Survey of 47 oilseed flax (Linum usitatissimum L.) growers to identify ways to expand its cultivation in France

Over the past 20 years in France, oilseed flax (Linum usitatissimum L.) has usually been grown on less than 20 000 ha per year while needs of the French linseed sector are estimated at around 30 000 ha per year. In this article we identify ways to increase the cultivation of oilseed flax in France. Forty-seven linseed producers were surveyed in January 2013. The aim of this survey was to (i) describe farming systems that currently grow oilseed flax (e.g. size, nature of production, farmers’ goals), (ii) identify farmers’ perceptions of strengths and weaknesses of oilseed flax in their farming systems, and (iii) test farmers’ reactions to scenarios created to facilitate expansion of oilseed flax cultivation. Surveys reveal a wide diversity of farming systems that produce linseed, from those specialised in crop production to those specialised in animal production. According to surveyed farmers, main advantages of oilseed flax are related to its compatibility with the farm work schedule. It is also useful agronomically, as it is a beneficial preceding crop and increases the duration of crop rotations. It is also a valuable alternative to crops that become difficult to grow. Its main drawback is that it provides insufficient profits. As for the expansion scenarios, cultivation of oilseed flax may increase if the linseed sector addresses genetic improvement (i.e. new oilseed flax cultivars), technical progress (especially in pest and disease management), and economic subsidies (e.g. a minimum price for linseed).


Introduction
Over the past 20 years in France, oilseed flax (Linum usitatissimum L.) has usually been grown on less than 20 000 ha per year, with mean annual production of 22 441 ± 13 842 t (Fig. 1).Consequently national demand is not satisfied as mean annual consumption of linseed was 45 848 t over the last five years (2009)(2010)(2011)(2012)(2013), nearly twice the amount of French production (Agreste, 2015).Demand is high and comes from the agricultural industry (mainly to produce linseed oil, Merrien et al., 2012;Zanetti et al., 2013) and the cattle-feed industry.The latter incorporates linseed into animal feed to produce feed enriched with the high omega-3 fatty acid content which linseed offers (e.g.Corino et al., 2014).
Cultivation of oilseed flax also has non-economic benefits, such as allowing farmers to diversify their crop rotations.Diverse crop rotations are known to reverse many of the negative socio-economic and environmental impacts of short rotations or rotations with few species, both of which are characteristics of modern crop-production systems (e.g.Davies et al., 2012).Long and diverse rotations with an appropriate crop sequence may increase yields due to "rotation effects" such as improved disease control nitrogen nutrition and water supply.For example, wheat yield after a flax crop was higher than that after wheat or barley (Kirkegaard et al., 2008).Less use of chemical inputs (particularly fungicides and mineral nitrogen fertilisers) than that for crops such as wheat is another benefit of growing oilseed flax; this reduction occurs for oilseed flax itself as well as for the following crop, especially wheat (Onidol, 2011).
However, despite "long-term breeding programs and indepth agronomic studies" (Zanetti et al., 2013), potential oilseed flax yields are difficult to achieve due to frequent unfavourable climatic conditions (e.g.frost, wind lodging), especially for spring-sown cultivars (Doré and Varoquaux, 2006).Moreover, certain steps of crop management are challenging such as harvest and straw management.For these reasons, farmers may consider its profits insufficient, especially compared to those of wheat (Merrien et al., 2012).
This mix of advantages and disadvantages may explain why French oilseed flax cultivation remains low and incon-sistent and depends on the economic context.Peaks in oilseed flax cultivation (Fig. 1) were due to subsidies of the Common Agricultural Policy of the European Union (1994,1999) and development of new markets (e.g.cattle feed in 2006-2007) (Labalette et al., 2011).
Currently, the French oilseed crop organisation (Onidol) and the French technical institute for oilseed crops (Cetiom) consider an area of 20 000-30 000 ha as an appropriate objective in France (Labalette et al., 2011).To achieve this objective, the main issue is to discover ways to encourage more farmers to produce more linseed.
Based on one face-to-face survey conducted in 2013 with each of 47 French linseed producers in the main production areas, the present study aims to identify ways to develop oilseed flax cultivation on their farms and to introduce this crop to other farms.

