PLENARY PAPERS

GRASSLANDS IN GALICIA - J. Piñeiro and A. González

Centro de Investigacións Agrarias de Mabegondo, Apartado 10, 15080 A Coruña, Spain

SUMMARY

The proportion of land devoted in Galicia to permanent and sown pastures in 1997 was 11.1 percent, while crops account for 14.2 percent, woodlands 27.3 percent, shrublands 39.9 percent and other uses 7.5 percent. Total geographical area is 2.9 million hectares.

Annual rainfall varies from 700 mm to more than 2 000 mm and is concentrated in late autumn, winter and spring. There is a summer drought of three to five months. Annual mean temperatures vary from 8.3ºC in the highest altitudes inland to 16ºC on the southwestern coast.

Animal production accounts for 57 percent of the final agrarian production. Dairy farming was the most dynamic activity, experiencing a 182 percent increase in the last three decades, going from 757 million kg of milk in 1962 to 2 137 million in 1996. The average size of the farms is 6.2 hectares, made up of 15.2 plots. Many farms, mainly the smallest, gave up milk production during the last decade leading to a big reduction in the number of dairy farms and to an increase in milk production per farm and animal on the bigger farms. Associated with the increased milk production, more land was devoted to sown grass-legume mixtures, going from 26 thousand hectares in 1965 to 126 thousand in 1995, an increase of 385 percent.

Italian, hybrid and perennial ryegrasses, cocksfoot, white and red clover are the main species used in pasture mixtures with a clearly defined trend in the reduction of cocksfoot at present. Permanent pastures are dominated by Yorkshire fog (Holcus lanatus L.), Agrostis sp., Cynosurus cristatus L., Plantago lanceolata L., etc., with a small presence of legumes.

The need on farms to increase forage production led to the transformation of shrublands into pastures. A relevant research programme was established in the seventies to study different techniques on land reclamation and the requirements for lime, phosphorus and potash in the shrublands, which are characterized by a very low pH, a high aluminium content and a very low phosphorus content.

The biggest change in farm practises was silage making, which was unknown in the fifties and is now common in modern dairy farms. The main problem is the poor fermentative quality of the grass silage. Due to the fact that maize is easier to ensile, it has contributed to more forage use of maize in dairy farms. Direct drilling of maize is a very recent technique which is becoming popular in the better dairy farms.

Despite the increase in the grass and forage acreage, the production of forage is scarce at farm level and farmers on average feed nearly 2 000 kg of concentrates per cow and lactation.

Keywords: species, mixtures, variety testing, nitrogen, land reclamation

INTRODUCTION

Galicia covers 2.96 million hectares, 5.8 percent of Spain's total surface and is located in Northwest Spain, with a population of 2.77 million people, 7.5 percent of the Spanish population.

The proportion of land devoted to permanent and sown pastures is 12.5 percent, while crops account for 12.8 percent, woodlands 27.3 percent, shrublands 39.9 percent and other uses 7.5 percent.

The natural soils are very acid, with Al dominating the cation exchange capacity, and are derived mainly from slates and schists (46 percent), granites (45 percent), basic rocks (5 percent) and sediments (4 percent). Limestone is present but is not very widespread.

Annual rainfall varies between 700 to more than 2 000 mm and is concentrated in late autumn, winter and spring. There is a summer drought of three to five months. Annual mean temperatures vary from 8.3ºC in the highest altitudes inland to 16ºC in the southwestern coast, going from 2.2 to 11.7ºC in the coldest month and from 15.4 to 22.6ºC in the hottest month.

It is a rolling and hilly country with elevations from sea level up to 2 124 m. It is part of the Humid Spain, an area of nearly 6 million hectares, which mainly includes the Northern Autonomic Communities of Galicia, Asturias, Cantabria, Vasque Country and part of Navarra.

GROSS AGRICULTURAL PRODUCT

The contribution of agriculture to the Galician Gross Product decreased from 29.1 percent in 1955 to 4.5 percent in 1985 (Table 1). Fisheries followed by agriculture and then industry, increased slightly while trade and services experienced a big increase. Associated with this tendency there was a dramatic reduction in the percentage of an agriculturally active population. In eight years it decreased from 38.7 percent in 1983 (Fernández, 1987) to 20.3 percent in 1991 but was still very high compared to Spain and the EU (García, 1993). The 1995 Galician Gross Product was 161 times greater than in 1955, while the corresponding Gross Agricultural Product was only 25 times greater.

Table 1. Changes in Galician Gross Agricultural Product from 1955 to 1995

Sources:
1) Fernández (1987);
2) Calculated from data published by IGE (1998).

THE FARMS

The total number of farms remained quite stable during the seventies and eighties. In the 1989 census, 358 886 farms recorded a figure slightly lower (93 percent) than in the 1972 census. The average farm is very small with a total of 6.2 hectares made up of an average of 15.2 separate plots. Only 10.6 percent of the farms are over 10 ha (IGE, 1997). Size and structure are the main limiting factors for farm management and development and help to explain the modern pronounced tendency of completely abandoning farming activities in the small farms because farmers retire without a successor, a trend which will continue in the near future. Twenty-two percent of the dairy farmers gave up milk production between 1987 and 1990, affecting overall the very small farms (Table 2). A recent study (López, 1998) shows that 23 percent of the smaller dairy farmers interviewed intend to abandon agriculture and that only 25 percent of the dairy farms have succession insured at present. Probably the number of farms in the 1999 census will reflect these trends.

Table 2. Galician dairy farms: percentage of dairy farms that gave up milk production between 1987 and 1990, according to milk yield (t) per farm and year.

Source: Posada, 1990.

LAND USE AND MAIN CROPS

In contrast to Dry Spain, Galicia is green due to the more luxurious pastures, which account for 11.1 percent of the total area (Table 3). Unfortunately the green colour is also due to the abundance of shrublands of mainly gorse (Ulex europaeus L., U. gallii Planchon and U. nanus Smith) and heath (Erica sp.) which account for nearly 40 percent. A significant part of this is capable of being converted into good pastures or crops.

Maize and potatoes are the main crops of the most common Galician rotations, derived from three basic rotations (Lloveras, 1982); 1) maize or potatoes/rye or wheat/turnips; 2) maize or potatoes/Italian ryegrass or oats; and 3) rye/fallow. Intensification decreases as the distance from the coast increases, going from two crops per year on the coast to one crop every two years in the Southeast, as happens in most of Dry Spain. Even though maize was cultivated as maize for grain, the main use at farm level was for animal feeding in the traditional farming. The plant is divided into three parts: 1) the above ear part of the plant was manually cut and fed as green forage in summer; 2) the grain as concentrate; and 3) the rest of the plant as straw after harvesting the grain. From the sixties onwards on the better sized farms, the whole plant is harvested for silage and the area of forage maize increased. The rotations played an important role in traditional agriculture, to meet animal needs all year round, complemented with hay from permanent pastures. With the adoption of silage making in the last 30 years, hay became less important in animal production and rotations were simplified, including more short rotation sown pastures.

Table 3. Land use in Galicia in 1997.

Source: AEA, 1998.

GRASSLANDS, FORAGES AND MILK PRODUCTION

The area covered by pastures in 1977 in Galicia accounted for 328 604 hectares, of which 208 594 (63 percent) are permanent and 120 010 (37 percent) are sown mixtures of grasses and clovers (AEA, 1998). Within Humid Spain, the sown mixtures are relatively more important in Galicia than in the other Autonomic Communities. While Galician sown grass-clover mixtures represented 87.7 percent of the mixtures sown in Humid Spain in 1995, permanent pastures represented only 30.4 percent. Differences in topography and soil lead to differences in type of pastures. Most of the permanent pastures are on slopes or on flat wet lands and Galicia is less hilly. Also, the other Humid Areas have more limestone based soils on the slopes, which are prone to being covered by pastures, while in Galicia the soils are predominantly based on granites, schists and slates, and the slopes are mainly covered by gorse or heath.

The sown grass-clover mixtures and forage maize acreage increased steadily from the fifties onwards, matching the increase in dairy milk production (Table 4). Permanent pasture areas increased from the fifties to the mid-seventies and stabilized from this date onwards. There was a decrease at the end of the thirties and the beginning of the forties as a consequence of the civil war (1936-1939) which led to planting of cereals in some permanent pasture lands. From this we can conclude that Galician farmers depended on their own farm grasslands and forages to base the spectacular development of animal production, mainly dairying, that has occurred during the past 35 years.

Nevertheless, data from 1 580 dairy farms of the Dairy Management Programme of the Regional Ministry of Agriculture, with an average size of 23.2 cows, showed that the milk from forages was as low as 20.9 percent in 1996 (Table 5), as a consequence of feeding 2 035 kg of concentrates per cow yielding 5 652 l of milk per year. Despite the programme target of reaching 40 percent of the milk from forages, there is a clearly defined negative tendency due to the good relationship between milk and concentrate price. In 1992 the milk from forages was 25.2 percent, still very low compared to the 45.9 percent obtained in the 4 269 dairy farms associated with the England and Wales Genus Milkminder Programme in the same year (Estévez, 1995). This big difference was mainly attributed to: 1) low use of nitrogen (Estévez, 1995); 2) low quality of pasture silages (Flores et al., 1990); and 3) in general, poor grassland management techniques. The average nitrogen application rates and silage fermentative quality are still low at present, despite efforts made by advisers of the main dairy cooperatives in promoting the best techniques for silage making. The structure of many farms, divided in several small independent pieces, makes grasslands management for grazing or silage a difficult task.

Table 4. Milk production vs. permanent pastures, sown grass-clover mixtures and forage maize in Galicia from 1935 to 1995.

Sources: Agrarian Statistics Bulletins of Ministries of Agriculture of Spain and Galicia.

PASTURE SPECIES

Permanent pastures

Permanent pastures are dominated by Yorkshire fog (Holcus lanatus L.), Agrostis sp., Cynosurus cristatus L., Plantago lanceolata L., etc., with a small presence of legumes like Trifolium repens L., Lotus corniculatus L. and Lotus uliginosus Schkuhr, etc.

Sown pastures

Although there are some previous positive experimental basis about the performance as pasture species of Lolium perenne L., Phleum pratense, L, Trifolium pratense L., and Medicago sativa, L. (González Pizarro, 1985; Hernández Robredo, 1933 and Blanco, 1945), the experimental basis for pasture species and mixtures was not developed until the fifties, as a consequence of the work carried out under the “Plan Agrícola de Galicia”, created in the late forties.

From 1948 until the late fifties a total of 50 pasture species were on trial in different localities of Galicia. As a conclusion of this work, Yepes (1961) recommended four main species (Lolium multiflorum Lam., Lolium perenne L., Dactylis glomerata L. and Phleum pratense L.) among the grasses, and four among the legumes (Trifolium incarnatum L., Trifolium pratense L., Trifolium repens L. and Medicago sativa L.).

Later, farmers' experience led to the decline in the usage of Phleum pratense L. and Trifolium incarnatum L.. Medicago sativa L. remained limited to some specific areas and was never widely used in Galicia despite the good, early results reported by Hernández Robredo (1933).

Table 5. Average results of 1 580 Galician dairy farms of the “Programa de Xestión de Leite” (Dairy Management Programme) in 1996.

Source: (Barbeyto, 1997).

Italian and perennial ryegrasses represented 91.3 percent of the Spanish seed market in the period 1993/94-1996/97 of pasture species eligible for sowing pastures in Galicia (Table 6). Obviously some of them are used to sow turfs and road margins but the figures give the trend of what is happening at farm level. Although there is no statistics on seed consumption at regional level, the seed purchased by a big cooperative that services the demand of about 9 500 farmers (Table 7) shows that the ryegrasses represent 90.4 percent of the total. Taking into account that the percentage of white clover, tall fescue and alfalfa used by the cooperative farmers is low, it can be concluded that the Spanish market species structure reflects the use of pasture seed in Galicia.

PASTURE MIXTURES

Recommendations published early in the present century (Hernández Robredo, 1915 and Eguileor, 1917) about pasture mixtures that contain from 2 to 14 components, chosen among 28 species, were based more on theoretical considerations than on experimental results, which were not available until the “Plan Agrícola de Galicia”, which was developed in the fifties. In the publications about mixtures made in the late fifties (Yepes, 1957) and early sixties (Fernández-Quintanilla, 1961; Yepes, 1963 and Azcárate, 1965), a big reduction in seed rate, number of components and number of species used is observed.

Table 6. Pasture species seed consumption (t/year) in Spain¹ of the species used in Galicia. Average for the period 1993/94-1996/97.

Table 7. Seed consumption (%) of the different pasture species used by the farmers of a big Galician cooperative in 1997.

Pasture mixtures in the sixties

The most common mixtures in the late sixties were:

Short-term rotation:

Italian ryegrass (4 kg/ha)
Red clover (10 kg/ha)
Cocksfoot (10 kg/ha)
Large leaf white clover (2 kg/ha)

Long-term rotation:

Perennial ryegrass (10 kg/ha)
Cocksfoot (10 kg/ha)
Medium leaf white clover (1 kg/ha)
Large leaf white clover (2 kg/ha)

It was estimated that in the late sixties and early seventies about 80 percent of the acreage was sown with the short-term rotation mixture and 20 percent with long-term rotation.

Changes from the sixties to the nineties

Farmers' and their advisers' experience, accumulated in the sixties, seventies and eighties, led mainly to the reduction of cocksfoot in the mixtures because of the lower palatability compared to other components and because of its tufted growth habit. Its wider use in the fifties and sixties was restricted to the areas with the driest summers because of the lower rainfall, higher temperature and more sandy soils. In the nineties, the continued reduction in the use of cocksfoot was also evident in these areas. Farmers prefer to renew the pastures more frequently than to sow cocksfoot. There was also a reduction in the use of red clover due probably, to its dominance in the second year in some cases, which added an extra difficulty to silage making, before silage additives were available in the Galician market.

