Possibilities of direct introgression from Brassica napus to B. juncea and indirect introgression from B. napus to related Brassicaceae through B. juncea.

The impact of genetically modified canola (Brassica napus) on biodiversity has been examined since its initial stage of commercialization. Various research groups have extensively investigated crossability and introgression among species of Brassicaceae. B. rapa and B. juncea are ranked first and second as the recipients of cross-pollination and introgression from B. napus, respectively. Crossability between B. napus and B. rapa has been examined, specifically in terms of introgression from B. napus to B. rapa, which is mainly considered a weed in America and European countries. On the other hand, knowledge on introgression from B. napus to B. juncea is insufficient, although B. juncea is recognized as the main Brassicaceae weed species in Asia. It is therefore essential to gather information regarding the direct introgression of B. napus into B. juncea and indirect introgression of B. napus into other species of Brassicaceae through B. juncea to evaluate the influence of genetically modified canola on biodiversity. We review information on crossability and introgression between B. juncea and other related Brassicaseae in this report.

Introduction

The global cultivation area of genetically modified (GM) canola (Brassica napus) has continuously increased each year, encompassing approximately 9.2 million hectares in 2012 (James 2012), with Japan importing most of its canola from Canada (Ministry of Finance 2013). While the production of GM canola should be increased, the impact of biodiversity by introgression from GM canola to related Brassicaceae has aroused public concern and led to a global debate (Wilkinson and Tepfer 2009). Introgression from GM canola derived from spilled seeds has also been considered and monitored in non-cultivating countries of GM canola, such as Japan (Aono , Mizuguti ), the United Kingdom (Claessen , 2005b, Crawrey and Brown 1995, 2004) and other countries (Devos , Pivard , Reuter , von der Lippe and Kowarik 2007a, 2007b).

A list of potential recipient species for B. napus is currently available, including their crossability and hybrid fitness (OECD 2012, OGTR 2011). B. rapa is the most likely recipient of introgression from B. napus (Scheffler and Dale 1994) and is one of the most highly distributed weed in Europe, Atlantic, Africa, Asia, North and South America, and Australia (Warwick ). Because B. rapa is an allogamous plant with self-incompatibility, its crossability, including interspecific and intergeneric hybridization, is generally higher than that of autogamous plants (OECD 2012). In fact, many spontaneous hybrids can be obtained by hybridization between B. napus and B. rapa (CFIA 1999, OECD 2012, OGTR 2011).

At present, 17 Brassicaceae species are recognized, including B. juncea, B. carinata, B. nigra, B. oleracea, B. fruticulosa, B. maurorum, B. tournefortii, Diplotaxis catholica, D. muralis, D. erucoides, Eruca sativa, Erucastrum gallicum, Hirschfeldia incana, Raphanus sativus, R. raphanistrum, Sinapis alba, and S. arvensis. These species have also been listed as recipients of B. napus (OECD 2012). B. juncea is considered to have the second highest crossability with B. napus after B. rapa (Scheffler and Dale 1994).

An interspecific hybrid is easily produced between B. juncea and B. napus by artificial pollination (Mason , Mohammad and Sikka 1940, Rao and Shivanna 1997, Tsuda ). H. incana and R. raphanistrum are hybrids that were generated by artificial pollination with B. napus (Kerlan ). However, the probability of introgression into these two species is low (OECD 2012) because of poor seed productivity and fertility of the F1 progeny of H. incana (Chadoeuf ) and R. raphanistrum (Guéritaine ) as recipients. R. sativus (Huang ) and S. arvensis (Moyes ) have extremely low crossability with B. napus and barely produce F1 hybrids with B. napus under natural conditions. Moreover, crossability between B. napus and 12 of its relatives, namely B. carinata, B. nigra, B. tournefortii, D. muralis, B. fruticulosa, B. maurorum, B. oleracea, E. sativa, D. catholica, D. erucoides, Erucastrum gallicum, and S. alba, is very low (OECD 2012) and thus, spontaneous hybridization will not occur. Therefore, introgression from B. napus to these 12 closely related species relatives is highly unlikely, resulting in the identification of B. rapa and B. juncea as major candidate recipients of introgression from B. napus.

