Marine Micropaleontology, 13 (1988) 239-263 239 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands
Extinction, Survivorship and Evolution of Planktic Foraminifera across the Cretaceous/Tertiary Boundary at E1 Kef, Tunisia
GERTA KELLER
Princeton University, Department of Geological and Geophysical Sciences, Princeton, NJ 08544 (U.S.A.)
(Revised and accepted November 25, 1987 )
Abstract
Keller, G., 1988. Extinction, survivorship and evolution of planktic foraminifera across the Cretaceous/Tertiary boundary at E1 Kef, Tunisia. Mar-Micropaleontol., 13: 239-263.
An expanded sediment record at E1 Kef shows that the K/T boundary extinctions of planktic foraminifera extend over an interval from 25 cm below the geochemical boundary (Ir anomaly) to 7 cm above. Species extinctions appear sequential with complex, large, ornate forms disappearing first and smaller, less ornate, forms surviving longer. The 14 species extinctions below the boundary appear unrelated to an impact event. Cretaceous species survivorship is greater than previously assumed. About 10 species survive (22%) into Subzone Pla (Globigerina eugubina). All Cretaceous survivors are small primitive forms which are generally smaller than their ancestors in Cretaceous sediments. Species evolution after the K/T event occurs in two pulses. The first new Paleocene species evolve in the basal black clay (Zone PO ) immediately after the major Cretaceous extinctions. Evolving species are small and primitive similar to Cretaceous survivors. The second pulse in species evolution occurs in the lower part of Subzone Plb with the appearance of larger more diverse species. The first major increase in carbonate sedimentation and productivity occurs at this time and signals the recoveyr of the ecosystem nearly 300,000 years after the K/T event. The species extinctions prior to the generally assumed impact event implied by the Ir anomaly, and the long recovery period of the ecosystem thereafter cannot be explained by a single impact, but suggest that multiple causes may be responsible such as climatic changes, volcanism, a sea level drop, production of warm saline bottom water and the chemical consequences associ- ated with increased salinity.
Introduction nannofossil and foraminiferal workers have discovered one additional biostratigraphic zone The E1 Kef section is located about 7 km west at the base of the Paleocene which is not gen- from the town of E1 Kef in northwest Tunisia erally present in deep-sea sections with more (Fig. 1 ). Nannofossil studies by Perch-Nielsen condensed K/T boundary sequences (Perch- (1979, 1981, a,b), Perch-Nielsen et al. (1982) Nielsen, 1979; Smit, 1982). Despite recognition and Romein and Smit (1981) and planktic for- of E1 Kef as perhaps the most complete K/T aminiferal studies by Salaj (1973, 1977) and boundary sequence, no detailed studies on Smit (1982) claim El Kef as the most complete planktic or benthic foraminifera have been K/T boundary section known to date. Both published.
0377-8398/88/$03.50 © 1988 Elsevier Science Publishers B.V. 240
:~, I - (j 2/ fl " Y I
Lower and Middle ,~;<" (Metlaoui Fro.) ~ upper Campanian and -~ lower Maastrichtlan L ~ upper Maastrichtian to "~- Paleocene (El Haria Fm.) Campanian and older "37 ---- ~_~0 zerte
~ : c/r Boundary
: small village nis
Ilnllllllll • studied section
: road
Fig. 1. Location map showing the K/T boundary section near E1 Kef, Tunisia along with regional distribution of sediments of Campanian to Eocene age,
This report and companion studies on benthic boundary ( P lb Subzone, Keller, in prep.; Pey- foraminifera (Keller, 1988a) and stable isotope pouquet et al., 1986). Thus, the K/T boundary stratigraphy (Keller and Lindinger, in prep. ) event profoundly affected faunas living on the attempt to elucidate the biotic turnover and pa- sea floor of the Tunisian continental platform. leoenvironmental conditions on the sea floor Planktic foraminifera were analyzed to ob- and in surface waters across the K/T boundary tain a complete history across the K/T bound- at E1 Kef. Benthic foraminifera show a 50% re- ary in a relatively protected shallow sea. The duction in species diversity at the K/T bound- surface record of faunal changes essentially ary and diversity remained 37% lower during mirrors that on the sea floor, but planktic for- deposition of the first 3 m of the sediment above aminifera suffered greater losses and were sub- the boundary (PO to Pla Zone ). Surviving spe- jected to more stressful conditions than benthic cies are generally tolerant of low 02 conditions foraminifera. Although planktic foraminifera and productivity was low. Increased productiv- show overall similarities with the record known ity and return to higher oxygen levels on the from the deep-sea, fundamental differences are seafloor resumed about 4 m above the K/T apparent; these, however, may be largely due to 241 the expanded section present at E1 Kef. The Biostratigraphy and lithology differences concern primarily the nature of spe- cies extinctions across the K/T boundary, spe- The K/T boundary section of El Kef is con- cies survivorship, and evolution of new species, tained within the E1 Haria Formation (Fig. 1 ). This report primarily deals with these aspects This formation spans upper Cretaceous to Pa- of the K/T boundary event. leocene sediments consisting of marls with spo- radic limestone intercalations (Burollet, 1966). The lithology of the K/T boundary interval is Material and methods characterized by five units which are discussed below in the context of the biostratigraphic zonation of Premoli Silva and Bolli (1973), Samples were obtained by the Cretaceous/ Berggren (1978), Herm (1981), and modified Paleocene Working Group of the International by Salaj (1977), Smit (1982) and Smit and Romein (1985). Further modifications of this Committee on Stratigraphy. Sample numbers zonal system are proposed here and illustrated used in this report represent original numbers in Fig. 2. given by the collecting team. The section was sampled at 20-cm intervals below the K/T UpperMaastrichtian boundary and continuous sampling at 5-cm in- tervals for the first 1.60 m above the boundary, A white-grey clayey marl with about 40% then at 20-cm intervals between 1.60-3.00 m CaCO3 and common burrows about 2-3 cm long and at 50-cm intervals between 3 and 10 m marks the uppermost Maastrichtian. The up- above the boundary. A total of 66 samples were permost 4.5 m of sediment examined of this unit analyzed for this study between 4 m below the do not contain Abathomphalus mayaroensis, the K/T boundary to 10 m above the boundary, index species of the latest Maastrichtian zone Samples were processed by standard micro- with the same name. Absence of this species in paleontological techniques. In general samples the E1 Haria Formation was earlier observed by were not easy to disagregate and repeated soak- Salaj (1980) and was also recently observed by ing in calgon with dilute H202 followed by sev- the present author in the Negev sections of Is- eral washings were necessary. Planktic rael and Brazos River section of Texas, andby foraminiferal preservation is generally good al- Jones et al. (1987) in the Braggs section of Al- though recrystallization of original calcite shells abama. The frequently observed absence of A. is evident in lower Paleocene sediments mayaroensis from uppermost Cretaceous sedi- Population counts were based on represent- ments of continental shells indicates that this ative splits (using an Otto microsplitter) of species disappears prior to the K/T boundary about 300 specimen of the size fractions > 63 at least in shelf regions. For this reason the new /lm and > 150/~m. Two size fractions were ana- Pseudotextularia deformis Zone is proposed for lyzed because of the variable species sizes. The the latest Maastrichtian (Fig. 2). smaller size fraction ( > 63/lm) best shows ev- KIT boundary olutionary trends of small species in late Cre- taceous and earliest Tertiary sediments. The The K/T boundary lies between the top of larger size fraction ( > 150/~m ) was expected to the white-grey clayey marl and the overlying 50 show trends of larger species, but no faunal cm thick black clay unit which has an average trends could be recognized. Faunal counts for of 4% CaCQ. A 2-3 mm thin rust colored fer- > 63 ,urn and > 150 ~m size fractions are listed ruginous layer at the base of the black clay in Tables I and II. Sample depths are listed as marks the boundary event. No burrows are ob- midpoints for continuous sample intervals, served through this layer or in the black clay. 242
