From a system dynamic perspective this could suggest that the studied farming systems have some buffering capacity to deal with disturbances . An example of this is the farm expansion in area and number of animals in many farming systems that compensates for the loss of farms from the system. From a methodological perspective, it could be argued that the participatory assessment of critical thresholds of challenges is easier than for system functions and resilience attributes. Critical thresholds of challenges are linked to important function indicators and resilience attributes and, therefore, may serve as warnings in the mental models of farming system stakeholders. Based on workshop results and further reflections, interactions between critical thresholds are expected to directly affect the economic viability at farm level, a central critical threshold observed in all farming systems .This gives another argument for monitoring income and other economic indicators in the monitoring frameworks such as the CMEF. The lack of a consistent pattern with regard to environmental thresholds indicates the importance of the local context. In all farming systems, exceeding the critical threshold for economic viability at farm level affects the attractiveness of the sector, the number of farm closures and the availability of farm successors, which in turn in about half of the case studies contribute to lower availability of labor and/or depopulation, which finally can reinforce low economic viability. Hence, a vicious cycle is initiated. This suggests that processes related to the economic and social domain can be driving dynamics of farming systems as well as being reinforced by those dynamics. This potentially can turn a relatively slow social process into a fast process. Social processes are therefore indeed important to monitor . This is already acknowledged in, for instance, in DE-Arable&Mixed, where participants emphasized the attractiveness of the area, hydroponic dutch buckets specifically regarding the development of infrastructure. Through its interactions with processes in other domains and levels, economic performance can be seen as an indirect driver as well as a warning signal for approaching critical thresholds in other domains and levels.
In all farming systems food production was perceived to directly impact economic viability. Therefore, from the perspective of many farming system actors participating in our workshops, focus on food production and economic viability , which are based on relatively fast and measurable processes , seems often more justified than focusing on the more slowly developing social functions such as providing an attractive countryside. However, this may be due to the fact that farmers were in most case studies the best represented stakeholder group, thus possibly masking the voices of other stakeholder groups that were represented less. In any case, social and environmental functions should not be overlooked as a focus on one domain will likely lead to missing important interactions with critical thresholds in other domains . For example, improving economic viability through scale enlargement and intensification, meaning fewer farms and often replacing labor by technology, often leads to a less attractive countryside. Regarding the environmental domain, focus on economic farm performance can even be dangerous as it could ignore externalized risk. For instance in UK-Arable and NL-Arable soil quality, the base of crop production and hence economic performance, was considered close to critical thresholds, while prohibition of certain crop protection products was seen as a challenge for the farming system, rather than the damage these products cause to surrounding ecosystems. Another example of externalized risk in one of our case studies is the pollution of water bodies in IT-Hazelnut. On their own, farmers may initially not have the willingness or capacity to look beyond the farm level. In IT-Hazelnut, farmers, through interaction with environmental actors, are now addressing these environmental issues. Building on this example, we argue that for instance societal dialogues and policy deliberations on improving sustainability and resilience need input from specific social and environmental actors, possibly even from outside the farming system. This seems necessary to counter-balance the bias towards economic performance at farm level by most of the participating farming system actors in most of our workshops. In the more remote case studies, e.g. DE-Arable&Mixed and BGArable, attractiveness of the area seems low anyway. Consequently, improving prices alone, for instance, may not improve the availability of the necessary labor, thus reducing the emphasis on economic performance. Extensive rural development seems necessary to maintain the functioning of these farming systems. Mitter and et al. , based on their mechanistic scenario development approach, expected no or negative developments regarding rural development in all future scenarios of EU agriculture.
