Social assessment of miscanthus cultivation in Croatia: Assessing farmers' preferences and willingness to cultivate the crop

Social aspects of miscanthus cultivation have been investigated in a limited way in the scientific literature. Adopting existing frameworks for social life‐cycle assessment enables assessments to include numerous social aspects; however, the relevance of these aspects depends on the local context. This study aims to identify the most relevant social aspects from the farmers' perspective using a previously proposed framework for the assessment of the stakeholder ‘farmer’. It is based on a case study for miscanthus production in Sisak Moslavina in Croatia. The existence of abandoned lands in Croatia presents an opportunity for the cultivation of miscanthus as a potential source of biomass for the production of bio‐based materials and fuels. The study seeks to assess the feasibility of cultivating miscanthus in the region, taking into account potential challenges and opportunities, as well as farmers' willingness to adopt the crop, and to understand the reasons behind land abandonment. We conducted a survey among 44 farmers in the region and used a scoring method to identify the most relevant social aspects. The aspects most valued by the farmers were health and safety, access to water, land consolidation and rights, income and local employment, and food security. Responses to the question of whether they would adopt the crop highlight the importance of an established market, good trading conditions and profitability of cultivation. The survey also enabled an understanding of farmers' preferences with respect to the production conditions of crops. The farmers regarded the provision of subsidies as one of the main factors that render a crop attractive. Opportunities for the adoption of the miscanthus cultivation include high yields and low input requirements. Barriers include land conflicts and land availability. Despite the opportunities for miscanthus development in the region, there are important challenges to consider for successful implementation of the crop.


| INTRODUCTION
According to the sustainability paradigm, the social dimension is one of the three pillars of sustainability, and is as relevant as its economic and environmental counterparts. Experience has clearly shown that neglecting social aspects from studies can lead to unexpected outcomes, such as was the case with the fuel versus food debate back in 2008, when political support encouraged the development of bioenergy. Public opposition arose as the promises of greenhouse gas (GHG) reduction were not met as expected. Furthermore, political support through incentives for liquid biofuels influenced the global demand for edible biomass, thus contributing to price fluctuations for example of maize. Another example of opposition to the development of emerging technologies due to their impact on local communities can be seen in the experience of biogas acceptance in several countries (Mazzanti et al., 2021).
There are various approaches for investigating the social dimension of production processes. A framework used for evaluating how production affects stakeholders involved along value chains is the social life-cycle assessment (S-LCA). In accordance with life cycle thinking, S-LCA assesses social impacts from cradle to grave. The main objective of such an analysis is to support decision making to improve the social conditions along the value chain of a product (Benoît & Mazijn, 2009). An S-LCA at process level is possible, for example, by considering the inputs required for each process unit in mass and the worker hours associated with each input. Marting Vidaurre et al. (2022) presented a generic analysis of the potential use of miscanthus for bioethanol production with the foreground system, which takes place in Croatia, and considering 49 countries and regions for international processes and 42 social indicators. In that assessment, it was possible to identify hotspots of potential social risks and human rights violations along the value chain. However, the consideration of country-level indicators only allowed a risk assessment to be performed and gave no information on the benefits or issues that may arise from miscanthus cultivation at a local level. To investigate such potential benefits or other issues, this article focuses on miscanthus cultivation at local level in the region of Sisak-Moslavina in Croatia and involves a survey of farmers in the region. The framework, proposed by the UNEP Guidelines for S-LCA and an adapted framework to assess the stakeholder 'farmers' proposed by Marting Vidaurre et al. (2020), served as a basis for this study.
Miscanthus is a perennial C 4 grass originating from Asia, which is adapted to a wide range of climates and soils, including those present in Europe . After successful establishment, miscanthus does not require any soil cultivation for a period of 20 years. In addition, after herbicide application during the establishment phase, no further plant protection is needed. Due to its high water and nutrient efficiency, it has low nitrogen fertiliser requirements. On account of its perenniality, the soil is not disturbed after planting, thus contributing to the accumulation of soil organic matter . In contrast to annual arable crops, miscanthus provides a retreat and nesting opportunity for birds and mammals . These factors significantly improve the environmental and economic performance of miscanthus in relation to annual crops. In addition, miscanthus has been investigated as a potential source of lignocellulosic material cultivated on low-quality soils (Xue et al., 2016), thus reducing the risk of competition with food or feed crops and also reducing the risk of indirect land use change (iLUC).
Technologies already exist for the extraction of cellulose and hemicellulose from lignocellulosic biomass and their conversion into bio-products such as bioethanol (Lask et al., 2021) and are commercially available in several regions of the world, such as Europe (e.g. Clariant) and the U.S. (Jin et al., 2019;Moriarty et al., 2016).
While a number of studies have tackled various social aspects related to the cultivation of miscanthus (Anand et al., 2019;Gallardo & Bond, 2011;La Rúa & Lechón, 2016;Panoutsou & Chiaramonti, 2020;Perrin et al., 2017;Raman et al., 2015), only a few have been dedicated exclusively to the assessment of the social dimension of this crop (Augustenborg et al., 2012;Bocquého & Jacquet, 2010;Panoutsou & Chiaramonti, 2020). Of these, most focus on socio-economic aspects, such as job creation and the adoption of miscanthus by farmers.
To achieve the targets proposed by the Renewable Energy Directive, Croatia aims to increase the share of renewable energy in the transport sector to 14% by 2030 (Kulišić et al., 2021). Biomass for advanced biofuels could contribute to achieving such targets. Croatia possesses a total area of 8.8 million ha (one third is water) (Lisjak et al., 2021;Roić, 2016), of which 2.7 million ha are estimated to be agricultural land (Lisjak et al., 2021). The existence and use of abandoned and unused agricultural land here is seen as an opportunity for job creation in areas of declining population and increasing unemployment, for example in the region of Sisak-Moslavina (Bilandžija et al., 2016). Due to its low input requirements and high yield potential in Croatia in the range of 14.5-25.8 t ha −1 (Bilandžija et al., 2018Voća et al., 2021), miscanthus offers an opportunity for the productive use of this land (Voća et al., 2019). Currently, there is a 30-ha miscanthus plantation in Popovača, Sisak-Moslavina, established to produce and sell rhizomes (personal communication: Faculty of Agriculture, University of Zagreb); however, a market for miscanthus biomass has not yet been established in Croatia.
The aim of this study is to identify context-relevant social/socio-economic aspects from the farmers' perspective in the county of Sisak-Moslavina in Croatia, and to understand both farmers' interest in adopting miscanthus and the potential opportunities and/or barriers for its cultivation in the region. The specific objectives of this study are (1) to understand the willingness or otherwise of farmers to adopt miscanthus cultivation, including the reasons for their (lack of) interest in the crop, (2) to understand farmers' preferences regarding the production conditions of crops, (3) to identify relevant social aspects for the local farmers and (4) to explore the reasons for the land abandonment in Croatia.
The contribution of this study to the existing body of literature lies in the identification of relevant aspects or categories from the social perspective involving the stakeholder 'farmer'. This set of social categories can be used as a basis for the impact assessment of the agricultural stage of bio-based value chains in the region.

