Prioritising key variables for assessing food system resilience in Finland

ABSTRACT
 Building food system resilience enables the system to buffer, adapt and transform in response to changes. Identification of key variables of resilience support actors and policymakers managing the food systems. The objective of this study was to identify the most important variables to assess food system resilience and actors with the best opportunities to take responsibility for preparing for disruptions. We operationalised the key variables of resilience in the Finnish food system and evaluated them using a Delphi expert method. We produced a framework for food system resilience, including three indispensable interacting levels of action and good practices within each supporting resilience building. In the prioritisation of key variables, diversity in production, versatile cooperation between actors based on trust, independence of external inputs, system understanding, and renewable domestic energy were considered the most important. Research and administration play a key role in producing information and implementing actions targeted especially at agriculture, where actions are expected to be the most efficient and impactful. In the whole food system, increasing transparency would help create trust and thus promote co-creation of sustainable practices. Platforms to obtain information about future risks and for co-creating solutions for building food system resilience are needed.


Introduction
Increasing uncertainty in society challenges the ability of food systems to provide healthy and affordable food and to safeguard the functioning of ecosystems (Gordon et al. 2017;Cottrell et al. 2019;Willett et al. 2019).Securing the food supply requires ensuring the production inputs and operations at all food system levels, including in the event of disruptions.Promoting food system resilience meets this challenge, allowing the system to buffer, adapt and transform in response to changes (Berkes et al. 2000;Folke et al. 2010).To manage the multi-level interactions of food system influencing resilience, there is a need to identify key variables and means to help food system actors and policymakers' direct actions at building resilience (Cabell and Oelofse 2012).
Recent global events such as avian influenza, droughts in southern Europe, prolonged Covid-19 pandemic and the Russian invasion of Ukraine have revealed many vulnerabilities in societies and food systems (Laborde et al. 2020;Fan et al. 2021;Latvala et al. 2022).Resilience is the key concept that characterises system functioning when disturbance occurs (Vogel et al. 2007;Mumby et al. 2014).Resilience refers to the system ability to maintain its functions and, if necessary, to adapt and reorganise in the face of disruptions (Holling 1973;Gunderson and Holling 2002;Folke 2006;Folke et al. 2010;Bullock et al. 2017).
Resilience theory originates from ecology (Holling 1973), and its theoretical structures help to understand the dynamics and functions of many types of socialecological systems, including food systems (Tendall et al. 2015;Bullock et al. 2017;Stone and Rahimifard 2018).Béné et al. (2016) emphasises acknowledging three system characteristics when considering resilience, including system ability to (1) absorb disturbances, (2) adapt through learning and (3) transform and make notable changes in structures and functions as a result of shocks and stresses.The operationalising of resilience within increasingly globalised and complex socioecological systems (SES) is however challenging due to its abstract and multidimensional nature (Cabell and Oelofse 2012;Callo-Concha and Ewert 2014).In addition, the lack of understanding of the current state of the system resilience and the necessary development measures hamper the development of measurable indicators (Carpenter et al. 2001;Cabell and Oelofse 2012).
In this study, we take a step towards quantitative assessment of food system resilience.The starting point is the elements of food system resilience, identified by Rimhanen et al. (2023).Our focus is on the Finnish food system case, representing an interesting and unique research target for the food system resilience due to its northern and isolated location and high self-sufficiency of agricultural production.In addition, a long common border with a country at war increases the susceptibility to many disturbances.
Food systems can be defined as 'the sum of actors and interactions along the food value chainfrom input supply and production of crops, livestock, fish, and other agricultural commodities to transportation, processing, retailing, wholesaling, and preparation of foods to consumption and disposal' (IFPRI 2023).In the Global Food Security Index in 2022, Finland was ranked number one receiving top scores for all evaluated measures, i.e. affordability, availability, quality and safety and sustainability and adaptation (GFSI 2022).The main pillar of the Finnish food system is domestic agriculture, which produces a large part of the raw materials for the food industry.The dairy and meat processing industries make up the two main sectors in the Finnish food industry, accounting together for 43% of the food industry's turnover in 2016 (Niemi and Väre 2019).
Under normal conditions, the Finnish food system can meet the needs of consumption well.Finnish agriculture is self-sufficient in the productions of most agricultural products, such as cereals, milk and meat (Jansik et al. 2021).However, the weak point of the Finnish food system, in terms of resilience, is the high dependence on external inputs, such as chemicals, fuels and additives, as well as the workforce, which threatens the functioning of food systems if problems with their availability exist (Lehikoinen et al. 2021).Recently, the rising prices of external inputs, with the crisis in agricultural profitability, have threatened the liquidity of farmers (Latvala et al. 2022).Tendall et al. (2015) define food system resilience as the 'capacity over time of a food system and its units at multiple levels, to provide sufficient, appropriate and accessible food to all, in the face of various and even unforeseen disturbances'.The literature focusing on quantitative operationalisation that aims to compile measurable variables for assessing resilience have increased in recent years.Biggs et al. (2012) identified policy-relevant principles for enhancing the resilience of ecosystem services.Cabell and Oelofse (2012) identified 'behaviourbased indicators' of resilience within agroecosystems in which social, ecological, system connectivity and diversity concepts were integrated.For the food supply chain, Stone and Rahimifard (2018) identified resilience elements, Tendall et al. (2015) conceptualised food system resilience and Meuwissen et al. ( 2019) developed a framework to assess resilience in the farming system.
A framework for Finnish food system resilience was introduced by Rimhanen et al. (2023) to promote a national foresight system to better prepare for shocks and disruptions.The study identified four key elements of resilience in the food system: (1) system thinking through science and communication; (2) redundancy of activities and networks; (3) diversity of production and partners; and (4) buffering strategies.In addition, three national means of enhancing resilience in the Finnish food system were identified, including domestic protein crop production, renewable energy production and job creation measures.
There is however still gap in the knowledge on which are the most important concrete ecological, social and economic variables and means that help actors manage food system resilience.We agree with Darnhofer et al. 2010 that farming systems, not to mention food systems, are too complex to be assessed using single methods and indicators.Acknowledging the uniqueness of different systems (Cabell and Oelofse 2012), we organise, operationalise and evaluate the key elements of food system resilience (Rimhanen et al. 2023) into key variables of resilience.We fill the gap in knowledge regarding which variables are the most important for assessing the food system resilience, and by whom the variables should be promoted in the Finnish food system.The specific research questions for this study were: (1) Which are the key variables and good practices to promote resilience in the Finnish food system, and how are they interlinked?(2) Which are the most important variables society should invest in to build food system resilience?(3) Which actors have the best opportunities to take responsibility for preparing for disruptions?

