Decision Support for Seascape Conservation and Ecosystem-Based Marine Management in the Northern Baltic Sea

Decision support for seascape conservation and ecosystem-based marine management in the northern Baltic Sea

ELINA VIRTANEN

ACADEMIC DISSERTATION To be presented, with the permission of the Faculty of Science of the University of Helsinki, for public examination in banquet room 303, Unioninkatu 33, on 19th of October 2020, at 12 o´clock.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 / HELSINKI 2020

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Author´s address: Elina Virtanen Marine Research Centre Finnish Environment Institute Latokartanonkaari 11 00790 Helsinki, Finland

Supervised by: Research Director Atte Moilanen Department of Geosciences and Geography University of Helsinki

Reviewed by: Professor Erik Bonsdorff Environmental and Marine Biology Faculty of Science and Engineering Åbo Akademi University

Senior research scientist Matt White Biodiversity Division, Department of Environment, Land, Water & Planning Arthur Rylah Institute for Environmental Research, Australia

Opponent: Professor Mary Wisz Section for Ocean Sustainability, Governance and Management World Maritime University Senior scientist at National Institute of Aquatic Resources Section for Ecosystem-Based Marine Management Technical University of Denmark

ISSN-L 1798-7911 ISSN 1798-7911 (print) ISBN 978-951-51-6576-3 (paperback) ISBN 978-951-51-6577-0 (pdf) http://ethesis.helsinki.fi

Unigrafia Oy, Helsinki 2020 2

“No blue, no green.” Sylvia Earle

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Virtanen E., 2020. Decision Support for Seascape Conservation and Ecosystem-Based Marine Management in the Northern Baltic Sea. University of Helsinki, Department of

Geosciences and Geography A87. Unigrafia Oy, Helsinki. 54 pages, 4 figures and 2 tables.

Abstract

Marine ecosystems are degrading around marine management. The role of the world at an unprecedented rate. Loss of management in sustaining marine biodiversity, population declines, invasion biodiversity is investigated and the of non-indigenous species, and change in applicability of methods developed in community composition are apparent in all terrestrial realm to marine environments is marine ecosystems. Various policies at evaluated. multiple management levels address these The case studies are located in the challenges with specific targets for good northern Baltic Sea, where multiple ecological and environmental status of stressors threaten marine biodiversity. The marine areas. While various policies, work relies on extensive species inventory directives and strategies are applicable at data from 140,000 underwater sites, global and regional levels, threats facing collected by the Finnish Inventory marine ecosystems in coastal areas are more Programme for the Underwater Marine localized. Thus, to achieve effective results, Environment (VELMU). Statistical conservation and management actions modelling was used in case studies (1) and should be designed and addressed locally, (4) to explain the distribution of species, and and carefully targeted to maximize cost- further in case studies (2) and (3) in efficiency and benefits for the marine describing hypoxia probabilities and the ecosystem. occurrence of ferromanganese concretions, In this thesis, four case studies are respectively. Further, key areas for developed which demonstrate how spatially conservation were identified with spatial explicit analyses can support seascape prioritization in case study (1). conservation, sustainable use of marine Based on the results, current marine areas, as well as effective management protected areas (MPAs) leave almost three- actions: (1) locate key areas for quarters of ecologically important species conservation, (2) pinpoint areas for effective occurrence areas unprotected. This nutrient abatement, (3) identify locations for highlights the need to further develop marine mineral extraction, and (4) estimate current MPA network, and the role of spatial potential future changes in key communities planning in guiding the allocation of marine with the projected declines in marine areas to human activities. Knowledge of environment. This thesis aims to show how unprotected key areas can be further utilized extensive data combined with appropriate to promote private seascape conservation spatial analysis paths together with cross- and sustainable use of marine areas. In case discplinary marine science can support study (2), areas naturally prone to hypoxia seascape conservation and ecosystem-based development were identified with spatial 2 analyses, borrowing concepts and to take place also in the Baltic Sea. Results methodologies from landscape ecology. The of case studies (1) and (3) can guide approach developed can be used to detrimental mining activities to ecologically optimally target nutrient abatement less valuable areas, where abundant measures to where they are most likely to be concretions can be found. efficient, as well as explain why some areas Spatially explicit analyses described in are more or less immune to nutrient case studies (1)–(4) can provide valuable abatement actions already taken. Case study support for seascape conservation and (4) further emphasizes that some areas ecosystem-based management and can give would benefit more from nutrient abatement further guidance for sustainable use of measures than others. Case study (3) marine areas. Finally, efficient management demonstrated that marine minerals, namely of marine areas requires the integration of ferromanganese concretions, are more local management actions into wider policy widespread than previously anticipated. As processes. Ecosystem-based marine spatial concretions hold high quantities of minerals planning needs to adopt place-based targeted by the emerging seabed mining management strategies and decisions that industry, there may be economic are actionable at various spatial scales and opportunities for such extraction activities can be implemented locally.

Keywords: ecosystem-based management, spatial prioritization, statistical modelling, species distribution modelling (SDM), seascape ecology, Marine Protected Areas (MPAs), systematic conservation planning (SCP), hypoxia, ferromanganese concretions

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Tiivistelmä

Meriekosysteemien tila heikkenee kehitettyjen työkalujen käytettävyyttä kiihtyvällä tahdilla ympäri maailman. Jo nyt meriympäristössä. kaikissa maailman merissä monimuotoisuus Tapaustutkimukset sijoittuvat pohjoi- hupenee, populaatiot pienenevät, vieraslajit selle Itämerelle, jossa meriympäristössä leviävät ja lajien yhteisörakenteessa kertaantuvat paineet uhkaavat meriluonnon tapahtuu muutoksia. Näitä haasteita monimuotoisuutta. Tutkimukset nojaavat ratkotaan monella eri poliittisella tasolla, ja laajaan vedenalaiseen inventointiaineistoon meren hyvälle ekologiselle tilalle pyritään 140,000 näytepisteeltä, jotka on kerätty asettamaan selkeitä tavoitteita. Monet Suomen vedenalaisen meriluonnon direktiivit, säädökset ja linjaukset ovat monimuotoisuuden inventointiohjelmassa globaaleja ja alueellisia, vaikka (VELMU). Tilastollista mallinnusta meriekosysteemejä kohtaavat uhat, käytettiin tapaustutkimuksissa (1) ja (4), erityisesti rannikolla, ovat hyvin paikallisia. joissa mallinnettiin lajien levinneisyyttä, ja Parhaiden tulosten saavuttamiseksi edelleen tapaustutkimuksissa (2) ja (3), direktiivien ja säädösten toimeenpanon joissa kuvattiin vastaavasti hapettomuuden pitäisi olla paikallisesti suunniteltuja ja todennäköisyyksiä ja mereisten mine- huolellisesti kohdennettuja siten, että raalien, rautamangaanisaostumien esiinty- merien käytön kustannustehokkuus ja mistä. Lisäksi tapaustutkimuksessa (1) meriekosysteemien säilyvyys voitaisiin tunnistettiin suojelulle tärkeitä alueita turvata. spatiaalisen suojelupriorisoinnin avulla. Tässä väitöskirjassa osoitetaan neljän Tulosten perusteella nykyiset tapaustutkimuksen keinoin, miten paikal- merisuojelualueet jättävät melkein kolme lisesti räätälöidyt spatiaaliset analyysit neljäsosaa ekologisesti merkittävien lajien voivat tukea tehokasta meren suojelua ja esiintymisalueista suojelematta. Tämä hallintaa: (1) paikallistamalla suojelun korostaa tarvetta kehittää edelleen nykyistä avainalueet, (2) osoittamalla alueet merensuojelualueiden verkostoa sekä tehokkaalle ravinteiden vähentämiselle, (3) aluesuunnittelun roolia toimintojen tunnistamalla kohteet mereisten mineraalien sijoittelussa merialueilla. Suojelematta louhinnalle ja (4) arvioimalla mahdolliset jääneitä alueita voidaan suositella muutokset avainyhteisöissä heikkenevän suojeltavaksi yksityisillä suojelualueilla ja meren tilan myötä. Tämä väitöskirja ottaa huomioon meren kestävässä käytössä. osoittaa, miten laajat aineistot ja spatiaaliset Toisessa tapaustutkimuksessa tunnistettiin analyysit yhdessä poikkitieteellisen luonnollisesti hapettomia alueita, lainaten merentutkimuksen kanssa voivat tukea käsitteitä ja menetelmiä maisema- meren suojelua ja ekosysteemilähtöistä ekologiasta. Tällä lähestymistavalla voidaan meren käytön hallintaa, ja mikä rooli merien kohdentaa toimenpiteitä ravinteiden käytön suunnittelulla on meren vähentämiseen alueille, joista niistä on monimuotoisuuden ylläpitämisessä. eniten hyötyä, ja toisaalta selittää miksi Väitöskirjan tavoitteena on myös arvioida jotkin alueet ovat immuuneja jo tehdyille alun perin terrestriselle puolelle vähentämistoimenpiteille. Neljännessä 4 tapaustutkimuksessa myös esitettiin, miten toisaalta alueille, joilla saostumia esiintyy eri alueet reagoivat eri tavoin ravinteiden runsaasti. vähentämiseen johtaviin toimenpiteisiin. Tapaustutkimukset 1–4 voivat tukea Kolmannessa tapaustutkimuksessa havain- päätöksentekoa, jotka liittyvät meren nollistettiin, miten mereiset mineraalit, tässä suojeluun ja ekosysteemilähtöiseen meren esimerkkinä rautamangaanisaostumat, ovat kestävään käyttöön. Merialueiden käytön huomattavasti laajemmalle levinneitä kuin tehokas hallinta vaatii myös aiemmin on luultu. Koska saostumat paikallistoimien integrointia laajempiin sisältävät suuria määriä kaivosalan politiikkaprosesseihin. Ekosysteemiläh- tavoittelemia mineraaleja, mereiselle töisen merialuesuunnittelun pitää omaksua kaivostoiminnalle saattaa olla tulevaisuu- strategioita ja päätöksiä, jotka ovat dessa taloudellisia edellytyksiä Itämerellä. toteutettavissa monella eri mittakaavan Tapaustutkimukset (1) ja (3) voivat ohjata tasolla, ja joita voidaan soveltaa paikallisesti mereistä kaivostoimintaa ekologiselta erilaisilla alueilla. kannalta vähiten arvokkaille alueille, ja

Asiasanat: ekosysteemilähestymistapa, meren käytön hallinta, spatiaalinen priorisointi, tilastollinen mallinnus, lajien levinneisyysmallinnus, merimaisemaekologia, merisuojelualueet, systemaattinen suojelusuunnittelu, hypoksia, rautamangaanisaostumat

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Acknowledgements

First and foremost, I would like to thank my color/graphics visualizations, Kirsi for your supervisor Atte Moilanen. As you Atte said know-how in, well, almost everything, Ale it: always collaborate with people smarter for your expertise in marine geology, and than you, it will take you a long way. And lastly, Kari and Sofia for your collaboration indeed, it has. Thank you Atte for your in remote sensing. A huge, collective thanks consistent support, guidance, mentorship goes also to various people in several and encouragement throughout this institutes, whom I´ve had the pleasure to dissertation (and beyond) and thank you for work with for the past few years. your friendship. I am also grateful to the pre-examiners Secondly, I would like to express my of this thesis, Professor Erik Bonsdorff and sincere gratitude to all the people who have senior research scientist Matt White, and to collected VELMU data for all these years, the thesis committee members, Tuuli especially in Parks and Wildlife Finland. Toivonen and Laura Uusitalo. Without your field efforts this thesis would In addition, I would like to thank several not exist. In addition, I would like to colleagues in the SYKE Marine Research acknowledge the whole VELMU project Centre for (non-)professional discussions, group, and all the people who have been and especially colleagues in the MMK unit involved in the project, in some way or for your support and company. A special another, irrespective of institute in question, thanks, asanteni, goes to the “Six Degrees and especially Markku Viitasalo, Wilma South” group – some moments to Viljanmaa and Penina Blankett for the remember! A huge thanks goes also to the VELMU coordination. unofficial support groups, namely “Cat I would like to thank my co-authors, therapy”, “Sister and her brother”, “Gangs particularly Make for your idealism, of Vaskio” and last, but not least, the enthusiasm and optimism (even criticism), shadow division of “Marine scientists above which have positively influenced the the water”. somewhat pessimistic side of me as a Lastly, I would like to acknowledge researcher. I would like to say a special financial support from the Finnish Inventory thank you to Juho for your cool head, Programme for the Underwater Marine tenacious attitude towards data Environment (VELMU), funded by the management, and your interest in model Ministry of the Environment and the optimization – it truly was a pleasure to SmartSea project (grants 292985 and work with you. I wish to also thank Antonia 314225) funded by the Academy of Finland for interesting science- and not-so-science- Strategic Research Council. related conversations (e.g. in Pharmarium and thereabouts), Alf for delightful discussions regarding benthic fauna and hypoxia (4.6 is the answer to everything), Laura for your inspirational energy towards science and your excellence in

Contents

Abstract ...... 2 Tiivistelmä ...... 4 Acknowledgements ...... 6 List of original publications ...... 8

1 Introduction ...... 10 1.1 Pathways to sustainable use of marine areas ...... 10 1.2 The Baltic Sea – multiple pressures ...... 12 1.3 Seascape conservation and ecosystem-based marine management ...... 13 1.3.1 Context of case study 1: Locate key areas for conservation ...... 13 1.3.2 Context of case study 2: Indicate areas for effective nutrient abatement ...... 14 1.3.3 Context of case study 3: Identify areas for marine mineral extraction...... 15 1.3.4 Context of case study 4: Consider expected change in key communities to adjust mitigation measures ...... 15 1.4 Support for seascape conservation and ecosystem-based marine management: spatial analyses ...... 16 1.5 Aims of this thesis ...... 17

2 Materials and methods ...... 20 2.1 Study area ...... 20 2.2 Data ...... 21 2.2.1 Data from below the surface ...... 21 2.2.2 Predictor variables ...... 23 2.2.3 Anthropogenic stressors ...... 23 2.3 Data pre-processing and modelling ...... 25 2.4 Spatial conservation prioritization ...... 27

3 Results and discussion ...... 29 3.1 Key areas for conservation ...... 29 3.2 Indicating areas for effective nutrient abatement ...... 31 3.3 Identifying locations for resource extraction ...... 33 3.4 Potential future changes in key communities ...... 35 3.5 Uncertainties and methodological challenges...... 38

4 Conclusions and future perspectives ...... 40 4.1 Applicability of results ...... 40 4.2 Spatial analyses in the marine realm ...... 42 4.3 Future perspectives ...... 43

References ...... 44

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

List of original publications

This thesis is based on the following publications:

I. Virtanen, E. A., M. Viitasalo, J. Lappalainen, and A. Moilanen (2018). Evaluation, gap analysis, and potential expansion of the Finnish marine protected area network, Frontiers in Marine Science, 5, 402, doi:10.3389/fmars.2018.00402.

II. Virtanen, E. A., A. Norkko, A. Nyström Sandman, and M. Viitasalo (2019). Identifying areas prone to coastal hypoxia – the role of topography. Biogeosciences 16, 3183–3195, doi:10.5194/bg-16-3183-2019.

III. Kaikkonen*, L., E. A. Virtanen*, K. Kostamo, J. Lappalainen, and A. T. Kotilainen (2019). Extensive coverage of marine mineral concretions revealed in shallow shelf sea areas. Frontiers in Marine Science 6, 541, doi:10.3389/fmars.2019.00541. *These authors contributed equally to this work

IV. Lappalainen, J., E. A. Virtanen, K. Kallio, S. Junttila, and M. Viitasalo (2019). Substrate limitation of a habitat-forming genus Fucus under different water clarity scenarios in the northern Baltic Sea. Estuarine, Coastal and Shelf Science 218, 31– 38, doi:10.1016/j.ecss.2018.11.010.

The publications will be referred to in the text by their roman numerals.

Author´s contribution I II III IV Original idea EV, AM, EV, MV EV, LK, EV, MV, JL MV KK, AK

Analyses EV, JL, AM EV, ANS EV, LK JL, EV, SJ, KYK

Manuscript EV, MV, JL, EV, MV, EV, LK, JL, EV, MV, preparation AM ANS, AN KK, AK, JL SJ, KYK

EV=Elina Virtanen, AM=Atte Moilanen, MV=Markku Viitasalo, JL=Juho Lappalainen, LK=Laura Kaikkonen, AN=Alf Norkko, ANS=Antonia Nyström Sandman, KK=Kirsi Kostamo, AK=Aarno Kotilainen, KYK=Kari Y. Kallio, SJ=Sofia Junttila

Abbreviations

BSAP Baltic Sea Action Plan BRT Boosted Regression Trees CBD Convention on Biological Diversity EBM Ecosystem-Based Management GES Good Ecological Status (WFD) GES Good Environmental Status (MSFD) HD Habitats Directive HELCOM Baltic Marine Environment Protection Commission IUCN The International Union for Conservation of Nature MPA Marine Protected Area MSFD Marine Strategy Framework Directive MSP Marine Spatial Planning MSPD Maritime Spatial Planning Directive SDM Species Distribution Modelling WFD Water Framework Directive Zeu Euphotic depth

List of figures

Fig 1 Case studies I–IV in the northern Baltic Sea, page 20 Fig 2 Density of inventory sites of the Finnish Inventory Programme for the Underwater Marine Environment (VELMU) 2004–2019, page 22 Fig 3 The current Marine Protected Areas (MPAs) and suggested MPA expansion candidates, page 30 Fig 4 Potential distribution areas of Fucus spp. under different water clarity scenarios, page 36

List of tables

Table 1 Predictor variables developed for modelling species and concretion distributions, and hypoxia probabilities, page 24 Table 2 Present euphotic depth and required change needed to achieve good ecological status (GES) as defined by the Water Framework Directive (WFD), page 37

