Two new effective semiselective crystal violet pectate media for isolation of Pectobacterium and Dickeya




Pectolytic bacteria, including Pectobacterium spp. and Dickeya spp., are best isolated on crystal violet pectate (CVP), a semiselective medium containing pectin. The source of pectin is essential, because pectolytic bacteria are not able to degrade all of them. The aims of this study were to identify a new pectin source and to perfect formulations of semiselective CVP media to isolate the pectolytic bacteria Pectobacterium spp. and Dickeya spp. from different environmental compartments (plants, soil and water). The AG366 pectin, selected after screening six different formulations, was incorporated into single-layer (SL-CVPAG366) and double-layer (DL-CVPAG366) CVP media. Both media were compared with those based on Bulmer, Sigma-Aldrich and Slendid-Burger pectins, using 39 Pectobacterium and Dickeya strains. All strains formed deep cavities on AG366-CVPs, whereas nine did not produce cavities on Bulmer or Sigma-Aldrich media. Recovery rates were similar on DL-CVPAG366, Sigma-Aldrich and Bulmer CVPs for a given taxon, and did not differ significantly between SL- and DL-CVPAG366. Pectolytic bacteria were successfully isolated on both media from field samples of diseased potatoes, carrots, tobacco, onions, radishes and ornamentals. AG366 is thus a high-performance pectin source for the elaboration of CVP media suitable to isolate Dickeya and Pectobacterium. It is also efficient for enrichment purposes in liquid medium. The validation of AG366 as an improved source of pectin to recover the polyphagous Pectobacterium and Dickeya in different environmental compartments is essential given the current worldwide emergence and recrudescence of these bacteria.


Pectolytic bacteria, particularly those belonging to the genera Pectobacterium (Gardan et al., 2003) and Dickeya (Samson et al., 2005), are among the main phytopathogens of important crops worldwide (Ma et al., 2007). Their pathogenicity lies essentially in their ability to secrete cell-wall-degrading enzymes (including pectinases, cellulases, proteases and xylanases), and therefore to macerate host tissue (Pérombelon, 2002): this infectious process accounts for the often polyphagous nature of these bacteria, and hence for their presence in a wide range of production systems. It may also be partly responsible for the expanding host range of some of these species, and/or for the worldwide emergence of new taxa (Laurila et al., 2008; Czajkowski et al., 2009; Parkinson et al., 2009; Sławiak et al., 2009; Tsror et al., 2009; Pitman et al., 2010; Toth et al., 2011).

Understanding the ecology and epidemiological potential of pectolytic bacteria requires methods to selectively detect, isolate and quantify them in their different habitats. Many of the pectolytic bacterial plant pathogens are indeed present in several environmental compartments, including complex matrices (soil, plants) containing large microbial populations. One of the most useful techniques to track these bacteria in the environment is to use semiselective media (Meneley & Stanghellini, 1976; Burr & Schroth, 1977), such as the crystal violet pectate (CVP) media (Cuppels & Kelman, 1974; Pérombelon & Burnett, 1991; Hyman et al., 2001). Their selectivity is based on the inclusion of crystal violet, which inhibits the growth of Gram-positive bacteria, and on the presence of pectin as the major carbon source. Pectolytic bacteria form characteristic cavities in the medium, because of their ability to metabolize pectin.

CVP-based media allow the isolation of taxa belonging to different species of the genera Pectobacterium and Dickeya and the recovery in a single step of various pectolytic bacteria present in the same sample. While genus-specific media exist, such as the NGM medium useful for Dickeya identification by targeting the ability of this group to produce indigoidine (Lee & Yu, 2005), their performance to isolate bacteria from plant or environmental samples (soil, water) is poor when secondary saprophytic microorganisms are present. Therefore, CVP remains the medium of choice for isolation of pectolytic bacteria from diseased plants.

