Successive legumes tested in a greenhouse crop rotation experiment modify the inoculum potential of soils naturally infested by Aphanomyces euteiches

Authors

  • A. Moussart,

    Corresponding author
    1. INRA, UMR1349 IGEPP, F-35653 Le Rheu
    2. Union Nationale Interprofessionnelle des Plantes riches en Protéines (UNIP), 11 rue de Monceau, CS 60003, F-75378 Paris, France
    Search for more papers by this author
  • M. N. Even,

    1. INRA, UMR1349 IGEPP, F-35653 Le Rheu
    2. Union Nationale Interprofessionnelle des Plantes riches en Protéines (UNIP), 11 rue de Monceau, CS 60003, F-75378 Paris, France
    Search for more papers by this author
  • A. Lesné,

    1. INRA, UMR1349 IGEPP, F-35653 Le Rheu
    2. Union Nationale Interprofessionnelle des Plantes riches en Protéines (UNIP), 11 rue de Monceau, CS 60003, F-75378 Paris, France
    Search for more papers by this author
  • B. Tivoli

    1. INRA, UMR1349 IGEPP, F-35653 Le Rheu
    Search for more papers by this author

E-mail: anne.moussard@rennes.inra.fr

Abstract

The consequence of 10 successive monocultural cycles involving different legume species/cultivars on the inoculum potential (IP) of soils naturally infested by Aphanomyces euteiches was investigated under greenhouse conditions. The results showed that the IP of a soil naturally infested by A. euteiches can be significantly modified not only by the non-host or host status of crop species but also by the level of resistance of the cultivar. Susceptible species/cultivars (pea, lentil and susceptible cultivars of vetch and faba bean) are very favourable to pathogen multiplication, and continuous cultivation of each of these increased the IP values of a soil with a moderate initial IP (from 1·9 to 3·5 after 10 cycles). Conversely, non-host species and resistant cultivars of vetch or faba bean contributed to reducing the IP values of soils irrespective of the initial IP (from 1·9 to 0·5 and from 4 to 2, respectively, after 10 cycles). Aphanomyces root rot severity values on the resistant legume species/cultivars were not affected by the successive cultural cycles. This study, which showed that the IP of A. euteiches in soil can be reduced by planting appropriate legume species and cultivars in greenhouse conditions, will be useful for defining better crop successions for legumes.

Introduction

Crop diversification is one of the practices used to develop a long-term strategy for disease management (Krupinsky et al., 2002). Diseases can be most effectively reduced or avoided with crop selection and crop rotation to non-host crops for pathogens that are soil- or residue-borne. Thus, increasing diversification of cereal cropping systems by alternating crops, such as oilseed, pulse and forage crops, reduce disease risks by enhancing the population and activity of beneficial soil organisms and lengthening the time between host crops so that pathogen populations have time to decline (Ledingham, 1961; Colbach et al., 1994). Halloran et al. (2008) studied the effect of rotation crops and rotation sequences on the incidence and severity of soilborne diseases in the following potato crop and showed that several rotation sequences can reduce diseases and economic risk when compared to continuous potato or the standard barley–potato rotations.

