Addressing biohazards to food security in primary production

This review addresses ways to prepare for and to mitigate effects of biohazards on primary production of crops and livestock. These biohazards can be natural or intentional introductions of pathogens, and they can cause major economic damage to farmers, the agricultural industry, society, and international trade. Agroterrorism is the intentional introduction of animal or plant pathogens into agricultural production systems with the intention to cause socioeconomic harm and generate public fear. Although few acts of agroterrorism are reported, the threat of agroterrorism in Europe is real. New concerns about threats arise from the rapid advancements in biotechnology and emerging technologies. FORSA, an analytical framework for risk and vulnerability analysis, was used to review how to prepare for and mitigate the possible effects of natural or intentional biohazards in agricultural production. Analyzing the effects of a biohazard event involves multiple scientific disciplines. A comprehensive analysis of biohazards therefore requires a systems approach. The preparedness and ability to manage events are strengthened by bolstered farm biosecurity, increased monitoring and laboratory capacity, improved inter-agency communication and resource allocation. The focus of this review is on Europe, but the insights gained have worldwide applications. The analytical framework used here is compared to other frameworks. With climate change, Covid-19 and the war in Ukraine, the supply chains are challenged, and we foresee increasing food prices associated with social tensions. Our food supply chain becomes more fragile with more unknowns, thereby increasing the needs for risk and vulnerability analyses, of which FORSA is one example.


Agriculture and food security
Agriculture plays a key role in maintaining welfare and social, economic, and political stability by providing essential food products. A secure food supply is one of the basic tenets of modern societies and what many citizens expect from their government. Food supply and food security are constantly facing risks related to weather, lack of input resources, diseases, and national or international conflicts as well as threats by intentional and unintentional actions (Fig. 1). Agricultural commodities and their production, storage, and distribution could become targets for attacks that aim to disrupt socioeconomic stability. Disruptions to agriculture could be, for example, the introduction of pathogens (parasites, bacteria, viruses, and fungi) into crops and livestock herds potentially resulting in disease outbreaks, transmission of infections to other hosts, or infestations of seed or plant material.

What is agroterrorism?
The term agroterrorism is used to describe the intentional release of plant pests or animal pathogens with the goal to cause economic damage, undermine social stability and generate public fear. This could be done by a state, a group, or an individual in the form of biowarfare, bioterror or biocrime (Chalk, 2004;Foxell, 2001;Latxague et al., 2007;Mumford et al., 2017). In this review, agroterrorism refers to any attack using a biological agent against any on-farm agricultural commodity. Acts of agroterrorism may be difficult to distinguish from unintentional or natural introductions of pathogens, but all of them require the same preparedness planning and management, at least until the origin of an introduction has been confirmed. The resulting damage can be considerable regardless of whether the introduction is deliberate, accidental, or natural (Table 1).
In the past, the focus of biosecurity measures in European agriculture was on managing natural or accidental pathogen introductions (Suffert, 2017). However, over the past two decades, agroterrorism has gained increased attention (Green et al., 2017;Suffert, 2017;Mackelprang & Friedman, 2021). There are also increasing concerns that the continuing rapid advancements in biotechnology, and emergence of 4 th industrial revolution technologies, could enable or facilitate the development of biological weapons and the creation of adapted harmful pathogens that could be used in agroterror attacks (Koblentz, 2020;Selgelid, 2009).
This review addresses ways to prepare for and mitigate possible effects of natural biohazards and agroterrorism on primary production of crops and livestock. The structure is based on an analytical framework called the FORSA model (Winehav & Nevhage, 2011). It was developed in Sweden as a tool in risk and vulnerability analysis of important societal activities and functions at local, regional, and governmental level. While the focus is on the situation in Europe, worldwide application of the approach presented here is possible. The usefulness of this and other approaches will be discussed.

FORSA-An analytical framework for risk and vulnerability analysis
A risk analysis consists of risk assessment, risk management, and risk communication (EU, 2002;FAO, 2007;IPPC, 2007). In risk assessment, risks are identified, analysed and evaluated in a systematic way and form the base for risk management. There are inconsistencies in the use of the terms risk analysis and risk assessment (cf. Anon, 2018;FAO, 2007), and here we follow the FAO outline (FAO, 2007). Analytical frameworks for risk analysis are used to define the steps in the analysis and their relations (examples are discussed in Sect. 3). The FORSA framework was developed by the Swedish Defence Research Agency (FOI) (Winehav & Nevhage, 2011), and consists of six successive steps (Fig. 2). Here, we use the framework to organise the information required to address potential agroterror acts and other biohazards to primary crop and livestock production intended for human consumption. The steps include: I. Description of the function for society itself, including critical dependencies on other functions. II. Identification and characterization of unwanted or unexpected events, or threats, and their respective likelihoods. III. Description of the effects of an event on the function that was identified in the previous step. It includes the identification of vulnerabilities, methods, options and capabilities to manage risk, along with the resources this requires. IV. Actions to manage an acute event, eliminate or reduce threats and vulnerabilities, and strengthen the ability to manage a crisis. V. The methods and data used in steps I-IV.   Conraths et al., 2009;Gethmann et al., 2020 VI. The decisions that are made on the ways to convey the results to stakeholders and to ensure that risk management actions are implemented.
This framework can be used to address both natural and intentional biohazards, and in the following sub-sections, actions in each step of the framework are described.

Step I-Function for society
The first step in the risk and vulnerability analysis describes the function for society (Fig. 2), which in this case is primary crop and livestock production for food, that may lead to diseases or other direct damage to livestock or crops. Threats, risks, vulnerabilities, and critical dependencies are identified along with the resources that are available to reduce serious disturbances.