Pasture mixtures in the nineties

Among the mixtures recommended in the nineties (Piñeiro and Pérez, 1993), the most commonly used are slight variations of the following:

Short-term rotation:

Biannual Italian or Italian type hybrid ryegrass (20 kg/ha)
Red clover (10 kg/ha)

Long-term rotation:

Mixture 1

Perennial ryegrass (30 kg/ha)
Medium or large leaf white clover (3 kg/ha)

Mixture 2

Perennial ryegrass (20 kg/ha)
Intermediate or perennial type hybrid ryegrass (10 kg/ha)
Medium or large leaf white clover (3 kg/ha)

Mixture 3

Perennial ryegrass (20 kg/ha)
Cocksfoot (10 kg/ha)
Medium or large leaf white clover (3 kg/ha)

Ryegrass seed rates are established on a diploid basis. A 40 percent increase is recommended for tetraploids.

VARIETY TESTING

Information on the value for cultivation and use of the varieties of the different pasture species was very scarce until the early seventies. It was not until this time that the earliest studies, started between 1965 and 1970, were available (Yepes and Pérez, 1971 and Zulueta, 1971). Some more trials continued between 1970 and 1975, but they did not correspond with the varieties marketed in Spain, i.e. there was no connection between the varieties sold in Spain and the varieties included in the trials.

After the establishment of the “Lista de Variedades Comerciales” (equivalent to National List or Catalogue in other countries) of Alfalfa in 1975, of Italian ryegrass, perennial ryegrass, hybrid ryegrass, cocksfoot, tall fescue and meadow fescue in 1976, and red and white clovers in 1985, a given variety of those species had to be included in the lists before being marketed in Spain. This was the tool that linked variety trials with the Spanish market.

The pasture varieties evaluation system for Humid Spain started in 1978 under an Agreement between the “Instituto Nacional de Semillas y Plantas de Vivero” (INSPV) (National Institute of Seeds and Nursery Plants), responsible for the National List and the “Instituto Nacional de Investigaciones Agrarias” (INIA) (National Institute of Agrarian Research). Recently, INIA was replaced by the different Autonomic Communities in order to adapt the Agreement to the new structure of the country. In Humid Spain the system started with three sites, while at present the trials are located in five places (Table 8), three of them in Galicia because of the high percentage of Galician sown mixtures within the whole wet region.

Between 1978 and 1998 a total of 540 varieties of Italian, hybrid and perennial ryegrasses, cocksfoot, tall fescue, white clover, red clover and alfalfa (Table 9) were tested in Spain in order to know their value for cultivation and use, and include them in the National List provided they performed well in the trials.

To be in the National List is a guarantee that the variety was tested and performed well, especially if the variety was recently included in the list. The information available is currently given in seminars oriented to technical people of the Advisory Service and Cooperatives and published in simple leaflets. Despite this, and due to the fact that a variety can be sold in Spain provided it is in the Common Catalogue of the European Union, 61 percent of the varieties marketed in Spain in 1998 of the species listed in Table 10 were not in the Spanish National List. This means that our farmers or the people that take decisions in their name, ignore the information available and rely on the variety features given by the owner, taking the risk of buying varieties rejected in the evaluation system because they did not meet the minimum standards imposed by the “Comisión Nacional de Estimación de Leguminosas y Pratenses” (National Commission to Estimate the Value of Leguminous and Pasture Species).

Table 8. Variety evaluation sites in Humid Spain (N and NW).

Table 9. Number of varieties tested in Humid Spain between 1978 and 1998.

The percentage of varieties sold in Spain and not included in the Spanish List increases with time and leads to a continuous decrease in the number of new varieties in trials. Two main reasons explain this tendency: 1) variety owners know that they can sell varieties in Spain provided they are in the European Union Catalogue because Spanish farmers buy varieties not included in the Spanish National List; and 2) the Spanish System for Evaluation of Pasture Varieties is slow, expensive and quite inefficient, and does not attract owners to test their varieties.Table 10. Number of varieties of different pasture species sold in Spain in 1999 and their status respective to the Spanish National List (NL).Table 9. Number of varieties tested in Humid Spain between 1978 and 1998.

Table 10. Number of varieties of different pasture species sold in Spain in 1999 andtheir status respective to the Spanish National List (NL).

LAND RECLAMATION

The demand for meat and milk in Spain increased as a consequence of the improvement of the Spanish Economy in the sixties. Within Spain, Galicia and the “Cornisa Cantábrica” (Cantabric coast countries) are the most suitable areas for intensive beef and, overall, milk production due to more luxurious pastures compared with Dry Spain. As shown in Table 4, Galician farmers increased the area devoted to pastures to meet the needs of their dairy animals in order to produce more milk to meet the increasing market demands. Part of these pastures were planted in the previous cultivated lands and part were established in shrublands, that accounted for 39.9 percent of the land in Galicia, as a good part of them are in good soils provided their chemical properties were corrected.

Although some farmers had their own way of converting shrublands into pastures in traditional agriculture, it was always seen as a difficult operation, taking into account their limited farm machinery and fertilizers. In the sixties, the recommendation to convert a shrubland into a pasture was to plough, fertilize mainly with farm manure, and sow rye or wheat as a pioneer crop for one or two years before the establishment of the pasture.

The need for more grassland areas to meet the increasing demand of increasing herds of better cows led to the establishment of a Research Programme in the seventies with three main objectives to have a sound experimental basis for the development of new pastures in the shrublands: 1) study the fertilizer requirements of the natural soils to establish good pastures; 2) study the different techniques available for land reclamation; and 3) establish animal production systems on these lands. These were the main research activities of the Centro de Investigaciones Agrarias de Mabegondo in the seventies and early eighties, which were complemented with studies on the use of nitrogen fertilizer in established grass-clover pastures.

Lime and fertilisers requirements

In general shrublands have a very low pH of about 5 (in H2O) and a very low P content. After a series of trials in schistic and granitic soils (Piñeiro et al., 1997) the general recommendation to establish a pasture on lands reclaimed from shrublands was:

40-50 kg/ha N
100-150 kg/ha P₂O₅
100 kg/ha K₂O
1 100 kg/ha CaO

Later, more detailed trials added more precision to the recommendations and showed that the problem of acidity was that Al dominated the cation exchange capacity (Mombiela, 1983). In general, these trials confirmed that the previous general recommendations were applicable to more soil types.

Land reclamation techniques

The experiments included three main techniques to directly transform shrublands into pastures: 1) complete cultivation; 2) minimum tillage; and 3) no tillage, with special emphasis in the control of gorse and bracken regrowth in the establishment and maintenance phases of the new pastures (Sineiro, 1997). These series of trials were the experimental basis to develop in 1978, a specific Research Farm on Shrublands using “complete cultivation” in the flatter areas for conservation and grazing, “minimum tillage” in the intermediate slopes for grazing and “no tillage” in the steepest places for grazing, where no machinery can be used.

NITROGEN USE

The average farm of the Galician Dairy Management Programme applies 82 kg/ha of N, a low rate compared to the rates used by northern European countries' standards that rely more on grass than on grass-clover swards. According to Estévez (1995) this is one of the reasons for the dependence of the farms on the use of more concentrates, but the problem of using higher rates of nitrogen is that the contribution of clover to yield decreases.

To have an experimental basis to take decisions about nitrogen use in the Galician grass-cloves swards, a series of trials was carried out in the late seventies and in the eighties with three main objectives: 1) effect of early nitrogen to extend the grazing season; 2) nitrogen rates and dates for silage production; and 3) seasonal response to nitrogen

Early nitrogen

A rate of 30-40 kg ha^-1, applied at different times from mid-January to mid-February was used in several small plot experiments carried out over six years at Mabegondo, on the coast at 100 m altitude, to identify the more appropriate date to apply the first nitrogen after the winter in order to stimulate early growth. Mid-February was the safest to have response to N, even though T-sum 200 was reached earlier in most experiments carried out in different years. It was concluded that by using 30-40 kg/ha, a grazeable yield can be obtained one month earlier than in plots without N (González, 1996).

Similar experiments were carried out in an inland hill area, at 600 m altitude. The safest date to get a response from early nitrogen was mid-March, a month later than near the coast (Mosquera and González, 1992).

Nitrogen rates for silage

Part of the spring grass is conserved as silage, a common technique in the more developed dairy farms. The response to N from 0 to 120 kg/ha varies from 15 kg DM per kg N in clover rich swards to 38 on grass swards or in the establishment year of grass-clover pastures (González, 1991). Repeated silage making in the same sward can lead to a depletion of potash in the soil which mainly affects the persistence of clover. Some studies to elucidate the N-K interaction were carried out (González, 1993a).

The rate of N also affects the crude protein (CP) content, which varied from 12 to 20 percent of the dry matter in an early cut of 30 days, regrowth as the N applied changed from 0 to 120 kg/ha, but CP content did not change when the mixture was harvested for silage after a longer interval (González, 1992). The application of 80 kg ha/ha, which is the highest rate recommended for silage, can give a yield up to 6 t/ha of DM with a crude protein content of 15 percent and a content on acid detergent fibre of 27 percent. This leads to the disappearance of white clover if repeated for two to three years in the same field (González, 1993b).

Seasonal distribution of nitrogen

Most of the studies in the seventies and eighties involved the comparison between grass alone with grass-clover swards in different sites with a range of yields that go from 2-4 to 12-14 t/ha/year. Grass-clover swards always produced more than grass alone swards and gave a less uneven seasonal distribution of the yield. In most situations, where N applications of around 250 kg ha/ha or over were used, the clover disappeared (González, 1986, 1992).

The response to N is higher in the spring than in the autumn, both in lowlands near the coast and in hill areas inland (González, 1994). Nitrogen is not recommended in summer and winter because there is no growth due to the drought and low temperatures, respectively. An exception can be made with Italian ryegrass that responds to nitrogen applied in the winter, overall on the coast.

REFERENCES

AEA, 1998. Anuario de Estatística Agraria 1997. Consellería de Agricultura, Gandería e Montes. Xunta de Galicia. Santiago de Compostela (Spain).

Barbeyto Nistal, F. 1997. Explotacións de vacún de leite en Galicia. Manexo técnico e resultados económicos. Año 1996. Consellería de Agricultura, Gandería e Montes. Santiago de Compostela (Spain).

Blanco. 1945. Los fleos. Sección de Publicaciones, Prensa y Propaganda. Ministerio de Agricultura. Madrid (Spain).

Eguileor, R. 1917. Establecimiento de prados. 74 pp. Servicio Agronómico. Sección de Lugo. Lugo (Spain).

Estevez Feijoo, E. 1995. O sector productivo do leite en Galicia. Cadernos da Area de Ciencias Agrarias, 11-54. Publicacións do Seminario de Estudos Galegos.

Fernandez, G. 1987. La agricultura gallega en 1986. Caja Rural Provincial de Orense. Servidio de Estudios, 728 pp. Orense (Spain).

Flores, G., Gastro, J. & Gonzalez Arraez, A. 1990. O ensilado nas explotacións leiteiras de Galicia. Memoria del Centro de Investigacións Agrarias de Mabegondo del año 1990, 188-194.

Garcia Pita, P. 1993. La población activa agraria: la realidad de un modelo social de economía familiar. El campo, 127, 231-240. Revista de Información Agraria del Banco de Bilbao Vizcaya.

Gonzalez Pizarro, J. de P. 1985. Plantas pratenses: Las alfalfas y los tréboles. Estudio agrícola-bromatológico, 226 pp. Santiago (Spain).

Gonzalez Rodriguez, A. 1986. El trébol blanco y el fertilizante nitrogenado como fuentes de nitrógeno para la pradera. Actas de la XXVI Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 1, 265-279.

Gonzalez Rodriguez, A. 1991. Efecto del manejo del primer año y de la aplicación de nitrógeno sobre la producción herbácea de una pradera mixta establecida en terrenos de monte. Investigación Agraria. Producción y Sanidad animales, 6(1), 9-20.

Gonzalez Rodriguez, A. 1992. Efecto de la aplicación de nitrógeno sobre el contenido de proteína bruta de praderas con y sin trébol blanco. Actas de la XXXII Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 190-193.

Gonzalez Rodriguez, A. 1993a. The response to nitrogen and potassium fertilizer in grass white clover sward in North West Spain. In: White Clover in Europe: State-of-the-Art, 105-107. REUR Techn. Ser. 29. FAO. Rome (Italy).

Gonzalez Rodriguez, A. 1993b. Influencia del nitrógeno y manejo sobre el crecimiento y la calidad de la pradera en primavera. Actas de la XXXIII Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 253-261.

Gonzalez Rodriguez, A. 1994. Respuesta a la aplicación de nitrógeno en una pradera de gramínea y trébol blanco para corte de ensilado. Actas de la XXXIV Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 249-252.

Gonzalez Rodriguez, A. 1996. Aplicación del primer nitrógeno anual para pastoreo temprano. Actas de la XXXVI Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 257-262.

Herbabdez Robredo, L. 1915. Instrucciones sobre prados. 12 pp. Granja Escuela Práctica de Agricultura Regional. La Coruña (Spain).

Hernandez Robredo, L. 1933. Pastos y Prados. 230 pp. Biblioteca Agropecuaria. Marín y G. Campo, S.L. Editores. Madrid (Spain).

