2. The Kolka-54 core

Kolka is situated in northwestern Latvia (Fig. 1). The Kolka-54 core was drilled in 1967. Sampling was undertaken by V. Nestor and V. Viira in 1978. The core is deposited in Riga in the store-house of the Geological Survey of Latvia.

Figure 1. Location of Aizpute-41 on a map showing distribution of early Wenlock (approximately riccartonensis Zone) sedimentary rock types and facies belts in the northern Baltic region (modified from Bassett, Kaljo & Teller, Reference Bassett, Kaljo, Teller, Holland and Bassett1989). The locations of other boreholes mentioned in the text are also indicated. Key to facies belts: 1 – nearshore, high energy shoals; 2 – shallow mid-shelf; 3 – deeper, outer shelf; 4 – deep-shelf depression.

The chitinozoan, graptolite and conodont collections, including all figured specimens, are housed in the Institute of Geology, Tallinn Technological University (prefix GIT).

According to Nestor & Einasto (Reference Nestor, Einasto, Raukas and Teedumäe1997), the Llandovery and Wenlock strata of the Kolka-54 core belong to the deep-water transitional facies belt, except for the upper part of the Rootsiküla Stage (upper Homerian), which is represented by lagoonal dolomites.

3. Graptolite biostratigraphy of the Kolka-54 core

The material available for study comprised both 70 mm diameter core samples and graptolites picked from residues after processing for chitinozoans. The latter material was in some cases highly fragmented, preventing identification to species level. Graptolites from the samples are illustrated in Figures 2–4. Stratigraphical ranges are shown in Figures 5 and 6.

Figure 2. Graptolites from the Kolka-54, core, Latvia. (a) Glyptograptus sinuatus (Nicholson); GIT 352-2326; 625.00 m. (b) Rhaphidograptus toernquisti (Elles & Wood); GIT 352-2315(1); 619.40 m. (c) Streptograptus exiguus (Lapworth); GIT 352-2290; 599.00 m. (d) ‘Monograptus’ austerus Törnquist sensu lato; GIT 352-2323; 623.30 m. (e) Glyptograptus enodis Packham; GIT 352-2313; 616.50 m. (f) Pseudorthograptus insectiformis (Nicholson); GIT 352-2315(2); 619.40 m. (g) Atavograptus atavus (Jones); GIT 352-2317; 621.30 m. (h) ‘Monograptus’ nobilis Törnquist; GIT 352-2315(1); 619.40 m. (i) Streptograptus johnsonae Loydell; GIT 352-2293; 601.00 m. (j) Normalograptus nikolayevi (Obut); GIT 352-2302(2); 605.80 m. (k) Monograptus riccartonensis Lapworth; GIT 352-2251; 556.20 m. (l) Oktavites excentricus (Bjerreskov); GIT 352-2282; 589.30 m. (m) Spirograptus turriculatus (Barrande); GIT 352-2293; 601.00 m. (n) Diversograptus ramosus Manck?; GIT 352-2279; 587.30 m. (o) Parapetalolithus altissimus (Elles & Wood); GIT 352-2291(2); 599.20 m. (p) Coronograptus gregarius (Lapworth); GIT 352-2315(2); 619.40 m. (q) Glyptograptus sp.; GIT 352-2318; 621.50 m. (r) Pristiograptus pergratus Přibyl; GIT 352-2291(1); 599.20 m. (s) Metaclimacograptus slalom Zalasiewicz; GIT 352-2305(1); 607.80 m. (t) Pristiograptus bjerringus (Bjerreskov) GIT 352-2293; 601.00 m. (u) Mediograptus vittatus (Štorch); GIT 352-2261; 566.60 m. (v) ‘Monograptus’ communis Lapworth; GIT 352-2312; 615.50 m. (w) Mediograptus flexuosus (Tullberg); GIT 352-2255(1); 560.00 m. (x, dd) Oktavites spiralis (Geinitz); (x) GIT 352-2287; 591.40 m; (dd) GIT 352-2274; 580.80 m. (y) Monograptus priodon (Bronn); GIT 352-2290; 599.00 m. (z) Mediograptus sp.; GIT 352-2244; 542.00 m. (aa) ‘Monograptus’ inopinus Törnquist; GIT 352-2302(2); 605.80 m. (bb) Streptograptus sp.; GIT 352-2280; 588.50 m. (cc) Cyrtograptus perneri Bouček; GIT 352-2227; 489.50 m. (ee) Monograptus flemingii (Salter); GIT 352-2192; 413.20 m. (ff) Demirastrites simulans (Pedersen); GIT 352-2305(2); 607.80 m. (gg) Cyrtograptus centrifugus Bouček or Cyrtograptus murchisoni (Carruthers); GIT 352-2262; 567.10 m. (hh) Demirastrites triangulatus (Harkness); GIT 352-2318; 621.50 m.

Figure 3. Graptolites from the Kolka-54, core, Latvia. (a) Glyptograptus tamariscus (Nicholson); GIT 565-5; 621.7–622.00 m. (b) Retiolites angustidens Elles & Wood; GIT 565-6; 563.00–563.30 m. (c) Cyrtograptus metatheca; GIT 565-7; 559.0–559.30 m. (d) Streptograptus wimani (Bouček); GIT 565-8; 578.90–579.20 m. (e) ‘Monograptus’ decipiens Törnquist, proximal theca; GIT 565-9; 603.40–603.70 m. (f) Pribylograptus sudburiae (Hutt); GIT 565-10; 614.20–614.55 m. (g) Streptograptus sp.; GIT 565-11; 603.40–603.70 m. (h–j) Colonograptus ludensis (Murchison); (h) GIT 565-12; 353.60–353.65 m; (i, j) GIT 565-13; 380.80–381.10 m. (k, q, r) Cyrtograptus lundgreni Tullberg. (k) ventral view of proximal end; GIT 565-14; 505.50–505.55 m; (q) proximal fragment; GIT 565-20; 510.50–510.55 m; (r) proximal theca; GIT 565-21; 435.0–435.55 m. (l) Barrandeograptus sp.; GIT 565-15; 559.0–559.30 m. (m) Cyrtograptus rigidus Tullberg; GIT 565-16; 519.80–519.85 m. (n, o) Mediograptus sp(p).; 549.60–549.65 m. (n) proximal end; GIT 565-17; (o) distal fragment; GIT 565-18. (p) Colonograptus praedeubeli Jaeger; GIT 565-19; 394.20–394.50 m. (s, t) Pristiograptus dubius var. A sensu Radzevičius (this may be Pristiograptus labiatus Urbanek; P. Štorch, pers. comm.); (s) GIT 565-22; 470.90–470.95 m; (t) GIT 565-23; 484.60–484.65 m. Scale bars represent 1 mm (a, b, h–j, m, o, s, t) or 100 μm (others).

Figure 4. Graptolites from the Kolka-54, core, Latvia. (a, c) Sokolovograptus parens Kozłowska-Dawidziuk; (a) GIT 565-24; 519.80–519.85 m; (c) GIT 565-26; 524.70–524.75 m. (b) Sokolovograptus? sp.; GIT 565-25; 530.0–530.30 m. (d, g) Gothograptus nassa (Holm); (d) GIT 565-27; 408.40–408.70 m; (g) GIT 565-30; 392.70–392.75 m. (e) Eisenackograptus eisenacki (Obut & Sobolevskaya); GIT 565-28; 451.50–452.30 m. (f, j) Sokolovograptus textor (Bouček & Münch); (f) GIT 565-29; 534.60–534.65 m; (j) GIT 565-33; 544.90–544.95 m. (h) Gothograptus kozlowskii Kozłowska-Dawidziuk; GIT 565-31; 427.40–427.70 m. (i) Pseudoplectograptus simplex Kozłowska-Dawidziuk; GIT 565-32; 445.30–445.35 m. Scale bars represent 1 mm (a–c, e, f, h–j) or 100 μm (d, g).

Figure 5. Stratigraphical ranges of graptolites through the Llandovery and lower Wenlock (Sheinwoodian) of the Kolka-54 core. See Figure 10 for lithological legend.

Figure 6. Stratigraphical ranges of graptolites through the upper Wenlock (Homerian) of the Kolka-54 core. See Figure 10 for lithological legend.

The lowest graptolite-bearing sample was close to the base of the Silurian, at 659.50–659.80 m, but this unfortunately yielded only the siculae and th1¹ of indeterminate biserials. The lowest sample to yield identifiable graptolites was at 625.0 m; the presence at this level of Glyptograptus sinuatus (Nicholson) (Fig. 2a) and Rhaphidograptus toernquisti (Elles & Wood) (Fig. 2b) indicates either the upper Rhuddanian cyphus Biozone or lower Aeronian triangulatus Biozone. Triangulate monograptid fragments appear in the 621.7–622.0 m sample, indicating the triangulatus Biozone. Demirastrites triangulatus (Harkness) (Fig. 2hh) itself occurs from 621.50 m to 612.00 m. By comparison with higher Aeronian biozones, the triangulatus Biozone is thus rather thick (at c. 10 m). The simulans Biozone (which occupies a similar stratigraphical position to the magnus Biozone, that is, between the triangulatus and leptotheca biozones) is recognized by the presence of the eponymous species (Fig. 2ff) at 607.80 m. ‘Monograptus’ inopinus Törnquist (Fig. 2aa), at 605.80 m indicates the leptotheca Biozone, while the occurrence of ‘Monograptus’ decipiens Törnquist (Fig. 3e) in the 603.40–603.70 m sample indicates the convolutus Biozone.

