34 LUNDELLIA DECEMBER, 2006

SPECIES OF PHILADELPHUS (HYDRANGEACEAE) FROM TRANS-PECOS TEXAS

Billie Lee Turner Plant Resources Center, The University of Texas Austin, Texas, 78712, USA Abstract: Three species of Philadelphus are accepted as occurring in Trans-Pecos, Texas: P. mearnsii, P. microphyllus, and P. serpyllifolius. Philadelphus mearnsii is known only from Culberson and El Paso counties; P. microphyllus and P. serpyllifolius are more widespread. Philadelphus microphyllus is treated as having three intergrad­ ing varieties: Philadelphus microphyllus var. argenteus, P. m. var. crinitus (C. L. Hitchc.) B. L. Turner, and P. m. var. microphyllus. Philadelphus serpyllifolius is treated as having two intergrading varieties: Philadelphus serpyllifolius var. intermedius B. L. Turner, and P. s. var. serpyllifolius. Maps showing their distributions in the area concerned are provided.

Keywords: Philadelphus, Hydrangeaceae, Trans-Pecos, Texas.

The taxonomic history of the Trans-Pe­ tative treatment compiled from the work of cos species of Philadelphus is checkered, to yet others. Were I to re-map the treatment say the least. Rydberg (1905) provided the based upon my own studies, I would rec­ first comprehensive treatment of the genus ognize but three species in the Trans-Pecos for the region positioning it in the family region as follows: Hydrangeaceae; this recently confirmed by DNA data (Soltis and Soltis, 1997). Yet, in spite of his splintering proclivities, he rec­ KEY TO THE TRANS-PECOS SPECIES ognized only a single species for Trans-Pe­ OF PHILADELPHUS cos, Texas, P. serpyllifolius. 1. Lower surfaces of!eaves without a vestiture of mi­ Correll and Johnston (1970) and John­ nute frizzy or pannose hairs. ston (1988), accepting Hu's (1954-1956) 2. Stamens 15-20; petals acute to broadly obtuse highly abiological treatment of the group, at the apex ...... 1. P. mearnsii recognized nine species for Trans-Pecos, 2. Stamens 25-60; petals rounded at the apex ...... 2. P. microphyllus Texas: Philadelphus argenteus Rydb., P. ar­ 1. Lower surfaces of leaves with appressed straight gyrocalyx Wooton, P. crinitus (C. L. stiff hairs, these underlain by a vestiture of minute Hitchc.) S. Y. Hu, P. hitchcockianus S.Y. Hu, frizzy or pannose hairs ...... 3. P. serpyllifolius P. mearnsii W. H. Evans ex Small & Rydb., P. microphyllus A. Gray, P. occidentalis A. 1. PHILADELPHUS MEARNSII W.H. Evans Nelson, P. palmeri Rydb., and P. serpyllifol­ ex Small & Rydb., N. Amer. fl. 22: 174. ius A. Gray. Hu's treatment was also ac­ 1905. cepted by Powell (1998), but with the ad­ monition that "The Trans-Pecos species of Philadelphus hitchcockianus S.Y. Hu, J. Ar­ Philadelphus require further study in sup­ nold Arbor. 37: 51. 1956. Fig. 1 port of the work by Hu (1954-56)." For the Atlas of the Vascular Plants of This species is known to Texas only Texas (Turner et al., 2003) I had to review from the Franklin Mts. (El Paso Co.: Wor­ anew the status of Philadelphus in Texas. thington 24987, SRSC, UTEP) and Guada­ Unfortunately, the present study was in­ lupe Mts. (Culberson Co.: numerous col­ complete at the time of the Atlas' prepara­ lections, SRSC, TEX). The single collection tion and I had to incorporate into it a ten- from the Franklin Mts. is sterile, but ap-

