The Ultramarine Lory (Ultramarine Lorikeet, Marquesas Lorikeet, Pihiti) Vini ultramarina is one of the most threatened insular Lory species. Endemic to the Marquesas Islands, where it probably once ranged throughout the archipelago, it has been nearly extirpated from all but the tiny island of Ua Huka. Given the vulnerability of a single population inhabiting one small island, and the decline of the species in recent years, establishment of the Ultramarine Lory on another less disturbed island within its historic range has been proposed as a conservation strategy. This paper describes a 1991 survey evaluating the status of the Ultramarine Lory in the Marquesas Islands, and three subsequent translocations of birds from the island of Ua Huka to the island of Fatu Hiva. Twenty-nine birds were relocated during the period from 1992 to 1994 at the request of the Delegation de L'Environnement, French Polynesia. A preliminary survey, prior to the third translocation, indicates that previously transferred birds are surviving and may be reproducing; an intensive survey is planned in 1997.

We assessed the impact of forest fire on the nesting success of the endemic Red-knobbed Hornbill Aceros cassidix in north Sulawesi, Indonesia. Pre- and post-fire breeding activity was monitored at 60 nest trees. Low rainfall and possible reductions in fruit abundance did not appear to have an effect on initial nest uptake or on subsequent chick production at non-burnt nests, which were comparable to previous years. However, there was an adverse effect on productivity at nests that directly experienced fire and the accompanying heat and smoke in the vicinity, resulting in a post-fire nest success (62%) that was significantly lower than in previous years (78.6–80%). Population recruitment dropped by an estimated 47% to 0.17 fledglings/breeding pair. Our data suggest that, given the reproductive profile of this species and the ability of fire to reinforce its own negative effects on forest habitats, Red-knobbed Hornbill populations may be vulnerable to the potential effects of repeated fires.

Identifying the fate of birds’ nests and the causes of breeding failure is often crucial for the development of conservation strategies for threatened species. However, collecting these data by repeatedly visiting nests might itself contribute to nest failure or bias. To solve this dilemma, automatic cameras have increasingly been used as a time-efficient means for nest monitoring. Here, we consider whether the use of cameras itself may influence hatching success of nests of the Black-tailed Godwit Limosa limosa at two long-term study sites in northern Germany. Annually between 2013 and 2019, cameras were used to monitor godwit nests. In 2014 and 2019, nests were randomly equipped with cameras or not, and nest survival checked independently of the cameras. Nest-survival models indicated that survival probabilities varied between years, sites and with time of the season, but were unaffected by the presence of cameras. Even though predation is the main cause of hatching failure in our study system, we conclude that predators did not learn to associate cameras with food either when the cameras were initially installed or after they had been used for several years. Cameras were thus an effective and non-deleterious tool to collect data for conservation in this case. As other bird species may react differently to cameras at their nests, and as other sets of predators may differ in their ability to associate cameras with food, the effect of cameras on breeding success should be carefully monitored when they are used in a new study system.

In 1986 a single Norfolk Island Boobook Owl Ninox novaeseelandiae undulata remained. As part of a re-establishment programme, two male New Zealand Moreporks N. n. novaeseelandiae were introduced, one of which survived to pair with the female in the wild and breed successfully. By 1995 the population numbered 12 or 13 individuals of which seven were second generation (F2). However, there were only two breeding pairs. As the 11 hybrids could not be sexed using morphometrics we developed a molecular method based on a recently described avian polymerase chain reaction (PCR)-based sexing technique. The population was found to contain six females and five males. A scarcity of mature males was established as the main factor slowing the recovery effort.

The Red-legged Kittiwake Rissa brevirostris is a small gull which is restricted to four breeding locations, all in the Bering Sea (Pribilof Islands, Bogoslof Island, Buldir Island, and the Commander Islands). In the mid-1970s, when the earliest counts were made at most sites, approximately 260,000 birds were present at breeding colonies. Subsequent counts indicated that populations at Bogoslof and Buldir had increased by the early 1990s. Data for the Commander Islands were inadequate to make strong conclusions about trends but there were no indications of substantial change between the 1970s and 1990s. In contrast, Red-legged Kittiwakes have declined in the Pribilof Islands by approximately 50% since the mid-1970s, a matter of concern because this island group once contained more than 80% of the world's population. Research is needed to determine causes for declines in Red-legged Kittiwakes and population monitoring should continue to provide a basis for conservation measures.

