Recent developments in the Upper Ordovician and lower Silurian conodont biostratigraphy in Estonia

The current state of and the latest developments in the conodont-based biostratigraphical studies of Upper Ordovician and lower Silurian strata in Estonia are discussed. An updated conodont zonation is proposed, based on successive first and/or last occurrences of taxa as interpreted from the stratigraphical record of Estonia as well as of other parts of the world. The zonation includes mainly units proposed by different authors at different times, with some revision by the author. Currently, 32 zones (with 14 subzones) and 3 informal units can be recognized in the Upper Ordovician to upper Wenlock (Homerian) of Estonia.


INTRODUCTION
Systematic studies of conodonts in Estonia started in the middle of the 1960s.Since then, a wealth of data about conodont taxonomy, palaeoecology, and biostratigraphy in Estonia has been collected.In accordance with the development of conodont studies in other areas, a conodont biozonation has been worked out and updated for the East Baltic (e.g.Viira & Männik 1997).Most of the biozones described elsewhere (e.g.Walliser 1964;Bergström 1971;Jeppsson 1997) are recognized in the East Baltic region.In addition, detailed studies of the Estonian conodont successions have provided valuable information for improving the resolution of the socalled international conodont standard (Viira 1999;Männik in press).Recognition of various ecological associations of conodonts has resulted in different local conodont zonations for shallow and deeper shelf environments of some stratigraphical intervals (Viira 1982).In recent years, studies of the conodont succession have obtained sufficient detail to locate and date several gaps in the Silurian sequence in Estonia (Jeppsson et al. 1994).The main aim of the present paper is to summarize the current state of, and the latest developments in conodont-based biostratigraphical studies of Upper Ordovician and lower Silurian strata in Estonia.

CONODONT ZONATION
The zonation discussed below is based on a succession of units proposed by different authors at different times, with some revisions by the author.The Upper Ordovician zonation corresponds mainly to that of Bergström (1971Bergström ( , 1983)); the Wenlock zonation was developed by Jeppsson (Jeppsson 1997;Jeppsson & Calner 2003); an updated version of the Telychian zonation is available in Männik (in press).At present, 32 zones (with 14 subzones) and 3 informal units can be recognized in the Upper Ordovician to upper Wenlock (Homerian) strata of Estonia (Fig. 1).

The Pygodus anserinus and Amorphognathus tvaerensis zones
The latest data indicate that the boundary between the Middle and Upper Ordovician in Estonia most probably correlates with the boundary between the Uhaku and Kukruse stages (Nõlvak et al. 2006).In the conodont succession this level lies in the middle part of the Pygodus anserinus Zone, in the uppermost Sagittodontina kielcensis Subzone (Männik 2003;Männik & Viira, 2005; Fig. 2).Probable correlations between the Ordovician conodont, chitinozoan, and graptolite zonations in Estonia can be found in Nõlvak et al. (2006).

The Pygodus anserinus Zone
The lower boundary of the Pyg.anserinus Zone coincides with the level where Pyg.serra is replaced by Pyg.anserinus, and its upper boundary is marked by the appearance of Amorphognathus tvaerensis.Two subzones, Lower and Upper, have been distinguished in this zone (Bergström 1971).Their boundary corresponds to the level at which Baltoniodus prevariabilis is evolutionally replaced by B. variabilis (Fig. 1).Later, based on the suggestion by Dzik (1978) that S. kielcensis is replaced by Amorphognathus inaequalis at the same level, Bergström (1983) proposed renaming the Lower and Upper subzones, respectively, as the S. kielcensis and A. inaequalis subzones.However, as these taxa occur sporadically, the boundary between the subzones can be identified mainly on the basis of changes in the Baltoniodus lineage.Unfortunately, as Dzik (1978) pointed out, early specimens of B. variabilis had an amorphognathiform (Pa) element barely distinguishable from that of its ancestor B. prevariabilis.For that reason precise location of the boundary in a section is often problematic.Data from Estonia agree with Dzik's conclusions: morphological changes in the B. prevariabilis-B.variabilis transitions are gradational, thus the exact position of the boundary between the subzones is often hard to define.
In Estonia, the Pyg.anserinus Zone correlates with the upper Uhaku and lower Kukruse stages; the boundary between the S. kielcensis and A. inaequalis subzones has been recognized in the lowermost Kukruse Stage (Männik 2003;Männik & Viira 2005; Fig. 2).The boundary interval of the Pyg.serra and Pyg.anserinus zones is characterized by the occurrence of S. kielcensis, and the uppermost part of the A. inaequalis Subzone of the Pyg.anserinus Zone by the occurrence of Eoplacognathus elongatus Bergström.Eoplacognathus elongatus ranges into the lower part of the overlying A. tvaerensis Zone.In all studied sections Scabbardella ex gr.altipes appears together with B. variabilis.In the Ruhnu-500 and Mehikoorma-421 core sections probable specimens of A. inaequalis have been identified in the upper A. inaequalis Subzone (Männik 2003;Männik & Viira 2005).

