Systematics of Nama ( Hydrophyllaceae )

Lemmonia has been maintained as a monotypic genus, despite the fact that it is morphologically very similar to several species of Nama. The major basis for this separation resides in the reportedly coalescent filament bases of Lemmonia as contrasted with the distinct filaments of Nama. Examination of corollas of Lemmonia with the scanning electron microscope has revealed that the filaments of the taxon are distinct. There is, then, no qualitative basis upon which to found Lemmonia, and the taxon is transferred to Nama.


INTRODUCTION
The genus Nama (Hydrophyllaceae) includes a diverse assemblage of some 50 annual and perennial species distributed primarily in the drier regions of western United States and Mexico.As hypothesized by Hitchcock (1933), the genus is structured of five sections, Arachnoidea, Cinerascentia, Zonolacus, Conanthus, and Nama (=Eunama of Hitchcock); the former three sections are monotypic, the fourth houses but three species, and sect.Nama shelters the remainder, an assorted community itself.Indeed, Chance and Bacon (1984) have found that sect.Nama comprises five species groups based on seed coat morphology, and in seed features species of sects.Conanthus and Zonolacus singly conform with two of those groupings.Species of sects.Arachnoidea and Cinerascentia, N. lobbii A. Gray and N. rothrockii A. Gray, respectively, form a sixth seed group, but their distinctive features argue that any relationship between the two or of either to the bulk of Nama is remote (Chance and Bacon 1984;Bacon, Bragg, and Hannan 1986), an argument countenanced by their flavonoid chemistry (Bacon, Fang, and Mabry 1986) and chromosome number (Bacon 1984;Constance 1963).In fact, accumulated evidence argues convincingly that these latter two species are best viewed as distinct phyletic elements within Hydrophyllaceae with affinities nearer to Eriodictyon than to Nama (Bacon, Fang, and Mabry 1986;Bacon, Bragg, and Hannan 1986).
Traditionally, the nearest relative of Nama has been identified as Eriodictyon.Nama lobbii is critical for this alignment since it is through this taxon that Hitchcock (1933) allied the genus with Eriodictyon.Acceptance of recent evidence, however, portraying N. lobbii as well removed from the bulk of Nama, severs the direct relationship of the two genera and requires that relatives of Nama proper be sought elsewhere within the family.While Hitchcock (1933) addressed directly only the perceived relationship of Nama and Eriodictyon, he did briefly note the monotypic genera Draperia and Lemmonia as close relatives of Nama.Although Nama and Draperia are chromosomally divergent (Constance 1963), their relationship has yet to be fully explored.Lemmonia, on the other hand, is chromosomally harmonious with Nama (Constance 1963).Moreover, Gray (1877) observed that the seeds of Lemmonia "are very like those of Nama demissa (sic) and of Conanthus" (=N.aretioides (H.& A.) Brand in sect.Conanthus of Hitchcock).Seed coat morphology has proven informative in discerning relationships within Nama proper (Chance and Bacon 1984) and in assessing relationships among N. lobbii, Eriodictyon and Turricula (Bacon, Bragg, and Hannan 1986).Therefore, seeds of Lemmonia have been examined using the scanning electron microscope (SEM) with the aim of evaluating seed coat features as they pertain to the relationship of the former taxon and Nama.Also, ultrastructural features of the Lemmonia seed coat revealed the need for re-evaluation of cell wall ultrastructure in one of the six seed groups of Nama as identified by Chance and Bacon (1984).Implications of these seed features as they bear on systematics-relationships of Nama are considered herein.