In the case of aphanomyces root rot of pea (Pisum sativum) caused by the oomycete Aphanomyces euteiches, managing plant disease risk by appropriate crop rotations is highly important because this soilborne disease is very destructive and there is no available control method other than avoidance of highly infested fields (Moussart et al., 2009). The pathogen depends on high soil moisture for infection, root rot development and rapid spread, and disease development is optimal between 16 and 28°C (Papavizas & Ayers, 1974). In favourable conditions, Cunningham & Hagedorn (1962) observed the rapid invasion of the root cortex as well as the appearance of the sexual stage (oospores) a few days after the infection. Disease often appears early in the spring on young pea plants and yield losses may be considerable. Oospores were reported to resist adverse conditions, such as alternate freezing and thawing in dry conditions (Sherwood & Hagedorn, 1962), and to survive in soils for 10 to 20 years in the absence of susceptible crops (Pfender & Hagedorn, 1983). Many crops have been reported to be susceptible to A. euteiches: pea (Wicker et al., 2001; Levenfors et al., 2003), faba bean (Lamari & Bernier, 1985; Salt & Delanay, 1985), vetch (Levenfors et al., 2003), lentil (Lamari & Bernier, 1985), alfalfa (Malvick et al., 1998), French bean (Salt & Delanay, 1985), red clover (Tofte & Smith, 1992), subterranean clover (Greenhalgh et al., 1985), barrel medic (Moussart et al., 2007) and several other plants (Chan & Close, 1987). Rotations with non-host crops for a sufficient period of time are required to reduce A. euteiches infestation to safe levels. Oyarzun et al. (1993) observed that root rot severity was correlated with the frequency of peas or legumes grown over a period of 18 years but did not give information on the level of resistance to A. euteiches of legume species, and their effect on the inoculum potential (IP) of soil, an index of potential disease activity as defined by Malvick et al. (1994). Recently, Moussart et al. (2008) screened several legume species/cultivars for their resistance to one A. euteiches French isolate, recovered from pea, under controlled conditions. They defined four categories: (i) susceptible legume species in addition to pea (lentil, alfalfa, French bean); (ii) legume species including both susceptible genotypes and highly resistant genotypes (common vetch, faba bean and clover); (iii) species with a very high level of partial resistance (chickpea); and (iv) species developing no symptoms (lupin). The authors also demonstrated that the aggressiveness not only depends on the plant species but also on the plant cultivar and concluded that the choice of the legume species or cultivar may considerably modify the risk of disease development in the field. Therefore, the aim of this paper was to study the impact of some of these legume species/cultivars on the dynamics of the IP of naturally infested soils under greenhouse conditions. More specifically, this study aimed at defining: (i) if legume species/cultivars with different levels of resistance to A. euteiches influence the IP of a moderately infested soil when monocropped through 10 cultural cycles; and (ii) if resistant or non-host species/cultivars can contribute to decreasing the IP of a highly infested soil.

Material and methods

This study was based on two experiments. The objective of the first experiment was to study the effect of susceptible and resistant legumes on the dynamics of IP using soil with an initially moderate IP. The objective of the second experiment was to study the effect of resistant legumes on the IP dynamics using soil with an initially high IP. Experiments 1 and 2 were conducted under greenhouse conditions between October 2006 and June 2007, and October 2007 and June 2008, respectively.

Soil infestation

A sandy silt soil, naturally infested with A. euteiches, was used for the experiments. It was collected in a field at the Union Nationale Interprofessionelle des Legumes Transformés (UNILET) farm, in Riec sur Belon (47°52′43·00′N; 3°42′30·03′W), Brittany, France. This field had a consistent cropping history of wheat, maize and pea since 1990 (each crop was cultivated in the whole area of the field every 3 or 4 years) and spatial distribution of A. euteiches inoculum in the soil had been characterized (Moussart et al., 2009). Two samples of this soil, with two different levels of IP, were collected in two patches. The first sample (Experiment 1) had a low level of IP (1·9) and the second (Experiment 2) had a very high level of IP (4·0). Each sample of soil was thoroughly homogenized before use.

Plant material

Seven legume species were used in this study: pea (Pisum sativum), faba bean (Vicia faba), lentil (Lens culinaris), lupin (Lupinus alba), common vetch (Vicia sativa), clover (Trifolium pratense) and alfalfa (Medicago sativa). For each of these species, the resistance to A. euteiches of various genotypes (cultivars and breeding lines) had previously been characterized in controlled conditions in infested vermiculite (Moussart et al., 2008), using the reference isolate RB84 originating from the field where the soil for this experiment was collected. Based on these results, one cultivar of pea, lentil, alfalfa (susceptible species), clover and lupin (resistant species) were chosen (Table 1). For common vetch and faba bean for which susceptible and resistant cultivars were identified, one cultivar of each type was selected for each species. A Poaceae species (perennial ryegrass, Lolium perenne) and a non-cultivated soil were used as non-host controls. In Experiment 1, the susceptible and resistant genotypes were used. In Experiment 2, only the resistant or weakly susceptible genotypes were used. In this last experiment, the susceptible pea cv. Baccara was used as control to monitor whether the high level of IP was maintained during the whole experiment.

Table 1. Cultivars of legume species and perennial ryegrass used in each experiment
Plant speciesCultivar (source)Susceptibility/resistance to Aphanomyces euteichesa (Moussart et al., 2008)Number of seeds per potExperiment 1 (low inoculum potential)Experiment 2 (high inoculum potential)
  1. aStudy performed in controlled conditions in artificially infested vermiculite.

