ORNITOLOGIA NEOTROPICAL 14: 15–28, 2003© The Neotropical Ornithological Society

POPULATION STATUS OF THE RUDDY-HEADED GOOSE (CHLOEPHAGA RUBIDICEPS) IN TIERRA DEL FUEGO AND

MAINLAND PATAGONIA (CHILE AND ARGENTINA)*

Jesper Madsen1, Ricardo Matus2, Olivia Blank2, Luis Benegas3, Gustavo Mateazzi4 & Daniel E. Blanco5

1National Environmental Research Institute, Department of Arctic Environment, Frederiksborgvej 399, P.O. Box 358, DK-4000 Roskilde, Denmark. E-mail: jm@dmu.dk

2José Robert 0289, Punta Arenas, Chile.3Museo de Ciencias Naturales e Historia, Rio Grande, Tierra del Fuego, Argentina.

4Dpto. Fauna y Ambientes Naturales, Dirección de Proteccion Ambiental, San Martin 1401, (9410) Ushuaia, Tierra del Fuego, Argentina.

5Wetlands International, 25 de Mayo 10° G, (1002) Buenos Aires, Argentina.

Resumen. – Situación de la población del Canquén (Cauquén) Colorado (Chloephaga rubidiceps)en Tierra del Fuego y Patagonia continental (Chile y Argentina). – La población del Canquén (Cau-quén) Colorado (Chloephaga rubidiceps) que cría en Tierra del Fuego (Chile y Argentina) y Chile continental einverna en el sur de Buenos Aires (Argentina) ha sufrido una seria disminución desde los 50s y reciente-mente su tamaño poblacional se calculaba en 300–400 individuos. Durante Diciembre 1999 y Marzo 2000,se realizaron censos y se estudió su distribución durante la cría. Los cauquenes fueron observados casiexclusivamente en vegas (mallines) de la estepa. Para el sector chileno, se utilizó una imagen satelitalLANDSAT-7, con el fin de localizar hábitat potenciales para la especie (cubriendo el 90% de estos). EnDiciembre 1999, se contaron 768 Cauquenes Colorados (634 adultos, 134 crías de 32 nidadas) y en Marzo2000, un total de 793 individuos (adultos y juveniles). El tamaño poblacional post-reproductivo se calculóen 900 individuos, siendo esta una estimación mínima, ya que, existen áreas potenciales aún no estudiadas.Los grupos familiares se concentraron en pocos sitios en Chile continental, los que se caracterizaron por lapresencia de humedales que ofrecían protección contra la predación de zorros. Este tipo de hábitat prácti-camente no existe en Tierra del Fuego, donde los cauquenes probablemente no nidifican por falta de sitiosreproductivos seguros. La cría exitosa estuvo limitada a escasos sectores, donde el éxito reproductivo esaltamente susceptible a las perturbaciones locales. Es imperativo implementar acciones concretas de con-servación para estos “sitios claves”. La introducción del zorro gris patagónico (Dusicyon griseus) en Tierradel Fuego aparece como el principal factor responsable de la disminución y escasa recuperación poblacio-nal. El éxito en la recuperación de esta especie dependerá a largo plazo de la implementación de medidaspara incrementar las oportunidades de reproducción en Tierra del Fuego y el continente, mediante la inun-dación temporal de vegas para reducir el acceso de los zorros a las áreas de nidificación.

Abstract. – The population of the Ruddy-headed Goose (Chloephaga rubidiceps) which breeds in Tierra delFuego (Chile and Argentina) and nearby mainland Chile and winters in the southern Buenos Aires prov-ince in Argentina has seriously declined since the 1950s. Recently, the population was estimated at 300–400individuals. During December 1999–January 2000 and March 2000, surveys of numbers and distribution______________*We dedicate this paper to the memory of Pablo Canevari († 22 March 2000) as a tribute to his strive forthe conservation of wetlands and waterbirds in South America.