| State-of-the-art
Farmers' willingness to adopt the crop, socio-economic aspects related to miscanthus and health and safety in agriculture have been addressed in the literature as described below.
1.1.1 | Willingness to cultivate miscanthus Augustenborg et al. (2012) conducted a survey of 172 Irish farmers to understand their willingness and opinions regarding energy crop production, including miscanthus. They concluded that the lack of interest in energy crops may be related to a lack of knowledge regarding the crops' benefits. They also attempted to identify the profile of a 'potential adopter' of energy crops. Potential adopters expressed a lack of trust in their current business strategies as a reason for introducing miscanthus cultivation. Furthermore, respondents belonging to a bioenergy organization had a significantly higher mean in their level of interest in miscanthus compared to non-members. The study discussed the need for either long-term contracts or government schemes to encourage farmers to adopt the crop. The main obstacle to miscanthus adoption identified was the lack of a guaranteed market. Other relevant obstacles were the lack of information regarding profit margins, lack of long-term guarantees and high upfront costs. Potential adopters were interested in both the economic and environmental benefits of the crop. 1.1.2 | Socio-economic aspects Bocquého and Jacquet (2010) performed an economic assessment using a net present value (NPV) analysis of miscanthus and switchgrass in comparison to a rotation of conventional annual crops in France. They emphasized the importance of including farmers' risk perception in the analysis when assessing perennial crops. Miscanthus requires high upfront costs. During the establishment period, farmers invest in planting material with no returns in the first 2 years. The irregularity of returns can lead to liquidity constraints during this phase. This represents a risk for farmers and affects their decision to adopt miscanthus. The assessment also considered two scenarios for supply contracts, since miscanthus requires long-term (20 years) commitment of the land. The first scenario considers a long-term price guarantee for the biomass and the second is related to an initial loan to cover establishment costs and liquidity requirements. The results of the study show that miscanthus and switchgrass are less profitable than a conventional annual crop rotation under the assessed scenarios and they highlight the relevance of considering the financing and liquidity constraints during the initial phase of the cultivation period and the need for long-term contracts for successful adoption of the crop by farmers. Panoutsou and Chiaramonti (2020) assessed the employment and profit prospects from cultivating miscanthus for energy production on marginal lands in Italy and Greece using an econometric model. They considered two value chains: combined heat and power (CHP) and fast pyrolysis bio oil (FPBO). They assumed an annual yield of 10 (low-yield scenario) to 15 DM tonnes ha −1 (high-yield scenario). The study concluded that small-scale value chains are more labour-intensive, leading to the creation of more jobs than in larger projects. Furthermore, profit is highly influenced by yield. Since a long-term commitment is required for miscanthus cultivation, the authors highlight the importance of trust building and collaboration with the local farming community for successful adoption of the crop.