MATERIALS AND METHODS
Mature, whole seeds removed from herbarium specimens (see Appendix) were mounted on aluminum stubs with double-stick carpet tape.For examination of internal features of cells forming the outermost testa layer (reticulum cells), seeds were sectioned free-hand with a single-edge razor blade, extracted for about two minutes in acetone to remove obscuring oils, and mounted as for whole seeds.Additionally, whole seeds were placed in a 1/4-dram screw-cap vial, covered with acetone and sonicated for 60-150 sec to remove the outer tangential wall and reveal radial wall features of cells composing the reticulum; sonicated seeds were mounted as for whole seeds.Specimens were coated to a thickness of approximately 50 nm in a Polaron E51 00 sputter coater using gold-palladium target and examined with a JEOL JSM 30-C SEM at an accelerating voltage of 15 kV.

RESULTS
Seeds of Lemmonia are dark brown to black and generally ovate-ovoid (Fig. 1).Seed size and shape are variable, however, as is the case in Nama and most other examined groups (see Chance and Bacon 1984).The outer testa is obviously cellular and surface patterning is shallowly reticulate as defined by Chance and Bacon (1984).Reticulum cells generally are elongate and tend to be oriented at right angles to the long axis of the seed (Fig. 1); however, some cells approach being as broad as long (Fig. 2) and orientation is obscured.Reticulum cell walls are undulate (Fig. 1, 2) and in section exhibit what appear to be pores in radial walls.Indeed, this impression can be gained in some sonicated seed preparations where sonication has removed much of the cell wall, leaving only fragments (Fig. 11).However, even in some sections the pores are clearly occluded by a thin wall (Fig. 5), suggesting that they are an artifact of the sectioning process and that walls are possessed of thickenings, instead.Examination of seeds sonicated for a shorter period reveals that radial walls do indeed include columnar thickenings (Fig. 12).Thickenings are irregularly placed and their positioning reflects the undulate nature of the wall.Seeds with outer testa cells exhibiting undulate radial walls that show pores in sections characterize one seed group in Nama (Chance and Bacon 1984).Species included in this grouping are N. demissum A. Gray (Fig. 3), N. depressum Lemmon ex A. Gray (Fig. 4), N. densum Lemmon, N. aretioides (Fig. 7), N. parviflorum (Greenm.)Const.(Fig. 8), N. dichotomum (R. & P.) Choisy, N. origanifolium H.B.K. (Fig. 9), N. pusillum Lemmon ex A. Gray (Fig. 10) and N. sericeum Willd.ex Roem.& Schult.In the light of results from Lemmonia, review of reticulum cell radial wall features in these species was mandated.Reexamination of seeds of species in this group confirms features as reported by Chance and Bacon (1984), including the presence of pores in seed sections (Fig. 6).Examination of sonicated seeds, however, shows that reticulum walls of these species include columnar thickenings (Fig. 13-18) rather than pores, and affirms that the pores as reported for this grouping by Chance and Bacon (1984) are indeed artifacts of sectioning.
Generally, wall structure is strikingly similar among examined Namas and between them and Lemmonia.Variations occur among species, nevertheless.Thickenings in N. depressum are not so pronouncedly distinct as in other species (Fig. 15) but are clearly present.Those of N. dichotomum (Fig. 18) are not so rounded nor as frequent and often they are broader than those of other taxa.
Wall structure of N. pusillum has yet to be fully confirmed.Sonication, generally removes completely the outer testa layer or leaves the collapsed outer tangential wall intact, obscuring radial walls.A few wall portions have been observed and I believe thickenings are present although weakly defined.However, additional study is needed before such can be stated with certainty.
Overall, seeds of Lemmonia are more similar, remarkably so, to those of N. demissum than to those of any other Nama.Seeds of both exhibit markedly undulate radial walls and their reticulum cells show similar variations in size and shape.Additionally, wall thickenings of the two are similar in shape, more so than with other species examined.All in all, seed features of these two taxa are strikingly alike.

