Heenan diversity.indd

New Zealand Journal of Botany, 2004, Vol. 42: 815–832 0028–825X/04/4205–0815 The Royal Society of New Zealand 2004 Diversity of Brassica (Brassicaceae) species naturalised in Canterbury, New Zealand hybridisation. One plant has a fl ow cytometry profi le and a high percentage of malformed pollen that are consistent with being the putative interspecifi c F1 hybrid B. napus × B. rapa. Another population in- Landcare Research cluded plants of the putative intraspecifi c hybrid B. rapa var. oleifera × B. rapa var. chinensis growing Lincoln, New Zealand with the two parental species.
Keywords naturalised plants; crop escapes; Bras-
sica; B. juncea; B. rapa; B. napus; B. oleracea; fl ow t Field surveys were undertaken of Brassi-
cytometry; hybrids; Canterbury fl ora c naturalised in Canterbury, investigating taxonom- ic diversity, morphological variation, distribution, and abundance of naturalised taxa, crop escapes, and crop–weed hybrids. As a result, six species, nine INTRODUCTION
varieties, and two natural hybrids were recognised The potential escape of genetically modifi ed (GM) as naturalised in Canterbury. B. rapa var. oleifera crops into the environment, hybridisation with wild was the most common taxon, while B. rapa var. relatives, and possible environmental effects of gene , B. rapa var. glabra, B. rapa var. rapa, B. transfer are increasingly being investigated world- var. oleifera, and B. oleracea var. acephala wide. The cultivation of GM crops in New Zealand is were less abundant and generally occured in small subject to a rigorous approval process, with applica- populations of only a few individuals in rural ar- tions being governed by the Hazardous Substances eas. Single plants of B. oleracea var. gongylodes and New Organisms Act and administered by the (kohl rabi) and B. napus var. napobrassica (swede) Environmental Risk Management Authority. The were collected. An additional species, B. juncea, is merits and risks of each GM crop application are recorded as a new naturalised plant in Canterbury. considered on a case-by-case basis.
Most of the Brassica populations sampled were We are seeking to evaluate environmental risks small, with 64% of the 107 populations observed due to gene fl ow from crops in New Zealand and comprising fewer than 10 plants.
to determine the ecological factors promoting gene Flow cytometry profi les were obtained for 168 fl ow by crop escape or hybridisation with wild rela- samples of B. juncea, B. napus, B. oleracea, and tives. Brassica was adopted as a case study because , and these usually confi rmed morphologi- of the numerous GM taxa available, and the fact that cal identifi cations to a particular species. However, several Brassica crops have been genetically engi- there were some identifi cation uncertainties with B. neered and fi eld trialled in New Zealand (Christey , as some plants had a fl ow cytometry profi le & Woodfi eld 2001). There is also an extensive lit- of that species but the morphology of either B. rapa erature on interspecific hybridisation. However, var. oleifera or B. oleracea var. acephala. Tetra- little information on the rates of crop escapes or of ploid plants of B. rapa (var. chinensis, var. oleif- the presence of compatible relatives exists in a New , and var. rapa) and B. oleracea var. acephala Zealand context, and this information is important were recorded. We also identifi ed two instances of for risk assessment. Brassica is also one of the most- studied GM crops in the world in regard to gene fl ow and invasiveness, and is considered to be a potential B04024; Online publication date 9 December 2004 worse-case scenario and, therefore, a critically im- Received 23 June 2004; accepted 19 August 2004 portant model system (Conner et al. 2003).
New Zealand Journal of Botany, 2004, Vol. 42 Our study is the fi rst step in addressing risk as- Plant identifi cation
sessment for Brassica by establishing accurate and Naturalised and crop species and varieties of Bras- contemporary information on the taxonomic and siscia a are usually reliably identifi ed by a specifi c set of morphological diversity, distribution, and abun- vegetative, fl oral, and fruit characters (e.g., Webb et dance of naturalised weed species, crop escapes, and al. 1988; Zhou et al. 2001). These include leaf shape, crop–weed hybrids. This biodiversity information is colour, and hairiness, petiole distinctiveness and a necessary prerequisite to ecological and molecular shape, open fl ower position relative to fl ower buds, studies for risk assessment of GM Brassica crops. petal length and colour, sepal and stamen orientation, We selected Canterbury in which to undertake a and fruit shape, orientation, and size. Specimens survey of naturalised Brassica, as Brassica crops were identifi ed in the fi eld on the basis of these are grown there for both seed certifi cation and forage morphological characters, and identifi cations were (Charlton & Stewart 2000; Stewart 2004), and sev- subsequently verifi ed by fl ow cytometry data.
eral species are known to be naturalised there (Healy To estimate male fertility of putative hybrid plants 1969; Webb et al. 1988). In this paper we report their pollen was stained with Alexander's Differen- on the current taxonomic diversity, morphological tial Stain (Alexander 1969) to establish whether the variation, distribution, and abundance of Brassica cytoplasm was normally developed. Five hundred taxa naturalised in Canterbury.
pollen grains per plant were counted. Pollen grains were considered normal if they were turgid and the cytoplasm fully stained red. Abnormal pollen grains were not rigid, having collapsed walls, lacking cy- toplasm, or the cytoplasm was shrunken, unstained, or only partially stained.
Nomenclature follows Zhou et al. (2001), a recent Field surveys of Brassica were undertaken during and comprehensive treatment of Br a completed September and October 2003 (Fig. 1) to coincide for the Flora of China series. The main differences with peak fl owering. We randomly selected 50 plots between Zhou et al. (2001) and Webb et al. (1988) in Canterbury using Arcview data for the region. are that wild turnip is treated in Zhou et al. (2001) as Plot dimensions were 3 × 3 km and their location B. rapa var. oleifera DC. rather than B. rapa subsp. was constrained to slopes of less than 10°, their sylvestris (L.) Janch., and intraspecifi c names are centres were greater than 2.5 km from the sea, they applied to B. napus, B. oleracea, and B. rapa.
did not overlap, were not in established urban areas, and had more than 3 km of roads. Within each plot Flow cytometry
we drove along all public roads recording Bras- Brassica species used for seed certifi cation, forage, sica plants on road verges, in and around drainage and with naturalised populations in Canterbury have ditches, channels, and natural watercourses, around different chromosome numbers and comprise dif- shelterbelts, and on wasteland. Samples were also ferent combinations of three genomes (A, B, and collected when travelling between plots. Aerial he- C) (U 1935). These genomes have different nuclear licopter surveys were undertaken of the Ashburton DNA content, which facilitates the reliable identi- and Rakaia rivers from the foothills to the coast and fi cation of Brassica species and hybrids using fl ow along the coast between these two rivers (Fig. 1), as cytometric analyses that estimate 2C nuclear DNA both fl ow through a large area of Brassica grown for content of individual plants (Sabharwal & Doležel seed certifi cation. We did not examine pastoral and 1993). This technique has previously been success- cropping land during this survey.
fully used to detect hybrids between B. napus and From each plot and collection sites between the B. rapa (Wilkinson et al. 2000; Jenkins et al. 2001). plots we collected plants for identifi cation, morpho- Freshly collected leaf material from wild populations logical examination, mapping distributions, and fl ow and cultivated specimens of known species and crop cytometric analysis. The number of individual plants cultivars were analysed. A hexaploid specimen of per population were counted. Herbarium vouchers Hieracium pilosella × H. caespitosum was used as representing plants collected during the fi eld survey a control standard of known ploidy. For each sample are deposited in the Allan Herbarium (CHR). Speci- the results are presented as the ratio of the Hieracium men searches were also undertaken at New Zealand value, which was standardised to 1.0.
herbaria (AK, CHR, and WELT) for other records Leaf material was fi nely chopped in 0.4 ml of of Brassica naturalised in Canterbury.
buffer for cell extraction and liberation. The cell Heenan et al.—Brassica Fig. 1 Distribution of Brassica species naturalised in Canterbury. Grey shading represents areas cultivated for
Brassica seed certifi cation crops 1990–2003.