The survey
We used the survey method of Merrien et al. (2013).From 21-28 January 2013, 47 farmers were interviewed for approximately 90 minutes each by students of AGRO-CAMPUS OUEST, a French university-level college, also known as a "Grande École".Farms were located in western France (Fig. 2  -Test of three scenarios for expanding oilseed flax cultivation on the farm. Each "scenario" was a hypothetical situation described in detail to the surveyed farmers to test their reactions about expansion of oilseed flax crops on their farms.After each scenario was described, surveyed farmers were asked three questions: Would this scenario incite you to expand cultivation of oilseed flax on your farm?What are the advantages of this scenario?What improvements (of the scenario) do you suggest?
The first scenario ("economic") tested whether or not under certain conditions, payment of subsidies by authorities would be sufficient for farmers to increase oilseed flax cultivation.The compensation would be calculated as the difference between the gross margin of oilseed flax and those of main crops in the region.The conditions would require that at least four different crops be grown on the farm and that fewer pesticides and nitrogen fertilisers be applied to oilseed flax than to the main crops.
The second scenario ("agronomic") tested whether or not better information about agronomic advantages of oilseed flax and solutions to overcome its management problems would be sufficient for farmers to increase oilseed flax cultivation.Agronomic advantages include "rotation effects" and "break-crop benefits".Solutions to overcome flax-management problems include a wide range of oilseed flax cultivars adapted to multiple soil and climate combinations; new pesticides to control flax lodging, diseases and pests; new equipment for successful flax harvest; and the guarantee of useful technical advice.
The third scenario ("alternative") tested whether or not oilseed flax is a valuable alternative to other conventional crops if they become difficult to grow (e.g.pesticide bans, low prices, emergence of new pests).
The open-source, free software LimeSurvey r was used to create the online question-and-answer survey and to collect data (LimeSurvey, 2014).

Data analysis
The raw data file included 47 rows (individuals) and 377 columns (variables).Data editing and the descriptive analyses (percentage distributions, averages, standard deviations, 95% confidence intervals) were performed using Microsoft Excel r .A Multiple Factor Analysis (Pagès, 2014) was performed with R software (R Core Team, 2012) using continuous variables (age of farmers, farm arable land, proportion of farm arable area under oilseed flax in 2011-2012, importance of seven objectives in making decisions about farm management, ranking of the three scenarios for expanding oilseed flax cultivation on the farm) and categorical variables (farm type (see below), farmers' projects divided into seven categories).As one farmer did not rank scenarios, dataset is made of 46 rows and 21 columns.
We used the typology for agricultural holdings of the European Commission's (EC) Farm Accounting Data Network (Commission regulation (EC) No. 1242/2008 of 8 December 2008) to classify surveyed farms according to their main production.This typology is defined by structural farm characteristics (e.g.crop areas, animal heads) and "standard output" of crop and animal products, which is a product's mean farmgate price (e ha −1 or e animal head −1 ) over a 5-year period in a given region (Regulation (EC) No. 1242/2008).A farm's overall "economic size" (in e) is calculated as the sum of the quantities of products it produces (e.g.ha, head) multiplied by their respective standard outputs.A farm is considered a specialised farm type if the economic size of one or more types of production exceeds two thirds of its overall economic size.Farm types include the following: "specialist field crops", "specialist horticulture", "specialist permanent crops", "specialist grazing livestock", "specialist granivores", "mixed cropping", "mixed livestock holdings", "mixed cropslivestock", and "non-classified holdings" (Regulation (EC) No. 1242/2008).Farms specialised in animal production are one of those farm types: "specialist grazing livestock", "specialist granivores", "mixed livestock holdings".
Among farms specialised in animal production, less than half were "specialist granivores" farms, whose main production is confined rabbits poultry and/or pigs.The other farms specialised in animal production were "specialist grazing livestock" and "mixed livestock holdings" farms (60% and 57% of farms specialised in animal production in Ille-et-Vilaine/Mayenne and Loire-Atlantique/Vendée, respectively), mainly dairy farms.

