Relationships between home-range size, sex and season with reference to the mating system of the Houbara Bustard Chlamydotis undulata undulata

Ibis (2004), 146, 314 322 Blackwell Publishing, Ltd. Relationships between home-range size, sex and season with reference to the mating system of the Houbara Bustard Chlamydotis undulata undulata YVES HINGRAT, 1,2 MICHEL SAINT JALME, 2 * FRÉDÉRIC YSNEL, 3 FREDÉRIC LACROIX, 1 JOHN SEABURY 1 & PIERRICK RAUTUREAU 1 1 Emirates Center for Wildlife Propagation, Department of Ecology, PO Box 47, 33250 Missour, Morocco 2 Muséum National d Histoire Naturelle, Département Ecologie et Gestion de la Biodiversité, UMR 5173 MNHN-CNRS Conservation des Espèces, Restauration et Suivi des Populations, Parc Zoologique de Clères, 76690 Clères, France 3 Musée Universitaire de Beaulieu, UMR 6553, Muséologie et Biodiversité, Université de Rennes 1, France This paper presents observations of Houbara Bustard Chlamydotis undulata undulata homerange, in the semi-desert steppe of eastern Morocco. Home-ranges were calculated from radiotagged birds using the fixed kernel method (LSCV). The mean annual home-range of males (17 km 2 ) was smaller than that of females (146 km 2 ; P < 0.001). The majority of male home-ranges had a unimodal distribution (86%), whereas 67% of female home-ranges were multimodal. Consequently, the amplitude of female movements was larger (mean: 44 km/13 km; P < 0.002). In spring, male home-ranges decreased in size around display sites (8 km 2 ) and many of them overlapped considerably. Display sites show an aggregative distribution (P < 0.001) and a single female reproductive range may contain the display sites of several males. The data fulfil the definition of an exploded lek social structure. In our study, C. u. undulata appears to be sedentary with relatively limited home-ranges. Successive years of fidelity to home-ranges indicate that adults are not nomadic. The Houbara Bustard is a polytypic Palearctic species inhabiting steppe and semi-desert areas with open and scattered vegetation (Johnsgard 1991). Previously considered as a single species divided into three subspecies, recent studies have proposed a taxonomic revision of the genus Chlamydotis (Gaucher et al. 1996, D Aloia 2001, Pitra et al. 2002, Broders et al. 2003). Two distinct species are currently recognized: the Asiatic species, Chlamydotis macqueenii (Gray 1832), which is distributed from the eastern part of the Nile valley to Mongolia (Vaurie 1965, Cramp & Simmons 1980), and the North African species Chlamydotis undulata. The latter species is divided into two subspecies, Chlamydotis u. fuerteventurae (Rothschild & Hartert 1894), only found in the Canary Islands, and Chlamydotis u. undulata (Jacquin 1784), which is restricted in range from northern Mauritania to the western part of the Nile valley in Egypt (Vaurie 1965, Cramp & Simmons 1980). During the last three decades, populations of Houbara Bustard are believed to *Corresponding author. Email: stjalme@wanadoo.fr have declined drastically throughout their entire range (Collar 1980, Goriup 1997, 1999). Over-hunting and severe habitat degradation are thought to be the two major causes of their decline (Johnsgard 1991, Combreau et al. 2001). Concern over the decline of Houbara populations in Morocco was an important factor leading to the establishment of the Emirates Center for Wildlife Propagation (ECWP) in Missour in 1996. The long-term goal of the ECWP is to secure self-sustaining wild populations of Houbaras within eastern Morocco (Lacroix 2003). To be effective, conservation efforts require a detailed ecological knowledge of the species involved. Through this holistic approach, the Houbara Bustard can be considered as an umbrella species, whose safety can help to ensure survival of the habitat. However, the conservation programme for Houbaras is hindered by the paucity of published data concerning movement patterns and habitat requirements (Combreau et al. 2000), which are key factors in the elaboration of conservation plans. Field studies on movements, migration patterns and survival rates of Houbaras using tracking 2004 British Ornithologists Union

Mating system of the Houbara Bustard 315 methods (Osborne et al. 1997, Launay & Combreau 1999, Launay 2000, Combreau et al. 2001) were all conducted on the Asian species, which is known to be a partial latitudinal migrant from the Nile Valley to Mongolia (Launay & Combreau 1999). One study focused on the assessment of its home-range (Combreau et al. 2000), but this concerned a reintroduced population in a fenced area. Very few studies have considered the two African subspecies. Collins (1984) observed that established male C. u. fuerteventurae showed a high fidelity to their display site during the breeding season and from year to year. No results for C. u. undulata have been provided, from general field surveys, regarding its home-range, movements and breeding strategy. On the basis of very few field observations in Morocco (Brosset 1961, Géroudet 1974, Goriup 1983, Haddane 1985), C. u. undulata is still