Assessing bird migrations

Abstract

Birds perform many types of migratory movements that vary remarkably both geographically and between taxa. Nevertheless, nomenclature and definitions of avian migrations are often not used consistently in the published literature, and the amount of information available varies widely between taxa. Although comprehensive global lists of migrants exist, these data oversimplify the breadth of types of avian movements, as species are classified into just a few broad classes of movements. A key knowledge gap exists in the literature concerning irregular, small-magnitude migrations, such as irruptive and nomadic, which have been little-studied compared with regular, long-distance, to-andfro migrations. The inconsistency in the literature, oversimplification of migration categories in lists of migrants, and underestimation of the scope of avian migration types may hamper the use of available information on avian migrations in conservation decisions, extinction risk assessments and scientific research. In order to make sound conservation decisions, understanding species migratory movements is key, because migrants demand coordinated management strategies where protection must be achieved over a network of sites. In extinction risk assessments, the threatened status of migrants and non-migrants is assessed differently in the International Union for Conservation of Nature Red List, and the threatened status of migrants could be underestimated if information regarding their movements is inadequate. In scientific research, statistical techniques used to summarise relationships between species traits and other variables are data sensitive, and thus require accurate and precise data on species migratory movements to produce more reliable results. It is in this context that this thesis aims to provide more clarity regarding types of avian movements and how to systematically gather and produce data on bird migrations. In Chapter 1, I explain in more detail the importance of avian migrations, the gaps in the existing literature, and the need to understand avian migrations and account for these movements in a range of applications. In Chapter 2, I take a holistic approach to review and reinterpret terminologies and definitions on avian migrations from a vast literature, to build a new movement typology framework. Using this typology, movements of avian taxa can be classified into six primary classes, such as to-and-fro, irruptive and nomadic migrations, and 16 secondary classes that are qualifiers of the primary classes, like dispersive, altitudinal and differential. I ii exemplify the use of the typology globally by assessing the movements of 13 bird species from different taxa and regions, and regionally by assessing the movements of 33 taxa of cuckoos in Australia. In Chapter 3, I use the typology from Chapter 2, together with a suite of geoprocessing techniques to build the most comprehensive and detailed dataset on migratory movements of global bird species (n=10,596) to date. The dataset includes 56 variables on avian movement types and magnitude, their mobility modes, location of occurrence and range. I also present the results of validations to test the quality of the data. In Chapter 4, I explore how presence data, which are already abundant and still rapidly growing for birds, can help us understand migration. I develop the seasonality index, an indicator of spatial and temporal variation in presence data, which could be reflecting species migratory movements. I calculate this index for 2368 species of the Western Hemisphere from eBird, the single largest citizen science program for birds. As expected, the presence of seasonality was high within to-and-fro and irruptive migrants (91 of 92 species) and random within nomadic and undefined migrants, partial migrants, and species with insufficient data (49% of the species). However, contrary to what was expected, a significant number of resident species presented seasonality (57% of the total number of residents). I highlight the need to further investigate the possibility that cryptic migrations are actually present within some of these ‘resident’ species. In Chapter 5, I explore bias in effort applied to collecting avian presence data globally using a dataset of more than 500 million records from over 11,000 species (including migrants and non-migrants). I obtain this dataset from the Global Biodiversity Information Facility, the largest single repository of avian presence data, which includes data from eBird and many other programs. The results showed that effort is correlated with 21 socioeconomic and biophysical variables, and I discuss ways to minimise spatial bias and ways to account for bias in existing data through statistical approaches. In Chapter 6, I provide a general discussion of the research undertaken in chapters 2 to 5 to explain their linkages, limitations, ideas for future work and pathway to impact.