Farming in Europe
The contribution of agriculture to maintaining private and public activity is significant and important for rural employment and development (Table 2). Agriculture also generates raw materials for businesses and secondary industries, employing millions of people. The food industry is the largest manufacturing industry in the EU (Food-DrinkEurope, 2020). Thus, any unexpected disruption event in the primary production would not only directly affect farmers, but it would also have indirect effects on the secondary industries, as well as on international trade.
Over the last century, farming in Europe has changed dramatically, including a trend towards fewer and larger, farming operations (Eurostat, 2018). In the past, agriculture was a highly localized industry where crops and animals were raised, harvested, and processed on-farm or within the same region. Today's farms are more specialized and focus on the production of few commodities (European Commission, 2013;Abson, 2019). Many goods and services, which were previously provided locally, are now provided by outside contractors and private companies. This has resulted in a shift to agricultural production where animals, grains, agricultural supplies, and food products are unevenly distributed throughout the world and transported over large distances (Greger, 2007;Niemi, 2012). The EU's Common Agricultural Policy (CAP) regulates production, prices, and markets (European Commission, 2018a).

Changing attitudes towards agriculture
The number of people living in rural areas has been decreasing since the 1950s (World Bank, 2019), so for many people, the connection between agriculture and food production has been lost. The result is a public that has a sense of security generated by a plentiful supply of safe and healthy food in stores, but with a limited appreciation for agriculture's role in society. The limited contact that people have with agriculture also means that fewer persons pursue careers within the agro-industry, food animal medicine or plant pathology (Dunning et al., 2009;Gullino, 2009;Kelly, 2005;MacDonald et al., 2009). If this leads to job vacancies, it could add to a country´s vulnerability to biohazards and agroterrorism. A lack of veterinarians or agriculture experts could delay the diagnosis of an emerging disease or the identification of an antagonistic release of a pathogen, allowing the disease to spread more widely.

Farm security and biosecurity
Many farms operate in an open manner with few physical security measures to prevent unauthorized access, especially to outlying fields and pastures (Crutchley et al.,   2007). Farm biosecurity refers to the measures to protect a farm against introduction of pathogens and to minimize their spread within a farm and beyond (CFIA, 2013). Farm biosecurity is important to consider when looking at vulnerability to agroterrorist attacks, particularly in intensive production systems. In recent years, few European countries have suffered major losses due to agricultural disease outbreaks, which can lull people into the false sense that their production systems are safe and protected (Chalk, 2004;Crutchley et al., 2007). These attitudes could be changed by increased knowledge about the benefits of implementing higher biosecurity levels compared to their costs (Kristensen & Jakobsen, 2011;Laanen et al., 2014). A high level of farm biosecurity, including new technologies, would limit potential introduction of a biological agent into a crop or group of animals (Jurdak et al., 2015;Hu et al., 2020;Kakani et al., 2020;Koblenz, 2020).

Legal frameworks
All EU members are required to have a national capacity to perform pest and disease risk assessments (PRA), as stated in the Animal Health Law and The Plant Health Law, and in the International Plant Protection Convention (IPPC) (EU, 2016a, b;IPPC, 1997). The PRA are the base for the development of risk management plans. These regulations apply regardless of whether outbreaks are intentional or not. Regulations and lists of pathogens are established with the aims to build awareness of potential risks and to prevent pathogen introductions and disease outbreaks (Table 3). Since many of the agents on these lists lack the desired characteristics of an agroterrorism agent (See Annex 1), additional lists of high-threat agricultural pathogens have been compiled (Wilson et al., 2000;Madden & Wheelis, 2003;Schaad et al., 2006;Suffert et al., 2009;Kamenidou et al., 2013;Suffert, 2017). It is impossible to produce a common short list of high-threat pathogens because the threat level varies depending on, for example, location, perpetrators, motives, and the intended target (Madden & Wheelis, 2003;Suffert et al., 2009). Sometimes biological information about pathogens is lacking, in which case list developers rely on expert opinion and educated guesses (Schaad et al., 2006). Despite these problems, creating lists of potential threat agents is an important aspect of preparedness to counter agroterrorism (Madden & Wheelis, 2003).

Step II-Unwanted event
The second step of a risk and vulnerability analysis, includes identification of specific unwanted events, and evaluation of their risks, vulnerabilities, and critical dependencies. The  Group, 2007 likelihood of an event, for example the intentional introduction of an animal or plant pathogen, and its consequences and uncertainties, is quantified. Risk matrices are useful to illustrate the relationships between likelihoods and consequences, based on current knowledge and assessments, especially when quantitative data is scarce or uncertain (Van der Fels-Klerx et al., 2018) (Fig. 3).

