by Alexei HERMAN, Dr. Sc. (Geol. & Min.), Geological Institute, Russian Academy of Sciences

What natural conditions did dinosaurs live in? Numerous remains of these reptiles were found in the regions with a tropical or subtropical climate and weak seasonality in the Mesozoic (251 - 65 Ma). That is why discovery of remains in the Arctic Regions--Chukotka and Alaska-interested paleontologists and generated a number of hypotheses explaining how this has become possible.

MYSTERIES OF THE CRETACEOUS

In 1961, US geologist Robert Liscomb found a multitude of dinosaur bones in the northern part of Alaska (70 °N) that he originally took for remains of not very ancient Cenozoic* mammals. Only in the mid-1980s an exceptional value of this discovery became clear: the bones turned out to be bones of dinosaurs that inhabited the Arctic Regions long ago, at the end of the Cretaceous**. Further paleontological excavations were carried out in this region by American paleontologists Anthony Fiorillo, Roland Gangloff and their associates who found thousands of bones for several years. These research works gave rise to disputes on the ecology of discovered animals: were they permanent residents of the Arctic Regions or migrated to the south in winter? What did they eat? How and at what speed did they move?

By now, at the site located in the low reaches of the Colville River in the Ocean Point area named the Liscomb Quarry after its discoverer, remains of cartilaginous and bony fish, Hesperornis birds, multituberculate, marsupial and placental mammals, dinosaurs--small and big theropods (carnivors from the Dromaeosaurus, Saurornitholestes, Troodon genera and tyrannosaurids

* The Cenozoic--a period in geological history that started about 65 mln years ago and lasts till now.--Ed.

** The Cretaceous--the latest period of the Mesozoic; began 145 mln years ago, lasted for 80 mln years.--Ed.

Distribution of land and sea basins in the Arctic Regions in the Late Cretaceous (68 Ma, polar projection).

Configuration of the modern shoreline is colored red.

Figures of dinosaurs indicate their habitats: the right figure--in Northern Alaska, the left one--in Chukotka.

close to the Albertosaurus genus), plant-eating pachy-cephalosaurids (Pachycephalosaurus genus), ceratopsids (Pachyrhinosaurus genus) and hadrosaurids or duckbilled dinosaurs (reptiles from the Edmontosaurus genus that reached 10 m in length) were discovered. Bones of the latter were dominant and belonged mostly to young animals. According to the present-day knowledge, layers where bones were found formed in the Campanian (70.6-83.5 Ma) and Maastrichtian (65.5-70.6 Ma) stages in the Late Cretaceous, i.e. about 68-75 mln years ago.

In other sites of Northern Alaska of the same age bones and teeth of dinosaurs, imprints of their skin and footprints were discovered. However, only the Liscomb Quarry can offer such a great variety and abundance of fossil material, even though it is located northwards as compared with other sites (80-82 °N, according to present-day reconstructions).

There is another site rich of dinosaur remains in other parts of the world--in the extreme north-east of Eurasia, in Chukotka. But ancient animals lived there later--in the second half of the Maastrichtian, about 65-68 Ma, and at a lower paleolatitude (70-72 °N). It is located in the eastern part of Koryak Highlands in the Kakanaut River basin and was discovered in 1988 by Russian scientists--zoologist and paleontologist Lev Nessov (Leningrad State University) and paleobotanist Lina Golovneva (later, Dr. Sc. (Biol.), Komarov Botanical Institute RAS, St. Petersburg). Later on, in 2007-2009, Golovneva and her team members together with the Belgian expert in reptiles Pascal Godefroit visited the Kakanaut site once again and found there the scales of the palaeoniscoid fishes, bones, teeth and dinosaur egg shells. The remains of these reptiles belonged to plant-eating dinosaurs (basal omithopods, hadrosaurids, ankylosaurids and neoceratopsids) and carnivorous dinosaurs (troodontids, dromaeosaurids, and tyrannosaurids). According to the scientists, two types of egg shells belong to hadrosaurids and theropods.

