Mem. Soc. Geol. /t., 20 (1979), 405-423. 9 ff., 4 tabb.

PALEOECOLOGICAL REMARKS ON TUE " LIMESTONE" OF SOUTHWESTERN SARDINIA (MIDDLE-UPPER ) ( * )

Memoria di MAURIZIO GNOLI (**), GrAN CLEMENTE PAREA (***). FRANCO Russo(**) & ENRICO SERPAGLI (**)

(presentata a Siena, nella Seduta /ematica del /8-/9 maggio 1979)

CO NTE NT ricco di fauna esclusivamente pelagica nelle parti p age superiori e decisamente tossico verso il fondo. lNTRODUCTION 405 Analisi statistiche ed esperimenti di laborato­ l"Ìo hanno portato gli scriventi a considerare il GEOLOGICAL REMARKS 406 particolare tipo di orientamento presentato dai SEDIMENTOLOGICAL REMARKS 407 fossili come un effetto del moto ondoso. I gusci ed i loro frammenti, sparsi su un fondo di fango Geometry of the "Orthoceras lenses" 407 fine, verrebbero disposti dalle onde a bande suc­ The problem of orientation 410 cessive che rappresenterebbero l'equivalente mec­ Laboratory experiments 411 canico delle increspature di fondo tipiche dei fon­ Orientation analysis of the orthoconic di sabbiosi. Potrebbe trattarsi di un mare epicontinentale 413 di profondità limitata tale da permettere al moto GENERAL FEATURES OF THE FAUNA 417 ondoso di agire sul fondo. Nektic organisms associated with the grap- tolites 417 Organisms possibly attached to floating ABSTRACT plants 417 A study of the fauna and of the sedimento­ ENVIRONMENTAL CONCLUSIONS 420 logica! characteristics of the "Orthoceras limestone" REFERENCES 422 has shown that these sediments were deposited in a normally oxygenated sea rich in an exclusively pelagic fauna in the upper parts but definitely KEY WORDS : Shelf environment, death as­ toxic towards the bottom. semblage, sedimentary structures, depth Statistica! analyses along with laboratory ex­ periments have led. the writers to consider that

indicators, Silurian. the particular type · of orientation found in the is the result of ware motion. The shells and their fragments, distributed on the fine mud of the bottom, must have been positioned by the RIAS SUNTO waves in successive belts which represent the mechanical equivalent of ripple marks, typical of Dallo studro della fauna e delle caratteristiche sandy bottoms. Possibly this was an epicontinental

sedimentologiche dei calcari a « Orthoceras » di sea of Iimited depth so that the ware motion Fluminimaggiore risulta che questi sedimenti si could affect the sea bottom. depositarono in un mare normalmente ossigenato

INTRODUCTION (*) Work supported by the Consiglio Naz. delle Ricerche, grants: 78.01881.66, IGCP Project n. 53: Ecostratigraphy and 78.00308.05. The objective of this study was to try (**) Istituto di Paleontologia dell'Università di to define the environmental conditions that Modena, Italy. controlled the deposition of the "Orthoceras (***) Istituto di Geologia dell'Università di Mo­ dena, Italy. limestone" in Southwestern Sardinia. 406 MAURIZIO GNOLI ET Alli

Both the fauna content and the litholog­ less flattened prismatic blocks surrounded ical and sedimentological characteristics of by non calcareous pelites. These limestone these limestone are taken into consideration blocks are generally not longer than a meter in this report. From the paleontological and their edges are slightly rounded. The point of view we have tned to determine pelites clearly have flown around the hard the environmental significance of the indi­ blocks which are often folded in various viduai taxa as well as that of the entire ways, sometimes very tightly. association. The sedimentological aspects On the basis of what can be observed of the study were directed towards evalua­ in the field, the breaking up of the calcare­ ting the energetic characteristics of the en­ ous layers and the incorporation of the vironment by means of the sedimentary fragments in the silt-clay matrix have been structures. In particular we based our con­ caused by intense tectonic activity. No signs clusions on the èistribution and orientation of deformation of the blocks in the plastic of the orthoconical shells of nautiloids. state were observed not even at the edges. Further information was obtained from The occasionally observable rounding of the lithological characteristics of the rocks. the edges of the limestone blocks, which is not pronounced however, was caused solely by dissolution and disintegration as­ GEOLOGICAL REMARKS sociated with surface weathering. The "Orthoceras limestones" in the A brief summary of the Silurian rocks is studied region are made,up of sparite or given here in order to piace in perspective microsparite having widely variable crystal the "Orthoceras limestone" dealt with in this sizes, even within the same thin section, paper. (generai between 30 and 300 microns). These Silurian marine sediments, which are sparites must be derived from the recrystal­ exposed in two main areas (lglesiente at lization of a very fine calcareous mud in SW an d Gerrei at SE), are represented chief­ which fragments and shells of orthoceratids ly by black shales rich in graptolites. Well - were scattered with a few bivalves and bedded black limestone with some gastropods as described in the para­ ("Orthoceras"), Cardiola, graptolites, gastro­ graph regarding the fauna. pods, brachiopods, ostracods and conodonts The orthoceratid shells have varying occurs- in the upper part of the sequence. sizes from a few millimeters to around 30 On the basis of conodonts, the "Orthoceras centimeters in length for whole specimens. Iimestone" was dated by SERPAGLI (1967, Some fragments of living chambers reach 1971) as Uppermost Wenlockian-Ludlovian a diameter of about 20 centimeters, hence, in age. considering an angle of expansion of approxi­ A complete Silurian stratigraphic record mately 2°, these chambers must have been is unknown in Sardinia since it is impossible part of a 2 meter long shell. to find continuous sequences of fossiliferous In the microspar surrounding the nauti­ marine strata. Furthermore, it is virtually loids, shell fragments of ali dimensions down impossibile to trace individuai beds over to a few microns are present. The overall long distances. Latera! facies changes and aspect is that of a completely random as­ intrusions of magmatic rocks, to which sociation of shells and shell fragments. Con­ mineralized veins are connected, greatly centrati o n of small shells in the "shadows" complicate the geologica! mapping of this of those having greater dimensions were regio n. never observed. Because of the intense tectonic activity Some chemical analyses gave values of and poor outcrops, it is impossible to Ca C03 close to 96-98%, which can certainly measure the thickness of the "Orth.oceras be considered a little in excess given the limestone" with any precision; however, the impossibility of eliminating with certainty overall thickness should be of the order of every spot or veinlet of sparry calcite. X-ray 30 meters. diffraction analyses of the small amount of The "Orthoceras limestones" in the Flu­ insoluble residue, after removal of the or­ minimaggiore area are found as more or ganic fraction, showed the presence of " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARiliNIA 407 quartz, mica, chlorite and/or montmoril­ cumulation of particular organisms, or sedi­ lonite and feldspar (l). ments derived from them, on certain areas Often the calcareous mud partialiy pen­ of the bottom particularly favorable to etrated the shelis, occupying only the low­ the organisms concerned. In a situation er part and decreasing in thickness towards of this type, there should exist a tran­ the apical portion (fig. l). The remaining sition belt in which the favorable con­ upper parts of the shelis are filled by ditions progressively disappear outwards blocky calcite cement. until they no longer exist. In this way lenticular accumulation would be produced whose dimensions are greater by one or SEDIMENTOLOGICAL REMARKS more orders of magnitude to those of the constituent elements. Various other ways of GEOMETRY OF THE "0RTHOCERAS LENSES" accumulation of Ienticular deposits can be hypothesized, ali of which, however, imply The term "Orthoceras lens", used by the the existence of sedimentary structures or first authors that delt with them {TARICCO, characteristics which change from the center 1921-1922, p. 13; GORTANI, 1922, p. 365; No­ to the margins because of the variation in VARESE & TARICCO, 1922, p. 318-319) and genetic conditions. SERPAGLI recently adopted also by (1970) On the contrary, the most evident aspect and SERPAGLI & GNOLI (1977), is purposely in ali of the blocks of "Orthoceras lime­ not used in this artide since the lenticular stone" examined is the uniformity of the form in any sedimentary body implies the characteristics aver the entire external sur­ existence of different environmental condi­ face and in the observable sections. In ad­ tions "between the thicker center and the dition, the observable blocks have sizes of ends. The difference in environmental con­ the same arder of magnitude of the shells ditions, far example, could be a different they contain and in certain cases the di­ distribution of the action of the current mensions of the blocks are less than those between the bottom and the sides of a of the enclosed shells. channel of which the kns represents the ·From these considerations it foliows that filling. lt is important to note here that if the Orthoceras formations must have origi­ "Orthoceras the so called lenses" represent naliy consisted of an alternation of calcare­ orthogonal sections of sedimentary bodies ous layers with pelitic layers that were formed by filling of channels, they should poor or completely lacking in carbonates. be elongated in a particular direction in the Within the calcareous layers the fossils form of ribbon shaped bodies and not be are not uniformly distributed in either the essentialiy equidimensional as they actual­ horizontal or vertical direction. ly are. More evident is that the larger shells are Another situation able to produce lentic­ more densely packed in certain levels, sep­ ular sedimentary bodies could be the ac- arated by variable thicknesses of micrites that are lacking or poor in fossils. A closer The authors thank G. GARUTI, lst. Mi­ (l) Dr. examination shows that the predominantly neralogia and Petrologia, Modena University, for having dune the X-ray dillraction analyses. micritic rock contains thin laminae very

