If you use this dataset, please cite Eyres et al. (2017); this paper also contains essential information of the rationale behind the dataset, although it used version 1 of the dataset (which only differed in not classifying extinct species and some minor mistakes fixed since). Here we provide a methods overview for the classification of migratory behaviour in birds, which is valid for both versions of the dataset. The dataset focuses on seasonal migration, i.e. movements causing an individual adult bird to be found in different locations over the course of one year, excluding everyday routine movements (e.g. foraging movements) and one-way dispersal movements by juveniles (natal dispersal) or adults (Newton 2008). Types or categories of migratory behaviour and different subcategories were recorded and classified with the Handbook of the Birds of the World (del Hoyo 1992-2013 and updates on the Handbook of the Birds of the World Alive website www.hbw.com, accessed until September 2016).

Main categories and subcategories of migratory behaviour within species

We distinguished these main categories of migratory behaviour (for examples, see Table A1 in Eyres et al. 2017), following Newton (2008):

Initially, we classified intraspecific variation in movement behaviours for each species, so that all different population- and individual-level movements were recorded in a presence-absence matrix of categories and subcategories. The categories are given above; each of the three categories designating movement behaviour (directional migratory, dispersive migratory, and nomadic) was further subdivided into the three different subcategories full, partial, and local. The subcategories were primarily intended to capture the degree of intraspecific variation, as each movement behaviour can occur as follows: - across all individuals and populations of a species (fully migratory or fully nomadic species), - just in some populations or parts of populations (partially migratory or partially nomadic species), or - at a local scale, i.e. only in or over some areas that constitute a small proportion of a species’ distribution (locally migratory or locally nomadic species).

Note that the distance of migratory movements was disregarded in our classification, except that very short-distance movements were classified as subcategory local (see Table S1 for examples). Several of the main categories were applied to a species if categories were each classified as only partially or locally in a given category, to capture the behaviours exhibited by each population (e.g. partially migratory species are often also partially resident). Species classified as fully within one category of migratory behaviour (e.g. full directional migrants) could be also classified as locally within another migratory category (e.g. local dispersive migrant or locally nomadic), but not as partially within another category. If a type of migratory behaviour occurred partially in a species, we did not apply the subcategory “local” in the same main category even if the local behaviour applied to a different population; similarly, if “full” applied for a category, none of the other two subcategories (“partial” and “local”) were applied for the same main category. Note that under our classification, a species described as nomadic within one season (usually the non-breeding season) is not partially nomadic, because the behaviour occurs seasonally; instead, it is classified as partially migratory in the “dispersive migration” category.

Resident behaviour was only distinguished into the subcategories full or partial:

A comment column was used to enter a brief summary of the descriptions of movements from the Handbook of the Birds of the World, which can therefore differ from the framework used here in their use of definitions to describe migratory behaviour (see e.g. the note on partially nomadic vs. dispersive behaviour above; examples for abbreviated comments and descriptions are in Table A1 of Eyres et al. 2017). We used both the family and species accounts originally from the printed book versions (del Hoyo 1992-2013), but updated most with the Handbook of the Birds of the World Alive webpage (www.hbw.com, accessed April 2016 – November 2017). In general, the mapped geographic distributions also supplied in the Handbook corroborated our categorization of species from the species accounts, but sometimes text descriptions did not match information apparent in the map, which was then mentioned in the migratory status comments.

Species-level classification of migratory behaviour

In a second step, we classified each species based on the available intraspecific data on movement behaviour. The final migratory behaviour for species with several “partial” subcategories was decided as consistently as possible by a single person for all species, on the basis of proportional representation of behaviours across different populations (see Table A1 in Eyres et al. 2017 for examples). In addition, there is an “uncertain” column in the dataset that was used to flag species where the decision among the main categories of behaviour was difficult because information was lacking to determine which category was predominant (e.g. they were classified as “partial” for more than one behaviour and from the description showed each behaviour to roughly equal extents; or the movement description indicated that movements or seasonal distributions were poorly or inadequately known).