TABLE I
Relative percentage of planktic foraminifera in the size fraction > 63/zm at El Kef, Tunisia; x = < 2%
El KEF+ TUNISIA .... , , ., , ...... - .... - .... - ..... ~3 .~m
Archeog~ob~gerina blO~l x x • x x GkJbigenna mon~ut hens,s x x ~ * x x x G~obi~e~ c4 ~nmOUthenSlS x x x x x x x x x • x ~ , G k)bigermellO R hex~amer~la • 2 x x ~ x 2 6 , • x x x ~ × R m~cceph~ta x x ~ ~ ~ , ~ ~ ~ x ~ , , R reich eli x x x x x x ~ , ~ R nog=a x x x x ~ x x x • 2 ~ 2 * • R rob~JSt a sp A ~ ' x 4 ,- 7 4 : e : 5 ~ X Rugot~ana ~bpennyl Loebl~Che!la hessi Vent rilabella eggen GIObOltun~nella havanenSl~ K ~ x x x x x , • ~ x x G petalc4dea , 2 , x x x × , • * , G~obolru~na aegyptla~ x × ~ x x x 2 x x , ~ , x G ar~ G faisostuan~ x x • x * x ~ ~ x x K G gansseq O lrlP;daOens,s ~ ~ ~ G stuarll x ~ ~ x × ~ ~ * x x ~ ] , , G s~uamformm p ~udog~mbelina ~slu¢~ta ! 7 25 15 ~0 14 20 7 1 : z 7 1 S t 4 • 2 12 24 1 : p kempens,s ~ x p pa~pebra ~ ~ 2 3 ? p punct u[at a , ) x x ~ ~ ~ ~ x ~ • Pseudmextuiarla eieOans P deform=s x x x x ~ x x x x ~ K x ~ x H elerOhelix glabrans x 4 2 3 3 x x 4 ~ 3 9 3 3 N g~losa '7 ~ 9 3 2 4 2 2 2 2 H slnata 5 11 9 12 23 7 23 a 12 b 21 16 24 I- ~ , * x , 9 H ame~ana * planog~obuh na blazoensls x x ~ x Racemiguembelma I ructi~sa R inte~med~a x x x x x x , ~ • ~ x , • R powel¢i x ~ , x × × × , , x , ~ ~ S~na ceno~a x Ouembelit rit Ct etacea 2 x x x ~ 2 2 2 x x x 4 18 1:' 9' ~4 95 85 67 B 4 '+ ~ r {0 ,'{~ "5 85 60 /2 5" G irregutaris G Inloha : 4 4 5 G miowaye~s~s Wo~ ringina homersl~nens,s ~ Chiloguemb~ltha danica 2 , × : ~ , 2 6 6 ~, Ch laurlca Globot~u ~anella cat av~aens=s ~ x x 3 x 2 , , 3 , ~ x x GCY~ocon~a oorusa '5 g 9 10 8 22 12 20 G d~bjerga~is Globigenna eugul~ na G. f riP, ga G mosey,n, Subbotina tnloculmo~das EogloO~gen n~ ~ila x E ~bulle ~es E Pemlspnaerlc~ E Slmpbclsslm a E launca E tet~gona GlobOrotaha atch~mpressa ~ 2 GI ~rRptes~ GI. polycamera G~ pr~t=qua O~ ~se~do~de$ ~ ps~buII S~Ua~F~IUS 232 m ~ ~ g m ~ ~ 561 +102cm 562 +107cm 267 ~ ~ ~ ~ ~ x ~ ~ 564 +ll7cm 172 ~ ~ ~ ~ ~ w ~ 565 +122cm !39 ~ ~ ~ x ~ Z 566 +127cm 266 • ~ ~ ~ • ~ m 567 +132cm 164 ~ ~ ~ ~ 569 +142cm 189 ~ ~ ~ " ~ ~ 570 +147cm 117 - m ~ ~ ~ x ~ ~ ~ ~ ~ ~ 571 +152cm 277 ~ x x ~ ~ ~ ~ ~ g ~ ~ ~ ~ ~ 572 +157cm 242 . • ~ ~ ~ ~ ~ m ~ m ~ - 374 +167cm 249 - ~ ~ ~ ~ ~ ~ ~ m 576 +177cm 108 ~ . ~ ~ ~ ~ ~ ~ ~ m 578 +182cm 251 ~ ~ ~ ~ ~ ~ ~ ~ ~ 580 +197cm 262 ~ ~ ~ ~ : ~ ~ ~ ~ ~ 582 +222cm 230 ~ ~ ~ ~ ~ g ~ ~ m ~ 586 +302cm 268 . m ~ ~ ~ ~ ~ ~ ..... m ~ 587 +350cm 5~9 +450¢~ 590 +500cm 591 +550cm 592 +600cm 59~ +650cm 297 .... ~ x x x M ~ 594 +700cm 312 x x ~ ~ ~ ..... 595 +750cm 596 +800em 389 ...... = g g ~ g 597 +850cm ~66 ...... ~ gg ~ g 598 +9oocm X54 ...... g ~ ~ ~ 599 +950cm 600 +lOOOcr t',O ¢.0 TABI,E II gx Relative percentage of planktic Foraminifera in the si~e fraction > 150 lln~ at El Kef, Tunisia; x = < 2~. El KEF, TUNISIA ...... +t ; '4~ + "~ ~ ~ rE~: ? + ~ > 150 /Am x = <2% Archeoglobigerina blowi x x x : : . . G. cf monmouthensis x x 2 2 x 3 x x x 3 x ~ i ~ Globigerinelloidesasia.era x x x 2 x x x x x 2 x : : : : : : : x ' G. multispina x x x x x : ! ~ i i i ! - x G..~,bcafinatus x × x G. volutus × 2 x x x 2 × ...... - • . Hedl~tgella holmdelensis × x x × ...... " " RugogloblgerinahantkenJnoide s x x 3 x x x 3 5 x x .... - - - ' . - : : : : : : : - . . • . R hexacamerata x 7 3 4 3 5 5 2 5 5 3 5 [ 1 ...... " " " " • " . .... , , . • ...... , , . , . R. macro~ephala x x x x x x x ...... ' • • R reichel~ x x x x ...... " " R. rugesa 3 x 4 x x × 4 2 3 ~ ~ ! : i ~ . . . ; i : x x : : : : : : ; • . . R scotti 3 2 2 2 2 3 x x x x x ;7 ~ ~ : i Rsp. A 3 × 4 2 4 5 9 8 3 2 ~ : : : : R subpennyJ x ,~ ~ i ! ~ i °'=o,..,o~...... , , , ~ ii i i:. i:- =. i :- =. ~ :. i i Globotruncana aegyptiaca 4 2 2 2 5 3 4 "• , " . • . " . • . • G, ama 4 5 2 3 3 7 ~3j ~ .... : : : ; : - G. ¢o~i6a x x x x x x 2 I2i : : : : : " : : : G falsosluart: x ...... " " G. du~ 4 x ~ 2 , . . . . : : ; " : G ganssen x : : : : : : : : : G trinidadensi' ~ : : : : ' " " ' G stuarti x 2 × x ~ 3 2 ~ : : : : i i i : i : : : : : : : : : G stuartiformis x • x ~ ~ ~ :...... : : : : : : " : " : Pseudoguembelinacos[u~ata ! 2 2 12 ~ 0 6 6 (~ ! 3 6 3 ~ 2 ~ ...... ' ' " " P costata 3 B t 0 6 7 : = 10 ~ :...... : : : : : ' : " : " : P kempensis × × x x x × x x x ~ , 3 ~. . • ..... - , : : : : : : : : : P. palpebra 9 9 9 9 10 14 l0 8 7 6 g 23 Ll!~ ~ : : : : : : : : : P punctu;ata ~ x x x x [2] :~ : : : : : : : : : Pseudotextu~ana elegans ~ x x 2 2 ~ x x x × x [1 ] ~ : : ; : : ; : : : Heterebalixglabrans , , 2 ~ • : i ! i ::. : . : . : H globulosa 8 6 i 0 8 6 9 3 " " " H. striata !8 2O 9 !4 i n4 22 1 15 !6 2 ' [2) H americana x x ' Gublerina cuvillier, x "~ ...... i : Planoglobulina brazoens,s ~ ~ 3 P. c~rseyae ss × Racem~guembeh na fructicosa x x x x x x x i R. intermedia x x x 4 : R.p.~lli x × x x × x × × Ventrilabella eggeri x x x x x x × Chiloguembelitria taunca Globigerina moskvini 30 10 x 16 32 G. ~leodosP~ [2 [4] 10 i 60 37 21 9 55 70 32 18 Eoglobigerina hemisphaerlca 14 39 x 5 6 3 E. t~urica 17 16 29 ~ 0 43 32 x E tetrago~a i2] r41 [3] 2 41 42 74 x x Subbotina triloculinoides 2 x x x x 8 5 x x C4~otalia ~teae~pa i 3 ) 3 2 3 3 x 5 3 7 ! 0 GI. pseudol~ltoides !!2] !6 [11] 7 × x x 4 × 4 55 Total No, counted 245 Datum Events This Report Smit,1982 Premoli Silva&BoUi, Herm et al; 1973; Berggren,1978 1981 G trinidadensis ~ c[~ G. pseudo- G. pseudo - bulloides bu Iloides G taurlca ~- ' G.pseudo- G.pseudo- G. b bulloides bulloides b~ taurica G. G preaequa S triloculinoides ~ taurica G eugubina "~ G mos,vini G. pseudobu Iloides ~.. =1 h ]Eoglobigerina Pl P1 Gconusa ~' ~1] spp. Eoglobigerina spp ~_ G daubjergensis G. G. a G. G. a eugubina eugubina G taurica ~= eugubina G midwayensis a eugubina G eugubina G conusa, Eedita ~== ::)0 b G. conusa ~/'///////~ G. fringa W hornerstownetlsls G.lringa, Garcheoc. alG.cretacea P(I G.cretacea EXTINCTION ~ L R deformis A. A. A. A. rnayaroensis ~' M 3 mayaroensis mayaroensis mayaroens. A. mayaroensis Fig. 2. Comparison of Late Maastrichtian and early Paleocene planktic foraminiferal biozonations with biozonation of this report. The base of the black clay layer contains posi- Lower Paleocene tive anomalies of Ir and Os (Kuslys and Kra- henbuhl, 1983). The thin ferruginous layer The lowermost Tertiary was originally di- shows a drop in CaCO3 to less than 1%, a max- vided into Globigerina eugubina (K/T bound- imum in organic carbon, and a negative excur- sion in ~1~C (Keller and Lindinger, in prep.), ary to FA G. pseudobulloides) and G. At the base of the black clay layer (sample 541 ) pseudobulloides Zones (FA of zonal marker to FAG. trinidadensis, Premoli Silva and Bolli, are numerous compressed spherules of sanidine 1973 ). Subsequently, these zones have been di- and hematite. These spherules were originally vided into subzones with a numerical system described by Smit (1982) as altered microtek- tites, but they are now shown to be authigenic beginning with PO at the boundary (Berggren, in origin (Izett, 1987). 