sect. A-A'

F1.�. l - Axial an d cross sect io n of a living chambcr of Kopaninoceras ? thyrsus (BARRANDE) showing the progressive wedging out towards the apex of the muddy infilling, x l. 408 MAURIZIO GNOLI ET ALli rich in small shells or fragments of shells. pebbles or various shells due to the action Such laminae are of relatively reduced of waves on heterogeneous mud bottom. lateral extension, at the most of the order Considering the limited lateral continuity of of 10 cm or slightly more (fig. 2). the laminae, the generai aspect and the ab­ The laminae or layers made-up of medium salute absence of cross laminations even on sized or large shells are much more exten­ a small scale that could indicate the pres­ sive laterally but it was not possibile to ence of currents, the agent that most prob­ determine their actual sizes because of the ably concentrated the remains must scarcity of observable blocks. The lamimie have been waves. formed by the thin concentrations of small The shells and shell fragments, as al­ fossil remains can not be the result of ready said, are in part filled with micrite, simple sedimentation since their lateral ex­ that constitutes the mass of the rock, and tension is too limited but rather must have in part with �parry calcite. · been caused by the reworking in situ of the The filling of the shells by the originai upper part of the sediment itself imme­ calcareous mud, more or less rich in frag­ diately after its deposition. ments, is in generai quite extensive; it some­ In fact, the laminae (or better, layers) times reached the innermost parts of the caused by generai variations in the condi­ sifuncle. tions of sedimentation have a considerable Sometimes it can also happen that whole continuity and extension. One can quote as large shells and small ones are telescoped an example the varves or laminae present together (fig. 3). lt quite frequently occurs in some calcareous layers of the Swabic that even the more fragil parts of the shells, which are continuous, with negli­ such as the living chamber and apical ex­ gible variations in thickness, for various tremity, are well preserved. The living kilometers. (Personal communication from chambers of the larger shells are typically Prof. A. SEILACHER). fractured by the compaction causing a re­ The concentrations of fossils described duction of as much as 30% of the originai here recall to mind the concentrations of thickness. On the basis of their particle

Fig. 2 - Weathered cross section of a slab of "Orthoceras" Iimestone. The plane-parallel arrangement of the laterally vanishing laminae of shell debris is clearly independent from the lenticular outer shape of the edge of the slab, x l. " " P.\LEOECOLOGICAL REMARKS ON THE ORTHOCERAS J.IMESTONE OF SARDINIA 409

Fig. 3- Orthoconic shells telescoped together with the guest shell occupying the living chamber (a) (cross scction, x 5) or the of the host (b) drawing after a photograph, x 1). size, these sediments can be classified as ditions above the bottom other than in the medium-fine silts. Visible traces of burrow­ mud itself. The Eh = O level, must there­ ing were not found in any of the fore have been located within the water of samples examined. This maight be the the sedimentary environment (above the result of two apposite situations: either bottom). burrowing animals were really absent in the As already said, the shells of orthoceratids sedimentary environment or they were pres­ are concentrated in preferred levels, where ent in such large quantities, with respect the shells are definitely ordered in a ho­ to the rate of sedimentation, that there was rizontal or very nearly horizontal position. time for them to homogenize completely Examination of an horizontal section of the sediment. Of these two hypotheses, how­ the orthoceratid plates and blocks shows ever, only the first is valid because of the that the shells are not uniformly distributed presence of the laminae previously discuss­ in the horizontal piane. In some plates they ed as well as of a certain orientation of the are apparently randomly distributed, while shells which will be described later. in others they show a certain orientation is The black colour of the rock which tends developed. to be altered in grey, (more or less light), In sufficiently extensive plates, belts one by oxidation must be due to organic matter found in which the shells are iso-oriented in the absence of iron monosulfides. Organic in alternation with other belts in which the matter (both carbonaceous and bituminous) shells are less clearly iso-oriented and often was found in the insoluble residue. trending perpendicularly to those of the In effect, many of the Orthoceras blocks surrounding area. Considering the relation­ have rusty spots evidently caused by oxida­ ship between the sizes of the shells and the tion of iron sulfides. . Closer observation consistency of their orientation, the shells showed that the sulfides were concentrated that more often have a more preferred in the living chambers near the openings of orientation are those that are more elongated. some shells where in fact, the rock has As a first approximation, this is equivalent actually a brown-yellowish color. This fact, to say those having an average length of which can be explained by the permanence 2-5 cm. of soft parts of the mollusk in many living Another observable characteristic in the chambers at least until the time of the first samples was that the orientation of the burial, proves the existence of reducing con- large shells and shell fragments (those hav- 410 MAURIZIO GNOLI ET ALli