Pea (Pisum sativum)Baccara (Florimond Desprez)Susceptible5××
Lentil (Lens culinaris)Anicia (Agri Obtentions)Susceptible10× 
Alfalfa (Medicago sativa)Comète (Laboulet)Susceptible15××
Faba bean (Vicia faba)Baraca (Agrovegetal)Susceptible5× 
 Melodie (Agri Obtentions)Resistant5××
Common vetch (Vicia sativa)Amethyste (Jouffray-Drillaud)Susceptible10× 
 Topaze (Jouffray-Drillaud)Resistant10××
Clover (Trifolium pratense)Merviot (Pickseed)Resistant15××
Lupin (Lupinus alba)Lublanc (Agri Obtentions)Resistant5××
Perennial ryegrass (Lolium perenne)Ventoux (Jouffray-Drillaud)Non-host15××

Culture conditions

The methodology of both experiments was the same. Seeds were sown into plastic pots (2 L, 20 cm diameter) containing soil to be tested, with low or high levels of IP. In order to favour the development of the root system and to obtain a high colonization of the soil by the different legume species, the number of seeds per pot was different depending on the plant species (Table 1). For each species, four replicates were used with 10 pots per replicate. Pots were arranged in a completely randomized design in the greenhouse. The temperature was maintained at 20 ± 5°C (16 h photoperiod) and the soil was kept moist by regularly watering to favour disease.

Disease severity and IP assessments

After 4 weeks, which allowed the disease to develop and sexual reproduction to take place on the necrotic tissues, soil and plants were removed. Disease severity was assessed on the root system of the plants of one pot of each species per replicate. For this, roots were carefully washed under running tap water and the disease index (DI) was scored according to the scale described by Moussart et al. (2008): 0 = no symptoms; 1 = traces of discoloration on the roots (<25%); 2 = discoloration of 25–50% of the roots; 3 = discoloration of 50–75% of the roots; 4 = discoloration of >75% of the roots; 5 = dead plant. The occurrence of oospores was also assessed by microscopic observation of diseased root tissues (×20 magnification). Subsequently, the IP of the soil in these pots was assessed with the test described by Moussart et al. (2009), using the susceptible pea cv. Baccara as baiting host.

For each species and controls of each replicate, roots in the remaining pots were extracted from the soil, cut into 5–8 mm pieces and mixed with their respective soil sample again. Five days later (a preliminary study had shown that oospores had no dormancy period), the same species was sown again into the soil sample of the same crop from the previous cycle. This process was repeated each month for a total of 10 months in order to simulate 10 cultural cycles.

Data analysis

An analysis of variance (anova) was carried out for the screening results, and means were compared, using a Newman–Keul’s test (< 0·05) in the General Linear Model procedure of sas (SAS Institute, 1999).

Results

Experiment 1: Effect of genotypes on severity and IP in a moderately infested soil

Evolution of disease severity on roots of legumes during 10 cycles

No symptoms were observed on the control (ryegrass) and on lupin during the ten cultural cycles. Typical symptoms were observed on the roots of the other legume species, but DIs which differed among species were relatively stable from one cycle to the next: the susceptible species/cultivars remained susceptible and the resistant species/cultivars remained resistant (Fig. 1). On pea, roots were entirely necrosed starting with the first cultural cycle, and most of the plants were dead (4 < DI < 5). On susceptible vetch (cv. Amethyste), DIs ranged from 3·4 to 3·7 from the second cycle onwards. On lentil, DI values were around 2·3 until the third cycle, after which values increased to levels ranging from 3·1 to 4·3. Typical oospores were observed in the root tissues of these three susceptible species from the first cultural cycle. Disease severity on the susceptible faba bean (cv. Baraca) ranged from 1·6 to 2·5 during the entire experiment. DI values of the resistant faba bean (cv. Melodie) and resistant vetch (cv.Topaze) ranged between 1 and 2. On alfalfa and clover, very weak symptoms were observed and DI values were lower than 1·0. Oospores were occasionally observed on the roots of susceptible and resistant faba bean and resistant vetch, but never on clover and alfalfa.

Figure 1.

 Evolution of disease severity on the roots of several legume species cultivated in a moderately infested soil over 10 cultural cycles, expressed as Disease Index (DI). The bars marked with the same letter within the same graph are not significantly different (Newman–Keul’s test, = 0·05).