IGE (Instituto Galego de Estatistica). 1998. Galicia en cifras. Anuario 1997. Instituto Galego de Estadística, 436 pp. Xunta de Galicia. Consellería de Economía e Facenda. Santiago de Compostela (Spain).

Lloveras, J. 1982. Cultivos y rotaciones de cultivos. El Campo, 86, 85-89.

Lopez Garrido, C., Flores Calvete, G., Barbeyto Nistal, F. & Lois Mosquera, M. 1998. Costes de producción de leche en Galicia. Pastos (in press).

Mombiela Muruzabal, F. 1983. Estudio de la fertilidad del suelo en Galicia. Apuntes históricos y problemática general de la investigación sobre la acidez y falta de fósforo. Cuadernos da Area de Ciencias Agrarias, 4, 75-118.

Mosquera, R. & Gonzalez Rodriguez, A. 1992. Efecto de la dosis y tiempo del primer nitrógeno anual a una pradera de raigrás y trébol. Actas de la XXXII Reunión Científica de la Sociedad Española para el Estudio de los Pastos, 185-189.

Piñeiro Andion, J. & Perez Fernandez, M. 1993. Mezclas pratenses para la España húmedad. 48 pp. Hojas Divulgadoras nº 8/92 HD. Secretaría General de Estructuras Agrarias. Ministerio de Agricultura, Pesca y Alimentación. Madrid (Spain).

Piñeiro Andion, J., Gonzalez Arraez, E. & Perez Fernandez, M. 1977. Acción del fósforo, potasio y cal en el establecimiento de praderas en terrenos procedentes de monte. In: III Seminario INIA/SEA sobre Pastos, Forrajes y Producción Animal, 53-83. Centro de Investigación y Desarrollo Agrario de Galicia. Instituto Nacional de Investigaciones Agrarias. Mabegondo. A Coruña (Spain).

Posada, C. 1990. El sector productor de leche en Galicia. Revista de estudios agrosociales, 154, 153-183.

Sineiro Garcia, F. 1977. Técnicas de transformación del monte en pasto. In: III Seminario INIA/SEA sobre Pastos, Forrajes y Producción Animal, 7-52. Centro de Investigación y Desarrollo Agrario de Galicia. Instituto Nacional de Investigaciones Agrarias. Mabegondo. A Coruña (Spain).

Yepes, V. 1961. Plan forrajero para Galicia. II Reunión Científica de la Sociedad Española para el Estudio de los Pastos.

Yepex, V. & Perez, M. 1971. Avance-resumen de las experiencias con especies de gramíneas y leguminosas pratenses. Pastos. 1(1), 38-52.

Zulueta, J. 1971. Variedades de gramíneas para los pastizales forestales de Galicia. Resultados del transcurso del primer invierno. Pastos, 1(1).

EFFECTS OF DROUGHT STRESS AND WINTER STRESS ON THE PERSISTENCE OF WHITE CLOVER - R.P. Collins

Legume Breeding Group, IGER, Plas Gogerddan, Aberystwyth, SY23 3EB, United Kingdom

SUMMARY

This paper discusses two of the main environmental constraints limiting the persistence of white clover in pastures worldwide, namely drought stress and winter stress. Persistence is defined as the maintenance of long-term agronomic yield and is considered in this paper to be mainly a function of stolon density. Consequently, the paper deals with the physiological and morphological effects of drought stress and winter stress on the vegetative persistence of the species.

As a crop, white clover tends to be excluded from areas subject to severe drought. In areas with a distinct dry season, indigenous populations can escape drought by adopting a summer-dormant life cycle and re-establishing themselves from seed. In temperate climates periods of drought can be particularly damaging to crop productivity, as they occur on an irregular basis. This limits the time available for plant adaptation to the dry conditions. The effects of drought stress on white clover are discussed under the following headings: (i) timing of crop development; (ii) efficiency of the root system in extracting and transporting water; (iii) effectiveness of the shoot in controlling transpiration; and (iv) ability of the plant to tolerate desiccation. Possible approaches to the development of improved drought tolerance in white clover are discussed within each section.

The ability to overwinter successfully is one of the most desirable attributes for perennial forage legumes, such as white clover and in temperate climates and is essential for survival in more marginal, northern areas of Europe and North America. Winter survival in the species is discussed under the following headings: (i) cold tolerance; (ii) cold resistance; (iii) adaptation; and (iv) acclimation.

There is at present a lack of real understanding of how the mechanisms of drought and cold tolerance operate at the molecular level. However, advances in molecular biology should result in improved knowledge of the basis of plant adaptation to the environment and of the links between specific biochemical changes and plant phenotype.

Keywords: drought stress, persistence, plant breeding, white clover, winter stress

ORIGINS AND RANGE OF ADAPTATION OF WHITE CLOVER

White clover (Trifolium repens) is the most agriculturally important pasture legume in temperate zones (Frame and Newbould, 1986). Its precise centre of origin is debated, but it has become widely adapted to regions from the Arctic to the subtropics and has even been reported at altitudes up to 6 000 m in the Himalayas (Williams, 1987a). Following human migration the species has become distributed worldwide in suitable climates due to its utilization in pastures. It is now cultivated chiefly in Northern Europe, North America, New Zealand and in humid coastal areas of Australia (Frame and Newbould, 1986; Williams, 1987a).

Although white clover has become adapted to a wide range of climatic conditions, certain morphological and physiological constraints limit its performance. Thus, the extension of the species into cool climates is restricted by its relative lack of cold tolerance, whilst its cultivation is not recommended in drought-prone areas (Turkington and Burdon, 1983). The basis of these constraints will be discussed in this paper.

PERSISTENCE OF WHITE CLOVER

White clover populations adapted to regular climatic stress, such as drought, tend to adopt an annual strategy and reproduce from seed (Harris, 1987). In moist, temperate climates, however, seedling recruitment in pastures appears to be of relatively minor importance (Williams, 1987a). Under favourable environmental conditions, therefore, the persistence of the species is primarily dependent on vegetative propagation through the spread of a network of stolons (Hollowell, 1966). Persistence, which has been defined as the maintenance of long-term agronomic yield (Annicchiarico, 1997), is thus regarded as a function of stolon density (Williams, 1987a) and comprises the following elements: (i) requirement for continual renewal of the stolon population by asexual reproduction; (ii) requirement for the survival of stolon roots; and (iii) ability to produce seed when environmental stress limits the capacity for asexual reproduction (Williams, 1987b). The emphasis in this paper is on the effects of drought stress and winter stress on the vegetative persistence of white clover. Environmental effects on the reproductive development of the species have been comprehensively reviewed by Thomas (1987).

It should be emphasized that the terms “persistence” and “yield” are not synonymous in the case of white clover. There was, until the development of modern cultivars, a negative correlation between persistence and short-term agronomic yield (i.e. the amount of leaf material produced) (Rhodes and Harris, 1979; Annicchiarico, 1997). This is because large-leaved cultivars, which produce high yields under appropriate management, generally have lower stolon densities than small-leaved cultivars, which are characterized by having a dense network of thin stolons (Rhodes et al., 1994). However, in cultivars recently produced in the UK and New Zealand, this negative relationship between persistence and short-term yield has been weakened and material with increased stolon density has been developed without sacrificing leaf size (Rhodes et al., 1994; Woodfield and Caradus, 1996).

DROUGHT STRESS

Of all the environmental variables affecting plant growth and development, drought stress is one of the most important (Parsons, 1982). There are numerous reports in the literature showing that drought stress limits crop yields; the response varies with the severity, duration and timing of this stress (Turner and Begg, 1978). In pasture plants, morphological responses such as reduction in leaf area, tillering and root growth are more sensitive to water stress than are physiological processes such as photosynthesis, stomatal behaviour and translocation (Parsons, 1982). For this reason, pasture plants such as white clover, whose agronomic yield consists of vegetative growth, are considered to be more susceptible to drought stress than cereals and grain legumes (Turner and Begg, 1978). In this section of the paper the morphological and physiological responses of white clover to drought stress are considered, with particular emphasis on approaches to the development of conventional cultivars or interspecific hybrids that can avoid or tolerate periods of drought.

Any discussion of the responses of plants to drought stress must, however, also consider soil and climatic factors, as it is evident that the type of drought resistance/avoidance mechanisms developed by plants will be partly determined by the environment in which they grow (Quizenberry, 1982). Many places in the world are subject to drought (defined as a period of time during which plant and/or soil water deficiencies affect the growth of plants), but the duration and intensity vary greatly between climatic zones (Quizenberry, 1982). Drought occurs via interactions between physical soil factors (e.g. soil type and depth) and various components of climate (e.g. variability in yearly precipitation, seasonal precipitation patterns, relative humidities, average temperatures etc.) (Quizenberry, 1982; Frame and Newbould, 1986).

Quizenberry (1982) identified three types of moisture environment associated with drought stress in plants: 'stored', 'variable' and 'optimal'. As a crop, white clover tends to be excluded from areas subject to severe drought (Foulds, 1978). For agricultural purposes white clover would chiefly be used in variable moisture environments, where crops are grown during the part of the year when precipitation is most likely to occur, or in optimal moisture environments, in which crops are grown in conditions of adequate soil moisture during most of the growing season, but in which occasional periods of drought occur (Quizenberry, 1982). In variable moisture environments, such as Mediterranean areas, indigenous white clover populations can escape drought by being winter-active and exhibiting summer dormancy (Williams, 1987a). Drought escape is related to the ability of the plant to flower early and set seed in response to dry summer conditions. Subsequently, the population can re-establish itself from soil seed reserves in the autumn (Caradus, 1989). This strategy is discussed further in section (i) below. Variable moisture environments can, of course, be transformed by irrigation, with cultivation shifted from winter-spring to spring-summer (Buddenhagen, 1990).

Optimal moisture environments (which would include the temperate climates in which white clover is chiefly cultivated) are characterized by short, irregular periods of drought, the imposition of which can be particularly damaging to crop productivity because of the inadequate amount of time available for plants to adjust to the dry conditions (Quizenberry, 1982). In this environment soil type becomes an important factor determining water availability to plants, with clay soils remaining wetter for longer than sandy soils and consequently allowing the process of drought acclimation to occur (Quizenberry, 1982).

Wilson (1981) categorized the morphological/physiological factors influencing crop water relations as follows: (i) timing of crop development; (ii) efficiency of the root system in extracting and transporting water; (iii) effectiveness of the shoot system in controlling transpiration; and (iv) ability of plants to tolerate desiccation. These factors are discussed in relation to white clover in the following sections.

Timing of crop development

White clover populations either indigenous to or naturalized in very drought-prone areas may escape drought by behaving as winter annuals. An example of this type of strategy is given by O'Brien (1970), who described a situation in which natural selection had acted on the genetic variation within naturalized white clover populations growing in the sub-tropical east coast of Australia to produce adapted ecotypes with heavy seeding capabilities. This region has a long, hot summer period and a shorter, cool winter period with substantial rainfall and the populations maintain themselves through the mechanism of re-seeding prior to the onset of drought stress. Selection for this strategy has been carried out on white clover in New Zealand, with favourable results in very drought-prone areas (Caradus, 1989).

Roots

When considering the effect of root system type on plant persistence in conditions of water stress, it is important to take into account both the timing of soil water deficits and the use of stored soil water in relation to crop phenology (Clarke and McCaig, 1993). Thus, where soil water reserves exist at depth, the ability of plants to produce deep, extensive root systems is likely to be advantageous (Wilson, 1981). Conversely, where moisture reserves are confined to the upper layers of soil, then rooting depth becomes less important than the ability of plants to produce an efficient extraction system in these surface layers (Wilson, 1981).

The root system of white clover consists of a taproot that develops from the seedling primary root and nodal roots that develop from stolon nodes (Caradus, 1990). The seedling taproot generally persists for 12 to 18 months, depending on environmental conditions (Caradus, 1990). The structure of white clover root systems differs between populations, varying from types with a high proportion of deep, vertically penetrating roots to those with a more shallow, fibrous form (Caradus, 1981). Although depth of rooting depends on soil type, cultivar and management, white clover is generally considered to be a shallow-rooted species, with most roots distributed in the top 10-20 cm of soil (Caradus, 1990). This distribution pattern is likely to be detrimental to the ability of the species to persist under drought. In addition, there is evidence that the growth of white clover root systems per se is adversely affected by drought conditions. For example, the production of new roots of white clover has been found to be greatly reduced under drought stress (Thomas, 1984; Stevenson and Laidlaw, 1985). Studies by Foulds (1978) comparing root system development in seedlings of three legume species, T. repens, Lotus corniculatus and Medicago lupulina, under moist and droughted experimental conditions, found that the depth of penetration of the root system of white clover was considerably more reduced by drought than that of the other species. Consequently, in deep, moist soils the shallow root system of white clover was found to be adequate to sustain vegetative growth and seed production but in conditions of drought an inability to penetrate the soil profile effectively was found not only to reduce dry matter yield but also to lead to lower seed production per plant and ultimately, to plant mortality (Foulds, 1978).

It is likely that recovery from drought would also be impeded in white clover due to the sensitivity of the process of root initiation to drought stress. Stevenson and Laidlaw (1985) described a typical scenario for white clover subjected to drought in an “optimal” moisture environment, in which soil dries from the surface downwards, which would result in a decline in the production of new roots. In the initial stages of drought, older roots would be able to supply the stolons with sufficient water, via acropetal movement within the stolon, to continue growth. However, if the drought continued for long enough, these roots would become progressively more drought stressed, resulting in stolon death. During recovery from drought the soil surface would increase in water content faster than the deeper soil layers, creating a suitably moist environment for root initiation. Stevenson and Laidlaw (1985) however, considered that root formation would be unlikely to occur at old nodes on white clover plants previously subjected to drought stress. As a result, root initiation would be temporarily greatly reduced.