| Health and safety
Although there is very little information on health and safety aspects specifically for the cultivation of miscanthus, the literature provides a general overview of on-farm health and safety. Health risks in agriculture have been related to accidents involving heavy machinery (Fargnoli et al., 2018), exposure to vibration, noise (Calvo et al., 2019) and dust (Rumchev et al., 2019), transmission of diseases from animals to humans (Mori & Roest, 2018), musculoskeletal disorders (Kumaraveloo & Lunner Kolstrup, 2018), and exposure to pesticides (Kim et al., 2017) and fertilizers (Ahmed et al., 2017;Hill et al., 2019), among others.
As with any other crop, miscanthus cultivation can affect the health and safety of the farmers in a number of ways. A low technological level of the farm is associated with more manual tasks that could lead to musculoskeletal disorders, for example, tillage using a hand hoe or bending over to perform manual planting (European Agency for Safety and Health at Work, 2012). According to Kok et al. (2019), the agriculture, forestry and fishing sector has been identified as the main sector reporting musculoskeletal disorders in the EU in 2015 (e.g. 60% of workers reporting back problems came from this sector). Musculoskeletal disorders can occur due to lifting or carrying heavy loads, awkward positions, stretching and bending during work, and so forth. In miscanthus cultivation, this could be related to planting operations, activities including the carrying of heavy loads on farms that use manual work instead of machinery for transport, driving long distances, handling machinery and so forth.
Mechanized operations, on the other hand, are more related to accidents involving tractors and agricultural machinery, for example entanglement in moving parts, contact with a power take-off (PTO) shaft, tractor rollovers and runovers (European Agency for Safety and Health at Work, 2012), and exposure to vibration, noise and dust. Accidents may occur more often during the harvest season due to tiredness of farmers working longer hours than usual. Miscanthus is harvested at two different times in the year depending on the use of the biomass. A green cut between mid-October and early November supplies biomass suitable for biogas production (Kiesel & Lewandowski, 2017;Mangold et al., 2019). A brown cut after winter in March favours combustion and animal bedding. Thus, if a farmer cultivates miscanthus for combustion purposes and maize at the same time, harvesting would take place at different times in the year: March for miscanthus and July-October for maize. In this way, miscanthus cultivation would not require any additional working hours for farmers during the maize harvesting season. However, if the miscanthus is harvested green, the harvest regime would coincide with that of silage maize (Winkler et al., 2020).
Compared to annual crops such as maize, miscanthus cultivation requires a high number of working hours for the initial planting of rhizomes, but does not need recultivation every year. This may lead to an overall lower risk of injuries associated with agricultural operations during crop establishment than for annual crops.
Exposure to chemical substances in agriculture also constitutes a risk to human health. Problems reported in the literature include exposure to pesticides (Kim et al., 2017) and fertilizers (Ahmed et al., 2017;Hill et al., 2019). In the case of miscanthus cultivation, pesticides are often not used at all during the entire productive period and herbicides are only required during the establishment phase (first 2-4 years of a 20-year plantation) . Fertilizers are also only applied up to once a year for the entire 20-year cultivation period (Winkler et al., 2020). Miscanthus cultivation has a significantly lower requirement for crop protection chemicals and fertilizers than other crops such as maize, and therefore lower health risks associated with the use of these substances.

| MATERIALS AND METHODS
The case study methodology was chosen to gain an appreciation of the local context since we consider it relevant to understand what determines farmers' willingness to cultivate the crop, together with their preferences and what they consider relevant for their households in terms of social aspects. The type of case study applied is exploratory one.

| Case study
Sisak-Moslavina County is located in the southern central part of Croatia and is part of the continental area of the country (see Figure 1), approximately 50 km away from the capital city Zagreb. Sisak-Moslavina is home to 147,216 inhabitants (Croatian Bureau of Statistics, 2021), representing 3.6% of Croatia's total population, and is sparsely populated (32.9 inhabitants/km 2 ). In 2018, approximately 23% of the population was over 65 years old (Croatian Bureau of Statistics, 2018), indicating a substantial presence of elderly people in the region.
As is the case for Croatia as a whole, the labour market in Sisak-Moslavina has recovered considerably in the last years. Whereas the unemployment rate for this county was 34% in 2013, it had fallen to 18% in 2019, which is however still much higher than at country level.
Sisak-Moslavina County has a predominantly agricultural landscape, though some regions such as the city of Sisak have been undergoing a process of de-ruralisation (Sisak Projekti, 2015). Sisak-Moslavina is one of the many regions in Croatia where important areas of agricultural land have remained unused for years (Voća et al., 2019). Table 1 shows the 17 social aspects with corresponding indicators considered in this study. The indicators served to improve the respondents' understanding of each social aspect for the scoring (see Section 2.3.3, Survey guideline-Scoring of aspects).

