Animal–substrate interactions preserved in ancient lagoonal chalk

Trace-fossil assemblages reflect the response of the benthos to sets of paleoenvironmental conditions during and immediately after sedimentation. Trace fossils have been widely studied in pelagic shelf and deep-sea chalk deposits from around the globe but never documented from ancient lagoonal chalk successions. Here we report the first detailed ichnologic analysis of a lagoonal chalk unit, using as an example the Upper Cretaceous Buda Formation from the Texas Gulf Coast Basin. In this unit, variable interconnection with the open ocean, accompanied by marked fluctuations in physicochemical parameters inherent to lagoonal circulation (e.g., salinity, hydrodynamic energy, bottom-water oxygenation), highly influenced the resultant trace-fossil content of the chalk. These lagoonal chalk deposits contain twenty ichnotaxa, displaying a clear dominance of Thalassinoides isp. and Chondrites isp., which are present in most of the bioturbated strata. The dominance of Thalassinoides isp., both in softgrounds as an element of the Cruziana Ichnofacies and in firmgrounds as a component of the Glossifungites Ichnofacies, highlights similarities with trace-fossil assemblages from shallow-water shelf-sea chalks. In contrast to both (open) shallow-water shelf-sea chalks and deep-sea chalks, the Buda Formation chalk exhibits more diverse assemblages and sharp fluctuations in ichnodiversity and ichnodisparity during relatively short periods of time. The increased ichnodiversity and ichnodisparity in this lagoonal chalk (in comparison with its open ocean counterparts) may reflect a complex interplay of taphonomic (i.e., incomplete bioturbation allowing preservation of shallow-tier trace fossils and ecologic (i.e., increased spatial environmental heterogeneity in the carbonate lagoonal setting) factors.

Ichnology, the study of organism-substrate interactions, has proved to be highly successful in refining paleoenvironmental interpretations, although not all depositional environments have received the same degree of attention 26,27 . In particular, numerous ichnologic studies have focused on chalk [19][20][21]28,29 , and the ichnologic content of shelf-sea and deep-sea chalks has been compared 19,30 . However, no prior studies have been dedicated to the ichnology of chalks deposited under lagoonal conditions (i.e., in partially to almost fully enclosed shallow shelf settings).
The west-central Texas Buda Formation (commonly referred to as the Buda Limestone; Supplementary  Fig. S1) comprises a lower Cenomanian chalk [31][32][33] deposited on the broad Cretaceous Comanche Shelf of the Texas Gulf Coast Basin [31][32][33][34] . This chalk is characterized by a matrix which is visibly dominated by coccoliths debris, partly cemented coccoliths, and well-preserved coccoliths (see Supplementary Fig. S2). The calcite cement, where abundant, may substantially obscure the coccolith particles in this unit ( Supplementary Fig. S2), which is a common phenomenon during the diagenetic evolution of some chalks 35,36 . The matrix of the Buda Formation can be classified either as "Microtexture 3" or as "Dispersed Clay Microtexture" sensu the "Pure" or "Impure" chalk microtexture classification of Saïag et al. 8 , depending on the CaCO 3 content (higher or less than 96 wt.%, respectively). Valencia et al. 32,33 reported an average CaCO 3 content of 85 wt.% for the west-central Texas Buda Formation, with values ranging from 60 to 99 wt.%. As a comparison, lower Cenomanian chalk units from the Paris Basin has an average CaCO 3 content of 75 wt.%, with values ranging from 52 to 95 wt.% 8 .
The Buda Formation was deposited under varying semi-restricted, slightly restricted to nearly open lagoonal conditions. The latter reflects a higher (but still partial) degree of connection with the open sea, largely controlled by the interplay of a wide and low-gradient paleotopography, together with topographic highs (e.g., Stuart City/ Sligo paleo-reefal complex, Terrel Arch, San Marcos Arch) and fluctuating relative sea levels, under possible eustatic control 32,33 . Evidence includes (1) the thinning or pinching out of the Buda Formation in the Lower Cretaceous Stuart City/Sligo paleo-reefal complex 37 ; (2) the broad and very low-gradient of the Buda Formation paleo-shelf 38,39 ; (3) the endemic nature of the Buda Limestone ammonites (Budaiceras and Faraudiella 40 ); (4) the occurrence of rapid salinity-and oxygen-fluctuating facies, e.g., the brackish-water Cribatina texana-bearing facies (WF1) rapidly transitioning to a normal-marine facies (WF2 32 ); (5) the highly-variable macrobenthos diversity and bioturbation intensity [32][33][34] ; (6) the occurrence of a benthic fauna dominated by mollusks (largely oysters), echinoderms, calcispheres, and green-algae (dasycladales and bryopsidales) 32-34 ; (7) a foraminiferal fraction dominated by benthonic species, with subordinated unkeeled planktonic communities, such as heterohelicids, favusellids and hedbergellids 32,33 ; and (8) the occurrence of proximal tempestites 32 , low-angle cross-bedding and oolitic facies 33 ; among other features [32][33][34]39 . This line of evidence not only testifies its lagoonal character and shallow-water nature but reflects a different extent of interconnection with the open ocean 32,33 .
