refer to refs. 16,50.Identification of resistance phenotypesThe phenotype composition of each study population was defined by individual plant responses to the four pathogen strains, where each response could be “susceptible = 0” or “resistant = 1”. For example, a phenotype “1111” refers to a plant resistant to all four pathogen strains. The diversity of distinct resistance phenotypes within populations was estimated using the Shannon diversity index as implemented in the vegan software package51. The Shannon diversity index for all four study groups was then analyzed using a linear model with class predictors population type (well-connected or isolated), infection history (healthy or infected), and their interaction.Analysis of population connectivity and infection history effects on host resistanceTo test whether host population resistance varied depending on connectivity (SH) and infection history, we analyzed the inoculation responses (0 = susceptible, 1=resistant) of each host-pathogen combination by using a logit mixed-effect model in the lme4 package52. The model included the binomial dependent variable (resistance-susceptible; 1/0), and class predictors population type (well-connected or isolated), infection history (healthy or infected), mildew strain (M1, M2, M3, and M4) and their interactions. Plant individual and population were defined as random effects, with plant genotype (sample) hierarchically nested under population. Model fit was assessed using chi-square tests on the log-likelihood values to compare different models and significant interactions, and the best model was selected based on AIC-values. P-values for regression coefficients were obtained by using the car package53. We ran all the analyses in R software54.The metapopulation modelWe model the ecological and co-evolutionary dynamics of host and pathogen metapopulations to understand key features of the experimental system that impact on the qualitative patterns observed. The structure and parameters in our model are therefore not estimated using experimental data, but rather are chosen to cover a range of possibilities (e.g., low vs high transmission rates, variation in trade-off shapes for fitness costs). We construct the metapopulations in two stages to account for relatively well and poorly connected demes. All demes are identical in quality (i.e., no differences in intrinsic birth or death rates between demes) and only differ in their connectivity. Our metapopulation consists of an outer network of 20 demes, equally spaced around the unit square (0.2 units apart), and a 7×7 inner lattice of demes at a minimum distance of 0.2 units from the outer network (Fig. 3A), giving a total of 69 demes. Demes that are separated by a Euclidean distance of at most 0.2 are then connected to each other. This means that populations near the centre of the metapopulation are highly connected, while those on the boundary of the metapopulation are poorly connected. This also has the effect of making connections between well and poorly connected demes assortative (i.e., well/poorly connected demes tend to be connected to well/poorly connected demes). We relax the assumption of assortativity in a second type of network by randomly reassigning connections between demes, while maintaining the same degree distribution. (i.e., the probability of two demes being connected is proportionate to their degree). While well connected demes still have more connections to other well connected demes than to poorly connected demes, they are not more likely to be connected to a well connected deme than by chance based on the degree distribution. In both types of network structure, we classify a deme as well-connected if it is in the top 20% of the degree distribution and poorly connected if it is in the bottom 20%.We model the genetics using a multilocus gene-for-gene framework with haploid host and pathogen genotypes characterized by (L) biallelic loci, where 0 and 1 represent the presence and absence, respectively, of resistance and infectivity alleles. Host genotype (i) and pathogen genotype (j) are represented by binary strings: ({x}_{i}^{1}{x}_{i}^{2}ldots {x}_{i}^{L}) and ({y}_{j}^{1}{y}_{j}^{2}ldots {y}_{j}^{L}). Resistance acts multiplicatively such that the probability of host (i) being infected when challenged by pathogen (j) is ({Q}_{{ij}}={sigma }^{{d}_{{ij}}}), where (sigma) is the reduction in infectivity per effective resistance allele and ({d}_{{ij}}={sum }_{k=1}^{L}{x}_{i}^{k}big(1-{y}_{j}^{k}big)) is the number of effective resistance alleles (i.e., the number of loci where hosts have a resistance allele but pathogens do not have a corresponding infectivity allele). Hosts and pathogens with more resistance or infectivity alleles are assumed to pay higher fitness costs, ({c}_{H}left(iright)) eq. (4) and ({c}_{P}left(jright)) eq. (5) with:$${c}_{H}left(iright)={c}_{H}^{1}left(frac{1-{{{{{rm{e}}}}}}^{frac{{c}_{H}^{2}}{L}{sum }_{k=1}^{L}{x}_{i}^{k}}}{1-{{{{{rm{e}}}}}}^{{c}_{H}^{2}}}right)$$
    (4)
    and$${c}_{P}left(jright)={c}_{P}^{1}left(frac{1-{{{{{rm{e}}}}}}^{frac{{c}_{P}^{2}}{L}{sum }_{k=1}^{L}{y}_{j}^{k}}}{1-{{{{{rm{e}}}}}}^{{c}_{P}^{2}}}right)$$
    (5)
    where (0 , < , {c}_{H}^{1},; {c}_{P}^{1},le, 1) control the overall strength of the costs (i.e., the maximum proportional reduction in reproduction (hosts) or transmission rate (pathogens)) and ({c}_{H}^{2},; {c}_{P}^{2}in {{mathbb{R}}}_{ne 0}) control the shape of the trade-off. When ({c}_{H}^{2},; {c}_{P}^{2}, < , 0) the costs decelerate (increasing returns) and when ({c}_{H}^{2},; {c}_{P}^{2}, > , 0) the costs accelerate the costs accelerate (decreasing returns) (Supplementary Fig. 4). This formulation, therefore, allows for a wide-range of trade-off shapes that may occur in nature.The dynamics of the (finite) host and pathogen populations are modelled stochastically using the tau-leap method with a fixed step size of (tau=1). For population (p), the mean host birth rate at time (t) for host (i) (6) is$${B}_{i}^{p}left(tright)=left(aleft(1-{c}_{H}left(iright)right)-q{N}_{p}left(tright)right){S}_{i}^{p}left(tright)$$
