The northern bald ibis Geronticus eremita: history, current status and future perspectives

Abstract The northern bald ibis Geronticus eremita was once widespread throughout the Middle East, northern Africa, and southern and central Europe. Habitat destruction, persecution and the impacts of pesticides have led to its disappearance from most of its former range. It disappeared from central Europe > 400 years ago, but has persisted as a relict and slowly growing breeding population in Morocco, where c. 700 wild birds of all ages remain. In Algeria, the last confirmed breeding was in 1984; in Turkey the fully wild population disappeared in 1989, but a population remains in semi-wild conditions. In Syria a small population was rediscovered in 2002, only to subsequently decline to functional extinction. Restoration programmes have been initiated independently in several locations, with over 300 free-flying birds resulting from reintroduction projects in Austria, Germany, Spain and Turkey, to restore both sedentary and fully migratory populations. Maintaining current efforts in Morocco remains a high conservation priority.


Introduction
T he northern bald ibis Geronticus eremita is one of the rarest birds. Although recently recategorized on the IUCN's Red List from Critically Endangered to Endangered, it has a precariously small wild population limited to a handful of breeding sites, and the main subpopulation has only recently recovered slightly (BirdLife, ). The northern bald ibis was probably once widespread across the Middle East, northern Africa, and southern and central Europe, with wintering populations as far south as Mauritania and Senegal, and from the Arabian Peninsula and African Red Sea coast to Eritrea and Ethiopia. Habitat destruction, persecution and the impacts of agricultural pesticides led to its disappearance from most of its former range (Hirsch, , ; Collar & Stuart, ). It was extirpated from central Europe .  years ago, but breeding populations persisted in Morocco, Algeria, Turkey and Syria into the th century. The only viable wild population is now in Morocco, but there are significant captive zoo populations, release projects and semi-wild populations. Here we review the species' current status throughout its former range, and look at conservation prospects, ongoing reintroductions and natural recolonizations. The geographical distribution of colonies, projects and sightings (up to and including ), and consolidated figures for all wild, semiwild, released and captive populations, are summarized in Fig.  and Table .

Indigenous range and population structure
Based on minor morphological distinctions and preliminary genetic work, two populations (eastern and western) have been identified, but the extent and timing of their separation appears to be minor and recent (Pegoraro et al., ; Wirtz et al., , ). All fully captive populations in Europe, Japan and North America, and those used in European release projects, derive from western origin birds (Böhm, , ), whereas the Turkish semi-wild birds are of local, eastern origin. Thus, although more work is needed, the two populations can be regarded as distinct management units.

