SHORT COMMUNICATIONS 465 KING, J. R. 1972. Variation in the songof the Rufous- MILLER, A. H., AND V. D. MILLER. 1968. The be- collared Sparrow, Zonotrichia capensis, in north- havioral ecology and breeding biology of the An- western Argentina. Z. Tierpsychol. 30:344-373. dean sparrow, Zonotrichia capensis. Caldasia 10: KROODSMA,D. E., AND J. R. BAYLIS. 1982. A world 83-l 55. survey of evidence for vocal learning in birds, p. MORELLO,J. 1958. La provincia fitogeograficade1 3 1l-337. In D. E. Kroodsma and E. H. Miller monte. Opera Lilloana 2: 11-155. [eds.], Acoustic communication in birds. Vol. 2. NOTTEBOHM,F. 1969. The songof the Chingolo, Zo- Academic Press,New York. notrichia capensis, in Argentina: description and LENZ, J. 1973. Nachtgesang eines Gartenrotsch- evaluation of a systemof dialects.Condor 7 1:299- wanzes(Phoenicurusphoenicurus). Omithol. Mitt. 315. 25199. NOTTEBOHM,F. 1975. Continental patterns of song LOWTHER,J. K., AND J. B. FALLS. 1968. Zonotrichia variability in Zonotrichia capensis: some possible albicollis, p. 1364-1392. In 0. L. Austin, Jr. [ed.], ecologicalcorrelates. Am. Nat. 109:605-624. Life histories of North American cardinals, gros- SAS INSTITUTE. 1985. SAS statistical softwear ver- beaks, buntings, towhees, finches, sparrows, and sion 5. Carey, NC. allies. U.S. Natl. Mus. Bull. 237. TheCondor 91:465-467 0 The CooperOrnithological Society 1989 MORPHOMETRICS, METABOLIC RATE AND BODY TEMPERATURE OF THE SMALLEST FLIGHTLESS BIRD: THE INACCESSIBLE ISLAND RAIL ’ PETERG. RYAN, BARRYP. WATKINS,AND W. ROY SIEGFRIED Percy Fitzpatrick Institute of African Ornithology, University of Cape Town, Rondebosch 7700, South Africa Key words: Atlantisia rogersi; body temperature; and October-November 1988. We measured,aged, and jlightlessness; Inaccessible Island; metabolic rate. sexedbirds (after Elliott 1957, Richardson 1984), and recordedlinear measureswith Vernier calipers(for bill The Inaccessible Island Rail Atlantisia rogersi is re- measurements,to the nearest 0.1 mm) or a stopped strictedto InaccessibleIsland (37”15’S, 12”3O’W), a 12- wing rule (for flattened wing chord and tarsus, to the km2 island in the Tristan da Cunha Island group, cen- nearest0.5 mm). Mass was recorded to the nearest0.2 tral South Atlantic Ocean. It is the only extant member g on a 100 g Pesolabalance. Student ’s t-tests were used of its genus (Olson 1973) and it is generally accepted to test the significanceof intersexual size differences. to be the smallest flightless bird in the world (Collar We estimated metabolic rates (MR) from oxygen and Stuart 1985), although there are few published (0,) consumption measured in translucent, airtight morphometric data (Lowe 1928, Hagen 1952). The chambers(volume 5 liter) using an open flow-through wingsand tail are greatlyreduced, and the feathershave system. Air from outside the laboratory was pumped few barbules, resulting in a soft, hair-like plumage through a regulatingflowmeter into the chamber, over unique among rails (Lowe 1928). Little is known of the the bird, and out through the following equipment: a ecologyof thejnaccessibleIsland Rail (Collar and Stuart silica gel drying tube, a Rotameter flowmeter, and a 1985). It is widespreadat InaccessibleIsland, occurring. silica gel/Carbosorb/silicagel tube leading to a Taylor in almost all habitats, but spendsmuch of its time in Servomex 0A570 paramagneticoxygen analyzer (after runs and burrows (Lowe 1928, Hagen 1952). We pre- Adams and Brown 1984). Incurrent air was used to set sent morphometric data for InaccessibleIsland Rails the oxygen analyzer at a standard 20.94 %O, every and examine their metabolic rate and body tempera- hour during experiments. The Rotameter flowmeter ture to assessthe extent to which their lifestyle has was calibrated against a Hastings Mass Flowmeter modified their energy metabolism. (Model ECPR- 1A). Esophagealtemperature (assumed to approximate MATERIALS AND METHODS core body temperature, T,) was measuredwithin 1 hr We caught InaccessibleIsland Rails using mist nets, of capture using a thermocouple inserted at least 15 scoopnets, or by hand during September-October1987 mm down the esophagus.We then placed test birds into a chamber and allowed them to become accus- tomed to their surroundingsfor 1 hr before taking read- ingsof flow rate, %O,, and chambertemperature. Read- I Received 28 June 1988. Final acceptance 16 Jan- ings were taken at 30-min intervals for 22 hr, after uary 1989. which time we again recorded esophagealtempera- 466 SHORT COMMUNICATIONS TABLE 1. Morphometrics of male and female InaccessibleIsland Rails. Males (n = 6) Females(n = 7) Significance Dimension (mm/g) R i SD Range R i- SD Range f,, PC Culmen length 21.4 0.4 20.7-22.0 19.8 0.2 19.5-20.1 9.01 0.001 Bill depth (at base) 1.5 0.2 7.4-7.8 6.9 0.4 6.2-7.5 3.06 0.05 Bill width (at base) 5.3 0.2 5.1-5.5 4.9 0.2 4.6-5.2 3.58 0.01 Tarsus ~ ’ 22.9 0.9 22.0-24.5 22.5 1.0 22.0-24.5 0.77 * Wing 54.3 0.5 54.0-55.0 52.1 0.7 51.0-53.0 6.41 ono”o1 Mass? 41.8 3.8 38.0-49.0 36.9 2.8 34.0-42.0 2.58 0.05 * ns= notsignificant. t n = 12, one sick female (mass26.2 g) omitted. tures, and birds were reweighed and released.During the experimental period, birds were subjectedto am- (28.8 kJ.dayml;Aschoffand Pohl 1970; 32.5 kJ,dayml; bient photoperiods and temperatures(12-24°C cham- Kendeiah et al. 1977). a reduction in MR tvnical of ber temperatures). groupsconsidered to ‘have relatively low BMRs (e.g., 0, consumption(at STPD) was calculated(Hill 1972) Caprimulgiformes and Coliiformes; Bartholomew and and converted to an energy equivalent assumingthat Trost 1970, Lasiewski et al. 1970). 1 liter 0, = 20.083 kJ. An estimate of standard met- Mean massloss (t SD) during the approximately 24- abolic rate (SMR) was calculated as the mean of the hr period of enforced starvation was 5.7 + 1.7 g (1 l- four lowest measures of energy consumption during 19% body mass), largely due to defecation and water the 22-hr test period. loss. This fairly substantial mass loss may have con- tributed to the low estimate of MR (cf. Prinzinger et al. 1981, Weathers et al. 1983). However, starvation RESULTS AND DISCUSSION and dehydration is unlikely to have been a major factor MORPHOMETRICS reducing MR because all periods of lowest 0, con- We captured 13 adult InaccessibleIsland Rails. Mean sumption occurred at night, irrespective of when ex- mass of six males and six females was 39.36 i 4.09 g periments were started (three each in the morning and (excludes one sick female, mass 26.2 g). Males were afternoon). Even the MRs of very small birds (< 20 g) significantlylarger than females in all measuresexcept only fall below the initial postabsorptive level after tarsus length (Table 1). Elliott (1957) suggestedthat fasts of 12 hr (Weathers et al. 1983). males were slightly larger than females, but gave no InaccessibleIsland Rail mean T, immediately after supportingdata. Our morphometric data are lessvari- capture (39.9 + 0.7”C, range = 39.3-41.1) was higher able than previously published measurements(Hagen than T, after the experiment and enforced fast (38.5 1952) possibly becauseall birds sampled were at least ir 0.5”C. ranae = 37.8-39.2). This orobablv is due 1 year old and all were measured by one person. Our partly to exet?ion during capture and partly-to star- culmen measurementsare slightlylonger than previous vation causinga reduction in T, (Prinzinger et al. 1981, measurementsprobably as a consequenceof usingdif- Shapiro and Weathers 1981). However, both measures ferent measuringtechniques. of T, are considerablylower than that predicted from an allometric relationship for all birds (42.5”C, McNab 1966). The fasting T, is typical of food-restricted (but METABOLIC RATE AND BODY TEMPERATURE still active) individuals of the Coliiformes (Bartholo- We measured metabolic rates and body temperatures mew and Trost 1970, Prinzinger et al. 1981). Like this for eight birds (five malesand three females).Metabolic order. the low T, of the InaccessibleIsland Rail may ratesgenerally were greatestimmediately after capture, result’ from a low MR. However, the thermal conducl suggestingthat capture and subsequent handling is tance (and possible selective advantage) of the soft, stressful.All periods used to estimate SMR (when 0, hair-like plumage of the InaccessibleIsland Rail war- consumption-waslowest) were during the night, in ac: rants investigating. cordancewith the InaccessibleIsland Rail being largely Plying Rahidae do not have particularly low BMRs diurnal (Hagen 1952, Collar and Stuart 1985, pers. (mean BMR 102.7% of allometric medictions. n = 5. observ.). However, experimental birds were active at range= 83-l 18%;calculated from Kkndeighet al. 1977; times throughout the night, and free-living birds were Brent et al. 1984), sugges.tinga relatjonshjp.between a heard calling up until 22:00 and after 04:OO. low MR and flightlessness.__....,.-- Calder and Dawson (1978) Estimates of SMR from two individuals (19.8 and rep~rt!t!To~wMRsfrom flightlessratites, but could not 30.3 kJ.day-I) were excluded from the analysis be- assesswhether this was a consequenceof flightlessness cause the birds remained active and did not attain a or taxonomic differences.Flightless birds tvnicallv do stable, low MR. Mass-specific0, consumption for the not have a lower metabolic-capacity than-do flying other six birds was 1.03 t 0.35 rn1.g~l.h~’ (range = birds (Brackenbury 1984), and thus it is likely that the 0.66-1.70) giving an estimated SMR of 19.5 f 5.4 k.I. InaccessibleIsland Rail’s island lifestyle has resulted day-’ (range = 12.5-25.5).