A multi-stakeholder participatory study identi ﬁ es the priorities for the sustainability of the small ruminants farming sector in Europe Animal The international journal of animal bioscience

The European small ruminants (i.e. sheep and goats) farming sector ( ESRS ) provides economic, social and envi- ronmental bene ﬁ ts to society, but is also one of the most vulnerable livestock sectors in Europe. This sector has diverse livestock species, breeds, production systems and products, which makes dif ﬁ cult to have a clear vision of its challenges through using conventional analyses. A multi-stakeholder and multi-step approach, including 90 surveys, was used to identifyand assess themain challenges for thesustainabilityof the ESRS to prioritize ac- tions.Thesechallengesandactionswereidenti ﬁ edbyESRSexpertsincludingfarmers,cooperatives,breeding associations, advisers and researchers of six EU countries and Turkey. From the 30 identi ﬁ ed challenges, the most relevantwereeconomy-relatedchallengessuchas ‘ uncertaintyofmeatandmilkprices ’ , ‘ volatility ofcommodity prices ’ , ‘ low farm income ’ , ‘ highsubsidy dependency ’ and ‘ uncertainty infuturechangesinsubsidies ’ resulting in ‘ a sector not attractive to young farmers ’ . Most of these challenges were beyond the farmer ’ s control and per-ceivedas dif ﬁ culttoaddress.Challenges were prioritizedusing anindex, calculatedbymultiplying therelevance and the feasibility to address measures. The identi ﬁ ed challenges had a similar priority index across the whole sector with small differences across livestock species (sheep vs goats), type of products (meat vs dairy) and in- tensi ﬁ cation levels (intensive vs semi-intensive vs extensive). The priorities were different, however, between socio-geographical regions (Southern vs Central Europe). Some of the top prioritized challenges were linked to aspectsrelatedtotheproduction systems( ‘ lowpromotion oflocalbreeds ’ and ‘ slowadaptability ofhighproduc-ingbreeds ’ )andmarketpractices( ‘ unfairtrade/lackoftraceability ’ ).Themajorityoftheprioritychallenges,how-ever,wereassociatedwith ade ﬁ cientknowledgeor trainingatfarmlevel( ‘ poor businessmanagement training ’ , ‘ lack of professionalization ’ , ‘ slow adoption of innovations ’ ), academia ( ‘ researchers do not address real problems ’ ) andsociety asawhole ( ‘ low consumereducation inlocalproducts ’ , ‘ low socialknowledgeaboutfarming ’ , ‘ poor recognition of farming public services ’ ). Thus, improved collaboration among the different stakeholders across the food chain with special implication of farmers, associations of producers, academia and governments isneededtofacilitateknowledgeexchangeandcapacitybuilding.TheseactionscancontributetomakeESRSeco-nomically more sustainable and to adapt the production systems and policy to the current and future societal needs in a more region-contextualized framework. ﬁ ed using a multi-stakeholder participatory approach. Economy-related challenges were perceived as highly relevant, but out of the farmer ’ s control. The challenges that were identi ﬁ ed as priority based on their relevance and feasibility to be addressed were 1) enhancing consumer and social awareness, 2) farmer ’ s capacity building, 3) promoting local breeds and 4) re-focusing research priorities. It was also noted that effective collaboration and knowledge transfer across the different stakeholders involved in this sector are critical to address these challenges.


