Molotra, Ubick & Griswold, 2011

Ubick, Darrell & Griswold, Charles E., 2011, The Malagasy Goblin Spiders of the New Genus Molotra (Araneae: Oonopidae), American Museum Novitates 2011 (3729), pp. 1-72 : 5-10

Molotra female genitalia

The female genitalia of Molotra are unusual in lacking a receptaculum and are described here in some detail. Females from three species are known and were examined with light microscopy, but only M. molotra was viewed with SEM.

Light microscopy shows that the genitalia are rather simple. In untreated specimens the larger structures can be seen through the cuticle, depending on its transparency, and show two pairs of apodemes and a median T-shaped process (figs. 198, 202, 204) to which are attached muscles (fig. 200). Digested specimens, of course, show this in greater detail. The genitalia of the three species are very similar and the only obvious difference at this magnification is the size of the T-shaped process (TP in figs. 199, 201, 203, 205), but not that of the anterior (AA) and posterior (PA) apodemes, which seem fairly uniform in size. Also visible in this preparation are some transverse ridges between the PA, most distinct in fig. 201.

The external genitalia are very simple. The gonopore is a wide opening with sclerotized margins, especially the outward projecting posterior one (figs. 206, 207). An abdomen was separated along the epigastric furrow, showing much detail. The anterior lip has a broad flat edge and, in its center a triangular, concave process (AMP of figs. 208, 211). This sclerite has two openings, a wide median one where it attaches to the epigastric scutum (AO), and a small pore at the distal, pointed end (figs. 209, 211). The posterior lip also has a median process (PMP), which also has a median opening (PO), albeit much wider (figs. 210, 212). Unlike the AMP, the PMP is convex and has a fine median longitudinal crack (fig. 212). Attached to both the AMP and PMP are membranous remnants that are presumably the walls of the uterus internus (figs. 209–213).

In the internal scans (figs. 214–219), the T-shaped process (TP) is prominent, with its large triangular head and broad muscle attachments. The attachments align with the broad anterior apodemes (AA) which have a similarly scalloped surface (figs. 216–218). The TP is associated with two structures. The first is a low transverse ridge immediately ventrad of the TP stalk, and whose membranous edges can be seen between the arrows in figure 217. The second is a triangular sclerite, whose edge is indicated by the membranous remnants of the uterus (UI in fig. 217), and to which the TP seems to be firmly attached. Based on position, this is the same AMP as described above, and its median pore is aligned with the TP stalk.

The posterior genitalia are dominated by the large foliate posterior apodemes (PA in fig. 216) and a series of five transverse ridges (fig. 219). The posteriormost ridge connects the posterior respiratory spiracles, and the next one encircles the bases of the PA. The three anterior ridges all seem to be connected. The anteriormost has two rounded lateral projections (fig. 217), which are also visible in anterior view at the sides of the PMP (fig. 210). The next two ridges are continuous with the bases of the PA. Between them lies a field of papillae (PF in fig. 215) and a cylindrical median structure (GAp in fig. 215), which is fused to the anteriormost ridge and opposes the base of the TP (fig. 219). The posterior median process (PMP) seems to be a rigid structure that includes the three anterior ridges and their attachments (fig. 218).

A digested abdomen was subjected to a sagittal section (fig. 226) and scans were made of both the left (figs. 220, 222, 224, 225) and right (figs. 221, 223) sides. One obvious observation is that the integrity of the PMP is evident in the images. The connections between the three anterior ridges and their attachment to PA and GAp clearly show that this is one rigid unit (fig. 225). Sectioned views also show that the PMP is attached to the postepigastric scutum by much thinner tissue that looks flexible (figs. 221, 225). Also revealed are the transverse pockets associated with the openings at the gonopore. The one associated with the anterior opening (AO) is indicated by the torn membrane ventrad of the PA (figs. 222, 223). The posterior opening (PO) goes into a pocket beneath the GAp and the field of papillae (figs. 220, 222). The dorsal wall of that pocket is sclerotized and formed by the PMP, whereas the posterior and ventral walls are membranous and connect the PMP to the posterior rim of the gonopore (fig. 230).

