Excavations of Middle Triassic strata at Monte San Giorgio (Switzerland–Italy) have famously recovered a range of reptiles whose affinities, morphology and preservation have been extensively described. The locality has also yielded a contemporaneous and equally diverse fish fauna that has yet to be described to the same degree. To address this imbalance, a taphonomic study was undertaken using the actinopterygian Saurichthys, a relatively abundant taxon found in the Besano Formation (latest Anisian–earliest Ladinian) and Cassina Beds of the Meride Formation (early Ladinian). Specimens from each horizon were scored for articulation and completeness across ten anatomical units, the resulting datasets being used to determine a taphonomic model and investigate preservational variation through time. Saurichthys showing moderate to high articulation and high completeness occur in both horizons but states of low articulation and moderate completeness are only present in the Besano Formation. The same feature is apparent in corresponding unit plots, suggesting different environmental conditions were present during deposition of each horizon, specifically those that reduced articulation and completeness during a prolonged residence on the sediment surface.
The excellent preservation, diversity and relative abundance of the fossilised biota from Monte San Giorgio (MSG) on the Switzerland–Italy border has led to its recognition as a UNESCO world heritage site in 2003 and 2010 that today is still a source of Middle Triassic vertebrates, invertebrates and plants (Fig. 1). These originate from five horizons, primarily the Besano Formation (latest Anisian–earliest Ladinian), but also the Cava Inferiore, Cava Superiore and Cassina Beds of the Meride Formation (early Ladinian), and the Kalkschieferzone (late Ladinian), deposited in an intra-platform basin on a shallow carbonate shelf at the north-western rim of the deeper Tethys Ocean at about 20 degrees north latitude (Bernasconi 1991; Röhl et al. 2001; Furrer 1995, 1999, 2003; Furrer and Vandelli 2014). Water depth is estimated to have been 30–130 m with the water column stratified into well oxygenated surface waters and predominantly but not continuously anoxic bottom conditions. The site is most famous for the skeletal remains of aquatic and terrestrial reptile fossils, representing the Nothosauridae, Pachypleurosauridae, Placodontia, Ichthyosauria, Protorosauria, Thalattosauria and Rauisuchia. The morphology of the majority has been described in detail, notably by Peyer (1931a, b, 1932, 1936a, b, 1937), Zangerl (1935), Kuhn-Schnyder (1964, 1974), Krebs (1965), Wild (1974), Rieppel (1989), Sander (1989) and Nosotti and Teruzzi (2008). Their taphonomy has also been investigated, highlighting a number of key preservational features and conditions in the depositional environment (Wild 1974; Sander 1989; Furrer 1995, 2003; Furrer and Vandelli 2014; Beardmore 2012; Beardmore et al. 2012).
Less well-known is the even more diverse array of fossilised fish (see overviews in Bürgin 1999; Furrer 2003; Etter 2002). To date, over 30 genera have been recorded from almost all major groups, such as Chondrichthyes, Actinistia, basal Actinopterygii and Neopterygii (Kuhn-Schnyder 1964; Schwarz 1970; Rieppel 1980, 1981, 1985, 1992; Bürgin et al. 1989; Bürgin 1990, 1992, 1996, 1999, 2004; Tintori 1992; Furrer 1995, 2003; Mutter 1998, 2004; Tintori and Lombardo 1999; Romano and Brinkmann 2009; Stockar 2010; Brinkmann and De Baets 2012; Lombardo et al. 2012). In contrast to the reptiles, studies of the fish are often very general. For example, Tintori’s (1992) description of preservation of the Besano Formation fish biota was undertaken as a whole and as one of several fish-bearing, black shales from the Middle–Upper Triassic of the Southern and Eastern Alps of Europe. Furthermore, the focus of this particular study was the degree of oxygenation in the bottom waters of respective depositional basins. Stockar’s (2010) report of fish preservation, although focused on specimens from the Cassina Beds, is also not specific to any particular vertebrate group.
A good number of very specific morphological studies regard the actinopterygian Saurichthys. Rieppel (1985, 1992) and Maxwell et al. (2015) provided a detailed description of the taxon, which is represented at MSG by six species: S. curionii, S. macrocephalus, S. costasquamosus, S. paucitrichus, S. breviabdominalis, S. rieppeli. Characters distinguishing these are noted as overall size, fin structure, skull proportions and features of the dentition (see Rieppel 1992; Maxwell et al. 2015). Investigations of the distribution of Saurichthys species through time has found that S. costasquamosus, S. paucitrichus, S. breviabdominalis and S. rieppeli occur in the Besano Formation with S. curionii and S. macrocephalus occurring in the succeeding Cassina Beds (Rieppel 1992; Maxwell et al. 2015). Reproductive strategies have also been discussed (Rieppel 1985; Bürgin 1990). With regard to taphonomy, Stockar (2010) provides comments on the preservation of Saurichthys, as the dominant fish taxon from the Cassina Beds. Renesto and Stockar (2009) described three examples of presumably adult Saurichthys from the Cassina Beds with embryonic skeletons, many of which show dark eye patches, preserved in the abdominal region of each. Excellent preservation of a further specimen of Saurichthys from the Cassina Beds has allowed the interrelationship of skeletal elements and soft tissue to be observed (Maxwell et al. 2013).
The above knowledge of Saurichthys, combined with a relatively high abundance of fossil specimens at MSG, makes it an ideal subject for a comprehensive preservational study of a fish taxon. This was undertaken with the primary aim of determining a general taphonomic model that could, secondarily, be tested against various parameters relating to the surrounding environment and morphology. For example, the occurrence of specimens in more than one fossiliferous horizon at the MSG locality, and in most cases from known beds, offers the opportunity for variation in preservation through time to be tested and the potential causes of any variation to be inferred. Saurichthys are also found in black shale and carbonate (dolomite and limestone) lithologies, allowing facies variation to be tested. In addition, although normally less than a metre long, the total body length of Saurichthys varies from a few centimetres to several decimetres depending on species and ontogenetic stage, which might also influence preservation (Rieppel 1992).