15

MADSEN ET AL.

were carried out in the breeding area. Ruddy-headed Geese were almost exclusively associated with wetmarshes (“vegas”) in the steppe zone. A LANDSAT-7 satellite scene covering the Chilean breeding rangewas used to locate potential habitats used by the geese. More than 90% of potential areas were surveyed.During December 1999–January 2000, 768 geese (634 adults and 134 young in 32 broods) were counted,and during March 2000, 793 adults and juveniles. The post-breeding population size was estimated at 900individuals, which was a minimum because potential areas north of the known breeding range remainedunsurveyed. Family groups were concentrated in few areas in mainland Chile. Brood-rearing sites werecharacterized by swamps and/or open water offering retreat in case of predation attempts by foxes. Suchhabitats are generally lacking in Tierra del Fuego. In Tierra del Fuego, the geese probably refrain fromnesting due to lack of predator-free breeding habitat. Successful reproduction is restricted to few areasmaking productivity highly sensitive to local perturbations. The effective protection of the areas of highbreeding success is imperative. The introduction of the grey fox (Dusicyon griseus) to Tierra del Fuego seemsto be the major factor causing the decline and subsequent lack of recovery of the population. In thelonger-term, the restoration of the population will require improved opportunities for breeding in Tierradel Fuego, most realistically achieved by reducing fox access to nesting and brood-rearing areas by tempo-rary flooding of vegas. Accepted 22 August 2002.

Key words: Chloephaga rubidiceps, conservation, Ruddy-headed Goose, grey fox, population size, Patagonia,Tierra del Fuego.

INTRODUCTION but several explanations have been suggested:

The Ruddy-headed Goose (Chloephaga rubidi-ceps) occurs in two populations in the south-ernmost South America: one sedentarypopulation in the Malvinas/Falkland Islandsand one migratory population which breedsin Tierra del Fuego (Chile and Argentina) andnearby mainland Chile, and winters in thesouthern Buenos Aires province, in Argen-tina. Since the 1950s, serious declines in abun-dance of the mainland population have beenreported, although the historical numericalevidence is weak (e.g., Rumboll 1979,Vuilleumier 1994, Canevari 1996). Based oncensuses in the breeding areas in 1985–1988,Vuilleumier (1994) estimated the populationat less than 1000 individuals. Intensified sur-veys carried out in the breeding as well as thewintering range during 1996–1997 resulted ina population estimate of 300–400 individuals,and only few breeding records were made(Gibbons et al. 1998). Hence, the availableinformation suggests that the population is inserious risk of extinction.

The reasons for the decline are unknown,

(1) increased predation of eggs and young bythe grey fox (Dusicyon griseus) (Rumboll 1975,Canevari 1996) which was introduced toTierra del Fuego in the 1950s (Jaksic & Yañez1983); (2) large-scale collection of eggs in thebreeding area during the 1940s–1970s andexcessive shooting in the wintering quarters,both actions resulting from geese consideredan agricultural pest (Weller 1975, Rumboll1979, Martin et al. 1986); (3) the disappea-rance of tall grass in the lowlands due to over-grazing by sheep (Fjeldså 1988); (4)competition with the more numerous goosespecies, the Ashy-headed Goose (C. polioce-phala) and, especially, the larger-sized andhence dominant Upland Goose (C. picta)(Vuilleumier 1994); and (5) sports shooting ofgeese in the breeding grounds and migratorystop-over areas (Canevari 1996, C. Vinci & R.Lini in litt.).

During late November 1999 to late March2000, we carried out a study of Ruddy-headedGeese in Fuego-Patagonia with the aim toupdate the status of the population and todescribe habitat utilization and behavior of

16

STATUS OF RUDDY-HEADED GEESE

the geese, which would hopefully contributeto an understanding of the reasons behindthe poor performance of the population.

In this paper we present an update of thenumbers and distribution of the mainlandpopulation of the Ruddy-headed Goose inthe breeding range, and we describe habitatcharacteristics of brood-rearing areas andfeeding areas of non-breeding birds. Finally,we review the current knowledge of potentialfactors affecting the breeding output of thepopulation, and we suggest options for man-agement and needs for further research.

METHODS

Study area. During the austral spring and sum-mer, the Ruddy-headed Geese occur in thesteppe zone of the northern part of Tierra

del Fuego and on the Chilean mainland ofPatagonia north of the Magellan Strait. Southof Punta Arenas, in Chile, the Ruddy-headedGeese also occur in grassland clearings inriver mouths in the forest/steppe ecotone.The undulating lowland steppe landscape isdissected by small rivers and streams formingsmall valleys and dotted, in the lowest partsand in depressions, with ponds, lagoons andtemporary flooded wetlands locally called“vegas” or “mallines” (Collantes & Faggi1999). Tussock-grass vegetation of Festucagracillima dominates the steppes, with varyingdegree of coverage by shrub plants, the mostabundant of which is Chiliotrichum diffusum,and patches of heathland with Empetrumrubrum. Vegas are common in lowlands andvalleys, with abundant grasses dominated byDeschampsia antarctica, Hordeum halophilum and

FIG. 1. Study area in mainland Patagonia (Chile) and on Tierra del Fuego (Chile and Argentina). Majorroads are shown by lines and the Argentina-Chile border by interrupted line.