In white clover, significant genetic variation has been observed for many root characters (Caradus, 1990), some of which might be expected to confer improved drought tolerance. Ennos (1985), for example, found genetic variation in seedling root system depth and a high heritablity for this character, within a collected population of white clover. These differences were related to dry matter yield under drought conditions, with genotypes from the 'long' selection group producing higher yields than those with shorter roots. Caradus (1981), however, considered that meaningful differentiation between types of root system cannot be obtained until various root traits have stabilized, which may require screening for up to two years after seedling germination.

Differences in white clover root system characteristics have been related to the performance of cultivars and ecotypes under drought stress (Caradus, 1981; Thomas, 1984; Woodfield and Caradus, 1987). Caradus (1981) compared the performance over a three year period of three cultivars with a range of root system types: Ladino (taprooted type i.e. with a high proportion of vertically penetrating roots), Kent Wild White (fibrous rooted type) and Grasslands Huia (intermediate type). Ladino, which had the greatest number of root tips at depth, was less visibly affected by drought stress than the other cultivars. Conversely, Annicchiarico and Piano (1998) found no evidence that a more developed root system contributed to improved forage yield of selections from within Ladino populations under drought. One explanation for this observation may be that strong correlations exist between root and shoot type in white clover, so that taprooted, Ladino-type populations tend to be large-leaved, with thick stolons and a low stolon density, whereas types with a more fibrous root system generally also have small leaves and a dense stolon network (Woodfield and Caradus, 1987). The existence of this correlation may result in a conflict, when considering selection for drought tolerance, between the improved water uptake associated with the deeper and more extensive root system of the taprooted type and the superior water conservation usually associated with small leaf size (Woodfield and Caradus, 1987; Annicchiarico and Piano, 1998). However, screening of a large number of genotypes of varied origin as part of a breeding programme in New Zealand has demonstrated the existence of some material with a larger proportion of taproot than expected from its leaf size (Caradus and Woodfield, 1986; Woodfield and Caradus, 1987), which suggests that selection for improved drought tolerance using smaller-leaved, more tap-rooted genotypes is a feasible breeding strategy.

Another approach to the improvement of drought tolerance in white clover by altering below-ground morphology is the production of hybrids between it and more drought tolerant species within the genus Trifolium. An example of this is the work of Meredith et al. (1995), populations from crosses between T. repens and T. ambiguum, who produced F₁, BC₁ and BC₂ an extremely persistent, rhizomatous, drought tolerant species (Coolbear et al., 1994). Further work has been carried out in this programme, resulting in the recent production of a BC₃ generation containing some of the rhizomatous growth habit of T. ambiguum. The amount of rhizome necessary to confer improved drought tolerance has not yet been determined, but populations to select plants for there appears to be sufficient variation within the BC₂ and BC₃ future experimentation (Michaelson-Yeates et al., 1997).

Control of transpiration

The rate at which soil water reserves are depleted will determine the time of onset and rate of intensification of drought (Wilson, 1981). Crop characteristics influencing this process are those relating to water uptake by the roots (discussed above) and water loss due to leaf transpiration. Numerous studies have shown that white clover does not control transpiration efficiently (Hart, 1987). The species appears to have poor stomatal control of leaf water loss, due to incomplete stomatal closure even under conditions of low leaf turgor (Turner, 1990a). Low cuticular resistance in white clover may also contribute to high rates of leaf water loss (Sánchez-Díaz and Sánchez-Marin, 1974). White clover cultivars have been shown to differ in their rates of transpiration, so that soil water deficits measured during drought were significantly lower with cv. Olwen than with cv. S184 for the same leaf area index (Thomas, 1984).

Anatomical studies comparing vascular tissue in stolons and petioles have suggested that water supply to white clover leaves is restricted by a relatively high hydraulic resistance in the petioles, due to their narrower xylem vessels (Turner, 1990a). The combination of high rates of transpiration and restricted water supply to leaves contributes to the widely-reported observation that white clover leaves wilt and senesce rapidly in dry conditions (Hart, 1987). The early senescence of leaves may be a useful response to drought in that it prevents further transpirational water loss and thus helps to preserve soil water content (Turner, 1990b). Turner (1990b) demonstrated that stolon tips were capable of maintaining turgor under conditions of water stress in pots because of osmotic adjustment (see section (iv) below). In addition, the dimensions of xylem vessels in stolons are such that they favour water flow to the stolon tips, rather than to leaves (Turner, 1990a).

There is some disagreement in the literature as to whether stolons or leaves show the greatest response to drought. Thomas (1984) found that drought had little effect on stolon number or length, but that leaf growth was affected. Turner (1991) reported that the absolute growth rate of stolons subjected to drought was consistently less than that of leaves and petioles. However, when the results of Turner (1991) were expressed in terms of yield, stolons were found to be less sensitive than leaves. The implication of these findings is that in drought conditions stolons stop growing, thereby preventing the production of new leaves. In addition, leaves on stressed stolons tend to wilt and die, whilst the stolons themselves are able to maintain turgor and survive for longer (Turner, 1991). The relative contributions of decreases in stolon initiation, number of leaves per stolon and individual leaf area to the reduction of total plant leaf area under drought have been calculated by Belaygue et al. (1996). These authors concluded that the size of the relative reductions in each component depended on the duration and intensity of the water deficit.

Desiccation tolerance

Precise mechanisms affecting tolerance, as opposed to avoidance, of drought stress in plants are as yet unclear (Wilson, 1981; Turner, 1990b). One of the physiological changes occurring in plants exposed to drought stress is the accumulation of solutes, either passively (due to a concentration effect caused by dehydration) or actively (due to the process of osmotic adjustment) (Parsons, 1982; Turner, 1990b). Osmotic adjustment is one way of maintaining cell turgor as water potential is reduced, allowing survival during periods of drought stress (Turner, 1990b). In addition, changes in cell wall elasticity may be involved in the maintenance of turgor (Parsons, 1982). Osmotic adjustment is considered to be of potential value in breeding for drought tolerance, as it does not 'cost' the plant greatly and does not affect photosynthesis as much as stomatal closure would (Parsons, 1982). Osmotic adjustment is sensitive to the rate at which drought stress develops and is greater when stress develops gradually (Morgan, 1984).

There is some evidence that osmotic adjustment occurs in white clover (Turner, 1990b), although only one cultivar, Olwen, was used in this study. Stolon tips were found to survive the levels of drought stress imposed in this experiment by maintaining turgor, whereas leaves wilted and died. High levels of proline, a substance frequently implicated in the response of plants to stress (Castonguay et al., 1997), were measured in these stolon tips, but also in the wilted leaves. The use of proline accumulation in white clover as a selection criterion in breeding for drought tolerance would therefore seem inadvisable (Turner, 1990b). Turner (1990b) suggested that the sugar alcohol pinitol, a substance found particularly in legumes, may be a more likely candidate to explain the difference in drought response between stolons and leaves, but offered no experimental evidence for this.

The way in which plants respond to drought stress is a complex topic. Understanding the response to drought of white clover, a species which interacts with a number of biotic and edaphic factors, requires consideration of many elements. In addition to those discussed above, other factors that have an impact on white clover's response to drought include cultivar leaf size, the frequency and height of defoliation, the presence of a companion grass and the nutrient status of the soil (Thomas, 1984). The effects of drought on nitrogen fixation in white clover have been discussed by Crush (1987).

WINTER STRESS

The ability to survive over winter is one of the most desirable attributes for perennial forage legumes, such as white clover, in temperate climates (Rhodes et al., 1994, Viiralt and Higuchi, 1996) and is essential for cultivars intended for use in marginal, northern areas of Europe (Junttila et al., 1990) and North America (Klebesadel, 1986). It is considered to be somewhat less important in white clover breeding programmes in New Zealand (Caradus, 1989). Maintenance of a network of live stolon during the winter is essential for vigorous spring growth (Harris et al., 1983), which is an important component of white clover's competitive ability (Collins et al., 1991).

Winter hardiness can be regarded as the ability of overwintering plants to be persistent under a range of winter stresses such as snow cover, attack by fungal pathogens, freezing damage and encasement by ice, but low temperature per se is often the dominant factor (McHugh et al., 1988). In temperate climates, overwintering pastures are periodically exposed to potentially damaging low temperatures (Fuller and Eagles, 1978). The critical temperature at which winter damage occurs varies both between and within species, but also depends on the degree of hardening carried out by plants (McHugh et al., 1988; Svenning et al., 1997). Thus, cold tolerance is a complex characteristic, which, although genetically determined, also exhibits considerable phenotypic plasticity (McHugh et al., 1988).

In this section of the paper the winter survival of white clover will be discussed using the definitions of Castonguay and Guckert (1996).

Cold tolerance

Cold tolerance has been defined as the constitutive ability of a plant to survive exposure to low temperature, frost or chilling with little or no injury (Castonguay and Guckert, 1996). For the reason discussed above, cold tolerance is a vital component of winter hardiness. Artificial screening tests have been widely used to assess the cold tolerance of white clover populations. In most of these tests plants have been directly exposed to low temperatures and the resulting damage has been evaluated (e.g. Caradus et al., 1989; Junttila et al., 1990; Eagles et al. values, 1994; Collins and Rhodes, 1995). Results are frequently presented as LT₅₀ (e.g. the temperature at which 50 percent mortality occurs in the sample population) calculated from transformed data (e.g. Collins and Rhodes, 1995). Tests involving the freezing of values appear to provide a segments of apical stolon and subsequent estimation of their LT₅₀ reasonable correlation with the winter survival rankings of populations in field plots (Junttila et al., 1990; Collins and Rhodes, 1995), although the LT₅₀ temperature itself does not necessarily relate to temperatures experienced in the field. Not surprisingly, studies of this nature have demonstrated the existence of interactions between population and freezing temperature for percentage of stolon apices damaged or killed. This observation reflects (a) the existence of intrinsic differences in cold tolerance between white clover populations, related to either their climatic origin, or their morphology, or both (e.g.genetic differences); and (b) differences in the extent to which populations have acclimatized to the experimental conditions (e.g. differences in phenotypic plasticity).

White clover populations of contrasting cold tolerance often differ morphologically. Thus, Caradus et al. (1989) found that large-leaved, erect cultivars tended to be less cold tolerant than small-leaved, prostrate ones. Similarly, Svenning et al. (1997) reported that the most cold tolerant population in their study was characterized by small leaf size, thin stolons and short internodes. However, in the study of Collins and Rhodes (1995), the most cold tolerant population also had the highest stolon weight/unit length, an indirect measure of stolon diameter. Clearly, caution is necessary before drawing conclusions about populations' cold tolerance from their morphology, since the latter may interact strongly with climatic origin (Eagles and Othman, 1988). As a general rule (with some exceptions), large-leaved populations with thick stolons are of Mediterranean origin, and thus lack intrinsic cold tolerance, whereas those from northern locations are small-leaved with thin stolons (Williams, 1987a). The influence of geographic origin is discussed further in section (iii).

Cold resistance

Cold resistanceis the ability of a plant to actively develop mechanisms of protection against the negative effects induced by low temperatures. Knowledge of the exact nature of the damage occurring at low temperatures and the adaptations required to maximize low temperature tolerance are considered to be fundamental to our understanding of forage legume survival in cold climates (Castonguay and Guckert, 1996). Unfortunately, although the precise mechanism of cold resistance in plants has been the subject of much research, it is not yet understood in detail. Production of “cryoprotective” compounds e.g. soluble carbohydrates, sugar alcohols and proline, may be one mechanism through which cold resistance is attained in some species, including white clover (Yelenosky and Guy, 1989; Røsnes et al., 1993; Svenning et al., 1997), although experimental evidence from other species has sometimes shown no direct link between solute accumulation and increased cold tolerance (Pollock et al., 1988; Thomas and James, 1993). Accumulation of cryoprotective compounds has been observed in white clover during winter (Røsnes et al., 1993; Turner and Pollock, 1998). This process is thought to contribute to cold resistance in two ways: (a) by directly causing depression of the freezing point and reduction of the osmotic potential of cell sap (Pollock et al., 1988); and (b) by stabilizing cell membranes and proteins during freeze-induced cell dehydration (Castonguay and Guckert, 1996).

In white clover, the synthesis and efficient use of starch reserves are considered to be important determinants of winter hardiness (Harris et al., 1983; Collins and Rhodes, 1995; Turner and Pollock, 1998). Turner and Pollock (1998) describe the functions of starch reserves in white clover as, firstly, to supply energy for basal metabolism if current rates of photosynthesis are low, and, secondly, to supply sucrose and other sugars for cryoprotection as described above. Cultivar differences in the starch content of white clover stolons during autumn and winter have been identified by many researchers, but it may be difficult to relate these differences directly to levels of cold tolerance because starch metabolism is altered in any case by the continued growth of some cultivars during the winter (Turner and Pollock, 1998). Volenec et al. (1996) have described the existence of inconsistencies between plant carbohydrate content and stress tolerance as an example of the difficulties of extrapolating from correlation to causation and have suggested a role for plant nitrogen reserves (in the form of 'vegetative storage proteins') in the stress tolerance of forage species. Clearly, more research is required before we fully understand the processes involved in cold resistance in white clover.

