Seven new species of Amynthas (Clitellata: Megascolecidae) and new earthworm records from Taiwan

Earthworm specimens collected in southern Taiwan consisted of seven new species of Amynthas and several previously known species, mostly widely distributed peregrines. The new species are A. nanrenensis of the octothecal A. corticis species group, A. monsoonus and A. huangi of the sexthecal A. aelianus species group, and four proandric octothecal species: A. chaishanensis, A. hengchunensis, A. kaopingensis and A. ailiaoensis. Amynthas chaishanensis has dorsal intrasegmental spermathecal pores, but the other three proandric species have dorsal, lateral or ventral intersegmental spermathecal pores, respectively. The proandric species are united by several features, including the enclosure of segment xi in a sac, as in A. formosae (also proandric), octothecal with spermathecae in vi–ix, spermathecal diverticula stalks generally kinked and often enclosed in membrane, and prostatic ducts divided polytomously into numerous small ductlets, which may be grouped into bundles of two to five. In A. ailiaoensis the prostatic duct trunk contains up to seven separate lumens in the ental half, surrounded by the circular muscle of the duct, while in A. chaishanensis the prostatic duct trunk contains about 40 small lumens. Of the previously known species in the collection, Pontodrilus litoralis and Metaphire houlleti are first reported from Taiwan. Additional locations for A. incongruus and A. robustus are given, and in the latter case the material appears to be the usual male‐sterile morph. Other species found are Pontoscolex corethrurus, Amynthas corticis, A. gracilis, Metaphire californica, and Polypheretima elongata.


Introduction
The earthworm fauna of Taiwan has received considerable attention in recent years, beginning with Shih et al.'s (1999) review of the species known from the island and continuing with several publications by a group at the Taiwan Endemic Species Research Institute (Tsai et al. 1999(Tsai et al. , 2000(Tsai et al. , 2001Shen et al. 2002Shen et al. , 2003aShen et al. , 2003b, in which 13 new species of Amynthas and Metaphire were described. While working on the literature review, H.-T. Shih made several collections over a period of 5 years with the members of the laboratory of H.-W. Chang of the National Sun Yat-sen University in Kaohsiung, Taiwan. In contrast to most of the previous collections, which were from the northern part of Taiwan, these were made in the south. Seven species in these collection lots were determined to be new by reference to published descriptions of Taiwan and other East Asian Amynthas.

Materials and methods
From 1996 to 2000, the earthworm fauna of Kaohsiung City, Kaohsiung County and Pingtung County (Figure 1) was surveyed. Collections were made by digging and hand sorting soils and litter, by collecting earthworms on the soil surface where available, and by permanent traps containing formalin. GPS readings were taken at many of the collection localities. Specimens were preserved by killing in alcohol and fixing in 10% formaldehyde, after which they were transferred to 70% ethanol. Internal anatomy was examined by dorsal dissection, and drawings prepared with a drawing tube mounted on a stereomicroscope. All materials are deposited at the National Museum of Natural Science, Taichung, Taiwan (NMNS). By convention, we use lower case Roman numerals to refer to segment numbers and Arabic numerals separated by diagonal lines to indicate segmental boundaries both external and internal. Setae are labeled with upper-case letters with the ventral-most of a side as A and the dorsal-most as Z, regardless of the actual number of setae in a segment.
We use the definition of the Megascolecidae offered by Jamieson et al. (2002), which is supported by a molecular analysis. It is identical to that of Blakemore (2000), the two systems differing greatly regarding definitions of the Acanthodrilidae, and the nonrecognition by Jamieson et al. (2002) of the Octochaetidae and Exxidae.

Etymology
This species is named after Nanrenshan, which means Mt. Nanren.
Male sexual system holandric, testes, funnels in ventrally joined sacs in x, xi. Seminal vesicles large in xi, xii, with dorsal lobe. Prostates xviii, two or three main lobes, ducts thick, muscular, join vasa deferentia distal to glandular portion; vasa deferentia non-muscular; genital marking glands lacking.