Materials and methods
We used a three-step Delphi expert evaluation process to define the variables of resilience in the Finnish food system.Delphi as one of the main methods in futures studies has been widely used in anticipating future developments (Linstone and Turoff 1975;Bell 1997).
The Delphi method consists of experts' judgement by means of successive iterations of a given questionnaire, to show convergence of opinions and to identify dissent or non-convergence.It has often been used also in food system foresight (see e.g.Rikkonen 2005;Rikkonen and Tapio 2009;Himanen et al. 2016;Rikkonen et al. 2023;Rimhanen et al. 2023) In the first step, we identified the key elements of resilience in the Finnish food system (Rimhanen et al. 2023) through in-depth interviews with food system actors close to the national security of supply.Based on the interview results, in step two, we proposed statements and tested the key determinants, constructed probable and preferred future views of the Finnish food system and identified the needs for improvements through an online survey, the results of which are reported in Rikkonen et al. (2023).This paper reports the third step of this Delphi process, the results of an online survey, including the operationalisation, evaluation and organisation of key variables and their prioritisation, and an assessment of the actors' responsibilities for enhancing food system resilience.
The Delphi expert panel included food system experts in primary production, the input and processing industry, retail, research, governance, policymaking and advisory groups as this study's target group (Figure 1).The expertise of the Delphi panel has been described in detail in Rikkonen et al. (2023).
Based on the first and second steps, we created a draft framework (Figure 2).We included this figure in the online survey and asked the participants to comment on it.Written comments from the experts were analysed and used to re-evaluate and finalise the figure.
Furthermore, based on the results of the first and second steps, we formulated variables to illustrate key concerns for enhancing resilience in the food system.In the online survey, we asked the experts to rate the importance of each variable on a scale of 1 to 5. The survey also included open-ended questions, and the responses to them were used to supplement and clarify the key variables.The answers to the open questions were qualitatively analysed by one reviewer, using content analysis (Elo et al. 2014).

Results and discussion
A total of 24 individuals responded to the survey.