The Buda Formation represents a unique opportunity to understand how ancient endobenthic communities responded to changing environmental parameters inherent to lagoonal circulation. Therefore, the aims of this study are to (1) document the trace-fossil content of the Buda Formation chalk, (2) compare the chalk ichnofauna of this unit with the trace-fossil content of chalk from shelf-and deep-sea settings, and (3) provide a characterization of lagoonal chalk trace-fossil assemblages that allows differentiating this more restricted setting from those present in chalk formed in the open sea. The underlying broad objective of this research is to expand the applications of ichnology to depositional environments that remain underexplored. Additionally, the outcome of this study illustrates the importance of integrating ichnologic and sedimentologic datasets in elucidating physicochemical parameters during deposition of ancient chalks in particular and carbonates in general.

Softground trace-fossil assemblages in nearly open
Firmground and hardground trace-fossil assemblages (Figs. 1, 2g,i, 4d-f and Supplementary  Figs. S3-S5). In addition to the softground ichnotaxa, overprinting firmground and hardground trace-fossil suites occur in the Buda Formation. Firmground trace fossils are abundant and distributed in the different depositional subenvironments, as well as within major (e.g., sequence boundary, maximum flooding surface) and minor (e.g., flooding surface) stratigraphic surfaces. Firmground trace fossils include Thalassinoides isp. with rare Gyrolithes isp. terminations (Fig. 2i), and ?Sinusichnus isp. (Fig. 4e). The hardground trace fossils, on the other hand, consist of bivalve borings 57 assigned to Gastrochaenolites isp. (Fig. 4d), the worm-produced boring 58 Trypanites isp. (Fig. 2g), and an indeterminate microboring only visible under back-scattered electron microscopy analysis (Fig. 4f). These bioeroding structures are usually sparse (yet locally covering up to 40% of the rock fabric) and largely linked to major stratigraphic surfaces, such as sequence boundaries and maximum flooding surfaces. As in the case of the firmground burrows, the bioerosion trace fossils are not linked to a specific lagoonal subenvironment but are more commonly distributed through all the shallower-water chalk deposits from central Texas.

Discussion
Lagoons are shallow-neritic water bodies developed on platforms that are generally protected by wide shallow seas, by reef trends, by sand-shoal barriers, or by islands 59,60 . Several studies in lagoons and nearshore settings have shown that physicochemical stress factors, such as freshwater discharge, hydrodynamic energy, sedimentation rate, water turbidity, and bottom-water oxygenation, have a considerable impact on the benthos, therefore influencing trace-fossil composition [61][62][63] . Substantial freshwater discharge and the resultant brackish-water conditions have been widely acknowledged as major stressors in marginal-marine systems 61,64,65 . In the Buda Formation, the brackish-water chalk interval is characterized by a depauperate Cruziana Ichnofacies (Fig. 1), with extremely low ichnodiversity and ichnodisparity (monospecific suite of Thalassinoides isp.), and low to moderate bioturbation (BI = 2-3), in agreement with the basic tenets of the brackish-water trace-fossil model 64,65 . In fact, ichnodisparity and ichnodiversity increase parallel to an increase in the degree of connection with the open sea.
