Chronostratigraphy and new vertebrate sites from the upper Maastrichtian of Huesca (Spain), and their relation with the K/Pg boundary

Accepted Manuscript Chronostratigraphy and new vertebrate sites from the upper Maastrichtian of Huesca (Spain), and their relation with the K/Pg boundary E. Puértolas-Pascual, I. Arenillas, J.A. Arz, P. Calvín, L. Ezquerro, C. García-Vicente, M. Pérez-Pueyo, E.M. Sánchez-Moreno, J.J. Villalaín, J.I. Canudo PII: S0195-6671(17)30487-1 DOI: 10.1016/j.cretres.2018.02.016 Reference: YCRES 3817 To appear in: Cretaceous Research Received Date: 9 November 2017 Revised Date: 24 January 2018 Accepted Date: 22 February 2018 Please cite this article as: Puértolas-Pascual, E., Arenillas, I., Arz, J.A., Calvín, P., Ezquerro, L., GarcíaVicente, C., Pérez-Pueyo, M., Sánchez-Moreno, E.M., Villalaín, J.J., Canudo, J.I., Chronostratigraphy and new vertebrate sites from the upper Maastrichtian of Huesca (Spain), and their relation with the K/ Pg boundary, Cretaceous Research (2018), doi: 10.1016/j.cretres.2018.02.016. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Chronostratigraphy and new vertebrate sites from the upper Maastrichtian of Huesca (Spain), and their relation with the K/Pg boundary E. Puértolas-Pascual 1,4 , I. Arenillas 5 , J.A. Arz 5 , P. Calvín 2 , L. Ezquerro 3 , C. García-Vicente 5 , M. PérezPueyo 1 , E.M. Sánchez-Moreno 2 , J.J. Villalaín 2 , J.I. Canudo 1 1 Grupo Aragosaurus-IUCA, Área de Paleontología, Dpto. Ciencias de la Tierra, Universidad de Zaragoza, 50009 Zaragoza, Spain [email protected]; [email protected]; manuppuey[email protected] 2 Dpto. de Física, Facultad de Ciencias, Universidad de Burgos, 09001 Burgos, Spain [email protected]; [email protected]; vill[email protected] 3 Grupo GEOTRANSFER, Área de Estratigrafía, Dpto. Ciencias de la Tierra, Universidad de Zaragoza, 50009 Zaragoza, Spain [email protected] 4 Faculdade de Ciências e TecnologiaGeoBioTec, Universidade Nova de Lisboa, 2829-526 Monte de Caparica, Portugal [email protected] 5 IUCA, Área de Paleontología, Dpto. Ciencias de la Tierra, Universidad de Zaragoza, 50009 Zaragoza, Spain [email protected]; [email protected]; [email protected] Abstract: The transitional-continental facies of the Tremp Formation within the South-Pyrenean Central Unit (Spain) contain one of the best continental vertebrate records of the Upper Cretaceous in Europe. This Pyrenean area is therefore an exceptional place to study the extinction of continental vertebrates across the Cretaceous/Paleogene (K/Pg) boundary, being one of the few places in Europe that has a relatively continuous record ranging from the upper Campanian to lower Eocene. The Serraduy area, located on the northwest flank of the Tremp syncline, has seen the discovery of abundant vertebrate remains in recent years, highlights being the presence of hadrosaurid dinosaurs and eusuchian crocodylomorphs. Nevertheless, although these deposits have been provisionally assigned a Maastrichtian age, they have not previously been dated with absolute or relative methods. This paper presents a detailed stratigraphic, magnetostratigraphic and biostratigraphic study for the first time in this area, making it possible to assign most vertebrate sites from the Serraduy area a late Maastrichtian age, specifically within polarity chron C29r. These results confirm that the vertebrate sites from Serraduy are among the most modern of the Upper Cretaceous in Europe, being very close to the K/Pg boundary. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Key words: K/Pg boundary; Maastrichtian; magnetostratigraphy; biostratigraphy; vertebrates; Pyrenees. Introduction Recognizing the K/Pg (Cretaceous/Paleogene) boundary in continental deposits is a complicated task due to several biases that affect the continental record (Barret et al., 2009; Butler et al., 2011; Smith et al., 2001; Mannion et al., 2011; Smith and McGowan, 2011; Upchurch et al., 2011). Even so, great efforts have been made in recent years to detect the continental K/Pg boundary and ascertain its relation with faunal and floral extinctions, especially in North America (e.g., Fastovsky and Sheehan, 2005; Archibald et al., 2010; Brusatte et al., 2015; and references therein), but also in Europe (Canudo et al., 2016; and references therein) and Asia (Jiang et al., 2011; and references therein). In the European scenario, the greatest difficulty in knowing how the vertebrate faunas were affected by the K/Pg extinction event is as a result of the fragmentary nature of the continental geological record during the Late Cretaceous and early Paleogene. Nonetheless, major advances have been made in the last few years, and new outcrops, mainly in Romania, France and Spain, are being discovered and datings carried out (Puértolas-Pascual et al., 2016; and references therein). In Eastern Europe (Romania), the Maastrichtian continental vertebrate assemblages have been examined and dated by biostratigraphy, magnetostratigraphy and radioisotopic techniques (Antonescu et al., 1983; Van Itterbeeck et al., 2005; Codrea et al., 2010, 2012; Panaiotu and Panaiotu, 2010; Bojar et al., 2011; Panaiotu et al., 2011; Vremir et al., 2014; Csiki- MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Sava et al., 2015). However, the vertebrate sites are located in excessively broad age ranges, the correlation between the different sites remains problematic and more accurate datings are required (Buffetaut and Le Loeuff, 1991; Gheerbrant et al., 1999; Codrea et al., 2012; Vremir et al., 2014). Similar concerns occur with respect to the Upper Cretaceous and lower Paleogene of France. Apart from the Provence area, where several biostratigraphic, magnetostratigraphic, chemostratigraphic and sedimentological studies with a good chronostratigraphic control have been carried out (Cojan et al., 2003; Cojan and Moreau, 2006), most of the northern Pyrenees still lack accurate datings or correlations (Buffetaut and Le Loeuff, 1991; Laurent et al., 2002). Therefore, despite the abundant Maastrichtian vertebrate fossil record recovered from southern France, only limited biostratigraphic data (Bessiére et al., 1980, 1989; Bilotte, 1985; Garcia and Vianey-Liaud, 2001; Marty, 2001) and one new magnetostratigraphic study (Fondevilla et al., 2016b) are available, and further studies and correlations are still necessary (Dinarès-Turell et al., 2014). The continental vertebrate record of the uppermost Cretaceous of Spain is one of the most complete and most studied in Europe (e.g., Company and Szentesi, 2012; Ortega et al., 2015; Pereda-Suberbiola et al., 2015; Canudo et al., 2016). Most of these vertebrate sites are located within the Tremp Basin, in the Pyrenees of Aragon and Catalonia (Spain), specifically in the Maastrichtian transitional and continental facies of the Tremp Formation. The Tremp Formation has been exhaustively prospected and studied, providing abundant new vertebrate fossil remains including dinosaurs, crocodylomorphs, testudines, mammals, fishes, amphibians and squamates (e.g., López-Martínez et al., 1999, 2001; Peláez-Campomanes et al., 2000; Riera et al., 2009; Pereda-Suberbiola et al., 2009; Blain et al., 2010; Cruzado-Caballero et al., 2010, MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT 2013, 2015; Puértolas et al., 2011; Marmi et al., 2012, 2016; Vila et al., 2012, 2013, 2015; Moreno-Azanza et al., 2014; Puértolas-Pascual et al., 2014, 2016; Sellés et al., 2014a, 2014b, 2016; Blanco et al., 2014, 2015a, 2015b, 2016, 2017; Company et al., 2015; Torices et al., 2015; Canudo et al., 2016). In addition to the record of the vertebrates themselves, there is a relatively continuous geological record ranging from the Maastrichtian to the end of the Thanetian (López-Martínez et al., 2006), which is probably the best dated and correlated in Europe for this time interval and which may contain the Cretaceous/Paleogene boundary. This makes the southern Pyrenees and the Tremp Basin one of the best areas in the world for studying vertebrate associations across the K/Pg boundary, allowing comparisons with the extinction patterns reported from other parts of the world (Brusatte et al., 2015; Csiki-Sava et al., 2015; Canudo et al., 2016; Puértolas-Pascual et al., 2016). Ever since the 1980s, therefore, a great effort has been put into dating the fossil vertebrate sites and searching for the K/Pg boundary within the transitional and continental deposits of this sector of the Pyrenees. Outstanding in this context are works on the biostratigraphy of rudists (Vicens et al., 2004), charophytes and palynomorphs (Feist and Colombo, 1983; Médus et al., 1988; Galbrun et al., 1993; López-Martínez et al., 2001; VillalbaBreva and Martín-Closas, 2011, 2013; Villalba-Breva et al., 2012; Vicente et al., 2015), foraminifers (López-Martínez et al., 2001; Díez-Canseco et al., 2014), on eggshells (Vila et al., 2011; Sellés et al., 2013; Sellés and Vila, 2015), magnetostratigraphy (Galbrun et al., 1993; Oms et al., 2007; Pereda-Suberbiola et al., 2009; Vila et al., 2011, 2012; Canudo et al., 2016; Fondevilla et al., 2016a) and dinosaur occurrences (Riera et al., 2009; Vila et al., 2016). MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT The Serraduy area, located in the Aragonese northwestern branch of the Tremp Basin, has been prospected by the Aragosaurus-IUCA research group of the University of Zaragoza for over 10 years. This has resulted in the discovery of around 40 new paleontological sites with hundreds of vertebrate remains. These findings include important specimens such as the holotype of the eusuchian crocodylomorph Agaresuchus subjuniperus (Puértolas-Pascual, Canudo and Moreno-Azanza, 2014) and the smallest hadrosaurid known in Europe to date (Company et al., 2015), probably a new dwarf taxon. Despite the importance and potential of these vertebrate sites, a chronostratigraphic framework for the Serraduy sector has not yet been provided. Serraduy is located between other areas with vertebrate sites such as Campo to the west and Arén to the east, corresponding to the northwestern-most branch of the Tremp Formation within the Tremp Basin. In these nearby sectors (Campo and Arén, Huesca), previous magnetostratigraphic studies have stated that the vertebrate sites of the Tremp Formation in these areas lie within magnetic polarity chrons C30n and C29r, being