reported to be apparently resident (as Heim de Balzac & Mayaud 1962). In addition, because the lek seems predominant among mating systems in bustards (Payne 1984, Tarboton 1989, Alonso et al. 2000, Jiguet et al. 2000, Morales et al. 2001), the Houbara Bustard has traditionally been classified as a lekking bird. Because of the scarcity of observations on the African Houbara Bustard C. undulata, hereafter referred to as Houbara, 5 years of field surveys were conducted, in order to increase our knowledge of its behavioural ecology. The aim of this paper is to describe the species s home-range and movements related to sex and season. These results were inferred from data collected using aerial and ground radiotracking. The addition of data concerning distribution of display sites and nest locations provided tools to explore the mating strategy and social structure of Houbara populations in eastern Morocco. METHODS Study area The study was carried out in an area of 40 000 km 2 of semi-desert steppe habitat centred at 33 07 N, 02 58 W, 210 km south-east of Fes. The climate in eastern Morocco is Mediterranean subdesertic with cold winters. Mean annual rainfall is less than 200 mm. Vegetation cover in the plains mainly comprises Chenopodiaceae such as Salsola spp., Hamada spp. and Compositae such as Artemisia herba-alba. In clayand silt-rich areas created by seasonal runoff waters, Salsola sp. is often associated with Atriplex spp. Drainage courses and wadis are characterized by Zizyphus lotus (Ramnaceae) and Retama sp. (Leguminosae). On the high plateaux and slopes, from the extreme east to the Oran region of Algeria, the vegetation is dominated by Stipa tenacissima (Graminea). Display and nest-sites were studied in a protected core area of 663 km 2 situated 20 km north of Missour, a piedmont characterized by slightly undulating plains from the Middle Atlas to the Moulouya river. This core area, named Al Baten, has been protected from hunting since 1996, and shelters a breeding population of Houbaras. Trapping and radiotracking Different trapping methods were used to catch Houbaras from 1997 to 2001: various nets, snares around a bait female on a male display site, a bow-net used on the nest, snares around the nest or surrounding a wildcaught chick, and using a specially equipped trained falcon (Launay et al. 1999, Seddon et al. 1999). Birds were banded with a unique metal leg band and radiotagged. Two different types of battery-powered transmitters were used: a backpack (43 g, 18-month duration, Telonics, Inc., Meza, AZ, USA) and a necklace (18 g, 36-month duration, HSL RI-2CSP, Holohil System Ltd, Ontario, Canada). Houbaras were monitored from the ground and via aerial telemetry. Ground radiotracking was undertaken by direct observation using a portable scanner-receiver and a three-element Yagi antenna (AF Antronics, Inc., Urbana, IL, USA), which can receive signals from up to 10 km on elevated sites. Aerial locations were taken using one two-element Yagi antenna fixed to each wing strut of an aeroplane (Maule-7 B-235). Lost birds were relocated using a four-element Yagi antenna (AF Antronics, Inc.) fixed on the right wing strut and pointed forward, which allows reception of signals from as far as 45 km at an altitude of 1800 m. Aerial and ground telemetry were both generally carried out weekly. Forty birds were trapped from February 1997 to April 2002. Of these, two females were killed by predators in the month following the capture, and seven individuals who were lost for several months were excluded from the analysis. Thus 31 birds (22 males and nine females) were located weekly, with a mean interval between locations of 8 days (sd = ±3), giving a total record of 2605 telemetry locations. The data from one female that was lost for several weeks in the breeding season were used only in the analysis of the non-breeding range. Thirty Houbaras were followed for at least one complete year with a mean of 84 locations (43 254) and a mean duration of 581 radiotracking days (332 1133). Of these, one

316 Y. Hingrat et al. male was followed for 1 year, 15 males and two females for a period that included two breeding seasons, and six males and six females during three successive breeding seasons. After trapping, marking and release, we often observed long escape flights at the release site. Most of the males trapped on their display site spent a few days far away from it without displaying. We considered that this uncommon behaviour was related to the stress of capture. Therefore, locations recorded between the release day and when the males reached their displaying site were removed from analysis. Site fidelity Initially, as recommended by Hooge and Eichenlaub (1997), we tested whether individuals exhibited a site fidelity pattern in their movement path. We used the site fidelity test, which is an extension of the Monte Carlo random walk test developed by Spencer et al. 1990) and modified by Hooge and Eichenlaub (1997). One thousand simulations were run for each bird, in order more accurately to reflect the random walk