Livestock farms evolve in a disturbed, uncertain, and challenging context. Climate change impacts the quality and yields of crops and pastures (Olesen and Bindi 2002) and livestock health (Gauly et al. 2013), which requires livestock farmers to develop adaptation and mitigation strategies (Rojas-Downing et al. 2017). Livestock products are sold on competitive globalized markets, and volatile prices result in a lack of economic viability for most farmers (Garrido 2016, Schulte et al. 2018). Besides environmental and economic instability, livestock farmers must also respond to new societal concerns about animal welfare and the environmental impacts of their practices (Caracciolo et al. 2016). This uncertain context challenges the ability of farms, as social-ecological systems, to cope with disturbances of variable intensity, which may result in consequences that remain over the long term (Dedieu and Ingrand 2010).
The concept of social-ecological resilience has gained interest in the field of agricultural science. Consequently, the resilience framework has been adapted to farms as complex and dynamic systems (Darnhofer 2014). A resilient farm is able to maintain its productive functions over the long term despite disturbances (e.g., sudden shocks, unpredictable “surprises,” slow-onset changes). Integrating the buffer capacity (i.e., ability to absorb disturbances without modifying the state of the system), adaptability (i.e., ability to adapt to current or potential future disturbances), and transformability (i.e., ability to make drastic changes and redesign a farm completely) enables farms to be resilient (Darnhofer 2014).
As in other fields of application (e.g., ecology, engineering sciences, psychology), the concept of resilience applied to agricultural systems remains abstract and multidimensional, and thus difficult to render operational (Cumming et al. 2005). The few studies that have assessed resilience in agroecosystems often focused on the evolution of objective performance indicators such as crop yields (Li et al. 2019) and economic profitability (Groot et al. 2016). However, these resilience indicators remained at the subsystem scale (e.g., field) and lacked a holistic view of farms as units that consist of multiple, interrelated subsystems (e.g., land, flock, farmer). These assessments also often ignored the social component of farms. Cabell and Oelofse (2012) developed an indicator framework that includes the social component by using indicators related to the ability to learn from experiences or to build human capital. This framework later served as a basis for the SHARP farm-resilience assessment tool (Diserens et al. 2018). However, this tool focuses on farm resilience to a single type of disturbance—climate change—whereas holistic approaches are required to assess farm resilience over the middle-to-long term and to a diversity of disturbances (e.g., economic crises, life risks) (Kaseva et al. 2019, Meuwissen et al. 2020).
Farmers lie at the heart of farms as social-ecological systems. Each farmer has his/her own perceptions of the world and thus his/her own subjectivity (i.e., cognitive ability to effectively self-evaluate their farm’s resilience). Because previous assessments of household/farm resilience have ignored this subjectivity, Jones and Tanner (2017) recommended assessing subjective resilience in addition to applying objective methods. Doing so requires giving relevance to farmers’ cognitive self-evaluation of their farms’ resilience over time. Farmers can be questioned about subjective farm resilience in many ways, each with its own advantages and biases, from fully qualitative approaches based on semi-structured interviews to more quantitative and structured approaches based on multiple-choice questions (Jones and Tanner 2017). For both approaches, collecting information about subjective resilience remains challenging due to ambiguities surrounding the term.
We studied the subjective resilience of organic dairy sheep farms that produce milk in the area of Roquefort cheese production in Aveyron, France. This case study of livestock farms was chosen for two main reasons:
In this study, we (i) monitored the evolution of farm structure and farming practices over long timeframes (5 years and more), (ii) described subjective resilience of organic dairy sheep farms to a variety of disturbances, and (iii) characterized relations among the evolution of farm structure, farmers’ practices, and subjective farm resilience to identify driving factors behind resilience.
We defined the resilience of organic dairy sheep farms as their ability to cope with all types of disturbances over the long term while successfully maintaining their productive functions (Walker et al. 2004). We built the dynamic component of the analytical framework using two bodies of literature: studies of livestock farming system trajectories and social-ecological resilience, especially farm resilience (Fig. 1). Both consider the evolution of farms over time as a process of responding to internal or external disturbances (Moulin et al. 2008, Darnhofer 2014). To understand this dynamic process further, it is recommended to identify stable periods (i.e., “consistent phases”) (Davoudi et al. 2012). Each period is characterized by temporal consistency in farm structure and the farmer’s management practices and objectives. In each period, the farmer relies mainly on the farm’s buffering capacity and occasionally on minor adaptations (e.g., purchasing fodder) to cope with disturbances. The period lasts until the farmer’s objectives and farm performances no longer match, or when the impacts of disturbances force the farmer to adapt or transform the farming system to obtain new temporal consistency. We thus considered farm trajectories as a sequence of farmer-defined periods that corresponded to relative stability in structure and management practices despite disturbances and changes. We did not study the transition between two periods in detail, but instead considered long-term trends of variables to identify key attributes of farm resilience (e.g., incorporating diversity (Dumont et al. 2020), increasing autonomy (Heiberg and Syse 2020)).
We built the subjective resilience component of our analytical framework based on the literature on job satisfaction (Green 2010) and global life satisfaction (Cohn et al. 2009). Between 2005–2016, the number of livestock farms in the European Union decreased by 37.6% (European Commission 2019). Key factors for this decrease include inadequate income and the working conditions of livestock farming (Maucorps et al. 2019), which decrease job satisfaction and often overall life satisfaction. Lack of job satisfaction is associated with an increase in quitting (Green 2010), which holds true for the farming sector. Thus, monitoring farmers’ satisfaction over the long term appears to be a simple way to determine subjective farm resilience to multiple disturbances. Armitage et al. (2012) demonstrated the potential of well-being, which is strongly related to job satisfaction (Green 2010) and overall life satisfaction (Cohn et al. 2009), to reflect subjective resilience holistically. Previous studies highlighted interactions between resilience and well-being (Greenhill et al. 2009, Armitage et al. 2012). Thus, we assumed that the evolution of farmers’ satisfaction with the condition of their land, flock, and economic and social life over time can be used as a proxy of farm resilience to multiple disturbances over the middle-to-long term. Thus, farms with satisfactory land and flock conditions and that provide a fair income and good working conditions over the long term are better able to maintain their productive functions despite disturbances. In contrast, farmers’ dissatisfaction compromises the long-term continuity of farming and progressively erodes the identity of farms as production units. We thus considered subjective farm resilience as a long-term trend that maintains or improves farmers’ satisfaction. We then related these trends to variables that represented farm structure and farmers’ management practice to identify driving factors behind subjective resilience.