1978; Smit, 1982). In this report a further mod- ification of this system is proposed and illus- Based on planktic foraminifers the K/T boundary is generally placed at the extinction trated in Fig. 2 along with earlier biozonations. of nearly all Cretaceous species. However, the expanded K/T boundary at E1 Kef clearly shows Zone PO ~ Subzones POa, (Guembelitria cre- that not all species went extinct simultane- tacea), POb (Globoconusa conusa) ously, but spread out over about 32 cm of sedi- The 50cm thick black clay unit marks the ment with major extinctions beginning 25 cm basal Paleocene. This unit contains both Cre- below the boundary (discussed below). Globo- taceous survivors and the evolution of the ear- truncana and Rugoglobigerina species extinc- liest Tertiary species, but is characterized by an tions most closely coincide with the lithological abundance of Guembelitria cretacea. Smit and geochemical boundary. (1982) defined an earliest Paleocene PO, or G. 246 cretacea Zone as ranging from the K/T bound- subzone comprises the main occurrence of the ary to the first appearance (FA) of Globigerina Globigerina taurica fauna. At E1 Kef this fauna minutula. (Note, the species illustrated by Smit is most common above the extinction of G. eu- ( 1982, pl. 3 ) as G. minutula is Globoconusa con- gubina (Table I ). It is therefore proposed that usa Chalilov). Based on the present faunal the definition of Subzone Plb (G. taurica) be analysis of E1 Kef it is proposed that Zone PO modified to include the main occurrence of the be modified to range from the K/T boundary to nominate species and that the subzone be di- the FA of Globigerina eugubina with a subdivi- vided into Plbl and Plb2 based on the extinc- sion into subzones a and b. Subzone POa re- tion of G. eugubina. mains Smit's original definition of G. cretacea Zone. Subzone POb, or Globoconusa conusa Subzone P l bl (Eoglobigerina spp.) ranges from the FA of G. conusa to the FA of G. Partial range subzone from entry of main eugubina. In the E1 Kef section G. conusa first proliferation of Eoglobigerina spp. to the ex- appears 22 cm above the K/T boundary and G. tinction of G. eugubina. At E1 Kef this subzone eugubina appears 52 cm above the boundary comprises 2 m of sediment from 3 to 5 m above near the top of the black clay unit. These two the K/T boundary. The fauna in this interval subzones provide further stratigraphic control is characterized by a rapid terminal decline in for the K/T boundary interval. Moreover, they the abundance of G. eugubina, peak abundance encompass the black boundary clay which ini- of Eoglobigerina spp., first abundant Chilo- tiates a profoundly changed ecosystem after the guembelitria taurica and Globigerina taurica, and K/T boundary event, the first appearance of G. pseudobuUoides. At E1 Kef Subzone Plbl, occurs within an interval of Subzone Pla (G. eugubina) rapidly rising carbonate (from 20 to 50% ) and Partial range subzone from FA Globigerina sharp increase in surface and bottom water pro- eugubina to first proliferation of Eoglobigerina ductivity as indicated by a positive ~ 13C shift of about 2%0 (Keller and Lindinger, in prep.). spp. The lower boundary of this subzone is here modified from Smit (1982) who based it on the Subzone Plb2 (Globigerina taurica) FA of G. conusa (G. minutula of Smit). Subzone Partial range subzone from the last appear- Pla comprises most of Globigerina eugubina ance of G. eugubina to the last appearance of G. Zone and is characterized by dominance of G. taurica. This subzone is characterized by peak eugubina common G. conusa and decreasing abundances of G. taurica and Chiloguembelitria abundance of Chiloguembelitria danica, taurica and common Globoconusa daubjergen- At E1 Kef about 250 cm of sediment represent sis. At Al Kef Subzone Plb2 encompasses 4 m Subzone Pla. The basal 50 cm of Subzone Pla ( + 5 to + 9 m) of relatively carbonate rich consist of dark grey clay with 6-10% CaCO3. ( ~ 45% ) marly sediments. Few burrows are apparent. This unit grades up- ward into grey clay-rich shale about 200 cm Sediment accumulation rates thick with an average 14% CaCO3. Above this No magnetostratigraphy has been done at E1 unit carbonate increases rapidly in increasingly Kef. However, age estimates can be obtained marly sediments, from sediment accumulation rates based on da- tum events of foraminifers dated elsewhere. The Subzones Plb K/T boundary lies within the reversed polarity This subzone ranges from the entry of com- interval between anomalies 29 and 30 (Berg- mon Eoglobigerina spp. to the extinction of G/o- gren et al., 1985 ). Age estimates for the bound- bigerina eugubina. As originally described by ary range from 66.2 to 66.8 Ma with the best age Salaj (1977) based on Tunisian sections this estimated by Haq (1983) at 66.4 Ma. The first 247 appearances of the index species Globigerina boundary. This decline occurs in stages with the eugubina and G. pseudobuUoides have been dated maximum number of species extinct in sample at 66.35 Ma and 66.10 Ma respectively (Berg- 541 at the K/T boundary. Total species diver- gren et al., 1985 ). Based on these ages average sity remains very low ( 10 species or less) in the sediment accumulation rates can be calculated, first 50 cm of black clay (Zone POa, b) above If we estimate the K/T boundary at 66.4 Ma, the K/T boundary. Above this level species di- then the first 52 cm of black clay above the versity increases to 15-20 species and remains boundary to the first appearance of G. eugubina at this level through Subzones P la (G. eugu- was deposited within 50,000 years, or a sedi- bina) and Plb. mentation rate of 1.04 cm/1000 years. The The number of last appearances of species in nearly 350 cm sediments of the G. eugubina Zone Fig. 3 illustrate the rapid and sequential change to the first appearance of G. pseudobuUoides re- occurring across the K/T boundary. In con- flect 250,000 years, or an average sedimenta- trast, few species disappear before or after the tion rate of 1.40 cm/1000 years. We do not have K/T boundary event. Furthermore, there are good estimates of sediment accumulation rates relatively few first appearances of species and for the carbonate rich marls above or below the these occur in two intervals. The first pulse in K/T boundary, evolutionary diversity occurs in the 50 cm thick black clay layer above the K/T boundary. In Species diversity this interval 10 species evolve. The second pulse occurs in the lower part of Subzone Plb (Eo- Total species diversity as indicator of envi- globigerinaspp. ) coincident with increased car- ronmental conditions is illustrated in Fig. 3 bonate sedimentation and surface productivity along with the number of first and last appear- (Keller and Lindinger, in prep. ). ances of species and percent carbonate in sedi- There appears to be a positive correlation be- ments. Total species diversity is based on the tween high species diversity or evolutionary actual number of species present in a sample turnover, high carbonate in sediments and po- plus species temporarily absent, but present in sitive 13C values. High species diversity corre- the samples immediately preceeding and suc- lates with high carbonate and high productivity, ceeding. The latter are represented by a line at low species diversity correlates with low car- the top of the histogram. This temporary ab- bonate and low productivity. This suggests that sence is considered to be due to rarity of species species diversity is strongly linked to surface or small sample size. productivity. During the latest Cretaceous species diver- sity of planktic foraminifera remains relatively Sequential nature of K/T boundary constant around 50 species. But during the early extinctions Tertiary the number of species averages 15-20. This represents an average decline of 78% in Cretaceous/Tertiary boundary extinctions at total species diversity as observed at E1 Kef, E1 Kef do not appear to have occurred geologi- Tunisia. In contrast, benthic species diversity cally instantaneously, nor do they correlate declined only 37% during the same time inter- precisely with the geochemical and lithological val (Keller, 1988a). Similar to benthic faunas boundary. Species extinctions begin between 25 the decline in planktic species diversity be- cm below the boundary event (Ir anomaly) and tween Cretaceous and Tertiary faunas is rapid end about 7 cm above. Although it is possible beginning in sample 539 25 cm below the K/T that bioturbation may have obscured the ex- boundary and reaches a maximum low of 7 spe- tinction record, this does not seem a major cies, or 86% decline in diversity, 7 cm above the problem in the El Kef section. Burrows of 2-3 248 EL KEF, TUNISIA PLANKTIC FORAMINIFERA ~'~ ~ SPECIES LAST FIRST CaCO3(%)1 ,,~ t~ i -- .