Alloberberis trifoliolata (Moric.) C.C. Yu & K.F. Chung

Berberis trifoliata Hartw. ex Lindl., Bot. Reg. 27 [= n.s., 4]: [Misc.] 68. 1841; M.J. Young, Familiar Lessons Bot. 152. 1873. TYPE: MEXICO. Chihuahua:Rocky hills near Chihuahua, 25 Mar 1885 (fl), C.G. Pringle 261 p. p. ( NEOTYPE, designated here: US [ US 00103926]!; ISONEOTYPES: P [P0232716 as image!], US [ US 00952515]!).

Mahonia trifoliolata var. glauca I.M. Johnst, J. Arnold Arbor. 31:190. 1950. TYPE: MEXICO. Chihuahua:Rocky hills near Chihuahua, 25 Mar 1885 (fl), C.G. Pringle 261 p. p. ( LECTOTYPE, designated here: US [ US 00103926]!; ISOLECTOTYPES: P [P0232716 as image!], US [ US 00952515]!).

Alloberberis trifoliolata (Moric.) C.C. Yu & K.F. Chung View in CoL , Taxon 66:1387. 2017.

Berberis trifoliolata Moric., Pl. Nouv. Amérique 113, t. 69. 1841.

TYPE: MEXICO [now U.S.A. Texas:]. Entre Laredo et Bejar , Mar 1828, J.L. Berlandier 1437 ( LECTOTYPE, designated here: G [ G00342708 as image!] ; ISOLECTOTYPES: G [ G00342706 as image!], G [ G00342707 as image!], GH [ GH00038685 as image!], NY [ NY00000019 as image!], P [ P00752258 as image!], P [ P00752259 as image!], P [ P00752260 as image!], WU [WU sheet no. 0064710 as image!] ).

In her treatment of Berberis L., Young (1873) recognized both “ B. trifoliata ” and “ B. trifoliolata, Torr. ” and provided them with very slightly different descriptions and distributions, although as she circumscribes them, they have overlapping morphology and geography and cannot really be distinguished. Both names antedate her flora, and both were published in 1841 as B. trifoliata Hartw. ex Lindl. and B. trifoliolata Moric. , respectively. The former name has been placed in synonymy under the latter ( Watson 1878), but priority cannot be established with certainty because while we know B. trifoliata was published in September 1841, only the year of publication is known for B. trifoliolata . Frequently the epithet of the former also is used mistakenly in place of the latter. Young’s (1873) treatment of Berberis appears to be derived from an earlier note by Buckley (1870a) where he discussed the “Three-Leaved Berberry. ( Berberis trifoliata .)” and was silent regarding B. trifoliolata . Elements of Buckley’s description of “ B. trifoliata ,” including its habit, distribution, and edible fruit, strongly suggest that it is the same as the description of “ B. trifoliolata, Torr. ” in Young’s flora.

Torrey (1857, 1859) also confounded “trifoliata ” and “trifoliolata ” when he treated Berberis from Texas and the Southwest. Invariably, however, he cited Moricand suggesting that his “ B. trifoliata ” was nothing more than a misreading of the epithet. Torrey (1857) treated a blue-berried species as “ Berberis trifoliata, Moricand, Pl. Amer. , t. 69?” and not only corrupted the epithet but misapplied the name to a species that he later described as B. fremontii Torr. [≡ Alloberberis fremontii (Torr.) C.C. Yu & K.F. Chung ]. In his contribution to the Botany of the Boundary Survey, Torrey (1859) wrote “ Berberis trifoliata [sic], Moric. Pl. Nuov. Amer. p. 113, t. 69 … Western Texas, and on hills near the Copper Mines, New Mexico; Bigelow.” This latter treatment is the source for part of the locality and the sole collector cited by Young (1873), viz. “Austin, thence west to New Mexico.– Bigelow.”

The protologue of Berberis trifoliata Hedw. ex Lindl. indicates that the name was based on plants grown in England from seed collected in Mexico near Hacienda del Espiritu Santo on the road from Zacatecas to San Luis Potosi by “Mr. Hartweg.” The seeds were then distributed by the Horticultural Society of London, but no original material has been traced. A neotype is designated here that fixes the application of the name to the glaucous-leaved form of Alloberberis trifoliolata found in both Mexico and the southwestern U.S.A.

Johnston (1950) intentionally redescribed Berberis trifoliata as Mahonia trifoliolata var. glauca . He indicated that “Pringle 261” was the type of this variety, but he did not specify where he examined material. Yu and Chung (2017) stated that a specimen at US was the “ holotype,” but this is incorrect, and their designation cannot be treated as a misused term ( Turland et al. 2018; Art. 9.10) because other provisions of the ICN are not met ( Turland et al. 2018; Arts. 7.11, 9.23). Furthermore, Yu and Chung (2017) failed to notice that each sheet of “Pringle 261” is comprised of two separate gatherings:the labels clearly state “Flowers, 25, March; fruit, 20, May.” The flowering material only is designated here as the lectotype of this varietal name.

Yu and Chung (2017) considered a specimen of “Berlandier 1437” in Geneva (G) to be the “ holotype ” of Berberis trifoliolata . The name, however, was published without a holotype and the designation by Yu and Chung also cannot be treated as a misused term for the reasons cited in the previous paragraph. The lectotype designated here is the same specimen that they incorrectly assumed was the holotype.