Key variables interacting at three levels of activity
The respondents emphasised the importance of considering the resilience of the food system holistically over a sufficiently long period and considering all relevant actors and their perspectives.'Systems thinking is not the prevailing practice [at the moment], but things are viewed very narrowly'.The respondents suggested that the goal of food system resilience defined in the draft should be extended from ensuring the stability of the food supply to ensuring the adequacy of food.The respondents stressed that flexibility should be increased in terms of both the quantity and quality of food.'During a crisis, you may have to cope with less and poorer nutrition'.This expansion was further reflected in the system boundaries and the agencies within the system, and in the final figure, we therefore considered the resilience measures relevant to all the actors in the food system, instead of emphasising primary production and the supply of food alone.
Based on the survey results, the view of the three interacting levels of action required for promoting resilience in the food system was confirmed (Figure 3).At the first level, agency-based measures implemented by all food system actors and driven by a wide range of entrepreneurial goals and regulatory constraints form the basis for building food system resilience.
According to the respondents, the current high input intensity of production and the high dependence on the use of chemical fertilisers and pesticides in agriculture increased the vulnerability of the food system.As a result, support for nutrient and energy independence, in particular the increase in nutrient recycling and domestic renewable energy production, was highlighted.The respondents emphasised the importance of increasing diversity at all levels of the food system, alongside ensuring adequate emergency stocks and systems which were considered to reduce the vulnerability of the food system, especially in the event of disruptions affecting the availability of imported inputs.In addition, the shortness and locality of the food value chain, as well as the decentralisation of food processing, were considered to facilitate the management of resilience.This would require the disbanding of the currently centralised food industry into networks of small and medium-sized entrepreneurial enterprises.A clear barrier identified by the respondents was the lack of vocational education for food processing.'We do have cooking schools, but not educational schools such as agricultural schools aimed at the food processor profession'.
Society's support for providing a framework for resilience was considered important.The National Emergency Supply Agency was highlighted to ensure the functionality of critical fields in the event of unexpected disruptions by maintaining emergency stocks.Securing back-up systems for data transmission and the distribution of electricity was considered important.To ensure the security of supply, keeping critical industries and infrastructure in national hands was considered critical in the responses.In addition, there was an  emphasis on influencing through legislation the structural development of agriculture, creating incentives to diversify the production structure and farm level, and decentralising the concentration of food processing.
The respondents also stressed the important role consumers played in food system resilience.Directing consumers towards a healthy and sustainable diet through food education was considered vital for achieving the environmental sustainability goals.
As the third level, social relations between actors and social justice were considered to be of primary importance in surviving the crisis.At the interface of the agency-based measures and society's support, actions promoting information exchange and cooperation in social and economic networks were considered necessary.The respondents emphasised the importance of the close cooperation of the National Emergence Supply Agency, the business community and the network of authorities to prevent disruption and adapt to change.
A partnership based on trust connects different levels of actors and promotes the emergence of a situational picture after a crisis, thus enabling the steering of the critical resources, for example.Such networks are also important for actors' co-learning and the emergence of innovations.They also increase the sense of fairness in the food system and promote mental wellbeing.The network of model farmers was mentioned as an example to promote peer learning and spur activities promoting resilience, but such a partnership should be expanded to various levels of the food system.
Enablers cutting across the three levels (Figure 3) promoting the resilience actions mentioned above were identified from the open-ended answers, including psychological factors important for motivation that were considered important to be aligned with the wider social value and reward system to promote resilience.'Activities that increase resilience should be motivating and in line with values, and of course also rewarding'.

The most important variables and means for building food system resilience
The experts' assessments of the importance of the variables for the resilience of the Finnish food system and the means which society should invest in enhancing food system resilience are shown in Tables 1 and 2, respectively.
The experts highlighted the importance of maintaining domestic production at the level of demand.Regarding the primary production and processing sector, the experts emphasised the importance of maintaining the number of farms at a sufficient level to secure the availability of expertise in agriculture and domestic raw materials.'The production structure has to be in order to secure the production in the event of a disruption and to be able to increase the production capacity if needed'.At the same time, the experts considered it important to educate citizens that a varied and healthy diet was a privilege and that a person could get by with less if necessary.
Increasing economic profitability, especially in primary production but also in the food industry, was considered a key factor in improving food system resilience.In terms of farm and processing entrepreneurship, the importance of decentralising operations, increasing diversity, sufficient investment and developing expertise was emphasised.In terms of primary production, the experts emphasised the importance of good soil growth condition, which would increase the resilience of cultivation to extreme weather events such as droughts and flooding and reduce the need for fertilisers and plant protection agents.
Regarding the entire food system, the importance of back-up systems for energy production was emphasised, as all food system operations were so dependent on energy.Likewise, capital and liquidity buffers were considered important.
According to the respondents, the decentralisation and localisation of food processing are important means of enhancing resilience, the promotion of which required support from the retail and political sectors.The functionality and flexibility of operations in the industry and retail sectors and logistic chains were also highlighted.For example: 'How flexibly can recipes be changed during the processing phase if necessary?And how does the retail sector respond to changes in the availability and composition of products after a failed harvest due to extreme weather, for example?' International trade agreements and negotiations were highlighted as important resilience means in the global food system.The experts also emphasised the importance of active imports.'If an accident or another serious disruption threatened the operation of domestic production on a large scale, imports would be vital for us'.