Morocco
Historically, breeding colonies occurred in the High and Middle Atlas Mountains and some Atlantic coastal areas north of the present colonies. The number of populations declined from  sites in  to  in  and only three in , as a result of habitat modification, human disturbance, hunting and pesticide use (Collar & Stuart, ). Thevenot et al. () documented a higher number of historical colonies because they treated some dispersed colonies as separate entities. Historical records suggest that the now extinct inland populations were migratory, wintering further south in Morocco and as far away as Mauritania and Mali (Bowden, ).
The remaining global wild population now occupies a small coastal strip in southern Morocco in the region of Agadir, where it breeds on sea cliffs in two main areas and is resident year-round. In  the Moroccan authorities designated the Parc National de Souss-Massa, a , ha coastal protected area specifically delimited to protect the nesting and feeding areas of the northern bald ibis south of Agadir. The other main breeding and feeding area is a coastal strip  km north of Agadir in the region of Tamri, a no-hunting area designated as a Site d'Interêt Biologique et Ecologique, but with no formal protection status. By  a total population of c.  birds was breeding in three subcolonies within the Parc National de Souss-Massa, and in one colony at Tamri (Bowden, ). There is some movement between these sites, which are c.  km apart (Bowden et al., ; El Bekkay et al., ).
A project, started in  as a collaboration between the Royal Society for the Protection of Birds and the Sociedad Española de Ornitología, and later Groupe de Recherche pour la Protection des Oiseaux au Maroc, working with Parc National de Souss-Massa, aims to increase the northern bald ibis population in Souss-Massa. Local fishers were hired and trained as wardens, and a range of local Northern bald ibis rural development projects were implemented in the area to develop grass-root interest in the species' conservation. Managing and maintaining the local wardening at both Parc National de Souss-Massa and Tamri has been the main intervention, and has been consolidated recently with Moroccan government support.
By  the population had declined to  pairs following the unexplained death of  birds in  (Bowden, ; Touti et al., ), but by  it had recovered to  breeding pairs (Oubrou & El Bekkay, ) and has continued to increase. By , there was a total of  individuals, with  pairs raising  fledglings. In ,  pairs bred and there was a post-breeding total population of  birds, including recently fledged juveniles. The total population at the end of the  breeding season was .  individuals, with a breeding adult population of  pairs (Oubrou & El Bekkay, ; Table ). These trends appear to be linked to measures taken by Parc National de Souss-Massa to engage local communities and develop wardening capacity. Intensive monitoring of breeding and feeding sites and the provision of  supplementary fresh water has improved breeding productivity (Smith et al., ). The population is apparently dependent on littoral steppe habitat and a traditional agriculture system where large areas are left fallow and grazed by sheep for several years at a time. This supports high biomass and diversity of the invertebrates and reptiles that are the main prey for the ibis (Aghnaj et al., ).
There is renewed hope that the birds may naturally recolonize former breeding sites in Morocco (as predicted in Böhm & Bowden, , ). In  breeding was confirmed for the first time in several decades in a small subcolony near Immsouane (Aourir et al., ),  km north of Tamri.

Algeria
In Algeria, most of the data on the historic distribution of the species are from the s when  colonies were known. Another colony was discovered in  in the north-western El Bayadh region, persisting until the late s (Fellous, ). Northern bald ibises in Algeria were migratory, arriving at breeding sites towards the end of winter in February and early March, and departing after breeding in late summer.
The Algerian colonies were historically estimated to be c. - birds, but signs of population decline were noted in the s (Collar & Stuart, ). Causes of decline included hunting, habitat loss caused by overgrazing, and droughts in the s and s, by which time ,  individuals remained. The northern bald ibis became extinct in Algeria during the late s or early s (Fellous, ).
Questionnaires circulated by Algeria's National Agency for Nature Conservation showed that in autumn  two birds were seen flying at Thiet Ould Moumen,  km north of the last known breeding site for this species in Algeria; there have not been any subsequent observations.

Turkey
The northern bald ibis was once present through large areas of south-eastern Turkey. By the first half of the s there were five documented colonies, one with as many as , birds (Akçakaya, ). This population underwent a dramatic decline from the late s, principally resulting from the use of DDT to combat insect pests (Hirsch, , ; Hatipoğlu, ). Following deaths and reproductive failure, the breeding population was reduced to  pairs by  (Arihan, ). In  Turkey's General Directorate of Nature Protection and National Parks (now Doğa Koruma ve Milli Parklar Genel Müdürlüğü) initiated a project in collaboration with the Society for the Protection of Nature, Turkey (the former BirdLife Partner), to save the remaining wild northern bald ibis population in Birecik. A breeding station was established  km north of the town of Birecik, and  birds were caught during - and taken into semi-captivity. A similar number of birds were left in the wild. The captive birds were initially enclosed throughout the year, but a decision was made to release them from the aviaries each spring (in mid February, around the dates they would have arrived back from migration), and they were encouraged to breed on the rock ledges and in nest boxes outside the aviaries, with supplementary food provided twice per day. This spring release practice continues today, and the free-flying birds forage in the surrounding area during the breeding season and until they are recaptured each July. This prevents migration and is believed to improve survival rates, although a few birds (mainly juveniles) have disappeared before recapture, and it is possible some of them have migrated south. Unfortunately, the remaining fully wild population continued to decline, and was declared extinct in  (Akçakaya ; Yeniyurt et al., ).
The semi-wild population has increased despite a number of sudden deaths of chicks and adult birds, notably during -, in part caused by contamination of supplementary food. By December  there were  birds, of all ages, in seasonal captivity at Birecik (Table ). Doğa Derneği (BirdLife Turkey) together with Doğa Koruma ve Milli Parklar Genel Müdürelüğü have monitored the breeding population since  (Yeniyurt et al., ), and are working to enhance veterinary conditions in the breeding station in Birecik, to increase the size of the semi-wild colony and thereby re-establish the migratory population at Birecik.
From  the release and tracking of juvenile birds from the Birecik Breeding Centre confirmed that the Turkish northern bald ibis had not lost the ability to migrate (Hatipoğlu, ). In  juveniles moved south, following a similar trajectory to that of the Palmyra birds, and one bird returned to the colony in spring , although the overwinter location of this individual was not identified (Lindsell et al.,  ). None of the other released birds have returned. The release programme was suspended in  when the Syrian civil war started, and satellite tags on three of the released birds stopped transmitting in circumstances suggesting hunting. Nevertheless, releases remain a high priority conservation action for the eastern flyway of the species.