LUNDELLIA 9:34-40. 2006 NUMBER 9 TURNER: PHILADELPHUS OF TRANS-PECOS, TEXAS 35

pears to have the vegetative characters of 2. PHILADELPHUS MICROPHYLLUS A. Gray, Philadelphus mearnsii, and is said to occur PL wright. 1: 54. 1852. Fig. 2 at an elevation of 5900 ft. The only other Hitchcock (1943) recognized eight sub­ collection of Philadelphus known to me species under the fabric of P. microphyllus, from the Franklin Mts. is Worthington three of these are accepted as occurring in 25113 (SRSC), collected at the "base of cliffs the Trans-Pecos; albeit recognized as vari­ on the NE side and immediately below An­ eties, as follows: thony's Nose," 6000-6300 ft. I take this specimen, because of its densely tomentose 1. Hypanthium densely white tomentose; hypanthium, to be P. microphyllus var. ar­ Franklin Mts .. 2a. P. m. var. argenteus genteus. In the Trans-Pecos at least, P. 1. Hypanthium sparsely appressed-ciliate mearnsii apparently occurs at somewhat to glabrous. lower elevations than P. microphyllus 2. Undersurfaces of the leaves densely (5900-6500 ft. vs. 6400-6800 ft., as deter­ pubescent with appressed, often mat­ mined from label data) with which it is ter hairs, the dermal layer scarcely sympatric, at least in the Guadalupes of visible; known only from the upper­ Culberson Co., Texas. No intergradation most slopes of Mount Livermore between the two species has been noted (7200-8000 ft), Jeff Davis Co from this locality among the 30 or more ...... 2b. P. m. var. crinitus specimens examined by me (LL, SRSC, 2. Undersurfaces of leaves sparsely to TEX). Indeed, the late Barton Warnock moderately pubescent with appressed (22805, 22806, SRSC) collected typical ele­ non-matted hairs, the dermal layer ments of both species in "South McKittrick clearly visible: widespread at mid el­ evations of Chisos, Davis, Del Norte, Canyon" on 28 May 1968. In short, the two and Guadalupe Mts taxa appear to be non-interbreeding species ...... 2c. P. m. var. microphyllus in nature. Philadelphus hitchcockianus is typified (in part) by a collection from Culberson 2a. PHILADELPHUS MICROPHYLLUS var. Co., Texas (Guadalupe Mts., above Mc­ ARGENTEUS (Rydb.) Kearney & Peebles, J. Washington Acad. Sci. 29: 480. 1939. Kittrick Canyon, Moore & Steyermark 3477 Philadelphus microphyllus subsp. argen­ [LECTOTYPE: A; ISOLECTOTYPES: GH, teus (Rydb.) C. L. Hitchc., Madrofio 7: 42. MO]) that Hitchcock considered to be typ­ 1943. ical P. mearnsii, as do I. Hu complicated the According to Frazier (pers. comm., typification of P. hitchcockianus by citing as email, 1995), "This is the common Phila­ "types" two elements (a fruiting element, delphus of southern New Mexico and ad­ Moore & Steyermark 3477) and a flowering jacent southwest Arizona and north-central element (New Mexico. Eddy Co.: Carlsbad Mexico." It is known to me by only a single Cave, 5 May 1924, Bailey s. n. US). Such collection from the Franklin Mts. of El Paso typification is contrary to the present Inter­ Co., as follows: "at base of cliffs on the NE national Code of Botanical Nomenclature side and immediately below Anthony's and I propose here to lectotypify the name Nose, 6000-6300 ft .... limestone cliffs and P. hitchcockianus with the aforementioned talus, shaded much of the day and some Moore and Steyermark collection 3477. Hu areas completely." 19 Jul 1995, Worthington also cited one additional collection in her 25113 (SRSC, UTEP). concept of P. hitchcockianus, this being a collection from Coahuila, Mexico (Steward 2b. PHILADELPHUS MICROPHYLLUS var. 1728), which is probably also P. mearnsii, crinitus (C. L. Hitchc.) B. L. Turner, stat. sensu Hitchcock and the present worker. nov. ( 36 LUNDELLIA DECEMBER, 2006

