A synthesis of recent tools and perspectives in migratory connectivity studies

Abstract

Migration movements connect breeding and non-breeding bird populations over the year. Such links, referred to as migratory connectivity, have important implications for migratory population dynamics as they dictate the consequences of localised events for the whole population network. This calls for concerted efforts to understand migration processes for large-scale conservation. Over the last 20 years, the toolbox to investigate connectivity patterns has expanded and studies now consider migratory connectivity over a broader range of species and contexts. Here, we summarise recent developments in analysing migratory connectivity, focusing on strategies and challenges to pooling various types of data to both optimise and broaden the scope of connectivity studies. We find that the different approaches used to investigate migratory connectivity still have complementary strengths and weaknesses, whether in terms of cost, spatial and temporal resolution, or challenges in obtaining large sample sizes or connectivity estimates. Certain recent developments offer particularly promising prospects: robust quantitative models for banding data, improved precision of geolocators and accessibility of telemetry tracking systems, and increasingly precise probabilistic assignments based on genomic markers or large-scale isoscapes. In parallel, studies have proposed various ways to combine the information of different datasets, from simply comparing the connectivity patterns they draw to formally integrating their analyses. Such data combinations have proven to be more accurate in estimating connectivity patterns, particularly for integrated approaches that offer promising flexibility. Given the diversity of available tools, future studies would benefit from a rigorous comparative evaluation of the different methodologies to guide data collection to complete migration atlases: where and when should data be collected during the migratory cycle to best describe connectivity patterns? Which data are most favourable to combine, and under what conditions? Are there methods for combining data that are better than others? Can combination methods be improved by adjusting the contribution of the various data in the models? How can we fully integrate connectivity with demographic and environmental data? Data integration shows strong potential to deepen our understanding of migratory connectivity as a dynamic ecological process, especially if the gaps can be bridged between connectivity, population and environmental models.

Keywords: Data pooling; Integrated modelling; Migration patterns; Movement ecology; Population dynamics; Transition probabilities.

Fig. 1

Number of peer-reviewed publications about migratory connectivity in the Web of Science Core Collection. The number of publications mentioning only one approach (dark purple), or data from at least two different types of approach either analysed separately (light purple) or integrated in a single framework (medium purple) are shown by year. Publications were selected based on the occurrence of selected word patterns in the title, abstract or keywords (search conducted on 02/11/2023; see Additional file 1). 7% of publications could not be classified using our methodology (grey). The absolute numbers are expected to be underestimated, but the relative proportions of analysis types are assumed to be representative. Data from Web of Science, provided by Clarivate. Web of Science and Clarivate are trademarks of their respective owners and used herein with permission

Fig. 2

Conceptual framework of migratory connectivity studies. Large arrows show which data types are used in which analysis, and what information can be obtained from these. Various types of data provide more or less direct information on the movement of individuals between sites or populations (1.a), which can be used either alone (2.a) or in combination (2.b) to investigate migratory connectivity at different scales and resolutions (3.a, b). Complementary data such as species abundance or occurrence (1.b) is increasingly included in combination to improve the precision and/or accuracy of analyses. Recent studies build on the output of migratory connectivity studies (3.a, b) to investigate the evolutionary causes and consequences of this connectivity in relation to additional environmental variables or population dynamics (3.c). The possibility of integrating such additional data at the same level as other complementary data (1.b) needs further investigation in order to consider migratory connectivity in its functional and dynamic aspects rather than as a static description