DISCUSSION
Emendation of characters defining the Nama seed group proposed by Chance and Bacon (1984) does not negate the grouping as proposed; in fact, it places it on a more solid basis, both conceptually and pragmatically, because wall thickenings are more readily explained than are pores.Moreover, Chance and Bacon (1984) were less than certain that N. demissum and N. dichotomum were to be included in this seed group, despite their obviously undulate walls, since seed sections of the two rarely yield clearly interpretable wall patterns and they were unable to observe pores in the walls of either taxon.As well, Chance and Bacon (1984) were concerned as to placement of N. origanifolium and N. sericeum within this group owing largely to the woody habit of both, but also due to the presence of what appeared to be thickenings traversing the lower reticulum cell walls in N. origanifolium (such thickenings characterize yet another seed group in Nama) ; they attributed these apparent thickenings to folds in the collapsed upper tangential wall where it contacted more prominent wall undulations.The fact that only radial wall thickenings were observed in all examined Nama species removes much of the organizational uncertainty voiced by Chance and Bacon (1984) and affirms their proposed grouping.
Lemmonia californica A. Gray is a small, prostrate, dichotomously branched, white-flowered annual distributed in the southern portion of California, adjacent Nevada, western Arizona, and Baja California.The taxon also is possessed of unique floral features; its corolla is campanulate and its stamens are basally coalescent " ... in such manner as to form a sort of annulus to the corolla-tube."(Gray 1877).Also, the taxon generally produces only four seeds.Gray (1877) noted that the genus was " ... somewhat related to Draperia among the Phacelieae, yet belonging properly to the Nameae." Since its establishment by Gray (1877), Lemmonia has been consistently maintained as a monotype (Gray 1880(Gray , 1886;;Peter 1892;Brand 1913;Jepson 1925Jepson , 1943;;Hitchcock 1933;Munz and Keck 1959;Constance 1963); this treatment is surprising since, except for the two woody taxa, the species of Nama in the above seed group are all very similar to Lemmonia; they occur in the same region, all are dichotomously branched annuals and four of the seven annuals have white flowers.Moreover, the annuals are generally small and, except for N. dichotomum, prostrate in habit.Indeed, even more markedly divergent elements have been included in Nama by many of the same authorities who accepted Lemmonia as distinct, e.g., Draperia (Gray 1862) and Turricula (Gray 1875(Gray , 1880(Gray , 1886;;Peter 1892;Brand 1913), the placement of the latter problematical even in recent years (Hitchcock 1933;Jepson 1943).(However, much ofthe seeming inconsistency in this regard almost certainly is due to the positioning of N. lobbii-see Bacon, Bragg, and Hannan [1986] and Bacon, Fang, and Mabry [1986] for a full discussion of this situation.)Gray's treatment of Lemmonia as a distinct genus seems consistent, nevertheless, since, owing to its connate styles, Gray continually held N.
aretioides to be generically distinct from Nama.Additionally, Gray (1875) considered ovule number as fundamental to certain groupings in his reorganization of Phacelia; to what extent this stance influenced his perception of the status of Lemmonia is unclear, but it seems reasonable to assume that, along with floral characters, the fewer, larger seeds of the latter, as compared to those of Nama, would have enhanced his evaluation.(Ovule-seed number may also account for Gray's suggestion that Lemmonia is "somewhat related" to Draperia, which also produces but four seeds; in all other aspects the two are distinct.)However, Constance (1949Constance ( , 1963) ) has shown that ovule number is generally an arbitrary character and not indicative of phyletic origin in Phacelia.Recent evidence from seed morphology regarding sect.Conanthus suggests that, in some instances at least, floral features may be misleading when used to adjudge phyletic status.In fact, Jepson has implied as much regarding floral features of Lemmonia, while accepting the genus, by noting (Jepson 1943, p. 284): "The tube of the corolla of Lemmonia is very short but distinctive and rather conspicuous.It is constricted at the top, at the point of stamen attachment.The bowl-shaped throat, markedly