Syria
The northern bald ibis was previously relatively common in the Al Badia landscapes of Syria (Serra et al., ). Little is known of the species' former distribution in Syria, or the factors driving declines. The birds once migrated through Jordan, Saudi Arabia, Yemen, Eritrea and Sudan, with stopover sites during the autumn migration between breeding sites in Syria (and Turkey, see above) and the wintering ground in the Ethiopian Highlands (Lindsell et al., ; Bowden, ). Hunting along the migration route through the western areas of the Arabian Peninsula (Jordan and Saudi Arabia) became unsustainable, and by the early s the Northern Bald Ibis was believed to be extinct in Syria (Aharoni, ; Safriel, ). This left the last known colony of eastern northern bald ibis at Birecik, Turkey, with ,  breeding pairs in the s (Bowden et al., ). However, the first well-documented observation in recent history was in , when an expedition of the Ornithological Society of the Middle East recorded  northern bald ibises in the Taiz wetlands in Yemen. Birds were also sighted in western and south-western Arabia from  (Brooks et al., ; Schulz & Schulz, ), giving rise to speculation that there was a hitherto undiscovered colony in the region.
In  a breeding colony of  birds was discovered in Syria's Palmyra region (Serra et al., ). Subsequent tagging and satellite-tracking of Syrian birds in  confirmed winter migration through Arabia to the Ethiopian highlands (Lindsell et al., ), c.  km from Addis Ababa, where sightings of northern bald ibises had been relatively common in the th century (Welch & Welch, ; Wondafrash, ).
Historical data (Welch & Welch, ) and satellitetracking, in combination with field surveys in , suggest young ibises disperse around the southern Red Sea basin to the Ethiopian highlands for - years before returning to their natal site in Syria, when close to sexual maturity. It is unknown to what extent juveniles rely on adults to guide them to optimum feeding sites, as is the case with the white stork Ciconia ciconia (Bildstein, ; Rotics et al., ), or whether the greater foraging time typical of juveniles mean they could struggle to find sufficient food (Brooks, ) and become too weak to migrate long distances (as for the white ibis Eudocimus albus ; Del Hoyo et al., ). In Arabia, where juveniles tended to winter (Lindsell et al., ; Serra, ), threats such as hunting and powerlines could be drivers of decline. On the adult wintering grounds in Ethiopia, disturbance and rainfall shortages are potential short-term threats, whereas pesticide use and the conversion of pastures to crops could be significant in the longer term (Serra et al., ).
Following the Palmyra colony's discovery, the site was protected, but migration losses further reduced the tiny population (Serra, , ). During - at least one adult failed to return from each migration, and of  juveniles leaving during - only seven returned ( Table ). Evidence suggests high mortality rates as a result of hunting in Syria and Saudi Arabia, in addition to electrocution on power lines (Serra et al., ). Consequently, the Syrian population declined from seven birds in  to five in ; fledged young failed to reach the wintering site in Ethiopia, reappearing in only very low numbers in subsequent years (Serra et al., ).
Unsustainably high annual mortality (c. %) of young birds outside the breeding grounds was the main reason for juveniles failing to return. Combined with breeding failures in  and  resulting from unknown causes, this reduced the colony to one pair and an unpaired female by . There was an attempted supplementation using Turkish birds in , when two tagged juveniles were released, with both successfully migrating to southern Saudi Arabia and surviving - weeks before dying (Bowden et al., ). In summer , during the Syrian civil war, only one pair returned to the site but did not breed. Three adults had been observed in - at their wintering grounds in Ethiopia, and one in , but only one of these birds returned to the breeding grounds in ; no free-flying birds were seen by late February  (Mahmud Abdallah & Yilma Dellelegn, pers. comm., ). The Syrian population now appears to be functionally extinct, although since  monitoring at the site has been extremely difficult because of security issues ( Table ).