| Survey
Survey was the selected method to collect data in this study.

| Selection of participants
Interviews were conducted in Sisak-Moslavina County from May to October 2021. The farmer registry contains about 6000 entries, from these a number of farmers were randomly selected and contacted. A total of 44 farmers participated in the questionnaire. Figure 2 shows the distribution of farmers registered in Sisak-Moslavina by total area in hectares. Approximately 75% of the farms possess less than 7.8 ha. The median is 4.19 and the mean value (μ) is 10 ha.
The sample mean X is 49.8 ha, much higher than the population mean, suggesting a bias towards farmers in possession of larger farm areas than the average smallscale farmers.

| Survey design and process
The survey was divided into three sections. The first section corresponds to the profiling of respondents. The second section serves to understand the farmers' attitudes towards miscanthus cultivation. The third section is related to the identification of the main social categories/ aspects for the farmers. The selection of the categories was based on the S-LCA framework proposed by the UNEP/ SETAC in the Methodological Sheets for Subcategories in S-LCA (Benoît Norris et al., 2013) and a framework proposed in a previous study for the assessment of the stakeholder 'farmer' (Marting Vidaurre et al., 2020). The survey can be found in Supplementary Information 1. The survey draft was discussed and modified with seven researchers from Germany and Croatia before distribution to improve understanding and avoid bias. Surveys were performed personally on site and via telephone. The data were collected in Croatian and translated into English for analysis.

| Survey guideline
The survey consisted of three parts.

Profiling of respondents
Farmers were asked to provide information on gender, education level, age, land size and type, crops cultivated, membership and interest in cooperatives, and years of ownership of the farm and land.

Farmers' attitudes towards miscanthus cultivation
Farmers were first asked how familiar they were with the crop. Then, they were asked to rate conditions of a crop that may make it attractive for farmers to cultivate on a scale from "very important" to "not important at all". To assess the willingness of farmers to adopt miscanthus, the questionnaire also included the closed question "If there was a market for miscanthus biomass, would you be willing to plant this crop?" and an open question asking them to provide reasons for their responses.

Scoring of aspects
Farmers were asked to express the extent to which they agree with a set of 1 to 2 statements, referred previously as indicators, regarding a social aspect using a 5-point Likert scale. The statements served to specify the meaning of each aspect, as the title of the aspect alone, for example "Health and Safety", would not have been enough to communicate the meaning to the farmers. After this, the farmers were asked to assign a score to each category on a scale from 1 to 10, 1 being "very important for my household" and 10 "not important at all for my household".

| Data analysis
The software SPSS was used for descriptive statistics and to perform Chi-square tests of independence between the variables "familiarity with the crop miscanthus" and T A B L E 1 Social aspects and indicators considered in this study. • In general, the requirements of the local community are taken into consideration and adequately addressed by the regional authorities 13. Local employment (employment for the local community)

Social aspects Indicators
• Farms in the region tend to hire residents of the region (for example, local farm workers) • I acquire agricultural inputs from local suppliers (products made in Croatia) provided this is possible "willingness to cultivate miscanthus". The data were not sufficient for a representative sample; however, a logistic regression model was used to identify differences between farmers with and without an interest in miscanthus. To understand whether the willingness to cultivate the crop can be explained by predictor variables, we used a logistic regression. Logistic regression is part of the generalized linear models (GLS). It is commonly used for the analysis of binary data (Agresti, 2015). Details of the model can be found in Supplementary information 2, section: Model and model results: Binomial Logit. A backward elimination process for the predictor variables was performed. The variables included were: gender, education level, age, household size, farm size and number of parcels. The variables with the highest p-value were then removed from the model one by one until only variables with a α < 0.05 remained. For the regression, "yes" and "if…" answers were grouped together into a single statement that the respondent would at least consider the possibility of cultivating miscanthus if the expected conditions were given. This decision to group the two responses was made because the farmers' lack of information regarding miscanthus plays an important role in how they answered the question. This is reflected in the 52.2% of respondents who provided "if…" answers.

| RESULTS
The results section is structured according to the research questions.