Pectin sources useful for CVP preparation should: (i) have a strong gelling capacity, to allow the medium to solidify rapidly when plating, (ii) result in deep and distinct cavities to make them clearly visible when metabolized by the target bacteria, and (iii) be nontoxic for bacteria, to allow high recovery rates for all pectolytic Dickeya and Pectobacterium species. Only a few pectin sources have been shown to combine these characteristics since the first pectin media were described (Cuppels & Kelman, 1974; O’Neill & Logan, 1975; Meneley & Stanghellini, 1976). Cother et al. (1980) developed a protocol for laboratory-scale pectin preparation, in order to overcome the scarcity of pectin sources convenient for CVP preparation. The quality of the resulting pectin was satisfying, but its preparation is time-consuming and produces only relatively small quantities in each run. Moreover, a standardized quality of this laboratory-scale pectin is difficult to ensure, as the stability and the quality of the raw material is difficult to verify. Since the mid-1970s, four commercial pectin sources (Sunkist, Bulmer, Slendid-Burger and Sigma-Aldrich) have been available (Cuppels & Kelman, 1974; Pérombelon & Burnett, 1991; Hyman et al., 2001) and used in various CVP preparations. The production of the first two of these sources (Sunkist and Bulmer) was discontinued about 10 years ago. The production process of the third source (Slendid-Burger), largely used in the food industry, was modified in 2004, when oranges were replaced by lemons as the raw material for pectin extraction; this made the resulting pectin inefficient for the detection of pectolytic bacteria. The Sigma-Aldrich pectin is thus the only source still available now; unfortunately, its use in CVP does not allow detection of all pectolytic strains, and results in small cavities from which isolation can be difficult (J.V.D. Wolf, unpublished data). Moreover, the process of commercial conditioning in small quantities (500 g maximum) means that the preparation of the media (e.g. pH adjustments) has to be adapted for each lot.

Identifying a new pectin source suitable for the isolation and detection of pectolytic bacteria therefore became essential, because of the recrudescence of Pectobacterium and Dickeya as major plant pathogens and the emergence of new pathogenic pectolytic species (Laurila et al., 2008; Czajkowski et al., 2009; Parkinson et al., 2009; Sławiak et al., 2009; Tsror et al., 2009; Pitman et al., 2010; Toth et al., 2011). This paper reports on a pectin product with good gelling properties, nontoxic for soft rot bacteria, and producing deep and distinct cavities with Dickeya and Pectobacterium species if applied in a CVP medium.

Materials and methods

Selective media development


Six new pectin formulations (Pectin-1 to Pectin-6) were tested for CVP preparation. Low-methoxyl pectins were preferred to high-methoxyl pectins, as the low pH required for gel formation of the latter is inadequate for use in bacteriological solid media (Pierce & McCain, 1992). All pectins tested were ones developed as food additives under EU designation E440; they are used as gelling agents in jams, pasteurized or sterilized beverages, or preserves. The selected pectins were provided as 100–200 g white to cream powder samples by their respective suppliers. The powder samples were stored at room temperature and below 65% relative humidity, as recommended by the manufacturers. Older stocks of Bulmer pectin (kind gift of M. Perramant, Bretagne Plants, Hanvec, France) and of Slendid-Burger pectin (M. Burger Enterprises, Madison, WI, USA), kept in the laboratory, as well as Sigma-Aldrich pectin (ref. P3850) bought from the commercial supplier (Sigma-Aldrich), were used as controls in most of the tests.

CVP preparation

CVP media were formulated as single-layer (SL-CVP) or double-layer (DL-CVP) media. The six new pectin sources were evaluated first using the single-layer formulation; the one retained after the first evaluation step (coded AG366) was then evaluated using both a single-layer and a double-layer formulation.

SL-CVP media were derived from the two-step medium CVP-S2 of Hyman et al. (2001). Two mixes, the crystal violet mix and the pectin mix, were prepared and autoclaved separately before being mixed together and poured. Ingredients of both mixes were introduced sequentially in the order of the component lists of the following recipe. The crystal violet mix was prepared in 500 mL distilled water and contained 1·02 g CaCl2·2H2O, 1 g tryptone, 5·0 g trisodium citrate, 2·0 g NaNO3, 4·0 g agar and 1·5 mL crystal violet (1% aqueous solution). Each ingredient of the crystal violet mix, except the agar, was dissolved by stirring the medium before adding the following one. The pectin mix contained 2·0 mL NaOH (5 m) and 18·0 g pectin in 500 mL distilled water. The water was heated to 80–100°C before ingredients were added, to allow the pectin to dissolve under stirring. After autoclaving (120°C for 15 min) and while still hot, the crystal violet mix was slowly poured into the pectin mix, while gently stirring the medium using a magnetic bar to avoid bubble formation. The final pH was verified to be c. 7·0, and the medium was immediately poured into Petri dishes in a laminar flow cabinet. It was then allowed to set and dry in that cabinet, at room temperature, for 24–48 h. CVP dishes were stored at 4°C; before use, they were dried with the lids ajar in the laminar flow cabinet to eliminate condensation.

