Metropolitan France is on the western edge of Europe. It is situated mostly between latitudes 41° and 51° N (Dunkirk [Dunkerque] is just north of 51°), and longitudes 6° W and 10° E. It lies within the northern temperate zone (Figures 1a and 1b).
The French Republic also has territories in North America, the West Indies, in the Caribbean Sea, South America, the southern Indian Ocean, the Pacific Ocean, and Antarctica. These territories have varying forms of government ranging from overseas départements to overseas collectivity. These overseas regions will not be considered in this study. Indeed, plant production and animal production systems in these regions are very diverse due to the range of climates. They strongly differ from the metropolitan conditions and each of them only represents a limited production, even though it may be of major local importance.
The European territory of France covers 547 030 square kilometres, the largest area among European Union members. France has a wide variety of landscapes, from coastal plains in the north and west to mountain ranges of Vosges and Jura in the east, Alps in the south-east, the Massif Central in the centre-south and Pyrenees in the south-west. At 4 810.45 metres above sea level, the highest point in Western Europe, Mont Blanc, is in the Alps on the border between France and Italy. It is also important to mention the Hercynian Massif Armoricain in Brittany; although it has no impact on the altitude and climate, the metamorphic structure of the soil induces peculiarities of the cultivated soil such as a high proportion of silt, a fairly low pH, and impermeable subsoil. This conditions the composition of the natural grasslands in this region and has induced several management practices to cope with these peculiarities.
France has extensive river systems such as the Seine which flows from the Morvan in Burgundy to its estuary in the Channel, the Loire flowing from Massif Central to its estuary on the Atlantic coast near Saint-Nazaire, the Garonne which flows from the Pyrenees to its estuary on the Atlantic coast near Bordeaux, the Rhone, which separates the Massif Central from the Alps and flows into the Mediterranean Sea at the Camargue delta and the Rhine river which separates France from Germany and which flows northwards. There are also several large rivers in Lorraine region (Meuse and Moselle), flowing to the North Sea. Corsica lies off the Mediterranean coast. The mountain regions as well as river valleys play a key role in the location of grasslands, especially permanent grasslands and associated animal production.
The north and northwest have a temperate climate, while a combination of maritime influences, latitude and altitude produce a variety of climates in Metropolitan France. In the south-east a Mediterranean climate prevails. In the west, the climate is predominantly oceanic with high rainfall, mild winters and cool to warm summers. Inland the climate becomes more continental with hot, stormy summers, colder winters and less rain. The climate of the Alps and other mountainous regions is mainly alpine, with the number of days with temperatures below freezing over 150 per year and snow cover lasting for up to six months. This clearly means that there is a wide range of agricultural production, especially plant production. It also means that there will be a large range of conditions for grasslands and this will influence their botanical composition, their potential biomass production and the optimum management.
With an estimated population of 65.8 million (as of 1 Jan. 2011), France is the twenty-first most populous country in the world. In 2003, France's natural population growth (excluding immigration) was responsible for almost all natural population growth in the European Union. The natural growth (excess of births over deaths) rose to 302 432 in 2006, its highest since the end of the baby boom in 1973. The total fertility rate rose to 2.01 in 2010, from a nadir of 1.68 in 1994.
As of 2008, the French national institute of statistics INSEE estimated that 11.8 million foreign-born immigrants and their direct descendants (born in France) lived in France, representing 19% of the country's population. More than 5 million are of European origin and about 4 million of Maghrebi origin. Immigrants aged 18–50 account for 2.7 million (10% of population aged 18–50) and 5 million for all ages (8% of population).
The largest cities in France, in terms of population, are Paris (11 836 970), Lyon (1 757 180), Marseille (1 618 369), Lille (1 163 934), Toulouse (1 118 472), Bordeaux (1 009 313), Nice (999 678), Nantes (768 305) and Strasbourg (641 853).
Rural depopulation is a perennial political issue. Over the period 1960–1999 fifteen rural départements experienced a decline in population. In the most extreme case, the population of Creuse fell by 24%.
Agriculture only employs 3.3% of the active manpower (to be compared with 5.3% at European level). In less than a century, agriculture lost most of its workers. Indeed, in 1900, more than 50% of French people worked in agriculture. During the twentieth century, the number of farms was divided by ten, decreasing from more than 5 million to 550 000 at the end of the century. Technical progress in all domains (machinery, mechanization) made this shift possible, as it was possible to produce more with less labour. Since the beginning of this century, the number of farms has been falling by 10 000 every year, half of them being in herbivore production.
Moreover, the mean age of farmers is increasing steadily. The proportion of young farmers (under 35 years) was 18% in 2000 and is only 13% in 2010. Most farmers work full time; however, 10% have a second activity. This proportion is increasing since 2000.
Work is mainly organized around the family structure, where the extra manpower comes from; 120 000 spouses and 74 000 relatives provide this extra resource. However, this resource is decreasing and is progressively replaced by salaried employees. Today 140 000 permanent employees and seasonal workers are recorded in agriculture. Farms devoted to fruits, vegetables and flower production are the main sectors for salaried employment. On the other hand, in animal farming, more than 90% of farms are based on family workers.
Women comprise a quarter of the professional farmers. They even contribute 30% of new operations. But wives work less and less on farms, having more often a job outside. This is extremely important for younger generations. Indeed, 75% of wives below 30 have their own job, while this proportion is only 35% for those older than 50 years.
French agriculture is the leader in Europe, with 18% of European agricultural production. However, it only contributes 1.4% of national GDP and employs 3.3% of the workers. French agriculture is a net exporter, with a net market balance of 9 billion euros.With agricultural production evaluated at 67 billion euros, subsidies excluded, in 2008, France is ranked first in EU27, both for plant and animal production (Table 1). It contributes 18% of the agricultural production of EU (estimated at 379 billion euros), and is followed by Germany and Italy (48 billion euros, 13% of European agricultural production). The first 15 members of EU have a powerful role, as they contribute 83% of European agricultural production. However, since 1998, the share of new member states has regularly increased. In 2008, Poland and Romania contributed 6 and 4% of EU production respectively.
However, in terms of employment, France was ranked fifth in 2007, while Poland and Romania together used 40% of the agricultural manpower of EU.
The crude added value of the agricultural, forestry and fisheries accounted for 37.3 billion euros in 2007 with 30 billion for agriculture alone. This is a very small share of the national economy. It has to be compared with the tertiary sector (77% of national GDP with 1 309 billion euros), industry (238 billion) or construction (110 billion). The share of agriculture has steadily declined. While it contributed 4% of GDP in 1980, this share was only 2% in 2007, with 1.4% for agriculture only. Despite an increase in volume, this significant reduction was due to a marked decrease of agricultural product prices over the last 25 years.
In 2008, France exported 50 billion euros of agri-food products (raw and processed), mainly to EU countries (for 36 billion euros). Imports accounted for a total of 40.7 billion euros, again mainly from EU (28 billion). As a consequence, the surplus, in 2008, reached 9.3 billion euros, mainly because of the increase in export of cereals. Beverages (wines and spirits), cereals, dairy products and living animals are the products where surpluses are largest.
Topography and geology
Relief in France is shown in Figure 1b. France’s geology is complex combining sedimentary, metamorphic and volcanic areas (Figure 2). As a consequence, in some areas, variation within relatively short distances is far greater than in many parts of Europe. Due to this geology, the topography is very varied, with flat lands in the sedimentary basins, hilly landscapes in the old metamorphic areas and mountainous landscapes in the recent metamorphic areas, such as the Alps and the Pyrenees. This has major consequences on the types of grasslands (permanent and temporary) and on the types of animal production.
The soil map of France is also a consequence of the geological origin and is influenced by climate and also by production and agronomic practices. A national soil survey was initiated, in partnership between INRA and the Ministry of Agriculture (GIS Sol). Thanks to this survey, some maps are available at the national level. In this section, 5 key parameters of French soils are presented which are relevant for grasslands and for forage production. They were obtained from Arrouays et al. (2012) and Ranjard (2012).
The pH (Figure 3) is mainly neutral or slightly alkaline in the main sedimentary basins, i.e. Seine River and most of Aquitaine, acidic in Brittany and Massif Central. The soil is very acidic in part of Aquitaine, in the regions named Landes. On the very sandy soils, the main vegetation is forestry, with a large forest of pine trees, well adapted to this poor and shallow soil with the presence of an alios layer (which is a cemented ferruginous layer in these sandy soils).
The available phosphorus content is very high in some regions, such as Brittany, northern France, Alsace and some sectors of the Massif Central (Figure 4). This P soil content largely indicates the agronomic practices over the last decades, where an excessive P fertilization was often the rule.
Stocks of organic carbon were also measured in this network (Figure 5), for the first 30 cm of topsoil. Stocks tend to be high at high altitudes because of the climate and the subsequent slow organic carbon degradation. Levels are also high in regions with a high proportion of forests or with predominant cattle production and a high proportion of grasslands. C stocks also depend on soil texture and are higher in clay soils.
The risk of erosion was estimated for every small agricultural region (Figure 6). This risk was estimated thanks to the Mesales model (Modèle d’évaluation spatiale de l’aléa d’érosion des sols) implemented by INRA. It combines several characteristics of soil (susceptibility to soil sealing and erosion), land (land use, slope) and climate (intensity and quantity of rainfall). Risk was estimated in five classes representing the probability that erosion occurs. The risk appears to be high in the north of France and Brittany, in the Alps and in the south-west (Garonne and Adour Valleys). In these environments, the presence of permanent grasslands will be extremely relevant to prevent such risks.
A last parameter was documented - soil microbial biomass (Figure 7) estimated from the microbial DNA. It was shown that the microbial biomass was significantly and positively related to the microbial DNA extracted from soils. The H zones correspond to areas where the quantity of microbial biomass is high and L zones where it is low. This map shows a heterogeneous distribution of soil microbial biomass, but organized along geographic gradients. Such a distribution shows a large influence of soil pedology and of land use (forestry, grasslands, crop production, vineyards). At this national scale, no influence of climate is detectable. The sub-zones correspond to small geographic areas showing homogeneous and extremely low microbial biomass. For Sub-L-zone-3 (Sologne), this is likely due to a sandy soil while, in sub-L-zone 1, the long history of wine production may explain this low microbial biomass.
France has a wide range of climates (Figure 8) from oceanic towards continental ones, along a gradient from West to East. The Mediterranean coast has a pronounced Mediterranean climate. The Alps and the Pyrenees show a climate typical of high altitudes.
The distribution of climates influences various weather parameters, which are relevant for grassland production and vegetation types. It can be illustrated by the mean annual duration of sunshine (Figure 9), the rainfall (Figure 10) and temperature (Joly et al., 2010) (Figure 11).
As a consequence of the topography, climate and soil characteristics, there is a very wide range of agro-ecological regions in France, and it would be misleading to summarize them in a few types.
However, a few simple trends may be identified, which determine the main type of land use.
In the mountains, the Alps, the Pyrenees, the Massif Central, Vosges and Jura, where the temperatures are low, with high rainfall, forests are prevalent, except at high altitudes, where only permanent grasslands are present and they are called alpages (in the Alps) and estives (in the Pyrenees). At lower altitudes, where the slopes are not too steep, forests are often mixed with grasslands, and this is where the transition from grasslands to forest occurs. At lower altitudes, permanent grasslands become predominant and are then progressively replaced by temporary grasslands and annual crops in the plains.
There are two major exceptions to this general and over-simplified pattern. The first one is the pine forest in the Landes (Aquitaine) where the pine plantations were established in the nineteenth century on poor sandy soils which were marshlands. The second exception is the Mediterranean forest where the wood production is very low but where forests protect the soil against erosion and preserve typical vegetation and biodiversity.
Many French TV and newspaper advertisements promoting cheese show dairy cows grazing green grasslands in peaceful and typical landscapes. It underlines that, in the mind of publicists and consumers, grasslands and animals are full parts of the supply chain that leads to production of quality cheese. The evidenced link between grassland and landscape also stresses the complex relationship between the quality of a product and the characteristics of the location where it is produced.
This small example illustrates the role of grasslands as a feed resource for ruminants, as critical for quality of animal products and as a key component of quality of landscape and territory structure. All these functions are actually ecosystem services, alongside the preservation of biodiversity, quality of water and carbon storage. Thus, any analysis carried out on grasslands and forage crops must take into account economic, social and environmental dimensions, all three being the pillars of sustainable development and sustainable agriculture.
Economic stakeholders of this supply chain are frequently confronted with the questions of perspective at short, mid and long term. It is the same for local councils as well as national and European authorities, but also for heads of the public research sector. Investigating the possible future of such a complex supply chain requires knowing in detail the driving forces and identifying ruptures. This needs an extensive analysis of the whole supply chain, considering its structure and relationships among compartments, describing the size of every compartment, the flux of matters and information, as well as the stakeholders and their motivations. Analysis of the time series makes it possible to detect trends and ruptures. Simultaneously, the geographic scale has to be taken into account, to include the large diversity of soil and climate conditions which play a key role for grasslands and forage crops. All this is the basis of a systemic analysis (Huyghe et al., 2005).
The present paper was built through such an analysis for the supply chain based upon grasslands and forage production. Indeed, grasslands and forage crops are not per se a supply chain. Under French conditions, and with the exception of the Alpine stratum, they only exist to feed herbivores and are there thanks to the presence of those herbivores.
At this very early stage, it is important to mention that, in France, the forage produced from grasslands and forage crops has no market value per se, as there is only a very limited trade of forage, mainly hay for horses or for pellet production through dehydration. Thus, their value is mainly determined by commercialization (valorization) through livestock, and so depends on animal management.
Changes in grasslands and forage crops, both regarding acreage or agronomic practices will fully depend on the size of the animal herds to be fed, of the characteristics and performance of these animals (genetic merit), but also on the other possible sources of feed and grains and concentrates for diet supplementation. The number of animals itself as well as their management will vary, often in relationship with the volume of animal products required to meet industry demands and consumers’ expectations, both in terms of quantity and quality. Industry and distribution will sometimes consolidate these expectations through specifications that may include grassland management and diet composition.
It is then clear that grasslands and forage crops may be analyzed through the prism of a long supply chain, which may be extended to consumption of animal products, and will be influenced by the numerous acting stakeholders.
Grasslands and forage crops are primarily an essential feed resource for herbivores, which contribute most of their daily diets. But it is an intermediate consumption which is easily substituted by other resources which would better meet the expectations of the farmers, considering their prices, the ease of use or their availability. As such, grasslands and forage crops are fully in a production logic.
Since 1979, a significant reduction in the number of dairy cows has been recorded (Figure 12). From a population of more than 7 million head, the present number of dairy cows is below 4 million (FAOSTAT figure for 2010 was 3.73 million). This is due to several factors, the main two being:
Dairy cows are now mainly located in the west of France (Brittany, Pays de Loire and Normandy) where the largest dairy industries are (Figure 15). In these regions, feeding is mainly based on annual forage crops (for silage) and grazed temporary grasslands. Permanent grasslands, when present, are mainly used by young stock, especially heifers. There are also dairy herds in the east of France (Lorraine, Jura and Franche-Comté, the Alps, Massif Central). In all these regions except Lorraine, milk is mainly processed for production of PDO (Protected Designation of Origin) cheeses, such as Comté (in Jura and Franche-Comté), Beaufort (in the Alps), and Cantal (in Massif Central). In the regions, the breeds are more diverse than in the west. Montbéliarde is very present in the East of France, while local breeds Abondance and Tarentaise are found in the Alps and Salers in Massif Central.
