Asoriculus castellarini (PASA, 1947)

Asoriculus castellarini (PASA, 1947)

Pl. 5, Fig. 4, Pl. 6, Fig. 3

Material. MN 17: Včeláre 3: right md with m1–m3 (as Episoriculus cf. castellarini in Fejfar and Horáček 1983), 4 fragments of left md (3 ramus md, 1 corpus md without dentition). MN 17/Q 1: Včeláre 3/1: 1 part of ramus mandibulae (coronoid process), 1 part of md (corpus mandibulae) with molar tooth. Včeláre 5: fragment of left md (coronoid process). Q 1: Včeláre 4E: fragment of right md (ramus md). Včeláre 4A/7: fragment of left md (corpus md) with m2–m3 (as Episoriculus cf. castellarini in Fejfar and Horáček 1983). Q 1/Q 2: Včeláre 6/7: right md with p4–m1 (alternatively as Neomys newtoni HINTON, 1911 ).

Description and comparison. Smaller size, more gracile teeth than Asoriculus gibberodon . Slight pigmentation of teeth, limited to the apical parts of the crown present in all specimens. Upper sigmoid notch not as shallow as in Asoriculus gibberodon , CdL in samples from Včeláre 3/1

Q

which

and

)

those

NM

Rec

and

(

OTUs

milleri

Neomys compared extant

of of range samples variation to into

,

)

castellarini falling

(cases

age of

1

Q

–

percentages

17

MN

–

late %

,

of)

those

class to character

age

17

MN

particular

–

15

MN for of obtained

samples.

)

data <( Asoriculus metric them of of basic margin

(comparisons cases

variation

of biometric

number

lower

–

*

n

under

of

.

Summary newtoni

fall or

.

N)>

(

. of

10 it

Table sample exceed and Včeláre 6/7 significantly exceeds variance span of A. gibberodon , as former corresponds to smallest individual of N. milleri , and latter to mean value of this species. Coronoid process leans in mesial direction, similarly to Neomys .

We were able to obtain biometric comparative information for 88 mandibular measurements (n = 3–12) of A. castellarini and A. gibberodon , for which the Shapiro-Wilk test showed normal distribution in 80 variables. Deviance from this state (p = 0.003–0.05) was observed in the following variables: Cr3, Cr5, RM1, RM2, mdLm2, Cd1Cd10, TrLm1, TalLm3. In five of them, values of the Q 1 and MN specimens differ distinctly: Cr3, Cr5, TalLm3 (Q 1> MN) and RM1, RM2 (Q 1 <MN). The difference was also indicated by the results of PCA4.

Until the late MN 17, our material is homogenous and comprised the original form A. gibberodon , and similarly to the European fossil record ( Rzebik-Kowalska 1998, 2002), this taxon represents the vast majority of our material. In contrast to our expectations, our samples did not follow the trend of enlarging body size throughout the MN 15 – Q 1 period. In most of the measurements, specimens from MN 15 exceeded the samples from MN 16 and MN 17 and reached values similar to the Q 1 material. Yet when the MN 15 specimens were excluded, the expected trend is confirmed. Similarly to our results, Popov (2003) also reported bimodality while comparing large sample of A. gibberodon from MN 17 Varshets, Bulgaria with other localities (MN 14 – Q 1).

As expected, our results showed a smaller size of Asoriculus in comparison with Q 2 and Q 4 Neomys , as well as extant N. milleri . Sometimes an overlap with N. milleri was recorded. However, some samples exceeded the maximum of N. milleri . A wide terminal part of the coronoid process was observed in samples from Ivanovce (MN 15), Včeláre 3 (MN 17) and Včeláre 4E (Q 1). The specimen from Ivanovce also exceeds the recent species in other aspects (pW, RmL), yet the distance between tip of upper facet and centre of upper sigmoid notch (lingual view, CdL) is shorter. This trend is observed in another 9 specimens (out of 11), and thus could be considered a significant distinguishing criterion between Asoriculus and Neomys . Similarly to Neomys , the width of the talonid of the first two lower molars exceeds the width of the trigonid, yet the m 3 in Asoriculus is more reduced (comp. smaller length and width of talonid).

The sample from Včeláre 3/1 (the earliest Q 1) possesses a combination of characteristics mentioned above. There is not an enhanced gap between the molar row and ascending ramus, the coronoid process points almost straight upwards (without leaning forwards, as is typical for Neomys ) and its spiculum is pronounced. The talonid of the last molar is small; the entoconid is relatively low in all molars. In terms of metric variables, this specimen is one of the largest, especially regarding the CdL. It corresponds well to the diagnosis and redescriptions of A. castellarini by van der Meulen (1973).

In contrast, the specimen from Včeláre 6/7 (late Q 1) shares more characteristics with Neomys than Asoriculus . Even though the spiculum is pronounced, and according to the position of alveoli, the last molar was positioned closely to the ascending ramus, based on the value of CdL and the high coronoid process pointing dorsally, it could alternatively be assigned to Neomys newtoni .

Apparently, Asoriculus gibberodon differs from Neomys newtoni significantly by the shape of the lower premolar (massive unicuspid with pronounced distal prolonging of the labial cingulum in Asoriculus , and wide bicuspid with significant occlusion edges of the cusps in Neomys ). Unfortunately, not a single specimen from all of the European Q 1 fossil record possesses this particular tooth, except for those from Atapuerca, Trinchera del Elephante, Spain ( Rofes and Cuenca-Bescos 2006), in which it corresponds clearly to the state in A. gibberodon , while at the same time, it differs from it by a dorsally tapered coronoid process and distinctly pronounced high position of the upper articulation condyle.

Comments. Unfortunately, our material is not sufficient to test the relevance of the hypothesis that A. castellarini represents a transient form between A. gibberodon and N. newtoni , nor to prove the trend of enlarging body size during the period of MN 17 – Q 1.

A similar issue stands behind the sceptical view on the validity of other forms described from various European Pliocene and Early Pleistocene sites: Soriculus kubinyii KORMOS, 1934 , Episoriculus tornensis JÁNOSSY, 1973 and Episoriculus borsodensis JÁNOSSY, 1973 . Considering the wide variability of Asoriculus gibberodon , these taxa are usually synonymised with this form ( Reumer 1984, Rzebik-Kowalska 1998).

Nevertheless, at least considering the span of variation among our OTU IV specimens, the tendencies conforming to expected transitional state between the Asoriculus and Neomys phenotypes seems to be clearly marked. Hence, we tentatively propose, as a provisional solution, to separate the late MN 17 – Q 1 populations as a distinct taxonomic unit, for which a prior name A. castellarini (PASA, 1947) is available.