DL-CVP media were prepared after Pérombelon & Burnett (1991) and Cazelles et al. (1995), with some modifications. The basal layer mix was prepared in 1 L distilled water by sequentially adding the ingredients according to the following recipe: 5·5 g CaCl2·2H2O, 1 g tryptone, 1·5 mL crystal violet (0·1% solution), 1·6 g NaNO3 and 15 g agar. After autoclaving for 15 min at 120°C and cooling to 45–50°C, the lower layer was poured in plates (15 mL/plate). The overlayer contained 2 mL EDTA 5·5% (pH 8·0), 2 mL NaOH (5 m) and 20 g pectin in 800 mL distilled water. The overlayer preparation was similar to that described for the pectin mix of the SL-CVP media, to dissolve the pectin and avoid lump formation. The overlayer mix was sterilized by autoclaving (120°C, 15 min). The pH was verified to be 7·0, and 7 mL medium were dispensed in the plates when the basal layer was firm and dry.

The Sigma and Bulmer pectins were used as references in DL-CVP formulations. The DL-CVP Sigma medium composition and preparation was identical to that described above. The DL-CVP Bulmer (Pérombelon & Burnett, 1991) was prepared as described by Cazelles et al. (1995). Briefly, the CaCl2·2H2O quantity in the basal layer was lowered to 3·7 g; 1 mL NaOH (1 m) was added; NaNO3 (1·6 g) was included in the overlayer mix instead of in the basal mix; 2 mL 5% EDTA replaced the 20 mL previously described; and 0·6 mL 1 m NaOH was added in the overlayer mix.

Pectolytic activity expression and recovery rate on the CVP media

Bacterial strain selection and bacterial culture preparation

Two collections of Pectobacterium and Dickeya strains originating from various hosts and belonging to different species were used (Table 1). The CVP evaluations were performed using calibrated pure bacterial suspensions. Bacterial colonies grown on nutrient broth agar (NBA) for 24–48 h were suspended in phosphate buffer (Na2HPO4·12H2O 0·27%, NaH2PO4·2H2O 0·04%). Bacterial concentrations were adjusted to 108 CFU mL−1, corresponding to an optical density of 0·1 at 600 nm wavelength. Two bacterial dilutions (103 and 102 CFU mL−1) were used in all tests.

Table 1.   Capacity of the tested crystal violet pectate (CVP) media to allow Pectobacterium spp. and Dickeya spp. strains from French and Dutch collections used in the study to develop cavities
Tested strainHost of originYear of isolationPectolytic capacity on different media
SL-CVPAG366DL-CVPAG366DL-CVP BulmerDL-CVP Sigma-Aldrich (ref. P3850)
  1. +: deep and distinct cavities (satisfactory). ±: very shallow cavities. −: no cavity. nt: not tested.

  2. aStrain originated from the V. Hélias collection. CFBP strains originated from the French collection of plant pathogenic bacteria. IPO strains originated from the PRI collection.

  3. bLow number of colonies compared to the control culture on NBA.

