<<

The (Late ) in the Northern Hemisphere

LENA B. GOLOVNEVA Komarov Botanical Institute of the Russian Academy of Sciences, 2 Popov str., St Petersburg 197376, Russia (e-mail: [email protected])

Abstract: This investigation of the Maastrichtian climate in the Northern Hemisphere is based on taxonomic and ecological studies of floras, leaf physiognomy and the distribution of dinosaurian . Fossil evidence indicates that the climate during Maastrichtian time was warm temperate at high and middle , and subtropical south of 40°N. Precipitation was relatively high (about 700-800 mm) and evenly distributed over the . The annual range of temperatures was similar to that of modern maritime , but the latitudinal gradient was lower than at present. At high latitudes cold-month mean temperatures were about 3-4°C and probably never dropped below 0°C for extended periods. It seems that these comparatively mild winter temperatures in polar regions were a result of the heating of these areas by warm oceanic .

The Maastrichtian is the youngest stage of tian) and the Kakanaut (Middle Maastrichtian) the Cretaceous Period, immediately preceding Formations (Golovneva 1994, 1995). The Maas- the Cretaceous- boundary. At that time trichtian flora of the Avgustovka from there was a great turnover in the Earth's biota was studied by Krassilov (1979). This and the development of terrestrial ecosystems assemblage is from the upper part of the was significantly affected by changes in climate. Krasnoyarka Formation and includes 16 spe- Thus, the reconstruction of the Maastrichtian cies. In Kazakhstan, Maastrichtian floras have palaeoclimate is of particular importance for the been found in the lower part of the Severozaysan evaluation of climatic change near this bound- Formation (Zaysan Depression), at the localities ary. To determine the climate, the taxonomic of Zhuvankara and Tajzhusgen (Shilin & and ecological characters of fossil land-plant Romanova 1978; Akhmetiev & Shevyreva assemblages and faunas were analysed. 1989), and at Ulken-Kalkan in the Ili River The conclusions presented in this paper are Basin (Makulbekov 1974). The Ulken-Kalkan based largely on the study of 12 fossil floral flora is rather poor and contains only nine assemblages from northeastern Asia and the . It is considered to be of Maastrichtian western part of from 35°N to ; this age is suggested by the presence of 70°N (Fig. 1). These floral assemblages were dinosaurian bones in the underlying conglomer- chosen because they are accurately dated and ates although the flora could equally be Early well described. In there are no rich in age. In Mongolia the Maastrichtian Maastrichtian macrofloral assemblages that can flora comes from deposits of the Nemegt be used for investigation of palaeoclimates. The Formation, which also contains a rich floras from northeast Russia and Kazakhstan . The Nemegt flora includes about 10 were studied on the basis of collections in the species although it has been studied only in a Komarov Botanical Institute, and other floras preliminary way (Krassilov & Makulbekov were analysed using published data. As a rule, 1995; Makulbekov 1995). Kogosukruk is the literature-based climatic inferences do not sig- northernmost Maastrichtian locality (70°N) in nificantly differ from evaluations based on (Spicer & Parrish 1987). It is dated to original collections. Late -Early Maastrichtian time by The northeastern Asia Maastrichtian floras marine and includes only five or six are known from the Koryak Upland, the island species. Numerous Early and Late Maastrich- of Sakhalin, Central Asia and Mongolia. In the tian leaf assemblages are known in the Western Koryak Upland, rich floral assemblages come Interior of North America from the deposits of from deposits of the Rarytkin (Late Maastrich- the Edmonton, Brazeau, Hell Creek, Lance,

From: HART, M. B. (ed.) Climates: Past and Present. Geological Society, London, Special Publications, 181, 43 54, 1-86239-075-4/$15.00 44 L. B. GOLOVNEVA

4O 'I II0 o /lO" 46o ° 14;0" 120 ~' II0+ tO <'

Inf .

,;2L .-

0" ~ 1£0* 160 ° f20* toe ~0"

• macrofloralremains C) dinosaur remains

Fig. 1. Distribution of the Maastrichtian floral and dinosaur remains in the Northern Hemisphere (continuous line indicates the approximate geographical extent of the sea and land masses, after Gorodnitskiy et al. (1978).