Dynamics of IP of soils cultivated with several legumes for 10 cultural cycles

From an initial value of 1·9, IP of soils cultivated with pea, lentil, susceptible vetch and susceptible faba bean significantly increased to 4·0, 3·3, 2·9 and 2·8, respectively, after the first cultural cycle and remained high until the last cultural cycle (Table 2). Conversely, IPs of soils cultivated with resistant species/cultivars (clover, resistant vetch, resistant faba bean and lupin), alfalfa and controls (grass and non-cultivated soil) tended to decrease slowly, although significant differences were sometimes only observed from the ninth to the tenth cycle. Comparison of the effect of the crop species during each cultural cycle (Table 2) showed that there were significant differences among IPs of soils cultivated with susceptible species/cultivars (pea, lentil, susceptible vetch, susceptible faba bean) and resistant species/cultivars (clover, lupin, resistant vetch and resistant faba bean) from the second cycle. Most notable were the very clear differences in IPs between soils cultivated with susceptible and resistant vetch (3·6 and 0·9, respectively, after the second cycle) and between soils cultivated with susceptible and resistant faba bean (3·2 and 0·6 after the second cycle, respectively). Between the third and the tenth cycle, these significant differences between IP of soils cultivated with susceptible and resistant hosts were always observed. In the case of alfalfa, this species was classified as susceptible in previous tests using infested vermiculite (Moussart et al., 2008), but was resistant in this experiment using naturally infested soil. The ryegrass and non-cultivated soil controls had the same effect on IP dynamics as resistant species/cultivars.

Table 2. Dynamics of the inoculum potential (IP) of a moderately infested soil cultivated with different crops during 10 cultural cycles (rows) and comparison of the effect of different crops on the IP after each cultural cycle (columns)
Plant speciesCultivarR/SaCultural cyclesb
012345678910
  1. aR/S: Resistance/susceptibility to Aphanomyces euteiches (study conducted under controlled conditions in artificially infested vermiculite; Moussart et al., 2008).

  2. bMeans followed by the same letter (lower case) within rows are not significantly different (Newman–Keul’s test, = 0·05). Means followed by the same letter (upper case) within columns are not significantly different (Newman–Keul’s test, = 0·05).

Pea (P. sativum)BaccaraS1·9 b4·0 aA4·2 aA3·9 aA4·4 aA4·2 aA4·0 aA3·7 aAB4·1 aA4·0 aA3·9 aA
Lentil (L. culinaris)AniciaS1·9 b3·3 aAB3·1 aB3·3 aA3·9 aA3·8 aAB3·4 aAB3·4 aAB3·8 aA4·1 aA3·7 aA
Vetch (V. sativa)AmethisteS1·9 c2·9 bAB3·6 abAB3·6 abA4·2 aA3·8 abAB4·0 abA4·0 abA3·9 abA3·9 abA3·4 abA
Faba bean (V. faba)BaracaS1·9 b2·8 aAB3·2 aB3·3 aA3·5 aA2·9 aB2·9 aB3·1 aB2·9 aB2·9 aB3·1 aA
Alfalfa (M. sativa)ComèteS1·9 a0·9 aD1·2 aCD1·4 aB0·5 aB0·8 aC1·3 aC1·0 aC0·4 aC1·1 aC1·4 aB
Clover (T. pratense)MerviotR1·9 a1·2 aCD1·9 aC1·8 aB0·9 aB1·3 aC1·0 aC1·3 aC0·7 aC0·6 aC0·9 aB
Vetch (V. sativa)TopazeR1·9 ab2·3 aBC0·9 bcdCD1·6 abcB1·5 abcB1·7 abcC1·3 abcC1·3 abcC1·1 abcdC0·6 cdC0·1 dB
Faba bean (V. faba)MelodieR1·9 a1·5 abCD0·6 bcD1·8 aB1·8 aB1·8 aC1·8 aC1·4 abcC1·1 abcC1·3 abcC0·3 cB
Lupin (L. alba)LublancR1·9 a0·8 abD0·6 bD1·5 abB1·0 abB1·3 abC1·0 abC0·8 abC0·8 abC0·3 bC0·6 bB
Perennial ryegrass (L. perenne)VentouxNon-host1·9 a0·4 cD0·6 bcD1·5 abcB1·7 abB0·9 abcC1·3 abcC1·4 abcC0·8 abcC1·3 abcC0·8 abcB
Control (non-cultivated soil)  1·9 a2·2 aBC1·8 aC1·9 aB1·8 aB1·5 aC1·3 aC0·8 aC0·6 aC1·1 aC0·8 aB