Based upon P. microphyllus subsp. crin­ Texas, as cited by Hitchcock (1943). I take itus C. L. Hitchc., Madroiio 7: 45. 1943. the two Arizona specimens (the Hinckley Philadelphus crinitus (C. L. Hitchc.) S. collections excluded) to be atypical forms. Y. Hu, J. Arnold Arb. 37: 60. 1956. of P. microphyllus var. microphyllus. Whether Philadelphus crinitus is deserv­ This variety is known in its typical form ing of formal nomenclatural recognition is only from specimens obtained from the up­ moot. It is superficially similar to P. serpyl­ per slopes of Mt. Livermore. Hitchcock lifolius in that the dense, often tangled ves­ (1943) proposed the taxon with the follow­ titure of the undersurfaces of its leaves sug­ ing comment: "The status of this entity is gest the latter, as noted by Hitchcock; he puzzling." He further noted that it was like­ also noted that "Since the flowers are so ly closely related to Philadelphus argyrocalyx alike those of the other subspecies of P. mi­ and/or P. argenteus, "but the pubescence of crophyllus [namely P. m. subsp. argenteus the calyx and leaves is so unique that it can­ and P. m. subsp. argyrocalyx] there is not not be combined with either." an adequate basis for according the plant I accept the taxon concerned as a weak­ specific recognition." I have not maintained ly delimited high elevational varietal ele­ its specific or subspecific status because of ment of Philadelphus microphyllus, this the variability of its leaf vestiture, as noted grading at lower elevations into the typical above. P. m. var. microphyllus, as attested to by the following intermediates: Jeff Davis Co.: 2c. PHILADELPHUS MICROPHYLLUS var. Northeastern Davis Mts., 11 mi W of high­ MICROPHYLLUS way 17 on Farm Road 1832, Nations Ranch, hillside W of Bear Cave Mt., 26 Aug 1985, This is the most widespread, commonly Larke & Poole 275 (SRSC); rocky ledge on encountered variety of Philadelphus micro­ NW slope of Sawtooth Mt., 29 Jun 1948, phyllus occurring in the Trans-Pecos (Fig. Turner s.n. (SRSC). These two localities are 2). As noted in the above, Hitchcock (1943) well separated from Mt. Livermore (Fig. 2) recognized nine infraspecific categories and are at lower elevations than found for within P. microphyllus, the oldest name typical P. m. var. crinitus; still the leaf ves­ within the complex concerned, the type col­ titure of both specimens appears to ap­ lected by Fendler near Santa Fe, New Mex­ proach that of the latter. Typical forms of ico in 1847. Hitchcock also provided a rel­ P. m. var. crinitus also occur on Sawtooth atively crude map showing the distribution Mt. (W side of Sawtooth, 4 Oct 1926, Palm­ of these taxa in the southwestern U.S.A. and er 31950, TEX). Mexico. With the exception of several du­ Hitchcock (1943) based his original de­ bious mostly Mexican taxa, it is likely that scription of Philadelphus microphyllus some of these are worthy of varietal rec­ subsp. crinitus upon only three collections. ognition in that they appear to have some Since that time numerous collections of the degree of morpho-geographical differentia­ taxon have been assembled (LL, SRSC, tion, and from his account and that of Hu, TEX), all from the upper slopes of Mount tend to intergrade with contiguous infra­ Livermore. Hu (1954-56) elevated the tax­ specific elements. Holmgren and Holmgren on to specific rank but confounded its de­ (1997), however, recognized P. microphyllus scriptive parameters and geographic bound­ "as one variable complex ranging from the aries with the addition of three problematic southern Rocky Mountains to northern collections from the Santa Catalina Mts. of Mexico and west to California," noting that southeastern Arizona, one of these (Hinck­ "The differences [among the infraspecific ley 50) is a citation error. The Hinckley categories] are not at all constant and bear specimen is actually from Mt. Livermore, no correlation with geographical distribu- NUMBER 9 TURNER: PHILADELPHUS OF TRANS-PECOS, TEXAS 37

r-·-· ----.---1 ! LOVING• WINKLER I I . . I I , - . ~iiiow~;.;"~~iCRANE! I

0 so 100 - - - - I FIG. 1. Distribution of Philadelphus mearnsii in Texas. tion." I think this observation an overstate­ Blue, Glass Mts." 2 Jun 1941, B. H. War­ ment. Probably, biological reality lies some­ nock W559 (HOLOTYPE: SRSC; ISOTYPES: where between the latter study and that of SRSC, TEX). Hitchcock. Most of the observations of Hu, The holotype bears an additional no­ however, can be ignored as not having tation by its collector, "Maple odor when much biological or populational validity. dry."