ing a diameter greater than 2-3 cm) deter­ consequence of mechanical action related mined that of the smaller shells around them. to fluid motion. However, what exactly are Altogether, the observations of sufficient­ the types of motion and the precise relation­ Iy extensive piates suggests of a periodic ships between the direction of motion of aiternation of belts in which the fossiis the fiuid and the preferred orientation of were oriented first in a particuiar direction the bodies that have undergone the action and than in a direction either perpendicular are not at ali known with precision. to the first or more or less random. The Some papers (see criticai review in PoT­ existence of such a periodic pattern Ieads TER & PETTIJOHN, 1963) report oniy on the one to think of the effect of an oscillatory measurements made on the oriented shells motion such as wave motion. In that case, and "do not provide any data that helps to the repetition of the belts with the fossils confirm or disprove the direction and the oriented Iongitudinally alternating with those sense of the current inferred as responsible oriented transversally, would be analogous for the particular orientation found. to the crests and the trough of ripple Results of more thorough laboratory ex­ marks. lt is evident that ripples of this periments evidence that various patterns of type would be formed on an incoherent orientation can be produced depending on sandy bottom and not on a bottom con­ the form of the bodies as well as the ex­ sisting of fine mud; however, a priori, it is perimental conditions. In their criticai re­ piausible that rigid bodies with a specific view POTTER & PETTIJOHN (1963, p. 37-40) weight oniy siigthiy greater than that of point out the uncertain value of the re­ water couid be oriented by wave motion in lationships that can be established between such a way as to develop a structure shell orientation and the current that has analogous to ripple marks on the bottom. caused the orientation. Aiso, the width of the belts observed in the More recently, KELLING & WILLIAMS (1967) orthoceratid plates (about 10 cm each, found that on a silty bottom, depending on thus - 20 cm from crest to crest of the the flow conditions of the current, she!ls hypothesized anaiogues to ripple marks) is of Mytilus may be distributed in varying of the same order of magnitude as the ways. ripple marks reported in the literature and BRENCHLEY & NEWALL (1970) affirmed on observed by one of the authors along various the basis of laboratory experiments that, with tracts of the ltalian coast at a depth the present state of knowledge, it is stili too of severa! meters. Aiso, the non negligible soon to arrive at a generai valid conclusion prevailing sense of orientation of cones may regarding the orientation that can be ex­ be considered analogous to the sligth asym­ pected for an association of fossiis or shells. metry which often characterizes the ripple Among the researchers who are involved made by the waves. To check the plausibility in studies of orthoconical shell orientation, of the interpretation proposed here, the KRINSLEY (1960), REYMENT (1971) and LAU· orientation of the shells was studied with BERE (1977) consider oniy the currents ca­ statistica! methods in order to test the pable of orienting the shells on the bottom. existence of prevailing orientations as well LAUFELD (1974) expiiciteiy considers wave as the objectivity of the subdivision into motion as a source of perturbation of the distinct belts. orientation caused by currents. Aiso KEY In addition, a series of Iaboratory ex­ (1954) probably refers to wave motion when periments on the effect of wave motion on comparing the orientations of the Ordovi­ eiongated bodies slightly heavier than water cian orthoceratids of St. Joseph's Island to were carried out. that of water-soaked or semi-floating pieces of wood that are distributed respectively THE PROBLEM OF NAUTILOID ORIENTATION parallel and perpendicular to the coast in lakes. Symmetric and asymmetric scattered In 1953 PETRANEK and KOMARKOVA held that elongated bodies showing a preferred orienta­ perpendicular wave motions, other than tion are relatively common in clastic sedi­ perpendicular currents, could be the cause ments. The preferred orientations are the of two perpendicular directions of Palaeozoic " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINIA 411 orthoceratid shells of the Bohemian Basin. some specimens of Turritella collected BALASHOV ( 1965) considered both w ave mo­ from the Adriatic Sea and some Ceritium tio n and currents as being responsible far were used. Their dimensions were about an orientation along a single direction. 30 mm in length with a maximum diameter Hence, from this examination of the of 8 mm. literature it results that (i) a considerable In the course of the experiments it was uncertainty persists regarding the exact re­ necessary to eliminate interference pro­ lationship between orientation of shells and duced by waves reflected from the two ex­ the motion of the water that acts an them treme walls of the basin both at the back and (ii) the fundamental cause of the orienta­ of the generating plate and in front of it. tion of shells is considered more or less The waves generated at the back of the plate to be the action of currents and only in were sufficiently neutralized with a very exceptional cases that of wave motion. permeabile and irregular bar made-up of a lt also seems that up to now the cases small accumulation of very angular rock considered are always those in which the fragments. shells have more or less the--- same type of The disturbances by the waves reflected orientation in all the surfaces examined. from the end wall of the thank were However, the orthoconids of Fluminimag­ neutralized by shifting a movable wall to giore are sparsely distributed in a discon­ make the lenght of the experimental tank tinuous way, specifically in parallel belts in multiple of the wave-lenght used. Thus, the which the elements tend to be alternatively experiments could be carried aut with stand­ distributed according to two orientations ing waves. The lenght of the waves used which are discernably perpendicular to each was 22 cm and the period 0.75 seconds. other. Also, some experiments were carried aut with progressive waves. Even though it was not possible to completely eliminate distur­ LABORATORY EXPERIMENTS bances caused by reflected waves, the results obtained were very similar (fig. 4d). In arder to experimentally study the ef­ As can be seen from figures 4b and 4c, fect of wave motion on elongated bodies the waves perceptibly oriented the barley lying an the bottom, a simple system was and turritellids parallel to the axes of the built to produce waves with controlled ripples in correspondence to their crests as characteristics in a small basin. well as in a preferred direction perpendic­ A tank of galvanized sheet iran was built ular to the crests in the troughs between with dimensions 200 X 25 cm and a depth them. of 15 cm. The experiments were carried Some differences can be seen when the aut with 6.5 cm of water in the basin. distributions of the grains of barley and the Waves were produced by moving a vertical shells of Turritella are compared with that plate back and forth in the tank. The plate of the « Orthoceras » shells (far example, was 10 cm high and 24.5 cm wide and im­ fig. 7). In the studied rocks the belts made­ merged in the water unti! a few mm of the up of the shells perpendicular to the belt bottom. To obtain the required motion, an itself are more evident. Probably, this electric engine was used to drive the orientation corresponds to the furrows lo­ vertical plate by means of a suitable handle cated between the crests of the ripples. and reducing gear. A meter to count the This difference is probably caused by the rotations was inserted between the reduc­ different characteristics of the substrate an ing gear and the plate so that the number which the shells were found in nature (fine of waves used far each experiment could soft mud) and that used far the shells of be calculated. Turritella and the grains of barley (smooth Grains of barley, 8 mm long and 3 mm sheet iran). lt is also worthwhile t o keep in diameter, were used as the elongated in mind that given their different size, the bodies. These were found to be suitable as "Orthoceras" shells should have a much regards both shape and specific weight. greater mutuai interaction in the belts where For the experiments with conica! bodies, they accumulated parallel to the structure 412 MAURIZIO GNOLI ET ALli " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINIA 413