~ ~, d DIVERSITY APPEARANCE APPEARANCE m ~ -597 E_ zo "592 50 . E 4o q l - 30 i .5~ ~5 ] > E 20 " (D ¢'-" .575 c0 ~_. :565 "-- 3 ('~ 10 _ -i C --- -1 L :555 "0 05 ~ ] -~0 1 , 3"--1 , v I I ~5 ~ 5~ I, I (') -20 533 0 50 ...... ~50""5 0 30 ' 6Or. No, Species No Species 1 Lindinger & Keller, 1988 Fig. 3. Species diversity, first and last appearances of species and percent carbonate in late Maastrichtian to early Paleocene sediments at El Kef, Tunisia. cm length were observed below the boundary in latest Cretaceous faunas and their extinction clay layer, but no bioturbation was observed is likely due to minor environmental perturba- across the 2-3 mm thick ferruginous layer at tions unrelated to the K/T boundary event. the base of the 50 cm thick black clay or whithin The K/T boundary extinctions begin in sam- the clay layer. Moreover, species extinctions ple 539, 25 cm below the boundary clay. Six spe- show a systematic sequential nature which is cies last appear in this sample. Four of these unlikely to be caused by bioturbation, species are morphologically similar with mul- Ranges of species are illustrated in Fig. 4. tiple globular chambers arranged in either bis- Cretaceous species found in early Tertiary sam- erial or multiserial fashion ( Ventrilabella eggeri, ples are marked with dots; for the most part Racemiguembelina fructicosa, Pseudotextularia these are considered reworked specimens. Spe- carseyae, P. elegans) (Fig. 4, Plate I). Eight cies with dots connected by a line are consid- species last appear in sample 540, a sample ered as Cretaceous survivors and will be spanning 0-5 cm below the K/T boundary. Ex- discussed later. Six species disappeared during tinctions at this level appear to affect primarily the latest Cretaceous between 1-4 m below the the globotruncanid and globotruncanellid spe- K/T boundary (Fig. 4 ). These species are rare cies (5 of 8 species) (Globotruncana trinidaden- 249 AG E MAASTRICH. PALEOC E NE ZONE P. deformis ~ PObI G.eugubina Pla I Plbl [ Plb2 IP1¢ DEPTHtm) ~ ~ ~ ~ ~ o ~ o~ ~~ b~ o~ o~ o~ t g g~ ~ " ~, -4 I b -I -4 -I "I "4 -I i -I .4 sample no. ~ ~ ~ § ~ ~ ~ ~ ~ ~ = ~= ~ o== :- , (D 0 O. ultr=mJcra L hess/ m m A. blowi ~ "~ R. subpennyi G. gansse~ m ~ ~ +~ H. holmdelensis "~ ~ v.~w~i ,,m ,,--, r" rr'] ..= ®~ (3. ~rti~'ms m R. fn~c.~ m m ~ Z O. uinidllcmis m mmm i O ...... I " (3 "rl ~-o G.~ O. momnouthensis • G. mulespina mmm mm m • ~ ,.G G. cuvilicri ~j~ Z G. falsostua~ m G. c~ica , Z ~ ~ ~ R. scotti -n ..,o~,. m ~ "=" R.m,~ __X) .~.~ R. hamkeninoides R. bn~z~nsis G. subcizcumnodifef m mmm. ~ ~ ~m m~ R. reicheli m m m ~ 0 17,. hcxacametata • . • • o ..... ~2 P. defo~mis . , P. punctu]ata mmm m m P palpebra • • ~ P. kempensis * * • • * H. O,b~, • ..- ~ H. 81olmlom , ..~ O. sul=~rinams ' , , • P. carseyaes.l. _s , m ~ ~ O. upera 03 Rusogtobigerina sp. A. - Z ~ H. su~am :: ; ,__~,a* "4 • H. navm'mensis ~ ()0 c~ C,O'cmvmc'~cnsiSdanica ~ ram, m .m, ,m. =" " 0 :. E. fiingl a= IN ml l m I ~ ~ G. tlifolia mm OL alcheocomp~ssa m m • m • • m m m ) "g t~ W. hon~rswwnensis • • i , o~ o .... . E, impli~faittla tim t m* m I mmm E. eobulloides • mm mm m m m G. eegubina am G. midwayensis m m • E. ~am'Jca m m G. daubjergensis ~ ~.~ E. h~-i~a ,, -~ C. irregularis G.c.~ modadni , ~ S. tl~locolinoides ' r.? 250 PLATEI 251 sis, G. contusa, G. aegyptiaca, Globotruncanella palpebra, P. kempensis, P. costata, Plate II, Fig havanensis, G. petaloidea, Plate I, Fig. 4). 4). In contrast, the smaller biserial heteroheli- Twelve species last appear in sample 541, a cid forms decline in abundance throughout the sample spanning 0-5 cm above the base of the K/T boundary extinction interval, but con- boundary clay. Major extinctions continue tinue to be present in the early Tertiary where among the morphologies previously affected: they disappear in the Globoconusa conusa four globotruncanid species (G. stuarti, G. [al- (POb) and G. eugubina (Pla) Subzones (Fig. sostuarti, G. conica, G. fornicata) and three spe- 4). In addition, Rugoglobigerina sp. A, a small cies of multichambered biserial or multiserial rugosa-like form also appears to survive into the morphology ( Racemiguembelina brazoensis, R. Paleocene. intermedia, R. powelli ). No species of the mul- Species range data from E1 Kef, Tunisia, thus tiserial group appears to survive this interval at show that the K/T boundary extinctions did not E1 Kef. In addition to these species extinctions occur simultaneously, but stretched out over a a new morphology, the rugoglobigerinids, are considerable time period (-25 to +7 cm). now also affected by extinctions (Rugoglobiger- Moreover, there appears to have been a system- ina scotti, R. robusta, R. macrocephala, R. hantkeninoides), atic elimination of species groups with common morphologies suggesting that successive niche Ten species disappear in sample 542, span- ning the interval 5-10 cm above thhe K/T destruction occurred over this time period. It boundary. The remainingrugoglobigerinidspe- may also indicate that species with certain cies disappear at this time (R. reicheli, R. ru- morphologies were more tolerant of environ- gosa, R. hexacamerata), as well as the last mental changes than other groups. Thus, the surving globotruncanid species (G. arca). In history of species extinctions at E1 Kef shows addition, a new group, the biserial pseudotex- the early disappearance of complex large bis- tularids and pseudoguembelinids are now erial to multiserial forms followed by globo- strongly affected by extinctions (Pseudotextu- truncanid and globotruncanellid species. Most laria deformis, Pseudoguembelina punctulata, P. of these species are relatively rare in the upper- PLATE I All Figures at 150× magnification (scale bar= 100 #m). All samples form P.deformis Zone. Species disappearancesin the El Kef section are as follows: Figs. 14-17, 25 cm belowthe K/T boundary:figs. 9-13, 5 cm belowthe K/T boundary:figs. 4-8 at the boundaryand figs. 1- 3 between5-10 cm above the boundary. 1 Rugoglobigerinascotti (Bronnimann),sample 540. 2. Rugoglobigerinarobusta Bronnimann, sample 540. 3. Rugoglobigerinamacrocephala Bronnimann,sample 540. 4. Globotruncanastuarti (Lapparent ), sample 540. 5. Planoglobulina brazoensis Martin, sample 539. 6. Globotruncanaaegyptiaca Nakkady, sample 540. 7, 8 Globotruncana arca (Cushman), sample 540. 9. Globotruncana contusa (Cushman), sample 539. 10. GlobotruncaneUahavanensis (Voorwijk),sample 539. 11. Globotruncanaarca (Cushman), sample 539. 12,13. GlobotruncaneUapetaloidea (Gandolfi),sample 540. 14. Ventrilabellaeggeri (Cushman), sample 539. 15. Racemiguembelina[ructicosa (Egger), sample 539. 16. Planoglobulina carseyae (Plummer),sample 539. 17. Pseudotextularia elegans (Rzehak), sample 539. __m__ / 253 most Maastrichtian of E1 Kef. These extinc- man (1984). However, this selective nature of tions are followed by the somewhat more robust species extinctions has not been demonstrated and more common rugoglobigerinid and smaller, before for the K/T boundary extinction event. simpler biserial pseudotextularid species ex- The K/T boundary extinctions are generally tinctions. The small biserial heterohelicids as believed to have occurred too rapidly for selec- well as some globigerinellids, hedbergellids and tive species extinctions to occur. This study of guembelitrids survived longest and appear to E1 Kef faunas shows that even in a very rapidly have been best adapted for survival, or catastrophically changing environment, the Plates I and II illustrate the characteristic theory ofselectivityinspeciesextinctionsholds. morphologies of species extinct across the K/T boundary. All scanning electron micrographs Speeies survivorship (SEM) were taken at the same magnification for Late Cretaceous species, except for the small Foraminiferal workers commonly report the K/T boundary survivors, to show size changes, near extinction of all planktic species at the Because K/T boundary survivors and early Cretaceous/Tertiaryboundary. However, a few Danian species are extremely small, these spe- authors have observed an impoverished relict cies are illustrated at over three times the mag- faunule composed of dwarfed Late Cretaceous nification of Late Cretaceous species {Plates I genera such as Rugoglobigerina, Planomalina, and II). The following characteristics are im- GlobigerineUoides and Hedbergella (Berggren, mediately obvious: large speciesdisappearear- 1965; Hofker, 1978; Blow, 1979). Although lier than small species, complex morphologies many other workers have observed the pres- disappear first, and simple morphologies sur- ence of small Cretaceous species in early Ter- vive longest. Thus, large, complex overspecial- tiary sediments, they have generally assumed ized morphologies are unsuited for adaptation that their presence was a result of bioturbation in a rapidly changing environment, whereas or reworking of older sediments. Reworked generalists with simple morphologies are best specimens are difficult and often impossible to suited for survival. This illustrates what has identify especially if they are not discolored or long been known in theory and was observed show differential preservation and are ob- among late Cretaceous invertebrates by Kauff- viously anomalous in age with the rest of the PLATE II Figs. 1-8 represent Late Cretaceous survivors, magnified 500 × (scale bar- 10 #m ). Figs. 9-17 represent species extinct about 7 cm above the K/T boundary and magnified 150× (scale bar--100 llm). Note the smaller size of Cretaceous survivors which are actually less than 1/3 the size of average Cretaceous species. 