Promoting resilience means in society
The experts considered it important to define clear criteria, target levels and high-level monitoring of resilience.The experts considered it important to assess objectively the current state of food system resilience and to maintain programmes that supported the building of sustainability and resilience.'When the current situation is well known, it is easier for the administration to allocate resources'.
Advance planning, the dissemination of information, diverse discussions and cooperation between the actions of special groups, and policies that promoted resilience were the most important means of promoting food system resilience measures.
According to the experts, planning in advance, information sharing, diverse discussions and cooperation between different groups of actors, as well as policies promoting resilience, were the most important means of promoting resilience measures in the food system.Incentives and regulations were needed at all levels of the food system, from producers to politicians, from the local level to the EU and global levels.The experts considered that the importance of food supply security should be emphasised as a crosscutting factor in different policy areas.'The markets are not a reliable basis for the security of food supply'.A tangible proposal from the experts was the creation of a roadmap for the security of the food supply in which the identified resilience factors could be included.
It is also important to identify policies that pull in different directions in terms of resilience to break down administrative barriers, to create logical goals that exceed sectoral boundaries, and produce coherent policies to support them.
The experts considered that a structural change was needed in society that promoted the withdrawal from the fossil economy.The external effects of production should be included in retail prices through carbon taxes, for example.It was considered that this would improve the profitability of domestic production, as long as profit was targeted at primary production, and would increase the resilience of the entire food system.According to the experts, domestic energy production subsidies should be further increased, and green tariffs created to encourage domestic electricity generation.
The participation of actors, cooperation and holistic system thinking created a strong foundation for increasing awareness.The experts considered it important that the food system functioned more as a united entity.This would enable a holistic examination of the components of resilience.'On the basis that we should have the idea that the system is functioning well when all its parts are functioning well'.The experts highlighted that in food policy, there was a need to create a common vision of what the resilient food system should look like.Policy decisions and guidance should then be compatible with this vision.'The current style of policies based on CAP periods included the ideas: "let's change as little as possible" and "let's hold on to the achieved gains" does not support resilience in the food system in the long term'.
Instead of increasing food production, the experts emphasised that it was important to promote the moderate and fair use of resources.For example, regarding energy, effective measures and functioning distribution networks supported energy saving.Promoting fairness in the food system was important, especially in production and distribution.'No one would have to hoard more than they needed or even start dealing in food on the black market'.
According to the experts, targeted research was needed to promote the food system's resilience.Sufficient resources must be allocated to research, as well as to the development of solutions and to the preparation of contingency plans.The implementation of the research results in practice and information transmission was likewise considered important ways to promote resilience.Increasing open discussion and dialogue between different actors was considered important.Tangible actions and experiences promoted understanding.'Often these words remain very abstract to people.But when they are opened, people will understand better'.
The experts highlighted the need for a change in the agricultural support systems in a way that would promote diversity in agriculture more than was currently the case.The experts considered that promoting resilience would require the decentralisation of food system operations.Among different production sectors in agriculture, the importance of livestock production was highlighted in Finland, due to favourable environmental conditions for grass production and large water resources.Regionally, more evenly distributed livestock production would be beneficial for efficient nutrient recycling.
Regarding primary production, the development of cultivation methods and the training of farmers were considered important.The diversification of cultivation, the functionality of the soil food web, and the nutrient and carbon cycles of the fields were considered to be particularly important.They were considered to promote the resilience of cultivation against flooding and droughts and to produce healthier plants.

