THE SIGNIFICANCE OF THE BARRANCO DEL GREDERO SECTION, (CARAVACA, SE SPAIN) IN THE K/T BOUNDARY DEBATE.

JAN SMIT
Dept Sedimentology, Vrije Universiteit Amsterdam, de Boelelaan 1085, 1081HV Amsterdam, Netherlands.
The Barranco del Gredero (Loma de Solana tectonic unit) section  has attracted scientists since the 1950's (a.o. Pablo Fallot, [1] J. Paquet, [2] G. van Veen [3] and A. von Hillebrandt ) because of its exceptional thickness, completeness and richness of well-preserved foraminifers (fig1).
The section remains in the top 5 of complete K/T sections and is still the subject of detailed investigations, e.g. [4] , [5, 6] The Milankovitch-style, well-bedded section from early-Maastrichtian to Lower Eocene, is over 225m thick (fig 2) (Jorquera fm., van Veen [3] developed in a bathyal, hemipelagic facies. By comparison, the same interval in the famous Gubbio section is only 10.5m thick. Analyses of mesopelagic fishteeth (Pat Doyle pers. comm.) suggest a depth of deposition of more than 1000m near the K/T boundary. The Loma de Solana is usually placed in the allochtonous Subbetic units, but the striking lithological similarities with the Agost, Relleu and Finestrat sections further east, demonstrate that the Loma de Solana unit is part of the par-authochtonous Prebetic, the southern margin of the Iberian continent. Van Veen [3] , p.116) was one of the first to recognize that the K/T boundary interval was uniquely well exposed in the barranco del Gredero.

Two mastersprojects of C. Hermes, measured the Gredero section in detail. [7] subdivided the Paleocene, and his student [8] recognized the presence of the P. eugubina Zone, but analysed only three samples from the K/T interval. Hermes (pers comm) and [8] mentioned a 10 clayinterval at the K/T boundary, but paid no further attention. [9] (fig 3) analysed the K/T interval in further detail, every cm in the K/T transitional interval, including the boundary clay (BCL). He found the red laminae (sample sm75-503), that most people believe it is the Chicxulub impact ejecta layer. A new biozone was established [9] between the top of the Maastrichtian and the base of the G. eugubina Zone; the G. cretacea   or the P0 zone, whose type locality therefore is in the Gredero section. A paleomagnetic analysis of the Gredero section was performed by G. Brunsmann (Univ Amsterdam) [10] (fig4). The magnetostratigraphy of the uppermost 100m of the Maastrichtian and the basal 70m of the Paleocene showed magnetochrons C31 to C27, very well comparable to the classic section at Gubbio, and the Agost section 100km further east [11] ). The same sample set was subjected to analyses by neutron activation in 1977 at Delft University. The results, obtained spring 1978, showed a strongly anomalous content of Cr, Ni, As, Sb, Zn in sample Sm75-503, 15-250 times enriched opposite background values, but not (yet) the iridium anomaly [12] . Nickel of course is often associated with Ni-iron meteorites, and values of 2000ppm (of sm75-503) are rare anywhere on earth. Around the same time, the Alvarez group published the find of anomalous iridium in Gubbio at the 1978 AGU fall meeting, by a similar, but more sensitive neutron activation analysis. Jan Hertogen of Gent university showed in 1979 the presence of anomalous Ir in Gredero as well, particularly in sample 75-503.            

The extremely short extinction interval <0.5cm), already called for a catastrophic extinction because in the preceding >100m thick interval very little happened in terms of planktic foraminiferal changes (only G. gansseri disappeared 10 m below K/T). The strong adaptive radiation of foraminifers within the first 50cm above the boundary clay, further strengthened a catastrophic cause. The iridium anomaly, by pointing at a large impact event, provided a plausible cause for the catastrophe.

Fig 3 Detailed lithological column of the K/T interval. (after [9])

Fig 2. Lithological column of the Loma de Solana section. (after [3])

The Gredero section remained a key player in the investigations that followed the launch of the impact-extinction theory.