Alagoas Curassow Mitu mitu is considered extinct in the wild, having previously inhabited a small area in north-eastern Brazil that has since been replaced by sugar cane farms. Around 50 birds possessing morphological features of this species are still alive in captivity in Brazil, all being descendants from a breeding programme started in 1979, using offspring from a single male and two females captured from the wild. However, this captive population also includes some hybrids with the congeneric Razor-billed Curassow M. tuberosa and their descendants. Furthermore, the validity of Alagoas Curassow as a species is questionable. We used two molecular markers to study the validity of this taxon as a species, to detect potential hybrids present in the stock, and to estimate genetic variability among the remnant specimens. The analysis of 760 base pairs from mitochondrial cytochrome b and control region sequences revealed that at least three of the 20 birds analysed had sequences identical to those of Razor-billed Curassow. The other 17 birds presented sequences that diverged 2.6% from Razor-billed Curassow. Moreover, a sample from an Alagoas Curassow museum skin collected from the wild in 1951 had cytochrome b sequences identical to those of the 17 birds. These results confirm the Alagoas Curassow as a valid species. DNA fingerprinting profiles of the 20 descendants from the Alagoas Curassow breeding population showed that this group of birds is depauperate in genetic variability. There was an increase in genetic variability of birds born after 1990, attributed to the hybrid mating of Alagoas Curassow with Razor-billed Curassow. We suggest that birds born before 1990 should be handled separately from the others. However, if a decrease in chick harvest among birds of this group is detected due to inbreeding depression or ageing, cross-breeding between this group and the group of birds most closely related to it should be considered in order to enrich the progeny with the Alagoas Curassow genome.

The breeding areas of the Critically Endangered Slender-billed Curlew Numenius tenuirostris are all but unknown, with the only well-substantiated breeding records being from the Omsk province, western Siberia. The identification of any remaining breeding population is of the highest priority for the conservation of any remnant population. If it is extinct, the reliable identification of former breeding sites may help determine the causes of the species’ decline, in order to learn wider conservation lessons. We used stable isotope values in feather samples from juvenile Slender-billed Curlews to identify potential breeding areas. Modelled precipitation δ2H data were compared to feather samples of surrogate species from within the potential breeding range, to produce a calibration equation. Application of this calibration to samples from 35 Slender-billed Curlew museum skins suggested they could have originated from the steppes of northern Kazakhstan and part of southern Russia between 48°N and 56°N. The core of this area was around 50°N, some way to the south of the confirmed nesting sites in the forest steppes. Surveys for the species might be better targeted at the Kazakh steppes, rather than around the historically recognised nest sites of southern Russia which might have been atypical for the species. We consider whether agricultural expansion in this area may have contributed to declines of the Slender-billed Curlew population.

The Faunal Reserve of the Lobéké area of south-eastern Cameroon is an important area for forest conservation. One day, it is hoped, it will be part of an international unit, in association with two protected areas in neighbouring countries (Dzanga-Ndoki National Park, Central African Republic and Nouabalé-Ndoki National Park, Congo). With a view to assessing the status of its avifauna in particular, 24 days were spent in three short surveys from 1997 to 1999. The main forest type is semi-evergreen, with an open canopy; the only natural savannas are small saline swamps. The total of 305 species of bird recorded includes a forest nightjar not yet identified (but more likely to be the rare Itombwe Nightjar Caprimulgus prigoginei than a new species) and the Dja River Warbler Bradypterus grandis. The latter is a species confined to Rhynchospora swamps and had not been re-located in Cameroon since it was first collected in 1914 west of the Dja river; the extent of suitable habitat in Lobéké makes this site the most important to date for its conservation. Other rare or little-known forest species recorded include Olive Ibis Bostrychia olivacea, Sandy Scops Owl Otus icterorhynchus, Zenker's Honeyguide Meligomon zenkeri, Tessmann's Flycatcher Muscicapa tessmanni and Yellow-capped Weaver Ploceus dorsomaculatus. Barred Owlet Glaucidium capense is locally common in open-canopy forest: this population was only recently discovered in central Africa and its taxonomic relationships have yet to be determined. We include a brief comparison with the avifauna of adjacent Dzangha-Ndoki National Park and Nouabalé-Ndoki National Park. A feature of the Lobéké avifauna is the presence of a few species normally associated with forest at higher altitudes (such as White-headed Wood Hoopoe Phoeniculus bollei, Uganda Woodland Warbler Phylloscopus budongoensis, Black-throated Apalis Apalis jacksoni), and perhaps absent from lower-lying Nouabalé-Ndoki.