Barriers to resilience
According to the respondents, the lack of understanding, knowledge and expertise about the food system as a whole and the necessary development measures hindered the development of resilience.'When there is no consensus on the importance of things, and it is not understood that everything actually affects everything!' In addition, ossified attitudes and opinions were considered obstacles to change.
The experts considered that there was a lack of policy instruments that promoted resilience, a result of low research funding and consequent research data, as well as insufficient cooperation between food system actors.The experts thought that the political confrontation of relevant topics hindered the development of resilience and prevented an objective examination of the necessary development needs.As examples of politicised issues, the experts raised chemicals and pesticides, where ideologies were perceived as dominating the discussion.Another significant factor was the injustice and even corruption experienced by some of the food system actors, which had reduced the willingness to cooperate and prevented the comprehensive development of the food system.
The seclusion of different parts and administration of the food system was considered to prevent the food system's comprehensive development in a more sustainable direction.Conflicts of interest between different actors were considered to threaten the promotion of resilience factors if cooperation and resources could not be directed to the right targets.'The incentives for individual actors are not always in line with the overall resilience of the food system'.Responses are on a 5-point Likert scale (1 = not at all important; 2 = little important; 3 = fifty-fifty; 4 = fairly important; 5 = very important).
According to the experts, due to the globality of the food systems, many aspects affected the Finnish food system that were difficult to directly influence.Operating in a global operating environment, internationality and free movement were felt to be somewhat in conflict with the maintenance of the national food supply security.However, the experts also emphasised that sufficient development measures would require international solutions.With the uncontrollability of the effects of other countries, EU institutions and internationally regulated world trade, the importance of examining the issues that could be influenced and the most effective ways to promote critical issues was highlighted.For example, to improve nutrient and energy supply in Finland, a stronger investment could be made in the development of biogas.
Regarding critical production inputs, Finland was considered to be too small a market to increase its domestic production.However, the experts considered that it was important to increase their production at the EU level because vitamins and amino acids were currently being imported from Asia, for example.
The experts identified short-sighted and rigid agricultural policy as one of the biggest obstacles to enhancing food system resilience.The ideology of continuous economic growth was found to be unsuitable for food production.The experts considered it especially problematic that in the Finnish production structure pig and poultry production were currently disconnected from the local biomass production through imported feed and nutrients.
In addition, the experts claimed that decision makers currently lacked the political will and courage to promote a trade policy that included the external effects of production, for example, through carbon tax solutions.Political ideas were considered to hinder the social discussion of the sustainability of the food system.'Emeritus professors should create more twists'.The experts emphasised that the current economic and market system was oriented towards the economy of scale maintained by the fossil economy, which contradicted the diversity, locality and decentralised solutions associated with resilience.
The experts criticised the current agricultural support system and its supervision.According to the experts, the current targeted subsidies steered the food system in the wrong direction in terms of resilience.The cheap price of food, which did not take the external effects of production into account, received criticism.
According to the experts, there was currently too much trust in the market as a guardian of food supply security.This was considered to be the result of shortsighted thinking and disruptions being perceived as only temporary problems.The experts considered that in a market economy, taking resilience into account usually increased costs, and it was thought to worsen financial profitability.'It is impossible to advance the resilience goals sufficiently through consumption demand.On the contrary, the current development seems to be towards an increasingly concentrated globalised system'.The experts also emphasised that the low profitability of primary production weakened the implementation of resilience measures.Low producer prices were considered especially to guide the choice of crop species and varieties, which was an obstacle to the diversification of cultivation.

Actor responsibilities in preparing for disruptions
The experts emphasised solidarity for preparing for disruptions in the food system.Due to the wide-ranging and surprising nature of the effects of disturbances, the responsibility for identifying new threats, black swans, lay especially with all sectors and actors.The experts emphasised that it would be important for each actor to recognise and develop their own operational possibilities and capabilities in preparing for disruptions.The experts emphasised the role of public administration and political decision makers in developing and ensuring preparedness for various disturbances.The experts considered it beneficial to clearly define responsibilities and recommended the utilisation of expert groups with representatives from a wide range of fields in the planning.
The experts considered that the cooperation of actors in the food chain was currently insufficient.This was because cooperation had been dictated by the pursuit of economic profit, leading to confrontation in pricing disputes.'Food security is not market power'.
The experts highlighted that the prerequisite for promoting resilience in the food system was the creation of reliable information and its implementation (Table 3).The experts considered it important that instead of individual competing projects, resilience research was assembled into broad programmes around different themes.'Research funding should be secured in a completely different way than the current project system, where the work input of brilliant researchers has to be allocated to applying for funding and office work'.
According to the experts, the public administration would benefit from a more thorough understanding of food system resilience.'To build resilience, regulation is needed, which requires research into relevant topics and methods'.Research, public administration and political decision makers therefore had joint responsibility in many mattersthey needed one another.Equally, officials and political decision makers should promote and facilitate research and development work with their actions and decisions.The importance of the choices made by sectoral associations to promote cooperation at a higher level than the market setting was also highlighted.In addition, guaranteeing the operations of civic and advisory organisations were considered important for spreading information widely.