Taphonomic methodologies accommodating the morphology of fish are not abundant or evident in literature. The analysis herein was undertaken using a modified version of a methodology described by Beardmore (2012) and Beardmore et al. (2012), used to investigate preservation of Middle Triassic reptiles from MSG possessing a typical tetrapod (‘Serpianosaurus-type’) body plan. In accordance with these studies, the skeleton of Saurichthys was divided into a series of anatomical units, each scored for articulation and completeness. In contrast to the original study on pachypleurosaurids, in which the anatomical divisions were the head, neck, dorsal vertebrae, tail, ribs and four limb units, the Saurichthys were divided into a head, axial skeleton (in which only the haemal and neural arches are present), caudal fin, scale row (up to six lines), left and right pectoral fin, left and right pelvic fin, dorsal fin and anal fin units to accommodate and utilise the main features of the piscean morphology: the ‘fish-type’ body plan [Fig. 2; see Rieppel (1992, table 1) for specific Saurichthys anatomical details]. In doing so, the number of units was increased from 9 to 10. In addition, the scoring system was modified so that the head and axial skeletal units were scored between 0 and 5, the scale rows 0 to 6, the caudal fin 0 to 4, the left and right pelvic fin 0 to 2, and the left and right pectoral, dorsal and anal fins 0 to 3, amounting to a maximum score of 36. Units that could not be scored because they were either missing from the edge of a slab or obscured by other parts of the skeleton were noted with an ‘X’ and the possible maximum score was reduced by the score for that particular unit (see Beardmore 2012; Beardmore et al. 2012).
The dataset of scores was used to produce a scatter plot of overall articulation versus completeness values for each specimen (‘percentage plots’) and scores of articulation and completeness for each unit were plotted on a bivariate bubble plot to identify variation within, and between, each anatomical region (‘unit plots’) (Beardmore 2012; Beardmore et al. 2012). Linear regression lines forced to pass through the ‘taphonomic origin’, that is, the position of a carcass on a percentage or unit plot in its fully articulated and complete state prior to post-mortem degradation, were fitted to the data in the percentage and unit plots to determine a T value, herein used as a measure of articulation and completeness, or loss thereof. The T value also defines three trends in the distribution of data that are used to interpret the taphonomic pathway of a carcass in the time between death and final burial under sediment (the ‘death–burial interval’). As the maximum score in the modified methodology herein varies across the ten anatomical units, the boundaries for trends, as originally described by Beardmore (2012), were also re-defined. Visually, Trend 1 is a horizontal distribution of data (the region of variable articulation–high completeness); Trend 2 is a diagonal distribution (from maximum values of articulation and completeness to the minimum); Trend 3 is a distribution between Trends 1 and 2. Actual values are more difficult to apply, but are determined in the strictest sense by dividing the maximum score for a particular unit into thirds, the uppermost being Trend 1, the lowermost Trend 2, and that between Trend 3. An exception is made herein for the pelvic fin units, which have a maximum score of 2 allowing only 2 trends (Trend 1 and Trend 2) to be defined; these fall between 1 to 2 and 0 to 1, respectively.
The data is further described using Pearson’s r2, reported by Excel and used to express, parametrically, the goodness of fit of the regression line forced to pass through the taphonomic origin, and the Spearman rank-order correlation coefficient (rs) calculated in PAST (PAlaeontological STatistics program; Hammer and Harper 2001) used as a non-parametric measure of the strength of correlation between articulation and completeness. Values for r2, rs and T are shown below, or within, the plot they relate to. Verbally, the r2 and rs values are described as low when below 0.35, moderate between 0.35 and 0.65, and high above 0.65 (Beardmore 2012; Beardmore et al. 2012).
All of the MSG Saurichthys included herein are housed in the Paläontologisches Institut und Museum der Universität Zürich, Switzerland (PIMUZ), and were collected during the bed by bed excavations under direction of Bernhard Peyer and Emil Kuhn-Schnyder between 1924 and 1975. Preservation is in black shale or laminated carbonate muds, the latter calcitic in the Cassina Beds, but dolomitised in the Besano Formation (Bernasconi 1991; Furrer 1995; Röhl et al. 2001). For the main analysis, Saurichthys specimens were included if they comprised more than ten individual elements, had at least two units that could be scored for completeness (but not necessarily articulation) with values of 1 or more on the scale devised, and the scoring of at least 7 of the 10 units was possible, even if the score was 0, as opposed to an ‘X’ (see above). The resulting dataset comprised 133 ‘complete’ Saurichthys, divided into 83 specimens from the Besano Formation and 50 from the Cassina Beds analysed separately to determine models of preservation for each horizon and any variation between. These datasets were also used to investigate the effect of size on preservation. A second dataset of 362 Saurichthys was compiled for an investigation of fossil distribution and preservational variation through time. For this investigation, the criterion specified only origination of a specimen from the bed by bed excavations at “P. 902/Mirigioli” (Besano Formation) and knowledge of the source bed (Kuhn-Schnyder 1974; Röhl et al. 2001; Furrer 2003); the majority of specimens were therefore ‘fragmentary’.
The percentage plot for Saurichthys (n = 133) shows a distribution of data from moderate to high articulation and high completeness, to poorly articulated and moderately complete (Fig. 3a). Data is sparse in the latter region and entirely absent in the regions of high articulation and completeness, poorly articulated and fully complete, and poorly articulated and incomplete. The T value calculated for these 133 Saurichthys is 45 %, a mid-Trend 3 distribution of data.
Preservation in each horizon
The percentage data for the Besano Formation (n = 83) covers most of the above described envelope with articulation ranging between 0–91 % and completeness 39–100 % (Fig. 3a). The data is distributed in the regions of near fully articulated and fully complete, and poorly articulated and incomplete with an absence of points in the region of low articulation and high completeness. This distribution is supported by a T value of 40 % (low Trend 3). The high r2 (0.69) infers a good fit of the data to the linear regression line, and the high rs (0.89), a strong relationship between articulation and completeness.