distribution (Hooge & Eichenlaub 1997). The starting point of each simulation was the trapping location of the bird. We calculated for each walk both the mean squared distance from the centre of activity (harmonic mean) and the linearity of the path. These two values were measures, respectively, of the data dispersion (R 2 ) and of directed movement (L). The actual movement path s values were then compared with the ranked values of the random walks to determine any significance. To be site-faithful, the animal s real locations should exhibit neither significant dispersion (R 2 ) nor significant linearity (L) (when 95% of random movement paths have higher R 2 and L values) (Hooge & Eichenlaub 1997). Home-range and movement estimations Many statistical home-range estimators are available in the literature (Mohr 1947, Jennrich & Turner 1969, van Winkle 1975, Macdonald et al. 1980, Schoener 1981, Anderson 1982, Worton 1987, Harris et al. 1990, White & Garrot 1990, Goldingay & Kavanagh 1993). All have different underlying assumptions, providing different operating characteristics, and can therefore produce dissimilar results for a specific set of data (Boulanger & White 1990). On the basis of our field observations, we decided to use the fixed kernel method with the cross-validated h calculated from a least-square cross-validation (LSCV). This method is widely recognized as an efficient estimate of the probability density distribution of locations, and allows us clearly to identify non-convex and multimodal features such as a home-range consisting of separate areas (Silverman 1986, Worton 1987, Seaman & Powell 1996). The choice of an accurate estimate of the shape and size of the home-range is of major importance because a slight difference in the position of home-range boundaries of neighbouring animals may have considerable implications for the interpretation of their social organization (Macdonald et al. 1980). Seaman and Powell (1996) have shown a sample size and data structure effect on the degree of smoothing, which can result in unexpected patterns for LSCV kernel-based estimates of seasonal vs. yearly home-ranges. Thus for multimodal homeranges identified with the kernel method, the overall size was calculated by adding together the areas from each mode. We calculated the fixed kernel, with probability density of 95%, 75% and 50%, for birds followed during at least one complete year. Estimation of annual home-ranges was performed on samples of more than 40 locations, as recommended by Seaman et al. (1999), who used simulations to test sample size effects on the kernel method. Given that birds showed fidelity to their range in successive years, home-ranges were calculated from all available data in order to attain a significant number of locations for each individual. Seasonal home-ranges were calculated for birds with samples of more than 15 fixes, in both breeding and non-breeding periods. These periods were set according to the displaying and laying periods observed in the field. For each bird, distances between successive locations were calculated. As an indicator of movement ability we used the maximum distance travelled by each individual between successive radio locations. Search for display sites and nests We searched for displaying males by using observation points distributed every 1.5 km along 12 transects regularly spaced in the core study area (663 km 2 ). At each point, two observers made a circular observation lasting up to 15 min with binoculars or a telescope. Surveys were conducted at least three times a week, from January to the end of May in 2000 and 2001. They were driven early in the morning and at the end of the afternoon, when males were on

Mating system of the Houbara Bustard 317 display sites (Saint Jalme & Van Heezik 1996). Counts were conducted only under similar climatic conditions (e.g. no wind and no clouds). Males sighted at least once were then monitored regularly throughout the season to obtain more precise coordinates in order to exclude double counts and to assess male fidelity to their display area. Because nesting females are very shy and cryptic, it was not possible to set a standard protocol in order to record the distribution of nests accurately in the study area. From February to June 2001, a survey of nest locations was carried out, using the help of local shepherds. In addition, as nesting females are reluctant to fly, and they adopt a characteristic behaviour (Gaucher 1995, Combreau et al. 2002), field workers were asked to pay attention to wild females encountered during surveys. Females exhibiting such behaviour were then followed at a distance of at least 800 m with a telescope until they returned to their nest. Dates of laying were estimated from the hatching date, considering an average incubation period of 23 days in the wild (Gaucher 1995, Combreau et al. 2002). Data analysis and statistics The coordinates of bird, nest and display site locations were recorded using a GARMIN 2+ GPS locator. Observations were plotted on a map of the study area using a Geographical