Fig. 4 - Laboratory experiments in a tank. a) Barley seeds and turritellids randomly scattered on the smooth bottom of the tank. b) The same as above after the action of 100 waves. Rhythmic concentra­ tions of fine debris washed out from seeds and shells mark the crests of the standing waves (lenght 22 cm). c) The same as above with barley seeds only. d) Experiment with progressive waves (approximatively the same waves lenght). e) Turritellids randomly scattered on sandy bottom (medium grained sand). f) The same as above after the action of 100 waves. The position of the crests of the standing waves is shown by markers. Note that after wave action turritellids and seeds tend to be more densely packed beneath each crest of the waves where they are parallel to the crest itself. In the area between the crests, shells and seeds appear unoriented or a t right angle, with the waves.

much more than would be the case for the within each single sample were considered. grains of barley. The reason for this was that the intense It is necessary at this point to emphasize tectonic dislocations which completely broke­ that these experimental studies were carried up the originai calcareous strata made it out with very simple equipment and that impossible to reconstruct the true geo­ the objective was not to obtain quantitative graphic orientation of the samples. data regarding either the depth to which a In each sample the azimuth of the growth wave of a particular length can act or the axis of the was measured with relationship between wave length and size respect to an arbitrary selected direction of the ripples and oriented bodies. The only of reference. These data were taken from correctly utilizable information from these the smooth surfaces of 3 plates coming experiments is that wave motion is actually from the localities Perd'e Fogu, Galemmu capable of orientating elongated cyndrical and Corti Baccas respectively, ·ali in the or conica} bodies according to a periodic Fluminimaggiore area. The surfaces exam­ pattern of alternating belts, within which ined were cut parallel to the strata and had the elongated bodies are distributed in a characteristic way. Such a distribution of conica} bodies is schematically represented --+ w ave in fig. 5. With conica} bodies of the same --+ motion size and shape, uniformly scattered on the bottom surface, a rose diagram re­ lated to a limited portion of the total � ...... ' l' field may assume the shape of one of the � diagrams represented in fig. 5. The shape ...... of an analogous rose diagram will be a Iittle l ' l more irregular and asymmetric when the conica! bodies are different in size, not ' l ' exactly at right angle each other, not uni­ formly scattered on the bottom surface and l possibly with a slightly prevailing sense due l ' to any perturbance whatever. The arms of the cross or the modal classes of the rose ' l ' diagrams must be in any case either parallel or perpendicular to the belts if they do exist. l ' l

0RIENTATION ANALYSIS OF THE ORTHOCONIC NAUTILOIDS i Statistica} analyses were employed in I arder to verify the existence and width of

the belts cointaining fossils seen to be al­ Fig. 5 · Schematical pattern of distribution of conic­ ternately oriented in the two preferred di­ a! bodies oriented by wave action. This figure shows how the shape of each rose diagram is con­ rections. In these analyses only the recip­ trolled by the size and orientation of the sampled roca} orientations of the orthoconic shell.� area. MAURIZIO GNOLI ET ALli 41 4 dimensions of some decimeters squared. The the standard deviation) which is related to fossil remains, of sizes aver a centimeter, the value of r by the following equation: had a density of about 10 specimens per 100 cm2• s = 2(1-r) Using the method of REYMENT (1971) for v quantitative analysis, the growth axis of each was considered as a unit The values of s, r and a mean for the vector whose origin was the apical part of three samples examined are given in table l. the orthocone. The following expression In arder to verify that the environmental gives the components of the mean vector characteristics did not change with time, in a series of observations: measurements on samples 2 and 3 were made on two different levels of stratification about 2 centimeters apart (2', 2", 3', 3"). To l N l N x = -- � cos CXi ; y = -- � sin CXi further verify the influence of the fossil size N N i=t i=t on orientation, specimens in sample l were analysed separately according to length where N = the number of fossils per sample (class la, � 20 mm; class lb > 20 mm). and (i = l, ... N) are the angles between cxi The circular distribution diagrams for a reference direction and that of the growth the shells in the plane examined are given axis of the orthocones. for each sample in fig. 6. In polar coordinates, the mean vector is The following considerations are suggest­ as follows: ed by the data of Table l: a) With time, the orientation of the fossils varied little y (26") or remained practically constant, at r = x2 + yz tan a v x least for the 4 periods represented by the two faces of the available plates; b) The The dispersion of a series of angular mea­ orientation of the fossils was not influenced surements can be expressed by mean of in any determined way by their size or their the mean angular deviation, s (analogous to length.

Fig. 6 - Rose diagrams of fossi! nautiloid orientations. l, sample n. l from Perd'e Fogu; la the same sam­ ple but taking in account only specimens � 20 mm in length 2'-2" upper and lower surfaces of sample n. 2 from Galemmu; 3'-3" upper and lower surfaces of sample n. 3 from Corti Baccas. " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINIA 415

TABLE l were divided into parallel strips, 5 cm wide,

Mean angle and angular deviation; N = number of perpendicular to the main moda! direction specimens in each sample; so = mean angular pointed out by the rose diagrams of fig. 6. deviation; r = mean angular vector; eo = mean The width of the strips was chosen as to angle. (See text on page 414). contain a sufficient number of specimens. In this analysis only specimens with length Sample o N s r more of 20 mm were used. For each strip the mean value of the l 101 60° 0.45 cosines of the angles !X; between the direc­ tion of each specimen and the main moda! la 59 65 o 0.34 direction was calculated, obtaining a certain value cos !X;. This value was used to draw lb 42 50° 0.61 the diagram of figure 7, for sample l. This diagram results a discontinuous line with peaks and lows situated at quite regular l 7 2 61 5 ° 0.50 distances, corresponding to the belts form­ ed by groups of strips in which the mean 2" 43 46 0.68 values of cos a either greater than or less

3 l 38 65° 0.35 TABLE 2 3" 40 63° 0.38 Randomness tests or "Rayleigh tests". (See text on this page).

Random orientation Sample N z From an examination of the polar dia­ hypothesis supported grams of figure 6 one sees that almost all the specimens are distributed in a semi­ 101 18.4 no 1% piane. The existence of a preferred direc­ la 59 7.3 no 1% tion, with a relatively limited dispersion, is particularly evident in the graphs relative lb 42 15.9 no 1% to samples l, la, lb and 2'. A statistica! verification of the non ran­ 2 r 61 15.1 no 1% domness of the distribution is supplied by 2" 43 20.1 no 1% Rayleigh's test if the hypothesis of a poly­ l modal distribution is excluded. When ap­ 3 38 4.8 no 1% plied to our data, this test, amply illustrat­ 3" ed by REYMENT ( 1971 ), gave the results re­ 40 5.8 no 1% ported in table 2 and confirms a non random distribution in all the samples. TABLE 3 In order to give more statistica! signif­ Tests for uniform distributions or "x·square tests" icance to a distribution of this type, the (see t ext o n this page). x2 test was applied. As seen from the values given in table 3, the hypothesis of a uniform Degrees 2 Hypothesis of uniform Sample N x distribution is significantly discarded. The of freedom distribution rejected leve! of significance falls to 5 to 10% only lO l ll 90.2 yes for sample 3. The comparative analysis of l% this slightly anomalous data, along with the la 59 26.1 yes l% unclear unimodality seen in figure 6 for lb 42 28.6 yes l% semples 3', 3" and 2', led us to verify the 2' 61 ll possibility of a distribution in homogeneous 39.1 yes l% zones having differing average directions 2" 43 ll 87.3 yes l% and density. For this purpose, the follow­ 3' 38 ll 17.6 yes 5-10% ing empirica! procedure was used: The sur­ faces of the samples considered (l and 3") 3" 40 ll 26.0 yes l% 416 MAURIZIO GNOLI ET ALli

than 0.71 = cos 45°, which represents the than cos 45°, the fossils are predominantly

mean theoretical value of ali the cos rxi oriented along the main modal direction. values. In fact, fossils which are orientated Whereas, in the belts characterized by a at 45" are in an intermediate position mean value of cos a less than cos 45•, the between parallelism and perpendicularity to fossils are predominantly orientated in a the main modal direction. In the belts direction perpendicular to the main modal caracterized by mean values of cos a greater direction. As seen from fig. 7, this analysis