Among the ten unit plots there is no consistently observed pattern of distribution although it is possible to divide the plots into three groups based on the T value and visual observation of points (Fig. 4). The first comprises only the head unit, which exhibits a distribution of data primarily in the region of variable articulation and high completeness but with some limited data indicating low articulation and variable completeness. The T value is relatively high (3.75) in comparison to the other units. The second group comprises the axial skeleton, scale row, caudal fin, dorsal fin and anal fin units, which share a wide distribution of data across most of the available field and moderate T values, when the possible maximum values of each unit is considered. It is noted that these plots also share a concentration of data at, or near, the maximum values of articulation and completeness. The final group is made up of the paired (pectoral and pelvic) fins, which exhibit a distribution of data described as Trend 2. This group could be further divided into the left and right pectoral fin units, sharing a concentration of data at minimum values of articulation and completeness (highlighting an absence of the unit in the majority), and the left and right pelvic fin units, showing a similar concentration of data along the diagonal between maximum to minimum values of articulation and completeness (Trend 2 in the strictest sense). The rs values for the unit plots, all above 0.5, suggest a moderate to good relationship between articulation and completeness. Although r2 values are harder to divide in the same way, the left and right pelvic, and left pectoral fin units have the highest values (0.9, 0.89 and 0.94, respectively) suggesting a better fit to the linear regression line than the head and scale row units in particular (0.26 and 0.37, respectively).
The distribution of the percentage data for the Cassina Bed (n = 50) covers a relatively smaller field by comparison to the Besano Formation data (Fig. 3a). The data is distributed in the region of moderate to high articulation with values for articulation ranging between 34–91 %, and completeness 47–100 %. Despite this, the supporting values for the Cassina Beds (T 40 %; r2 0.69; rs 0.82) are similar to those returned for the Besano Formation.
The ten unit plots for the Cassina Beds can be divided into four groups (Fig. 5). The first comprises the head and scale row units, in which data occurs only in the variable articulation and high completeness region and maximum T values are returned. Although the axial skeleton and caudal fin units also show a focus of data in the same region, they are separated from the head and scale row units to form a second group based on the presence of data representing other (lower) score combinations. The third group comprises the paired (pectoral and pelvic) fins, which exhibit strong Trend 2 distributions and low T values between 0.06 and 0.27. The final group comprises the unpaired dorsal and anal fins, which are more similar to each other than either is to any other unit. The plots can also be divided into two groups based on the rs values: the left and right pectoral, and left and right pelvic fins all have high values above 0.9, suggesting excellent relationships between articulation and completeness, while the scale row, axial skeleton, caudal fin, dorsal fin and anal fin units have low to moderate values between 0.2 and 0.45; no value was returned for the head and scale row units. This division can also be made for the r2 values, with the former group having high values above 0.8 and the latter between 0 and 0.45.
Preservational variation relating to size was investigated by plotting total body lengths of Saurichthys separately against articulation and completeness (Fig. 6a, b). Specimens from each horizon cover a range of measurements from approximately 5–80 cm. In the Besano Formation, the majority are in the low to mid-range: between 5 and 50 cm. In the Cassina Beds the range is narrower and higher, between 20 and 50 cm. In terms of preservation, for specimens of a similar size, the plots indicate higher values of both articulation and completeness for Saurichthys in the Cassina Beds compared to the Besano Formation (Fig. 6).
The Saurichthys from P. 902, 75 of the 83 Besano Formation specimens, are preserved in four main lithologies: black shale (20 specimens), and finely laminated- (7), laminated- (38) and massive dolomite (9) beds (see sections in Bernasconi 1991; Röhl et al. 2001). Values of articulation and completeness for specimens in laminated dolomite are distributed most widely, covering almost the entire envelope described above for MSG Saurichthys as a whole and the Besano Formation specimens (compare Figs. 3a, 7). Data for the other three lithologies is encompassed within this but only the finely laminated dolomite data has any focus in distribution, the data for this being found above 50 % articulation and 75 % completeness. Any further concentrations or patterns are not apparent (Fig. 7).
Variation at P. 902 (Besano Formation)
Articulation and completeness data for the 362 specimens was plotted as a percentage plot (Fig. 3b). Although this includes 83 specimens from the main analysis, there was no compensation in this second analysis for units marked with an X, therefore the position of some points differs slightly. The distribution of data more strongly follows a diagonal line from maximum to minimum values of articulation and completeness (compare Fig. 3a, b) with the majority of data focused in the region of low articulation and moderate to low completeness. Very few points indicate high articulation and high completeness, and data is entirely absent from the region of moderate to low articulation and high completeness. This distribution is supported by a calculated T value of 9 % (Trend 2).
The second part of this analysis investigated the preservation of specimens among beds in the Besano Formation at P. 902. Preservational values of articulation and completeness were plotted separately against bed number, of which there are 183 beds in the standard section at P. 902 (Rieber 1973; Röhl et al. 2001). Specimens in the dataset originate from 70 beds, with most sourcing no more than 12 (Fig. 8a). The exceptions are beds 148 (31 specimens), 152 (54, the highest for the section), 153 (25), 158 (14), 161 (18) and 164 (19). Data shows better articulation and completeness, and a broader range of values, in the upper compared to the lower part of the formation (Fig. 8b, c). Values for the fragmentary specimens show as much variation as the complete specimens, but the former occupies the lower half and the latter, the upper part, of the range for articulation and completeness for each bed.
Specific preservational features
During the scoring of units for Saurichthys, a number of repeated preservational features were observed:
Isolation of the head unit
Examination of the 362 Saurichthys in the P. 902 dataset found that many of the fragmentary specimens comprised of isolated head units but also a number of more complete specimens in which the head as a coherent unit was separated from, but still associated with, the remainder of the body (Figs. 9a, 10). Skulls in isolation were present in 59 of the beds that yielded specimens herein with no more than 12 in any one bed (Fig. 8d). It was also noted that in specimens showing only limited disarticulation of the head, the elements that separated were the opercles, the posterior elements of the skull and/or the lower jaw elements (Fig. 2).
Black circular or oval shaped spots within the orbital fenestrae of the skull were noted in specimens from 10 beds in the upper part of the Besano Formation with no more than 12 in any one bed (Figs. 8e, 9b). Most of these specimens were of small size and considered to be juvenile (36 of 41 exhibiting this feature). In contrast, the number of specimens from the Cassina Beds with the same feature was much lower (2 of 50) and these were of a larger, probably adult, size.
Fins showing full, or near full, articulation as a unit can be separated from the remainder of the body by a distance of a few centimetres (Fig. 9c). The feature is most commonly seen in the unpaired dorsal and anal fins, but has also been observed less frequently in the paired pectoral and pelvic fins. The left and right of the pectoral fin units are more often entirely absent, which is unusual when all other units are at least moderately complete and/or articulated.
