Lower-latitude mammals as year-round residents in Eocene Arctic forests J. Eberle1, H. Fricke2, and J. Humphrey3 1University of Colorado Museum and Department of Geological Sciences, University of Colorado, Boulder, Colorado 80309, USA 2Department of Geology, Colorado College, Colorado Springs, Colorado 80903, USA 3Department of Geology and Geological Engineering, Colorado School of Mines, Golden, Colorado 80401, USA ABSTRACT of plants to discriminate between carbon isotopes (Farquhar et al., 1989). The Arctic is undergoing rapid warming, but the impact on the The result is higher isotopic ratios for plants living in hot, dry, and nutri- biosphere, in particular on large terrestrial mammals, is not clear. ent-poor settings. Moreover, carbon isotope fractionation is taxon specifi c; Among the best deep time laboratories to assess biotic impacts of for example, gymnosperms typically have higher carbon isotope ratios Arctic climate change, early Eocene (ca. 53 Ma ago) fossil assem- than angiosperms (Heaton, 1999). blages on Ellesmere Island, Nunavut (~79°N), preserve evidence of Although oxygen in mammalian bioapatite has sources in ingested forests inhabited by alligators, tortoises, and a diverse mammalian water and atmospheric oxygen, isotopic variations in ingested water play fauna most similar to coeval lower-latitude faunas in western North a primary role in corresponding variations in isotopic ratios of bioapa- America. By analyzing carbon and oxygen isotope ratios of mam- tite (Kohn, 1996). Oxygen isotope ratios of ingested waters from streams, malian tooth enamel, we show that large herbivores were year-round lakes, and leaves vary in response to environmental factors, particularly inhabitants in the Arctic, a probable prerequisite to dispersal across aridity, as 16O is preferentially incorporated into the vapor phase during northern high-latitude land bridges. If present-day warming contin- evaporation. The other major control on oxygen isotope ratios of surface ues, year-round occupation of the Arctic by lower-latitude plants and waters is the hydrological history of air masses that supply precipitation. animals is predicted. Preferential incorporation and removal of 18O into condensate from cool- ing air masses results in a decrease in oxygen isotope ratios of precipita- INTRODUCTION tion as air masses move from tropical sources to polar sinks (Rozanski et In the Arctic, the most abundant large mammals today are caribou al., 1993). Latitudinal temperature gradients, and thus latitudinal gradients that survive primarily by migrating south up to 1000 km to overwinter in isotopic ratios of precipitation, vary with season, and larger seasonal below the Arctic Circle. There is, however, evidence that Arctic regions temperature ranges at high latitudes result in larger seasonal isotopic are changing, with temperatures rising at almost twice the global average. ranges in precipitation (Fricke and O’Neil, 1999). This climate change begs the question of how large terrestrial mammals will respond: will they migrate or live year-round in the Arctic? METHODS From a paleontological perspective, the study of how ancient biota Arctic fossils for this study are from the Margaret Formation, Eureka responded to past global warming is a way to address this question. One Sound Group, on Ellesmere Island. An early Eocene (late Wasatchian; such warm interval occurred during the early Eocene (ca. 50–55 Ma ca. 53 Ma ago) age is inferred for this fauna based upon fossil mammals ago), when mean annual temperatures in the Arctic may have been 40 (Dawson et al., 1993) and palynology (Norris and Miall, 1984). Rocks °C warmer than today (Sluijs et al., 2006). The Arctic contained swamp comprise coarsening-upward cycles of interbedded sandstone, siltstone, forests of predominantly deciduous conifer (McIver and Basinger, 1999) mudstone, and coal, interpreted as a proximal delta front to delta plain and a warm-weather fauna of alligators, tortoises, lizards, snakes, and a environment with abundant channels and coal swamps (Miall, 1986). diversity of mammals (Estes and Hutchison, 1980; Eberle and McKenna, Macrofl oral assemblages indicate a mostly deciduous swamp forest of 2002). As the relevant fossil-bearing rocks on Ellesmere Island were ~77° conifer and broadleaf angiosperms (McIver and Basinger, 1999), con- N paleolatitude (Irving and Wynne, 1991), this environment underwent trasting with predominantly angiosperm forests at mid-latitudes (Wing months of continuous sunlight as well as darkness. et al., 1995). The early Eocene Arctic mammal fauna shares most genera with Arctic mammal fossils are rare; only ~80 Coryphodon, 40 tapiroid, coeval mid-latitude faunas thousands of kilometers to the south in western and 7 brontothere specimens are documented. We sampled tooth enamel North America (Eberle and McKenna, 2002). These similarities suggest from 9 Coryphodon and 5 perissodactyls (n = 3 for tapirs; n = 2 for bron- that animals from lower latitudes were either adapted