New Zealand Journal of Botany, 2004, Vol. 42 suspension was then fi ltered through a 30-µm mesh RESULTS
into a sample tube and incubated for 2 min before the addition of 1.6 ml of DAPI stain. The buffer Field surveys
and stain are included in the proprietary Partec T kit Over half (58%) of the 50 plots surveyed did not (CyStain UV precise T (05-5003)). The cell suspen- contain any naturalised Brassica species, although sion was passed through a fl ow-cytometer (Partec a number of Brassica specimens were sometimes Ploidy Analyser PA-II) with the detector operating at collected from just outside the plot boundaries (Fig. 355 nm. A minimum of 5000 nuclei were counted for 1, 2); several plots had more than one population. In each sample, at a rate of about 30 nuclei per second. total we identifi ed 107 populations during the fi eld Data analysis was performed using PA-II's Partec surveys with more than half (56%) of these occur- FloMax software, and Gaussian curves were fi tted ring outside the plots. The survey covered 3050 km to the histograms.
by car, and an additional 166 km were travelled by helicopter during the Rakaia and Ashburton river surveys. Of the 107 populations recorded during the Plant material used for chromosome counts (Table survey, 27% were represented by one plant, 57% had 1) was grown in a greenhouse at Landcare Research, less than 5 plants, and 64% fewer than 10 plants (Fig. Lincoln. Voucher specimens are deposited in the 3). Ten (9.3%) of the populations had two or more Allan Herbarium (CHR).
species or varieties of Brassica (Table 2).
Mitotic chromosome preparations were made from root tips that had been pretreated with a mixture Flow cytometry and cytology
of 70 ppm cycloheximide and 250 ppm 8-hydroxy- We obtained chromosome counts and fl ow cyto- quinoline, fi xed in methanol-chloroform-propionic metric profi les for eight control samples of known acid (6:3:2), hydrolysed in 1 M hydrochloric acid, species and crop cultivars (Table 1). These provide macerated in 5% pectinase, and stained with Feulgen the basis on which we made identifi cation decisions and lactic-acetic-orcein. These techniques follow of fi eld-collected material using the ratio values of Dawson (1993).
the fl ow cytometry profi les. On the basis of fl ow Table 1 Controls for the fl ow cytometric analyses. *, Brassica values standardised as a ratio of the Hieracium
control (1.0).
Flow cytometry Chromosome Brassica type Indian mustard Rakaia, site P. B. Heenan & R. G. FitzJohn B. napus var. napus A. V. Stewart B. oleracea Portabello, P. N. Johnson var. capitata L.
B. oleracea A. V. Stewart var. acephala ‘Maris Kestrel' B. oleracea var. tetraploid kale cultivated A. V. Stewart acephala tetraploid B. rapa var. oleifera Portabello, P. N. Johnson B. rapa var. rapa A. V. Stewart ‘Barabas' tetraploid B. ‘Ceres Hunter' A. V. Stewart 2n = c. 20 CHR 569884 (rapa var. oleifera × B. rapa var. glabra)Hieracium pilosella × control R. A. Bicknell H. caespitosum

Heenan et al.—Brassica Fig. 2 Brassica rapa distribution.
New Zealand Journal of Botany, 2004, Vol. 42 Fig. 3 Population sizes for Brassica species naturalised in Canterbury. The rectangle areas for each graph sum to
Heenan et al.—Brassica Fig. 4 Flow cytometry data for individual plants assigned to four species and one putative hybrid. The large symbols
represent control plants for which chromosome counts were done. For each taxon the mean and standard error are indicated. Brassica values standardised as a ratio of the Hieracium control (1.0).
Table 2 Sympatry of Brassica species and varieties.
Site
B.r. oleifera B. oleracea East Ashburton 1 East Ashburton 2 cytometry profi les the samples were generally reli- counts documented here (Table 1) are the same as ably assigned to a species (Fig. 4), but within each those reported previously (e.g., Gómez-Campo & species we were not able to discriminate to varietal Hinata 1980).
rank (e.g., B. rapa; Fig. 5, 6). However, within spe- Flow cytometry profi les were obtained for 169 cies we identifi ed tetraploid plants of B. rapa (var. plants of Brassica, with 28 of these assigned to chinensis, var. oleifera, and var. rapa) and B. olera- B. juncea, 31 to B. napus, 16 to B. oleracea, and a var. acephala. For each species, the chromosome 93 to B. rapa. The remaining one sample had an New Zealand Journal of Botany, 2004, Vol. 42 Fig. 5 Flow cytometry data for individual plants assigned to different varieties of B. rapa. For each taxon the mean
and standard error are indicated.
unusual profi le, representing the F1 hybrid between naturalised plants of Brassica in Canterbury. Webb B. napus × B. rapa (Fig. 4, 6). In this study the only et al. (1988) provide a key and formal descriptions to uncertainty in using fl ow cytometric profi les for all of the species reported here, but not for varieties identifi cation was that tetraploid B. rapa (2n = 4x = within a species.
40; genome AAAA) and B. juncea (2n = 4x = 36; genome AABB) have very similar fl ow cytometry Brassica juncea (L.) Czern. var. juncea
profi les (Fig. 4, 6) and these species were best dis- Indian or brown mustard tinguished by their morphology.
REPRESENTATIVE SPECIMENS: west of Tinwald, P. B. Heenan & R. G. FitzJohn 64/8, 16 Oct 2003, CHR 568671; east of Ashburton, P. B. Heenan & R. G. FitzJohn 36/2, 14 Oct 2003, CHR 568562; NE of Rakaia, P. B. Heenan & R. G. FitzJohn 25/2, 14 Oct Six species and nine varieties (including the un- 2003, CHR 568543; NE of Rakaia, P. B. Heenan & named B. rapa var. rapa × B. rapa var. glabra hy- R. G. FitzJohn 17/3, 7 Oct 2003, CHR 568531; Ash brid), as well as two natural hybrids, are recognised burton, P. B. Heenan & D. Peltzer 104/1, 10 Nov as naturalised in Canterbury, and information for 2003, CHR 568670; near Methven, P. B. Heenan each of these is presented below. The taxonomic & R. G. FitzJohn 92/1, 30 Oct 2003, CHR 568653; treatment of Brassica for the Flora of China (Zhou Ashburton, P. B. Heenan & R. G. FitzJohn 65/1, et al. 2001) includes identifi cation keys and descrip- 16 Oct 2003, CHR 568623; SE of Ashburton, P. B. tions at the rank of species and variety, and these Heenan & R. G. FitzJohn 7/2/1, 14 Oct 2003, CHR were particularly useful for the identifi cation of 568556.