    (6)
    where (a) is the maximum per-capita birth rate, (q) is the strength of density-dependent competition on births, ({N}_{p}left(tright)={S}_{i}^{p}left(tright)+{I}_{icirc }^{p}left(tright)) is the local host population size, ({S}_{i}^{p}left(tright)) and ({I}_{icirc }^{p}left(tright)={sum }_{j=1}^{n}{I}_{{ij}}^{p}left(tright)) are the local sizes of susceptible and infected individuals of genotype (i), and ({I}_{{ij}}^{p}left(tright)) is the local size of hosts of genotype (i) infected by pathogen (j). Host mutations occur at an average rate of ({mu }_{H}) per loci (limited to at most one mutation per time step), so that the mean number of mutations from host type (i) to ({i}^{{prime} }) is ({mu }_{H}{m}_{i{i}^{{prime} }}{B}_{i}^{p}left(tright)), where ({m}_{i{i}^{{prime} }}=1) if genotypes (i) and ({i}^{{prime} }) differ at exactly one locus, and is 0 otherwise.The mean local mortalities for susceptible and infected individuals are (b{S}_{i}^{p}left(tright)) and (left(b+alpha right){I}_{{ij}}^{p}left(tright)), respectively, where (b) is the natural mortality rate and (alpha) is the disease-associated mortality rate. The average number of infected hosts that recover is (gamma {I}_{{ij}}^{p}left(tright)), where (gamma) is the recovery rate.The mean number of new local infections of susceptible host type (i) by pathogen (j) eq. (7) is:$${INF}_{{ij}}^{p}left(tright)=beta left(1-{c}_{P}left(jright)right){Q}_{{ij}}{S}_{i}^{p}left(tright){Y}_{j}^{p}left(tright)$$
    (7)
    where (beta) is the baseline transmission rate and ({Y}_{j}^{p}left(tright)) is the local number of pathogen propagules following mutation and dispersal. Pathogen mutations occur in a similar manner to host mutations, with mutations from type (j) to ({j}^{{prime} }) occurring at rate ({mu }_{P}{m}_{j{j}^{{prime} }}{I}_{circ j}^{p}left(tright)) where ({mu }_{P}) is the mutation rate per loci (limited to at most one mutation per timestep) and ({I}_{circ j}^{p}left(tright)={sum }_{i=1}^{n}{I}_{{ij}}^{p}left(tright)) is the local number of pathogen (j.) Following mutation, the local number of pathogens of type (j) eq. (8) is:$${W}_{j}^{p}left(tright)={I}_{circ j}^{p}left(tright)left(1-{mu }_{P}Lright)+{mu }_{P}{m}_{j{j}^{{prime} }}{I}_{circ j}^{p}left(tright)$$
    (8)
    Pathogen dispersal occurs following mutation at a rate of (rho) between connected demes, given by the adjacency matrix ({G}_{{pr}}), with ({G}_{varSigma p}) the total number of connections for deme (p). The mean local number of pathogen propagules following mutation and dispersal eq. (9) is therefore:$${Y}_{j}^{p}left(tright)={W}_{j}^{p}left(tright)left(1-rho {G}_{varSigma p}right)+rho mathop {sum }limits_{r=1}^{{M}_{varSigma }}{G}_{{pr}}{W}_{j}^{r}left(tright)$$
    (9)
    We focus our parameter sweep on: (i) the structure of the network (assortative or random connections); (ii) the strength (left({c}_{H}^{1},; {c}_{P}^{1}right)) and shape (left({c}_{H}^{2},; {c}_{P}^{2}right)) of the trade-offs; (iii) the transmission rate (left(beta right)); and (iv) the dispersal rate (left(rho right)), fixing the remaining parameters as described in Supplementary Table 1 (preliminary investigations suggested they had less of an impact on the qualitative outcome) and conducting 100 simulations per parameter set. For each simulation we initially seed all populations with the most susceptible host type and place the least infective pathogen type in one of the well-connected populations to minimize the risk of early extinction. We then solve the dynamics for 10,000 time steps (preliminary investigations indicated this was a sufficient period for the metapopulations to reach a quasi-equilibrium in terms of overall resistance). We calculate the average level of resistance (proportion of loci with a resistance allele) between time steps 4001 and 5000 (transient dynamics) and over the final 1000 time steps (long-term dynamics) for well and poorly connected demes, categorized according to whether the disease is present in (infected) or absent from (uninfected) the local population at a given time point and discarding simulations where the pathogen is driven globally extinct.We compare the mean level of resistance in infected/uninfected poorly/well-connected populations across all simulations to the empirical results. We say that a simulation is a qualitative ‘match’ for the empirical findings if: (i) in poorly connected demes, the infected populations are on average at least 5% more resistant than uninfected populations; and (ii) in well-connected demes, the uninfected populations are on average at least 5% more resistant than infected populations. In other words, if ({R}_{{CS}}) is the mean resistance for a population with connectivity (C) ((C=W) and (C=P) for well and poorly connected demes, respectively) and infection status (S) ((S=U) and (S=I) for uninfected and infected populations, respectively), then a parameter set is a qualitative ‘match’ for the empirical findings if ({R}_{{WU}} > 1.05{R}_{{WI}}) and (1.05{R}_{{PI}}, > , 1.05{R}_{{PU}}). If these criteria are not met, then the parameter set is a qualitative ‘mismatch’ for the empirical findings. The model is not intended to be a replica of an empirical metapopulation, but rather is used to reveal the key factors which lead to qualitatively similar distributions of resistance and disease incidences observed in the study of the Åland islands. Hence, the purpose of the model is to determine which biological factors are likely to be crucial to the patterns observed herein.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More