Adaptation

Adaptation has been defined as a response to long-term changes resulting in heritable genetic alterations which are stable and long lasting within the population. The variability and high degree of outbreeding in white clover strongly favour genetic shifts in response to natural or artificial selection pressures (Gibson and Hollowell, 1966). Natural selection for winter survival is effective through the elimination of insufficiently cold tolerant genotypes by winter kill (Klebesadel, 1986). This process may occur slowly in perennial plant populations comprising mainly cold tolerant genotypes, but genetic shift is likely to be more rapid in populations that are marginally cold tolerant and where mortality is high (Klebesadel, 1986). In addition, genetic shifts would be accelerated if a perennial species were to adopt an annual strategy and rely to a greater extent on regeneration from seed (Klebesadel, 1986). Gibson and Hollowell (1966) considered that white clover in cold, northern areas of the USA behaves mostly as a biennial or short-lived perennial, adopting the usual perennial strategy in more favourable, southern areas. Rhodes (pers. comm.) also observed that white clover populations at high altitudes in the Swiss Alps were more profusely flowering than would be expected for their leaf size, indicating a greater reliance in this material on seedling recruitment under stressful environmental conditions.

Experimental evidence supporting the hypothesis that natural selection for winter hardiness may occur relatively quickly under suitable conditions is provided by Swedish research carried out in the 1930s using German and Danish white clover cultivars grown in Sweden for two years before being harvested for seed (Williams, 1987a). After only one generation of seed production the cultivars under selection were reported to produce considerably higher yields than the original stocks at this site. Winter hardiness was considered to be the dominant character undergoing natural selection. A pan-European research project is currently in operation under the auspices of the EU COST 814 Programme entitled “Overwintering and Spring Growth of White Clover”, in which the possibility of the occurrence of genetic shift for winter hardiness in two cultivars grown at a range of locations throughout Europe is one of the topics being investigated (Collins et al., 1996).

Considerable genetic variation exists within white clover for winter hardiness, the pattern of which can be related to gradations in temperature (Williams, 1987a). Within Europe, for example, populations of Mediterranean origin are often winter active and lacking in cold tolerance, whereas northern populations tend to be winter dormant and cold tolerant (Williams, 1987a). However, localized climatic conditions at the site of origin may have a considerable influence on the physiological characteristics of white clover populations (Ollerenshaw and Haycock, 1984; Klebesadel, 1986; Junttila et al., 1990), so that growth at low temperatures and cold tolerance are not invariably negatively correlated. Sufficient genetic variation exists within populations in northern latitudes to enable selection to be carried out for the agronomically desirable combination of adequate cold tolerance and active growth at low, positive temperatures (Ollerenshaw and Haycock, 1984; Rhodes et al., 1994; Svenning et al., 1997). The interaction of factors such as altitude, snow cover and degree of exposure to dehydrating winds provides an array of microenvironments for populations to become adapted to in sites at similar latitudes. Consequently, the existence of populations from the same latitude exhibiting contrasting physiological responses to the main environmental factors which influence the survival strategies of plants during winter, namely daylength and temperature, is not unexpected (e.g. Boller and Nösberger, 1983; Rhodes and Collins, 1993).

Acclimation

Acclimation is a response induced by environmental changes that results in a phenotypic alteration over a single generation time, without any compositional change in the genetic complement. The successful winter survival of cultivars of white clover in cold climates is dependent not only on their intrinsic cold tolerance but also on their ability to acclimate to low temperatures. This ability is termed “cold hardening” and has been investigated in white clover by many researchers (e.g. Junttila et al., 1990; Røsnes et al., 1993; Sandli et al., 1993; Collins and Rhodes, 1995; Guinchard et al., 1997; Svenning et al., 1997; Turner and Pollock, 1998). The timing of cold hardening/dehardening (e.g. the stability of cold tolerance during the winter) has important consequences for winter survival. White clover populations appear to differ substantially in the temperatures at which these processes occur (Røsnes et al., 1993; Eagles et al., 1994; Svenning et al., 1997). A comparison of the rate of cold hardening in four white clover populations by Svenning et al. (1997) showed that it was fastest in the population from the most northerly location. In all populations examined in that study, cold hardening was most rapid in the first week of treatment in a controlled environment, with only minor increases in subsequent weeks. Rates of dehardening were also investigated in the same populations and there were indications of the existence of interaction between temperature and population for this process. A quantitative analysis of rates of cold hardening/ dehardening similar to that carried out by Gay and Eagles (1991) for Lolium spp. is required for contrasting white clover populations in order to compare the kinetics of these processes in a mathematically rigorous way.

In addition to temperature, daylength plays an important role in determining rates of cold hardening/dehardening and again, there is some evidence of variation between populations in their response to this factor (Eagles et al., 1994). In a comparison between the cultivars AberCrest, AberHerald and Grasslands Huia after hardening for 14 days at 2^°C and then exposure to higher temperatures, it was found that AberCrest maintained its level of cold tolerance in a short (8 hours) daylength treatment, whereas the other cultivars dehardened. In a longer daylength treatment, AberCrest showed some dehardening, but was still more cold tolerant than the other material (Eagles et al., 1994). This photoperiodic response of AberCrest is consistent with its observed lack of growth during the short days of a north European winter (Rhodes and Collins, 1993).

As mentioned above, growth at low temperatures and cold tolerance are often, though not always, negatively correlated. Cessation of growth at low temperatures and the onset of cold hardening result in an accumulation of soluble carbohydrates due to a reduction in assimilate utilization (Pollock et al., 1988). The time course of cold hardening and the accumulation of carbohydrates and other cryoprotective compounds have been analysed in white clover by Røsnes et al. (1993) and Guinchard et al. (1997), and significant increases in sugars were found in stolons and leaves during hardening. Svenning et al. (1997) reported a positive correlation between stolon sucrose content and cold tolerance during both the hardening and dehardening of a Norwegian population and the UK cultivar AberHerald. However, Turner and Pollock (1998) observed sucrose accumulation in a range of cultivars during autumn in the UK at temperatures too high to cause cold hardening. These authors proposed that sucrose accumulation may not represent a hardening response but is, rather, a consequence of changes in the ratio of supply to demand of assimilate resulting from the changes in temperature, daylength and irradiance which occur during the onset of winter. Levels of putative cryoprotective compounds such as pinitol and proline have also been shown to increase greatly during the cold hardening process (Røsnes et al., 1993; Sandli et al., 1993; Svenning et al, 1997; Turner and Pollock, 1998). However, it may not be possible to link this accumulation directly to cold hardening, as Sandli et al. (1993) observed an accumulation of proline during hardening at 6^°C, but no further increase when hardening was intensified at 0.5°C.

Caution is required when extrapolating from acclimation studies carried out in controlled environments to field conditions. Guinchard et al. (1997) found that the partitioning of stolon reserves between soluble and insoluble carbohydrates during cold acclimation in controlled environments was greatly influenced by the type of treatment imposed. Thus, stolons exposed to alternating periods of frost and thaw in controlled environments had similar carbohydrate profiles to those sampled in the field, whereas exposure to continuous chilling without frost produced markedly different carbohydrate profiles.

Major biochemical changes evidently occur during cold hardening/dehardening in white clover, however, as in the case of drought resistance, the establishment of precise causal relationships between compounds and processes has so far proved to be elusive.

CO-ACCLIMATION TO COLD AND DROUGHT STRESS

There are certain parallels between cold and drought stress, both in terms of the damage they inflict on plant cells and in terms of plant responses to this damage. The concept of co-acclimation to cold and drought stress has been developed in response to numerous examples where seasonal variation in resistance to one stress is accompanied by increased resistance to other stresses (Steponkus, 1980). However, although it is possible to gather evidence suggestive of a common basis for resistance to many environmental stresses, inconsistencies emerge which tend to diminish the validity of such generalizations (Steponkus, 1980). For example, a study conducted by Thomas and James (1993) on Lolium perenne demonstrated no direct relationship between freezing tolerance and a range of drought-induced expressions of osmotic potential, water soluble carbohydrate, proline and total amino acid contents.

It is generally accepted that cold acclimated tissue has a lower water content at full turgor than non-acclimated tissue (Yelenosky and Guy, 1989). Evidence in support of this has been produced for white clover in a study by Røsnes et al. (1993), in which stolon water content was found to decrease in the initial stages of cold hardening. A relationship between geographic origin and stolon water content was identified in the Norwegian populations used in that experiment, such that the cold tolerant northern material contained less tissue water than the less tolerant southern populations during hardening (Røsnes et al., 1993). However, as the winter progressed, tissue water contents remained static, or even increased slightly, even though the cold tolerance of all populations increased during this period.

Guinchard et al. (1997) concluded that an important consequence of cold acclimation in white clover was the improved ability of the plant to avoid freeze-induced cellular dehydration. However, the authors did not establish a precise relationship between plant water status and cold tolerance as they did not measure changes in cold tolerance during acclimation. The effect of cold acclimation on plant water relations in white clover is clearly a topic that requires detailed investigation.

CONCLUSION

The genetic bases of cold and drought tolerance have been discussed in some detail by Castonguay et al. (1997), who considered that progress in plant breeding in these areas through classical approaches has been relatively slow. Among the difficulties facing the plant breeder are the complex multigenic nature of these traits and a lack of real understanding of the mechanisms of tolerance at the molecular level (Castonguay et al., 1997). However, recent advances in molecular biology and genetic engineering should facilitate the transfer of desirable genes between taxonomically unrelated species (Castonguay et al., 1997), which will complement existing breeding programmes. In addition, these advances will lead to a greater understanding of the molecular basis of plant adaptation to the environment and the links between specific biochemical changes and plant phenotype (Castonguay et al., 1997). Further knowledge of the nature of the stress-regulation of gene expression is necessary before we fully understand the bases of tolerance of the economically important stresses of cold and drought.

REFERENCES

Annicchiarico, P. 1997. Indirect selection for persistence and seed yield in large-leaved white clover. Plant Breeding, 116, 192-194.

Annicchiarico, P. & Piano, E. 1998. Root development and drought tolerance in white clover. Proceedings of 5th. Congress of the European Society for Agronomy, (in press).

Belaygue, C., Wery, J., Cowan, A.A. & Tardieu, F. 1996. Contribution of leaf expansion, rate of leaf appearance, and stolon branching to growth of plant leaf area under water deficit in white clover. Crop Science, 36, 1240-1246.

Boller, B.C. & Nösberger, J. 1983. Effects of temperature and photoperiod on stolon characteristics, dry matter partitioning, and nonstructural carbohydrate concentration of two white clover ecotypes. Crop Science, 23, 1057-1062.

Buddenhagen, I.W. 1990. Legumes in farming systems in Mediterranean climates. In: The role of legumes in the farming systems of the Mediterranean areas (Eds A.E. Osman, M.H. Ibrahim & M.A. Jones). Kluwer, Dordrecht, The Netherlands. pp. 3-29.

Caradus, J.R. 1981. Root growth of white clover (Trifolium repens L.) lines in glass-fronted containers. New Zealand Journal of Agricultural Research, 24, 43-54.

Caradus, J.R. 1989. Improving white clover persistence by breeding. In: Legumes in farming systems (eds P. Plancquaert & R. Haggar). Kluwer, Dordrecht, The Netherlands. pp. 84-92.

Caradus, J.R. 1990. The structure and function of white clover root systems. Advances in Agronomy, 43, 1-46.

Caradus, J.R. & Woodfield, D.R. 1986. Evaluation of root type in white clover genotypes and populations. Proceedings of DSIR Plant Breeding Symposium, 1985. pp. 322-325.

Caradus, J.R., Mackay, A.C. & Van den Bosch, J. 1989. Intraspecific variation for frost hardiness in white clover. Journal of Agricultural Science, Cambridge, 112, 151-157.

Castonguay, Y. & Guckert, A. 1996. Adaptation of forage legumes to cold climates. In: Proceedings of the 16th European Grassland Federation Meeting (Eds G. Parente, J. Frame & S. Orsi). ERSA, Gorizia, Italy. pp. 911-917.

Castonguay, Y., Laberge, S., Nadeau, P. & Vézina, L.-P. 1997. Temperature and drought stress. In: Biotechnology and the improvement of forage legumes (Eds B.D. McKersie & D.C.W. Brown). CAB International, Wallingford, UK. pp.175-202.

Clarke, J.M. & McCaig, T.N. 1993. Breeding for efficient root systems. In: Plant breeding: principles and prospects (Eds M.D. Hayward, N.O. Bosemark & I. Romagosa). Chapman & Hall, London, UK. pp. 485-499.

Collins, R.P. & Rhodes, I. 1995. Stolon characteristics related to winter survival in white clover. Journal of Agricultural Science, Cambridge, 124, 11-16.

Collins, R.P., Glendining, M.J. & Rhodes, I. 1991. The relationships between stolon characteristics, winter survival and annual yields in white clover (Trifolium repens L.). Grass and Forage Science, 46, 51-61.

Collins, R.P., Connolly, J., Fothergill, M., Frankow-Lindberg, B.E., Guckert, A., Guinchard, M.-P., Lüscher, A., Nösberger, J., Rhodes, I., Robin, C., Stäheli, B. & Stoffel, S. 1996. Variation in the overwintering and spring growth of white clover cultivars in the cool, wet areas of Europe. In: Proceedings of the 16^th European Grassland Federation Meeting (Eds G. Parente, J. Frame & S. Orsi). ERSA, Gorizia, Italy. pp. 201-204.

Coolbear, P., Hill, M.J. & Efendi, F. 1994. Relationships between vegetative and reproductive growth in a four year old stand of Caucasian clover (Trifolium ambiguum M. Bieb.) cv. Monaro. Proceedings of the Agronomy Society of New Zealand, 24, 77-82.

Crush, J.R. 1987. Nitrogen fixation. In: White clover (Eds M.J. Baker & W.M. Williams). CAB International, Wallingford, UK. pp. 185-201.