Anthony Fiorillo and his colleagues believe: dinosaurs of Northern Alaska lived in the high Arctic latitudes all year round, at a distance less than 1,000 km from the North Pole, and were adapted to life in terms of a long polar night. According to L. Nessov, L. Golovneva and P. Godefroit, ancient reptiles of Chukotka were also permanent residents of the Arctic Regions, and at least some of them reproduced there. In the estimation of these scientists, plenty of remains of young hadrosaurids and small carnivorous dinosaurs, as well as a great dis-

tance to regions where evergreen vegetation existed in the Late Cretaceous, make it possible to assume: the reptiles did not migrate to the south and survived winters being hypoactive. To prove a settled way of life of the animals, scientists refer to the fact that together with remains of adults, there were found teeth of youngsters, including newly hatched ones. Long-term migrations would not leave time to breed the fry and, besides, would make the dinosaurs leave their conventional habitat.

According to L. Nessov, at that time climate of the Arctic Regions was not as severe as today, and mild winters allowed animals to live in high latitudes all year round. In the Cretaceous the so-called Okhotsk-Chukotka Volcanic Belt stretched along the eastern border of the Asia. The scientist assumed that in winter dinosaurs could migrate to the volcanic belt neighboring regions and spend cold seasons there on comfortable warm ground of volcanic calderas or in a hot water of underground springs. In the cold season they could eat leaves of evergreen cycadalean plants, fallen foliage of woody plants and detritus*. Summer temperatures are assumed to be high enough for incubation of eggs. Nessov also believed that Chukotka dinosaurs could use warm volcanic soils for nesting. Some present-day primitive tropical species of weed-infested hens (Megalopodiidae) inhabiting the Sunda Isles used to do so these days.

Other advocates of the settled way of life of the dinosaurs believe: "polar" reptiles were warm-blooded (homoiothermic) animals and, like present-day mammals, could survive the severe conditions of relatively cold winters. To confirm this idea, scientists refer to the data of observations that the remains of cold-blooded (poikilo-thermic) animals--amphibians, turtles, lizards and crocodiles--were never found in the Northern Alaska and Kakanaut sites. This may mean that the latest Cretaceous dinosaurs were able to survive the Arctic climate that turned out to be too severe for typical cold-blooded animals.

To estimate reliability of the opinions above, we should study paleobotanical data and find answers to two questions: what was the food supply of Arctic dinosaurs and what can a geological record tell us about the Arctic climate in the latest Cretaceous?

VEGETATION AND CLIMATE

Fortunately, numerous imprints of leaves and leafy shoots were found both in Northern Alaska and Chukotka in the same layers as remains of dinosaurs. This means that we have valuable paleobotanic evidence depicting vegetation that surrounded the Arctic dinosaurs.

From the late 1970s, fossil plants of Northern Alaska were studied by the English paleobotanist and paleocli-matologist Robert Spicer. He found out: they were of a low taxonomic diversity (only 10-12 species), and relatives of the present-day Metasequoia and Taxodium--taxodioid conifers belonging to the Parataxodium genus and represented by small deciduous trees that shed their leafy shoots at the end of the vegetation period--were dominating. The understory was represented by horsetails, ferns and herbaceous angiosperms with surface parts dying at the end of the growing season; in addition to horsetails, water bodies were also inhabited by aquatic flowering plants Quereuxia.

Fossil plants discovered at the Kakanaut dinosaur site have been studied by L. Golovneva. It turned out that the discovered vegetation was substantially different from one typical of Northern Alaska: it was more diverse (about 50 species) and thermophilic, which is of little wonder, as it grew more than 1,000 km southwards. It is interesting to note that probable evergreen plants--relatives of the present-day Cycas--were discovered there. They were studied and described by L. Golovneva and Valentin Krassilov (Dr. Sc. (Geol. & Min.), Borissyak Paleontological Institute, RAS, Moscow). In addition to thermophilic plants, the Kakanaut flora included numerous deciduous ginkgoalean plants as well as deciduous and evergreen conifers, with taxodioid plants, especially Sequoia, being the most frequent. Angiosperms (about 30 species) were also numerous and represented by deciduous taxa. Mosses, horsetails and ferns were rather rare.

Fossil plants give us information on the ancient climatic conditions, both qualitative (warm or cold, humid or dry climate) and quantitative characteristics. For the last 20 years, a new calculation method CLAMP (Climate-Leaf Analysis Multivariate Program) was created and tested. It was first proposed in 1993 by US paleobotanist Jack Wolfe and significantly improved by English expert Robert Spicer and the author of this article*. The core of this method is to determine a statistic correlation between morphological character of leaves and climatic parameters. Land plants interact with the atmosphere in the process of gas and water exchange and have to adapt to local natural conditions in order to survive. Since similar climatic types impose similar physical limitations on the plants, different species of plants growing in similar climatic conditions but different by space and time develop convergent morphological char-

* Detritus--particles of a dead organic matter suspended in water. A typical element of biocoenotic environment of aquatic communities.--Ed.