The genetic value of livestock breeds is under constant survey and it is possible to analyze the gain in genetic value. Dissemination of the genetic progress is ensured by artificial insemination, which is now the most common practice in dairy herds. In beef cattle, dissemination of the genetic gain is mainly made by purchase of bulls.
The genetic improvement of milk production has been regular and very strong for most breeds. Genetic gains depend on dissemination of the best bulls and on financial investment. Thus, this very much depends on the herd size of a given breed. As illustrated in Figure 14 for France, the genetic gain, expressed here in kg milk/year has been much higher for Friesian-Holstein than for Brune, a breed with a low number of cows in France.
Breeding indices may vary among countries. They all combine milk production, milk quality, animal fertility and adaptation. In Europe, most breeding schemes do not take into account the ability of the animal breeds to utilize various sources of feed, and especially grasslands through grazing. However, several studies from Ireland compared breeds for their ability to do so (Prendiville et al., 2007). An economic assessment showed that dual-purpose breeds, such as Montbeliarde, performed better from an economic point of view than highly specialized breeds (Friesian Holstein) under most economic scenarios (Evans et al., 2004). The Friesian Holstein presented a major weakness in its low fertility.
Suckling cows and other cattle
The number of suckling cows has regularly increased and it now exceeds the number of dairy cows (Figure 15). As a consequence, the number of other cattle, i.e. young animals and beef cattle has first declined and has then been stable over the last two decades.
The main breeds of suckling cows are Charolaise (a breed originating from the north-east of Massif Central), Limousine (originating from the north-west of Massif Central) and Blonde d’Aquitaine, which is a breed created through crossing in the late 1960s, in order to maximise meat production. The genetic merit of the animals is regularly measured, through their liveweight gain and their slaughter weight. A regular increase is reported for the weight at slaughtering.
The feed of the suckling cows and their followers is mainly based upon grazing during the season (from early March till November). They make good use of the permanent grasslands which are abundant in the regions where these animals are mainly kept.
Sheep and goats
The number of sheep is regularly declining after a peak in early 1980. It is now close to 8 million head (the 2010 FAOSTAT figure was 7.98 million), but it is necessary to distinguish between dairy ewes, whose number is fairly stable and herds for meat production which are declining.
The dairy herd is located in a few regions, mainly Aveyron, Pays Basque and Corsica. The milk is processed to produce PDO cheeses which are very famous and much exported. It is worth mentioning Roquefort (for Aveyron), Ossau Iraty (for Pays Basque) and Brocciu (for Corsica).
The animals for milk and meat production graze temporary and permanent grasslands. For milk ewes, alfalfa hay is a major part of the diet during winter. In the case of meat sheep, it is worth mentioning transhumance, i.e. the seasonal migration of livestock to summer pastures. Even though the number of animals is limited, transhumance enables use to be made of a peculiar feed resource and contributes to the preservation of these fragile environments and their biodiversity.Goats are only kept for milk production. The number of goats was lowest in 1994 and is now regularly increasing to reach slightly more than 1.3 million head (the 2010 FAOSTAT figure was 1.35 million). The main breeds are Alpine Chamoisée and Saanen. Two thirds of the goats are in Poitou-Charentes and in neighbouring départements. The peculiarity of goat production is that most of the animals are fed indoors all year long and are fed with either farm-produced forage (mainly alfalfa hay) or purchased feed (dehydrated forage, cereals, protein concentrate). This was documented in detail by a survey run in Poitou-Charentes in 2003 by BTPL (Bureau technique pour la Production Laitière) which analysed in detail a large set of goat farms. On average, they had 234 goats and produced 193 000 litres of milk. The mean arable area was 100 ha and 20% of the acreage was used to produce feed for the goat herd. Alfalfa, distributed as hay, contributed 59% of the acreage of forage crops and grasslands, for a mean acreage of 11.7 ha (Figure 16). This very large contribution of alfalfa to goat (and sheep) diets is relevant to the geographic distribution of artificial grasslands in France.
The horse population in France has increased rapidly in recent years, but the total number of animals is still slightly below 1 million head (the 2010 FAOSTAT figure was lower, at 453 000). Horses for meat or for traction have disappeared. However, horses for recreation and sport are more and more numerous. They accounted for a total population of 900 000 head in France, which is comparable to 1 million head in Germany in 2010 and 950 000 head in Great Britain (Table 2 ).
This livestock industry presents two peculiarities when compared to the other herbivores. The first peculiarity is that only some live on a farm (Figure 17). Most operations have only one or a few animals. This means that owners are not involved in the same farmers groups and thus may not have access to the same level of information regarding grassland management. Moreover, the feed quality for horses requires special attention regarding its hygienic quality. Indeed, horses are very susceptible to dust and to the presence of mycotoxins, with possibilities of a disease named Chronic Obstructive Pulmonary Disease, which severely reduces the activity of animals. It is characterized by variable clinical signs ranging from exercise intolerance to mucus secretion or chronic cough, to expiratory dyspnea (Lowell, 1990; Mair and Derksen, 2000). Such symptoms are frequently met when hay is fed to horses, and it is especially severe when some grass species such as velvet grass, Holcus lanata, are present in swards. An adequate answer was given by the production of round-baled silage which is well-eaten by horses (Spordnly and Nilsdotter-Linde, 2011).
The second peculiarity is that the horses are not in the same regions as the main ruminant herds. Thus, they exploit pastures with a different botanical and chemical composition. Horse-grazed pastures may also be in regions specialized in grain production and these paddocks will then play a key role for biodiversity preservation in these landscapes. In France (Figure 18 ), a lot of horses are in Normandy, where there has long been a tradition of raising horses for the army in the past (this is the origin of the National Stud - Haras nationaux) and presently for racing and leisure. But there are large horse populations in the South-West where ruminant numbers, especially cattle, have dramatically dropped.
Production of meat and milk
As shown in Figure 19, the production of beef meat decreased strongly before 1990. This was due to the great reduction in the number of dairy cows, leading to the lower number of followers. It was thereafter stable, as a consequence of the balance between dairy and suckling cows. However, this stable trend was strongly impacted by the BSE crisis in 2000. During recent years, meat production has increased again. A significant portion of young cattle is not slaughtered in France. Indeed, there is a strong increase in the number of young animals (over 160 kg) exported to northern Italy, where they are fattened in specialized operations and slaughtered. On average over the last decade, 350 000 young animals were exported, 66% of them being exported to Italy.
The situation is different for sheep meat production and goat meat production, the volume being very small for the latter (Figure 20). For sheep meat, production constantly declined since 1984, accordingly with herd numbers. Self-sufficiency (ratio of internal production vs. internal consumption) of the French market for sheep meat is low. It averaged 51% over the last ten years (Office de l’Elevage, SCEES data).
Figure 21 illustrates the total quantity of milk and cow milk collected by the industry since 1981. Cow milk represents the main part of the milk collected by the industry, even if the contribution of other milks (ewe and goat) tends to increase.
There has been a severe decrease since 1985 due to the milk quota. The French reference was set at 23.8 million tons of cow milk and this was reached in 1990. The production remained very stable and slightly below this quota till 2011, when, after a shortage of dairy products, a large boost in production was recorded.
On the other hand, sheep and goat milks showed a constant increase over the same period with more than a two-fold increase in the quantity of goat milk collected by the industry over the three decades (Figure 22). This increase was achieved with fairly constant numbers of animals, showing the tremendous increase in animal productivity.
It is important to notice that the grasslands give a very positive image to the dairy products. This may be seen through two aspects. The first one is the fact that most of the production systems leading to PDO (Protected Designation of Origin)and IGP (Protected Geographical Indication) products include that the animal nutrition must be mainly based upon farm-produced forage and include mainly, or only, perennial swards, i.e. temporary and permanent grasslands. This is especially the case for Comté cheese, which is the largest French PDO of dairy products, where silage has been banned in the last version of the specifications.
The second aspect showing the importance of grasslands for the consumers is that grasslands are widely used for the iconography and labeling of dairy products. Three components are very often gathered: a cow grazing a green sward with a human presence (either a person or a house in the background). This structure of the images has been very common over the last 50 years (Figure 23). It is interesting to notice that, in recent years, some of these pictures have changed slightly to include an extra component, biodiversity, usually figured through flowers (Figure 24).
This iconography and the presence of grasslands and landscape have been extensively studied in the case of Saint-Nectaire, a famous cheese from Massif Central. A considerable body of research in semiology performed with all stakeholders has been carried out by Yves Michelin (2008) (Research Unit Metafort in Clermont-Ferrand). This study clearly identified the links between the object (such as grasslands), the value given by the stakeholders and the image of the object (which is on the packaging) (Figure 25). On all these labels, cows are looking to the left. In communication, looking to the left is a synonym of tradition. So, the association of cows and grasslands on these labels clearly means that grassland-based products are the traditional ones, with all the expected qualitative attributes.
When starting this section on grasslands and forage crops, it is essential to clearly define grasslands, forage crops and the various sources of fodder.
The word “Forage” (Fourrage in French) comes from an old German word from the 15th century, ‘Födar’ which meant straw, the straw being distributed to animals in stables as feed or litter. The word Forage can thus be defined as the raw material which is harvested from whole plants and distributed to animals as the basis of their feed diet.
In order to analyze production and to better identify the area of grasslands and forage crops, statistical databases from the French Ministry of Agriculture were used. There are five main categories: 1) roots and tubers, 2) annual forage crops, 3) artificial grasslands, 4) temporary grasslands, and 5) permanent grasslands. This nomenclature and the recording procedures have been constants since the middle of the 20th century making it possible to draw long time series.
The category “Roots and Tubers” includes fodder beets (Beta vulgaris L.) which gather all beets used to feed animals, fodder cabbages (Brassica oleracea L. var. viridis), forage turnips (Brassica rapa L. var. rapa) and Jerusalem artichokes (Helianthus tuberosus L.).
“Annual forages” mainly include fodder maize which is primarily harvested for silage. A minor proportion of it is distributed as green fodder or dried (dehydrated). When maize grains are ensiled, for instance to feed pigs, maize is then considered as a cereal crop. Crimson clover (Trifolium incarnatum L.) and ryegrass (Lolium perenne L. and Lolium multiflorum Lam.) are also taken into account as forage crops when the crop duration is less than one year, or as intermediate crops. When ryegrass swards are present over at least one whole year, they are considered as temporary grasslands.
Non-permanent grasslands include both temporary grasslands and so-called artificial grasslands.
“Artificial grasslands” are swards only including perennial forage legumes and grown during one to five years. The terminology ‘Artificial grasslands’ (prairies artificielles in French) first appeared in the literature in the middle of the 18th century. The commonest perennial forage legumes are alfalfa (or lucerne) (Medicago sativa L.), red clover (Trifolium pratense L.), sainfoin (Onobrychis viciaefolia Scop.), black medic (Medicago lupulina L.), birdsfoot trefoil (Lotus corniculatus L.). But other Trifolium, such as Trifolium michelianum are minor fodders.
“Temporary grasslands” are sown grasslands, grown during one to five years, and sown with grass species alone or in mixture with legumes. When sown grasslands are more than six years old, they are then considered as permanent grasslands.
“Permanent grasslands” mainly include native perennial forage species, but often also host many annual native species. They include sown grasslands which are more than six years old and natural, non-sown grasslands. Among the natural grasslands, we can separate the productive ones, yielding more than 1 500 Forage Units (i.e. a forage yield which is likely to cover the feed requirements of one Livestock Unit during six months) and the low productive ones such as alpine pastures. This latter category produces less than 1 500 Forage Units and often shows the presence of woody vegetation.
However, behind this unique description of permanent grasslands, there are different interpretations according to the points of view of farmers, agronomists, ecologists or administrative services. Indeed, a permanent grassland may be described through various elements, which will be given different weights by the various listed stakeholders: age, use, ecosystem services, which include production, structural characteristics, vegetation composition. This will be developed later. The various categories of temporary and permanent grasslands are precisely listed in Table 3.
Table 4 presents changes of area of the various types of grasslands and forage crops from 1960 to 2010 in France.
The analysis of data shows some key points of the changes of the forage systems over the last three decades.
Even if the areas have changed over this period and if the share among the various categories has been modified, the areas of grasslands and forage crops are very important in absolute terms and are a key component of the French agricultural landscapes in most regions. It presently contributes more than 45% of the arable lands. Indeed, when not considering the non-productive permanent grasslands, the grasslands and forage crops occupy more than 10.5 million ha. The total of arable lands in France amounts to a total of 18.3 million ha, while the whole of the SAU (“surface agricole utile“) of the farms reaches 28.1 million ha. It clearly underlines the importance of grasslands and forage crops in France.
Roots and tubers
The decline in acreage of forage roots and tubers has been regular over the last three decades. It is very pronounced for fodder beets, even though an increasing interest has been expressed by farmers over recent years, thanks to the very large increase in the genetic merit of the varieties. The last three decades are actually the end of a long decrease that started at the end of the Second World War. In 1960, there were still more than 750 000 ha. The breeding of the monogerm varieties in fodder beet did not stop this decreasing acreage. This is mainly explained by the difficulty in harvesting and handling the roots and the work load related to the distribution of this feed to the animals, and this despite the very high feeding value of this forage resource (energy value, sanitary quality).
Annual forage crops
Among all annual forage crops, silage maize is the main crop. The increase of silage maize acreage was very strong before 1980 and the acreage seems to have plateaued since the early 1980s. Indeed, the area in silage maize in 1994 was back to the value of 1984 and has not changed a lot since that time, despite some variation among years. However, it is important to be cautious when discussing these data, as they are national values and they hide large differences among regions with significant increases in some regions, such as Brittany and Normandy (at the expense of permanent grasslands). The accuracy of the value is debatable as the acreage in maize may have a dual purpose, being harvested either for grain production or for silage production. This possibility is due to the fact that, till the early 1990s, there was no variety specially dedicated to silage production. Since that time, more and more breeding programs are especially devoted to silage maize and moreover, there are special criteria for variety registration, taking into account total biomass production and feeding value through measurement of the biochemical composition. The national list for silage maize was opened in 1986. This dual purpose was and is of significance for the farmers as they can adjust the area harvested as silage to the actual need of their animal herds, the rest being harvested as grains. It is also important to consider that silage maize was the only type of grassland and forage crop to get direct support as for all the cereal, oil and protein crops before the creation of the Single Farm Payment. However, it is interesting to notice that the area of maize did not increase a lot after the CAP (Common Agricultural Policy) reform in 1992 and did not decrease after the change of the support. This may suggest that they reached the maximum acreage in relation with the corresponding animal production (type of animals, animal performance), the soil and climate constraints of the regions and the choice of the farmers. It is also important to take into consideration that dairy cattle are the main user of silage maize and that the size of this herd is limited through the milk quota system, the number of dairy cows decreasing while the mean cow performance is regularly increasing.
It is worth mentioning that in recent years, farmers, advisors and researchers have explored new species to produce forage stocks. Two may be reported here as having some potential.
Aftificial grasslands and pure forage legumes
When considering data from the last four decades, the most striking feature is the constant decline of pure legumes, which are mainly alfalfa. Moreover, 25% of the alfalfa acreage is used for the crop drying (dehydration) industry (Figure 27).
