High Stress Late Maastrichtian Paleoenvironment: Inference from Planktonic Foraminifera in Tunisia

Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 www.elsevier.com/locate/palaeo

High stress late Maastrichtian paleoenvironment: inference from planktonic foraminifera in Tunisia

Sigal Abramovich *, Gerta Keller

Department of Geosciences, Princeton University, Princeton, NJ 08544, USA

Received 10 July 1999; accepted 9 August 2001

Abstract

High resolution (V5^10 kyr) planktonic foraminiferal analysis at Elles, Tunisia, reveals major changes in the structure of the Tethyan marine ecosystem during the upper Maastrichtian. During the first 1.5 Myr of the late Maastrichtian (68.3^66.8 Ma) relatively stable environmental conditions and cool temperatures are indicated by diverse planktonic foraminiferal populations with abundant intermediate and surface dwellers. A progressive cooling trend between V66.8^65.45 Ma resulted in the decline of globotruncanid species (intermediate dwellers). This group experienced a further decline at the climax of a rapid warm event about 300 kyr before the K^T boundary. At the same time relative abundances of long ranging dominant species fluctuated considerably reflecting the high stress environmental conditions. Times of critical high stress environments during the late Maastrichtian, and particularly at the K^T boundary, are indicated by low species diversity and blooms of the opportunistic genus Guembelitria at warm^ cool transition intervals. During the last 100 kyr of the Maastrichtian rapid cooling is associated with accelerated species extinctions followed by the extinction of all tropical and subtropical species at the K^T boundary. ß 2002 Elsevier Science B.V. All rights reserved.

Keywords: planktonic foraminifera; biostratigraphy; upper Maastrichtian; Tunisia; paleoecology; paleoclimate

1. Introduction Canudo et al., 1991; Keller and Benjamini, 1991; Olsson and Liu, 1993; Keller et al., 1995, 1997; Until recently most studies of Maastrichtian Pardo et al., 1996; Appellaniz et al., l997; Lucia- planktonic foraminiferal populations that aimed ni, 1997; Molina et al., 1998). During the last few to describe the nature of the Cretaceous^Tertiary years, it has become increasingly evident that the (K^T) boundary mass extinction focused in extra- nature of this mass extinction cannot be e¡ec- ordinary detail on the K^T transition itself, which tively addressed until the background variations represents at best a few hundred thousand years of populations and their relationship to environ- (e.g. Smit, 1982, 1990; Keller, 1988, 1989a, 1993; mental parameters in the Maastrichtian ecosystem are understood (e.g. Keller, 1996). To date, only a few stratigraphically well constrained studies of a * Corresponding author. Tel.: +1-609-258-4117; more complete Maastrichtian faunal record have Fax: +1-609-258-1671. been published, and these concern faunas from E-mail address: [email protected] (S. Abramovich). the Brazos River, Texas (Keller, 1989b), Denmark

PALAEO 2753 1-5-02 146 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

(Schmitz et al., 1992), the Southern Ocean (Hu- ber, 1992), mid-latitude South Atlantic (Li and Keller, l998a), Tunisia (Li and Keller, 1998b), and the Negev, Israel (Abramovich et al., l998). These studies have shown that the deterioration of Cretaceous planktonic foraminiferal populations began in the late Maastrichtian and accelerated over the last few hundred thousand years of the Maastrichtian. The onset of the decline was accompanied by blooms of the opportunistic genus Guembelitria which demonstrate the instability of the pelagic ecosystem during the last few million years of the Cretaceous (Abramovich et al., 1998). The faunal decline climaxed at the K^ T boundary with the mass extinction of all specialized tropical^subtropical planktonic foraminiferal species, whereas the cosmopolitan and eco- logically generalist species survived into the Danian (for a summary see Keller et al., 2002). Recent studies of the Maastrichtian stable isotopic records from DSDP Sites 525A and 21 (Li and Keller, 1998a,c) leave no doubt that global Fig. 1. Maastrichtian paleogeography of Tunisia (after Burol- climatic deterioration occurred during the Maas- let, 1967). Note the paleogeographic location of the Elles sec- trichtian, and was accompanied by signi¢cant sea tion between the emergent zone of the Kasserine Island and El Kef section. level £uctuations (see also Barrera, 1994; Barrera et al., 1997; Barrera and Savin, l999). Further- more, the timing of these events strongly suggests that the Late Cretaceous faunal decline was trig- Most studies of Maastrichtian and Paleogene gered by environmental changes. This suggests sections in Tunisia have concentrated on the that a bolide impact at the K^T boundary was K^T boundary transition at the El Kef stratotype not the sole causal mechanism for the K^T mass (e.g. early studies summarized by Salaj (1980) and extinction. Further detailed high resolution re- later studies by Keller et al. (1995)). High resolu- cords that evaluate the Maastrichtian ecosystem tion microfossil and geochemical studies have stress are still needed if we are to understand the generally focused on the 50^100-cm interval be- relationship between environment and biotic per- low and above the K^T boundary. Maastrichtian turbations leading up to the K^T boundary. studies are few and generally of a stratigraphic Among the regions of high potential for obtaining nature (Salaj, l980) with the exception of Neder- high resolution records are the sections from the bragt (l991) who studied the Heterohelicidae at northern part of Tunisia. El Kef. A more complete quantitative study of The Upper Cretaceous sequence of northern the entire Maastrichtian planktonic foraminiferal Tunisian was deposited in a series of basins which fauna and various geochemical analyses (stable surrounds a tectonically emergent zone, the Kas- isotopes, mineralogy, trace element geochemistry) serine Island (Fig. 1), the major source for clastic at El Kef and Elles has recently been published material into the basins (Adatte et al., 1998; Ben- by Li and Keller (1998b) and Li et al. (1999, salem, 1998). The sediments within the northwest- 2000). ern basin are characterized by abundant pelagic The objective of this study is to evaluate the microfaunas deposited on the outer shelf to upper late Maastrichtian faunal turnover at a section slope (Li and Keller, 1998b). located near the hamlet of Elles in north-central

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 147

Tunisia (Fig. 1). In this section planktonic fora- clean foraminiferal residue was recovered, which miniferal populations were quantitatively ana- was oven-dried at 50‡C. Planktonic foraminiferal lyzed at 10^20-cm intervals, or on average one tests are recrystallized, but preservation of test sample every V5^10 kyr, for the last million morphology is very good. Population counts for years of the Maastrichtian. planktonic foraminifera are based on random sample splits using a micro-splitter. From each sample approximately 250^300 planktonic fora- 2. Location, lithology and methods minifera were picked from each of two size fractions ( s 63 Wm and s 150 Wm Background Data The Elles section is located in the Karma valley Set1). Only the larger s 150-Wm size fraction of near the hamlet of Elles, about 75 km southeast the uppermost meter of the section was analyzed of the city of El Kef and the nearby K^T bound- for this report; the smaller s 63-Wm size fraction ary stratotype section. The Maastrichtian se- was studied by Keller et al. (2002). These two quence exposed in the Karma valley was depos- size fractions were analyzed in order to obtain ited at middle to outer shelf depths and relatively statistically signi¢cant representations of the close to the emergent zone of the Kasserine Island smaller and larger species. At the same time, (Fig. 1). This position explains the high terrige- the quantitative study of two population splits nous content in Maastrichtian sediments at Elles, reduces the bias in ¢rst and last appearances and the unusually high sedimentation rate (e.g. due to the Signor^Lipps e¡ect (Signor and Lipps, V4 cm/1000 yr, Li and Keller, 1998b; Adatte et 1982). Species identi¢cations are based on de- al., 2002). The very high sediment accumulation scriptions of Smith and Pessagno (1973), Robas- rate at Elles, which exceeds that of the El Kef zynski et al. (1984), Caron (l985) and Nederbragt section (Li and Keller, 1998b), has great potential (1991). for accurately evaluating the timing of environmental changes during the late Maastrichtian and particularly during the last million years of 3. Biostratigraphy the Maastrichtian. Lithologically, the upper Maastrichtian in In previous studies, the upper Maastrichtian of northern Tunisia is represented by the El Haria low to middle latitude was generally assigned to Formation. Upper Maastrichtian sediments at El- the Abathomphalus mayaroensis Zone (Caron, les consist of relatively uniform marly shales with l985; Robaszynski and Caron, 1995). More re- a 25-cm-thick resistant layer of bioturbated marly cently, the interval spanning the last 200^300 limestone at about 26 m below the K^T bound- kyr has been assigned to the Plummerita hantke- ary. The K^T transition is marked by a 20-cm- ninoides Zone (Masters, l984; Pardo et al., 1996). thick channelized cross-bedded bioclastic layer Li and Keller (l998a) have further subdivided the (foraminiferal packstone), which underlies a 1- late Maastrichtian A. mayaroensis Zone based on cm dark gray clay and thin rust-colored layer. Cretaceous planktonic foraminiferal assemblages The rusty layer contains an iridium anomaly and the paleomagnetic time scale of DSDP Site and Ni-rich spinels which mark the K^T bound- 525A, and applied this chronology to the Tuni- ary (Adatte et al., 2002). A total of 191 samples sian sections at El Kef and Elles (Li and Keller, were collected from the last 31 m of the Maas- 1998b). Their zonal scheme subdivides the late trichtian. Samples were collected at 10-cm inter- Maastrichtian into four zones labeled CF1 to vals between 31.2 and 26.5 m, and at 20-cm in- CF4 (CF = Cretaceous Foraminifera) as applied tervals through the rest of the section. However, in this study (Fig. 2). near the boundary interval samples were collected at 1-cm intervals. In the laboratory, samples were disaggregated and washed through a s 63-Wm sieve until a 1 http://www.elsevier.com/locate/palaeo

PALAEO 2753 1-5-02 148 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

Fig. 2. Later Maastrichtian planktonic foraminiferal biozonation used in this study and comparison with the zonations of Li and Keller (l998b) and Robaszynski and Caron (l995).

3.1. Plummerita hantkeninoides zone Negev, Israel, 8 m (Canudo et al., l99l; Abramo- (CF1, 65^65.3 Ma) vich et al., 1998).

This zone is de¢ned by the total range of the 3.2. Pseudoguembelina palpebra zone nominate species. The zone was ¢rst de¢ned by (CF2, 65.3^65.45 Ma) Masters (1984), based on sections in Egypt. Pardo et al. (1996) calibrated the age and the duration of The base of this zone is de¢ned by the LA (last this zone as spanning the last 200^300 kyr of the appearance) of Gansserina gansseri, and the top Maastrichtian based on the paleomagnetic record by the FA (¢rst appearance) of Plummerita hant- at Agost, Spain. At Elles, biozone CF1 spans the keninoides. At Elles zone CF2 spans 12 m (from top 9.6 m of the Maastrichtian. This thickness 21.6 to 9.6 m) and represents the highest sedimen- represents the highest sedimentation rate (V3.2 tation rate (V8 cm/1000 yr) measured to date for cm/1000 yr) measured to date for this biozone. this interval. At the nearby El Kef section zone In contrast, zone CF1 at El Kef is 6 m thick, at CF2 is only 3 m thick due to a local fault that Agost in Spain 3.8 m, and at Hor Hahar in the truncates the base of the zone (Li and Keller,

Fig. 3. Planktonic foraminiferal species census data at Elles arranged in order of last appearances. Note that two-thirds of the upper Maastrichtian species are rare and sporadically present. These are tropical to subtropical species with large complex morphologies. Among these species, extinctions are progressive with 17 species disappearing during the late Maastrichtian. Sediment reworking is evident within the 50^100 cm below the K^T boundary where 12 of these species temporarily reappear.

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 149

PALAEO 2753 1-5-02 150 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

1998b). Based on the paleomagnetic records at imentological factors. Outside its preferred habi- Site 525A and Agost, Zone CF2 spans about tat, ecological stress may result in shortened spe- 150 kyr (Li and Keller, 1998a), though this corre- cies ranges or sporadic occurrences. Species which lation is tentative and needs further calibration at are rare and only sporadically present may not be other localities. encountered in standard biostratigraphic analysis and therefore result in bias towards shortened 3.3. Pseudoguembelina hariaensis zone species ranges, also known as the Signor^Lipps (CF3, 65.45^66.8 Ma) e¡ect (Signor and Lipps, l982). But species ranges in the sediment record can also be biased towards This zone is de¢ned by the FA of Pseudoguem- prolonged ranges beyond that of the biological belina hariaensis at the base and the LA of Gans- occurrence by bioturbation and reworking. serina gansseri at the top. At the Elles section, the A total of 75 planktonic foraminiferal species FA of P. hariaensis almost coincides with the LA were identi¢ed in upper Maastrichtian sediments of G. gansseri (20 cm below). The absence of the at Elles. About 25 of these species have relatively index species may be interpreted as either repre- continuous occurrences throughout the upper senting a hiatus, or ecological exclusion due to Maastrichtian (Fig. 3). They are generally abun- adverse environmental conditions. Since there dant, small and simple morphotypes characteristic are no structural or sedimentological changes of ecological generalists that tolerate a wide range (fault, lithological changes, hardground, and bio- of environments across latitudes (e.g. hedbergell- turbation), or abrupt faunal changes, indicative of ids, heterohelicids, guembelitrids). But two-thirds a hiatus, we prefer the latter explanation, though of the species are only sporadically present and the possibility of a hiatus cannot be excluded. generally few in numbers. These species have gen- This interpretation is supported by the rarity of erally more complex morphologies (e.g. trochospi- P. hariaensis in the part of its stratigraphic range ral, multiserial), narrow tolerance limits, and gen- below Zone CF2, which was also observed at El erally restricted to lower latitude environments. Kef (Nederbragt, 1991; Li and Keller, 1998b). Similar species range patterns were observed at For this reason we have not subdivided Zones El Kef (Li and Keller, 1998b), and the Negev CF3 and CF4 based on the FA of P. hariaensis. (Abramovich et al., 1998). Of particular interest in the species census data 3.4. Racemiguembelina fructicosa zone at Elles is the progressive disappearance of 17 (CF4, 66.8^68.3 Ma) species (mostly globotruncanids) in the upper Maastrichtian preceding the K^T boundary ex- This zone is de¢ned by the FA of Racemiguem- tinction event (e.g. Abathomphalus mayaroensis, belina fructicosa at the base and the FA of Pseu- Archaeoglobigerina blowi, Gansserina gansseri, doguembelina hariaensis at the top. Only the upper Gansserina wiedenmayeri, Globotruncanita pettersi, part of the Zone is included in this study. The Globotruncanita angulata, Globotruncana duepe- lower part of the zone was studied by Li and blei, Globotruncana falsostuarti, Globotruncana in- Keller (1998b). signis, Globotruncana ventricosa, Globigerinelloides multispina, Globotruncanella citae, Planoglobulina multicamerata, Pseudoguembelina excolata, Rosita 4. Species ranges contusa, Rosita plicata, and Rosita wal¢shensis). A similar late Maastrichtian decrease in species rich- The temporal range of a species is de¢ned from ness (presumably due to extinction or high stress its evolutionary ¢rst appearance to its last appear- environment) was observed at El Kef (Li and Kel- ance or extinction. During its range, a species may ler, 1998b) and the Negev (Abramovich et al., only be intermittently or sporadically present, or 1998). However, at Elles 12 of these species tem- have a truncated range as compared with other porarily reappear in the uppermost meter of the localities due to ecological, preservational or sed- Maastrichtian (upper Zone CF1, see Fig. 3), after

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 151

Fig. 4. Planktonic foraminiferal species richness patterns of the late Maastrichtian at Elles, in terms of cumulative and non-cumulative species richness (black line marks ¢ve points running average). Non-cumulative species richness is further divided based on three depth ranked foraminiferal groups: surface, intermediate, and deep dwellers. Cumulative species richness shows a gradual decline with accelerated species extinctions during the last meter below the K^T boundary, and re£ects the evolutionary response to global environmental stress. Non-cumulative species richness shows a decreasing trend with strong sinusoidal variations mainly within intermediate dwellers that re£ect the local response to changes in climate and watermass strati¢cation.

prolonged absence. There are two likely explana- Though it is often di⁄cult to separate reworked tions for this phenomenon. (1) The reappearance species, especially if reworking of extant species is of these species represents their true evolutionary involved, there are some clues, including poor ranges, whereas their observed early extinction is preservation as compared with in situ specimens, an artifact of rare species (e.g. Signor^Lipps ef- occurrences restricted to speci¢c intervals marked fect, rare and sporadic occurrences that could be by bioturbation, hardground, and lithological missed in the sample analyzed). This is a real pos- changes. In addition, reworked sediments gener- sibility for any environmentally sensitive species ally preserve large, robust and thick-shelled speci- in stressed environments. (2) Sediment reworking mens as a result of transport and winnowing. In by currents and redeposition into younger sedi- contrast, isolated occurrences due to the Signor^ ments is a very common way to produce the ob- Lipps e¡ect show no preservation preference ^ served pattern of isolated species occurrences. large and small, fragile and robust specimens

PALAEO 2753 1-5-02 152 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

Table 1 Relative percent abundances of planktonic foraminifera in the 63^150-Wm size fraction at Elles, Tunisia Depth (m) below the KTB 4.20 4.00 3.80 3.60 3.40 3.20 3.00 2.80 2.60 2.40 2.20 2.00 1.80 1.60 1.40 1.20 Biozones CF1 A. blowi A. cretacea Guembelitria cretacea 1662357344112511 G. trifolia 134x644164 21311 Globigerinelloides aspera 12xx21131x121111 G. prehillensis x1 xxx G. volutus x1 1 G. subcarinatus 123x32132xx1 x11 G. rosebudensis xx 21 11 11 xx1 G. multispina G. yaucoensis 4224321222222211 Abathomphalus mayaroensis Globotruncana aegyptica x1x G. orientalis x1x1 G. arca x1x G. esnehensis xx G. falsostuarti G. mariei x1x1 x G. insignis x G. ventricosa G. rosetta xx Globotruncanita conica G. angulata G. pettersi G. stuarti G. stuartiformis x G. petaloidea xx x x G. citae G. havanensis x121xx1x11x Gublerina cuvilieri x G. acuta Gansserina gansseri G. wiedenmayeri Hedbergella 49885697510458654 L. glabrans 221312443322x223 H. globulosa 28 20 19 19 12 19 18 22 22 18 18 16 13 16 24 29 H. carinata 6244141453433445 H. labellosa xxx1xx1 H. navarroensis 10 19 17 16 18 15 23 9 17 14 11 8 21 15 7 10 H. dentata 15 15 15 11 19 11 15 15 14 14 21 25 30 23 28 13 H. planata 11122 x 11 H. punctulata Plummerita hantkeninoides 1xxx1x P. reicheli Pseudoguembelina costulata 16 13 14 20 17 16 13 18 15 19 28 19 16 16 16 22 P. hariaensis x 1xx x1 P. kempensis 2111211212112111 P. palpebra xx x x x Pseudotextularia deformis 1xxx P. intermedia P. elegans x 1112x1 x xx111 Rugoglobigerina hexacamerata 1x 11111xx12 x1

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 153

Table 1 (continued) Depth (m) below the KTB 4.20 4.00 3.80 3.60 3.40 3.20 3.00 2.80 2.60 2.40 2.20 2.00 1.80 1.60 1.40 1.20 Biozones CF1 R. macrocephala xx R. rotundata R. pennyi R. rugosa 221113121124x333 R. scotti xxx11x1 x1 1 1 R. cf scotti Total number counted 412 264 331 339 325 241 270 302 331 229 323 299 270 554 335 336 x= 6 1%, xx = very rare. should be represented. Most of the suspect spe- ord due to adverse environmental £uctuations. cies, which are reappearing in the uppermost Cu mulative species richness thus expresses the evo- meter at Elles well after their last continuous lutionary response of species to long-term changes occurrences, show characteristics of reworked in the environment over time. specimens. In addition, ¢eld and laboratory sedimentological and mineralogical observations in- 5.1. Cumulative species richness dicate various features characteristic of reworking and current transport, including cross-bedding, The cumulative species richness pattern at Elles size gradation, and the presence of a foraminiferal shows a gradual decline through the upper Maas- packstone (see Adatte et al., 2002). These features trichtian followed by accelerated species extinc- strongly indicate that the reappearance of the sus- tions during the top 1 m below the K^T boundary pect species (stippled interval, Fig. 3) is an artifact (Fig. 4). The lower part of the section (lower of sediment reworking. CF4^3 interval, part I) shows a diverse planktonic foraminiferal community with a maximum of 63 species. This species community gradually de- 5. Species richness creased to 53 species by the upper Zone CF2 (part II). Species richness remained low during Species richness, or the number of species the transition between CF2 and CF1 (part III) present within a sample, is the simplest measure and decreased by three species in the middle to of diversity and an important tool for understand- upper Zone CF1 (part IV). A rapid decrease in ing £uctuations in the structure of an ecosystem. species richness by 18 species occurred in the There are two measures of species richness: simple uppermost Maastrichtian, last meter below the species richness and cumulative species richness. K^T boundary (part V), followed by the extinc- Simple species richness is the number of species tion of all remaining tropical and subtropical spe- physically present in any given sample. It measures cies at or near the K^T boundary (Fig. 4, Tables 1 the species £ux in and out of a given ecosystem. As and 2 Background Data Set1). This cumulative such it is a measure of ecological variability, of species richness pattern re£ects the decreased evo- climatic changes, of short-term availability of nu- lutionary diversity during the late Maastrichtian trients, of oxygen and salinity £uctuations. In and the accelerated rate of extinction approaching short, it is an index of climatic, ecological, geo- the K^T boundary event. Similar patterns were graphic and overall environmental variability. In observed from sections at El Kef and the Negev, contrast, cumulative species richness is the number Israel (Li and Keller, l998b; Abramovich et al., of species theoretically present in any given sample. l998). High biotic stress induced by rapidly chang- It assumes that a species is present from its ¢rst ing climatic conditions near the end of the Maas- evolutionary appearance to its extinction and tricthian are the likely causal factors for this spe- ignores any temporary absence in the sediment rec- cies richness decrease (Li and Keller, l998c).

PALAEO 2753 1-5-02 154 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

Table 2 Relative percent abundances of planktonic foraminifera in the size fraction s 150 Wm at Elles, Tunisia Depth (m) below the KTB 0.18 0.13 0.08 0.04 0.03 0.02 0.01 0.00 Biozones CF1 Archaeoglobigerina blowi xxxxx A. cretacea Globigerinelloides aspera 1111x1 G. prehillensis x4xx G. rosebudensis x G. multispina G. subcarinatus xx1 x G. volutus G. yaucoensis 2x x Abathomphalus mayaroensis Globotruncana aegyptica 1111x12 G. arca 22323557 G. orientalis 31112322 G. dupeublei xxxxxx G. esnehensis x1111211 G. ventricosa x11xx1 G. angulata xx1 G. falsostuarti xxxx1 G. mariei 43112332 G. insignis xxxx G. rosetta 111 Globotruncanita conica 1x111 G. pettersi xxx G. stuarti x11 xx11 G. stuartiformis 2x21x4x2 G. petaloidea 1111xxxx G. citae G. havanensis 3123233 Gublerina cuvilieri 1x Gublerina acuta 1 Gansserina gansseri G. wiedenmayeri xxxx Heterohelix globulosa 34 38 37 46 37 28 35 30 H. punctulata x1xx H. labellosa 43224468 L. glabrans 111 1xx Planoglobulina brazoensis x11 P. carseyae x1 x 1 P. acervulinoides xxx P. multicamerata xx xx Plummerita hantkeninoides x1x P. reicheli x Pseudoguembelina costulata 58764234 P. excolata P. hariaensis 33245432 P. kempensis 1xx11x P. palpebra 31122321 Pseudotextularia deformis 31336777 P. intermedia x1x1x P. elegans 871288644 Racemiguembelina fructicosa 1x x 1 R. powelli 11 x1211

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 155

Table 2 (continued) Depth (m) below the KTB 0.18 0.13 0.08 0.04 0.03 0.02 0.01 0.00 Biozones CF1 Rosita contusa x R. patelliformis R. plicata R. wal¢schensis x Rugoglobigerina hexacamerata 23x11142 R. miliamerensis x1 R. macrocephala 31212132 R. rotundata 11111 R. pennyi xx 1 R. rugosa 8 131210111065 R. scotti 53243245 R. cf scotti 1 Total number counted 314 351 330 312 309 269 272 293 x= 6 1%

5.2. Non-cumulative species richness by evaluating the non-cumulative species richness (total number) patterns based on groups living at speci¢c depth habitats (e.g. surface mixed layer, intermediate or The non-cumulative species richness (total thermocline depth, and deep dwellers below ther- number) shows a very di¡erent pattern from mocline, Fig. 4). Depth habitats of planktonic that of the cumulative species richness (Fig. 4). foraminifera, inferred from stable isotopic data Where the latter shows a linear acceleration of of individual species, provide a solid basis for the evolutionary decrease, the former shows a de- reconstructing species life habitats and strati¢ca- creasing trend with strong sinusoidal variations. tion of the water column (D’Hondt and Arthur, The maximum species diversity (V50^54 species) l995; Li and Keller, 1998b). Species richness is found in the lowermost part of the CF4^3 in- depth habitat patterns show that surface and terval, but decreased rapidly to 37^40 species in deeper dwellers remained relatively stable the upper CF4^3. In the lower part of Zone CF2, throughout the upper Maastrichtian. Though at species richness values at ¢rst increased by seven the end of the Maastrichtian (uppermost meter), species, then decreased to 32^35 species in the the species richness of surface dwellers decreased. upper CF2 to lower CF1 Zones. In the lower- Thus, the non-cumulative diversity £uctuations middle CF1, species richness temporarily noted above primarily occurred within the inter- increased by ¢ve species, and then decreased rap- mediate dwellers which lived within the thermo- idly to 30^35 species towards the K^T boundary cline layer. In today’s oceans a well-developed prior to the extinction of all tropical and subtrop- thermocline layer characterized by high produc- ical forms (for changes across the K^T transition tivity supports the highest diversity of planktonic see Keller et al., 2002). These broad-based sinus- foraminifera. This is largely due to the high nu- oidal variations likely re£ect changes in climate trient content recycled from deeper waters, and and watermass strati¢cation, though no climate hence an abundant food supply (e.g. phytoplank- data based on stable isotopes are available at ton and other zooplankton, Hemleben et al., this time. 1989). The strength of the thermocline layer primarily depends on the temperature gradient and 5.3. Non-cumulative species richness therefore re£ects mostly climate and oceanic cir- (depth ranked) culation. The depth ranked species richness curves thus also re£ect signi¢cantly accelerating biotic More environmental information can be gained stress induced by climate changes.

PALAEO 2753 1-5-02 156 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 157

6. Relative abundance changes 6.2. Guembelitrids

6.1. Heterohelicids Two separate Guembelitria blooms (G. cretacea and G. trifolia) occurred in the upper Maas- The planktonic foraminiferal populations of the trichtian as observed in the s 63-Wm size fraction upper Maastrichtian at Elles are typical of the (Fig. 5). The ¢rst Guembelitria bloom occurred in Tethyan pelagic upper slope and outer shelf envi- the interval between the upper part of Zone CF4^ ronments (Figs. 5 and 6). Long-ranging biserial 3 and the lower part of Zone CF2, with the com- heterohelicid species dominate (generally s 70%) bined relative abundance of the two species reach- in the two size fractions analyzed ( s 63 Wm and ing a maximum of 8.5% (see Fig. 7). The second s 150 Wm). Heterohelicid populations in the Guembelitria bloom occurred in Zone CF1 where smaller ( s 63 Wm) size fraction comprise mainly it reached a maximum level of 13.6%. These four species (Heterohelix globulosa, Heterohelix blooms are not observed at the nearby El Kef dentata, Heterohelix navarroensis, and Pseudo- section (Li and Keller, 1998b), probably due to guembelina costulata), whereas in the larger the lower sample resolution. ( s 150 Wm) size fraction three species dominate In previous studies of the Maastrichtian, Guem- (H. globulosa, Pseudotextularia elegans, and Pseu- belitria has been reported in low abundances in dotextularia deformis). Hence, with the exception outer shelf and deep sea environments (D’Hondt of H. globulosa, heterohelicid species are speci¢c and Keller, 1991; Keller, 1996), and relatively to the smaller size fraction. All other Maastrich- high abundances in shallow shelf regions (e.g. tian taxa are much less abundant, especially in the Brazos River (Keller, 1989a,b), Stevns Klint smaller ( s 63 Wm) size fraction (Fig. 5). (Schmitz et al., 1992), Seldja, Tunisia (Keller et Within the heterohelicids of the s 63 Wm size al., 1998). High Guembelitria abundances in the fraction, Heterohelix dentata gradually decreased pelagic marine environment have been described in the lower part of Zone CF2 (from V30% to so far only from the earliest Danian (Smit, 1982, V10%), and increased again towards the upper 1990; Keller and Benjamini, 1991; Schmitz et al., part of CF1 (V30%). Pseudoguembelina costulata 1992; Keller et al., 1993, 1995; Pardo et al., decreased in the upper part of CF4^3 (V30 to 1996), and in the upper Maastrichtian of the Ne- V10%, Fig. 5), and increased again (V30%) in gev (Abramovich et al., 1998). the lower part of CF2. Heterohelix navarroensis Guembelitria are ecological opportunists as in- signi¢cantly increased from the lower to the upper dicated by the very low N13C values associated part of CF4^3 (from V5% to V20%, Fig. 5). The with the early Danian blooms that indicate a relative abundance of Heterohelix globulosa £uc- drastic reduction in primary productivity (Keller, tuated in Zone CF4^3, and stabilized in CF2 with 1996; Keller et al., 2002). The presence of two high average values of V40%. The terminal de- Guembelitria blooms in the upper Maastrichtian cline of H. globulosa began in the upper part of at Elles re£ects the instability of the planktonic CF1 (V20%). In the larger ( s 150 Wm) size frac- population structure at this time. The relative tion the most signi¢cant biotic change within the strength of environmental stress is indicated by heterohelicids is the sharp abundance decline of the magnitude of the Guembelitria blooms. By Pseudotextularia deformis in the upper part of this measure, the environmental stress at Elles is CF2 which coincided with the sharp increase in lower than in the Negev where coeval Guembeli- Pseudotextularia elegans (Fig. 6). tria blooms reached a maximum of V80% as

Fig. 5. Relative abundances of planktonic foraminifera in the smaller ( s 63 Wm) size fraction at Elles. Note that only about ¢ve species dominate the assemblages and four of these species are biserial morphotypes and the ¢fth is a hedbergellid. There are short-term (Milankovitch) £uctuations and long-term trends. Intervals of environmental crises during the late Maastrichtian can be identi¢ed based on two separate blooms of the opportunistic Guembelitria species (G. cretacea, and G. trifolia) in zones CF1 and CF3^4.

PALAEO 2753 1-5-02 158 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 159 compared with 13.6% at Elles (Abramovich et al., induced by instability in the Tethyan tropic belt 1998). Thus the biotic response to the environ- environment preceding the K^T boundary (Fig. 4). mental stress induced by climatic perturbation in di¡erent parts of the Tethys may be of the same nature, but of di¡erent magnitude. 7. Late Maastrichtian paleoenvironment in Tunisia

6.3. Globotruncanids The biotic response to climatic perturbations is re£ected by relative abundance changes in species In the larger ( s 150 Wm) size fraction, major populations, species census data and species rich- biotic changes took place between the upper ness patterns (both cumulative and non-cumula- CF4^3 and the upper CF2 Zones when many glo- tive). At Elles all of these parameters show signi¢- botruncanids decreased in relative abundance cant variations during the late Maastrichtian and (Gansserina gansseri, G. wiedenmayeri, G. arca, a progressively accelerating decrease in species G. orientalis, G. mariei, and G. havanensis, richness during the last 300 kyr of the Maastrich- Fig. 6). This decrease demonstrates that the upper tian. In addition, biotic stress is indicated by the Maastrichtian decline of globotruncanids was ex- decline in relative abundances of dominant spe- pressed not only by decreased species richness, cies, sporadic blooms of opportunistic species, but also by decreased population abundance. A and abundance £uctuations in populations of similar decrease in globorotaliids was observed long-ranging species. Interpreting these records in the Negev sections, Israel (Abramovich et al., in terms of speci¢c climate changes, such as 1998). However, because globorotaliid are large warm and cool events, variations in temperature species and generally few to rare, this decrease is gradients and watermass strati¢cation, remains a not evident when counting only the smaller ( s 63 challenge. In Tunisian sections, this task is further Wm) size fraction (Fig. 5), and hence Li and Keller complicated by oxygen isotope records of fora- (1998b) who analyzed only the smaller size frac- minifera that are often compromised by diage- tion, failed to note it. netic alteration and therefore allow no tempera- Globotruncanids are usually regarded as inter- ture interpretations (e.g. El Kef, Keller and mediate to deep dwellers and geographically lim- Lindinger, l989). ited to the Tethyan tropical^subtropical belt dur- To date, the most detailed and least diageneti- ing the Cretaceous. All globotruncanid species cally compromised stable isotope paleotempera- were extinct before or at the K^T boundary. ture record for the late Maastrichtian is from mid- The limited distribution of globotruncanids and dle latitude DSDP Site 525A. This record their characteristic complex morphologies illus- indicates that relatively cool temperatures pre- trate both their speci¢c requirements for stable vailed during the ¢rst 3 Myr of the late Maas- undisturbed environmental conditions (mainly trichtian (CF4^3 interval) with intermediate water well strati¢ed watermass with constant food sup- temperatures about 6^7‡C cooler than in the early ply, and stable temperatures) and their specialized Maastrichtian and surface temperatures cool but life strategy as many modern analogous forms of variable (Li and Keller, l998a,c). This long-term today’s tropics indicate (Caron and Homewood, cooling trend continued throughout the CF4^3 1983; Leckie, 1989; Keller et al., 1995; Keller, interval and reached minimum temperatures in 1996). Therefore, their decline during the late the upper part of CF3 coincident with a hiatus Maastrichtian at Elles demonstrates biotic stress at El Kef at 65.5 Ma (Li and Keller, 1998c). At

Fig. 6. Relative abundances of planktonic foraminifera in the larger ( s 150 Wm) size fraction at Elles. Note that biserial taxa are still dominant in this size fraction, but rugoglobigerinids are also common. Globotruncanids (mainly tropical complex morphologies) are signi¢cantly more common than in the smaller size fraction, but decrease signi¢cantly near the top of CF3^4 and CF2.

PALAEO 2753 1-5-02 160 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 161

Elles, the environmental parameters of planktonic the last 100 kyr of the Maastrichtian temperatures foraminifera for the equivalent late Maastrichtian dropped rapidly, and returned to the previous low interval (CF4^3) indicate a well-strati¢ed water level. The biotic e¡ect of this warm event on the column characteristic of a cooler climate in which Tethyan ecosystem is well demonstrated at Elles deeper water habitats are well populated, partic- (Fig. 7). Globotruncanids, which began to decline ularly in the lower part of the CF4^3 interval. in the upper part of the CF4^3 interval, experi- These conditions are mainly demonstrated by enced another decline during the maximum warm- the high species richness and the relative abun- ing (lower to middle part of Zone CF1) when dance of species with complex trochospiral shells their combined relative abundance decreased by (globotruncanids), and biserial species that inhab- another 10%, and the species richness decreased it the intermediate to deeper waters (e.g. at or by ca. ¢ve species. The same response was also below the thermocline, Fig. 7). However, towards observed in the Negev sections (Abramovich et the upper part of the Zone CF4^3 interval globo- al., 1998). Heterohelix dentata (a biserial inter- truncanids began to decline (from V30% to mediate dweller) was also a¡ected by this event. V20%, see Fig. 7) and species richness decreased During the warm interval the relative abundance by 10 species, resulting in depleted intermediate of this species decreased by V30%, and with the water communities (Fig. 4). This decline may re- return to cooler temperatures the relative abun- £ect a decrease in water temperatures below the dance increased again. The two species of the ge- optimal tolerance range for globotruncanid spe- nus Pseudotextularia (P. elegans, and P. deformis) cies. In general, the geographic restriction of also seem to respond to this climatic change. most globotruncanid species to low latitudes, their When the relative abundance of the surface dwell- relatively low abundance at middle latitudes, and er, P. deformis, decreased near the base of Zone near absence in higher latitudes re£ects a narrow CF1, the relative abundance of the intermediate temperature tolerance range. Alternatively, it is dweller P. elegans increased (Fig. 7). The £uctua- possible that the onset of the subsequent warming tion in the relative abundances of other major in CF2^1 (Li and Keller, l998a,c) began earlier in species, such as Pseudoguembelina costulata, Ru- the Tethys resulting in decreased thermal gra- goglobigerina spp., Heterohelix navarroensis, and dients and hence habitats for subsurface popula- Hedbergella spp., does not seem to correspond to tions. Stable isotope data of the eastern Tethys major temperature changes. are still needed to evaluate which scenario is It is interesting to note that some species, which most likely. respond in the same way to climatic changes, do In middle to high southern latitudes, the long- not share the same habitat. Also, some species term late Maastrichtian cooling was interrupted that share the same habitat show opposite re- by a short-term warming of 3^4‡C in both surface sponses to climatic changes. For example, both and intermediate waters (e.g. Sites 690 and 525A, the intermediate dwelling globotruncanids and Barrera, 1994; Li and Keller, 1998a,c). This the surface dweller Pseudotextularia deformis de- warming began 450 kyr before the K^T boundary clined during the maximum warming. But the in- (base of Zone CF2), reached a maximum at 65.3 termediate dweller Heterohelix globulosa increased Ma (base of Zone CF1) and ended 100 kyr before during the warming, whereas the intermediate the K^T boundary (Li and Keller, 1998c). During dweller Heterohelix dentata decreased. This dem-

Fig. 7. Summary of relative abundances of dominant planktonic foraminiferal species and genera from the two size fractions ( s 150 Wm and s 63 Wm) at Elles. Abundance changes re£ect planktonic foraminiferal response to climatic instability. Shaded interval marks the late Maastrichtian warm event (age estimate based on biostratigraphic correlation with the warm event at Site 525A, Li and Keller, 1998c). Note the species turnover within the biserial population in response to the warm event; the biserial species Heterohelix dentata, and Pseudotextularia deformis decreased, whereas the biserial species Heterohelix globulosa, Pseudotex- tularia elegans and Pseudotextularia costulata increased. Globotruncanids also decreased preceding and during the warm event.

PALAEO 2753 1-5-02 162 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 onstrates that a given change in the environment (2) Non-cumulative species richness data (ac- can have di¡erent e¡ects on species that share the tual number of species present), which re£ect en- same habitat depth zone. Consequently, species vironmental changes, show major broad-based that inhabit the same depth habitats do not nec- £uctuations that mirror climate changes and asso- essarily have the same requirements for resources, ciated variations in ocean circulation, thermal or the same tolerance for environmental changes. gradients and watermass strati¢cation. The Elles For example, the intermediate dwelling heterohe- records indicate that environmental changes were licids are considered to have a high tolerance for particularly strong during the last 500 kyr of the low oxygen conditions (Premoli Silva and Boers- Maastrichtian. ma, l989; Keller, 1993; Barrera and Keller, 1994), (3) Non-cumulative depth ranked (surface, in- but intermediate dwelling globotruncanids clearly termediate, deep dwellers) species richness data do not, they decrease or disappear during times of reveal the particular habitat most severely a¡ected expanded low oxygen conditions (e.g. Almogi-La- by ongoing climate and environmental changes. bin et al., 1993). The small triserial Guembelitria, At Elles, the intermediate dwellers living within best known as disaster species that thrived after thermocline depths drove both the long-term evo- the mass extinction of tropical and subtropical lutionary trends and the response to short-term species at the K^T boundary, generally thrived environmental changes. Surface and deeper dwell- at times of major biotic stress associated with er groups remained relatively stable. major climate transitions. In addition to the K^ (4) Species census data (individual species T boundary, two blooms of these disaster species ranges) also reveal environmental £ux. At Elles, were observed during the upper Maastrichtian at 17 species (intermediate dwellers) gradually disap- Elles and in the eastern Tethys (Abramovich et peared during the upper Maastrichtian and re£ect al., l998). the increasing environmental stress during the last 2 Myr of the Cretaceous. (5) Globotruncanids are the most sensitive in- 8. Conclusions dicators of Maastrichtian climate changes. This group of thermocline dwellers (intermediate) Upper Maastrichtian planktonic foraminiferal thrived during relatively cool climates and a assemblages at Elles reveal major faunal turnovers well-strati¢ed water column. But their tempera- that re£ect the global long-term climate cooling ture tolerance limit is relatively low and they are followed by rapid and extreme warming between prone to extinctions when temperatures are either 200 and 400 kyr before the K^T boundary, and a too cool or too warm. return to a cooler climate during the last 100 kyr (6) Guembelitrids are useful indicators for en- of the Maastrichtian. Although there are still few vironmental extremes and high biotic stress, par- studies that evaluate the association of speci¢c ticularly at transitions from warm to cold climates Cretaceous species with particular environmental and vice versa. During these times, Guembelitria changes, preliminary results are encouraging and species responded with opportunistic blooms that promise that this approach will reveal much about re£ect the high stress and unstable planktonic the environmental changes and biotic stresses that population structure. accompany major climate transitions. From the Elles study a number of preliminary conclusions can be reached. Acknowledgements (1) Cumulative species richness data (total number of species living), which re£ect evolution- We gratefully acknowledge Dr. Ben Haj Ali, ary trends, show a gradual decrease in diversity Director of the Geological Survey of Tunisia, and through the last 3 Myr of the Maastrichtian with Dr. Ben Salem for their tremendous logistical and a rapid decrease during the last 100 kyr of the transportation support for the field excursion and Maastrichtian. Workshop of May l998. Fieldwork was done

PALAEO 2753 1-5-02 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164 163 jointly with Thierry Adatte and Wolfgang Stin- Direction du Service Ge¤ologique, Tunis, l998, Abstracts, pp. nesbeck and we gratefully acknowledge their 11^13. Burollet, P.F., 1967. General geology in Tunisia. In: Martin, collaboration and discussions. We thank the L. (Ed.), Guidebook to the geology history of Tunisia, Pe- reviewers Francis Robaszynski and one anon- troleum Exploration Society of Libya, 9th Annual Field ymous reviewer for their helpful comments, and Conference, 67 pp. Chaim Benjamini for discussions. This study was Canudo, J.I., Keller, G., Molina, E., 1991. Cretaceous/Teriary supported by NSF-INT 95-04309. boundary extinction pattern and faunal turnover at Agost and Caravaca, SE Spain. Mar. Micropaleontol. 17, 319^341. Caron, M., 1985. Cretaceous planktic foraminifera. In: Bolli, H.M., Saunders, J.B., Perch-Nielsen, K. (Eds.), Plankton References Stratigraphy. Cambridge University Press, Cambridge, pp. 17^86. Abramovich, S., Almogi-Labin, A., Benjamini, C., 1998. De- Caron, M., Homewood, P., 1983. Evolution of early planktic cline of the Maastrichtian pelagic ecosystem based on foraminifers. Mar. Micropaleontol. 7, 453^462. planktic foraminiferal assemblage changes: Implications D’Hondt, S., Keller, G., 1991. Some patterns of planktic fo- for the terminal Cretaceous faunal crisis. Geology 26, 63^66. raminiferal assemblage turnover at the Cretaceous^Tertiary Adatte, T., Keller, G., Li, L., Stinnesbeck, W., Zaghbib-Turki, boundary. Mar. Micropaleontol. 17, 77^118. D., 1998. Climate and sea level £uctuations across the K^T D’Hondt, S., Arthur, M.A., 1995. Interspecies variation in boundary in Tunisia: Warm and humid conditions linked to stable isotopic signals of Maastrichtian planktonic forami- the Deccan volcanism? In: International Workshop on Cre- nifera. Paleoceanography 10, 123^135. taceous^Tertiary Transition, O⁄ce National des Mines, Di- Hemleben, C., Spindler, M., Anderson, O.R., 1989. Modern rection du Service Ge¤ologique, Tunis, l998, Abstracts, pp. 7^ Planktonic Foraminifera. Springer Verlag, New York. 10. Huber, B.T., 1992. Paleobiogeography of Campanian^Maas- Adatte, T., Keller, G., Stinnesbeck, W., 2002. Late Cretaceous trichtian foraminifera in the southern high latitudes. Pa- to Early Paleocene climate and sea-level £uctuations. Pa- laeogeogr. Palaeoclimatol. Palaeoecol. 92, 325^360. laeogeogr., Palaeoclimatol., Palaeoecol. (this volume). Keller, G., 1988. Extinction, survivorship and evolution of Almogi-Labin, A., Bein, A., Sass, E., 1993. Late Cretaceous plalnktic foraminifera across the Cretaceous/Teriary bound- upwelling system along the southern Tethys margin (Israel): ary at El Kef Tunisia. Mar. Micropaleontol. 13, 239^263. Interrelationship between productivity, bottom water envi- Keller, G., 1989a. Extended period of extinctions across the ronments, and organic matter preservation. Paleoceanogra- Cretaceous/Teriary boundary in planktonic foraminifera of phy 8, 671^690. continental shelf section: Implications for a impact and vol- Appellaniz, E., Baceta, J.I., Benrnaola-Bilbao, G., Nunez-Be- canism theories. Geol. Soc. Am. Bull. 101, 1408^1419. lelu, K., Orue-Etxebarria, X., Payros, A., Pujalte, V., Robin, Keller, G., 1989b. Extended Cretaceous/Teriary boundary ex- E., Rocchia, R., 1997. Analysis of uppermost Cretaceous- tinctions and delayed population change in planktonic fora- lowermost Tertiary hemipelagic successions in the Basque minifera from Brazos River, Texas. Paleoceanography 4, Country (western Pyrenees): evidence for a sudden extinc- 287^332. tion of more than half planktic foraminiferal species at the Keller, G., Lindinger, M., 1989. Stable isotopes, TOC and K/T boundary. Bull. Soc. Geol. France 168 (6), 783^793. CaCO3 records across the Cretaceous-Tertiary boundary Barrera, E., 1994. Global environmental changes preceding the at El Kef, Tunisia. Palaeogeogr. Palaeoclimatol. Palaeoecol. Cretaceous^Tertiary boundary: Early^upper Maastrichtian 73, 243^265. transition. Geology 22, 877^880. Keller, G., Benjamini, C., 1991. Paleoenvironment of the east- Barrera, B., Keller, G., 1994. Productivity across the Creta- ern Tethys in the Early Paleocene. Palaios 6, 439^464. ceous/Teriary boundary in high latitudes. Geol. Soc. Am. Keller, G., 1993. The Cretaceous/Teriary boundary transition Bull. 106, 1254^1266. in the Antarctic Ocean and its glocal implications. Mar. Barrera, E., Savin, S.M., Thomas, E., Jones, C.E., 1997. Evi- Micropaleontol. 21, 1^45. dence for thermohaline-circulation reversals controlled by Keller, G., Barrera, E., Schmitz, B., Matsson, E., 1993. Grad- sea-level change in the latest Cretaceous. Geology 25, 715^ ual mass extinction, species survivorship, and long term en- 718. vironmental changes across the Cretaceous-Teriary boun- Barrera, E., Savin, S.M., 1999. Evolution of late Campanian^ dariy in high latitudes. Geol. Soc. Am. Bull. 105, 979^997. Maastrichtian marine climates and oceans. In: Barrera, E., Keller, G., Li, L., MacLeod, N., 1995. The Cretaceous/Terti- Johnson, C.C. (Eds.), Evolution of the Cretaceous Ocean- ary boundary stratotype section at El Kef, Tunisia: How Climate System. Geological Society of America Special Pa- catastrophic was the mass extinction? Palaeogeogr. Pa- per 332, Boulder, CO, pp. 245^282. laeoclimatol. Palaeoecol. 119, 221^254. Bensalem, H., 1998. The El Haria Tunisian K/T transition Keller, G., 1996. The K/T mass extinction in planktic foramin- serie: Bibliographic survey. In: International Workshop on ifera biotic constraints for catastrophe theories. In: Mac- Cretaceous^Tertiary Transition, O⁄ce National des Mines, Leod, N., Keller, G. (Eds.), The Cretaceous^Tertiary Mass