Remarks
Amynthas nanrenensis keys to the ''diffringens'' species group (which should be known as the corticis species group, as diffringens is a junior synonym) in Sims and Easton (1972). Blakemore (2003) provides a detailed synonymy for A. corticis (Kinberg, 1867), including many names indicated as questionable or uncertain in the synonymy, and a diagnosis of the species. The diagnosis is brief, based on the location, number and spacing of spermathecal pores (0.3 body circumference apart), the variable presence of paired or variable small genital markings near the spermathecal and male pores, and the simple or incised condition of the intestinal caeca. The only character on which the present material differs from Blakemore's (2003) diagnosis is the narrower spacing of the spermathecal pores. However, in view of the other differences between A. nanrenensis and Pheretima diffringens (Baird, 1869) (5A. corticis) as described in detail in Gates (1972) it is not the widespread peregrine ''species'' composed entirely of a set of parthenogenetic morphs, for which a hermaphroditic sexual population is as yet unknown (Gates 1972). Gates (1959) reported two specimens from Taiwan (American Museum of Natural History 3575) in which male function was present, and he remarked that these resemble a hypothetical hermaphroditic ancestor of A. corticis. These specimens share with A. nanrenensis the larger number of setae per segment than A. corticis, but differ in having spermathecal pores near mid-lateral, different arrangement of genital markings in the male field, presence of genital markings in the spermathecal segments, and hearts in xi. Male function is clearly present in A. nanrenensis, with iridescent male funnels and iridescent spermathecal diverticulum chambers, the latter indicating that mating has taken place and sperm has been received. Spermathecal segment genital markings are lacking in A. nanrenensis, unlike A. corticis which generally has them. Genital marking glands are absent, another separation from A. corticis, and cannot, as has been done in other cases, be ascribed to parthenogenetic degradation of sexual characters.
The complete absence of hearts in x and xi is unusual, and not found in Gates' (1972) material, where hearts of x are ''usually aborted'' but hearts are uniformly present in xi. The missing hearts might be taken as a developmental abnormality, but the next species also shows deletion of some hearts. Amynthas nanrenensis differs from all other known Amynthas with spermathecae in vi-ix and ventrally placed spermathecal pores with respect to the lack of hearts in x and xi, plus the arrangement of genital markings in the male field. Absence of hearts in x and xi distinguishes it from most other Amynthas, the nearest in this respect being the similarly octothecal Vietnamese A. primadamae (Michaelsen, 1934), in which the contractile hearts are also in xii and xiii only, with a pair of small lateral hearts enclosed in the testes sacs in xi. Amynthas nanrenensis has oesophageal hearts in xi, has more narrowly placed spermathecal and male pores, and has a very differently shaped spermathecal diverticulum. It may be necessary to place A. primadamae in the synonymy of A. corticis based on its spermathecal battery and pore spacing, if Blakemore's diagnosis of the latter is sufficient, and modification of hearts is not considered important.

Etymology
The species is named for the tropical monsoon forest in Nanrenshan which is unusual for such a northern latitude.