late Maastrichtian in age (Pereda-Suberbiola et al., 2009; Canudo et al., 2016). On the basis of works of magnetostratigraphy (Fondevilla et al., 2016a) and biostratigraphy (Díez-Canseco et al., 2014) on the more eastward-lying Isona sector of the Tremp syncline, however, some authors have detected the possible presence of important hiatuses in some areas of the Tremp Basin. These gaps reveal that most of the succession and vertebrate content in that area correlates to the early Maastrichtian (mostly chron C31r), suggesting an older age (Fondevilla et al., 2016a) for many vertebrate sites than previously thought (Vila et al., 2012). In accordance with these new datings (Pereda-Suberbiola et al., 2009; Díez-Canseco et al., 2014; Canudo et al., 2016; Fondevilla et al., 2016a), the chronostratigraphic study of the areas of Campo-Serraduy-Arén may thus acquire greater relevance, given that the MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT lower part of the Tremp Formation exposed in Campo and Arén contains the only continental record of chron C30n in the whole Tremp syncline (Fondevilla et al., 2016a). To test all these hypotheses, we here describe for the first time the magnetostratigraphy, biostratigraphy and a preliminary study of the fossil vertebrate assemblage of the Serraduy area. According to our biostratigraphic and magnetostratigraphic results, most of the “lower red unit” of the Tremp Formation up to the top would be included within magnetic polarity chron C29r, and most of the vertebrate sites would therefore have a late Maastrichtian age, being located very close to the K/Pg boundary. Unfortunately, the paleomagnetic and biostratigraphic data from the lower half of the studied sections are not conclusive enough to give a specific age or reveal the presence of hiatuses in this area. Geographical and geological context The studied area is located in the Aragonese part of the Tremp Basin within the Pyrenean range (Serraduy area, Huesca, Spain) (Fig. 1A, B). The Pyrenees are a 430-km-long east-west-oriented continental collisional fold-and-thrust belt, located in the northeastern Iberian Peninsula between France and Spain (Fig. 1A); they formed as the result of the oblique collision and compressive episodes between the Iberian microplate and the European plate. This process took place during the Alpine orogeny, from Late Cretaceous until early Miocene times (GarridoMegías and Ríos, 1972; Puigdefabregas and Souquet, 1986; Muñoz, 1992; Ardèvol et al., 2000; Sibuet et al., 2004; Teixell, 2004). The Tremp Basin is located within the South-Pyrenean Central Unit or SPCU (Séguret, 1972), which corresponds with the central sector of the Southern Pyrenees (between the Nogueras thrust fault in the north and the Sierras Marginales frontal thrust in the south) (Fig. 1A). Several syn-sedimentary synclines (Ager, Tremp, Coll de Nargo and MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Vallcebre) associated with the emplacement of south-verging thrust-sheets developed during the Late Cretaceous, acting as different sub-basins (Fondevilla et al., 2016a; Oms et al., 2016). The fossil remains studied here are from the Arén and Tremp Formations (Mey et al., 1968) within the Tremp Basin and the northern flank of the east-west-oriented Tremp syncline. The Tremp Formation, which forms part of what is informally known as the “Garumnian” facies (Leymerie, 1862), was deposited during the Upper Cretaceous–Paleocene, when the Pyrenees Basin was completely filled by coastal and continental deposits due to the end-Cretaceous marine regression (Rosell et al., 2001), representing the last infilling episode of the South-Pyrenean Basin (Mey et al., 1968; López-Martínez et al., 1999; Oms et al., 2016). The SPCU has an extension of 5000 km 2 , of which the Tremp Formation is estimated to encompass about 1000 km 2 (López-Martínez et al., 2006). This formation crops out in the central and western part of the SPCU, reaching a thickness of about 900 m in the depocenter near the locality of Tremp (López-Martínez et al., 1999). In the northern areas, such as the Tremp syncline, the bottom of the Tremp Formation is underlain by and laterally interdigitated with upper Campanian–Maastrichtian mixed-platform marine deposits that correspond to the beach, barrierisland and deltaic sandstones of the Arén Formation (Fig. 1C) (Ardèvol et al., 2000). In the southern outcrops, such as the Àger syncline, the Arén Formation is replaced by more calcareous deposits corresponding with the limestones Les Serres Formation (Souquet, 1967; LópezMartínez et al., 2006). Above, the Tremp Formation is overlain by Ilerdian (lower Eocene) marine sediments of the alveoline limestones Cadí Formation, or marly deposits laterally equivalent to the Figols Group (Fonnesu, 1984; Eichenseer and Luterbacher, 1992; LópezMartínez et al., 1999, 2006). MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT The sedimentary succession of the Tremp Formation can be linked with two main stages of compressive tectonics (Puigdefàbregas and Souquet, 1986). The first stage occurred during the late Santonian–late Maastrichtian and was characterized by active tectonics, particularly intense during the late Santonian and Campanian, which caused the inversion of the previous Mesozoic rift structures and the development of a foreland basin. During this period, the basin was filled by mainly siliciclastic deposits which progressively passed to marine facies towards the west, where the open sea was located. The second stage occurred during the late Maastrichtian–early Eocene, this being a period of smooth tectonics and almost uniform subsidence represented predominantly by carbonate-marl deposits. The Tremp Formation has been divided into different local units by several authors. Cuevas (1992), and later Pujalte and Schmitz (2005), divided the series into five formations and four members, elevating the Tremp Formation to the category of group. However, the classification of the Tremp Formation as a group is not widespread within the literature and some authors indicate that the boundaries between the formations of this group can be confusing (Riera et al., 2009). For this reason, Galbrun et al. (1993) and Rosell et al. (2001) divided the Tremp Formation into informal units with a wider regional rank. The correspondence between the different units of each author is as follows (Fig. 2): “grey unit” or “Grey Garumnian” of Rosell et al. (2001) (Posa Formation according to Cuevas, 1992; Unit 1 according to Galbrun et al., 1993); “lower red unit” or “Lower Red Garumnian” of Rosell et al. (2001) (Conques Formation and Talarn Formation according to Cuevas, 1992; Unit 2 according to Galbrun et al., 1993); “Vallcebre limestones” and lateral equivalents of Rosell et al. (2001) (Suterranya Formation and St. Salvador de Toló Formation according to Cuevas, 1992; Unit 3 according to Galbrun et al., 1993); “upper red unit” or “Upper Red Garumnian” of Rosell et al. (2001) (Esplugafreda MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT On the other hand, the correlation of the “lower red unit” is based on five characteristic packages of sandstones, two of which can be recognized in the whole area and permit the correlation between the Larra and Barranco Serraduy profiles (Fig. 3A, B). Throughout the whole area, the first correlation level (L1 in Figs. 3 and 4) presents numerous dinosaur ichnites at the base and is always associated with a purple-greyish mudstone (Fig. 3A). In the western area, L1 comprises a tabular body of brownish fineto medium-grained sandstones with parallel lamination that grade upwards to massive sandstones with bioturbation on the top. The underlying tabular body of grey-purple mudstones exhibits root bioturbation, vegetal remains and fossil vertebrates. Towards the southeast, L1 passes laterally into a thick body of greyish mediumto coarse-grained sandstones, which is composed of tabular and channeled strata. These beds are interfingered with purple bioturbated mudstones and show hadrosaurid dinosaur tracks at the base (Fig. 3A). The second level (L2 in Figs. 3 and 4) comprises fineto medium-grained massive sandstones with intense pedogenization and bioturbation in the western zone; L2 shows a negative-upwards trend. L2 corresponds to greyish sandstones that are fineto very coarsegrained with trough cross-bedding, cross-lamination and ripples. This level displays sharp variations in thickness between the Larra and Isábena River outcrops. Towards the east, in the Barranco Serraduy outcrop, the variation in thickness continues to increase, and L2 forms a group of strata (~7 m thick) with a coarsening-upwards trend located 5 m above L1. The presence of hadrosaurid ichnites is also common at the base. Level 3 is a cm-thick bed of brown fine-grained massive sandstones with scarce lateral continuity. L3 is located near to the Isábena River outcrop (Fig. 3B) and is not observed in other areas. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT The fourth correlation level, which is up to one meter thick, is constituted by brown medium-grained sandstones with parallel lamination that passes vertically into cross-lamination. L4 presents a characteristic small slump (~20 cm thick) at the base and massive bioturbated sandstones (~15 cm thick) in the upper part. This level is easily recognizable in both western and eastern areas, from the Isábena River outcrop to near Barranco Serraduy (Fig. 3B). The fifth level (L5 in Figs. 3 and 4A) is a fining-upward bed composed of 20 cm of greyish microconglomerates with slight channel geometry at the base and 60 cm of fine-grained massive sandstones with bioturbation. In the western area, from the Larra to Isábena River outcrops (Figs. 4A, B), this level presents good lateral continuity (Fig. 3B) and shows spectacular hadrosaurid dinosaur ichnites at its base. On the basis of the vertical arrangement of the guide levels, chronostratigraphic refinement is possible for the studied sediments. Comparison of the thickness of the “lower red unit” between the Larra and Barranco Serraduy profiles shows clear variations, the succession being thicker in the latter (Fig. 3A). Accordingly, the average sedimentation rate for the Barranco Serraduy succession was slightly higher than for the Larra section. Consequently, the correlation results also allow us to constrain the vertical position of paleontological sites. This new correlation reveals that the most recent Cretaceous vertebrate remains correspond to dinosaur tracks and bones in L5 (the Camino de Rin 2 site) near the Larra section, which is located ca. 15 m lower than the Danian “Vallcebre limestones” (Fig. 3A). Sedimentological interpretation Characterization of the sedimentary environments requires exhaustive sedimentological analysis in order to establish and interpret correctly the different facies associations not studied MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT in this work. Even so, the Serraduy area deposits may correspond with the evolution from coastal to continental environments, an interpretation previously proposed by several authors for the Tremp Basin (e.g., Eichenseer, 1987; Díaz-Molina, 1987; Cuevas, 1992; Rosell et al., 2001; López Martínez et al., 2006; Díaz-Molina et al., 2007; Oms et al., 2007, 2016; Riera et al., 2009; Villalba-Breva et al., 2012; Díez-Canseco et al., 2014; Canudo et al., 2016; Fondevilla et al., 2016a). The sedimentary features, especially the sedimentary structures and grain-size distribution, indicate that the Arén Formation in this area corresponds with a barrier-island or deltaic environment. The presence of m-scale coarsening-upwards sequences, facies associations and stacking patterns in the studied interval are similar to those described by Navarrete et al. (2013) for barrier-island and washover fan deposits interbedded within mudflat lagoonal deposits. The Tremp Formation mainly represents terrestrial environments, but the presence of planktonic foraminifers (see below) indicates continuous entrances of marine water into the more protected areas. Thus, the lowermost subunit (“grey unit”) has been interpreted as a transitional marine-to-continental environment connecting tidal systems with the barrier island. The “lower red unit” is predominantly composed of reddish-brownish and greyish-darkish mudstones representative of back-barrier mudflats, whereas the brownish tabular and erosive sandstones represent fluvial channels and their overbank deposits in the floodplains. Thus, frequent waterlevel oscillations and cyclic flooding of the mudflat area can be inferred from the sedimentary features. Several characteristics, such as mottling, oxide haloes and crusts, resulted from the migration and differential accumulation of iron, also indicating common water-table oscillations. In this context, the reddish colors of the mudstones suggest frequent subaerial exposure, probably in low water-level events. The vertical bioturbation and carbonate precipitation in the MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT traces indicate the existence of vegetation with root penetration in search of the water level during dry periods. Darkish and greyish mudstones were deposited under anoxic conditions that favored the preservation of organic matter. This facies indicates the occurrence of high waterlevel periods, in which the mudflat areas were flooded. Isolated and anastomosed channeled sandstone bodies with freshwater charophytes (see below) indicate the existence of low-energy, meandering fluvial channels. Tabular, poorly sorted sandstone bodies and bioturbation traces filled with sands in mudstones reveal sharp flooding events related to high-energy water discharges. These floods occurred as a consequence of the overflow in the fluvial channels and the floodplain. The “Vallcebre limestones” represent the establishment of an extensive freshwater lake. Material and methods In order to ensure the replicability of this research, all the paleontological material figured in this study, including the vertebrate remains and foraminifers, is properly labeled with MPZ abbreviations (Museo Paleontológico de la Universidad de Zaragoza) and housed in the Museo de Ciencias Naturales de la Universidad de Zaragoza (Zaragoza, Spain). The methodology applied in this study (magnetostratigraphy and biostratigraphy) is detailed and explained in the corresponding section. Magnetostratigraphy Paleomagnetic sampling and laboratory procedures 115 levels were sampled as part of the Serraduy magnetostratigraphic study, 21 from the Arén Formation (SB) and 94 from the Tremp Formation (SR). Both magnetostratigraphic MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT profiles were generated in the vicinity of the Barranco Serraduy (BS) stratigraphic section. SB consists of 6 levels of blue-grey marls from the Campo Member, and 14 levels of sandstones with one level of grey marls from the Arén Formation. In the SR profile, 17 levels correspond to the “grey unit” of marls and calcarenites of the Tremp Formation. Another 73 levels are defined as an alternation of red, grey and versicolor mudstones, with some levels of sandstones in the socalled “lower red unit”. Finally, 4 levels of the Paleocene “Vallcebre limestones” were sampled at the top of the SR section (Fig. 5). The complete SB profile was sampled with a portable gas-powered and water-cooled drill and directly oriented in the field with a magnetic compass and an inclinometer, providing from 1 to 3 samples per level, each divisible into 1 to 3 standard-sized specimens. In the SR profile, 62 levels were drilled with a portable electrical water-cooled drill, hand samples (blocks) were taken from 29 levels and in 3 levels both drilled and hand samples were collected. These were oriented in situ with a magnetic compass. Hand samples were collected because of how easily broken up (being disaggregated) the finest materials corresponding to the “lower red unit” were. They were consolidated with sodium silicate dissolved in distilled water to try to make the consolidator percolate to the interior of each piece. Once hardened, about 3 cubes per block were sectioned with a disc cutter, maintaining the face perpendicular to the strike line and parallel to the dip line, both oriented in the field, as the marker for the paleomagnetic analysis. The sampled levels in the SB profile were established each 1 m from the SB01 to SB06 marls, every 2–4 m at the beginning of the sandstones, and every 6 m afterward, due to the homogeneity of the materials. The sampled SR profile levels were separated by 1 m whenever possible. In total a sequence of 173 m was sampled. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Thermal (Th) and alternating field (Af) demagnetizations were carried out in the paleomagnetic laboratory of the University of Burgos, using a 755 superconducting magnetometer (2G) with an alternating field inductor demagnetizer system (for automatic Af), a TD48-DC (ASC) oven and a LDA3 (Agico) alternating field demagnetizer (for manual Af). A total of 198 samples (1 to 3 samples per level) were demagnetized with different stepwise temperatures and applied alternating fields according to the sample lithology. 145 of these were Th-demagnetized, heating up to 400–575°C for marls, sandstones, calcarenites and limestones, and up to 475–675°C for mudstones (several samples of all lithologies were heated up to 675°C in order to check their magnetic behaviors), 39 with automatic Af, 14 with manual Af trying to improve the accuracy of the method, and an Af protocol with an initial thermal step of 130°C to delete the part of the signal carried by goethite (all Af up to 100 mT). Principal component analysis (PCA) and great circle (GC) analysis were performed with Remasoft 3.0 software (Chadima and Hrouda, 2006). Virtual Geomagnetic Poles (VGPs) were calculated, through the isolated paleomagnetic directions considered primary. In cases where overlapping prevents the isolation of stable paleomagnetic components, GCs were calculated. Together with the stratigraphic column, the VGP latitudes obtained from PCA paleomagnetic directions are symbolized with a point, whereas for the primary components verified by GC a bar occupies the status corresponding to normal or inverse latitude (-90° to 0° or 0° to 90°) (Fig. 5). In addition, rock-magnetic measurements were carried out at the University of Burgos with a variable field translation balance (VFTB). Powdered whole-rock specimens from 14 representative samples from all lithologies were submitted to experiments on IRM acquisition and backfield curves, hysteresis loops and strong field magnetization versus temperature (Ms-T) MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT curves. Analysis of these measurements was performed with RockMagAnalyzer 1.0 software (Leonhardt, 2006). Paleomagnetic behavior The natural remanent magnetization (NRM) behavior was analyzed separately in accordance with the lithology because of the big variations among lithologies. Sandstones and green/violet mudstones in general show low NRM intensities (0.1–0.5 mA/m) and a heterogeneous paleomagnetic behavior, being paleomagnetically unstable (it is not possible to isolate a reliable paleomagnetic component). In marls and red beds two paleomagnetic components can be identified on the basis of the unblocking temperature ranges and the coherence of the directional data. Blue-grey marls These rocks appear below the Arén Formation (Campo Member), at the bottom of the Tremp Formation (“grey unit”) and intercalated with continental sediments of the Tremp Formation (“lower red unit”) (Fig. 5). Marls of the Campo Member (samples SB01-SB06 and SB11) show homogeneous paleomagnetic behavior, with NRM intensities between 1.1 and 1.78 mA/m, and display two different components in thermal (Th) demagnetization. A low-temperature component MB (unblocking temperatures between 250/300–450 ºC) with a northwards direction and positive inclinations (Fig. 6A) is isolated in all samples. This component does not go to the origin in some samples (Fig. 6B, C), going systematically to the southern quadrant with negative inclination (Fig. 6B). Great circle analysis (Fig. 6E) allows us to infer this high-temperature component (up to 500 ºC) with negative inclination (component MA), but this cannot be isolated MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT because of spurious component formation during heating. Alternating field (Af) demagnetization diagrams show only one recognizable component (6/15–40/60 mT), which corresponds with component MB (Fig. 6A) with a slight overlapping with MA when this appears, as can be observed in the