Vol. 9.

TORREYA

February, 1909

NORTH AMERICAN ROSE RUSTS *

BY J. C. ARTHUR

No. 2.

From the days of Schweinitz, that is, the times of the first studies of American fungi, down to the near present, all rusts upon roses in North America had been placed under two species, i. e., Phragmidium speciosum, a strictly American form, and P. subcorticium, a cosmopolitan form. The latter name has many synonyms, P. mucronatum having been especially popular, but the earliest and consequently the rightful name appears to be P. disciflorum, and therefore will be used in this paper.

In 1876 Peck vaguely called attention in his twenty-eighth Report of the Botanist of the New York State Musuem (page 86) to a variation in teliospores that he had observed. His words are "American specimens generally have the spores more opaque, and with two or three more septa than the typical form. This variant form might be called var. Americanum." The variety was placed under P. mucronatum. Two years ago Dietel published an extended taxonomic study of the genus Phragmidium in Hedwigia, and five months later a supplementary article in the same journal (44: 112–132, 330-346). In these two articles Dietel established and well defined four new species of Phragmidium inhabiting American roses, and one new species of Cacoma, C. Rosae-gymnocarpae, from California. This comprises all important taxonomic work upon rose rusts of America up to the present time.

In pursuing the study of American rusts for systematic presentation in the forthcoming North American Flora the genus Phragmidium has been reached, and I desire to give in this

* Read before the Botanists of the Central States, at the Madison meeting, March 29, 1907. Illustrated with the aid of the McManes fund.

[No. 1, Vol. 9, of TORREYA, comprising pages 1-20, was issued January 26, 1909.]

III

paper some of the more interesting results that have come to light pertaining to the forms on roses.

Very little has been learned about the Californian Cacoma.

It

s clearly an aecial stage of the type of Cacoma nitens on Rubus, and like it may belong to the genus Gymnoconia. But as no hint has yet been secured regarding the telial stage, the assignment to any other than a form-genus is hazardous.

3

FIG. I. Spores of the three species of Phragmidium on rose having slender eliospores 1, P. americanum, 2, P. Rosae-setigerae, 3, P. Rosae-californicae ; I, aeciospore, II, urediniospore, III, teliospore.

:

The characters of the rust which has been called Phragmidium speciosum, such as the non-gelatinous pedicels of the teliospores, the large, compact telia, found on the stems, and the absence of a uredinial stage, show that it does not accord with true members of the genus Phragmidium, and justify its separation

under the name Earlea speciosa, made some two years ago. This rust occurs upon any and all species of roses in North America, both wild and cultivated, and extends throughout the United States and southern Canada. Its omnivorous and adaptable habits are in marked contrast with the fastidious and restricted habits of all true species of Phragmidium on roses found in the same region.

In carefully going over the available material of American rose rusts, properly assignable to the genus Phragmidium, the old world species, P. disciflorum and all the species erected by Dietel

FIG. 2. Spores of the three species of Phragmidium on rose having stout teliospores: 4, P. Rosae-arkansanae, 5, P. montivagum, 6, P. disciflorum; I, aeciospore, II, urediniospore, III, teliospore.

are confirmed, as common in North America, together with one additional species now to be described. In defining these species, characters have been drawn from all three stages of the rust, aecial, uredinial, and telial. The new species may be characterized as follows:

Phragmidium montivagum Arthur, n. sp.'

Pycnia amphigenous, gregarious and often confluent, in small groups surrounded by aecia or on spots opposite the aecia, inconspicuous, subcuticular, 80-112 μ in diameter by 30-35 fl high.

Aecia hypophyllous and petiolicolous, 0.4-1.5 mm. across, solitary, or in irregular groups, often confluent over areas 5-10 mm. long, applanate; paraphyses abundant, conspicuous, surrounding each individual sorus, noticeably taller than the sporemass, spatulate-capitate or clavate, 12-25 μ by 50-70 μ, wall evenly thin, 1-1.5 μ; aeciospores globoid or broadly ellipsoid, 16–19μ by 21-26, wall medium thin, 1.5-2 μ, rather sparsely but distinctly

verrucose.