Overall, the characteristics of the Buda Formation trace-fossil assemblages indicate that endobenthic colonization of this chalk occurred during low-energy periods, as indicated by the generalized scarcity of vertical dwelling burrows and escape trace fossils, the abundance of horizontally oriented endobenthic structures, and the overall high bioturbation intensity 66,67 . Nevertheless, this was not the case for the proximal tempestites facies. Therein, (1) the dominant vertical components of the Thalassinoides isp. networks, (2) the occurrence of unknown vertical trace fossils, and (3) the relatively low bioturbation intensity of these event layers (BI = 1-3) compared to the slowly settled background deposits (BI = 2-5), likely reflect that, despite most endobenthic colonization occurred during a waning stage of the flow, energy was persistently high near the bed 67 . The rate of sediment accumulation is considered to have a major impact on the colonization window and the consequent bioturbation intensity 29 . Rapid sedimentation frequently reduces or inhibits bioturbation 68 , as in the case of the tempestites, whereas slow rates of sedimentation represent the ideal scenario for intense burrowing activity 69 . In stratigraphic sequences, sediment accumulation rates are controlled by the interplay between sediment supply and accommodation 70 . Unlike in siliciclastic sediments, sediment supply in carbonate systems may not only be controlled by changes in physical energy (e.g., storms, turbidity currents) but by the productivity of the carbonate factory 71 . In the Buda Formation, the carbonate factory was largely controlled by changing climate conditions, sea level, and variable water turbidity 32 . The more productive times (i.e., periods with increased sedimentation rate) were characterized by high sea levels, higher temperatures, and quieter and less turbid waters 32 . Interestingly, however, the most bioturbated facies are those associated with the most carbonate productive and less clay-diluted cycles (Lechuguilla and Love Station members 32 ) (i.e., cycles characterized by increased sedimentation, thus somehow contradicting the influence of sedimentation rate on the colonization window). Nevertheless, the relatively reduced bioturbation intensity in the slowly settled and clay-richer depositional cycle (Red Light Member 32 ) may not have been a response to changes in sedimentation rates but a consequence of other significant coeval operating stress factors, such as elevated water turbidity, which is known as an important stress factor for the benthic communities 72 .
Oxygen content has been widely regarded as a major environmental stress factor on the benthos 29,73 . From an ichnologic perspective, progressive dysoxia is generally reflected by decreasing (1) ichnodiversity, (2) bioturbation intensity, (3) burrow size, and (4) burrow penetration 74 . These organism responses, however, may have been induced by several other physicochemical parameters (e.g., substrate consistency) or taphonomic factors, rather than decreasing oxygenation levels 75 . In the Buda Formation, dominantly oxic bottom-water conditions are expressed by the overall abundant bioturbation intensity and ichnodiversity. Nevertheless, facies with reduced bioturbation intensity and ichnodiversity, and concomitant occurrences of (locally) abundant pyrite and smallsized Chondrites (e.g., CF2b; Fig. 3c), might reflect an endobenthic response to times of oxygen-depleted (dysoxic) bottom-waters 73,76 .
Some of the above-mentioned factors (e.g., dysoxia) may also operate in open shelf-and deep-sea environments. Hence, it is not surprising that shelf-sea chalks (epicratonic) and deep-sea chalks (bathyal-abyssal) 19,30 show some similarities in ichnofaunal content with the Buda Formation 21, [28][29][30] (Table 1; Supplementary  Table S2). Shelf-and deep-sea chalks are composed of trace-fossil suites largely attributable to the Seilacherian Zoophycos Ichnofacies, with abundant occurrences of Chondrites, Zoophycos, and Planolites, as well as common Teichichnus (Fig. 5a,b; Table 1, Supplementary Table S2). However, the deep-sea chalk deposits differ from their shelf counterparts by the lower occurrences of decapod burrows, such as Ophiomorpha and Thalassinoides; the latter being the dominant ichnogenus in shelf-sea settings (Fig. 5a,b www.nature.com/scientificreports/ Moreover, among the recurring chalk ichnogenera, Palaeophycus, Phycosiphon, Taenidium, and Trichichnus are consistently absent or rare in deep-sea chalk deposits (Fig. 5a,b; Table 1; Supplementary Table S2). Skolithos, on the other hand, seems to be more common in deep-sea chalks (Fig. 5a,b). Shelf-sea chalks also show some differences between their bathymetrical end-members (Supplementary Table S2). For example, chalks accumulated in shallow-water environments (e.g., inner shelf) contain trace-fossil suites dominated by Thalassinoides with very rare or absent Zoophycos, attributable