Aveyron is located on the southern border of the Massif Central, France. It is the main sheep milk production region in France, where Roquefort cheese is produced. Roquefort cheese was the first PDO product in France in 1925. Although the PDO is supposed to ensure high and stable milk prices for farmers, conventional Roquefort cheese production has encountered many difficulties, including a 15% decrease in production from 2007–2017 (Spelle and Daudé 2019). At the same time, costs of conventional milk production have increased (Institut de l’élevage (IDELE) 2018). Faced with this unstable situation, conventional farmers have increasingly converted their farms to organic farming (Fig. 2b), in part due to a higher and increasing price for sheep milk (Fig. 2a). This trend has increased the percentage of organic dairy sheep farms in the region (Fig. 2c).
Along with this expansion, the organic dairy sheep sector faced a number of disturbances. Like many other livestock farms in France, organic dairy sheep farms in Aveyron experienced a sudden increase in input prices after the economic crisis of 2008–2009. The climate in Aveyron is Mediterranean, which limits land productivity, and climate change has resulted in frequent summer droughts, which decreased pasture yields and worsened the farms’ already low self-sufficiency in fodder (Vial 2017). These farms also experienced the recent arrival of wolves in the area, along with unexpected international political disagreements (e.g., the Chinese embargo on Roquefort cheese in 2017 threatened production and the market (Boffet 2017)).
We interviewed 36 organic dairy sheep farmers in Aveyron (Fig. 3) in winter 2017–2018. Interviewers came from several organizations (i.e., a Chamber of Agriculture, two farmers’ associations, an agricultural high school, and a research institute) involved in the “Résilait” project (Institut de l’agriculture et de l’alimentation biologiques (ITAB) 2017). The status of the interviewers varied (i.e., researchers, advisers, veterinarians, interns, and students). Interviewers helped design the survey guide, from defining the list of variables of interest to formulating the questions. All interviewers were trained in using the survey guide to ensure homogeneity in the survey process.
Each organization was responsible for finding farms within its own network and for contacting farmers. To consider only farms with a “routine” organic system, farms had to have converted to organic at least 5 years before the survey. The organizations had known the sampled farms for several years and had sometimes monitored their technical and/or financial characteristics. The organizations selected only farms that delivered milk to dairies (i.e., not those that processed their own milk) and that had proven their ability to maintain their productive functions despite the disturbances they had encountered since they began their conversion to organic. For example, these organizations knew the extent to which farms had been affected technically and economically by the 2017 summer drought and how they had coped with that event. Ultimately, the 36-farm sample represented 38% of the 95 local organic dairy sheep farms that had been certified organic for at least 5 years (Agence Bio 2018). Until this study, data on organic dairy sheep farms had been limited to a 15-farm network in the Massif Central (De Boissieu et al. 2020). Compared with these, the farms surveyed in this study had fewer permanent workers and livestock units (LU) but more utilized agricultural area (UAA) (Table 1).
During face-to-face interviews, farmers first divided the history of their farms into several stable periods on a timeline. Starting from their conversion to organic farming, farmers thus identified different stable periods. Farmers then added values to the timeline for four farm structure variables and 20 agricultural practice variables (Table 2). The variables were defined by the group of researchers, advisers, teachers, and farmers (other than those who were surveyed) who contributed to the project. They are standard agricultural variables necessary to understand the overall functioning of livestock farms and were used in previous studies of the vulnerability and resilience of organic dairy farms in France (Bouttes et al. 2019, Perrin et al. 2020a).
Farmers also added values to the timeline to indicate their level of satisfaction with each of four dimensions defined by the project group. For each stable period, farmers were asked how satisfied they were at that time and rated this level on a scale of 1–4 (“very unsatisfied”, “unsatisfied”, “satisfied,” or “very satisfied”, respectively) to clearly distinguish satisfactory from unsatisfactory situations and decrease the attractiveness of a “neutral” answer (Krosnick et al. 2001). Farmers were not provided with any indicators (e.g., yield, income) to support their assessment beyond the names of the dimensions:
For each stable period, the four dimensions were summed to calculate the response variable “overall satisfaction” (OverallSat), which was used as a proxy for farm resilience:
Overall Satisfaction= Land Satisfaction +Animal Satisfaction+ Economic Satisfaction+ Social Satisfaction
This summing created a dynamic and complex proxy for resilience that is consistent with long-term observations of social-ecological systems such as farms, in accordance with Carpenter et al. (2005). Each satisfaction dimension had the same weight when calculating overall satisfaction. Even though a farmer’s relative preferences in the four dimensions may have changed over time, weighting the four dimensions for each stable period would have made data collection far more complex.
To assess subjective farm resilience to multiple disturbances, farmers described the disturbances they had experienced since their conversion to organic farming and specified the years when these disturbances had occurred. The interviewer classified disturbances into six types: climatic (e.g., droughts, hailstorms, wet years with poor harvests), economic (e.g. economic crises, increases in input prices), health related (e.g., disease outbreaks, abortions, parasitism), organizational (e.g., departure of an associate, personal health problem), technical (e.g., machine breakdown), and “other.” Interviewers also provided a timeline that specified the main disturbances in Aveyron over the previous 10 years. It was used as a reference to identify whether farmers cited, and thus experienced, these reference disturbances (Fig. 4). Farmers were interviewed at their home or workplace for 2–3 hours, which was the most suitable format for collecting the data and reconstructing the farm history to accurately reflect farmers’ statements and descriptions of stable periods.