Descriptive statistics of the 47 surveyed farms
Mean arable land area (±1 standard deviation) of the 47 farms was 142 ± 62 ha, with larger farms in Eureet-Loir/Loiret (194 ± 49 ha) and smaller farms in Ille-et-Vilaine/Mayenne (94 ± 38 ha) (Tab.1).These differences are partly explained by the presence (or not) of livestock.Mean arable land area of farms specialised in crop production (148 ha) was higher than that of farms specialised in animal production (123 ha).As mentioned, all farms in Eureet-Loir/Loiret specialise in crop production whereas farms specialised in animal production occur more frequently in the two other production areas.
Unlike arable land area the mean number of workers per farm (2.1 ± 1.1) varied somewhat among production areas as farms with animal production had more workers per ha (Tab.1).The mean number of workers per ha equalled 0.023, 0.012 and 0.013 for specialised animal production, specialised crop production, and "mixed crops-livestock" farms, respectively.
Regardless of farm type and size, few farmers (13%) used only their own equipment for farm work; the others relied on service-supply agencies and/or cooperative associations for the use of agricultural equipment.
Contracting was common in the sample as 74% of surveyed farmers produced at least one crop under contractual arrangements in 2011-2012 (because surveys occurred in January 2013 we asked this question for the last complete cropping season; Fig. 4).Contractual arrangements were generally prepared and signed between farmers and an agricultural cooperative (here, Cavac and Terrena) or an agricultural company (here, Groupe Soufflet).In 2011-2012, all surveyed farmers in Eure-et-Loir/Loiret produced at least one crop under contractual arrangement, which contrasts with those in Ille-et-Vilaine/Mayenne where only 40% of farmers produced one or more crops under contractual arrangements.Among the 32 farmers who grew (or planned to grow) oilseed flax in 2012-2013 only ten (31%) had (or planned to have) contracts for this crop.
Regarding the economic dimension of farms, gross operating surplus helps analyse the economic profitability of an agricultural holding without considering its investment policy, financial management or exceptional charges.A worker (farmer, farm worker) is a person who works full-time on a farm for one year.† Gross operating surplus is "the balance of the trading account for companies" and is equal to "value added minus payroll and other taxes on production and plus operating grants" (INSEE, 2014).Four farmers did not provide these data.(133 782 ± 94 927 e) but lower in Ille-et-Vilaine/Mayenne (85 333 ±40 283 e).This is partly related to farm type as mean gross operating surplus of "specialist grazing livestock" -a farm type more common in Ille-et-Vilaine/Mayenne than in the two other production areas -was the lowest (87 111 e).
For the four other farm types it varied from 96 411 e ("specialist field crops") to 230 850 e ("mixed livestock holdings").

Crop areas in 2011-2012
In 2011-2012, cultivation of cereals (especially, winter wheat) was much higher than that of other crops: from a mean of 52% of arable area in Loire-Atlantique/Vendée to 62% of arable area in Ille-et-Vilaine/Mayenne (Tab.2).Cultivation of oilseed rape was relatively high in Eure-et-Loir/Loiret (23% of arable area), lower in Ille-et-Vilaine/Mayenne (15% of arable area) and rare in Loire-Atlantique/Vendée (3% of arable area) while sunflower was present only in this last area.Unsurprisingly, fodder crops (i.e.grasslands, silage maize) were rarely grown in Eure-et-Loir/Loiret (0.3% of arable area) but represented 15-24% of arable area in the two other production areas.Except in grasslands no farmers adopted intercropping.Oilseed flax cultivation varied from 5% (Ille-et-Vilaine/Mayenne) to 9% (Eure-et-Loire/Loiret) of arable area.