Discussion
In this paper, we have operationalised the key elements into key variables of resilience in the Finnish food system and evaluated them using a Delphi expert method.The study compiles the food system experts' views on the key variables to assess and promote food system resilience.An analysis of the results of the expert process produced a framework for food system resilience that included three indispensable interacting levels of action and good practices within each level supporting resilience building.In addition, a prioritisation of the key variables for assessing food system resilience and assessment of actors with the best prerequisites to take responsibility for preparing for disruptions was produced.

Limitations of the study
The Delphi method as one of the main methods in futures studies is an expert method, and the results base on the expert panel composition.Kuusi (1999) argues that the method for selecting the Delphi panel is one of the most critical phases of a Delphi study.The Delphi facilitator should consider in his/her actor analysis the most important stakeholders, most important substance (the competence of experts) as well as the terms of delivering information in a Delphi process (information policy).The selection process of an expert panel should be done as explicit as possible.We identify that a majority of experts came from the background of research and development, and primary production.Also, from the educational background point of view, natural sciences dominated the panel composition.However, during the Delphi research process we were able to commit altogether 72 food system experts from various stakeholder and interest groups to answer in interviews and surveys, and as a result this study of resilience brings to the table a quite broad food system expert community future views on the needs for resilience development.
The identification of the elements of resilience and their derived variables is based on in-depth interviews with food system experts in which the management of disturbances and risks affecting the food system were discussed (Rimhanen et al. 2023).The poor profitability of primary production, the current high intensity and dependency on imported inputs of agricultural production, especially the increased use of chemical fertilisers and pesticides in the food system, were currently increasing vulnerability in the Finnish food system.The decentralisation of agricultural production and processing was also considered risky.The seclusion of different food system functions was considered to prevent the development of a mutual holistic understanding, hindering the co-creation of good practices that built resilience (Rikkonen et al. 2023;Rimhanen et al. 2023).
We agree with Darnhofer et al. (2010) that farming systems alone, much less food systems, are too complex to be comprehensively assessed with the aid of a list of indicators that inevitably fail to account for multiple pressures and feedback loops.However, through our expert method, we operationalised and integrated social, ecological and economic aspects that determined resilience and thus identified critical points for building food system resilience.We follow Cabell and Oelofse's (2012) interpretation according to which the presence of these key variables suggests resilience in the system, and correspondingly, their absence suggests vulnerability.
Promoting diversity has been clearly identified as a powerful means of enhancing resilience (Biggs et al. 2012).Moreover, in our study, the importance of increasing diversity at all levels of the food system was highlighted throughout the expert process.In normal conditions, increasing diversity in production and networks may decrease a system's efficiency (Jin et al. 2021).However, when considering increasing disturbances and uncertainty, the ability of diversity to provide multiple back-ups and to promote learning and adaptation are essential for food system resilience (Rimhanen et al. 2023).
Food systems provide several positive and negative outcomes, such as food security, employment and greenhouse gases.Activities to promote resilience involve normative choices that can lead to making compromises for different outcomes and stakeholders (Zurek et al. 2022).For agriculture, in the current difficult economic situation, the diversification of production would advantage both resilience and profitability.In recent years, the market demand for special crops such as oil and protein crops has increased (Latvala et al. 2022).Biological diversity alone is not however enough, but more diversity is needed in the regional production systems, for different sizes of companies, and in the courses of action to secure the performance of food systems in the event of disturbances.The literature distinguishes between two types of diversity.Functional diversity refers to a variety of provided ecosystem services (Flynn et al. 2009), manifested for example in a diversity of inputs, outputs and sources of income.Whereas response diversity provides a range of responses to different disturbances (Kahiluoto et al. 2014), for example, the different growth rates of different crops ensure harvested yield, and different actors react differently to market disturbances.The food system experts rated the diversity of crops, varieties and rotation the most important means in which society should invest to enhance resilience.In practice, this will entail a transition to an agroecological approach, e.g.integrated and organic farming (Hendrickson et al. 2008;Reganold and Wachter 2016).The consequences of the intensification of agriculture and especially the absence of crop rotation is strongly observed today (Peltonen-Sainio et al. 2017).The need for development has been identified especially on farms specialising in onesided crop production using purchased inputs.Many experts consider the inflexible nature of the Common Agricultural Policy does not steer in this direction (Peltonen-Sainio et al. 2015, 2016).Some experts in Finland have even proposed that the current direction of agricultural policy may be contrary to what will be needed to build more resilience in the Finnish food system (Rikkonen et al. 2023).