Eagles, C.F. & Othman, O.B. 1988. Variation in growth of overwintered stolons of contrasting white clover populations in response to temperature, photoperiod and spring environment. Annals of Applied Biology, 112, 563-574.

Eagles, C.F., Rhodes, I. & Collins, R.P. 1994. Growth and cold hardiness of white clover populations. In: Proceedings of the 13th Trifolium Conference (Ed. B.R. Christie). pp. 73-74.

Ennos, R.A. 1985. The significance of genetic variation for root growth within a natural population of white clover (Trifolium repens). Journal of Ecology, 73, 615-624.

Foulds, W. 1978. Response to soil moisture supply in three leguminous species. I. Growth, reproduction and mortality. New Phytologist, 80, 535-545.

Frame, J. & Newbould, P. 1986. Agronomy of white clover. Advances in Agronomy, 40, 1-88.

Fuller, M.P. & Eagles, C.F. 1978. A seedling test for cold hardiness in Lolium perenne L. Journal of Agricultural Science, Cambridge, 91, 217-222.

Gay, A.P. & Eagles, C.F. 1991. Quantitative analysis of cold hardening and dehardening in Lolium. Annals of Botany, 67, 339-345.

Gibson, P.B. & Hollowell, E.A. 1966. White clover. US Department of Agriculture Handbook 314. US Government Printing Office, Washington DC, USA.

Guinchard, M.P., Robin, C., Grieu, P. & Guckert, A. 1997. Cold acclimation in white clover subjected to chilling and frost: changes in water and carbohydrate status. European Journal of Agronomy, 6, 225-233.

Harris, W. 1987. Population dynamics and competition. In: White clover (Eds M.J. Baker & W.M. Williams). CAB International, Wallingford, UK. pp. 203-297.

Harris, W., Rhodes, I. & Mee, S.S. 1983. Observations on environmental and genotypic influences on the overwintering of white clover. Journal of Applied Ecology, 20, 609-624.

Hart, A.L. 1987. Physiology. In: White clover (Eds M.J. Baker & W.M. Williams). CAB International, Wallingford, UK. pp. 125-151.

Hollowell, E.A. 1966. White clover, Trifolium repens L. Annual or perennial? In: Proceedings of 10th. International Grassland Congress. pp.184-187.

Junttila, O., Svenning, M.M. & Solheim, B. 1990. Effects of temperature and photoperiod on frost resistance of white clover (Trifolium repens) ecotypes. Physiologia Plantarum, 79, 435-438.

Klebesadel, L.J. 1986. Natural selection may modify introduced white clover toward superior winterhardiness. Agroborealis, 18, 13-19.

McHugh, S., Humphreys, M.O., McNeilly, T. & Johnson, M.S. 1988. Variation in cold tolerance in Festuca rubra L. Grass and Forage Science, 43, 305-312.

Meredith, M.R., Michaelson-Yeates, T.P.T., Ougham, H.J. & Thomas, H. 1995. Trifolium ambiguum as a source of variation in the breeding of white clover. Euphytica, 82, 185-191.

Michaelson-Yeates, T.P.T., Holdbrook-Smith, K., Abberton, M.T., Meredith, M.R. & Marshall, A.H. 1997. Introgression of genes into white clover from related species. In: Seeds of progress. Occasional Symposium of the British Grassland Society No. 31 (Ed. J.R. Weddell). BGS, Reading, UK. pp. 86-87.

Morgan, J.M. 1984. Osmoregulation and water stress in higher plants. Annual Review of Plant Physiology, 35, 299-319.

O'Brien, A.D. 1970. White clover (Trifolium repens L.) in a subtropical environment on the east coast of Australia. Proceedings of 11th. International Grassland Congress. pp.165-168.

Ollerenshaw, J.H. & Haycock, R. 1984. Variation in the low temperature growth and frost tolerance of natural genotypes of Trifolium repens L. from Britain and Norway. Journal of Agricultural Science, Cambridge, 102, 11-21.

Parsons, L.R. 1982. Plant responses to water stress. In: Breeding plants for less favourable environments (Eds M.N. Christiansen & C.F. Lewis). Wiley, New York, USA. pp.175-192.

Pollock, C.J., Eagles, C.F. & Sims, I.M. 1988. Effect of photoperiod and irradiance changes upon development of freezing tolerance and accumulation of soluble carbohydrate in seedlings of Lolium perenne grown at 2^°C. Annals of Botany, 62, 95-100.

Quizenberry, J.E. 1982. Breeding for drought resistance and plant water use efficiency. In: Breeding plants for less favourable environments (Eds M.N. Christiansen & C.F. Lewis). Wiley, New York, USA. pp.193-212.

Rhodes, I. & Collins, R.P. 1993. Temperature and daylength control of spring growth in white clover. In: Crop adaptation to cool, wet climates. Proceedings of COST 814 Workshop (Eds D. Wilson, H. Thomas & K. Pithan). Guyot, Brussels, Belgium. pp.175-180.

Rhodes, I. & Harris, W. 1979. The nature and basis of differences in sward composition of grass-clover mixtures. In: Changes in sward composition and productivity. Occasional Symposium of the British Grassland Society No.10 (Eds A.H. Charles & R.J. Haggar). BGS, Reading, UK. pp. 55-60.

Rhodes, I., Collins, R.P. & Evans, D.R. 1994. Breeding white clover for tolerance to low temperature and grazing stress. Euphytica, 77, 239-242.

Røsnes, K., Junttila, O., Ernsten, A. & Sandli, N. 1993. Development of cold tolerance in white clover (Trifolium repens L.) in relation to carbohydrate and free amino acid content. Acta Agriculturae Scandinavica, 43, 151-155.

Sánchez-Díaz, M. & Sánchez-Marin, M. 1974. Resistance of leaf tissue to desiccation and stomatal closure in lucerne (Medicago sativa L.) and white clover (Trifolium repens L.) in relation to water stress. Annales de Edafologia y Agrobiologia, 33, 743-754.

Sandli, N., Svenning, M.M., Røsnes, K. & Junttila, O. 1993. Effect of nitrogen supply on frost resistance, nitrogen metabolism and carbohydrate content in white clover (Trifolium repens). Physiologia Plantarum, 88, 661-667.

Steponkus, P.L. 1980. A unified concept of stress resistance in plants? In: Genetic engineering of osmoregulation (Eds D.W. Rains, R.C. Valentine & A. Hollaender). Plenum Press, New York, USA. pp. 235-255.

Stevenson, C.A. & Laidlaw, A.S. 1985. The effect of moisture stress on stolon and adventitious root development in white clover (Trifolium repens L.). Plant and Soil, 85, 249-257.

Svenning, M.M., Røsnes, K. & Junttila, O. 1997. Frost tolerance and biochemical changes during hardening and dehardening in contrasting white clover populations. Physiologia Plantarum, 101, 31-37.

Thomas, H. 1984. Effects of drought on growth and competitive ability of perennial ryegrass and white clover. Journal of Applied Ecology, 21, 591-602.

Thomas, H. & James, A.R. 1993. Freezing tolerance and solute changes in contrasting genotypes of Lolium perenne L. acclimated to cold and drought. Annals of Botany, 72, 249-254.

Thomas, R.G. 1987. Reproductive development. In: White clover (Eds M.J. Baker & W.M. Williams). CAB International, Wallingford, UK. pp. 63-123.

Turkington, R. & Burdon, J.J. 1983. The biology of Canadian weEds 57. Trifolium repens L. Canadian Journal of Plant Science, 63, 243-266.

Turner, L.B. 1990a. Water relations of white clover (Trifolium repens): water potential gradients and plant morphology. Annals of Botany, 65, 285-290.

Turner, L.B. 1990b. The extent and pattern of osmotic adjustment in white clover (Trifolium repens L.) during the development of water stress. Ibid., 66, 721-727.

Turner, L.B. 1991. The effect of water stress on the vegetative growth of white clover (Trifolium repens L.): comparison of long-term water deficit and a short-term developing water stress. Journal of Experimental Botany, 42, 311-316.

Turner, L.B. & Pollock, C.J. 1998. Changes in stolon carbohydrates during the winter in four varieties of white clover (Trifolium repens L.) with contrasting hardiness. Annals of Botany, 81, 97-107.

Turner, N.C. & Begg, J.E. 1978. Responses of pasture plants to water deficits. In: Plant relations in pastures (Ed. J.R.Wilson). CSIRO, Melbourne, Australia. pp. 50-66.

Viiralt, R. & Higuchi, S. 1996. Freezing tolerance of white clover (Trifolium repens L.) cultivars. Proceedings of the 14th. Trifolium Conference (Ed. T. Phillips). pp. 33-35.

Volenec, J.J., Ourry, A. & Joern, B.C. 1996. A role for nitrogen reserves in forage regrowth and stress tolerance. Physiologia Plantarum, 97, 185-193.

Williams, W.M. 1987a. Adaptive variation. In: White clover (Eds M.J. Baker & W.M. Williams). CAB International, Wallingford, UK. pp. 299-321.

Williams, W.M. 1987b. Genetics and breeding. Ibid. pp. 343-419.

Wilson, D. 1981. Breeding for morphological and physiological traits. In: Plant breeding II (Ed. K.J. Frey). Iowa State University Press, USA. pp. 233-290.

Woodfield, D.R. & Caradus, J.R. 1987. Adaptation of white clover to moisture stress. Proceedings of the New Zealand Grassland Association, 48, 143-149.

Woodfield, D.R. & Caradus, J.R. 1996. Challenges and future directions in white clover improvement. Proceedings of the 14th. Trifolium Conference (Ed. T. Phillips). p.3.

Yelenosky, G. & Guy, C.L. 1989. Freezing tolerance of citrus, spinach and petunia leaf tissue. Plant Physiology, 89, 444-451.

INTERCROPPING AND BICROPPING WITH WHITE CLOVER - G. Bergkvist^[1] and R.O. Clements^[2]

ABSTRACT

White clover (Trifolium repens L.) is commonly intercropped with cereals as a way to establish pastures and leys, but the aim of intercropping white clover with arable crops can also be to maintain or improve soil fertility, reduce weed problems and/or to decrease the incidence of pests and diseases. White clover that is allowed to become perennial can be interseeded annually with cash crops in a bicropping system. Traits that make white clover attractive as an intercrop in arable systems include its ability to fix nitrogen to the benefit of the subsequent crop, its perenniality and the horizontal growth of the stolons, which in combination with the horizontally arranged leaves makes white clover able to efficiently cover the soil surface while maintaining a low height. The rate of net mineralization of nitrogen from decomposing white clover tissue is high. In northern temperate climates, white clover needs to be undersown to be able to produce any substantial biomass in a cereal:cereal crop sequence. The grain yield of winter cereals and dense crops of spring cereals are not greatly affected by undersown white clover. A negative effect of undersown white clover on the grain yield of maize (Zea mays L.) and of sparse stands of spring cereals can be avoided by keeping the seed rate of the clover low and by delaying the undersowing in relation to the sowing of the cereal. Crops sown after the incorporation of white clover into the soil are likely to benefit from the residual effect of the clover, but much nitrogen can be lost by leaching. If a pure stand of white clover is defoliated and winter wheat (Triticum aestivum L.) is directly drilled into the clover understorey, it is possible to achieve a whole crop silage yield approximately equal to that in conventionally grown crops. Since this yield can be achieved with a lower rate of N fertilizer and with less use of herbicides, fungicides and insecticides, the cost of production is lower in the bicropping system than in the conventional system. Grain yields are generally smaller in the bicropping system than in the conventional system. White clover bicropped with maize needs to be suppressed by mechanical or chemical means to reduce the effects of competition on the maize. Work to develop bicropping systems is progressing.

INTRODUCTION

Intercropping is usually defined as growing more than one crop simultaneously in the same field (Andrews and Kassam, 1976). This definition includes most pastures and leys, but generally the literature on intercropping deals with growing crops simultaneously for root, tuber or seed production (Willey, 1979). Usually only legume-cereal intercropping systems are considered (e.g. Ofori and Stern, 1987). White clover (Trifolium repens L.) and other forage crops can be intercropped with cereals as a way to establish pastures and leys, which is a common practice to improve economic returns. However, forage crops can also be of interest as intercrops in arable cropping systems as a means to maintain or improve soil and water quality (e.g. Hargrove, 1991), reduce weed problems (e.g. Heyland and Merkelbach, 1991) and/or to decrease the incidence of pests and diseases (e.g. Costello and Altieri, 1995). Environmental concerns have encouraged development of practices that require less use of chemicals and mineral fertilizer and/or that conserve soil fertility. Such practices often include crops other than cash crops to assist in the maintenance of the main crop and soil. Crops included in this manner are often called catch, cover or green manure crops depending on their intended function. Bicropping is used here to describe a special case of intercropping where an intercrop is allowed to become perennial and where the main crop is established annually into the perennial crop. This review emphasizes the use of white clover as an undersown crop in cereals and as a component in bicropping systems.

WHITE CLOVER AS AN INTERCROP

Traits that make white clover attractive as an intercrop in arable systems include its perenniality and its ability to fix nitrogen (N₂) to the benefit of a subsequent crop. As white clover is perennial, it can make efficient use of the whole growing season by including late fixation. This is of particular value in areas with autumn and early spring for growth and N₂ mild winters and where there are problems with soil erosion (Scott et al., 1987). When land is tilled before sowing in spring, N bound in white clover tissues will quickly become available to the new crop (Wallgren and Lindén, 1994).