Remarks
In Sims and Easton (1972) this worm keys to the sieboldi group. Easton (1981) transferred A. sieboldi (Horst, 1883 to Metaphire, so the species group name may no longer be appropriate. On the other hand, the male pores of the type of M. sieboldi (National Natuurhistorisch Museum, Leiden, Netherlands, cat. no. 1825) are subapical on blunt cones surrounded by elevated circular lips leaving wide openings. Through this opening the cone is clearly visible. The circular trough surrounding the cone is wholly confined to the body wall, which shows no trace of bulging into the coelom. This leaves one in grave doubt about the validity of assigning A. sieboldi to Metaphire. Gates (1975, p. 7) wrote, ''Presence or absence of copulatory chambers is too vague. The really important character is whether the male pores are superficial or invaginate. In the latter case, whether in slight transverse slits or much deeper spaces still confined to the parietes or whether thick-walled copulatory chambers deeply penetrating into coelomic cavity (cf Gates 1972, p. 150)''. In Easton (1981) no details are given in support of the transfer, and he further commented on M. riukiuensis that it was uncertain if the male pores were in seminal grooves or copulatory pouches, so it was unclear if it should be placed in Amynthas or Metaphire. If the definition of ''copulatory pouch'' is so vague, then a critical review of the character and assignments of species to genera based on the character are clearly needed. We support following the suggestion of Gates (1975) to better characterize the status of various types of non-superficial male pores. For now we support restricting Metaphire to those species distinguishable from Pheretima only by the absence of nephridia from the spermathecal ducts (Sims and Easton 1972). This would require the presence of well-developed copulatory pouches protruding into the coelom (as in Pheretima), but leaves unclear what to do with species whose copulatory pouches are entirely intramural and could thus be distinguished from Pheretima.
Pending the outcome of these issues, the species group could be renamed the aelianus group after A. aelianus (Rosa, 1892), that being the first in the species group list in Sims and Easton (1972). This group should also include six recently described species from Taiwan with spermathecal pores in 6/7/8/9: A. binoculatus Tsai, Shen and Tsai, 1999, A. fenestrus Shen, Tsai and Tsai, 2003, A. sexpectatus Tsai, Shen and Tsai, 1999, A. tayalis Tsai, Shen and Tsai, 1999, A. tenuis Shen, Tsai and Tsai, 2003, and A. tungpuensis Tsai, Shen and Tsai, 1999. Amynthas monsoonus differs from them all in lacking the anterior two pairs of hearts, like A. nanrenensis. It also has a different genital marking pattern from its Taiwanese sexthecal congeners. The missing hearts suggest that it is more closely related to A. nanrenensis than to the aelianus group members. No one has tested the hypothesis that the spermathecal battery is evolutionarily more conservative than details of the circulatory system, and there is evidence to the contrary. The locations of hearts are widely conserved among Amynthas, across great variation in other characters, particularly the numbers and locations of spermathecae. Amynthas monsoonus and the previous two species are quite unusual in having lost the hearts of x, or x and xi. These three are very similar with respect to other somatic characters and spermathecal morphology. Therefore it seems possible that their similarity is due to common ancestry, and that some species groups defined by spermathecal batteries could be polyphyletic or paraphyletic (addition or deletion of a pair would remove a taxon from its clade, rendering the latter paraphyletic). Advocates of sexual characters as indicators of phylogenetic relationships include most of the classical authors (prominent among them Michaelsen and Stephenson). Promotion of somatic characters for this purpose is one of the central themes of Gates' work (Gates 1972), but the question will not be decided without recourse to a third and independent set of characters. Nucleic acid sequence data are an obvious choice.
A suggestion that A. monsoonus is very similar to A. carnosus (Goto and Hatai, 1899) discounts the more anterior location of the three or four pairs of spermathecae in the latter, as well as its possession of genital markings in segments xviii and xix, greater numbers of setae per segment, lack of genital marking glands, and the very different spermathecal morphology. Blakemore's (2003) diagnosis of A. carnosus and subsequent remarks all place its first pair of spermathecal pores in 5/6. Amynthas huangi sp. nov.

Etymology
This species is named after Mr. Chung-Chi Huang who helped the collection work extensively.
Male sexual system holandric, testes, funnels in ventral paired sacs in x, xi. Seminal vesicles small in xi, xii, without dorsal lobe. Prostates large xviii, deeply lobed; ducts thick, muscular, short; vasa deferentia join duct at duct-glandular portion junction; vasa deferentia non-muscular; prostatic duct flanked by large sessile glandular masses on body wall.
Ovaries in xiii. Paired spermathecae in vii-ix; ampulla ovoid, large; diverticulum large flat ovate mass composed of tightly folded tubular chamber, short slender straight stalk ( Figure 2F); no nephridia on spermathecal ducts; genital marking glands with long stalks meeting body wall in vi-viii next to spermathecal ducts.

Remarks
The male pores are clearly not within intra-coelomic copulatory pouches, such as characterize Pheretima s.s. and perhaps Metaphire. In the present case, and in some species described below, the male pores are within slight folds of the body wall. In the absence of additional evidence supporting transfer to Metaphire, we assign this species to Amynthas. Sims and Easton (1972) stated that in the absence of spermathecae, it is not possible to distinguish a Pheretima from a Metaphire. In light of the fact that Pheretima all have intracoelomic copulatory pouches appearing as domes of tissue (usually muscular in appearance) partially separable from the body wall, this must also be a characteristic of Metaphire, or Sims and Easton (1972) were wrong. We are open to both possibilities, but to date no one has adequately addressed this question. In our experience, there exist species with large intramural copulatory pouches within a thickened body wall of xviii, and these consistently fall in Metaphire. Such structures appear to us not homologous to the intracoelomic pouches of Pheretima, but we could be mistaken. Based on this we prefer to restrict Metaphire to those species with well-characterized copulatory pouches and no nephridia on the spermathecal ducts (Sims and Easton 1972), excluding those whose pores lie within wrinkles or seminal grooves, under small flaps, or within shallow indentations.
Amynthas huangi belongs to the aelianus species group, in which it is most similar to A. taipeiensis (Tsai, 1964). However, the differences are many: smaller size than A. taipeiensis, fewer setae, no setal enlargement ventro-anteriorly, male pore area different, hood or flap over male pores present, colour different, intestinal origin in xvi not xv, seminal vesicles lack dorsal lobes, prostatic ducts short and straight, not coiled or bent, diverticulum chamber coiled with straight stalk versus stalk kinked in A. taipeiensis, and no genital marking glands in A. taipeiensis. Note that genital marking glands are present in xviii even though no externally visible genital markings are present. This suggests that the genital markings are hidden under the hoods partially obscuring the male pores. Furthermore, the spermathecal segment genital markings must be deep in the pore slits or even within the pores themselves, out of view.