Medicine Bow, Denver, Raton and Vermejo analytical procedure called CLAMP (climate leaf Formations ( 1935; Dorf 1942a,b; Bell analysis multivariate program; Wolfe 1993). 1949; Shoemaker 1966; Rouse, 1967; Johnson This program uses 31 leaf morphological char- 1992). Many localities contain rich and well- acters to obtain eight climate variables: mean described floral remains. The flora of the Ripley annual temperature, cold-month mean tempera- Formation (Early Maastrichtian) at 35°N is the ture, warm-month mean temperature, mean southeastern-most occurrence considered (Berry annual precipitation, mean month growing 1925). season precipitation, mean growing season pre- Several characters of the fossil floras were cipitation, precipitation during the three con- examined, including their general diversity, the secutive driest months and length of the growing relative abundance of different taxonomic and season. The application of this method has been ecological groups, and the ratio between ever- described in detail elsewhere (Wolfe 1993; Her- green and deciduous (Table 1). man & Spicer 1996; Kovach & Spicer 1996). Palaeoclimatic analysis of the structure of plant Only floras that included more than 15-20 assemblages and putative tolerances of living species of woody dicotyledonous species were relatives has produced only a generalized inter- scored for CLAMP analysis. pretation of the Maastrichtian climate. Mor- For a more accurate evaluation of the phological characters of angiosperm leaves Maastrichtian climate and verification of the correlate especially well with environmental CLAMP data, the distribution of the vertebrate factors (Richards 1952; Wolfe 1979). Physiog- fauna, primarily , was analysed and nomic analysis of leaves of dicotyledonous oxygen-isotope data were surveyed. These data woody can provide not only general were compiled from various publications (e.g. inferences on climate but also rather precise Teis & Nidin 1973; Savin 1977; Yasamanov quantitative estimations of climatic factors 1978; Jerzykiewicz & Russell 1991; Weishampel (Bailey & Sinnot 1915; Dilcher 1973; Wolfe 1991; Nessov 1995). By comparing evidence 1979; Wolfe & Upchurch 1987). In this study I from independent sources, more precise infer- applied a recently developed physiognomic ences of the climate can be produced. Table 1. Taxonomic and ecological diversity of the Maastrichtian floral assemblages

Number of species Morphological characters of woody angiosperm leaves

Entire- margined groups karl- Woody- Aquatic- leaves Leaf size (Wolfe. 1993) Reference Fossil tude- Gene- angio- angio- Coni- Cycado- Ever- Deci- used floras (°N) ral sperms sperms fers phytes n % microl micro2 micro3 mesol meso2 meso3 green duous for calculations

> Kogosukruk 70 10 1 1 1 - 2 - 100 - 1 Spicer & Parrish (1997) ..] Rarytkin 63 68 30 7 I0 - 5 8 30 3 14 24 18 5 9 Golovneva (1994) Kakanaut 63 43 23 6 7 2 1 3 15 9 34 37 9 4 6 Golovneva (1994) ~Z Edmonton 53 35 17 8 5 1 2 5 32 - 24 47 22 4 3 Bell (1949) --] Sakhalin 49 17 5 2 3 3 4 2 40 - 6 36 33 2 4 Krassilov (1979) Zaisan 48 22 11 1 8 1 2 22 12 38 21 23 5 3 Shilin & Romanowl (1978) Z Hell Creek 46 33 30 2 ll 31 I II 28 31 - 3 Shoemaker (1996) O Ulken-Kalkan 44 9 5 1 1 - -- 50 50 - 1 Makulbekov (1974) Nemegt 43 7 2 3 2 - 100 I 2 Makulbekov (1995) Lance 43 50 29 7 2 4 15 53 3 44 36 21 2 1 Dorf (1942h) 7~ Medicine Bow 40 39 31 2 2 - 3 20 66 - 19 61 13 2 Dorf (1942a) ~Z Ripley 35 54 36 2 6 - 5 22 63 16 36 22 10 6 Berry (1925) 46 L. B. GOLOVNEVA

Maastrichtian floral assemblages represented by both evergreen and deciduous taxa. The number of species generally The Maastrichtian floras from various parts of decreases from north to south. This is accom- the Northern Hemisphere are characterized by panied by an increased percentage of evergreen widely varying species composition, despite the . In the northern regions deciduous general similarity of the gymnosperms. The Taxodiaceae dominated the conifers. In north- taxonomic composition of the angiosperms eastern Asia they prevailed in floras up to differs considerably on both generic and specific approximately 45°N, including those at the levels. During Maastrichtian time, the floras of Zaysan and Ulken-Kalkan localities in eastern each area