Heenan et al.—Brassica Fig. 6 Flow cytometry profi les of Brassica taxa investigated. *, Hieracium pilosella × H. caespitosum control.
New Zealand Journal of Botany, 2004, Vol. 42 DISTRIBUTION: Previously recorded from the North G. FitzJohn 13/2, 2 Oct 2003, CHR 568523; near Island in North Auckland (Webb et al. 1988). Col- Methven, P. B. Heenan & R. G. FitzJohn 56/1, 16 lected during this study from Canterbury where Oct 2003, CHR 568581; near Waimate, P. B. Heenan it is known from near Ashburton and Mt Somers. & R. G. FitzJohn 0/1/2, 23 Oct 2003, CHR 568641; B. juncea is almost certainly recently naturalised west of Christchurch, P. B. Heenan & R. G. FitzJohn in Canterbury and is reasonably common in the 56/1/1, 21 Oct 2003, CHR 568635; SE of Ashburton, Ashburton area, although it is not considered to be P. B. Heenan & R. G. FitzJohn 48/1/1, 14 Oct 2003, a common crop (Stewart 2004).
CHR 568557; near Dromore, P. B. Heenan & R. NOTES: B. juncea is distinguished by the rosette and G. FitzJohn 26/1, 14 Oct 2003, CHR 568545; near cauline leaves being subsessile or petiolate and not Ashburton, P. B. Heenan & R. G. FitzJohn 36/3, amplexicaul or auriculate, and bright yellow petals 14 Oct 2003, CHR 568560; Leeston, P. B. Heenan usually 0.7–1.1 mm long. B. juncea var. juncea is & R. G. FitzJohn 21/1, 7 Oct 2003, CHR 568538; naturally variable in a number of leaf characters north of Amberley, P. B. Heenan & R. G. FitzJohn (Zhou et al. 2001), and we recognise three different 22/1, 9 Oct 2003, CHR 568540; west of Ashburton, forms in Canterbury. The most common of these P. B. Heenan & R. G. FitzJohn 99/1, 30 Oct 2003, has green and toothed or lobed leaves. We have also CHR 568663; west of Rakaia, P. B. Heenan & R. G. made collections of a form with anthocyanin color- FitzJohn 45/1, 14 Oct 2003, CHR 568570.
ation in the leaves (e.g., CHR 568652) and others DISTRIBUTION: Occurs in the North Island and South with dissected leaves (e.g., CHR 568592).
Island where it is locally established on coasts, and is an occasional escape from cultivation on road- Brassica napus L.
sides and in cultivated fi elds (Webb et al. 1988). In NOTES: B. napus is distinguished by the leaves often Canterbury B. napus var. napus occurs in scattered being glaucous, open fl owers below the buds, and and small populations across the Canterbury Plains, petals usually 10–16 mm long and pale yellow or and it is generally not locally abundant. The largest creamy yellow. Within B. napus we identifi ed two population seen during the survey occurred near varieties as naturalised in Canterbury. Webb et al. Methven and comprised about 100 plants that were (1988) did not distinguish rape and swede, referring growing on a roadside verge with B. oleracea var. both to B. napus L.
acephala.
NOTES: B. napus var. napus varies considerably in Brassica napus var. napobrassica (L.) Rchb.
leaf and infl orescence characters and is sometimes swede diffi cult to identify using traditional morphological REPRESENTATIVE SPECIMEN: north of Mayfield, P. B. characters (e.g., using the keys provided by Webb Heenan & R. G. FitzJohn 98/1, 30 Oct 2003, CHR et al. (1988) and Zhou et al. (2001)). The leaves of some plants approach B. rapa in being green, semi- DISTRIBUTION: In Canterbury we collected it from a coriaceous, and with a prominent amplexicaul or single locality near Mayfi eld.
auriculate cauline leaf base, whereas other plants are more like B. a in that the leaves are glaucous, NOTES: B. napus var. napobrassica is distinguished from B. napus var. napus by its swollen root.
coriaceous, and with only slightly amplexicaul or auriculate bases. The height of fl ower buds to open Brassica napus L. var. napus
weed rape, rape fl owers, petal length, and petal colour are also unreli- able characters, and can approach and indeed overlap REPRESENTATIVE SPECIMENS: Lake Forsyth, E. Edgar & with either B. rapa (buds below open fl owers, petals K. R. West, 17 Sep 1975, CHR 276561; near Waima- 7–10 mm long, bright yellow) or B. oleracea (buds kariri River gorge, D. R. Given, 13 Oct 1976, CHR above open fl owers, petals 15–25 mm long, creamy 286329; NE of Rakaia, P. B. Heenan & R. G. Fitz- yellow). Fortunately, fl ow cytometry provides excel- John 18/1, 7 Oct 2003, CHR 568535; near Pleasant lent data for the identifi cation of B. napus (Fig. 4).
Point, P. B. Heenan & R. G. FitzJohn 88/1, 23 Oct 2003, CHR 568647; SW of Timaru, P. B. Heenan & Brassica nigra (L.) W.D.J.Koch black mustard
R. G. FitzJohn 60/1/3, 23 Oct 2003, CHR 568645; east of Dromore, P. B. Heenan & R. G. FitzJohn REPRESENTATIVE SPECIMEN: near Lincoln, Canterbury, 6/1/1, 14 Oct 2003, CHR 568550; SW of Glentunnel, T. Kirk, CHR 3431.
P. B. Heenan & R. G. FitzJohn 11/1, 30 Sep 2003, DISTRIBUTION: Occurs in the North Island in North CHR 568517; NW of Darfi eld, P. B. Heenan & R. Auckland, and in the South Island in Canterbury Heenan et al.—Brassica and Otago (Webb et al. 1988). In Canterbury it is Brassica rapa L.
known only from a single collection made by Kirk NOTES: B. rapa is distinguished by the open fl owers near Lincoln; this was probably made between Nov overtopping the buds; the basal leaves are green, 1881 and April 1884 when Kirk is known to have and the petals are 6–10 mm long. Within B. rapa we collected in the area (Hamlin 1965). It was not col- identifi ed four varieties and an unnamed intraspecifi c lected during this study.
hybrid as naturalised in Canterbury. Webb et al. NOTES: B. nigra is most easily recognised by the (1988) did not distinguish the different varieties of short (up to 25 mm long) and 4-angled fruit that B. rapa, although they did refer to wild turnip (var. are strongly appressed to the rachis, and the short oleifera) as B. rapa subsp. sylvestris.
pedicels (3–5 mm long).
Brassica rapa var. chinensis (L.) Kitam.
Brassica oleracea L.
pak choi, Chinese cabbage NOTES: B. oleracea is distinguished by glabrous, REPRESENTATIVE SPECIMENS: SW of Springston, P. B. glaucous, and coriaceous leaves, open fl owers well Heenan & R. G. FitzJohn 17/2/1, 17 Sep 2003, CHR below the buds, erect sepals, and petals usually 568503; south of Methven, P. B. Heenan & R. G. 15–25 mm long and creamy yellow. Within B. ol- FitzJohn 50/1/1, 14 Oct 2003, CHR 568568; SE of eracea we identifi ed two varieties as naturalised in Ashburton, P. B. Heenan & R. G. FitzJohn 7/3/1, Canterbury. Webb et al. (1988) did not distinguish 14 Oct 2003, CHR 568559; east of Ashburton, P. B. the different varieties of B. oleracea.