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

1 Introduction

Around the world, marine ecosystems are now not only feasible but also imminent deteriorating at an unprecedented rate (Dunn et al. 2018). Intact seascapes, or (Worm et al. 2006, Halpern et al. 2019). “marine wilderness” areas, are found only in Loss of biodiversity, population declines, less accessible areas, at high seas and invasions of non-indigenous species, and extreme latitudes (Jones et al. 2018). changes in community composition are In order to control these ecologically apparent in all marine ecosystems (Halpern harmful processes and to steer the use of et al. 2008, Halpern et al. 2019). Moreover, marine resources in a sustainable way, a changing marine environment rearranges necessary actions should be sought at food-webs and shifts distribution ranges of global, regional and local levels to curb various species (Sunday et al. 2012, Rocha negative environmental trends in marine et al. 2015, Molinos et al. 2016). Direct and ecosystems. indirect causes for the degradation include (and are not limited to) fisheries exploitation 1.1 Pathways to sustainable use (Jackson et al. 2001), physical habitat of marine areas destruction/alteration (Lotze et al. 2006, Airoldi et al. 2008, van Denderen et al. Steps have already been taken at multiple 2019), pollution (Islam and Tanaka 2004), levels of management to improve the status ocean acidification (Fabry et al. 2008), of the marine environment. The idea of eutrophication (Crain et al. 2009, Reusch et sustainable use of marine areas – and in al. 2018), hypoxia (Breitburg et al. 2018) general marine management – is to protect and global warming (Harley et al. 2006, and enhance marine biodiversity, and to Poloczanska et al. 2016, Jonsson et al. ensure the delivery of ecosystem services 2018). for the benefit of the society (Elliott 2011). As the anthropogenic capacity to Good Ecological Status (GES) of marine industrialize and economize the ocean waters supports the capacity to deliver grows, increasing human activities in the ecosystem services, which translates marine realm are posing severe threats to directly to economic benefits (Nieminen et marine ecosystems (Halpern et al. 2019). al. 2019). Decline of land-based resources acts as the In Europe, the management of aquatic catalyst for commercial interests on marine environments is orchestrated by various materials, food and space (Lester et al. 2018, directives. The cornerstone of conservation Nyström et al. 2019). Shallow, coastal areas is the Habitat Directive (HD) (Directive are shaped by various human activities and 92/43/EEC), which aims to protect habitats recent technological advances have (Annex I) and species (Annex II) that are propelled the exploitation of even the most either biogeographically unique or in danger remote parts of the ocean (Ramirez-Llodra of disappearing. Areas are designated under et al. 2011). Oceans have become a new protection in the Natura 2000 network based economic frontier, and costly endeavours, on the listed habitats and species in annexes such as mining of deep-sea minerals, are I and II (Evans 2012). The objective of the 10

Water Framework Directive (WFD) considers only certain indicator species for (Directive 2000/60/EC) is “good ecological determining GES, and lacks holistic status” of the European surface waters, ecosystem indicators, and HD focuses on calling for mitigation of eutrophication. The certain species and habitats only, which do Marine Strategy Framework Directive not necessarily indicate a well functioning (MSFD) aims to achieve Good marine ecosystem (Moss 2008, Voulvoulis Environmental Status (GES) of the EU´s et al. 2017). A framework that considers marine waters by 2020 (Directive ecosystems in a holistic way and integrates 2008/56/EC). MSFD is also the first ecological and socio-economic objectives legislative instrument that ensures the into management is needed (Rouillard et al. protection of marine biodiversity in its 2018). entirety (MSFD 2012). The overall goal is to Implementation of environmental and maintain marine biodiversity, regulate water policies has been promoted with the human activities and to ensure the concept of Ecosystem-based Management sustainable use of marine areas. On a (EBM) (or ecosystem approach to regional level, the Baltic Sea Action Plan management). There is no single definition (BSAP) integrates diverse management of EBM, but it constitutes of policies and measures to restore the good ecological management actions aiming to restore and status of the marine environment by 2021, enhance ecosystem health and resilience, set by a regional sea convention, the Baltic and to conserve biodiversity, while at the Marine Environment Protection Commis- same time delivering the services, goods and sion (HELCOM) (HELCOM 2007). benefits required by the society (Atkins et al. Although the HELCOM BSAP goals are 2011, Rouillard et al. 2018). EBM does not broader, the ecological objectives are just strive to define management strategies similar to MSFD descriptors, and thus can for certain components of the ecosystem, support the corresponding environmental but for the entire ecosystem. actions of MSFD (de Grunt et al. 2018). In Claims on marine space – driven by the 2014, EU adopted the Maritime Spatial need for food, materials, resources or Planning Directive (MSPD) (Directive infrastructure – require clear spatial visions 2014/89/EU), designed to support the on how activities should be distributed in implementation of MSFD, and urged the order to maintain and manage marine member states to develop transparent ecosystems. Marine Spatial Planning (MSP) marine spatial plans by the end of 2021 is a process where overlapping interests of (MSPD 2014). different stakeholders are coordinated and However, these nature, water and tied together to make well informed marine directives have not been successful decisions for the sustainable use of marine in halting the declining trend of the state of resources and conservation of marine marine ecosystems (EEA 2015). One reason biodiversity. MSP integrates in a holistic is that the water and nature directives do not manner marine governance instruments target the structure and functioning of the related to the use of sea space from various whole marine ecosystem or overall sectors (Douvere 2008, Ehler 2009). While biodiversity. For instance, the WFD it is generally accepted that EBM needs to

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 be integrated to MSP in order to achieve by various anthropogenic activities, such as both ecological and socio-economic objecti- infrastructure development, commercial ves, various environmental problems are fishing and maritime traffic, which can lead still being tackled separately (Elmgren et al. to marked changes in species richness and 2015), and decisions about the allocation of community composition (Korpinen et al. marine space are based on single-sector 2012, Sundblad and Bergström 2014, objectives (Douvere 2008). Bringing all Sagerman et al. 2020). Interests of economic sectors together, EBM-MSP can form a sectors are also on the rise related to, for mechanism for cross-sectoral collaboration, instance, marine mineral resource extrac- integrating conflicting requirements of tion, which can have negative impacts on various stakeholders, without jeopardizing marine ecosystems, especially in shallow the protection and condition of marine water environments (Kaikkonen et al. ecosystems (Bigagli 2015, Jones et al. 2018). 2016). The Baltic Sea has also experienced offshore and coastal ecosystem-level changes, the disappearance of top predators 1.2 The Baltic Sea – multiple and macrophytes, and altered foodwebs, pressures driven by detrimental human activities, such The Baltic Sea is a semi-enclosed, shallow as overfishing and coastal eutrophication coastal sea with steep vertical and horizontal (Torn et al. 2006, Österblom et al. 2007, environmental gradients. The basin is young Casini et al. 2008, Moksnes et al. 2008, from the geological and ecological Eriksson et al. 2011). Moreover, rapid perspective, and post-glacical processes are colonization, invasion, and expansion by still undergoing (Leppäranta and Myrberg non-indigenous species has altered the 2009, Snoeijs-Leijonmalm et al. 2017). ecosystem function and composition The Baltic Sea hosts a relatively small (Norkko et al. 2012, Jormalainen et al. 2016, variety of species of marine and freshwater Kotta et al. 2016). origin, of which only a few are endemic Projected environmental changes (Bonsdorff 2006, Ojaveer et al. 2010). further imply declining salinity levels, Currently, the Baltic Sea suffers from warming, and a worsening eutrophication eutrophication and increasing anthro- status (Meier et al. 2011a, Meier et al. pogenic disturbance (Vahtera et al. 2007, 2011b, Meier et al. 2012a, Meier et al. Conley et al. 2011, Korpinen et al. 2012, 2014). Such drastic changes, if realized, will Sundblad and Bergström 2014, Andersen et have profound effects on the distributions of al. 2015, Andersen et al. 2017). Hypoxia is various species, which already live at the also one of the well-known problems of the limits of their environmental tolerance Baltic Sea, occurring in central deep basins (Vuorinen et al. 2015, Takolander et al. and in coastal zones, enhanced by the recent 2017a, Jonsson et al. 2018, Kotta et al. pace of excess anthropogenic nutrient 2019). Although large uncertainties in such loading (Conley et al. 2002, Conley et al. projections remain, rigorous adoption of 2011, Jokinen et al. 2018). BSAP measures would lead to improved In addition, the Baltic Sea is impacted environmental status of the Baltic Sea,

despite the negative effects of climate of different policies that aim to mitigate change (Meier et al. 2018, Saraiva et al. anthropogenic pressures. 2019, Wåhlström et al. 2020). Together the In this thesis, four case studies are history and future place the Baltic Sea under developed which demonstrate how spatially multiple threats, with cascading and explicit analyses can support seascape interacting effects on marine ecosystem conservation and effective management (Korpinen et al. 2012, BACC 2015). This actions. Motivations, challenges addressed, challenge requires cross-border and solutions suggested in the case studies management strategies, as well as are briefly explained below. integrative, local management actions. 1.3.1 Context of case study 1: Locate 1.3 Seascape conservation and key areas for conservation ecosystem-based marine A key aspect in safeguarding marine management biodiversity is the designation of Marine Various policies and directives are operated Protected Areas (MPAs). MPAs contribute and implemented at the regional level, with, to EBM and are perceived as an optimal way for instance, targets set for the entire Baltic to safeguard marine biodiversity (Lester and Sea, or for individual basins, such as the Halpern 2008, Edgar et al. 2014). Especially Baltic Proper or the Gulf of Finland. no-take reserves have proven to support However, problems may be more localized marine biodiversity and ecosystem in many coastal sea areas: for example functionality (Halpern and Warner 2002, nutrient discharges can sometimes be traced Lester et al. 2009, Halpern 2014). The to a certain point-source (HELCOM 2018a), design of MPAs must be ecologically sediment loads can be linked to a certain efficient to ensure the implementation of dredging site (Bolam et al. 2006, Fettweis et various conservation objectives, set by al. 2011), and resuspension from international policies (Edgar et al. 2014). recreational boating can impact a single bay International and regional agreements (Sagerman et al. 2020). require nations to designate areas under To reach effective and cost-efficient protection, and for instance the Natura 2000 outcomes, implementation should be network aims to protect key habitats and carefully targeted at the local level to threatened species. The MSFD also states maximize benefits for the marine that marine biodiversity should be protected ecosystem. The extent of management and maintained (MSFD 2012). In 2010, actions required depends on the scale of Convention on Biological Diversity (CBD) activities and processes causing problems adopted a strategic plan to safeguard for marine ecosystems, as well as on the biodiversity, known as the Aichi target, physical complexity of the area in question. which stated that: “By 2020, at least 17% of Thus, management measures should be terrestrial and inland water and 10% of tailored and optimized to effectively tackle coastal and marine areas, especially areas local challenges, and spatially explicit of particular importance for biodiversity solutions should be sought to reach the goals and ecosystem services, are conserved

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 through effectively and equitably managed, outside the current MPAs could be ecologically representative and well- identified to well-informed expansion, to connected systems of protected areas and reach ambitious goals of, for instance, CBD other effective area-based conservation post-2020 biodiversity strategy (EEA 2020). measures (OECMs) and integrated into the wider landscape and seascape” (CBD 1.3.2 Context of case study 2: Indicate 2010). areas for effective nutrient Conservation should thus be abatement implemented through a network of ecologically coherent, well-managed and The main goal of MSFD has been the GES connected MPAs, and designated areas of marine waters by 2020, which, based on should be qualitatively and quantitatively the current knowledge (e.g. Korpinen et al. adequate and representative (CBD 2010, 2018), will not be reached. One of the main HELCOM 2010, 2016). The post-2020 targets of MSFD in the Baltic Sea and the Global Biodiversity Framework by CBD is HELCOM BSAP has been the reduction of expected to scale up conservation efforts, eutrophication and resulting hypoxia. and call for (up to) 30 % protection of land Biogeochemical processes contributing and sea areas by 2030, as it is evident that to hypoxia formation are well-known and the conservation goals set in 2010 will not are often associated with high be reached by 2020 (EEA 2020). anthropogenic nutrient loading and high Having a functioning network of MPAs primary productivity as well as strong presupposes that key areas are conserved. temperature or salinity stratification However, designation of MPAs is not (Bonsdorff et al. 1997, Conley et al. 2011). necessarily based on site-specific know- Nutrient loading and hypoxia are connected ledge of habitats and species, and can rely through internal loading of nutrients from on ad hoc decisions (Agardy et al. 2011). anoxic sediment, creating a vicious circle of Furthermore, conserving only certain eutrophication (Vahtera et al. 2007). habitats or individual species at the expense Moreover, physical conditions, such as of overall marine biodiversity does not complexity of coastal areas or heterogenous guarantee the long-term persistence or stabi- archipelago limiting lateral movement of lity of ecosystems (Stevens and Connolly water, often create opportunity for hypoxia 2004, Jackson and Lundquist 2016). to develop (Conley et al. 2009, Rabalais et Unfortunately, data of sufficient breadth and al. 2010, Breitburg et al. 2018, Fennel and quality for competent evaluation of the Testa 2019). Ecological consequences of success of MPAs has been largely missing, lack of oxygen vary from dysfunctioning and consequently analysis paths for MPA benthic communities to mass mortality of evaluation have been variable. benthic animals (Vaquer-Sunyer and Duarte Suitable tools combined with solid data 2008, Norkko et al. 2015, Gammal et al. can enable the estimation of the ecological 2017). coherence of MPA networks and the However, challenges remain in identification of gaps in protection. projecting spatial and temporal variability of Moreover, key areas for conservation hypoxia in coastal environments.

Hydrodynamic-biogeochemical models economically most profitable areas, it is have mostly been developed for the entire imperative to identify areas that are Baltic Sea (Eilola et al. 2009, Neumann ecologically sensitive to extraction 2010, Meier et al. 2011a, Meier et al. activities. If extraction of marine resources 2012a). While these models are useful at is steered so that impacts on marine regional and basin scales, their horizontal biodiversity become minimized while resolution (usually 2 to 3 nautical miles) is economic benefits are maintained, the too coarse for guiding effective, local sustainability of future resource utilization management actions in coastal areas, could be improved. especially within archipelago. Finding alternative ways to pinpoint 1.3.4 Context of case study 4: Consider areas prone to coastal hypoxia in coastal expected change in key areas are necessary. If nutrient abatement communities to adjust mitigation measures could be directed cost-efficiently measures to areas most urgently needed – and avoided Both MSFD and WFD call for improved in areas naturally problematic where status of marine waters and aim to control abatement measures most probably fail – eutrophication. The role of MSPD is to environmental and economic benefits could support both directives to achieve their be maximized. objectives (MSPD 2014). The goal of

national marine spatial planning is to 1.3.3 Context of case study 3: Identify identify and evaluate the current needs for areas for marine mineral extraction marine space, and a critical part of the MSP One of the aims of MSPD is to support process is analyzing future conditions (Ehler “Blue Growth”, i.e. sustainable economic 2009). With the projected environmental growth and use of resources in the marine change in the marine environment, areas (MSPD 2014). As the pool of land- integration of the temporal dimension with based resources drains, extraction of sea- spatial aspects would benefit planning. floor materials becomes economically Understanding the consequences of viable (Jouffray et al. 2020). The demand expected changes to future marine for raw materials is on the rise, and untapped ecosystems is necessary, both for spatial mineral potential is of interest to the seabed conservation measures and mitigation of mining industry (Hannington et al. 2017). eutrophication. As habitat-forming species For instance, mineral deposits hold large have an important role in ecosystem quantities of commercially exploitable structure and functioning, assessing the metals, such as iron, manganese and cobalt impacts of environmental change on their (Kuhn et al. 2017). spatial distributions is essential. Scenario- The environmental impacts of seafloor based methods are useful for assessing the mining can be substantial, and planning of effects and intensity of environmental mineral extraction needs to consider not changes, such as consequences of decresing only the actual locations of the resource, but salinity on species ranges (Jonsson et al. also adjacent areas (Kaikkonen et al. 2018). 2018). Thus, in addition to locating the Eutrophication is related to vertical 15

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 light availability in the water column, which use the explanative species-environment in turn influences the maximum depth of models to identify potential distributions in plant growth. For instance, depth- present time and/or similar region, and penetration of Fucus spp., one of the most projected SDMs extend the species- important keystone species in the Baltic Sea, environment relationship to the future has been used as a biological indicator for and/or novel geographies (Araújo et al. ecological status in WFD. Estimating how 2019). increasing turbidity limits occurrences of In the terrestrial realm, the use of SDMs such habitat-forming species is essential for has proliferated for the past decades, and estimating how communities might adapt to SDMs have been used to address a wide mitigation measures, and to focus future array of theoretical and applied questions, management actions to optimal areas. including conservation management (Guisan et al. 2013), climate change impacts 1.4 Support for seascape (and adaptation) (Willis et al. 2015, Hällfors et al. 2016), and risk assessments (Jiménez- conservation and ecosystem- Valverde et al. 2011). However, the based marine management: development of marine SDMs has lagged spatial analyses behind their terrestrial counterparts. There Seascape conservation and sustainable use are various reasons for this, such as of marine areas requires suitable tools, of deficiencies in biological data collection, which many fall under the realm of lack of information about environmental geographic data science. A key class of predictor variables, temporal mismatches methods is statistical modelling, where between environmental and biological data, models are used to explain the relationships sampling biases, or insufficient resolution of between observations and background hydrodynamic/biogeochemical surrogates variables. From an ecological perspective, (Robinson et al. 2011, Robinson et al. 2017). one useful framework is Species Moreover, process-based and monitoring Distribution Modelling (SDM), which studies have a long history in marine combines species observations with science, and small-scale and time series environmental characteristics. SDMs draw analyses have predominated, which has correlative conclusions about a species and contributed to the lack of spatial data in the its habitat (ecological niche) and use that marine realm. information to predict species occurrence The development of geographic marine patterns across landscapes (or seascapes) data science (marine GIS) is only now (Elith and Leathwick 2009). The use of evolving, supported by novel marine SDMs can be roughly categorized to: (1) mapping techniques (Brown et al. 2011), explanation, (2) prediction and (3) and rapid advances in understanding spatial projection. Explanative SDMs investigate patterns, gradients, scales and structures in the statistical relationship of species with its the marine environment and seascape environment and develop hypotheses of the (Pittman 2017). Also, with the rise of MSFD environmental factors that explain the and MSPD – and the economic, social and distribution of the species. Predictive SDMs ecological analyses needed for their 16

implementation – demand for georeferenced In terms of EBM and conservation marine data has increased. This has further planning, widely utilized tools include for promoted the need to formulate a holistic, instance Marxan (Ball et al. 2009) and cross-disciplinary view of the whole marine Zonation (Moilanen et al. 2005), which are ecosystem, where ecological and human capable of identifying priority areas for dimensions become integrated, thereby protected area development. Zonation has supporting for instance, EBM-MSP. also been used, e.g., in ecological impact Only for the last decade (or so) has there avoidance and conflict resolution for been a surge of spatially-explicit studies in renewable energy development (Santangeli the marine environment. Based on a recent et al. 2018), biodiversity offsets (Moilanen review by Robinson et al. (2017), a large et al. 2020) and habitat restoration fraction of marine SDM applications have (Thomson et al. 2009). With the race to concentrated on conservation planning, implement MSFD, there has been a assessing the impacts of climate change and corresponding rush in the development of spread of invasive species, or rather DSTs specific for marine environments traditionally, modelling biogeographical (Stelzenmüller et al. 2013, Pınarbaşı et al. ranges of marine species (Embling et al. 2017). However, a recent review concluded 2010, Verbruggen et al. 2013, do Amaral et that tools are not widely utilized, with al. 2015, Weatherdon et al. 2016, Weinert et explanations varying from the complexity of al. 2016). However, modelling is only the DSTs to the lack of output details (Janßen et first step which needs to be taken before al. 2019). Various spatial methods – integrating knowledge into decisions. although commonly applied in the terrestrial Decision Support Tools (DSTs) have realm – are not always easily adopted to the been developed to inform decision making marine environment, as the transferability of and spatially explicit planning. DSTs can such tools is largely dependent on the integrate large amounts of data, including availability of suitable data. the ecological and societal dimensions, contrast alternative planning options, and 1.5 Aims of this thesis enable the evaluation of effectiveness of different management strategies. For Seascape conservation and ecosystem-based instance, Integrated Valuation of Ecosystem management, also in terms of MSP, requires Services and Tradeoffs, InVEST, quantifies detailed information on ecological, societal ecosystem services produced under and economic factors. One science-related different scenarios (Sharp et al. 2018), the impediment has been the lack of adequate end-to-end ecosystem model Atlantis georeferenced data (Martin and Hall-Arber explores the full spectrum of processes that 2008, Cornu et al. 2014). EBM-MSP is affect natural ecosystems, including mostly about what type of activities can be oceanography, ecology, economy and regulated to occur where and when. As society (Fulton et al. 2011), and the Cumu- marine ecosystems, resources, and human lative Impact Assessment Tool evaluates the activities are inherently place-based, all effects of human activities on ecosystem management decisions and strategies should components (Halpern et al. 2008). be of spatial and temporal nature. Therefore,