* See: A. Herman. "Paleobotany and Paleoclimates of the Earth: Reminiscing About the Future", Science in Russia". No. 1, 2006.--Ed.

acteristics. Consequently, leaf morphology of fossil plants can be used to determine climatic characteristics of the past.

This approach was implemented in the methodology of the above-mentioned multivariate statistical analysis: CLAMP makes it possible to estimate several quantitative parameters of the ancient climate using climatic signals encoded in the morphology of woody dicot leaves. The improved method is based on the so-called canonical correspondence analysis (CANOCO) that determines correlation between 31 morphological characteristics of the dicot leaves (their shape, size, etc.) and 11 climatic variables (mean annual, warm month mean and cold month mean temperatures, precipitations during three consecutive wettest and driest months, duration of the growing season, etc.). As a result, it is possible to estimate temperature and humidity variables for the fossil flora as well as possible quantitative errors. The CLAMP can give an answer to the following question: if this or that ancient flora would grow at the present time, in what climatic conditions we would expect to observe it?

Applying this method to the ancient plants of the Kakanaut Flora, L. Golovneva estimated likely variables of the climate the Chukotka dinosaurs lived in. According to her estimates, the mean annual temperature was about 10 °C, the warm month mean temperature was 19 °C, the cold month mean temperature -3 °C, growing season lasted 6.3 months, and the mean monthly growing season precipitation was 98 mm. In tempera-

tures this ancient climate was similar to that of the present-day London, but was more humid.

It is true that scientists cannot use the CLAMP to estimate paleoclimatic variables of the latest Cretaceous in Northern Alaska "directly": the matter is that at high paleolatitudes (over 80 °N) the diversity of woody dicot leaves is not high enough to apply this method. That is why we had to estimate latitudinal gradients of paleotemperatures using CLAMP technique on several lower latitude floras of the Northern hemisphere, which are rich in angiosperms, and then, using these gradients, estimate likely paleotemperature variables of the paleo-climate experienced by dinosaurs of Northern Alaska. According to such approach it was established that 70 mln years ago at 82 °N the mean annual temperature was about 6.3 °C, the warm month mean temperature was 14.5 °C, and the cold month mean temperature was 2.0 °C below zero.

THE PROBLEM OF OVERWINTERING

As it was already mentioned before, all experts in dinosaurs who studied their remains discovered in Northern Alaska and Chukotka stick to the point that the reptiles lived in the Arctic Regions all year round. However, the data on the ancient flora and paleoclimate of high latitudes of the Arctic Regions make us doubtful. It is clear that the problem of overwintering in Northern Alaska is the most interesting issue for scientists: compared with their Chukotka relatives, dinosaurs of Northern Alaska lived at a higher paleolatitude in the environ-

ment of an impoverished vegetation. Another point of interest is a lifestyle of plant-eating reptiles, since if they had enough sources of heat and nutrition, carnivorous animals eating them hardly had any problems.

Analysis of the problem of nutrition of the plant-eating Arctic dinosaurs should be commenced with the peculiarities of a light regime in high latitudes. The present-day Arctic Regions are characterized by an apparent seasonality: the polar day with a constant sunlight is replaced here by a long winter night. According to the present knowledge, Earth obliquity in the Late Cretaceous was the same as today. This means that a permanent summer day at the 82 °N (paleolatitude of the Liscomb Quarry) lasted almost five months. It is no wonder that a plenty of sunlight during warm summers had resulted in rapid growth of vegetation providing dinosaurs with an abundant forage. At the 72 °N (paleolatitude of the Kakanaut site), a summer day was shorter, around three months, but it along with a warm growing season was also favorable for growth of plants. Obviously the abundance of food attracted plant-eating dinosaurs to high latitudes, and they were followed by carnivorous animals.

As for the polar night at 82 °N, it lasted over three months, from November till early February, with 2-3 weeks of twilight before and after each night. At the 72 °N there was no multiweek continuous winter darkness, but twilight period lasted about two months. So, it is no wonder that all woody plants of Northern Alaska were leaf - or twig-shedding, and surface parts of herbaceous forms died for a winter season; this was likely caused by both negative winter temperatures and absence of sunlight during polar nights. This definitely supports the idea that dinosaurs of Northern Alaska had a dramatic lack of food supply in winter. Deciduous plants dominated in Chukotka, but there were also evergreen plants there, and hence the winter food problem was less severe for the Kakanaut dinosaurs.

