Redescription of the Far Eastern brook lamprey Lethenteron reissneri (Dybowski, 1869) (Petromyzontidae)

Abstract Nonparasitic Lethenteron reissneri (Dybowski) is redescribed based on four syntypes (two adults and two ammocoetes) from the Onon and Ingoda rivers, Russia, and 15 topotypic specimens (seven metamorphosing ammocoetes and eight ammocoetes) from the Onon River system, Russia and Mongolia. Posterial teeth were not mentioned in the original description, but Berg (1931) stated that they were sometimes absent, which he later (Berg 1948) changed to usually absent, based on material (some of which we have re-identified as parasitic Lethenteron camtschaticum) from far outside of the type locality. The latter view has been widely accepted by subsequent authors. Unfortunately, the poor condition of the two adult syntypes did not permit verification of this character. However, a row of posterials was clearly visible in six of the seven topotypic metamorphosing ammocoetes and indicates their usual presence in the species. The first full description of the ammocoetes, including pigmentation, is provided. The present study restricts the distribution of Lethenteron reissneri to the Shilka and Songhua river systems within the Amur River basin, until a more geographically comprehensive study is undertaken. Additionally, in this study, feeding versus non-feeding at the adult stage, are considered to be valid taxonomic characters at the species level.


Introduction
The Northern Hemisphere lamprey genus Lethenteron was originally erected as a subgenus of Entosphenus by Creaser and Hubbs (1922) and elevated to generic rank by Jordan et al. (1930), without justification. Vladykov and Follett (1967) accepted this action and defined members of Lethenteron as possessing posterial teeth in a single curved row and with the outer laterals (exolaterals) absent. Renaud (2011) recognized one parasitic species, L. camtschaticum (Tilesius, 1811) (i.e., the stem species) and six nonparasitic species, L. alaskense Vladykov & Kott, 1978, L. appendix (DeKay, 1842, L. kessleri (Anikin, 1905), L. ninae Naseka, Tuniyev & Renaud, 2009, L. reissneri (Dybowski, 1869), and L. zanandreai (Vladykov, 1955) (i.e., the satellite species) in the genus, along with two additional undescribed nonparasitic species from Japan, and noted that the taxonomic limits of the nonparasitic L. reissneri were unclear. Authors have either treated these satellites of L. camtschaticum as distinct species (e.g., Potter et al. 2015) or as synonyms of this species. For example, Artamonova et al. (2011aArtamonova et al. ( , 2011b and Makhrov et al. (2013) have, respectively, suggested that L. ninae, L. reissneri, and L. kessleri are synonyms of L. camtschaticum and that mode of life (i.e., parasitism versus nonparasitism) is not a valid criterion for specific distinctiveness. This is not a new concept. McPhail and Lindsey (1970), in discussing L. camtschaticum (reported as Lampetra japonica (von Martens, 1868)) from the Northwest Territories, Yukon and Alaska, suggested that it may not be specifically distinct from L. appendix (reported as L. lamottenii (Lesueur, 1827)) from eastern North America, but preferred to retain them as distinct pending a review of all relevant species from both continents (i.e., North America and Eurasia). Renaud et al. (2009) proposed that the key to resolving the issue of whether parasitism or nonparasitism constitute valid taxonomic characters at the species level may be to conduct common garden experiments (i.e., rearing two putative species under the same conditions from the zygote to the adult stage) that seek to elucidate the triggering mechanism for trophic interactions, or the lack thereof, in the adult stage. In the absence of such experiments, however, we follow here conventional taxonomy and continue to recognize the above nonparasitic species as distinct. Additionally, prior to synonymizing species, it is critical to compare any extant type material as well as the original descriptions, and, unfortunately, this has not been done on a consistent basis by authors. The original description by Dybowski (1869) of Petromyzon reissneri was very short, incomplete and based on material from the upper Amur River basin, Russia (type locality: Onon and Ingoda rivers). A number of authors conducted morphological studies on what they referred to as Lampetra reissneri and anadromous L. japonica (= Lethenteron camtschaticum) from the lower Amur River basin (Berg 1931, 1948, Nikol'sky 1956, Abakumov 1960, Hensel 1963. However, we are aware of only two studies since the original description, to have examined lamprey material, which they treated as a single species, from the upper Amur (Karasev 1987, Yamazaki et al. 2006. The Amur River basin is divided into five or six welldefined zoogeographic districts  or ecoregions (Abell et al. 2008), respectively, based on the distributional disparities of freshwater fish species along its 4,370-4,510 km length. Thus, it is critical to examine lamprey material from the upper rather than lower Amur, to objectively evaluate the characteristics of L. reissneri and prevent the inclusion of another species in the redescription. Karasev (1987) studied 52 adult lamprey from the Ingoda and the Shilka rivers, but only reported a few morphometric characters and the tooth formula of the infraoral lamina. Yamazaki et al. (2006) studied 26 ammocoetes from the Onon and 29 from the Ingoda rivers and determined that they possessed 65-73 trunk myomeres, but made no mention of their body pigmentation. Yamazaki et al. (2006) examined an adult syntype of L. reissneri (Museum für Naturkunde, Berlin; ZMB 7118) and determined it had 69 trunk myomeres, but made no mention of its dentition. Although Berg (1931) reported that L. reissneri possessed 56-67 trunk myomeres, that range was taken from Jordan and Hubbs (1925) based on Japanese material (Himeji and Lake Biwa, Honshu Island and Sapporo, Hokkaido Island), which they identified as Entosphenus mitsukurii (Hatta 1901). Berg (1931) further stated that the lower labial teeth (= posterial teeth) were sometimes absent in L. reissneri. He subsequently (Berg 1948) changed this to usually absent based on material from the Asian Pacific coast between the Anadyr estuary and the Shangshi River, distant from the type locality. Hubbs and Potter (1971) highlighted the inconsistency between Berg's (1948) statement of usually no lower labial teeth and their observations. Nevertheless, Reshetnikov (2002) repeated Berg's (1948) statement in the widely cited Atlas of freshwater fishes of Russia. The purpose of this study is to provide a full description of the species based on the examination of type and topotypic material of ammocoetes, metamorphosing ammocoetes and adults, and in view of the inconsistency between the observations of Berg (1931Berg ( , 1948 and Hubbs and Potter (1971), re-examine material that Berg studied to determine how often posterials are absent.