B. juncea was introduced from China before the 9th century (Hoshikawa 1998) and eventually became a naturalized species of Japan (Shimizu ). B. rapa was also introduced to Japan in the 8th century (Hoshikawa 1998). Weedy B. rapa is also derived from cultivated varieties, and is distributed as an escaped plant from fields in Japan (Osada 1980). Based on these historical accounts, the Japanese government has exempted B. rapa and B. juncea from its list of protected native plants, because biodiversity is based on the Cartagena protocol, despite the occurrence of spontaneous hybridization with B. napus. However, indirect introgression from B. napus to related Brassicaceae may still occur with B. rapa and/or B. juncea acting as bridge species.

Hybridization between D. tenuifolia and B. napus has not been reported, although B. rapa can hybridize with D. tenuifolia (Salisbury 1989), which in turn, suggests that some traits of B. napus were transferred into D. tenuifolia through B. rapa. We consider this case as a typical example of indirect introgression.

Because B. juncea is widely distributed around the world, yet centrally focused on Asia, this species has the potential role of acting as a mediator of introgression from B. napus to related members of the Brassicaceae, as observed in the case of B. rapa in America and Europe. Several research groups have previously investigated crossability, and spontaneous hybridization frequency between B. napus and B. rapa has been reported and reviewed by many research groups (CFIA 1999, OECD 2012). However, the information about B. juncea and B. napus is not sufficient (CFIA 2012, Warwick ). This review will perform two tasks. First, we will review and summarize introgression and crossability between B. juncea and B. napus. Second, we will show and discuss the potential of indirect introgression through B. juncea by crossability between B. juncea and other members of the Brassicaceae.

Introgression from B. napus to B. juncea

Spontaneous hybridization frequency

Spontaneous hybridization frequency is greatly influenced by the field experimental design, such as the size of the pollen donor and the isolated distance from the donor (Scheffler , Tsuda ). This section summarizes the effects of distance between the donor and the recipient on spontaneous hybridization frequency (Table 1).

Table 1

Spontaneous hybridization frequency between B. juncea and B. napus

Cross-combinations	Cultivation condition	Frequency (%)	References
B. juncea × B. napus	Mixed cultivation	0.13–5.91	Tsuda et al. (2012a), Heenan et al. (2007), Jørgensen et al. (1998), Huiming et al. (2007), Liu et al. (2010), Bing et al. (1991, 1996)
	Isolated cultivation (1.0–27.5 m)	0.00–0.05	Tsuda et al. (2012a)
B. napus × B. juncea	Mixed cultivation	1.10–1.30	Heenan et al. (2007), Jørgensen et al. (1998), Bing et al. (1991, 1996)

In mixed cultivation experiments, spontaneous hybridization frequency varied from 0.13 to 5.91% in B. juncea × B. napus (Bing , 1996, Heenan , Huiming , Jørgensen , Liu , Tsuda ). On the other hand, the frequency in B. napus × B. juncea ranged from 1.1 to 1.3% (Bing , 1996, Heenan , Jørgensen ). Spontaneous hybridization in Brassicaceae under isolated cultivation, wind-pollination (Becker , McCartney and Lacey 1991), and insect-pollination (Free and Spencer- Booth 1963, Free 1993) have been previously reported. However, spontaneous hybridization is generally difficult because B. juncea is a self-compatible species (Ohsawa and Namai 1987, Tsuda ) reported that hybridization frequency under isolated conditions between the donor and the recipients was 0.05% at a distance of 1 m and 0.03% at a distance of 17.5 m, and no hybrids were observed under more distant isolated conditions, ranging from 20 to 27.5 m. These results also indicate the difficulty of hybridization when B. juncea is isolated from B. napus. In Japan, it has been previously reported that the flowering period of B. juncea and B. napus overlap (Matsuo and Itoh 2001). However, when the flowering period of the donor and the recipient completely overlapped, the spontaneous hybridization frequency of B. juncea × B. napus was not high compared with that of B. rapa × B. napus (Tsuda ). Based on these reports, it has thus been estimated that spontaneous hybridization results in low hybrid production.

Hybrid production by artificial pollination

Several research groups have examined seed productivity and hybrid productivity by artificial pollination (Table 2). It is desirable to evaluate hybrid productivity on the basis of production efficiency of true hybrids per pollination because false hybrids can be produced by artificial pollination in Brassicaceae (Kakizaki 1925, Nishi and Hiraoka 1962, Nishi , Terao 1934), especially in B. juncea × B. napus (Tsuda ) and B. napus × B. juncea (Mohammad and Sikka 1940). It has been determined that false hybrids are generated by pseudogamy, which is a type of apomixis (Nishi and Hiraoka 1962). False hybrids do not present any morphological characteristics and genomic regions of its pollen parent (Ammitzbøll and Jørgensen 2006, Nishi and Hiraoka 1962, Tsuda ). Therefore, we identified features that could assist in distinguishing true hybrids from false hybrids.

