Scatella melanderi (Cresson 1935)

Scatella melanderi (Cresson)

Parascatella melanderi Cresson, 1935:358.—Wirth, 1965:757 [catalog].

Scatella (Parascatella) melanderi.—Sturtevant and Wheeler, 1954:179.

DIAGNOSIS.—Specimens of this species are easily distinguished from those of similar congeners by the following combination of characters: mesofrons mostly dull, pollinose, weakly differentiated from surrounding parafrons; antennal arista very long, almost equal to twice the combined length of other antennal segments; face from interfoveal carina to oral margin unicolorous, pollinose, grayish brown; wing distinctly infuscated, brown; white spots in wing membrane contrasting distinctly; general coloration of abdomen grayish brown to blackish brown.

DESCRIPTION.—Medium-sized shore flies, length 3.07 to 3.91 mm; generally dull, pollinose, grayish brown to grayish blue, abdomen darker in color and shinier.

Head (Figure 59): Mesofrons brown, more or less concolorous with disc of mesonotum, at most subshiny, usually heavily pollinose, vestiture weakly differentiated from parafrons; parafrons pollinose, usually darker in color than mesofrons, grayish charcoal or slightly blackish blue anteriorly; fronto-orbits brown, pollinose; dorsalmost postocular setae weakly developed, at most slightly larger than postocellar setae and much smaller than ocellars; 2 pair of large, lateroclinate fronto-orbital setae. Antennal segments mostly unicolorous, brownish black to black; third segment more pubescent, appearing darker from some angles; arista long (Figure 58), length subequal to twice the combined length of first 3 antennal segments, micropectinate above on basal two-thirds. Face except for antennal fovea mostly unicolorous, grayish brown; antennal fovea lighter in color, whitish gray; facial setae extending obliquely ventrad from interfoveal carina to posteroventral corner of face and those along oral margin distinctly larger, remaining facial setae uniformly smaller; interfoveal carina with distinct dorsal carina. Gena lighter in color than face, with distinct greenish tinges but also with light tan areas.

Thorax: Generally grayish brown to gray, mostly pollinose, dull, darker dorsally; mesonotum with faint indication of vittae along major setal tracts, slightly darker brown; scutellum more or less concolorous with mesonotum or slightly darker. Mesonotum and pteropleuron mostly concolorous, both more or less concolorous with gena; lower portion of sternopleuron grayer and front coxa distinctly grayer, contrasting with rest of pleural areas. Legs generally dark, grayish black; femora with slight greenish color; tibiae and tarsi darker, especially dorsally; ventral surface of tarsi often quite pale, tawny. Wing (Figure 54) infuscated, brown; larger white spots subrectangular and very distinct from surrounding infuscation; white areas in cell R3 slightly apicad of alignment of posterior crossvein.

Abdomen: Generally unicolorous, dark, greenish brown, but overlaid with grayish pollinose vestiture; posterior segments tending to become darker and less pollinose; often subshiny; abdominal segments 2–4 more or less subequal in length in male specimens.

TYPE-MATERIAL.—Holotype male, labeled: “Tacoma, Wash./TYPE Parascatella MELANDERI E. T. Cresson, Jr. 6526 (red).” The holotype is in good condition, is pinned directly, and is in the Academy of Natural Sciences of Philadelphia, type number 6526. Cresson's original description also lists 4 male paratopotypes, all collected by A. L. Melander.

SPECIMENS EXAMINED.—158 specimens (79 , 79 ). CANADA. BRITISH COLUMBIA: Vancouver, Point Grey (CNC, USNM). UNITED STATES. CALIFORNIA: San Mateo Co., Memorial Park USNM); Santa Cruz Co., Capitola (CA), Davenport (CA, USNM), Santa Cruz (CA, USNM); Sonoma Co., Stillwater Cove (USNM). OREGON: Marion Co., Silver Falls State Park–South Silver Creek Falls (CA); Tillamook Co., Neskowin (WNM). WASHINGTON: Clallam Co., Sequim (USNM), Sequim Bay (USNM); Grays Harbor Co., Hoquiam (USNM); Lewis Co., Chehalis (ANSP); Mason Co., Lake Cushman (USNM); Pacific Co., Ilwaco (ANSP, USNM); Pierce Co., La Grande, 3 mi N (WNM), Tacoma (ANSP, USNM).

GEOGRAPHIC DISTRIBUTION (Figure 60).—Scatella melanderi occurs along the coast of the Pacific Northwest from Vancouver, British Columbia, in the north to Santa Cruz County, California, in the south. This distribution pattern overlaps that of S. marinensis, although we do not have specific collecting records of S. melanderi occurring with that species.

NATURAL HISTORY.—Mathis has collected specimens of this species along freshwater streams that eventually drain into the Pacific Ocean. Sweeping through grasses and other emergent vegetation associated with protected, moist areas, especially near embankments, yielded better catches, although specimens were not collected commonly anywhere. The specimens from Vancouver, British Columbia, were collected by Dr. J. R. Vockeroth at Point Grey and were labeled with a habitat description as follows: “On seepage on earth cliff.”