Materials and methods
Material was examined from the Academy of Natural Sciences, Philadelphia (ANSP), Natural History Museum, London (BMNH), Canadian Museum of Nature Fish Collection, Ottawa (CMNFI), Naturhistorisches Museum Wien, Vienna (NMW), Zoological Institute, Russian Academy of Sciences, St. Petersburg (ZIN), and Museum für Naturkunde, Berlin (ZMB). Note that collection dates for ZIN collections use the Gregorian rather than the Julian calendar. The characters examined in the ammocoetes and adults follow the method of Renaud (2011). Evaluation of the lateral line neuromast pigmentation in both ammocoetes and adults was made on the dorsal aspect of the branchial region. We additionally examined in ammocoetes the pigment pattern of the posterior fleshy part of the tail as in Richards et al. (1982); their caudal ridge, and the pigmentation in adults of the gular region and apex of the second dorsal fin following Vladykov and Kott (1978). TL, total length. Syntypes (for locality see also map in Fig. 1) BMNH 1871.7.19.37, 1 adult, reported as 120 mm TL in Regan (1911), but only 90+ mm when measured by CBR in 2010, Russia: Onon River, Transbaikalia (also known as Dauria), purchased from Museum Godeffroy (Museum Godeffroy, Hamburg, Germany, existed 1861 to 1885); ZMB 7118, 1 adult, 117.1 mm TL, Russia: Onon River, B.N. Dybowski; NMW 78112, 2 ammocoetes, 106.5-122.2 mm TL, Russia: Onon and Ingoda rivers, upper Amur River basin, received from B.N. Dybowski on 1 Feb. 1870.