| Demographic characteristics
Of the total respondents, about 91% are men and 9% women. This is related to the higher presence of men in the registries used to contact the farmers. With regard to education level, 95.5% have at least a high school education, 34% have an additional higher education, while 4.5% did not respond to this question. Most of the respondents (59%) are between 41 and 60 years old, 27% are between 20 and 40 years old, and only 14% are older than 60. The farmers are mainly medium-sized landholders (54.5%), 29.5% have small farms and 16% have large farms (including rented land). The minimum farm area in the sample was 3 ha and the maximum 241 ha. The farm size categories given in Supplementary Information 2, Table SI-2, with definitions of sizes ranging from 'very small' to 'large', were taken from (Eurostat, 2016). Of the 44 farmers interviewed, 95.5% possess arable land, the remaining 4.5% did not answer this question.
To understand how fragmented the land of the farmers under study is, we use the relation 'number of ha/number of parcels' (Hartvigsen, 2014;McPherson, 1982). The mean for this relation in the sample is 2.73 ha/parcel (median 1.61 ha/parcel) and corresponds to the fragmentation of land use and not to the fragmentation of the property, since the total farm area includes rented land.
In response to the question of how much land the farmers dedicate to their own consumption, 18% answered more than 20 ha. This is because in many cases they use the land to feed their livestock and sell the meat, and this was considered as their "own consumption." On the other hand, 68% designated more than 80% of their land to the cultivation of products for the market.
With regard to rented land, 34% of the farms rent in more than 50% of the land they cultivate, and 31% rent in between 11% and 50% of the land. By contrast, 41% do not rent out any land and only 7% rent out 10%-50% of their land (50% did not provide a response for this question). This could mean that farmers in the sample tend to rent in significant amounts of land while renting out is less common. The farmers have very little unused land; only one farmer reported to have 3 ha of unused land.
Supplementary Information 2, Table SI-2 gives full details of the farmer profiling.

| Familiarity with the crop miscanthus
As we anticipated that few people in the region know miscanthus, the first question in this section asked about the farmers' familiarity with the crop using a 5-point scale from "I have never heard of it" to "Very familiar, I have grown miscanthus." Only 2.3% (1 farmer) had grown the crop, 31.8% said they were familiar with it in the sense that they could recognize it, 18.2% said they had heard about it and had an idea what it is, 25% had heard about it but did not know what it is, and 20.5% had never heard of it. This means that 79.5% had at least heard of it. This high percentage of farmers aware of the crop was unexpected and could be related to programmes introducing miscanthus to farmers in the region in recent years (2015-2018). Figure 3 shows that the most relevant aspects for a farmer to be willing to cultivate a specific crop are: the provision of subsidies, low requirement for agricultural inputs, easiness of cultivation, no additional machinery required and high returns. The least relevant aspects are aesthetic of the landscape and that the household has traditionally grown the crop. About 60% of the respondents considered the periodicity of recultivation important.

F I G U R E 3
Farmers' responses to the question: What makes a crop attractive? Data presented in absolute values.

| Farmers' willingness to cultivate miscanthus including reasons
A further question in the survey was: "If there was a market for miscanthus biomass, would you be willing to plant this crop? Why?" Out of 44 respondents, 29.5% said they would not plant miscanthus and 63.6% said they would or provided "if answers" ("yes" 11.4%, "if…" 52.2%) and 6.9% said "I don't know" or did not answer. Table 2 shows the number of farmers who answered "yes," "no" or "if…" in relation to their familiarity with the crop.
To test whether these two variables are associated, a Chi-square test was performed. The variable "familiarity with miscanthus" was transformed into a binary variable, assuming that responses from "very familiar…" to "I have heard of it, I have an idea" means a farmer is familiar with miscanthus, and that "I have heard of it, but I don't know what it is" and "I have never heard of it" represent those who are not familiar with the crop. The variable "willingness of farmers to adopt the crop" can take the two values "no" or "yes/if." This results in a 2 × 2 contingency table. The Chi-square test shows an association between the variables with a p-value of 0.041. A V-Cramer test revealed that the association between the variables has a value of 0.327, which is interpreted as a moderate level of association.
A further statistical analysis was performed taking the "willingness of farmers to adopt the crop" as response variable and demographic variables (gender, education level, age, household size, farm size and number of parcels) as predictor variables, however the model did not show any explanatory variables with an α < 0.05.
The reasons provided by the farmers for whether they would plant the crop or not are shown in Table 3.

| Other aspects
Regarding the questions related to abandoned lands, almost all respondents were aware of their existence (91%), however the farmers provided different reasons for the non-use of such lands. 10% attributed it to poor soil health, 83% to land ownership conflicts, 66% to low labour capacity, and 13% to other reasons including politics, law, non-secured trade deals and lack of interest in such land. Section 3.7 explores the reasons behind the abandonment of land in Croatia.
Most of the respondents (95%) regarded land consolidation as positive for the region, due to the current high fragmentation. However, 75% agreed or slightly agreed that there has been progress in this aspect in the region, while 10% were neutral and 15% disagreed or slightly disagreed that there is a trend toward consolidating land in the region.
Of the respondents, none belonged to a cooperative and only 29% expressed an interest in being part of one. The reasons provided for their lack of interest in cooperatives included: "It does not work in Croatia," "due to politics," "bad politics," "I did not have the time." Cooperation may however become important, for example when individual farmers cannot afford new machinery required for growing and harvesting the crop. This lack of trust in cooperatives could act as a barrier to the successful development of miscanthus in the region. Successful implementation of energy crops requires the cooperation of local farmers (Augustenborg et al., 2012;Bocquého & Jacquet, 2010;Panoutsou & Chiaramonti, 2020). ranking of the aspects; the order given results from the scoring. Another aspect included by the respondents was lack of land.