The next highest graptolite-bearing sample (at 601.00 m), only 2.4 m above the convolutus Biozone sample, yielded abundant Streptograptus johnsonae Loydell (Fig. 2i), together with Spirograptus turriculatus (Barrande) (Fig. 2m) and Pristiograptus bjerringus (Bjerreskov) (Fig. 2t) indicating the johnsonae Subzone of the lower part of the Telychian turriculatus Biozone. There is interpreted to be an unconformity in the section with the uppermost Aeronian and lower Telychian not represented in the core. The middle Telychian is highly condensed, samples at 599.20 m and 599.0 m yielding species known from the crispus–lower griestoniensis biozones. Pristiograptus pergratus Přibyl (Fig. 2r), in the lower of these two samples, has been recorded previously only from the crispus Biozone. The lowest sample attributable to the spiralis Biozone is at 591.40 m; Oktavites excentricus (Bjerreskov) (Fig. 2l), indicating the middle of this biozone, occurs at 589.30 m. A ventrally curved Streptograptus (Fig. 2bb), broader (maximum DVW 0.8 mm) than S. kaljoi Loydell, Männik & Nestor (maximum DVW 0.6 mm) and with apparently simpler distal thecae, occurs both 0.2 m below and 0.8 m above the O. excentricus-bearing level. The presence of Streptograptus wimani (Bouček) (Fig. 3d) in the 578.9–579.2 m and 575.00–575.30 m samples indicates the lower lapworthi Biozone.

Unfortunately, the next graptolitic samples, between 570.60 m and 568.0 m, yielded only indeterminate material or specimens of long-ranging taxa such as Retiolites and Monograptus priodon (Bronn). A robust, tightly coiled cyrtograptid proximal fragment (either Cyrtograptus centrifugus Bouček or C. murchisoni Carruthers) (Fig. 2gg) occurs at 567.10 m; Mediograptus vittatus Štorch (Fig. 2u), recorded previously only from the murchisoni Biozone, occurs in the next sample, at 566.60 m. The lowest sample attributable to the firmus Biozone is at 560.0 m; Monograptus riccartonensis Lapworth (Fig. 2k), indicating the riccartonensis Biozone, occurs in samples at 556.90 m and 556.20 m.

It is difficult to erect a graptolite biostratigraphy for the samples from higher in the Sheinwoodian for a number of reasons. The biozonal index species Monograptus belophorus Meneghini ( = Monograptus flexilis Elles) is not represented and there is only one specimen of Cyrtograptus rigidus Tullberg (Fig. 3m; at 519.80 m). Several taxa, for example, the various species of ‘Mediograptus’, are in need of taxonomic revision, preventing meaningful identification of the mostly fragmentary, but otherwise well-preserved, isolated specimens encountered (e.g. Fig. 3n, o). These are referred to as ‘Mediograptus’ spp. in the range charts (Figs 5, 6). They are abundant in the 549.60 m, 545.0–545.30 m, 544.90 m and 542.0 m samples at a stratigraphical level consistent with the antennularius Biozone of Rickards (Reference Rickards1967) and White et al. (Reference White, Barron, Barnes and Lintern1992).

The retiolitid Sokolovograptus textor (Bouček) (Fig. 4f, j) has a known stratigraphical range commencing in the upper riccartonensis Biozone (in which it was described as Plectograptus? bouceki by Rickards, Reference Rickards1967) and is common in the upper Sheinwoodian (e.g. Kozłowska-Dawidziuk, Reference Kozłowska-Dawidziuk1995). In the Kolka-54 core its first appearance is at 544.90 m in strata yielding abundant ‘Mediograptus’ fragments and its range extends into the lower lundgreni Biozone (Fig. 6). The Cyrtograptus perneri Biozone has not been recognized. It may lie somewhere between 519.80 m and 510.50 m.

The lowest sample that can be assigned to the lundgreni Biozone is at 510.50 m. This yielded a proximal fragment of Cyrtograptus lundgreni Tullberg (Fig. 3q), exhibiting the characteristic, widely spaced, hooked and spinose thecae of this species. The base of the Homerian (if one takes this at the base of the lundgreni Biozone; see Zalasiewicz & Williams, Reference Zalasiewicz and Williams1999, p. 276 for discussion) thus lies between 519.80 m (sample with C. rigidus) and 510.50 m (lowest C. lundgreni). The highest (indeterminate species) Cyrtograptus fragment occurs at 430.8–431.1 m; the lundgreni Biozone is thus the thickest biozone (at a minimum of c. 80 m) in the Kolka-54 core. At the top of the lundgreni Biozone (430.8–431.1 m sample), Gothograptus kozlowskii Kozłowska-Dawidziuk (Fig. 4h) appears. This distinctive species occurs also at 427.40–427.70 m. Above this, the 415.20 m and 414.90 m samples yielded Pristiograptus dubius (Suess) and the 413.20 m sample Monograptus flemingii (Salter).

The lowest Gothograptus nassa (Holm) (Fig. 4d) fragments occur at 408.40–408.70 m. In the 394.20–394.50 m sample, the orientation of the virgella combined with the narrowness of the sicula enable identification of a damaged proximal end as Colonograptus praedeubeli (Jaeger) (Fig. 3p). The presence of G. nassa (Fig. 4g) in the 392.70 m and 390.0–390.40 m samples indicates that strata from 394.50–390.0 m can be assigned to the praedeubeli Biozone. Graptolites are very sparse higher in the Homerian and are represented only by fragmentary isolated material. Rapidly widening proximal fragments occur at 380.8–381.1 m and 353.60 m and are assigned to Colonograptus ludensis (Murchison) (Fig. 3h–j), indicating the uppermost Homerian ludensis Biozone.

4. Conodont biostratigraphy of the Kolka-54 core

In total, 104 samples were processed and picked for conodonts. All samples, except one from the uppermost part of the studied interval (the 356.60–357.00 m sample; Fig. 8), yielded conodonts. More than 24,000 identifiable specimens were found. The weight of samples varied from 500 g up to 1.4 kg, the number of specimens per kilogram of rock from a few tens (in the lower Raikküla Stage and in the Rootsiküla Stage) up to 4800 specimens (in the 575.0–575.3 m sample). The richest samples were from the Adavere Stage from which most samples (excluding those from the lowermost and uppermost parts of the stage) yielded more than 1000 conodont specimens per kilogram of rock. All figured conodont specimens are deposited in collection GIT 566. Stratigraphical ranges are shown in Figures 7 and 8.

Figure 7. Distribution of conodonts in the lower part of the Kolka-54 core. From left to right: regional stage, depth (in metres), lithological log of the core, sample, sample interval, distribution of taxa (solid line – continuous occurrence of a taxon; dotted line – sporadic occurrence of a taxon; filled circle – confident identification of a taxon; unfilled circle – problematical identification of a taxon), subzone, zone, superzone/zonal group. Abbreviations: Dist. – Distomodus, K. – Kockelella, O. s. – Ozarkodina sagitta, Pand. – Panderodus, Ps. – Pseudooneotodus, Pt. – Pterospathodus, Pt. am. – Pterospathodus amorphognathoides, Pterosp. am. angul. – Pterospathodus amorphognathoides angulatus, Pterosp. am. am. – Pterospathodus amorphognathoides amorphognathoides, Pt. cell. – Pterospathodus celloni, Pt. e. – Pterospathodus eopennatus, U. – Upper. 1 – Pterospathodus eopennatus Superzone; 2 – Pterospathodus amorphognathoides lithuanicus Biozone; 3 – Lower Pterospathodus pennatus procerus Biozone; 4 – Pterospathodus pennatus procerus Superzone.

Figure 8. Distribution of conodonts in the upper part of the Kolka-54 core. From left to right: regional stage, depth (in metres), lithological log of the core, sample, sample interval, distribution of taxa (solid line – continuous occurrence of a taxon; dotted line – sporadic occurrence of a taxon; filled circle – confident identification of a taxon; unfilled circle – problematical identification of a taxon), subzone, zone, superzone. See Figure 10 for lithological legend. Datum 1 and 1.5 refer to the Mulde Event.

The most common taxa in the faunas are those with coniform elements, particularly Panderodus. The almost continuous occurrence of Dapsilodus, but also the rare occurrences of several taxa known from nearer shore environments (e.g. Apsidognathus), and lack of Ozarkodina polinclinata (Nicoll & Rexroad) and some other taxa, indicate that the strata studied formed in distal, deeper shelf environments. The rare, sporadic occurrences of many stratigraphically diagnostic taxa complicates considerably the dating of some intervals in the core.

The conodont biostratigraphy of the strata below the level of appearance of Distomodus staurognathoides (Walliser) is highly problematical in the Kolka-54 core. In this interval the conodont fauna is dominated by simple-cone taxa, mainly by Panderodus ex gr. equicostatus (Rhodes) (Fig. 9a, e, f) and Dapsilodus sp. n. R (Fig. 9b, g). However, the occurrence of Oulodus? cf. panuarensis Bischoff (Fig. 9o) and Ozarkodina ex gr. excavata (Branson & Mehl) (Fig. 9i, n) in the 640.80–641.00 m sample (Fig. 7) indicates that these strata are not older than late Rhuddanian; both taxa are known to appear in the uppermost Juuru Stage in the Baltic region (e.g. Nestor et al. Reference Nestor, Einasto, Männik and Nestor2003). The only identifiable Distomodus in this interval comes from the 654.50–654.80 m sample. The Pa elements found here (e.g. Fig. 9j) are morphologically quite similar to D. combinatus Bischoff (Bischoff, Reference Bischoff1986, pl. 7, fig. 19, pl. 8, figs 1–7), although the processes, particularly the anterior and posterior ones, on our specimens are shorter. The specimens from the Kolka-54 core (identified as D. ex gr. combinatus on Fig. 7) probably are from the earlier part of the D. combinatus lineage.