to the Cruziana Ichnofacies. Deeper-water shelf-sea chalks (e.g., outer  www.nature.com/scientificreports/ shelf, distal epeiric basins), in contrast, exhibit trace-fossil suites with abundant occurrences of Zoophycos that illustrate the Zoophycos Ichnofacies. Taking into consideration the abundance of Thalassinoides in both shallow-and deep-shelf settings (Supplementary Table S2), the scarcity/absence of Zoophycos seems to be critical in discriminating between these two environments on ichnologic grounds. In general, and somewhat unsurprising given their close bathymetric proximity, the trace-fossil assemblages of the Buda Formation are like those of the shallow-water shelf-sea chalks (Table 1, Supplementary Table S2), as indicated by trace-fossil suites dominated by the ichnogenera Thalassinoides and Chondrites, and lacking Zoophycos, therefore illustrating the Cruziana Ichnofacies. Likewise, the common occurrences of the substratecontrolled Glossifungites and Trypanites ichnofacies reflect similarities with ichnofaunas recorded in shallowwater shelf-sea chalks. Nevertheless, the lagoonal Buda Formation chalk shows ichnologic features that differ from those of (open) shallow-water shelf-sea deposits, which may be attributable specifically to its lagoonal character. These features include (1) overall higher ichnodiversity and ichnodisparity levels in lagoon deposits, and (2) abrupt vertical changes in trace-fossil content, reflecting different expressions of the Cruziana Ichnofacies. These differences in trace-fossil content between chalk formed under open marine conditions and chalk formed in a lagoonal setting reflect a complex interplay of ecologic and taphonomic factors.
In comparison with chalk formed under open marine conditions, the studied lagoon chalk trace-fossil assemblages typically show higher ichnodiversity and ichnodisparity. Average ichnodiversity and ichnodisparity in shelf chalks are 9 and 7, respectively (n = 24; Supplementary Table S2). Of all these units, the highest ichnodiversity and ichnodisparity have been recorded in the Upper Cretaceous Austin Chalk (19 bioturbation ichnotaxa and 14 categories of architectural design), which is the only one showing similar ichnodiversity and ichnodisparity levels to those of the Buda Formation (20 ichnospecies and 17 categories of architectural design, respectively). In addition, the Buda Formation contains the bivalve trace fossils Lockeia siliquaria and Protovirgularia isp., and the burrowing anemone trace fossils Bergaueria isp. and ?Conichnus isp., all previously unrecorded in chalk anywhere. Though the Buda Formation deposits are dominantly normal marine in character (except by the brackish-water WF1), the fact that ichnodiversity and ichnodisparity levels are higher in lagoonal settings than in the open sea seems to be inconsistent with current ideas regarding ichnology of marginal-marine environments. However, this apparent anomaly can be explained by invoking an interplay of taphonomic and ecologic controls. Ichnodiversity in shelf and deep-sea chalk is typically relatively low. However, this low level may reflect taphonomic overprint due to the extremely high bioturbation degree in these deposits 108 , resulting in ichnofabrics of low fidelity showing a preservational bias towards elite trace fossils 144 . Intense bioturbation may have allowed preferential preservation of deep-tier trace fossils, preventing preservation of those emplaced in shallow tiers, therefore artificially decreasing ichnodiversity 145 . In other words, intense bioturbation is masking the true diversity levels of animal activity in these open marine settings. Although the degree of bioturbation in the studied lagoonal chalk is overall relatively high, not all intervals have suffered biogenic reworking to the same extent that those formed under fully marine conditions. This allowed keeping open a taphonomic window that promoted the preservation of shallow-tier trace fossils, therefore contributing to a diversity increase that reflects enhanced ichnologic fidelity.
In addition to this taphonomic overprint, spatial heterogeneity may have played a role in promoting overall high ichnodiversity and ichnodisparity levels in the lagoonal Buda Formation. Increased environmental heterogeneity along a broad spectrum of spatial scales is known to enhance diversity 146,147 . Specifically, carbonate lagoons may have been host to a complex mosaic of habitats contributing to both within-and between-community variability 148 . A comprehensive study dealing with geographic patterns of marine benthos biodiversity along the European coasts showed that lagoons tend to be considerably higher in species density and diversity than open coast systems 149 . Furthermore, the relatively high depositional depth variance within the Buda Formation, which include very-shallow subtidal to deep subtidal lagoonal settings (above SWB) 32,33 , maybe also responsible for increased diversity in this unit 150 .