Pectobacterium spp.
 P. atrosepticum (Pa)
  86.20aSolanum tuberosum1986++++
  104aSolanum tuberosum2003++++
  96.1aSolanum tuberosum1996++++
  100TaSolanum tuberosum2003++++
  IPO 1987Solanum tuberosum1990+nt+ 
 P. carotovorum spp. carotovorum (Pcc)
  98.1aSolanum tuberosum1998++++
  93.13aSolanum tuberosum1993++++
  87.25aSolanum tuberosum1987+b+++
  IPO 1250Solanum tuberosum?+nt++
  IPO 1962Solanum tuberosum2001+nt+
  IPO 1957Solanum tuberosum2001+nt+
  IPO 1949Solanum tuberosum2001+nt++
  IPO 1955Solanum tuberosum2001+nt++
  IPO 1957Solanum tuberosum2001+nt+
  IPO 2160Solanum tuberosum2006+nt+
  IPO 2155Solanum tuberosum2006+nt+
  IPO 2161Solanum tuberosum2006+nt+±
  IPO 2175Solanum tuberosum2006+nt±+
  IPO 2145Solanum tuberosum2006+nt++
  IPO 2164Solanum tuberosum2006nt
  IPO 2153Solanum tuberosum2006+nt+
  IPO 2209Solanum tuberosum2006+nt±±
  IPO 2151Solanum tuberosum2006+nt+
  IPO 2208Solanum tuberosum2006+nt++
 P. carotovorum spp. odoriferum (Pco)
  CFBP 3261Allium porrum1982+b+++
  CFBP 1878Cichorium intybus1978++b++b
 P. betavasculorum (Pb)
  CFBP 2122Beta vulgaris cv. saccharata1972++++
  CFBP 1520Helianthus annuus?++++
  SF 142.2aCynara scolymus1986++++
 P. wasabiae (Pw)
  CFBP 3308Eutrema wasabi1985++++
Dickeya spp.
 D. chrysanthemi bv. parthenii
  CFBP 1270Parthenium argentatum1957++++
 D. chrysanthemi bv. chrysanthemi
  CFBP 3262Cichorium intybus1981++++
  CFBP 2048Chrysanthemum morifolium1956++++
 D. dianthicola
  CFBP 1805Kalanchoe blossfeldiana1977++++
  IPO 1991Solanum tuberosum1990+nt
  99.21aSolanum tuberosum1999++++
  RNS1-1aCichorium intybus2004+++±
  07.1EaSolanum tuberosum2007++++
 D. solani
  RNS08.23.3.1AaSolanum tuberosum2005++++

Preliminary evaluation of the pectins

Bacterial growth and cavity formation were rated on SL-CVP dishes of all six pectins, incubated for 48–72 h at 27°C after plating 50 μL bacterial suspension of Pectobacterium atrosepticum (Pa) 100T, Pectobacterium carotovorum spp. carotovorum (Pcc) 98.1 or Dickeya sp. (Dck) 99.21 (Table 1). Each bacterial suspension was plated in duplicate, and the experiment was performed twice. DL-CVP Bulmer medium was used as the positive control.

Pectolytic expression capacity of Pectobacterium and Dickeya species on CVPAG366

The ability of 30 Pectobacterium spp. and nine Dickeya spp. reference strains to produce cavities on the new CVP formulations was evaluated in comparison with the Sigma and the Bulmer DL-CVP media, after 24–72 h incubation at 27°C (Table 1).

Comparison of recovery rates

SL-CVPAG366 and DL-CVPAG366 formulations, prepared from the best pectin source among the six screened initially, were evaluated for recovery rate on four Pa (86.20, 104, 96.1 and 100T), three Pcc (98.1, 93.13 and 87.25) and three Dck strains (99.21, RNS08.23.3.1A and RNS1-1). Bacterial suspensions calibrated to 103 and 102 CFU mL−1 were plated on five replicate dishes of each CVPAG366. DL-CVP prepared from Sigma and Bulmer were used as comparative reference selective media, and NBA (0·3% beef extract, 0·5% bactopeptone, 1·5% agar) was used as a nonselective control medium. Colony counts were performed after 24–72 h incubation at 27°C, and used to calculate recovery rates. The performance of DL-CVPAG366 was compared to that of SL-CVPAG366 on one hand, and to Bulmer and Sigma DL-CVP on the other hand, using t-tests for paired samples on bacterial count data.


Medium preparation

CVP media with satisfactory gelling capacity, comparable to the reference Sigma and Bulmer media, and suitable for plating, were prepared from all six pectin formulations. The strong pH decrease noticed during autoclaving, resulting from pectin dissolution, required the initial pH of the medium to be raised. NaOH quantities used to adjust the final pH to 7·0 varied depending on the pectin type. The pH of the pectin mixes prepared with Pectin-5 and Pectin-6 needed to be initially adjusted to pH 11·0 to reach neutral to slightly alkaline conditions (pH 7·3 and 7·8, respectively), adequate for Pectobacterium and Dickeya growth, after autoclaving. The pectin mixes had to be autoclaved in large bottles to avoid overflowing when boiling, a phenomenon probably caused by pectin viscosity.

The media prepared from Pectin-1, Pectin-2, Pectin-3, Pectin-4 and Pectin-5 were totally translucent after plating, while that prepared from Pectin-6 gave a very light precipitate, comparable to that in Bulmer CVP. This phenomenon, probably the result of CaCl2 precipitation, did not affect the visualization of cavities.