In fact, this situation is the continuation of a long history. And it is important to measure the changes over a much longer period, which is illustrated in Figure 28.
The area of pure forage legume stands started to increase in the mid 1700s, when agronomists realized that their presence in cereal rotations was a very effective way of increasing grain production, thanks to a better nitrogen status of the grain crops. Sainfoin was the first forage legume to be grown, introduced from Switzerland to Burgundy. And very rapidly, red clover and Lucerne were also used. The cultivation of alfalfa was made possible by the discovery of the first Flemish types, obtained through hybridization of two types of materials. One of them was the alfalfa populations initially introduced from the Arabian peninsula to the warm regions of the Mediterranean basin. They had a good yield potential but showed a low persistency because of a poor frost resistance and winter survival due to their low fall dormancy. The second group is composed by wild populations of the sub-species Medicago sativa ssp falcata, whose plants are prostrated, yellow-flowered, with a very good frost resistance due to its fall dormancy. The hybridization generated populations which were very well adapted to the French and European climate, offering the possibility to produce large quantities of forage. This forage, with good quality, was particularly of interest for the armies to feed horses…. This encouraged the production in fertile regions and had strong impacts on the farming systems, especially leading to the disappearance of an old system, named ‘vaine pâture’, where all ruminants were allowed to graze on all fields once cereals crops had been harvested. The presence of forage legumes in those rotations induced a large increase in soil fertility and especially nitrogen fertility, making it possible to achieve much higher grain yields. It is also interesting to notice that the availability of more forages in these farms induced the development of the sheep production. From this period, a sheep breed well adapted to these conditions was selected. It is the breed Ile de France. It was obtained by crossing Merino type, for production of wool, and English breeds (Dishley or Leicester) for animal conformation and meat production. When using Merino, the farmers introduced the genetic ability to breed almost year-round. This makes it possible to get lambs in September, while the normal season is in spring (J. Bouix, pers. comm.). This ability made it possible to make use of the stubbles of grain crops and also the forage legume crops, where it is possible to obtain several cuts of hay.
The expansion of forage legumes was reported in 1788 by Gilbert, who wrote a precise report on legume cultivation in Bassin Parisien for the Royal Agricultural Academy, where he described the presence of alfalfa, sainfoin and red clover, depending on the soil characteristics.
Statistical data from 1887 indicated that the acreage of forage legumes cultivated for hay reached 1.6, 2.8 and 3.1 million ha. in 1842, 1862 and 1882 respectively.
Until 1946, forage legumes and alfalfa acreages remained fairly constant. Alfalfa reached about 1.2 million ha., red clover 500 000 ha. and sainfoin 600 000 ha., amounting to a total of 2.3 million ha. (Figure 29). This large area of forage legumes grown for hay-making already existed in the previous century. Forage legumes had a key role in the sustainability of the farming systems both as a source of quality hay, rich in proteins, and as a source of nitrogen either through their nitrogen residues when ploughed or through manures.
After the Second World War, the whole of agriculture was mobilized to reinforce the production of foodstuffs that were dramatically lacking in Europe. Alfalfa, as a source of forage, increased to support animal production and especially milk production. The maximum acreage was achieved in 1961.
Right after this maximum, a sudden decline started. The reason for this decline resided in the deep changes experienced in France with the quick expansion of industry and the subsequent labour requirement. With agriculture and rural areas being the main possible labour sources, a strong ‘exodus’ started, leading to a strong reduction of most activities that were very labour-demanding, such as hay making. This is the exact time when the dehydration industry started, which was an industrial answer to the labour constraint. The development of this industry was supported by a special Common Agricultural Policy for dried forages. In France, Spain and Italy it was mainly used for alfalfa dehydration, while in northern Europe this was mainly dedicated to grass dehydration (Figure 30 ).
It was also the time when the use of nitrogen fertilizers became widespread. From that time, farmers understood that N-fertilizers were much easier to use as a source of nitrogen than legumes, as their effect was immediate. The cost of these fertilizers was also low despite the high quantity of fossil energy they require for their synthesis. Farmers used them more and more and excess of fertilizer use was not rare. After the first oil crisis in 1973, the use of N-fertilizers was improved and farm advisers insisted on a sustainable use of all kinds of nitrogen sources including organic manures and legumes.
The rate of legume decline slowed down after 1973. This decline occurred both in the regions specialized in dehydration and where alfalfa was used on-farm as hay. There were several reasons:
The mid-1970s were marked by a strong development of forage silage. In the early days, the machines had limited power and the capability of fine chopping was reduced. There was no practice of pre-drying. In those conditions, maize was well adapted because of the high grain proportion. Most perennial forage grasses were satisfactorily adapted because of their high content in water soluble carbohydrates (WSC), while alfalfa is lacking this compound. The consequence is that there is no quick acidification of the ensiled material and thus a poor microbial stability of the forage stocks. This situation is presently changing, with a stabilization of the acreage and this is due to a better understanding of the agronomic services due to legumes in a rotation and to technical innovations, such as pre-drying and round-bales of silage. In this case, at higher dry matter content, the need for acidification is not so important and the chemical composition of alfalfa is compatible with this limited acidification. An increasing proportion of alfalfa is grown as mixtures with grasses, mainly cocksfoot and tall fescue (these areas were not considered in the statistical survey of artificial grasslands). However, in the case of mixtures, the WSC of the companion grass may cover the need for bacterial fermentation and the consequent acidification.
Thus, the alfalfa acreage was influenced by various forces: biomass production, consumption, labour requirements and availability of other feeds, some of them being far away from the alfalfa crop itself. As will be seen later in this profile, some forces are still to be considered today.
As can be seen in Table 4, the decrease of temporary grasslands between 1979 and 1989 was stopped at the end of the 1990s. A significant increase is noticed over the last decade, with about 500 000 ha of additional temporary grasslands.
Among the temporary grasslands, the share between pure grasses and mixtures between grasses and legumes is difficult to estimate. Indeed, the statistical survey was based upon farmers’ claims. One of the last national estimates, given in 2005, was 58%. According to rural advisors in Brittany and Pays de Loire, in 2010, more than two thirds of the temporary grasslands were sown mixtures between grasses and legumes, mainly white clover. The general trend of mixtures sowings can be approached through the markets of forage legume seeds (see below). However, these statistical data (acreage, seeds) do not provide any information on the actual amount of legumes in the harvested biomass and on the farming practices, especially nitrogen fertilization, applied to these temporary grasslands.
Temporary grasslands are used either for grazing or for producing forage stocks. In the case of grazing, the main species are grasses with a high persistency, such as perennial ryegrass (in the West) and tall fescue or cocksfoot in the driest regions. For the production of stocks, the farmers tend to reduce the duration of the grasslands and use species with low persistency. In such a case, mixtures of hybrid ryegrass and clovers are more commonly met (Figure 31).
Temporary and artificial grasslands may contribute to a low percentage of the total acreage of forage crops and grasslands. It is especially the case in mountainous or semi-mountainous regions such as Lozère (Figure 32). But, in such environments where there are large climatic variations between years, the sustainability relies upon the possibilities of building stocks of forage for the winter periods and for the summer months, i.e. when there is no more feed because of summer drought. As a consequence, in such environments, temporary and artificial grasslands are essential for an optimized use of the permanent grasslands, even those with a low productivity.
Permanent grasslands (Surfaces Toujours en Herbe (STH), in French), over the last forty years, contributed 62 to 69% of the total forage crops and grasslands acreage (Figure 33). The non-productive permanent grasslands (less than 1500 FU per ha. and per year) are significant and very stable. However, in absolute values, the permanent grasslands lost 2 million ha. over the last three decades. This is a very important area and this decrease was not homogeneous over the whole territory.
Half of the 2 million ha. of permanent grasslands lost over the last three decades have returned to bushes and forests. This mainly occurred in mountainous regions or in the periphery of the Massif Central. A significant part was also lost in the mixed animal – crop farms. In these cases, mainly located in Brittany, Normandy, Pays de Loire, Poitou-Charentes and Lorraine, the permanent grasslands were turned into arable lands and grain crop production. In coastal regions, the wetlands are very unique ecosystems where permanent grasslands play a key role for preservation of these environments. The areas of permanent grasslands also varied significantly and a lot of initiatives were taken to make better use of these environments and the role of grasslands therein. A case study of the Atlantic marshlands is presented later.
The general trends that were observed on the acreage of grasslands and forage crops are due to the conjunction of three main factors:
These various elements on the share of lands among the main categories of forage crops and grasslands underline the importance of the ‘natural’ resources, if we compare this supply chain to most of the other ones. This situation has consequences in two main domains:
Table 5 presents the distribution on the national territory of the various types of forage crops and grasslands in 2002. The values are presented in ha.
The share among the various types of forage crops and grasslands varies among the French administrative regions. They are illustrated by the various maps presented below.
Artificial grasslands, i.e. pure legumes, are located in Champagne-Ardennes (for the dehydration industry), in Midi-Pyrénées and Poitou-Charentes. For these last two administrative regions, the importance of pure legume grasslands, mainly alfalfa, is associated with the types of ruminants that are using this feed resource. Indeed, these two regions host most of the sheep and goats, alfalfa being very well adapted to meet the nutritional requirements of these two types of ruminants.
In the four western regions (Haute and Basse-Normandy, Brittany and Pays de Loire) the acreage of annual forage crops and temporary grasslands is such that these four regions gather 60% of the acreage of annual forage crops (mainly silage maize) and 41% of the temporary grasslands. Among these four administrative regions, Basse Normandy and to a lesser extent Pays de Loire also host large areas of permanent grasslands, even though this category quickly declines in Basse Normandy. Permanent grasslands are predominantly present in marginal and mountainous regions (Burgundy, Limousin, Auvergne, Midi-Pyrénées and Rhone-Alpes) and also in Lorraine. Eventually, the non-productive permanent grasslands are present in Mediterranean regions and in the south of the Massif Central.
Non-productive permanent grasslands such as wetlands, rangelands and alpine grasslands are of course unevenly distributed among regions (Figure 34). On this figure can be seen the share of non-productive grasslands. Their percentage presence is low in the regions of intensive animal production of the North West of France as well as in the regions with a high proportion of productive permanent grasslands of the north of Massif Central (Limousin and Auvergne). In the administrative regions of Midi-Pyrénées (Pyrenees) and Rhone-Alpes (Alps) they contribute 25% of the total area of grasslands and forage crops. This proportion reaches 70 to 85% in the Mediterranean regions (PACA and Languedoc-Roussillon) and in Corsica. This is due to the environmental characteristics of these regions. As such, they will benefit from agri-environmental measures dedicated to these types of ecosystems. However, it is very important to stress the key role of the limited area of temporary or artificial grasslands in those regions. Indeed, these sown plots will be essential for the sustainability of the production systems, through the production of stocks and as such will make it possible to use the non-productive grasslands well, and thus ensure the other ecosystem services generated by the grasslands, which, for instance, very often show a very large and rich plant, insect and animal biodiversity.
Beyond the presentation of the geographic distribution of the various types of grasslands, Figure 35 shows a synthesis by administrative region. It shows a division according to a diagonal line from Bordeaux region to Ardennes. To the north of this line, the share of annual forages and sown grasslands is very important, while to the south there is a predominance of permanent grasslands. This division will also be visible for animal production, with dairy cattle in the North and beef cattle as well the milk and meat sheep in the south.
The variations of acreage of the various types of forage crops and grasslands show very large geographic heterogeneities. If roots and tubers are regressing in all regions, annual forage crops increased in regions where there is an important dairy production. This was studied in detail in the Seine Basin for silage maize. Figure 36 shows that the production of silage maize expanded outside the Seine Basin, especially the North West (the Normandy region). This is the expression of the specialisation of these territories, with rotations of annual grain crops in the middle and animal production in the outside regions.
The strong decline of artificial grassland is very pronounced in the regions where it was a very minor crop. It stays significant in Poitou-Charentes and Midi-Pyrénées. It has decreased in Champagne-Ardennes where it grew with the development of the dehydration industries. But, due to the reform of the Common Organization of the Agricultural Markets (CMO) dedicated to dried forages, the area progressively concentrated to a very limited region.
This has been analyzed in detail by Schott et al. (2009), with special focus on the Seine basin (Figure 37). As shown, it can be seen that the reduction in alfalfa acreage led to a concentration of this production in a small area where the acreage exceeds 5% of the total agricultural area, while it disappeared from the rest of the Basin. This raises a major question regarding the agronomic and environmental benefits which could be generated by the presence of alfalfa in these production systems mainly targeting grain production.
The geographic distribution of forage production led to a typology of the metropolitan agricultural regions. Various illustrations have been proposed.
Rouquette and Pfimlin (1995) proposed to identify 700 natural regions. An illustration of this typology is given in Figure 38. And it has been later applied at the European scale by Pflimlin (2010). When not taking into account non-forage production, Emile (1996) proposed to mainly consider 4 categories for forage crops and grasslands: zones of forage crops, herbage regions, wet mountains, rangelands. In these various categories, the importance and role of these various types of feed resources vary as a function of the soil and climate conditions, and on the need to produce feed stocks for winter and summer periods, when there is little possibility for grazing. It can easily be concluded that this will be in strong interaction with the type of animal production and structure of the holdings.
On the basis of a local analysis of crops (including forage crops), temporary and permanent grasslands and other land uses, including forestry, Véron and Brunet (2003) proposed a landscape typology. The main land uses led to the identification of 14 categories (Figure 39). This typology resulted in an innovative structure of the French territory (Figure 40). Mountainous regions appear in categories A2 and A3, with a few basins in A1 and others, with a high density of forests in C5 and C4. In regions with a high proportion of permanent grasslands, A1 type is predominant. On the opposite, in Brittany, Pays de Loire and Nord-Pas de Calais, with an intensive cattle production, the proportion of crops, including silage maize is high. It is in these regions that the decline of permanent grasslands is presently the most pronounced and this strongly affects their landscape identity.
There is very little data available on the agronomic practices applied by farmers to temporary and permanent grasslands. It is possible to use case studies. A national survey was performed in 1998, on the initiative of SCEES (Service Central des Enquêtes et Etudes Statistiques.) for all grasslands to better know the actual practices. This survey was part of a large project aimed at estimating in real time the production of grasslands and thus, estimating the possible deficit in comparison to the long-term average. It does not make it possible to extrapolate the production to an individual farm as the local variation in soil characteristics may differ, as well as the agronomic practices.
Two points were particularly investigated, as they strongly influence the production, i.e. number of harvests (cutting or grazing) and nitrogen fertilization, either organic or mineral.
Number of harvests (cuts or grazing)
Figures 41 and 42 show that the mean number of harvests (cuts or grazing) on permanent grasslands is low, being below or equal to three in most regions, except in northern Brittany, the piedmont of Pyrenees and in the Ardennes, where up to six harvests were recorded..
For temporary grasslands, the number of harvests is higher, exceeding six in wet regions with a climate favourable to growth of grassland species. It is especially true in Brittany and Normandy. It is in these regions, that the differences between permanent and temporary grasslands are the largest.
Mean nitrogen fertilization rates on permanent and temporary grasslands are presented in Figures 43 and 44. On permanent grasslands, there were contrasting situations between northern and western regions where the mean annual fertilizations were estimated between 60 and 100 kg N/ha and the rest of the country where they were estimated to be below 60 kg N/ha. The Piedmont of Pyrenees was very different with much higher rates, which were relevant with the number of harvests.