equal-area demagnetization diagram (Fig. 6B), but it is noteworthy that the overlapped component is almost trending to the origin, preventing recognition of the presence of two components. The NRM intensity is lower in the marls of the Tremp Formation (SR samples), around 0.4–0.5 mA/m. A low-temperature component (250–400 ºC) is recognizable, and with some exceptions it goes to the origin, mainly in the basal marls (SR01-SR17); this component shows the same behavior as the already described MB component for the SB samples. However, a few samples (SR-01A, SR26-1) show an overlapping of components either (i) in the definition of great circles (Fig. 6E), indicating the presence of a high component (Fig. 6C), or (ii) by a clusterend showing a south declination and negative inclination (Fig. 6D). Therefore, most samples of SB and SR marls show the low-temperature MB component (between 250/300 ºC and 350/400 ºC) with normal polarity (positive inclination towards the north). This component does not go to the origin because of the presence of a high-temperature component (up to 350/400 ºC) with reversed polarity, which cannot be isolated because of the formation of spurious components but is clearly evidenced by analyzing the great circles. According to the NRM behavior (low coercivity and unblocking temperatures between 150–500? ºC), and the magnetic rock properties (Fig. 6F), both components are carried by magnetite. Thermomagnetic curves show a major growth of magnetic minerals (magnetite according to the Curie temperature in the cooling curve) in agreement with that observed in the NRM. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Red mudstones Red and orange mudstones appear throughout the Tremp Formation, mainly in the upper section. Most samples show low–medium NRM intensities ranging from 0.1 to 1.8 mA/m, but some of them have higher intensities around 3 mA/m. Low-intensity samples (~0.2–0.8 mA/m) usually show a single component (Fig. 7A) with unblocking temperatures from 350/550 ºC up to 620 ºC; however, the end of the component is obliterated because of a spurious component generated during heating. In some samples, generally those of a higher intensity (> 1 mA/m), component A overlaps with an intermediate temperature component between 350 ºC and 550 ºC (Fig. 7B, C); this overlapping component has low inclination and does not go to the origin. In these samples, component RA can be observed at temperatures up to 500 ºC (Fig. 7B). Finally, SR53 (Fig. 7D) shows high intensity and a single component with positive inclination toward the north, which can be interpreted as component RA according to its unblocking temperatures (550–625 ºC). The unblocking temperatures and high coercivity (Fig. 7) point to hematite as the carrier of component RA. This is in agreement with the rock magnetism experiments (Fig. 7E), which are characterized by a high-coercivity magnetic phase with Curie temperatures over 600 ºC. Differences between the cooling and the heating in the thermomagnetic curve indicate the growth of magnetic minerals (probably magnetite or maghemite) during heating, at temperatures above 600 ºC. Interpretation of the paleomagnetic components Carbonatic rocks show the presence of two components with different unblocking temperatures. Component B, carried by magnetite, is characterized by low to intermediate MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT unblocking temperatures (300–450 ºC) and does not go to the origin. Several works (e.g. Juárez et al., 1994; Villalaín et al., 1994; Osete et al., 2007) evidence the presence of low to intermediate unblocking temperatures (below 450–500 ºC) for diagenetic secondary magnetite, and a high-temperature component (above 450 ºC) corresponding with primary magnetite. The unblocking temperatures of secondary minerals are usually lower than those of primary ones. This is because of the small size of secondary minerals, which range from the superparamagnetic to the stable single domain (see Jackson and Swanson-Hysell, 2014). This suggests a secondary origin for component B found in the marls, but does not ensure a primary origin for the hightemperature component since the presence of two secondary magnetizations is also possible. The paleopole reference for the Late Cretaceous of Iberia from the Lisbon Volcanics (Van der Voo and Zijderveld, 1971) corresponds to an expected direction for the section location of D=1.08º and I=47.39º. In spite of the low dip of the studied materials, the mean direction of component B is in better agreement before than after the bedding correction (BBC and ABC respectively, Fig. 8B). This fact agrees with a secondary origin for this component. We can thus consider component B to be a chemical remanent magnetization (CRM). This was acquired probably during the early diagenesis, but after the tilting of the series (note that the tilting is Maastrichian–early Paleocene [Simó et al., 1985], slightly postdating the age of the rocks). Component B partially obliterates the high-temperature component (A), which cannot be calculated because of the growth of magnetic minerals during heating. However, this is clearly evidenced by the demagnetization great circles. Comparisons between the demagnetization great circles (Fig. 8C) calculated in carbonates (the NW–SE GC) and in red beds (the NE–SW GC) are coherent with the calculated MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT hadrosauroids in Eurasia (Sellés et al., 2014a). The recovered eggshells from Serraduy have been identified up until the lower-mid part of the "lower red unit" of the Tremp Formation (172-i/04/f paleontological site). If their assignment to Spheroolithus europaeus is confirmed, these eggshells would therefore be even more modern than those recovered at the Pont d'Orrit locality. Theropods The presence of theropods within the Serraduy area is not very abundant, representing between 1% and 4% of the identified dinosaur remains. These theropod bones have been found from the “grey unit” to the middle part of the “lower red unit” of the Tremp Formation, thus constituting the youngest reliable record of non-avian theropods in the Iberian Peninsula and one of the youngest records in Europe. The most important record corresponds to two isolated teeth that belong to two different taxa, a medium–large form and a small-sized theropod. The first specimen (MPZ 2017/804; Fig. 11F) corresponds to a medium–large tooth with serrated carinae, which was recovered at the top of the "grey unit" of the Tremp Formation, very close to the 172-i/04/e paleontological site. This tooth is very similar to Morphotype 1 described by Torices et al. (2015) in the Spanish sites of Blasi (Huesca, upper Maastrichtian), Montrebei (Lleida, upper Campanian–lower Maastrichtian) and Laño (Burgos, upper Campanian–lower Maastrichtian). Due to its limited diagnostic value, this morphotype has been assigned to Theropoda indet. (López-Martínez et al., 2001; Torices et al., 2004, 2015; Pereda-Suberbiola et al., 2015). The second specimen corresponds to a small tooth with smooth carinae, which was recovered at the Larra 4 paleontological site within the "lower red unit" of the Tremp Formation. This tooth is very similar to the teeth assigned to Coelurosauria indet. from Blasi (Huesca, upper Maastrichtian), Montrebei (Lleida, upper Campanian–lower Maastrichtian), Laño (Burgos, upper Campanian–lower Maastrichtian) and MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Vicari 4 (Lleida, upper Campanian) (López-Martínez et al., 2001; Torices et al., 2004, 2015; Pereda-Suberbiola et al., 2015). The other remains, also recovered in the “grey unit” and the “lower red unit” of the Tremp Formation, correspond to a possible cervical vertebra of an avian theropod (CruzadoCaballero et al., 2012) and fragmentary long bones and a vertebral fragment that may correspond to undetermined theropods. Nevertheless, for the proper taxonomic assignation of these remains, further detailed studies will be necessary. Sauropods Among the dinosaur remains, the presence of sauropods is the scarcest, amounting to around 1% of the identified remains. The most important item is a proximal left femur fragment (MPZ 99/143; Fig. 11E) assigned to Titanosauria indet. (Canudo, 2001; Vila et al., 2012). This femur was recovered in the "grey unit" of the Tremp Formation, representing one of the youngest sauropods yet documented in Eurasia (Canudo, 2001; Vila et al., 2012; Sellés et al., 2016). Other possible sauropod remains consist of a caudal vertebra from the Barranco de Serraduy 4 site (Cruzado-Caballero et al., 2012) and a proximal fragment from a big autopodial bone from the Camino de Fornons 1 site. However, due to the fragmentary nature of these bones, their assignment to Sauropoda still remains doubtful. For the proper assignation of these specimens, further studies as well as the recovery of new remains will be necessary. Both these remains appeared in the middle part of the "lower red unit", so if their assignment to Sauropoda is confirmed, they would be more modern than the femur, extending the presence of sauropods to chron C29r, as already seen in other sectors of the Tremp Basin (Sellés et al., 2016). Testudines MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Another common clade, comprising about 5% of the remains found in Serraduy, is Testudines. The presence of testudines has been recognized from the top of the Arén Formation to the last deposits with vertebrates of the “lower red unit” of the Tremp Formation. This clade is represented entirely by isolated and disarticulated plates, in most cases preventing a more accurate identification or classification in the preliminary study, and rendering further systematic studies necessary. Nevertheless, when the plates are well preserved, it is possible to observe their smooth and brilliant ornamentation crossed by very fine dichotomic sulci, suggesting highly vascularized shell bones. This characteristic ornamentation pattern is widely used to recognize bothremydids (e.g., de Lapparent de Broin and Murelaga, 1996; Murelaga and Canudo, 2005; Marmi et al., 2012), so most of these plates are assigned to Bothremydidae indet. In addition, better-preserved plates allowed more accurate anatomical identification, and a left xiphiplastron and a right mesoplastron belonging to