to living in polar totheres). A bulk sample was taken by drilling enamel from the base to regions or were migratory. We use oxygen and carbon isotope ratios of the top of the tooth or tooth fragment. For six Coryphodon teeth, multiple fossil tooth enamel to test the hypothesis that some Eocene lower-latitude samples were taken along the tooth’s length to determine seasonal varia- mammals lived year-round in the Arctic. In so doing, our study provides tions in isotope ratios. Oxygen and carbon isotope ratios were measured a deep time analog from which to predict the consequences of imminent for the carbonate component of tooth enamel (see the GSA Data Reposi- global warming on today’s Arctic biota. tory1 for details of sample preparation, isotopic analyses, and diagene- sis). Stable isotope ratios are reported as δ13C and δ18O values, where δ = Stable Isotopes, Behavior, and Ecology (Rsample/Rstandard – 1) × 1000‰, and the standard is Vienna Peedee belem- Animals record the isotopic characteristics of ancient landscapes nite for carbon and Vienna standard mean ocean water for oxygen. Arctic when they ingest organic material and drink from surface water reservoirs isotopic data are compared to published data for three late Wasatchian and then form bioapatite [Ca5(PO4, CO3)3(OH, CO3)]. Carbon in the car- bonate phase of bioapatite is related to ingested organic material such as 1GSA Data Repository item 2009119, details of sample preparation, isotope plants in the case of herbivores (Kohn, 1996). In turn, carbon isotope com- analyses, and diagenesis; tables of stable isotope data from bulk samples of Arctic positions of Eocene plants varied primarily in response to environmental Coryphodon and perissodactyls, as well as sequential samples of Arctic Cory- phodon; and Fig. DR1 (changes in seasonal range in δ18O of precipitation with conditions and plant-specifi c differences in carbon isotope discrimination latitudinal differences), is available online at www.geosociety.org/pubs/ft2009. during photosynthesis. Environmental conditions infl uence the opening of htm, or on request from [email protected] or Documents Secretary, GSA, leaf stomata, which control the concentration of CO2 and thus the ability P.O. Box 9140, Boulder, CO 80301, USA. © 2009 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. GEOLOGY,Geology, June June 2009; 2009 v. 37; no. 6; p. 499–502; doi: 10.1130/G25633A.1; 3 fi gures; Data Repository item 2009119. 499 (ca. 53 Ma ago) Coryphodon teeth from the Willwood Formation, Bighorn ~6 months of growth Basin, Wyoming (Fricke et al., 1998). Strata consist of channel sands, Winter Fall Summer Specimen No. 23 fl oodplain siltstones, and mudstones, with variable paleosol development UM96744 22 B characterized by red beds and carbonate nodules that imply seasonal fl uc- UM88232 21 tuation of water table (Kraus, 2001). UM88268 20 , ‰, VSMOW)‰, , 3 19 RESULTS AND DISCUSSION δ18O (CO , ‰, VSMOW) 3 18 17 19 21 23 O (CO O -8 18 17 Isotopic Differences Among Arctic Herbivores δ 05101520253035404550 δ18 δ13 -10 Figure 1 shows O and C for bulk samples of tooth enamel -10 Sample distance from base of tooth (mm) from Arctic Coryphodon and perissodactyls (Table DR1). Overlapping -11 δ18 O among Arctic Coryphodon and perissodactyls suggest a humid -12 -12 , ‰, VPDB) 3 early Eocene High Arctic, consistent with sedimentology (e.g., abun- , ‰, VPDB) 3 -13 dant coal seams and absence of red beds; Miall, 1986) and swamp forest -14 18 C (CO -14 fl ora. Humidity controls the degree to which leaves are enriched in O 13 C (CO δ 16 A 13 C over surface water, due to preferential evaporation of O. In dry cli- -16 δ -15 mates, δ18O of vegetation is higher and more variable, and because dif- Figure 2. A: Oxygen and carbon isotope ratios for mid-latitude Cory- ferent browsers consume different plant species, there is less likelihood phodon (Fricke et al., 1998) have positive correlation coeffi cient of of overlap in their δ18O values. 0.32. VPDB—Vienna Peedee belemnite; VSMOW—Vienna standard Mid-latitude studies reveal that Coryphodon has lower δ13C than mean ocean water. B, C: Oxygen and carbon isotope ratios versus coexisting herbivores, a difference that, along with lower variability in tooth position (distance from base) for mid-latitude samples. Cyclic δ18 isotopic variations from low to high occurring over ~50 mm are con- O than other taxa, probably refl ects a semiaquatic lifestyle (Secord et sistent with summer to early winter changes in isotope ratios of in- 13 al., 2008). Arctic Coryphodon tends to have lower δ C than Arctic peris- gested water and plants (Table DR2; see footnote 1). sodactyls, suggesting similar niche partitioning. ~6 months of tooth growth -8 Specimen No. Winter Fall Summer Figure 1. Stable isotope CMN 52851 15 -9 CMN 52852 data from bulk samples 14 of Arctic Coryphodon CMN 52853 -10 CMN 52854 re 13 ratu (gray squares, n = 7) and CMN 52855 pe , ‰, VSMOW) ‰, , m 3 te perissodactyls (black CMN 52856 12 n -11 es i 3 diamonds, n = 5).