On temporary grasslands, fertilization rates were higher, as expected in a more intensified production regime. The highest rates, aggregated at the level of a homogeneous region, were met in Normandy, with more than 100 kg N/ha. On the opposite, it is in the administrative region Centre, that the lowest fertilization rates were met.
Altogether, these fertilization rates are relevant with the production potentials linked to the climate, and especially the summer drought.
On the basis of the agronomic practices presented above, using the growth model STICS (Simulateur mulTIdisciplinaire pour les Cultures Standard [Brisson N. et al., 1998]) parameterized for grasslands and long term databases for climatic data, a model simulating the dry matter production was established. This model, named ISOP (Informations et Suivi Objectif des Prairies), makes it possible to know, the dry matter production all year around. Based upon the mean agronomic practices and long term weather record from 1982 to 1996, the mean dry matter potentials were calculated for every region.
The mean dry matter yield (Figures 45 and 46) shows high yield potential in mountainous regions, in the Piedmont of Pyrenees and in the North and West of France. In all these regions, the high summer rainfalls explain the high potential. For each region, the mean dry matter yield of temporary grasslands exceeded the mean dry matter yield of the permanent grasslands. This may be explained in the modelling approach by the differences in the levels of intensification (number of cuts and mean fertilization rate).
Estimates of dry matter yield are produced every decade and the data are available on the Website of the Ministry of Agriculture (http://agreste.agriculture.gouv.fr/conjoncture/grandes-cultures-et-fourrages/prairies/). The estimates are given in percentage of the long term average calculated on the 15 years period of reference. These estimates do not take into account the changes in management practices and especially nitrogen fertilization or irrigation.
To illustrate this feature, Figure 47 presents the case of estimated cumulated dry matter yield for permanent grasslands in 2011, a year that was marked by a very severe water deficit in spring and early summer, especially in the west and south of France, while July and August had more rainfall than the long term average. As expected, the estimated dry matter yields showed a very marked reduction. It is also possible at the end of the season to make a final evaluation. This is illustrated again for the dry year 2011, and compared to another very dry year (2003) and 2009, a very wet year (Figure 48).
Another estimate of the production potential may be obtained from the computation of data collected from the variety registration trial network. Based upon the fact that the control varieties remain fairly constant and that the agronomic practices are very standard between years and locations, it is possible to draw general information on the yield potential of the various species between French sites and years (Figure 49). This graph shows the difference between years, with lower production in dry years and differences between species. The species with a low cutting frequency (alfalfa, tall fescue and cocksfoot) yielded more than perennial ryegrass which was harvested more often.
A long term survey of dry matter production was carried out on alfalfa in Champagne-Ardennes, i.e. the region of dehydration factories. In spite of a large variation between years, there was a trend towards a slow increase in dry matter yield (Figure 50). This may be explained by a slow genetic progress, especially on disease resistance. A better crop management has been implemented, where there has been a tendency to have more cuts per year in order to get higher protein content. This management tends to slow down the increase in dry matter yield. Among the changes in crop management, two items may be reported:
It could also be suggested that this is partly explained by the effect of climate change.
Estimation of protein production
To give an overall view of forage production, it is proposed to estimate the protein production from forage, and precisely from the various types of forage crops and grasslands. This estimate was made using the acreage of 2010 from Agreste and estimations of protein content in the various roughages proposed by INRA. Of course, it is necessary to be extremely cautious when discussing these data, because they are general data, based upon mean yields and mean protein contents. It does not take into account any dry matter loss at harvest or when distributing feed to animals.
The total quantity of protein produced from forage crops and grasslands is very large. It may be estimated at 8.6 Mt of proteins for a total dry matter yield of 95 Mt (Table 6, Figure 52). This amount of protein is equivalent to 21 t of soya bean or 17 Mt of soya bean cakes at 48% protein. The permanent grasslands contribute 48% of the available dry matter and 42% of the protein produced from forage crops and all grasslands. These values were 52 and 45 respectively, 10 years ago (in 2002), and these variations are due to the significant increase in the acreage of temporary grasslands and the stabilization of forage legumes, while the area of permanent grasslands is still decreasing.
This quantity of protein is a main resource of ruminant feed for ruminants. Indeed, it has to be compared to the quantity of proteins from protein-rich plant commodities. The quantity of these commodities used in France was estimated by UNIP (Union Nationale Interprofessionnelle des Plantes Riches en Protéines - a technical institute dedicated to annual protein crops) and the amount of protein was estimated to be a total of 3.39 Mt. In France, the self-sufficiency of protein-rich grains or cakes was 61% in 2009/2010, while it was only 33% for EU27, for a total volume of 22.6 Mt including 15.5 Mt soybean. In 2011-2012, feed concentrates used for cattle contributed 15.6% of the 20.28 Mt of feed supplements produced or used in France (SNIA-SYNCOPAC). Because of the differences in protein contents among the concentrates supplemented to the various animals, it is then possible to estimate that one third of the protein used by the feed compounding industry is for supplementation of ruminant diets.
Thus, if the grains which are directly used on farms are not considered, and especially proteins from cereals and grain legumes, it may be estimated that 85% of the proteins fed to herbivores are produced by forage crops and grasslands.
Season dynamics and variation in dry matter productioni and feeding value of permanent grasslands
A review of production and feeding value of permanent grasslands was recently made by Baumont et al. (2012).
The analysis of a network of 190 grasslands surveyed in 2009 and 2010 (Launay et al., 2011) provided a set of data for feeding value and dry matter production of a large group of permanent grasslands used in the main cattle and sheep production systems, for either meat or milk production, and over a large range of soil and climate conditions, from the Atlantic coast to higher grasslands up to 1200 m (except for the Mediterranean region and Alps). The 190 grasslands were selected from a population of 1480 grasslands studied by Farrie et al. (2011). The sampling was done according to 5 criteria: 1) management regime (grazing only, cutting only, combination of grazing and cutting), 2) management intensity (stocking rate, grazing duration, number of cuts), 3) nitrogen fertilization intensity (either mineral or organic), 4) animal production (cattle/sheep) and 5) geographic region.
A large variability
For the whole set, the mean annual dry matter production was estimated as the sum of the production at the end of spring (P2, measured at about 1 180 °C.d from 1st February), summer and autumn regrowth measured after 7 to 8 weeks of regrowth after the previous cut). The average production estimated in 2009 and 2010 was 6.2 t DM/ha. This production is comparable with the estimates from Agreste (2011) that reported 5.2 t DM/ha in 2009 and 4.9 t DM/ha in 2010. The difference may be explained by the fact than the sampling protocol implemented by Launay et al. (2011) did not include the edge effects and loss during harvest. A large variation was detected; indeed, 25% of the plots produced less than 4.2 t DM/ha and 25% more than 8.1 t DM/ha.
The spring production contributed 75% of the annual production (Figure 53). At the end of spring, the biomass available for harvest ranged between 5.5 and 6.5 t DM/ha and this was comparable with the records made in the 60s and 70s in the productive and fertilized permanent grasslands in Normandy (Pin au Haras) and Massif Central (Orcival) (Demarquilly et al., unpublished). This corresponds to a spring growth of 500 kg DM/ha per 100 °C.d. This mean value was also measured by Delaby et al. (unpublished) in early spring between 1990 and 2010 in Normandy (Pin au Haras). Summer and autumn productions were more contrasting. During the two years of survey, the regrowths were too low to justify harvest in more than 25% of the grasslands.
Changes in organic matter digestibility (OMD, estimated with pepsin-cellulase digestibility (Aufrère et al., 2007)) and protein content (%N * 6.25) during the seasons have been surveyed and were comparable with data published by INRA (2007) for permanent grasslands in plains and piedmont with in vivo measurements with sheep, in the 1960s and 1970s (Figure 54). In the set of grasslands, the mean OMD value in early spring was 77% and 64% in late spring. These values correspond to a grazing stage and late heading stage respectively in the above-mentioned tables. These values are also comparable to data provided by the Swiss tables (Daccord et al., 2006) for the G type grasslands with a high proportion of grasses. The decrease of OMD between 600 and 1000°C.d was on average 2.1 points per 100 °C.d. This value was also found by Baumont et al., 2012 in the project “Prairies AOP”. As expected the reduction of OMD was much quicker during the reproductive phase (Duru et al., 2008) than during the vegetative phase (-0.4 point of digestibility between 125 and 650°C.d measured at INRA Le Pin by Delaby (unpublished)). During the vegetative phase, there is no change in morphological composition (only leaves). The reduction in digestibility is only due to organ ageing and the increased proportion of sheath (Vignau-Loustau and Huyghe, 2008). However, as for the production data, the variation between grasslands is very large. Indeed, in early spring, 25% of the grasslands have an OMD which was higher than 80% while 25% were lower than 75%. At the end of spring, 25% were higher than 67% while 25% were lower than 62%. For the summer and autumn regrowth, the mean OMD of this network was similar to the value reported in the INRA tables, the variation being even larger than in spring.
For the protein content, the mean value in early spring was 17% (170 g/kg DM), this value being similar to the value given in the INRA tables. But for late spring, the mean value was 1% lower than the reference value in the INRA tables, but is in agreement with the Swiss data for G type grasslands. On average, the reduction in protein content was 1.2% per 100 °C.d. As for production and digestibility, the variability between grasslands was very large, especially in late spring when 25% had a protein content lower than 8%, while 25% exceeded 10.5%.
Relationships between production and feeding value differentiate the forage value of permanent grasslands
The combined analysis of dry matter production and feeding value (Figure 55) confirms the negative relationship between dry matter production and feeding value which was established on sown grasslands (Huyghe et al., 2008). Moreover, it can be seen that there is a negative relationship between dry matter production and the stability of feeding value during the spring. The most productive grasslands are those where the feeding value rapidly declined during the spring. Indeed, for the grasslands where a dry matter production of 6 t DM/ha was achieved in late spring, a drop of 15 units of digestibility was observed during the spring. In contrast, this drop was reduced to 10 units when the maximum biomass was only 4 t DM/ha. The large increase in biomass and subsequent decrease in digestibility was due to the production of a lot of stems. It clearly corresponds to a given type of grass species, which correspond to the capture type (Ansquer et al., 2009).
As a consequence, 5 main characteristics differentiated the forage value of permanent grasslands (Figure 55):
As can be seen and as expected the most productive grasslands are those with a high proportion in grasses, and as a consequence with the lowest proportion in legumes and other dicots. Among the grass species, those belonging to functional types A, B and b are commonly met in the fertile environments and under high competition for resources (Cruz et al., 2010) characterize the most productive grasslands. On the other hand, the functional type C, with species adapted to poor environments thanks to a conservation strategy, is more abundant in grasslands with low production and low quality in early spring, but a more stable quality.
High forage quality, especially in late spring was associated with a high proportion of legumes. Grasslands with the highest legume proportions had the highest OMD and protein contents. This was already clearly demonstrated for multispecies sown swards (Delaby et al., 2007, Baumont et al., 2008, Huyghe et al., 2008). The stability of the feeding value in spring depends on the proportion of other dicots and on C type grasses. The positive role of the other dicots on feeding value was found in plain grasslands where the digestibility of these leafy species changed little. The C type grasses, such as red fescue, are typical of altitude grasslands. The digestibility of these altitude grasslands dominated by red fescue was lower in early spring than in all other types of grasslands. In a trial where digestibility measurements were performed till 1500 °C.d, digestibility of C type grasslands became higher than A and b types (Michaud et al., 2011a).
Even if a convergence of traits between dicot (including legumes) and grass species in a given grassland was demonstrated (Ansquer et al., 2009, Duru et al., 2010), this national survey showed the need to consider both the proportion of legumes and other dicots along with grasses in order to better explain dry matter production and feeding value of grasslands. This was consistent with the findings of Daccord et al. (2006) for Swiss grasslands.
Based upon the vegetation of permanent grasslands, characterized both after exclosure and after use, and which depended upon the environmental conditions and the management practices, it was possible to establish a typology of permanent grasslands with 19 categories. They are defined in Figure 56.
For all these types, dry matter content, production of dry matter, digestibility and protein content were measured. The values are presented in Figures 57 to 60. Clearly some types had a fairly stable composition, both in dry matter content, protein and digestibility, while others showed a very contrasting composition. This corresponded to the most productive grasslands, which have a high production in late spring, due to heading and the production of stems. Grasslands from the wetlands of the Atlantic coast showed poor composition and a significant production in spring.
Based upon these data and the categories of the national typology for permanent grasslands, it was possible to calculate the range of production and quality for the various types, as well as the variation within types (Table 7).
Comparison of feeding value of permanent and temporary grasslands
There are few reports on the comparison between temporary grasslands and permanent grasslands in a given region.
Such a study is available for Normandy. It was performed by Houssin and Pavie (2010) in an R&D project dedicated to grasslands and forage harvests in real conditions. In order to avoid any bias due to losses during harvests, all forages were barn-dried. The number of fields and paddocks harvested varied from 2 to 10.
The data clearly showed that the feeding value of permanent grasslands was lower than the temporary grasslands sown with mixtures of grasses and legumes, both for the protein content and the energy value (Table 8). The dry matter productions were not precisely documented in this study run in farm conditions, but it was also higher for the sown temporary grasslands. Alfalfa, in these conditions showed its ability to produce high protein content and confirmed its low digestibility, especially in the first cut. This has been well-documented in research work (see review by Julier et al., 2003), and it fully underlines the relevance of including quality traits as a criterion for characterization of varieties during the process of registration.
These data are consistent with the miscellaneous data collected in other regions.
From such an analysis, it would be very dangerous and useless to conclude that there is little interest in preserving permanent grasslands. It must first be taken into consideration that they provide other ecosystem services, and especially biodiversity preservation and preservation of soil quality, especially the carbon stored in such soils.
So, the real question is the need to optimize the contribution of permanent grasslands in the animal production systems and as such the combination and complementarity of permanent grasslands, temporary grasslands and forage crops. The possible role of a permanent grassland will depend on its forage value, i.e. production and feeding value. It depends on the vegetation composition which will be related to soil characteristics, climate, location, but also the management practice. As a consequence, the choice of species for the temporary grasslands must take into account the possibility to better valorise the potential of the permanent grasslands. Doing so, it is then possible to maximize the production function and the preservation of the environment.
The management of forage crops and grasslands, and especially the harvest, is essential to make good use of the feed resource and to contribute to the economic performance of the animal production. In this section, the share of the various modes of harvest and the changes in machinery for the mechanical harvest and storage of forage are analyzed.
Grazing is the main mode of harvest of the available dry matter. In a synthesis published in 2001, and often validated since then, Le Gall et al. (2001) showed that grazing was a very effective way to reduce production costs in dairy production (Figure 61). These values are means estimated for various countries, not considering any financial support to milk production. This clearly demonstrates that for ruminants, grazed grass is the cheapest source of energy and protein among all feed resources available for ruminants. It clearly explains why grazing is more widely used in beef cattle production than in dairy production, because of the strong economic constraints on the beef production activity.
However, under French climatic conditions, the duration of the grazing period is limited, both in winter and summer. In winter, it may be impossible to graze because of the lack of feed resource to be grazed and by the soil structure. In summer, in many regions, it is not possible because of the lack of feed due to summer drought. A longer grazing period, especially with an early start, was targeted in many projects, with some success, for instance with the possibility of using winter cereals, or stockpiling in the field. On average, the grazing period is longer for suckling cows and heifers than lactating dairy cows. However, it is clear that in French conditions, as in most west-European conditions it is not possible to feed productive dairy cows all year long with grazing only.
