Pemphigus obesinymphae

The North American aphid Pemphigus populi-transversus Riley (Hemiptera: Aphididae: Pemphiginae) forms galls on the leaves of the eastern cottonwood (Populus deltoides; Salicaceae) throughout the eastern and central United States. From the 1940s to the 1990s, Dr. Robert Sokal and his associates published extensively on character and geographical variation in P. populi-transversus in the central and eastern United States (citations summarized in Sokal et al. 1991). One of their more intriguing observations was that most populations surveyed exhibited two distinct gall types, which they named ‘‘elongate’’ and ‘‘globular,’’ based on their shapes (Bird et al. 1979; Sokal and Riska 1981). The aphids forming the two types differed at several allozyme loci and in the seasonal production of sexuals, with the sexuals of the elongate-type appearing in the fall and those of the globular-type appearing in the spring (Faith 1979a,b; Rhomberg 1980; Setzer 1980). Faith (1979a,b) hypothesized that the unusual gall polymorphism could reflect these underlying differences in life cycles, suggesting a rare case of incipient allochronic speciation.

The relationship between P. populi-transversus and a closely related aphid species, P. obesinymphae Aoki, has been of the most confusing among North American Pemphigus species (Aoki and Moran 1994; Blackmon and Eastop 1994). Pemphigus obesinymphae was originally described in the southwestern United States on Populus fremontii (a cottonwood that is likely the sister taxon of P. deltoides; Eckenwalder 1996). Pemphigus obesinymphae was found to share many morphological and life-history characters with the globular-gall morph of P. populi-transversus, including the globular gall shape and the spring sexuals noted by Faith (1979a,c). Then, in groundwork for this study, we found that the globular morph is in actuality a distinct population of P. obesinymphae, and not a morph of P. populi-transversus. We clarify this relationship here to coordinate our study with the existing literature on P. populi-transversus. Hereafter, to limit further confusion, we will recognize the species that forms the elongate gall alone as P. populi-transversus. The species that forms the globular gall is P. obesinymphae, and is distributed across the United States on two hosts: the eastern cottonwood P. deltoides (on which it is sympatric with P. populi-transversus) and the Fremont cottonwood (on which P. populi-transversus is absent).

This analysis is necessary because of the complex distribution of P. obesinymphae: as we have described, the species consists of two lineages, one sympatric with P. populi-transversus on P. deltoides, and one restricted to southwestern North America on Populus fremontii (Aoki and Moran 1994; Abbot and Moran 2002).

In early spring, a first burst of growth occurs from embryonic leaves that have overwintered in the buds. These ‘‘early’’ leaves are followed by a ‘‘late’’ flush of summer leaves, most of which are newly initiated on the growing shoots. Pemphigus populi-transversus foundresses, having overwintered as eggs, form galls on the early, spring leaves. Pemphigus obesinymphae, by contrast, returns to P. deltoides after the first flush of spring leaves, and forms galls on the late flush of summer leaves. Having appeared later on trees than P. populi-transversus, it also remains longer, well into the fall after P. populi-transversus has departed for its secondary hosts (Bird et al. 1979; Fig. 1).

As expected, there was no evidence that P. populi-transversus forms galls on Fremont cottonwoods, despite a westerly distribution on P. deltoides that overlaps with the eastern boundary of P. fremontii’s range (Fig. 2). Overall, the distribution of P. populi-transversus is in complete accord with the surveys of Sokal and associates.

Pemphigus populitransversus

[Drawings of the two "morphs" of P populitransversus appear on page 2 of the pdf]

On a single, isolated tree in West Point, Georgia, Sokal discovered the simultaneous occurrence of two apparently distinct types of galls, either "globular" or "elongate" in form (see Fig. 1). The latter were more cylindrical than the former and generally were smaller. Field examination of contents of some of the galls suggested differences in the biology of the inhabitants. The elongate galls contained many alates (in addition to nymphs and one stem mother each), while the globular galls held few if any alates, but possessed a stem mother and nymphs.

Pemphigus obesinymphae Aoki & Moran

Galls are formed on the petioles of late flushing leaves of Populus fremontii (Aoki & Moran 1994), and P. deltoides (Abbot 2009), usually close to the leaf blade. When mature they are almost spherical, 8-16 mm in diameter, with the lips of the transverse slit sometimes bulging outwards, and a circular ostiole about 1.2 mm in diameter somewhere along the slit. First instar progeny of the fundatrix in the gall are heavily sclerotised with well-developed fore legs and attack predators (Moran 1993). Abbott (2009) demonstrated using microsatellite DNA analysis that there was a high level of clonal admixture within galls, and Grogan et al. (2010) investigated its consequences. Alatae (BL 1.9-2.5 mm) fly in August-October, presumably migrating like P. populitransversus to cruciferous hosts (N. Chen et al. 2009), on which overwintering occurs. Sexuparae return to Populus in May of the following year, and the eggs hatch only a short time after oviposition. Apparently widely distributed in USA (Ohio, Tennessee, Arizona, Utah, Texas), and probably in northern Mexico. N. Chen et al. (2009) recorded molecular and phenological differences between P. obesinymphae and P. populitransversus on cabbage in Texas. This species was confused for many years with P. populitransversus (q.v.), and its distribution and range of both primary and secondary hosts requires further confirmation.