Table 2

Hybrid production efficiency between B. juncea and B. napus by artificial pollination

Cross-combinations	Number of seeds/pollination		References	Number of hybrids/pollination		References
	Range	Average		Range	Average
B. juncea × B. napus	0.21–16.50	4.96	Sharma and Singh (1992), Frello et al. (1995), Heenan et al. (2007), Tsuda et al. (2011), Mathias (1985), Bing et al. (1991, 1996), Choudhary and Joshi (1999), GhoshDastidar and Varma (1999), Mohammad and Sikka (1940)	0.00–13.5	4.05	Choudhary and Joshi (1999), Tsuda et al. (2011), Mohammad and Sikka (1940), Sabharwal and Doelžel (1993), Rao and Shivanna (1997), Mason et al. (2011a)
B. napus × B. juncea	0.01–26.20	1.99	Frello et al. (1995), Heenan et al. (2007), Choudhary and Joshi (1999), Yamagishi and Takayanagi (1982), GhoshDastidar and Varma (1999), Mohammad and Sikka (1940)	0.00–0.32	0.07	Choudhary and Joshi (1999), Mohammad and Sikka (1940), Sabharwal and Doelžel (1993), Mason et al. (2011a)

The range and average of hybrid production efficiency are shown in Table 2. Mohammad and Sikka (1940) could not obtain hybrids from B. juncea × B. napus, whereas Rao and Shivanna (1997) could obtain 13.5 seeds per pollination. Despite an average of 4.05 seeds per pollination in B. juncea × B. napus cross (Table 2), hybrid production efficiency may vary among cultivars and experimental conditions. On the other hand, the progeny seeds produced only 0.32 seeds per pollination at maximum and 0.07 hybrid/pollination on average in B. napus × B. juncea. Therefore, introgression from B. napus to B. juncea showed a higher potential than the direction of B. juncea to B. napus. The cross of B. juncea × B. napus should be assessed with the assumption that false hybrids may occur. Furthermore, low seed productivity of the F1 hybrid may be a major limiting factor in the extension of introgression from B. napus to B. juncea.

Hybrid production by tissue culture techniques

Ovary, ovule, and embryo culture procedures have been used to rescue the hybridized embryos of Brassicaceae (Bajaj 1990, Kaneko ). Tissue culture has also been utilized in the production of hybrids between B. napus and B. juncea (Bajaj , Sacristán and Gerdemann 1986, Sharma and Singh 1992, Zhang ). Their range and average hybrid production efficiencies are presented in Table 3. In the cross B. juncea × B. napus, the average hybrid production efficiency ranged from 24.9 to 43.4% in each culture technique, and hybrids of B. napus × B. juncea were stably produced, within the range of 10.2–39.8%. A higher efficiency was observed in B. juncea × B. napus than in B. napus × B. juncea, but the efficiency of B. napus × B. juncea was sufficient to produce F1 hybrids. Therefore, the production of hybrids between B. juncea and B. napus by tissue culture was considered stable.

Table 3

Proportion of hybrid plants to cultured tissues

Cross-combinations	Culture method	Range (%)	Average (%)	References
B. juncea × B. napus	Ovary	0.4–67.6	43.4	Bajaj et al. (1986), Sharma and Singh (1992)
	Ovule	30.4–49.5	42.4	Bajaj et al. (1986)
	Embryo	4.2–73.0	24.9	Bajaj et al. (1986), Zhang et al. (2003)
B. napus × B. juncea	Ovary	5.4–14.9	10.2	Bajaj et al. (1986)
	Ovule	37.5–42.0	39.8	Bajaj et al. (1986)
	Embryo	0.8–60.9	22.6	Sacristán and Gerdemann (1986), Bajaj et al. (1986), Zhang et al. (2003)

Fertility of hybrids and their progenies

Introgression may be affected by the productivity of hybrids and their progenies. To evaluate the productivity of hybrid progenies, seed fertility of F1 derived from hybridization between B. juncea and B. napus and their self-progenies, as well as backcross progenies, were examined.