  • in

    Global hotspots for soil nature conservation

    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

  • in

    Global soil map pinpoints key sites for conservation

    Johnson, N. et al. (eds) Global Soil Biodiversity Atlas (EU, 2016).
    Google Scholar 
    FAO et al. State of Knowledge of Soil Biodiversity — Status, Challenges and Potentialities (FAO, 2020).
    Google Scholar 
    Cameron, E. K. et al. Nature Ecol. Evol. 2, 1042–1043 (2018).PubMed 
    Article 

    Google Scholar 
    van den Hoogen, J. et al. Nature 572, 194–198 (2019).PubMed 
    Article 

    Google Scholar 
    Phillips, H. R. P. et al. Science 366, 480–485 (2019).PubMed 
    Article 

    Google Scholar 
    Guerra, C. A. et al. Nature https://doi.org/10.1038/s41586-022-05292-x (2022).Article 

    Google Scholar 
    Moore, J. C. & de Ruiter, P. C. Energetic Food Webs: An Analysis of Real and Model Ecosystems (Oxford Univ. Press, 2012).
    Google Scholar 
    Wolters V. et al. Bioscience 50, 1089–1098 (2000).Article 

    Google Scholar 
    Schimel, J. P. & Schaeffer, S. M. Front. Microbiol. 3, 348 (2012).PubMed 
    Article 

    Google Scholar 
    IPCC. In Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change Impacts, Adaptation, and Vulnerability: Summary for Policymakers (eds Shukla, P. R. et al.) 50 (Cambridge Univ. Press, 2022).
    Google Scholar 
    Chenu, C. et al. Soil Till. Res. 188, 41–52 (2019).Article 

    Google Scholar 
    Liang, C., Schimel, J. P. & Jastrow, J. D. Nature Microbiol. 2, 17105 (2017).PubMed 
    Article 

    Google Scholar 
    Hannula, S. E. & Morriën, E. Geoderma 413, 115767 (2022).Article 

    Google Scholar  More

  • in

    Ecological risk and health risk analysis of soil potentially toxic elements from oil production plants in central China

    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