Cryptantha pterocarya (Torrey) Greene var. pterocarya forma pseudocycloptera M.E. Mabry & M.G. Simpson, 2016

Mabry, Makenzie E., Dowdy, Regina A., Simpson, Lee M., Rebman, Jon P. & Simpson, Michael G., 2016, Taxonomy of the winged popcorn flower: Cryptantha pterocarya (Boraginaceae), Phytotaxa 253 (2), pp. 97-130 : 114-116

Cryptantha pterocarya (Torrey) Greene var. pterocarya forma pseudocycloptera M.E. Mabry & M.G. Simpson , forma nov.

Type:— USA. California, Imperial County , In-Ko-Pah Mountains, Jacumba Natural Area, along road ascending to Valley of the Moon, granite substrate, 1000 m, 32.63944˚ N, -116.10111˚ W, 13 April 2001, Rebman 7221, with D. Silverman, B. Lauri, Oscar Solis (holotype SD 155821!, isotypes RSA 700474!, UC 1787754!, UCR 155804!).

Diagnosis: — Cryptantha pterocarya var. pterocarya forma pseudocycloptera is similar to the typical form of Cryptantha pterocarya var. pterocarya ( forma pterocarya ) in having prominently winged (wing generally ≥ 0.6 mm on one side) nutlets, the wings typically lobed and not extending to the base of the nutlet body. It differs from the typical form in having homomorphic nutlets, with all four nutlets winged.

Etymology: —The epithet “ pseudocycloptera ” means “false cycloptera,” in reference to its resemblance and past confusion with Cryptantha pterocarya var. cycloptera [= C. cycloptera ].

Paratypes: —See Appendix 1.

Interestingly, in the typical form of C. pterocarya var. pterocarya , the odd nutlet of heteromorphic specimens shows a resemblance to C. utahensis in shape (both lance-ovate), sculpturing (both papillate-tuberculate with often spinulose tubercles especially toward the nutlet apex), and in having a sharp “knife-like” margin ( Fig. 6E ). Cryptantha utahensis was included in series Pterocaryae by Johnston (1925) along with C. oxygona and C. pterocarya (including vars. cycloptera , pterocarya , and later stenoloba, the last described by Johnston 1939). However, Cryptantha utahensis typically has only one nutlet per fruit (rarely two, and then homomorphic) and never exhibits any appreciable marginal wing. It is because of these clear differences in morphology that we did not include C. utahensis in our comparative analyses.

Cryptantha p. var. purpusii warrants continued recognition at the varietal level. Variety purpusii is discrete among the other varieties or forms of C. pterocarya in having a significantly narrower nutlet wing ( Fig. 5 ; 6 ; 9A ). Variety purpusii also generally has a smaller nutlet body length, more closely approaching that of C. oxygona and the “truncata” form ( Fig. 8E ), but lacking any significant difference. Variety purpusii has a relatively narrow range, found in the Argus Mountains and along the escarpment of the Sierra Nevada and Peninsular Ranges of California and Baja California , plus a few populations in mountains of western Nevada ( Fig. 12A ). This geographic continuity provides additional evidence for its continued recognition as a taxon. However, variety purpusii represents a greater challenge to characterize, as it is quite variable in nutlet morphology. Some samples approach var. pterocarya in being heteromorphic with consimilar nutlets having a relatively broad wing width (e. g., Fig. 6A ). Other samples are homomorphic and have a narrow nutlet wing (e.g., Fig. 6C, D ), in some more of a marginal ridge, these approaching C. oxygona ( Fig. 6F ). However, Cryptantha oxygona is different from all C. pterocarya varieties and forms in its significantly larger corolla diameter ( Fig. 8D ), although with some overlap; this feature alone generally clearly distinguishes C. oxygona from the varieties and forms of C. pterocarya , warranting its continued recognition as a separate species. In identification of C. p. var. purpusii we used the previous designation of a wing width less than 0.5 mm ( Kelley et al. 2012). One specimen measured 0.55 mm in wing width, but had a small nutlet body length and therefore was classified as C. p. var. purpusii . We suggest, from results of this analysis, that C. p. var. pterocarya and var. stenoloba (see below) be distinguished as having a wing width 0.6 mm and larger and C. p. var. purpusii as having a wing width less than 0.6 mm, a smaller nutlet body, and located in a more restricted range.