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 in order to maintain marine ecosystems in Contributions of studies to this thesis are as good condition a key question is where areas follows: worth conserving are, and where anthropogenic activities – and mitigation measures – Paper I is the first comprehensive estimation should be located. of key marine biodiversity areas in Finland, This dissertation has multiple broad and it synthetizes a large quantity of aims: (1) show how extensive data biological and anthropogenic information. combined with suitable (spatial) analysis The study tests the transferability of can support sustainable, ecosystem-based methods developed in terrestrial realm to marine management; (2) highlight the marine realm with a large quantity of intrinsic part sea governance plays in underwater data, shows an analysis path for sustaining marine biodiversity; and (3) identifying priority areas for conservation, reaffirm the applicability and transferability evaluates the effectiveness of the current of tools developed in the terrestrial realm to MPA network, and suggests optimal MPA marine environments. More specifically, expansion sites. this dissertation seeks to find answers to: Paper II provides a novel way to predict and • How to identify priority areas for identify areas prone to coastal hypoxia, conservation and sustainable sea without data on currents, stratification, governance? (I-IV) biological variables, or complex biogeochemical models. By borrowing concepts • How to determine locations for cost- and methods from landscape ecology, this efficient nutrient abatement measures, study quantifies the facilitating role seafloor maximizing the benefits for the marine complexity has in the formation of coastal environment (II)? hypoxia. The study provides a straight- forward approach for identifying areas cost- • How to recognize areas for the effectively for nutrient abatement measures. economic resource potential of marine minerals while at the same time Paper III uses statistical modelling to avoiding impacts on biodiversity? (I, localize marine resources, applied to the III) estimation of the distribution of ferromanganese concretions. The role of concretions • If management actions prove to be in ecosystem functioning is still unknown, effective – or for some reason fail, how and as concretions hold high quantities of will alternative futures look like, from commercially exploitable metals, they are of the perspective of marine biodiversity? great interest to the mining industry. This (IV) study contributes to the role sea governance has in impact avoidance, and to the steering of the economic usage of marine resources towards sustainability.

Paper IV Demonstrates with scenario modelling how potential future changes will affect key marine communities. This is demonstrated with increasing and diminish- hing water clarity scenarios, as water transparency is one of the most important factors that structure shallow water marine assemblages. How will functionally important keystone species, such as bladderwrack, Fucus spp., respond to changes in light availability, and thus to eutrophication?

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

2 Materials and methods

2.1 Study area mouths, where freshwater enters the sea. Turbidity gradient follows similar patterns, All four case studies are focused on the as transport of dissolved and particulate northern Baltic Sea, covering the territorial organic matter from rivers and high on-site waters and exclusive economic zone of primary productivity elevates water Finland. Case study II also covers the turbidity and limits underwater light Stockholm archipelago (Figure 1). availability in the inner archipelago. In The Finnish marine environment is offshore, outer areas water clarity on characterized by strong environmental average increases with lower primary gradients of salinity, turbidity and exposure. productivity and higher water exchange Surface salinity ranges from 7 PSU in the between adjacent basins. Summertime southwestern, outer archipelago and reaches cyanobacteria blooms may however at times almost zero in the northernmost part of the decrease water transparency also in the Gulf of Bothnia, as well as near the river offshore areas.

Figure 1. Case studies I–IV in the northern Baltic Sea.

Glacial erosion and deposition have species richness, habitat and functional formed the Finnish seabed to be diversity in the Baltic Sea decrease from geologically diverse and patchy, with a south to north, and are higher in shallow heterogeneous mixture of various substrate marine areas, compared to deep, dark types. Glacial and post-glacial sediments seafloors (Bonsdorff and Pearson 1999, consist mainly of till, clays, silts and fine- HELCOM 2012, Viitasalo et al. 2017). grained sediments. The crystalline bedrock can be characterized by tectonic lineaments 2.2 Data and fracture zones, evident for instance in 2.2.1 Data from below the surface the Archipelago Sea, where deep, underwater “canyons” crisscross the seabed Studies I, III and IV utilize data from (Kaskela et al. 2012, Kaskela and Kotilainen underwater inventories by the Finnish 2017). Inventory Programme for the Underwater Finnish marine waters are rather Marine Environment, VELMU. Since 2004, shallow, with a mean depth of only 50 m, VELMU has collected information on with the deepest parts (299 m) located species, communities and habitats using southwest from Åland Islands. The most mainly scientific diving and video northern part, Bothnian Bay, is shallow and observation methods. Visited sites range low-saline, with exposed shorelines and from enclosed, inner archipelago areas to comparatively monotonic geomorphology. exposed sites in the outer archipelago, as Moving south, the Kvarken in the middle of well as deep environments with soft seabed the Gulf of Bothnia acts as a substrates. Inventories have been carried out biogeographical barrier between north and mostly based on random stratified sampling, south. Continuing further south from the although some targeted inventories have Kvarken, salinity levels increase, followed fixed, systematic patterns, for topography and geomorphology becomes instance for the purpose of delineating more complex, and over 50,000 islands dot habitat types of Habitat Directive Annex I the Archipelago Sea (Viitasalo et al. 2017), (Kaskela and Rinne 2018). creating one of the most complex In 2019, ~160,000 sites had already archipelago systems in the world. The been visited (Figure 2). Underwater videos southern part, Gulf of Finland, resembles form the bulk of the data; ~100,000 sites, geomorphologically the Archipelago Sea, explored with drop-video or remotely and is also heavily burdened with operated vehicle, 60,000 sites dived, and eutrophication, human-induced pressures, additional ~10,000 locations investigated and hypoxia (Raateoja and Setälä 2016, with other methods (fish larvae sampling Korpinen et al. 2018). Together this sites, benthos and geological sediment geomorphological and environmental samples). complexity creates a variety of habitats for In the scientific diving method, a diver benthic organisms. Benthic communities are observes the coverage (%) of all a mixture of species of freshwater and macrophytes, sessile benthic invertebrates, marine origin and are less diverse than and different bottom substrates along ∼100 “true” marine assemblages. In general, m long dive transects, every horizontal 10 m

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Figure 2. The map on the left shows where VELMU inventories have taken place between 2004 and 2019, represented as density of underwater inventory sites per 10 km2. The upper right panel shows the count of VELMU inventory sites collected from different depth zones, with the two main VELMU methods, scientific diving (Dive) and video observation methods (Video). “Dive” includes all the VELMU inventory methods where species identification is possible to the species level. The lower right panel represent VELMU inventory years 2004–2019 and the count of data collected based on the dive and video inventory methods. or vertical 1 m, from inspection squares of zander), based on VELMU fish larvae 1, 2, or 4 m2. Drop-videos record samplings, was also used in study I approximately 20 m2 of seabed, and (Kallasvuo et al. 2016). In addition, eight coverages of species and seabed substrates Habitats Directive marine habitats are analyzed later from the videos. Overall, associated with “marine environments” this extensive data offers an exceptional were used in study I: Baltic esker islands base for exploring questions regarding (1610), boreal Baltic islets (1620), boreal spatial ecology, conservation science, Baltic narrow inlets (1650), coastal lagoons ecosystem-based management and (1150), estuaries (1130), large shallow inlets changing environment. and bays (1160), sand banks (1110), and In this thesis, VELMU data was used in reefs (1170). Habitats were based on case studies I, IV (species data), and III existing models and expert delineations (ferromanganese concretions). Existing data reported for the EU in 2013 (EEA 2013, on fish reproduction areas (perch, smelt, Rinne et al. 2014, Kaskela and Rinne 2018). 22

2.2.2 Predictor variables 2.2.3 Anthropogenic stressors In the modelling, to draw any conclusions SDMs describe the ecological niche of a about habitat preferences of species, or the species, which is related to environmental conditions where concretions form, tolerances and habitat preferences (section information about the marine environment 1.5). A major challenge is to determine how is required. Information available included, anthropogenic activities (such as coastal for instance, bathymetry, nutrient construction) change the inhabiting concentration, wave forcing, temperature, environment of species, as monitoring data salinity, euphotic depth, oxygen variability before and after the activity is seldom and seabed substrates (studies I, III and IV). available. Moreover, how intensities of In study II, measures describing resulting impacts are defined, causing either seafloor ruggedness and complexity were destruction, degradation or impairment, derived from bathymetry, such as: depends both on species and the habitat in bathymetric position indices (BPI) with question. Therefore estimates of cumulative varying search radii, depth-attenuated wave impacts on marine biodiversity are usually exposure (SWM(d)), topographic shelter based on expert knowledge (HELCOM index (TSI), arc-chord rugosity (ACR) and 2018b). Because of difficulties in vector ruggedness measure (VRM). BPIs quantifying subtle or indirect effects of measures the bathymetric surface ratio human activities on the marine environment, higher/lower in relation to surrounding only activities leading to severe seabed environments, SWM(d) estimates wave modification, i.e. habitat loss and habitat force, TSI differentiates wave directions and degradation, were considered in the spatial sheltering effects of islands, and ACR and prioritization of study I (section 2.4). VRM describe seascape rugosities. Activities categorized as such were capital For the scenario modelling study IV, and maintenance dredging, proximity of euphotic depth (Zeu) – the depth where harbours, and areas reserved for resource radiation has dropped to 1% of the surface extraction and deposition of dredged radiation levels – was derived from Envisat- materials. Data was collated from national MERIS (Medium Resolution Imaging databases and transformed into pressure Spectrometer) satellite images for the layers following Sundblad and Bergström summer periods (May–September) 2003– (2014) with minor modifications.

2011. The calculation of Zeu layer was based on optical models with concentrations of total suspended matter, chlorophyll-a, humic substances as well as sun altitude angle and specific inherent absorption and scattering coefficients. All predictors utilized in studies I–IV are summarized in Table 1.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Table 1. Predictor variables developed for modelling species and concretion distributions, and hypoxia probabilities. Predictor Unit Explanation Study Bathymetry m Depth information I, II, III, IV Bathymetric Position Index An estimate of a higher topographic features than I, II, III Index (BPI) with varying the surrounding environment, search radius 0.1, search radii 0.2, 0.4, 0.8, 2, 4, 10, 20 km Bottom temperature ◦C Temperature (average, min, max) near the seabed I (1 m) and temperature difference during the growing season Bottom and surface PSU Salinity near the seabed (1 m) and in the surface I, III, IV salinity (1 m), corrected with the effects of rivers Chlorophyll a µg l−1 Mean chlorophyll a concentration in surface III waters (0–5 m) during the growing season Colored Dissolved m−1 Yellow substance; optically measurable I, IV Organic Matter (CDOM) component of the dissolved organic matter in the water Depth Attenuated Wave Index Fetch + average wind speed + depth I, II, III, Exposure (SWM(d)) IV Distance to sandy shores m Closest distance to sandy shore I Euphotic depth m Euphotic depth and ± 50 % deviations from the IV present with 10 % intervals Geographical area Index Geographical location of study area as an integer II value value Iron content µg l−1 Cumulative and average concentration of soluble III iron in the water column during 2004–2015 Oxygen variability, mg l−1 Continuous oxygen (average, min) content, I, III frequent and occasional % probability of frequent and occasional hypoxia −1 hypoxia with O2 thresholds 2 and 4.6 mg l Rocky, rock, sandy and % The proportion of rocky (boulders and stones, I, III, IV soft substrates 0.1–3 m), rock, sandy and soft (gravel, sand, silt, mud, clay; <60 mm) substrates Seascape rugosity: arc- Index Both measures evaluate surface ruggedness, ACR II, III chord rugosity (ACR) and using a ratio of surface area, and VRM ratioa of vector ruggedness cell center, local slope and aspect measure (VRM) Secchi depth m Secchi depth I Share of sea proportional % Proxy for the complexity of archipelago; search I, III to land area radius 1, 5, and 10 km Slope ◦ Slope of the seabed I, III Topographical shelter Index Sheltering effect of topography I, II, III (TSI) Total nitrogen and mg l−1 Total nitrogen and phosphorous content in the I, III phosphorous content water column Turbidity FNU Turbidity due to suspended material I

2.3 Data pre-processing and ristic curve (AUC), a measure of detection modelling accuracy of true and false positives and negatives (Jiménez-Valverde and Lobo To generalize the relationship between 2007). AUC values above 0.9 indicate species, hypoxia, and concretions with their excellent, of 0.7–0.9 indicate good and surrounding environments, the modelling below 0.7 indicate poor predictions. method Gradient Boosting Machine and In study II, oxygen profile data was extended functions from Boosted harvested from national, environmental data Regression Trees (BRT) were utilized portals of Hertta (Finland) and SHARK (Friedman et al. 2000, Breiman 2017) (for (Sweden). Only August and September clarity, denoted only as BRT from hereon). 2000–2016 were considered, as seasonal In study I, modelling relied mainly on hypoxia occurs usually in late summer when dive data, and video sites were used only for water temperatures are higher (Conley et al. clearly identifiable species. Additional 2011). Ecologically meaningful limits to national data repository, Hertta, was used hypoxia were defined to be O2 < for modelling invertebrate distributions, and 2 mg L−1 and <4.6 mg L−1. The former is a for modelling macrophyte absences from threshold where coastal organisms start to deep seafloors. Most of the VELMU dive show severe symptoms of oxygen and video data are limited to rather shallow deficiency (Diaz and Rosenberg 1995, Diaz depths (typically 0 to 30 m). Thus, enough and Rosenberg 2008, Vaquer-Sunyer and samples do not exist from deep areas (below Duarte 2008), and the latter has been 50 m). In order to avoid artefacts, a estimated to be a minimum safe limit for randomized absence dataset of benthic species survival and functioning in benthic invertebrate samples (Ekman, Ponar, Van communities (Norkko et al. 2015). Veen and other grab samples for As there exists no reference values for soft sediment sampling) for areas deeper severity of hypoxia based on the frequency than 50 m was utilized during the modelling of hypoxic events, a site was categorized as process. These sites were used only as “occasionally hypoxic”, if it experienced absences in macrophytes models, as habitat hypoxia at least once during the study constraints and lack of light limit the period. If hypoxia was recorded in ≥ 20% of distribution of macrophytes at such depths. the visits, it was categorized as “frequently Randomized subsets of data (50–80%) hypoxic”. This was considered ecologically were used to train the marine SDMs and relevant, as species can develop symptoms tuning of model parameters in general was already from short exposure to oxygen dependent on sample size and the deficiency (Villnäs et al. 2012, Norkko et al. prevalence of species, affecting the choice 2015). The actual oxygen concentrations in of learning rate. Higher tree complexities the sediment, where benthic species live, are required slower learning rates and vice versa anyway probably lower than concentrations (common vs. rare species). Performances of 1 m above the seafloor where the “bottom” SDMs were estimated with deviance water samples were taken. Four hypoxia explained, and the cross validated Area models were trained based on the Under the Receiver Operating Characte- ecologically meaningful thresholds, and 25