The disturbances experienced by social-ecological systems, the response to these disturbances, and the capacity for adaptive action are common elements when applying the concepts of vulnerability and resilience (Adger 2006) to farms. Martin et al. (2017) developed a method to assess the vulnerability of farms to climatic and economic variability. We adapted this method to the resilience framework. The method consists of performing one linear regression for each variable of each farm throughout the stable periods the farmer identified. We used the lm() function (fitting linear models) of the stats package of R software to perform the linear regressions (Ballesteros 2008). This first step generated both intercepts and slopes (called “I.<Variable name>” and “Ev.<Variable name>”, respectively; Table 2). The slopes showed how farm structure, farmers’ practices, and farmers’ overall satisfaction evolved over time. The intercepts estimated the state of the farm and the farmer during the period of conversion to organic farming. These mathematical parameters were then included in several iterations of partial least squares (PLS) analysis and sparse PLS (sPLS) analysis performed with the pls() and spls() functions of the mixOmics package of R software (Append. 1) (Tenenhaus 1998, Lê Cao et al. 2008). This method provided a holistic approach to relate the response variable—evolution of overall satisfaction—to explanatory variables that were divided into intercepts and slopes. The PLS results indicated the extent to which models were predictive (Q² values) and explanatory (R² values). We always performed the sPLS after PLS to obtain clearer and more robust results. The sPLS, which selects variables by introducing LASSO penalization (Lê Cao et al. 2008), improved the quality of the models. The PLS and sPLS enabled us to assess the dynamic aspect of the evolution of farm structure and farmers’ practices over time (i.e., more than 5 years) and relate it to the evolution of the proxy of resilience.
Farmers in the sample identified two to four stable periods that lasted 1–25 years each over a 5- to 33-year timeframe (Fig. 4). For each farm, the first stable period included the conversion to organic (Conversion Period (CP)), as the retrospective survey started on the conversion date, and the last stable period ended in winter 2017–2018, the time of the survey (Survey Period (SP)). If necessary, farmers identified up to two intermediate periods (IP1 and IP2, respectively) between the CP and SP. Farms had to respond to a variety of disturbances over this long timeframe. Disturbances that farmers had experienced during the 10 years preceding the survey were compared with those reported by advisers, which were used as a reference. The type of disturbance mentioned most often (80 times) was climatic. Summer droughts in Aveyron had been severe during the previous 20 years, and farmers considered those in 2003 (9), 2007 (12), 2011 (21), and 2017 (23) as major disturbances. The drought in 2017 concerned mainly farms in southern Aveyron and had been compounded by late frosts that damaged lucerne fields. The second-most mentioned type of disturbance was economic (57), particularly two main events: the global economic crisis in 2008 (17) and delays in common agricultural policy payments in 2016 (24). The third-most mentioned disturbance was related to flock health (46), which were most frequent in 2017 (13) due to predation by wolves in plateau areas. As organizational and technical disturbances are components of the daily management of farms, they were mentioned consistently among years and stable periods. For the 36 farms, one stable period changed to another 72 times, and most changes (60%) did not coincide with disturbances. For the remaining 40% of changes, most (62%) coincided with a combination of two or more disturbances.
During the SP, farmers’ overall satisfaction was high, with a mean (± 1 standard deviation) score of 11.96 ± 1.69 (out of 16) at the survey period (Fig 5a). In addition, 25 of those farms had a score ≥ 12 (i.e., high satisfaction with the overall situation) in the same period. The mean evolution of overall satisfaction from the CP to SP was positive but variable: 0.11 ± 0.26/year. After the CP, overall satisfaction increased on the majority of farms (20 out of 36) after the conversion period (Fig.5b). Overall satisfaction decreased for eight farms (-0.34 to -0.03 for the value of the evolution of overall satisfaction). The remaining eight farms maintained the same level of overall satisfaction over time (i.e., presented a null value for the evolution of farmers’ overall satisfaction). Evolution in the economic, animal, social, and land dimensions of satisfaction differed in their contributions to the mean evolution of overall satisfaction (64%, 27%, 9%, and 0%, respectively) (Fig. 5c).
For farm structure, the number of ewes increased in 19 farms (Table 3), remained the same in nine farms, and decreased in eight farms. Across the 19 farms that increased their number of ewes, the increase was on average by 20.2 ± 23.1 ewes/year. Two farms drove the increase in both mean and variability because their farmers purchased a second farm and merged a second flock into the first one in the SP. The number of LU also tended to increase, but the number of permanent workers tended to remain the same, which resulted in an increasing trend in LU per worker. The UAA increased on 21 farms (maximum = +36.3 ha/year) and remained the same on 14 farms. As increases in UAA and flock size were proportional, the mean stocking rate remained the same (0.00 ± 0.02 LU/ha/year). The workload per hectare also tended to increase.
For farming practices, the mean duration of the grazing period decreased slightly (-0.04 months/year), but it was driven by five farms that shortened their grazing period the most (-0.60 months/year). The duration of the grazing period remained the same on most farms (29/36) and increased on only two farms. The start date of lambing remained the same on 16 farms, was advanced on 14 farms, and was delayed on six farms. Concentrate-use efficiency differed more among farms. After conversion, 21 farmers decreased the mean amount of concentrates distributed to ewes (by -115.5 g/L milk/year), five farmers provided the same amount, and the remaining 10 farmers increased this amount. The mean evolution of concentrate use efficiency thus decreased but remained highly variable (-11.8 ± 23.6 g/L milk/year). This trend had little influence, however, on the large amount of concentrates distributed in this context, which reached a mean of 564 ± 263 g/L milk/year during the SP. Ewe productivity increased on 27 farms, remained the same on four farms, and decreased on five farms. Mean ewe productivity thus increased by 4.5 ± 5.6 L milk/ewe/year.