Captive populations
The northern bald ibis has been kept in zoos since the s. The first few birds were imported from the eastern population, but none survived for longer than a few weeks, probably because of poor husbandry. The first surviving northern bald ibises were juveniles imported from Morocco to Basel Zoo in the early s. They became founders of the captive northern bald ibis population (Böhm, ), and now all northern bald ibises in European zoos derive from the western population.
Overall there were  successful imports up to , and a further two during -. Approximately  birds were imported in  and  to Zoo Rheine and Zoo Rabat. All captive birds can be separated into three bloodlines, termed Basel stock, Rheine stock and Rabat stock. All three lines have been mixed in most collections, and currently most zoos keep descendants of at least two bloodlines (Böhm, ).
The captive population increased over  years from  in  to . , birds by , and , registered in the studbook by . This increase is a result of breeding success (c. - offspring per year), low mortality and the longevity of the species.
Since  juveniles have been donated from European Endangered Species Programme sources to three research projects. The Konrad Lorenz Research Station, Austria, received  northern bald ibises from seven locations during - to establish a sedentary semi-captive colony ( Table ),  of which subsequently died. Since then, the colony recovered sufficiently to require exchanges only with another nearby sedentary semi-captive colony in Tierpark Rosegg. The latter was set up with  birds in - (Table ). During - a total of  chicks from nine different zoos were donated to the release project Waldrappteam;  have died while free flying and  returned to captivity because they were injured. Proyecto Eremita in Spain received  birds during the initial stage (-), and in total  birds from  different zoos (including Jerez Zoo);  had died by .
Each year - new colonies are founded in European zoos. Besides the European Endangered Species Programme there are two further studbooks, one in Japan and one in North America, with fewer birds ( and , respectively). There are an estimated  additional non-studbook birds of unknown parentage in other collections.

European research and reintroduction projects
The northern bald ibis was native in Europe until the Middle Ages (Kumerloeve, ; Pegoraro, ), but had become extirpated by , possibly because of habitat loss, climate change and direct persecution (Schenker, ; Böhm & Pegoraro, ). It is likely the European population was migratory, an idea reinforced by Conrad Gessner () in his Vogelbuch description of northern bald ibis distribution, behaviour and fledging time. Centuries later, the northern bald ibis had become a mythical creature until it was rediscovered by scientists at the end of the th century (Rothschild et al., ).
Three hundred years after the extinction of the northern bald ibis in Europe, in the s, European zoos started to keep the species, with the first captive breeding successes in the s at Basel Zoo (Mendelssohn, ). From the late s improved husbandry resulted in the , birds now held in  European zoo colonies (Böhm et al., ).