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 estimation of model predictive do not exist in the training data. Thus, performances relied on the ability to predictions outside the range where discriminate a hypoxic site from an oxic one observations have been collected (be it and simply with the percent correctly either presence or absence), may be over- or classified (Freeman and Moisen 2008). underestimations (Elith et al. 2010). This In study III, ferromanganese concre- was corrected in study IV with information tion data from VELMU inventories were about historical conditions, or “retrospective used to build models describing concretion environment”. Depth-penetration of Fucus distributions and abundances. For the spp. was remarkably deeper 100 years ago abundance models, all coverages (0–100 %) (Torn et al. 2006). To inform models in the were used in the analyses, whereas in the model building with the past conditions, i.e. distribution models, four coverage the historical depth-penetration of Fucus thresholds were developed, as the detection spp., a subset of presence observations was accuracy may vary depending on the duplicated and used as pseudo-presences. observation method in question. Four Zeu was multiplied by 1.25 and 1.5 to thresholds were: >0.1% (all presence represent same sites as already observed in observations), >10% (abundant concre- the inventories, but with an increased water tions), >50% (substantial cover) and >70% transparency based on historical data. (major concretion fields). Estimation of In studies I–IV models were concretion models relied on AUC and true extrapolated to the full seascape at a skill statistics (TSS) scores (Allouche et al. resolution of 20 m and in studies II and III 2006). For the concretion abundance models spatial predictions were repeated 10 times (percent coverages 0.1–100 %) the with randomly shuffled training datasets. In coefficient of determination (R2) and mean studies II, III and IV, probability absolute error were calculated. predictions were dichotomized into binary In study IV, data pre-processing presence/absence classes. Although this followed similar patterns as in studies I-III. flattens the information content, it also Fucus spp. (F. vesiculosus and F. radicans) facilitates the interpretation of results and is are clearly identifiable species from both needed for management purposes. dives and videos, thus no selection between Dichotomization cut-offs are based on the the two methods were made. However, only confusion matrix, i.e., how well the model a randomly chosen 25% of the targeted captures true/false presences or true/false video inventories was used in the modelling. absences. Usually the threshold is defined to As in study I, to correct the inventory bias maximize the agreement between observed from shallow areas, benthic invertebrate and predicted distributions. Widely used samples from depths 17–286 m were added thresholds, such as 0.5, can be arbitrary to the fitting dataset as known Fucus spp. unless the threshold equals prevalence of absences. presences in the data, i.e., the frequency of Scenario modelling may face a problem occurrences (how many presences of the of “environmental novelty”, meaning that total dataset) (Liu et al. 2005). In study II model extrapolation does not work well if and III, the cut-off was based on an expected future environmental conditions agreement between predicted and observed 26

prevalence and thus represents a analyses. conservative estimate. In study IV the An important first part of Zonation threshold was chosen to deliver equal analyses is assigning weights for features sensitivity and specificity, meaning positive going into spatial prioritization analysis. As observations are just as likely to be wrong as a starting point, features can be equally negative ones (Freeman and Moisen 2008). weighted, although there are several reasons for elevating weights, such as species characterized as ecosystem engineers or 2.4 Spatial conservation species holding economic value (Lehtomäki prioritization and Moilanen 2013). In study I, a In study I, key areas for conservation were hierarchical way of assigning weights was identified with the decision-support tool adopted, in which relative aggregate weights Zonation. Technically, Zonation operates on 3:1:1 were assigned to species, HD habitats high-dimensional spatial data, concerning and habitat types based on 2018 threatened for instance biodiversity features (habitats, status assessment of IUCN Red List of species, ecosystem services), costs, threats, Ecosystems, respectively. Weights were or connectivity (Kareksela et al. 2013, inclined towards species, as the number of Kukkala and Moilanen 2017, Verhagen et species was higher than that of habitats in al. 2017, Virtanen et al. 2020). Zonation the analysis. Negative weights, for features produces a balanced ranking across the thought to impact negatively on the landscape, by iteratively removing cells that ecological value of a site, were assigned to can be lost with smallest aggregate loss for non-indigenous species (e.g. zebra mussel) biodiversity. From a management and marine pressures, such as maintenance perspective, areas receiving high rank dredging and resource extraction. values are key areas from conservation point Zonation requires information about of view - hosting various highly weighted how features are balanced during the and rare species, habitats and habitat types - analysis runs. Aggregation of biodiversity and lowest degraded, pressurized areas, value was done with the additive benefit holding less ecological value, where function, where feature performances are management activities could be directed to, tracked along individual species-area or where human activities could be allowed curves, aiming to minimize aggregate with minimized loss for biodiversity expected extinction risk (Moilanen 2007). (Moilanen et al. 2005, Kareksela et al. This is justified in situations where input 2013). data can be seen to act as surrogates for Zonation was used in identifying key factors not directly represented by available areas for conservation and in evaluating the data. As an output, Zonation produced a ecological coherence of current MPAs, and priority rank map, where cells were ordered further suggesting expansion areas with respect to each other. The ranking does complementing the present MPA network. not quantify solution quality in any absolute Marine SDMs (section 2.3), HD habitats, sense. Rather, directly associated fish reproduction areas, and pressures performance curves summarize the (section 2.2) were used as inputs into the conservation coverage that would be 27

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 achieved in any top priority fraction selected (groups of grid cell) or area networks, to the from the priority rank maps. Key level of individual features (Moilanen and conservation value hotspots were identified Kujala 2014). LSM was used to identify by combining the priority rank outputs, good-quality habitat patches outside the describing the relative ranking balanced existing MPA network and evaluate the across features, and the weighted range-size quality of already established MPAs. Each rarity map, the weighted sum that individual habitat and MPA was evaluated emphasizes locations having many features based on the mean rank – the average of in them. Integration of these two emphasizes pixel-specific rank values from the priority species richness and ecosystem function rank map, and feature density of area i (FDi) compared to the priority rank map. – the feature distribution sum of the area Connectivity is an integral part of divided by the distribution sum expected if spatial analyses, and an important all features were evenly distributed across component of spatial prioritization in the seascape: marine environments, the relevance of it ��∗� �� = , where depending on however, species and the ��∗�� environment in question (Virtanen et al. DSi = distribution sum of focal area i, C = 2020). Connectivity was induced into number of effective cells in the whole analyses using two basic options, matrix seascape, Ai = number of cells in the focal connectivity and edge removal, with the area, and TDS = total distribution sum of all general objective of accomplishing features across the entire study area. aggregation that would facilitate the Finally, illustrative, potential logistics of management decisions. Matrix candidates were identified to complement connectivity identifies and enables the existing MPA network, based on a connectivity of similar and adjacent habitats hierarchic prioritization that specifically (Lehtomäki et al. 2009). A decay distance of accounts for the present MPA network. For 200 m was set for matrix connectivity illustration, potential MPA expansion between different Habitat Directive Annex I candidates were identified taking the highest habitats, elevating priorities of for instance ranked 3 % of areas outside the present reefs and underwater parts of islets. Edge MPA network, then filtering out areas less removal promotes maintenance of structural than 1 km2 in size to emphasize large continuity of prioritized areas, as cells are expansion areas, leading to a proposal for an ranked and removed from the edges of 1 % net expansion of Finnish marine remaining areas. protected sea network. Post-processing options of Zonation were used to estimate the quality of Habitat Directive Annex I habitats and each existing MPAs (HELCOM MPAs, national parks, Nature 2000 sites, nature reserves, private MPAs, Ramsar sites). After prioritization runs, landscape mask analysis (LSM) enables the evaluation of pre-specified areas 28

3 Results and discussion

3.1 Key areas for conservation unconstrained spatial prioritization run, “clean slate solution”, shows where the Case study I identified key areas for highest concentrations of marine conservation, evaluated the ecological biodiversity features are. High priorities coherence of the Finnish MPA network, and correspond to ecologically highly relevant suggested potential expansion candidates to areas, and host comparatively many rare and fix its gaps in protection. For this, SDMs threatened species, functionally important were built for alga, bryophytes, vascular (highly weighted) species and habitats, well- plants and invertebrates, and together these connected habitat complexes, and species- SDMs represent over 200 species and ~100 rich environments. taxa: (i) most common and widespread High priority areas found by this species (e.g., clasping-leaf pond- analysis can be characterized as shallow, weed Potamogeton perfoliatus), (ii) key and diverse environments, with a favourable habitat-forming species (e.g., bladder- amount of light and limited anthropogenic wrack Fucus spp.), (iii) threatened species disturbance. Areas worth mentioning (e.g., Baltic water-plantain Alisma wahlen- include shallow bays and river estuaries in bergii), (iv) rare or sparsely occurring the northern Bothnian Bay, pristine reef species (e.g., eelgrass Zostera marina), (vi) environments in the northeastern parts of non-indigenous species (e.g., zebra Åland main island, sandbanks in the mussel Dreissena polymorpha) and (vii) Archipelago Sea, diverse islet and reef threatened habitat types based on 2018 environments west from the Hanko threatened status assessment based on IUCN Peninsula and species-rich shallow bays in Red List of Ecosystems (e.g., dominating the Gulf of Finland. Establishing a de novo benthic habitats characterized by red algae) MPA network from this “clean slate (Kontula and Raunio 2019). solution”, would lead to high conservation The SDMs performed generally well, gains, as 80 % of the distributions of marine with median deviance explained 71–87 % biodiversity features would become on withheld data and AUC values above 0.7 covered. for all models. Models were based on best However, as MPAs have already been underwater data available and on modelling established in the Finnish sea areas, a more methods prominent in broad SDM literature realistic approach would be the further (e.g. Elith et al. 2010, Robinson et al. 2011, development of the existing MPA network Guisan et al. 2013, Breiner et al. 2015, with highest-quality expansion sites, Morán-Ordóñez et al. 2017, Norberg et al. efficiently filling ecological and 2019). geographical gaps in protection. As it turns The SDMs developed, as well as spatial out, the present MPA network misses out on layers for HD habitats, fish reproduction key species and habitats, as on average only areas and anthropogenic stressors (section 27 % of the distributions of the marine 2.2), were used as input data for identifying biodiversity features are located inside the key areas for conservation. The current MPA network (Fig. 5b in study I). 29

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

This is not surprising, as at the time of MPA In general, new areas were suggested establishments there was limited knowledge further away from areas pressurized by of underwater species and habitats. various human activities, such as cities and An expansion of the MPA network by harbors. Geographically, a large part of the only one percentage point in area (815 km2) individual MPA expansion candidates were would double the mean marine biodiversity identified around the Åland main island, in feature coverage (Figure 6B and areas relatively less impacted by Supplementary Figure S1 in study I). This eutrophication and anthropogenic activities. suggests that well-informed MPA These areas were also identified as expansion has the potential to considerably ecologically relevant in a recent national improve the ecological performance of the survey (Rinne et al. 2019), and sustain, for existing MPA network. An illustrative set of instance, high occurrence rates of Fucus MPA expansion candidates were identified, spp. (Rinne and Salovius-Laurén 2019). complementing the current MPA network: Concentration of new MPA expansion adjacent areas close to current MPAs as well candidates around the Åland main island is as independent new MPAs (Figure 3). further supported by a recent biophysical

Figure 3 The current Marine Protected Areas (MPAs) and suggested MPA expansion candidates. Figure redrawn from study I. 30

modelling study, aiming to maximize of MPAs within the Baltic Sea area. connectivity between HELCOM MPAs, by Surprisingly, only a few prior attempts exist. Jonsson et al. (2020), where MPA expansion For instance, Sundblad et al. (2011) candidates mostly coincide with the results evaluated the ecological coherence of MPAs of this study. Although studies were based based on recruitment habitats of common on different methods and data, this fish species, and Jonsson et al. (2020) compatibility most likely results from the estimate the connectivity of HELCOM fact that high connectivity correlates with MPAs based on biophysical modelling. This biodiversity, as high quality habitats tend to scarcity of MPA research is most likely both export and receive dispersing linked to the lack of detailed data on species propagules (Jonsson et al. 2020, Virtanen et and habitats, and suitable analysis paths for al. 2020). comprehensive evaluation of MPAs, as also A peculiarity for Finland is private suggested by the HELCOM ecological water ownership. A large part of coastal coherence assessment of the Baltic Sea waters is owned by private land owners, as MPAs (HELCOM 2016). well as municipalities, cities and in some Based on the findings of study I the cases private enterprises. As much as 71 % majority of ecologically most important of the MPA expansion candidates are areas is located outside the current MPA located on private waters. With a limited network. Consequently, the role of MSP in amount of state-owned area available for safeguarding marine biodiversity becomes MPA expansion, private marine protection elevated, as decisions on the use of marine as well as “other effective area-based space need to consider important areas conservation measures” (OECMs) should outside legal protection, including many be promoted, to reach the goals of CBD’s privately-owned areas. post-2020 strategy (EEA 2020). Private marine conservation may increase the total 3.2 Indicating areas for effective area under protection, increase nutrient abatement environmental awareness, and enhance the dialogue, and co-operation, between the As biogeochemical modelling of hypoxia is private sector, key stakeholders and challenging in coastal environments, study conservation management. However, II tested if proxies describing seafloor designation and implementation of private complexity could explain the small-scale MPAs depends on the capacity and variation of coastal hypoxia and identify willingness to protect and manage MPAs, locations naturally prone to hypoxia and on the resources of conservation development. The importance of the institutions to monitor the effectiveness of physical morphology of the seabed in private conservation actions (Bottema and hypoxia formation is intuitively obvious and Bush 2012, Farmer et al. 2017, Drescher and has been suggested by several authors (Diaz Brenner 2018). and Rosenberg 1995, Virtasalo et al. 2005, Relying on the extensive VELMU data, Rabalais et al. 2010, Conley et al. 2011), but study I is the first comprehensive has nevertheless not been tested with actual assessment of the ecological effectiveness data.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Recognizing that the enclosed nature of not directly dependent on depth. Hypoxia seafloors facilitates hypoxia formation, was common in shallow and moderate simple topographic parameters were depths of 10–45 m, and, for instance in the developed for hypoxia models in the Archipelago Sea, deep (60–100 m) complex archipelagos of Finland and “channels” are normoxic, as strong currents Sweden. A surprisingly large fraction (∼ 80 keep them oxygenated throughout the year %) of hypoxia occurrences could be (Virtasalo et al. 2005). Instead steep, explained by topographical parameters isolated, and sheltered depressions become alone. Areas identified as prone to hypoxia more easily hypoxic. The relatively high were characterized by low exposure to wave contribution of topographic shelter also forcing, high topographic shelter from indicates that height of islands creates surrounding land areas and isolation from shelters for wind-induced mixing of water, the open sea, all probably contributing to contributing to hypoxia formation. This was longer water residence times in seabed the case for example in the archipelago areas depressions. Deviations from this pattern are of western Gulf of Finland, where high most likely to be caused by directional, islands surround the enclosed water bodies strong currents or by high nutrient loading (Fig. 4, study II). and elevated primary production, either Areas topographically prone to hypoxia improving or worsening the oxygen status, represent less than 25 % of the studied respectively. Major nutrient sources, such seascapes, and were concentrated on the as rivers, cities or intensive agricultural western Gulf of Finland, the Finnish areas, potentially also induce hypoxia Archipelago Sea, the Stockholm formation. However, in extremely complex archipelago and western Gulf of Finland. archipelago areas, such as the ones in These areas are partly isolated from the deep Finland and Sweden, physical factors areas of the central Baltic Sea and are limiting lateral and vertical movement of characterized by complex topography. In water probably facilitate, and in some areas contrast, around 10 % of areas in the eastern even dictate, the development of hypoxia. Gulf of Finland were vulnerable to

The most influential predictors, occasional, moderate hypoxia (O2 < 4.6 mg −1 averaged across models, were depth- L ) but less to severe hypoxia (O2 < 2 mg attenuated exposure (SWM(d)), followed by L−1). This may be at least partly caused by depth, and BPIs identifying wider sinks the intermittent transport of anoxic waters (Fig. 4, study II). This indicates that severe from the central Baltic Sea, along the Gulf oxygen deficiency is more likely to develop of Finland, into the shallow archipelago in sheltered areas, where water movement is areas of the south-eastern Finland (Alenius limited. Such areas are also usually afflicted et al. 2016). by internal loading of phosphorus from Although hypoxic areas represent rather sediments (Puttonen et al. 2014, Puttonen et small geographical entities, even small- al. 2016), although phosphorus can also be sized hypoxic depressions, especially if released from oxic sediments when organic forming a ‘hypoxic network’, releasing matter decomposition is high (Walve et al. nutrients into the water, may degrade the 2018). It is notable that coastal hypoxia was ecological status of a whole coastal area.

Ecological repercussions of even short- energy and nontidal systems, such as large termed hypoxia may be profound to shallow bays and semienclosed sea areas ecosystem functioning (Villnäs et al. 2013). elsewhere in the world. Currently, ecologically most important areas are located in rather shallow 3.3 Identifying locations for environments (study I). These shallow areas resource extraction are vulnerable to projected negative effects resulting from climate change, as water Study III explored the distribution of temperatures are on the rise, thus ferromanaganese concretions, which are at accelerating deoxygenation (Meier et al. the moment a “data deficient” habitat type 2011a, Breitburg et al. 2018). Oxygen in the assessment of threatened habitat types deficiency has also been projected to based on the IUCN Red List of Ecosystems develop faster in shallow, coastal systems (Kontula and Raunio 2019). Moreover, the than in the open sea (Gilbert et al. 2010, ecological role of concretions as a potential Altieri and Gedan 2015). Seasonal hypoxia biogenic habitat remains undecided. may thus become an even more recurrent Concretions are known to be widespread in phenomenon in shallow areas above the the coastal waters, but more research efforts thermocline in late summer. have been invested in deep-sea concretions Results of study II are generally in line (Gazis et al. 2018, Peukert et al. 2018). with prior research, confirming that coastal Concretions are of interest to the seabed hypoxia is a common phenomenon in the mining industry, as they contain Baltic Sea (Conley et al. 2011), but economically valuable and commercially ecologically relevant hypoxia may be more exploitable metals (Hannington et al. 2017). common than previously anticipated. In study III, concretions were found Although extensive biogeochemical models adundantly from almost all sea basins, have been developed for the main basins of forming distinct belts extending from the the Baltic Sea (Meier et al., 2011b, 2012a, Bothnian Bay to the Gulf of Finland. In the 2014), previous estimates of coastal hypoxia Kvarken and the Gulf of Finland, have relied on point observations (Conley et concretions form extensive fields. al. 2009, Conley et al. 2011), as According to the results, at least 11 % of the biogeochemical modelling of hypoxia has Finnish seafloors host suitable its limitations in complex coastal areas. This environments for concretions to form. These study proposes a new approach for findings show a much larger extent of modelling coastal hypoxia, without data on concretions than previously reported currents, stratification, or biological (Glasby et al. 1997, Yli-Hemminki et al. variables, and without convoluted 2016). To put this into geographical context, biogeochemical models, requiring intensive the projected distribution of concretion computational power, especially when run fields is larger than the total coverage of all even on moderate resolution 3D grids. The marine Habitat Directive Annex I Habitats, approach developed here would be useful which jointly cover only 6 % of the Finnish for targeting local nutrient abatement sea area (I). measures and is applicable in other low- Concretions were recorded in depths of

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

0–75 m together with various seafloor types, 2009), and it has been suggested that ranging from mud to rock. Denser concretions in the Gulf of Finland contain concretion fields were mostly related to mid 10 times more phosphorus than anoxic areas depths (in the Baltic Sea context) close to (Savchuk 2000, Lehtoranta and Pitkänen the slopes of the deeper basins. This could 2003). Climate change is projected to be caused by a specific chemical worsen the oxygen status of the Baltic Sea environment prevailing in these areas, such (Meier et al. 2011a), which may have an as hypoxic water originating from the deep effect on the rate how fast concretions basins occasionally flushing the slopes. dissolve, and consequently to the rate of Concretions are also more easily observed in phosphorus release from concretions. areas, where sediment accumulation rates Concretions are reported here to occur are low, and impacts from wave exposure extensively in the Finnish sea areas. high (Glasby et al. 1997, Zhamoida et al. Scattered observations have also been 2007). As this study was based on visual reported on the fringes of deeper basins in observations only, concretions most Sweden, Estonia and Russia, spanning areas probably were not detected in areas where hundreds of kilometers long within the waters are turbid or in environments where Bothnian Sea and the Gulf of Finland sedimentation rates leave concretions (EMODnet 2019). While the results of study buried. This has one important implication: III can not be used to evaluate the vertical if a large part of concretions is buried under thickness of concretion fields, the sheer sediment, concretions may be even more distribution range of concretions may make common and widespread than reported in them rather tempting for economic this study. purposes. Shallow-water concretions are not Frequently hypoxic areas seemed to be yet industrially exploited, but experimental devoid of concretions, whereas the opposite extraction has already taken place in the was observed for areas suffering from eastern Gulf of Finland (Zhamoida et al. occasional, moderate hypoxia (hypoxia 2017). This raises questions about models developed in study II). This can be environmental effects of extensive explained by the fact that in anoxic and exploitation of concretion fields, and other severely hypoxic conditions, concretions types of seabed mining, in particularly dissolve (Zhamoida et al. 2007, Yli- sensitive sea areas such as the Baltic Sea in Hemminki et al. 2016). In contrast, the general. proximity of areas with oscillating hypoxia In order to examine the economic facilitates concretion growth by the resource potential of ferromanganese transport of dissolved nutrients (Glasby et concretions, the biogeochemical and al. 1997). This explains why concretions ecological risks and potential impacts of tend to form on slopes and edges of larger large-scale extraction activities must be depressions bordering anoxic areas, in close assessed (Kaikkonen et al. 2018). proximity to large hypoxic areas with Ferromanganese concretions may serve as potentially high phosphorus releases. biogenic habitats for epibiotic species, but Concretions also deposit high further research would be required. While concentrations of phosphorus (Baturin there are no explicit studies of the

relationship between concretions and (Figure 4). Decrease in water clarity would individual species or biological diversity, it in turn lead to marked losses of Fucus spp. is obvious that they form three-dimensional, distribution. In the most extreme scenario, relatively stable structures, which might be where water transparency decreases by 50 populated by a variety of organisms that %, the distribution extent of Fucus spp. would not occur in areas mostly covered by would narrow down by 59–100 % in the soft sediments. Mining activities could be Kvarken, 55–70 % in the Gulf of Finland, detrimental to the ecological status of such 37–66 % in the Bothnian Sea, and 24–53 % habitats. in the southwestern parts of the Archipelago Sea (Figure 4). Moreover, steep profiles of 3.4 Potential future changes in shorelines and underwater parts of island key communities prevail in some areas of the inner archipelago, such as in the western Gulf of In study IV, potential changes in the Finland, which undermines the horizontal distribution of habitat-forming Fucus spp. expansion of Fucus spp. with increasing was modelled under different water clarity light, compared to gently sloping,