Etymology
This species is named after Hengchun Peninsula, Pingtung County, Taiwan, where it was discovered.
Male sexual system proandric, testes, funnels in ventrally joined sac in x. Seminal vesicles large in xi, with small dorsal lobe; seminal vesicles, other contents of xi enclosed in thin sac. Prostates large in xviii, four main lobes, each lobe served by two to five small ductlets radiating fan-like from ental end of prostatic duct; ducts stout, straight, muscular, narrowing towards body wall; vasa deferentia join duct at duct-glandular portion junction; vasa deferentia non-muscular.

Remarks
Amynthas hengchunensis is similar to A. formosae (Michaelsen, 1922) with respect to proandry, the enclosure of the contents of xi in a sac, being octothecal in vi-ix, the general form of the spermathecae, the body size, the intestinal origin, and the very large number of setae in the anterior segments. In contrast, A. hengchunensis has many more setae, especially in the post-clitellate segments, and has spermathecal pores more ventrally placed. Other differences (A. hengchunensis features given) include the male field possessing a flap partially covering the male porophore, much shorter caeca, and lack of membranous covering of the diverticulum stalk. This species is the second proandric Amynthas known from Taiwan, with three more described below. Amynthas hengchunensis also has a different prostatic duct structure from that encountered in most Amynthas, the only one similar to it being A. formosae. The ordinary Amynthas prostatic duct has a multiple approximately dichotomous branching pattern, with occasional trichotomies. In the present species, and the next three described below, numerous ductlets of equal size join the large prostatic ducts at the same point, creating a single polytomy. This polytomy may form from two or more groups of ductlets, or as in A. chaishanensis, which is described later in this paper, one undivided fan of ductlets.

Description
Slight dorsal-anterior dark pigmentation, male field with conical porophores composed of concentric rings, innermost ring surrounding semi-circular protrusion from within slight indentation; this protrusion placed laterally to a smaller round protrusion of lower elevation; male pores not seen but probably on the smaller protrusion or between the two protrusions. Thirty setal follicles between porophores, 27 present in these; porophore apices 8 mm apart, or 0.25 body circumference. Spermathecal pores intrasegmental at equators of presetal annulus in each of segments vi-ix, subdorsal, 0.06-0.08 circumference apart from dorsal side.
Prostates composed of four or five main lobes, each connected by numerous very small ductlets to large muscular prostatic duct; prostatic ducts consisting of thick circular muscle layer surrounding spongy tissue composed of numerous very small tubules.

Remarks
The single type specimen examined is missing the gut, hearts and male organs from ix-xvii. Additional type material formerly in the Zoologisches Institut Hamburg (cat. no. 9309) was discovered to be missing when requested for examination, the bottle and label being present but no material within. Thus little more than the above could be learned regarding internal anatomy. It is now clear that the spermathecal pores are intrasegmental and very close to the mid-dorsal line, in contrast to representations made in Gates (1959) for other material. The prostatic duct structure of the type is clearly generally similar to A. hengchunensis and the other proandric species described below. Gates' (1959) material may belong to a separate species, because the spermathecal pores are given as ''well towards mL'', rather than nearly mid-dorsal dorsal and intrasegmental as in A. formosae. Amynthas yuhsi as described by Tsai (1964) is indistinguishable from the remains of the type of A. formosae, so it is a junior synonym.

Etymology
This species is named after the combination of prefixes, ''kaoping'', of its localities, Kaohsiung and Pingtung Counties.
Male pores minute at posterior end of seminal grooves extending from centre of ovate to rounded angular genital pad longitudinally orientated from 17/18 to equator of xviii, surrounded by epidermal folds, lateral folds closest to male pores enlarged to form flap adjacent to or partially covering genital pad ( Figure 3A); one specimen with paired oval genital markings presetal xvii slightly median to male pore line; 32-40 setae between male pores, pores 0.32 circumference apart on setal line 22. Spermathecal pores dorsal in 5/6/7/ 8/9, 0.29-0.32 circumference apart dorsally.
Male sexual system proandric, testes, funnels in ventrally joined sac in x. Seminal vesicles large in xi, with small fine-textured dorsal lobe; seminal vesicles, other contents of xi enclosed in thin sac. Prostates large in xviii, three to five main lobes, each lobe served by two to five small ductlets radiating fan-like from ental end of prostatic duct; ducts stout, straight, muscular, narrowing towards body wall; vasa deferentia join duct at ductglandular portion junction; vasa deferentia non-muscular.