Figure 9. Selected conodonts from the Kolka-54 core section. (a, e, f) Panderodus ex gr. equicostatus (Rhodes); (a) GIT 566-1, lateral view of aequaliform element, 648.20–648.40 m; (e) GIT 566-2, furrowed face of falciform element, 648.20–648.40 m; (f) GIT 566-3, unfurrowed face of high-based graciliform element, 648.20–648.40 m. (b, g) Dapsilodus sp. n. R; (b) GIT 566-4, inner lateral view of high-based element, 659.50–659.80 m; (g) GIT 566-5, lateral view of symmetrical element, 659.50–659.80 m. (c, d, h) Aspelundia? expansa Armstrong; (c) GIT 566-6, inner lateral view of M element, 618.10–618.50 m; (d) GIT 566-7, inner lateral view of Pb? element, 618.10–618.50 m; (h) GIT 566-8, inner lateral view of Sc element, 614.20–614.55 m. (i, n) Ozarkodina ex gr. excavata (Branson & Mehl); (i) GIT 566-9, inner lateral view of M element, 640.80–641.00 m; (n) GIT 566-10, inner lateral view of Sc element, 640.80–641.00 m. (j) Distomodus ex gr. combinatus Bischoff; GIT 566-11, upper view of Pa element, 654.50–654.80 m. (k–m) Aspelundia? fluegeli (Walliser); (k) GIT 566-12, lateral view of Pa? element, 601.90–602.00 m; (l) GIT 566-13, inner lateral view of Sc element, 601.90–602.00 m; (m) GIT 566-14, inner lateral view of M element, 601.90–602.00 m. (o) Oulodus? panuarensis Bischoff; GIT 566-15, inner lateral view of Sc element, 640.80–641.00 m. (p, r) Pterospathodus amorphognathoides angulatus (Walliser); (p) GIT 566-16, outer lateral view of Sc₁ element, 588.90–589.20 m; (r) GIT 566-17, inner lateral view of Pa element, 593.10–593.30 m. (q, t) Apelundia? fluegeli ssp. n.; (q) GIT 566-18, inner lateral view of Pb? element, 575.00–575.30 m; (t) GIT 566-19, inner lateral view of M element, 575.00–575.30 m. (s, u) Pterospathodus celloni (Walliser); (s) GIT 566-20, (u) GIT 566-21, inner lateral views of Pa elements, 584.90–585.20 m. (v) Panderodus sp. n. N; GIT 566-22, unfurrowed face of falciform element, 568.60–568.80 m. (w) Pseudooneotodus tricornis Drygant; GIT 566-23, upper view, 563.00–563.30 m. (x) Pterospathodus amorphognathoides lithuanicus Brazauskas; GIT 566-24, inner lateral view of Pa element, 581.90–582.20 m. (y) Pterospathodus pennatus procerus (Walliser); GIT 566-25, upper view of Pa element, 575.00–575.30 m. (z) Panderodus cf. langkawiensis Igo & Koike; GIT 566-26, unfurrowed face of low-based graciliform element, 559.00–559.30 m. (aa) Pterospathodus amorphognathoides amorphognathoides Walliser; GIT 566-27, upper view of Pa element, 578.90–579.20 m. (bb) Ozarkodina excavata (Branson & Mehl); GIT 566-28, inner lateral view of Pa element, 419.80–420.10 m. (cc) Kockelella amsdeni Barrick & Klapper; GIT 566-29, upper view of Pa element, 530.00–530.30 m. (dd) Pseudooneotodus linguicornis Jeppsson; GIT 566-30, upper view, 434.60–434.90 m. (ee, ii) Walliserodus sp. n. C; (ee) GIT 566-31, inner lateral (ee1) and outer lateral (ee2) views of deboltiform element, 419.80–420.10 m; (ii) GIT 566-32, inner lateral view of multicostatiform element, 419.80–420.10 m. (ff) Ozarkodina sagitta sagitta (Walliser); GIT 566-33, lateral view of Pa element, 436.70–437.00 m. (gg) Ozarkodina bohemica longa Jeppsson; GIT 566-34, lateral (gg1) and lower (gg2) views of Pa element, 362.30–362.60 m. (hh) Dapsilodus praecipuus Barrick; GIT 566-35, inner lateral view of high-based element, 427.40–427.70 m. (jj) Kockelella ortus absidata Barrick & Klapper; GIT 566-36, lateral view of Pa element, 402.80–403.10 m. Scale bars represent 100 μm.

In the Kolka-54 core, Aspelundia? expansa Armstrong (Fig. 9c, d, h) appears in the 621.70–622.00 m sample and, above this level, is continuously present up to the 601.00–601.20 m sample (Fig. 7). The appearance of A.? expansa normally marks the lower boundary of the A.? expansa Biozone (Armstrong, Reference Armstrong1990). However, based on published data on the distribution of conodonts and other faunas in the Baltic region, the real level of the lower boundary of this biozone in the Kolka-54 core should most probably be looked for at a lower level, close to the lowermost O.? panuarensis in the basal Raikküla Stage (Fig. 7). This conclusion is based on the following: (1) in the Aizpute-41 core, A.? expansa appears in the topmost Belonechitina postrobusta chitinozoan Biozone and O.? panuarensis in the lowermost Euconochitina electa chitinozoan Biozone; in the graptolite biostratigraphy, both appear in the upper cyphus Biozone (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003); (2) in the Heimtali and Põltsamaa cores, O.? panuarensis makes its first appearance again in the E. electa chitinozoan Biozone, although the lowermost specimen of A.? expansa in the Heimtali core was found at a considerably higher level (Nestor et al. Reference Nestor, Einasto, Männik and Nestor2003); (3) in the Ikla and Ruhnu-500 cores the oldest specimens of A.? expansa again come from the E. electa chitinozoan Biozone; based on data from the Ruhnu-500 core, this is from the upper Coronograptus cyphus graptolite Biozone (Kaljo, Männik, Nestor in Põldvere, Reference Põldvere2003; Nestor et al. Reference Nestor, Einasto, Männik and Nestor2003). Considering these data, it is evident that in the Kolka-54 core only the upper part of A.? expansa's stratigraphical range has been recognized and, in this paper, the lower boundary of the A.? expansa Biozone (and of the Aspelundia? Superzone) is therefore tentatively drawn below the level of the lowermost O.? panuarensis (Fig. 7). Such an interpretation of this boundary agrees with data from New South Wales. According to Bischoff (Reference Bischoff1986), in that region O.? panuarensis and Aspelundia? expansa (identified by Bischoff as Oulodus planus planus (Walliser)) appear almost at the same level, with the stratigraphically lowest specimens stated to be from the C. cyphus graptolite Biozone, the same graptolite biozone as their first occurrences in the Baltic region.

The reason for the stratigraphically late appearance of Aspelundia? in the Kolka-54 core is not clear, but it may be an artefact resulting from the rarity of conodonts in the interval from 620 m to 645 m; the number of specimens per kilogram of rock is generally between ten and 40 and the faunas are strongly dominated by (up to 95%) or in many cases are exclusively of taxa with coniform elements. Starting from the level of appearance of Aspelundia?, the number of specimens per kilogram of rock increases considerably and in most samples above this level reaches several hundreds. Also, it is evident that if there is too great an interval between samples (generally more than 3 m) and sample size is limited, this hampers the biostratigraphical resolution of conodonts considerably (see also below; Figs 7, 8).

The lowermost A.? fluegeli (Fig. 9k–m) was found in the 607.60–607.70 m sample, well into the D. staurognathoides conodont Biozone. Previously, the oldest specimens of A.? fluegeli reported from the Baltic region were from the Conochitina alargada chitinozoan Biozone (Nestor et al. Reference Nestor, Einasto, Männik and Nestor2003), and the oldest identifiable specimens of D. staurognathoides were also from the C. alargada chitinozoan Biozone but above the level of appearance of A.? fluegeli (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003). In the Kolka-54 core, both of these taxa again have their lowest occurrence in the C. alargada chitinozoan Biozone: identifiable D. staurognathoides about 7 m below the lowermost A.? fluegeli (Fig. 7). These data indicate that our ideas about the succession of appearances of these taxa in a sequence, and the conodont biozonation proposed for this interval (e.g. Männik, Reference Männik2007a), need further studies and that the conodont biozonation of the Rhuddanian and Aeronian is still problematical.

The lower boundary of the Pterospathodus eopennatus ssp. n. 2 Biozone lies between the 601.00–601.20 m and 596.90–597.20 m samples (Fig. 7). The P. eopennatus ssp. n. 1 Biozone cannot be identified in the Kolka-54 core. Most probably it corresponds to a part of the 2.8 m thick unsampled interval between these samples. In all studied Baltic sections, the appearance of the Pterospathodus faunas is accompanied by considerable changes in the taxonomic composition of conodont assemblages and a sharp increase in the number of taxa and specimens in samples (Männik in Põldvere, Reference Nestor, Einasto and Loydell2003; Männik, Reference Männik2008). In the Kolka-54 core the number of conodont specimens increases almost ten times: from about 345 specimens per kilogram of rock below this level up to 3300 specimens per kilogram above it.

The appearance of P. amorphognathoides angulatus (Walliser) (Fig. 9p, r) in the 593.10–593.30 m sample indicates that the lower boundary of the P. a. angulatus conodont Biozone (and that of the P. celloni conodont Superzone sensu Männik, Reference Männik2007b) lies below this level (Fig. 7). In the same sample the lowermost P. celloni (Walliser) (Fig. 9s, u) was found. As in the Aizpute-41 core (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003), in the Kolka-54 core P. a. lennarti Männik has not been found and P. a. lithuanicus Brazauskas (Fig. 9x) is rare (occuring in only one sample, 581.90–582.20 m; Fig. 7). For this reason, the upper boundary of the P. a. angulatus Biozone is drawn tentatively above the uppermost sample yielding P. a. angulatus. The lower boundary of the P. a. amorphognathoides Biozone is marked by the appearance of P. a. amorphognathoides Walliser (Fig. 9aa) in the 578.90–579.20 m sample (Fig. 7). The interval between this sample and that with the uppermost P. a. angulatus probably corresponds to the P. a. lennarti and P. a. lithuanicus biozones.