The Amorphognathus tvaerensis Zone
The A. tvaerensis Zone was defined by Bergström (1971) as the interval from the first appearance of A. tvaerensis up to that of A. superbus.The A. tvaerensis Zone was subdivided into three subzones (in ascending order): Baltoniodus variabilis, B. gerdae, and B. alobatus (Bergström 1971).Recently it was demonstrated, at least in Estonia, that the disappearance of A. tvaerensis is followed by an interval where Amorphognathus is missing or is too rare to be found in samples available from a core (corresponds approximately to the upper Haljala and Keila stages).Also, A. superbus appears in the uppermost Oandu Stage and is preceded by an interval with A. ventilatus (= main part of the Oandu Stage) (Männik 2003(Männik , 2004;;Männik & Viira 2005).Currently, the A. tvaerensis Zone in Estonia is considered to correspond to the upper Kukruse and lower Haljala (= the Idavere Substage) stages (Fig. 2), up to Datum 1 sensu Männik (2004) of the Mid-Caradoc Event.All three subzones of the A. tvaerensis Zone are well represented by their nominal taxa and easy to recognize in Estonia.

The Mid-Caradoc Event interval
In the conodont succession, the Mid-Caradoc Event is considered to correspond to the interval from the level of disappearance (considerable decrease in abundance?) of A. tvaerensis (= Datum 1 of the event) up to the level of disappearance of the Semiacontiodus lineage (= Datum 5) (Männik 2004).Five datums (levels of major changes in conodont faunas) have been recognized in the event interval.These datums have proved useful in regional stratigraphy for recognition and correlation of several informal units (Männik 2003(Männik , 2004;;Männik & Viira 2005).

"The uppermost Baltoniodus alobatus range"
This informal unit corresponds to the interval between Datum 1 (level of disappearance of A. tvaerensis) and Datum 2 (level of disappearance of B. alobatus) of the Mid-Caradoc Event, and has been recognized in several sections in Estonia (Männik 2003(Männik , 2004;;Männik & Viira 2005; Fig. 1)."The uppermost B. alobatus range" correlates with the upper part of the Haljala Stage (Fig. 2).