Information System (ARCVIEW 3.2 Environmental Systems Research Institute, Inc. 1996). Distances between locations, site fidelity tests and home-range estimates were calculated in ARCVIEW using the Animal Movement Extension (Hooge & Eichenlaub 1997). Mean home-range sizes were compared between sexes, and breeding and non-breeding seasons. Mean maximum distances were compared between sexes and types of distribution (uni- or multimodal). Because of small sample sizes, all comparisons were performed using the non-parametric Mann Whitney U-test with SYSTAT 7.0. Spatial distributions of nests and display sites were tested with the nearest neighbour method (Clark & Evans 1954). RESULTS Site fidelity The site fidelity test rejected the null hypothesis for the 31 individuals included. Observed movements were more constrained than random movement paths (Monte Carlo simulation, n = 1000, P < 0.05). Therefore, Houbaras did not move at random, but did demonstrate fidelity to certain areas during their travels. Thus, estimation of their home-range can be investigated. Movements and distribution shape of locations Nineteen males (86%) presented a unimodal distribution of locations (unimodal range: UMR), against only three of the nine females monitored. In samples that show a multimodal distribution (multimodal range: MMR), the maximum distance travelled is nearly the greatest distance between modes. Consequently, the amplitude of their movements was larger than those of unimodally distributed ones. In males, the mean maximum distance of MMR birds (47 km) was significantly larger than that of UMR birds (8 km; Table 1). In females, the mean maximum distance of MMR birds (58 km) was also significantly larger than that of UMR birds (16 km, Table 1). Because of the low representation of MMR among Table 1. Mean maximum distances travelled (km) by Houbara Bustards, according to sex and type of home-range (uni- and multimodal). All birds Unimodal home-range Multimodal home-range P1 Female n 9 3 6 mean ± sd 44 ± 36 17 ± 4 58 ± 35 P < 0.05 range 13 114 13 21 19 114 Male n 22 19 3 mean ± sd 13 ± 15 8 ± 6 47 ± 46 P < 0.006 range 2 51 2 22 42 51 P2 P < 0.002 ns ns P1: comparison between uni- and multimodal home-ranges using the Mann Whitney U-test. P2: comparison between sexes using the Mann Whitney U-test.

318 Y. Hingrat et al. Figure 1. Maximum distances travelled by Houbara Bustards (km) during an annual cycle, according to sex (male:, female: ). Figure 2. Occurrence of Houbara Bustard s home-range sizes (fixed kernel 95%, LSCV), according to sex (male:, female: ). males (14%) and the high proportion among females (67%), in the overall population the mean maximum distance travelled by females (44 km) appeared to be significantly greater than that of males (13 km; Table 1, Fig. 1). In each type of distribution the difference between sexes was not significant. Home-range The mean annual home-range size of males (17 km 2 with 95% estimate density) was significantly smaller than that of females (146 km 2 ; Table 2, Fig. 2). Male home-ranges were still significantly smaller than those of females for the fixed kernel estimations at 75% and 50% density (Table 2, Figs 2 & 3). The mean breeding home-ranges of females (108 km 2 with 95% estimate density) was significantly larger than that of males (8 km; Table 2). The difference between sexes was also significant for the fixed kernel at 75% and for the core area (50%). In males, during the reproductive season, homerange size decreased significantly around the display site (Fig. 4, Table 2). The mean breeding home-range (95%) appeared to be significantly smaller than in the non-breeding season (8 vs. 26 km 2 ; Mann Whitney: U = 376, P = 0.003). The 75% and 50% fixed kernels both decreased significantly during breeding from 10 to 2 km 2 (U = 393, P = 0.001), and from 4 to 1 km 2 (U = 376, P = 0.003), respectively. In the breeding season, the males core area was relatively small (0.1 4 km 2 ) and encompassed the display site (Fig. 1). Annual ranges as well as breeding ranges of males overlapped widely (Figs 1 & 2). In females, there was no significant difference in home-range sizes between seasons for the three probability density estimations (Table 2). Females with well-separated seasonal territories established breeding ranges around male aggregates. Males and females Table 2. Annual, breeding and non-breeding home-ranges (HR) of male and female Houbara Bustards, calculated (km 2 ) with the fixed kernel method for the three probability densities of 95%, 75% and 50%. Density estimate Male Female n Mean ± sd Range n Mean ± sd Range P Annual HR 95% 22 17 ± 22 2 98 8 146 ± 97 30 317 0.001 75% 22 5 ± 6 0.5 30 8 54 ± 35 12 126 0.001 50% 22 2 ± 3 0.2 15 8 20 ± 11 2 37 0.001 Breeding HR 95% 22 8 ± 8 1 31 8 108 ± 100 28 338 0.001 75% 22 2 ± 2 0.3 9 8 33 ± 22 8 69 0.001 50% 22 1 ± 1 0.1 4 8 14 ± 12 2 37 0.001 Non-breeding HR 95% 22 26 ± 27 3 89 9 105 ± 73 20 210 0.001 75% 22 10 ± 11 1 39 9 45 ± 32 12 96 0.001 50% 22 4 ± 5 0.3 19 9 16 ± 11 5 35 0.001 P: comparison between sexes using the Mann Whitney U-test.