17

MADSEN ET AL.

Festuca magellanica, and forbs like Juncusscheuchzerioides, Carex vallis-pulchrae and Calthasagittata (Collantes & Faggi 1999). In somevegas, re-seeding has taken place with alienPoa spp. and other grasses (Moore 1983). Invalleys with poor drainage and slowly flowingwaters, shallow swamps with submerged veg-etation dominated by Hippuris vulgaris areformed. Along ponds, swamps and streams, awet Juncaceae dominated community forms atransition zone with the Festuca-Poa spp. dom-inated grassland.

Surveys. Surveys were carried out by theauthors in two periods: 25 November to 13December 1999, with additional observa-tions during 28–30 January 2000, and 14–24March 2000. The surveys covered the steppezone of Tierra del Fuego (north of latitude54ºS) and the Chilean mainland from SanJuan in the south to Punta Delgada andCiaike-San Jorge (south of latitude 52°S) inthe north (Fig. 1). The aim of the first surveywas to cover the distribution and numbers ofRuddy-headed Geese during the breedingperiod, while the second survey was carriedout to describe post-breeding distribution andnumbers.

The surveys focused on all areas knownfrom the censuses in 1996–97 (Gibbons et al.1998) and 1992–98 (Argentina only) (Benegas1997), with a bigger effort done in the SanJuan, San Gregorio, Springhill-Cerro Som-brero areas in Chile, and at Estancia Cullen,Estancia Los Flamencos and TF1-Las Viole-tas in Argentina (Fig. 1).

To make a more complete coverage of theChilean part of the range, a scene from theLANDSAT-7 remote sensing satellite takenon 18 September 1999 was obtained from theUS Geological Surveys (data granuleSC:L70RWRS.002:2000230269 LANDSAT-7Level-OR WRS-SCENE V002, hhtp://edcwww.cr.usgs.gov/landdaac) to locate hith-erto unknown potential sites. On 18 Septem-

ber 1999, the study area was largely free ofclouds, except for the hills east of a line fromCerro Sombrero to Porvenir in Tierra delFuego. In the field we took several positionswith a handheld GPS (Garmin 12), for subse-quent correction of the satellite imagery togeographical coordinates.

Based on the satellite imagery, classifica-tion of vegetation zones was carried out byvisual inspection of areas with differentreflections. From 18 September to earlyDecember 1999, some of the vegas becamedrier. As a result, the satellite imagery showedthe maximal extent of wet areas. Areas of wet-lands (swamps, wet and dry vegas, respec-tively) were digitized for the San Gregorioarea (580 km2) to obtain a relative distributionof different types of habitat.

During the March 2000 survey and withthe purpose of locating hitherto unknownpotential sites, a vegetation map was used forthe Argentine side of the island (M. B. Collan-tes unpubl.).

Counts of geese were made using binocu-lars and telescopes (20–60 x zoom). EachRuddy-headed Goose locality was positionedwith the help of GPS. In the Chilean range allbirds were plotted on a detailed print of thesatellite imagery.

In randomly selected areas with a goodoverview, we described habitat use by Ruddy-headed Geese, assigning to each individualthat was observed while feeding a given habi-tat type and the estimated distance to thenearest body of open water (river, stream,pond, lagoon, coast) (Madsen et al. in prep.).

Searches for hitherto unknown sites weretargeted using the vegetation characterizationfrom the satellite imagery for Chile and thevegetation map for Argentina coupled withour observational description of habitat useby Ruddy-headed Geese. Within the knownrange, the vast majority of potential sites werevisited. To check potentially suitable areasoutside the known range of the Ruddy-hea-

18

STATUS OF RUDDY-HEADED GEESE

ded Goose, trips were made to areas south ofRio Grande, northwest of San Juan, and toRio Verde northwest of Punta Arenas.

Brood records. During December 1999 (andsupplemented in January 2000), brood sizesof goose families were recorded, and theapproximate age (in weeks) of the young wasestimated based on size and plumage devel-opment, using a classification developed fornorthern Hemisphere true geese (Anserini)(Owen 1980). Upland Goose young developflank feathers at an age of four weeks andfledge at an age of about 70 days (Summers1983). As this development is similar to that

of true geese of the same size, we can safelyassume a similar timing in Ruddy-headedGeese. From the estimated age, the date offirst egg laying was then calculated, assumingan average clutch size of five eggs, an incuba-tion period of 30 days and an interval of twodays between each egg laid (Summers 1983,Summers & McAdam 1993). To compare ourdata with those from previous years, weincluded records of broods with age estima-tion from 1992, 1997, and 1998 (R. Matusunpubl., Benegas 1997).

During March 2000, we attempted torecord the number of juveniles in the flocksof Ruddy-headed Geese.

FIG. 2. Habitat use by non-breeding individuals and family groups of Ruddy-headed Geese (Chloephagarubidiceps) expressed as the number of groups (group size ranging from one to six individuals) observed inrelation to the distance of the group from open water.

19

MADSEN ET AL.

Fox abundance. To obtain a crude index of foxabundance, we recorded the number of foxesand fox dens/cubs observed during day tripsin December 1999. The daily sum of thenumber of adult foxes sighted was used tocompare relative abundances in Tierra delFuego and the Chilean mainland.

RESULTS

Habitat use. During December 1999, non-breeding Ruddy-headed Geese (single individ-uals, pairs, groups of pairs) foraged almostexclusively in the wet zones of vegas (93% ofindividuals); only 4% were observed in dryvegas and 3% in the Festuca steppe. Sixty-eightpercent foraged within a distance of fivemeters from open water (Fig. 2). Fifty-eightpercent were observed in relation to streams,32% in relation to ponds.

Families were observed exclusively in thewet vegas, and 84% foraged within fivemetres of open water. Eighty-seven percentwere observed in relation to ponds/swamps/coasts and 13% in relation to streams.

During March 2000 in Tierra del Fuego,75% of Ruddy-headed Geese (single individu-als, pairs, groups of pairs) were foraging inwet vegas. Ten percent were sighted in dryvegas and another 15% were registered forag-ing in the Festuca steppe and in sheep grazedfens very close to farm buildings.

In the San Gregorio area (total 580 km2),swamps made up 1% of the total area; wetvegas 2%, and dry vegas 3%. The remaining94% were covered by various types of drysteppe vegetation and barren ground.

Coverage of surveys. Based on our gained know-ledge of habitat use, we searched for geese in

TABLE 1. Counts of Ruddy-headed Geese (Chloephaga rubidiceps) and Upland Geese (C. picta) in mainlandPatagonia (Chile) and in Tierra del Fuego (Chile and Argentina), 25 November–13 December 1999, withadditional data from January 2000. The number of non-breeding Upland Geese was not systematicallycounted and therefore not included in the table.

Ruddy-headed Goose Upland Goose

No. adults No. broods No. youngs No. broods No. youngsMainland Chile

San GregorioOazy Harbour-Puesto Panteón-Ciaikea

San Juanb

Punta DelgadaSubtotal

Chilean part of Tierra del FuegoSpringhill-Cerro SombreroBahia InutilSubtotal

Argentine part of Tierra del FuegoTF1-Ea. Flamencos-Los ChorillosEa. Cullen-Pampa BetaSubtotal

Total

222508817377

2233

226

201131634

2505030

202

00032

1070250

132

202

000

134

2102124

2

2

00026

7905589

7

7

00096

aSurveyed 30 January 2000 (R. Matus).bThe number of broods and young as surveyed 28 January 2000 (R. Matus).

20

STATUS OF RUDDY-HEADED GEESE

all wet vegas that could clearly be identifiedon the satellite imagery. We estimate thatmore than 90% of all swamps and wet vegaswere visited in the Chilean range. Thus, com-pared with the 1996–97 survey, the 1999 cov-erage was improved considerably. In SanGregorio, approximately 20% of the area wascovered in 1996–97 compared to 1999; inSpringhill-Cerro Sombrero 40%, while in SanJuan 100%. The Punta Delgada and Ciaike-San Jorge areas were not visited in 1996–97.

For the Argentine side of Tierra delFuego the coverage was not assessed indetail, but in view of good local knowledge(Benegas 1997), we assume that coverage wasclose to complete.

Numbers and distribution. During December1999–January 2000, the number of adultRuddy-headed Geese counted was 570 non-breeding birds and 32 pairs with young, total-

ing 634 adults. A total of 134 young was re-gistered, so the total number counted was768 Ruddy-headed Geese (Table 1).

During March 2000, the number of adultRuddy-headed Geese counted was 779; 14young were registered in mainland Chile andTierra del Fuego, yielding a total of 793 indi-viduals (Table 2). In the same month, the dis-covery of new localities with Ruddy-headedGeese in northern Chilean Tierra del Fuego(along the road from Cerro Sombrero to SanSebastián and to the Cullen area) and northof San Gregorio (San Jorge) in mainlandChile yielded 108 additional individuals.Assuming that these geese were also there inDecember 1999, the difference in numberscounted in December 1999 and March 2000was 83 individuals. The difference in num-bers of young/juveniles between the two sur-veys is probably primarily due to difficultiesin field identification of juveniles in March.

TABLE 2. Counts of Ruddy-headed Geese (Chloephaga rubidiceps) in mainland Patagonia (Chile) and inTierra del Fuego (Chile and Argentina), 14–24 March 2000.

No. adults No. broods No. youngsMainland Chile

San GregorioOazy Harbour-Puesto Panteón-CiaikeSan Jorgea

San JuanPunta DelgadaSubtotal

Chilean part of Tierra del FuegoSpringhill-Cerro SombreroCullen areaa

Road Cerro Sombrero-San SebastiánBahia Inutilb

SubtotalArgentine part of Tierra del Fuego

TF1-Ea. Flamencos-Los ChorillosEa. Cullen-Pampa BetaSubtotal

Total

246437330329

3066041-

407

41243779

21

20

000-0

0006

81

5014

000-0

00014

aNot surveyed in December 1999–January 2000.bNot surveyed in March 2000.

21

MADSEN ET AL.

The December 1999–January 2000 censusshowed that the majority of Ruddy-headedGeese was concentrated around San Gregorio(49% of total numbers), Springhill-CerroSombrero (29%) and San Juan (15%) (Table1; Fig. 3). During March 2000, the majoritywas concentrated in the Chilean part of Tierradel Fuego (51%) and at San Gregorio (39%),whereas only 1% was registered at San Juan(Table 2; Fig. 4).

During December 1999–January 2000,twenty-five families were observed in the SanGregorio area, five in San Juan and two inEstancia Las Vegas south of Springhill.Hence, in most areas in Tierra del Fuego andat Punta Delgada there was no evidence ofbreeding activity. Lack of breeding was con-firmed by the observation of generally passivebehavior of pairs; in San Juan, most Ruddy-

headed Geese were flocking with UplandGeese and Ashy-headed Geese.

For comparison, the number of broods ofUpland Geese observed during December1999–January 2000 is included in Table 1.This species showed similar geographicalvariation in breeding success as the Ruddy-headed Geese.

Brood sizes and time of egg-laying. The averagebrood size in the 32 Ruddy-headed Goosebroods observed in December 1999–January2000 was 4.4 ± 2.8 (mean ± SD) (range 1–12).The estimation of time of egg-layingshows that the first pairs started duringmid September, with a peak during thefirst two weeks of October (Fig. 5).Observations from earlier years showsimilar distribution. Late egg-laying dates

FIG. 3. Distribution of Ruddy-headed Geese (Chloephaga rubidiceps) during the survey in December 1999,supplemented by observations in January 2000.

22

STATUS OF RUDDY-HEADED GEESE

are likely to represent replacement clutches.

Abundance of foxes. Grey foxes were frequentlyregistered in both Tierra del Fuego and main-land Chile. During December 1999, the aver-age daily number of foxes for the three day’strip in the Argentine part of Tierra del Fuegowas 10; in the Chilean part of Tierra delFuego, six foxes (three day trips) and in theChilean mainland eight foxes (five day trips).Hence, there appeared not to be major differ-ences in abundance between areas.

DISCUSSION

Population estimate. The numbers of 768 and793 Ruddy-headed Geese observed duringthe two surveys in December 1999–January2000 and March 2000 represent a minimum

population size for two reasons. Firstly, weprobably missed some birds within theknown range due to the dispersion of birdsin small groups in correlation with thedistribution of suitable habitat in somepart of the surveyed range. Secondly,we failed to locate all sites outside the knownrange. With an approximate coverage of 90%of all suitable habitat within the knownrange, the estimated summer population isapproximately 900 individuals includingyoung birds. This remains a minimum popu-lation estimate, because we may still miss sitesoutside the known breeding range, for exam-ple in the river catchment areas north of SanGregorio (Chile) and in the southern SantaCruz province (Argentina) (S. Imberti pers.com.).

The population estimate based on the

FIG. 4. Distribution of Ruddy-headed Geese (Chloephaga rubidiceps) during the survey in March 2000.

23

MADSEN ET AL.

1999–2000 survey almost triples the previousestimate of 300–400 individuals based on sur-veys in breeding areas in 1996–97 and winter-ing areas in 1996–97 and 1999 (Gibbons et al.1998, Blanco et al. unpubl.). The apparentincrease represents probably improvedcoverage of the breeding range and not a gen-uine population increase, since numbers inbreeding areas with good coverage in 1996–97 were similar to those in 1999–2000 (seealso Vuilleumier 1994 for 1985–1988 and1993).

The question of where the additional indi-viduals spend the winter is open. From obser-vations in Tierra del Fuego and mainlandChile, it is documented that the majority ofbirds leave in late April/early May and reap-pear in late August. However, few individualshave been observed in San Gregorio in July(R. Matus unpubl.). Possibly, more birds stay

in the newly discovered areas. Furthermore,not necessarily all geese spend the winter inthe Buenos Aires province but somewhereelse, not yet discovered.

Habitat use. Our observations of habitat useshowed that Ruddy-headed Geese forage inthe narrow wet zone of vegas. During brood-rearing, the zone which is used is even nar-rower, probably because they need the prox-imity to open water as a refuge for the youngin case of attempts of predation by terrestrialpredators. In three cases where we observedGrey foxes approaching Ruddy-headedGoose families feeding close to a pond, thefamilies ran to the water edge and swam tothe other side of the pond (Madsen et al. inprep.).

Characteristics of brood-rearing areas. Observa-

0

5

10

15

20

25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Weeks after 1 September

Num

ber o

f clu

tche

s

1992199719981999

FIG. 5. Timing of laying of first egg of Ruddy-headed Geese (Chloephaga rubidiceps), extrapolated from theage of young in broods observed during 1992–1999.

24

STATUS OF RUDDY-HEADED GEESE

tions of behavior and spacing of familiesshowed that parental Ruddy-headed Geesedefend a territory, not only intraspecificallybut also against Upland Geese. Hence, in astudy area in San Gregorio families of Uplandand Ruddy-headed Geese partitioned thesuitable marshes with close access to openwater (Madsen et al. in prep.). Summers(1983) showed that Upland Geese maintaineda territory from the time of pre-nesting tofledging of the young, although some pairsmoved from nesting to nearby post-hatchingareas. Assuming that Ruddy-headed Geesemaintain the same territory throughoutbreeding, or only move a short distancebetween nesting and post-hatching territories,we can compare areas with successful hatch-ing with areas without hatching success.Brood-rearing areas (and assumed nestingareas) are characterized by wet vegas withponds, swamps or estuaries with islets thatprovide protection against fox predation dur-ing nesting and brood-rearing. In most sitesin Tierra del Fuego, these habitat features arelacking (except for Estancia Las Vegas) andthere, geese stayed along streams or edges oflakes (with no islets). Some of the pairsshowed territorial behavior (aggression andextreme alertness by males) but with no signof nesting activity. It is possible that pairs duenot attempt to nest at all because of highpredator abundance. In Brent Geese (Brantabernicla) it is known that in years with highabundance of Arctic foxes (Alopex lagopus) inthe breeding areas, the geese do give up theattempt to nest, except for islets free fromfoxes (Madsen et al. 1992, Spaans et al. 1998).

In order to reveal causal relationshipsaffecting nesting success, there is an immi-nent need for studies of nesting behavior andhabitat use during late September throughOctober.

Causes of decline and lack of recovery. Grey foxesare abundant in the mainland as well as in

Tierra del Fuego and successful breeding byRuddy-headed Geese seems to be achievedonly in sites that offer safe brood-rearing(and probably nesting) areas. Before theintroduction of the Grey fox to Tierra delFuego (Jaksic & Yañez 1983), geese wereprobably not so dependent on sites with suchcharacteristics, because no natural terrestrialpredator existed in the steppe zone (exceptfor aboriginal Indians). It is therefore highlylikely that the introduction of foxes has beenthe main cause of the decline and subsequentlack of recovery of the Ruddy-headed Goose.The observation that Upland Geese showedthe same geographical difference in breedingsuccess as Ruddy-headed Geese indicates thatthe fox predation problem may be more gen-eral.

With regard to the other suggested causesof decline, they may very well have been con-tributing factors. The large-scale egg collec-tion, which primarily took place in Argentina,probably aggravated the reduction. However,despite the stop of egg collection in the1970s, the population did not recover, sug-gesting that one or more other factors werecontrolling the population size (Vuilleumier1994). The disappearance of the tall-grasssteppe vegetation due to sheep grazing, espe-cially along rivers and around lakes, may havecaused losses of nest cover (Fjeldså 1988).On the other hand, sheep grazing and ranch-ing have facilitated goose usage partlybecause sheep have opened much habitat andpartly because the introduction of pasturegrasses such as Poa pratensis has lead toimproved food quality (Summers & Grieve1982).