The F1 hybrid from B. juncea (AABB, 2n = 36) × B. napus (AACC, 2n = 38) showed a genome composition of AABC (Sabharwal and Doležel 1993, Tsuda ) or AAB (Sabharwal and Doležel 1993). Most of the hybrids had chromosomes and an AABC genome composition (Choudhary and Joshi 1999, Tsuda ). Pollen fertility of F1 ranged from 0 to 35.6% (Choudhary and Joshi 1999, Frello , Heenan , Liu , Sandhu and Gupta 2000, Song and Qiang 2003), with the highest fertility (78%) observed in nine samples of the F1 hybrid. On the other hand, the F1 hybrid from B. napus × B. juncea showed a genome composition of AABC or AABBC (Sabharwal and Doležel 1993), and pollen fertility was 95.4% (Heenan ). Therefore, male fertility of F1 derived from B. juncea × B. napus was lower than that of F1 from B. napus × B. juncea. However, only one report (Heenan ) described male fertility of F1 from B. napus × B. juncea, prompting further investigations to assess the occurrence of introgression. Investigations on the mechanisms underlying the variations in genome composition in F1 hybrids from each pollen parent of B. juncea and B. napus are warranted. Because the A–C chromosomal pairing affinity is higher than that of A–B and B–C in the F1 hybrid with AABC genome composition from B. juncea × B. napus (Mason ), hybrids of the combinations of the A genome derived from B. juncea and the C genome derived from B. napus might easily occur during hybridization between B. juncea and B. napus. Female fertility of the F1 hybrids was assessed by comparative analysis of the productivity of BC1 seeds using backcrossing (Table 4). Seed productivity of BC1 was remarkably reduced compared with that of both parents (Choudhary and Joshi 2001, Frello , Heenan , Kirti , Liu , Mathias 1985, Roy 1984, Schelfhout , Song ), which was indicative of the degradation of female and male fertility of F1.

Table 4

Production efficiencies of progenies in F1 by backcrossing and selfing

Cross-combinations	Seeds/pollination	References
F1 produced from B. juncea × B. napus
Backcrossing
B. juncea × F1	0.65–1.9	Song et al. (2010), Frello et al. (1995)
F1 × B. juncea	0.06–1.2	Song et al. (2010), Liu et al. (2010), Heenan et al. (2007)
B. napus × F1	NRa	Schelfhout et al. (2006), Prakash and Chopra (1990)
F1 × B. napus	0–1.6	Mathias (1985), Liu et al. (2010), Heenan et al. (2007)
Selfing (F1 × F1)	NR	Liu et al. (2010), Schelfhout et al. (2006), Roy (1984), Choudhary and Joshi (2001)
F1 produced from B. napus × B. juncea
Backcrossing
B. juncea × F1	NR	Schelfhout et al. (2006)
F1 × B. juncea	NR	Schelfhout et al. (2006), Kirti et al. (1995)
B. napus × F1	NR	Schelfhout et al. (2006)
F1 × B. napus	NR	Schelfhout et al. (2006)
Selfing (F1 × F1)	NR	Schelfhout et al. (2006), Roy (1984)

NR: not reported, progeny seeds were obtained, but the number of seed sets per pollinated flower was not described.

Fertility of BC1, BC2, and BC3 generations from F1 (B. juncea × B. napus) × B. juncea was reported by Song and Tsuda . Seed productivities of BC2 and BC3 were promptly recovered to the same level as those of the parent lines (Table 5). However, seed productivity of BC1 [(B. juncea × B. napus) × B. napus] was not reported. It is possible that BC2 produced by reciprocal cross-pollination between BC1 and B. napus was extremely low.

Table 5

Production efficiencies of progenies from BC1 derived from B. juncea × B. napus

Cross-combinations	Seeds/pollination	References
BC1 from B. juncea × F1
B. juncea × BC1	5.6–10.6	Song et al. (2010), Tsuda et al. (2012b)
B. juncea × BC2	5.8–16.0	Song et al. (2010), Tsuda et al. (2012b)
B. juncea × BC3	15.5–16.0	Song et al. (2010)
BC1 from F1 × B. juncea
BC1 × B. juncea	3.8	Song et al. (2010)
BC2 × B. juncea	3.9–15.0	Song et al. (2010)
BC3 × B. juncea	9.7–15.2	Song et al. (2010)
B. juncea × B. juncea	6.2–16.2	Song et al. (2010), Tsuda et al. (2012b)

On the other hand, F2 to F7 derived from F1 (B. juncea × B. napus) were obtained (Roy 1984), whereas F3 was not obtained from F1 ( B. napus × B. juncea) (Liu , Roy 1980).

