THE GENUS GVENTHERA ANDR. IN BESS. (BRASSICACEAE, BRASSICEAE)

GÓMEZ-CAMPO, C. (2003). The genus Guenthera Andr, in Bess. (Brassicaceae, Brassiceae). Anales Jard. Bot. Madrid 60(2): 301-307. A group of nine species -now included in Brassica— differ from all the other species in several characters, mainly in the stylar portion of their pistils always without seed primordia. Also in their branched subterranean stem (caudex) with several leaf rosettes, their leaves entire to deeply pinnatifid but never pinnatisect, their shallowly notched cotyledons and their flattened, elliptic or ovoid seed contour. It is suggested to include these species under the generic denomination Guenthera Andr, in Bess. New ñames for the species and subspecies are provided, as well as a determination key for the species.

The seedless stylar portion of Brassica nigra L. has always been a cause of confusion.In fact, the creation of sect.Melanosinapis DC. (under Sinapis) was largely endeavored to fit this odd species.However, it is evident that, apart from the lack of a beak, all other characters showed by B. nigra are those of Brassica, particularly those observable in sect.Micropodium DC. which includes for instance B. fruticulosa, B. cossoniana, B. maurorum, B. spinescens, etc. Everything suggests that B. nigra is a true Brassica, simply renouncing to its fertile beak during the evolutionary proccess leading to a strong shortening of its fruit.
When the species corresponding to the Pomel's groups Brassicaria and Nasturtiops are considered together and B. elongata is added, the resulting complex of species shows a series of important characters in common (table 1).In our opinion they deserve the aplication of a generic denomination (Guenthera).
DIFFERENTIAL CHARACTERS Among all the other differential characters, the absolute absence of seeds -or even seed primordia-within the stylar portion of the pistil should be emphasised (character 5 in table 1).The absence or presence of a seeded beak has always been considered a relevant character to distinguish genera within the tribe Brassiceae, where each option occurs in aproximately one half of the 55 existing genera.Species with and without a beak had been maintained together up to the present only in Brassica and Diplotaxis.The existence of a seeded beak (heteroarthrocarpy) is both phylogenetically and taxonomically important (G6MEZ-CAMPO, 1999b) because it is a singular morphogenetic achievement which appears exclusively within the evolutionary radiation of the tribe Brassiceae, while it is completely absent in all other cruciferous tribes.Guenthera (as Eruca, Sinapidendron, Vella, etc.) is anterior to this evolutionary development while Brassica (as Erucastrum, Coincya, Raphanus, etc.) is posterior.We avoid to call a sterile stylar portion "beak", since this is the general case with fruits in the Crucifer family.
The absence of a beak shows correlations with at least five other significant characters.The presence of a caudex is commonplace in Guenthera but never observable in Brassica.A caudex is a vertical buried stem branched under the soil surface, with leaf rosettes at soil level, at the tip of each branch.Leaf scars or leaf remainings demonstrate its caulinar condition.A caudex is like a buried bush and it represents an intermediate situation between chamaephytes and mono-rossette hemicriptophytes.The unique exception is the annual habit showed by Guenthera amplexicaulis.However, all other characters of this species are those of Gunthera and, particularly, G. amplexicaulis shows strong affinities to G. dimorpha.Apart from this case, Gunthera species are hemicriptophytes with several rosettes, although there might be a single one in the juvenile stage or sometimes in the adult stage (mostly in G. elongata).The scapiform habit for floriferous stems is very extended though species with scarcely branched leafy aerial stems also exist.Rosette-forming habit is already conspicuous in the seedling development, a stage where Guenthera is already easy to distinguish from Brassica.
The typical lyrate-pinnatisect leaf silhouette -so common in Brassica and most other related genera with seeded beak-, can never be found in Guenthera.Guenthera leaves are normally entire, lobed or pinnatifid, often deeply so.Only within the high variability present in the leaves of the G. repanda subspecies it is possible to find individual cases approaching a pinnatisect contour.POMEL (1860) already observed clear differences in cotyledon shape which can be "emarginate or not".Such differences are very patent and they were quantified by G6MEZ-CAMPO & TORTOSA (1974).In other words, the cotyledon notch in Guenthera is much shallower than in Brassica.The cotyledon petiole is also shorter.With respect to the contour of mature seeds, overlappings may exist among the species of both genera (BENGOECHEA & G6MEZ-CAMPO, 1975), but a tendency to sphericity is always much more marked and constant within the species of Brassica.
Cromosome number n = 11 is present in Guenthera and absent in Brassica.Chromosome number n = 10 exists in both genera but the deep morphological differences among the species belonging to Brassica and Guenthera (as for instance Brassica rapa and Guenthera repanda both with n = 10) suggest that it is a mere coincidence.
Not all molecular studies with Brassica include Guenthera species.However, dendrograms obtained by WARWICK & BLACK, (1991,1993) (2002) with sequences of nuclear genes do include a number of species here ascribed to Guenthera.Constantly, they appear clearly separated from typical Brassica species and, significantly, they are often closer to other taxa with asperm stylar portion belonging to genera as Eruca, Diplotaxis, etc.On the other hand, differences showed by G. elongata and G. amplexicaulis are still deeper, a fact that is correlated to other morphological trends and reinforces the distribution into sections proposed below.
Brassica balearica Koch has a seeded beak and its polyploid nucleus includes the n = 9 genome shown by B. oleracea and other close Mediterranean relatives (SNOGERUP & PERS- SON, 1983).However, its dwarfing facies and compressed seeds suggest that it might be the product of a possible intergeneric cross Brassica x Guenthera.Further cytogenetic studies would be necessary to elucidate this case.The position of Brassica somalensis Hedge & Miller-a suffruticose plant (without a caudex) with asperm stylar portion and ellipsoidal seeds-is more difficult to interpret.HEDGE & MILLER (1977) with c-DNA.by PRADHAN & al. (1992) with c-DNA and m-DNA, or by INABA & NISHIO

TABLE 1 MAIN
DIFFERENCES BETWEEN THE GENERA GUENTHERA AND BRASSICA , more rarely biennial (G.elongata, partly) or annual (G.amplexicaulis).Floriferous stems leafy or leafless growing annually from the rosettes.Leaves entire, sinuate, or deeply pinnatifid, never pinnatisect.Silique with a prominent mid-nerve in each valve and a stylar portion always sterile.Mature seeds somehow flattened, ovoid, ellipsoidal or subglobose, rarely globose.Cotyledons only very slightly notched.Chromosome number In = 20 or 22, with some cases of polyploidy.Three sections are distinguished and are defined by the characters which are expressed in the proper place.
Hemicriptophytes with a caudex (vertical subterranean branched stem) with several leaf * Besser explicitly attributes the generic name and description to Andr6anszky.rosettes