Two subzones, Lower and Upper, were recognized in the P. a. amorphognathoides Biozone by Männik (Reference Männik2007b). The boundary between these subzones is defined as the level of disappearance of A.? fluegeli ssp. n., the youngest known representative of the Aspelundia? lineage. In the Kolka-54 core, A.? fluegeli ssp. n. (Fig. 9q, t) occurs in two samples, the higher of which is at 568.60–568.80 m (Fig. 7). From the same sample the uppermost probable specimens of P. a. amorphognathoides in the core have been identified. In general, the distribution of P. a. amorphognathoides in the Kolka-54 core is similar to that in the Aizpute-41 core (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003; conodont biostratigraphy revised in Männik, Reference Männik2007c), that is, the taxon becomes rare, or is missing, in the Upper Subzone of the P. a. amorphognathoides Biozone. This is due to ecology: P. a. amorphognathoides is common in nearshore environments (Männik, Reference Männik1998). In deeper, graptolite-bearing environments it becomes rare and is replaced by P. pennatus procerus (Walliser) (Fig. 9y). In the Kolka-54 core P. pennatus procerus appears in the Lower Subzone of the P. a. amorphognathoides Biozone (Fig. 7).

Two samples, from 566.40–566.70 m and 563.00–563.30 m, probably come from the Upper Subzone of the P. a. amorphognathoides Biozone. The position of the upper boundary of this subzone, and, accordingly, also of the upper boundary of the P. a. amorphognathoides Biozone, in the Kolka-54 core is problematical. However, considering the results of the highly detailed analysis of faunal changes during the Ireviken Event carried out by Jeppsson (Reference Jeppsson, Brett and Baird1997a), it is quite possible that the continuous occurrence of Pseudooneotodus tricornis Drygant (Fig. 9w) up to and including the 563.00–563.30 m sample indicates that this sample lies below Datum 1 of the event (and, accordingly, below the upper boundary of the P. a. amorphognathoides Biozone) (Fig. 7). From the same sample comes the uppermost Panderodus sp. n. N (Fig. 9v), indicating also that Datum 3 of the Ireviken Event, marking the boundary between the Ps. bicornis and P. pennatus procerus superzones sensu Jeppsson (Reference Jeppsson1997b), lies between this and the next sample. Thus, at least two biozones, the Lower and Upper Ps. bicornis biozones, correspond to the 3.7 m thick unsampled interval between the 563.00–563.30 m and 559.00–559.30 m samples (Fig. 7). Also, as the Llandovery/Wenlock boundary in its type section at Hughley Brook correlates approximately with Datum 2 of the Ireviken Event (Jeppsson, Reference Jeppsson1997b), the series boundary in the Kolka-54 core should be looked for in the same interval.

A single specimen of Panderodus cf. langkawiensis Igo & Koike (Fig. 9z) occurring with P. pennatus procerus in the 559.00–559.30 m sample indicates that this sample most probably comes from the Lower P. pennatus procerus Biozone (Fig. 7). This is the highest sample with P. pennatus procerus and D. staurognathoides. These data indicate that the Upper P. pennatus procerus and the Lower Kockelella ranuliformis conodont biozones evidently correspond to the unsampled interval between this and the overlying (555.00–555.30 m) sample. However, as D. staurognathoides becomes very rare in the topmost part of its range (e.g. Jeppsson & Männik, Reference Jeppsson and Männik1993) and as our samples are small, it is quite possible that, in reality, the boundary between the Lower and Upper K. ranuliformis biozones lies higher in the core.

The strata above this level, up to the upper Jaagarahu Stage, are dominated by simple-cone taxa: Pseudooneotodus, Panderodus and Decoriconus. Dapsilodus is quite common. Ramiforms are represented mainly by Ozarkodina excavata (Branson & Mehl) (Fig. 9bb). Although the coniform fauna is quite diverse in this interval, current knowledge of these taxa does not allow reliable recognition in the Kolka-54 core of any of the conodont biozones described from the Wenlock by Jeppsson (Reference Jeppsson1997b). However, a single specimen of Kockelella amsdeni Barrick & Klapper (Fig. 9cc) in the 530.00–530.30 m sample suggests that this sample comes from the Lower K. walliseri Biozone. In North America the range of K. amsdeni coincides with that of Ozarkodina sagitta rhenana and with part of the range of K. walliseri (Barrick & Klapper, Reference Barrick and Klapper1976). According to Jeppsson (Reference Jeppsson1997b), the Lower K. walliseri Biozone is the only unit characterized by co-occurrences of K. walliseri and O. s. rhenana. Unfortunately, no specimens of O. s. rhenana and K. walliseri have been found in the Kolka-54 core samples.

The next level well dated by conodonts lies in the upper Jaagarahu Stage. Based on the occurrence in the 436.70–437.00 m sample of Pseudooneotodus linguicornis Jeppsson (Fig. 9dd) and of O. s. sagitta (Walliser) (Fig. 9ff), this level definitely lies in the O. s. sagitta Biozone. A single broken Pa element of O. s. sagitta in the 448.00–448.30 m sample suggests that the lower boundary of the biozone most probably lies below this sample (Fig. 8).

Ps. linguicornis is continuously present up to the 419.8–420.10 m sample. This sample also contained the uppermost specimens of Walliserodus sp. n. C (Fig. 9ee, ii). According to Jeppsson & Calner (Reference Jeppsson and Calner2003), both of these taxa disappear at Datum 1.5 of the Mulde Event, in the middle part of Subzone 0 of the Ozarkodina bohemica longa Biozone. Starting from Datum 1.5 of the Mulde Event the number of conodont specimens in samples decreases considerably in the Kolka-54 core. In the strata from just below the O. s. sagitta Biozone through to the lowermost O. b. longa Biozone (below Datum 1.5) the number of conodonts varied from several hundreds up to more than 1500 (in the 448.00–448.30 sample), whereas above this interval the number of specimens rarely reaches more than just a few tens, and in many samples it is less than ten. Only in three samples from the uppermost part of the studied interval of the core (the 355.00–355.20 m, 351.80 m and 350.10–350.40 m samples) does the number of specimens reach more than 200 or 300.

By the definition of Jeppsson & Calner (Reference Jeppsson and Calner2003), Datum 1 of the Mulde Event, and the boundary between the O. s. sagitta and O. b. longa biozones, corresponds to the level of extinction of O. s. sagitta. At this level also K. ortus ortus (Walliser), Dapsilodus praecipuus Barrick and D. sparsus Barrick became extinct. As all of these taxa are very rare and have a sporadic distribution in the Kolka-54 core, recognition of Datum 1 is difficult, but most probably it lies (and is tentatively drawn in Fig. 8) above the 427.40–427.70 m sample from which the uppermost specimen of D. praecipuus (Fig. 9hh) was found.

According to Calner & Jeppsson (Reference Calner and Jeppsson2003), the upper boundary of the O. b. longa Biozone corresponds to the level of appearance of Kockelella ortus absidata Barrick & Klapper, the nominal taxon of the following conodont biozone. In the Kolka-54 core K. o. absidata (Fig. 9jj) occurs only in the 402.80–403.10 m sample, probably indicating that the level of this sample lies already in the K. o. absidata Biozone (Fig. 8). This agrees with the graptolite dating of these strata, although O. b. longa (Fig. 9gg) appears higher in the section, in the 399.80–400.00 m sample (Fig. 8). Based on data from Gotland, the lower boundary of the K. o. absidata Biozone lies very close to the boundary between the G. nassa and Co.? praedeubeli graptolite biozones (Calner & Jeppsson, Reference Calner and Jeppsson2003). In the Kolka-54 core the only specimen of K. o. absidata comes from an undated interval between these graptolite biozones (Figs 6, 8). Accordingly, it is most probable that the lowermost O. b. longa in the Kolka-54 core is not from the lowermost part of the total stratigraphical range of this taxon, evidently due to small sample size. Five subzones were defined in the O. b. longa Biozone (Calner & Jeppsson, Reference Calner and Jeppsson2003) but, due to inadequate information, these units cannot be recognized in the studied section. The conodont biostratigraphy of the strata above the level of appearance of K. o. absidata is highly problematical.

In the uppermost part of the studied interval, in the 359.00–359.20 m and 350.10–350.40 m samples, the youngest specimens of Walliserodus sp. known from the East Baltic region have been found (Fig. 8). Previously (Jeppsson & Calner, Reference Jeppsson and Calner2003), it was considered that in the Baltic basin Walliserodus became extinct during the Mulde Event (at Datum 1.5).

5. Chitinozoan biostratigraphy of the Kolka-54 core

The assemblages of lower Silurian chitinozoans in the Kolka 54 core are similar to those of the Ohesaare (Nestor, Reference Nestor1994) and Ruhnu (Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003) cores, with the argillaceous limestones and predominantly argillaceous marlstones containing abundant chitinozoans.

Chitinozoan ranges and biozones (mainly interval zones) are shown in Figures 10 and 11. Most of the identified taxa are illustrated in Figures 12–15. Alongside the well-known taxa are many new and obscure forms, which are left in open nomenclature. Because of the argillaceous nature of the entombing sedimentary rocks, some of the chitinozoans are flattened.

Figure 10. Distribution of chitinozoans in the lower part of the Kolka-54 core. From left to right: regional stage, depth (in metres), lithological log of the core, sample, sample interval, distribution of taxa. Note that the barren interval (Interzone) in the nodular argillaceous marlstones in the Juuru Stage is characterized by red coloration.

Figure 11. Distribution of chitinozoans in the upper part of the Kolka-54 core. From left to right: regional stage, depth (in metres), lithological log of the core, sample, sample interval, distribution of taxa, biozone. See Figure 10 for lithological legend. W – Wenlock; Ld – Ludlow.