Pathogens and diseases of concern
Natural disease outbreaks can be predicted to some extent, and they have similarities with previous outbreaks (Treadwell et al., 2003). This knowledge is used in conventional disease risk analyses, which are the base for disease response plans. Annex 2 gives examples of pathogens that constitute serious threats to food security. In case of an intentional pathogen release, the predictabilities of an outbreak are replaced by uncertainties through the calculated actions of the perpetrator who can select the agent, target, location, and the scale and timing of the release Treadwell et al., 2003). For example, a pathogen could be released in several geographically distant locations to initiate multiple, simultaneous outbreaks Mackelprang & Friedman, 2021). Manipulations that bypass conventional biosecurity measures and exploit existing vulnerabilities may thus cause an outbreak that plays out differently from a natural outbreak. This could overwhelm a nation's response capacity, leading to the uncontrolled dispersal of a pathogen and spread of the disease ). Very few plant pathogens pose a risk to human health. However, some fungi, such as the cereal pathogens Fusarium spp., which causes fusarium head blight and ear rot, and Claviceps purpurea, which causes ergot, produce mycotoxins that can result in acute and chronic disease or death in humans and animals. The risk of mycotoxins entering the food chain is low due to good detection methods, and regular testing of harvested grains, food, and feed for mycotoxin contamination (Alshannaq & Yu, 2017;Paterson, 2006). On the other hand, use of mycotoxin-producing fungi as an agroterror weapon can potentially have human and animal health effects and result in economic damage as infested grains can become unfit for consumption.
Many livestock diseases, such as Rift Valley Fever, Congo Crimean Haemorragic Fever and Avian Influenza, are zoonotic which means that they can be transmitted between animals and humans. However, human-to-human transmission of these diseases is rare. Therefore, an agroterror attack using a zoonotic pathogen is unlikely to cause a widespread outbreak among people, but farm workers, veterinarians, and persons having direct contact with infected animals face the risk of becoming ill (Hueston & Singleton, 2010).

Dilemmas with synthetic biology
Recent discoveries and developments in molecular genetics and synthetic biology have allowed for precise changes in the genome of living organisms. These advances have led to research that increases our understanding of how pathogens infect hosts which has aided in development of more effective vaccines and increased the speed at which vaccines can be manufactured in response to disease outbreaks (DiEuliis et al., 2017). However, with the advent of increasingly sophisticated methods for altering organisms at the molecular level, there are concerns about the potential to use artificially created or genetically altered biological agents as weapons (Drew & Mueller-Doblies, 2017). The need to balance the advancement of beneficial life sciences research with the need to prevent further development of biological weapons is known as the 'dual-use dilemma' (Atlas & Dando, 2006;Selgelid, 2009), and research that is intended for benefit, but which also could be applied to cause harm, is referred to as 'dual-use research of concern' (DURC) (Imperiale & Casadevall, 2015).
Openly sharing information is vital for the scientific community, but it also provides would-be proliferators of biological weapons easy access to new knowledge and techniques (Anand, 2018). There are numerous scientific papers describing research that is considered as DURC (WHO, 2010;Valles & Bernacchi, 2014), and several examples of how it can be addressed scientifically and ethically (Atlas & Dando, 2006;Selgelid, 2009;Imperiale & Casadevall, 2015;Selgelid, 2016;Drew & Mueller-Dobies, 2017;Anand, 2018;Imperiale & Casadevall, 2020). In addition, the risks for malicious use of new techniques and emerging technologies such as the use of drones, digital equipment, and 3D-printing requires preventive actions from policymakers (Bajema, 2018;Koblenz, 2020

Dispersal and long-distance movement of pathogens
The natural processes of pathogen dispersal, especially by wind and vectors, may be difficult to control. In addition, human activities may also contribute to pathogen dispersal. Long-distance movements of agricultural commodities increase the possibilities for pathogen dispersal beyond the original site of introduction (Crutchley et al., 2007;Dunn, 1999;Wu et al., 2021). For animals, the stress of prolonged transport can have an immunosuppressive effect and increase their susceptibility to disease (Greger, 2007). A study tracking the movements of Swedish livestock hauliers showed that vast areas are covered in short periods of time (Olofsson et al., 2014). A highly contagious pathogen could then easily be dispersed over a very large area before animals show disease symptoms (Olofsson et al., 2014). When examining long-distance movements, the vehicles, boots, clothing, and other equipment that can transmit pathogens between fields and farm sites need attention, in addition to the commodities themselves.

Identification of vulnerabilities
A deliberately introduced pathogen can disperse quickly through animal or plant populations in a country or region. Usually, agricultural animals are not naturally exposed to many infectious diseases and will be neither immune nor vaccinated against these diseases. The resulting lack of herd immunity increases their vulnerability to deliberate attacks (Ungerer & Rogers, 2006), as would limited stocks of vaccines. Similarly, certain plant pathogens and pests are currently not found in Europe, and little attention is paid to integrating resistance to them in the varieties grown. In case of an introduction these crops will be susceptible. Although several plant pathogens may be managed by pesticides, their availability depends on whether a product is approved in a country (EU, 2009;European Chemicals Agency, 2021), if it is approved to be used in the crop at risk, and if the quantity in stock covers the need in an outbreak situation. Vulnerability to disease outbreaks increases with large or dense populations of susceptible plants or animals. The trend toward increased farm size, homogenous crop distribution, and regional concentration of production often results in large populations of susceptible hosts in small areas (Dorea et al., 2017;Foxell, 2001;Gilligan, 2008;McDonald & Stukenbrock, 2016;Meadows et al., 2018). Widespread use of crop varieties with a narrow genetic base increases vulnerability to new pathogens or new variants of known pathogens (McDonald & Stukenbrock, 2016;Zhan et al., 2015). The vulnerability of crops to all types of biohazards also increases if there is too much reliance on previous success in managing a disease. For example, while the viruses causing virus yellows of sugar beet were managed with neonicotinoids, the development of alternative management methods, such as virus-resistant cultivars, was neglected. After the use of neonicotinoids in the EU was banned, virus yellows has re-emerged as an immediate threat to sugar beet cultivation (Hossain et al., 2021).
The ability to quickly identify the pathogen is crucial for early and appropriate control measures in the event of a deliberate or unintentional introduction. It can become extremely costly if the identification of a pathogen is delayed or fails due to a lack of training (Dorea et al., 2017;Hammond et al., 2016;Wright et al., 2016). Veterinarians, plant pathologists, and other agricultural experts often have limited experience in identification and management of exotic diseases because they are seldom found Thelaus et al., 2017). Adequate monitoring and surveillance systems are needed so that a pathogen does not go undetected, and disease outbreaks are managed in time.