Mating system of the Houbara Bustard 319 Figure 3. Annual home-ranges (fixed kernel 75%) of 18 male (thin lines) and four female Houbara Bustards (bold lines) with multimodal ranges (females A, B, C and D). Figure 5. Breeding home-ranges of two female Houbara Bustards (female X with an unimodal range and Y with a multimodal range) during two successive reproductive seasons (fixed kernel 95% 2001: thin lines; 2002: bold lines). Successive nesting sites of female X (nest 2001: ; nest 2002: ), successive nesting sites of female Y (nest 2001: ; nest 2002: ) and locations of males display sites (+) within the study area (663 km 2 ) of Al Baten are shown. Figure 4. Breeding home-ranges of 18 male Houbara Bustards (fixed kernel 95% including the 50% core area) and their display sites ( ). Locations of display sites of other males (+) and nest sites ( ) within the study area (663 km 2 ) of Al Baten are also shown. followed during more than one breeding season were faithful to their breeding range, although females that used the same breeding area from one year to another (Fig. 5) did change their nest-sites. Displaying males Fifty-three display sites were identified in 2000 and 40 in 2001. These counts give us a minimum display site density of 0.08 birds/km 2 in 2000 and 0.06 birds/km 2 in 2001 on the Al Baten core study area (663 km 2 ), with a respective mean distance to the nearest neighbour of 0.86 km (sd = ±1.01) and 0.96 km (sd = ±0.89). Display sites were not randomly distributed; a tendency towards clumping existed in 2000 (R = 4.515 10 6, z = 14.569, P < 0.001) as well as in 2001 (R = 4.641 10 6, z = 12.099, P < 0.001; Fig. 4). In 2001, display observations occurred from 17 January to 25 May. Twelve out of the 14 males marked in 2000 and 2001 were seen displaying solitarily on the same site during the whole breeding season of 2002. Among these males, which showed a high fidelity to their site, only one started displaying in December 2001, whereas all the others started in the first 15 days of January 2002. From this it would appear that the breeding season of males lasted from January to May inclusive. Nests Twenty-five nests were found during the 2001 breeding season, and these were significantly aggregated (R = 5.18 10 6, z = 9.565, P < 0.001; Fig. 4). Because the search effort was not uniform across the study area (i.e. higher shepherd density near

320 Y. Hingrat et al. villages), the number of nests and their distribution must be interpreted with care. To obtain more information on the distribution of nesting areas, we plotted the 59 nests recorded since 1996 by the ECWP (Fig. 4). Nests showed an aggregated distribution (R = 6.096 10 6, z = 14.694, P < 0.001). The first laying was estimated at 16 February. Later nests found at the end of June probably contained replacement clutches. Radiotracking showed that multimodal females reached their laying area from December to April. From these results, we can say that the females breeding season lasted from the beginning of January to the end of June. This arbitrary partitioning of seasons, based on our observations, was made for the purposes of seasonal home-range estimation. It must be kept in mind that these periods may vary in time because of climatic variations from year to year. In addition, although male breeding periods are quite synchronous, female ones are not. DISCUSSION We identified differences between males and females in the use of space, both in the shape and size of their home-ranges. Females show a tendency to move more (mean maximum distance of 44 km) and to use larger areas (146 km 2 ) than males (means of 13 km and 17 km 2 ). Sixty-three per cent of females had multimodal home-range features with the same marked seasonal use of each mode. Females use different areas in the breeding and the wintering season. This could be linked to male density and/or resource availability, as related to mating and rearing of chicks. Nevertheless, where females showed fidelity to their breeding area from year to year, the nesting site did change, and factors leading to its selection are still unknown. Another characteristic of female distribution is the non-convex shape of the home-range with a breeding fixed kernel isopleth (95%) overlapping the males home-ranges and including several display areas, suggesting that females may visit them for mating. Display sites appeared to be non-randomly distributed, with a mean