Twisting of carcasses
The orientation of a carcass relative to the sediment surface is commonly inconsistent along the length of the body. For example, the caudal fins normally lie in lateral orientation so that each lobe rests on the sediment surface but the head is not restricted to any particular angle given its overall barrel-shape morphology and the limited stabilisation provided by the pectoral fins, which can lie flat against the body. Interestingly, the body by the dorsal and anal fins tends to lie laterally so that these fins are flat against the sediment in the same way as the caudal fins but in a limited number of specimens the dorsal and anal fins appear to have folded over the body as if this region has been rolled. Such movement does not seem to affect the preservation of one or both fins, which are still at least moderately articulated and complete, highlighting the coherence of the soft tissue supporting the rays that comprise the fin units. In a few specimens the axial skeleton and the lateral scale rows can also corkscrew as if half of the body was rolled several times relative to the other in opposing directions (Fig. 9d). Again, these specimens do not show extensive disarticulation or loss of completeness.
Localised disruption of abdominal elements
The location of disarticulation in the axial skeleton tends to occur at particular points in the body, normally at the anterior-most end of the axial skeleton behind or close to the skull (Fig. 9a) and/or the mid-region of the body posterior to the dorsal and anal fins. Only elements in the near vicinity of these points are affected, being scattered randomly and without obvious alignment; those beyond are often still fully articulated. ‘Ballooning’ of the scale rows, bulging symmetrically out from the midline of the body, is also noted. Although a thorough investigation is still necessary there seems to be a greater number of specimens exhibiting these features in the Besano Formation (at least 25 were recorded) compared to the Cassina Beds (approximately 10).
Features of the scale row and axial skeletal units
Highly disarticulated axial skeletal units can show articulated chains of approximately 10–20 neural and haemal arches (Fig. 9c) (Maxwell et al. 2013). In fully articulated Saurichthys, the scale rows tend to form bands that run the full length of the dorsal surface and three quarters of the lateral and ventral surfaces. These effectively form an outline of the body between which the axial skeleton runs. In some specimens, the axial skeleton shows such extensive curvature that either end of relatively long disarticulated segments ‘break out’ of the body outline formed by the scale rows but without disturbing it in any way (Fig. 9c).
The taphonomic data for the main MSG Saurichthys dataset returned a moderate Trend 3 distribution and T value of 45 % (Fig. 3a; Beardmore 2012; Beardmore et al. 2012) inferring disarticulation of skeletal elements was accompanied by loss of completeness in many cases. The notable absence of the left and right pectoral fins supports this, however, evidence at several analytical levels suggests the loss of articulation and completeness was not simultaneous. At specimen level there is an occurrence of three data points representing entirely disarticulated Saurichthys (Fig. 3a), whose skeletal completeness has remained high enough for the scattered elements from each designated unit to be seen in the approximate regions occupied and relative to each other during life. At unit level, the head and several of the fins have been observed as fully or near fully articulated but separated from the remainder of the body. In short, skeletons tend to show some degree of articulation, which varies across the entire range of values, but completeness in most is maintained at or above a moderate level. This conclusion is supported by the findings of Renesto and Stockar (2009, p. 324) who suggested ‘the presence of mostly complete and articulated fossils’ in the Cassina Beds and described the three specimens of Saurichthys with embryos as mainly articulated and complete. Tintori (1992, p. 397) also described the vertebrates of the Besano Formation as ‘wholly articulate’ and suggested preservation of fish could be divided into those that were fully and only partially articulated.
The loss of articulation observed can potentially occur at any time during the death-burial interval, but, during transport and floating, it is automatically accompanied by loss of completeness. Simultaneous loss of articulation and completeness in Saurichthys, as stated above, does not occur in the majority of cases, suggesting arrival at the sediment–water interface was probably relatively soon after death. Alternatively, it has been considered that the high completeness might instead reflect the resistance of the integument to decay, which might prevent any elements that did become disarticulated from separating from the remainder of the carcass. However, a rapid transition to the sediment surface is more likely due to the heavy weight of the skull. Transition through the water column itself can also be inferred as steady, or with carcasses landing long axis parallel to the sediment surface, as penetration into substrate is known in only a few specimens and limited to the rostrum of the skull. Only after arrival at the sediment surface, during the residence phase of the death–burial interval, did the reduction of skeletal articulation and completeness observed in Saurichthys take place, as was also noticed for pachypleurosaurids (Beardmore 2012; Beardmore et al. 2012). In quiet bottom water conditions, the progression of disarticulation can occur passively as soft tissue decays, allowing elements to settle into a stable position by gravity. It can be inferred very generally that the greater the disarticulation of a skeleton the longer a carcass was exposed before final burial under sediment. The fact that Saurichthys show almost the entire range of articulation values suggests that the duration of the residence phase also varied considerably (Fig. 3a-b). Unusually, there is evidence of more active mechanisms of disarticulation in Saurichthys involving additional processes that initiated or exacerbated disarticulation. For example, the above described disruption to elements in the abdominal region could be caused through weakening of the tissues in the gut region as decay progressed, resulting in deformation of the scale rows by gravitational collapse however the localised disarticulation and scattering of elements in the immediate area cannot be explained by gravity alone. Current activity has been inferred at various beds at MSG but would affect the whole carcass rather than focused regions (Sander 1989; see also below). The disruption could be the effect of initial bloating by accumulation of decay gas in the pectoral area just after death in the warm surficial water, as noted by Schäfer (1972), Stockar (2010) and discussed by Maisey (1991). However, as the deposition of the integral carcass occurred in deeper water (30–130 m water deep), further bloating was not possible due to high hydrostatic pressure and probably cold and anoxic bottom water (see Reisdorf et al. 2012, 2014).