Heenan & R. G. FitzJohn 30/1, 14 Oct 2003, CHR Brassica oleracea var. acephala DC.
kale 568551; west of Tinwald, P. B. Heenan & R. G. FitzJohn 64/23, 16 Oct 2003, CHR 568598; west REPRESENTATIVE SPECIMENS: NE of Methven, P. B. of Tinwald, P. B. Heenan & R. G. FitzJohn 54/2/2, Heenan & R. G. FitzJohn 57/1, 16 Oct 2003, CHR 30 Oct 2003, CHR 568665.
568582; south of Temuka, P. B. Heenan & R. G. FitzJohn 80/1, 23 Oct 2003, CHR 568639; north of DISTRIBUTION: In Canterbury B. rapa var. chinensis Belfast, P. B. Heenan & R. G. FitzJohn 76/1, 21 Oct occurs in scattered and small populations across the 2003, CHR 568634; NE of Rakaia, P. B. Heenan & Canterbury Plains.
R. G. FitzJohn 25/1, 14 Oct 2003, CHR 568544; NOTES: B. rapa var. chinensis is distinguished by east of Ashburton, P. B. Heenan & R. G. FitzJohn the usually simple, entire, broadly elliptic basal 36/1, 14 Oct 2003, CHR 568561; NE of Rakaia, P. and lower cauline leaves and the petiole is usually B. Heenan & R. G. FitzJohn 17/6, 7 Oct 2003, CHR distinct, fl eshy and without wings.
568534; NE of Methven, P. B. Heenan & R. G. Fitz- John 93/1, 30 Oct 2003, CHR 568660.
Brassica rapa var. glabra Regel
DISTRIBUTION: In Canterbury B. oleracea var. aceph- ala occurs in scattered and small populations across REPRESENTATIVE SPECIMEN: near Prebbleton, P. B. the Canterbury Plains.
Heenan & R. G. FitzJohn 107/1, 17 Jun 2003, CHR NOTES: B. oleracea var. acephala can be confused with some forms of B. napus (see notes under B. DISTRIBUTION: In Canterbury we consider B. rapa var.
napus var. napus). Older herbarium collections of B. glabra to be a persistent crop escape in the area in oleracea (e.g., from near Amberley (CHR 286321) which it was cultivated as well as colonising adjacent and Lyttelton (CHR 68340)) are not able to be de- soil heaps; it has also been seen but not collected in termined to a particular variety due to the lack of a drainage ditch near Halswell, Christchurch (PBH pers. obs.).
glabra is distinguished by having Brassica oleracea var. gongylodes L.
kohlrabi NOTES: B. rapa var. more than 10 basal leaves, the leaves being obovate REPRESENTATIVE SPECIMEN: NW of Ashburton, P. B. to oblanceolate, and the petioles are fl attened and Heenan & R. G. FitzJohn 42/2, 14 Oct 2003, CHR with incised or toothed wings.
568567.
DISTRIBUTION: Collected during this study from a Brassica rapa var. oleifera DC.
single site near Ashburton.
NOTES: B. oleracea var. gongylodes is distinguished REPRESENTATIVE SPECIMENS: Wainono Lagoon, R. Ma- from other varieties of B. oleracea by the swollen son 9512, 28 Dec 1962, CHR 126856; Middleton, stem that produces the vegetable kohlrabi.
Christchurch, A. J. Healy 64/463, 10 Dec 1964, New Zealand Journal of Botany, 2004, Vol. 42 CHR 152088; New Brighton, Christchurch, B. P. J. Charing Cross, P. B. Heenan & R. G. FitzJohn 12/2, Molloy, 26 Oct 1970, CHR 212056; Prebbleton, W. 30 Sep 2003, CHR 568521; south of Rangiora, P. B. R. Sykes 411/91, 20 Oct 1991, CHR 474040; west Heenan & R. G. FitzJohn 16/2, 2 Oct 2003, CHR of Methven, P. B. Heenan & R. G. FitzJohn 4/4/1, 568527; west of Tinwald, P. B. Heenan & R. G. 16 Oct 2003, CHR 568576; west of Tinwald, P. B. FitzJohn 54/1/1, 16 Oct 2003, CHR 568586.
Heenan & R. G. FitzJohn 64/31, 16 Oct 2003, CHR DISTRIBUTION: Scattered individuals and small popu- 568603; west of Springston, P. B. Heenan & R. G. lations of fewer than fi ve plants of B. rapa var. rapa FitzJohn 17/4/5, 17 Sep 2003, CHR 568506; south are often seen in rural areas of Canterbury. We locat- of Springston, P. B. Heenan & R. G. FitzJohn 7/1, ed only one signifi cant population of more than 100 17 Sep 2003, CHR 568513; SW of Lincoln, P. B. plants, and this occurred on disturbed and sparsely Heenan & R. G. FitzJohn 2/3, 17 Sep 2003, CHR vegetated roadside gravels in an isolated area with 568500; outskirts of Christchurch, P. B. Heenan & no obvious seed source such as seed storage silos or R. G. FitzJohn 2/2/1, 21 Oct 2003, CHR 568624; farmyards; this population probably originated from outskirts of Halswell, Christchurch, P. B. Heenan seed spilt from farm machinery or a truck transport- & R. G. FitzJohn 73/3, 21 Oct 2003, CHR 568632; ing seed. We were not able to tell if this population south of Rangitata, P. B. Heenan & R. G. FitzJohn was naturally reproducing or whether new plants , 23 Oct 2003, CHR 568650; near Waimate, P. B. Heenan & R. G. FitzJohn 83/1, 23 Oct 2003, were germinating from the soil seed bank.
CHR 568642; near Leeston, P. B. Heenan & R. G. NOTES: Plants of B. rapa var. rapa are distinguished FitzJohn 21/2, 7 Oct 2003, CHR 568539; south of by their large and fl eshy taproot. They usually grow Woodend, P. B. Heenan & R. G. FitzJohn 24/2/1, 9 as biennials although may grow as annuals when Oct 2003, CHR 568542.
autumn-germinated seed is vernalised during the following winter. There is considerable variation in DISTRIBUTION: Occurs in the North Island and South Island where it can be a common and persistent weed the size, shape, and colour of the fl eshy taproot in of pastures, roadsides, waste places, and gardens wild-collected material.
(Webb et al. 1988). In our survey of Canterbury, Brassica rapa L. var. rapa × B. rapa var. glabra
B. rapa var. oleifera occured in scattered and small Regel
populations in rural areas where it was generally not locally abundant. It is most abundant and common REPRESENTATIVE SPECIMENS: SE of Ashburton, P. B. in and around the periphery of urban areas.
Heenan & R. G. FitzJohn 33/1, 14 Oct 2003, CHR 568558; west of Kirwee, P. B. Heenan & R. G. NOTES: B. rapa var. oleifera is distinguished by a FitzJohn 21/3/1, 30 Sep 2003, CHR 568516; west slender and cylindric taproot, having few basal of Kirwee, P. B. Heenan & R. G. FitzJohn 21/2/1, leaves (less than 10), and petioles that are slender 30 Sep 2003, CHR 568515; west of Charing Cross, and not winged.