Metamorphosing ammocoetes
Seven topotypic metamorphosing ammocoetes from two tributaries to the Onon River examined: six from Barh River, Mongolia (ANSP 185410) and one from Ilya River,  undetermined †Extremely dessicated and contorted specimen; total length cannot be determined accurately. Table 2.
Characteristics re-examined in adults of Lethenteron reissneri (syntypes) and those identified by Berg (1931) as that species. Pigmentation coverage as follows: -< 1%; +++ ≥ 75%. b, bicuspid; u, unicuspid.  Ammocoetes (Table 3) Morphometrics and number of trunk myomeres were determined in two syntypic ammocoetes (NMW 78112, Fig. 3), but specimens extremely faded with virtually no pigmentation. Both possessed prominent streak of dark pigmentation along midline of tail region (Fig. 3). This limited information was augmented by the study of eight topotypic ammocoetes (ANSP 185410 and ZIN uncatalogued) in which virtually all characters studied were recorded.

Discussion
In the original description of Petromyzon reissneri written in German, Dybowski (1869) did not give the number of adults examined and the diagnosis was: 120-140 mm in TL; supraoral lamina with one blunt tooth at each end; infraoral lamina with six blunt teeth, the lateralmost being bicuspid; second dorsal fin almost three times higher than the first dorsal fin and higher than the body depth; dorsal body surface dark and lower surface whitish. Additional information was that spawning occurs in June and that ammocoetes are widespread and abundant, while adults are extremely rare. The small size of adults, the very high second versus first dorsal fins, and the strong bicoloration of the body points to the diagnosis being based on spawning individuals.