Step III-Analysis of event
The third step in the analysis describes possible consequences of an unwanted event affecting the function for society, the abilities to mitigate them, and the resources that are available or needed. Vulnerabilities and critical dependencies are rated on scales that express the seriousness of the consequences, and the abilities to mitigate or manage an event. These scales are often qualitative descriptions based on current knowledge, experience from other events, and expert judgements or knowledge elicitations. For evaluation and comparison of risk scenarios, the qualitative scales can be made quantitative or semi-quantitative by ranking e.g., likelihoods, impacts, or frequencies in time intervals on nominal (e.g., good to poor) or ordinal scales (e.g., 1-5) (Léger et al., 2017;Robb, 2017;EFSA PLH Panel et al., 2018;Van der Fels-Klerx et al., 2018). Another way is to create scores based on exposure, sensitivity and adaptive capacity (Fontaine & Steinemann, 2009). Descriptions of consequences in categories such as the specific effects on health, environment, economy, and extent and duration of the event would be informative. However, to be useful they must be established in advance, and collaboratively between all actors (Utne et al., 2008). The consequences of a crisis become difficult to handle when it is uncertain what the hazards will be (Frykmer et al., 2018;Utne et al., 2008).
The mitigation and management of an unintentional or deliberate introduction of a disease requires a systematic and systems-oriented approach, even if the introduction is limited (Anand, 2018). In this step, further analysis of the consequences is possible, and the risk matrix from step II can be updated with new facts (Fig. 3). The continued analysis will be based on the estimates of consequences in this step.

Economic effects
Animal and plant diseases can cause considerable losses (FAO, 2016;Oerke, 2006;Savary et al., 2019). Even limited outbreaks of diseases, most of them natural, have resulted in substantial economic losses (Table 1). The losses incurred during agroterrorism attacks would be similar since the measures taken to control a disease outbreak are the same regardless of its cause (Hugh- Jones & Brown, 2006).
An agroterror attack or other biohazard in primary crop or livestock production will result in quantitative and qualitative losses for the farmer(s) involved. In addition to these losses, a biohazard event will have socio-economic consequences on three levels: direct, indirect, and for international trade (Chalk, 2004). The measures to manage the indirect losses vary depending on the introduced pathogen, and they can be costly, even if the affected commodity's contribution to the economy is small. However, in the long-term, the gains from appropriate management should outweigh the costs (Wheelis et al., 2002). The losses from a disease outbreak to secondary industries and the society's economy are expected to be similar to those of the affected primary industry, or larger (Blake et al., 2003;Petterson & Widell, 2010).
The direct and indirect economic losses, other than those experienced by the farmer, are seldom presented and, if so, mostly in qualitative terms. It is particularly hard to derive the estimates of indirect losses. Sampling routines and intensities are not designed to estimate loss to society (Madden et al., 2007), and losses to other actors in the food production chain are more difficult to disentangle (Yudelman et al., 1998). The process to identify abilities and resources that are needed for the mitigation of an unwanted event and to foresee vulnerabilities would be strengthened with advancements in the identification and quantification of the direct and indirect losses (Madden & Wheelis, 2003;Chalk, 2004;Suffert et al., 2009;Gamliel & Fletcher, 2017a) (Fig. 4).
International trade losses arise when trade embargoes are imposed by a country´s trading partners (Chalk, 2004). Every member of the World Trade Organization (WTO) has the right to impose import restrictions to protect public, animal, and plant health, and decide on an appropriate level of protection (Suffert et al., 2009). The detection of a disease in one country can result in import restrictions or bans on agri-food products by other countries. Such bans can take months or years to resolve and trade volumes may never return to those existing prior to the outbreak (Hueston & Singleton, 2010). The severity of trade implications due to an animal or plant disease outbreak depends on the size of the export markets for the affected country (Kahn, 2019;Keygene, 2019;Tozer & Marsh, 2012).
When public confidence in food security is undermined, it has a significant impact on consumer trust and demand (Evans, 2006), and results in reduced consumption of the affected products, even if there is no real cause for concern to human health (Clarke & Rinderknecht, 2011). For example, domestic chicken consumption dropped by up to 30% in some European and Asian countries after outbreaks of Avian Influenza (Kraipornsak, 2010;Valceschini, 2006). Spanish cucumber growers were severely affected when consumption and export of their cucumbers halted after they were wrongly suspected of being the source of an Escherichia coli outbreak in Germany in 2011, which was ultimately traced to contaminated fenugreek seeds (Trigonella foenum-graecum) for sprouts from Egypt (Burger, 2012). Consumer demand for the affected foodstuff could be reduced to a greater extent if a food safety scare resulted from agroterrorism than if it were due to a natural disease outbreak (Just et al., 2009;Turvey et al., 2010).
Food prices would increase if an accidental or deliberate introduction of a disease or pest results in a food supply shortage (Linacre et al., 2005;McDonald, 2013). In mediumor low-income countries this could cause consumer discontent, and political instability (Arezki & Brückner, 2011). To Reduce revenue deficits at agribusiness and -industry -manufacturers receiving agricultural goods -industries supplying products or services to farmers -tourism, hotel and restaurant industries what degree increased imports could compensate for a shortage would depend on trade arrangements, with food security being higher in countries with an openness to trade (Dithmer & Abdulai, 2017). The most vulnerable countries would be low-income countries and countries with a high proportion of GDP coming from agriculture, economic dependence on cash crops, insufficient self-sufficiency for basic food items, undersized border control, low farm biosecurity combined with intensive farming, and few resources for extension systems and monitoring (Anand, 2018;Dithmer & Abdulai, 2017;Linacre et al., 2005;Sundström et al., 2014).