Interspecific competition for foodresources is probably not a major problemoutside brood-rearing areas. We recordedmany suitable feeding areas which were notused or only extensively used by eitherRuddy-headed Geese or Upland Geese.However, our observation suggest that at

25

MADSEN ET AL.

present the available suitable area for brood-rearing (and probably nesting) is restricted tothe wet vegas with swamps and ponds, prima-rily found in San Gregorio. At one study sitein San Gregorio, the marsh was partitionedbetween Upland and Ruddy-headed Geeseand territories were maintained by interspe-cific aggression (Madsen et al. in prep.). In thiscritical part of the annual cycle, interspecificcompetition for space may therefore be sig-nificant locally, potentially controlling thenumber of successful breeding pairs.

There exists no concrete informationabout past and present hunting pressure onRuddy-headed Geese, but indiscriminateshooting and persecution in the winteringquarters are likely to affect the species (Cane-vari 1996, Blanco et al. unpubl.). We can onlyassume that at the present low population sizeany shooting will add significantly to naturalmortality. Therefore, even the kill of tens ofgeese is likely to have a negative effect onpopulation size.

‘All eggs in one basket’. San Gregorio appears tobe the major area where substantial reproduc-tion takes place, while in Tierra del Fuego,reproduction is only sporadic. Despite severalyears of intensive monitoring since 1992, onlytwo broods have been recorded in the Argen-tine sector of Tierra del Fuego (Benegas 1997and unpubl.). We have no information aboutdispersal patterns within the population, butit is realistic to suggest that numbers in Tierradel Fuego can be maintained only with immi-gration of birds from the mainland. Hence,this may represent a case of a source-sinkrelationship, and the short-term future popu-lation development will depend on continuedsuccessful breeding in especially San Grego-rio.

Management implications. The major anthropo-genic threats to the Ruddy-headed Geese dur-ing the breeding period at present appears to

be disturbance through human activity in thenesting and brood-rearing areas. Potentiallydisturbing activities include ranching, pipelinemaintenance work, recreational activitiesand sport shooting. Because the mostsensitive areas are so small, it appears thatthere is scope for taking effective protectionmeasures to mitigate these negative effects bythe creation of reserves with access regula-tions.

However, to restore the population to thelevels it had in the 1940s and 1950s, ourobservations suggest that it will be necessaryto find means to reduce fox predation of eggsand young. This can be achieved either bycontrol of foxes or habitat managementwhich will reduce fox access. The former maybe possible in some specific areas; however,the control of foxes on large scale, e.g., inTierra del Fuego, will require a massive cam-paign, to be coordinated between Chile andArgentina. Habitat management thereforeseems to be the most realistic way forward forthe immediate future. Reduced fox access canbe achieved by temporary flooding of vegasby means of stems or dams, which will createislets and swamps with no fox access. Waterlevels should be maintained high at least frombefore egg-laying (preferably from the estab-lishment of territories) to fledging of young,i.e., from approximately late August to lateDecember). This type of water managementis actually actively practiced in some farmareas to maintain wet and productive grass-lands for sheep, for example in Estancia LasVegas south of Springhill, the only area withsuccessfully breeding Ruddy-headed Geese inTierra del Fuego during this study. Severalother areas have similar potential. To becomeoperational, such habitat modification workwill require active collaboration with land-owners who need to understand potentialconservation benefits as well as economicincentives in relation to sheep grazing oppor-tunities.

26

STATUS OF RUDDY-HEADED GEESE

CONCLUSION

The Fuego-Patagonian population of Ruddy-headed Goose was estimated at approxi-mately 900 individuals. Successful reproduc-tion is restricted to few areas making theproductivity highly sensitive to local pertur-bations. Effective protection of the areas withgood breeding success in San Gregorio andSan Juan (mainland Chile) is imperative. Res-toration of the population will requireimproved opportunities for breeding inTierra del Fuego, most realistically achievedby reducing fox access to nesting and brood-rearing areas by temporary flooding of vegas.