PALAEO 2753 1-5-02 164 S. Abramovich, G. Keller / Palaeogeography, Palaeoclimatology, Palaeoecology 178 (2002) 145^164

Extinction: Biotic and Environmental Changes. W.W. Nor- development of Heterohelicidae (planktic foraminifera). Mi- ton, New York, pp. 63^100. cropaleontology 37, 329^372. Keller, G., Adatte, T., Stinnesbeck, W., Stuben, D., Kramar, Olsson, R.K., Liu, C., 1993. Controversies on the placement of U., Berner, Z., Li, L., vonSalisPerch-Nielsen, K., 1997. The the Cretaceous^Paleogene boundary and the K/T mass ex- Cretaceous^Teriary transition on the shallow Sharan plat- tinction of planktonic foraminifera. Palaios 8, 127^139. form of southern Tunisia. Geobios 30, 951^975. Pardo, A., Ortiz, N., Keller, G., 1996. Latest Maastrichtian Keller, G., Adatte, T., Stinnesbeck, W., Stuben, D., Kramar, foraminiferal turnover and its environmental implications U., Berner, Z., Li, L., von Salis Perch-Nielsen, K., 1998. The at Agost, Spain. In: MacLeod, N., Keller, G. (Eds.), Creta- Cretaceous-Tertiary transition on the shallow Sharan plat- ceous^Tertiary Boundary Mass Extinction: Biotic and En- form of southern Tunisia. Geobios 30 (7), 951^975. vironmental Changes. W.W. Norton, New York, pp. 139^ Keller, G., Adatte, T., Stinnesbeck, W., Luciani, V., Karoui, 172. N., Zaghbib-Turki, D., 2002. Paleoecology of the Creta- Premoli Silva, I., Boersma, A., 1989. Atlantic Paleogene plank- ceous-Tertiary mass extinction in planktic foraminifera. Pa- tonic foraminiferal bioprovincial indices. Mar. Micropaleon- laeogeogr. Palaeoclimatol. Palaeoecol. 178, 257^259. tol. 14, 357^371. Leckie, R.M., 1989. A paleogeographic model for the early Robaszynski, F., Caron, M., Gonzales Donoso, J.M., Won- evolutionary history of planktonic foraminifera. Palaeo- ders, A.A.H., 1984. Atlas of Late Cretaceous Globotrunca- geogr. Palaeoclimatol. Palaeoecol. 73, 107^138. nids. Rev. Micropale¤ontol. 26, 145^305. Li, L., Keller, G., 1998a. Maastrichtian climate productivity Robaszynski, F., Caron, M., 1995. Foraminife'res planctoni- and faunal turnovers in planktic foraminifera in South At- ques du Cre¤tace: Commentaire de la zonation Europe-Me¤d- lantic DSDP Sites 525A and 21. Mar. Micropaleontol. 33, iterrane¤. Bull. Soc. Geol. France 166, 681^692. 55^86. Salaj, J., 1980. Microbiostratigraphie du Cre¤tace¤ et du Pale¤o- Li, L., Keller, G., 1998b. Diversi¢cation and extinction in ge'ne de la Tunisie septentrionale et orientale (Hypostrato- Campanian^Maastrichtian planktic foraminifera of North- types tunisiens). Institut Geologique de Dionyz Stu¤ra, Bra- western Tunisia. Ecl. Helv. 91, 75^102. tislava. Li, L., Keller, G., 1998c. Abrupt deep-sea warming at the end Schmitz, B., Keller, G., Stenvall, O., 1992. Stable isotope and of the Cretaceous. Geology 26, 995^999. foraminiferal changes across the Creatceous-Teriary bound- Li, L., Keller, G., Stinnesbeck, W., 1999. The Late Campanian ary at Stevns Klint, Denmark: Arguments for long term and Maastrichtian in northern Tunisia: palaeoenvironmen- oceanic instability before and after bolide impact event. Pa- tal inferences from lithology, macrofauna and benthic fora- laeogeogr. Palaeoclimatol. Palaeoecol. 96, 233^260. minifera. Cretac. Res. 20, 231^252. Signor, P., Lipps, J., 1982. Sampling bias, gradual extinction Li, L., Keller, G., Adatte, T., Stinnesbeck, W., 2000. Late patterns and catastrophes in the fossil record. In: Sliver, Cretaceous sea level changes in Tunisia: A multi-disciplinary L.T., Schultz, P.H. (Eds.), Geological Implications of Im- approach. J. Geol. Soc. London 2, 447^458. pacts of Large Asteroids and Comets on the Earth. Geo- Luciani, V., 1997. Planktonic foraminiferal turnover across the logical Society of America, Special Paper 190, Boulder, CO, Cretaceous^Teriary boundary in the Vajont valley (Southern pp. 291^296. Alps, northern Italy). Cretac. Res. 18, 799^821. Smit, J., 1982. Extinction and evolution of planktic forami- Masters, B.A., 1984. Comparision of planktonic foraminifersat nifera after a major impact at the Cretaceous/Tertiary the Cretaceous-Teriary boundary from the El Haria shale boundary. Earth Planet. Sci. Lett. 74, 155^170. (Tunisia) and the Esna shale (Egypt). Proceedings of the Smit, J., 1990. Meteorite impact, extinctions and the Creta- 7th Exploration Seminar, March, 1984, Cairo, Egypt. Egyp- ceous/Tertiary boundary. Geol. Mijnb. 69, 187^204. tian General Petroleum Corporation, Cairo, pp. 310^324. Smith, C.C., Pessagno, E.A., Jr., 1973. Planktonic foraminifera Molina, E., Arenillas, I., Arz, J.A., 1998. Mass extinction in and stratigraphy of Corsicana Formation (Maastrichtian), planktic foraminifera at the Cretaceous^Tertiary boundary north-central Texas. Contribution from the Cushman Foun- in subtropical and temperate latitudes. Bull. Soc. Geol. dation for Foraminiferal Research, Special Publication 12, France 169, 351^372. 5^66. Nederbragt, A., 1991. Late Cretaceous biostratigraphy and

PALAEO 2753 1-5-02