seemed to have developed indepen- Central Asia. Evergreen appeared dently and belonged to different phytogeogra- only in southern Gobi (Nemegt Fm.). In North phical units (Golovneva 1994). America the distribution of deciduous Taxodia- In all floras, except the Alaskan, angiosperm ceae extended as far south as . In more taxa prevail over ferns, conifers and cycado- southern areas conifers were not abundant and phytes. The relative abundance of ferns is low in were represented mainly by evergreen Taxodia- comparison with other floras. In many ceae, and Araucariaceae. The Cretaceous floras ferns amounted to 30% of the distribution of conifers in the Northern Hemi- total diversity and apparently formed non-forest sphere can, therefore, be interpreted as an communities on swampy or poor, disturbed indicator of a rise in temperature and a decrease soils. Increased abundance of ferns in floras is in rainfall from north to south. usually correlated with epochs of climatic Most Maastrichtian angiosperms are repre- warming (Samylina 1974). During Maastrich- sented by extinct taxa. This impedes use of tian time the reduced number of species and nearest living relatives for climatic determina- their lower relative abundance (Table 1) were the tions. Leaves of angiosperms in the Maastrich- consequence of the evolutionary replacement of tian floras of the Northern Hemisphere are on fern communities by angiosperms, rather than a average of moderate size and gradually became result of climatic cooling. However, the rare smaller southwards. This may be related to more occurrence of ferns along with a small number of favourable moist conditions at high-mid-lati- thin layers may indicate an insufficiently tudes. At the taxonomic level this reduction is moist climate. reflected by the slow disappearance of plata- The remains of cycadophytes and in noids and trochodendroids. The range of leaf the Maastrichtian deposits occur only at high sizes and scarcity of drip-tips indicate relatively latitudes, where they are rather abundant in moderate precipitation (Wolfe 1990). In north- some localities. These plants had already become ern regions leaves of angiosperms are character- extinct by time in mid-low ized mainly by their dentate margins and thin latitudes. On the whole, extant cycadophytes are texture. Lobed and simple pinnately veined thermophilic plants. Throughout Cretaceous leaves with entire margins and thick texture time they were typical elements of tropical and become more abundant to the south. These data subtropical floras. In the Cretaceous temperate indicate warmer and drier climates in southern floras an increased ratio of cycadophytes to areas by comparison with more northern ones other taxa has always been considered to (Wolfe 1979; Wolfe & Upchurch 1987). The indicate climatic warming. Krassilov (1985) percentage of entire-margined species varies proposed a cycadophyte index to evaluate from 15% at 63°N to 65% at 40°N. Accord- climatic changes during Early and mid-Cretac- ingly, leaf margin analysis (Wolfe 1979) allows eous time. By Late Cretaceous time cycadophyte us to conclude that the latitudinal range of mean diversity had gradually decreased at all latitudes annual temperatures ranged from 5°C to 20°C. and, strange as it may seem, northern temperate Thus, the Maastrichtian climate can be regions became one of their last refuges. evaluated as generally warm temperate in north- Vakhrameev (1991) ascribed this phenomenon ern regions and subtropical at mid-low lati- to a milder and more humid climate in these tudes, with moderate humidity decreasing regions. Thus, for Maastrichtian time the slightly southwards. latitudinal distribution of cycadophytes cannot be the climatic indicator, but within northern CLAMP palaeoclimatic inferences territories their relative abundance can be used to evaluate climates of different Cretaceous ages The results of the CLAMP analysis are in (Herman & Lebedev 1991; Golovneva 1994). accordance with evidence derived from the Conifers in the high- Maastrichtian general analysis of fossil floral assemblages floras of the Northern Hemisphere are usually (Table 2). CLAMP predictions suggest that the Table 2. Climate data jor Maastrichtian time obtained fi'om CLAMP analyses