P. B. Heenan & R. G. FitzJohn 12/3, 30 Sep 2003, Brassica rapa L. var. rapa
turnip CHR 568519; SW of Oxford, P. B. Heenan & R. G. FitzJohn 14/1, 2 Oct 2003, CHR 568525; SW REPRESENTATIVE SPECIMENS: Dromore, P. B. Heenan of Temuka, P. B. Heenan & R. G. FitzJohn 26/1/1, & R. G. FitzJohn 27/1, 14 Oct 2003, CHR 568548; 23 Oct 2003, CHR 568648; SW of Springston, P. Halswell, Christchurch, P. B. Heenan & R. G. Fitz- B. Heenan & R. G. FitzJohn 17/4/1, 17 Sep 2003, John 73/1, 21 Oct 2003, CHR 568630; near Dun- CHR 568510; SE of Ashburton, P. B. Heenan & R. sandel, P. B. Heenan & R. G. FitzJohn 79/1, 21 G. FitzJohn 7/2/5, 14 Oct 2003, CHR 568554.
Oct 2003, CHR 568638; NW of Ashburton, P. B. Heenan & R. G. FitzJohn 8/1/1, 14 Oct 2003, CHR DISTRIBUTION: Canterbury, scattered localities.
568564; near Rakaia, P. B. Heenan & R. G. FitzJohn NOTES: Turnip-rape is derived from crossing turnips 19/1, 7 Oct 2003, CHR 568537; north of Mayfi eld, (B. rapa var. rapa) with related Asiatic leaf veg- P. B. Heenan & R. G. FitzJohn 59/1, 16 Oct 2003, etables (B. rapa var. glabra) (Charlton & Stewart CHR 568584; near Waimate, P. B. Heenan & R. G. 2000). These are fast-growing and non-bulb-produc- FitzJohn 39/1/1, 23 Oct 2003, CHR 568640; west ing leafy turnips that produce high leaf quality and of Methven, P. B. Heenan & R. G. FitzJohn 53/1, have excellent regrowth after being browsed. The 16 Oct 2003, CHR 568578; west of Methven, P. B. higher leaf density and better regrowth of turnip-rape Heenan & R. G. FitzJohn 4/5/1, 16 Oct 2003, CHR cultivars may mean that they are able to compete and 568579; west of Methven, P. B. Heenan & R. G. persist among other naturalised plants better than FitzJohn 4/1/1, 16 Oct 2003, CHR 568573; west of other cultivars and varieties of B. rapa other cultivars and varieties of Heenan et al.—Brassica Cultivars of turnip-rape grown in Canterbury Heenan & R. G. FitzJohn 64/9, 16 Oct 2003, CHR include ‘Pasja' and ‘Ceres Hunter', and the wild-col- 568617. lected specimens are very similar to these cultivars in that the leaves are oblanceolate and with distinctly DISTRIBUTION: Known from a single site near Tinwald, winged petioles.
NOTES: The basal and lower cauline leaves of the Brassica rapa var. chinensis × B. rapa var. oleifera mediterranean mustard hybrid are intermediate between those of the putative REPRESENTATIVE SPECIMENS: Timaru, A. J. Healy 56/255, parents. The basal leaves of the hybrids are simple 20 Dec 1956, CHR 92184; Addington, Christchurch, and with the margin dentate, crenate, or sinuate, they A. J. Healy 59/261, 20 Mar 1959, CHR 121496; are distinctly petiolate, and the base of the petiole is Wetheral, North Canterbury, A. J. Healy 59/343, 13 sometimes slightly auriculate or amplexicaul (Fig. Apr 1959, CHR 121623; Papanui, Christchurch, A. J. 7). B. a var. chinensis has simple, broadly elliptic, Healy 59/665, 29 Dec 1959, CHR 122772; Linwood, entire, and distinctly petiolate leaves, and B. rapa Christchurch, A. J. Healy 63/358, 3 Dec 1963, CHR var. oleifera has lyrate-pinnatifi d leaves with sinuate 143862; Kaianga, Waimakariri River, A. J. Healy or toothed margins and 1–3 pairs of lateral lobes, and 64/127, 13 Aug 1964, CHR 143966; Rangiora, North the petiole is auriculate or amplexicaul at the base. Canterbury, A. J. Healy 59/351, 13 Apr 1959, CHR Pollen stainability of these hybrids varies between 121625; New Brighton, Christchurch, A. E. Wright 97% and 100%. About 20 plants of this putative 6002, 4 Nov 1983, CHR 396815.
hybrid were identifi ed in the fi eld near Tinwald. Intraspecifi c crosses among the different botani- DISTRIBUTION: Occurs in the North Island in North Auckland, South Auckland, Taranaki, and Welling- cal varieties of B. rapa are easily undertaken and ton, and in the South Island in Canterbury and Marl- particularly fertile, and hybrid plants typically have borough (Webb et al. 1988). Known in Canterbury pollen stainability of 73–97%, and usually above from Timaru and Christchurch, and several areas just 90% (Olsson 1954, and references therein). In par- to the north of Christchurch. B. tournefortii was not ticular, wild turnip (var. oleifera) crosses easily and collected during this study.
spontaneously with cultivated turnip (var. rapa) and turnip-rape (e.g., Davey 1939).
Natural hybrids
Brassica napus × B. rapa
Habitats
In this study we observed that the species and variet-
REPRESENTATIVE SPECIMEN: SW of Springston, Canter- bury, P. B. Heenan & R. G. FitzJohn 17/1/1, 17 Sep ies of Brassica that are naturalised in Canterbury are 2003, CHR 568502.
mainly weeds of disturbed roadside verges, drainage ditches, and hedges, at the approaches to culverts DISTRIBUTION: A single plant was identified near and bridges, at the bases of telegraph poles and road signs, on gravel ballast along railway tracks, NOTES: A single plant of this hybrid was identified and in wasteland and on soil stockpiles. They are by fl ow cytometry (Fig. 4, 6). This hybrid is prob- particularly common along road verges where they ably between B. napus var. napus and B. rapa var. most frequently occur in the narrow zone of loose oleifera, both of which occur in the Springston area. gravel between the consolidated road and the usu- Pollen of the putative hybrid plant stained with ally densely grassed berm. At these sites Brassica Alexander's Differential Stain showed considerable usually occur in association with other introduced irregularity in pollen development, with 75% of the annual and short-lived perennial dicotyledon weeds grains having abnormal or aborted cytoplasm.
such as Capsella bursa-pastoris (L.) Medik., Ce- rastium glomeratum Thuill., Lepidium didymus L., Brassica rapa var. chinensis × B. rapa var.
Sisymbrium offi cinale (L.) Scop., Stellaria media (L.) Cirillo., Vicia sativa L., and Viola arvensis Mur- REPRESENTATIVE SPECIMENS: west of Tinwald, P. B. ray. This suite of short-lived species refl ects regular Heenan & R. G. FitzJohn 64/13, 16 Oct 2003, CHR seasonal disturbance and exposed soil and open 568620; west of Tinwald, P. B. Heenan & R. G. habitat for Brassica to establish and mature. Bras- FitzJohn 64/10, 16 Oct 2003, CHR 568588; west sica rarely occur in stable fertile habitats dominated of Tinwald, P. B. Heenan & R. G. FitzJohn 64/14, by introduced grasses, except where they established 16 Oct 2003, CHR 568604; west of Tinwald, P. B. synchronously.