Bothremydidae from the Rim 2 site at the top of the Arén Formation have been recognized (Murelaga and Canudo, 2005). Crocodylomorphs Representing about 4% of the recovered bone remains, crocodylomorphs are one of the most representative taxa in Serraduy. This group of archosaurs is mainly represented by isolated teeth, although some ichnites, a eusuchian vertebra, osteoderm fragments and a complete skull have also been found (Fig. 13E). All the recovered remains have been assigned to Eusuchia. These have a highly extended stratigraphic distribution, remains being found from the top of the Arén Formation up until the last levels with vertebrates before the K/Pg boundary. The most important taxon corresponds to the complete skull of the eusuchian crocodylomorph originally erected in Serraduy with the name of Allodaposuchus subjuniperus MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT (Puértolas-Pascual et al., 2014) (MPZ 2012/288; Fig. 13E). This taxon was later assigned to the new genus Agaresuchus and included within Allodaposuchidae (Narváez et al., 2016), a clade of endemic European eusuchian crocodylomorphs with a record until the K/Pg boundary. Agaresuchus subjuniperus was recovered in one of the last Maastrichtian sandstone strata (Amor 3 site), so this taxon may represent the last and youngest record of Allodaposuchidae before the K/Pg extinction event (Puértolas-Pascual et al., 2014). As regards isolated teeth, at least two different morphotypes have been distinguished. The first morphotype (Fig. 13A) corresponds to a slender conical tooth ornamented with wellmarked longitudinal ridges, which was found at the top of the Arén Formation (Barranco de Extremadura site). These tooth have been assigned to cf. Thoracosaurus (Puértolas-Pascual et al., 2016). This marine genus belonging to Gavialoidea is typical of the Upper Cretaceous–lower Paleocene of Europe and North America, which is consistent with its presence within the shallow marine facies of the Arén Formation. Teeth of the second morphotype (Fig. 13B, C) have been recovered from the “grey unit” up until the last levels with vertebrates within the “lower red unit” of the Tremp Formation. This morphotype corresponds to generalist conical teeth with an ornamentation that varies from smooth to gently longitudinally-ridged enamel. This generalist morphology is widely distributed within Crocodylomorpha and has little taxonomic value (e.g., Prasad and Broin, 2002; Turner, 2006; Andrade and Bertini, 2008; Buscalioni et al., 2008). However, as this dental morphology is also typical of Allodaposuchidae, the most common clade in Europe during the Campanian–Maastrichtian, this morphotype has been tentatively assigned to cf. Allodaposuchidae. Additionally, several isolated teeth similar to those present in A. subjuniperus were recovered in the same site where the holotype was recovered, so these teeth have been assigned to cf. Agaresuchus subjuniperus. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Other bone remains correspond to osteoderm fragments and a procoelous dorsal vertebra (Fig. 13D) that have been assigned to Eusuchia indet. Further evidence of Crocodylomorpha is the presence of about five tracks (MPZ 2012/832) composed of scratch marks and one pedal impression located on a fluvial channel deposit in the uppermost part of the “lower red unit” of the Tremp Formation (Serraduy Norte site, chron C29r). The scratch marks resemble Characichnos whereas the pes track has been assigned to cf. Crocodylopodus (Vila et al., 2015). Amphibians Due to their small size, amphibian remains have only been recovered by washing and sieving techniques. In the area of Serraduy, one of the paleontological sites with the greatest potential for the study of macrovertebrates and microvertebrates is Larra 4, located within the "lower red unit" of the Tremp Formation. This site is located in a dark grey lutite layer with a high organic content, where vegetal remains (wood and amber fragments), macrovertebrates (dinosaurs, crocodylomorphs and testudines) and microvertebrates are highly abundant. Most of the microvertebrate remains are very fragmentary and need a more thorough systematic study for their proper identification. Nevertheless, a very preliminary study of the micropaleontological content allowed us to identify several remains that may correspond with amphibians. The most outstanding remains are several distal parts of humeri. In spite of the low taxonomic value of the humerus (Evans and Milner, 1993), the large and spherical humeral ball shifted laterally and a rather long ulnar epicondyle allow us tentatively to assign these specimens to Discoglossidae indet., being very MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT similar to other humeri assigned to this clade in other sites within the Tremp Basin (Blain et al., 2010; Blanco et al., 2016). Discussion Considering component A as primary (see “Magnetostratigraphy” section), the local magnetic stratigraphy of the Serraduy section can be correlated with the geomagnetic polarity time scale (GPTS) (Gradstein et al., 2012). According to our results, the “grey unit” and the “lower red unit” of the Tremp Formation can be assigned to a reverse polarity chron (Fig. 5). The presence of the planktonic foraminifer P. hariaensis in sample SR35 indicates that this level has a maximum age of 67.3 Ma (maximum age range of the species), or younger if it is reworked. The age range for this species is between 67.3 and 66.0 Ma, its lowermost and uppermost occurrences being coincident respectively with the upper part of C30n and the K/Pg boundary in the middle part of C29r. According to López-Martínez et al. (2006) and DíezCanseco et al. (2014), the “Vallcebre limestones unit” and lateral equivalents are late Danian in age. Because this biostratigraphic information indicates that the K/Pg boundary is located between the “lower red unit” and the “Vallcebre limestones unit”, all the reverse polarity section between levels SR35 and SR90 can thus only correspond with chron C29r (Fig. 5). As a result, the K/Pg boundary can be located within the last 25 m of the “lower red unit”, between the last horizon with dinosaur remains and the “Vallcebre limestone unit” (Fig. 5). Correlating the profile of Barranco Serraduy with Larra, located further west, the K/Pg boundary can be located with more precision within the last 5 m of the “lower red unit” (Figs. 3, 5). Therefore, the upper section of the “lower red unit” is well defined as reverse polarity, pointing to the C29r (Fig. 5). However, some inconsistencies can be observed in the middle part MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT of this formation. SR37, above the level marked by P. hariaensis, has reversed polarity (indicating C29r), whereas level SR53 shows normal polarity. Therefore, one part of this paleomagnetic information must be wrong. As regards its paleomagnetic properties, level SR35 has similar behavior to samples from the uppermost levels, with a non-zero ending cluster up to 500 ºC with reversed polarity. Otherwise, component RA of level SR53 (Fig. 7D) has been defined by its unblocking temperature range; however, it is possible that this component corresponds with the CRM defined as RB. The lithology of both samples can also be analyzed: the paleomagnetic component defined in SR35 is more reliable than that defined in SR53 because the former is sampled in red mudstones, similar to the upper section where coherent paleomagnetic components appear, whereas SR53 is in a level of shale located between sandstones. The greater porosity of sandstones could have favored chemical processes in level SR53. In the light of these considerations, we consider that the section between levels SR37 and SR90 belongs to subchron C29r. The lower section of the Tremp Formation also shows reversed polarities (see “Magnetostratigraphy” section), so at the beginning it is possible to ascribe this to subchron C29r (Fig. 14). However, Fondevilla et al. (2016a) provide evidence of the presence of a major hiatus affecting chrons C31n, C30r and C30n in the Isona section located in the eastern sector of the Tremp syncline (Fig. 14), probably related to an abrupt migration of the basin depocenter. This would imply the presence of consecutive deposits associated with chrons C31r and C29r, with a hiatus lacking most of the upper Maastrichtian. Nevertheless, according to the magnetostratigraphic works of Pereda-Suberbiola et al. (2009) on Arén, and Canudo et al. (2016) on Campo, this major hiatus seems not to have affected the most western sectors of the Tremp Basin (Fig. 14). Therefore, the sections of Campo and Arén acquire greater relevance because MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT the “lower red unit” exposed there is the only continental record of chron C30n in the whole Tremp syncline (Fondevilla et al., 2016a). In summary, the magnetostratigraphic data presented in this work show that both the marls and red beds of the Tremp Formation in the Serraduy area are of reverse polarity. Biostratigraphic data ensure that the upper part of this formation (the last 60 m of the “lower red unit”) belong to chron C29r. However, although this cannot be fully confirmed for the lower part of the Tremp Formation (the “grey unit” and beginning of the “lower red unit”), these units probably also belong to the same chron C29r, unless there exists a hiatus such as that observed by Fondevilla et al. (2016a) in the Isona section. It should be pointed out that this hiatus has not been observed in Campo and Arén (Pereda-Suberbiola et al., 2009; Canudo et al., 2016), the sections closest to the Serraduy area. Conclusions In this work, a chronostratigraphic framework for the vertebrate sites of the Arén and Tremp Formations within the Serraduy sector of the Tremp Basin is proposed for the first time. The joint study of stratigraphy, field correlations, magnetostratigraphy and biostratigraphy has allowed most of the vertebrate sites in this area to be dated to within chron C29r, making this one of the areas with dinosaur sites closest to the K/Pg boundary anywhere in Europe. In addition, a complete faunal list of the taxa recovered in the Serraduy area is presented. This shows a great diversity of theropods, sauropods and hadrosaur dinosaurs, eusuchian crocodylomorphs, testudines, amphibians