10 cm

• modal directton

0.9 parallelism t 1ncrease 0.8

0.6 l �erpendicularity ' �ncrease 0.5

10 20 30 40 cm

Fig. 7 - Fossi! distribution in the sample l. Diagram helps to emphasize belts where fossils tend to be alternately parallel and perpendicular to main moda) direction, (see text on page 415). Rose diagrams are related to fossils distribution in each corresponding belt. " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINJA 417 demonstrates the existence of 10-15 cm wide mostly represented by orthoconic species belts in which fossils are alternately orient­ with no ornamentation other than growth ed in directions parallel or perpendicular lines. Transversally (Hemicosmorthoceras, to the belt itself. Although the size of Plagiostomoceras) or longitudinally (Kiono­ sample 3 is relatively small, the data ob­ ceras) ornamented forms are also present tained are in complete agreement with that as well as rare cyrthoceraconic species. of sample l. Sample 2 was not used for Such animals may have been ;nektic or this type of analysis since its small surface nektobenthic in their mode of life, inha­ area was not sufficient to guarantee signif­ biting surface waters. According to WATKINS icant results. In the lower part of fig. 7, & BERRY ( 1977) Silurian cephalopods respond the rose diagrams relative to the orienta­ to the same hydrographic controls affecting tion of the fossils in each single belt are graptolites because, for instance, in the given. In the belts where fossils tend to Ludlow series of Wales and the Welsh be parallel to the main modal direction, a Borderland their large scale density trends certain polarity is present. directly parallel those of graptolites.

Aptycopsis. - These problematic fossils GENERAL FEATURES OF THE FAUNA of Silurian age belonging to the genus Aptycopsis Barrande, 1872, have been con­ The only fossils until now recorded in sidered from time to time to be carapaces the Silurian black shale sequence of south­ of phyllo'pods, carapaces of phyllocarids or western Sardinia come from black lime­ more recently, opercula of nautiloids (Tu­ stones interbedded with shales. A few REK, 1978). In any case a nektic or planktic graptolites were also recorded from the rpode of life can be supported. underlying shales (TARICCO, 1920-1921) but Phylloceratids species in association with never illustrated. graptolites already have been supported by A total of about 100 taxa were distin­ JONES & WOODWARD (1888-1889) and WATKINS guished at varying taxonomic levels. A rank­ & BERRY {1977). ed list of the most abundant taxa in the Conodonts. - They are not very good "Orthoceras limestone" is presented in tools to reconstruct ancient eenvironments table 4. because there stili exists disagreement con­ Cephalopods followed by graptolites cerning the originai mode of life of the dominate the fauna in terms of abundance. conodont animai. In fact, two different Cephàlopods are also dominant in terms of ecologie models have been proposed to ex­ taxonomic diversity, being represented by 38 taxa compared to only 3 for gastropods. plain the distribution.patterns of conodonts, the pelagic depth stràtification model and To evaluate the meaning of this statement it must be, however, kept in mind that the nektobenthic mode!. According to KLAP· cephalopods, conodonts and ostracods are PER & BERRICK (1978), who have recently debated the problem, it is unlikely that the only well studied groups of the as­ conodonts were infaunal or sessile epi­ semblage while no taxonomic treatment faunal benthos because there is not much exists for graptolites, bivalves, gastropods congruence between lithofacies and cono­ and brachiopods. Toghether with graptolites, which were dont biofacies. Also ALDRIGE (1976, p. 102) recently noted that there is no evidence that a major group of zooplankton in Silurian conodonts fitted into the structure of true seas, organisms of nektic habit as well as . communi,ties. the probable alternafives others possibly attached to floating plants So are that they were nektobenthic or neritic occur in the assemblage. pelagic animals (either planktic or nektic).

NEKTIC ORGANISMS ASSOCIATED WITH THE GRAPTOLITES 0RGANISMS POSSIBLY A TT ACHED TO FLOATING PLANTS Cephalopods. - They are apparently the more common fossils of ali the assemblage The vast majority of bivalves, gastropods (38 taxa · SERPAGLI & GNOLI, 1977) and are and brachiopods occur as molds or casts

27 Graptolites Bi val ves

"Monograptus priodon" * CaJ'dium subarcuatum

* Cardium sp. ind. Brachiopods * Avicula cf.semiauriculata * Avicula sp. ind. Merista passa · Cardiola bohemica * ?"Orthis" sp. Cardiola gibbosa * ?"Pentamerus" sp. Cardiola interrupta * ?"Atrypa" sp. Maminka sp.

* Lunulicardium sp. Cephalopods Slava sp. Hemicosmorthoceras laterculum Dualina sp.

Hemicosmorthoceras semiannulatum Spanila $P· Hemicosmorthoceras sp.

Kopaninoceras jucundum Gastropods Kopaninoceras ? thyrsus Platyceras sp. currens Platyostoma sp. Michelinoceras grande Spirina sp. Michelinoceras sp.l * ? "Pleurotomaria" sp. Michelinoceras sp.2 Michelinoceras sp.3 Ostracodes Plagiostomoceras gruenewaldti

Plagiostomoceras cf. pleurotomum Beyrichia reticulata "Parasphaerorthoceras" sp.B s. Ristedt Aparchites pygmaeus "Parasphaerorthoceras" sp.H s. Ristedt Aparchites grecai "Parasphaerorthoceras" sp.J s. Ristedt Richteria lamarmorai "Parasphaerorthoceras" sp.K s. Ristedt Entomozoe meneghinii Protobactrites ? sp. Entomozoe ichnusae "Sphaerorthoceras" beatum Entomozoe zoppii "Sphaerorthoceras" curvum Entomozoe subreniformis "Sphaerorthoceras" teicherti Entomozoe (?)amygdaloid�s "Sphaerorthoceras" sp. Entomozoe (?)parvula .� /rrosphaerorthoceras gregale Entomozoe sp. ind. fin

Merocycloceras declive * Cypridina tirrenica doricum Balbozoe (?)bohemica Parakionoceras originale Balbozoe (?)italica Arionoceras affine Balbozoe (?)capellini Arionoceras submoniliforme Balbozoe (?)lanceolata Columenoceras sp. cf. C. columen