Figure 12. Chitinozoans from the Kolka-54 core. (a) Belonechitina postrobusta (Nestor); GIT 546-1; 660.50–661.0 m. (b) Belonechitina aspera (Nestor); GIT 546-2; 660.50–661.0 m. (c) Spinachitina fragilis (Nestor); GIT 546-3; 660.50–661.0 m. (d) Plectochitina nodifera (Nestor); GIT 546-4; 659.50–659.80 m. (e) Cyathochitina campanulaeformis (Eisenack); GIT 546-5; 660.50–661.0 m. (f) Euconochitina electa (Nestor); GIT 546-6; 643.0–643.30 m. (g) Cyathochitina calix (Eisenack); GIT 546-7; 640.80–641.0 m. (h) Plectochitina spongiosa Achab; GIT 546-8; 659.50–659.80 m. (i) Ancyrochitina bifurcaspina Nestor; GIT 546-9; 643.0–643.30 m. (j) Conochitina iklaensis Nestor; GIT 546-10; 643.0–643.30 m. (k) Conochitina alargada Cramer; GIT 546-11; 618.10–618.50 m. (l) Spinachitina maennili (Nestor); GIT 546-12; 632.70–633.0 m. (m) Cyathochitina kuckersiana (Eisenack); GIT 546-13; 623.60–623.90 m. (n) Conochitina edjelensis Taugourdeau; GIT 546-14; 621.70–622.0 m. (o) Conochitina elongata Taugourdeau; GIT 546-15; 621.70–622.0 m. (p) Ancyrochitina convexa Nestor; GIT 546-16 (HT- Ch 248/9795); 621.70–622.0 m. (q) Ancyrochitina ramosaspina Nestor; GIT 546-17; 618.10–618.50 m. (r) Conochitina malleus Van Grootel (nomen nudum); GIT 546-18; 614.60–614.65 m. (s) ‘Vitreachitina’ sp.; GIT 546-19; 603.40–603.70 m. (t) Ancyrochitina rumbaensis Nestor; GIT 546-20; 601.90–601.95 m. (u) Conochitina cf. leviscapulae Mullins & Loydell; GIT 540-21; 603.40–603.70 m. (v) Belonechitina oeselensis Nestor; GIT 546-22; 603.40–603.70 m. (w) Conochitina praeproboscifera Nestor; GIT 546-23; 601.0–601.05 m. (x) Eisenackitina dolioliformis Umnova; GIT 546-24; 596.90–597.20 m. (y) Conochitina visbyensis Laufeld; GIT 546-25; 601.0–601.05 m. Scale bars represent 50 μm.

Figure 13. Chitinozoans from the Kolka-54 core. (a) Bursachitina conica (Taugourdeau & de Jekhowsky, sensu Mullins & Loydell, Reference Mullins and Loydell2001); GIT 546-26; 603.40–603.70 m. (b) Rhabdochitina sp. A; GIT 546-27; 596.90–597.20 m. (c) Conochitina leptosoma Laufeld; GIT 546-28; 601.0–601.05 m. (d) Eisenackitina causiata Verniers; GIT 546-29; 596.90–597.20 m. (e) Eisenackitina inanulifera Nestor; GIT 546-30; 596.90–597.20 m. (f) Belonechitina cavei Mullins & Loydell; GIT 546-31; 596.90–597.20 m. (g) Conochitina emmastensis Nestor; GIT 546-32; 588.90–589.0 m. (h) Angochitina longicollis Eisenack; GIT 546-33; 584.90–585.20 m. (i) Bursachitina nestorae Mullins & Loydell; GIT 546-34; 596.90–597.20 m. (j) Ramochitina costata (Umnova); GIT 546-35; 581.90–582.20 m. (k) Ancyrochitina mullinsi Nestor; GIT 546-36; 578.90–579.20 m. (l) Bursachitina nana (Nestor); GIT 546-37; 568.60–568.80 m. (m) Anthochitina primula Nestor; GIT 546-38; 575.0–575.30 m. (n) Ramochitina ruhnuensis (Nestor); GIT 546-39; 575.0–575.30 m. (o) Conochitina acuminata Eisenack; GIT 546-40; 575.0–575.30 m. (p) Conochitina flamma Laufeld; GIT 546-41; 575.0–575.30 m. (q) Rhabdochitina sp. B; GIT 546-42; 568.60–568.80 m. (r) Belonechitina cf. meifodensis Mullins & Loydell; GIT 546-43; 588.90–589.0 m. (s) Conochitina proboscifera Eisenack; GIT 546-44; 581.90–582.20 m. (t) Calpichitina densa (Eisenack); GIT 546-45; 578.90–579.20 m. (u) Plectochitina pachyderma (Laufeld); GIT 546-46; 575.0–575.30 m. (v) Ramochitina nestorae Grahn; GIT 546-47; 566.40–566.70 m. (w) Margachitina banwyensis Mullins; GIT 546-48; 578.90–579.20 m. (x) Plectochitina magna (Nestor); GIT 546-49; 568.60–568.80 m. (y) Belonechitina sp. A; GIT 546-50; 588.90–589.20 m. (z) Margachitina margaritana (Eisenack); GIT 546-51; 566.40–566.70 m. (aa) Conochitina aff. tuba Eisenack; GIT 546-52; 566.40–566.70 m. (ab) Belonechitina sp. B; GIT 546-53; 552.30–552.60 m. (ac) Linochitina? sp. GIT 546-54; 540.0–540.30 m. Scale bars represent 50 μm.

Figure 14. Chitinozoans from the Kolka-54 core. (a) Conochitina claviformis Eisenack; GIT 546-55; 552.30–552.60 m. (b) Conochitina mamilla Laufeld; GIT 546-56; 552.30–552.60 m. (c) Conochitina tuba Eisenack; GIT 546-57; 545.0–545.30 m. (d) Lagenochitina sp.; GIT 546-58; 540.0–540.30 m. (e) Ancyrochitina paulaspina Nestor; GIT 546-59; 533.0–533.60 m. (f) Conochitina aff. flamma Laufeld; GIT 546-60; 540.0–540.30 m. (g) Bursachitina sp. A; GIT 546-61; 537.30–537.60 m. (h) Sphaerochitina sp. A; GIT 546-62; 517.0–517.30 m. (i) Ramochitina spinosa (Eisenack); GIT 546-63; 514.0–514.30 m. (j) Conochitina armillata Taugourdeau & Jekhowsky; GIT 546-64; 537.30–537.60 m. (k) Ancyrochitina gutnica Laufeld; GIT 546-65; 533.0–533.60 m. (l) Cingulochitina cingulata (Eisenack); GIT 546-66; 530.0–530.30 m. (m) Bursachitina sp. B; GIT 546-67; 527.70–528.0 m. (n) Ramochitina martinssoni (Laufeld); GIT 546-68; 517.0–517.20 m. (o) Ramochitina sp. A; GIT 546-69; 514.0–514.30 m. (p) Belonechitina sp. C; GIT 546-70; 506.60–506.70 m. (q) Ancyrochitina plurispinosa Nestor; GIT 546-71; 496.0–496.30 m. (r) Cingulochitina crassa Nestor; GIT 546-72; 489.0–489.30 m. (s) Cingulochitina baltica Nestor; GIT 546-73; 486.20–486.50 m. (t) Linochitina odiosa Laufeld; GIT 546-74; 486.20–486.50 m. (u) Eisenackitina spongiosa Swire; GIT 546-75; 496.0–496.30 m. (v) Conochitina fortis Nestor; GIT 546-76; 492.60–493.0 m. (w) Conochitina argillophila Laufeld; GIT 546-77; 489.30–489.60 m. (x) Conochitina pachycephala Eisenack; GIT 546-78; 486.20–486.50 m. (y) Conochitina aff. proboscifera Eisenack; GIT 546-79; 476.0–476.40 m. Scale bars represent 50 μm.

Figure 15. Chitinozoans from the Kolka-54 core. (a) Bursachitina sp.; GIT 546-80; 484.60–484.65 m. (b) Conochitina subcyatha Nestor; GIT 546-81; 486.20–486.50 m. (c) Sphaerochitina sp. B; GIT 546-82; 486.20–486.50 m. (d) Cingulochitina gorstyensis Sutherland; GIT 546-83; 472.20–472.70 m. (e) Conochitina linearistriata Nestor; GIT 546-84; 473.80–474.10 m. (f) Plectochitina obuti Nestor; GIT 546-85; 473.80–474.10 m. (g) Conochitina sp. B; GIT 546-86; 455.40–455.50 m. (h) Calpichitina sp.; GIT 546-87; 468.50–468.80 m. (i) Belonechitina sp. D; GIT 546-88; 462.20–462.40 m. (j) Eisenackitina sp. A; GIT 546-89; 459.10–459.30 m. (k) Ramochitina sp. B; GIT 546-90; 459.10–459.30 m. (l) Conochitina cribrosa Nestor; GIT 546-91; 445.30–445.35 m. (m) Ramochitina uncinata Laufeld; GIT 546-92; 422.80–423.0 m. (n) Sphaerochitina concava Laufeld; GIT 546-93; 430.80–431.10 m. (o) Conochitina cf. argillophila Laufeld; GIT 546-94; 427.40–427.70 m. (p) Linochitina sp. A; GIT 546-95; 422.80–423.0 m. (q) Clathrochitina sp.; GIT 546-96; 422.80–423.0 m. (r) Linochitina erratica Eisenack; GIT 546-97; 422.80–423.0 m. (s) Ramochitina sp. C; GIT 546-98; 422.80–423.0 m. (t) Ramochitina tabernaculifera Laufeld; GIT 546-99; 408.40–408.70 m. (u) Eisenackitina sp.; GIT 546-100; 416.0–416.30 m. (v) Rhabdochitina sera Nestor; GIT 527-36; 366.70–366.75 m. (w) Sphaerochitina sp.; GIT 546-101; 408.40–408.70 m. (x) Conochitina sp. A; GIT 546-102; 405.50–405.55 m. (y) Ramochitina sp.; GIT 546-103; 380.80–381.0 m. (z) Sphaerochitina lycoperdoides Laufeld; GIT 546-104; 361.50–361.55 m. Scale bars represent 50 μm.