This posterior pocket is interesting because, except for size, it closely resembles the receptaculum described for Silhouettella loricatula (Roewer) by Burger et al. (2006). In their figure 7b, the posterior opening of the gonopore also enters a cavity with a membranous ventral wall that connects to a sclerotized structure which bears the papillae and GAp. Given the similarity in structure, the posterior pocket of M. molotra is the probable homolog (vestige) of the receptaculum.

Figure 227 shows the gonopore with its margins viewed on end. The main (median) opening is indicated by the membranous remnants of the uterus, and the anterior and posterior openings are shown by the upper and lower arrows, respectively. The sectioned view shows that these openings (arrows in fig. 228) lead into pockets. A reconstruction of the genitalia (fig. 230) gives the anterior component in green and posterior in brown, with sclerotized sections darkly pigmented.

The slitlike openings at the female gonopore combined with the broad flat dorsal lobes of the male palp suggest that the two structures couple. Without sliding too far down the slippery slope of speculation, this fascinating morphology makes it hard to avoid some discussion. First, as this species does not have external epigynal openings, insemination needs to take place within the gonopore. This means that the embolus and embolar tube need to penetrate, as does the closely connected dorsal lobe. Second, as the posterior pocket is homologous with the typical receptaculum, it seems safe to assume that insemination takes place here (as opposed to the anterior pocket, which seems to be a blind cavity on the opposite side of the gonopore).

Figure 229 shows a male M. molotra in oblique apical view holding the left palp outward. The image is at the same magnification as the female genitalia (fig. 230) and orientated so as to maximize the fit between the posterior gonopore margin of the female and the groove formed by the dorsal lobe of the male. A comparison of the embolar region to the female genitalia shows that the shape and curvature of the dorsal lobe would fit the posterior opening (fig. 230). The placement of the ventral lobe is ambiguous, as it may or may not need to enter the gonopore. If it does enter the gonopore, the obvious possibility is the anterior pocket. As this lobe is capable of expansion, doing so while inserted may function as a locking mechanism. The joining of the genitalia as described requires that the spider pair is positioned venter to venter and facing in the same direction. This is the same copulatory position described for S. loricatula ( Burger and Carrera, 2011: figs. 2–4), except that its male orients the dorsal surface of the bulb against the female’s abdomen, whereas the morphology of M. molotra requires the prolateral surface.

RELATIONSHIPS The relationship of Molotra to other oonopids is not clear. Our search through the literature and the images available on the PBI site has not turned up any obvious sister group of the genus. The unusual genitalia of Molotra , which clearly appear to be derived, suggest a relation among species with similar modifications: a broadly lobed embolar region in the male and a reduced receptaculum in the female. One possibility is Xyphinus Simon , where the male palp does have broad terminal lobes (see figures in Deeleman-Reinhold, 1987) and the female genitalia have recently been shown to lack a receptaculum ( Burger, 2010b). However, unpublished images of the Xyphinus palp (made available by Y. Kranz-Baltensperger) show many differences from Molotra , such as a larger number and more complex arrangement of the terminal lobes and an enlarged tibia and patella with prolateral tubercles. The female genitalia of Xyphinus also differ from Molotra in having external modifications, such as cavities and scapelike structures, and a very different internal arrangement ( Burger, 2010b: fig. 1A, B). Finally, the many somatic differences of Xyphinus (for example, the absence of leg spines, the enlarged pedicel, and various modifications in carapace and abdomen shape and ornamentation) further suggest that it is not closely related to Molotra .

Another place to seek relatives of Molotra , is among somatically similar species. The species of the M. molotra group closely resemble some of the silhouettelloids found on Madagascar. The group “silhouettelloids” as used here refers to those genera that have genitalia similar to that described for Silhouettella loricatula (Roewer) by Burger et al. (2006); see also Álvarez- Padilla et al., in press. For comparison we present a typical silhouetelloid, Tolegnaro sagani Álvarez-Padilla et al. , in press (figs. 7, 9, 11). The female genitalia are recognized by a large oval receptaculum (fig. 7), which is missing in Molotra (fig. 8). Despite this difference, closer examination of Molotra female genitalia shows much similarity with the silhouettelloids. Both show a basic division of the genitalia into a strongly sclerotized anterior and posterior component, which together form a locking mechanism for the gonopore. Both also have:

(1) a distinct to robust central T-shaped process, TP (figs. 7, 8, 199, 203, 205.);