 
Smit and Klaver (13] discovered the tiny (50-500µm) 'sanidine' spherules in sm75-503, that later became known as microkrystites, now best explained as condensation droplets from the hot impact vapour cloud. Smit (1977) initially regarded these as "gypsum nodules", because gypsum is abundant in these pelagic sediments. Gypsum, unfortunately, has the same refraction index as sanidine, so they went unnoticed for several years. Gerard Klaver (NITG/TNO) recognized the quench texture within the sanidine spherules, because he had observed similar textures in the chilled crust of pillow-lava basalts on the island of Bonaire [14] There remained a problem: K-spar is characteristic of K-rich, very viscous igneous rocks, not the ideal environment for producing quench crystals, on the contrary. Also in Furlo and Petriccio (Italy) were these sandine spherules found in the Ir-rich layer, but in these localities accompanied by dark clay-rich spherules, full of quench crystals of magnesioferrite, rich in iridium. These were much more mafic, consistent with the quench crystallinity, but hard to reconcile with K-spar spherules, assumed that they are from the same source. The Cr-rich magnesioferrite and chromite quench crystals also occur in 75-503 in great abundance (fig5), but are seemingly floating in the clay matrix, as if they were formed directly by condensation. Close inspection of 75-503 shows that the spinels occur in clusters, outlining flattened spherules. The same textures in the sanidine spherules were also found in spherules of pure smectite (Fonte d'Olio, Bidart), in arsenopyrite spherules from Zumaya, and goethite from New Zealand and Tetri-Tskaro. It became clear that the Kspar quench crystals were pseudomorph alteration products of another precursor, preferably a mafic mineral. The high ?18O (+25%o) of the sanidine [15] confirmed the low temperature origin. The precursor mineral was discovered at DSDP site 577, Shatsky rise Pacific; a Ca-rich augite, clinopyroxene (fig6).
Figure 6 SEM graphs of skeletal Ca-rich clinopyroxene of DSDP site 577, Shatsky Rise Pacific (left). Small crystals of K-spar (sandine) from Caravaca (sm75-503), pseudomorph arranged after cpx(right)

Figure 5. SEM graph of skeletal Ni-Cr-rich spinels from sample 75-503, Gredero section

The next important discovery made at the Gredero section, are the chondritic ratios of all the platinum group elements (PGE) [16] , an important indicator that the Ir anomaly was not from a terrestrial (i.e. volcanic) source, as suggested by some. It is remarkable that the magnetic residues of 75-503, consisting primarily of chromite and magnesioferrite, are highly enriched in Ir and Cr (table1)
Wolbach et al [17] determined an anomalous amount of soot in 75-503, indicative of large wildfires just after impact. It is curious that at the "twin" site Agost not a trace of soot has been determined, while all other characteristics (lithology, biotic changes, geochemistry) are almost identical.

Last but not least, sample 75-503 yielded an important clue to the composition of the impacting bolide, by the finding of the anomalous Cr-isotope values [18].

The concentration profile of both Cr and Ir is almost identical in all the K/T sites where both Cr and Ir were measured, both in continental and marine sites. The strongest peak in the ejecta layer (Sm75-503). The Caravaca ejecta layer holds the K/T boundary record in terms of Cr concentration. It is safe to say that whatever the source of the iridium, it must also have been the source for the excess Cr. The 53Cr/52Cr ratio on earth is everywhere the same, whether from the core, mantle or crust (epsilon e earth=0). Extraterrestrial matter has a different 53Cr/52Cr. Sample 75-503 yielded a strong negative eCr ratio, identical to carbonaceous chondrites, unlike ordinary chondrites and asteroids that have a positive eCr ratio. This finding excludes any hypothesis for a terrestrial origin of Cr and Ir at K/T.

            Stable isotope analyses have played a major role in establishing the magnitude of the K/T mass-extinctions. T. Romein [19] was the first to analyse stable isotopes in detail across the K/T boundary in the Gredero section. He was the first to find the short-term strong negative(2.5 o/oo ?13C) anomaly, that later became known as the "Strangelove ocean" condition. In these conditions the vertical ?13C gradient in the oceans temporarily disappeared during deposition of the boundary clay. The ?18O profile of Caravaca [20] suggests that sea surface temperatures have risen considerably at the base of the boundary clay. Kaiho[21] and others confirmed these anomalies in detail, and demonstrated that the vast majority of Cretaceous species above the K/T boundary are reworked specimens.

Benthic foraminiferal [22] communities collapsed suddenly directly above the K/T boundary. Only a few taxa remained under anoxic conditions. The benthic taxa must have temporarily migrated elsewhere, because the majority returned to the Gredero sea floor when oxygenation improved.

Fig 4 Detailed lithological column of the Gredero section, from 95m below to 85m above the K/T boundary. The magnetostratigraphy is from G. Brunsmann (after [23]). Important planktic FAD/LAD are indicated

Burrowing organisms likewise temporarily disappeared in the region, probably due to lack of oxygen on the seafloor. The ejecta layers in the sections of Agost, Relleu and Gredero are almost completely intact, not bioturbated. Frequent burrows can be observed in the top Maastrichtian (zoophycos, chondrites, thalassinoides), but those, scavengers of the seafloor, are filled with dark boundary clay, not with ejecta layer debris, which showed that they became active substantially after deposition of the ejecta layer, in contrast with similar borrows from Italy (Furlo, Petriccio) and Bidart., that often contain the debris of ejecta (spherules).