distance to the nearest neighbour of 0.9 km. In the Canary Islands, the distance between neighbouring males was about 1 km, with a breeding range of 0.6 km 2 (Collins 1984). In Morocco, the annual home-range of males is larger, but significantly concentrated around the display site during the breeding season (core area of about 1 km 2 ). Males display in the morning and at the end of the afternoon and spend the remainder of the day resting or foraging in the vicinity of their site. However the term territory, widely used in mating system studies, would be misleading in the case of the Houbara Bustard. In fact, it is not uncommon to find nearest neighbours foraging together when not displaying. This behaviour has also been described in the Great Bustard Otis tarda (Morales & Martin 2002). Males should be considered as partially territorial during a part of the day and only during the 5 months of the breeding season in a small area, the display site. Consequently, and considering the aggregation of display sites, the breeding ranges and annual ranges of neighbours overlapped widely. In lek mating systems, males display conspicuously from small display territories, but do not control resources that are necessary for reproduction, and do not provide parental care (Bradbury & Gibson 1983). This definition of lek includes both classical leks, in which male display sites are densely clustered, and exploded leks, in which display sites are more dispersed (Emlen & Oring 1977). However, the differences between classical and exploded leks are continuous rather than categorical and thus should be treated with caution (Höglund & Alatalo 1995). In fact, as has been shown in other bustard species in which the exploded lek strategy was recognized (Tarboton 1989, Alonso et al. 2000, Jiguet et al. 2000, Morales et al. 2001), the dispersed nature of display arenas is compensated by the open habitat and visually striking courtship behaviour that characterizes the family. Although Collins (1984) observed a nonrandom distribution of display sites, visits by females for mating and no male parental care, he excluded the exploded lek hypothesis to explain the Canarian Houbara Bustard mating system. Because Collins (1984) did not observe overlapping between male ranges and because they were not always able to see the displays of their neighbours, he argued that the apparent clumping of males indicates no more than a shortage of habitat. In other studies, the Asian species C. macqueenii has been considered as polygynous or promiscuous, but its mating system is still unknown (Ponomavera 1983, Launay & Paillat 1990, Launay & Loughland 1995). One of the four criteria suggested by Bradbury (1981) to distinguish leks from other mating systems is the assumption that the display sites of males contain no significant resources required by females except the males themselves. In our Houbara population, the role of resources in the male s choice of display site has not yet been clarified. However, as has been shown in the Little Bustard Tetrax tetrax, limits between resource defence polygyny and extreme exploded or resource-based

Mating system of the Houbara Bustard 321 leks are thin and unclear (Jiguet et al. 2000). All the spatial parameters measured, and behaviours observed, are consistent with those expected from the definition of an exploded lek (Emlen & Oring 1977). Nevertheless, more studies are necessary to understand the role of resource availability and natal dispersion in the social structure of the population. The African Houbara Bustard appears to be mainly sedentary with relatively limited home-ranges, but local movements linked to reproduction (especially by females) must be treated with caution. Recently, the term partial-migration has been proposed to explain seasonal movements of Great Bustards between wintering, lekking, nesting and chick-rearing areas (Alonso et al. 2000, Morales et al. 2000). This notwithstanding, successive years of fidelity to homeranges indicate that adults are not nomadic. This result is very important in terms of population management. Interannual fidelity to a relatively small home-range means that juvenile dispersal would be the only source of birds for colonization of new areas or for exchanges between adjacent populations. Understanding postjuvenile dispersion should be the next step of our study. We would like to thank Their Highnesses Sheikh Zayed bin Sultan Al