Experiment 2: Effect of resistant genotypes on disease severity and IP in a highly infested soil

Evolution of disease severity on the roots

Because of defoliation due to an insecticide application 2 days before the end of the cycle, no data were obtained for the seventh cycle. On pea, typical symptoms were observed with numerous oospores in the tissues and high DI values during the 10 cultural cycles (from 4·3 to 4·8; Fig. 2). Only few symptoms were observed on the other legume species, and oospores were rarely present in the root tissues. From the first to the tenth cycle, DI of resistant faba bean (cv. Melodie) and resistant vetch (cv. Topaze) ranged between 1·0 and 2·0. On alfalfa and clover, DI values ranged between 1·0 and 2·0 until the third and fourth cycle. During the following cultural cycles, only a few weak symptoms were observed. During the whole experiment, no symptoms were observed on the roots of lupin and ryegrass.

Figure 2.

 Evolution of disease severity on the roots of several legume species cultivated in a highly infested soil over 10 cultural cycles, expressed as Disease Index (DI). The bars marked with the same letter within the same graph are not significantly different (Newman–Keul’s test, = 0·05).

Dynamics of IP

As expected, IP of soil cultivated with pea were maintained at a very high level (3·6 to 4·5) from the first to the last cultural cycle (Table 3). The dynamics of IP of soils cultivated with resistant species/cultivars and ryegrass was the same as that of the non-cultivated soil. In these soils, IP significantly decreased from the third or fourth cultural cycle to reach values ranging from 1·4 to 2·8 at the end of the experiment, depending upon the species. The difference between the IP of soils of these crops and those of the susceptible pea crop were only significant from the sixth cultural cycle.

Table 3. Dynamics of the inoculum potential (IP) of a highly infested soil cultivated with different crops during 10 cultural cycles (rows) and comparison of the effect of different crops on the IP after each cultural cycle (columns)
Plant speciesCultivarR/SaCultural cyclesb
012345678910
  1. aR/S: resistance/susceptibility to Aphanomyces euteiches (study conducted under controlled conditions in artificially infested vermiculite; Moussart et al., 2008).

  2. bMeans followed by the same letter (lower case) within rows are not significantly different (Newman–Keul’s test, P = 0·05). Means followed by the same letter (upper case) within columns are not significantly different (Newman–Keul’s test, = 0·05).

Pea (P. sativum)BaccaraS4 abc4·5 abA4·4 abA3·8 bcA4·1 abcA4·0 abcA4·3 abcA3·6 cA4·3 abcA4·7 aA4·0 abcA
Alfalfa (M. sativa)ComèteS4 ab4·3 aAB3·8 abcCD3·4 bcdAB3·6 abcdAB3·4 bcdAB3·4 bcdB2·7 dB3·0 cdB3·3 bcdB2·8 dB
Clover (T. pratense)MerviotR44·2 aAB4·1 aABC3·4 abAB2·9 bcB2·7 bcBC2·9 bcBC2·5 bcB2·0 cB3·0 bcBC2·2 bcBC
Vetch (V. sativa)TopazeR4 a4·0 aABC4·2 aAB3·8 aA3·1 bAB1·8 dC2·2 cdC2·2 cdB2·0 cdB2·9 bcBC2·6 bcdBC
Faba bean (V. faba)MelodieR4 ab4·2 aAB4·4 aA3·1 bcBC2·8 bcB2·2 cBC3·0 bcBC1·8 bB2·1 cB2·2 cCD1·9 cBCD
Lupin (L. alba)LublancR4 a3·9 aC3·9 aBCD2·8 bC2·7 bB2·1 bcdBC2·4 bcC1·8 cdB1·6 cdB1·7 cdD1·4 dD
Perennial ryegrass (L. perenne)VentouxNon-host4 a4·3 aAB4·1 aABC3·3 abcABC3·4 abAB2·8 bcdABC2·5 bcdC1·7 dB2·6 bcdB2·2 cdCD2·5 bcdBC
Control (non-cultivated soil)  4 a3·7 abC3·7 abD3·3 abcABC3·1 bcAB3·3 abcAB2·5 cdC2·1 deB2·7 cdB2·7 cdBCD1·7 eCD