New Zealand Journal of Botany, 2004, Vol. 42 Brassica naturalised in Canterbury (Healy 1969; Webb et al. 1988).
Brassica rapa var. oleifera is the most common and widespread naturalised Brassica in New Zea- land. In Canterbury it is primarily an urban and urban-periphery weed and occurs where there is frequent disturbance due to road maintenance, more intensively managed roadside verges and drainage ditches, and ongoing land development (e.g., hous- ing and industrial subdivision). B. rapa may also occur more frequently around the urban-periphery where market gardening has perhaps provided a seed source and seed bank for subsequent inva- sions. This may in part explain the prevalence of naturalised B. rapa on the periphery of Christ- church, for example in the vicinity of Halswell, Lincoln, Prebbleton, and Springston. There are Fig. 7 Leaf profi les of the putative B. rapa intraspecifi c generally few substantial populations of B. rapa
hybrid and its parents. A, var. chinensis; B, var. chinensis
var. oleifera in rural areas due to the established × var. oleifera; C, var. oleifera. Scale bar = 10 mm.
and stable dense sward of introduced grasses along road verges and berms, drainage ditches, and fence lines. In New Zealand, B. napus has previously been considered to be mainly a coastal species (e.g., Allan 1940; Webb et al. 1988), and the reasonable number of inland Canterbury occurrences reported here for B. napus var. napus may be a result of Historical and current distributions
breeding for increased hardiness (Charlton & Stew- The present study recognises six species of Bras- art 2000).
sica naturalised in Canterbury. B. rapa is the most Our study of Brassica naturalised in Canter- variable and common species with var. chinensis, bury differs from early accounts that described the diploid var. oleifera, tetraploid var. oleifera, and naturalised plants of Christchurch (Armstrong 1872) var. rapa occurring in the wild. Population sizes and Ashburton County (Smith 1904). Among the vary, but most rural populations of var. chinensis naturalised plants listed for these areas were sev- and var. rapa comprised only a few individuals (Fig. eral species of Brassica and Sinapis, a genus to 3). Populations of B. rapa var. oleifera ranged from which some species of Brassica are often assigned. a few individuals to several thousand plants, and Armstrong (1872) and Smith (1904) both recorded the tetraploid var. oleifera plants usually occurred B. napus, B. oleracea, and B. rapa (including its in populations that mainly comprised diploid var. synonym B. campestris). In the Christchurch area, oleifera. B. napus var. napus and B. oleracea var. Armstrong (1872) reported that these three species acephala are less abundant and generally occur were observed as occasional individuals but were in small populations of only a few individuals in not common on cultivated land. Armstrong (1872) rural areas, and only single plants of B. oleracea also recorded Sinapis arvensis L. (= Brassica kaber var. gongylodes (kohlrabi) and B. napus var. na- (DC.) L.C.Wheeler) as being very abundant along pobrassica (swede) were collected. An additional roadsides in the Christchurch area. In the Ashburton species, B. juncea, is a new naturalised plant record area, Smith (1904) also listed Brassica alba (L.) for Canterbury where it occurs near Ashburton and Rabenh. (= Sinapis alba L.), Sinapis arvensis (in- Mt Somers. Although B. juncea has been considered cluding Brassica sinapistrum Boiss.), and Sinapis to be an uncommon crop in Canterbury (Stewart nigra L. (= Brassica nigra). These historical records 2004), its presence as a naturalised plant at a number of S. arvensis and B. nigra, as well as the more of sites suggests it has naturalised relatively quickly recent records of B. tournefortii, are particularly and is likely to become more common in the future. interesting as although these species are known from Although not collected during the present study, B. Canterbury (Webb et al. 1988), and are common and tournefortii and B. nigra are additional species of often invasive weeds of cultivated areas, roadsides, Heenan et al.—Brassica and waste places overseas (Holm et al. 1979), we bank are needed to determine their role in the es- did not fi nd them during this survey. Unfortunately, tablishment of naturalised Brassica.
there are no known herbarium specimens to verify the identifi cations of Armstrong (1872) and Smith Flow cytometry and morphological variation
Flow cytometry proved particularly useful in dis- It is important to note that the present survey tinguishing the species of Brassica naturalised in examined only roadsides, waterways, and waste Canterbury, as B. juncea, B. napus, B. oleracea, and places and did not survey pastoral and cropping land. B. rapa all have different chromosome numbers and In this regard we sometimes observed volunteer nuclear DNA contents (Fig. 4, 6). The only problem plants of Brassica in pastures, but did not record encountered was in distinguishing B. juncea (2n = this. Interestingly, we did not record any popula- 4x = 36; genomes AABB) from tetraploid B. rapa tions of volunteers spreading from their paddock (2n = 4x = 40; genomes AAAA). However, they are into adjacent roadside verges and drainage ditches. reliably identifi ed by their leaf morphology (Webb et On agricultural land in Canterbury, B. rapa var. al. 1988; Zhou et al. 2001). A number of tetraploid oleifera has been considered to be a common weed plants of B. rapa var. oleifera were detected by fl ow in the past (Hilgendorf & Calder 1952; Symons cytometry, and these usually occurred among popu- 1956). Smith (1904) considered that "B. campestris lations of diploid B. rapa var. oleifera. The origin [= B. rapa] and B. sinapistrum [= B. kaber] have of these plants is diffi cult to determine, although long been great scourges, and have in certain years they are most likely to be crop escapes as tetraploid taken almost complete possession of large cultivated fodder turnips are grown in Canterbury (Charlton areas." Later, Healy (1969) observed on cultivated & Stewart 2000; Stewart 2004). Tetraploid turnips land on the Canterbury Plains that B. rapa (as B. (e.g., B. rapa var. rapa ‘Barabas' and ‘Marco') are campestris) was one of the most widespread and also grown in New Zealand (Charlton & Stewart abundant species. Today, B. rapa var. oleifera is not 2000), and we identifi ed four tetraploid turnips by considered to be a signifi cant weed of agricultural fl ow cytometry. We also collected several plants of land (Bourdôt et al. 1999), and in wheat and barley tetraploid kale ( oleracea var. acephala) growing crops in Canterbury it occurs as volunteers at a rate amongst diploid kale, and this is of interest as tetra- of <0.5 plant/m2 (Bourdôt et al. 1998). B. napus often ploid kale is not considered to have been grown in grows as a volunteer in agricultural areas where it Canterbury (Stewart 2004). Either there is tetraploid occurs at a rate of 1 plant/m2 in wheat and barley kale grown or these plants are autotetraploids.
fi elds (Bourdôt et al. 1998), and it often occurs as We encountered some identifi cation diffi culties a volunteer in crops grown in rotation with oilseed involving B. napus. Firstly, some naturalised plants rape (Bourdôt et al. 1999).