Thus, even if grazing is an essential component, stocks are compulsory and are an important component of the sustainability and economic viability of animal production. The quantity of forage stocks and their quality have to be defined in conjunction with the whole animal production system.
According to Renaud (2002), areas used for stocks were:
Changes in harvesting machinery
The book written by Renaud reviews the long history of machinery for forage harvesting throughout the agricultural ages. Details of all the stages will not be covered here, but it is necessary to recall that the first mowing machine was created by McCormick in 1830 and improved by Hussey in 1847. The first silages were produced in Bavaria in 1861 with whole plants of maize and in Normandy in 1872 with green forages and maize. No chopper was used and the results were unsatisfactory. The first tools for haymaking pulled by animals were used in 1875. The real changes occurred when tractors were used.
The first balers, producing low density bales, were used in 1948 and the first round baler, with a fixed chamber, was used in 1974. These machines introduced a real change in forage harvesting, as a single person could now carry out the task. In 1988, the first system of foil wrapping was developed and the machines combining round-baling and wrapping entered the French market. In 1985, the balers producing high-density square bales were available.
The first prototypes of mechanical forage harvesters were demonstrated in 1950 in the USA and the first harvesters, almost identical to present machines were available in 1953. They became self-propelled in 1973.
For mowing, there are three types of conditioning equipment, which contribute either to a quicker drying and/or a better digestibility. These types of conditioners are:
For storing silage, mole or bunker silos are the main types. They are now included in the national plan for managing the pollution from agriculture (PMPOA in French). Collecting the liquids produced by silages is compulsory to reduce pollution. It means that concrete platforms are needed for an adequate control of the losses.
For barn-drying systems the first machine was created in France in 1956, while the very first systems worldwide were created in 1924 in Oxford and in 1935 in the USA. The first French system was a ventilated tower. In 1960, ITCF and CNEEMA, two technical institutes, conceived a ventilated barn drying system that could be used for bulk material and for small bales, while the first system adapted to cylindrical round bales was created in 1997. Today, in France, no national statistical data are available to precisely document the number of barn-drying systems. The number of barn-drying systems has been surveyed in the west of France (Normandy, Brittany and Pays de Loire) in the project led by Houssin and Pavie (2010) from Institut de l’Elevage. Data are presented in Figure 62. Even if the number is still low in absolute terms, it clearly shows that this type of forage drying and storing system meets farmers’ expectations.
The drying in-barn has consequences on forage quality, as it better preserves the forage and all its constituents. This has been documented by a trial in Normandy run by J.R. Peccate in 2008. As shown in Table 9, the main effect in on the preservation of the energy value of the hay and as a consequence a better valorisation of the protein.
Quantifying the sales of forage harvesting machines
The number of machines sold every year was analysed by the SECIMA (Syndicat des Entreprises de Commerce International de Matériel Agricole) and SYGMA (Syndicat Général des Constructeurs de Tracteurs et Machines Agricoles) and also by Axema (Union of Agro-equipment Industry).
The number of ensiling machines sold every year shows little variation between years (around 300 units), similarly the number of large square balers was very constant, with a mean value of 386 over the last 6 years. The number of mowing machines is more variable, but is around 2 500 units every year over the last two decades. For these three types of machines, the main trend is to see an increasing size.
The most interesting point is the analysis of the round balers and the transition from small low density bales to round-balers. These machines are essential for the forage harvest and storage and they illustrate the trend in mechanization. Figure 63 shows the very sudden transition between the two types of machine, with the disappearance of small low-density bales, which were very labour demanding. At the same time the number of round balers expanded very quickly from 1978 to 1985 on all French farms. The size of the market is now fairly stable.
The press chamber may be fixed or variable. Machines with a variable chamber presently (2011) account for 82% of the total market (it was 77% in 2003).
On the basis of the data available for 8 trademarks (out of the 13 available at that time), that contributed 81% of the total market for 2001, the share of the fixed/variable chambers in the French market is presented in Figure 64. It can easily be seen that the round balers are mainly located in the regions of grasslands and herbivore farming (this justifies the presence of this section) in the west of France, the mountainous regions (Pyrenees, Massif Central, Alps, Jura and Vosges), Lorraine and the north of France. Analysis of the types of machines also shows a very strong geographic pattern. Indeed, West and North West France mainly have machines with a variable chamber, while by contrast, mountainous regions have fixed chamber machines. The reason is the machine weight; the ones with a fixed chamber are not so heavy and are more convenient for use in difficult environments.
Many grass species may host endophytic fungi. These fungi have been studied in detail, mainly in USA, New Zealand and Australia, mainly because animals eating endophyted grasses may develop toxicosis, such as ryegrass stagger, induced by some of the alkaloids produced during the symbiotic process. This induces severe negative economic consequences, as it may lead to animal death. These countries have also developed research to identify endophytic strains which are free of the deleterious alkaloids.
In Europe and France, research devoted to endophytes is much more limited, because of the limited number of toxicosis situations. Raynal (1991) reported difficulties in the centre of France on beef cattle grazing tall fescue paddocks where more than 50% of the tillers were contaminated. Another situation was reported in Brittany during summer 1997 on dairy cows on a ryegrass paddock where 30% of the tillers were infected by endophytes (Bony et al., 1998). Eventually, some cases of toxicosis were observed on horses being fed with straw harvested from a field of seed production of an endophyted turf variety.
Fungi, belonging to the genus Neothyphodium are symbionts of perennial grasses and contribute to improving their resistance to drought, their nitrogen nutrition and their resistance to voles and insects. This is of critical importance in environments with severe biotic and abiotic constraints, such as New Zealand, Australia and Midwest USA. Most of these characteristics are related to the presence of alkaloids. There are two groups of alkaloids which may be found in cases of symbiosis between perennial grasses and Neotyphodium. Peramin and lolime are active against insects, while ergovaline and lolitrem B have a strong effect on vertebrates, including ruminants. In ruminants, the consumption of feed with endophytes producing ergovaline and lolitrem B has a strong and quick effect on blood circulation (Boivin et al., 1998) and functioning of intestines. Ergovaline reduces serum blood prolactin and increases body temperature, while severe symptoms of neuro-muscular disorders, named ryegrass staggers may be induced by lolitrem B. If recent, symptoms rapidly disappear when animals graze or eat forage without endophyte and toxic alkaloids. By contrast, in cases of prolonged exposure to high doses of alkaloids (ergovaline and lolitrem B), severe dry gangrenes on tail and legs may be seen which may lead to animal death. This is considered as a main cause of financial losses in many countries. Fortunately, the impact on French ruminant populations is very limited.
Neotyphodium endophyte may be detected in stems, epidermis and sheath of leaves as well as inflorescences and seeds. Contaminated seeds are the only way of dispersal of Neotyphodium. Thus, it is maintained at high level, when the presence of endophyte confers a strong competitive advantage, while it tends to stay at low level or to disappear in environments with little stress. In tall fescue, the endophyte species is N. coenophialum Morgan Jones 1 Gams, while it is N. uncinatum for meadow fescue, the symbiosis with this species inducing no production of ergovaline and lolitrem B, and N. lolii Latch, Christensen & Samuels for perennial ryegrass.
In the next section, elements related to the presence of these fungi in the various grass species of French permanent grasslands and the regulation related to the marketing of certified seeds, are presented.
Endophytes in French permanent grasslands
The first report of the occurrence of endophytes in grassland species in France was provided by Latch et al. (1987) who reported that among 28 samples of seeds of perennial ryegrass harvested in permanent grasslands in France between 1967 and 1974, 25 accessions hosted N. lolii endophyte for a mean infection rate of 43%. On commercial seed lots, between 1986 and 1988, when there was no regulation on the presence of endophytes in commercial seeds, G. Raynal and R. Champion (GEVES) showed that:
Among the 547 natural perennial ryegrass populations collected in French permanent grasslands and roadsides, a sub-set of 262 perennial ryegrass accessions were selected for evaluation of the occurrence of endophytes, and estimation of the contamination rate (Ravel, 1997). This study showed that 188 out of the 262 populations were contaminated with endophytes, i.e. 72% of the subset.
The distribution map (Figure 65) shows that endophyted populations may be met over the whole territory, even though some regions, such as Brittany, seem to be free of endophytes.
The map of infection rates shows that they are on average fairly low and that a geographic structure does exist (Figure 66). Indeed, infection rates are much higher in dry and hot regions, such as the Mediterranean region. Populations from regions with less biotic and abiotic stress have lower infection rates. These lower proportions in stress-free environments are consistent with the lack of competitive advantages of endophyted plants in those environments and the likely loss of endophyte during seed production. A mutation – selection model was developed which explains this polymorphic and stable state.
In 2001, Leyronas and Raynal summarized the presence of Neotyphodium endophytes in a range of grass species in France and confirmed the high rate of occurrence in natural populations all over the country. Tall fescues collected up to an altitude of 1600 m exhibited a contamination rate of more than 90%. They also found endophytes in Lolium perenne, Lolium multiflorum Lam., Lolium rigidum Gaudin, Agrostis sp., Festuca ovina L., and Molinia coerulea (L.) Moench. Naffaa (1998) identified endophytes in 17 grass species in grasslands from Puy de Dôme. Among these species, there were one Agrostis, two Bromus, one Dactylis, nine fescues, one Holcus, one Koeleria, one Melica and one Poa. This study was extended to the whole Auvergne region by Guillaumin et al. (2000). They screened 63 grass species. 49% of the samples were infected by endophytes, with on average low infection rates. The highest rates were found for the various fescues, reaching 100% for samples of meadow fescues.
Thus, even though endophytes are found in many situations, infection rates are often low. It is especially true for perennial ryegrass in permanent grasslands. These low rates as well as the total absence of endophytes in temporary grasslands (see below) explain why there are very few cases of animal toxicity.
Regulation on the presence of endophytes in commercial seeds in registered varieties in France
Because of the risks of toxicity induced by endophytes, their presence in commercial seeds was regulated. The CTPS (Comité Technique Permanent de la Sélection) decided in 2000 to prohibit the presence of endophytes in the seed lots of the varieties proposed for registration, and this means that the marketing of endophyte free seeds secures the farmers against any accident due to the presence of alkaloids in the harvested forage. However, at the international level, research was and is conducted to select strains of endophytes which are safe for animals, such as U2 strain or AR1. These strains show a different pattern of alkaloids, where there is little or no ergovaline or lolitrem B. This was for instance documented by Oliveira et al. (2003) who showed, during a three year study in Spain that when inoculating perennial ryegrass with a native endophyte strain free of lolitrem B, there was also no accumulation of ergovaline at toxic levels. More research in New Zealand and the USA confirmed this Spanish paper.
As a consequence, it is important for regulation to encourage innovation. Thus, CTPS decided in 2008 to allow varieties with endophytes for turf. For forage varieties, endophytes are prohibited, except when the breeder can demonstrate that the forage produced from such varieties is free of alkaloid and does not induce any physiological impact for ruminants.
Except for a limited market for hay, dehydrated pellets are the only types of forage feed with a real market and market price, in France and in Europe. There is a special Common Market Organization for dehydrated forages in Europe. In Northern Europe, it is mainly dehydration of grasses, while in France and in Europe, it is mainly alfalfa.
Production and consumption of dehydrated forages has experienced strong variation over the last four decades (Figure 67). From 1970 till mid-1990, production strongly increased, due to strong support for this industry producing a protein-rich feed. Consumption carried on increasing till the late 1990s. Since 1993, production has decreased due to a reduction in the area of alfalfa dedicated to this production. The changes in alfalfa acreage are due to the modification in the CMO, with less support to the industry and direct support to the farmers, the subsidy being included in the farm direct support.
Between 1990 and 2005, consumption was very stable and declined in later years. This is probably due to strong competition in the protein product for supplementation of animal diets with the cakes from oilseed rape.
In 2000, when there was a large alfalfa production, France was the second worldwide exporter with 303 025 t, while the first exporter was Canada (402 464 t). The main customer of the French industry was Benelux with 43% of the exported amounts between 1990 and 2000 (Figure 68).
The seed sector is a key component of the input part of the supply chain. This is true from the economic point of view. It is also an important component because it is a major element for dissemination of technical improvement of management practices for temporary grasslands. Knowledge from the seed industry will influence the choice of species and mixture of species as well as the choice of varieties to be sown. This section will not include silage maize and will focus on the amounts of seeds marketed in France, on the seed exchanges (import and export), the development of the seed market for species mixtures and will conclude with some regulatory aspects.
Quantity of seeds sold in France
Over the last 30 years, the trend in volumes of seeds was very stable for most grass species, even though there was strong inter-annual variation (Table 10). Perennial ryegrass was the only species to show a large increase in the volumes of seeds sold in the French market, this increase being very important between 1975 and 1985 and this is mainly due to the fact that the data combined both forage and turf types.
Indeed, in perennial ryegrass and tall fescue, the amounts did not differentiate forage and turf. They have been explicitly separated since the creation of the national turf catalogue in 1990. Before that date, the only option to differentiate for these two uses would be on the basis of packaging. On the basis of the data available since 1991, the share of turf is 46.4% for perennial ryegrass and 55% for tall fescue.
The volumes of perennial ryegrass have shown a very large decrease in 2008 and 2009. This is explained by the fact that during these two years the good weather conditions did not induce the need for large sowing of temporary grasslands. It is also due to the increase in tall fescue and cocksfoot that are more persistent than perennial ryegrass, even under grazing management, especially after a dry summer.
The variation between years was much larger for Italian ryegrass (Figure 69). This erratic curve is fully explained by the role of Italian ryegrass in the animal farms. It is only used to produce feed stocks over very short periods of time. Thus, the sowings of Italian ryegrass are decided on the basis of the feed stocks available on the farms. As such, more Italian ryegrass is sown when the stocks are low, and this is very often the case after dry years, such as 2003.
It is very interesting to notice changes in the market for forage legume seeds. After a slow decline between 1980 and 1990, which was related to the decrease in artificial grasslands (pure alfalfa), the demand has remained stable since 2000 and over the last 10 years, the volumes have increased strongly, especially over the last three years. This means that more and more legumes are sown in mixtures with grasses in temporary grasslands, a practice that the statistical data does not show. From 2007 to 2011, alfalfa showed a 50% increase. It is also worth noting that there has been a significant increase in the market for seed of fodder pea and vetch. They are grown in mixtures with cereals. Most of the crops are then harvested for the production of silage at the milky or waxy phenological stage of the companion cereal, and only a very small part of the crops are harvested for grain production.
The analysis of the registered varieties shows that the set of varieties considerably increased with a widest range of variation for traits which are essential for adaptation, such as heading date or reheading.
The set of varieties registered on the national catalogue and their characteristics are presented on a website recently developed (http://www.herbe-book.org/). This website is available for farmers and rural advisors and helps in selecting the most adapted varieties for the soil and climate conditions and for the anticipated management practices. Recent genetic gains have been achieved to improve disease resistance, dry matter production and its distribution over seasons, these characters being taken into account during variety testing before registration on the national catalogue. Feeding quality has long been taken into account for alfalfa, first through protein content and, in 2008, through NDF (Neutral Detergent Fibre) content, which is a good indicator of digestibility. For grasses, the feeding value will be taken into account from 2013 onwards. Three components will be considered: protein, water soluble carbohydrates and NDF. For all these components, the scientific literature showed that there was quantitative or qualitative (in the case of WSC) variation among varieties or accessions, for a similar level of dry matter production.