All Pectobacterium and Dickeya strains used in the initial assessments were able to grow on each of the CVP media. However, Pectin-1 could not be degraded by these bacteria, since cavities were observed in none of the cultures, even though colonies developed. Only very shallow cavities were obtained with the four pectins, Pectin-2, Pectin-3, Pectin-4 and Pectin-5. The very limited depth of these cavities did not allow reliable identification of pectolytic colonies in the presence of numerous saprophytic colonies, making these pectin sources inadequate for detection and isolation of soft rot Pectobacterium sp. and Dickeya sp. Only one pectin source, Pectin-6, allowed both Pectobacterium sp. and Dickeya sp. strains to form deep cavities. This pectin (now referenced as AG366, Agdia Biofords SARL) was consequently the only useful source for comparative tests with Sigma, Slendid-Burger and Bulmer pectins. The corresponding single- and double-layer CVP media developed with this pectin source were named SL-CVPAG366 and DL-CVPAG366, respectively.

Pectolytic activity of Pectobacterium and Dickeya on SL-CVPAG366 and DL-CVPAG366 media

All strains tested were able to form deep cavities, similar to those obtained on the reference CVP (Sigma and Bulmer) media, on both SL-CVPAG366 and DL-CVPAG366 (Table 1; Fig. 1). The quality of the cavities obtained on both new media (SL-CVPAG366 and DL-CVPAG366) was similar, irrespective of bacterial taxon or host of origin (potatoes, ornamentals, witloof chicory) of the strain. Additional tests conducted with Pa 100T, Pcc 98.1 and Dck 99.21 showed that cavities obtained on SL-CVPAG366 were as deep as those observed on Slendid-Burger CVP-S2 (Hyman et al., 2001) (Fig. 2); unfortunately, these comparisons could not be extended to more bacterial strains, as the stock of Slendid-Burger pectin had run out.

Figure 1.

 Cavity formation on various crystal violet pectate (CVP) media inoculated with Pectobacterium carotovorum spp. carotovorum strain 98.1 and incubated at 27°C for 48 h.

Figure 2.

 Cavity formation in Slendid-Burger CVP-S2 (a) and the new SL-CVPAG366 (b) inoculated with Pectobacterium carotovorum spp. carotovorum and incubated at 27°C for 48 h.

SL-CVPAG366 performed better than the reference pectin sources when used in additional tests on the second (IPO) collection (Table 1) of 18 pectolytic strains, including one Pa, 16 Pcc and one Dck strain. Seventeen of these strains developed deep and clear cavities on SL-CVPAG366, whereas only 10 of them did on Sigma-Aldrich and Bulmer DL-CVP; among the remaining seven strains, five were not able to develop any cavities on either of the reference media, and two developed only shallow, unclear cavities (Table 1). The last strain, IPO 2164, did not form cavities on any of the CVP media tested; possibly it was a non-pectolytic contaminant. The superiority of the CVPAG366 formulations was also confirmed on a second Dck strain, RNS1-1 from witloof chicory, which generated much shallower cavities on Sigma-Aldrich DL-CVP than on either SL-CVPAG366 or DL-CVPAG366 (Table 1). Overall, cavity size tended to be smaller on all the DL-CVP media, where pits required 48–72 h incubation at 27°C to appear, compared with 24–48 h for SL-CVPAG366 (Fig. 1). This is advantageous for isolation, as it reduces the risk of liquefaction of the medium at high bacterial densities.

Recovery rate comparisons

Recovery rates on the various CVP formulations were slightly lower for Pa than for Dck or Pcc (Fig. 3). The performance of DL-CVPAG366 was similar to that of the Sigma-Aldrich DL-CVP and Bulmer DL-CVP references (t values 0·675 and 0·446, respectively, 8 d.f., > 0·05; Fig. 3). The high recovery rates on DL-CVPAG366 reflect the lack of toxicity of the AG366 pectin, an essential condition for development of Pectobacterium spp. and Dickeya spp.

Figure 3.

 Average recovery rate and standard deviations of four Pectobacterium atrosepticum, two Pectobacterium carotovorum spp. carotovorum and three Dickeya sp. strains on various semiselective crystal violet pectate (CVP) media.