The uppermost Ordovician
The Amorphognathus ventilatus Zone Dzik (1999) recognized an interval with A. ventilatus between the ranges (zones) of A. tvaerensis and A. superbus.Later, the A. ventilatus Zone was identified in several core sections in Estonia (Männik 2003;Männik & Viira 2005), with its lower boundary (the level of appearance of A. ventilatus) coinciding with Datum 4 of the Mid-Caradoc Event.The upper boundary of the zone is marked by the appearance of A. superbus (Fig. 1).Previously in Estonia, an interval almost identical to the A. ventilatus Zone was correlated with the Ozarkodina aff.rhodesi-Icriodella cf.superba Zone (Viira 1974; = I. superba-Plectodina sp.Subzone of the A. superbus Zone in Viira & Männik 1997; Fig. 2).The zone corresponds to the main part of the Oandu Stage.
The Amorphognathus superbus Zone Bergström (1971) defined the A. superbus Zone as a unit corresponding to the interval from the first appearance of A. superbus up to the first appearance of A. ordovicicus.In Bergström's scheme the A. superbus Zone follows directly the A. tvaerensis Zone.As we know now, a considerable interval separates the ranges of A. tvaerensis and A. superbus (see above; Fig. 1).In the lower part of this interval Amorphognathus is very rare or absent; in its upper part A. ventilatus occurs.Besides, according to Dzik (1999), A. superbus is not directly followed by A. ordovicicus, but these taxa are evolutionally linked by a form called by Dzik Amorphognathus sp.n., and the interval between the A. superbus and A. ordovicicus zones can be considered as a separate unit.In the zonal scheme proposed by Viira (1974) the interval identified here as the A. superbus Zone corresponds to the main part of the "Ambalodus triangularis frognoeyensis" Zone, and in Viira & Männik (1997) to the upper part of the A. superbus Zone (including the Hamarodus europaeus Subzone; Fig. 2).
The A. superbus Zone corresponds to the uppermost Oandu(?), Rakvere, and Nabala stages, possibly the uppermost part of the Nabala Stage excluded (see below the discussion of the A. ordovicicus Zone).Amorphognathus complicatus Bergström occurs in this zone, particularly in its lower part (e.g.Männik 2003; Fig. 1).

The Amorphognathus ordovicicus Zone
Originally, this zone was defined as corresponding to the total range of A. ordovicicus (Bergström 1971).Amorphognathus ordovicicus was considered to be the last member in the evolutionary lineage of Amorphognathus, which became extinct at the Ordovician-Silurian boundary.However, as Dzik (1999) demonstrated, A. ordovicicus is not the last species in the lineage but is followed by A. duftonus Rhodes.So far, A. duftonus has not been identified in Estonia, but the uppermost Ordovician conodont faunas of the Valga-10 and Ruhnu-500 core sections (Porkuni Stage, its lower part -the Bernati Member -excluded) differ considerably from that of the underlying strata with A. ordovicicus (Männik 2001(Männik , 2003)).In both sections the Ordovician strata above the Bernati Member yield very poor faunas, mainly dominated by Noixodontus girardeauensis.Amorphognathus is missing in these strata (in the Valga-10 core) or is extremely rare and represented by unidentifiable fragments (in the Ruhnu-500 core).
In the Mehikoorma-421 core section, the lowermost specimens of A. ordovicicus are found at 270.00-270.10m, in the upper part of the Saunja Formation (Männik & Viira 2005).The species is almost continuously present in all samples above this level.In the light of the present data it is likely that in Estonia A. ordovicicus appears already in the upper Nabala Stage and not in the lowermost Vormsi Stage as reported earlier (Männik 1992).However, a single M element of Amorphognathus, identical to that of Amorphognathus sp.n.Dzik (1999), was found in the uppermost sample from the Saunja Formation (at 269.10-269.20 m).This may indicate that the lowermost part of the A. ordovicicus Zone (corresponding to the upper Nabala Stage), as identified in the Mehikoorma-421 core section, is an equivalent to the Amorphognathus sp.n. interval sensu Dzik (1999).In Estonia, the A. ordovicicus Zone corresponds to the Vormsi (or the uppermost Nabala?, see above), Pirgu, and lower Porkuni stages (Fig. 2).

"The Noixodontus interval"
This informal unit corresponds to the uppermost Ordovician strata, to the Edole Member of the Kuldiga Formation (Porkuni Stage; Fig. 2), and has thus far been recognized only in the Valga-10 and Ruhnu-500 core sections in southern Estonia (Männik 2001(Männik , 2003)).Conodonts are extremely rare in this unit.The most common species is N. girardeauensis.Single specimens of Decoriconus, Panderodus, Walliserodus, and unidentifiable fragments of Amorphognathodus may occur sporadically.