Nahyan and Sheikh Mohammed bin Zayed Al Nahyan, Mr Mohammed Al Bowardi and Mr Jacques Renaud for funding and supporting the research programme. Special thanks are given to all those involved in data collection in the field: Moussa Marou Dodo, Nicolas Orhant, Pierre-Marie Beranger, Eric Le Nuz, Charles Facon, Albane Du Boisgueheneuc and Joseph Le Cuziat. We would also like to thank Marie Christine Eybert for her comments on the manuscript and Alain Canard for his helpful participation. Many thanks to Dr Brendan Prendiville for improving the English text. REFERENCES Alonso, J.C., Morales, M.B. & Alonso, J.A. 2000. Partial migration, and lek and nesting area fidelity in female Great Bustards. Condor 102: 127 136. Anderson, D.J. 1982. The home range: a new nonparametric estimation technique. Ecology 63: 103 112. Boulanger, J.G. & White, G.C. 1990. A comparison of home range estimators using Monte Carlo simulation. J. Wildl. Manage. 54: 310 315. Bradbury, J.W. 1981. The evolution of leks. In Alexander, R.D. & Tinkle, D.W. (eds) Natural Selection and Social Behavior: 138 169. New York: Chiron Press. Bradbury, J.W. & Gibson, R.M. 1983. Leks and mate choice. In Bateson, P. (ed.) Mate Choice: 109 138. Cambridge: Cambridge University Press. Broders, O., Osborne, T. & Wink, M. 2003. A mtdna phylogeny of bustards (family Otididae) based on nucleotide sequences of cytochrome b-gene. J. Ornithol. 144: 176 185. Brosset, A. 1961. Ecologie des oiseaux du Maroc Oriental. Travaux l Institut Scientifique Chérifien, Séries Zool. 22: 16 19, 22 23, 40 41, 114 117, 120 123, 138 139. Clark, P.J. & Evans, F.C. 1954. Distance to the nearest neighbor as a measure of spatial relationships in populations. Ecology 35: 445 453. Collar, N.J. 1980. The world status of the Houbara: a preliminary review. In Coles, C.L. & Collar, N.J. (eds) The Houbara Bustard: 12. Poole: Sydenhams. Collins, D. 1984. A Study of the Canarian Houbara Bustard (Chlamydotis undulata fuertaventurae) with Special Reference to its Behaviour and Ecology. M.Phil. thesis, University of London. Combreau, O., Gelinaud, G. & Smith, T.R. 2000. Home range and movements of Houbara Bustards introduced in the Najd Pediplain in Saudi Arabia. J. Arid. Environ. 44: 229 240. Combreau, O., Launay, F. & Lawrence, M. 2001. An assessment of annual mortality rates in adult-sized migrant Houbara Bustards (Chlamydotis [undulata] macqueenii). Anim. Conserv. 4: 133 141. Combreau, O., Qiao, J., Lawrence, M., Gao, Y., Yang, W. & Launay, F. 2002. Breeding success in a Houbara Bustard Chlamydotis [undulata] macqueenii population on the eastern fringe of the Jungar Basin, People s Republic of China. Ibis 144 (on line): E45 E46. Cramp, S. & Simmons, K.E.L. 1980. The Birds of the Western Palearctic, Vol 2. Oxford: Oxford University Press. D Aloia, M.A. 2001. Studies on the population structure of the Houbara Bustard Chlamydotis undulata in the Middle East with DNA analysis techniques. Zool. Mid. East 22: 25 35. Emlen, S.T. & Oring, L.W. 1977. Ecology, sexual selection, and the evolution of mating systems. Science 197: 215 223. Gaucher, P. 1995. Breeding biology of the Houbara Bustard Chlamydotis undulata undulata in Algeria. Alauda 63: 291 298. Gaucher, P., Paillat, P., Chappuis, C., Saint Jalme, M., Lotfikhah, F. & Wink, M. 1996. Taxonomy of the Houbara Bustard Chlamydotis undulata subspecies considered on the basis of sexual display and genetic divergence. Ibis 138: 273 282. Géroudet, P. 1974. Notes marocaines sur la parade nuptiale de l Outarde Houbara Chlamydotis undulata. Oiseau Rev. Fr. Ornithol. 44: 154 155. Goldingay, R.L. & Kavanagh, R.P. 1993. Home range estimates and habitat of the Yellow-billed Glider (Petaurus australis) at Waratah Creek, New South Wales. Wildl. Res. 20: 387 404. Goriup, P.D. 1983. The Houbara Bustard in Morocco. Al-Areen: ICBP. Goriup, P.D. 1997. The world status of the Houbara Bustard Chlamydotis undulata. Bird Conserv. Int. 7: 373 397. Goriup, P.D. 1999. Houbara bustard conservation makes progress. Oryx 33: 184 187. Gray, J.E. 1832. Otis macqueenii. Harswickes illustrated. Indian Zool. 2: no pagination. Haddane, B. 1985. The Houbara Bustard in Morocco: a brief review. Bustard Stud. 3: 109 112. Harris, S., Cresswell, W.J., Forde, P.G., Trewhella, W.J., Woolard, T. & Wray, S. 1990. Home range analysis using radio-tracking data, a review of problems and techniques particularly as applied to the study of mammals. Mammal Rev. 20: 97 123.