Uredinia hypophyllous, numerous, scattered, round, small, about 0.1 mm. or less across, soon naked, inconspicuous; paraphyses numerous and noticeable, encircling the sorus, cylindrical or slightly clavate, 9-11 by 45-64 μ, wall thin, about Iμ, slightly thicker above on outer side of curve; urediniospores obovate-globoid, 16-19μ by 19-23, wall pale yellow, rather thin, 1-1.5, closely verrucose-echinulate.

Telia hypophyllous, at first arising from the uredinia, numerous, thickly scattered, 0.1-0.5 mm. across; paraphyses none; teliospores cylindrical, 24-29 by 64-96, usually rounded below and narrowed above, cells 6-9, closely and rather moderately verrucose, apex usually with a conical subhyaline papilla7-10 μ long; pedicel rugose when dry, upper half 7-9 μ in diam, eter, lower part swelling in water to 15-30 μ at broadest part.

On Rosa Sayi Schw., Cummins, Albany Co., Wyo., July 26, 1895, Aven Nelson 1499 (type), Crow Creek, Albany Co., Wyo., August 12, 1903, Aven Nelson 8913, Belt Mountains, Mont.,

* Pycniis amphigenis, in greges dispositis, inconspicuis, 80-112 μ diam., 30–35 μ altis.

Aeciis hypophyllis vel petiolicolis, 0.4-1.5 mm. latis, saepe confluentibus, applanatis; paraphysibus conspicuis, marginalibus; aeciosporis subglobosis vel ellipsoideis, 16-19 × 21-25 μ; episporio subhyalino, 1.5-2 μ crasso, verruculoso.

Urediniis hypophyllis, numerosis, minutis, rotundatis; paraphysibus cylindraceis vel clavatis, marginalibus; urediniosporis obovato-globosis, 16-19 19-23μ; episporio dilute flavo, 1-1.5 μ crasso, verrucoso-echinulato.

Teliis hypophyllis, numerosis, sparsis; teliosporis cylindraceis, 24-29 X 64-96 μ, verrucosis, 5-8-septatis, loculo terminali apiculo conoideo hyalino 7-10 u longo ornato; pedicello supra 7-9 μ diam., infra incrassato, oblanceolato vel ellipsoideo, 15-30 μ late.

In foliis Rosa Sayi, Cummins, Wyoming, Julo 26, 1895, Aven Nelson, 1499.

September, 1889, F. W. Anderson; and also on related species of hosts from Colorado and Utah northward in the Rocky Mountains.

Of the rose rusts in North America belonging to the restricted genus Phragmidium there are now to be recognized six valid species, all indigenous but one. Space does not permit, and the needs of this discussion do not require the full characterization to be given for each species, but the following key, when taken in connection with hosts and geographical data, will provide some aid to those persons who desire to determine their collections. Teliospores slender, 8-11-celled.

Walls of aecio- and urediniospores thin, 1-1.5 μ.

Teliospores long, 80-100 μ.

Teliospores very long, 90-130 μ.

Walls of aecio- and urediniospores thick, 2-3μ.

Teliospores long, 90-112 u.

Teliospores stout, 5-9-celled.

1. P. americanum Diet.

2. P. Rosae-setigerae Diet.

3. P. Rosae-californicae Diet.

These six species of Phragmidium have a most interesting distribution, both as to hosts and territory. The one species coming. from Europe occurs chiefly upon thick-leaved roses of the dog and cabbage rose sections, Rosa canina and R. Gallica, their allies and hybrids, and appears to follow wherever these roses are cultivated. It is known throughout the United States from the Atlantic to the Pacific, northward into Canada, and south

ward into Mexico and Central America. It does not appear to have passed over to any native rose.

The distribution of the five indigenous species is shown by the accompanying chart. P. americanum inhabits the northeastern region along the Atlantic coast from Maryland northward and north of the great lakes, chiefly on Rosa blanda, R. lucida, R. Sayi, and certain cultivated varieties derived from these. Rosae-setigerae is only known upon Rosa setigera and R. carolina, extending nearly throughout the region of the hosts from central New York to central Nebraska. P. Rosac-arkansanac is only known

P.