Austria and Germany: developing reintroduction methodologies
In  Alpenzoo Innsbruck started hand raising and releasing northern bald ibis chicks (Thaler et al., ). The Konrad Lorenz Research Station (a core facility of the University of Vienna) in Grünau im Almtal established the first European free-flying sedentary (but not selfsustaining) northern bald ibis colony in  (Kotrschal, ). This population has maintained - breeding pairs over the last  years ( Table ) and was a starting point for the experimental northern bald ibis restoration projects in Europe. In  hand rearing was stopped, and no summer supplementary food was provided while the birds remained free flying. During - the birds were enclosed in the aviary during the dispersal period (July-September). Since  (Kotrschal, ). The birds use nearby foraging areas, spending the summer in the neighbouring valley, and returning to the aviary in autumn. The fledglings do not migrate, but up to % of them (in -) undergo juvenile dispersal, flying north/north-east and coming back to Grünau at the end of the summer. From  Tierpark Rosegg in Carinthia has also managed a small sedentary, seasonally free-flying northern bald ibis colony, which in  comprised  adult pairs ( Table ). Numbers released and post-release survival during - are presented in Supplementary Tables  and .

Austria and Germany: reintroduction of a migratory subpopulation
In  the project Waldrappteam initiated a -year feasibility study (-) to develop methods for the establishment of migratory northern bald ibis colonies using human-led migration techniques, involving training birds to follow human foster parents in ultra-light aircraft (Fritz et al., ).
During -, a total of  chicks from European zoos have been taken for hand rearing (Fritz et al., ). Before release, contact with the birds is limited to human foster parents to avoid habituation to other humans. During the first weeks, hand rearing takes place at a zoo. Shortly before fledging, the birds are transferred to a field camp in the area where they will be released. Flight training takes  months, starting immediately after fledging, with gradually increased flights up to  km. The human-led migration starts mid August, when the young birds enter a state of migratory restlessness (Bairlein et al., ). The human-led migration routes cross the Alps to a wintering area in Southern Tuscany (Italy), with daily flight stages of up to  km and  hours (Fritz et al., ).
In , the first juveniles hatched in the wild and followed conspecifics from Burghausen to the wintering site in Tuscany (Fritz et al., ). Since then, an increasing proportion migrates independently. Spring migration starts in late March, with the oldest birds departing first. Autumn  migration starts at the beginning of August, but birds regularly remain north of the Alps up to mid September. In the unusually mild winter of  most of the birds delayed migration to late December, resulting in significant mortality (Fritz et al., ). In ,  juveniles followed conspecifics to the wintering site, and one group of seven juveniles and two adults reached the wintering area within  days. On one day the birds covered over  km.
In  the feasibility study evolved into a -year reintroduction project involving eight partners from Germany, Austria and Italy, with significant co-financing from the European Union under the LIFE+ programme. The aim was to establish three migratory breeding colonies with a total of at least  birds (Fritz et al., ). A follow-up project has been proposed, aiming to achieve a viable population by .
By the end of , there were  free-flying, migrating northern bald ibises ( breeding pairs; Table ), belonging to two established breeding colonies at Burghausen, southern Germany, and Kuchl, Austria, with a third breeding colony establishing from  in Ueberlingen at Lake Constance, southern Germany. A major challenge for the project are losses caused by illegal hunting during the autumn migration (Fritz, ). During the feasibility study (-) up to % of the annual losses were attributable to illegal hunting in Italy (Fritz, ; Fritz et al., ). Since , illegal hunting accounts for % of the annual losses, with electrocution at medium voltage power poles causing % of the mortality (Supplementary Table ). Criminal and civil cases brought against an Italian northern bald ibis hunter raised awareness of the issues in Europe (Fritz et al., , ). Since  a total of  juveniles fledged in the wild at the two breeding sites (Table ), a mean of . per nest.
Details of annual releases of - birds are presented in Supplementary Table . Overall annual post-release survival in the first year is c. %. All released birds are individually marked and tracked (Fritz et al., ).