Cropping sequences
Excluding grasslands, we identified 133 main cropping sequences on surveyed farms (Fig. 5).Despite this diversity, cropping sequences were dominated by wheat, which is consistent with its high coverage of arable areas.Oilseed flax was included 32% of the cropping sequences.Durations of cropping sequences varied from 1-7 years (mean = 3.3 years).In this sample, monocultures (mainly maize fields) were rare, D604, page 5 of 14   comprising less than 5% of the 133 cropping sequences and 80% of cropping sequences lasted 2-4 years.Two-year cropping sequences alternated wheat with another crop such as maize (47%), oilseed flax (19%) and rape (16%).Three-year cropping sequences were also based on wheat: 61% alternated wheat and barley (or triticale) with rape (or another crop such as maize or oilseed flax) while 39% alternated wheat and ryegrass with maize (or another crop).In order of importance, 4-year rotations alternated wheat with (i) rape and another crop (oilseed flax or maize), (ii) sunflower and another crop (maize, oilseed flax or rape), and (iii) other crops such as oilseed flax and maize.

Farmers' crop management strategies
To understand how farmers manage their crops, we asked them to classify ten objectives (plus one open-response: "other" objective) that drive winter-wheat management.Winter wheat was chosen because it was grown on all farms.
The most important objectives farmers expressed were related to wheat grain quality and maximum yield (Fig. 6).
Preserving the environment, reducing operating costs, and reducing the risk of crop failure remained important objectives.Others, such as reducing labour time, remaining "on top" technically, and making life easier did not play a key role for wheat management strategies.Surveyed farmers' agricultural practices thus seemed intensive (seeking maximum yield) but oriented toward sustainability (i.e.preserving the environment, reducing operating costs, seeking the highest quality) and relatively traditional (i.e.no interest in the most recent cropping practices, aversion to risk-taking).This is consistent with the typical yields reported by farmers (Tab.3) and their relatively low participation rate (26%) in European Union agrienvironmental schemes.

Incentives and disincentives for linseed
production on the 47 surveyed farms

Cropping sequences including oilseed flax
Thirty-two farmers farmers grew (or intended to grow) oilseed flax in 2012-2013, 12% of them, all located in Loire-Atlantique/Vendée for the first time.Among farmers who did not grow oilseed flax in 2012-2013, 87% had decided definitively to stop growing the crop, of whom 60% had already stopped the previous year.
Oilseed flax could follow wheat in 93% of cropping sequences and may be followed by wheat in 96% of cropping sequences (Tab.5).Surveyed farmers clearly considered this the best cropping practice.Fifty-seven percent of 28 farmers (those who grew (or intended to grow) oilseed flax in 2012-2013 and mentioned cropping sequences including oilseed flax) consider wheat the best preceding crop for oilseed flax (other preceding crops cited were maize, sunflower, and grasses), mainly due to the appropriate compatibility between the work schedules of oilseed flax and wheat.Sixty-eight percent of the farmers considered wheat the best crop to follow oilseed flax, mainly due to the impact of oilseed flax roots on soil structure.Crops often mentioned as not recommended before or after oilseed flax were oilseed crops (mainly rape and sunflower) and maize (data not shown).