| Scoring of relevant social aspects
As described in Section 2, statements were used to put the social aspect in context. Farmers' responses in this regard are provided in Supplementary Information 2,

| Literature review on the reasons for land abandonment in Croatia
Abandonment of rural and urban properties has been taking place in several regions of Croatia (Lončar & Pavić, 2020) for various reasons. One important reason is related to demographic changes, such as emigration of the population, fall in fertility rate and depopulation induced by war activities (Lončar & Pavić, 2020). It has been estimated that the net migration rate in the twentieth century was −1.2 million people (Lončar & Pavić, 2020;Nejašmić, 2014). A number of emigration events occurred in that period, including migration overseas in the first half of the century, internal movement of people during the Yugoslavia period, and war-induced displacement in the 1990s during the Croatian war of independence (Homeland War) (Lončar & Pavić, 2020). A more recent migration event took place after the accession of Croatia to the European Union in 2013 (Lončar & Pavić, 2020). In rural areas, emigration was also caused by urbanization, as people moved to bigger cities (Lončar & Pavić, 2020).
Besides emigration, a number of other reasons contributed to the abandonment of properties in rural areas and to the abandonment of agricultural land and socalled uncultivated lands. For instance, issues with the registration of state-owned agricultural land has posed a challenge for efficient disposal of land in the country. It is estimated that about 36% of the agricultural land was owned by the state in Croatia in the last decade (Obradović & Bačić, 2016). During the socialist era, agricultural land was managed by large-scale collective state-owned farms (Hartvigsen, 2014). In 1989, Croatia, as well as other central and eastern European countries, started a transition from a planned to a market economy (Hartvigsen, 2014). In this period, socially owned land became property of the state (Lisjak et al., 2021;Mikuš, 2021) and the land could be disposed in other forms of land tenure (Lisjak et al., 2021;Obradović & Bačić, 2016). To perform such disposal of land, new legislation including the Law of Agricultural Land (1991) were passed (Obradović & Bačić, 2016). This law was amended several times (Lisjak et al., 2021) in the last three decades and has been poorly implemented (Obradović & Bačić, 2016). One reason for this is the lack of registries for state-owned land during the socialist era, when changes in land ownership were often not recorded. This led to differences between the existing registers (containing information of the previous owners to that of the socialist era) and the current situation in the fields. There are also marked differences between the land registry and the cadastre (Croatia conducts a dual property registration system). All of this has led to an inefficient disposal of land and is often a source of conflict. In the 10 years up to 2011, efforts were made by the government to register the ownership of the socalled unregulated state-owned agricultural land and to harmonize the information contained in the land registry, the cadastre and the situation on the fields (Obradović & Bačić, 2016). Despite progress in these efforts, in the year 2021 more than half of the state-owned agricultural land remained uncultivated (Lisjak et al., 2021). An inefficient and non-transparent state-owned disposal of land is a relevant obstacle to adequate agrarian reform and agricultural markets (Mikuš, 2021). In addition, there are also areas in private ownership that remain uncultivated (Lisjak et al., 2021). According to Bilandžija et al., 2016, 61,000 ha of agricultural land in Sisak-Moslavina county have been abandoned. Other authors state that the county has 33,733 ha uncultivated public and 57,412 ha uncultivated private land (Pfeifer et al., 2016). The difference in such estimations suggests a lack of a reliable source of information on the amount of abandoned land in Croatia.
Another aspect contributing to land abandonment is the high fragmentation of agricultural land (Baričević et al., 2016;Hartvigsen, 2014). On average, family farms in Croatia have size of 2.7 ha split into 5.3 separated parcels (Baričević et al., 2016). According to Tomić et al., 2018, parcels in the cadastre show high fragmentation of land in the various regions of Croatia (ranging from 0 to 152.7 parcels per ha). According to MacPherson 1982, the negative consequences of excessive land fragmentation form an obstacle to the modernization of agriculture, and lead to both higher resource requirements (labour, fuel, etc.) in comparison to larger fields and high costs of alleviation (Hartvigsen, 2014). Eventually, high fragmentation can lead to the abandonment of agricultural land.
Another reason for land abandonment is the presence of landmines in some regions of Croatia (Musa et al., 2017). According to the Croatian government, in January 2021 about 249 km 2 were still suspected to be contaminated with landmines, 42 km 2 of which are located in Sisak-Moslavina county (Ministry of the Interior and Ministry of Defense of the Republic of Croatia, 2021).

| DISCUSSION
The aim of this study was to gain an understanding of both the willingness of farmers in the region of Sisak-Moslavina in Croatia to adopt miscanthus and the potential opportunities and barriers to the development of the crop in the region. Based on a framework developed through a literature review in a previous study (Marting Vidaurre et al., 2020), this work aims to identify context-relevant social aspects from the farmers' perspective in order to provide a guide for a planned and careful implementation of miscanthus, taking into consideration aspects that are relevant for the local community.