Interestingly, the distribution of C. oxygona (not illustrated) parallels that of C. p. var. purpusii in occurring mostly along the escarpment of the upper peninsular range (as far south as northern San Diego County) and along the escarpment of the southern Sierra Nevada (as far north as Yosemite), but is more widespread than var. purpusii in the Tehachapi Range, the central-northern transverse range, and along the (mostly) eastern slopes of the southern coastal ranges of California . It appears to be restricted to California , with some populations near the border with western Nevada.

Cryptantha pterocarya var. stenoloba also warrants continued recognition at the varietal level. It is discrete from other varieties and forms in having a significantly longer calyx ( Fig. 7A–C ; 8A ), a significantly longer sepal length: width ratio ( Fig. 8C ), a significantly longer consimilar nutlet body ( Fig. 8E ), and longer consimilar nutlet body length: width ratio ( Fig. 8F ; see Fig. 7F ), these last two features not recognized previously. This taxon has the narrowest range of any variety or form, occurring near the Colorado River basin in southeastern California , western Arizona, and southern Nevada, and near the Virgin River basin in southeastern Nevada and northwestern Arizona ( Fig. 12A ). Variety stenoloba most likely represents a diverging taxon of var. pterocarya as it shares many morphological similarities in nutlet morphology, the nutlets heteromorphic, the consimilar nutlets with a broad wing ( Fig. 9A ), and an essentially absent wing at the base ( Fig. 9B, D ; see Fig. 7E, F ). Given its discreteness, it is possible that C. p. var. stenoloba could be recognized at the species level in the future. But, because of the overlap it shows in some features, we elect to treat it at the rank of variety at this time. It should definitely be included in local floristic treatments, where it is often missing (e.g., Kelley & Wilken 1993; Kelley et al. 2012). We cite four specimens ( RSA 774935, RSA 782075, SDSU 20929, UCR 214870, all from San Bernardino County), which constitute new state records for California .

We noticed no significant differences in consimilar nutlet wing notch depth, except for C. p. var. stenoloba , which has a significantly deeper, but highly overlapping, notch depth relative to all other taxa or forms ( Fig. 9E ). We have no basis for believing that C. p. var. stenoloba represents the elusive C. p. var. pectinata . Although we were unable to physically examine the presumed (although confused; see earlier quote by Johnston 1925) type specimen of C. p. var. pectinata , our measurements of the digital scan of this specimen (Parry 168-169, NY 01111827) shows it to have relatively wide and short sepals, the length measuring 3.7–4.7 mm, well below that of variety stenoloba . Still, it is intriguing that this type specimen of var. pectinata , which might show some resemblance in wing notch depth to var. stenoloba , was collected near St. George, Utah in the Virgin River Valley, very near known populations of var. stenoloba (see Fig. 12A ). However, based on variation of this feature observed in this study, we reject C. p. var. pectinata as a valid taxon, as other taxonomists have in recent treatments. Lastly with regard to this feature, Cryptantha oxygona and C. pterocarya var. purpusii have significantly smaller notch depths relative to other taxa or forms, but not to one another ( Fig. 9E ). This is not unexpected given the relatively smaller wing width of these two taxa ( Fig. 5 , 6A–D, F ).

Cryptantha pterocarya var. cycloptera and the “truncata ” form are similar to one another in together having a significantly longer consimilar nutlet wing base ( Fig. 9B ), a longer odd nutlet wing base ( Fig. 9D ), and a longer gynobase stalk length ( Fig. 9F ), although note that in the last two features, var. cycloptera and the “truncata ” form are statistically different from one another, with the “truncata ” form intermediate to var. cycloptera and the other taxa/ forms. We emphasize the clarification that only C. p. var. cycloptera and the “truncata ” form have any appreciable nutlet basal wing and any appreciable gynobase stalk. In addition, we note that from our qualitative observations only C. p. var. cycloptera and the “truncata ” form have an upper stem vestiture composed of only appressed trichomes; the other forms have appressed and spreading trichomes. From these data, we propose that C. pterocarya var. cycloptera be re-elevated to species level (as C. cycloptera (Greene) Greene ), using a taxonomic species concept ( Cronquist 1978, 1988). Interestingly, Cronquist (1984) also implied the distinctiveness of var. cycloptera by suggesting that varieties pterocarya , purpusii, and stenoloba be classified together within a subspecies (not described), separate from var. cycloptera ( Cronquist 1984, p. 257) . The geographic range of Cryptantha cycloptera overlaps considerably with C. pterocarya , but is more restricted, occurring only in the southern half of California , southern Nevada, most of Arizona, southern Utah, western New Mexico and Texas, and Baja California and Sonora, Mexico ( Fig. 12B ).

