632 ShortCommunications [Auk, Vol. 103

Recovery of the EndangeredMauritius Kestrel from an Extreme Population Bottleneck

STANLEY A. TEMPLE Departmentof WildlifeEcology, University of Wisconsin,Madison, Wisconsin 53706 USA

The MauritiusKestrel (Falco punctatus) is a severely that a circular area of about 28 ha approximatelyde- endangeredspecies (King 1981)whose entire popu- scribesa kestrel'sterritory size. lation in the wild reached a low of only 2 pairs in By systematicallyinspecting the remaining habitat 1974 (Temple 1974, 1977). Temple (1978) speculated for all potential nesting cliffs (basedon characteris- that a spontaneous "tradition-shift" (sensuWilson tics of 1974-1977 nests) and plotting nonoverlap- 1975: 13) in nest-siteselection may have been initi- ping, circular 28-ha territories around each usable ated in the kestrelpopulation in 1974.This shift from site, I estimatedthat only about 18 cliff-nestingpairs nesting primarily in tree cavitiesto nesting in holes could be accommodated in the available habitat. If and nicheson cliffs was initiated by the single suc- nonbreeders,whose 120-hahome rangesoverlapped cessfullybreeding pair in 1974 and seemsto have by 50%, filled in all areasnot occupiedby breeders, permitted the kestrels to reproduce more successful- the carryingcapacity of the remaining habitat around ly. They had previously sufferedheavy nest losses the Black River Gorgeswould be approximately 66 (mostlikely to introducedmacaques, Macaca fascicu- kestrels. laris).All young kestrelsfledged since 1974 were from Between 1973 and 1977 the size of the kestrel pop- sites on cliffs and subsequentlyadopted cliffs for ulation was monitored closely. Becauseof the small nesting when they establishedtheir own breeding population size, total countsof individuals were pos- territories. There are no records of tree nests since sible. I counted 4 birds at the end of the 1973 breed- 1976 (Collar and Stuart 1985). The overall improve- ing season,7 in 1974,7 in 1975,and, with the help ment in reproductive successthat seemsto have re- of S. D. McKelvey, 11 birds in 1976. In 1977, P. Tre- suited from cliff-nestinghas allowed the kestrelpop- frey (pets. comm.) counted 15 birds. Between 1978 ulation to experience a recovery from the severe and 1983no systematiccounts of the total population bottleneck of 1974. were made, but in 1984 a thorough censusof the There hasbeen much interestand concernamong population resulted in an estimatedpopulation size conservationbiologists over the ability of extremely of about 50 kestrels(N. C. Fox pets. comm.). depleted populationsto rebound from major bottle- Using only the population data from 1973 to 1977 necks(e.g. Schonewaid-Coxet ai. 1983),but there has and the methodsdescribed by Caughley (1977), I cal- been relatively little documentationof the fates of culatedan annual populationgrowth rate (r) of 0.35 natural populationsduring the coursesof their re- for the kestrel. Given this population parameter and coveries.Here, I discussthe recovery of the Mauri- an estimatedcarrying capacityof 66 birds, it was pos- tius Kestrel population from a major bottleneckand sible to constructa population-growthmodel that show that its subsequentgrowth seemsto have fol- predictedthe population'ssize in eachyear since1973. lowed a pattern that approximatesa logistic growth Similar approacheshave been used to model the re- Curve. coveriesof other endangeredspecies (e.g. Miller et The Mauritius Kestrel population is now restricted al. 1974,Barclay and Cade 1983). I used the logistic to a potential range of approximately4,000 ha of model of population growth, as describedby Tanner remnant forest habitat surroundingthe Black River (1978), Gorgesin the southwestof the island of Mauritius K N, in the Indian Ocean (Temple 1977). The ecological 1 + Ae-" basisfor this habitat restrictionis the kestrel'shighly where Nt is the population size in year t, K is the specializedforaging ecology, which allowsit to hunt carrying capacity of the available habitat, A is the successfullyonly in certain types of forests(Temple value [(K - No)/No], e is the base of natural loga- 1986). rithms, r is the population growth rate, and t is the By plotting repeated sightings of the 7 individ- number of years since 1974 (Fig. 1). Fox's 1984 esti- ually recognizablekestrels that comprisedthe wild mateof 50 birdssuggests that the population'sgrowth population from 1973 to 1975 and creating everted is, indeed, following a pattern that approximatesthe polygonsenclosing sighting points, I was able to de- logisticgrowth model generatedfrom the 1973-1977 termine the approximatesizes of kestrel home ranges population data. and territories. Three pairs of kestrels had annual The wild population of the Mauritius Kestrel has home ranges that averaged approximately 120 ha survived an extreme reduction in population size and (range97-142 ha), and adjacentpairs appearedto tol- is recovering its numbers at a rate that suggestsit erate about a 50% overlap (range 42-70%) in their will reach the carrying capacityof its habitat within home ranges. All observedterritorial disputesoc- 20 yr of emergingfrom a majorbottleneck. The wild curred within a 300-m radiusof nest sites,suggesting population is now substantiallyinbred, with an in- July1986] ShortCommunications 633