(Zeu) scenarios, deviating from the present illuminated seafloors, which are typical for up to ± 50 % with 10 % intervals. Evaluated instance in parts of the outer archipelago of against validation data, the base model the Bothnian Sea and the Archipelago Sea. performance was good, with AUC 0.924 Achieving GES of surface waters as (SE ± 0.003) and TSS 0.69. Euphotic depth defined by WFD would lead to positive was the most influential predictor (28 %), change in the water clarity, consequently followed by depth (27 %), surface salinity benefiting Fucus spp. and other macroalgae (18 %), unstable seabed substrates (17 %) living on hard substrates. However, large and depth attenuated wave exposure (10 %). variation exists in the eutrophication status In general, although Fucus spp. could between different WFD coastal types. penetrate deeper with increasing water National targets for GES have been set for clarity, the availability of suitable substrates each WFD coastal types (Aroviita et al. limits vertical colonization. Consequently, 2012). To achieve GES, Zeu should increase proportional increases in the horizontal by 7–59 %, depending on the coastal type extent of Fucus spp. are larger in the outer (Table 2). than in the inner archipelago, due to the The largest improvement in water availability of suitable hard substrates. For clarity is required in the Gulf of Finland instance, the southwestern parts of the (45–59 %) and in the Southwestern Archipelago Sea and the outer archipelago archipelago (33–50 %), whereas in the outer of the Gulf of Finland had the greatest parts of the Bothnian Sea and Kvarken, the potential for gaining new Fucus spp. change needed to reach GES would be only distribution areas with increasing water 7–12 %. For instance, the Gulf of Finland clarity, as suitable substrates prevailed and Archipelago Sea suffer from deeper. Changes in the inner archipelago eutrophication and high water turbidity, and were less marked, as the proportional share as a result, Fucus spp. and other macroalgal of soft sediment types becomes higher species are presently not able to utilize the

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Figure 4. Relative potential distribution area of Fucus spp. compared to predicted present 2003–2011 distribution for the WFD coastal water types using three different methods: original area predicted by SDM, substrate correction method and reef layer method. Original SDM predicts area as is, substrate correction applies correction using substrate data from random videos and reef method extracts only Fucus spp. areas that are located on reefs. The scenarios on x-axis present the change of Zeu from the present state in percentages. Figure redrawn from study IV.

full breadth of their potential occurrence Thus, such areas would benefit from area due to water turbidity (Rinne and targeted nutrient abatement measures, since Salovius-Laurén 2019). habitat gains would be largest. This would Decreasing water turbidity has already not only benefit Fucus spp., but also the limited the depth penetration of Fucus spp. flora and fauna associated with Fucus spp. in the western Gulf of Finland, as the lower belts. At the other end of the spectrum is limit of Fucus spp. distribution has shifted Kvarken and the Bothnian Sea, which are in further towards shallow waters, as shown by comparatively good state and suffer less monitoring studies (Ruuskanen 2016). from eutrophication. 36

Table 2. Mean euphotic depth (Zeu) in 2003–2011, required change needed to achieve good ecological status (GES) of Zeu as defined by Water Framework Directive (WFD) and potential Fucus spp. distribution gains (%) if GES of Zeu is achieved. Table modified from study IV.

WFD coastal types Mean GES of Zeu (m) Fucus spp.

Zeu (and change distribution gain (m) needed to achieve (+ %) if GES of

2003 - it as %) Zeu is achieved 2011

Gulf of Finland (inner) 6.0 9.6 (+ 59 %) > 57–125 %

Gulf of Finland (outer) 7.9 11.4 (+ 45 %) 80–106 %

Southwestern archipelago (inner) 6.6 9.8 (+ 50 %) 18–124 %

Southwestern archipelago (middle) 8.9 11.8 (+ 33 %) 80–164 %

Southwestern archipelago (outer) 10.6 14.0 (+ 33 %) 196–302 %

Bothnian Sea (inner) 7.1 9.2 (+ 29 %) 8–69 %

Bothnian Sea (outer) 9.9 10.8 (+ 9 %) 19–24 %

Kvarken (inner) 6.5 7.0 (+ 7 %) 9–25 %

Kvarken (outer) 9.0 10.0 (+ 12 %) 18–26 %

Consequently, greatest distribution (2017b) showed that Fucus spp. may be losses due to worsening eutrophication vulnerable to low salinity, especially if could be seen exactly there, as Fucus spp. subjected to high temperatures even for and other macroalgal species are currently relatively short periods. able to utilize the full breadth of their The necessity to preserve Fucus spp. in potential distribution zone. This is also the Kvarken and Bothnian Sea is further supported by the findings of a recent study emphasized by the recent declines of Fucus by Rinne and Salovius-Laurén (2019), spp. in other sea areas, especially in the where Fucus spp. were found to be in outer Archipelago Sea (Vahteri and relatively good status in the Bothnian Sea Vuorinen 2016), where the potential for and northern parts of the Åland Sea, with Fucus spp. growth may be hampered by the higher occurrence rates and deeper depth- high exposure gradient, grazing pressure by penetration. Idotea balthica, and by competition with It is also notable that Fucus spp. filamentous algae (or a combination of communities in the Kvarken and Bothnian these) (Berger et al. 2003, Nilsson et al. Sea will probably be sensitive to projected 2004, Jonsson et al. 2006). Another oceanographic changes induced by climate plausible reason for the inability of Fucus change (Andersson et al. 2015, Vuorinen et spp. to recolonize its former distribution al. 2015). For instance, Jonsson et al. (2018) areas in the outer Archipelago Sea is limited demonstrated that Fucus spp. habitats are connectivity. High habitat fragmentation expected to shrink dramatically due to and consequent habitat isolation in these declining salinity levels and consequent areas may exceed the relatively short habitat fragmentation, and Takolander et al. dispersal abilities of Fucus spp., which

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 usually are less than 10 km (Jonsson et al. missing environmental predictors (e.g. 2018). wintertime ice scouring leading to habitat Thus, considering these current and destruction), and temporal uncertainty (e.g. future progressions, the most viable habitat outdated species inventories). areas for Fucus spp. populations in the Subjective sampling may have caused future may well be in the Kvarken and in the taxonomic survey errors or biases in the Bothnian Sea, if projected declines in reported percent cover of species (or salinity conditions will not be realized. concretions). This is an intrinsic problem of all underwater inventory programmes operating in aquatic environments, where 3.5 Uncertainties and water clarity challenges visual methodological challenges interpretation. However, in situations where Modelling has always the problem of species identification is not 100 % reliable, uncertainty around it, as no model can fully a diver takes a sample of the species in describe the complexity and dynamics of the question and does the identification later. In natural world. Models are only as good as the case of video data, this is of course not the data underlying them. Especially in possible. Moreover, taxonomic identi- marine environments, assembling a fication to the level of species is not always representative set of reliable species possible from videos (except for occurrence data can be challenging. In this macroscopic species) and reported percent thesis, an unusually large amount of spatial coverages should be interpreted with some data was used for modelling the current and caution. Consequently, only a small part of future distribution of species, occurrences of video data was utilized in models developed ferromanganese concretions and in studies I, III and IV. In study III, probabilities of hypoxia development. subjective sampling uncertainties were Models were based on information sampled addressed by varying percent thresholds of using standardized methods from tens of reported concretions, and by stacking thousands of sites visited, where species (or predictions from replicate data sets used for concretions) were either recorded present model building (as also in study II). (with percent cover assessed) or absent. Locational uncertainties may arise from Thus, the breadth and amount of data was errors in georeferencing, resulting from substantial for developing statistically inaccurate precision of GPS positioning and sound models. Nonetheless, there remain boat movement. Positioning accuracy also various sources of errors inherent in the decreases with depth. This is a problem if data, which should be acknowledged. the maximum error in location where Uncertainties associated with the spatial species is identified exceeds the resolution data arise from interpretation errors (e.g. of environmental predictors. This was not subjective sampling), locational uncertainty the case with VELMU data, as the locational (e.g. inaccurate georeferencing), sampling error does not exceed the predictor grain biases (e.g. fewer samples in deeper, size of 20 m. Locational uncertainties only offshore areas), varying sampling intensities become an issue for fine-scale predictions (e.g. gridded vs. random observations), (e.g. couple of meters), and in models built

with smaller sample sizes (Mitchell et al. military restrictions. The predictors used 2017). here cover a wide breadth of environmental Opportunistic sampling strategies may factors affecting species distribution and not represent the true species-environment habitat preferences (Table 1). Model relationship. VELMU inventories have performances throughout this thesis were mostly followed random stratified sampling high, which also suggests that predictors strategies, except for targeted inventories, work at the seascape scale. and have concentrated less on deep (>50m) The SDMs capture the species- areas. Sampling biases and varying environment relationship in the absence of sampling intensities were dealt with in the disturbances. Coastal areas are mostly modelling by using random subsets of data shaped by various human activities, and by (I–IV), creating pseudo-absences to less pressures they are causing on organisms, visited areas, such as deep (>50m) offshore which are not necessarily captured by the areas (I and IV), by handling sampling abiotic predictors derived from water differences between Sweden and Finland quality monitoring studies. Moreover, (and WFD areas), by treating study areas as species inventories are more inclined separate area in the model building (II), and towards areas with less influence from by addressing spatial autocorrelation by human presence. Emphasis should be placed introducing a residual autocovariate term to on collecting species data from both pristine final models (III) (Crase et al. 2012). and disturbed environments, and on the Temporal uncertainties may have derivation of near real-time (e.g. satellite- compromised model validity in locations based) environmental data. where the suitability of the environment for In a few years´ time, remote sensing species occurrence has changed will probably revolutionize the field of considerably after species was observed. marine ecological modelling in a similar Community compositions, species ranges, manner as in the terrestrial realm, where habitats and environments may also change remotely derived predictors have over time, and thus models may not fully significantly improved understanding of represent the changed conditions. To species distributions (He et al. 2015). High- describe species-environment relationship resolution earth observation missions, such correctly, synoptic high-resolution as Sentinel-2 (10–60 m), already provide environmental (predictor) data, long-term near-real time data on marine areas, and biodiversity monitoring and physiological include variables such as water clarity, experiments, as well as accurate information enabling the development of more refined about how threats (pressures, stressors) marine species distribution models. High modify habitats, would be desirable, but is temporal and spatial resolution of remotely unfortunately rarely available. sensed products will probably also enable Inadequate relevance of available the before/after analyses of intensities and predictor variables may pose challenges for extents of destructive human activities, fine-scale, spatially explicit models. The which will make visible the effects of bathymetry and substrate information are anthropogenic activity on the ecological often inaccurate, due to the lack of data and state of the marine environment.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

4 Conclusions and future perspectives

Seascape conservation and ecosystem-based planning decisions guiding the allocation of marine management require spatially sea area for human activities becomes explicit data of areas worth conserving to elevated. support decision-making. This thesis aimed Study I serves as a basis for identifying to show how extensive data combined with priority areas for spatial management suitable spatial analysis can support measures, including establishment of new seascape conservation and marine MPAs, and demonstrates the contribution of management, and to reaffirm the spatial prioritization to MSP. The results of applicability of spatial analytical methods study I can be and have already been used in developed in the terrestrial realm to marine various ways to promote conservation and environments. In terms of applications, this sustainable use of sea areas. For instance, thesis aimed to identify key conservation the key areas for conservation were used in areas, pinpoint locations for efficient the Finnish version of CBD EBSA nutrient abatement measures and reveal (Ecologically or Biologically Significant areas suitable for economic resource Areas) (Johnson et al. 2018), called EMMA extraction. (ecologically significant underwater marine areas) (Lappalainen et al. 2020), which was further used in the development of national 4.1 Applicability of results marine spatial plans according to the EU In study I, the ecological coherence of Marine Spatial Planning Directive. The Finnish MPA network was evaluated with results of study I could also act as a stimulus spatial prioritization. Current MPAs leave for promoting private MPAs, encourage almost three-quarters of ecologically and private owners to protect their waters, and functionally important species occurrence facilitate the conceptualization of private areas unprotected, as in the past MPAs have marine conservation. been designated without much knowledge In study II, areas naturally prone to of underwater marine life. This suggests that hypoxia were identified using spatial the Finnish MPA network would benefit analyses, borrowing concepts from from further development. Expansion of the landscape ecology. Based on the results, MPA cover by just 1%, from 10 to 11% area seafloor complexity facilitates, and even coverage using ideal expansion candidate dictates, hypoxia development in enclosed, sites would lead to extremely high relative sheltered areas, where lateral movement of conservation gains, as the mean water is limited. Deviations from this conservation coverage of marine pattern are a result of either strong mixing biodiversity feature cover would be doubled due to directional currents or high external (study I, Fig. 6A-B). As the most promising nutrient loading, which may improve or expansion candidates are located on private worsen the oxygen status, respectively. The waters, the need for spatial measures beyond hypoxia modelling approach gives a state governed MPA network expansion is practical baseline for various hypoxia- apparent. Especially the role of spatial related studies and can support 40

biogeochemical hypoxia models. needs thorough investigation, as concretions Developed hypoxia models can be used to may serve as biogenic habitats for various target nutrient abatement measures to species. Only by combining sufficient locations, where they are most likely to be ecological, geological and technological efficient. The results of study II can also knowledge can environmentally sustainable explain why some areas are immune to marine resource governance be achieved. nutrient abatement actions already taken. Study IV showed that some areas would For instance, in areas naturally prone to benefit more from nutrient abatement severe hypoxia, and strong internal loading, measures than others. Although Fucus spp. measures focusing on limiting external could penetrate deeper with increasing nutrient loads may prove futile. In contrast, water clarity, the availability of suitable nutrient abatement could be much more substrates limits vertical colonization in effective in areas burdened by external some areas. Due to the current loading but topographically less prone to eutrophication status in the Archipelago Sea hypoxia. These findings emphasize the role and Gulf of Finland, the most viable of sea governance: how should nutrient populations of Fucus spp. may well be in the abatement measures be targeted cost- future in the Bothnian Sea and in the efficiently, to maximize benefits for the Kvarken, if declines in salinity conditions marine environment? Decisions are are not realized. In these areas decreases in especially needed to conserve the remaining water clarity would lead to marked losses of pristine marine areas and to rehabilitate Fucus spp. and ecological functionality of ecosystems already suffering from the associated communities. This implies eutrophication. that Fucus spp. communities of these Study III demonstrated that northern areas are especially vulnerable to ferromanganese concretions are more further eutrophication, caused by projected widespread than previously anticipated, environmental change. occurring in over 11 % of the Finnish marine Together these studies demonstrate that areas. However, these modelling results are cross-disciplinary spatial analyses can both based on visual inventories, suggesting that support decisions regarding marine concretions can plausibly occur even more conservation and sustainable use of marine widespread than reported here, as areas and can also complement the success concretions can also be buried in sediments. of other modelling methodologies (e.g. Because concretions hold high biogeochemical modelling) in complex concentrations of minerals targeted by the coastal areas. Further, efficient management emerging seabed mining industry, there may of marine areas requires integration of local be economic opportunities for such management actions to wide-ranging policy extraction activities to take place also in the processes. Ecosystem-based marine Baltic Sea. Results of studies I and III could management needs to adopt and implement guide detrimental mining activities to areas place-based management decisions that act holding less ecological value, and to areas at various spatial scales, operating at global where concretions are abundantly found. (international policies and conventions), However, the ecological role of concretions regional (EU directives, HELCOM),

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87 national (laws, decrees) and local (land/sea in the marine realm can be temporally more use zoning) levels. dynamic than in terrestrial environments. For instance, turbidity and salinity near river estuaries are highly fluctuating and depend 4.2 Spatial analyses in the marine on freshwater outflow. In such instances, the realm ecological tolerance of a species to the This thesis has provided verifiable support environmental predictor should be for the fact that SDMs can be highly evaluated over a long period, e.g. from water operational also in the marine realm. quality monitoring surveys or from satellite- However, the success of marine SDMs is derived environmental products. Especially dependent on the characteristics of species for perennial species, the environmental occurrence data available (e.g. density and conditions also outside the main growing design of survey data, detectability of season should be considered in determining species, rarity of species, sample breadth of the niche of the species. the total species range) and on the relevance As a recommendation, time and effort of environmental predictors. Relevant should be reserved to the quest for relevant proximal and distal environmental factors environmental predictors, as in the marine which regulate the occurrence of species realm predictors usually rely on 3D have a larger influence on the success of hydrodynamic-biogeochemical models at a SDMs than the type of modelling method grain and extent size (few nautical miles, chosen, and success of SDMs varies more basin-scale) not necessarily useful for between different species than between developing fine-scale SDMs. For instance, different modelling methods. various 3D models leave out shallow areas Based on the results of this thesis, due to the complexity of coastal and modelling species distributions is feasible in archipelago zones, although these are the marine environments where distinctive areas where various marine organisms live ecological niches, i.e. environmental (I, IV) in close interaction with heterogeneity, confine species occurrences, anthropogenic influences. following partly from strong horizontal SDMs are almost always a means to an (increasing/declining by latitude/longitude), end, not the goal itself, as SDMs are used for vertical (increasing/declining by depth) and instance in conservation planning, risk distance-based gradients (increasing/ assessments, and in understanding species declining by distance). Marine SDMs also invasions and range shifts (Glardon et al. benefit from predictors regulating species 2008, Martínez et al. 2015, Giakoumi et al. distributions at different spatial scales, 2016, Oh et al. 2017, Duarte et al. 2018). varying from local (e.g. substrate type) and This thesis has also illustrated that marine seascape scales (e.g. turbidity) to regional SDMs are useful in conservation planning scales (e.g. salinity). and in the evaluation of the ecological While various environmental predictors coherence of MPAs. Still, analyses could be in the marine realm have terrestrial further expanded by adding information on analogues (e.g. bathymetry model vs. ecosystem services, economics and marine elevation model), environmental predictors connectivity dependent on species traits (cf.