Cold- Warm Mean Mean Three Mean month month Mean monthly growing consecutive Length of the annual mean mean annual growing season season driest months growing Fossil Latitue temperature temperature temperature precipation precipation precipation precipation season floras (°N) (°C) (°C) (°C) (mm) (mm) (mm) (ram) (months)

Rarytkin 63 11 4 19 722 129 1410 305 6.9 Kakanaut 63 10 3 19 414 98 948 181 6.3 Edmonton 53 12 5 19 804 141 1586 335 7.1 Sakhalin 49 14 8 20 892 136 1214 293 8 Zaisan 48 11 4 19 515 106 1065 230 7 Hell Creek 46 12 6 19 899 131 1094 302 7 Lance 43 14 8 21 574 ll9 1216 242 8 Medicine Bow 40 17 13 23 933 172 1946 348 9.7 Ripley 35 17 11 23 498 ll5 1126 204 9.2 SD 1.8 3.3 3.1 430 23 280 70 l.l 48 L. B. GOLOVNEVA

climate in Maastrichtian time was warm tempe- macrofloral localities are Kakanaut from the rate (sensu Wolfe 1979, mean annual tempera- Koryak Upland and Kogosukruk in Alaska. ture (MAT) about 10-14°C) at high and mid- The Maastrichtian position of the North Pole is latitudes and subtropical south of about 40°N assumed to be significantly nearer the Pacific (MAT about 15-17°C). At high and mid- than it is today, and not far from the latitudes cold-month mean temperatures were coast of northeastern (Smith & Briden apparently about 3-8°C, and warm-month mean 1978; Howarth 1981). Accordingly, these local- temperatures about 19-20°C. To the south of ities were farther north during Maastrichtian 40°N, winter temperatures increased to 10-13°C time and well within the Circle. it follows and summer temperatures to 20-23°C. The from different continental reconstructions that estimated annual range of temperatures is about Kakanaut was at about 75-80°N and Kogosuk- 12-16°C and, therefore, similar to that of ruk at about 75-85°N. Of these two floras the modern maritime climates. This is consistent Kakanaut flora is more diverse, and includes with the proximity of many of the fossil localities more than 40 species. This number of species is to sea coasts during Maastrichtian time, sufficient for valid estimation of climate by although today they are situated in the con- CLAMP. Mean annual temperature is determined tinental interior (Fig. 1). Predicted precipitation as 10°C and cold-month mean temperature is relatively high in all observed areas. It about 3°C. The Kogosukruk locality contains averaged about 1500-1700mm and was fairly a poorer floral assemblage, which is character- evenly distributed over the year. Periods of ized by strong dominance of deciduous conifers drought are not detectable by CLAMP, but the and rare occurrence of angiosperms, represented three driest months were slightly less moist than only by 3-4 species. The depauperation of this the mean monthly precipitation. It seems that flora by comparison with other Cretaceous the drier period corresponded to winter, because Alaskan floras was previously interpreted as a the mean monthly precipitation of the growing dramatic drop in diversity during Maastrichtian season was higher than that of the drier period. time. The mean annual temperature for this area On the whole, the Maastrichtian climate of was estimated to be about 5°C (Parrish et al. northern latitudes was characterized by low 1987; Spicer & Parrish 1987), implying a seasonal variation in precipitation and moderate frequent occurrence of winter temperatures seasonality in temperature. Above the Arctic below 0°C. However, assemblages from Circle, the light regime was an additional the North Slope of Alaska (Frederiksen 1991) seasonal factor that controlled the structure of and from other localities in Alaska and Arctic vegetation. However, the influence of this factor Canada (Tabbert 1967; Rouse & Srivastava is uncertain. According to the relationship 1972, Felix & Burbridge 1973) usually involve between temperature and precipitation, the more than 30 species of angiosperms, repre- Maastrichtian climate is close to the modern sented by approximately 15 genera. In taxo- warm temperate and subtropical climates of the nomic composition and diversity these eastern coast of North America from Maryland assemblages are very similar to pollen assem- to northern Florida. blages from localities further south, in and Saskatchewan, and also to assemblages of Discussion the Koryak Upland and Yenisei depression in northern Siberia (Samoylovich 1977). Palyno- Temperature floras of these regions are characterized by the dominance of Aquilapollenites pollen and by the CLAMP temperature predictions are in good presence of other characteristic genera from the agreement with general climate inferences and Triprojectacites (Samoylovich 1977; - data based on leaf margin percentages. The ten 1984). Close resemblance of spore and pollen evaluation of temperatures in polar areas is of assemblages implies that macrofloras were also particular importance because the Maastrichtian similar in composition and diversity. Hence, the climate of these regions differed significantly depauperation of the Maastrichtian Kogosuk- from modern conditions. Palaeobotanical evi- ruk flora may not reflect a change in climate. dence suggests that the winters were mild. Perhaps, it could be attributable to a tapho- Sometimes temperatures could drop to 0°C at nomic factor instead. This leads to the conclu- mid-latitudes and even below 0°C at high sion that the climate in polar areas during latitudes, but the magnitude of minimum winter Maastrichtian time was similar to that indicated temperatures and duration of frost periods are by the flora of the Kakanaut locality, i.e. cold- uncertain. month mean temperatures were not lower than The most northerly known Maastrichtian 3-4°C. This assumption is confirmed by the MAASTRICHTIAN CLIMATE IN THE NORTHERN HEMISPHERE 49

Table 3. Diversity of vertebrate faunas of the North Pacific region during Maastrichtian time (on the basis of data from Spicer & Parrish (1987), Nessov & Golovneva (1990), Wheishampel (1991) and Nessov (1997)

Area or locality Latitude Number of Main vertebrate groups (°N) dinosaur genera

Alaska (Kogosukruk) 70 5 , dinosaurs, Yukon Territory, Canada 66 1 dinosaurs Northwest Territories, 64 2 dinosaurs Canada Koryak Upland (Kakanaut) 63 2 , plesiosaurs, dinosaurs Amur region 49-50 4 dinosaurs, , Kazakhstan (Zaysan) 48 8 shells Sakhalin 47 1 dinosaurs Kazakhstan (Ulken-Kalkan) 44 5 dinosaurs Southern Gobi (Nemegt) 43 13 crocodiles, turtles, dinosaurs Southwestern Canada and 40-53 ~20 fish, , turtles, crocodiles, northwestern USA , dinosaurs, mammals Southern USA < 40 ~26 fish, amphibians, turtles, crocodiles, lizards, dinosaurs, mammals