K= 66 INDIVIDUALS 3-40 in Bird conservation1 (S. A. Temple, Ed.). a-datafrom Temple --'"•'""'ø'"-'ø•¸ Madison, Univ. Wisconsin Press. b--datafromTemple &McKelvey CAUGHLEY,G. 1977. Analysisof vertebratepopula- c-data fromTrefrey o tions. Chichester,England, J. Wiley and Sons. COLLAR,N.J., • S. N. STUART. 1985. Threatened birds of Africa and related islands. ICBP/IUCN

• 3c growth calculated as: Red Data Book, part 1, 3rd ed. Cambridge, En- gland, Intern. Council for Bird Preserv.and In- tern. Union for Conserv. of Nature and Natural Resources. •o •o COO?ER,J.E. 1979. An oviduct adenocarcinoma in a Mauritius Kestrel (Falco punctatus).Avian Pa- thol. 8: 187-191. --, C. G. JONES,& A. W. OWADALLY. 1981. Mor- FJ.õ.1. Observed and predJ.cted sJ.ze of the Mau- bidity and mortality in the Mauritius Kestrel.Pp. rJ.tJ. us Kestrel populatJ.on followinõ a severe popula- 31-35 in Recentadvances in the study of raptor tion bottleneck in 1974. diseases(J.E. Cooperand A. G. Greenwood,Eds.). Keighley, England, Chiron Publ. breeding coefficient, F (Wright 1922), of at least KINe, W. B. 1981. Endangeredbirds of the world: 0.125, where F = 1/2N• and N, was the number of the ICBP Red Data Book. Washington, D.C., breedersin the 1974 bottleneckpopulation. Inbreed- Smithsonian Inst. Press. ing depressionis commonlyencountered in popu- LOVEJOY,T. E. 1978. Genetic aspectsof dwindling lations that have passedthrough sucha severebot- populations.Pp. 275-280 in Endangeredbirds: tleneck (Schonewald-Cox et al. 1983), and Cooper management techniques for preserving threat- (1979) and Cooper et al. (1981) have wondered ened species(S. A. Temple, Ed.). Madison, Univ. whether mortality ratesin a smallcaptive population Wisconsin Press. of Mauritius Kestrels may have been elevated be- MILLER, R. S., D. B. BOTKIN, & R. MENDELSSOHN. 1974. causeof inbreeding depression. The Whooping Crane (Grus americana)popula- At presentany reductionthat may have occurred tion of North America. Biol. Conserv. 6: 106-111. in the fitness of individuals in the wild population SCHONEWALD-Cox, C. M., S. M. CHAMBERS,B. MAC- apparentlyhas not interferedwith a relativelyrobust BRYDE, & W. L. THOMAS. 1983. Genetics and rate of population growth. This growth is an en- conservation.Menlo Park, California, Benjamin/ couragingindication that the Mauritius Kestrelmay Cummings Publ. Co. have survived a severepopulation bottleneck with- TANNER•J. T. 1978. Guide to the study of animal out acquiringthe potentiallyheavy genetic loads that populations. Knoxville, Univ. TennesseePress. can lead to extinction in such situations. Lovejoy TEMPLE,S.A. 1974. Wildlife in Mauritius today. Oryx (1978) speculatedthat populations that have sur- 13: 584-591. vived previousbottlenecks may be in better condi- 1977. The status and conservation of en- tion to survive subsequentones becausethey have demic kestrels on Indian Ocean islands. Pp. 74- eliminatedmany deleterious alleles during eachsuc- 92 in Proceedingsof the ICBP World Conference cessfulpassage through a bottleneck.Species like the on Birdsof Prey (R. D. Chancellor,Ed.). London, Mauritius Kestrel that colonized remote oceanic is- Intern. Council for Bird Preservation. lands almost certainly have passedthrough bottle- 1978. Manipulating behavioral patterns of necks in the course of their evolution on the island. endangeredbirds: a potential managementtech- Nonetheless,with a potentialwild populationof only nique. Pp. 435-446 in Endangered birds: man- 66 individuals, the Mauritius Kestrel will remain an agement techniques for preserving threatened endangeredspecies until conservationmeasures, such species(S. A. Temple, Ed.). Madison, Univ. Wis- as translocations of birds to areas of new habitat on consin Press. Mauritius or elsewhere (Temple 1981), allow the 1981. Applied island biogeographyand the population to reach a larger size. conservationof endangered island birds of the My work on the MauritiusKestrel has been funded Indian Ocean. Biol. Conserv. 20: 147-161. by the International Council for Bird Preservation, 1986. Foraging ecology of the Mauritius the World Wildlife Fund (U.S. Appeal), and the New Kestrel (Falcopunctatus). Biotropica in press. York ZoologicalSociety. The Mauritius Ministry of WILSON,E. O. 1975. Sociobiology:the new synthe- Agriculture and Natural Resourcesfacilitated my sis. Cambridge, Massachusetts,Belknap Press. work on Mauritius. WRIGHT,S. 1922. Coefficientsof inbreeding and re- LITERATURE CITED lationship. Amer. Natur. 56: 330-338.

BARCLAY,J., & T. CADE. 1983. Restoration of the Per- Received31 fanuary 1986, accepted18 February1986. egrine Falcon in the eastern United States. Pp.'