Jonsson et al. 2020). examples of such research, and presently Most importantly, adding accurate and available work has concentrated on timely information of the pressures and modelling range shifts for a small number of disturbances that marine species and species (Jonsson et al. 2018, Kotta et al. habitats face, would improve the accuracy 2019). Modelling climate change impacts of the description of the state of the marine for a broad range of species could provide ecosystem. Currently, most threat layers that important insight into the sensitivity, enter spatial prioritization are constructed resilience, and extinction risk of species in based on expert opinion without support relation to projected changes, as well as to from empirical data, and the effects of potential changes in the functionality of pressures on species and habitats, such as marine ecosystems. intensity and longevity, remain unclear. While SDMs predict the occurrences of Impacts of the pressures resulting from individual species, recent methodological various human activities could be gathered advances have enabled the simultaneous from various sources, such as scientific modelling of joint responses of multiple reviews and meta-analyses, technical species to the environment. One of these reports and environmental impact methods is hierarchical modelling of species assessments. Each pressure and disturbance communities, which integrates (partially layer could be individually coupled to correlated) community-level responses to species and habitats (e.g., via SDMs), due to the environment, information on species the differences in responses of species and traits, biotic interactions and phylogenetic habitat to various pressures. Together these relationships across various spatio- improvements would lead to a much more environmental scales (Ovaskainen et al. realistic depiction of key conservation areas 2017). Topical questions could include: how and would be of utility to decision making similarities between marine communities around spatial conservation and marine depend on environmental similarity and management. geographical distance, or how much variation in marine species communities is explained by species traits and biotic 4.3 Future perspectives interactions across varying spatial scales. The Baltic Sea is changing rapidly, as heat Identification of potential new waves, declining salinity levels, increasing conservation areas and marine biodiversity hypoxia and eutrophication reshape marine prioritization are of use also in ecological ecosystems and habitable environments, as impact avoidance, where ecologically demonstrated by coupled oceanographic- harmful activities are avoided in high- hydrodynamic biogeochemical modelling priority areas and directed to areas of less (Belkin 2009, Meier et al. 2011a, Meier et ecological value. For instance, “inverse al. 2012b, Andersson et al. 2015, BACC spatial conservation prioritization” can be 2015, Humborg et al. 2019). Thus, used to identify potential areas for economic modelling species ranges conditional on development, while at the same time projected change should be a research limiting environmental effects of the priority. Nevertheless, there are scarce development activity (Kareksela et al.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

2013). One of the next continuations from section). Interest in the addition of marine here could be to optimize potential areas for ecosystem services into conservation offshore wind farms, by combining the planning analysis is also apparent. Currently existing underwater knowledge with the the concepts of marine ecosystem services economic feasibility of offshore wind and marine ecosystem accounting are being energy, together with the societal and developed, and certain habitat types and ecological impacts of such infrastructure functions of species groups are being tied to development. specific ecosystem services, after which the Another broadly useful continuation of economic value of the ecosystem service study I would be the inclusion of set of provision can be calculated. Inclusion of species- and habitat-specific impacts of ecosystem services would advance spatial pressure and disturbance layers resulting planning processes and promote sustainable from various human activities (see previous marine management.

References

Agardy, T., di Sciara, G.N. and Christie, P. pressures. Estuarine, Coastal and Shelf (2011) Mind the gap Addressing the Science 161, 88-92. shortcomings of marine protected areas Andersson, A., Meier, H.E.M., Ripszam, M., through large scale marine spatial planning. Rowe, O., Wikner, J., Haglund, P., Eilola, K., Marine Policy 35(2), 226-232. Legrand, C., Figueroa, D., Paczkowska, J., Airoldi, L., Balata, D. and Beck, M.W. (2008) Lindehoff, E., Tysklind, M. and Elmgren, R. The gray zone: relationships between habitat (2015) Projected future climate change and loss and marine diversity and their Baltic Sea ecosystem management. Ambio applications in conservation. Journal of 44 Suppl 3(Suppl 3), 345-356. Experimental Marine Biology and Ecology Araújo, M.B., Anderson, R.P., Márcia Barbosa, 366(1-2), 8-15. A., Beale, C.M., Dormann, C.F., Early, R., Alenius, P., Myrberg, K., Roiha, P., Lips, U., Garcia, R.A., Guisan, A., Maiorano, L., Tuomi, L., Pettersson, H. and Raateoja, M. Naimi, B., O’Hara, R.B., Zimmermann, N.E. (2016) Gulf of Finland Physics: The Gulf of and Rahbek, C. (2019) Standards for Finland assessment. Raateoja, M. and Setälä, distribution models in biodiversity O. (eds), pp. 42-57. assessments. Science Advances 5(1), Allouche, O., Tsoar, A. and Kadmon, R. (2006) eaat4858. Assessing the accuracy of species Aroviita, J., Hellsten, S., Jyväsjärvi, J., distribution models: prevalence, kappa and Järvenpää, L., Järvinen, M., Karjalainen, the true skill statistic (TSS). Journal of S.M., Kauppila, P., Keto, A., Kuoppala, M. Applied Ecology 43(6), 1223-1232. and Manni, K. (2012) Ohje pintavesien Altieri, A.H. and Gedan, K.B. (2015) Climate ekologisen ja kemiallisen tilan luokitteluun change and dead zones. Global Change vuosille 2012–2013− päivitetyt Biology 21(4), 1395-1406. arviointiperusteet ja niiden soveltaminen. Andersen, J.H., Carstensen, J., Conley, D.J., Atkins, J.P., Burdon, D., Elliott, M. and Gregory, Dromph, K., Fleming‐Lehtinen, V., A.J. (2011) Management of the marine Gustafsson, B.G., Josefson, A.B., Norkko, environment: Integrating ecosystem services A., Villnäs, A. and Murray, C. (2017) Long‐ and societal benefits with the DPSIR term temporal and spatial trends in framework in a systems approach. Marine eutrophication status of the Baltic Sea. Pollution Bulletin 62(2), 215-226. Biological Reviews 92(1), 135-149. BACC (2015) Second assessment of climate Andersen, J.H., Halpern, B.S., Korpinen, S., change for the Baltic Sea basin, Murray, C. and Reker, J. (2015) Baltic Sea SpringerOpen. biodiversity status vs. cumulative human Ball, I., Possingham, H. and Watts, M. (2009) Spatial conservation prioritisation: 44

Quantitative methods and computational Breitburg, D., Levin, L.A., Oschlies, A., tools. Moilanen, A., Wilson, K. and Grégoire, M., Chavez, F.P., Conley, D.J., Possingham, H. (eds), pp. 185-195, Oxford Garçon, V., Gilbert, D., Gutiérrez, D., University Press, Oxford, UK. Isensee, K., Jacinto, G.S., Limburg, K.E., Baturin, G. (2009) Geochemistry of Montes, I., Naqvi, S.W.A., Pitcher, G.C., ferromanganese nodules in the Gulf of Rabalais, N.N., Roman, M.R., Rose, K.A., Finland, Baltic Sea. Lithology and Mineral Seibel, B.A., Telszewski, M., Yasuhara, M. Resources 44(5), 411-426. and Zhang, J. (2018) Declining oxygen in the Belkin, I.M. (2009) Rapid warming of large global ocean and coastal waters. Science marine ecosystems. Progress in 359(6371). Oceanography 81(1-4), 207-213. Brown, C.J., Smith, S.J., Lawton, P. and Berger, R., Henriksson, E., Kautsky, L. and Anderson, J.T. (2011) Benthic habitat Malm, T. (2003) Effects of filamentous algae mapping: A review of progress towards and deposited matter on the survival of Fucus improved understanding of the spatial vesiculosus L. germlings in the Baltic Sea. ecology of the seafloor using acoustic Aquatic Ecology 37(1), 1-11. techniques. Estuarine Coastal and Shelf Bigagli, E. (2015) The EU legal framework for Science 92(3), 502-520. the management of marine complex social– Casini, M., Lövgren, J., Hjelm, J., Cardinale, M., ecological systems. Marine Policy 54, 44-51. Molinero, J.-C. and Kornilovs, G. (2008) Bolam, S., Rees, H., Somerfield, P., Smith, R., Multi-level trophic cascades in a heavily Clarke, K., Warwick, R., Atkins, M. and exploited open marine ecosystem. Garnacho, E. (2006) Ecological Proceedings of the Royal Society B: consequences of dredged material disposal in Biological Sciences 275(1644), 1793-1801. the marine environment: a holistic CBD (2010) Aichi Biodiversity Targets. assessment of activities around the England Convention on Biological Diversity. and Wales coastline. Marine Pollution Conley, D.J., Björck, S., Bonsdorff, E., Bulletin 52(4), 415-426. Carstensen, J., Destouni, G., Gustafsson, Bonsdorff, E. (2006) Zoobenthic diversity- B.G., Hietanen, S., Kortekaas, M., Kuosa, H., gradients in the Baltic Sea: continuous post- Markus Meier, H.E., Müller-Karulis, B., glacial succession in a stressed ecosystem. Nordberg, K., Norkko, A., Nürnberg, G., Journal of Experimental Marine Biology and Pitkänen, H., Rabalais, N.N., Rosenberg, R., Ecology 330(1), 383-391. Savchuk, O.P., Slomp, C.P., Voss, M., Bonsdorff, E., Blomqvist, E.M., Mattila, J. and Wulff, F. and Zillén, L. (2009) Hypoxia- Norkko, A. (1997) Coastal eutrophication: Related Processes in the Baltic Sea. Causes, consequences and perspectives in the Environmental Science & Technology Archipelago areas of the northern Baltic Sea. 43(10), 3412-3420. Estuarine, Coastal and Shelf Science 44, 63- Conley, D.J., Carstensen, J., Aigars, J., Axe, P., 72. Bonsdorff, E., Eremina, T., Haahti, B.M., Bonsdorff, E. and Pearson, T.H. (1999) Variation Humborg, C., Jonsson, P., Kotta, J., in the sublittoral macrozoobenthos of the Lannegren, C., Larsson, U., Maximov, A., Baltic Sea along environmental gradients: A Medina, M.R., Lysiak-Pastuszak, E., functional-group approach. Australian Remeikaite-Nikiene, N., Walve, J., Journal of Ecology 24(4), 312-326. Wilhelms, S. and Zillen, L. (2011) Hypoxia Bottema, M.J.M. and Bush, S.R. (2012) The Is Increasing in the Coastal Zone of the Baltic durability of private sector-led marine Sea. Environmental Science & Technology conservation: A case study of two 45(16), 6777-6783. entrepreneurial marine protected areas in Conley, D.J., Humborg, C., Rahm, L., Savchuk, Indonesia. Ocean & Coastal Management 61, O.P. and Wulff, F. (2002) Hypoxia in the 38-48. Baltic Sea and basin-scale changes in Breiman, L. (2017) Classification and regression phosphorus biogeochemistry. Environmental trees, Routledge. Science & Technology 36(24), 5315-5320. Breiner, F.T., Guisan, A., Bergamini, A. and Cornu, E.L., Kittinger, J.N., Koehn, J.Z., Nobis, M.P. (2015) Overcoming limitations Finkbeiner, E.M. and Crowder, L.B. (2014) of modelling rare species by using ensembles Current practice and future prospects for of small models. Methods in Ecology and social data in coastal and ocean planning. Evolution 6(10), 1210-1218. Conservation Biology 28(4), 902-911.

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Crain, C.M., Halpern, B.S., Beck, M.W. and Barrett, N.S., Becerro, M.A., Bernard, Kappel, C.V. (2009) Understanding and A.T.F., Berkhout, J., Buxton, C.D., managing human threats to the coastal Campbell, S.J., Cooper, A.T., Davey, M., marine environment. Annals of the New Edgar, S.C., Forsterra, G., Galvan, D.E., York Academy of Sciences 1162(1), 39-62. Irigoyen, A.J., Kushner, D.J., Moura, R., Crase, B., Liedloff, A.C. and Wintle, B.A. (2012) Parnell, P.E., Shears, N.T., Soler, G., Strain, A new method for dealing with residual E.M.A. and Thomson, R.J. (2014) Global spatial autocorrelation in species distribution conservation outcomes depend on marine models. Ecography 35(10), 879-888. protected areas with five key features. Nature de Grunt, L.S., Ng, K. and Calado, H. (2018) 506(7487), 216-+. Towards sustainable implementation of EEA (2013) Reporting under Article 17 of the maritime spatial planning in Europe: A peek Habitats Directive (period 2007-2012), into the potential of the Regional Sea EIONET. European Topic Centre on Conventions playing a stronger role. Marine Biological Diversity. Policy 95, 102-110. EEA (2015) State of the environment Report Diaz, R. and Rosenberg, R. (1995) Marine 2015 European Environment Agency, benthic hypoxia: A review of its ecological Copenhagen. effects and the behavioural response of EEA (2020) Europe´s state of the environment benthic macrofauna. Oceanography and 2020: change of direction urgently needed to marine biology 33, 245-303. face climate change challenges, reverse Diaz, R.J. and Rosenberg, R. (2008) Spreading degradation and ensure future prosperity. dead zones and consequences for marine Ehler, C.a.F.D. (2009) Marine Spatial Planning: ecosystems. Science 321(5891), 926-929. a step-by-step approach toward ecosystem- do Amaral, K.B., Alvares, D.J., Heinzelmann, L., based management. 53. Borges-Martins, M., Siciliano, S. and Eilola, K., Meier, H.E.M. and Almroth, E. (2009) Moreno, I.B. (2015) Ecological niche On the dynamics of oxygen, phosphorus and modeling of Stenella dolphins cyanobacteria in the Baltic Sea; A model (Cetartiodactyla: Delphinidae) in the study. Journal of Marine Systems 75(1), 163- southwestern Atlantic Ocean. Journal of 184. Experimental Marine Biology and Ecology Elith, J., Kearney, M. and Phillips, S. (2010) The 472, 166-179. art of modelling range-shifting species. Douvere, F. (2008) The importance of marine Methods in Ecology and Evolution 1(4), 330- spatial planning in advancing ecosystem- 342. based sea use management. Marine Policy Elith, J. and Leathwick, J.R. (2009) Species 32(5), 762-771. Distribution Models: Ecological Explanation Drescher, M. and Brenner, J.C. (2018) The and Prediction Across Space and Time. practice and promise of private land Annual Review of Ecology, Evolution, and conservation. Ecology and Society 23(2). Systematics 40(1), 677-697. Duarte, B., Martins, I., Rosa, R., Matos, A.R., Elliott, M. (2011) Marine science and Roleda, M.Y., Reusch, T.B., Engelen, A.H., management means tackling exogenic Serrão, E.A., Pearson, G.A. and Marques, unmanaged pressures and endogenic J.C. (2018) Climate change impacts on managed pressures – A numbered guide. seagrass meadows and macroalgal forests: an Marine Pollution Bulletin 62(4), 651-655. integrative perspective on acclimation and Elmgren, R., Blenckner, T. and Andersson, A. adaptation potential. Frontiers in Marine (2015) Baltic Sea management: Successes Science 5, 190. and failures. Ambio 44(3), 335-344. Dunn, D.C., Van Dover, C.L., Etter, R.J., Smith, Embling, C.B., Gillibrand, P.A., Gordon, J., C.R., Levin, L.A., Morato, T., Colaço, A., Shrimpton, J., Stevick, P.T. and Hammond, Dale, A.C., Gebruk, A.V., Gjerde, K.M., P.S. (2010) Using habitat models to identify Halpin, P.N., Howell, K.L., Johnson, D., suitable sites for marine protected areas for Perez, J.A.A., Ribeiro, M.C., Stuckas, H. and harbour porpoises (Phocoena phocoena). Weaver, P. (2018) A strategy for the Biological Conservation 143(2), 267-279. conservation of biodiversity on mid-ocean EMODnet (2019) EMODnet Geology, 2019. ridges from deep-sea mining. Science Marine Minerals, © EMODnet Geology, Advances 4(7), eaar4313. European Commission. Edgar, G.J., Stuart-Smith, R.D., Willis, T.J., Eriksson, B.K., Sieben, K., Eklöf, J., Ljunggren, Kininmonth, S., Baker, S.C., Banks, S., L., Olsson, J., Casini, M. and Bergström, U. 46