The Distomodus kentuckyensis Zone
This zone was introduced by Nicoll & Rexroad (1968) as the "Icriodina irregularis" Zone and renamed later by Cooper (1975) as the D. kentuckyensis Zone ("I.irregularis" is the Pa element of D. kentuckyensis).In Estonia, it corresponds to the Juuru and lowermost Raikküla stages (Fig. 2) and is mainly represented by poor conodont faunas dominated by several longranging coniform-bearing taxa (Nestor et al. 2003).As the zonal taxon is very rare in the studied sections (only a few specimens have been found, provisionally identified as D. cf.kentuckyensis), the zone can be defined only tentatively.In the Estonian sections, the Silurian strata below the level of appearance of Aspelundia have been assigned to the D. kentuckyensis Zone.

The Aspelundia Superzone
This Superzone corresponds to the interval from the level of appearance of Aspelundia up to the level of appearance of Pterospathodus eopennatus Männik.It correlates with the main part of the Raikküla Stage and the lowermost Adavere Stage in Estonia (Figs 1, 2).Three zones, As. expansa, As. fluegeli, and Distomodus staurognathoides, discussed in Nestor et al. (2003), are recognized in the Aspelundia Superzone.The lower boundaries of all three zones correspond to the levels of appearance of the nominal taxa.
The Rhuddanian-Aeronian zonation based on Aspelundia was first proposed by Armstrong (1990) for Greenland and can be applied in several other regions (e.g.Timan-northern Ural region and Siberia).Judging from the co-occurrences of conodonts and graptolites, it seems that Aspelundia appears in the uppermost Rhuddanian, in the cyphus graptolite Zone (Loydell et al. 2003).This fits well with the data from Greenland (Armstrong 1990).The data in Loydell et al. (2003) also suggest that the lower boundary of the As.fluegeli Zone evidently lies in the uppermost leptotheca graptolite Zone and that of the D. staurognathoides Zone probably in the convolutus graptolite Zone.

The Lower Pseudooneotodus bicornis Zone
A detailed conodont zonation for the uppermost Telychian and Sheinwoodian and its applicability in different regions is presented in Jeppsson (1997).The zonal boundaries in the lower part of this interval correspond to the datums of the Ireviken Event.
The Lower Ps. bicornis Zone forms the lower part of the Ps.bicornis Superzone.Its lower and upper boundaries coincide, respectively, with Datum 1 and Datum 2 of the Ireviken Event.Datum 2 correlates with the Llandovery-Wenlock boundary in its type section at Leasows (Welsh Borderland) (Aldridge et al. 1993).Recent studies of the graptolite sequence (Männik et al. 2002;Loydell et al. 2003) show that Datum 2 of the event, and accordingly also the Llandovery-Wenlock boundary, does not coincide with the base of the centrifugus Zone but lies higher in the section, in the (upper) murchisoni Zone.In Estonia, the Lower Ps. bicornis Zone corresponds to the lower Jaani Stage (Fig. 2).

Wenlock
The conodont zonation for the interval from the uppermost Telychian (Lower Ps. bicornis Zone) to upper Homerian (Ctenognathodus murchisoni Zone), revised by Jeppsson, includes 17 zones (Jeppsson 1997;Calner & Jeppsson 2003).Most of these zones have been recognized in Estonia.The identification of Jeppsson's zones in Estonia is, as a rule, complicated due to the limited size of samples available from core sections.Besides, some zones evidently correspond to gaps in many of the studied sections, particularly in the outcrop area (e.g.Jeppsson et al. 1994).Often, only the superzones proposed by Jeppsson (1997) can be defined.