The Spinachitina fragilis Biozone assemblage is typical of the lowermost Silurian. In the Ohesaare core the corresponding biozone was named the Ancyrochitina laevaensis Biozone (Nestor, Reference Nestor1994), but the latter species was not found in Aizpute and Ruhnu, although it is present in the Kolka-54 core. In addition to S. fragilis Nestor (Fig. 12c), the biozonal assemblage also contains Plectochitina nodifera (Nestor) (Fig. 12d), Belonechitina postrobusta (Nestor) (Fig. 12a), B. aspera (Nestor) (Fig. 12b), Plectochitina spongiosa Achab (Fig. 12h) and abundant Cyathochitina campanulaeformis (Eisenack) (Fig. 12e). The overlying, red-coloured strata (the Rozeni Member) represent an Interzone, which is empty of chitinozoans.

The Belonechitina postrobusta Biozone, which has been recognized in many regions (see Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995), was not identified in the Kolka-54 core. However, it is not excluded that it is actually present in this core, within the unsampled interval between 643.30 m and 647.30 m. The B. postrobusta Biozone can be represented by a condensed interval; for instance, in the Ruhnu core it is represented by only one sample.

The Euconochitina electa Biozone is treated here in a restricted sense, as in its former upper part the Spinachitina maennili Biozone (of Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995) has been distinguished. The most important species associated with E. electa (Nestor) (Fig. 12f) are Conochitina iklaensis Nestor (Fig. 12j), Ancyrochitina bifurcaspina Nestor (Fig. 12i) and Cyathochitina calix (Eisenack) (Fig. 12g).

The Spinachitina maennili Biozone encompasses the upper part of the Slitere Member and most of the Kolka Member in the Kolka-54 core, but is quite thin in some other sections, corresponding only to the upper part of the Kolka Member (Nestor, Reference Nestor1994). S. maennili (Nestor) (Fig. 12l) has not been found in the shallower water sections of Estonia (Nestor, Reference Nestor1998). The associated species in the S. maennili Biozone of the Kolka-54 core are the same as in the underlying biozone, including E. electa. In most of the studied East Baltic sections, Ancyrochitina ramosaspina Nestor (Fig. 12q) occurs together with S. maennili, indicating that this biozone is present also in shallower water sections where the biozonal species is lacking (Nestor, Reference Nestor1994).

Ancyrochitina convexa Nestor (Fig. 12p) is a rather rare species in the East Baltic cores and found seldom in other regions. It appears to be similar in stratigraphical occurrence to Conochitina edjelensis Taugourdeau (Fig. 12n) and Conochitina elongata Taugourdeau (Fig. 12o). As the latter species occurs more commonly than C. edjelensis, this new biozone is named after C. elongata. The abundant occurrence of Cyathochitina kuckersiana (Eisenack) (Fig. 12m) is quite characteristic of this interval (see Nestor, Reference Nestor1994). C. elongata is known from many areas of the world, but has never been used before as a biozonal species.

Conochitina alargada Cramer (Fig. 12k) is the index species of a global biozone (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995). In addition to the index species, the C. alargada Biozone contains numerous Conochitina iklaensis, C. elongata, C. edjelensis and Spinachitina maennili. A newcomer is ‘Vitreachitina’ sp. (Fig. 12s) in the uppermost sample from this interval (603.40–603.70 m). The biozone corresponds to the middle and upper parts of the Raikküla Stage.

Interzone (II), which is devoid of chitinozoans in many East Baltic cores and separates the C. alargada and Eisenackitina dolioliformis biozones, is missing in the Kolka-54 core. Most likely this indicates an unconformity in this section; the lower Stimulograptus sedgwickii graptolite Biozone, established in the Aizpute-41 core (see Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003), is also missing in this interval in Kolka-54.

The lower boundary of the Eisenackitina dolioliformis Biozone is sharp, probably because of the underlying gap in the section. Seven species appear together at the base of the Adavere Stage: the index species E. dolioliformis Umnova (Fig. 12x), Ancyrochitina rumbaensis Nestor (Fig. 12t), Conochitina leviscapulae Mullins & Loydell (Fig. 12u), Belonechitina oeselensis Nestor (Fig. 12v), Conochitina praeproboscifera Nestor (Fig. 12w), Bursachitina conica (Taugourdeau & de Jekhowsky, sensu Mullins & Loydell, Reference Mullins and Loydell2001) (Fig. 13a) and Conochitina emmastensis Nestor (Fig. 13g).

The next sample, at a depth of 601.0–601.05 m, yielded also Conochitina visbyensis Laufeld (Fig. 12y), Conochitina leptosoma Laufeld (Fig. 13c) and Eisenackitina causiata Verniers (Fig. 13d). Rhabdochitina sp. A (Fig. 13b), Eisenackitina inanulifera Nestor (Fig. 13e), Belonechitina cavei Mullins & Loydell (Fig. 13f) and Bursachitina nestorae Mullins & Loydell (Fig. 13i) appear in the middle of the biozone (in the 596.90–597.20 m sample). The uppermost sample from the biozone (593.10–593.30 m) contained only a few chitinozoans, including Ancyrochitina primitiva Eisenack. The E. dolioliformis Biozone corresponds to the lower part of the Adavere Stage and is recognized worldwide (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995).

Angochitina longicollis Eisenack (Fig. 13h), indicative of the succeeding biozone, appears in the Kolka-54 core, together with Belonechitina cf. meifodensis Mullins & Loydell (Fig. 13r) and Belonechitina sp. A (Fig. 13y). It is important to note that the questionably identified specimens of A. longicollis in the lowermost Telychian of the Ohesaare (Nestor, Reference Nestor1994) and Ruhnu (Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003) cores have been re-examined and identified as Angochitina cf. hansonica Soufiane & Achab, Reference Soufiane and Achab2000. Six chitinozoan species have their last occurrences in this biozone, in the middle part of the Adavere Stage. The A. longicollis Biozone is known worldwide (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995; Grahn, Reference Grahn2005, Reference Grahn2006).

Chitinozoan species diversity in the Conochitina proboscifera Biozone of the Kolka-54 core is lower than that in Estonian cores (Nestor, Reference Nestor2005). In the Kolka-54 core, besides the index species (Fig. 13s), Ancyrochitina mullinsi Nestor (Fig. 13k) and Ramochitina costata (Umnova) (Fig. 13j) appear. C. proboscifera is a dominant species for this and the next three biozones, but large numbers of E. causiata and A. longicollis are also found. In the upper part of the C. proboscifera Biozone, Calpichitina densa (Eisenack) (Fig. 13t) and Margachitina banwyensis Mullins & Loydell (Fig. 13w) appear. The M. banwyensis Biozone (see Mullins & Loydell, Reference Mullins and Loydell2001) is not differentiated in the Kolka-54 core, as the eponymous species is represented here by only a few, questionably identified, specimens. C. proboscifera is an easily identified and widely distributed species, but as a biozonal species it has only been used in East Baltic cores (Nestor, Reference Nestor1994, Reference Nestor2005; Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003).

The Conochitina acuminata Biozone was first defined by Mullins & Loydell (Reference Mullins and Loydell2001) in the Banwy River section and is well distinguished also in the Estonian cores (Nestor, Reference Nestor2005). At the base of the biozone, Ramochitina ruhnuensis (Nestor) (Fig. 13n), Anthochitina primula Nestor (Fig. 13m), Conochitina flamma Laufeld (Fig. 13p) and Plectochitina pachyderma (Laufeld) (Fig. 13u) appear along with the index species (Fig. 13o). In the middle and uppermost parts of the biozone, Bursachitina nana (Nestor) (Fig. 13l), Ramochitina nestorae Grahn (Fig. 13v), Rhabdochitina sp. B (Fig. 13q) and Plectochitina magna (Nestor) (Fig. 13x) make their appearance. According to Laufeld (Reference Laufeld1974), typical populations of Conochitina acuminata Eisenack occur only in the Lower Visby Beds and below. This corresponds well with data from the Estonian cores (Nestor, Reference Nestor2005) and from the Banwy River section, where the range of this species correlates with the Cyrtograptus lapworthi graptolite Biozone (Mullins & Loydell, Reference Mullins and Loydell2001).

The lower boundary of the Margachitina margaritana Biozone has caused much debate (Mullins & Loydell, Reference Mullins and Loydell2001; Loydell & Nestor, Reference Loydell and Nestor2005; Nestor, Reference Nestor2005). In the Kolka-54 core, M. margaritana (Eisenack) (Fig. 13z) appears above the base of the Wenlock. In addition to the index species, Conochitina aff. tuba Eisenack (Fig. 13aa) is the only newcomer in this biozone. The disappearance levels of two species are the most important events for this biozone: that of Ramochitina nestorae Grahn at the Llandovery/Wenlock boundary (Grahn, Reference Grahn1998; Mullins & Aldridge, Reference Mullins and Aldridge2004; Nestor, Reference Nestor2005) and that of A. longicollis which identifies the upper boundary of the biozone (Nestor, Reference Nestor1994, Reference Nestor2005). This biozone and the next Interzone correspond to most of the Ireviken Event (Nestor, Einasto & Loydell, Reference Nestor, Einasto and Loydell2002), within which 11 species disappear in the Kolka-54 core.