Harrisoceras vibrayei Conodonts Metarmenoceras ? meneghinii Acontiodus obliquicostatus "Orthoceras" cf. richteri Hindeodella equidentata Oocerina abdita Spathognathodus incl.inclinatus Oonoceras plebeium Ozarkodina media Galtoceras ? sardoum Trichonodella excavata Michelinoceras subconoideum Plectospathodus extensus "Parasphaerorthoceras" ? subcyprium Neoprioniodus excavatus * "Orthoceras" bronni Kockelella variabilis * "Orthoceras" dulce Ozarkodina ziegl. ziegleri ·* "Orthoceras" transiens Trichonodella inconstans * "Orthocl'!ras" p()tens Lonchodina greilingi "Parasphaerorthoceras" sp.F s.Ristedt Lingonodina salopia Hemicosmorthoceras sp. s. Ristedt Lingonodina silurica *Ophidioceras sp. Neoprioniodus multiformis Spathognathodus sagitta Incertae sedis (probably cephalopods) Ozarkodina edithae Aptychopsis sp. Ozarkodina fundamentata Spathognathodus primus Ligonodina sp. TABLE 4 List of species recorded from Middle-Upper Silurian "Orthoceras" Iimestone. Data from MENEGHINI ( 1857, 1880), CANAVARI (1899), TEICHMVLLER (1931), GORTANI (1922 ), COMASCHI-CARIA (1949), SERPAGLI (1967, 1971), RISTEDT (1968), GNOLI & SERPAGLI (1977), SERPAGLI & GNOLI (1977). Specimens preceeded by (*) are of un­ certain determination. " " PALEOECOLOGJCAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDJNJA 41 9

and in nearly ali cases, features of the shell Brachiopods. - A few articulate bra­ exterior have been impressed into the in­ chiopods belonging to the genera Merista, ternai mold, and muscle scars and hinge Orthis, "Pentamerus " and "Atrypa " some­ structures have been obliterated. This is times occur in the assemblage. An epi­ possible only in the case of relatively thin planktic mode of life for small brachiopods shells; this characteristic acquires a par­ has been suggested severa! times either for ticular importance for its paleoecological forms ( AGER, 1962) or for Palaeozoic significance. ones (ZIEGLER et al., 1968; THAYER, 1974). Further reports of brachiopods which may Bivalves. - Eight to ten species of prae­ have lived attached to seaweed, are due, cardiacean bivalves, among which members among others, to RUEDEMANN (1934 ), SPJELD­ of the genera Cardiola, Slava and Dualina NAES (1967) and HAVLICEK (1967). have been recorded in consistent associa­ tion with cephalopods and graptolites. Al­ Ostracodes. - Twenty species of ostra­ though never abundant, these species occur codes have been reported by CANAVARI (1899) in an sections of the studied area and seem from about 2, 100 kilograms of rocks. The to be characteristic throughout Silurian fauna, which needs a revision, is formed by deposits of the British Isles, continental members of the genera Aparchites, Richte­ and North . This distribu­ ria, Entomozoe, Pseudoentomozoe, Balbozoe, tion pattern supports evidence for a pelagic Kloedenia and "Cypridina". Marine ostra­ mode of life among praecardiaceans. Ac­ codes were mostly crawlers, burrowers and cording to WATKINS (1978, p. 47) who care­ near-bottom swimmers but it is very likely fully studied bivalve ecology in the Silurian that some forms may have had an epiplank­ of the Welsh Borderland, this type of as­ tic mode of !ife. In fact (fide BENSON, 196 1. sociation occupied a quiet-water, deep shelf p. 58) members of the genus "Cypridina", position where the oxygen conditions were as well as a few other fossi! forms, were apparently low. predominantly pelagic. Also ELOFSON (1941 ), studying Holocene marine ostracodes from Also BABIN & ROMBARDET ( 1974, p. 41) agree with an epiplanktic mode of !ife fol­ , recovered some pelagic forms. lowed by a simple embedding into the mud­ dy bottom of a low turbulence environment As seen from the paleoecological signi­ for a bivalve assemblage in a similar lime­ ficance of the fauna examined here and as stone ( « Ampelites >> with graptolites, ostra­ already outlined by SERPAGLI & GNOLI (1977, codes, cephalopods, etc.) from Upper Silur­ p. 157), typically benthic fauna is completely ian of Normandy. In that assemblage, be­ lacking in the assemblage; ali the mention­ sides praecardiaceans, ambonychiaceans, pte­ ed fossils can be interpreted as pelagic riaceans and modiomorfoceans also are pres­ organisms. ent. Finally, it is interesting to note that Good preservation of fossils belonging severa! specimens of Cardiola have b.een to a broad spectrum of forms further sug­ recorded and illustrated by BARRANDE ( 1868, gests lack of scavenging animals and boring 1870) fixed to the shells of nautiloids (pl. sponges and algae, or strong aerobic bacter­ ial decay of the organisms lying dead on 372, fig. l; pl. 396, fig: 16; pl. 313, fig. 3; pl. 343', fig. 17). the sea floor represents an additional evidence of toxic conditions above the Gastropods. - Although no identifiable bottom. alga! remains were found, it is highly prob­ To conclude, we can assume that the able that the few small gastropods of the Upper Silurian shelly fossils lived as epi­ assemblage were attached to floating plants, bionts to other pelagic animals or as epi­ where these herbivores would also be able plankton on seaweed which subsequently to find their food supply. Members of the settled to the ocean floor. Although no genus Platiceras have been already report­ identifiable vegetable matter was found, it ed as epibionts; BROOKES, KNIGHT et al. (1960, seems to be very likely, that the high con­ p. 240) found them attached to crinoid centration of organic matter found in the calyces. limestone is at least partly related to floating 420 MAURIZIO GNOLI ET ALli plants. The habits of this epiplanktic as­ shell fragments of various sizes concentrat­ semblage are tentatively reconstructed in ed along distinctly plane-parallel preferred figure 8. surfaces can be interpreted as clastic en­ richments related to the wave action on a muddy bottom rich in fossil remains.

ENVIRONMENTAL CONCLUSIONS Also, wave motion is the only agent ca­ pable of moving the sediments on the bot­ An important conclusion that can be tom without causing an overall remixing drawn from the observations and discussion of the water of the basin and hence with­ presented in this paper is that the shells out causing an enrichment of oxygen in the from the Fluminimaggiore "Orthoceras lime­ lower strata, thereby maintaining euxinic stone " are incorporated in evenly bedded conditions near the bottom. and not lenticular shaped limestone bodies. An oscillatory wave motion also is the Sedimentary bodies attributable to accu­ only agent capable of causing the small mulation by current action. Equally in­ shells to deeply penetrate the larger ones dicative of the lack of currents are the and of settling the shells and shell frag­ sedimentary structures observable on a ments in a tight packing without causing small scale and the lithological character­ any sorting. istics; also rule out current, the fine particle In addition, the orientation of the ortho­ size, the large quantity of organic matter cones and their fragments in the limestone (both carbonaceous and bituminous) along is very similar to that produced by waves with the absence of any type of current in an experimental tank and the distribu­ bedding are ali characteristics of a tranquil tion of the graptolites recalls that observed environment. in segments of reeds oriented by wave mo­ The discontinuous laminae, sometimes tion near the beach of a lake (fig. 9). present in the limestone, made-up only of The sea in which the studied rocks were

