Cairo Tehran Giza Lagos Dakar Maputo Luanda Riyadh Kuwait Jeddah Aleppo Ibadan Douala Nairobi Baghdad Algiers Abidjan Kinshasa Damascus Mogadishu Alexandria Casablanca Addis Ababa Johannesburg Dar es Salaam Iran Algeria Mali Libya Sudan Chad Niger Angola Egypt Saudi Arabia Ethiopia Dem. Rep. Congo Nigeria Iraq Tanzania Zambia South Africa Namibia Mauritania Kenya S. Sudan Morocco Yemen Mozambique Somalia Botswana Congo Oman Madagascar Cameroon Gabon Zimbabwe Ghana Syria Guinea Uganda Central African Rep. Côte d'Ivoire Senegal Burkina Faso Benin Eritrea Somaliland Tunisia Malawi Turkey Liberia Jordan Togo Afghanistan W. Sahara Sierra Leone Israel Lesotho Burundi United Arab Emirates Italy Rwanda Spain Djibouti Kuwait Eq. Guinea Guinea-Bissau Qatar Greece Swaziland Gambia Lebanon Pakistan Cyprus Palestine N. Cyprus Reunion Mauritius Comoros Madeira Bahrain Turkmenistan Malta São Tomé and Principe Seychelles Legend l arg e prot ec t ed areas likelihood of poisoning Value H i gh : 0 . 998477 Low : 2 . 18697e - 009 poisoning likelihood 0 500 1 , 000 1 , 500 2 , 000 250 K i l om e t e rs map T11 African vulture hotspot mapping 24 M ay 2017 map proj ec t i on: Lam be rt Azi m ut h Equal A rea ( cent ra l m eri d i an 20 deg ree s) CREDITS: Coordination: Ralph Buij (Wageningen University & Research), Corinne Kendall (North Carolina Zoo), Ara Monadjem (University of Swaziland). Data collation: Lutfor Rahman & Lou Luddington (Habitat Info). Analysis & map production: Rob Davies (Habitat Info). Finance: The vulture surveys, data gathering and habitat and threat modeling were funded by the following organisations through Wageningen University & Research (which also contributed resources): Dutch Ministry of Economic Affairs, WWF-Netherlands, UNEP-CMS Raptors MoU, North Carolina Zoo, Fondation LePal Nature, Quagga Foundation, Stichting Vogelpark Avifauna, Stichting Koninklijke Rotterdamse Diergaarde, Detroit Zoological Society, and Stichting Wildlife. Through The Peregrine Fund this project benefited greatly from access to the ESRI Grant Scheme. Data on vultures were contributed or facilitated by the following individuals: Yilma D Abebe, Hichem Azafzaf, Laila Bahaa El Din, Neil & Liz Baker, Clive R Barlow, Keith Bildstein, Claire Bracebridge, Andy Branfield, Erik & Asaph Brohaugh, Joost Brouwer, Chris Brown, Evan Buechley, Ralph Buij & Barbara Croes, Andre Botha, Mike Cadman, Alazar Daka Rufo, Rob Davies, Maria Diekmann, Nina Farwig, Oliver Fox, Toby Galligan, Beckie Garbett, Ashwell Glasson, Roi Harel, Stratton Hatfield, Ohad Hatzofe, Joseph Heymans, Constant Hoogstad, Mawdo J Jallow, Walter Jubber, Gregory Kaltenecker, Adam Kane, Chris Kelly, Alan & Meg Kemp, Corinne Kendall, Holger & Claire Kolberg, Bernard & Antje Madden, Glyn Maude, John Mendelsohn, Mike McGrady, Ara Monadjem, Campbell Murn, Ran Nathan, Karin Nelson, Stoyan Nikolov, Darcy Ogada, Steffen Oppel, Louis Phipps, Bram Piot, Thomas Rabeil, Sascha Rösner, Lizanne Roxburgh, Volker Salewski, Andrea Santangeli, Dana Schabo, Orr Spiegel, Lindy Thompson, Simon Thomsett, Dirk van Stuyvenberg, Rien van Wijk, Munir Virani, Tim Wacher, Kerri Wolter (VULPRO) and numerous other African Raptor Databank observers; and by the following organisations: AFRICAN RAPTOR DATABANK, AFRICAN IMPACT, BIRDLIFE INTERNATIONAL & NATURESERVE, BIRDLIFE BULGARIA (BSPB), BIRDLIFE TUNISIA (AAO), BOISE STATE UNIVERSITY, CITES (MIKE DATABASE), ENDANGERED WILDLIFE TRUST, HAWK CONSERVANCY TRUST, HAWK MOUNTAIN SANCTUARY, INTERNATIONAL UNION FOR CONSERVATION OF NATURE (AFRICAN ELEPHANT DATABASE & REDLIST MAPS), ISRAEL NATURE & PARKS AUTHORITY, NATURAL HISTORY MUSEUM (TRING), MOVEBANK, NIOKOLO-KOBA CITIZEN SCIENCE PROJECT, NORTH CAROLINA ZOO, RAPTORS BOTSWANA, RARE AND ENDANGERED SPECIES TRUST, ROYAL SOCIETY FOR THE PROTECTION OF BIRDS, TANZANIAN BIRD ATLAS, THE PEREGRINE FUND, UNIVERSITY OF UTAH, VULPRO, WEST AFRICAN BIRD DATABASE, WILDLIFE ACT and WILDLIFE CONSERVATION SOCIETY. M ETHODS : We adde d g e o- re fe re nc i ng i nform a t i on t o 444 records i n t he poi s oni ng s dat abase mai nt ai ne d by The Pereg r i ne Fund ( Darcy Og ada) a nd Endang e re d Wi l d l i fe Trus t ( L i za nne Roxburg h & And re Bot ha) . Lou Ludd i ng t on assi st ed wi t h t hi s w ork. We adde d four fi el ds t o t he dat a t abl e: re sol ut i on ( 7 c l asses as pe r ARDB ) , g eoreference d by , vul t ures ki l l ed ( y / n) a nd a dat e- t i me fi el d t o e na b l e h ot s pot m a ppi ng over t i me. Th e se e m pi ri cal dat a of wh e re poi s oni ng of vul t ure s a nd ot her wi l d l i fe has ac t ua l l y ocurre d ac ross A fr i ca coul d be us e d i n a s e parat e i nduc t i ve model l i ng exe rc i se us i ng t he same expl a na t ory var i abl es a nd M axent s oft wa re t hat we re us e d i n t he vul t ure d i st r i b ut i on model l i ng . A l i ke l i h ood of poi s oni ng i s pre se nt ed i n t he m a p. The model y i el ded g ood re sul t s wi t h A rea Unde r t he Curve ( AUC) val ue of 0 . 846 for t est records ( 15% of dat a we re hel d back for eval ua t i on purpos e s ) . A si ng l e var i abl e poach10km whi ch was a c rude i nd i cat or of b ot h el e ph a nt a nd rh i no poach i ng com bi ne d made t he l arg e s t c ont r i b ut i on t o t he m odel , fol l ow e d by g rasscove r ( more poi s oni ng i n g ras s l a nd habi t at s) a nd perce nt non- Chr i st i an a re as ( more poi s oni ng i n Chr i st i an areas) . I nd i cat ors of l i v est ock farm i ng a nd of d i st ance t o t ra d i t i ona l med i c i ne m arke t s d i d not m a ke i m port a nt c ont r i b ut i ons t o t he m odel s ug g e s t i ng t hat i nt e nt i ona l poi s oni ng of vul t ures as s e nt i ne l s may be t he bi ggest curre nt d r i ver of poi s oni ng ( b ut i t may be more re port ed t han ot her t ypes) . But a not her var i abl e, d i st ance t o l arg e prot ec t ed are a ( >1000km 2 ) , had t he g re at est pe rm ut at i on i m port ance , i . e. i f t hi s vari abl e was ad j us t ed i t cause d m os t pe rt urbance . C l ear l y from t he map more vul t ure poi s oni ng ocurs i n t he v i c i ni t y of l arg e prot ec t ed a re as a nd t h ere a re s om e si mi l ar i t i es t o t he combi ne d poi s oni ng t h re at map whi ch was a deduc t i ve m odel .