美国版science原版科学杂志2018年合集全51本.pdf 134页

127 145 CONTENTS Responding to light 13 JULY 2018 • VOLUME 361 • ISSUE 6398 D o w n l o a d e d 120 f r o m FEATURE 130 ERADICATION GENOMICS— NEWS 120 LIQUID SUNSHINE LESSONS FOR PARASITE CONTROL Ammonia made from sun, air, and water Genomic surveillance could could turn Australia into a renewable help achieve targets for the IN BRIEF energy superpower By R. F. Service elimination of tropical diseases ▶ VIDEO By J. A. Cotton et al. 112 News at a glance 132 PEROVSKITE FERROELECTRICS IN DEPTH GO METAL FREE 115 ICE REVEALS A MESSENGER FROM INSIGHTS Metal-free perovskites exhibit o n A BLAZING GALAXY ferroelectric properties rivaling J u Neutrino astronomy is born as IceCube those of BaTiO By W. Li and L.-J. Ji 3 l PERSPECTIVES y and other instruments find likely home ▶ RESEARCH ARTICLE P. 151 1 5 of particle By D. Clery 124 WHEN PERSISTENCE DOESN’T PAY , 2 0 ▶ RESEARCH ARTICLES PP. 146 147; PODCAST Rats, mice, and humans all invest 133 JENS CHRISTIAN SKOU 1 more time in a foraging task than is (1918–2018) 8 116 CHINA ANNOUNCES NEW FLOTILLA OF in their interest By S. F. Brosnan A pioneer in the biochemistry of SPACE SCIENCE MISSIONS ▶ REPORT P. 178 membrane proteins By P. Nissen Probes will watch for colliding black A I L holes, image Earth’s magnetosphere, 125 MANY ROADS TO CONVERGENCE POLICY FORUM A R T and monitor the sun By D. Nor Plant genomes highlight complex S U 134 CITIZEN SCIENCE, PUBLIC POLICY A mechanisms behind evolutionary H New research models may T U 117 DIGITAL CHEMICAL TEST IMPRESSES O convergence By L. G. Nagy S benefit from policy modifications D, Giant database shows promise for ▶ RESEARCH ARTICLE P. 144 R By C. J. Guerrini et al. A O B replacing animal studies N O By V. Zainzinger 127 FEMTOSECOND STRUCTURAL I T BOOKS ET AL. C E PHOTOBIOLOGY T O R 118 A ‘GENE DRIVE’ MAKES ITS DEBUT Time-resolved crystallography 137 TO BE A BEE P T S IN MAMMALS reveals how bacteriorhodopsin uses A charming account celebrates the A O C ; CRISPR-based gene-spreading strategy light to drive chemistry By K. Moffat insects’ idiosyncrasies and the people L. A promises to speed development of passionate about protecting them T ▶ RESEARCH ARTICLE P. 145 E Y L engineered mice By J. Cohen By R. Winfree G O N 128 DOUBLE TROUBLE AT THE P. ) 119 MOLECULAR ‘BARCODES’ REVEAL BEGINNING OF LIFE 138 LEARNING FROM DIFFERENT P O T LOST WHALE HUNTS Dual-spindle assembly in early embryos DISCIPLINES M O R F DNA method can rapidly identify can compromise mammalian development Conversations spark connections ( : S T species, shows toll of ancient cultures By A. P. Zielinska and M. Schuh as scientists search for inspiration in I D E on biodiversity By E. Pennisi and M. Price ▶ REPORT P. 189 other fields By B. Uzzi R C 108 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713TOC.indd 108 7/11/18 10:36 AM 132 151 166 All-organic perovskites Satellites locate nuclear test LETTERS 145 STRUCTURAL DYNAMICS 178 EVOLUTIONARY COGNITION 139 BRAZIL NATURALIZES Retinal isomerization in Sensitivity to “sunk costs” in mice, NON-NATIVE SPECIES bacteriorhodopsin captured by a rats, and humans B. M. Sweis et al. femtosecond x-ray laser P. Nogly et al. ▶ PERSPECTIVE P. 124 By M. F. G. Brito et al. RESEARCH ARTICLE SUMMARY; FOR FULL TEXT: 181 PLANT SCIENCE 139 FREE SATELLITE DATA KEY /10.1126/science.aat0094 ▶ PERSPECTIVE P. 127 Ethylene-gibberellin signaling underlies TO CONSERVATION adaptation of rice to periodic flooding By G. M. Buchanan et al. NEUTRINO ASTROPHYSICS T. Kuroha et al. 146 Multimessenger observations of a ▶ PODCAST 140 FUNDING AGENCIES CAN flaring blazar coincident with high- PREVENT HARASSMENT 186 GREENHOUSE GASES energy neutrino IceCube-170922A By L. L. Iversen and M. Bendixen The IceCube Collaboration et al. Assessment of methane emissions from the U.S. oil and gas supply D RESEARCH ARTICLE SUMMARY; FOR FULL o 140 TECHNICAL COMMENT ABSTRACTS chain R. A. Alvarez et al. w TEXT: /10.1126/science.aat1378 n l o a 147 Neutrino emission from the 189 EARLY DEVELOPMENT d e direction of the blazar TXS d Dual-spindle formation in zygotes f r RESEARCH 0506+056 prior to the IceCube- keeps parental genomes apart in early o m 170922A alert IceCube Collaboration mammalian embryos J. Reichmann et al. ▶ NEWS STORY P. 115 ▶ PERSPECTIVE P. 128 IN BRIEF 141 From Science and other journals 151 FERROELECTRICITY Metal-free three-dimensional perovskite DEPARTMENTS RESEARCH ARTICLES ferroelectrics H.-Y. Ye et al. 111 EDITORIAL ▶ PERSPECTIVE P. 132 Spain’s good news By Amaya Moro-Martín 144 PLANT SCIENCE Phylogenomics reveals multiple losses 156 SYNTHETIC BIOLOGY 198 WORKING LIFE of nitrogen-fixing root nodule symbiosis Programming self-organizing Heed the call to change By Tracy Evans M. Griesmann et al. multicellular structures with synthetic RESEARCH ARTICLE SUMMARY; FOR FULL TEXT: cell-cell signaling S. Toda et al. o /10.1126/science.aat1743 n J ON THE COVER ▶ PERSPECTIVE P. 125 u l y REPORTS Artist’s impression of 1 the IceCube Neutrino 5 , 162 PHYSICS 2 0 Phase transitions in a programmable Observatory in 7 1 1 8 L quantum spin glass simulator Antarctica. Spherical L 139 I E digital optical modules N R. Harris et al. C (DOMs), each about M C / 35 cm in diameter, are M O 166 GEODETIC MONITORING C K. positioned up to 2.5 km C The rise, collapse, and compaction of O deep in the ice. More T S I Mt. Mantap from the 3 September 2017 ; L. than 5000 DOMs make up a cubic-kilometer A North Korean nuclear test T. Wang et al. T detector weighing more than a billion tons. E G The DOMs detect the faint flash of light N A W 171 ORGANIC CHEMISTRY . created when a high-energy neutrino T ; E Deconstructive fluorination of cyclic interacts with the ice. See pages 115, 146, C N E and 147. Image: Jamie Yang and Savannah I amines by carbon-carbon cleavage C S / J. B. Roque et al. Guthrie/IceCube/NSF L E K C I B C. 174 EARLY OCEAN ) Science Staff 110 T F Late inception of a resiliently E New Products 194 L P oxygenated upper ocean W. Lu et al. Science Careers 195 O T M O R F E S I SCIENCE (ISSN 0036-8075) is published weekly on Friday, except last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Periodicals mail W K postage (publication No. 484460) paid at Washington, DC, and additional mailing offices. Copyright © 2018 by the American Association for the Advancement of Science. The title SCIENCE is a registered trademark of the AAAS. Domestic C individual membership, including subscription (12 months): $165 ($74 allocated to subscription). Domestic institutional subscription (51 issues): $1808; Foreign postage extra: Mexico, Caribbean (surface mail) $55; other countries O L (air assist delivery): $89. First class, airmail, student, and emeritus rates on request. Canadian rates with GST available upon request, GST #125488122. Publications Mail Agreement Number 1069624. Printed in the U.S.A. C ( : Change of address: Allow 4 weeks, giving old and new addresses and 8-digit account number. Postmaster: Send change of address to AAAS, P.O. Box 96178, Washington, DC 20090–6178. Single-copy sales: $15 each plus shipping S T and handling; bulk rate on request. Authorization to reproduce material for internal or personal use under circumstances not falling within the fair use provisions of the Copyright Act is granted by AAAS to libraries and others who I D E use Copyright Clearance Center (CCC) Pay-Per-Use services provided that $35.00 per article is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923. The identification code for Science is 0036-8075. Science is indexed in the R C Reader’s Guide to Periodical Literature and in several specialized indexes. SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 109 Published by AAAS DA_0713TOC.indd 109 7/11/18 10:36 AM BOARD OF REVIEWING EDITORS (Statistics board members indicated withS) 1200 New York Avenue, NW, Washington, DC 20005 Adriano Aguzzi, U. Hospital Zürich David Lazer, Harvard U. Clarendon House, Clarendon Road, Cambridge, UK CB2 8FH Takuzo Aida, U. of Tokyo Stanley Lemon, U. of North Carolina at Chapel Hill Leslie Aiello, Wenner-Gren Foundation Ottoline Leyser, U. of Cambridge Judith Allen, U. of Manchester Wendell Lim, U. of California, San Francisco Sebastian Amigorena, Institut Curie Marcia C. Linn, U. of California, Berkeley Meinrat O. Andreae, Max Planck Inst. Mainz Jianguo Liu, Michigan State U. Editor-in-Chief Jeremy Berg Paola Arlotta, Harvard U. Johan Auwerx, EPFL Luis Liz-Marzán, CIC biomaGUNE Executive Editor Monica M. Bradford News Editor Tim Appenzeller David Awschalom, U. of Chicago Jonathan Losos, Harvard U. Deputy Editors Lisa D. Chong, Andrew M. Sugden(UK), Valda J. Vinson, Jake S. Yeston Clare Baker, U. of Cambridge Ke Lu, Chinese Acad. of Sciences Nenad Ban, ETH ZÜrich Christian Lüscher, U. of Geneva Franz Bauer, Pontificia Universidad Católica de Chile Fabienne Mackay, U. of Melbourne Research and Insights Ray H. Baughman, U. of Texas at Dallas Anne Magurran, U. of St. Andrews DEPUTY EDITOR, EMERITUS Barbara R. Jasny SR. EDITORS Gemma Alderton(UK), Caroline Ash(UK), Julia Fahrenkamp-Uppenbrink(UK), Carlo Beenakker, Leiden U. Oscar Marín, King’s College London Kamran Behnia, ESPCI Pamela J. Hines, Stella M. Hurtley(UK), Paula A. Kiberstis, Marc S. Lavine(Canada), Steve Mao, Ian S. Osborne(UK), Beverly A. Charles Marshall, U. of California, Berkeley Yasmine Belkaid, NIAID, NIH Purnell, L. Bryan Ray, H. Jesse Smith, Jelena Stajic, Peter Stern(UK), Phillip D. Szuromi, Sacha Vignieri, Brad Wible, Laura M. Zahn Philip Benfey, Duke U. Christopher Marx, U. of Idaho ASSOCIATE EDITORS Michael A. Funk, Brent Grocholski, Priscilla N. Kelly, Tage S. Rai, Seth Thomas Scanlon(UK), Keith T. Smith(UK) Gabriele Bergers, VIB C. Robertson McClung, Dartmouth College ASSOCIATE BOOK REVIEW EDITOR Valerie B. Thompson LETTERS EDITOR Jennifer Sills LEAD CONTENT PRODUCTION EDITORS Harry Jach, Lauren Bradley Bernstein, Massachusetts General Hospital Rodrigo Medellín, U. of Mexico Kmec CONTENT PRODUCTION EDITORS Amelia Beyna, Jeffrey E. Cook, Amber Esplin, Chris Filiatreau, Cynthia Howe, Catherine Wolner Peer Bork, EMBL Graham Medley, London School of Hygiene SR. EDITORIAL COORDINATORS Carolyn Kyle, Beverly Shields EDITORIAL COORDINATORS Aneera Dobbins, Joi S. Granger, Jeffrey Hearn, Lisa Chris Bowler, École Normale Supérieure Tropical Med. Johnson, Maryrose Madrid, Jerry Richardson, Alice Whaley(UK), Anita Wynn PUBLICATIONS ASSISTANTS Ope Martins, Nida Masiulis, Ian Boyd, U. of St. Andrews Jane Memmott, U. of Bristol Dona Mathieu, Ronmel Navas, Hilary Stewart(UK), Alana Warnke, Brian White EXECUTIVE ASSISTANT Jessica Slater ADMINISTRATIVE Emily Brodsky, U. of California, Santa Cruz Tom Misteli, NCI, NIH SUPPORT Janet Clements(UK), Ming Yang (UK) Ron Brookmeyer, U. of California, Los Angeles (S) Yasushi Miyashita, U. of Tokyo Christian Büchel, UKE Hamburg Richard Morris, U. of Edinburgh Dennis Burton, The Scripps Res. Inst. Alison Motsinger-Reif, NC State U. (S) Carter Tribley Butts, U. of California, Irvine News Daniel Neumark, U. of California, Berkeley Gyorgy Buzsaki, New York U. School of Medicine NEWS MANAGING EDITOR John Travis INTERNATIONAL EDITOR Martin Enserink DEPUTY NEWS EDITORS Elizabeth Culotta, David Grimm, Blanche Capel, Duke U. Kitty Nijmeijer, TU Eindhoven Eric Hand, David Malakoff, Leslie Roberts SR. CORRESPONDENTS Daniel Clery(UK), Jeffrey Mervis, Elizabeth Pennisi ASSOCIATE Mats Carlsson, U. of Oslo Helga Nowotny, Austrian Counci EDITORS Jeffrey Brainard, Catherine Matacic NEWS WRITERS Adrian Cho, Jon Cohen, Jennifer Couzin-Frankel, Jocelyn Kaiser, Ib Chorkendorff, Denmark TU Rachel O’Reilly, U. of Warwick Kelly Servick, Robert F. Service, Erik Stokstad(Cambridge, UK), Paul Voosen, Meredith Wadman INTERNS Katie Langin, Frankie James J. Collins, MIT Harry Orr, U. of Minnesota Schembri, Matt Warren CONTRIBUTING CORRESPONDENTS Warren Cornwall, Ann Gibbons, Mara Hvistendahl, Sam Kean, Eli Kintisch, Robert Cook-Deegan, Arizona State U. Pilar Ossorio, U. of Wisconsin Kai Kupferschmidt(Berlin), Andrew Lawler, Mitch Leslie, Eliot Marshall, Virginia Morell, Dennis Nor hanghai), Charles Piller, Lisa Coussens, Oregon Health Science U. Andrew Oswald, U. of Warwick Tania Rabesandratana(London), Emily Underwood, Gretchen Vogel(Berlin), Lizzie Wade(Mexico City) CAREERS Donisha Adams, Alan Cowman, Walter Eliza Hall Inst. Isabella Pagano, Istituto Nazionale di Astrofisica Roberta Croce, VU Amsterdam Rachel Bernstein(Editor) COPY EDITORS Julia Cole (Senior Copy Editor), Cyra Master (Copy Chief) ADMINISTRATIVE SUPPORT Jeff L. Dangl, U. of North Carolina Margaret Palmer, U. of Maryland Meagan Weiland Tom Daniel, U. of Washington Steve Palumbi, Stanford U. Chiara Daraio, Caltech Jane Parker, Max Planck Inst. Cologne Nicolas Dauphas, U. of Chicago Giovanni Parmigiani, Dana-Farber Cancer Inst. (S) Executive Publisher Rush D. Holt Frans de Waal, Emory U. Samuel Pfaff, Salk Inst. for Biological Studies Stanislas Dehaene, Collège de France Matthieu Piel, Institut Curie Publisher Bill Moran Chief Digital Media Officer Josh Freeman Robert Desimone, MIT Kathrin Plath, U. of California, Los Angeles DIRECTOR, BUSINESS STRATEGY AND PORTFOLIO MANAGEMENT Sarah Whalen DIRECTOR, PRODUCT AND CUSTOM PUBLISHING Will Schweitzer Claude Desplan, New York U. Martin Plenio, Ulm U. MANAGER, PRODUCT DEVELOPMENT Hannah Heckner BUSINESS SYSTEMS AND FINANCIAL ANALYSIS DIRECTOR Randy Yi DIRECTOR, BUSINESS Sandra DÍaz, Universidad Nacional de CÓrdoba Albert Polman, FOM Institute for AMOLF Dennis Discher, U. of Penn. OPERATIONS ANALYST Eric Knott ASSOCIATE DIRECTOR, PRODUCT MANAGMENT Kris Bishop ASSOCIATE DIRECTOR, INSTITUTIONAL LICENSING Elvira Poloczanska, Alfred-Wegener-Inst. Gerald W. Dorn II, Washington U. in St. Louis SALE Geoffrey Worton SENIOR SYSTEMS ANALYST Nicole Mehmedovich SENIOR BUSINESS ANALYST Cory Lipman MANAGER, BUSINESS Jennifer A. Doudna, U. of California, Berkeley Philippe Poulin, CNRS OPERATIONS Jessica Tierney BUSINESS ANALYSTS Meron Kebede, Sandy Kim, Jourdan Stewart FINANCIAL ANALYST Julian Iriarte Bruce Dunn, U. of California, Los Angeles Jonathan Pritchard, Stanford U. ADVERTISING SYSTEM ADMINISTRATOR Tina Burks SALES COORDINATOR Shirley Young DIRECTOR, COPYRIGHT, LICENSING, SPECIAL PROJECTS William Dunphy, Caltech David Randall, Colorado State U. Emilie David DIGITAL PRODUCT ASSOCIATE Michael Hardesty RIGHTS AND PERMISSIONS ASSOCIATE Elizabeth Sandler RIGHTS, CONTRACTS, Christopher Dye, U. of Oxford Sarah Reisman, Caltech AND LICENSING ASSOCIATE Lili Catlett RIGHTS PERMISSIONS ASSISTANT Alexander Lee Todd Ehlers, U. of TÜbingen Félix A. Rey, Institut Pasteur Jennifer Elisseeff, Johns Hopkins U. Trevor Robbins, U. of Cambridge MARKETING MANAGER, PUBLISHING Shawana Arnold SENIOR ART ASSOCIATES Paula Fry ART ASSOCIATE Kim Huynh Tim Elston, U. of North Carolina at Chapel Hill Amy Rosenzweig, Northwestern U. Barry Everitt, U. of Cambridge Mike Ryan, U. of Texas at Austin DIRECTOR, INSTITUTIONAL LICENSING Iquo Edim ASSOCIATE DIRECTOR, RESEARCH DEVELOPMENT Elisabeth Leonard Vanessa Ezenwa, U. of Georgia Mitinori Saitou, Kyoto U. SENIOR INSTITUTIONAL LICENSING MANAGER Ryan Rexroth INSTITUTIONAL LICENSING MANAGERS Marco Castellan, Chris Murawski Ernst Fehr, U. of ZÜrich Shimon Sakaguchi, Osaka U. SENIOR OPERATIONS ANALYST Lana Guz MANAGER, AGENT RELATIONS CUSTOMER SUCCESS Judy Lillibridge Michael Feuer, The George Washington U. Toren Finkel, NHLBI, NIH Miquel Salmeron, Lawrence Berkeley Nat. Lab WEB TECHNOLOGIES TECHINICAL DIRECTOR David Levy PORTFOLIO MANAGER Trista Smith PROJECT MANAGER Dean Robbins DEVELOPERS Liana Kate Fitzgerald, U. of Massachusetts Nitin Samarth, Penn. State U. Birke, Elissa Heller, Ryan Jensen Peter Fratzl, Max Planck Inst. Potsdam Jürgen Sandkühler, Medical U. of Vienna Elaine Fuchs, Rockefeller U. Alexander Schier, Harvard U. DIGITAL MEDIA DIRECTOR OF ANALYTICS Enrique Gonzales MULTIMEDIA MANAGER Sarah Crespi MANAGING WEB PRODUCER Kara Estelle- Eileen Furlong, EMBL Wolfram Schlenker, Columbia U. Powers DIGITAL PRODUCER Jessica Hubbard VIDEO PRODUCER Chris Burns SOCIAL MEDIA PRODUCER Brice Russ Jay Gallagher, U. of Wisconsin Susannah Scott, U. of California, Santa Barbara DIGITAL/PRINT STRATEGY MANAGER Jason Hillman QUALITY TECHNICAL MANAGER Marcus Spiegler DIGITAL PRODUCTION MANAGER Lisa Daniel Geschwind, U. of California, Los Angeles Vladimir Shalaev, Purdue U. Stanford ASSISTANT MANAGER DIGITAL/PRINT Rebecca Doshi SENIOR CONTENT SPECIALISTS Steve Forrester, Antoinette Hodal, Karl-Heinz Glassmeier, TU Braunschweig Beth Shapiro, U. of California, Santa Cruz Ramon Gonzalez, Rice U. Lori Murphy, Anthony Rosen CONTENT SPECIALISTS Jacob Hedrick, Kimberley Oster Elizabeth Grove, U. of Chicago Jay Shendure, U. of Washington Brian Shoichet, U. of California, San Francisco DESIGN DIRECTOR Beth Rakouskas DESIGN MANAGING EDITOR Marcy Atarod SENIOR DESIGNER Chrystal Smith DESIGNER Christina Nicolas Gruber, ETH ZÜrich Kip Guy, U. of Kentucky College of Pharmacy Robert Siliciano, Johns Hopkins U. School of Aycock GRAPHICS MANAGING EDITOR Alberto Cuadra GRAPHICS EDITOR Nirja Desai SENIOR SCIENTIFIC ILLUSTRATORS Valerie Altounian, Taekjip Ha, Johns Hopkins U. Medicine Chris Bickel, Katharine Sutliff SCIENTIFIC ILLUSTRATOR Alice Kitterman INTERACTIVE GRAPHICS EDITOR Jia You SENIOR GRAPHICS Christian Haass, Ludwig ilians U. Uri Simonsohn, U. of Penn. SPECIALISTS Holly Bishop, Nathalie Cary PHOTOGRAPHY MANAGING EDITOR William Douthitt PHOTO EDITOR Emily Petersen Sharon Hammes-Schiffer, U. of Illinois at Lucia Sivilotti, U. College London IMAGE RIGHTS AND FINANCIAL MANAGER Jessica Adams Urbana-Champaign Alison Smith, John Innes Centre Wolf-Dietrich Hardt, ETH ZÜrich Richard Smith, U. of North Carolina at Chapel Hill (S) SENIOR EDITOR, CUSTOM PUBLISHING Sean Sanders: ASSISTANT EDITOR, CUSTOM PUBLISHING Jackie Oberst: Louise Harra, U. College London Mark Smyth, QIMR Berghofer ASSOCIATE DIRECTOR, BUSINESS DEVELOPMENT Justin Sawyers: science_advertising@ ADVERTISING PRODUCTION OPERATIONS Michael Hasselmo, Boston U. Pam Soltis, U. of Florida MANAGER Deborah Tompkins SR. PRODUCTION SPECIALIST/GRAPHIC DESIGNER Amy Hardcastle SR. TRAFFIC ASSOCIATE Christine Hall Jian He, Clemson U. John Speakman, U. of Aberdeen DIRECTOR OF BUSINESS DEVELOPMENT AND ACADEMIC PUBLISHING RELATIONS, ASIA Xiaoying Chu: +86-131 6136 3212, xchu@ Martin Heimann, Max Planck Inst. Jena COLLABORATION/CUSTOM PUBLICATIONS/JAPAN Adarsh Sandhu + 81532-81-5142 asandhu@ EAST COAST/E. CANADA Laurie Faraday: Carl-Philipp Heisenberg, IST Austria Tara Spires-Jones, U. of Edinburgh Ykä Helariutta, U. of Cambridge Allan C. Spradling, Carnegie Institution for Science FAX WEST COAST/W. CANADA Lynne Stickrod: , FAX MIDWEST Jeffrey Dembksi: Janet G. Hering, Eawag Eric Steig, U. of Washington x3005, Steven Loerch: x3006 UK EUROPE/ASIA Roger Goncalves: TEL/FAX +41 43 243 1358 JAPAN Kaoru Sasaki (Tokyo): + 81 (3) Kai-Uwe Hinrichs, U. of Bremen Paula Stephan, Georgia State U. 6459 4174 ksasaki@ David Hodell, U. of Cambridge V. S. Subrahmanian, U. of Maryland GLOBAL SALES DIRECTOR ADVERTISING AND CUSTOM PUBLISHING Tracy Holmes: +44 (0) 1223 326525 CLASSIFIED advertise@ SALES Lora Hooper, UT Southwestern Medical Ctr. at Dallas Ira Tabas, Columbia U. MANAGER, US, CANADA AND LATIN AMERICA SCIENCE CAREERS Claudia Paulsen-Young: EUROPE/ROW SALES Sarah Lelarge SALES Fred Hughson, Princeton U. Sarah Teichmann, U. of Cambridge Randall Hulet, Rice U. ADMIN ASSISTANT Kelly Grace +44 (0)1223 326528 JAPAN Miyuki Tani(Osaka): +81 (6) 6202 6272 mtani@ CHINA/ Xiaoying Chu: Auke Ijspeert, EPFL Shubha Tole, Tata Inst. of Fundamental Research +86-131 6136 3212, xchu@ GLOBAL MARKETING MANAGER Allison Pritchard DIGITAL MARKETING ASSOCIATE Aimee Aponte Akiko Iwasaki, Yale U. Wim van der Putten, Netherlands Inst. of Ecology Stephen Jackson, USGS and U. of Arizona Bert Vogelstein, Johns Hopkins U. AAAS BOARD OF DIRECTORS, CHAIR Susan Hockfield PRESIDENT Margaret A. Hamburg PRESIDENT-ELECT Steven Chu TREASURER Carolyn Seema Jayachandran, Northwestern U. David Wallach, Weizmann Inst. of Science N. Ainslie CHIEF EXECUTIVE OFFICER Rush D. Holt BOARD Cynthia M. Beall, May R. Berenbaum, Rosina M. Bierbaum, Kaye Husbands Kai Johnsson, EPFL Jane-Ling Wang, U. of California, Davis (S) Fealing, Stephen P.A. Fodor, S. James Gates, Jr., Michael S. Gazzaniga, Laura H. Greene, Robert B. Millard, Mercedes Pascual, Peter Jonas, Inst. of Science Technology Austria David Waxman, Fudan U. William D. Provine Matt Kaeberlein, U. of Washington Jonathan Weissman, U. of California, San Francisco William Kaelin Jr., Dana-Farber Cancer Inst. Chris Wikle, U. of Missouri (S) SUBSCRIPTION SERVICES For change of address, missing issues, new orders and renewals, and payment questions: 866-434-AAAS (2227) or FAX Mailing addresses: AAAS, P.O. Box 96178, Washington, DC 20090-6178 or AAAS Member Services, 1200 New York Avenue, NW, Washington, DC 20005 Daniel Kammen, U. of California, Berkeley Terrie Williams, U. of California, Santa Cruz Abby Kavner, U. of California, Los Angeles Ian A. Wilson, The Scripps Res. Inst. (S) INSTITUTIONAL SITE LICENSES REPRINTS: Author Inquiries COMMERCIAL INQUIRIES PERMISSIONS , Masashi Kawasaki, U. of Tokyo Timothy D. Wilson, U. of Virginia permissions@ AAAS Member Central Support . V. Narry Kim, Seoul Nat. U. Yu Xie, Princeton U. Robert Kingston, Harvard Medical School Science serves as a forum for discussion of important issues related to the advancement of science by publishing material on which a consensus has been reached as Etienne Koechlin, École Normale Supérieure Jan Zaanen, Leiden U. well as including the presentation of minority or conflicting points of view. Accordingly, all articles published in Science—including editorials, news and comment, and book Alexander Kolodkin, Johns Hopkins U. Kenneth Zaret, U. of Penn. School of Medicine reviews—are signed and reflect the individual views of the authors and not official points of view adopted by AAAS or the institutions with which the authors are affiliated. Thomas Langer, U. of Cologne Jonathan Zehr, U. of California, Santa Cruz INFORMATION FOR AUTHORS See Mitchell A. Lazar, U. of Penn. Maria Zuber, MIT 110 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Masthead.indd 110 7/11/18 10:25 AM EDITORIAL Spain’s good news ast month finally brought good news for science innovations based on the generation of new knowledge. in Spain. Newly elected prime minister Pedro Sán- Relying on the private sector to achieve the 2020 target of chez named Spanish astronaut and aeronautical 3% GDP investment in RD (proposed by the European engineer Pedro Duque as head of a new Ministry Union in 2000) while drastically reducing public invest- for Science, Innovation, and Universities. This will ment is unrealistic. hopefully put Spanish research and development The creation of Spain’s new Ministry for Science, In- L(RD) back on the country’s political agenda after novation, and Universities acknowledges the important two legislatures in which science languished under the role of basic research and the synergy that should exist Amaya Moro- Ministry of Economy. Given the minority position of the between research and higher education in the sciences. Martín is ruling party in both chambers It should be further supported an astrophysicist of Congress, Duque will face by the inclusion of universi- and treasurer extraordinary challenges to Madrid Innovation Campus, ties in Spain’s Science, Tech- D of the Spanish o increase the budget. But the Villaverde district, nology and Innovation Law w n Association for l creation of this new ministry Madrid, Spain. (to strengthen research at a o a d the Advancement e is an encouraging first step. national level) and by con- of Science. d Over the past 7 years, Spain crete measures to encourage f r amayamoromartin o m has suffered from insufficient mobility between research RD spending, setting its and educational institutions. @ entire scientific enterprise The latter should include back by two decades. In 2016, addressing the role of the RD represented only 1.2% National Agency for Quality of the gross domestic prod- Assessment and Accreditation uct (GDP), compared to 2% in feeding academic inbreed- for the European Union; half ing in Spanish universities (af- of this was accounted for by fecting 70% of professors; less the private sector. During this than 2% are foreigners). decline, most public research The new government cer- institutions were left nearly tainly needs to invest more o bankrupt [including the Span- in RD, but there are urgent n J ish National Research Council measures that can be imple- u l “Spain’s prosperity will only y (with 139 centers), the Span- mented quickly that require 1 4 , ish Institute of Oceanography, come about through innovations more political will than fund- 2 the Carlos III Health Insti- ing. These include ending bu- 0 1 tute, and the Research Cen- based on…new knowledge.” reaucratic processes imposed 8 O P tre for Energy, Environment by the former Ministry of Fi- S E R and Technology (CIEMAT)]. Research projects became nance on all research activities (contracts, procurements) C - A G paralyzed by new bureaucratic processes; scientific talent that have paralyzed the system; spending all appropriated E I R O thinned out dramatically; employment declined to pre- RD funds (only 50% has been spent so far); ending the N O T carious levels (plagued by irregular short-term contracts); year-long delays on announcing calls for research; ending R E B L and the average age of tenured academics increased to 53 limitations on continually renewing contracts (unavoid- A ) T years. From 2010 to 2014, Spain lost 27,358 RD jobs— able with the current 3-year grant system); exempting H G I R P had funding been maintained at 2009 levels, about 61,940 research-related contracts from lengthy competitive O T such jobs would have been created. bidding processes; and ending the Prior Intervention ( ; O T Sadly, the previous Spanish government’s vision of how Requirement that prevents research institutions from O H P K RD can change the economy was seriously flawed— executing their budgets (some institutions have reported C O it underestimated the importance of basic research, 50% losses). These practices were intended to sequester T S Y oversimplified knowledge transfer, minimized the time RD funds to reduce spending. M A L A needed to bring ideas to market, and ignored how pub- The previous government compromised Spain’s future / N A lic investment attracts private investment. The new Sán- by creating a deficit in innovation and discovery. Spain L A G A chez government must move away from this ill-conceived can no longer be trapped in this failed economic model. I R A model. Spain’s prosperity will only come about through – Amaya Moro-Martín M ) T E S N I ( : S T I 10.1126/science.aau6630 D E R C SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 111 Published by AAAS DA_0713Editorial.indd 111 7/11/18 10:20 AM A repudiation of established research … in what appears to be an effort to curry favor with commercial interests. “ Public health scholars in a letter to the Trump administration, criticizing its attempts to alter ” NEWS a World Health Assembly resolution supporting breastfeeding to favor baby formula. I N B R I E F Wellcome seeks long shots Edited by Jeffrey Brainard RESEARCH FUNDING | The world’s second biggest private funder of medi- CONSERVATION cal research, the London-based charity Wellcome Trust, announced this week Dams may threaten endangered apes that it will provide £250 million for unconventional research that could transform science or health. The money is earmarked for ideas shut out from traditional funding because they are less likely to succeed, and for projects with D o researchers who lack a background in w n l the life sciences but bring other exper- o a d e tise. Called the Wellcome Leap Fund, d it will be set up as a subsidiary of the f r o m Wellcome Trust with an independent board; the search for a CEO will start shortly and its first research programs could start in late 2020. The fund, to be established over 5 years, will account for about 5% of Wellcome’s total spending Flooding could during that period. jeopardize 1500 chimpanzees in a Death from nerve agent probed conservation preserve in Guinea. CHEMICAL WEAPONS | A second incident involving a Novichok nerve agent in o the United Kingdom has left one victim n J onservationists are ing proposals to build two hydro- dead—and set abuzz the community of u l y electric dams, one in West Africa and another in Indonesia, scientists who study countermeasures to 1 2 , chemical weapons. On 30 June, a couple because both could reduce habitat for critically endangered 2 fell ill in Amesbury, U.K., about 0 1 nonhuman primates—western chimpanzees and Tapanuli 12 kilometers north of Salisbury, where, 8 orangutans, respectively. In Guinea, a management board was 4 months earlier, ex-spy Sergei Skripal Cfinalizing plans this week to construct a 294-megawatt facility and his daughter Yulia spent weeks in Moyen-Bafing National Park that wildlife experts say could kill up in intensive care after exposure to a Novichok agent first developed by the to 1500 western chimpanzees (Pan troglodytes verus), a subspecies Soviet Union decades ago. U.K. authori- whose population has fallen by more than 80% over the past 20 years. ties say it’s too early to say how the Flooding from the dam and associated human activity will force some incidents are related. Dawn Sturgess, of them to move, primatologists working with the International Union 44, died on 8 July; her partner, Charlie for Conservation of Nature report. Because chimpanzees are territo- Rowley, 45, remains critically ill. Investigators have not revealed whether rial, notes one of the scientists, Rebecca Kormos of the University of the agent that poisoned them is the same California, Berkeley, deadly warfare will likely break out when one compound used against the Skripals and group is pushed onto another’s turf. Kormos and colleagues have if so, whether it is from the same batch. urged Guinea’s government to look for alternative sites for the dam or Novichok compounds stored in pow- der form are thought to be stable, but alternative energy sources. A dam planned for Sumatra in Indonesia degrade quickly when exposed outdoors. M faces similar criticism; the Tapanuli orangutan subspecies (Pongo A The notion that the same agent might M E E tapanuliensis), discovered only last year, has a remaining population have lingered in the environment for N A Y L months is “incredible, and deeply dis- A K of just 800 apes. : turbing,” a U.S.-based chemical defense O T O researcher says. H P 112 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsInBrief.indd 112 7/11/18 10:35 AM A high-fat diet may have helped Ötzi the Iceman survive at high altitudes. D o w n l o a d e d f r o BIOANTHROPOLOGY ancient DNA, proteins, and other chemicals p in his m Iceman’s last meal analyzed stomach contents. Fat residues, which came from the red deer and ibex he ate, made up a remarkably high share of the total, esearchers have provided a detailed chemical analysis of about 50%, they determined; although one meal doesn’t reveal the last meal eaten by Ötzi the Iceman—the 5300-year-old a lifetime’s diet, a high-fat diet may have given Ötzi the energy man found frozen and naturally mummified in the Italian he needed to survive at high altitudes. The iceman’s last meal Alps in 1991—and found it was rich in fat. As described was balanced with grains from einkorn wheat and traces of Rin previous accounts, Ötzi had a completely full stomach a genus of toxic ferns called bracken. Researchers say he may when he was shot with an arrow and died. In the new study, have consumed the bracken as medicine to treat parasites published 12 July in Current Biology, scientists quantified the previously identified in his intestines. o n J u l y BY THE NUMBERS the health, well-being, and economy of 1 Puerto Rico’s stats agency saved ,2 Puerto Rico and its people for decades to 2 SCIENCE POLICY | Puerto Rico has backed come,” she says. 0 1 away from a controversial plan to elimi- 40 8 nate the Puerto Rican Institute of Statistics Global science body launched (PRIS), the agency that vets data and sta- Countries in which people posted tistical analyses about the island. Governor SCIENTIFIC SOCIETIES | A new interna- on social media during the first Ricardo Rosselló had proposed bringing tional body aiming to be the “global voice week of July celebrating the PRIS under the control of the island’s for science” was launched at a meeting in first International Day o BTQ+ I Department of Economic Development Paris last week. The inauguration of the People in Science, Technology, L L E and Commerce and ultimately outsourcing International Science Council (ISC) follows Engineering, and Math on 5 July. D A M its responsibilities to the private sector, a decision last year to merge two existing A S . M as part of a broad reorganization plan. bodies representing the social and natural \ C A Critics—including AAAS, which pub- sciences—the International Social Science R U E \ lishes Science—argued that the bill would Council and the International Council for M U 500 E undermine PRIS’s political independence Science—to pool resources and break down S U M Y and the trustworthiness of Puerto Rico’s barriers between disciplines. The ISC, which G million O statistics. On 30 June, the island’s leg- will focus on supporting research related L O E A islature approved the reorganization of to international issues such as poverty and Tons of insects consumed by birds H C R the economic department but p climate change, counts more than 180 orga- per year, according to an estimate in A L O PRIS in its current form. The move won nizations as members, including the U.S. a study that examined habitats R Y T H praise from Lisa LaVange, president of National Academy of Sciences and The Royal around the world and noted the T U O the American Statistical Association in Society of the United Kingdom. At the 4 July importance of bird populations for S : O Alexandria, Virginia. “Having independent, meeting, South African mathematician Daya keeping insect pests in check T O H strong, and transparent statistics will serve Reddy was elected ISC’s first president. (The Science of Nature). P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 113 Published by AAAS DA_0713NewsInBrief.indd 113 7/11/18 10:36 AM NEWS | IN BRIEF Ancient tools found help reach the Paris climate agreement scandals. Wheeler, a polished Washington goal of keeping global warming to less than insider who was confirmed as EPA deputy ARCHAEOLOGY | Hominins lived 2°C. The church originally planned to start administrator in April, is expected to forge more than 2 million years ago, according divesting from companies by 2020 if they ahead with many of Pruitt’s initiatives, to new evidence that may push back the didn’t pass muster, but leaders this week including efforts to roll back climate and earliest dates for hominins living outside voted to delay that by 3 years to allow more wetlands protection rules. But his arrival Africa by 250,000 years. If the dates hold, time to pressure companies to change their likely marks the demise of one dream of they will boost a minority view that a key ways. The climate divestment would be the EPA foes: undermining the so-called endan- human ancestor, Homo erectus, could have largest by any church fund. germent finding, a 2009 legal opinion that arisen in Asia, not Africa. In this week’s obligates EPA to regulate carbon dioxide Nature, researchers report that they recov- Carrier ends monkey flights emissions. Wheeler has called the finding ered and analyzed 96 stone points, flakes, “settled law.” and cores dropped in a gully at Shangchen ANIMAL RESEARCH | The Russian carrier in north-central China’s windswept Loess AirBridgeCargo Airlines will no longer Newborn screening list amended Plateau. Until now, the oldest known transport nonhuman primates for research site for hominins outside of Africa was purposes, leaving Air France as the only PUBLIC HEALTH | Health and Human Dmanisi, Georgia, where fossil hominins major airline still carrying the animals. Services (HHS) Secretary Alex Azar last and stone tools date to 1.85 million years. A representative of the Moscow-based week added spinal muscular atrophy The Shangchen site lacks volcanic miner- company disclosed the decision in a 3 July (SMA) to the list of disorders for which the als, which provide the gold standard for email to People for the Ethical Treatment government recommends that newborns be radiometric dating methods. Instead, the of Animals (PETA) in Norfolk, Virginia. screened. States implement the screening D o scientists used paleomagnetic dating— In late June, PETA urged its supporters and generally follow the federal govern- w n l which detects known reversals in Earth’s to ask AirBridgeCargo to stop transport- ment’s lead; with the addition