The Ozarkodina sagitta rhenana Superzone
This superzone includes the O. s. rhenana and Lower Kockelella walliseri zones (Jeppsson 1997).The lower superzone boundary is defined by the appearance of O. s. rhenana (Fig. 1).In Estonia, the lower boundary of the O. s. rhenana Superzone, and of the O. s. rhenana Zone, lies close to the boundary between the Jaani and Jaagarahu stages (Fig. 2).The lower boundary of the Lower K. walliseri Zone is marked in Estonia by the appearance of K. walliseri in the lowermost part of the Jaagarahu Stage.The zone is characterized by the co-occurrence of O. s. rhenana and K. walliseri, and its upper boundary corresponds to the level of disappearance of O. s. rhenana.

The Kockelella walliseri Zonal Group
This unit was defined by Jeppsson (1997) as corresponding to the interval of total range of K. walliseri and comprising the Lower K. walliseri Zone (characterized by co-occurrence of K. walliseri and O. s. rhenana; forms the upper part of the O. s. rhenana Superzone; see above) and the Upper K. walliseri Superzone with three zones (in ascending order, the Middle K. walliseri Zone, K. patula Zone, and the uppermost K. walliseri range).The strata of the Lower K. walliseri Zone are probably the youngest beds preserved below the major gap between the Jaagarahu and Rootsiküla stages in the outcrop area in Estonia (Jeppsson et al. 1994).The zones of the Upper K. walliseri Superzone cannot be identified in Estonia on the basis of the data available but, evidently, the unit correlates roughly with the middle of the Jaagarahu Stage (Fig. 2).The main problem seems to be too rare occurrence of K. patula, the nominal species of the K. patula Zone (= middle zone in the K. walliseri Superzone; Fig. 1), in the sections.According to L. Jeppsson (pers. comm. 2006), at least 30 kg of rock is needed to find one identifiable specimen of K. patula on Gotland.

The Kockelella ortus ortus Superzone
This superzone corresponds to the interval from the level of disappearance of K. walliseri up to the level of disappearance of Ozarkodina sagitta sagitta (Jeppsson 1997) (to Datum 1 of the Mulde Event, Jeppsson & Calner 2003).Three units (in ascending order), the post-K.walliseri interregnum, K. o. ortus Zone, and O. s. sagitta Zone, form the K. o. ortus Superzone.In Estonia, the strata corresponding to the K. o. ortus Superzone are preserved only in the southwesternmost sections.The lower boundary of the superzone lies in the upper Jaagarahu Stage (Männik 2003; Fig. 2) and probably coincides with the boundary between the Middle and Upper Riksu beds of the Riksu Formation sensu Nestor et al. (2001).

The post-Kockelella walliseri interregnum
This unit corresponds to the interval from the level of disappearance of K. walliseri up to the level of appearance of K. o. ortus (Jeppsson 1997).On the basis of the data from the Riksu core (SW Saaremaa; Nestor et al. 2001), it lies in the upper Jaagarahu Stage.In the paper by Nestor et al. (2001), the post-K.walliseri interregnum was evidently included into the K. walliseri Zone as its upper part.

The Kockelella o. ortus Zone
The boundaries of this zone are marked by the appearance of K. o. ortus (lower one) and O. s. sagitta (upper one) (Jeppsson 1997).In Estonia, the lower boundary of the zone is defined in the Riksu core in the upper part of the Jaagarahu Stage (in the Upper Riksu Beds; Nestor et al. 2001).The upper boundary of the K. o. ortus Zone, based on the data from the Ruhnu-500 core section (Männik 2003), lies also in that stage (Fig. 2).

The Ozarkodina sagitta sagitta Zone
The zone corresponds to the interval of total range of O. s. sagitta (Jeppsson 1997;Jeppsson & Calner 2003).In Estonia, the O. s. sagitta Zone correlates with the uppermost part of the Jaagarahu Stage (Männik 2003; Fig. 2).So far it has been recognized only in the Ruhnu-500 core section but it is possible that at least the lowermost part of the O. s. sagitta Zone is preserved also in the Ohesaare core.The occurrence of Pseudooneotodus linguicornis at 163.55 m in the Ohesaare core (Männik, unpublished data) indicates that the uppermost strata of the Sõrve Formation in that section may correspond to the O. s. sagitta Zone.Pseudooneotodus linguicornis, described in Calner & Jeppsson (2003), has a very short range.According to Jeppsson (Jeppsson in Calner & Jeppsson 2003), this taxon appears just below the lower boundary of the O. s. sagitta Zone and becomes extinct during the Mulde Event.