The sharp changes in trace-fossil content in the Buda Formation chalk are evidenced, for example, by the abrupt vertical replacement of the depauperate, monospecific suite of the Cruziana Ichnofacies occurring in the brackish-water deposits (WF1) by the more diverse expression of this ichnofacies that characterizes the overlying more marine and pervasively bioturbated deposits formed in less restricted lagoonal settings (WF2; Fig. 1), therefore most likely reflecting a lower incidence of the salinity stressor. Another example is noted with the sharp switch from deposits (CF2c) having the impoverished proximal Cruziana Ichnofacies to the overlying highly bioturbated deposits (CF4b) containing the archetypal Cruziana Ichnofacies (Supplementary Figs. S3-S5), due to a substantial upward increase in bottom-water oxygenation. This vertical variability in the different expressions of the Cruziana Ichnofacies in the Buda Formation may represent another distinguishing feature of lagoonal chalks, as a reflection of the rapidly fluctuating physicochemical conditions in this marginal marine setting 151 .

Conclusions
The ichnologic content of the lagoonal chalk deposits from the Buda Formation is similar to those of the (open) shallow-water shelf-sea chalks, as represented by ubiquitous Thalassinoides isp. and Chondrites isp., together with the absence or scarcity of Zoophycos isp. Nevertheless, the studied lagoonal chalk displays higher both ichnodiversity and ichnodisparity than its open ocean counterpart. This relatively richer endobenthic community in the Buda Formation chalk may be the result of increased environmental variability in the lagoonal realm and the effect of differential taphonomic processes in comparison to the open ocean chalks.

Methods
Material. This ichnologic study is based on Buda Formation sections previously assessed in sedimentologic and stratigraphic studies by Valencia et al. 32,33 in the west-central Texas region. It comprises the analysis of seven outcrops (A-1, SA-1, DR1, 16, DR17, DRD-B, DR33-B), and one cored-section ("Shanklin, W.R. 2"), where nine main sedimentary facies were described 32,33 (see Supplementary Table S1 and Supplementary Fig. S1 for facies summary and outcrops/well-core locations, respectively). A total of ten rock slabs representative of most sedimentary facies were collected and polished at the facilities of the University of Saskatchewan. Material lacking a good trace fossil-host rock color contrast was photographed and digitally treated using Adobe Photoshop software (version 21.2.3) to enhance trace-fossil recognition, by using techniques developed by Dorador and Rodriguez-Tovar 152,153 , where adjustments of image levels, contrast/brightness, and vibrance were implemented. In addition, selected rock samples were prepared for back-scattered electron microscopy analysis (SEM), using the Phenom-World Phenom XL scanning electron microscope at the Texas A&M University at College Station.
Ichnologic analysis. The ichnologic characterization of the studied material (outcrops, cores, and polished rock-slabs) included ichnotaxonomic descriptions, tiering analysis, and semi-quantitative estimation of the bioturbation intensity via the Bioturbation Index (BI) of Taylor and Goldring 154 . This index is a widely applied bioturbation intensity scheme that comprises seven categories. BI = 0 indicates no bioturbation (0%). BI = 1 is represented by sparse bioturbation (1-4%) with few discrete trace fossils. BI = 2 represents low bioturbation (5-30%). BI = 3 indicates moderate bioturbation (30-61%), with rare overlapping of the traces. BI = 4 indicates high bioturbation (61-90%), with common overlapping of the traces. BI = 5 typified sediment with intense bioturbation (91-99%), with abundant overlapping a limited reworking. BI = 6 illustrates fully bioturbated (100%) and totally reworked sediment, due to repeated overprinting of the biogenic structures. For the degree of bioerosion, a percentage of area occupied by borings was quantitatively assessed in photographed material, using a combination of the Adobe Photoshop and ImageJ software for boring-area calculations following the method in Cao et al. 155 . This method includes (1) delineation of the boring/bioeroded area using Adobe Photoshop's "lasso tool" and painting it in black; followed by the (2) measurements of the "total area" and "bioeroded area" (in pixels) using the "wand (tracing) tool" in ImageJ; and (3) Table S2). Each of the recorded trace-fossil occurrences has been checked and the ichnotaxonomy revised to use a consistent classification framework.

Data availability
All data generated or analysed during this study are included in this published article [and its supplementary information files].