of the fatal, o a d e magnetic field that are recorded in ancient ing primates, mostly macaques, bound for inherited disease, which can be detected by d rock—to date the layers of sediment con- research labs. More than 188,000 people a genetic test, HHS now suggests screen- f r o m taining the stone tools. The oldest layers emailed the airline. It had been shipping ing for 35 such conditions. SMA destroys date from 2.1 million years ago. animals from China to Chicago O’Hare motor neurons in the brain stem and spi- International Airport. Airline officials did nal cord. In its most common, most severe Church backs climate action not respond to emails requesting comment. form, children die in the first few years of Most primates imported to the United life. In December 2016, the Food and Drug CLIMATE SCIENCE | The Church of States for research originate ; oth- Administration (FDA) approved the first England will, by 2023, begin to divest from ers arrive from Vietnam and Mauritius. In therapy, nusinersen (Spinraza), made by companies that aren’t taking enough action 2014, the last year for which the U.S. Fish Biogen of Cambridge, Massachusetts. In against climate change, leaders agreed this and Wildlife Service made such informa- February, an advisory committee had voted week. The decision follows other steps by tion available, 23,426 research primates 8 to 5 to recommend that Azar add the dis- the church to influence the energy indus- were imported. ease to the screening panel, but dissenters try by using its £12 billion investment worried about the lack of long-term data o fund, which contains about £125 million on the drug’s effectiveness and side effects n New leader at EPA J in shares of large oil and gas companies. (Science, 29 June, p. 1385). u l y In 2015, the church divested £12 million of SCIENCE POLICY | Andrew Wheeler, a 1 2 , its assets from tar sands and coal projects. former energy industry lobbyist and U.S. 2 0 Ontario head shrinks science Last year, the church teamed up with the Senate aide, became acting head of the 1 London School of Economics to create the Environmental Protection Agency (EPA) this SCIENCE POLICY | Researchers in Ontario, 8 Transition Pathway Initiative, which has week, replacing Scott Pruitt, who resigned Canada’s most populous province, are been evaluating efforts by companies to on 5 July after becoming engulfed in ethics concerned by a sudden shift in the govern- ment’s approach to science and evidence-based policy following the elec- tion of populist Doug Ford as the province’s premier. Since taking power on 29 June, Ford, the brother of the late Toronto Mayor Rob Ford, has canceled the province’s car- bon cap-and-trade system; eliminated the position of minister of science and innova- tion in his Cabinet; and fired the province’s D N chief scientist, who had been on the job C- N only since November 2017. “It’s unfortunate Y- B to see Premier Ford shift robust evidence C C / R and science-based policy down his list of K C I L priorities,” said Kathleen Walsh, director F / E P of policy at the science campaign group P O H Evidence for Democracy in Ottawa. R E G R I B : SCIENCEMAG.ORG/NEWS O T O Canterbury Cathedral is the seat of the Church of England, which is mulling divesting funds over climate change. Read more news from Science online. H P 114 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsInBrief.indd 114 7/11/18 10:36 AM I N D E P T H The IceCube Laboratory at the South Pole operates a neutrino detector 1.5 kilometers beneath the ice. ASTROPHYSICS Ice reveals a messenger from a blazing galaxy D o w n l Neutrino astronomy is born as IceCube and other instruments find likely home of particle o a d e d f r o m By Daniel Clery Earth come from? Known as ultrahigh-en- rect it will be revolutionary, extraordinary,” ergy cosmic rays, these particles have a mil- says Eli Waxman of the Weizmann Institute f astronomers are right, a ghostly lion times more energy than has ever been of Science in ovot, Israel. But, he adds, particle that lit up an instrumented produced in an earthbound particle accel- “an extraordinary result needs extraordi- swathe of ice beneath the South Pole on erator, but what boosts them to such colossal nary support, and the support is not quite 22 September last year was a messenger energies is unknown. Suspects have included extraordinary yet.” from a distant galaxy. The particle was neutron stars, gamma-ray bursts, hyperno- Completed in 2010, the IceCube neu- Ia neutrino, electrically neutral and al- vae, and the radiation-spewing black holes at trino detector snares these elusive par- most massless, which means its path could the center of some galaxies, but whatever the ticles in a cubic kilometer of Antarctic E be traced back to the extragalactic event source, high energy neutrinos are a likely by- ice. When a neutrino hits a nucleus in the C N E I that created it. Cued by IceCube, the Ant- product (Science, 31 July 2015, p. 465). If the frozen water molecules, other particles fly C S / L arctic detector, the orbiting Fermi Gamma- IceCube team is right, blazars could be the off in recoil; as they decelerate, they emit E K o n C I ray Space Telescope found that the neutrino first confirmed source of these cosmic rays. light called Cherenkov radiation, which B . J u C likely came from a far off blazar, a hugely Researchers note, however, that the some of IceCube’s 5160 light detectors may l Y B y D E bright source of radiation powered by a link between the neutrino and the blazar pick up. Based on the location, timing, and 1 2 T P , A supermassive black hole. isn’t rock-solid. “It’s a very mouthwatering brightness of the detected light, research- 2 D 0 A ) Astronomers have long been tantalized by observation and I very much hope it will be ers can reconstruct the path and energy of 8 1 1 0 8 2 the prospect of using neutrinos, which move confirmed,” says Pierre Sokolsky of the Uni- the neutrino. ( 7 4 1, at almost the speed of light and rarely inter- versity of Utah in Salt Lake City. “If their Most of the neutrinos detected by IceCube 1 6 3 . act with other matter, to learn about violent interpretation of those observations is cor- originated nearby, spawned by cosmic rays L O V, cosmic events. The new find- hitting Earth’s upper atmo- E C ing, reported on pages 146 and sphere. IceCube researchers N E I C 147, could mark the founding eliminate those using a variety S, A polar light show with cosmic origins . L A T event of neutrino astronomy. The detection, and computed path, of neutrino IceCube-170922A. Each circle repre- of methods, leaving the very E T The detection also triggered a sents one of IceCube’s spherical light detectors in the Antarctic ice: size indicates the few, very high energy neutri- N A R powerful example of another brightness detected, from earliest (in dark blue), to latest (in yellow). nos, above 30 trillion electron G ) C I new trend, multimessenger volts (TeV). In 2013, the IceCube H P A astronomy, in which tele- team first revealed a handful of R G ( ; scopes and other instruments such events, arguing that their F S N / studied the flaring blazar in all high energies and other prop- E B U parts of the electromagnetic erties showed they must have C E C I F, spectrum, from gamma rays to Side view come from outside of our gal- L O radio waves. axy (Science, 22 November 2013, W N I A neutrino-producing blazar p. 920). The detector contin- T R A M could also help solve a decades- ues to bag about a dozen high ) O T old mystery in astronomy: energy neutrinos a year; when O H P ( Where do the extremely high it gets a clean track with a : S T I energy protons and other nu- Nanoseconds 125 m well-defined direction, other D E R clei that occasionally bombard 0 500 1000 1500 2000 2500 3000 telescopes scramble to see if C SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 115 Published by AAAS DA_0713NewsInDepth.indd 115 7/11/18 10:33 AM NEWS | IN DEPTH there is an obvious cosmic source—until now, without success. In 2016 IceCube’s operators set up an alert service, with the hope of getting more telescopes at different wavelengths involved in the hunt. Then, last September, IceCube got lucky. A detected neutrino, dubbed IceCube-170922A and calculated to have an energy of 290 TeV, offered a relatively clear track back into space. An automatic alert went out less than a minute later. Several observatories initially didn’t see anything unusual. Six days later, the Fermi team reported the satellite had found that a blazar, known as TXS 0506+056 and just 0.1° away from the neutrino track suggested by IceCube, was especially bright, having started flaring a few months earlier. Soon, more than a dozen telescopes had studied the blazar. Blazars, like quasars, are distant SPACE cosmic beacons powered by supermassive D o black holes, which generate intense radia- w n l tion and fire jets of particles from their poles. China announces new flotilla o a d e Blazars are exceptionally bright, astrono- d mers believe, because their jets happen to be f r of space science missions o m aimed straight at Earth. IceCube and the other observers estimate the probability that the neutrino path and Probes will watch for colliding black holes, image Earth’s the blazar coincided by chance is roughly magnetosphere, and monitor the sun one in 740. Physicists and astronomers, however, aren’t usually convinced that two phenomena are connected until there’s no By Dennis Nor Xiong Shaolin, an astrophysicist at CAS’s more than a one in 3.5 million, or 5-sigma, Institute of High Energy Physics in Beijing, probability of a coincidence. hina’s ambitious human space mis- and his colleagues identified a month after IceCube researchers also went back sions get most of the headlines, but the U.S. Laser Interferometer Gravitational- through almost a decade of data to see its fledgling space science program is Wave Observatory announced its historic whether an excess of high-energy neutri- quietly gaining strength. The Chinese detection of gravitational waves in Febru- o nos had streamed from the same location Academy of Sciences (CAS) last week ary 2016. They proposed putting two satel- n J before. They found a period of 150 days in Cconfirmed plans to launch four new lites into orbit on opposite sides of Earth u l y late 2014 and early 2015 when IceCube de- scientific satellites beginning in 2020. Com- that together could watch the entire sky 1 2 , tected around 13 more neutrinos than nor- ing on the heels of four successful missions, for gamma rays emanating from the events 2 mal from that spot. It’s not yet clear whether including one devoted to x-ray astronomy that generate gravitational waves. Funding 0 1 TXS 0506+056 was flaring at that time, but and another that demonstrated quan- for technical studies arrived a few months 8 “the archival event was much more inter- tum entanglement over a record-setting later, and the mission has jumped to the esting” than the recent detection, says Ice- 1200 kilometers, these “phase 2” projects front of the launch queue, with a date of Cube Principal Investigator Francis Halzen will examine areas including solar physics 2020. “When you have this kind of oppor- of the University of Wisconsin in Madison. and the hunt for electromagnetic signals as- tunity you can’t handle it like a normal mis- Sokolsky and Waxman agree that Ice- sociated with gravitational waves. sion, with selection and review taking 10 or Cube’s September 2017 detection should Given that China’s space science program 20 years,” Xiong says. strengthen the project’s longstanding bid only started about 10 years ago, the length- So far, gamma rays and other electro- to massively increase the size of the instru- ening track record “is impressive, but there magnetic signals have only been detected ment, which would also increase how many are still not many missions given that it’s a from one kind of gravitational-wave source, a neutrinos it can detect and improve its big country with a big science community,” neutron star merger, but they yielded a trove S pointing accuracy. Since IceCube was built, says Xin Wu, a China-born physicist at the of detail about the enigmatic event. Astro- A C , the team has found the ice is clearer than University of Geneva in Switzerland who physicists are still debating whether black S C I S previously thought, so they believe they can collaborates on China’s astrophysics mis- hole mergers, the other confirmed source Y H P Y make IceCube 10 times bigger while only sions. “There is pent-up demand” among of gravitational waves, also produce electro- G R E doubling the number of light detectors, Chinese space scientists, he says. magnetic emissions. The GECAM team is N E H matching the $280 million cost of its origi- CAS broke with tradition for one of betting that they do—and that much can be G I H nal construction. The team is about to start the new missions, the Gravitational Wave learned from the signals. “I think probably F O E experiments to test that. “With a 10-times High-energy Electromagnetic Counterpart we will find something,” Xiong says. T U T I T bigger detector, the answer [to whe igh All-sky Monitor (GECAM). It fast-tracked GECAM’s observations will complement S N I : energy neutrons come from] would be clear selection and development to take advan- those of another phase 2 mission, the Ein- E G A and obvious,” Waxman says. j tage of a new scientific opportunity, which stein Probe (EP), which will survey the sky M I 116 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsInDepth.indd 116 7/11/18 10:33 AM Two satellites, launching in 2020, will watch for gamma TOXICOLOGY rays from the violent birth of gravitational waves. for the low energy x-rays associated with Digital chemical test impresses violent phenomena such as gamma ray bursts and black hole collisions. Combining Giant database shows promise for replacing animal studies GECAM, EP, and gravitational wave obser- vations “will allow us to better understand gamma ray burst astrophysics,” says Ik By Vanessa Zainzinger with similar structures can have similar Siong Heng, an astrophysicist at the Uni- health effects, such as being an irritant versity of Glasgow in the United Kingdom. oxicologists this week unveiled a digi- (Science, 12 February 2016, p. 651). China’s space scientists have long tar- tal chemical safety screening tool that In their new paper, Hartung’s team goes geted another area: solar physics. Only the could greatly reduce the need for six bigger. First, the researchers expanded United States produces more papers in the common animal tests. Those tests ac- their database to 10 million chemical struc- field than China. “But [China’s] papers used count for nearly 60% of the estimated tures by adding information from the pub- data from missions developed by Japan and T3 million to 4 million animals used lic database PubChem and the U.S. National the U.S. and elsewhere,” says Gan Weiqun, annually in risk testing worldwide. Toxicology Program. Next, they compared a solar physicist at CAS’s Purple Mountain The computerized tool—built on a mas- the structures and toxicological proper- Observatory in Nanjing. He says China’s sive database of molecular structures and ties of every possible pair of compounds solar scientists have been pushing for their existing safety data—appears to match, and in their database—a total of 50 trillion own mission for 40 years; they’ve finally sometimes improve upon, the results of ani- comparisons—creating a vast similarity D o gotten the nod for the Advanced Space- mal tests for properties such as skin sensi- map that groups compounds by structure w n l based Solar Observatory (ASO-S). “It’s very tization and eye irritation, the researchers and effect. Finally, they tested the model: o a d e important for us to make original contribu- reported in the 11 July issue of Toxicologi- They asked it to predict a randomly cho- d tions in terms of hardware and data,” Gan cal Sciences. But it also has limitations; for sen chemical’s toxicological profile by link- f r o m says. He explains that ASO-S will be the first instance, the method can’t reliably evaluate ing it to similar “neighbors” on the map, space observatory to monitor the sun’s mag- a chemical’s risk of causing and compared the results netic field while watching for solar flares cancer. And it’s not clear to six actual animal tests of and the titanic blasts known as coronal how open regulatory agen- the compound. mass ejections. The simultaneous observa- cies will be to adopting a “This won’t On average, the computa- tions could yield clues to how those erup- nonanimal approach. be the end of tional tool reproduced the tions are triggered. Still, “We’re really ex- animal test results 87% of The last mission included in the phase 2 cited about the potential animal testing.” the time. That’s better than list was identified as a priority years ago of this model,” says toxico- Mike Rasenberg, animal tests themselves can (Science, 22 July 2016, p. 342). The Solar logist Nicole Kleinstreuer, European Chemicals Agency do, Hartung says: In review- wind Magnetosphere Ionosphere Link Ex- the deputy director of a ing the literature, his group plorer (S ), a joint CAS and European center that evaluates alternatives to animal found that repeated animal tests replicated o Space Agency mission, will pioneer a new testing at the National Institute of Environ- past results just 81% of the time, on average. n J technique for imaging Earth’s magneto- mental Health Sciences (NIEHS) in Dur- “This is an important finding,” Hartung says, u l y sphere. Previous satellites have made point ham, North Carolina. Kleinstreuer, who was because regulators often expect alternative 1 2 , measurements as they traveled through not involved in the work, adds that using methods to animal testing to be reproduc- 2 the magnetosphere. But scientists recently “big data … to build predictive models is an ible at the 95% threshold—a standard even 0 1 learned that collisions between particles of extremely promising avenue for reducing the animal tests aren’t meeting. 8 the solar wind and stray particles of Earth’s and replacing animal testing.” The screening method has weaknesses. atmosphere produce low energy x-rays that Most developed nations require new Although it can predict simple effects such light up the magnetosphere. By watch- chemicals that enter commerce to undergo as irritation, more complex endpoints such ing these x-rays, S will capture its at least some safety testing. But the long- as cancer are out of its reach, says Mike dynamic behavior. standing practice of exposing rabbits, rats, Rasenberg, who heads the ECHA’s Compu- Last week’s announcement bodes well and other animals to chemicals to evaluate tational Assessment Dissemination unit. for China’s space science program beyond risks is facing growing public objections “This won’t be the end of animal testing,” the next four missions. The 4 billion yuan and cost concerns, helping spur a hunt he predicts. Still, Rasenberg thinks Euro- ($605 million) phase 2 budget includes sup- for alternatives. pean regulators will accept the method for port for development on future missions, One approach is to use what is known simple endpoints. particularly the enhanced X-ray Timing about the safety of existing chemical In the United States, the NIEHS center and Polarimetry mission (Science, 9 March, compounds to predict the risks posed is validating the method, and Kleinstreuer p. 1085), an ambitious international project by new chemicals with similar molecu- says if it passes, she could see U.S. regula- led by Chinese scientists to study black holes, lar structures. Two years ago, a team led tors adopting it for some purposes. In the neutron stars, and magnetars. by Thomas Hartung of the Johns Hop- meantime, the researchers are moving to China’s planetary exploration and astro- kins Bloomberg School of Public Health make the tool available to companies that naut programs will continue to make history; in Baltimore, Maryland, took a step to- might want to screen products before sub- later this year or early next, for example, it ward that goal by assembling test data mitting them for regulatory review. j plans to land the first probe on the far side of on 9800 chemicals regulated by the the moon. But the future of its space science European Chemicals Agency (ECHA) in Vanessa Zainzinger is a science journalist efforts also seems assured. j Helsinki. They then showed that chemicals in the United Kingdom. SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 117 Published by AAAS DA_0713NewsInDepth.indd 117 7/11/18 10:33 AM The genome editor CRISPR can be used to engineer female lab mice that have increased odds of passing down a specific gene to offspring. on during meiosis, the cell division process that helps creates sperm or eggs. The chro- mosomes naturally swap DNA during meio- sis, and during those exchanges the cell only allows HDR. The experiment did not work in males, likely because spermatogonia go through mitotic cell division before meiosis, stymy- ing HDR. But in females, the gene drive succeeded. It copied the coat color modi- fying gene to the partner chromosome in many eggs, which would significantly raise the odds of offspring inheriting it. In one GENOME EDITING mouse, 79% of her eggs ended up with the color modifying gene on both chromosomes. If she mated with a male without the gene, A ‘gene drive’ makes its about 90% of her pups would inherit the D o gene. (Flies have a different process of em- w n l bryogenesis, one that boosts the efficiency o a d debut in mammals e of HDR and allows it to work in both sexes.) d Cooper and her colleagues write that this f r CRISPR-based gene-spreading strategy promises o m system of “active genetic elements” could to speed development of engineered mice speed the creation of mice that have sev- eral introduced or crippled genes. Michael Wiles, who heads tech evaluation and de- By Jon Cohen discussion. “It’s a very good study and it’s of velopment at the Jackson Laboratory in Bar pretty high significance,” says Gaétan Burgio, Harbor, Maine—one of the world’s largest esearchers have used CRISPR, the a mouse geneticist at the John Curtin School commercial producers of engineered mice— genome editing tool, to speed the of Medical Research in Canberra, Australia, says the method could be “very useful,” as inheritance of specific genes in who tweeted a series of comments about the many human diseases are caused by aberra- mammals for the first time. Demon- report. “Nothing is really known about gene tions in several genes, and producing mice strated in lab-reared insects several drives in rodents. We all assumed the effi- that genetically mimic those diseases is slow Ryears ago, this controversial “gene- ciency would be the same as in flies, but it and laborious. “I’m now getting requests to o drive” strategy offers the ability to quickly turns out to be very different.” make mice with six modifications, and the n J spread a gene throughout an entire species. In a variation of the earlier fly work, the breeding time becomes phenomenal,” Wiles u l y It has sparked dreams of deploying lethal UCSD researchers built their gene drive by says. With a gene drive like this, what takes 1 2 , genes to eradicate pests such as malaria- engineering female mice to carry the gene 5 years could be done in one, he says. 2 carrying mosquitoes—and now, perhaps, for the DNA-cutting enzyme Cas9, one of Even though this new work aims only to 0 1 crop-damaging, disease-causing mammals CRISPR’s two components; they engineered engineer lab mice, Kevin Esvelt, an evolu- 8 such as rabbits, mice, and rats. males to carry a gene for the other compo- tionary biologist at the Massachusetts Insti- The new research aims to create novel nent, the guide RNA (gRNA) that shuttles tute of Technology Media Lab in Cambridge, strains of lab mice, not wipe out wild popu- Cas9 to a specific target on a genome, plus says it concerns him. He believes a gene lations, and it shows that gene drives work a gene that modifies coat color. Breeding drive that could be released into the wild less efficiently in rodents than in insects. the altered mice created pups that had the should include a kill switch to shut it down Still, Paul Thomas, a molecular geneticist at genes for both CRISPR components on dif- and restore animals to their natural state. the University of Adelaide in Australia who ferent chromosomes. “It’s troubling to see that the study does not is working on similar experiments, calls it After Cas9 makes its cut, a cell repairs the explicitly mention safeguards,” Esvelt says. an “important first step towards develop- damage—and how it does so is key to the The new UCSD gene drive, however, ment of gene drive technology in mammals.” success of a gene drive. The cell can either likely would stop spreading in a mouse pop- The study was posted 4 July on bioRxiv, an reconnect the severed strands of DNA or ulation after a few generations. Because the online site for preprints, by a team at the Uni- bridge the gap by slotting in chunks of new genes for Cas9 and the gRNA ride on dif- versity of California, San Diego (UCSD), led DNA, a process called homology directed ferent chromosomes, they would gradually R I by geneticist Kimberly Cooper. The research- repair (HDR). A gene drive harnesses HDR become separated and the drive would lose M E D ers, among them Ethan Bier and Valentino to insert a new gene—in this case, a copy of effectiveness. In the preprint, Cooper and M E K Gantz, who 3 years ago showed that CRISPR the coat-color modifier gene. But cells natu- her colleagues stress the continuing chal- R O G / could create an efficient gene drive in flies rally prefer to simply reconnect the severed lenges of creating an efficient gene drive for M O C (Science, 24 April 2015, p. 442), have submit- DNA, which thwarts a gene drive. wild mammals. “[T]he optimism and con- K. C O ted the study to a peer-reviewed journal that The UCSD team exploited a fundamental cern that gene drives may soon be used to T S I : asked them not to speak with media. But biological phenomenon to force cells to- reduce invasive rodent populations in the O T O it has already triggered plenty of scientific ward HDR. They manipulated Cas9 to turn wild is likely premature,” they conclude. j H P 118 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsInDepth.indd 118 7/11/18 10:33 AM NEWS | IN DEPTH ARCHAEOLOGY Molecular ‘barcodes’ reveal lost whale hunts DNA method can rapidly identify species, shows toll of ancient cultures on biodiversity By Elizabeth Pennisi and Michael Price applied their method to 5000 bone frag- Romans fished tuna from villages dating ments from 38 sites in New Zealand span- to between 400 B.C.E. and 425 C.E. Cryp- ccording to myth, the Maori people ning the past 20,000 years. “It’s bones that tic texts referring to “sea monsters” or arrived in New Zealand riding the never would have been identified,” says “ram fishes,” plus large bones suspected back of a whale; today, whales fig- Rodrigues, who was not part of this study. to be whale (and often shaped into tools), ure prominently in Maori art and Seersholm pinpointed 110 near-perfect suggested the Romans also hunted coast- stories. Now it appears that Maori matches to known species, as the group hugging gray and right whales. But these Aancestors may also have systemati- reported this week in the Proceedings of behemoths don’t ply those waters today. cally hunted the animals. the National Academy of Sciences. Until Rodrigues’s team combined tried-and- Archaeological sites in New Zealand hold now, archaeologists usually assumed that true barcoding with collagen fingerprinting, few recognizable whale bones, but a new whale bones at sites in New Zealand were a technique that matches amino acids in technique for detecting species from short from pilot whales that had been scavenged samples’ connective tissue to known animal snippets of DNA in bone scraps has enabled after they beached. But the DNA barcodes families. They found that 10 of the 11 bones D o researchers to identify the remains of five identified orca, true dolphins, Cuvier’s belonged to whales—three gray, three right, w n l whale species in early New Zealand settle- beaked whales, fin whales, and southern one fin, one pilot, one sperm, and one dol- o a d e ments, including smaller or slow-moving right whales. “We think they we unting phin. The preponderance of gray and right d species, the likely targets for whales, which are reliably found f r o m hunters. The same “DNA meta- close to shore and thus could have barcode” method showed that the been hunted, raises the possibil- early New Zealanders may also ity of a forgotten 2000-year-old have decimated populations of Roman whaling industry, the a native parrot. And on the op- researchers suggest this week posite side of the world, a team in the Proceedings of the Royal used a simpler barcode method to Society B. That’s about 1000 years suggest that the Romans hunted before Basque whalers, long whales too, potentially push- considered the first commercial ing the first known commercial whalers, emerged. whaling industry back by about a If the Romans did hunt gray thousand years. and right whales from Gibraltar, o By revealing species from once- the authors note, they may have n J meaningless bone fragments, bar- played a role in the species’ disap- u l y codes are “opening new windows pearance from the eastern North 1 2 , into historical ecosystems” says Atlantic. But other scholars say 2 Ana Rodrigues, an ecologist at it’s more likely that the Romans 0 1 the University of Montpellier in scavenged beached whales. 8 France. And the new data are The identification of the 11th changing views. “For a long time, Scientists retrieved the DNA signatures of whales and other species from bone presumed whale bone threw the we thought that ancient cultures fragments like these in New Zealand. researchers for a loop: It belonged were technologically not advanced to an elephant. That’s “why it’s enough to significantly influence animal the smaller whales,” driving the mammals good to use these methods,” Rodrigues says. populations, but more and more evidence into shallow water with boats and then har- “Who knows how many other bones out is accumulating that they actually did,” says pooning them, Seersholm says. the ave been misidentified?” S E C evolutionary biologist Michael Hofreiter of The barcodes also suggested the early Barcoding has limitations. It can’t iden- N E I C the University of Potsdam in Germany. New Zealanders hunted fur seals, sea lions, tify a species if its sequence isn’t included S F O At Curtin University in Perth, Australia, and elephant seals. And the method identi- in the barcode database, which is rap- Y M E DNA researcher Michael Bunce, graduate fied several terrestrial species including the idly expanding but still incomplete. And D A - - - C student Frederik Seersholm, and colleagues kakapo, a now-endangered ground-dwelling barcodes can’t yet show how abundant a A L A have been assessing the impact of human New Zealand parrot. Further analysis of species was—only that it was present. N O I - - - T arrival in New Zealand in about 750 B.C.E. kakapo DNA showed that its genetic diver- But Rodrigues, for one, is in awe of bar- A N E They developed a way to cheaply and easily sity plunged once New Zealand was settled. coding’s power. “It’s like a miracle,” she says. H T F determine the DNA barcodes of hundreds “It really shows that high impact humans “You send them a little bit of dust and they O S G of bones by high throughput sequencing had on the biodiversity,” Seersholm says. send you back results.” j N I D E of blends of ground-up bones. Barcodes, In the Gibraltar region by the Mediter- E C O named after the ubiquitous labels on items ranean Sea, Rodrigues and her colleagues Michael Price is a freelance journalist R P : O in stores, are short stretches of DNA that took a focused approach, analyzing 11 puta- and a magazine editor at San Diego T O H uniquely identify a species. Bunce’s team tive whale bones from five sites. There, the State University in California. P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 119 Published by AAAS DA_0713NewsInDepth.indd 119 7/11/18 10:33 AM FEATURES D o w n l o a d e d f LIQUID r o m SUNSHINE Ammonia made from sun, air, o n and water could turn Australia into J u l a renewable energy superpower y 1 4 , 2 0 1 8 By Robert F. Service Yet with a small population and few ways is breathing nitrogen in and breathing am- in Sydney, Brisbane, to store or export the energy, its renewable monia out,” MacFarlane says, beaming like and Melbourne, Australia bounty is largely untapped. a proud father. Y T E That’s where MacFarlane comes in. For Companies around the world already I C O S the past 4 years, he has been working on produce $60 billion worth of ammonia C I H he ancient, arid landscapes of Aus- a fuel cell that can convert renewable elec- every year, primarily as fertilizer, and P A R tralia are fertile ground for new tricity into a carbon-free fuel: ammonia. MacFarlane’s gizmo may allow them to G O T growth, says Douglas MacFarlane, Fuel cells typically use the energy stored make it more efficiently and cleanly. But he O H P L a chemist at Monash University in chemical bonds to make electricity; has ambitions to do much more than help A Y O R in suburban Melbourne: vast for- MacFarlane’s operates in reverse. In his farmers. By converting renewable electric- E H T ests of windmills and solar panels. third-floor laboratory, he shows off one of ity into an energy-rich gas that can easily F O More sunlight per square meter the devices, about the size of a hockey puck be cooled and squeezed into a liquid fuel, W O L strikes the country than just about and clad in stainless steel. Two plastic tubes MacFarlane’s fuel cell effectively bottles L E F / Tany other, and powerful winds buf- on its backside feed it nitrogen gas and wa- sunshine and wind, turning them into a N O T fet its south and west coasts. All told, Aus- ter, and a power cord supplies electricity. commodity that can be shipped anywhere R O M tralia boasts a renewable energy potential Through a third tube on its front, it silently in the world and converted back into N E V E of 25,000 gigawatts, one of the highest in exhales gaseous ammonia, all without the electricity or hydrogen gas to power fuel T S : the world and about four times the planet’s heat, pressure, and carbon emissions nor- cell vehicles. The gas bubbling out of the O T O installed electricity production capacity. mally needed to make the chemical. “This fuel cell is colorless, but environmentally, H P 120 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsFeatures.indd 120 7/11/18 10:27 AM NEWS A component in a reverse fuel cell uses Energy in Pullenvale. “It’s the bridge to a Most is used as fertilizer. Plants crave ni- renewable power to knit together whole new world.” trogen, used in building proteins and DNA, water and nitrogen to make ammonia. First, however, the evangelists for renew- and ammonia delivers it in a biologically able ammonia will have to displace one of available form. Haber–Bosch reactors can MacFarlane says, ammonia is as green as the modern world’s biggest, dirtiest, and churn out ammonia much faster than nat- can be. “Liquid ammonia is liquid energy,” most time-honored industrial processes: ural processes can, and in recent decades he says. “It’s the sustainable technology something called Haber–Bosch. the technology has enabled farmers to feed we need.” the world’s exploding population. It’s esti- Ammonia—one nitrogen atom bonded THE AMMONIA FACTORY, a metallic metrop- mated that at least half the nitrogen in the to three hydrogen atoms—may not seem olis of pipes and tanks, sits where the red human body today comes from a synthetic like an ideal fuel: The chemical, used in rocks of Western Australia’s Pilbara Desert ammonia plant. household cleaners, smells foul and is toxic. meet the ocean. Owned by Yara, the world’s Haber–Bosch led to the Green Revolu- But its energy density by volume is nearly biggest producer of ammonia, and com- tion, but the process is anything but green. double that of liquid hydrogen—its primary pleted in 2006, the plant is still gleaming. It It requires a source of hydrogen gas (H ), 2 competitor as a green alternative fuel— is at the technological vanguard and is one which is stripped away from natural gas or and it is easier to ship and distribute. “You of the largest ammonia plants in the world. coal in a reaction using pressurized, super- can store it, ship it, burn it, and convert it Yet at its core are steel reactors that still use heated steam. Carbon dioxide (CO ) is left 2 back into hydrogen and nitrogen,” says Tim a century-old recipe for making ammonia. behind, accounting for about half the emis- Hughes, an energy storage researcher with Until 1909, nitrogen-fixing bacteria made sions from the overall process. The second manufacturing giant Siemens in Oxford, U.K. “In many ways, it’s ideal.” D o Researchers around the globe are chasing w n l the same vision of an “ammonia economy,” o a d e and Australia is positioning itself to lead it. d “It’s just beginning,” says Alan Finkel, Aus- f r o m tralia’s chief scientist who is based in Can- berra. Federal politicians have yet to offer any major legislation in support of renew- able ammonia, Finkel says, perhaps un- derstandable in a country long wedded to exporting coal and natural gas. But last year, the Australian Renewable Energy Agency declared that creating an export economy for renewables is one of its priorities. This year, the agency announced AU$20 million in initial funds to support renewable export technologies, including shipping ammonia. o In Australia’s states, politicians see re- n J newable ammon a potential source u l y of local jobs and tax revenues, says Brett 1 4 , Cooper, chairman of Renewable Hydrogen, 2 a renewable fuels consulting firm in Syd- 0 1 ney. In Queensland, officials are discussing 8 creating an ammonia export terminal in Australia’s windy coasts offer a bounty of energy, which it might one day export as a carbon-free fuel. the port city of Gladstone, already a hub for shipping liquefied natural gas to Asia. most of the ammonia on the planet. But that feedstock, N , is easily separated from air, 2 In February, the state of South Austra- year, German scientist Fritz Haber found which is 78% nitrogen. But generating the lia awarded AU$12 million in grants and a reaction that, with the aid of iron cata- pressure needed to meld hydrogen and loans to a renewable ammonia project. lysts, could split the tough chemical bond nitrogen in the reactors consumes more And last year, an international consortium that holds together molecules of nitrogen, fossil fuels, which means more CO . The 2 announced plans to build a US$10 billion N , and combine the atoms with hydrogen emissions add up: Ammonia production 2 combined wind and solar plant known to make ammonia. The reaction takes brute consumes about 2% of the world’s energy A I L as the Asian Renewable Energy Hub in force—up to 250 atmospheres of pressure and generates 1% of its CO . A R 2 T S Western Australia state. Although most of in the tall, narrow steel reactors—a process Yara is taking a first step toward greening U A H the project’s 9000 megawatts of electric- first industrialized by German chemist Carl that process with a pilot plant, set to open T U O ity would flow through an undersea cable Bosch. The process is fairly efficient; about in 2019, that will sit next to the existing Pil- S , D R to power millions of homes in Indonesia, 60% of the energy put into the plant ends bara factory. Instead of relying on natural A O B N some of that power could be used to gen- up being stored in the ammonia’s bonds. gas to make H , the new add-on will feed 2 O I T erate ammonia for long-distance export. Scaled up to factories the size of Yara’s, the power from a 2.5-megawatt solar array into C E T “Ammonia is the key enabler for exporting process can produce vast amounts of am- a bank of electrolyzers, which split water O R P T renewables,” says David Harris, research monia. Today, the facility makes and ships into H and O . The facility will still rely on S 2 2 A O director for low-emissions technologies at 850,000 metric tons of ammonia per year— the Haber–Bosch reaction to combine the C : O Australia’s Commonwealth Scientific and more than double the weight of the Empire hydrogen with nitrogen to make ammonia. T O H Industrial Research Organisation (CSIRO) State Building. But the solar-powered hydrogen source P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 121 Published by AAAS DA_0713NewsFeatures.indd 121 7/11/18 10:27 AM N EWS | FEATURES A green way to make ammonia cuts total CO2 emissions from the process Reverse fuel cells can use renewable power to make ammonia from air and water, a far more roughly in half. environmentally friendly technique than the industrial Haber–Bosch process. Renewable ammonia Other projects are following suit. The could serve as fertilizer—ammonia’s traditional role—or as an energy-dense fuel. state of South Australia announced plans in February to build a AU$180 million am- Industrial ammonia monia plant, again relying on electrolyzers Most of the world’s ammonia is synthesized using Haber–Bosch, a century-old process that powered by renewable energy. Slated to is fast and fairly efficient. But the factories emit vast amounts of carbon dioxide (CO ). 2 open in 2020, the plant would be a regional CO2 source of fertilizer and liquid ammonia, which can be burned in a turbine or run H2 CO2 through a fuel cell to make electricity. The Steam Output supply of liquid energy will help stabilize Natural gas reformation NH3 Efciency the grid in South Australia, which suffered CO2 a debilitating blackout in 2016. N2 Ammonia made this way should attract Air Air separation unit High temperature buyers in places such as the European and pressure Union and California, which have cre- ated incentives to buy greener fuels. And as the market grows, so will the distribu- Gentler reactions tion routes for importing ammonia and the D A reverse fuel cell uses renewable electricity to drive a chemical reaction that makes ammonia. Water technologies for using it, Harris says. By + o reacts at the anode to make hydrogen ions (H ), which migrate to the cathode where they react with then, fuel cells like MacFarlane’s could be w nitrogen (N ) to form ammonia. The reaction is efficient, but slow. n l o 2 ready to displace Haber–Bosch itself—and a d e the half-green approach to ammonia pro- d duction could become fully green. f r o Wind m Power INSTEAD OF APPLYING fearsome heat and Output pressure, reverse fuel cells make ammonia NH3 Efciency by deftly wrangling ions and electrons. As Solar O2 CO2 in a battery being charged, charged ions flow between two electrodes supplied with electricity. The anode, covered with a catalyst, splits water molecules into O , 2 hydrogen ions, and electrons. The protons flow through an electrolyte and a proton- Liquid Separator permeable membrane to the cathode, while electrolyte membrane the electrons make the journey through e– e– o a wire. At the cathode, catalysts split N n 2 J molecules and prompt the hydrogen ions u l y and electrons to react with nitrogen and 1 4 , make ammonia. 2 N2 + + At present, the yields are modest. At room 0 1 H O H H 2 temperature and pressure, the fuel cell reac- 8 tions generally have efficiencies of between Anode Cathode 1% and 15%, and the throughput is a trickle. But MacFarlane has found a way to boost efficiencies by changing the electrolyte. Low temperature and pressure In the water-based electrolyte that many groups use, water molecules sometimes react with electrons at the cathode, stealing Ammonia made near farms electrons that would otherwise go into mak- Fertilizer ing ammonia. “We’re constantly fighting having the electrons going into hydrogen,” MacFarlane says. To minimize that competition, he opted NH3 for what’s called an ionic liquid electro- lyte. That approach allows more N2 and E C Ammonia power plant N less water to sit near the catalysts on the E I C Transportation S cathode, boosting the ammonia produc- / N A To market I tion. As a result, the efficiency of the fuel N Ammonia is more than fertilizer. The gas liquefies N U 2 O cell skyrocketed from below 15% to 60%, T L easily under light pressure and chilling, and can be A. V transported to power plants to generate carbon-free he and his colleagues reported last year in : H C 2 I Cracking Energy Environmental Science . The re- H electricity. It can also be “cracked” into H , a valuable P 2 A energy source for fuel cell vehicles. Fuel cell vehicle sult has since improved to 70%, MacFarlane R G 1 22 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713NewsFeatures.indd 122 7/11/18 10:27 AM says—but with a tradeoff. The ionic liquid Soloveichik, a manager in Washington, D.C., switched on, the reactor will “crack” am- in his fuel cell is goopy, 10 times more for the DOE’s Advanced Research Projects monia into its two constituents: H , to be 2 viscous than water. Protons have to slog Agency-Energy program on making renew- gathered up for sale, and N , to waft back 2 their way to the cathode, slowing the rate able fuels, agrees. “To make [green] ammonia into the air. of ammonia production. “That hurts us,” is not hard,” he says. “Making it economically That reactor is basically a larger version MacFarlane says. on a large scale is hard.” of Giddey’s membrane reactor, operating To speed things up, MacFarlane and his in reverse. Only here, gaseous ammonia is colleagues are toying with their ionic liq- HOWEVER DISTANT, the prospect of Asia- piped into the space between two concen- uids. In a study published in April in ACS bound tankers, full of green Australian am- tric metal tubes. Heat, pressure, and metal Energy Letters, they report devising one monia, raises the next question. “Once you catalysts break apart ammonia molecules rich in fluorine, which helps protons pass get ammonia to market, how do you get and push hydrogen atoms toward the more easily and speeds ammonia produc- the energy out of it?” asks Michael Dolan, tube’s hollow core, where they combine to tion by a factor of 10. But the production a chemist at CSIRO Energy in Brisbane. make H that’s sucked out and stored. 2 rate still needs to rise by orders of magni- The simplest option, Dolan says, is to Ultimately, Dolan says, the reactor will tude befo is cells can meet targets, set use the green ammon fertilizer, like produce 15 kilograms per day of 99.9999% for the field by the U.S. Department of En- today’s ammonia but without the carbon pu ydrogen, enough to power a few fuel ergy (DOE), that would begin to challenge penalty. Beyond that, ammonia could be cell cars. Next month, he plans to demon- Haber–Bosch. converted into electricity in a power plant strate the reactor to automakers, using it Next to Monash University, Sarb Gid- customized to burn ammonia, or in a tradi- to fill tanks in a Toyota Mirai and Hyundai dey and his colleagues at the Clayton tional fuel cell, as the South Australia plant Nexo, two fuel cell cars. He says his team is offices of CSIRO Energy are making am- plans to do. But currently, ammonia’s high- in late-stage discussions with a company to D o monia with their “membrane reactor.” It est value is as a rich source of hydrogen, build a commercial pilot plant around the w n l relies on high temperatures and modest used to power fuel cell vehicles. Whereas technology. “This is a very important piece o a d e pressures—far less than those in a ammonia fertilizer sells for about $750 a of the jigsaw puzzle,” Cooper says. d Haber-Bosch reactor—that, compared to ton, hydrogen for fuel cell vehicles can go Beyond 2030, Japan will likely import f r o m MacFarlane’s cell, boost throughput while for more than 10 times that amount. between $10 billion and $20 billion of sacrificing efficiency. The reactor designs In the United States, fuel cell cars seem hydrogen each year, according to a re- call for a pair of concentric long metallic all but dead, vanquished newable energy road- tubes, heated to 450°C. Into the narrow gap by battery-powered ve- map recently published between the tubes flows H , which could be hicles. But Japan is still by Japan’s Ministry of 2 “It looks like there’s made by a solar- or wind-powered electro- backing fuel cells heav- Economy, Trade and lyzer. Catalysts lining the gap split the H2 ily. The country has spent enough interest to get Industry. Japan, Singa- molecules into individual hydrogen atoms, more than US$12 billion pore, and South Korea which modest pressures then force through on hydrogen technology this industry started.” have all begun discus- the atomic lattice of the inner tube wall as part of its strategy David Harris, CSIRO Energy sions with Australian of- to its hollow core, where piped-in N mol- to reduce fossil fuel im- ficials about setting up 2 ecules await. A catalytically active metal ports and meet its commitment to reduce ports for importing renewably produced o such as palladium lines the inner surface, CO emissions under the Paris climate ac- hydrogen or ammonia. “How it all comes n 2 J splitting the N and coaxing the hydrogen cord. Today the country has only about together economically, I don’t know,” Har- u 2 l y and nitrogen to combine into ammonia— 2500 fuel cell vehicles on the road. But by ris says. “But it looks like there’s enough 1 4 , much faster than in MacFarlane’s cell. So far 2030 Japanese officials expect 800,000. interest to get this industry started.” 2 only a small fraction of the input H reacts And the nation is eyeing ammon a way Cooper knows how he wants it to end. 0 2 1 in any given pass—another knock to the to fuel them. Over coffee on a rainy morning in Sydney, 8 reactor’s efficiency. Converting hydrogen into ammonia he describes his futuristic vision for renew- Other approaches are in the works. At only to convert it back again might seem able ammonia. When he s ts, he can see, the Colorado School of Mines in Golden, strange. But hydrogen is hard to ship: It maybe 30 years down the road, Australia’s researchers led by Ryan O’Hayre are devel- has to be liquefied by chilling it to temper- coast dotted with supertankers, docked oping button-size reverse fuel cells. Made atures below –253°C, using up a third of its at offshore rigs. But they wouldn’t be fill- from ceramics to withstand high operat- energy content. Ammonia, by contrast, liq- ing up with oil. Seafloor powerlines would ing temperatures, the cell can synthesize uefies at –10°C under a bit of pressure. The carry renewable electricity to the rigs from ammonia at record rates—about 500 times energy penalty of converting the hydrogen wind and solar farms on shore. On board, faster than MacFarlane’s fuel cell. Like Gid- to ammonia and back is roughly the same one device would use the electricity to de- dey’s membrane reactors, the ceramic fuel as chilling hydrogen, Dolan says—and salinate seawater and pass the fresh water cells sacrifice some efficiency for output. because far more infrastructure already to electrolyzers to produce hydrogen. An- Even so, O’Hayre says, they still need to im- exists for handling and transporting am- other device would filter nitrogen from the prove production rates by another factor of monia, he says, ammonia is the safer bet. sky. Reverse fuel cells would knit the two 70 to meet the DOE targets. “We have a lot That last step—stripping hydrogen off together into ammonia for loading on the of ideas,” O’Hayre says. ammonia molecules—is what Dolan and tankers—a bounty of energy from the sun, Whether any of those approaches will his colleagues are working on. In a cavern- air, and sea. wind up being both efficient and fast is still ous metal wa ouse on the CSIRO cam- It’s the dream that nuclear fusion never unknown. “The community is still trying to pus that has long been used to study coal reached, he says: inexhaustible carbon-free figure out what direction to go,” says Lauren combustion, two of Dolan’s colleagues are power, only this time from ammonia. “It Greenlee, a chemical engineer at the Uni- assembling a 2-meter-tall reactor that is can never run out, and there is no carbon versity of Arkansas in Fayetteville. Grigorii dwarfed by a nearby coal reactor. When in the system.” j SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 123 Published by AAAS DA_0713NewsFeatures.indd 123 7/11/18 10:27 AM INSIGHTS PERSPECTIVES D o w n l o a d e d f r o m o n J u l y 1 2 , 2 0 1 8 COGNITION costly to individuals and society. Comparing responses across different species is key to When persistence doesn’t pay understanding how these biases evolved and will help to both predict when they will man- ifest and offer suggestions as to how to offset Rats, mice, and humans all invest more time in a foraging them. However, determining how the sunk task than is in their interest cost effect evolved has not been easy, be- cause the bias is not consistently expressed in other species (7–9). By Sarah F. Brosnan1,2 viduals invest more time or resources in an Studying the sunk cost bias is challenging K C O outcome than it is worth given the potential because it is difficult to separate investment T S R E eople routinely make bad decisions. gain, presumably because they focus on hind- from future expectations, especially in field T T U Far from being random, however, sight evaluation of irrecoverable costs rather studies. Lab experiments can more easily dis- H S / Z these bad decisions are often predict- than prospective gains (5). On page 178 of tinguish between these factors, but lab exper- E R R E able, occurring reliably in specific con- this issue, Sweis et al. (6) report a clever ex- iment setups can differ substantially across I T U G texts (1–3). One set of such suboptimal periment demonstrating the sunk cost bias species, particularly between humans and O D Pdecisions are cognitive biases, wherein in rats, mice, and humans. The findings also other species, making comparisons difficult. L A R I G individuals make decisions that predictably provide insight into why results from previ- S 1Neuroscience Institute and Language Research Center, U violate rationality or their own best interests ous studies have been so variable. S E 2 J Georgia State University, Atlanta, GA 30302, USA. National : without a logical reason for doing so (4). One Cognitive biases can be substantial imped- O Center for Chimpanzee Care, MD Anderson Cancer Center, T O such bias is the sunk cost bias, in which indi- iments to optimal decision-making and are Bastrop, TX 78602, USA. Email: sbrosnan@ H P 124 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Perspectives.indd 124 7/11/18 10:32 AM When foraging, animals must decide whether to viding information about how much future EVOLUTION continue in the current patch or move on. Staying too investment remained to minimize informa- long is an example of the sunk cost bias, which tion uncertainty. Importantly, they provided Many roads Sweis et al. demonstrate in rats, mice, and humans. the humans and the rodents with the same procedures and information. Sweis et al. overcome the latter problem by Of course, there were differences among to convergence using the same basic procedure, based on a species. Differences in training between the foraging task, across all three species. rodent labs required control tests to verify Plant genomes highlight In their experiment, subjects had a lim- that behavior was truly the same. Humans, ited time budget during which they rotated unlike rodents, are not consistently food complex mechanisms behind through four reward options that differed in motivated, and so they “foraged” for enter- evolutionary convergence how well the subjects liked them (subjects tainment, in the form of movie clips, instead. showed consistent preferences across the Such adaptations to account for species-typ- study). At each transition, subjects first had ical behavior and differing environments are By László G. Nagy to decide whether to commit to the available nearly unavoidable in comparative work, but option or move to the next option. If they it is critical that the fundamental procedure any plants form specialized symbi- committed, they could not get the reward is the same and that the limitations of one’s otic root structures, called nodules, until they had waited for a predetermined study are acknowledged (10), as in Sweis et al. that harbor beneficial associations period of time; the time remaining was indi- For instance, the authors chose film clips be- with nitrogen-fixing bacteria in cated to the subject throughout the waiting cause, like the differently flavored food, they the genera Rhizobium or Frankia period. Subjects could leave at any point dur- differ in value but not handling time, which M(see the photo). How this nitrogen- D o ing the waiting period, without penalty, and allows for a direct comparison across the four fixing root nodule (NFN) symbiosis arose w n l move to the next option in the rotation. choice options. repeatedly during plant evolution is an age- o a d e A subject not showing the sunk cost bias Why are we susceptible to the sunk cost old mystery: It shows signatures of conver- d should leave quickly if a less preferred re- bias? In some situations, it may be difficult gence (the repeated emergence of similarity f r o m ward is paired with a long wait time. None- to evaluate future costs; in these cases, past during evolution) yet builds on similar gene theless, subjects of all species were more efforts can be a reasonable proxy. If this heu- sets in phylogenetically distant plants. On committed to sticking with their choices ristic worked well enough most of the time, page 144 of this issue, Griesmann et al. (1) than their preferences warranted, thus dem- there would not have been strong selection sequenced the genomes of 10 plant species onstrating a sunk cost bias. Indeed, this bias pressure to evolve a more sophisticated strat- to reveal the genetic correlates of the ori- grew stronger the longer they waited; that egy. Indeed, many human decision-making gin and loss of NFN symbiosis. Their work is, it increased with increasing investment. biases likely result from a trade-off between reveals intricate gain and loss patterns of However, it was unaffected by the time spent the costs of acquiring and processing infor- symbiosis-associated genes, calling for new making the choice initially. Supplemental mation and the costs of getting it wrong (11). models to explain convergent evolution. work indicates that there is a distinct deci- Understanding this evolutionary trade-off NFN symbiosis has an immense impact sion phase during which sunk costs do not can help us to better understand why hu- on plant growth and global nitrogen cy- accumulate, even when this phase is not ex- mans make the decisions that we do and, ul- cling but occurs only in four related plant o plicitly included as a separate period in the timately, suggest ways of improving human orders, the Fabales, Fagales, Cucurbitales, n J experimental design. The authors argue that decision-making (12). Sweis et al.’s adaptable and Rosales, collectively known as the NFN u l y this is evidence for two independent valua- design is likely to prove helpful in future clade. Only a few phylogenetically diverse 1 2 , tion processes, one assessing whether to ac- experiments investigating sunk cost biases members of the NFN clade form symbio- 2 cept an offer and the other—the one that is across a variety of species and contexts. j sis, making NFN symbiosis a phylogeneti- 0 1 susceptible to sunk costs—assessing whether cally patchy character (2, 3). Despite its 8 REFERENCES to continue investing in the choice. phylogenetic patchiness, nodule formation 1. G. Gigerenzer, in Bounded Rationality: The Adaptive The evidence for two different valuation Toolbox, G. Gigerenzer, R. Selten, Eds. (MIT Press, 1999), builds on surprisingly similar gene sets (4), systems working together offers the possibil- pp. 37–50. with some key symbiosis-associated genes, ity of a comp ensive explanation for the 2. D. Kahneman, P. Slovic, A. Tversky, Eds., Judgment Under such as NODULE INCEPTION (NIN) and Uncertainty: Heuristics and Biases (Cambridge Univ. sunk cost effect that integrates cognition, be- Press, ed. 1, 1982). RHIZOBIUM-DIRECTED POLAR GROWTH havior, and neural architecture across species 3. R. H. Thaler, The Winner’s Curse: Paradoxes and Anomalies (RPG), being widely conserved across and contexts. Moreover, these results hint at of Economic Life (Free Press, New York, 1991). plants. Besides NIN and RPG, another 4. M. G. Haselton, D. Nettle, P. W. Andrews, in The Handbook of some of the reasons for the previous dispa- 290 genes were recently found to be up- Evolutionary Psychology (Wiley, 2005), pp. 724–746. rate results. Perhaps wait times were not long 5. R. Thaler, J. Econ. Behav. Organ. 1, 39 (1980). regulated in nodules of Medicago truncatula enough to get past the assessment period, or, 6. B. M. Sweis et al.,Science 361, 178 (2018). and Parasponia andersonii, two plant spe- more speculatively, decisions may change if 7. H. R. Arkes, P. Ayton, Psychol. Bull. 125, 591 (1999). cies that independently evolved the ability 8. F. Hoeffler, Why humans care about sunk costs while subjects do not know how long they must animals don’t. An evolutionary explanation , paper pre- to form symbiosis. This is surprising, given wait or if they are otherwise engaged in an sented at the Royal Economic Society Annual Conference, that these species diverged ~100 million active task (such as lever pressing) during the Nottingham, UK, 20 April 2006. years ago and belong to two separate or- 9. P. Magalhães, K. Geoffrey White, J. Exp. Anal. Behav. 105, waiting period. 339 (2016). ders, the Fabales and Rosales, respectively A key reason for Sweis et al.’s success is the 10. M. Smith, J. Watzek, S. F. Brosnan, Int. J. Comp. Psychol. 31, (4). How can this level of genetic similarity degree to which they worked to provide a rel- uclapsych_ijcp_37777 (2018). be reconciled with the phylogenetic patchi- 11. J. Watzek, S. F. Brosnan, Cognition 178, 109 (2018). evant, comparable methodology. They chose 12. R. H. Thaler, C. R. Sunstein, Nudge: Improving Decisions a context, foraging, that had ecological rel- About Health, Wealth, and Happiness (Yale Univ. Press, evance to these species in the context of sunk ed. 1, 2008). Synthetic and Systems Biology Unit, Biological Research Center, Hungarian Academy of Sciences, 6726 Szeged, costs and came up with a clever way of pro- 10.1126/science.aau3144 Hungary. Email: lnagy@ SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 125 Published by AAAS DA_0713Perspectives.indd 125 7/11/18 10:32 AM INSIGHTS | PERSPECTIVES ness of nodulation? Griesmann et al. out- a more complex scenario involving homol- bilaterian ancestor (9). In this case, both lined three possible evolutionary scenarios ogy that dates back to early, nonnodulating function and genetics a omologous, but to describe the extant distribution of NFN precursors. evidence for this is blurred by divergence symbiosis across the plant tree of life: It Convergently evolved traits involving a and/or lineage-specific refinements, overall could be described by up to 16 independent similar genetic background a ard to ex- leading to apparent convergence. evolutionary origins, a single origin with plain. Although the multiple-origins model Conversely, latent homologies are de- multiple losses, or an intermediate scenario is often the phylogenetically most parsi- velopmental potentials (for example, gene combining homologous and convergently monious for such traits, the likelihood of regulatory circuits) that can be deployed evolved components of the underlying ge- parallel changes affecting tens to hundreds (through co-option or exaptation) for new netics. The intermediate scenario invokes a of the same genes is so low that multiple in- functions by simple genetic changes. Be- predisposition event, the gain of a genetic dependent origins seem implausible. More cause a few mutations (for example, in element (gene, regulatory interaction, or complex models, involving deep homolo- regulatory sequences) are sufficient for other), that made descendant lineages more gies (7), latent homologies (8), predisposi- their expression in a new context, latent likely to evolve NFN symbiosis (5). tion events, or parallel co-option of genes homologies reduce the mutational target Griesmann et al. sequenced the genomes have been proposed to explain the evolu- size for evolution and can thus promote of 10 carefully selected plant spe- convergence. Mechanistically, pre- cies to cover several clades that disposition events can comprise the independently gained the ability to evolution of latent homologies that, form NFN symbiosis. The authors as hypothesized for the NFN clade embarked on a sizable challenge (5), lead to a higher likelihood for to identify genomic evidence for convergence to occur. D o a postulated predisposition event, Teasing apart the possible mecha- w n l either for independent gains or for nisms behind convergently evolved o a d e independent losses. Searching mul- traits remains a substantial chal- d tiple genomes for genetic changes lenge even in the era of genomics. f r o m underlying the evolution of a trait is It nevertheless appears that case like finding a needle in a haystack, studies and models are emerging although incorporating the gain to explain the pervasive occur- and loss patterns of the trait across rence of convergence across the the phylogeny can help narrow the tree of life. NFN symbiosis rep- search space down (6). Indeed, the resents one particularly intricate authors found several genes, such case of phylogenetically patchy as NIN and RPG, that were lost in traits. The phylogenomic compari- clades that lost the ability to form sons by Griesmann et al. provide NFN symbiosis. Such losses have im- important clues about the genetics portant implications for genetic en- of the repeated emergence of NFN gineering of NFN symbiotic plants. symbiosis and the genes that are o In nature, the symbiosis can be lost repeatedly lost in secondarily non- n J in nitrogen-rich habitats, where the symbiotic clades, although some u l y cost of forming symbiosis outweighs ambiguity remains as to the con- 1 2 , its benefits, making symbiosis-asso- vergent origins. With the increas- 2 ciated genes dispensable. In agricul- ing accessibility of whole-genome 0 1 tural settings, however, long-term sequences, such comparisons will 8 fertilizer application might lead to A colored scanning electron micrograph of a bean plant root nodule likely play an increasing role in similar losses of the ability to nodu- shows nitrogen-fixing Rhizobium bacteria (green) embedded in plant dissecting the genetic background late, which could be exploited with tissue, forming a symbiotic relationship (magnification ¤50). of complex multigenic traits, such the development of engineered NFN as nodulation, and decoding long- symbiotic crop plants in regions where fer- tion of phylogenetically patchy traits, like standing evolutionary questions, including tilizer availability is limited. NFN symbiosis. Although these have been the many roads that can lead to evolved Another 52 gene families showed parallel used to denote various, sometimes overlap- similarity in disparate branches of the tree expansions in multiple symbiotic subclades ping processes, they refer to at least two of life. j of the NFN clade and contained known distinct evolutionary mechanisms. Traits REFERENCES nodulation-associated genes of Medicago, are deeply homologous if homology is not 1. M. Griesmann et al.,Science 361, eaat1743 (2018). E although most gene family expansions were evident at the level of the entire trait but C 2. M. J. Telford, G. E. Budd, H. Philippe, Curr. Biol. 25, R876 R U specific to individual subclades. These ob- historical continuity of shared function ex- (2015). O S E servations would make NFN symbiosis a ists for at least some of the genes involved 3. L. G. Nagy, G. M. Kovács, K. Krizsán, Biol. Rev. 10.1111/ C N E brv.12418 (2018). I classic example of convergent evolution: in its development. Arthropod and verte- C 4. R. van Velzen et al.,Proc. Natl. Acad. Sci. U.S.A. 115, E4700 S / R independent acquisitions of similar func- brate eyes are textbook examples of deep (2018). E N S 5. D. E. Soltis et al.,Proc. Natl. Acad. Sci. U.S.A. 92, 2647 S tions via different evolutionary routes, with homologies: They develop by largely nonho- I E (1995). M dissimilar genetic backgrounds. However, mologous mechanisms but build on homol- H 6. L. G. Nagy et al.,Mol. Biol. Evol. 34, 35 (2017). C S G many symbiosis-associated genes predate ogous genes, such as opsins, developmental 7. N. Shubin, C. Tabin, S. Carroll, Nature 457, 818 (2009). E V 8. L. G. Nagy et al.,Nat. Commun.5, 4471 (2014). E the origin of the NFN clade and are uni- transcription factors, and photoreceptor T S 9. N. Shubin, C. Tabin, S. Carroll, Nature 388, 639 (1997). : versally up-regulated during nodule forma- cells that can be traced back to primor- O T O tion in independent clades (4), suggesting dial photosensory organs of a cnidarian or H 10.1126/science.aau2409 P 126 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Perspectives.indd 126 7/11/18 10:32 AM CHEMISTRY Femtosecond structural photobiology Time-resolved crystallography reveals how bacteriorhodopsin uses light to drive chemistry By Keith Moffat molecular dynamics calculations; and Synchrotron x-ray sources at which such structural probes with low spatial and experiments were pioneered are restricted bsorption of a photon in the visible temporal resolution. Spectroscopic data to a time resolution of ~100 ps (6). Nogly region of the solar spectrum is the provide a simple linear model for the reac- et al. now access isomerization on the fem- fundamental step in the photochem- tion path. Explicit structural information tosecond to picosecond time scale by us- ical reactions that drive biological on the femtosecond time scale is needed to ing the extremely intense, femtosecond, processes such as imaging, photo- develop a more detailed picture. hard x-ray pulses from the Linac Coherent Asynthesis, and sensory perception. Nogly et al. now provide these data in Light Source at Stanford, the world’s first All known photoreceptors are proteins that the face of major experimental and ana- hard x-ray free–electron laser (XFEL) (7, 8). contain a chromophore: a small organic lytical difficulties. The structural changes However, XFELs are based on noise ampli- molecule that absorbs in the visible region that occur during isomerization are small; fication and generate x-ray scattering data of the spectrum (1). The bacterial integral few atoms move, and those that do move that are prone to systematic errors. Each membrane protein bacteriorhodopsin (bR) do not move very far. The time-dependent tiny bR crystal is destroyed by the extremely D o uses retinal as its chromophore (see the changes in x-ray scattering that constitute intense XFEL pulse immediately after gen- w n l figure), capturing the energy in a photon the raw data are therefore also small, yet erating its diffraction pattern. o a d e to drive retinal isomerization Nogly et al. present differ- d and pump protons across the ence electron density maps f r o m membrane. bR is a readily ac- Tracking femtosecond retinal changes that minimize systematic error cessible analog to the visual Difference electron density maps after the pump laser pulse show the retinal by directly comparing the x-ray rhodopsins in, for example, structure in bacteriorhodospin (bR) before to isomerization (top) and during scattering from the light struc- the human eye. On page 145 isomerization (bottom). ture (at a known, variable time of this issue, Nogly et al. (2) delay after the pump pulse) used time-resolved x-ray crys- Time bin one (49 to 40 6fs) Negative and dark structure (before the tallography (3, 4) to determine Small structural changes in the diference C20 pump pulse). Such difference intermediate structures of bR initial electronic excited density Schif base maps a ighly sensitive to at near-atomic resolution on state occur after a photon is small structural differences abosorbed but before bR reJned the femtosecond to picosecond isomerization has occurred. C13 C14 intermediates (9). To obtain atomic struc- time scales at which isomeriza- bR dark state tures, the authors converted tion occurs. the difference maps into time- o The photochemistry of reti- dependent, extrapolated den- n J nal in bR differs substantially sity maps and further averaged u Time bin two (457 to 646 fs) l C20 y from that of free retinal. In- In the J state, isomerization these maps into time bins cen- 1 2 , teractions between the retinal is half complete, with tered on the anticipated time 2 Schif base Positive and its protein surroundings a torsion angle of 90° about of um concentration of 0 1 diference tightly constrain the chromo- the C13=C14 bond. C13 density the I, J, and K spectroscopic 8 phore and confer specificity C14 states. They then refined a in trans-to-cis isomerization single atomic structure in each about the C13=C14 bond, high extrapolated density map. quantum yield, and efficient Two effects may complicate proton pumping. These constraints must their accurate measurement is essential. the accuracy of refinement. First, in any be very selective. Isomerization requires The authors’ visible laser pump–hard x-ray dynamic experiment, an individual time carefully choreographed, ultrafast atomic probe experiments are analogous to mak- delay—here, a time bin—typically contains motions that have evolved to direct the re- ing a movie one frame at a time. A light a mixture of structures (3, 4). Second, ab- action along the desired path and inhibit pulse from a visible laser—the pump—ini- sorption of a photon by retinal generates all undesired reactions, such as isomeriza- tiates the reaction in the bR molecules in a a vibrationally excited molecule far from tion about a different bond or rapid rever- tiny crystal. After a controlled time delay, equilibrium, with unusual bond lengths sion to the ground state. the crystal is illuminated by an extremely and angles. Reassuringly, the single refined How is this molecular choreography intense pulse from an x-ray laser—the structures are chemically plausible and achieved? Extensive earlier studies (5) probe. By obtaining probe data over all consistent with both calculations (2) and E have explored this question by using fem- crystal orientations, the authors determine the previously proposed aborted bicycle C N E I tosecond optical spectroscopies; quantum the structure of the bR molecules at that pedal model for isomerization (10, 11). C S / L mechanical, molecular mechanics, and time delay. By varying the time delay, the The difference maps are of interest both E K C I “light” structures are determined frame for the atomic motions they reveal and for B . C : by frame—the reaction path. If the visible those they do not reveal. Certain protein C I Department of Biochemistry and Molecular Biology and H laser pulse is omitted, no reaction occurs, atoms comprising the retinal pocket in the P Institute for Biophysical Dynamics, University of Chicago, A R Chicago, IL 60615, USA. Email: moffat@ and the “dark” structure is obtained. dark do not move; they impose constraints G SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 127 Published by AAAS DA_0713Perspectives.indd 127 7/11/18 10:32 AM INSIGHTS | PERSPECTIVES on retinal motion throughout isomerization (see the figure). Other atoms and groups move in concert, restricting isomerization to the C13=C14 bond. A counter-ion cluster consisting of Asp85, Asp212, and a key buried water molecule (Wat402) stabilizes the posi- tive charge on the Schiff base in the dark state. Even before isomerization in the electronically excited state, this cluster is disrupted, and specific hydrogen bonds are weakened. The positive charge on the Schiff base is redistributed along the polyene chain. The alternation of single and double bonds along the polyene chain is perturbed and at least partial single-bond character EMBRYOGENESIS conferred on the C13 to C14 bond (a double bond in the dark state) to facilitate its twist- Double trouble at the ing as isomerization begins. In bR, isomerization appears to be half complete (with a torsion angle of 90° about the C13=C14 bond) at ~500 fs. Isom- beginning of life D o erization has been studied with femtosec- w n Dual-spindle assembly in early embryos can l ond time-resolved crystallography in two o a d e other proteins: photoactive yellow protein, compromise mammalian development d whose small chromophore contains only a f r o m single double bond about which isomeriza- tion is also half complete at ~500 fs (12), By Agata P. Zielinska and Melina Schuh Reichmann et al. labeled the maternal and a photoswitchable fluorescent protein, and paternal chromosomes in different in which a larger chromophore must ex- very human life begins with the fer- colors by taking advantage of distinct DNA ecute molecular contortions to accomplish tilization of an egg (1). Once the egg sequences (5 ) in the parental chromosomes, isomerization in ~1 ps (13). It is likely that and the sperm have fused, the paren- which came from different mouse strains electronic and structural similarities ex- tal chromosomes need to be united. (6). To follow in detail how the chromo- ist in the earliest steps of the elementary To this end, the egg and sperm chro- somes are united, zygotes have to be imaged chemical process of isomerization in differ- Emosomes are first packaged into two at very high spatial and temporal resolu- ent chromophores. separate membrane-enclosed nuclei. These tion. However, embryos are light sensitive, Challenging questions remain. Is the na- nuclei then slowly move toward each other which has hindered a detailed analysis in ture of the cis-to-trans isomerization path- and break down in the center of the fertil- the past. The authors overcame this limi- o way in the physiologically important visual ized egg, called the zygote. Only then are tation by using an innovative light-sheet n J rhodopsins simply the reverse of the trans- the maternal and paternal chromosomes microscope, which illuminates the embryo u l y to-cis pathway seen here in bR? Are the united—but not quite. Surprisingly, the pa- selectively in the region of interest but not 1 2 , structures and charge distributions of the rental chromosomes do not mix immediately in adjacent regions, as is the case with stan- 2 retinal pockets in the trans and cis species but instead occupy distinct territories in dard microscopy approaches (7). This re- 0 1 in bR and visual rhodopsins closely similar? the zygote throughout the first cellular divi- duces the amount of light that the embryo 8 Nogly et al. show that the techniques are sion (2, 3). How the autonomy of parental is exposed to. Moreover, this method is fast firmly in place to address these questions. j genomes is retained after fertilization has and allowed the authors to reconstruct the remained unclear. On page 189 of this issue, entire volume that is occupied by the chro- REFERENCES Reichmann et al. (4) used elegant microscopy mosomes with unprecedented spatial and 1. A. Möglich, X. Yang, R. A. Ayers, K. Moffat, Annu. Rev. Plant methods to illuminate this special moment, temporal resolution. Biol. 61, 21 (2010). 2. P. Nogly et al., Science 361, eaat0094 (2018). when the parental chromosomes first meet in The authors tracked the maternal and 3. K. Moffat, Chem. Rev. 101, 1569 (2001). live mouse zygotes, and follow how the chro- paternal chromosomes as they first met. In 4. K. Moffat, in X-ray Free Electron Lasers , U. Bergmann, V. mosomes become distributed as the zygote addition, they imaged microtubules, the pro- K. Yachandra, J. Yano, Eds. (Royal Society of Chemistry, divides. Their findings reveal an unexpected teinaceous fibers that form the spindle ap- 2017), vol. 18, p. 107. 5. C. Wickstrand, R. Dods, A. Royant, R. Neutze, Biochim. mechanism that keeps the parental genomes paratus that captures, aligns, and distributes Biophys. Acta 1850, 536 (2015). apart during the first division of the embryo: the chromosomes equally between the two 6. V. Srajer et al., Science 274, 1726 (1996). The male and female chromosomes each as- daughter cells of the dividing zygote. Surpris- 7. P. Emma et al.,Nat. Photonics 4, 641 (2010). semble their own chromosome separation ingly, Reichmann et al. found that the mater- 8. J. Spence, IUCr J. 4, 322 (2017). 9. R. Henderson, K. Moffat, Acta Cryst. B27, 1414 (1971). machineries. This increases the probability nal and paternal chromosome masses each ) 4 ( 10. A. Warshel, Z. T. Chu, J. Phys. Chem. B . 105, 9857 (2001). that chromosomes are separated into multi- assemble a separate spindle structure that . L A 11. P. Altoè, A. Cembran, M. Olivucci, M. Garavelli, Proc. Natl. ple, unequal groups, which may compromise autonomously initiates chromosome align- T E Acad. Sci. U.S.A. 107, 20172 (2010). N embryo development and give rise to sponta- ment (see the images). Later, the two spindles N 12. K. Pande et al.,Science 352, 725 (2016). A M neous miscarriage. merge into a single spindle. However, the ma- H 13. N. Coquelle et al.,Nat. Chem. 10, 31 (2018). C I E ternal and paternal chromosomes remain in R : S separate regions of the merged spindle and O Max Planck Institute for Biophysical Chemistry, Göttingen T O 10.1126/science.aau3200 37077, Germany. Email: melina.schuh@mpibpc.mpg.de do not mix. H P 128 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Perspectives.indd 128 7/11/18 10:32 AM Immunofluorescence staining of a mouse zygote permissive environment for the formation of beginning of life is not as clear cut as was pre- (left image) shows parallel alignment of two mitotic two independent spindle structures. viously assumed. How we define the “begin- spindles (green). This is overlaid (right image) The presence of centrosomes in human zy- ning of life” not only has ethical implications with microtubule organizing centers (magenta), sites gotes could facilitate the assembly of a single but also practical repercussions for fertility of microtubule binding to chromosomes (gray), bipolar spindle. However, it is still unclear treatments. This is because in some countries, and chromosomes (blue). whether spindle assembly in human zygotes such as Germany (14), parental nuclei are al- is primarily centrosome driven, or whether lowed to merge and hence life is allowed to Whether the spatial separation of paren- they also assemble two distinct spindles from “begin” in IVF clinics only in a very limited tal chromosomes has any advantages for the chromosome surfaces. Interestingly, spin- number of in vitro–cultured zygotes, all of developing mammalian embryo is unclear. dle assembly in unfertilized human eggs is which have to be transferred to the mother. However, the fact that zygotes have a dual largely chromosome driven (13), which raises This policy increases the rate of multiparous spindle creates a previously unforeseen the possibility that also in human zygotes, pregnancies, which are associated with se- source of potential error. The final task the even in the presence of centrosomes, chro- vere risks to the health of the mother and her zygote has to accomplish before it divides mosome-driven spindle assembly may occur. children. Additionally, because identification is to align the two spindle axes in parallel In many countries, the law considers that of the most promising fertilized eggs has to to each other, so that the two spindles can new human life begins when parental chro- legally occur before the parental genomes merge into a compact dual structure. If the mosomes are united upon fertilization (14). have merged (so that embryos with merged poles of the spindles fail to align and merge, This work is a reminder that the definition genomes, defined as being human life, are the genetic material of the zygote could be of the unification of parental genomes as the not discarded), embryologists have to select pulled into three or four direc- embryos for implantation at the tions instead of two (see the zygotic stage. Because our under- D o figure). Reichmann et al. dem- standing of early human embryo- w Together, but still apart n l onstrate that spindle misalign- genesis is still poor, it is difficult o Upon fertilization, parental chromosomes need to be united. Unexpectedly, in a d e ment leads to the formation of mice, male and female chromosomes are kept apart on separate mitotic spindles to select at this early stage the d multinucleated embryos that embryos that have the high- during the first division of the zygote. This increases the probability of forming f r o m have more than one nucleus per multinucleated cells. est chance of dividing correctly, cell. It is easy to envision that implanting, and giving rise to a First mitosis in mammals such an undesired partition- healthy pregnancy. Considering ing of the DNA would have a Polar body that there is now comp ensive negative impact on the fidel- Fertilization evidence that the parental DNA ity of subsequent cell divisions Maternal Sperm does not mix in live zygotes, this and hence might compromise chromosomes study highlights that a dialogue embryo development. Indeed, is needed to reshape the legal experience from in vitro fertil- Paternal chromosomes definition of when life begins, so ization (IVF) clinics shows that Formation that it is supported by up-to-date of two Old model Current model early human embryos cultured scientific knowledge. By revisit- nuclei in vitro before implantation fre- ing this definition, we could in- o quently have multiple nuclei in crease the chances of a healthy n Zygote Nuclear J their cells (8) and then fail to pregnancy among couples who u l envelope y develop further (9). otherwise struggle to conceive. j 1 2 breakdown , To evaluate the implications of REFERENCES 2 these findings for human infertil- 0 1 1. D. Clift, M. Schuh, Nat. Rev. Mol. Cell Biol. ity, it is important to investigate Mitotic spindle 14, 549 (2013). 8 Early stages 2. W. Mayer et al.,J. Cell Biol .148, 629 whether cells with multiple nu- of spindle (2000). clei in human embryos are also assembly 3. C. van de Werken et al.,Nat. Commun.5, caused by defects during dual- 5868 (2014). Correct axis Incorrect axis 4. J. Reichmann et al.,Science 361, 189 spindle merging, as reported for (2018). mouse embryos in this study. In alignment alignment 5. S. Henikoff, K. Ahmad, H. S. Malik, Science 293, 1098 (2001). humans, the sperm delivers a Mature 6. Y. Miyanari, C. Ziegler-Birling, M. E. Torres- centrosome into the zygote (10, frst Padilla,Nat. Struct. Mol. Biol.20, 1321 11). Centrosomes are microtu- mitotic (2013). spindle 7. P. Strnad et al.,Nat. Methods 13, 139 bule-organizing centers that help (2016). to generate a bipolar spindle in 8. H. Balakier, K. Cadesky, Hum. Reprod.12, 800 (1997). somatic cells (12). However, in 9. A. Egashira et al.,J. Reprod. Dev .61, 595 mice, the sperm does not carry First (2015). a centrosome, and the zygotic mitotic 10. G. Manandhar, C. Simerly, G. Schatten, Hum. Reprod.15, 256 (2000). spindle needs to assemble with- anaphase 11. E. L. Fishman et al.,Nat. Commun. 9, 2210 E out the help of centrosomes (1). (2018). C 12. M. Bettencourt-Dias, D. M. Glove, Nat. Rev. N Reichmann et al. demonstrate E One nucleus Multinucleated I Mol. Cell Biol.8, 451 (2007). C S 13. Z. Holubcova et al.,Science 348, 1143 / that chromosomes play a key per cell cells I A (2015). S E role in driving spindle assembly D. 14. M. Kreyenfeld, D.Konietzka, Childlessness N : in the mouse zygote. The chro- Two-cell in Europe: Contexts, Causes, and C I H mosome-based spindle assembly embryo Consequences (Springer, 2017). P A R mechanism seems to create a 10.1126/science.aau3216 G SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 129 Published by AAAS DA_0713Perspectives.indd 129 7/11/18 10:32 AM INSIGHTS | PERSPECTIVES INFECTIOUS DISEASES genomics is a mirror image of conservation genomics. Although the latter aims to under- Eradication genomics— stand populations to preserve biodiversity and ultimately reverse species declines, eradi- cation genomics studies the genetics of popu- lessons for parasite control lations that are deliberately being reduced by human activity, to enhance our ability to de- Genomic surveillance could help achieve targets press the populations further and ultimately remove them completely. For some time, for the elimination of tropical diseases population genetic data have been used in conservation biology to understand changes in endangered populations (6), but we expect By James A. Cotton, Matthew Berriman, tional resources to areas of high or stubborn that analysis of eukaryotic pathogen popula- Love Dalén, Ian Barnes prevalence. However, monitoring pathogen tions could result in some important, more populations by conventional techniques is subtle lessons from conservation genomics. arge-scale programs are seeking to con- challenging in many of the environments For example, that isolated subpopulations are trol or eliminate infectious diseases where these diseases flourish, particularly harder to conserve implies that eradication with the greatest impact on global as only limited diagnostics are available will be easier if pathogen populations can be health. Many of these efforts target (2). Monitoring will likely become particu- isolated from one another. In small popula- the neglected tropical diseases (NTDs) larly difficult as prevalence is reduced (3). tions, stochastic changes in allele frequencies Lthat disproportionately affect the lives The power of molecular data to understand due to rando th or survival of individu- D o of the poor. Often the aim is to eradicate the pathogen populations is already widely ap- als can swamp the natural selection of benefi- w n l causative pathogens. The idea—inspired by preciated: Genetics can help confirm sus- cial genetic variation; the increased impact of o a d e the success of smallpox eradication in the pected cases, discriminate between strains, this genetic drift can make it much harder for d 1960s and 1970s—is that a large-scale, but identify new infectious agents, understand small populations to adapt to changing en- f r o m time-limited, effort could eliminate a disease the evolutionary history of populations, and vironments. Furthermore, inbreeding within for all successive generations, resulting in shed light on transmission. For example, in small isolated populations is expected to be an enormous payoff both financially and in the polio eradication campaign, a global net- deleterious (7). Looking for genomic signa- improved health. Here, we discuss the value work sequenced part of every virus isolated tures such as reduced genetic diversity and of genomic approaches to support disease from around 200,000 stool specimens per longer sequences of homozygosity due to in- eradication efforts, particularly by analogy year to identify transmission links and track breeding may therefore identify populations with how conservation genomics is support- the disappearance of particular genotypes close to elimination (8) (see the figure). ing efforts to prevent extinctions. (4). Genetic data identified polio outbreaks One difficulty is that levels of genome-wide Although there are fungal, viral, and bac- caused by oral vaccine–derived revertant diversity in species can be strongly affected by terial NTDs, most are caused by infections life history and ancient demographic events, with parasites (particularly protozoa and as well as recent changes in population size. helminths), including the protozoa that We may thus need a good picture of pathogen o n “Genomic surveillance cause sleeping sickness, leishmaniasis, and populations before control measures have an J Chagas disease, or worms that cause diseases impact. Therefore, collecting and conserving u could thus be a powerful l y such as schistosomiasis, hookworm disease, samples for genetic analysis should be an 1 5 and river blindness. Only two diseases—hu- tool for tracking… urgent consideration for control programs. , 2 man smallpox and the cattle disease rin- Guinea worm is the parasite closest to eradi- 0 parasite populations.” 1 derpest—have been successfully eradicated; cation; cases have declined from several mil- 8 both of these campaigns relied on the avail- lion per year in the 1980s to 30 in 2017 (9). ability of highly effective and long-lasting strains, highlighting a need to gradually However, Guinea worm infections were re- vaccines that are not available for any of the withdraw serotypes from the vaccine (5). ported in Chad in 2011 after a decade with parasitic NTDs. Control methods vary be- Whereas the most comp ensive genetic no reported cases. Coincident with this re- tween the different diseases and can include data are for prokaryotes and viruses, most emergence were infections in dogs, although controlling parasite carriers (vectors), better NTDs are caused by eukaryotic parasites that Guinea worm was not previously considered sanitation, and improvements to health sys- at least occasionally reproduce sexually. Most zoonotic (10). We assume that Guinea worm tems to allow faster detection and treatment. viruses and bacteria reproduce clonally, and evolved during the 10-year gap, but the lack However, the main approach has been large- exchange genetic material unpredictably, so of suitable samples predating 2011 makes scale treatment of people at risk of infection, molecular epidemiology can be restricted to it impossible to exclude the possibility that whether they actually have the disease or not. tracking the abundance of different lineages worms migrated from elsewhere during this The size of these interventions is astounding: through space and time. Recombination in time. Understanding the dog infections is key More than 1 billion people were treated for sexual eukaryotes makes each region of the to the eradication program: Although there NTDs in 2016, with drugs donated by manu- genome an independent genetic marker, so were only 25 human cases reported world- facturers. Nonetheless, this represents only genomic data for a small number of individ- wide in 2016, more than 1000 dog infections 63% of the treatment needed to fully cover uals can reveal a great deal about the popula- were identified in Chad that year. The need to the populations at highest risk (1). tion those individuals come from. Genomic retrospectively obtain samples is clear. There Key to the success of NTD control is mea- surveillance could thus be a powerful tool are no historical Guinea worm samples stored suring changes in pathogen populations, to for tracking changes in the demography or ensure that interventions are on course to abundance of parasite populations. Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, achieve elimination goals and to direct addi- In many ways, the study of eradication Cambridge CB10 1SA, UK. Email: james.cotton@sanger.ac.uk 130 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Perspectives.indd 130 7/11/18 10:32 AM in a way that is likely to keep DNA intact, but ential distribution of microsatellites—repeat the genome, or signs of a strong bottleneck there may be ways to circumvent the need for sequences in DNA—and other molecular in the absence of reduced prevalence are well-p DNA. Recent developments in markers. However, reductions in cost make indications that a population is evolving in paleogenomic methods, including improve- whole-genome sequencing the most appro- response to drug treatment or other con- ments in the efficiency of building genomic priate tool for the future, as genome-wide trol measures. This approach is being used libraries and the use of enzymatic treatments data are more powerful, and more compa- to track artemisinin resistance in human to reduce the impact of DNA damage, allow rable between populations and pathogens, malaria parasites in Asia, where population access to the highly fragmented and chemi- and sequencing avoids the need to develop genomics approaches have played a key role cally modified genomic information recov- marker panels. At least for most key NTD in identifying the major locus responsible for ered from archival specimens, and even from pathogens, basic genomic resources such as reduced efficacy of the drug and understand- environmental sources such as archaeologi- annotated reference genomes are already ing the genetic architecture of this trait (14). cal specimens or sediments (11, 12). A precise available. Genome-wide data also enable the Clearly, the highest priorities for control picture of the historical population structure impact of control measures on pathogen pop- programs focus on delivering effective con- and size of these parasites will allow us to ulations to be understood beyond changes in trol measures and maintaining funding and parameterize models of inbreeding and gene distribution or numbers, as genomic data are political support. However, the rewards for flow, improving the accuracy of inferences sensitive to changes at functional loci and collecting genomic data, and establishing about the effect of control measures on cur- could reveal evolutionary changes of epide- pretreatment baselines through the analysis rent populations (13). miological relevance, such as the emergence of archival samples, will be increasingly evi- For many species, researchers have started of drug resistance. The emergence or sudden dent as programs mature. There will be many exploring the genetics of populations using spread of new pathogen genotypes, signa- challenges in establishing genomic surveil- DNA “fingerprints” obtained from the differ- tures of strong recent selection on parts of lance programs: the need to be comp en- D o sive and coordinated, obtaining access to w n l both prospective and archival samples, un- o a d e Loss of diversity in a declining population derstanding the epidemiology and biology of d the pathogens, to name a few. In some cases, As control efforts reduce the size of a population, the chance that an individual is descended from closely f r o m related parents increases, leading to a decline in genetic diversity and heterozygosity. theory will need to be developed to explain the population genetics of NTD pathogens, Declining population many of which have genetic systems distinct 0.5 from those of obligately sexual, diploid ani- 0.0 Heterozygosity mals. Translation of biomedical science to 0.5 clinical practice has necessitated “bench to Genetic diversity bedside” collaborations, and similarly, eradi- 0.0 cation genomics will require collaboration between technical experts in molecular bi- s ology and paleogenomics, as well as groups l a u with expertise in the biology, control, and d i v i epidemiology of particular NTDs. The speed d o n i f and scope of changes in NTD pathogen pop- n o J r e ulations happening today is new territory u b l y m u for science. The ambitious goals set for NTD 1 N 5 , control will require pathogen populations to 2 Allele 1 Allele 1 inheritance shrink at an unprecedented rate. However, 0 1 Allele 2 Allele 2 inheritance success is not guaranteed (15) and will re- 8 Generation quire constant vigilance. j Constant population REFERENCES 0.7 1. Reaching a Billion: Fifth Progress Report on the London Declaration on NTDs (Uniting to Combat Neglected 0.0 Tropical Diseases, 2017). 0.5 2. P. J. Hotez et al.,PLOS Negl. Trop. Dis.10, e0003895 (2016). 3. P. Klepac et al.,Philos. Trans. R. Soc. London B Biol. Sci. 0.0 368,(2013). 4. O. M. Diop et al.,MMWR Morb. Mortal. Wkly. Rep. 66, 538 (2017). 5. N. C. Grassly, Philos. Trans. R. Soc. London B Biol. Sci. 368, s (2013). l a 6. F. W. Allendorf et al.,Nat. Rev. Genet. 11, 697 (2010). u E d C i 7. B. Charlesworth et al., Genet. Res. 74, 329 (1999). N v i E d 8. F. Abascal et al., Genome Biol. 17, 251 (2016). I C n S i 9. Guinea worm wrap-up #253 (Centers for Disease Control / f N o A r and Prevention, 2018); M e pdfs/news/health_publications/guinea_worm/wrap- R b E T m up/253.pdf. T I u K. N 10. M. L. Eberhard et al.,Am. J. Trop. Med. Hyg. 90, 61 (2014). A Y 11. M.-T. Gansauge et al.,Nat. Protoc.8, 737 (2013). B D 12. N. Rohland et al.,Philos. Trans. R. Soc. London B Biol. Sci. E T 370,(2015). P A 13. D. Díez-del-Molino et al. Trends Ecol. Evol. 33, 176 (2018). D A : Generation 14. O. Miotto et al.,Nat. Genet.47, 226 (2015). C I H 15. C. J. M. Whitty, Clin. Med. 14, 419 (2014). P Diagrams show inheritance of two alleles in simulated populations of 12 diploid individuals that declines or stays A R 10.1126/science.aar6609 G constant over eight generations. SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 131 Published by AAAS DA_0713Perspectives.indd 131 7/11/18 10:32 AM INSIGHTS | PERSPECTIVES FERROELECTRICITY Perovskite ferroelectrics go metal free Metal-free perovskites exhibit ferroelectric properties rivaling those of BaTiO3 By Wei Li1,2 and Li-Jun Ji2 Many hybrid organic–inorganic perovskites should enable wide potential applications in have been synthesized by replacing the single environments where conventional oxide ma- erroelectric materials, which have spon- atomic X site with bitopic organic linkers terials dominate. In addition, their multiple taneous electric polarization that can be such as formate and dicynamide groups (8). polarization directions enable facile rotation switched with an external electric field, Can this strategy extend to the B site? The through electrical-field manipulation, which have applications as capacitors, sen- perovskite framework is constructed via co- should make the poling process easier (10). sors, and data-storage devices (1). Since ordination bonding in oxides and hybrid The substantial organic content of the Fthe discovery of high-performance systems. Formation of coordination bonds metal-free perovskites makes these com- BaTiO and LiNbO , ferroelectric perovskite requires that a ligand donate lone-pair elec- pounds soft, so they respond to stress dif- 3 3 oxides have dominated industrial applica- trons to the orbitals of transition metals. ferently compared with oxides and hybrid tions (2). However, these ceramics are expen- Likewise, for forming hydrogen (H) bonds, perovskites, in which stronger ionic and co- sive to produce as high-quality thin films, and electronegative atoms with lone-pair elec- ordination bonding forces exist. This facili- D they often contain heavy metals such as Pb2+, trons are needed as acceptors. In this sense, tates facile growth of high-quality thin films o which raises health and environmental con- these two different bonding interactions that are critical for managing electric dipoles w n l cerns. Replacing heavy metals with organic exhibit some geometrical similarity. As the through strain engineering. These new metal- o a d e components in the perovskite lattice was B site requires spherical symmetry to bond free perovskite ferroelectrics are simple to d synthesize, low cost, and lightweight. f r o m The marriage between symmetry and H Ferroelectric origins in perovskites bonding in these metal-free compounds ex- Ferroelectricity in perovskites is induced by discrete structural changes that occur at phase transitions. tends the fields of perovskites and ferroelec- trics to a new realm. However, some exciting Polar diamine Ti4+ O2– X– aspects need further exploration. Bonding in- cation2+ + teractions between NH and halides extend 4 beyond H bonding, and their electronic na- ture needs to be further unveiled. Some of the transitions in these metal-free perovskites are 2+ Ba strongly first-order and ferroelastic, which leads to substantial shear strains and dis- NH4+ tinctive twin-wall dynamics (11). Ferroelastic Perovskite oxides Metal-free perovskites o Ferroelectricity primarily arises from the Ferroelectricity in the materials of Ye et al. twin walls show exotic physical properties n 4+ J displacement of the Ti ion from the center primarily stems from the order-disorder of beyond the bulk material (12), so knowledge u l of the TiO octahedron in BaTiO . the polar diamine cation. y 1 6 3 of the twin-wall nature in these metal-free 2 , perovskites would enrich their potential for 2 considered an alternative (3), but hybrid or- with six neighboring X sites for forming an high-density memory devices (13). Finally, 0 1 ganic–inorganic perovskites have not shown octahedron, replacing it with a molecular the existence of chirality in these molecular 8 comparable ferroelectricity and stability (4). cation that requires an octahedral H-bonding perovskites complicates the alignment of On page 151 of this issue, Ye et al. (5) report mode to bind with appropriate X-site accep- electric dipoles in their crystal lattices, so metal-free perovskites that can compete with tors could result in metal-free perovskite understanding its coupling with ferroelectric ferroelectric oxides (see the figure). compounds. Indeed, the first metal-free ordering could lead to unexpected properties Perovskites have a general formula ABX , perovskite, (piperazinium)(NH Cl )·H O, was at the bulk and nanoscale levels (14). j 3 4 3 2 where A and B are large and small metal cat- synthesized by assembling the NH4+ with REFERENCES AND NOTES ions, respectively, and X is a bitopic anion chlorides through H-bond interactions (9). 1. J. F. Scott, Science 315, 954 (2007). (for example, O2–) that coordinates to B. The Guided by these molecular design princi- 2. Y.-M. You et al.,Science 357, 306 (2017). structural topology can be maintained if the ples, Ye et al. have designed a new class of ferro- 3. B. Saparov, D. B. Mitzi, Chem. Rev. 116, 4558 (2016). 4. W. Li et al.,Nat. Rev. Mater. 2, 16099 (2017). A- and X-site ions are replaced with organic electric metal–free perovskites. The inclusion 5. H.-Y. Ye et al.,Science 361, 151 (2018). molecules within certain size ranges (6). The of noncentrosymmetric or chiral organic 6. G. Kieslich et al., Chem. Sci. 5, 4712 (2014). obvious starting point would be substituting amine cations imposed the polar symmetry 7. D. Weber, Z. Naturforsch. B 33, 1443 (1978). 8. W.-J. Xu et al., CrystEngComm 18, 7915 (2016). the large A-site metal cation, and the success- that rendered the desired ferroelectric order- 9. C. A. Bremner et al.,J. Am. Chem. Soc. 124, 10960 (2002). ful structural characterizations of MAPbX3 ing in the perovskite lattice. The prototypi- 10. J. Harada et al .,Nat. Chem. 8, 946 (2016). 11. E. K. H. Salje, Annu. Rev. Mater. Res .42, 265 (2012). (where MA is methylammonium, and X is cal compound MDABCO–NH I (MDABCO, E 4 3 12. S. Cherifi-Hertel et al.,Nat. Commun.8, 15768 (2017). C N E Cl, Br, or I) opened up such possibilities (7). bis-N -methyl-N 9-diazabicyclo[2.2.2]octo- 13. J. F. Scott, C. A. Paz de Araujo, Science 246, 1400 (1989). I C S 14. S. Seki et al.,Science 336, 19 8 (2012). / nium) exhibits a high Curie temperature (T of L c E K 1 Center of Rare Earth and Inorganic Functional Materials, C 448 K) and substantial spontaneous polariza- ACKNOWLEDGMENTS I B National Institute for Advanced Materials, Nankai University, 2 C. 2 tion (P of 22 µC/cm ), which compare well with We acknowledge funding from the National Natural Science : s C Tianjin 300350, China. School of Physics, Huazhong I the properties of commercial BaTiO . Such Foundation of China (no.. H University of Science and Technology, Wuhan 430074, China. 3 P A Email: wl276@ notable T and P values of MDABCO–NH I 10.1126/science.aat5729 R G c s 4 3 132 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Perspectives.indd 132 7/11/18 10:32 AM RETROSPECTIVE scientific rigor. Whereas Skou was the classi- cal biochemist, Jørgensen became a molecu- Jens Christian Skou (1918–2018) lar biologist. The two groups complemented each other in both style and results. I first met Skou in 1991 when I was a A pioneer in the biochemistry of membrane proteins master’s student in Aarhus University’s de- partment of chemistry. In Jens Nyborg’s lab, we were working on the EF-Tu protein and By Poul Nissen the same activity. Leaving aside his medical transfer RNA by crystallography. Skou came career plans, he instead devoted himself to to request the use of our French press homog- ens Christian Skou, a pioneer of bio- research. Skou and a slowly growing research enizer to disrupt cell membranes and isolate + + membrane research and molecular community characterized the enzyme, guided Na ,K -ATPase from shark rectal gland. I physiology, died on 28 May, just a by the excellent controls of tractable sub- used the device often, so people directed few months shy of his 100th birth- strates and specific inhibitors, and through- him to me. Crystallization of the sodium- day. The Danish physiologist dis- out his career, this system would define his potassium pump seemed an unreal possibil- Jcovered the enzyme known as the work. He remained at Aarhus University and ity at that time, but the encounter led to our sodium-potassium adenosine triphospha- continued to do research even after his retire- first discussions about how to proceed. + + tase (Na ,K -ATPase), or sodium-potassium ment in 1988. It wasn’t until 2007 that I talked again at pump, which exports sodium ions from In 1957, Skou published his first paper length with Skou. Our group, which included + + and imports potassium ions into cells. Al- on the Na ,K -ATPase enzyme. In it, he biophysicist Jesper Vuust Møller, had deter- D most all modern biochemistry, cell biology, speculated that the enzyme might be the mined the Ca2+-ATPase crystal structure and o and physiology textbooks showcase the long-sought sodium-potassium pump that then (with physiologist Bente Vilsen, also at w n + + l sodium-potassium pump, which accounts maintains steep electrochemical gradients Aarhus) the first crystal structure of Na ,K - o a + + d e for about 20 to 30% of the energy used by for Na and K across the cell membrane ATPase. Skou had invited himself to see it. I + + d cells through hydrolysis of ATP in human was still a newcomer to the Na ,K -ATPase f r o m and animal bodies and up to 70 to 80% in community, and I was quite nervous about the brain. In recognition of this first dis- hosting this giant of the field. Fortunately, covery of an ion-transporting ATPase, Skou Skou set me at ease by drawing me into a shared the 1997 Nobel Prize in Chemistry chat about sea trout fishing. with chemists Paul D. Boyer and John E. I learned that Skou was a man of the sea. Walker for their work on the mechanism of He grew up by the shores of Limfjorden, and ATP synthesis. his research ideas had materialized at Woods + + Skou grew up in the coastal town of Lem- Hole. His favorite sources for Na ,K -ATPase vig at Limfjorden in Northern Jutland, Den- studies came from crab and shark, and local mark. He earned his medical degree from fishermen we is most trusted suppliers. A the University of Copenhagen in 1944, and devoted fly fisher of sea trout, he spent time before any consideration of a research career on the shores of the beautiful peninsula of o had even struck him, he practiced medicine Mols across the Bay of Aarhus, whe e had n J at a small hospital in the town of Hjørring, a perfectly located summer cottage. u l y also in Northern Jutland. In 1947, to help his People would often describe Skou as stub- 1 2 , career in medicine along by earning a doc- born, but he was a modest man and a clear 2 torate degree, Skou moved to the then very and enables the majority of transmembrane supporter of the social equality that emerged 0 1 young Aarhus University to do research in its transport and signaling processes, includ- in Denmark in the 20th century. His wife, 8 small department of physiology. His work fo- ing the action potentials of firing neurons. Ellen Margrethe, was active in politics and cused on the mechanism of anesthetics. This He proposed that the pump should span a respected voice on the public health care brought him to lipid monolayers as models the cell membrane, an idea that at the time system. He was also a strong proponent of of the nerve cell membrane, and then to a was heavily disputed. Despite insight just work-family balance and would always finish research visit to the marine biology research gained from the myoglobin structure that work at the lab in the afternoon to join his station in Woods Hole, Massachusetts, in made the hypothesis seem unlikely, Skou family. His Nobel Prize gave him the author- 1953. The trip was a transformative moment persisted, and later in the 1960s and early ity to advocate giving researchers the free- of his career. Unlike his small Aarhus depart- 1970s, when better models of biomem- dom he had been given in his early career to ment, Woods Hole was a large and dedicated branes were presented and the first struc- pursue creative ideas and follow his curios- research environment with intense scholar ture of a membrane protein was visualized, ity. He wished that today’s younger genera- interactions and a huge library of scientific his theory was shown to be true. tions could have the same leeway to explore journals and books. There, Skou came across Aarhus University came of age, and in the that had allowed him to develop his original the curious observations of an ATPase activ- mid-1960s when physiologist Peter Leth Jør- discoveries. Jens Christian Skou personified ity in the nerves of giant squid. He took these gensen joined the department, a highly pro- the creativity and courage of a great scientist, D studies back home to Aarhus and applied ductive research community emerged with and he was a quiet and modest role model to R A A them to a more manageable source of nerves a critical spark of creativity and drive. Skou us all. He will live on in our research, discus- G D L E from local crabs, which he found displayed and Jørgensen’s internal competition and sions, and memories. j J K N fierce discussions were well-known; neither E ACKNOWLEDGMENTS R E O of them would prioritize compromises and S Danish Research Institute of Translational Neuroscience I am grateful to M. A. Skydsgaard and many colleagues for shar- : O “hygge”—a Danish social concept of relaxed ing their insights and memories. T (DANDRITE), Department of Molecular Biology and Genetics, O H Aarhus University, Aarhus, Denmark. Email: pn@mbg.au.dk tolerance, comfort, and contentment—over 10.1126/science.aau5275 P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 133 Published by AAAS DA_0713Perspectives.indd 133 7/11/18 10:32 AM INSIGHTS potential disputes using Creative Commons P O L I C Y F O RU M licenses, which allow creators to retain their copyrights while permitting others to copy CITIZEN SCIENCE and use their works. A similar tool is not widely available for Citizen science, public policy patents. In most countries, the legal inven- tor of a patentable discovery is one who con- tributes to its conception, and in the United New research models may benefit from policy modifications States, only inventors may apply for patents. Employers typically require their scientist employees to assign any rights to future in- By Christi J. Guerrini, Mary A. Majumder, promote public support of scientific activi- ventions as a condition of employment, in- Meaganne J. Lewellyn, Amy L. McGuire ties, professional associations have emerged cluding decisions to apply for patents and in the United States, Europe, and Australia to license inventions. By contrast, patent assign- itizen science initiatives that support support citizen science efforts, including con- ments do not appear to be a typical condition collaborations between researchers sideration of policy interactions. Meanwhile, of volunteering in citizen science projects. and the public are flourishing. As a re- the U.S. government recently passed legisla- Yet, it is possible that some citizen scientists sult of this enhanced role of the public, tion that supports agency use of citizen sci- will make contributions that support claims citizen science demonstrates more di- ence and crowdsourcing to conduct projects of sole or co-inventorship under domestic Cversity and flexibility than traditional that advance their missions (2). patent laws (4). For example, Sharon Terry, science and can encompass efforts that have a self-described citizen scientist who helped D o no institutional affiliation, are funded en- INTELLECTUAL PROPERTY discover the gene responsible for her chil- w n l tirely by participants, or continuously or sud- Citizen scientists are usually not paid for dren’s rare disease, pseudoxanthoma elasti- o a d e denly change their scientific aims. But these their contributions, although they might ben- cum, is a co-inventor on U.S. patents related d structural differences have regulatory impli- efit from participation in other ways. When to that discovery (5). Challenges may arise if f r o m cations that could undermine the integrity, citizen scientist volunteers are not com- the citizen science inventor who has retained safety, or participatory goals of particular cit- pensated, their work is not subject to most IP rights exercises them to exclude projects izen science projects. Thus far, citizen science employment-related laws and practices that from using the patented inventions  or dis- appears to be addressing regulatory gaps govern their scientific collaborators. Some agrees with co-inventors on whether to pur- and mismatches through voluntary actions employment-related laws and practices have sue patents or license inventions to others. of thoughtful and well-intentioned practitio- implications for ownership of IP, including Contracts can be used to clarify rights ners. But as citizen science continues to surge copyrights and patents. Although domestic IP and establish expectations related to pat- in popularity and increasingly engage diver- laws generally do not reach beyond national ent and other IP rights. Projects might take gent interests, vulnerable populations, and borders, international treaties have harmo- a page from employment practices and sensitive data, it is important to consider the nized basic IP protections. Thus, application require participants to assign future pat- long-term effectiveness of these private ac- of many U.S. protections to citizen science ent rights to project leaders as a condition tions and whether public policies should be contexts is generalizable to other countries, of participation. However, such a practice o adjusted to complement or improve on them. though some important differences persist. seems incompatible with models that view n J Here, we focus on three policy domains that U.S. copyright protection extends to au- citizen scientists as respected partners in the u l y are relevant to most citizen science projects: thors of every original work fixed in a tan- research process. The online citizen science 1 4 , intellectual property (IP), scientific integrity, gible medium. An exception exists for a game FoldIt, for example, does not require 2 and participant protections. “work made for hire” that is prepared by an patent assignments. Instead, FoldIt’s IP pol- 0 1 Although the definitional bounds of citizen employee within the scope of employment or icy provides that “players who contributed to 8 science are debated, there is general consensus by agreement as a specially ordered contribu- the discovery will be considered co-inventors that citizen science encompasses scientific en- tion to a collective work (3). Because citizen for any discovery produced through play” deavors in which individuals without specific scientists are volunteers, any copyrightable and issues of ownership will be handled at scientific training participate as volunteers in works that they develop in the course of a a later date by the University of Washington, one or more activities relevant to the research project—for example, photographs, writings, where the game was developed and is man- process other than (or in addition to) allow- and creative selections or arrangements of aged (6). It remains to be seen how well this ing personal data or specimens to be collected scientific data—likely do not constitute works approach will work in practice (7). from them. These activities might take place for hire absent valid agreements otherwise. Alternatively, contracts can incorporate at any point during the research process and Rather, the volunteer who created the works advance commitments by citizen scientists include participation in study design, data is the copyright “author” and owner. Other and professional scientists to cede steward- collection and analysis, and dissemination of countries do not recognize an automatic ship of patents to nonprofit organizations results (1). They might even encompass the transfer of ownership to employers or com- that support research. Ms. Terry’s advocacy entirety of the research process where, as in missioning parties (3). organization, for example, is the assignee of coordinated self-experimentation, the role of Where projects do not require copyright the patented inventions that she helped dis- professional scientists is minimal. assignment as a condition of volunteering, cover (5). However, it is unclear whether this Recognizing the potential for citizen sci- a citizen scientist who retains ownership of model would be readily accepted in contexts ence to advance scientific knowledge and his or her works under governing domestic where citizen scientists do not have funding law can refuse to grant permission to publish or other leverage to negotiate IP control. them, which could disrupt the scientific pro- Citizen scientists and professional sci- Baylor College of Medicine, Center for Medical Ethics and Health Policy, 1 Baylor Plaza, Houston, TX 77030, USA. cess or prevent the dissemination of findings. entists might forego patenting discoveries Email: guerrini@ However, projects can and often do avoid resulting from their collaborations or grant 134 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713PolicyForum.indd 134 7/11/18 10:37 AM each other nonexclusive, royalty-free cross-li- may prioritize other aspects of their par- reliable data on par with those produced by censes to any discoveries. Although the devel- ticipation. Also, there may be fewer op- professionals (12). But it is not yet clear how opment of standardized agreements for each portunities for professional scientists to well the best practices that have been de- of these scenarios seems achievable, support address knowledge gaps and otherwise act veloped address the many flavors of citizen for such agreements may be limited where in- as a check on quality if their involvement in science or are being followed by its practitio- terest in commercialization is high (7). One- projects is minimal. ners. These questions should be studied, but way material transfer agreements, which are Conflicts of interest can also undermine in the meantime, there may be opportunities a common vehicle for transferring research research integrity. Individuals who engage to promote integrity through policy. materials between institutions, might be in citizen science can have biases stemming For example, recently enacted U.S. legisla- adapted to promote sharing by and with citi- from their alliances with private, nonprofit, tion aimed at promoting citizen science and zen scientists. But in some cases, they may and political organizations, as well as their crowdsourcing projects by federal agencies add unnecessary complexity to what might involvement in lawsuits (9, 10). They may provides that all data collected through such otherwise be straightforward transactions, also have biases based on their perceptions efforts should be made publicly available and they are notoriously difficult to monitor of how they or their community might be where appropriate and to the extent possible and enforce. harmed or benefited by particular findings. (2). Such data accessibility creates opportuni- In sum, although contractual approaches Further, and especially with respect to re- ties to investigate questionable or poor-qual- have limits, the ability to tailor contracts search on politically charged topics, it is not ity data and assess fitness for use through to circumstances makes them more practi- beyond the realm of possibility that citizen independent examination (10). The law also cal than legislative approaches that would scientists might be recruited to provide bad requires agencies to “make all practicable ef- change the legal rules to produce different data to “improve” results or even sabotage re- forts” to ensure that participants adhere to federal research misconduct policies, which D o include sanctions for fabricating and falsi- w n l fying data (2). For studies conducted with o a d e federal funds, the penalties for research mis- d conduct—the most severe being a funding f r o m ban—will not apply to most citizen scientists because they do not seek such funding. How- ever, their interests will be affected if profes- sional collaborators lose the federal support that makes possible their work. These regulations will not reach citizen sci- ence projects that are not conducted with U.S. federal employees or not federally funded. Journals that publish citizen science research help fill this gap through the peer review process and by placing pressure on authors to make their study data publicly available. o Although these policies do not affect citizen n Citizen scientists J science projects that are uninterested in u collect samples l y in a U.S. traditional publication, some projects may 1 4 , national park. have commitments to openness that match 2 or even exceed journal requirements. For ex- 0 1 ample, DIY publishes protocols 8 default outcomes—assuming that consensus search, although careful monitoring will help and data on project wikis. on optimal outcomes even could be reached. ferret out such attacks. A related problem is presented when citi- As a policy matter, resources are probably Of course, research conducted exclusively zen scientists have conflicts of interest that better spent encouraging transparency and by professional scientists can also be un- are not disclosed. Many national regulations negotiation of IP terms at the outset of citizen dermined by personal biases and conflicts and journals have adopted requirements of science collaborations. of interest, but institutional rules, funding conflicts reporting, but they may not reach stipulations, regulatory procedures, and pro- citizen scientists who do not qualify as inves- RESEARCH INTEGRITY fessional norms work to compel their identi- tigators or authors, although they facilitate Concerns about the quality of data contrib- fication and disclosure whether the research research in meaningful ways. To promote uted by citizen scientists and the soundness is conducted in for-profit, nonprofit, or gov- accountability, such requirements should of their collection, reporting, and analyti- ernmental settings. Citizen science projects extend to the disclosure of relevant conflicts cal techniques have long been raised (8, that are not embedded within institutions, held by citizen scientists. Problems will arise, 9). Although professional scientists are not are self-funded, or are otherwise outside of however, when disclosure is incompatible immune from quality transgressions, meth- regulatory control may operate without these with volunteer terms that allow anonymous odological rigor is central to their training traditional safeguards. participation. We support project discretion and professional advancement, and formal Citizen science projects have adopted to offer anonymity, especially where sensitive mechanisms exist for holding professional strategies to promote research integrity, and information is involved or citizen scientists scientists accountable for the quality of many guides, tools, and templates are avail- could be subjected to intimidation or harm S P their work. Volunteers, on the other hand, able to support projects from the planning if identified. In such cases, consideration N : O may not experience similar external pres- stages through evaluation (11, 12). Studies should be given to permitting disclosure at T O H sures to ensure research integrity (10) and indicate that citizen scientists can produce the aggregate level (10). P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 135 Published by AAAS DA_0713PolicyForum.indd 135 7/11/18 10:37 AM INSIGHTS | POLICY FORUM PARTICIPANT PROTECTIONS of funding source. Like the Common Rule, If it is determined that additional regula- As uncompensated volunteers, citizen scien- however, FDA research subject protections tion is warranted, it is unlikely to occur at the tists are unable to rely on traditional labor do not explicitly account for risks and ben- federal level given that a proposal to extend laws to prevent and redress harm because efits associated with supporting scientific the Common Rule to additional kinds of re- they do not qualify as protected workers. The initiatives in ways other than serving as re- search was recently considered and rejected. Common Rule, which applies to all research search subjects. At a minimum, however, guidance should be involving human subjects that is conducted For those citizen science initiatives out- developed by citizen science practitioners or supported by the U.S. government, is di- side the scope of these federal regulations and participants in collaboration with ethi- rected at protecting research participants. and not conducted in the few states that cists to aid unregulated projects that cannot For studies governed by the Common Rule extend these kinds of protections to all re- afford independent IRB review in conducting and not within an exemption, institutional search regardless of funding source, there risk-benefit assessments. Such guidance also review boards (IRBs) must ensure that any may be no legal requirement to evaluate or may be useful to traditional IRBs evaluating risks to subjects are minimized and reason- disclose potential risks to citizen scientists. regulated projects in which the role of citi- able in relation to anticipated benefits. Stud- Individuals owe a duty of ordinary care to zen scientists is multi-faceted or ambiguous. ies that take place in other countries may avoid foreseeable injury to others, includ- In addition, modes of risk assessment and still be subject to the Common Rule or simi- ing general volunteers, but application of management that are complementary or al- lar national requirements, so consideration this common law doctrine is highly context- ternative to those required by the Common of the Common Rule’s provisions may be specific and may be unavailable to volun- Rule and FDA regulations, including but not broadly instructive. teers for any number of reasons. Moreover, limited to ethics reviews conducted by non- Importantly, Common Rule protections some projects may require waivers of lia- traditional IRBs and citizen ethicists, should are directed toward “subjects” of research bility as a condition of volunteering, which be evaluated to understand the citizen sci- D o from whom identifiable data or biospeci- are not allowed in the context of federally ence contexts in which they could be useful. w n l mens are collected. They regulated research. As citizen science becomes more prevalent, o a d e do not explicitly authorize Some unregulated proj- it is increasingly important that policy oppor- d IRBs to consider risks or ects, like Sage Bionetworks’ tunities to support participants and practitio- f r “…citizen scientists o m benefits to citizen scientists mPower study of Parkin- ners are identified and responsibly pursued. who facilitate research in are both subjects son’s disease, which collects In the end, some policy adjustments may other ways. Depending on data through participants’ be in the best interests of society, but to be the setting and study de- and facilitators of mobile phones, voluntarily successful, they must appreciate the distinct sign, examples of specific research...” close this gap by securing ethos of citizen science and be guided by its risks to citizen scientists independent IRB review of diverse stakeholders. j could include expectations their protocols, although REFERENCES AND NOTES of overwork, requirements to assume finan- this can be cost-prohibitive for many projects. 1. J. L. Shirk et al.,Ecol. Soc. 17, 29 (2012). cial burdens, or vulnerability to harassment Others have proposed ethics evaluations by 2. Crowdsourcing and Citizen Science Act, Public Law 114- by others. Examples of specific benefits, on “citizen ethicists” who critique experiments 329 (6 January 2017), codified at 15 U.S.C. § 3724. 3. S. J. Marsnik, R. M. Lorentz, “Who owns employee works? the other hand, could include training in and post their opinions online for potential Pitfalls in a globally distributed work environment,” in scientific techniques, access to tools and participants to review, as well as download- Managing the Legal Nexus Between Intellectual Property and Employees: Domestic and Global Contexts , L. J. o data for personal use, or opportunities for able ethics toolkits geared toward citizen sci- n J Oswald, M. A. Pagnattaro, Eds. (Elgar, 2015), chap. 8. coauthorship. Where citizen scientists are entists (15). Meanwhile, do-it-yourself biology 4. T. Scassa, H. Chung, “Managing intellectual property u l y rights in citizen science” (Commons Lab, Woodrow Wilson both subjects and facilitators of research, projects may voluntarily adhere to a commu- 1 International Center for Scholars, 2015). 4 , IRBs may interpret their jurisdiction to in- nity code of ethics, which includes directives 2 5. S. F. Terry et al.,Nat. Rev. Genet.8, 157 (2007). clude risks and benefits associated with both to pursue peaceful purposes and adopt safe 6. FoldIt, Terms and services of consent v. 3 (8 March 2013); 0 1 roles. However, in some contexts, citizen practices. Until recently, the community also . 8 7 L. S. Vertinsky, Fl. State Univ. Law Rev. 41, 1067 (2014). scientists are only research facilitators and supported an online forum through which 8. A. C. J. W. Janssens, P. Kraft, PLOS Med.9, e1001328 (2012). not also subjects, or their roles may be am- professional biosafety experts answered 9. Rise of the citizen scientist. Nature 524, 265 (2015). 10. D. B. Resnik, K. C. Elliott, A. K. Miller, Environ. Sci. Policy 54, biguous (13). Federal protections also do not questions posed by citizen scientists. How- 475 (2015). explicitly contemplate risks and benefits to ever, some projects may fail to engage in any 11. A. Wiggins et al., “Data management guide for public par- communities, although community-based kind of risk assessment or management. ticipation in scientific research” (DataONE PPSR Working Group, 2013); research review processes, including com- Of the three policy domains discussed, steps/accept/DataONE-PPSR-DataManagementGuide. munity IRBs and advisory committees, have participant protections require the most pdf. 12. M. Kosmala, A. Wiggins, A. Swanson, B. Simmons, Front. emerged to complement traditional IRB re- immediate attention given that they impli- Ecol. Environ. 14, 551 (2016). view to ensure that involved communities cate direct physical harms, yet collective 13. R. Gellman, “Crowdsourcing, citizen science, and the law: are engaged in and directly benefit from pro- action focused on assessing those harms legal issues affecting federal agencies” (Commons Lab, Woodrow Wilson International Center for Scholars, 2015). posed research and that study designs are relative to other risks and benefits specific 14. M. A. Rothstein, J. T. Wilbanks, K. B. Brothers, J. Law Med. culturally appropriate. to citizen science has thus far been limited. Ethics 43 , 425 (2015). 15. A. D. Marcus, “The ethics of experimenting on yourself,” Many U.S.-led citizen science initiatives How citizen science projects are making Wall Street J. (24 October 2014); are not covered by the Common Rule be- these risk-benefit calculations, and accord- the-ethics-of-experimenting-on-yourself-1414170041. cause they are not federally conducted or ing to what processes, should be empirically ACKNOWLEDGMENTS supported (14). A small number may never- investigated. Then, whether these data jus- Development of this manuscript was funded by the U.S. National theless be covered by research subject pro- tify extending current legal protections to Human Genome Research Institute grant KO1- HG009355. The tections adopted by the U.S. Food and Drug unregulated citizen science, or creating new authors declare no financial conflicts of interest and thank C. Administration (FDA), which apply to clini- policy frameworks altogether, merits study Lee for research assistance. cal investigations that support applications by policy-makers in partnership with the for products regulated by FDA, regardless citizen science community. 10.1126/science.aar8379 136 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713PolicyForum.indd 136 7/11/18 10:37 AM B O O K S e t a l . ECOLOGY To be a bee A charming account celebrates the insects’ idiosyncrasies and the people passionate about protecting them By Rachael Winfree e have bees to thank for some of the better features of our world, from the fruits and vegetables we D o eat to the world’s flowering plants, The sweat bee Augochloropsis sumptuosa is one of the roughly 20,000 known bee species. w n l which radiated alongside them o a d e Wduring the Cretaceous. Nearly 90% Who knew, for example, that “dumbledore” by weighing a bee and a wasp before and d of plant species require pollinators (an added is a historical name for the bumblebee? The after dipping each in flour. f r o m benefit for anyone who suffers from wind- book’s bibliography reflects the diversity of Hanson also has a sharp eye for the id- borne pollen allergies). Humans even seem to topics covered, ranging from Virgil and Tol- iosyncratic—which he doesn’t have to look share aesthetic preferences with bees, insofar stoy to the scientific literature and . far to find among bees or among bee en- as the shapes, colors, and scents of the flow- In one of the more surprising chapters thusiasts. In the first category, he describes ers we enjoy originally evolved to appeal to (even to an ecologist who studies bees), Han- bees that nest inside active termite mounds these and other pollinators. son presents evidence for the key role played and others that construct “elaborate vases” Given the necessity of bees, the signs that by honey bees in human evolution. He begins made of flower petals. Still others carry they are in trouble are concerning. Although with the honeyguide, an African bird species pollen on their abdomens, making each m a l a d i e s o f t h e d o m e s t i c a t e d that feeds on bee hives but needs bee “look like she’s wearing a tiny, brightly honey bee (Apis mellifera ), such as another, larger species to conduct colored apron” (as compared with most colony collapse disorder, have cap- the initial raid on the hive. The other bees, “whose pollen loads look more tured the most public attention, honey badger was long thought to like stockings pulled up high on their back o d e c l i n e s i n t h e w o r l d ’ s 2 0 , 0 0 0 be the honeyguide’s original mu- legs.”) In the second, are a retired insurance n J wild bee species have the potential tualist partner. Bringing some ba- salesman who gives bees away as presents u l y to be even more ecologically cata- sic biology to bear, Hanson points and an alfalfa farmer who slows his car to 1 2 , strophic. In Buzz, Thor Hanson out that honey badgers don’t like a crawl while driving so as to avoid killing 2 deftly presents the predicament to climb trees and in addition even a single bee. 0 1 8 Buzz of bees today, in part by recount- are nearsighted, hard of hearing, The only criticism I have of Buzz, and The Nature and ing conversations with scientists and nocturnal. By contrast, the it is not a major one, is that the book un- Necessity of Bees and practitioners who are leading Thor Hanson Hadza people of Tanzania eagerly derplays how little we actually know about the effort to understand and con- Basic Books, 2018, seek out honeyguides and har- bees. The chapter covering the decline and serve bees. In only 304 pages, he 304 pp. vest honey from wild honey bee possible extinction of certain North Ameri- successfully covers a wide range hives—so much so that honey con- can bumblebee (Bombus) species quite of topics, from bee evolution and natural his- stitutes 15% of the calories in the Hadza diet. rightly focuses on the extent and causes of tory to the economic value of pollination. The availability of honey may even have fur- these declines. It doesn’t mention, though, R Like all good naturalists, Hanson has a gift thered the evolution of our unusually large that Bombus is among the best-studied of K C I for seeing the world from his subject’s point brains, the book reveals. the world’s 443 bee genera. The conserva- L F / B of view. At one point, he predicts the nest site Hanson is an upbeat and often humor- tion status of most bee species is simply un- A L G preferences of a bumblebee queen (a boot, it ous guide, with a DIY attitude, the latter of known. In the United States, for example, N I R O turns out). At another, he empathizes with which sometimes lands him in odd situa- there is no federal list of rare or declining T I N O bees that won’t pollinate alfalfa because the tions (such as sitting in a McDonald’s, us- bee species, and ad hoc assessment of which M D flowers’ anthers bop them on the head. ing tweezers to separate the bee-pollinated species are in trouble has been left to re- N A Y In addition to his interest in entomology, from the non–bee-pollinated components searchers and the Xerces Society for Inver- R O T N Hanson also has an interest in etymology. of a Big Mac). In another chapter, after tebrate Conservation. E V N reading that branched, pollen-collecting Overall, however, if you have time to read I E E B The reviewer is at the Department of Ecology, Evolution, hairs are a distinctive trait that separates one book on what is happening with mod- S G and Natural Resources, Rutgers University, New Brunswick, S bees from their evolutionary forebears the ern bees, you couldn’ t do better than Buzz. j U : NJ 08901, USA, and is on the board of directors of the Xerces O wasps, he tests the hypothesis that minute T Society for Invertebrate Conservation, Portland, OR 97232, O H USA. Email: rwinfree.rutgers@ particles adhere better to bees than wasps 10.1126/science.aat9250 P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 137 Published by AAAS DA_0713Books.indd 137 7/11/18 10:21 AM INSIGHTS | BOOKS FILM Learning from dif erent disciplines Conversations spark connections as scientists search for inspiration in other fields By Brian Uzzi ologist Victoria Orphan describes how tiny spans are increasing, social media and fren- microbes, too small to see with the naked zied news cycles may lead to the perception f you knocked on Heaven’s door, and God eye, can “eat” greenhouse gases and there- that life is shorter and less fulfilling. greeted you, what question would you fore are a potential source of big solutions Science is replete with examples of mys- ask? What is the nature of human con- for climate change. Orphan then travels to teries that were unlocked through the ac- sciousness, and how can it be expanded? Boulder, Colorado, where she meets physi- cumulation of sometimes small and often Where does the Universe begin and end? cist Jun Ye, who is obsessed with measur- separate advances that came from varied IWhat is time, and why isn’t it ing time with a clock so precise perspectives. The process of discovery is un- constant? What causes deviance? that it only loses one second in certain. What looks like a breakthrough may And further still, if you received The Most Unknown accuracy every million years. prove to be a dead end, whereas accidents an answer, could you decipher Ian Cheney, director As the two scientists converse, may provide a result—a piece of a larger it? Or would the answer only be Vice/Motherboard, 2018, they make a potential connec- puzzle—that suddenly makes all the other a clue leading to the next clue? 88 minutes. Available tion: The secret to the success separate pieces make sense collectively. D o w The Most Unknown, a docu- on Netf ix, August 2018. of greenhouse-eating microbes On an uninhabited island off the coast of n l mentary film in the Simons may be their particular experi- Puerto Rico, Seth meets Yale psychologist o a d e Foundation Science Sandbox series, takes ence with time. At the bottom of the ocean, Laurie Santos, who studies deviance—theft, d the viewer on a fantastical journey of nine strong gravitation waves bend and slow to be exact. Santos is specifically interested f r o m scientists as they intrepidly knock on Heav- down time, reveals Ye. The change isn’t a in the conditions under which monkeys in en’s door. Each asks a profound question in a lot, but it may be a clue to understanding the wild risk taking a five-finger discount different way. But all are making a difference the microbes’ specific dietary requirements. of succulent treats. In a moment of levity, an accident messes up the experiment to be filmed, but the presence of the documen- tary crew reveals something unexpected about monkey behavior: Even if the re- searcher is looking away, attempted theft is less likely to occur when there are potential third-party witnesses. A serendipitous mis- take gives Santos her next clue. Woven into the film’s electrifying tapestry o of ideas, persons, and places is the story of n J the hard work of science. Great discovery, we u l y see, comes part and parcel with emotional 1 2 , frustration and disappointment, unusual 2 hours that buck circadian rhythms, tight 0 1 spaces, cold water, and dark places. 8 In The Most Unknown, Cheney connects apparently disparate journeys of discovery, illustrating where interdisciplinary team- work can fill in a scientist’s blind spots. Microbiologist Jennifer Macalady collects samples from a water source in the Frassasi caves in Italy. When seen through this lens, science can be viewed as a big import-export business in how science explains the Universe, the From Boulder, Ye travels to Sussex, United of ideas. Conventional, well-understood world we live in, and the worlds within us. Kingdom, where the world’s most accurate ideas in one area, when brought into an- Although the filmmaker, Ian Cheney, is clock maker comes face to face with neuro- other scientific domain, are suddenly seen not a scientist, he makes a point that is now scientist Anil Seth, who studies why humans in a new light. well documented scientifically: Scientists can’t keep time without reference points Newton’s observations that “standing from different disciplines who immerse of activity. In Seth’s laboratory, subjects in on the shoulders of giants” is key to sci- themselves in one another’s work and ex- magnetic resonance imaging machines are entific discovery may only have been half change ideas solve the hardest problems asked to guess the runtime of video clips. right. As documented in The Most Un- (1). Midway through the film, for example, They find that the human perception of time known, scientists also have much to learn California Institute of Technology geobi- is linked to the activity level in the video. from their contemporaries. j A When persons in a video a ighly active M A REFERENCE R The reviewer is at the Kellogg School of Management, or objects move quickly, subjects underesti- O M 1. B. Uzzi, S. Mukherjee, M. Stringer, B. Jones, Science 342, A the McCormick School of Engineering, and the R mate the length of the video and vice versa 468–472 (2013). B A Northwestern Institute on Complex Systems (NICO), : for low-activity videos. One potential impli- O Northwestern University, Evanston, IL 60208, USA. T O Email: uzzi@ cation of his work is that although our life 10.1126/science.aau1692 H P 138 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Books.indd 138 7/11/18 10:21 AM Brazil may remove LETTERS restrictions on the trade of introduced species such as this American bullfrog. D o w n l o a REFERENCES d e Edited by Jennifer Sills (Clarias gariepinus), and American bullfrog d (Lithobates catesbeianus) have established 1. J. R. S. Vitule et al., Nature 513, 315 (2014). f r 2. J. R. S. Vitule et al.,Science 347 , 1427 (2015). o m Brazil naturalizes populations in some localities (7), and the 3. F. A. Bockmann et al.,Science 360 , 865 (2018). approval of this new ordinance will permit 4. Portaria Nº 103 (4 May 2018); . non-native species their free trade and rearing across Brazil. org/Recursos/Arquivos/editor/portariasmc103.pdf This retrogression conflicts with several [in Portuguese]. 5. F. M. Pelicice et al., Conserv. Lett. 7 , 55 (2014). Brazil’s national policies are putting the Aichi Biodiversity Targets, especially the 6. K. L. Speziale et al.,Biol. Inv.14, 1609 (2012). country’s megadiversity at risk (1–3). After one related to the prevention, control, or 7. A. O. Latini, D. C. Resende, V. B. Pombo, L. Coradin, passing a 2016 law that will put constraints eradication of non-native species (8). Brazil Espécies exóticas invasoras de águas continentais no Brasil (Ministério do Meio Ambiente, 2016) on biodiversity research (3), the Ministry harbors the most diverse aquatic biota [in Portuguese]. of Agriculture, Livestock, and Supply has in the world (9), and it is imperative that 8. D. P. Lima-Junior et al.,Ambio 47, 427 (2018). taken another controversial action: A new local authorities take appropriate measures 9. A. A. Padial et al.,Biodivers. Conserv. 26, 243 (2017). 10. V. M. Azevedo-Santos et al.,Biodivers. Conserv. 26, 1752 ordinance proposes that introduced aquatic that value and preserve native biodiversity. (2017). species in Brazil should be considered Basic research and knowledge produced o 10.1126/science.aau3368 “native” (4), including invasive species in n by scientists (10) should play a vital role in J the Neotropics (1, 5). Classifying introduced u l these decisions. y aquatic species as if they are indigenous 1 1 2 Marcelo Fulgêncio Guedes Brito, * André Free satellite data , to Brazil could potentially cause even 2 Lincoln Barroso Magalhães, Dilermando 2 more introductions and lead to the loss of 0 1 3 key to conservation Pereira Lima-Junior, Fernando Mayer ecosystem services and functions, as well as 4 5 8 Pelicice, Valter M. Azevedo-Santos, Diego traditional knowledge about native spe- 6 Biodiversity is in crisis, with extinction Azevedo Zoccal Garcia, Almir Manoel cies (6). Moreover, Brazil shares some large 7 8 rates orders of magnitude higher than Cunico, Jean Ricardo Simões Vitule river basins (such as Paraguai, Paraná, and 1 Programa de Pós-Graduação em Ecologia e background levels (1). Underfunded con- Amazon) with other countries. Therefore, Conservação, Universidade Federal de Sergipe, servationists need to target their limited Brazil will become a major source of non- São Cristóvão, SE, 49100-000, Brazil. 2 Programa resources effectively. Over the past decade, de Pós-Graduação em Tecnologias para o native species for other countries in South Desenvolvimento Sustentável, Universidade Federal satellite remote sensing has revolutionized America. The rate of introductions in Brazil de São João Del Rei, Ouro Branco, MG, 36420-000, our ability to monitor biodiversity globally, will likely outpace the research investigating Brazil. 3 Laboratório de Ecologia e Conservação de and is now used routinely, especially by their negative effects (1, 3). Ecossistemas Aquáticos, Universidade Federal do nongovernmental organizations, to detect Mato Grosso, Pontal do Araguaia, MT, 78698-000, This is not the first time that political Brazil. 4 Núcleo de Estudos Ambientais, Universidade changes, set priorities, and target conser- decisions have tried to categorize non-native Federal do Tocantins, Porto Nacional, TO, 77500- vation action. The U.S. Geological Survey species as native. In 2009, the Brazilian 000, Brazil. 5 Universidade Estadual Paulista “Júlio (USGS) unlocked high-resolution Landsat de Mesquita Filho,” Botucatu, SP, 18618-970, Brazil. Congress proposed a law that intended to 6 Laboratório de Ecologia de Peixes e Invasões data in 2008 (2), making data available 7 “naturalize by decree” several non-native Biológicas, Universidade Estadual de Londrina, online (3), and the Copernicus program 1 L L 7 I fishes to foster aquaculture development Londrina, PR, 86057-970, Brazil. Programa de from the European Commission subse- E N Pós-Graduação em Aquicultura e Desenvolvimento C (5). The most recent ordinance is based on quently made their data available as well M C Sustentável, Universidade Federal do Paraná, Palotina, / M a 2016 law that considers introduced spe- 8 (4). These resources have been instrumen- O PR, 85950-000, Brazil. Laboratório de Ecologia e C K. cies with established populations as part of Conservação, Setor de Tecnologia, Departamento tal to biodiversity research. Assessments of C O de Engenharia Ambiental, Universidade Federal do T the Brazilian genetic heritage. Non-native environmental changes such as deforesta- S I : Paraná, Curitiba, PR, 81531-970, Brazil. O species such as the Malaysian giant prawn tion are now readily available. The current T *Corresponding author. O H (Macrobrachium rosenbergii), African catfish Email: marcelictio@ spatial and spectral resolution of Landsat P SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 139 Published by AAAS DA_0713Letters.indd 139 7/11/18 10:25 AM INSIGHTS | LETTERS data make them appropriate to many con- Funding agencies can 4. R. E. Bell, L. S. Koenig, Science 358, 1223 (2017). servation applications, and although they 5. E. Marín-Spiotta, Nature 557, 141 (2018). are not always ideal, pragmatic researchers prevent harassment 10.1126/science.aau3979 with limited resources use them regularly. Conservationists have already called for Harassment and lack of physical safety in these data to remain free (5). Consequently, fieldwork and laboratories exists across a TECHNICAL COMMENT ABSTRACTS the news that USGS may charge for data range of disciplines (1, 2). Editorials and (6) is deeply troubling. # stories have recently highlighted Comment on “Designing river flows to USGS has recently convened an advisory that research is often conducted under improve food security futures in the Lower committee to determine whether users “macho” conditions in which harassment, Mekong Basin” would be prepared to pay for increased bullying, and unsafe work environments John G. Williams spectral and spatial resolution images (7). are common (3, 4). In response, codes of Sabo et al. (Research Articles, 8 December Requiring users to pay would put these conduct for researcher safety are on the 2017, p. 1270) use sophisticated analyses of images beyond the reach of conservation- rise (3, 5). However, national research flow and fishery data from the Lower Mekong ists. It would halt time-series analyses funds, private funding organizations, and Basin to design a “good” hydrograph that, that have been useful in monitoring the monitoring agencies rarely require that if implemented by planned hydropower effects of climate change, land-cover the recipients of their grants implement dams, would increase the catch by a factor of change, and ocean surfaces, likely hinder- codes of conduct or safety standards (2). 3.7. However, the hydrograph is not imple- ing the achievement of the Sustainable Opportunities for cultural change should mentable, and, if it were, it would devastate the Development Goals (8). We urge the USGS rest not only with individual scientists, fishery. Further, the analyses are questionable. to reconsider their position and continue teams, and professional societies. Funding Full text: /10.1126/science.aat1225 D o to provide data from the Landsat program agencies should share the responsibility. w n Comment on “Designing river flows to l freely to all users. The cost of ensuring researcher safety o a 1 1 improve food security futures in the Lower d G. M. Buchanan, * A. E. Beresford, M. should be part of the overall budget, and e Mekong Basin” d 2 3 Hebblewhite, F. J. Escobedo, H. M. De predefined safety standards should prevent f r 4 5 6 Ashley S. Halls and Peter B. Moyle o m Klerk, P. F. Donald, P. Escribano, L.P. situations in which harassment could 7 8 9 occur (2). For example, when companies or The designer flow regime proposed by Sabo Koh, J. Martínez-López, N. Pettorelli, 10 4 7 institutions need scientists to do con- et al. (Research Articles, 8 December 2017, A. K. Skidmore, Z. Szantoi, K. Tabor, M. Wegmann,11 S. Wich12 tracted monitoring work, bidding prices p. 1270) to support fisheries in the Lower 1 RSPB Centre for Conservation Science, Royal often determine whom they select. Unless Mekong Basin fails to account for important Society for the Protection of Birds, Edinburgh, EH12 funding agencies require safety standards, ecological, political, and economic dimen- 9DH, UK. 2Wildlife Biology Program, Department of Ecosystem and Conservation Sciences, W.A. Franke such bidding prices will always favor sions. In doing so, they indicate that dam College of Forestry and Conservation, University low-cost solutions that neglect safety. As impacts can be easily mitigated. Such an of Montana, Missoula, MT 59812, USA. 3 Faculty of another example, when principal investiga- action would serve to increase risks to food Natural Sciences and Mathematics, Universidad del and livelihood futures in the basin. Rosario, Bogotá, DC, 11122, Colombia. 4 Department tors (PIs) write applications, they should of Geography and Environmental Studies, budget for training and counseling to Full text: /10.1126/science.aat1989 Stellenbosch University, Stellenbosch 7602, South prevent and address harassment. Such o Africa. 5 BirdLife International, David Attenborough Response to Comments on “Designing river n Building, Pembroke Street, Cambridge, CB2 3QZ, measures would be more widespread if flows to improve food security futures in the J UK. 6CAESCG, University of Almería, Cañada de u funding agencies acknowledged them. l 7 Lower Mekong Basin” y San Urbano s/n 04120 Almería, Spain. Betty Funding agencies have the power to 1 Gordon Moore Center for Science, Conservation 2 G. W. Holtgrieve, M. E. Arias, A. Ruhi, , 8 participate in changing the culture by International, Arlington, VA 22202, USA. BC3- 2 V. Elliott, So Nam, Peng Bun Ngor, T. A. 0 Basque Centre for Climate Change, Scientific requiring codes of conduct for accept- 1 Campus of the University of the Basque Country, able behavior from their grant recipients. Räsänen, J. L. Sabo 8 48940, Leioa, Spain. 9 Institute of Zoology, Zoological Sabo et al. presented an empirically derived Society of London, Regent’s Park, London, NW1 Forcing researchers and companies to 4RY, UK. 10 University of Twente, Faculty of Geo- incorporate safety standards into grant algorithm defining the socioecological Information Science and Earth Observation, 7500 proposals and assignment bids will increase response of the Tonle Sap Dai fishery in the AE Enschede, Netherlands. 11 Institute of Geography awareness about harassment and stressful Cambodian Mekong to basin-scale variation and Geology, 97074 Würzburg, Germany. 12School of Natural Sciences and Psychology, Liverpool John working environments. Only through full in hydrologic flow regime. Williams suggests Moores University, Liverpool, L33AF, UK. support from the broad spectrum of players that the analysis leading to the algorithm is *Corresponding author. Email: graeme.buchanan@ involved in science will it be possible to cre- flawed because of the large distance between .uk ate an inclusive and responsible culture that the gauge used to measure water levels REFERENCES ensures safe workspaces. (hydrology) and the site of harvest for the 1. J. DeVos et al., Cons. Biol. 29, 452 (2015). Lars L. Iversen1,2* and Mette Bendixen3 fishery. Halls and Moyle argue that Sabo et al.’s 2. C. Woodcock et al.Science 320, 1011 (2008). 1School of Life Sciences, Arizona State University, findings are well known, and contend that the 3. USGS, Earth Explorer (/). 4. European Commission, Copernicus (http://copernicus. Tempe, AZ 85281, USA. 2Center for Macroecology, algorithm is not a comp ensive assess- eu/). Evolution and Climate, National Museum of Natural ment of sustainability. We argue that Williams’ 5. W. Turner et al.,Biol. Conserv.182, 173 (2015). Sciences, University of Copenhagen, Denmark. critique stems from a misunderstanding 6. G. Popkin, Nature 556, 417 (2018). 3 Institute of Arctic and Alpine Research, University of 7. USGS, Landsat Advisory Group undertakes a Landsat Cost Colorado, Boulder, CO 80309, USA. about our analysis; further clarification of the Recovery Study (2018); *Corresponding author. Email: lliversen@bio.ku.dk analysis is provided. We regret not citing more landsat-advisory-group-undertakes-a-landsat-cost- of the work indicated by Halls and Moyle, yet recovery-study. REFERENCES we note that our empirical analysis provides 8. Sustainable Development Goals (https://sustainable 1. K. B. Clancy et al.,PLOS ONE 9, e102172 (2014). /?menu=1300). 2. M. A. Rinkus et al.,Soc. Nat. Res. additional new insights into Mekong flow- 10.10802018.1471177 (2018). fishery relationships. 10.1126/science. aau2650 3. Nat. Ecol. Evol. 1, 1787 (2017). Full text: /10.1126/science.aat1477 140 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713Letters.indd 140 7/11/18 10:25 AM RESEARCH ◥ for the Tonle Sap–Great Lake system, and perhaps TECHNICAL COMMENT for other reaches of the lower Mekong with large floodplains. Implementing the “good” flow regime would en- SUSTAINABILITY tail managing releases from the most downstream dam, well upstream from the Tonle Sap–Mekong Comment on “Designing river flows to confluence. During the high-flow period, releases would be large and the floodplain would be in- undated. At the end of the high-flow period, improve food security futures in the releases from the dam would be cut to create the desired decrease in stage. Downstream from Lower Mekong Basin” the dam, water would drain from the flood- plain to the channel as soon as releases were cut, reducing the rate of decrease in stage (and John G. Williams discharge) in the channel farther downstream. The opposite would happen, more strongly be- Sabo et al. (Research Articles, 8 December 2017, p. 1270) use sophisticated analyses of cause of downstream overland flow, at the end flow and fishery data from the Lower Mekong Basin to design a “good” hydrograph that, if of the low-flow season, prolonging the transition implemented by planned hydropower dams, would increase the catch by a factor of 3.7. from dry to wet conditions. The situation is too However, the hydrograph is not implementable, and, if it were, it would devastate the complex for simple analysis, but based on hydro- fishery. Further, the analyses are questionable. dynamic modeling, about 35 km3 of water flowed onto the left bank of the Mekong upstream and D o n a methodological tour de force, Sabo et al. w n l (1) relate components of the flow regime of o a the Mekong River, quantified as stage at Stung d e Treng, to catch in the important Dai fishery on d f r Ithe Tonle Sap River, Cambodia. They use these o m empirical relationships to design a “good” flow regime proposed as an “ecological objective func- tion” for the operation of controversial dams planned for the Mekong River. They project that this flow regime would increase yield of the Dai fishery by a factor of 3.7. However, simple flow routing considerations show that the “good” re- gime could not be implemented, and if it were, it would devastate the Dai fishery. Sabo et al. also depict the historical flow re- gime and a “bad” regime as plots of stage over time (Fig. 1). The plot of the “good” regime is a o roughly rectangular wave, with long low-flow pe- n J riods connected to shorter high-flow periods by u l y 1 transitions ~15 days long. The transitions are more 5 gradual in the “bad” regime. , 2 Stung Treng is several hundred kilometers 0 1 upstream from the Tonle Sap–Mekong conflu- 8 ence. The Tonle Sap connects the Mekong to a large natural lake and adjacent floodplains that support large populations of many fishes (2). Water flows up the Tonle Sap to the Great Lake during high flows in the Mekong, and back down the Tonle Sap to the Mekong during low flows. Flow in the Mekong is a major driver of stage in the Great Lake, but the Great Lake receives ~40% of its water from tributaries and rainfall (3), and the stage records differ substantially (Fig. 2). Sabo et al. state that “…we used estimated hydrologic drivers of the historical bag net, or ‘Dai,’ fishery on the Tonle Sap River—the largest commercial fishery in the Mekong—to design bet- ter fisheries futures by comparing designed flows to current and pre-dam (natural-flow) regimes,” and they hypothesize that “Flow variation (high and low) may drive production by controlling redox conditions in floodplain soils, …” This im- plies that they intend their “good” flow regime Fig. 1. Illustration of designed flows, including a “good” design and a “bad” design, compared Retired consultant. Email: jgwill@ to the reconstructed historical flow regime. [Copied from Sabo et al. (1)] Williams, Science 361, eaat1225 (2018) 13 July 2018 1 of 2 RESEARCH | TECHNICAL COMMENT Fig. 2. Stage of the Mekong (17 years); the analysis was not specified in ad- River at Stung Treng (solid vance (5); other important factors such as habitat line) and the Great Lake cover, sedimentation, and net primary produc- at Kampong Luong (dashed tion (6) were ignored; and it seems improbable line) for two flood cycles. that one drastic change in the flow regime would The flood pulse in the Great be good for the fishery whereas a modestly dif- Lake lags that at Stung Treng, ferent drastic change would be bad (7). Intuitively, and high-frequency variation variation in stage in the Great Lake seems more is suppressed. [Data from likely to influence catch in nets targeting fishes Sameng Preap, Mekong River migrating out of the lake than does variation in Commission] stage in the Mekong River hundreds of kilometers upstream. Despite the sophistication of the methods used to develop empirical relationships, predictions based on them depend on the assumption of cau- sality and the assumption that all else will remain more or less equal, which clearly will not be true if the proposed dams are built and the hydrograph on the opposite side from the Tonle Sap confluence hydrograph to ~15 days would truncate the pe- is radically altered. Complex computer-based in 2002 (4), so flows between the channel and riod when the nets might be used by a factor of analyses can be marvelously powerful tools, but floodplain are substantial. Thus, the proposed rect- about 11. Until the channel enlarged in response they should be used carefully, and it is always angular wave would degrade to something more to the higher flows, the flow probably would be necessary to think clearly about the results. D o like the “bad” flow regime as it moved downstream. too swift to use the fine mesh nets needed to w n l o Regarding the Tonle Sap, the average annual catch the small fishes that make up the bulk of REFERENCES 3 a outflow from the Great Lake exceeds 75 km (3). the Dai fishery catch (2), even if it were possible 1. J. L. Sabo et al., Science 358, eaao1053 (2017). d e Draining this volume in 15 days would require a to anchor the nets. 2. A. S. Halls, B. R. Paxton, N. Hall, N. Peng Bun, L. Lieng, d f 3 −1 N. Pengby, N. So, “The Stationary Trawl (Dai) Fishery of the r constant discharge of ~58,000 m s , which is Sabo et al. “hypothesized that high fisheries o m Tonle Sap-Great Lake, Cambodia,” MRC Technical Paper No. 32 almost 6 times the high flow reported in (2) and yields are driven by measurable attributes of hy- (Mekong River Commission, Phnom Penh, Cambodia, 2013). greater than current peak flows in the Mekong. drologic variability, and that these relationships 3. M. Kummu, J. Sarkkula, Ambio 37, 185–192 (2008). This discharge could not plausibly be carried by can be used to design and implement future flow 4. H. Fujii et al., Int. J. River Basin Manage. 1, 253–266 (2003). the existing Tonle Sap channel. regimes that improve fisheries yield through 5. A. Gelman, E. Loken, The Garden of Forking Paths: The Dai fishery uses what are essentially an- control of impending hydropower operations.” Why Multiple Comparisons Can Be a Problem Even When There Is No “Fishing Expedition” or “p-Hacking ” and the chored trawls, with mouths facing upstream, that Indeed, the size of the flood pulse in the Great Research Hypothesis Was Posited Ahead of Time (Department of depend on current to function. The nets target Lake is a good predictor of the Dai fishery catch Statistics, Columbia University, 2013); fishes migrating out of the Great Lake to the (2). Sabo et al. found associations between several ~gelman/research/unpublished/p_hacking.pdf. Mekong River with the falling limb of the hydro- other attributes of variability in stage at Stung 6. M. E. Arias et al., Ecol. Modell. 272, 252–263 (2014). 7. J. P. Ioannidis, PLOS Med. 2, e124 (2005). graph, as the Great Lake drains over 5+ months Treng and the catch, but whether any are “hy- (2). If somehow the “good” flow regime could be drologic drivers” of catch remains an open ques- 25 January 2018; accepted 25 May 2018 implemented, shortening the falling limb of the tion because the flow-catch records are short 10.1126/science.aat1225 o n J u l y 1 5 , 2 0 1 8 Williams, Science 361, eaat1225 (2018) 13 July 2018 2 of 2 RESEARCH ◥ The “Good” design flow of (1) features large, TECHNICAL COMMENT almost rectangular, flood pulses, punctuated by long dry seasons, and very rapid transitions— approximately 15 days, versus 4 to 5 months under SUSTAINABILITY natural conditions. Power production is the pri- mary function of the Mekong dams, which are Comment on “Designing river flows to designed and financed according to predefined operating procedures. Existing (and planned) dams generate downstream flows that are the improve food security futures in the inverse of the “Good” design flow (i.e., diminished flood pulse and elevated dry-season flow). The Lower Mekong Basin” “Good” flow would compromise the economic viability of dams in the LMB (11). We estimate that Cambodia alone would not have necessary 1 2 3 Ashley S. Halls and Peter B. Moyle * storage (approximately 28 km ) to generate the “Good” design flow regime (Table 1). Creating The designer flow regime proposed by Sabo et al. (Research Articles, 8 December 2017, the “Good” design flow would require coordina- p. 1270) to support fisheries in the Lower Mekong Basin fails to account for important tion of dam operations by all six countries in the ecological, political, and economic dimensions. In doing so, they indicate that dam Mekong Basin, only three of which have a major impacts can be easily mitigated. Such an action would serve to increase risks to food stake in the Lower Mekong River fisheries (Table 1). and livelihood futures in the basin. Even if the flow could somehow be generated, the designer flow regime is an extrapolation of D o abo et al. (1) propose that dams in the Lower Sabo et al. also found this flood index to be a an empirical model of a pattern that has never w n l Mekong Basin (LMB) could produce an significant driver of CPUE, together with a com- been observed in the LMB. The rapid changes in o a artificial (“designer”) hydrograph to mitigate posite index of dry-season anomalies. However, flow between dry and wet seasons threaten fish d e impacts on downstream fisheries that feed the authors appear to have disregarded the MRC populations upon which the Dai and other fisheries d f r Smillions of people. Modeling studies that findings (5): “…previous work suggests two hypo- in the LMB depend (12) by removing migration o m predict responses of tropical river fisheries to theses to explain how dams diminish production cues, hampering upstream spawning migrations, hydrological variation are not new, even in the of inland fisheries: (i) by reducing connectivity and sweeping downstream-drifting eggs and larvae LMB (2–4). A study similar to (1) was recently and dispersal, and (ii) by reducing primary pro- past suitable nursery habitats. Stranding and undertaken by the Mekong River Commission ductivity through the entrapment of sediment diminished reproductive success under rapidly (MRC) (5) using a shorter segment of the same supply and associated nutrient delivery from changing water levels will further diminish fish CPUE (catch per unit of effort) time series for the headwaters to downstream nursery habitats” (1). catch (13). Other flow-dependent sectors of the Cambodian Dai fishery. Hydrology was measured Sabo et al. explain that their positive relation- economy, such as flood recession agriculture (rice in the Tonle Sap River where the fishery takes ship between CPUE and FPExt “…may in part be farming) and navigation, would also be affected place, not several hundred kilometers away at related to increased catchability of fish at high by these new flow regimes. Stung Treng as in (1). water levels.” However, catchability declines in This MRC study showed that a flood index, large floods because fish density decreases. There- REFERENCES termed “FPExt” (flood pulse extent) (1), is the fore, fishermen in tropical rivers typically fish 1. J. L. Sabo et al., Science 358, eaao1053 (2017). o primary driver of fisheries yields; fish growth during the receding flood and dry seasons rather 2. A. S. Halls, R. L. Welcomme, River Res. Appl. 20, 985–1000 n (2004). J increases exponentially with the flood index, than peak flood (7). u l y 3. A. S. Halls, G. P. Kirkwood, A. I. Payne, Ecohydrol. Hydrobiol. 1, 1 likely reflecting improved feeding opportunities The model of Sabo et al. also does not take into 323–339 (2001). 5 (6). The response was modeled “allowing pre- account other serious impacts of dams besides 4. R. L. Welcomme, D. Hagborg, Environ. Biol. Fishes 2, 7–24 (1977). , 5. A. S. Halls et al., “The Stationary Trawl (Dai) Fishery of the 2 dictions to be made of how…fishery yield is like- modifying downstream flows, such as creating 0 Tonle Sap-Great Lake, Cambodia,” MRC Technical Paper No. 32 1 ly to vary under different flooding conditions, barriers to fish migrations. Passage through dams (Mekong River Commission, Vientiane, Lao PDR, 2013). 8 whether natural, or modified [by] climate change increases mortality and diminishes reproductive 6. R. L. Welcomme, River Fisheries (FAO Fisheries Technical and/or water management projects [dams]” (5). success in fish populations. Dams trap nutrient- Paper No. 262, Food and Agriculture Organisation of the United Nations, 1985). bearing sediments important for maintaining 7. D. D. Hoggarth et al., Management Guidelines for Asian Floodplain primary production and fish habitat, thereby River Fisheries (FAO Fisheries Technical Paper No. 384, Food and 1Aquae Sulis Research 2 ., Midway House, Turleigh, diminishing water quality (8, 9). The barrier Agriculture Organisation of the United Nations, 1999). Wiltshire BA15 2LR, UK. Department of Wildlife, Fish, and 8. G. Marmulla, Dams, Fish and Fisheries (FAO Fisheries Technical effects of dams alone could end important fish Conservation Biology, Center for Watershed Sciences, Paper No. 419, Food and Agriculture Organisation of the United University of California, Davis, CA 95616, USA. migrations in the LMB, irrespective of the avail- Nations, 2001). *Corresponding author. Email: pbmoyle@ ability of designer flows (10). 9. G. M. Kondolf et al., Sci. Total Environ. 625, 114–134 (2018). Table 1. Estimates of live storage in existing and planned dams in the LMB and estimated annual fish catch in each LMB country. Estimates are 3 from (14) and (15), respectively. The necessary live storage to generate the “Good” design flow proposed by Sabo et al. (1) is approximately 28 km . Country Live storage (km3) Proportion of total live storage Total annual catch (tonnes) Proportion of total annual catch Lao PDR 57.5 68% 208,503 8% Cambodia 18.9 22% 481,537 20% Thailand 5.4 6% 911,485 37% Vietnam 3.2 4% 852,823 35% Total 85.0 100% 2,454,348 100% Halls and Moyle, Science 361, eaat1989 (2018) 13 July 2018 1 of 2 RESEARCH | TECHNICAL COMMENT 10. A. S. Halls, M. Kshatriya, “Modelling the Cumulative Barrier and 12. A. F. Poulsen et al., “Fish Migrations of the Lower Mekong River PDR, 2009); Passage Effects of Mainstream Hydropower Dams on Migratory Basin: Implications for Development, Planning and BDP/BDP2-Regional-Hydropower-Sector-Review-5-Mar-09.pdf. Fish Populations in the Lower Mekong Basin” (Mekong River Environmental Management” (Mekong River Commission, 15. K. G. Hortle, “Consumption and Yield of Fish and Other Aquatic Commission, Vientiane, Lao PDR, 2009). Phnom Penh, Cambodia, 2006). Animals from the Lower Mekong Basin” (Mekong River 11. MRC, “Assessment of Basin-wide Development Scenarios: 13. R. L. Welcomme, A. S. Halls, Ecohydrol. Hydrobiol. 1, 313–321 Commission, Vientiane, Lao PDR, 2007). Main Report. Basin Development Plan Programme, Phase 2” (2001). (Mekong River Commission, Vientiane, Lao PDR, 2011); 14. C. Yermoli, “MRC Hydropwer Database. Hydropower Sector Review for the Joint Planning Process. Basin Development Plan 2 February 2018; accepted 25 May 2018 BDP-Assessment-of-Basin-wide-Dev-Scenarios-2011.pdf. Programme, Phase 2” (Mekong River Commission, Vientiane, Lao 10.1126/science.aat1989 D o w n l o a d e d f r o m o n J u l y 1 5 , 2 0 1 8 Halls and Moyle, Science 361, eaat1989 (2018) 13 July 2018 2 of 2 RESEARCH ◥ ing that the production of organic matter and TECHNICAL RESPONSE nutrients in floodplains during the dry season supports the exceptionally high productivity of flood-pulse ecosystems observed worldwide (4–6). SUSTAINABILITY Governing mechanisms and the magnitude of nutrient and organic transfer from floodplains to Response to Comments on “Designing fishes remain active areas of research, but current information suggests that their importance to fishery productivity is substantial [e.g., (7–9)]. river flows to improve food security The multivariate autoregressive modeling ap- proach used by Sabo et al. reveals a statistical futures in the Lower Mekong Basin” relationship between multiple aspects of hydro- logic variance at the Stung Treng gauging station— namely, water height at um flood and 1 2 3 4 5 6 G. W. Holtgrieve *, M. E. Arias , A. Ruhi , V. Elliott , So Nam , Peng Bun Ngor , duration of the low-flow period—and annual fish 7 8,9,10 T. A. Räsänen , J. L. Sabo catches at the Dai fishery on the Tonle Sap River. Stung Treng was chosen because it had the longest Sabo et al. presented an empirically derived algorithm defining the socioecological period of record, which was important for quan- response of the Tonle Sap Dai fishery in the Cambodian Mekong to basin-scale variation in tifying pre-dam hydrologic conditions. The gaug- hydrologic flow regime. Williams suggests that the analysis leading to the algorithm is ing station in the Tonle Sap River at Prek Kdam flawed because of the large distance between the gauge used to measure water levels is the closest to the fishery, but the water level (hydrology) and the site of harvest for the fishery. Halls and Moyle argue that Sabo et al.’s record was too short to carry out the full analysis. D o findings are well known and contend that the algorithm is not a comp ensive assessment The distance between the gauge station and the w n l of sustainability. We argue that Williams’ critique stems from a misunderstanding about fishery is irrelevant, however, as long as the hy- o a our analysis; further clarification of the analysis is provided. We regret not citing more of the drology at the reference location and at the fish- d e work indicated by Halls and Moyle, yet we note that our empirical analysis provides ery location are consistently related to each other. d f r additional new insights into Mekong flow-fishery relationships. The two records need not be identical in shape or o m magnitude in order to be valid for analysis and forecasting, as systematic differences in hydrology abo et al. (1) presented an empirical anal- of the impacts of dam development on capture between the locations are inherently captured in ysis that mathematically defines the socio- fisheries in the Mekong River system. the empirical analysis. ecological response of an ecosystem (fish In his Comment, Williams (2) contends that The data presented in figure 1 of Williams (2) catch) to an increasingly human-controlled this empirical analysis is flawed because of the are from the Tonle Sap Lake itself at Kampong Sphysical ecosystem driver worldwide (hydrol- distance between the river gauging station used in Luong, yet despite the issues inherent in attempt- ogy). This analysis yielded three critical findings: the analysis and the location of the fishery, and ing to compa eight of water in a lake basin (i) The timing of river discharge, as much as the that the “good design” flow scenario presented to height in a flowing river for only 2 years, the magnitude, controls fish catches in the Tonle Sap in Sabo et al. is thus unable to be implemented. data show a high degree of systematic coherence. River Dai fishery; (ii) the two aspects of the hy- Both questions are addressed below, which we The more appropriate comparison between the drology most closely related to fish catch—water believe arise from Williams’ misunderstanding Mekong stage at Stung Treng and the Tonle Sap o height during the flood and duration of the low- of the analysis presented in Sabo et al. In a second River stage at Prek Kdam is given in Fig. 1 here. We n J flow period—have been changing with existing Comment, Halls and Moyle (3) note that previ- have also extended the comparison to a 27-year u l y 1 dam development in the direction opposite to ous research by Halls and others has identified period to better represent interannual variability. 5 what promotes fish catches; and (iii) a “designed” flood magnitude and duration as important to This figure shows two important aspects of the sys- , 2 hydrograph has the potential to maintain or even Dai fishery catches, which were not cited by tem hydrology: (i) the strong coherence between 0 1 improve fish catches without the constraint of Sabo et al. They also question the use of the Stung the two records (Fig. 1A), and (ii) the nonlinear 8 maintaining a completely natural hydrologic re- Treng gauging station on the Mekong River for relationship on the falling limb of the hydrograph, gime. Together, these results provide a new per- our analysis as opposed to the more proximate with stage decreasing faster in the Mekong than spective on managing flow regimes in heavily Prek Kdam gauging station on the Tonle Sap at Prek Kdam (Fig. 1B). The former validates the developed rivers and thus provide directions for River. Lastly, Halls and Moyle raise concerns use of Stung Treng as the reference station in our future research, while simultaneously presenting about implementation of a designed-flow regime analysis. The latter indicates that because of a potentially viable approach for mitigating some and messaging to regional policy-makers that landscape morphometry, declining river stage at designed-flow regimes are a complete solution to Stung Treng will precede and be more rapid than 1 environmental impacts arising from dam develop- declines in stage in the Tonle Sap River. Because School of Aquatic and Fishery Sciences, University of Washington, Seattle, WA 98105, USA. 2Department of Civil and ment in the Mekong. We appreciate the contri- this effect is inherently captured in our analysis, Environmental Engineering, University of South Florida, Tampa, butions of Williams (2) and Halls and Moyle (3) it renders the calculation by Williams moot, as the FL 33620, USA. 3Department of Environmental Science, Policy, and the opportunity to discuss these issues. “good design” presented in Sabo et al. is for the and Management, University of California, Berkeley, CA 94720, It is indeed unfortunate that Sabo et al. did not hydrograph at Stung Treng, not at the Tonle Sap USA. 4Betty and Gordon Moore Center for Science, Conservation International, Arlington, VA 22202, USA. 5Mekong River cite the full complement of research showing the River, and it is wrong to assume that the two Commission Secretariat, Ban Sithane Neua, Sikhottabong relationship between fishery catches and flood- stations should have identical hydrographs. In District, Vientiane 01000, Lao PDR. 6Fisheries Administration, pulse magnitude and duration, as we agree that short, Williams has made an incorrect assump- No. 186, Preah Norodom Boulevard, Khan Chamcar Morn, 7 there is substantial literature in support of this tion that our “good design” hydrograph is to be P.O. Box 582, Phnom Penh, Cambodia. Consultant, Rinkelitie 3A2, P.O. Box 02580, Siuntio, Finland. 8Futu 2O, Knowledge result. Beyond this, however, Sabo et al. also dis- applied at any location in the basin; it is applica- Enterprise Development, Arizona State University, Tempe, AZ cerned that the magnitude and duration of the ble at Stung Treng only, with the response of fish 85287, USA. 9Julie Ann Wrigley Global Institute of Sustainability, low-flow period is equally as important to fish- catch applicable at the Dai fishery only. Arizona State University, Tempe, AZ 85287, USA. 10Faculty of eries as the high-flow period. This more novel Any meaningful assessment of the implemen- Ecology, Evolution and Environmental Science, School of Life Sciences, Arizona State University, Tempe, AZ 85287, USA. and unexpected result is nonetheless consistent tation of designed-flow scenarios in the Mekong *Corresponding author. Email: gholt@ with other research and general theories suggest- will involve a complex modeling exercise that Holtgrieve et al., Science 361, eaat1477 (2018) 13 July 2018 1 of 2 RESEARCH | TECHNICAL RESPONSE suite of futures to local policy-makers to better highlight critical cross-sector interdependencies. Finally, as rapid economic development pro- ceeds in the Lower Mekong Basin, an altered flow regime is one of many ecosystem changes likely to have an impact on fish production and fish- eries. Sabo et al.’s focus on flow regime does not imply that other potential impacts—for example, barriers to migration, reduced sediment supply, and eutrophication—are unimportant or need not be considered. The potential for unintended ecosystem consequences of a non-natural flow regime must also be further researched and fully considered in the decision-making process. Designing flow regimes is one area where there is scope for management action, however, and Sabo et al. provide a method to robustly quantify the relationship between flow regime and fish- eries catches in a way that managers can use. With this, there is now means for formal trade- off evaluation useful to policy-makers and a poten- tial path toward mitigating some of the impacts D o that dam development will have on lower Mekong w n l fisheries. Exactly how designed flow regimes o a might be implemented in the real world will d e require substantial additional research and d f r stakeholder engagement. o m REFERENCES AND NOTES 1. J. L. Sabo et al., Science 358, eaao1053 (2017). 2. J. G. Williams, Science 361, eaat1225 (2018). 3. A. S. Halls, P. B. Moyle, Science 361, eaat1989 (2018). 4. P. B. Bayley, Bioscience 45, 153–158 (1995). 5. J. W. Junk, P. B. Bayley, R. E. Sparks, Can. Spec. Publ. Fish. Aquat. Sci. 106, 110–127 (1989). 6. R. L. Welcomme, Inland Fisheries Ecology and Management Fig. 1. Hydrograph coherence and nonlinearity in the Lower Mekong Basin. (A) River stage (Fishing News Books, 2001). at Stung Treng on the Mekong River and Prek Kdam on the Tonle Sap River from 2 January 1986 7. G. W. Holtgrieve et al., PLOS ONE 8, e71395 (2013). to 31 December 2008. The correlation coefficient (Pearson r) of log10-transformed stage is 8. D. Lamberts, J. Koponen, Ambio 37, 178–184 (2008). 9. S. B. Correa, K. Winemiller, Oecologia 186, 1069–1078 (2018). 0.89. (B) The same data plotted against each other, showing a linear relationship in water height between the two stations on the rising limb of the hydrograph but a nonlinear relationship on o ACKNOWLEDGMENTS the falling limb. n Supported by MacArthur Foundation award 100717 and NSF J grant DEB-1541694 (G.W.H.); NSF grant CBET-1204478 (J.L.S.); u l y NSF grant EAR-1740042 (J.L.S. and G.W.H.); the National Socio- 1 considers, at least, movement of water across the gation, biodiversity protection, and electricity Environmental Synthesis Center (SESYNC) with funding from NSF 5 landscape and through time, the number and production—must also be taken into account , 2 grant DBI-1052875 (A.R.); and MacArthur Foundation awards 100718 placement of dams throughout the watershed, when considering how river hydrology might 0 1 and 15-108455-000-CSD to Conservation International. All authors operation rules for electricity production, coor- be managed in the future. This was obviously thank the Mekong River Commission and the Inland Fisheries 8 dination (or lack thereof) of dam operations, and beyond the scope of Sabo et al., and continued Research and Development Institute (IFReDI) of Cambodia for access to data. future climate. Similarly, as Halls and Moyle (3) future research is needed to provide a more com- note, multiple other societal objectives beyond p ensive multisectoral analysis. The goal of 14 February 2018; accepted 25 May 2018 fisheries—such as rice production, flood miti- such an analysis should be to present a broader 10.1126/science.aat1477 Holtgrieve et al., Science 361, eaat1477 (2018) 13 July 2018 2 of 2 Modeling tissues with synthetic biology RESEARCH Toda et al., p. 156 I N SCIENCE J O U R NA L S Edited by Stella Hurtley PLANT SCIENCE How rice defeats the floodwaters eepwater rice varieties grow taller when flooded, in a growth response driven by the plant hormones gibberellin and ethyl- Dene. This keeps the leaves above the water. Kuroha et al. identified the genes underlying this phenotype, which encode a component of the gibberellin biosynthetic pathway and its regulatory ethylene-responsive transcription factor. This genetic relay drives growth of the plant stem internodes in response to flooding. Modern cultivated deepwater rice, which has been domesticated for adaptation to the monsoon season of Bangladesh, emerged from the genetic variation found in wild rice strains over a broader geographic region. —PJH A farmer tends flooded deepwater rice fields in Bangladesh. Science, this issue p. 181 PHYSICS simulating problems in materi- by using deformation from the suggest that this discrepancy Simulating correlated als physics. —ISO nuclear test and the subsequent exists because current inven- Science, this issue p. 162 collapse of Mount Mantap. The tory methods miss emissions electron systems test occurred at a depth of that occur during abnormal Correlated electron systems are about half a kilometer, with an operating conditions. These data, generally difficult to simulate GEODETIC MONITORING explosive yield around 10 times and the methodology used to because of limited capabilities Nuclear testing under that of the Hiroshima explosion. obtain them, could improve and S of computational resources. —BG verify international inventories of E G A Harris et al. used a D-Wave the radar Science, this issue p. 166 greenhouse gases and provide a M I Y T chip based on a large array North Korea conducted its sixth better understanding of mitiga- T E G of superconducting elements underground nuclear weapons tion efforts outlined by the Paris / N GREENHOUSE GASES O to simulate the phases of a test in September 2017. The Agreement. —HJS S N A ’ I D complex magnetic system. They seismic waves generated from A leaky endeavor Science, this issue p. 186 R A tuned the amount of frustration the test allow for triangulation Considerable amounts of the H C I R within the lattice and varied the and explosive yield estimates. greenhouse gas methane leak ; . L A T effective transverse magnetic However, Wang et al. show that from the U.S. oil and natural EVOLUTIONARY COGNITION E A field, which revealed phase tran- synthetic aperture radar (SAR) gas supply chain. Alvarez et al. D The impact of time O T S. sitions between a paramagnetic, should be added to the arsenal reassessed the magnitude of ) P an ordered antiferromagnetic, of techniques used to detect this leakage and found that in wasted O T M and a spin-glass phase. The and characterize nuclear tests. 2015, supply chain emissions The amount of time already O R F ( results compare well to theory SAR tracks deformation from were ~60% higher than the spent on a task influences : S T I for this spin-glass problem, space, which resulted in a better U.S. Environmental Protection human choice about whether D E R validating the approach for constraint of source parameters Agency inventory estimate. They to continue. This dedicated C SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 141 Published by AAAS RESEARCH | IN SCIENCE JOURNALS time, known as the “sunk cost,” thus offers a potential mecha- reduces the likelihood of giving nistic explanation for division up the pursuit of a reward, even errors frequently observed in IN OTHER JOURNALS Edited by Caroline Ash and Jesse Smith when there is no indication human embryos in the fertility of likely success. Sweis et al. clinic. —SMH show that this sensitivity to Science, this issue p. 189; time invested occurs similarly see also p. 128 in mice, rats, and humans (see the Perspective by Brosnan). All three display a resistance HEPATITIS C VIRUS to giving up their pursuit of a Hampering HCV reward in a foraging context, but only after they have made the transmission decision to pursue the reward. No hepatitis C virus (HCV) —SNV vaccine is currently available. Science, this issue p. 178; Furthermore, evidence from see also p. 124 studies in nonhuman primates suggests that any futu uman HCV vaccine would be unlikely EARLY DEVELOPMENT to induce complete immunity. It takes two to tango Major et al. examined whether Fusion of egg and sperm lowered HCV titers potentially combines the genetic material resulting from an imperfect of both parents in one cell. In vaccine might still stem HCV mammals, including humans, transmission in people who each parental genome is initially inject drugs. They measured confined in a separate pro- the HCV RNA from infected nucleus. For the new organism human plasma retained in con- to develop, the two genomes taminated needles and syringes. must be spatially coordinated Their mathematical model so that the first embryonic divi- combining these measurements sion can create two cells that with published data on HCV combine both genomes in one viral kinetics suggested that a nucleus. Reichmann et al. found partially effective vaccine could EMERGING THERAPIES untreated mice. —PAK that at the beginning of the reduce the HCV transmission Gene editing gets a Nat. Commun. 10.1038/s41467-018- first division, two microtubule risk from sharing contaminated 04894-2 (2018). spindles organize the maternal needles and syringes. —CAC head start and paternal chromosomes and Sci. Transl. Med. 10, eaao4496 (2018). The development of gene- CANCER subsequently align to segre- editing technologies into gate the parental genomes in therapies for human disease Finding the precise drug parallel (see the Perspective by CELL BIOLOGY is an exciting prospect. A Treating cancer according E C R U Zielinska and Schuh). Failure crucial question is whether to the molecular makeup of O S An inflammatory way to E of spindle alignment led to there are advantages to tumor cells is challenging. C N E activate AMPK I two-celled embryos with more correcting disease-causing How can we find a target that C S / than one nucleus per cell. Dual- In its role as a sensor of energy mutations before rather than will kill tumor cells and match N G I S spindle assembly in the zygote status, the kinase AMPK is after birth, and whether this it to an effective drug? Alvarez E D S T activated by phosphorylation approach is even feasible. In et al . analyzed the protein- K ; L mediated by the tumor sup- a proof-of-concept study in signaling networks in human B M E / pressor LKB1. Antonia and mice, Ricciardi et al. accom- gastroenteropancreatic neuro- G R E Baldwin found that AMPK plished successful in utero endocrine tumors (GEP-NETs) B N E L activation could also be medi- correction of a hemoglobin to find master regulators of L E / ated by a pathway involving gene mutation that causes tumor survival. Subsequent N N A TAK1, a kinase associated with b-thalassemia, a serious screening of GEP-NET–derived M H C I inflammatory pathways, and its blood disorder. They injected cells with various drugs E R / N target, IKK. IKK phosphorylated nanoparticles containing confirmed that their top hit, I A S AMPK independently of LKB1. gene-editing machinery (tri- entinostat, inhibited GEP-NET S O H / Combining an IKK inhibitor with plex-forming peptide nucleic growth in vivo. This approach A V O I the cancer drug phenformin acids and single-stranded has the potential to provide S A T improved the drug’s ability to donor DNA) intravenously unbiased precision therapy for R A ) T kill LKB1-deficient cancer cells, into midgestational mouse many cancers, especially rare F E L highlighting a potential treat- fetuses. After birth, the cancers, such as GEP-NETs. M O R F ment for cancers lacking this treated mice showed sus- —GKA ( : S T Dual spindles separate maternal and tumor suppressor. —WW tained amelioration of anemia Nat. Genet. 10.1038/s41588-018- I D E Sci. Signal. 11, eaan5850 (2018). R paternal chromosomes in parallel. and survived longer than 0138-4 (2018). C 142 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS occupy about 20% of a cell’s volume, mTORC1 activity tunes MICROBIOTA cytoplasmic crowding and Cell-to-cell action in cholera thus influences cell physiology. —LBR ut bacteria are competitive. They jostle Cell 10.1016/j.cell.2018.05.042 for niches and nutrients by means of a (2018). variety of mechanisms. Gram-negative bacteria devote a lot of resources Gto the comparatively large (up to CARBON CAPTURE 600 nm) type 6 secretion system (T6SS) Help wanted used to stab neighboring bacterial and Preventing global surface air host cells. Contrary to expectations, Fast temperatures from rising too et al. discovered that commensal bacteria high will require emissions to be reduced and CO to be are not always protective against T6SS- 2 toting pathogens. Vibrio cholerae T6SS removed (captured) from contributes to host death in the fly model the air. Many ways to do this of cholera. Acetobacter pasteurianus is a have been proposed, although common, normally harmless, commensal none to date are both tech- in the fly gut. If A. pasteurianus is removed, nologically and economically surprisingly, V. cholera is enfeebled, too. feasible, so new and better This effect seems to be mediated by signal- strategies are needed. Rau et ing through the immune deficiency (IMD) al. suggest that generating H2 pathway of the fly. The IMD is “primed” by with a combination of saline the commensal organism. —CA water electrolysis and mineral Proc. Natl. Acad. Sci. U.S.A . 10.1073/ weathering, powered by elec- pnas.1802165115 (2018). tricity not derived from fossil fuels, could greatly increase energy generation and CO2 The pathogenicity of Vibrio cholerae (illustrated at left) removal, at a lower cost than is reduced by removal of a commensal bacterium. methods involving biomass energy plus carbon capture and storage. This approach QUANTUM OPTICS by the lack of single-photon couple them in an optic fiber. also would allow carbon to Single photons for sources operating at telecom The results demonstrate the be sequestered as long-lived wavelengths. Dibos et al. show basis of an architecture to ocean alkalinity rather than as optic fibers that single erbium ions doped develop long-distance quan- concentrated CO . —HJS 2 Being robust, fast, and able to in a solid-state host material tum optical networks. —ISO Nat. Clim. Change 8, 621 (2018) . encode information in a number produce single photons at just Phys. Rev. Lett. 120, 243601 (2018). of different ways, single photons the right wavelength for long- are ideal carriers of quantum distance transmission. They use THERMOELECTRICS CELL BIOLOGY information. Exploiting the vast a silicon-based nanophotonic Lower symmetry for optic fiber network to route crystal in close proximity to a mTORC1 jams cell traffic single photons between distant doped ion to enhance the extrac- The cytoplasm of a cell is higher performance points, however, is hampered tion of single photons and to stuffed with large molecules Thermoelectric materials inter- and organelles. Such crowding change heat and electricity. may cause some molecules Lowering thermal conductivity to aggregate, which could in while maintaining electrical ) turn affect their biological conductivity is important for 8 1 0 function. To monitor cytoplas- developing promising ther- 2 ( 1 0 mic crowding, Delarue et al. moelectric materials. Li et al. 6 3 4 2, made a probe from bacte- depart from the usual strategy 0 2 1. rial proteins that form large of using high-symmetry T T E scaffolds linked to fluorescent materials by moving from L . V E proteins. In yeast and human cubic to rhombohedral GeTe, R . S Y cells, the probes revealed that which ends up boosting the H P , L. the protein kinase com- thermoelectric performance. A T plex mTORC1 (mechanistic The performance metric called E S O target of rapamycin complex the figure of merit is 2.4 at 600 B I D M. 1) enhanced cytoplasmic K. The general strategy may be . A : viscosity. It did so by increas- applicable to other materials, N O I T ing production of ribosomes providing another pathway to A R T while inhibiting their degra- improve performance. —BG S U L L Schematic illustration of silicon waveguides and a YSO crystal dation. Because ribosomes Joule 2, 976 (2018) I SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 143 Published by AAAS RESEARCH ALSO IN SCIENCE JOURNALS Edited by Stella Hurtley PLANT SCIENCE SYNTHETIC BIOLOGY are fairly unreactive when they track parasite populations and Genomic traces of Engineering multilayered are not strained in a tight ring. their evolution in the context of Roque et al. now report that a control programs. They also dis- symbiosis loss cellular structures silver salt can cleave C–C bonds cuss how conservation genomics A symbiosis between cer- The ability to program the manu- in unstrained cyclic amines such can inform us about the dynam- tain bacteria and their plant facture of biological structures as pyrrolidines and piperidines. ics of extinction. —GKA hosts delivers fixed nitrogen may yield new biomaterials or Paired with an electrophilic Science, this issue p. 130 to the plants. Griesmann et synthetic tissues and organs. fluorine source in aqueous solu- al . sequenced several plant Toda et al. engineered mam- tion, the silver first oxidizes the a genomes to analyze why nitro- malian “sender” and “receiver” carbon adjacent to the nitrogen. HUMAN GENETICS gen-fixing symbiosis is irregularly cells with synthetic cell surface Ring-opening fluorination of the Methylation patterns in scattered through the evolution- ligands and receptors that b carbon then proceeds by an ary tree (see the Perspective by controlled gene regulatory apparent radical mechanism. early embryos Nagy). Various genomes carried circuits based on Notch signal- The reaction offers a versatile The methylation state of an traces of lost pathways that ing. Programming the cells to means of introducing fluorine embryo can be influenced could have supported nitrogen- express cell adhesion molecules to structural motifs common in by its environment, although fixing symbiosis. It seems that and other regulatory molecules pharmaceutical research. —JSY whether this affects a person’s this symbiosis, which relies on enabled spontaneous forma- Science, this issue p. 171 health later in life is unclear. multiple pathways and complex tion of multilayered structures, A mother’s diet can influence interorganismal signaling, is sus- like those that form during genomic regions that vary EARLY OCEAN ceptible to selection and prone embryonic development. The greatly in methylation state, to being lost over evolutionary three-layered structures even The rise of oxygen known as metastable epialleles time. —PJH showed regeneration after injury. To understand the evolution of (MEs). Kessler et al. performed Science, this issue p. 144; —LBR the biosphere, we need to know genome-wide screening on 687 see also p. 125 Science, this issue p. 156 how much oxygen was present candidate MEs in preimplanta- in Earth’s atmosphere during tion embryos of North American most of the past 2.5 billion years. Caucasians. Compared with ME FERROELECTRICITY NEUTRINO ASTROPHYSICS However, there are few proxies patterns previously observed Neutrino emission from a Perovskites go organic sensitive enough to quantify O2 in a Gambian population and in The perovskite structure at the low levels present until Chinese embryos, regions of the flaring blazar accommodates many differ- slightly less than 1 billion years embryonic genome that are sen- Neutrinos interact only very ent combinations of elements, ago. Lu et al. measured iodine/ sitive to season of conception weakly with matter, but giant making it attractive for use in calcium ratios in marine car- harbored many MEs and atypical detectors have succeeded in a wide variety of applications. bonates, which are a proxy for methylation patterns. Thus, MEs detecting small numbers of Building perovskites out of only dissolved oxygen concentrations are sensitive to genetic and envi- astrophysical neutrinos. Aside organic compounds is appeal- in the upper ocean. They found ronmental features. —PJB from a diffuse background, ing because these materials that a major, but temporary, rise Sci. Adv. 10.1126/sciadv.aat2624 only two individual sources tend to be flexible, fracture- in atmospheric O occurred at (2018). 2 have been identified: the Sun resistant, and potentially around 400 million years ago and a nearby supernova in easier to synthesize than their and that O levels underwent 2 1987. A multiteam collaboration inorganic counterparts. Ye et al. a step change to near-modern THYMUS detected a high-energy neutrino describe a previously unknown values around 200 million years Metabolic signaling event whose arrival direction family of all-organic perovskites, ago. —HJS was consistent with a known of which they synthesized 23 Science, this issue p. 174 circuits in thymocytes blazar—a type of quasar with a different family members (see Cell differentiation is often relativistic jet oriented directly the Perspective by Li and Ji). accompanied by metabolic INFECTIOUS DISEASES along our line of sight. The bla- The compounds are attractive changes. Yang et al. report that zar, TXS 0506+056, was found as ferroelectrics, including one Tracking parasite decline generation of double-positive to be undergoing a gamma-ray compound with properties close Neglected tropical diseases, (DP) thymocytes, which have flare, prompting an extensive to the well-known inorganic fer- such as malaria, are often both CD4 and CD8 antigens multiwavelength campaign. roelectric BaTiO . —BG caused by parasites, the on their surface, from double- 3 Motivated by this discovery, the Science, this issue p. 151; population dynamics of which negative (DN) thymocytes IceCube collaboration examined see also p. 132 are difficult to track. Control coincides with dynamic regula- lower-energy neutrinos detected programs that are aimed at tion of glycolytic and oxidative over the previous several years, eradicating these diseases would metabolism. Given the central finding an excess emission at ORGANIC CHEMISTRY benefit from an ability to monitor role of mTORC1 (mechanistic the location of the blazar. Thus, A silver cleaver splits parasite populations to assess target of rapamycin complex 1) blazars are a source of astro- efficacy and tailor responses. signaling in regulating metabolic physical neutrinos. —KTS cyclic amines In a Perspective, Cotton et al. changes, the authors examined Science, this issue p. 147, p. 146 Carbon-carbon single bonds discuss the use of genomics to the role of the mTORC1 pathway 143-B 13 JULY 2018 • VOL 361 ISSUE 6398 SCIENCE Published by AAAS DA_0713ISIO.indd 144 7/11/18 10:20 AM in thymocyte development. They conditionally deleted RAPTOR, a key component of the mTORC1 complex. Loss of RAPTOR impaired the DN-to-DP transition, but unexpectedly also perturbed the balance between ab and gd T cells. —AB Sci. Immunol. 3 , eaas9818 (2018). STRUCTURAL DYNAMICS Look fast Organisms from bacteria to humans sense and react to light. Proteins that contain the light-sensitive molecule retinal couple absorption of light to conformational changes that produce a signal or move ions across a membrane. Nogly et al . used an x-ray laser to probe the earliest structural changes to the retinal chromophore within microcrystals of the ion pump bacteriorhodopsin (see the Perspective by Moffat). The excited-state retinal wiggles but is held in place so that only one double bond of retinal is capable of isomerizing. A water molecule adjacent to the proton-pumping Schiff base responds to changes in charge distribution in the chromophore even before the movement of atoms begins. —MAF Science, this issue p. 145; see also p. 127 SCIENCE 13 JULY 2018 • VOL 361 ISSUE 6398 143-C Published by AAAS DA_0713ISIO.indd 145 7/11/18 10:20 AM RESEARCH ◥ RESULTS: We sequenced the genomes of sev- RESEARCH ARTICLE SUMMARY en nodulating species belonging to the Fagales, Rosales, and Cucurbitales orders and the legume subfamily Caesalpinioideae. We complemented PLANT SCIENCE this dataset by sequencing three genomes of non- nodulating species from the Cucurbitales and Phylogenomics reveals multiple losses from the legume subfamilies Cercidoideae and Papilionoideae. Using a genome-wide phyloge- of nitrogen-fixing root nodule symbiosis nomic approach, we found that all legume genes with a characterized role ◥ ON OUR WEBSITE in NFN symbiosis are con- ilian Griesmann, Yue Chang, Xin Liu, Yue Song, Georg Haberer, Read the full article served in nodulating spe- Matthew B. Crook, Benjamin Billault-Penneteau, Dominique Lauressergues, at . cies with one exception. We Jean Keller, Leandro Imanishi, Yuda Purwana Roswanjaya, Wouter Kohlen, org/10.1126/ observed larger numbers Petar Pujic, Kai Battenberg, Nicole Alloisio, Yuhu Liang, Henk Hilhorst, science.aat1743 of order-specific gene fam- Marco G. Salgado, Valerie Hocher, Hassen Gherbi, Sergio Svistoonoff, Jeff J. Doyle, ily expansions that, solely Shixu He, Yan Xu, Shanyun Xu, Jing Qu, Qiang Gao, Xiaodong Fang, Yuan Fu, because of their phylogenetic distribution, may Philippe Normand, Alison M. Berry, Luis G. Wall, Jean-Michel Ané, include genes contributing to multiple gains or Katharina Pawlowski, Xun Xu, Huanming Yang, Manuel Spannagl, Klaus F. X. Mayer, subsequent refinements of the symbiosis. In Gane Ka-Shu Wong, Martin Parniske*, Pierre-Marc Delaux*, Shifeng Cheng* parallel, we discovered signatures of multiple independent loss-of-function events for the gene D o encoding the indispensable NFN symbiosis w n INTRODUCTION: Access to nutrients such as phyletic NFN clade. However, only 10 of the l nitrogen is required for plant growth. Legumes 28 plant families within this clade contain regulator NODULE INCEPTION (NIN) in 10 of o a 13 genomes of nonnodulating species within d and nine additional plant families benefit from species engaged in the NFN symbiosis. Even e d the NFN clade. The pattern suggests at least the nitrogen-fixing root nodule (NFN) symbio- within these 10 families, most genera do not f r o m sis, in which roots develop nodules that intra- form this symbiosis. The NFN symbiosis re- eight independent losses of NFN symbiosis. cellularly host nitrogen-fixing bacteria. In this quires the coordinated function of more than CONCLUSION: We found that multiple inde- mutually beneficial symbiosis, the bacteria con- 30 essential genes. Presence of this symbiosis pendent losses of NFN symbiosis occurred in vert atmospheric nitrogen into ammonium and in related families suggests that a genetic change the four orders of the NFN clade. These results deliver it to the host plant. NFN symbiosis thus in the ancestor of the NFN clade enabled evo- suggest that NFN symbiosis has previously been enables plant survival under nitrogen-limiting lution of NFN symbiosis in this clade. The more common than currently evident and that conditions in terrestrial ecosystems. In agricul- scattered distribution of functional NFN sym- this symbiosis is subject to an underestimated ture, this symbiosis reduces reliance on nitro- biosis across the clade has led to the question adverse selection pressure. ▪ gen fertilizer, thus reducing the costs, ecological of whether NFN symbiosis evolved multiple impact, and fossil fuel consumption attendant times independently in a convergent manner on large-scale application of fertilizers. or was lost multiple times regardless of the number of times it arose. Fossil data have o The list of author affiliations is available in the full article online. RATIONALE: Molecular phylogenies show been unable to answer this question. Here *Corresponding author. Email: chengshf@ (S.C.); n J that NFN symbiosis is restricted to four angio- we used molecular evidence to ask how the pierre-marc.delaux@lrsv.ups-tlse.fr (P.-M.D.); parniske@lmu.de u l (M.P.) y sperm orders—Fabales, Fagales, Cucurbitales, current pattern of plant species with NFN sym- Cite this article as M. Griesmann et al., Science 361, eaat1743 1 5 , and Rosales—that together form the mono- biosis evolved. (2018). DOI: 10.1126/science.aat1743 2 0 1 8 Phylogenomics and evolution of NFN symbiosis. Genome sequencing of nodulating and nonnodulating species combined with 27 previously available genomes resulted in a dataset spanning the NFN clade and species outside the NFN clade as an outgroup. Orthogroups were identified and filtered following three phylogenetic patterns.This genome-wide analysis identified two genes involved in NFN symbiosis, NIN and RHIZOBIUM-DIRECTED POLAR GROWTH (RPG), that were lost in most nonnodulating species.The occurrence of multiple losses (red crosses) of NFN symbiosis suggests an adverse selection pressure. Griesmann et al., Science 361, 144 (2018) 13 July 2018 1 of 1 RESEARCH ◥ that form nodules (referred to here as “nodulating RESEARCH ARTICLE species”) and these do not form a monophyletic group; moreover, within 9 of these 10 families, most genera do not form NFN symbiosis (7). In PLANT SCIENCE addition to this scattered distribution, a further unsolved mystery that surrounds the evolution Phylogenomics reveals multiple of NFN symbiosis is its diversity at multiple levels: Legumes (Fabales) and the nonlegume Parasponia (Rosales) (8) form nodules with rhizobia, whereas losses of nitrogen-fixing root species of the actinobacterial genus Frankia in- fect actinorhizal plants from eight plant families nodule symbiosis of the orders Fagales, Cucurbitales, and Rosales (9). A diversity of infection mechanisms has been described (9), and root nodule structures display 1,2 3,4 3,4 3,4 2 ilian Griesmann , Yue Chang , Xin Liu , Yue Song , Georg Haberer , wide variations (5, 7). The most parsimonious 5 1 6 hypothesis to explain the restricted and yet Matthew B. Crook , Benjamin Billault-Penneteau , Dominique Lauressergues , 6 7 8 8 scattered distribution pattern of such diverse Jean Keller , Leandro Imanishi , Yuda Purwana Roswanjaya , Wouter Kohlen , 9 10 9 3,4 11 NFN symbioses predicted a genetic change in Petar Pujic , Kai Battenberg , Nicole Alloisio , Yuhu Liang , Henk Hilhorst , Marco G. Salgado12, Valerie Hocher13, Hassen Gherbi13, Sergio Svistoonoff13, the ancestor of the NFN clade, a predisposition Jeff J. Doyle14, Shixu He3,4, Yan Xu3,4, Shanyun Xu3,4, Jing Qu3,4, Qiang Gao3,15, event, that enabled the subsequent independent 3,15 3,4 9 10 7 evolution of NFN symbiosis specifically and ex- Xiaodong Fang , Yuan Fu , Philippe Normand , Alison M. Berry , Luis G. Wall , 16,17 12 3,4 3,18 clusively in this clade (the multiple-origins hypo- D Jean-Michel Ané , Katharina Pawlowski , Xun Xu , Huanming Yang , o 2 2,19 3,20,21 1 thesis), along with a number of losses (5, 7, 10–12). w Manuel Spannagl , Klaus F. X. Mayer , Gane Ka-Shu Wong , Martin Parniske *, n l 6 3,4 Recent quantitative phylogenetic modeling studies o Pierre-Marc Delaux *, Shifeng Cheng * a supported scenarios with independent gains and d e switches between the nonnodulating and nod- d The root nodule symbiosis of plants with nitrogen-fixing bacteria affects global nitrogen f r ulating states during the evolution of the NFN o cycles and food production but is restricted to a subset of genera within a single clade of m clade (11, 13). However, none of these analyses flowering plants. To explore the genetic basis for this scattered occurrence, we sequenced provide direct evidence or the molecular causes the genomes of 10 plant species covering the diversity of nodule morphotypes, bacterial of the specific gains and losses that explain the symbionts, and infection strategies. In a genome-wide comparative analysis of a total of distribution of NFN symbiosis in extant genera. 37 plant species, we discovered signatures of multiple independent loss-of-function events Exploring the genetic basis underlying the in the indispensable symbiotic regulator NODULE INCEPTION in 10 of 13 genomes of evolutionary dynamics of the NFN symbiosis in nonnodulating species within this clade. The discovery that multiple independent losses plants will improve our understanding of the di- shaped the present-day distribution of nitrogen-fixing root nodule symbiosis in plants versity of symbiotic associations observed in ex- reveals a phylogenetically wider distribution in evolutionary history and a so-far-underestimated tant taxa and the ecosystems they inhabit and selection pressure against this symbiosis. potentially provide keys to engineer it in crops and to predict the stability of this trait over long itrogen is one of the main requirements survival under nitrogen-limiting conditions. In evolutionary times. Here we used a genome-wide o for plant growth. Members of the legume agriculture, this independence from chemical comparison including genomic and phylogenomic n J family (Fabaceae, order Fabales) and of nitrogen fertilizer reduces costs and fossil fuel methods (14) to address the long-standing con- u l y 1 nine additional plant families benefit from consumption imposed by the Haber-Bosch pro- undrum of the evolution of NFN symbiosis and 5 Nsymbiosis with nitrogen-fixing bacteria, cess (2). Since the discovery of the NFN symbi- identify the underlying genetic players (15, 16). , 2 either phylogenetically diverse rhizobia or species osis, with rhizobia in 1888 and with Frankia in 0 1 8 Results of the genus Frankia, which a osted inside 1895 (3, 4), it has been unclear why it is restricted plant cells found within nodules—specialized to only a limited number of flowering plant species. Genome sequencing in the NFN clade host-derived lateral organs typically found on A major scientific step forward in our under- Sequenced genomes of nodulating species were roots. In this mutualistic nitrogen-fixing root standing was the reorganization of the phyloge- only available for a few agriculturally relevant nodule (NFN) symbiosis, intracellular bacteria netic tree of angiosperms in 1995, which revealed legume species belonging to a single subfamily convert atmospheric nitrogen into ammonium that plants forming the NFN symbiosis are re- (Papilionoideae), all derived from a single pre- by means of the enzyme nitrogenase (1). This stricted to the Fabales, Fagales, Cucurbitales, dicted evolutionary origin of NFN symbiosis, with “fixed nitrogen,” delivered to the host plant, is and Rosales orders, which together form the NFN no representation either of taxa representing an essential building block for amino acids, DNA, clade (5, 6). However, this reorganization also possible additional origins within legumes or of RNA, tetrapyrroles such as chlorophyll, and many raised new questions, because only 10 of the 28 nonlegume nodulating species widely accepted other molecules. This symbiosis enables plant plant families within the NFN clade contain plants as representing multiple additional origins (17–19). 1Faculty of Biology, Genetics, LMU Munich, Großhaderner Strasse 2-4, 82152 Martinsried, Germany. 2Plant Genome and Systems Biology (PGSB), Helmholtz Center Munich–German Research Center for Environmental Health, 85764 Neuherberg, Germany. 3BGI-Shenzhen, Shenzhen 518083, China. 4China National GeneBank, BGI-Shenzhen, Shenzhen 518120, China. 