The Ozarkodina bohemica longa Zone
The zone corresponds to the main part of the upper Wenlock (middle Homerian) Mulde Event (Calner & Jeppsson 2003).It includes five subzones, identified as subzones 0 to 4. Datum 1 of the event, at which the relatively diverse fauna of the O. s. sagitta Zone disappears, is considered as the lower boundary of the O. bohemica longa Zone.According to Jeppsson (in Jeppsson & Calner 2003, p. 139): "The succeeding fauna lacks zonal taxa but otherwise is closely related to the first fauna with O. bohemica longa.Both are markedly less diverse and strongly dominated by ramiforms, in most collections by O. excavata.They are therefore separated as Subzone 0 (zero) and Subzone 1 of the O. b. longa Zone."The most characteristic event in the conodont succession at Datum 2 of the Mulde Event, at the boundary between subzones 1 and 2 of the O. b. longa Zone, is the change of the dominant taxon.The faunas between datums 1 and 2 were dominated by O. excavata, which at Datum 2 was replaced by Panderodus equicostatus (Jeppsson & Calner 2003).In Subzone 3, ramiforms, particularly O. excavata, regained dominance.Subzone 4 has a more balanced conodont fauna.
In Estonia, the Mulde Event interval has been determined in the Ohesaare (see Jeppsson & Calner 2003 and references therein) and Ruhnu-500 core sections.In the latter core, Datums 1 and 2 of the Mulde Event, and accordingly the lower boundary of subzones 0 and 2 of the O. b. longa Zone, were recognized (Männik 2003).Due to the lack of adequate information (evidently because of too small samples) Datum 1.5 cannot be identified, and subzones 0 and 1 of the O. b. longa Zone are not distinguished in the Ruhnu-500 core.The upper part of the O. b. longa Zone (subzones 3, 4) has not been recognized in Estonia yet.Because the strata above the sample with the Subzone 2 fauna were not studied in the Ruhnu-500 core, the upper part of the zone may still be represented there.In Estonia, the O. b. longa Zone corresponds to the uppermost Jaagarahu Stage and the lower part of the Rootsiküla Stage (Fig. 2).

The Ctenognathodus murchisoni Zone
The lower boundary of this zone corresponds to the level of appearance of C. murchisoni (Fig. 1).The C. murchisoni Zone was originally distinguished in Estonia as a regional unit for shallow-water environments and considered to correspond to the upper part of the O. bohemica Zone recognized in the offshore environments (Viira 1982).Later, Jeppsson (in Calner & Jeppsson 2003) recognized the C. murchisoni Zone on Gotland and showed that on the basis of other taxa characteristic of the zone (O.b. longa and morphologically very distinct O. confluens densidentatus), the unit can be defined in offshore environments where C. murchisoni is rare or missing.In Estonia, the C. murchisoni Zone is easy to identify and corresponds, in general, to the Rootsiküla Stage (Fig. 2).