presence of taxa from the subtropical families Yukon Territories of Canada (Weishampel Loranthaceae, Nyssaceae, and Ola- 1991), as well as from the Koryak Upland of caceae in pollen assemblages at high latitudes northeast Russia (Nessov & Golovneva 1990). (Samoylovich 1977). The disappearance of sub- The diversity of the dinosaur fauna decreases tropical thermophilic elements is observed only northwards, and is presumed to be linked not in the Khatanga-Lena palynological province in only to diminishing temperatures, but also to a northeastern Siberia. It is not yet clear if this decreasing number of localities at high latitudes. depauperation of palynofloras is associated with However, a drop in diversity to the north of c. proximity to the North Pole or with more severe 55-60°N (from 5-10 genera to 1-3 genera) can winters in the Siberian continental interior. be probably related to the former position of the Comparison of the predicted Maastrichtian Arctic Circle and long dark polar nights. polar climate with the database developed by Taxonomically, only one biogeographical bor- Wolfe (1993) shows that climatic characteristics der in the distribution of dinosaurs in North of the Kakanaut locality are closest to those of America has been identified. This is marked by several sites in Maryland, including Battle Creek the appearance of sauropods south of 35~0°N Cypress Swamp State Park. Forests consisting of (Lehman 1987). In Maastrichtian time, sauro- deciduous Taxodiaceae, which were similar to pods were also distributed in tropical regions of the modern genera and , , and southern Europe were very common plant communities of high (Weishampel 1991). This boundary corresponds latitudes in Maastrichtian time, and sometimes to that between warm temperate and subtropical they were dominant in the vegetation. Thus, the climates, which is inferred from floral data, and predicted polar climatic values do not contradict also corresponds to the boundary between the the putative temperature tolerance of the Maas- Aquilapollenites and Normapolles palynological trichtian plants. provinces (Samoylovich 1977; Batten 1981, However, the discovery of dinosaurian re- 1984). The fauna of mid- and high latitudes is mains in polar regions was not expected, because taxonomically very similar, but high-latitude dinosaurs have been previously regarded as assemblages were more depauperate. The de- animals adapted to warm climates. In the North creased diversity reflects, in particular, the Pacific region dinosaurian and other vertebrate absence (or rarity) of amphibians and non- faunas are associated with most of the same dinosaurian , which were abundant and localities as the floral remains (Table 3). taxonomically diverse further south (Clemens & Maastrichtian high-latitude dinosaurs have been Nelms 1993). Among the large reptiles that lived reported from the North Slope of Alaska (Spicer within the Arctic Circle during Maastrichtian & Parrish 1987), the Northwest Territories and time were plesiosaurs and dinosaurs represented 50 L.B. GOLOVNEVA by the herbivorous hadrosaurs and ceratopsians, capacity to hibernate (Clemens & Nelms 1993). as well as by carnivorous carnosaurs and small In this case the occurrence of dinosaurs cannot terapods. This rich fauna of reptiles could not be used as a climatic indicator because their survive long severe winters. Small modern temperature tolerance is uncertain. It is most reptiles of the temperate zone can hibernate in likely that they were not fully endothermic and shelters during the winter, but large dinosaurs used some other physiological mechanisms (e.g. could not do the same because of their size gigantothermy) to maintain their body tempera- (Parrish et al. 1987). Hence, polar dinosaurs ture (Paladino & Spotila 1994). In any case, the were probably active throughout the winter. presence of reptiles with naked skin indicates Most recent reptiles need average winter tem- that winter temperatures in polar regions were peratures to be not less than 10-12°C to sustain generally above freezing and never dropped the necessary level of activity, although they can below 0°C for long periods. endure slight frosts for short periods (McKenna The summarized data obtained from oxygen- 1980; Hutchison 1982). If dinosaurs were simple isotope analyses indicate that mean annual ectothermic animals and had temperature toler- temperatures were about 18-22°C (Douglas & ances similar to those of other reptiles, we Savin 1973; Saito & Van Donk 1974) or 20-25°C should assume that mean winter temperatures (Frakes et al. 1994) at tropical latitudes; about during Maastrichtian time were also about 10- 11-16°C in mid-latitudes (Teis & Nidin 1973; 12°C or above. These estimates of cold period Yasamanov 1978; Frakes et al. 1994); and about temperatures are much higher than the estima- 10°C at high latitudes (Teis & Nidin 1973; tions inferred from plant physiognomy. Wolfe Boersma 1984; Frakes et al. 1994). The Arctic (1987) suggested that the deciduousness of high- Ocean is generally believed to have lacked a latitude forests resulted from the polar light permanent ice cap in Maastrichtian time (Clark regime rather than from lower in temperature. 1974). He estimated mean annual temperatures of Evidence coming from oxygen-isotope data polar regions to be about 10-15°C. However, for high and mid-latitudes is therefore in good the gradual change in abundance of species with agreement with climate predictions derived from entire-margined leaves along a latitudinal gra- plant physiognomy. dient, which is not accompanied by an abrupt change close to the former Arctic Circle, and the Precipitation presence of a relatively large numbers of these species (20-30%) at high latitudes, indicates that CLAMP predictions for precipitation in Maas- deciduousness was determined primarily by trichtian time are high (about 1500-1700mm) seasonality in temperature, rather than by the and indicate an even distribution throughout the light regime. It seems that some entire-margined year. No latitudinal differences in rainfall were species could have been evergreen and survived revealed by CLAMP. This amount of precipitation the polar winter in a dormant state. corresponds to a humid climate and closed- Some investigators have supposed that polar canopy forest vegetation. However, general dinosaurs, like caribou today, could have physiognomic analysis of fossil floral assem- migrated southwards to escape cold winters blages, scarcity of drip-tips and range of leaf size (Hotton 1980). However, because of the lower suggest only moderate humidity that decreased thermal gradient in Maastrichtian time by slightly southwards. Coal layers are insignificant comparison with the present, the dinosaurs (usually several centimetres) in the Maastrich- would have had to cover a vast distance of tian deposits at high latitudes and almost 2500-4000km to reach significantly warmer disappear in the mid-latitudes. The existence of environments. The predominance of juvenile large dinosaurs is consistent with open vegeta- individuals in dinosaurian assemblages contra- tion and, therefore, semi-arid conditions (Wing dicts this concept of southward migration & Tiffney 1987). In its turn, disturbance of (Nessov & Golovneva 1990; Clemens & Nelms vegetation by large must have fa- 1993). It is highly improbable that hatchlings voured the formation of open landscapes. It is and young dinosaurs were involved in such long- probable that the precipitation values derived distance annual migrations. fronl CLAMP analysis are somewhat overesti- These conclusions indicate that dinosaurs mated. This overestimation can be caused by were able to remain in polar areas all year taphonomic factors because fossil assemblages round and tolerated a cool, dark winter. It is are primarily formed by plants from river possible that they differed metabolically from valleys, lake shores and other lowland environ- modern reptiles. Smaller turtles and crocodiles ments. In these conditions the magnitude of did not live so far north, although they had the moisture content is always much greater than for MAASTRICHTIAN CLIMATE IN THE NORTHERN HEMISPHERE 51