Absence vs. presence of posterials in Lethenteron reissneri
Dybowski (1869) did not mention posterials in his brief original description of L. reissneri. Regan (1911) examined a 120 mm adult syntype from the Onon River (BMNH 1871.7.19.37) and identified it as Lampetra planeri (Bloch, 1784), a species that does not possess posterial teeth according to the key he provided in the same publication. Unfortunately, the condition of this syntype, as well as the one in Berlin (ZMB 7118), was such that the presence or absence of posterials could not be ascertained by us. However, the presence of a row of posterials in all topotypic metamorphosing ammocoetes in which this character was discernible, albeit of uncertain numbers, showed that posterials are the predominant if not constant characteristic of this species. This contrasts with the assumption in the recent literature (e.g., Reshetnikov 2002) that posterials in L. reissneri are sometimes (Berg 1931) or usually (Berg 1948) absent. Berg (1931) did not examine any material from the type locality, but recognized L. reissneri as a distinct species. He stated that in this species, the posterial teeth (he called them lower labial or infralabial teeth) were either entirely absent or present as a complete row and he provided drawings of the oral disc, showing these respective conditions [Berg (1931): Pl. VI, Fig. 2, ZIN 15747 and Pl. VI, Fig. 3, ZIN 15734] in two specimens from the Sedanka River, Russia. He documented five cases of L. reissneri without posterials (ZIN 14457, 15078, 15547, 15747, and an uncatalogued male from the Anadyr River). We were able to locate three of these specimens (ZIN 14457, 15078, 15747). Our re-examination revealed the following counts of posterials in a complete row: 24 unicuspid teeth (Fig. 4a); 18 unicuspid, 3 bicuspid, 1 unicuspid teeth (Fig. 4b); and 18 unicuspid teeth (Table 2). In fact, ZIN 14457 (Shangshi River, People's Republic of China) possessed two rows of posterials (Fig. 4a). It is unknown why these numerous posterial teeth were overlooked by Berg; however, it is clear that an absence of posterials in the drawings of the oral disc [Berg (1931): Pl. VII, Fig. 1, ZIN 14457 and Pl. VI, Fig. 2, ZIN 15747] is incorrect. Berg (1948) partially corrected his 1931 comment by stating that non-keratinized posterials could be seen on the left side of the oral disc of specimen ZIN 14457 at 20× magnification. Based on its September collection date, strong dentition with nine teeth on the transverse lingual lamina and the fact that it was collected from brackish waters, ZIN 15078 (Fig. 4b) from the Samarga River, Russia was re-identified as a young adult L. camtschaticum on its feeding migration to the sea. The specimen from the Sedanka River (ZIN 15747) was likewise re-identified as L. camtschaticum based on the presence of nine strong teeth on its transverse lingual lamina. However, we agree with Berg (1931) that the adult from the Shangshi River, Songhua River system (ZIN 14457) is L. reissneri on the basis of its weaker dentition and lower number of teeth (i.e., five) on its transverse lingual lamina. Additionally, the L. reissneri specimen from the Shangshi River has an unpigmented second dorsal fin whereas the L. camtschaticum from the Sedanka River has a blotch on this fin (Table 2). Karasev (1987) reported on seven morphometric characters comparable to those we examined (Table 1) along with the tooth formula of the infraoral lamina in 52 Table 3. Characteristics in ammocoetes of Lethenteron reissneri from upper Amur River basin and Shangshi River. Pigmentation coverage is as follows: -, absent to trace; +, 1% to < 25%; ++, 25% to < 75%; +++, ≥ 75%. The number in parentheses indicates the number of specimens with this condition.   (2) spade-like (2) spade-like (6) spade . The values recorded in the six morphometric characters that were expressed as percentages of TL for the best preserved syntype (ZMN 7118, 117.1 mm, Table 1) fall within the ranges reported by Karasev (1987). Unfortunately, it is unclear from the reported infraoral lamina counts (i.e., 1-4-1 and 1-5-1) whether the lateralmost teeth were enlarged relative to the internal ones. Regardless, the total number of teeth (6-7) is very similar to the only infraoral lamina count (6) in a syntype (ZMB 7118, Table  2). Karasev (1987) observed two lampreys attached to specimens of Leuciscus waleckii (Dybowski, 1869) in the Olengui River (tributary to the Ingoda River) in June 1960 and noted these were the only cases of lamprey attacking fishes in rivers of Transbaikalia observed over a period of years and that an unspecified number of examined adults had empty intestines. We do not consider these observations evidence of lamprey parasitism because attachment does not necessarily lead to feeding and especially since June, as noted above, is the spawning period in the upper Amur River basin and lampreys do not feed during this period (Potter et al. 2014). We believe additionally, that the small size difference between the largest ammocoetes (ZIN uncatalogued, 171 mm TL, Table 3) and adults [182 mm TL, Karasev (1987)] from the upper Amur River system is not conclusive evidence that L. reissneri feeds post metamorphosis and we presume that ammocoetes exceeding 182 mm TL exist. In fact, the 221.5 mm TL ammocoete and 146 mm TL spawning female from the Shangshi River, Songhua River system (ZIN 14457, Tables 1 and 3), that we identify as L. reissneri as did Berg (1931), indicate as in nonparasitic species that ammocoete total lengths exceed those of adults, even allowing for shrinkage in spawning individuals (Vladykov and Kott 1978: fig. 12). The total lengths reported by Dybowski (1869) for the spawning adults (120-140 mm) from the upper Amur are similar to that of the spawning female from the Shangshi River (146 mm) and indicate narrow variation across a wide distributional range (Fig. 1). The absolute fecundity in 15 females collected from the Ingoda River in April-May 1969 and 1970 was estimated to be 1,720-3,360 eggs/female and the egg diameter varied between 0.68 and 0.84 mm (Karasev 1987); a low fecundity expected in a nonparasitic species (Potter et al. 2014).
The number of trunk myomeres reported for 18 adults of L. reissneri from Japan and Sakhalin Island by Vladykov and Kott (1978) was 57-63. This is clearly different from the counts of 70-72 based on adult specimens from the upper Amur River system (syntype) and Songhua River system (Shangshi River, Table 2) and may refer to another species. Therefore, we restrict the distribution of L. reissneri to the Shilka and Songhua river systems, within the Amur River basin (Fig. 1) until a more geographically comprehensive study is undertaken. Yamazaki et al. (2006) gave counts of 65-73 trunk myomeres for 55 topotypic ammocoetes from the Onon and Ingoda rivers. This is similar to the counts of 66-70 reported here (Table 3) based on two syntypes and eight topotypes.
The presence of a dark midline streak on the posterior fleshy part of the tail in ammocoetes of L. reissneri was distinct from the pigmentation patterns reported by Richards et al. (1982) for ammocoetes of Entosphenus macrostomus (Beamish, 1982), E. tridentatus (Gairdner in Richardson, 1836), Lampetra ayresii (Günther, 1870), and L. richardsoni Vladykov & Follett, 1965. The same pattern of dark midline streak appears in Fig. 7b in Vladykov (1950) depicting the tail region of Lethenteron appendix (reported as Entosphenus lamottenii) and in Fig. 2 in Naseka et al. (2009) depicting the tail region of L. ninae. Additionally, Renaud et al. (in press) found the same pigmentation pattern in ammocoetes of L. alaskense and L. camtschaticum. This dark midline streak appears therefore to be a diagnostic characteristic of Lethenteron, but its presence still requires confirmation in L. kessleri.