CONCLUDING NOTES Zonation and evolution of faunas
The boundaries of the zones discussed above have been defined by the appearances or disappearances of taxa.The duration of the zones (thickness in the section) is highly variable, due to different rates of the evolution of taxa and of faunas in general (e.g.Jeppsson 1998).Changes in conodont faunas appear in a specific order and are mostly cyclic (Ziegler & Lane 1987).The last authors recognized low-and high-diversity intervals in their cycles.This agrees with Jeppsson's model of Silurian oceanic episodes and events (Jeppsson 1990).According to Jeppsson, low-diversity intervals correspond to secundo episodes and high-diversity intervals to primo episodes.Secundo episodes are as a rule characterized by long-ranging taxa that do not allow detailed biostratigraphy (e.g. the Spirodden Secundo Episode in the Rhuddanian to lower Aeronian corresponds mainly to the D. kentuckyensis Zone).Primo episodes are usually represented by taxonomically variable, rapidly evolving faunas, easy to use in highresolution biostratigraphy (e.g. the Telychian Snipklint Primo Episode faunas; Männik 2001, in press).Rapid and distinct changes in the successions of conodonts during an oceanic event allow the greatest resolution in biostratigraphy and reliable high-precision correlations between different regions (Jeppsson 1997(Jeppsson , 1998;;Jeppsson & Männik 1993).Detailed characterization of the general composition of faunas (co-occurrences, changes in the dominating taxon or taxa, etc.) provides additional valuable information and, in some cases, contributes to further subdivision of the zonal units (e.g.subzones in the O. b. longa Zone, Jeppsson & Calner 2003; possible subzones in the O. s. rhenana Zone, Jeppsson 1997Jeppsson , pers. comm. 1999).Moreover, detailed information about the succession of populations allows subdivision of the intervals poor in conodonts and, in some cases, provides criteria for tracing biostratigraphical boundaries from one facies into another.
As the studies by L. Jeppsson have shown, only collections of adequate size enable proper zonation.It means that often very large samples have to be processed (more than 100 kg of rock).However, the time and effort expended on such studies will not be wasted.A good example is the conodont zonation for the Wenlock.For a long time, the Wenlock was one of the most problematic intervals in Silurian conodont biostratigraphy.In most of the sections studied, faunas appeared to be composed mainly of long-ranging taxa.Other potentially biostratigraphically useful taxa (e.g.Kockelella) occurred sporadically and were too poorly represented to allow reliable biostratigraphical subdivisions.But, after detailed studies of outcrops in Gotland and processing of hundreds of samples (with an average weight from several tens of kilograms to more than 100 kg) from these sections, adequate information about conodont succession in the Wenlock became available and a high-resolution zonation was possible (Jeppsson 1997;Jeppsson & Calner 2003).As demonstrated by Jeppsson, this zonation can be applied in many sections of Wenlock age all over the world, provided that adequate information is available.Although the samples processed from the Estonian core sections are limited in size (usually less than 1 kg) and, as a result, often very poor in conodonts, the thorough characterization of faunas from each zonal unit provided by Jeppsson makes it possible to recognize these units in Estonia.The difficulties with identification of some of the zonal units defined by Jeppsson in Estonia are evidently caused by the lack of adequate data (mainly, too small size of samples available from the core) (e.g.zones in the K. walliseri Zonal Group) or by gaps (e.g. the Mulde Event interval in most of the Estonian sections).

The boundary between the Adavere and Jaani stages, and the Llandovery-Wenlock boundary
The boundary between the Adavere and Jaani stages in Estonia has traditionally been considered to correlate with the Llandovery-Wenlock boundary (e.g.Nestor 1997).The stage boundary, marked by a bentonite, has been drawn at 345.8 m in the Ohesaare core section (Aaloe 1970, p. 244) and considered to coincide with the lower boundary of the centrifugus graptolite Zone.Until recently, it was believed that the base of this graptolite zone corresponds to the Llandovery-Wenlock boundary as it was defined in its type section (Holland 1980).Aldridge et al. (1993) show that the series boundary in its type section lies very close to, or coincides with, Datum 2 of the Ireviken Event.According to recent biostratigraphical and geochemical studies, the boundary between the Adavere and Jaani stages (as defined in the Ohesaare core, see above) correlates with a bentonite at 121.03 m (about 8 m below Datum 2, and 7.2 m below Datum 1, of the Ireviken Event) in the Viki core (Männik et al. 2002 and references therein).It is evident from the data above that (1) the Llandovery-Wenlock boundary is younger than the centrifugus graptolite Zone (lies in the murchisoni graptolite Zone; Männik et al. 2002) and (2) the lowermost Jaani Stage is of Telychian age (Fig. 2).