30

r..) o

10 & ~ present 0

-lO -

I I I I l , I ,,_ N 80° 600 40° 20° 0° 20° Latitude

Fig. 2. Temperature gradients for present (continuous line, after Barron (1983)) and the Maastrichtian stage (dotted line). upland areas. It seems that the vegetation during resulted from heating of polar areas by warm Maastrichtian time would be a mosaic of open oceanic (Nessov & Golovneva 1990; woodland on uplands and closed-canopy forests Nessov 1997). The temperature of the bulk of along river valleys. Wolfe (1990) estimated the deep oceanic waters is now about 2°C. These mean annual precipitation at mid-latitudes to be water masses are formed as the result of the about 700-900 mm, deriving this value from the descent of cold, dense water at high latitudes. estimated precipitation for the driest months. Later they appear near the western margins of This approach seems more appropriate for at low latitudes as upwellings of cold estimating precipitation during Maastrichtian waters containing a large amount of nutrients. time. In tropical and subtropical areas only a thin surface layer of water is warmed. Thus, the Latitudinal temperature gradient modern vertical ocean circulation is determined by cold high-latitude downwellings. On the whole, the Maastrichtian climate was However, during Maastrichian time the pic- characterized by high average temperatures and ture was different. The temperatures of deep high equability both in precipitation and seaso- ocean waters reached 10-14°C (Boersma 1987). nal range of temperature. Oxygen-isotopic and These waters were probably heated at low palaeobotanical data indicate that temperatures latitudes in shallow epicontinental seas. As a were relatively lower in the tropics and signifi- result of surface evaporation the salinity and cantly higher at high latitudes by comparison density of these waters increased, leading to their with conditions today. The latitudinal thermal descent to the bottom and into the deeper parts gradient was less pronounced (Fig. 2). The of the ocean. These warm downwellings trans- difference between mean annual temperatures ferred the heat from the sea surface to the depths of polar and tropical areas during Maastrichtian of the ocean. At present, similar downwellings time was about 10°C. This is 4-5 times less than are observed in the Mediterranean and Red today. Seas, but the influence of this type of down- The greatest differences between the Maas- welling on the heat balance of the is not trichtian and modern climates are observed at great, because the areal extent of these seas is high latitudes. The much higher Arctic tempera- comparatively small. In Cretaceous time, epi- tures during Maastrichtian time may have continental seas in subtropical arid regions were 52 L.B. GOLOVNEVA widespread. As a result, vertical oceanic circula- COUVERING, J. A. (eds). Catastrophes and Earth tion was determined not by cold downwellings at History. The New Uniformitarianism. Princeton high latitudes, but by warm downwellings at low University Press, Princeton, N J, 247-278. latitudes. At high latitudes the warm deep waters CLARK, D. L. 1974. Late Mesozoic and Early cores from the . Geology, 2, to the surface. During Campanian time, 4144 these upwellings are well marked by findings of CLEMENS, W. A. & NELMS, L. G. 1993. Paleoecological hesperornidids, because the distribution of these implications of Alaskan terrestrial vertebrate in the Cretaceous period was related to the fauna in latest Cretaceous time at high paleolati- occurrence of water masses of high productivity tudes. Geology, 21, 503-506. (Nessov & Yarkov 1993). DILCnER, D. L. 1973. The floras of south- Warm upwellings at high latitudes heated eastern North America. IN: GRAHAM, A. (ed.) both the surface waters and the atmosphere. At Vegetation and Vegetational History of Northern that time the ocean could have been an Latin America, Elsevier, Amsterdam, 39-59. DORy, E. 1942a. Stratigraphy and palaeontology of the accumulator of heat, storing it at depth for a Fox Hills and Lower Medicine Bow Formations long time and then redistributing it from the of southern Wyoming and northeastern Color- mid-latitudes to the poles. This model could ado. Carnegie Institution of Washington Publica- explain the low temperature gradients during tion, 580, 1-78. Maastrichtian time and the general equability of -- 1942b. Flora of the in its the climate. type locality, Niobrara County, Wyoming. Car- negie Institution of Washington Publication, 580, 79-159. The research described in this paper was made possible DOUGLAS, J. G, & gAVIN, S. M. 1973. Oxygen and in part by Grant N 96-04-50335 and in part by Grant carbon isotope analyses of the Cretaceous and N 96-05-65843 from the Russian Foundation of Tertiary from the central North Fundamental Research. I would like to thank R. A. Pacific. Initial Reports of the Deep Sea Drilling Spicer for his help in obtaining CLAMP data, and I. A. Project, 17. US Government Printing Office, Mershchikova for her helpful editorial advice. Washington, DC, 591 605. FELIX, CH. J. & BURBRIDGE, P. P. 1973. A Maastrich-