1. Guembelitria cretacea {Cushman), POa, G. cretacea Subzone, sample 544. 2. Globigerinelloides subcarinatus (Bronnimann), Pla, G. eugubina Subzone, sample 552. 3, 4. GlobotruncaneUa caravacaensis Smit, Pla, G. eugubina Subzone, sample 556. 5. Heterohelix navarroensis Loeblich, POb, G. conusa Subzone, sample 548. 6. Heterohelix striata (Ehrenberg), POb, G. conusa Subzone, sample 548. 7, 8. Globigerinelloides aspera, Bolli, Pla, G. eugubina Subzone, sample 558. 9. Pseudotextularia deformis (Kikoine), P. deformis Zone, sample 541. 10. Pseudoguembelinapunctulata {Cushman), P. deforrnis Zone, sample 538. 11. Pseudoguembelina palpebra (Bronnimann and Brown), P. deformis Zone, sample 538. 12. Pseudoguembelina kempensis (Esker), P. deformis Zone, sample 538. 13. Pseudoguembelina costata (Carsey), P. de[ormis Zone, sample 538. 14. Rugoglobigerina rugosa (Plummer), P. deformis Zone, sample 539. 15. Rugoglobigerina reicheli (Bronnimann), P. deformis Zone, sample 539. 16,17. Rugoglobigerina hexacamerata (Bronnimann), P. deformis Zone, sample 539. 254 faunal assemblage. The early Danian fauna at and Gartner (1986) reported that a significant E1 Kef clearly contains some intervals with re- number of nannoplankton species also sur- worked Cretaceous specimens, particularly glo- vived the K/T boundary at Brazos River, Texas. botruncanids (Fig. 4). However, many samples Some of the ten Cretaceous planktic forami- contain the smaller heterohelicid, pseudotex- niferal species which appear to have survived tularid and hedbergellid species without other into the early Tertiary are illustrated in Plate characteristic Cretaceous species. This sug- II. Their disappearances in early Tertiary sed- gests that at least some of these species may be iments occurs gradually over about 300,000 Cretaceous survivors, years or 300 cm of sedimentation at E1 Kef, Ranges of ten species believed to be Creta- Tunisia. Of these species Heterohelix globulosa ceous survivors are illustrated in Fig. 4 and and Globigerinelloides subcarinatus disappear at numbered 1-10 with lines connecting their oc- the top of POb (G. conusa) Subzone. Hetero- currences (dots) in specific samples above the helix planata, Globigerinelloides volutus and K/T boundary. These species are believed to be Pseudotextularia costulata disappear in the Cretaceous survivors because of their frequent lower part of Subzone Pla (G. eugubina). Glo- presence in early Tertiary samples in absence bigerinelloidesaspera and Rugoglobigerina sp. A of other characteristic Cretaceous species, com- disappear within Subzone Pla. The remaining parable preservation to early Tertiary fauna, three Cretaceous survivors (H. navarroensis, H. and generally smaller overall size when com- striata,G. cretacea, G. caravacaensis) disappear pared with the same species in Cretaceous sed- near the top of Subzone P la, or in the early part iments. Moreover, the same ten species have of Subzone Plb (Fig. 4). also been found in coeval early Tertiary sedi- It is possible that other small Cretaceous spe- ments from Brazos River, Texas (Keller, un- cies also survived into the early Tertiary, as for published data). A graphic correlation plot instance minute specimens of Rugoglobigerina based on first and last appearances of species hexacamerata,Heterohelixglabrans, Pseudotex- at E1 Kef plotted against the same species at tulariacostata and P. kempensis. These species Brazos River indicates that these species dis- have also been found in early Tertiary sedi- appeared at about the same time in both re- ments of both E1 Kef, Tunisia and Brazos River, gions. This coincidence is unlikely ~o be the Texas, but their disappearances in these sec- result of reworking of Cretaceous sediments. It tions are not coeval. It will be necessary to study strongly suggests that more species survived the other sections in different parts of the world in K/T boundary event than generally assumed, order to confirm the nature and extent of sur- A preliminary ~:~Canalysis of some Brazos River vivorship of these species. survivors (H. striata, G. cretacea) reveals nearly A few observations on the nature of Creta- a 2¢i~ difference in values before and after the ceous survivorship can be made however from K/T boundary (Keller and Lindinger, unpub- this study of E1 Kef faunas. The Cretaceous lished data) This isotopic difference clearly survivors are all small species which are not shows that the early Tertiary specimens are not common in late Cretaceous faunas except for from reworked Cretaceous sediments. E1 Kef H. globulosa and H. striata. Without exception species have not been analyzed isotopically be- the Cretaceous survivors are of a basic primi- cause insufficient specimens are present, tive morphology consisting of small globular Species survivorship among calcareous nan- chambers arranged either in uniserial coiled noplankton across the K/T boundary has been fashion (globigerinellids, hedbergellids), or observed earlier by Thierstein (1982) from biserial (heterohelicids and pseudotextularids) deep-sea sections and by Perch-Nielsen and triserial fashion (such as Guembelitria). ( 1981a,b ) from El Kef, Tunisia. Recently, Jiang Apertures are simple and there is little wall or- 255 namentation (Plate II ). The simple morphol- ent author also studied the Carvaca section, but ogy and small size imply that these species were found preservation of planktic foraminifers too generalists able to tolerate a wide range of en- poor for positive species identification for the vironmental conditions. During the late Cre- purpose of evolutionary studies. Smit's obser- taceous when niche partitioning was at its vation of a monospecific assemblage of Guem- height, as seen in the highly diverse faunas, this belitria cretacea at the base of the Tertiary at E1 ecologic tolerance appears to have been a Kef could not be confirmed in this study (see hindrance to developing large populations of Fig. 4, Table I). individual species. But after the destruction of First appearances (FA) and ranges of new the late Cretaceous ecosystem, this tolerance early Tertiary species at E1 Kef are shown in aided their survival. Fig. 4 and illustrated in Plate III. It is apparent To understand why some Cretaceous species that species evolution was most active imme- survived it will be necessary to study their pop- diately after the K/T boundary event in the 50 ulations in late Cretaceous and early Tertiary cm of black clay overlying the 2-3 mm thick deposits by means of quantitative stratigraphy, ferruginous boundary layer. The black clay layer morphometrics and stable isotopes. Stable iso- represents deposition over about 50,000 years tope analysis of survivors in relationship to non- (K/T boundary to FAG. eugubina). The next survivors is expected to yield clues to their ecol- major pulse in species evolution occurs about ogic habitat and changes in watermass strati- 250,000 years later (FA G. eugubina 66.35 Ma fication across the K/T boundary, to FAG. pseudobulloides 66.1 Ma) in Subzone Plb, (Eoglobigerina spp.) during a rapid in- Early Tertiary evolution crease in carbonate deposition. The earliest new Tertiary species (Guembel- The evolution of the early Tertiary fauna after itria trifolia, Chiloguernbelitria danica, Globi- the K/T boundary extinctions is still largely a gerinafringa) appear in sample 542, 7 cm (about mystery. The problem lies primarily in the 7000 years) above the K/T boundary. Both G. paucity of sections with a complete sedimen- trifolia and Ch. danica evolved from G. cretacea tary record and high sedimentation rates. For and specimens of similar morphology can be instance, K/T boundary sections from the deep- found occasionally among G. cretacea popula- sea generally lack Zone PO and frequently also tions of the late Cretaceous. Therefore, its ev- Subzone Pla (G. eugubina). Thus, 50,000- olutionary first appearance may occur earlier. 