Cingulochitina bouniensis Verniers, which was not found in the Kolka-54 core, has been identified above the M. margaritana Biozone in the Banwy River section (Mullins & Loydell, Reference Mullins and Loydell2001), as well as in the Aizpute-41 core (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003) and Ruhnu core (Nestor, Reference Nestor2005). The Interzone between the M. margaritana and Conochitina mamilla biozones is characterized by a gradual disappearance of species, but Belonechitina sp. B (Fig. 13ab) and Conochitina claviformis Eisenack (Fig. 14a) appear in the topmost part of this interval. The latter is one of the most abundant and long-ranging Silurian species (Nestor, Reference Nestor1994, Reference Nestor2007).

Conochitina mamilla Laufeld (Fig. 14b) is the index species of the next biozone. It is less common than, and is sometimes difficult to differentiate from, C. claviformis. The biozone contains only a few, mostly long-ranging species. This and the next biozone have been established and used up to now only in East Baltic chitinozoan biostratigraphy (Nestor, Reference Nestor1994), correlating there with the middle of the Jaani Stage.

The diversity of species increases considerably in the Conochitina tuba Biozone. In addition to C. tuba Eisenack (Fig. 14c), Ancyrochitina paulaspina Nestor (Fig. 14e) and Calpichitina acollaris appear in its lowermost part, followed by Lagenochitina sp. (Fig. 14d), Linochitina? sp. (Fig. 13ac), Conochitina aff. flamma (Fig. 14f), Bursachitina sp. A (Fig. 14g) and Conochitina armillata Tougourdeau & Jekhowsky (Fig. 14j). In the uppermost part, Ancyrochitina gutnica Laufeld (Fig. 14k) appears. This biozone embraces the upper part of the Jaani Stage.

Many newcomers (11 taxa) occur in the Cingulochitina cingulata Biozone. One after another there appear Cingulochitina cingulata (Eisenack) (Fig. 14l), Bursachitina sp. B (Fig. 14m), Linochitina odiosa Laufeld (Fig. 14t), Ramochitina martinssoni (Laufeld) (Fig. 14n), Sphaerochitina sp. A (Fig. 14h), Ramochitina spinosa (Laufeld) (Fig. 14i), Ramochitina sp. A (Fig. 14o), Belonechitina sp. C (Fig. 14p), Plectochitina? sp., Ancyrochitina plurispinosa Nestor (Fig. 14q) and Ramochitina uncinata Laufeld (Fig. 15m). The biozonal index species is widely distributed around the world (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995).

Eisenackitina spongiosa (Fig. 14u), originally described from the Coalbrookdale Formation, Shropshire (Swire, Reference Swire1990), is the index species of the next biozone. In the middle Wenlock beds of the East Baltic, Nestor (Reference Nestor1994) identified Eisenackitina lagena, which Eisenack (Reference Eisenack1968) described originally from graptolitic erratic boulders of early Ludlow age. High-resolution SEM study has revealed differences in the ornamentation of the vesicle wall of these two species that are usually identical in overall shape. Like the East Baltic Wenlock species, E. spongiosa has a spongy ornamentation of the vesicle wall (Nestor, Reference Nestor1994, this paper), whereas E. lagena has a porous vesicle wall, covered with granules or tubercles (Nestor, Reference Nestor2007). Thus, it is not the E. lagena Biozone, but the E. spongiosa Biozone that occurs above the C. cingulata Biozone in the East Baltic Wenlock succession. In addition to the index species, the biozone yields an abundant assemblage of transitional species, while C. claviformis is still the most numerous. In addition to C. cingulata, C. crassa Nestor (Fig. 14r) and C. baltica Nestor (Fig. 14s) appear in the upper part of the biozone, together with Conochitina fortis Nestor (Fig. 14v) and Conochitina argillophila Laufeld (Fig. 14w).

In the Ohesaare core the interval between the appearance level of Conochitina pachycephala Eisenack (Fig. 14x) and Conochitina subcyatha (Fig. 15b) is more than 20 m (Nestor, Reference Nestor1994), whereas in the Kolka-54 core these (biozonal) species appear at the same level. In the lower–middle part of this joint biozone appear Conochitina aff. proboscifera Eisenack (Fig. 14y), C. linearistriata Nestor (Fig. 15e), Cingulochitina gorstyensis Sutherland (Fig. 15d) and Plectochitina obuti Nestor (Fig. 15f). Sphaerochitina sp. B (Fig. 15c), Bursachitina sp. (Fig. 15a), Conochitina sp. B (Fig. 15g), Calpichitina sp. (Fig. 15h), Belonechitina sp. D (Fig. 15i), Eisenackitina sp. A (Fig. 15j) and Ramochitina sp. B (Fig. 15k) are newcomers. C. pachycephala is the index taxon of a global biozone (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995), while the geographical distribution of C. subcyatha is restricted to the East Baltic (Nestor, Reference Nestor1994), Wales (Verniers, Reference Verniers1999) and the Brabant Massif, Belgium (Verniers et al. Reference Verniers, Van Grootel, Louwye and Diependaele2002).

In the Kolka-54 core the Conochitina cribrosa Biozone embraces also the Sphaerochitina indecora Biozone, which was erected for strata above the C. cribrosa Biozone in the Ohesaare core (Nestor, Reference Nestor1994). According to Laufeld (Reference Laufeld1974), Sphaerochitina is a facies-controlled genus, not occurring in deep-water sections. Nevertheless, some species of Sphaerochitina, S. concava Laufeld (Fig. 15n) and Sphaerochitina sp., do appear in this biozone in the Kolka-54 core. Within the stratigraphical range of C. cribrosa Nestor (Fig. 15l), many Wenlock chitinozoan species disappear, including Margachitina margaritana. The C. cribrosa Biozone, recognized only in East Baltic cores, is characterized also by the appearance of Conochitina cf. argillophila Laufeld (Fig. 15o), Conochitina sp. A (Fig. 15x), Linochitina erratica (Eisenack) (Fig. 15r), Linochitina sp. A (Fig. 15p), Clathrochitina sp. (Fig. 15q), Eisenackitina sp. (Fig. 15u), Ramochitina tabernaculifera (Laufeld) (Fig. 15t), Ramochitina sp. C (Fig. 15s) and Rhabdochitina sera Nestor (Fig. 15v).

Sphaerochitina lycoperdoides Laufeld (Fig. 15z) is the index species of the uppermost chitinozoan biozone in the Wenlock. In addition to long-ranging species, Ancyrochitina cf. ansarviensis Laufeld and Ramochitina sp. (Fig. 15y) are also present. The number of chitinozoans in samples varies, but decreases upwards within the biozone. The index species itself is rare, being sometimes represented by uncharacteristic specimens. The S. lycoperdoides Biozone has been established also in other East Baltic cores, for example, Ohesaare, Ventspils, Pavilosta and Gussev-1 (Nestor, Reference Nestor2007). In more offshore sections, some Wenlock species (C. claviformis, C. pachycephala, C. tuba) range without interruption up to the Ludlow, but in the Kolka-54 and Ohesaare cores, all Wenlock species disappear in the uppermost part of the Rootsiküla Stage (Nestor, Reference Nestor2007). There is an interval, lacking chitinozoans, which has previously (Nestor, Reference Nestor1994) been distinguished as Interzone (V). The S. lycoperdoides Biozone has been included in the Silurian chitinozoan global biozonation scheme (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995).

6. Correlation of graptolite and conodont biozones

As the lowest graptolite biozone to be identified confidently is the Demirastrites triangulatus Biozone, we begin our discussion of graptolite–conodont biozone correlation (and of graptolite–chitinozoan biozone correlation, below) at the base of the Aeronian.

The lower and middle Demirastrites triangulatus graptolite Biozone correlates with the upper part of the Aspelundia? expansa conodont Biozone. The lowest Distomodus staurognathoides occurs in the 614.60–614.80 m sample, in the upper part of the D. triangulatus graptolite Biozone. This is at a significantly lower stratigraphical level than previously recorded. In the Aizpute-41 core, the base of the D. staurognathoides conodont Biozone lay within the Lituigraptus convolutus graptolite Biozone, more than two graptolite biozones higher than its Kolka-54 level (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003, their fig. 17). Distomodus was, however, rare and poorly preserved in the Aizpute-41 core. It is clear that the D. staurognathoides conodont Biozone represents a very long period of time, comprising almost all of the Aeronian, plus the early part (guerichi–turriculatus graptolite zones) of the Telychian.

Wide conodont sample spacing in the lower Telychian part of the Kolka-54 core means that the base of the Pterospathodus eopennatus ssp. n. 2 Biozone (between 601.00–601.20 m and 596.90–597.20 m) cannot be tied to the graptolite biostratigraphy with any great precision. It is clearly above the Streptograptus johnsonae Subzone of the Spirograptus turriculatus Biozone, which is consistent with its level in the Aizpute-41 core (base of Streptograptus sartorius graptolite Biozone: Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003, their fig. 17).

The base of the P. amorphognathoides angulatus conodont Biozone (and P. celloni conodont Superzone sensu Männik, Reference Männik2007b) in the Kolka-54 core lies between graptolitic samples of middle Telychian (crispus to lower griestoniensis graptolite Zone) age and the base of the Oktavites spiralis graptolite Biozone. This is consistent with its position in the Aizpute-41 core, in which the base lay in the middle of the crenulata graptolite Biozone (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003, their fig. 17).

The base of the P. a. amorphognathoides conodont Biozone is at the same level as that of the Cyrtograptus lapworthi graptolite Biozone in the Kolka-54 core. This is in agreement with data from the Ohesaare (Loydell, Kaljo & Männik, Reference Loydell, Kaljo and Männik1998), Aizpute-41 (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003) and Ruhnu (Kaljo, Männik in Põldvere, Reference Nestor, Einasto and Loydell2003) cores, indicating consistent correlation of these two important biozones throughout the East Baltic region. As noted above, recognition of the top of the P. a. amorphognathoides conodont Biozone is problematical in the Kolka-54 core. It would appear that the top of this biozone, plus the entire Lower and Upper Ps. bicornis biozones, lie within the upper Cyrtograptus murchisoni and/or Monograptus firmus graptolite Biozone(s).