(2011) Effects of altered offshore food webs Giakoumi, S., Guilhaumon, F., Kark, S., Terlizzi, on coastal ecosystems emphasize the need A., Claudet, J., Felline, S., Cerrano, C., Coll, for cross-ecosystem management. Ambio M., Danovaro, R. and Fraschetti, S. (2016) 40(7), 786. Space invaders; biological invasions in Evans, D. (2012) Building the European Union’s marine conservation planning. Diversity and Natura 2000 network. Nature Conservation Distributions 22(12), 1220-1231. 1, 11. Gilbert, D., Rabalais, N.N., Díaz, R.J. and Zhang, Fabry, V.J., Seibel, B.A., Feely, R.A. and Orr, J. (2010) Evidence for greater oxygen decline J.C. (2008) Impacts of ocean acidification on rates in the coastal ocean than in the open marine fauna and ecosystem processes. ICES ocean. Biogeosciences 7(7), 2283-2296. Journal of Marine Science 65(3), 414-432. Glardon, C.G., Walters, L.J., Quintana-Ascencio, Farmer, J.R., Ma, Z., Drescher, M., Knackmuhs, P.F., McCauley, L.A., Stam, W.T. and Olsen, E.G. and Dickinson, S.L. (2017) Private J.L. (2008) Predicting risks of invasion of landowners, voluntary conservation macroalgae in the genus Caulerpa in Florida. programs, and implementation of Biological Invasions 10(7), 1147-1157. conservation friendly land management Glasby, G., Emelyanov, E., Zhamoida, V., practices. Conservation Letters 10(1), 58-66. Baturin, G., Leipe, T., Bahlo, R. and Fennel, K. and Testa, J.M. (2019) Bonacker, P. (1997) Environments of Biogeochemical Controls on Coastal formation of ferromanganese concretions in Hypoxia. Annual Review of Marine Science the Baltic Sea: a critical review. Geological 11(1), 105-130. Society, London, Special Publications Fettweis, M., Baeye, M., Francken, F., Lauwaert, 119(1), 213-237. B., Van den Eynde, D., Van Lancker, V., Guisan, A., Tingley, R., Baumgartner, J.B., Martens, C. and Michielsen, T. (2011) Naujokaitis-Lewis, I., Sutcliffe, P.R., Monitoring the effects of disposal of fine Tulloch, A.I.T., Regan, T.J., Brotons, L., sediments from maintenance dredging on McDonald-Madden, E., Mantyka-Pringle, suspended particulate matter concentration in C., Martin, T.G., Rhodes, J.R., Maggini, R., the Belgian nearshore area (southern North Setterfield, S.A., Elith, J., Schwartz, M.W., Sea). Marine Pollution Bulletin 62(2), 258- Wintle, B.A., Broennimann, O., Austin, M., 269. Ferrier, S., Kearney, M.R., Possingham, H.P. Freeman, E.A. and Moisen, G. (2008) and Buckley, Y.M. (2013) Predicting species PresenceAbsence: An R Package for distributions for conservation decisions. Presence-Absence Model Analysis. Journal Ecology Letters 16(12), 1424-1435. of Statistical Software 23(11), 1-31. Halpern, B.S. (2014) Conservation: Making Friedman, J., Hastie, T. and Tibshirani, R. (2000) marine protected areas work. Nature Additive logistic regression: a statistical view 506(7487), 167-168. of boosting (with discussion and a rejoinder Halpern, B.S., Frazier, M., Afflerbach, J., by the authors). The annals of statistics 28(2), Lowndes, J.S., Micheli, F., O’Hara, C., 337-407. Scarborough, C. and Selkoe, K.A. (2019) Fulton, E.A., Link, J.S., Kaplan, I.C., Savina‐ Recent pace of change in human impact on Rolland, M., Johnson, P., Ainsworth, C., the world’s ocean. Scientific Reports 9(1), Horne, P., Gorton, R., Gamble, R.J. and 11609. Smith, A.D. (2011) Lessons in modelling and Halpern, B.S., Walbridge, S., Selkoe, K.A., management of marine ecosystems: the Kappel, C.V., Micheli, F., Agrosa, C., Bruno, Atlantis experience. Fish and Fisheries 12(2), J.F., Casey, K.S., Ebert, C., Fox, H.E., Fujita, 171-188. R., Heinemann, D., Lenihan, H.S., Madin, Gammal, J., Norkko, J., Pilditch, C.A. and E.M.P., Perry, M.T., Selig, E.R., Spalding, Norkko, A. (2017) Coastal Hypoxia and the M., Steneck, R. and Watson, R. (2008) A Importance of Benthic Macrofauna Global Map of Human Impact on Marine Communities for Ecosystem Functioning. Ecosystems. Science 319(5865), 948. Estuaries and Coasts 40(2), 457-468. Halpern, B.S. and Warner, R.R. (2002) Marine Gazis, I.Z., Schoening, T., Alevizos, E. and reserves have rapid and lasting effects. Greinert, J. (2018) Quantitative mapping and Ecology Letters 5(3), 361-366. predictive modeling of Mn nodules' Hannington, M., Petersen, S. and Krätschell, A. distribution from hydroacoustic and optical (2017) Subsea mining moves closer to shore. AUV data linked by random forests machine Nature Geoscience 10(3), 158-159. learning. Biogeosciences 15(23), 7347-7377. 47

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Harley, C.D., Randall Hughes, A., Hultgren, Jackson, J.B., Kirby, M.X., Berger, W.H., K.M., Miner, B.G., Sorte, C.J., Thornber, Bjorndal, K.A., Botsford, L.W., Bourque, C.S., Rodriguez, L.F., Tomanek, L. and B.J., Bradbury, R.H., Cooke, R., Erlandson, Williams, S.L. (2006) The impacts of climate J. and Estes, J.A. (2001) Historical change in coastal marine systems. Ecology overfishing and the recent collapse of coastal Letters 9(2), 228-241. ecosystems. Science 293(5530), 629-637. He, K.S., Bradley, B.A., Cord, A.F., Rocchini, Jackson, S.E. and Lundquist, C.J. (2016) D., Tuanmu, M.-N., Schmidtlein, S., Turner, Limitations of biophysical habitats as W., Wegmann, M. and Pettorelli, N. (2015) biodiversity surrogates in the Hauraki Gulf Will remote sensing shape the next Marine Park. Pacific Conservation Biology generation of species distribution models? 22(2), 159-172. Remote Sensing in Ecology and Janßen, H., Göke, C. and Luttmann, A. (2019) Conservation 1(1), 4-18. Knowledge integration in Marine Spatial HELCOM (2007) Baltic Sea Action Plan. Planning: A practitioners' view on decision HELCOM (ed). support tools with special focus on Marxan. HELCOM (2010) Towards an ecologically Ocean & Coastal Management 168, 130-138. coherent network of well-managed Marine Jiménez-Valverde, A. and Lobo, J.M. (2007) Protected Areas – Implementation report on Threshold criteria for conversion of the status and ecological coherence of the probability of species presence to either–or HELCOM BSPA network: Executive presence–absence. Acta Oecologica 31(3), Summary., Baltic Sea Environment 361-369. Proceedings No. 124A. Jiménez-Valverde, A., Peterson, A.T., Soberón, HELCOM (2012) Baltic Sea Environment J., Overton, J.M., Aragón, P. and Lobo, J.M. Proceedings No. 130. (2011) Use of niche models in invasive HELCOM (2016) Ecological coherence species risk assessments. Biological assessment of the Marine Protected Area Invasions 13(12), 2785-2797. network in the Baltic, Baltic Sea Johnson, D.E., Froján, C.B., Turner, P.J., Environment Proceedings No. 148. Weaver, P., Gunn, V., Dunn, D.C., Halpin, HELCOM (2018a) Sources and pathways of P., Bax, N.J. and Dunstan, P.K. (2018) nutrients to the Baltic Sea, Baltic Sea Reviewing the EBSA process: improving on Environment Proceedings No. 153. success. Marine Policy 88, 75-85. HELCOM (2018b) State of the Baltic Sea – Jokinen, S.A., Virtasalo, J.J., Jilbert, T., Kaiser, Second HELCOM holistic assessment 2011- J., Dellwig, O., Arz, H.W., Hänninen, J., 2016, Baltic Sea Environment Proceedings Arppe, L., Collander, M. and Saarinen, T. 155. (2018) A 1500-year multiproxy record of Humborg, C., Geibel, M.C., Sun, X., McCrackin, coastal hypoxia from the northern Baltic Sea M., Mörth, C.-M., Stranne, C., Jakobsson, indicates unprecedented deoxygenation over M., Gustafsson, B., Sokolov, A., Norkko, A. the 20th century. Biogeosciences 15, 3975- and Norkko, J. (2019) High Emissions of 4001. Carbon Dioxide and Methane From the Jones, K.R., Klein, C.J., Halpern, B.S., Venter, Coastal Baltic Sea at the End of a Summer O., Grantham, H., Kuempel, C.D., Shumway, Heat Wave. Frontiers in Marine Science 6, N., Friedlander, A.M., Possingham, H.P. and 493. Watson, J.E.M. (2018) The Location and Hällfors, M.H., Liao, J., Dzurisin, J., Grundel, R., Protection Status of Earth’s Hyvärinen, M., Towle, K., Wu, G.C. and Diminishing Marine Wilderness. Current Hellmann, J.J. (2016) Addressing potential Biology 28(15), 2506-2512.e2503. local adaptation in species distribution Jones, P.J.S., Lieberknecht, L.M. and Qiu, W. models: implications for conservation under (2016) Marine spatial planning in reality: climate change. Ecological Applications Introduction to case studies and discussion of 26(4), 1154-1169. findings. Marine Policy 71(Supplement C), Islam, M.S. and Tanaka, M. (2004) Impacts of 256-264. pollution on coastal and marine ecosystems Jonsson, P.R., Granhag, L., Moschella, P.S., including coastal and marine fisheries and Åberg, P., Hawkins, S.J. and Thompson, approach for management: a review and R.C. (2006) Interactions between wave synthesis. Marine Pollution Bulletin 48(7-8), action and grazing control the distribution of 624-649. intertidal macroalgae. Ecology 87(5), 1169- 1178. 48

Jonsson, P.R., Kotta, J., Andersson, H.C., Assessment. , Finnish Environment Institute Herkül, K., Virtanen, E., Sandman, A.N. and and Ministry of the Environment, Helsinki. Johannesson, K. (2018) High climate Korpinen, S., Laamanen, M., Suomela, J., velocity and population fragmentation may Paavilainen, P., Lahtinen, T. and Ekebom, J. constrain climate-driven range shift of the (2018) Suomen meriympäristön tila 2018, key habitat former Fucus vesiculosus. Suomen ympäristökeskus, Suomen Diversity and Distributions 24(7), 892-905. ympäristökeskus. Jonsson, P.R., Moksnes, P.-O., Corell, H., Korpinen, S., Meski, L., Andersen, J.H. and Bonsdorff, E. and Nilsson Jacobi, M. (2020) Laamanen, M. (2012) Human pressures and Ecological coherence of Marine Protected their potential impact on the Baltic Sea Areas: New tools applied to the Baltic Sea ecosystem. Ecological Indicators 15(1), 105- network. Aquatic Conservation: Marine and 114. Freshwater Ecosystems n/a(n/a). Kotta, J., Nurkse, K., Puntila, R. and Ojaveer, H. Jormalainen, V., Gagnon, K., Sjöroos, J. and (2016) Shipping and natural environmental Rothäusler, E. (2016) The invasive mud crab conditions determine the distribution of the enforces a major shift in a rocky littoral invasive non-indigenous round goby invertebrate community of the Baltic Sea. Neogobius melanostomus in a regional sea. Biological Invasions 18(5), 1409-1419. Estuarine, Coastal and Shelf Science 169, 15- Jouffray, J.-B., Blasiak, R., Norström, A.V., 24. Österblom, H. and Nyström, M. (2020) The Kotta, J., Vanhatalo, J., Jänes, H., Orav-Kotta, Blue Acceleration: The Trajectory of Human H., Rugiu, L., Jormalainen, V., Bobsien, I., Expansion into the Ocean. One Earth 2(1), Viitasalo, M., Virtanen, E., Sandman, A.N., 43-54. Isaeus, M., Leidenberger, S., Jonsson, P.R. Kaikkonen, L., Venesjärvi, R., Nygård, H. and and Johannesson, K. (2019) Integrating Kuikka, S. (2018) Assessing the impacts of experimental and distribution data to predict seabed mineral extraction in the deep sea and future species patterns. Scientific Reports coastal marine environments: Current 9(1), 1821. methods and recommendations for Kuhn, T., Wegorzewski, A., Rühlemann, C. and environmental risk assessment. Marine Vink, A. (2017) Deep-Sea Mining, pp. 23-63, Pollution Bulletin 135, 1183-1197. Springer. Kallasvuo, M., Vanhatalo, J. and Veneranta, L. Kukkala, A.S. and Moilanen, A. (2017) (2016) Modeling the spatial distribution of Ecosystem services and connectivity in larval fish abundance provides essential spatial conservation prioritization. information for management. Canadian Landscape Ecology 32(1), 5-14. Journal of Fisheries and Aquatic Sciences Lappalainen, J., Kurvinen, L. and Kuismanen, L. 74(5), 636-649. (2020) Suomen ekologisesti merkittävät Kareksela, S., Moilanen, A., Tuominen, S. and vedenalaiset meriluontoalueet (EMMA) – Kotiaho, J.S. (2013) Use of inverse spatial Finlands ekologiskt betydelsefulla marina conservation prioritization to avoid undervattensmiljöer (EMMA). Autio, R. biological diversity loss outside protected (ed), Suomen ympäristökeskus, Helsinki. areas. Conserv Biol 27(6), 1294-1303. Lehtomäki, J. and Moilanen, A. (2013) Methods Kaskela, A. and Rinne, H. (2018) Vedenalaisten and workflow for spatial conservation Natura -luontotyyppien mallinnus Suomen prioritization using Zonation. Environmental merialueella, Geologian tutkimuskeskus. Modelling & Software 47, 128-137. Kaskela, A.M. and Kotilainen, A.T. (2017) Lehtomäki, J., Tomppo, E., Kuokkanen, P., Seabed geodiversity in a glaciated shelf area, Hanski, I. and Moilanen, A. (2009) Applying the Baltic Sea. Geomorphology 295, 419- spatial conservation prioritization software 435. and high-resolution GIS data to a national- Kaskela, A.M., Kotilainen, A.T., Al-Hamdani, scale study in forest conservation. Forest Z., Leth, J.O. and Reker, J. (2012) Seabed Ecology and Management 258(11), 2439- geomorphic features in a glaciated shelf of 2449. the Baltic Sea. Estuarine, Coastal and Shelf Lehtoranta, J. and Pitkänen, H. (2003) Binding of Science 100, 150-161. phosphate in sediment accumulation areas of Kontula, T. and Raunio, A. (eds) (2019) the eastern Gulf of Finland, Baltic Sea. Threatened Habitat Types in Finland 2018. Hydrobiologia 492(1-3), 55-67. Red List of Habitats – Results and Basis for

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Leppäranta, M. and Myrberg, K. (2009) Physical Brunnabend, S.-E., Dieterich, C., Frauen, C., Oceanography of the Baltic Sea, Springer- Friedland, R., Gröger, M., Gustafsson, B.G., Verlag, Berlin-Heidelberg-New York. Gustafsson, E., Isaev, A., Kniebusch, M., Lester, S., Stevens, J., Gentry, R., Kappel, C., Kuznetsov, I., Müller-Karulis, B., Omstedt, Bell, T., Costello, C., Gaines, S., Kiefer, D., A., Ryabchenko, V., Saraiva, S. and Maue, C. and Rensel, J. (2018) Marine Savchuk, O.P. (2018) Assessment of spatial planning makes room for offshore Eutrophication Abatement Scenarios for the aquaculture in crowded coastal waters. Baltic Sea by Multi-Model Ensemble Nature communications 9(1), 1-13. Simulations. Frontiers in Marine Science 5, Lester, S.E. and Halpern, B.S. (2008) Biological 440. responses in marine no-take reserves versus Meier, H.E.M., Eilola, K. and Almroth, E. partially protected areas. Marine Ecology (2011b) Climate-related changes in marine Progress Series 367, 49-56. ecosystems simulated with a 3-dimensional Lester, S.E., Halpern, B.S., Grorud-Colvert, K., coupled physical‑biogeochemical model of Lubchenco, J., Ruttenberg, B.I., Gaines, the Baltic Sea. Climate Research 48(1). S.D., Airamé, S. and Warner, R.R. (2009) Meier, H.E.M., Hordoir, R., Andersson, H.C., Biological effects within no-take marine Dieterich, C., Eilola, K., Gustafsson, B.G., reserves: a global synthesis. Marine Ecology Hoglund, A. and Schimanke, S. (2012a) Progress Series 384, 33-46. Modeling the combined impact of changing Liu, C., Berry, P.M., Dawson, T.P. and Pearson, climate and changing nutrient loads on the R.G. (2005) Selecting thresholds of Baltic Sea environment in an ensemble of occurrence in the prediction of species transient simulations for 1961-2099. Climate distributions. Ecography 28(3), 385-393. Dynamics 39(9-10), 2421-2441. Lotze, H.K., Lenihan, H.S., Bourque, B.J., Meier, H.M., Müller-Karulis, B., Andersson, Bradbury, R.H., Cooke, R.G., Kay, M.C., H.C., Dieterich, C., Eilola, K., Gustafsson, Kidwell, S.M., Kirby, M.X., Peterson, C.H. B.G., Höglund, A., Hordoir, R., Kuznetsov, and Jackson, J.B. (2006) Depletion, I. and Neumann, T. (2012b) Impact of degradation, and recovery potential of climate change on ecological quality estuaries and coastal seas. Science indicators and biogeochemical fluxes in the 312(5781), 1806-1809. Baltic Sea: a multi-model ensemble study. Martin, K.S. and Hall-Arber, M. (2008) The Ambio 41(6), 558-573. missing layer: Geo-technologies, Mitchell, P.J., Monk, J. and Laurenson, L. (2017) communities, and implications for marine Sensitivity of fine-scale species distribution spatial planning. Marine Policy 32(5), 779- models to locational uncertainty in 786. occurrence data across multiple sample sizes. Martínez, B., Arenas, F., Trilla, A., Viejo, R.M. Methods in Ecology and Evolution 8(1), 12- and Carreño, F. (2015) Combining 21. physiological threshold knowledge to species Moilanen, A. (2007) Landscape Zonation, distribution models is key to improving benefit functions and target-based planning: forecasts of the future niche for macroalgae. Unifying reserve selection strategies. Global Change Biology 21(4), 1422-1433. Biological Conservation 134(4), 571-579. Meier, H.E.M., Andersson, H.C., Arheimer, B., Moilanen, A., Franco, A.M.A., Early, R.I., Fox, Donnelly, C., Eilola, K., Gustafsson, B.G., R., Wintle, B. and Thomas, C.D. (2005) Kotwicki, L., Neset, T.-S., Niiranen, S., Prioritizing multiple-use landscapes for Piwowarczyk, J., Savchuk, O.P., Schenk, F., conservation: methods for large multi- Węsławski, J.M. and Zorita, E. (2014) species planning problems. Proceedings of Ensemble Modeling of the Baltic Sea the Royal Society B-Biological Sciences Ecosystem to Provide Scenarios for 272(1575), 1885-1891. Management. Ambio 43(1), 37-48. Moilanen, A. and Kujala, H. (2014) Zonation Meier, H.E.M., Andersson, H.C., Eilola, K., spatial conservation planning framework and Gustafsson, B.G., Kuznetsov, I., Müller- software v. 4.0, User Manual. Helsinki, Karulis, B., Neumann, T. and Savchuk, O.P. Finland: University of Helsinki. (2011a) Hypoxia in future climates: A model Moilanen, A., Kujala, H. and Mikkonen, N. ensemble study for the Baltic Sea. (2020) A practical method for evaluating Geophysical Research Letters 38(24). spatial biodiversity offset scenarios based on Meier, H.E.M., Edman, M.K., Eilola, K.J., spatial conservation prioritization outputs. Placke, M., Neumann, T., Andersson, H.C., 50

