234
could offer a way to further improve governance. In particular, an increased emphasis on human factors as causes of accidents seems warranted, since human error and
performance become more important as other causes are reduced.
Concerning fi sheries , while the complexity of sources is low, the high uncertainty of some important ecosystem effects in combination with sociopolitical controversies is clearly obstructing governance efforts. One response so far has been to
apply a precautionary approach, if not in political decisions on quotas at least in
preceding science-based advice. More important are the relatively new arrangements for stakeholder participation (Linke et al. 2016 ; Sellke et al. 2016 ). In this
case, it seems most important at present to ensure full implementation of the policies in place, which to some extent were recently ( 2014 ) renewed in the EU, in order
to see if that will be adequate in relation to stated objectives. We consider two
dimensions to be particularly important; fi rst, to apply the principle of maximum
sustainable yields within the frame of the EAM and the precautionary approach, as
well as to phase out discards and subsidies, and, second, to further regionalise
decision- making and to improve stakeholder participation.
Similarly, when it comes to invasive alien species (IAS) (Smolarz et al. 2016 ),
recent policies have been adopted (EU 2014 ). While uncertainty in terms of ecosystem effects of IAS is very high, risk management measures, for example, to better
control ballast water, seem well founded and relatively unproblematic to implement, as long as international cooperation works smoothly. Still, if an invasive species has high fi tness in the Baltic Sea ecosystem, even quite small implementation
defi cits might cause large problems, in particular over time. Nevertheless, in our
view, a critical point seems to be to ensure an ambitious and broad implementation
of the new regulation in its three dimensions of prevention, early warning and rapid
response and management. Possibly, this could be achieved if, or when, the IMO
Ballast Water Management Convention enters into force.
Regarding chemicals , it is much more diffi cult than in the other cases to obtain
suffi cient knowledge. Present risk assessments, that are affl icted with a number of
shortcomings, and cocktail effects, amongst other things, are extremely diffi cult to
evaluate (Karlsson and Gilek 2016 ). There are several science-based methods for
coping with uncertainty, for instance, by using precautionary default values for
exposure and toxicity when data is missing and by applying alternative decisionmaking criteria, such as maximin criteria (Karlsson 2010 ; Udovyk and Gilek 2013 ),
but present regulatory frameworks in the EU and the nation states around the Baltic
Sea have seldom used such approaches (Linke et al. 2016 ). Improved environmental risk governance in this case would presume vast regulatory reforms in the EU
and amongst parties to the Helsinki Convention. We consider it important, fi rst, to
fully reverse the burden of proof for decision-making, meaning, for example, that a
producer or user of a substance should show that legal requirements for safety are
met so that agencies do not have to prove risks beyond a reasonable doubt. Second,
regulatory reforms are needed to better coordinate environmental (e.g. the Marine
Strategy Framework Directive, MSFD) and polluter-oriented policy approaches
(such as the REACH regulation) (cf. Karlsson and Gilek 2016 ).
M. Gilek and M. Karlsson
could offer a way to further improve governance. In particular, an increased emphasis on human factors as causes of accidents seems warranted, since human error and
performance become more important as other causes are reduced.
Concerning fi sheries , while the complexity of sources is low, the high uncertainty of some important ecosystem effects in combination with sociopolitical controversies is clearly obstructing governance efforts. One response so far has been to
apply a precautionary approach, if not in political decisions on quotas at least in
preceding science-based advice. More important are the relatively new arrangements for stakeholder participation (Linke et al. 2016 ; Sellke et al. 2016 ). In this
case, it seems most important at present to ensure full implementation of the policies in place, which to some extent were recently ( 2014 ) renewed in the EU, in order
to see if that will be adequate in relation to stated objectives. We consider two
dimensions to be particularly important; fi rst, to apply the principle of maximum
sustainable yields within the frame of the EAM and the precautionary approach, as
well as to phase out discards and subsidies, and, second, to further regionalise
decision- making and to improve stakeholder participation.
Similarly, when it comes to invasive alien species (IAS) (Smolarz et al. 2016 ),
recent policies have been adopted (EU 2014 ). While uncertainty in terms of ecosystem effects of IAS is very high, risk management measures, for example, to better
control ballast water, seem well founded and relatively unproblematic to implement, as long as international cooperation works smoothly. Still, if an invasive species has high fi tness in the Baltic Sea ecosystem, even quite small implementation
defi cits might cause large problems, in particular over time. Nevertheless, in our
view, a critical point seems to be to ensure an ambitious and broad implementation
of the new regulation in its three dimensions of prevention, early warning and rapid
response and management. Possibly, this could be achieved if, or when, the IMO
Ballast Water Management Convention enters into force.
Regarding chemicals , it is much more diffi cult than in the other cases to obtain
suffi cient knowledge. Present risk assessments, that are affl icted with a number of
shortcomings, and cocktail effects, amongst other things, are extremely diffi cult to
evaluate (Karlsson and Gilek 2016 ). There are several science-based methods for
coping with uncertainty, for instance, by using precautionary default values for
exposure and toxicity when data is missing and by applying alternative decisionmaking criteria, such as maximin criteria (Karlsson 2010 ; Udovyk and Gilek 2013 ),
but present regulatory frameworks in the EU and the nation states around the Baltic
Sea have seldom used such approaches (Linke et al. 2016 ). Improved environmental risk governance in this case would presume vast regulatory reforms in the EU
and amongst parties to the Helsinki Convention. We consider it important, fi rst, to
fully reverse the burden of proof for decision-making, meaning, for example, that a
producer or user of a substance should show that legal requirements for safety are
met so that agencies do not have to prove risks beyond a reasonable doubt. Second,
regulatory reforms are needed to better coordinate environmental (e.g. the Marine
Strategy Framework Directive, MSFD) and polluter-oriented policy approaches
(such as the REACH regulation) (cf. Karlsson and Gilek 2016 ).
M. Gilek and M. Karlsson