As we have already mentioned, the cold month mean temperature in the latest Cretaceous calculated for the 82 °N was --2.0 °C. Probably, in the course of this month temperatures could drop down to --8-10 °C for a short time--for some days or weeks. And it is very doubtful that big dinosaurs of Northern Alaska could survive such low temperatures. Today, some reptiles, for example, small turtles, are able to survive winter periods with negative temperatures in a state of a very low activity--literally in a frozen state. But what is now suitable for small animals that can easily find shelter in the underground holes, was absolutely unlikely for big (up to 9-10 m in length) Arctic dinosaurs that were not able to dig vast shelter holes in the ground and go into hibernation. And winter hibernation out of a secure shelter in the midst of predators, including warm-blooded ones, could result in a very dramatic consequence... The hypothesis that the reptiles used ground and water bodies heated by volcanoes for their overwintering and egg incubation seems to be rather extravagant and devoid of serious proofs; the more so especially as no volcanic activity is recorded in Northern Alaska of those times.

As we have already mentioned before, some scientists who supported the idea on the settled way of life of Arctic dinosaurs believe that the latter could survive cold winters owing to their homoiothermy. However, popular hypothesis on the homoiothermy of reptiles is lacking a reliable proof, since all arguments of its advocates are based on indirect and poorly substantiated facts or assumptions (review of the reasoning is presented in the book by English paleontologist David Norman The Illustrated Encyclopedia of Dinosaurs). Anyway, no proof of homoiothermy of Northern Alaska reptiles was found. On the contrary, no signs of wool were discovered in the imprints of the scaly skin of edmontosaurs that could speak for homoiothermy of these animals.

But if we even assume that Arctic dinosaurs were warm-blooded and could survive cold winters, this will

Distribution of dark and light hours in the high latitudes of the Arctic Regions during a year. Latitudes of 72° and 82 °N are colored red.

hardly solve the problem of overwintering. The point is that warm-blooded animals need much more food than cold-blooded ones: they "burn" most of food to maintain a constant body temperature and "heat the atmosphere" due to a more intense thermal exchange with the environment. A remarkable example of "food wastefulness" was given by Sergei Severtsov, Dr. Sc. (Biol.), in his work of 1941: "An ermine that needs to get meat equivalent to the half of its weight daily and a viper that spends half of the year in anabiosis and in summer needs only two or three field voles a week are a good example of different life energies. Crocodiles in a zoo were happy with 36 g of meat a day. Panthers and leopards from the same weight category receive 3.5-4 kg of meat daily". In general, we can say that warm-blooded animals need approximately 10 times more food than cold-blooded ones.

As for the food, it was scarce in winter in the Arctic Regions. As we remember, vegetation was represented by leaf - and twig-shedding plants, i.e. in winter there were no green leaves and fresh leafy shoots. To eat tree waste and dried herbs, ferns and lichens (as reindeers do) is quite difficult, if at all possible: at the end of the Cretaceous the climate in the Arctic Regions was humid, characterized by heavy precipitation during the whole year. This means that in winter the ground together with everything eatable on it were covered with a thick snow layer. An assumption that plant-eating dinosaurs of Northern Alaska could eat aquatic plants (their remains have been discovered in the Upper Cretaceous of the region) for several winter months is incredible, since in terms of temperatures below zero backwaters and lakes were covered with ice. Of course, one may assume that in winter reptiles used to eat bark of trees and shrubs, wood pulp and timber fungi. But it sounds odd, taking into account size and number of these animals.

Besides, such drastic change in the dinosaur "diet" would inevitably cause changes in their bone structure. The point is the following. Recently, reliable data on the biology of dinosaurs of Northern Alaska were obtained by histologist Anasuya Chinsami from the University of Capetown (Republic of South Africa), who studied bone microstructure of the Edmontosaurus hadrosaurs. She discovered two interesting facts. First, no traces of seasonal slowdown of bone growth--specific "growth rings" marking unfavorable periods characterized by the lack of nutrients--were detected in the bones found and studied so far. Therefore, dinosaurs did not suffer from significant changes in the "diet" or "fasting" periods due to the lack of fresh vegetable food. Secondly, fossil bones of these animals are well "vascularized" and have a lot of small channels serving as routes for blood vessels, which speaks for a high speed of bone growth. The scientist discovered this very peculiarity when he studied bones of another plant-eating dinosaur of the Pachyrhinosaurus genus. This means that young animals grew very fast. Taking into account apparently seasonal climate of Northern Alaska in the latest Cretaceous and deciduous nature of the plants that surrounded edmontosaurids, we can make a conclusion: these animals migrated seasonally--moved southwards following retreating warm days where they could find enough food for normal life. By the end of summer or beginning of autumn young reptiles were strong and big enough to migrate for long distances together with adults.