expanding above the tube, is twice as long as the tube.The ovary and capsule, described by Gray (Proc.Am .Acad. 12: 162; Bot.Cal.2:468) as 2-celled, are really I-celled with closely parietal narrow placentae.In no way does Lemmonia differ in fruit from the species of Nama as a whole.Indeed, Lemmonia californica resembles Nama pusillum so closely that for the generic validity of Lemmonia.one must look to the corolla character newly described above and to the peculiar structure joining the bases of the filaments.In all other essential respects, in habit, aspect, foliage, inflorescence and capsules, the similarity of Lemmania calif arnica to Nama pusil!um is remarkable.Lemmania as a genus rests mainly on two slight technical characters-that of the corollatube and the dilation-joined filaments." In addition to morphological similarities, chromosome number of both Lemmonia and Nama is n = 7 (see Constance 1963;Bacon 1974Bacon , 1984;;).Hydrophyllaceae have been thoroughly surveyed chromosomally (Cave and Constance 1942, 1944, 1947, 1950, 1959;Constance 1963), and n = 7 is found elsewhere in the family only in one species of Phacelia, P. racemosa (Kell.)Brandeg., and one of Nemophila, N. phacelioides Nutt.ex Barton (see Constance 1963); the number is clearly derived in each of these taxa and its origin in each certainly is phyletically unrelated to that in either Nama or Lemmonia.Indeed, N. phacelioides has the largest chromosomes yet encountered in Hydrophyllaceae (Con- stance 1963) while those of Nama and Lemmonia are small.Conformity of seed features of Lemmonia to those of the aforementioned Namas is apparent.It seems unlikely, given the degree of agreement in other features, that these similarities are simply fortuitous.Indeed, seed features among the few hydrophyllaceous taxa examined in depth, Nama, Turricula and Eriodictyon (Chance and Bacon 1984;Bacon, Bragg, and Hannan 1986), accentuate the unique organization exhibited by these species.Moreover, preliminary observations have been completed on seeds of Phacelia (25 spp.), Emmenanthe (I sp.), Ellisia (I sp.) and Tricardia (1 sp.), and testa organization in all is distinct as contrasted with that in the taxa discussed herein.Seed features confirm Gray's observation of the similarity of Lemmonia seeds and those of N. demissum and N. aretioides; indeed, seed similarities of the former and N. demissum would portray the two as more closely allied than are Lemmonia and N. pusillum, as might be inferred from the statement of Jepson (1943).However, the uniformity in testa organization among all species examined herein poses somewhat ofa paradox; it strongly supports the alliance of Lemmonia with Nama but also places the generic status of Lemmonia in doubt.
My intent is not to suggest that seed coat features are the ultimate arbiters of taxonomic status in Hydrophyllaceae.However, in those few genera studied in depth, all in Gray's Nameae, seed coat indications correspond remarkably with other features of morphology and cytology.(Seed coat features of examined species of Phacelia correlate strongly with subgenera and species groups proposed by Constance [1963].)Therefore, results of those studies might be instructive in evaluating the taxonomic significance of seed features of Lemmonia.Indeed, a similar situation exists in Turricula and Eriodictyon (Bacon, Bragg, and Hannan 1986).Seeds of Turricula are, essentially, identical to those found in Eriodictyon species; on the basis of seed features alone, Turricula could easily be accommodated in Eriodictyon.However, the former is a malodorous, herbaceous perennial while the latter is an aromatic, shrubby genus.Additionally, Turricula has a chromosome complement of n = 13, while Eriodictyon is uniformly n = 14 (see Constance 1963).Seed features would affirm that the two are to be allied while morphological and cytological features support their accepted status.In the case of Lemmonia and Nama there is no chromosomal or morphological basis supporting of generic maintenance of the former.I tend to the opinion that Lemmonia cannot be maintained as distinct from Nama.Nevertheless, I have no field experience with Lemmonia and hesitate to alter its status until seeds of other potential relatives of it and Nama, including Draperia, have been examined.But whatever its status, seed features argue cogently that relationships of the taxon lie with the mentioned seed group in Nama.