Discussion

As demonstrated for other pathosystems (Hall & Phillips, 1992; Gilligan et al., 1996; Ratnadass et al., 2012), this study showed that the IP of a soil infested by A. euteiches can be significantly modified by the choice of the crop species (host or non-host) grown within the rotation. To the authors’ knowledge, these results demonstrate for the first time that in a short-term greenhouse crop rotation experiment, the dynamics of IP also depends on the resistance of the cultivar. Susceptible species/cultivars (pea, lentil and susceptible cultivars of vetch and faba bean) are very favourable to pathogen multiplication and therefore increased the IP values of a soil with a moderate initial IP. Furthermore, these crops maintained the IP in a soil with high initial IP. Conversely, non-host species or resistant cultivars of vetch or faba bean, contributed to a reduction in IP values of soils irrespective of the initial IP, as for the non-cultivated soil. This reduction was quite rapid with regard to the biology of oospores, which are commonly considered to be highly resistant propagules (Sherwood & Hagedorn, 1962), able to survive in soil for many years (Papavizas & Ayers, 1974; Pfender & Hagedorn, 1983). This quite rapid reduction could be due to a decrease in oospore population in soil because of the artificial conditions used in the experiments, especially soil structure disruption. In this study, disease severity values on the different legume species/cultivars were not affected by the successive cultural cycles and were in agreement with the levels of resistance previously identified by Moussart et al. (2008), except for the particular case of alfalfa. Indeed, these authors had shown that in an artificial substrate (vermiculite), cv. Comète used in this study as well as all alfalfa genotypes assessed was susceptible to the reference pea isolate of A. euteiches (RB84). In the present study conducted in a naturally infested soil, only few symptoms were present on the roots of the alfalfa cultivar and no sexual reproduction was observed in the root tissues. This last observation is in accordance with the fact that to the authors’ knowledge, no aphanomyces root rot has ever been observed on alfalfa crops in France. Levenfors et al. (2003) had already observed the poor agreement between the results obtained in vermiculite and those obtained in naturally infested soil. As suggested by Salt & Delanay (1985), the microflora or other biotic or abiotic soil factors may play an important role in root rot expression, and consequently could explain the lack of symptoms observed on alfalfa with naturally infested French soils. In contrast, aphanomyces root rot was reported in North American alfalfa fields, but the disease is due to specific isolates (Malvick et al., 2009).

Results obtained in this experiment need to be confirmed in field trials for a number of years under a variety of climatic conditions, different soil types and different levels of infestation. Moreover, this study should be done with different populations/strains of A. euteiches, which may respond differently to rotations. Further investigations should also be done in order to predict the impact of alfalfa grown in an infested field. In addition, research is needed to specify if the dynamics of IP is due to solely quantitative changes in the population of A. euteiches in the soil, or whether genetic changes occur among isolates in response to the pressure of the different species/cultivars making them more or less virulent. Such research has already been done in the case of take-all disease on cereals (G. graminis var. tritici), for which it was shown that some specific crops may induce the pathogen population either towards high or low aggressive strains (Lebreton et al., 2004).

If results obtained in this study are validated in long-term field experiments, the methodology described, based on continuous short crop cycles under greenhouse conditions, could allow prediction of the dynamics of the IP of a soil infested by A. euteiches and the subsequent risk for pea crops. Indeed, as shown by Moussart et al. (2008), there was a highly significant relationship between soil IP and disease severity on pea plants during the cropping season, with a significant lower risk when IP is lower than 1·5. Such experiments should be considered in order to better define crop successions to ensure the durability of a crop system based on legumes. Thus, according to the present results, pea, lentil and susceptible cultivars of common vetch and faba bean should be avoided in infested fields whatever the level of IP. Conversely, lupin, clover and resistant cultivars of vetch or faba bean can be grown in these infested fields.

Acknowledgements

The authors thank B. Le Delliou from the Union Nationale Interprofessionelle des Legumes Transformés (UNILET) for the authorization to collect the soil. They also thank Mr Marget and Mrs Raffiot (INRA Dijon) for providing faba bean cultivars, and breeders who have provided cultivars of lentil, common vetch and lupin. The authors are grateful to Professor Sabine Banniza (University of Saskatoon, Canada) for critical comments on the manuscript. This work was supported by INRA and UNIP (Union Nationale Interprofessionnelle des Plantes riches en Protéines).

Ancillary