Introduction
The European small ruminants (i.e. sheep and goats) farming sector (ESRS) contributes to many of the Sustainable Development Goals described by the United Nations (Animal Task Force [ATF], 2019). This sector produces 6% of meat and 3% of milk production in the European Union (EU) representing an important share of the total livestock production in some European countries such as UK, Spain, Italy, France and Greece (Food and Agriculutre Organization of the United Nations [FAOSTAT], 2020). The ESRS supports 1.5 million farmers, plus members of the food chain, and provides economic and social cohesion for the most disadvantaged and depopulated rural areas (ATF, 2019). Furthermore, this sector delivers multiple environmental services such as wildfire prevention through grazing in dry regions during the hot summer season, naturalization of marginal areas and maintenance of landscape and biodiversity (Rook et al., 2004). All these ecosystem services, along with the quality of their products and ways of living in rural areas, contribute to the cultural heritage of many EU countries (Parente and Bovolenta, 2012).
Despite the importance of ESRS, it is one of the most vulnerable livestock sectors in Europe (Pulina et al., 2018). In recent decades, the ESRS suffered a steady decrease in animals and farms (mostly sheep) because of economic and structural difficulties (FAOSTAT, 2020). One key driver has been the constant decline in sheep and goat meat demand derived from changes in the consumers' habits. Young consumers living in urban areas prefer fast and easy-to-make recipes and diets with less fat and less red meat (Montossi et al., 2013;Mandolesi et al., 2020). Moreover, the ageing farmers population and the lack of new entrants provide little hope to the future of the ESRS (Pulina et al., 2018), unless the main constraints of the sector are identified and successfully addressed. Given the complexity of the ESRS, it is exposed to many challenges that are interconnected and apply to different levels (i.e. farm, regions, countries and EU). This means that farmers and stakeholders often have limited control over these challenges (Castel et al., 2011). Also, the diversity of livestock, products and level of intensification in the ESRS within and across countries makes the challenges and the identification of feasible solutions difficult (Ruiz-Morales et al., 2019).
Farm sustainability encompasses economic, social, cultural and environmental aspects (Lebacq et al., 2013). Views and perceptions of relevant players need to be considered to achieve significant outputs when analysing complex and multi-level systems. This requires multi-way interaction and co-production of knowledge between researchers, decision-makers and other beneficiaries of science (Francis and Goodman, 2010). In the case of the ESRS, farmers and industry views are critical for any analysis. Multi-stakeholder and participatory approaches are commonly used to integrate all existing views to identify and frame problems and solutions in socio-ecosystems in farming systems (Thiele et al., 2011;Martin-Collado et al., 2013).
In this context, we carried out a wide international multistakeholder study to provide a broad vision and new insights about current and future challenges that compromise the sustainability of the ESRS. This work was part of the H2020 iSAGE project 'Innovation for Sustainable Sheep and Goats in Europe' (www.isage.eu) which aimed to enhance the sustainability, competitiveness and resilience of the ESRS through collaboration between industry and academia. Specifically, the objectives of this study were 1) to identify the main challenges affecting the ESRS sustainability, 2) to assess whether these challenges are common across the production systems and 3) to prioritize the challenges according to both their relevance and the feasibility to be addressed. The results of this study can help the different members of the food chain and policy makers to prioritize the effort when addressing the main constraints of the ESRS.

Multi-stakeholder participatory approach
This study used an international multi-stakeholder and multi-step participatory process as previously validated in similar studies (Thiele et al., 2011;Martin-Collado et al., 2013). All 34 partners participating in iSAGE were involved in every phase of the process (Supplementary  Table S1). Partners included farmers' groups and cooperatives, breeding associations, National livestock associations, governmental research institutions and universities from six EU countries and Turkey (Table 1). These partners ensured the identified challenges were relevant to the real situation on the ground of the ESRS. The countries represented by the iSAGE industry partners group collectively accounted for 80% of the European sheep and 85% of the European goat censuses (EUROSTAT, 2018). The multi-stakeholder participatory approach included seven steps ( Fig. 1): 1) definition of the framework, 2) identification of challenges, 3) shortlisting of challenges, 4) assessment of the relevance of challenges, 5) assessment of the feasibility to address the challenge, 6) prioritization of challenges and 7) identification of the stakeholders needed to take action.
Step 1. Definition of the analysis framework The framework defined the boundaries of the analysis and aimed to analyse the challenges of ESRS through the central role that farmers have in the development of the sector. SWOT analysis is a widely used tool to explore the Strengths, Weaknesses, Opportunities and Threats of an organization with the final aim of adjusting its internal behaviour with the surrounding environment (Martin-Collado et al., 2013). Following the SWOT principles, challenges were divided into internal and external. Internal challenges (or weaknesses) were mostly controlled by the farmer, and external challenges (or threats) were mostly beyond the farmer's control such as environment and market factors.
Steps 2 and 3. Identification and shortlisting of challenges Challenges were identified using two consecutive steps. First, a preliminary list of challenges was prepared from a review of EU reports on 'The current situation and future prospects for the sheep and goat sector in the EU' [2017[ /2117], 'The future of the sheep meat and goat meat sectors in Europe' (IP/B/AGRI/IC/2007_043), 'The role of agriculture in rural development today' [Ares (2011) 1350301] and on iSAGE Deliverable 1.1 (www.isage.eu). Farmers' views were included in the process using semi-structured interviews of 47 sheep and goat farmers across 6 countries (Finland, France, Greece, Italy, Spain and UK). The interviews covered three general topics: 1) local and environmental changes affecting farm sustainability, 2) strengths, weaknesses, opportunities and threats and 3) future priorities for the development of ESRS.
Second, the soundness and completeness of the preliminary list of challenges were discussed by 45 iSAGE experts representing all project partners in a workshop in Bilbao, Spain. During the workshop, attendees were randomly distributed in three groups with similar representation of researchers, ESRS experts and organizations of farmers. The main outputs were shared in a closing debate and a final list of challenges was discussed and agreed.
Steps 4 and 5. Assessment of challenge relevance and feasibility to address This phase involved an online survey of 90 ESRS experts. This expert panel included representatives of all iSAGE partners plus the members of the National Advisory Committees of the seven countries participating in the project (Table 1). The National Advisory Committees were established at the beginning of iSAGE project and members were selected on the basis of having a wide knowledge in national and regional sector policy and practice relevant to the industries in their countries. The proportion of experts from each country was approximately equivalent to the country contribution to the European small ruminants census, with more representing sheep than goats, but having a similar representation the intensive, semi-intensive and extensive production systems. The group of experts included farmers groups (n = 14, including farmers, cooperatives, meat producers and cheese producing companies), breeding associations (n = 14, including technicians and farm advisers), researchers (n = 35, including experts on nutrition, production systems, genetics, health, reproduction, climate change and consumer behaviour) and National Advisory Committees (n = 27, including experts from agriculture departments, veterinarians and coordinators of R&D programs). The survey included three questions assessing experts' views on 1) the Relevance (R), 2) the Feasibility (F) to address each challenge and 3) the main stakeholder(s) that should be involved in addressing the challenges. Participants scored the relevance and feasibility to address the challenges using a scale for the Relevance (being 1 very low and 5 very high) and the Feasibility question (being 1 very difficult and 5 very easy).
Steps 6 and 7. Challenge prioritization and identification of the actors needed to take action Challenges were prioritized using a priority index (PI) calculated by multiplying the relevance and the feasibility for each challenge. Then, participants indicated which stakeholders would have to act to address each challenge from the following list: 1) government, 2) associations of producers, 3) farmers, 4) consumers, 5) retailers, 6) academia and 7) processing industry. Participants could select as many stakeholders as they considered necessary with the subsequent weight dilution as the number of choices increased.