We surveyed birds in two remnant patches of montane evergreen forest landscapes differing in intensity of habitat fragmentation, land use patterns and development. Present landscape configurations in Mae Tuen and Om Koi show that both became heavily fragmented (Table 1, see also Figure 1) between 1954 and 1996. The low abundance at Om Koi of large frugivores, such as Brown Hornbills Ptilolaemus tickelli and Great Hornbills Buceros bicornis, and their lack at Mae Tuen, are probably effects of prolonged fragmentation.

Hen Harrier Circus cyaneus and Short-eared Owl Asio flammeus are open-country birds of prey with overlapping distributions. Although both species face similar conservation threats across their ranges, work to date has largely been undertaken at a national scale with few attempts to collate and assess factors relevant to their conservation at an international scale. Here we use an expert knowledge approach to evaluate the impact of conservation threats and the effectiveness of conservation strategies for each species across Europe. We report results of responses to a questionnaire from 23 Hen Harrier experts from nine countries and 12 Short-eared Owl experts from six countries. The majority of responses for both species reported declines in breeding numbers. The perceived impact of threats was broadly similar for both species: ecological factors (predation, extreme weather and prey availability), changes in land use (habitat loss and agricultural intensification) and indirect persecution (accidental nest destruction) were considered to be the greatest threats to breeding Hen Harrier and Short-eared Owl. Short-eared Owl experts also highlighted lack of knowledge and difficulties associated with monitoring as a major conservation challenge. Despite broad-scale similarities, geographical variation was also apparent in the perceived importance of conservation threats, with some threats (such as direct persecution, large-scale afforestation or habitat degradation) requiring country-specific actions. Implementation of different conservation strategies also varied between countries, with the designation of protected areas reported as the most widespread conservation strategy adopted, followed by species and habitat management. However, protected areas (including species-specific protected areas) were perceived to be less effective than active management of species and habitats. These findings highlight the overlap between the conservation requirements of these two species, and the need for collaborative international research and conservation approaches that prioritise pro-active conservation strategies subject to continued assessment and with specific conservation goals.

Systematic bird observations were performed during autumn 2006 in the migration bottleneck area of Zeit Bay, Egypt (27.73°N, 33.51°E), situated on the coast of the Gulf of Suez, opposite southern Sinai. Owing to the strong northerly winds in that area, considerable pressure exists to construct wind farms. A total of 145,432 soaring birds, including 134,599 storks and 9,376 raptors over a period of 453.6 hours were observed between 20 August and 29 October by two teams working daily in two shifts. White Stork Ciconia ciconia and European Honey Buzzard Pernis apivorus were the most numerous species (91.4% and 5.7% respectively) of all soaring birds observed. As both species migrate early, relatively few birds were recorded passing through in October. A total of 57,179 soaring birds were observed within a range of 2.5 km from the observers, and of these 32,248 were flying at heights up to 200 m. Few individuals of the rapidly declining breeding populations of Egyptian Vulture Neophron percnotperus (‘Endangered’) and Lesser Kestrel Falco naumanni (‘Vulnerable’) passed (8 and 33 respectively). Pallid Harriers Circus macrourus (‘Near Threatened’) however, were more numerous (100) than at other migration sites along the East African-West Asian Flyway (Israel, southern tip of Sinai, Suez and Bab-el-Mandeb). Western Marsh Harrier Circus aeruginosus and Montagu’s Harrier Circus pygargus were also more numerous than at these sites. Some other large migrant species, observed at Suez, Bab el-Mandeb or in Israel, were seen only in small numbers or not at all at Zeit Bay in autumn. Among them, the Steppe Eagle Aquila nipalensis being the best known example of loop migration, entering Africa at Suez and Bab el-Mandeb, is for this reason not expected in autumn at Zeit Bay, while in spring it is a regular migrant. The number of Black Storks Ciconia nigra observed accounted for 4.8% of the flyway population, that of White Storks for 33.3%, Great White Pelicans Pelecanus onocrotalus for a further 3.5% and European Honey Buzzards for 0.8%. The extrapolation of the number of birds passing through the study area during the migration season, suggests that it was used by 92% of the White Stork, 12% of the Black Stork, 5% of the Great White Pelican, 4% of the European Honey Buzzard and 2% of the Pallid Harrier flyway populations in autumn 2006.