The high ranking of increasing the production of critical inputs in Finland, with nutrient cycling and promotion of domestic renewable energy production, reflects the identified current vulnerabilities of the Finnish food system (Rimhanen et al. 2023).As was experienced in 2021-2023, the rising prices of energy greatly affected the economic conditions of farms in Finland (Termonen 2022), highlighting the importance of energy security to food system resilience.Being a net importer of electricity, and integrated into the wider electricity system in Europe, external disturbances in energy markets also extend to consumer prices in Finland.Distributed renewable energy production has been recognised as a central means of strengthening energy security (TEM 2023).Especially, there is growing interest in integrating aspects of energy production more closely with agriculture and the food processing industry.Finnish farmers have some experience of farm-scale energy production, yet widespread integration of biogas production and solar photovoltaic (PV) generation into the wider energy system remains to be developed, despite being an oft-discussed topic in Finland.On-farm energy production would have the benefit of reducing the impact of electricity price fluctuations, as well as of acting as a grassroots back-up system during energy outages.Furthermore, integrating agricultural energy production with other local energy systems would support the resilience of the local energy security while simultaneously helping overcome the economic barriers that are often experienced with agricultural energy initiatives, for example, biogas facilities (Rasi et al. 2022).
A holistic understanding is important for ensuring that food system actors are aware of the underlying socio-ecological processes and factors influencing the decision making of different actors.The common understanding feeds the cooperation based on trust and commits different actors to cooperation to achieve a common goal.This also enables learning from past experiences and improves preparedness for future changes, thus enhancing recovery, adaption and even transformation (Gunderson and Holling 2002;Folke 2006;Folke et al. 2010;Cabell and Oelofse 2012).
The need to improve the understanding of the complex food system emerged strongly in the expert method.Farmers and other practitioners, advisors, educators, developers and researchers need up-to-date information about possible risks and potential solutions, as well as joint platforms for exchanging experiences and ideas on building food system resilience.The knowledge of local actors is a vital complement to research knowledge for the co-creation of feasible and acceptable solutions for local conditions (Berkes et al. 2000;Shava et al. 2010).In Finland, ongoing national communication projects provide platforms for sharing information, increasing communication and for the multiactor co-creation of feasible local solutions (Kike 2023;Muru 2023).
Research and administration were considered to have the most possibilities to take responsibility for preparing for disruptions.Preparation for future uncertainty requires knowledge of the risks and their consequences that research can offer, while administration can provide carrots or sticks through taxation or incentives to promote sustainable practices, for example.Regarding surprising new black swans, the importance and responsibility of all actors for recognising sensitive signals were emphasised.Increasing the understanding of all actors and engaging them in working together for a common goal emphasises the importance of communication.
The means to build resilience are especially important for agriculture, which has been found to be the most vulnerable part of the Finnish food system (Rimhanen et al. 2023).It is important to highlight that any further preparatory actions on farms are likely to come with added costs to Finnish farmers, who have long been struggling with poor profitability (Latvala et al. 2022).While preparatory actions should be applied where they are the most cost-effective and impactful, the benefits will be collected by all food system actors as added resilience.A just allocation of the costs in preparatory efforts between different food system actors is therefore a prominent topic for further research.
While the discussion of resilience and self-sufficiency in the EU food system has increased, the European Commission has also taken measures to prepare for and respond to potential threats to global food supply and food security.The European Food Security Crisis preparedness and response mechanism (EFSCM) assesses how to improve cooperation between the public and private sectors and evaluate risks when crises arise (EC 2022a).Furthermore, The European Commission's proposal for a Critical Entities Resilience Directive (CER) aims to improve the resilience of society's critical services, and it covers critical entities such as energy and drinking water in societies' resilience more broadly (EC 2022b).In the Finnish context, the National Emergency Supply Agency has a long tradition of organising the Finnish security of supply on a voluntary basis between private and public sector cooperation.In terms of its starting points and principles, the CER is partly different from the Finnish security of supply system, which has been created for the needs of Finnish society.The CER Directive covers 10 sectors that are broader in scope than those of the more targeted Finnish security of supply organisation.Moreover, the participation of operators identified in sectors or pools critical to the security of supply is voluntary in Finland, whereas the CER Directive introduces separate obligations for the identified critical actors (NESA 2023).The harmonisation efforts between the national and EU levels for these new resilience mechanisms can therefore be seen as key development points in the near future as the implementation proceeds in the EU area.
It seems that the power and responsibility for taking responsibility for preparing for disruptions are distributed currently in different ways in the food system.In Finland, the dominant position of trade in the food system is often emphasised, but in this study, trade was considered to have the fewest opportunities to lead preparation for disruptions.
Taking into account the better profitability of trade compared to other actors in the food system, their bigger role in participating in the costs would probably be considered fair.Increasing transparency will be important for building trust and thus increasing the sense of fairness in the food system.

