On two cases of male dimorphism in dwarf spiders (Araneae: Linyphiidae)

Abstract. Diplocephalus cristatus (Blackwall, 1833) is confirmed as a dimorphic species, having two morphs: cristatus and foraminifer. This view was first proposed by Georgescu (1969), but not supported in the literature. Diplocephalus foraminifer (O. Pickard-Cambridge, 1875), D. bicephalus (Simon, 1884), D. rectilobus (Simon, 1884), D. foraminifer thyrsiger (Simon, 1884) and D. arvernus (Denis, 1948) are here considered junior synonyms of D. cristatus (Blackwall, 1833). Diplocephalus bicephalus belongs to the morph cristatus, D. rectilobus, D. thyrsiger and D. arvernus to the morph foraminifer. A lectotype (♂) is designated for Diplocephalus bicephalus Simon, 1884; the paratype female of D. bicephalus was incorrectly identified and actually belongs to Dicymbium nigrum (Blackwall, 1834). Savignia harmsi Wunderlich, 1980 is another dimorphic species, with the two strongly differing male morphs: typica and cor.

For a long time, dimorphic erigonid spiders were not recognised as such and considered separate species. Only when morph differences were small, for example in the size of the cephalic tubercle or the post-ocular sulci, were these sometimes considered variations. Holm (1979: p. 269) wrote about Pelecopsis mengei (Simon, 1884): "The males occur in two different forms, the one, which is the most frequent, with a high cephalic lobe and large sulcal orifice, the other with lower lobe and with much smaller orifice. … As no intermediate forms have been found and moreover, the two types of males have quite similar palpal tibiae and bulbs and are found together, the males of P. mengei seem to be dimorphic". Similarly, Bosmans & Abrous (1992) considered the specimens of Pelecopsis oranensis (Simon, 1884) with small and large postocular sulci, but having identical palpal tibiae and bulbs, as morphs of the same species. Diplocephalus marijae Bosmans, 2010 from Spain is another species occurring in two morphs (Bosmans et al. 2010). The decision to recognise species as being dimorphic is not easy or consequent. Roberts (1987) proposed Troxochrus scabriculus (Westring, 1851) and T. cirrifrons (O. Pickard-Cambridge, 1871) to be one, dimorphic species, but this is not followed in the World Spider Catalog (2018) where they are still considered two separate species. On the contrary, when the same author (Roberts 1987) proposed Diplocepha lus connatus Bertkau, 1889 andD. jacksoni (O. Pickard-Cambridge, 1904) to be forms of the same species, this opinion was accepted in the World Spider Catalog (2018).
The best documented case of dimorph linyphiid spiders is that of Oedothorax gibbosus (Blackwall, 1841) and O. tuberosus (Blackwall, 1841) having very different cephalic tubercles and because of that considered separate species in older identifi-cation books (e.g., Locket & Millidge 1953, Wiehle 1960, Palmgren 1976. After a detailed study of the male palps of several Oedothorax species, Bosmans (1985) concluded that all palpal sclerites of O. gibbosus and O. tuberosus were completely identical and the two species names were synonymized. De Keer & Maelfait (1988) provided further evidence from breeding experiments. They reared spiderlings from the same egg sac and obtained both forms. In our opinion, when palpal sclerites are identical, specimens should be treated as belonging to the same species.
The aim of the present paper is to discuss two more cases of male dimorphism in Linyphiidae.

Material and methods
The material studied was collected by the authors or loaned from museum collections. Species were examined by mean of a Nikon SMZ1270 stereo microscope. Details of male palps and female epigynes were studied with an Olympus CH-2 microscope with a drawing tube. Left palps are illustrated.
Male palps were detached and transferred to glycerol for examination under the microscope. Female epigynes were excised using sharpened needles. These were then transferred to clove oil for examination under the microscope. Later, palps and epigynes were returned to 70% ethanol.