and probably pterosaurs. The presence of dinosaurs (ichnites and bones) in the highest levels of the series has pinpointed the range of the K/Pg boundary to the last 5 m of the “lower red unit” within the MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Tremp Formation, below the “Vallcebre limestones”. This suggests a high abundance of hadrosaurid dinosaurs, eusuchian crocodylomorphs, amphibians and testudines just before the great extinction event of the Late Cretaceous. Although everything points to a late Maastrichtian age for the studied deposits, the lower half of the section unfortunately shows unclear paleomagnetic signals and inconclusive biostratigraphic content, so it has been assigned an undetermined polarity. The magnetostratigraphic results also seem to indicate the presence of a reverse polarity chron in the lower half of the section, yet we do not have the biostratigraphic data to be able to assign it to a specific chron (C31r, C30r or C29r). For this reason, the continuity of the lower part of the series or the presence of possible hiatuses cannot be determined. Further studies in the adjacent outcrops located between Serraduy and Campo (e.g. Rin or Larra sections) and between Serraduy and Arén (e.g. Iscles section) could be crucial to achieve more accurate knowledge of the chronostratigraphic framework of the northwestern-most branch of the Tremp Basin. In conclusion, these results show the great paleontological potential of the Serraduy area, which is one of the few and most important places in the world for studying, within continental deposits, the great extinction event which affected planet Earth at the end of the Cretaceous. Acknowledgements This work was supported by the Spanish Ministry of Economy and Competitiveness (grant numbers CGL2014-53548-P, CGL2015-64422-P and CGL2017-85038-P), cofinanced by the European Regional Development Fund; and by the Department of Education and Science of the Aragonese Government (grant numbers DGA groups H54 and E05), cofinanced by the MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT European Social Fund (ESF). The paleomagnetic study was possible thanks to the complementary grants (beneficiaries of FPU, grant number CGL2010-16447/BTE: Brief Stays and Temporary Transfers, year 2015) supported by the Spanish Ministry of Culture, Education and Sports; and the Laboratory of paleomagnetism of the University of Burgos (Spain). Eduardo Puértolas Pascual is the recipient of a postdoctoral grant (SFRH/BPD/116759/2016) funded by the Fundação para a Ciência e Tecnologia (FCT-MCTES). Special thanks to R. Silva, V. Burriel, M. Gallego, A. Alonso, C. Nuñez, J. Galán, J. Parrilla, J.C. García and X. Pereda for their help in the field work; and J. Larrañaga and G. Martín for the discovery of the Larra paleontological sites. Many thanks to P. Cruzado for her helpful comments. Rupert Glasgow edited the text in English. References Álvarez-Sierra, M.A., Arribas, M.E., Ardévol, L., Civis, J., Daams, R., Krauss, S., López-Martínez, N., De La Peña, A., Soler, R., Vianey-Liaud, M., 1994. El límite CretácicoTerciario en la sección de Fontllonga (cuenca de Ager, provincia de Lérida), in: Jaca (Spain), II Congreso Del Grupo Español Del Terciario, Comunicaciones, 23–26. Andrade, M.B., Bertini, R.J., 2008. Morphology of the dental carinae in Mariliasuchus amarali (Crocodylomorpha, Notosuchia) and the pattern of tooth serration among basal Mesoeucrocodylia. Arquivos do Museu Nacional (rio de Janeiro) 63–82. Antonescu, E., Lupu, D., Lupu, M., 1983. Correlation palinologique du Crétacé terminal du sud-est des Monts Metaliferi et des Depressions de Haţeg et de Rusca Montană. Anuarul Institutului de Geologie şi Geofizică 59, 71–77. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Csiki-Sava, Z., Buffetaut, E., Ősi, A., Pereda-Suberbiola, X., Brusatte, S.L., 2015. Island life in the Cretaceous-faunal composition, biogeography, evolution, and extinction of land-living vertebrates on the Late Cretaceous European archipelago. ZooKeys 1. doi:https://doi.org/10.3897/zookeys.469.8439 Cuevas, J.L., 1992. Estratigrafía del “garumniense” de la Conca de Tremp: Prepirineo de Lérida. Acta geológica hispánica 27, 95–108. Cushman, J.A., 1926. The foraminifera of the Velasco shale of the Tampico embayment. Bulletin of the American Association of Petroleum Geologists 10(6), 581–612. Cushman, J.A., 1928. Foraminiferes du Stampien du Bassin de Paris. Bulletin de la Societe des Sciences de Seine-et-Oise 9, 47–63. Cushman, J.A., 1933. Some new foraminiferal genera. Contributions from the Cushman Laboratory for Foraminiferal Research 9(2), 32–38. Cushman, J.A., 1938. Cretaceous species of Gümbelina and related genera. Contributions from the Cushman Laboratory for Foraminiferal Research 14(1), 2–28. Dalbiez, F., 1955. The genus Globotruncana in Tunisia. Micropaleontology 1(2), 161– 171. de Lapparent de Broin, F., Murelaga, X., 1996. Une nouvelle faune de chéloniens dans le Crétacé supérieur européen. Comptes rendus de l’Académie des sciences. Série 2. Sciences de la terre et des planètes 323, 729–735. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Díaz-Molina, M., 1987. Sedimentación sintectónica asociada a una subida relativa del nivel del mar durante el Cretácico superior (Fm. Tremp, provincia de Lérida). Estudios Geológicos 43, 69–93. Díaz-Molina, M., Kälin, O., Benito-Moreno, M.I., López-Martínez, N., Vicens, E., 2007. Depositional setting and early diagenesis of the dinosaur eggshell-bearing Aren Fm at Bastus, Late Campanian, south-central Pyrenees. Sedimentary geology 199, 205–221. Díez-Canseco, D., Arz, J.A., Benito, M.I., Díaz-Molina, M., Arenillas, I., 2014. Tidal influence in redbeds: A palaeoenvironmental and biochronostratigraphic reconstruction of the Lower Tremp Formation (South-Central Pyrenees, Spain) around the Cretaceous/Paleogene boundary. Sedimentary Geology 312, 31–49. doi:https://doi.org/10.1016/j.sedgeo.2014.06.008 Dinares-Turell, J., Fondevilla, V., Oms, O., Le Loeuff, J., Vila, B., Estrada, R., Riera, V., 2014. Magnetostratigraphic dating of the Maastrichtian terrestrial successions from the Pyrenees: State of the art and new data from Aude (France). J. Marmi, O. Oms, B. Vila,A. Galobart, R. Estrada, & J. Dinares-Turell (Eds.), Paleontologia i Evolució, mem. especial 68. Donze, P., Porthault, B., Thomel, G., Villoutreys, O. de, 1970. Le Sénonien inférieur de Puget-Théniers (Alpes-Maritimes) et sa microfaune. Geobios 3, 81–82. d'Orbigny, A., 1839. Foraminiferes. In: de la Sagra, R. (Ed.), Histoire physique et naturelle de l'Ile de Cuba. A. Bertrand, Paris, France, 224. Ehrenberg, C.G., 1839. Uber die Bildung der Kreidefelsen und des Kreidemergels durch unsichtbare Organismen. physikalische abhandlungen der Koniglichen akademie der wissenschaften zu Berlin, 59–147. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Eichenseer, H., 1987. Facies geology of late Maastrichtian to Early Eocene coastal and shallow marine sediments, tremp-graus basin, northeastern spain. Arb. Mus. Geol. Pal. Univ. Tübingen 237. Eichenseer, H., Luterbacher, H., 1992. The marine paleogene of the tremp region (NE Spain)-depositional sequences, facies history, biostratigraphy and controlling factors. Facies 27, 119–151. doi:10.1007/BF02536808 El Naggar, Z.R., 1966. Stratigraphy and Planktonic Foraminifera of the Upper Cretaceous-Lower Tertiary: Bulletin of the British Museum (Natural History). Geology, Supplement, 2, 291. Evans, S.E., Milner, A.R., 1993. Frogs and Salamanders from the Upper Jurassic Morrison Formation (Quarry Nine, Como Bluff) of North America. Journal of Vertebrate Paleontology 13, 24–30. Fastovsky, D.E., Sheehan, P.M., 2005. The extinction of the dinosaurs in North America. Gsa Today 15, 4–10. Feist, M., Colombo, F., 1983. La limite Cretace-Tertiaire dans Ie nord-est de l’Espagne, du point de vue des charophytes. Ceol. medil 10, 303–326. Fondevilla, V., Dinarès-Turell, J., Oms, O., 2016a. The chronostratigraphic framework of the South-Pyrenean Maastrichtian succession reappraised: Implications for basin development and end-Cretaceous dinosaur faunal turnover. Sedimentary Geology 337, 55–68. doi:10.1016/j.sedgeo.2016.03.006 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Fondevilla, V., Dinarès-Turell, J., Vila, B., Le Loeuff, J., Estrada, R., Oms, O., Galobart, À., 2016b. Magnetostratigraphy of the Maastrichtian continental record in the Upper Aude Valley (northern Pyrenees, France): Placing age constraints on the succession of dinosaurbearing sites. Cretaceous Research 57, 457–472. doi:10.1016/j.cretres.2015.08.009 Fonnesu, F., 1984. Estratigrafía física y análisis de facies de la secuencia de Figols entre el río Noguera Pallaresa e Iscles:(provs. de Lérida y Huesca). Universidad Autónoma, Departamento de Estratigrafía y Geología Histórica. Galbrun, B., Feist, M., Colombo, F., Rocchia, R., Tambareau, Y., 1993. Magnetostratigraphy and biostratigraphy of Cretaceous-Tertiary continental deposits, Ager Basin, Province of Lerida, Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 102, 41– 52. doi:10.1016/0031-0182(93)90004-3 Garcia, G., Vianey-Liaud, M., 2001. Dinosaur eggshells as biochronological markers in Upper Cretaceous continental deposits. Palaeogeography, Palaeoclimatology, Palaeoecology 169, 153–164. doi:10.1016/S0031-0182(01)00215-2 Garrido-Megías, A., Ríos, L., 1972. Síntesis geológica del Secundario y Terciario entre los ríos Cinca y Segre (Pirineo central de la vertiente surpirenaica, provincias de Huesca y Lérida). Boletín Geológico y Minero de España 83, 1–47. Gheerbrant, E., Codrea, V., Hosu, A., 1999. Découverte de vertébrés dans les Calcaires de Rona (Thanétien ou Sparnacien), Transylvanie, Roumanie: les plus anciens mammiferes cénozoiques d’Europe Orientale. Birkhäuser. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Gómez-Gras, D., Roigé, M., Fondevilla, V., Oms, O., Boya, S., Remacha, E., 2016. Provenance constraints on the Tremp Formation paleogeography (southern Pyrenees): Ebro Massif VS Pyrenees sources. Cretaceous Research 57, 414–427. doi:10.1016/j.cretres.2015.09.010 Gradstein, F.M., Ogg, J.G., Schmitz, M., Ogg, G., 2012. The geologic time scale 2012. elsevier. Jackson, M., Swanson-Hysell, N.L., 2012. Rock magnetism of remagnetized carbonate rocks: another look. Geological Society, London, Special Publications 371, 229. doi:10.1144/SP371.3 Jiang, X., Liu, Y., Ji, S., Zhang, X., Xu, L., Jia, S., Lü, J., Yuan, C., Li, M., 2011. Dinosaur-bearing strata and K/T boundary in the Luanchuan-Tantou Basin of western Henan Province, China. Science China Earth Sciences 54, 1149. doi:10.1007/s11430-011-4186-1 Juárez, M.T., Osete, M.L., Meléndez, G., Langereis, C.G., Zijderveld, J.D.A., 1994. Oxfordian magnetostratigraphy of the Aguilón and Tosos sections (Iberian Range, Spain) and evidence of a pre-Oligocene overprint. Physics of the Earth and Planetary Interiors 85, 195–211. doi:10.1016/0031-9201(94)90017-5 Laurent, Y., Bilotte, M., Le Loeuff, J., 2002. Late Maastrichtian continental vertebrates from southwestern France: correlation with marine fauna. Palaeogeography, Palaeoclimatology, Palaeoecology 187, 121–135. doi:10.1016/S0031-0182(02)00512-6 Leonhardt, R., 2006. Analyzing rock magnetic measurements: The RockMagAnalyzer 1.0 software. Computers & Geosciences 32, 1420–1431. doi:10.1016/j.cageo.2006.01.006 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Leymerie, A., 1862. Aperçu géognostique des Petites Pyrénées et particulièrement de la montagne d’Ausseing. Bull. Soc. Geol. Fr. 19, 1091–1096. Liebau, A., 1973. El Maastrichtiense lagunar (Garumniense) de Isona. Libro-Guıa, XIII Coloquio Europeo Micropaleontologıa 87–112. Llompart, C., 1979. Yacimiento de huellas de pisadas de reptil en el Cretácico prepirenaico. Acta geológica hispánica 14, 333–336. López-Martinez, N., Ardevol, L., Arribas, M.E., Civis, J., Gonzalez-Delgado, A., 1998. The geological record in non-marine environments around the K/T boundary (Tremp Formation, Spain). Bulletin de la Societe Geologique de France 169, 11. López-Martínez, N., Arribas, M.E., Robador, A., Vicens, E., Ardèvol, L., 2006. Los carbonatos danienses (Unidad 3) de la FM Tremp (Pirineos Sur-Centrales) : paleogeografía y relación con el límite Cretácico-Terciario. Revista de la Sociedad Geológica de España 19, 233– 255. López-Martínez, N., Fernández-Marrón, M.T., Valle, M.F., 1999. The succession of vertebrates and plants across the Cretaceous-Tertiary boundary in the Tremp Formation, Ager valley (south-central Pyrenees, Spain). Geobios 32, 617–627. López-Martınez, N., Canudo, J.I., Ardèvol, L., Suberbiola, X.P., Orue-Etxebarria, X., Cuenca-Bescós, G., Ruiz-Omeñaca, J.I., Murelaga, X., Feist, M., 2001. New dinosaur sites correlated with Upper Maastrichtian pelagic deposits in the Spanish Pyrenees: implications for the dinosaur extinction pattern in Europe. Cretaceous Research 22, 41–61. doi:10.1006/cres.2000.0236 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Mannion, P.D., Upchurch, P., Carrano, M.T., Barrett, P.M., 2011. Testing the effect of the rock record on diversity: a multidisciplinary approach to elucidating the generic richness of sauropodomorph dinosaurs through time. Biological Reviews 86, 157–181. doi:10.1111/j.1469185X.2010.00139.x Marmi, J., Blanco, A., Fondevilla, V., Dalla Vecchia, F.M., Sellés, A.G., Vicente, A., Martín-Closas, C., Oms, O., Galobart, À., 2016. The Molí del Baró-1 site, a diverse fossil assemblage from the uppermost Maastrichtian of the southern Pyrenees (north-eastern Iberia). Cretaceous Research 57, 519–539. doi:10.1016/j.cretres.2015.06.016 Marmi, J., Luján, Á.H., Riera, V., Gaete, R., Oms, O., Galobart, À., 2012. The youngest species of Polysternon: A new bothremydid turtle from the uppermost Maastrichtian of the southern Pyrenees. Cretaceous Research 35, 133–142. doi:10.1016/j.cretres.2011.12.004 Marty, D., 2001. Sedimentology, Paleoecology and Stratigraphy of the Palustrine Facies Rognacien at the K/T-boundary in Southern France. Geologisch-Paläontologisches Institut. Masriera, A., Ullastre, J., 1983. Essai de synthèse stratigraphique des couches continentales de la fin du Crétacé des Pyrenées catalanes (NE de lEspagne). Géologie Méditerranéenne 20, 283–290. Médus, J., Feist, M., Rocchia, R., Batten, D.J., Boclet, D., Colombo, F., Tambareau, Y., Villatte, J., 1988. Prospects for recognition of the palynological Cretaceous Tertiary boundary and an iridium anomaly in nonmarine facies of the eastern Spanish Pyrenees: a preliminary report. Newsletters on Stratigraphy 18, 123–138. doi:10.1127/nos/18/1988/123 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Mey, P.H.W., Nagtegaal, P.J.C., Roberti, K.J., Hartevelt, J.J.A., 1968. Lithostratigraphic subdivision of post-Hercynian deposits in the south-central Pyrenees, Spain. Leidse Geologische Mededelingen 41, 221–228. Moreno-Azanza, M., Bauluz, B., Canudo, J.I., Puértolas-Pascual, E., Sellés, A.G., 2014. A re-evaluation of aff. Megaloolithidae eggshell fragments from the uppermost Cretaceous of the Pyrenees and implications for crocodylomorph eggshell structure. Historical Biology 26, 195– 205. doi:10.1080/08912963.2013.786067 Muñoz, J.A., 1992. Evolution of a continental collision belt: ECORS-Pyrenees crustal balanced cross-section. Thrust tectonics 235–246. Murelaga, X., Canudo, J.I., 2005. Descripción de los restos de quelonios del Maastrichtiense superior de Aren y Serraduy (Huesca). Geogaceta 38, 51–54. Nagtegaal, P.J.C., Van Vliet, A., Brouwer, J., 1983. Syntectonic coastal offlap and concurrent turbidite deposition: The Upper Cretaceous Aren sandstone in the South-Central Pyrenees, Spain. Sedimentary Geology 34, 185–218. doi:10.1016/0037-0738(83)90086-6 Nakkady, S.E., 1950. A new foraminiferal fauna from the Esna Shales and Upper Cretaceous Chalk of Egypt. Journal of Paleontology, 675–692. Narváez, I., Brochu, C.A., Escaso, F., Pérez-García, A., Ortega, F., 2016. New Spanish Late Cretaceous eusuchian reveals the synchronic and sympatric presence of two allodaposuchids. Cretaceous Research 65, 112–125. doi:10.1016/j.cretres.2016.04.018 Navarrete, R., Rodríguez-López, J.P., Liesa, C.L., Soria, A.R., Veloso, F. de M.L., 2013. Changing physiography of rift basins as a control on the evolution of mixed siliciclastic– MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT carbonate back-barrier systems (Barremian Iberian Basin, Spain). Sedimentary Geology 289, 40– 61. doi:10.1016/j.sedgeo.2013.02.003 Nederbragt, A.J., 1991. Late Cretaceous Biostratigraphy and Development of Heterohelicidae (Planktic Foraminifera). Micropaleontology 37, 329–372. https://doi.org/10.2307/1485910 Ogg, J.G., Hinnov, L.A., Huang, C., 2012. Chapter 27 - Cretaceous, in: The Geologic Time Scale. Elsevier, Boston, 793–853. https://doi.org/10.1016/B978-0-444-59425-9.00027-5 Oms, O., Dinarès-Turell, J., Vicens, E., Estrada, R., Vila, B., Galobart, À., Bravo, A.M., 2007. Integrated stratigraphy from the Vallcebre Basin (southeastern Pyrenees, Spain): New insights on the continental Cretaceous−Tertiary transition in southwest Europe. Palaeogeography, Palaeoclimatology, Palaeoecology 255, 35–47. doi:10.1016/j.palaeo.2007.02.039 Oms, O., Fondevilla, V., Riera, V., Marmi, J., Vicens, E., Estrada, R., Anadón, P., Vila, B., Galobart, À., 2016. Transitional environments of the lower Maastrichtian South-Pyrenean Basin (Catalonia, Spain): The Fumanya Member tidal flat. Cretaceous Research 57, 428–442. doi:10.1016/j.cretres.2015.09.004 Oms, O., Pérez-Cano, J., Fondevilla, V., Anadón, P., Ibáñez-Insa, J., Rejas, M., Fernandez-Turiel, J.L., Dinarès-Turell, J., Pardo, P., Estrada, R., 2014. On the search of the KPG boundary in the terrestrial record in Europe: preliminary geochemical data from the Puig Pedrós section (Tremp basin, Catalonia, Spain). Abstract book Reconstructing the terrestrial endCretaceous paleoenvironments in Europe (Tremp, 2014). Paleontologia i Evolució Special memoir 7, 74. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Ortega, F., Bardet, N., Barroso-Barcenilla, F., Callapez, P.M., Cambra-Moo, O., Gómez, V.D.-, Díaz, V.D., Domingo, L., Elvira, A., Escaso, F., García-Oliva, M., Gómez, B., Houssaye, A., Knoll, F., Marcos-Fernández, F., Martín, M., Mocho, P., Narváez, I., García, A.P.- , Peyrot, D., Segura, M., Serrano, H., Torices, A., Vidal, D., Sanz, J.L., 2015. The biota of the Upper Cretaceous site of “Lo Hueco” (Cuenca, Spain). Journal of Iberian Geology 41. Osete, M.-L., Gialanella, P.-R., Gómez, J.J., Villalaín, J.J., Goy, A., Heller, F., 2007. Magnetostratigraphy of Early–Middle Toarcian expanded sections from the Iberian Range (central Spain). Earth and Planetary Science Letters 259, 319–332. doi:10.1016/j.epsl.2007.04.048 Panaiotu, A.G., Ciobănete, D., Panaiotu, C.G., Panaiotu, C.E., Csiki, Z., 2011. New palaeomagnetic data from the Haţeg Basin, Romania, in: Abstract Book, 8th Romanian Symposium of Paleontology, Bucharest (Romania). Ed. Ars Docendi, Bucharest, 84–85. Panaiotu, C.G., Panaiotu, C.E., 2010. Palaeomagnetism of the Upper Cretaceous Sânpetru Formation (Haţeg Basin, South Carpathians). Palaeogeography, Palaeoclimatology, Palaeoecology 293, 343–352. doi:10.1016/j.palaeo.2009.11.017 Peláez-Campomanes, P., López-Martínez, N., Álvarez-Sierra, M.A., Daams, R., 2000. The Earliest Mammal of the European Paleocene: The Multituberculate Hainina. Journal of Paleontology 74, 701–711. Pereda-Suberbiola, X., Canudo, J.I., Cruzado-Caballero, P., Barco, J.L., LópezMartínez, N., Oms, O., Ruiz-Omeñaca, J.I., 2009. The last hadrosaurid dinosaurs of Europe: A new lambeosaurine from the Uppermost Cretaceous of Aren (Huesca, Spain). Comptes Rendus Palevol 8, 559–572. doi:10.1016/j.crpv.2009.05.002 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Vila, B., Sellés, A.G., Brusatte, S.L., 2016. Diversity and faunal changes in the latest Cretaceous dinosaur communities of southwestern Europe. Cretaceous Research 57, 552–564. doi:10.1016/j.cretres.2015.07.003 Villalaín, J.J., Osete, M.L., Vegas, R., García-Dueñas, V., Heller, F., 1994. Widespread Neogene remagnetization in Jurassic limestones of the South-Iberian palaeomargin (Western Betics, Gibraltar Arc). Physics of the Earth and Planetary Interiors 85, 15–33. doi:10.1016/00319201(94)90005-1 Villalba-Breva, S., Martín-Closas, C., 2013. Upper Cretaceous paleogeography of the Central Southern Pyrenean Basins (Catalonia, Spain) from microfacies analysis and charophyte biostratigraphy. Facies 59, 319–345. doi:10.1007/s10347-012-0317-1 Villalba-Breva, S., Martín-Closas, C., 2011. A characean thallus with attached gyrogonites and associated fossil charophytes from the Maastrichtian of the Eastern Pyrenees (Catalonia, Spain). Journal of Phycology 47, 131–143. doi:10.1111/j.1529-8817.2010.00947.x Villalba-Breva, S., Martín-Closas, C., Marmi, J., Gomez, B., Fernández-Marrón, M.T., 2012. Peat-forming plants in the Maastrichtian coals of the Eastern Pyrenees. Geologica Acta: an international earth science journal 10. doi:http://dx.doi.org/10.1344/105.000001711 Vogler, J., 1941. Ober-Jura und Kreide von Misol. Beitrage zur Geologie von Niederlandisch-Indian 4, 243–293. Voorwijk, G.H., 1937. Foraminifera from the upper Cretaceous of Habana, Cuba. Akad. Wet., Amsterdam, Sect. Sci. 40, 190–198. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Vremir, M., Bălc, R., Csiki-Sava, Z., Brusatte, S.L., Dyke, G., Naish, D., Norell, M.A., 2014. Petreşti-Arini – An important but ephemeral Upper Cretaceous continental vertebrate site in the southwestern Transylvanian Basin, Romania. Cretaceous Research 49, 13–38. doi:10.1016/j.cretres.2014.02.002 Figure captions Figure 1. Geographical and geological situation of the Serraduy sector (Huesca, Spain). A. location of the Pyrenees within the Iberian Peninsula; B. geological map of the Serraduy sector; the red stars indicate the areas with the highest concentration of vertebrate sites; the red rectangles indicate the magnetostratigraphic sections (SB, SR) studied in this work; C. composite stratigraphic section of the studied profiles and stratigraphic vertebrate distribution. Figure 2. Lithostratigraphic subdivision of the Tremp Formation according to different authors. Modified from Cuevas (1992) and Riera (2010). Figure 3. A. correlation panel of the stratigraphic profiles and vertebrate sites studied in this work; B. detailed geological map of the Serraduy sector and guide level locations. Figure 4. Landscape views of the main areas with outcrops and paleontological vertebrate sites in the Serraduy sector. A. outcrop of the Larra stratigraphic section located on the west side of the Isábena River; B. succession of the Arén and Tremp Formation in the sector located west of the Isábena River; C. succession of the Arén and Tremp Formation in the Barranco Serraduy sector located east of the Isábena River. Red stars point to the main areas with vertebrate sites; the MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT white dotted lines mark the contact between different units; and the colored dotted lines point to the location of the different guide levels. Figure 5. Lithology, paleontological site positions, and the proposed magnetostratigraphy from the Barranco Serraduy section. VGP latitude logs along the Arén Formation (SB) and Tremp Formation (SR) profiles are shown by circles (black or white) when the polarity has been calculated from paleomagnetic directions obtained by principal component analysis (PCA) or by bars (white) when it has been obtained by great circle analysis (GCA). Figure 6. A to D. demagnetization diagrams, in geographic coordinates, showing the paleomagnetic behaviors in representative samples of carbonatic rocks; A to B. samples from the SB and SR sections showing an overlap between two components with opposite direction; D. sample of the SR section with a normal polarity component at low temperatures and a dispersed high temperature cluster; E. equal area projection of the MB component and the demagnetization great circles calculated in these rocks, before and after bedding correction (BBC and ABC respectively). Note the path from normal polarity to reverse polarity followed by all samples; F. representative rock magnetic experiments of this lithology. From left to right, acquisition of the isothermal remanent magnetization (IRM), non-corrected and corrected hysteresis loop, and thermomagnetic curve. The low coercivity may indicate the presence of magnetite as the main magnetic mineral, as can be observed in the IRM and hysteresis loops; the thermomagnetic curve shows an important growth of magnetite up to 400ºC. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Figure 7. A to D. demagnetization diagrams showing, in geographic coordinates, the paleomagnetic behaviors in representative samples of red beds; A. components RA and RB with the same polarity; B and C. component RB does not go to the origin and a non-zero end cluster reveals component RA with reversed polarity; in SR78-2B, RA is partially demagnetized before the growth of the magnetic mineral during heating (up to 620ºC); D. RA and RB with normal polarity; E. representative rock magnetic experiments of this lithology. From left to right, IRM, hysteresis loop and thermomagnetic curve. The high coercivity observed in the IRM and in the hysteresis loop and the presence of a magnetic phase with Curie temperatures above 620ºC indicate the presence of hematite as the main magnetic phase. The higher magnetization of the cooling curve indicates the growth of magnetite during heating. Figure 8. Equal-area projection of component A (A) and component B (B) with their respective Fisher means (Fisher, 1953), before and after bedding correction (BBC and ABC respectively); C. calculated demagnetization great circles (GC) and mean direction of both components; note that both components overlap with the GC mean intersection, component A being almost coincident with its mean. n: number of samples. NW-SW GCs correspond with carbonates and NE-SW GCs with red beds. Figure 9. Some representative planktonic and benthic foraminifer species identified in the Barranco Serraduy section. From left to right and from top to bottom: Laeviheterohelix glabrans (MPZ 2018/25), Heterohelix globulosa (MPZ 2018/23), Guembelitria blowi (MPZ 2018/28), Hedbergella flandrini (MPZ 2018/24), Ventilabrella eggeri (MPZ 2018/22), Pseudoguembelina hariaensis (MPZ 2018/21), Globigerinelloides praevolutus (MPZ 2018/27), Globotruncana MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT bulloides (MPZ 2018/26), Contusotruncana fornicata (MPZ 2018/20), Favusella washitensis (MPZ 2018/18), Gyroidinoides sp. (MPZ 2018/19). Figure 10. Vertebrae of Hadrosauridae indet. from Serraduy. A. caudal (juvenile MPZ 2017/796 from AM2 site); B. caudal (adult MPZ 2017/797 from LAR3 site); C. caudal (juvenile MPZ 2017/798 from LAR3 site); D. cervical (adult? MPZ 2017/799 from LAR3 site). In posterior/anterior, dorsal, lateral and ventral views respectively. Scale bar = 3cm. Figure 11. Dinosaur remains from Serraduy. A. coracoid of Hadrosauridae indet. (MPZ 2017/800 from LAR2 site) in medial and lateral views; B. phalanx of Hadrosauridae indet. (MPZ 2017/801 from LAR3 site) in dorsal, lateral and ventral views; C. tibia (distal fragment) of Hadrosauridae indet. (MPZ 2017/802 from LAR3 site) in anterior, posterior and distal views; D. femur (proximal fragment) of Hadrosauridae indet. (MPZ 2017/803 from LAR3 site) in anterior, lateral, posterior, medial and proximal views; E. femur (proximal fragment) of Titanosauria indet. (MPZ 99/143 from femur site) in posterior view; F. teeth of Theropoda indet. (MPZ 2017/804 from 172-i/04/e site) in lingual and labial view (white boxes show the detailed denticles). Scale bar without number = 3cm. Figure 12. Dinosaur ichnites of Hadrosauropodus indet. from Serraduy (Pedregal site). Figure 13. Crocodylomorph (Eusuchia) remains from Serraduy. A. teeth of cf. Thoracosaurus (MPZ 2017/806 from EXT site); B. teeth of cf. Allodaposuchidae (MPZ 2017/807 from DL3 site); C. teeth of cf. Allodaposuchidae (MPZ 2017/808 from PED site); D. dorsal vertebra of MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Eusuchia indet. (MPZ 2017/805 from LAR3B site); E. skull holotype of Agaresuchus subjuniperus (MPZ 2012/288) (from AM3 site). Figure 14. Chronostratigraphic framework with indication of lithostratigraphy, paleomagnetic data, correlation and the K/Pg transition in the main Maastrichtian South Pyrenean continental sections with magnetostratigraphic data. Table 1. Vertebrate faunal list for Serraduy (Huesca, Spain), upper Maastrichtian. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Site Abbreviation Taxa Site Abbreviation Taxa 172-i/04/a 172-i/04/a Hadrosauridae indet. Camino Rin 2 CRIN2 Hadrosauridae? indet. Theropoda? indet. Pterosauria? indet. 172-i/04/b 172-i/04/b Dinosauria indet. Dolor 1 DL1 Hadrosauridae? indet. 172-i/04/c 172-i/04/c Dinosauria indet. Dolor 2 DL2 Dinosauria indet. Avialae? indet. Bothremydidae indet. 172-i/04/d 172-i/04/d Dinosauria indet. Dolor 3 DL3 Hadrosauridae indet. Avialae? indet. Bothremydidae indet. cf. Allodaposuchidae 172-i/04/e 172-i/04/e Hadrosauridae indet. Theropoda indet. Barranco Extremadura EXT Dinosauria indet. Hadrosauridae indet. Bothremydidae indet. cf. Thoracosaurus 172-i/04/f 172-i/04/f Hadrosauridae indet. Fornons 1 F1 Dinosauria indet. Hadrosauridae indet. Amor 1 AM1 Dinosauria indet. Hadrosauridae indet. Bothremydidae indet. Fornons 2 F2 Dinosauria indet. Hadrosauridae? indet. Amor 2 AM2 Hadrosauridae indet. Bothremydidae indet. Fornons 3 F3 Dinosauria indet. Theropoda? indet. Amor 3 AM3 Dinosauria indet. Hadrosauridae? indet. Bothremydidae indet. Agaresuchus subjuniperus Larra 1 LAR1 Vertebrata indet. Eusuchia indet. Barranco Serraduy 1 BS1 Dinosauria indet. Hadrosauridae indet. Larra 2 LAR2 Dinosauria indet. Ornithopoda indet. Hadrosauridae indet. Bothremydidae indet. cf. Allodaposuchidae Barranco Serraduy 2 BS2 Dinosauria indet. Hadrosauridae indet. Larra 3 LAR3 Dinosauria indet. Hadrosauridae indet. Eusuchia indet. Barranco Serraduy 3 BS3 Vertebrata indet. Larra 4 LAR4 Hadrosauridae indet. Coelurosauria indet. Bothremydidae indet. cf. Allodaposuchidae Discoglossidae indet. Barranco Serraduy 4 BS4 Dinosauria indet. Hadrosauridae indet. Sauropoda indet. Bothremydidae indet. Eusuchia indet. Larra 5 LAR5 Hadrosauridae indet. Barranco Serraduy 5 BS5 Hadrosauridae indet. Bothremydidae indet. Larra 6 LAR6 Dinosauria indet. Bothremydidae indet. Camino Fornons 1 CF1 Dinosauria indet. Hadrosauridae indet. Theropoda? indet. Sauropoda? indet. Osteichthyes indet. Bothremydidae indet. Pedregal PED cf. Allodaposuchidae Camino Fornons 2 CF2 Hadrosauridae indet. cf. Allodaposuchidae Rin 1 y 2 RIN1-2 Dinosauria indet. Color COL Dinosauria indet. Hadrosauridae indet. Bothremydidae indet. San Cristobal SCRI Dinosauria indet. Hadrosauridae indet. Camino Rin 1 CRIN1 Theropoda? indet. Sierra de Sis 1 SIS1 Dinosauria indet. Hadrosauridae indet. Bothremydidae indet. Sierra de Sis 2 SIS2 Dinosauria indet. Hadrosauridae indet. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT