What

Project Coordinator PERSEUS stands for Policy-oriented marine Environmental G Reygondeau ([email protected]), JO Irisson, Hellenic Centre for Marine Research Research in the Southern EUropean Seas and is a large-scale C Albouy*, SD Ayata, F Benedetti, C Guieu, S Gasparini, P Koubbi e-mail: [email protected] research project funded by the European Commission’s 7th Laboratoire d'Océanographie de Villefranche (LOV) www.perseus-net.eu Framework Research Programme (FP7) under the theme * Université du Québec à Rimouski “Oceans of Tomorrow” OCEAN 2011-2013. The total project budget is nearly €17 million, of which almost €13 million (or 76%) is funded as a research grant under FP7.

Ecoregionalisation of the Mediterranean Sea

Summary

The discipline of biogeography aims to study the spatial distribution of species in relation with environmental and geographical gradients. More recently, this approach has been used to partition geographical areas according to biological, physical and chemical features. This novel geographical framework is helpful for ecological monitoring or conservation purposes as it outlines ecological discontinuities. Nonetheless, in a context where an ecosystem-based approach is advised for almost all ecological management, no geographical framework was proposed based on all ecosystem components (from environmental conditions to the spatial distributions of species). In this study, based on the most comprehensive dataset gathered on the Mediterranean sea and on newly developed mathematical tools, we propose an objective 3D mapping of biogeochemical regions (based on more than 10 environmental variables) and ecoregions (based on more than 1200 species modelled distributions, from phytoplankton to top predators) in that basin. In addition, we evaluate the weights of various anthropogenic pressures on each ecoregion, along three categories: climate change pressure, fisheries pressure and direct anthropogenic pressure (such as pollution, oil spills, etc.).

Data

Clustering PERSEUS Consortium Partners

Environment Why 22 variables : Tº, salinity, chloro-Marine life’s at stake. Human activity, such as overexploitation k-means c-means hclust complete hclust Ward 1. HELLENIC CENTER FOR MARINE RESEARCH (HCMR) 2. MIDDLE EAST TECHNICAL UNIVERSITY (METU) phyll but also depth of the of fisheries resources, pollution related to urbanisation and 3. INSTITUT FRANCAIS DE RECHERCHE POUR L’EXPLOITATION DE LA MER (IFREMER) euphotic layer, strength of the industrial activities, oil spills, alien species introduction, 4. AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS (CSIC) 5. CONSORZIO NAZIONALE INTERUNIVERSITARIO PER LE SCIENZE DEL MARE (CoNISMa) thermocline, … litter, and even tourism, are all creating serious and soon, Estimation ofDEthe optimal number ofGEOLOGIE clusters 6. INSTITUTUL NATIONAL CERCETARE-DEZVOLTARE PENTRU SI GEOECOLOGIE MARINA-GEOECOMAR (GeoEcoMar) Spatialised on a 0.2º grid and 25 irreversible threats to our seas. Combined with the impacts 7. PLAN BLEU POUR L’ENVIRONNEMENT ET LE DEVELOPPEMENT EN MEDITERRANNEE (Plan Bleu) 8. COSNAV ENGINEERING SRL of global warming and other natural hazards, our seas are vertical levels. 9. UNIVERSITA TA MALTA (UMT-IOI-MOU) fundamentally changing and the consequences are just now 10. EIR SYMVOULOI ANAPTYXIS ETAIREIA PERIORISMENIS EFTHYNIS (EIR) Grouped in epi, meso, bathypela11. BC3 BASQUE CENTRE FOR CLIMATE CHANGE - KLIMA ALDAKETA IKERGAI (BC3) starting to be fully understood. The new knowledge, insights,Superposition gic layers and bottom. of ESPANOL DE OCEANOGRAFIA (IEO) 12. INSTITUTO 13. UNIVERSITAT DE BARCELONA (UB) Sources : MedAtlas, QuickSCAT, strategies and policies resulting from PERSEUS will becluster boundaries 14. UNIVERSITAT POLITECNICA DE CATALUNYA (UPC) valuable on an international scale towards combating these 15. CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE (CNRS) GEBCO, and various papers. 16. UNIVERSITE D’AIX-MARSEILLE (UnivMed/COM) pressures and saving our seas. 17. UNIVERSITE PIERRE ET MARIE CURIE - PARIS 6 (UPMC) Species (Reflect physical features 18. UNIVERSITE PAUL SABATIER TOULOUSE III (UPS-LA) 19. CENTRO EURO-MEDITERRANEO PER I CAMBIAMENTI CLIMATICI SCARL (CMCC) Occurrence record of 1281 species in the flow) Where 20. CONSIGLIO NAZIONALE DELLE RICERCHE (CNR) in 4 groups: phytoplankton, zooplankton, fish, others (mammals, sharks, turtle) 21. AGENZIA NAZIONALE PER LE NUOVE TECNOLOGIE,L’ENERGIA E LO SVILUPPO ECONOMICO The marine ecosystems of the Mediterranean and Black SOSTENIBILE (ENEA) Sources : Copepod, IOBIS, GBIF, Pangea, FishBase, SeSame, Alvain et al 2008, 22. JRC -JOINT RESEARCH CENTRE- EUROPEAN COMMISSION (JRC IES) Seas, together representing the SES, are the focus of the 23. ISTITUTO NAZIONALE DI OCEANOGRAFIA E DI GEOFISICA SPERIMENTALE OGS (OGS) Albouy et al 2012 24. STAZIONE ZOOLOGICA ANTON DOHRN (SZN) project’s research. PLYMOUTH MARINE (PML) Anthropogenic impact 18 variables from Halpern et al 2008 Watershed 25. algorithm and LABORATORY definition of spatially coherent clusters = regions