Pemphigus obesinymphae, new species

[A drawing of the gall appears on page 8 of the pdf]

Gall (Fig. 3d). The galls are formed on the petioles of late flushing leaves; most are adjacent to the leaf blade. The shape is roughly spherical, with the lips of the gall slit sometimes bulging outward. Diameters of mature galls ranged from 8 to 16 (mean 13.8). Walls of mature galls are approximately 1.5 thick. The gall slit is irregular in shape and extends along the equator of the gall for about 1/3 of the circumference, with the petiole and leaf blade at the poles. The lips remain closed along most of their length until the emergence of the adult fundatrigeniae. However, each gall has a single, round ostiole, 1.2 in diameter; this may occur anywhere along the slit.

Material examined: All collections of P. obesinymphae were from petiole galls on Populus fremontii Watson in Arizona and Utah, USA, and were taken between 15 July 1991 and 31 July 1992. Arizona locations include: Santa Cruz County: Sonoita, Canelo (Audubon Research Ranch), Patagonia; Cochise County: Portal (Southwest Field Station); Graham County: Thatcher; Pima County: Tucson (Molino Basin, Santa Catalina Mtns.). The Utah localities include Salt Lake County: Salt Lake City (City Creek Canyon) and Davis County: Uintah (Weber River). Some alates that emerged from galls were confined in vials, and their 1 st instar offspring were obtained.

The holotype is an alate fundatrigenia, containing embryos, collected from Populus fremontii by N. A. Moran at Patagonia on 13 October 1991. The paratypes are two adult fundatrices collected at Patagonia on 13 October 1991, two fundatrices collected at Patagonia on 8 September 1991 and 19 October 1991, one fundatrix collected at Portal on 16 September 1991, five emigrants collected at Patagonia on 25 August 1991, six emigrants collected at Patagonia on 13 October 1991, ten 1st instar fun datrigeniae, 14 2nd instar fundatrigeniae, and one 3rd instar fundatrigenia collected at Patagonia on 13 October 1991, and 12 1st instar exules deposited by emigrants collected at Patagonia on 25 August 1991. All the paratypes were collected from P. fremontii by N. A. Moran. The holotype and eight paratypes will be deposited in the U. S. National Museum, Beltsville, Maryland, U.S.A. An additional 16 paratypes will be deposited at the British Museum of Natural History, London and the Canadian National Museum.

Pemphigus obesinymphae, n. sp. is closely related to Pemphigus knowltoni Stroyan and, especially, to Pemphigus populitransversus Riley. We hypothesize that the three constitute a monophyletic species group within the genus Pemphigus. Both P. obesinymphae and P. populitransversus form galls only on Populus fremontii, whereas P. knowltoni uses only Populus angustifolia. When adult fundatrigeniae are not present, the presence of a gall ostiole and morphological features of the 1 st and 2nd instar fundatrigeniae can be used in order to distinguish P. obesinymphae and P. populitransversus in galls on Populus fremontii (Table 1).

[Table 1 shows that obesinymphae is known from Populus fremontii in AZ and UT and likely in NM, CO and northern Mexico, is found on petioles typically near the leaf blade, has an ostiole as well as a slit, colonizes Populus fremontii in the spring and galls late flushing leaves; that populitransversus is known from Populus fremontii throughout North America, is found near the center or elsewhere on petioles of first flush leaves, colonizes its host in fall, has no ostiole; that knowltoni is known from Populus angustifolia in UT, possibly also in ID, CO, WY, MT, galls leaf midribs, colonizes first flush leaves of its host in the fall]

Biology: The galls of P. obesinymphae contained aphids of only two generations: the fundatrix and fundatrigeniae, as is typical of Pemphigus. No apterous fundatrigeniae were found in galls of obesinymphae, although such forms are usual in both Pemphigus spirothecae Passerini and Pemphigus monophagus Maxson, other species that are known to have long-lived galls and defenders or suspected defenders (Aoki and Kurosu, 1986, 1988). The adult fundatrigeniae possessed embryos resembling typical Pemphigus exules. Therefore, this species probably alternates between two different host-plant taxa. Despite considerable searching and attempted transfers to several suspected plants (Chenopodium and Poinsettia species), its secondary hosts are yet unknown. The host alternation shows peculiar seasonal timing: fundatrigeniae fly from galls much later in the season than in other Pemphigus, overwintering occurs on secondary hosts, and sexuparae return to P. fremontii in May of the following year, after the first leaves have expanded. Sexuparae were collected from 5-20 May at Patagonia in 1992; they produced sexual offspring of typical appearance for the genus. Hatching fundatrices were collected from the same clusters of sexuales, suggesting that the eggs hatch within a short period rather than overwintering. Hatching fundatrices then establish galls on late flushing leaves; gall establishment was observed at the Patagonia site from 15 May-l June 1992. The first fundatrigeniae began to be deposited in some galls on 10 June. This life cycle recalls that of Schlechtendalia chinensis (Bell), which also produces sexuparae in spring and which also has eggs that hatch without prolonged dormancy or overwintering (Takada, 1991).