To be marketed in France, forage varieties of the list of species listed in the directive 66/401/CE must be registered either on the French or on the European common catalogues, with the associated studies for DUS (Distinctness, Uniformity, Stability) and VCUS (Value of Cultivation Use and Sustainability).
For the major species, varieties registered on the French national catalogue contribute most of the markets (Figure 70). This share in 2002 and 2011 is very high for alfalfa, white clover, bromegrass, cocksfoot, tall fescue, hybrid and perennial ryegrass. The share is much lower for Italian ryegrass, where the price of seeds is a criterion of farmers’ choice, alongside the agronomic value. The share was very low for timothy and meadow fescue, these species being minor in the French climatic conditions.
Forage seed production in France
Figure 71 illustrates the acreage devoted to seed production of forage grasses, amenity grass and forage legumes. This time-series was initiated in 1992, which is the year when it became possible to separate without ambiguity grass seed production devoted to forage and to amenity.
This figure shows that there are large variations between years. This is due to the rapid adaptation to market demands and especially to excess of material, both for internal market and exports. It also very much depends on the availability of seeds from other European countries with large seed production activity. It is particularly the case for Netherlands and Denmark, where one coop, DLF, is one of the major seed producers worldwide.
It is also clear that acreage of seed production for varieties registered for amenity use increased, as they reached and then exceeded the acreage devoted to forage-type varieties in 2000.
The mean area devoted to seed multiplication of forage and amenity species over the last 5 years in France reached 18 887 ha for forage legumes, 10 590 ha for forage grasses and 5 413 ha for amenity grasses, for a total of 34 890 ha.
The decline in the last five years is due to 1) the low demand, 2) the strong competition for land use with other commodities, and especially the cereals whose market prices were very high and 3) the difference in financial competitiveness with cereal production, this difference being larger in France than in the Netherlands or Denmark.
The acreage dedicated to forage and turf seed production reached 31 101 ha for the harvest 2011. It was distributed in 12 967 ha of forage and turf grass species, 16 205 ha of perennial forage legumes, 1 755 ha of annual forage legumes and 174 ha. of other forage crops (fodder cabbage and radish). These acreages showed a significant decrease in comparison with 2000 when 24 673 ha of perennial forage legumes and 21 800 ha of grasses (11 278 of forage varieties and 10 522 of turf varieties) were harvested. The share between the species is presented in Table 11. The main variations are due to Italian ryegrass and hybrid ryegrass among the forage varieties, perennial ryegrass and red fescue among the turf varieties and alfalfa and red clover among the legumes.
When focusing on quantities at the species level, it was decided not to take into account fodder pea and vetches. Indeed, these large-seeded species have very high seed yield per hectare and would bias the share of volume. It appears that, on average over the period 1997-2009, perennial ryegrass is the more important species, with 47% of the total production (Figure 72). However, this is biased by the fact that for this species it was not possible to separate uses for forage and for turf. It is followed by Italian ryegrass and meadow grass (Poa pratensis). Seed production and the seed market for tall fescue and cocksfoot are still limited, although these species are well adapted to dry summers and gain more interest in Europe as a consequence of climate change.
Among legumes, alfalfa (Medicago sativa) seed production is by far the largest. The environmental conditions, as well as the landscape structure are well adapted to alfalfa seed production. Indeed, this species requires a compulsory insect pollination for flower triggering. In the European environments, the wild pollinators, mainly wild bees, nesting in the non-perturbed soils of field margins, are able to ensure the pollination, while in other environments, such as the North-West of USA and Western Canada (Alberta) where a lot of alfalfa seeds are produced, it is necessary to supply pollinators, mainly megachiles. This is then a major cost. As a consequence, the European landscape structure provides a source of competitiveness for alfalfa seed production.
The areas of seed production (Figure 73) are mainly located in the west of France, mainly in the Loire Valley, with grasses and legumes. There is also a significant acreage in Picardy and Champagne, with grasses and red clover. This geographic concentration is explained by two features. The first one is the soil and climate conditions which are favourable to either grass or legume species. The second feature is the specialisation of the farmers involved in seed multiplication, which requires a peculiar know-how and the presence of seed companies and the adapted seed cleaning equipment. The geographic location of the seed companies in France is among the most southern ones in Europe. This makes it possible for the companies to have a strategy of marketing right after the harvest for grassland seeding in the autumn. As a consequence, this reduces the storage volumes and costs.
Import of forage seeds
Imports of forage seeds are massive for some species, especially grasses (Table 12). It is particularly important for perennial ryegrass, which is one of the top two species for market volumes. Among the forage legumes, the imports are very important for white clover, whose imports contribute to more than 90% of the markets.
Exports of forage seeds
Exports of forage seeds are small in volume. They are only significant for alfalfa, whose export volumes vary between 1 500 and 4 500 tons a year (Figure 74). The main markets for exports are Italy, Germany and Denmark. The variations between years are mainly due to the market size in Italy, the exports towards Denmark and Germany being fairly stable. The competitiveness of the French producers on these markets is due to the quality of the seeds, the absence of weeds and the low production costs. Indeed, the weather conditions in the main production areas (Loire Valley and Poitou-Charentes) are convenient to achieve satisfactory yields and good seed quality. However, the large variations in yield between years are difficult to handle for farmers. In those environments, the small size of the fields (on average, less than 5 ha) and the presence of many edges ensure the presence of wild pollinators and as a consequence a good pollination of this crop without additional pollinator hives.
Self-sufficiency of French forage seed sector
The self-sufficiency rate was calculated as the ratio (Sales + Export – Import)/Sales. This calculation does not take into account the variation of stocks for a given season. Such variations could explain fluctuations for a given year but do not modify the long term trends.
The self-sufficiency rates are very different for the various species. For white clover, the situation has been very stable over the last three decades, as most of the French market is covered by imports from either Northern Europe (Denmark and Netherlands) or from New Zealand. On the other hand, the long term trend for perennial ryegrass has been drastically modified over the same period. In 1982, for this species, nearly all seeds were imported. As can be seen in Figure 75, the self-sufficiency rate reached 62% in 2003 and even 76% in 2007. The large variations during the last few years were due to large variations in wheat price, which became a major competitor with ryegrass seed production.
To explain this long term change in seed self-sufficiency in perennial ryegrass, it is necessary to consider the changes which occurred in the whole supply chain. Indeed, at the beginning of the 1980s a large effort was put into genetic research and breeding in the public sector (INRA - National Institute for Agronomic Research, Genetics and Plant Breeding division) and by private breeding companies who worked together under the auspices of ACVF (Association des Créateurs de Variétés Fourragères). Research projects were developed together by public and private sectors. They focused on assembling a large collection of perennial ryegrass accessions from all over France and enriched by accessions from the rest of Europe. Several breeding programs were initiated, targeting some key traits for adaptation to French climatic conditions. As a consequence, the main breeding objectives were the forage yield and its distribution over seasons, disease resistance, and late heading in order to maximize the duration of the grazing season without heading. In the meanwhile, production techniques for seed production adapted to the French climate were established by seed companies in partnership with a dedicated technical institute.
Marketing seed mixtures
In France, for the forage species as listed in the directive 66/401/CE, and until 1 February 2004, seeds had been marketed without mixtures. Since 2004, mixtures of varieties are allowed. The French regulation defined that:
Three categories of ecosystem services (inputs, outputs and non-merchant) have been described by Le Roux et al. (2008) in a scientific treatise devoted to agriculture and biodiversity. All three categories can be illustrated by processes and functions occurring in grassland ecosystems (Amiaud and Carrère, 2012). Feed and animal production have been documented in the previous chapters.
In this section, focus is on analyzing three environmental services: carbon storage, soil fertility and biodiversity preservation. Carbon storage and soil fertility were classified as supporting services according to Reid (2005).
Grasslands and carbon storage
Carbon is stored in the soil as an organic form. This organic matter originates from plant organs (leaves, stems, roots,…), dead organs, root secretions of organic molecules and from microbial biomass (Ranjard, 2012). However, the organic matter is progressively mineralized. Farming activities may favour carbon storage through either more organic matter being incorporated into the soil or a slower mineralization. A higher primary production (more senescence of aerial organs and roots), more crop residues and animal feces will increase the restitution of organic matter to the soil. A decrease in mineralization rate will be achieved through a modification of the composition of the organic matter, a modification of land use and of agronomic practices (Arrouays et al., 2002). Some practices, such as turning an annual crop into a perennial crop (such as temporary grasslands), will cumulate several effects and favour carbon storage, by incorporating more roots and, thanks to the absence of soil tillage, stabilizing soil aggregates and the subsequent protection of organic matter (Balesdent and Besnard, 1998). Conversion of ploughed lands into permanent grasslands favours carbon flux to the soil and this was estimated at 0.50 ± 0.25 t/ha/year of C, on a 20 year period. Beyond that period, the accumulation is getting lower. However, storage is twice lower than the losses which occur when a grassland is turned into a cultivated field. More environmental benefits will be lost when turning a grassland into a field, such as loss of biodiversity or nitrogen leaching. Temporary or artificial grasslands have intermediate carbon storage (Schuman et al., 2002) and their conversion to permanent grasslands will induce a more limited additional carbon storage (Loiseau and Soussana, 1996).
The European project, GreenGrass, which gathered 18 research teams from 9 countries showed the ability of grasslands to be on average a sink of 2.4 t C/ha/year (Soussana et al., 2007a). However, a large share of this carbon is temporarily stored, for instance as forage and stored feed. Once considering the emissions of methane and nitrous oxide (on average 20% of the CO2 atmospheric sink), the balance of greenhouse gas (GHG) attributed to grasslands gave a sink of 0.9 t CO2/ha/year, with a large variation between sites and years. It was also shown that grasslands with the highest number of harvests (cutting or grazing) had the lowest sink potential for GHG (Soussana et al., 2007 b). In reality, a compromise has to be found as from the literature (Freibauer et al., 2002), it can be seen that a reduction of fertilization of the very intensive grasslands as well as a moderate intensification of the very extensively managed ones were leading to a higher C storage. Highly fertilized grasslands have a lot of grass species with a very competitive strategy, which are able to rapidly recycle their nutrients and this may lead to a rapid degradation of the organic matter during litter decomposition (Wedin and Tilman, 1996). On the other hand, grassland swards on unfertile soils mainly include species with a conservative strategy characterized by a slow turn-over of nutrients, inducing a slow degradation of the litter. In such a case, the carbon storage is increased because of the slower decomposition of the organic matter (Post and Kwon, 2000). However, when species with high agronomic value are introduced, such as after over-seeding of forage legumes, the carbon storage may be increased by 0.1 to 3 t C/ha/year thanks to a higher primary production of aerial parts and roots (Thornley et al., 1991; Conant et al., 2001). Similarly, a regular grazing will stimulate root growth because of the higher turn-over of the aerial parts. This will stimulate the accumulation of carbon storage in grazed grasslands. This could be further stimulated by the animal dungs (Hassink and Neeteson, 1991).
Nevertheless, the potential for soil carbon storage highly depends on the management and harvest regime (grazing and grazing management, cuts, silage production,…). Indeed, Klumpp et al. (2009) showed that at 1000 m altitude in Massif Central, the mean soil carbon storage was 2 t C/ha/year with a mean stocking rate of 0.5 LU/ha without fertilization, but highly depended on the annual rainfall regime. Under 1 LU/ha and 210 kg N/ha of annual fertilization, the net accumulation of the ecosystem was much lower (Figure 77).
Among permanent grasslands, changes in management could lead to contrasting effects, in strong interactions with variates such as previous paddock management. It is often mentioned that additional research is needed to precisely document the potential of soil carbon storage in grasslands, in interaction with animal husbandry.
Grasslands and soil fertility: relationships between plants and micro-organisms
The recent advances on the priming effect (PE) showed that the decomposition of soil organic matter (SOM) was reduced by the activity and dynamics of microbial populations which have still to be identified. Priming effect is the ability of decomposers to degrade soil organic matter by using energy of organic matter recently released by plants. It was shown by Fontaine et al. (2011) that fungi were key players of PE. It was also shown that grasslands were a bank of nutrients. Indeed, fungi produce organic stocks when a lot of soluble nutrients are available and thanks to the PE, release them into the soil solution in case of nutrient shortage. This mechanism would explain the synchrony between the availability of soluble nutrients, which is conditioned by the microbial activity and the potential of plants to uptake the nutrients. Alternatively, in the case of highly perturbed environments (cultivated soils, overgrazed pastures), the reduction in fungi activity would lead to a loss of SOM and nitrate leaching.
It was established that according to grazing intensity, permanent grasslands could move towards two contrasting systems as far as species diversity and carbon storage are concerned. Klumpp and Soussana (2009) studied this relationship on grasslands with a change in the grazing management. Dynamics of plant and microbe communities, aerial and root biomass, and litter were studied over two years. Changes in grazing management induced a cascade of events. It was shown that roots of plants with a slow growth, well adapted to low grazing pressure, inhibit Gram+ bacteria, litter decomposition, nitrogen availability and as such the development of species with a rapid growth. The intensification of grazing induces a large root mortality of plants with a slow growth, and thus releases all these inhibitions and induces a C release in soil litters. Moreover, grazing intensification generates a loss of nutrients through a leaching which favours the priming effect and, in the long term, loss of Carbon.
Similarly, management fertilization which orientates botanical composition influences microbial communities and thus soil mineralization capacities. Benizri and Amiaud (2005) showed that microbial communities from soils of grasslands which were fertilized with 120 kg N/ha/year significantly differed from those of unfertilized soils (Figure 78). This difference existed throughout the season of growth and depended on the sward botanical composition. Bacteria communities preferentially oxidized nitrogen compounds in paddocks with little plant diversity and dominated by cocksfoot and Trisetum flavescens. In non-fertilized grasslands with a more diverse vegetation where Agrostis stolonifera and Convolvulus arvensis were dominant, bacteria communities preferentially used carbohydrates.
Grasslands and biodiversity
The link between grasslands and biodiversity is important and complex.
It is important because grasslands play a key role in preserving biodiversity and could be further mobilized to preserve common and heritage biodiversity. This importance is due to the fact that grasslands contribute a major part of the agricultural land in France and to the large diversity of grasslands, thus offering a huge heterogeneity which is a component favourable to biodiversity, from the local scale to the territory scale.
It is complex because the link is both direct and indirect and because it has to be considered at the three organizational levels of biodiversity: diversity of the communities, diversity of the species within communities, genetic diversity within species.
In the present section, the direct and indirect links between grasslands and biodiversity are analyzed by looking at the underlying mechanisms. The aim is also to identify the management practices which are favourable and could be mobilized to further enhance or protect the biodiversity. The relationship between botanical diversity and production is not investigated. This is an item related to the biodiversity of the grasslands ecosystem.
Grassland and hosted biodiversity
The grassland ecosystem hosts many plant species, which are either the main source of forage or which are simply present and contribute to the common diversity.
The diversity of flora has been extensively documented in many grasslands and a dedicated web site has been created by the group de Sylvain Plantureux (INPL Nancy) and provides a huge resource on the plant diversity in grasslands and on the plant community (http://eflorasys.inpl-nancy.fr/index.php/en/pages/display/home).
Two detailed surveys of French permanent grasslands are relevant to this section and will be presented here.