3. PHILADELPHUS SERPYLLIFOLIUS A. Gray, Similis P. s. var. serpyllifoliae A. Gray, sed superficies Pl. wright. 1: 77. 1852. Fig. 3. foliorum pubescentiam opressam habent, et plantae in solis calcareis (non igneis) adsunt. Two varieties of this species are recog­ nized for the Trans-Pecos, as follows: Low divaricately branched SHRUBS to 1 m high. LEAVES ovate, 3-nervate, bicolored, 1. Upper surfaces of leaves with strictly appressed 10- 15 mm long, 5- 8 mm wide, the upper hairs; limestone soils of eastern Brewster Co. and eastward ...... 3a. P. s. var. intermedius surfaces with appressed pubescence. CALYX 1. Upper surfaces of leaves with erect hairs, the latter ca. 7 mm long; lobes ca. 3 mm long. PET­ few or numerous; igneous soils, Davis Mts. and ALS broadly oval, 7-9 mm long, 6-8 mm westward ...... 3b. P. s. var. serpyllifolius wide. STAMENS 18- 20.

3a. PHILADELPHUS SERPYLLIFOLIUS var. in­ ADDITIONAL SPECIMENS EXAMINED: termedius B. L. Turner, var. nov. TEXAS. Brewster Co.: Glass Mts., lime­ stone rocks of Gilliland Canyon, Old Blue TYPE: TEXAS. Brewster Co.: "Infre­ Ranch, 15 Jun 1987, Maddux 124 (SRSC); quent on lower limestone slopes of Old Old Blue Mt., 25 Jun 1945, Muller 8149 38 LU N DELLIA DEC EMBER, 2006

L PAso"·J::::-'H'uosp--0----7----..,. ----,...--• : """•rs '"'h . ETH;·': ,... . .-:ULBERSON; • LOVING j WINKLER ! . I I .

-·.:...·-·-·-·'·-·-·WARD""••~•ICRANI

. I "o'·. ··'t; , <> • •. /""' 1 -' .. i_,; i ... ( .«,... (•. ..-~ !,..;

0- --,. -- FIG. 2. Distribution of Philadelphus microphyllus in Texas: var. argenteus (large dot); var. cri­ nitus (circles); var. microphyllus (small dots).

(TEX); W of Old Blue, limestone canyons Hu (1954-56) placed the type (and sev­ and bluffs, 20 Jun 1978, Powell 3289 (TEX); eral additional collections) of this variety limestone hills, S of Old Blue Mt., "pine­ within her concept of Philadelphus serpylli­ juniper assoc." 20 Aug 1979, Powell 3480 folius. She further confounded circumscrip­ (SRSC); " Limestone hills on John Pate tion of the latter by including in this col­ Ranch, ca. 12 mi SE of Alpine, near Bird lections from central Texas (Blanco and Mt.; Pinus remota dominant." 19 Jul 1986, Kendall counties) that clearly belong to P. Powell et al. 5245 (SRSC); "Infrequent texensis S. Y. Hu. In my initial cogitation shrub in Mc Rae Canyon on C. F. Cox on the biological status of P. s. var. inter­ Ranch in limestone soil." 31Jul1953, War­ medius, I considered placing the present nock 11414 (SRSC); slopes of Old Blue, taxon under the fabric of P. texensis, the rocky limestone slopes, 18 Jun 1941, Rose­ latter reduced to varietal status under P. ser­ Innes & Moon 1231 (TEX). Pecos Co.: lime­ pyllifolius. Indeed, Hu herself noted that stone slopes of Sierra Madera, 22 mi S of both P. serpyllifolius and P. texensis "are Ft. Stockton, 26 May 1949, Warnock & similar in appearance and their leaves are McVaugh 8699 (SRSC). strigose and lanate beneath." The latter tax­ on (and P. s. var. intermedius) is readily dis­ The variety is named for its interme­ tinguished from typical P. s. var. serpyllifol­ diate stance between Philadelphus serpylli­ ius by the vestiture on its upper leaf surfaces folius var. serpyllifolius and P. s. var. tex­ (consisting of appressed hairs vs. erect), and ensis, both geographically and morpholog­ its proclivity for calcareous soils (vs. volca­ ically. nic or igneous). All of the above cited spec- NUMBER 9 TURNER: PHILADELPHUS OF TRANS-PECOS, TEXAS 39

...... W::-- --- I

FIG. 3. Distribution of Philadelphus serpyllifolius in Texas: var. intermedius(circles); var. serpyl lifolius (dots).