The sPLS model with the highest and most stable quality on component 1 (i.e., Q² > the ad hoc threshold of 0.0975; R² = 0.27, the highest obtained) predicted the evolution of overall satisfaction mainly with three explanatory variables: estimated overall satisfaction during the CP (i.e., the intercept predicted by the model), the evolution of ewe productivity and the evolution of the duration of the grazing period (association scores of -0.59, 0.52, and -0.47, respectively) (Fig. 6). Thus, increasing ewe productivity and decreasing the duration of grazing increased overall satisfaction in our sample.
As an illustration of these results, farmers’ overall satisfaction increased the most (+0.65/year) for F31, after starting at a high estimated level during the CP (10.58) (Table 4). This increase was related to a larger increase in ewe productivity than the sample’s mean increase (+18.46 and +4.55 L milk/ewe/year, respectively) and a larger decrease in the duration of the grazing period than the sample’s mean decrease (-0.29 and -0.04 months/year, respectively). This is characteristic of the main results of the sPLS. In contrast, overall satisfaction decreased (-0.15/year) for F21, despite starting at a high level during the CP (12.56). Ewe productivity decreased (-2.89 L milk/ewe/year), unlike the mean increase for the sample. The duration of its grazing period was long (8.00 months) and remained the same over time.
Overall, the results showed an increasing trend in farm size (i.e., ewes, LU, and UAA) and a relationship between increasing ewe productivity and overall satisfaction. However, three farms did not follow this trend completely. On F2, F9, and F13, the number of ewes decreased (-6.7, -4.8, and -20.0/year, respectively), and ewe productivity decreased on F2 and F9 but increased on F13 (+4.0/year). For all three farms, the decrease in the number of ewes was a direct consequence of the farmers’ strategies. When these farmers divided their farms’ histories into stable periods, they specified that the main objectives that had guided their management decisions included “simplifying the system,” “having free time,” “reducing flock size while producing a living for three households.” As a result, these evolutions resulted in an increase in overall satisfaction (which increased from 9/16 to 14/16 from CP to SP for both F2 and F9) or a slight decrease in an already high overall satisfaction (from 13 to 12 for F13).
Our study is the first to relate the evolution of subjective farm resilience to changes in farm structure and farming practices on organic dairy sheep farms in France and to identify driving factors behind this subjective resilience. Farmers’ overall satisfaction was used to provide subjectivity to the resilience assessment, which has been a major disadvantage of resilience assessments to date (Jones 2019). Estimating overall satisfaction as the sum of individual satisfaction with their land, flock, economics, and social situation enabled the farm to be considered in a holistic manner, with its diversity of subsystems and performances (Darnhofer et al. 2012). Using slope to characterize the evolution of explanatory and response variables over long timeframes addressed the dynamic aspect of resilience, as required in any resilience assessment (Carpenter et al. 2001). The wide timeframes defined by our sampling protocol (at least 5 years under organic farming) smoothed situational biases and avoided an overly positive assessment of non-resilient farms. A farmer may have overestimated his/her level of satisfaction for a given period, but if difficulties persisted, they would have emerged in the long term and been reflected in the evolution of overall satisfaction.
Nevertheless, our results should be interpreted and generalized with caution. Assessing resilience remains a complex task (Quinlan et al. 2016), and using a single indicator based on farmers’ remembered levels of satisfaction resulted in some cognitive biases:
The driving factors of subjective farm resilience assessed via farmers’ satisfaction we identified highlighted an orientation of organic sheep milk production toward a form of conventionalization, with organic farming becoming a slightly modified version of modern conventional agriculture (De Wit and Verhoog 2007). This phenomenon is highlighted among other things by the specialization of farms and the intensification of production (Guthman 2004), which generally implies increased reliance on off-farm inputs, e.g., livestock feed. We observed a dominant trend toward an increase in ewe productivity and a high consumption of concentrates per L milk and a sharp decrease in the duration of the grazing period on a few farms. For example, on the one hand, F31 experienced the largest increase in farmer overall satisfaction, but on the other hand, it consumed a lot more concentrates per L of milk compared with a network of French organic sheep farms (336 and 223 g/L milk, respectively; De Boissieu et al. 2018) although productivity levels were also higher. Further increases in ewe productivity would require breeding progress and increasing the energy and protein density of the ewe diet, which is eased by distributing concentrates. Thus, this productivity paradigm could progressively promote increased dependence on feed inputs in organic sheep milk production, which was clearly identified as an indicator of conventionalization of organic animal production (Darnhofer et al. 2010). Conventionalization was also evident in flock management. The starting date of lambing (the fourth-strongest variable associated with the evolution of overall satisfaction on sPLS component 1; see Append. 2) was progressively advanced and moved away from the “natural” behavior of ewes. Some sheep farmers implemented that practice to produce more milk outside the “natural” milking season to obtain higher milk prices. The consequence is a lengthening of the indoor feeding period, which has the highest rate of concentrate distribution to compensate for hay’s lower feed value compared with grazed pasture. The three farms that had followed alternative pathways to the increasing trend in farm size (i.e., ewes, LU, and UAA) did not differ much from the general trend regarding agricultural practices with, e.g., the duration of grazing decreasing on one farm and remaining stable on the two others. Conventionalization thus appears as an indirect outcome of driving factors that lead to higher farmers’ satisfaction.
