scientific correspondence humidity and climate variations related to 9. Minnis, P., Ayers, J. K. & Weaver, S. P. Surface-Based Observations 40 11,12 of Contrail Occurrence Frequency over the U. S., April 1983–April the North Atlantic oscillation , showed Rhea that none of them, taken individually, is a 1994 (NASA Reference Publication 1404, 1997). Wolf 10.Jensen, E. J. & Toon, O. B. Geophys. Res. Lett. 19, 1759–1762 (1992). Coyote 30 good explanation for the observed positive 11.Hurrell, J. W. Science 269, 676–679 (1995). Pony trend in cirrus occurrence and its regional 12.Mächel, H., Kapala, A. & Flohn, H. Int. J. Climatol. 18, 1–22 (1998). Fox distribution. 13.Hartmann, D. L., Ockert-Bell, M. E. & Michelsen, M. L. J. Clim. Budgerigar d

5, 1281–1304 (1992). t 20

s Raven

· Humming- 14.Warren, S. G., Hahn, C. J., London, J., Chervin, R. M. & Jenne, E

The trends in cirrus ‘amount when pre- /

o R. L. Global Distribution of Total Cloud Cover and Cloud Type c o sent’ over the previously defined regions of l · Pigeon E Turkey Amounts over the Ocean (NCAR Technical Note TN-317 + STR, 10 Most high fuel consumption are ǁ1.9% and Boulder, Colorado, 1988). mammals Penguin ǁ4.2% for land and ocean, respectively 15.Hahn, C. J., Warren, S. G. & London, J. J. Clim. 8, 1429–1446 Stork Penguin (1995). (Table 1). The combination of a large posi- 0 Supplementary information is available on Nature’s World-Wide 0.001 0.01 0.1 1 10 100 1,000 tive trend in cirrus occurrence associated Web site (http://www.nature.com) or as paper copy from the with a negative trend in the cirrus amount London editorial office of Nature. Body mass (kg) when present is consistent with an increas- Figure 1 Factorial increases in metabolic rates above ing amount of condensation-trail cirrus resting minimum for rheas and athletic mammals at covering a small portion of an otherwise the aerobic maximum and for other avian at clear sky at the cirrus cloud level. This High metabolic rates the highest rates available. Triangles, flying ; results in a trend in cirrus coverage (cirrus squares, walking and running birds; circles, swim- occurrence ǂ cirrus amount when pre- in running birds ming birds; and diamonds, running mammals. sent) of 0.0% averaged over land, rising to 2.9% over the continental regions with The ability to increase metabolic rate during (the maximum aerobic metabolic rate most air traffic. Over ocean, these values locomotion has been important in the divided by the minimum resting rate2) of are 2.3% and 1.0%, respectively, because of structural evolution and evolutionary suc- these relatively inactive birds is more than the aforementioned contribution of the cess of both birds and mammals. Greater 3.5 times those of typical mammals1, and North Pacific Ocean to the global increase endurance capabilities are conferred directly equals those of some of the most capable in cirrus occurrence. Both radiative trans- by greater maximal metabolic rates, which mammalian endurance runners (Fig. 1). fer calculations and Earth Radiation Bud- vary between species. These maximal rates Although the rheas’ factorial increase in get Experiment measurements of the net are known for many mammals1 but have not anaerobic metabolism is 1.7 times greater change in radiation associated with high, been determined for birds. We have mea- than the highest previously reported avian thin clouds13 suggest a 0.2 W mǁ2 cloud sured oxygen consumption in a large flight- values (Fig. 1), we believe that their aerobic radiative forcing per 1% of condensation less bird, the rhea, Rhea americana, while it scope is representative of birds in general. trail or high-level cloudiness. I therefore was running on an inclined treadmill, and Previous avian values do not provide evi- estimate that the trend in cirrus amount find an upper limit to aerobic metabolism dence of being upper limits; they are proba- over the continental regions of high air that is 36 times greater than the minimum bly below species’ maxima, but to unknown traffic (2.9%, as discussed above) would resting rate, a factorial increase exceeding extents. Nonetheless, the minimum meta- correspond locally to an additional cloud that reported for nearly all mammals. bolic increase required for flight3 of 15 times radiative forcing of about 0.7 W mǁ2 We trained two female rheas (average the resting value exceeds the aerobic power between 1982 and 1991. A larger forcing mass 21.8 kilograms) for two years to run of a typical mammal by a factor of 1.5, and would be expected if calculated from the on a treadmill while wearing clear plastic most avian species can fly. Even birds with beginning of commercial jet traffic until hoods. We then measured linear increases more limited endurance, such as the pen- the present time. However, we cannot rule in rates of oxygen uptake while the birds guin and domestic fowl, can increase their out the possibility that the observed change ran up a 16° incline. The maximum rate of aerobic metabolism by as much as a typical in cirrus amount is not due solely to avia- oxygen uptake was 2.85 millilitres per kilo- mammal4,5. We suggest that, on average, the tion, or that other climate processes have gram of body mass per second, reached at a aerobic scope of birds exceeds those of not masked an even larger aviation impact running speed of 4.0 metres per second. mammals by a factor of two or more. on cirrus cloudiness. For instance, a previ- There were no further increases in oxygen The structural basis of the rheas’ aero- ous analysis14 showed that other regions uptake when running speeds were increased bic power indicates that respiratory experienced large positive trends in cirrus to 4.7 metres per second, and plasma lactate structure–function relationships, previously amount between 1952 and 1981. concentrations indicated an increasing undetermined for birds, are quantitatively Olivier Boucher reliance on anaerobic glycolysis to provide similar to those of mammals throughout Laboratoire d’Optique Atmosphérique, metabolic energy at these higher speeds. most of the respiratory system, and are U.F.R. de Physique, Université de Lille–I, Rates of lactate accumulation (4.4 to 6.8 consistent with the concept of reasonable 59655 Villeneuve d’Ascq Cedex, France mM minǁ1) and peak concentrations (24.8 design6 (Table 1). Rheas, like aerobic e-mail: [email protected] mM) matched values reported for mam- mammals, achieve maximal high rates of 1. Seinfeld, J. H. Nature 391, 837–838 (1998). mals running at speeds above their aerobic oxygen uptake with existing cardiovascular 2. Minnis, P. et al. Geophys. Res. Lett. 25, 1157–1160 (1998). maximum. Remarkably, the aerobic scope and muscular structures, rather than by 3. Schumann, U. Meteorol. Z. 5, 4–23 (1996). 4. Jensen, E. J. & Toon, O. B. Geophys. Res. Lett. 4, 249–252 (1997). Table 1 Mitochondrial volume densities and capillary densities of rhea muscles 5. Hahn, C. J., Warren, S. G. & London, J. Edited Synoptic Cloud Muscle Bird A Bird B Reports from Ships and Land Stations Over the Globe, V (%) N (mm–2) V (%) N (mm–2) 1982–1991, NDP026B (Carbon Dioxide Information Analysis v a v a Center, Oak Ridge National Laboratory, Tennessee, 1996). Iliotibialis lateral 10.1 803 6.5 787 6. Baughcum, S. L., Henderson, S. C., Tritz, T. G. & Pickett, D. C. Gastrocnemius 11.8 864 7.9 642 Scheduled Civil Aircraft Emission Inventories for 1992: Database Pectoralis 4.2 411 4.2 469 Development and Analysis (NASA CR-4700, 1996). 7. Mortlock, A. M. & van Alstyne, R. Military, Charter, Unreported Humerotriceps 4.3 619 4.9 508 Domestic Traffic and General Aviation: 1976, 1984, 1992, and Mean mitochondrial volume densities (Vv) and capillary densities (Na) for leg muscles (iliotibialis and gastrocnemius) 2015 Emission Scenarios (NASA CR-207639, 1998). were 9.04% and 774 mm–2, respectively; values for wing muscles (pectoralis and humerotriceps) were 4.37% and 8. Energy Statistics and Balances of Non-OECD Countries, 1994–1995 502 mm–2, respectively, for the rheas. Values for equivalent leg muscles of similarly sized mammals: for a 28-kg dog, –2 –2 124–125 (International Energy Agency, Organization for Vv ǃ 9.1% and Na ǃ 884 mm ; for a 26-kg goat, Vv ǃ 3.4% and Na ǃ 719 mm . Economic Cooperation and Development, Paris, 1997).