(2) large, broadly foliate posterior apodemes, PA (fig. 218);

(3) PA attachment via anterior and posterior transverse ridges (fig. 219);

(4) a field of papillae between the ridges (figs. 9, 10, 215); and

(5) papillae size and shape similar to that in some silhouettelloids (figs. 8, 11, 215);

If a relationship to the silhouettelloids is suggested by female morphology, what about the male? At first glance, the Molotra male has a very different embolar region (fig. 12) from that of the silhouettelloid, which has several terminal attenuations (fig. 11). However, there are still some potential homologies. Both have liplike prolateral surfaces on the embolar region, a basic division into dorsal and ventral components (lobes), an embolus attached to the dorsal lobe, a subapical embolar opening, and a ventral lobe basally creased (wrinkled) and presumably expansible.

Although the association of Molotra with the silhouettelloids seems reasonable, a specific sister-group relationship is not obvious. Also not certain is the polarity of the different genitalic states. Is the relative simplicity of the genitalia (presence of fewer parts) a primitive state or derived as a consequence of fusion and reduction? If the absence of the receptaculum is derived on the basis of a structural loss, then the Molotra palp may likewise be derived, and represent a fusion and reduction of lobes. If this trend is extrapolated within Molotra , the derived state would be an embolar region with the shortest and most fused lobes, as in M. milloti (fig. 6).

MAP 1. Map of Madagascar showing the distribution of Molotra species.

And conversely, the palp with the largest number of structures would be the most primitive state, as M. tsingy (fig. 5). This latter conclusion is difficult to accept given the apparently unique (autapomorphic) states in that species, such as the palpal rotation and loss of the sternal cavities. That the palp in M. tsingy may have a combination of plesiomorphies and apomorphies is another possibility.

BIOGEOGRAPHY

Given the rarity of Molotra in our samples, few conclusions may be drawn here. The 90 specimens examined represent less than 1% of the total Malagasy oonopids available for study. Only the type species is abundant, with the remaining five species represented by only 13 specimens, or about 0.1% of the collection. The relative abundance of M. molotra is curious, given that the other localities were sampled in a similar manner. Perhaps there is a seasonal factor here, as this species was the only one sampled in December, or an altitudinal one, as it comes from a slightly higher elevation (1300 m) than the others (<1100 m).

The species distributions are plotted on map 1. The three groups are geographically disjunct and occupy different biotic regions. The four species of the M. molotra group are allopatric to parapatric and confined to a small portion of northeastern Madagascar, in wet montane rainforests from ~ 1000–1300 m elevation.

M. tsingy is from the midwestern part of the island, a tropical dry forest in karst terrain (Tsingy). The two known males resemble those in the M. molotra group, except for size-related differences. In addition to their smaller size, these males also have paler pigmentation, shorter leg spines and reduced tuberculation, including the absence of the carapace knob and scutopedicel ridges (table 1). The reduced size and form of M. tsingy are most likely adaptations to life in their drier habitat. A similar pattern is known from another Malagasy oonopid, Malagiella , which also includes a small pale species from the more xeric western part of the country ( Ubick and Griswold, 2011).

M. milloti is known from a single male specimen collected by Millot on the northwest of the island, at the coastal resort area of Ankify. A single collection from a coastal locality is suspicious and suggests the possibility that this species may be introduced. This is supported by its unusual somatic morphology, which is very different from both Molotra and other known Malagasy oonopids (table 1).

Figures

FIGS. 227–230. Molotra molotra, new species, male and female genitalia. (PBI_01990). Female (227, 228, 230), male (229, 230). 227. Composite image of gonopore with anterior and posterior sections separated and rotated. Anterior section (upper image), posterior view, arrow to anterior opening. Posterior section, anterior view, arrow to posterior opening. 228. Left sagittal section, laterodorsal view, with arrows showing anterior and posterior openings. 229. Left palpus, attached to body, in anteriolateral view. 230. Reconstructions of female and male genitalia in hypothesized precopulatory positions. Anterior median process shown in green, posterior process in brown, cuticle section in gray. AA = anterior apodeme, DL = dorsal lobe, E = embolus, GAp = globular appendix (median knob), PA = posterior apodeme, PF = papillae field, TP = T-shaped process, UI = uterus internus, VL = ventral lobe. Scale bars: 50 µm.