We believe we have elucidated the original basis for the transfer in rank of C. cycloptera to a variety of C. pterocarya by Macbride (1916). In that treatment, Macbride stated (with regard to this taxon):

“However, a study of the ample material in the Gray Herbarium seems to prove conclusively that it is, at best, only a geographical variety of C. pterocarya . In the first place, the ventral face of the nutlets may or may not be smooth in either of the proposed species. Secondly, all of the nutlets may be winged and yet the wings not extend across the base, as for example in the plants collected at Grand Junction, Colorado, by Alice Eastwood. When one considers the fact that the ventral faces of these winged nutlet are rough, one is puzzled as to whether the plants are more nearly related to C. pterocarya or to C. cycloptera . However, it must be noted that it is only the southwestern material that can be referred to C. cycloptera . It seems advisable, therefore, to consider C. cycloptera as a variety of C. pterocarya and to include in this variety all specimens that have four winged nutlets, irrespective of whether the wing extends across the base. Although the species ranges from Washington to Utah and southern California , the variety apparently largely replaces it, in the interior of the Southwest.”

In our opinion, Macbride overemphasized nutlet heteromorphism over the presence of a basal wing (i.e., the wing completely encircling the nutlet body) in evaluating these taxa. Also, he was not aware of the gynobase stipe that is obviously correlated with a basally positioned wing. Nor did he consider stem pubescence in his evaluation. The Alice Eastwood specimen he mentions we believe to be Eastwood 5126 (CAS 26648, collected near Grand Junction, Colorado; see Fig. 11 ), a specimen that we verified in this study to be a “pseudocycloptera” form, the homomorphic form of C. pterocarya var. pterocarya . As mentioned earlier, this has been an ongoing source of confusion in identification (even originally by the authors of this article), in that all homomorphic forms of C. pterocarya var. pterocarya were often identified as C. cycloptera , as that is the feature singled out in many keys. We do, however, agree with Macbride that the ventral surface of the nutlet of C. pterocarya is quite variable in sculpturing. Although we did not quantify this feature, we believe that ventral surface sculpturing is not consistent with respect to any taxon or form in the complex.

The “truncata” form is similar to C. cycloptera in having a consimilar nutlet wing base significantly longer than all other taxa and forms ( Fig. 9B ; note that in the homomorphic C. cycloptera all four similar nutlets were averaged for this feature). The stem pubescence of the “truncata” form is also like that of C. cycloptera , having solely appressed trichomes. The “truncata” form is intermediate between C. cycloptera and all other varieties and forms with respect to odd nutlet wing base ( Fig. 9D ; again, in the homomorphic C. cycloptera all four, similar nutlets were averaged for this feature) and gynobase stalk length ( Fig. 9F ). From these analyses, it is clear that the “truncata” form is much more similar to C. cycloptera than to any other varieties or forms. We have considered the option of treating the “truncata” form as a variety of C. cycloptera , but are hesitant to do so for three reasons. First, the geographic range of the “truncata” form is more restricted than, but overlapping and nested within that, of C. cycloptera , with no clear discontinuity ( Fig. 12B ). Second, the “truncata” form of the nutlet shows a fair amount of variation in the odd nutlet basal wing width and gynobase stalk length, although the quantitative analyses do show significant differences ( Fig. 9D, F ). Third, we have found specimens of a typical cycloptera form in which one or more fruits of the “truncata ” form were found on the same plant (Damrel 1623-B, ASU 237494) and we have found a specimen of a typical “truncata” form in which one or more fruits of the typical cycloptera form were found on the same plant (Pinkava 10939, ASU 166212); see Appendix 3. The fact that nutlets representing the two forms can be found on the same individual, suggests that the two forms of C. cycloptera may not be fully discrete. However, this is uncommon; almost all specimens of these forms are invariant within an individual with respect to nutlet features. Therefore, we believe at this stage that this form warrants formal recognition as a form of C. cycloptera .

Figures

FIGURE 5. Cryptantha pterocarya var.purpusii, Purpus 5433(Holotype).A. Fruit with calyx and four nutlets.B. Nutlets of fruit removed; note heteromorphism.C. One of three consimilar, winged nutlets in dorsal (left), ventral (middle), and side (right) views.D. Odd, unwinged nutlet in dorsal (left), ventral (middle), and side (right) views. Scale bars = 1 mm.