| Willingness of farmers to adopt miscanthus cultivation
A key finding is that the majority of the farmers (more than 50%) stated that they would plant the crop under specific conditions. In relation to the "no" responses (30%), this suggests a general openness of farmers to the crop, given certain conditions. A chi-square test revealed an association between the variables "willingness of farmers to adopt the crop" and "familiarity with miscanthus." This indicates that unfamiliarity with the crop influences farmers' response. The reasons provided by farmers who said they would not plant the crop included satisfaction with their current business, their advanced age, no land availability, and no market for the crop. Farmers who said they would cultivate it, if certain conditions were given, stated mainly profitability and good market conditions as the required context. Other aspects mentioned were land leasing, no need for additional machinery and the need for collaboration within the local community. These results imply that if there was a market for the crop and it was profitable, farmers would potentially be interested in its cultivation. However, there may still be some barriers to its adoption and development, for instance the advanced age of farmers. The ageing of farmers and lack of young farmers is a challenge for agriculture both in Croatia and the European Union in general. In 2016, the percentage of farm managers younger than 40 years old was 11% in the EU and slightly less than 11% in Croatia (Eurostat, 2018). For this reason, supporting the generational renewal is one of the 10 objectives of the Common Agricultural Policy (CAP, 2023). Young farmers are more likely to adopt miscanthus than older farmers because its cultivation involves a long-term investment (with higher initial costs during the first few years, and yielding profits from the second or third year until the end of the 20-year period).

| Farmer's preferences concerning the production conditions of crops
In our study, the main crop production conditions that farmers rated as attractive were: the provision of subsidies, low agricultural input requirements, no need for additional machinery, ease of cultivation, and high returns. A lack of subsidies for miscanthus cultivation could thus be understood as a barrier for its adoption in Croatia. Furthermore, if a farmer is in possession of low-grade arable land and wants to convert this land to miscanthus cultivation, he/she would stop receiving the subsidies granted for cultivating another crop on that land. The importance of subsidies in miscanthus cultivation has been previously mentioned in the literature. In a SWOT analysis of the resource efficiency of miscanthus, Panoutsou and Chiaramonti (2020) refer to the lack of subsidies for the crop as a weakness or threat. Bocquého and Jacquet (2010) emphasize the need for either a loan during the establishment period or a long-term price guarantee. Augustenborg et al. (2012) state that long-term contracts and/or government schemes are prerequisites for successful adoption of miscanthus. As a perennial crop, one specific feature of miscanthus is the high investment necessary for establishment in the first year combined with a lack of revenue from the biomass until the second to third year. Subsidies would reduce the risk of this initial investment and encourage farmers to cultivate the crop.
Favourable conditions for miscanthus adoption compared to annual arable crops can be seen in the lower requirement for agricultural inputs and the fewer working steps required in the long term.

| Identification of relevant social aspects for the local farmers
Part of this study was the identification of the social aspects most relevant for the farmers in the region. These were: health and safety, access to water, land rights, income, land consolidation, local employment and food security. We recommend using such aspects as a starting point for the social analysis of the implementation of agricultural projects in the region. For example, this could be useful for an analysis of the foreground system in a S-LCA of a biobased value chain in the region of Sisak-Moslavina. The scope of this study covered the identification of relevant aspects only. Therefore, a very brief assessment of each aspect is included here.

| Health and safety
Due to its nutrient recycling capacity, miscanthus is not vulnerable to pests or diseases and efficiently suppresses weeds after the establishment phase. In practice, the crop is cultivated without fertilization. The lower demand for fertilizers and plant protection chemicals and the lower frequency of application reduce the health threats to farmers in comparison to annual crops.

| Access to water
The survey results (see Supplementary information 2, Figure SI-1) show that the farmers do not anticipate water scarcity or a worsening of the water quality in the region. Nevertheless, they value water highly in their scoring of social aspects. This aspect is relevant when assessing agricultural activities since irrigation in agriculture is the main human activity leading to water stress (Pfister & Bayer, 2014).
An assessment of the water demand of miscanthus in the region considering hybrid type, local conditions (e.g. weather, soil) and number of hectares dedicated to the crop would be important to understand its potential impact on the local water bodies and water supply. Taking the hybrid type into account is crucial because wateruse efficiency varies according to miscanthus hybrid. Furthermore, hybrids differ in their water use strategy, for example, Miscanthus sinensis regrows faster after a drought period than Miscanthus giganteus as the latter keeps its stomata open, maintaining a green canopy. Studies on various miscanthus genotypes have shown that the crop has a high water-use efficiency (7.8-9.2 g DM kg H 2 O −1 ) compared with C 3 and some other C 4 plants . However, it requires considerable amounts of water to achieve a high productivity. As long as the water demand is met through precipitation, the impact of miscanthus on water availability should not be critical.