The first one by Plantureux and Amiaud (2008) focused on the permanent grasslands in the Vosges, a mountainous region in the east of France (Figure 79). This survey showed a wide range of variation in the number of plant species, from 25 to 66, with the other dicots being the most abundant and explaining most of the variations. However, it is interesting to notice that the abundances in all three groups are highly correlated. In a second survey (Figure 80), on a larger set of grasslands, they studied the number of species according to their mode of propagation. They especially demonstrated that a large proportion of species was seed propagated, while the number of vegetatively propagated species was lower and tended to plateau.
These interesting surveys did not take into account the type of grasslands. This downside was corrected in the work of Launay et al. (2011) who described flora in each of the 19 types of permanent grasslands. From this survey, it can be seen (Figure 81) that all types were very similar for the mean number of plant species, except the PA1 (Grasslands in high altitude in dry and acidic conditions with red fescue and Agrostis) which had on average more plant species but with a large variance among paddocks of this type. However, PO1 (Western grasslands wet and well fertilized with meadow grass and velvet grass) had significantly less plant species than the average. Among those plant species, some have heritage values, and special efforts have been devoted to protect them. It is worth mentioning an initiative taken by the Conservatoire Regional des Espaces Naturels in the Poitou-Charentes region which targeted the protection of one site, la Côte Belet at Pamproux, Deux-Sèvres, where up to 23 species of orchids have been identified on only a few hectares (Figure 82). The open structure of the grasslands is needed for these species to set and grow. Sheep grazing at a low stocking rate were established to maintain open areas and avoid invasion by shrubs.
The survey by Launay et al. (2011) also identified the taxa and identified the species which are visited by pollinating insects (Figure 83). The three types of grasslands of the West Atlantic wetlands have a significantly lower proportion of species favourable to pollinators. However, the proportion of such species varied a lot within and among the types of permanent grasslands. This may be explained by the fact that the pollinators also strongly react to the environment of the grasslands and especially the presence of edges. It was indeed shown in old studies that the proportion of pollinators was highly correlated with the proportion of grasslands which were in direct contact with refuge areas. As this was not available in the study of Launay et al. (2011), it may explain the variation within a given type of grassland.
As documented above for the pollinators, grasslands will host a large animal diversity, and especially the insects. There are very numerous examples of such a diversity which underlines the presence of insects as a consequence of the presence of a given species. Among these examples can be mentioned the dusky large blue (Phengaris nausithous) which preferentially visits a plant species Sanguisorbe officinale which is very often present in permanent grasslands or the wild bees which are only or mainly feeding on grassland dicots.
Another example is the consequence of the presence of Centaurea jacea, this plant species being visited by many insects, both butterflies and pollinators. This clearly means that beyond the number of plant species, the presence of peculiar plant species is very important to increase the hosted biodiversity.
The heterogeneity of the grassland swards will facilitate a broad plant diversity and the associated animal diversity.
In France, special actions were undertaken to preserve a heritage bird species, the corn crake, Crex crex. This species is breeding in the wet permanent grasslands along the rivers, where it sets its nest right on the soil. Large populations used to be met along the Saone and the Loire rivers. It is also present in the wetlands along the Atlantic coast (Figure 84). However, the populations declined by 40% between 1983 and 1992 (Broyer, 1996). The studies run by the ONCFS (Office National de la Chasse et de la Faune Sauvage) showed that the main cause of the decline was the date of the first cut in these grasslands. Due to this early date, the nests were destroyed or exposed to predators before hatching. After discussion with the farmers who became partners of the project, they delayed the cut by two weeks on average. This delay had little impact on the quality of the harvested forages (Broyer, 1997) and was sufficient to secure hatchings. Since this agreement, the corn crake populations in the Loire valley and the Saone valley are increasing again (Broyer, 2000).
Biodiversity associated with grasslands and associated fixed elements
Grasslands are usually associated with fixed elements of the landscapes, such as ponds or hedges. In close interactions with the grasslands, and if they are properly managed, these elements will favour the presence of biodiversity, both plant and animal. As a consequence, this biodiversity is indirectly linked to grasslands.
Ponds are the most convenient source of water for the grazing animals in the permanent grasslands and were very abundant in the west of France, except when the soil is shallow and the subsoil is limestone. Ponds welcome of lot of diversity and in the case of the wetlands of the Atlantic coast, species include the white swan.
The gradient of plant biodiversity around hedges was studied by Petit et al. (2008) as a key to understanding the dynamics of weeds in annual grain crops.
Hedges are a very important component of landscapes with a high proportion of grasslands, as illustrated on Figure 85. This was actually a key component of the hedged farmland system, which was very common over the western part of France. In this traditional landscape, the hedges were used to keep the animals in the grasslands, both temporary and permanent ones, protecting the crops. They were also a source of woods, with the presence of oak trees regularly cut at 2 m high and named in French ‘chêne tétard’.
Hedges favour the presence of birds. A survey run by V. Barret in 2007 documented the presence of three common Columbidae (Turtle dove, collared dove and wood pigeon) in a hedged farmland of the Pays de Loire region (La Chambaudière, Vendée), over three spring periods. As expected and as it can be seen in Figure 86, the birds are totally associated with the hedges.
It is very obvious that the distribution of the hedges in a territory, their abundance and their management will be key factors affecting the biodiversity. The increase in the mean size of the paddocks affects the lineage of hedges, reducing their density. This was precisely documented by V. Bretagnolle and his group at CNRS Chizé, Niort, for the zone-atelier of the Niort region, this zone covering a total of 50 000 ha. Based upon aerial photographs and field survey (over the last 15 years), he showed that in 30 years (from 1980 to 2010), the mean field acreage was multiplied by 10, and the grasslands which were scattered all over the territory and tightly mixed with the field crops tended to be less frequent and to cluster in some zones of the territory.
This territory was especially studied for the presence of the little bustard, Tetrax tetrax. This bird, which was common in the West plains and in the Camargue region, strongly declined as a consequence of the modification of the landscape. Indeed, this bird needs both cereal field and grasslands, as the nestlings and the young birds only eat grasshoppers which are only present in grasslands. The intensification of farming on the grasslands, leading to a lower abundance of grasshoppers, and the looser geographic link between field crops and grasslands explained the decrease in the population of little bustards. Work was undertaken with the farmers in this region to study and then facilitate the presence of grasshoppers in the grasslands and also to protect the nests. As a consequence, over the last few years, a slight increase has been recorded. Similarly, in the Champeigne tourangelle area (Centre region) where the little bustard used to be present, but with a strong decline due to the disappearance of grasslands, the farmers started to drill permanent cover with grass and legume species and to manage them in an adapted way. This was done in a concerted agri-environmental measure. Thanks to this concerted action, the population started increasing (Bolotte, 2010).
Genetic diversity in grasslands species
When analyzing biodiversity, little attention is usually paid to the genetic diversity available within the species and how this diversity is shaped by the environment (soil and climate) and the management practices.
This has been documented for two grass species in France when landraces and local accessions were collected as a resource for plant breeding programs. This was undertaken in partnership between INRA and an association of forage plant breeders, gathered in ACVF (Association des Créateurs de Variétés Fourragères). The first large scale collection was done for perennial ryegrass and the material was extensively described by Balfourier and Charmet (1991) and Charmet et al. (1990). The collection of this material and its multi-site evaluation were parts of the concerted effort that was implemented in the early 1990s to set, in France, breeding programs dedicated to perennial ryegrass. This contributed to the self-sufficiency of the French market for perennial ryegrass seeds (see above).
A second case study is offered by fescues and especially the species complex of red fescues, some species of this complex being used for turf (Festuca nigrescens (Figure 87), Festuca rubra ssp. litoralis, rubraand fallax, Festuca ovina). A large collection of material, with more than 400 accessions, was collected all over France and studied in several sites under spaced plant design and turf swards (Sampoux and Huyghe, 2009).
Material belonging to this species complex was collected over a large number of sites covering a wide range of environmental conditions in France (Figure 88)
The material comprised a collection of populations from several taxa of the ‘fine-leaved fescue’ lineage, namely section Aulaxyper (Festuca rubra s.l.) and subsection Festuca (Festuca ovina s.l.) of the subgenus Festuca. This collection was assessed for several traits related to vegetative growth, investment in seed production, morphology and phenology. Environmental parameters that best differentiate taxon-realized niches were identified by multivariate ordination and logistic regression. Canonical correlation analyses were performed to relate plant traits and ploidy-level variation to these environmental parameters. Taxon differentiation was assessed along the first canonical directions of the plant descriptor (plant-trait and ploidy-level) space.
The climatic summer water balance, soil texture and land use were identified as the main environmental parameters differentiating taxon-realized niches. The several land uses were grazed or mown meadows, wasteland which was scarcely or unexploited poor swards and diverse fallow lands, and roadsides. Canonical correlation analyses revealed associations between plant traits and these environmental parameters, independently of ploidy-level variation. More specifically, the production of long and abundant rhizomes appeared to be an efficient adaptation to poor climatic summer water balance. In contrast, the ploidy level was rather weakly associated with environmental parameters, probably because all this material was collected in valleys or at low altitudes. The marked prevalence of high ploidy levels in the collection suggests that they could have a competitive advantage in the range of environmental conditions investigated. Some diploid species of this complex are present in the Pyrenees, such as Festuca eskia and Festuca gautieri but were not considered in this study.
Adaptive trait diversity appears to contribute more than ploidy-level variation to the present diversity of realized niches for fine-leaved fescue taxa, although plant-trait variations were associated to some extent with ploidy-level variation. However, cytotypes with high ploidy levels could have been efficient colonizers and competitors, favouring the expansion of the lineage and leading to their present prevalence. Furthermore, adaptive trait diversification may have been an efficient factor in niche diversification for the present cytotypes with high ploidy levels.
This material, fully characterized, for its turf value and for seed production, is presently used as genetic resources by the breeders.
These two examples clearly illustrate the importance of the various types of grasslands, from intensively managed to extensively managed ones, as a reservoir of genetic diversity which have to be considered as cases of in situ conservation of genetic resources. The study carried out on fine leaved fescues also showed how the land use shaped the genetic diversity.
The improvement of pasture resources may use several opportunities but must fulfill the requirements of sustainable agriculture. This especially requires a high economic and environmental performance. Among the various environmental issues, special attention must be paid to the energy aspects, as most agricultural production, and especially animal production, shows a fairly low energy efficiency. The presence of grasslands and forage crops in the French landscapes is considered as a benefit for the environment and its diversity, including the biodiversity. As a consequence, any improvement of the pastures resources must take into account the need to preserve these benefits.
Three main issues may be considered to improve these resources:
The first one is the combined use of permanent, temporary grasslands and annual forage crops, including the less productive ones. Over long periods of time, most work has tended to oppose these two groups of grasslands, trying to better use them independently. Even though it is possible to make the best of each of them independently (see below), it is above all necessary to consider them as a unique resource. This means for instance the possibility to define and manage the temporary grasslands as a function of the available resources from permanent grasslands, but also to define them in order to maximize the use of permanent grasslands, but also the security of use of this resource. This is especially the case in the harshest environments where the variation between years may be large. In such environments, the temporary grasslands, which may contribute only a minor share of the resources, must be defined to optimally use the permanent grasslands. Such a situation is common found in the South of France in the Causse du Larzac, where alongside the rangelands, small acreages of alfalfa are harvested and secure the whole forage and production systems.
Similarly, more attention must be paid to the coupling between plant and animal production. France used to be a country of mixed farming, where animals and crops used to be met on every farm. Progressively, farms and territories became more and more specialized in either crop or animal production. This especially resulted in nitrogen surpluses in animal farming territories, where annual forage crops play a key role as areas for manure spreading and massive uses of mineral nitrogen fertilizers and pesticides in crop farming territories. The re-coupling is a complex issue and requires the implementation of new territory organization with innovative interactions among farms with specialized productions.
The second key issue is related to the implantation of multi-species swards in temporary grasslands, and in the management of permanent grasslands to maintain or enrich their species diversity. In such multi-species, the main source of nitrogen fertility is the symbiotic fixation of nitrogen by the legumes and as a consequence, the proportion of legumes and the stability of this proportion are key features. In temporary grasslands, as mentioned earlier and as shown by the seed market, there is a regular trend to have a larger proportion of species-rich grasslands. However, the management is still not fully adapted with a high use of mineral nitrogen fertilization, even on the temporary grasslands with legume species. The main reason reported by farmers is the risk of low biomass production. Thus, it is first a question of dissemination of knowledge on the optimum management and performance of such temporary grasslands with large species diversity. However, there are still a lot of open fundamental and applied research questions. Among these issues, the optimum species and variety composition and the sowing doses must be defined to meet the objectives of economic and environmental performances and the optimum management must be adapted. The adequate decision tool kits must be identified. As a consequence of this field of opportunities, it may be questioned whether the range of species used is well adapted to this main use in multi-species swards. This questions the organization of the seed industry as well as the seed regulations. However, the present EU directive 66/401/CE which defines the list of species whose seeds must be certified is very broad and offers a range of possibilities. This also questions the breeding objectives of the grass and legume forage species. The breeding objectives, as well as the registration criteria mainly take into account the use in pure stand. It is necessary to re-visit these questions to investigate how this could be modified to take into account such a change in the composition of grasslands. It includes the plant and sward architecture, the weight given to disease resistance, the disease pressure being lower in multi-species swards than in pure stands, and the quality components.
The third key issue is the reduction of losses. Only a part of the available biomass is used by the animals in grazed swards and when harvested in hay or in silage a significant share of the available biomass is lost. In the French research programs, attention is paid to a better use of the available biomass through grazing. This includes an adequate management, using decision tool kits for an optimum management of the whole set of paddocks that is available at the scale of the farms. It also includes the possibility to combine several animal species in order to optimally use the available biomass. Indeed, the various types of herbivores do not use the sward in the same way (height of defoliation, homogeneity of grazing). Moreover, the grazing by various herbivores may induce an heterogeneity of swards within paddocks and among paddocks that may be favourable to hosted biodiversity.
In the domain of grasslands and forage crops, the main research and development organizations are:
AFPF – Association Française pour la Production Fourragère
GNIS – Groupement National Interprofessionnel des Semences
ACVF – Association des Créateurs de Variétés Fourragères (this association gathers all forage crop breeders)
GEVES – Groupement d’étude et de contrôle des variétés et des semences
Idele – Institut de l’Elevage
Arvalis – Institut du Végétal
Coop de France Deshydratation
BTPL – Bureau Technique de Promotion Laitière
APCA – Assemblée Permanente des Chambres d’Agriculture
Agreste (various dates). The national statistical agency for agriculture and fisheries. Source of various statistical data. < http://agreste.agriculture.gouv.fr/ >.
Amiaud, B. & Carrère, P. 2012. La multifonctionnalité de la prairie pour la fourniture de services écosystémiques. Fourrages (in press).
Ansquer, P., Duru, M., Theau, J-P. & Cruz P. 2009. Convergence in plant traits between species within grassland communities simplifies their monitoring. Ecological Indicators 9, 1020-1029.
Arrouays, D., Antoni, V., Bardy, M., Bispo, A., Brossard, M., Jolivet, C., Le Bas, C., Martin, M., Saby, N., Schnebelen, N., Villanneau, E. & Stengel, P. 2012. Fertilité des sols : conclusions du rapport sur l’état des sols de France. Innovations Agronomiques 21 (à paraître).