FIGURE 6. A–D. Cryptantha pterocarya var. purpusii, showing variation in nutlet morphology.A. Specimen with heteromorphic nutlets, odd nutlet unwinged (left) and consimilar nutlets with relatively broad wing (right) (RSA 730967). B. Specimen with heteromorphic nutlets, odd nutlet marginally ridged (left), consimilar nutlets (right) more narrowly winged, winged strongly lobed (SD 91861). C–D. Specimens with homomorphic nutlets, nutlet margin ridged. (C, SDSU 18624; D, RSA 728905). E. Nutlet of Cryptantha utahensis (SDSU 17286), shown for comparison with odd, unwinged nutlet of C. pterocarya; dorsal view (left), ventral view (right). F. Nutlet of Cryptantha oxygona (RSA 717219), shown for comparison with certain forms of C. pterocarya var. purpusii; dorsal view (left), ventral view (right). Scale bars = 1 mm.

FIGURE 7. A–C. Cryptantha pterocarya var. stenoloba, showing elongate calyx and nutlets. A, C. Maguire & Blood 4466 (Holotype), images from scanned herbarium sheet. B. Fruits of additional specimen of C. p. var. stenoloba (RSA 782075). D. Cryptantha p. var. pterocarya (NY 00335248, Holotype) for comparison.Note shorter nutlets and considerably shorter calyx.E,F.Cryptantha p. var.stenoloba (GH barcode 00096072, Holotype). E. Four nutlets of a single fruit, showing heteromorphism, with three consimilar nutlets winged only on sides and one odd, unwinged nutlet. F. One of consimilar nutlets, dorsal view (left) and single odd nutlet in dorsal (middle) and ventral (right) views. G–I. Comparison of nutlet form with other varieties, all to the same scale as “F.” G. Cryptantha p. var. pterocarya (NY 00335248, Holotype). H. Cryptantha pterocarya var. cycloptera [C. cycloptera] (CAS 202, barcode 0006884, Holotype). I. Cryptantha p. var. purpusii, (JEPS 2548, Holotype). Scale bars = 1 mm.

FIGURE 8. Box plots of characters analyzed for Cryptantha pterocarya var. cycloptera, C.oxygona, C. p. “pseudocycloptera” form, C. p. var. pterocarya, C. p. var. purpusii, C. p. var. stenoloba, and C. p. “truncata” form.A. Calyx body length in fruit (mm). B. Sepal width (mm). C. Sepal length:width ratio. D. Corolla limb diameter (mm). E. Consimilar nutlet body length (mm). F. Consimilar nutlet length: width ratio. Taxa that are significantly different from all other taxa and forms are indicated with double asterisks (**) (p<0.01). Pairs of taxa that are significantly different from all other taxa or forms except from one another are indicated by two plus symbols (++) (p<0.01) or a single plus symbol (+ or +’) (p<0.05). The “†” symbol indicated statistical difference between all other taxa and forms of C. pterocarya but not with C. oxygona. Outliers are indicated by a single small asterisk and extreme outliers, greater than 1.5 times the data are indicated by a small open circle. Stastistical differences determined via ANOVA Tukey post hoc test.

FIGURE 9. Box plots of additional characters analyzed for the same taxa as in Figure 9. A. Consimilar nutlet maximum wing width (mm). B. Consimilar nutlet wing base (mm). C. Odd nutlet lateral wing width (mm). D. Odd nutlet wing base (mm). E. Consimilar nutlet wing (maximum) notch depth (mm). F. Gynobase stalk length (mm). Symbols and analyses as in Figure 9.

FIGURE 11. Distribution map of: 1) Cryptantha pterocarya var. pterocarya specimens verified to identity and measured (black squares), verified to identify but not measured (black dots), and not verified but included for geographic range mapping (black triangles); and of 2) the “pseudocycloptera” form verified to identity and measured (red squares) and verified to identity but not measured (red circles). Map data from ©Google 2015, INEGI Data.

FIGURE 12. A. Distribution map from specimens of: 1) Cryptantha pterocarya var. purpusii verified to identity and measured (black squares) and verified to identity but not measured (black dots); and 2) C. pertocarya var. stenoloba verified to identity and measured (red squares) and verified to identity but not measured (red circles). B. Distribution map from specimens of: 1) Cryptantha pterocarya var. cycloptera [=C. cycloptera] verified to identity and measured (black squares), verified to identify but not measured (black dots), and not verified but included for geographic range mapping (black triangles); and 2) the “truncata” form verified to identity and measured (red squares) and verified to identity but not measured (red circles). Map data from ©Google 2015, INEGI Data.