300,000 years of the earliest Tertiary history is Woodringina hornerstownensis, Eoglobigerina either missing or condensed into a few centi- edita, Globoconusa conusa and Globorotalia ar- meters of sediment. The most expanded history cheocompressa first appear 17-22 cm ( ~ 20,000 for this interval is present at E1 Kef, Tunisia years) above the K/T boundary. Three new where 50 cm of black clay records Zone PO and species appear near the top of the black clay 2.5 m record Subzone Pla. A similarly high layer (G.eugubina, E. eobulloides, E. sedimentation rate is observed at Brazos River, simplicissima). Texas (Keller, 1988b). Both sections represent Species populations remain relatively stable shelf sequences deposited at about 150-350 m and there is little evolutionary diversity during depth. Thus, early Tertiary radiation can be deposition of the 250 cm of sediment represent- uniquely studied in two widely separated re- ing Subzone Pla (G. eugubina). Only two spe- gions of the world, cies appear at this time (G. midwayensis and Smit {1982 ) earlier studied planktic forami- Eoglobigerina taurica). niferal evolution in early Tertiary sediments at The second pulse in evolutionary diversity in Caravaca, Spain and E1Kef, Tunisia. The pres- Subzone Plbl (Eoglobigerina spp. ) marks the 256 PLATE III 257 entry of Globorotalia pseudobuUoides and re- except among biserial and multiserial forms lated species (G. moskvini, G. pseudobulloides (D'Hondt and Keller, 1987). subquadratus), the diversification of the G.(E.) The second pulse in species diversification tauricagroup (E. tetragona, E. hemisphaerica), occurs under more favorable environmental the first appearances of Chiloguembelitria taur- conditions evidenced by a return of carbonate ica, Globoconusa daubjergensis, Subbotina tril- in sediment to pre-K/T levels, higher produc- oculinoides, and Globorotalia preaequa. The tivity as seen by a 2%o positive 13C shift, warmer remaining Cretaceous survivors and G. eugu- more stable water temperatures and increased bina become extinct at this time. water mass stratification (Keller and Lindin- Thus, early Danian faunas show two major ger, in prep. ). The morphologies evolving at this pulses of species diversification. The first pulse time are larger and more complex globigerine is recorded in the 50 cm of black clay above the and globorotalid forms. This implies that more K/T boundary and immediately follows the equitable environmental conditions returned after the G. eugubina Subzone (Pla) or about major Cretaceous species extinctions. This ev- olutionary diversification occurs at a time of 300,000 years after the K/T boundary event. extremely low species diversity, less than 5% Population turnover CaC03 in sedimens, very low productivity as in- dicated by low 13C values and generally cool Quantitative faunal analysis reveals trends in fluctuating temperatures (Keller and Lindin- species dominance that yield clues to the nature ger, in prep. ). The species that evolved under of faunal assemblages and stability of the en- these stressful conditions were very small (lack vironment. However, late Cretaceous faunas of nutrients? ) and of simple morphology with with their wide variations in species sizes yield small globular chamber arranged in uniserial unique problems. To observe dominant species coiled fashion (Globorotalia archeocompressa, trends in both small and larger forms both the Globigerina eugubina), biserial ( Woodringina > 63 pm and > 150 pm size fractions were ana- hornerstownensis), triserial (Guembelitria tri- lyzed. It was found that the > 63 pm size frac- folia), quadriserial (Globoconusa conusa) or tion provided the most informative data on globigerinid. Although the basic morphologies faunal turnover and evolution across the K/T of evolving species is the same as those of Cre- boundary. Figure 5 illustrates this data for the tacous survivors there is no obvious ancestor- most common species. The commonly studied descendent relationship between these groups > 150 pm size fraction was found to bias to- PLATE III Figs. 1-7, 11 are magnified 300× (scale bar-100/lm). Figs. 8-10, 12-20 are magnified 500× (scale bar-10 gm). Note, in actual size species of Subzone Plb are nearly twice as large as those of Subzones Pla and PO. 1. Globigerinamoskvini Shutskaya, Plbz E. taurica Subzone sample 596. 2. Globigerinapseudobulloides Plummer, Pla, G. pseudobuUoides Subzone, sample 599. 3. Chiloguembelitriataurica Morozova, Pla, G. pseudobuUoides Subzone, sample 599. 4, 5. Globigerina(Eoglobigerina) taurica Morozova, Plb2, E. taurica Subzone, sample 591. 6, 7,11 Globigerina(Eoglobigerina) hemisphaerica Morozova, Plb~, E. taurica Subzone, sample 592. 8-10. Globigerinaeugubina Luterbacher and Premoli Silva, Pla, G. eugubina Subzone, sample 565. 12-14. Globoconusaconusa Chalilov, POb, G. conusa Subzone, sample 548. 15. Woodringinahornerstownensis Olsson, Pla, G. eugubina Subzone, sample 553. 16,17 Globigerinafringa Subbotina, POb, G.conusa Subzone, sample 548. 18. Globigerinacf edita Subbotina, POb, G. conusa Subzone, sample 548. 19,20 Globorotaliaarcheocompressa Blow, Pla, G. eugubina Subzone, sample 554. 258 ,,,:~d '~ EL KEF, TUNISIA - ~ Planktic Foraminifera t3 " 599 --bPo i ~" 7.0 50 -59o ~ 4+0 i o,.. 3,0 --c "586 b • ~ o.s :sso ~. e , Ill :t'l. ~530 i 20 ,',°o°J0 Fig. 5. Relative population abundances of common species in the size traction > 63 #m. Note, only small biseria] or planispira] species are common in uppermost Maastrichtian sediments. In the early Paleocene species dominance of any one species is short lived each being successively replaced by another (competing?) species. Low species diversity and short lived popu- lations indicate highly unstable stressed conditions. wards large species without showing recogniz- bers of juveniles present which could not be able trends ( Fig. 6 ). For spot samples the > 105 identified to species level with the light micro- #m size fraction was also analyzed for the latest scope. Thus, for biostratigraphic resolution, Cretaceous and early Tertiary Subzone Plb. ecologic and evolutionary studies the > 63/lm The results were not significantly different from size fraction was found to provide the best re- the smaller size fraction ( > 63 #m). The still sults for the K/T boundary interval. Neverthe- smaller size fraction (>32 /tin)was also ex- less, major faunal successions can also be amined, but was found to be impractical for recognized in the > 150 #m size fraction as il- quantitative studies because of the large num- lustrated by Gerstel et al. ( 1986, 1987 ). Species 259 ------~--600 E. ~-~ 9 EL KEF, TUNISIA Ip ) - 8 Planktic Foraminifera ~ 7 >150#m -- T 5 JJ-- ~ 3 III -r-- / ~- 1.0 "~ { / ALL PLANKTIC FORAMINIFERA Q..:#- o (:~ o ~ ----i o ------ {D -1.0 o ~0 -3 ~..i1~_~o o ~ n- -4 Fig. 6. Relative population abundancesi!ltlttl of common species in the size fraction > 150 #m. Note the similarity in common species below the K/T boundary, absence of foraminifers above the boundary and return of species in the > 150 #m size fraction in Subzone Plb. The latter represents a major loss of information when compared with Fig. 5. abundance curves illustrated in Figs. 5 and 6 this size fraction in the latest Cretaceous. have error bars of _+ 1.5% as tested by duplicate All species in the earliest Tertiary assem- faunal counts, blages are smaller than 150/~m. The fauna after Figure 5 illustrates that among small late the K/T event is dominated by Guembelitria Cretacous species the heterohelicid, pseudo- cretacea through the 50 cm black clay and 50 guembelinid and globigerinellid survivors were cm dark grey clay. Only Globoconusa conusa co- most abundant. However, the abundance of exists in relative abundance at this time. The each species