were initially identifi ed by their morphology as B. The presence of Brassica along some road verg- napus var. napus, as they had glaucous leaves, open es in rural areas may be due to farming practices, flowers below buds, and petals 10–16 mm long whereby farm machinery used for drilling and har- (Webb et al. 1988; Zhou et al. 2001). However, these vesting seed is the prime dispersal agent while in plants had fl ow cytometry profi les consistent with B. transit between farm paddocks along public roads. rapa, which usually has green leaves, open fl owers This explanation was offered on one occasion by a overtopping buds, and petals 6–10 mm long (Webb farmer to explain the origin of three plants estab- et al. 1988; Zhou et al. 2001). Other plants had the lished on a road verge near Geraldine. Wind disper- morphology of B. rapa but fl ow cytometry profi les sal may also be locally important as it was also of B. napus. The origin of these plants is unknown, reported to us that in the Geraldine area during 2002 although plant breeders, in improving disease re- a very strong north-west wind blew seed from a cut sistance and production, have crossed B. napus and but not harvested crop of B. rapa var. chinensis B. rapa and then backcrossed the progeny to one of throughout a rural neighbourhood subsequently the parental species (e.g., Namai et al. 1980; Olsson producing hundreds of plants in gardens and on & Ellerström 1980). Depending on the traits being roadsides. An alternative and possibly important selected for and the inheritance of other characters, seed source may be provided by the soil seed bank. the resulting plants could comprise the chromo- At a number of sites it is possible that plants origi- some number and hence fl ow cytometry profi le of nated from seed in the seed bank rather than from one parental species but have inherited some of the recently dispersed or distributed seed. Studies of morphological characteristics of the other parental farming practices relevant to Brassica and the seed species. Several such hybrid cultivars have been New Zealand Journal of Botany, 2004, Vol. 42 raised in New Zealand including, for example, the % (Jenkins et al. 2001). The reciprocal interspecifi c turnip ‘Kapai' and the swede ‘Kiri' (Palmer 1983).
cross B. napus ♀ × B. rapa ♂ gave 10.8% hybrid Secondly, we also encountered difficulties in progeny (Palmer 1962).
identifying number ( ol- The morphology of the B. napus × B. rapa hy- eracea var. acephala) or ( napus var. napus), brid was very much like other plants of B. rapa var. as some of the morphological characters contra- oleifera collected from Canterbury, and without the dicted the fl ow cytometry profi les. B. oleracea has identifi cation of the putative hybrid by fl ow cytom- petiolate or sessile upper cauline leaves (compared etry it would have been very diffi cult to detect using with auriculate or amplexicaul in B. napus), the pet- leaf, fl ower, and infl orescence characters. B. rapa als are 15–25 mm long (cf. 10–16 mm long), and the var. oleifera is naturalised in the general area where sepals and stamens erect (cf. spreading) (Webb et al. the hybrid was collected but B. napus was not seen 1988; Zhou et al. 2001). Interspecifi c hybridisation at all, although it is grown in the vicinity as a fodder between these two species by plant breeders has crop. The possible origin of this hybrid is that it is also been used to improve disease resistance and derived from hybridisation of a crop B. napus with production (e.g., Palmer 1983), and this is the most a naturalised B. rapa var. oleifera. If this is the case, likely explanation to account for discrepancies be- this hybrid provides evidence of gene fl ow from a tween the fl ow cytometry profi les and morphological crop to a wild population and this may represent the early formation of a crop–weed hybrid complex in In regard to the problems of identifi cation involv- the area.
ing B. napus, this species is an allotetraploid (2n = A population from near Tinwald was particularly 4x = 38) derived from B. oleracea (2n = 2x = 18) × unusual in that it contained three different forms of B. rapa (2n = 2x = 20) (U 1935). Many of the mod- B. rapa (diploid var. oleifera, tetraploid var. oleif- ern B. napus cultivars have been artifi cally bred by era, and var. chinensis), B. juncea, and the putative crossing selected forms of the two parental species intraspecifi c hybrid B. rapa var. oleifera × B. rapa (Olsson & Ellerström 1980). Therefore, the hybrid var. chinensis. Plants representing the putative hy- origin of B. napus may also account for some of the brid B. rapa var. oleifera × B. rapa var. chinensis morphological variation present in naturalised B. na- were distinguished by the lower cauline leaves being pus and the diffi culty of distinguishing some forms simple, ovate, and entire or dentate, the petiole being of that species from either B. rapa or B. oleracea.
distinct, and the auricles being small and incon- spicuous (Fig. 7). These plants had a fl ow cytom- etry profi le consistent with being an intraspecifi c B. Flow cytometry identifi ed a single plant with a nu- rapa cross, and are therefore unlikely to be derived clear DNA content consistent with being an F1 hybrid from B. rapa × B. juncea, also a known and fertile of B. napus × B. a as the fl ow cytometry profi le of cross (Choudhary et al. 2002). At the Tinwald site this plant is intermediate between those of the puta- the putative hybrid plants were directly associated tive parents (Fig. 4, 6). From this it can be inferred with two subpopulations that included plants of B. that this hybrid plant is probably a sesquiploid with a rapa var. chinensis and B. rapa var. oleifera grow- chromosome number of 2n = 29, this being interme- ing only a few centimetres apart, and hybrid plants diate between the chromosome numbers of B. rapa were not found to occur in other subpopulations that (2n = 20) and B. napus (2n = 38). A chromosome contained only one of the putative parents.
number of 2n = 29 was obtained for F1 plants from the interspecifi c cross B. napus and B. rapa by Lu et al. (2001). The hybrid plant documented here has a low level of stainable pollen (25%), and this is at the SUMMARY AND CONCLUSIONS
lower end of the range (12.9–89.3%) of pollen stain- This study has shown that there is considerable taxo- ability reported from a set of interspecifi c crosses nomic and morphological diversity of Brassica spe- between different cultivars of B. napus and B. rapa cies and varieties naturalised in Canterbury, and that (Lu et al. 2001). In New Zealand, in experimental crop escapes are an important part of this diversity. fi eld/crop conditions, natural hybrids between B. However, there is considerable disparity between rapa var. oleifera ♀ (wild turnip) and B. napus ♂ phenotype and genotype in some species, particu- (rape) have previously been documented to occur at larly B. napus. Much of this disparity is most likely frequencies of 87.9 % (Palmer 1962) and 0.06–2.13 to be due to the breeding and selection of particular Heenan et al.—Brassica traits through the hybridisation of B. napus with B. ACKNOWLEDGMENTS
rapa and B. oleracea. Some of the traits selected by We thank Peter Johnson, Alan Stewart, Ross Bicknell, plant breeders may offer certain advantages for the and Evan Johnston for providing plant material. Tony naturalisation of Brassica including, for example, Conner, Linda Newstrom, Arthur Healy, Duane Peltzer, increased hardiness in B. napus and the improved and Bill Sykes for discussion and/or advice; Steve Ferris and resprouting ability of turnip-rape (B. for assistance with maps; and Ross Bicknell for assistance with the fl ow cytometry. Linda Newstrom, Duane Peltzer, In Canterbury, the distribution of naturalised Bill Lee, Carol West, and Christine Bezar are thanked for Brassica would best be described as sparse and critically reading the draft manuscript.
infrequent, and each population usually comprised only a few individuals. In rural areas many of the small populations appear to be ephemeral and com- REFERENCES
prise casual crop escapes that do not form persistent natural populations. The lack of recent collections Allan, H. H. 1940: A handbook of the naturalised fl ora of of B. nigra and B. tournefortii suggests that these New Zealand. Wellington, Government Printer.