322 Y. Hingrat et al. Heim de Balzac, H. & Mayaud, N. 1962. Les Oiseaux du Nord-Ouest de l Afrique. Paris: Paul Lechevalier. Höglund, J. & Alatalo, R.V. 1995. Leks. Princeton: Princeton University Press. Hooge, P.N. & Eichenlaub, B. 1997. Animal Movement Extension to Arcview, Version 1.1. Anchorage, AK, USA: Alaska Biological Science Center, U.S. Geological Survey. Jacquin. 1784. Psophia undulata. Beytr. Gesch. Vogel. 24: 24, plate 9. Jennrich, R.I. & Turner, F.B. 1969. Measurements of noncircular home range. J. Theor. Biol. 22: 227 237. Jiguet, F., Arroyo, B. & Bretagnolle, V. 2000. Lek mating systems: a case study in the Little Bustard Tetrax tetrax. Behav. Process. 51: 63 82. Johnsgard, P.A. 1991. Bustards, Hemipodes, and Sandgrouses. Birds of Dry Places. Oxford: Oxford University Press. Lacroix, F. 2003. The Emirates Center for Wildlife Propagation: developing a comprehensive strategy to secure a selfsustaining population of Houbara Bustards in eastern Morocco. Houbara News 5: 2. Launay, F. 2000. Tracking the Houbara. World Birdwatch 22: 18 21. Launay, F. & Combreau, O. 1999. Annual migration of Houbara Bustard Chlamydotis undulata macqueenii from the United Arab Emirates. Bird Conserv. Int. 9: 155 161. Launay, F., Combreau, O., Aspinall, S.J., Loughland, R.A., Gubin, B. & Karpov, F. 1999. Trapping of breeding Houbara Bustard (Chlamydotis undulata). Wild. Soc. Bull. 27: 603 608. Launay, F. & Loughland, R. 1995. Breeding system of Houbara Bustard Chlamydotis undulata macqueeni: preliminary observations. Sandgrouse 35: 14 17. Launay, F. & Paillat, P. 1990. A behavioural repertoire of the adult Houbara Bustard (Chlamydotis undulata macqueenii). Rev. Ecol. (Terre Vie) 45: 65 88. Macdonald, D.W., Ball, F.G. & Houg, N.G. 1980. The evaluation of home range size and configuration using radio tracking data. In Amlaner, C.J. & Macdonald, D.W. (eds) A Handbook on Biotelemetry and Radiotracking: 405 424. Oxford: Pergamon Press. Mohr, C.O. 1947. Table of equivalent populations of North American small mammals. Am. Midl. Nat. 37: 223 249. Morales, M.B., Alonso, J.C., Alonso, J.A. & Martin, E. 2000. Migration patterns in male Great Bustards (Otis tarda). Auk 117: 493 498. Morales, M.B., Jiguet, F. & Arroyo, B. 2001. Exploded leks: what bustards can teach us. Ardeola 48: 85 88. Morales, M.B. & Martin, C.A. 2002. Otis tarda Great Bustard. BWP Update 4: 217 232. Osborne, P.E., Launay, F. & Gliddon, D. 1997. Wintering habitat use by Houbara Bustards Chlamydotis undulata in Abu Dhabi and implications for management. Biol. Conserv. 81: 51 56. Payne, R.B. 1984. Sexual selection, lek and arena behavior, and sexual size dimorphism in birds. Ornith. Monogr. 33: 52. Pitra, C., Lieckfeldt, D., Frahnert, S. & Fickel, J. 2002. Phylogenetic relationships and ancestral areas of the Bustards (Gruiformes: Otididae), inferred from mitochondrial DNA and nuclear intron sequences. Mol. Phylogenet. Evol. 23: 63 74. Ponomavera, T.S. 1983. Reproductive behaviour and distribution of Houbara Bustards on their breeding grounds. Zool. Mag. 64: 592 603. Rothschild, W. & Hartert, E. 1894. On a new bustard from the Palaearctic region. Nov. Zool. 1: 689. Saint Jalme, M. & Van Heezik, Y. 1996. Propagation of the Houbara Bustard. London: Keegan, Paul. Schoener, T.W. 1981. An empirically based estimate of home range. Theor. Pop. Biol. 20: 281 325. Seaman, E.D., Millspaugh, J.J., Kernohan, B.J., Brundige, G.C., Raedeke, K.J. & Gitzen, R.A. 1999. Effects of sample size on kernel home range estimates. J. Wildl. Manage. 63: 739 747. Seaman, E.D. & Powell, R.A. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology 77: 2075 2085. Seddon, P.J., Launay, F., Van Heezik, Y. & Al Bowardi, M. 1999. Methods for live trapping Houbara Bustards. J. Field Ornithol. 70: 169 181. Silverman, B.W. 1986. Density Estimations for Statistics and Data Analysis. London: Chapman & Hall. Spencer, S.R., Cameron, C.G. & Swihart, R.K. 1990. Operationally defining home range: temporal dependence exhibited by hispid cotton rats. Ecology 71: 1817 1822. Tarboton, W. 1989. Breeding behaviour of Denham s Bustards. Bustard Stud. 4: 160 169. Vaurie, C. 1965. The Birds of the Palearctic Fauna. London: Witherby. White, G.C. & Garrot, R.A. 1990. Analysis of Wildlife Radio- Tracking Data. San Diego, CA: Academic Press. van Winkle, W. 1975. Comparison of several probabilistic home range models. J. Wildl. Manage. 39: 118 123. Worton, B.J. 1987. A review of models of home range for animal movement. Ecol. Model. 38: 277 298. Received 27 January 2003; revision accepted 11 November 2003.