declined immediately after the K/ subsequent rise in G. eugubina is associated with T boundary to less than 2% and is therefore too a permanent decline in G.cretacea suggesting small to show on Fig. 5. Among the > 150/lm niche competition or destruction. Globigerina fauna this group, except globigerinellids, is also eugubina, G. conusa and C. danica dominate most abundant (Fig. 6). Most surprising how- Subzone Pla (G. eugubina). These rapid suc- ever, is the high species diversity, but overall cessive species replacements suggest highly low abundance among the larger species of Rug- stressed unstable environmental conditions re- oglobigerina and Globotruncana. Thus, no ad- lated to low surface productivity {Keller and ditional information was gained by analyzing Lindinger, in prep. }. Similarly stressed condi- 260 tions were observed by Gerstel et al. (1987) tinctions start before the K/T boundary event? from DSDP Site 577 who concluded that low The early disappearance of complex species productivity adversely affected faunal stability, suggests a changing environment preceding and Dramatic productivity changes in the early thus unrelated to, a K/T boundary impact. Danian have been observed by other authors There is also selectivity in species extinctions including Hsu and McKenzie (1985) and Za- before and after the K/T boundary event (Ir choset al. (1986). anomaly). Complex larger species disappear The first major faunal change in the early before smaller less ornate species and primitive Tertiary occurs during Subzone Plbl (Eoglo~ small species survive longest. This selectivity bigerina spp., Fig. 5 ) and marks the first recov- in species extinctions suggests sequential elim- ery of the ecosystem after the K/T boundary ination of less tolerant species. event. Major environmental changes are im- Habitats of Cretaceous species are relatively plied by the increased evolutionary diversity and unknown; however, a clue can be obtained from the first appearance of larger species ( > 150 ~m, oxygen isotope ranking of planktic foramini- see Fig. 6, Plate I). Perhaps the most signifi- fera. Isotopically light species generally live in cant change at this time was a major increase surface waters whereas isotopically heavier in surface water productivity as indicated by in- species live in deeper waters (Boersma and creased carbonate sedimentation and a 2(:i~, Shackleton, 1981; Keller, 1985). Unfortu- positive shift in 1:~C. Although carbonate sedi- nately, not all late Cretaceous species have been mentation reached pre-K/T boundary levels, isotopically ranked, but work by Boersma and surf'ace productivity remained well below. Ox- Shackleton (1983) and Keller and Thierstein ygen isotope ratios indicate that temperatures (unpublished data) imply that large complex were more stable, but remained cool at this time species including globotruncanids live in rela- (Keller and Lindinger, in prep. ). The change tively deeper waters whereas rugoglobigerinids in faunal diversity also suggests an increase in and pseudotextularids are surface dwellers. This watermass stratification, suggests that pre-K/T boundary extinctions largely affected deeper dwelling planktic spe- Discussion and conclusions cies first and surface dwellers survived longest. This implies changes in the water mass struc- The Cretaceous/Tertiary boundary section ture, temperature or chemical conditions of' at E1 Kef, Tunisia illustrates that high resolu- subsurface waters prior to the K/T boundary tion stratigraphic control is possible in conti- event. nental shelf sections with high sedimentation A clue to pre-K/T boundary species extinc- rates. But it is not clear whether the faunal se- tions may also be found in the convergent oxy- quence is representative of open oceanic con- gen isotopic trends of bottom and surface waters ditions where the K/T boundary interval is at E1 Kef (Keller and Lindinger, in prep. ). This much more condensed and hence stratigraphic convergent trend is primarily due to more neg- resolution is lost. Detailed studies of deep-sea ative benthic oxygen isotope values suggesting sequences (on acm scale or less) will be nec- an influx of warm saline bottom water. A sim- essary to correlate and compare continental ilar convergent oxygen isotopic trend was ob- shelf with oceanic environments, served by E. Barrera (pers. comm., 1987) from The K/T boundary extinctions at E1 Kef be- DSDP Sites 689,690 and Seymour Island. This gin about 25 cm below the lithologic and geo- implies that warm saline bottom waters may chemical boundary (Ir anomaly and base of have affected oceanic circulation globally. In- black clay layer) and continue through 7 cm creased production of saline waters may have above the boundary. Why did major species ex- occurred in low latitude marginal seas (Brass 261 et al., 1982 ) possibly triggered by a lowered eus- sitive shift of about 2 %o in 13C. Productivity, tatic sea level (Peypouquet et al., 1986; Keller, however, stabilized below pre-K/T boundary 1988a). It is likely that increased salinity and levels at this time. its chemical consequences are responsible for The Cretaceous/Tertiary boundary section the pre-K/T boundary species extinctions, of E1 Kef has shown this mass extinction event The K/T boundary event marked by Ir and to be more complex than previously thought. Os anomalies (Kushlys and Krahenbuhl, 1983 ), Although the findings of this study clarify many maximum total organic carbon, a dramatic de- aspects of the K/T boundary event, such as the crease in CaCOa to less than 1% and drop in nature of extinctions and species survivorship, surface productivity (Keller and Lindinger, in subsequent radiation of early Tertiary species p~ep. ) also coincides with maximum species ex- and the recovery of the ecosystem, many new tinctions. Primarily surface dwellers are af- questions are raised. For instance, why did the fected at this time (rugoglobigerinids, ecosystem not recover from the K/T boundary pseudotextularids). This event appears to be shock for so long? Can a single impact explain unrelated to, but superimposed upon the ad- such a long recovery period? The effects of an verse environmental trend and associated spe- impact event, with an increase in atmospheric cies extinctions that began during the late dust and possible warming due to the green- Maastrichtian. The geologically instantaneous house effect, are likely to be much shorter. At geochemical changes at the boundary point to- the present time there is evidence (Ir anomaly, wards a catastrophic event such as a large ex- shocked quartz) for only one impact event at traterrestrial impact (Alvarez et al., 1980), but the K/T boundary. An immediate but short large scale volcanism cannot be excluded lasting surface water warming seems to be as- (McLean, 1985; Officer and Drake, 1985). socited with the K/T boundary event at E1 Kef El Kef data show that the ecosystem did not (Lindinger and Keller, in press) which lends recover for nearly 300,000 years after the K/T support to the theory of a greenhouse effect, but boundary event (Subzone Plb, FA G. pseudo- does not explain the long recovery period. bulloides) and full recovery was not reached un- Were there repeated shocks to the ecosystem til about 1.5 m.y. after the boundary event before and after the K/T boundary event as for (Zachos and Arthur, 1986; Arthur et a1,1987; instance from a major episode of volcanism or Gerstel et al., 1987). During this time faunal multiple impacts? El Kef data of pre-K/T assemblages remain simple, primitive and of boundary extinctions and long period of de- very low diversity. Generally low oxygen con- pressed environmental conditions suggests that ditions prevailed on the ocean floor (Peypou- multiple causes were involved, but there is no quet et al,. 