Statistical analyses
Data distribution, regarding to challenge relevance, feasibility to address and PI, was assessed using the Shapiro-Wilk test resulting on a non-normal distribution (summary statistics in Supplementary   Table S2). Therefore, pair-wise contrasts were performed for each challenge using the Kruskal-Wallis non-parametric ANOVA with H adjusted for ties to evaluate the following factors: type of product [dairy (n = 61) vs meat (n = 29)], livestock species [sheep (n = 60) vs goats (n = 30)], production system [intensive (n = 25) vs semi-intensive (n = 31) vs extensive (n = 34)] and geographical region [Southern (n = 69) vs Central Europe (n = 21)]. Southern Europe was represented by Spain, Italy, Greece and Turkey, whereas Central Europe was represented by France, UK and Finland. To compare challenges in terms of relevance, feasibility to address and PI, a repeated measures mixed-model was performed considering the answers from each stakeholder as repeats. When significant effects were detected, means were compared by Fisher's protected LSD test. Effects were declared significant at P < 0.05. All statistical analyses were conducted using SPSS software (IBM Corp., Version 21.0, NY, USA).

European research priorities for the small ruminants sector
A final analysis compared the priorities identified in the present study with the EU research programs related to ESRS. Funding programs were found by searching the European Commission website (www. cordis.europa.eu) limiting the search to projects funded through the Framework Program 7 (FP7) and Horizon 2020 (H2020). These programs represent the main EU research funding over the last decade. The keyword combination for small ruminants was ('sheep' OR 'goat*' OR 'small ruminant*'). A similar search was done for the cattle sector (keywords: 'cow*' OR 'cattle' OR 'large ruminant*') that was used as a control. All projects identified were classified according to the thematic area based on the 'Domain of Application' described in the EU website. The same approach was used to search for publications in Web of Sciences. All the research publications published in the last 10 years were considered with the following keyword combinations in the title 'technology OR innovation', 'health OR welfare' and 'environment* OR climate change' along with the keywords described before to define the ruminant typologies.

Identification and grouping of challenges
A total of 30 challenges were identified by the multi-stakeholder participatory approach ( Table 2). The 13 internal challenges identified were classified into four categories ('Farmer level', 'Farming system', 'Sector level' and 'Overarching'), and the 17 external challenges were assigned into 6 categories ('Production factors', 'Environment', 'Market', 'Society', 'Policy' and 'Science').