Given these facts, using farmers’ satisfaction as a proxy of subjective farm resilience might incompletely reflect the long-term resilience of farms as social-ecological systems; identified driving factors of satisfaction are sometimes at odds with indicators and factors of social-ecological resilience previously reported in the literature. Practices compromising a responsible use of local resources are at odds with the resilience indicators “Coupled with local natural capital” and “Globally autonomous and locally interdependent” proposed by Cabell and Oelofse (2012). Increasing ewe productivity and feed distribution in stalls while remaining highly reliant on purchased concentrates and decreasing the use of pasture corresponds to what Darnhofer et al. (2016) called the “pathology of ‘command and control’”, which is identified as a threat to farm resilience. Shortening the grazing period, which often occurs when pastures are used less often, changes how urine and manure are spread on a farm and thus increases imbalances in soil fertility among farm fields. In the long term, it decreases a farm’s ability to ecologically self-regulate (Cabell and Oelofse 2012). A shorter grazing period can also encourage woody plants to encroach on areas that are suitable only for grazing, which are then abandoned to focus on producing feed in the most productive pastures. This process can degrade the traditional legacy of landscape management in Aveyron (Cabell and Oelofse 2012). Although dairy intensification and higher productivity and efficiency can increase income, they have social and environmental costs (Clay et al. 2019) In the long term, the trend toward conventionalization and the related productivity paradigm could affect farms’ ability to cope with shocks, as observed by Sinclair et al. (2014) for dairy cattle farms. This calls for caution when using farmers’ satisfaction as a proxy of subjective farm resilience.
Using the evolution of farmers’ overall satisfaction as a proxy for farm resilience allowed us to focus on the subjective dimension of resilience and highlight its driving factors, mainly increasing ewe productivity and reducing the duration of the grazing season, enabling organic dairy sheep farms to cope with a variety of disturbances. We also identified biases in the use of this resilience proxy. The emphasis on subjectivity in a particular context of high-priced agricultural products might induce optimistic biases. Using satisfaction might also overlook the long-term resilience of the social-ecological systems that farms are. Satisfaction as a proxy for resilience also highlighted contradictions between farmers’ discourse and the evolution of practices implemented on the farms. In a sample of 128 farms (including the 36 farms in the present study), we observed that farmers identified the search for autonomy, the reduction of pressure on farm resources, and the search for diversity as resilience factors (Perrin et al. 2020b). The context of high milk prices and the observed evolution of agricultural practices on organic dairy sheep farms in Aveyron did not match those assertions. Thus, assessing farm resilience based on the evolution of farmers’ satisfaction requires caution in interpreting the results. Long-term subjective assessments should be supplemented with more objective long-term monitoring of slow variables that reflect farm resilience (e.g., soil organic matter content, flock genetic potential against drought, parasite and disease outbreaks, farmers’ knowledge).
The driving factors increasing subjective resilience highlighted in this study are relatively specific to French dairy sheep farms in a context of high milk prices due to organic and PDO certification. These results may not be relevant for European countries with lower milk prices, such as Greece or Spain. Nevertheless, these findings reflect the trend of conventionalization of the organic sector observed outside of France (Ramos García et al. 2018) and can raise concrete policy issues. Sensitizing farmers to resilience issues that result from this conventionalization should be on the political agenda in rural regions where maintaining agricultural activities is a concern.
We thank the farmers for welcoming us to their farms and taking the time to answer our questions. This study was funded by the French Ministry of Research (Ph.D. fellowship of A.P.) and the Résilait project. We thank Alexandre Bancarel, Stéphane Doumayzel, Mathilde Durand, Olivier Patout, Camille Andrieu, and the groups of students coordinated by Séverine Cassel for conducting the interviews. We also thank Catherine Experton and Jerôme Pavie as Résilait project leaders.
The data/code that support the findings of this study are available on request from the corresponding author, [A.P].
The data/code are not publicly available as they contain information that could compromise the privacy of research participants.
A part of the code and of the statistical analysis that support the findings of this study are available in the 'supplementary materials'
Adger, W. N. 2006. Vulnerability. Global Environmental Change 16(3):268-281. https://doi.org/10.1016/j.gloenvcha.2006.02.006
Agence Bio. 2018. Aller plus loin: Les données brutes, données départementales. In “Les chiffres clés.” Agence Bio, Paris, France. [online] URL: https://www.agencebio.org/vos-outils/les-chiffres-cles/
Allen, C. R., H. E. Birge, D. G. Angeler, C. A. Arnold, B. C. Chaffin, D. A. DeCaro, A. S. Garmestani, and L. Gunderson. 2018. Quantifying uncertainty and trade-offs in resilience assessments. Ecology and Society 23(1): 3. https://doi.org/10.5751/ES-09920-230103
Armitage, D., C. Béné, A. T. Charles, D. Johnson, and E. H. Allison. 2012. The interplay of well-being and resilience in applying a social-ecological perspective. Ecology and Society 17(4): 15. https://doi.org/10.5751/ES-04940-170415
Ballesteros, S. 2008. Le modèle linéaire avec R : fonction lm(). Muséum national d’histoire naturelle, Paris, France. [online] URL: https://rug.mnhn.fr/semin-r/PDF/semin-R_lm_SBallesteros_110308.pdf