Diagnosis
Lethenteron reissneri is distinguished from parasitic L. camtschaticum by not feeding as an adult, reaching 182 mm compared to 625 mm TL, 5 versus 9-18 teeth on the transverse lingual lamina. It possesses 70-72 trunk myomeres in adults compared to 54-60 in L. zanandreai (this species has been re-assigned to the genus Lampetra; see Potter et al. (2015) for justification) and 58-62 in L. ninae; absence of pigmentation in the gular region in adults compared to presence of pigmentation in L. appendix; 6-7 infraoral teeth compared to 6-11 (mean 8.6) in L. alaskense.
Lethenteron kessleri (type locality: Tom' River and tributary, Kirgizka River, Ob' River basin, near Tomsk, Russia) is poorly defined and may be a junior synonym of L. reissneri as suggested by Yamazaki et al. (2006) on the basis of similarities in their number of trunk myomeres, allozyme alleles, and mtDNA genes CO I and cyt b sequences in larval material from the type localities (although only in a very general sense in the case of L. kessleri, as the material was from the upper Ob' River basin, in the Irtysh and Uba rivers, Kazakhstan). This synonymy will require corroboration using adults and a re-examination of the former's type material, which these authors did not conduct. It is the present authors' intention to do so in the near future. The two undescribed Lethenteron spp. from Japan are distinguished from L. reissneri on the basis of trunk myomeres (Yamazaki et al. 2006): 49-62 (Lethenteron sp. S) and 51-66 (Lethenteron sp. N).

Distribution
Shilka (Russia and Mongolia) and Songhua (People's Republic of China) river systems within the Amur River basin (Fig. 1).

Conclusion
This study establishes in L. reissneri the usual presence of a row of posterial teeth based on the examination of topotypic metamorphosing ammocoetes from the upper Amur River basin. The long and widely held perception that these teeth were either sometimes or usually absent in the species was based on the observations of Berg (1931Berg ( , 1948. However, re-examination of material Berg (1931) reported to be without posterial teeth showed that these were in fact present as a complete row in the species. Lethenteron reissneri possesses in adults an oral disc 3.9-5.1% of TL, a broad supraoral lamina with one cusp at either end, an endolateral tooth formula 2-2-2, an infraoral lamina with 6-7 teeth (one or both lateralmost teeth bicuspid and the rest unicuspid), 3 rows of anterials, no exolateral rows (but up to two exolateral teeth may be found on one side), 1-2 rows of posterials, the first has 24 unicuspid teeth, a transverse lingual lamina with an enlarged median cusp and two cusps on either side, a spade-like and strongly pigmented (i.e., +++) caudal fin, unpigmented (i.e., -) lateral line neuromasts and gular region, 70-72 trunk myomeres and reaches total lengths of 120-182 mm. The ammocoetes have a spade-like caudal fin, dark midline streak on the posterior fleshy part of the tail, unpigmented (i.e., -) lateral line neuromasts and lower and ventral branchial regions, weakly pigmented (i.e., +) upper lip, moderately pigmented (i.e., ++) upper branchial region, lower lip, and caudal fin, strongly pigmented (i.e., +++) cheek and area between the upper lip and cheek, a moderately pigmented (i.e., ++) bulbous middle prong of the tongue precursor and unpigmented (i.e., -) areas lateral to the elastic ridge, 65-73 trunk myomeres and reach 221.5 mm TL. It undergoes metamorphosis in August at 139.5-164 mm TL and spawns between April and June at 120-146 mm TL.