species have not been successful in establishing Alexander, M. P. 1969: Differential staining of aborted themselves as naturalised species in Canterbury. In and nonaborted pollen. Stain Technology 44: contrast, the collections of B. juncea made during this survey indicate that this species has become Armstrong, J. F. 1872: On the naturalised plants of established as a fully naturalised plant in a relatively the province of Canterbury. Transactions and short time. The most common Br a is wild turnip Proceedings of the New Zealand Institute 36: (B. rapa var. oleifera), and the largest populations of this variety usually occur on the urban periphery Bourdôt, G.; Hurrell, G. A.; Saville, D. J. 1998: Weed where there is regular disturbance.
fl ora of cereal crops in Canterbury, New Zealand. Our fi eld observations suggest that disturbance New Zealand Journal of Crop and Horticultural Science 26: 233–247.
and the open habitats that this often creates are among the most crucial factors in the establishment Bourdôt, G.; Conner, T.; Jenkins, T. 1999: Ecological of naturalised Brassica. We also suspect that farm- risks and managerial consequences of Roundup ReadyTM oilseed rape in New Zealand. Lincoln, ing practices such as seed harvesting and transport AgResearch and Crop & Food Research. 67 p.
and the movement of farm machinery along roads are important in the distribution of Brassica in the Charlton, D.; Stewart, A. 2000: Pasture and forage plants rural landscape.
for New Zealand. Grassland Research and Prac- tice Series No. 8. 74 p.
Sympatry is not a common occurrence and there were few populations with more than two species or Choudhary, B. R.; Joshi, P.; Rama Rao, S. 2002: Cytoge- varieties. Nevertheless, the recognition of interspe- netics of Brassica juncea x Brassica rapa hybrids and patterns of variation in the hybrid derivatives. cifi c and intraspecifi c hybridism confi rms that gene Plant Breeding 121: 292–296.
fl ow can occur in two ways in the species of Brassica naturalised in Canterbury: (1) between crop plants Christey, M.; Woodfi eld, D. 2001: Coexistence of geneti- that have naturalised (B. rapa var. chinensis) and cally modifi ed and non-genetically modifi ed crops. Unpublished Crop & Food Research Confi dential their naturalised wild relatives (B. rapa var. oleif- Report No. 427. Crop & Food Research, Lincoln, era); (2) between naturalised wild plants (B. rapa New Zealand.
var. oleifera) and crop plants (B. napus). The two Conner, A. J.; Glare, T. R.; Nap, J.-P. 2003: The release of instances of hybridism reported here are considered genetically modifi ed crops into the environment. to be a signifi cant fi nding given the few occurences Part II. Overview of ecological risk assessment. of sympatry (Table 1) and small numbers of plants The Plant Journal 33: 19–46.
that usually occur in naturalised populations (Fig. Davey, V. M. 1939: Hybridisation in Brassicae and the 3). As B. rapa var. oleifera is the most widespread occasional contamination of seed stocks. Annals and common naturalised species, it is not surprising of Applied Biology 26: 634–636.
that it is involved in the two instances of hybridisa- Dawson, M. I. 1993: Contributions to a chromosome tion. It is anticipated that further surveys of Brassica atlas of the New Zealand fl ora — 31. Clematis are very likely to identify additional occurrences of (Ranunculaceae). New Zealand Journal of Botany 31: 91–96.
New Zealand Journal of Botany, 2004, Vol. 42 Gómez-Campo, C.; Hinata, K. 1980: A check list of Palmer, T. P. 1962: Population structure, breeding system, chromosome numbers in the tribe Brassiceae. inter-specifi c hybridisation and allopolyploidy. In: Tsunoda, S.; Hinata, K.; Gómez-Campo, C. Heredity 17: 278–283.
ed Brassica crops and wild allies. Tokyo, Japan Palmer, T. P. 1983: Forage brassicas. In: Wratt, G. S.; Scientifi c Societies Press. Pp. 51–63.
Smith, H. C. ed. ed Plant breeding in New Zealand. Hamlin, B. 1965: Itinerary of Thomas Kirk's botanical Wellington, DSIR. Pp. 63–70.
expeditions. Records of the Dominion Museum Sabharwal, P. S.; Doležel, J. 1993: Interspecifi c hybridiza- 5: 93–100.
tion in Brassica: application of fl ow cytometry Healy, A. J. 1969: The adventive fl ora of Canterbury. In: for analysis of ploidy and genome composition in Knox, G. A. ed e The natural history of Canterbury. hybrid plants. Biologia Plantarum 35: 169–177.
Wellington, Reed. Pp. 261–333.
Smith, W. W. 1904: Plants naturalised in the County of Ashburton. Transactions and Proceedings of the Hilgendorf, F. W.; Calder, J. W. 1952: Weeds of New New Zealand Institute 36: 203–225.
Zealand and how to control them. 5th ed. Christ- church, Whitcombe & Tombs. 260 p.
Stewart, A. V. 2004: A review of Brassica species, cross- pollination and its implications for pure seed Holm, L.; Pancho, J. V.; Herberger, J. P.; Plucknett, D. L. production in New Zealand. Proceedings of the 1979: A geographical atlas of world weeds. New New Zealand Agronomy Society 32: 63–82.
York, John Wiley.
Symons, F. L. 1956: Weed problems in South Canterbury. Jenkins, T. E.; Conner, A. J.; Frampton, C. M. 2001: Proceedings of the 9th New Zealand Weed Control Investigating gene introgression from rape to Conference. Pp. 7–12.
wild turnip. New Zealand Plant Protection 54: U, N. 1935: Genome analysis in Brassica with special reference to the experimental formation of B. na- Lu, C.; Shen, F.; Hu, K. 2001: Heterosis in interspecifi c pus and peculiar modes of fertilisation. Japanese hybrids between Brassica napus and B. rapa. Journal of Botany 7: 389–452.
SABRAQ Journal of Breeding and Genetics 33: Webb, C. J.; Sykes, W. R.; Garnock-Jones, P. J. 1988: Flora of New Zealand. Vol. IV. Christchurch, Botany Division, DSIR. 1365 p.
Namai, H.; Sarashima, M.; Hosoda, T. 1980: Interspecifi c and intergeneric hybridization breeding in Japan. Wilkinson, M. J.; Davenport, I. J.; Charters, Y. M.; Jones, In: Tsunoda, S.; Hinata, K.; Gomez-Campo, C. A. E.; Allainguillaume, J.; Butler, H. T.; Mason, D. ed Brassica crops and wild allies. Tokyo, Japan C.; Raybould, A. F. 2000: A direct regional scale Scientifi c Societies Press. Pp. 191–203.
estimate of transgene movement from genetically modifi ed oilseed rape to its wild progenitors. Mo- Olsson, G. 1954: Crosses within the campestris group of lecular Ecology 9: 983–991.
the genus Brassica. Hereditas 40: 398–419.
Zhou, T.; Lu, L.; Yang, G.; Al-Shehbaz, I. A. 2001: Bras- Olsson, G.; Ellerström, S. 1980: Polyploidy breeding in sicaceae (Cruciferae). In: Wu, Z.; Raven, P. H. ed. Europe. In: Tsunoda, S.; Hinata, K.; Gomez-Cam- Flora of China volume 8. Beijing, Science Press, po, C. ed. ed Brassica crops and wild allies. Tokyo, and St. Louis, Missouri Botanical Garden Press. Japan Scientifi c Societies Press. Pp. 167–190.
Pp. 1–193.

Source: https://richfitz.github.io/pubs/Heenan-2004-815.pdf