Arrouays, D., Balesdent, J., Germon, P.A., Jayet, J.F., Soussana, J.F. & Stengel, P. 2002. Contribution à la lutte contre l’effet de serre. Stocker du carbone dans les sols agricoles de France. Expertise scientifique collective. INRA France. 334 pp.
Aufrère, J., Baumont, R., Delaby, L., Pecatte, J.-R., Andrieu, J., Andrieu, J.-P. & Dulphy, J.-P. 2007. Prévision de la digestibilité des fourrages par la méthode pepsine-cellulase. Le point sur les équations proposées. INRA Productions Animales 20, 129-136.
Balesdent, J. & Besnard, E. 1998. The dynamics of carbon in particle-size fractions of soil in a forest cultivation sequence. Plant and Soil 201, 49-57.
Balfourier, F. & Charmet, G. 1991. Relationships between agronomic characters and ecogeographical factors in a collection of French perennial ryegrass populations. Agronomie 11, 645-657.
Baumont, R., Aufrère, J., Niderkorn, V., Andueza, D., Surault, F., Pecatte, J.R., Delaby, L. & Pelletier, P. 2008. La diversité spécifique dans le fourrage : conséquences sur la valeur nutritive. Fourrages 194, 189-206.
Baumont, R., Michaud, A. & Delaby, L. 2012. Services fourragers des prairies permanentes : production d’herbe et valeur alimentaire pour les ruminants. Fourrages (in press).
Benizri, E. & Amiaud, B. 2005. Relationship between plants and soil microbial communities in fertilized grasslands. Soil Biology and Biochemistry 37, 2055-2064.
Boivin, R. Bony, S. & Emile, J.C. 1998. Effect of feeding endophyte-infected tall fescue on the ruminal blood flow in sheep. Mycotox 98, Toulouse, France, p 627.
Bollotte, E. 2010. La gouvernance locale en Champeigne tourangelle : clef de voûte de l'élaboration et de la contractualisation des MAET ''Culture Outarde''. Fourrages 202, 125-130.
Bony, S. Collin, E., Perret du Cray, G., Ravel, C. & Delatour, P. 1998. Endophyte toxicosis : observations on an outbreak of ‘ryegrass staggers’ in a dairy cow herd in France. Mycotox 98, Toulouse, France, p 628.
Brisson, N., Mary, B., Ripoche, D., Jeuffroy, M.H., Ruget, F., Nicoullaud, B., Gate, P., Devienne-Barret, F., Antonioletti, R., Durr, C., Richard, G., Beaudoin, N., Recous, S., Tayot, X., Plenet, D., Cellier, P., Machet, J.M. Meynard, J.M. & Delecolle, R. 1998. STICS: a generic model for the simulation of crops and their water and nitrogen balance. I. Theory and parameterization applied to wheat and corn. Agronomie 18, 311-346.
Broyer, J. 1996. Les « fenaisons centrifuges », une méthode pour réduire la mortalité des jeunes Râles des genêts et Cailles des blés. Revue d’Ecologie (Terre Vie), 51, 269-276.
Broyer, J. 1997. Incidence des fenaisons tardives sur la valeur nutritive des fourrages dans les prairies inondables de la vallée de la Saône. Fourrages 150, 225-234.
Broyer, J. 2000. Le Râle des genêts, Éveil éditeur, coll. « Approche », Saint-Yrieix-sur-Charente, 106 p.
Charmet, G., Balfourier, F. & Bion, A. 1990. Agronomic evaluation of a collection of French perennial ryegrass populations – Multivariate classification using genotype x environment interactions. Agronomie 10, 807-823.
Conant, R.T., Paustian, K. & Elliot, E.T. 2001. Grassland management and conversion into grassland. Effect on soil carbon. Ecological Applications 11, 343-355.
Cruz, P., Theau, J.-P., Lecloux, E., Jouany, C. & Duru, M. 2010. Typologie fonctionnel de graminées fourragères pérennes : une classification multitraits. Fourrages 201, 11-17.
Daccord, R., Wyss, U., Kessler, J., Arrigo, Y., Rouel, M., Lehmann, J. & Jeangros, B. 2006. Apports alimentaires recommandés et tables de la valeur nutritive des aliments pour les ruminants. Chapitre 13. Valeur nutritive des fourrages, 18p. On line publishing. Station de rech. Agroscope Liebefeld-Posieux ALP, Posieux.
Delaby, L., Pecatte, J.R., Aufrère, J. & Baumont, R. 2007. Description et prévision de la valeur alimentaire de prairies multi-espèces – Premiers résultats. Renc. Rech. Ruminants 14, 249.
Duru, M., Cruz, P. & Theau, J.P. 2008. Un modèle générique de digestibilité des graminées des prairies semées et permanentes pour raisonner les pratiques agricoles. Fourrages 193, 79-102.
Duru, M., Cruz, P. & Theau, J.P. 2010. A simplified method for characterising agronomic services provided by species-rich grasslands. Crop & Pasture Science, 61, 420-433.
Emile, J.C. 1996. Demain, quelles prairies, et avec quel matériel végétal, pour les systèmes de production de ruminants? Fourrages 147, 223-236.
Evans, R.D., Dillon, P., Wallace, M. & Garrick, D.J. 2004. An economic comparison of dual-purpose and Holstein-Friesian cow breeds in a seasonal grass-based system under different milk production scenarios. Irish Journal of Agricultural and Food Research 43, 1-16.
Farrie, J.P., Launay, F., Pottier, E., Michaud, A., Baumont, R. & Plantureux, S. 2011. L’utilisation des prairies permanentes au travers d’une enquête nationale sur leur place dans les systèmes d’alimentation. Rencontres Recherches Ruminants 18, 233-236.
Fontaine, S., Henault, C., Aamor, A., Bdioui, N., Bloor, J.M.G., Maire, V., Mary, B., Revaillot, S. & Maron, P.-A. 2011. Fungi mediate long-term sequestration of carbon and nitrogen in soil through their priming effect. Soil Biology and Biochemistry 43, 86-96.
Freibauer, A. & Rounsewell, M.D.A. et al., 2002. Background paper on carbon sequestration in agricultural soils under Article 3.4 of the Kyoto Protocol. Contract Report N° 2001.40.CO001 within the framework of the Communication on “EU policies and measures to reduce greenhouse gas emission: Towards a European Climate Change Programme (ECCP)”, COM (2000) 88, Working Group Sinks, Subgroup Soils. 50 pp.
Guillaumin, J.-J., Frain, M., Pichon, N. & Ravel, C. 2000. Survey of the fungal endophytes in wild grass species of the Auvergne Region (Massif Central). In The Grassland Conference 2000. pp 85-92.
Hassink, J. & Neeteson, J.J. 1991. Effect of grassland management on the amounts of soil organic N and C. Netherland Journal of Agricultural Science 39, 225-236.
Houssin, B. & Pavie, J., 2010. Le séchage en grange. Pour récolter le meilleur de l’herbe. Des expériences en Normandie. Institut de l’Elevage, Paris, France, 76 pages.
Huyghe, C., Baumont, R. & Isselstein, J. 2008. Plant diversity in grasslands and feed quality. Grassland Science in Europe 14, 375-386.
Huyghe, C., Duru, M., Peyraud, J.L., Lherm, M., Gensollen, V., Bournoville, R. & Couteaudier, Y. 2005. Prairies et cultures fourragères: au carrefour des logiques de production et des enjeux environnementaux. INRA Editions, 209 p.
INRA, 2007. Alimentation des bovins, ovins et caprins, Besoins des animaux – Valeur des aliments Tables INRA 2007 (J. Agabriel coord.), Editions Quae, 312 p.
Joly, D., Brossard, T., Cardot, H., Cavailhes, J., Hilal, M. & Wavresky, P. 2010. Les types de climats en France, une construction spatiale. Cybergeo : European Journal of Geography [En ligne], Cartographie, Imagerie, SIG, article 501, mis en ligne le 18 juin 2010, URL : < http://cybergeo.revues.org/23155 ; DOI : 10.4000/cybergeo.23155 >.
Julier, B., Ecalle, C., Lila, M., Emile, J.C., Huyghe, C., Guines, F. & Briand, M. 2003. Eléments pour une amélioration génétique de la valeur énergétique de la luzerne. Fourrages 173, 49-62.
Klumpp, K., Fontaine, S., Attard, E., Le Roux, X., Gleixner, G. & Soussana, J.F. 2009. Grazing triggers soil carbon loss by altering plant roots and their control on soil microbial community. J. of Ecology 97, 876-885.
Klumpp, K. & Soussana, J.F. 2009. Using functional traits to predict grassland ecosystem change: a mathematical test of the response-and-effect trait approach. Global Change Biology 15, 2921-2934.
Latch, G.C.M., Potter, L.R. & Tyler, B.F. 1987. In,cidence of endophytes in seeds from collections of Lolium and Festuca species. Ann. Appl. Biol. 11, 59-64.
Launay, F., Baumont, R., Plantureux, S., Farrie, J.-P., Michaud, A. & Pottier, E. 2011. Prairies Permanentes : des références pour valoriser leur diversité. Ed. Institut de l’Elevage. 128 pages.
Le Gall, A., Faverdin, P., Thomet, P. & Vérité, R. 2001. Le pâturage en Nouvelle-Zélande: des idées pour les régions arrosées d'Europe. Fourrages 166, 137-164.
Le Roux, X., Barbault, R., Baudry, J., Burel, F., Doussan, I., Garnier, E., Herzog, F., Lavorel, S., Lifran, R., Roger-Estrade, J., Sarthou, J.P. & Trommeter, M. 2008. Agriculture et biodiversité. Valoriser les synergies. Expertise Scientifique Collective INRA, pp 175.
Leyronas C. & Raynal, G. 2001. Presence of Neotyphodium-like endophytes in European grasses. Ann. Appl. Biol. 139, 119-127.
Liljenstolpe, C. 2009. Horses in Europe. Swedish University of Agricultural Sciences, Uppsala, 28 pages.
Loiseau, P., Soussana, J.F. et al., 1996. Evolution des stocks de matières organiques sous prairies : quantification, évolution, modélisation. Les Dossiers de l'Environnement INRA 10, 57-77.
Lowell, F.C. 1990. Observations on heaves. An asthma-like syndrome in the horse. Allergy Proc.11, 147-150.
Mair, T.S. & Derksen, F.J. 2000. Chronic Obstructive Pulmonary Disease: a review. Equine Vet. Educ.12, 35-44.
Michaud, A., Andueza, D., Picard, F., Plantureux, S. & Baumont, R. 2011. The seasonal dynamics of biomass production and herbage quality of three grasslands with contrasting functional compositions. Grass and Forage Science, DOI: 10.1111/j.1365-2494.2011.00821.
Michelin, Y. 2008. L’approche sémiologique au service de la mise en évidence du lien produit agricole – paysage : l’exemple de l’AOC Saint-Nectaire. Nouveaux Actes Sémiotiques, < http://revues.unilim.fr/nas/document.php?id=2413 >.
Naffaa, W. 1998. Les endophytes des graminées : variabilité et co-adaptation avec leurs hôtes. Thèse de Doctorat, Université Blaise Pascal.
Pecatte, J.R. 2008. Valeur alimentaire de foins ventilés issus de prairies multispécifiques. Fourrages 195, 354-356.
Petit, S., Thenail, C., Chauvel, B., Le Cœur, D. & Baudry, J. 2008. Les apports de l’écologie du paysage pour comprendre la dynamique de la flore adventice. Innovations Agronomiques 3, 49-60.
Pflimlin, A. 2010. Europe laitière. Valoriser tous les territoires pour construire l’avenir. Editions La France Agricole, 309 pages.
Plantureux, S. & Amiaud, B. 2008. Intérêt des prairies à flore complexe pour le préservation de la biodiversité. Journées AFPF, 26-27 mars 2008, Paris.
Post, W.M. & Kwon, K.C. 2000. Soil carbon sequestration and land-use change: processes and potential. Global Change Biology 6, 317-327.
Ranjard, L. 2012. L'apport des techniques de la biologie moléculaire à la connaissance de la biodiversité microbienne dans les sols et de ses fonctions. Innovations Agronomiques 21 (in press).
Ravel, C. 1997. Les champignons endophytes des graminées : distribution et co-évolution avec leurs espèces hôtes. Thèse de Doctorat, Université Blaise Pascal, 120 pages.
Raynal, G. 1991. Observations françaises sur les Acremonium, champignons endophytes des graminées fourragères. Fourrages 126, 225-237.
Reid, W.A. (Ed). 2005. Living beyond our means. Natural assets and human well-being. Millenium Ecosystem Assesment, 28 p. < http://www.maweb.org/documents/document.429.aspx.pdf >.
Renaud, J. 2002. Récolte des fourrages à travers les âges. France Agricole Editions. 416 pages.
Sampoux, J.P. & Huyghe, C. 2009. Contribution of ploidy-level variation and adaptive trait diversity to the environmental distribution of taxa in the ‘fine-leaved fescue’ lineage (genus Festuca subg. Festuca). Journal of Biogeography 36, 1978-1993.
Schott, C., Mignolet, C. & Benoît, M. 2009. Agriculture du bassin de la Seine. Plaquette du programme PIREN-Seine, Agence de l’Eau Seine-Normandie. 79 p.
Schuman, G.E., Janzen, H.H. & Herrick, J.E. 2002. Soil carbon dynamics and potential carbon sequestration by rangelands. Environmental Pollution 116, 392-396.
Soussana, J.F., Allard, V., Pilegaard, K., Ambus, P., Amman, C., Campbell, C., Ceschia, E., Clifton-Brown, J., Czobel, S., Domingues, R., Flechard, C., Fuhrer, J., Hensen, A., Horvath, L., Jones, M., Kasper, G., Martin, C., Nagy, Z., Neftel, A., Raschi, A., Baronti, S., Rees, R.M., Skiba, U., Stefani, P., Manca, G., Sutton, M., Tubaf, Z. & Valentini, R. 2007a. Full accounting of the greenhouse gas (CO2, N2O, CH4) budget of nine European grassland sites. Agriculture, Ecosystems and Environment 121, 121-134.
Soussana, J.F., Fuhrer, J., Jones, M. & Van Amstel, A. 2007b. The greenhouse gas balance of grasslands in Europe. Agriculture, Ecosystems and Environment 121, 1-4.
Spordnly, R. & Nilsdotter-Linde, N. 2011. L'ensilage des prairies temporaires en Suède : un développement réussi. Fourrages 206, 107-118.
Véron, F. & Bernard-Brunet, J. 2003. Utilisation des données satellitales à l’échelle nationale pour apprécier la place de l’herbe dans les paysages cantonaux et les enjeux environnementaux qui en résultent. Ingénierie 33, 35-44.
Véron, F. & Brunet, J.B. 2003. Importance des prairies et enjeux environnementaux associés, analysés à l’échelle cantonale. Journée AFPF, 28 Octobre 2003.
Vignau-Loustau, L. & Huyghe, C. 2008. Stratégies fourragères. Editions La France Agricole, 336 pages.
Wedin, D.A. & Tilman, D. 1996. Influence of nitrogen loading and species composition on the carbon balance of grasslands. Science 274, 1720-1723.
This profile was written by:
[The profile was drafted in the period January-August 2012 and was lightly edited by J.M. Suttie and S.G. Reynolds in August 2012].