Farmers' strategies for oilseed flax crop management
It is important to know that 66% of the 32 surveyed farmers who grew (or intended to grow) oilseed flax in 2012-2013 considered that they had mastered oilseed flax crop management, while 25% still considered themselves novices.As for wheat, we asked farmers to classify ten objectives (plus one open-response: "other" objective) that drive oilseed flax crop management.
For the 32 farmers that grew (or intended to grow) oilseed flax in 2012-2013, the most important objectives were related to grain quality and maximum yield (Fig. 7), as observed for wheat management Typical attainable oilseed flax yields reported by farmers were 2.0 ± 0.4 t ha −1 for winter-sown cultivars (25 mentions) and 2.1 ± 0.6 t ha −1 for spring-sown cultivars (2 mentions).Of these farmers, only 9% directly sowed oilseed flax; the rest decompacted soil and/or performed stubble cultivation and/or ploughed fields, then prepared the seedbed for oilseed flax (most used an integrated mechanical seed drill that combines a rotary harrow and a sowing machine) (Fig. 8).Soil decompaction was uncommon (16% of farmers).It was performed using a deep cultivator such as a subsoiler, a soil loosener, or a heavy cultivator.The aim of this operation is to restore damaged soil by eliminating the plough pan and to separate soil particles to increase pore space.Stubble cultivation was performed by 50% of the farmers, generally using disc harrows and tined cultivators in one or more passes (Fig. 8).The purpose of this operation is (i) to break and bury harvest residues, (ii) to manage weeds (it creates a false seedbed that promotes germination and emergence of weed seeds which are mechanically or chemically destroyed before sowing oilseed flax), and (iii) to remove compaction.Ploughing was less common (only 25% of farmers) (Fig. 8).Winter oilseed flax cultivars were sown from mid-September to mid-October.
Diseases, weeds, and pests were managed as follows: -For disease control, 94% of farmers used pesticides, in 1-3 passes, mostly in autumn.The most frequently cited disease, stem break, is caused by the pathogenic fungus Kabatiella lini (Gruzdeviene et al., 2008).Farmers often mentioned that treatments may not be successful due to difficulty accessing fields in autumn.-For weed control, all farmers used herbicides, in 1-3 passes, mostly in autumn.Oilseed flax does not compete well with weeds (Ehrensing, 2008).Like for disease control, 22% of farmers mentioned that treatments may not be successful due to difficulty accessing fields in autumn.-For pest control, less than 25% of farmers used pesticides.
Lodging was a significant concern for farmers: 69% always used a plant growth regulator.These farmers often mentioned that plant growth regulation is a known effect of a fungicide they use for it.
After harvest and field retting, 78% of farmers removed straw from fields and usually sold it as a co-product.Even though straw management period is a crucial one (retting is a long process that delays future agricultural practices such as stubble cultivation), farmers appreciated this additional income.

Why surveyed farmers grow or stop growing oilseed flax
Among suggested reasons for growing oilseed flax, two were not chosen by farmers: improvement of soil organic matter content by oilseed flax straw and use of linseeds as a feed supplement for animals (Fig. 9).Several organisational and agronomic reasons were chosen by most farmers; for instance: -Oilseed flax crop management is well adapted to the farm work schedule and allows farmers to spread their workloads, especially sowing and harvesting which occur in different seasons than those of wheat.-Introducing oilseed flax into rotations diversifies and increases durations of cropping sequences.
Some reasons were seen as constraints that hinder oilseed flax expansion to other farms.Oilseed flax was not perceived as a crop that (i) requires few inputs; (ii) provides a competitive gross margin; (iii) is well adapted to drained fields, or (iv) protects soil against erosion.
As for reasons to stop growing oilseed flax, many farmers had no opinion, knowing better why they grow it than why they stop (or would stop) (Fig. 10).Nonetheless, 77% of farmers identified the lower profitability of linseed is its main weakness.Among "other" reasons, farmers also cited problems during harvesting, inconsistent yields and crop sensitivity to climate (e.g.freezing).

Scenarios: how to expand oilseed flax cultivation
Of surveyed farmers, 43% considered the agronomic scenario the most favourable for expanding oilseed flax cultivation on their farms or additional farms, followed by the economic scenario (37%) and the alternative scenario (20%) (Tab.6).Differences appeared among production areas.In Eure-et-Loir/Loiret more than 70% of farmers ranked the economic scenario last.In Ille-et-Vilaine/Mayenne, 56% ranked the agronomic scenario first, while 44% ranked it last.