Methods in Ecology and Evolution 11, 794- and Ovaskainen, O. (2019) A comprehensive 803. evaluation of predictive performance of 33 Moksnes, P.O., Gullström, M., Tryman, K. and species distribution models at species and Baden, S. (2008) Trophic cascades in a community levels. Ecological Monographs temperate seagrass community. Oikos 89(3), e01370. 117(5), 763-777. Norkko, J., Gammal, J., Hewitt, J.E., Josefson, Molinos, J.G., Halpern, B.S., Schoeman, D.S., A.B., Carstensen, J. and Norkko, A. (2015) Brown, C.J., Kiessling, W., Moore, P.J., Seafloor Ecosystem Function Relationships: Pandolfi, J.M., Poloczanska, E.S., In Situ Patterns of Change Across Gradients Richardson, A.J. and Burrows, M.T. (2016) of Increasing Hypoxic Stress. Ecosystems Climate velocity and the future global 18(8), 1424-1439. redistribution of marine biodiversity. Nature Norkko, J., Reed, D.C., Timmermann, K., Climate Change 6(1), 83-88. Norkko, A., Gustafsson, B.G., Bonsdorff, E., Morán-Ordóñez, A., Lahoz-Monfort, J.J., Elith, Slomp, C.P., Carstensen, J. and Conley, D.J. J. and Wintle, B.A. (2017) Evaluating 318 (2012) A welcome can of worms? Hypoxia continental-scale species distribution models mitigation by an invasive species. Global over a 60-year prediction horizon: what Change Biology 18(2), 422-434. factors influence the reliability of Nyström, M., Jouffray, J.B., Norström, A.V., predictions? Global Ecology and Crona, B., Søgaard Jørgensen, P., Carpenter, Biogeography 26(3), 371-384. S.R., Bodin, Ö., Galaz, V. and Folke, C. Moss, B. (2008) The Water Framework (2019) Anatomy and resilience of the global Directive: total environment or political production ecosystem. Nature 575(7781), compromise? Science of The Total 98-108. Environment 400(1-3), 32-41. Oh, B.Z., Sequeira, A.M., Meekan, M.G., MSFD (2012) Contribution of the Marine Ruppert, J.L. and Meeuwig, J.J. (2017) Strategy Framework Directive (2008/56/EC) Predicting occurrence of juvenile shark to the implementation of existing obligations, habitat to improve conservation planning. commitments and initiatives of the Member Conservation Biology 31(3), 635-645. States or the EU at EU or international level Ojaveer, H., Jaanus, A., MacKenzie, B.R., in the sphere of environmental protection in Martin, G., Olenin, S., Radziejewska, T., marine waters. Telesh, I., Zettler, M.L. and Zaiko, A. (2010) MSPD (2014) Maritime spatial planning for the Status of biodiversity in the Baltic Sea. PLOS sustainable development and growth of ONE 5(9), e12467. Europe’s maritime areas. Ovaskainen, O., Tikhonov, G., Norberg, A., Neumann, T. (2010) Climate-change effects on Guillaume Blanchet, F., Duan, L., Dunson, the Baltic Sea ecosystem: A model study. D., Roslin, T. and Abrego, N. (2017) How to Journal of Marine Systems 81(3), 213-224. make more out of community data? A Nieminen, E., Ahtiainen, H., Lagerkvist, C.-J. conceptual framework and its and Oinonen, S. (2019) The economic implementation as models and software. benefits of achieving Good Environmental Ecology Letters 20(5), 561-576. Status in the Finnish marine waters of the Peukert, A., Schoening, T., Alevizos, E., Köser, Baltic Sea. Marine Policy 99, 181-189. K., Kwasnitschka, T. and Greinert, J. (2018) Nilsson, J., Engkvist, R. and Persson, L.-E. Understanding Mn-nodule distribution and (2004) Long-term decline and recent evaluation of related deep-sea mining recovery of Fucus populations along the impacts using AUV-based hydroacoustic and rocky shores of southeast Sweden, Baltic optical data. Biogeosciences 15(8), 2525- Sea. Aquatic Ecology 38(4), 587-598. 2549. Norberg, A., Abrego, N., Blanchet, F.G., Adler, Pınarbaşı, K., Galparsoro, I., Borja, A., F.R., Anderson, B.J., Anttila, J., Araújo, Stelzenmüller, V., Ehler, C. and Gimpel, A. M.B., Dallas, T., Dunson, D., Elith, J., (2017) Decision support tools in marine Foster, S.D., Fox, R., Franklin, J., Godsoe, spatial planning: Present applications, gaps W., Guisan, A., O'Hara, B., Hill, N.A., Holt, and future perspectives. Marine Policy 83, R.D., Hui, F.K.C., Husby, M., Kålås, J.A., 83-91. Lehikoinen, A., Luoto, M., Mod, H.K., Pittman, S.J. (2017) Seascape ecology, John Newell, G., Renner, I., Roslin, T., Soininen, Wiley & Sons. J., Thuiller, W., Vanhatalo, J., Warton, D., Poloczanska, E.S., Burrows, M.T., Brown, C.J., White, M., Zimmermann, N.E., Gravel, D. García Molinos, J., Halpern, B.S., Hoegh- 51

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Guldberg, O., Kappel, C.V., Moore, P.J., Finnish marine area, northern Baltic Sea. Richardson, A.J. and Schoeman, D.S. (2016) Ambio 49(1), 118-129. Responses of marine organisms to climate Robinson, L.M., Elith, J., Hobday, A.J., Pearson, change across oceans. Frontiers in Marine R.G., Kendall, B.E., Possingham, H.P. and Science 3, 62. Richardson, A.J. (2011) Pushing the limits in Puttonen, I., Kohonen, T. and Mattila, J. (2016) marine species distribution modelling: Factors controlling phosphorus release from lessons from the land present challenges and sediments in coastal archipelago areas. opportunities. Global Ecology and Marine Pollution Bulletin 108(1-2), 77-86. Biogeography 20(6), 789-802. Puttonen, I., Mattila, J., Jonsson, P., Karlsson, Robinson, N.M., Nelson, W.A., Costello, M.J., O.M., Kohonen, T., Kotilainen, A., Lukkari, Sutherland, J.E. and Lundquist, C.J. (2017) A K., Malmaeus, J.M. and Rydin, E. (2014) Systematic Review of Marine-Based Species Distribution and estimated release of Distribution Models (SDMs) with sediment phosphorus in the northern Baltic Recommendations for Best Practice. Sea archipelagos. Estuarine, Coastal and Frontiers in Marine Science 4, 421. Shelf Science 145, 9-21. Rocha, J., Yletyinen, J., Biggs, R., Blenckner, T. Raateoja, M. and Setälä, O. (2016) The Gulf of and Peterson, G. (2015) Marine regime Finland Assessment, Finnish Environment shifts: drivers and impacts on ecosystems Institute. services. Philosophical Transactions of the Rabalais, N.N., Diaz, R.J., Levin, L.A., Turner, Royal Society B: Biological Sciences R.E., Gilbert, D. and Zhang, J. (2010) 370(1659), 20130273. Dynamics and distribution of natural and Rouillard, J., Lago, M., Abhold, K., Röschel, L., human-caused hypoxia. Biogeosciences 7(2), Kafyeke, T., Mattheiß, V. and Klimmek, H. 585-619. (2018) Protecting aquatic biodiversity in Ramirez-Llodra, E., Tyler, P.A., Baker, M.C., Europe: How much do EU environmental Bergstad, O.A., Clark, M.R., Escobar, E., policies support ecosystem-based Levin, L.A., Menot, L., Rowden, A.A., management? Ambio 47(1), 15-24. Smith, C.R. and Van Dover, C.L. (2011) Man Ruuskanen, A. (2016) The occurrence and and the Last Great Wilderness: Human monitoring of macroalgae in the coastal Impact on the Deep Sea. PLOS ONE 6(8), waters of Uusimaa. The description of e22588. national macrophyte monitoring and Reusch, T.B.H., Dierking, J., Andersson, H.C., application in Uusimaa 1993–016. . Bonsdorff, E., Carstensen, J., Casini, M., Sagerman, J., Hansen, J.P. and Wikström, S.A. Czajkowski, M., Hasler, B., Hinsby, K., (2020) Effects of boat traffic and mooring Hyytiäinen, K., Johannesson, K., Jomaa, S., infrastructure on aquatic vegetation: A Jormalainen, V., Kuosa, H., Kurland, S., systematic review and meta-analysis. Ambio Laikre, L., MacKenzie, B.R., Margonski, P., 49(2), 517-530. Melzner, F., Oesterwind, D., Ojaveer, H., Santangeli, A., Butchart, S.H., Pogson, M., Refsgaard, J.C., Sandström, A., Schwarz, G., Hastings, A., Smith, P., Girardello, M. and Tonderski, K., Winder, M. and Zandersen, Moilanen, A. (2018) Mapping the global M. (2018) The Baltic Sea as a time machine potential exposure of soaring birds to for the future coastal ocean. Science terrestrial wind energy expansion. Ornis Advances 4(5). Fennica 95(1), 1-14. Rinne, H., Björklund, C., Hämäläinen, J., Saraiva, S., Markus Meier, H.E., Andersson, H., Häggblom, M. and Salovius-Laurén, S. Höglund, A., Dieterich, C., Gröger, M., (2019) Rapport nr 153: Mapping Marine Hordoir, R. and Eilola, K. (2019) Baltic Sea Natura 2000 habitats in Åland Final report. ecosystem response to various nutrient load Rinne, H., Kaskela, A., Downie, A.L., Tolvanen, scenarios in present and future climates. H., von Numers, M. and Mattila, J. (2014) Climate Dynamics 52(5), 3369-3387. Predicting the occurrence of rocky reefs in a Savchuk, O.P. (2000) Studies of the assimilation heterogeneous archipelago area with limited capacity and effects of nutrient load data. Estuarine Coastal and Shelf Science reductions in the eastern Gulf of Finland with 138, 90-100. a biogeochemical model. Boreal Rinne, H. and Salovius-Laurén, S. (2019) The Environment Research 5(2), 147-163. status of brown macroalgae Fucus spp. and Sharp, R., Tallis, H., Ricketts, T., Guerry, A., its relation to environmental variation in the Wood, S., Chaplin-Kramer, R., Nelson, E., Ennaanay, D., Wolny, S., Olwero, N., 52

Vigerstol, K., Pennington, D., Mendoza, G., Torn, K., Krause-Jensen, D. and Martin, G. Aukema, J., Foster, J., Forrest, J., Cameron, (2006) Present and past depth distribution of D., Arkema, K., Lonsdorf, E., Kennedy, C., bladderwrack (Fucus vesiculosus) in the Verutes, G., Kim, C.K., Guannel, G., Baltic Sea. Aquatic Botany 84(1), 53-62. Papenfus, M., Toft, J., Marsik, M., Vahtera, E., Conley, D.J., Gustafsson, B.G., Bernhardt, J., Griffin, R., Glowinski, K., Kuosa, H., Pitkanen, H., Savchuk, O.P., Chaumont, N., Perelman, A., Lacayo, M., Tamminen, T., Viitasalo, M., Voss, M., Mandle, L., Hamel, P., Vogl, A.L., Rogers, Wasmund, N. and Wulff, F. (2007) Internal L., Bierbower, W., Denu, D. and Douglass, J. ecosystem feedbacks enhance nitrogen- (2018) InVEST 3.6.0 User’s Guide. The fixing cyanobacteria blooms and complicate Natural Capital Project, Stanford University, management in the Baltic Sea. Ambio 36(2- University of Minnesota, The Nature 3), 186-194. Conservancy, and World Wildlife Fund. Vahteri, P. and Vuorinen, I. (2016) Continued Snoeijs-Leijonmalm, P., Schubert, H. and decline of the bladderwrack, Fucus Radziejewska, T. (2017) Biological vesiculosus, in the Archipelago Sea, northern oceanography of the Baltic Sea, Springer Baltic proper. Boreal Environment Research Science & Business Media. 21, 373-386. Stelzenmüller, V., Lee, J., South, A., Foden, J. Walve, J., Sandberg, M., Larsson, U. and and Rogers, S.I. (2013) Practical tools to Lannergren, C. (2018) A Baltic Sea estuary support marine spatial planning: a review and as a phosphorus source and sink after drastic some prototype tools. Marine Policy 38, 214- load reduction: seasonal and long-term mass 227. balances for the Stockholm inner archipelago Stevens, T. and Connolly, R.M. (2004) Testing for 1968-2015. Biogeosciences 15(9), 3003- the utility of abiotic surrogates for marine 3025. habitat mapping at scales relevant to van Denderen, P.D., Bolam, S.G., Friedland, R., management. Biological Conservation Hiddink, J.G., Norén, K., Rijnsdorp, A.D., 119(3), 351-362. Sköld, M., Törnroos, A., Virtanen, E.A. and Sunday, J.M., Bates, A.E. and Dulvy, N.K. Valanko, S. (2019) Evaluating impacts of (2012) Thermal tolerance and the global bottom trawling and hypoxia on benthic redistribution of animals. Nature Climate communities at the local, habitat, and Change 2(9), 686-690. regional scale using a modelling approach. Sundblad, G., Bergstrom, U. and Sandstrom, A. ICES Journal of Marine Science 77(1), 278- (2011) Ecological coherence of marine 289. protected area networks: a spatial assessment Vaquer-Sunyer, R. and Duarte, C.M. (2008) using species distribution models. Journal of Thresholds of hypoxia for marine Applied Ecology 48(1), 112-120. biodiversity. Proceedings of the National Sundblad, G. and Bergström, U. (2014) Shoreline Academy of Sciences of the United States of development and degradation of coastal fish America 105(40), 15452-15457. reproduction habitats. Ambio 43(8), 1020- Weatherdon, L.V., Ota, Y., Jones, M.C., Close, 1028. D.A. and Cheung, W.W. (2016) Projected Takolander, A., Cabeza, M. and Leskinen, E. scenarios for coastal First Nations’ fisheries (2017a) Climate change can cause complex catch potential under climate change: responses in Baltic Sea macroalgae: A management challenges and opportunities. systematic review. Journal of Sea Research PLOS ONE 11(1). 123, 16-29. Weinert, M., Mathis, M., Kroencke, I., Neumann, Takolander, A., Leskinen, E. and Cabeza, M. H., Pohlmann, T. and Reiss, H. (2016) (2017b) Synergistic effects of extreme Modelling climate change effects on benthos: temperature and low salinity on foundational Distributional shifts in the North Sea from macroalga Fucus vesiculosus in the northern 2001 to 2099. Estuarine, Coastal and Shelf Baltic Sea. Journal of Experimental Marine Science 175, 157-168. Biology and Ecology 495, 110-118. Verbruggen, H., Tyberghein, L., Belton, G.S., Thomson, J.R., Moilanen, A., Vesk, P., Bennett, Mineur, F., Jueterbock, A., Hoarau, G., A.F. and Nally, R.M. (2009) Where and Gurgel, C.F.D. and De Clerck, O. (2013) when to revegetate: a quantitative method for Improving transferability of introduced scheduling landscape reconstruction. species’ distribution models: new tools to Ecological Applications 19(4), 817-828. forecast the spread of a highly invasive seaweed. PLOS ONE 8(6), e68337. 53

DEPARTMENT OF GEOSCIENCES AND GEOGRAPHY A87

Verhagen, W., Kukkala, A.S., Moilanen, A., van Wåhlström, I., Höglund, A., Almroth-Rosell, E., Teeffelen, A.J.A. and Verburg, P.H. (2017) MacKenzie, B.R., Gröger, M., Eilola, K., Use of demand for and spatial flow of Plikshs, M. and Andersson, H.C. (2020) ecosystem services to identify priority areas. Combined climate change and nutrient load Conservation Biology 31(4), 860-871. impacts on future habitats and eutrophication Viitasalo, M., Kostamo, K., Hallanaro, E.-L., indicators in a eutrophic coastal sea. Viljanmaa, W., Kiviluoto, S., Ekebom, J. and Limnology and Oceanography 65, 2170- Blankett, P. (eds) (2017) Meren aarteet, 2187. Gaudeamus. Yli-Hemminki, P., Sara-Aho, T., Jørgensen, K.S. Willis, S., Foden, W., Baker, D., Belle, E., and Lehtoranta, J. (2016) Iron–manganese Burgess, N., Carr, J., Doswald, N., Garcia, concretions contribute to benthic release of R., Hartley, A. and Hof, C. (2015) Integrating phosphorus and arsenic in anoxic conditions climate change vulnerability assessments in the Baltic Sea. Journal of Soils and from species distribution models and trait- Sediments 16(8), 2138-2152. based approaches. Biological Conservation Zhamoida, V., Grigoriev, A., Gruzdov, K. and 190, 167-178. Ryabchuk, D. (2007) The influence of Villnäs, A., Norkko, J., Hietanen, S., Josefson, ferromanganese concretions-forming A.B., Lukkari, K. and Norkko, A. (2013) The processes in the eastern Gulf of Finland on role of recurrent disturbances for ecosystem the marine environment. Geological Survey multifunctionality. Ecology 94(10), 2275- of Finland Special Papers 45, 21-32. 2287. Zhamoida, V., Grigoriev, A., Ryabchuk, D., Villnäs, A., Norkko, J., Lukkari, K., Hewitt, J. Evdokimenko, A., Kotilainen, A.T., Vallius, and Norkko, A. (2012) Consequences of H. and Kaskela, A.M. (2017) Increasing Hypoxic Disturbance on Benthic Ferromanganese concretions of the eastern Communities and Ecosystem Functioning. Gulf of Finland – Environmental role and PLOS ONE 7(10), e44920. effects of submarine mining. Journal of Virtanen, E.A., Moilanen, A. and Viitasalo, M. Marine Systems 172, 178-187. (2020) Marine connectivity in spatial Österblom, H., Hansson, S., Larsson, U., Hjerne, conservation planning: analogues from the O., Wulff, F., Elmgren, R. and Folke, C. terrestrial realm. Landscape Ecology 35(5), (2007) Human-induced trophic cascades and 1021-1034. ecological regime shifts in the Baltic Sea. Virtasalo, J.J., Kohonen, T., Vuorinen, I. and Ecosystems 10(6), 877-889. Huttula, T. (2005) Sea bottom anoxia in the Archipelago Sea, northern Baltic Sea— Implications for phosphorus remineralization at the sediment surface. Marine Geology 224(1), 103-122. Worm, B., Barbier, E., Beaumont, N., Duffy, J., Folke, C., Halpern, B., Jackson, J., Lotze, H., Micheli, F., R Palumbi, S., Sala, E., Selkoe, K., Stachowicz, J. and Watson, R. (2006) Impacts of Biodiversity Loss on Ocean Ecosystem Services. 314(5800), 787-790. Voulvoulis, N., Arpon, K.D. and Giakoumis, T. (2017) The EU Water Framework Directive: From great expectations to problems with implementation. Science of The Total Environment 575, 358-366. Vuorinen, I., Hänninen, J., Rajasilta, M., Laine, P., Eklund, J., Montesino-Pouzols, F., Corona, F., Junker, K., Meier, H.M. and Dippner, J.W. (2015) Scenario simulations of future salinity and ecological consequences in the Baltic Sea and adjacent North Sea areas–implications for environmental monitoring. Ecological Indicators 50, 196- 205. 54