imens of P. s. var. intermedius have ap­ 3b. PHILADELPHUS SERPYLLIFOLIUS var. pressed hairs on their upper leaf surfaces SERPYLLIFOLIUS except for Warnock 11414 (SRSC) which has at least a few erect hairs, these occurring The type of Philadelphus serpyllifolius on the upper third of the blade. Were fu­ ( C. Wright 1100, GH) is from the upper ture workers to combine P. s. var. inter­ part of Limpia Canyon, Jeff Davis Co., Tex­ medius with P. texensis, and these treated at as, first collected in 1851 and lectotypified the varietal level, its correct epithet would by Hu (1954-56). be var. texensis, this validated by Hu's de­ The species is relatively common in scription of P. texensis var. coryi S. Y. Hu, volcanic soils of the Davis Mts. (Fig. 3 ), but also a taxon of central Texas (Bandera Co.), is not known to occur in calcareous soils, and known only by its type. as noted under the discussion of Philadel­ Philadelphus texensis, sensu Hu (1954- phus serpyllifolius var. intermedius. Philadel­ 56, in part), is sympatric with P. ernestii S.Y. phus serpyllifolius has been said to occur in Hu, the latter also confined to central Tex­ New Mexico, but such records are probably as. The latter species is readily distinguished based upon type material collected by C. from the former by the vestiture on the un­ Wright from Trans-Pecos, Texas. der surfaces of its leaves, these lacking the Finally, I view the term subspecies as a frizzy hairs characteristic of P. serpyllifolius. clustering category and/or an infraspecific Except for this singular feature the two spe­ category to denote degree of biological dif­ cies are very similar. ferentiation, much as the term subgenus is LUNDELLIA 40 DECEMBER, 2006.

used for generic classificatory purposes (cf. Hitchcock, C.L. 1943. The xerophyllus species of Turner and Nesom, 2000). Varieties are Philadelphus of southwestern North America. Ma­ usually viewed as closely related allopatric drano 7: 35-56. populational categories that intergrade over Holmgren, P.K. and N.H. Holmgren. 1997. Phila­ delphus, Pp. 10-12 in Cronquist et al. (eds.) In­ a relatively short distance near or in regions termountain FL 3. of contact. Further, varieties do not nor­ Hu, S.Y. 1954-56. A monograph of the genus Phil­ mally grow together, for allopatric infraspe­ adelphus. J. Arnold Arbor. 35: 275 333; 36: 52- cific intergradation is a populational phe­ 109; 37: 15-90. nomenon, as clearly noted by Turner ( 1956, Johnston, M.C. 1988. The Vascular Plants of Texas, an updating. Austin: Published by the author. 1984) and yet others. Powell, A.M. 1998. Trees and Shrubs of the Trans­ Pecos and Adjacent Areas. Austin: University of ACKNOWLEDGEMENTS Texas Press. Rydberg, P.A. 1905. Philadelphus, in N. Amer. Fl. 22: I am grateful to my colleague Guy Ne­ 162-175. som for the Latin diagnosis, and to Mike Soltis, D.E. and P.S. Soltis. 1997. Phylogenetic re­ lationships in Saxifragaceae sensu lato: a compar­ Powell, Martin Terry, and Jackie Poole for ison of topologies based on 18S rDNA data and their reviews and helpful comments. The rbcL sequences. Amer. J. Bot. 84: 504-522. dot maps are based upon herbarium spec­ Turner, B.L. 1956. A cytotaxonomic study of the ge­ imens housed at SRSC, and TEX; thanks are nus Hymenopappus (Compositae) Rhodora 58: extended to both institutions for permis­ 691-694. sion to study the same. ---. 1984. Taxonomy of the genus Aphanoste­ phus. Phytologia 56: 81-101. ---. and G.L. Nesom. 2000. Use of variety and LITERATURE CITED subspecies and new varietal combinations for Sty­ rax platanifolius (Styracaceae). Sida 19: 257-262. Correll, D.S. and M.C. Johnston. 1970. Manual of ---, H. Nichols, G. Denny, and 0. Doron. 2003. the Vascular Plants of Texas. Renner: Texas Re­ Atlas of the Vascular Plants of Texas, Vols. 1 and search Foundation. 2, Sida Bot. Misc. 24.