PHYLOGENETIC RELATIONSHIP.—Scatella melanderi is apparently the sister species to S. triseta and S. marinensis; however, no apotypic character state has been discovered to establish this relationship.

Evolutionary Considerations

Most recent classifications divide the subfamily Ephydrinae into two tribes: Ephydrini and Scatellini (Wirth, 1965, 1968). Ephydrini is characterized by several synapotypies (Mathis, in preparation) and is undoubtedly monophyletic. Scatellini, however, as presently recognized, is an assemblage of convenience. Genera included are largely the residue of the subfamily Ephydrinae that cannot be placed in Ephydrini. It is likely that these will be arrayed into several tribes once their relationships are clarified, which is beyond the scope of the present paper.

Various apparently monophyletic assemblages of genera can be recognized, however, within the tribe Scatellini. One such group, comprising genera that are similar in appearance to the genus Scatella Robineau-Desvoidy, include (besides the latter) Limnellia Malloch, Neoscatella Malloch (here treated as part of Scatella), Parascatella Cresson, and Scatophila Becker (and perhaps Apulvillus Malloch, not studied). With the exception of Scatophila, these genera plus Lamproscatella Hendel were previously treated as an informal group (Wirth, 1948) and a key was provided for their identification. We are excluding Lamproscatella from further consideration here because it appears to be more closely related to another assemblage of genera (Mathis, in prep.).

The assemblage related to Scatella is characterized as follows.

1. Wing pattern: The wing pattern appears to be a simple morphocline in which the plesiotypic state is a hyaline wing, characteristic of most members of Ephydrinae. At the next level, the wing becomes lightly infuscated, grayish to tan, with a pattern of white areas arranged as follows: cell R1 with one white spot, sometimes subquadrate, more or less aligned or slightly apicad of posterior crossvein; cell R3 with two white spots on either side of posterior crossvein; discal cell with one white spot, often subdivided, near the apical end; cell M2 with one basal white area, frequently large and irregular in shape and/or subdivided; cell M4 with one to two white areas, usually in apical one-half. Some inter- and intraspecific variation occurs, usually by the addition of white spots.

2. Gonite shape: Typically the gonites are paired structures extending from the sides of the aedeagus, connecting the latter with the hypandrium. In the Scatella assemblage, the hypandrium becomes rudimentary or is lacking, and the gonite is modified to form an angulate structure with a stout “arm” that extends along the same plane as the aedeagus and a simpler, narrow ventral process. The apices of each ventral, gonal process fuse, to form a loop through which the aedeagus projects. This condition becomes more pronounced in the genera immediately related to Scatella, such as Parascatella.

3. Aedeagal apodeme: The shape and attachments of this structure appear to form a duo morphocline. The generalized condition in Ephydrinae and throughout most of the family is for the aedeagal apodeme to be attached basally to the hypandrium as in Ephydra (Wirth, 1971) and Paracoenia (Mathis, 1975), and for the structure to be well developed but laterally flattened. From this condition, the structure becomes rudimentary. In males of Scatophila and Limnellia, the aedeagal apodeme still remains attached to the base of the aedeagus and to the rudimentary hypandrium or to the fused apices of the ventral gonal processes. But in males of Scatella and Parascatella, the apodeme is detached from the hypandrium or ventral gonal processes and is only loosely attached to the base of the aedeagus.

The cladogram (Figure 61) plus the accompanying list of supportive character evidence (Table 1) summarizes the relationships among the species and species-groups of Parascatella and Scatella herein treated as we understand them at present.

Our concept of Parascatella, as characterized above, is a conveniently recognized, monophyletic group. The species and species-groups comprising the genus are closely related as evidenced by the synapotypic character states that establish the monophyly of the genus (characters 1–2). Additionally, the species and species-groups are very similar to each other, making the genus an easily recognizable taxon.

Using these criteria, we are excluding from Parascatella the species included in what we are designating as the triseta group, preferring to include them in Scatella sensu lato. We suggest that the latter genus is also monophyletic as recognized by the character states indicated on the cladogram (characters 1', 3–4). We are treating the triseta group as an informal species-group because its basis is patristic, lacking characterization by an apotypic character state. The monophyly of the remaining species of Scatella (including Neoscatella) is clearly evident, however, being established by three synapotypies (characters 24–26).

The reason we are treating the species of the triseta group here is because they have been included in Parascatella by most students of the family, either when Parascatella was a distinct genus (Cresson, 1935; Wirth, 1965, 1968) or a subgenus of Scatella (Sturtevant and Wheeler, 1954). Also, the triseta group is closely related to Parascatella, and we had already gathered the basic data pursuant to a taxonomic treatment and did not feel a second paper would be any more effective in communicating our results of that group.

The first major dichotomy in the phylogeny of Parascatella is the differentiation of the brunnea and pilifera groups. The monophyly of the pilifera group is indicated by many synapotypies (characters 6–8), while the brunnea group is defined by only one, subjectively determined apotypic character state (character 5).