Biggs, R., M. Schlüter, D. Biggs, E. L. Bohensky, S. BurnSilver, G. Cundill, V. Dakos, T. M. Daw, L. S. Evans, K. Kotschy, A. M. Leitch, C. Meek, A. Quinlan, C. Raudsepp-Hearne, M. D. Robards, M. L. Schoon, L. Schultz, and P. C. West. 2012. Toward principles for enhancing the resilience of ecosystem services. Annual Review of Environment and Resources 37(1):421-448. https://doi.org/10.1146/annurev-environ-051211-123836
Boffet, L. 2017. Le roquefort n’est plus le bienvenu en Chine. Franceinfo: 3 occitanie. [online] URL: https://france3-regions.francetvinfo.fr/occitanie/aveyron/rodez/roquefort-n-est-plus-bienvenu-chine-1325557.html
Bouttes, M., N. Bize, G. Maréchal, G. Michel, M. S. Cristobal, and G. Martin. 2019. Conversion to organic farming decreases the vulnerability of dairy farms. Agronomy for Sustainable Development 39(2): 19. https://doi.org/10.1007/s13593-019-0565-3
Bradburn, N. M., L. J. Rips, and S. K. Shevell. 1989. Answering autobiographical questions: the impact of memory and inference on surveys. Science 236(4798):157-161. https://doi.org/10.1126/science.3563494
Bradford, R. A., J. J. O’Sullivan, I. M. van der Craats, J. Krywkow, P. Rotko, J. Aaltonen, M. Bonaiuto, S. De Dominicis, K. Waylen, and K. Schelfaut. 2012. Risk perception—issues for flood management in Europe. Natural Hazards and Earth System Sciences 12(7):2299-2309. https://doi.org/10.5194/nhess-12-2299-2012
Cabell, J. F., and M. Oelofse. 2012. An indicator framework for assessing agroecosystem resilience. Ecology and Society 17(1): 18. https://doi.org/10.5751/ES-04666-170118
Caracciolo, F., G. Cicia, T. Del Giudice, L. Cembalo, A. Krystallis, K. G. Grunert, and P. Lombardi. 2016. Human values and preferences for cleaner livestock production. Journal of Cleaner Production 112:121-130. https://doi.org/10.1016/j.jclepro.2015.06.045
Carpenter, S. R., F. Westley, and M. G. Turner. 2005. Surrogates for resilience of social-ecological systems. Ecosystems 8(8):941-944. https://doi.org/10.1007/s10021-005-0170-y
Carpenter, S., B. Walker, J. M. Anderies, and N. Abel. 2001. From metaphor to measurement: resilience of what to what? Ecosystems 4(8):765-781. https://doi.org/10.1007/s10021-001-0045-9
Clay, N., T. Garnett, and J. Lorimer. 2019. Dairy intensification: drivers, impacts and alternatives. Ambio 49:35-48. https://doi.org/10.1007/s13280-019-01177-y
Cohn, M. A., B. L. Fredrickson, S. L. Brown, J. A. Mikels, and A. M. Conway. 2009. Happiness unpacked: positive emotions increase life satisfaction by building resilience. Emotion 9(3):361-368. https://doi.org/10.1037/a0015952
Cumming, G. S., G. Barnes, S. Perz, M. Schmink, K. E. Sieving, J. Southworth, M. Binford, R. D. Holt, C. Stickler, and T. Van Holt. 2005. An exploratory framework for the empirical measurement of resilience. Ecosystems 8(8):975-987.
Darnhofer, I. 2014. Resilience and why it matters for farm management. European Review of Agricultural Economics 41(3):461-484. https://doi.org/10.1093/erae/jbu012
Darnhofer, I., D. Gibbon, and B. Dedieu. 2012. Farming systems research: an approach to inquiry. Pages 3-31 in I. Darnhofer, D. Gibbon, and B. Dedieu, editors. Farming systems research into the 21st century: the new dynamic. Springer Netherlands, Dordrecht, The Netherlands. https://doi.org/10.1007/978-94-007-4503-2_1
Darnhofer, I., T. Lindenthal, R. Bartel-Kratochvil, and W. Zollitsch. 2010. Conventionalisation of organic farming practices: from structural criteria towards an assessment based on organic principles. A review. Agronomy for Sustainable Development 30(1):67-81. https://doi.org/10.1007/978-94-007-0394-0_18
Davoudi, S., K. Shaw, L. J. Haider, A. E. Quinlan, G. D. Peterson, C. Wilkinson, H. Fünfgeld, D. McEvoy, L. Porter, and S. Davoudi. (Davoudi, S., and L. Porter, editors.) 2012. Resilience: a bridging concept or a dead end? “Reframing” resilience: challenges for planning theory and practice. Interacting traps: resilience assessment of a pasture management system in northern Afghanistan. Urban resilience: what does it mean in planning practice? Resilience as a useful concept for climate change adaptation? The politics of resilience for planning: a cautionary note. Planning Theory and Practice 13(2):299-333. https://doi.org/10.1080/14649357.2012.677124
De Boissieu, C., E. Morin, O. Patout, T. Taurignan, N. Rivemale, L. Drochon, S. Doumayzel, and G. Noubel. 2020. Les exploitations ovines laitières du massif central en agriculture biologique. Résultats campagne 2018. Collectif BioRéférences et Pôle AB massif central, Lempdes, France.
De Boissieu, C., O. Patout, N. Rivemale, L. Drochon, S. Doumayzel, and G. Noubel. 2018. Les exploitations ovines laitières du massif central en agriculture biologique. Résultats campagne 2016. Collectif BioRéférences. Pôle AB massif central et Pôle AB Massif central, Lempdes, France.
Dedieu, B., and S. Ingrand. 2010. Incertitude et adaptation : cadres théoriques et application à l’analyse de la dynamique des systèmes d’élevage. INRAE Productions Animales 23(1): 81–90.
De Wit, J., and H. Verhoog. 2007. Organic values and the conventionalization of organic agriculture. NJAS - Wageningen Journal of Life Sciences 54(4):449-462. https://doi.org/10.1016/S1573-5214(07)80015-7
Diserens, F., J. Choptiany, D. Barjolle, B. Graeub, C. Durand, and J. Six. 2018. Resilience assessment of Swiss farming systems: piloting the SHARP-Tool in Vaud. Sustainability 10(12):4435. https://doi.org/10.3390/su10124435
Dumont, B., L. Puillet, G. Martin, D. Savietto, J. Aubin, S. Ingrand, V. Niderkorn, L. Steinmetz, and M. Thomas. 2020. Incorporating diversity into animal production systems can increase their performance and strengthen their resilience. Frontiers in Sustainable Food Systems 4: 109. https://doi.org/10.3389/fsufs.2020.00109
European Commission. 2019. Agri-environmental indicator- livestock patterns. Statistics explained. Eurostat, Luxembourg City, Luxembourg.
Eurostat. 2017a. Eurostat statistics explained. Glossary utilised agricultural area (UAA). Eurostat, Luxembourg City, Luxembourg.
Eurostat. 2017b. Eurostat statistics explained. Glossary livestock unit (LSU). Eurostat, Luxembourg City, Luxembourg.
FranceAgriMer. 2020. Données économiques, agricoles et alimentaires. Enquête mensuelle laitière. Lait de Brebis. FranceAgriMer, Libourne, Montreuil, Volx, France.
Garrido, A. 2016. Agricultural markets instability: revisiting the recent food crises. Routledge, London, UK. https://doi.org/10.4324/9781315676265
Gauly, M., H. Bollwein, G. Breves, K. Brügemann, S. Dänicke, G. Daş, J. Demeler, H. Hansen, J. Isselstein, S. König, M. Lohölter, M. Martinsohn, U. Meyer, M. Potthoff, C. Sanker, B. Schröder, N. Wrage, B. Meibaum, G. von Samson-Himmelstjerna, H. Stinshoff, and C. Wrenzycki. 2013. Future consequences and challenges for dairy cow production systems arising from climate change in Central Europe—a review. Animal 7(05):843-859. https://doi.org/10.1017/s1751731112002352
Green, F. 2010. Well-being, job satisfaction and labour mobility. Labour Economics 17(6):897-903. https://doi.org/10.1016/j.labeco.2010.04.002
Greenhill, J., D. King, A. Lane, and C. MacDougall. 2009. Understanding resilience In South Australian farm families. Rural Society 19(4):318-325. https://doi.org/10.5172/rsj.3220.127.116.118
Groot, J., J. Cortez-Arriola, W. Rossing, R. Améndola Massiotti, and P. Tittonell. 2016. Capturing agroecosystem vulnerability and resilience. Sustainability 8(11): 1206. https://doi.org/10.3390/su8111206
Guthman, J. 2004. The trouble with “organic lite” in California: a rejoinder to the “conventionalisation” debate. Sociologia Ruralis 44(3):301-316. https://doi.org/10.1111/j.1467-9523.2004.00277.x
Hassan, E. 2005. Recall bias can be a threat to retrospective and prospective research designs. Internet Journal of Epidemiology 3: 2.
Heiberg, E. J., and K. L. Syse. 2020. Farming autonomy: Canadian beef farmers reclaiming the grass through management-intensive grazing practices. Organic Agriculture 10:471-486. https://doi.org/10.1007/s13165-020-00291-6
Institut de l’élevage (IDELE). 2018. Les chiffres clés du GEB. Ovins 2018. Productions de lait et de viande. Institut de l’élevage, Laxou, France.
Institut de l’agriculture et de l’alimentation biologiques (ITAB). 2017. RESILAIT : Résilience des systèmes laitiers biologiques ; optimisation des facteurs de compétitivité et mise au point de systèmes plus efficients dans la gestion des risques à venir. Institut technique national dédié à la recherche en agriculture biologique. ITAB, France. [online] URL: http://www.itab.asso.fr/programmes/resilait.php
Jones, L. 2019. Resilience isn’t the same for all: comparing subjective and objective approaches to resilience measurement. Wiley Interdisciplinary Reviews: Climate Change 10(1): e552. https://doi.org/10.1002/wcc.552
Jones, L., and T. Tanner. 2017. “Subjective resilience”: using perceptions to quantify household resilience to climate extremes and disasters. Regional Environmental Change 17(1):229-243. https://doi.org/10.1007/s10113-016-0995-2
Kaseva, J., S. J. Himanen, and H. Kahiluoto. 2019. Managing diversity for food system resilience. Pages 1-32 in D. Barling, editor. Advances in food security and sustainability. Elsevier, Cambridge, Massachusetts, USA; London, UK. https://doi.org/10.1016/bs.af2s.2019.07.001
Krosnick, J. A., A. L. Holbrook, M. K. Berent, R. T. Carson, W. M. Hanemann, R. J. Kopp, R. C. Mitchell, S. Presser, P. A. Ruud, V. K. Smith, W. R. Moody, M. C. Green, and M. Conaway. 2001. The impact of “no opinion” response options on data quality. Public Opinion Quarterly 66(3):371-403. https://doi.org/10.1086/341394
Lê Cao, K.-A., D. Rossouw, C. Robert-Granié, and P. Besse. 2008. A sparse PLS for variable selection when integrating omics data. Statistical Applications in Genetics and Molecular Biology 7(1): 35. https://doi.org/10.2202/1544-6115.1390
Li, M., C. A. Peterson, N. E. Tautges, K. M. Scow, and A. C. M. Gaudin. 2019. Yields and resilience outcomes of organic, cover crop, and conventional practices in a Mediterranean climate. Scientific Reports 9(1): 12283. https://doi.org/10.1038/s41598-019-48747-4
Martin, G., M.-A. Magne, and M. San Cristobal. 2017. An integrated method to analyze farm vulnerability to climatic and economic variability according to farm configurations and farmers’ adaptations. Frontiers in Plant Science 8: 1483. https://doi.org/10.3389/fpls.2017.01483
Maucorps, A., A. Münch, S. Brkanovic, and B. Schuh. 2019. The EU farming employment: current challenges and future prospects. European Parliament, Policy Department for Structural and Cohesion Policies, Brussels, Belgium.