Some problems for consideration in future
The revision of the latest developments in conodont biostratigraphy in Estonia revealed several problems in need of special attention in future studies.1. Owing to ecological changes in faunas, it is difficult to trace several datums of the Ireviken Event and zonal boundaries in distal offshore environments (e.g.Loydell et al. 2003).Additional studies of those changes are needed.2. The zonation in the Rhuddanian-Aeronian strata needs to be elaborated.Because of the rare, sporadic occurrence of conodonts in most of the sections studied, the exact levels of appearances of zonal taxa (i.e. A. expansa, A. fluegeli, and D. staurognathoides) in the succession are not known.3. Some lineages contain transitional forms between the taxa used to define the zonal boundaries.As said above, late forms of B. prevariabilis are difficult to distinguish from the early forms of B. variabilis.Gradual morphological transition seems to occur between Aspelundia expansa and As.fluegeli.Special morphological studies are needed to find distinct criteria for separating taxa in these lineages.4. Several zonal boundaries are still problematic in Estonia.The boundary between the A. superbus and A. ordovicicus zones has long been considered to lie in the lowermost Vormsi Stage (Männik 1992).
The recent data from the Mehikoorma-421 core section, however, show that A. ordovicicus may appear as low as the upper Nabala Stage (Männik & Viira 2005).Still, it might be that the boundary between these zones was identified correctly in the lowermost Vormsi Stage, and the upper Nabala Stage corresponds to the Amorphognathus sp.n. interval sensu Dzik (1999).5.The succession of changes during the end-Ordovician Event, one of the most throughly investigated intervals, is still poorly known, mainly due to a lack of good sections.Additional information could be obtained from detailed studies of core sections of the offshore part of the Baltic basin (e.g.Valga-10 and Ruhnu-500).6.A serious problem, which is rarely addressed explicitly, is whether the observed pattern in the distribution of faunas in a section is a result of biological changes (evolutionary and extinction events) or an artefact of the structure of the stratigraphical record.It has been established that species ranges in shelf strata reflect beside biological processes also preservation bias imposed by predictable patterns of deposition and erosion in response to eustatic sea-level fluctuations.Therefore, the ranges, used to erect biostratigraphic zones and on which evolutionary lineages and bioevents are based, must be interpreted in the light of this bias to exclude occurrences that are explicable by sequence architecture (Barrick & Männik 2005).
It is evident that some additional studies would increase the resolution of the conodont zonation.The most promising targets are the event intervals.At present, several datums, not yet used to define zonal units, can be employed effectively in the regional high-resolution stratigraphy.The current resolution of conodont biostratigraphy provides an opportunity to analyse the sedimentary history of the (eastern) Baltic palaeobasin in detail, particularly with respect to sequence stratigraphy.Preliminary results of such studies have already been published (Harris et al. 2004(Harris et al. , 2005)).On the other hand, consideration of sequence stratigraphy in biostratigraphical studies is essential to avoiding misinterpretations of palaeontological data and allows selection of appropriate sections for defining the boundaries of regional and global stratigraphical units.

________________________________________________________________________________________________________Fig. 2 .
Fig. 2. Correlation of the Upper Ordovician-lower Silurian condont zonation with regional stages and with the conodont zonation proposed earlier (in Viira & Männik 1997).Conodont zonation: in the Upper Ordovician part, zones are indicated in the lefthand column and subzones in the right-hand column; in the Silurian part, zonation for distal (deep-water) environments is indicated in the left-hand column and zonation for proximal (shallow-water) environments is shown in the right-hand column.Dashed lines indicate probable levels of zonal boundaries.H., Hamarodus; I., Icriodella; P., Plectodina.For other abbreviations refer to Fig. 1.
This zone, introduced byJeppsson (in Calner &  Jeppsson 2003), has not yet been recognized in Estonia.It corresponds to the uppermost part of the Mulde Event and correlates with the topmost Halla Formation on Gotland.In the majority of sections studied in Estonia, this interval most probably corresponds to a gap.