The notion that both mechanisms at EU and farming system level are not wired to address rural development, shows how the low attractiveness of an area can persist once it has come about. Avoiding exceedance of critical thresholds without further adaptation or transformation, implies a performance at or below the current low to moderate levels for most system function indicators and resilience attributes . A potential exceedance of a critical threshold in the coming ten years is expected to lead to negative developments for most system function indicators and resilience attributes. Negative developments of function indicators are expected in the economic, social as well as the environmental domain. On average, across all farming systems, we did not observe any differences in the magnitude of the effect between domains for function indicators. This consistent development confirms the idea that the different domains are interacting. The consistent expected developments for function indicators and resilience attributes after exceeding critical thresholds suggest a perceived interaction between them. One could argue that a system needs resources to react to shocks and stresses , especially for adaptation and transformation. These resources can only be adequately realized when there is an enabling environment and when system functions are performing well. The other way around, resilience attributes can be seen as “resources” to support system functions on the way to more sustainability. For instance, existing diversity of activities and farm types makes visible what works in a specific situation, openness of a system helps to timely introduce improved technologies, and connection with actors outside the farming system may help to create the enabling environment for innovations to improve system functioning . Impact of challenges is primarily experienced at the farm level,resulting in the disappearance of farms from the farming system. In multiple case studies , participants indicated that identified critical thresholds would be perceived differently among farmers. As mentioned before, farm closure generally leads to a less attractive countryside, a long-term process that is currently not perceived the most important issue in most studied farming systems, according to stakeholder input. Increasing farm size could be seen as a solution to compensate for the loss of farms and farmers in the farming system. Increasing the farm size is often associated with the advantage of economies of scale. For multiple farming systems in our study , production margins are low, which could further stimulate this thinking. However, from the farm level perspective, beyond a certain size, further economies of scale are not realized in some of the studied farming systems, i.e. there are limits to growth dependent on the rural context. In BE-Dairy, for instance, increasing farm size seems to be limited due to environmental standards. In ES-Sheep, further reduction of the farmer population is perceived to be harming the farming system, e.g. through reduction of facilities such as farmer networks, agricultural research initiatives, etc., but also hospitals, schools, etc. Besides, to further increase farm size, farmers in ES-Sheep depend on extra labor that is not available because of low attractiveness of the countryside, bato bucket while investment in labor saving technology does not pay off with the current market prices.
This is an example of the reflection of Kinzig et al. that a seemingly reversible threshold becomes irreversible because a certain management option to reverse processes is not available anymore. Based on Fig. 1, we argue that this specific example may be true for more farming systems where a lack of labor force is experienced and investment in labor saving technology are not likely to pay off . The importance of the social domain of farming systems makes us argue that indicators in this domain should be monitored. The option for countries in CAP2021-27 to shift 25% of the budget from income support to rural development provides the opportunity to adapt policies and investments to rural development needs. For instance for the more remote farming systems such as DE-Arable&Mixed and BGArable. We argue that a large shift of budget across the two pillars is already an indication of the perceived need to improve rural living conditions and can thus be used for monitoring. Although relating to economic values, the allocation of budget to rural development can thus be seen as the importance that is attributed to support processes in the social domain. Caution is needed however, as Pillar II also supports processes related to the environmental domain. Surveys among experts at national and regional level that record how much of the budget should be shifted from pillar I to II is a further step in assessing the performance of farming systems in the social domain. This implies introducing subjectivity in the CMEF on the evaluation side, while the choice of the parameter is defined objectively, i.e. externally. Jones remarks that objectively defined and subjectively evaluated resilience assessments are relatively robust, easy and quick, while the limitations lay mainly in having to deal with bias, priming and social desirability. Other possibilities for objectively defined and subjectively evaluated indicators may lie in including indicators on living conditions and quality of life in rural areas based on Eurofound studies . These type of indicators also have the advantage of being entirely in the social domain, i.e. they don’t indirectly refer to economic values such as the shift in budget from Pillar I to Pillar II as discussed above.
A common reflection in the discussion section so far is that having adequate system resources seems essential for stimulating system resilience attributes and dealing with challenges. In cases of low farming system resilience, building system resources may initially depend largely on external resources. This implies a role for regional, national and EU government bodies, i.e. a pro-active role for actors in the institutional domain outside the farming system. Given the tendency to focus on economic performance at farm level, external resources in the form of economic subsidies should be increasingly conditional regarding environmental and social functioning of the farming system. The emphasis on resources for building resilience is also acknowledged in several recent resilience frameworks , for instance with regard to knowledge and innovation systems . To elaborate on the example of AKIS, we argue that, rather than only monitoring and evaluating the amount of budget and the number of people that benefit from improved AKIS , also the amount of this resource and stakeholders’ access to it should be known and evaluated regularly. Similarly, other social and institutional resources need to be monitored next to economic and environmental resources. Given the challenges regarding assessing and discussing critical thresholds in workshops , all identified critical thresholds could be seen as “Thresholds of potential concern” . In our case these TPCs would express the concerns of a selection of farming system stakeholders. TPCs can be seen as a set of evolving management goals that are aimed at avoiding critical thresholds that are expected, e.g. from experiences in other systems, but are not known. In case thresholds are considered beforehand as TPC’s, Q-methodology may be an interesting participatory method to define which TPC deserves most priority. Estimating main functions of a system by assessing critical thresholds as TPCs, reduces the presence of clear sustainability goals. This makes the threshold assessment less dependent on externally determined values and criteria than most sustainability assessments . Implicitly, the goal is to avoid a decline in sustainability and resilience levels of the current system, which may give the participating system actors the trust to provide details, expose interrelatedness between sustainability domains, and also come up with solutions. Regarding the latter, it should be noted that avoiding exceedance of critical thresholds does not automatically imply that a system is steering away from mediocre performance.