A further explanation for the disruption to the abdomen, and other anatomical regions, could be the apparent fixing of parts of the carcass to substrate, specifically the scale rows that effectively form an outline of the body, and the mismatching axial skeleton that breaks through it in several specimens (Fig. 9). The method of fixing was potentially microbial mats growing at the sediment–water interface or substances seeping from the decaying carcass (Bickart 1984; Briggs 1995; Furrer 1999; Renesto and Stockar 2009; Stockar 2010; Stockar et al. 2013; Orr et al. in review); both of these processes cannot immediately adhere a carcass to substrate but take days or weeks to form to the extent that body parts would remain fixed in place (Bickart 1984; Stockar 2010). The fact that the feature is evident in Saurichthys in the first place is due to the position of the axial skeleton in many specimens. That is, the normally relatively straight axial skeleton shows extensive curvature, suggesting it was subject to a process affecting the entire length of the body, but focused in the axial skeleton and probably the soft tissue surrounding these elements, and that was in effect over a sustained period of time (Fig. 9). The force exerted was in the end sufficient to snap the axial skeleton into several sections that each show curvature. Strong, but continuous, curvature has been observed in other vertebrates from MSG, notably the protorosaurs Tanystropheus and Macrocnemus, in which the long neck and sometimes tail curl back to meet the trunk region (Peyer 1937; Wild 1974; Beardmore 2012). Although the cause is still subject to discussion (see Reisdorf and Wuttke 2012), and beyond the scope of this study, in the case of the MSG Saurichthys acquisition of curvature in the axial skeleton can be concluded as post mortem. A carcass had to become fixed to the sediment and a sufficient shortening force exerted on the axial skeleton to cause it to bend and eventually break. All of the above preservational features were observed in several specimens from both the Besano Formation and Cassina Beds, suggesting that the residence phase at the sediment surface was crucially a minimum of several days, but was more likely to be on the week or even month timescale.
Despite the above described commonalities, there are enough differences in the results of the semi-quantitative analyses to suggest there is preservational variation between specimens from each horizon. At the highest percentage level, the strong overlap of data suggests that the same states of moderate to high articulation and high completeness occur in specimens from both horizons, but states of low articulation and moderate completeness are only present in the Besano Formation Saurichthys (Fig. 3a). The extension of the preservational field by the Besano Formation specimens, beyond that of the Cassina Beds, is also observed between corresponding unit plots, most obviously, in the head and scale row units (compare Figs. 3, 4). T values for units are also consistently lower for the Besano Formation. Together, these features imply not only a greater range in the loss of articulation and completeness but also that loss was able to progress much further in Saurichthys from the Besano Formation to result in specimens with low, or no, articulation and moderate completeness. The strong Trend 2 (T = 9 %) distribution for the dataset of 362 Saurichthys (Fig. 3b) suggests reduction of articulation and completeness could progress further still, to the stage where units, notably the head, became disarticulated from the remainder of the body and, subsequently, were isolated. In short, the datasets used herein and the results they returned seem to represent steps in a single taphonomic path from mostly articulated and complete skeletons (Cassina Beds), to poorly articulated and moderately complete skeletons (the more ‘complete’ Besano Formation specimens), and eventually disarticulated skeletons or isolated units (‘fragmentary’ specimens from the Besano Formation). Whether the latter step is also evident in specimens from the Cassina Beds was beyond the scope of this study but is potentially testable in the future as fragmentary specimens from the horizon are also available in the PIMUZ collection.
What facilitated the reduction of articulation and completeness through the steps of the taphonomic pathway is almost certainly the surrounding environmental conditions. Both horizons were deposited in the same subsiding, intra-platform basin (the Monte San Giorgio Basin), under a stratified water column, restricted in the most part from more open marine conditions, and with influxes of allochthonous sediments and organisms from surrounding carbonate platforms during monsoonal storms (Rieber 1973; Sander 1989; Bernasconi 1991; Furrer 1995, 2003; Röhl et al. 2001; Stockar 2010; Beardmore 2012; Beardmore et al. 2012; Stockar et al. 2013). Conditions with enough strength and intensity to exacerbate disarticulation and cause loss of completeness, and that could have varied between the Besano Formation and Cassina Beds, are therefore limited. One difference is water depth, estimated to have been well below maximal storm base (30–130 m) for the Besano Formation and just below maximal storm wave base (30–40 m) for the lower Meride Formation (Bernasconi 1991; Furrer 1995; Röhl et al. 2001; Stockar et al. 2013). Another difference is the sedimentation rate, which is much lower in the Besano Formation than for the Cassina Beds, the latter comprising more numerous and thicker beds of allochthonous carbonate mud and sand. Based on old chronostratigraphic data, the compacted sedimentation rate for the Besano Formation was estimated to 5 mm/kyr (Bernasconi 1994; Furrer 1995) however, newer radiometric dating of volcanic ash layers in the middle Besano Formation (242.1 ± 0.6 Myr, Mundil et al. 1996, 2010) and in the Cassina Beds (240.63 ± 0.13 Ma; Stockar et al. 2012) give a compacted sedimentation rate of about 15 mm/kyr for the Besano Formation, and 32 mm/kyr for the laminite facies of the Cassina Beds (Stockar et al. 2012).
A commonality in previous taphonomic studies, specifically of MSG reptiles, is the suggestion of weak bottom currents. Sander (1989) inferred this feature, based on the alignment of individual bones and skeletal postures in the pachypleurosaurids Neusticosaurus and Serpianosaurus. Beardmore (2012) went on to propose a taphonomic model for these taxa, in which the slower rate of sedimentation in the Besano Formation led to a prolonged residence on the sediment surface in Serpianosaurus when current activity impacted on the completeness of the carcasses. By comparison, residence was shorter for Neusticosaurus resulting in their better articulation and, especially, completeness. The taphonomic model for the protorosaur Tanystropheus also involved a combination of three main factors: sedimentation rate, current activity and size (Beardmore 2012). The latter was highly variable, from 50 to 450 cm, and when correlated with preservation revealed a complicated ‘goldilocks’ relationship whereby the largest specimens were better preserved due to the decreased effect of current activity on their larger and heavier elements, as were the smaller specimens that were more rapidly buried by sediment due to the smaller size of elements comprising their skeletons, compared to moderately sized specimens between that suffered the effects of current activity during prolonged residence on the sediment surface. For Saurichthys, a combination of factors, specifically sedimentation rate, residence duration and current activity, is also indicated. That is, the higher rate of sedimentation would have allowed more rapid burial of Saurichthys preserved in the Cassina Beds compared to those in the Besano Formation. The variable but generally prolonged residence on the sediment surface, especially for the Besano Formation Saurichthys, is supported by several preservational features, notably the greater progression of disarticulation with only a moderate loss of elements. Direct evidence of current activity, such as alignment of skeletal elements is not apparent, even where elements have been disarticulated and scattered; parallel orientation of body parts, such as that described for the neck and tail in Serpianosaurus (Beardmore et al. 2012), was not attainable due to the overall less flexible morphology of Saurichthys. There was also no greater loss of articulation or completeness from smaller Saurichthys skeletons compared to those of larger size (Fig. 6). If currents were present in either the Besano Formation or Cassina Bed setting they were not particularly intense or consistent, but are considered the most probable cause for the limited loss of completeness and the rolling and twisting observed in carcasses.