5Department of Microbiology, Weber State University, Ogden, UT 84408-2506, USA. 6Laboratoire de Recherche en Sciences Végétales (LRSV), Université de Toulouse, CNRS, UPS, 24 chemin de Borde Rouge, Auzeville, BP42617, 31326 Castanet Tolosan, France. 7LBMIBS, Departamento de Ciencia y Tecnología, Universidad Nacional de Quilmes, CONICET, R. Saénz Peña 352, B1876BXD Bernal, Argentina. 8Laboratory for Molecular Biology, Wageningen University, Droevendaalsesteeg 1, 6708 PB Wageningen, Netherlands. 9Université Lyon 1, Université de Lyon, CNRS, Ecologie Microbienne, UMR 5557, Villeurbanne, 69622 Cedex, France. 10Department of Plant Sciences, University of California, Davis, Davis, CA 95616, USA. 11Laboratory for Plant Physiology, Wageningen University, Droevendaalsesteeg 1, 6708 PB Wageningen, Netherlands. 12Department of Ecology, Environment and Plant Sciences, Stockholm University, 106 91 Stockholm, Sweden. 13French National Research Institute for Sustainable Development (IRD), UMR LSTM (IRD–INRA–CIRAD–Université Montpellier–Supagro), Campus International de Baillarguet, TA A-82/J, 34398, Montpellier Cedex 5, France. 14Section of Plant Breeding and Genetics, School of Integrated Plant Science, Cornell University, Ithaca, NY 14853, USA. 15BGI Genomics, BGI-Shenzhen, Shenzhen 518083, China. 16Department of Agronomy, University of Wisconsin–Madison, Madison, WI 53706, USA. 17Department of Bacteriology, University of Wisconsin–Madison, Madison, WI 53706, USA. 18James D. Watson Institute of Genome Sciences, Hangzhou 310058, China. 19TUM School of Life Sciences Weihenstephan, Technical University of Munich, 85354 Freising, Germany. 20Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9, Canada. 21Department of Medicine, University of Alberta, Edmonton, AB T6G 2E, Canada. *Corresponding author. Email: chengshf@ (S.C.); pierre-marc.delaux@lrsv.ups-tlse.fr (P.-M.D.); parniske@lmu.de (M.P.) Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 1 of 11 RESEARCH | RESEARCH ARTICLE Conversely, sequenced genomes of nonnodulating evolutionary origins of NFN symbiosis, according culated a separate phylogeny and subsequently species were restricted to the Fagales, Rosales, and to the multiple-gains hypothesis (5, 7, 10, 11). inferred orthologs for all genes of the reference Cucurbitales and did not include nonnodulating These 10 sequenced genomes, together with 18 species, Medicago truncatula (16). We selected legume taxa (table S1). To overcome this sampling other genomes from the NFN clade and 9 gen- groups for which orthologs were absent in the bias, which restricted the phylogenomic analysis, omes from other flowering plants as an out- nine species outside of the NFN clade but were we sequenced de novo the genomes of seven group (Fig. 1), were compared to detect molecular retained in nodulating species (fig. S3). To ob- nodulating species belonging to the Cucurbitales, traces supporting any of the three postulated tain a candidate set for manual validation from Fagales, and Rosales orders and the Caesalpi- events in the evolution of NFN symbiosis: (i) the total of 29,213 orthologous groups, we used nioideae subfamily of the Fabaceae, representing predisposition to evolve it, (ii) multiple inde- an automated filtering approach with relaxed a possible second origin of NFN symbiosis in pendent gains, and (iii) multiple independent criteria, which allowed for the absence of ortho- legumes. Three nonnodulating species from the losses of NFN symbiosis. logs in a small subset of nodulating species NFN clade were also sequenced, notably Nissolia (16). This step was necessary to avoid the loss of schotii, a papilionoid legume that has lost the The putative predisposition event putative candidates owing to missed gene models ability to form the NFN symbiosis and which did not involve NFN clade-specific resulting in false negatives, as they often occur in therefore provides insights on the genomic con- gene gains automated gene-prediction pipelines. Our re- sequences of losing the symbiosis (Fig. 1 and The predisposition event postulated by Soltis et al. laxed filter identified a total of 31 orthologous fig. S1). For each species, 144 to 381 Gb of Illumina in 1995 (5) may be based on the acquisition of candidate groups (table S30). All of these can- reads were obtained, covering the estimated one or several genes or sequence modifications didate groups underwent an iterative manual genomes at least 189-fold and up to 1113-fold, specific to the NFN clade. This acquisition would curation, including a search for missed gene with the resulting scaffold N50 length between be also consistent with a single-origin hypothesis, models and recalculation of phylogenies and 96 kb and 1.18 Mb and an average genome in which NFN symbiosis in all taxa is predicted orthologs (16). completeness of 96% (Fig. 1, figs. S1 and S2, and as a homologous trait. Genes acquired during the Not a single candidate gene was identified D o tables S2 to S29). Altogether, the genomes of predisposition event are expected to be specific that matched the evolutionary pattern expected w n l species sequenced here represent six families to the NFN clade and present in all nodulating for predisposition-related genes, suggesting that o a and most known nodule anatomy types and root species. To search for genes following this evol- genes gained in the most recent common an- d e infection pathways; include hosts for the main utionary pattern, we identified gene families cestor of the NFN clade have been conserved or d f r classes of nodule-inducing symbiotic nitrogen- across all 37 plant genomes in our dataset using lost irrespectively of the symbiotic state of the o m fixing a-proteobacteria, b-proteobacteria, and the Orthofinder pipeline (20). For each of the lineages. If the predisposition indeed occurred, actinobacteria; and cover six to seven independent resulting 29,433 gene family clusters, we cal- this result indicates that it did not involve the Fig. 1. Genome features of species used in this study. Genome statistics are shown as pictograms for species used in this study, with nodulating species highlighted by blue sectors. Species names are shown as four-letter abbreviations o at the outer circle, with their n J u taxonomic order color coded as l y shown in the top-right legend. 1 Newly sequenced species are in 5 , 2 0 bold blue letters. The next two 1 circles show, as pie charts, the 8 proportion of complete BUSCO genes [see (68)] detected in the genome assembly (light gray) and the percentage of assembled sequence relative to the estimated genome size (dark gray), respectively. Scaffold N50 values are depicted as bubble charts (black) capped to a al N50 of 1 Mb to reduce graphical biases by finished genomes assembled to pseudochromosomes. Note that even for assemblies in this study with low contiguity, the BUSCO results suggest that the gene space has been well covered (tables S4 and S5). The innermost circle represents the genome size by proportional chromosome pictograms. Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 2 of 11 RESEARCH | RESEARCH ARTICLE acquisition of genes but rather the co-option However, similar to the NRT1/PTR family, all of role of NIN in NFN symbiosis in actinorhizal of existing genes and their corresponding the enriched clusters also expanded outside the host plants. pathways. NFN clade. Inside the NFN clade, these clusters Confirmation of NIN absence expanded beside the hypothesized gain of NFN Gene family dynamics is compatible symbiosis nodes at many additional nodes (table by microsynteny with multiple gains S31). In our survey of all independent gene family In addition to the presence of NIN in all the Multiple molecular mechanisms leading to the expansions, we did not identify any cluster that nodulating species in our dataset, synteny anal- convergent evolution of a trait have been iden- displayed parallel expansions, one possibility ysis revealed the conservation of the syntenic tified (21). Deep homology (22), the independent among others that would indicate convergent blocks surrounding the NIN locus across the recruitment of a homologous gene set for the recruitment for the independent gains of NFN NFN clade (Fig. 2). By contrast, 10 of 13 non- development of nonhomologous traits, has been symbiosis (7, 16). If NFN symbiosis indeed evolved nodulating species of the Fabales, Fagales, Cu- proposed for the NFN symbiosis (7, 23). Indeed, multiple times, our genome-wide analysis re- curbitales, and Rosales underwent partial (four several genes initially identified for their sym- vealed hundreds of clade-specific candidate genes species) or complete (six species) deletions of biotic role in legumes were later found to also that, together with gene co-option, may have NIN from the conserved genomic block (Fig. 2). play a symbiotic role in Fagales (24–27), Rosales played a role in the putative independent evolu- The legume family is divided into six subfamilies, (28, 29), and Cucurbitales (30). An alternative tions of this trait. two of which include nodulating species (39). mechanism for the evolution of new traits is gene According to previous estimates (7), Cercis (a family expansion, as exemplified by the parallel Genomic evidence for multiple losses member of Cercidoideae, which is sister to most diversification of the zinc-finger transcription fac- Testing homologies of a trait shared by multiple or all other legumes) and Castanospermum tor family in the evolution of a dominant yeast taxa typically involves assessing the trait in these (a member of a clade sister to the crown group form in fungi (31) or the acquisition of strigolactone species and inferring origins of the trait once of papilionoid legumes, in which nearly all perception in parasitic plants (32). We hypothe- homology is accepted or rejected. However, in- members form the NFN symbiosis) both rep- D o sized that if NFN symbiosis evolved multiple formation on the origin of a trait, and thus its resent lineages in which NFN symbiosis never w n l times, independent expansions in the same gene homology, can also be obtained from taxa lacking occurred. By contrast, Nissolia and its sister o a family might have been involved. Given that our the trait, by distinguishing primary absence (the genus Chaetocalyx are nonnodulating genera d e dataset covers six to seven of the predicted in- taxon never had the trait) from secondary loss nested within the nodulating crown group of d f r dependent gains of NFN symbiosis, it allowed us of the trait (23). It has been demonstrated that papilionoids and thus have been predicted to o m to search for gene families whose evolutionary genes involved in a specific biological process represent secondary loss of NFN symbiosis patterns are consistent and would support the are lost following the loss of that trait, a process (11, 40). Our synteny approach allowed the multiple-gains hypothesis (7). We analyzed Ortho- known as gene coelimination (35–37). To test discovery of the complete absence of NIN from finder clusters for copy-number variation to iden- the multiple-losses hypothesis, we searched the the genome of N. schottii (Fig. 2). In addition, tify gene family expansion events at each of these ortholog groups calculated above for an evolu- we confirmed the absence of NIN in three nodes (Fig. 2) (16). Multiple alternative models tionary pattern that retained orthologs in all Chaetocalyx species (fig. S5). Because NIN was have been proposed for the independent gain nodulating species but lost them in nonnodulat- present in the most recent common ancestor of NFN symbiosis. In one scenario (7), a single ing ones. To filter and confirm the list of can- of the NFN clade and conserved in nodulating gain before the radiation of the legume family didate orthologous groups, we used the same species, the absence of this gene in the Nissolia- has been suggested that correlates with the ex- two-step process combining an automated pipe- Chaetocalyx lineage represents a loss that cor- pansion of 33 clusters in our analysis (Fig. 2). To line with relaxed criteria for nodulating species relates with, and is sufficient to explain, the loss test whether any of these clusters were en- followed by a careful manual curation and evalua- of NFN symbiosis. Cercis canadensis harbored o riched with differentially expressed genes (DEGs) tion step with stringent criteria (presence in all only a NIN pseudogene remnant in the genomic n J in nodule tissue versus root tissue, we derived for nodulating species required). The automated block, whereas the genome of C. australe com- u l y 1 M. truncatula (Medicago) gene expression data pipeline with relaxed criteria resulted in a list pletely lacked NIN (Fig. 2). These results demon- 5 for both conditions from the gene expression atlas of 121 candidate groups (table S32). During our strate three independent losses of NIN in the , 2 (33) and tested for an enrichment of Medicago manual confirmation and refinement, we rejected legume family. Similarly, the synteny analyses 0 1 DEGs in each cluster (table S31). We only found 31 of these candidate groups because orthologs confirmed a minimum of three independent 8 one such cluster that belongs to the nitrate were absent from one or more nodulating species. losses in nonnodulating Rosales and two in the transporter family NRT1/PTR (34). However, Another 62 candidate groups were rejected be- Cucurbitales (Fig. 2). Together, the diversity of this gene family appears to also be expanded at cause orthologs of more than 50% of nonnodulat- NIN deletions in the nonnodulating species is nodes not related to independent gains within ing species were present. A weak phylogenetic indicative of at least eight independent evolu- and outside of the NFN clade (table S31). An signal did not allow for inferring a reliable tionary events that led to the loss of NIN function alternative model proposes two independent orthology for 27 candidate groups. A single (Fig. 3). Following the multiple-gains hypothesis gains within the legumes: the first at the most gene, NODULE INCEPTION (NIN ), was con- (7), NFN symbiosis was predicted to have evolved recent common ancestor of the papilionoids and firmed to be present in the genomes of all the independently at least six times in the species the clade to which Castanospermum australe nodulating species and in the genome of species space sampled here (Fig. 2). Loss of NIN pro- belongs (21-291 enriched-expanded clusters), or at outside the NFN clade, and absent from most vides an alternative model with at least eight the base of the papilionoids (4-32), and the second nonnodulating ones in the NFN clade (Fig. 2 and independent losses of NFN symbiosis (Fig. 3). gain at the base of the caesalpinioid and mimosoid fig. S4). Forward genetic screens in the legumes Thus, the current distribution of NFN symbiosis clade (7-92). This clade could alternatively com- Lotus japonicus (27), Pisum sativum (38), and might be a combination of these two comple- prise two independent gains for Chamaecrista M. truncatula (24) identified NIN as indispens- mentary and not mutually exclusive models. fasciculata (39-710) and mimosoids (37-723). able for the two developmental aspects of the Loss of RPG Larger numbers of expanded gene families were NFN symbiosis: initiation of root nodule develop- observed for the predicted events in the Rosales, ment and the formation of the plant structure In parallel with these multiple confirmed losses Fagales, and Cucurbitales (Fig. 2 and table S31). facilitating intracellular uptake of bacteria (27). of NIN, the synteny analysis also confirmed the Taken together, we found 52 gene family clusters Furthermore, RNA interference–based suppres- presence of NIN in three nonnodulators of the that were enriched with differentially expressed sion of NIN expression in Casuarina glauca NFN clade (Fig. 2). We hypothesized that the loss genes in Medicago nodules and expanded mul- (Fagaceae, Fagales) impaired nodule formation of genes other than NIN may explain the non- tiple times at proposed independent gain nodes. (25), consistent with a conservation of the nodulating state of these species and that such Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 3 of 11 RESEARCH | RESEARCH ARTICLE genes may have been missed by the specific reverse genetics as critical for NFN symbiosis GROWTH (RPG) (41), which is present outside and stringent criteria for the detection of the in legumes (Fig. 3 and table S33). Among these the NFN clade, is missing in N. schottii and 11 presence-absence patterns. In addition to NIN, genes, 20 were conserved in nodulating species other nonnodulating species from the four we determined the presence-absence pattern and most nonnodulating ones (figs. S6 to S25). orders of the NFN clade (Fig. 3). These losses of 21 genes that were identified by forward and By contrast, RHIZOBIUM-DIRECTED POLAR were confirmed by microsynteny analyses for D o w n l o a d e d f r o m o n J u l y 1 5 , 2 0 1 8 Fig. 2. Gene family expansions and contractions in the NFN clade. In colored in green. NIN gene IDs are shown above the gene symbol. The the left panel, a species phylogeny of the used dataset is depicted. For synteny analysis upholds orthologous relationships drawn from the each node, the numbers of gene families showing expansions (+, blue) phylogenetic analysis and supports the absence of NIN in several species and contractions (−, red) are given. Blue boxes point out nodes hypothesized by verifying the existence of contiguous NIN regions without NIN genes. to be positions of independent gain events of the NFN symbiosis, including Enlarged gene models are only shown for fragmented NIN genes in all suggested alternative models in the given dataset (7). For example, comparison to the full M. truncatula NIN gene. Blocks with no fill and green among the Fabales, the dataset could comprise one, two, or three fill represent parts absent and present, respectively, when compared to independent gains. A black arrow marks the base of the NFN clade. The Medicago NIN. Insertions and deletions (INDEL) and premature stop right panel depicts syntenic relationships of the NIN region. NIN genes are codons (*) are symbolized by a vertical red line. Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 4 of 11 RESEARCH | RESEARCH ARTICLE all nonnodulating species (fig. S26). In the will be difficult to identify by phylogenomic anal- reflected a putative neo functionalization for NFN M. truncatula rpg mutant, infection threads ysis, because of the rapid evolution and expan- symbiosis–specific functions (table S34). are still present but their structure is abnormal sionary dynamics of these gene families (45). In Discussion (41), indicating that RPG, similar to NIN , is re- contrast to other symbiosis-relevant genes involved quired for proper infection-thread progression. in infection, NIN and RPG are only known to have In recent decades, the favored model to explain However, in contrast to nin mutants, nodules are NFN symbiosis–specific functions, whereas the the scattered occurrence of NFN symbiosis in formed on rpg mutant roots (41). Among the mutation of other genes may have more pleiotropic flowering plants predicted a single predisposi- nodulating species, RPG is absent in the papilionoid effects. For example, the key signaling components tion event at the base of the NFN clade followed Arachis ipaensis (Fig. 3). In the genus Arachis , SYMRK, CCaMK, and CYCLOPS are involved in by up to 16 origins, even though the occurrence infection threads were not observed; instead, rhizo- both NFN symbiosis and arbuscular mycorrhizal of multiple losses was never excluded (7, 11). bia appear to infect nodules intercellularly (42). symbiosis, the most widespread symbiosis in Our genome-wide comparative analysis did Polymorphism in RPG may represent an inter- land plants (46). Mutations in any of these three not detect gene gains specific to the NFN clade mediate step on an evolutionary path toward the genes would also affect arbuscular mycorrhizal and maintained in all nodulating species. Such loss of this symbiosis in Arachis , a genus in which symbiosis. Illustrating this dual selection pressure, genes would have been ideal candidates for either NFN symbiosis is described as a labile trait (43, 44). retention of these genes has been described in the the predisposition event (in the multiple-gains Absence of RPG in Arachis also explains why genus Lupinus, which lost the arbuscular my- hypothesis) or the evolution of NFN symbiosis the genome-wide comparative phylogenomic corrhizal symbiosis but retained NFN symbiosis itself in the hypothesis that NFN symbiosis evolved approach did not identify this gene, given that (37, 47, 48). Given that both NIN and RPG are only once in the most recent common ancestor the pipeline required the candidate genes to be present in species outside the NFN clade (Fig. 3 of the NFN clade (the single-gain hypothesis). present in all nodulating species. RPG was one and figs. S4 and S6), their consistent losses in This indicates that this step involved either fast- of the genes rejected for not fulfilling this crit- nonnodulating species suggest the shift in con- evolving genes that were not captured by our erion. Among nonnodulating species, Juglans straints on sequence evolution specifically in the phylogenomics pipeline or subtler genetic changes. D o regia (Fagales), Ziziphus jujuba , and Prunus NFN clade, which might be mirrored by Evolutionary developmental genetics in plant, w n l persica (Rosales) have lost RPG but retained NIN, signature of relaxed or positive selection on both fungal, and animal systems have revealed that o a suggesting that additional mutations might be genes at the base of the NFN clade. We even more than gains of genes, new traits often d e causative for the loss of NFN symbiosis in these investigated the selective pressure acting on the arise from the rewiring of existing gene networks d f r species. Some of the candidate mutation targets, NFN clade for these two genes using the PAML via gains or losses of cis regulatory elements lead- o m for example, lysine motif (LysM) receptors in- package (49). Results did not reveal a statisti- ing to the co-option of ancestral genes (50). A sim- volved in the perception of symbiotic signals cally significant positive or relaxed selection ilar mechanism may have acted in the most recent produced by nitrogen-fixing nodulating rhizobia, occurring in the NFN clade that would have common ancestor of the NFN clade. Fig. 3. Phylogenetic pattern of NFN symbiosis–related genes. The chronogram contains nodulating (blue) and nonnodulating (gray) species from all four orders of the NFN clade (blue circle with dot), to o which NFN symbiosis is limited. n J Nine species outside the NFN u l y clade are included as outgroups 1 5 at the top. The absence or , 2 presence of entire or 0 1 fragmented copies of 21 8 symbiosis genes are indicated by white, black, and gray boxes, respectively. Stars indicate independent losses of NIN. The independent loss or fragmentation of NIN correlates with the absence of nodules after the emergence of the NFN clade. RPG is lost or fragmented in even more nonnodulating species than NIN and also in the nodulating species A. ipaensis and Mimosa pudica. Asterisks indicate species sequenced for this study. INF, genes required for infection; NOD, genes involved in nodule organogenesis and regulation; CSG, genes required for both NFN symbiosis and arbuscular mycorrhiza symbiosis; mya, million years ago. Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 5 of 11 RESEARCH | RESEARCH ARTICLE Co-option of different, or homologous, in nitrogen-fixation—involving infection, building Deep genome sequencing was performed for most the case of deep homology, genetic compo- nodules to host bacteria, and providing carbon of the species, with at least 110-fold coverage after nents may lead to the convergent evolutions of to feed them—outweighs the benefit to the plant. a stringent data filtering and the highest sequenc- nonhomologous traits in multiple species (50). In most terrestrial habitats, nitrogen is limiting ing depth reached 535-fold in cleaned data. Deep homology has been invoked for the evolu- (54), suggesting that the scale should be tipped The overview statistics of data production are tion of NFN symbiosis, during either the pre- toward the conservation of NFN symbiosis once summarized in table S2 and fig. S1. disposition or the following putative multiple this complex trait evolves. In nitrogen-rich hab- To minimize sequencing errors and reduce gains (7). Our results support this hypothesis, itats, NFN symbiosis is known to be inhibited in genome assembly artifacts, several quality con- given that all the genes characterized for their legumes (55). Long-term fertilizer application trol steps were taken to filter out low-quality involvement in NFN symbiosis in legumes were would make NFN symbiosis–specific genes super- sequencing reads: already present in the most recent common fluous, leading to their eventual mutational in- Removal of N-rich reads: Reads that contained ancestor of the NFN clade and that we did not activation and loss. In addition to this abiotic more than 10% of “N”s bases or polyA structure detect genes specific to the NFN clade that were constraint, NFN symbiosis may be undermined were removed. conserved in all nodulating species (Figs. 2 and 3). by “cheating” bacteria that gain entry into root Removal of low-quality reads: Reads in which For the putative multiple gains of NFN symbiosis, nodules and are fed by the plant but do not 40% of the bases were low-quality (quality scores it cannot be excluded that gene gains were also deliver nitrogen (56–58). Cheaters may there- ≤7) were filtered out. involved in addition to co-option of ancestral fore imbalance the trade-off between the costs Filtering of reads with ≥10 nt aligned to the pathways. We identified hundreds of such and benefits of the association, as already pro- adapter sequences: Adaptor sequences were lineage-specific candidate genes (Fig. 2). How- posed on the basis of patterns of legume NFN aligned to read1 and read2 using a dynamic ever, considering that most of the predicted gains symbiosis in Africa (59). This would result in programming approach; if the aligned fragments in our analysis are located in terminal taxa that the loss of NFN symbiosis being adaptive, thus from read1 or read2 were reversed complementary are known to accumulate orphan genes and providing an ecological explanation for the oc- to each other, the pair was also removed. D o species-specific duplications in comparative ap- currence of this symbiosis in a few flowering Filtering of small-insert-size reads with insert w n l proaches, it can be anticipated that only a subset plant species. In this context, the finding that size (170 to 800 bp): The overlapping length be- o a of them participated in the evolution of NFN NIN participates in shaping the root microbiome tween read1 and read2 is ≥10 bp; 10% mismatch d e symbiosis (in the multiple-gains hypothesis) or beyond NFN symbiosis makes it a target of was allowed. d f r in lineage-specific refinements of the trait (in adaptive selection against this symbiosis (60). Filtering of PCR duplicates: If the PE read1 o m the single-gain hypothesis). Engineering biological nitrogen fixation in and read2 were 100% identical, these reads were Our results validate another hypothesis: mul- crops remains a goal of plant synthetic biologists, treated as duplicates and only one was retained. tiple independent losses of NFN symbiosis in the with the aim of improving food production in Trimming of read ends: The low-quality bases four orders of the NFN clade. In a classical model developing countries in which the application of from read ends (5′–5 bp, 3′–8 bp) were directly of evolution, if the number of losses necessary to nitrogen fertilizer is limited by economic and trimmed. explain the distribution of a trait in a given clade infrastructural constraints. Our results supporting This filtering process was carried out using an outnumbers the predicted gains, multiple gains the occurrence of multiple independent losses in-house Perl program. After filtering, on average, will be favored over multiples losses to explain indicates that the apparent selection against 150-fold sequencing coverage was generated. For the distribution of the trait. In legumes, up to NFN symbiosis must be taken into account by each species, clean reads were then passed to the six gains of NFN symbiosis were predicted (7), projects whose aim is to improve legumes and, genome assembler pipeline for de novo genome whereas the clear losses that we identified in even more, when considering the engineering assembly. C. canadensis, C. australe, and N. schotii now of nitrogen fixation in other crops. The statistics of clean data are also summar- o argue for a single origin before the radiation of ized in table S2. n J u Materials and methods the family (Fig. 3). Beyond legumes, the number l Plant material and sample preparation Genome assembly y 1 of validated losses of NFN symbiosis is consistent 5 with a single gain of this symbiosis in the most The origin of the plant material used for DNA or To optimize the strategy for genome assembly, a , 2 recent common ancestor of the NFN clade, even RNA extraction in this study is summarized in genome survey is necessary to estimate the gen- 0 1 though it does not reject the possible occurrence data S2. ome complexity. Some genomes are abundant 8 of multiple gains. The recent identification of loss Methods for plant growth, DNA extractions, in repetitive content and/or maintain a high het- of NFN symbiosis in the Rosales Trema orientalis and RNA extractions are described in the sup- erozygosity rate through genome k-mer–analysis. brings further support to this hypothesis (51). plementary materials online. A k-mer refers to a continuous sequence with k Besides NFN symbiosis, the single or multiple base pairs, typically extracted from the reads origin(s) of other traits, such as the evolution of Genome sequencing (thus, shorter than the read length, e.g., 17 bases complex multicellularity in fungi, are currently Figure S1 describes the overall strategy and per k-mer). If an “ideal” sequencing dataset is debated with the accumulating evidence of mul- results of the dataset production in this study. produced from a randomly whole-genome shotgun tiple losses demonstrated by the loss of associated Whole-genome sequencing for the 10 genomes process without sequencing errors or coverage essential genes (52). Thus, multiple gains and mul- was performed using Illumina sequencing tech- bias, the start positions of reads along the genome tiple losses are not mutually exclusive scenarios to nology (HISEq. 2000 and HISEq. 4000) at BGI- will follow Poisson distribution (61). Supposing explain the evolution of complex traits such as Shenzhen. Hierarchical library construction that the read length is far shorter than the genome NFN symbiosis. This also suggests that reduction, strategy was applied that typically included size, the k-mer can be regarded as randomly similarly to the evolution of complexity, might be multiple paired-end libraries with insert sizes of generated from the genome and their occurrence a major driver of the phenotypic diversity ob- 170, 250, 350, 500, and 800 bp and mate-pair (sequencing depth) also is expected to be Poisson served in extant organisms (36, 53). libraries with insert sizes of 2, 5, 10, and 20 kb. distributed (fig. S3A). Based on this assumption, The fixation of loss-of-function alleles of NIN Most of the paired-end and mate-pair libraries the genome size can be estimated as (62): (either complete loss or pseudogenization) in were prepared from large genomic fragments, nonnodulating species provides the genetic ex- typically of size 20 to 40 kb, or even larger. For Genome size ¼ k mernumber planation for the loss of NFN symbiosis in 10 some species, more small-insert-size PE libraries Average sequencing depth species representing eight nonnodulating line- were constructed to complement the limited ages. Fixation of such alleles requires ecological mate-pair libraries. The library construction For a “normal” diploid genome, the k-mer fre- conditions in which the cost of symbiotic for each species is summarized in table S36. quency produced from adequate reads would Griesmann et al., Science 361, eaat1743 (2018) 13 July 2018 6 of 11 RESEARCH | RESEARCH ARTICLE follow Poisson distribution. For genomes which A custom repeat library was constructed for each ( ). are either repeat-rich or highly heterozygous, species by careful self-training. To construct the An optimal core protein set was collected from an additional peak either indicative of highly re- repeat custom library, we first collected the min- several closely related species for homolog-based petitive content (typically twofold depths of the iature inverted repeat transposable elements gene prediction for each species, for example, gene main peak) or of high heterozygosity (63) (typi- (MITEs) from many closely-related species and models from the model plants like Arabidopsis cally half depths of the main peak) is expected created a lineage-specific custom library by MITE- thaliana and Oryza sativa, as well as from some next to the “main peak” (indicative of the normal hunter (70) with default parameters. For the well-annotated legumes like Medicago and Glycine diploid genome) from the frequency distribution, prediction of long terminal repeats (LTRs), we max. Default parameters were used to run or some even more complex scenarios either used LTRharvest (71) integrated in Geno ls MAKER-P with all integrated annotation sources caused by the unexpected non-canonical genomic (version 1.5.8) (72), defining LTR in the length of and to produce the final set of gene models for characteristics (e.g., degree of heterozygosity, 1.5 to 25 kb, with two terminal repeats ranging each species. The number of gene models for complexity of the size and distribution of re- from 100 to 6000 bp with ≥99% similarity. Ele- each species is summarized in table S26, and petitive content, etc.) coupled with the use of ments with intact PPT (poly purine tract) or PBS detailed statistics are summarized in table S28. different k-mer sizes. (primer binding site) were necessary to define BUSCO evaluation suggests complete and reli- LTR, which were identified by LTRdigest (71) able gene annotation for all newly sequenced K-mer statistics and distributions are using a eukaryotic tRNA library ( . genomes (tables S4 and S5). presented in fig. S2 and tables S6 to S25. /), whereas elements without appropriate HMMER-based engineer InterproScan (version During de novo genome assembly, we tried dif- PPT or PBS location were removed. To remove 5.11) (80) was used to predict gene function from ferent k-mers (from 23- to 33-mer) to construct false positives such as local gene clusters and several functional databases. The motifs and contigs, and the best k-mer (with the largest tandem local repeats, 50-bp flanking sequences domains of genes were determined by searching contig N50 length) was selected for the final run. on both sides of the LTRs of each candidate ele- against protein databases. An integrated gene Owing to differences in genome complexities be- ment were aligned using MUSCLE (73) with de- functional annotation is summarized in table S29 D o tween species, multiple genome assemblers were fault parameters; if the identity was greater than for all species. w n l applied to achieve the optimal assembly result. As or equal to 60%, the LTR element was considered o a d described in fig. S1, SOAPdenovo2 (version 2.04) as a false positive and removed. LTR elements Transcriptome sequencing e (64) was the most frequently used assembler and nested with other inserted, but unrelated, com- RNA samples were sequenced for seven species d f r Platanus (version 1.2.4) (65) for highly hetero- ponents were also removed. Exemplars were built to assist gene prediction in this study. The over- o m zygous genomes. After several rounds of assembly using a cutoff of 80% identity in 90% of element view of total RNA samples with various tissues is evaluations regarding contig contiguity and ge- length from an all versus all BLASTn search. summarized in table S37. For each RNA sample, nome completeness, the best assemblies (largest Terminal repeat retrotransposon in miniature a pair-end library with insert size of ~200 bp was contig N50 and highest BUSCO gene mapping (TRIM) libraries, with lengths of 70 to 500 kb, constructed following the manufacturer protocol. rate) were selected for the downstream gap- were built following a similar prediction strategy. Libraries were barcoded and pooled together closing step by Gapcloser (version 1.2) (64). For Furthermore, the genomic sequence was masked as input to the Illumina Hiseq 400