Spain: reintroduction of a sedentary population
In  a joint project between Zoobotánico Jerez and the Andalusian government established Proyecto Eremita, to develop techniques for the release of captive-bred northern bald ibises (López & Quevedo, ), with the aim to establish a sedentary, free-flying colony in southern Spain. The release area in south-western Cádiz province has coastal and inland cliffs, and extensive foraging areas, allowing the birds to feed all year round. Birds are hand-reared by human foster parents wearing black shirts and ibis-shaped cycle-helmets (Quevedo et al., ; Quevedo, ). Once a free-flying hand-reared group of ibises was established in the area, groups of parent-reared juveniles were kept in a neighbouring release aviary for several months (October-January), then released to join the free-flying birds. This technique facilitates the integration of additional birds coming from collaborating zoos that are members of the European Endangered Species Programme for the northern bald ibis.
Annual releases of - birds during - (Table ) demonstrated the ability of captive bred northern bald ibises to survive after release and to establish a wild colony in the region of La Janda, Cádiz, south-west Spain. At present post-release survival in the first year is c. %, with the main causes of mortality being predation by eagle owls Bubo bubo and electrocution (Supplementary Table  ; López & Quevedo, ). The first nesting of released birds was in  at Tajo Barbate, a coastal cliff  km from the release site, where breeding occurred annually until . In  a new breeding site was established beside a busy road at Barca de Vejer,  km from the release site, and this has become the main colony (López & Quevedo, ). In  a new breeding site was established in Castilnovo, a coastal tower,  km from the release site. By  there were two active breeding sites ( Table ) within  km of the release site, with a total of  birds in the wild, including  breeding pairs (Supplementary Table ). The four breeding sites were naturally colonized by northern bald ibises; no boxes or ledges were provided. All birds are individually marked, and some carry transmitters. From the hand-reared groups, some juvenile dispersal, mainly southwards, has been recorded, in contrast with the wild-hatched juveniles that have remained in the area with their parents and the rest of the group. In July  the Environment Ministry of Spain formally approved the work as a northern bald ibis reintroduction project. The project has two phases: () the consolidation of the existing colonies, and () applying the techniques refined by Proyecto Eremita to establish a second population of northern bald ibises in Cabo de Gata Natural Park,  km away, also in Andalusia. To date, Proyecto Eremita has succeeded in establishing a sedentary population without supplementary feeding or other intensive interventions. The population is not yet self-sustaining, so supplementation with additional birds is continuing, improving genetic variability and aiming for a minimum of  birds and  breeding pairs by .

Conservation progress
Management of the remaining population in coastal Morocco has generated the first sustained increase in the natural wild population in historical times. Although the northern bald ibis was categorized as Critically Endangered in the IUCN Red List up to , the positive trend in Morocco since  justified recategorization from Critically Endangered to Endangered (BirdLife, ). Management by the Moroccan National Parks authority has been key to this success, along with regulation of development pressures and increasing tourism. Maintaining the traditional agricultural areas and preventing disturbance around the breeding sites will be ongoing challenges to support this largely resident northern bald ibis population.
By  the last wild migratory breeding northern bald ibis pair had disappeared from the Middle East, although they survive in semi-wild conditions and retain the ability to migrate, and thus could serve as a source of founders for the re-establishment of a migratory eastern population. The lessons learned from the European reintroduction projects provide cause for optimism that sedentary populations (Spain) and even migratory populations (Austria) can be reestablished, and could inform the reintroduction of species with similarly migratory and sedentary phenotypes.
In  the International Advisory Group for the Northern Bald Ibis was founded. Its principal aims are to facilitate good communication and information exchange amongst all parties working with the northern bald ibis. The group distributes new and accumulated knowledge about the species by organizing periodic meetings, publishing newsletters and the meeting proceedings, and maintaining its own website with a download section for reports. The group's activities have been recognized by the Agreement on the Conservation of African-Eurasian Migratory Waterbirds (AEWA) that initiated and sponsored the first Species Action Plan for the northern bald ibis in . A revision of this action plan was produced from the st Meeting of the AEWA Northern Bald Ibis International Working Group in  in Saudi Arabia, and finalized in  (Bowden, ). It states that because of significant progress in methodology over the past decade, conservation translocation has become a viable approach to improve the status of the species in the wild (Bowden, ). The two European release projects (LIFE+ project and Proyecto Eremita) have so far developed largely independently of the AEWA process, but have demonstrated the potential to develop reintroduction projects within the more recent former range in Algeria, Turkey and Syria. The first steps for this were outlined in the updated implementation plan for Algeria (AEWA, ). The AEWA Northern Bald Ibis Working Group is composed of one nominated senior government representative plus a national expert from each of the range states, with a regional coordinator for the eastern and western populations, and an overall coordinator. Experts, usually from the International Advisory Group for the Northern Bald Ibis, are invited to periodic meetings in line with the main issues on the agenda for the implementation of the action plan. The Moroccan government hosted the most recent meeting in November , at which -year implementation plans were developed and agreed for each country (AEWA, ). Significant progress has thus been made both in maintaining and steadily increasing the small western population in Morocco, and in developing the methodology and prospects for establishing a sedentary population in southern Spain and possibly also a migratory population in central Europe.