| Land rights and land consolidation
The literature review performed in this study provides some important indications of the complex situation of land issues in the region. Land conflicts can be a major barrier to agricultural development in general and, as a consequence, to miscanthus cultivation. In addition, the amount of land the farmers in the sample rent in or out indicates that most farmers lease the land they work. This is often a major barrier to miscanthus adoption, since leasing contracts are usually only short term and do not allow for long-term investment in a perennial crop. Miscanthus cultivation would require farmers to have access to the land for the entire 20-year cultivation period, which can only be achieved by long-term leasing contracts or farmers possessing their own land.
With respect to the high fragmentation of land in the region, miscanthus could be an interesting option for fragmented parcels. Winkler et al. (2020) stated that miscanthus could be an alternative to annual crops for cultivation on relatively small parcels and awkward land shapes. On awkwardly shaped land, agricultural operations tend to be time-consuming and costly for all arable crops. As soil cultivation for miscanthus cultivation is required only once in a 20-year period, the amount of operations and therefore time spent and costs incurred over a long-term period are significantly reduced.

| Local employment and income
An assessment of the potential of miscanthus cultivation to contribute to local employment should take into account the potential scarcity of agricultural workers in the region. Approximately 66% of respondents in this study suggested that such a scarcity was one of the main problems related to the non-use of agricultural lands.
With regard to income, the long establishment phase in miscanthus cultivation means farmers need to make an initial investment with no returns in the first 2 years. This could be a barrier to adoption of the crop. A solution to overcome this barrier could be granting farmers economic support during this phase, for example via loans (Bocquého & Jacquet, 2010) or subsidies. Given an existing market for miscanthus biomass, alternative strategies could be adopted by farmers to deal with high up-front costs. One of these could be the gradual adoption of the crop, so that future establishment of miscanthus is financed by already productive parcels.

| Food security
Miscanthus is a non-food crop and its cultivation on arable land has been controversially discussed due to the risk of indirect land use change (iLUC). However, its cultivation on marginal lands unsuitable for food production can reduce iLUC risks. Currently, the law in Croatia allows miscanthus cultivation on low-grade land (Croatian Parliament, 2018), minimizing the risk of its cultivation on arable land designated for food production.

| Limitations
The logistic regression model could be improved by collecting more data and increasing the sample size, as this is the main limitation of this study.
A limitation to the scoring of social aspects in this study is the consideration of a single stakeholder only. Ideally, a range of stakeholders' perspectives need to be taken into account, since important human rights aspects can only be identified by involving different perspectives. This could be the case, for example, for forced or child labour, as this information could be difficult to obtain from farmers.

| CONCLUSIONS
Opportunities for miscanthus development in Sisak Moslavina are related to the high yields of the crop in the region compared to other regions in Europe. The low inputs required can be seen as a characteristic that could make miscanthus cultivation attractive to farmers. In addition, the crop requires fewer working steps over the long term compared to annual arable crops, with tillage operations taking place only at the beginning of the 20-year period. It could provide an additional source of income for the farmer if adopted as an on-farm diversification strategy for plots that are not suitable for growing food crops. Despite these opportunities, the successful development of miscanthus in the region faces several barriers. These are related to the absence of (guaranteed) markets-a key aspect if farmers are to become familiar with the crop and consider its cultivation. Miscanthus is a new crop in the region and is not well known by farmers. This lack of knowledge influences farmers' attitudes towards it, as they are unaware of the advantages and disadvantages of adopting it. Problems related to ownership and disposal of land in the region indicate that land is not readily available to farmers. The high fragmentation of land in the region is an additional barrier for agriculture in general, since it renders modernization difficult and generates higher costs for farmers. Considering the high initial investment required for the establishment phase of miscanthus, subsidies are another key aspect for the successful development of the crop in the region but are currently not provided for miscanthus. Farmers renting land will not be willing to make long-term investments for miscanthus cultivation, as it requires a 20-year land commitment. Such long-term investments will of course not be perceived in the same way by young farmers as by elderly farmers, in turn highlighting the importance of generational renewal programmes in agriculture.
This study also highlights the social aspects regarded as relevant by the farmers interviewed. It reveals that economic aspects, such as profitability, are of utmost relevance for farmers when deciding to adopt a crop, but are not the only aspects of agricultural projects that affect their lives. In addition to income and employment, the farmers in this study mainly valued health and safety features, access to water and land, and food security. Therefore, these aspects should also be considered when assessing potential positive or negative impacts of agricultural projects in Sisak Moslavina.

ACKNOWLEDGMENTS
This study was supported by the Bio-Based Industries Joint Undertaking under the European Union's Horizon 2020 Research and Innovation Programme, grant agreement no. 745012 (GRACE project). This reflects only the authors' view and the European Commission is not responsible for any use that may be made of the information it contains. The authors thank Maja Rapić for her contribution to the research. The authors are grateful to Nicole Gaudet for proofreading the manuscript.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in Dryad at http://doi.org/10.5061/dryad.2z34t mprn.