Indirect introgression from B. napus to other Brassicaceae through B. juncea

Spontaneous hybridization between B. juncea and B. napus has been extensively examined (Bing , 1996, Heenan , Huiming , Jørgensen , Liu et al 2010, Tsuda ). However, Bing , 1996) reported that spontaneous hybridization of B. nigra × B. juncea and S. arvensis × B. juncea was unsuccessful, and no other reports on spontaneous hybridization between B. juncea and other members of the Brassicaceae have been published.

On the other hand, numerous reports have described the production of inter- and intra-specific hybrids by artificial pollination. Crossability and hybrid production efficiency between B. juncea and seven other species were evaluated previously, and the results are listed in Table 6. Progeny seeds were obtained from all cross-combinations, except S. pubescens × B. juncea. On the other hand, hybrid plants were not obtained in cross-combinations of B. oleracea × B. juncea, B. juncea × B. oleracea, and S. arvensis × B. juncea. Differences between seed/pollination and hybrid/pollination showed the occurrence of false hybrids in inter- and intra-specific crossings (Table 6).

Table 6

Interspecific and intergeneric cross compatibility by artificial pollination between B. juncea and other members of the Brassicaceae

Species	Parent status	Seeds/pollination		References	Hybrids/pollination		References
		Range	Average		Range	Average
B. nigra	Pollen	0.00–10.20	2.11	Mohammad and Sikka (1940), Bing et al. (1991, 1996), Prasad et al. (1997), GhoshDastidar and Varma (1999), Morinaga (1934)	0.03–0.10	0.06	Rao and Shivanna (1997), Bing et al. (1991, 1996)
	Seed	0.00–0.01	0.01	Mohammad and Sikka (1940), Bing et al. (1991, 1996), GhoshDastidar and Varma (1999)	0.01	0.01	Bing et al. (1991, 1996)
B. rapa	Pollen	0.00–2.96	0.39	Sharma and Singh (1992), Rhee et al. (1997), GhoshDastidar and Varma (1999), Kakizaki (1925), Song and Qiang (2003), Choudhary and Joshi (1999), Takeshita et al. (1980), Mohammad and Sikka (1940)	0.00–5.66	0.85	Rao and Shivanna (1997), Choudhary and Joshi (1999), Mohammad and Sikka (1940)
	Seed	0.00–10.40	0.07	Kakizaki (1925), GhoshDastidar and Varma (1999), Prasad et al. (1997), Choudhary and Joshi (1999), Yamagishi and Takayanagi (1982), Mohammad and Sikka (1940), Nishi et al. (1964), Rhee et al. (1997), Takeshita et al. (1980)	0.00–0.50	0.00	Choudhary and Joshi (1999), Mohammad and Sikka (1940), Nishi et al. (1964)
B. oleracea	Pollen	0.00–0.08	0.02	Rao and Shivanna (1997), Kakizaki (1925), GhoshDastidar and Varma (1999)	0.00	0.00	Kakizaki (1925)
	Seed	0.00–0.12	0.03	Yamagishi and Takayanagi (1982), Nishi et al. (1964), Kakizaki (1925), GhoshDastidar and Varma (1999)	0.00	0.00	Nishi et al. (1964)
B. carinata	Pollen	0.22–0.75	0.48	Sharma and Singh (1992), Mason et al. (2011b), GhoshDastidar and Varma (1999)	0.02–0.2	0.48	Mason et al. (2011b), Rao and Shivanna (1997)
	Seed	0.04	0.04	GhoshDastidar and Varma (1999), Getinet et al. (1997), Mason et al. (2011a)	–	–	–
R. sativus	Pollen	0.00–0.31	0.09	Kakizaki (1925), Rhee et al. (1997)	–	–	–
	Seed	0.00–0.51	0.14	Nishi et al. (1964)	0.00–0.06	0.03	Nishi et al. (1964)
S. arvensis	Pollen	0.07	0.07	Bing et al. (1991, 1996)	0.03	0.03	Bing et al. (1991, 1996)
	Seed	0.02	0.02	Bing et al. (1991, 1996)	0.00	0.00	Bing et al. (1991, 1996)
S. pubescens	Pollen	0.15	0.15	Inomata (1991)	0.05	0.05	Inomata (1991)
	Seed	0.00	0.00	Inomata (1991)	–	–	–

Although artificial pollination of S. pubescens × B. juncea did not generate seeds or hybrids, a few seeds or hybrid plants were obtained by artificial pollination of B. juncea × S. pubescens. However, these hybrids were barely produced under natural conditions (Inomata 1991). Furthermore, hybridization between S. pubescens and B. napus has not been reported in the literature. It is thus possible that S. pubescens could be a new candidate of indirect introgression from B. napus.