Relevance of challenges
Despite the large range of variation observed across stakeholders (values from 1 to 5) in their perception of the relevance for each challenge (Supplementary Table S2), a moderate variation was observed on the average relevance across challenges (average 2.80-4.28). Significant differences were observed between challenges for their perceived relevance (P < 0.001, Table 3). All challenges, except 'parasites and infectious diseases', had scores above 3, indicating a medium-to-high relevance. Eight out of the top 10 challenges were external and four out of the five most relevant challenges were related to economy. The average relevance of internal challenges was the same for sheep and goats and productions systems but differed across geographical regions and type of products. The average relevance of external challenges only differed across farming systems. Overall, internal challenges were perceived as more relevant in Southern Europe and in the dairy sector while external challenges were perceived as more relevant in extensive than in intensive or semi-intensive systems.
The dairy systems were perceived to be more affected by internal challenges related to the structure of the sector (e.g. 'sector fragmentation' and 'low cooperation between farmers') and the farming system (e.g. 'low promotion of local breeds', 'low integration of livestock and agriculture' and 'low adaptability of high producing breeds' or 'slow adoption of innovations'). Experts from dairy systems also perceived higher relevance for 'low social knowledge about farming', 'low consumer education in local products', 'lack of traceability' and 'climate change threats', whereas the 'low consumer demand' was perceived as more relevant for meat production.
There was a similar overall perception of relevance for the external challenges across types of animal product (meat or milk). Sheep production was perceived to be more affected than goat production by 'high subsidy dependency', 'low female involvement', 'volatility of commodity prices' and 'low consumer demand, whereas goat production systems perceived higher relevance for 'low integration of livestock and agriculture', 'slow adoption of innovations', 'climate change threats' and 'limited access to land'. The average relevance of the external challenges decreased as the system's intensification increased. In particular, internal challenges such as 'low competitiveness' and 'low female involvement' and external challenges such as 'unfair trade/lack of traceability', 'climate change threats', 'poor recognition of farming public services', 'limited access to land' and 'wildlife conflicts' were perceived as more relevant in extensive and semi-extensive than in intensive production systems. The perception of relevance of many internal challenges was higher in Southern than Central Europe (P < 0.001), and in particular, 'lack of professionalization', 'slow adoption of innovations', 'sector fragmentation', 'low promotion of local breeds' and 'low adaptability of high producing breeds'.

Feasibility to address challenges
The perception of the feasibility to address the challenges showed a large range of variation across stakeholders (values from 1 to 5, Supplementary Table S2), whereas the average variation across challenges varied from 1.70 to 3.01. None of the 30 challenges investigated were perceived as easy or very easy to address ( Fig. 2) but significant differences were noted across challenges on the feasibility to be addressed (P < 0.001, Supplementary Table S3). The challenge 'researchers do not address real problems' was perceived as the easiest to address (score 3), whereas most challenges had a score between 2 and 3, indicating a medium-high difficulty for being addressed. Five external challenges were considered as very difficult to address (score below 2), most of them being external challenges related to market factors (e.g. 'volatility of commodity prices', 'uncertainty of meat and milk prices' and 'market monopolized'), as well as 'low farm income' and 'climate change threats'. The challenge 'not attractive to young farmers' was perceived as not only the most relevant internal challenge but also the most difficult to address. Contrary, addressing internal challenges related to farming practices ('parasite and infectious diseases', 'low adaptability of high productive breeds', 'low promotion of local breeds', 'poor business management training', 'slow adoption of technology' and 'lack of professionalization') and addressing external challenges related to education and sciences ('researchers do not address real problems', 'low consumer education in local products', 'environmental policy against intensification', 'low social knowledge about farming' and 'EU policy without scientific evidence') were perceived as more feasible. Overall, type of product, livestock species and level of intensification had minor effect on the overall perception of the feasibility to address the challenges. Most external challenges were perceived easier to address in Southern than in Central Europe.

Prioritization of challenges
The PI values for some challenges showed the maximum potential range of variation (from 1 to 25, Supplementary Table S2), but the average PI for each challenge showed a smaller variation (from 6.28 to 10.91) indicating that the most relevant challenges tended to be difficult to address and vice-versa (Fig. 2). Moreover, significant differences were noted across challenges for their PI (P < 0.001, Table 4). The top 10 priority challenges had equal representation between internal and external categories. Most of them were related to low awareness/training at different levels such as farm, consumers, researchers and society as a whole. On the contrary, the lowest PI values corresponded to four external challenges about 'climate change threats', 'market monopolized' and 'wildlife conflicts'. The PI of challenges differed between geographical regions while just minor differences were found across production systems, type of products and animal species. Southern Europe had higher PI values for several internal and external challenges mostly related to the sector's structure and insufficient farmer and consumer awareness, while Central Europe had higher PI for 'parasite and infectious diseases'.