A final point is the consideration that some of the patterns in preservation are artefacts of excavation. The results of the bed-by-bed investigation of preservation indicated articulation and completeness were higher, and that the scatter of values increased to cover almost the entire range of each, in the upper part of the Besano Formation. The patterns initially support the suggestion of variation on this scale, however the greater surface area of strata excavated at P. 902 in the upper part (240 m2 relative to 90 m2 in the lower part) could potentially have resulted in recovery of a higher number of specimens. Why complete specimens are more common in the upper part of the formation is still unclear. A study of sedimentology and microfacies does however document that the upper part of the Besano Formation is dominated by finer carbonate (mudstone), whereas the middle part is characterised by coarser carbonates (wackestone and packstone) intercalated with muddy black shales (Röhl et al. 2001). The black shale is proposed to have a low sedimentation rate interrupted by rare events with deposition of coarser material (detached turbidites, Bernasconi 1994) while the rate for the dolomitic mudstone of the upper Besano Formation was much higher. These features should be investigated further as a possible explanation of preservational differences.
The investigations of taphonomy in Saurichthys from Monte San Giorgio indicate the following:
The taphonomic model for Saurichthys began with death at some level within the stratified water column of the Monte San Giorgio Basin, followed by minimal floating in the water column and a rapid transition to the sediment–water interface. Reduction in the condition of carcasses occurred during the residence phase on the seabed.
The Saurichthys lay on the sediment surface for variable and often prolonged lengths of time. This is attested to by adhesion of carcasses to the sediment surface, the curvature of the axial skeleton sufficient to break it, and the damage to the abdominal region, which are all features that take several days or weeks to develop.
Disarticulation occurred as decay progressed but loss of completeness was the result of environmental features. The end result for Saurichthys was either a fully disarticulated skeleton or the isolation of parts of the carcass, notably the head unit, from the remainder of the body.
Preservational differences exist between specimens from the Besano Formation and Cassina Beds of the Meride Formation that can be attributed to differences in the depositional settings for each horizon. That is, specimens in the Besano Formation underwent a longer residence on the sediment surface, due to a slower sedimentation rate, and in turn were exposed to weak bottom currents for longer, resulting in the greater variability and progression in the loss of articulation and completeness. Weak bottom currents were probably also responsible for the rolling of carcasses and potentially exacerbated disarticulation in the long term.
Future work should focus on the investigation of small scale, bed-level preservational variation in the Besano Formation at P. 902 to identify more conclusively the extent of variation and in turn any causal features. There is the potential for a similar study to be undertaken for specimens from the Cassina Beds at the same scale.
The study importantly highlights how using more ‘complete’ specimens alone presents only half of the taphonomic story and that the information from ‘fragmentary’ material should also be sought.
Beardmore, S. R. (2012). The skeletal taphonomy of vertebrates from the Triassic and Jurassic. PhD dissertation, University College Dublin, Ireland, pp. 272.
Beardmore, S. R., Orr, P. J., Manzocchi, T., & Furrer, H. (2012). Death, decay and disarticulation: A method of modelling the skeletal taphonomy of marine reptiles demonstrated using Serpianosaurus (Reptilia; Sauropterygia). Palaeogeography, Palaeoclimatology, Palaeoecology, 337–338, 1–13.
Bernasconi, S. M. (1991). Geochemical and microbial controls on dolomite formation and organic matter production/preservation in anoxic environments: A case study from the Middle Triassic Grenzbitumenzone, Southern Alps (Ticino, Switzerland). PhD dissertation, Geological Institute of the ETH, Zürich, Switzerland, pp. 196.
Bernasconi, S. M. (1994). Geochemical and microbial controls on dolomite formation in anoxic environments: A case study from the Middle Triassic (Ticino, Switzerland). Contributions to Sedimentology, 19, 1–109.
Bickart, K. J. (1984). A field experiment in avian taphonomy. Journal of Vertebrate Paleontology, 4, 525–535.
Briggs, D. E. G. (1995). Experimental taphonomy. Palaios, 10, 539–550.
Brinkmann, W., & De Baets, K. (2012). Palaeontological Museum of the University of Zurich: Exhibition guide (p. 108). Zürich: Paläontologisches Museum der Universität Zürich.
Bürgin, T. (1990). Reproduction in Middle Triassic actinopterygians; complex fin structure and evidence of viviparity in fossil fishes. Zoological Journal of the Linnean Society, 100, 379–391.
Bürgin, T. (1992). Basal ray-finned fishes (Osteichthyes; Actinopterygii) from the Middle Triassic of Monte San Giorgio (Canton Tessin, Switzerland). Schweizerische Paläontologische Abhandlungen, 114, 1–164.
Bürgin, T. (1996). Diversity in the feeding apparatus of perleidid fishes (Actinopterygii) from the Middle Triassic of Monte San Giorgio (Switzerland). In G. Arratia & G. Viohl (Eds.), Mesozoic fishes-systematics and paleoecology (pp. 555–565). München: Pfeil.
Bürgin, T. (1999). Middle Triassic marine fish faunas from Switzerland. In G. Arratia & H.-P. Schultze (Eds.), Mesozoic fishes 2 systematics and fossil record (pp. 481–494). München: Pfeil.
Bürgin, T. (2004). Eosemionotus ceresiensis sp. nov., a new semionotiform fish (Actinopterygii, Halecostomi) from the Middle Triassic of Monte San Giorgio (Southern Switzerland). In G. Arratia & A. Tintori (Eds.), Mesozoic fishes—3 systematics, paleoenvironments and biodiversity (pp. 239–251). München: Pfeil.
Bürgin, T., Rieppel, O., Sander, M. P., & Tschanz, K. (1989). The fossils of Monte San Giorgio. Scientific American, 260, 50–57.
Etter, W. (2002). Monte San Giorgio: Remarkable Triassic marine vertebrates. In D. J. Bottjer, W. Etter, J. W. Hagadorn, & C. M. Tang (Eds.), Exceptional fossil preparation: A unique view on the evolution of marine life (pp. 221–242). New York: Columbia University Press.