Fig. 8 · Hypothetical reconstruction, not to scale, of the "Orthoceras" limestone environment with schematical representation of the orientation of the shells on the bottom. " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINIA 42 1

deposited must have been practically "Orthoceras limestone " also contains repre­ devoid of terrigenous supply with the ex­ sentatives of various other types of or­ ception of slight amounts of fine silt and ganisms and is very rich in specimens. clay as evidenced by the almost pure lime­ Since specimens of every size are present, stone whose impurities are essentially re­ indicating that death occured at ali stages presented by carbonaceous and bituminous of development, it is reasonable to assume organic matter. The high content of organic that the taphacoenosis almost entirely rep­ materia! and the complete absence of bio­ resents the biocoenosis without successive turbation indicates that the depositional en­ enrichments or impoverishments of some vironment must have been anoxic. species on the basis of their size or shape. The presence of iron sulfides near the Ali the organisms represented by the openings of various shells is a further in­ fossi! remains had without doubt a pelagic dication that the alteration of the soft parts mode of life or at least can very reasonably must have occurred in strictly euxinic con­ be considered as such. Completely lacking ditions. are remains of exclusively benthic organ­ As previously stated, the fauna of the isms, both epibionts and endobionts.

Fig. 9- Rhabdosomes of Monograptus from the "Orthoceras" limestone of Fluminimaggiore (a) showing a pattern close to that of an accumulation of segments of reeds in shallow water near the beach of Vico lake (Centrai ltaly) (b). 422 MAURIZIO GNOLI ET ALli

Finally, ali evidence leads to the con­ ALDRIDGE R. J. (1976) - Comparison of macrofossil clusion that the organisms from the "Ortho­ communities and conodont distribution in the British Silurian. Geo!. Assoc. of Canada, Special ceras limestones " of Fluminimaggiore lived Paper 15, 91-104, 3 text-figs, Toronto. in a normally oxygenated sea rich in pelagic BABIN C. & ROMBARDET M. (1974) - Mollusques Bi­ fauna in its upper parts and fell after death va/ves du Silurien supérieur et de l'extréme in the euxinic environment near the bot­ base du Dévonien en Normandie. Annales de tom where their remains accumulated and la Soc. Geo!. du Nord, 94, 19-45, pls. 1-6, Lille. BALASHOV from time to time were weakly agitated by S. G. (1968) - endoceratoids of URSS. Leningrad University Press, 1-276, wave motion. pls. 1-51, 23 text-figs (in Russian), Leningrad. One can immagine such a sea to be BA NDEL K. ( 1972) - Palokologie und Palaegeographie analogous, for example, to the present Black im Devon und Unter-Karbon der zentralen Kar­ Sea. lt is not possible to make a quanti­ nischen Alpen. Palaeontographica Abt. A, 141, 1-117, pls. 1-10, Stuttgart. tative evaluation of the rea! depth of such BANDEL K. (1974) - Deep-water limestones from the a sea since the depth at which wave action - of the Carnic Alps, Au­ can have an effect depends on the wave stria. Spec. Publs. Int. Ass. Sediment. l, 93-115. size and this, in turo, depends on the fetch BARRANDE J. (1865-1877) - Systeme Silurien du cen­ of the winds which generate them, and fair tre de la Boheme: l. Reach. Pal., v. 2, Classe versus storm conditions, and both of these de Mollusques, Ordre des Céphalopodes; l' ser., pls., 1-107 (1865); ZO ser., pls. 108-244 (1866 ); 3° ser., parameters are unknown. However the en­ pls. 245-350 (1868); 4° ser., pls. 351-460 (1870); vironment in which the orthoceratids had suppl. et sér. tard., pls. 461-544, Texte III (1870); been living can be immagined as an epicon­ Texte IV (1877); Texte V (1877); 4o ser., Distri­ tinental sea of limited depth. bution des Cephalopodes ( 1870), Prague. BENSON H. Considering a thickness of the order of R. (1961) - Ecology of Ostracode As­ semblages. Treatise on Invertebrate Paleontol­ twenty to thirty meters for the calcareous ogy, 56-63, Geo!. Soc. of America and Univ. of sequence and its uppermost Wenlokian-Lud­ Kansas Press. lovian age (time lap se corresponding to BERRY W. B. H. & Boucor A. J. (1967) - Pelecypod­ around 15 million years), the velocity of Graptolite association in the 0/d World Silurian. Geo!. Soc. America Bull., 78, 1515-1522, l fig., sedimentation must have been of the order Boulder. of l Boubnoff (l IJ./year). This velocity BRENCHLEY P. J. & NEWALL G. (1970) · Fiume ex­ (FISHER, 1969) is at the lower known limit periments on the orientation and transport of for calcareous sediments in shallow water. models and shell va/ves. Pal. Pal. Pal., 7, 185-220, Aside from the imprecision in the evalua­ 12 figs., Amsterdam. CANAVARI tion of the total thickness, which could be U. (1899 ) - La fauna dei calcari nerastri con Orthoceras e Cardiola di Xea S. Antonio in underestimated, this very low rate really Sardegna. Paleontogr. ltalica, 5, 187-210, 2 pls., depend on the fact that the carbonate was Pisa. derived only from nektic organisms with COMASCHI CARIA l. (1949 ) · l fossili della Sardegna. contributions from benthic fauna and flora Istituto per gli Studi Sardi, 1-410, Cagliari. lacking completely. ELOFSON O. (1941) - Zur Kenntnis der mariner Ostra­ Finally, it must be pointed out that the coden Schwedens, mit besenderer Berncksich­ tigung des Skagerraks. Uppsala Univ. Zoo!. Bi­ present contribution support the hypothesis, drag., 19, 215-534, Uppsala. (1967 ), already supposed by BERRY & BoucoT FISCHER A. G. (1969) - Geologica/ time-distance rate; HOLLAND (1971 ), VAI (1972, 1979), of a shelf The Bubnoff Unit. Geo!. Soc. Amer. Bull., 80, sea upon a stable structural platform al least 549-552, l fig., Boulder. during the second half of the Silurian. GNOLI M. & SERPAGLI E. (1977) - Silurian cephalopods of the Meneghini collection (1857). Boli. Soc. Paleont. Italiana, 16, (2), 137-142, 2 text-figs, 1 tab., Modena. Manoscritto consegnato il maggio 18 1979. GORTANI M. ( 1922) - Osservazioni sul P aleozo ico del­ Ultime bozze restituite il 6 giugno 1980. la Sardegna. Boli. Soc. Geo!. Italiana, Il, 362-369, l tab., Roma. GREENWOOD J. A. & DuRANO D. (1955) - The distribu­ tion of length and components of the sum of REFERENCES N random vectors. Ann. Math., Statist., 26, 233-248, Baltimore. V. AGER D. (1962) - The occurrence of pedunculata HAVLICEK V. (1967) - Brachiopoda of the Suborder brachiopods in soft sediments. Geo!. Mag., 99, Strophomenidina in Czechoslovakia. Rozopravy 184-186, Oxford. Ustred Ust. Geo!., 33, 1-235, Prague. " " PALEOECOLOGICAL REMARKS ON THE ORTHOCERAS LIMESTONE OF SARDINIA 423