P. obesinymphae exhibits four peculiar features: First, the life cycle is dramatically altered (Moran, 1993), with the result that galls occur only on the late leaves that expand later in the season as the shoots elongate. Most Pemphigus initiate galls on the early leaves; however, Bird et al. (1979) and Faith (1979) reported a dimorphism in the timing of gall initiation within P. populitransversus, with one gall type restricted to first flush leaves and a second restricted to late leaves. Second, 1 st instar nymphs of the 2nd generation attack predators (Moran, 1993) and have the legs well developed and the tergites heavily sclerotized. Defensive 1 st instar nymphs have been reported in Pemphigus dorocola (Aoki, 1978) and P. spirothecae (Aoki and Kurosu, 1986; Foster, 1990), but their tergites are not so strongly sclerotized as in P. obesinymphae. Third, galls have an ostiole, through which defenders enter and exit the gall (Moran, 1993). Ostioles are characteristic of galls of Pemphigus dorocola (Aoki, 1978) and P. spirothecae (Foster, 1990), but are not known from other American Pemphigus. Fourth, 2nd and 3rd instar nymphs of the 2nd generation are usually much swollen, and look as if they were wingless adults. Such swollen nymphs are not known from galls of any other Pemphigus species. Why they are so swollen remains to be investigated.

Pemphigus populitransversus [Felt's image shows the "globular" morph of P populitransversus now considered part of P obesinymphae]

Oval petiole galls with transverse openings and near middle of leaf stem, aphid, summer

Pemphigus populitransversus [the "globular morph"]

Gall initiation by Pemphigus populitransversus was found to occur at two clearly separate times during the summer of 1975. Gall initiation is first detected as a slight swelling and bending of the petiole at the point where a fundatrix nymph has begun feeding. In this early stage, the young fundatrix is not enclosed and appears as a small black spot in the pocket formed by the swelling and bending of the petiole. Transverse galls first began forming, along with oblique galls, around May 15 on cottonwoods in the vicinity of Stony Brook. At this time, the early leaves of the tree were expanding and provided suitable sites for gall initiation. The trees were checked carefully on successive days, and after two or three days, there was no further gall initiation. Then, on June 15, more transverse galls began forming on the newly formed late leaves. This dimorphism in the timing of gall initiation resulted in a set of galls on the early leaves and a set of galls on the late leaves of the shoots.

A dimorphism in gall shape, life history, and aphid morphology has been described for this species by Senner and Sokal (1974). The morph they refer to as "elongate" produces galls that are more elongate in shape than those of the other morph called "globular." The question immediately arose whether the presence of early and late transverse galls was related to the globular-elongate dimorphism. An analysis of the early and late transverse galls confirms the suspected relationship (Table 1). The late galls are larger than the early galls, and more spherical, as shown by the ratios of gall dimensions. The values in Table 1 resemble those of the sample from West Point, Georgia, shown in Table 3 of Senner and Sokal (1974). The early and late galls were significantly different for each character. Corroborating tests were run following methods developed by Senner and Sokal (1974). The gall dimensions and their ratios split a mixed sample of 71 galls into early and late classes fairly well. Only two galls would be misclassified by time of gall formation when the sample was split according to gall depth and only 4 galls would be so misclassified on the basis of length/width ratio. It is highly probable that the early and late galls correspond to the elongate and globular morphs, respectively.

PEMPHIGUS POPULI-TRANSVERSUS Riley [photos show the morphotype now known as P obesinymphae]

The details of the life history of this species are very well known because of the studies of Jones and Gillette (1918), who have shown that it alternates between poplars and the roots of various cruciferous plants. The winged migrants (fundatrigeniae) leave the galls (Fig. 48) on poplars in spring or summer and fly to cruciferous plants, where the sexuparae are eventually produced. The latter return to the trunks and branches of poplars in spring and produce the sexual forms which produce eggs from which the stem mothers hatch. The stem mothers start tile galls, each of which produces a large number of winged females or migrants.

We have been unable to separate the cotypic specimens of P. rubi Thomas (1879), which are sexuparae, from sexuparae of P. populi-transversus Riley. The similarity in their antennae, which are quite different from the antennae of other sexuparae taken in Illinois, seems to us quite indicative that they are the same species.

Data associated with our Illinois material on Populus are as follows: Danville (Sept. 18), Elizabethtown (October 12), Golconda (October 13), Havana (.June 21), Mahomet (October 2), Mt. Carmel (May 26). Pekin ( July 22), 1928; Parmer City (September 22), Herod (Sept. 27), Oakwood (October 13), Starved Rock State Park (July 5), 1929; Fairmount (September 28), Watson (Nov. 13), 1930.

In addition, the sexuparae have been taken on roots of Lepidium virginicum at Mt. Carmel on May 26, 1928. Previously recorded from Illinois by Davis (1910) and Gronemann (1930).