Within the pilifera group, the major weakness in the proposed phylogeny is the suggested relationship between P. balioptera and P. penai and the remaining species. We suspect these two species are closely related and that they are more closely related to those species having a setal comb along the posteroventral margin of the front leg (P. glabra, P. spinicrus, P. lanicrus, and P. pilifera) than to the species with a large white spot in cell R3 of male specimens (P. apicalis, P. hirticrus, P. semicinerea, and P. semipolita). A second lineage, that of P. semicinerea and P. semipolita, lacks characterization by apotypic character states, but the large size of specimens of these species may be synapotypic.

Because species of Parascatella reproduce sexually, speciation is likely to occur only when populations are isolated. Accordingly, where species are now sympatric, we assume that they were once disjunct and that the zone of overlap is secondary. Of the ten species presently comprising the pilifera group, eight frequent similar aquatic habitats in the provinces of Jujuy and Salta in northwestern Argentina and adjacent areas of Chile (Figure 62). It is not uncommon to collect several species, including sister species at the same locality. For example, the sister species P. apicalis and P. hirticrus are known from collection data to frequently occur together. We can speculate, therefore, that during the recent past the parent populations of these two species were separated by a barrier, allowing differentiation (vicariance), and that the present extensions of their overlapping distributions developed subsequently (dispersal). During the differentiation of the pilifera group, these events apparently occurred repeatedly, some undoubtedly concurrently, to result in the speciation pattern now evident for the group.

According to Simpson (1975), the region now occupied by these species, the Altiplano, is “a broad relatively flat high plateau stretching from southern Peru to northern Argentina.” The following historical description of the plateau, summarized largely from Simpson (1975), provides a basis for further discussion.

The plateau arose initially from the sea during the Cretaceous but remained low throughout much of the Tertiary. During the Miocene, the plateau was raised further, but not until the mid-Pliocene to possibly the Pleistocene did significant uplift occur. Deep surface deposits now covering the Altiplano are from lacustrine sediments that were probably deposited during the Pleistocene when large lakes covered the area. Today, the Altiplano is a dry, barren steppe with characteristic non-arborescent vegetation known as puna. There is, however, a slight but noticeable transition in the vegetation from open tussock grass shrubland in the dryer west to meadowlike grass scrub formations in the wetter east.

Despite the apparent uniformity of the area now, several studies (F. Vuilleumier, 1968, 1969, 1971) have documented the occurrence of many plant and animal genera with several sympatric species or taxa that have distributional boundries at the same specific localities within this region. Simpson (1975) suggested that during the recent past, populations of many taxa were isolated in at least three major areas and possibly several minor ones. The third major area Simpson discussed is located south of glacial Lago Michin, where the distributions of eight species of Parascatella appear to be centered (Figure 62).

During the Pleistocene, this area was subject to glacial climates (summarized in Vuilleumier, 1971) that lowered both the snowline and timberline and resulted in increased opportunities for migration and contact for organisms inhabiting high-altitude environments along the eastern and western slopes. During the interglacial periods, like the present, however, the climate was considerably dryer. Intermontane valleys lying in the rainshadow would act as barriers due to the drying up and partitioning of aquatic habitats. Ecological barriers as described coupled with physiographic ones, such as the paralleling mountain ranges of this region, would be particularly effective in isolating populations of aquatic organisms, such as Parascatella members.

Dating these events presents difficulties, especially as greater precision is sought. For our purposes, it is sufficient to establish rough estimates. Because the final uplift of the Altiplano dates to the end of the Tertiary, the age of the high plateau habitats and their endemic biotas is limited to the late Tertiary and Quaternary. From this, we suggest that the pilifera group arose and differentiated within the last ten million years.

The zoogeography of Parascatella can be explained satisfactorily by reference to Figure 62, without recourse to a complicated exposition. Perhaps this simple pattern is an artifact of our data base, our interpretation, or is due to the relative recency of the genus.

Evidence for our explanation is as follows. Parascatella is known to occur only along the western cordilleran system of South America below 10° south latitude, where present-day climatic conditions are temperate. The distribution of the genus is fairly continuous along this system, where they have apparently differentiated. Differentiation can be explained by vicariating events accompanying the changing geography and climate of the Altiplano during the ebb and flow of glaciation and the general uplift of the area in general.

We speculate that the ancestral stock, from which Parascatella arose, originated in South America. The South American fauna is by far more diverse (Wirth, 1968) and all genera closely related to Parascatella except Apulvillus are represented there. The latter genus probably arose more recently in the Pacific Islands.

Following differentiation of Parascatella from the ancestral lineage, which also gave rise to the Scatella-Neoscatella complex, two further radiations occurred, both in the Neotropics. The brunnea group, presently known from only P. brunnea, developed in the Tierra del Fuego region of the southern Andes. Radiation in the higher elevation habitats of the Andes Mountains produced most of the known species of Parascatella.

In summary, we suggest that Parascatella arose in South America from an ancestor that was common to it and to the genus Scatella. Parascatella then underwent two major radiations: the brunnea group developing in the Tierra del Fuego region and the pilifera group radiating in the Altiplano from southern Peru to northern Argentina and Chile.