NATURE | VOL 397 | 7 JANUARY 1999 | www.nature.com © 1999 Macmillan Magazines Ltd 31 scientific correspondence qualitative adjustment. Rhea heart mass 3. Ellington, C. P. J. Exp. Biol. 160, 71–91 (1991). our DNA analysis, including the interesting 4. Brackenbury, J. H. & El-Sayed, M. S. J. Exp. Biol. 117, 349–355 (267 grams) conforms to that predicted for (1985). one proposed by Davis. The title assigned to a mammal of the same body size and aero- 5. Kooyman, G. L. & Ponganis, P. J. J. Exp. Biol. 195, 199–209 (1994). our study was misleading in that it repre- bic power. Volume densities of mitochon- 6. Weibel, E. R., Taylor, C. R. & Hoppeler, H. Proc. Natl Acad. Sci. sented only the simplest explanation of our dria in leg muscles are similar to those of USA 88, 10357–10361, (1991). molecular findings: namely, that Thomas 7. Maina, J. N. & King, S. A. J. Anat. 163, 67–73 (1989). mammals of the same size and aerobic 8. Dubach, M. Respir. Physiol. 46, 43–60 (1981). Jefferson, rather than one of the Carr broth- capacity, whereas those in the relatively 9. Piiper, J. & Scheid, P. in Comparative Physiology (eds Bolis, L., ers, was likely to have been the father of inactive flight muscles are only half as large Schmidt-Nielsen, K. & Maddrell, S. H. P.) 161–185 (North- Eston Hemings Jefferson. Holland, 1973). and correspond to values reported for less 10.Taylor, C. R., Dmi’el, R., Fedak, M. & Schmidt-Neilsen, K. It had been suggested to us earlier (by aerobic mammals (Table 1). Am. J. Physiol. 51, 772–776 (1981). Herbert Barger, who also helped to recruit Given that 30% of the rhea’s body mass the descendants of Field Jefferson who par- consists of leg musculature, rates of oxygen ticipated in our study) that Isham Jefferson, uptake by mitochondria per millilitre dur- son of Thomas Jefferson’s brother Ran- ing locomotion at the aerobic maximum The Thomas Jefferson dolph, might have been the father of one or appear to fall within the range reported for more of Sally Hemings’ children. Barger’s mammals. Capillary densities in the rhea paternity case proposal was based on a statement5 that leg and flight muscles, like the mitochon- Isham was reared by Thomas Jefferson; he drial volume densities, also parallel values The DNA analysis of Y-chromosome haplo- suggested that Isham could have been at reported for the muscles of athletic and less types used by Foster et al.1 to evaluate Monticello or at Snowden (Snowden was mobile mammals. Thomas Jefferson’s alleged paternity of across the James River from Scottsville, In contrast to the apparent conservation Eston Hemings Jefferson, the last child of which is about 20 miles from Monticello) of structure–function relationships in most his slave Sally Hemings, is impressive. How- when Eston Hemings was conceived. But it of the respiratory system, our results sug- ever, the authors did not consider all the is not known for certain that Isham was at gest that there are basic differences in the data at hand in interpreting their results. Monticello at that time, whereas it is docu- performance of the lungs of birds and No mention was made of Thomas Jeffer- mented that Thomas Jefferson was. From mammals. We could not measure the lung son’s brother Randolph (1757–1815), or of the historical knowledge we have, we can- volumes of the rheas directly, but in birds his five sons2,3. Sons of Sally Hemings con- not conclude that Isham, or any other these are normally slightly more than half ceived by Randolph (or by one of his sons) member of the Jefferson family, was as likely of those of mammals of the same mass7,8. would produce a Y-chromosome analysis as Thomas Jefferson to have fathered Eston The rheas therefore achieved maximum identical to that described by Foster et al. Hemings. oxygen flux rates, equal to those of the most Further collaborative data (for example, the We know from the historical and the aerobic mammals of their size, using lungs whereabouts of any of those who might have DNA data that Thomas Jefferson can neither that are probably only