Recovery rates did not differ significantly between single-layer and double-layer CVPAG366 formulations (= 1·234, 8 d.f., > 0·05; Fig. 3). The lower recovery rate of Pcc on SL-CVPAG366 was the result of the poor performance on this medium of one of the two strains tested, Pcc 87.25 (34%, vs. 78·4% for Pcc 93.13). The few comparisons conducted between CVPAG366 and Slendid CVP-S2, carried out with three strains (Pa 100T, Pcc 98.1 and Dck 99.21), showed similar recovery rates on both media, >70% in all cases (data not shown).


All characteristics required for the semiselective medium CVP to be used for isolation of Pectobacterium and Dickeya spp. (firmness to allow plating; formation of deep cavities by pectin degradation by soft rot bacteria; transparency to allow easy visualization of cavities; adequate recovery rate related to nontoxicity of the pectin) were fulfilled using AG366 as the pectin source. Only shallow cavities were formed on CVP prepared with the other five sources tested. AG366 will be commercially available soon, and distributed by Agdia, Biofords SARL.

Use of CVP plating remains a key technique for the detection, isolation and quantification of pectolytic bacteria from diseased plant samples (Palacio-Bielsa et al., 2006; Luzzatto et al., 2007; Tsror et al., 2009) or from the environment, such as soil or water (Laurila et al., 2008). SL-CVPAG366 and DL-CVPAG366 gave very satisfactory results in isolating pectolytic bacteria from a range of diseased host plants, including potato and ornamentals such as Syngonium, Lilium, Aconitum, Hosta, Zantedeschia, Ornithogalum and several flower bulbs, in various laboratories to which pectin samples were sent for testing (L. Tsror, ARO, Gilat Research Center, M.P. Negev 85280, Israel; I. Yedidia, ARO (Agricultural Research Organisation), The Volcani Center, Bet-Dagan 50250, Israel; personal communications). Both SL-CVPAG366 and DL-CVPAG366 formulations have been successfully used in the laboratories of the current study for isolation of bacteria from diseased potatoes, carrots, tobacco, onions and radishes. The fact that CVPAG366 allowed more pectolytic Pectobacterium spp. and Dickeya spp. strains to form characteristic cavities than did any other pectin source tested (including the reference Sigma and Bulmer ones) strongly suggests that this pectin can be degraded by pectolytic enzymes produced by a large range of Dickeya (including the emerging species ‘D. solani’) and Pectobacterium strains.

DL-CVPAG366 is probably better suited than SL-CVPAG366 to samples presenting large populations of soft rot bacteria, because of the slower cavity development. DL-CVPAG366, although more time-consuming and cumbersome to prepare, is safer, as the crystal violet mix can be poured at a lower temperature, reducing the risk of inhalation of vapour of this hazardous compound. The selectivity of SL-CVPAG366 and DL-CVPAG366 were very satisfactory, allowing Pectobacterium spp. and Dickeya spp. strains to be recovered, and preventing saprophytic bacteria from overgrowing the plates (V. Hélias, unpublished data). However, these media do not allow Pectobacterium spp. to be distinguished from Dickeya spp., as strains of the two genera form similar cavities on CVPAG366. Bacteria can be further characterized by plating of cavity-forming bacteria on NGM medium (Lee & Yu, 2005), which discriminates Dickeya spp. from Pectobacterium spp. Alternatively, colonies may be characterized by PCR-amplification, ELISA or biochemical assays. A further characterization of bacteria on plates may also be performed using a bio-PCR procedure (Hyman et al., 1997).

The AG366 pectin can also be used for enrichment in a liquid medium prepared according to recipes described by Gorris et al. (1994) and Meneley & Stanghellini (1976). This liquid medium (LEMAG366: 0·375% MgSO4·7H2O, 0·1% (NH4)2SO4, 0·1% K2HPO4, 20 μL NaOH 5 n and 0·17% pectin AG366) is useful for bacterial multiplication prior to isolation on CVP media, or for direct detection with serological or molecular tools (data not shown). This pectin source is therefore very valuable for the detection, isolation and epidemiological tracking of plant pathogenic pectolytic bacteria.


This work was supported by interprofessional fundings including the seed potato professionals [Groupement National Interprofessionnel des Semences et Plants (GNIS)], the ware potato professionals [Comité National Interprofessionnel de la Pomme de Terre (CNIPT)] including growers, cooperatives, retailers and ARVALIS-Institut du Végétal. We thank M. Benigni and S. Leignez from the National Federation of witloof chicory growers [Association des Producteurs d’Endives de France (APEF)] for providing witloof chicory strains.