Opinions of the three scenarios
Sixty-six percent of the 44 farmers with an opinion considered the economic scenario a favourable solution for promoting linseed production: gross margin compensation would encourage farmers to increase cultivation of this crop (Tab.7).This is in agreement with Meynard et al. (2014) who suggested that actors of oilseed fax sector adjust price of oilseed flax to price of dominant crops.However, in Eure-et-Loir/Loiret, more than 66% of farmers rejected this scenario, in contrast with those from the two other areas.Main weaknesses of this scenario were the agri-environmental restrictions as farmers would like to receive financial aid without any governmental control over their agricultural practices.Farmers also considered over-dependency on public aid as dangerous.Some considered the economic scenario unrealistic since its success depends on technical and genetic improvements that do not yet exist.
Since the agronomic scenario assumes that the main problems of oilseed flax cultivation are resolved, 66% of the   7. Strengths and weaknesses mentioned by at least three surveyed farmers of three scenarios for expanding oilseed flax cultivation on their farms (or to introduce oilseed flax to additional farms).Scenarios are as follows: under agri-environmental conditions, payment of compensation for the income foregone and additional costs incurred (economic scenario); better information on agronomic advantages of oilseed flax and highlighting solutions to overcome crop management difficulties (agronomic scenario); oilseed flax cultivation as a valuable alternative to other conventional crops if they become difficult to grow (alternative scenario).

Scenario
Strengths

Conclusion
Forty-seven French linseed producers were surveyed to identify ways to increase oilseed flax cultivation on their farms and to introduce it to other farms.This was achieved by characterising farming systems that currently crop oilseed flax.Farm types were diverse, from farms specialised in field-crop production to farms specialised in animal production.The mean Individuals are not plotted to enhance graphical display.Objectives for making decisions about farm management are as follows: O1, to get more free time; O2, to improve quality of life; O3, to preserve the environment; O4, to improve social recognition of farmer profession; O5, to increase farm income; O6, to increase farm income stability; O7, to facilitate farm succession.Importance of these objectives in making decisions about farm management varies from 1 (not considered at all) to 4 (crucial).Other continuous variables are as follows: AGE, age of farmers; FAA, farm arable area; FLAX, proportion of farm arable area under oilseed flax in 2011-2012; AGR, ranking of the agronomic scenario; ALT, ranking of the alternative scenario; ECO, ranking of the economic scenario.Ranking of the three scenarios varies from 1 (the least favourable) to 3 (the most favourable).Farm types are as follows: FT_1, specialist field crops; FT_4, specialist grazing livestock; FT_5, specialist granivores; FT_7, mixed livestock holdings; FT_8, mixed crop-livestock.area of arable land on surveyed farms was 142 ha, and oilseed flax covered < 10% of that, on average (compared to > 50% for cereals).Oilseed flax was mainly included in short cropping sequences (< 4 years long).Farmers' crop-management practices seemed intensive but oriented toward sustainability and relatively conventional.
According to surveyed farmers, oilseed flax was interesting mainly because it was compatible with the farm work schedule (i.e.farmers can spread their workloads) and diversified crop production.It was agronomically useful because it was a beneficial preceding crop and increased the durations of crop rotations.It was also a valuable alternative to conventional crops if they become difficult to grow.Finally it guaranteed commercial outlets.One drawback was that it was not a low-input crop.Because oilseed flax yields are usually low and highly variable, profitability was the main reason why farmers stopped (or would stop) growing flax.To conclude, expansion of oilseed flax cultivation would occur if the linseed sector focuses on the following: -Genetic improvement by developing new oilseed flax cultivars that are more productive and more resistant against biotic and abiotic stresses.-Technical progress that offers new solutions for pests diseases and problems during harvest.-Subsidies to ensure a minimum price (or a minimum gross margin) to linseed producers.
) in three main production areas: Eureet-Loir/Loiret (n = 7), Ille-et-Vilaine/Mayenne (n = 10) and Loire-Atlantique/Vendée (n = 30).The names of the farmers had been provided by two agricultural cooperatives (Cavac, Vegam), one agricultural company (Ets Dutertre), and one linseed producer's association (Graine de Lin 28).All farmers surveyed had grown oilseed flax (four for the first time in 2012-2013 and 15 who did not grow it in 2012-2013).Consequently, in 2012-2013, 32 farmers grew (or planned to grow -surveys occurred in January 2013 after farmers seeded winter-sown cultivars but before they seeded spring-sown cultivars) oilseed flax.This sample of oilseed flax growers is clearly not representative of French agriculture since only 0.3% of French farms grow oilseed flax: specifically 1.9% (155 farms) in Eure-et-Loir/Loiret, 0.9% (169 farms) in Ille-et-Vilaine/Mayenne, and 1.7% (213 farms) in Loire-Atlantique/Vendée (2010 agricultural census in Agreste, 2015).The survey covered the following points: -General information about the farmer: age, personal goals (e.g. have more free time, earn more money, preserve the environment), projects for the farm (e.g.farm expansion, field drainage, field irrigation, purchase of agricultural equipment).-Description of the farming system: legal status of the farm, number of workers, areas of crops, yields of main crops, sizes and production of livestock populations, description of the main fields (e.g.surface area, advantages, disadvantages).-Focus on the oilseed flax crop: reasons to grow oilseed flax, reasons to stop growing oilseed flax (if any), oilseed flax yields, description of rotations including oilseed flax, description of oilseed flax crop management.