Political, health-related, and social effects
In addition to the economic impact of a biohazard, there will be political, health-related, and social effects. A terrorist attack on agriculture could change the political climate in a country. It is likely to affect the public's trust in their government's preparedness to manage biohazards and the ability to provide adequate quality control of their food supply, resulting in a decrease in consumer confidence in the safety of their food (Wilson et al., 2000;Cupp et al., 2004;Evans, 2006;Crutchley et al., 2007;Chalk, 2010;Hueston & Singleton, 2010).
After an agroterror attack, the physical health risks to the public are likely to be smaller than the mental health effects (Hueston & Singleton, 2010). A terrorist act resulting in a disease outbreak would be a highly distressing experience for those directly impacted by the event (Peck et al., 2005;Hueston & Singleton, 2010). Many people lack knowledge about how pathogens cause disease, so there is likely to be widespread anxiety and fear around attacks with pathogens (Crutchley et al., 2007). The health effects are likely to strain public health care systems, particularly in rural areas (Evans, 2006).
If the goal of terrorism is to weaken society by creating fear and feelings of vulnerability, the social effects on farmers, the extended agricultural sectors, communities and first responders dealing with the outcome of an attack should be recognized (Evans, 2006). Following the UK FMD outbreak, producers and communities were socially isolated due to closure of roads and schools, cancellation of events, and limited access to community support networks. Consequently, tension and long-lasting conflicts were seen in rural communities (Royal Society of Edinburgh, 2002; Becker, 2010).

Step IV-Actions
The fourth step in the risk and vulnerability analysis addresses actions to manage an acute and unexpected event, eliminate or reduce threats and vulnerabilities, and strengthen the ability to manage a crisis. The estimates of risks and consequences in the previous steps help to identify necessary and possible actions, and their order of priority. The goal of all actions is to reduce the likelihood of an event, its consequences, or both (Fig. 3). At the farm level, the actions would address farm biosecurity. At the national level they would include surveillance and diagnosis, and at national and international levels, regulations and risk assessments. All actions will be considered from both economic and political angles since it is difficult to estimate the costs of the consequences (Utne et al., 2008).
The key to successful short-and long-term management of a disease outbreak or an agroterror attack would be a joint and multisectoral response.

Farm biosecurity
Farms with a high level of biosecurity are more difficult targets and, in case of an event, the pathogen spread between farms may be slowed. An important step would be to increase farmers' knowledge and understanding of disease risks, and how responsibilities are shared between farmers and authorities (Higgins et al., 2016;Renault et al., 2020). If several channels and means for information distribution are used, the chances of reaching most farmers are increased .
According to EU regulations, all member states must establish national control plans (EU, 2016a). A biosecurity program for farmers includes training in prevention of disease spread, assessment of farms by a veterinarian with training in biosecurity and giving advice to farmers on how to improve their farm biosecurity.

Pest and disease risk assessment (PRA)
An important step in preparedness and response planning for agroterrorism is to determine which pathogens, out of hundreds, that pose the greatest threat, since it is not feasible to develop prevention and response plans for all Schaad et al., 2006). A country's preparedness level would increase by adding and updating lists of high threat agroterror agents to their agricultural security policies and standards. The lists of high-threat pathogens would be useful as guidance for*: • development of counter-terrorism strategies to prevent and deter the use of specific pathogens as weapons, • research into more rapid and sensitive methods of detection and diagnosis of target diseases, • risk assessments and epidemiological studies of disease spread, • development of forensic biology methods that could help differentiate between deliberate and natural introductions * (Madden & Wheelis, 2003) Disease and pest risk assessments require regular revisions by individual nations in order to adhere to the Animal Health and the Plant Health Laws, respectively (EU, 2016a, b). Some advocate that disease and pest risk assessments should include agroterrorism-related factors, such as perpetrator motives and capability, and target vulnerabilities, since an agroterrorist attack may have a course that deviates from what is expected or known from naturally occurring disease outbreaks Suffert, 2017).