In the special case of altitudinal migration, species that exclusively perform altitudinal migration within the breeding range were usually classified as species-level residents with “partial” subcategories applying for the “dispersive” and the “resident” category and the “altitudinal” column flagged; if the altitudinal migration caused the species’ distribution to have large distinct seasonal areas mapped, e.g. Himalayan birds migrating down and southwards into the adjacent lowlands, the species was classified as species-level dispersive with “partial” subcategories applying for the “dispersive” and the “resident” category and the “altitudinal” column flagged (see Phylloscopidae example species in Table A1, Eyres et al. 2017).

The “unknown” category of movement behaviour was only applied to very few species where migratory behaviour really was completely unclear, e.g. those only known from a few museum specimen, and the “uncertain” column was used to flag only cases where there was reason to doubt a particular classification as any of the main categories. For some species, the “movement” section in the Handbook only said “no information”, which was generally entered in the comment column. These species were usually mapped as resident in the Handbook itself and often had restricted ranges or were part of higher taxa consisting exclusively of residents; in these cases, species were classified as resident without flagging in the “uncertain” column (as any large-scale movement behaviour is highly unlikely for these species), and the comment “presumed resident” was added with a qualifier in brackets (e.g. “New Guinea” for species endemic to New Guinea, or “HBW map year-round only”).

The database contains the full list of 10596 species as given by the IOC version 3.1 (Gill & Donsker 2012). In version 2, we also classified the movement behaviour of 155 extinct species; these are flagged in a “fully extinct” column, and were mostly species officially declared extinct, plus a few that have not been documented in decades but are listed as “critically endangered (possibly extinct)” (IUCN Red List at www.iucnredlist.com, accessed April-September 2016). Further, we used a “partially extinct” column to flag a further 49 species affected by the extinction of whole populations or across entire geographic entities (typically islands); this was to highlight that migratory behaviour of extinct populations for these species might be unknown or badly known, and might have been different from the extant populations used as the basis of our classification.

We also classified species as fully or partially marine, to highlight seabirds and waterbirds whose geographic distributions are fully or partially mapped at sea. Marine species were defined as spending most of their daily movements in marine environments, typically foraging in any marine habitat from tidal flats to pelagic or open sea, so that geographic distributions are mapped at sea (at least partially). Fully marine species were defined to also breed in marine environments, usually along coasts or on oceanic islands, and very little of their distribution was usually mapped on land. Partially marine species were defined as occurring seasonally in terrestrial or freshwater environments, so these may breed on land close to or far from the coast. There were 252 fully marine species and a further 117 partially marine species (numbers for version 2 of the dataset).

Examples of detailed descriptions and how they were classified

Below, we give some examples of how descriptions of movements in the Handbook were interpreted in our database, and summarize the key parameters for each categorization. In general, if a behaviour was described as occurring “possibly”, “maybe”, “likely”, or similar, we added the description in the comment column but did not use the information for categorization; stronger words like “probably” or “apparently” were often used for categorization, but usually the uncertainty column was then flagged. Descriptions that were too vague (“migration / movements suspected / suggested” or “may show migration / movements”) were recorded in the comments but disregarded for categorization (but note that “some seasonal / dispersive / nomadic / irregular movements” was recorded as locally dispersive or locally nomadic). This rule did not apply to the description of resident behaviour, as in many cases the movement text only said “apparently”, “presumably”, or “probably” resident or sedentary; if there was no further information given, the species was classified as resident.

Directional migration: regular seasonality of movements and clear directionality of seasonal movement.

Dispersive migration: regular seasonality of movements and no clear directionality of seasonal movement.

Nomadism: irregular movements that are not seasonal and no clear directionality of movement.

References

Del Hoyo, J. 1992-2013. Handbook of the Birds of the World. Vol. 1-16 and Special Vol. Lynx Edicions, Barcelona.

Eyres, A., Böhning-Gaese, K. & Fritz, S. A. 2017 Quantification of climatic niches in birds: adding the temporal dimension. Journal of Avian Biology online (doi: 10.1111/jav.01308)

Gill, F. & Donsker, D. 2012. IOC World Bird List (v. 3.1). Available at www.worldbirdnames.org. Accessed September 2012.

Newton, I. 2008. The Migration Ecology of Birds. Academic Press, London.



RS-eco/traitdata documentation built on Oct. 29, 2022, 7:52 p.m.