Furrer, H. (1995). The Kalkschieferzone (Upper Meride Limestone) near Meride (Canton Ticino, Southern Switzerland) and the evolution of a Middle Triassic intraplatform basin. Eclogae Geologicae Helvetiae, 88, 827–852.
Furrer, H. (1999). New excavations in marine Middle Triassic Fossil-Lagerstaetten at Monte San Giorgio (Canton Ticino, Southern Switzerland) and the Ducan mountains near Davos (Canton Graubuenden, Eastern Switzerland). In 3rd international symposium on lithographic limestones. Rivista Museo Civico Scienze "Naturali Enrico Caffi", supplemento 20, 85–88.
Furrer, H. (2003). Der Monte San Giorgio im Südtessin—vom Berg der Saurier zur Fossil-Lagerstätte internationaler Bedeutung. Neujahrsblatt der Naturforschenden Gesellschaft in Zürich, 206, pp. 64.
Furrer, H., & Vandelli, A. (2014). Guide to the museum of fossils from Monte San Giorgio Meride. Fondazione del Monte San Giorgio, pp. 128, ISBN 978-88-940067-2-8.
Hammer, Ø., Harper, D. A. T., & Ryan. P. D. (2001). PAST: Paleontological Statistics Software package for education and data analysis. Palaeontologia Electronica, 4, http://palaeo-electronica.org/2001_1/past/issue1_01.htm.
Krebs, B. (1965). Ticinosuchus ferox nov. Gen. nov. sp. Schweizerische Paläontologische Abhandlungen, 81, 1–140.
Kuhn-Schnyder, E. (1964). Die Wirbeltierfauna der Trias der Tessiner Kalkalpen. Geologische Rundschau, 53, 393–412.
Kuhn-Schnyder, E. (1974). Die Triasfauna der Tessiner Kalkalpen. Neujahrsblatt der Naturforschenden Gesellschaft in Zürich, 176, 1–119.
Lombardo, C., Renesto, S., Tintori, A., & Tona, D. (2012). A new species of Sangiorgioichthys (Actinopterygii, Semionotiformes) from the Kalkschieferzone of Monte San Giorgio (Middle Triassic; Meride, Canton Ticino, Switzerland). Bollettino della Società Paleontologica Italiana, 51, 203–212.
Maisey, J. G. (Ed.). (1991). Santana fossils. An illustrated Atlas. Neptune City, NJ: TFH.
Maxwell, E. E., Furrer, H., & Sánchez-Villagra, M. R. (2013). Exceptional fossil preservation demonstrates a new mode of axial skeleton elongation in early ray-finned fishes. Nature Communications. doi:10.1038/ncoms3570.
Maxwell, E. E., Romano, C., Wu, F., & Furrer, H. (2015). Two new species of Saurichthys (Actinopterygii: Saurichthyidae) from the Middle Triassic of Monte San Giorgio, Switzerland, with implications for character evolution in the genus. Zoological Journal of the Linnean Society, 173, 887–912. doi:10.1111/zoj.12224.
Mundil, R., Brack, P., Meier, M., Rieber, H., & Oberli, F. (1996). High resolution U–Pb dating of the Middle Triassic volcaniclastics: Time-scale calibration and verification of tuning parameters for carbonate sedimentation. Earth and Planetary Science Letters, 141, 137–151.
Mundil, R., Palfy, J., Renne, P. R., Brack, P. (2010). The Triassic timescale: New constraints and a review of geochronological data. In S. G. Lucas (Ed.), The Triassic timescale (pp. 41–60). Special Publications of the Geological Society of London, 334.
Mutter, R. J. (1998). Zur systematischen Stellung einiger Bezahnungsreste von Acrodus georgii sp. nov. (Selachii, Hybodontoidea) aus der Grenzbitumenzone (Mittlere Trias) des Monte San Giorgio (Kanton Tessin), Schweiz. Eclogae Geologicae Helvetiae, 91, 513–519.
Mutter, R. J. (2004). The “perleidiform” family Colobodontidae: A review. In G. Arratia & A. Tintori (Eds.), Mesozoic fishes—3 systematics, paleoenvironments and biodiversity (pp. 197–208). München: Pfeil.
Nosotti, S., & Teruzzi, G. (2008). I rettili di Besano-Monte San Giorgio Natura. Rivista di Scienze Naturali, 98, 1–99.
Orr, P. J., Adler, L. B., Beardmore, S. R., Furrer, H., McNamara, M. E., Peñalver-Mollá, E., & Redelstorff, R. (In review) Stick ‘n’ peel: explaining unusual patterns of disarticulation and loss of completeness in fossil vertebrates. Palaeogeography, Palaeoclimatology, Palaeoecology
Peyer, B. (1931a). Die Triasfauna der Tessiner Kalkalpen. I. Einleitung. Schweizerische Paläontologische Abhandlungen, 50, 1–5.
Peyer, B. (1931b). Die Triasfauna der Tessiner Kalkalpen. III. Placodontia. Schweizerische Paläontologische Abhandlungen, 51, 1–25.
Peyer, B. (1932). Die Triasfauna der Tessiner Kalkalpen. V. Pachypleurosaurus edwardsii Cornalia spec. Schweizerische Paläontologische Abhandlungen, 52, 1–18.
Peyer, B. (1936a). Die Triasfauna der Tessiner Kalkalpen. XI. Hescheleria ruebeli nov. gen. nov. spec. Schweizerische Paläontologische Abhandlungen, 58, 1–48.
Peyer, B. (1936b). Die Triasfauna der Tessiner Kalkalpen. X. Clarazia schinzi nov. gen. nov. spec. Schweizerische Paläontologische Abhandlungen, 57, 1–61.
Peyer, B. (1937). Die Triasfauna der Tessiner Kalkalpen. XII. Macrocnemus bassani Nopsca. Schweizerische Paläontologische Abhandlungen, 59, 1–140.
Reisdorf, A. G., Anderson, G. S., Bell, L. S., Klug, C., Schmid-Röhl, A., Röhl, H. J., et al. (2014). Reply to “Ichthyosaur embryos outside the mother body: Not due to carcass explosion but to carcass implosion” by van Loon (2013). Palaeobiodiversity and Palaeoenvironments, 94, 487–494.