Cross-pollination between B. juncea and other relatives, such as B. oxyrrhina (Bijral and Sharma 1999b, Katiyar and Chamola 2007, Salisbury 1989), B. gravinae (Nanda Kumar ), D. erucoides (Bhat , Inomata 1998) as male and female parents, D. catholica (Banga ), D. berthautii (Bhat ), D. tenuifolia (Salisbury 1989), Sinapidendein fruticulosa (Herberd and McArthor 1980) as female parent, R. raphanistrum (Kamala 1983), E. sativa (Bijral and Sharma 1999a), Moricandia arvensis and B. cossoneana (Herberd and Mcarthor 1980) as male parent, produced seeds by artificial pollination, although the data on seed/pollinated were not reported. The following nine species, namely B. tounefortii (GhoshDastidar and Varma 1999, Goyal , Lokanadha and Sarla 1994), D. virgata (Herberd and McArthor 1980, Inomata 1994, 2003) as male and female parents, D. siifolia (Ahuja , Batra ), D. siettiana (Sarmah and Sarla 1995), Erucastrum abyssinicum (Rao , Sarma and Sarla 1997) as a female parent, Erucustrum virgatum (Inomata 2001), Orychophragmus violaceus (Li ), Crambe abyssinica (Wang and Luo 1998), and Enarthrocarpus lyratus (Gundimeda ) as a male parent, produced hybrids with B. juncea by tissue culture. However, these species have not been reported for hybrid production by artificial pollination. Moreover, hybridization between B. juncea and B. fruticulosa (used as male and female parents), D. siifolia (used as a male parent), E. sativa, and M. arvensis (used as a female parent) have not been reported by artificial pollination or tissue culture.

In this review, we summarized the possibility of introgression from B. napus to a closely related species through B. juncea. There has been no report of crossability and productivity of hybrids between B. juncea and other Brassicaceae species, except for the six species described above.

Conclusions and Remarks

B. juncea is currently considered as the species that generated hybrids with B. napus by spontaneous hybridization, artificial pollination, and tissue culture. Scheffler and Dale (1994) reported that B. juncea shows the second highest crossability with B. napus after B. rapa. Moreover, investigations on Hortus siccus revealed that the flowering periods of B. juncea and B. napus in Japan completely overlap (Matsuo and Itoh 2001) and that B. juncea is distributed in all prefectures of Japan (Abe , Hokkaido 2010, Kanai , Konta , MAFF 2012, MLIT 2005, National Museum of Nature and Science 2000, Tsuda and Tabei 2014). Under these conditions, some feral GM canola cultivars have been identified each year since it was initially reported in 2003 (MAFF 2012). Therefore, spontaneous hybridization between B. juncea and B. napus is possible. However, hybrids between B. juncea and GM canola have not been discovered, although hybrids between B. rapa and GM canola have been observed around feral GM canola (MAFF 2012). It is possible that B. juncea is an autogamous plant, and this trait minimizes cross-pollination under natural conditions. In fact, the maximum spontaneous hybridization between B. juncea and B. napus was 5.91% in mixed planting (Heenan ), and spontaneous hybrids were not detected among plants separated at distances of 20 m (Tsuda ). Moreover, Devos concluded that feral GM canola has not invaded areas outside its cultivated fields; these plants also occupy ruderal habitats because of its low spontaneous hybridization frequency and fertility. Introgression between crop and feral and between feral and wild relatives could therefore be very difficult.

On the other hand, Di reported that F1 hybrids from wild B. juncea × GM canola showed higher fertility than that reported previously (Bing , 1996, Frello , Tsuda ). Di also reported that the vigorous vegetative and reproductive growth of wild B. juncea resulted in higher fertility rates of the F1 hybrids. Because wild B. juncea consists of multiple genotypes, discussion on introgression potential should also consider genotypic variation, and further investigations should be conducted.