HOLLAND C. K. (1971) · Silurian faunal provinces? in RUEDEMANN R. (1934 ) · plankton of North MIDDLEMISS F. A., RAWSON P. F. & NEWALL G. America. Geol. Soc. America Mem., 2, 1-106, Wa­ (eds.) Faunal provinces in space and lime. shington. Geol. Journ. Sp. Is., 4, 61-67. ScHWARZACHER W. (1963) · Orientation of crinoids by JONES WOODWARD T. R. & H. (1888-1889 ) · A mono­ current action. J. Sed. Petrol., 33 (3), 580-586, graph of the British Palaeozoic Phyllopoda. Pa­ figs. 1-5, Tulsa. laeontogr. Soc. Monographs, 1-208, London. SERPAGLI E. (1967) · Prima segnalazione di Conodon­ KAY M. (1945 ) · Paleogeographic and palinspastic ti nel Siluriano della Sardegna e relative osser­ maps. Bull. Am. Ass. Petrol. Geol., 29, 426450, vazioni stratigrafiche. Accad. Naz. Lincei, Rend. 12 figs. Menasha. Classe Sci. fis. matem. e naturali, fase, 6, ser. 8, KELLING G. & WILLIAMS P. F. (1967 ) · Fiume studies v. 42, 856-858, Roma. of the reorientation of pebbles and shells. SERPAGLI E. (1970) · Uppàmost Wenlockian-Upper Journ. of Geol., 75, 243-267, 19 figs., Chicago. Ludlovian (Silurian) Conodonts from Western KINDLE · E. M. (1938) A pteropod· record of current Sardinia. Boli. Soc. Paleontol. Italiana, 9 (1), direction. J. Paleont., 12, 5't5:St6, l fig., Tulsa. 76-96 ,pls. 21-24, text-fig. l, Modena. KLAPPER G. & BARRICK J. E. (1978) · Conodont ecology: SERPAGLI E. & GNOLI M. (1977 ) · Upper Silurian Ce­ pelagic versus benthic. Lethaia, 15-23, 3 text­ 11, phalopods from Southwestern Sardinia. Boli. figs, Osio. Soc. Paleont. Italiana, 6 (2), 153-196, pls. 1-9, KRINSLEY 1 D. (1960) · Orientation of orthoceracone text-figs 1-13, tab. 1-3, Modena. cephalopods a t Lemont, Illinois. J. Sedimentary SHEELIAN P. M. (1977 ) . Ordovician and Silurian Petrol., 30, (2), 321-323, l text-fig., Tulsa. brachiopods from graptolitic shales and related LAUFELD S. (1974) · Preferred orientation of ortho­ deep-water argillaceous rocks. Lethaia, 201- conic nautiloids in the Ludlovian Hemse Beds 10, 203, Osio. of Gothland. Geologiska Forhandlingar, 96, 157- SPJELDNAES 162, Stockholm. N. (1967 ) · The palaeogeography of the Tethyan region during the Ordovician: in Aspects LouBERE P. (1977 ) · Orientation of orthocones in the of Tethyan Biogeography Syst. Assoc. Pubi., 7, English Lake "Dis trict based on field observa­ 45-57, London. tions and experimental work in a fiume. J. Sed. Petrol., 47 (1), 419427, figs 1-2, Tulsa. TARICCO M. (1920-1921) · Sul Paleozoico del Fluminese MENEGHINI G. (1857) · Paléontologie de l'Ile de Sar­ (Sardegna). Boli. R. Com. Geol. d'Italia, 48 (6), daigne (in LAM.�RMORA A., Voyage en Sardaigne). 1-21, Roma. 1-584, A-H , lmprimerie Royal Turin. TEICHMiiLLER R. (1931) · Zur Geologie d es Tyrrenis MENEGHINI G. ( 1880) · Nuovi fossili siluriani di Sar­ Gebietes. Abh. Gesellschaft d. Wiss. Z. Gottin­ degna. Reale Ace. Lincei, S�;rie 3° Mem. Classe gen, Matem.-Phys. Klasse, III F, 3 1-124, 3 pls., se. fis. mat. e naturali, 5, 1-13, l tav., Roma. 47 text-figs, Berlin. MILLER R. L. & KAHN J. S. (1962) · Statistica[ analysis THAYER C. W. (1974 ) · Marine paleoecology in the in the geologica[ sciences. pp. 1-483, J. Wiley & Upper Devonian of New York. Lethaia, 7, 121-155, Sons Inc., New York, London. 24 text-figs, Osio. NAGLE J. S. (1967 ) · Wave and current orientation TuREK V. (1974) · Povtdmky k sedimentaci "orto­ of shells. J. Sed. Petrol., 37, (4), 1124-1138, 7 figs, cerovych vdpenctl" Barrandienu. Casopis pro Tulsa. mineralogii a geologii, ro�. 19, 165-169, l text-fig. PETRANEK KoMARKOWA (1953) · J. & E. Orientace schrà­ TUREK V. (1978) · Biologica[ and stratigraphical si­ nek hlavonozcu ve vapencich Barrandienu a jeji gnificance of the Silurian nautiloid Aptychopsis. paleogeograficky vyznam. Sbornik. Ustredniho Lethaia, 11 (2), 127-138, 15 text-figs, Osio. ustavu geologického, 20, 129-148, pls. 1-2, Prague. VAI G. B. (1972 ) · Evidence of Silurian in the English summary in Geol. Soc. America Mem. Apuane Alps, (Tuscany, Italy). Giornale di Geo­ 67, v. 2, 850-851, 1957. logia, 38, 349-372, 3 pls., Bologna. POTTER P. E. & PETTI.IOHN F. J. (1963) · Paleocurrents VAI G. B. (1979 ) · Sedimentary evironment of Devon­ and basin analysis. pp. 1-296, Springer-Verlag, Berlin-Heidelberg. ian pelagic limestones in the Southern Alps. Lethaia, 4, (in press). REYMENT R. A. (1971) · Introduction to quantitative 12, paleoecology. Elsevier Pubi. Company, 1-226, Am­ WATKINS R. (1978) · Bivalve ecology in a Silurian sterdam-London-New York. shelf environment. Lethaia, U, 41-56, 11 text-figs, Osio. RHOADS D. C. & MORSE J. W. (1971) · Evolutionary and ecologie significance of oxygen deficient WATKINS R. & BERRY W. B. N. (1977 ) · Ecology of marine basins. Lethaia, 4, 413428, 5 text-figs., Late Silurian fauna of graptolites and associated Osio. organisms. Lethaia, 10, 267-286, 16 text-figs, Osio.

RISTEDT H. (1968) · Zur Revision der . ZIEGLER A. M., COCKS L. R. M. & BAMBACH R. K. Akad. Wiss. u. Literatur, Abhandl. d. Math.-Na­ (1968) · The composition and structure of Low­ turwiss. Klasse, 4, 213-297, pls. 1-5, 4 text-figs., er Silurian marine communities. Lethaia, l, 1-27, Wiesbaden. Osio.

A contribution to IGCP Project No. 53

l