The stoloniferous growth habit is another trait that makes white clover useful as an intercrop in arable systems. When undersown in a cereal the horizontal growth of the stolons will keep the white clover from becoming as tall as the cereal, thus severe shading of the cereal and entanglement at harvest are avoided. The horizontal growth of the stolons also makes white clover able to cover the soil surface efficiently. The leaves are arranged horizontally and concentrated at about the same height. Consequently, white clover is an efficient competitor with weeds that germinate after the white clover is established, but may also cause management challenges when annuals are established in it (Galloway and Weston, 1996).

MINERALIZATION OF WHITE CLOVER

White clover tissue has a high N concentration and thus much N can mineralize in a short time if the conditions are favourable. Marstorp and Kirschmann (1991) found that shoots of white clover mineralized as fast as shoots of subterranean clover (Trifolium subterraneum L.) and black medic (Medicago lupulina L.). Shoots from red clover (T. pratense L.) and Persian clover (T. resupinatum L.) required a period of net N immobilization before net mineralization started and once started, occurred at a slower rate than with shoots of white clover. However, comparisons of rate of mineralization among species are much affected by the stage of development at sampling. Net mineralization from leaves is faster than from stems (Quemada and Cabrera, 1995) and in many species, like red clover and lucerne, there is a large increase in the proportion of biomass in stems during development (Wilman and Altimimi, 1994). In white clover, the proportion of stolon biomass does not increase and the proportion of leaves does not decrease to the same extent. The concentration of N in white clover shoots is, however, somewhat affected by the development of inflorescence with a relatively low N content. Still, the properties of white clover shoots change less with stage of development than the properties of shoots of red clover and lucerne (Wilman and Altimimi, 1994). When all crops were harvested in their bud or early flowering stage and the proportion of leaves were similar among the crops, Wivstad (1997) measured a larger short-term net mineralization of N from white clover than from red clover, but smaller than from yellow sweetclover (Melilotus officinalis Lam.).

Perennial legumes have a relatively larger root biomass than annual legumes, especially in the autumn. Kirschmann (1988) also recorded a higher concentration of N in the roots of perennial than in roots of annual legumes. When sampling in October after sowing in May, he found that the roots of white clover contained about 30 percent of the N in the plants, while the roots of subterranean clover, black medic and Persian clover all contained less than 11 percent of the N in the plants. The rate of release of N from decomposing legume roots is faster than from stems, but slower than from leaves (Wivstad, 1997). Much N is lost also from roots of growing white clover. Laidlaw et al. (1996) estimated a loss per day that was about ten times larger than the loss from roots of perennial ryegrass (Lolium perenne L.).

WHITE CLOVER UNDERSOWN IN CEREALS

The incorporation of white clover and other legumes as green manure is a very important source of N in organic farming systems. The use of catch or cover crops in both conventional and organic systems for purposes other than fixation of N₂ might also benefit from including legumes, as a pure stand or in mixtures, to improve the residual effect. However, care must be taken so that the original purpose of the crop is not jeopardized, e.g. to prevent N from leaching or reduce soil erosion. To fulfil their purposes without diminishing economic returns, green manure, catch and cover crops often have their main growth during periods of the year when the fields are not used for cash crops. In a cereal:cereal crop sequence, especially in northern temperate climates, white clover must be undersown to produce any substantial biomass during the period from the harvest of one cereal crop to incorporation before sowing the next crop (Stute and Posner, 1993). The introduction of undersown crops reduces the possibilities for weed control; thus it is important, especially in organic farming, that the undersown crop reduces weed problems by competition.

The establishment of white clover and its effect on grain yield

It is unlikely that any substantial direct transfer of N will occur from white clover to the cereal in which it is undersown (McNeill and Wood, 1990). Mineralization of decomposing tissues is generally the major pathway of N transfer and in most circumstances very little white clover tissue will die during the lifetime of a short-lived cereal. Therefore, white clover is not likely to cause any yield advantages due to increased N supply to the cereal in which it is undersown. However, a significant yield increase after undersowing of white clover is possible as reported by Breland (1996a). Breland explains a significant increase in grain yield of spring barley (Hordeum vulgare L.) in one of his experiments as being caused by a large senescence of white clover between early heading and the milkstage of barley. In all his experiments the grain yields were small and the seed rate of undersown white clover very large. Thus, in the experiments in which no large senescence of white clover occurred, grain yields of spring cereals were smaller with white clover than without.

Spring cereals

Generally, the grain yield of a dense crop of spring cereals is little affected by an undersowing of white clover (e.g. Wallgren and Lindén, 1994). In a less dense crop, like in the experiments by Breland (1996a), a significant reduction in grain yield can be avoided by increasing the seed rate of the cereal, or more practically, by keeping the seed rate of the undersown crop low (Kvist, 1992). Delaying the undersowing in relation to the sowing of the main crop also reduces the effect on grain yield. However, the undersowing should not be delayed beyond the emergence of the cereal because of the risk of causing damage to the cereal in the seeding operation and because of the risk that the undersown crop will fail to establish (Ohlander et al., 1996).

Maize

Crops that establish slowly, like maize (Zea mays L.), are likely to suffer more from competition with undersown white clover than spring cereals. As with spring cereals, delaying the undersowing will reduce the effect of competition on maize. Scott et al. (1987) found that the grain yield of maize was not largely affected by white clover established when the maize was between 0.15 and 0.30 m high. To be able to perform effective weed control, Stute and Posner (1993) delayed the undersowing until the maize was 0.30-0.60 m. With this late undersowing, white clover and all other intercrops tested, died at canopy closure. They concluded that this late undersowing of intercrops is not a potential method of introducing green manures into continuous maize production in Wisconsin.

Winter cereals

Due to competition effects from the cereal there is a great uncertainty involved in the establishment of white clover in winter cereals in the spring (Schultheiss and Opitz v. Boberfeld, 1994). In experiments conducted by Heyland and Merkelbach (1991) in southern Germany, white clover became well established when undersown in winter wheat (Triticum aestivum L.) in the autumn or early spring. However, white clover stands were thin if undersown when the wheat plants had five side shoots and failed completely when undersown at the beginning of stem elongation. Regardless of time of undersowing, the grain yield of the winter wheat was never reduced by the undersowing of white clover. Undersowings are most likely to be successful when the anticipated grain yields are small as in organic farming. Thus, Schmitt and Dewes (1996), who worked with organic farming, were successful in establishing white clover in winter wheat after weed harrowing late in the spring. In spite of vigorous growth of clover, grain yields were still not notably affected. In experiments where white clover is undersown in cereals at different rates of N fertilizer, the amount of white clover is generally smaller at the higher rates (e.g. Bakermans and van der Zweerde, 1972). Reynolds et al. (1994) suggested that production at low N levels is not limited by access to light and a larger total biomass can therefore be produced when a legume is intercropped with cereals as long as other production factors are supplied in sufficient amounts.

In areas with serious frost in the winter there is a great risk of poor overwintering of white clover if it is sown late in the autumn. Laidlaw and McBride (1992) showed that young plants of white clover do not withstand low temperatures as well as older plants. White and Scott (1991) avoided this problem by sowing white clover and other legumes a month before the sowing of winter wheat or rye. In their experiments the white clover still caused no reduction in grain yield, but with red clover and lucerne the grain yield of wheat was substantially reduced.

The effect on weeds

The relative time of emergence is very important in deciding the outcome of competition between forage crops undersown in cereals and weeds. In pot trials, Kvist (1992) found that the reduction in weed biomass caused by undersown Italian ryegrass (Lolium multiflorum Lam.) was 65 percent less when the undersowing was delayed six days in relation to the sowing of the cereal than without the delay. However, his field experiments were dominated by weeds, which are likely to have emerged as early as the undersown crop, e.g. Chenopodium album L. Thus, he found that the undersowing of forage crops in spring cereals did not have any substantial effect on weed growth until after the harvest of the cereal.

In winter cereals no tillage is needed in the spring and thus there is a greater likelihood of an undersown crop emerging earlier than many of the weeds. In experiments by Hartl (1989), white clover broadcasted in the spring in winter wheat reduced the dry matter of weeds that had died by the time of harvest by 9 percent and the living weeds by more than 70 percent, indicating that the effect was largely on weeds germinating when the white clover was well established. Heyland and Merkelbach (1991) found that not only the amount of weeds at the harvest of wheat, but also the non-wheat biomass, was smaller with white clover undersown in the autumn or early spring than without undersown clover. With Italian ryegrass undersown in the autumn, the non-wheat biomass was larger than in the control and in spite of a much larger biomass of Italian ryegrass than white clover the amount of weeds was about equal.

The growth of the weeds after the harvest of the cereal is generally substantially reduced by undersown forage crops. It is only when the forage crops are poorly established that there is vigorous growth of weeds in the autumn (Schultheiss and Opitz v. Boberfeld, 1994).

EFFECTS OF INCORPORATING WHITE CLOVER AFTER THE YEAR OF ESTABLISHMENT

Grain yield of subsequent crops

The incorporation of a white clover crop, by e.g. ploughing under, generally results in an increased yield of the subsequent crop (e.g. Wallgren and Lindén, 1994). However, there is a large variation in this effect among experiments, which can be explained by differences in amount of white clover incorporated, climate, soil texture and management. In one experiment Breland (1996a) recorded an increase in grain yield of spring cereals corresponding to a N fertilizer rate of about 60 kg ha^-1 after incorporation of white clover in the previous autumn, but in another experiment with similar amounts of white clover the yield increase was negligible. The difference between the experiments was explained by a larger leakage of N on the coarser textured soil. However, as an average over several experiments in Sweden, Wallgren and Lindén (1994) reported an increase in the grain yield of spring barley of about 10 and 7 percent after incorporation of white clover late in autumn and in spring, respectively. The increase in N uptake of the barley was about 30 kg ha^-1, which was similar to the content of N in the white clover before incorporation. In more southern latitudes white clover can be expected to grow vigorously during autumn and thus have potential for a larger effect on the grain yield of the subsequent crop. In Germany, Dachler and Köchl (1994) recorded a yield increase of spring cereals corresponding to a N fertilizer rate of about 60 kg ha^-1 after incorporation of white clover late in autumn.

There are not many examples in the literature where no or negative effects of incorporation of white clover has been recorded. However, the effect of white clover incorporated before sowing, on grain yield of winter wheat can be quite small, which is probably due to the large losses of N through leaching following the incorporation of legumes early in the autumn (Stopes et al., 1996). Faris et al. (1986) report no effect of incorporation of white clover, but in their experiments N fertilizer had no impact on yield until the fourth year of the experiments. During the first three years, other factors limited grain yield. There is some risk that the white clover residue will have a phytotoxic effect on annuals established within some weeks after incorporation (Breland, 1996b). However, according to Weston (1990) large seeded and fast germinating species perform better under conditions with high residue than small seeded slow germinating species do. The phytotoxity can thus be a problem in vegetable production (Stirzaker and Bunn, 1996), but with large seeded annuals it seems to be less of a problem and can instead be used in the control of small seeded weeds (White et al., 1989).

N losses to the environment

The leaching of N to surface and ground water constitutes a major risk after incorporation of legumes (Francis et al., 1992). Wallgren and Lindén (1994) observed that neither red nor white clover was able to reduce the content of mineral N in the soil during autumn compared to unploughed ground without an undersown crop, while perennial ryegrass reduced the N content of the soil substantially. Francis et al. (1992) recorded a much larger leakage of N with a longer rather than shorter fallow period between the incorporation of a white clover/ ryegrass pasture and the sowing of the next crop, when the new crop was established in the spring. However, if legumes are incorporated before the sowing of winter cereals the risk of leaching is large even with a short fallow period (Stopes et al., 1996). The rate of mineralization is high when the soil is warm and soil moisture sufficient (Cassman and Munns, 1980), which is a common situation during early autumn in Europe. Consequently, Stopes et al. (1996) recorded a loss of 102 kg N ha^-1 through leaching when red clover was incorporated prior to sowing of winter wheat, while only 26 kg ha^-1 was leached when the red clover was incorporated the following spring. The corresponding figures with ryegrass were 18 and 4 kg ha^-1, respectively. The results of the studies by Francis et al. (1992) and Stopes et al. (1996) indicate that a white clover crop should not be ploughed until spring because of the large risk of leaching during winter.

If legumes are cut and the residue left on the soil surface there may be large volatile losses of ammonia, but these losses are small if the legumes are ploughed under (Glasener and Palm, 1995). Another cause of loss is denitrification. Substantial amounts of N can be lost on water-saturated soils. Aulakh et al. (1991) found negligible losses of N through denitrification when 60 percent of the pore space was filled with water and about 15 mg N kg^-1 soil when 90 percent of the pore space was filled with water. When legume residues were included, the losses increased a further three to six times.

THE BICROPPING SYSTEMS

White clover:winter wheat

Concerns relating to environmental problems and excess production from intensive cereal production caused Jones and Clements (1993) to devise a method of growing cereals where much of the N requirement is supplied by a permanent, perennial white clover understorey. Subsequent work by Clements et al. (1997) and Clements and Donaldson (1997) refined the system to make it reliable and profitable for whole-crop silage production and investigated how the system could be developed for cereal grain production. The system also enables major savings in agrochemical use to be made.

The system that has evolved is straightforward and simple to adopt. Initially a sward of pure white clover is established and is defoliated in the autumn either by machine and ensiled or grazed by sheep. The cereal is then directly drilled into the clover understorey, which becomes permanent and perennial. In spring there is a need to apply a small amount of N, around 50 kg ha^-1, or to apply slurry containing about that amount of N. The cereal and clover develop together and are cut as whole crop silage (i.e. cut at the soft dough stage of development of the cereal ear). The clover understorey recovers very quickly following harvest and is then cut and ensiled or grazed in the autumn again and another winter cereal crop is directly drilled in order to repeat the cycle. Instead of being cut for silage, the crop can be left to develop so that the cereal matures and is harvested for grain.