1986; Keller, 1988a). Stable oxygen evidence of multiple impacts or major volca- and carbon isotopic data indicate highly stressed nism. The present study suggests a multi-cause environmental conditions with low surface pro- scenario including pre-K/T boundary extinc- ductivity, generally cool fluctuating tempera- tions possibly related to production of warm sa- tures and a very low surface to bottom line bottom waters and its geochemical temperature gradient (Keller and Lindinger, in consequences, a geologically instantaneous ca- prep. ). The post-K/T boundary recovery starts tastrophy at the K/T boundary possibly caused in Subzone Plbl just before the first appear- by a bolide impact followed by prolonged envi- ance of Globorotalia pseudobulloides, dated at ronmental stress perhaps due to a disrupted cli- 66.35 Ma. The recovery is relatively rapid with matic regime and CO2 fluctuations. Further an increase in carbonate sedimentation to pre- detailed geochemical and faunal studies of other K/T levels and a major increase in surface and K/T boundary sections with continuous sedi- bottom water productivity as indicated by a po- mentation, high sedimentation rates and good 262 faunal preservation will be necessary to test this Haq, B.U, 1983. Jurassic to Recent nannofossilBiochron- scenario, ology: an update. In: B.U. Haq (Editor), Nannofossil Biostratigraphy, Benchmark Papers in Geology. Hutchinson and Ross, Stroudsburg, Pa., 78: pp. 358- Acknowledgements 378. Herm, D., Hillebrandt, A.V. and Perch-Nielsen, K., 1981. I am grateful to Drs. K. Perch-Nielsen and T. Die Kreide/Terti~ir-Grenze im Lattengebirge (NSrd- liche Kalkalpen) in mikropal~iontologischer Sicht. Geol. Romein for providing samples from E1 Kef and Bay., 82: 319-344. to Drs. J. Salaj, R. Thunell and H. Thierstein Hofker,J., Sr., 1978. Analysis of a large succession of sam- for helpful critical reviews and comments of this pies through the Upper Maastrichtian and Lower Ter- manuscript, tiary of drillhole 47.2, Shatsky Rise, Pacific, Deep Sea Drilling Project, J. Foraminiferal Res., 8: 46-75. Hsu, K.J. and McKenzie, J.A., 198. A "Strangelove" ocean References in the earliest Tertiary. In: E.T. Sunquist and W.S. Broecker (Editors), The Carbon Cycle and Atmos- pheric CO2: Natural Variations Archean to Present. Am. Arthur, M.A., Zachos, J.C. and Jones, D.S., 1987. Primary Geophys. Union Monogr., 32: 487-492. productivity and the Cretaceous/Tertiary boundary Izett, G.A., 1987. Anthigenic "spherules" in K-T boundary event in the oceans. Cretaceous Res., 8: 43-54. sediments at Caravaca, Spain and Ratou Basin, Colo- Berggren, W.A., 1965. Paleocene -- a micropaleontologist's rado and New Mexico, may not be impact derived. Geol. point of view. Bull. Am. Assoc. Pet. Geol., 44: 1473-84. Soc. Am. Bull. 99: 78-86. Berggren, W.A., 1978. Recent advances in Cenozoic plank- Jiang, M.J. and Gartner, S., 1986. Calcareous nannofossil tonic foraminiferal biostratigraphy, biochronology and succession across the Cretaceous/Tertiary boundary at biogeography: Atlantic Ocean. Micropaleontology, 24: Braggs, Alabama. Geology, 15:311-315. 337-370. Jones, D.S. Mueller, P.A., Bryan, J.R., Dobson, J.P., Chan- Berggren, W.A., Kent, D.V. and Flynn, J.J., 1985. Jurassic nell, J.E.T., Zachos, J.C. and Arthur, M.A., 1987. Biotic, to Paleogene: Part 2, Paleogene geochronology and geochemical andpaleomagneticchangesacrosstheCre- chronostratigraphy. In: N.J. Snelling (Editor), The taceous/Tertiary boundary at Braggs, Alabama. Geol- Chronology of the Geologic Record. Geol. Soc. Lond. ogy, 15: 311-315. Mem., I0: 141-195. Kauffman, E., 1984. The fabric of Cretaceous marine ex- tinctions. In: W.A. Berggren and J.A. Van Couvering Blow, W.H., 1979. The Cainozoic Globigerinidae. Brill, (Editors), Catastrophes and Earth History. Princeton Leiden, 1413 pp. (3 vols.). Univ. Press, Princeton, pp. 151-237. Boersma, A. and Shackleton, N.J., 1981. Oxygen and car- Keller, G., 1985. Depth stratification of planktonic fora- bon isotope variations and planktonic foraminiferal minifers in the Miocene Ocean. Geol. Soc. Am. Mem., depth habitats. Late Cretaceous to Paleocene, Central 163: 177-194. Pacific, DSDP Sites 463 and 465, Leg 65. In: Initial Re- Keller, G., 1988a. Biotic turnover in benthic foraminifera ports of the Deep Sea Drilling Project, 62. U.S. Govern- across the Cretaceous/Tertiary boundary at El Kef, ment Printing Office, Washington, D.C., pp. 513-526. Tunisia. Palaeogeog. Palaeoclimatol., Pataeoecol., 66: Brass, G.W., Southam, J.R. and Peterson, W.H., 1982. 153-171. Warm saline bottom water in the ancient ocean. Nature, Keller, 1988b. The K/T boundary mass extinctions in the 296: 620-623. western Interior Seaway (Brazos River, Texas). GSA Burollet, P.F., 1966. Contribution ~ l'~tude stratigraphique Annual Meeting, Abstr. with Programs, 1988. de la Tunisie centrale. Am. Mines Geol., Tunis, 18: 1- Keller, G. and Lindinger, M., in prep. Stable isotope, TOC 352. and CaCO~ record across the Cretaceous/Tertiary D'Hondt, S. and Keller, G., 1987. Patterns of evolution in boundary at E1 Kef, Tunesia. Palaeogeogr., Palaeocli- some earliest Paleocene lineages of planktic foramini- matol., Palaeoecol., in prep. fera. In: Geol. Soc. Am. Annu. Meet., Abstr. with Pro- Kuslys, M. and Krahenbuhl, U., 1983. Noble metals in Cre- grams, 19 ( 7 ): 639. taceous/Tertiary sedimens from El Kef. Radiochim. Act, Gerstel, J., Thunell, R.C., Zachos, J.C. and Arthur, M.A., 34: 139-141. 1986. The Cretaceous/Tertiary boundary event in the McLean, D.M., 1985. Mantle degassing induced dead ocean North Pacific: Planktonic foraminiferal results from in the Cretaceous-Tertiary transition. In: The Carbon Deep Sea Drilling Project Site 577, Shatsky Rise. Pa- cycle and Atmospheric CO~; Natural variation Archean leoceanography, 1 (2): 97-17. to Present. Geophys. Monogr., 32. Am. Geophys. Union, Gerstel, J., Thunell, R. and Ehrlich, R., 1987. Danian faunal Washington, D.C., pp. 493-503. succession: Planktonic foraminiferal response to a Officer, C.B. and Drake, C.L., 1985. Terminal Cretaceous changing marine environment. Geology, 15: 665-668. environmental events. Science, 227:1161-1167. 263 Perch-Nielsen, K., 1979. Calcareous nannofossils at the K/ topes. Proc. K. Ned. Akad Wet, Ser. B, 83 (3): 295-313. T boundary in Tunisia. In: W.K. Christensen and R.G. Salaj, J., 1973. Proposition pour des hypostratotypes du Bromley {Editors), K/T Boundary Events. Univ. Co- Danian et du PaldocSne (localitd type de la formation penhagen, 1979, 1: 238-243. El Haria, Le Kef, Tunisie septentrionale ). Notes Serv. Perch-Nielsen, K., 1981a. Nouvelle observations sur les G~ol. Tun., 59-66. nannofossiles calcaires ~ la limite Cr~tac~/Tertiaire pros Salaj, J., 1977. The type sections of the Cretaceous and Pa- de E1 Kef, Tunisie. Cah. Micropal$ontol., 3: 25-36. leocene of Tunisia proposed for the hypostratotypes and Perch-Nielsen, K., 1981b. Les coccolithes du Pal~oc~ne pros their application for correlation with planktonic zones de E1 Kef, Tunisie et leurs anc~tres Cah. Micropal~on- of Libya. Abstr. 2nd Symp. Geol. Libya, Tripoli, p. 21. tol., 3: 7-23. Salaj, J., 1980. Microbiostratigraphie du Cr~tac~ et du Pa° Perch-Nielsen, K., McKenzie, J. and Q. He, 1982. Biostra- l~og~ne de la Tunisie septentrionale et orientale (Hy- tigraphy and isotope stratigraphy and the "cata- postratotypestunisiens).Guds, Bratislava, pp. 1-238. strophic" extinctions of calcareous nannoplankton at the Smit, J., 1982. Extinction and Evolution of planktonic for- Cretaceous/Tertiary boundary. Geol. Soc. Am. Spec. aminfera after a major impact at the Cretaceous/Ter- Pap., 190: 353-371. tiary boundary. Geol. Soc. Am. Spec. Pap., 190: 329- Peypouquet, J.P., Grousset, F. and Mourgniart, P., 1986. 352. Paleoceanography of the Mesogean Sea based on ostra- cods of the northern Tunisian continental shelf between Smit, J. and Romein, A.J.T., 1985. A sequence of events the Late Cretaceous and Early Paleogene. Geol. across theCretaceous-Tertiaryboundary. Earth Planet. Rundsch., 75(1 ): 159-174. Sci. Lett., 74: 155-170. Premoli Silva, I. and Bolli, H.M., 1973. Late Cretaceous to Thierstein, H.R, 1982. Terminal Cretaceous plankton ex- Eocene planktonic foraminifera and stratigraphy of Leg tinctions: a critical assessment. Geol.Soc. Am. Spec. 15 sites in the Caribbean Sea. In: Initial Reports of the Pap., 190: 385-399. Deep Sea Drilling Project, 15. U.S. Government Print- Zachos, J.C. and Arthur, M.A., 1986. Paleoceanography of ing Office, Washington, D.C., pp. 499-547. the Cretaceous-Tertiary boundary event: Inferences Romein, A.J.T. and Smit, J., 1981. The Cretaceous/Ter- from stable isotopic and other data. Paleocanography, tiary boundary: calcareous nannofossils and stable iso- 1: 5-26.