Reisdorf, A. G., Bux, R., Wyler, D., Benecke, M., Klug, C., Maisch, M.W., Fonaro, P., & Wetzel, A. (2012). Float, explode or sink: Post-mortem fate of lung-breathing marine vertebrates. In M. Wuttke & A. G. Reisdorf (Eds.), Palaeobiodiversity and palaeoenvironments (pp. 67–82), Special issue 92: Taphonomic processes in terrestrial and marine environments.
Reisdorf, A. G., & Wuttke, M. (2012). Re-evaluating Moodie’s opisthotonic-posture hypothesis in fossil vertebrates. Part I: Reptiles—The taphonomy of the bipedal dinosaurs Compsognathus longipes and Juravenator starki from the Solnhofen Archipelago (Jurassic, Germany). In M. Wuttke & A. G. Reisdorf (Eds.), Palaeobiodiversity and palaeoenvironments (pp. 119–168), Special issue 92: Taphonomic processes in terrestrial and marine environments.
Renesto, S., & Stockar, R. (2009). Exceptional preservation of embryos in the actinopterygian Saurichthys from the Middle Triassic of Monte San Giorgio, Switzerland. Swiss Journal of Geosciences, 102, 323–330.
Rieber, H. (1973). Ergebnisse paläontologisch-stratigraphischer Untersuchungen in der Grenzbitumenzone (Mittlere Trias) des Monte San Giorgio (Kanton Tessin, Schweiz). Eclogae Geologicae Helvetiae, 66(3), 667–685.
Rieppel, O. (1980). A new coelacanth from the Middle Triassic of Monte San Giorgio, Switzerland. Eclogae Geologicae Helvetiae, 73, 921–939.
Rieppel, O. (1981). The hybodontiform sharks from the Middle Triassic of Monte San Giorgio, Switzerland. Neues Jahrbuch Geologie und Paläontologie, Abhandlungen, 161, 324–353.
Rieppel, O. (1985). Die Gattung Saurichthys (Pieces; Actinopterygii) aus der Mittleren Triass des Monte San Giorgio, Kanton Tessin, Schweiz. Schweizerische Paläontologische Abhandlungen, 108, 1–103.
Rieppel, O. (1989). A new pachypleurosaurid (Reptilia, Sauropterygia) from the Middle Triassic of Monte San Giorgio, Switzerland. Philosophical Transactions of the Royal Society of London, Series B, 323, 1–73.
Rieppel, O. (1992). A new species of the genus Saurichthys (Pieces; Actinopterygii) from the Middle Triassic of Monte San Giorgio (Switzerland), with comments on the phylogenetic interrelationships of the Genus. Palaeontographica Abteilung A, 221, 63–94.
Röhl, H. J., Schmid-Röhl, A., Furrer, H., Frimmel, A., Oschmann, W., & Schwark, L. (2001). Microfacies, geochemistry and palaeoecology of the Middle Triassic Grenzbitumenzone from the Monte San Giorgio (Canton Ticino, Switzerland). Geologia Insubrica, 6, 1–13.
Romano, C., & Brinkmann, W. (2009). Reappraisal of the lower actinopterygian Birgeria stensioei Aldinger, 1931 (Osteichthyes; Birgeriidae) from the Middle Triassic of Monte San Giorgio (Switzerland) and Besano (Italy). Neues Jahrbuch für Geologie und Paläontologie Abhandlungen, 252, 17–31.
Sander, P. M. (1989). The Pachypleurosaurids (Reptilia; Nothosaurier) from the Middle Triassic of Monte San Giorgio (Switzerland) with the description of a new species. Philosophical Transactions of the Royal Society of London, Series B, 325, 561–666.
Schäfer, W. (1972). Ecology and paleoecology of marine environments. Chicago: The University of Chicago Press.
Schwarz, W. (1970). Birgeria stensiöi Aldinger. Schweizerische Paläontologische Abhandlungen 89, 1–93.
Stockar, R. (2010). Facies, depositional environment, and palaeoecology of the Middle Triassic Cassina Beds (Meride Limestone, Monte San Giorgio, Switzerland). Swiss Journal of Geosciences, 103, 101–119.
Stockar, R., Adatte, T., Baumgartner, P. O., & Föllmi, K. B. (2013). Palaeoenvironmental significance of organic facies and stable isotope signatures: The Ladinian San Giorgio Dolomite and Meride Limestone of Monte San Giorgio (Switzerland, WHL UNESCO). Sedimentology Special Issue: Alpine Sedimentology, 60, 239–269.
Stockar, R., Baumgartner, P. O., & Condon, D. (2012). Integrated Ladinian bio-chronostratigraphy and geochrononology of Monte San Giorgio (Southern Alps, Switzerland). Swiss Journal of Geosciences, 10, 85–108.
Tintori, A. (1992). Fish taphonomy and Triassic anoxic basins from the Alps: A case history. Rivista Italiana di Paleontologia e Stratigrafia, 97, 393–408.
Tintori, A., & Lombardo, C. (1999). Late Ladinian fish faunas from Lombardy (North Italy): Stratigraphy and paleobiology. In G. Arratia & H.-P. Schultze (Eds.), Mesozoic fishes 2 systematics and fossil record (pp. 495–504). München: Pfeil.
Wild, R. (1974). Tanystropheus longobardicus (Bassani). Schweizerische Paläontologische Abhandlungen, 92, 1–162.
Zangerl, R. (1935). Pachypleurosaurus edwardsii Cornalia sp. Osteologie—Variationsbreite—Biologie. Schweizerische Paläontologische Abhandlungen, 56, 1–80.
We would like to thank T. Brühwiler, C. Egli, J. Huber, H. Lanz, L. Pauli (all PIMUZ), C. Obrist (Rickenbach BL) and U. Oberli (St. Gallen) for their excellent preparation of the specimens. B. Jost (PIMUZ) drew Fig. 1. We also appreciate useful comments by two anonymous reviewers. This project was funded through the Swiss National Science Foundation (SNF) Sinergia Grant CRSII3-136293 to M. Sánchez-Villagra, H. Furrer, and W. Salzburger.
Editorial handling: D. Marty.
About this article
Cite this article
Beardmore, S.R., Furrer, H. Taphonomic analysis of Saurichthys from two stratigraphic horizons in the Middle Triassic of Monte San Giorgio, Switzerland. Swiss J Geosci 109, 1–16 (2016). https://doi.org/10.1007/s00015-015-0194-z
- Skeletal taphonomy