Identification of the stakeholders that need to take action
The combined participation of several stakeholders (2.8 on average) was perceived to be required to address the internal challenges (Table 5). Among them, the highest percentage of influence was attributed to farmers (27%), associations of producers (26%) and governments (23%), followed by academia (14%), while minor involvement (<5%) was needed from the processing industry, consumers and retailers. An average of 3.2 stakeholders was needed to address the external challenges: the government was considered the most responsible to address them (31%), followed by association of producers (16%), farmers (15%), academia (12%), processing industry (11%), retailers (8%) and consumers (8%). However, some challenges required the involvement of specific stakeholders such as the processing industry for 'low education in local products' and 'lack of traceability', academia and consumers for 'low social knowledge about farming', 'poor recognition of farming public services' and 'researcher do not address real problems'. Across the 30 identified challenges, it was perceived that farmers only compiled an average of 20% of the total responsibility/capacity to address the challenges concerning farm sustainability, whereas the remaining 80% would involve external stakeholders.

European research priorities for the small ruminants sector
A total of 332 EU-funded research projects directly or indirectly connected with the ruminant sector were identified (Fig. 3). One-third Table 3 Relevance of the challenges for the European small ruminants sector across type of products, species, production systems and regions. a-c For the production system, means within the same row with different superscript differ at P < 0.05. 1 Relevance was based on a 1-5 scale. Within this column, means with different superscript differ at P < 0.05. 2 Abreviations: Int., intensive; Semi., Semi-intensive; Ext., Extensive; *** P < 0.001; ** P < 0.01; * P < 0.05.

Discussion
Collaborative multi-stakeholder process Our study involved a wide stakeholder platform including sheep and goats farmers, breeding associations, cooperatives, advisers, meat producers, cheese producing companies, representatives of national governments within Agriculture departments from seven countries, international farming training organizations and academics specialized in nutrition, genetics, health, reproduction management, climate change and consumers behaviour. These stakeholders prioritized challenges using a multi-step process which gave opportunity to identify needs for capacity building, knowledge gaps and research (Thiele et al., 2011;Drago-Severson and Blum-DeStefano, 2018). Working in heterogeneous and informal groups revealed disparity of perspectives among stakeholders about the current and future trends in the ESRS and the implications of potential adaptation measures. Despite the diversity of stakeholders involved in this study, there was an even representation from the different farm typologies and socio-cultural/regional contexts to have a balanced vision of the ESRS. This approach ensured that most relevant economic, social, environmental, policy and technical attributes were considered (Paraskevopoulou et al., 2020). Previous analyses of ESRS include the perception of individual stakeholders such as European Commission reports, academic studies focusing on a range of specific issues such as consumer preferences (Montossi et al., 2013), dairy products market (Pulina et al., 2018;Ruiz-Morales et al., 2019), technicaleconomic performances (Castel et al., 2011;Mena et al., 2017) or describing the sector situation within specific countries (Sossidou et al., 2013). The present study is one of the first international multistakeholder participatory diagnoses of the problems and giving particular attention to the views of farmers and farming-related institutions. This exercise is an example of 'a structured communicative process of linking scientists with selected stakeholders that are relevant for the research problem at hand' (Welp et al., 2006).

Relevance of the challenges
The challenges perceived as most relevant for the sustainability of ESRS were mainly linked to external economic aspects. These challenges are mostly beyond the control of farmers and farmers' institutions, since only 20% of the influence was assigned to the farmers. This situation reflects the weak position of farmers to deal with the most critical challenges for the ESRS sustainability. Therefore, different stakeholders, especially governmental institutions and retailers, need to be involved in strategic decisions.
The volatility and uncertainly of inputs and commodity prices were identified as relevant challenges for the sustainability of ESRS. While some variation in prices is considered to be normal in wellfunctioning markets, volatility becomes problematic when price fluctuations are large and unpredictable. The ESRS deals with higher price volatility levels than other livestock sectors because of three main factors: 1) meat and dairy products from small ruminants have variable and seasonal demand (e.g. peaks of consumption in Easter, Sacrifice Festivals and Christmas). Alternatively, production cycles are long and cannot be adapted easily causing short-term volatility.
2) The small ruminants sector is sensitive to natural factors such as adverse weather conditions and plant and animal diseases, and seasonal changes in the animal's physiology make output uncertain across the year (Montossi et al., 2013). 3) Following World Trade Organization agreement, EU agricultural sectors are more open to global markets and hence more vulnerable to international macro-economic changes such as food, feeds and oil prices or currency exchange rates (Westhoek et al., 2011). One recent example of macro-volatility was the COVID-19 pandemic that caused major market disruptions in the ESRS because the closure of the hospitality sector, which is considered the main selling point for small ruminants products. Moreover, the lower rates received by the farmers did not lead to reduced shelf prices and ultimately higher demand as previously reported (Ben-Kaabia and Gil, 2007).
Because of these economics constraints, the production costs in many European farms are higher than the total revenue (up to 52% in dairy sheep), making them highly dependent on EU Common Agricultural Policy (CAP) subsidies (Pulina et al., 2018). Our study revealed that market volatility was perceived as more relevant for the sheep than for the goat sector, possibly because it has higher seasonality in the offer-demand market. In this sense, there are mixed views about how to deal with market volatility in the European livestock sector (Moraine et al., 2014). Since its establishment in 1962, the EU CAP sought to stabilize markets and ensure a fair standard of living through market interventions measures. Over time, however, CAP reforms (1992, 2003, 2009, 2013 and 2020) have gradually reduced market interventions towards a more open market to increase farm incomes. The Priority Index for each challenge was calculated as the product of its relevance (1-5 scale) and its ease to be addressed (1-5 scale). Within this column, means with different superscript differ at P < 0.05. 2 Abreviations: Int., intensive; Semi., Semi-intensive; Ext., Extensive; *** P < 0.001; ** P < 0.01; * P < 0.05.

Feasibility to address challenges
Several authors (De Rancourt et al., 2006;Castel et al., 2010;Castel et al., 2011;Montossi et al., 2013;Pulina et al., 2018) proposed different strategies to tackle the economic challenges of the sheep and goat sectors such as: 1) increasing farm size and intensification, 2) decreasing seasonality of production, 3) implementing new technologies in reproduction, management and breeding strategies, 4) promoting feed selfsufficiency, 5) promoting cooperatives and relationships between farmers and food industries and 6) improving product diversification and differentiation. Most of these strategies focus on the economic and technical aspects of farming, and our findings agreed with these studies on the high relevance of economic challenges.
Our study revealed a substantial variability in the perceived feasibility to address the identified challenges. Therefore, our priority assessment, which also included the feasibility of addressing the challenges, found that many of the most relevant challenges were external. This makes it difficult to address since they are intimately linked to the low flexibility of the traditional production systems (De Rancourt et al., 2006) and market idiosyncrasy (Ben-Kaabia and Gil, 2007). Our study also revealed that these issues are perceived as more challenging to address in Southern than Central Europe. This could be due to higher feeding costs derived from a lower availability of perennial pastures, insufficient sector structure (e.g. not so developed cooperative organizational systems and farmers associations) and market idiosyncrasy (e.g. asymmetric price transmission thought the food chain) in Southern Europe (De Rancourt et al., 2006;Ben-Kaabia and Gil, 2007;Mena et al., 2017;Ruiz-Morales et al., 2019).
Several internal challenges were perceived as non-difficult to address ('parasite and infectious diseases', 'low adaptability of high productive breeds', 'low promotion of local breeds', 'poor business management training', 'slow adoption of innovations' and 'lack of professionalization'). Solving these challenges was likely perceived as more achievable as they rely on the successful implementation of technology and innovations, health treatments, disease surveillance, appropriate breeding programmes and farmer training initiatives. It has been described that the successful implementation of these strategies can improve farm productivity and sustainability, offering career opportunities for young people (Pulina et al., 2018). Furthermore, several external challenges related to education and sciences were also considered achievable such as 'researchers do not address real problems', 'low consumer education in local products', 'environmental policy against intensification' and 'low social knowledge about farming'. These observations justify the need to consider, not only the relevance but also the feasibility to address the challenges during the prioritization process. Considering both the relevance and feasibility to address the challenges, this study identified a number of prioritized challenges that could be grouped in four areas as follows.

Enhancing consumer and social awareness
Our analysis showed that several aspects related to consumer and society knowledge on sheep and goat farming (i.e. 'low consumer education in local products', 'low social knowledge about farming' and 'poor recognition of public services of farming') are of great importance for the ESRS. These results point out the existing miscommunication between rural and urban societies and the different perceptions on the production systems and food attributes (Mceachern and Seaman, 2005). Our study also revealed that this lack of knowledge is more relevant in Southern than in Central Europe and for dairy than meat systems, possibly because of higher product diversification (e.g. cheese, yogurt, meat) in comparison to meat production systems (De Rancourt et al., 2006).
Previous SWOT analyses have also identified similar external challenges for the ESRS because of declining consumption of animal products, and ruminant products in particular, even if they may have been produced and transformed in a sustainable manner (Pulina et al., 2018;Ruiz-Morales et al., 2019). This might indicate that societal knowledge on the ways that ruminant foods are produced is poor (Montossi et al., 2013). Consumers prefer local or national products because they are considered fresher, tastier and of better quality (Chambers et al., 2007). However, consumers also prefer cheaper options which are often imported products and affect the market of local sheep and goat products (Chambers et al., 2007;Montossi et al., 2013). Product diversification and commercialization of easy-to-cook foods (e.g. pre-cooked recipes, new meat cuts and dairy products) along with marketing campaigns can, to some extent, stimulate consumption of small ruminants products (Mandolesi et al., 2020). However, providing more information to consumers about animal rearing conditions, feeding and animal welfare, previous to lamb consumption, positively influenced consumer acceptability, contributing to improved satisfaction of their expectations (Napolitano et al., 2007;Font i Furnols et al., 2011). As a result, product information about high standards of animal welfare or more 'natural' production systems is starting to be used as a marketing strategy (Blokhuis et al., 2013). In our study, the assessment of stakeholders that should take action in addressing the reduced consumer demand and social awareness in local products included governments, association of producers and consumers themselves.
In recent years, social awareness about how food is produced has increased livestock farming (Blokhuis et al., 2013). After a systematic review of 80 publications, Clark et al. (2016) reported that the public is concerned about animal welfare in modern production systems with variations in relation to age, gender, education and familiarity with farming. As a result, consumers adopted self-justifications to avoid animal products . This has led the ESRS to work on a cross-sectoral communication with the public in recent years in order to correct existing misconceptions about animal husbandry (Blokhuis et al., 2013). This information deficit-approach assumes that better knowledge within the public about farming will increase the acceptance of husbandry systems. Although such strategies may increase transparency with regard to animal farming, they have largely failed to increase acceptance for existing systems (Montossi et al., 2013). In some cases, such strategies provoke the opposite reactions and people can get even more critical towards livestock farming due to the new information they received (McNeill, 2014). Reasons may lie in different reference frames and values between experts and laypeople about farming.
The challenge 'poor recognition of public services of farming' was perceived as a more relevant and priority in extensive than in intensive production systems. Several studies indicate that sheep and goat semiextensive and extensive farming systems provide multiple ecosystem services that improve environmental diversity, landscape conservation, fire prevention, carbon sequestration and employment in rural areas (Bernués et al., 2011;Sanderson et al., 2013;Rodríguez-Ortega et al., 2014;Herrero et al., 2015). These public services justify further EU public support for a 'Greening CAP' paying for ecosystem services (Rodríguez-Ortega et al., 2018). However, our study showed that the role of ESRS for environmental and social services is not perceived equally across the society. Therefore, more collaboration between government, farmers and food industry would guarantee effective knowledge transfer through the food chain.

Farmer's capacity building
Insufficient sheep and goat farmer professionalization and farm modernization have often been identified as key weaknesses of the ESRS. Previous studies based on SWOT analyses (De Rancourt et al., 2006;Pulina et al., 2018), technical-economic indicators (Castel et al., 2011;Mena et al., 2017) and country-specific observations (Castel et al., 2010;Sossidou et al., 2013) highlighted that farmers need more capacity building beyond the farm gate through training, developing public advisory services, promoting innovation and modernization and better product commercialization. Our study reflected this situation and highlighted insufficient business management training, slow adoption of innovations on-farm and lack of professionalization. Our study also revealed that these challenges are a higher priority in intensive production systems from Southern Europe (mostly dairy), compared to the more extensive grazing system from Central Europe which are mostly devoted to sheep meat production.
This observation highlights the need to enhance specific skills such as artificial milk feeding, milking, insemination and product diversification in intensive dairy systems (Thornton, 2010) or pasture management and monitoring of technical-economic indicators in extensive systems (Castel et al., 2011). Our study revealed that improved farmer's skills is perceived as easy to solve and should be given high priority. This aspect can even be considered as an opportunity if farmers, associations of producers, governments and academia work together to develop comprehensive training programs tailored at a regional level to address the needs of local farmers and new entrants (Ruiz-Morales et al., 2019). Programs that build farmer capacity developed by regional governments and technical courses promoted by farmers associations and cooperatives could aid to address this problem by encouraging the innovation uptake by the farmers. For example, recently and as part of the iSAGE project, we have shown that implementation of smart-farming technology based on individual animal data minimizes unproductive periods and assists farmers in their decision-making, increasing productivity in dairy goats farms (Belanche et al., 2019). These innovations promote a step change when farmers have sufficient knowledge beyond the farm gate for business management training and risks assessment based on the use of technical-economic indicators (Gaspar et al., 2008).