Fig. 2 .
Fig. 2. Location of French communes where the 47 farms surveyed in January 2013 are located (square, one farm in the commune; circle, two farms in the commune; diamond, three farms in the commune).Names of French departments are followed by their oilseed flax surface areas in 2013 (Agreste, 2015).Darker shading of the departments indicates higher oilseed flax cultivation.

Fig. 3 .
Fig. 3. Classification of 47 surveyed farms according to their farm type in each of three production areas.Farm types are as follows: plain grey bars, specialist field crops; thick grey hatched bars, specialist grazing livestock; thin grey hatched bars, specialist granivores; grey dotted bars, mixed livestock holdings; grey dashed bars, mixed crop-livestock.

Fig. 4 .
Fig. 4. Percentage of 47 surveyed farmers, grouped by production area, according to whether they produce crops under contractual arrangements in 2011-2012.Legend is as follows: plain dark bars, none; plain grey bars, for one or more crops except oilseed flax; thick grey hatched bars, for one or more crops and oilseed flax; grey dotted bars, for oilseed flax only.

Fig. 6 .
Fig. 6.Classification of the importance of ten objectives (plus one open-response: "other" objective) in making decisions about winterwheat management for 47 surveyed farmers.Legend is as follows: plain dark bars, crucial; plain grey bars, important; thick grey hatched bars, not important; grey dotted bars, not considered at all; white bars, no response.

Fig. 7 .
Fig. 7. Classification of the importance of ten objectives (plus one open-response: "other" objective) in making decisions about oilseed flax crop management for 32 surveyed farmers that grew (or intended to grew) it in 2012-2013.Legend is as follows: plain dark bars, crucial; plain grey bars, important; thick grey hatched bars, not important; grey dotted bars, not considered at all; white bars, no response.

Fig. 8 .
Fig. 8. Description of 32 oilseed flax management strategies reported by surveyed farmers who grew (or intended to grow) it in 2012-2013.Each row represents main agricultural practices of oilseed flax crop management.Contrary to the others, the penultimate line describes crop management for a spring-sown cultivar.Solid grey boxes indicate that farmers manage these operations, while thick grey hatched boxes indicate that farmers sometimes manage these operations.Boxes surrounded by a dotted line represent operations performed together.For stubble cultivation, the number in parentheses is the number of tractor passages (if >1).For nitrogen fertilisation, the number is the nitrogen application rate in kg N ha −1 (if mentioned).For plant growth regulator, F means that the farmer indicates the plant growth regulator is a fungicide.B, buried straw; C, chopped straw; CC, cover crop; CK, crosskill roller; CP, cultipacker; DH, disk harrow; H, harrow; R, straw removal; RH, rotary harrow; Rho, rotary hoe; RL, roller; SC, S-tine cultivator; SCu, stubble cultivation.

