Posted on

Data collection

There has been no systematic collection of data for this Atlas as took place, for example, in the Southern African Bird Atlas Project (Harrison 1997). Instead I contacted all persons I knew of who had made ornithological observations in Mauritania and neighbouring areas and I also studied as much as possible of the ornithological literature of the country and surroundings. By these means I obtained data from 122 sources of which 86 furnished data from Mauritania. These sources are identified on the Sources page. The sources were of several types: published books and articles, unpublished reports, database dumps from organizations and emailed notes or tables from individual or groups of observers. In 2007, a colleague and I designed a data entry form of 27 columns, each row of which would contain one record. This was translated into French by Raphaël Bussière. A few sources completed the data entry form with records. I transcribed records from the other sources to data entry forms, often by automated procedures controlled by macros. The sources table gives the number of records so derived from each source, which varied from 1 to 17,186. The collection of data went on from March 2007 to December 2017.

From the data of each observation I derived a point record that included the latitude and longitude of the observation. Most of these were accurate to within 1 minute though about 1% of the records were of lesser precision. However, all records used were judged to be sufficiently precise to be acceptable for the plotting accuracy of the maps (30 minutes). In 16 cases of rare birds no record was judged to be within these accuracy bounds so no map was plotted for those species. In addition to the records indicating the observed presence of a bird, other records were created, as justified, to indicate the existence of evidence for Holarctic wintering (13.5% more records) and for breeding (2.0% more records).

Species distribution maps

From the point records I obtained a composite file covering all species which included, for each species: scientific name followed by latitude and longitude of each record and a code to indicate if the record was for presence only or for evidence of breeding or Holarctic wintering. This file was input to the software package DMAP ( which produced one distribution map for each species. On each map is plotted in green the half-degree squares where at least one record of the presence of the species has been obtained. The number of records is not taken into account as this depends mainly upon the number of observers. The green square is overwritten, if the evidence exists, to show with a blue circle evidence of Holarctic wintering as well as the location of breeding evidence with a red circle.

Holarctic wintering is defined as a record in December or January (Browne 1982) of a known Holarctic migrant. Most Holarctic migrants to Mauritania come from northern Eurasia (Palaearctic) but some individuals from North America (Nearctic), e.g. Turnstone Arenaria interpres (Isenmann 2006) and Northern Wheatear Oenanthe oenanthe (Bairlein et al 2011). Breeding evidence is indicated with two categories: “strong” and “weaker”. “strong” means that a reliable observer has reported eggs, chicks or newly fledged young, or that there have been published reports of breeding at a given site over several years. “weaker” means that only such behaviour as nest building, injury-feigning, territorial display or territorial song has been observed, or that there have been isolated published reports of breeding without further evidence being provided.

These maps can be viewed for 517 species via links on the Maps page. The links are in the same order as, and next to, the page number of the species in Isenmann et al 2010. Thus, the first column is page number, the second column scientific (Latin) name with a link to the corresponding map, the third column French name and the fourth column English name. 16 species have no map because no point record of sufficient accuracy could be obtained. Also new records have given rise to maps for 23 species not included in Isenmann et al 2010.

Website: Created with Microsoft FrontPage and LeechFTP

Reliability of reports

As with anyone compiling data from many sources, I have been faced with the question of their reliability. The latest data included in this Atlas, those for the second version gathered in 2016 and 2017, have been subject to the closest scrutiny. I produced distribution maps of these data for each species and compared each map with that for the same species in the first version of the Atlas. For any records that caused wide discrepancies between the two maps, I contacted the source and asked for evidence of the identification. If satisfactory evidence (usually photographic) was supplied I accepted the record; otherwise deleted it.

A few decades ago, B. Lamarche published a bird atlas covering Mauritania and Mali (Lamarche 1993) and annotated checklists of the birds of these countries (Lamarche 1980-1981, Lamarche 1988). Doubts about the reliability of some of these publications appeared (Borrow & Demey 2004, p. 16, Fry & Keith 2004, p. xiv-xv) and were reinforced regarding the Mali data (Dowsett-Lemaire & Dowsett 2005). Isenmann et al. 2010 did not retain 68 species reported by Lamarche in Mauritania because they were considered “inherently unlikely, unsupported by a description or seen by one observer only”. Of these species, other observers have since confirmed the presence in Mauritania of a few (e.g. Schulz & Walther 2013, Brouwer et al. 2013). Lamarche has not provided me with any point records but I could infer some from his publications. However, because of the large number of species reported by Lamarche in Mauritania that have still not been confirmed by other observers, the above mentioned conclusions for Mali, and the fact that Lamarche has apparently never published confirmatory details of his observations, I decided to exercise caution and to omit all of his reports from this website, including those mentioned in Isenmann et al 2010.

There are some other records included in the compilation of this Atlas that, in my opinion, can be criticized on the basis of the criteria of Isenmann et al 2010 cited above. My resources have not allowed me to give all records submitted to me adequate scrutiny and screening. I believe that a Mauritanian Bird Records Committee should be established to authenticate records, as has been done in many other countries.


Many thanks to the numerous observers named on the Sources page who provided the records on which this Atlas is based. Several other observers provided records for areas close to Mauritania in Senegal, Mali and Western Sahara (Morocco). Their data will influence predicted distributions and their contributions will be acknowledged if and when such distributions are included. Warm thanks also to Alan Morton, creator of the DMAP software (, who unfailingly furnished advice and guidance when I needed it and to Raphaël Bussière for translating into French data entry forms and instructions.

Peter Browne, Ottawa, Canada