Meuwissen, M., W. Paas, T. Slijper, I. Coopmans, A. Ciechomska, J. Deckers, W. Vroege, E. Mathijs, B. Kopainsky, H. Herrera, R. Finger, Y. D. Mey, P. M. Poortvliet, P. Nicholas-Davies, M. Vigani, D. Maye, J. Urquhart, A. Balmann, F. Appel, K. Termeer, P. Feindt, J. Candel, M. Tichit, F. Accatino, S. Senni, E. Wauters, I. Bardají, B. Soriano, C.-J. Lagerkvist, G. Manevska-Tasevska, H. Hansson, C. Gavrilescu, and P. Reidsma. 2020. Report on resilience framework for EU agriculture Work Performed by P1 (WU) in cooperation with all partners. SURE-Farm, Wageningen, The Netherlands. [online] URL: https://www.surefarmproject.eu/wordpress/wp-content/uploads/2018/02/SURE-Farm_Deliverable-D1.1-Resilience-Framework.pdf
Milestad, R., and I. Darnhofer. 2003. Building farm resilience: the prospects and challenges of organic farming. Journal of Sustainable Agriculture 22(3):81-97. https://doi.org/10.1300/J064v22n03_09
Moulin, C.-H., S. Ingrand, J. Lasseur, S. Madelrieux, M. Napoléone, J. Pluvinage, and V. Thénard. 2008. Comprendre et analyser les changements d’organisation et de conduite de l’élevage dans un ensemble d’exploitations : propositions méthodologiques. Pages 181-196 in B. Dedieu, E. Chia, B. Leclerc, C. H. Moulin, and M. Tichit, editors. L'élevage en mouvement : flexibilité et adaptation des exploitations d'herbivores, Editions Quae, Versailles, France.
Olesen, J. E., and M. Bindi. 2002. Consequences of climate change for European agricultural productivity, land use and policy. European Journal of Agronomy 16(4):239-262. https://doi.org/10.1016/S1161-0301(02)00004-7
Perrin, A., M. S. Cristobal, R. Milestad, and G. Martin. 2020a. Identification of resilience factors of organic dairy cattle farms. Agricultural Systems 183: 102875. https://doi.org/10.1016/j.agsy.2020.102875
Perrin, A., R. Milestad, and G. Martin. 2020b. Resilience applied to farming: organic farmers’ perspectives. Ecology and Society 25(4): 5. https://doi.org/10.5751/es-11897-250405
Quinlan, A. E., M. Berbés-Blázquez, L. J. Haider, and G. D. Peterson. 2016. Measuring and assessing resilience: broadening understanding through multiple disciplinary perspectives. Journal of Applied Ecology 53(3):677-687. https://doi.org/10.1111/1365-2664.12550
Ramos García, M., G. I. Guzmán, and M. González De Molina. 2018. Dynamics of organic agriculture in Andalusia: moving toward conventionalization? Agroecology and Sustainable Food Systems 42(3):328-359. https://doi.org/10.1080/21683565.2017.1394415
Raphael, K. 1987. Recall bias: a proposal for assessment and control. International Journal of Epidemiology 16(2):167-170. https://doi.org/10.1093/ije/16.2.167
Rojas-Downing, M. M., A. P. Nejadhashemi, T. Harrigan, and S. A. Woznicki. 2017. Climate change and livestock: impacts, adaptation, and mitigation. Climate Risk Management 16:145-163. https://doi.org/10.1016/j.crm.2017.02.001
Schulte, H. D., O. Musshoff, and M. P. M. Meuwissen. 2018. Considering milk price volatility for investment decisions on the farm level after European milk quota abolition. Journal of Dairy Science 101(8):7531-7539. https://doi.org/10.3168/jds.2017-14305
Simon, M., S. M. Houghton, and K. Aquino. 1999. Cognitive biases, risk perception, and venture formation: how individuals decide to start companies. Journal of Business Venturing 15(2):113-134. https://doi.org/10.1016/s0883-9026(98)00003-2
Sinclair, K., A. Curtis, E. Mendham, and M. Mitchell. 2014. Can resilience thinking provide useful insights for those examining efforts to transform contemporary agriculture? Agriculture and Human Values 31(3):371-384. https://doi.org/10.1007/s10460-014-9488-4
Spelle, C., and C. Daudé. 2019. Chiffres clés 2017 des produits sous signes de la qualité et de l’origine. Produits laitiers AOP et IGP. Version corrigée de juin 2019. Institut National de l'Origine et de la Qualité (INAO), Paris, France.
Tenenhaus, M. 1998. La régression PLS: théorie et pratique. Editions Technip, Paris, France.
Toro-Mujica, P., A. García, A. G. Gómez-Castro, R. Acero, J. Perea, V. Rodríguez-Estévez, C. Aguilar, and R. Vera. 2011. Technical efficiency and viability of organic dairy sheep farming systems in a traditional area for sheep production in Spain. Small Ruminant Research 100(2-3):89-95. https://doi.org/10.1016/j.smallrumres.2011.06.008
Toro-Mujica, P., A. García, A. Gómez-Castro, J. Perea, V. Rodríguez-Estévez, E. Angón, and C. Barba. 2012. Organic dairy sheep farms in south-central Spain: typologies according to livestock management and economic variables. Small Ruminant Research 104(1-3):28-36. https://doi.org/10.1016/j.smallrumres.2011.11.005
Vial, M. 2017. État des lieux des fermes ovins lait bio en Aveyron. Fédération Nationale de l'Agriculture Biologique, France. [online] URL: https://www.produire-bio.fr/articles-pratiques/etat-lieux-fermes-ovins-lait-bio-aveyron/ https://doi.org/10.1016/s0992-5945(20)30149-5
Walker, B., C. S. Holling, S. R. Carpenter, and A. P. Kinzig. 2004. Resilience, adaptability and transformability in social-ecological systems. Ecology and Society 9(2): 5. https://doi.org/10.5751/ES-00650-090205