Future perspectives
There is a need to evaluate the suitability of potential and former breeding sites and associated feeding areas for both the western (Morocco, Algeria) and eastern (Syria, Turkey, Yemen, Iraq) populations. This is especially needed in Morocco where the in situ population increase means there is potential for natural recolonization beyond the currently occupied sites. In addition, there is evidence that birds from the Spanish project can cross to Morocco, with sightings of lone birds in both  and , and multiple observations in  and  (Muñoz & Ramírez, ; Bowden et al., ).
Every year some birds at the Birecik Breeding Station, Turkey, disappear before enclosure; they may not be dead but, rather, migrating south. Release trials of satellitetracked birds have shown promising results for potential reestablishment of a fully wild Turkish population, although because birds inevitably pass through Syria where civil conflict means personal firearms are ubiquitous, this may become viable only after the region has stabilized politically.
The captive population of western origin is managed through studbooks and continues to increase slowly, although individual captive colonies need careful breeding management as some bloodlines are already over-represented and aviary space is becoming limited. However, the age structure and sex ratio indicate that the captive population is still viable and . % of the birds are of optimal breeding age. This means sufficient genetically appropriate birds can be made available for reintroduction projects, but cooperation between holders is essential to control inbreeding and maintain genetic variability.
Finally, our understanding of methodology for the reintroduction of this species, for both sedentary and migratory populations, has advanced significantly since . The positive trend for the remaining wild Moroccan population is a result of the measures taken by the Moroccan government together with partner NGOs. The recent signs of recolonization of former areas, if maintained, will be an important consolidation and a positive step in the recovery of the species. Meanwhile, pressures on current feeding and breeding habitats are increasing. Interventions such as wardening, water provisioning and law enforcement to prevent illegal construction activities thus remain a top priority and may need further reinforcement. Availability of viable management options such as reintroduction offers significant opportunities for the species and provides hope that the re-establishment of further self-sustaining populations is possible.
Acknowledgements We thank the numerous funders and supporters of the enormous body of work this paper encompasses although these are too many to mention individually. Two anonymous referees helped develop earlier drafts of this paper, and Amina Fellous commented on the Algeria text. We are unable to mention all the field workers that have contributed to gathering the information presented, but due to the harsh conditions and the difficulties on the ground, we would particularly like to acknowledge the full team of wardens in Morocco who have dedicated their lives to protecting and monitoring the birds. The team continues to do this, often in very tough field conditions, and Mohammed El Gadrouri and Abdallah Essamar deserve special mention having both been fully involved for 25 years. Likewise, in Syria, the wardens and rangers there have served often through extraordinarily difficult times, well beyond the call of duty, and although many more have been involved, we would like to pay special tribute to Mahmud S. Abdullah and Ahmed Khaber Abdallah.
Author contributions Lead on compiling the information and writing: CB, CGRB, PJS; contribution of updated information from the respective range states and projects, and revisions: TH, WO, MEB, MAQ, JF, CY, JML, JFO, DF, MU; overall coordination: CGRB.

Conflicts of interest None.
Ethical standards This work abided by the Oryx guidelines on ethical standards.