Karakelang, largest of the Talaud Islands, Indonesia, was visited in 1999 with the aim of producing population estimates of globally threatened and Restricted-range bird species. Density estimates were calculated from primary forest and secondary habitats using the variable circular plot method. Eight of the nine restricted-range or threatened species resident on Karakelang were recorded and density estimates calculated for five of these. Two species of rail are endemic to Karakelang, Talaud Bush-hen Amaurornis magnirostris and Talaud Rail Gymnocrex talaudensis. The bush-hen was encountered in secondary habitat but occurred at higher densities in primary forest and the estimated population was 2,350–9,560 birds. The less vocal, cryptic G. talaudensis was noted just twice, in primary forest close to rivers. Both have small populations, are threatened by habitat degradation, hunting, and possibly predation by introduced rats and should be classified as threatened. Population densities of the Endangered, endemic Red-and-blue Lory Eos histrio had remained stable since the last survey in 1997. They occurred at higher densities in primary forest, but were also common in human-made habitats. The estimated population was 8,230–21,400 birds, threatened by habitat loss and trapping for the wild bird trade. Although more commonly encountered in forest, the Near Threatened Blue-naped Parrot Tanygnathus lucionensis was found at higher densities in secondary habitats, and the population was estimated at 8,130–20,700 birds. The Restricted-range Blue-tailed Imperial Pigeon D. concinna was very common on Karakelang (14,500–27,700 birds), whilst the Vulnerable Grey Imperial Pigeon D. pickeringii was recorded rarely; both species were more frequently recorded in primary forest. The endemic Talaud Kingfisher Halcyon enigma occurred at low densities (5,290–8,690 birds), in primary forest and adjacent disturbed areas. It is heavily dependent on primary forest and threatened by habitat loss, and should be classified as Vulnerable. Approximately 350 km2 of primary forest on Karakelang is protected, 250 km2 as a wildlife reserve. However, at present, management is absent and forest is threatened by agricultural encroachment, illegal logging, and fire. Given that all endemic and threatened species were encountered more frequently or occurred at higher densities in primary forests, future conservation efforts should target these protected areas. Management of reserves should involve local stakeholders from government and representatives of island communities. Wildlife trade is a major threat to Eos histrio and strict enforcement of the species' protected status should include monitoring of roost sites, patrols of harbours and markets in Talaud, Sulawesi and the Philippines, and the control of Philippine fishing boats involved in the wild bird trade. Further species-specific research and monitoring is also required.

The Ivory-billed Woodpecker Campephilus principalis has suffered from destruction of its habitat over the whole of its range, being last recorded in eastern Cuba (Ojito de Agua) in 1987. The only chance for the species was that the last-known birds had moved to adjacent areas, or that other populations existed in other mountainranges. Between February and May 1993 an extensive search for the species and for suitable habitat was carried out in mountain areas of eastern Cuba. In the surroundings of Ojito de Agua no suitable large old-growth forests remain, and evidence for the absence of Ivory-billed Woodpeckers from the area is discussed. In the region surveyed, only around Pico la Bayamesa in the Sierra Maestra does an almost untouched forest still stand, but this is probably outside the species's original range. Ojito de Agua and Pico la Bayamesa should nevertheless be fully protected. The last known population of Ivory-billed Woodpecker appears to have become extinct in 1989 or 1990. Scaling of bark from dead pines, formerly considered a sure sign of the species's presence, may be caused by weather damage.

Following a recent assessment of the distribution and habitat use of Gurney’s Pitta in Myanmar (Burma), further extensive surveys were undertaken in 2010, 2011 and 2012. These have extended the species’ known altitudinal limit to between 250 m and 300 m asl and its latitudinal limit to above 12.5°N, around 80 km north of the northernmost historical record, although the species was recorded far less frequently at higher altitudes and latitudes. Birds were recorded in a range of forested habitats, from intact primary forest to secondary and bamboo forest, with no significant difference between major forest types in the likelihood of occurrence. Niche envelope modelling (MaxEnt) suggested a total range size in Myanmar of 3,379 km2, and did not identify any potentially suitable areas in adjacent parts of Thailand. The species’ preference for warmer, wetter areas on flat ground, conditions ideal for growing oil palm and rubber, suggest that its distribution is likely to contract in the near future. The entire range of Gurney’s Pitta in Myanmar falls within the part of the country most suitable for commercial oil palm production, although the projected yields within its range are low to moderate. Field surveys found evidence of rapid recent deforestation and high levels of hunting and trapping in many parts of the region. The species’ range in Myanmar does not overlap with any protected areas. The protection of southern Myanmar’s biodiversity will require substantial investment by foreign conservation interests, sympathetic land-use planning and the strengthening of environment legislation. Protection of extensive tracts of lowland forest within the range of Gurney’s Pitta, particularly the proposed Lenya National Park and the adjacent Ngawun and Htaung Pru Reserve Forests, is urgently needed. Conserving these areas will also protect populations of other globally threatened bird and mammal species.