References tian age microflora from Arctic Canada. Geoscience and Man, 7, 31 38. AKHMETIEV,M. A. & SHEVYREVA,N. S. 1989. The flora FRAKES, L. A., PROBST, J. & LUDWIG, W. 1994. of the Tsagaian-type near Zaysan Lake (eastern Latitudinal distribution of pateotemperature on Kazakhstan). Izvestiya Akademii Nauk SSSR, land and sea from to middle Seriya Geologicheskaya, 1989(6), 80-90. [in Rus- . Comptes Rendus de l'Academie des sian]. Sciences, Serie II, 318, 1209-1218. BAILEY, I. W. ~ SINNOT, E. W. 1915. A botanical index FREDERIKSEN, N. 0. 1991. Pollen zonation and of Cretaceous and Tertiary climates. Science, 41, correlation of Maastrichtian marine beds and 831-834. associated strata, Ocean Point dinosaur locality, BARRON, E. J. 1983. A warm, equable Cretaceous: the North Slope, Alaska. US Geological Survey nature of the problem. Earth-Science Reviews, 18, Bulletin, 1990-E, 1-24 305-338. GOLOVNEVA, L. B. 1994. Maastrichtian-Danian floras BATTEN, D. J. 1981. Stratigraphic, palaeogeographic of Koryak Upland. Proceedings of Komarov and evolutionary significance of Late Cretaceous Botanical Institute, 13, 1 148. [in Russian]. and early Tertiary Normapollespollen. Review of -- 1995. Environmental changes and patterns of Palaeobotany and Palynology 35, 125-137. floral during the Cretaceous-Tertiary -- 1984. Palynology, climate, and the develop- transition in Northeastern Asia. Paleontological ment of Late Cretaceous floral provinces in the Journal, 29(2a), 36-49. Northern Hemisphere; a review. Geological Jour- GORODNITSK1Y,A. M., ZONENSHAIN,L. P. & MIRLIN, E. nal, special issue, ll, 127-164. G. 1978. Reconstruction o1" Positions of the BELL, W. A. 1949. Uppermost Cretaceous and Continents during the (on the Basis floras of western Alberta. Canada of Palaeomagnetic and Geological Data). Nauka, Geological Survey Bulletin , 13, 1-23 I. Moscow [in Russian]. BERRY, E. W. 1925. The flora of the . HERMAN, A. B. ~ LEBEDEV, E. L. 1991. Stratigraphy US Geological Survey, Professional Paper, 136, and Flora of the Cretaceous Deposits of North- 1-94. eastern Kamchatka. Nauka, Moscow [in Russian]. -- 1935. A preliminary contribution to the floras ---- ~ SPICER, R. A. 1996 Palaeobotanical evidence of the Whitemud and Ravenscrag Formations. for a warm Cretaceous Arctic Ocean. Nature, 380, Canada Geological Survey Memoirs, 182, 1-65. 330-333. BOERSMA, A. 1984. Campanian through Paleocene HOTTON, N. 1980. An alternative to dinosaur en- paleotemperature and carbon isotope sequence dothermy: the happy wanderers. In: THOMAS, R. and the Cretaceous-Tertiary boundary in the D. K, & OLSON, E. C. (eds) A Cold Look at the Ocean. In: BERGGREN, W. A. ~ VAN Warm-blooded Dinosaurs. American Association MAASTRICHTIAN CLIMATE IN THE NORTHERN HEMISPHERE 53