How

This map highlights that the features are stronger and better defined near the surface. The main frontal structures are highlighted by the analysis (Ligurian, Alboran…).

Who & When

The PERSEUS project is implemented by 53 Partners (over 300 scientists) from 21 countries over the course of 4 years, from 1 January 2012 until 31 December 2015.

Clean Seas by 2020

The definition of the envelope is based on the Mahalanobis distance from the point of interest (blue) to the cloud of points Developing defining the region/species (black and red). science-based Therefore it takes the covariance structure policy recommendations toward into account (shape of the cloud).

Good Environmental Status (GES) in the Mediterranean Black Seas The probabilities are estimatedand through permutations, following the NonParametric Probabilistic Ecological Niche (NPPEN) algorithm (Beaugrand et al 2011).

Anthropogenic impacts

Ecosystem level regions are defined by combining per-group ecoregions boundaries. Within those regions, potential threats are mapped and quantified along 3 groups: climate related (Tº, pH, …), fisheries related (total catch, bycatch, …) and direct pressures (pollution, …). Coastal ecosystems are most affected, notably Tunisia and Northern Adriatic. Climate change threats are the only ones concentrated in the open sea. They will affect regions of high planktonic diversity. The worst prognosis for combined effects is for the regions offshore of Egypt, Lebanon and Greece.

Ecoregions

Fish 1 1 0

Biogeochemicals regions of the epipelagic layer 2 3 4 5 6 7 8 9 10 11 12 Biogeochemicals regions of the mesopelagic layer 2 3 4 5 6 7 8 9 10 11 12 Biogeochemicals regions of the bathypelagic layer 2 3 4 5 6 7 8 9 10 11 12 13 Index of stability of each cluster 10

20

30

40

50

60

70

80

90 100

Others

index of cluster stability

1

Latitude

Direct anthropogenic

Longitude

Fisheries

Biogeochemical regions

Transect

January 2012 – December 2015

Climate

Cluster stability is low in those dynamic frontal regions, as well as in the strong vertical convection regions (Gulf of Lion, 42° to 44°N from 0 to 700 m).

Diversity (log10)

43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.

Used in two ways: ‣ Environmental envelopes of each cluster and computation, for each point of the map, of the probability of belonging to that cluster; ‣ Environmental envelope of the occurrences of each species and computation of the probability of presence at any location.

Total ecoregions

Clusters built from database data can be checked against real-world cruise data at the surface (BOUM cruise - bottom left) and show good agreement in the boundaries between regions.

Ecoregions

Phytoplankton

Variables relevant for each layer (epi, meso, bathypelagic and bottom) are fed to the clustering algorithm. This results in biogeochemical regions which present distinct environmental profiles. They cannot be simply considered as water masses (identified by T/S diagram) but are biotopes defined by multivariate environmental intervals. Once such clusters are defined, their environmental envelopes can be extracted and reprojected on the full-size dataset (0.2º, 25 layers) giving a 3D map of the region.