Promoting local breeds
Local breeds are a major component of animal farm biodiversity, despite the low census of some autochthonous breeds, which represents a weakness (Castel et al., 2010;Montossi et al., 2013;Ruiz-Morales et al., 2019). Adding value to local breeds contributes to their conservation. 'Low promotion of local breeds' and 'low adaptability of high producing breeds' were challenges perceived as moderately relevant, but were considered as two of the easiest to address through appropriate breeding programs (Martin-Collado et al., 2013) and therefore are considered a strategic priority.
Local dairy goat breeds are traditionally reared in semi-extensive systems using marginal territories (Pulina et al., 2018) and have a highly variable production (150 to 500 kg/lactation). Despite their improvement through recent breeding and management programs (Belanche et al., 2019), they are still far from foreign breeds (800-950 kg/lactation) such as Saanen or Alpine (Pulina et al., 2018). The dairy sheep industry is based on local breeds and crossbreeds raised under more intensive systems than dairy goats, but with increasing importance of foreign breeds such as Assaf and East Friesian (Pulina et al., 2018). As a result, the census of some local dairy breeds in Europe is a growing cause for concern.
However, this change in the use of breeds is not always accompanied with improvements in feeding, sanitary programs, milking automation, reproduction intensification, management or genetic selection, aspects that can severely limit their productivity (Miller and Lu, 2019).
The European meat sheep sector is dominated by local rustic breeds (e.g. Merino, Rasa Aragonesa, Welsh Mountain or Scottish Blackface) which often produce branded local products (e.g. Ternasco de Aragon, Welsh lamb, Scottish lamb). However, the European sheep meat production has decreased by 34% in the last decade causing imports of lamb meat mostly from New Zealand and Australia (FAOSTAT, 2020). This, along with the substantial decrease in lamb consumption, has led to a crisis in the EU meat sheep sector (Montossi et al., 2013). Our study identified that both challenges ('low promotion of local breeds' and 'low adaptation of high producing breeds') were perceived as more relevant for the dairy than for the meat sector and in Southern than in Central Europe. These observations could be justified because dairy animals are more sensitive than meat animals to the environmental factors such as feed scarcity and heat stress typical from Southern European countries (Ramón et al., 2016).
To solve this situation, the European Commission allocates resources for the implementation of regional programs to promote local breeds. These resources are normally for breeding associations and farmers to subsidize some breeding programs costs such as milk controls or artificial inseminations to improve productivity (Ruiz-Morales et al., 2019). This investment has allowed substantial increase in productivity in some local breeds such as the Lacaune sheep in France or the Murciano-Granadina goats in Spain (Belanche et al., 2019). However, our study revealed that this financial support seems insufficient and requires a stronger collaboration between governments, farmers, producers and academia. A recent report from the European Parliament (A8-0064/ 2018) called on the Commission to take measures to step up support for the keeping and promoting further sheep and goat local breeds.

Re-focusing research priorities
Despite the multi-stakeholder process used in this study aimed to maximize the adherence between the identified challenges and the real situation of the ESRS, we are aware that the slight overrepresentation of research institutions in this process could magnify some of their concerns, such as 'researchers do not address real problems'. Our study identified this challenge as a priority across species, production systems, type of products and regions, mostly because it was perceived as the easiest to solve. Although in the last 10 years a total of 117 EU research projects have been funded to address different aspects related to small ruminants, along with 28 102 research publications worldwide, our study revealed that this sector has the perception that researchers do not satisfactorily deal with relevant issues at farm level. Most EU projects supporting the ESRS focused on two specific research areas ('health' and 'climate change and environment') which are aspects of great social concerns, but that as our multi-stakeholder survey revealed are not the priority for the ESRS. Similar overrepresentation of these two research areas in comparison with the area 'technology OR innovation' was also identified in the research publications related to small ruminants. On the contrary, cattle-related funded projects had a more even distribution across research areas and provided more support to areas such as 'food and natural resources' and 'industrial technologies'. These aspects are more focused on production elements that likely lead to improving farm profitability through innovations (Thornton, 2010). These findings show that in order to improve the ESRS perception on the usefulness of public research, which will likely facilitate their confidence on governmental agencies and their participation in research activities, the views from ESRS (including farmers, associations of producers, breeders and cooperatives) should be more effectively communicated to governments and funding agencies, so they can be incorporated into future research funding programs and agricultural policies (Blokhuis et al., 2013).

Conclusion
The proposed multi-stakeholder participatory analysis identified the ESRS main challenges. The challenges perceived as most relevant were related to economic aspects but were mostly beyond the farmer's control (external) and difficult to address. The implementation of a priority index facilitated the prioritization of challenges by considering the relevance and the feasibility to be addressed. The results of our study suggest that knowledge development at farm level (i.e. farmer's capacity building), consumers (i.e. increase consumer knowledge on food products), society (i.e. increasing social recognition of the ESRS role on delivering public services) and academia (i.e. increase applied research focused on key sector challenges) are the priorities. Promoting local breeds was also identified as a priority to preserve diversity and cultural heritage. Effective collaboration and knowledge transfer across the different stakeholders involved in the ESRS are needed to address these challenges. Moreover, the different perception across regions in Europe suggests that the upcoming policies should be more contextualized for different regions than they have been before.

Supplementary materials
Supplementary data to this article can be found online at https://doi. org/10.1016/j.animal.2020.100131.

Ethics approval
Surveys were anonymized.

Data and model availability statement
Data are available on request to the corresponding author.