Conclusions
As both sudden unexpected disturbances and long-term changes are expected to become more common in the future, it is important to understand which are the key variables for assessing food system resilience.By rating the importance of variables for assessing the food system resilience, this study revealed that diversity of production, trust and versatile cooperation between actors, independence of external inputs, system understanding and domestic renewable energy are the most important variables for assessing food system resilience in the Finnish context.This is an important step in the development of measurable resilience indicators.Further developing these variables into socio-ecological quantitative and qualitative indicators enables the development and monitoring of the resilience of food systems.
Promoting efficient nutrient cycling and soil health through diverse crop rotation will reduce the dependence on external inputs in agriculture.Integrating renewable energy production into agriculture and food processing would reduce impacts from electricity price fluctuation and improve local energy security.Also, increasing transparency would help create trust and thus promote joint co-creation towards sustainable practices.These activities will make food systems more resilient in Finland, but lessons can also be learned in other countries, especially those whose food systems are strongly based on self-sufficient agriculture.
We created a framework in which we organised good practices at three levels of action: agency-based measures; social and economic networks; and society's support.This structure facilitates the discussion of the complex and multidimensional topic.Preparing for disruptions and changes requires knowledge of risks and opportunities, and consequences and the means to manage them.In addition, ability to renew and transform is needed.Research and administration play a key role in producing information and implementing actions, targeted especially at agriculture, where actions are expected to be the most efficient and impactful.Platforms for actors to obtain information about future risks and potential adaptation means, and for exchanging experiences and co-creating locally acceptable solutions for building food system resilience, are required.Any further preparatory means on farms are likely to come with added costs to farmers, whose poor profitability prevents the implementation of new practices.As the benefits of such means will be collected by all food system actors as added resilience and considering the better economic success of other actors in the food system, especially retail, their more significant investments in the costs will be justified.Future research into food system resilience should therefore focus on a just allocation of the costs in resilience means.Furthermore, determining the actual indicators and their appropriate level, as well as the relationship between resilience and effectiveness, would promote the management and monitoring of resilience.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
We gratefully acknowledge funding from the Academy of Finland DEFORFO project [grant number 335648] and TREFORM [grant number 339830].

Notes on contributors
Dr. Karoliina Rimhanen is a Research Scientist at the Natural Resources Institute Finland (Luke).She has a Ph.D. in agroecology from the University of Helsinki, Finland.Her research has focused on resilience, climate change adaptation and multiactor approach in agriculture and food systems.
Kalle Aro works as a Research Scientist at the Natural Resources Institute Finland (Luke).His research pertains policy making and stakeholder participation in the context of low carbon transition, with emphasis on bioenergy and agriculture.Aro graduated an M.Sc.(Environment, Energy and Society) from the University of Stavanger and currently pursues a doctoral degree at Tampere University.
Dr. Pasi Rikkonen is a Principal Research Scientist and futurist at the Natural Resources Institute Finland (Luke).Dr. Rikkonen received his doctorate from the Turku School of Economics.He has conducted numerous future-oriented studies and participatory scenario exercises within the fields of agriculture, food and bioenergy.He has been an expert member of several future-oriented working groups of the Ministry of Agriculture and Forestry and Parliament of Finland.

Figure 1 .
Figure 1. Background information of Delphi expert panel, n = 25 (in educational background and in background reference group, a respondent was able to answer more than one option).

Figure 2 .
Figure 2. The draft framework of the main levels of activity and the practices within them to enhance resilience in the Finnish food system drafted based on in-depth interviews (first Delphi round).

Figure 3 .
Figure 3. Enhancing resilience in the food system through good practices at three levels of action: agency-based measures; social and economic networks; and society's support.Good practices within each level of activity are presented in the box on the right.

Table 1 .
Expert assessment of the importance of variables for food system resilience in Finland.

Table 2 .
Expert assessment of the importance of the means which society should invest in enhancing food system resilience.

Table 3 .
Expert assessment of actors with best prerequisites to take responsibility for preparing for disruptions.