Fieldwork on the distribution, status and ecology of birds was conducted in the northern Sierra Madre mountain range, Luzon, Philippines, during March–May 1991 and March-May 1992. The findings show the area to be one of the most important for conservation of threatened species of birds in all Asia. The results are here combined with evidence from earlier surveys by other searchers. Fourteen threatened and 18 near-threatened species are now known from the area. This paper reports on all the threatened and near-threatened resident species of the island of Luzon, with special emphasis on their occurrence in the Sierra Madre mountains. In addition, the paper treats species with very limited global distribution that breed in Luzon, and lists species of forest birds endemic to the Philippines that have not previously been reported from the Sierra Madre mountains. Maps show the known sites for 17 species of special concern for conservation. New data on the altitudinal distribution of threatened and near-threatened species suggest that it is essential to protect primary forest at all elevations.

The Northern Silvery Grebe Podiceps juninensis is an aquatic bird species considered Vulnerable to extinction in Ecuador and Endangered in Colombia. Globally the species is considered Near Threatened. Herein we report recent changes in distribution and population trends for the subpopulation in the northern Andes (Ecuador, Colombia), which is small and isolated. The Northern Silvery Grebe has undergone two documented extirpations, and at least one colonisation event, over the last three decades. More than 90% of the subpopulation relies on two wetlands, La Mica and Colta lakes in Ecuador, with the latter facing an accelerated sedimentation process which jeopardizes the fate of the wetland, and the species, in the mid-term. The population of La Mica lost 80% of its birds in 2000 after the damming of the lake, but then remained stable over our census period (2004–2013). In contrast, the Colta population results from a recent colonisation in the early 2000s, and has experienced an exponential growth since then (annual population multiplication rate: 1.464). The growth of this population is so fast that it must include an important contribution of immigrants. The temporal coincidence between the drop in population size at lake La Mica and the colonisation of lake Colta suggests that the Colta population was founded by immigrants from La Mica. Finally, the absence of current reports of the species at wetlands in Colombia suggests that the species remains at very low numbers, bolstering the shrinkage in northern distribution.

Cyclonic storms (often called hurricanes, typhoons, or cyclones) often cause population declines in vulnerable bird species, and the intensity of these storms appears to be increasing due to climate change. Prior studies have reported short-term impacts of hurricanes on avifauna, but few have examined long-term impacts. Over two decades (1993–2018), we periodically surveyed a subspecies of West Indian Woodpecker Melanerpes superciliaris nyeanus on San Salvador, a small island in The Bahamas, to determine its distribution on the island, habitat use, and effects of hurricanes on abundance and population size. We conducted passive and playback surveys, supplemented with mist-netting. Woodpeckers were found only in the northern part of San Salvador, despite extensive surveys throughout other accessible areas of the island. Birds occupied areas with taller coppice adjacent to sabal palm Sabal palmetto groves, which were used for nesting. After hurricanes with >160 kph winds passed over San Salvador, woodpecker densities declined to 35–40% of pre-hurricane densities, but generally recovered back to pre-hurricane densities within 2–3 years. Based on an estimated density of woodpeckers within a ~1,400 ha occupied area, we calculated a population size of approximately 240 individuals (CI = 68-408). However, the population declined to far lower numbers immediately following hurricanes. Under IUCN Red List criteria, M. s. nyeanus classifies as ‘Critically Endangered’, and could be especially sensitive to future hurricanes if they occur at a high enough frequency or intensity to prevent the population from rebounding. Given the small size, isolation, and vulnerability of this population, we recommend preservation of the core habitat, continued monitoring, and further research. Our study shows that small, threatened bird populations can be resilient to the effects of hurricanes, but increased intensity of hurricanes, in combination with other threats, may limit this resilience in the future.

The status of the Comoro endemic birds is briefly reviewed. Field data were gathered during expeditions in 1981, 1983, 1985 and 1989.