ACKNOWLEDGMENTS

This study was part of a project of WetlandsInternational and was carried out under theauspices of the Convention on MigratorySpecies (CMS); Project No. 1200-98-53: Con-servation Action for the Ruddy-headedGoose in the austral region of Argentina andChile. The project was coordinated by Wet-lands International and was conducted bylocal researchers from Chile and Argentina,with the support of the Museo de CienciasNaturales e Historia de Río Grande and theDirección de Protección Ambiental of Tierradel Fuego, in Argentina. We are grateful toCMS and the organizations and specially toPablo Canevari (Wetlands International), forhis advice and important support. Ib KragPetersen and Geoff Groom (National Envi-ronmental Research Institute, Denmark) arethanked for technical support with regard tosatellite imagery handling. Francisco Salesand Luis Leaniz are thanked for technicalassistance. Marta Collantes (CEVEG-CONICET, Argentina) made available theunpublished vegetation map of northernTierra del Fuego (Argentina). Comments byFrancois Vuilleumier and Jon Fjeldsåimproved the manuscript.

REFERENCES

Benegas, L. G. 1997. El Cauquen Cabeza Coloradaen Tierra del Fuego (Arg.). Museo de CienciasNaturales e Historia, Municipalidad de RioGrande, Rio Grande, Argentina.

Canevari, P. 1996. The austral geese (Chloephagaspp.) of southern Argentina and Chile: a reviewof its current status. Gibier Faune Sauvage,Game Wildl. Sci. 13: 355–366.

Collantes, M. B., & A. M. Faggi. 1999. Loshumedales del Sur de Sudamérica. Pp. 15–25 inMalvárez, A. I. (ed.). Tópicos sobre humedalessubtropicales y templados de Sudamérica.UNESCO-ORCYT-MAB, Montevideo, Uru-guay.

Fjeldså, J. 1988. Status of birds of steppe habitatsof the Andean zone and Patagonia. ICBPTech. Bull. 7: 81–95.

Gibbons, J. E., R. Matus, Y. A.Vilina, D. E. Blanco,S. Zalba, & C. Belenguer. 1998. Desarrollo deun plan de conservación para el CauquénCabeza Colorada (Chloephaga rubidiceps), en laregión austral de Argentina y Chile. Internalreport. Wetlands International, Buenos Aires,Argentina and CONAF, Santiago, Chile.

Humphrey, P. S., D. Bridge, P. W. Reynolds, & R. T.Peterson. 1970. Birds of Isla Grande (Tierradel Fuego). Smithsonian Institution, Washing-ton, D.C..

Jaksic, F., & J. L.Yañez. 1983. Rabbit and fox intro-ductions in Tierra del Fuego: history andassessment of the attempts at biological con-trol of the rabbit infestation. Biol. Conserv. 26:367–374.

Madsen, J., T. Bregnballe, & A.. Hastrup. 1992.Impact of the arctic fox Alopex lagopus on nest-ing success of geese in Southeast Svalbard,1989. Polar Res. 11: 35–39.

Martin, S. I., N. Tracanna, & R. Summers. 1986.Distribution and habitat use of Sheldgeesepopulations wintering in Buenos Aires Prov-ince, Argentina. Wildfowl 37: 55–62.

Moore, D. M. 1983. Flora of Tierra del Fuego.Anthony Nelson, Oswestry, UK.

Owen, M. 1980. Wild geese of the world. B. T.Batsford, London, UK.

Rumboll, M. A. E. 1975. Notas sobre Anseri-formes: el Cauquén de Cabeza Colorado

27

MADSEN ET AL.

(Chloephaga rubidiceps), una nota de alarma. Hor-nero 11: 315–316.

Rumboll, M. A. E. 1979. El estado actual de Choe-phaga rubidiceps. Act. Zool. Lilloana 34: 153–154.

Spaans, B., H. Blijleven, I. U. Popov, M. E. Rykhli-kova, & B. S. Ebbinge. 1998. Dark-belliedBrent Geese Branta bernicla bernicla foregobreeding when arctic foxes Alopex lagopus arepresent during nest initiation. Ardea 86: 11–20.

Summers, R. W. 1983. The life cycle of the UplandGoose Chloephaga picta in the Falkland Islands.Ibis 125: 524–544.

Summers, R. W., & J. H. McAdam. 1993. TheUpland Goose. Bluntisham Books, Hunting-

don, UK.Summers, R. W., & A. Grieve. 1982. Diet, feeding

behavior and food intake of the Upland Goose(Chloephaga picta) and Ruddy-headed Goose (C.rubidiceps) in the Falkland Islands. J. Appl. Ecol.19: 783–804.

Vuilleumier, F. 1994. Status of the Ruddy-headedGoose Chloephaga rubidiceps (Aves, Anatidae): aspecies in serious danger of extinction inFuego-Patagonia. Rev. Chil. Hist. Nat. 67: 341–349.

Weller, M. W. 1975. Habitat selection by waterfowlof Argentine Isla Grande. Wilson Bull. 87: 83–90.

28