UNIVERSITY OF PLYMOUTH (UoP) STICHTING DELTARES (DELTARES) UNIVERSITEIT UTRECHT (UU) UNIVERSITE DE LIEGE (Ulg) UNIVERSITY OF THE AEGEAN (AEGEAN) NATIONAL AND KAPODISTRIAN UNIVERSITY OF ATHENS (NKUA) PANEPISTIMIO KRITIS (UNIVERSITY OF CRETE) (UoC ECPL) THE CYPRUS RESEARCH AND EDUCATIONAL FOUNDATION (CyI) UNIVERSITY OF CYPRUS (OC-UCY) NACIONALNI INSTITUT ZA BIOLOGIJO (NIB) INSTITUTE OF OCEANOGRAPHY AND FISHERIES (IOF) ISRAEL OCEANOGRAPHIC AND LIMNOLOGICAL RESEARCH LIMITED (IOLR) UNIVERSITY OF HAIFA (HU) BLACK SEA NGO NETWORK (BSNN) SOFIISKI UNIVERSITET SVETI KLIMENT OHRIDSKI (DMG-SU) INSTITUT PO BIORAZNOOBRAZIE I EKOSISTEMNI IZSLEDVANIYA BALGARSKA AKADEMIYA NA NAUKITE (IBER - BAS) INSTITUTE OF OCEANOLOGY - BULGARIAN ACADEMY OF SCIENCES (IO-BAS) INSTITUTUL NATIONAL DE CERCETARE-DEZVOLTARE MARINA GRIGORE ANTIPA (NIMRD) ISTANBUL UNIVERSITY (IU) A.O. KOVALEVSKIY INSTITUTE OF BIOLOGY OF SOUTHERN SEAS (IBSS) MARINE HYDROPHYSICAL INSTITUTE - UKRAINIAN NATIONAL ACADEMY OF SCIENCES (MHI) ODESSA NATIONAL I.I. MECHNIKOV UNIVERSITY (ONU) P.P. SHIRSHOV INSTITUTE OF OCEANOLOGY OF RUSSIAN ACADEMY OF SCIENCES (SIO-RAS) IVANE JAVAKHISHVILI TBILISI STATE UNIVERSITY (TSU) INSTITUT NATIONAL DE RECHERCHE HALIEUTIQUE (INRH) CLU srl ECOLOGIC INSTITUT gemeinnützige GmbH (ECOLOGIC) SAROST SA

The innovation in PERSEUS is that, for the first time, the scientific community has been given a specific mandate The probability occurrence of each recommendations species is mappedfor through envelope (niche) modelling. Diversity is to develop of science-based policy estimated as the sum of the of occurrence. those probabilities are used as the data for the better governance of probabilities the SES marine environment. Finally, By clustering, which delineates ecoregions per taxonomic group. merging natural and socio-economic sciences, PERSEUS Lower patterns areand explained willtrophic fill thelevel knowledge gaps design by an temperature effective and (high temp, high diversity) and food availability (high Chla, governance low diversity). Fish diversity mostly demersal. Diversity and regionalisation of other, innovative framework, based onissound scientific knowledge, to help achieve ‘Good Environmental in Gibraltar). larger species, reflects their behaviour (migrationStatus’, through 42. line with objectives and principles of the EU’s MSFD.

Zooplankton

BIOGEOCHEMICAL REGIONS

26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

Envelope modelling

www.perseus-net.eu

1307

Dr. Evangelos Papathanassiou

TAKE HOME ‣ 12-15 distinct regions ‣ strong vertical differentiation ‣ biochemical regions well defined by fronts ‣ ecology refines biochemical regions ‣ differences in distribution/diversity between “trophic” groups ‣ anthropogenic impact uneven, affects high diversity regions

Next

‣ Better, more ecological, grouping of species ‣ Multiple niche models ‣ Creation of a total environmental impact index

Biblio Albouy et al (2012) Global Change Biology, 18(10):2995–3003 Alvain et al (2008) Global Biogeochemical Cycles, 22, GB3001 Beaugrand et al (2011) Marine Ecology Progress Series, 424:175-190 BOUM cruise (2010) Biogeoscience, 7 Halpern et al (2008) Science, 319(5865):948-952