for the Advancement of Science Selected Sympo- physiology of large living reptiles provide insights sium, 28, 311-350. into the evolution of endothermy and paleophy- HOWARTH, M. K. 1981. Palaeogeography of the siology of extinct dinosaurs? Paleontological Mesozoic. In: Greenwood, P. H. (ed.) The Society Special Publication, 7, 261 274. Evolving Earth. Cambridge University Press, PARRISH, J. M., PARRISH, J. T., HUTCHISON, J. H. & Cambridge. 197-220. SPICER, R. A. 1987. Late Cretaceous vertebrate HUXCHISON, J. H. 1982. , crocodilian and fossils from the North Slope of Alaska and champsosaur diversity changes in the Cenozoic implication for dinosaur . Palaios, 2, of the north-central region of western United 377-389. States. Palaeogeography, Palaeoecology. Palaeo- RICHARDS, P. W. 1952. The Tropical Rain Forest. climatology, 37, 149-164. Cambridge University Press, Cambridge. JERZYKIEWICZ, T. & RUSSELL, D. A. 1991. Late RousE, G. E. 1967. A late Cretaceous plants assem- Mesozoic stratigraphy and of the blage from east-central British Columbia. Cana- Gobi Basin. Cretaceous Research, 12, 345-377. dian Journal of Earth Sciences, 4, 1185-1197. JOHNSON, K. R. 1992. Leaf-fossil evidence for extensive & SRWASTAVA, S. K. 1972. Palynological floral at the Cretaceous Tertiary zonation of Cretaceous and Early Tertiary rocks boundary, North , USA. Cretaceous of the Bonnet Plume formation Northeastern Research, 13, 91-117. Yukon. Canadian Journal of Earth Sciences, 9, KOVACH, W. L. & Spicer, R. A. 1996. Canonical 1163-1179, correspondence analysis of leaf physiognomy: a SAITO, T. & VAN DONK, J. 1974. Oxygen and carbon contribution to the development of a new isotope measurements of Late Cretaceous and palaeoclimatological tool. Palaeoclimates, 2, Early Tertiary Foraminifera. Micropaleontology, 125-138. 20, 152-177. KRASSILOV, V. A. 1979. Cretaceous Flora of Sakhalin. SAMOYLOVJCH, S. P. 1977, New scheme of floristic Nauka, Moscow [in Russian]. differentiation of Northern Hemisphere during -- 1985. The Cretaceous Period. Evolution of the Late Senonian. Palaeontological Journal, 1977(3), Earth's Crust and Biosphere. Nauka, Moscow [in 118-127 [in Russian] Russian] SAMVUNA, V. A. 1974. The Early Cretaceous floras of -- & MAKULBEKOV, N. M. 1995. Maastrichtian the Northeast of USSR (on the problem of aquatic plants from Mongolia. Paleontological development of Cenophyte floras). Komarovskie Journal, 29(2a), 119 140. chtenija, 27, 1-56. [in Russia] LEHMAN, T. M. 1987. Late Maastrichtian paleoenvir- SAVIN, S. M. 1977. The history of the earth's surface onments and dinosaur biogeography in the temperature during the past 100 million . Western Interior of North America. Palaeogeo- Annual Review of Earth and Planetary Sciences, 5, graphy, Palaeoecology, Palaeoclimatology, 60, 319-355. 189-217. SHILIN, P. V. & ROMANOVA, E. V. 1978. The Senonian MAKULBEKOV, N. M. 1974. The Late Cretaceous flora flora of Kazakhstan. Nauka, Alma-Ata [in Rus- of Ulken-Kalkan (the Ilijskaja depression). In: sian] KOJAMKULOVA, B. S. et al. (eds.) The Fauna and SHOEMAKER, R. E. 1966. Fossil leaves of the Hell Creek Flora from Mesocenozoic of South Kazakhstan. and Tullock Formations of Eastern . Materials on the History of Fauna and Flora of Palaeontographica, l19B, 54 75. Kazakhstan, 6, 108 120 [in Russian]. SMn-H, A. G. & BRIDEN, J. C. 1978. Mesozoic and 1995. Florogenesis in central Asia at the Ceno-_oic Paleocontinental Maps. Cambridge Uni- Cretaceous Palaeogene boundary. International versity Press, Cambridge. Cot!ference: Evolution of Ecosystems, Abstracts, SPICER, R. A. & PARRISH, J. T. 1987. Plant megafossils, PIN RAN, Moscow, 57. vertebrate remains, and paleoclimate of the MCKENNA, M. C. 1980. Eocene paleolatitude, climate Kogosukruk Tongue (Late Cretaceous), North and mammals of Ellesmere Island. Palaeogeogra- Slope, Alaska. US Geological Survey Circular, phy, Palaeoecology, Palaeoclimatology, 30, 349- 998, 47-48. 362. TABBERT, R. L. 1967. Cretaceous pollen and spores NESSOV, L. A. 1995. Dinosaurs of Northern Eurasia: from the Ivishak river area arctic Alaska. Review New Data about Assemblages, Ecology and Pa- of Palaeobotany and Palynology, 2, 8-9. laeobiography. St Petersburg University, St Pe- TEIs, R. V. & NIDIN, D. P. 1973. Palaeothermometry tersburg [in Russian]. and O.xTgen-isotopic Composition of Organogenous 1997. Cretaceous Nonmarine Vertebrates of Cm'bonate Rocks. Nauka, Moscow [in Russian]. Northern Eurasia. St Petersburg [in Russian]. VAKHRAMEEV, V. A. 1991. and Cretaceous -- & GOLOVNEVA, L. B. 1990. History of the flora, Floras and Climates of the Earth. Cambridge vertebrates and climate in the late Senonian of the University Press, Cambridge. north-eastern Koryak Uplands. In: Krassilov, V. WHSHAMPEL, D. B. 1991. Dinosaur distribution. In: A. (ed.) Continental Cretaceous o.1 the USSR. WElSHAMPEt, D. B., et al. (eds.) The Dinosauria. Vladivostok, 191-212 [in Russian]. University of California Press, Berkeley, 63-140. -- & YARKOV, A. A. 1993. Hesperornitids in Wmo, S. L. & TWFNEY, B. H. 1987. Interactions of Russia. Russian Ornithological Journal, 2, 37-54. angiosperms and herbivorous . In: Funs, PALADINO, V. F. & SPOTILA, J. S. 1994. Does the E. M. et al. (eds) The Origins of Angiosperms and 54 L.B. GOLOVNEVA

their Biological Consequence. Cambridge Univer- 1993. A method of obtaining climatic para- sity Press, Cambridge. 203-224. meters from leaf assemblages. US Geological WOLVE, J. A. 1979. Temperature parameters of humid Survey Bulletin, 2040, 1-71. to mesic forests of eastern Asia and their relation & UPCI4URCH, G. R. 1987. North American to forests of other regions of the Northern nonmarine climates and vegetation during the Hemisphere and Australasia. US Geological Late Cretaceous. Palaeogeography, Palaeoecol- Survey, Professional Paper, 1106, 1-37. ogy, Palaeoclimatology, 61, 33-77. 1987. Late Cretaceous-Cenozoic history of YASAMANOV, N. A. 1978. Landscape and Climatic deciduousness and terminal Cretaceous event. Conditions during the Jurassic, Cretaceous and Paleobiology, 13, 215-226. Palaeogene in Southern USSR. Nedra, Moscow -- 1990. Palaeobotanocal evidence for a marked [in Russian]. temperature increase following the Cretaceous/ Tertiary boundary. Nature, 343, 153-156.