Remarks
Another octothecal proandric species with dorsal intersegmental spermathecal pores, A. kaopingensis is closest to A. hengchunensis in all respects, including details of internal anatomy such as the structure of the ductlets of the prostates, the membrane enclosing segment xi, and the presence of septum 8/9. It differs from A. hengchunensis in the features of the male field, a more dorsal placement of spermathecal pores, and the structure of the spermathecal diverticulum. There seems to be considerable morphological unity among A. kaopingensis, A. hengchunensis and A. formosae.
The male field of A. kaopingensis has seminal grooves, a feature not previously reported in the literature of Taiwan earthworms, but well known among Korean Amynthas (Kobayashi 1936;Hong and James 2001;) and species from other parts of Asia: A. glabrus (Gates, 1932), A. japonicus (Horst, 1883), A. papilio (Gates, 1930), A. plantoporophoratus (Thai, 1984), and A. riukiuensis (Ohfuchi, 1957). The locations of these species include Myanmar, Japan, Vietnam, and the Ryukyu Islands. The grooves are commonly formed within an otherwise flat genital pad of varying shape, but in A. riukiuensis they are formed by folds on the male field. It is not clear if all are descended from a common ancestral type, or if there could be two or more independent evolutions of seminal grooves in Amynthas. In any case, none of the species with seminal grooves also has copulatory pouches, and therefore these do not belong to Metaphire.

Etymology
This species is named after the Ailiao River, of the locality Wutai.
Male sexual system proandric, testes, funnels in ventrally joined sac in x. Seminal vesicles large in xi, with large dorsal lobes; seminal vesicles, other contents of xi sometimes enclosed in thin sac. Prostates large in xviii, seven main lobes, each lobe served by two to three small ductlets radiating fan-like from ental end of prostatic duct ( Figure 3D); ducts stout, muscular, narrowing towards body wall; vasa deferentia join duct at duct-glandular portion junction; vasa deferentia non-muscular.

Remarks
The fourth octothecal proandric Amynthas of Taiwan is much more similar to Gates' (1959) material from Chao-Chow (5Chaojhou), Pingtung and Green Mountain (5Yangmingshan), Taipei than the others. Differences from Gates' material are few. It is possible that these are the same species. A. ailiaoensis has paired genital markings within a shallow invagination of the male field, lacks septum 8/9, has much shorter caeca, and the spermathecal pores are at or above mid-lateral rather than ''well towards mL'' which might mean below mid-lateral but close. Many small nematodes were found in the body cavity, mainly around the caeca, prostates and seminal vesicles.

Etymology
This species is named after the locality Chaishan (5Mt. Chai), Kaohsiung City, Taiwan.
Male sexual system proandric, testes, funnels in ventrally joined sacs in x. Seminal vesicles large in xi, with dorsal lobe, enclosed within sac containing all segmental contents. Prostates xviii, numerous deeply incised main lobes covering xvii-xx, ducts short, thick, muscular, straight; numerous ductlets from glandular portion form undivided fan-shaped array, ental portion of prostatic duct with approximately 40 very small lumens, one larger lumen probably sperm duct; vasa deferentia join at duct-glandular portion junction; vasa deferentia non-muscular.
Ovaries in xiii. Paired spermathecae in vi-ix; ampulla large ovate sac, duct stout but flaccid, half length of ampulla, diverticulum stalk long convoluted kinks enclosed within membrane, chamber terminal ovate knob; diverticulum axis shorter than ampulla axis ( Figure 3G); no nephridia on spermathecal ducts.