We can produce two more "indirect pieces of evidence" in favor of seasonal migration of Arctic dinosaurs. First, plant-eating mammals such as reindeers or caribou used to move from one place to another simply, because they eat away all food and have to find other pastures. In other words, herds of these animals are forced to move all the time--so why could not the Arctic dinosaurs move northwards in spring and backwards in autumn? As Russian geologist Nikolai Chumakov noted in the mid-1990s, in the Cretaceous pasture lands located along the shores of the longitudinal Western Interior Seaway in the North America could serve as perfect

Microstructure of bones: a--reptiles, b-dinosaurs. Structure of bones, belonging to dinosaurs, big mammals and birds is alike--they have more blood vessels than bones of reptiles. This fact is treated as an evidence of homoiothermy of dinosaurs, but there is another explanation that seems to be more real: this specific characteristic shows that these animals grew very fast.

migration routes. Secondly, the closest present-day relatives of dinosaurs--birds--migrate seasonally following "warmth and food". Might it be that birds inherited this lifestyle from their four-footed Mesozoic ancestors?

A logical question arises: how far did dinosaurs of Northern Alaska migrate? Speaking of present-day four-footed Arctic animals, cariboos are champions of longdistance migrations, every year they cover up to 5,500 km. A. Fiorillo and R. Gangloff reasonably believe: cariboos are not similar to dinosaurs, and it would be incorrect to take them as a model illustrating ancient way of life, but still, this example is important. Evergreen cycadalean shrubs and conifers, as well as aquatic plants, typical of Chukotka about 1,000 km to the south from the habitat of Alaska dinosaurs, could be a winter nutrition base for plant-eating reptiles, and Kakanaut ankylosaurids hardly were good "walkers", judging by their body structure. On the other hand, estimated (using the Kakanaut fossil leaves morphology) temperature of the coldest month of 3 °C does not seem warm enough for the survival of big reptiles. With a certain conventionality we can assume: to reach regions with an evergreen vegetation and suitable climate, dinosaurs of Northern Alaska should have covered 1,200-1,300 km southwards, approximately to the latitude of Southern Alaska.

Egg shells were found in the Kakanaut site; as for Northern Alaska, they were not found until now. One can assume that Alaska dinosaurs, like Chukotka ones, reproduced in summer and in the Arctic Regions, but this hypothesis gives rise to some questions we are unable to answer at the moment: were the Alaskan summers warm enough (with the temperature of the hottest month up to 14.5 °C) to ensure proper development of eggs of cold-blooded animals of those times, and were summer pasture seasons long enough to ensure birth and breeding of baby dinosaurs? In view of these questions, a hypothesis offered by some paleontologists seems reasonable: dinosaurs could use decaying and heat-emitting plants for their nesting, and their hatchling grew very fast and reached size allowing them to avoid attacks of small predators and migrate with elder reptiles.

At the very end of the Cretaceous, dinosaurs became extinct everywhere. The prevailing but rather disputable hypothesis assumes a large-scale impact event that could happen at the Cretaceous/Palaeogene boundary--fall of one or more asteroids that led to a catastrophic extinction of animals and plants. The extinction could be triggered by a long period of darkness and quick climatic cooling due to a global distribution of dense clouds of dust and sulphate aerosols as well as catastrophic forest fires. However, supporters of a nonmigratory way of life of ancient reptiles in the Arctic Regions point out that these animals were well adapted both to the dark polar night and cold Arctic climate--so these factors could not lead to their extinction. But if the arguments presented in this article are correct, and Arctic dinosaurs migrated seasonally towards south in winter, the said contradiction becomes irrelevant. In other words, cooling in the latest Cretaceous could easily provoke extinction of these reptiles. But what was the reason of this cooling-terrestrial or extraterrestrial reasons is another question.