Some native birds (Cyanolanius madagascarinus, Dicrurus fuscipennis on Ngazidja, Treron australis on Mwali, Accipiter francesae, Otus rutilus, Leptosomus discolor on Ndzuani, Dicrurus waldenii on Maore) are so rare that they are certainly highly endangered. The altitudinal stenotopy of a few other species makes them also very vulnerable, e.g. Zosterops mouroniensis in the tree-heath zone on Mount Karthala above 1,750 m. A preliminary analysis of habitat preference on Ngazidja suggests a relatively large flexibility of most forest species between different types of forest, including those with altered vegetation structure, but demonstrated their absence in plantations without large trees and/or natural shrubs. Comparison of density of species in Ngazidja and Mwali and an overview of the avifaunal composition on all four islands are the basis for a discussion on colonization history (with possible extinctions), historical change in avifaunal composition, the suitability of each island for the different species and the completeness of their avifauna. The results are checked for possible threats and necessary actions for conservation. Competition by introduced species does not yet seem to be a problem. But importation of foreign animals must be banned and hunting must be regulated (especially for pigeons). Habitat loss (overexploitation of the forest) will soon become a serious problem for the stenotopie forest birds. A suggestion is made for delimiting a national park on Ngazidja and Mwali. Most controversial is the development of tourism. For long-lasting results, education of local people is necessary, together with full liaison with the authorities.

Migratory birds are critically dependent on adequate wintering habitats for their long-term survival. Cantabrian farmland, a mixed agricultural landscape extending across the coastal lowlands of northern Spain, constitutes an important wintering area for many short-distance migrants coming from central-western and northern Europe. Unfortunately, the Common Agricultural Policy of the European Union and national afforestation schemes have resulted in a massive replacement of farmland by pine Pinus spp. and eucalypt Eucalyptus sp. plantations. This work assesses the importance of Cantabrian farmland as wintering grounds for short-distance European migrants and for wintering species that originate in nearby woodlands. We examined the seasonal changes in passerine bird populations in the Cantabrian region and used winter ringing recoveries obtained in the area to evaluate the contribution made by European migrants to winter populations. Bird communities were surveyed along 299 500-m long transects distributed between 67 farmland patches, 67 lowland forests and 14 upland forests. Winter assemblages were more diverse and species more abundant in farmland than in lowland or upland forests, whereas these differences were smaller in the spring. Bird numbers in farmland tripled in winter, numbers increasing by about 6.9 million birds compared to breeding populations. Most of this increase was accounted for by species that also bred in the region and that considerably increased their abundance (65.6% of all wintering birds, with the Chaffinch Fringilla coelebs responsible for 31.4% of the total increase) and by five exclusively wintering species (34.4%, with the Meadow Pipit Anthus pratensis accounting for 25.2% of the total increase). The main bulk of this increase is caused by the influx of European migrants. The importance of halting the current spread of eucalypt plantations (which increased over 400% over the past 30 years) and of applying more effective agri-environment schemes to achieve appropriate farmland conservation is discussed.

The identification of threats to migratory species of conservation concern and the relevance of protected areas for them is often biased towards breeding areas. The European Roller Coracias garrulus is a long-distance migrant experiencing a pronounced decline throughout its breeding range, which has been attributed to the degradation of open agricultural habitats. However, its conservation status in non-breeding areas in Africa remains unstudied. Land cover change is a major threat affecting migratory birds in their wintering grounds, therefore identifying important areas for their protection at this stage is a priority. Here we used occurrence data during the wintering season and ecological niche models to identify key land cover and areas used by Rollers in Africa. First, we used 33 filtered locations from six satellite-tracked birds breeding in Spain to describe suitable wintering areas for the Spanish population (westernmost part of the Eurasian breeding range). We also used 1,167 occurrence data in southern Africa from open-access databases and bird atlases to characterise the overall wintering range of the species. The Spanish population occupied a relatively small area in the north-western part of southern Africa, and a narrow range of land covers. Open grassland, less steep areas and those with sparse tree cover are correlated with suitability. In all, 18.06% of suitable wintering areas for the Spanish population overlapped with protected areas. The overall population of Rollers occupied a wider area and range of land cover. Tree cover was the most important variable affecting suitability, with areas without tree cover being the least suitable. We found that 9.58% of suitable wintering areas for the overall population overlapped with protected areas. Our results suggest that Rollers from different origins (breeding populations) use separate, but overlapping, wintering areas and may have different habitat requirements, and therefore, population-specific conservation strategies in these areas might be needed to fully protect the species.