Fig. 9 .
Fig. 9. Reasons why 47 surveyed farmers decided to grow oilseed flax.Legend is as follows: plain dark bars, agree; plain grey bars, disagree; thick grey hatched bars, no opinion.

Fig. 10 .
Fig. 10.Reasons why 47 surveyed farmers decided to stop (or would decide to stop) growing oilseed flax.Legend is as follows: plain dark bars, agree; thick grey hatched bars, no opinion.

Fig. 11 .
Fig. 11.Representation of continuous variables (a) and categorical variables (b) on the first plane issued from Multiple Factor Analysis.Individuals are not plotted to enhance graphical display.Objectives for making decisions about farm management are as follows: O1, to get more free time; O2, to improve quality of life; O3, to preserve the environment; O4, to improve social recognition of farmer profession; O5, to increase farm income; O6, to increase farm income stability; O7, to facilitate farm succession.Importance of these objectives in making decisions about farm management varies from 1 (not considered at all) to 4 (crucial).Other continuous variables are as follows: AGE, age of farmers; FAA, farm arable area; FLAX, proportion of farm arable area under oilseed flax in 2011-2012; AGR, ranking of the agronomic scenario; ALT, ranking of the alternative scenario; ECO, ranking of the economic scenario.Ranking of the three scenarios varies from 1 (the least favourable) to 3 (the most favourable).Farm types are as follows: FT_1, specialist field crops; FT_4, specialist grazing livestock; FT_5, specialist granivores; FT_7, mixed livestock holdings; FT_8, mixed crop-livestock.Farmers' projects (0, no; 1; yes) are as follows: P1, change in farm size; P2, change in type and intensity of animal production; P3, change in type and intensity of crop production; P4, change in crop management; P5, change in farm territory; P6, change in farm equipment; P7, change in work organisation.
Fig. 11.Representation of continuous variables (a) and categorical variables (b) on the first plane issued from Multiple Factor Analysis.Individuals are not plotted to enhance graphical display.Objectives for making decisions about farm management are as follows: O1, to get more free time; O2, to improve quality of life; O3, to preserve the environment; O4, to improve social recognition of farmer profession; O5, to increase farm income; O6, to increase farm income stability; O7, to facilitate farm succession.Importance of these objectives in making decisions about farm management varies from 1 (not considered at all) to 4 (crucial).Other continuous variables are as follows: AGE, age of farmers; FAA, farm arable area; FLAX, proportion of farm arable area under oilseed flax in 2011-2012; AGR, ranking of the agronomic scenario; ALT, ranking of the alternative scenario; ECO, ranking of the economic scenario.Ranking of the three scenarios varies from 1 (the least favourable) to 3 (the most favourable).Farm types are as follows: FT_1, specialist field crops; FT_4, specialist grazing livestock; FT_5, specialist granivores; FT_7, mixed livestock holdings; FT_8, mixed crop-livestock.Farmers' projects (0, no; 1; yes) are as follows: P1, change in farm size; P2, change in type and intensity of animal production; P3, change in type and intensity of crop production; P4, change in crop management; P5, change in farm territory; P6, change in farm equipment; P7, change in work organisation.

Table 1 .
Descriptive statistics of 47 farms growing oilseed flax in western France.

Table 3 .
Mean and 1 standard deviation (SD) of typical yields (t ha −1 ) of four main crops reported by 47 surveyed farmers.

Table 4 .
95% confidence intervals of surface area (ha) and percentage of oilseed flax on farms of 47 surveyed farmers in 2011-2012 in three main production areas.

Table 5 .
The 22 cropping sequences mentioned by the 28 ‡ surveyed farmers who grew (or intended to grow) oilseed flax in 2012-2013.

Table 6 .
Ranking of three scenarios for expanding/introducing oilseed flax cultivation on a farm from the most (rank 1) to the least favourable (rank 3).One farmer did not rank scenarios.

Table
and Weaknesses n