Surveillance and diagnosis
In recent years, counterterrorism efforts have become a political and national security priority. Unlike other acts of terrorism such as bombings and hijackings, which are carried out overtly, an act of agroterrorism is most likely to be a covert operation (Farsang et al., 2013). If the act is clandestine and because most biological agents are invisible, an attack may remain undetected until signs or symptoms of disease appear in exposed plants or animals at the end of the incubation period (several days to weeks), or until the detection threshold in affected plants or animals is reached (Madden & Wheelis, 2003;Zadoks & Schein, 1979). This makes early detection of an agroterror attack challenging (Farsang et al., 2013). In light of ongoing climate changes, unexpected disease outbreaks and introductions of pathogens are foreseeable, and their detection will require similar attentiveness.
In the EU, there are monitoring and surveillance programs for animal diseases, and plant pests and diseases (European Commission, 2008;EFSA et al., 2020b). An animal disease notification system (ADNS) is established, and quarantine priority plant pests are identified based on their potential for social, economic, and environmental damage (European Commission, 2012; EU, 2019). In addition, most countries have their own monitoring programs that provide evidence that certain pathogens and pests are absent from the country, and baseline prevalence information, and serve to detect the introduction of pathogens so that actions can be taken to prevent further spread . Preparedness for plant disease outbreaks would be strengthened by a global disease surveillance system (Carvajal-Yepes et al., 2019). For economic reasons, active surveillance programs are not possible for all animal diseases or quarantine plant pests. Combined statistical and epidemiological models can be used when designing surveillance systems based on pathogen characteristics (Parnell et al., 2017), but no program is completely reliable.
Successful surveillance for high-threat pathogens depends on farmers, veterinarians, and plant protection specialists, and notifications of new disease outbreaks to the authorities are essential early detection tools (Elbers et al., 2010). Early detection therefore hinges on the ability to recognize the signs and symptoms of unexpected diseases. If a disease has not been seen for many years, the signs may not be recognized, and may instead be assumed to be signs of endemic disease (Elbers et al., 2010). It is of utmost importance that those who work on the front lines of agricultural health, have education in disease recognition to ensure rapid detection Mackelprang & Friedman, 2021).
It is mandatory to report suspected high-threat pathogen infections to responsible authorities before any action can be taken. The reporting of suspect disease cases should be uncomplicated and could be encouraged by financial compensation, or fines upon un-reported outbreaks. If producers must pay for a pest eradication, they will be encouraged to adopt preventive measures (Centner & Ferreira, 2012;Wilkinson et al., 2011).
Early detection and identification of a pathogen minimizes delay and the costs of management actions and increases the likelihood of success (Karlsson et al., 2013). EU regulations require that Member States have sufficient capacity for diagnosis of animal and plant pathogens and pests (EU, 2016a, b). In addition, there are requirements for having National Reference Laboratories and designated EU Community Reference Laboratories for diagnosing animal and plant pathogens and pests (EU, 2017;European Commission, 2018b). At these laboratories, it should be possible to identify most pathogens that could be used in an agroterror attack or introduced by neglect of the regulations. In the United States, the National plant disease diagnostic network (NPDN) is a valuable platform for information exchange, collaboration between diagnostic laboratories, and other activities that strengthen national biosecurity (Stack et al., 2014).
In addition to the identification of pathogens and diseases, there are methods for determining if a disease outbreak is of natural or malicious origin, and whether an organism was genetically modified or not (Elworth et al., 2020;Lewis et al., 2020;Pilch et al., 2020). Unless a perpetrator's goal is to claim responsibility for an action, technologies that can attribute perpetrators to malicious actions, may have a deterrent effect, and lead to conviction of crime or diplomatic penalties (Lewis et al., 2020).
During a crisis event, the capacity of individual laboratories could quickly be exceeded by the large number of samples to be analysed. To address this problem, laboratory networks that strengthen capacity and capabilities through, for example, sharing sample analysis duties can be formed. This was useful in Sweden during naturally occurring outbreaks of anthrax in 2008 and 2011 and during the Covid-19 pandemic in 2020-2021 (Lindberg & Melin, 2016;SVA, 2021). Other important areas for enhancing bio-preparedness in national laboratories include ring trials, maintaining competence, sample transport and delivery, inter-agency communication and sharing of data (Thelaus et al., 2017).

Response to unwanted events
An efficient response to an agroterror attack or unintentional introduction requires a multidisciplinary and multisectoral approach that includes private and government health services, law enforcement agencies and crisis management services (Anand, 2018;Farsang et al., 2013). If an agricultural pathogen is introduced into a country, regardless of how and why, the response must be quick to contain the disease (EU, 2016a; Gamliel & Fletcher, 2017a). For some pathogens, the contingency plans for control measures are laid out by EU directives, while for others they are developed on a national level. To reduce the likelihood of disease spread, authorities can, for example, set up zones around an infected holding where animal, plant, vehicle, and equipment movements are restricted (European Commission, 2020c). During an outbreak, and depending on its nature and extent, collaboration with and assistance from others, such as police, customs, transportation authorities, food authorities, and county and municipalities, are needed to contain the disease. An efficient response requires that all actors are prepared to work collaboratively, and that good relationships and communication mechanisms between law enforcement, and plant, animal and human health officials are developed before an incident occurs (FBI, 2008;CDC, 2011;Thelaus et al., 2017;Anand, 2018). It is important to test agricultural disease contingency plans, regularly, through exercises involving these authorities.

Step V-Reporting
The fifth step in a risk and vulnerability analysis describes the work process. It includes who participated in the work, what sources of information were used, and their reliability. This step is important for future analyses and revisions. It is preferably described in the beginning of a written report, even if it cannot be completed before the completion of steps I-IV (Winehav & Nevhage, 2011).
The number of risks is immense, even within a limited function for society, and quantifications are required for prioritizing potential mitigation strategies (Robb, 2017). When there is a lack of data behind an analysis, as when analysing a potential agroterror event, only likelihoods with high uncertainty and estimated frequencies can be presented. The uncertainty of the likelihood and how to address it, needs to be communicated. The prospects of the recipients understanding the results increases if descriptions and quantifications of risks and consequences are uniform and if communication is targeted to different end user groups (Utne et al., 2008;Gilioli et al., 2017;EFSA PLH Panel et al., 2018;Månsson et al., 2019;Lin et al., 2017;Rydmark et al., 2020).