half as large. This been involved at conception) are needed to be definitely excluded nor solely implicated supports the general belief that avian lungs confirm that Jefferson did indeed father his in the paternity of illegitimate children with provide relatively more function per unit slave’s last child, as claimed in the title. We his slave Sally Hemings. When we embarked volume than mammalian lungs8,9. know Thomas Jefferson was there, but how on this study, we knew that the results could Although the aerobic limits of rheas and about Randolph Jefferson and his sons? not be conclusive, but we hoped to obtain athletic mammals are similar, the metabolic David M. Abbey some objective data that would tilt the power available in practice, and their func- Health Science Center, University of Colorado, weight of evidence in one direction or tional needs, are not. Unlike dogs, horses 1512 Teakwood Court, Fort Collins, another. We think we have provided such and other athletic mammals that sustain Colorado 80525, USA data and that the modest, probabilistic high metabolic rates for hours during pre- e-mail: [email protected] interpretations we have made are tenable at dation and migration, rheas do little or no present. sustained running and are poor at dissipat- If the data of Foster et al. are accurate, then E. A. Foster*, M. A. Jobling†, P. G. Taylor†, ing metabolic heat loads10. Rheas have any male ancestor in Thomas Jefferson’s P. Donnelly‡, P. de Knijff§, R. Mieremet§, apparently not been under strong selective line, white or black, could have fathered T. Zerjal¶, C. Tyler-Smith¶ pressures like those that promoted the aero- Eston Hemings. Plantations were inbred *6 Gildersleeve Wood, Charlottesville, bic power of extant running mammals. communities, and the mixing of racial types Virginia 22903, USA Large flightless birds lead fairly inactive was probably common. As slave families e-mail: [email protected] lives, and may have lost the ability to fly were passed as property to the owner’s off- †Department of Genetics, primarily because of a lack of predation. spring along with land and other property, University of Leicester, Adrian Building, Why rheas have an aerobic power that it is possible that Thomas Jefferson’s father, University Road, Leicester LE1 7RH, UK greatly exceeds their apparent functional grandfather or paternal uncles fathered a ‡Department of Statistics, needs remains a puzzle. male slave whose line later impregnated University of Oxford, South Parks Road, Matthew W. Bundle*, Hans Hoppeler†, another slave, in this case Sally Hemings. Oxford OX1 3TG, UK Ruth Vock†, June M. Tester*, Sally herself was a light mulatto, known §MGC Department of Human Genetics, Peter G. Weyand* even at that time to be Thomas Jefferson’s Leiden University, PO Box 9503, *Museum of Comparative Zoology, wife’s half sister3,4. 2300 RA Leiden, The Netherlands Harvard University, Gary Davis ¶Department of Biochemistry, Bedford, Massachusetts 01730, USA Evanston Hospital, 2650 Ridge Avenue, University of Oxford, South Parks Road, e-mail: [email protected] Evanston, Illinois 60201, USA Oxford OX1 3QU, UK

†Department of Anatomy, 1. Foster, E. A. et al. Nature 396, 27–28 (1998). University of Bern, Foster et al. reply — It is true that men of 2. Mayo, B. & Bear, J. A. Jr Thomas Jefferson and his Unknown Buhlstrasse 26, CH 3000 Bern 9, Switzerland Randolph Jefferson’s family could have Brother (Univ. Press of Virginia, Charlottesville, 1981). fathered Sally Hemings’ later children. 3. Brodie, F. M. Thomas Jefferson: An Intimate History (Norton, 1. Seeherman, H. J., Taylor, C. R., Maloiy, G. M. O. & Armstrong, New York, 1974). R. B. Respir. Physiol. 44, 11–23 (1981). Space constraints prevented us from 4. Randall, W. S. Thomas Jefferson: A Life (Holt, New York, 1993). 2. Lasiewski, R. C. & Dawson, W. R. Condor 69, 13–23 (1967). expanding on alternative interpretations of 5. History of Todd County, Kentucky (1884).

32 © 1999 Macmillan Magazines Ltd NATURE | VOL 397 | 7 JANUARY 1999 | www.nature.com