A forgotten case of dimorphism
Males and females of Diplocehalus cristatus (Blackwall, 1833) were first described by Blackwall (1833) from England as Walckenaeria c. Subsequently, O. Pickard-Cambridge (1875) described Erigone foraminifera Pickard-Cambridge, 1875 from France. Differences between the two species were based on differently shaped cephalic lobes. Pickard-Cambridge (1875: p. 208) stated that "E. foraminifera is also allied to E. cristata, but the very different form of the caput and its cleft … will distinguish it at once".
Later, Simon (1884Simon ( , 1926 also gave considerable importance to the shape of cephalic lobes in describing several (sub) species in the genus Prosoponcus: P. bicephalus Simon, 1884, P. rectilobus Simon, 1884 an P. thyrsiger Simon, 1884. None of these species was matched to a female, except for P. bicephalus bicephalus; yet the female of the latter species appeared to be that of Dicymbium nigrum (Blackwall, 1834). Denis (1948) added one more species to this species group, D. arvernus, from France, Auvergne. He stated that this species was closely related to D. foraminifer, but slightly differed in the shape of cephalic lobe. Georgescu (1969) was the first author to propose that D. cristatus, D. foraminifer, D. bicephalus, D. rectilobus and D. thyr siger all belong to the same species, occurring in two morphs: cristatus and foraminifer. She also included Diplocephalus cras silobus (Simon, 1884) in that list, but the conformation of the male palp of the latter species is completely different (cf., Millidge 1979, Pesarini 1996. The suggestion by Georgescu has not been followed and in the World Spider Catalog (2018), these names are currently listed as separate species. We have been able to re-examine the material of all these Diplocephalus species, including the types of D. bicephalus and D. rectilobus, and can confirm Georgescu's proposal.

Comments on the type material
The only material of Diplocephalus bicephalus (originally as Prosoponcus b.) that is available in MNHNP is the male,  which is therefore designated as the lectotype. The two accompanying females belong to Dicymbium nigrum (Blackwall, 1834), and Simon's figure 674 (Simon 1884) obviously shows the epigyne of this species. The only material of Prosoponcus rectilobus available in the MNHNP is the male holotype.

Comments on the synonymy
We first became interested in the Diplocephalus cristatus complex, while studying specimens collected by Antony Russell-Smith from Lefkada, Greece (Fig 2). A number of males and females were collected from near a spring and they are illustrated in Fig. 2. Having tried to identify these specimens, we found out that their palps and epigynes were completely similar to those of the common European species D. cristatus, but the males had very different cephalic lobes. Further research showed a clear match with D. foraminifer and D. arvernus, as illustrated by Deltshev (1985), Denis (1948) and Georgescu (1969) -compare above the section "A forgotten case of dimorphism".
Figs 1b-c, e-f, h-i, k-l and n, o show the male palps and palpal tibiae of respectively D. cristatus, D. rectilobus, D. fora minifer, D. bicephalus and D. arvernus. Detailed examinations of all palpal sclerites and palpal tibiae revealed no differences. Simon (1926: p. 495) already wrote about D. rectilobus in a footnote: "Peut-être une forme ou variété de D. cristatus". Thus, in our opinion, the males of D. cristatus occur in two morphs: viz., Figs 1a, d show the morph cristatus with a low cephalic lobe, and Figs 1g, j, m and 2a show the morph fora minifer with a high cephalic lobe.

Distribution and habitat
Specimens of Diplocephalus morph cristatus occur all over Europe (Nentwig et al. 2017). On the contrary, the morph foraminifer has a much smaller distribution: northern Spain, southern France, Switzerland and northern Italy in the western part of Europe, and Bosnia and Hercegovina, Macedonia, Montenegro, Bulgaria, Serbia and Romania in the eastern part (Nentwig et al. 2017).
Material examined. SPAIN: Granada: Baza, 5 )) 1 (, pitfalls in dry riverbed, 12.xi.1990, L. Zarcos coll. (CRB). Comments. At first glance, the five studied males appear to belong to different species, because their prosomas have very different shapes (Figs 4a-c, 5a-c). However, their palp conformation is identical (cf . Figs 3e and 3d). The first morph ( Fig.  4a-b) has a nose-like projection carrying the anterior median eyes, like in Savignia frontata Blackwall, 1833. Apparently, because of this resemblance Wunderlich (1980) placed the species in the genus Savignia. The second morph (Fig. 4c-d) has a completely different cephalic lobe in the form of a large, rounded lobe, heart-shaped in the anterior view (Fig. 4d). For this morph, the name cor (Latin for heart) is herein proposed. If this morph was found first, the species would probably have been described in Diplocephalus. Distribution. S. harmsi was described from both sexes from Spain, in the province of Malaga (Wunderlich 1980). It was recollected from the neighbouring province of Granada. It was not yet recorded since the original description (Morano et al. 2014).