Step VI-Communication and continuation
The last step in the FORSA framework is where risk management actions identified in previous steps are evaluated and feed-back is given. This leads to decisions about how to proceed within the function for society, the use and communication of results to stakeholders, what actions to take and their implementation in order to minimize risks for unwanted events and mitigate those that may occur. To ensure that results are transferred, and actions are taken, assignments and responsibilities are given to named persons or organizations. The decisions and actions in step VI are valuable for step I in future analyses of the same or related functions.
A complicating factor in the case of an unintentional or agroterror event is that several actors will work in parallel. Plant or animal health authorities will manage the disease outbreak, diagnose and confirm its cause, determine its origin and thereby assist the law enforcement authorities in the identification of the perpetrator (Lewis et al., 2020). Simultaneously, security and law enforcement authorities will investigate criminal or terror implications. Public health authorities will also be involved if an attack is carried out using a zoonotic agent or a toxin-producing plant pathogen. The effectiveness of often-overlapping investigations could be limited by differences in cultures of information sharing and performance of work tasks, legal hurdles, vocabulary and definitions, and poor communication and understanding between agricultural health, public health, and law enforcement officials. This could lead to prolonged disease outbreaks or failures to gather evidence to identify and prosecute a perpetrator (CDC, 2011;Knutsson et al., 2012).
To build trust and improve working relationships and understanding between those who would be involved in the event of an agroterror attack, people from intelligence, law enforcement, diagnostic laboratory networks, public, animal and plant health could be brought together in seminars, workshops, collective improvisations, and table-top exercises (Frykmer et al., 2018;Lindberg & Melin, 2016;Mårtensson et al., 2013;Stack et al., 2014). All efforts to promote inter-agency cooperation, such as functioning communication, understanding of differences in work routines, and clarified expectations, are necessary when managing acts of agroterrorism and biohazard events. At the international level, there are initiatives for capacity building. A good example of efforts to improve resilience to agroterrorism is the OIE-FAO-INTERPOL project containing the bio-threat detection module SET (Vasconcelos Gioia et al., 2021).

Other frameworks or approaches for risk and vulnerability analysis
The overall goal with a risk and vulnerability analysis is to understand the nature of risks, strengthen crisis preparedness, reduce vulnerabilities, and increase the abilities to manage crises (Winehav & Nevhage, 2011). There are often separate analytical frameworks and models for risk analysis, and vulnerability. Analytic and conceptual vulnerability models are tools for understanding hazards and identifying risk management approaches (Fontaine & Steinemann, 2009;Luers et al., 2003;Turner et al., 2003). The analytical frameworks described in Fig. 5 have similar structures and are applicable to agricultural production. Their differences seem larger due to the labelling of the steps. An analysis of the activities in individual steps may show more similarities. The largest differences are whether the frameworks are cyclic, if and how the function for society (I) is addressed, and how reporting and communication (VI) take place. While some are shown as a linear flow (e.g., van Tuyll, 2013;Gamliel et al., 2017b), it should not imply an absence of feed-back of outcomes or experiences for the following event. Cyclic frameworks imply an ongoing process with built-in learning components. Several disciplines will be involved in a risk and vulnerability analysis and most of the frameworks presented here have built-in flexibility to manage this, as well as to include new actors or components in an ongoing analysis.
The frameworks presented by the World Organisation for Animal Health (OIE) and International Plant Protection Convention (IPPC) are informative concerning the risks for animals and plants (IPPC, 2007;OIE, 2019). Their advantage is the use of globally agreed standards for risk analysis.  (FAO, 2007) is cyclic. It is focused on risk management and less on the identification of unwanted events in the function for society of concern. C: The risk management guidelines by ISO (Anon, 2018) are aimed for managing risks by building risk management frameworks in organizations. It is a circular framework, and an example of where risk analysis is described as part of the risk assessment. D: The risk management framework for plant biosecurity describes in eight steps the activities and responsibilities in the management process following an outbreak in 'real time' (Gamliel et al., 2017b). E and F: Risk and vulnerability analyses are also found in engineering sciences (Aven, 2007) and in the Dutch National Security Strategy (van Tuyll, 2013). G: A common analytical framework developed for benefit cost analysis in risk assessments by Ayyub et al. (2007), addresses the importance of quantifications and inclusion of interdependencies in complex systems. Created with BioRe nder. com These frameworks have three steps 1) Hazard identification, 2) Risk assessment, and 3) Risk management, using the OIE terminology. Communication is continuous as in the ISO framework (Anon, 2018). The socioeconomic impacts are included, i.e., benefit cost analyses of the alternatives for risk mitigation. For analysis of unwanted events in steps I-III of the FORSA framework, it is worth considering using the outlines from IPPC and OIE.
Unlike some of the analytical frameworks for risk and vulnerability analysis, the FORSA approach is proactive. It starts by describing the function for society at risk instead of with the unwanted event. It provides background and helps to identify possible unwanted events by scrutinizing the function itself. The circular structure aids in developing action plans and strengthens the ability to handle and prepare for crises. With the built-in feedback loop, the learning outcome from one analysis feeds into, and facilitates, a recurring one. A stronger focus may be placed on specific steps, depending on the overall aim of the analysis, and returns to previous steps are possible. In a FORSA risk and vulnerability analysis, several predefined scales and assessment criteria are used, and the sources of information and uncertainties of the assessments are reported. The systems that typically are analysed with this framework are more narrowly defined than the system addressed in this review.