Remarks
This proandric octothecal species with intrasegmental spermathecal pores keys to the rimosus-group in Sims and Easton (1972), all members of which are from Myanmar. Amynthas rimosus (Gates, 1931) is holandric and its spermathecal pores are ventral. The true affinities of A. chaishanensis lie with the other Taiwanese proandric species, and particularly with A. formosae. Amynthas chaishanensis has numerous prostatic ductlets of equal size joining the large prostatic ducts at the same point, creating a single undivided fan of ductlets. This is in contrast to the other species described here and A. formosae, all of whose ductlets gather into small bundles prior to joining the main prostatic duct.
Amynthas chaishanensis is one of two known proandric Amynthas species with dorsal intrasegmental pores on vi-ix, the other being A. formosae. It is further distinguished by having hearts only in xii and xiii but A. formosae's hearts are unknown, beyond having the last hearts in xiii. However, Gates (1959) noted the presence of hearts in x-xiii in his material and did not note this as a distinction between his material and A. formosae. Other differences from A. formosae are many more setae in the anterior segments, fewer setae between the male pores, slightly more widely spaced (from mid-dorsal line) spermathecal pores, caeca originating in xxvii, lack of pseudovesicles in xii and xiii, and prostate glands divided into numerous main lobes, rather than only 2.

Remarks
The material is very probably A. corticis, a widespread peregrine species or complex of asexual morphs. For a detailed proposed synonymy see Blakemore (2003).
Male sexual system holandric, testes, funnels in paired ventral sacs in x, ventrally joined sac in xi. Seminal vesicles acinous, large in xi, xii, with small uniform dorsal lobe. Prostates small xviii, three main lobes, ducts straight, muscular, vasa deferentia join duct at ductglandular portion junction; vasa deferentia non-muscular; stalked genital marking glands xviii.
Ovaries in xiii. Paired spermathecae in vi-viii; ampulla pear-shaped, duct shorter than ampulla, diverticulum chamber small ovate, stalk slender with one or two short kinks, ental third of stalk and chamber iridescent with sperm; no nephridia on spermathecal ducts.

Remarks
This species keys to the peregrine A. gracilis (Kinberg, 1867). The only other sexthecal Taiwan Amynthas known to date to have the spermathecae located in vi-viii is A. wangi (Shen et al. 2003b). Differences between A. gracilis and A. wangi are primarily in the male reproductive organs and the locations of genital markings. Genital markings are present in the spermathecal segments in A. wangi, but not in this material of A. gracilis, and in A. wangi on xvii in line with the male porophores, rather than in xviii medial to the male pores. The testes sacs of A. wangi are paired, while on the other hand the testes sacs in x of A. gracilis are joined ventrally.

Remarks
This worm's morphology is consistent with the original description, and information given in Gates (1959) for other A. incongruus material collected on Taiwan.

Description
Unpigmented, spermathecal pores lateral, male pores on porophores with one genital marking on each, large conical mid-ventral marking on xviii. GMs in vii, ix presetal just median to spermathecal pores. Genital marking glands mushroom-shaped with thick stalks, one or two at each spermathecal duct, one for mid-ventral GM in xviii, two at each prostatic duct.

Remarks
This is probably the male-sterile A. robustus, as most particulars agree with previous data on the species. The only exception is the mid-ventral genital marking on xviii, a condition never seen in material examined by Gates (1972).

Remarks
This is a new record for the Taiwanese earthworm fauna.

Discussion
There is considerable work to be done refining our understanding of relationships within Amynthas. If nothing else can be concluded from the new species reported here, it is clear that the species groups relied on for the last 30 years are mainly matters of convenience for purposes of constructing a key to the genus. Gates (1972) emphasized the need for basing classifications on somatic characters as well as sexual, but here we have presented some examples where the two conflict. Do we prefer to give emphasis (implicit weighting) to spermathecal battery over heart configuration-which would place A. nanrenensis in the corticis species group, excluding A. monsoonus with its similar reduction in number of hearts-or unite these three irrespective of their differing spermathecal batteries? Similarly, do we weight andry and prostatic duct structure in preference to hearts, placing the proandric octothecal species reported here together, and consider loss of hearts in A. chaishanensis to be homoplasious with respect to A. nanrenensis and A. monsoonus? On the simple basis of number of shared derived character states, we would favour weight given to andry and prostatic ducts in the latter example.
Two newly discovered Amynthas from the Philippines (Y. Hong and S. W. James, unpublished data) are proandric but do not have the same prostatic duct structure as the Taiwan proandric species, and are quadrithecal. The Philippine species have a different and unique structure of the prostatic duct, which is modified to form a sheath over a small penis visible through the outer secondary male pores. Thus when one enlarges the scope of investigation into relationships within the genus, characters that served well locally are not as reliable globally. The massive task of revising Amynthas will be made easier if all future species descriptions are as complete as possible and the describers alert to details in structures too often given brief accounts or no details at all.