The single sample (559.00–559.30 m) identified as belonging probably to the Lower P. pennatus procerus conodont Biozone is from a level within the firmus graptolite Biozone. In the Aizpute-41 core, Loydell, Männik & Nestor (Reference Loydell, Männik and Nestor2003) tentatively placed the Lower/Upper P. pennatus procerus conodont Biozone boundary within the upper murchisoni graptolite Biozone. The Kolka-54 core data suggest that this boundary probably lies higher in the graptolite biozonation, within the Monograptus firmus Biozone. In the Ruhnu core (Kaljo, Männik in Põldvere, Reference Nestor, Einasto and Loydell2003), the base of the P. p. procerus conodont Biozone lies in a condensed interval between 448.70 m (upper Cyrtograptus murchisoni graptolite Biozone) and 447.0 m (Monograptus riccartonensis Biozone).

The 530.00–530.30 m sample suggested to be from the Lower K. walliseri Biozone is within the ‘middle Wenlock’ graptolitic interval, between the Monograptus riccartonensis and Cyrtograptus lundgreni graptolite biozones.

The next confidently recognized conodont biozone is the O. s. sagitta Biozone, which clearly correlates with part of the upper Cyrtograptus lundgreni graptolite Biozone. Jeppsson (Reference Jeppsson1997b, fig. 3) had proposed that the O. s. sagitta conodont Biozone and C. lundgreni graptolite Biozone were broadly correlative. Unfortunately, few biostratigraphically useful graptolites were recovered from between 430.8 m (Cyrtograptus lundgreni graptolite Biozone) and 408.70 m (Gothograptus nassa graptolite Biozone). The base of the Ozarkodina bohemica longa conodont Biozone is tentatively placed above the 427.40–427.70 m sample. Jeppsson & Calner (Reference Jeppsson and Calner2003, fig. 2) show the base of the O. b. longa conodont Biozone as correlating with a level high in the C. lundgreni graptolite Biozone (within the Testograptus testis Subzone). The Kolka-54 data are consistent with this proposal. As noted above, the conodont biozonation of the Kolka-54 core above the appearance of O. b. longa is problematical and for this reason we do not discuss possible correlations with the graptolite biostratigraphy.

7. Correlation of graptolite and Chitinozoan biozones

The lowest sample assignable to the basal Aeronian triangulatus graptolite Biozone lies within the upper part of the newly erected Conochitina elongata chitinozoan Biozone. The base of the Conochitina alargada chitinozoan Biozone lies within the triangulatus graptolite Biozone in the Kolka-54 core at a level similar to that at which it was tentatively identified in the Aizpute-41 core (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003, p. 215) and confidently recognized in the Ruhnu core (Kaljo, Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003). The biozone includes also the simulans to convolutus biozones in the Kolka-54 core as it did also in the Aizpute-41 core (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003). The base of the Eisenackitina dolioliformis Biozone lies above a stratigraphical gap and thus, as in other East Baltic sections, the level of the true base of this biozone with respect to the graptolite biozonation is unknown.

As in the Aizpute-41 core, the base of the Angochitina longicollis chitinozoan Biozone is close to that of the spiralis graptolite Biozone. The base of the Conochitina proboscifera Biozone in the Aizpute-41 core was in the upper Oktavites spiralis graptolite Biozone (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003); this is also consistent with the Kolka-54 biostratigraphical data and first appearance of C. proboscifera in the Banwy River section, Wales (Mullins & Loydell, Reference Mullins and Loydell2001). In the Ruhnu core, however, the base of the C. proboscifera chitinozoan Biozone appears to be slightly lower, in the lower or middle part of the O. spiralis graptolite Biozone (Kaljo, Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003).

Conochitina acuminata Eisenack was not found in the Aizpute-41 core samples. In the Banwy River section the acuminata Biozone correlated with the lower lapworthi graptolite Biozone (Mullins & Loydell, Reference Mullins and Loydell2001). The base of the acuminata biozone is at the base of the lower lapworthi Biozone in the Kolka-54 core, as it is also in the Ohesaare core (base between 353.70 m and 354.50 m; Nestor, Reference Nestor2005; see Loydell, Kaljo & Männik, Reference Loydell, Kaljo and Männik1998 for graptolite biostratigraphy).

The first appearance of Margachitina margaritana lies 0.4 m above the lowest graptolitic sample to yield a robust Cyrtograptus (centrifugus or murchisoni). This is the stratigraphical level at which M. margaritana typically appears in the East Baltic. Loydell & Nestor (Reference Loydell and Nestor2005), however, illustrated stratigraphically older material of M. margaritana, from the Telychian O. spiralis graptolite Biozone of the Ventspils core, demonstrating the diachronous appearance of this species.

The base of an ‘interzone’, lacking stratigraphically diagnostic chitinozoans, correlates approximately with the base of the firmus graptolite Biozone in the Kolka-54 core, as it does also in the Ohesaare and Aizpute-41 cores (Loydell, Männik & Nestor, Reference Loydell, Männik and Nestor2003, p. 225) and Ruhnu core (Kaljo, Nestor in Põldvere, Reference Nestor, Einasto and Loydell2003).

Previous integrated biostratigraphical studies in the East Baltic region have terminated in the lower Sheinwoodian, so discussion here of stratigraphically higher graptolite–chitinozoan biozone correlation focuses on comparison with the ‘global Chitinozoa biozonation’ of Verniers et al. (Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995) and with the work of Verniers (Reference Verniers1999), who erected a chitinozoan biostratigraphy for the Wenlock of Builth area of Wales and integrated this with the graptolite biozonation of Zalasiewicz & Williams (Reference Zalasiewicz and Williams1999) using samples from the same localities and horizons.

Conochitina mamilla occurs in only two samples in the Builth area (Verniers, Reference Verniers1999, fig. 2). The lower of these (sample K27) is immediately below the lowest sample (K28) to yield the biozonal index Monograptus riccartonensis (Zalasiewicz & Williams, Reference Zalasiewicz and Williams1999, fig. 6); the other C. mamilla-bearing sample in the Builth region was from within the M. riccartonensis graptolite Biozone. In the Kolka-54 core the first C. mamilla occur in the 552.30–552.60 m sample above the M. riccartonensis-bearing core sample at 556.20 m, but below the lowest Mediograptus-rich sample at 549.60 m. C. mamilla ranges into the ‘middle Wenlock’ graptolitic interval. Thus the Kolka-54 and Builth area occurrences of C. mamilla are at broadly similar stratigraphical levels.

Cingulochitina burdinalensis Verniers, index species of the succeeding chitinozoan biozone in the Builth area, is not present in the Kolka-54 samples. Conochitina tuba is present, however; its first appearance is at 545.0–555.30 m, towards the top of the Mediograptus-rich interval in the Kolka-54 core. As noted above, this Mediograptus-rich level has been termed previously the antennularius Biozone; this includes the upper riccartonensis Biozone of standard usage. Verniers’ lowest C. tuba (from sample N33C) is from the lowest sample assigned by Zalasiewicz & Williams (Reference Zalasiewicz and Williams1999) to their Pristiograptus dubius Biozone, which immediately overlies the riccartonensis Biozone. In the Ohesaare core, the base of the tuba Biozone lies immediately below a sample bearing Monograptus flexilis, but above the dubius Interzone (Loydell, Kaljo & Männik, Reference Loydell, Kaljo and Männik1998). Thus the base of C. tuba chitinozoan Biozone appears to lie at a level low within the ‘middle Wenlock’ with slight variation between localities as to the precise stratigraphical level.

The lowest Cingulochitina cingulata occurs low in the dubius Biozone in the Builth area, somewhat lower stratigraphically than suggested by Verniers et al. (Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995, fig. 2). Its first appearance is probably a little higher stratigraphically in the Kolka-54 core than in the Builth area, but is still within the middle Wenlock (includes dubius and rigidus graptolite biozones of Zalasiewicz & Williams, Reference Zalasiewicz and Williams1999). Eisenackitina spongiosa was not recorded in the Builth area. The succeeding pachycephala and subcyatha biozones are combined in the Kolka-54 core as the two index species first appear at the same level (486.20–486.50 m), in the lower third of the lundgreni graptolite Biozone. In the Builth area, Conochitina pachycephala makes its first appearance only a little lower, in the upper Sheinwoodian rigidus graptolite Biozone (at the same level as shown by Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995, fig. 2), while Conochitina subcyatha appears a little higher, close to the base of the lundgreni graptolite Biozone. The base of the Conochitina cribrosa chitinozoan Biozone is at the same stratigraphical level in both the Kolka-54 core and the Builth area, in the upper lundgreni graptolite Biozone. The cribrosa chitinozoan Biozone was the highest Wenlock chitinozoan biozone recognized in the Builth area; neither of the upper Homerian samples studied by Verniers (Reference Verniers1999) yielded Sphaerochitina lycoperdoides. The S. lycoperdoides Biozone is the highest Wenlock chitinozoan biozone recognized within the ‘global Chitinozoa biozonation for the Silurian’ (Verniers et al. Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995). From the limited Kolka-54 core graptolite data, the lycoperdoides Biozone appears to correlate with the praedeubeli to ludensis graptolite biozones. Overall, the correlations between graptolite and chitinozoan biozones within the Wenlock of the Kolka-54 core are similar to those proposed by Verniers et al. (Reference Verniers, Nestor, Paris, Dufka, Sutherland and Van Grootel1995) in their ‘global Chitinozoa biozonation’.