Nitrogen fertilizer applications are greatly reduced using this system and yields of silage exceeding 20 t ha^-1 dry matter have been obtained. Slurry can also be utilized fully by the system and the presence of vegetation throughout the year reduces the risk of slurry running off the field surface. The need to use insecticides in the bicropping system is greatly reduced. This is because aphid numbers are usually low (Clements and Donaldson, 1997). Slugs may damage cereal seedlings, but can be controlled by tactical molluscicide use. Use of direct drilling machines that have a press wheel to close the drill slot, makes the environment less attractive for slugs and reduces the likelihood of damage by them.

The clover understorey deflects rain splash and greatly reduces the progress of splash-borne diseases (e.g. Septoria spp.). Fungicide is not usually needed to control Septoria in bicropped cereals, but if it is needed at all the dose rate can be halved. Late season lodging caused by eyespot (Peudocercosproella) is a problem in the third successive wheat crop. Probably the best solution to the problem of stem-base diseases is to not grow more than two successive winter wheat crops, but to use a break crop e.g. oats (Avena sativa L.) in the third year.

The dense clover/cereal canopy makes it difficult for broad-leaved weeds to survive, but grass weeds (Poa spp. particularly P. trivialis) have caused major losses. However, the grass weeds are easy to control for the following year by applying paraquat, to which clover is resistant in the autumn before drilling the cereal. Cleavers (Galium spp.), particularly spreading from hedgerows, if allowed to encroach, could be a problem since as yet no selective herbicide is available.

Soil erosion and the associated problems of phosphate and particulate run-off into water courses are virtually eliminated by the bicropping system, because the perennial and dense plant cover retains its grip on the soil, year round. Earthworm populations build-up to high levels aided by the lack of soil disturbance, abundance of organic matter, the insulating effects of the permanent crop cover and low inputs of agrochemicals. Although there has been no scientific evaluation of the situation, the presence of large populations of earthworms together with the cover provided by bicropping probably enhances the survival of many wildlife species.

The system is well suited to the production of whole crop silage especially on mixed farms where the option of leaving the crop to grow for grain production would sometimes be very attractive. Grain yields have often been variable and disappointingly low compared to conventional crops.

There is considerable scope to improve cereal grain yields further, without increasing inputs. Unpublished results from Bergkvist indicate that if the system is intensified by introducing a slight tillage operation before sowing the wheat, it is possible to produce large grain yields. On the other hand, with this intensification, weed problems and problems with e.g. Septoria are likely to increase. The weeds, both broad-leaved and grasses, can be effectively controlled by applying diflufenikan + isopruteron when the wheat has about two leaves (Bergkvist, unpublished). This was achieved with some damage to the clover, but without killing it. The modified system can be of particular value in countries where the use of paraquat is not allowed.

There is a need to investigate the use of non-cereals to be used as break crops and to develop the system for organic systems.

White clover:maize

Maize grown from silage is becoming increasingly more popular in Europe for several reasons e.g. the advent of more cold-tolerant varieties, the realization by dairy farmers that maize can be dealt with by contractors, maize can withstand and utilize large amounts of slurry and finally there is the benefit of growing a crop for silage, harvested at a distinctly different time of year to other forages, allowing a better spread of activity on the farm.

Unfortunately, the maize crop may engender some serious problems of environmental pollution. Partly these are a consequence of the long period of time during the year where there is no or very little crop cover. There are also associated problems of phosphate loss and pollution, since phosphates bind to soil particulates and cause eutrophication of watercourses. Several authors have attempted to grow understoreys within maize or to use over-winter cover crops either to try and capture excess N or to prevent soil erosion. Most frequently in Europe, grasses or other non-legumes, e.g. rape, have been used for this purpose (e.g. Froment, 1998; Schroder, 1998). However in the USA legumes may be grown under maize to prevent soil erosion (Hartwig, 1983). More recently, white clover has been used as an understorey (Ammon and Bohren, 1996; Ammon et al., 1995) in work in Switzerland and Germany. White clover has the substantial advantage over other potential understorey crops that it is perennial and its stoloniferous habit makes it ideal for reducing or preventing soil erosion.

The challenge in developing a white clover:maize bicropping system is that spring sown crops in general have considerable difficulty in competing with white clover grown as a perennial understorey (Williams and Heyes, 1991). The slow early growth of maize accentuates the problem further. To avoid severe competition from the clover, mechanical or chemical control of the growth of the clover is necessary (Ammon et al., 1995).

REFERENCES

Ammon, H.U. & Bohren, C. 1996. Maize seeding in living catch crop. Vegetation management and weed control by living or dead mulch and mechanical interrow cutting. X^e Collogue International sur le biologie des mauvaises herbs, Dijon, September 1996, 387-392.

Ammon, U.H., Gariboy, S. & Bohren, C. 1995. The use of dead or living mulch in maize and its suppression with herbicides. 9^th EWRS (European Weed Research Society) Symposium Budapest 1995: Challenges for weed science in a changing Europe, 527-534.

Andrews, D.J. & Kassam, A.H. 1976. The importance of multiple cropping in increasing world food supplies. In: Multiple cropping, American society of agronomy, Special publication No 27. Madison. 1-10.

Aulakh, M.S., Doran, J.W., Walters, D.T. & Power, J.F. 1991. Legume residue and soil water effects on denitrification in soils of different textures. Soil Biology and Biochemistry 23, 1161-1167.

Bakermans, W.A.P. & Zwerde van der, H. 1972. Growing green-manure under wheat. Bedrijfsontwikkeling 3 (9), 809-811.

Breland, T.A. 1996a. Green manuring with clover and ryegrass catch crops undersown in small grains: crop development and yields. Acta Agriculturae Scandinavica. Section B, Soil and Plant Science 46, 30-40.

Breland, T.A. 1996b. Phytotoxic effects of fresh and decomposing cover crop residues. Norwegian Journal of Agricultural Sciences 10, 355-362.

Cassman, K.G. & Munns, D.N. 1980. Nitrogen mineralization as affected by soil moisture, temperature, and depth. Soil Science Society of America Journal 44, 1233-1237.

Clements, R.O. & Donaldson, G. 1997. Clover and cereal – low input bi-cropping. Farming and Conservation, August 1997: 12–14.

Clements, R.O., Donaldson, G., Purvis, G. & Burke, J. 1997. Clover: cereal bi-cropping. Optimising cereal inputs. Its scientific basis. Aspects of Applied Biology 50, 467-479.

Costello, M.J. & Altieri, M.A. 1995. Abundance, growth rate and parasitism of Brevicoryne brassicae and Myzus persicae (Homoptera: Aphididae) on broccoli grown in living mulches. Agriculture, Ecosystems & Environment 52, 187-196.

Dachler & Kochl. 1994. Die Wirkung verschiedener Gründüngungspflanzen auf Ertrag und Erlös der Folgekultur. Bodenkultur 45 (2), 139-146.

Faris, M.A., Smith, D.L. & Coulman, B.E. 1986. Plough down effects of different forage legume species, cultivars, cutting strategies and seeding rates on the yields of subsequent crops. Plant and Soil 95, 419-430.

Francis, G.S., Haynes, R.J., Sparling, G.P., Ross, D.J. & Williams, P.H. 1992. Nitrogen mineralization, nitrate leaching and crop growth following cultivation of a temporary leguminous pasture in autumn and winter. Fertilizer Research 33, 59-70.

Froment, M. 1998. Undersown Italian ryegrass cover crops for forage maize. Grass Farmer 60, 14.

Galloway, B.A. & Weston, L.A. 1996. Influence of cover crop and herbicide treatment on weed control and yield in no-till sweet corn (Zea mays L.) and pumpkin (Cucurbita maxima Duch). Weed Technology 10, 341-346.

Glasener, K.M. & Palm, C.A. 1995. Ammonia volatilization from tropical legume mulches and green manurees on unlimed and limed soils. Plant and Soil 177, 33-41.

Hartl, W. 1989. Influence of Undersown Clovers on Weeds and on the Yield of Winter Wheat in Organic farming. Agriculture, Ecosystems and Environment 27, 389-396.

Hargrove, W.L. (Ed.). 1991. Cover crops for clean water. Soil and water conservation society, Ankeny, Iowa, USA. 197 pp.

Hartwig, N.L. 1983. Crown vetch and no tillage crop production for soil erosion control. Penn. State. Univ. Extension Service, Pennsylvania 16802 USA.

Heyland, K.U. & Merkelbach, H. 1991. Die Möglichkeiten des Einsatzes von Untersaaten zur Unkrautunterdrückung sowie Konkurrenzwirkungen von Unkraut und Untersaat auf die Ertragsbildung des Winterweizens. Bodenkultur 42 (4), 347-359.

Jones, L. & Clements, R.O. 1993. Development of a low input system for growing wheat (Triticum vulgare) in a permanent understory of white clover (Trifolium repens). Annals of Applied Biology 123, 109-119.

Kirschmann, H. 1988. Shoot and root growth and nitrogen uptake by six green manure legumes. Acta Agriculturae Scandinavica 38, 25-31.

Kvist, M. 1992. Catch crops undersown in spring barley - competitive effects and cropping methods. Crop production Science 15. Uppsala. Sweden. 210 pp.

Laidlaw, L.S., Christie, P. & Lee, H.W. 1996. Effect of white clover cultivar on apparent transfer of nitrogen from clover to grass and estimation of relative turnover rates of nitrogen in roots. Plant and Soil 179, 243-253.

Laidlaw, L.S. & McBride, J. 1992. The effect of time of sowing and sowing method on production of white clover in mixed swards. Grass and Forage Science 47, 203-210.

Marstorp, H. & Kirschmann, H. 1991. Carbon and nitrogen mineralization and crop uptake of nitrogen from six green manure legumes decomposing in soil. Acta Agriculturae Scandinavica 41, 243-252.

McNeill, A.M. & Wood, M. 1990. Fixation and transfer of nitrogen by white clover to ryegrass. Soil Use and Management 6, 84-86.

Ofori, F. & Stern, W.R. 1987. Cereal-legume intercropping systems. Advances in Agronomy 41, 41-90.

Ohlander, L., Bergkvist, G., Kvist, M. & Stendahl, F. 1996. Yield of catch crops and spring barley as affected by time of undersowing. Acta Agriculturae Scandinavica. Section B, Soil and Plant Science 46, 161-168.

Quemada, M. & Cabrera, M.L. 1995. Carbon and nitrogen mineralized from leaves and stems of four cover crops. Soil Science Society of America Journal 59, 471-477.

Reynolds, M.P., Sayre, K.D. & Vivar, H.E. 1994. Intercropping wheat and barley with N fixing legume species: a method for improving ground cover, N use efficiency and productivity in low input systems. Journal of Agricultural Science, Cambridge 123, 175-183.

Schmitt, L. & Dewes, T. 1996. Influence of clover understorey on canopy structure, yield and quality of winter wheat in a low input farming system. In: Legumes in sustainable farming systems. Occasional symposium no. 30 British Grassland society. Reading.

Schroder, J. 1998. Towards improved nitrogen management in silage maize production on sandy soils. Dissertation LU-2417. Landbouwuniversiteit, Wageningen. 223 pp.

Schultheiss, U & Opitz v. Boberfeld, W. 1994. Untersuchungen zur gegenseitigen Beeinflussung der Deckfrucht Wintergerste und verschiedenen Untersaaten. Bodenkultur 45(1), 25-35.

Scott, T.W., Pleasant, J.Mt., Burt, R.F. & Otis, D.J. 1987. Contributions of ground cover, dry matter, and nitrogen from intercrops and cover crops in a corn polyculture system. Agronomy Journal 79, 792-798.

Stirzaker, R.J. & Bunn, D.G. 1996. Phytotoxity of ryegrass and clover cover crops, and a lucerne alley crop for no-till vegetable production. Biological Agriculture and Horticulture 13, 83-101.

Stopes, C., Millington, S. & Woodward, L. 1996. Dry matter and nitrogen accumulation by three leguminous green manure species and the yield of a following wheat crop in an organic production system. Agriculture, Ecosystem and Environment 57, 189-196.

Stute, J.K. & Posner, J.L. 1993. Legume cover crop options for grain rotations in Wisconsin. Agronomy Journal 85, 1128-1132.

Wallgren, B. & Linden, B. 1994. Effects of catch crops and ploughing times on soil mineral nitrogen. Swedish Journal of Agricultural Research 24, 67-75.

Weston, L.A. 1990. Cover crop and herbicide influence on row crop seedling establishment in no-tillage culture. Weed Science 37, 674-679.

White, J.G. & Scott, T.W. 1991. Effects of perennial forage-legume living mulches on no till winter wheat and rye. Field Crops Research 28, 135-148.

White, R.H., Worsham, A.D. & Blum, U. 1989. Allelopathic potential of legume debris and aqueous extracts. Weed Science 37, 674-679.

Willey, R.W. 1979. Intercropping - its importance and research needs. Part 1: Competition and yield advantages. Field Crop Abstracts 32, 1-10.

Williams, E.D. & Heyes, M.J. 1991. Growing cereals in a white clover (Trifolium repens) crop. Journal of Agricultural Science, Cambridge, 117, 23-27.

Wilman, D., Altimimi, M.A.K. 1984. The in-vitro digestibility and chemical composition of plant parts in white clover, red clover and lucerne during primary growth. Journal of the Science of Food and Agriculture 35, 133-138.

Wivstad, M. 1997. Green-manure crops as a source of nitrogen in cropping systems. Dissertation. Agraria 34. Swedish University of Agricultural Sciences, Uppsala.