Final remarks
If the societal costs of crop and animal diseases were acknowledged, this information could justify direction of more resources for surveillance and action plans to manage serious events. In general, plant disease management has a lower public profile than animal health (Ilbery et al., 2012;Waage & Mumford, 2008;Wilkinson et al., 2011). Animals also represent a higher investment value than annual crops (Waage & Mumford, 2008). The effects and consequences of an unwanted event within agricultural primary production would vary depending on i) whether it is accidental or intentional, ii) the magnitude of the damage, iii) the capability to maintain production capacity of the affected commodity and ensure food security, and iv) the extent of the outbreak when detected. Clearly structured frameworks are useful for analysing unwanted events, their effects and what actions that would be required for reducing their damage. Here we used the FORSA framework as one approach to analyse threats to animal and crop production in agriculture.
A risk and vulnerability analysis that addresses activities within agricultural production soon becomes multidisciplinary. Even an analysis of a narrowly defined function within agricultural production, covers several disciplines, involving farmers, local and national authorities (including police and customs), plant and animal health experts, politicians, and legislators. The drivers for emergence of new pathogens need to be identified and analysed on a multidisciplinary scale (Anand, 2018;Richardson et al., 2016). Due to this complexity, a systems approach is required in the risk and vulnerability analysis itself and in all decisions (political, practical, legislative) concerning risk management. As an example, the BSE crisis, during 1988-2012 with a feed-transmitted epidemic in cows and zoonotic transfer to humans (variant Creutzfeldt-Jakob disease) through contaminated beef or beef products (Salman et al., 2012), showed that a systems approach was necessary to capture all the complexities that challenged the control efforts.
Multidisciplinarity involves the challenge of differences in the professional languages within the group working with a risk and vulnerability analysis. Working through the steps in a well-structured framework facilitates the mutual understanding, and abilities to communicate between disciplines. The understanding of terms, definitions, organizational structures, and differences in working cultures would be promoted by active efforts to improve inter-agency collaboration. All actors should be familiar with the individual steps in analyses and assessments at the onset of the work process, and an ideal way of testing management strategies and inter-agency collaboration is to have real-time exercises (Knutsson, 2017;Mackelprang & Friedman, 2021).
Communication that fails when facing a real biohazard could create more problems than the hazard itself. As an example, the police and customs, and national control agencies need the capability to identify and investigate agroterror crime as well as negligence of regulations. There may be a lack of impetus for further improvements, since few events have been linked to agroterrorism, and it could be difficult to maintain a certain alert level if the events are rare.
To improve the preparedness and ability to manage both natural and intentional biohazards, risk management options include bolstering farm-level biosecurity, increasing laboratory capacity, and developing and improving methods for diagnosis and early detection. For a country to be better prepared, agroterror or biohazard scenarios should be included in disease risk analyses, and inter-agency communication enhanced. There is a need for willingness and motivation to allocate resources, and actions to improve national appreciation for domestic agricultural production.
There will always be a demand for personnel with the special skills to identify pathogens and diagnose diseases. This emphasizes the need to maintain high levels in education and regular training of personnel. Education in fundamental elements, such as in-field or on-farm disease diagnosis, pathogen biology and disease epidemiology, should include the ability to handle exotic diseases, disasters and agroterrorism (Chomel & Marano, 2009;Dunning et al., 2009;Fletcher et al., 2020;Kelly, 2005;MacDonald et al., 2009;Mackelprang & Friedman, 2021).
There is a need for openness for unexpected events. One can prepare for the emergence of new plant and animal diseases, zoonoses, and pandemics, but not predict when and where they will appear (Meslin, 1997;Webster & Walker, 2003;WHO, 2004;Ali et al., 2017). Taking the expected behaviour of other plant and animal pathogens for granted may come with surprises, as will the effects of climate change (Shaw & Osbourne, 2011;Casadevall, 2017;Garret et al., 2021). With the Covid-19 pandemic, war in Ukraine and persistent global supply chain problems during 2020-2022, the context for risk and vulnerability analyses exemplified by the FORSA framework has become more complicated. We foresee more difficulties with the supply of necessary inputs in agriculture, but also food processing, including supply disruptions of pesticides, fertilizers, minerals and vitamins. The Ukraine war is associated with more expensive and uncertain supplies of cereals and inputs such as fertilizers.
The risk of an agroterror attack is low but not negligible. However, the potential for agroterror events will increase as the 4th industrial revolution technologies are implemented in agriculture (e.g., Koblentz, 2020). These technological systems may advance digitization processes, be commercially available and able to be repurposed for malicious use (e.g., Bajema, 2018). Major economic fallout is foreseen as a result of an agroterrorist act. Other consequences include erosion of public trust in government to protect national security and provide safe food and increased public health costs. Human lives could also be at risk if the chosen pathogen is zoonotic or produces toxin.
We conclude that agroterrorism must be included when threats to society are considered. Our food supply chain becomes more fragile with more unknowns, thereby increasing the needs for risk and vulnerability analyses of which FORSA is one example. European countries have strong agricultural protection systems in place and coupled with efficient diagnostic laboratory networks, these can respond to natural emergencies of crop and livestock diseases. There are, however, several areas where improvements could be made to enhance the level of preparedness for both natural and intentional spread of pathogens. Although biased towards Europe and conditions in the EU, the principles presented here have applications world-wide.

Annex 1 Characteristics of agroterrorism agents
In order to be considered a major threat for antagonistic use pathogens must possess certain characteristics*: • Pathogenic for livestock, poultry, or food crops • Highly infectious and contagious in the target population • Easy to acquire or produce • Easy to disseminate • Good ability to survive in the environment • Low level of immunity in the target population • Produces a predictable pattern of clinical disease in the target population • Attributable to a natural outbreak, ensuring plausible deniability • Not harmful to the perpetrator • Produces extremely negative biological, social and/or economic effects • Used in previous attacks or previously developed as part of bioweapons program *Summarized from: Wilson et al., 2000;Clarke & Rinderknecht, 2011;Menrath et al., 2014. Annex 2 Examples of pathogens that constitute serious threats to food security Author contribution The paper was initiated by Anders Kvarnheden, and written by Annika Djurle, Beth Young, and Jim Nygren. All coauthors contributed with comments and text to the manuscript.
Funding Open access funding provided by Swedish University of Agricultural Sciences.

Conflicts of interest
The authors declared that they have no conflicts of interest.
Research involving human and animal rights Data involving human or animal subjects are present in publications and other publicly available sources.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.